Advances in Protein Chemistry, Volume 59: Protein Folding in the Cell (Advances in Protein Chemistry)

ADVANCES IN PROTEIN CHEMISTRY Volume 59 Protein Folding in the Cell This . Page Intentionally Left Blank ADVANCES I...

Author: Arthur Horwich

42 downloads 1397 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

ADVANCES IN PROTEIN CHEMISTRY Volume 59 Protein Folding in the Cell

This . Page Intentionally Left Blank

ADVANCES IN PROTEIN CHEMISTRY EDITED BY FREDERIC M. RICHARDS

DAVID S. EISENBERG

Department of Molecular Biophysics and Biochemistry Yale University New Haven, Connecticut

Department of Chemistry and Biochemistry University of California, Los Angeles Los Angeles, California

John Kuriyan Department of Molecular Biophysics Howard Hughes Medical Institute Rockefeller University 1230 York Avenue New York, NY 10021

VOLUME 59

Protein Folding in the Cell EDITED BY ARTHUR HORWICH Howard Hughes Medical Institute Yale University School of Medicine New Haven, Connecticut

San Diego London Boston New York Sydney Tokyo Toronto

∞ This book is printed on acid-free paper.

C 2002 by ACADEMIC PRESS Copyright

All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the ﬁrst page of a chapter in this book indicates the Publisher’s consent that copies of the chapter may be made for personal or internal use of speciﬁc clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. 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. Copy fees for pre-2000 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0065-3233/02 $35.00 Explicit permission from Academic Press is not required to reproduce a maximum of two ﬁgures or tables from an Academic Press chapter in another scientiﬁc or research publication provided that the material has not been credited to another source and that full credit to the Academic Press chapter is given.

Academic Press An Elsevier Science Imprint 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.academicpress.com

Academic Press Harcourt Place, 32 Jamestown Road, London NW1 7BY, UK http://www.academicpress.com Library of Congress Catalog Card Number: 0065-3233 International Standard Book Number: 0-12-034259-6 PRINTED IN THE UNITED STATES OF AMERICA 01 02 03 04 05 06 SB 9 8 7 6 5

4

3

2

1

CONTENTS

PREFACE

.

.

.

.

.

.

.

.

.

.

.

.

xi

Hsp70 Chaperone Machines MATTHIAS P. MAYER , DIRK BREHMER, CLAUDIA S. GA¨ SSLER, AND BERND BUKAU I. II. III. IV. V.

Introduction . . . . . Chaperone Activities of Hsp70 . . Mechanism of Action . . . . The Targeting Activity of Co-chaperones Outlook . . . . . . References . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

1 2 8 28 36 37

. . . . .

45 46 56 70 71

Allostery and Protein Substrate Conformational Change during GroEL/GroES-Mediated Protein Folding HELEN R. SAIBIL, ARTHUR L. HORWICH, AND WAYNE A. FENTON I. II. III. IV.

Introduction . . . . . . . Structure of GroEL and Its Functional Complexes Polypeptide Folding . . . . . . Conclusions . . . . . . . References . . . . . . . . v

. . . . .

vi

CONTENTS

Type II Chaperonins, Prefoldin, and the Tubulin-Specific Chaperones NICHOLAS J. COWAN AND SALLY A. LEWIS I. II. III. IV. V. VI. VII. VIII. IX. X.

Introduction . . . . . . Discovery of Type II Chaperonins and Early Functional Studies . . . . . Subunits and Assembly . . . . Target Range and Specificity of CCT . . Cycling of Target Proteins by CCT . . Genetics . . . . . . . Structure of Type II Chaperonins . . Prefoldin . . . . . . . Tubulin-Specific Chaperones . . . Conclusion . . . . . . . References . . . . . . .

.

.

.

73

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

74 75 79 81 83 84 88 93 98 98

Structure and Function of the Small Heat Shock Protein/α-Crystallin Family of Molecular Chaperones ROB VAN MONTFORT, CHRISTINE SLINGSBY, AND ELIZABETH VIERLING I. II. III. IV. V. VI. VII.

Introduction . . . . . . . . Diversity of the sHsps and Their Expression Patterns X-Ray Structural Analysis . . . . . . Dynamic Nature of the sHsp Oligomer . . . sHsp Chaperone Activity . . . . . . Potential sHsp Substrates . . . . . . Conclusions . . . . . . . . References . . . . . . . . .

. . . . . . . .

105 107 116 124 127 138 146 147

. .

157 158

Structure, Function, and Mechanism of the Hsp90 Molecular Chaperone LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU I. II.

Introduction . . . . Domain Structure and Function

. .

. .

. .

. .

vii

CONTENTS

III. IV. V. VI. VII. VIII. IX.

ATP Binding and Hydrolysis by Hsp90 Are Essential in Vivo . . . . . . Conformational Changes in Hsp90 Accompanying the ATPase Cycle . . . . . . Hsp90 ATPase Inhibitors—A New Class of Antitumor Drugs . . . . . . Interaction with Co-chaperones . . . . Regulation of ATP Binding and Hydrolysis in the Client-Protein Activation Pathway . . . Interactions with Alterations of Client Proteins by Hsp90 . . . . . . . . Conclusion . . . . . . . . References . . . . . . . .

.

.

164

.

.

166

. .

. .

171 172

.

.

175

. . .

. . .

176 180 180

The Proteasome: A Supramolecular Assembly Designed for Controlled Proteolysis ¨ , BARBARA KAPELARI, PETER ZWICKL, ERIKA SEEMULLER AND WOLFGANG BAUMEISTER I. II. III. IV.

Introduction . . . . The 20S Proteasome . . . Activators of the 20S Proteasome Conclusions . . . . References . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

187 188 202 213 213

. . . . . . . .

223 224 227 228 229 237 238 239

Hsp70 Proteins in Protein Translocation MICHAEL T. RYAN AND NIKOLAUS PFANNER I. II. III. IV. V. VI. VII.

Introduction . . . . . . . . Protein Translocation into Mitochondria and ER . Cytosolic Hsp70s Are Involved in Protein Translocation Hsp70 and Its Cofactors . . . . . . Lumenal Hsp70s and Protein Translocation . . Other Roles for Hsp70s in Protein Translocation . Conclusion . . . . . . . . . References . . . . . . . . .

viii

CONTENTS

Prolyl Isomerases FRANZ X. SCHMID I. II. III. IV. V. VI. VII. VIII. IX. X. XI. XII.

Perspective . . . . . . . . . Properties of Prolyl Peptide Bonds . . . . Prolyl Isomerizations in Protein Folding . . . Examples . . . . . . . . . Cis/trans Isomerizations at Nonprolyl Peptide Bonds Prolyl Isomerizations in Folded Proteins . . . Prolyl Isomerases . . . . . . . Prolyl Isomerases as Catalysts of in Vitro Protein Folding . . . . . . . . The Trigger Factor . . . . . . . Catalysis of Prolyl Isomerization during de Novo Protein Folding . . . . . . . . Cellular Functions of Prolyl Isomerases . . . Concluding Remarks . . . . . . . References . . . . . . . . .

. . . . . . .

244 244 246 250 253 255 256

. .

261 264

. . . .

267 268 273 274

Catalysis of Disulfide Bond Formation and Isomerization in Escherichia coli MARTIN W. BADER AND JAMES C. A. BARDWELL I. II. III. IV. V. VI. VII. VIII. IX.

Introduction . . . . . . . . De Novo Formation of Disulfide Bonds in E. coli: The Discovery of DsbA . . . . . . DsbA Is the Most Oxidizing Disulfide Catalyst . . DsbB Provides the Periplasm with Oxidizing Power . Correcting Wrong Disulfide Bonds in the Periplasm: Disulﬁde Bond Isomerization by DsbC . . . DsbD Provides Reducing Equivalents in a Highly Oxidizing Environment . . . . . . Dsb Proteins and Cytochrome c Maturation . . Disulfide Bond Formation Does Not Interfere with Disulfide Isomerization . . . . . . Concluding Remarks . . . . . . . References . . . . . . . . .

.

283

. . .

284 286 290

.

292

. .

296 297

. . .

298 298 299

ix

CONTENTS

N-Glycan Processing and Glycoprotein Folding E. SERGIO TROMBETTA AND ARMANDO J. PARODI I. II. III. IV. V. VI. VII. VIII.

Introduction . . . . . . . . N -Glycan Processing in the Endoplasmic Reticulum . Glycoprotein Reglucosylation . . . . . Chaperones and Protein Folding in the Endoplasmic Reticulum . . . . . . . . . Interaction of Glycoproteins with Calnexin and Calreticulin . . . . . . . . Calnexin and Calreticulin Are Lectins Specific for Monoglucosylated Oligosaccharides . . . N-Glycan Processing and Glycoprotein Degradation . Summary and Future Perspectives . . . . References . . . . . . . . .

. . .

303 304 308

.

320

.

321

. . . .

323 330 332 333

Functional Genomic Approaches to Understanding Molecular Chaperones and Stress Responses KEVIN J. TRAVERS, CHRISTOPHER K. PATIL, AND JONATHAN S. WEISSMAN I. II. III. IV. V. VI.

Introduction . . . . . . . . Historical Perspective . . . . . . . Functional and Genomic Analysis of the Unfolded Protein Response . . . . . . . Strategy for Identifying UPR Targets . . . . An Overview of Unsupervised Search Strategies . UPR as a Case Study in the Comparison of Supervised versus Unsupervised Searches . . . . . References . . . . . . . . .

. .

345 346

. . .

353 363 370

. .

375 379

. .

391 395

The Yeast Prion [PSI + ]: Molecular Insights and Functional Consequences TRICIA R. SERIO AND SUSAN L. LINDQUIST I. II.

Overview . . Reversible Curing

. .

. .

. .

. .

. .

. .

. .

x

CONTENTS

III. IV. V. VI.

Separating Prion Initiation and Propagation Conformational Replication in Vitro . . Functional Consequences of the [PSI + ] State Conclusion . . . . . . . References . . . . . . .

. . . . .

. . . . .

. . . . .

400 402 406 409 409

Clp ATPases and Their Role in Protein Unfolding and Degradation JOEL R. HOSKINS, SUVEENA SHARMA, B. K. SATHYANARAYANA, AND SUE WICKNER I. II. III. IV. V. VI. VII.

Introduction . . . . . . . . . Clp ATPase Family of Proteins . . . . . . Chaperone Activity of Clp ATPases and Their Participation in Proteolysis . . . . . . . Structure of Clp ATPases: Alone and with Partner Proteases . . . . . . . . Mechanism of Action of Clp ATPases as Chaperones and as Components of Degradation Machinery . . Clp ATPase Specificity Factors . . . . . . Summary . . . . . . . . . . References . . . . . . . . . .

AUTHOR INDEX SUBJECT INDEX

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

413 414 415 416 419 424 425 426 431 481

PREFACE

Of the myriad structural and catalytic roles of proteins in the cell, that of assisting others to fold is surely one of the most remarkable and is the subject of this volume of Advances in Protein Chemistry. The early studies of Anﬁnsen and others identiﬁed a role for protein disulﬁde isomerase in catalyzing folding of RNase A, but the notion of broadly provided assistance to protein folding through the ofﬁces of noncovalent binding and release of nonnative conformations was not recognized until the studies on Hsp70 proteins by Pelham and co-workers in the mid-1980s, which suggested an action of ATP-dependent disaggregation/dissociation of protein multimers, for example, under heat-shock conditions. Subsequent studies in the late 1980s resolved this action to the level of binding of monomeric polypeptide chains, for example, entering and passing through biological membranes. Additional observations that Hsp60/GroEL ring assemblies could mediate ATP-driven folding to the native state produced an even broader view of what these “chaperone” components could accomplish, and what we have seen over the past ten years is an exciting progression of cell biologic and mechanistic understanding of both these two chaperone families and additional families. This volume, the brainchild of two major ﬁgures in protein chemistry, David Eisenberg and Peter Kim, covers our current understanding of these cellular components/machines, with an emphasis on their mechanisms of action. In addition, there is a chapter on protein misfolding in prion disease in a yeast system, with some attention given to the inﬂuence of molecular chaperones on that process, and one on a system analysis of response to ER stress. With many important structural and mechanistic discoveries lying just behind us, it was time for a volume dealing with these components. xi

xii

PREFACE

Thus, a reader of this volume with little or no basic knowledge in this area will get a comprehensible quantum of information. It is a labor of love produced by a group of leaders in this ﬁeld who were generous enough with their time to sit down and take stock of their respective areas of study. In all cases, the treatments are lucid, penetrating to the major questions about the respective components. The order of presentation is somewhat arbitrary, but the initial chapters seek to cover the major molecular chaperone systems. Hsp70 and Hsp60/GroEL chaperonin families are presented ﬁrst by Mayer, Brehmer, G¨assler, and Bukau and by Saibil, Horwich, and Fenton, respectively, reviewing both structural and mechanistic aspects, including the actions of ATP and co-chaperones at these machines. A chapter on the eukaryotic cytosolic chaperonin, CCT, by Cowan and Lewis follows, reviewing its essential role in actin and tubulin folding as well as the roles of such co-chaperone components as prefoldin and the tubulin cofactors. Van Montfort, Slingsby, and Vierling then present structural and mechanistic information on small heat shock proteins. Then Pearl and Prodromou review the contemporary understanding of the Hsp90 chaperone system involved crucially in biogenesis of signal transducing receptors and kinases in metazoans. Somewhat dislocated, Hoskins, Sharma, Sathyanarayana, and Wickner deal in a ﬁnal chapter with prokaryotic members of the Hsp100 family, ClpA and ClpX ring hexamers, and their role in ATP-dependent protein unfolding/degradation. Just proximal to that, Serio and Lindquist deal with yeast prion formation and the role of eukaryotic Hsp104 in that process. Following the Pearl and Prodromou chapter, Zwickl, Seem¨uller, Kapelari, and Baumeister present a review of eukaryotic proteasome action, where the 19S cap structure may have unfolding actions akin to those of Hsp100 proteins. Ryan and Pfanner examine the action of Hsp70 proteins in mediating translocation of polypeptides into both ER and mitochondrial systems, focusing informatively on models of ratcheting and pulling mechanics. Schmid then presents a stunning review of the biology of proline isomerization and enzymes that catalyze that process, many of which appear to exhibit chaperone-like polypeptide binding activities. The prolyl isomerases, increasingly, appear to play a role not only in de novo protein folding but also in signal transduction mechanisms. Bader and Bardwell review the remarkable Dsb pathway of disulﬁde bond formation in the bacterial periplasm. Trombetta and Parodi then cover the pathway of glycoprotein folding in the ER that employs the calnexin/calreticulin system and the UDP-glucose:glycoprotein glucosyltransferase chaperone that Parodi originally discovered. Travers, Patil, and Weissman then deal with the ER

PREFACE

xiii

quality control system known as the unfolded protein response pathway (UPR), reviewing the approaches and results from microarray screens. This volume thus provides both a broad and deep look at cellular components that assist protein folding in the cell. The availability of outlines at the beginning of each chapter should allow the reader to travel quickly to subjects of greatest interest, but the writing styles of this collective of authors make a front-to-back reading of any of the chapters thoroughly enjoyable. While this ﬁeld is just at a waypoint in its progress, a reader overviewing this collective of developments cannot help but be stunned to know that these components were barely recognized ten years ago! Happy reading. Arthur Horwich

This . Page Intentionally Left Blank

Hsp70 CHAPERONE MACHINES ¨ By MATTHIAS P. MAYER, DIRK BREHMER, CLAUDIA S. GASSLER, and BERND BUKAU Institute for Biochemistry and Molecular Biology, University of Freiburg, Hermann-Herder-Str. 7, 79104 Freiburg, Germany

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Chaperone Activities of Hsp 70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Substrate Speciﬁcity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Protein Folding Processes Assisted by Hsp70 . . . . . . . . . . . . . . . . . . . . . . . . . III. Mechanism of Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. General Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The ATPase Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Substrate Binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. The Targeting Activity of Co-chaperones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. DnaJ Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Bag Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Hip and Hop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Chip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. HspBP1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 2 3 8 8 10 21 28 28 31 34 35 36 36 37

I. INTRODUCTION The ubiquitous 70-kDa heat shock proteins (Hsp70s) assist an extraordinarily broad spectrum of folding processes within the entire life span of proteins. These processes include the folding of newly synthesized and misfolded proteins, the prevention and reversion of protein aggregation, the translocation of organellar and secretory proteins across membranes, the assembly and disassembly of oligomeric structures, and the control of the biological activity of regulatory proteins. Hsp70s have housekeeping functions in the cell in which Hsp70s are built-in components of folding and signal transduction pathways, and quality control functions in which Hsp70s proofread the folding status of proteins and repair misfolded conformers. All these activities appear to be driven by the basic property of Hsp70s to interact with short hydrophobic peptide segments of protein substrates in an ATP-dependent fashion. The broad spectrum of cellular functions of Hsp70 proteins is achieved through three strategies. First, the ampliﬁcation and diversiﬁcation of hsp70 genes in evolution 1 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0096-5332/02 $35.00

2

MATTHIAS P. MAYER ET AL.

has generated functionally specialized Hsp70 chaperones. Second, a large set of co-chaperones are selectively recruited by Hsp70 chaperones to fulﬁll speciﬁc cellular functions. Third, Hsp70 homologs cooperate with other chaperone systems to broaden their activity spectrum. Hsp70 proteins with their co-chaperones and cooperating chaperones thus constitute a complex network of folding machines. This chapter describes the molecular basis of this network. Particular emphasis is given to the DnaK system of Escherichia coli as it is the best understood Hsp70 system, and to the mechanistic differences between Hsp70 family members. II. CHAPERONE ACTIVITIES OF Hsp 70 A. Substrate Specificity The ensemble of Hsp70 proteins binds promiscuously to non-native polypeptides, such as the entire spectrum of heat-denatured proteins of E. coli cell extracts (Mogk et al., 1999), and selectively to folded proteins such as clathrin, transcription factors (including σ 32, HSF, steroid hormone receptors), kinases (including Raf, eIF2α-kinase), and DNA replication proteins (including λP, RepE, RepA). These substrates are unrelated in sequence and structure and represent a large spectrum of folding conformers ranging from completely unfolded nascent chains emerging at ribosomes and translocation pores to native regulatory proteins. On closer inspection, only the Hsp70 proteins with general chaperone functions have such a broad substrate spectrum in their cellular environment, including E. coli DnaK, yeast Ssa1 to Ssa4 and Kar2 (BiP), as well as mammalian Hsc70 and Hsp70. Other Hsp70 homologs are highly specialized, such as E. coli HscA (Hsc66), yeast mitochondrial Ssq1, and yeast cytosolic Ssb1 and Ssb2, which interact with proteins related to iron–sulfur cluster–containing proteins (Hoff et al., 2000; Lutz et al., 2001; Voisine et al., 2001) and nascent polypeptide chains or ribosomal components (Pfund et al., 1998), respectively. The differences between general and specialized Hsp70s are poorly understood, but may include the selective recruitment of DnaJ co-chaperones and subtle differences in the substrate-binding cavity. Their broad substrate speciﬁcity implies a rather degenerative binding motif for at least the Hsp70 homologs with general chaperone functions. Peptide substrates of the DnaK and BiP homologs are enriched in hydrophobic residues (Flynn et al., 1991; Blond-Elguindi et al., 1993; Gragerov et al., 1994) and, as shown by NMR for DnaK, are bound

Hsp70 CHAPERONE MACHINES

3

to the chaperone in an extended conformation (Landry et al., 1992). Peptide library screening revealed for DnaK that typical binding sites in substrates consist of a core of ﬁve consecutive residues enriched in hydrophobic amino acids and ﬂanking regions enriched in basic amino acids (Rudiger ¨ et al., 1997a; Rudiger ¨ et al., 1997b). Binding sites for eukaryotic Hsp70 homologs are similar, although some differences exist (Rudiger ¨ and Bukau, unpublished results). Such binding sequences occur frequently within protein sequences, e.g., approximately one DnaK binding site for every 36 residues, and are typically buried within the hydrophobic core of the folded proteins (R¨udiger et al., 1997b). These features imply that substrates, whether non-native or native, must expose at least one binding site at their surface to be recognized by Hsp70s. It is conceivable that the folding state of the substrate is a major determinant for the number of Hsp70s bound per substrate molecule. In the case of denatured ﬁreﬂy luciferase only one or two (at most) Hsp70 molecules associate with a substrate molecule at a given time (Schr¨oder and Bukau, unpublished results). In the case of proteins that are translocating into the endoplasmic reticulum, an in vitro analysis using a membrane-free system revealed that as many as six to seven molecules of the luminal BiP can associate with a single nascent polypeptide chain of prepro-α-factor (165 residues) (Matlack et al., 1999). It is unclear how such massive association of Hsp70 proteins affects the folding process of the substrate in the endoplasmic reticulum. B. Protein Folding Processes Assisted by Hsp70 Most information on the mechanism by which Hsp70s and their cochaperones assist protein folding comes from analyses of the in vitro folding of denatured model proteins such as ﬁreﬂy luciferase. In such analyses the denatured proteins are transferred to conditions that are permissive for chaperone-dependent folding, e.g., after dilution of the denaturant. The role of Hsp70s in the folding of denatured proteins can be divided into three related activities: the prevention of aggregation, the promotion of folding to the native state, and the solubilization and refolding of aggregated proteins. In the cellular milieu, Hsp70s exert these activities in the quality control of misfolded proteins (Nguyen et al., 1989; Morimoto et al., 1994; Hartl, 1996; Glover and Lindquist, 1998; Goloubinoff et al., 1999; Mogk et al., 1999) and the co- and posttranslational folding of newly synthesized proteins (Nelson et al., 1992; Hendrick et al., 1993; Frydman et al., 1994; Hansen et al., 1994; Pfund et al., 1998; Deuerling et al., 1999; Teter et al., 1999). Probably mechanistically related but less understood is the role of Hsp70s in the disassembly of protein complexes such as clathrin coats (Ungewickell, 1985; Greene

4

MATTHIAS P. MAYER ET AL.

and Eisenberg, 1990), RepA dimers (Wickner et al., 1991a; Wickner et al., 1991b; Wickner et al., 1992), and the nucleoprotein complex that initiates the replication of bacteriophage λ DNA (Liberek et al., 1988; Alfano and McMacken, 1989). A more complex folding situation exists for the Hsp70-dependent control of regulatory proteins since several steps in the folding and activation process of these substrates are assisted by multiple chaperones (Pratt and Toft, 1997). The role of Hsp70s in protein translocation is discussed in detail in another chapter of this volume. 1. Prevention of Aggregation Hsp70 proteins in cooperation with DnaJ co-chaperones prevent the aggregation of folding intermediates through the association with exposed hydrophobic patches of the substrate molecules which shields them from intermolecular interactions (“holder” activity). Optimal aggregation prevention usually requires a stoichiometric excess of Hsp70 with respect to the substrate, substoichiometric concentrations of a DnaJ co-chaperone with respect to Hsp70, and ATP (Gragerov et al., 1992; Schr¨oder et al., 1993; Ziemienowicz et al., 1993; Szabo et al., 1994; Ziemienowicz et al., 1995; Freeman and Morimoto, 1996; Michels et al., 1997; Kim et al., 1998). Some DnaJ homologs such as DnaJ of E. coli can prevent aggregation by themselves through ATP-independent transient and rapid association with the substrates (see below) (Schr¨oder et al., 1993; Szabo et al., 1996). Hsp70s share this ability to promiscuously prevent protein aggregation with most other chaperone systems including small HSPs, GroEL, and Hsp90. However, only members of the Hsp70 family with general chaperone functions have such general holder activity. 2. Folding to Native State Hsp70s with DnaJ co-chaperones and, in the case of DnaK homologs, GrpE co-chaperones assist some, but not all, non-native intermediates to fold to the native state (“folder” activity). These folding intermediates may have been prevented from aggregation by association with Hsp70 or other chaperones. Several reports have shown that substrates bound to sHSPs, GroEL, Hsp90, SecB, and ClpA can be transferred to Hsp70 systems for refolding, thereby forming a network of cooperating chaperones (Buchberger et al., 1996; Freeman and Morimoto, 1996; Ehrnsperger et al., 1997; Veinger et al., 1998; Knoblauch et al., 1999). Such substrate transfer probably involves soluble intermediates which kinetically partition between the different chaperone systems (Buchberger et al., 1996). It is not known how frequently substrates become transferred from various other chaperones onto Hsp70 in vivo and how important such transfers are for the folding process.

Hsp70 CHAPERONE MACHINES

5

The mechanism by which Hsp70 chaperones assist the folding of nonnative substrates is still unclear. Hsp70-dependent protein folding in vitro occurs typically on the time scale of minutes or longer. Substrates cycle in an ATP-dependent fashion between chaperone-bound and -free states until the ensemble of molecules has reached the native state (Szabo et al., 1994; Buchberger et al., 1996). For ﬁreﬂy luciferase it has been shown that the molecules dissociated from DnaK are in non-native conformations which are competent for binding to GroEL (Buchberger et al., 1996). These ﬁndings are consistent with at least two alternative mechanisms. In the ﬁrst mechanism Hsp70s play a rather passive role. Through repetitive substrate binding and release cycles they keep the free concentration of the substrate sufﬁciently low to prevent aggregation while allowing free molecules to fold to the native state. In the second mechanism, the binding and release cycles induce local conformational changes in the substrate, e.g., untangling a misfolded β-sheet, thereby overcoming kinetic barriers for folding to the native state. The energy of ATP may be used to induce such conformational changes. Although lacking in direct experimental support, the second mechanism would provide a satisfying explanation for the observed ability of Hsp70s to disaggregate small aggregates on their own (see below). Furthermore, it would explain why among all the different chaperones that prevent folding intermediates from aggregation, only very few, in particular Hsp70s and chaperonins such as GroEL and CCT/TRiC, can assist their folding to the native state. It is interesting that DnaK and GroEL show differences in their substrate speciﬁcity. They cannot replace one another in assisting the folding of some substrates, including denatured β-galactosidase folded by the DnaK system (Ayling and Baneyx, 1996), ﬁreﬂy luciferase folded by the Hsp70 and CCT/TriC systems but not by the GroEL system (Frydman et al., 1992; Buchberger et al., 1996), rhodanese folded by the GroEL system (Langer et al., 1992), and actin and tubulin folded by the CCT/TRiC system (Ursic and Culbertson, 1991; Frydman et al., 1992; Yaffe et al., 1992). These differences in the folding potential reﬂect the different modes of action of chaperonins and Hsp70 chaperones, i.e., the demonstrated ability of chaperonins to unfold substrates globally (Shtilerman et al., 1999) and the postulated ability of Hsp70s to unfold substrates locally. 3. Protein Disaggregation A subset of Hsp70 homologs solubilizes and refolds aggregated proteins. The ﬁrst report for such activity was for a bi-chaperone system of Saccharomyces cerevisiae, consisting of the Hsp70 system, Ssa1 and Ydj1, and the Hsp104 member of the Hsp100 chaperone family (Glover and

6

MATTHIAS P. MAYER ET AL.

Lindquist, 1998). Both chaperone systems cooperated to solubilize heataggregated proteins in vivo and heat-aggregated ﬁreﬂy luciferase and β-galactosidase in vitro (Glover and Lindquist, 1998). A highly efﬁcient disaggregating activity with broad substrate speciﬁcity was also reported for the E. coli DnaK system and the Hsp104 homolog ClpB (Goloubinoff et al., 1999; Mogk et al., 1999). Disaggregating activities were also described for the DnaK system and ClpB of Thermus thermophilus (Schlee et al., 2001) and the mitochondrial Hsp70 system and Hsp78 of S. cerevisiae (Krzewska et al., 2001). The functional cooperation between the Hsp70 and Hsp104 proteins may require physical interaction between both chaperones, as has been suggested for the S. cerevisiae system (Glover and Lindquist, 1998). But it is not excluded that the two chaperones act separately and perhaps sequentially in the disaggregation reaction. It is an intriguing ﬁnding that in case of the E. coli DnaK/ClpB bi-chaperone system, the requirement for ClpB is alleviated below a threshold size of the aggregates. Small aggregates (2- to 3-mers) of glucose-6-phosphate dehydrogenase were efﬁciently solubilized and subsequently refolded by the DnaK system alone. Larger aggregates (up to 10-mers) could be solubilized only partially by the DnaK system, which then was necessary in a high molar excess. In the presence of ClpB these larger aggregates were disaggregated more efﬁciently even at stoichiometric concentrations of DnaK (Diamant et al., 2000). This result indicates that the DnaK system has limited capacity to solubilize small protein aggregates. It is possible that the postulated ability of DnaK to induce local conformational changes in substrates is involved in this disaggregation activity. Furthermore, these ﬁndings indicate that in a disaggregation and refolding reaction, a cooperating mode of action, perhaps through physical interaction between Hsp70 and Hsp104, and separate modes of action of each chaperone alone go hand in hand. So far we lack all mechanistic insights into this fascinating process. 4. Control of Regulatory Proteins An increasing number of regulatory proteins of eukaryotes are known to be controlled in their biological activity through the transient association with Hsp70. These proteins include nuclear receptors (steroid hormone receptors), kinases (Raf, eIF2α-kinase, CyclinB1/Cdk1) and transcription factors (HSF, c-Myc, pRb). Several features emerge that are common to many of these chaperone–substrate complexes. First, formation of active complexes depends on the interaction with Hsp70 already during the de novo synthesis of the substrate. Second, the association of the chaperone is sensitive to substrate-speciﬁc ligands (e.g., steroid,

Hsp70 CHAPERONE MACHINES

7

heme) or substrate phosphorylation. Third, the complexes involve not only the Hsp70 system but also the Hsp90 system, co-chaperones such as Cdc37 and p23, and additional proteins including immunophilins (Buchner, 1996; Hunter and Poon, 1997; Pratt and Toft, 1997; Uma et al., 1997; Toft, 1999; Uma et al., 1999). For some regulatory proteins, e.g., c-Src-kinase, Cdc28-kinase and NO-synthase, only the interaction with Hsp90 and/or Cdc37 has been demonstrated (Xu and Lindquist, 1993; Garc´ıa-Gardena et al., 1998; Bender et al., 1999; Xu et al., 1999; Ferrell et al., 2000; O’Keeffe et al., 2000), but an interaction with the Hsp70 system is not excluded. Formation of a chaperone complex containing Hsp70, Hsp90, and accessory proteins may be the general mechanism to control the activity of regulatory proteins; in addition, this complex may constitute a general protein folding machine in the cytoplasm (Pratt et al., 1999; Hartson et al., 2000). It was proposed that Hsp90, Hsp70, Hsp40, p60/Hop, and p23 exist in a complex, the “foldosome,” even in the absence of added substrates. Such a complex was found in reticulocyte lysate (Sanchez et al., 1990; Perdew and Whitelaw, 1991; Smith et al., 1993; Hutchison et al., 1994b; Dittmar et al., 1998). However it is not clear whether this complex is necessarily preassembled as there may be several substrates present in the reticulocyte lysate. For the steroid homone receptors it has been shown that the interaction with Hsp70 and a DnaJ protein is essential for the complex formation with Hsp90. In the absence of Hsp70 and Hsp90 or at low concentrations thereof, the substrate does not attain its ligand-binding–competent conformation (Picard et al., 1990; Hutchison et al., 1994a; Pratt and Toft, 1997; Smith, 1998). Most in vitro data suggest that Hsp70 leaves the complex after the initial assembly process and that Hsp90 together with p23 and immunophilins holds the receptor in a conformation competent for steroid binding until the arrival of the hormone (Dittmar and Pratt, 1997; Frydman and H¨ohfeld, 1997; Buchner, 1999). It has been suggested that Hsp70 serves at least two purposes: to facilitate the assembly of the Hsp90 complex and to stepwise open the hormone binding pocket to make it competent for ligand binding (Morishima et al., 2000b; Morishima et al., 2001). There is also evidence that after the assembly process a DnaJ protein is necessary to repress the basal transcriptional activity of the receptor ( Johnson and Craig, 2000), then after hormone binding, a second co-chaperone of Hsp70, the anti-apoptotic protein Bag-1M (Rap46/Hap46), is able to downregulate the transcriptional activity of the receptor and thereby antagonize Hsp70 (Schneikert et al., 1999; Schneikert et al., 2000). All these data suggest that the steroid hormone receptor complex must be highly dynamic, with a number of factors including Hsp70 dissociating and reassociating continuously. There

8

MATTHIAS P. MAYER ET AL.

is indeed evidence that the complex decays with a relatively short half-life of 5 min and reassembles continuously (Smith, 1993; Smith, 1998). A similar situation seems to exist for the eIF2α-kinase (heme-regulated inhibitor HRI). The de novo folding of this kinase depends on the cotranslational action of Hsp70 and Hsp90 (Uma et al., 1997; Uma et al., 1999). After the kinase has reached the mature state, it remains associated with Hsp70 and Hsp90 in a repressed state until it is activated by the dissociation of heme or by phosphorylation (Uma et al., 1997; Donze and Picard, 1999; Uma et al., 1999). It remains to be established whether the scheme drawn for nuclear receptors and eIF2α-kinase is also valid for the other Hsp70- and Hsp90-regulated substrates. What renders these proteins substrates for chaperones for a prolonged time after de novo synthesis, and why do so many of them belong to signal transduction pathways? The dependence of these proteins on chaperone power during de novo synthesis argues for slow, complicated folding pathways. As molecular switches they may depend on extensive conformational changes and interdomain communications during the folding and activation process which may require chaperone assistance. Furthermore, cells may have a particular need for repression of the activity of signal transducers in the uninduced state since untimely active transducers may be fatal for the cell. Finally, the survival of cells after stress treatment may require a connection between the signal transduction pathways and the chaperones which themselves can sense stress and regulate the heat shock response accordingly. III. MECHANISM OF ACTION A. General Aspects Hsp70 homologs share the same overall structure, consisting of an N-terminal ATPase domain of 45 kDa and a C-terminal substrate-binding domain of at least 25 kDa, which is further subdivided into a β-sandwich subdomain of 15 kDa and a C-terminal α-helical subdomain (Figs. 1 and 3a). Sequence alignment of the meanwhile more than 500 different Hsp70 sequences in the databases has revealed a number of differences which may be arbitrary remnants of evolution in some cases but may translate into relevant functional differences in other cases. For most of these differences, a clear structure–function relationship has not been established; thus only a few striking examples are given for illustration. The most apparent diversity is observed in the Hsp110 and Hsp170 subfamilies of Hsp70 chaperones, with molecular weights of 110 kDa and 170 kDa, respectively (Easton et al., 2000). These two purely eukaryotic

Hsp70 CHAPERONE MACHINES

9

FIG. 1. Domain structure of the Hsp70 and Hsp70-related protein families. Schematic representation of the ATPase domain (∼385 residues, dark gray), a small linker, and the substrate-binding domain, which is subdivided into the β-sheet (∼150 residues, light gray) and the α-helical subdomains (∼100 residues, gradient). The white boxes in the Hsp110 and Hsp170 families represent insertions of unknown function.

homologs share a considerable degree of similarity in the N-terminal ATP-binding domain with the other Hsp70 chaperones but have extensive insertions in the C-terminal domains (Fig. 1). Interestingly, in mouse two variants of the Hsp110 were found, Hsp105α and Hsp105β, where part of the insertion is removed by splicing. Depending on the environmental conditions, one or both variants are produced (Yasuda et al., 1995). On the other extreme is the Ssz1 protein of S. cerevisiae which has a lower molecular weight (62 kDa) than usually found for Hsp70 proteins (Gautschi et al., 2001). The homology of Ssz1 to the classical Hsp70 proteins is low with ∼30% overall identity and 12% in the substrate binding domain. The closest homolog in Schizosaccharomyces pombe has only 40% identity. More subtle differences are also observed within the classical Hsp70 chaperones. The Hsp70 members from archaea and gram-positive eubacteria and the HscC homolog of E. coli lack a stretch of about 25 amino acids of unknown function. ATP binding to the ATPase domain of Hsp70 proteins decreases the afﬁnity of the substrate-binding domain for substrates 5- to 85-fold (Theodorakis et al., 1989; Palleros et al., 1993; Schmid, 1993; McCarty and Walker, 1994; Pierce et al., 1997; Takayama et al., 1999; Mayer et al., 2000) (Fig. 2). This decrease in afﬁnity is due to increases in the koff of Hsp70–substrate complexes by 2 to 3 orders of magnitude (Schmid et al., 1994; Theyssen et al., 1996; Mayer et al., 2000), and in the kon for substrate binding by approximately 50-fold (Schmid et al., 1994; Pierpaoli et al., 1997). The ATPase cycle of Hsp70 thus consists of an alternation between the ATP state with low afﬁnity and fast exchange rates for substrates, and the ADP state with high afﬁnity and low exchange rates for substrates. The molecular mechanism of the ATPase and substrate binding/release cycles have been analyzed in detail only for a few Hsp70 homologs, including DnaK and HscA (Hsc66) of E. coli, DnaK of T. thermophilus, Ssa1 of S. cerevisiae, bovine Hsc70, and BiP of hamster and S. cerevisiae.

10

MATTHIAS P. MAYER ET AL.

FIG. 2. Model of the chaperone cycle of DnaK. The rates shown are those determined for E. coli DnaK (Theyssen et al., 1996; Packschies et al., 1997; Russell et al., 1998; Slepenkov and Witt, 1998a; Slepenkov and Witt, 1998b; Laufen et al., 1999; Russell et al., 1999). The substrate association and dissociation rates have been determined using a high-afﬁnity peptide (Schmid et al., 1994; Mayer et al., 2000).

These proteins show differences in their cycles that have implications for their chaperone activities. We will use the available data to provide for these homologs a comparative description of the mechanistic features of the ATPase cycles, the substrate binding features, the coupling mechanism, and the regulation by co-chaperones. B. The ATPase Cycle 1. Nucleotide Association The ATPase domain is homologous to the structures of actin, hexokinase, and glycerokinase (Fletterick et al., 1975; Kabsch et al., 1990; Hurley et al., 1993), and consists of two large globular subdomains (I and II), each further divided into two small subdomains (A and B; Figs. 3a and b, see color insert). The subdomains are separated by a deep cleft at the bottom of which the nucleotide binds in complex with one Mg2+ and two K+ ions, involving both subdomains and the connecting helices (Flaherty et al., 1990). The X-ray structures of the bovine Hsc70 ATPase domain complexed with several adenosine nucleotides (ADP, AMPPNP, ATP to mutant proteins) revealed that the adenosine

Hsp70 CHAPERONE MACHINES

11

nucleotide is positioned in the active site by interactions with two β- and γ -phosphate-binding loops and a hydrophobic adenosine-binding pocket (Flaherty et al., 1990; O’Brien and McKay, 1993). Additional important contributions to tight nucleotide binding are contacts with the Mg2+ ion which is coordinated by several side chains of Hsc70, forming a hydrophilic pocket. The phosphate- and ribose-coordinating residues are highly conserved throughout the Hsp70 family, and even to some extent in actin and hexokinase. In contrast, adenine coordination seems to be relatively weak and is not conserved. This explains that at least the E. coli DnaK and HscC homologs which were analyzed in this respect are not entirely ATP-speciﬁc but can bind and hydrolyze other nucleotides as well (Liberek et al., 1991b; Kluck, Mayer, and Bukau, unpublished results). Hsp70 proteins have a much higher afﬁnity for ATP than for ADP. The K d values of E. coli and T. thermophilus DnaK and bovine brain Hsc70 for ATP are in the submicromolar range (Gao et al., 1994; Ha and McKay, 1994; Theyssen et al., 1996; Klostermeier et al., 1998), implying that at most conditions in vivo, these chaperones are predominantly in the ATPbound state which starts the functional chaperone cycle. This may not be the case for all Hsp70 chaperones. Escherichia coli HscA and yeast Ssb, for example, have K d values for ATP in the high micromolar range, which could lead to a nucleotide-free state under certain cellular conditions (Lopez-Buesa et al., 1998; Silberg and Vickery, 2000). Further dissection of the ATP-binding process revealed that it is composed of at least two steps: the formation of an initial weak complex followed by an isomerization (Ha and McKay, 1995; Theyssen et al., 1996; Slepenkov and Witt, 1998a; Slepenkov and Witt, 1998b). The ﬁrst step is very rapid, with rates of 1.3–7 ×105 M−1 s−1 (Ha and McKay, 1995; Russell et al., 1998) equivalent to ≥400 s−1 (τ ≤ 1.7 ms) at the cellular ATP concentration of 3 mM, implying that the association of ATP is not rate-limiting for the functional cycle of Hsp70 proteins. The isomerization step is slower [kiso = 0.2–1.5 s−1; τ = 0.5 − 3 s; Fig. 2 (Ha and McKay, 1995; Theyssen et al., 1996; Russell et al., 1999)] and probably reﬂects the precise lockingin of the nucleotide into the nucleotide-binding pocket. It requires the correct positioning of one Mg2+ and two K+ ions in the active site and does not occur with several ATP analogs (e.g., ATPγ S, AMPPNP). It is kinetically coupled to the release of bound substrate and therefore of crucial importance for the chaperone cycle. 2. ATP Hydrolysis a. Mechanism of Catalysis. ATP hydrolysis is the rate-limiting step in the ATPase cycle of most Hsp70 proteins investigated except HscA and T. thermophilus DnaK (Gao et al., 1993; McCarty et al., 1995; Karzai

12

MATTHIAS P. MAYER ET AL.

and McMacken, 1996; Theyssen et al., 1996; Pierpaoli et al., 1997; Klostermeier et al., 1998). The rates of hydrolysis are very low in the ground state, ranging between 3 × 10−4 to 1.6 × 10−2 s−1, respectively (Ha et al., 1999). Structural investigations and analysis of mutant proteins of bovine Hsc70 suggested a mechanism of hydrolysis (Ha et al., 1999). Accordingly, Mg2+ and ATP bind initially in a conformation not correctly aligned for hydrolysis, probably with an H2O molecule inserted between the Mg2+ ion and the γ -phosphate. Rearrangement of the γ -phosphate leads to a tight β,γ -bidentate complex with the Mg2+ ion. In this position, Lys-71 of the ATPase domain (Hsc70 numbering) is proposed to align a H2O or OH− molecule for in-line attack on the terminal phosphate, resulting in γ -phosphate cleavage. Efﬁcient catalysis requires the correct positioning of the Mg2+ ion, a task that necessitates an electrostatic surrounding which may be established by FIG. 3. Structural aspects of Hsp70 chaperones and their interaction with cochaperones. (a) Model of the structure of full-length Hsp70: Ribbon representation of the structure of the DnaK ATPase domain [modeled based on the structure of the ATPase domain of bovine Hsc70 (Flaherty et al., 1990) using SWISS-MODEL (Peitsch, 1995; Peitsch, 1996; Guex and Peitsch, 1997); rotated 180◦ compared to the standard view] with bound ADP+Pi (as ball-and-stick model) and the substrate-binding domain of E. coli DnaK (PDB entry code 1DKY; Zhu et al., 1996) with bound peptide substrate (as stick model). Indicated are residues with important features as mentioned in the text. A hypothetical orientation of the two subdomains is shown, the connection between the two domains being indicated by a dashed line. The isolated substrate-binding domain rotated by 90◦ is shown in addition. Indicated are the helices C, D, E which, together with helix B, build up the lid over the substrate-binding pocket. (b) Surface representation of the DnaK APTase domain and the DnaJ J domain of E. coli: The structure of the ATPase domain is rotated by 180◦ according to the standard view; the molecular surface and the charge distribution were calculated by the program GRASP (Nicholls et al., 1993). Left panel: Marked in green are the residues in the J domain involved in the interaction with DnaK, identiﬁed by NMR (Greene et al., 1998); marked in magenta are the residues in the ATPase domain (Y145, N147, D148, R167, E217, V218) important for the interaction with the J domain, identiﬁed by mutant analysis (G¨assler et al., 1998; Suh et al., 1998). Right panel: Surface representation colored by the electrostatic potential. The dashed lines indicate a hypothetical mode for the interaction between the two domains. (c) Complexes of Hsp70 chaperones and their nucleotide exchange factors: Left panel: X-Ray structure of the complex of the DnaK ATPase domain (shown as a ribbon model) and the GrpE dimer (shown in a molecular surface representation with marked N and C termini, PDB entry code 1DKG; Harrison et al., 1997). The subdomains of the ATPase domain, the two potential salt bridges (b, c), and the hydrophobic patch (a) shown to restrict nucleotide release (Brehmer et al., 2001) are indicated. Right panel: X-Ray structure of the complex of the human Hsc70 ATPase domain (shown as a ribbon model) and a fragment of Bag-1, PDB entry code 1HX1 (shown in a molecular surface representation together with a ribbon model with marked N and C termini). The subdomains of the ATPase domain and the potential salt bridge (c) are indicated.

Hsp70 CHAPERONE MACHINES

13

the binding of two K+ ions near the Mg2+ ion. Although the hydrolysis mechanism has so far only been investigated for Hsc70, it is most likely conserved within the Hsp70 family since the catalytic Lys-71 and all other phosphate- and Mg2+-coordinating amino acids are highly conserved. b. Stimulation by Substrates and DnaJ Proteins. The hydrolysis of ATP triggers the closing of the substrate-binding cavity and the locking-in of associated substrates. In a thermodynamically coupled process, substrates stimulate the hydrolysis of ATP typically 2- to 10-fold (Flynn et al., 1989; Jordan and McMacken, 1995; McCarty et al., 1995; Takeda and McKay, 1996; Theyssen et al., 1996; Barouch et al., 1997). This stimulation by substrates is too low to drive the functional cycle of Hsp70 chaperones. Instead, the activity of DnaJ co-chaperones is required for productive tight coupling of ATP hydrolysis with substrate association (Wawrzynow ´ et al., 1995; Karzai and McMacken, 1996; Barouch et al., 1997; Misselwitz et al., 1998; Laufen et al., 1999; Mayer et al., 1999). DnaJ co-chaperones are a heterogeneous class of multidomain proteins that share a conserved stretch of ∼70 residues often located at the N terminus (Fig. 4). This so-called J domain is essential for the interaction of DnaJ proteins with their Hsp70 partner proteins. The mechanism of DnaJ proteins is best understood for the E. coli member, DnaJ, and will be described here in some detail. Escherichia coli DnaJ and protein substrates synergistically stimulate the rate of ATP hydrolysis >1000-fold (Liberek et al., 1991a; Karzai and McMacken, 1996; Laufen et al., 1999; Russell et al., 1999). Peptide substrates are poorly effective in this synergistic action, suggesting a mechanism to avoid jamming of the DnaK chaperone with peptides (Laufen et al., 1999). The action of DnaJ requires both the binding of protein substrates to the central hydrophobic pocket of DnaK’s substrate-binding cavity and the structural coupling between DnaK’s ATPase- and substratebinding domains which transmits the substrate binding event to the catalytic center (Laufen et al., 1999). Furthermore, DnaJ’s coupling activity requires the ability to interact with DnaK through its J domain (Wall et al., 1994; Karzai and McMacken, 1996; Laufen et al., 1999) (see Fig. 3b) and with substrates through its C-terminal substrate-binding domain. DnaJ interacts with substrates in a rapid and transient fashion (Wawrzynow ´ and Zylicz, 1995; Gamer et al., 1996; Pierpaoli et al., 1997; Pierpaoli et al., 1998a). Its substrate-binding motif consists of a hydrophobic core of approximately eight residues enriched for aromatic and large aliphatic hydrophobic residues and arginine (Rudiger ¨ et al., 2001). Although this motif shows differences from DnaK’s binding motif, DnaJ and DnaK share the majority of binding sites in peptide sequences (Rudiger ¨ et al.,

14

MATTHIAS P. MAYER ET AL.

FIG. 4. DnaJ protein family. (a) Domain structure of the three DnaJ subfamilies: The different domains are marked in the following way: J, J domain; G/F, Gly-Phe-rich region; Zn, Zn2+-binding domain. The numbering belongs to a speciﬁc member of each subfamily, e.g., human DjA1/Hdj-2 (type I), human DjB1/Hdj-1 (type II), and bovine DjC5/Csp and human kinesin light chain (KLC) (type III). (b) Secondary structure representations of DnaJ protein domains: NMR structure of the J domain of E. coli DnaJ (PDB entry code 1XBL; Szyperski et al., 1994; Pellecchia et al., 1996); marked is the conserved HPD motif as a ball-and-stick model. NMR structure of E. coli DnaJ Zn-binding domain (PDB entry code 1EXK; Martinez-Yamout et al., 2000); marked are the two Zn2+ ions coordinated by four Cys residues shown as a ball-and-stick model. X-Ray structure of the yeast Sis1 substrate-binding domain (PDB entry code 1C3G; Sha et al., 2000); indicated are the two subdomains. The homology between DnaJ and Sis1 starts at Gly-210 marked in the ﬁgure as G210 (compare Rudiger ¨ et al., 2001).

Hsp70 CHAPERONE MACHINES

15

2001). Because DnaJ does not differentiate between L- and D-amino acids, it is not restricted by backbone contacts, allowing it to scan hydrophobic protein surfaces (Feifel et al., 1998; Rudiger ¨ et al., 2001). On the basis of the available data, a model for the action of DnaJ has been proposed (Karzai and McMacken, 1996; Bukau and Horwich, 1998; Laufen et al., 1999). Accordingly, DnaJ starts the functional cycle of the DnaK system by rapid and transient association with the substrate through its scanning function, although the cycle may also start by the association of DnaK · ATP with substrates. The substrate can then be transferred onto DnaK in a two-step process involving the transient interaction of the J domain with DnaK · ATP and the association of the substrate with the open substrate-binding cavity of DnaK. It is an open question whether, at this stage, DnaJ and DnaK both contact the substrate at distinct hydrophobic patches, thereby forming a transient ternary complex, or whether the substrate is transferred from DnaJ to DnaK in a handover mechanism involving a single hydrophobic binding site. Through an interdomain communication, the association of substrates and the interaction with the J domain lowers the activation energy for the hydrolysis of ATP by DnaK. According to this model, DnaJ can act catalytically, consistent with experimental in vitro observations (Liberek et al., 1995; McCarty et al., 1995; Pierpaoli et al., 1998b) and the substoichiometric concentration of DnaJ with respect to DnaK in vivo (Bardwell et al., 1986; Neidhardt, 1996; Tomoyasu et al., 1998). Recent studies provided insights into the structural basis of this coupling mechanism. One important step forward was the elucidation of the NMR structure of the J domain (Fig. 4b). The J domain has a stable hydrophobic core built up by three α-helices (Szyperski et al., 1994; Pellechia et al., 1996; Qian et al., 1996). The amphipathic, antiparallel helices II and III form a short coiled-coil that is stabilized by helix I which extends from the C terminus of helix III to the center of helix II. A fourth helix is positioned at the end of helix III on the same side of the coiledcoil as helix I but perpendicular to the plane formed by helices II and III. The ﬂexible disordered loop connecting helices II and III contains the universally conserved histidine–proline–aspartic acid (HPD) motif essential for the functional interaction with Hsp70 chaperones (Feldheim et al., 1992; Wall et al., 1994). The ﬂexibility of the loop suggests that binding to DnaK involves an induced ﬁt–like mechanism. Using NMR and 15N-labeled J domains, Landry and co-workers (Greene et al., 1998) were able to map the residues in the J domain that interact with the ATPase domain of DnaK (Tyr-6, Ser-13, Glu-20, Ile-21, Arg-22, Ala-24, Lys-26, Arg-27, Leu-28, Met-30, Tyr-32, His-33, Asp-35, Tyr-54, and Thr-58). On the side of DnaK, genetic and

16

MATTHIAS P. MAYER ET AL.

biochemical analysis identiﬁed a conserved channel on the lower back side of the standard view of the ATPase domain of DnaK responsible for the functional interaction with DnaJ (G¨assler et al., 1998; Suh et al., 1998). Figure 3b shows two surface representations of the J domain and of the ATPase domain turned 180◦ with respect to the standard representation, indicating the surface potential and the interacting residues of the J domain and the channel. Combined available data suggest that the mode of interaction between the J domain and the ATPase domain relies on electrostatic forces and operates by an induced-ﬁt mechanism that brings the HPD motif in close proximity to the catalytic center of the ATPase domain. The rearrangements in the catalytic center that facilitate γ -phosphate cleavage are not yet clear. The synergistic stimulation of ATP hydrolysis by substrates and DnaJ proteins has also been observed in other Hsp70 systems, including the action of auxilin and clathrin on Hsc70 (Barouch et al., 1997), of Sec63 and substrates on BiP (Misselwitz et al., 1998; Misselwitz et al., 1999), and of HscB and IscU on HscA (Silberg and Vickery, 2000). Thus, the coupling activity of DnaJ proteins appears to be a mechanism that is conserved at least in some Hsp70 chaperones. However, the exact mechanism of coupling may be subject to evolutionary variation, as suggested for the BiP–Sec63 situation. It has been claimed that the J domain of Sec63 is sufﬁcient to stimulate ATP hydrolysis by BiP 50-fold, in contrast to the J domain of E. coli DnaJ which requires substrates for stimulation of DnaK’s ATPase, thereby allowing the trapping of substrates. An even more dramatic difference exists for the T. thermophilus DnaK system which lacks this coupling mechanism since ATP binding does not induce substrate release nor do DnaJ and substrates induce ATP hydrolysis (Klostermeier et al., 1998; Klostermeier et al., 1999; Groemping et al., 2001). The molecular basis for these mechanistic differences is unclear. 3. Nucleotide Dissociation a. Basic Mechanism. The next step in the ATPase cycle, the release of ADP and Pi, allows the subsequent rapid binding of ATP and, consequently, the release of bound substrates and reestablishment of the starting point of the chaperone cycle. Nucleotide dissociation is thus a crucial step in the cycle, it is therefore not surprising that this step is highly regulated and subject to strong evolutionary variation. The dissociation of bound nucleotides requires an opening of the nucleotide binding cleft. Such movement has been observed for E. coli DnaK and bovine Hsc70 when their respective nucleotide exchange factors, GrpE and Bag-1, are bound. In the absence of such factors, the equilibrium between the closed and the opened states is probably the determining factor for the intrinsic rates of nucleotide dissociation. For

Hsp70 CHAPERONE MACHINES

17

DnaK, the ADP dissociation rates are very low, being 0.004–0.035 s−1 and 0.0004–0.0014 s−1 in the absence and presence of Pi, respectively (Ha and McKay, 1994; Ha and McKay, 1995; Theyssen et al., 1996; Klostermeier et al., 1998; Russell et al., 1998; Slepenkov and Witt, 1998a; Brehmer et al., 2001). Pi is thus rate-limiting for the dissociation of ADP, which implies that Pi release precedes ADP release. The ADP dissociation rate is 2- to 60-fold faster than the rate of hydrolysis in the unstimulated cycle (McCarty et al., 1995; Theyssen et al., 1996), nevertheless, it becomes ratelimiting when the hydrolysis of ATP is stimulated by DnaJ and substrates. For bovine Hsc70, the dissociation rates for ADP ± Pi are about 20-fold higher than those of DnaK ± Pi (Ha and McKay, 1994; Ha and McKay, 1995; Brehmer et al., 2001). Even more dramatic differences exist for HscA, which has dissociation rates for ADP ± Pi that are 700-fold higher than those of DnaK (Silberg and Vickery, 2000; Brehmer et al., 2001). What is the structural basis for these strong kinetic differences between Hsp70 homologs in nucleotide dissociation? Although the modeled ATPase domain structures of DnaK, HscA, and Hsc70 are almost completely superimposable, they show subtle differences which allow classiﬁcation of the entire Hsp70 family into three subfamilies with E. coli DnaK, E. coli HscA, and human Hsc70 as prototypes (Brehmer et al., 2001). First, variations exist in an exposed loop in subdomain IIB near the nucleotide binding cleft (Fig. 3c). DnaK subfamily members share a particularly long loop (A276–R302 in E. coli DnaK) with a subfamilyspeciﬁc sequence. Members of the Hsc70 subfamily share a loop with a subfamily-speciﬁc sequence whose tip is four residues shorter, while members of the HscA subfamily share a loop that is ten residues shorter and less conserved in sequence. Second, variations exist in the interface of the nucleotide binding cleft. DnaK proteins contain a hydrophobic patch (L257–V59 of DnaK) at the top of the cleft and two putative salt bridges (E264–R56, upper; E267–K55, lower) which are mainly responsible for the polarity of the interface (a, b, c in Fig. 3c). In contrast, Hsc70 proteins lack a DnaK-like hydrophobic patch and the upper salt bridge, while HscA proteins lack all three elements. The loop and the salt bridges constitute a device that allows rapid association of ATP and slow dissociation of ATP and ADP ± Pi. Mutation of these elements in DnaK gradually converts this homolog into Hsc70 and HscA with respect to nucleotide exchange (Brehmer et al., 2001). The salt bridges together with the hydrophobic contact probably function in a mousetraplike fashion to allow tight closure of the nucleotide binding cleft. The role of the loop is less obvious. Given the considerable size of its exposed part, it has been speculated that it may reach over to subdomain IB of the ATPase domain of DnaK, thereby acting as a latch (Brehmer et al., 2001).

18

MATTHIAS P. MAYER ET AL.

It is intriguing that the mutational alterations of the salt bridges and the loop have signiﬁcant effects on the chaperone activity of DnaK. Even small disturbances of the kinetics of nucleotide release, and hence of substrate release, were found sufﬁcient to affect DnaK’s activity as chaperone. Fine-tuning of the ATPase motor of DnaK at the level of nucleotide exchange is therefore of major functional importance, rendering nucleotide exchange a prime target for regulatory devices. The large variability within the Hsp70 family in nucleotide exchange rates is likely to contribute to the functional diversiﬁcation of Hsp70 chaperone systems. b. Stimulation by Nucleotide Exchange Factors. The differences between Hsp70 homologs in nucleotide dissociation go along with differences in the regulation of this step. DnaK homologs that have the described trap to prevent nucleotide dissociation require the nucleotide exchange factor, GrpE, for chaperone activities in vivo and in vitro (Alfano and McMacken, 1989; Zylicz et al., 1989; Langer et al., 1992; Schr¨oder et al., 1993; Skowyra and Wickner, 1993; Wyman et al., 1993). HscA homologs which lack this device and have high nucleotide dissociation rates neither interact with GrpE nor appear to have their own exchange factor (Silberg et al., 1998; Brehmer et al., 2001). Mammalian Hsc70 homologs that have intermediate dissociation rates possess chaperone activity without exchange factor. However, they can interact with a nucleotide exchange factor, Bag-1, perhaps to fulﬁll special chaperone activities (see below). The two known nucleotide exchange factors for Hsp70 proteins, GrpE and Bag-1, are entirely speciﬁc for their particular partner chaperones (Brehmer et al. 2001); furthermore, they have entirely different structures and mechanisms, although they generate the same conformational open state of their target chaperone (Harrison et al., 1997; Briknarova et al., 2001; Sondermann et al., 2001). These proteins may therefore have been generated by convergent functional evolution as in the case of the nucleotide exchange factors for G proteins (Cherﬁls and Chardin, 1999). Recently, yet another unrelated protein, Sls1/PER100, was proposed to be a nucleotide exchange factor for the ER resident Hsp70 homolog BiP (Kabani et al., 2000). If the Sls1 protein turns out to stimulate nucleotide dissociation as well, this would be another independently evolved nucleotide exchange factor of the Hsp70 system. i. GrpE. GrpE has a molecular weight of 22 kDa and forms a stable dimer in solution (Sch¨onfeld et al., 1995). The X-ray structure of GrpE in complex with the ATPase domain of DnaK was solved to 2.8 A˚ resolution (Harrison et al., 1997) (Fig. 3c). The GrpE mutant protein used for crystallization was truncated, missing the ﬁrst 33 residues, in addition, it had the Gly-122 to Asp exchange of the GrpE280 mutant protein (Ang et al.,

Hsp70 CHAPERONE MACHINES

19

1986; Wu et al., 1994). The N terminus of the truncated GrpE monomer ˚ Both helices constitutes an α-helix with the remarkable length of 100 A. of the GrpE dimer lie in the same plane and interact tightly without forming a classical coiled-coil. The N-terminal helix is connected over a rather unstructured loop to two additional short α-helices and a small β-sheet domain composed of six short β-strands. In addition to the N-terminal helix, the dimer interface is formed by the two short helices of each monomer which together build up a four-helix bundle. In the crystal structure GrpE forms an asymmetric bent dimer whereby only one GrpE monomer contacts the ATPase domain of DnaK through ﬁve major contact sites including an exposed loop identiﬁed biochemically as essential for stable GrpE–DnaK interaction (Buchberger et al., 1994). The bent structure of the GrpE dimer prevents it from interacting with two DnaK molecules at the same time, consistent with the GrpE2 : DnaK stoichiometry of the complex identiﬁed in solution (Sch¨onfeld et al., 1995) (Fig. 3c). GrpE binds to the ADP-bound and nucleotide-free states of DnaK with high afﬁnity (K d = 1 nM). Addition of ATP leads to the dissociation of the complex (Zylicz et al., 1987; Sch¨onfeld et al., 1995). As described above, the superposition of the structure of the nucleotide-free ATPase domain of DnaK in complex with GrpE with the structure of the ADPand Pi-bound ATPase domain of Hsc70 (Flaherty et al., 1990) shows that subdomain IIB in DnaK is rotated outward by 14◦ , thereby opening up the nucleotide binding cleft. This rotation displaces residues Ser-274, ˚ The corresponding amino Lys-270, and Glu-267 of DnaK by 2 to 3 A. acids in Hsc70 are involved in the coordination of the adenine and the ribose rings of the bound ADP. Furthermore, residues that constitute the hydrophobic patch and the upper salt bridge (E264–R56) interact with GrpE (Harrison et al., 1997). Figure 3c (left panel) shows how GrpE loosens the ATPase interface by interfering with the hydrophobic contact and the salt bridges. The mechanism by which GrpE accelerates the nucleotide release thus seems to rely on an active opening of the nucleotide binding cleft. Furthermore, GrpE stabilizes the open conformation of the nucleotide binding pocket, which facilitates the rapid binding of ATP to the nucleotide-free state of DnaK (Brehmer et al., 2001). In addition, it has been proposed that the long N-terminal α-helices together with the missing residues of the N terminus reach to the substrate-binding domain, thereby inﬂuencing substrate binding and release (Harrison et al., 1997). Experimental evidence for such an acceleration of substrate release by GrpE has also been reported (Harrison et al., 1997; Mally and Witt, 2001). However, an intensive investigation of the release kinetics of peptide and protein substrates could not conﬁrm this possibility (Brehmer, Mayer, and Bukau, unpublished results).

20

MATTHIAS P. MAYER ET AL.

ii. Bag. Bag proteins form a heterogeneous family of multidomain proteins which share the Bag domain frequently located at their C termini (Takayama et al., 1999; Thress et al., 2001). This domain is essential and sufﬁcient for the stimulation of nucleotide exchange by the mammalian homologs Hsc70 and Hsp70 (G¨assler et al., 2001). The nucleotide exchange activity of Bag proteins has been ﬁrst described for the Bag-1 homolog by H¨ohfeld and co-workers (H¨ohfeld and Jentsch, 1997; L¨uders et al., 1998; L¨uders et al., 2000b). Recent kinetic measurements using ﬂuorescent-labeled nucleotides showed that Bag-1 stimulates dissociation of ADP from Hsc70 and Hsp70 up to 100-fold in the absence of inorganic phosphate and 600-fold in its presence (G¨assler et al., 2001). Interestingly, Bag-1 does not stimulate the dissociation of ATP and is strongly affected by Pi which increases the apparent K d approximately 17-fold. This is in contrast to the interaction of GrpE with DnaK which also stimulates ATP dissociation and is not affected by Pi. These observations suggest that the mechanisms by which GrpE and Bag-1 stimulate nucleotide release are different and that Bag-1 plays a more passive role than GrpE. During the intrinsic ﬂuctuations between the open and closed states Bag-1 may only bind and stabilize the open conformation of the ATPase domain of Hsc70. The opening frequency of the ATPase domain most likely depends on the nucleotide bound, with Hsc70 · ADP > Hsc70 · ADP + Pi > Hsc70 · ATP. In contrast, the structural and biochemical data indicate that GrpE binds to both the open and the closed conformations and actively shifts the equilibrium to the open state. According to such a mechanism, GrpE would be less inﬂuenced by the nucleotide that is bound to DnaK. The comparison of the two crystal structures is consistent with this model (Harrison et al., 1997; Sondermann et al., 2001). The mammalian Bag domain of Bag-1 forms a three-helix bundle (Briknarova et al., 2001; Sondermann et al., 2001). It associates as a monomer with subdomains IB and IIB of the Hsc70 ATPase through electrostatic interactions involving the conserved residues Glu-212, Asp-222, Arg-237, and Gln-245 in Bag-1. Although Bag-1 contacts residues in Hsc70 that differ from those contacted by GrpE in DnaK, both exchange factors cause the same 14◦ outward rotation of subdomain IIB through a hinge. Two residues in the ATPase domain of Hsc70, Arg-261 and Glu-283, which have been identiﬁed in the crystal structure, are important for Bag-1 binding (Sondermann et al., 2001). While Arg-261 is conserved in members of the DnaK and Hsc70 subfamilies except for the ER-resident BiP homologs (replaced by Lys), the Glu-283 residue is conserved only in the Hsc70 subfamily. These differences contribute to the exclusive speciﬁcity of GrpE for DnaK and Bag-1 for Hsc70 (Brehmer et al., 2001).

Hsp70 CHAPERONE MACHINES

21

C. Substrate Binding 1. Structure of the Substrate-Binding Domain The substrate-binding domain shows high sequence conservation within the Hsp70 family, although differences of unclear functional relevance exist. Most structural information on this domain exists for E. coli DnaK. The X-ray structure of the substrate-binding domain was solved in complex with a heptameric peptide substrate at 2.0 A˚ resolution (Fig. 3a) (Zhu et al., 1996). The peptide-binding moiety (residues 389 to 607) forms a sandwich of two four-stranded β-sheets with four upwardprotruding loops (two inner and two outer loops) and two helices, A and B, which are packed against the inner loops (L1,2 and L4,5). The substrate-binding cavity is formed by the β-sheets 1 and 2 and the loops, L1,2 and L3,4. Helix B constitutes a lid which closes the cavity through electrostatic contacts to the outer loops L3,4 and L5,6. ATP binding to DnaK must at least open this lid to allow substrate release. The distal part of helix B together with helices C, D, and E forms a hydrophobic core of unknown function. Not included in the crystallized fragment were the C-terminal 30 residues of DnaK for which no function has yet been assigned. This peptide stretch was shown by NMR structure analysis of the C-terminal 100 residues of DnaK to be largely unstructured (Bertelsen et al., 1999). In prokaryotic Hsp70 proteins this C terminus is conserved only inasmuch as it is highly charged, but in eukaryotic Hsp70 proteins the C terminus ends with the highly conserved EEVD motif which they share with the Hsp90 proteins. Deletion or alteration of this motif to AAAA decreased the stability of Hsp70 against proteolytic digest in both nucleotide states. At the same time these mutations increased the basal ATPase rate of Hsp70 up to 2.5-fold and decreased afﬁnity for substrates (Freeman et al., 1995). As the isolated ATPase domain has the same ATPase rate as the full-length protein, these data suggest that the EEVD motif is involved in interdomain communication. Consistent with this interpretation is the fact that the intrinsic ﬂuorescence of the single Trp residue in the ATPase domain, which is indicative of interdomain communication (Buchberger et al., 1995; Ha and McKay, 1995), was decreased signiﬁcantly (Freeman et al., 1995). Interestingly, a number of eukaryotic co-chaperones interact via their tetratricopeptide repeats (TPR) with the very same EEVD motif (see below), but it is not known whether these interactions affect the interdomain communication. The substrate peptide which has been cocrystallized with the substratebinding domain of DnaK is contacted over a stretch of ﬁve residues by two types of interactions (Zhu et al., 1996; Rudiger ¨ et al., 1997a). Hydrogen bonds between the backbone of the cavity-forming loops

22

MATTHIAS P. MAYER ET AL.

(L1,2 and L3,4) and the backbone of the substrate mediate the recognition of the extended peptide conformation. Van der Waals interactions of hydrophobic side chains lining the substrate-binding cavity with the substrate side chains account for DnaK’s preference for hydrophobic peptide segments. The central substrate residue, referred to as position 0, reaches down into a deep hydrophobic pocket of the substrate-binding cavity which accommodates large hydrophobic amino acids, especially Leu (Zhu et al., 1996). This central substrate residue is completely buried from solvent. The adjacent residues of the bound substrate at positions −1 and +1 are only partially buried from solvent. They point upward and contact Met-404 and Ala-429, which form a hydrophobic arch bridging over the substrate peptide backbone. The surface potential of the substrate-binding domain surrounding the cavity is mainly negative, which explains the contribution of positive charges in peptides to the afﬁnity for DnaK. The architectural constraints imposed by the substrate-binding cavity agree well with the experimentally determined binding motif of DnaK (Rudiger ¨ et al., 1997b). By providing essential backbone contacts, the cavity allows DnaK to recognize only peptides composed of L-amino acids but not peptides composed of D-amino acids (Feifel et al., 1998; Ru¨ diger et al., 2001). Furthermore, DnaK has different afﬁnities for peptides with identical amino acid composition but authentic or inverse sequence (Rudiger ¨ et al., 2001). This implies that a peptide exhibiting lower afﬁnity for DnaK in the authentic sequence than a peptide in the inverse sequence cannot simply “turn around” and bind to DnaK with the N terminus on the other side. DnaK–peptide interactions therefore require a speciﬁc direction of the backbone. The evolutionary conservation of the binding cavity (Rudiger ¨ et al., 1997a) suggests that this feature is conserved within the Hsp70 family. This directionality of the binding mode may have consequences for substrate recognition, e.g., by Hsp70 homologs which interact with translocating polypeptides in an oriented fashion with respect to their N and C termini. Three structural elements of the cavity are of prime importance for substrate binding: the α-helical lid, the arch enclosing the substrate peptide and the central hydrophobic pocket. Removal of the α-helical lid by truncation of the C terminus starting at a hinge point within helix B decreases the afﬁnity of DnaK for peptide and protein substrates 3- to 7-fold, primarily by increasing the dissociation rates (Mayer et al., 2000; Pellecchia et al., 2000). Similarly, replacement of the lid residue Glu-543 in bovine Hsc70 (corresponding to Asp-540 in DnaK), which forms a salt bridge with the outer loop residue Arg-467, by Lys leads to a 3- and 4.5-fold increase in K d and koff (Ha et al., 1997). These data are consistent

Hsp70 CHAPERONE MACHINES

23

with data obtained by surface plasmon resonance spectroscopy for the ER homolog BiP (Misselwitz et al., 1998). Surprisingly, the total removal of the α-helical lid had no more dramatic effects on substrate binding by DnaK than the weakening of the electrostatic interactions of the helix B with the outer loops (Ha et al., 1997; Mayer et al., 2000). Therefore, the contacts seem to be the major contribution of the lid to the stability of the Hsp70–substrate complex. Disturbing the arch by alteration of Met-404 to Ala or Ala-429 to Trp also decreases the afﬁnity of DnaK for substrates 2- to 4-fold in the ADP state, with the major effects being on the dissociation rates (Mayer et al., 2000). Alterations in the arch also confer changes in the speciﬁcity of DnaK for some peptide substrates (Rudiger ¨ et al., 2000). This ﬁnding is interesting with regard to the fact that the arch-forming residues are the only substrate-contacting residues of the cavity which are subject to considerable evolutionary variation. It is therefore possible that the functional specialization of Hsp70 chaperones involves adaptive variation of the arch and the substrate speciﬁcity. The main contribution to the binding afﬁnity for substrates, however, comes from the central hydrophobic pocket. Introducing a steric hindrance in this pocket by alteration of Val-436 to Phe increased the dissociation equilibrium constants of DnaK 20- and 40-fold for protein and peptide substrates, respectively (Mayer et al., 2000). Interestingly, this decrease in afﬁnity was caused solely by generally decreased association rates while the dissociation rates in ADP and ATP states were identical to those of wild-type DnaK. Substrates may therefore bind to the mutant protein in a way very similar to that of wild-type DnaK, suggesting that the hydrophobic pocket has a considerable degree of ﬂexibility and that substrates bind in an induced-ﬁt–like manner. Such a high degree of ﬂexibility of the hydrophobic pocket was also suggested by the NMR structures of various DnaK and Hsc70 substrate-binding domain fragments (Wang et al., 1998; Morshauser et al., 1999; Pellecchia et al., 2000). In the ﬁrst two structures (Wang et al., 1998; Morshauser et al., 1999) the distal part of helix B (532–553 and 531–540) forms a loop which binds as substrate into the substrate-binding cavity. Probably owing to intramolecular constraints, the hydrophobic pocket is wider in these structures than in the crystal structure. In the third NMR structure the residue Phe-426 which contributes to forming the hydrophobic pocket shows a high degree of freedom, moving in and out of the pocket and only becoming ﬁxed upon substrate binding (Pellecchia et al., 2000). Despite the structural ﬂexibility of the pocket, introducing a bulky side chain into the hydrophobic pocket is fatal for chaperone activity in vitro and in vivo. Even high concentrations of the DnaK-V436F mutant

24

MATTHIAS P. MAYER ET AL.

could not refold chemically denatured luciferase or rescue a dnaK strain for growth at 40◦ C (Mayer et al., 2000). Other mutations affecting the conformation of the hydrophobic pocket, especially the positioning of Phe-426, also decreased the afﬁnity for substrates and rendered the respective proteins inactive in vivo and in vitro (Montgomery et al., 1999) 2. Importance of Substrate Affinity for Chaperone Activity The above discussion has mainly considered the interaction of Hsp70s with substrates in the ADP state. However, Hsp70s encounter substrates in the ATP state and enclose them on ATP hydrolysis. This raises the question of whether it is functionally relevant to investigate the substrate afﬁnity and speciﬁcity in the ADP state. This question was also fueled by a study that showed by surface plasmon resonance spectroscopy that the BiP homolog binds to all peptides tested when ATP and the J domain of Sec63 are present (Misselwitz et al., 1998). The authors concluded that Hsp70 proteins act as an unspeciﬁc trap for any substrate peptide provided that a J domain is present to stimulate ATP hydrolysis. This unspeciﬁc trap mechanism is difﬁcult to reconcile with the high conservation of the substrate-binding cavity which creates a hydrophobic environment for the bound substrate. An additional argument against this mechanism comes from the analysis of DnaK mutant proteins which differ in their afﬁnities for protein substrates (Mayer et al., 2000). For all mutant proteins, the K d s of complexes between several unrelated peptides and DnaK · ATP and those between these peptides and DnaK · ADP are correlated. This indicates that the architecture of the substratebinding cavity is similar in both nucleotide states, hence measurements of substrate afﬁnities in the ADP state are functionally relevant. These ﬁndings furthermore triggered the proposal that the substrate-binding domain of DnaK alternates permanently between the open and the closed conformation, and that ATP acts by increasing the frequency of transition to and the duration of the open state. This proposed mechanism is similar to the observed behavior of ion channels (Franciolini and Petris, 1988). The analysis of the DnaK mutant proteins also revealed that the synergistic stimulation of the ATPase activity by substrate and DnaJ depends for all mutant proteins on their afﬁnity for the substrate. Finally, the efﬁciency of refolding of chemically denatured luciferase by these DnaK mutant proteins correlated well with their afﬁnity for substrates in the ATP state (Mayer et al., 2000). The afﬁnity for substrates is therefore a crucial determinant for the locking-in process and the chaperone activity of DnaK. This does not rule out the possibility of cellular situations in which Hsp70s are “forced” into an association with almost any substrate. BiP may be in such a situation when targeted through the action of Sec63 to the translocating polypeptides emerging at translocation pores.

Hsp70 CHAPERONE MACHINES

25

3. The Coupling Mechanism: Nucleotide-Controlled Opening and Closing of the Substrate-Binding Cavity a. Conformational Changes in the Substrate-Binding Cavity. All available structures of the substrate-binding domain are believed to represent the high-afﬁnity state of the Hsp70 chaperones. Since the architectural constraints of the binding cavity do not permit the dissociation of substrates through a lateral sliding mechanism, an opening of the lid, the arch, and the cavity must occur to allow substrate release. Association and dissociation of peptide and protein substrates are observed in the ADP state of Hsp70 chaperones (Flynn et al., 1989; Gragerov et al., 1994; Schmid et al., 1994; Wawrzynow ´ and Zylicz, 1995; Banecki and Zylicz, 1996; Gamer et al., 1996; Pierpaoli et al., 1997; Misselwitz et al., 1998), e.g., with association rates of peptide–DnaK · ADP complexes of 0.003–0.08 s−1 at 1 μM DnaK (Pierpaoli et al., 1998a). This demonstrates that transitory opening of the substrate binding cavity also occurs, albeit slowly, in the absence of ATP. Christen and co-workers investigated the association of several related peptides to DnaK · ADP and found similar association rates for all peptides (Pierpaoli et al., 1998a). They concluded that substrate association to DnaK · ADP is limited by the opening rate of the substrate-binding cavity. This opening rate in the ADP status of DnaK is limited by the lid and the arch, as revealed by DnaK mutant analysis (Mayer et al., 2000). Both structural elements act independently in the control of substrate binding since the combination of lid truncation and arch alterations had only additive effects. In the ATP state, the association of ﬂuorescent peptide substrates to DnaK · ATP occurs at a rate of 105 –106 M−1 s−1 (Schmid et al., 1994; Gisler et al., 1998; Pierpaoli et al., 1998a; Mayer et al., 2000), and of a ﬂuorescent peptide to Hsc70 · ATP at a slower rate of 700 M−1 s−1 (Takeda and McKay, 1996). Although these association rates are orders of magnitude higher than those in the ADP state, they are far from diffusion-controlled (109 M−1 s−1). This feature is consistent with the existence of a slow isomerization step, e.g., in the substrate needed to generate a ﬂuorescent signal of the bound substrate. Alternatively, the substrate-binding cavity may sometimes be closed even in the ATP state. The mechanism of the nucleotide-controlled opening and closing of the substrate-binding domain is still a matter of debate. Based on crystallographic evidence, Hendrickson and co-workers suggested that ATP binding may induce an upward movement of helix B starting at a hinge point (residues 536 to 538 of DnaK) located above the cavity (Zhu et al., 1996). This movement would disrupt the interactions of helix B with the outer loops and consequently destabilize the outer loops and open up the substrate-binding cavity. However, recent structural and biochemical

26

MATTHIAS P. MAYER ET AL.

data indicate a more complicated mechanism. The β-sheet domain of the NMR structure of the DnaK fragment 386–561 is superimposable to the crystal structure, although this fragment lacks the distal part of helix B including the contacts to the outer loops (Wang et al., 1998). Deletion of the helical lid at the proposed hinge point increases the dissociation rates for bound substrates only approximately 5-fold compared to the 400- to 2500-fold increase in the dissociation rate induced by ATP in case of wild-type DnaK (Schmid et al., 1994; Mayer et al., 2000). Furthermore, lidless mutant proteins of DnaK and BiP and Hsc70 have retained substantial activity as chaperones in vivo and in vitro and still respond to ATP binding with a release of bound substrates (Ungewickell et al., 1997; Misselwitz et al., 1998; Tokunaga et al., 1998; Mayer et al., 2000). Even the isolated β-sheet domain lacking all helices A to E is not in an open conformation (Pellecchia et al., 2000). An alternative opening mechanism which we propose involves the short loop between helix A and helix B as a hinge. This hypothesis takes into account the fact that helix B is not as tightly bound to the β-sheet domain as originally assumed. It has hydrophilic residues at the interphase between the helices and the β-sheets which is unusual for a corelike stable structure. A third alternative was suggested by Zuiderweg and co-workers (Morshauser et al., 1999). They proposed that the opening of the substrate-binding cavity is initiated by a rotational movement of helix A around its length axis, which in turn would move the helical lid sideward away from the substrate-binding cavity. This hypothesis is based on the observation that the substrate-binding domain contains a hydrophobic depression near helix A to which a peptide was bound in their NMR structure. This depression is proposed to allow the binding of a hydrophobic side chain of helix A whereby the rotated conformation is stabilized (Wang et al., 1998; Morshauser et al., 1999). b. Structural Coupling between the ATPase Domain and the SubstrateBinding Domain. Little is known about the coupling mechanism that brings together all the events that control the frequency of opening and closing of the substrate-binding pocket: the DnaJ- and substratestimulated ATP hydrolysis which stabilizes the closed state, and the ATPstimulated stabilization of the open state. For discussion of the coupling mechanism, two aspects can be distinguished: the mechanics and the energetics of coupling. The mechanics of coupling considers the contact sites between the ATPase and substrate-binding domains which mediate the nucleotide-dependent interdomain communication. The energetics of coupling considers the necessary energy transmissions. Binding of ATP frees energy which is necessary to surmount the activation barrier from the closed to the open conformation of the substrate-binding

Hsp70 CHAPERONE MACHINES

27

domain, while binding of substrate and DnaJ delivers the energy to overcome the activation barrier for ATP hydrolysis. Several mutational alterations have been identiﬁed that affect the coupling mechanism. Some of them are clearly involved in ATP binding (Thr-13 to Ser, Ala, or Val in Hsc70; Sousa and McKay, 1998) or lower the activation barrier for the transition to the open substrate-binding domain conformation (Glu-543 to Lys in Hsc70; Ha et al., 1997). Other residues were found in sites where such an obvious effect on the energetics of coupling could not easily be implicated. These residues may therefore be necessary for the transmission of the signals (see Fig. 3a for the mutated sites). The strongest coupling defects were detected in DnaK mutant proteins with alterations in the highly conserved segment, referred to as the linker, which connects the ATPase domain with the substrate-binding domain. The exchanges VLLL(389–392) to AAAA and LL(390, 391) to DD completely abolish the interdomain communication in the sense that ATP fails to release bound substrates and substrates with DnaJ fail to stimulate ATP hydrolysis (Laufen et al., 1999). Interestingly, this linker was found in two different positions in the crystal structure of the substratebinding domain: an extended position as shown in Fig. 3a and a position bound to a hydrophobic depression formed by the β-sheet domain proximal to helix A. The existence of these two structural variants led Hendrickson and co-workers to suggest that the linker may undergo an ATP-dependent movement in and out of this depression as part of a process that allows the ATPase domain to dock onto the substrate-binding domain. The linker may, in addition, be directly involved in the stimulation of ATP hydrolysis. Additional residues implicated in the coupling mechanism were identiﬁed in the substrate-binding domain in close proximity to the hydrophobic depression [K414 (Montgomery et al., 1999), G455 (Buchberger et al., 1999), N415 (Mayer and Bukau, unpublished results)] and in the ATPase domain in spatial proximity to the linker in a conserved channel that is also involved in binding of the J domain (G¨assler et al., 1998). Mutational alterations of these residues resulted in moderate to strong coupling defects. From these data an initial working hypothesis for a “Ping-Pong” mechanism can be envisioned. In the ATP state the linker is bound to the hydrophobic depression and allows further contacts between the neighboring segments of the substratebinding domain and the lower back side of the ATPase domain. These contacts lead to an open state of the substrate-binding cavity. On substrate binding, the linker is displaced from the hydrophobic depression and associates with the region near the channel of the ATPase domain. The additional binding of DnaJ in the channel (see Section III.B.2.b) then synergistically stimulates ATP hydrolysis. Subsequent ADP dissociation and ATP binding reestablishes the starting conformation.

28

MATTHIAS P. MAYER ET AL.

This mechanism is supported by the NMR structure of the β-sheet domain of DnaK lacking all helices (Pellecchia et al., 2000). Peptide binding to this domain induces changes in the chemical shift that were largest in β-strand 3, including F426 of the substrate-binding cavity, down to the region around K414 and N415, which are in close proximity to the hydrophobic depression. Zuiderweg and co-workers also implicated the hydrophobic depression of the substrate-binding domain in the coupling mechanism (Wang et al., 1998). IV. THE TARGETING ACTIVITY OF CO-CHAPERONES The collective of co-chaperones acts mainly to target Hsp70 partner proteins to speciﬁc substrates or to mediate the targeted release of Hsp70-bound cargo. This regulatory activity provides a basis for the functional diversity of Hsp70 chaperone systems in the cell. It relies on the ability of most co-chaperones (except for GrpE proteins) to connect Hsp70 proteins with their target proteins, including substrates and other chaperones, through physical association with both the chaperone and the targets. We describe here the targeting strategies of the known cochaperones. A. DnaJ Proteins The activity of DnaJ proteins in mediating the ATP hydrolysisdependent locking of substrates into the binding cavity of Hsp70 proteins is essential for almost all chaperone activities of Hsp70 proteins. The only known exception is the activity of the mitochondrial Hsp70 homolog of yeast, Ssc1, in protein import. Ssc1 is targeted to translocating proteins through the Tim44 component of the translocation pore (Neupert, 1997; Pfanner et al., 1997; see the chapter by Ryan and Pfanner in this volume for detail). Cells usually encode a broad spectrum of DnaJ proteins which differ in their domain composition and cellular functions. Six DnaJ homologs exist in E. coli, 20 in S. cerevisiae, 33 in C. elegans, 34 in D. melanogaster, and presumably 44 in human cells (Venter et al., 2001). The number of Hsp70 homologs in these organisms is smaller than the number of DnaJ homologs, e.g., three in E. coli and around ten in the other organisms. Some Hsp70 homologs cooperate with several DnaJ homologs in vitro (Ungewickell et al., 1995; Terada et al., 1997; Lu and Cyr, 1998b; Terada and Mori, 2000) and in vivo (Ueguchi et al., 1995; Kelley and Georgopoulos, 1997a; Yan and Craig, 1999; Genevaux et al., 2001), while others are restricted to cooperation with only one particular DnaJ homolog (Silberg et al., 1998). At a ﬁrst approximation, the ability of a Hsp70 homolog to cooperate with several distinct DnaJs indicates that it plays multiple roles in cellular processes.

Hsp70 CHAPERONE MACHINES

29

The functional relationship between the diverse Hsp70 and DnaJ homologs is complex and not well understood. It is known, however, that the J domain is an important determinant for the speciﬁcity of the Hsp70–DnaJ interactions. The J domain is subject to sequence variations within the DnaJ family, e.g., within helices II and III (Hennessy et al., 2000), which mediate the contacts to Hsp70 proteins. Some of the residues identiﬁed in the E. coli J domain to be involved in the physical interaction with the ATPase domain of DnaK are not conserved (Greene et al., 1998; Hennessy et al., 2000), which renders them candidate residues for providing speciﬁcity to the interaction between DnaJ and Hsp70 proteins. In a different study, a swap of the J domains between the yeast homologs Sis1 and Sec63 leads to a nonfunctional Sec63 protein that could be converted into a functional protein by only three amino acid exchanges (Schlenstedt et al., 1995). The residues that were mutated in Sec63 are not homologous to the DnaK-contacting residues identiﬁed in the J domain of E. coli DnaJ, suggesting that different surface patches can be involved in J-domain–Hsp70 interactions. While these domain swap experiments show that some DnaJ proteins exhibit high speciﬁcity for particular Hsp70 partner proteins, in other cases the activity of DnaJ proteins in Hsp70-dependent folding reactions can be replaced by heterologous DnaJ proteins and chimeric DnaJ proteins with swapped J domains (Campbell et al., 1997; Kelley and Georgopoulos, 1997b; Yan and Craig, 1999). At present we can only speculate that the sequence variations within the J domains embody a speciﬁcity code for the Hsp70–DnaJ interactions that is sophisticated enough to establish functional crosstalk between Hsp70 systems in some cases and prevent undesired crosstalk between Hsp70 systems in other cases. The members of the DnaJ family have been subdivided according to their domain composition (Cheetham and Caplan, 1998; Kelley, 1998) (see Fig. 4a). An attempt has been made to introduce a consistent and unifying nomenclature of the mammalian DnaJ proteins that takes the domain structure into account (Ohtsuka and Hata, 2000), and in the following we will add the newly assigned names accordingly. Class I proteins share all four domains characteristic for E. coli DnaJ. DnaJ is composed of the N-terminal J domain, followed by a glycine- and phenylalanine-rich region (G/F region), a Zn-binding domain, and a C-terminal domain. Class II proteins have, in addition to the J domain, only two of the other domains and usually lack the Zn-binding domain. Class III proteins share with DnaJ only the J domain. While the J domain is always at the N terminus of the mature protein in the ﬁrst two classes, it may be at any place within the sequence of class III proteins. The class III family is the most diverse and contains proteins with additional distinct motifs or domains, e.g., transmembrane helices (E. coli DjlA, yeast Sec63, human DjC9/hSec63, yeast Mdj2), tetratricopeptide

30

MATTHIAS P. MAYER ET AL.

repeats (TPR; mouse DjC2/Zrf1/Mida1, human DjC3/hp58, and DjC7/hTpr2), and cysteine-rich regions which are polypalmitoylated (cysteine string proteins). Some of the members of classes I and III (e.g., yeast Ydj1, mammalian DjA1/Hdj2, mammalian DjC11/Mdg11) contain a C-terminal CaaX-box (C, cysteine; a, aliphatic amino acid; X, any amino acid) that is modiﬁed by a farnesyl group in vivo and serves for speciﬁc membrane localization. One member of class III, the kinesin light chain, even has two J domains (Tsai et al., 2000). The question of how DnaJ proteins interact with substrates and mediate their transfer onto Hsp70 partner proteins is not answered for any of the three classes of DnaJ proteins. Some DnaJ homologs have broad substrate speciﬁcity, such as E. coli DnaJ and yeast Ydj1, while others have more restricted substrate spectra. In particular the DnaJ proteins of class III may either bind a restricted number of substrates, such as the clathrin-speciﬁc auxilin or the kinesin light chain, or they may not bind substrates themselves but rather are positioned in close proximity to substrates. The latter seems to be the case for Dj1A in the plasma membrane of E. coli (Clarke et al., 1997; Kelley and Georgopoulos, 1997a), Sec63 at the translocation pore in the ER (Corsi and Schekman, 1996; Rapoport et al., 1996), and cysteine string proteins on the surface of neurosecretory vesicles (Buchner and Bundersen, 1997). With the exception of the G/F region, which may serve as a linker between domains and the J domain, most domains of class I and II proteins have been implicated in the interaction with substrates. Fragments comprising both the Zn-binding domain and the C-terminal domain of E. coli DnaJ are able to bind non-native proteins and prevent their aggregation but cannot assist their refolding by DnaK (Szabo et al., 1996). Both domains are thought to contribute to substrate binding since the deletion of either domain affects the binding of at least a subset of substrates (Banecki et al., 1996; Szabo et al., 1996; Lu and Cyr, 1998a). For the class II proteins which lack the Zn-binding domain, it has to be the C-terminal domain which alone is responsible for substrate binding. Structural information is available for the J domains of E. coli DnaJ, human DjB1/Hdj-1, and polyomavirus T-antigen (Szyperski et al., 1994; Pellecchia et al., 1996; Qian et al., 1996; Berjanskii et al., 2000; see Section III.B.2.b), the Zn-binding domain of E. coli DnaJ (residues 121– 209; Martinez-Yamout et al., 2000), and the C-terminal domain of yeast Sis1 (residues 171–352) which has homology to the C-terminal domain 1Ohtsuka and Hata assigned the name DjB9 to rat Mdg1/mouse mDj7, which classiﬁes it as a class II DnaJ proteins. However, this protein contains a predicted transmembrane domain N-terminal of the J-domain, no classical G/F domain, and no other homology to DnaJ but the J domain; thus, we feel that it belongs to class III and assigned it the new name DjC11.

Hsp70 CHAPERONE MACHINES

31

of DnaJ (Sha et al., 2000). All structures are unique to DnaJ proteins and have not been found in other structures listed in databases. The Zn-binding domain consists of a segment of 76 amino acids characterized by four repeats of the motif C-X-X-C-X-G-X-G (Bardwell et al., 1986). NMR spectroscopy revealed a V-shaped extended hairpin topology consisting of three pairs of antiparallel β-strands separated by the two Zn-binding sites (Fig. 4b) (Martinez-Yamout et al., 2000). Given this topology, the N and C termini are on the same side of the structure and the ﬁrst and the last C-X-X-C-X-G-X-G motif of the sequence form the ﬁrst Zn-binding site while the two middle C-X-X-C-X-G-X-G motifs form the second Zn-binding site. Removal of the Zn2+ ions causes the unfolding of the structure, demonstrating the importance of the Zn ions for the structural stability of the domain. The isolated domain precipitated in the presence of hydrophobic peptides and potential substrate-binding elements could therefore not be elucidated. The functional role for the Zn-binding domain remains obscure. Most structural information on how a DnaJ protein may bind substrates comes from the 2.7-A˚ crystal structure of the C-terminal substratebinding domain (residues 171–352) of the class II homolog Sis1 of S. cerevisiae (Sha et al., 2000). It consists of two highly similar domains each formed by a sandwich of two β-sheets and a short α-helix (Fig. 4b). The C terminus of the second domain extends into a short α-helix which is involved in the dimerization of Sis1 in the crystals. Although Sis1 and other DnaJ homologs form dimers and higher-form oligomers in solution, it is not clear whether a dimer is the active form. The structure of the Sis1 fragment did not reveal an obvious substrate-binding cavity. However, the authors propose a small hydrophobic depression in domain I as peptide binding site for two reasons. First, the analogous hydrophobic depression in domain II is occupied by the aromatic side chain of a phenylalanine residue. Second, in the crystal packing the side chain of a proline residue from an adjacent Sis1 molecule was inserted into this hydrophobic depression (Sha et al., 2000). An alternative proposal for a substrate-binding site of Sis1 is the C-terminal short α-helix, which in the crystal structure mediates the dimerization of Sis1. This helix has properties that qualify it for such a function, e.g., its exposure and hydrophobicity (Rudiger ¨ et al., 2001). B. Bag Proteins Bag proteins differ from GrpE proteins by virtue of their ability to associate with ligands other than Hsp70 proteins. Although not shown experimentally, it is generally assumed that Bag proteins use this ability

32

MATTHIAS P. MAYER ET AL.

to connect the nucleotide dissociation–dependent delivery of Hsp70bound cargo with cellular processes. These cellular processes seem to be highly diverse since Bag homologs have been implicated by genetic two-hybrid screens (Takayama et al., 1995) and coimmunoprecipitation (Zeiner and Gehring, 1995; Wang et al., 1996) in signal transduction processes related to cellular stress responses, including apoptosis, proliferation, and development. Bag proteins have been found in complexes with the antiapoptotic Bcl-2, the protein kinase Raf, the transcription factor c-Jun, the receptors for vitamin D, androgen and glucocorticoids (Takayama et al., 1995; Zeiner and Gehring, 1995; Bardelli et al., 1996; Wang et al., 1996; Clevenger et al., 1997; Zeiner et al., 1997; Schneikert et al., 1999), and the proteasome (L¨uders et al., 2000a). Whatever the functions of Bag proteins are in the above cellular processes, it is clear from the work of several laboratories that Bag proteins are not strictly essential for the chaperone activity of Hsp70 proteins in folding of non-native proteins (Takayama et al., 1997; Zeiner et al., 1997; Bimston et al., 1998; L¨uders et al., 1998; Takayama et al., 1999; Nollen et al., 2000) (G¨assler et al., 2001). This is in sharp contrast to the essential role of GrpE in the chaperone cycle of DnaK. On the contrary, Bag proteins and, in particular, Bag-1M (see below) have been proposed to act as negative regulators of the Hsp70 and Hsc70 homologs in a more general sense. This proposal was ﬁrst based on studies revealing that Bag-1 at a variety of concentrations inhibits the chaperone activity of Hsp70 and Hsc70 in the folding of denatured luciferase and β-galactosidase in vitro and in vivo (Takayama et al., 1997; Zeiner et al., 1997; Bimston et al., 1998; Nollen et al., 2000). A new mechanistic twist was given to this proposal by native gel experiments indicating that Bag-1 inhibits the functional cycle of Hsc70 by uncoupling substrate binding from ATP binding (Bimston et al., 1998). However, in a recent kinetic study no such uncoupling activity was identiﬁed for Bag-1 (G¨assler et al., 2001). Furthermore, this latter study showed that at physiological concentrations of inorganic phosphate (which slows down the nucleotide dissociation from Hsc70 and Hsp70 5- to 10-fold), Bag-1 at low concentrations stimulates slightly (up to 1.3-fold) the luciferase folding activity of Hsp70 or Hsc70 with DjB1/Hdj-1 or DjA1/Hdj-2 in vitro. Since the cellular concentration of Bag-1 is rather low (3% of the Hsc70 concentration; Kanelakis et al., 1999), it might be that the inhibitory effects of Bag-1, observed in vitro in the absence of Pi and in vivo when Bag-1 is overproduced, are restricted to special environmental and metabolic conditions of the cell. An attractive hypothesis is that in the cell, most Bag proteins are associated with their protein partners and exert their regulatory role only locally. The family of Bag proteins is composed of the homologs Bag-1, Bag-2, Bag-3 (CAIR/BIS), Bag-4 (SODD), Bag-5, and Bag-6 (BAT3/Scythe)

Hsp70 CHAPERONE MACHINES

33

FIG. 5. Domain structure of Hsp70 co-chaperones. Individual domains/modules are represented by differently shaded boxes. The known structural features and functions of domains are indicated. The following abbreviations for the different domains were used: NLS, nuclear localization sequence; TRSEEX, Thr-Arg-Ser-Glu-Glu-Xaa repeat motif; Ub, ubiquitin-like domain; Bag, Bag homology region; WW, Trp-Trp domain; TPR, tetratricopeptide repeat; GGMP, Gly-Gly-Met-Pro repeat motif; +/−, charged region; U box, U box motif of E4 ubiquitin ligases; DnaK, interaction site for DnaK; 70, interaction site for Hsp70; 90, interaction site for Hsp90.

(Takayama et al., 1999; Thress et al., 2001), most of which have been identiﬁed by two-hybrid screens and genome mining of human cDNA libraries (Takayama et al., 1999; Thress et al., 2001). All homologs by deﬁnition contain a Bag domain, but otherwise show a high degree of difference in their domain composition; e.g., four of these homologs do not code for any of the other domains found in Bag-1 (Fig. 5). Bag-1

34

MATTHIAS P. MAYER ET AL.

is produced in the cell as four translation initiation variants, Bag-1L (50 kDa), Bag-1M (46 kDa), Bag-1p33 (33 kDa), and Bag-1S (29 kDa) (Yang et al., 1998) (Fig. 5). Bag-1M is identical to Rap46, a protein found in the cytosol of higher eukaryotes (Takayama et al., 1995; Zeiner and Gehring, 1995; H¨ohfeld and Jentsch, 1997). The N-terminal segment, which exists only in Bag-1L, is rich in basic amino acids that have been implicated in DNA binding. Bag-1L is indeed found mainly in the nucleus (Takayama et al., 1998). The second region, present in Bag-1L and Bag-1M, contains a stretch of six TRSEEX repeats (threonine, arginine, serine, glutamic acid, glutamic acid, any amino acid; TRS is not in all repeats identical). It has been suggested that these repeats stimulate transcriptional processes in an unspeciﬁc fashion in analogy to an acidic activation domain (Zeiner et al., 1999). However, there is also evidence that these TRSEEX repeats are involved in down regulation of the transcriptional activity of the glucocorticoid receptor (Crocoll et al., 2000). Common to all four Bag-1 variants is a ubiquitin-like domain (H¨ohfeld and Jentsch, 1997; Brehmer et al., 2001; Sondermann et al., 2001; G¨assler et al., 2001) that was proposed to mediate the interaction of Bag-1 with the proteasome. Such interaction has been demonstrated by coimmunoprecipitation experiments (L¨uders et al., 2000a). It is tempting to speculate that the association of Bag-1M with the proteasome allows it to dissociate substrates from a complex with Hsp70 for degradation by the proteasome. Such a role would directly link the Hsp70 activity to proteolysis. Alternatively, Bag-1 may act as an antitargeting factor that prevents Hsp70 from binding to substrates of the proteasome since this may interfere with the degradation process. C. Hip and Hop Hip has been identiﬁed in a yeast two-hybrid screen as protein of 43 kDa which interacts with the ATPase domain of human Hsc70 (H¨ohfeld et al., 1995) (Fig. 5). Analytical ultracentrifugation showed that Hip forms an elongated dimer (Velten et al., 2000) which associates with two Hsc70 molecules. Hip stabilizes the ADP state of Hsp70 and thereby prevents the release of Hsp70-bound substrates (H¨ohfeld et al., 1995; H¨ohfeld and Jentsch, 1997). It has been claimed that this stabilizing activity is responsible for the observed ability of Hip to stimulate the chaperone activity of Hsp70 in luciferase folding in vitro 2- to 3-fold (H¨ohfeld et al., 1995; L¨uders et al., 1998). However, it is unclear of how this may work, considering that Hip was used in a 75- to 250-fold excess (Hip dimer) over luciferase. There are also reports that Hip does not affect the Hsc70- and DjB-1/Hdj-1-mediated refolding of denatured luciferase (Gebauer et al., 1997).

Hsp70 CHAPERONE MACHINES

35

The 60-kDa protein Hop (yeast Sti1) has been identiﬁed in a genetic screen for proteins involved in the regulation of the heat shock response in yeast (Nicolet and Craig, 1989) and was subsequently found as a component of the progesterone receptor complex (Smith et al., 1993) (Fig. 5). Hop was shown to stimulate the ATPase and chaperone activity of Hsp70 in vitro. This stimulation was proposed to result from Hop’s acceleration of ADP release and ATP binding (Gross and Hessefort, 1996) and thereby accelerated substrate release. However, neither the ATP association rate nor the ADP and substrate dissociation rates were determined in the presence of Hop, and it is therefore not yet clear how Hop enhances Hsp70 chaperone function. Both Hip and Hop are found in complexes of Hsp70, Hsp90, and the steroid hormone receptor and are proposed to stabilize the interaction between Hsp70 and Hsp90 (Prapapanich et al., 1996). This proposal was strongly supported by structural data. Hip and Hop contain several TPR domains of different lengths (Frydman and H¨ohfeld, 1997). TPR domains bind to the EEVD motif present at the C termini of most eukaryotic cytosolic Hsp70s and Hsp90s. The recent elucidation of the crystal structures of the two TPR domains of Hop, each in complex with a peptide containing the EEVD motifs of Hsp70 and Hsp90, respectively, and biochemical data show that these motifs allow the speciﬁc binding to the TPR domains (Scheuﬂer et al., 2000). The remaining sequences of Hip and Hop lack homology to other proteins and their functional signiﬁcance remains obscure. It has been hypothesized that Hip and Hop stabilize the chaperone– steroid hormone receptor complex by inﬂuencing the ATPase cycles of Hsp70 and Hsp90 and/or by acting as scaffolding proteins that assemble Hsp70 and Hsp90 in an ordered complex on the receptor. Alterantively, it has been proposed based on in vitro data on complex assembly and stability (Smith, 1998) that Hsp70 with its DnaJ co-chaperone interacts ﬁrst with the receptor, whereupon Hop links the incoming Hsp90 to Hsp70 and promotes the transfer of the receptor to Hsp90 through its action as nucleotide exchange factor. In vitro reconstitution experiments showed that Hop enhances the maturation rate of the receptor (Morishima et al., 2000a), which is consistent with any of the models. However, whatever the exact roles of Hip and Hop are, neither is essential for the formation of the complex in vitro (Rajapandi et al., 2000). D. Chip The 35-kDa protein Chip is the most recently discovered co-chaperone of eukaryotic Hsp70s (Fig. 5). Chip inhibits the in vitro chaperone activity of Hsc70 and Hsp70 in preventing the aggregation of denatured

36

MATTHIAS P. MAYER ET AL.

rhodanese and refolding of denatured luciferase. Chip inhibits the ATPase activity of Hsc70 without inﬂuencing ADP dissociation (Ballinger et al., 1999). How Chip acts on the ATPase activity and whether this inhibitory activity is responsible for the effects of Chip on the chaperone activity of Hsc70 and Hsp70 are unclear. Chip contains TPR domains similar to Hop and the U-box motif of E4-ubiquitin ligases, suggesting a function in the ubiquitin tagging of proteins for degradation by the proteasome (Ballinger et al., 1999). Both in vivo and in vitro, increasing concentrations of Chip induce the ubiquitination of two Hsc70 substrates investigated, the glucocorticoid receptor and the cystic ﬁbrosis transmembrane conductance regulator (CFTR), and promoted their degradation by the proteasome (Connell et al., 2001; Meacham et al., 2001). Chip therefore shares with Bag-1 the potential to connect the activity of Hsp70 chaperones with proteolysis. E. HspBP1 Another recently identiﬁed co-chaperone is Hsp70-binding protein 1 (HspBP1) (Raynes and Guerriero, 1998) that binds to the ATPase domain of Hsp70 (Fig. 5). HspBP1 was shown to modulate the ATPase cycle of Hsp70 by moderately inhibiting ATP binding and thereby decreasing the DjB1/Hdj-1-stimulated ATPase activity of Hsp70 2-fold. This modulation of the ATPase cycle is proposed to be responsible for the observed strong inhibition of the chaperone activity in vitro as measured by refolding denatured luciferase. There are no sequence similarities with any known proteins and neither structural nor biochemical properties have been described. V. OUTLOOK The Hsp70 protein family and their co-chaperones constitute a complex network of folding machines that is utilized by cells in many ways. Despite considerable progress in the elucidation of the mechanistic basis of these folding machines, important aspects remain to be solved. With respect to the Hsp70 proteins, it is still unclear whether their activity to assist protein folding relies on the ability to induce conformational changes in the bound substrates, how the coupling mechanism allows ATP to control substrate binding, and to what extent the sequence variations within the family, such as the Hsp110 and Hsp170 proteins, translate into variations of the mechanism. With respect to the action of cochaperones, we lack a molecular understanding of the coupling function of DnaJ proteins and of how co-chaperones target their Hsp70 partner

Hsp70 CHAPERONE MACHINES

37

proteins to substrates. Furthermore, it can be expected that more cellular processes will be discovered that depend on the chaperone activity of Hsp70 chaperones. REFERENCES Alfano, C., and McMacken, R. (1989). J. Biol. Chem. 264, 10709–10718. Ang, D., Chandrasekhar, G. N., Zylicz, M., and Georgopoulos, C. (1986). J. Bacteriol. 167, 25–29. Ayling, A., and Baneyx, F. (1996). Protein Sci. 5, 478–487. Ballinger, C. A., Connell, P., Wu, Y., Hu, Z., Thompson, L. J., Yin, L.-Y., and Patterson, C. (1999). Mol. Cell Biol. 19, 4535–4545. Banecki, B., Liberek, K., Wall, D., Wawrzyn´ow, A., Georgopoulos, C., Bertoli, E., Tanfani, F., and Zylicz, M. (1996). J. Biol. Chem. 271, 14840–14848. Banecki, B., and Zylicz, M. (1996). J. Biol. Chem. 271, 6137–6143. Bardelli, A., Longati, P., Alberto, D., Goruppi, S., Schneider, C., Ponzetto, C., and Comoglio, P. M. (1996). EMBO J. 15, 6205–6212. Bardwell, J. C., Tilly, K., Craig, E., King, J., Zylicz, M., and Georgopoulos, C. (1986). J. Biol. Chem. 261, 1782–1785. Barouch, W., Prasad, K., Greene, L., and Eisenberg, E. (1997). Biochemistry 36, 4303–4308. Bender, A. T., Silverstein, A. M., Demady, D. R., Kanelakis, K. C., Noguchi, S., Pratt, W. B., and Osawa, Y. (1999). J. Biol. Chem. 274, 1472–1478. Berjanskii, M. V., Riley, M. I., Xie, A., Semenchenko, V., Folk, W. R., and Van Doren, S. R. (2000). J. Biol. Chem. 275, 36094–36103. Bertelsen, E. B., Zhou, H., Lowry, D. F., Flynn, G. C., and Dahlquist, F. W. (1999). Protein Sci. 8, 343–354. Bimston, D., Song, J., Winchester, D., Takayama, S., Reed, J. C., and Morimoto, R. I. (1998). EMBO J. 17, 6871–6878. Blond-Elguindi, S., Cwirla, S. E., Dower, W. J., Lipshutz, R. J., Sprang, S. R., Sambrook, J. F., and Gething, M.-J. H. (1993). Cell 75, 717–728. Brehmer, D., Rudiger, ¨ S., G¨assler, C. S., Klostermeier, D., Packschies, L., Reinstein, J., Mayer, M. P., and Bukau, B. (2001). Nat. Struct. Biol. 8, 427–432. Briknarova, K., Takayama, S., Brive, L., et al. (2001). Nat. Struct. Biol. 8, 349–352. Buchberger, A., G¨assler, C. S., B¨uttner, M., McMacken, R., and Bukau, B. (1999). J. Biol. Chem. 274, 38017–38026. Buchberger, A., Schr¨oder, H., B¨uttner, M., Valencia, A., and Bukau, B. (1994). Nat. Struct. Biol. 1, 95–101. Buchberger, A., Schr¨oder, H., Hesterkamp, T., Sch¨onfeld, H.-J., and Bukau, B. (1996). J. Mol. Biol. 261, 328–333. Buchberger, A., Theyssen, H., Schr¨oder, H., McCarty, J. S., Virgallita, G., Milkereit, P., Reinstein, J., and Bukau, B. (1995). J. Biol. Chem. 270, 16903–16910. Buchner, E., and Bundersen, C. B. (1997). Trends Neurosci. 20, 223–227. Buchner, J. (1996). FASEB J. 10, 10–19. Buchner, J. (1999). Trends Biochem. Sci. 24, 136–141. Bukau, B., and Horwich, A. L. (1998). Cell 92, 351–366. Campbell, K. S., Mullane, K. P., Aksoy, I. A., et al. (1997). Genes & Development 11, 1098– 1110. Cheetham, M. I. E., and Caplan, A. J. (1998). Cell Stress Chap. 3, 28–36. Cherﬁls, J., and Chardin, P. (1999). Trends Biochem. Sci. 24, 306–311.

38

MATTHIAS P. MAYER ET AL.

Clarke, D. J., Holland, I. B., and Jacq, A. (1997). Mol. Microbiol. 25, 933–944. Clevenger, C. V., Thickman, K., Ngo, W., Chang, W. P., Takayama, S., and Reed, J. C. (1997). Mol. Endocrinol. 11, 608–618. Connell, P., Ballinger, C. A., Jiang, J., Wu, Y., Thompson, L. J., H¨ohfeld, J., and Patterson, C. (2001). Nat. Cell Biol. 3, 93–96. Corsi, A. K., and Schekman, R. (1996). J. Biol. Chem. 271, 30299–30302. Crocoll, A., Schneikert, J., Hubner, S., Martin, E., and Cato, A. C. (2000). Kidney Int. 57, 1265–1269. Deuerling, E., Schulze-Specking, A., Tomoyasu, T., Mogk, A., and Bukau, B. (1999). Nature 400, 693–696. Diamant, S., Peres Ben-Zvi, A., Bukau, B., and Goloubinoff, P. (2000). J. Biol. Chem. 275, 21107–21113. Dittmar, K. D., Banach, M., Galigniana, M. D., and Pratt, W. B. (1998). J. Biol. Chem. 273, 7358–7366. Dittmar, K. D., and Pratt, W. B. (1997). J. Biol. Chem. 272, 13047–13054. Donze, O., and Picard, D. (1999). Mol. Cell Biol. 19, 8422–8432. Easton, D. P., Kaneko, Y., and Subjeck, J. R. (2000). Cell Stress & Chaperones 5, 276–290. Ehrnsperger, M., Gr¨aber, S., Gaestel, M., and Buchner, J. (1997). EMBO J. 16, 221–229. Feifel, B., Sch¨onfeld, H.-J., and Christen, P. (1998). J. Biol. Chem. 273, 11999–12002. Feldheim, D., Rothblatt, J., and Schekman, R. (1992). Mol. Cell Biol. 12, 3288–3296. Ferrell, K., Wilkinson, C. R. M., Dubiel, W., and Gordon, C. (2000). Trends Biochem. Sci. 25, 83–88. Flaherty, K. M., Deluca-Flaherty, C., and McKay, D. B. (1990). Nature 346, 623–628. Fletterick, R. J., Bates, D. J., and Steitz, T. A. (1975). Proc. Natl. Acad. Sci. USA 72, 38–42. Flynn, G. C., Chappell, T. G., and Rothman, J. E. (1989). Science 245, 385–390. Flynn, G. C., Pohl, J., Flocco, M. T., and Rothman, J. E. (1991). Nature 353, 726–730. Franciolini, F., and Petris, A. (1988). Experientia 44, 183–188. Freeman, B. C., and Morimoto, R. I. (1996). EMBO J. 15, 2969–2979. Freeman, B. C., Myers, M. P., Schumacher, R., and Morimoto, R. I. (1995). EMBO J. 14, 2281–2292. Frydman, J., and H¨ohfeld, J. (1997). Trends Biochem. Sci. 22, 87–92. Frydman, J., Nimmesgern, E., Erdjument-Bromage, H., Wall, J. S., Tempst, P., and Hartl, F.-U. (1992). EMBO J. 11, 4767–4778. Frydman, J., Nimmesgern, E., Ohtsuka, K., and Hartl, F. U. (1994). Nature 370, 111–117. Gamer, J., Multhaup, G., Tomoyasu, T., McCarty, J. S., Rudiger, ¨ S., Sch¨onfeld, H.-J., Schirra, C., Bujard, H., and Bukau, B. (1996). EMBO J. 15, 607–617. Gao, B., Greene, L., and Eisenberg, E. (1994). Biochemistry 33, 2048–2054. Gao, B., Yumiko, E., Greene, L., and Eisenberg, E. (1993). J. Biol. Chem. 268, 8507–8513. Garc´ıa-Gardena, G., Fan, R., Shah, V., Sorrentino, R., Cirino, G., Papapetropoulos, A., and Sessa, W. C. (1998). Nature 392, 821–824. G¨assler, C. S., Buchberger, A., Laufen, T., Mayer, M. P., Schr¨oder, H., Valencia, A., and Bukau, B. (1998). Proc. Natl. Acad. Sci. USA 95, 15229–15234. G¨assler, C. S., Wiederkehr, T., Brehmer, D., Bukau, B., and Mayer, M. P. (2001). J. Biol. Chem. 276, 32538–32544. Gautschi, M., Lilie, H., F¨unfschilling, U., Mun, A., Ross, S., Lithgow, T., Rucknagel, P., and Rospert, S. (2001). Proc. Natl. Acad. Sci. USA 98, 3762–3767. Gebauer, M., Zeiner, M., and Gehring, U. (1997). FEBS Lett. 417, 109–113. Genevaux, P., Wawrzynow, A., Zylicz, M., Georgopoulos, C., and Kelley, W. L. (2001). J. Biol. Chem. 276, 7906–7912.

Hsp70 CHAPERONE MACHINES

39

Gisler, S. M., Pierpaoli, E. V., and Christen, P. (1998). J. Mol. Biol. 279, 833–840. Glover, J. R., and Lindquist, S. (1998). Cell 94, 73–82. Goloubinoff, P., Mogk, A., Peres Ben Zvi, A., Tomoyasu, T., and Bukau, B. (1999). Proc. Natl. Acad. Sci. USA 96, 13732–13737. Gragerov, A., Nudler, E., Komissarova, N., Gaitanaris, G., Gottesman, M., and Nikiforov, V. (1992). Proc. Natl. Acad. Sci. USA 89, 10341–10344. Gragerov, A., Zeng, L., Zhao, X., Burkholder, W., and Gottesman, M. E. (1994). J. Mol. Biol. 235, 848–854. Greene, L. E., and Eisenberg, E. (1990). J. Biol. Chem. 265, 6682–6687. Greene, M. K., Maskos, K., and Landry, S. J. (1998). Proc. Natl. Acad. Sci. USA 95, 6108– 6113. Groemping, Y., Klostermeier, D., Herrmann, C., Veit, T., Seidel, R., and Reinstein, J. (2001). J. Mol. Biol. 305, 1173–1183. Gross, M., and Hessefort, S. (1996). J. Biol. Chem. 271, 16833–16841. Guex, N., and Peitsch, M. C. (1997). Electrophoresis 18, 2714–2723. Ha, J.-H., Hellman, U., Johnson, E. R., Li, L., McKay, D. B., Sousa, M. C., Takeda, S., Wernstedt, C., and Wilbanks, S. M. (1997). J. Biol. Chem. 272, 27796–27803. Ha, J.-H., Johnson, E. R., McKay, D. B., Sousa, M. C., Takeda, S., and Wilbanks, S. M. (1999). In Molecular Chaperones and Folding Catalysts. Regulation, Cellular Function and Mechanism (B. Bukau, ed.), pp. 573–607. Harwood Academic Publishers, Amsterdam. Ha, J.-H., and McKay, D. B. (1994). Biochemistry 33, 14625–14635. Ha, J.-H., and McKay, D. B. (1995). Biochemistry 34, 11635–11644. Hansen, W. J., Lingappa, V. R., and Welch, W. (1994). J. Biol. Chem. 169, 26610–26613. Harrison, C. J., Hayer-Hartl, M., Di Liberto, M., Hartl, F.-U., and Kuriyan, J. (1997). Science 276, 431–435. Hartl, F. U. (1996). Nature 381, 571–580. Hartson, S. D., Irwin, A. D., Shao, J., Scroggins, B. T., Volk, L., Huang, W., and Matts, R. L. (2000). Biochemistry 39, 7631–7644. Hendrick, J. P., Langer, T., Davis, T. A., Hartl, F. U., and Wiedmann, M. (1993). Proc. Natl. Acad. Sci. USA 90, 10216–10220. Hennessy, F., Cheetham, M. E., Dirr, H. W., and Blatch, G. L. (2000). Cell Stress & Chaperones 5, 347–358. Hoff, K. G., Silberg, J. J., and Vickery, L. E. (2000). Proc. Natl. Acad. Sci. USA 97, 7790–7795. H¨ohfeld, J., and Jentsch, S. (1997). EMBO J. 16, 6209–6216. H¨ohfeld, J., Minami, Y., and Hartl, F. U. (1995). Cell 83, 589–598. Hunter, T., and Poon, R. Y. C. (1997). Trends Cell Biol. 7, 157–161. Hurley, J. H., Faber, H. R., Worthylake, D., Meadow, N. D., Roseman, S., Pettigrew, D. W., and Remington, S. J. (1993). Science 259, 673–677. Hutchison, K. A., Dittmar, K. D., Czar, M. J., and Pratt, W. B. (1994a). J. Biol. Chem. 269, 5043–5049. Hutchison, K. A., Dittmar, K. D., and Pratt, W. B. (1994b). J. Biol. Chem. 269, 27894–27899. Johnson, J. L., and Craig, E. A. (2000). Mol. Cell Biol. 20, 3027–3036. Jordan, R., and McMacken, R. (1995). J. Biol. Chem. 270, 4563–4569. Kabani, M., Beckerich, J.-M., and Gaillardin, C. (2000). Mol. Cell Biol. 20, 6923–6934. Kabsch, W., Mannherz, H. G., Suck, D., Pai, E. F., and Holmes, K. C. (1990). Nature 347, 37–44. Kanelakis, K. C., Morishima, Y., Dittmar, K. D., Galigniana, M. d., Takayama, S., Reed, J. C., and Pratt, W. B. (1999). J. Biol. Chem. 274, 34134–34140. Karzai, A. W., and McMacken, R. (1996). J. Biol. Chem. 271, 11236–11246.

40

MATTHIAS P. MAYER ET AL.

Kelley, W. L. (1998). Trends Biochem. Sci. 23, 222–227. Kelley, W. L., and Georgopoulos, C. (1997a). Mol. Microbiol. 25, 913–931. Kelley, W. L., and Georgopoulos, C. (1997b). Proc. Natl. Acad. Sci. USA 94, 3679–3684. Kim, S., Schilke, B., Craig, E. A., and Horwich, A. L. (1998). Proc. Natl. Acad. Sci. USA 95, 12860–12865. Klostermeier, D., Seidel, R., and Reinstein, J. (1998). J. Mol. Biol. 279, 841–853. Klostermeier, D., Seidel, R., and Reinstein, J. (1999). J. Mol. Biol. 287, 511–525. Knoblauch, N. T. M., Rudiger, ¨ S., Sch¨onfeld, H.-J., Driessen, A. J. M., Schneider-Mergener, J., and Bukau, B. (1999). J. Biol. Chem. 274, 34219–34225. Krzewska, J., Langer, T., and Liberek, K. (2001). FEBS Lett. 489, 92–96. Landry, S. J., Jordan, R., McMacken, R., and Gierasch, L. M. (1992). Nature 355, 455–457. Langer, T., Lu, C., Echols, H., Flanagan, J., Hayer, M. K., and Hartl, F. U. (1992). Nature 356, 683–689. Laufen, T., Mayer, M. P., Beisel, C., Klostermeier, D., Reinstein, J., and Bukau, B. (1999). Proc. Natl. Acad. Sci. USA 96, 5452–5457. Liberek, K., Georgopoulos, C., and Zylicz, M. (1988). Proc. Natl. Acad. Sci. USA 85, 6632– 6636. Liberek, K., Marszalek, J., Ang, D., Georgopoulos, C., and Zylicz, M. (1991a). Proc. Natl. Acad. Sci. USA 88, 2874–2878. Liberek, K., Skowyra, D., Zylicz, M., Johnson, C., and Georgopoulos, C. (1991b). J. Biol. Chem. 266, 14491–14496. Liberek, K., Wall, D., and Georgopoulos, C. (1995). Proc. Natl. Acad. Sci. USA 92, 6224– 6228. Lopez-Buesa, P., Pfund, C., and Craig, E. A. (1998). Proc. Natl. Acad. Sci. USA 95, 15253– 15258. Lu, Z., and Cyr, D. M. (1998a). J. Biol. Chem. 273, 5970–5978. Lu, Z., and Cyr, D. M. (1998b). J. Biol. Chem. 273, 27824–27830. Lu¨ ders, J., Demand, J., and H¨ohfeld, J. (2000a). J. Biol. Chem. 275, 4613–4617. Lu¨ ders, J., Demand, J., Papp, O., and H¨ohfeld, J. (2000b). J. Biol. Chem. 275, 14817–14823. Lu¨ ders, J., Demand, J., Sch¨onfelder, S., Frien, M., Zimmermann, R., and H¨ohfeld, J. (1998). Biol. Chem. 379, 1217–1226. Lutz, T., Westermann, B., Neupert, W., and Herrmann, J. (2001). J. Mol. Biol. 307, 815–825. Mally, A., and Witt, S. N. (2001). Nat. Struct. Biol. 8, 254–257. Martinez-Yamout, M., Legge, G. B., Zhang, O., Wright, P. E., and Dyson, H. J. (2000). J. Mol. Biol. 300, 805–818. Matlack, K. E. S., Misselwitz, B., Plath, K., and Rapoport, T. A. (1999). Cell 97, 553–564. Mayer, M. P., Laufen, T., Paal, K., McCarty, J. S., and Bukau, B. (1999). J. Mol. Biol. 289, 1131–1144. Mayer, M. P., Schr¨oder, H., Rudiger, ¨ S., Paal, K., Laufen, T., and Bukau, B. (2000). Nat. Struct. Biol. 7, 586–593. McCarty, J., and Walker, G. (1994). J. Bacteriol. 176, 764–780. McCarty, J. S., Buchberger, A., Reinstein, J., and Bukau, B. (1995). J. Mol. Biol. 249, 126– 137. Meacham, G. C., Patterson, C., Zhang, W., Younger, J. M., and Cyr, D. M. (2001). Nat. Cell Biol. 3, 100–105. Michels, A. A., Kanon, B., Konings, A. W. T., Ohtsuka, K., Bensaude, O., and Kampinga, H. H. (1997). J. Biol. Chem. 272, 33283–33289. Misselwitz, B., Staeck, O., Matlack, K. E. S., and Rapoport, T. A. (1999). J. Biol. Chem. 274, 20110–20115. Misselwitz, B., Staeck, O., and Rapoport, T. A. (1998). Mol. Cell 2, 593–603.

Hsp70 CHAPERONE MACHINES

41

Mogk, A., Tomoyasu, T., Goloubinoff, P., Rudiger, ¨ S., R¨oder, D., Langen, H., and Bukau, B. (1999). EMBO J. 18, 6934–6949. Montgomery, D. L., Morimoto, R. I., and Gierasch, L. M. (1999). J. Mol. Biol. 286, 915–932. Morimoto, R., Tissieres, A., and Georgopoulos, C. (1994). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Morishima, Y., Kanelakis, K. C., Murphy, P. J., Shewach, D. S., and Pratt, W. B. (2001). Biochem. J. 40, 1109–1116. Morishima, Y., Kanelakis, K. C., Silverstein, A. M., Dittmar, K. D., Estrada, L., and Pratt, W. B. (2000a). J. Biol. Chem. 275, 6894–6900. Morishima, Y., Murphy, P. J. M., Li, D.-P., Sanchez, E. R., and Pratt, W. B. (2000b). J. Biol. Chem. 275, 18054–18060. Morshauser, R. C., Hu, W., Wang, H., Pang, Y., Flynn, G. C., and Zuiderweg, E. R. P. (1999). J. Mol. Biol. 289, 1387–1403. Neidhardt, F. E. (1996). Eschrichia coli and Salmonella ASM Press, Washington, D.C. Nelson, R. J., Ziegelhoffer, T., Nicolet, C., Werner-Washburne, M., and Craig, E. A. (1992). Cell 71, 97–105. Neupert, W. (1997). Annu. Rev. Biochem. 66, 863–917. Nguyen, V. T., Morange, M., and Bensaude, O. (1989). J. Biol. Chem. 264, 10487–10492. Nicholls, A., Bharadwaj, R., and Honig, B. (1993). Biophys. J. 64, A166–A166. Nicolet, C., and Craig, E. (1989). Mol. Cell Biol. 9, 3638–3646. Nollen, E. A. A., Brunsting, J. F., Song, J., Kampinga, H. H., and Morimoto, R. I. (2000). Mol. Cell Biol. 20, 1083–1088. O’Brien, M. C., and McKay, D. B. (1993). J. Biol. Chem. 268, 24323–24329. O’Keeffe, B., Fong, Y., Chen, D., Zhou, S., and Zhou, Q. (2000). J. Biol. Chem. 275, 279–287. Ohtsuka, K., and Hata, M. (2000). Cell Stress & Chaperones 5, 98–112. Packschies, L., Theyssen, H., Buchberger, A., Bukau, B., Goody, R. S., and Reinstein, J. (1997). Biochemistry 36, 3417–3422. Palleros, D. R., Reid, K. L., Shi, L., Welch, W. J., and Fink, A. L. (1993). Nature 365, 664–666. Peitsch, M. C. (1995). Bio/Technology 13, 658–660. Peitsch, M. C. (1996). Biochem. Soc. Trans. 24, 274–279. Pellecchia, M., Montgomery, D. L., Stevens, S. Y., Vander Kooi, C. W., Feng, H., Gierasch, L. M., and Zuiderweg, E. R. P. (2000). Nat. Struct. Biol. 7, 298–303. Pellecchia, M., Szyperski, T., Wall, D., Georgopoulos, C., and Wuthrich, ¨ K. (1996). J. Mol. Biol. 260, 236–250. Perdew, G. H., and Whitelaw, M. L. (1991). J. Biol. Chem. 266, 6708–6713. Pfanner, N., Craig, E. A., and H¨onlinger, A. (1997). Annu. Rev. Cell Dev. Biol. 13, 25–51. Pfund, C., Lopez-Hoyo, N., Ziegelhoffer, T., Schilke, B. A., Lopez-Buesa, P., Walter, W. A., Wiedmann, M., and Craig, E. A. (1998). EMBO J. 17, 3981–3989. Picard, D., Khursheed, B., Garabedian, M. J., Fortin, M. G., Lindquist, S., and Yamamoto, K. R. (1990). Nature 348, 166–168. Pierce, D. W., Hom-Booher, N., and Vale, R. D. (1997). Nature 388, 338. Pierpaoli, E. V., Gisler, S. M., and Christen, P. (1998a). Biochemistry 37, 16741–16748. Pierpaoli, E. V., Sandmeier, E., Baici, A., Sch¨onfeld, H.-J., Gisler, S., and Christen, P. (1997). J. Mol. Biol. 269, 757–768. Pierpaoli, E. V., Sandmeier, E., Sch¨onfeld, H.-J., and Christen, P. (1998b). J. Biol. Chem. 273, 6643–6649. Prapapanich, V., Chen, S., Nair, S., Rimerman, R., and Smith, D. (1996). Mol. Endocrinol. 10, 420–431. Pratt, W. B., Silverstein, A. M., and Galigniana, M. D. (1999). Cell Signal 11, 839–851.

42

MATTHIAS P. MAYER ET AL.

Pratt, W. B., and Toft, D. O. (1997). Endocr. Rev. 18, 306–360. Qian, Y. Q., Patel, D., Hartl, F. U., and McColl, D. J. (1996). J. Mol. Biol. 260, 224–235. Rajapandi, T., Greene, L. E., and Eisenberg, E. (2000). J. Biol. Chem. 275, 22597–22604. Rapoport, T. A., Jungnickel, B., and Kutay, U. (1996). Annu. Rev. Biochem. 65, 271–303. Raynes, D. A., and Guerriero, Jr., V. (1998). J. Biol. Chem. 273, 32883–32888. Ru¨ diger, S., Buchberger, A., and Bukau, B. (1997a). Nat. Struct. Biol. 4, 342–349. Ru¨ diger, S., Germeroth, L., Schneider-Mergener, J., and Bukau, B. (1997b). EMBO J. 16, 1501–1507. R¨udiger, S., Mayer, M. P., Schneider-Mergener, J., and Bukau, B. (2000). J. Mol. Biol. 304, 245–251. R¨udiger, S., Schneider-Mergener, J., and Bukau, B. (2001). EMBO J. 20, 1–9. Russell, R., Jordan, R., and McMacken, R. (1998). Biochemistry 37, 596–607. Russell, R., Wali Karzai, A., Mehl, A. F., and McMacken, R. (1999). Biochemistry 38, 4165– 4176. Sanchez, E. R., Faber, L. E., Henzel, W. J., and Pratt, W. B. (1990). Biochemistry 29, 5145– 5152. Scheuﬂer, C., Brinker, A., Bourenkov, G., Pegoraro, S., Moroder, L., Bartunik, H., Hartl, F.-U., and Moareﬁ, I. (2000). Cell 101, 199–210. Schlee, S., Groemping, Y., Herde, P., Seidel, R., and Reinstein, J. (2001). J. Mol. Biol. 306, 889–899. Schlenstedt, G., Harris, S., Risse, B., Lill, R., and Silver, P. A. (1995). J. Cell Biol. 129, 979–988. Schmid, D., Baici, A., Gehring, H., and Christen, P. (1994). Science 263, 971–973. Schmid, F. X. (1993). Annu. Rev. Biophys. Biomol. Struct. 22, 123–143. Schneikert, J., H¨ubner, S., Langer, G., Petri, T., J¨aa¨ ttel¨a, M., Reed, J., and Cato, A. C. B. (2000). EMBO J. 19, 6508–6516. Schneikert, J., Hubner, S., Martin, E., and Cato, A. C. (1999). J. Cell Biol. 146, 929–940. Sch¨onfeld, H.-J., Schmidt, D., Schr¨oder, H., and Bukau, B. (1995). J. Biol. Chem. 270, 2183–2189. Schr¨oder, H., Langer, T., Hartl, F.-U., and Bukau, B. (1993). EMBO J. 12, 4137–4144. Sha, B., Lee, S., and Cyr, D. M. (2000). Structure 15, 799–807. Shtilerman, M., Lorimer, G. H., and Englander, S. W. (1999). Science 284, 822–825. Silberg, J. J., Hoff, K. G., and Vickery, L. E. (1998). J. Bacteriol. 180, 6617–6624. Silberg, J. J., and Vickery, L. E. (2000). J. Biol. Chem. 275, 7779–7786. Skowyra, D., and Wickner, S. (1993). J. Biol. Chem. 268, 25296–25301. Slepenkov, S. V., and Witt, S. N. (1998a). Biochemistry 37, 1015–1024. Slepenkov, S. V., and Witt, S. N. (1998b). Biochemistry 37, 16749–16756. Smith, D. F. (1993). Mol. Endocrinol. 7, 1418–1429. Smith, D. F. (1998). Biol. Chem. 379, 283–288. Smith, D. F., Sullivan, W. P., Marion, T. N., Zaitsu, K., Madden, B., McCormick, D. J., and Toft, D. O. (1993). Mol. Cell Biol. 13, 869–876. Sondermann, H., Scheuﬂer, C., Scheider, C., H¨ohfeld, J., Hartl, F.-U., and Moareﬁ, I. (2001). Science 291, 1553–1557. Sousa, M. C., and McKay, D. B. (1998). Biochemistry 37, 15392–15399. Suh, W.-C., Burkholder, W. F., Lu, C. Z., Zhao, X., Gottesman, M. E., and Gross, C. A. (1998). Proc. Natl. Acad. Sci. USA 95, 15223–15228. Szabo, A., Korzun, R., Hartl, F. U., and Flanagan, J. (1996). EMBO J. 15, 408–417. Szabo, A., Langer, T., Schr¨oder, H., Flanagan, J., Bukau, B., and Hartl, F. U. (1994). Proc. Natl. Acad. Sci. USA 91, 10345–10349.

Hsp70 CHAPERONE MACHINES

43

Szyperski, T., Pellecchia, M., Wall, D., Georgopoulos, C., and Wuthrich, ¨ K. (1994). Proc. Natl. Acad. Sci. USA 91, 11343–11347. Takayama, S., Bimston, D. N., Matsuzawa, S.-i., Freeman, B. C., Aime-Sempe, C., Xie, Z., Morimoto, R. I., and Reed, J. C. (1997). EMBO J. 16, 4887–4896. Takayama, S., Krajewski, S., Krajewski, M., et al. (1998). Cancer Res. 58, 3116–3131. Takayama, S., Sato, T., Krajewski, S., Kochel, K., Irie, S., Millan, J. A., and Reed, J. C. (1995). Cell 80, 279–284. Takayama, S., Xie, Z., and Reed, J. C. (1999). J. Biol. Chem. 274, 781–786. Takeda, S., and McKay, D. B. (1996). Biochemistry 35, 4636–4644. Terada, K., Kanazawa, M., Bukau, B., and Mori, M. (1997). J. Cell Biol. 139, 1089–1095. Terada, K., and Mori, M. (2000). J. Biol. Chem. 275, 24728–24734. Teter, S. A., Houry, W. A., Ang, D., Tradler, T., Rockabrand, D., Fischer, G., Blum, P., Georgopoulos, C., and Hartl, F. U. (1999). Cell 97, 755–765. Theodorakis, N. G., Zand, D. J., Kotzbauer, P. T., Williams, G. T., and Morimoto, R. I. (1989). Mol. Cell Biol. 9, 3166–3173. Theyssen, H., Schuster, H.-P., Bukau, B., and Reinstein, J. (1996). J. Mol. Biol. 263, 657– 670. Thress, K., Song, J., Morimoto, R. I., and Kornbluth, S. (2001). EMBO J. 20, 1033–1041. Toft, D. O. (1999). In Molecular Chaperones and Folding Catalysts. Regulation, Cellular Function and Mechanism (B. Bukau, ed.), pp. 313–327. Harwood Academic Publishers, Amsterdam. Tokunaga, M., Kato, S., Kawamura-Watabe, A., Tanaka, R., and Tokunaga, H. (1998). Yeast 14, 1285–1295. Tomoyasu, T., Ogura, T., Tatsuta, T., and Bukau, B. (1998). Mol. Microbiol. 30, 567–581. Tsai, M.-Y., Morﬁni, G., Szebenyi, G., and Brady, S. T. (2000). Mol. Biol. Cell 11, 2161–2173. Ueguchi, C., Shiozawa, T., Kakeda, M., Yamada, H., and Mizuno, T. (1995). J. Bacteriol. 177, 3894–3896. Uma, S., Hartson, S., Chen, J.-J., and Matts, R. L. (1997). J. Biol. Chem. 272, 11648–11656. Uma, S., Thulasiraman, V., and Matts, R. L. (1999). Mol. Cell Biol. 19, 5861–5871. Ungewickell, E. (1985). EMBO J. 4, 3385–3391. Ungewickell, E., Ungewickell, H., Holstein, S. E., Lindner, R., Prasad, K., Barouch, W., Martin, B., Greene, L. E., and Eisenberg, E. (1995). Nature 378, 632–635. Ungewickell, E., Ungewickell, H., and Holstein, S. E. H. (1997). J. Biol. Chem. 272, 19594– 19600. Ursic, D., and Culbertson, M. R. (1991). Mol. Cell. Biol. 11, 2629–2640. Veinger, L., Diamant, S., Buchner, J., and Goloubinoff, P. (1998). J. Biol. Chem. 273, 11032–11037. Velten, M., Villoutreix, B. O., and Ladjimi, M. M. (2000). Biochemistry 39, 307–315. Venter, J. C., Adams, M. D., Myers, E. W., et al. (2001). Science 291, 1304–1351. Voisine, C., Cheng, Y., Ohlson, M., Schilke, B., Hoff, K., Beinert, H., Marszalek, J., and Craig, E. (2001). Proc. Natl. Acad. Sci. USA 98, 1483–1488. Wall, D., Zylicz, M., and Georgopoulos, C. (1994). J. Biol. Chem. 269, 5446–5451. Wang, H., Kurochkin, A. V., Pang, Y., Hu, W., Flynn, G. C., and Zuiderweg, E. R. P. (1998). Biochemistry 37, 7929–7940. Wang, H.-G., Takayam, S., Rapp, U. R., and Reed, J. C. (1996). Proc. Natl. Acad. Sci. USA 93, 7063–7068. Wawrzyn´ow, A., Banecki, B., Wall, D., Liberek, K., Georgopoulos, C., and Zylicz, M. (1995). J. Biol. Chem. 270, 19307–19311. Wawrzyn´ow, A., and Zylicz, M. (1995). J. Biol. Chem. 270, 19300–19306.

44

MATTHIAS P. MAYER ET AL.

Wickner, S., Hoskins, J., and McKenney, K. (1991a). Proc. Natl. Acad. Sci. USA 88, 7903– 7907. Wickner, S., Hoskins, J., and McKenny, K. (1991b). Nature 350, 165–167. Wickner, S., Skowyra, D., Hoskins, J., and McKenney, K. (1992). Proc. Natl. Acad. Sci. USA 89, 10345–10349. Wu, B., Ang, D., Snavely, M., and Georgopoulos, C. (1994). J. Bacteriol. 176, 6965–6973. Wyman, C., Vasilikiotis, C., Ang, D., Georgopoulos, C., and Echols, H. (1993). J. Biol. Chem. 268, 25192–25196. Xu, Y., and Lindquist, S. (1993). Proc. Natl. Acad. Sci. USA 90, 7074–7078. Xu, Y., Singer, M. A., and Lindquist, S. (1999). Proc. Natl. Acad. Sci. USA 96, 109–114. Yaffe, M. B., Farr, G. W., Miklos, D., Horwich, A. L., Sternlicht, M. L., and Sternlicht, H. (1992). Nature 358, 245–248. Yan, W., and Craig, E. A. (1999). Mol. Cell. Biol. 19, 7751–7758. Yang, X., Chernenko, G., Hao, Y., Ding, Z., Pater, M. M., Pater, A., and Tang, S. C. (1998). Oncogene 17, 981–989. Yasuda, K., Nakai, A., Hatayama, T., and Nagata, K. (1995). J. Biol. Chem. 270, 29718–29723. Zeiner, M., Gebauer, M., and Gehring, U. (1997). EMBO J. 16, 5483–5490. Zeiner, M., and Gehring, U. (1995). Proc. Natl. Acad. Sci. USA 92, 11465–11469. Zeiner, M., Niyaz, Y., and Gehring, U. (1999). Proc. Natl. Acad. Sci. USA 96, 10194–10199. Zhu, X., Zhao, X., Burkholder, W. F., Gragerov, A., Ogata, C. M., Gottesman, M., and Hendrickson, W. A. (1996). Science 272, 1606–1614. Ziemienowicz, A., Skowyra, D., Zeilstra-Ryalls, J., Fayet, O., Georgopoulos, C., and Zylicz, M. (1993). J. Biol. Chem. 268, 25425–25431. Ziemienowicz, A., Zylicz, M., Floth, C., and H¨ubscher, U. (1995). J. Biol. Chem. 270, 15479– 15484. Zylicz, M., Ang, D., and Georgopoulos, C. (1987). J. Biol. Chem. 262, 17437–17442. Zylicz, M., Ang, D., Liberek, K., and Georgopoulos, C. (1989). EMBO J. 8, 1601–1608.

ALLOSTERY AND PROTEIN SUBSTRATE CONFORMATIONAL CHANGE DURING GroEL/GroES-MEDIATED PROTEIN FOLDING By HELEN R. SAIBIL,* ARTHUR L. HORWICH,†, ‡ and WAYNE A. FENTON† *Department of Crystallography, Birkbeck College London, Malet Street, London, WC1E 7HX, UK; and †Department of Genetics and ‡Howard Hughes Medical Institute, Yale School of Medicine, 295 Congress Ave., New Haven, CT 06510

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Structure of GroEL and Its Functional Complexes . . . . . . . . . . . . . . . . . . . . . . . . A. Structural Changes on Nucleotide and GroES Binding . . . . . . . . . . . . . . . . B. The GroEL Folding Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Allostery Drives the Alternating Actions of the Two Rings . . . . . . . . . . . . . . D. ATP Binding Initiates GroES Binding and Signaling between the Rings . . E. Peptide Binding to the Apical Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Polypeptide Folding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. What Binds to GroEL? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Changes in Polypeptide Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. The Role of the Folding Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

45 46 48 50 52 53 54 56 56 64 69 70 71

I. INTRODUCTION Chaperonins are large double-ring complexes that assist the folding to native form of a wide variety of proteins in the cytosol of all cells and in mitochondria and chloroplasts of eukaryotes. In addition to playing an essential role in folding newly translated and newly imported proteins, they serve a vital role in the recovery of proteins after heat shock or other stresses. Under normal conditions, they may also play a role in maintaining native conformations, returning to the native state proteins that have become unfolded during the course of their action. The Escherichia coli chaperonin system, comprising GroEL and its co-chaperonin GroES, has been extensively studied as the paradigm for chaperonin-mediated, ATP-dependent protein folding. Cell biological, biochemical, biophysical, and structural studies have contributed to a growing understanding of the detailed mechanics and dynamics of this protein-folding machine (for other reviews, see Fenton and Horwich, 1997; Sigler et al., 1998; Grantcharova et al., 2001). Two facets of the chaperonin reaction that have been less well understood are the allosteric effects observed during GroEL’s functional cycle and the detailed fate of non-native substrate 45 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0096-5332/02 $35.00

46

HELEN R. SAIBIL ET AL.

polypeptide during binding and folding. Recent results are shedding light on these problems and are discussed in this chapter. II. STRUCTURE OF GroEL AND ITS FUNCTIONAL COMPLEXES GroEL is a 58-kDa protein that assembles into a cylindrical 800-kDa oligomer formed of two heptameric rings stacked back-to-back (Sigler et al., 1998; Ranson et al., 1998). The structure of the GroEL 14-mer and the fold of a subunit are shown in Figure 1a, with the secondary structure and domain organization indicated on the protein sequence (Fig. 1b). Three domains have been recognized in each subunit, designated equatorial, intermediate, and apical. The assembly is mainly held together by the equatorial domain, which is formed of the N- and C-terminal portions of the chain. This mainly α-helical domain contains

FIG. 1. GroEL structure. (a) One subunit is outlined on a surface-rendered view of a cryo-EM map of the GroEL oligomer (left), and enlarged in the atomic structure representation on the right. The two inter-ring contacts formed by each subunit are numbered 1 and 2 (left) and c1 and c2, with the charged residues in each shown in space-ﬁlling format (right). An ATP molecule is shown in its binding site in the equatorial domain, which also contains the N and C termini (N, C). The ﬂexible hinge positions between domains are labeled h1 and h2. The white residues on the inner surface of the apical domain are hydrophobic residues involved in substrate binding. (Based on Fig. 1 of Roseman et al., 1996.) (b) The sequence of E. coli GroEL is shown with the secondary structure elements coded by domain: equatorial, black, residues 1–133 and 409–548; intermediate, light gray, residues 134–190 and 377–408; apical, dark gray, residues 191–376.

FIG. 1. (continued )

48

HELEN R. SAIBIL ET AL.

the nucleotide-binding site and forms most of the intra-ring contacts via β-strand contacts near the termini. Twenty-four residues at the C terminus are not seen in the crystal structure owing to disorder. Although there thus appears to be a channel through the GroEL oligomer, these disordered residues have been visualized by small-angle neutron scattering and form a barrier that separates the two rings (Thiyagarajan et al., 1996), so that each ring contains a large cavity open at the apical aspect. The equatorial domains also form the interface between the rings, with charged residues lying at the ends of two of the α-helices forming the two salt-bridge contacts between a given subunit of one ring and two adjacent subunits in the other ring. The apical domains have an αβ fold, and exposed hydrophobic residues on the surface facing the cavity form the substrate and GroES binding sites. Connecting the major equatorial and apical domains at exposed hinge points is a smaller intermediate domain, formed of two α-helices coiled around each other, capped by a small β-sheet. A. Structural Changes on Nucleotide and GroES Binding The two ﬂexible hinge regions in each subunit, together with the loosely packed, porous nature of the assembly, allow the oligomer to take up a remarkably wide repertoire of conformations in its ATPase and GroES binding cycle. The range of conformations in the functional cycle has been described at low resolution by cryoelectron microscopy (cryoEM) (Chen et al., 1994; Roseman et al., 1996; White et al., 1997), shown in the series of cryo-EM structures in Figure 2. In unliganded (apo) GroEL (Fig. 2a), the structure is relatively compact, and the substrate binding sites (Fig. 1a) line the central cavity. Nucleotide binding brings about hinge rotations at the ends of the intermediate domains, such that the ADP- and ATP-bound states (Fig. 2b and c, respectively) are elongated, and the apical domains are twisted, with more asymmetry in the ATP form. This twisting makes the apical substrate binding sites somewhat less accessible. GroES binds only to a nucleotide-bound ring, and this binding brings about a dramatic further change in conformation in GroES-bound rings (Fig. 2d and e), where the GroEL apical domains rise and twist further to form an enclosed chamber capped by GroES. The crystal structure of the bullet-shaped GroEL–GroES–ADP7 complex (Fig. 3) provided a major advance in understanding the mechanism of assisted protein folding. In the GroES-bound ring, the apical domains are twisted 90◦ relative to those in the open ring. This large twist radically changes the character of the surface lining the cavity. Whereas the open ring is lined with exposed hydrophobic residues that bind

GroEL/GroES-MEDIATED PROTEIN FOLDING

49

FIG. 2. Cryo-EM maps of GroEL complexes with different nucleotides and GroES: (a) apo GroEL, (b) GroEL–ADP, (c) GroEL–ATP, (d) GroEL-GroES-ADP, (e) GroEL– GroES–ATP. The nucleotide concentration in these samples was 5 mM, and the maps are at 30 A˚ resolution. Nucleotide and GroES binding result in concerted, large rotations about the two hinge regions in each subunit. Reproduced from Roseman et al. (1996) by permission of Cell Press.

non-native proteins, once GroES is bound, these sites are buried in the wall of the chamber by the large twist. The implication is that unfolded or misfolded proteins with exposed hydrophobic surface, initially bound to the hydrophobic sites on an open GroEL ring, are evicted from the binding sites but, at the same time, are trapped inside an encapsulated, now hydrophilic chamber. The GroEL–GroES–ADP7 crystal structure also provided important insights into the GroEL ATPase mechanism. The opening-out motion of GroEL subunits on binding GroES involves a downward rotation of the intermediate domain. This motion moves intermediate domain helix M over the nucleotide-binding pocket in the equatorial domain, closing the site and inserting Asp-398 to coordinate the bound Mg2+, in addition to Asp-87. Asp-398 plays an important role in catalyzing ATP hydrolysis, perhaps by activating a bound water molecule (Ditzel et al., 1998), and its mutation to Ala reduces the hydrolytic rate more than 60-fold (Rye et al., 1997).

50

HELEN R. SAIBIL ET AL.

FIG. 3. The crystal structure of GroEL-GroES-ADP. In this view based on the structure of Xu and colleagues (1997), the front half of the complex is not shown, so that the internal cavities are exposed. GroES (dark gray) is bound at the top. The hydrophobic binding residues are shown in light-colored, space-ﬁlling format. In the lower ring, they are facing the open cavity, but in the GroES-bound ring, they are either buried in the subunit interface or occluded by GroES. The enclosed chamber is lined by hydrophilic residues. The extended loops of GroES that make the contacts are mobile and disordered in free GroES.

B. The GroEL Folding Cycle Along with these structural descriptions of intermediates in the chaperonin cycle, biochemical studies have provided a mechanistic view of the pathway, illustrated in Figure 4 (Fenton and Horwich, 1997; Sigler et al., 1998; Rye et al., 1999). The asymmetric GroEL–GroES complex (ﬁrst panel) is likely to be the predominant physiological state, given the observed ratio of GroES to GroEL (∼2 : 1) and normal nucleotide concentrations in the cell. When ATP binds to the open GroEL ring, several things happen in succession. The conformation of the open ring

GroEL/GroES-MEDIATED PROTEIN FOLDING

51

FIG. 4. Scheme of the functional cycle. The physiologic acceptor state for non-native polypeptide binding is likely the GroEL–GroES–ADP complex (ﬁrst panel). Binding of ATP and GroES displaces the polypeptide into the now hydrophilic cis folding cavity, capped by GroES. At the same time, the GroES and ADP in the opposite ring are discharged (second panel). In the rate-limiting step (bold arrow), ATP is hydrolyzed, weakening the GroEL–GroES interaction, and folding proceeds (third panel). The open ring is now available for ligand binding, and GroES is primed for release. In the ﬁnal panel, new ATP and GroES binding to the lower ring encapsulates the newly bound polypeptide to form a new cis folding complex. Simultaneously, the previous complex is discharged, and the polypeptide is released, either folded to a native or near-native state or still in a non-native state that can rebind to the chaperonin.

changes, displacing the hydrophobic binding sites and thus weakening the binding of polypeptide substrate, but initiating the binding of GroES to that ring. GroES binding drives a further major conformational change in what is now the encapsulated cis folding cavity, releasing the polypeptide from the apical binding domains into the cavity to begin folding (second panel). At the same time, conformational changes transmitted to the opposite ring discharge its bound GroES. ATP hydrolysis in the cis ring (the slow step in the overall reaction) reduces the afﬁnity of that ring for GroES, priming it for release, while the substrate protein continues to fold (third panel). The binding of another non-native polypeptide and ATP to the open trans ring discharges the cis complex (fourth panel), releasing the substrate into solution either folded to the native state (or a state committed to it) or still in a nonnative conformation that can rebind to the chaperonin to repeat the process. Simultaneously, another GroES has bound to form a new cis folding-active complex on what had been the trans ring. Thus, the rings

52

HELEN R. SAIBIL ET AL.

of GroEL alternate as folding-active chambers, consuming one set of seven ATP molecules per folding cycle. C. Allostery Drives the Alternating Actions of the Two Rings The double-ring structure provides the possibility for an intricate set of allosteric interactions. It was noted early in the biochemical analysis of GroEL that ATP binding and hydrolysis show positive cooperativity within a ring; this implies that the accompanying conformational changes are concerted. In addition, ATP binding in one ring lowers the afﬁnity of the second ring for ATP in a negatively cooperative mechanism. This behavior has been analyzed in terms of nested cooperativity, combining concerted and sequential models (Yifrach and Horovitz, 1994, 1995). At its simplest, this treatment describes the double-ring system of GroEL as being in equilibrium between three sequential states: TT, TR, and RR. The T state of a ring has a low afﬁnity for ATP, the R state has a high afﬁnity, only the R states hydrolyze ATP, and the change between these states within a ring is concerted. In reality, of course, the situation is more complex. Excluding K+ and Mg2+ ions, GroEL has three signiﬁcant ligands—GroES, non-native substrate polypeptide, and ATP—not just one (ATP), and it is clear that they interact with the chaperonin to modulate the allostery. For example, reduced denatured α-lactalbumin, which is not folded by GroEL, binds to the chaperonin and affects the allosteric transitions of ATP hydrolysis. These data have been interpreted to mean that the T state has a high afﬁnity for unfolded protein relative to that of the R state, in contrast to their relative afﬁnities for ATP (Yifrach and Horovitz, 1996). Adding GroES to the system further complicates the analysis. Interestingly, when several variant GroELs with different levels of allostery were tested in protein refolding assays, the folding of mouse dihydrofolate reductase, a nonstringent substrate requiring only ATP for refolding, was signiﬁcantly affected by the differences in both intra- and inter-ring cooperativity. Slower folding was observed both for mutants with increased positive intra-ring cooperativity and for those with increased negative inter-ring cooperativity. On the other hand, the refolding of MDH, a stringent substrate requiring both ATP and GroES, was not affected at all (Yifrach and Horovitz, 2000). A recent kinetic analysis (Cliff et al., 1999) using tryptophan-containing variants of GroEL has identiﬁed several additional states of the system, which have been interpreted to correspond to those in the polypeptide binding–GroES encapsulation–polypeptide release/folding process outlined in the ﬁrst two panels of the mechanism in Figure 4. Further structural studies have begun to illuminate some of these states.

GroEL/GroES-MEDIATED PROTEIN FOLDING

53

D. ATP Binding Initiates GroES Binding and Signaling between the Rings Biochemical and kinetic studies, including changes in ﬂuorescent reporters (e.g., Jackson et al., 1993; Gibbons et al., 1996) and ATPase cooperativity (Cliff et al., 1999; Yifrach and Horovitz, 1994, 1995), indicate that there is a major conformational change induced by ATP. While this change has been observed in solution by cryo-EM (cf. Fig. 2), crystal structures have not revealed the ATP-bound conformation of GroEL despite the presence of ATPγ S in the binding pocket of all subunits in one structure (Boisvert et al., 1996). It appears that conditions in the crystal prevent or reverse the conformational change. Recently, cryo-EM analysis of an ATP-bound state has been extended to 10 A˚ resolution (Ranson et al., submitted). At this resolution, the domain atomic structures can be docked into the EM density, and the hinge rotations and intersubunit ˚ ATP binding to one interactions can be analyzed to an accuracy of 2–3 A. ring induces a complex series of domain movements in both rings, resulting in an extended, asymmetric structure, as shown in cartoon form in Figure 5. The primary event appears to be a downward rotation of

FIG. 5. Diagrams of the domain rotations induced by ATP binding, based on the structure of D398A GroEL–ATP (Ranson et al., submitted). (a) Three subunits are shown to indicate the switch made by the intermediate domain (triangle in upper subunits). The salt bridge (asterisk) made by the intermediate domain to the adjacent apical domain (light gray) in apo GroEL is replaced by one to the equatorial domain (black) in GroEL– ATP. The ring interface is distorted by tilts of the equatorial domains in ATP. (b) End view from outside the oligomer. ATP binding causes a large anticlockwise twist of the apical domains. The change in salt bridge is also indicated (asterisk).

54

HELEN R. SAIBIL ET AL.

the intermediate domain (triangle in Fig. 5a). This movement breaks the salt bridge between E386 on the intermediate domain and R197 on the adjacent apical domain (asterisk, Fig. 5a) and replaces it with a new contact on the adjacent equatorial domain (upper ring, Fig. 5a). (A similar downward movement of the intermediate domain was observed in the ADP bullet crystal structure, but the rotation angle was larger, so that E386 made a new contact farther down the equatorial domain helix.) It is proposed that the intermediate domain movement is propagated in a concerted manner around the ring, so that all seven sites bind ATP cooperatively. The apical domains, released from the constraint of the R197 salt bridge, rotate upward and twist 25◦ in the plane of the ring. Surprisingly, this twist is in the opposite direction (counterclockwise as viewed from outside the oligomer) to that present in the ADP bullet complex. Therefore, the apical domains, along with the substrate binding sites, must go through an even larger excursion than previously thought in order to reach the GroES-bound position, 90◦ clockwise from their orientation in apo GroEL. Such large excursions during the critical steps when a GroEL ring containing bound substrate is exposed to ATP and GroES may be important in the mechanism of encapsulation and release. It has been proposed that domain twisting and separation causes forced unfolding of the bound substrate (Shtilerman et al., 1999), but this remains equivocal (see below). In addition to the movements within the ATP-bound ring, the other ring adopts a different conformation, as negative cooperativity ensures that the oligomer becomes (or remains) asymmetric. This information must be transmitted across the inter-ring interface, and the GroEL–ATP structure reveals that this interface is distorted by tilts in the equatorial domains (Fig. 5a). Of the two inter-ring contacts, one (c1 in Fig. 1) is shortened and the other lengthened; overall, the interface is widened (Roseman et al., 2001). This signaling is essential for the release of GroES and thus the opening of the chamber and release of polypeptide, which is triggered by ATP binding to the trans ring after ATP hydrolysis in the cis ring (Fig. 4, last panel). E. Peptide Binding to the Apical Domain A cryo-EM study of the GroEL–MDH complex showed substrate density caught between the apical domains, although it was rotationally averaged by the imposition of 7-fold symmetry in the three-dimensional reconstruction (Chen et al., 1994). The mutagenesis study of Fenton and colleagues (1994) made it clear that the set of hydrophobic residues labeled in Figure 1, comprising the H and I helices and an underlying

GroEL/GroES-MEDIATED PROTEIN FOLDING

55

extended segment, is essential for substrate binding. The ﬁrst detailed view of the interaction of the binding site with non-native peptides came about serendipitously, in a crystal form of an isolated apical domain in which an N-terminal tag was found lodged on the binding surface of a neighboring molecule in the crystal (Buckle et al., 1997). Seven residues were bound in an extended conformation, mainly by hydrophobic interactions with the H and I helices, but also by some polar ones. More recently, a strongly binding peptide has been selected by phage display, and crystals containing it bound to an isolated apical domain and to GroEL have been analyzed (Chen and Sigler, 1999). The peptide formed a β-hairpin that interacted with the same H and I helices through both hydrophobic interactions and hydrogen bonds. Remarkably, the binding peptides so far examined bind to the GroEL apical domain in a very similar manner to that of the GroES mobile loop. In Figure 6, the major part of the apical domain is shown with three binding peptides: the GroES mobile loop (Xu et al., 1997), the N-terminal tag peptide

FIG. 6. Mode of peptide binding to the GroEL apical domain. Residues 197–335 of the apical domain are shown as a ribbon diagram, with three binding peptides (black) overlaid. The hydrophobic binding residues identiﬁed by mutagenesis (Fenton et al., 1994) are shown in light-colored, space-ﬁlling format. Also shown are side chains of the binding peptides that interact with the apical domain.

56

HELEN R. SAIBIL ET AL.

(Buckle et al., 1997), and the strongly binding peptide (Chen and Sigler, 1999). They all bind as an extended chain or β-hairpin to a groove between two helices, reminiscent of the MHC class I antigen-binding site. All these peptides are unstructured in solution, but adopt the extended conformation when bound to the apical domain. These structures provide some rationale for the lack of speciﬁcity in polypeptide binding by GroEL. The binding peptides and, to some extent, the apical domain binding site itself all exhibit conformational ﬂexibility, suggesting that the binding groove can accept a wide variety of polypeptide sequences. The strongly binding peptide (Chen and Sigler, 1999) makes the most extensive contacts to this groove and thus has the strongest binding afﬁnity of the peptides examined. On the other hand, the mutagenesis study of Fenton and colleagues (1994) also identiﬁed a lower extended segment of hydrophobic residues as essential for substrate binding (Fig. 6). Either the peptides so far investigated do not exhibit the full repertoire of binding or, conceivably, the lower sites play a role at some crucial intermediate stage in the reaction cycle. III. POLYPEPTIDE FOLDING Although a great deal has been learned about the nucleotide cycle of the GroEL/GroES system, its structural correlates, and the ways in which GroES and substrate protein binding and release are modulated, our understanding of the changes that occur in the structure of the substrate polypeptide itself during a folding reaction remains minimal. Current experiments are yielding new insights into several aspects of this problem, including the nature of polypeptide binding, the effects of binding on polypeptide structure, and changes in polypeptide structure during the course of the folding reaction. A. What Binds to GroEL? 1. Model Peptide Studies Polypeptide binding to GroEL has generally been described as an interaction between exposed hydrophobic groups or regions on nonnative proteins and the extensive hydrophobic surface of the inside of the apical domains of the GroEL cavity. Indirect biochemical data have supported the role of hydrophobicity in determining binding of polypeptides, and afﬁnity panning experiments with short peptide segments have identiﬁed hydrophobic ones as having the highest afﬁnity for GroEL. The recent structures of GroEL and apical domain “mini-chaperones”

GroEL/GroES-MEDIATED PROTEIN FOLDING

57

complexed with peptides, described in detail above, provide the ﬁrst direct observation of such interactions. Several biophysical studies have further examined the nature of polypeptide binding by using different sets of model peptides and determining the effects of directed changes in them on binding to GroEL. One noticeable aspect of the bound peptides or segments in the crystal structures is that each is in an extended β-conformation. Although not unexpected for the one peptide derived from the mobile loop of GroES, this ﬁnding does raise the question of whether GroEL favors binding extended β-strands rather than α-helices. Earlier work had suggested that α-helical peptides could be bound, but the data are limited to a few examples. For example, a 13-mer peptide derived from the helical N terminus of rhodanese, a GroEL substrate, formed an α-helix when bound to GroEL, as determined by NMR spectroscopy, even though it was unstructured in solution (Landry and Gierasch, 1991). A similar result has been obtained using an apical domain mini-chaperone to bind this peptide (Kobayashi et al., 1999). Using ﬂuorescence anisotropy, Brazil and colleagues (1997) have shown that a peptide that can be forced to assume a helical conformation by internal disulﬁde bonds binds to GroEL only in its oxidized, helical conformation. In a similar, more recent study, Preuss and colleagues (1999) examined sets of 14-mer peptides forced to assume (or not assume) an α-helical conformation in solution by virtue of an N-terminal “template.” The helical version bound more strongly than the nonhelical one, but only if the predicted helix was amphipathic. It was not determined in this case whether the peptides remained helical (or nonhelical) when bound to GroEL (but see below). The signiﬁcant effect of the template structure on the afﬁnities of the attached peptides is a further concern in interpreting these results. An NMR study by Wang and colleagues (1999) examined variants of the 13-mer peptide derived from the N terminus of rhodanese, mentioned above. Versions synthesized with all D-amino acids and with alternating D- and L-amino acids were compared with the native peptide. Although none was structured in aqueous solution, the all L- (native) and all D-peptides formed right- and left-handed helices, respectively, both in 20% triﬂuoroethanol solution and when bound to GroEL. Equivalent line broadenings indicated that these peptides bound with about equal afﬁnities. Interestingly, the mixed DL-peptide, which could not form a helix, bound to GroEL as well as its helical relatives, based on line broadening, but with a pattern of trNOEs that was unique and inconsistent with an α-helix. Moreover, the helical and nonhelical peptides appeared to compete for the same (or at least overlapping) binding sites on GroEL. Finally, Chatellier and colleagues (1999) carried out afﬁnity panning

58

HELEN R. SAIBIL ET AL.

with phages bearing variants of the 34-residue cellulose binding domain of cellobiohydrolase I with mutations in the seven nonsequential recognition residues, using an apical domain mini-chaperone. A number of peptides were recovered with signiﬁcant afﬁnity for the mini-chaperone. Although the native structure of the cellulose-binding domain is a ﬂat surface, the variant residues could be mapped onto either an extended β-strand, similar to the GroES mobile loop peptide, or an amphipathic α-helix. Which structural motif was represented in the bound peptide was not determined, however. Interestingly, a mini-chaperone bearing the mutation corresponding to Y203E in GroEL, which is in the extended segment implicated in polypeptide binding, bound the selected peptides with much less afﬁnity than did the wild-type mini-chaperone, conﬁrming the role of the extended segment in substrate interaction. Several reports have also examined the role of amphipathicity in peptide binding. Brazil and colleagues (1997) showed that a peptide with a hydrophobic face bound to GroEL, while one with a hydrophilic face did not. Preuss and colleagues (1999) and Wang and colleagues (1999) both arranged the residues in a test peptide so that it would be either amphipathic or not when in a helical conformation. In both cases, the amphipathic form had higher afﬁnity for GroEL, whether it was a helix in solution (Preuss et al., 1999) or became helical only after binding to GroEL (Wang et al., 1999). Wang and colleagues extended their study to two peptides with a tendency to form a β-strand conformation. Here, only the peptide with alternating hydrophobic and hydrophilic residues, presenting a hydrophobic face as a β-strand, bound signiﬁcantly to GroEL. NMR analysis of the peptide–GroEL complex showed trNOEs characteristic of an extended β-conformation for the peptide. Thus, the data from model peptides suggest that both α-helices and β-sheets, as well as other less well deﬁned secondary structures, can be bound by GroEL, with the ability to present a hydrophobic face in any secondary structural context playing the key role in determining afﬁnity. This result and the ﬁnding that afﬁnity panning does not identify particular high-afﬁnity amino acid motifs are consistent with the nature of the cavity face of the apical domains of GroEL, which provide a hydrophobic surface without obvious binding pockets or other selective structures. The limitations of model peptide studies preclude extrapolating these analyses too far, however. The size of the peptides used is one problem, as detailed above. In addition, these peptides seem to be too short to interact simultaneously with multiple GroEL apical domains, an important characteristic of the binding of non-native substrate proteins

GroEL/GroES-MEDIATED PROTEIN FOLDING

59

(see below). This both limits their binding strength and prevents them from reporting on cooperative effects of the binding reaction. The stoichiometry of binding is also poorly deﬁned for most of the peptides. Whereas only a single molecule of non-native substrate protein typically binds to a GroEL double ring, multiple peptide molecules appear to be binding in the few cases examined. For example, Preuss and colleagues (1999) estimated that there were as many as 75–80 binding sites on each GroEL molecule for the amphipathic helical peptide they tested, that is, ﬁve or six per subunit! This raises many questions about the signiﬁcance of the interactions detected, the validity of apparent peptide binding afﬁnities, and the interpretation of other biochemical parameters. Indeed, the relevance of the observed peptide binding to that of substrate proteins has only rarely been tested by asking whether it is modulated by ATP or ATP/GroES binding. When it has been, by Preuss and colleagues (1999) and Wang and colleagues (1999), for example, the results have been equivocal, suggesting a mixture of speciﬁc, relevant binding and nonspeciﬁc interactions. 2. Polypeptide Binding Although it would be desirable to extend studies of binding to GroEL to larger polypeptides and especially to known substrates for chaperonindependent folding, some of the most informative techniques, particularly structure determination by X-ray crystallography or NMR, have not yet been possible for GroEL–polypeptide complexes. Nevertheless, more indirect techniques, such as ﬂuorescence anisotropy, protease digestion, and hydrogen/deuterium exchange monitored by mass spectrometry and NMR, have been applied to polypeptides bound to GroEL or during a folding reaction in efforts to understand the nature of polypeptide binding. These techniques have been used to examine both smaller polypeptides whose folding pathways in the absence of GroEL are well established and larger ones that do not fold efﬁciently, if at all, without assistance by the chaperonin system. A number of early studies (reviewed in Fenton and Horwich, 1997; Coyle et al., 1998) established that a wide range of folding intermediates can interact with GroEL, from ones very early in the folding pathway to late ones which may be native-like. For any given polypeptide, binding could be exclusively by early or by late species, or apparently by an ensemble of states. The overall interpretation of these results has been that GroEL is promiscuous in binding non-native polypeptides, recognizing exposed hydrophobic surface but not clearly discriminating on the basis of the presence of, or potential for, speciﬁc secondary structure.

60

HELEN R. SAIBIL ET AL.

More recent studies have further addressed the means of polypeptide binding to GroEL. Investigations of the binding and folding of β-lactamase, not normally a GroEL substrate, have shown that GroEL can bind two distinct non-native forms of this protein. When a βlactamase variant lacking cysteine residues was denatured with guanidinium chloride and diluted into a solution with GroEL, most (90%) bound to the chaperonin (Gervasoni et al., 1998a). If refolding was allowed to proceed in the absence of GroEL for 30 seconds, only 20% of the β-lactamase bound, even though only 25% had refolded, based on activity. This suggests that only early folding intermediates, presumably collapsed, largely unstructured forms, are recognized by GroEL. On the other hand, tryptophan ﬂuorescence measurements indicated that some structure had formed rapidly after dilution, at least in the vicinity of this residue (Gervasoni and Pl¨uckthun, 1997). Despite this, deuterium-exchange experiments on the β-lactamase/GroEL complex (Gervasoni et al., 1998a) revealed no protected protons in the species formed by direct dilution into GroEL-containing buffer, whereas dilution in the absence of GroEL led to the production of species with highly protected protons within 30 seconds of folding. In contrast to these results, when fully folded variant β-lactamase was destabilized by incubation at 42◦ C, it became inactive and bound to GroEL, but now in a native-like conformation that protected 18 protons (Gervasoni et al., 1998b). Further investigation by partial proteolysis and MALDI mass spectrometry of the thermally destabilized species formed from wild-type β-lactamase and bound by GroEL suggested that its structure was only partially disrupted (i.e., near-native, although inactive) with loss of structure in the C-terminal helix and exposure of the hydrophobic surface of a ﬁve-stranded β-sheet. The protease protection pattern of the GroEL-bound species was interpreted to mean that the C-terminal helix and two loops connecting strands in the exposed β-sheet were the primary sites of interaction with GroEL. One point that is not clear from these studies is how the two bound species of β-lactamase are related to each other. Both appear to be productive of the native state when the complexes are incubated with ATP at room temperature. On the other hand, later intermediates from chemically denatured β-lactamase do not bind to GroEL (Gervasoni et al., 1998a), so it seems unlikely that the early intermediate form gives rise directly to a late species resembling the thermally denatured one. Conversely, extensive thermal denaturation does not generate a more unfolded species similar to the early intermediate. These observations suggest that the two types of denaturation populate different folding pathways and raise the possibility that folding de novo (e.g., after translation or translocation) may be different from

GroEL/GroES-MEDIATED PROTEIN FOLDING

61

refolding from non-native states produced by thermal or other stress, at least for certain proteins, and that some proteins might even depend on the chaperonin system for stress recovery, although they do not require it for initial folding. Other studies examining the range of non-native forms bound by GroEL followed the thermal denaturation and renaturation (at lower temperature) of citrate synthase in the presence and absence of GroEL (Grallert et al., 1998; Grallert and Buchner, 1999). Here, GroEL appeared to bind both a native-like, but inactive, dimeric species formed early in the thermal inactivation and a more slowly forming monomeric species, the latter with higher afﬁnity. Both GroEL-bound forms were productive of the native, active enzyme when incubated with GroES and ATP, and analysis of the kinetics and concentration dependence of activity regain suggested that the monomeric species gave rise to the inactive dimeric one during refolding. In an in vivo attempt to discern the nature of polypeptides that bind to GroEL, Houry and colleagues (1999) carried out a pulse-chase study on intact E. coli, isolating GroEL-bound polypeptides by coimmunoprecipitation with anti-GroEL antibodies and analyzing them by twodimensional gel electrophoresis. As expected from earlier experiments (Horwich et al., 1993; Ewalt et al., 1997), a limited number (250–300) of cytoplasmic polypeptides bound to GroEL. Most were larger than 20 kDa and smaller than 60 kDa, although a signiﬁcant fraction (∼20%) were larger than 60 kDa. During the in vivo chase period, a majority of the 20–60-kDa proteins were released completely from GroEL, although about a third remained bound to a varying extent. Many of the >60-kDa proteins also remained with GroEL during the chase period. Interestingly, the retained 20–60-kDa proteins were largely released by incubation of the lysates with GroES and ATP, while only a fraction of the >60-kDa ones were, suggesting that at least some of these were “deadend” species. To better understand what distinguished binding from nonbinding species, Houry and colleagues isolated a number of individual proteins following immunoprecipitation and electrophoresis and identiﬁed them by mass spectrometry. Of the 52 identiﬁed polypeptides, 24 had known or predicted (by homology) structures and 13 of these contained two or more αβ domains. The authors suggested that multiple αβ domains are the hallmark of GroEL substrates and that the hydrophobic interfaces between the α-helices and β-sheets forming these domains may provide, when disrupted, sites of favorable, high-afﬁnity interactions with the GroEL apical domains. They further proposed that such interactions might stabilize a nascent β-sheet by providing a ﬂexible helical partner (the GroEL apical binding sites), avoiding both intramolecular

62

HELEN R. SAIBIL ET AL.

domain swapping and intermolecular aggregation. It remains to be seen whether this analysis holds up as more putative GroEL substrates are identiﬁed and their structures become available. It also must be pointed out that none of the identiﬁed polypeptides has been shown to be a stringent GroEL substrate in vitro, i.e., one that is dependent on the full complement of GroEL, GroES, and ATP for refolding. 3. How Are Polypeptides Bound? One signiﬁcant difference between model peptides and substrate polypeptides of GroEL is their relative size. A consequence of this is the likelihood that a polypeptide interacts with more than one apical domain binding site at a time, increasing its afﬁnity and making allosteric effects possible. To establish whether substrate polypeptides indeed interact with multiple GroEL apical domains and whether there is a minimum number required for both stable binding and efﬁcient folding, Farr and colleagues (2000) developed a covalent version of GroEL in which the seven subunits of a ring were joined at the DNA level to become subdomains of a 400-kDa chaperonin ring. A mutation (V263S) previously shown to disrupt substrate binding was made in one or more individual apical domains in varying order around the ring, and the variant covalent GroELs were tested for function in vivo, for the ability to bind three well characterized substrate proteins (MDH, rhodanese, and Rubisco) and GroES, and for the ability to refold MDH. Four consecutive wild-type apical domains were required for function in vivo and for binding substrates in vitro in initial tests. Cleavage of the links between the subunits relaxed the in vitro requirement without changing the order of the subunits, such that three consecutive wild-type domains provided near-normal binding of both MDH and Rubisco (Fig. 7). In contrast, rhodanese required only two wild-type domains in any order for stable binding, while GroES bound relatively well when only a single wild-type subunit was present. MDH refolding on GroES and ATP addition followed the same pattern as binding, with three consecutive normal subunits sufﬁcient to give refolding kinetics and yield similar to those of wild-type GroEL. Several conclusions can be reached from this study. First, multiple GroEL apical domains indeed bind to substrate polypeptides, both in vivo and in vitro, and this ability appears to be crucial to the cellular function of GroEL. This is supported by the additional demonstration that non-native Rubisco bound to a GroEL containing an apical cysteine residue made disulﬁde crosslinks to multiple GroEL subunits. Second, different substrates have different requirements for the apical domains, stringent substrates such as MDH and Rubisco being the most

GroEL/GroES-MEDIATED PROTEIN FOLDING

63

FIG. 7. Binding of non-native Rubisco and MDH to GroEL complexes with different numbers and arrangements of mutant subunits. Chemically denatured Rubisco or MDH was diluted into buffer containing the indicated mutant complexes, the binary complexes were isolated by gel ﬁltration, and the amounts of bound substrate protein were quantitated. The data are presented relative to binding by wild-type, noncovalent GroEL. In the diagrams to the left, the composition and arrangement of the complexes are indicated, with ﬁlled circles representing mutant (V263S) subunits. Reproduced from Farr et al. (2000) with permission.

64

HELEN R. SAIBIL ET AL.

demanding. Other substrates, exempliﬁed by rhodanese, may have relaxed requirements, but whether this reﬂects reduced stringency in folding, different size, or particular domain arrangements in the non-native state is unclear. Third, refolding has the same requirements as binding, at least for MDH and rhodanese (G. Farr, unpublished results), not a surprising ﬁnding in light of the fact that the majority of both MDH and rhodanese molecules must bind to and release from GroEL multiple times before complete folding is achieved. Finally, GroES binding is the least demanding interaction, perhaps reﬂecting the cooperative action of ATP and GroES in bringing about the major domain movements that result in its binding. In sum, multivalent hydrophobic interactions are clearly critical to both binding and folding actions of the chaperonin. 4. Allostery in Polypeptide Binding Although allostery in the GroEL folding cycle is usually considered in relation to the ATPase cycle of the machine (see above), polypeptide binding also has the possibility of providing allosteric signals that direct the folding reaction. As mentioned above, when the stoichiometry of model peptide binding has been estimated, it often appears to be several per subunit. Small proteins that bind transiently to GroEL, such as lysozyme, may have a stoichiometry of one per ring; even some larger substrates, such as citrate synthase, may show such binding. For a number of stringent GroEL substrates, however, including MDH, Rubisco, rhodanese, and ornithine transcarbamylase, only one non-native molecule can bind to a double ring, implying allosteric control of polypeptide binding to the opposite ring. Another effect of this allostery is seen in the 50- to 100-fold acceleration of the dissociation of the cis folding complex by ATP and non-native polypeptide binding to the opposite ring (Rye et al., 1999). Are these the result of speciﬁc ring-to-ring communication following polypeptide binding, as described above for ATPmediated negative cooperativity, or are they simply the consequence of a “bulging” apical cavity on one ring distorting the normal equatorial communication with the other and forcing a conformational change that restricts polypeptide binding? Because stringent substrates that interact multivalently with GroEL in particular show this effect, perhaps it reﬂects the engagement of multiple apical domains simultaneously. B. Changes in Polypeptide Structure Another intriguing facet of polypeptide–GroEL interaction is the effect of binding on the structure of the non-native substrate. It is conceivable that the chaperonin, like other cellular chaperones, is a passive

GroEL/GroES-MEDIATED PROTEIN FOLDING

65

entity, simply binding folding intermediates to prevent their aggregation, then releasing them in an unchanged conformation to attempt to fold. A more plausible hypothesis, however, given its multivalency and the large motions of its polypeptide-binding domains during a folding cycle, is that the chaperonin actively inﬂuences the conformation of polypeptides bound to it. This would account for its apparent ability to rescue folding proteins from kinetic traps or other reversibly misfolded or aggregation-prone (or even aggregated) conformations. There are two points in the folding cycle where GroEL could inﬂuence substrate structure—one at the initial polypeptide binding state, and the other when polypeptide is released from the apical domains into the cis cavity upon ATP and GroES binding. Effects on polypeptide structure at each of these points of action have been examined for several substrates. 1. Changes during Binding It has already been pointed out in discussing the binding of model peptides that several that are unstructured in aqueous solution adopt characteristic structures when bound to GroEL. These bound structures are similar to those produced by structure-stabilizing solvents or found in the proteins from which the peptides were derived. On the other hand, there is one report (Preuss and Miller, 1999) of a peptide, stabilized as a nonamphipathic α-helix in solution by a templating residue, that becomes disordered on GroEL binding, even though related amphipathic peptides remain helical. Two mechanisms by which such folding or unfolding could take place have generally been considered—thermodynamic and catalytic (see Fenton and Horwich, 1997, for details). The ﬁrst reﬂects the effect of binding to GroEL on the equilibria among various folding states of a molecule in solution. If one state (or a set of states) binds more tightly than others, the equilibrium will shift toward that conformation, whether it is more or less folded than the overall solution state. Such an explanation seems likely for the behavior of the ﬁrst set of peptides, which assume a conformation at GroEL that is consistent with their natural propensity. The catalytic mechanism, on the other hand, suggests that some of the binding energy between a substrate and GroEL is used to drive a conformational change in the substrate. Such active unfolding could provide a means for GroEL to rescue a folding polypeptide from a kinetic trap. Although this mechanism could account for the production of the unstructured binding state of the latter α-helical peptide described above, the presence of the templating group in this molecule makes interpretation of this experiment equivocal.

66

HELEN R. SAIBIL ET AL.

For polypeptide substrates of GroEL, there is limited evidence for unfolding on binding, particularly by the catalytic mechanism. This reﬂects, at least in part, limited knowledge of the structural details of bound states, as well as the possibility that ensembles of states are bound to GroEL, each contributing uniquely to observed properties such as circular dichroism or ﬂuorescence. The most convincing evidence for a catalytic unfolding mechanism is from a deuterium-exchange experiment with barnase, a 6-kDa protein (Zahn et al., 1996). Even though it does not bind stably to GroEL, barnase undergoes more rapid hydrogen/deuterium exchange in the presence of catalytic amounts of GroEL, even in its normally buried core region, implying that transient binding to GroEL has globally unfolded this protein. On the other hand, pulsed deuterium exchange during a lysozyme refolding reaction in the presence of GroEL showed no unfolding of protected domains, despite GroEL enhancement of the folding rate (Coyle et al., 1999). A number of stable binary complexes have also been examined by deuterium exchange, but none has shown a similar global exchange coincident with binding. Some are already in a fully exchangeable state when bound following dilution from denaturant, whereas others exhibit modest levels of protection from exchange that do not change while in the complex. In the case of human dihydrofolate reductase in particular, even cycling off and onto GroEL in the presence of ATP did not lead to exchange of the protected amide protons of the GroEL-bound species (Groß et al., 1996). In most other cases, however, it has not been possible to observe non-native species in the absence of GroEL owing to aggregation or other irreversible changes, so the effect of binding on the folded state of the protected domains is not known. In the case of Rubisco, for example, the metastable folding intermediate in solution appears to have the same level of protection from exchange as the GroEL-bound species (Shtilerman et al., 1999), but it is unclear whether the same amide groups (and hence the same structure) are protected in the two forms of this protein. Somewhat more evidence exists for the action of the thermodynamic mechanism with polypeptides. Several small, nonstringent substrates, including RNase T1, β-lactamase, dihydrofolate reductase, and barstar, show this effect. It has been particularly well studied for RNase T1, where the presence of GroEL shifts the equilibrium between two non-native states toward the less folded one with higher afﬁnity for the chaperonin without changing the microscopic rate constants for their interconversion (Walter et al., 1996). Although this mechanism has not been clearly observed during the binding of stringent substrates, the kinetics of MDH refolding has been modeled by assuming that binding to GroEL shifts an

GroEL/GroES-MEDIATED PROTEIN FOLDING

67

equilibrium between non-native monomers and aggregating multimers toward the folding-competent monomers, thus improving the efﬁciency of refolding (Ranson et al., 1995). As a complication in these analyses, the multidomain nature of many GroEL substrates and the multivalency of GroEL binding suggest that both mechanisms may be occurring simultaneously and/or locally, so that it becomes difﬁcult to evaluate their contribution to the overall folding reaction. For example, in the DHFR refolding experiment mentioned above, while the protected core apparently remained intact upon rebinding to GroEL, suggesting no unfolding of this region, other properties of the folding intermediate were somewhat different from those of the initially bound state (Groß et al., 1996). These included tryptophan ﬂuorescence and iodide quenching and the ability to bind the hydrophobic dye anilinonaphthalenesulfonic acid. Did these changes reﬂect local structural rearrangements as a result of rebinding to GroEL, as the authors suggested, or had the chaperonin selected a particular nonnative state from the refolding mixture that was locally different from those available for binding during the initial dilution from denaturant? More detailed examination of folding intermediates will be necessary to address such questions. 2. Changes during Initiation of Folding The other step in the folding cycle where the interaction between GroEL and a substrate polypeptide makes an unfolding action possible is at the point of ATP/GroES-driven release of polypeptide from the apical binding sites into the cis cavity to begin folding. Structural data (see above) indicate that there are signiﬁcant changes in the apical domains and polypeptide-binding regions on initial ATP binding and massive changes in these regions accompanying subsequent GroES binding (Fig. 2). Because the polypeptide binding sites move so much relative to one another and because stringent substrates (at least) bind to multiple sites, it is attractive to hypothesize that the movement stretches the segments of polypeptide between the bound portions, destabilizing or completely removing any secondary structure present there before the bound segments are released. If the affected region has been misfolded, this “stretching on the rack” might unfold it sufﬁciently to attempt refolding when the polypeptide is released into the cavity (Shtilerman et al., 1999). Moreover, if the misfolded structure represented a kinetic trap on the folding landscape, repeated cycles of binding, stretching, and release, referred to as “iterative annealing,” might allow this region, and hence the polypeptide, eventually to escape the trap and fold to its native state (Todd et al., 1996).

68

HELEN R. SAIBIL ET AL.

However appealing it may be, there have been few experiments addressing this hypothesis. One result that could be interpreted to support this mechanism is in the study by Rye and colleagues (1997), where changes in the anisotropy of tryptophans in chaperonin-bound Rubisco were followed on stopped-ﬂow addition of ATP and GroES. Using a single-ring version of GroEL (SR1) to force complete conversion of SR1– Rubisco binary complexes to stable cis folding-active chambers, a rapid initial fall in anisotropy (t1/2 ∼1 sec) was observed, followed by a slower rise at a rate similar to that of the recovery of activity. The initial decrease appears to reﬂect a release of some (or all) of the six Rubisco tryptophans from local constraints on their motion. Although this could result from the GroES/ATP binding-directed unfolding of a tryptophan-containing secondary structure in the bound Rubisco, it is equally plausible that it represents the release of some of the tryptophans from hydrophobic binding contacts with the walls of the apical domains. In an effort to obtain more direct evidence for unfolding, Shtilerman and colleagues (1999) examined tritium exchange during Rubisco folding by GroEL. They ﬁrst determined that fully tritium-exchanged denatured Rubisco retained 12 highly protected tritiums (exchange times of >30 min) when diluted into either a buffer that prevented refolding or the same buffer containing a stoichiometric amount of GroEL. These tritiums could reﬂect the presence of one or more stable secondary structures in the GroEL-bound protein. Within 5 seconds of GroES/ATP addition, most (9.5) of the tritiums exchanged with the solvent. Because this happened more rapidly than ATP hydrolysis and complex dissociation (t1/2 ∼13 sec under these conditions), it seems likely to be a result of GroES/ATP binding and Rubisco release into the cavity rather than of complex turnover and substrate rebinding. The authors interpreted this rapid exchange as evidence for active unfolding of the substrate as a result of the domain movements accompanying ATP and GroES binding. Certain aspects of the experimental design and interpretation make this conclusion somewhat questionable, however. The inherent asymmetry of the GroEL folding cycle presents one problem. When GroES and adenine nucleotide are added to a GroEL–substrate complex, the GroES can bind to either the same ring as polypeptide (cis) or the opposite one (trans) with about equal afﬁnity. The cis complex releases polypeptide into the cavity and begins productive folding, but the trans complex remains unchanged until ATP hydrolysis and another round of ATP and GroES binding releases the ﬁrst GroES to form a new complex (see Fig. 4). Consequently, only about half of the initial GroEL– Rubisco complexes would have formed a cis GroES complex and released

GroEL/GroES-MEDIATED PROTEIN FOLDING

69

Rubisco in the ﬁve- to twelve-second interval tested; thus, the maximum observable exchange should have been only six tritiums. An alternative explanation for the exchange data is also possible. The crystal structures of peptides bound to GroEL show clearly that, in addition to a variety of hydrophobic interactions, there are hydrogen bonds between the peptide backbone and the apical binding domains. Similar interactions between tritium-labeled Rubisco and the binding domains would protect any tritiums involved from exchange in the binary complex. Once the protein was released into the cis cavity, however, these would be rapidly exchanged. This might even account for the apparent release of tritiums from the trans complexes, given that ATP binding in one ring affects the orientation of the apical domains of the other (Ranson et al., submitted). MDH, another stringent GroEL substrate, has also been examined by hydrogen exchange, here using D2O and HPLC–MS to monitor deuterium content (Chen et al., 2001). When diluted from deuterated denaturant into protic buffer containing SR1, a number of deuterons were partially protected from exchange, although none to the degree of the Rubisco tritiums in the previous study. Rapid peptide mapping showed that thirteen of these were found in an apparent “core” region comprising part of the Rossmann fold of the mature protein. Using pulsed deuterium exchange of SR1–MDH complexes at various times after initiating a folding reaction with GroES and ATP, the authors showed that these same protons largely remained protected on release of MDH into the cis cavity, arguing against a global unfolding action driven by the apical domain movements. Some increased exchange was noted, but it was widely distributed in peptides throughout MDH, a ﬁnding more consistent with the release of backbone hydrogen bonds to the apical binding sites. In support of the conclusion that MDH is not generally unfolded upon GroES binding, preliminary ﬂuorescence anisotropy experiments with tryptophan-containing variants of MDH show no rapid drop in anisotropy when GroES and ATP are added to SR1–MDH binary complexes (J. Cochrane and W. Fenton, unpublished). Thus, even if forced unfolding accounts for a fraction of the tritium exchange seen in the Rubisco experiment, it seems clear that other substrates do not experience such an action as part of their folding cycle. C. The Role of the Folding Cavity Although it is well established that the GroES-encapsulated cis chamber of a GroEL–substrate complex is the critical space in which folding

70

HELEN R. SAIBIL ET AL.

occurs, at least for stringent substrates, the role that the cavity itself plays in the refolding reaction is unclear, beyond providing an “Anﬁnsen cage” to ensure that folding proceeds in the absence of other, potentially interfering (by aggregation, for example) molecules (Ellis and Hartl, 1996). The crystal structure shows that the walls of the chamber are polar and hydrophilic, in contrast to the hydrophobic nature of the open, polypeptide binding state. Does this play anything more than a passive role in stabilizing exposed polar segments of folding intermediates and destabilizing exposed hydrophobic ones, thus favoring a native-like state? In addition, the size of the cavity is clearly important in determining the size of GroEL substrates, but does conﬁnement have another role? Even moderate-sized substrates may interact with the chamber walls, as shown by the increased correlation time of GFP conﬁned to SR1 after folding (Weissman et al., 1996). Because the folding pathways of most GroEL substrates are not known, detailed investigation of this question has not been possible. In the case of lysozyme, however, a nonstringent substrate with a well deﬁned folding pathway, Coyle and colleagues (1999) have examined refolding in the presence and absence of GroEL by stoppedﬂow ﬂuorescence and pulsed deuterium exchange. GroEL alone accelerates folding about 1.5-fold, but there is no domain unfolding and no apparent change in the folding pathway or mechanism, which involves domain docking as its ﬁnal rate-limiting step. The authors suggest that minor structural changes on GroEL binding lead to a reorganization of non-native interactions in the α-domain without wholesale unfolding of the native-like core domain structures. This rearranged α-domain interface can then dock productively with the β-domain to form active enzyme. An alternative, however, is that the GroEL cavity itself increases the probability of productive domain interactions, and hence increases the folding rate, by restricting the folding space and orientations available to the bound folding intermediate. IV. CONCLUSIONS Advances in structural biology are revealing additional states of the chaperonin machine that represent essential steps in the protein folding reaction. At the same time, biochemical and biophysical approaches are addressing fundamental questions regarding the physical states and folding trajectories of substrate polypeptides within the privileged environment of the folding-active chaperonin complex. Together, these techniques should provide new insights into this critical cellular process.

GroEL/GroES-MEDIATED PROTEIN FOLDING

71

REFERENCES Boisvert, D. C., Wang, J., Otwinowski, A., Horwich, A. L., and Sigler, P. B. (1996). Nature Struct. Biol. 3, 170–177. Brazil, B. T., Cleland, J. L., McDowell, R. S., Skelton, N. J., Paris, K., and Horowitz, P. M. (1997). J. Biol. Chem. 272, 5105–5111. Buckle, A. M., Zahn, R., and Fersht, A. R. (1997). Proc. Natl. Acad. Sci. USA 94, 3571–3575. Chatellier, J., Buckle, A. M., and Fersht, A. R. (1999). J. Mol. Biol. 292, 163–172. Chen, J., Walter, S., Horwich, A. L., and Smith, D. L. (2001). Nature Struct. Biol. 8, 721–728. Chen, S., and Sigler, P. B. (1999). Cell 99, 757–768. Chen, S., Roseman, A. M., Hunter, A. S., Wood, S. P., Burston, S. G., Ranson, N. A., Clarke, A. R., and Saibil, H. R. (1994). Nature 371, 261–264. Cliff, M. J., Kad, N. M., Hay, N., Lund, P. A., Webb, M. R., Burston, S. G., and Clarke, A. R. (1999). J. Mol. Biol. 293, 667–684. Coyle, J. E., Jaeger, J., Grosse, M., Robinson, C. V., and Radford, S. E. (1998). Folding Des. 3, R93–R104. Coyle, J. E., Texter, F. L., Ashcroft, A. E., Masselos, D., Robinson, C. V., and Radford, S. E. (1999). Nature Struct. Biol. 6, 683–690. Ditzel, L., L¨owe, J., Stock, D., Stetter, K.-O., Huber, H., Huber, R., and Steinbacher, S. (1998). Cell 93, 125–138. Ellis, R. J., and Hartl, F. U. (1996). FASEB J. 10, 20–26. Ewalt, K. L., Hendrick, J. P., Houry, W. A., and Hartl, F. U. (1997). Cell 90, 491–500. Farr, G. W., Furtak, K., Rowland, M. B., Ranson, N. A., Saibil, H. R., Kirchhausen, T., and Horwich, A. L. (2000). Cell 100, 561–573. Fenton, W. A., and Horwich, A. L. (1997). Protein Sci. 6, 743–760. Fenton, W. A., Kashi, Y., Furtak, K., and Horwich, A. L. (1994). Nature 371, 614–619. Gervasoni, P., Gehrig, P., and Pl¨uckthun, A. (1998a). J. Mol. Biol. 275, 663–675. Gervasoni, P., and Pl¨uckthun, A. (1997). FEBS Lett. 401, 138–142. Gervasoni, P., Staudenmann, W., James, P., and Pl¨uckthun, A. (1998b). Biochemistry 37, 11660–11669. Gibbons, D. L., Hixson, J. D., Hay, N., Lund, P., Gorovits, B. M., Ybarra, J., and Horowitz, P. M. (1996). J. Biol. Chem. 271, 31989–31995. Grallert, H., and Buchner, J. (1999). J. Biol. Chem. 274, 20171–20177. Grallert, H., Rutkat, K., and Buchner, J. (1998). J. Biol. Chem. 273, 33305–33310. Grantcharova, V., Alm, E. J., Baker, D., and Horwich, A. L. (2001). Curr. Opin. Struct. Biol. 11, 70–82. Groß, M., Robinson, C. V., Mayhew, M., Hartl, F. U., and Radford, S. E. (1996). Protein Sci. 5, 2506–2513. Horwich, A. L., Low, K. B., Fenton, W. A., Hirshﬁeld, I. N., and Furtak, K. (1993). Cell 74, 909–917. Houry, W. A., Frishman, D., Eckerskorn, C., Lottspeich, F., and Hartl, F. U (1999). Nature 402, 147–154. Inbar, E., and Horovitz, A. (1997). Biochemistry 36, 12276–12281. Jackson, G. S., Staniforth, R. A., Halsall, D. J., Atkinson, T., Holbrook, J. J., Clarke, A. R., and Burston, S. G. (1993). Biochemistry 32, 2554–2563. Kobayashi, N., Freund, S. M., Chatellier, J., Zahn, R., and Fersht, A. R. (1999). J. Mol. Biol. 292, 181–190. Landry, S. J., and Gierasch, L. M. (1991). Biochemistry 30, 7359–7362. Preuss, M., Hutchinson, J. P., and Miller, A. D. (1999). Biochemistry 38, 10272–10286. Preuss, M., and Miller, A. D. (1999). FEBS Lett. 461, 131–135.

72

HELEN R. SAIBIL ET AL.

Ranson, N. A., Dunster, N. J., Burston, S. G., and Clarke, A. R. (1995). J. Mol. Biol. 250, 581–586. Ranson, N. A., White, E. H., and Saibil, H. R. (1998). Biochem. J. 333, 233–242. Ranson, N. A., Farr, G. W., Roseman, A. M., Gowen, B., Fenton, W. A., Horwich, A. L., and Saibil, H. R. (2001). Submitted. Roseman, A., Chen, S., White, H., Braig, K., and Saibil, H. (1996). Cell 87, 241–251. Roseman, A. M., Ranson, N. A., Gowen, B., Fuller, S. D., and Saibil, H. R. (2001). J. Struct. Biol. 135, 115–125. Rye, H. S., Burston, S. G., Fenton, W. A., Beechem, J. M., Xu, Z., Sigler, P. B., and Horwich, A. L. (1997). Nature 388, 792–798. Rye, H. S., Roseman, A. M., Chen, S., Furtak, K., Fenton, W. A., Saibil, H. R., and Horwich, A. L. (1999). Cell 97, 325–338. Sigler, P. B., Xu, Z., Rye, H. R., Burston, S. G., Fenton, W. A., and Horwich, A. L. (1998). Annu. Rev. Biochem. 67, 581–608. Shtilerman, M., Lorimer, G. H., and Englander, S. W. (1999). Science 284, 822–825. Tanaka, N., and Fersht, A. R. (1999). J. Mol. Biol. 292, 173–180. Thiyagarajan, P., Henderson, S. J., and Joachimiak, A. (1996). Structure 4, 79–88. Todd, M. J., Lorimer, G. H., and Thirumalai, D. (1996). Proc. Natl. Acad. Sci. USA 93, 4030–4035. Walter, S., Lorimer, G. H., and Schmid, F. X. (1996). Proc. Natl. Acad. Sci. USA 93, 9425– 9430. Wang, Z., Feng, H., Landry, S. J., Maxwell, J., and Gierasch, L. M. (1999). Biochemistry 38, 12537–12546. Weissman, J. S., Rye, H. S., Fenton, W. A., Beechem, J. M., and Horwich, A. L. (1996). Cell 84, 481–490. White, H. E., Chen, S., Roseman, A. M., Yifrach, O., Horovitz, A., and Saibil, H. R. (1997). Nature Struct. Biol. 4, 690–694. Xu, Z., Horwich, A. L., and Sigler, P. B. (1997). Nature 388, 741–750. Yifrach, O., and Horovitz, A. (1994). J. Mol. Biol. 243, 397–401. Yifrach, O., and Horovitz, A. (1995). Biochemistry 34, 5303–5308. Yifrach, O., and Horovitz, A. (1996). J. Mol. Biol. 255, 356–361. Yifrach, O., and Horovitz, A. (2000). Proc. Natl. Acad. Sci. USA 97, 1521–1524. Zahn, R., Perrett, S., Stenberg, G., and Fersht, A. R. (1996). Science 271, 642–645.

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES By NICHOLAS J. COWAN and SALLY A. LEWIS Department of Biochemistry, NYU Medical Center, New York, New York 10016

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Discovery of Type II Chaperonins and Early Functional Studies . . . . . . . . . . III. Subunits and Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. CCT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Thermosome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Target Range and Specificity of CCT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Cycling of Target Proteins by CCT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Structure of Type II Chaperonins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Structure of CCT and the CCT/α-Actin Binary Complex . . . . . . . . . . . . B. Structure of the Thermosome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Prefoldin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Biochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Role of Prefoldin In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX. Tubulin-Specific Chaperones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The Tubulin Folding Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Tubulin Cofactors in S. cerevisiae and S. pombe . . . . . . . . . . . . . . . . . . . . . . C. Postfolding Functions of These Chaperones . . . . . . . . . . . . . . . . . . . . . . . X. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

73 74 75 76 78 79 81 83 84 84 87 88 89 90 92 93 96 97 98 98

I. INTRODUCTION Chaperonins are a specialized but ubiquitous group of molecular chaperones that are characterized by a multisubunit ring structure. They exist in all kingdoms of life, in all cases using cycles of ATP binding and hydrolysis to generate allosteric changes that promote the correct folding of bound target protein. Chaperonins have been classiﬁed into two groups (Horwich and Willison, 1993; Kim et al., 1994). Type I chaperonins are present in prokaryotes (exempliﬁed by GroEL/GroES), in mitochondria (exempliﬁed by Hsp60/hsp10), and in chloroplasts (exempliﬁed by Rubisco subunit binding protein), while Type II chaperonins include the chaperonin present in archaea (exempliﬁed by the thermosome) and the chaperonin present in the cytosol of eukaryotes (referred to here as CCT, for cytosolic chaperonin-containing TCP-1, but which is also termed c-cpn or TriC). Two principal features distinguish the two types of chaperonin. First, Type I chaperonins possess sevenfold 73 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0096-5332/02 $35.00

74

NICHOLAS J. COWAN AND SALLY A. LEWIS

symmetry, while Type II chaperonins form eight- or nine-membered rings. Second, Type I chaperonins function in conjunction with a cochaperonin that is an essential participant in the reaction cycle. No functionally similar co-chaperonin is known to participate in the Type II chaperonin reaction cycle. Prefoldin is a recently discovered chaperone that exists in archaea and in eukaryotes, but has no counterpart in prokaryotes. It is a heterohexameric molecule with an apparent molecular mass of about 200 kDa. Prefoldin binds to nascent polypeptide chains that will ultimately require facilitated folding by CCT, and delivers them to the chaperonin, with which it interacts. The principal targets for facilitated folding by CCT in cooperation with prefoldin are the cytoskeletal proteins actin and tubulin. The actin monomer assembles into microﬁlaments, while the subunit that forms microtubules is the tubulin heterodimer, which consists of a single α- and a single β-tubulin polypeptide. Though actin can be folded to the native state via one or more cycles of ATP-dependent interaction with CCT, this is not the case for either α- or β-tubulin. Tubulin subunits released from CCT are assembled into α/β heterodimers by interaction with several tubulin-speciﬁc chaperones known as cofactors in a reaction that depends on GTP hydrolysis by the cofactor-bound tubulin. II. DISCOVERY OF TYPE II CHAPERONINS AND EARLY FUNCTIONAL STUDIES The discovery of the chaperonin GroEL/GroES in bacteria (Bochkareva et al., 1988; Goloubinoff et al., 1989a; Goloubinoff et al., 1989b; Hemmingsen et al., 1988; Hendrix, 1979; Hohn et al., 1979) and its homologs in mitochondria (Cheng et al., 1989; Lubben et al., 1990; Ostermann et al., 1989; Viitanen et al., 1992) and chloroplasts (Viitanen et al., 1990) during the period from 1979 to 1992 strongly suggested that a functionally related molecule might exist in the cytosol of eukaryotes. This suspicion was reinforced by the observation that expression in Escherichia coli of cDNAs encoding the exclusively eukaryotic proteins actin and tubulin led to the production of completely insoluble inclusion bodies in the bacterial host cells. On the other hand, in vitro translation of the identical cloned sequences in a eukaryotic cell extract (for example, from rabbit reticulocyte lysate) led to the production of native protein (Cleveland et al., 1978). The mouse t-complex polypeptide 1 (TCP-1) seemed a good candidate as a constituent of a eukaryotic chaperonin homolog because it bears a distant amino acid sequence similarity with GroEL and the mitochondrial chaperonin Hsp60 (Gupta, 1990). Indeed, a 900-kDa tubulin-containing complex puriﬁed from reticulocyte lysate contained TCP-1 as a component (Lewis et al., 1992; Yaffe et al.,

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

75

1992). Moreover, the thermophilic archaebacterium Sulfolobus shibitae was shown to contain a heat shock protein (TF55) with 40% amino acid sequence identity to mouse TCP-1 (Trent et al., 1991). TF55 formed a double toroid and was able to bind unfolded substrates in vitro, thus strongly implicating it as a functional chaperonin molecule. The biochemical isolation and characterization of the cytosolic chaperonin from a eukaryotic source were initially accomplished in two ways. First, an in vitro folding assay was developed that exploited the deposition of insoluble actin inclusion bodies in E. coli. These inclusion bodies could be made selectively radioactive by including 35S-methionine and rifampicin in the bacterial culture (Studier et al., 1990), and are relatively easy to purify; solubilization in guanidine or urea then yields a probe of sufﬁciently high speciﬁc radioactivity for use in biochemical folding assays. Such assays allowed the puriﬁcation from reticulocyte lysate— already known to contain all that is necessary to produce native actin by in vitro translation—of the component responsible for the facilitated folding of β-actin (Cowan, 1998; Gao et al., 1992). The puriﬁed protein migrated on gel ﬁltration with a molecular mass of >500 kDa and contained multiple distinct polypeptides, one of which was identiﬁed immunologically as TCP-1. Most persuasively, it appeared as a double toroid in the electron microscope, where its orientation on the grid depended on the presence or absence of ATP, implying a striking structural change depending on its nucleotide-bound state (Gao et al., 1992) (Fig. 1). Second, based on the assumption that TCP-1 was indeed a component of the eukaryotic chaperonin, a monoclonal antibody was used to follow its puriﬁcation from a crude extract of bovine testis tissue, where expression of TCP-1 was known to be relatively abundant. This resulted in the isolation of a double toroidal molecule with the ability to facilitate the in vitro folding of denatured ﬁreﬂy luciferase and, ostensibly, tubulin (Frydman et al., 1992; but see Section IX). An alternative and convenient puriﬁcation protocol for CCT from rabbit reticulocyte lysate that does not involve anion exchange chromatography or afﬁnity chromatography on ATP–agarose has since been described (Norcum, 1996). III. SUBUNITS AND ASSEMBLY In common with the polypeptides of Type I chaperonins, each subunit of Type II chaperonins is organized into three recognizable domains: an equatorial domain that contains the ATP-binding site and forms the inter-ring contacts, an apical domain that contains the sites of interaction with target protein, and an intermediate domain that forms a hingelike connection between the two (Ditzel et al., 1998; Klumpp et al., 1997;

76

NICHOLAS J. COWAN AND SALLY A. LEWIS

FIG. 1. Appearance of CCT in the electron microscope. CCT puriﬁed from rabbit reticulocyte lysate was examined in the absence (a) or presence (b) of ATP. Reprinted from Gao et al. (1992), with permission.

Waldmann et al., 1995). The extent of amino acid sequence identity between Type I and Type II chaperonins is only 15–25%, implying a very early divergence from a common ancestor occurring more than two billion years ago (Gutsche et al., 1999; Kubota et al., 1994; Kubota et al., 1995a). A. CCT Each ring of CCT contains eight different but related polypeptides, designated α, β, γ , δ, ε, η, and ζ [Kubota et al., 1994; Kubota et al., 1995a; Kubota et al., 1995b; CCT1-8 in yeast, reviewed in Stoldt et al. (1996)]; these polypeptides are about 30% identical in sequence. The duplication of genes encoding the individual subunits and their sequence divergence must have been an early event in the evolution of eukaryotes because homologous subunits among different eukaryotic species share a higher degree of similarity compared with different subunits from the same organism. Sequence similarities between Type I chaperonins and CCT are greatest in the ATP-binding domain, less in the intermediate

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

77

domain, and nonexistent in the apical domain (Kim et al., 1994; Kubota et al., 1994; Lewis et al., 1992). The latter observation probably reﬂects the relatively restricted target range and hence specialization of CCT (compared with Type I chaperonins) discussed below. The mouse genes encoding the subunits of CCT have all been cloned and sequenced, and their patterns of expression examined in different tissues, as well as in response to chemically induced stress and as a function of the cell cycle (Kubota et al., 1999a; Kubota et al., 1999b; Kubota et al., 2000; Yokota et al., 1999; Yokota et al., 2000). Though the level of expression varies widely among tissues, it is relatively similar among the eight subunits, suggesting a tight coregulation in order to maintain a constant ratio for association into the hexadecamer (Kubota et al., 1999b). The subunits within CCT are arranged with a ﬁxed geometry, with one of each type of subunit in each ring. Deﬁnition of the order of subunits within each ring of CCT has been facilitated by the existence within cells of CCT microcomplexes that contain subsets of these subunits. These microcomplexes, which are present at about 5% of the level of intact chaperonin, can be fractionated by gel ﬁltration and behave with apparent molecular masses in the range 120–250 kDa. By a combination of sucrose gradient fractionation, gel ﬁltration, nondenaturing gel electrophoresis, and Western blotting using subunit-speciﬁc antibodies, it has been possible to deﬁne the composition of microcomplexes containing two or more subunits, and hence the overall pattern of neighbor-to-neighbor contacts within the CCT ring (Liou and Willison, 1997) (Fig. 2). Whether the subunits are arranged in this order

FIG. 2. Location of subunits in each of the two stacked rings of CCT. Reprinted from Liou and Willison (1997), with permission.

78

NICHOLAS J. COWAN AND SALLY A. LEWIS

clockwise or counterclockwise as viewed from the top of the ring is unknown. Also yet to be elucidated is the question of which pairs of subunits make the weaker inter-ring contacts. In any case, these data, and the biochemical and structural properties of CCT isolated from testis or reticulocyte lysate, are consistent with the existence of CCT as a single molecular species with a unique structure. However, immunocytochemical observations using subunit-speciﬁc antibodies have been interpreted in terms of the existence of heterogeneous populations of chaperonin particles within neuronal cells (Roobol et al., 1995; Roobol and Carden, 1999); these data could conceivably be attributable to reactions with microcomplexes. Neither translation of subunits in vitro nor their presentation as labeled, denatured target protein to CCT results in the generation of detectable subunit/CCT binary complexes; thus, it seems unlikely that these polypeptides are themselves targets for CCT-mediated folding. However, the assembly of newly synthesized CCT is very inefﬁcient on translation of subunits in vitro. It has been suggested that newly synthesized subunits are incorporated into preexisting CCT complexes via a nucleotide-dependent disassembly reaction that produces single rings, and that these single rings serve as templates for the assembly of the second ring (Liou et al., 1998). There is also a report that one or more unknown protein factors may function in the K+- and ATP-dependent assembly of CCT (Roobol et al., 1999). The contribution of these reactions (if any) to the overall mechanism of CCT-mediated folding is not known. B. Thermosome “Thermosome” is the term used to describe CCT homologs in the archaea, a group of prokaryotes thought to be descended from the organism that is the ancestor of the eukaryotic cytosol. Archaea and eukaryotes are more similar to each other than to eubacteria in their translation and folding machinery, the relationship between their chaperonins being a prime example of this phenomenon. The thermosome from methanogens contains only a single subunit, but the majority of thermosomes are assembled from two homologous subunits, α and β, which share about 55% amino acid sequence identity. The nucleotide and temperature conditions that dictate the assembly, dissociation, or polymerization of thermosomes in vitro vary enormously depending on the species of origin; data describing these conditions have recently been reviewed elsewhere (Gutsche et al., 1999). The fact that

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

79

the thermosome is a very abundant protein and that it can polymerize in vitro suggests that it may play a direct cytoskeletal role in the archaea. IV. TARGET RANGE AND SPECIFICITY OF CCT Estimates of the spectrum of target proteins that undergo facilitated folding by the Type I chaperonin GroEL/GroES vary quite widely, but in any event appear to involve minimally about 5% of newly synthesized proteins in E. coli (Lorimer, 1996). Since most of these target molecules share little or no structural homology, recognition of folding intermediates is considered to be promiscuous, i.e., dependent on some general feature of folding intermediates such as exposed hydrophobic surfaces which can interact with the chaperonin. Given this scenario, it seemed plausible that the cytosolic chaperonin, like its Type I relatives, might also recognize and facilitate the folding of a broad range of eukaryotic target molecules. On the other hand, the abundance of CCT in tissues such as liver (which is very actively engaged in protein synthesis) is relatively low. This fact, together with the slow turnover rate of CCT (compared with GroEL/GroES) in vitro, implies that there is simply not enough CCT to participate in the facilitated folding of more than a small subset of newly synthesized protein. In addition to this quantitative argument, several experimental lines of evidence point directly or indirectly to a relatively narrow target range for CCT. First, an analysis of CCT binary complexes by mass spectrometry resulted in the identiﬁcation of actins and tubulins as the principal components bound to the chaperonin (Hynes et al., 1996); indeed, when cultured cells are pulse-labeled with [35S]-methionoine, actins and tubulins account for the overwhelming majority of CCT-associated proteins (Yaffe et al., 1992). Second, CCT has a much higher afﬁnity for actin and tubulin target proteins in vitro compared to other proteins of noncytoskeletal origin (Melki and Cowan, 1994). Third, temperature-sensitive mutants of CCT in Saccharomyces cerevisiae all display cytoskeletal phenotypes at the nonpermissive temperature, implicating cytoskeletal proteins as the principal (or at least the most sensitive and essential) target proteins of CCT (Chen et al., 1994; Ursic and Culbertson, 1991; Vinh and Drubin, 1994). Although CCT can form a binary complex with a wide range of unfolded proteins in vitro (Melki et al., 1997; Melki and Cowan, 1994; Tian et al., 1995b), this does not mean that this happens in vivo. For example, the prototypical Type I chaperonin GroEL binds about 50% of the polypeptides present in an unfractionated mixture of denatured,

80

NICHOLAS J. COWAN AND SALLY A. LEWIS

labeled E. coli proteins, yet the fraction of proteins that require facilitated folding by GroEL/GroES in vivo is thought to be close to 5% (Lorimer, 1996). On the other hand, it has been suggested, based on immunoprecipitation reactions, that CCT might interact with a broad range (accounting for 9–15%) of newly synthesized eukaryotic proteins (Feldman and Frydman, 2000; McCallum et al., 2000; Thulasiraman et al., 1999). There is also evidence that some proteins other than actins and tubulins fold via interaction with CCT. These include Gα-transducin (Farr et al., 1997), cyclin E (Won et al., 1998), and the von Hippel–Landau tumor suppressor protein VHL (Feldman et al., 1999). Moreover, translation in vitro of myosin heavy and light chains has identiﬁed an intermediate in the biogenesis of the heavy meromyosin subunit (HMM) of skeletal muscle myosin that contains all three myosin subunits and CCT, from which partially folded HMM can be released in an ATP-dependent reaction. Other as yet unknown cytosolic protein(s) are also apparently required for the completion of the myosin folding reaction (Srikakulam and Winkelmann, 1999). The fact that neither detectable native actin nor tubulin (nor many other proteins) is produced on expression in E. coli appeared somewhat paradoxical, given the apparent promiscuity of Type I chaperonins such as GroEL/GroES. At least a partial resolution of this paradox became apparent when an attempt was made to fold actin (or tubulin) in vitro in a GroEL/GroES-mediated reaction. While labeled, denatured actin or tubulins could form a binary complex with Type I chaperonins, these target proteins underwent endless cycles of ATP-dependent release and rebinding to the chaperonin without ever partitioning to the native state. Only on inclusion of CCT in the reaction mixture did the folding intermediates cycling on GroEL/GroES transfer to the cognate chaperonin and subsequently acquire their native conformation (Tian et al., 1995b) (Fig. 3). This experiment demonstrated that CCT is speciﬁcally tailored for the facilitated folding of these cytoskeletal proteins, suggesting that they have coevolved in the cytosol of eukaryotes. The unique ability of CCT to productively fold actins and tubulins gives further weight to the argument that CCT is specialized for the facilitated folding of relatively few cytosolic proteins. The target range of the archaeal Type II chaperonins remains to be investigated. No natural in vivo targets have been identiﬁed to date, although there are reports of facilitated folding in vitro of several thermophilic enzymes by the thermosomes from Sulfolobus solfataricus (Guagliardi et al., 1994) and Methanococcus thermolithotrophicus (Furutani et al., 1998).

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

81

FIG. 3. Only CCT (but not GroEL) can productively fold actin or tubulin. Binary complexes formed in vitro between GroEL and unfolded β-actin (A, B) or α-tubulin (C) were incubated in the presence of GroES, ATP (A, B) or ATP, GTP and tubulin-speciﬁc chaperones (C), and an eightfold molar excess of CCT (B, C). After the incubation times (in minutes) shown, the reaction products were analyzed by nondenaturing gel electrophoresis. Arrows (top to bottom) show the location of CCT binary complexes, GroEL binary complexes, and either native actin or native tubulin heterodimers. Adapted from Tian et al. (1995b), with permission.

V. CYCLING OF TARGET PROTEINS BY CCT In the case of the best-studied Type I chaperonin, GroEL/GroES, folding occurs via multiple rounds of binding to and ATP-dependent release from the chaperonin (Bukau and Horwich, 1998). Only a fraction of target protein molecules—of the order of 20%—becomes committed to the native state while sequestered under the cap provided by the co-chaperonin GroES. On release, therefore, the majority of target molecules do not reach the native state; they either rebind to another chaperonin molecule, where a further “attempt” is made at productive folding, become bound to other chaperones, or follow a degradation pathway. The successive binding and ATP-dependent release of nonnative forms by chaperonin is known as cycling. Whether or not target proteins bound to Type II chaperonins undergo similar cycles of binding and release has been a matter of controversy. On the one hand, CCT-generated tubulin intermediates can be efﬁciently trapped by the inclusion in the folding reaction of a molar excess of

82

NICHOLAS J. COWAN AND SALLY A. LEWIS

Hsp60, which is not competent to productively fold tubulin (Gao et al., 1994). Furthermore, when CCT/β-actin binary complexes are incubated with ATP in the presence of excess Type I chaperonin trap, there is a large reduction in the yield of native product (Tian et al., 1995b). In contrast, addition of an excess of a GroEL mutant, G337S/I349E (a mutation that affects the GroEL apical domain), to a rabbit reticulocyte cell-free translation cocktail did not measurably reduce the yield of native product (Frydman and Hartl, 1996). Although intended to trap target proteins in this experiment, it has been shown that the G337S/I349E mutant does in fact efﬁciently discharge its target protein in the presence of ATP (Farr et al., 1997). Further studies done with a more effective GroEL trap, D87K, dramatically suppressed the yield of native product. Similar data were obtained in an authentic in vivo context, i.e., on microinjection of this trap into Xenopus oocytes (Farr et al., 1997) (Fig. 4). Notwithstanding questions regarding the effectiveness of different mutant traps in sequestering cycling intermediates, it is worth sounding a

FIG. 4. Microinjection of the GroEL trap D87K into X. laevis oocytes prevents newly translated actin from reaching the native state: Hatched bars represent the amount of radiolabeled native actin present after ﬁve minutes of labeling. Note that when the trap is injected almost no more native actin is produced (right-hand graph). Reprinted from Farr et al. (1997), with permission.

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

83

note of caution concerning folding experiments done either in vivo or in whole cell extracts. Such extracts contain the entire spectrum of chaperones present in the cytosol, and their presence is likely to affect the partitioning of target proteins. For example, the presence of prefoldin— a relatively abundant component in reticulocyte lysate that functions by delivering target proteins speciﬁcally to CCT and whose properties are discussed in Section VIII—seems likely to direct cycling folding intermediates to CCT and away from traps introduced to capture them. Taken together, the evidence points to a cycling of intermediates by CCT analogous to that which occurs in the case of Type I chaperonins. One reason that this controversy persists is the fact that CCT-mediated folding in vitro proceeds much less efﬁciently than either in vivo or in a cell-free translation cocktail. In the former case, attainment of the native state by target protein requires many more cycles of binding and release from CCT. This relative inefﬁciency is almost certainly due to the nonvectorial nature of in vitro folding reactions in which denatured target protein is suddenly diluted from denaturant so that it can assume a large number of unnatural and nonproductive conformations. These would not occur in a translational context. In the in vivo situation it is clear that productive folding by CCT requires only one or few rounds of ATP hydrolysis (Farr et al., 1997; Siegers et al., 1999). Nevertheless, we feel that the cycling of target protein is a crucial property of CCT and all chaperonins, allowing them to unload mutant or irrevocably denatured target protein and providing an automatic mechanism for releasing folded or folding proteins. Indeed, chaperones in general appear to function by releasing and rebinding target proteins, allowing a kinetic partition of folding intermediates among the chaperone proteins present in the cell, and ultimately leading to their folding, subcellular sorting, or degradation (Bukau et al., 2000; Bukau et al., 1996). VI. GENETICS The genetics of CCT has been investigated in S. cerevisiae. (In this organism the subunit composition of CCT is the same as in mammals, but the corresponding genes are numbered rather than given Greek letters.) Expression of six of the eight subunits [CCT1(α) (Ursic and Culbertson, 1991), CCT2(β) (Chen et al., 1994), CCT3(γ ) (Chen et al., 1994), CCT4(δ) (Vinh and Drubin, 1994), CCT6(ζ ) (Lin et al., 1997), and CCT8(θ) (Esser et al., 1999)] has been ablated by gene disruption experiments, which show that each is essential for survival. Conditional mutants of four of the subunits (CCT1–4) were found to produce growth arrest and cytoskeletal defects when shifted to growth at nonpermissive

84

NICHOLAS J. COWAN AND SALLY A. LEWIS

temperatures (Chen et al., 1994; Miklos et al., 1994; Ursic and Culbertson, 1991; Ursic et al., 1994; Vinh and Drubin, 1994). These phenotypes are consistent with what we know about the biochemistry and target range of mammalian CCT, and can be understood in terms of deﬁcits in tubulin and/or actin folding. Interestingly, some of the conditional mutations affect the actin cytoskeleton more than microtubules, while others have the reverse effect. For example, the cct4-1 mutant displayed actin defects similar to those found in actin mutants, while the tubulin cytoskeleton appeared normal (Vinh and Drubin, 1994); on the other hand, cct1-245 and cct2-326 lead to abnormal microtubules when expressed, but resulted in no defect in actin-mediated secretion (Miklos et al., 1994). These data suggest that speciﬁc CCT subunits interact differentially with these two major CCT target proteins, an inference that has been further substantiated by structural studies (see Section VII). VII. STRUCTURE OF TYPE II CHAPERONINS A. Structure of CCT and the CCT/α-Actin Binary Complex Because CCT is assembled from eight different polypeptides, the prospect of engineering a host/vector system for the expression of recombinant chaperonin is a daunting one; hence, all studies of CCT have thus far depended on material puriﬁed from a eukaryotic source such as mouse or bovine testis or rabbit reticulocyte lysate. Even if sufﬁcient material could be puriﬁed from these tissues, the heterooligomeric nature of the particle might make crystallization extremely challenging. Structural analyses of CCT have therefore been conﬁned to studies by cryoelectron microscopy. Nonetheless, such analyses have proved very useful in identifying some of the unique characteristics of this chaperonin. Three-dimensional reconstruction of apo-CCT and ATP-CCT has been obtained at 28 A˚ resolution (Fig. 5). Binding of ATP induces large conformational changes compared to apo-CCT (Llorca et al., 1998) and generates an asymmetric particle in which one ring has a slightly different conformation from the apo-CCT ring. Binding of ATP causes an upward displacement of the equatorial domain toward the interior of the cavity, and a rotation of the apical domains such that they point toward the cylinder axis. This suggests that, as in the case of the thermosome, the helical protrusions present at the tips of the apical domains can act as an iris diaphragm, closing the interior space in a manner analogous to the lid provided by the binding of GroES to GroEL. Indeed, there is an almost perfect ﬁt between the ATP ring of CCT as determined by cryoelectron

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

85

FIG. 5. Three-dimensional reconstructions of apo-CCT and ATP-CCT. Side, top, and tilted views are shown; additionally, bottom views are shown for the markedly asymmetric ATP-CCT particle. Reprinted from Llorca et al. (1999b), with permission.

microscopy and the enclosed structure of the thermosome from Thermoplasma acidophilum in the Mg-ADP-AIF3 form, which mimics the ATPbound state. In either case, there is a resulting enlargement of the enclosed cavity. Thus, it is likely that for all Type II chaperonins, the need for interaction with a co-chaperonin is abrogated (Llorca et al., 1999b). The structure of the binary complex formed between α-actin and CCT has also been examined by cryoelectron microscopy (Llorca et al., 1999a; Marco et al., 1994). This revealed that actin binds within the CCT cavity as a rod-shaped mass attached at either end to the apical domains of two

86

NICHOLAS J. COWAN AND SALLY A. LEWIS

CCT subunits placed 1,4 with respect to each other (Llorca et al., 1999a). The structure of the CCT-bound actin corresponds to an “opened up” version of native actin, in which the tip of each of the two major domains is bound to a CCT subunit, suggesting that the bound target protein has a quasi-native conformation. No particles with target protein bound to both rings were observed, even at a substrate : CCT ratio of 10 : 1. Indeed, the binding of β-tubulin to one ring of CCT has been found to reduce the binding afﬁnity of target protein to the second ring by two orders of magnitude (Dobrzynski et al., 2000). These structural and biochemical data directly contradict the claim (based on a quantitative analysis of labeled β-actin or β-tubulin bound to CCT) that both rings of CCT are occupied in the target protein/CCT binary complex (Melki et al., 1997). It has been suggested that the upward and outward movement of apical domains that accompanies target protein binding may be part of a mechanism of inter-ring signaling that prevents target protein binding in the opposite ring. The subunits within the CCT ring responsible for binding to α-actin target protein were identiﬁed by reacting binary complexes formed between CCT and either full-length actin or subdomains derived from it with subunit-speciﬁc monoclonal antibodies. This analysis showed that actin binds to CCT via two speciﬁc interactions: In either case, the small domain binds to CCTδ, but the large domain binds to either CCTβ or CCTε (Fig. 6). Because the organization of subunits within the CCT ring is unique, it follows that the binding of actin to CCT is both subunit-speciﬁc and geometry-dependent (Llorca et al., 1999a). This subunit-speciﬁc binding of actin to CCT has been corroborated by biochemical experiments in which 35S-labeled β-actin generated by translation in vitro was subjected to immunoprecipitation with subunit-speciﬁc anti-CCT antibodies using buffer conditions that disrupt CCT into its constituent monomers. The interactions observed, i.e., between CCTα, CCTβ, CCTε, and CCTθ, are to at least some extent consistent with the interactions deduced from cryoelectron microscopy of apo-CCT–α-actin (Hynes and Willison, 2000). An attempt has been made to gain insight into those domains within actin and α- and β-tubulin that interact with CCT by testing the ability of truncated forms of these proteins to bind to the chaperonin. Three hydrophobic domains within actin, corresponding to residues 125–179, 244–285, and 340–375, were found to interact with CCT in a cooperative manner. A region within α- and β-tubulins is also apparently involved in recognition by CCT, and contains a distribution of hydrophobic residues that is similar to region 244–285 in actin (Rommelaere et al., 1999). In an alternative approach, a β-actin peptide array was screened for CCT-binding sequences; these experiments implicated three surface regions rich in charged and polar residues in binding to CCT (Hynes

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

87

FIG. 6. Modes of interaction of actin with CCT. (A–H) Two-dimensional average images of different classes of CCT/α-actin (A, B, E, F) or α-actin subdomain 4 immunocomplexes with an anti-CCT-δ antibody (A–D) or an anti-CCT-α antibody (E–H). (I) Structure of actin showing the large and small domains and their division into four subdomains (subs 1–4). ( J) Two possible modes of interaction of actin with CCT. Reprinted from Llorca et al. (1999a), with permission.

and Willison, 2000). Taken together, the structural and biochemical data point to a relatively constrained (rather than promiscuous) mode of target protein recognition by CCT, consistent with the limited number of known target proteins for this chaperonin.

B. Structure of the Thermosome The structure of the thermosome from T. acidophilum has recently been reported (Ditzel et al., 1998), as has that of the isolated apical domain of its α subunit (Klumpp et al., 1997). The intact thermosome is spherical, with a diameter of 16.4 nm and a height of 15.8 nm

88

NICHOLAS J. COWAN AND SALLY A. LEWIS

(Fig. 7, see color insert). The data conﬁrm the overall similarity in domain structure between Type I and Type II chaperonins, but there are two striking differences that distinguish them (reviewed in Gutsche et al., 1999). First, though the intra-ring contacts closely resemble those in Type I chaperonins, the inter-ring contacts differ betwen the two groups: In the thermosome, the inter-ring contacts are such that the subunits form α–α and β–β pairs. Second, the apical domains of the thermosome differ from GroEL in that they contain an insertion of 28 amino acids that form a unique 25 A˚ helical protrusion. This feature is common to all Type II chaperonins, and provides a relatively well conserved hydrophobic surface patch that is consistent with a role in the binding of target protein. In addition, these helical protrusions are capable of modulating access of target proteins to the central cavity in a manner that parallels the closure of the GroEL central cavity by the binding of GroES. This built-in feature of Type II chaperonins presumably obviates the requirement for a GroES-like homolog in Type II chaperonin-mediated folding reactions. VIII. PREFOLDIN Prefoldin is a chaperone protein that was discovered in two different ways. In one approach, the S. cerevisiae homolog of prefoldin, termed GIM (for genes involved in microtubule biogenesis), was identiﬁed in a genetic screen for mutants that were synthetically lethal with tub4-1, which encodes a mutated yeast γ -tubulin (Geissler et al., 1998). The ﬁve gene products identiﬁed in this screen, termed GIM1-5, were shown by coimmunoprecipitation to exist in a single complex. Disruption of any of these genes resulted in a similar phenotype, i.e., sensitivity to cold and supersensitivity to the microtubule-depolymerizing drug benomyl, thus implicating the GIM complex in tubulin biogenesis (Fig. 8). Independently, prefoldin was isolated biochemically by virtue of its ability to bind unfolded actin, with which it forms a relatively stable binary complex that can be detected on nondenaturing gels. This latter property allowed the conventional puriﬁcation of prefoldin from a soluble extract of bovine testis tissue (Vainberg et al., 1998). The puriﬁed protein migrates as a symmetrical peak with an apparent molecular mass of about 200 kDa on gel ﬁltration, and yields a single band upon nondenaturing gel electrophoresis. However, on denaturing gel electrophoresis, prefoldin is resolved into six polypeptides in the molecular mass range 14– 25 kDa, termed prefoldin 1–6. Each polypeptide is present in the complex in equal ratios. Sequence comparisons of prefoldin subunits showed that prefoldin homologs exist throughout eukarya and in the archaea. The archaeal homolog of prefoldin from Methanobacterium thermoautotrophicum, MtGimC, is a hexamer of 87 kDa and consists of two α and

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

89

FIG. 8. Phenotype of yeast cells lacking prefoldin subunit pfd5. The actin cytoskeleton is largely conﬁned to small cortical patches; a large proportion (examples highlighted by arrows) of pfd5 mutant cells fails to segregate their nuclei to the bud upon cell division. Reprinted from Vainberg et al. (1998), with permission.

four β subunits each of which is predicted to contain extended coiled coils. Reconstitution experiments with the archaeal homolog suggested that the two α subunits form a core dimer onto which the four β subunits assemble. The two archaeal subunits can form heterocomplexes with yeast Gim subunits, and MtGimβ can partially complement yeast mutants that lack Gim1 and 4 (Leroux et al., 1999). The evolutionary conservation of prefoldin implies that this chaperone might play a general role in protein folding. This expectation was borne out by functional analyses, as described below.

A. Biochemistry Prefoldin binds to unfolded, but not to native, β-actin (Vainberg et al., 1998). This interaction, which results in the formation of a relatively stable β-actin/prefoldin binary complex, is nucleotide-independent. When the prefoldin/β-actin binary complex is incubated with CCT, there is a rapid and nucleotide-independent transfer of the target protein to the chaperonin, without the formation of a stable prefoldin/β-actin/CCT ternary complex. The nucleotide-independent transfer of β-actin target protein between prefoldin and CCT takes place efﬁciently even in

90

NICHOLAS J. COWAN AND SALLY A. LEWIS

the presence of an excess of a trap (Hsp60) designed to capture any non-native target molecules that might be released into solution. These data imply that the transfer of β-actin from prefoldin to the chaperonin occurs via a direct mechanism that does not involve the discharge of target protein into solution. Even when this reaction is repeated in the presence of ATP, in which case Hsp60 will act as a competitive inhibitor of folding, prefoldin-bound β-actin is transferred to CCT and partitions to the native state as efﬁciently as in a parallel experiment lacking Hsp60 (Vainberg et al., 1998). This experiment demonstrated the functional importance of prefoldin in transferring folding intermediates speciﬁcally to CCT. By the same criteria, both α- and β-tubulin are targets for binding to prefoldin, consistent with the genetic evidence described above. The efﬁcient transfer of target protein from prefoldin to CCT even in the presence of a trap implied that prefoldin might be interacting directly with the chaperonin. This indeed turns out to be the case: When prefoldin is passed over a column of CCT bound to ATP-agarose, the appearance of prefoldin in the efﬂuent is signiﬁcantly delayed, and a large proportion remains bound to the column in comparison to a parallel control column containing bound GroEL, with which prefoldin does not interact (Vainberg et al., 1998) (Fig. 9). In support of this ﬁnding, about 10% of cytosolic GimC (yeast prefoldin) can be coimmunoprecipitated with CCT (Siegers et al., 1999). The interaction between CCT and prefoldin is apparently not strong enough to allow the puriﬁcation of the complex. B. Role of Prefoldin In Vivo The experiments described above establish a general role for prefoldin in facilitated protein folding: In the crowded molecular environment that exists within living cells, the binding of polypeptides to prefoldin promotes productive folding by escorting them to CCT. This conclusion implies that binding of target protein to prefoldin should precede binding to CCT. A careful kinetic analysis of in vitro actin translation reactions shows that this is indeed the case (Hansen et al., 1999) (Fig. 10). When β-actin is translated for a brief period and further translational initiation is blocked by the addition of edeine and 7-MeGMP, the ﬁrst β-actin–containing species that appear are complexed with prefoldin. Furthermore, prefoldin binds stably to nascent β-actin chains when they contain on the order of 145 amino acids; under these conditions, the β-actin is in a relatively unfolded conformation as judged by its resistance to proteolytic digestion. Prefoldin is found associated with actin-translating polyribosomes and remains bound to the target

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

91

FIG. 9. Prefoldin binds to CCT, but not to GroEL. (A, B) Columns of ATP-agarose were loaded with either GroEL (A) or CCT (B). Prefoldin was applied to these columns together with bovine serum albumin (included as an internal control), and the column efﬂuents monitored by SDS–PAGE and by Western blotting with an anti-prefoldin antibody (C). Note the delayed appearance of prefoldin in (B) compared with (A), and the recovery of prefoldin in the eluate from the ATP-agarose/CCT column, but not in the eluate from the ATP-agarose/GroEL column (C). Reprinted from Vainberg et al. (1998), with permission.

protein until it is delivered to CCT (Hansen et al., 1999). Consistent with these data, an analysis of the actin folding pathway in yeast showed that newly synthesized actin is not captured by an introduced trap, so is efﬁciently channeled by GimC (prefoldin) from the ribosome directly to CCT. Only when GimC (prefoldin) expression is ablated is the trap effective (Siegers et al., 1999).

FIG. 10. Kinetic analysis of β-actin translation in rabbit reticulocyte lysate. Triangles: β-actin bound to prefoldin; closed circles: β-actin bound to CCT; squares: native monomeric β-actin; open circles: β-actin bound to an unknown protein. Reprinted from Hansen et al. (1999), with permission.

92

NICHOLAS J. COWAN AND SALLY A. LEWIS

Whether CCT as well as prefoldin binds to ribosome-bound nascent polypeptide chains is a matter of controversy. There are several reports that it does (Frydman and Hartl, 1996; Frydman et al., 1994; McCallum et al., 2000; Thulasiraman et al., 1999), but these are vigorously disputed (Eggers et al., 1997) on the grounds that the immunoprecipitation reactions used to demonstrate the interaction were insufﬁciently well controlled. In any event, this does not seem to be a vital issue, since CCT can probably productively fold only full-length actin and tubulin. In vitro experiments suggest that prefoldin functions not only to deliver newly synthesized polypeptides to CCT, but to direct cycling folding intermediates back to CCT (Vainberg et al., 1998) and possibly to direct stressed proteins back to CCT for refolding. These roles for prefoldin have yet to be established in vivo. Because several lines of evidence point to a restricted target range for CCT, it seems plausible that the target range of prefoldin is equally constrained. It has been shown that α- and β-tubulins follow maturation pathways similar to β-actin upon translation in vitro (Hansen et al., 1999; Vainberg et al., 1998), but as yet there are no well-studied interactions between prefoldin and noncytoskeletal target proteins. We assume that prefoldin functions by targeting nascent chains that are obligate substrates for chaperonin-mediated folding to CCT. This conclusion is consistent with the phenotypes of S. cerevisiae cells harboring deletions of one or more prefoldin subunits: In each case, the cells show impaired functions of the actin- and tubulin-based cytoskeleton (Siegers et al., 1999; Vainberg et al., 1998). These cells behave as if their CCT were impaired, showing that prefoldin acts in vivo as an enhancer of CCT function. These observations and the fact that prefoldin is found in all organisms (including the archaea, which have a Type II chaperonin) suggest that it has an ancient and important role in assisting the chaperonin. However, many archaea lack an hsp70/dnaK homolog; in these organisms it seems likely that prefoldin additionally fulﬁlls the typical hsp70 role as the main chaperone for ribosome-bound nascent polypeptide chains. To what extent the functions of prefoldin and hsp70 overlap in other organisms could be tested genetically, but so far these experiments have not been done. IX. TUBULIN-SPECIFIC CHAPERONES Both α- and β-tubulin are GTP-binding proteins: α-Tubulin binds GTP nonexchangeably and does not hydrolyze its bound nucleotide, while the binding of GTP to β-tubulin is exchangeable, and hydrolysis of its

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

93

GTP is coupled to the polymerization of the α/β heterodimer into the microtubule polymer (Mitchison and Kirschner, 1986). As in the case of actin, tubulins are obligate substrates for CCT: It has never proved possible to obtain either native actin or tubulin following their denaturation. Although denatured actin can be folded to the native state by interaction with CCT alone in the presence of ATP (Gao et al., 1992), early in vitro folding experiments showed that this is not the case for either α- or β-tubulin. Neither inclusion of GTP in CCT-mediated tubulin folding reactions nor the addition of native carrier tubulin heterodimers as a potential reservoir with which newly folded heterodimers might exchange (but see Section IX. A.1) remedies this failure. Nonetheless, both α- and β-tubulin target proteins efﬁciently form a binary complex with CCT, and each undergoes multiple cycles of ATP-dependent binding and release from the chaperonin without partitioning to the native state (Gao et al., 1994). On the other hand, both α- and β-tubulin can be translated in vitro in reticulocyte lysate—a relatively rich source of CCT— so as to yield polymerization-competent heterodimers (Cleveland et al., 1978). It therefore seemed likely that one or more additional components were present in lysate that were required for the proper folding of the tubulin heterodimer (Zabala and Cowan, 1992). Furthermore GTP hydrolysis was shown to be essential for tubulin dimer formation in in vitro translation reactions (Fontalba et al., 1993). Indeed, it transpired that the addition of two crude cellular fractions (distinguished by their elution characteristics upon anion exchange chromatography), plus native tubulin heterodimers, ATP and GTP, was required in order to generate polymerization-competent tubulin heterodimers in CCTcontaining in vitro folding reactions (Gao et al., 1993). Further fractionation of these crude fractions ultimately deﬁned the components that participate in the de novo formation of tubulin heterodimers (Tian et al., 1996; Tian et al., 1997). This pathway, which occurs downstream of CCT and is essential for the de novo formation of properly folded tubulin heterodimers, is described below. A. The Tubulin Folding Pathway As a result of one or more cycles of interaction with CCT, α-tubulin acquires a quasi-native state (termed IQ for intermediates, quasi-native) deﬁned by its relative resistance to proteolytic digestion (compared with the same target protein bound to CCT in the absence of ATP) and by the fact that, like native α-tubulin, it contains nonexchangeably bound GTP (Tian et al., 1995a). Presumably, a similar kind of intermediate is produced in the case of β-tubulin, but in neither case are these folding intermediates competent to form native tubulin heterodimers. Rather,

94

NICHOLAS J. COWAN AND SALLY A. LEWIS

a total of ﬁve tubulin-speciﬁc chaperones, known as cofactors A–E, have been shown to participate in the post-CCT formation of the tubulin heterodimer. Fortunately, in many cases, interaction of one or more cofactors with IQ intermediates results in the generation of a relatively stable species with a characteristic mobility on nondenaturing gels. The generation of these intermediates could therefore be used to track many of these cofactors through multiple conventional puriﬁcation dimensions, eventually resulting in their isolation as homogeneous proteins (Tian et al., 1996; Tian et al., 1997). The overall scheme whereby tubulin heterodimers are formed de novo is outlined in Figure 11 (see color insert) (Bhamidipati et al., 2000; Cowan and Lewis, 1999). Quasi-native intermediates generated as a result of ATP-dependent interaction with CCT are captured and stabilized by cofactors A or D (in the case of β-tubulin) or by cofactors B or E (in the case of α-tubulin). Eα and Dβ interact with each other, forming a multimolecular complex; C enters this complex, forming a supercomplex that releases native tubulin heterodimers upon GTP hydrolysis. The GTP-hydrolyzing entity in the supercomplex has been shown to be β-tubulin itself, by several criteria. First, none of the cofactors contains recognizable GTP binding site motifs within their amino acid sequences. Second, the cofactors do not bind labeled GTP in vitro. Third, generation of the supercomplex in the presence of [α-32P]GTP followed by UV crosslinking and analysis of the products by SDS PAGE shows β-tubulin to be the only radioactive species (Tian et al., 1997). Finally, the hydrolytic reaction proceeds with equal efﬁciency when the GTP analogs dGTP or ddGTP are substituted for GTP in the folding reaction; the ability of β-tubulin to hydrolyze these analogs as efﬁciently as GTP is a characteristic of this GTPase (Tian et al., 1999). GTP hydrolysis thus accompanies the release of native heterodimers from the supercomplex. We can therefore think of GTP hydrolysis acting as a switch for the release of heterodimers from the supercomplex if cofactors have a much lower afﬁnity for GDP-tubulin than for GTP-tubulin. FIG. 7. Side view of the hexadecameric thermosome from T. acidophilum (A), compared with the side view of the GroEL–GroES–(ADP)7 complex (B). Domains are colored in red (equatorial), green (intermediate), and yellow (apical). Within each complex, domains of aligned subunits are highlighted in blue (equatorial), light blue (intermediate), and violet (apical). Bound ADP is in yellow. Reprinted from Ditzel et al. (1998), with permission. FIG. 11. The tubulin folding pathway. CCT is shown in orange, prefoldin is in yellow, and cofactors A–E are denoted by red letters. Polymerization of native, heterodimerized GTP-tubulin into a microtubule is depicted at the right. Reprinted from Bhamidipati et al. (2000), with permission.

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

95

This notion is borne out experimentally: When tubulin and cofactor D are mixed with equal quantities of [α-32P]GTP and [α-32P]GDP, only labeled GTP is found in association with the complexes formed (Tian et al., 1999). The fact that tubulin must hydrolyze GTP for it to be released from chaperones suggests that this reaction might function as a quality-control step. The continuous monitoring (through folding and refolding) of tubulin’s capacity to properly hydrolyze GTP might serve to protect the cell from the consequences of hydrolysis-incompetent tubulin, whose incorporation into microtubules might otherwise have drastic physiologic consequences even at low levels (Lewis et al., 1997; Tian et al., 1999). This idea is supported by experiments with putative GTPase-defective β-tubulin mutants which show that these mutant proteins are neither incorporated into microtubules in vivo nor released from cofactor complexes in vitro (Zabala et al., 1996). In addition to participating in the generation of de novo tubulin heterodimers, cofactors D and E can interact directly with the native heterodimer. When they do so, the dimer is disrupted: cofactor D binds the β-subunit, while cofactor E binds to the α-subunit. In either case, the remaining subunit, bereft of its partner, rapidly decays (within a time scale of a few minutes at 30◦ C) to a non-native state that is no longer competent to form heterodimers and that is capturable by chaperonin (Tian et al., 1997). This result illustrates the fact that neither α- nor β-tubulin can exist in a stable native form in mutual isolation and explains why it has never proved possible to isolate either α- or β-tubulin as individual entities. It also goes some way toward explaining the function of the tubulin cofactors: Their interaction with CCT-generated folding intermediates is required to stabilize them in a nonminimal energy state and bring them together in such a way as to form the native, stable heterodimer. That the cofactors act as a heterodimerization machine for tubulin is established by experiments showing that tubulin heterodimerization cannot occur spontaneously (Tian et al., 1999). The ability of cofactors D and E to disrupt the tubulin heterodimer in vitro implies that overexpression of these cofactors in vivo might have a disruptive effect on the microtubule cytoskeleton. This expectation is borne out experimentally: Cells overexpressing either of these cofactors lose most or all of their microtubules (Bhamidipati et al., 2000; Martin et al., 2000). It makes sense, therefore, that regulatory mechanisms should exist to ensure the balanced availability of these cofactors within living cells. Interaction of cofactors with native tubulin is also regulated by the small G protein Arl2 (Bhamidipati et al., 2000). Of the ﬁve cofactors that are participants in the generation of tubulin heterodimers, structural information is thus far available for only one,

96

NICHOLAS J. COWAN AND SALLY A. LEWIS

the S. cerevisiae homolog of cofactor A, Rbl2. This protein is so named because its overexpression rescues yeast cells from the otherwise lethal overexpression of β-tubulin (rescue β-tubulin lethality) (Archer et al., 1995), a fact that may reﬂect a function in maintaining the balance of available β-tubulin by storing it as a complex with cofactor A (Fanarraga et al., 1999). The crystal structure reveals that Rbl2 exists as a slightly convex molecule and has a surface that is predominantly hydrophilic in character (Steinbacher, 1999). This distinguishes it from the majority of chaperone proteins, which present relatively hydrophobic patches that recognize and bind to unfolded proteins. Rbl2 and mammalian cofactor A share only about 30% amino acid sequence identity; nevertheless, the fact that cofactor A can restore viability to yeast cells overexpressing β-tubulin (Archer et al., 1995) suggests that the yeast and mammalian proteins share a high degree of structural homology (Cowan and Lewis, 1999).

B. Tubulin Cofactors in S. cerevisiae and S. pombe Tubulins are among the most highly conserved proteins known, and it is therefore not surprising that homologs of the cofactors that participate in the biogenesis of heterodimers in mammals should exist in yeast cells (Table I). On the other hand, it is somewhat surprising that, while TABLE I Tubulin-Specific Chaperone Proteins Mammalian protein

S. cerevisae homolog

S. pombe homolog

Cofactor A (Gao et al., 1994; Melki et al., 1996) Cofactor B (Tian et al., 1997)

Rbl2p (Archer et al., 1998; Archer et al., 1995) Alf1p (Feierbach et al., 1999; Tian et al., 1997) Cin2p (Hoyt et al., 1997; Hoyt et al., 1990; Stearns et al., 1990) Cin1p (Hoyt et al., 1997; Hoyt et al., 1990; Stearns et al., 1990) Pac2p (Hoyt et al., 1997)

Alp31p (Radcliffe et al., 2000)

Cofactor C (Tian et al., 1996)

Cofactor D (Tian et al., 1996)

Cofactor E (Tian et al., 1996)

Arl2 (regulatory protein) (Bhamidipati et al., 2000)

Cin4p (Hoyt et al., 1997; Hoyt et al., 1990; Stearns et al., 1990)

Alp11p (Radcliffe et al., 1999; Radcliffe and Toda, 2000) Unknown

Alp1p (Hirata et al., 1998)

Alp21p (Radcliffe et al., 1999) Sto1p (Grishchuk and McIntosh, 1999) Alp41p (Radcliffe et al., 2000)

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

97

mammalian cofactors are essential for the assembly of new heterodimers in vitro, deletion of the corresponding genes is not lethal in S. cerevisiae. However, mutations in these genes do result in microtubule phenotypes such as supersensitivity to the microtubule poison benomyl or the depolymerizing drug nocodazole, as well as to chromosome instability (which is characterized by the unstable inheritance of minichromosomes and chomosome fragments). Since the cofactors are not essential in S. cerevisiae, but the tubulin heterodimer is, it is clear that tubulin in this organism must be able to form heterodimers in the absence of cofactors. One possible explanation for this is that the sequences of the S. cerevisiae tubulins are the most divergent known. Saccharomyces cerevisiae cells contain only relatively simple microtubule arrays that participate in only three well deﬁned and temporally distinct cellular process: mitosis, meiosis, and nuclear fusion. Perhaps this reduced selective pressure on their tubulin genes has allowed the evolution of spontaneously dimerizing tubulin in this organism. In contrast to S. cerevisiae, the microtubule cytoskeleton in the ﬁssion yeast Schizosacaromyces pombe is more complex and the amino acid sequences of the tubulin polypeptides are closer to mammalian sequences. The ﬁssion yeast homologs of cofactors B, D, and E are essential for survival in this organism. It seems highly likely that the cofactors are also essential for life in ciliates, ﬂagellates, and multicellular organisms, in which the sequence of tubulins are also highly conserved. C. Postfolding Functions of These Chaperones As well as acting in de novo tubulin heterodimerization (see Section IX), some of the tubulin-speciﬁc chaperones also act together as a GTPase activator (GAP) for native tubulin (Tian et al., 1999). Although none of the tubulin-speciﬁc chaperone proteins (cofactors) has any measurable intrinsic GTPase activity, when native tubulin and either cofactors C and D or tubulin and cofactors C, D, and E are mixed, GTP hydrolysis ensues. In this reaction, the cofactors behave with Michaelis–Menton kinetics, with GTP-tubulin as substrate and Pi as product. The cofactors have a K m for tubulin of about 0.1 μm, which is about 200-fold lower than the critical concentration for polymerization of heterodimers into microtubules. It follows, therefore, that GTP hydrolysis in this reaction is not a result of cofactors acting as microtubule associated proteins and promoting microtubule polymerization with its associated GTP hydrolysis. Rather, GTP hydrolysis by β-tubulin in this reaction is a result of the direct interaction of cofactors with native tubulin dimers

98

NICHOLAS J. COWAN AND SALLY A. LEWIS

(Tian et al., 1999). This GAP activity is clearly the same reaction that releases tubulin heterodimer from the cofactor-containing supercomplex (see above and Fig. 11). The difference is that in the latter case the input for the reaction is chaperoned tubulin subunits and in the former case it is native heterodimers. The cofactor GAP activity converts GTP-tubulin, which is polymerization-competent, to GDP-tubulin, which is not. The fact that this reaction appears to be regulated by the small G protein Arl2 (Bhamidipati et al., 2000) suggests that it could be used by the cell for the spatial or temporal control of microtubule dynamics. However, it may prove difﬁcult to dissect out the folding and postfolding activities of these chaperones in vivo. X. CONCLUSION Many of the experiments described in this chapter give us a qualitative feel for how Type II chaperonins might facilitate protein folding. Imaging of CCT/actin complexes and experiments on actin and tubulin folding in vitro suggest that target proteins bind to chaperonin in a surprisingly native-like state, probably consisting of only one or a few kinetically trapped folding intermediates. The trapped intermediate changes conformation by virtue of the binding of different domains to different subunits of CCT, and possibly also through the force transmitted to the target protein by allosteric changes in CCT as ATP is exchanged and/or hydrolyzed. Target protein is then released and folding proceeds. In the case of actin, CCT binding may serve to open up an actin folding intermediate (Schuler et al., 2000) to allow the binding of ATP (Llorca et al., 1999a). In the case of tubulin, the folding intermediates probably already have bound nucleotide (Tian et al., 1995a). Our task for the future is to prove or disprove these ideas in a rigorous and quantitative way. Other major goals include imaging the interaction of prefoldin with chaperonin, establishing the range of target proteins for these two chaperones and delineating the functions of their archaeal homologs. REFERENCES Archer, J. E., M. Magendantz, L. R. Vega, and F. Solomon (1998). Formation and function of the Rbl2p-beta-tubulin complex. Mol. Cell. Biol. 18, 1757–1762. Archer, J. E., L. R. Vega, and F. Solomon (1995). Rbl2p, a yeast protein that binds to beta-tubulin and participates in microtubule function in vivo. Cell 82, 425–434. Bhamidipati, A., S. A. Lewis, and N. J. Cowan (2000). ADP ribosylation factor-like protein 2 (Arl2) regulates the interaction of tubulin-folding cofactor D with native tubulin. J. Cell. Biol. 149, 1087–1096.

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

99

Bochkareva, E. S., N. M. Lissin, and A. S. Girshovich (1988). Transient association of newly synthesized unfolded proteins with the heat-shock GroEL protein. Nature 336, 254–257. Bukau, B., E. Deuerling, C. Pfund, and E. A. Craig (2000). Getting newly synthesized proteins into shape. Cell 101, 119–122. Bukau, B., T. Hesterkamp, and J. Luirink (1996). Growing up in a dangerous environment: A network of multiple targeting and folding pathways for nascent polypeptides in the cytosol. Trends Cell Biol. 6, 480–486. Bukau, B., and A. L. Horwich (1998). The Hsp70 and Hsp60 chaperone machines. Cell 92, 351–380. Chen, X., D. S. Sullivan, and T. C. Huffaker (1994). Two yeast genes with similarity to TCP-1 are required for microtubule and actin function in vivo. Proc. Natl. Acad. Sci. USA 91, 9111–9115. Cheng, M. Y., F. U. Hartl, J. Martin, R. A. Pollock, F. Kalousek, W. Neupert, E. M. Hallberg, R. L. Hallberg, and A. L. Horwich (1989). Mitochondrial heat-shock protein hsp60 is essential for assembly of proteins imported into yeast mitochondria. Nature 337, 620–625. Cleveland, D. W., M. W. Kirschner, and N. J. Cowan (1978). Isolation of separate mRNAs for alpha- and beta-tubulin and characterization of the corresponding in vitro translation products. Cell 15, 1021–1031. Cowan, N. J. (1998). Mammalian cytosolic chaperonin. Methods Enzymol. 290, 230–241. Cowan, N. J., and S. A. Lewis (1999). A chaperone with a hydrophilic surface. Nature Struct. Biol. 6, 990–991. Ditzel, L., J. Lowe, D. Stock, K. O. Stetter, H. Huber, R. Huber, and S. Steinbacher (1998). Crystal structure of the thermosome, the archaeal chaperonin and homolog of CCT. Cell 93, 125–138. Dobrzynski, J. K., M. L. Sternlicht, I. Peng, G. W. Farr, and H. Sternlicht (2000). Evidence that beta-tubulin induces a conformation change in the cytosolic chaperonin which stabilizes binding: Implications for the mechanism of action. Biochemistry 39, 3988– 4003. Eggers, D. K., W. J. Welch, and W. J. Hansen (1997). Complexes between nascent polypeptides and their molecular chaperones in the cytosol of mammalian cells. Mol. Biol. Cell 8, 1559–1573. Esser, K., B. Scholle, and G. Michaelis (1999). Disruption of six open reading frames on chromosome X of Saccharomyces cerevisiae reveals a cluster of four essential genes. Yeast 15, 921–933. Fanarraga, M. L., M. Parraga, K. Aloria, J. del Mazo, J. Avila, and J. C. Zabala (1999). Regulated expression of p14 (cofactor A) during spermatogenesis. Cell Motil. Cytoskeleton 43, 243–254. Farr, G. W., E. C. Scharl, R. J. Schumacher, S. Sondek, and A. L. Horwich (1997). Chaperonin-mediated folding in the eukaryotic cytosol proceeds through rounds of release of native and nonnative forms. Cell 89, 927–937. Feierbach, B., E. Nogales, K. H. Downing, and T. Stearns (1999). Alf1p, a CLIP-170 domain-containing protein, is functionally and physically associated with alphatubulin. J. Cell. Biol. 144, 113–124. Feldman, D. E., and J. Frydman (2000). Protein folding in vivo: The importance of molecular chaperones [see comments]. Curr. Opin. Struct. Biol. 10, 26–33. Feldman, D. E., V. Thulasiraman, R. G. Ferreyra, and J. Frydman (1999). Formation of the VHL–elongin BC tumor suppressor complex is mediated by the chaperonin TRiC. Mol. Cell 4, 1051–1061.

100

NICHOLAS J. COWAN AND SALLY A. LEWIS

Fontalba, A., R. Paciucci, J. Avila, and J. C. Zabala (1993). Incorporation of tubulin subunits into dimers requires GTP hydrolysis. J. Cell Sci. 106, 627–632. Frydman, J., and F. U. Hartl (1996). Principles of chaperone-assisted protein folding: Differences between in vitro and in vivo mechanisms [see comments]. Science 272, 1497–1502. Frydman, J., E. Nimmesgern, H. Erdjument-Bromage, J. S. Wall, P. Tempst, and F. U. Hartl (1992). Function in protein folding of TRiC, a cytosolic ring complex containing TCP-1 and structurally related subunits. EMBO J. 11, 4767–4778. Frydman, J., E. Nimmesgern, K. Ohtsuka, and F. U. Hartl (1994). Folding of nascent polypeptide chains in a high molecular mass assembly with molecular chaperones [see comments]. Nature 370, 111–117. Furutani, M., T. Iida, T. Yoshida, and T. Maruyama (1998). Group II chaperonin in a thermophilic methanogen, Methanococcus thermolithotrophicus. Chaperone activity and ﬁlament-forming ability. J. Biol. Chem. 273, 28399–28407. Gao, Y., R. Melki, P. D. Walden, S. A. Lewis, C. Ampe, H. Rommelaere, J. Vandekerckhove, and N. J. Cowan (1994). A novel cochaperonin that modulates the ATPase activity of cytoplasmic chaperonin. J. Cell Biol. 125, 989–996. Gao, Y., J. O. Thomas, R. L. Chow, G. H. Lee, and N. J. Cowan (1992). A cytoplasmic chaperonin that catalyzes beta-actin folding. Cell 69, 1043–1050. Gao, Y., I. E. Vainberg, R. L. Chow, and N. J. Cowan (1993). Two cofactors and cytoplasmic chaperonin are required for the folding of alpha- and beta-tubulin. Mol. Cell. Biol. 13, 2478–2485. Geissler, S., K. Siegers, and E. Schiebel (1998). A novel protein complex promoting formation of functional alpha- and gamma-tubulin. EMBO J. 17, 952–966. Goloubinoff, P., J. T. Christeller, A. A. Gatenby, and G. H. Lorimer (1989a). Reconstitution of active dimeric ribulose bisphosphate carboxylase from an unfolded state depends on two chaperonin proteins and Mg-ATP. Nature 342, 884–889. Goloubinoff, P., A. A. Gatenby, and G. H. Lorimer (1989b). GroE heat-shock proteins promote assembly of foreign prokaryotic ribulose bisphosphate carboxylase oligomers in Escherichia coli. Nature 337, 44–47. Grishchuk, E. L., and J. R. Mclntosh (1999). Sto 1p, a ﬁssion yeast protein similar to tubulin folding cofactor E, plays an essential role in mitotic microtubule assembly. J. Cell Sci. 112, 1979–1988. Guagliardi, A., L. Cerchia, S. Bartolucci, and M. Rossi (1994). The chaperonin from the archaeon Sulfolobus solfataricus promotes correct refolding and prevents thermal denaturation in vitro. Protein Sci. 3, 1436–1443. Gupta, R. S. (1990). Sequence and structural homology between a mouse T-complex protein TCP-1 and the ‘chaperonin’ family of bacterial (GroEL, 60–65 kDa heat shock antigen) and eukaryotic proteins. Biochem. Int. 20, 833–841. Gutsche, I., L. O. Essen, and W. Baumeister (1999). Group II chaperonins: New TRiC(k)s and turns of a protein folding machine. J. Mol. Biol. 293, 295–312. Hansen, W. J., N. J. Cowan, and W. J. Welch (1999). Prefoldin-nascent chain complexes in the folding of cytoskeletal proteins. J. Cell. Biol. 145, 265–277. Hemmingsen, S. M., C. Woolford, S. M. van der Vies, K. Tilly, D. T. Dennis, C. P. Georgopoulos, R. W. Hendrix, and R. J. Ellis (1988). Homologous plant and bacterial proteins chaperone oligomeric protein assembly. Nature 333, 330–334. Hendrix, R. W. (1979). Puriﬁcation and properties of groE, a host protein involved in bacteriophage assembly. J. Mol. Biol. 129, 375–392. Hirata, D., H. Masuda, M. Eddison, and T. Toda (1998). Essential role of tubulin-folding cofactor D in microtubule assembly and its association with microtubules in ﬁssion yeast. EMBO J. 17, 658–666.

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

101

Hohn, T., B. Hohn, A. Engel, M. Wurtz, and P. R. Smith (1979). Isolation and characterization of the host protein groE involved in bacteriophage lambda assembly. J. Mol. Biol. 129, 359–373. Horwich, A. L., and K. R. Willison (1993). Protein folding in the cell: Functions of two families of molecular chaperone, hsp 60 and TF55-TCP1. Phil. Trans. R. Soc. Lond. B, Biol. Sci. 339, 313–325; discussion 325–316. Hoyt, M. A., J. P. Macke, B. T. Roberts, and J. R. Geiser (1997). Saccharomyces cerevisiae PAC2 functions with CIN1, 2 and 4 in a pathway leading to normal microtubule stability. Genetics 146, 849–857. Hoyt, M. A., T. Stearns, and D. Botstein (1990). Chromosome instability mutants of Saccharomyces cerevisiae that are defective in microtubule-mediated processes. Mol. Cell Biol. 10, 223–234. Hynes, G., C. W. Sutton, S. U, and K. R. Willison (1996). Peptide mass ﬁngerprinting of chaperonin-containing TCP-1 (CCT) and copurifying proteins. FASEB J. 10, 137– 147. Hynes, G. M., and K. R. Willison (2000). Individual subunits of the eukaryotic cytosolic chaperonin mediate interactions with binding sites located on subdomains of betaactin. J. Biol. Chem. 275, 18985–18994. Kim, S., K. R. Willison, and A. L. Horwich (1994). Cystosolic chaperonin subunits have a conserved ATPase domain but diverged polypeptide-binding domains. Trends Biochem. Sci. 19, 543–548. Klumpp, M., W. Baumeister, and L. O. Essen (1997). Structure of the substrate binding domain of the thermosome, an archaeal group II chaperonin. Cell 91, 263–270. Kubota, H., G. Hynes, A. Carne, A. Ashworth, and K. Willison (1994). Identiﬁcation of six Tcp-1-related genes encoding divergent subunits of the TCP-1-containing chaperonin. Curr. Biol. 4, 89–99. Kubota, H., G. Hynes, and K. Willison (1995a). The chaperonin containing t-complex polypeptide 1 (TCP-1). Multisubunit machinery assisting in protein folding and assembly in the eukaryotic cytosol. Eur. J. Biochem. 230, 3–16. Kubota, H., G. Hynes, and K. Willison (1995b). The eighth Cct gene, Cctq, encoding the theta subunit of the cytosolic chaperonin containing TCP-1. Gene 154, 231–236. Kubota, H., S. Matsumoto, S. Yokota, H. Yanagi, and T. Yura (1999a). Transcriptional activation of mouse cytosolic chaperonin CCT subunit genes by heat shock factors HSF1 and HSF2. FEBS Lett. 461, 125–129. Kubota, H., S. Yokota, H. Yanagi, and T. Yura (1999b). Structures and co-regulated expression of the genes encoding mouse cytosolic chaperonin CCT subunits. Eur. J. Biochem. 262, 492–500. Kubota, H., S. Yokota, H. Yanagi, and T. Yura (2000). Transcriptional regulation of the mouse cytosolic chaperonin subunit gene Ccta/Tcp-1 by Staf family zinc ﬁnger proteins. J. Biol. Chem. 275, 28641–28648. Leroux, M. R., M. Fandrich, D. Klunker, K. Siegers, A. N. Lupas, J. R. Brown, E. Schiebel, C. M. Dobson, and F. U. Hartl (1999). MtGimC, a novel archaeal chaperone related to the eukaryotic chaperonin cofactor GimC/prefoldin. EMBO J. 18, 6730–6743. Lewis, S. A., G. Tian, and N. J. Cowan (1997). The alpha and beta tubulin folding pathways. Trends Cell Biol. 7, 479–485. Lewis, V. A., G. M. Hynes, D. Zheng, H. Saibil, and K. Willison (1992). T-Complex polypeptide-1 is a subunit of a heteromeric particle in the eukaryotic cytosol [see comments]. Nature 358, 249–252. Lin, P., T. S. Cardillo, L. M. Richard, G. B. Segel, and F. Sherman (1997). Analysis of mutationally altered forms of the Cct6 subunit of the chaperonin from Saccharomyces cerevisiae. Genetics 147, 1609–1633.

102

NICHOLAS J. COWAN AND SALLY A. LEWIS

Liou, A. K., E. A. McCormack, and K. R. Willison (1998). The chaperonin containing TCP-1 (CCT) displays a single-ring mediated disassembly and reassembly cycle. Biol. Chem. 379, 311–319. Liou, A. K., and K. R. Willison (1997). Elucidation of the subunit orientation in CCT (chaperonin containing TCP1) from the subunit composition of CCT microcomplexes. EMBO J. 16, 4311–4316. Llorca, O., E. A. McCormack, G. Hynes, J. Grantham, J. Cordell, J. L. Carrascosa, K. R. Willison, J. J. Fernandez, and J. M. Valpuesta (1999a). Eukaryotic type II chaperonin CCT interacts with actin through speciﬁc subunits. Nature 402, 693–696. Llorca, O., M. G. Smyth, J. L. Carrascosa, K. R. Willison, M. Radermacher, S. Steinbacher, and J. M. Valpuesta (1999b). 3D reconstruction of the ATP-bound form of CCT reveals the asymmetric folding conformation of a type II chaperonin. Nat. Struct. Biol. 6, 639–642. Llorca, O., M. G. Smyth, S. Marco, J. L. Carrascosa, K. R. Willison, and J. M. Valpuesta (1998). ATP binding induces large conformational changes in the apical and equatorial domains of the eukaryotic chaperonin containing TCP-1 complex. J. Biol. Chem. 273, 10091–10094. Lorimer, G. H. (1996). A quantitative assessment of the role of the chaperonin proteins in protein folding in vivo. FASEB J. 10, 5–9. Lubben, T. H., A. A. Gatenby, G. K. Donaldson, G. H. Lorimer, and P. V. Viitanen (1990). Identiﬁcation of a groES-like chaperonin in mitochondria that facilitates protein folding. Proc. Natl. Acad. Sci. USA 87, 7683–7687. Marco, S., J. L. Carrascosa, and J. M. Valpuesta (1994). Reversible interaction of beta-actin along the channel of the TCP-1 cytoplasmic chaperonin. Biophys. J. 67, 364–368. Martin, L., M. L. Fanarraga, K. Aloria, and J. C. Zabala (2000). Tubulin folding cofactor D is a microtubule destabilizing protein. FEBS Lett. 470, 93–95. McCallum, C. D., H. Do, A. E. Johnson, and J. Frydman (2000). The interaction of the chaperonin tailless complex polypeptide 1 (TCP1) ring complex (TRiC) with ribosome-bound nascent chains examined using photo-cross-linking. J. Cell. Biol. 149, 591–602. Melki, R., G. Batelier, S. Soulie, and R. C. Williams, Jr. (1997). Cytoplasmic chaperonin containing TCP-1: Structural and functional characterization. Biochemistry 36, 5817– 5826. Melki, R., and N. J. Cowan (1994). Facilitated folding of actins and tubulins occurs via a nucleotide-dependent interaction between cytoplasmic chaperonin and distinctive folding intermediates. Mol. Cell Biol. 14, 2895–2904. Melki, R., H. Rommelaere, R. Leguy, J. Vandekerckhove, and C. Ampe (1996). Cofactor A is a molecular chaperone required for beta-tubulin folding: Functional and structural characterization. Biochemistry 35, 10422–10435. Miklos, D., S. Caplan, D. Mertens, G. Hynes, Z. Pitluk, Y. Kashi, K. Harrison-Lavoie, S. Stevenson, C. Brown, B. Barrell, et al. (1994). Primary structure and function of a second essential member of the heterooligomeric TCP1 chaperonin complex of yeast, TCP1 beta. Proc. Natl. Acad. Sci. USA 91, 2743–2747. Mitchison, T. J., and M. W. Kirschner (1986). Beyond self-assembly: from microtubules to morphogenesis. Cell 45, 329–342. Norcum, M. T. (1996). Novel isolation method and structural stability of a eukaryotic chaperonin: The TCP-1 ring complex from rabbit reticulocytes. Protein Sci. 5, 1366– 1375. Ostermann, J., A. L. Horwich, W. Neupert, and F. U. Hartl (1989). Protein folding in mitochondria requires complex formation with hsp60 and ATP hydrolysis. Nature 341, 125–130.

TYPE II CHAPERONINS, PREFOLDIN, AND THE TUBULIN-SPECIFIC CHAPERONES

103

Radcliffe, P. A., D. Hirata, L. Vardy, and T. Toda (1999). Functional dissection and hierarchy of tubulin-folding cofactors in ﬁssion yeasts. Mol. Biol. Cell 10, 2987–3001. Radcliffe, P. A., and T. Toda (2000). Characterisation of ﬁssion yeast alp 11 mutants deﬁnes three functional domains within tubulin-folding cofactor B. Mol. Gen. Genet. 263, 752–760. Radcliffe, P. A., L. Vardy, and T. Toda (2000). A conserved small GTP-binding protein Alp41 is essential for the cofactor-dependent biogenesis of microtubules in ﬁssion yeast. FEBS Lett. 468, 84–88. Rommelaere, H., M. De Neve, R. Melki, J. Vandekerckhove, and C. Ampe (1999). The cytosolic class II chaperonin CCT recognizes delineated hydrophobic sequences in its target proteins. Biochemistry 38, 3246–3257. Roobol, A., and M. J. Carden (1999). Subunits of the eukaryotic cytosolic chaperonin CCT do not always behave as components of a uniform hetero-oligomeric particle. Eur. J. Cell Biol. 78, 21–32. Roobol, A., J. Grantham, H. C. Whitaker, and M. J. Carden (1999). Disassembly of the cytosolic chaperonin in mammalian cell extracts at intracellular levels of K+ and ATP. J. Biol. Chem. 274, 19220–19227. Roobol, A., F. E. Holmes, N. V. Hayes, A. J. Baines, and M. J. Carden (1995). Cytoplasmic chaperonin complexes enter neurites developing in vitro and differ in subunit composition within single cells. J. Cell Sci. 108, 1477–1488. Schuler, H., U. Lindberg, C. E. Schutt, and R. Karlsson (2000). Thermal unfolding of G-actin monitored with the DNase l-inhibition assay stabilities of actin isoforms. Eur. J. Biochem. 267, 476–486. Siegers, K., T. Waldmann, M. R. Leroux, K. Grein, A. Shevchenko, E. Schiebel, and F. U. Hartl (1999). Compartmentation of protein folding in vivo: Sequestration of nonnative polypeptide by the chaperonin-GimC system. EMBO J. 18, 75–84. Srikakulam, R., and D. A. Winkelmann (1999). Myosin II folding is mediated by a molecular chaperonin. J. Biol. Chem. 274, 27265–27273. Stearns, T., M. A. Hoyt, and D. Botstein (1990). Yeast mutants sensitive to antimicrotubule drugs deﬁne three genes that affect microtubule function. Genetics 124, 251– 262. Steinbacher, S. (1999). Crystal structure of the post-chaperonin beta-tubulin binding cofactor Rbl2p. Nature Struct. Biol. 6, 1029–1032. Stoldt, V., F. Rademacher, V. Kehren, J. F. Ernst, D. A. Pearce, and F. Sherman (1996). Review: The Cct eukaryotic chaperonin subunits of Saccharomyces cerevisiae and other yeasts. Yeast 12, 523–529. Studier, F. W., A. H. Rosenberg, J. J. Dunn, and J. W. Dubendorff (1990). Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185, 60–89. Thulasiraman, V., C. F. Yang, and J. Frydman (1999). In vivo newly translated polypeptides are sequestered in a protected folding environment. EMBO J. 18, 85–95. Tian, G., A. Bhamidipati, N. J. Cowan, and S. A. Lewis (1999). Tubulin folding cofactors as GTPase activating proteins: GTP hydrolysis and the assembly of the alpha/beta tubulin heterodimer. J. Biol. Chem. 274, 24054–24058. Tian, G., Y. Huang, H. Rommelaere, J. Vandekerckhove, C. Ampe, and N. J. Cowan (1996). Pathway leading to correctly folded beta-tubulin. Cell 86, 287–296. Tian, G., S. A. Lewis, B. Feierbach, T. Stearns, H. Rommelaere, C. Ampe, and N. J. Cowan (1997). Tubulin subunits exist in an activated conformational state generated and maintained by protein cofactors. J. Cell. Biol. 138, 821–832. Tian, G., I. E. Vainberg, W. D. Tap, S. A. Lewis, and N. J. Cowan (1995a). Quasi-native chaperonin-bound intermediates in facilitated protein folding. J. Biol. Chem. 270, 23910–23913.

104

NICHOLAS J. COWAN AND SALLY A. LEWIS

Tian, G., I. E. Vainberg, W. D. Tap, S. A. Lewis, and N. J. Cowan (1995b). Speciﬁcity in chaperonin-mediated protein folding. Nature 375, 250–253. Trent, J. D., E. Nimmesgern, J. S. Wall, F. U. Hartl, and A. L. Horwich (1991). A molecular chaperone from a thermophilic archaebacterium is related to the eukaryotic protein t-complex polypeptide-1 [see comments]. Nature 354, 490–493. Ursic, D., and M. R. Culbertson (1991). The yeast homolog to mouse Tcp-1 affects microtubule-mediated processes. Mol. Cell. Biol. 11, 2629–2640. Ursic, D., J. C. Sedbrook, K. L. Himmel, and M. R. Culbertson (1994). The essential yeast Tcp1 protein affects actin and microtubules. Mol. Biol. Cell 5, 1065–1080. Vainberg, I. E., S. A. Lewis, H. Rommelaere, C. Ampe, J. Vandekerckhove, H. L. Klein, and N. J. Cowan (1998). Prefoldin, a chaperone that delivers unfolded proteins to cytosolic chaperonin. Cell 93, 863–873. Viitanen, P. V., G. H. Lorimer, R. Seetharam, R. S. Gupta, J. Oppenheim, J. O. Thomas, and N. J. Cowan (1992). Mammalian mitochondrial chaperonin 60 functions as a single toroidal ring. J. Biol. Chem. 267, 695–698. Viitanen, P. V., T. H. Lubben, J. Reed, P. Goloubinoff, D. P. O’Keefe, and G. H. Lorimer (1990). Chaperonin-facilitated refolding of ribulosebisphosphate carboxylase and ATP hydrolysis by chaperonin 60 (groEL) are K+ dependent. Biochemistry 29, 5665– 5671. Vinh, D. B., and D. G. Drubin (1994). A yeast TCP-1-like protein is required for actin function in vivo. Proc. Natl. Acad. Sci. USA 91, 9116–9120. Waldmann, T., E. Nimmesgern, M. Nitsch, J. Peters, G. Pfeifer, S. Muller, J. Kellermann, A. Engel, F. U. Hartl, and W. Baumeister (1995). The thermosome of Thermoplasma acidophilum and its relationship to the eukaryotic chaperonin TRiC. Eur. J. Biochem. 227, 848–856. Won, K. A., R. J. Schumacher, G. W. Farr, A. L. Horwich, and S. I. Reed (1998). Maturation of human cyclin E requires the function of eukaryotic chaperonin CCT. Mol. Cell Biol. 18, 7584–7589. Yaffe, M. B., G. W. Farr, D. Miklos, A. L. Horwich, M. L. Sternlicht, and H. Sternlicht (1992). TCP1 complex is a molecular chaperone in tubulin biogenesis [see comments]. Nature 358, 245–248. Yokota, S., H. Yanagi, T. Yura, and H. Kubota (1999). Cytosolic chaperonin is up-regulated during cell growth. Preferential expression and binding to tubulin at G(1)/S transition through early S phase. J. Biol. Chem. 274, 37070–37078. Yokota, S. I., H. Yanagi, T. Yura, and H. Kubota (2000). Upregulation of cytosolic chaperonin CCT subunits during recovery from chemical stress that causes accumulation of unfolded proteins. Eur. J. Biochem. 267, 1658–1664. Zabala, J. C., and N. J. Cowan (1992). Tubulin dimer formation via the release of alphaand beta-tubulin monomers from multimolecular complexes. Cell Motil. Cytoskeleton 23, 222–230. Zabala, J. C., A. Fontalba, and J. Avila (1996). Tubulin folding is altered by mutations in a putative GTP binding motif. J. Cell. Sci. 109, 1471–1478.

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEIN/α-CRYSTALLIN FAMILY OF MOLECULAR CHAPERONES By ROB VAN MONTFORT,* CHRISTINE SLINGSBY,* and ELIZABETH VIERLING† *Department of Crystallography, Birkbeck College, Malet Street, London WC1E 7HX, United Kingdom, and †Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, AZ 85721-0106

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Diversity of the sHsps and Their Expression Patterns . . . . . . . . . . . . . . . . . . . . A. Sequence Relationships of the sHsps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. sHsp Oligomeric Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Complexity of sHsp Expression Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . III. X-Ray Structural Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Secondary and Tertiary Structure of the sHsp Monomer . . . . . . . . . . . . . B. The Dimeric Building Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Oligomeric Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Location of Conserved Sequence Motifs . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Structural Properties of Other sHsps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Dynamic Nature of the sHsp Oligomer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Subunit Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Temperature-Induced Structural Changes . . . . . . . . . . . . . . . . . . . . . . . . . C. Phosphorylation-Induced Structural Changes . . . . . . . . . . . . . . . . . . . . . . D. Transition to Insoluble Aggregates In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . V. sHsp Chaperone Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Model for sHsp Chaperone Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. sHsp/Substrate Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Substrate Binding Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Conformation of the Substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Refolding of sHsp-Bound Substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Insoluble sHsp/Substrate Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Potential sHsp Substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Interaction with the Cytoskeleton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Dominant Mutations in α-Crystallin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Substrates Implicated by Loss-of-Function Mutants . . . . . . . . . . . . . . . . . . D. Other Interacting Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. sHsp Membrane Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105 107 107 109 114 116 118 119 119 120 121 124 124 125 125 127 127 127 130 132 134 135 138 138 139 140 142 143 144 146 147

I. INTRODUCTION The small heat shock proteins (sHsps) and the structurally related vertebrate eye lens α-crystallins have been the poor cousins in the family 105 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0096-5332/02 $35.00

106

ROB VAN MONTFORT ET AL.

of molecular chaperones, and remain the least understood both structurally and functionally. For simplicity, throughout this chapter the term sHsps will be used inclusively to refer to both sHsps and α-crystallins. This family of virtually ubiquitous stress proteins, with monomeric masses of ∼16–40 kDa, are deﬁned by a conserved C-terminal domain of approximately 90 amino acids referred to as the α-crystallin domain (de Jong et al., 1998). The monomeric units of almost all members of the family assemble into large oligomers of 9 to >24 subunits, depending on the speciﬁc protein. The sHsps are very efﬁcient at binding denatured proteins, and current models propose that sHsps function as molecular chaperones to prevent irreversible protein aggregation and insolubilization (Derham and Harding, 1999; Groenen et al., 1994; Haslbeck et al., 1999; Horwitz, 1992; Jaenicke and Creighton, 1993; Jakob et al., 1993; Jakob and Buchner, 1994; Lee and Vierling, 2000; Merck et al., 1993a). Many experiments provide evidence that increased expression of sHsps in different cell types can increase tolerance to a variety of stresses, including heat, salt, drugs, and oxidants (Arrigo, 1998; Arrigo and Landry, 1994). Thus, the sHsps appear to be capable of binding a variety of unstable proteins and thereby provide an important cellular protection mechanism that operates during stress and in prevention of cataract in the eye lens. Numerous reviews discuss the sHsp family from many standpoints, and the reader is referred to several of these for extensive details that cannot be covered here. An excellent summary of earlier literature with a focus on the mammalian proteins and cell biology was prepared by Arrigo and Landry (1994). Recent papers concerning the potential role of sHsps in human disease and eye lens structure are summarized by several authors (Ciocca et al., 1993; Clark and Muchowski, 2000; Dillmann, 1999; Groenen et al., 1994; Horwitz, 2000; van dan IJssel et al., 1999; Welsh and Gaestel, 1998). Discussion of the evolution and diversity of sHsps in microbes, animals, and plants can be found in a number of papers and reviews (Caspers et al., 1995; de Jong et al., 1993; 1998; Vierling, 1997; Waters, 1995; Waters et al., 1996). Derham and Harding (1999) have written comprehensively about the molecular chaperone activity of the α-crystallins. Interpretation of NMR studies of sHsp structure and substrate interactions is presented in Carver and Lindner (1999), and MacRae (2000) concisely summarizes mutation analysis of sHsp function. Earlier books concerning molecular chaperones include multiple chapters on proteins in the sHsp family (Ehrnsperger et al., 1998; Gething, 1997; Lorimer and Baldwin, 1998); and recently an entire volume devoted to this family has been completed, providing further discussion and references on speciﬁc aspects of sHsp biology and biochemistry (Arrigo and M¨uller, 2001).

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

107

From the outset it should be made clear that the function of sHsps as chaperones is a working model. Whether the chaperone properties that sHsps exhibit in vitro can provide a mechanistic basis for all their cellular activities remains to be shown. The diversity of sHsp structure and expression patterns is immense, and their activities in vivo may easily involve multiple mechanisms. Thus, one goal of this chapter is to clarify the diversity within the sHsp family and to describe evidence indicating that sHsps have many different substrates and affect a wide range of cellular functions. Second, owing to the solution of the X-ray structure of two sHsps—Hsp16.5 from the hyperthermophilic archeon Methanococcus jannaschii (MjHsp16.5) (Kim et al., 1998a) and, very recently, Triticum aestivum (wheat) Hsp16.9 (TaHsp16.9) (van Montfort et al., 2001)—it is possible to begin, for the ﬁrst time, a structural analysis of variation in sHsp oligomeric structure and to interpret available functional data in a structural context. Finally, bringing together biochemical studies of the chaperone properties of these proteins in light of the new structural data allows us to identify areas ripe for experimentation and to enjoy some speculations. II. DIVERSITY OF THE SHSPS AND THEIR EXPRESSION PATTERNS A. Sequence Relationships of the sHsps The sHsp family was essentially ﬁrst deﬁned by Ingolia and Craig (1982), who recognized that sHsps from Drosophila and the vertebrate eye lens α-crystallins contained homologous C-terminal domains of the order of 90 amino acids in length. Genes encoding clearly similar C-terminal domains have now been recognized in all organisms, with the exception of a few bacteria (de Jong et al., 1998; Kapp´e et al., 2001; Munchbach et al., 1999). Outside the conserved α-crystallin domain the sHsps have a variable-length N-terminal domain and a short C-terminal extension. The N-terminal domains of sHsps show little or no similarity, except between evolutionarily closely related proteins. The sHsp family is therefore different from the Hsp70 and Hsp90 chaperone families, in which there is considerable sequence identity throughout the protein even between eukaryotic and prokaryotic members. The number of sHsps found in different organisms varies signiﬁcantly. Most prokaryotes contain one or two sHsps, except species of Rhizobium, some of which have at least ten sHsp genes (Kapp´e et al., 2001; Munchbach et al., 1999; Vierling, 1997). Saccharomyces cerevisiae has two sHsp genes—Hsp26 and Hsp42, one of the largest family members

108

ROB VAN MONTFORT ET AL.

owing to an extended N terminus (Petko and Lindquist, 1985; Wotton et al., 1996). The completion of various eukaryotic genome projects has increased the number of proteins classiﬁed as sHsps, revealing a greater complexity in some organisms than originally thought. In humans, only ﬁve sHsps, including αA- and αB-crystallin (de Jong et al., 1998), had been described, until recently when an additional four transcribed sequences were identiﬁed (Boelens et al., 1998; Kapp´e et al., 2001; Krief et al., 1999; Suzuki et al., 1998). The genome of Caenorhabditis elegans encodes 16 proteins classiﬁed as sHsps (Ding and Candido, 2000a), and Drosophila melanogaster has 12 sHsp genes (S´ebastien et al., 2001). Arabidopsis thaliana, a representative of higher plants, which were recognized early as having large numbers of sHsp genes (Vierling, 1991), has 19 genes encoding sHsps and another 25 genes with putative α-crystallin domains (Scharf et al., 2001). The sHsp family is quite unusual in plants and warrants some additional discussion to put the structure of TaHsp16.9 in its evolutionary context. The plant sHsp lineage appears to have evolved independently after the divergence of plants and animals (de Jong et al., 1998; Waters, 1995; Waters et al., 1996). In other organisms sHsps are found in the cytoplasm under most conditions and at times in the nucleus, but plants express both cytosolic sHsps and speciﬁc isoforms targeted to intracellular organelles. There are at least two evolutionarily distinct forms of sHsp in the cytosol, referred to as class I and class II proteins, which diverged more than 400 million years ago (Vierling, 1991; Waters and Vierling, 1999b). They are on the order of 50% identical in the α-crystallin domain but have essentially no identity in the N-terminal domain. TaHsp16.9 is a member of the class I cytosolic plant sHsps. Three separate gene families encode mitochondrial, chloroplast, and endoplasmic reticulum– localized sHsps, each containing appropriate targeting signals for these organelles. Analysis of the array of Arabidopsis sHsps indicates the possibility of additional plant sHsp gene families (Scharf et al., 2001). The organelle proteins appear to have arisen through duplication of nuclear genes, most likely of the class I type or its precursor (Waters and Vierling, 1999a). This pattern of evolution is distinct from that of the Hsp70 genes, which are also found in multiple cellular compartments (Boorstein et al., 1994). Notably, Hsp70s found in chloroplasts and mitochondria are clearly derived from prokaryotic endosymbiotic Hsp70s, indicating a more ancient origin than the plant sHsp organelle forms. The organelle sHsp forms appear to be unique to plants, with the exception of Hsp22 in Drosophila, which was recently reported to be located in mitochondria (Morrow et al., 2000).

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

109

The diversity in sHsp primary structure is illustrated in part by Figure 1, which aligns the amino acid sequences of some of the sHsps for which considerable structural and/or functional data are available. The alignment includes TaHsp16.9 and Mj Hsp16.5, the two proteins for which there are X-ray structures; Pisum sativum (pea) Hsp18.1, a close homolog of the wheat protein; the two α-crystallin isoforms, αA and αB; human and mouse sHsps, and Saccharomyces cerevisiae Hsp26. The common α-crystallin domain is considered to extend from the beginning of β-strand β2 to the end of β-strand β9 (indicated above and below the alignment). The N-terminal regions, which vary from 44 amino acids in Mj Hsp16.5 to ∼93 residues in ScHsp26 cannot be aligned to reveal any meaningful similarity between all eight of these proteins, although similarities between the closely related proteins are evident. The length of the C-terminal extensions varies from 12 residues in human Hsp20 to 36 residues in mouse Hsp25. Sequence alignments of the α-crystallin domain and C-terminal extension of a more complete spectrum of sHsps have been carefully analyzed by de Jong and colleagues (Caspers et al., 1995; de Jong et al., 1998). These authors identify three regions that appear highly conserved even in a vast majority of distantly related sHsps (Fig. 1). The ﬁrst region overlaps strand β7 and is generalized as F-x-R-polar-aromatic-x-L-P. The second region, containing primarily β9, is polar-G-V-L-polar-aliphaticpolar-aliphatic-P-basic. The last motif includes β10 and can be viewed as basic-x-I-x-I/V. Beyond these few consensus residues, other conserved motifs appear to be speciﬁc either to a lineage comprising animal sHsps, α-crystallins, and a subset of prokaryotic sHsps, including Escherichia coli IbpA and IbpB, or to a lineage comprising plant and fungal sHsps as well as sHsps from other prokaryotes, such as Methanococcus and Synechocystis. Indeed, phylogenetic analysis suggests that sHsps may have diverged along these lines into two groups very early in evolution (de Jong et al., 1998). This sequence divergence notwithstanding, very notable conserved features of the α-crystallin domain are its hydrophobicity proﬁle and predicted β-strand secondary structure. Together with the sequence motifs, these features can be used to identify members of the sHsp family. B. sHsp Oligomeric Organization Another deﬁning feature of almost all sHsps is their capacity to assemble into large oligomeric complexes. The number of subunits in the complexes varies depending on the sHsp, as listed for representative

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

111

proteins in Table I. Two features of these oligomers are notable. First, they are generally large. With the exception of a few variant sHsps, all of the native proteins comprise at least 9 subunits, and some more than 30. Second, the proteins seem to be clustered into two different groups—sHsps capable of forming regular, monodisperse assemblies, such as the plant, yeast, and Methanococcus proteins, and sHsps that form almost a continuum of oligomeric structures with a variable number of subunits, such as the mammalian proteins and E. coli IbpB. This structural difference in the sHsp oligomers appears to parallel the two suggested evolutionary lineages described above. In many cases the exact size of the oligomers has not been determined by primary methods (for discussion of accuracy in mass determination, see Sch¨onfeld and Behlke, 1998). Despite extensive efforts and the use of a single species of recombinant polypeptides, these latter sHsps have resisted puriﬁcation in a monodisperse oligomeric form. This heterogeneity may well be important for function, but presents considerable difﬁculties for many types of biochemical analyses. The oligomeric structures listed in Table I are all derived from puriﬁed recombinant proteins, typically expressed in E. coli. However, in many organisms multiple sHsps are found in the same cellular compartment and coassembly of sHsps into heterooligomers is often observed. A primary case in point is the α-crystallins; αA- and αB-crystallin are complexed together in the eye lens at an αA : αB ratio of 3 : 1, forming polydisperse aggregates with an average molecular mass of 800 kDa ←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− FIG. 1. Amino acid sequence alignment of diverse sHsps indicating major structural features. Sequences are Triticum aestivum (wheat) TaHsp16.9 (wHSP 16.9; S21600), Pisum sativum (pea) PsHsp18.1 (pHSP18.1; P19243), human αA-crystallin (halphaA: P02489), human αBcrystallin (halphaB; P02511), human Hsp20 (hHSP20; O14558), human Hsp27 (hHSP27; P04792), mouse Hsp25 (mHSP25; P14602), Saccharomyces cerevisiae Hsp26 (yHSP26; M23871), and Methanococcus jannaschii MjHsp16.5 ( jHSP16.5; U67483). Alignment based on de Jong et al. (1998), modiﬁed to include the structural data. Secondary structural features of Mj Hsp16.5 are indicated below the alignment and those of TaHsp16.9 above the alignment. The α-crystallin domain corresponds to the region from β2 to β9. Asterisks above and below the alignment indicate the start of the ordered region in the six subunits of TaHsp16.9 with unresolved N termini, or of the start of the ordered region for all 24 subunits of the Mj Hsp16.5 in the corresponding crystal structures. Conserved sequence motifs as identiﬁed by de Jong et al. (1998) are underlined below the Mj Hsp16.5 sequence. The conserved Arg residue mutated in human genetic disorders is shown in bold italics. Phosphorylation sites in the mammalian sHsps are underlined and in bold. Peptides implicated in substrate binding are underlined within the appropriate protein sequence and include: PsHsp18.1 residues 1–11 and 75–93 (Lee et al., 1997); αB-crystallin residues 57–69 and 93–107 (Sharma et al., 1997), 73–82 and 93–103 (Sharma et al., 1998b), and 61–65 (Smith et al., 1996); αA-crystallin residues 50–54 and 79–88 (Sharma et al., 1998b), 32–37 and 72–76 (Smith et al., 1996), and 12–21 and 71–88 (Sharma et al., 2000). Note that some of these sites overlap within the underlined regions.

112

TABLE I Sizes of sHsp Oligomers Produced as Recombinant Proteins Amino acids

Subunits per oligomer (size)

Methoda

References

IbpB (B45245)

142

? (2000 kDa) ? (13- and 190-nm particles)

SEC SEC, EM, LS

Hsp16.5 (U67483)

147

24

X-Ray structure

Veinger et al., 1998 Shearstone and Baneyx, 1999 Kim et al., 1998a

Hsp16.3 (A42651)

144

9

Equil. AUC, LS

Chang et al., 1996

Hsp16.9 (S74956)

146

∼20

Equil. AUC

Hsp17.0 (AAC79726)

142

23–26 (400–450 kDa)

Sed. vel.

K. Friedrich and E. Vierling, unpublished Michelini and Flynn, 1999

Hsp27 (P04792) αA-Crystallin (P02489)

199 173

Human

αB-Crystallin (P02511)

175

Mouse

Hsp25 (JN0679)

209

Rat

Hsp20 (P97541)

162

∼32 (9–22-nm particles) ∼32 (650 kDa) >30 (700–800 kDa) ∼32 (650 kDa) 25–39 ∼32 ∼32 2 and >12 (470 kDa)

SEC, EM SEC SEC SEC SEC Sed. vel. AUC, EM Sed. vel. AUC SEC, crosslinking

Haley et al., 2000 Sun et al., 1997 Merck et al., 1992 Sun et al., 1997 Haley et al., 1998 Behlke et al., 1991 Ehrnsperger et al., 1999 van de Klundert et al., 1998

Species Prokaryotic Escherichia coli

Methanococcus jannaschii Mycobacterium tuberculosis Synechocystis sp. PCC6803 Thermotoga maritima Mammalian Human Human

Protein (Acc number)

Plant Triticum aestivum Pisum sativum Pisum sativum Other eukaryotic Saccharomyces cerevisiae Caenorhabditis elegans

Hsp16.9 (S21600)

151 158 152

12 12 12 12

X-Ray structure Equil. AUC Equil. AUC Equil. AUC

van Montfort et al., 2001 van Montfort et al., 2001 Lee et al., 1995 Lee et al., 1995

Hsp18.1 (P19243) Hsp17.8 (AAA33670) Hsp26 (AAA66914)

214

24

STEM

Hsp12.6 (U92044) Hsp12.2 (S44755) Hsp12.3 (CAA92770) Hsp16.2 (P06581)

110 110 109 143

24 1 4 4 14 and 24

SEC SEC, crosslinking SEC, crosslinking SEC, crosslinking Sed. vel. AUC

T. Suzuki and E. Vierling, unpublished Ehrnsperger et al., 1999 Leroux et al., 1997a Kokke et al., 1998 Kokke et al., 1998 Leroux et al., 1997b

a Method abbreviations: SEC, size exclusion chromatography; EM, electron microscopy; STEM, scanning transmission electron microscopy; AUC, analytical ultracentrifugation; Equil., sedimentation equilibrium; Sed. vel., sedimentation velocity; LS, light scattering.

113

114

ROB VAN MONTFORT ET AL.

(de Jong et al., 1998; Groenen et al., 1994). αB-Crystallin is also found in assemblies with Hsp27 in other tissues (Kato et al., 1992; 1993; 1994b; Zantema et al., 1992) and can be assembled in vitro into oligomers with Hsp25 (Merck et al., 1993a). Multiple isoforms of class I proteins in plants are present in the same oligomers in vivo, as are the plant class II proteins (Helm et al., 1997; Jinn et al., 1995). In other cells where multiple sHsps occur, they are also likely to assemble into heterooligomers. Interestingly, there are now three examples in which more than one type of sHsp oligomeric complex exists in a single cell. The nitrogenﬁxing, plant symbiotic bacterium Rhizobium japonicum is unusual among bacteria in having at least ten sHsps that can be classiﬁed in two different sequence groups (Munchbach et al., 1999). One group strongly resembles the E. coli IbpA and B proteins, while the other group is more diverse, albeit similar to yeast and plant sHsps. Whereas proteins in each sequence group will assemble into oligomers with one another, oligomers containing proteins from both groups do not form (Studer and Narberhaus, 2000). Similarly, in the higher plant cytosol, class I and II proteins are found in separate oligomers; even in vitro they will not coassemble (Helm et al., 1997; K. Friedrich, E. Basha, and E. Vierling, unpublished observations). Sugiyama and colleagues (2000) have found two distinct sHsp oligomeric complexes in muscle cells. Hsp27, Hsp20, and αB-crystallin interact in one complex, while HspB2 and HspB3 are found in separate oligomers. The distinct sHsp complexes in all these organisms emphasize the potential diversity of sHsp activities and indicate that they may serve specialized functions. C. Complexity of sHsp Expression Patterns A common feature of sHsps in a wide range of organisms is that they increase in abundance during heat stress, in some cells reaching >1% of total cell protein (Arrigo and Landry, 1994; Vierling, 1991). However, the function of sHsps is not restricted to heat stress and eye lens structure, as evidenced by their many other modes of induction and sites of expression. In mammals sHsps are found under normal physiological conditions in several cell types, but are most conspicuously present in muscle tissues (Kato et al., 1992; Krief et al., 1999; Sugiyama et al., 2000). They also accumulate in nervous system tissues, including motor and primary sensory neurons in the adult rat central nervous system (Plumier et al., 1997) as well as neurons and satellite cells in the peripheral nervous system (Loones et al., 2000; Yamamoto et al., 2001). Likewise, in addition to having a critical role in the eye, αB-crystallin is present at considerable levels in the heart, skeletal muscle, and kidney, and can be induced by

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

115

heat (de Jong et al., 1998; Kato et al., 1993; Klemenz et al., 1991; 1993; Longoni et al., 1990). Accumulation of sHsps is also observed in certain muscular (Vicart et al., 1998) and neurological diseases (Iwaki et al., 1993; Lewis et al., 1999; Van Noort et al., 1995), during stress or injury to the brain (Allen and Chase, 2001; Kato et al., 1994a; Plumier et al., 1996), and in cancer cells or other cells in growth state transitions (Ciocca et al., 1993; Hitotsumatsu et al., 1996; Minowada and Welch, 1995). In these cells sHsps may contribute to the disease and/or represent a failed protective mechanism. sHsps can also block apoptosis and may modulate cellular redox state (Arrigo, 1998; Bruey et al., 2000; Charette et al., 2000; Garrido et al., 1999; Pandey et al., 2000). Developmental regulation of sHsps is also common and has been documented in many different organisms. In Drosophila (Amin et al., 1991; Arrigo and Pauli, 1988; Glaser et al., 1986; Joanisse et al., 1998b; Klemenz and Gehring, 1986; Landry et al., 1991; Marin and Tanguay, 1996), C. elegans (Ding and Candido, 2000a; 2000b; Linder et al., 1996), and plants (Almoguera et al., 1998; Carranco et al., 1997; Prieto-Dapena et al., 1999; Waters et al., 1996; Wehmeyer et al., 1996; Wehmeyer and Vierling, 2000), speciﬁc subsets of sHsp genes have been shown to be controlled by hormonal or other discrete developmental signals and are often expressed in a tissue-speciﬁc fashion. Additional studies describe sHsp expression during development of organisms as diverse as mammals (Gernold et al., 1993; Shakoori et al., 1992; Tanguay et al., 1993; Walsh et al., 1997), Xenopus (Heikkila et al., 1997; Lang et al., 1999; Ohan and Heikkila, 1995), Dictyostelium discoideum (Moerman and Klein, 1997), Schistosoma mansoni (de Jong et al., 1998), Bacillus subtilis (Henriques et al., 1997), and the brine shrimp Artemia (Jackson and Clegg, 1996). Even in S. cerevisiae Hsp26 is regulated both by heat and development, accumulating to high levels in spores (Kurtz et al., 1986). The precise function of sHsps in these contexts remains unknown. In two cases, genetic analysis indicates that sHsp expression is required for normal embryonic development. Inhibition of expression of the C. elegans sHsp SEC-1 by antisense RNA arrested nematode development at a stage shortly after fertilization (Linder et al., 1996), and deletion of the Drosophila sHsp gene Eﬂ led to embryonic lethality (Kurzik-Dumke and Lohmann, 1995). The potential roles of sHsps in prokaryotes is also wide-ranging and perhaps mirrors the variation in their sHsp sequences (de Jong et al., 1998; Kapp´e et al., 2001; Vierling, 1997). Many, though not all, are induced by heat shock, consistent with the idea that protection from high-temperature damage represents an ancient and conserved function of the sHsps. Depending on the organism, they are also observed

116

ROB VAN MONTFORT ET AL.

to be induced in stationary phase, during sporulation, and in response to a broad range of stress conditions including oxidative and osmotic stress, expression of foreign proteins, low pH, and anaerobiosis. In both Bacillus subtilis and mycobacteria the sHsp is associated with the cell wall (Cunningham and Spreadbury, 1998; Henriques et al., 1997). The mycobacterial proteins were also the ﬁrst prokaryotic members of the family to be identiﬁed because they act as immunodominant antigens during infection (Booth et al., 1993; Verbon et al., 1992). Altogether, the variation in sHsp expression patterns indicates that these proteins participate in many cellular processes. III. X-RAY STRUCTURAL ANALYSIS The polydispersity of many sHsp assemblies has made it difﬁcult to obtain high-resolution structural information on members of this protein family. But, as noted above, structures of both prokaryotic and eukaryotic sHsps that form regular oligomeric assemblies have recently been solved, and they provide exciting insights into conserved and diverse aspects of the structural organization of the whole protein family. Kim and colleagues (1998a) reported the structure of Hsp 16.5 from Methanococcus jannaschii at 2.9 A˚ resolution, and we have recently completed the 2.65-A˚ structure of Hsp 16.9 from the cytosol of wheat (Triticum aestivum) (van Montfort et al., 2001). Figures 2–4 illustrate important structural

FIG. 2. Space-ﬁlling model of TaHsp16.9 (left) and Mj Hsp16.5 (right) oligomers. ˚ while Mj Hsp16.5 is roughly TaHsp16.9 is a disklike 12-mer with a radical dimension of 95 A, ˚ Molecular graphics created according to Kraulis spherical, with a radial dimension of 120 A. (1991).

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

117

FIG. 3. Comparison of the monomer structures of TaHsp16.9 (left) and Mj Hsp16.5 (right). Ribbon diagram shows the α-crystallin domain in lighter gray and the N-terminal domain in darker gray. The structure of a complete TaHsp16.9 monomer is shown. The six TaHsp16.9 monomers with unresolved N termini would be truncated prior to α3. Molecular graphics created as in Figure 2.

FIG. 4. Ribbon diagram of the dimer structure of TaHsp16.9 (left) and Mj Hsp16.5 (right). Note the nonequivalent angles of the C-terminal tails of the two TaHsp16.9 monomers; in contrast, the Mj Hsp16.5 dimer is fully symmetrical. Molecular graphics created as in Figure 2.

118

ROB VAN MONTFORT ET AL.

features of these sHsps. The two proteins form different oligomeric assemblies (Table I and Fig. 2). Mj Hsp16.5 is a 24-subunit, hollow spherical complex with an outer diameter of 120 A˚ and an inner diameter of 65 ˚ The complex has 432 symmetry and all monomers are equivalent. In A. contrast, TaHsp16.9 is a dodecameric disk comprising three tetramers organized to form two six-membered rings. The oligomer is approximately 95 A˚ in diameter and 55 A˚ wide with 32 symmetry. It has a hole ˚ The presence of a hollow cavity in both through its center of about 25 A. proteins must, however, be interpreted with caution. In the Mj Hsp16.5 structure 32 residues at the N terminus were not resolved, leaving unaccounted for on the order of 20% of the protein mass, which could ﬁll a signiﬁcant portion of the interior space. In fact, cryoelectron microscopy studies indicate that the interior cavity does contain protein density (Haley et al., 2000). Likewise, although the entire chains for six of the twelve TaHsp16.9 subunits were resolved in the oligomer, for the other six subunits 42 N-terminal residues were disordered. Thus, it is unclear how large any “interior space” is in either protein. A. Secondary and Tertiary Structure of the sHsp Monomer Both structures give an excellent view of the fold of the ∼90 amino acid α-crystallin domain (Fig. 3), which extends from β2 to β9 as marked in the alignment in Figure 1. The sequence identity between these two proteins in this region is <25%, but the fold of this domain in the two structures can be superimposed with a root mean square difference in ˚ The domain consists of a β-sandwich of Cα-coordinates of about 1.5 A. two antiparallel β-sheets similar to the fold of the unrelated Fc fragment of immunoglobulin, although the exact folding topology is different (Deisenhofer, 1981). The ﬁrst β-sheet in the sHsps comprises strands β2, β3, β9, and β8, and the opposing sheet, strands β7, β5, and β4 (Fig. 3). A 21- (wheat) or 19-aa (Mj Hsp16.5) loop between β5 and β7 extends out from the β-sandwich and contains β-strand β6. The dominant β structure of the protein was long predicted from CD spectroscopy of several members of the sHsp family (Dudich et al., 1995; Merck et al., 1993b; Sun et al., 1997), as well as from secondary structure analysis of sHsp sequences (Caspers et al., 1995; de Jong et al., 1998) and, more recently, site-directed spin-labeling studies (Berengian et al., 1999; Mchaourab et al., 2000). This same β-sandwich fold is most likely present in all sHsps, and relative positions of the β-strands in other members of the family can be predicted from the sequence alignment in Figure 1. Outside the α-crystallin domain Mj Hsp16.5 and TaHsp16.9 show both intriguing similarities and differences. The C-terminal extension is

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

119

similar in length (12 and 15 aa, respectively) and conformation, being primarily extended and containing a short, three-residue β10 strand. In the N-terminal region the sequence divergence is reﬂected in structural differences. Although only a small portion of this domain is present in the Mj Hsp16.5 structure (aa 33–44), this segment forms another β-strand that binds antiparallel to β7. In contrast, the same segment forms a short α-helix (α3) in the six fully ordered subunits of TaHsp16.9 (Fig. 3). The remaining N-terminal residues of TaHsp16.9 include a 3/10 helix (α1)and another α-helix (α2) connected by random coil segments. Because TaHsp16.9 is the only sHsp for which there is a complete Nterminal structure, it is hard to extrapolate from this to other sHsps. However, CD studies again suggest that other sHsps have a minor α-helical component which must be outside the α-crystallin domain (Merck et al., 1993b; Sun et al., 1997).

B. The Dimeric Building Block In addition to conservation of the fold of the α-crystallin domain, the basic building block of the M. jannaschii and T. aestivum oligomers is a dimer composed of two monomers related by a twofold axis (Fig. 4). Within the resolved structures, the dimer interface forms the strongest and most extensive contacts in the oligomer. Dimer interface contacts are formed primarily with the extended loop between β-strands β5 and β7. The β6 strand in the loop incorporates into the ﬁrst β-sheet of the partner monomer where it binds antiparallel to β2 to form a ﬁve-stranded sheet. On the order of 15–20% of the solvent-accessible surface (∼1600 A˚ 2) of the monomer is buried in this interface in either protein. In TaHsp16.9 27 intermolecular hydrogen bonds and extensive hydrophobic contacts stabilize the dimer. Although the basic dimer structure is the same in the two proteins, the two monomers of Mj Hsp16.5 dimer have identical conformations, while they are nonidentical in the TaHsp16.9 structure. In each TaHsp16.9 dimer one monomer has a structured N terminus and the other monomer has the unresolved N terminus. In addition, the C-terminal extensions of the TaHsp16.9 monomers are in two different orientations, which is discussed further below relative to assembly of the oligomer.

C. Oligomeric Interactions A major feature stabilizing the TaHsp16.9 oligomer is the six ordered N-terminal arms. In the oligomeric disk, the N-terminal arm from a

120

ROB VAN MONTFORT ET AL.

monomer in one ring and a monomer from the other ring intertwine in a pairwise fashion, creating an unusual knot structure (van Montfort et al., 2001). The three knots formed by this interaction link the oligomer together by hydrophobic contacts and hydrogen bonding between the N-terminal α2 helices. The disorder of the N-terminal arms in Mj Hsp16.5 prevents a detailed analysis of their role in the stabilization of the Mj Hsp16.5 assembly. Interactions of the C-terminal extensions are clearly critical in stabilizing the oligomer, and the C-terminal contact surfaces are virtually identical in the bacterial and wheat proteins. First, the β10 strand of the tail extends the β-sheet of another dimer by binding to β4. In addition to intermolecular hydrogen bonds, the side chains Ile-147 and Ile-149 of TaHsp16.9 and the comparable Ile-144 and Ile-146 of Mj Hsp16.5 make extensive hydrophobic contacts with a hydrophobic groove formed by strands β4 and β8 on one monomer of the interacting dimer. Although these interactions are the same, differences in the orientation of the C-terminal extension relative to the α-crystallin domain allow assembly of the two different oligomeric symmetries. In Mj Hsp16.5 all the C-terminal extensions are identically oriented, and the 24-subunit oligomer is basically spherical and can be viewed as composed of four trimers of dimers. The TaHsp16.9 dodecamer can be considered as formed from two disks, with each disk composed of a trimer of dimers, and assembly achieved in part by asymmetry in the orientation of the C-terminal extensions. There is a 30◦ difference in the angle of the C-terminal extensions of the two monomers in the TaHsp16.9 dimer (Fig. 4). The angle arises from the bend in a hinge region comprising residues 137–141 which follow strand β9. This allows the C terminus of one monomer from each dimer to contact a dimer in the same disk, while the C terminus of the other monomer binds to a dimer in the other disk holding the whole dodecameric assembly together. D. Location of Conserved Sequence Motifs The three conserved sequence motifs identiﬁed by de Jong and colleagues (1998) are all located in the three-dimensional structures (Fig. 1). The two regions in the α-crystallin domain, F-x-R-polar-aromaticx-L-P and polar-G-V-L-polar-aliphatic-polar-aliphatic-P-basic, comprise β-strands β7 and β9, respectively. In the TaHsp16.9 and Mj Hsp16.5 oligomers these regions are primarily involved in maintaining the structural integrity of the α-crystallin domain; however, β7 in TaHsp16.9 also interacts with its N-terminal arm and with the dimerization loop of

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

121

another monomer in the assembly. The third motif, basic-x-I-x-I/V, includes β-strand β10 and is the part of the C-terminal extension that locks into the hydrophobic groove of the α-crystallin domain of another monomer in both sHsp assemblies. E. Structural Properties of Other sHsps There is little doubt that the basic β-sandwich fold of the α-crystallin domain is conserved throughout the sHsp protein family. The β-sheet structure is consistent with structural prediction, CD spectroscopy, and site-directed spin-labeling studies (de Jong et al., 1998; Koteiche and Mchaourab, 1999). What other aspects of sHsp structure, particularly of the higher order structure, can be predicted based on the existing X-ray data? The virtually identical dimer foundation of the M. jannaschii and wheat proteins suggests that this could be the fundamental building block of all sHsps. However, the critical loop between β5 and β7, which contains β-strand β6 and participates in the majority of dimer contacts, is a region of both length and sequence heterogeneity (Fig. 1; Caspers et al., 1995; de Jong et al., 1998). This loop is 19 and 21 residues long in the M. jannaschii and wheat sHsps, respectively, but signiﬁcantly shorter in human and other mammalian sHsps. A loop region similar to Mj Hsp16.5 and TaHsp16.9 may be an excellent predictor of the presence of the same basic dimer structure. For example, subunits of S. cerevisiae Hsp26 most likely form dimers similar to Mj Hsp16.5 and TaHsp16.9. In contrast, a dimer interface formed by monomers with a very different length loop region would likely involve different contacts. There are experimental data to support a dimer substructure in mammalian sHsps (Ehrnsperger et al., 1999; Feil et al., 2001; Lambert et al., 1999; Merck et al., 1992; Rogalla et al., 1999). Recently, Feil and colleagues (2001) proposed a novel dimeric interface for the mammalian proteins based on X-ray solution scattering data of an αB-crystallin N-terminal truncation mutant. In their model the loop region forms a strand–turn strand motif, and interaction of two monomers creates a four-stranded, antiparallel, intersubunit β-sheet. The two α-crystallin domain β-sandwiches are oriented in a V shape and the dimer has an increased surface area compared to the Mj Hsp16.5 or TaHsp16.9 dimers. It will be interesting to conﬁrm this alternative dimer structure in intact, wild-type oligomers. A trimeric substructure has been proposed for Hsp16.3 from Mycobacterium tuberculosis. The MtHsp16.3 oligomer was imaged by cryoelectron microscopy as a ˚ From these three-lobed particle with an edge dimension of ∼100 A.

122

ROB VAN MONTFORT ET AL.

images and the estimated molecular mass of the oligomer (Table I), it was proposed to be composed of nine subunits organized as a trimer of trimers. MtHsp16.3 is also predicted to have a smaller loop between β-strands β5 and β7 that might organize the subunits differently (de Jong et al., 1998). Involvement of the C-terminal extension in oligomeric contacts can be predicted to be similar in most sHsps. The requirement of this region for oligomerization is consistent with a large body of mutagenesis data (Andley et al., 1996; Bova et al., 2000; Derham and Harding, 1999; Merck et al., 1992) and with differences in the oligomeric structure of certain variant sHsps. For example, mammalian Hsp20, which lacks the C-terminal extension motif, forms primarily dimers alone in solution, although at higher concentrations heterogeneous oligomers of ∼470 kDa can occur (van de Klundert et al., 1998; Table I). Similarly, in Hsp12.6 from C. elegans the C-terminal extension is completely absent and the protein is reported to be monomeric (Leroux et al., 1997a). Comparison of the Mj Hsp16.5 and TaHsp16.9 oligomers illustrates how interactions of the C-terminal extension can be conserved despite overall differences in oligomer form, and suggests that these contacts are a common feature of the protein family. The difference in oligomer organization is possible due to the hinge region between the α-crystallin domain and the C-terminal extension, which permits different orientations of the dimeric building blocks, even within a single oligomer. With the exception of human Hsp20, all of the proteins aligned in Figure 1 can be predicted to have a C-terminal extension with a short β-strand and the basic-x-I-x-I/V motif whose hydrophobic residues interact with the hydrophobic groove between β4 and β8 of another monomer. This same motif is found in many other sHsps as well (de Jong et al., 1998). The ability to vary the orientation of the C-terminal extension gives considerable ﬂexibility in building oligomers, and even polydisperse sHsps could utilize this contact in oligomerization. Haley and colleagues (2000) put forth the idea of ﬂexible linkers in their discussions of the heterogeneity of sHsp structures. This assembly feature is analogous to the VP1 protein of simian virus, which also has a ﬂexible C-terminal extension that allows variability of packing while maintaining equivalent subunit–subunit interfaces (Liddington et al., 1991; Stehle et al., 1996). Considering the possible conservation of C-terminal contacts, it is relevant to note that NMR studies have deﬁned highly mobile amino acid residues at the C terminus of mammalian sHsps, termed a ﬂexible C-terminal “tail” (Carver and Lindner, 1999). The presence of this

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

123

mobile region is fully consistent with conservation of C-terminal interactions in the oligomer, since the ﬂexible region involves residues after the basic-x-I-x-I/V motif. In the alignment in Figure 1 the ﬁnal eight or ten residues of αA- and αB-crystallin, respectively, were found to be highly mobile. Likewise, the last 16 residues of human Hsp27 and 18 residues of mouse Hsp25 are ﬂexible (Carver and Lindner, 1999). Flexible extensions of this length are absent from the M. jannaschii and wheat sHsps. The N-terminal region of all sHsps no doubt contributes to oligomeric organization and may actually determine the possible geometry of the oligomer. sHsp N-terminal truncation mutants fail to assemble properly (Leroux et al., 1997b) and the N terminus of some sHsps expressed alone forms ill-deﬁned aggregates (Merck et al., 1992). This latter observation may correlate with the disordered structure of the N termini seen in both TaHsp16.9 and Mj Hsp16.5. Caenorhabditis elegans Hsp12.2 and 12.3 have almost no N-terminal region and form tetramers (although they also lack the C-terminal extension) (Kokk´e et al., 1998). Further understanding of the quaternary organization of diverse sHsps will require resolution of structures from additional family members. Although considerable effort has been directed toward the eye lens α-crystallins, the best preparations remain polydisperse (Table I) and fail to produce crystals. The oligomeric structure of a few sHsps has been explored by cryoelectron microscopy. As already mentioned, EM images of MtHsp16.3 supported the conclusion that this protein is a trimer of trimers (Chang et al., 1996). Stewart and colleagues (Haley et al., 1998; 2000) imaged recombinant human αB-crystallin and human Hsp27. The Hsp27 oligomers ranged in diameter from 90 to 220 A˚ and could not be analyzed in any detail. The αB-crystallin particles were also variable, from 80 to 180 A˚ in diameter, with an estimated 25–39 subunits. By choosing particles of ∼150 A˚ (∼32 subunits), they were able to produce a low-resolution, three-dimensional reconstruction that shows an irregular shell-like structure with a hollow cavity in the center, reminiscent of Mj Hsp16.5 quaternary organization. Cryoelectron microscopy of S. cerevisiae Hsp26 also reveals a shelllike structure (Haslbeck et al., 1999). Because ScHsp26 oligomers are monodisperse, 24-subunit particles, a higher resolution reconstruction is in progress. However, despite having a regular number of subunits, the ScHsp26 oligomer appears to exist in two slightly different size classes, an observation that may explain continued difﬁculty in obtaining diffraction-quality crystals of this protein (H. Saibil, personal communication).

124

ROB VAN MONTFORT ET AL.

IV. DYNAMIC NATURE OF THE SHSP OLIGOMER A. Subunit Exchange As discussed, standard analytical techniques show that most sHsps are oligomeric in their native state; purely monomeric forms are observed only after denaturation. A number of experiments, however, demonstrate that sHsp oligomeric structure is highly dynamic and that the large oligomeric assemblies are in rapid equilibrium with a smaller species. Oligomeric rearrangement can easily be observed by monitoring sHsp subunit exchange between oligomers. Oligomeric rearrangements under native conditions were ﬁrst documented by van den Oetelaar and colleagues (1990), who mixed puriﬁed αA- and αB-crystallin and analyzed the resulting complexes by isoelectric focusing. They showed that temperature-dependent formation of a heterogeneous population of mixed oligomers occurred in a period of hours. Recently, ﬂuorescence energy transfer has been used as a more sensitive method to investigate subunit exchange. Puriﬁed αA-crystallin labeled with two different ﬂuorophores exchanged subunits with a rate constant of 0.075 min−1 at 37◦ C, reaching equilibrium within 4 hours (Bova et al., 1997). Between pH 6.5 and 9.5 little change in exchange rate was observed, whereas temperature had a dramatic effect on rate. There was virtually no exchange at 3◦ C, while the rate at 42◦ C (heat shock temperatures in mammals) was 4.2-fold higher than that at 37◦ C. Interaction with saturating amounts of denatured polypeptides also decreased the exchange rate, by only 15% in the case of a small protein (insulin, 3 kDa) but up to 45% with ovotransferrin (40 kDa). Similarly αA- and αB-crystallin readily exchange in a similar time frame and exchange increases with temperature (Sun and Liang, 1998; Sun et al., 1998), and likewise αA-crystallin and Hsp27 can exchange (Bova et al., 2000). In further studies with wild-type and truncated sHsps Bova and colleagues (2000) concluded that the unit of exchange was larger than a monomer, because smaller assemblies formed by speciﬁc truncation mutants showed no exchange. Subunit exchange is not limited to sHsps that form polydisperse oligomers. Rapid subunit rearrangements also occur in sHsps with deﬁned quaternary structures such as TaHsp16.9, as documented by formation of heterooligomers between TaHsp16.9 and the related PsHsp18.1 under native conditions (van Montfort et al., 2001). In total these data indicate that sHsp oligomers are in rapid equilibrium with a smaller subassembly, which is not seen under standard analytical conditions. A similar subunit exchange has been reported between dimers of the tetrameric bacterial chaperone SecB (Topping et al., 2001) and monomers of the dimeric βB2-crystallin (Slingsby and

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

125

Bateman, 1994). Coleman and colleagues (1994) have documented rapid exchange between subunits of dimeric ornithine decarboxylase and suggest that this exchange facilitates regulation by antizyme. In contrast, there is no evidence for a rapid equilibrium between subunits of the 14-subunit chaperone GroEL (A. Horwich, personal communication), or in a number of other multisubunit enzymes (Distefano et al., 1990; Perry et al., 1992; Wente and Schachman, 1987). The potential importance of subunit exchange to sHsp function is discussed further below. B. Temperature-Induced Structural Changes Because sHsps are proposed to function at elevated temperatures in vivo, it is signiﬁcant to consider temperature effects on sHsp structure. There are considerable data documenting structural changes of sHsps upon heating. The most interesting transitions to consider are those that occur at physiological temperatures and which may have mechanistic implications, although transitions at higher temperatures provide further information on biophysical properties of the proteins. One very clear generalization arises from these studies: Treatment at moderately high temperatures, in the range that induces stress in vivo, leads to structural changes that expose hydrophobic sites. This was ﬁrst shown by Raman and Rao (1994) for α-crystallin with pyrene ﬂuorescence. This conformational change in α-crystallin occurs at 30–40◦ C, a transition temperature also observed by differential scanning calorimetry (Walsh et al., 1991). Similar heat-induced exposure of hydrophobic sites at physiological temperatures has since been documented further for α-crystallin (Das and Surewicz, 1995), plant PsHsp18.1 (Lee et al., 1997), mycobacterial Hsp16.3 (Yang et al., 1999), E. coli IbpB (Shearstone and Baneyx, 1999), and Synechocystis Hsp16.6 (T¨or¨ok et al., 2001). Temperature-induced structural changes appear to be extreme in some members of the sHsp family and lead to actual dissociation of the oligomer. This is best documented for S. cerevisiae Hsp26, where sizeexclusion chromatography performed at different temperatures showed dissociation of the 24-subunit oligomer to a smaller species, potentially tetrameric, at >35◦ C (Haslbeck et al., 1999). In contrast, α-crystallin appears to remain oligomeric, although otherwise altered in conformation (Haslbeck et al., 1999; Raman and Rao, 1997). C. Phosphorylation-Induced Structural Changes There is an extensive literature on phosphorylation of α-crystallins and of mammalian sHsps (Gaestel, 2001; Groenen et al., 1994; Kantorow and

126

ROB VAN MONTFORT ET AL.

Piatigorsky, 1998). Mammalian sHsps are very rapidly and speciﬁcally phosphorylated in response to stress, cytokines, and growth factors, indicating that phosphorylation is an important regulatory modiﬁcation. It should be noted that αB-crystallin also becomes phosphorylated on exposure to the same inducers, but to a lesser extent, and αA-crystallin is not phosphorylated in response to these stimuli (Ito et al., 1997; Kato et al. 1998; van den IJssel et al., 1998; Voorter et al., 1989). Phosphorylation occurs primarily on serine residues within R-x-x-S kinase recognition sequences in the nonconserved N-terminal domain (Fig. 1) and results from activation of a MAP kinase cascade through subsequent stimulation of MAPKAP kinases 2 and 3 (Kato et al., 1998; Knauf et al., 1994; Rouse et al., 1994). The position of the phosphorylation sites is not conserved in all paralogous sHsps, which is not surprising given the lack of similarity of the N-terminal sequences. Additional phosphorylation sites are seen in the α-crystallin domain of human αA-crystallin, and recently a cGMPdependent threonine phosphorylation site was described in the human Hsp27 α-crystallin domain (Butt et al., 2001) (Fig. 1). Phosphorylation alters sHsp oligomeric structure, with the number of sites phosphorylated and total percentage phosphorylation affecting the equilibrium of different-sized species. Several groups have concluded phosphorylation results in stabilization of a dimeric form (Kato et al., 1994b; Lambert et al., 1999; Lavoie et al., 1995; Mehlen et al., 1995), although a tetrameric form has also been suggested (Rogalla et al., 1999). This shift toward sHsp dissociation is consistent with the importance of N-terminal interactions in stabilizing sHsp oligomeric structure. Thus, phosphorylation offers a mechanism for regulating available binding surfaces on the sHsp. The importance of sHsp phosphorylation for in vivo protection of cell function has been conﬂicting, potentially reﬂecting different cellular requirements for the activities of oligomeric and dissociated sHsps in different cell types (Rogalla et al., 1999). The difference in the effect of phosphorylation on chaperone activity in vitro is more difﬁcult to explain. Rogalla and colleagues (1999) report phosphorylation inhibits chaperone activity, while previous reports indicated it had no effect (Knauf et al., 1994). One possibility is that the earlier sHsp preparations were contaminated with phosphatases that removed the phosphates prior to the assay. This mode of regulation does not appear to be conserved throughout the sHsp family, however. There is no indication that plant, yeast, or bacterial sHsps are similarly regulated by phosphorylation, and these sHsps lack consensus phosphorylation sites. One can speculate that homeothermic organisms required this kind of mechanism to control sHsp oligomerization, while heat-induced conformational changes are

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

127

sufﬁcient in organisms with little control over internal temperatures. In addition, while the α-crystallins in the lens are also phosphorylated, thus modiﬁed, they still associate into high-molecular-weight structures (Wang et al., 1995), although molecular mimicry of three sites by mutagenesis suggest that some decrease in size occurs (Ito et al., 2001). D. Transition to Insoluble Aggregates In Vivo Although sHsp oligomers are readily isolated from the soluble cell fraction even when expressed in E. coli (Arrigo and Welch, 1987; Lee and Vierling, 1998), a universally observed and still poorly understood phenomenon is the stress-induced transition of sHsps from soluble oligomers to a salt and nonionic detergent–insoluble form. In mammals (Arrigo et al., 1988; Cuesta et al., 2000; Djabali et al., 1997; Kampinga et al., 1995; Lavoie et al., 1995; van de Klundert and de Jong, 1999), Drosophila (Leicht et al., 1986), chicken (Collier et al., 1988; Collier and Schlesinger, 1986; Miron et al., 1991), plants (Neumann et al., 1987; Nover et al., 1989; Osteryoung and Vierling, 1994), Synechocystis (Horvath et al., 1998; K. Giese, E. Basha, and E. Vierling, unpublished), and other organisms (Katayama-Fujimura et al., 1987), after more severe stress sHsps fractionate with cellular organelles, membranes, and cytoskeleton, but then return to the soluble fraction on recovery. In plants this transition actually involves formation of highly ordered, cytosolic complexes approximately 40 nm in diameter that have been characterized in detail by Nover and colleagues (Kirschner et al., 2000). These complexes, termed “heat shock granules,” contain plant class I and II sHsps as well as other proteins, including possibly Hsp70. They have also been reported to contain stored, untranslated mRNAs (Nover et al., 1989). Their assembly is highly speciﬁc and involves at least some ordered incorporation of components. It is not clear if completely analogous structures exist in vertebrates, although “stress granules” containing sHsps have been described (Arrigo et al., 1988; Collier et al., 1988; Kedersha et al., 1999). In vertebrate cells insolubilization is more typically explained by cytoskeletal associations (see below). V.

SHSP

CHAPERONE ACTIVITY

A. Model for sHsp Chaperone Activity After the chaperone activity of both Hsp70 and GroE were described (see Bukau and Horwich, 1998; Gething and Sambrook, 1992; for review), it was but a short leap of faith to consider that other Hsps might

128

ROB VAN MONTFORT ET AL.

have chaperone activity. This hypothesis led to experiments showing that sHsps functioned in vitro to interact with non-native proteins and to prevent their irreversible aggregation and insolubilization. Early studies with crude cell fractions from plants already suggested that sHsps could protect other proteins from heat-induced insolubilization (Jinn et al., 1989). The ﬁrst work with puriﬁed components was reported by Horwitz (1992) using bovine lens α-crystallin and several model substrates, and experiments were extended immediately to recombinant mouse Hsp25, human Hsp27, and αB-crystallin (Jakob et al., 1993; Merck et al., 1993a). Several now well-established assays have been used to demonstrate sHsp protection of a range of protein substrates (Buchner et al., 1998b; Horwitz et al., 1998b; Lee, 1995; Lee and Vierling, 1998). The most common assay is prevention of formation of light-scattering aggregates of model protein substrates exposed to heat- or reductioninduced denaturation. Protection of proteins from insolubilization is also assayed by differential centrifugation. In addition, the effects on the half-time of enzyme inactivation and enhanced recovery of active enzymes diluted from denaturant have been tested. The ability to interact with partially denatured proteins has now been conﬁrmed further with recombinant αA-crystallin (Andley et al., 1996; Sun et al., 1997) and for sHsps from plants (Collada et al., 1997; Jinn et al., 1995; Lee et al., 1995; Lee et al., 1997; Smykal et al., 2000; Young et al., 1999), bacteria (including Mj Hsp16.5) (Chang et al., 1996; Kim et al., 1998b; Michelini and Flynn, 1999; Roy et al., 1999; Shearstone and Baneyx, 1999; Veinger et al., 1998), and other organisms (Fernando and Heikkila, 2000; Leroux et al., 1997b; Liang et al., 1997). A current model for sHsp chaperone activity is diagrammed in Figure 5 considering interaction with a heat-sensitive protein. The sHsp, depicted as the TaHsp16.9 structure, is oligomeric at optimal growth temperatures for the organism. As discussed earlier, increased temperature leads to exposure of hydrophobic sites, possibly due to a shift in the equilibrium distribution toward a smaller subspecies caused by destabilizing, for example, interface hydrogen bonds or “melting” of structure such as α-helix in interdomain contacts. Increased accessibility of hydrophobic patches could also result from an increased rate of subunit exchange or a more subtle, uncharacterized structural alteration. The end result is that substrate is granted exposure to one or more hydrophobic sites per sHsp monomer. The sHsp cannot integrate into preformed substrate aggregates, but must be present during substrate denaturation. Note that in the cell, substrate binding to sHsps will occur even in the presence of other chaperones, and speciﬁc substrates may preferentially

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

129

FIG. 5. Model for sHsp chaperone activity. The sHsp oligomer (TaHsp16.9 shown here) is in rapid equilibrium with a smaller species (possibly a dimer). Heat-denatured substrates bind hydrophobic sites exposed on the sHsp subunits to form soluble sHsp/substrate complexes, preventing formation of insoluble aggregates of denatured proteins. The sHsp/substrate complexes may also be in rapid equilibrium, and when dissociated, the denatured substrate can be picked up and refolded in an ATP-dependent fashion by the Hsp70 or DnaK (plus cochaperone) machinery. Note that sHsp/substrate complexes can also become larger and insoluble, and the fate of these latter complexes is unknown.

bind to sHsps. The sHsp/substrate complexes interact to form higher order aggregates, which can be soluble, but also in some cases insoluble. These sHsp substrate interactions are proposed to be energyindependent. The soluble sHsp/substrate complexes, from which the substrate could be reversibly binding and dissociating, “present” substrate to Hsp70 or DnaK and co-chaperones (in eukaryotic or prokaryotic cells, respectively) for ATP-dependent folding to the ﬁnal native state. In prokaryotes, the additional refolding activity of GroE may be required for some substrates. The fate of insoluble sHsp/substrate complexes is poorly understood. Experimental evidence for this model is discussed below.

130

ROB VAN MONTFORT ET AL.

B. sHsp/Substrate Complexes A remarkable feature of the sHsps compared to Hsp70 and GroE is their large binding capacity for a variety of chemically and heatdenatured model protein substrates. For example, α-crystallin protected an equal weight of alcohol dehydrogenase and twice its weight of βL-crystallin, from insolubilization (Horwitz, 1992). Similarly, plant PsHsp18.1 bound twice its weight of malate dehydrogenase (Lee et al., 1997), and yeast ScHsp26 protected an equal weight of citrate synthase (Haslbeck et al., 1999). This large capacity for substrate binding indicates that there is more than one substrate interaction site per sHsp oligomer and is consistent with the presence of one to two bis-ANS or 1,5AZNS binding sites per α-crystallin monomer, as estimated by Sharma and colleagues (1998a; 1998b). However, capacity can be signiﬁcantly less, depending on the sHsp/substrate combination. Stable sHsp/substrate complexes have been characterized by electrophoresis, size-exclusion chromatography, and electron microscopy (Ehrnsperger et al., 1997; Haslbeck et al., 1999; Lee et al., 1997; Lee and Vierling, 1998; Rao et al., 1993; T¨or¨ok et al., 2001; Fig. 6). Complexes formed with heat-denaturing substrates are typically large and heterogeneous, and apparent complex size increases with increasing ratio of substrate to sHsp (Lee et al., 1997). Bound substrates are inactive and protease-susceptible. The large size of sHsp/substrate complexes suggests that substrates bind to the exterior of the sHsp oligomer; or if the oligomer dissociates before substrate binding, some type of reoligomerization/aggregation must occur. Haslbeck and colleagues (1999) report exactly the latter phenomenon with ScHsp26 and citrate synthase. ScHsp26 dissociates at substrate-binding temperatures, but ScHsp26/citrate synthase complexes imaged by cryoelectron microscopy showed unusually regular, globular assemblies with an estimated 3.6-fold increase in particle volume compared to ScHsp26 alone. In order to demonstrate that binding was to an intact oligomer, it would be necessary to invoke conformational changes to a “subunit” that increased surface hydrophobicity without perturbing the oligomer, which is a difﬁcult process to observe when the subunit exchange rate is fast. Although a lack of ATP requirement for sHsp/substrate interactions has been observed in several studies (Haslbeck et al., 1999; Horwitz, 1992; Jakob et al., 1993; Lee et al., 1995; Lee and Vierling, 2000; Merck et al., 1993a), ATP effects have been reported by others, primarily in experiments with α-crystallin. Clark and colleagues (Muchowski and Clark, 1998) found that 3.5-mM ATP enhanced, by twofold, citrate synthase reactivation after dilution from denaturant in the presence of αB-crystallin.

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

131

FIG. 6. sHsp/substrate complexes observed by size exclusion chromatography. PsHsp18.1 (1 μM oligomer) mixed with 1 μM MDH (A), 1 μM Luc (B), or no substrate (C) was either maintained at room temperature or heated at 45◦ C for 60 min (MDH or no substrate) or 22 min (Luc), then separated at room temperature by size exclusion chromatography. Traces show absorbance at 220 nm versus elution time. Elution position of size standards indicated at the top. For further details, see Lee et al. (1997).

132

ROB VAN MONTFORT ET AL.

ATP also enhanced twofold the ability of αB-crystallin to suppress citrate synthase aggregation during heating. ADP and nonhydrolyzable analogs had no effect. Indirect evidence for ATP binding was also obtained from ATP inhibition of trypsin or chymotrypsin cleavage at speciﬁc sites in αB-crystallin (Muchowski et al., 1999a; 1999b), and earlier work characterized temperature-dependent ATP binding to α-crystallin (Palmisano et al., 1995). In contrast, ATP suppression of tobacco Hsp18 chaperone activity was recently reported (Smykal et al., 2000). These data have all been interpreted as demonstrating that ATP binding alters sHsp conformation. As the sHsps have no deﬁned ATP-binding sites, considerable further work is required to understand the mechanism and importance of these ATP effects. sHsp/substrate complexes are very stable; at room temperature they appear largely unchanged after 24 hours or more and are not disrupted by salt, nonionic detergent, or ATP. However, indirect evidence has suggested that bound substrate is in rapid equilibrium with unbound substrate, as depicted in the model (Fig. 5). Jakob and colleagues (1993) suggested an equilibrium of substrate binding and release based on two observations. First, they saw a small and slow, but reproducible, increase in substrate aggregation during continued incubation of sHsp/substrate complexes at high temperature. Second, reactivation of proteins diluted from denaturant in the presence of sHsps proceeds over a relatively long time scale, suggesting reversible interaction of the sHsp with substrate. Veinger and colleagues (1998) also suggested that denatured MDH and E. coli IbpB are in equilibrium. C. Substrate Binding Site Several studies have attempted to identify substrate interaction sites on sHsps/α-crystallins, but our understanding of these interactions is at best fragmentary. A working hypothesis has been that hydrophobic interactions are responsible for sHsp–substrate binding. The involvement of hydrophobic substrate binding sites in chaperone action are well characterized for GroE and DnaK. Using insulin and other substrates that can be denatured in the absence of heat, Rao and colleagues (1998) observed that chaperone activity of α-crystallin was increased at higher temperatures correlated with exposure of hydrophobic sites, and they developed the hypothesis that heat-induced conformational changes were required for chaperone activity. The correlation between temperature-induced increases in hydrophobicity and enhanced chaperone activity has now been further documented (Das and Surewicz, 1995; Datta and Rao, 1999; Haslbeck et al., 1999; van Boekel et al., 1999; Yang

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

133

et al., 1999). One should take note, however, that other studies report enhanced activity in assays involving temperatures above 50◦ C. As these are nonphysiological temperatures that perturb secondary structure, the mechanistic relevance of these assays is questionable. With the solution of the X-ray structure of TaHsp16.9, it is now possible to place biochemically detected, putative substrate binding sites on an sHsp structure. Lee and colleagues (1997) used bis-ANS binding to identify exposed hydrophobic peptides of PsHsp18.1, a close homolog of TaHsp16.9. Bis-ANS binding was enhanced by incubation of the sHsp at 45◦ C, a temperature at which the protein will protect many different substrates in vitro, and which corresponds to the temperature of induced thermotolerance in plants in vivo (Hong and Vierling, 2000; Vierling, 1991). Exposure of hydrophobic sites was reversed when the protein was returned to 22◦ C and increased bis-ANS binding was blocked when heat-denatured MDH was bound to the sHsp, consistent with the interpretation that the bis-ANS binding sites are involved in substrate interactions. Two peptides from endoproteinase Arg-C digestion of PsHsp18.1 were found to have incorporated bis-ANS, the N terminus (MSLIPSFFSGR) and a peptide within the α-crystallin domain spanning β4 and parts of β3 and β5 (ADLPGLKKEEVKVKVEVEDDR). In an alignment with PsHsp18.1 the corresponding N-terminal peptide in TaHsp16.9 is shorter (MSLIV) while the second peptide falls in a highly conserved region and is almost identical (ADLPGLKKEEVKVEVEDGN) (Fig. 1). The hydrophobic residues in these peptides are not accessible on the exterior of the protein, consistent with the idea that a conformational change or oligomer dissociation is necessary to activate substrate binding properties of sHsps. However, it cannot be ruled out that the disordered N-terminal arms from six of the subunits are accessible in the oligomeric structure. There is some congruence between these putative TaHsp16.9 substrate binding sites and suggested substrate sites on αA- and αB-crystallin (Fig. 1). Bis-ANS binding studies found the N-terminal region of αB-crystallin to be accessible (Smulders and de Jong, 1997). Alcohol dehydrogenase was chemically crosslinked to peptides in αB-crystallin (Sharma et al., 1997), partly overlapping the PsHsp18.1 peptide. An N-terminal peptide and peptides within the ADH crosslinked region of αB-crystallin bound bis-ANS, and substrate binding reduced bis-ANS binding at these sites (Sharma et al., 1998b). Hydrogen–deuterium exchange identiﬁed three accessible peptides in αB-crystallin, two in the N-terminal domain and one which also overlaps the C-terminal PsHsp18.1 peptide (Smith et al., 1996). The same region, plus an N-terminal peptide, was also found to crosslink melittin (Sharma et al.,

134

ROB VAN MONTFORT ET AL.

2000). Considering hydrophobic residues within these regions, Ile-3, Phe-13, Val-56, Phe-57, Val-69, -71, and -80 are all at least partially surfaceexposed in TaHsp16.9; but exposure of additional hydrophobic residues would require structural rearrangement. Site-directed spin labeling concurs that the putative substrate binding peptides in α-crystallin are buried (Koteiche et al., 1998; Koteiche and Mchaourab, 1999). Even α-crystallin, which does not dissociate to a stable suboligomeric species, shows increased subunit exchange at elevated temperatures (Bova et al., 1997), which would also effectively make hydrophobic sites more available for substrate binding. The ﬂexible C-terminal “tail” of mammalian sHsps does not seem to be involved directly in substrate binding (Carver et al., 1994; Lindner et al., 2000). Many sHsps without ﬂexible tails are very effective chaperones. However, the tail appears to confer additional solubility to those sHsps with this feature, and thus can contribute indirectly to chaperone activity (Carver and Lindner, 1998). Using mutagenesis to obtain further evidence for the importance of individual residues or domains in substrate binding is complicated by perturbations of the oligomeric structure that in turn compromise chaperone activity (Derham et al., 2001; Horwitz et al., 1998a; Kumar and Rao, 2000; for review, see Derham and Harding, 1999, and MacRae, 2000). In total, studies have suggested residues in the α-crystallin domain are important for substrate interactions. Although conformational changes, or a shift in the sHsp subunit equilibrium or exchange rate, appear essential for chaperone activity, mutations that disrupt oligomeric assembly (Derham and Harding, 1999; Leroux et al., 1997b) or sHsp variants that do not form large oligomers (Table I) (Kokke et al., 1998) are often not active in chaperone assays. It is possible that natural and engineered sHsp truncation variants are simply missing the appropriate constellation of hydrophobic sites for optimal binding of substrate. However, it also seems reasonable that effective protein solubilization requires the sHsp to reassemble in some way after substrate binding, although not necessarily in a strictly ordered fashion. This reassembly process would most likely involve interactions also required for oligomer formation. D. Conformation of the Substrate Based on interactions with model substrates, an idea of the substrate conformation recognized by sHsps has been derived. First, α-crystallin does not interact with stable unfolded hydrophobic proteins such as α-lactalbumin and α-casein (Carver et al., 1995), which readily interact with DnaK and SecB. Plant PsHsp18.1 also failed to interact with reduced, carboxymethylated α-lactalbumin at room temperature (G. J. Lee and

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

135

E. Vierling, unpublished). Raman and colleagues (1997) suggested that α-crystallin binds to a molton-globule–like intermediate on the refolding pathway of denatured-reduced lysozyme. Lindner and colleagues (1997) concluded αB-crystallin interacts with proteins in a disordered molten globule state, when the protein is close to large-scale precipitation. The ability of different sHsps to retard heat-induced enzyme inactivation or to facilitate enzyme refolding from denaturant varies signiﬁcantly, giving a mixed picture of how sHsps interact with different folding intermediates (Buchner et al., 1998a; 1998b). In total, there is evidence for binding to both early and late unfolding intermediates. E. Refolding of sHsp-Bound Substrates The formation of stable, soluble sHsp/substrate complexes is proposed to be advantageous in the eye lens, a tissue that can no longer perform biosynthetic or repair functions, serving to maintain lens clarity (Groenen et al., 1994; Horwitz, 2000). In other cell types, however, it would seem important that such complexes be removed from the cell, either by release and reactivation of the substrate or by substrate degradation (Waters et al., 1996). While denatured, sHsp-bound proteins have not yet been shown to be preferentially degraded, a few studies indicate that they are good substrates for refolding by the ATP-dependent chaperone Hsp70 (or DnaK). Ehrnsperger and colleagues (1997) reported that heat-inactivated citrate synthase bound to murine Hsp25 could be reactivated to 15% of initial activity after 3 hours in the presence of bovine Hsp70 and ATP. However, this represented only a 2.5-fold stimulation over spontaneous refolding. Much higher yields of reactivated substrate have been obtained using complete reticulocyte lysate and ATP to reactivate ﬁreﬂy luciferase (Luc). Luc heat-denatured in the presence of plant PsHsp18.1 was refolded with more than 80% recovered activity in some assays (Lee et al., 1997; Lee and Vierling, 2000; Fig. 7). The active components of the lysate were found to be Hsp70 and co-chaperones DnaJ (Lee and Vierling, 2000). Interestingly, the plant PsHsp18.1/Luc complexes appeared to be equally good substrates for Luc refolding by Hsp70s from different eukaryotes, as well as by E. coli DnaK (Fig. 7), arguing that speciﬁc interactions between the sHsp and folding system are not required. The ability of sHsps to maintain substrates in a folding competent state was further demonstrated with MDH and lactate dehydrogenase (LDH) denatured in the presence of E. coli IbpB (Veinger et al., 1998), and with MDH in the presence of cyanobacterial Hsp16.6 (also called Hsp17) from Synechocystis (T¨or¨ok et al., 2001). In the latter experiments, GroEL and ES in addition to the DnaK system were required for optimal MDH refolding, while the DnaK

136

ROB VAN MONTFORT ET AL.

FIG. 7. Refolding of PsHsp18.1-bound Luc by reticulocyte lysate and eukaryotic chaperones or by the DnaK system. Panel A: Hsp18.1/Luc complexes (formed with 1 μM Luc and 12 μM PsHsp18.1 monomers) were added to Hsc70 (), Hsc70/Hdj1 (䉬), Hsc70/Ydj1 ( ), Hsc70/Hdj1/Ydj1 (䉱), or 50% reticulocyte lysate (䊊), and allowed to refold at 30◦ C. Concentrations in refolding mix were: 25 nM PsHsp18.1/Luc, 1.5 μM Hsc70, 0.30 μM Hdj1, 0.30 μM Ydj1, or 0.15 μM each Hdj1 and Ydj1 when used in combination. Panel B: 1 μM Luc was heated in the presence of 1 μM PsHsp18.1 (PsHsp18.1/Luc, 䊉), 2 μM DnaK/0.4 μM DnaJ/0.8 μM GrpE (KJE+Luc, 䉱), or 0.21 mg/mL bovine IgG (IgG+Luc, 䊊) at 42◦ C for 8 min in the absence of ATP. Samples containing 25 nM Luc were then added to 1.5 μM DnaK/0.3 μM DnaJ/0.6 μM GrpE and ATP and allowed to refold at 30◦ C. Adapted by permission from Lee and Vierling (2000); copyright by American Society of Plant Physiologists.

system alone operated on LDH. Hsp70-dependent refolding of Luc and citrate synthase bound to α-crystallin was also recently demonstrated, though with lower efﬁciency (Wang and Spector, 2000). It should be noted that these refolding systems are optimized to see an effect, and that the concentration of refolding components used is in large excess to the sHsp/substrate complexes. The sHsp/substrate complexes are also preformed under optimal conditions. However, some data support this model in vivo. Interaction of Hsp70 and an sHsp in renaturation of heat-denatured Luc was implicated in experiments performed by Forreiter and colleagues (1997). They showed that Luc expressed in Arabidopsis cells remained more active and recovered faster from heat denaturation in vivo when protoplasts were ﬁrst transfected with plasmids directing the expression of Hsp70 and an sHsp. Lee and Vierling (2000) could show that total reticulocyte lysate, which has the stoichiometry and approximate concentrations of folding components in whole cells, reactivated added Luc (heat-denatured in the lysate) more effectively when supplemented with PsHsp18.1 (Fig. 8). Luc denatured

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

137

FIG. 8. Luciferase refolding and interaction with PsHsp18.1 in reticulocyte lysate. Panel A: PsHsp18.1 enhances the ability of reticulocyte lysate to refold Luc. In (A), 1 μM Luc in the presence of ATP was mixed with 50% reticulocyte lysate (RL, 䉱) or 0.5 μM Hsp18.1/50% reticulocyte lysate (Hsp18.1+RL, 䊉), heated for 15 min at 42◦ C, then shifted to 30◦ C to allow refolding. Panel B: PsHsp18.1 binds heat-denatured Luc in the presence of other chaperones. Samples similar to those in Panel A containing Luc and reticulocyte lysate (RL) supplemented with or without Hsp18.1, were heated for 15 min at 42◦ C in the presence of ATP. After heating, Luc was immunoprecipitated with anti-Luc IgG (αLuc) or nonspeciﬁc rabbit IgG (mock). Immunoprecipitates were analyzed by Western blot for the presence of Luc, PsHsp18.1, or Hsc70. Bands below Luc and Hsp70 are IgG heavy chains. Adapted by permission from Lee and Vierling (2000); copyright by American Society of Plant Physiologists.

138

ROB VAN MONTFORT ET AL.

in the total lysate was also recovered bound to the sHsp, showing that denatured proteins interact with sHsps even in the presence of other chaperones (Fig. 8). Although Lee and Vierling (2000) and the model in Figure 5 suggest that direct interaction of the refolding machinery with sHsp/substrate complexes is not required, two groups have provided evidence that sHsps can be found in complexes containing Hsp70 and Hsp110 (an Hsp70-related protein), consistent with cooperation of these components in refolding reactions. Wang and colleagues (2000) found Hsp110 in a complex with Hsp70 and Hsp25 in mammalian cells, and could form a complex of the three proteins in vitro. When used for afﬁnity chromatography, Neurospora Hsp30 bound Hsp70 and Hsp88 (a member of the Hsp110 family) (Plesofsky-Vig and Brambl, 1998). Activity of these complexes has not been tested. Integration of the Hsp70 and sHsp systems is also implied by the presence of both genes in a single operon in the bacterium Thermotoga maritima (Michelini and Flynn, 1999). F. Insoluble sHsp/Substrate Complexes The model in Figure 5 includes formation of both soluble and insoluble complexes of sHsp and substrate. The formation of insoluble sHsp/substrate complexes is consistent with the in vivo transition of sHsps to an insoluble, structure-bound form under many stress conditions as discussed above. At present we can provide only speculative explanations for this insolubility in the context of the chaperone model of sHsp function. From in vitro studies, it is clear that the ability of sHsps to keep substrates soluble is dependent on the sHsp-to-substrate ratio, the rate of substrate denaturation, and other factors; in vitro conditions can be manipulated to cause precipitation of sHsp and substrate, as well as to maintain substrate solubility. Thus, insolubilization could result from a type of “overload” of the soluble binding capacity of the sHsps. Since in vivo there is good evidence that the insolubilization is reversible, this leads to the intriguing question of the mechanism of resolubilization, and whether this is also a function of Hsp70 systems, or if additional components are required. Alternatively, sHsp insolubilization in vivo could result from interaction with insoluble components in the cell. VI. POTENTIAL SHSP SUBSTRATES A considerable gap remains between in vitro studies of sHsp chaperone function and our understanding of how sHsps function in vivo. An important outstanding question is: With what substrates do sHsps interact in vivo? The variations in sHsp structure and expression are consistent

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

139

with the conclusion that different sHsps act on a variety of substrates in different organisms and cell types. As is the case for GroE and DnaK (Houry et al., 1999; Mogk et al., 1999), sHsps are likely to have multiple substrates in a single cell. The ability to stabilize many different proteins is supported by the number of substrates that can be protected in vitro, and in vivo, by coimmunoprecipitation and other experiments. Progress in identifying proteins that interact with sHsps in mammalian systems and other organisms is discussed below. A. Interaction with the Cytoskeleton Dramatic alterations in the cytoskeleton in response to heat stress have been characterized in many organisms, and thus the cytoskeleton represents a logical target for protection and repair. The expression of sHsps is also common in cells undergoing extensive developmental remodeling that require reorganization of the cytoskeletal system. In this context the sHsps could be considered as being expressed in response to what Wistow (1993) has termed the “stress of morphogenesis.” Data now support interaction of sHsps with both the actin and intermediate ﬁlament networks. In vertebrate cells sHsps are proposed to modulate the structure of the actin cytoskeleton, and this activity is linked to the protective function of sHsps (Arrigo and Landry, 1994; Huot et al., 1995; Lavoie et al., 1993a; 1993b; 1995; Minowada and Welch, 1995; Piotrowicz and Levin, 1997). This conclusion comes from experiments in which cells transfected or stably transformed to express sHsps show correlated changes in the stability of the actin network, and often colocalization of sHsps and actin stress ﬁbers. sHsp phosphorylation is reported to be essential for actin stabilization during stress, linking stabilization to the multiple signal transduction pathways that induce sHsp phosphorylation. Despite the in vivo support for sHsp/actin interaction, the biochemistry of this interaction has barely been explored. Puriﬁed turkey gizzard sHsp inhibited actin polymerization by 80% at an sHsp monomer:actin molar ratio of 1 : 4 and also reduced actin low shear viscosity (Miron et al., 1988; 1991). The sHsp was proposed to act as a barbed-end actin capping protein. Benndorf and colleagues (1994) prepared phosphorylated and nonphosphorylated sHsps from Ehrlich ascites tumor cells and found that only a nonphosphorylated mixture of monomers and dimers was effective in suppressing actin polymerization, with 90% inhibition at an sHsp:actin ratio of 1 : 1. Isolated recombinant Hsp25 was reported to be nonfunctional in the assay. Thus, sHsps may interact with both actin monomers and ﬁlaments in a phosphorylation-dependent manner.

140

ROB VAN MONTFORT ET AL.

Several groups report that sHsps associate with Type III intermediate ﬁlament proteins (IFPs) in both lenticular and nonlenticular tissues, and have proposed that sHsps modulate IFP assembly. Colocalization of sHsps and IFPs was observed initially in Drosophila cells during heat stress (Leicht et al., 1986) and subsequently in diseased cells (Lowe et al., 1992; Wisniewski and Goldman, 1998). Djabali and colleagues (1997) reported that αB-crystallin was associated with IFPs in NIH3T3 cells only after stress, while Perng and colleagues (1999) saw association already in unstressed astrocytoma and MCF7 cells. Cardiac α-crystallin also colocalizes with desmin in rat heart myoﬁbrils (Longoni et al., 1990). Interactions have been further supported by coimmunoprecipitation and ligand blotting (Djabali et al., 1997; Nicholl and Quinlan, 1994; Perng et al., 1999). In vitro, recombinant α-crystallins will associate with ureasolubilized vimentin or glial ﬁbrillary acidic protein and prevent formation of ﬁlaments at a molar ratio of 0.1 : 1 (α : IFP) (Nicholl and Quinlan, 1994), although higher levels of Hsp25 and αB-crystallin were required in another report (1 : 2; sHsps : IFP), and still did not achieve complete solubilization (Perng et al., 1999). sHsps will also associate with intact ﬁlaments of several different IFPs in copelleting assays (Djabali et al., 1997; Nicholl and Quinlan, 1994; Perng et al., 1999) and prevent gel formation of ﬁlament solutions (Perng et al., 1999). Consistent with heat exposure of sHsp-binding sites, binding to intact ﬁlaments was increased at elevated temperatures within the physiological range (Djabali et al., 1997; Perng et al., 1999). These data suggest that, as with actin, the sHsp alters the ﬁlament assembly/disassembly equilibrium. B. Dominant Mutations in α-Crystallin The relationship of sHsps and cytoskeletal proteins has recently gained further support from genetic evidence linking αB-crystallin to an autosomal dominant desmin-related myopathy (Vicart et al., 1998). Desminrelated myopathy is characterized by adult onset and accumulation of aggregates of desmin, a Type III IFP, and can develop due to either mutation in the desmin gene or mutation of a conserved Arg residue (see Fig. 1) in αB-crystallin (Arg-120 to Gly). Cells transfected with the mutant αB-crystallin also accumulate desmin/αB aggregates. Interestingly, mutation at the same position in αA-crystallin (Arg-116 to Cys) is linked to autosomal dominant congenital cataract (Litt et al., 1999). Thus, mutation at this residue clearly disturbs normal function of α-crystallin. The conserved Arg residue linked to cataract and myopathy can now be considered in the context of the two determined sHsp structures. However, the structural contacts made by this residue differ signiﬁcantly

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

141

in Mj Hsp16.5 (Arg-107) and TaHsp16.9 (Arg-108), although the Arg residue is located in equivalent positions in β-strand β7 in the α-crystallin domain. Arg-107 in Mj Hsp16.5 makes an intramolecular bifurcated hydrogen bound with the carbonyl oxygen atom of Gly41, stabilizing the loop between β-strands β1 and β2. The situation is quite different in TaHsp16.9 as the β1 strand is absent. Instead, an intermolecular contact is formed with the guanidinium group of Arg-108 making a salt bridge with Glu-100 in the dimerization loop of the other monomer of the dimer. Relating these observations to potential contacts in α-crystallins is difﬁcult. αA-Crystallin does not appear to have a similar β1 β-strand (Koteiche and Mchaourab, 1999), making a contact like that in Mj Hsp16.5 unlikely, and the topological equivalent of Glu-100 in the TaHsp16.9 dimerization loop is not conserved in αA-crystallin. In total, the differences in potential contacts of this Arg residue do not provide a clear picture of its structural role or the reason for its conservation. Several studies have examined the effects of the Arg mutations on in vitro chaperone activity of αA- and αB-crystallin (Bova et al., 1999; Cobb and Petrash, 2000b; Der et al., 1999; Kumar et al., 1999). The mutation clearly alters the structure of α-crystallin at all levels. The mutant protein has decreased β-sheet structure, reduced heat stability, altered protease sensitivity, an increased molecular mass, and an even more irregular structure compared to wild type. In vitro it was not only ineffective in keeping α-lactalbumin, alcohol dehydrogenase, insulin, citrate synthase, and aldose reductase soluble in chaperone assays, but was found to aggregate with the denaturing proteins in some studies. The rate of subunit exchange was also reduced in the αA mutant (Cobb and Petrash, 2000b). Further consistent with the disease pathology, mutant αB was signiﬁcantly less effective in preventing assembly of glial ﬁbrillary acidic protein and was incorporated into insoluble material in the assembly assay. The mutant protein also failed to prevent gel formation of intact ﬁlaments (Der et al., 1999). In addition to these data, αB-crystallin accumulation marks reactive astrocytes in a wide variety of neurological disorders—speciﬁcally, intermediate ﬁlament-based glial inclusion bodies. For example, Alexander’s disease is characterized by the presence of Rosenthal ﬁbers, cytoplasmic inclusions in astrocytes that contain glial ﬁbrillary acidic protein (an IFP) in association with sHsps (Head and Goldman, 2000). Genetic analyses show that most patients carry nonconservative mutations in glial ﬁbrillary acidic protein (Brenner et al., 2001). Not only do drastic mutations leave this IFP unable to carry out its normal management of cellular interactions, but overexpression of the mutant glial ﬁbrillary acidic protein results in unfolded protein in need of a chaperone (Quinlan, 2001).

142

ROB VAN MONTFORT ET AL.

C. Substrates Implicated by Loss-of-Function Mutants In two model genetic systems sHsp-deletion mutants exhibit phenotypes consistent with a role in enhanced survival of heat stress, and protection of speciﬁc functions has been proposed. Disruption of Neurospora crassa hsp30 is lethal when the mutant is maintained at high temperature under carbohydrate limitation in the presence of a nonmetabolizable glucose analog (Plesofsky-Vig and Brambl, 1995). Biochemical studies suggest that the mutant has altered glucose metabolism and defective mitochondrial protein import (Plesofsky et al., 1999; Plesofsky and Brambl, 1999). The mechanism for these effects is unknown. The cyanobacterium Synechocystis sp. PCC6803 has a single sHsp gene that is not necessary for normal growth, but is needed for proper adaptation to high temperature (Lee et al., 1998; 2000). The mutant has altered photosynthetic membrane structure and the heat sensitivity of oxygen evolution is enhanced. Constitutive expression of an sHsp was also reported to provide cellular thermotolerance and thermal protection to the photosynthetic apparatus in cyanobacteria (Nakamoto et al., 2000). In parallel to these observations, the higher plant chloroplast-localized sHsp has been suggested to protect the oxygen-evolving capacity of photosystem II (Downs et al., 1999; Heckathorn et al., 1998), although others have disputed this effect (H¨arndahl and Sundby, 2001). Speciﬁc target proteins were not identiﬁed in these studies. Both Neurospora and Synechocystis offer the possibility for further genetic analysis of sHsp structure and function in vivo. Phenotypes associated with deletion of sHsp genes in other model organisms have been elusive. The E. coli IbpB and IbpA genes appear nonessential and their deletion has little effect on viability at high temperature (Thomas and Baneyx, 1998). An ibpA, ibpB deletion somewhat enhances the phenotype of a dnaK756 mutant at high temperature, consistent with the proposed interaction of the sHsp and Hsp70 systems. In contrast to the absence of a loss-of-function phenotype, several studies have reported that high-level expression of heterologous sHsps enhances survival of E. coli to high-temperature treatment (Fernando and Heikkila, 2000; Linder et al., 1996; Muchowski and Clark, 1998; Soto et al., 1999; Yeh et al., 1997). Escherichia coli cell viability after heat stress was also used as an assay to test the function of αB-crystallin mutants, although discrimination between mutants was difﬁcult (Muchowski et al., 1999b). In S. cerevisiae deletion of ScHsp26 alone or in combination with ScHsp42 leads to no measurable defect in tolerance to a variety of stresses (Petko and Lindquist, 1985; Susek and Lindquist, 1989; Wotton et al., 1996; E. Vierling, unpublished observations). Deletion of ScHsp26 also

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

143

has no effect on the phenotype of Hsp70 or Hsp104 mutations. A weak actin interaction has been inferred from the ability of overexpressed ScHsp26 to suppress cytoskeletal defects induced when human Vpr is expressed in yeast, but this could be an indirect effect (Gu et al., 1997). Even extensive screening for mutations that would be lethal in a hsp26, hsp42 background has not been productive (E. Vierling, unpublished data). Other loss-of-function sHsp mutants are worth mentioning, although they are not yet connected to potential substrates. As already mentioned, sHsp mutations have been associated with embryo lethality in both Drosophila and C. elegans (Kurzik-Dumke and Lohmann, 1995; Linder et al., 1996). In mouse, targeted disruption of the αA-crystallin gene induces cataract and cytoplasmic inclusion bodies containing the αB-crystallin (Brady et al., 1997). Surprisingly, given the wide distribution of αB in different tissues, αB knockout mice have no overt phenotype (Brady and Wawrousek, 1997). Phenotype of Hsp25 knockout mice has not yet been reported. Another link of sHsps to human disease is the apparent requirement of the Mycobacterium tuberculosis Hsp16.3 protein for effective growth of the pathogen in macrophages (Yuan et al., 1998). Hsp16.3 is actively expressed in virulent strains during infection and is induced in culture by low oxygen and in stationary phase. It is suggested to be necessary for stability of the cell wall and thereby may also contribute to persistence of the dormant pathogen. Krief and colleagues (1999) note that the newly described human sHsp, cvHsp, maps to a chromosomal interval associated with a cardiomyopathy, and cvHsp is most prominently expressed in heart muscle, suggesting that a defect in this sHsp may be the cause of the disease. Loss-of-function sHsp mutations have not yet been identiﬁed in plants, and the number of sHsp genes may well mask phenotypes of single mutants. D. Other Interacting Proteins Several Hsps, including sHsps, have now been implicated in modulating apoptosis (Xanthoudakis and Nicholson, 2000). sHsps have been shown to act as negative regulators of apoptosis by blocking cytochrome c–dependent activation of caspases (Bruey et al., 2000; Garrido et al., 1999; Pandey et al., 2000). Ability to interfere with apoptosis provides an appealing model to explain how sHsps could modulate drug resistance and correlate with poor prognosis in some human tumors. In cell extracts Hsp27 will coimmunoprecipitate with endogenous or supplemental cytochrome c. Interaction appears to be speciﬁc to the holocytochrome

144

ROB VAN MONTFORT ET AL.

form (Bruey et al., 2000). Recombinant Hsp27 could not replace endogenous Hsp27 in caspase regulation in cell-free extracts, suggesting that the cellular sHsp is somehow modiﬁed. Direct interaction of Hsp27 and cytochrome c has not been studied in vitro. It will be interesting to see if further research can deﬁne direct binding of Hsp27 and cytochrome c. Forming another connection to muscle diseases, the recently described human HspB2 (also called MKBP) was discovered associated with myotonic dystrophy protein kinase (DMPK), which is mutated in the autosomal dominant disease myotonic dystrophy (Suzuki et al., 1998). HspB2 as well as Hsp27 and αB-crystallin appear to enhance DMPK activity in vitro and partially protect DMPK activity during heating. HspB2 is increased in myotonic dystrophy patients, and is found in a complex with HspB3, but not with other sHsps in muscle. A diversity of other proteins have also been suggested to associate with sHsps by yeast two-hybrid interactions, coimmunoprecipitation, and other experimental approaches. These include aggrecan, a condroitin sufate proteoglycan (Zheng et al., 1998); ubc9, a ubiquitin-conjugating enzyme (Joanisse et al., 1998a), a transacting DNA-binding protein involved in the response to glucocorticoids, hGMEB1 (Th´eriault et al., 1999); and a proteasomal subunit (C8/α7) (Boelens et al., 2001). Kudva and colleagues (1997) suggest interaction with Aβ1–42 amyloid peptide, based on inhibition of in vitro amyloidogenesis. Hsp27 from feline heart tissue was puriﬁed as an inhibitor of c-Src kinase in vitro (Kasi and Kuppuswamy, 1999). Suggesting a role in translational control during heat stress, Cuesta and colleagues (2000) found that Hsp27 could bind eIF4G during heat stress and could prevent cap-dependent translation. The idea that sHsps have multiple substrates in cells is also supported by observation in plants, yeast, and mammals, where there is evidence that multiple different proteins associate with sHsps during heat stress (Ehrnsperger et al., 1997; Haslbeck et al., 1999; Jinn et al., 1995; Young et al., 1999). However, none of these latter studies identiﬁed any of the interacting proteins. E. sHsp Membrane Interactions In addition to their potential to interact with non-native proteins in a chaperone capacity, a recent hypothesis postulates that sHsps interact with membranes to reduce high-temperature–induced membrane ﬂuidity (T¨or¨ok et al., 2001; Vigh et al., 1998). Membranes are clearly an important cellular component perturbed by heat, and a need for membrane stabilization is logical. Changes in membrane composition and correlated ﬂuidity in S. cerevisiae and Synechocystis change the “set point”

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

145

for Hsp expression, with expression initiated at lower temperatures in cells with more ﬂuid membranes (Carrat`u et al., 1996; Chatterjee et al., 2000; Horvath et al., 1998). Vigh and colleagues suggest that increased membrane ﬂuidity at high temperature activates a signal contributing to Hsp gene activation and that binding of sHsps and possibly other Hsps to membranes acts to restabilize the membrane, which would then deactivate the transcriptional signal. The interaction of α-crystallins with membranes has been studied for quite some time, because the small fraction of α-crystallin associated with lens ﬁber cell plasma membranes increases with aging and onset of cataract. Association is seen only for α-crystallin; other lens crystallins are not found in membrane preparations (Liang and Li, 1992). Interpretations of different binding studies agree that interactions with the membrane are hydrophobic and do not involve the ﬂexible C-terminal tail (Chandrasekher and Cenedella, 1997; Cobb and Petrash, 2000a; Ifeanyi and Takemoto, 1991; Mulders et al., 1985; Zhang and Augusteyn, 1994). Interactions with the membrane surface (Chandrasekher and Cenedella, 1997) as well as penetration into the fatty acid core of the bilayer have been suggested (Cobb and Petrash, 2000a). Whether these interactions represent a function resulting from positive evolutionary selection or are the result of the dynamic structure of α-crystallin and the presence of exposed hydrophobic surfaces remains open for debate. Direct interaction of sHsps other than α-crystallin with membranes has been reported only for Synechocystis Hsp16.6 (T¨or¨ok et al., 2001). Puriﬁed recombinant protein interacted with lipid vesicles, as monitored by ﬂuorescence anisotropy, to reduce lipid ﬂuidity and also increased the surface pressure of lipid monolayers. There was evidence for both lipid speciﬁcity and membrane penetration of the sHsp. Thylakoids isolated from cells deleted for Hsp16.6 were more ﬂuid membranes than wildtype cells, even in the presence of very little sHsp. Ability to stabilize directly photosynthetic membranes by lipid interaction could explain data supporting a role for sHsps in photosystem II protection, but the speciﬁcity of this effect has not been very well documented (H¨arndahl and Sundby, 2001; Heckathorn et al., 1998; Lee et al., 2000; Nakamoto et al., 2000). sHsp association with membranes in many organisms has often been inferred from biochemical fractionation using differential centrifugation. These studies need to be interpreted with caution, as it is clear that sHsps will form large aggregates with different cellular components, resulting in detergent-insoluble structures. In most cells, immunolocalization studies place sHsps in the soluble cellular compartment, sometimes in granule structures or associated with the cytoskeleton as already

146

ROB VAN MONTFORT ET AL.

discussed. Prominent membrane staining is not observed in eukaryotic cells. The possibility that a portion of the sHsp pool in many cells interacts with membrane lipids or proteins, however, cannot be ruled out. VII. CONCLUSIONS The sHsps are a diverse family of proteins, which are related by possession of a common, deﬁned α-crystallin structural domain. While virtually all of these proteins share the capacity to act as chaperones by binding non-native proteins in an ATP-independent manner, evolutionary arguments and biochemical data indicate that they can interact with many different proteins to inﬂuence a potentially wide range of functions in different cells and organisms. In this regard, they are similar to other chaperones which, by the ability to bind diverse substrates, act to protect and/or to regulate multiple cellular processes. The three-dimensional structure of two sHsps, one prokaryotic and one eukaryotic, deﬁnes the β-sandwich structure of the conserved α-crystallin domain. This domain appears to be a structural organizing feature that accommodates assembly of different higher order oligomers, depending on the composition of the ﬂanking, highly variable N-terminal region and the C-terminal extension. Disordered N-terminal regions are present in both structures and, along with a “hinged” C-terminal extension, are predicted to facilitate the formation of oligomers with different stoichiometries. In the two known structures the most stable subunit interface appears to be that of a dimer, a feature that may or may not be found in other members of the family. In any case, the complexity of other interactions within the oligomer, along with the importance of the variable N-terminus in oligomer assembly, argues that only a subset of oligomeric interactions is conserved across the family. One such conserved interaction is the binding of the C-terminal I-x-I/V motif in a hydrophobic groove of the α-crystallin domain β-sandwich. Our understanding of the mechanism of sHsp action is limited by a combination of minimal knowledge of physiological substrates and of sHsp structures. For the two sHsps with a deﬁned structure, interactions with only a few of the model substrates have been examined; no physiological substrates have been identiﬁed. Common features of all sHsps, however, indicate that a dynamic equilibrium of the oligomer with some type of subassembly is very likely important for function, as dissociation leads to exposure of hydrophobic sites. The data in fact suggest that the oligomer may be a convenient storage form, keeping subunit hydrophobic binding surfaces minimally available until the equilibrium shifts in favor of the dissociated form, or a change in the subunit exchange occurs, activating the protein. Activation of protein function by

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

147

dissociation or alteration of protein complexes is not new. What is new is the possibility that for sHsps this is a heat-sensitive process, potentially ﬁne tuned such that the temperature of sHsp “activation” matches the destabilization temperature of physiological substrates. It is interesting to consider whether temperature plays an important role in the function of other chaperones. sHsps are also expressed in many cells in the absence of heat stress, however, and are presumably active. In mammalian cells exposure of buried surfaces can be regulated by phosphorylation in response to many different signals. Further, in the absence of temperature- or phosphorylation-regulated dissociation, rapid subunit exchange would still allow sHsps to interact with substrates. The conserved features of sHsps, their increasing connection with disease, and the evolutionary expansion of the family in many organisms, such as higher plants, clearly indicate that sHsps have provided a selective advantage over time in many contexts. Difﬁculty in associating phenotypes with these chaperones in some instances is most likely explained by a redundancy of the cellular protection machinery, in which a variable balance of chaperones and proteases can act to ensure cell survival. The chaperone model for sHsp function provides a basic framework to explain the many proposed sHsp/protein interactions and potential functions. The diversity of the sHsp family, however, indicates that care must be taken in generalizing biochemical properties and activities across different family members. Nonetheless, we now have a ﬁrmer structural foundation on which to design future experiments to build a biochemical mechanism of action. ACKNOWLEDGMENTS E.V. thanks all members of her laboratory for continued enthusiasm for the sHsps, but especially Drs. Gary Lee and Eman Basha, whose untiring efforts made biochemical and structural analysis of the plant sHsps possible. Funds for this research have come from grants from the National Institutes of Health (GM-42762) and the U.S. Department of Agriculture (NRICGP) to E.V. and from the Medical Research Council, London, to C.S. and R.vM. For preparation of this review, E.V. is grateful for the support of the National Science Foundation (POWRE Award), the Guggenheim Foundation, and the Dutch National Science Foundation. We also thank many colleagues in the ﬁeld who shared and discussed both published and unpublished data, and Drs. K. Giese and W. de Jong for comments on the manuscript.

REFERENCES Allen, G. V., and Chase, T. (2001). Neuroscience 102, 603–614. Almoguera, C., Prieto-Dapena, P., and Jordano, J. (1998). Plant J. 13, 437–446. Amin, J., Mestril, R., and Voellmy, R. (1991). Mol. Cell. Biol. 11, 5937–5944.

148

ROB VAN MONTFORT ET AL.

Andley, U. P., Mathur, S., Griest, T. A., and Petrash, J. M. (1996). J. Biol. Chem. 271, 31973–31980. Arrigo, A. P. (1998). Biol. Chem. Hoppe Seyler 379, 19–26. Arrigo, A.-P., and Landry, J. (1994). In The Biology of Heat Shock Proteins and Molecular Chaperones (R. Morimoto, A. Tissieres, and C. Georgopolous, eds), pp. 335–373. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Arrigo, A.-P., and Muller, ¨ W. E. G. (2001). Small Stress Proteins. Springer-Verlag, Heidelberg. Arrigo, A. P., and Pauli, D. (1988). Exp. Cell Res. 175, 169–183. Arrigo, A.-P., and Welch, W. J. (1987). J. Biol. Chem. 262, 15359–15369. Arrigo, A.-P., Suhan, J. P., and Welch, W. (1988). Mol. Cell. Biol. 8, 5059–5071. Behlke, J., Lutsch, G., Gaestel, M., and Bielka, H. (1991). FEBS Lett. 288, 119–122. Benndorf, R., Hayess, K., Ryazantsev, S., Wieske, M., Behlke, J., and Lutsch, G. (1994). J. Biol. Chem. 269, 20780–20784. Berengian, A. R., Parfenova, M., and Mchaourab, H. S. (1999). J. Biol. Chem. 274, 6305– 6314. Boelens, W. C., Croes, Y., and de Jong, W. W. (2001). Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 1544, 311–319. Boelens, W. C., Van Boekel, M. A. M., and de Jong, W. W. (1998). Biochim. Biophys. Acta. Protein Struct. Mol. Enzymol. 1388, 513–516. Boorstein, W. R., Ziegelhoffer, T., and Craig, E. A. (1994). J. Mol. Evol. 38, 1–17. Booth, R. J., Williams, D. L., Moudgil, K. D., Noonan, L. C., Grandison, P. M., McKee, J. J., Prestidge, R. L., and Watson, J. D. (1993). Infect. Immun. 61, 1509–1515. Bova, M. P., Ding, L. L., Horwitz, J., and Fung, B. K. K. (1997). J. Biol. Chem. 272, 29511– 29517. Bova, M. P., Mchaourab, H. S., Han, Y., and Fung, B. K. K. (2000). J. Biol. Chem. 275, 1035–1042. Bova, M. P., Yaron, O., Huang, Q. L., Ding, L. L., Haley, D. A., Stewart, P. L., and Horwitz, J. (1999). Proc. Natl. Acad. Sci. USA 96, 6137–6142. Brady, J. P., Garland, D., Duglas-Tabor, Y., Robinson, W. G., Jr., Groome, A., and Wawrousek, E. F. (1997). Proc. Natl. Acad. Sci. USA 94, 884–889. Brady, J. P., and Wawrousek, E. F. (1997). Invest. Opthalmol. Vis. Sci. 38, 4352–4352. Brenner, M., Johnson, A. B., Boespﬂug-Tanguay, O., Rodriguez, D., Goldman, J. E., and Messing, A. (2001). Nature Genet. 27, 117–120. Bruey, J. M., Ducasse, C., Bonniaud, P., Ravagnan, L., Susin, S. A., Diaz-Latoud, C., Gurbuxani, S., Arrigo, A.-P., Kroemer, G., Solary, E., and Garrido, C. (2000). Nat. Cell Biol. 2, 645–652. Buchner, J., Ehrnsperger, M., Gaestel, M., and Walke, S. (1998a). Methods Enzymol. 290, 339–349. Buchner, J., Grallert, H., and Jakob, U. (1998b). Methods Enzymol. 290, 323–338. Bukau, B., and Horwich, A. L. (1998). Cell 92, 351–366. Butt, E., Immler, D., Meyer, H. E., Kotlyarov, A., Laass, K., and Gaestel, M. (2001). J. Biol. Chem. 276, 7108–7113. Carranco, R., Almoguera, C., and Jordano, J. (1997). J. Biol. Chem. 272, 27470–27475. Carratu, ` L., Franceschelli, S., Pardini, C. L., Kobayashi, G. S., Horvath, I., Vigh, L., and Maresca, B. (1996). Proc. Natl. Acad. Sci. USA 93, 3870–3875. Carver, J. A., and Lindner, R. A. (1998). Int. J. Biol. Macromol. 22, 197–209. Carver, J. A., and Lindner, R. (1999). Prog. Retinal & Eye Res. 18, 431–462. Carver, J. A., Aquilina, J. A., Cooper, P. G., Williams, G. A., and Truscott, R. J. W. (1994). Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 1204, 195–206. Carver, J. A., Guerreiro, N., Nicholls, K. A., and Truscott, R. J. W. (1995). Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 1252, 251–260.

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

149

Caspers, G.-J., Leunissen, J. A. M., and de Jong, W. W. (1995). J. Mol. Evol. 40, 238–248. Chandrasekher, G., and Cenedella, R. J. (1997). Exp. Eye Res. 64, 423–430. Chang, Z., Primm, T. P., Jakana, J., Lee, I. H., Serysheva, I., Chui, W., Gilbert, H. F., and Quiocho, F. A. (1996). J. Biol. Chem. 271, 7218–7223. Chatterjee, M. T., Khalawan, S. A., and Curran, B. P. (2000). Microbiology 143, 3063–3068. Charette, S. J., Lavoie, J. N., Lambert, H., and Landry, J. (2000). Mol. Cell. Biol. 20, 7602– 7612. Ciocca, D. R., Oesterreich, S., Chamness, G. C., McGuire, W. L., and Fugua, S. A. W. (1993). J. Nat. Cancer Inst. 85, 1558–1570. Clark, J. I., and Muchowski, P. J. (2000). Curr. Opin. Struct. Biol. 10, 52–59. Cobb, B. A., and Petrash, J. M. (2000a). J. Biol. Chem. 275, 6664–6672. Cobb, B. A., and Petrash, J. M. (2000b). Biochemistry 39, 15791–15798. Coleman, C. S., Stanley, B. A., Viswanath, R., and Pego, J. (1994). J. Biol. Chem. 269, 3155–3158. Collada, C., Gomez, L., Casado, R., and Aragoncillo, C. (1997). Plant Physiol. 115, 71–77. Collier, N. C., Heuser, J., Levy, M. A., and Schlesinger, M. J. (1988). J. Cell Biol. 106, 1131–1139. Collier, N. C., and Schlesinger, M. J. (1986). J. Cell Biol. 103, 1495–1507. Cuesta, R., Laroia, G., and Schneider, R. J. (2000). Genes Dev. 14, 1460–1470. Cunningham, A. F., and Spreadbury, C. L. (1998). J. Bacteriol. 180, 801–808. Das, K. P., and Surewicz, W. K. (1995). FEBS Lett. 369, 321–325. Datta, S. A., and Rao, C. M. (1999). J. Biol. Chem. 274, 34773–34778. Deisenhofer, J. (1981). Biochemistry 20, 2361–2370. de Jong, W. W., Caspers, G.-J., and Leunissen, J. A. M. (1998). Int. J. Biol. Macromol. 22, 151–162. de Jong, W. W., Leunissen, J. A., and Vooter, C. E. (1993). Mol. Biol. Evol. 10, 103–126. Der, P., Muchowski, P. J., Van den Ijssel, P., Wu, G. J. S., Hutcheson, A. M., Clark, J. I., and Quinlan, R. A. (1999). J. Biol. Chem. 274, 33235–33243. Derham, B. K., and Harding, J. J. (1999). Prog. Retinal & Eye Res. 18, 463–509. Derham, B. K., Van Boekel, M. A., Muchowski, P. J., Clark, J. I., Horwitz, J., HepburneScott, H. W., De Jong, W. W., Crabbe, M. J., and Harding, J. J. (2001). Eur. J. Biochem. 268, 713–721. Dillmann, W. H. (1999). Ann. NY Acad. Sci. 874, 66–68. Ding, L., and Candido, E. P. M. (2000a). J. Biol. Chem. 275, 9510–9517. Ding, L., and Candido, E. P. M. (2000b). Biochem. J. 349, 409–412. Distefano, M. D., Moore, M. J., and Walsh, C. T. (1990). Biochemistry 29, 2703–2713. Djabali, K., De N´echaud, B., Landon, F., and Portier, M. M. (1997). J. Cell Sci. 110, 2759– 2769. Downs, C. A., Ryan, S. L., and Heckathorn, S. A. (1999). J. Plant Physiol. 155, 488–496. Dudich, I. V., Zav’yalov, V. P., Pfeil, W., Gaestel, M., Zav’yalova, G. A., Denesyuk, A. I., and Korpela, T. (1995). Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 1253, 163– 168. Ehrnsperger, M., Buchner, J., and Gaestel, M. (1998). In Molecular Chaperones in the Life Cycle of Proteins (A. L. Fink and Y. Goto, eds.), pp. 533–575. Marcel Dekker, New York. Ehrnsperger, M., Gr¨aber, S., Gaestel, M., and Buchner, J. (1997). EMBO J. 16, 221–229. Ehrnsperger, M., Lilie, H., Gaestel, M., and Buchner, J. (1999). J. Biol. Chem. 274, 14867– 14874. Feil, I. K., Malfois, M., Hendle, J., van der Zandt, H., and Svergun, D. I. (2001). J. Biol. Chem. 276, 12024–12029. Fernando, P., and Heikkila, J. J. (2000). Cell Stress & Chaperones 5, 148–159. Forreiter, C., Kirschner, M., and Nover, L. (1997). Plant Cell. 9, 2171–2181.

150

ROB VAN MONTFORT ET AL.

Gaestel, M. (2001). In Small Stress Proteins (A.-P. Arrigo and W. E. G. M¨uller, eds.), 2001. Springer-Verlag, Heidelberg. Garrido, C., Bruey, J. M., Fromentin, A., Hammann, A., Arrigo, A. P., and Solary, E. (1999). FASEB J. 13, 2061–2070. Gernold, M., Knauf, U., Gaestel, M., Stahl, J., and Kloetzel, P.-M. (1993). Dev. Genet. 14, 103–111. Gething, M.-J. (1997). Guidebook to Molecular Chaperones and Protein Folding Catalysts. Oxford University Press, New York. Gething, M.-J., and Sambrook, J. (1992). Nature 355, 33–45. Glaser, R. L., Wolfner, M. F., and Lis, J. T. (1986). EMBO J. 5, 747–754. Groenen, P. J. T. A., Merck, K. B., de Jong, W. W., and Bloemendal, H. (1994). Eur. J. Biochem. 225, 1–19. Gu, J. R., Emerman, M., and Sandmeyer, S. (1997). Mol. Cell. Biol. 17, 4033–4042. Haley, D. A., Bova, M. P., Huang, Q. L., Mchaourab, H. S., and Stewart, P. L. (2000). J. Mol. Biol. 298, 261–272. Haley, D. A., Horwitz, J., and Stewart, P. L. (1998). J. Mol. Biol. 277, 27–35. Haslbeck, M., Walke, S., Stromer, T., Ehrnsperger, M., White, H. E., Chen, S. X., Saibil, H. R., and Buchner, J. (1999). EMBO J. 18, 6744–6751. H¨arndahl, U., and Sundby, C. (2001). Physiol. Plant. 111, 273–275. Head, M. W., and Goldman, J. E. (2000). Neuropath. Appl. Neurobiol. 26, 304–312. Heckathorn, S. A., Downs, C. A., Sharkey, T. D., and Coleman, J. S. (1998). Plant Physiol. 116, 439–444. Heikkila, J. J., Ohan, N., Tam, Y., and Ali, A. (1997). Cell. Mol. Life Sci. 53, 114–121. Helm, K. W., Lee, G. J., and Vierling, E. (1997). Plant Physiol. 114, 1477–1485. Henriques, A. O., Beall, B. W., and Moran, C. P., Jr. (1997). J. Bacteriol. 179, 1887–1897. Hitotsumatsu, T., Iwaki, T., Fukui, M., and Tateishi, J. (1996). Cancer 77, 352–361. Hong, S.-W., and Vierling, E. (2000). Proc. Natl. Acad. Sci. 97, 4392–4397. Horvath, I., Glatz, A., Varvasovszki, V., T¨or¨ok, Z., Pali, T., Balogh, G., Kovacs, E., N´adasdi, L., Benkovic, S., Joo, B., and Vigh, L. (1998). Proc. Natl. Acad. Sci. USA 95, 3513–3518. Horwitz, J. (1992). Proc. Natl. Acad. Sci. USA 89, 10449–10453. Horwitz, J. (2000). Sem. Cell Dev. Biol. 11, 53–60. Horwitz, J., Bova, M. P., Huang, Q.-L., Ding, L., Yaron, O., and Lowman, S. (1998a). Int. J. Biol. Macromol. 22, 263–269. Horwitz, J., Huang, Q. L., Ding, L. L., and Bova, M. P. (1998b). Methods Enzymol. 290, 365–383. Houry, W. A., Frishman, D., Eckerskorn, C., Lottspeich, F., and Hartl, F. U. (1999). Nature 402, 147–154. Huot, J., Lambert, H., Lavoie, J. N., Guimond, A., Houle, F., and Landry, J. (1995). Eur. J. Biochem. 227, 416–427. Ifeanyi, F., and Takemoto, L. (1991). Exp. Eye Res. 53, 305–308. Ingolia, T. D., and Craig, E. A. (1982). Proc. Natl. Acad. Sci. USA 79, 2360–2364. Ito, H., Okamoto, K., Nakayama, H., Isobe, T., and Kato, K. (1997). J. Biol. Chem. 272, 29934–29941. Ito, H., Kamei, K., Iwamoto, I., Inaguma, Y., Nohara, D., and Kato, K. (2001). J. Biol. Chem. 276, 5346–5352. Iwaki, T., Iwaki, A., Tateishi, J., Sakaki, Y., and Goldman, J. E. (1993). Am. J. Pathol. 143, 487–495. Jackson, S. A., and Clegg, J. S. (1996). Dev. Growth Diff. 38, 153–160. Jaenicke, R., and Creighton, T. E. (1993). Curr. Biol. 3, 234–235. Jakob, U., and Buchner, J. (1994). Trends Biochem. Sci. 19, 205–211.

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

151

Jakob, U., Gaestel, M., Engel, K., and Buchner, J. (1993). J. Biol. Chem. 268, 1517–1520. Jinn, T.-L., Chen, Y.-M., and Lin, C.-Y. (1995). Plant Physiol. 108, 693–701. Jinn, T. L., Yeh, Y. C., Chen, Y. M., and Lin, C.-Y. (1989). Plant Cell Physiol. 30, 463–469. Joanisse, D. R., Inaguma, Y., and Tanguay, R. M. (1998a). Biochem. Biophys. Res. Commun. 244, 102–109. Joanisse, D. R., Michaud, S., Inaguma, Y., and Tanguay, R. M. (1998b). J. Biosci. 23, 369–376. Kampinga, H. H., Brunsting, J. F., Stege, G. J. J., Burgman, P. W. J. J., and Konings, A. W. T. (1995). Exp. Cell Res. 219, 536–546. Kantorow, M., and Piatigorsky, J. (1998). Int. J. Biol. Macromol. 22, 307–314. Kapp´e, G., Leunissen, J. A. M., and de Jong, W. W. (2001). In Small Stress Proteins (A.-P. Arrigo and W. E. G. Muller, ¨ eds.). Springer-Verlag, Heidelberg. Kapp´e, G., Verschuure, P., Philipsen, R. L. A., Staalduinen, A. A., van de Boogaart, P., Boelens, W. C., and de Jong, W. W. (2001). Biochim. Biophys. Acta 1520, 1–6. Kasi, V. S., and Kuppuswamy, D. (1999). Mol. Cell. Biol. 19, 6858–6871. Katayama-Fujimura, Y., Gottesman, S., and Maurizi, M. R. (1987). J. Biol. Chem. 262, 4477–4485. Kato, H., Liu, Y., Kogure, K., and Kato, K. (1994a). Brain Res. 634, 235–244. Kato, K., Goto, S., Hasegawa, K., Shinohara, H., and Inaguma, Y. (1993). Biochim. Biophys. Acta 1175, 257–262. Kato, K., Hasegawa, K., Goto, S., and Inaguma, Y. (1994b). J. Biol. Chem. 269, 11274–11278. Kato, K., Ito, H., Kamei, K., Inaguma, Y., Iwamoto, I., and Saga, S. (1998). J. Biol. Chem. 273, 28346–28354. Kato, K., Shinohara, H., Goto, S., Inaguma, Y., Morishita, R., and Asano, T. (1992). J. Biol. Chem. 267, 7718–7725. Kedersha, N. L., Gupta, M., Li, W., Miller, I., and Anderson, P. (1999). J. Cell Biol. 147, 1431–1442. Kim, K. K., Kim, R., and Kim, S. H. (1998a). Nature 394, 595–599. Kim, R., Kim, K. K., Yokota, H., and Kim, S. H. (1998b). Proc. Natl. Acad. Sci. USA 95, 9129–9133. Kirschner, M., Winkelhaus, S., Thierfelder, J. M., and Nover, L. (2000). Plant. J. 24, 397–411. Klemenz, R., and Gehring, W. (1986). Mol. Cell. Biol. 6, 2011–2019. Klemenz, R., Andres, A.-C., Fr¨ohli, E., Sch¨afer, R., and Aoyama, A. (1993). J. Cell Biol. 120, 639–645. Klemenz, R., Fr¨ohli, E., Steiger, R. H., Sch¨afer, R., and Aoyama, A. (1991). Proc. Natl. Acad. Sci. USA 88, 3652–3656. Knauf, U., Jakob, U., Engel, K., Buchner, J., and Gaestel, M. (1994). EMBO J. 13, 54–60. Kokke, B. P. A., Leroux, M. R., Candido, E. P. M., Boelens, W. C., and de Jong, W. W. (1998). FEBS Lett. 433, 228–232. Koteiche, H. A., Berengian, A. R., and Mchaourab, H. S. (1998). Biochemistry 37, 12681– 12688. Koteiche, H. A., and Mchaourab, H. S. (1999). J. Mol. Biol. 294, 561–577. Kraulis, P. J. (1991). J. Appl. Crystallogr. 24, 946–950. Krief, S., Faivre, J.-F., Robert, P., le Douarin, B., Brument-Larignon, N., Lefr`ere, I., Bouzyk, M. M., Anderson, K. M., Greller, L. D., Tobind, F. L., Souchet, M., and Bril, A. (1999). J. Biol. Chem. 274, 36592–36600. Kudva, Y. C., Hiddinga, H. J., Butler, P. C., Mueske, C. S., and Eberhardt, N. L. (1997). FEBS Lett. 416, 117–121. Kumar, L. V. S., Ramakrishna, T., and Rao, C. M. (1999). J. Biol. Chem. 274, 24137–24141.

152

ROB VAN MONTFORT ET AL.

Kumar, L. V. S., and Rao, C. M. (2000). J. Biol. Chem. 275, 22009–22013. Kurtz, S., Rossi, J., Petko, L., and Lindquist, S. (1986). Science 231, 1154–1157. Kurzik-Dumke, U., and Lohmann, F. (1995). Gene 154, 171–175. Lambert, H., Charette, S. J., Bernier, A. F., Guimond, A., and Landry, J. (1999). J. Biol. Chem. 274, 9378–9385. Landry, J., Chretien, P., Laszlo, A., and Lambert, H. (1991). J. Cell. Biochem. 147, 93–101. Lang, L., Miskovic, D., Fernando, P., and Heikkila, J. J. (1999). Dev. Genet. 25, 365–374. Lavoie, J. N., Gingras-Breton, G., Tanguay, R. M., and Landry, J. (1993a). J. Biol. Chem. 268, 3420–3429. Lavoie, J. N., Hickey, E., Weber, L. A., and Landry, J. (1993b). J. Biol. Chem. 268, 24210– 24214. Lavoie, J. N., Lambert, H., Hickey, E., Weber, L. A., and Landry, J. (1995). Mol. Cell. Biol. 15, 505–516. Lee, G. J. (1995). Methods Cell Biol. 50, 325–334. Lee, G. J., Pokala, N., and Vierling, E. (1995). J. Biol. Chem. 270, 10432–10438. Lee, G. J., Roseman, A. M., Saibil, H. R., and Vierling, E. (1997). EMBO J. 16, 659–671. Lee, G. J., and Vierling, E. (1998). Methods Enzymol. 290, 350–365. Lee, G. J., and Vierling, E. (2000). Plant Physiol. 122, 189–197. Lee, S., Owen, H. A., Prochaska, D. J., and Barnum, S. R. (2000). Curr. Microbiol. 40, 283–287. Lee, S., Prochaska, D. J., Fang, F., and Barnum, S. R. (1998). Curr. Microbiol. 37, 403–407. Leicht, B. G., Biessmann, H., Palter, K. B., and Bonner, J. J. (1986). Proc. Natl. Acad. Sci. USA 83, 90–94. Leroux, M. R., Ma, B. J., Batelier, G., Melki, R., Peter, E., and Candido, M. (1997a). J. Biol. Chem. 272, 12847–12853. Leroux, M. R., Melki, R., Gordon, B., Batelier, G., and Candido, E. P. M. (1997b). J. Biol. Chem. 272, 24646–24656. Lewis, S. E., Mannion, R. J., White, F. A., Coggeshall, R. E., Beggs, S., Costigan, M., Martin, J. L., Dillmann, W. H., and Woolf, C. J. (1999). J. Neurosci. 19, 8945–8953. Liang, J. J. N., and Li, X.-Y. (1992). Exp. Eye Res. 54, 719–724. Liang, P., Amons, R., MacRae, T. H., and Clegg, J. S. (1997). Eur. J. Biochem. 243, 225–232. Liddington, R. C., Yan, Y., Moulai, J., Sahil, R., Benjamin, T. L., and Harrison, S. C. (1991). Nature 354, 278–284. Linder, B., Jin, Z. J., Freedman, J. H., and Rubin, C. S. (1996). J. Biol. Chem. 271, 30158– 30166. Lindner, R. A., Carver, J. A., Ehrnsperger, M., Buchner, J., Esposito, G., Behlke, J., Lutsch, G., Kotlyarov, A., and Gaestel, M. (2000). Eur. J. Biochem. 267, 1923–1932. Lindner, R. A., Kapur, A., and Carver, J. A. (1997). J. Biol. Chem. 272, 27722–27729. Litt, M., Kramer, P., LaMorticella, D. M., Murphey, W., Lovren, E. W., and Weleber, R. G. (1999). Hum. Mol. Genet. 7, 471–474. Longoni, S., Lattonen, S., Bullock, G., and Chiesi, M. (1990). Mol. Cell. Biochem. 97, 121–128. Loones, M.-T., Chang, Y., and Morange, M. (2000). Cell Stress & Chaperones 5, 291–305. Lorimer, G. H., and Baldwin, T. O. (1998). Molecular Chaperones. Academic Press, New York. Lowe, J., McDermott, H., Pike, I., Spendlove, I., Landon, M., and Mayer, R. J. (1992). J. Pathol. 166, 61–68. MacRae, T. H. (2000). Cell. Mol. Life Sci. 57, 899–913. Marin, R., and Tanguay, R. M. (1996). Chromosoma 105, 142–149.

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

153

Mchaourab, H. S., Berengian, A. R., and Koteiche, H. A. (2000). Biochemistry 36, 14627– 14634. Mehlen, P., Mehlen, A., Guillet, D., Preville, X., and Arrigo, A.-P. (1995). J. Cell. Biochem. 58, 248–259. Merck, K. B., de Haard-Hoekman, W. A., Oude Essink, B. B., Bloemendal, H., and de Jong, W. W. (1992). Biochim. Biophys. Acta Gene Struct. Expression 1130, 267–276. Merck, K. B., Groenen, P. J. T. A., Voorter, C. E. M., De Haard-Hoekman, W. A., Horwitz, J., Bloemendal, H., and De Jong, W. W. (1993a). J. Biol. Chem. 268, 1046–1052. Merck, K. B., Horwitz, J., Kerstein, M., Overkamp, P., Gaestel, M., Bloemendal, H., and De Jong, W. W. (1993b). Mol. Biol. Rep. 18, 209–215. Michelini, A. T., and Flynn, G. C. (1999). J. Bacteriol. 181, 4237–4244. Minowada, G., and Welch, W. (1995). J. Biol. Chem. 270, 7047–7054. Miron, T., Vancompernolle, K., Vandekerckhove, J., Wilchek, M., and Geiger, B. (1991). J. Cell Biol. 114, 255–261. Miron, T., Wilchek, M., and Geiger, B. (1988). Eur. J. Biochem. 178, 543–553. Moerman, A. M., and Klein, C. (1997). Exp. Cell Res. 237, 149–157. Mogk, A., Tomoyasu, T., Goloubinoff, P., Rudiger, ¨ S., R¨oder, D., Langen, H., and Bukau, B. (1999). EMBO J. 18, 6934–6949. Morrow, G., Inaguma, Y., Kato, K., and Tanguay, R. M. (2000). J. Biol. Chem. 275, 31204– 31210. Muchowski, P. J., and Clark, J. I. (1998). Proc. Natl. Acad. Sci. USA 95, 1004–1009. Muchowski, P. J., Hays, L. G., Yates, J. R., III, and Clark, J. I. (1999a). J. Biol. Chem. 274, 30190–30195. Muchowski, P. J., Wu, G. J. S., Liang, J. J. N., Adman, E. T., and Clark, J. I. (1999b). J. Mol. Biol. 289, 397–411. Mulders, J. W., Stokkermans, J., Leunissen, J. A., Benedetti, E. L., Bloemendal, H., and de Jong, W. W. (1985). Eur. J. Biochem. 152, 721–728. Munchbach, M., Nocker, A., and Narberhaus, F. (1999). J. Bacteriol. 181, 83–90. Nakamoto, H., Suzuki, N., and Roy, S. K. (2000). FEBS Lett. 483, 169–174. Neumann, D., zur Nieden, U., Manteuffel, R., Walter, G., Scharf, K.-D., and Nover, L. (1987). Eur. J. Cell Biol. 43, 71–81. Nicholl, I. D., and Quinlan, R. A. (1994). EMBO J. 13, 945–953. Nover, L., Scharf, K.-D., and Neumann, D. (1989). Mol. Cell. Biol. 9, 1298–1308. Ohan, N. W., and Heikkila, J. J. (1995). Dev. Genet. 17, 176–184. Osteryoung, K. W., and Vierling, E. (1994). J. Biol. Chem. 269, 28676–28682. Palmisano, D. V., Groth-Vasselli, B., Farnsworth, P. N., and Reddy, M. C. (1995). Biochim. Biophys. Acta 1246, 91–97. Pandey, P., Farber, R., Nakazawa, A., Kumar, S., Bharti, A., Nalin, C., Weichselbaum, R., Kufe, D., and Kharbanda, S. (2000). Oncogene 19, 1975–1981. Perng, M. D., Cairns, L., Van den Ijssel, P., Prescott, A., Hutcheson, A. M., and Quinlan, R. A. (1999). J. Cell Sci. 112, 2099–2112. Perry, K. M., Pookanjanatavip, M., Zhoa, J., Santi, D. V., and Stroud, R. M. (1992). Protein Sci. 1, 800. Petko, L., and Lindquist, S. (1985). Cell 45, 885–894. Piotrowicz, R. S., and Levin, E. G. (1997). J. Biol. Chem. 272, 25920–25927. Plesofsky-Vig, N., and Brambl, R. (1995). Proc. Natl. Acad. Sci. USA 92, 5032–5036. Plesofsky-Vig, N., and Brambl, R. (1998). J. Biol. Chem. 273, 11335–11341. Plesofsky, N., and Brambl, R. (1999). Biochim. Biophys. Acta Mol. Cell Res. 1449, 73–82. Plesofsky, N., Gardner, N., Lill, R., and Brambl, R. (1999). Biol. Chem. 380, 1231–1236.

154

ROB VAN MONTFORT ET AL.

Plumier, J. C. L., Armstrong, J. N., Landry, J., Babity, J. M., Robertson, H. A., and Currie, R. W. (1996). Neuroscience 75, 849–856. Plumier, J. C. L., Hopkins, D. A., Robertson, H. A., and Currie, R. W. (1997). J. Comp. Neurol. 384, 409–428. Prieto-Dapena, P., Almoguera, C., Rojas, A., and Jordano, J. (1999). Plant Mol. Biol. 39, 615–627. Quinlan, R. (2001). Nature Genet. 27, 10–11. Raman, B., Ramakrishna, T., and Rao, C. M. (1997). FEBS Lett. 416, 369–372. Raman, B., and Rao, C. M. (1994). J. Biol. Chem. 269, 27264–27268. Raman, B., and Rao, C. M. (1997). J. Biol. Chem. 272, 23559–23564. Rao, C. M., Raman, B., Ramakrishna, T., Rajaraman, K., Ghosh, D., Datta, S., Trivedi, V. D., and Sukhaswami, M. B. (1998). Int. J. Biol. Macromol. 22, 271–281. Rao, P. V., Horwitz, J., and Zigler, J. S., Jr. (1993). Biochem. Biophys. Res. Commun. 190, 786–793. Rogalla, T., Ehrnsperger, M., Preville, X., Kotlyarov, A., Lutsch, G., Ducasse, C., Paul, C., Wieske, M., Arrigo, A. P., Buchner, J., and Gaestel, M. (1999). J. Biol. Chem. 274, 18947–18956. Rouse, J., Cohen, P., Trigon, S., Morange, M., Alonso-Llamazares, A., Zamanillo, D., Hunt, T., and Nebreda, A. R. (1994). Cell 78, 1027–1037. Roy, S. K., Hiyama, T., and Nakamoto, H. (1999). Eur. J. Biochem. 262, 406–416. Scharf, K.-D., Siddique, M., and Vierling, E. (2001). Cell Stress & Chaperones 6, 225–237. Sch¨onfeld, H. J., and Behlke, J. (1998). Methods Enzymol. 290, 269–296. S´ebastein, M., Genevi`eve, M., Julie, M., and Tanguay, R. M. (2001). In Small Stress Proteins (A.-P. Arrigo and W. E. G. Muller, ¨ eds.), 2001. Springer-Verlag, Heidelberg. Shakoori, A. R., Oberdorf, A. M., Owen, T. A., Weber, L. A., Hickey, E., Stein, J. L., Lian, J. B., and Stein, G. S. (1992). J. Cell. Biochem. 48, 277–287. Sharma, K. K., Kaur, H., and Kester, K. (1997). Biochem. Biophys. Res. Commun. 239, 217–222. Sharma, K. K., Kaur, H., Kumar, G. S., and Kester, K. (1998a). J. Biol. Chem. 273, 8965– 8970. Sharma, K. K., Kumar, G. S., Murphy, A. S., and Kester, K. (1998b). J. Biol. Chem. 273, 15474–15478. Sharma, K. K., Kumar, R. S., Kumar, G. S., and Quinn, P. T. (2000). J. Biol. Chem. 275, 3767–3771. Shearstone, J. R., and Baneyx, F. (1999). J. Biol. Chem. 274, 9937–9945. Slingsby, C., and Bateman, O. A. (1994). Exp. Eye Res. 58, 761–764. Smith, J. B., Liu, Y. Q., and Smith, D. L. (1996). Exp. Eye Res. 63, 125–127. Smulders, R. H. P. H., and de Jong, W. W. (1997). FEBS Lett. 409, 101–104. Smykal, P., Masin, J., Hrdy, I., Konop´asek, I., and Z´arsky, V. (2000). Plant J. 23, 703– 713. Soto, A., Allona, I., Collada, C., Guevara, M.-A., Casado, R., Rodriguez-Cerezo, E., Aragoncillo, C., and Gomez, L. (1999). Plant Physiol. 120, 521–528. Stehle, T., Gamblin, S. J., Yan, Y., and Harrison, S. C. (1996). Structure 4, 165–182. Studer, S., and Narberhaus, F. (2000). J. Biol. Chem. 275, 37212–37218. Sugiyama, Y., Suzuki, A., Kishikawa, M., Akutsu, R., Hirose, T., Waye, M. M., Tsui, S. K., Yoshida, S., and Ohno, S. (2000). J. Biol. Chem. 275, 1095–1104. Sun, T.-S., Das, B. K., and Liang, J. J. N. (1997). J. Biol. Chem. 272, 6220–6225. Sun, T.-S., and Liang, J. J. N. (1998). J. Biol. Chem. 273, 286–290. Sun, T.-X., Akhtar, N. J., and Liang, J. J. N. (1998). FEBS Lett. 430, 401–404. Susek, R. E., and Lindquist, S. L. (1989). Mol. Cell. Biol. 9, 5265–5271.

STRUCTURE AND FUNCTION OF THE SMALL HEAT SHOCK PROTEINS

155

Suzuki, A., Sugiyama, Y., Hayashi, Y., Nyu-i, N., Yoshida, M., Nonaka, I., Ishiura, S., Arahata, K., and Ohno, S. (1998). J. Cell Biol. 140, 1113–1124. Tanguay, R. M., Wu, Y., and Khandjian, E. W. (1993). Dev. Genet. 14, 112–118. Th´eriault, J. R., Charette, S. J., Lambert, H., and Landry, J. (1999). FEBS Lett. 452, 170– 176. Thomas, J. G., and Baneyx, F. (1998). J. Bacteriol. 180, 5165–5172. Topping, T. B., Woodbury, R. L., Diamond, D. L., Hardy, S. J. S., and Randall, L. L. (2001). J. Biol. Chem. 276, 7437–7441. T¨or¨ok, Z., Goloubinoff, P., Tsvetkova, N. M., Glatz, A., Balogh, G., Varvasovszki, V., Los, D. A., Vierling, E., Crowe, J. H., and Vigh, L. (2001). Proc. Natl. Acad. Sci. 98, 3098–3103. van Boekel, M. A. M., De Lange, F., De Grip, W. J., and de Jong, W. W. (1999). Biochim. Biophys. Acta Protein Struct. Mol. Enzymol. 1434, 114–123. van dan Ijssel, P., Norman, D. G., and Quinlan, R. A. (1999). Curr. Biol. 9, R103–R109. van de Klundert, F. A. J. M., and de Jong, W. W. (1999). Eur. J. Cell Biol. 78, 567–572. van de Klundert, F. A. J. M., Smulders, R. H. P. H., Gijsen, M. L. J., Lindner, R. A., Jaenicke, R., Carver, J. A., and de Jong, W. W. (1998). Eur. J. Biochem. 258, 1014–1021. van den Ijssel, P. R. L. A., Overkamp, P., Bloemendal, H., and de Jong, W. W. (1998). Biochem. Biophys. Res. Commun. 247, 518–523. van den Oetelaar, P. J. M., van Someren, P. F. H. M., Thomson, J. A., Siezen, R. J., and Hoenders, H. J. (1990). Biochemistry 29, 3488–3493. van Montfort, R. L. M., Basha, E., Friedrich, K. L., Slingsby, C., and Vierling, E. (2001). Crystal structure and assembly of a eukaryotic small heat shock protein. Nat. Struct. Biol. in press. van Noort, J. M., Van Sechel, A. C., Bajramovic, J. J., El Ouagmiri, M., Polman, C. H., Lassmann, H., and Ravid, R. (1995). Nature 375, 798–801. Veinger, L., Diamant, S., Buchner, J., and Goloubinoff, P. (1998). J. Biol. Chem. 273, 11032–11037. Verbon, A., Hartskeerl, R. A., Schuitema, A., Kolk, A. H. J., Young, D. B., and Lathigra, R. (1992). J. Bacteriol. 174, 1352–1359. Vicart, P., Caron, A., Guicheney, P., Li, Z. L., Pr´evost, M. C., Faure, A., Chateau, D., Chapon, F., Tom´e, F., Dupret, J. M., Paulin, D., and Fardeau, M. (1998). Nature Genet. 20, 92–95. Vierling, E. (1991). Annu. Rev. Plant Physiol. Plant Mol. Biol. 42, 579–620. Vierling, E. (1997). Acta Physiol. Plant. 19, 539–547. Vigh, L., Maresca, B., and Harwood, J. L. (1998). Trends Biochem. Sci. 23, 369–374. Voorter, C. E. M., de Haard-Hoekman, W. A., Roersma, E. S., Meyer, H. E., Bloemendal, H., and de Jong, W. W. (1989). FEBS Lett. 259, 50–52. Walsh, D., Li, Z., Wu, Y., and Nagata, K. (1997). Cell. Mol. Life Sci. 53, 198–211. Walsh, M. T., Sen, A. C., and Chakrabarti, B. (1991). J. Biol. Chem. 266, 20079–20084. Wang, K., Ma, W., and Spector, A. (1995). Exp. Eye Res. 61, 115–124. Wang, K. Y., and Spector, A. (2000). Eur. J. Biochem. 267, 4705–4712. Wang, X. Y., Chen, X., Oh, H. J., Repasky, E., Kazim, L., and Subjeck, J. (2000). FEBS Lett. 465, 98–102. Waters, E. (1995). Genetics 141, 785–795. Waters, E. R., Lee, G. J., and Vierling, E. (1996). J. Exp. Bot. 47, 325–338. Waters, E. R., and Vierling, E. (1999a). Proc. Natl. Acad. Sci. USA 96, 14394–14399. Waters, E. R., and Vierling, E. (1999b). Mol. Biol. Evol. 16, 127–139. Wehmeyer, N., Hernandez, L. D., Finkelstein, R. R., and Vierling, E. (1996). Plant Physiol. 112, 747–757.

156

ROB VAN MONTFORT ET AL.

Wehmeyer, N., and Vierling, E. (2000). Plant Physiol. 122, 1099–1108. Welsh, M. J., and Gaestel, M. (1998). Ann. NY Acad. Sci. 851, 28–35. Wente, S. R., and Schachman, H. K. (1987). Proc. Natl. Acad. Sci. USA 84, 31–35. Wisniewski, T., and Goldman, J. E. (1998). Neurochem. Res. 23, 385–392. Wistow, G. (1993). Trends Biochem. Sci. 18, 301–306. Wotton, D., Freeman, K., and Shore, D. (1996). J. Biol. Chem. 271, 2717–2723. Xanthoudakis, S., and Nicholson, D. W. (2000). Nat. Cell Biol. 2, E164–E165. Yamamoto, M., Fan, L. Y., Wakayama, T., Amano, O., and Iseki, S. (2001). Anat. Rec. 262, 213–220. Yang, H. M., Huang, S. F., Dai, H. Z., Gong, Y. D., Zheng, C. X., and Chang, Z. Y. (1999). Protein Sci. 8, 174–179. Yeh, C. H., Chang, P. F. L., Yeh, K. W., Lin, W. C., Chen, Y. M., and Lin, C. Y. (1997). Proc. Natl. Acad. Sci. USA 94, 10967–10972. Young, L.-S., Yeh, C.-H., Chen, Y.-M., and Lin, C.-Y. (1999). Biochem. J. 344, 31–38. Yuan, Y., Crane, D. D., Simpson, R., Zhu, Y. Q., Hickey, M. J., Sherman, D. R., and Barry III, C. E. (1998). Proc. Natl. Acad. Sci. USA 95, 9578–9583. Zantema, A., Verlaan-De Vries, M., Maasdam, D., Bol, S., and Van der EB, A. (1992). J. Biol. Chem. 267, 12936–12941. Zhang, W. Z., and Augusteyn, R. C. (1994). Curr. Eye Res. 13, 225–230. Zheng, J., Luo, W., and Tanzer, M. L. (1998). J. Biol. Chem. 273, 12999–13006.

STRUCTURE, FUNCTION, AND MECHANISM OF THE Hsp90 MOLECULAR CHAPERONE By LAURENCE H. PEARL and CHRISOSTOMOS PRODROMOU Section of Structural Biology, Institute of Cancer Research, Chester Beatty Laboratories, 237 Fulham Road, London SW3 6JB, UK

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Domain Structure and Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The N-Terminal Nucleotide-Binding Domain . . . . . . . . . . . . . . . . . . . . . . III. ATP Binding and Hydrolysis by Hsp90 Are Essential in Vivo . . . . . . . . . . . . . . IV. Conformational Changes in Hsp90 Accompanying the ATPase Cycle . . . . . . V. Hsp90 ATPase Inhibitors—A New Class of Antitumor Drugs . . . . . . . . . . . . . VI. Interaction with Co-chaperones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Regulation of ATP Binding and Hydrolysis in the Client-Protein Activation Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Interactions with Alterations of Client Proteins by Hsp90 . . . . . . . . . . . . . . . . IX. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

157 158 161 164 166 170 173 175 176 180 181

I. INTRODUCTION The Hsp90 proteins are ubiquitous molecular chaperones found in eubacteria and all branches of eukarya, but apparently absent from the archaea. In eukaryotes, a functional gene for a cytoplasmic Hsp90 isoform is absolutely essential for viability under all conditions (Borkovich et al., 1989; Cutforth and Rubin, 1994), whereas the bacterial homolog HtpG is dispensable for normal growth (Spence et al., 1990). Cytoplasmic Hsp90 is an abundant protein found at levels as high as 1% of the total soluble cell protein, even under nonstressed conditions (Welch and Feramisco, 1982). Isolated Hsp90 shows an ability to slow the aggregation of denatured proteins in vitro (Wiech et al., 1992), but the relevance of this activity to Hsp90 function in vivo remains to be established. In vivo, Hsp90 interacts with proteins that have already attained a high degree of tertiary structure, and appears to be involved in late-stage maturation and activation of these “client” proteins rather than their initial folding. The known client proteins include steroid hormone receptors ( Joab et al., 1984), helix– loop–helix transcription factors (Wilhelmsson et al., 1990), tyrosine and serine/threonine kinases (Aligue et al., 1994; Cutforth and Rubin, 1994; Dai et al., 1996; Opperman et al., 1981; Stancato et al., 1993) and tumor suppressors (Chen et al., 1996a; Sepehrnia et al., 1996). Comprehensive 157 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0096-5332/02 $35.00

158

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

lists of client proteins can be found in Buchner (1999), Csermely et al. (1998), and Pratt (1998). The involvement of Hsp90 in many signaling and cell regulatory pathways gives it a role as a capacitor for morphological evolution by buffering cryptic genetic variation (Rutherford and Lindquist, 1998). Thus, Hsp90 is believed to maintain the “normal” behavior of mutated signaling proteins whose altered phenotype becomes manifest only when Hsp90 function is compromised. In addition to the cytoplasmic Hsp90 found in all eukaryotes, higher eukaryotes also possess distinct isoforms localized to the endoplasmic reticulum (ER) and to mitochondria. GRP94/GP96 (or endoplasmin) serves as a molecular chaperone in the endoplasmic reticulum, where it is implicated in the maturation of proteins such as p185ErbB2, nascent immunoglobulins, MHC class II, collagen, HSV-1 glycoproteins, and thyroglobulin (Chavany et al., 1996; Ferreira et al., 1994; Kuznetsov et al., 1994; Melnick et al., 1992; Navarro et al., 1991; Schaiff et al., 1992). GRP94 is also implicated in binding peptides destined for presentation by MHC class I molecules (Ishii et al., 1999; Nieland et al., 1996; Sastry and Linderoth, 1999; Srivastava et al., 1986; 1998; Wearsch and Nicchitta, 1997; Wearsch et al., 1998). TRAP1 (previously Hsp75) has recently been shown to have a predominantly mitochondrial localization (Felts et al., 2000), although no role in mitochondrial function has yet been established. Originally, TRAP1 was identiﬁed as a protein interacting with type I tumor necrosis factor receptor (Song et al., 1995) and retinoblastoma (Rb) protein (Chen et al., 1996a), roles that are difﬁcult to reconcile with its mitochondrial localization. In this chapter we discuss the current knowledge of the structure and biochemistry of the Hsp90 family of molecular chaperones, with particular attention to the emerging understanding of the role of its ATPase activity and the opportunities this presents for anti-chaperone drug development. We go on to discuss the various co-chaperones that are known to interact with Hsp90, examining their roles in Hsp90 function as well as the complicated picture of dynamic and varied Hsp90/co-chaperone complexes that is emerging. Finally, we discuss the current understanding of the interaction of Hsp90 with its client proteins, the possible origins of client-protein speciﬁcity, and the change of state in the client proteins that is achieved by their involvement in the ATP-dependent chaperone cycle of Hsp90. II. DOMAIN STRUCTURE AND FUNCTION A full understanding of the biochemistry of Hsp90 will come only with a detailed knowledge of its atomic structure, and considerable effort has been expended in the last few years by our lab and others to achieve

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

159

FIG. 1. Domain structure of the Hsp90 family.

this goal. We have succeeded in crystallizing the heat-shock protein 90 (Hsp90) isoform of budding yeast (Prodromou et al., 1996) and the Escherichia coli homolog HtpG, but so far these have failed to provide sufﬁcient diffraction for structure determination. Analysis of sequence conservation and proteolysis studies of Hsp90s have indicated the presence of at least three identiﬁable domains (Fig. 1). The N-terminal domain consisting of residues 1–220 (yeast) is very highly conserved in sequence among the Hsp90 family, the yeast Hsp90 sharing 62% sequence identity with human Hsp90α/β and 38% identity with the E. coli HtpG protein. This domain binds adenine nucleotides and is essential for the ATP-dependent function of the chaperone in vivo, and its high-resolution crystal structure has been determined for the yeast HSP82 (Prodromou et al., 1997b) and human Hsp90α isoforms (Stebbins et al., 1997). The N-terminal domain is structurally homologous to N-terminal domains of bacterial topoisomerases (Wigley et al., 1991), the MutL family of DNA-repair proteins (Ban et al., 1999), and to the central histidine kinase domains of several bacterial signal transduction proteins (Bilwes et al., 1999; Tanaka et al., 1998). Sequences homologous to the Hsp90 N-terminal domain have also been identiﬁed in sacsin (SACS), a large protein associated

160

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

with autosomal recessive spastic ataxia of Chalevoix-Saguenay (ARSACS) (Engert et al., 2000). In the endoplasmic reticulum–speciﬁc Hsp90 homologs GRP94/gp96 (or endoplasmin), the conserved nucleotide-binding domain is preceded by an N-terminal region of ∼70 residues. Of these, residues ∼3–21 are very hydrophobic and probably constitute the signal for ER import; the KDEL signal required for retention in the ER is found at the extreme C terminus of GRP94. The role of the remaining residues of this N-terminal extension is unknown. A much shorter N-terminal extension is found in those cytoplasmic Hsp90 isoforms from higher eukaryotes, whose expression is strongly upregulated by heat shock. This shorter region contains conserved Ser/Thr residues, which are found to be phosphorylated by the DNA-dependent protein kinase (DNA-PK) involved in the regulation of DNA double-strand break repair (Walker et al., 1985). The effect of this phosphorylation on Hsp90 function (if any) is unknown; nor is it clear if it plays any role in vivo. However, the presence of these potential sites of phosphorylation only in higher eukaryotes that have DNA-PK, but not in the lower eukaryotes that lack it, does suggest some signiﬁcance. In the eukaryotic Hsp90s, the N-terminal nucleotide-binding domain is connected to the remainder of the protein by a highly charged and proteolytically sensitive segment that is variable both in length and composition between different species and between different isoforms in the same species. This “linker” region is considerably shorter in the bacterial HtpGs and in the mammalian mitochondrial homolog TRAP1. Deletion or mutation of the charged linker does not affect ATP binding and hydrolysis by yeast Hsp90 in vitro (L. H. Pearl and C. Prodromou, unpublished data) and is dispensable for Hsp90-dependent cell viability in vivo (Louvain et al., 1996). This segment appears to function as a simple covalent tether, connecting the N-terminal domain to the rest of the protein. Nonetheless, the conserved presence of this extended linker in eukaryotic Hsp90s, but not in bacterial HtpGs or TRAP1, and the conservation of its general nature suggest some subtle role speciﬁc to the eukaryotic cytoplasmic Hsp90s that remains to be detected. At the opposite end of the protein, proteolysis and yeast two-hybrid screens (Young et al., 1998) have identiﬁed a C-terminal domain of ∼12kDa to which two essential functions can be ascribed. The Cterminal region provides a strong inherent dimerization interface, which is essential for function. The C-terminal–most region of this domain, which is present only in the eukaryotic cytosolic Hsp90s, provides the binding site for a subset of Hsp90 co-chaperones, some of which possess tetratricopeptide repeat (TPR) domains (Chen et al., 1998; Young et al., 1998). Structural studies of some of these TPR domains (Das et al., 1998;

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

161

Scheuﬂer et al., 2000) have shown that tandem repeats of this degenerate 34-residue motif form consecutive helical hairpins, generating a concave peptide-binding groove. Binding to Hsp90 by these TPR-domain co-chaperones is mediated by a short sequence (MEEVD) occurring at the extreme C terminus of Hsp90, which binds speciﬁcally in the groove of the TPR domain (Carrello et al., 1999; Scheuﬂer et al., 2000). While the C-terminal “tag” is essential for TPR-domain binding, at least for the TPR-domain co-chaperone p60/Hop/Sti1, other regions of Hsp90 are also involved (Chen et al., 1998; Prodromou et al., 1999). The binding of TPR proteins and CDC37 has been shown to be mutually exclusive. However, the binding site of CDC37 is not the same as that for TPR proteins and is either very close to or overlapping that for TPR proteins (Owens-Grillo et al., 1996a; 1996b; Silverstein et al., 1998). Furthermore, the binding sites for different TPR proteins are not identical, but overlap (Ramsey et al., 2000). Between the N-terminal nucleotide-binding domain and the C-terminal dimerization/TPR-binding domain is the central region of ∼45 kDa. Limited proteolysis studies of this region (L. H. Pearl and C. Prodromou, unpublished data) identiﬁed a stable ∼35-kDa segment or middle domain, which can be expressed independently in bacteria as a soluble folded product. This region has been crystallized, and diffraction data to ∼3.5-A˚ resolution have been obtained. Circular dichroism and NMR spectroscopic analysis show a high content of α-helical structure in this domain. Although the middle domain comprises the majority of the protein, its role in Hsp90 chaperone function is not known. Most likely, it provides the binding site for client proteins associated with Hsp90, but until recently few experimental data justiﬁed that assertion. Recently, a highly conserved segment (327–340 in human Hsp90β) has been implicated in binding to protein kinase B/Akt (Sato et al., 2000) which had been implicated as an Hsp90 client by its sensitivity to Hsp90 inhibitors (Clarke et al., 2000). Sequence patterns in the middle domain (amino acid residues Met402–Leu423 in Hsp90α) characteristic of coiled-coil or “leucine-zipper” structure have been noted (Nemoto et al., 1995), but the role of these putative leucine zippers remains unclear. Whatever the role of the middle domain of Hsp90, it is nonetheless essential to function, and a point mutation within this region (E381K in yeast HSP82) generates a temperature-sensitive loss of viability in yeast (Nathan and Lindquist, 1995). A. The N-Terminal Nucleotide-Binding Domain While the N-terminal domain represents less than a third of the Hsp90 protein, the elucidation of its structure (Prodromou et al., 1997a; 1997b;

162

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

Stebbins et al., 1997) has provided one of the major advances in understanding the biochemistry of these chaperones. The N-terminal domain consists of a highly curved β-sheet exposed on one face, with the opposite face covered by a set of α-helices (Fig. 2A,B, see color insert). The strands in the β-sheet are primarily antiparallel except between strands b and h. Strands d, e, f, and g form a greek key motif, with no intervening secondary structural elements. In the yeast Hsp90 structure, we observed a dimerization interface between the free C-terminal ends of the N-terminal domain construct, which generated a narrow channel formed by the curved faces of the exposed β-sheet. The dimensions of the channel corresponded roughly to those of a hexa- or heptapeptide in an extended conformation, and suggested a possible role for the N terminus in peptide binding. This idea was consistent with observations that Hsp90 and GRP94 could bind peptide antigens (Wearsch et al., 1998) and was subsequently supported by direct observation of peptide binding by the isolated N-terminal domain (Scheibel et al., 1998; Young et al., 1997). Set against this interpretation, the regions of the structure involved in formation of this channel were among the least conserved, whereas the exposed helical faces of the N-terminal domain dimer were very highly conserved. Furthermore, no signiﬁcant dimerization of the isolated N-terminal domain in solution can be detected either by gel ﬁltration or crosslinking (Prodromou et al., 2000), and no such strand-swap interface was observed in the structure of the human Hsp90α N domain, which was a few residues shorter (Stebbins et al., 1997). In hindsight, it is likely that this strand-swap dimerization interface was an artifact of the crystallization of an isolated domain and has no biological relevance. The helical face of the N-terminal domain contains a deep pocket bounded on two sides by the helices formed from amino acid residues 28–50 and 85–94, and on the other sides by the end of the helix and loop from residues 117–124 and the loop from residues 81–85 (Fig. 2C). The base of the pocket is formed by residues Ile-77, Asp-79, Val-136, Ser-138, Thr-171, and Ile-173, whose side chains project up from the buried face of the β-sheet. At the time of our original study, we observed electron density in this pocket, which we were unable to interpret in terms of any known component of the crystallization buffer, although part of it could be reﬁned with some success as a bound glycerol molecule. In studies of the N-terminal domain of human Hsp90α (Stebbins et al., 1997), this pocket was found to be the binding site for the antitumor antibiotic geldanamycin, a drug previously shown to achieve its effect by impairing the Hsp90-dependent activation of key proliferative molecules such as Raf-1 (Schulte et al., 1995; Stancato et al., 1997) and cyclin-dependent protein kinases (Stepanova et al., 1996). Binding to Hsp90 was interpreted as geldanamycin acting as a peptide mimetic, imitating the

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

163

conformation of a β turn and implicating the N-terminal domain as a participant in client-protein binding. Analysis of the shape of the pocket in terms of peptide binding suggested possible speciﬁcity for the indole ring of tryptophan, which is usually buried in the core of proteins. Thus, it was suggested that Hsp90 might act to bind exposed hydrophobic regions of client proteins containing tryptophan, in an analogous manner to the binding of hydrophobic peptides containing leucine by Hsp70 molecular chaperones (Stebbins et al., 1997), and that geldanamycin achieved its effect by preventing such binding. Initially, no resemblance was recognized between the structures of Hsp90 N-terminal domains and any other known protein structure present in the public structural databases. However, at the sequence level, several conserved sequence motifs common to the N-terminal domains of Hsp90s, Type II and Type IV DNA topoisomerases and bacterial DNA gyrase B protein, and the MutL family of DNA mismatch repair proteins were recognized (Bergerat et al., 1997). At the time, only the N-terminal domain structure of DNA gyrase B was known (Wigley et al., 1991); however, the coordinates for this structure had not been deposited in the Protein Databank and thus were not found in the structural similarity searches employed in the N-domain studies. These sequence motifs corresponded to residues involved in the binding and hydrolysis of ATP in DNA gyrase B (Ali et al., 1993), suggesting the possibility that the corresponding domains in Hsp90 and MutL might also be involved in the binding of adenine nucleotides. In fact, ATP binding by both these types of protein had been controversial, especially so in the case of Hsp90. In early studies of Hsp90 (Nadeau et al., 1992; 1993) a high level of ATPase activity had been observed, but was subsequently shown to be due to a contaminating kinase. In vitro studies of highly puriﬁed Hsp90 had failed to observe any ATPase activity or binding to ATP analogs (Jakob et al., 1996). On the basis of the common Hsp90/Gyrase/MutL motifs, we cocrystallized the N-terminal domain of yeast Hsp90 with ATP, the nonhydrolyzable ATP analog ATP-γ S, or ADP, all in the presence of Mg2+, and determined the structures (Prodromou et al., 1997a). In all cases, an adenine nucleotide with at least two phosphates was clearly present in the deep pocket identiﬁed as the geldanamycin-binding site in the human N-terminal domain structure (Fig. 3A, see color insert). The nucleotide-binding site is unusual as compared to those previously seen in kinases and other ATP-dependent chaperones such as Hsp70 (Flaherty et al., 1994). The nucleotide itself adopts a “closed” conformation in which the C8 position of the ﬁve-membered imidazole ring is shielded by the C5′ methylene and α-phosphate, thereby preventing binding of the C8-modiﬁed ATP analogs used in earlier in vitro binding studies ( Jakob

164

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

et al., 1996). Direct nucleotide interactions with the protein are limited to a single hydrogen bond from the side chain of Asp-79 to the exocyclic amino group of the adenine; a hydrogen bond from the side chain of Asn-92 to the 2′ hydroxyl of the ribose; hydrogen bonds from the main chain carboxyl group of Gly-121, the main chain amide group of Phe-124, and the side chain amide group of Asn-37 to an oxygen of the α-phosphate of ATP; and ﬁnally hydrogen bonds from the side chains of Asn-37 and Lys-98 to an oxygen of the β-phosphate of ATP. All other interactions are via solvent molecules, several of which are observed in the same positions in the unliganded structure, and via a Mg2+ ion which links the α- and β-phosphates of ATP and the side chain of Asn-37. (Fig. 3B) Electron density for the γ -phosphate of ATP-γ S was never observed, indicating that the γ -phosphate of ATP is disordered and its binding in the protein is due to residues outside the N-terminus. One immediate consequence of these observations was the realization that the geldanamycin and adenine nucleotide-binding sites totally overlapped, and that geldanamycin must therefore be a competitive inhibitor of the binding of adenine nucleotides to Hsp90 rather than peptides, as had been suggested (Stebbins et al., 1997). Toft and colleagues reached similar conclusions in biochemical studies, where they were able to demonstrate that Hsp90s would bind ATP-agarose, but only when attached via the γ -phosphate rather than the more conventional C8-ATPagarose previously used (Jakob et al., 1996), and that geldanamycin could compete with this binding (Grenert et al., 1997). Adenine nucleotides bind to Hsp90 with K ds in the range of 30–100 μM, depending on the nucleotide and the method used (Panaretou et al., 1998; Prodromou et al., 1997a; 2000), and binding is absolutely dependent on Mg2+. While this concentration is low compared to that of many ATP-binding proteins, it is well below cellular ATP concentrations (2–5 mM) and would ensure ATP saturation. Again, unlike some ATPases, Hsp90 is exquisitely speciﬁc to adenine and will not detectably bind GTP or GDP (L. H. Pearl and C. Prodromou, unpublished results). III. ATP BINDING AND HYDROLYSIS BY Hsp90 ARE ESSENTIAL IN VIVO Structural and biochemical studies clearly established the presence of an ATP-binding function in the N-terminal domain of Hsp90 (Grenert et al., 1997; Prodromou et al., 1997a). The availability of a speciﬁc competitor of ATP binding to Hsp90 then allowed reliable measurement of the Hsp90-speciﬁc ATPase activity of puriﬁed Hsp90 (Panaretou et al., 1998). This had previously been difﬁcult to quantify against the background of nonspeciﬁc ATPase activities copurifying with Hsp90 (Fig. 4A). The rate of ATP hydrolysis measured in isolated Hsp90 is slow

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

165

FIG. 4. ATPase activity of Hsp90. (a) Titration of ATPase activity in puriﬁed recombinant yeast Hsp90 by addition of geldanamycin. ATPase activity was measured using a continuous enzyme-coupled assay, as previously described (Panaretou et al., 1998), which avoids end-product inhibition by ADP. (b) Hsp90 ATPase activity is stimulated at higher temperatures to a greater degree than expected, suggesting some mechanism of activation by heat shock.

compared with those observed in some other ATP-dependent chaperones. Puriﬁed yeast Hsp90 displays a turnover of ∼0.4 mol min−1 mol−1 at 30◦ C, which increases at higher temperatures more than expected from simple thermodynamics (Fig. 4B), suggesting a process of heat activation in Hsp90. ATPase activities have also been reported for the E. coli HtpG protein (Panaretou et al., 1998) and for human/mouse TRAP1 (Felts et al., 2000), and are of the same order of magnitude. The low rate of inherent ATPase activity observed in vitro raised concerns over the relevance of this activity in vivo. The essential requirement of both ATP binding and hydrolysis to Hsp90 function in vivo was subsequently demonstrated by mutagenesis studies in yeast, where functional Hsp90 is required for viability. In these studies, Asp-79 and Glu-33, which were implicated by structural studies in nucleotide binding and hydrolysis, respectively, were mutated. ATP binding and ATPase activities of the mutant proteins were determined in vitro, and the functionality in vivo was determined by the ability of the mutants to confer viability in yeast in the absence of a wild-type Hsp90 allele (Obermann et al., 1998; Panaretou et al., 1998). Consistent with its role in providing the only direct contact between the protein and the adenine base, all mutations of Asp-79 abolished detectable nucleotide binding and caused loss of viability in vivo. Mutation of Glu-33 also caused loss of viability in vivo. Unlike the Asp-79 mutants, the E33A mutation did not signiﬁcantly diminish nucleotide binding to Hsp90, but effectively abolished its ATPase activity, indicating that both binding and hydrolysis of ATP are essential for the in vivo action of Hsp90 (Fig. 5).

166

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

FIG. 5. In vivo requirement for ATP binding and hydrolysis by Hsp90. (a) Sector plate showing growth on selective uracil/5-FOA medium of S. cerevisiae strain PP30 harboring a single Hsp90 allele, either wild-type (W+), with an extant mutation in the main adeninebinding residue Asp-79 (D79N, D79W, D79R), or explicitly reverted to wild-type by a second site-speciﬁc mutation (D79N → W+, D79W → W+, D79R → W+). Only alleles that retain Asp-79 (and retain ATP-binding in vitro) are viable in vivo (Panaretou et al., 1998). (b) As in (a) but for mutation of the catalytic residue Glu-33. Again, only wild-type or reverted wild-type with a glutamic acid are viable in vivo, and retain hydrolytic activity in vitro.

IV. CONFORMATIONAL CHANGES IN Hsp90 ACCOMPANYING THE ATPASE CYCLE The functional requirement to both bind and subsequently hydrolyze ATP suggests that Hsp90 undergoes an ATPase cycle in the same way as the Hsp70/DnaK and Hsp60/GroEL chaperone families. As with those systems, the key to understanding the mechanism of client-protein activation by Hsp90 is the deﬁnition of the structural states that accompany each stage in the Hsp90 ATPase cycle and how the transitions between these are coupled to changes in the conformation of the client protein. However, unlike Hsp70 and GroEL, relatively little is known about Hsp90, states. Hsp90 forms a stable complex with the co-chaperone p23/Sba1 (Fang et al., 1998; Johnson and Toft, 1994; 1995) in the presence of nonhydrolyzable ATP analogs, such as AMP-PNP. Such a stable complex can also be formed by incubation with ATP in the presence of molybdate ions in which a stable Hsp90–ADP–MoO2− 4 complex is progressively formed. Molybdate has long been included in Hsp90 buffers, prior to

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

167

establishment of ATP dependence, having been empirically found to stabilize Hsp90 complexes with steroid hormone receptors (Pratt, 1998; Prat and Dittmar, 1998). Biochemically, the state of Hsp90 induced by these ATP mimics is less hydrophobic than other states (Sullivan et al., 1997), but the signiﬁcance of this observation is unclear. This distinct “ATP-bound” state of Hsp90 is clearly different from the putative states of Hsp90 in the absence of nucleotide, in the presence of ADP, or in the presence of inhibitors such as geldanamycin and radicicol, none of which permit binding of p23/Sba1. Considerable insight into the nature of the ATP-bound state has come from crosslinking studies of Hsp90 mutants lacking the C-terminal region that mediates inherent dimerization (Prodromou et al., 2000). These mutants remain monomeric in the absence of nucleotide or in the presence of ADP, geldanamycin, or radicicol, but dimerize readily in the presence of AMP-PNP, a nonhydrolyzable ATP analog. AMP-PNPdependent dimerization is efﬁciently inhibited by geldanamycin and radicicol, conﬁrming that this effect is a function of binding to the characterized ATP-binding site in the N-terminal domain. By introducing single cysteine mutations into surface positions on the N-terminal domain of yeast Hsp90 (which lacks any endogenous cysteine residues), we were able to label these site-speciﬁcally with pyrene. Consequently, it was possible to show in the full-length Hsp90 that it is the N-terminal domains themselves that are brought into close contact as a result of nucleotide triphosphate binding. Thus, two extreme conformational states of the Hsp90 dimer can now be deﬁned: a “tense” state in which ATP binding constrains the N-terminal domains into dimeric association, and a “relaxed” state (which appears to be the default in the absence of nucleotide triphosphate) in which the N-terminal domains are free and dissociated (Fig. 6). The kinetic parameters of some of the steps in the ATPase cycle that interconverts the tense and relaxed states have recently been determined (Weikl et al., 2000). In this study the rate-limiting step was identiﬁed as the hydrolysis of ATP bound to Hsp90 that has already undergone a conformational change “trapping” the bound nucleotide. At ﬁrst sight, the identiﬁcation of ATP hydrolysis as the rate-limiting step would appear to contradict our observation that the rate-limiting step in ATP turnover is the association of the N-termini in the Hsp90 dimer once ATP is bound (Prodromou et al., 2000), as is also the case for DNA gyrase B (Ali et al., 1995). This apparent contradiction is resolved with the realization that there are in fact two closely coupled conformational changes required along the productive path to ATP hydrolysis. The ﬁrst of these involves the closure of a “lid” segment in the N-terminal domain

168

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

FIG. 6. ATPase-coupled N-terminal dimerization in Hsp90. Binding of ATP to an Hsp90 dimer promotes association of the N-terminal nucleotide-binding domains into a “tense” state, which in turn promotes ATP hydrolysis. The ADP thus generated does not favor N-terminal association and allows relaxation into the “relaxed” state, completing the cycle (From Prodromou et al., 2000, with permission).

of Hsp90 (residues 100–121) over the bound nucleotide, in much the same way as occurs in DNA gyrase B and MutL (Ali et al., 1995; Ban et al., 1999). This conformational change would greatly reduce the efﬁciency of nucleotide exchange, and most probably constitutes the observed “committed” complex (Weikl et al., 2000). Lid closure as a result of ATP binding, which is yet to be directly observed in an Hsp90 structure, exposes a hydrophobic patch on the underside of the lid segment that is believed to mediate the association of the N-terminal domains within the Hsp90 dimer (Fig. 7, see color insert). Association of the Ntermini, which constitutes the second conformational step, is directly coupled to the ATPase reaction. Thus, Hsp90 mutants lacking the constitutive C-terminal dimerization interface, and for which N-terminal association is therefore an intermolecular reaction, hydrolyze ATP at only one-ﬁfth the rate of full-length dimeric Hsp90 for which N-terminal association is intramolecular (Prodromou et al., 2000). Furthermore, mutations in the N-terminus of Hsp90 that enhance N-terminal association increase ATP turnover, while mutations that diminish N-terminal association decrease ATP turnover. Thus the step following nucleotide commitment and identiﬁed as rate-limiting (Weikl et al., 2000) is not a single process but comprises the substeps of N-terminal association, which is truly rate-limiting, followed by ATP hydrolysis. Observation of Hsp90 dimers using rotary-shadowing electron microscopy has shown an essentially elongated structure in which the C-termini formed a dimeric interaction while the N-termini were dissociated, as in the relaxed state described above. Subsequently, formation of structures in which both N- and C-termini appeared to be in proximity was observed following severe heat shock or, less effectively, in the

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

169

presence of ATP (Maruya et al., 1999). Using AMP-PNP, we have now been able to obtain clear images of this ATP-bound tense state of yeast Hsp90 in negative-stain electron microscopy. (Fig. 8). These images show a closed structure of ∼150–160 A˚ external diameter consisting of four globular segments encompassing a stain-ﬁlled lumen of ∼30–40 A˚ diameter. Single-particle reconstruction based on these images is in progress. Dimeric association of the N-terminal domains on binding of nucleotide triphosphates, coupled with the inherent dimerization at the C-terminus, suggests that Hsp90 operates a molecular “clamp” whose opening and closing is coupled to the binding and hydrolysis of ATP (Prodromou et al., 2000). The formation of the tense N-terminally dimerized state observed in the electron micrographs requires ATP binding and is thus blocked by Hsp90 inhibitors such as geldanamycin and radicicol. Although the literature is not totally clear, in the most detailed studies binding of client protein appears not to be prevented by geldanamycin, but activation of the bound client protein is blocked (Schneider et al., 1996). This suggests strongly that it is in association with the ATP-induced tense state of Hsp90 that client proteins undergo the change of their state that allows the particular modiﬁcation or association required for their particular activation process. The nature and location of the client protein interaction with Hsp90 are obscure. Deletion studies suggest that the C-terminal region, implicated in dimerization and co-chaperone recruitment, was also involved in client-protein interaction (Shaknovich et al., 1992; Sullivan and Toft,

FIG. 8. Electron micrograph of “closed” Hsp90 dimer. Class-averaged image of one view of Hsp90 complexed with the nonhydrolyzable ATP analog AMP-PNP, in negative stain electron microscopy. Hsp90 adopts a closed toroidal conformation of ∼150 A˚ external diameter with a stain-ﬁlled central cavity. The image was produced by Dr. Ulrich Gohlke, Birkbeck College, London.

170

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

1993). As these studies primarily measured “loss of function” in clientprotein binding, it is impossible to distinguish whether the effects resulted from the absence of Hsp90 regions directly involved in molecular interaction with clients or a requirement for dimerization in client-protein binding and/or activation. In vitro studies of binding of peptides and nonclient denatured proteins to isolated Hsp90 fragments have complicated the issue still further. Consequently, binding sites apparently have been identiﬁed with different properties in the N- and C-terminal domains (Scheibel et al., 1998; Young et al., 1997), implicating the charged linker region, which is dispensable for conferring viability in vivo (Scheibel et al., 1999) and for ATPase activity in vitro (L. H. Pearl and C. Prodromou, unpublished data). It is tempting to suggest that the closed structure formed by dimerization of the N-terminal domains may indeed function as a holding “clamp,” the client proteins (or a part of them) being somehow trapped between the pair of jaws in the ATP-bound state, then released following hydrolysis. Something very similar to this occurs in DNA gyrase B, whose nucleotide-binding domain is structurally homologous to Hsp90 and which also possesses a strong C-terminal dimerization domain (Wigley et al., 1991). Here, the regions between the N- and the C-terminal domain provide positively charged DNA-binding surfaces which are believed to close around a DNA duplex on ATP binding and release it following hydrolysis (Kampranis et al., 1999). However, the potential analogy with Hsp90 is far from perfect, as the bound DNA in the gyrase B system undergoes no change of conformation during this cycle, whereas the Hsp90-bound client protein almost certainly does. Nonetheless, a role for the jaws of the clamp in binding the client protein is an attractive possibility. The clamp structure of the ATP-bound state would certainly be consistent with the observation that it is usually only one domain from a multidomain protein that requires interaction with Hsp90 (Hartson et al., 1998; Stancato et al., 1996). Thus the dependent domain in a client protein could be bound between the jaws of the chaperone, while other domains could remain free, protruding from either face of the complex. However, in the absence of direct evidence, such models remain speculative. V. Hsp90 ATPASE INHIBITORS—A NEW CLASS OF ANTITUMOR DRUGS Geldanamycin and the related herbimycin A are fungal antibiotics whose antitumor activity was originally attributed to direct inhibition of protein kinases such as v-Src, Raf-1, or ErbB2 (Fukazawa et al., 1991; 1994; Schnur et al., 1995a; 1995b; Taniguchi et al., 1993). Subsequently, it was shown that geldanamycin did not exert its main effect by directly

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

171

inhibiting these kinases, but instead reduced their active cellular levels and promoted their degradation (Miller et al., 1994b; Mimnaugh et al., 1996; Schneider et al., 1996; Schulte et al., 1995). The direct molecular targets of geldanamycin were then shown to be Hsp90 (and GRP94), and the reduction in active levels of the target kinases was found to result from disruption of their productive interaction with Hsp90 which is essential for their activation (Chavany et al., 1996; Miller et al., 1994a; Whitesell et al., 1994). Geldanamycin has since been shown to bind to the N-terminal domain of Hsp90 (Grenert et al., 1997; Roe et al., 1999; Stebbins et al., 1997) and to be a potent competitive inhibitor of ATP binding to Hsp90 both in vitro (K d ≈ 1.2 μM) (Roe et al., 1999) and in vivo. The observation that an inhibitor of ATP binding can block Hsp90-dependent activation of client proteins in vivo is strong supportive evidence for the essential ATP dependence of Hsp90 function. The elucidation of the mechanism of action of geldanamycin and the identiﬁcation of a number of key anticancer target proteins as Hsp90 clients have awakened a considerable interest in Hsp90 as an antitumor drug target. Target-validation and clinical trials of geldanamycin derivatives are underway in the UK and US (e.g., Clarke et al., 2000). Detailed molecular characterization of the interactions of geldanamycin with the unusual ATP-binding site in the N-domain of Hsp90 (Roe et al., 1999; Stebbins et al., 1997) has allowed the rationalization of structure/activity relationships for a large series of geldanamycin derivatives (Roe et al., 1999). These were originally developed in ignorance of the true target molecule (Schnur et al., 1995a; 1995b), but provide important guidance for future development of this class of inhibitors. Radicicol (or monorden), a fungal antibiotic unrelated to geldanamycin, was identiﬁed as an apparent inhibitor of Raf-1 kinase activity (Soga et al., 1998), but was later found to bind to Hsp90 (Roe et al., 1999; Soga et al., 1999). Initial studies suggested that radicicol bound to a site on Hsp90 different from that of geldanamycin (Sharma et al., 1998). However, crystallographic studies showed that radicicol binds in the nucleotide-binding pocket in the N-terminal domain of Hsp90, and is a competitive inhibitor of ATP binding with nanomolar (19 nM) afﬁnity (Roe et al., 1999) (Fig. 9; see color insert for Fig. 9c). The ability of two quite distinct classes of natural products to bind to the ATP-binding site in Hsp90 with high speciﬁcity and afﬁnity bodes well for the future identiﬁcation and development of novel Hsp90 inhibitors in high-throughput screens. Several other molecules have recently been suggested as Hsp90 inhibitors, although their sites and speciﬁcity of action (if any) are far less well characterized than geldanamycin and radicicol. Coumarins, such as novobiocin, are known as inhibitors of bacterial type II DNA topoisomerases such as DNA gyrase, binding to the ATP-binding site in

172

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

FIG. 9. Hsp90 inhibitors. (a) Chemical structure of the classic Hsp90 inhibitor geldanamycin. A modiﬁed version in which the 17 position is replaced by an allylamino group has lower toxicity and is in current clinical trials. (b) Chemical structure of the unrelated Hsp90 inhibitor radicicol (monorden).

the N-domain of the Gyr B protein (Ali et al., 1993). Novobiocin has recently been found to inhibit Hsp90 (Marcu et al., 2000). However, the apparent site of action on Hsp90 appears not to be the nucleotidebinding N-domain which is structurally homologous to that of Gyr B. Instead, novobiocin binding occurs within the C-terminal region at an ATP-sensitive site that is only accessible in recombinant Hsp90 constructs lacking the well characterized N-domain. The signiﬁcance of this observation is unclear.

FIG. 2. N-Terminal domain of yeast Hsp90. (a) Secondary structure cartoon of the crystal structure of the yeast Hsp90 N-terminal domain color ramped from the protein N terminus (blue) to the C terminus of the domain at residue 215 (red). The strands of the twisted β-sheet are labeled as in Prodromou et al. (1997b). (b) As in (a) but rotated 90◦ around the vertical: (c) as in (a) but rotated 180◦ around the vertical. FIG. 3. Adenine-nucleotide binding to the N-terminal domain. (a) Secondary structure cartoon (as in Fig. 1c), with ADP (stick model) bound in the pocket formed by the helical face of the domain. (b) Detail of the ATP/ADP-binding site. The bound nucleotide is shown as a CPK-colored ball-and-stick model, and the many water molecules bound in the site are shown as red spheres. Hydrogen bonds are indicated by broken yellow rods, and the ligand interactions of the magnesium ion by broken blue rods. FIG. 7. Lid movements accompanying ATP binding. (a) Conformation of the lid segment (residues 100–118 in yeast Hsp90) in the ADP-bound form observed in crystal structures (Prodromou et al., 1997a). Residue Thr-101 is buried in this open-lid conformation while Ala-107 is fully exposed. (b) Model of the closed-lid conformation believed to be favored by ATP binding. In this conformation, Thr-101 becomes fully exposed, while Ala-107 becomes buried against a hydrophilic patch formed by Asn-40 and Asp-43. Mutation of Thr-101 to the more hydrophobic isoleucine disfavors lid closure and results in diminished N-terminal association and ATPase activity. Mutation of Ala-107 to the more hydrophilic asparagine favors lid closure and enhances N-terminal association and ATPase activity (Prodromou et al., 1997a).

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

173

VI. INTERACTION WITH CO-CHAPERONES At least ten proteins have now been identiﬁed by genetic and/or biochemical means that interact with yeast Hsp90 and are not obviously client proteins. That is to say, they do not depend on transient association with Hsp90 for modiﬁcation or assembly into a complex devoid of Hsp90 that represents their “activated” form. Although these Hsp90associated proteins are often referred to as Hsp90 “co-chaperones,” the roles they play in Hsp90-dependent client-protein activation are far from clear. Indeed, in vivo studies in yeast suggest that the genes for many of these putative co-chaperones can often be deleted without loss of viability (Bohen, 1998; Dolinski et al., 1997; Duina et al., 1996a; 1996b; Fang et al., 1998; Knoblauch and Garabedian, 1999). A subset of these Hsp90-associating proteins (Hsp70/Ssa1, Hsp40/Ydj1, Hip) includes the components of other chaperone systems that cooperate with Hsp90 in the activation of at least some client proteins (reviewed in Pratt and Dittmar, 1998; Smith, 2000). One further class of Hsp90-associating proteins consists of an Hsp90-binding domain formed by tetratricopeptide repeats (TPR) (see below) coupled to a peptidylprolyl isomerase domain (Cyp40/Cpr6/7 FKBP52—generally classed as immunophilins) or to a phosphoprotein phosphatase domain (PP5/Ppt1). The binding site on Hsp90 for associated proteins with TPR-domains is localized to a peptide sequence MEEVD which occurs at the extreme C-terminus of cytoplasmic Hsp90s (Chen et al., 1998), but is absent from the bacterial HptG FIG. 9C. Hsp90 inhibitors. (c) Details from crystal structures of yeast Hsp90 N-terminal domain complexed with ADP (left), geldanamycin (middle), and radicicol (right). A detailed comparison of the binding of these compounds can be found in (Roe et al., 1999). FIG. 10. Structural basis for TPR-domain binding. Structure of the middle TPR domain of Hop bound to a peptide corresponding to the C terminus of Hsp90 (Scheuﬂer et al., 2000). The peptide binds in the groove formed by the superhelical arrangement of αhelices in the TPR domain, making speciﬁc interactions involving the side chains of the MEEVD sequence and the α-carboxyl of the terminal aspartic acid. FIG. 11. Hsp90-interacting region in PKB/Akt. Secondary structure cartoon of the catalytic domain of cAMP-dependent protein kinase (protein kinase A). The region identiﬁed in the homologous protein kinase B as binding directly to Hsp90 (Sato et al., 2000) is highlighted in yellow, with the most strongly interacting region in red. No crystal structure is yet available for PKB/Akt, but the strong homology to PKA allows a reliable mapping onto the known PKA structure (Knighton et al., 1991). These interacting regions are particularly signiﬁcant in the biochemistry of protein kinases regulated by Ser/Thr phosphorylation, as it includes the activation segment ( Johnson et al., 1996). Phosphorylation of a threonine in this segment (pThr) generates an interaction with a basic residue on the C-helix (blue) which promotes a shift in the position of the N-terminal domain bringing the catalytic residues into the correct alignment for phosphoryl transfer.

174

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

homologs and from the eukaryotic TRAP1 and ER-speciﬁc GRP94 homologs. The structural basis for the speciﬁc interaction between this C-terminal tail sequence and TPR domains has recently been deﬁned (Scheuﬂer et al., 2000) (Fig. 10; see color insert). Two Hsp90-associating proteins Hop/Sti1 and p50cdc37 appear to act as scaffold or “adaptor” proteins. Hop/Sti1 consists of at least three TPR domains and is able to bind Hsp70/Ssa1 via its N-terminal TPR domain and Hsp90 via its C-terminal TPR domain (Chen et al., 1996b), and p50cdc37 binds Hsp90 via its C-terminal half (Silverstein et al., 1998) and a wide range of protein kinase client proteins via its N-terminus (Grammatikakis et al., 1999; Silverstein et al., 1998; Stepanova et al., 1996). Until recently, it had generally been assumed that the Hsp90 dimer presented a single binding site for TPR-domain co-chaperones (OwensGrillo et al., 1996a; 1995). This model was based on the observation that binding of Hop/Sti1 and immunophilins such as Cyp40 to Hsp90 was mutually exclusive (Hoffmann and Handschumacher, 1995; Ratajczak and Carrello, 1996). Hop/Sti1 and p50cdc37 were also found to be mutually exclusive, suggesting that, while it did not possess any obvious TPR domain, p50cdc37 bound at or very close to this site (Silverstein et al., 1998). Thus the conventional model for Hsp90 co-chaperone complexes shows an Hsp90 dimer binding a single molecule of one of Hop/Sti1, an immunophilin, or p50cdc37. However, quantitative studies using puriﬁed proteins in vitro have revealed a more complex situation. Titration of Hsp90 with the yeast immunophilin Cpr6 or with Sti1 followed by isothermal calorimetry (Prodromou et al., 1999) showed a clear binding stoichiometry of one Hsp90 molecule per Cpr6 or Sti1 rather than the ratio of two Hsp90s per co-chaperone expected from the mutual exclusion data. This apparent contradiction was resolved by the observation that Sti1 itself is a stable dimer under most conditions and binds as such to the Hsp90 dimer, simultaneously occupying both available TPR-binding sites and excluding Cpr6 from binding. Conversely, Cpr6 or an isolated TPR domain, although monomeric, will prevent binding of the dimeric and bidentate Sti1 even if they occupy only one of the available TPR-binding sites. The presence of two independent TPR-domain binding sites in the Hsp90 dimer raises the possibility of a complex mixture of Hsp90based complexes potentially containing homo- and heterocombinations of TPR-domain co-chaperones. Such a situation has been observed in the yeast system with the binding of Cpr7, a Hsp90-binding TPR-domain immunophilin and Cns1, an essential multiple TPR-domain protein whose overexpression suppresses ts cpr7 and hsp82 − alleles (Dolinski et al., 1998; Marsh et al., 1998; Nathan et al., 1999). Immunoprecipitation of Cpr7 brings down Hsp90 and Cns1 (Marsh et al., 1998), suggesting that Cns1

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

175

and Cpr7, which can bind Hsp90 independently, are simultaneously present in a Cns1–Hsp902–Cpr7 complex. Another immunophilin-like protein, XAP2, has recently been shown to be part of the dioxin receptor–Hsp90 heterocomplex (Carver and Bradﬁeld, 1997; Ma and Whitlock, 1997; Meyer et al., 1998). This cochaperone has been shown to protect the ligand-free receptor against ubiquitination and following ligand activation was shown to regulate its subcellular localization by cytoplasmic retention (Kazlauskas et al., 2000). These functions were dependent on the stable formation of a Hsp90/p23/XAP2/dioxin receptor complex. This observation may explain why cells treated with geldanamycin, which would block p23 binding, show an increased rate of degradation of Hsp90 client proteins by proteolysis (Miller et al., 1994b; Mimnaugh et al., 1996; Schneider et al., 1996; Schulte et al., 1995). The p23/Sba1/Wos2 proteins are small acidic proteins with long highly charged C-terminal tails of repetitious sequence. P23/Sba1 associates with Hsp90 only in the presence of ATP (or ADP–MO2− 4 or AMP-PNP) (Fang et al., 1998; Johnson and Toft, 1994; 1995), and binding is thus dependent on an intact N-terminal domain. However, p23/Sba1 does not simply bind to the ATP-loaded N-terminal domain, but rather to the N-terminally associated dimeric structure that results from ATP binding (Prodromou et al., 2000). The role of p23/Sba1 has been far from clear and, as with several other co-chaperones such as some cyclophilins, it can be deleted in yeast with only mild impairment of viability (Bohen, 1998). In in vitro reconstructions of Hsp90-dependent steroid–hormone receptor activation (Dittmar and Pratt, 1997; Grenert et al., 1999; Kosano et al., 1998), p23 was found not to be essential, but its presence correlated with increased “yield” of activated receptor. The effect of p23 on client activation has recently been attributed to a role for p23 in enhancing the efﬁciency of release of bound client protein following ATP hydrolysis (Young and Hartl, 2000). VII. REGULATION OF ATP BINDING AND HYDROLYSIS IN THE CLIENT-PROTEIN ACTIVATION PATHWAY Immunoprecipitation studies of steroid–hormone receptor association with Hsp90 in reticulocyte lysates (Smith et al., 1995) have revealed an apparently ordered series of chaperone complexes. Early complexes contain Hsp70 and its co-chaperones, but lack Hsp90. Intermediate complexes retain Hsp70 and co-chaperones as well as Hsp90, which is coupled to Hsp70 via a common interaction with the p60/Hop scaffold protein (Chen et al., 1996b). Subsequently, late complexes can be

176

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

isolated in which p60/Hop and the Hsp70 subcomplex have gone, but in which other co-chaperones such as the TPR-domain immunophilins and p23 are present. Studies with Sti1 and Cpr6, the yeast homologs of p60/Hop and the TPR-immunophilin Cyp40, respectively, suggested that the ability of Hsp90 to bind ATP and go through its conformational cycle is regulated by its association with co-chaperones (Prodromou et al., 1999). Thus, p60/Hop/Sti1 inhibits ATP binding and hydrolysis by Hsp90 in the intermediate complexes in which Hsp70-bound client proteins are recruited to the Hsp90-based complex. Subsequent replacement of p60/Hop/Sti1 by immunophilins such as Cpr6 in the late complexes allows Hsp90 to bind ATP and progress through the nucleotide-coupled conformational cycle that is concomitant with client protein activation (Prodromou et al., 2000). What controls the rate of transition between the intermediate and late complexes remains to be determined. The function of other Hsp90 co-chaperones such as the immunophilins, Cns1, or the recently identiﬁed Hch1 (Nathan et al., 1999) is still unclear, but it would not be surprising to ﬁnd that these also play some role in modulating the ATPase-coupled chaperone cycle. Apart from regulation by co-chaperones, Hsp90 from higher eukaryotes is known to be a phosphoprotein, and its activity, as well as that of its co-chaperones, may also be modulated by kinases and phosphatases (Mimnaugh et al., 1995; Morano and Thiele, 1999). VIII. INTERACTIONS WITH ALTERATIONS OF CLIENT PROTEINS BY Hsp90 Compared with the Hsp70 and Hsp60/GroEL classes of molecular chaperones, remarkably little is known about the nature of the interaction between Hsp90 and its client proteins, and even less about the changes in client protein state that interaction engenders. On the basis of in vitro studies of Hsp90 in which it decreases the amount of aggregation of some denatured proteins (Wiech et al., 1992), it had been widely assumed that Hsp90, like GroEL, interacts with client proteins in a partly unfolded or “molten globule” state. However, there is little evidence that Hsp90 interacts with unfolded proteins in vivo or participates in de novo folding or refolding following denaturation under heat shock (Nathan et al., 1997). Furthermore, this in vitro activity is completely nucleotideindependent, and therefore difﬁcult to reconcile with the in vivo function of Hsp90, which is clearly nucleotide-dependent (Obermann et al., 1998; Panaretou et al., 1998). A general role in binding nonspeciﬁcally to improperly folded proteins implicated by the in vitro studies is also difﬁcult to reconcile with the

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

177

speciﬁc association of particular client proteins with Hsp90 that has been widely observed. The basis for this widespread but nonetheless speciﬁc client protein range remains one of the great conundrums in Hsp90 biochemistry. Hsp90 client proteins can be roughly partitioned into three classes: (1) protein kinases, (2) transcription factors, and (3) others. In yeast, the simplest organism in which an elaborated and essential Hsp90–co-chaperone system has been characterized, at least eight protein kinases have been identiﬁed (Wee1, Sch9, Gcn2, Cdc2, Cdc28, Cak1, Mps1, and Ste11) which have been shown experimentally to be obligate clients of the Hsp90 and/or Cdc37 system (AbbasTerki et al., 2000; Aligue et al., 1994; Donze and Picard, 1999; Farrell and Morgan, 2000; Louvain et al., 1998; Morano and Thiele, 1999; Munoz and Jimenez, 1999; Schutz et al., 1997). In contrast, clear homologs of steroid hormone receptors, which are the best studied class of Hsp90 client proteins in mammalian systems, cannot be identiﬁed in yeast. Consequently, in yeast at least, the Hsp90 system can reasonably be considered as a specialized chaperone for protein kinases, although Hsp90-dependent proteins such as mammalian steroid hormone receptors can utilize the yeast system with success (Nathan and Lindquist, 1995). Paradoxically, it is the steroid hormone receptors, which appear to be a much later evolutionary addition to the protein clientele of Hsp90, that have been most studied in terms of their Hsp90-dependent activation. Some concern must therefore be expressed that our current understanding of client protein activation by the Hsp90 system has been unduly biased by that emphasis. As a class, the Hsp90 protein kinases present one of the best opportunities for identifying a uniﬁed mechanism for Hsp90 client protein activation. Studies of several Hsp90-dependent kinases have shown that it is the common catalytic domain whose activation depends on Hsp90, rather than any of the attendant regulatory domains that are found in many kinases (Bijlmakers and Marsh, 2000; Hartson et al., 1998), but the nature of the interaction has been obscure. Recently, protein kinase B/Akt, which was implicated as a potential Hsp90 client by its marked downregulation in tumor cell culture treated with geldanamycin derivatives (Clarke et al., 2000), was shown to interact directly with Hsp90 (Sato et al., 2000). More detailed mapping of the interaction sites identiﬁed a segment in PKB/Akt (229–309) that was sufﬁcient for interaction with Hsp90. Although no crystal structure of PKB/Akt is yet available, homology with related protein kinases of known structure identiﬁes this interacting segment as one of the most important. The segment identiﬁed in protein kinases which comprises the ﬁrst part of the C-terminal α-helical domain, incorporates the catalytic loop and the early part of the activation segment. In many protein kinases, catalytic activity and

178

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

substrate speciﬁcity are exquisitely coupled to conformational changes within this segment and to its relative orientation to the N-terminal β-sheet domain which provides the ATP-binding site and an essential catalytic residue (Johnson et al., 1996). The high degree of conservation of parts of this segment may explain the ability of a wide range of protein kinases to interact with Hsp90, but it is yet to be demonstrated that the same regions in Hsp90 and other protein kinases mediate this interaction (Fig. 11, see color insert). It seems highly likely that the role of Hsp90 in kinase activation will be to facilitate the conformational switch between the inactive and active conformations of this segment, and/or to maintain a conformationally unstable inactive conformation of this segment in an activation-competent state. Many protein kinases that depend on Hsp90 for activation also interact with p50cdc37, which is able to bind to Hsp90 independently of kinases and appears therefore to function as a scaffold protein (Grammatikakis et al., 1999; Silverstein et al., 1998; Stepanova et al., 1996). It is by no means clear whether all protein kinases that interact with Hsp90 do so via p50cdc37 (this has not yet been shown for PKB/Akt), nor is it clear that all protein kinases that interact with p50cdc37 are consequently dependent on Hsp90 for their activation (Hunter and Poon, 1997). The ability of p50cdc37 to interact with a range of different protein kinases suggests that it recognizes generic features of protein kinases, but the location of these is yet to be determined. Hsp90 is directly involved in the activation of transcription factors including steroid hormone receptors and a variety of basic helix–loop– helix receptors, the most studied of which are the steroid hormone receptors. Within this class, known Hsp90 clients include receptors for glucocorticoid (GR) (Sanchez et al., 1985), progesterone (PR) (Catelli et al., 1985; Schuh et al., 1985), estrogen (ER) (Joab et al., 1984; Redeuilh et al., 1987), androgen (Joab et al., 1984; Veldscholte et al., 1992), mineralcorticoid (Joab et al., 1984; Rafestin-Oblin et al., 1989; Veldscholte et al., 1992), and v-erbA (related to the thyroid hormone receptor (Privalsky, 1991). These unliganded receptors exist in Hsp90 heterocomplexes which are essential for the ligand-binding domain to obtain hormone binding activity (Bresnick et al., 1989; Hutchison et al., 1992). Of the nonsteroid Hsp90-dependent transcription factors, the best studied is the dioxin or aryl hydrocarbon receptor (DR or Ah receptor), which mediates induction of aryl hydrocarbon hydroxylase, a cytochrome P450 involved in xenobiotic metabolism (Poellinger et al., 1992). DR differs from steroid receptors in that its DNA-binding domain consists of a basic helix– loop–helix (bHLH) domain rather than the double zinc-ﬁnger structure of the steroid hormone receptors, and on ligand activation forms as

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

179

heterodimer with Arnt (aryl hydrocarbon receptor nuclear translocator) protein in the nucleus (Hankinson, 1994; Poellinger et al., 1992) rather than the homodimer formed by steroid receptors. As with steroid hormone receptors, however, the transcriptional activity of DR is regulated in an Hsp90 ligand-dependent manner (Carver et al., 1994; Hankinson, 1994; Poellinger et al., 1992; Whitelaw et al., 1993). Other bHLH transcription factors that are Hsp90-dependent include Sim (McGuire et al., 1995); MOP1, 3, and 4 (Hogenesch et al., 1997); and MyoD and E12 (Shaknovich et al., 1992; Shue and Kohtz, 1994). Studies of progesterone and glucocorticoid receptors identiﬁed a series of other proteins as components of Hsp90-based receptor heterocomplexes. These include immunophilins such as the FK506-binding proteins FKBP51 and FKBP52 (Smith et al., 1993a; 1993b; 1990; Tai et al., 1986; 1992), the cyclosporin-binding immunophilin Cyp40 (Ratajczak et al., 1990), Hsp70 (Kost et al., 1989), p48/Hip (Prapapanich et al., 1996; Smith and Toft, 1993; Smith et al., 1995), p60/Hop (Smith et al., 1992; 1993c), and p23 ( Johnson et al., 1994). Observations of the time course of heterocomplex formation in reticulocyte lysates (Smith, 1993) have revealed that some of the components were present at different stages, indicating a dynamic protein complex. In the activation of progesterone receptor, PR is initially bound to Hsp70 and is subsequently incorporated into a heterocomplex with Hsp90, Hsp70, Hip, and Hop, which represents the intermediate stage. This complex “matures” with the loss of Hsp70, Hip, and Hop, and binding of immunophilins, ATP and p23. The cell-free assembly of PR into mature complex at 30◦ C takes ∼3 min and PR dissociates from the mature complex after ∼5 min (Smith et al., 1995). It is only in this mature complex that the bound receptor is competent to bind steroid hormone and take on the transcriptionally active conformation. The lifetime of the ATP-bound mature complex in reticulocyte lysates is comparable to the inherent rate of ATP hydrolysis (Prodromou et al., 1999) of yeast Hsp90 (∼0.2 min−1 at 30◦ C), raising the possibility that the ATPase activity of Hsp90 acts as a clock determining the length of time that an apo-receptor will be sensitive to a hormonal signal. Binding of hormone steroid receptors requires a conformational change in the structure of the apo-receptor to allow access to a hydrophobic pocket in the ligand-binding domain (Giannoukos et al., 1999; Hutchison et al., 1992; Stancato et al., 1996; Xu et al., 1998). It is most likely this open conformation of the ligand-binding pocket that is maintained within the ATP-bound “mature” Hsp90 heterocomplex. Although co-chaperones p60/Hop, p48/Hip, and Hsp40 have been implicated in the efﬁcient assembly of receptor heterocomplexes, only the

180

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

bona ﬁde ATP-dependent chaperones Hsp90 and Hsp70 are absolutely required to open the hydrophobic pocket and allow steroid binding (Morishima et al., 2000). We have recently shown that the Hsp90 dimer itself undergoes a conformational transition on binding ATP, from a “relaxed” state in which the N-terminal domains are free, to a “tense” state in which the N-terminal domains are closely associated (Prodromou et al., 2000). The ATP-bound “tense” state of Hsp90 is thus associated with the receptive conformation of steroid hormone receptors in the heterocomplexes. It is tempting to suggest that the dimeric ATP-dependent molecular “clamp” provided by Hsp90 could interact directly with two distinct regions on the steroid receptor ligand-binding domain, holding them apart and thereby permitting access of the hormone to the hydrophobic pocket. Eventually, the Hsp90-bound ATP will be hydrolyzed, causing Hsp90 to relax and release the activated (or unactivated) steroid receptor. However, Hsp90 alone does not appear able to induce the the hormone-receptive conformation in the receptor, but maintains this state for several minutes once it is established by Hsp70. This difference in behavior between Hsp70 and Hsp90 may be reﬂected in their very different inherent ATPase activities, so that the low-geared rapid cycle of Hsp70 allows it to exert sufﬁcient force to change the client protein conformation for a short period, whereas Hsp90 with a highgeared slow cycle is able to hold the metastable conformation of the client once passed on by Hsp70.

IX. CONCLUSION Our understanding of the biochemistry of Hsp90 has undergone an enormous change in the last ﬁve years, and a consensus view of many aspects of its structure and mechanism is starting to emerge. However, many of the key phenomena associated with this system remain obscure. Most notably, the roles of the expanding set of co-chaperones and the nature of the interaction between Hsp90 and the client proteins are still very poorly understood

ACKNOWLEDGMENTS We are very grateful to Peter Piper, Barry Panaretou, Mark Roe, John Ladbury, Giuliano Siligardi, Ronan O’Brien, Helen Saibil, Ulrich Gohlke, Shradha Singh, Chris Richardson, and other colleagues and collaborators, who have made our work in this ﬁeld such an exciting and stimulating process. We gratefully acknowledge the ﬁnancial support of The Wellcome Trust.

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

181

REFERENCES AbbasTerki, T., Donze, O., and Picard, D. (2000). FEBS Lett. 467, 111–116. Ali, J. A., Jackson, A. P., Howells, A. J., and Maxwell, A. (1993). Biochemistry 32, 2717–2724. Ali, J. A., Orphanides, G., and Maxwell, A. (1995). Biochemistry 34, 9801–9808. Aligue, R., Akhavannik, A., and Russell, P. A. (1994). EMBO J. 13, 6099–6106. Ban, C., Junop, M., and Yang, W. (1999). Cell 97, 85–97. Bergerat, A., de Massay, B., Gadelle, D., Varoutas, P. C., Nicolas, A., and Forterre, P. (1997). Nature 386, 414–417. Bijlmakers, M. J., and Marsh, M. (2000). Mol. Biol. Cell 11, 1585–1595. Bilwes, A. M., Alex, L. A., Crane, B. R., and Simon, M. I. (1999). Cell 96, 131–141. Bohen, S. P. (1998). Mol. Cell. Biol. 18, 3330–3339. Borkovich, K. A., Farrelly, F. W., Finkelstein, D. B., Taulien, J., and Lindquist, S. (1989). Mol. Cell. Biol. 9, 3919–3930. Bresnick, E. H., Dalman, F. C., Sanchez, E. R., and Pratt, W. B. (1989). J. Biol. Chem. 264, 4992–4997. Buchner, J. (1999). Trends Biochem. Sci. 24, 136–141. Carrello, A., Ingley, E., Minchin, R. F., Tsai, S., and Ratajczak, T. (1999). J. Biol. Chem. 274, 2682–2689. Carver, L. A., and Bradﬁeld, C. A. (1997). J. Biol. Chem. 272, 11452–11456. Carver, L. A., Jackiw, V., and Bradﬁeld, C. A. (1994). J. Biol. Chem. 269, 30109–30112. Catelli, M. G., Binart, N., Jung-Testas, I., Renoir, J. M., Baulieu, E. E., Feramisco, J. R., and Welch, W. J. (1985). EMBO J. 4, 3131–3135. Chavany, C., Mimnaugh, E., Miller, P., Bitton, R., Nguyen, P., Trepel, J., Whitesell, L., Schnur, R., Moyer, J., and Neckers, L. (1996). J. Biol. Chem. 271, 4974–4977. Chen, C. F., Chen, Y. M., Dai, K., Chen, P. L., Riley, D. J., and Lee, W. H. (1996a). Mol. Cell. Biol. 16, 4691–4699. Chen, S. Y., Prapapanich, V., Rimerman, R. A., Honore, B., and Smith, D. F. (1996b). Mol. Endocrinol. 10, 682–693. Chen, S. Y., Sullivan, W. P., Toft, D. O., and Smith, D. F. (1998). Cell Stress Chapter. 3, 118–129. Clarke, P. A., Hostein, I., Banerji, U., Stefano, F. D., Maloney, A., Walton, M., Judson, I., and Workman, P. (2000). Oncogene 19, 4125–4133. Csermely, P., Schnaider, T., Soti, C., Prohaszka, Z., and Nardai, G. (1998). Pharmacol. Ther. 79, 129–168. Cutforth, T., and Rubin, G. (1994). Cell 77, 1027–1036. Dai, K., Kobayashi, R., and Beach, D. (1996). J. Biol. Chem. 271, 22030–22034. Das, A. K., Cohen, P. T. W., and Barford, D. (1998). EMBO J. 17, 1192–1199. Dittmar, K., and Pratt, W. B. (1997). J. Biol. Chem. 272, 13047–13054. Dolinski, K., Muir, S., Cardenas, M., and Heitman, J. (1997). Proc. Natl. Acad. Sci. USA 94, 13093–13098. Dolinski, K. J., Cardenas, M. E., and Heitman, J. (1998). Mol. Cell. Biol. 18, 7344–7352. Donze, O., and Picard, D. (1999). Mol. Cell. Biol. 19, 8422–8432. Duina, A. A., Chang, H. C., Marsh, J. A., Lindquist, S., and Gaber, R. F. (1996a). Science 274, 1713–1715. Duina, A. A., Marsh, J. A., and Gaber, R. F. (1996b). Yeast 12, 943–952. Engert, J. C., Berube, P., Mercier, J., Dore, C., Lepage, P., Ge, B., Bouchard, J. P., Mathieu, J., Melancon, S. B., Schalling, M., Lander, E. S., Morgan, K., Hudson, T. J., and Richter, A. (2000). Nat. Genet. 24, 120–125. Fang, Y., Fliss, A. E., Rao, J., and Caplan, A. J. (1998). Mol. Cell. Biol. 18, 3727–3734. Farrell, A., and Morgan, D. O. (2000). Mol. Cell. Biol. 20, 749–754.

182

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

Felts, S. J., Owen, B. A., Nguyen, P., Trepel, J., Donner, D. B., and Toft, D. O. (2000). J. Biol. Chem. 275, 3305–3312. Ferreira, L. R., Norris, K., Smith, T., Hebert, C., and Sauk, J. J. (1994). J. Cell. Biochem. 56, 518–526. Flaherty, K. M., Wilbanks, S. M., DeLuca-Flaherty, C., and McKay, D. B. (1994). J. Biol. Chem. 269, 12899–12907. Fukazawa, H., Li, P. M., Yamamoto, C., Murakami, Y., Mizuno, S., and Uehara, Y. (1991). Biochem. Pharmacol. 42, 1661–1671. Fukazawa, H., Uehara, Y., Murakami, Y., Mizuno, S., Hamada, M., and Takeuchi, T. (1994). FEBS Lett. 340, 155–158. Giannoukos, G., Silverstein, A. M., Pratt, W. B., and Simons, S. S., Jr. (1999). J. Biol. Chem. 274, 36527–36536. Grammatikakis, N., Lin, J.-H., Grammatikakis, A., Tsichlis, P. N., and Cochran, B. H. (1999). Mol. Cell. Biol. 19, 1661–1672. Grenert, J. P., Johnson, B. D., and Toft, D. O. (1999). J. Biol. Chem. 274, 17525–17533. Grenert, J. P., Sullivan, W. P., Fadden, P., Haystead, T. A. J., Clark, J., Mimnaugh, E., Krutzsch, H., Ochel, H.-J., Schulte, T. W., Sausville, E., Neckers, L. M., and Toft, D. O. (1997). J. Biol. Chem. 272, 23843–23850. Hankinson, O. (1994). Trends Endocrinol. Metabol. 5, 240–244. Hartson, S. D., Ottinger, E. A., Huang, W., Barany, G., Burn, P., and Matts, R. L. (1998). J. Biol. Chem. 273, 8475–8482. Hoffmann, K., and Handschumacher, R. E. (1995). Biochem. J. 307, 5–8. Hogenesch, J. B., Chan, W. K., Jackiw, V. H., Brown, R. C., Gu, Y. Z., Pray-Grant, M., Perdew, G. H., and Bradﬁeld, C. A. (1997). J. Biol. Chem. 272, 8581–8593. Hunter, T., and Poon, R. Y. C. (1997). Trends Cell Biol. 7, 157–161. Hutchison, K. A., Czar, M. J., Scherrer, L. C., and Pratt, W. B. (1992). J. Biol. Chem. 267, 14047–14053. Ishii, T., Udono, H., Yamano, T., Ohta, H., Uenaka, A., Ono, T., Hizuta, A., Tanaka, N., Srivastava, P. K., and Nakayama, E. (1999). J. Immunol. 162, 1303–1309. Jakob, U., Scheibel, T., Bose, S., Reinstein, J., and Buchner, J. (1996). J. Biol. Chem. 271, 10035–10041. Joab, I., Radanyi, C., Renoir, M., Buchou, T., Catelli, M.-G., Binart, N., and Mester, J. (1984). Nature 308, 850–853. Johnson, J. L., Beito, T. G., Krco, C. J., and Toft, D. O. (1994). Mol. Cell. Biol. 14, 1956–1963. Johnson, J. L., and Toft, D. O. (1994). J. Biol. Chem. 269, 24989–24993. Johnson, J. L., and Toft, D. O. (1995). Mol. Endocrinol. 9, 670–678. Johnson, L. N., Noble, M. E. M., and Owen, D. J. (1996). Cell 85, 149–158. Kampranis, S. C., Bates, A. D., and Maxwell, A. (1999). Proc. Natl. Acad. Sci. USA 96, 8414–8419. Kazlauskas, A., Poellinger, L., and Pongratz, I. (2000). J. Biol. Chem. 275, 41317–41324. Knighton, D. R., Zheng, J., Ten Eyck, L. F., Ashford, V. A., Xuong, N.-H., Taylor, S. S., and Sowadski, J. M. (1991). Science 253, 407–413. Knoblauch, R., and Garabedian, M. J. (1999). Mol. Cell. Biol. 19, 3748–3759. Kosano, H., Stensgard, B., Charlesworth, M. C., McMahon, N., and Toft, D. (1998). J. Biol. Chem. 273, 32973–32979. Kost, S. L., Smith, D. F., Sullivan, W. P., Welch, W. J., and Toft, D. O. (1989). Mol. Cell. Biol. 9, 3829–3838. Kuznetsov, G., Chen, L. B., and Nigam, S. K. (1994). J. Biol. Chem. 269, 22990–22995. Louvain, J.-F., Abbas-Terki, T., and Picard, D. (1998). Mol. Biol. Cell. 9, 3071–3083. Louvain, J.-F., Warth, R., and Picard, D. (1996). Proc. Natl. Acad. Sci. USA 93, 13937–13942.

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

183

Ma, Q., and Whitlock, J. P., Jr. (1997). J. Biol. Chem. 272, 8878–8884. Marcu, M. G., Chadli, A., Bouhouche, I., Catelli, M., and Neckers, L. M. (2000). J. Biol. Chem. 275, 37181–37186. Marsh, J. A., Kalton, H. M., and Gaber, R. F. (1998). Mol. Cell. Biol. 18, 7353–7359. Maruya, M., Sameshima, M., Nemoto, T., and Yahara, I. (1999). J. Molec. Biol. 285, 903–907. McGuire, J., Coumailleau, P., Whitelaw, M. L., Gustafsson, J. A., and Poellinger, L. (1995). J. Biol. Chem. 270, 31353–31357. Melnick, J., Aviel, S., and Argon, Y. (1992). J. Biol. Chem. 267, 21303–21306. Meyer, B. K., Pray-Grant, M. G., Vanden Heuvel, J. P., and Perdew, G. H. (1998). Mol. Cell. Biol. 18, 978–988. Miller, P., DiOrio, C., Moyer, M., Schnur, R. C., Bruskin, A., Cullen, W., and Moyer, J. D. (1994a). Cancer Res. 54, 2724–2730. Miller, P., Schnur, R. C., Barbacci, E., Moyer, M. P., and Moyer, J. D. (1994b). Biochem. Biophys. Res. Commun. 201, 1313–1319. Mimnaugh, E. G., Chavany, C., and Neckers, L. (1996). J. Biol. Chem. 271, 22796–22801. Mimnaugh, E. G., Worland, P. J., Whitesell, L., and Neckers, L. M. (1995). J. Biol. Chem. 270, 28654–28659. Morano, K. A., and Thiele, D. J. (1999). EMBO. J. 18, 5953–5962. Morishima, Y., Kanelakis, K. C., Silverstein, A. M., Dittmar, K. D., Estrada, L., and Pratt, W. B. (2000). J. Biol. Chem. 275, 6894–6900. Munoz, M. J., and Jimenez, J. (1999). Mol. Gen. Genet. 261, 242–250. Nadeau, K., Das, A., and Walsh, C. T. (1993). J. Biol. Chem. 268, 1479–1487. Nadeau, K., Sullivan, M. A., Bradley, M., Engman, D. M., and Walsh, C. T. (1992). Protein Sci. 1, 970–979. Nathan, D. F., and Lindquist, S. (1995). Mol. Cell. Biol. 15, 3917–3925. Nathan, D. F., Vos, M. H., and Lindquist, S. (1997). Proc. Natl. Acad. Sci. USA 94, 12949– 12956. Nathan, D. F., Vos, M. H., and Lindquist, S. (1999). Proc. Natl. Acad. Sci. USA 96, 1409–1414. Navarro, D., Qadri, I., and Pereira, L. (1991). Virology 184, 253–264. Nemoto, T., Ohara-Nemoto, Y., Ota, M., Takagi, T., and Yokoyama, K. (1995). Eur. J. Biochem. 233, 1–8. Nieland, T. J., Tan, M. C., Monne-van Muijen, M., Koning, F., Kruisbeek, A. M., and van Bleek, G. M. (1996). Proc. Natl. Acad. Sci. USA 93, 6135–6139. Obermann, W. M. J., Sondermann, H., Russo, A. A., Pavletich, N. P., and Hartl, F. U. (1998). J. Cell. Biol. 143, 901–910. Opperman, H., Levinson, W., and Bishop, J. M. (1981). Proc. Natl. Acad. Sci. USA 78, 1067–1071. Owens-Grillo, J. K., Czar, M. J., Hutchinson, K. A., Hoffman, K., Perdew, G. H., and Pratt, W. B. (1996a). J. Biol. Chem. 271, 13468–13475. Owens-Grillo, J. K., Hoffman, K., Hutchinson, K. A., Yem, A. W., Deibel, M. R. J., Handschumacher, R. E., and Pratt, W. B. (1995). J. Biol. Chem. 270, 20479–20484. Owens-Grillo, J. K., Stancato, L. F., Hoffmann, K., Pratt, W. B., and Krishna, P. (1996b). Biochemistry 35, 15249–15255. Panaretou, B., Prodromou, C., Roe, S. M., O’Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1998). EMBO J. 17, 4829–4836. Poellinger, L., Gottlicher, M., and Gustafsson, J. A. (1992). Trends Pharmacol. Sci. 13, 241– 245. Prapapanich, V., Chen, S., Nair, S. C., Rimerman, R. A., and Smith, D. F. (1996). Mol. Endocrinol. 10, 420–431. Pratt, W. B. (1998). Proc. Soc. Exp. Biol. Med. 217, 420–434.

184

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

Pratt, W. B., and Dittmar, K. D. (1998). Trends Endocrinol. Metabol. 9, 244–252. Privalsky, M. L. (1991). J. Biol. Chem. 266, 1456–1462. Prodromou, C., Panaretou, B., Chohan, S., Siligardi, G., O’Brien, R., Ladbury, J. E., Roe, S. M., Piper, P. W., and Pearl, L. H. (2000). EMBO J. 19, 4383–4392. Prodromou, C., Piper, P. W., and Pearl, L. H. (1996). Proteins: Struct. Funct. Genet. 25, 517–522. Prodromou, C., Roe, S. M., O’Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1997a). Cell 90, 65–75. Prodromou, C., Roe, S. M., Piper, P. W., and Pearl, L. H. (1997b). Nature Struct. Biol. 4, 477–482. Prodromou, C., Siligardi, G., O’Brien, R., Woolfson, D. N., Regan, L., Panaretou, B., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1999). EMBO J. 18, 754–762. Rafestin-Oblin, M. E., Couette, B., Radanyi, C., Lombes, M., and Baulieu, E. E. (1989). J. Biol. Chem. 264, 9304–9309. Ramsey, A. J., Russell, L. C., Whitt, S. R., and Chinkers, M. (2000). J. Biol. Chem. 275, 17857–17862. Ratajczak, T., and Carrello, A. (1996). J. Biol. Chem. 271, 2961–2965. Ratajczak, T., Hlaing, J., Brockway, M. J., and Hahnel, R. (1990). J. Steroid Biochem. 35, 543–553. Redeuilh, G., Moncharmont, B., Secco, C., and Baulieu, E. E. (1987). J. Biol. Chem. 262, 6969–6975. Roe, S. M., Prodromou, C., O’Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1999). J. Med. Chem. 42, 260–266. Rutherford, S. L., and Lindquist, S. (1998). Nature 396, 336–342. Sanchez, E. R., Toft, D. O., Schlesinger, M. J., and Pratt, W. B. (1985). J. Biol. Chem. 260, 12398–12401. Sastry, S., and Linderoth, N. (1999). J. Biol. Chem. 274, 12023–12035. Sato, S., Fujita, N., and Tsuruo, T. (2000). Proc. Natl. Acad. Sci. USA 97, 10832–10837. Schaiff, W. T., Hruska, K. A., Jr., McCourt, D. W., Green, M., and Schwartz, B. D. (1992). J. Exp. Med. 176, 657–666. Scheibel, T., Siegmund, H. I., Jaenicke, R., Ganz, P., Lilie, H., and Buchner, J. (1999). Proc. Natl. Acad. Sci. USA 96, 1297–1302. Scheibel, T., Weikl, T., and Buchner, J. (1998). Proc. Natl. Acad. Sci. USA 95, 1495–1499. Scheuﬂer, C., Brinker, A., Bourenkov, G., Pegoraro, S., Moroder, L., Bartunik, H., Hartl, F. U., and Moareﬁ, I. (2000). Cell 101, 199–210. Schneider, C., Sepp-Lorenzino, L., Nimmesgern, E., Ouerfelli, O., Danishefsky, S., Rosen, N., and Hartl, F. U. (1996). Proc. Natl. Acad. Sci. USA 93, 14536–14541. Schnur, R. C., Corman, M. L., Gallaschun, R. J., Cooper, B. A., Dee, M. F., Doty, J. L., Muzzi, M. L., DiOrio, C. I., Barbacci, E. G., Miller, P. E., et al. (1995a). J. Med. Chem. 38, 3813–3820. Schnur, R. C., Corman, M. L., Gallaschun, R. J., Cooper, B. A., Dee, M. F., Doty, J. L., Muzzi, M. L., Moyer, J. D., DiOrio, C. I., Barbacci, E. G., et al. (1995b). J. Med. Chem. 38, 3806–3812. Schuh, S., Yonemoto, W., Brugge, J., Bauer, V. J., Riehl, R. M., Sullivan, W. P., and Toft, D. O. (1985). J. Biol. Chem. 260, 14292–14296. Schulte, T. W., Blagosklonny, M. V., Ingui, C., and Neckers, L. (1995). J. Biol. Chem. 270, 24585–24588. Schutz, A. R., Giddings, T. H., Steiner, E., and Winey, M. (1997). J. Cell. Biol. 136, 969–982. Sepehrnia, B., Paz, I. B., Dasgupta, G., and Momand, J. (1996). J. Biol. Chem. 271, 15084– 15090.

STRUCTURE, FUNCTION, AND MECHANISM OF THE

Hsp90

185

Shaknovich, R., Shue, G., and Kohtz, D. S. (1992). Mol. Cell. Biol. 12, 5059–5068. Sharma, S. V., Agatsuma, T., and Nakano, H. (1998). Oncogene 16, 2639–2645. Shue, G., and Kohtz, D. S. (1994). J. Biol. Chem. 269, 2707–2711. Silverstein, A. M., Grammatikakis, N., Cochran, B. H., Chinkers, M., and Pratt, W. B. (1998). J. Biol. Chem. 273, 20090–20095. Smith, D. F. (1993). Mol. Endocrinol. 7, 1418–1429. Smith, D. F. (2000). Semin. Cell. Dev. Biol. 11, 45–52. Smith, D. F., Albers, M. W., Schreiber, S. L., Leach, K. L., and Deibel, M. R., Jr. (1993a). J. Biol. Chem. 268, 24270–24273. Smith, D. F., Baggenstoss, B. A., Marion, T. N., and Rimerman, R. A. (1993b). J. Biol. Chem. 268, 18365–18371. Smith, D. F., Faber, L. E., and Toft, D. O. (1990). J. Biol. Chem. 265, 3996–4003. Smith, D. F., Stensgard, B. A., Welch, W. J., and Toft, D. O. (1992). J. Biol. Chem. 267, 1350–1356. Smith, D. F., Sullivan, W. P., Marion, T. N., Zaitsu, K., Madden, B., McCormick, D. J., and Toft, D. O. (1993c). Mol. Cell. Biol. 13, 869–876. Smith, D. F., and Toft, D. O. (1993). Mol. Endocrinol. 7, 4–11. Smith, D. F., Whitesell, L., Nair, S. C., Chen, S., Prapapnich, V., and Rimerman, R. A. (1995). Mol. Cell. Biol. 15, 6804–6812. Soga, S., Kozawa, T., Narumi, H., Akinaga, S., Irie, K., Matsumoto, K., Sharma, S. V., Nakano, H., Mizukami, T., and Hara, M. (1998). J. Biol. Chem. 273, 822–828. Soga, S., Neckers, L. M., Schulte, T. W., Shiotsu, Y., Akasaka, K., Narumi, H., Agatusuma, T., Ikuina, Y., Murakata, C., Tamaoki, T., and Akinaga, S. (1999). Cancer Res. 59, 2931– 2938. Song, H. Y., Dunbar, J. D., Zhang, Y. X., Guo, D., and Donner, D. B. (1995). J. Biol. Chem. 270, 3574–3581. Spence, J., Cegielska, A., and Georgopoulos, C. (1990). J. Bacteriol. 172, 7157–7166. Srivastava, P. K., Deleo, A. B., and Old, L. J. (1986). Proc. Natl. Acad. Sci. USA 83, 3407–3411. Srivastava, P. K., Menoret, A., Basu, S., Binder, R. J., and McQuade, K. L. (1998). Immunity 8, 657–665. Stancato, L. F., Chow, Y.-H., Hutchinson, K. A., Perdew, G. H., Jove, R., and Pratt, W. B. (1993). J. Biol. Chem. 268, 21711–21716. Stancato, L. F., Silverstein, A. M., Gitler, C., Groner, B., and Pratt, W. B. (1996). J. Biol. Chem. 271, 8831–8836. Stancato, L. F., Silverstein, A. M., Owens-Grillo, J. K., Chow, Y. H., Jove, R., and Pratt, W. B. (1997). J. Biol. Chem. 272, 4013–4020. Stebbins, C. E., Russo, A. A., Schneider, C., Rosen, N., Hartl, F. U., and Pavletich, N. P. (1997). Cell 89, 239–250. Stepanova, L., Leng, X. H., Parker, S. B., and Harper, J. W. (1996). Genes Dev. 10, 1491– 1502. Sullivan, W., Stensgard, B., Caucutt, G., Bartha, B., McMahon, N., Alnemri, E. S., Litwack, G., and Toft, D. O. (1997). J. Biol. Chem. 272, 8007–8012. Sullivan, W. P., and Toft, D. O. (1993). J. Biol. Chem. 268, 20373–20379. Tai, P. K., Albers, M. W., Chang, H., Faber, L. E., and Schreiber, S. L. (1992). Science 256, 1315–1318. Tai, P. K., Maeda, Y., Nakao, K., Wakim, N. G., Duhring, J. L., and Faber, L. E. (1986). Biochemistry 25, 5269–5275. Tanaka, T., Saha, S. K., Tomomori, C., Ishima, R., Liu, D. J., Tong, K. I., Park, H., Dutta, R., Qin, L., Swindells, M. B., Yamazaki, T., Ono, A. M., Kainosho, M., Inouye, M., and Ikura, M. (1998). Nature 396, 88–92.

186

LAURENCE H. PEARL AND CHRISOSTOMOS PRODROMOU

Taniguchi, M., Uehara, Y., Matsuyama, M., and Takahashi, M. (1993). Biochem. Biophys. Res. Commun. 195, 208–214. Veldscholte, J., Berrevoets, C. A., Brinkmann, A. O., Grootegoed, J. A., and Mulder, E. (1992). Biochemistry 31, 2393–2399. Walker, A. I., Hunt, T., Jackson, R. J., and Anderson, C. W. (1985). EMBO J. 4, 139–145. Wearsch, P. A., and Nicchitta, C. V. (1997). J. Biol. Chem. 272, 5152–5156. Wearsch, P. A., Voglino, L., and Nicchitta, C. V. (1998). Biochemistry 37, 5709–5719. Weikl, T., Muschler, P., Richter, K., Veit, T., Reinstein, J., and Buchner, J. (2000). J. Mol. Biol. 303, 583–592. Welch, W. J., and Feramisco, J. R. (1982). J. Biol. Chem. 257, 14949–14959. Whitelaw, M. L., Gottlicher, M., Gustafsson, J. A., and Poellinger, L. (1993). EMBO J. 12, 4169–4179. Whitesell, L., Mimnaugh, E. G., De Costa, B., Myers, C. E., and Neckers, L. M. (1994). Proc. Natl. Acad. Sci. USA 91, 8324–8328. Wiech, H., Buchner, J., Zimmermann, R., and Jakob, U. (1992). Nature 358, 169–170. Wigley, D. B., Davies, G. J., Dodson, E. J., Maxwell, A., and Dodson, G. (1991). Nature 351, 624–629. Wilhelmsson, A., Cuthill, S., Denis, M., Wikstrom, A. C., Gustafsson, J. A., and Poellinger, L. (1990). EMBO J. 9, 69–76. Xu, M., Dittmar, K. D., Giannoukos, G., Pratt, W. B., and Simons, S. S. (1998). J. Biol. Chem. 273, 13918–13924. Young, J. C., and Hartl, F. U. (2000). EMBO J. 19, 5930–5940. Young, J. C., Obermann, W. M. J., and Hartl, F. U. (1998). J. Biol. Chem. 273, 18007–18010. Young, J. C., Schneider, C., and Hartl, F. U. (1997). FEBS Lett. 418, 139–143.

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY DESIGNED FOR CONTROLLED PROTEOLYSIS ¨ By PETER ZWICKL, ERIKA SEEMULLER, BARBARA KAPELARI, and WOLFGANG BAUMEISTER Department of Molecular Structural Biology, Max-Planck Institute for Biochemistry, Am Klopferspitz 18a, 82152 Martinsried, Germany

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. The 20S Proteasome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Occurrence and Subunit Composition of 20S Proteasomes . . . . . . . . . . . B. Structural Features of the 20S Proteasome . . . . . . . . . . . . . . . . . . . . . . . . . . C. Catalytic Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Processing and Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Proteolytic Activity and Degradation Products . . . . . . . . . . . . . . . . . . . . . . . III. Activators of the 20S Proteasome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The Archaeal and Bacterial AAA ATPases . . . . . . . . . . . . . . . . . . . . . . . . . . B. The 19S Regulatory Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. The PA28 Activator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

187 188 188 190 193 196 199 202 202 203 212 213 213

I. INTRODUCTION The ubiquitin–proteasome pathway [for reviews see Hershko and Ciechanover (1998); Varshavsky et al. (2000)] is the major route used by eukaryotic cells for disposing of misfolded or damaged proteins and for controlling the life-span of regulatory proteins (Kirschner, 1999). The 26S proteasome is a huge molecular machine of approximately 2.5 MDa which degrades protein substrates by an energy-dependent mechanism [for reviews see Coux et al. (1996); Voges et al. (1999)]. It comprises two subcomplexes, the 20S core particle and one or two regulatory complexes, the 19S caps [for reviews see Tanaka (1998); DeMartino and Slaughter (1999); Gorbea and Rechsteiner (2000)]. The 20S complex allows the proteolytic action to be conﬁned to a nanocompartment in which substrates, sequestered from the cellular environment, undergo degradation (Lupas et al., 1997b). The 19S regulatory complex recruits substrates marked for degradation and prepares them for translocation into the 20S core complex (Lupas et al., 1993; Larsen and Finley, 1997). The sequence of events encountered by a substrate until it is ﬁnally degraded is reﬂected by a linear arrangement of functional modules within the 45-nm supramolecular assembly. 187 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0096-5332/02 $35.00

188

PETER ZWICKL ET AL.

Whereas the structure and function of the 20S proteasome have been elucidated in great detail [for review see Baumeister et al. (1998)], the 19S regulator is understood only dimly at present. Structural studies are hampered by the low intrinsic stability of this assembly, which makes it notoriously difﬁcult to obtain homogeneous preparations. Nevertheless, analyses of yeast, Drosophila, and human 26S proteasomes have revealed a common set of 17 to 18 subunits, although a few species-speciﬁc differences have been found (Tanaka, 1998; Glickman et al., 1998b; Ferrell et al., 2000; H¨olzl et al., 2000; Verma et al., 2000). These subunits can be assigned to two subcomplexes (Glickman et al., 1998a): the “base” part, which comprises an array of six paralogous AAA ATPases and is sufﬁcient to support the degradation of (partially) unfolded proteins, and the “lid” part, which provides the link to the ubiquitin system, which, in turn, confers selectivity (Pickart, 2000; Wilkinson, 2000). Consistent with the lack of a ubiquitin system in archaea and bacteria (Ruepp et al., 2000), minimal homohexameric AAA ATPase complexes are sufﬁcient for regulating prokaryotic 20S proteasomes (Horwich et al., 1999; Wickner et al., 1999; Schmidt et al., 1999a; Zwickl et al., 2000a). In this chapter, we review the current knowledge of the structure, assembly, and function of the 20S proteasome and its regulators in prokaryotic and eukaryotic cells. II. THE 20S PROTEASOME A. Occurrence and Subunit Composition of 20S Proteasomes As the number of sequenced genomes grows, a progressively clearer picture of the species distribution of proteasomes is emerging. In general, it appears that proteasomes are ubiquitous and essential in eukaryotes (Heinemeyer, 2000), ubiquitous but not essential in archaea (Ruepp et al., 1998), and rare and nonessential in bacteria in which other energydependent proteases abound (Knipfer and Shrader, 1997). While only one α-subunit gene and one β-subunit gene have been found in archaeal genomes, a few species contain a second α or a second β gene (Zwickl et al., 2000b; Maupin-Furlow et al., 2000). To what extent these additional genes are expressed and the subunits incorporated into proteasome particles remains to be investigated. In the halophilic archaeon Haloferax volcanii three proteasomal genes have been found, one coding for a β subunit and two coding for α-type subunits. 20S Proteasomes isolated from Haloferax cells contain two distinct particle populations, one built of α 1 and β subunits and one containing all

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

189

three subunits (α 1, α 2, and β) (Wilson et al., 1999). Other archaea, such as Pyrococcus furiosus and Pyrobaculum aerophilum, were found to have a single α-type and two β-type subunits (Bauer et al., 1997; Mallick et al., 2000). Whereas both β-type subunits from P. furiosus share the N-terminal active-site threonine, this residue is replaced in one of the two β-type subunits from P. aerophilum by an alanine (Zwickl et al., 2000b; MaupinFurlow et al., 2000). Therefore, one of the two β-type subunits is inactive. Archaeal proteasomes containing both active and inactive β-type subunits would have a reduced number of active sites, an evolutionary precedent for eukaryotic proteasomes, which have only 6 active sites as compared to 14 in the archaetypal Thermoplasma acidophilum proteasome. In bacteria, genuine proteasomes of the α+β-type have hitherto been found only in species belonging to the order Actinomycetales. With the exception of Rhodococcus erythropolis, where the 20S proteasomes are built of two α- and two β-type subunits (Tamura et al., 1995), proteasomes from all other species are composed of a single α- and a single β-type subunit (De Mot et al., 1999; Zwickl et al., 2000b). Many other bacteria contain a gene, called hslV in Escherichia coli, which encodes a protein that is closely related to the proteasomal β-type subunits (Lupas et al., 1994) and forms a ring-shaped dodecamer (Bochtler et al., 1997). Occurrence of proteasomes and the HslV protease seems to be mutually exclusive; either one or the other, or none is present in the same species (De Mot et al., 1999; Zwickl et al., 2000b). The HslV protease associates directly, i.e., without an intercessory α ring, with the HslU ATPase to form the ATP-dependent protease HslVU (Rohrwild et al., 1996; Kessel et al., 1996). The exact structure of the HslVU protease complex has been a matter of some debate recently. It is unlikely that the structure of the E. coli complex described by Bochtler and colleagues represents a physiologically relevant form (Bochtler et al., 2000a; 2000b; Song et al., 2000). Two more recent crystal structures of the HslVU complex, one from E. coli (Wang et al., 2001) and one from Haemophilus influenzae (Sousa et al., 2000), show, in agreement with results from two electron microscopy studies (Rohrwild et al., 1997; Ishikawa et al., 2000), a quite different mode of interaction between the HslV protease and the HslU ATPase. In eukaryotes, such as Saccharomyces cerevisiae, Caenorhabditis elegans, Drosophila melanogaster, and Oryza sativa, seven different α and seven different β genes have been found (Hughes, 1997; Bouzat et al., 2000; Sassa et al., 2000). In higher eukaryotes with an adaptive immune system, γ -interferon stimulates expression of three additional β-type subunits, which can replace the closely related, constitutively expressed active β-type subunits (Rock and Goldberg, 1999; Rechsteiner et al., 2000).

190

PETER ZWICKL ET AL.

This results in a modulation of proteolytic speciﬁcity. Interestingly, a superfamily of 23 proteasomal genes has been identiﬁed in Arabidopsis thaliana, which can be grouped into 14 distinct subfamilies encoding isoforms (Parmentier et al., 1997; Fu et al., 1998a). The functional signiﬁcance of these isoforms is unclear at present. Alignment of all the available sequences of proteasome subunits shows that the two families, α and β, originate from a gene duplication event, which must have taken place early in evolution (Zwickl et al., 1992; Hughes and Yeager, 1997; Bouzat et al., 2000; Zwickl et al., 2000b). B. Structural Features of the 20S Proteasome In spite of the differences in subunit complexity, the quaternary structure is highly conserved in 20S proteasomes from all three domains of life—archaea, bacteria, and eukaryotes. The 28 subunits, 14 of the α type and 14 of the β type, are grouped into four seven-membered rings, which collectively form a barrel-shaped complex with a length of 15 nm and a diameter of 11 nm (Fig. 1A). The two adjacent β-subunit rings enclose the central cavity (CC) with a diameter of approximately 5 nm, which harbors the active sites. It is connected via two narrow constrictions with two slightly smaller outer cavities, the “antechambers” (AC), which are formed jointly by one α and one β ring. An axial pore in the α rings gives access to the antechambers (see below and Fig. 1B). In prokaryotic proteasomes, built of 14 α and 14 β subunits, the rings are homomeric, and the complex is described by an α 7β 7β 7α 7 stoichiometry. Correspondingly, the stoichiometry of eukaryotic 20S proteasomes is α 1–7β 1–7β 1–7α 1–7. Each of the 14 different subunits is present in two copies within one complex and occupies a precisely deﬁned position. Thus the multiple axes of symmetry of the Thermoplasma proteasome are reduced to a C2 symmetry in the eukaryotic proteasome; each of the 14 different eukaryotic subunits is found twice within the complex and occupies well deﬁned positions (Schauer et al., 1993; L¨owe et al., 1995; Kopp et al., 1997; Groll et al., 1997). To indicate subunit positions within the rings, a systematic nomenclature for proteasome subunits has been proposed: Subunits are numbered α1 to α7 and β1 to β7, and those related by C2 symmetry are distinguished by the prime symbol [α1′ to α7′ and β1′ to β7′ (Fig. 1C)]. In mammals, the complexity of proteasomes is further enhanced by the fact that after γ -interferon stimulation three of the constitutive subunits (β1, β2, and β5) can be replaced by closely related subunits to form immunoproteasomes; according to the new nomenclature the inducible subunits are termed β1i, β2i, and β5i (Groll et al., 1997). Since only three subunits (β1 or β1i, β2 or β2i, β5 or β5i)

FIG. 1. Structure of the 20S proteasome. (A) Low-resolution model (1.2 nm) of the 20S proteasome derived from the crystal structure of the Thermoplasma proteasome (L¨owe et al., 1995). The α subunits form the heptameric outer rings; the β subunits, the inner rings. (B) The same structure cut open along the sevenfold axis to display the two antechambers (AC) and the central chamber (CC) with the 14 active sites (marked in black). The channel openings at the two ends of the cylinder are 1.3 nm in diameter. (C) Schematic representation of the arrangement of subunits within eukaryotic 20S proteasomes (Groll et al., 1997). Open boxes represent proteolytically inactive subunits; ﬁlled boxes, proteolytically active subunits. The single C2 symmetry axis is shown. (D) Fold of α and β subunits of the Thermoplasma proteasome (L¨owe et al., 1995). A pair of ﬁvestranded β sheets is ﬂanked on both sides by α-helices. Helices (H) and strands (S) are numbered HO to H5 and S1 to S10. The β subunits lack helix HO, which occupies the cleft on one side of the β-sheet sandwich in the α subunits. The active-site threonine (Thr-1) of β subunits is shown in a ball-and-stick representation.

192

PETER ZWICKL ET AL.

of the β-type subunits present in a single proteasome particle are proteolytically active, eukaryotic proteasomes contain a total of 6 active sites, whereas prokaryotic proteasomes, built of identical copies of β subunits, have 14 active sites (see below). Another feature that appeared to distinguish eukaryotic from prokaryotic proteasomes relates to the path of substrate entry into the inner cavities. Although the crystal structure of the Thermoplasma proteasome has revealed the existence of an axial channel (∼1.3 nm wide) in the α rings (L¨owe et al., 1995)—and electron microscopy of 20S proteasomes incubated with gold-labeled substrates has demonstrated that this channel is in fact the entry site (Wenzel and Baumeister, 1995)—there was no obvious passageway for substrate entry or exit in the structure of the yeast 20S proteasome (Groll et al., 1997). Meanwhile, it has become clear that the yeast particle had been crystallized in a closed form in which the N-terminal tails of the α-subunits interdigitate to form a plug; these tails are disordered in the archaeal proteasome, and therefore the gate appears to be open. Deletion of the nine N-terminal residues of one subunit (α3) of the yeast proteasome results in disorder of a number of other subunits and creates an opening which is very similar to that seen in the Thermoplasma proteasome (Groll et al., 2000a). While eukaryotic wild-type proteasomes are almost “latent,” this mutant proteasome has a much enhanced peptidase activity, probably because access is facilitated. In vivo, associations with regulatory complexes, e.g., the 19S regulatory complex, the PA28 activator, or the PA26 activator, are known to activate 20S proteasomes; all these regulatory complexes interact with the terminal α rings and are likely to function by mechanisms that open the gate for substrate uptake (or product release) (see Sections III.B.3 and III.C) (Pickart and VanDemark, 2000; Rechsteiner et al., 2000; Whitby et al., 2000). As anticipated from their sequence similarity, the (noncatalytic) α- and the (catalytic) β-type subunits have the same fold (L¨owe et al., 1995; Groll et al., 1997): a four-layer α+β structure with two antiparallel ﬁve-stranded β sheets, ﬂanked on one side by two, and on the other side by three α-helices (Fig. 1D). In the β-type subunits, the β-sheet sandwich is closed at one end by four hairpin loops and open at the opposite end to form the active-site cleft; the cleft is oriented toward the inner surface of the central cavity. In the α-type subunits an additional helix formed by an N-terminal extension crosses the top of the β-sheet sandwich and ﬁlls this cleft. Initially, the proteasome fold was believed to be unique; however, it turned out to be prototypical of a new superfamily of proteins referred to as Ntn (N-terminal nucleophile) hydrolases (Brannigan et al., 1995). Beyond the common fold, members of this family share the

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

193

mechanisms of the nucleophilic attack and self-processing (Dodson and Wlodawer, 1998; Oinonen and Rouvinen, 2000). C. Catalytic Mechanism Site-directed mutagenesis and the crystal structure analysis of a proteasome-inhibitor complex identiﬁed the amino-terminal threonine (Thr1) of Thermoplasma β subunits as both the catalytic nucleophile and the primary proton acceptor (Seemuller ¨ et al., 1995; L¨owe et al., 1995): The hydroxyl group of Thr1 attacks the carbonyl carbon of the scissile peptide bond, while its amino group serves as the primary proton acceptor, which enhances the nucleophilicity by stripping the proton from the side-chain hydroxyl; for steric reasons a water molecule is likely to mediate the proton transfer (Fig. 2). Such a “single-residue” active site is a characteristic feature of Ntn hydrolases, as is the autocatalytic removal of a prosequence, which is necessary to expose the nucleophilic group. For this autocatalytic cleavage Thr1Oγ of the proteasome β-subunit precursor adds to the carbonyl carbon of Gly-1, and a water molecule is supposed to fulﬁll the base function of the Thr1 α-amino group in the mature enzyme (Fig. 3) (Ditzel et al., 1998). Both Gly-1 and Thr1 are invariant residues of all “active” proteasome subunits. In other Ntn hydrolases the nucleophile is provided either by a threonine, cysteine, or serine residue, and the residues at the −1 position are variable. In proteasomes, replacement of the active-site threonine by serine does not alter the rates of hydrolysis of small ﬂuorogenic peptide substrates, but the mutant is signiﬁcantly slower in degrading larger peptides and proteins (Seemuller ¨ et al., 1995; Maupin-Furlow et al., 1998; Kisselev et al., 2000). Self-processing of β-subunit precursors is inefﬁcient with serine at the +1 position (Seemuller ¨ et al., 1996; Chen and Hochstrasser, 1996; Heinemeyer et al., 1997). Conversely, substitution of the threonine by a cysteine yields proteasomes that are unable to cleave any substrate, but the self-cleavage reaction remains unaffected by this mutation (Seemuller ¨ et al., 1996). Besides the N-terminal threonine, several other residues of proteasome β subunits (Glu17, Lys33, Asp166) are required for the proteolytic activity, although their precise contribution to the catalytic mechanism awaits further clariﬁcation (L¨owe et al., 1995; Seemuller ¨ et al., 1996; Chen and Hochstrasser, 1996; Schmidtke et al., 1996; Heinemeyer et al., 1997; Maupin-Furlow et al., 1998; Mayr et al., 1998). The catalytic role of the N-terminal threonine of β subunits was soon extended to eukaryotic proteasomes by the observation that binding of lactacystin, a Streptomyces metabolite, to the mammalian β-type subunit β5/X irreversibly inhibits its proteolytic activity

194

FIG. 2. Schematic diagram of the catalytic mechanism of 20S proteasomes. A proton transfer from the hydroxyl group of Thr1 of β subunits to its own terminal amino group initiates the nucleophilic attack (I). As a result of the nucleophilic addition to the carbonyl carbon of the scissile peptide bond, a tetrahedral intermediate is formed (II). By an N O acyl rearrangement, an ester is formed (the acyl enzyme); and the amino-terminal cleavage product is released (III). Finally, hydrolysis of the acyl enzyme yields the carboxyl-terminal cleavage product and frees the enzyme for another reaction cycle (IV).

FIG. 3. Schematic diagram of the self-processing reaction. Thr1Oγ of immature β subunits adds to the carbonyl carbon of Gly-1, which results in a cyclic tetrahedral intermediate (I, II). The C N bond is then cleaved and an ester compound is formed (III). Hydrolysis of the ester yields the propeptide and the mature β subunit with Thr1 at its N terminus (IV). 195

196

PETER ZWICKL ET AL.

(Fenteany et al., 1995). Meanwhile, a large number of mutational studies with other archaeal (Maupin-Furlow et al., 1998), bacterial (Mayr et al., 1998), yeast (Chen and Hochstrasser, 1996; Arendt and Hochstrasser, 1997; Heinemeyer et al., 1997; Groll et al., 1999), and mammalian β-type subunits (Schmidtke et al., 1996; Salzmann et al., 1999) have conﬁrmed a common proteolytic mechanism for all kinds of proteasomes. It remains enigmatic, however, why in eukaryotic proteasomes four out of seven β subunits lack the N-terminal active-site threonine, and therefore are presumably inactive. D. Processing and Assembly Proteolytically active β-type subunits are synthesized in an inactive precursor form containing N-terminal extensions of variable lengths, the propeptides, which must be removed posttranslationally to allow the formation of active sites. This process is tied in with the assembly of the 20S proteasome in such a manner that activation is delayed until assembly is complete and the active sites are sequestered from the cellular environment. Cleavage of the propeptide proceeds autocatalytically, relying on the active-site threonine, and the invariant glycine at position −1 appears to be the prime determinant of the cleavage site (Schmidtke et al., 1996; Seemuller ¨ et al., 1996; Chen and Hochstrasser, 1996). It is as yet unclear whether the reaction is intramolecular (cis) or intermolecular (trans) or whether both mechanisms apply. Experiments with Thermoplasma proteasomes in which inactive mutant β subunits were found to be correctly processed in the presence of wild-type β subunits were indicative of an intermolecular mechanism (Seemuller ¨ et al., 1996). An inactive mutant form of the mammalian subunit β1i/LMP2, in contrast, is not correctly processed, in spite of being incorporated into the complex along with active neighbors (Schmidtke et al., 1996). This ﬁnding would argue against an intermolecular mechanim of processing, which is indeed difﬁcult to reconcile with our current understanding that processing occurs at a late stage of proteasome assembly. Interestingly, mature eukaryotic proteasomes contain partially processed “precursor” subunits; the inactive subunits β 6 and β 7, both synthesized in a precursor form, have their N termini removed. However, cleavage of these subunits (and of inactive mutant subunits) occurs at some distance upstream of the consensus cleavage site of the active subunits (Gly-1/Thr1), leaving short propeptide remnants. The existence of such processing intermediates has led to the proposal of a two-step model for precursor processing: The propeptide is trimmed in

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

197

size in a ﬁrst intermolecular step, and completely removed in a second intramolecular step (Schmidtke et al., 1996). Several reports, in particular a recent study combining mutational, biochemical, and crystallographic data, demonstrated that processing of naturally inactive or mutionally inactivated subunits indeed relies on the nearest neighbor in the complex carrying an active site (Schmidtke et al., 1996; Heinemeyer et al., 1997; Ditzel et al., 1998; Groll et al., 2000b). Whether such an intermolecular cleavage merely serves to remove a bulky protrusion from inactive subunits or such a trimming step is linked in a more intricate manner to the subsequent intramolecular cleavages is as yet not clear. The pathways of 20S proteasome assembly and also the roles of the β propeptides appear to differ somewhat among archaea, bacteria, and eukaryotes. The α subunits of Thermoplasma and of other archaea spontaneously form seven-membered rings in the absence of β subunits (Zwickl et al., 1994; Maupin-Furlow et al., 1998; Wilson et al., 2000). β Subunits, in turn, remain monomeric, unprocessed, and inactive, and do not even fold properly in the absence of α subunits. Thus, it appears that the α rings serve as a template upon which the β subunits assemble. Coexpression of both archaeal genes yields fully assembled and functional proteasomes, irrespective of the presence or absence of the relatively short (6 or 10 residues) β-subunit propeptides. Proteasomes can even be reassembled after dissociation, also demonstrating that the propeptides are not required for assembly (Grziwa et al., 1994; Wilson et al., 1999). In contrast, neither α nor β subunits of the bacterial Rhodococcus proteasome assemble by themselves, but complexes are formed as soon as both types of subunits are allowed to interact. This suggests that assembly starts with the formation of an α/β heterodimer, although halfproteasomes, built of one ring of α subunits and one ring of β-subunit precursors, are the ﬁrst intermediates which have been captured so far. Dimerization of half-proteasomes, which are still inactive, triggers processing of the β-precursor subunits, and fully assembled, proteolytically active proteasomes are formed. Although the β propeptides are not essential for the formation of Rhodococcus proteasomes, they have a strong effect on assembly efﬁciency. The 65-residue propeptide of Rhodococcus β 1, which can be provided in trans, has two functions: It supports the initial folding of the β subunit, and it promotes the ﬁnal maturation of proteasomes after the docking of two half-proteasomes (Zuhl ¨ et al., 1997). The chaperone-like manner by which the propeptide accelerates subunit folding resembles that of the well studied propeptides of some extracellular bacterial proteases, such as α-lytic protease and subtilisin [see for example Baker et al. (1993)].

198

PETER ZWICKL ET AL.

Assembly of eukaryotic proteasomes is obviously a more complicated process. The correct positioning of each of the 14 different, but closely related subunits must be orchestrated. Since some of the eukaryotic α-type subunits are capable of assembling into homomeric ring structures similar to Thermoplasma α subunits, it appears possible that α rings have a scaffolding function for the incorporation of β subunits (Gerards et al., 1997; Yao et al., 1999b). Detailed studies of precursor complexes of mouse 20S proteasomes with subunit-speciﬁc antibodies indeed identiﬁed intermediates that contain all α-type subunits and a subset of “early” β-type subunits (β2, β3, and β4). Incorporation of the four remaining β subunits yields half-proteasomes and triggers their dimerization; subsequent autocatalytic processing and conformational changes are supposed to complete proteasome maturation (Nandi et al., 1997). Previously described 13S–16S precursor complexes are probably complete or incomplete half-proteasome intermediates (Yang et al., 1995; Schmidtke et al., 1997). The eukaryotic β propeptides are not only highly variable in length and sequence, but their importance for the assembly process varies. Those of yeast β5/Doa3 (Chen and Hochstrasser, 1996; Heinemeyer et al., 1997) and mammalian β5i/LMP7 (Cerundolo et al., 1995), both more than 70 residues long, are essential for subunit incorporation, suggesting an intramolecular chaperone function resembling that of the Rhodococcus β1 propeptide. Of less importance are the propeptides of the human β1i/LMP2 and yeast β2 subunits, which were shown to improve the efﬁciency of subunit incorporation (Groettrup et al., 1997; Schmidt et al., 1999b; J¨ager et al., 1999). Interestingly, propeptides do not seem to have a function in correct positioning of β-type subunits, as the replacement of the propeptide of β5i/LMP7 with the propeptide of β2i/LMP2 does not change the position of the β5i/LMP7 subunit within the particle (Schmidtke et al., 1996). Recently, removal of the propeptides of yeast β1, β2, and β5 has shown that another critical function of β propeptides might be to protect the catalytic threonine against Nα-acetylation (Arendt and Hochstrasser, 1999; J¨ager et al., 1999; Kimura et al., 2000). While the efﬁcient assembly of prokaryotic proteasomes proceeds autonomously, assembly of eukaryotic proteasomes appears to require extrinsic maturation factors. The chaperone protein Hsc73 was found to associate with mammalian proteasome precursor complexes, but not with mature proteasomes (Schmidtke et al., 1997). Similarly, Ump1, a small protein contained in half-proteasome precursor complexes of yeast proteasomes, is supposed to have a chaperone-like function. Ump1

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

199

is essential for maturation of wild-type proteasomes but not for mutants lacking the β5 propeptide, suggesting a special interaction between Ump1 and the propeptide (Ramos et al., 1998). Mammalian homologs of Ump1 have recently been described (Burri et al., 2000; Witt et al., 2000). E. Proteolytic Activity and Degradation Products When tested with short ﬂuorogenic peptides, proteasomes display “chymotrypsin-like” (cleavage after hydrophobic residues), “trypsin-like” (cleavage after basic residues), “peptidylglutamylpeptide-hydrolyzing” (PGPH) activity (cleavage after acidic residues), as deﬁned on the basis of the P1 residue of the substrate (the residue located N-terminally of the scissile bond) (Cardozo, 1993; Orlowski and Wilk, 2000). Two additional activities, a “branched-chain amino acid–preferring” activity (BrAAP) and a “small neutral amino acid–preferring” activity (SNAAP), have been described for mammalian proteasomes (Orlowski et al., 1993). Each of these types of activity has been associated with distinct β-type subunits through a number of mutagenesis, inhibition, and X-ray diffraction studies (Hilt et al., 1993; Enenkel et al., 1994; Heinemeyer et al., 1993; Chen and Hochstrasser, 1996; Heinemeyer et al., 1997; Arendt and Hochstrasser, 1997; Dick et al., 1998; McCormack et al., 1998; Groll et al., 1999): The chymotryptic, tryptic, and PGPH activities have been assigned to the subunits β5, β2, and β1, respectively. BrAAP activity resides in both β5 and β1, and SNAAP activity correlates with β2. The barrel-shaped architecture of the 20S proteasome allows substrate proteins to be degraded in a processive manner, i.e., without the release of degradation intermediates (Akopian et al., 1997). It is noteworthy that the cleavage sites found when longer peptides or proteins are used as substrates do not reﬂect the aforementioned speciﬁcities (Wenzel et al., 1994; Ehring, et al., 1996). The Thermoplasma proteasome, for example, when assayed with short ﬂuorogenic peptides displays solely a “chymotryptic” activity, but chymotryptic cleavage sites are found only rarely in degradation products from polypeptides (Wenzel et al., 1994). This suggests that residues beyond P1 (P2, P3, P4, . . .) may determine the site of cleavage [for review see Orlowski and Wilk (2000)]. Moreover, the association with regulatory complexes has been reported to modulate the nature of the degradation products of the 20S proteasome (Kisselev et al., 1999b; Emmerich et al., 2000). In any case, the complex rules are by no means fully understood, and proteins are cleaved in an apparently nonspeciﬁc manner (Dolenc et al., 1998).

200

PETER ZWICKL ET AL.

Nevertheless, the peptide products generated by the 20S proteasome fall into a relatively narrow size range of 6–10 amino acid residues. This observation led to the proposal that proteasomes may possess an intrinsic molecular ruler. At the time it was considered that the distance between neighboring active sites could provide the mechanistic basis for such a ruler (Wenzel et al., 1994). Indeed, the crystal structure of the Thermoplasma proteasome revealed a distance of 2.8 nm between neighboring active sites, which corresponds to a hepta- or octapeptide in an extended conformation, and seemed to provide strong evidence in support of the molecular ruler hypothesis (L¨owe et al., 1995). On the other hand, more comprehensive recent analyses of product lengths, while in agreement with an average length of eight residues (±1 residue), showed larger size variations, which are difﬁcult to reconcile with a purely geometrybased ruler which should yield products more focused in length (Kisselev et al., 1998; Nussbaum et al., 1998; Kisselev et al., 1999b). Moreover, a reduction in the number of active sites to four or two in mutant yeast proteasomes had little effect on the size of the peptides that were generated (Dick et al., 1998). It is therefore unlikely that the distance between active sites is a major determinant of the product size. Studies with synthetic peptides varying in length but displaying the same (repetitive) pattern of cleavage sites indicated that below a certain threshold in length (<12–14 residues) degradation is decelerated, possibly because the products have a higher probability of exiting the proteolytic nanocompartment (Dolenc et al., 1998). Although they might reenter and be degraded further, this appears to be a slow and inefﬁcient process, and therefore products smaller than 12–14 residues accumulate. It should be noted here that a diffusion-controlled mechanism was considered as another option when the molecular ruler hypothesis was originally put forward (Wenzel et al., 1994). Recently, a cyclical “bite–chew” mechanism for protein breakdown was proposed for the eukaryotic proteasome (Kisselev et al., 1999a). The model suggested that certain active sites allosterically regulate one another. Substrates of the chymotrypsin-like site stimulate the PGPH activity 5- to 35-fold and additionally activate the second chymotrypsin-like site. Moreover, substrates of the PGPH sites were reported to inhibit the chymotrypsin-like activity. These allosteric effects were proposed to be indicative of an ordered, cyclical mechanism for protein degradation. The model was challenged when it was found that Ritonavir, an inhibitor of human immunodeﬁciency virus-1 protease, inhibits the chymotrypsinlike activity and stimulates the trypsin-like activity of the proteasome (Schmidtke et al., 1999). Moreover, it was shown that the mutual regulation of the chymotrypsin-like and PGPH activities by their substrates,

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

201

as suggested by the bite–chew model, was not affected by selective inhibitors of the respective active sites (Schmidtke et al., 2000). The data of Schmidtke and colleagues suggest that proteasome substrates can bind not only to an active but also to a noncatalytic site, leading to the proposal of a kinetic two-site modiﬁer model (Schmidtke et al., 2000). More recently, it was shown that a selective inhibition of the PGPH activity by α ′ , β ′ -epoxyketone inhibitor did not inhibit protein degradation and the inhibition of the chymotryptic activity by PGPH substrates did not require the binding to the PGPH site or hydrolysis of the PGPH substrate. These ﬁndings support the existence of an intersubunit regulation through binding to noncatalytic site(s) (Myung et al., 2001). The immune system has taken advantage of the fact that proteasomes generate peptides with an average length of 8–12 residues. However, in the cellular environment the peptide products of the proteasome are not stable and the vast majority undergo further cleavage by peptidases for completion of the catabolic conversion of proteins into free amino acids (Rock and Goldberg, 1999). Nevertheless, some peptides escape further degradation and are presented by MHC class I molecules on the cell surface, eliciting an immune response. It has been shown that leucine aminopeptidase, which is upregulated by γ -interferon, assists in antigen production by N-terminal trimming of peptides generated by the proteasome (Beninga et al., 1998). On the other hand, thimet oligopeptidase, which is downregulated by γ -interfereon, mediates the destruction of antigenic peptides. Thus, γ -interferon appears to increase the supply of peptides by stimulating their generation and decreasing their destruction (York et al., 1999). Recently, two cytosolic proteases, puromycinsensitive aminopeptidase and bleomycin hydrolase, have been shown to remove N-terminal amino acids from the vesicular stomatitis virus nucleoprotein; in conjunction with proteasomal cleavage at the C terminus, the immunogenic epitope is generated (Stoltze et al., 2000). Obviously, the proteasome system alone is not sufﬁcient to generate the necessary supply of MHC class I antigens (Stoltze et al., 2000). Apart from the immune system, there is little use for the degradation products released by the proteasome, and several peptidases exist which complete their conversion into free amino acids. One of them is tricorn peptidase, originally found in Thermoplasma cells, where it exists in the form of a giant icosahedral complex of approximately 15 MDa (Tamura et al., 1996; Walz et al., 1997). It has been shown in in vitro studies that tricorn in cooperation with an array of aminopeptidases converts oligopeptides of 8–12 residues efﬁciently into free amino acids (Tamura et al., 1998). Tricorn peptidase has been found in a number of archaeal and bacterial species, but it is not ubiquitous; functional homologs are

202

PETER ZWICKL ET AL.

found in other species (Tamura et al., 2001). In eukaryotes inhibition of the 20S proteasome upregulates tripeptidylpeptidase II, another huge peptidase complex which is also regarded as a functional homolog of tricorn peptidase (Geier et al., 1999; Wang et al., 2000). III. ACTIVATORS OF THE 20S PROTEASOME A. The Archaeal and Bacterial AAA ATPases The sequencing of the genome of the methanogenic archaeon Methanococcus jannaschii revealed the existence of a gene, S4, with a high sequence similarity to the ATPases of the eukaryotic 19S regulator (Bult et al., 1996). The deduced 50-kDa protein has an N-terminal coiled-coil, a hallmark of proteasomal AAA ATPases, and a C-terminal AAA domain. The protein was expressed in E. coli and puriﬁed as a 650-kDa complex with nucleotidase activity. When mixed with proteasomes from T. acidophilum, degradation of substrate proteins was stimulated up to 25-fold; hence the complex was named PAN, for proteasome-activating nucleotidase (Zwickl et al., 1999). As expected, the Methanococcus PAN also stimulates the endogenous proteasome from Methanococcus, but it remains notoriously difﬁcult to isolate a stable protease–ATPase complex formed by both molecules (Wilson et al., 2000). Methanococcus PAN was recently shown to recognize ssrA-tagged GFP and mediate its energy-dependent unfolding and subsequent translocation into the 20S proteasome for degradation (Benaroudj and Goldberg, 2000). The ssrA tag is an 11-residue, hydrophobic sequence which is added to the C terminus of incompletely translated proteins (Keiler et al., 1996), thus targeting them for degradation by ATP-dependent proteases (Karzai et al., 2000). Interestingly, the ssrA tag is found in many bacteria but not in archaea (Keiler et al., 2000). This sugests that the hydrophobic nature of the peptide, which was shown to be essential for recognition by the ClpA and ClpX ATPases in E. coli (Karzai et al., 2000), is sufﬁcient for the recognition by the Methanococcus PAN complex. Homologs of Methanococcus PAN have been found in most but not all archaeal genomes. No PAN homolog exists, for example, in Thermoplasma acidophilum (Ruepp et al., 2000), Thermoplasma volcanium (Kawashima et al., 2000), and the more distantly related Pyrobaculum aerophilum (S. Fitz-Gibbon, personal commun.). Therefore, the role of PAN in activating the 20S proteasome must be assumed by different molecules in these organisms, most likely more divergent members of the AAA ATPase family. The complete sequence of the T. acidophilum genome

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

203

revealed three candidate proteins: namely, VAT, a two-domain AAA ATPase which is closely related to yeast Cdc48 and human p97; and two one-domain AAA ATPases, VAT2 and Lon2 (Ruepp et al., 2000). While the latter proteins have not yet been characterized, a chaperonelike activity was demonstrated for Thermoplasma VAT (Golbik et al., 1999). Still, it remains to be shown that VAT can stimulate the ATPdependent degradation of proteins by the Thermoplasma 20S proteasomes. In this context it is noteworthy that the eukaryotic homologs of VAT (Cdc48, p97) had previously been implicated in the degradation of substrate proteins via the ubiquitin–proteasome pathway (Johnson et al., 1995; Ghislain et al., 1996; Dai et al., 1998; Mayr et al., 1999; Koegl et al., 1999; Hoppe et al., 2000). Unlike the N-terminal coiled-coil found in PAN and the 19S ATPases, both VAT and p97/Cdc48 have an N-terminal substrate binding domain which is built of a double-ψ β-barrel and a novel six-stranded β-clam fold (Coles et al., 1999; Zhang et al., 2000a). A recurring feature of all bacteria possessing genuine 20S proteasomes is the existence of a gene, ARC, encoding a more distant member of the AAA family of ATPases, which is found upstream of the proteasome operons (Nagy et al., 1998; De Mot et al., 1999). The recombinant ARC ATPase from Rhodococcus erythropolis is a complex of two six-membered rings with ATPase activity; like other proteasomal ATPases, it has an N-terminal coiled-coil domain (Wolf et al., 1998). Although a functional interaction between the ARC complex and the Rhodococcus 20S proteasome has not yet been demonstrated, it is likely that it has a role in ATP-dependent protein degradation. B. The 19S Regulatory Complex 1. Cellular Functions of the Regulatory Complex In eukaryotic cells, the 20S proteasome assembles with one or two 19S regulatory complexes (RC) in an ATP-dependent manner to form the 26S proteasome (Hough et al., 1987; Ganoth et al., 1988; Driscoll and Goldberg, 1990; Peters et al., 1994). The 26S holoenzyme is the most downstream element of the ubiquitin–proteasome pathway. It is an abundant complex, both in the nucleus and in the cytoplasm, and studies with proteasome inhibitors indicate that the bulk of cellular proteins, 80–90%, are degraded via this pathway (Rock et al., 1994; Craiu et al., 1997). It is not only used as disposal machinery for misfolded proteins [termed DRiPs, for defective ribosomal products (Yewdell et al., 1996; Schubert et al., 2000)] and proteins damaged by stress or aging, but it also controls the life-span of many short-lived regulatory proteins

204

PETER ZWICKL ET AL.

(Coux et al., 1996; Hershko and Ciechanover, 1998; Ciechanover and Schwartz, 1998; Bonifacino and Weissman, 1998; Kirschner, 1999; Hicke, 1999; Plemper and Wolf, 1999). Interestingly, the 26S proteasome does not only degrade proteins completely, but also activates certain transcription factors by limited proteolytic processing of precursors, e.g., human NFκB and NFκB2 (Palombella et al., 1994; Heusch et al., 1999), Drosophila Ci (Maniatis, 1999), or yeast Spt23 and Mga2 (Hoppe et al., 2000). In addition to its function in ubiquitin-dependent proteolysis, the 26S proteasome was also found to mediate ubiquitin-independent degradation of several proteins, e.g., ODC, c-jun, IκBα, p21Cip1 [see recent overviews by Pickart (1997); Murakami et al. (2000); Verma and Deshaies (2000)]. Moreover, a not yet fully understood nonproteolytic role of proteasomes in nucleotide excision repair has been described recently (Russell et al., 1999a; Jelinsky et al., 2000; Ortolan et al., 2000). 2. Subcomplexes of the 19S Regulator Slightly differing sets of RC subunits have been reported for different organisms (Udvardy, 1993; DeMartino et al., 1994; Dubiel et al., 1995; Tanaka, 1998; Glickman et al., 1998b; H¨olzl et al., 2000), and the abundance of factors that have been described to associate to the RC in a tissue- and development-speciﬁc manner with variable afﬁnities makes it difﬁcult to draw the line between interacting factors, transiently bound subunits, or integral components of the RC. However, as the mass of the Drosophila RC, approximately 890 kDa, determined by scanning-transmission electron microscopy (STEM) measurements is in good agreement with the summed masses of the 18 individual subunits (932 kDa) identiﬁed on two-dimensional gels (H¨olzl et al., 2000), it seems likely that the catalog of integral subunits is now complete. The catalogs of the Drosophila and yeast 19S subunits are identical, except for a single subunit (p37A, see below), which seems to be absent from yeast 26S proteasomes. The 19S regulator of the yeast proteasome can be dissociated into two subcomplexes—the base and the lid—which are located proximally and distally with respect to the 20S core (Fig. 4) (Glickman et al., 1998a). The complex formed by the base and the 20S proteasome is sufﬁcient for the degradation of nonubiquitylated protein substrates, but does not mediate degradation of ubiquitylated substrate proteins. From the location of the base in the 26S complex, it had been inferred to have a role in substrate unfolding, acting as a “reverse chaperone,” and in controlling the gate in the coaxially apposed α-rings of the 20S particle (Lupas et al., 1993). Recently, a chaperone-like activity was demonstrated for the base (Braun et al., 1999; Strickland et al., 2000) and, likewise, for

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

205

FIG. 4. Model of the 26S proteasome. Overall average of 461 particles of negatively stained 26S complexes obtained from Drosophila melanogaster embryos. The 26S proteasome is composed of the barrel-shaped central 20S complex (two α and two β rings) and two 19S complexes, laterally attached. The 19S complexes, which confer speciﬁcity and regulation to protein degradation, are built from base and lid subcomplexes.

the ancestral PAN complex (Benaroudj and Goldberg, 2000). Both have the ability to recognize proteins destined for degradation, to unfold them in an ATP-dependent manner, and to assist in their translocation into the proteolytic core complex. The degradation signals remain to be deﬁned; possibly, some local unfolding with an exposure of hydrophobic segments is sufﬁcient for targeting substrates. As only the 26S holoenzyme, but not the 20S-base complex, degrades ubiquitylated proteins (Glickman et al., 1998a; Thrower et al., 2000), it appears that recognition and binding of ubiquitin-tagged substrates are mediated by the eight subunits of the lid subcomplex. Very little is known about the molecular mechanisms of substrate shuttling from the E3 ubiquitin–ligase complexes, where the ﬁnal step of ubiquitylation takes place, to the lid subcomplex of the 26S proteasome. Free diffusion of ubiquitylated substrates to the 26S proteasomes would be consistent with the rather high binding constants observed in vitro (Baboshina and Haas, 1996). On the other hand, subcomplexes or subunits of the RC, such as the lid or S5a/Rpn10, could serve as substrate carriers. Indeed, free lid complexes exist and can be isolated from human erythrocytes as stable particles (Henke et al., 1999; Kapelari et al., 2000). Recently, it has been reported that E2 ubiquitin-conjugating enzymes (Lommel et al., 2000), E3 ubiquitin–ligases such as Ubr1 and Ufd4 (Xie and Varshavsky, 2000), and E3 ubiquitin–ligase complexes such as SCF and APC (Verma et al., 2000) interact directly with subunits of the RC. Moreover, there has

206

PETER ZWICKL ET AL.

been evidence that molecular chaperones were also involved in protein degradation by the 26S proteasome (Lee et al., 1996; Bercovich et al., 1997). It was found recently that the Hsp70 co-chaperone Chip stimulates ubiquitylation and degradation of certain substrate proteins by the 26S proteasome (Connell et al., 2001; Meacham et al., 2001). Bag-1, another Hsp70-interacting protein, was also shown to stimulate substrate ubiquitylation and to bind directly to the 26S proteasome (L¨uders et al., 2000). Thus two distinct molecular machineries, the molecular chaperones and the 26S proteasome, seem to interact physically and cooperate in cellular protein quality control (Wickner et al., 1999). The molecular understanding of the steps involved in transferring ubiquitylated substrates from the lid to the base, and eventually into the 20S core, is fragmentary at best. One prerequisite for substrates to enter the proteolytic center is the removal of the multiubiquitin chain. In yeast, the deubiquitylating enzyme Doa4 was shown to interact physically and functionally with the proteasome, supporting the model that Doa4 removes ubiquitin from proteolytic substrates en route to the 20S complex (Papa et al., 1999). Mammalian and Drosophila RCs were shown to contain a deubiquitylating subunit with a molecular mass of approximately 37 kDa (Lam et al., 1997; H¨olzl et al., 2000). So far, however, this activity has only been reported to decrease substrate degradation rates, suggesting that it has an editing function, rescuing incompletely ubiquitylated substrates (Lam et al., 1997; Thrower et al., 2000). This raises the question of when exactly a substrate is irreversibly committed to destruction. Obviously, recognition and binding of ubiquitylated proteins to the 26S proteasome alone are not enough to determine their fate. Possibly, there is a step at which the 19S RC undergoes a remodeling or a conformational change that prevents an escape of substrates. 3. The Subunits of the Base The base is composed of the two largest subunits of the 26S proteasome, S1/Rpn2 and S2/Rpn1, and six paralogous AAA ATPases. The ubiquitin-chain binding subunit S5a/Rpn10 was originally assigned to the base, but there is now consensus that it is located at the base–lid interface. The ATPases perform and regulate some of the basic functions of the 26S proteasome (Lupas et al., 1993; Larsen and Finley, 1997). First, they are required for the ATP-dependent assembly of 26S proteasomes from the 19S RC and the 20S core (Armon et al., 1990; DeMartino et al., 1994; Peters et al., 1994; Hoffman and Rechsteiner, 1996; Verma et al., 2000). Second, as attachment of the base complex to the 20S core is sufﬁcient to activate peptidase activity, the ATPases are believed to be involved in the gating of the α-ring channel and thus in controlling

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

207

access to the proteolytic core. Such a function is in agreement with recent crystallographic studies revealing a gated channel formed by the N termini of the α subunits (Groll et al., 2000a; Whitby et al., 2000). Third, as the base complex and its evolutionary ancestor PAN have been shown to have chaperone-like activity in vitro, the ATP-dependent unfolding of substrates can be attributed to the ATPases, acting in a “reverse chaperone” mode (Lupas et al., 1993; Zwickl and Baumeister, 1999). Finally, it is possible that unfolding and translocation are mechanically coupled (a “pushing” mechanism), and therefore substrate translocation is dependent on the ATPase activity. The six ATPases (S4/Rpt2, S6a/Rpt5, S6b/Rpt3, S7/Rpt1, S8/Rpt6, and S10b/Rpt4) contain one approximately 230–amino acid AAA module, which contains the Walker A and B sequences characteristic of all P-loop ATPases (Walker et al., 1982), and the so-called second region of homology (SRH), a conserved stretch of residues common to all members of the AAA family of ATPases (Confalonieri and Duguet, 1995; Beyer, 1997; Swafﬁeld and Purugganan, 1997; Neuwald et al., 1999). Members of this superfamily act in assembly, remodeling, and unfolding of proteins and DNA. The hallmark of all proteasomal AAA ATPases, including PAN and ARC, is an N-terminal coiled-coil region, which has been proposed to mediate the binding of substrate proteins (Wang et al., 1996; Zhang et al., 2000b). Moreover, the coiled-coils appear to promote interactions between individual ATPases, creating special pairs of ATPases (S4/Rpt2–S7/Rpt1, S6/Rpt3–S8/Rpt6, S6′ / Rpt5–S10b/Rpt4) (Richmond et al., 1997). Crosslinking studies in conjunction with immunoprecipitation and Western blotting indicate the following ring-assembly scenario: The aforementioned heterotetrameric complex assembles with the residual ATPase pair, yielding a heterohexameric ATPase ring (Ferrell et al., 2000; Hartmann-Petersen et al., 2001; Rechsteiner, 1998). In mammalian and insect cells, the S6a/Rpt5 and S10b/Rpt4 ATPases constitute, together with the p27 protein, the “modulator,” a 220-kDa protein complex which facilitates the assembly of the 26S proteasome from RCs and the 20S core (DeMartino et al., 1996; Adams et al., 1998b; 1998a; Hastings et al., 1999). A homolog of the p27 protein, called non-ATPase subunit 2 (Nas2), found in yeast, turned out not to be essential (Watanabe et al., 1998); in line with this observation, a modulator complex was not detectable in yeast cells under normal growth conditions (Russell et al., 1999b). All of the 19S ATPase genes are essential, as shown by deletion analysis in ﬁssion and budding yeast (Gordon et al., 1993; Ghislain et al., 1993; Russell et al., 1996). Sitedirected mutations in the Walker A motif of individual yeast ATPase subunits resulted in different phenotypes, indicating that the Rpt ATPases,

208

PETER ZWICKL ET AL.

despite their high sequence similarity, are not functionally redundant (Seeger et al., 1996; Rubin et al., 1998; Fu et al., 1999). The two largest subunits of the base complex, S1/Rpn2 and S2/Rpn1, have signiﬁcant sequence similarities (∼20% identity) and are likely to have a common ancestor (Tsurumi et al., 1996). Both contain leucinerich–like repeats (LRR) at their C termini, similar to repeats found in the BimE subunit of the anaphase-promoting complex (APC) or cyclosome (Lupas et al., 1997a), as well as motifs rich in alternating lysine (K) and glutamate (E) residues (KEKE) (Realini et al., 1994); both motifs are supposed to be involved in protein–protein interactions. S1/Rpn2 was shown to bind to the ATPases S8/Rpt6 and S10b/Rpt4 (Richmond et al., 1997), consistent with its presence in the base complex. In addition, S1/Rpn2 interacts with the lid subunit S14/Rpn12, indicating that it has a linker function. Mutants of the S1 homolog in S. cerevisiae, Rpn2, strongly inhibit degradation of multiple ubiquitylated proteins, causing the accumulation of primarily ubiquitin conjugates but also of free ubiquitin chains (DeMarini et al., 1995; Yokota et al., 1996); this suggests that S1/Rpn2 has a role in the tethering of multiubiquitylated substrates and/or in promoting their deubiquitylation. S2/Rpn1 appears to interact with the ATPases S4/Rpt2 and S7/Rpt1, with S5a/Rpn10 and S5b, as well as with the lid subunits S11/Rpn9, S13/Rpn11, and S14/Rpn12 (Gorbea et al., 2000; Ferrell et al., 2000). Its susceptibility to proteolytic attack suggests that it is at an exposed location (Haracska and Udvardy, 1996). S. pombe cells deleted for the S2 homolog are characterized by slow growth, sensitivity to canavanine, and an accumulation of ubiquitin conjugates (Hampton et al., 1996). S5a/Rpn10, one of the best-studied subunits of the RC, shares sequence similarity with the p44 subunit of the basal transcription factor IIH (Aravind and Ponting, 1998). A conserved hydrophobic stretch [LAL(M)AL] within its C-terminal region has been shown to bind Lys48–linked ubiquitin chains in vitro as efﬁciently as the full-length recombinant subunit or puriﬁed 26S proteasomes (Deveraux et al., 1994; Ferrell et al., 1996; Baboshina and Haas, 1996; Haracska and Udvardy, 1997; Fu et al., 1998b; Young et al., 1998). Although yeast cells deleted for this subunit are viable (van Nocker et al., 1996; Glickman et al., 1998a), its deletion is synthetically lethal with mutations in other RC subunits (Wilkinson et al., 2000). Strains of the plant Physcomitrella patens deleted for the S5a/Rpn10 gene are viable but are under developmental arrest. Paradoxically, the N-terminal region appears to be more critical for the degradation of substrates than the C-terminal ubiquitin-binding region (Fu et al., 1998b; Girod et al., 1999). As the deletion of S5a/Rpn10 facilitates dissociation of the RC into base and lid, it is assumed to be located at the interface; this is consistent

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

209

with observations that it interacts with multiple subunits in the base and in the lid (Ferrell et al., 2000). Interestingly, S5a/Rpn10 is the only subunit of the 19S complex that is also present in signiﬁcant amounts in free monomeric form (Haracska and Udvardy, 1995; van Nocker et al., 1996). 4. The Subunits of the Lid The lid has been isolated as a discrete complex from yeast and from human erythrocytes (Glickman et al., 1998a; Henke et al., 1999); it comprises eight subunits, S3/Rpn3, Rpn5, S9/Rpn6, S10a/Rpn7, S11/Rpn9, S12/Rpn8, S13/Rpn11, and S14/Rpn12 (Table I). Intriguingly, each of the individual subunits has a counterpart with extended sequence similarities in the COP9 signalosome, a multiprotein complex involved in signal transduction (Seeger et al., 1998). Six of the subunits contain a PCI (proteasome, COP9, initiation factor 3) and two subunits, an MPN domain (Mpr 1 and Pad1 in the N terminus). It is likely, therefore, that both complexes have evolved from a common ancestor. Moreover, ﬁve subunits of the eIF3 complex and one subunit of the basal transcription factor IIH are closely related to subunits of the lid complex (Hofmann and Bucher, 1998; Aravind and Ponting, 1998). In addition to the RC subunits detected in all eukaryotic organisms investigated so far, a few subunits seem to be species-speciﬁc. A member of the ubiquitin carboxyl-terminal hydrolases (UCH) (Wilkinson, 2000) has been found in Drosophila (p37A) (H¨olzl et al., 2000), S. pombe (Uch2) (Li et al., 2000), and human (UCH37) (Li et al., 2001), but not in S. cerevisiae 26S proteasomes. The Drosophila p37A subunit was mapped to the base–lid interface in close vicinity of the putative location of S5a/Rpn10 (H¨olzl et al., 2000). In a two-hybrid analysis the human Uch37 was found to interact with the lid subunit S14/Rpn12 and another protein called Uch37 interacting protein 1 (Uip 1), which has so far not been identiﬁed as a component of the 26S proteasome (Li et al., 2001). Rpn4/Son1 has thus far been found only in S. cerevisiae 26S proteasome preparations (Fujimuro et al., 1998) and has previously been identiﬁed as a component (Ufd5) of the ubiquitin fusion degradation pathway (UFD) (Johnson et al., 1995). More recently, Rpn4 was suggested to be involved in the activation of transcription of proteasomal genes (Mannhaupt et al., 1999; Ng et al., 2000), as well as base excision and nucleotide excision repair genes ( Jelinsky et al., 2000), supporting a proposed function of 26S proteasomes in regulation of nucleotide excision repair (Russell et al., 1999a; Lommel et al., 2000). Rpn13 was identiﬁed as a new subunit of the yeast 26S proteasome by mass spectrometric analysis of puriﬁed 19S cap complexes, but no homolog has been detected in other species (Verma et al., 2000). The

TABLE I Subunits of the 19S Regulatory Complex a 19S subunits (human/ 19S Mass yeast) subcomplex (kDa) S3 Rpn3 p55 Rpn5 S9 Rpn6 S10a Rpn7 S11 Rpn9 S12 Rpn8 S13 Rpn11 S14 Rpn12

Lid

S4 Rpt2 S6ab Rpt5 S6b Rpt3 S7 Rpt1 S8 Rpt6 S10bb Rpt4 S1 Rpn2 S2 Rpn1

Base

Lid Lid Lid Lid Lid Lid Lid

Base Base Base Base Base Base Base

61 60 53 52 47 50 46 49 43 46 37 38 35 34 30 32 52 49 48 48 48 48 47 52 46 45 44 49 106 104 99 109

S5a Rpn10

Base/lid interface

41 30

UCH37 Uch2pc S5b — — Rpn4

Base/lid interface

37

a Abbreviations:

Sequence domain/ motifd

Related subunits in bacterial/ archaeal complexes

Related subunits in other eukaryotic complexes COP9–CSN3

PCI COP9–CSN4 PCI COP9–CSN2, elF3–p48 PCI, 9 dileucine repeats COP9–CSN1 PCI, KEKE COP9–CSN7 PCI COP9–CSN6, elF3–p47 MPN, KEKE COP9–CSN5, elF3–p40 MPN, Cys box of UBPs COP9–CSN8 PCI ARC/PAN AAA ARC/PAN AAA ARC/PAN AAA ARC/PAN AAA ARC/PAN AAA ARC/PAN AAA Cyclosome/APC–BimE LRR-like motif; KEKE Cyclosome/APC-BimE LRR-like motif; KEKE TFIIH-p44 N-terminal conserved domain I; C-terminal in vitro Ub-binding site 4 UCH blocks

50 — — 60

9 dileucine repeats C2H2-ﬁnger, 2 acidic domains

APC: anaphase-promoting complex; ARC: AAA ATPase-forming ring-shaped complexes; eIF3: eukaryotic initiation factor 3; PAN: proteasome-activating nucleotidase; TFIIH: basal transcription factor IIH. b These ATPases form the modulator. c Only found in S. pombe without homolog in S. cerevisiae. d Sequence motifs: : ANK (ANKyrin-like) repeats; : AAA (ATPases associated with a variety of cellular activities) domain or CAD (conserved ATPase domain) with Walker A and B motifs; KEKE (motif rich in alternating lysine (K) and glutamate (E) residues); cys box of UBPs : LRR (leucine-rich repeat)-like motif; : MPN (found in (ubiquitin-processing enzymes); : PCI (for proteasome, COP9 and initiation factor 3) Mpr1 and Pad1 in the N terminus) motif; motif; : UCH (ubiquitin C-terminal hydrolases) block.

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

211

subunit S5b has only been found in 26S proteasomes from human erythrocytes and is characterized by nine dileucine repeats, which have been implicated in trafﬁcking of a variety of transmembrane proteins (Deveraux et al., 1995; Dubiel et al., 1995). The subunit S5b is distantly related to p55 (Gorbea and Rechsteiner, 2000), a subunit of bovine 26S proteasomes. 5. Structural Features On electron micrographs, 26S proteasomes appear as elongated, dumbbell-shaped particles. Samples from a variety of different organisms all show a mixture of 20S core particles “capped” with either one or two RC(s); the single-capped particles have a length of 30 nm, and the double-capped, 45 nm (Peters et al., 1993; Yoshimura et al., 1993; Fujinami et al., 1994). Averages obtained from negatively stained preparations reveal a characteristic “dragon head” motif (Rechsteiner, 1998) with the RCs facing in opposite directions in the double-capped particles (Fig. 4). Thus they appear to reﬂect the underlying C2 symmetry of the core particle; a more rigorous analysis of interimage variations has shown, however, a small but signiﬁcant deviation from exact C2 symmetry. Moreover, the 19S caps appear not to be in a ﬁxed position with respect to the 20S core particles, but to undergo a peculiar up-and-down (“wagging”) movement (Walz et al., 1998); so far, the functional relevance of this movement is not clear. Furthermore, it cannot be ruled out that the observed wagging describes the real movements inadequately; adsorption of the 26S complex to the carbon ﬁlm prior to negative staining may severely restrain movements of the 20S and 19S subcomplexes with respect to each other. In fact, a large variety of states is observed, with unsupported 26S complexes embedded in vitreous ice. Among them are double-capped complexes that appear nearly mirror-symmetric, which is indicative of a rotary movement between the pseudo-sevenfold α ring of the 20S particle and the pseudo-sixfold ATPase ring in the base part of the 19S RC (Kapelari et al., in preparation). Such a symmetry mismatch has analogies in other rotating molecular machines. For the bacterial ClpAP system it has been shown that small rotational increments of 8.6 degrees are sufﬁcient to bring sixand sevenfold rings into (pseudo-)equivalent positions (Beuron et al., 1998). Whether or not a rotary movement has a role in the unfolding and translocation of target proteins is currently a matter of speculation. It is noteworthy in this context that the interaction between the 20S and 19S subcomplexes of the proteasome is a rather weak one and that the number of identiﬁed speciﬁc pairwise interactions between α subunits and ATPase subunits is remarkably small (Gerlinger et al., 1997; Zhang et al., 2000b; Satoh et al., 2001; Hartmann-Petersen et al., 2001).

212

PETER ZWICKL ET AL.

C. The PA28 Activator The PA28 activator or 11S regulator is an ATP-independent activator of the 20S proteasome, which greatly stimulates the hydrolysis of small peptides, but not of denatured or ubiquitylated proteins (Dubiel et al., 1992; Ma et al., 1992). PA28 is a predominantly cytosolic 200-kDa complex formed by two different but related 28-kDa subunits, PA28α and PA28β, which are believed to form a heteroheptamer (Zhang et al., 1999). Mice disrupted for the PA28β gene lack both the PA28β and the PA28α proteins, indicating that PA28 functions as heterooligomer in vivo (Preckel et al., 1999). The PA28 complex can bind to both ends of the 20S proteasome, as shown by electron microscopy (Gray et al., 1994; Koster et al., 1995). Immunoprecipitation experiments indicate that 20S proteasomes, PA28, and the 19S regulator can form “hybrid proteasomes” (Hendil et al., 1998), which are able to degrade ODC in an ATP-dependent, but ubiquitin-independent reaction (Tanahashi et al., 2000). To our knowledge the existence of hybrid proteasomes has not been proved by direct visualization. When expressed in E. coli, in the absence of PA28 β subunits, PA28 α subunits assemble into a heptameric particle whose crystal structure has been determined (Knowlton et al., 1997). The PA28 heptamer is a cone-shaped particle formed by bundles of α-helices; it is traversed by a central channel that has a 3-nm opening on the side proximal and a 2-nm opening on the side distal to the proteasome (Knowlton et al., 1997). In vitro studies have shown that PA28 can enhance the generation of antigenic peptides by inducing dual substrate cleavages by the 20S proteasome, which suggests an in vivo role for PA28 in antigen processing (Dick et al., 1996; Groettrup et al., 1996). Consistent with these data, genes for PA28α and PA28β are found only in organisms with adaptive immune systems (Tanaka and Kasahara, 1998). The cytokine γ -interferon induces expression of both PA28 subunits as well as both subunits of the transporter associated with antigen processing (TAP1 and TAP2) and the immunoproteasome subunits β1i, β2i, and β5i (Fruh ¨ and Yang, 1999). The PA28γ protein, also called Ki antigen, is a nuclear protein which shares high sequence similarity with the PA28α and PA28β proteins, but is not present in the predominantly cytosolic PA28 complex (Rechsteiner et al., 2000). PA28γ is also present in organisms lacking an adaptive immune system, such as Caenorhabditis elegans and Drosophila, and is considered to be an evolutionary precursor of the PA28α and PA28β proteins (Kandil et al., 1997; Murray et al., 2000; Masson et al., 2001). In vitro experiments have shown that PA28γ can oligomerize and activate

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

213

the proteasome (Zhang et al., 1998a; 1998b) and is induced in higher eukaryotes by γ -interferon, although to a much lesser extent than PA28α and PA28β (Ahn et al., 1995; Jiang and Monaco, 1997). Mice disrupted for PA28γ show growth retardation, but have no obvious abnormalities in their immune system, which suggests that PA28γ has a function different from that of the PA28 complex (Murata et al., 1999). PA26, a more distantly related protein of the PA28 family, was identiﬁed in Trypanosoma brucei (To and Wang, 1997; Yao et al., 1999a). PA26 forms a 170-kDa heptameric complex, which activates the peptidolytic activity of 20S proteasomes isolated from Trypanosoma, rat, and yeast (Yao et al., 1999a; Whitby et al., 2000). The crystal structure of a complex formed by Trypanosoma PA26 and yeast 20S proteasomes shows that complex formation induces conformational changes in the α subunits, which result in an opening of the gate to the interior of the 20S proteasome (Whitby et al., 2000). IV. CONCLUSIONS Structure, assembly, and enzymatic mechanism of the 20S complex have been elucidated, but the functional organization of the 26S proteasome is understood only dimly at present. The constituent subunits of the 19S complex have been identiﬁed; however, speciﬁc functions have only been assigned to a few. The molecular details of the distinct steps of substrate recognition, unfolding, and translocation en route to the 20S core complex where degradation takes place have remained elusive so far. More evidence for interactions between chaperones and the 26S proteasome has emerged recently, lending support to the notion that protein folding and protein degradation “machines” cooperate in the quality control of cellular proteins. REFERENCES Adams, G. M., Crotchett, B., Slaughter, C. A., DeMartino, G. N., and Gogol, E. P. (1998a). Biochemistry 37, 12927–12932. Adams, J., Behnke, M., Chen, S. W., Cruickshank, A. A., Dick, L. R., Grenier, L., Klunder, J. M., Ma, Y. T., Plamondon, L., and Stein, R. L. (1998b). Bioorg. Med. Chem. Lett. 8, 333–338. Ahn, J. Y., Tanahashi, N., Akiyama, K., Hisamatsu, H., Noda, C., Tanaka, K., Chung, C. H., Shibmara, N., Willy, P. J., Mott, J. D., et al. (1995). FEBS Lett. 366, 37–42. Akopian, T. N., Kisselev, A. F., and Goldberg, A. L. (1997). J. Biol. Chem. 272, 1791–1798. Aravind, L., and Ponting, C. P. (1998). Protein Sci. 7, 1250–1254. Arendt, C. S., and Hochstrasser, M. (1997). Proc. Natl. Acad. Sci. USA 94, 7156–7161. Arendt, C. S., and Hochstrasser, M. (1999). EMBO J. 18, 3575–3585.

214

PETER ZWICKL ET AL.

Armon, T., Ganoth, D., and Hershko, A. (1990). J. Biol. Chem. 265, 20723–20726. Baboshina, O. V., and Haas, A. L. (1996). J. Biol. Chem. 271, 2823–2831. Baker, D., Shiau, K., and Agard, D. A. (1993). Curr. Opin. Cell. Biol. 5, 966–970. Bauer, M. W., Bauer, S. H., and Kelly, R. M. (1997). Appl. Eviron. Microbiol. 63, 1160– 1164. Baumeister, W., Walz, J., Zuhl, ¨ F., and Seemuller, ¨ E. (1998). Cell 92, 367–380. Benaroudj, N., and Goldberg, A. L. (2000). Nat. Cell Biol. 2, 833–839. Beninga, J., Rock, K. L., and Goldberg, A. L. (1998). J. Biol. Chem. 273, 18734–18742. Bercovich, B., Stancovski, I., Mayer, A., Blumenfeld, N., Laszlo, A., Schwartz, A. L., and Ciechanover, A. (1997). J. Biol. Chem. 272, 9002–9010. Beuron, F., Maurizi, M. R., Belnap, D. M., Kocsis, E., Booy, F. P., Kessel, M., and Steven, A. C. (1998). J. Struct. Biol. 124, 179–188. Beyer, A. (1997). Protein Sci. 6, 2043–2058. Bochtler, M., Ditzel, L., Groll, M., and Huber, R. (1997). Proc. Natl. Acad. Sci. USA 94, 6070–6074. Bochtler, M., Hartmann, C., Song, H. K., Bourenkov, G. P., Bartunik, H. D., and Huber, R. (2000a). Nature 403, 800–805. Bochtler, M., Hartmann, C., Song, H. K., Ramachandran, R., and Huber, R. (2000b). Nature 408, 668. Bonifacino, J. S., and Weissman, A. M. (1998). Annu. Rev. Cell Dev. Biol. 14, 19–57. Bouzat, J. L., McNeil, L. K., Robertson, H. M., Solter, L. F., Nixon, J. E., Beever, J. E., Gaskins, H. R., Olsen, G., Subramaniam, S., Sogin, M. L., and Lewin, H. A. (2000). J. Molec. Evol. 51, 532–543. Brannigan, J. A., Dodson, G., Duggleby, H. J., Moody, P. C., Smith, J. L., Tomchick, D. R., and Murzin, A. G. (1995). Nature 378, 416–419. Braun, B. C., Glickman, M., Kraft, R., Dahlmann, B., Kloetzel, P. M., Finley, D., and Schmidt, M. (1999). Nat. Cell Biol. 1, 221–226. Bult, C. J., White, O., Olsen, G. J., Zhou, L., Fleischmann, R. D., Sutton, G. G.,, Blake, J. A., FitzGerald, L. M., Clayton, R. A., Gocayne, J. D., Kerlavage, A. R., Dougherty, B. A., Tomb, J. F., Adams, M. D., Reich, C. I., Overbeek, R., Kirkness, E. F.,, Weinstock, K. G., Merrick, J. M., Glodek, A., Scott, J. L., Geoghagen, N. S. M., and Venter, J. C. (1996). Science 273, 1058–1073. Burri, L., Hockendorf, J., Boehm, U., Klamp, T., Dohmen, R. J., and Levy, F. (2000). Proc. Natl. Acad. Sci. USA 97, 10348–10353. Cardozo, C. (1993). Enzyme Protein 47, 296–305. Cerundolo, V., Kelly, A., Elliott, T., Trowsdale, J., and Townsend, A. (1995). Eur. J. Immunol. 25, 554–562. Chen, P., and Hochstrasser, M. (1996). Cell 86, 961–972. Ciechanover, A., and Schwartz, A. L. (1998). Proc. Natl. Acad. Sci. USA 95, 2727–2730. Coles, M., Diercks, T., Liermann, J., Gr¨oger, A., Rockel, B., Baumeister, W., Koretke, K. K., Lupas, A., Peters, J., and Kessler, H. (1999). Curr. Biol. 9, 1158–1168. Confalonieri, F., and Duguest, M. (1995). Bioessays 17, 639–650. Connell, P., Ballinger, C. A., Jiang, J. H., Wu, Y. X., Thompson, L. J., H¨ohfeld, J., and Patterson, C. (2001). Nat. Cell Biol. 3, 93–96. Coux, O., Tanaka, K., and Goldberg, A. L. (1996). Annu. Rev. Biochem. 65, 801–847. Craiu, A., Gaczynska, M., Akopian, T., Gramm, C. F., Fenteany, G., Goldberg, A. L., and Rock, K. L. (1997). J. Biol. Chem. 272, 13437–13445. Dai, R. M., Chen, E. Y., Longo, D. L., Gorbea, C. M., and Li, C. C. H. (1998). J. Biol. Chem. 273, 3562–3573. De Mot, R., Nagy, I., Walz, J., and Baumeister, W. (1999). Trends Microbiol. 7, 88–92.

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

215

DeMarini, D. J., Papa, F. R., Swaminathan, S., Ursic, D., Rasmussen, T. P., Culbertson, M. R., and Hochstrasser, M. (1995). Mol. Cell. Biol. 15, 6311–6321. DeMartino, G. N., Moomaw, C. R., Zagnitko, O. P., Proske, R. J., Chu-Ping, M., Afendis, S. J., Swafﬁeld, J. C., and Slaughter, C. A. (1994). J. Biol. Chem. 269, 20878–20884. DeMartino, G. N., Proske, R. J., Moomaw, C. R., Strong, A. A., Song, X., Hisamatsu, H., Tanaka, K., and Slaughter, C. A. (1996). J. Biol. Chem. 271, 3112–3118. DeMartino, G. N., and Slaughter, C. A. (1999). J. Biol. Chem. 274, 22123–22126. Deveraux, Q., Jensen, C., and Rechsteiner, M. (1995). J. Biol. Chem. 270, 23726–23729. Deveraux, Q., Ustrell, V., Pickart, C., and Rechsteiner, M. (1994). J. Biol. Chem. 269, 7059– 7061. Dick, T. P., Nussbaum, A. K., Deeg, M., Heinemeyer, W., Groll, M., Schirle, M., Keilholz, W., Stevanovic, S., Wolf, D. H., Huber, R., Rammensee, H. G., and Schild, H. (1998). J. Biol. Chem. 273, 25637–25646. Dick, T. P., Ruppert, T., Groettrup, M., Kloetzel, P. M., Kuehn, L., Koszinowski, U. H., Stevanovic, S., Schild, H., and Rammensee, H. G. (1996). Cell 86, 253–262. Ditzel, L., Huber, R., Mann, K., Heinemeyer, W., Wolf, D. H., and Groll, M. (1998). J. Mol. Biol. 279, 1187–1191. Dodson, G., and Wlodawer, A. (1998). Trends Biochem. Sci. 23, 347–352. Dolenc, I., Seemuller, ¨ E., and Baumeister, W. (1998). FEBS Lett. 434, 357–361. Driscoll, J., and Goldberg, A. L. (1990). J. Biol. Chem. 265, 4789–4792. Dubiel, W., Ferrell, K., and Rechsteiner, M. (1995). Mol. Biol. Rep. 21, 27–34. Dubiel, W., Pratt, G., Ferrell, K., and Rechsteiner, M. (1992). J. Biol. Chem. 267, 22369– 22377. Ehring, B., Meyer, T. H., Eckerskorn, C., Lottspeich, F., and Tampe, R. (1996). Eur. J. Biochem. 235, 404–415. Emmerich, N. P. N., Nussbaum, A. K., Stevanovic, S., Priemer, M., Toes, R. E. M., Rammensee, H. G., and Schild, H. (2000). J. Biol. Chem. 275, 21140–21148. Enenkel, C., Lehmann, H., Kipper, J., Guckel, R., Hilt, W., and Wolf, D. H. (1994). FEBS Lett. 341, 193–196. Fenteany, G., Standaert, R. F., Lane, W. S., Choi, S., Corey, E. J., and Schreiber, S. L. (1995). Science 268, 726–731. Ferrell, K., Deveraux, Q., van Nocker, S., and Rechsteiner, M. (1996). FEBS Lett. 381, 143–148. Ferrell, K., Wilkinson, C. R. M., Dubiel, W., and Gordon, C. (2000). Trends Biochem. Sci. 25, 83–88. Fruh, ¨ K., and Yang, Y. (1999). Curr. Opin. Immunol. 11, 76–81. Fu, H. Y., Doelling, J. H., Arendt, C. S., Hochstrasser, M., and Vierstra, R. D. (1998a). Genetics 149, 677–692. Fu, H. Y., Doelling, J. H., Rubin, D. M., and Vierstra, R. D. (1999). Plant J. 18, 529–539. Fu, H. Y., Sadis, S., Rubin, D. M., Glickman, M., van Nocker, S., Finley, D., and Vierstra, R. D. (1998b). J. Biol. Chem. 273, 1970–1981. Fujimuro, M., Tanaka, K., Yokosawa, H., and Tohe, A. (1998). FEBS Lett. 423, 149–154. Fujinami, K., Tanahashi, N., Tanaka, K., Ichihara, A., Cejka, Z., Baumeister, W., Miyawaki, M., Sato, T., and Nakagawa, H. (1994). J. Biol. Chem. 269, 25905–25910. Ganoth, D., Leshinsky, E., Eytan, E., and Hershko, A. (1988). J. Biol. Chem. 263, 12412– 12419. Geier, E., Pfeifer, G., Wilm, M., Lucchiari-Hartz, M., Baumeister, W., Eichmann, K., and Niedermann, G. (1999). Science 283, 978–981. Gerards, W. L., Enzlin, J., Haner, M., Hendriks, I. L., Aebi, U., Bloemendal, H., and Boelens, W. (1997). J. Biol. Chem. 272, 10080–10086.

216

PETER ZWICKL ET AL.

Gerlinger, U. M., Guckel, ¨ R., Hoffmann, M., Wolf, D. H., and Hilt, W. (1997). Mol. Biol. Cell 8, 2487–2499. Ghislain, M., Dohmen, R. J., Levy, F., and Varshavsky, A. (1996). EMBO J. 15, 4884–4899. Ghislain, M., Udvardy, A., and Mann, C. (1993). Nature 366, 358–362. Girod, P. A., Fu, H. Y., Zryd, J. P., and Vierstra, R. D. (1999). Plant Cell 11, 1457–1471. Glickman, M. H., Rubin, D. M., Coux, O., Wefes, I., Pfeifer, G., Cjeka, Z., Baumeister, W., Fried, V. A., and Finley, D. (1998a). Cell 94, 615–623. Glickman, M. H., Rubin, D. M., Fried, V. A., and Finley, D. (1998b). Mol. Cell. Biol. 18, 3149–3162. Golbik, R., Lupas, A. N., Koretke, K. K., Baumeister, W., and Peters, J. (1999). Biol. Chem. 380, 1049–1062. Gorbea, C., and Rechsteiner, M. (2000). In Proteasomes: The World of Regulatory Proteolysis (W. Hilt and D. H. Wolf, eds.), pp. 91–128. Eurekah.com/Landes Bioscience, Georgetown. Gorbea, C., Taillandier, D., and Rechsteiner, M. (2000). J. Biol. Chem. 275, 875–882. Gordon, C., McGurk, G., Dillon, P., Rosen, C., and Hastie, N. D. (1993). Nature 366, 355–357. Gray, C. W., Slaughter, C. A., and Demartino, G. N. (1994). J. Mol. Biol. 236, 7–15. Groettrup, M., Soza, A., Eggers, M., Kuehn, L., Dick, T. P., Schild, H., Rammensee, H. G., Koszinowski, U. H., and Kloetzel, P. M. (1996). Nature 381, 166–168. Groettrup, M., Standera, S., Stohwasser, R., and Kloetzel, P. M. (1997). Proc. Natl. Acad. Sci. USA 94, 8970–8975. Groll, M., Bajorek, M., K¨ohler, A., Moroder, L., Rubin, D. M., Huber, R., Glickman, M. H., and Finley, D. (2000a). Nat. Struct. Biol. 7, 1062–1067. Groll, M., Ditzel, L., L¨owe, J., Stock, D., Bochtler, M., Bartunik, H. D., and Huber, R. (1997). Nature 386, 463–471. Groll, M., Heinemeyer, W., J¨ager, S., Ullrich, T., Bochtler, M., Wolf, D. H., and Huber, R. (1999). Proc. Natl. Acad. Sci. USA 96, 10976–10983. Groll, M., Kim, K. B., Kairies, N., Huber, R., and Crews, C. M. (2000b). J. Am. Chem. Soc. 122, 1237–1238. Grziwa, A., Maack, S., P¨uhler, G., Wiegand, G., Baumeister, W., and Jaenicke, R. (1994). Eur. J. Biochem. 223, 1061–1067. Hampton, R. Y., Gardner, R. G., and Rine, J. (1996). Mol. Biol. Cell 7, 2029–2044. Haracska, L., and Udvardy, A. (1995). Eur. J. Biochem. 231, 720–725. Haracska, L., and Udvardy, A. (1996). Biochem. Biophys. Res. Commun. 220, 166–170. Haracska, L., and Udvardy, A. (1997). FEBS Lett. 412, 331–336. Hartmann-Petersen, R., Tanaka, K., and Hendil, K. B. (2001). Arch. Biochem. Biophys. 386, 89–94. Hastings, R. A., Eyheralde, I., Dawson, S. P., Walker, G., Reynolds, S. E., Billett, M. A., and Mayer, R. J. (1999). J. Biol. Chem. 274, 25691–25700. Heinemeyer, W. (2000). In Proteasomes: The World of Regulatory Proteolysis (W. Hilt and D. H. Wolf, eds.), pp. 48–70. Eurekah.com/Landes Bioscience, Georgetown. Heinemeyer, W., Fischer, M., Krimmer, T., Stachon, U., and Wolf, D. H. (1997). J. Biol. Chem. 272, 25200–25209. Heinemeyer, W., Gruhler, A., M¨ohrle, V., Mahe, Y., and Wolf, D. H. (1993). J. Biol. Chem. 268, 5115–5120. Hendil, K. B., Khan, S., and Tanaka, K. (1998). Biochem. J. 332, 749–754. Henke, W., Ferrell, K., Bech-Otschir, D., Seeger, M., Schade, R., Jungblut, P., Naumann, M., and Dubiel, W. (1999). Mol. Biol. Rep. 26, 29–34. Hershko, A., and Ciechanover, A. (1998). Annu. Rev. Biochem. 67, 425–479.

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

217

Heusch, M., Lin, L., Geleziunas, R., and Greene, W. C. (1999). Oncogene 18, 6201–6208. Hicke, L. (1999). Trends Cell Biol. 9, 107–112. Hilt, W., Enenkel, C., Gruhler, A., Singer, T., and Wolf, D. H. (1993). J. Biol. Chem. 268, 3479–3486. Hoffman, L., and Rechsteiner, M. (1996). J. Biol. Chem. 271, 32538–32545. Hofmann, K., and Bucher, P. (1998). Trends Biochem. Sci. 23, 204–205. H¨olzl, H., Kapelari, B., Kellermann, J., Seemuller, ¨ E., Sumegi, M., Udvardy, A., Medalia, O., Sperling, J., Muller, ¨ S. A., Engel, A., and Baumeister, W. (2000). J. Cell Biol. 150, 119–129. Hoppe, T., Matuschewski, K., Rape, M., Schlenker, S., Ulrich, H. D., and Jentsch, S. (2000). Cell 102, 577–586. Horwich, A. L., WeberBan, E. U., and Finley, D. (1999). Proc. Natl. Acad. Sci. USA 96, 11033–11040. Hough, R., Pratt, G., and Rechsteiner, M. (1987). J. Biol. Chem. 262, 8303–8313. Hughes, A. L. (1997). Immunogenetics 46, 82–92. Hughes, A. L., and Yeager, M. (1997). Bioessays 19, 777–786. Ishikawa, T., Maurizi, M. R., Belnap, D., and Steven, A. C. (2000). Nature 408, 667–668. J¨ager, S., Groll, M., Huber, R., Wolf, D. H., and Heinemeyer, W. (1999). J. Mol. Biol. 291, 997–1013. Jelinsky, S. A., Estep, P., Church, G. M., and Samson, L. D. (2000). Mol. Cell. Biol. 20, 8157–8167. Jiang, H., and Monaco, J. J. (1997). Immunogenetics 46, 93–98. Johnson, E. S., Ma, P. C., Ota, I. M., and Varshavsky, A. (1995). J. Biol. Chem. 270, 17442– 17456. Kandil, E., Kohda, K., Ishibashi, T., Tanaka, K., and Kasahara, M. (1997). Immunogenetics 46, 337–344. Kapelari, B., Bech-Otschir, D., Hegerl, R., Schade, R., Dumdey, R., and Dubiel, W. (2000). J. Mol. Biol. 300, 1169–1178. Karzai, A. W., Roche, E. D., and Sauer, R. T. (2000). Nat. Struct. Biol. 7, 449–455. Kawashima, T., Amano, N., Koike, H., Makino, S., Higuchi, S., Kawashima-Ohya, Y., Watanabe, K., Yamazaki, M., Kanehori, K., Kawamoto, T., Nunoshiba, T., Yamamoto, Y., Aramaki, H., Makino, K., and Suzuki, M. (2000). Proc. Natl. Acad. Sci. USA 97, 14257–14262. Keiler, K. C., Shapiro, L., and Williams, K. P. (2000). Proc. Natl. Acad. Sci. USA 97, 7778– 7783. Keiler, K. C., Waller, P. R., and Sauer, R. T. (1996). Science 271, 990–993. Kessel, M., Wu, W., Gottesman, S., Kocsis, E., Steven, A. C., and Maurizi, M. R. (1996). FEBS Lett. 398, 274–278. Kimura, Y., Takaoka, M., Tanaka, S., Sassa, H., Tanaka, K., Polevoda, B., Sherman, F., and Hirano, H. (2000). J. Biol. Chem. 275, 4635–4639. Kirschner, M. (1999). Trends Biochem. Sci. 24, M42–M45. Kisselev, A. F., Akopian, T. N., Castillo, V., and Goldberg, A. L. (1999a). Mol. Cell 4, 395– 402. Kisselev, A. F., Akopian, T. N., and Goldberg, A. L. (1998). J. Biol. Chem. 273, 1982–1989. Kisselev, A. F., Akopian, T. N., Woo, K. M., and Goldberg, A. L. (1999b). J. Biol. Chem. 274, 3363–3371. Kisselev, A. F., Songyang, Z., and Goldberg, A. L. (2000). J. Biol. Chem. 275, 14831–14837. Knipfer, N., and Shrader, T. E. (1997). Mol. Microbiol. 25, 375–383. Knowlton, J. R., Johnston, S. C., Whitby, F. G., Realini, C., Zhang, Z. G., Rechsteiner, M., and Hill, C. P. (1997). Nature 390, 639–643.

218

PETER ZWICKL ET AL.

Koegel, M., Hoppe, T., Schlenker, S., Ulrich, H. D., Mayer, T. U., and Jentsch, S. (1999). Cell 96, 635–644. Kopp, F., Hendil, K. B., Dahlmann, B., Kristensen, P., Sobek, A., and Uerkvitz, W. (1997). Proc. Natl. Acad. Sci. USA 94, 2939–2944. Koster, A. J., Walz, J., Lupas, A., and Baumeister, W. (1995). Mol. Biol. Rep. 21, 11–20. Lam, Y. A., Xu, W., DeMartino, G. N., and Cohen, R. E. (1997). Nature 385, 737–740. Larsen, C. N., and Finley, D. (1997). Cell 91, 431–434. Lee, D. H., Sherman, M. Y., and Goldberg, A. L. (1996). 16(N9), 4773–4781. Li, T., Duan, W., Yang, H., Lee, M.-K., Mustafa, F. B., Lee, B.-H., and Teo, T.-S. (2001). FEBS Lett. 488, 201–205. Li, T. W., Naqvi, N. I., Yang, H. Y., and Teo, T. S. (2000). Biochem. Biophys. Res. Commun. 272, 270–275. Lommel, L., Chen, L., Madura, K., and Sweder, K. (2000). Nucl. Acids Res. 28, 4839–4845. L¨owe, J., Stock, D., Jap, B., Zwickl, P., Baumeister, W., and Huber, R. (1995). Science 268, 533–539. L¨uders, J., Demand, J., and H¨ohfeld, J. (2000). J. Biol. Chem. 275, 4613–4617. Lupas, A., Baumeister, W., and Hofmann, K. (1997a). Trends Biochem. Sci. 22, 195–196. Lupas, A., Flanagan, J. M., Tamura, T., and Baumeister, W. (1997b). Trends Biochem. Sci. 22, 399–404. Lupas, A., Koster, A. J., and Baumeister, W. (1993). Enzyme Protein 47, 252–273. Lupas, A., Zwickl, P., and Baumeister, W. (1994). Trends Biochem. Sci. 19, 533–534. Ma, C. P., Slaughter, C. A., and DeMartino, G. N. (1992). J. Biol. Chem. 267, 10515–10523. Mallick, P., Goodwill, K. E., Fitz-Gibbon, S., Miller, J. H., and Eisenberg, D. (2000). Proc. Natl. Acad. Sci. USA 97, 2450–2455. Maniatis, T. (1999). Genes Dev. 13, 505–510. Mannhaupt, G., Schnall, R., Karpov, V., Vetter, I., and Feldmann, H. (1999). FEBS Lett. 450, 27–34. Masson, P., Andersson, O., Petersens, U. M., and Young, P. (2001). J. Biol. Chem. 276, 1383–1390. Maupin-Furlow, J. A., Aldrich, H. C., and Ferry, J. G. (1998). J. Bacteriol. 180, 1480–1487. Maupin-Furlow, J. A., Wilson, H. L., Kaczowka, S. J., and Ou, M. S. (2000). Frontiers in Bioscience 1, D837–D866A. Mayr, J., Seemuller, ¨ E., Muller, ¨ S. A., Engel, A., and Baumeister, W. (1998). J. Struct. Biol. 124, 179–188. Mayr, P. S. M., Allan, V. J., and Woodman, P. G. (1999). Eur. J. Cell Biol. 78, 224–232. McCormack, T. A., Cruikshank, A. A., Grenier, L., Melandri, F. D., Nunes, S. L., Plamondon, L., Stein, R. L., and Dick, L. R. (1998). Biochemistry 37, 7792–7800. Meacham, G. C., Patterson, C., Zhang, W. Y., Younger, J. M., and Cry, D. M. (2001). Nat. Cell Biol. 3, 100–105. Murakami, Y., Matsufuji, S., Hayashi, S., Tanahashi, N., and Tanaka, K. (2000). Biochem. Biophys. Res. Commun. 267, 1–6. Murata, S., Kawahara, H., Tohma, S., Yamamoto, K., Kasahara, M., Nabeshima, Y., Tanaka, K., and Chiba, T. (1999). J. Biol. Chem. 274, 38211–38215. Murray, B. W., Sultmann, H., and Klein, J. (2000). Scand. J. Immunol. 51, 571–576. Myung, J., Kim, K. B., Linsten, K., Dantuma, N. P., and Crews, C. M. (2001). Mol. Cell 7, 411–420. Nagy, I., Tamura, T., Vanderleyden, J., Baumeister, W., and De Mot, R. (1998). J. Bacteriol. 180, 5448–5453. Nandi, D., Woodward, E., Ginsburg, D. B., and Monaco, J. J. (1997). EMBO J. 16, 5363– 5375.

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

219

Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. (1999). Genome Res. 9, 27–43. Ng, D. T. W., Spear, E. D., and Walter, P. (2000). J. Cell Biol. 150, 77–88. Nussbaum, A. K., Dick, T. P., Keiholz, W., Schirle, M., Stevanovic, S., Dietz, K., Heinemeyer, W., Groll, M., Wolf, D. H., Huber, R., Rammensee, H. G., and Schild, H. (1998). Proc. Natl. Acad. Sci. USA 95, 12504–12509. Oinonen, C., and Rouvinen, J. (2000). Protein Sci. 9, 2329–2337. Orlowski, M., Cardozo, C., and Michaud, C. (1993). Biochemistry 32, 1563–1572. Orlowski, M., and Wilk, S. (2000). Arch. Biochem. Biophys. 383, 1–16. Ortolan, T. G., Tongaonkar, P., Lambertson, D., Chen, L., Schauber, C., and Madura, K. (2000). Nat. Cell Biol. 2, 601–608. Palombella, V. J., Rando, O. J., Goldberg, A. L., and Maniatis, T. (1994). Cell 78, 773–785. Papa, F. R., Amerik, A. Y., and Hochstrasser, M. (1999). Mol. Biol. Cell 10, 741–756. Parmentier, Y., Bouchez, D., Fleck, J., and Genschik, P. (1997). FEBS Lett. 416, 281–285. Peters, J. M., Cejka, Z., Harris, J. R., Kleinschmidt, J. A., and Baumeister, W. (1993). J. Mol. Biol. 234, 932–937. Peters, J. M., Franke, W. W., and Kleinschmidt, J. A. (1994). J. Biol. Chem. 269, 7709–7718. Pickart, C. M. (1997). FASEB J. 11, 1055–1066. Pickart, C. M. (2000). Trends Biochem. Sci. 25, 544–548. Pickart, C. M., and VanDemark, A. P. (2000). Nat. Struct. Biol. 7, 999–1001. Plemper, R. K., and Wolf, D. H. (1999). Trends Biochem. Sci. 24, 266–270. Preckel, T., Fung-Leung, W. P., Cai, Z. L., Vitiello, A., Salter-Cid, L., Winquvist, O., Wolfe, T. C., Von Herrath, M., Angulo, A., Ghazal, P., Lee, J. D., Fourie, A. M., Wu, Y., Pang, J., Ngo, K., Peterson, P. A., Fruh, K., and Yang, Y. (1999). Science 286, 2162–2165. Ramos, P. C., H¨ockendorff, J., Johnson, E. S., Varshavsky, A., and Dohmen, R. J. (1998). Cell 92, 489–499. Realini, C., Rogers, S. W., and Rechsteiner, M. (1994). FEBS Lett. 348, 109–113. Rechsteiner, M. (1998). In Ubiquitin and the Biology of the Cell ( J. M. Peters, J. R. Harris, and D. Finley, eds.), pp. 147–189. Plenum Press, New York. Rechsteiner, M., Realini, C., and Ustrell, V. (2000). Biochem. J. 345, 1–15. Richmond, C., Gorbea, C., and Rechsteiner, M. (1997). J. Biol. Chem. 272, 13403–13411. Rock, K. L., and Goldberg, A. L. (1999). Annu. Rev. Immunol. 17, 739–779. Rock, K. L., Gramm, C., Rothstein, L., Clark, K., Stein, R., Dick, L., Hwang, D., and Goldberg, A. L. (1994). Cell 78, 761–771. Rohrwild, M., Coux, O., Huang, H. C., Moerschell, R. P., Yoo, S. J., Seol, J. H., Chung, C. H., and Goldberg, A. L. (1996). Proc. Natl. Acad. Sci. USA 93, 5808–5813. Rohrwild, M., Pfeifer, G., Santarius, U., Muller, ¨ S. A., Huang, H. C., Engel, A., Baumeister, W., and Goldberg, A. L. (1997). Nat. Struct. Biol. 4, 133–139. Rubin, D. M., Glickman, M. H., Larsen, C. N., Dhruvakumar, S., and Finley, D. (1998). EMBO J. 17, 4909–4919. Ruepp, A., Eckerskorn, C., Bogyo, M., and Baumeister, W. (1998). FEBS Lett. 425, 87–90. Ruepp, A., Graml, W., Santos-Martinez, M. L., Koretke, K. K., Volker, C., Mewes, H. W., Frishman, D., Stocker, S., Lupas, A. N., and Baumeister, W. (2000). Nature 407, 508– 513. Russell, S. J., Reed, S. H., Huang, W., Friedberg, E. C., and Johnston, S. A. (1999a). Mol. Cell 3, 687–695. Russell, S. J., Sathyanarayana, U. G., and Johnston, S. A. (1996). J. Biol. Chem. 271, 32810– 32817. Russell, S. J., Steger, K. A., and Johnston, S. A. (1999b). J. Biol. Chem. 274, 21943–21952. Salzmann, U., Kral, S., Braun, B., Standera, S., Schmidt, M., Kloetzel, P. M., and Sijts, A. (1999). FEBS Lett. 454, 11–15.

220

PETER ZWICKL ET AL.

Sassa, H., Oguchi, S., Inoue, T., and Hirano, H. (2000). Gene 250, 61–66. Satoh, K., Sasajima, H., Nyoumura, K., Yokosawa, H., and Sawada, H. (2001). Biochemistry 40, 314–319. Schauer, T. M., Nesper, M., Kehl, M., Lottspeich, F., M¨uller-Taubenberger, A., Gerisch, G., and Baumeister, W. (1993). J. Struct. Biol. 111, 135–147. Schmidt, M., Lupas, A. N., and Finley, D. (1999a). Curr. Opin. Chem. Biol. 3, 584–591. Schmidt, M., Zantopf, D., Kraft, R., Kostka, S., Preissner, R., and Kloetzel, P. M. (1999b). J. Mol. Biol. 288, 117–128. Schmidtke, G., Emch, S., Groettrup, M., and Holzh¨utter, H. G. (2000). J. Biol. Chem. 275, 22056–22063. Schmidtke, G., Holzh¨utter, H. G., Bogyo, M., Kairies, N., Groll, M., de Giuli, R., Emch, S., and Groettrup, M. (1999). J. Biol. Chem. 274, 35734–35740. Schmidtke, G., Kraft, R., Kostka, S., Henklein, P., Frommel, C., L¨owe, J., Huber, R., Kloetzel, P. M., and Schmidt, M. (1996). EMBO J. 15, 6887–6898. Schmidtke, G., Schmidt, M., and Kloetzel, P. M. (1997). J. Mol. Biol. 268, 95–106. Schubert, U., Anton, L. C., Gibbs, J., Norbury, C. C., Yewdell, J. W., and Bennink, J. R. (2000). Nature 404, 770–774. Seeger, M., Gordon, C., Ferrell, K., and Dubiel, W. (1996). J. Mol. Biol. 263, 423–431. Seeger, M., Kraft, R., Ferrell, K., Bech-Otschir, D., Dumdey, R., Schade, R., Gordon, C., Naumann, M., and Dubiel, W. (1998). FASEB J. 12, 469–478. Seemuller, ¨ E., Lupas, A., and Baumeister, W. (1996). Nature 382, 468–470. Seemuller, ¨ E., Lupas, A., Stock, D., L¨owe, J., Huber, R., and Baumeister, W. (1995). Science 268, 579–582. Song, H. K., Hartmann, C., Ramachandran, R., Bochtler, M., Behrendt, R., Moroder, L., and Huber, R. (2000). Proc. Natl. Acad. Sci. USA 97, 14103–14108. Sousa, M. C., Trame, C. B., Tsuruta, H., Wilbanks, S. M., Reddy, V. S., and McKay, D. B. (2000). Cell 103, 633–643. Stoltze, L., Schirle, M., Schwarz, G., Schr¨oter, C., Thompson, M. W., Hersh, L. B., Kalbacher, H., Stevanovic, S., Rammensee, H. G., and Schild, H. (2000). Nat. Immunol. 1, 413–418. Strickland, E., Hakala, K., Thomas, P. J., and DeMartino, G. N. (2000). J. Biol. Chem. 275, 5565–5572. Swafﬁeld, J. C., and Purugganan, M. D. (1997). J. Molec. Evol. 45, 549–563. Tamura, N., Lottspeich, F., Baumeister, W., and Tamura, T. (1998). Cell 95, 1–20. Tamura, N., Pfeifer, G., Baumeister, W., and Tamura, T. (2001). Biol. Chem. 382, 449–458. Tamura, T., Nagy, I., Lupas, A., Lottspeich, F., Cejka, Z., Schoofs, G., Tanaka, K., De Mot, R., and Baumeister, W. (1995). Curr. Biol. 5, 766–774. Tamura, T., Tamura, N., Cejka, Z., Hegerl, R., Lottspeich, F., and Baumeister, W. (1996). Science 274, 1385–1389. Tanahashi, N., Murakami, Y., Minami, Y., Shimbara, N., Hendil, K. B., and Tanaka, K. (2000). J. Biol. Chem. 275, 14336–14345. Tanaka, K. (1998). Biochem. Biophys. Res. Commun. 247, 537–541. Tanaka, K., and Kasahara, M. (1998). Immunol. Rev. 163, 161–176. Thrower, J. S., Hoffman, L., Rechsteiner, M., and Pickart, C. M. (2000). EMBO J. 19, 94–102. To, W. Y., and Wang, C. C. (1997). FEBS Lett. 404, 253–262. Tsurumi, C., Shimizu, Y., Saeki, M., Kato, S., DeMartino, G. N., Slaughter, C. A., Fujimuro, M., Yokosawa, H., Yamasaki, M., Hendil, K. B., Toh-e, A., Tanahashi, N., and Tanaka, K. (1996). Eur. J. Biochem. 239, 912–921. Udvardy, A. (1993). J. Biol. Chem. 268, 9055–9062.

THE PROTEASOME: A SUPRAMOLECULAR ASSEMBLY

221

van Nocker, S., Sadis, S., Rubin, D. M., Glickman, M., Fu, H., Coux, O., Wefes, I., Finley, D., and Vierstra, R. D. (1996). Mol. Cell. Biol. 16, 6020–6028. Varshavsky, A., Turner, G., Du, F. Y., and Xie, Y. M. (2000). Biol. Chem. 381, 779–789. Verma, R., Chen, S., Feldman, R., Schieltz, D., Yates, J., Dohmen, T., and Deshaies, R. J. (2000). Mol. Biol. Cell 11, 3425–3439. Verma, R., and Deshaies, R. J. (2000). Cell 101, 341–344. Voges, D., Zwickl, P., and Baumeister, W. (1999). Annu. Rev. Biochem. 68, 1015–1068. Walker, J. E., Saraste, M., Runswick, M. J., and Gay, N. J. (1982). EMBO J. 1, 945–951. Walz, J., Erdmann, A., Kania, M., Typke, D., Koster, A. J., and Baumeister, W. (1998). J. Struct. Biol. 121, 19–29. Walz, J., Tamura, T., Tamura, N., Grimm, R., Baumeister, W., and Koster, A. J. (1997). Mol. Cell 1, 59–65. Wang, E. W., Kessler, B. M., Borodovsky, A., Cravatt, B. F., Bogyo, M., Ploegh, H. L., and Glas, R. (2000). Proc. Natl. Acad. Sci. USA 97, 9990–9995. Wang, J., Song, J. J., Franklin, M. C., Kamtekar, S., Im, Y. J., Rho, S. H., Seong, I. S., Lee, C. S., Chung, C. H., and Eom, S. H. (2001). Structure 9, 177–184. Wang, W., Chevray, P. M., and Nathans, D. (1996). Proc. Natl. Acad. Sci. USA 93, 8236–8240. Watanabe, T. K., Saito, A., Suzuki, M., Fujiwara, T., Takahashi, E., Slaughter, C. A., DeMartino, G. N., Hendil, K. B., Chung, C. H., Tanahashi, N., and Tanaka, K. (1998). Genomics 50, 241–250. Wenzel, T., and Baumeister, W. (1995). Nat. Struct. Biol. 2, 199–204. Wenzel, T., Eckerskorn, C., Lottspeich, F., and Baumeister, W. (1994). FEBS Lett. 349, 205–209. Whitby, F. G., Masters, E. I., Kramer, L., Knowlton, J. R., Yao, Y., Wang, C. C., and Hill, C. P. (2000). Nature 408, 115–120. Wickner, S., Maurizi, M. R., and Gottesman, S. (1999). Science 286, 1888–1893. Wilkinson, C. R. M., Ferrell, K., Penney, M., Wallace, M., Dubiel, W., and Gordon, C. (2000). J. Biol. Chem. 275, 15182–15192. Wilkinson, K. D. (2000). Sem. Cell Dev. Biol. 11, 141–148. Wilson, H. L., Aldrich, H. C., and MaupinFurlow, J. (1999). J. Bacteriol. 181, 5814–5824. Wilson, H. L., Ou, M. S., Aldrich, H. C., and Maupin-Furlow, J. (2000). J. Bacteriol. 182, 1680–1692. Witt, E., Zantopf, D., Schmidt, M., Kraft, R., Kloetzel, P. M., and Kr¨uger, E. (2000). J. Mol. Biol. 301, 1–9. Wolf, S., Nagy, I., Lupas, A., Pfeifer, G., Cejka, Z., M¨uller, S. A., Engel, A., De Mot, R., and Baumeister, W. (1998). J. Mol. Biol. 277, 13–25. Xie, Y. M., and Varshavsky, A. (2000). Proc. Natl. Acad. Sci. USA 97, 2497–2502. Yang, Y., Fruh, ¨ K., Ahn, K., and Peterson, P. A. (1995). J. Biol. Chem. 270, 27687–27694. Yao, Y., Huang, L., Krutchinsky, A., Wong, M. L., Standing, K. G., Burlingame, A. L., and Wang, C. C. (1999a). J. Biol. Chem. 274, 33921–33930. Yao, Y., Toth, C. R., Huang, L., Wong, M. L., Dias, P., Burlingame, A. L., Cofﬁno, P., and Wang, C. C. (1999b). Biochem. J. 344, 349–358. Yewdell, J. W., Anton, L. C., and Bennink, J. R. (1996). J. Immunol. 157, 1823–1826. Yokota, K., Kagawa, S., Shimizu, Y., Akioka, H., Tsurumi, C., Noda, C., Fujimuro, M., Yokosawa, H., Fujiwara, T., Takahashi, E., Ohba, M., Yamasaki, M., DeMartino, G. N., Slaughter, C. A., Toh-e, A., and Tanaka, K. (1996). Mol. Biol. Cell 7, 853–870. York, I. A., Goldberg, A. L., Mo, X. Y., and Rock, K. L. (1999). Immunol. Rev. 172, 49–66. Yoshimura, T., Kameyama, K., Takagi, T., Ikai, A., Tokunaga, F., Koide, T., Tanahashi, N., Tamura, T., Cejka, Z., Baumeister, W., Tanaka, K., and Ichihara, A. (1993). J. Struct. Biol. 111, 200–211.

222

PETER ZWICKL ET AL.

Young, P., Deveraux, Q., Beal, R. E., Pickart, C. M., and Rechsteiner, M. (1998). J. Biol. Chem. 273, 5461–5467. Zhang, X. D., Shaw, A., Bates, P. A., Newman, R. H., Gowen, B., Orlova, E., Gorman, M. A., Kondo, H., Dokurno, P., Lally, J., Leonard, G., Meyer, H., van Heel, M., and Freemont, P. S. (2000a). Mol. Cell 6, 1473–1484. Zhang, Z. G., Clawson, A., Realini, C., Jensen, C. C., Knowlton, J. R., Hill, C. P., and Rechsteiner, M. (1998a). Proc. Natl. Acad. Sci. USA 95, 2807–2811. Zhang, Z. G., Krutchinsky, A., Endicott, S., Realini, C., Rechsteiner, M., and Standing, K. G. (1999). Biochemistry 38, 5651–5658. Zhang, Z. G., Realini, C., Clawson, A., Endicott, S., and Rechsteiner, M. (1998b). J. Biol. Chem. 273, 9501–9509. Zhang, Z. S., Torii, N., Furusaka, A., Malayaman, N., Hu, Z. Y., and Liang, T. J. (2000b). J. Biol. Chem. 275, 15157–15165. Z¨uhl, F., Seemuller, ¨ E., Golbik, R., and Baumeister, W. (1997). FEBS Lett. 418, 189–194. Zwickl, P., and Baumeister, W. (1999). Nat. Cell Biol. 1, E97–E98. Zwickl, P., Baumeister, W., and Steven, A. (2000a). Curr. Opin. Struct. Biol. 10, 242–250. Zwickl, P., Goldberg, A. L., and Baumeister, W. (2000b). In Proteasomes: The World of Regulatory Proteolysis (W. Hilt and D. H. Wolf, eds.), pp. 8–20. Eurekah.com/Landes Bioscience, Georgetown. Zwickl, P., Grziwa, A., P¨uhler, G., Dahlmann, B., Lottspeich, F., and Baumeister, W. (1992). Biochemistry 31, 964–972. Zwickl, P., Kleinz, J., and Baumeister, W. (1994). Nat. Struct. Biol. 1, 765–770. Zwickl, P., Ng, D., Woo, K. M., Klenk, H.-P., and Goldberg, A. L. (1999). J. Biol. Chem. 274, 26008–26014.

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION By MICHAEL T. RYAN* and NIKOLAUS PFANNER† *Department of Biochemistry, La Trobe University, 3086 Melbourne, Australia, and †Institut fur ¨ Biochemie und Molekularbiologie, Universitat ¨ Freiburg Hermann-Herder-Strasse 7, 79104 Freiburg, Germany

I. II. III. IV. V.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protein Translocation into Mitochondria and ER . . . . . . . . . . . . . . . . . . . . . . . Cytosolic Hsp70s Are Involved in Protein Translocation . . . . . . . . . . . . . . . . . . Hsp70 and Its Cofactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lumenal Hsp70s and Protein Translocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Identiﬁcation of ER and Mitochondrial Hsp70 Proteins . . . . . . . . . . . . . . B. mtHsp70 and BiP are Concentrated at the Preprotein Exit Sites of Translocation Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Models of Lumenal Hsp70 Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Other Roles for Hsp70s in Protein Translocation . . . . . . . . . . . . . . . . . . . . . . . VII. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

223 224 227 228 229 229 230 231 237 238 239

I. INTRODUCTION Compartmentalization of proteins within the organelles of a eukaryotic cell serves to coordinate and separate various metabolic processes. Except for a few proteins of the mitochondria and chloroplast, all proteins are encoded by nuclear genes and at least begin synthesis in the cytosol. During or following their synthesis, many proteins are targeted to various organelles. All organelles contain receptor components that serve to recognize the appropriately targeted protein and, in doing so, ﬁlter out proteins that do not belong to the compartment (Schatz and Dobberstein, 1996). Although the import of proteins into the nucleus and perhaps peroxisomes involves their transport across large membrane openings, other organelles contain relatively narrow channels. The Sec61 translocation complex of the endoplasmic reticulum (ER) forms ribosome-free channels of ∼20 A˚ (Hanein et al., 1996), while the mitochondrial outer membrane translocation component, Tom40, forms channels of ∼22 A˚ (Hill et al., 1998). The diameter of the mitochondrial inner membrane translocation channel has been estimated as being even smaller (Schwartz and Matouschek, 1999). As a result, proteins directed into these particular organelles translocate across these membrane channels in an unfolded conformation. This was demonstrated by Eilers and Schatz (1986), who 223 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

224

MICHAEL T. RYAN AND NIKOLAUS PFANNER

showed that a folded protein domain fused to a mitochondrial targeting signal was prevented from translocating into the matrix and remained exposed to the cytosol. A similar prerequisite for proteins to be unfolded in order to translocate into the ER has also been shown (M¨uller and Zimmerman, 1988). This requirement was also observed in vivo, where a fusion protein consisting of a mitochondrial targeting signal fused to the cytosolic protein dihydrofolate reductase (DHFR) could be imported into mitochondria when expressed in yeast cells but not in cells incubated with methotrexate, which causes DHFR to fold stably (Wienhues et al., 1991). Such ﬁndings lead to a question: How are these proteins unfolded in order to translocate across the organellar membrane(s)? Over the last three decades, much information has been gathered about the processes of protein translocation, particularly those of mitochondria and the ER. II. PROTEIN TRANSLOCATION INTO MITOCHONDRIA AND ER For most proteins of the ER, translocation is coupled with protein synthesis. The synthesis of the growing polypeptide from the ribosome docked onto the ER is sufﬁcient to drive the forward movement of the preprotein into the organelle (Fig. 1; Walter and Johnson, 1994; Matlack et al., 1998). However, it was found that some ER-targeted proteins, especially those found in yeast, could be imported into this organelle in a posttranslational fashion. Thus, in such cases the ribosome can play no part in driving preprotein translocation. Indeed an in vitro ER protein import assay has been developed that utilizes fully synthesized preproteins (Hansen et al., 1986; Rothblatt and Meyer, 1986; Waters and Blobel, 1986), and at least one ER-directed preprotein has been shown to be folded in the cytosol before its translocation into the organelle (Paunola et al., 1998). It seems that for yeast, posttranslational import utilizes the same ER Sec61 translocation channel as that employed in the cotranslational system; however, additional factors are also recruited to the complex (Fig. 1; Panzer et al., 1995). The majority of mitochondria-destined preproteins seem to be imported after completion of polypeptide synthesis (Pfanner et al., 1997; Herrmann and Neupert, 2000). Preproteins are targeted to the outer membrane, often via N-terminal signals, where they bind to translocase components of the outer membrane (TOM). Preproteins subsequently translocate across the general import pore (Ryan and Pfanner, 1998). Extra complexity is introduced into matrix-imported preproteins because they must also translocate across the inner membrane (Fig. 2). This is achieved via interactions of the preproteins with the translocase

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

225

FIG. 1. Protein translocation into the ER. The translocation of preproteins into the ER lumen is most often performed in a cotranslational manner. The ribosome docks and forms a seal over the Sec61 complex that forms the translocation channel. The resumption of polypeptide synthesis is sufﬁcient to drive the forward movement of the preprotein into the lumen. For the posttranslational import of preproteins into the lumen, the action of cytosolic Hsp70 and its DnaJ-like cofactor is involved in maintatining the preprotein in an import-competent state prior to translocation. Following recognition and initial translocation at the Sec61–Sec63 complex, the Sec63p-associated lumenal form of Hsp70, BiP, binds the preprotein. Through ATP-dependent cycles of binding and release, BiP drives the forward movement of the preprotein into the lumen.

of the inner membrane (TIM). The TIM23 complex, containing Tim23 and Tim17, translocates all matrix-directed preproteins (Rassow et al., 1999; Bauer et al., 2000). Here, the mitochondrial membrane potential is essential for import, and it was hypothesized that the negatively charged interior of the matrix may exert an electrophoretic effect to drive the translocation of preproteins with typical positively charged N-terminal targeting signals (Schleyer et al., 1982; Martin et al., 1991). In addition, the membrane potential was also found to be important in maintaining Tim23 in a dimerized form and was involved in regulating its gating (Bauer et al., 1996). However, the membrane potential alone is not sufﬁcient to drive preprotein translocation, since it was found that

226

MICHAEL T. RYAN AND NIKOLAUS PFANNER

FIG. 2. Protein translocation into mitochondria. Following synthesis, mitochondrialtargeted preproteins are directed to the translocase components of the outer membrane (TOM). Cytosolic Hsp70 and its DnaJ-like cofactor may bind to the preprotein, thereby preventing it from aggregating. Following translocation across the general import pore (GIP) of the outer membrane (OM), matrix-destined preproteins cross the intermembrane space (IMS) and engage with the TIM23 complex of the inner membrane (IM). The preprotein translocates across the inner membrane in a process requiring a membrane potential and mtHsp70. Tim44-associated mtHsp70 binds to the preprotein, and with ATP hydrolysis and the action of its cofactor, Mge1, the preprotein is driven across the translocation channels and into the matrix.

while the membrane potential could draw the targeting signal of a preprotein into the matrix, it was not sufﬁcient to translocate the remainder of the polypeptide (Martin et al., 1991; Ungermann et al., 1996; Neupert, 1997; Voos et al., 1999). Early studies showed that ATP was needed for preprotein import into mitochondria; however, there was some debate regarding whether this

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

227

requirement was external (i.e., cytosolic) or internal (matrix). It was concluded that while some preproteins utilize cytosolic ATP (Pfanner and Neupert, 1986; Chen and Douglas, 1987; Eilers et al., 1987; Wachter et al., 1994), matrix ATP is crucial for the import of perhaps all preproteins (Stuart et al., 1994; Wachter et al., 1994). Speciﬁc depletion of ATP from the mitochondrial matrix results in preproteins being able to cross the outer membrane, but they undergo only partial translocation across the inner membrane TIM23 complex (Hwang et al., 1991; Rassow and Pfanner, 1991). ATP was also shown to be required for the posttranslational import of preproteins into the ER lumen (Hansen et al., 1986; Rothblatt and Meyer, 1986; Waters and Blobel, 1986). The requirement was found to be subsequent to preprotein binding to the ER membrane (Sanz and Meyer, 1989). III. CYTOSOLIC Hsp70s ARE INVOLVED IN PROTEIN TRANSLOCATION During this time, a revolution in the ﬁeld of protein folding was taking place. Pelham (1986) suggested that stress or heat-shock proteins (Hsps) were involved in binding to unfolded or partially denatured proteins and were involved in their assembly/disassembly. Members of this group include the Hsp70 family, and homologs exist in many of the cellular compartments of the eukaryotic cell. It was found that not only did Hsp70 proteins bind to nascent polypeptides, but also cytosolic forms of Hsp70 and its cofactors were involved in the import of some preproteins into both the ER and mitochondria. The posttranslational import of yeast prepro-alpha factor into the ER was stimulated by the addition of a yeast postribosomal supernatant. This assay was used to purify two cytosolic members of the yeast Hsp70 family (Chirico et al., 1988). Similarly, while the import of a preprotein translated from wheat germ lysate could not be imported into yeast mitochondria, addition of yeast cytosolic Hsp70 (as well as another unidentiﬁed component) could restore this defect (Murakami et al., 1988). The in vitro studies implicating Hsp70 in preprotein translocation were also complemented through genetic means using yeast cells. Although deletion of all four genes encoding cytosolic Hsp70s is lethal to the cell, the expression of one gene is sufﬁcient for cell survival (WernerWashburne et al., 1987). One of the four Hsp70 members was expressed on a single-copy vector, under the control of a galatose-inducible promoter, in a mutant strain disrupted in all chromosomal copies of the cytosolic Hsp70 genes (Deshaies et al., 1988). Transfer of cells from galactose-to glucose-based media led to the repression of Hsp70 synthesis

228

MICHAEL T. RYAN AND NIKOLAUS PFANNER

and a gradual reduction in the level of this protein within the cell. Analysis of cytosolic extracts revealed an accumulation of the ER-targeted protein prepro-alpha factor. A mitochondrially targeted preprotein, the precursor of the β subunit of the F1F0-ATPase also accumulated in the cytosol (Deshaies et al., 1988). These studies revealed, for the ﬁrst time, a requirement for cytosolic Hsp70 in preprotein translocation. How does Hsp70 function? It is thought that cytosolic Hsp70 molecules, like other molecular chaperones, bind to newly synthesized preproteins and prevent their aggregation, thereby maintaining preproteins in an import-competent state (Brodsky, 1996). Immunodepletion of Hsp70 from rabbit reticulocyte lysate, from which preproteins are ﬁrst synthesized prior to incubation with organelle, led to a disruption in protein import into mitochondria. Addition of Hsp70 following translation did not restore this defect. Only when Hsp70 was present during translation did the preprotein remain competent for import into mitochondria (Terada et al., 1995). Preproteins can bypass the requirement of cytosolic Hsp70 in preprotein import by urea treatment before their addition to organelles (Chirico et al., 1988). IV. Hsp70 AND ITS COFACTORS Polypeptide binding and release by Hsp70 is coupled to its nucleotidebinding state. Furthermore, Hsp70 proteins often act in concert with cofactors involved in the regulation of its ATPase cycle. These cofactors thereby indirectly contribute to the process of polypeptide binding and release (see chapter by Mayer et al., in this volume). In eukaryotes, cytosolic homologs of the co-chaperone DnaJ, termed Ydj1p in yeast and dj2 in mammals, have been implicated in preprotein import, indicating that they act with cytosolic forms of Hsp70 (Atenico and Yaffe, 1992; Caplan et al., 1992; Terada et al., 1997). Indeed, Ydj1p can stimulate the release of preproteins from cytosolic Hsp70 (Cyr et al., 1992). A recent crosslinking analysis revealed that cytosolic Hsp70, along with another cytosolic chaperonin, TriC, binds to ER-targeted preproteins following the release of other cytosolic factors, NAC and SRP, from the polypeptide. Subsequently, the preprotein was spontaneously released from Hsp70, indicating that Hsp70 acts more as a molecular chaperone than as a speciﬁc targeting factor (Plath and Rapoport, 2000). For mitochondria, another independently interacting cytosolic factor, termed mitochondrial import stimulation factor (MSF), is also important for guiding preproteins to its surface (Mihara and Omura, 1996). Studies by Mihara and coworkers have indicated that cytosolic Hsp70

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

229

acts in concert with MSF in directing some preproteins to particular TOM receptors of the mitochondrial outer membrane (Hachiya et al., 1995; Komiya et al., 1996). Under the conditions used, MSF was found to bind to most, if not all, preproteins translated in the reticulocyte lysate, while the additional binding to Hsp70 was dependent on the state of the preprotein itself. An MSF–preprotein complex was targeted to a receptor complex containing the receptor Tom70, while an MSF–Hsp70–preprotein complex bypassed this interaction and bound directly to a different receptor subcomplex composed of Tom20 and Tom22 that sits at the general import pore (Hachiya et al., 1995). Curiously, no ATP dependence was observed for Hsp70 function in this process (Komiya et al., 1996). An involvement of cytosolic DnaJ homologs in this process was also not reported. The in vivo relevance of MSF for mitochondrial protein import is not yet known. V. LUMENAL Hsp70s AND PROTEIN TRANSLOCATION A. Identification of ER and Mitochondrial Hsp70 Proteins The genes encoding the major forms of yeast mitochondrial matrix Hsp70, mtHsp70, and the ER lumenal Hsp70, BiP (or Kar2p), are essential for cell viability (Craig et al., 1989; Normington et al., 1989; Rose et al., 1989). Because Hsp70s are ATPases and bind unfolded polypeptides, a connection between the requirements of both preprotein unfolding and the hydrolysis of lumenal ATP in preprotein translocation could be made. A preprotein arrested in the mitochondrial membranes and exposed to the matrix was associated with mtHsp70 (Ostermann et al., 1990; Scherer et al., 1990). Yeast cells harboring a temperature-sensitive mutant form of mtHsp70 accumulated mitochondrially targeted preproteins in the cytosol under nonpermissive conditions (Kang et al., 1990). Preincubation of preproteins with urea was able to restore their import into mitochondria containing impaired mtHsp70, demonstrating that mtHsp70 was involved in unfolding of preproteins (Kang et al., 1990; Gambill et al. 1993). Likewise for the ER, it was observed that mutants of BiP that led to cells’ being temperature-sensitive for growth or the direct depletion of BiP from cells blocked the translocation of a number of ER-destined preproteins and resulted in their accumulation in the cytosol (Vogel et al., 1990). Although the preprotein of invertase was found on the cytosolic face of the ER, its signal sequence was proteolytically removed, indicating that it had most likely engaged with the translocation machinery but could

230

MICHAEL T. RYAN AND NIKOLAUS PFANNER

not be completely translocated into the lumen (Nguyen et al., 1991). Thus, BiP acted at a later step in the preprotein translocation pathway. Preproteins arrested at the ER translocation site could be crosslinked to BiP, conﬁrming that BiP interacts directly with incoming polypeptides (M¨usch et al., 1992; Sanders et al., 1992). While mtHsp70 and BiP are intimately involved in protein translocation, they also act as typical molecular chaperones where they are involved in the folding of newly imported proteins and in stabilizing proteins under conditions of cell or organellar stress (Kang et al., 1990; Gething and Sambrook, 1992; Voisine et al., 1999; Molinari and Helenius, 2000). B. mtHsp70 and BiP are Concentrated at the Preprotein Exit Sites of Translocation Channels Hsp70s are soluble proteins and therefore must somehow be situated at the preprotein exit sites of membrane translocation channels in order to bind incoming preproteins and facilitate their import (Figs. 1 and 2). Sec63p, an integral membrane protein found at ER translocation sites, is required for preprotein translocation. Both genetic and biochemical analyses indicated that Sec63p and BiP interact (Scidmore et al., 1993; Brodsky and Schekman, 1993). This interaction was later found to occur through the DnaJ-homologous region, termed a J-motif, found in the lumenal domain of Sec63p (Corsi and Schekman, 1997). A stable interaction between BiP and Sec63p can occur in the presence of ATP but not ATPγ S, indicating that ATP hydrolysis is necessary (Brodsky and Schekman, 1993). Indeed, a complex formed between puriﬁed BiP and an Sec63 J-domain fusion protein was found to have converted radiolabeled ATP to ADP (Corsi and Schekman, 1997). It was also shown that while preproteins can be bound to the Sec61 translocation machinery in the absence of ATP, Sec63p-associated BiP released them in an ATP-dependent fashion (Lyman and Schekman, 1997). Matlack and colleagues (1997) elegantly showed that a seven-component yeast translocation complex, including Sec63p and devoid of lipid, was sufﬁcient for preprotein binding and partial translocation. The additional presence of BiP and ATP resulted in polypeptide movement completely through the channel and its release at the trans side of the complex. Mitochondrial Tim44 is a major binding site for mtHsp70, and a stable complex can be found in the absence of ATP (Kronidou et al., 1994; Rassow et al., 1994; Schneider et al., 1994; Voos et al., 1996). About 2% of total mtHsp70 can be in contact with Tim44, while the remainder is in the soluble portion of the matrix. Tim44 is a peripheral

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

231

membrane protein attached to the matrix side of the mitochondrial inner membrane and in contact with members of the TIM23 complex. Disruption of the gene encoding Tim44 results in cell death, while Tim44 mutations lead to a reduction in mitochondrial protein translocation (Maarse et al., 1992; Scherer et al., 1992; Blom et al., 1993; Horst et al., 1993; Moro et al., 1999). Tim44 also contains a short sequence with a weak similarity to part of the J-motif that may bind to mtHsp70, and this region is essential for its function (Rassow et al., 1994; Merlin et al., 1999). The interaction of Tim44 with mtHsp70 has been shown to require the ATPase domain of mtHsp70, while the substrate-binding domain of the chaperone plays only a stabilizing role in this interaction (Krimmer et al., 2000). It seems that mtHsp70 contains distinct binding sites for Tim44 and substrate. This is supported by studies showing that a mutant form of mtHsp70 does not efﬁciently associate with Tim44, but still binds polypeptides (von Ahsen et al., 1995; Voisine et al., 1999). Indeed, a complex between mtHsp70, Tim44, and preprotein has been detected (Horst et al., 1993, 1996; Berthold et al., 1995). In contrast to the ER, a mitochondrial homolog of the co-chaperone and nucleotide exchange factor GrpE, termed Mge1, is present in the matrix and is involved in the cycling of the mtHsp70 ATPase by releasing ADP from mtHsp70 (Dekker and Pfanner, 1997; Miao et al., 1997). Like Tim44, Mge1 can form a stable complex with mtHsp70 and a translocating preprotein (Voos et al., 1994). C. Models of Lumenal Hsp70 Action It is generally accepted that lumenal Hsp70 proteins bind to preproteins and assist in their translocation across membranes in a process requiring ATP hydrolysis. The mechanism of how Hsp70 exerts its action has been the subject of intense interest ( Jensen and Johnson, 1999; Pilon and Schekman, 1999; Herrmann and Neupert, 2000; Matouschek et al., 2000), and two different models have been proposed (Fig. 3). 1. The Brownian Ratchet Model In the Brownian ratchet model (Simon et al., 1992), Hsp70 molecules act to support unidirectional translocation of preproteins in a somewhat passive way. Brownian motion describes the random thermal motion of a system—in this case, the preprotein. If a preprotein is in transit at the translocation channel, Brownian motion will cause it to oscillate in an unbiased way. However, the binding of Hsp70 to the incoming preprotein at the exit site of the translocation channel prevents its backsliding. Hsp70 bound to preprotein is then released from its anchor (e.g., Sec63p

232

MICHAEL T. RYAN AND NIKOLAUS PFANNER

FIG. 3. Models of lumenal Hsp70 action in protein translocation. (I) In the Brownian ratchet model, the preprotein enters the lumen and associates with membrane-bound Hsp70. Hydrolysis of ATP results in the preprotein substrate being more tightly bound

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

233

or Tim44) through the turnover of ATP. The empty binding site allows another soluble Hsp70 molecule to bind to the anchor. Spontaneous protein unfolding at the cis side of the membrane allows the preprotein to translocate further into the lumen where it can bind to additional Hsp70 molecules. This model suggests that Hsp70 binding does not accelerate preprotein unfolding. Rather, a ratchet facilitates the translocation of segments of the preprotein into the matrix owing to its spontaneous unfolding at the mitochondrial surface coupled with biased Brownian movements that enable additional Hsp70 molecules to be recruited to the incoming polypeptide. 2. The Pulling Model In contrast to the Brownian ratchet, the pulling model suggests that Hsp70 exerts a force that accelerates the forward movement of the preprotein. The model is based on proteolytic mapping experiments indicating that the conformations of the ATP- and ADP-bound forms of Hsp70s differ (Liberek et al., 1991; Buchberger et al., 1995; von Ahsen et al., 1995). The membrane anchor to which Hsp70 is bound could serve as a molecular fulcrum and convert the energy released from these conformational changes into a driving force or “power stroke” sufﬁcient to pull a bound preprotein through the translocation channel (Glick, 1995; Pfanner and Meijer, 1995; Pilon and Scheckman, 1999). 3. Experimental Evidence for Models Evidence for both models exists, and while the action of mtHsp70 remains controversial, most evidence indicates that BiP acts as a ratchet. Using puriﬁed proteins in surface plasmon resonance assays, it was shown that the Sec63p J-domain could bind and activate BiP to bind peptide substrates in its immediate vicinity in the presence of ATP. In this manner, BiP could bind to a wide range of substrates, including some that it would not normally bind to on its own (Misselwitz et al., 1998). Such promiscuous binding would be favorable to preprotein translocation in ←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− to Hsp70. Hsp70 bound to preprotein is released from its membrane anchor, and additional molecules of Hsp70 recruited to the translocation channel can bind to the incoming preprotein. The preprotein is translocated into the lumen by now biased Brownian movements. (II) The Pulling model differs from the Brownian ratchet model in that the hydrolysis of ATP is coupled to a conformational change in the substrate-binding domain of Hsp70. Since Hsp70 is anchored at the membrane, this conformational change results in force that leads to the preprotein being pulled through the translocation channel. Release of Hsp70 from its membrane anchor and binding of additional Hsp70 molecules act to accelerate the translocation of the preprotein into the lumen.

234

MICHAEL T. RYAN AND NIKOLAUS PFANNER

which BiP could immediately bind to the incoming preprotein following Sec63p-mediated activation. Interestingly, it was found that when BiP was bound to the J-domain of Sec63p in the presence of ADP, it could not bind substrates (Misselwitz et al., 1999). Mutants of BiP that were defective in substrate binding but still capable of ATP hydrolysis also could not bind to the J-domain. It was suggested that if no substrate were available following BiP’s activation, then BiP would recognize Sec63p as a substrate (Misselwitz et al., 1999). In the presence of peptide substrate, BiP would bind only transiently to Sec63p, requiring only activation, before it bound to the incoming polypeptide substrate. This surprising ﬁnding suggested that BiP might operate as a ratchet in preprotein translocation, since a transient interaction with Sec63p may not be sufﬁcient for it to act as a membrane fulcrum in preprotein pulling. Matlack and colleagues (1999) further investigated this proposal. A ratchet, or trapping, model suggests that multiple molecules of BiP bind to the incoming preprotein and prevent backsliding, and this therefore favors unidirectional translocation. The translocating preprotein, prepro-alpha factor, was released from a puriﬁed Sec complex and subjected to sucrose density gradient centrifugation under conditions where an interaction with BiP was maintained. It was found that multiple molecules (perhaps 6 or 7) of BiP were associated with the 165–amino acid preprotein. To test if backsliding could occur, the preprotein was partially imported into vesicles containing the reconstituted Sec machinery along with internalized BiP. Following ATP depletion, which prevents further BiP binding to the substrate and a gradual release of any bound BiP, the preprotein became more accessible to externally added proteases, indicating that it was sliding out of the channel (Matlack et al., 1999). The most direct evidence for a ratchet activity in preprotein translocation was observed by replacing vesicle-enclosed BiP with that of antibodies that recognize the preprotein used in translocation studies. Antibodies speciﬁc for the N-terminal region of the preprotein could support the import of the preprotein into the vesicles. By internalizing a combination of other antibodies that recognized various regions of the preprotein, the translocation efﬁciency was further increased. While the optimal efﬁciency of translocation was ∼2-fold less than when BiP was employed, it was argued that BiP has the added advantage that it sits at the translocation exit site and binds to the preprotein at more sites (Matlack et al., 1999). For mitochondria, the Brownian ratchet model has been supported through a number of different studies. For example, it has been demonstrated that some preproteins can oscillate while in the translocation

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

235

machineries (Ungermann et al., 1994). In addition, an import defect due to a mutation in mtHsp70 could be suppressed by the overexpression of another matrix chaperone, Hsp78. In this case, it was thought that since Hsp78 can compensate for mtHsp70 inactivation and yet does not bind to Tim44, a pulling action could be discarded (Schmitt et al., 1995; Neupert, 1997). However, such conclusions have been interpreted differently in a separate study, suggesting that Hsp78 acts at the mutant mtHsp70 and partially protects its function (Moczko et al., 1995). The ﬁnding that Tim44 forms dimers and recruits two molecules of Hsp70 to the TIM23 complex extended the ratchet model to encompass a “hand-over-hand” mode of action of mtHsp70 (Moro et al., 1999). It was suggested that the cooperative sequential binding of these bound mtHsp70s to the preprotein could increase the total efﬁciency of translocation. The evidence for pulling is based initially on the fact that the hydrolysis of matrix ATP and the interaction of mtHsp70 with Tim44 are crucial for the import of both artiﬁcial and native preproteins containing folded protein domains that require unfolding prior to their translocation. Examples include dihydrofolate reductase, barnase, and the hemebinding domain of cytochrome b2 (Eilers and Schatz, 1986; Rassow et al., 1990; Glick et al., 1993; Voos et al., 1993; Wachter et al., 1994; G¨artner et al., 1995; Matouschek et al., 1997). Interestingly, if preproteins are long enough to access Tim44-bound mtHsp70 but contain folded and stabilized cytosolic protein domains, these domains are pressed against the outer membrane so they cannot be accessible to added protease. Disruption of the Tim44–mtHsp70 interaction results in the cleavage of such domains from the rest of the preprotein that stretches into the matrix (Schwarz et al., 1993; Voisine et al., 1999; Matouschek et al., 2000). One feature that may distinguish the two models is the rate of import of preproteins containing folded protein domains. The ratchet model suggests that spontaneous protein unfolding is linked to import, while the pulling model would imply that the import rate is actually faster than spontaneous protein unfolding. In contrast to the results of Gaume and colleagues (1998), studies have indicated that unfolding of preproteins and their subsequent import can be signiﬁcantly faster than their spontaneous unfolding in solution (Glick et al., 1993; Matouschek et al., 1997; Lim et al., 2001). In a recent study, the use of rabbit reticulocyte lysate for the translation of preproteins was discarded, and puriﬁed preproteins were instead employed for import studies (Lim et al., 2001). Since cytosolic chaperones could not be acting, the involvement of mitochondria in promoting unfolding of preproteins could be more closely scrutinized than before.

236

MICHAEL T. RYAN AND NIKOLAUS PFANNER

A comparison was made between preproteins containing different extension lengths between a mitochondrial targeting signal and the folded domain of DHFR. Binding to mtHsp70 requires a stretch of at least 60 amino acids reaching from the outer membrane, across both translocation channels and into the matrix (Ungermann et al., 1994; Matouschek et al., 1997), while 50 amino acids is long enough to span the translocation machineries but not contact mtHsp70 (Rassow et al., 1990). With a preprotein containing 47 amino acids before DHFR [b2(1-47)DHFR], and hence incapable of directly accessing mtHsp70 following its initial insertion, a delay of 2–4 minutes prior to import was observed (Lim et al., 2001). This delay corresponds to the spontaneous unfolding of DHFR in solution and additionally suggests that active unfolding by the TOM machinery did not take place as has been previously observed (Mayer et al., 1995). Other preproteins that contained longer extensions and could directly contact mtHsp70 did not have such a lag phase, and their import was in fact faster that the spontaneous unfolding of DHFR. These results were in contrast to those of Gaume and colleagues (1998) and were attributed to a higher overall import rate observed under the conditions employed by Lim and coworkers (2001). In mitochondria containing a defective mtHsp70 that binds preproteins but does not efﬁciently engage with Tim44, the import of b2(1-47)-DHFR was very poor. Unfolding of the preprotein with urea largely restored this import defect. Pulling cannot operate in this system to accelerate unfolding because the preprotein could not contact mtHsp70 until it spontaneously unfolded at the mitochondrial surface. Preproteins containing longer extensions could directly contact mtHsp70 and were unfolded at a faster rate than spontaneous unfolding (Lim et al., 2001). These results are consistent with a model for Hsp70-mediated pulling of preproteins. 4. A Unified Model: Pulling and Holding Protein import studies using mitochondria containing a mutant of Tim44 that aggregated under nonpermissive conditions revealed that loosely folded preproteins could be imported but preproteins containing tightly folded domains could not (B¨omer et al., 1998). Since binding of preproteins to mtHsp70 still took place, this work demonstrated that a trapping function performed by soluble mtHsp70 could be sufﬁcient for the import of some preproteins. However, an active association between mtHsp70 and Tim44 is crucial for the import of other preproteins requiring unfolding of their tightly folded domains. A similar conclusion was made with studies employing a mutant of mtHsp70, termed Ssc1-2, that lacks a stable association with Tim44 (Gambill et al., 1993; Voos et al., 1993, 1996; Schneider et al., 1994). Ssc1-2 could, however, still bind preproteins, and in fact bound them more efﬁciently and for a

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

237

longer time than wild-type mtHsp70 (Voisine et al., 1999). Nevertheless, while this mutant Hsp70 could translocate loosely folded preproteins, it could not translocate preproteins containing folded protein domains (Voos et al., 1993; Voisine et al., 1999). Preprotein holding or trapping forms part of the ratcheting mechanism proposed for mtHsp70 (Voisine et al., 1999). However, suppressors of the mutant Hsp70 strain that restored its previously impaired binding to Tim44 also restored the unfolding of a preprotein containing the folded heme-binding domain of cytochrome b 2 into mitochondria (Voisine et al., 1999). Although binding of the soluble mtHsp70 to incoming preproteins could occur independently of Tim44 and despite enhanced trapping of the preprotein (which would enhance biased Brownian movements), it was not sufﬁcient for the translocation of preproteins containing folded cytosolic domains (Voisine et al., 1999). Thus, an interaction of mtHsp70 with Tim44 is essential for the efﬁcient import of these preproteins. These results imply that while the holding, or trapping, action of mtHsp70 can be sufﬁcient for the translocation of a group of preproteins, it does not account for the translocation of all preproteins. An additional pulling function by mtHsp70 is therefore likely. Since we do not know the folded state of many mitochondrially targeted preproteins in the cytosol, holding may in fact be the dominant action of mtHsp70 for the translocation of preproteins. In some cases, such as the import of the authentic mitochondrial preprotein, precytochrome b 2, the pulling action by mtHsp70 bound to Tim44 is necessary. Insight into how preproteins are unfolded at translocation sites has come from studies by Huang and colleagues (1999). The interaction of N-terminal targeting regions of preproteins with the translocation channels and mtHsp70 can lead to unraveling of the folded domains of proteins in a manner that is distinct from protein unfolding in solution. In fact, it has been suggested that because mitochondria can alter the unfolding pathway, this mechanism could accelerate spontaneous unfolding rates in a manner that could still be compatible with a ratcheting mechanism for mtHsp70 (Matouschek et al., 2000). However, it has also been argued that accelerated spontaneous unfolding cannot account for the unfolding of all preproteins (Matouschek et al., 2000; Lim et al., 2001). VI. OTHER ROLES FOR Hsp70s IN PROTEIN TRANSLOCATION While Hsp70s bind directly to preproteins and are essential to their translocation, it seems that a population of these molecules may be involved in less direct protein translocation activities. Brodsky and

238

MICHAEL T. RYAN AND NIKOLAUS PFANNER

colleagues (1995) suggested that at least for yeast, BiP and Sec63p are important for both co- and posttranslational import systems. Since it is thought that translation of the polypeptide by the ribosome docking tightly to the Sec61 channel is sufﬁcient for translocation in this pathway, BiP may perform a role that differs from its role in posttranslational import. Indeed, Hamman and colleagues (1998) proposed that BiP seals the Sec61 translocation channel and thereby maintains the permeability barrier by preventing the unregulated movement of ions between the cytosol and the ER lumen. A mammalian homolog of Sec63p has recently been found and is associated with the Sec61p complex that is involved in the cotranslational import of preproteins (Tyedmers et al., 2000). BiP and Sec63p have also been implicated in the translocation of tail-anchored preproteins that are inserted into the ER membrane. Signiﬁcant lengths of these polypeptides are translated prior to their targeting and ER translocation (Tyedemers et al., 2000). In mitochondria, no such seal by mtHsp70 is required because a membrane potential exists even when mtHsp70 does not interact with Tim44 (Voos et al., 1993; Rassow et al., 1994; Schneider et al., 1994; Moczko et al., 1995). However, mtHsp70 has been reported to bind to regions of the TIM23 complex independent of Tim44 (B¨omer et al., 1997). MtHsp70 may play a regulatory role at the TIM23 translocation channel in addition to its direct interaction with preproteins. A preprotein–translocase complex has been isolated in which the preprotein is stably arrested in the TIM23 complex, while mtHsp70 is absent (Dekker et al., 1997). We propose that interaction of mtHsp70 with the TIM23 complex triggers an opening of the channel such that the preprotein can slide back and forth. Binding of mtHsp70 to the preprotein then renders translocation unidirectional. VII. CONCLUSION Hsp70 proteins are essential components of the protein translocation systems of the ER and mitochondria. A general rather than a speciﬁc role of cytosolic Hsp70 is carried out in assisting preproteins in the cytosol. Improper contacts and aggregation of preproteins in the cytosol can be prevented through the binding of cytosolic Hsp70s to these preproteins and keeping them competent for translocation during their targeting to the appropriate organelle. Inside the organelles, Hsp70 has been recruited to the translocation sites in order to perform a more direct role in preprotein translocation. Depending on the association and the folded state of the preprotein, Hsp70 may perform an active role in pulling the preprotein into the organelle, while in some cases a more passive role as ratchet or trap is sufﬁcient.

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

239

The role of Hsp70s in protein translocation is not conﬁned to the ER and mitochondria. Hsp70 proteins located in the plant cell cytosol and inside the chloroplast have been implicated in protein translocation into this organelle (Vothknecht and Soll, 2000). In addition, a role for cytosolic DnaJ in the import of peroxisomal import suggests that Hsp70 is also involved in this system (Hettema et al., 1998). Future studies on the role of Hsp70s in the translocation of proteins into these other organelles may provide some general rules for Hsp70 function that can be applied to all protein translocation systems. REFERENCES Atenico, D. P., and Yaffe, M. P. (1992). Mol. Cell. Biol. 12, 283–291. Bauer, M. F., Sirrenberg, C., Neupert, W., and Brunner, M. (1996). Cell 87, 33–41. Bauer, M. F., Hofmann, S., Neupert, W., and Brunner, M. (2000). Trends Cell Biol. 10, 25–31. Berthold, J., Bauer, M. F., Schneider, H. C., Klaus, C., Dietmeier, K., Neupert, W., and Brunner, M. (1995). Cell 81, 1085–1093. Blom, J., Kubrich, ¨ M., Rassow, J., Voos, W., Dekker, P. J. T., Maarse, A., Meijer, M., and Pfanner, N. (1993). Mol. Cell. Biol. 13, 7364–7371. B¨omer, U., Meijer, M., Maarse, A. C., Dekker, P. J. T., Pfanner, N., and Rassow, J. (1997). EMBO J. 16, 2205–2216. B¨omer, U., Maarse, A. C., Martin, F., Geissler, A., Merlin, A., Sch¨onﬁsch, B., Meijer, M., Pfanner, N., and Rassow, J. (1998). EMBO J. 17, 4226–4237. Brodsky, J. L. (1996). Trends Biochem. Sci. 21, 122–126. Brodsky, J. L., and Schekman, R. (1993). J. Cell Biol. 123, 1355–1363. Brodsky, J. L., Goeckeler, J., and Schekman, R. (1995). Proc. Natl. Acad. Sci. USA 92, 9643–9646. Buchberger, A., Theyssen, H., Schr¨oder, H., McCarty, J. S., Virgalitta, G., Mikereit, P., Reinstein, J., and Bukau, B. (1995). J. Biol. Chem. 270, 16903–16910. Caplan, A. J., Cyr, D. M., and Douglas, M. G. (1992). J. Biol. Chem. 267, 18890–18895. Chen, W.-J., and Douglas, M. G. (1987). Cell 49, 651–658. Chirico, W. J., Waters, M. G., and Blobel, G. (1988). Nature 332, 805–809. Corsi, A. K., and Schekman, R. (1997). J. Cell Biol. 137, 1483–1493. Craig, E. A., Kramer, J., Shilling, J., Werner-Washburne, M., Holmes, S., Kosic-Smithers, J., and Nicolet, C. M. (1989). Mol. Cell. Biol. 9, 3000–3008. Cyr, D. M., Lu, X., and Douglas, M. G. (1992). J. Biol. Chem. 267, 20927–20931. Dekker, P. J. T., and Pfanner, N. (1997). J. Mol. Biol. 270, 321–327. Dekker, P. J. T., Martin, F., Maarse, A. C., B¨omer, U., M¨uller, H., Guiard, B., Meijer, M., Rassow, J., and Pfanner, N. (1997). EMBO J. 16, 5408–5419. Deshaies, R. J., Koch, B. D., Werner-Washburne, M., Craig, E. A., and Schekman, R. (1988). Nature 332, 800–805. Eilers, M., and Schatz, G. (1986). Nature 322, 228–232. Eilers, M., Oppliger, W., and Schatz, G. (1987). EMBO J. 6, 1073–1077. Gambill, B. D., Voos, W., Kang, P. J., Miao, B., Langer, T., Craig, E. A., and Pfanner, N. (1993). J. Cell Biol. 123, 109–117. G¨artner, F., Voos, W., Querol, A., Miller, B., Craig, E. A., Cumsky, M. G., and Pfanner, N. (1995). J. Biol. Chem. 270, 3788–3795.

240

MICHAEL T. RYAN AND NIKOLAUS PFANNER

Gaume, B., Klaus, C., Ungermann, C., Guiard, B., Neupert, W., and Brunner, M. (1998). EMBO J. 17, 6497–6507. Gething, M.-J., and Sambrook, J. (1992). Nature 355, 33–45. Glick, B. S. (1995). Cell 80, 11–14. Glick, B. S., Wachter, C., Reid, G. A., and Schatz, G. (1993). Protein Sci. 2, 1901–1917. Hachiya, N., Mihara, K., Suda, K., Horst, M., Schatz, G., and Lithgow, T. (1995). Nature 376, 705–709. Hamman, B. D., Hendershot, L. M., and Johnson, A. E. (1998). Cell 92, 747–758. Hanein, D., Matlack, K. E. S., Jungnickel, B., Plath, K., Kalies, K.-U., Miller, K. R., Rapoport, T. A., and Akey, C. W. (1996). Cell 87, 721–732. Hansen, W., Garcia, P. D., and Walter, P. (1986). Cell 45, 377–406. Herrmann, J. M., and Neupert, W. (2000). Biochim. Biophys. Acta. 1459, 331–338. Hettema, E. H., Ruigrok, C. C. M., Koerkamp, M. G., van den Berg, M., Tabak, H. F., Distel, B., and Braakman, I. (1998). J. Cell Biol. 142, 421–434. Hill, K., Model, K., Ryan, M. T., Dietmeier, K., Martin, F., Wagner, R., and Pfanner, N. (1998). Nature 395, 516–521. Horst, M., Jen¨o, P., Kronidou, N. G., Bolliger, L., Oppliger, W., Scherer, P., Manning-Krieg, U., Jascur, T., and Schatz, G. (1993). EMBO J. 12, 3035–3041. Horst, M., Oppliger, W., Feifel, B., Schatz, G., and Glick, B. S. (1996). Protein Sci. 5, 759–767. Huang, S., Ratliff, K. S., Schwartz, M. P., Spenner, J. M., and Matouschek, A. (1999). Nat. Struct. Biol. 6, 1132–1138. Hwang, S. T., Wachter, C., and Schatz, G. (1991). J. Biol. Chem. 266, 21083–21089. Jensen, R. E., and Johnson, A. E. (1999). Curr. Biol. 9, R779–782. Kang, P. J., Ostermann, J., Shilling, J., Neupert, W., Craig, E. A., and Pfanner, N. (1990). Nature 348, 137–143. Komiya, T., Sakaguchi, M., and Mihara, K. (1996). EMBO J. 15, 399–407. Krimmer, T., Rassow, J., Kunau, W.-H., Voos, W., and Pfanner, N. (2000). Mol. Cell. Biol. 20, 5879–5887. Kronidou, N. G., Oppliger, W., Bolliger, L., Hannavy, K., Glick, B. S., Schatz, G., and Horst, M. (1994). Proc. Natl. Acad. Sci. USA 91, 12818–12822. Liberek, K., Marszalek, J., Ang, D., and Georgopoulos, C. (1991). Proc. Natl. Acad. Sci. USA 88, 2874–2878. Lim, J. H., Martin, F., Guiard, B., Pfanner, N., and Voos, W. (2001). EMBO J. 20, 941–950. Lyman, S. K., and Schekman, R. (1997). Cell 88, 85–96. Maarse, A. C., Blom, J., Grivell, L. A., and Meijer, M. (1992). EMBO J. 11, 3619–3628. Martin, J., Mahlke, K., and Pfanner, N. (1991). J. Biol. Chem. 266, 18051–18057. Matlack, K. E., Misselwitz, B., Plath, K., and Rapoport, T. A. (1999). Cell 97, 553–564. Matlack, K. E. S., Mothes, W., and Rapoport, T. A. (1998). Cell 92, 381–390. Matlack, K. E., Plath, K., Misselwitz, B., and Rapoport, T. A. (1997). Science 277, 938–941. Matouschek, A., Azem, A., Ratliff, K., Glick, B. S., Schmid, K., and Schatz, G. (1997). EMBO J. 16, 6727–6736. Matouschek, A., Voos, W., and Pfanner, N. (2000). EMBO Reports 1, 404–410. Mayer, A., Neupert, W., and Lill, R. (1995). Cell 80, 127–137. Merlin, A., Voos, W., Maarse, A. C., Meijer, M., Pfanner, N., and Rassow, J. (1999). J. Cell. Biol. 145, 961–972. Miao, B., Davis, J. E., and Craig, E. A. (1997). J. Mol. Biol. 265, 541–552. Mihara, K., and Omura, T. (1996). Trends Cell Biol. 6, 104–108. Misselwitz, B., Staeck, O., and Rapoport, T. A. (1998). Mol. Cell 2, 593–603.

Hsp70 PROTEINS IN PROTEIN TRANSLOCATION

241

Misselwitz, B., Staeck, O., Matlack, K. E. S., and Rapoport, T. A. (1999). J. Biol. Chem. 274, 20110–20115. Moczko, M., Sch¨onﬁsch, B., Voos, W., Pfanner, N., and Rassow, J. (1995). J. Mol. Biol. 254, 538–543. Molinari, M., and Helenius, A. (2000). Science 288, 331–333. Moro, F., Sirrenberg, C., Schneider, H. C., Neupert, W., and Brunner, M. (1999). EMBO J. 18, 3667–3675. M¨uller, G., and Zimmerman, R. (1988). EMBO J. 7, 639–648. Murakami, H., Pain, D., and Blobel, G. (1988). J. Cell Biol. 107, 2051–2057. M¨usch, A., Wiedmann, M., and Rapoport, T. A. (1992). Cell 69, 343–352. Neupert, W. (1997). Annu. Rev. Biochem. 66, 863–917. Nguyen, T. H., Law, D. T. S., and Williams, D. B. (1991). Proc. Natl. Acad. Sci. USA 88, 1565–1569. Normington, K., Kohno, K., Kozutsumi, Y., Gething, M.-J., and Sambrook, J. (1989). Cell 57, 1223–1236. Ostermann, J., Voos, W., Kang, P. J., Craig, E. A., Neupert, W., and Pfanner, N. (1990). FEBS Lett. 277, 281–284. Panzer, S., Dreier, L., Hartmann, E., Kostka, S., and Rapoport, T. A. (1995). Cell 81, 561–570. Paunola, E., Suntio, T., Jamsa, E., and Makarow, M. (1998). Mol. Biol. Cell 9, 817–827. Pelham, H. R. B. (1986). Cell 46, 959–961. Pfanner, N., and Meijer, M. (1995). Curr. Biol. 5, 132–135. Pfanner, N., and Neupert, W. (1986). FEBS Lett. 209, 152–156. Pfanner, N., Craig, E. A., and H¨onlinger, A. (1997). Annu. Rev. Cell Dev. Biol. 13, 25–51. Pilon, M., and Schekman, R. (1999). Cell 97, 679–682. Plath, K., and Rapoport, T. A. (2000). J. Cell Biol. 151, 167–178. Rassow, J., and Pfanner, N. (1991). FEBS Lett. 293, 85–88. Rassow, J., Hartl, F.-U., Guiard, B., Pfanner, N., and Neupert, W. (1990). FEBS Lett. 275, 190–194. Rassow, J., Maarse, A. C., Krainer, E., Kubrich, ¨ M., M¨uller, H., Meijer, M., Craig, E. A., and Pfanner, N. (1994). J. Cell Biol. 127, 1547–1556. Rassow, J., Dekker, P. J. T., van Wilpe, S., Meijer, M., and Soll, J. (1999). J. Mol. Biol. 286, 105–120. Rose, M. D., Misra, L. M., and Vogel, J. P. (1989). Cell 57, 1211–1221. Rothblatt, J. A., and Meyer, D. I. (1986). Cell 44, 619–628. Ryan, M. T., and Pfanner, N. (1998). Biol. Chem. 379, 289–294. Sanders, S. L., Whitﬁeld, K. M., Vogel, J. P., Rose, M. D., and Schekman, R. W. (1992). Cell 69, 353–365. Sanz, P., and Meyer, D. I. (1989). J. Cell Biol. 108, 2102–2106. Schatz, G., and Dobberstein, B. (1996). Science 271, 1519–1526. Scherer, P. E., Krieg, U. C., Hwang, S. T., Vestweber, D., and Schatz, G. (1990). EMBO J. 9, 4315–4322. Scherer, P. E., Manning-Krieg, U. C., Jen¨o, P., Schatz, G., and Horst, M. (1992). Proc. Natl. Acad. Sci. USA 89, 11930–11934. Schleyer, M., Schmidt, B., and Neupert, W. (1982). Eur. J. Biochem. 125, 109–116. Schmitt, M., Neupert, W., and Langer, T. (1995). EMBO J. 14, 3434–3444. Schneider, H.-C., Berthold, J., Bauer, M. F., Dietmeier, K., Guiard, B., Brunner, M., and Neupert, W. (1994). Nature 371, 768–774. Schwarz, E., Seytter, T., Guiard, B., and Neupert, W. (1993). EMBO J. 12, 2295–2302.

242

MICHAEL T. RYAN AND NIKOLAUS PFANNER

Schwartz, M. P., and Matouschek, A. (1999). Proc. Natl. Acad. Sci. USA 96, 13086–13090. Scidmore, M. A., Okamura, H. H., and Rose, M. D. (1993). Mol. Biol. Cell 4, 1145–1159. Simon, S. M., Peskin, C. S., and Oster, G. F. (1992). Proc. Natl. Acad. Sci. USA 89, 3770–3774. Stuart, R. A., Gruhler, A., van der Klei, I., Guiard, B., Koll, H., and Neupert, W. (1994). Eur. J. Biochem. 220, 9–18. Terada, K., Kanazawa, M., Bukau, B., and Mori, M. (1997). J. Cell Biol. 139, 1089–1095. Terada, K., Ohtsuka, K., Imamoto, N., Yoneda, Y., and Mori, M. (1995). Mol. Cell. Biol. 15, 3708–3713. Tyedmers, J., Lerner, M., Bies, C., Dudek, J., Skowronek, M. H., Haas, I. G., Heim, N., Nastainczyk, W., Volkmer, J., and Zimmermann, R. (2000). Proc. Natl. Acad. Sci. USA 97, 7214–7219. Ungermann, C., Guiard, B., Neupert, W., and Cyr, D. M. (1996). EMBO J. 15, 735–744. Ungermann, C., Neupert, W., and Cyr, D. M. (1994). Science 266, 1250–1253. Vogel, J. P., Misra, L. M., and Rose, M. D. (1990). J. Cell Biol. 110, 1885–1895. Voisine, C., Craig, E. A., Zufall, N., von Ahsen, O., Pfanner, N., and Voos, W. (1999). Cell 97, 565–574. von Ahsen, O., Voos, W., Henninger, H., and Pfanner, N. (1995). J. Biol. Chem. 270, 29848–29853. Voos, W., Gambill, D., Guiard, B., Pfanner, N., and Craig, E. A. (1993). J. Cell Biol. 123, 119–126. Voos, W., Gambill, B. D., Laloraya, S., Ang, D., Craig, E. A., and Pfanner, N. (1994). Mol. Cell. Biol. 14, 6627–6634. Voos, W., von Ahsen, O., Muller, ¨ H., Guiard, B., Rassow, J., and Pfanner, N. (1996). EMBO J. 15, 2668–2677. Voos, W., Martin, H., Krimmer, T., and Pfanner, N. (1999). Biochim. Biophys. Acta 1422, 235–254. Vothknecht, U. C., and Soll, J. (2000). Biol. Chem. 381, 887–897. Wachter, C., Schatz, G., and Glick, B. S. (1994). Mol. Biol. Cell 5, 465–474. Walter, P., and Johnson, A. E. (1994). Annu. Rev. Cell Biol. 10, 87–119. Waters, M. G., and Blobel, G. (1986). J. Cell Biol. 102, 1543–1550. Werner-Washburne, M., Stone, D. E., and Craig, E. A. (1987). Mol. Cell. Biol. 7, 2568–2577. Wienhues, U., Becker, K., Schleyer, M., Guiard, Tropschug, M., Horwich, A. L., Pfanner, N., and Neupert, W. (1991). J. Cell Biol. 115, 1601–1609.

PROLYL ISOMERASES By FRANZ X. SCHMID Biochemisches Laboratorium, Universitat ¨ Bayreuth, D-95440 Bayreuth, Germany

I. Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Properties of Prolyl Peptide Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Prolyl Isomerizations in Protein Folding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Fast and Slow Refolding Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Impact of Prolyl Isomerization Reactions on Protein Folding . . . . . . . . . C. Multiple Prolyl Isomerizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Ribonuclease A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Ribonuclease T1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Cis/trans Isomerizations at Nonprolyl Peptide Bonds . . . . . . . . . . . . . . . . . . . . VI. Prolyl Isomerizations in Folded Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Prolyl Isomerases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Discovery and Assays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Cyclophilins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. FK506 Binding Proteins (FKBPs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Parvulins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Substrate Speciﬁcities of the Prolyl Isomerases . . . . . . . . . . . . . . . . . . . . . G. Prolyl Isomerases as Domains of Larger Proteins . . . . . . . . . . . . . . . . . . . VIII. Prolyl Isomerases as Catalysts of in Vitro Protein Folding . . . . . . . . . . . . . . . . . A. Acceleration of Proline-Limited Folding Steps . . . . . . . . . . . . . . . . . . . . . B. Catalysis of RNase T1 Folding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Autocatalytic Folding of a Prolyl Isomerase . . . . . . . . . . . . . . . . . . . . . . . . IX. The Trigger Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Catalysis of Proline-Limited Folding by the Trigger Factor . . . . . . . . . . . B. Chaperone Properties of the Trigger Factor . . . . . . . . . . . . . . . . . . . . . . . C. In Vivo Functions of the Trigger Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . X. Catalysis of Prolyl Isomerization during de Novo Protein Folding . . . . . . . . . . XI. Cellular Functions of Prolyl Isomerases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Prolyl Isomerases in the Periplasm of E. coli. . . . . . . . . . . . . . . . . . . . . . . . B. Prolyl Isomerases as Mediators of Transmembrane Signaling . . . . . . . . . C. Interaction of Cyp18 with HIV-1 Capsid Protein . . . . . . . . . . . . . . . . . . . . D. Other Functions of FKBPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Pin 1 and Phosphorylation-Dependent Prolyl Isomerization . . . . . . . . . F. Function of Immunophilins in Protection from Oxidative Stress . . . . . . G. Peptide Bond Isomerization and Ca2+ Binding . . . . . . . . . . . . . . . . . . . . XII. Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

243 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

244 244 246 246 247 248 250 250 251 253 255 256 256 257 257 258 259 259 260 261 261 263 264 264 265 265 266 267 268 268 269 270 270 271 272 273 273 274

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

244

FRANZ X. SCHMID

I. PERSPECTIVE In folded proteins the peptide bonds are usually in the trans conformation, which, for nonprolyl bonds,1 is much less strained than the energetically unfavorable cis conformation. For the peptide bonds that precede proline (prolyl bonds), however, the energy difference between the cis and trans states is small, and therefore cis prolyl bonds are found rather frequently in folded proteins. These cis prolyl bonds create a problem for protein folding. The incorrect trans forms predominate in the unfolded or nascent protein molecules, and the trans → cis isomerizations are intrinsically slow reactions because rotation about a partial double bond is required. Incorrect prolyl isomers in a protein chain strongly decelerate its folding. This is clearly seen for small single-domain proteins. Many of them refold within a few milliseconds when they contain correct prolyl isomers; but when incorrect isomers are present, folding usually requires seconds to minutes. Prolyl isomerases, such as the cyclophilins and FK506 binding proteins, are ubiquitous proteins, which are found in all organisms and all cellular compartments. In bacteria a prolyl isomerase is associated with the ribosome. The functions of prolyl isomerases in the cell are not conﬁned to accelerating protein folding. Rather, the recognition of cis and/or trans prolyl bonds and the catalysis of their cis ⇀ ↽ trans isomerization might be a common means for switching between alternative functional states of folded proteins. In fact, some prolyl isomerases can modulate the activities of growth factor receptors and ion channels; others discriminate between phosphorylated and unphosphorylated forms of Ser–Pro and Thr–Pro bonds. The functions of prolyl isomerases are discussed in a book (Galat and Riviere, 1998) and several reviews (Schmid, 1993; Schmid et al., 1993; Fischer, 1994; Galat and Metcalfe, 1995; Heitman, 1997; Schmid, 1998; Andreeva et al., 1999; Fischer and Schmid, 1999; Galat, 1999; G¨othel and Marahiel, 1999; Lu et al., 1999a; Schiene and Fischer, 2000; Balbach and Schmid, 2001). II. PROPERTIES OF PROLYL PEPTIDE BONDS The peptide bond shows considerable double bond character, and the distance between the carbonyl carbon and the nitrogen is 0.15 A˚ (15 pm) 1To facilitate reading I use the terms cis and trans proline for proline residues preceded by a cis or a trans peptide bond in the folded protein; “nativelike” and “incorrect, nonnative” denote whether or not a particular prolyl peptide bond in an unfolded state shows the same conformation as in the native state. Further, I use the expression “isomerization of Xaa” for the isomerization of the peptide bond preceding Xaa. Peptide bonds preceding proline are referred to as ‘prolyl bonds,’ and those preceding residues other than proline are termed as “nonprolyl bonds.” The folding reactions that involve Xaa–Pro isomerizations as rate-limiting steps are called “proline-limited” reactions.

PROLYL ISOMERASES

245

FIG. 1. Isomerization between the cis and trans forms of an Xaa–Pro peptide bond.

shorter than expected for a C–N single bond (Schulz and Schirmer, 1979). As a consequence, peptide bonds are planar, and the ﬂanking Cα atoms can be in either the trans or the cis conformation (equivalent to dihedral angles ω of 180◦ and 0◦ , respectively). For peptide bonds preceding residues other than proline the cis state is strongly disfavored and cis contents between 0.11 and 0.48% were found for several nonprolyl peptide bonds by NMR spectroscopy (Scherer et al., 1998). It is therefore not surprising that nonprolyl cis peptide bonds are rarely found in native, folded proteins (Stewart et al., 1990; Macarthur and Thornton, 1991). Only 49 cis nonprolyl peptide bonds were found in a survey of 747 protein structures (U. Reimer, personal communication). Interestingly, carboxypeptidase A contains three of them (Rees et al., 1983). The peptide bonds that precede proline (prolyl bonds) (Fig. 1) are much more often found in the cis conformation because here the cis and trans conformations differ only slightly in energy. In short unstructured peptides cis contents of 10–30% are frequently observed (Cheng and Bovey, 1977; Grathwohl and Wuthrich, ¨ 1981; Reimer et al., 1998). The actual cis/trans ratio depends on the size and chemical nature of the ﬂanking amino acids. In folded proteins the conformational state of a prolyl bond is usually well deﬁned, because in most cases only one of the two conformations (cis or trans) can be accommodated in the folded structure. Of 1435 nonredundant protein structures in the Brookhaven protein database, 43% contain at least one cis peptidyl–prolyl bond (Reimer et al., 1998), and 7% of all prolyl peptide bonds in folded proteins are cis (Stewart et al., 1990; Macarthur and Thornton, 1991). Prolyl cis ⇀ ↽ trans isomerizations are slow reactions with time constants between 10 and 100 s (at 25◦ C) because they involve the rotation around a partial double bond. The barrier to isomerization (80–100 kJ/mol) is enthalpic in nature; the activation entropy is almost zero. This suggests that the aqueous solvent is not reorganized in the activated state of isomerization. Amide bond isomerizations are faster in nonpolar solvents than in water (Drakenberg et al., 1972), probably because the transition state

246

FRANZ X. SCHMID

for isomerization is less polar than the ground state. In the transition state the peptide bond is twisted, and thus the resonance between the carbonyl group and the nitrogen is lost (Stein, 1993). The partial double bond character and hence the high barrier to rotation can also be diminished by N protonation. Amide bonds are preferentially protonated at the carbonyl oxygen, but the nitrogen can also be protonated by very strong acids. Prolyl isomerization is in fact well catalyzed by a solution of acetic acid in acetic anhydride (Steinberg et al., 1960) or by ≥7 M HClO4 (Schmid and Baldwin, 1978). In summary, prolyl isomerization is slow because the resonance energy of the C–N partial double bond must be overcome. The reaction is decelerated when the resonance is increased, such as in solvents that donate a hydrogen bond or a proton to the carbonyl oxygen. It is accelerated when the resonance is decreased, particularly by N protonation. The chemical and mechanistic aspects of peptide bond isomerization have been reviewed (Stein, 1993; Fischer, 2000). III. PROLYL ISOMERIZATIONS IN PROTEIN FOLDING A. Fast and Slow Refolding Species In 1973, Garel and Baldwin (1973) discovered that unfolded ribonuclease A (RNase A) consists of fast-folding UF molecules that refold in less than a second and slow-folding US molecules that require several minutes to fold to completion. Such UF and US species were subsequently detected in the folding of many other proteins (Kim and Baldwin, 1982; Schmid, 1992; Schmid et al., 1993). Two years later, Brandts and co-workers (1975) suggested that the fast- and slow-folding molecules differ in the cis/trans isomeric state of one or more Xaa–Pro peptide bonds. In their “proline hypothesis” they assumed that in the native protein (N) each prolyl peptide bond is usually in a deﬁned conformation, being either cis or trans in every molecule. After unfolding (N → UF, Scheme 1), however, these bonds become free to isomerize slowly until cis/trans equilibria as in short peptides are established (in the UF ⇀ ↽ USi reaction; cf. Scheme 1).

unfolding

N −−−−→

UF

Pro isomerization

−− ⇀ ↽ −− −− −− −− −− −− −− − − USi

SCHEME 1. Kinetic model for the coupling between protein unfolding and prolyl isomerization.

PROLYL ISOMERASES

247

A mixture of species is thus created, which consists of a single unfolded form with correct prolyl isomers (UF) and one or more unfolded species with incorrect prolyl isomers (USi ). The UF molecules with the nativelike prolyl isomers refold directly (and often quickly) to the native conformation. USi molecules, however, refold slowly, because refolding is coupled with the reisomerizations of the incorrect prolyl bonds. B. Impact of Prolyl Isomerization Reactions on Protein Folding As outlined above, proline-limited reactions often show time constants of 10–100 s at 25◦ C and activation energies of 80–100 kJ/mol. In folding, prolyl isomerization is usually coupled with conformational reactions, which can modulate the rates and activation energies. The unfolded forms U with correct and incorrect prolyl isomers give rise to parallel folding reactions. Their relative amplitudes reﬂect the relative populations of the various U species, provided that direct folding is much faster than proline-limited folding and that partially folded intermediates with incorrect prolines do not accumulate during folding. US species with more than one incorrect proline may enter alternative refolding routes, and the rank order of the reisomerizations may change with the refolding conditions. In the process of protein unfolding, ﬁrst the ordered structure breaks down in the “conformational unfolding” step (N → UF in Scheme 1), then the prolyl isomerizations take place in the unstructured protein chains (UF ⇀ ↽ UiS in Scheme 1) with kinetic properties as known from the studies of small peptides. Therefore, proline-limited reactions are usually detected in the unfolding reaction of a protein, e.g., by slow refolding assays in “double-jump” experiments. In these experiments samples are withdrawn from an unfolding protein at different time intervals after the inititation of unfolding and transferred to refolding conditions to measure the amplitudes of fast and slow refolding. These amplitudes reﬂect the relative amounts of fast- and slow-folding species present at the time of sample withdrawal (Brandts et al., 1975; Schmid, 1986). The double-jump procedure reliably identiﬁes prolyl isomerizations when conditions can be found under which the conformational and the proline-limited events are well separated in both the unfolding step and the refolding assay. It should be noted that the assay cannot be used when conformational folding (UF → N) in the second (refolding) step shows a rate similar to that of the UF ⇀ ↽ US reactions. In this case further prolyl isomerizations compete with UF → N in the refolding assays and thus inﬂuence the measured kinetics.

248

FRANZ X. SCHMID

Partially folded intermediates usually tolerate incorrect prolines; therefore, in refolding, ordered structure can already form while some prolines are still in their nonnative states. Near the end of the folding process, however, correct prolines are required, and therefore the ﬁnal steps of folding are limited in rate by the prolyl isomerizations (Cook et al., 1979; Schmid and Blaschek, 1981; Schmid, 1992; Schmid et al., 1993). The extent of conformational folding possible with incorrect prolyl isomers depends on their location in the structure and on the folding conditions. Prolyl isomerization and conformational folding are thus interdependent. Incorrect isomers in the chain can decelerate its folding, and, at the same time, rapid chain folding can affect the equilibrium and the kinetics of prolyl isomerization. Generally, incorrect prolyl bonds in exposed or ﬂexible chain regions will not interfere strongly with conformational folding, and several proteins retain conformational heterogeneity at prolines, even in the folded state (see Section VI). A simple ordered US → UF → N refolding path in which the incorrect prolines must isomerize before conformational folding can begin (as envisaged in the original proline hypothesis) is followed only under conditions where the fully folded protein is only marginally stable. Here a single incorrect kink in the protein backbone (as introduced by a nonnative prolyl bond) is sufﬁcient to destabilize partially folded structure so strongly that folding intermediates can no longer form. Structure formation in conformational folding going hand-in-hand with prolyl peptide bond isomerization is a key feature of slow folding and is of central importance for understanding the role of prolyl isomerases in these processes. The refolding of antibody fragments provides insight into the intricate interrelationship between conformational folding, prolyl isomerization, and protein association. In the folding of Fab fragments prolyl isomerization occurs after subunit association (Lilie et al., 1995), but in the folding of the CH3 domains prolyl isomerization has to occur before these domains can associate (Thies et al., 1999). C. Multiple Prolyl Isomerizations To understand the effects of many trans prolines or of a mixture of cis and trans prolines on the folding kinetics of large proteins, we must consider that cis → trans isomerizations are intrinsically about 5–10 times faster than trans → cis isomerizations and show small amplitudes only (because the trans state is favored in the unfolded molecules), and that probably not all prolines are important for the folding kinetics. Staphylococcal unclease contains six prolines, but only three of them are ratelimiting for the slow refolding reactions of this protein (Maki et al., 1999).

PROLYL ISOMERASES

249

When proteins with many prolines are unfolded, the slow-folding species (US) form more rapidly because every isomerization at an important proline leads to a slow-folding species. The rates of the individual isomerizations are thus additive. In contrast, the rate of proline-limited refolding decreases with an increasing number of incorrect prolines (Creighton, 1978) because only the species with all prolines in the nativelike isomeric state can complete folding. The rate of direct folding usually decreases with increasing size of a protein, and therefore the direct and proline-limited folding reactions are often not well separated for large proteins. This complicates the kinetic analysis of such folding reactions. Carbonic anhydrase provides a good example of a protein with many prolines. It contains 15 trans and 2 cis prolines. After long-term denaturation virtually all molecules have nonnative prolines and refold slowly (Fransson et al., 1992). More than half of the slow-folding molecules form rapidly after conformational unfolding in a reaction that is about tenfold faster than expected for a single prolyl isomerization (Kern et al., 1995). This increase in rate is probably caused by the additive and independent isomerizations of the many trans prolines of carbonic anhydrase. The remainder of the slow-folding molecules are then created slowly in a reaction that probably reﬂects the cis ⇀ ↽ trans isomerizations of one or both cis prolines. Brieﬂy denatured carbonic anhydrase molecules contain correct prolines, but, in fact, only a fraction of them can refold rapidly because the formation of molecules with incorrect isomers continues early in refolding and competes with the direct refolding reaction, which is only marginally faster than the multiple cis ⇀ ↽ trans isomerizations. Addition of the prolyl isomerase cyclophilin had an interesting effect on the refolding of the brieﬂy denatured molecules. At the onset of the experiment it catalyzed cis/trans isomerization in the unfolded molecules and thus increased the fraction of molecules with incorrect prolines, but at the same time it accelerated their slow refolding (Kern et al., 1995). A similar behavior was found for the single-chain Fv fragment of an antibody, which contains four trans and two cis prolines (J¨ager and Pl¨uckthun, 1997). At this point it should be noted that not all slow steps in protein folding are prolyl isomerizations. The very slow refolding of large proteins is often limited in rate by other events, such as slow conformational rearrangements, domain-pairing reactions, or subunit associations. An extreme example is provided by the Escherichia coli alkaline phosphatase. This protein requires days to complete folding, but, clearly, this very slow refolding reaction is not related to prolyl isomerization (Dirnbach et al., 1999).

250

FRANZ X. SCHMID

IV. EXAMPLES In this section RNase A and RNase T1 are used as examples to illustrate the role of prolyl isomerizations for the unfolding and refolding of small single-domain proteins. Bovine pancreatic RNase A is selected because the history of the proline hypothesis and its experimental veriﬁcation are closely related with this protein. The mechanism of RNase T1 folding is described because it is one of the major in vitro systems for investigating the function of prolyl isomerases as catalysts of proline-limited protein folding. A. Ribonuclease A RNase A contains four prolines. Pro42 and Pro117 are trans and Pro93 and Pro114 are cis in the native protein. Garel and Baldwin (1973) discovered that unfolded RNase A consists of fast-folding (UF) and slowfolding (US) species. Brandts and co-workers (1975) developed the slowrefolding assays to measure the UF ⇀ ↽ US equilibration and suggested the proline hypothesis on the basis of these results. The quantitative analyses of the folding kinetics of RNase A (Hagerman and Baldwin, 1976; Nall et al., 1978) demonstrated that they can be described by a US ⇀ ↽ UF ⇀ ↽ N three-state mechanism, but also provided ﬁrst evidence for a coupling between conformational folding and prolyl isomerization. The US ⇀ ↽ UF reaction in unfolded RNase A was found to be independent of denaturants (Schmid and Baldwin, 1979a) and catalyzed by a strong acid (Schmid and Baldwin, 1978), as expected for a prolinelimited process. The ﬁrst partially folded intermediate (IN) with a nonnative prolyl isomer (a trans Pro93) was also discovered for RNase A (Cook et al., 1979; Schmid and Blaschek, 1981). The conversion of IN to N in the ﬁnal slow step of folding is limited in rate by the trans → cis isomerization of Tyr92–Pro93 (Schmid et al., 1986; Sendak et al., 1996). The nativelike character of the folding intermediate IN of RNase A is reﬂected in its partial catalytic activity and in a strong protection of its amide protons from exchange with the solvent (Schmid and Baldwin, 1979b; Brems and Baldwin, 1985; Udgaonkar and Baldwin, 1988, 1990). The number of unfolded species and the isomeric states of their prolines remained unclear for a long time. The kinetic analyses of the Baldwin, Brandts, and Schmid groups (for reviews see Kim and Baldwin, 1982, 1990) had already indicated that the US species are heterogeneous and consist of two minor UIS and the major UII S species (which refolds via the nativelike intermediate IN). A reevaluation of the folding kinetics of RNase A by the Scheraga group suggested that only 6% of all

PROLYL ISOMERASES

251

unfolded molecules (the Uvf species) contain native prolyl isomers and refold very fast. A further 18% (the original UF species) refold rapidly but contain an incorrect trans Pro114 (Houry et al., 1994). UII S contains an incorrect trans Pro93. Earlier mutant work had already ascribed a crucial role to Pro93 in the folding reaction from UII S via IN to N (Schultz et al., 1992). The Scheraga group combined single and double-mixing unfolding and refolding experiments on the wild-type protein and on four Pro→Ala variants to establish a kinetic mechanism for the folding of RNase A. It includes basically all 16 possible isomeric states of the unfolded protein (Dodge et al., 1994; Dodge and Scheraga, 1996; Houry and Scheraga, 1996). B. Ribonuclease T1 RNase T1 is a small single-domain protein of 104 amino acids (Martinez-Oyanedel et al., 1991) with two disulﬁde bonds (Cys2–Cys10 and Cys6–Cys103). Like RNase A, it contains two trans (Pro60 and Pro73) and two cis (Pro39 and Pro55) prolines, but otherwise the two proteins are unrelated in their structures. RNase T1 is strongly stabilized by NaCl, and it can fold to a nativelike conformation even in the absence of its disulﬁde bonds, provided that ≥1 M NaCl is present (Oobatake et al., 1979; Pace et al., 1988; Mayr and Schmid, 1993; M¨ucke and Schmid, 1994a). As in the case of RNase A, the conformational unfolding reaction is followed by slow prolyl isomerizations in the denatured protein chains, and at equilibrium at least four different unfolded species coexist, three of which contain incorrect trans isomers of Pro39 and/or Pro55 (Kiefhaber et al., 1990a, 1992; Mayr et al., 1994; M¨ucke and Schmid, 1994a, 1994b) (Scheme 2). Since the trans form is favored over cis in unfolded proteins, the species with two incorrect (trans) isomers (U39t 55t ) predominates at equilibrium, and only 2–4% of all unfolded molecules remain in the U39c 55c state with Pro39 and Pro55 in the correct cis state. There are also 39c two species with single incorrect isomers (U39t 55c and U55t ), which are populated to approximately 10–20% each. The two trans prolines do not contribute measurably to the folding of RNase T1 (Schindler et al., 1996). The refolding of RNase T1 is not simply a reversal of its unfolding. Rather, the different unfolded species of Scheme 2 follow individual refolding paths to the native protein. The slow-folding species with one or 39c 39t two incorrect prolyl isomers (U39t 55c , U55t , and U55t ) ﬁrst rapidly regain most of their secondary structure (in the range of milliseconds), then they complete folding in slow reactions that are limited in rate by the isomerizations of the incorrect prolyl isomers. The major unfolded

252

FRANZ X. SCHMID

k43

k32

k34

k23

k21

k12

↼ −− −− −− −− −− −− −− −− − − −− ⇁

↼ −− −− −− −− −− −− −− −− − − −− ⇁

−− ⇀ −− ⇀ N39c −− −− −− −− −− −− −− −− − −U39c −− −− −− −− −− −− −− −− − −U39c 55c ↽ 55t 55c ↽

k21

k32

k12

−− ⇀ U39t −− −− −− −− −− −− −− −− − −U39t 55t 55t ↽ k23

SCHEME 2. Kinetic model for the unfolding and isomerization of RNase T1. This model is valid for unfolding only. The superscript and the subscript indicate the isomeric states of prolines 39 and 55, respectively, in the correct, nativelike cis (c) and in the incorrect, nonnative trans (t) isomeric states. For example, U39t 55c is an unfolded species with Pro39 in the incorrect trans state and Pro55 in the correct cis state. In the denatured protein the two isomerizations are independent of each other; therefore, the scheme is symmetric with identical rate constants in the horizontal and vertical directions, respectively. At 25◦ C and 6.0 M GdmCl, pH 1.6, k43 = 0.49 s−1 , k12 = 2.0 × 10−3 s−1 , k21 = 22.6 × 10−3 s−1 , k23 = 8.7 × 10−3 s−1 , k32 = 50.5 × 10−3 s−1 (Mayr et al., 1996).

species of RNase T1, which has two incorrect isomers (U39t 55t ), can enter two alternative folding pathways, and the distribution of refolding molecules on these two pathways is determined by the relative rates of the trans → cis isomerizations of Pro39 and Pro55. The proline-coupled folding mechanism of RNase T1 is discussed further in Balbach and Schmid (2001). At equilibrium only 2–4% of all unfolded RNase T1 molecules are in theU39c 55c state with the correct cis Pro39 and cis Pro55. To characterize their fast refolding reaction Mayr and colleagues (1996) used a stoppedﬂow double-mixing technique. The ﬁrst mixing was used to produce a transient high concentration of U39c 55c by a short unfolding pulse, and the second mixing was used to initiate the refolding of these molecules. In addition, the duration of the unfolding step was varied to monitor the concentrations of all species of the wild-type protein in Scheme 2 as a function of time. Conformational folding and the reisomerizations of the prolyl peptide bonds are similarly interrelated in the folding of RNase A and RNase T1. In both cases rapid partial folding is possible in the presence of nonnative proline isomers, and the prolyl isomerizations determine the rates of the ﬁnal events of folding. The partially ordered structure in the intermediates affects the kinetics of the subsequent isomerizations. In RNase T1, the trans → cis isomerization at Pro39 is strongly decelerated in a folding intermediate (Kiefhaber et al., 1990a), but in the folding

PROLYL ISOMERASES

253

of pancreatic RNase A prolyl isomerization of Pro93 is accelerated (in the folding intermediate IN). In a folding intermediate of dihydrofolate reductase, prolyl isomerization was proposed to be accelerated by intramolecular catalysis (Texter et al., 1992). Folding intermediates are less well ordered than native proteins, and they are populated only for a short time during refolding. Therefore, it has remained extremely difﬁcult to determine their structures at high resolution. For an intermediate of RNase T1 with an incorrect trans Pro39, this was achieved by combining time-resolved NMR experiments with 2D-NOESY spectroscopy (Balbach et al., 1999). These experiments were performed with the S54G/P55N variant of RNase T1 (Kiefhaber et al., 1990b), which has only one cis proline (Pro39). The folding mechanism of this variant is thus greatly simpliﬁed. Of all unfolded molecules (U39t), 85% contain a single incorrect proline (a trans Pro39) and refold in two steps on a U39t → I39t → N pathway. First the intermediate I39t, which still contains the incorrect trans Pro39, forms in less than a second, then converts slowly to the native protein N. The second reaction is limited in rate by the trans → cis isomerization of Pro39 and, at 10◦ C, shows a time constant of about 8000 s. The I39t intermediate is thus long-lived enough to apply the above-mentioned kinetic 2D-NOESY spectroscopy to determine those NH protons in I39t that are already surrounded by a native environment and thus show native distances to protons in close vicinity (Balbach et al., 1999). Surprisingly, amide protons in nonnative environments are not only located close to the incorrect trans Tyr38–Pro39 bond in I39t, but also occur in parts of the protein that are distant from Pro39 in the folded structure of I39t. Although the folding intermediate with a trans Pro39 is a nativelike species, it is much less stable than the fully folded protein with a cis Pro39. This destabilization is obviously not structurally conﬁned to the local environment of Pro39 but involves several regions throughout the entire molecule. V. Cis/trans ISOMERIZATIONS AT NONPROLYL PEPTIDE BONDS Cis/trans isomerism is not conﬁned to prolyl bonds. Cis peptide bonds to residues other than proline (cis “nonprolyl” bonds) are, however, extremely rare in folded proteins because the trans form is strongly favored over cis. In short unstructured peptides 99.5–99.9% of nonprolyl peptide bonds are in the trans state (Scherer et al., 1998). Proteins that contain nonprolyl cis peptide bonds in their native states must therefore undergo trans → cis isomerizations of these bonds in virtually all refolding molecules.

254

FRANZ X. SCHMID

For RNase T1 the cis-Pro39→Ala mutation generated a cis nonprolyl peptide bond (Mayr et al., 1994). This cis bond (between Tyr38 and Ala39) in P39A-RNase T1 reduced the protein stability by about 20 kJ/mol and caused a major change in the folding mechanism. In the unfolding of P39A-RNase T1, conformational unfolding is followed by the cis→trans isomerization of the Tyr38–Ala39 bond. This reaction shows a time constant of 730 ms and is thus about 60-fold faster than the corresponding isomerization of the Tyr38–Pro39 bond in wild-type RNase T1 (Odefey et al., 1995). This explains why the P39A mutation is so strongly destabilizing. The thermodynamic coupling between conformational unfolding and the cis ⇀ ↽ trans reaction at Ala39 shifts the overall unfolding equilibrium strongly toward the unfolded molecules with trans Ala39. Unfolded P39A-RNase T1 molecules with the Tyr38–Ala39 bond still in the nativelike cis conformation (produced by a very short 100-ms unfolding pulse) refold to the native state with a time constant of 290 ms (in 1.0 M GdmCl, pH 4.6, 25◦ C). This capacity for fast refolding is lost rapidly, and, after 3 s of unfolding, virtually all protein molecules contain an incorrect trans Tyr38–Ala39 bond. Refolding is now 1600-fold decelerated to a time constant of 480 s because Tyr38–Ala39 trans → cis reisomerization is very slow and determines the overall rate of refolding. Comparison with the folding kinetics of wild-type RNase T1 indicates that the Tyr38–Pro39 and Tyr38–Ala39 isomerizations differ predominantly in the rate of the cis → trans and not of the trans → cis reaction. TEM-1 β-lactamase from E. coli contains a cis peptide bond between Glu166 and Pro167, which is located in a large loop near the active site. It is required for the catalytic activity of β-lactamase (Strynadka et al., 1992), and its trans → cis isomerization is a slow step in refolding (Vanhove et al., 1995, 1996, 1998a, 1998b). Similar to Pro39 of RNase T1, the cis character of the 166–167 bond is retained when Pro167 of β-lactamase is replaced by a Thr residue, and the trans → cis isomerization of the Glu166–Thr167 peptide bond becomes rate-limiting for the refolding of the P167T variant (Vanhove et al., 1996). It is interesting to note that cis peptide bonds between the residues equivalent to Glu166 and Pro167 of TEM-1 β-lactamase are found in many β-lactamases, but these cis peptide bonds are not always prolyl bonds. In the β-lactamase PC1 from Staphylococcus aureus a cis Glu–Ile bond is found at this position (Herzberg, 1991). The changes in the folding kinetics of PC1 β-lactamase after the Ile167Pro mutation suggested that the trans → cis reisomerization of the Glu166–Ile167 peptide bond is in fact the slowest and thus rate-determining step in the reactivation of wild-type PC1 β-lactamase (Wheeler et al., 1998).

PROLYL ISOMERASES

255

Together, the results for RNase T1 and β-lactamase show that the trans → cis isomerizations of nonprolyl peptide bonds control the folding of proteins with such cis bonds. The trans → cis isomerizations of prolyl and nonprolyl bonds show similar rates, the reverse cis → trans isomerizations are 50–100-fold faster, however, for the nonprolyl peptide bonds. VI. PROLYL ISOMERIZATIONS IN FOLDED PROTEINS In crystal structures of folded proteins the prolyl peptide bonds are generally either cis or trans in every molecule. There is, however, an increasing number of exceptions to this rule, and cis/trans equilibria have been found, in particular by 2D-NMR spectroscopy in solution. Examples include staphylococcal nuclease (Evans et al., 1987), insulin (Higgins et al., 1988), calbindin (Chazin et al., 1989; K¨ordel et al., 1990), scorpion venom Lqh-8/6 (Adjadj et al., 1997), human interleukin-3 (Feng et al., 1997), and the TB6 domain of human ﬁbrillin-1 (Yuan et al., 1997; Yuan et al., 1998). In folded staphylococcal nuclease, cis/trans equilibria exist at Pro117 as well as at Pro47 (Evans et al., 1987; Loh et al., 1991), and the two equilibria seem to be independent of each other. For Pro117 the cis conformer can be stabilized by binding of Ca2+ ions and the inhibitor thymidine 3′ ,5′ -phosphate (Alexandrescu et al., 1990) as well as by the H124L mutation. At the same time this mutation increased the stability by about 6kJ/mol (Alexandrescu et al., 1990; Truckses et al., 1996). The K116G variant contains a predominantly trans Gly116–Pro117 bond (Hodel et al., 1993). Pro to Gly mutations at Pro117 or at both Pro47 and Pro117 in the H124L variant also led to trans bonds, increased the stability in a nonadditive fashion, and inactivated the protein (Truckses et al., 1996). Cis/trans equilibria were also detected for TB6 (at Pro2073 and Pro2103). Two conformers of this protein domain are in slow exchange on the NMR time scale, and for both the solution structure could be solved. Interestingly, in the major conformer P2073 is trans and P2103 is cis (Yuan et al., 1997); reciprocally, in the minor conformer P2073 is cis and P2103 is trans (Yuan et al., 1998). Thus, it seems that the two cis/trans equilibria are coupled, although the prolines are remote from each other and in different secondary structure elements in folded TB6. The hyperactive variant SC-55494 of human interleukin-3 (hIL-3) shows at least two major conformations (Wu and Raleigh, 1998). This heterogeneity and the activity were unchanged in variants with single

256

FRANZ X. SCHMID

mutations at the four individual proline positions. In the double mutant P30A/P31A and the quadruple mutant P30A/P31A/P33A/P37A, the conformational heterogeneity was lost and the activity 2–4-fold decreased. Lqh-8/6 is a 38-residue protein venom of the scorpion Leiurus quinquestriatus hebraeus with a minor cis-proline and a major trans-proline conformer, which interconvert slowly (Adjadj et al., 1997). cis ⇀ ↽ trans equilibria at Xaa–Pro bonds might be fairly common for proteins with dynamic conformations (which cannot be crystallized) or for ﬂexible chain regions (which are not well deﬁned in crystal structures). For obvious reasons such proteins or protein segments are strongly underrepresented in the protein structure database. As shown for TB6 and the folding intermediate of RNase T1, the effects of a prolyl cis/trans isomerization in a folded protein are not necessarily conﬁned to its local environment. Rather, it can change the protein conformation in distant regions, a prerequisite for a possible function of prolyl isomerization as a long-range slow conformational switch, which can be ﬁnetuned by prolyl isomerases. VII. PROLYL ISOMERASES A. Discovery and Assays The ﬁrst enzyme to catalyze prolyl isomerization was identiﬁed by Fischer and colleagues (1984). This discovery was possible because they developed an ingenious assay for prolyl isomerases based on the conformational speciﬁcity of chymotrypsin. This protease cleaves a chromogenic reporter group from a tetrapeptide (such as succinyl-Ala-AlaPro-Phe-4-nitroanilide) only when the Ala–Pro bond of this peptide is in the trans conformation. In aqueous solution the assay peptide exists as a 90:10 mixture of molecules with the Ala–Pro bond in trans and cis, respectively. Therefore, in the presence of a high concentration of chymotrypsin, 90% of the peptide molecules are cleaved within the dead time of manual mixing. Hydrolysis of the remaining 10% is slow because it is limited in rate by the cis → trans isomerization of the Ala–Pro bond. Acceleration of this reaction serves as a sensitive probe for prolyl isomerase activities. Kofron and colleagues (1991; 1992) improved this assay. They increased the fraction of the cis isomer (and thus the sensitivity of the assay) from 10% up to 70% by dissolving the assay peptide in an anhydrous mixture of triﬂuoroethanol and LiCl. Other isomer-speciﬁc proteases were introduced to widen the range of peptides available

PROLYL ISOMERASES

257

for the assay (Fischer, 1994). Alternative assays employ ﬂuorescing peptides, peptides with a modiﬁed amino acid incorporated prior to proline (Garcia-Echeverria et al., 1992, 1993), or NMR methods (H¨ubner et al., 1991; Kern et al., 1995). A variant of the peptide-based assay avoids the coupling with isomerspeciﬁc proteolysis ( Janowski et al., 1997). The absorbance at 330 nm of succinyl-Ala-Phe-Pro-Phe-4-nitroanilide decreases slightly when the Phe and the nitroanilide moieties rearrange on the cis → trans isomerization at the Phe–Pro bond of the peptide. Because of the small change in signal, the sensitivity of the assay is limited, but it is the method of choice for assaying prolyl isomerases sensitive to proteases. Recently, Kullertz ¨ and Fischer developed a new prolyl isomerase assay for high-troughput screening of up to 20,000 samples per day. This assay employs a disulﬁde-bonded cis peptide with internally quenched ﬂuorescence. Cis/trans isomerization is initiated by rapid reductive ring opening and followed by the increase in ﬂuorescence (G. Kullertz ¨ and G. Fischer, unpublished). B. Nomenclature There is no common nomenclature for prolyl isomerases. The systematic name peptidylproline cis,trans-isomerase (E.C. 5.2.1.8) is shortened to prolyl isomerase, proline isomerase, PPI, or PPIase. Sometimes the expression “rotamase” is used. Prolyl isomerases belong to three structurally diverse families: the cyclophilins, the FK506 binding proteins, and the parvulins. Usually, they are abbreviated to Cyp, FKBP and Parv, followed by the molecular mass and, sometimes, an indicator for the subcellular localization. Cyp18cy thus denotes the cytoplasmic cyclophilin with a molecular mass of 18 kDa. This cyclophilin is also often referred to as CypA. C. Cyclophilins In 1989 the original prolyl isomerase was found to be identical with cyclophilin (Cyp), a high-afﬁnity receptor for the immunosuppressive drug cyclosporin A (CsA) (Handschumacher et al., 1984; Fischer et al., 1989; Takahashi et al., 1989). Cyclophilins are ubiquitous proteins that are present in all organisms and all subcellular compartments. In yeast, there are seven different forms of cyclophilin, which are found in the cytosol, the mitochondria, the endoplasmic reticulum, and the nucleus. Yeast cells with disruptions in all these cyclophilins are viable, but show several growth defects (Dolinski et al., 1997). In the Drosophila melanogaster

258

FRANZ X. SCHMID

genome, 17 cyclophilin homologs were found by a BLAST search. The genome of the nematode Caenorhabditis elegans contains 11 cyclophilin genes. An alignment and comparison of 133 different cyclophilins can be found in an article by Galat (1999). The basic cyclophilin domain as present in Cyp18 consists of a central eight-stranded antiparallel β-sheet structure and two α-helices. The core structures of the human (Takahashi et al., 1989; Ke et al., 1991; Ke, 1992; Konno et al., 1996) and E. coli proteins (Konno et al., 1996) are almost identical, and 75 Cα atoms within the β-sheet and the helices can be superimposed with a root mean square (r.m.s.) deviation of ˚ Stronger deviations are found in the loops at the protein periph0.61 A. ery, caused primarily by several deletions and insertions. Structures of the enzyme complexed with cyclosporin A and with proline-containing tripeptides were determined by a combination of X-ray crystallography and two-dimensional NMR spectroscopy (Kallen et al., 1991; Kallen and Walkinshaw, 1992; Mikol et al., 1994). The substrate peptide is bound in a long groove at the surface of the β-sheet. The Ala–Pro bond of the peptide is in the cis conformation and the proline ring is inserted into a hydrophobic pocket. Residues identiﬁed by NMR to participate in CsA binding cluster at the peptide-binding site, indicating that CsA binds at the prolyl isomerase active site as a competitive inhibitor. Recently, crystal structures were determined for SnuCyp20, a component of the small nuclear ribonucleoprotein particle (Reidt et al., 2000) and for a cyclophilin from Plasmodium falciparum (Peterson et al., 2000). D. FK506 Binding Proteins (FKBPs) The FK506 binding proteins (FKBPs) are also a very large family of proteins. Like the cyclophilins, they occur as small single-domain proteins (of about 12 kDa) or as domains of large proteins and protein assemblies. The ﬁrst FKBP12 was discovered in 1989 in human T cells (Harding et al., 1989; Siekierka et al., 1989). Although unrelated in sequence to the cyclophilins, FKBP12s also bind to immunosuppressants (FK506 and rapamycin, but not CsA) and catalyze prolyl isomerization in oligopeptides (Harding et al., 1989; Siekierka et al., 1989; Kallen et al., 1991; Kallen and Walkinshaw, 1992; Spitzfaden et al., 1992; Mikol et al., 1994) and in protein folding (Tropschug et al., 1990). The FKBPs share the ubiquitous distribution with the cyclophilins, and four FKBPs are found in yeast (Dolinski et al., 1997). The structures of FKBP12s were solved by NMR (Michnick et al., 1991; Moore et al., 1991) and by X-ray crystallography (Van Duyne et al., 1993) in the free form, in complexes with the inhibitors FK506 and rapamycin (Van Duyne

PROLYL ISOMERASES

259

et al., 1991), and in a ternary complex with rapamycin and the binding domain of the FKBP12-rapamycin–associated protein FRAB (Liang et al., 1999). FKBP12 is composed of a ﬁve-stranded antiparallel β sheet, a short α-helix, and a fairly high amount of aperiodic structure (Van Duyne et al., 1993). The pipecolic amide moiety of FK506, which is thought to mimic the proline residue of the substrate, is bound in a hydrophobic pocket of FKBP, presumably at the active site (Van Duyne et al., 1993). E. Parvulins The parvulins form the third family of prolyl isomerases. With only 92 residues, E. coli parvulin (Rahfeld et al., 1994b) is one of the smallest enzymes. It shows no sequence homology with cyclophilins or FKBPs. The prolyl isomerase activity of parvulin is high, and toward the substrate succinyl-Ala-Leu-Pro-Phe-4-nitroanilide it is as active as the cyclophilins of E. coli (Rahfeld et al., 1994a, 1994b). Parvulin domains have been found in many proteins, such as PrsA from Bacillus subtilis ( Jacobs et al., 1993), SurA from E. coli (Eisenstark et al., 1992), PrtM from Lactococcus lactis (Haandrikman et al., 1989), human Pin1 (Lu et al., 1996; Ranganathan et al., 1997), and Ess1 from yeast (Hanes et al., 1989; Hani et al., 1995). In both Pin1 and Ess1, a parvulin domain is linked with a WW domain. Arabidopsis thaliana contains a Pin1 homolog without a WW domain (Landrieu et al., 2000). The structures of Pin1 (Ranganathan et al., 1997) and human parvulin 14 (Sekerina et al., 2000) have been solved. They reveal that although the parvulin and FKBP12 domains are unrelated in sequence, they show similar three-dimensional structures. Several of the proteins with parvulin domains seem to be involved in protein maturation, particularly of secreted proteins, and it has been suggested that mutations cause misfolding and defects in the function of other proteins. The SurA protein is located in the periplasm of E. coli and participates in early stages of the maturation of proteins of the outer membrane (Lazar and Kolter, 1996; Rouviere and Gross, 1996; Wild et al., 1996). SurA is also involved in the pathway that signals the accumulation of unfolded proteins from the periplasm to the cytoplasm (Missiakas and Raina, 1997). Pin1 is an essential regulator protein of the cell cycle (Hunter, 1998). The possible functions of SurA and Pin1 are further discussed in Section XI. F. Substrate Specificities of the Prolyl Isomerases The cyclophilins and the FKBPs show different substrate speciﬁcities (Harrison and Stein, 1990; Stein, 1993; Fischer, 1994). The cyclophilins

260

FRANZ X. SCHMID

are highly tolerant with regard to the nature of the amino acids that ﬂank the proline in a tetrapeptide, and the speciﬁcity constant kcat /KM varied by less than one order of magnitude when different amino acids were inserted at these positions. The FKBPs show a much higher speciﬁcity, and kcat/KM varied more than 1000-fold when the nature of the amino acid prior to proline was changed. The kcat/KM value is particularly high for hydrophobic residues at this position, such as leucine and phenylalanine (kcat /K M = 620 × 103 M−1 s−1 ) and very low for glutamate (kcat /K M = 0.6 × 103 M−1 s−1 ) (Harrison and Stein, 1990). Parvulin resembles the FKBP12 and also prefers hydrophobic residues in the position preceding proline in the assay peptides (Rahfeld et al., 1994a, 1994b). Pin1 shows a very high speciﬁcity for phosphorylated Ser–Pro sequences (Yaffe et al., 1997). The functions of Cyp18 and FKBP12 in immunosuppression do not involve their prolyl isomerase activity or the catalysis of a protein folding reaction. Rather, the inhibitors CsA and FK 506 form tight 1 : 1 complexes with Cyp18 and FKBP12, respectively, which, in turn, bind to the protein phosphatase calcineurin and thereby inhibit the transport to the nucleus of a cytosolic component of the transcription factor NF-AT in T cells. The molecular mechanisms of these immunosuppressive processes have been reviewed (Liu et al., 1991; Fischer, 1994; Galat and Metacalfe, 1995). Cytosolic Cyp18 (CypA) seems not to be essential for mammals. Mice with a disruption of both genes for CypA (ppia−/−) were viable in a robust background. A sensitive mouse strain (C57BL/6) with this disruption died soon after birth. CypA-deﬁcient mice have a normally developed immune system; they are, however, prone to allergic diseases. Interestingly, in activated T cells from these mice calcineurin-dependent signaling is tenfold less sensitive to disruption by CsA than in normal T cells, providing strong support for the mechanism of immunosuppression via calcineurin inhibition (Colgan et al., 2000) G. Prolyl Isomerases as Domains of Larger Proteins Domains that are homologous to Cyp18, FKBP12, or parvulin are found in many proteins. The NinA protein (26 kDa) from an insect eye contains, in addition to the Cyp18 domain, a hydrophobic sequence, which serves as a membrane anchor (Stamnes et al., 1991; Rutherford and Zuker, 1994). FKBP25 of Legionella pneumophila (also called the Mip protein) is localized at the outer membrane surface of this pathogenic organism. Its C-terminal domain is homologous to FKBP12, and its N-terminal domain (residues 1–106) is probably required for attaching the protein to the cell membrane (Fischer et al., 1992; Ludwig et al., 1994; Wintermeyer et al., 1995). The infectivity of L. pneumophila is strongly

PROLYL ISOMERASES

261

reduced when FK506 binds to this FKBP25. A 22-kDa FKBP, homologous to FKBP25, is found in the periplasm of E. coli (Rahfeld et al., 1996). Cyp40 and FKBP51/FKBP52 (also known as Hsp56) bind to the heat shock protein Hsp90 in glucocorticoid receptor complexes (Kieffer et al., 1992, 1993). The two proteins share a related tetratricopeptide sequence that mediates the association with other components of the receptor complex. Cyp40 (Cpr7 in yeast) and FKBP52 compete in their interaction with Hsp90. Their roles for the function of steroid receptor complexes are still unknown, but it seems unlikely that their prolyl isomerase activities are involved in folding events of the complex or its subunits (Owens-Grillo et al., 1995; Barent et al., 1998; Duina et al., 1998). The mild growth defect of yeast caused by deletion of Cpr7 is suppressed by CNS1 (cyclophilin 7 suppressor 1) overproduction (Dolinski et al., 1998). CNS1 also contains a tetratricopeptide motif and binds to HsP90, but, unlike Cpr7, it is an essential protein in yeast (Marsh et al., 1998). There is evidence that Cyp40 and FKBP52 function as molecular chaperones (Bose et al., 1996; Freeman et al., 1996). FKBP60 and FKBP65 are ER-resident proteins with four FKBP domains and a Ca2+-binding motif. They are active in the peptide assay as well as in the catalysis of folding of type III collagen (Zeng et al., 1998; Shadidy et al., 1999). SlyD of E. coli combines an amino-terminal FKBP domain with a carboxyl-terminal His-rich domain, which binds metal ions. It was discovered because it is involved in cell lysis after phage infection (Maratea et al., 1985; Roof et al., 1997), but also because it binds tightly to immobilized Ni2+ and thus copuriﬁes with His-tagged proteins (Wulﬁng ¨ et al., 1994). The prolyl isomerase activity of SlyD is regulated by Ni2+ binding to the His-rich domain (Hottenrott et al., 1997). The trigger factor of E. coli also contains a FKBP domain. This prolyl isomerase locates to the ribosome of bacteria and is an excellent catalyst of protein folding. It is described in Section IX. Additional proteins with cyclophilin- or FKBP-related domains have been described (Anderson et al., 1993; Fischer, 1994; Galat and Metcalfe, 1995; Heitman, 1997; Galat and Riviere, 1998). VIII. PROLYL ISOMERASES AS CATALYSTS OF in Vitro PROTEIN FOLDING A. Acceleration of Proline-Limited Folding Steps The ﬁrst evidence for a catalysis of protein folding by a prolyl isomerase (porcine cytoplasmic Cyp18) came from experiments with the

262

FRANZ X. SCHMID

immunoglobulin light chain, porcine ribonuclease (RNase) and the S-protein fragment of bovine RNase A (Lang et al., 1987). These folding reactions were, however, accelerated only to a modest extent, and the slow refolding of RNase and of thioredoxin could not be accelerated at all (Lin et al., 1988). It is now clear that these proteins are not substrates for prolyl isomerases, because prolyl isomerization occurs in partially folded intermediates and thus the prolyl peptide bonds are inaccessible. Meanwhile, catalysis of proline-limited steps by several prolyl isomerases was observed in the folding of many proteins, including barnase (Matouschek et al., 1990), carbonic anhydrase (Freskg˚ard et al., 1992; Kern et al., 1994), β-lactamase (A. Lejeune, unpublished results), chymotrypsin inhibitor C12 (Jackson and Fersht, 1991), yeast iso-2-cytochrome c (Veeraraghavan and Nall, 1994; Veeraraghavan et al., 1995), the immunoglobulin light chain (Lang et al., 1987; Lilie et al., 1995), staphylococcal nuclease (Veeraraghavan et al., 1997; Maki et al., 1999), and trp aporepressor (Mann et al., 1995). Cis → trans and trans → cis isomerizations are catalyzed equally well in the folding of these proteins. The extent of acceleration differs strongly in these folding reactions. It depends to some extent on the conditions and on the nature of the prolyl isomerase employed. The dominant factor is, however, the accessibility for the prolyl isomerase of the proline residues in the refolding protein molecule. The C40A/C82A/P27A variant of barstar provides an illustrative example. This protein folds via a partially folded, almost nativelike intermediate (Killick et al., 1998) with an incorrect trans Pro48. The refolding reaction, which is limited in rate by Pro48 trans → cis isomerization, is extremely well catalyzed by Cyp18 (kcat /K M = 254,000 M−1 s−1 ). Pro48 is solvent-exposed in the folded protein as well as in the folding intermediate, and therefore remains accessible for Cyp18 throughout the folding process (Golbik et al., 1999). The role of the accessibility of the prolines was also studied for iso-2-cytochrome c. In aqueous buffer the folding of this protein is barely catalyzed by cyclophilin. When, however, GdmC1 is added in increasing, but still nondenaturing concentrations, catalysis is markedly improved, presumably because the denaturant destabilizes folding intermediates and thus improves the accessibility of the prolyl bonds (Veeraraghavan et al., 1995). In the maturation of the collagen triple helix, prolyl and hydroxyprolyl isomerizations are rate-limiting steps (B¨achinger et al., 1980; Buevich et al., 2000) and accelerated by Cyp18 (B¨achinger, 1987; Davis et al., 1989). For mouse dihydrofolate reductase, the overall formation of the catalytically active protein is accelerated by cyclophilin, but it is not the

PROLYL ISOMERASES

263

interconversion between a late intermediate and the native protein in the ﬁnal step that is catalyzed by the prolyl isomerase (von Ahsen et al., 2000). B. Catalysis of RNase T1 Folding The folding of RNase T1 is determined by the two trans → cis isomerizations of Pro39 and Pro55 (see Section IV.B), and the accessibility of these two prolyl bonds is very important for the catalysis of slow refolding. Pro55 is solvent-exposed in the native protein and presumably also in folding intermediates, and therefore all prolyl isomerases catalyze the isomerization at Pro55 very well. Pro39 is buried in the native protein, and its trans → cis isomerization is only marginally accelerated because rapid conformational folding renders this proline less accessible for prolyl isomerases. Accordingly, catalysis at Pro39 is strongly improved when partially folded intermediates are destabilized by unfavorable mutations or by breaking of the two disulﬁde bonds (as in the reduced and carboxymethylated RCM form of RNase T1) (M¨ucke and Schmid, 1992). The refolding of RNase T1 is catalyzed by prolyl isomerases of all families, and, in general, their activities were found to be similar (Schmid et al., 1993). In the absence of the disulﬁde bonds (in the RCM form), RNase T1 is unfolded in aqueous buffer, but folds reversibly to a nativelike ordered conformation when ≥1 M NaCl is added (M¨ucke and Schmid, 1992). The RCM form of the S54G/P55N Variant [RCM-(-P55)RNase T1] is a particularly simple model protein for studying catalyzed protein folding because it involves a single trans → cis siomerization only at Pro39) and because the access to this proline is not impaired by premature structure formation, as in the presence of the disulﬁde bonds. Since all RCM forms are unfolded in aqueous buffer, assisted and catalyzed unfolding and refolding could be studied in the absence of denaturants simply by varying the NaCl concentration (M¨ucke and Schmid, 1994a, 1994b; Scholz et al., 1997b). This is important because several prolyl isomerases are sensitive to denaturants, such as guanidinium chloride or urea. Prolyl isomerases are enzymes. In protein folding they catalyze cis ⇀ ↽ trans isomerizations in both directions (M¨ucke and Schmid, 1992) and show equal efﬁciencies in unfolding and refolding experiments under identical conditions near the midpoint of the unfolding transition. They carry no information about the isomeric states of the prolyl peptide bonds in the protein substrates. The native isomer is selected by the refolding protein itself simply because the molecules

264

FRANZ X. SCHMID

with native prolyl isomers complete refolding rapidly and are thus no longer substrates for the isomerase. The prolyl isomerases catalyze isomerizations only at prolyl bonds and not at nonprolyl peptide bonds. The refolding of the P39A variant of RNase T1, which is limited in rate by the very slow trans → cis reisomerization of the Tyr38–Ala39 bond (see Section IV.B), is not catalyzed by cyclophilins, FKBPs, or parvulins. These enzymes are also unable to catalyze amide bond isomerizations in the proline-free model peptide Ala-Ala-Tyr-Ala-Ala (Scholz et al., 1998b). C. Autocatalytic Folding of a Prolyl Isomerase Because they are folding enzymes, prolyl isomerases should, in principle, be able to catalyze their own folding. The folding of human cytosolic FKBP12 is indeed an autocatalytic process for the mature protein and, more pronounced, for a variant with an amino-terminal extension of 16 residues (Scholz et al., 1996; Veeraraghavan et al., 1996). Native FKBP contains seven trans prolyl peptide bonds, and the cis → trans isomerizations of some or all of them determine the rate of its folding. In an autocatalytic reaction the product catalyzes its own formation; therefore, the rate should increase with reactant concentration. This was indeed observed in the folding of the FKBP12 variant with the amino-terminal extension: It was more than 10-fold accelerated when the protein concentration was increased from 0.05 to 10 μM (Scholz et al., 1996). The refolding of parvulin is also an autocatalytic process (Scholz et al., 1997b). In contrast, the refolding of E. coli cyclophilin CypA is not autocatalytic. This protein refolds in two reactions. The molecules with correct prolines refold very rapidly (with a time constant of about 1.5 ms), and the molecules with incorrect prolines refold slowly (with a time constant of 25 s) in a proline-limited reaction. This reaction, however, cannot be accelerated in an autocatalytic fashion by increasing CypA concentration. It could also not be catalyzed by adding native CypA or other prolyl isomerases (Ikura et al., 2000), probably because the prolines of CypA itself are inaccessible for catalysis during folding. IX. THE TRIGGER FACTOR The trigger factor was discovered by Crooke and Wickner in 1987 when they searched for cytosolic components involved in the export of secretory proteins from E. coli (Crooke and Wickner, 1987; Lill et al., 1988). They found that trigger factor binds to and thus stabilizes the exportcompetent form of a precursor protein, proOmpA. The trigger factor is an abundant soluble protein, which associates with the large subunit

PROLYL ISOMERASES

265

of the ribosome. The functions of the trigger factor in vivo remained unclear since the export of proOmpA appeared to function normally in an E. coli strain, in which the trigger factor was depleted to about 5% of its original level (Guthrie and Wickner, 1990). In 1995, Fischer and co-workers (Stoller et al., 1995) discovered that trigger factor is a ribosome-associated prolyl isomerase. The enzymatic activity originates from a central FKBP domain, which encompasses residues 142–251 (Callebaut and Mornon, 1995; Hesterkamp and Bukau, 1996; Stoller et al., 1996). When protein translation is arrested, nascent protein chains can be crosslinked to the trigger factor at the ribosome (Valent et al., 1995; Hesterkamp et al., 1996). The N-terminal domain of the trigger factor (residues 1–118) mediates the interaction with the ribosome (Hesterkamp et al., 1997); the function of the C-terminal domain is unknown. A. Catalysis of Proline-Limited Folding by the Trigger Factor Because it is located at the ribosome and interacts with newly formed protein chains, trigger factor might accelerate prolyl isomerizations in the de novo folding of nascent proteins. In fact, trigger factor catalyzes the in vitro folding of RCM-RNase T1 extremely well and shows a higher activity in protein folding than in assays with proline-containing tetrapeptides. Addition of only 2.5 nM trigger factor doubles the folding rate of RCM-(-P55)RNase T1, and, in the presence of 20 nM trigger factor, this folding reaction is 14-fold accelerated. This remarkable catalytic efﬁciency of the trigger factor as a folding enzyme is reﬂected in a speciﬁcity constant kcat /K M of 1.1 × 106 M−1 s−1 . This is almost 100-fold higher than the respective value for human FKBP12 (Scholz et al., 1997a). The trigger factor behaves like a classical enzyme. The initial rates of catalyzed folding show saturation behavior and obey the Michaelis– Menten equation. Trigger-factor–catalyzed folding is characterized by a KM value of 0.7 μM for the refolding protein substrate and a kcat value of 1.3 s−1 for the catalyzed folding reaction (Scholz et al., 1997a). For the catalysis by Cyp18 of the trans → cis prolyl isomerization in a tetrapeptide, KM and kcat values of 220 μM and 620 s−1, respectively, were obtained (Kern et al., 1995). This comparison indicates that the remarkable activity of the trigger factor as a folding catalyst does not originate from a high turnover number, but from a high afﬁnity for the protein substrate. B. Chaperone Properties of the Trigger Factor Permanently unfolded proteins are strong, competitive inhibitors of trigger-factor–catalyzed folding. One of those, reduced and

266

FRANZ X. SCHMID

carboxymethylated bovine α-lactalbumin (RCM-La), inhibits the trigger factor of Mycoplasma genitalium with a K I value of 50 nM. This binding of inhibitory proteins is independent of proline residues. Unfolded RCM-tendamistat (an α-amylase inhibitor with 74 residues) binds to the trigger factor with equal afﬁnity in the presence and in the absence of its three proline residues (Scholz et al., 1998a). The catalysis of folding of RCM-(-Pro55)-RNase T1 occurs at Pro39. The good inhibition by a nonfolding variant of RNase T1 that lacks Pro39 showed that this proline is dispensable for substrate binding (Scholz et al., 1998a). The isolated FKBP domain of the trigger factor is fully active as a prolyl isomerase toward a short tetrapeptide (Stoller et al., 1996), but in protein folding its activity is about 800-fold reduced; moreover, this low residual activity of the FKBP domain is no longer inhibited by RCM-La (Scholz et al., 1997a). The high enzymatic activity in protein folding requires the intact trigger factor. Activity could not be restored by fusing either the N- or the C-terminal region to the catalytic FKBP domain (Zarnt et al., 1997). Together, these results suggest that the high-afﬁnity binding site for unfolded proteins is probably distinct from the catalytic site of the trigger factor. The binding interface extends over several domains of the intact trigger factor or requires the interaction of these domains. The additional protein binding site(s) on the intact trigger factor probably decelerate the dissociation of the protein substrate from the trigger factor, and the low kcat value of 1.3 s−1 may reﬂect a change in the rate-limiting step from bond rotation (in tetrapeptide substrates) to product dissociation (in protein substrates). Additionally, it is quite likely that some of the binding events are nonproductive because the reactive prolyl peptide bonds are not positioned correctly within the prolyl isomerase site. Indeed, a lowering of both KM and kcat, as observed for the intact trigger factor, points to nonproductive binding of a substrate to an enzyme (Fersht, 1985). The trigger factor thus seems to have properties of a folding enzyme and of a chaperone. The cooperation of both functions is required for its very high catalytic efﬁciency in protein folding. Trigger factor improved the efﬁciency of renaturation of glyceraldehyde-3-phosphate dehydrogenase by acting as a chaperone in this folding reaction (Huang et al., 2000), and the overexpression of trigger factor increased the yield of functional recombinant proteins after expression in E. coli (Nishihara et al., 2000). C. In Vivo Functions of the Trigger Factor Bacillus subtilis contains only two prolyl isomerases—PPIB, which is a cytosolic cyclophilin, and trigger factor. Disruption of the corresponding genes individually or in combination had no effect on viability of

PROLYL ISOMERASES

267

B. subtilis in rich medium. In poor medium, however, particularly in the absence of amino acids, the double mutant strain grew very slowly. This indicates that prolyl isomerases become essential for growth under starvation condition, but it is still unclear how this relates with the catalytic activities of PPIB and trigger factor (G¨othel et al., 1998). In E. coli the trigger factor is nonessential for growth between 15 and 42◦ C in rich and poor medium. Similarly, mutations in DnaK, a chaperone of the Hsp70 class, do not affect growth under normal conditions. In combination, however, the inactivations of DnaK and trigger factor caused synthetic lethality. This suggests that these two proteins cooperate in vivo, possibly as chaperones in the folding of newly synthesized protein chains (Deuerling et al., 1999; Teter et al., 1999). X. CATALYSIS OF PROLYL ISOMERIZATION DURING de Novo PROTEIN FOLDING Early evidence for an involvement of prolyl isomerization and of prolyl isomerases in cellular folding came from studies on the folding of collagen. The formation of the collagen triple helix is limited in rate by successive prolyl isomerizations both in vitro and in vivo (Engel and Prockop, 1991). The maturation of collagen in chicken embryo ﬁbroblasts is retarded when CsA is present, possibly because CsA inhibits the cyclophilin-catalyzed folding of collagen in the endoplasmic reticulum (Steinmann et al., 1991). A similar effect was found for the folding of luciferase in rabbit reticulocyte lysate (Kruse et al., 1995). Proteins targeted to the mitochondrial matrix must unfold outside the mitochondria, cross the two mitochondrial membranes, then refold in the matrix. To study this refolding reaction, Rassow (Rassow et al., 1995) and Matouschek (Matouschek et al., 1995) used a synthetic precursor protein in which the presequence of subunit 9 of the Neurospora crassa F1F0-ATPase was linked to mouse cytosolic dihydrofolate reductase (Su9-DHFR). Unfolded Su9-DHFR was rapidly imported into the matrix, the presequence was cleaved off, and the DHFR moiety refolded inside the mitochondria. When measured by the resistance against proteinase K, the refolding of DHFR showed half-times of about 5 min in yeast mitochondria at 30◦ C as well as in N. crassa mitochondria at 25◦ C. In both cases folding was about 5-fold decelerated when the mitochondria had been preincubated with 2.5–5 μM CsA to inhibit the prolyl isomerase activity of mitochondrial cyclophilin. The refolding of DHFR was also retarded in mitochondria derived from yeast or N. crassa mutants that lacked a functional mitochondrial cyclophilin, and the kinetics of Su9DHFR refolding in the mutant mitochondria were almost identical to the refolding kinetics in the presence of CsA in the wild-type mitochondria.

268

FRANZ X. SCHMID

This suggests that the mitochondrial cyclophilins acted as prolyl isomerases and thereby catalyzed protein folding in organello. XI. CELLULAR FUNCTIONS OF PROLYL ISOMERASES After their discovery in 1984 (Fischer et al., 1984), a huge number of prolyl isomerases have been found in all organisms and all subcellular compartments. It is clear that not all these proteins are involved in de novo protein folding. Rather, they are assumed to participate in many cellular functions. The following section discusses several of these functions. A. Prolyl Isomerases in the Periplasm of E. coli The periplasm of E. coli contains prolyl isomerases of all three families. PPIA is a member of the cyclophilin family (Liu and Walsh, 1990; Hayano et al., 1991), FkpA or FKBP26 (Horne and Young, 1995) contains a FKBP12-like domain, and SurA (Rouviere and Gross, 1996) and PpiD (Dartigalongue and Raina, 1998) show parvulin domains. SurA is a soluble protein, whereas PpiD is attached to the inner membrane with its catalytic domain exposed to the periplasmic space. It has often been assumed that proteins of the bacterial outer membrane need assistance by chaperones and folding enzymes when they pass through the periplasm and insert into the membrane. In fact, SurA and FkpA (together with the chaperone Skp) were identiﬁed in E. coli as multicopy suppressors of defects that were caused by a deliberate overexpression of unfolded proteins in the periplasm. In addition, overexpression of surA and fkpA downregulated the sigma E pathway, which signals the unfolded-protein response from the periplasm to the cytosol (Missiakas et al., 1996). SurA is required for the proper maturation of outer membrane proteins, such as the porins LamB, OmpA, OmpC, and OmpF (Lazar and Kolter, 1996; Missiakas et al., 1996; Rouviere and Gross, 1996). The other parvulin-type prolyl isomerase of the periplasm, PpiD, was discovered as a multicopy suppressor of mutations in surA. In combination, mutations in surA and ppiD are synthetically lethal (Dartigalongue and Raina, 1998). These genetic experiments provide very strong evidence that the periplasmic isomerases are involved in protein folding and membrane insertion. Pl¨uckthun and co-workers searched for proteins that improve folding of recombinant proteins in the periplasm by coexpressing a poorly folding single-chain Fv antibody fragment in a phagemid together with a library of E. coli proteins. In this selection they identiﬁed Skp (Bothmann and Pl¨uckthun, 1998) and FkpA (Bothmann and Pl¨uckthun, 2000;

PROLYL ISOMERASES

269

Ramm and Pl¨uckthun, 2000). Overexpression of FkpA strongly improved the functional expression of several Fv fragments, even of ones without cis prolines. Although FkpA shows a high prolyl isomerase activity toward proteins in vitro, this suggests that the improvement of protein folding in the periplasm by FkpA does not necessarily require its prolyl isomerase activity (Bothmann and Pl¨uckthun, 2000; Ramm and Pl¨uckthun, 2000). Behrens and colleagues (2001) investigated SurA both in vivo and in vitro. All regions of this large protein contribute to its function in vivo. Prolyl isomerase activity could be demonstrated for one of the two parvulin domains, but, as in the case of FkpA, this activity seems to be unimportant for SurA function in vivo. SurA has a chaperone function in vitro, which requires the N-terminal part of the protein to be present. Together, these results are puzzling. On the one hand, the genetic screens repeatedly and clearly identiﬁed periplasmic prolyl isomerases as folding helpers; but on the other hand, the enzymatic activity they share seems to be dispensable for their functions. All these large proteins probably combine chaperone and prolyl isomerase functions. It is possible that their chaperone function is more important in general, or at least under the conditions used to investigate these proteins in vitro and in vivo. B. Prolyl Isomerases as Mediators of Transmembrane Signaling Both FKBP12 and Cyp18 have been identiﬁed as regulators of ion channels. The tetrameric ryanodine receptor (RyR) is the major Ca2+release channel in the membrane of the sarcoplasmic reticulum of muscle cells. Every RyR monomer binds one FKBP12 at a proline-containing hydrophobic region. FKBP12 coordinates gating of the four RyR subunits (Marx et al., 1998; Ondrias et al., 1998). Phosphorylation of the RyR subunits by protein kinase A dissociates the FKBP12 molecules and thus affects channel opening (Marx et al., 2000). Transgenic mice that lack the gene for FKBP12 show cardiac defects and altered functions of the ryanodine receptor (Shou et al., 1998). FKBP12 also binds to growth factor receptors, such as those for epidermal growth factor (EGF-R) (Lopez-Ilasaca et al., 1998) and transforming growth factor β (TGFβ-R, type I) (Wang et al., 1994, 1996). Huse and colleagues solved the crystal structure of the cytoplasmic domain of the type I TGFβ-R in complex with FKBP12 (Huse et al., 1999). In this unphosphorylated and hence inactive complex, FKBP12 bound near the regulatory GS region of the TGFβ-R domain. It is thought that the

270

FRANZ X. SCHMID

interaction of FKBP12 with EGF-R and type I TGFβ-R inhibits signaling by binding and stimulation of receptor dephosphorylation (Wang et al., 1996; Chen et al., 1997; Lopez-Ilasaca et al., 1998; Schiene and Fischer, 2000). Other experiments suggest that FKBP12 exerts its effect by suppressing TGFβ-R endocytosis (Yao et al., 2000). Disruption of FKBP12 binding to TGFβ-R by immunosuppressive drugs, however, did not affect TGFβ-signaling in primary ﬁbroblasts and thymocytes from wildtype or FKBP12-deﬁcient mice (Bassing et al., 1998). FKBP12 was also reported to bind to the inositol 1,4,5-trisphosphate receptor (Cameron et al., 1997). C. Interaction of Cyp18 with HIV-1 Capsid Protein The human cyclophilins CypA and CypB bind to the capsid protein of human immunodeﬁciency virus type 1 (HIV-1), and this interaction is required for infection of cells by HIV-1 (Franke et al., 1994; Thali et al., 1994; Braaten et al., 1996). Inhibition of CypA with CsA before infection or inside the infected cell leads to less infectious HIV-1 virions (Franke et al., 1994; Sherry et al., 1998). The crystal structures of CypA with the amino-terminal domain (Yoo et al., 1997) or a 25–amino acid fragment (Zhao et al., 1997) of the capsid protein show that Gly89 and Pro90 of the capsid protein are important for this interaction (Schutkowski et al., 1996). The Gly89–Pro90 bond is trans in these complexes (Vajdos et al., 1997; Zhao et al., 1997), in contrast to the cis conformation observed for other CypA–peptide complexes. All mutations of residues that contribute to the hydrophobic pocket, where the proline-containing peptide substrates and CsA bind, resulted in a drastically reduced CypA binding to the viral HIV-1 polyprotein Gag and inhibited incorporation into virions (Braaten et al., 1997). D. Other Functions of FKBPs FKBP12 is enriched in the brain, and the immunosuppressant FK506 promotes the functional recovery of neurons from injury. This function is also mediated by nonimmunosuppressive variants of FK506 and thus seems not to involve FKBP12/FK506/calcineurin complexes (Snyder et al., 1998). The neurotrophic action of FK506 is also present in cultured neuronal cells from FKBP12 knockout mice; therefore, it was suggested that this exciting pharmacological effect of FK506 is mediated by FKBP52 (as part of the steroid hormone receptor complex) and not by FKBP12

PROLYL ISOMERASES

271

(Gold et al., 1999). Interestingly, FKBP12 has been reported to stimulate sperm motility, an observation that certainly merits further investigation (Walensky et al., 1998) E. Pin 1 and Phosphorylation-Dependent Prolyl Isomerization The discoveries of yeast Ess1/PTF1 (Hanes et al., 1989; Hani et al., 1995) and human Pin1 (Lu et al., 1996) linked two hitherto separate ﬁelds: prolyl cis/trans isomerization and phosphorylation-dependent signal transduction. Ess1 and Pin1 are both essential proteins; they consist of a prolyl isomerase domain of the parvulin type and a WW domain. Pin1 was discovered as an interaction partner of the NIMA protein kinase, which is an essential component for mitosis in Aspergillus nidulans. This suggested strongly that Pin1 itself exerts an essential function in cell cycle control (Lu et al., 1996). In the crystal structure Pin1 displays two Arg residues (Arg68 and Arg69) at the prolyl isomerase site, which led to the suggestion that Pin1 recognizes phosphorylated peptide or protein substrates (Ranganathan et al., 1997). Indeed, shortly afterward it was shown that Pin1 binds to phosphoserine–proline (pSer– Pro) and phosphothreonine–proline (pThr–Pro) in mitotic phosphoproteins. In peptides with Ser–Pro or Thr–Pro sequences, Pin1 showed a very high speciﬁcity for the phosphorylated forms (pSer–Pro and pThr–Pro) and catalyzed their isomerizations up to 1300-fold better than those in the unphosphorylated peptides (Yaffe et al., 1997). Ser– Pro and Thr–Pro are the substrates for a multitude of proline-directed protein kinases and phosphatases, which among others, regulate the eukaryotic cell cycle. Phosphorylation strongly decreases the backbone dynamics at Ser/Thr–Pro sites. Attachment of a phosphate group at Thr– Pro reduces the rate of its isomerization eightfold (Schutkowski et al., 1998). With this high speciﬁcity for pSer/pThr–Pro, Pin1 binds to a considerable number of phosphorylated proteins in vitro (Shen et al., 1998). It is not yet clear whether all these interactions are relevant for the function of Pin1 in vivo. In the cell cycle Pin1 interacts with the protein phosphatase Cdc25 (Shen et al., 1998). In its active, phosphorylated form Cdc25 could be depleted from Xenopus eggs by binding to Pin1coated beads; and, as a consequence, Cdc2/cyclinB, a key mitotic regulator, became hyperphosphorylated and thus inactivated (Crenshaw et al., 1998). The depletion of Pin1 from Xenopus eggs led to a defect in the replication checkpoint. This defect could be reversed by adding wild-type Pin1, but not an isomerase-inactive mutant (Winkler

272

FRANZ X. SCHMID

et al., 2000). The Pin1 WW domain binds to pSer–Pro or pThr–Pro containing peptides (Lu et al., 1999b), and in intact Pin1 a doubly phosphorylated peptide derived from the C-terminal domain of RNA polymerase II (CTD) (Morris et al., 1999) was found associated with the WW domain (Verdecia et al., 2000). This binding might facilitate the catalysis of processive prolyl isomerizations in substrates with many pSer/pThr–Pro sequences (Lu et al., 1999b). Interestingly, the Arabidopsis thaliana Pin1 does not contain a WW domain (Landrieu et al., 2000). An alternative or additional explanation for the observed role of Pin1/Ess1 in cell cycle control is provided by the strong link that exists in particular between Ess1 and gene transcription. Ess1 interacts with the regulatory C-terminal domain of RNA polymerase II (CTD) in vitro and in vivo (Morris et al., 1999). CTD consists of many repeats of the heptapeptide sequence YSPTSPS, which can be heavily phosphorylated. Five out of six multicopy suppressors of ts mutants of ess1 are involved in transcription; the sixth suppressor is the major cytoplasmic cyclophilin CypA (Ar´evalo-Rodr´ıguez et al., 2000; Wu et al., 2000). F. Function of Immunophilins in Protection from Oxidative Stress Interesting links exist between cytosolic Cyp18 and protection against oxidative damage. Tropschug and co-workers (J¨aschke et al., 1998) identiﬁed a speciﬁc interaction between human Cyp18 and Aop1, which is a thiol-speciﬁc antioxidant protein from human T cells. In vitro the Aop1–Cyp18 complex was able to protect a target protein better against oxidative inactivation than Aop1 alone. Familial amyotrophic lateral sclerosis (FALS) is caused by mutations (such as V148G) in Cu/Zn superoxide dismutase-1 (SOD). Cells expressing mutant V148G-SOD are more sensitive to cell death, and this sensitivity is further enhanced in the presence of immunosuppressive as well as nonimmunosuppressive cyclosporin derivatives (Lee et al., 1999). This increased sensitivity to cell death could be reduced by overexpression of human wild-type cytosolic Cyp18, but not of an active-site mutant with diminished prolyl isomerase activity. It is possible that increased oxidative damage of proteins (as in the presence of a mutated SOD) leads to enhanced protein turnover and thus to an increased need for a prolyl isomerase, such as Cyp18. It is also possible that Cyp18 is required for the folding of SOD itself (Lee et al., 1999). Other possible roles of prolyl isomerases in stress response are described by Andreeva and colleagues (1999).

PROLYL ISOMERASES

273

G. Peptide Bond Isomerization and C a 2+ Binding C-type mammalian lectins contain a highly conserved sequence motif of a cis proline ﬂanked by two polar residues, which act as efﬁcient ligands for a Ca2+ ion, but only when the proline is in cis. In the apo form the proline is largely in trans, and the afﬁnity for Ca2+ is very low. Ng and Weis found that fast Ca2+ binding and release are both kinetically coupled with slow prolyl isomerization. Since carbohydrate binding requires the presence of bound Ca2+, the coupling with prolyl isomerization might provide a timer for downregulating afﬁnity. Several C-type lectins with carbohydrate ligands are endocytosed; the ligand and the Ca2+ are discharged at the low pH in endocytotic vesicles. The subsequent cis → trans prolyl isomerization might ensure that the receptor remains in the lowafﬁnity state for an extended time, long enough for recycling the empty lectin to the cell surface (Ng and Weis, 1998). A similar interrelation between Ca2+ binding, carbohydrate binding, and peptide bond isomerization was detected for concanavalin A, a plant lectin. In this case the cis/trans isomerization of the Ala207–Asp208 peptide bond is coupled with Ca2+ and carbohydrate binding. It is thought that Ca2+ binding to Asp10 and Tyr12 forces the intervening residue, Thr11, to drive the Ala207–Asp208 isomerization, which, in turn, brings Asp208 into its carbohydrate binding position (Bouckaert et al., 2000). A further example for a coupling between Ca2+ binding and prolyl isomerization is provided by prothrombin. Isomerization seems to be important for membrane binding of this protein (Evans and Nelsestuen, 1996). XII. CONCLUDING REMARKS Protein folding can be extremely fast, and some proteins fold to their native state within a few milliseconds. Trans ⇀ ↽ cis peptide bond isomerizations complicate the folding process and decelerate it, sometimes by more than 1000-fold. Nevertheless, cis peptide bonds occur frequently in folded proteins, mainly before proline and occassionally before other amino acid residues. Prolyl isomerization and conformational folding are coupled: Incorrect prolines lower the stability of folding intermediates and partial folding can modulate isomerization rates. Prolyl isomerases catalyze prolyl isomerizations in protein folding, provided the prolines are accessible. Cis prolines are very well suited to introduce tight turns (such as type VIβ turns) into proteins, but this cannot be the sole reason for their widespread occurrence. Evidence increases that cis/trans heterogeneity at prolyl bonds exists not only in the course of protein folding, but also

274

FRANZ X. SCHMID

in folded proteins. The structural consequences of cis ⇀ ↽ trans isomerization are not conﬁned to the local environment of the prolyl bond, but can lead to conformational changes in distant regions of the protein. Prolyl cis/trans isomerization might thus constitute a very simple molecular device for switching between different functional states of a protein at an intrisically low rate (in the time range of seconds to minutes). Prolyl isomerases could affect such switching reactions in a dual manner. They can increase the switching rate by their catalytic function, and/or they can stabilize one of the two states by differential binding to the cis or trans form. This suggests that both binding and catalysis are important for the biological function of prolyl isomerases. First evidence for a regulatory function of native-state prolyl isomerization comes from the parvulins Pin1 and Ess1, which are speciﬁc for pSer/pThr–Pro sequences and possibly link regulation by prolyl isomerization to regulation by phosphorylation/dephosphorylation. Similarly, the correlation between prolyl isomerization and Ca2+ binding in the mammalian lectins might indicate that prolyl isomerization could also be connected with Ca2+ signaling. In the future we need to elucidate which of the many interactions between prolyl isomerases and target proteins are functionally important in the cell and how prolyl isomerase domains of large proteins contribute to the overall functions of these proteins. At the molecular level the kinetic and structural consequences of native-state prolyl isomerization must be elucidated at high resolution to understand how these isomerizations can contribute to cellular regulation, and how they are modulated by prolyl isomerases and by other processes, such as phosphorylation. ACKNOWLEDGMENTS I thank J. Balbach, M. Jacob, A. Martin, R. Maier, D. Perl, M. Zeeb, and T. Oas for many discussions and their critical reading of the manuscript. This paper is dedicated to Prof. Rainer Jaenicke on the occasion of his seventieth birthday. I thank him for his continuing support and encouragement.

REFERENCES Adjadj, E., Naudat, V., Quiniou, E., Wouters, D., Sautiere, P., and Craescu, C. T. (1997). Eur. J. Biochem. 246, 218–227. Alexandrescu, A. T., Hinck, A. P., and Markley, J. L. (1990). Biochemistry 29, 4516–4525. Anderson, S. K., Gallinger, S., Roder, J., Frey, J., Young, H. A., and Ortaldo, J. R. (1993). Proc. Natl. Acad. Sci. USA 90, 542–546. Andreeva, L., Heads, R., and Green, C. J. (1999). Int. J. Exp. Pathol. 80, 305–315. Ar´evalo-Rodr´ıguez, M., Cardenas, M. E., Wu, X. Y., Hanes, S. D., and Heitman, J. (2000). EMBO J. 19, 3739–3749. B¨achinger, H.-P. (1987). J. Biol. Chem. 262, 17144–17148.

PROLYL ISOMERASES

275

B¨achinger, H. P., Bruckner, P., Timpl, R., Prockop, D. J., and Engel, J. (1980). Eur. J. Biochem. 106, 619–632. Balbach, J., and Schmid, F. X. (2000). In Protein Folding: Frontiers in Molecular Biology (R. H. Pain, ed.), pp. 212–249. Oxford University Press, Oxford. Balbach, J., Steegborn, C., Schindler, T., and Schmid, F. X. (1999). J. Mol. Biol. 285, 829–842. Barent, R. L., Nair, S. C., Carr, D. C., Ruan, Y., Rimerman, R. A., Fulton, J., Zhang, Y., and Smith, D. F. (1998). Mol. Endocrinol. 12, 342–354. Bassing, C. H., Shou, W., Muir, S., Heitman, J., Matzuk, M. M., and Wang, X. F. (1998). Cell Growth Differ. 9, 223–228. Behrens, S., Maier, R., de Cock, H., Schmid, F. X., and Gross, C. A. (2001). EMBO J. 20, 285–294. Bose, S., Weikl, T., B¨ugl, H., and Buchner, J. (1996). Science 274, 1715–1717. Bothmann, H., and Pl¨uckthun, A. (1998). Nat. Biotechnol. 16, 376–380. Bothmann, H., and Pl¨uckthun, A. (2000). J. Biol. Chem. 275, 17100–17105. Bouckaert, J., Dewallef, Y., Poortmans, F., Wyns, L., and Loris, R. (2000). J. Biol. Chem. 275, 19778–19787. Braaten, D., Ansari, H., and Luban, J. (1997). J. Virol. 71, 2107–2113. Braaten, D., Franke, E. K., and Luban, J. (1996). J. Virol. 70, 3551–3560. Brandts, J. F., Halvorson, H. R., and Brennan, M. (1975). Biochemistry 14, 4953–4963. Brems, D. N., and Baldwin, R. L. (1985). Biochemistry 24, 1689–1693. Buevich, A. V., Dai, Q. H., Liu, X. Y., Brodsky, B., and Baum, J. (2000). Biochemistry 39, 4299–4308. Callebaut, I., and Mornon, J. P. (1995). FEBS Lett. 374, 211–215. Cameron, A. M., Nucifora, F. C., Fung, E. T., Livingston, D. J., Aldape, R. A., Ross, C. A., and Snyder, S. H. (1997). J. Biol. Chem. 272, 27582–27588. Chazin, W. J., K¨ordel, J., Drakenberg, T., Thulin, E., Brodin, P., Grundstr¨om, T., and Fors´en, S. (1989). Proc. Natl. Acad. Sci. USA 86, 2195–2198. Chen, Y. G., Liu, F., and Massague, J. (1997). EMBO J. 16, 3866–3876. Cheng, H. N., and Bovey, F. A. (1977). Biopolymers 16, 1465–1472. Colgan, J., Asmal, M., and Luban, J. (2000). Genomics 68, 167–178. Cook, K. H., Schmid, F. X., and Baldwin, R. L. (1979). Proc. Natl. Acad. Sci. USA 76, 6157–6161. Creighton, T. E. (1978). J. Mol. Biol. 125, 401–406. Crenshaw, D. G., Yang, J., Means, A. R., and Kornbluth, S. (1998). EMBO J. 17, 1315–1327. Crooke, E., and Wickner, W. (1987). Proc. Natl. Acad. Sci. USA 84, 5216–5220. Dartigalongue, C., and Raina, S. (1998). EMBO J. 17, 3968–3980. Davis, J. M., Boswell, B. A., and B¨achinger, H. P. (1989). J. Biol. Chem. 264, 8956–8962. Deuerling, E., Schulze-Specking, A., Tomoyasu, T., Mogk, A., and Bukau, B. (1999). Nature 400, 693–696. Dirnbach, E., Steel, D. G., and Gafni, A. (1999). J. Biol. Chem. 274, 4532–4536. Dodge, R. W., Laity, J. H., Rothwarf, D. M., Shimotakahara, S., and Scheraga, H. A. (1994). J. Protein Chem. 13, 409–421. Dodge, R. W., and Scheraga, H. A. (1996). Biochemistry 35, 1548–1559. Dolinski, K., Muir, S., Cardenas, M., and Heitman, J. (1997). Proc. Natl. Acad. Sci. USA 94, 13093–13098. Dolinski, K. J., Cardenas, M. E., and Heitman, J. (1998). Mol. Cell. Biol. 18, 7344–7352. Drakenberg, T., Dahlqvist, K.-I., and Fors´en, S. (1972). J. Phys. Chem. 76, 2178–2183. Duina, A. A., Marsh, J. A., Kurtz, R. B., Chang, H. C. J., Lindquist, S., and Gaber, R. F. (1998). J. Biol. Chem. 273, 10819–10822.

276

FRANZ X. SCHMID

Eisenstark, A., Miller, C., Jones, J., and Leven, S. (1992). Biochem. Biophys. Res. Commun. 188, 1054–1059. Engel, J., and Prockop, D. J. (1991). Annu. Rev. Biophys. Biomol. Struct. 20, 137–152. Evans, P. A., Dobson, C. M., Kautz, R. A., Hatfull, G., and Fox, R. O. (1987). Nature 329, 266–268. Evans, T. C., Jr., and Nelsestuen, G. L. (1996). Biochemistry 35, 8210–8215. Feng, Y., Hood, W. F., Forgey, R. W., Abegg, A. L., Caparon, M. H., Thiele, B. R., Leimgruber, R. M., and McWherter, C. A. (1997). Protein Sci. 6, 1777–1782. Fersht, A. (1985). Enzyme Structure and Mechanism. W. H. Freeman, New York. Fischer, G. (1994). Angew. Chem. Int. Ed. 33, 1415–1436. Fischer, G. (2000). Chem. Soc. Rev. 29, 119–127. Fischer, G., Bang, H., Ludwig, B., Mann, K., and Hacker, J. (1992). Mol. Microbiol. 6, 1375–1383. Fischer, G., Bang, H., and Mech, C. (1984). Biomed. Biochim. Acta 43, 1101–1111. Fischer, G., and Schmid, F. X. (1999). In Molecular Biology of Chaperones and Folding Catalysts (B. Bukau, ed.), pp. 461–489. Harwood Academic Publishers, New York. Fischer, G., Wittmann-Liebold, B., Lang, K., Kiefhaber, T., and Schmid, F. X. (1989). Nature 337, 476–478. Franke, E. K., Yuan, H. E. H., and Luban, J. (1994). Nature 372, 359–362. Fransson, C., Freskgard, P. O., Herbertsson, H., Johansson, A., Jonasson, P., Martensson, L. G., Svensson, M., Jonsson, B. H., and Carlsson, U. (1992). FEBS Lett. 296, 90–94. Freeman, B. C., Toft, D. O., and Morimoto, R. I. (1996). Science 274, 1718–1720. Freskg˚ard, P. O., Bergenhem, N., Jonsson, B. H., Svensson, M., and Carlsson, U. (1992). Science 258, 466–468. Galat, A. (1999). Arch. Biochem. Biophys. 371, 149–162. Galat, A., and Metcalfe, S. M. (1995). Prog. Biophys. Molec. Biol. 63, 67–118. Galat, A., and Riviere, S. (1998). Peptidyl-Prolyl cis/trans Isomerases. Oxford University Press, Oxford. Garcia-Echeverria, C., Kofron, J. L., Kuzmic, P., Kishore, V., and Rich, D. H. (1992). J. Am. Chem. Soc. 114, 2758–2759. Garcia-Echeverria, C., Kofron, J. L., Kuzmic, P., and Rich, D. H. (1993). Biochem. Biophys. Res. Commun. 191, 70–75. Garel, J. R., and Baldwin, R. L. (1973). Proc. Natl. Acad. Sci. USA 70, 3347–3351. Golbik, R., Fischer, G., and Fersht, A. R. (1999). Protein Sci. 8, 1505–1514. Gold, B. G., Densmore, V., Shou, W., Matzuk, M. M., and Gordon, H. S. (1999). J. Pharmacol. Exp. Ther. 289, 1202–1210. G¨othel, S. F., and Marahiel, M. A. (1999). Cell. Molec. Life Sci. 55, 423–436. G¨othel, S. F., Scholz, C., Schmid, F. X., and Marahiel, M. A. (1998). Biochemistry 37, 13392–13399. Grathwohl, C., and Wuthrich, ¨ K. (1981). Biopolymers 20, 2623–2633. Guthrie, B., and Wickner, W. (1990). J. Bacteriol. 172, 5555–5562. Haandrikman, A. J., Kok, J., Laan, H., Soemitro, S., Ledeboer, A. M., Konings, W. N., and Venema, G. (1989). J. Bacteriol. 171, 2789–2794. Hagerman, P. J., and Baldwin, R. L. (1976). Biochemistry 15, 1462–1473. Handschumacher, R. E., Harding, M. W., Rice, J., and Drugge, R. J. (1984). Science 226, 544–547. Hanes, S. D., Shank, P. R., and Bostian, K. A. (1989). Yeast 5, 55–72. Hani, J., Stumpf, G., and Domdey, H. (1995). FEBS Lett. 365, 198–202. Harding, M. W., Galat, A., Ueling, D. E., and Schreiber, S. L. (1989). Nature 341, 758–760. Harrison, R. K., and Stein, R. L. (1990). Biochemistry 29, 3813–3816.

PROLYL ISOMERASES

277

Hayano, T., Takahashi, N., Kato, S., Maki, N., and Suzuki, M. (1991). Biochemistry 30, 3041–3048. Heitman, J. (1997). In Guidebook to Molecular Chaperones and Protein-Folding Catalysts, (M. J. Gething, ed.), pp. 359–367. Oxford University Press, Oxford, UK. Herzberg, O. (1991). J. Mol. Biol. 217, 701–719. Hesterkamp, T., and Bukau, B. (1996). FEBS Lett. 385, 67–71. Hesterkamp, T., Deuerling, E., and Bukau, B. (1997). J. Biol. Chem. 272, 21865–21871. Hesterkamp, T., Hauser, S., Lutcke, ¨ H., and Bukau, B. (1996). Proc. Natl. Acad. Sci. USA 93, 4437–4441. Higgins, K. A., Craik, D. J., Hall, J. G., and Andrews, P. R. (1988). Drug Design and Delivery 3, 159–170. Hodel, A., Kautz, R. A., Jacobs, M. D., and Fox, R. O. (1993). Protein Sci. 2, 838–850. Horne, S. M., and Young, K. D. (1995). Arch. Microbiol. 163, 357–365. Hottenrott, S., Schumann, T., Pl¨uckthun, A., Fischer, G., and Rahfeld, J. U. (1997). J. Biol. Chem. 272, 15697–15701. Houry, W. A., Rothwarf, D. M., and Scheraga, H. A. (1994). Biochemistry 33, 2516–2530. Houry, W. A., and Scheraga, H. A. (1996). Biochemistry 35, 11719–11733. Huang, G. C., Li, Z. Y., Zhou, J. M., and Fischer, G. (2000). Protein Sci. 9, 1254–1261. Hu¨ bner, D., Drakenberg, T., Fors´en, S., and Fischer, G. (1991). FEBS Lett. 284, 79–81. Hunter, T. (1998). Cell 92, 141–143. Huse, M., Chen, Y. G., Massague, J., and Kuriyan, J. (1999). Cell 96, 425–436. Ikura, T., Hayano, T., Takahashi, N., and Kuwajima, K. (2000). J. Mol. Biol. 297, 791–802. Jackson, S. E., and Fersht, A. R. (1991). Biochemistry 30, 10436–10443. Jacobs, M., Andersen, J. B., Kontinen, V., and Sarvas, M. (1993). Mol. Microbiol. 8, 957–966. J¨ager, M., and Pl¨uckthun, A. (1997). FEBS Lett. 418, 106–110. Janowski, B., W¨ollner, S., Schutkowski, M., and Fischer, G. (1997). Anal. Biochem. 252, 293–307. J¨aschke, A., Mi, H., and Tropschug, M. (1998). J. Mol. Biol. 277, 763–769. Kallen, J., Spitzfaden, C., Zurini, M. G. M., Wider, G., Widmer, H., Wuthrich, ¨ K., and Walkinshaw, M. D. (1991). Nature 353, 276–279. Kallen, J., and Walkinshaw, M. D. (1992). FEBS Lett. 300, 286–290. Ke, H. M. (1992). J. Mol. Biol. 228, 539–550. Ke, H. M., Zydowsky, L. D., Liu, J., and Walsh, C. T. (1991). Proc. Natl. Acad. Sci. USA 88, 9483–9487. Kern, D., Kern, G., Scherer, G., Fischer, G., and Drakenberg, T. (1995). Biochemistry 34, 13594–13602. Kern, G., Kern, D., Schmid, F. X., and Fischer, G. (1994). FEBS Lett. 348, 145–148. Kieffer, L. J., Seng, T. W., Li, W., Osterman, D. G., Handschumacher, R. E., and Bayney, R. M. (1993). J. Biol. Chem. 268, 12303–12310. Kieffer, L. J., Thalhammer, T., and Handschumacher, R. E. (1992). J. Biol. Chem. 267, 5503–5507. Kiefhaber, T., Grunert, H. P., Hahn, U., and Schmid, F. X. (1990b). Biochemistry 29, 6475– 6480. Kiefhaber, T., Grunert, H. P., Hahn, U., and Schmid, F. X. (1992). Proteins: Struct. Funct. Genet. 12, 171–179. Kiefhaber, T., Quaas, R., Hahn, U., and Schmid, F. X. (1990a). Biochemistry 29, 3061– 3070. Killick, T. R., Freund, S. M. V., and Fersht, A. R. (1998). FEBS Lett. 423, 110–112. Kim, P. S., and Baldwin, R. L. (1982). Annu. Rev. Biochem. 51, 459–489. Kim, P. S., and Baldwin, R. L. (1990). Annu. Rev. Biochem. 59, 631–660.

278

FRANZ X. SCHMID

Kofron, J. L., Kuzmic, P., Kishore, V., Colonbonilla, E., and Rich, D. H. (1991). Biochemistry 30, 6127–6134. Kofron, J. L., Kuzmic, P., Kishore, V., Gemmecker, G., Fesik, S. W., and Rich, D. H. (1992). J. Am. Chem. Soc. 114, 2670–2675. Konno, M., Ito, M., Hayano, T., and Takahashi, N. (1996). J. Mol. Biol. 256, 897–908. K¨ordel, J., Forsen, S., Drakenberg, T., and Chazin, W. J. (1990). Biochemistry 29, 4400– 4409. Kruse, M., Brunke, M., Escher, A., Szalay, A. A., Tropschug, M., and Zimmermann, R. (1995). J. Biol. Chem. 270, 2588–2594. Landrieu, I., De Veylder, L., Fruchart, J. S., Odaert, B., Casteels, P., Portetelle, D., Van Montagu, M., Inze, D., and Lippens, G. (2000). J. Biol. Chem. 275, 10577–10581. Lang, K., Schmid, F. X., and Fischer, G. (1987). Nature 329, 268–270. Lazar, S. W., and Kolter, R. (1996). J. Bacteriol. 178, 1770–1773. Lee, J. P., Palfrey, H. C., Bindokas, V. P., Ghadge, G. D., Ma, L., Miller, R. J., and Roos, R. P. (1999). Proc. Natl. Acad. Sci. USA 96, 3251–3256. Liang, J., Choi, J., and Clardy, J. (1999). Acta Crystallogr. Sect. D, Biol. Crystallogr. 55, 736– 744. Lilie, H., Rudolph, R., and Buchner, J. (1995). J. Mol. Biol. 248, 190–201. Lill, R., Crooke, E., Guthrie, B., and Wickner, W. (1988). Cell 54, 1013–1018. Lin, L.-N., Hasumi, H., and Brandts, J. F. (1988). Biochim. Biophys. Acta 956, 256–266. Liu, J., Farmer, J. D., Lane, W. S., Friedman, J., Weissman, I., and Schreiber, S. L. (1991). Cell 66, 807–815. Liu, J., and Walsh, C. T. (1990). Proc. Natl. Acad. Sci. USA 87, 4028–4032. Loh, S. N., Mcnemar, C. W., and Markley, J. L. (1991). Tech. Protein Chem. II, 275–282. Lopez-Ilasaca, M., Schiene, C., Kullertz, ¨ G., Tradler, T., Fischer, G., and Wetzker, R. (1998). J. Biol. Chem. 273, 9430–9434. Lu, K. P., Hanes, S. D., and Hunter, T. (1996). Nature 380, 544–547. Lu, P. J., Wulf, G., Zhou, X. Z., Davies, P., and Lu, K. P. (1999a). Nature 399, 784–788. Lu, P. J., Zhou, X. Z., Shen, M., and Lu, K. P. (1999b). Science 283, 1325–1328. Ludwig, B., Rahfeld, J., Schmidt, B., Mann, K., Wintermeyer, E., Fischer, G., and Hacker, J. (1994). FEMS Microbiol. Lett. 118, 23–30. Macarthur, M. W., and Thornton, J. M. (1991). J. Mol. Biol. 218, 397–412. Maki, K., Ikura, T., Hayano, T., Takahashi, N., and Kuwajima, K. (1999). Biochemistry 38, 2213–2223. Mann, C. J., Shao, X., and Matthews, C. R. (1995). Biochemistry 34, 14573–14580. Maratea, D., Young, K., and Young, R. (1985). Gene 40, 39–46. Marsh, J. A., Kalton, H. M., and Gaber, R. F. (1998). Mol. Cell. Biol. 18, 7353–7359. Martinez-Oyanedel, J., Choe, H.-W., Heinemann., U., and Saenger, W. (1991). J. Mol. Biol. 222, 335–352. Marx, S. O., Ondrias, K., and Marks, A. R. (1998). Science 281, 818–821. Marx, S. O., Reiken, S., Hisamatsu, Y., Jayaraman, T., Burkhoff, D., Rosemblit, N., and Marks, A. R. (2000). Cell 101, 365–376. Matouschek, A., Kellis, J. T., Serrano, L., Bycroft, M., and Fersht, A. R. (1990). Nature 346, 440–445. Matouschek, A., Rospert, S., Schmid, K., Glick, B. S., and Schatz, G. (1995). Proc. Natl. Acad. Sci. USA 92, 6319–6323. Mayr, L. M., Odefey, C., Schutkowski, M., and Schmid, F. X. (1996). Biochemistry 35, 5550– 5561. Mayr, L. M., and Schmid, F. X. (1993). Biochemistry 32, 7994–7998. Mayr, L. M., Willbold, D., R¨osch, P., and Schmid, F. X. (1994). J. Mol. Biol. 240, 288–293.

PROLYL ISOMERASES

279

Michnick, S. W., Rosen, M. K., Wandless, T. J., Karplus, M., and Schreiber, S. L. (1991). Science 252, 836–842. Mikol, V., Kallen, J., and Walkinshaw, M. D. (1994). Proc. Natl. Acad. Sci. USA 91, 5183– 5186. Missiakas, D., Betton, J. M., and Raina, S. (1996). Mol. Microbiol. 21, 871–884. Missiakas, D., and Raina, S. (1997). Trends Biochem. Sci. 22, 59–63. Moore, J. M., Peattie, D. A., Fitzgibbon, J., and Thomson, J. A. (1991). Nature 351, 248– 250. Morris, D. P., Phatnani, H. P., and Greenleaf, A. L. (1999). J. Biol. Chem. 274, 31583–31587. Mucke, ¨ M., and Schmid, F. X. (1992). Biochemistry 31, 7848–7854. Mucke, ¨ M., and Schmid, F. X. (1994a). Biochemistry 33, 14608–14619. Mucke, ¨ M., and Schmid, F. X. (1994b). J. Mol. Biol. 239, 713–725. Nall, B. T., Garel, J.-R., and Baldwin, R. L. (1978). J. Mol. Biol. 118, 317–330. Ng, K. K. S., and Weis, W. I. (1998). Biochemistry 37, 17977–17989. Nishihara, K., Kanemori, M., Yanagi, H., and Yura, T. (2000). Appl. Environ. Microbiol. 66, 884–889. Odefey, C., Mayr, L. M., and Schmid, F. X. (1995). J. Mol. Biol. 245, 69–78. Ondrias, K., Marx, S. O., Gaburjakova, M., and Marks, A. R. (1998). Ann. NY Acad. Sci. 853, 149–156. Oobatake, M., Takahashi, S., and Ooi, T. (1979). J. Biochem. 86, 55–63. Owens-Grillo, J. K., Hoffmann, K., Hutchison, K. A., Yem, A. W., Deibel, M. R., Handschumacher, R. E., and Pratt, W. B. (1995). J. Biol. Chem. 270, 20479–20484. Pace, C. N., Grimsley, G. R., Thomson, J. A., and Barnett, B. J. (1988). J. Biol. Chem. 263, 11820–11825. Peterson, M. R., Hall, D. R., Berriman, M., Nunes, J. A., Leonard, G. A., Fairlamb, A. H., and Hunter, W. N. (2000). J. Mol. Biol. 298, 123–133. Rahfeld, J. U., Rucknagel, K. P., Schelbert, B., Ludwig, B., Hacker, J., Mann, K., and Fischer, G. (1994a). FEBS Lett. 352, 180–184. Rahfeld, J. U., Rucknagel, ¨ K. P., Stoller, G., Horne, S. M., Schierhorn, A., Young, K. D., and Fischer, G. (1996). J. Biol. Chem. 271, 22130–22138. Rahfeld, J.-U., Schierhorn, A., Mann, K.-H., and Fischer, G. (1994b). FEBS Lett. 343, 65–69. Ramm, K., and Pl¨uckthun, A. (2000). J. Biol. Chem. 275, 17106–17113. Ranganathan, R., Lu, K. P., Hunter, T., and Noel, J. P. (1997). Cell 89, 875–886. Rassow, J., Mohrs, K., Koidl, S., Barthelmess, I. B., Pfanner, N., and Tropschug, M. (1995). Mol. Cell Biol. 15, 2654–2662. Rees, D. C., Lewis, M., and Lipscomb, W. N. (1983). J. Mol. Biol. 168, 367–387. Reidt, U., Reuter, K., Achsel, T., Ingelﬁnger, D., Luhrmann, R., and Ficner, R. (2000). J. Biol. Chem. 275, 7439–7442. Reimer, U., Scherer, G., Drewello, M., Kruber, S., Schutkowski, M., and Fischer, G. (1998). J. Mol. Biol. 279, 449–460. Roof, W. D., Fang, H. Q., Young, K. D., Sun, J. L., and Young, R. (1997). Molec. Microbiol. 25, 1031–1046. Rouviere, P. E., and Gross, C. A. (1996). Gene Develop. 10, 3170–3182. Rutherford, S. L., and Zuker, C. S. (1994). Cell 79, 1129–1132. Scherer, G., Kramer, M. L., Schutkowski, M. U. R., and Fischer, G. (1998). J. Am. Chem. Soc. 120, 5568–5574. Schiene, C., and Fischer, G. (2000). Curr. Opin. Struct. Biol. 10, 40–45. Schindler, T., Mayr, L. M., Landt, O., Hahn, U., and Schmid, F. X. (1996). Eur. J. Biochem. 241, 516–524.

280

FRANZ X. SCHMID

Schmid, F. X. (1986). In Enzyme Structure Part L ( C. H. W. Hirs and S. N. Timasheff, eds.), pp. 71–82. Academic Press, New York. Schmid, F. X. (1992). In Protein Folding (T. E. Creighton, ed.), pp. 197–241. Freeman, New York. Schmid, F. X. (1993). Annu. Rev. Biophys. Biomol. Struct. 22, 123–143. Schmid, F. X. (1998). In Molecular Chaperones in the Life Cycle of Proteins (A. L. Fink and Y. Goto, eds.), pp. 361–389. Marcel Dekker, New York. Schmid, F. X., and Baldwin, R. L. (1978). Proc. Natl. Acad. Sci. USA 75, 4764–4768. Schmid, F. X., and Baldwin, R. L. (1979a). J. Mol. Biol. 133, 285–287. Schmid, F. X., and Baldwin, R. L. (1979b). J. Mol. Biol. 135, 199–215. Schmid, F. X., and Blaschek, H. (1981). Eur. J. Biochem. 114, 111–117. Schmid, F. X., Graﬂ, R., Wrba, A., and Beintema, J. J. (1986). Proc. Natl. Acad. Sci. USA 83, 872–876. Schmid, F. X., Mayr, L. M., Mucke, ¨ M., and Sch¨onbrunner, E. R. (1993). Adv. Protein Chem. 44, 25–66. Scholz, C., M¨ucke, M., Rape, M., Pecht, A., Pahl, A., Bang, H., and Schmid, F. X. (1998a). J. Mol. Biol. 277, 723–732. Scholz, C., Rahfeld, J., Fischer, G., and Schmid, F. X. (1997b). J. Mol. Biol. 273, 752–762. Scholz, C., Scherer, G., Mayr, L. M., Schindler, T., Fischer, G., and Schmid, F. X. (1998b). Biol. Chem. 379, 361–365. Scholz, C., Stoller, G., Zarnt, T., Fischer, G., and Schmid, F. X. (1997a). EMBO J. 16, 54–58. Scholz, C., Zarnt, T., Kern, G., Lang, K., Burtscher, H., Fischer, G., and Schmid, F. X. (1996). J. Biol. Chem. 271, 12703–12707. Schultz, D. A., Schmid, F. X., and Baldwin, R. L. (1992). Protein Sci. 1, 917–924. Schulz, G. E., and Schirmer, R. E. (1979). Principles of Protein Structure. Springer Verlag, New York. Schutkowski, M., Bernhardt, A., Zhou, X. Z., Shen, M., Reimer, U., Rahfeld, J. U., Lu, K. P., and Fischer, G. (1998). Biochemistry 37, 5566–5575. Schutkowski, M., Drewello, M., Wollner, S., Jakob, M., Reimer, U., Scherer, G., Schierhorn, A., and Fischer, G. (1996). FEBS Lett. 394, 289–294. Sekerina, E., Rahfeld, J. U., Muller, ¨ J., Fangh¨anel, J., Rascher, C., Fischer, G., and Bayer, P. (2000). J. Mol. Biol. 301, 1021–1035. Sendak, R. A., Rothwarf, D. M., Wedemeyer, W. J., Houry, W. A., and Scheraga, H. A. (1996). Biochemistry 35, 12978–12992. Shadidy, M., Caubit, X., Olsen, R., Seternes, O. M., Moens, U., and Krauss, S. (1999). Biochim. Biophys. Acta 1446, 295–307. Shen, M., Stukenberg, P. T., Kirschner, M. W., and Lu, K. P. (1998). Genes Develop. 12, 706–720. Sherry, B., Zybarth, G., Alfano, M., Dubrovsky, L., Mitchell, R., Rich, D., Ulrich, P., Bucala, R., Cerami, A., and Bukrinsky, M. (1998). Proc. Natl. Acad. Sci. USA 95, 1758–1763. Shou, W. N., Aghdasi, B., Armstrong, D. L., Guo, Q. X., Bao, S. D., Charng, M. J., Mathews, L. M., Schneider, M. D., Hamilton, S. L., and Matzuk, M. M. (1998). Nature 391, 489–492. Siekierka, J. J., Hung, S. H. Y., Poe, M., Lin, C. S., and Sigal, N. H. (1989). Nature 341, 755–757. Snyder, S. H., Sabatini, D. M., Lai, M. M., Steiner, J. P., Hamilton, G. S., and Suzdak, P. D. (1998). Trends Pharmacol. Sci. 19, 21–26. Spitzfaden, C., Weber, H. P., Braun, W., Kallen, J., Wider, G., Widmer, H., Walkinshaw, M. D., and Wuthrich, ¨ K. (1992). FEBS Lett. 300, 291–300.

PROLYL ISOMERASES

281

Stamnes, M. A., Shieh, B. H., Chuman, L., Harris, G. L., and Zuker, C. S. (1991). Cell 65, 219–227. Stein, R. L. (1993). Adv. Protein Chem. 44, 1–24. Steinberg, I. Z., Harrington, W. F., Berger, A., Sela, M., and Katchalski, E. (1960). J. Am. Chem. Soc. 82, 5263–5279. Steinmann, B., Bruckner, P., and Supertifurga, A. (1991). J. Biol. Chem. 266, 1299–1303. Stewart, D. E., Sarkar, A., and Wampler, J. E. (1990). J. Mol. Biol. 214, 253–260. Stoller, G., Rucknagel, ¨ K. P., Nierhaus, K., Schmid, F. X., Fischer, G., and Rahfeld, J.-U. (1995). EMBO J. 14, 4939–4948. Stoller, G., Tradler, T., Rucknagel, ¨ J.-U., and Fischer, G. (1996). FEBS Lett. 384, 117–122. Strynadka, N. C., Adachi, H., Jensen, S. E., Johns, K., Sielecki, A., Betzel, C., Sutoh, K., and James, M. N. (1992). Nature 359, 700–705. Takahashi, N., Hayano, T., and Suzuki, M. (1989). Nature 337, 473–475. Teter, S. A., Houry, W. A., Ang, D., Tradler, T., Rockabrand, D., Fischer, G., Blum, P., Georgopoulos, C., and Hartl, F. U. (1999). Cell 97, 755–765. Texter, F. L., Spencer, D. B., Rosenstein, R., and Matthews, C. R. (1992). Biochemistry 31, 5687–5691. Thali, M., Bukovsky, A., Kondo, E., Rosenwirth, B., Walsh, C. T., Sodroski, J., and G¨ottlinger, H. G. (1994). Nature 372, 363–365. Thies, M. J. W., Mayer, J., Augustine, J. G., Frederick, C. A., Lilie, H., and Buchner, J. (1999). J. Mol. Biol. 293, 67–79. Tropschug, M., Wachter, E., Mayer, S., Sch¨onbrunner, E. R., and Schmid, F. X. (1990). Nature 346, 674–676. Truckses, D. M., Somoza, J. R., Prehoda, K. E., Miller, S. C., and Markley, J. L. (1996). Protein Sci. 5, 1907–1916. Udgaonkar, J. B., and Baldwin, R. L. (1988). Nature 335, 694–699. Udgaonkar, J. B., and Baldwin, R. L. (1990). Proc. Natl. Acad. Sci. USA 87, 8197–8201. Vajdos, F. F., Yoo, S., Houseweart, M., Sundquist, W. I., and Hill, C. P. (1997). Protein Sci. 6, 2297–2307. Valent, Q. A., Kendall, D. A., High, S., Kusters, R., Oudega, B., and Luirink, J. (1995). EMBO J. 14, 5494–5505. Van Duyne, G. D., Standaert, R. F., Karplus, P. A., Schreiber, S. L., and Clardy, J. (1993). J. Mol. Biol. 229, 105–124. Van Duyne, G. D., Standaert, R. F., Schreiber, S. L., and Clardy, J. (1991). J. Am. Chem. Soc. 113, 7433–7434. Vanhove, M., Lejeune, A., Guillaume, G., Virden, R., Pain, R. H., Schmid, F. X., and Frere, J. M. (1998b). Biochemistry 37, 1941–1950. Vanhove, M., Lejeune, A., and Pain, R. H. (1998a). Cell Mol. Life Sci. 54, 372–377. Vanhove, M., Raquet, X., and Frere, J.-M. (1995). Proteins: Struct. Funct. Genet. 22, 110–118. Vanhove, M., Raquet, X., Palzkill, T., Pain, R. H., and Frere, J. M. (1996). Proteins: Struct. Funct. Genet. 25, 104–111. Veeraraghavan, S., Holzman, T. F., and Nall, B. T. (1996). Biochemistry 35, 10601–10607. Veeraraghavan, S., and Nall, B. T. (1994). Biochemistry 33, 687–692. Veeraraghavan, S., Nall, B. T., and Fink, A. L. (1997). Biochemistry 36, 15134–15139. Veeraraghavan, S., Rodriguez-Gdiharpour, S., MacKinnon, C., Mcgee, W. A., Pierce, M. M., and Nall, B. T. (1995). Biochemistry 34, 12892–12902. Verdecia, M. A., Bowman, M. E., Lu, K. P., Hunter, T., and Noel, J. P. (2000). Nat. Struct. Biol. 7, 639–643. von Ahsen, O., Lim, J. H., Caspers, P., Martin, F., Schonfeld, H. J., Rassow, J., and Pfanner, N. (2000). J. Mol. Biol. 297, 809–818.

282

FRANZ X. SCHMID

Walensky, L. D., Dawson, T. M., Steiner, J. P., Sabatini, D. M., Suarez, J. D., Klinefelter, G. R., and Snyder, S. H. (1998). Mol. Med. 4, 502–514. Wang, T. W., Donahoe, P. K., and Zervos, A. S. (1994). Science 265, 674–676. Wang, T. W., Li, B. Y., Danielson, P. D., Shah, P. C., Rockwell, S., Lechleider, R. J., Martin, J., Manganaro, T., and Donahoe, P. K. (1996). Cell 86, 435–444. Wheeler, K. A., Hawkins, A. R., Pain, R., and Virden, R. (1998). Proteins: Struct. Funct. Genet. 33, 550–557. Wild, J., Rossmeissl, P., Walter, W. A., and Gross, C. A. (1996). J. Bacteriol. 178, 3608–3613. Winkler, K. E., Swenson, K. I., Kornbluth, S., and Means, A. R. (2000). Science 287, 1644– 1647. Wintermeyer, E., Ludwig, B., Steinert, M., Schmidt, B., Fischer, G., and Hacker, J. (1995). Infect. Immun. 63, 4576–4583. Wu, W. J., and Raleigh, D. P. (1998). Biopolymers 45, 381–394. Wu, X. Y., Wilcox, C. B., Devasahayam, G., Hackett, R. L., Ar´evalo-Rodr´ıguez, M., Cardenas, M. E., Heitman, J., and Hanes, S. D. (2000). EMBO J. 19, 3727–3738. W¨ulﬁng, C., Lombardero, J., and Pl¨uckthun, A. (1994). J. Biol. Chem. 269, 2895–2901. Yaffe, M. B., Schutkowski, M., Shen, M. H., Zhou, X. Z., Stukenberg, P. T., Rahfeld, J. U., Xu, J., Kuang, J., Kirschner, M. W., Fischer, G., Cantley, L. C., and Lu, K. P. (1997). Science 278, 1957–1960. Yao, D., and Dore, J. J., Jr., and Leof, E. B. (2000). J. Biol. Chem. 275, 13149–13154. Yoo, S., Myszka, D. G., Yeh, C.-y., McMurray, M., Hill, C. P., and Sundquist, W. I. (1997). J. Mol. Biol. 269, 780–795. Yuan, X., Werner, J. M., Knott, V., Handford, P. A., Campbell, I. D., and Downing, A. K. (1998). Protein Sci. 7, 2127–2135. Yuan, X., Downing, A. K., Knott, V., and Handford, P. A. (1997). EMBO J. 16, 6659–6666. Zarnt, T., Tradler, T., Stoller, G., Scholz, C., Schmid, F. X., and Fischer, G. (1997). J. Mol. Biol. 271, 827–837. Zeng, B. F., Macdonald, J. R., Bann, J. G., Reck, K., Gambee, J. E., Boswell, B. A., and Bachinger, H. P. (1998). Biochem. J. 330, 109–114. Zhao, Y. D., Chen, Y. Q., Schutkowski, M., Fischer, G., and Ke, H. M. (1997). Structure 5, 139–146.

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION IN Escherichia coli By MARTIN W. BADER and JAMES C. A. BARDWELL Department of Molecular Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan 48109-1048

I. II. III. IV. V. VI. VII. VIII. IX.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . De Novo Formation of Disulfide Bonds in E. coli: The Discovery of DsbA . . . DsbA Is the Most Oxidizing Disulfide Catalyst . . . . . . . . . . . . . . . . . . . . . . . . . DsbB Provides the Periplasm with Oxidizing Power . . . . . . . . . . . . . . . . . . . . . Correcting Wrong Disulfide Bonds in the Periplasm: Disulﬁde Bond Isomerization by DsbC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DsbD Provides Reducing Equivalents in a Highly Oxidizing Environment . . Dsb Proteins and Cytochrome c Maturation . . . . . . . . . . . . . . . . . . . . . . . . . . . Disulfide Bond Formation Does Not Interfere with Disulfide Isomerization Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

283 284 286 290 292 296 297 298 298 299

I. INTRODUCTION The folding of proteins into their three-dimensional structure is essential for their biological function. For proteins that contain disulﬁde bonds, formation of these bonds is often an important step in the folding reaction. The presence of one or more disulﬁde bonds is crucial to the maintenance of the folded state of many secretory proteins. In contrast, cytosolic proteins form disulﬁde bonds only as part of their catalytic cycle, and are not stabilized by these bonds. The classic experiments conducted by Anﬁnsen and co-workers proved that all the information for the three-dimensional structure of a protein is encoded by its amino acid sequence (Anﬁnsen, 1973) Bovine pancreatic RNase A, a 124-residue protein that contains four disulﬁde bonds in its native state, was used as a model protein (Sela et al., 1959; White, 1961). RNase A denatures readily and its disulﬁde bonds are reduced by incubation in urea and 2-mercaptoethanol. After removal of urea and 2-mercaptoethanol by dialysis, a very slow but nearly complete recovery of catalytic activity is observed. Thus, Anﬁnsen concluded that it is possible to refold denatured proteins into their active state in the test tube. Based on this observation, he further noted that “the information for the assumption of the native secondary and tertiary structure is contained in the amino acid sequence itself” (Anﬁnsen et al., 1961). 283 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

284

MARTIN W. BADER AND JAMES C. A. BARDWELL

Native RNase A contains four disulﬁde bonds and the correct formation of these bonds is a key step during refolding of the protein (Anﬁnsen et al., 1961). Accordingly, if fully denatured and reduced RNase A is allowed to form disulﬁde bonds under denaturing conditions, e.g., in the presence of 8 M urea, a mixture of randomly oxidized RNase A molecules, called scrambled RNase, is obtained. Further, on removal of the denaturant, scrambled RNase is basically inactive, suggesting that incorrect disulﬁde bonds had “locked” the enzyme in numerous misfolded conformations. However, if scrambled RNase A is dialyzed against a buffer that contains a small amount of the reductant 2-mercaptoethanol, enzymatic activity is restored. This provided the ﬁrst evidence that a lowmolecular-weight reductant could reduce nonnative disulﬁde bonds and allow reformation of native disulﬁde bonds. Nevertheless, the ten hours needed for full restoration of RNase A activity under these conditions seemed much too long for efﬁcient disulﬁde bond formation in vivo. It takes only two minutes for the cell to synthesize RNase A, and its folding is complete within minutes, not hours. This discrepancy led Anﬁnsen and co-workers to postulate and later to identify an enzymatic activity that greatly accelerates reactivation of RNase A in liver microsomes (Goldberger et al., 1963). The protein associated with this activity, protein disulﬁde isomerase (PDI), was the ﬁrst protein folding catalyst found. Recently, signiﬁcant advances have been made in understanding how proteins fold in a cellular environment, and a large number of other proteins have been identiﬁed which assist the proper folding of proteins inside the cell. These proteins fall into two classes. First, there are true catalysts, such as PDI, which accelerate rate-limiting steps during protein folding. Such rate-limiting steps include the correct formation of disulﬁde bonds and the isomerization of proline residues in proteins. Molecular chaperones comprise a second class of proteins that assist the folding process. Chaperones prevent nonproductive reactions such as aggregation or premature folding of proteins, they also promote the folding and unfolding of proteins. The mechanisms of chaperone action will not be further discussed here, but excellent reviews are available in the literature (Beissinger and Buchner, 1998; Ellis and Hartl, 1999). This chapter deals with folding catalysts, in particular with catalysts that are essential for disulﬁde bond formation in Escherichia coli. Nevertheless, some of these catalysts contain chaperone activity, demonstrating that these two activities are sometimes found within the same protein molecule II. De Novo FORMATION OF DISULFIDE BONDS IN E. coli: THE DISCOVERY OF DsbA In 1991, Bardwell and co-workers reported the identiﬁcation of DsbA, which they found to be involved in the formation of disulﬁde bonds

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

285

in vivo (Bardwell et al., 1991). Mutants in dsbA exhibit a severe defect in the oxidative folding of several E. coli proteins. Prior to the discovery of DsbA, it was widely believed that the formation of disulﬁde bonds occurred spontaneously. However, the isolation of DsbA showed that disulﬁde bond formation in the cell depends on the presence of a catalyst. Using DsbA as a model catalyst, many crucial questions concerning disulﬁde bond formation were answered. Some advantages that make DsbA a good model protein are its small size (21 kDa) and its single function—the oxidation of disulﬁde bonds—in contrast to the multifunctional, multidomain 57-kDa PDI. For these reasons, rapid progress has been made in analyzing how oxidative folding of proteins is catalyzed within the cell. DsbA was identiﬁed by the use of a disulﬁde indicator protein, MalF-β-galactosidase (Bardwell et al., 1991). This fusion protein lacks β-galactosidase activity when present in a wild-type E. coli background that is competent in forming disulﬁdes. The dsbA null mutant was isolated by selecting for a Lac+ phenotype in a cell that expressed the gene for the fusion protein but lacked the wild-type lacZ gene. The restoration of β-galactosidase activity in the dsbA null mutant is most likely due to reduction of cysteine residues of β-galactosidase. β-Galactosidase is a cytosolic protein that does not normally contain disulﬁde bonds. It was suggested that on fusion with the inner membrane protein MalF, a portion of β-galactosidase is exported into the periplasm. This part of the enzyme then becomes sensitive to thiol oxidation, which leads to inactivation of β-galactosidase in a wild-type strain background. However, in a strain such as a dsbA null mutant where disulﬁdes fail to form, β-galactosidase remains reduced and active. Loss of DsbA causes a severe defect in the formation of disulﬁde bonds in many secretory proteins including OmpA, alkaline phosphatase, and β-lactamase (Bardwell et al., 1991). Since the absence of DsbA causes a general lack of disulﬁde bonds in periplasmic proteins, it was concluded that DsbA is the major catalyst of disulﬁde bond formation in the periplasm. This key role of DsbA also explains why dsbA null mutants exhibit pleiotropic phenotypes. Commonly observed phenotypes of dsbA null mutants are loss of motility and increased sensitivity to DTT, benzylpenicillin, and metal ions such as Hg2+ and Cd2+ (Dailey and Berg, 1993; Missiakas et al., 1993; Rensing et al., 1997; Stafford et al., 1999). These phenotypes are due to the loss of the oxidase activity of DsbA, which results in a failure to form disulﬁde bonds in many periplasmic proteins. For instance, the loss of motility is due to the absence of a disulﬁde bond in the P ring protein (FlgI) of ﬂagella (Dailey and Berg, 1993). Another phenotype of dsbA null mutants in pathogenic bacteria is attenuated virulence. Because so many virulence factors contain

286

MARTIN W. BADER AND JAMES C. A. BARDWELL

disulﬁde bonds, DsbA is important for the disease-causing properties of enteropathogenic and uropathogenic E. coli, Vibrio cholera, and Shigella flexneri (Donnenberg et al., 1997; Jacob-Dubuisson et al., 1994; Peek and Taylor, 1992; Watarai et al., 1995; Yu, 1998).

III. DsbA IS THE MOST OXIDIZING DISULFIDE CATALYST The 2.0-A˚ crystal structure revealed that DsbA contains a thioredoxinlike fold (Martin et al., 1993). The thioredoxin fold includes a central β-sheet formed by four antiparallel β-strands. The central β-sheet is ﬂanked by a perpendicular helix and two helices on the opposite side (Martin, 1995). Compared to thioredoxin, DsbA contains an additional β-strand in the central β-sheet and the insertion of a 65-residue helical domain (Fig. 1). Such insertions are commonly observed within the thioredoxin family (Martin, 1995; McCarthy et al., 2000). Most members of the thioredoxin superfamily are involved in disulﬁde exchange reactions, and contain a redox-active CXXC motif in their active site. The CXXC motif participates in disulﬁde exchange reactions by going through reversible cycles of oxidation and reduction. In this motif, the

FIG. 1. The crystal structure of DsbA. DsbA contains a thioredoxin-like fold including the insertion of an α-helical domain. The arrow indicates the location of the active-site disulﬁde bond.

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

287

N-terminal cysteine is solvent-exposed, making it the reactive species in disulﬁde exchange reactions. Despite their common structures, thioredoxin and DsbA fulﬁll different functions and exist in different cellular compartments. While thioredoxin acts as a reductant of disulﬁde bonds in the cytosol (for review see Rietsch and Beckwith, 1998), DsbA introduces disulﬁde bonds into newly synthesized proteins during their translocation to the periplasm. Why does DsbA act as a donor of disulﬁde bonds? The disulﬁde bond formed by the CXXC motif of DsbA is highly reactive. Thus, oxidized DsbA will react rapidly with thiols, resulting in their oxidation. For example, DsbA reacts about a 1000-fold faster with reduced glutathione (GSH) than does a normal protein disulﬁde (Zapun et al., 1993). The extremely oxidizing nature of DsbA becomes evident from its equilibrium constant with glutathione (K ox), which is very small, 0.1 mM, indicating that it will strongly tend to oxidize thiols. K ox is given by the following equation (Wunderlich and Glockshuber, 1993; Zapun et al., 1993):

K ox =

[GSH]2 [DsbAox ] [GSSG][DsbAred ]

From the equilibrium constant with glutathione and the standard redox potential of the GSSG/GSH pair, the redox potential of DsbA can be calculated. The redox potential of DsbA is −120 mV, making it the most oxidizing disulﬁde bond known. For comparison, the redox potential of thioredoxin is −270 mV, and therefore much more reducing. The small equilibrium constant of DsbA with glutathione demonstrates that the disulﬁde bond formed by DsbA is highly unstable. The stability of a particular disulﬁde bond corresponds to the extent to which a protein is stabilized by this bond. In other words, the more stable the disulﬁde bond, the more stable the protein conformation. In the case of DsbA, its unstable disulﬁde bond should therefore destabilize the protein conformation. This is indeed observed, since the reduction of DsbA’s disulﬁde bond leads to stabilization of its folded conformation by 4.5 kcal/mol (Zapun et al., 1993). This is unusual, since disulﬁde bonds normally stabilize proteins. Yet, it is in agreement with the in vivo function of DsbA as a donor of disulﬁde bonds. What causes the reduced form of DsbA to be more stable than its oxidized form? In the CXXC motif of DsbA, the N-terminal cysteine 30 is solvent-exposed and has a very low pKa of ∼3.5, in contrast to the pKa of ∼9.0 for cysteines commonly found in proteins (Nelson and Creighton, 1994). The pKa is the pH at which the group is half ionized. Consequently, at physiologic pH,

288

MARTIN W. BADER AND JAMES C. A. BARDWELL

cysteine 30 of DsbA is fully deprotonated and found as a thiolate anion carrying a negative charge. It is the stabilization of this negative charge that accounts for the difference in stability between the reduced and the oxidized form of DsbA (Grauschopf et al., 1995; Nelson and Creighton, 1994). This stabilization effect makes the reduction of DsbA very favorable and drives the reaction between oxidized DsbA and reduced substrate proteins. The ﬁnding that cysteine 30 of DsbA has such a low pKa requires that its deprotonated form be stabilized by residues in the vicinity. For instance, histidine 32, which lies within the CXXC motif, plays an important role in determining the redox properties of DsbA (Grauschopf et al., 1995). Mutation of histidine 32 leads to a dramatic decrease of redox potential, making DsbA a less potent donor of disulﬁde bonds. Three crystal structures of DsbA histidine 32 mutants have been solved (Guddat et al., 1997a). The structures of these mutant proteins do not show any signiﬁcant change in the overall fold, although their equilibrium constants with glutathione (K ox) are up to 1000-fold less oxidizing than wild-type DsbA. Apparently, these large differences in K ox cannot be explained by signiﬁcant structural changes between the individual proteins. It was therefore suggested that electrostatic interactions, which are absent in the mutant proteins, stabilize the thiolate anion of cysteine 30, thus causing its extremely low pKa. Support for this model comes from comparison of the crystal structures of reduced and oxidized DsbA (Guddat et al., 1998). According to this study, the structure of reduced DsbA reveals potential hydrogen bonds between residues around cysteine 30, which are absent in the oxidized structure. This hydrogen bonding network includes the backbone amide of histidine 32, cysteine 33, and the side chains of cysteine 33. Further, histidine 32 moves toward cysteine 30 on the reduction of DsbA, bringing this residue within hydrogen bond distance. Since cysteine 30 is located at the N terminus of an α-helix, the thiolate of cysteine 30 is also stabilized by favorable interactions with the partial positive charge of the helix dipole. These data, taken together, suggest that DsbA’s highly oxidizing nature arises from a few electrostatic interactions that favor the very low pKa of cysteine 30. Further insights into how DsbA’s redox properties are determined come from comparison between DsbA and thioredoxin. The redox potential of thioredoxin is −270 mV as compared to −120 mV for DsbA. This makes thioredoxin a much more reducing catalyst than DsbA, and appears to suit the in vivo function of thioredoxin as a reductant of disulﬁde bonds in the cytosol. As in DsbA, the residues that lie within the CXXC motif of thioredoxin strongly inﬂuence the redox potential of the

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

289

catalyst. For instance, if the active site of thioredoxin (CGPC) is changed to match the sequence of DsbA (CPHC), the redox potential of the mutant thioredoxin increases to −200 mV, making it more oxidizing ( Jonda et al., 1999). More support for the importance of central residues of CXXC motifs in determining the redox properties of the catalyst comes from in vivo studies with thioredoxin. Debarbieux and Beckwith fused an export signal to the N terminus of thioredoxin, causing its export to the periplasm (Debarbieux and Beckwith, 1998). Such a construct only partially promotes the formation of disulﬁde bonds of OmpA and alkaline phosphatase in a dsbA null background. However, if the active site of this thioredoxin construct is mutated to resemble the active site of DsbA, thus making it more oxidizing, the folding yields of OmpA and alkaline phosphatase are nearly indistinguishable from a DsbA+ strain (Debarbieux and Beckwith, 2000). Thus, thioredoxin can be turned into a DsbA-like catalyst by simply exchanging the two residues within its CXXC motif. As the major oxidant in the periplasm, DsbA introduces disulﬁde bonds into newly translocated proteins. The folding of these proteins competes with DsbA-mediated formation of their disulﬁde bonds. Consequently, premature folding would mask cysteine residues normally involved in forming disulﬁde bonds, thereby preventing their oxidation by DsbA. Thus, DsbA must interact rapidly with proteins while they are still unfolded or only partially folded in order to gain access to cysteine residues before they get buried upon folding. It is therefore reasonable to propose that DsbA speciﬁcally interacts noncovalently with unfolded substrate proteins. Indeed, there is good experimental evidence that it does so. (1) DsbA reacts about 10–25 times faster with unfolded hirudin than it does with DTT (Wunderlich et al., 1993). (2) Disulﬁde exchange between DsbA and a peptide derived from residues 4–31 of BPTI occurs up to 1000-fold faster than between glutathione and DsbA (Darby and Creighton, 1995). (3) Further, disulﬁde exchange between the model peptide and DsbA occurs via a mixed disulﬁde between the peptide and DsbA. This mixed disulﬁde bond is more stable than the disulﬁde bond between DsbA and glutathione, indicating the presence of additional stabilizing interactions with the peptide. (4) Moreover, a study by Frech and colleagues reported the isolation of a mixed disulﬁde complex between a DsbA variant lacking the second cysteine and a ribonuclease T1 variant (Frech et al., 1996). These authors showed that the conformation of DsbA is stabilized by 4.7 kJ/mol in the mixed disulﬁde complex. This strongly suggests that DsbA interacts noncovalently with substrate proteins. A potential substrate binding site of DsbA has been deduced from its crystal structure (Guddat et al., 1997b); however, no structure of a

290

MARTIN W. BADER AND JAMES C. A. BARDWELL

complex between DsbA and a peptide has been solved to date. DsbA exhibits a potential hydrophobic peptide binding groove below the active site disulﬁde, which was modeled according to the NMR structure of a mixed disulﬁde complex of thioredoxin and a target peptide (Qin et al., 1995). According to this model, conserved uncharged residues around the active site of DsbA are involved in peptide binding. Another very recent approach to study the interaction of DsbA with peptides was carried out with a model peptide containing a bromine-substituted alanine (Couprie et al., 2000). DsbA was speciﬁcally crosslinked to this peptide via its reactive cysteine. Preliminary calorimetric and NMR analysis revealed stabilization of DsbA due to peptide binding, further supporting the importance of noncovalent interactions between DsbA and substrate proteins. IV. DsbB PROVIDES THE PERIPLASM WITH OXIDIZING POWER Following the transfer of disulﬁde bonds from DsbA to substrate proteins, the active site of DsbA must be reoxidized in order to go through another catalytic cycle. This is accomplished by the inner membrane protein DsbB, which is responsible for keeping the disulﬁde bond of DsbA in an oxidized state. The dsbB gene was originally identiﬁed by using the same selection that was used to isolate dsbA (Bardwell et al., 1993). Like dsbA mutants, dsbB mutants show a strong defect in the formation of disulﬁde bonds in periplasmic proteins such as OmpA, β-lactamase, and alkaline phosphatase. In an independent approach, the dsbB gene was isolated by a genetic screen for multicopy suppressors of DTT sensitivity (Missiakas et al., 1993). The rationale behind this latter screen is that a protein participating in an oxidative pathway should confer resistance to the reductant DTT. Such screens have been used by the same authors to isolate more genes belonging to the Dsb family (see below). Early genetic evidence suggested that DsbA and DsbB participate in the same pathway (Bardwell et al., 1993), and that DsbB is responsible for the reoxidation of DsbA’s active site disulﬁde bond. For instance, dsbB null mutants accumulate DsbA in a reduced state in the periplasm, while DsbA is found in a mostly oxidized form in a wild-type background. This ﬁnding implies an important function for DsbB in reoxidizing DsbA. Evidence that DsbA and DsbB may interact directly comes from the isolation of a dsbA mutant that lacks cysteine 33 and therefore contains only the reactive cysteine 30 of its CXXC motif (Guilhot et al., 1995; Kishigami et al., 1995). This mutant dsbA, C33Y, was identiﬁed by screening a library of randomly mutagenized DsbA-expression plasmids for a

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

291

disulﬁde-negative phenotype in a DsbA+-strain background. When overexpressed, dsbA C33Y displays a dominant negative phenotype, causing a severe defect in the formation of disulﬁde bonds. The dominant negative phenotype is suppressed when DsbB is co-overexpressed. This suggests that DsbA C33Y titrates out all the cellular DsbB, thus causing the disulﬁde-negative phenotype. This was further supported by the isolation of a complex between DsbA C33Y and DsbB. Apparently, the reactive cysteine 30 of DsbA crosslinks to DsbB, thus leading to the formation of a mixed disulﬁde complex between the two proteins. The complex cannot resolve due to the lack of cysteine 33 in DsbA C33Y, and, as a consequence, DsbA C33Y inhibits DsbB activity. Based on its sequence, DsbB was predicted to be an inner membrane protein, which was conﬁrmed using the alkaline phosphatase fusion approach (Jander et al., 1994). DsbB was shown to contain four transmembrane domains, which are connected by two periplasmic loops. Furthermore, DsbB possesses four highly conserved cysteines, which are essential for its activity in reoxidizing DsbA. One pair of cysteines is found as a CXXC motif in the ﬁrst periplasmic loop of DsbB. Apart from this motif, DsbB has no other similarity to thioredoxin, making it rather unlikely that DsbB belongs to the thioredoxin superfamily. The most important criterion for a protein to be a member of this family is the presence of a common fold rather than a CXXC motif (Martin, 1995). The location of DsbB in the inner membrane led to the speculation that DsbB donates electrons generated by disulﬁde bond formation to the respiratory chain (Bardwell, 1994). Consistent with this, heme- or quinone-depleted cells accumulate DsbA and the CXXC motif of the DsbB motif in a reduced state (Kobayashi and Ito, 1999; Kobayashi et al., 1997). Nevertheless, the ultimate source of oxidizing power for oxidative protein folding remained unclear, as did the mechanism whereby DsbB drives disulﬁde bond formation. Only recently has it been shown how DsbB interacts with the electron transport system (Bader et al., 1998, 1999, 2000). In a highly puriﬁed in vitro system, DsbB activity requires the presence of either cytochrome bo or bd oxidase, each of which acts as a terminal step in the respiratory chain by transferring electrons from ubiquinone to molecular oxygen. Ubiquinone was identiﬁed as an intermediate between DsbB and the two cytochrome oxidases. Thus, DsbB catalyzes the ﬂow of electrons from DsbA to ubiquinone, whence electrons are passed on to cytochrome oxidases. The ﬁnding that DsbB directly interacts with quinones also helps explain why disulﬁde bond formation is not impaired in the absence of molecular oxygen. Under these conditions, ubiquinones get replaced by menaquinones, which serve as mobile electron carriers

292

MARTIN W. BADER AND JAMES C. A. BARDWELL

FIG. 2. The DsbA–DsbB pathway. DsbA is the immediate donor of disulﬁde bonds in the periplasm and cycles between its oxidized and reduced state. DsbB is directly involved in the reoxidation of DsbB. Electrons ﬂow from DsbA via DsbB to ubiquinone and further on to respiratory complexes, which catalyze electron transfer to oxygen. Under anaerobic conditions, menaquinone serves as an electron carrier between DsbB and terminal complexes.

between complexes of the anaerobic electron transport chain (Wallace and Young, 1977). Biochemical and genetic evidence suggests that, under anaerobic conditions, DsbB switches to utilizing menaquinone as its immediate electron acceptor (Bader et al., 1999). This is supported by the observation that a ubiA–menA double mutant lacking both ubiquinones and menaquinones accumulates DsbA in a reduced form in vivo, while DsbA is mostly oxidized in a ubiA single mutant (Kobayashi et al., 1997). Electrons therefore ﬂow from DsbB to menaquinones and further on to terminal complexes such as fumarate reductase. Taken together, the ability of DsbB to interact with either ubiquinones or menaquinones ensures efﬁcient disulﬁde bond formation under nearly all growth conditions (Fig. 2). V. CORRECTING WRONG DISULFIDE BONDS IN THE PERIPLASM: DISULFIDE BOND ISOMERIZATION BY DsbC By screening for E. coli mutants that display a DTT-hypersensitive phenotype, Missiakas and co-workers identiﬁed three additional Dsb proteins: DsbC, DsbD, and DsbG (Missiakas et al., 1994, 1995). DsbC was also isolated from a multicopy plasmid library by its ability to confer resistance to high levels of DTT. Unlike dsbA and dsbB mutants, the effect of a dsbC null mutant on the growth of E. coli is not very strong (Rietsch et al., 1996). There is, however, good evidence that DsbC acts

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

293

to isomerize incorrectly formed disulﬁde bonds. This becomes evident when eukaryotic proteins containing multiple disulﬁde bonds are targeted to the E. coli periplasm (Rietsch et al., 1996). For instance, the yield of native urokinase, a eukaryotic protein containing 12 disulﬁde bonds, is undetectable in a dsbC null mutant. On the other hand, the folding yield of the E. coli alkaline phosphatase, which contains only two disulﬁde bonds, is lowered by a mere 15% in a dsbC null background. Most periplasmic E. coli proteins contain only one or two disulﬁde bonds, perhaps explaining why disulﬁde isomerization appears less important for prokaryotes than it is in eukaryotes. The crystal structure of DsbC, which has been solved to a resolution of 1.9 A˚ (McCarthy et al., 2000), shows DsbC to be a homodimer consisting of two separate domains, an N-terminal dimerization domain and a C-terminal thioredoxin domain (Fig. 3). In the dimer, two β-strands from each N-terminal domain interact with the corresponding strands of the opposite molecule to form an overall V-shaped structure. The two C-terminal thioredoxin-like domains form the bulk of the arms of the V and include two redox active CXXC motifs facing the inside of the V. As in DsbA, the N-terminal cysteine residue of the CXXC motif is solvent-exposed, making it the reactive species in disulﬁde exchange reactions. In addition, the disulﬁde bond formed by DsbC displays an equilibrium constant with glutathione (K ox) of 0.12 mM, making DsbC highly reactive and only slightly less oxidizing than DsbA.

FIG. 3. The crystal structure of DsbC. DsbC forms a V-shaped homodimer. The monomer consists of a C-terminal thioredoxin domain and an N-terminal dimerization domain. The two domains are joined via a linker helix. The monomers interact via two consecutive β-strands, which form two extended β-sheets in the dimer.

294

MARTIN W. BADER AND JAMES C. A. BARDWELL

FIG. 4. Different mechanisms of DsbA and DsbC. (Top) DsbA randomly oxidizes cysteine residues, resulting in the net formation of disulﬁde bonds. The presence of three or more cysteines in a substrate protein may cause formation of incorrect disulﬁde bonds on DsbA-mediated oxidation. (Bottom) DsbC functions as an isomerase. DsbC has to be in its reduced state in order to attack incorrect disulﬁde bonds. Note that this cycle does not lead to the net formation of a disulﬁde bond.

DsbC isomerizes disulﬁde bonds in vivo more efﬁciently than does DsbA (Sone et al., 1997). During its catalytic cycle, DsbC forms a mixed disulﬁde bond with its substrate protein, thus freeing it from an incorrect disulﬁde bond (Fig. 4). This mixed disulﬁde complex will resolve once a more stable and hence more nativelike disulﬁde bond is “found” in the target protein. On the other hand, DsbA rapidly oxidizes proteins, which does not necessarily result in the formation of correct disulﬁde bonds. In contrast to DsbA, no net formation of disulﬁde bonds occurs during the catalytic cycle of DsbC. This is because the second active site cysteine 101 does not directly participate in the isomerization mechanism drawn

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

295

in Figure 4. Therefore, cysteine 101 might be important only for the resolution of kinetically trapped complexes between DsbC and target proteins. Indeed, such off-pathway intermediates seem to occur because mutations that alter cysteine 101 lead to a dramatic decrease of the folding yield of urokinase (Rietsch et al., 1996). The attack of cysteine 101 on such a trapped mixed disulﬁde leads to the oxidation of the CXXC motif of DsbC. In any case, only reduced DsbC is capable of attacking incorrect disulﬁde bonds. Therefore, there is a need to keep DsbC in a reduced state in vivo. This is accomplished by the inner membrane protein DsbD, which ensures a steady-state level of reduced DsbC in the cell (see below). The mechanisms of DsbA and DsbC action are consistent with the observation that a mixed disulﬁde bond between DsbC and the model peptide BPTI 4-31 is 40- to 100-fold more stable than the corresponding complex between the model peptide and DsbA (Darby et al., 1998). The higher stability of DsbC in complex with peptides might explain why DsbC, but not DsbA, acts as an isomerase. As an isomerase, DsbC has to scan through many possible disulﬁde intermediates until a more native disulﬁde bond is formed in the substrate protein. It is therefore necessary that the mixed disulﬁde bond between DsbC and its substrate protein be more stable than the DsbA–protein complex. The latter has to be resolved rapidly to free DsbA for another cycle of disulﬁde bond formation after its oxidation by DsbB. The higher stability of a mixed disulﬁde complex between peptide and DsbC is likely to result from enhanced peptide binding by DsbC. The inside of the V-like structure of DsbC is covered with uncharged and hydrophobic residues, forming a potential substrate binding surface (McCarthy et al., 2000). Further support for the hypothesis that the interior of DsbC is responsible for peptide binding by DsbC comes from the observation that the dimeric nature of DsbC is essential for its function as an isomerase in vitro (Sun and Wang, 2000). In contrast to wild-type DsbC, monomeric DsbC also lacks chaperone activity in vitro. Chaperones often interact nonspeciﬁcally with hydrophobic regions of a protein in order to prevent aggregation. Therefore, the loss of chaperone activity of the monomer is probably due to the destruction of the extended uncharged surface are of DsbC on monomerization. A second disulﬁde isomerase called DsbG exists in the periplasm. DsbG was identiﬁed by employing the same DTT-hypersensitivity screen used to identify DsbC (Andersen et al., 1997). In an independent approach, DsbG was identiﬁed by homology to DsbC (Bessette et al., 1999). Like DsbC, DsbG is a dimer and is kept in a reduced state in vivo. DsbG has isomerase activity in vivo (Bessette et al., 1999). It is not clear what the

296

MARTIN W. BADER AND JAMES C. A. BARDWELL

substrate speciﬁcities for DsbC and DsbG are. Do the two isomerases act on the same set of misoxidized protein substrates, or do they act on different, nonoverlapping sets of substrates? The redundancy of two isomerases in E. coli is surprising, since most secreted E. coli proteins possess just one or two disulﬁde bonds. On the other hand, the two proteins also act as chaperones in vitro and this activity might be an important part of their in vivo roles (Chen et al., 1999; Shao et al., 2000). VI. DsbD PROVIDES REDUCING EQUIVALENTS IN A HIGHLY OXIDIZING ENVIRONMENT It is important for the cell to keep the two isomerases DsbC and DsbG in a reduced state since only their reduced forms are able to attack incorrect disulﬁde bonds. Reduction of DsbC and DsbG is carried out by an inner membrane protein called DsbD (Fig. 5) (Missiakas et al., 1995; Rietsch et al., 1996). Accordingly, a dsbD null mutant accumulates DsbC and DsbG in their oxidized forms, while the two isomerases are mainly found in their reduced forms in a wild-type background. The ﬁnding that DsbC and DsbG are kept reduced in the overall oxidizing environment

FIG. 5. The E. coli isomerization pathway. DsbC and DsbG are kept in a reduced state in vivo by the inner membrane protein DsbD. DsbD provides reducing equivalents in the periplasm by transferring electrons from the cytosol across the membrane.

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

297

of the periplasm requires the constant ﬂow of reducing equivalents to the periplasm. Genetic evidence suggests that reducing equivalents originate from the cytosolic thioredoxin system and are transferred across the membrane via DsbD (Chung et al., 2000; Stewart et al., 1999). DsbD was shown to consist of three domains: an N-terminal 16-kDa domain (α), a transmembrane domain (β), and the N-terminal thioredoxin domain (γ ) (Fig. 5). Recently, a possible mechanism whereby DsbD transfers electrons through the membrane has been proposed. This mechanism includes consecutive disulﬁde exchange between six conserved cysteines, two of which are found in each domain of DsbD (Katzen and Beckwith, 2000). According to this model, electrons are transferred from cytosolic thioredoxin via the transmembrane domain (β) of DsbD to the γ -domain and further on to the α-domain. From there, electrons are ﬁnally passed on to DsbC, which was concluded from the isolation of a mixed disulﬁde complex between DsbC and the α-domain. Taken together, these data suggest that DsbD allows electron passage from the cytosol to the periplasm, thus keeping DsbC and DsbG in their reduced states, a prerequisite for their isomerase activities. VII. Dsb PROTEINS AND CYTOCHROME c MATURATION By transferring reducing equivalents to the periplasm, DsbD also plays a role distinct from disulﬁde bond isomerization. Mutants in dsbD, previously named dipZ, were shown to display a defect in the maturation of c-type cytochromes (Crooke and Cole, 1995). During anaerobic growth, E. coli synthesizes one or more of ﬁve different c-type cytochromes, depending on the available electron acceptor. An important step during the maturation of c-type cytochromes is the attachment of a covalently bound heme moiety to the apoprotein via two thioether linkages. The cysteine residues that participate in such thioether bonds have to be reduced before heme attachment can occur (Fabianek et al., 2000). This poses a fundamental problem since cysteines in periplasmic proteins are efﬁciently oxidized to disulﬁdes by DsbA. To reverse the DsbA-mediated oxidation of cysteine residues involved in heme binding, DsbD/DipZ provides reducing power to the periplasm. Reduction of c-type cytochromes requires the presence of yet another Dsb protein, which is called DsbE (CcmG) and acts in the same pathway with DsbD (Fabianek et al., 1998). Accordingly, the CXXC motif of DsbE is kept in a reduced state by DsbD. It was suggested that DsbE transfers electrons to c-type cytochromes either directly or via an additional protein called CcmH (Fabianek et al., 1999; Reid et al., 1998). Thus, the active-site cysteines of c-type cytochromes are kept in a reduced state before heme attachment can occur.

298

MARTIN W. BADER AND JAMES C. A. BARDWELL

VIII. DISULFIDE BOND FORMATION DOES NOT INTERFERE WITH DISULFIDE ISOMERIZATION DsbD is central in providing the periplasm with reducing power from the cytosol, which is important for such different cellular processes as cytochrome c biogenesis and disulﬁde isomerization. In the periplasm, the isomerization pathway, which includes DsbC, DsbG, and DsbD (Fig. 5), reduces nonnative disulﬁde bonds formed by the DsbA–DsbB pathways. How are these two systems separated from each other? Any crosstalk between them would be destructive and result in futile cycles of mutual reduction and oxidation. Indeed, recent genetic evidence suggests that the isomerization pathway does not interfere with the oxidation pathway. This line of evidence is connected to the methods used in the isolation of the dsbD gene. DsbD mutants were identiﬁed by their ability to partially suppress the phenotypes associated with a dsbA null mutation (Missiakas et al., 1995). The suppression depends on the presence of the dsbC gene (Rietsch et al., 1996). Thus, in the dsbA–dsbD null mutant, DsbC accumulates in its oxidized state and can therefore serve as a net donor of disulﬁde bonds by rescuing some of the phenotypes associated with a dsbA null mutant. Although DsbC seems to play a role similar to DsbA in a dsbA–dsbD null background, it is important to note that its reoxidation under such conditions does not depend on the presence of DsbB, but rather on the oxidant cystine present in rich media. Based on this observation, Rietsch and colleagues (1998) concluded that DsbB is unable to reoxidize DsbC in vivo and thus discriminates DsbC from DsbA. As a consequence, the two pathways do not interfere with each other, allowing them to carry out their opposite functions of oxidizing and reducing disulﬁde bonds. IX. CONCLUDING REMARKS The past years have shown that disulﬁde bond formation is actively catalyzed in the cell. We now have a very detailed understanding of the different electron pathways during oxidative protein folding in prokaryotes. DsbB and DsbD are the central membrane components which provide the periplasm with oxidizing and reducing equivalents, respectively, by linking disulﬁde bond formation to cell metabolism. The soluble periplasmic proteins DsbA, DsbC, and DsbG interact with target proteins and function as disulﬁde catalysts. DsbA is part of the oxidative pathway, which is responsible for the net introduction of disulﬁde bonds. The reoxidation of DsbA is linked to electron transport by DsbB. On the other hand, DsbC and DsbG isomerize incorrect disulﬁde bonds and, in order to do so, have to be kept in a reduced state by DsbD. The

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

299

coexistence of two distinct pathways in the periplasm guarantees a redox environment that favors the correct formation of disulﬁde bonds. A pathway analogous to the DsbA–DsbB system was recently shown to operate in the endoplasmic reticulum of eukaryotes (Frand and Kaiser, 1998; 1999; Pollard et al., 1998). This pathway includes the membranebound protein ERO1, which keeps protein disulﬁde isomerase (PDI) in an oxidized state in vivo. Consequently, PDI plays a role somewhat similar to that of DsbA in catalyzing the formation of disulﬁde bonds during the folding of secreted proteins. The mechanisms driving oxidative protein folding in eukaryotes are described in more detail in another chapter in this volume. Although we now possess a broad picture of the distinct electron pathways during oxidative protein folding in prokaryotes, it remains to be seen how exactly electrons are passed through the central membrane proteins DsbB and DsbD. It is therefore necessary to analyze the mechanism of these two proteins in vitro. Knowing the mode of interaction between DsbB and its immediate electron acceptor ubiquinone will certainly increase our understanding of how DsbB passes off electrons to the electron transport chain. On the other hand, the mechanism of how DsbD shuttles electrons through the membrane is less clear. Additional cofactors that play a role in electron transfer from the cytosol to the periplasm might be necessary for DsbD activity. Future work on the two disulﬁde isomerases DsbC and DsbG might show how these proteins interact with their native substrate proteins. It is of great interest to identify particular in vivo substrates for these catalysts. The interaction of DsbC and DsbG with their native substrates can then be studied in vivo and in vitro, and in conjunction with structural information, might lead to a better understanding of the mechanisms that drive disulﬁde isomerization. Ten years after the discovery of the ﬁrst Dsb protein, much progress has been made, but there is still much to learn about the Dsb protein family. Multiple genetic, biochemical, and structural approaches will be required to solve the remaining questions concerning disulﬁde metabolism in prokaryotes. REFERENCES Andersen, C. L., Matthey-Dupraz, A., Missiakas, D., and Raina, S. (1997). Mol. Microbiol. 26, 121–132. Anﬁnsen, C. B. (1973). Science 181, 223–229. Anﬁnsen, C. B., Haber, E., Sela, M., and White, F. H. (1961). Proc. Natl. Acad. Sci. USA 47, 1309–1314. Bader, M., Muse, W., Ballou, D. P., Gassner, C., and Bardwell, J. C. (1999). Cell 98, 217–227.

300

MARTIN W. BADER AND JAMES C. A. BARDWELL

Bader, M., Muse, W., Zander, T., and Bardwell, J. (1998). J. Biol. Chem. 273, 10302–10307. Bader, M. W., Xie, T., Yu, C. A., and Bardwell, J. C. (2000). J. Biol. Chem. 275, 26082–26088. Bardwell, J. C. (1994). Mol. Microbiol. 14, 199–205. Bardwell, J. C., Lee, J. O., Jander, G., Martin, N., Belin, D., and Beckwith, J. (1993). Proc. Natl. Acad. Sci. USA 90, 1038–1042. Bardwell, J. C., McGovern, K., and Beckwith, J. (1991). Cell 67, 581–589. Beissinger, M., and Buchner, J. (1998). Biol. Chem. 379, 245–259. Bessette, P. H., Cotto, J. J., Gilbert, H. F., and Georgiou, G. (1999). J. Biol. Chem. 274, 7784–7792. Chen, J., Song, J. L., Zhang, S., Wang, Y., Cui, D. F., and Wang, C. C. (1999). J. Biol. Chem. 274, 19601–19605. Chung, J., Chen, T., and Missiakas, D. (2000). Mol. Microbiol. 35, 1099–1109. Couprie, J., Vinci, F., Dugave, C., Quemeneur, E., and Moutiez, M. (2000). Biochemistry 39, 6732–6742. Crooke, H., and Cole, J. (1995). Mol. Micorbiol. 15, 1139–1150. Dailey, F. E., and Berg, H. C. (1993). Proc. Natl. Acad. Sci. USA 90, 1043–1047. Darby, N. J., and Creighton, T. E. (1995). Biochemistry 34, 3576–3587. Darby, N. J., Raina, S., and Creighton, T. E. (1998). Biochemistry 37, 783–791. Debarbieux, L., and Beckwith, J. (1998). Proc. Natl. Acad. Sci. USA 95, 10751–10756. Debarbieux, L., and Beckwith, J. (2000). J. Bacteriol. 182, 723 –727. Donnenberg, M. S., Zhang, H. Z., and Stone, K. D. (1997). Gene 192, 33–38. Ellis, R. J., and Hartl, F. U. (1999). Curr. Opin. Struct. Biol. 9, 102–110. Fabianek, R. A., Hennecke, H., and Thony-Meyer, L. (1998). J. Bacteriol. 180, 1947–1950. Fabianek, R. A., Hennecke, H., and Thony-Meyer, L. (2000). FEMS Microbiol. Rev. 24, 303 –316. Fabianek, R. A., Hofer, T., and Thony-Meyer, L. (1999). Arch. Microbiol. 171, 92–100. Frand, A. R., and Kaiser, C. A. (1998). Mol. Cell 1, 161–170. Frand, A. R., and Kaiser, C. A. (1999). Mol. Cell 4, 469 –477. Frech, C., Wunderlich, M., Glockshuber, R., and Schmid, F. X. (1996). EMBO J. 15, 392–398. Goldberger, R. F., Epstein, C. J., and Anﬁnsen, C. B. (1963). J. Biol. Chem. 238, 628–635. Grauschopf, U., Winther, J. R., Korber, P., Zander, T., Dallinger, P., and Bardwell, J. C. (1995). Cell 83, 947–955. Guddat, L. W., Bardwell, J. C., Glockshuber, R., Huber-Wunderlich, M., Zander, T., and Martin, J. L. (1997a). Protein Sci. 6, 1893–1900. Guddat, L. W., Bardwell, J. C., and Martin, J. L. (1998). Structure 6, 757–767. Guddat, L. W., Bardwell, J. C., Zander, T., and Martin, J. L. (1997b). Protein Sci. 6, 1148 –1156. Guilhot, C., Jander, G., Martin, N. L., and Beckwith, J. (1995). Proc. Natl. Acad. Sci. USA 92, 9895–9899. Jacob-Dubuisson, F., Pinkner, J., Xu, Z., Striker, R., Padmanhaban, A., and Hultgren, S. J. (1994). Proc. Natl. Acad. Sci. USA 91, 11552–11556. Jander, G., Martin, N. L., and Beckwith, J. (1994). EMBO J. 13, 5121–5127. Jonda, S., Huber-Wunderlich, M., Glockshuber, R., and Mossner, E. (1999). EMBO J. 18, 3271–3281. Katzen, F., and Beckwith, J. (2000). Cell 103. Kishigami, S., Kanaya, E., Kikuchi, M., and Ito, K. (1995). J. Biol. Chem. 270, 17072–17074. Kobayashi, T., and Ito, K. (1999). EMBO J. 18, 1192–1198.

CATALYSIS OF DISULFIDE BOND FORMATION AND ISOMERIZATION

301

Kobayashi, T., Kishigami, S., Sone, M., Inokuchi, H., Mogi, T., and Ito, K. (1997). Proc. Natl. Acad. Sci. USA 94, 11857–11862. Martin, J. L. (1995). Structure 3, 245–250. Martin, J. L., Bardwell, J. C., and Kuriyan, J. (1993). Nature 365, 464 –468. McCarthy, A. A., Haebel, P. W., Torronen, A., Rybin, V., Baker, E. N., and Metcalf, P. (2000). Nat. Struct. Biol. 7, 196–199. Missiakas, D., Georgopoulos, C., and Raina, S. (1993). Proc. Natl. Acad. Sci. USA 90, 7084– 7088. Missiakas, D., Georgopoulos, C., and Raina, S. (1994). EMBO J. 13, 2013–2020. Missiakas, D., Schwager, F., and Raina, S. (1995). EMBO J. 14, 3415–3424. Nelson, J. W., and Creighton, T. E. (1994). Biochemistry 33, 5974–5983. Peek, J. A., and Taylor, R. K. (1992). Proc. Natl. Acad. Sci. USA 89, 6210–6214. Pollard, M. G., Travers, K. J., and Weissman, J. S. (1998). Mol. Cell 1, 171–182. Qin, J., Clore, G. M., Kennedy, W. M., Huth, J. R., and Gronenborn, A. M. (1995). Structure 3, 289–297. Reid, E., Eaves, D. J., and Cole, J. A. (1998). FEMS Microbiol. Lett. 166, 369–375. Rensing, C., Mitra, B., and Rosen, B. P. (1997). J. Bacteriol. 179, 2769–2771. Rietsch, A., and Beckwith, J. (1998). Annu. Rev. Genet. 32, 163–184. Rietsch, A., Belin, D., Martin, N., and Beckwith, J. (1996). Proc. Natl. Acad. Sci. USA 93, 13048–13053. Sela, M., White, F. H., and Anﬁnsen, C. B. (1959). Biochim. Biophys. Acta 31, 417–426. Shao, F., Bader, M. W., Jakob, U., and Bardwell, J. C. (2000). J. Biol. Chem. 275, 13349– 13352. Sone, M., Akiyama, Y., and Ito, K. (1997). J. Biol. Chem. 272, 10349–10352. Stafford, S. J., Humphreys, D. P., and Lund, P. A. (1999). FEMS Microbiol. Lett. 174, 179– 184. Stewart, E. J., Katzen, F., and Beckwith, J. (1999). EMBO J. 18, 5963–5971. Sun, X. X., and Wang, C. C. (2000). J. Biol. Chem. 275, 22743–22749. Wallace, B. J., and Young, I. G. (1977). Biochim. Biophys. Acta 461, 84–100. Watarai, M., Tobe, T., Yoshikawa, M., and Sasakawa, C. (1995). Proc. Natl. Acad. Sci. USA 92, 4927–4931. White, F. H. (1961). J. Biol. Chem. 236, 1353–1359. Wunderlich, M., and Glockshuber, R. (1993). Protein Sci. 2, 717–726. Wunderlich, M., Otto, A., Seckler, R., and Glockshuber, R. (1993). Biochemistry 32, 12251–12256. Yu, J. (1998). Infect. Immun. 66, 3909–3917. Zapun, A., Bardwell, J. C., and Creighton, T. E. (1993). Biochemistry 32, 5083 –5092.

This . Page Intentionally Left Blank

N-GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING By E. SERGIO TROMBETTA and ARMANDO J. PARODI Instituto de Investigaciones Biotecnologicas, ´ Universidad de San Mart´ın, (1650) San Martin, Pcia. de Buenos Aires, Argentina

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. N-Glycan Processing in the Endoplasmic Reticulum . . . . . . . . . . . . . . . . . . . . III. Glycoprotein Reglucosylation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The UDP-Glc:Glycoprotein Glucosyltransferase . . . . . . . . . . . . . . . . . . . . B. Recognition of Nonnative Glycoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Chaperones and Protein Folding in the Endoplasmic Reticulum . . . . . . . . . V. Interaction of Glycoproteins with Calnexin and Calreticulin . . . . . . . . . . . . . VI. Calnexin and Calreticulin Are Lectins Specific for Monoglucosylated Oligosaccharides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Differential Binding to Calnexin and Calreticulin In Vivo . . . . . . . . . . . . B. Consequences of Glycoprotein–Calnexin/Calreticulin Interactions . . . C. The Role of Glycoprotein–Calnexin/Calreticulin Interactions . . . . . . . VII. N-Glycan Processing and Glycoprotein Degradation . . . . . . . . . . . . . . . . . . . . VIII. Summary and Future Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

303 304 308 309 314 320 321 323 324 324 328 330 332 333

I. INTRODUCTION Glycosylation of asparagine residues (N-glycosylation) in the endoplasmic reticulum (ER) presents two basic differences with respect to glycosylation of Ser/Thr residues (O-glycosylation) in either the cytosol or the Golgi apparatus. In N-glycosylation, a preassembled oligosaccharide containing 14 monosaccharide units is transferred “en bloc” to nascent polypeptide chains, whereas in O-glycosylation, either a variable number of monosaccharide units are sequentially transferred (Golgi O-glycosylation) or a singlet of such units is added (cytosolic O-glycosylation). Also, N-glycosylation occurs on nascent polypeptides, whereas O-glycosylation is a posttranslational event. After being transferred to nascent chains in the ER, N-oligosaccharides are processed all along the secretory pathway. Oligosaccharide remodeling in the Golgi not only presents a broad variation among different cell types, but also N-oligosaccharides on the same or different glycoproteins may be differently processed within the same cell, yielding mature glycoproteins bearing glycan moieties with widely different structures. This diversity reﬂects the roles of protein-linked oligosaccharides in mature glycoproteins, which are mainly involved in recognition phenomena. On the other hand, N-glycan processing in the ER is similar in all cells, and, as 303 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

304

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

will be described below, these reactions relate to the process of proper folding and assembly common to all glycoproteins. Addition of N-oligosaccharides has several effects on glycoprotein folding, at least two of which may be observed not only in vivo but also in folding assays performed in test tubes with pure glycoproteins: N-Oligosaccharides provide bulky, highly hydrophilic groups that help to maintain glycoproteins in solution during the folding process and, also, the presence of glycans modulates protein conformation by forcing amino acids close to the linking Asn unit to be in proximity with the water–glycoprotein interphase. This chapter will not deal with these effects, but will describe instead how N-glycan processing participates in the folding and quality control mechanisms operating in the ER. II. N-GLYCAN PROCESSING IN THE ENDOPLASMIC RETICULUM N-Glycosylation is initiated in most eukaryotic cells by the transfer, en bloc, of an oligosaccharide from a lipid (dolichol) derivative to polypeptidechainsastheyentertheERlumen.The precursor oligosaccharide has the composition Glc3Man9GlcNAc2 and the structure depicted in Figure 1 (1–3). This is a remarkably conserved structure in nature, as the same glycan is transferred in mammalian, plant, and fungal cells. Exceptions to this common pattern are seen in certain protozoa: Trypanosomatids transfer Man6GlcNAc2, Man7GlcNAc2, or Man9GlcNAc2, depending on the species (4 ), and Glc3Man5GlcNAc2 is transferred in Tetrahymena pyriformis (5 ) (Fig. 1). On the other hand, the complete canonical compound is transferred in another protozoan, Euglena gracilis (6 ). The consensus glycosylation sequence (Asn-X-Ser/Thr, where X may be any amino acid except for Pro) is necessary but is not always glycosylated. Apparently, N-glycosylation, which occurs when the Asn unit involved is about 10–12 amino acids from the inner ER membrane surface, is hindered if the polypeptide chain rapidly adopts a secondary structure (7, 8). Cell-free assays showed that, except for the enzyme from trypanosomatids, all other oligosaccharyltransferases transfer the triglucosylated glycan about 20- to 25-fold faster than Man9GlcNAc2 (9). This property, together with the strict distance between the receptor amino acid and the ER inner surface required for glycan transfer, determines a severe underglycosylation of glycoproteins synthesized by mutant cells unable to glucosylate the lipid-linked oligosaccharide (10). Folding is severely impaired in many (but not all) underglycosylated glycoproteins. The external Glc unit (residue n, Fig. 1) is immediately removed from the protein-linked oligosaccharide by glucosidase I (GI), a

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

305

FIG. 1. Structure of oligosaccharides. Structure of oligosaccharides transferred in wild-type mammalian, plant, fungal, and trypanosomatid cells. a–n indicate the order of addition of the monosaccharide units in the in vivo synthesis of Glc3Man9GlcNAc2P-P-dolichol. The entire oligosaccharide is synthesized in mammalian, plant, and fungal cells. Oligosaccharides formed by trypanosomatids that synthesize Man6GlcNAc2, Man7GlcNAc2, and Man9GlcNAc2 lack residues i–n, j –n and l–n, respectively. The oligosaccharide transferred in Tetrahymena pyriformis lacks residues h–k.

membrane-bound α(1-2)-glucosidase (Fig. 2) (11). The substrate donor (Glc3Man9GlcNAc2 -P-P-dolichol) is apparently not degraded by this enzyme, whereas the glycan is immediately deglucosylated following transfer, suggesting that GI might not be freely mobile in the ER membrane, but is probably intimately associated with the oligosaccharyltransferase in such a way that accessibility to the substrate donor is precluded. Further deglucosylation mediated by glucosidase II [GII, an α(1-3)-glucosidase] may remove the remaining Glc units (residues l and m in Fig. 1). GII is a soluble heterodimer composed of catalytic subunit and a noncatalytic subunit that contains an ER retrieval sequence at its C terminus (12, 13). Deglucosylation is independent of the conformation of the glycoprotein substrates and is a rapid process that occurs cotranslationally. Depending on the species, one or two Man units may be removed in the ER. Mammalian mannosidases I and II excise Man units i and k, respectively (Fig. 1) (14). Only one ER α-mannosidase occurs in Saccharomyces cerevisiae that yields the same product as mammalian α-mannosidase I

306

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

FIG. 2. Oligosaccharide processing reactions occurring in the ER in mammalian cells. N, M, G, Dol, Prot, OST, GI, GII, MI, MII, and GT stand for GlcNAc, Man, Glc, protein, dolichol, glucosidase I, glucosidase II, mannosidase I, mannosidase II, and UDPGIc:glycoprotein glucosyltransferase, respectively.

(15 ). Interestingly, a single mutation in S. cerevisiae ER α-mannosidase (R273L) renders the enzyme capable of degrading Ma9GlcNAc2 to Man5GlcNAc2 (16 ). Mannosidases appear to be absent in the ER in Schizosaccharomyces pombe (17 ). As will be described below, a severe disturbance not only in glycoprotein folding but also perhaps in the disposal of misfolded species could occur if a similar mutation occurred in the functionally equivalent mammalian ER α-mannosidase I. An additional N-glycan processing reaction occurring in the ER lumen is the transient reglucosylation of glycoproteins after complete deglucosylation by GI and GII to generate Glc1Man9GlcNAc2, Glc1Man8GlcNAc2, and Glc1Man7GlcNAc2 as reaction products (18–21), a reaction catalyzed by the UDP-Glc:glycoprotein glucosyltransferase (GT). The structure of the ﬁrst glycan is identical to that produced by partial deglucosylation of Glc3Man9GlcNAc2. GII is also responsible for the in vivo deglucosylation of the monoglucosylated compounds formed by GT. This enzyme is present not only in mammalian but also in plant, fungal (S. pombe, Mucor rouxii), and protozoan cells (21). Saccharomyces cerevisiae

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

307

FIG. 3. Oligosaccharide processing reactions occurring the ER in T. cruzi (A) and S. cerevisiae (B). For further indications, see the legend to Figure 2.

is the only organism known so far to lack GT (22, 23). For the sake of comparison, ER N-glycan processing reactions occurring in a trypanosomatid protozoans such as Trypanosoma cruzi and the yeast S. cerevisiae are depicted in Figure 3. Trypanosomatids have a GII-like activity capable of degrading both Glc1Man6–9GlcNAc and Glc2Man6–9GlcNAc in vitro (24). The fact that trypanosomatids do not possess oligosaccharides with two glucoses suggests that the only function of GII in these protozoa is to counteract the action of GT. Trypanosomatids lack GI activity (24), as these microorganisms do not form triglucosylated N-glycans. Interestingly, despite the many variations in the N-glycosylation pathway adopted by trypanosomatids, they seem to have selectively preserved the structures and enzymes required for transient reglucosylation (4). Several properties of GT will be described below, as this enzyme is a key sensor of the folding status of glycoproteins. The Golgi apparatus of species belonging to the chordate phylum, including placental and marsupial mammals, birds, reptiles, amphibians, and ﬁsh, as well as the Mollusca, in which it was detected in three distinct

308

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

classes, display an endomannosidase activity that degrades glucosylated N-glycans yielding Glc1–3Man (plus Man6–8GlcNAc2, depending on the substrate used) (25). This enzyme appears to ensure a complete removal of Glc units that would hinder formation of complex-type glycans. III. GLYCOPROTEIN R EGLUCOSYLATION The addition of a terminal glucose residue to protein-linked oligosaccharides was discovered while studying N-glycan processing in trypanosomatid cells pulse-chased with [14C]Glc, since, as mentioned above, unglucosylated oligosaccharides are transferred to proteins in these protozoa (18, 26). These in vivo observations were then extended to other eukaryotes, and it was soon established that transient reglucosylation of completely deglucosylated, protein-bound high-mannose–type oligosaccharides was a general glycoprotein processing reaction (19, 20 ). Since glucose is not normally found on mature glycoproteins, transient reglucosylation was intriguing, and the function of such apparently futile cycles of glucose addition and removal was not understood at the time. The reglucosylation of endogenous glycoproteins observed in vivo was reproduced in cell-free systems and the activity was localized by subcellular fractionation to the ER, where, as mentioned above, deglucosylated glycoproteins that are subject to reglucosylation are formed by the action of GI and GII (20, 21). When an in vitro assay for exogenous glycoprotein reglucosylation was developed, the substrate glycoprotein had to be denatured in order to be an acceptor in the reglucosylation reaction (21). This requirement was initially thought to reﬂect an inadequate accessibility of high-mannose–type N-glycans in native glycoproteins. However, the glycopeptide Man9GlcNAc2–Asn was a very poor acceptor for reglucosylation, indicating that oligosaccharide accessibility could not be the only limiting factor for the lack of reglucosylation of native glycoproteins (27 ). The poor recognition of short peptides was also observed for soybean agglutinin and phytohemagglutinin tryptic glycopeptides having 9 and 12 amino acids, respectively (27 ). A more heterogeneous mixture of tryptic glycopeptides derived from thyroglobulin was also not glucosylated by GT, unless held linked to the rest of the protease-treated protein by disulﬁde bonds (27 ). As several glycoproteins were tested as substrates for GT in vitro, it was noticed that none was an acceptor in its native state, even when the oligosaccharides were highly exposed, while all had to be denatured to be reglucosylated (27 ). These observations raised the radical notion that glycoproteins had to be devoid of a native conformation to be recognized as substrates by GT. Based on this unique property, it was suggested that

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

309

the enzyme (and monoglucosylated oligosaccharides) could be somehow involved in quality control and glycoprotein folding in the ER (21, 27, 28). The involvement of a protein recognition element in glycosylation reactions was not without precedent. While most glycosyltransferases display speciﬁcity at the carbohydrate level (sugar acceptor, sugar donor, and linkage formed), some glycosyltransferases also recognize the protein to which the substrate oligosaccharide is bound. The UDP-N-acetylglucosamine:lysosomal enzyme N-acetylglucosamine phosphotransferase selectively initiates formation of the Man 6-P epitope in lysosomal glycoproteins (29). Also, the UDP-GalNAc:glycoprotein hormone N-acetylgalactosaminyltransferase selectively adds a GalNAc residue to secreted hormones (30). The substrates recognized by GT in vitro did not appear to share any common feature, except for their nonnative conformation, as opposed to the other two glycosyltransferases, which maintain their rather narrow selectivity toward their respective glycoprotein substrates (lysosomal enzymes or glycoprotein hormones) when assayed in vitro. The requirement for nonnative conformations implied that a wide range of substrates could be reglucosylated in vivo, since most glycoproteins could potentially be recognized as substrates during their folding in the ER. Indeed, both soluble and membranebound glycoproteins could be reglucosylated in microsomes (31). Also, structural analysis of N-glycans from T. cruzi cells grown in the presence of GII inhibitors revealed that 50–60% of them had the monoglucosylated epitope (28), indicating that a substantial proportion of newly synthesized glycoproteins underwent transient glucosylation in vivo. A. The UDP-Glc:Glycoprotein Glucosyltransferase The reglucosylating enzyme was puriﬁed from rat liver and S. pombe (22, 32). The activity was present in low levels and very sensitive to proteases in crude extracts, residing in a single polypeptide of 170 kDa, signiﬁcantly larger than most glycosyltransferases, (22, 32, 33). While glycosidases and glycosyltransferases in the secretory pathway are type II membrane proteins, GT and its counteracting enzyme GII are so far the only soluble carbohydrate-processing enzymes described in the secretory pathway. The isolated GT had a neutral pH optimum and the same requirement for Ca2+ observed in crude extracts (22, 32, 33). Calcium could be partially replaced by Mn2+ but not by Mg2+, which is utilized by most other glycosyltransferases. Although the maximum activity is obtained in vitro with relatively high Ca2+ concentrations (5–10 mM), such a requirement is compatible with its location in the ER, which is

310

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

a dynamic intracellular reservoir of Ca2+. The requirement for Ca2+ by GT is not common to other glycosyltransferases, particularly those that are highly homologous to its C-terminal domain (see below). Another glucosyltransferase of the ER synthesizes the same structure produced by GT (Fig. 1), but on the precursor dolichol-P-Poligosaccharide (3). This enzyme is a membrane protein (denominated Alg6p in S. cerevisiae) that utilizes dolichol-P-Glc as the glucosyl donor, with a β-anomeric conﬁguration, which is in turn made on the cytosolic leaﬂet of the ER membrane from UDP-Glc and dolichol-P (31). In contrast, glycoprotein reglucosylation uses UDP-Glc as the sugar donor, and other sugar nucleotides (UDP-Gal, ADP-Glc, and TDP-Glc) are not substrates for GT (22, 32). The ER membrane contains a speciﬁc transporter for UDP-Glc from its site of synthesis (the cytosol) into the lumen of the ER, where the sugar nucleotide serves as substrate for glycoprotein reglucosylation (34). Further, a UDPase was recently described in the ER of mammalian cells (35) that probably facilitates transport of UDP-Glc into the lumen by an antiport mechanism, where entrance of sugar nucleotides into the ER or Golgi lumen is coupled to exit of the corresponding nucleoside monophosphates (36). The Glc residue is speciﬁcally transferred by GT to the terminal Man residue in the α(1-3)–α(1-2) branch (residue g in Fig. 1) (19, 21), and oligosaccharides containing a reduced number of Man residues on the α(1-6) branch (Man8GlcNAc2 and Man7GlcNAc2) are glucosylated at lower rates (50% and 15%, respectively) than the complete oligosaccharide (Man9GlcNAc2) (27 ). The decreased recognition of smaller oligosaccharides could be due to increased ﬂexibility of the α(1-3)–α(1-2) branch in the smaller structures as compared to Man9GlcNAc2, which appears to adopt a more rigid conformation (37). It is worth mentioning that GII also shows a decreasing activity toward monoglucosylated glycans of decreasing Man content (38 ). The cDNA for GT has been cloned (and its encoding protein characterized) from S. pombe, Drosophila melanogaster, rat, and human liver (33, 39–41). Homologs from other organisms can also be identiﬁed in databases (Table I). Identiﬁcation and sequencing of GT cDNAs conﬁrmed that the enzyme is widely expressed in eukaryotic organisms, and revealed several structural and functional properties. The cDNAs code for proteins of around 1500 amino acids, in agreement with the size of isolated enzymes (Table I). The primary sequences encode an N-terminal signal sequence and the mature proteins are devoid of transmembrane domains, in agreement with their behavior as soluble proteins (31). They contain C-terminal KDEL-related sequences

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

311

TABLE I Summary of GT and Related Protein Sequences Organism

Size (amino acids)

Accession number (Genebank)

GT activity

Reference

Schizosaccharomyces pombe Drosophila melanogaster Rattus norvegicus Homo sapiens (HUGT1) Homo sapiens (HUGT2) C. elegans 1 C. elegans 2 A. thaliana N. crassa S. cerevisiae Kre5p

1447 1548 1527 1555 1516 1493 1377 1674 1501 1365

S63669 Q09332 AAF67072 AAF66232 AAF66233 T16404 T19214 AAF26036 AC067938 BVBYK5

Yes Yes Yes Yes Noa n.d.b n.d. n.d. n.d. no

39 33 40 41 41 Direct Direct Direct Direct 42

aEvidence for lack of activity in HUGT2 is not as compelling as that for S. cerevisiae Kre5p. bn.d., not determined.

for retrieval to the ER. Moreover, several N-glycosylation consensus sequences are present, some of which appear to be glycosylated since concanavalin A–sepharose columns speciﬁcally retained the S. pombe, rat liver, and D. melanogaster enzymes (22, 32, 33). The ﬁrst GT cDNA described was that from D. melanogaster (33). It was originally cloned from an expression library using antibodies against an extracellular protein from conditioned media of Drosophila cell cultures. However, its primary sequence and the fact that it was later found to localize to the ER suggested, as conﬁrmed biochemically, that it corresponded to the D. melanogaster homolog of GT. The cDNAs for S. pombe and rat liver GT were cloned using sequence information derived from peptides obtained from the respective puriﬁed enzymes (39, 40). The cloning of GT from S. pombe revealed that it contained a heat shock response element in its 5′ untranslated region, and expression of its mRNA was indeed found to be induced under conditions that promote accumulation of misfolded proteins in the ER, such as heat shock and addition of Ca2+ ionophores, reducing agents, or tunycamicyn. Induction under similar conditions was also observed for the human GT, emphasizing the involvement of GT in quality control and protein folding in the ER (see below) (41). As several cDNAs coding for GT became available (see Table I), it was clear that they displayed only moderate conservation (30–40% identity, 40–60% similarity) along the entire protein (a higher conservation was observed among mammalian GTs, as human and rat homologs are ∼90%

312

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

identical) (40, 41). The C-terminal portions ∼270 amino acids) display the highest conservation between GTs (65–85% identity) (Fig. 4). This domain shares homology with bacterial glucosyl and galactosyl transferases that use UDP-Glc or UDP-Gal as sugar donors to generate α-linkages with retention of the anomeric conﬁguration, as GT does, but they differ in the acceptor speciﬁcity and bonds formed. The bacterial enzymes are ∼300 amino acids long, suggesting that the highly conserved C-terminal domain of GT corresponds to a complete glucosyltransferase domain. Several residues are highly conserved within this domain between bacterial enzymes and GT, and were shown to be required for reglucosylating activity in vivo and in vitro (Fig. 4). A recombinant fragment of 37 kDa from the C-terminal domain of rat GT displayed some residual activity and could be photoafﬁnity-labeled with UDP-Glc analogs, suggesting that this C-terminal glucosyltransferase is an autonomous domain within the protein (40). A stretch of ∼200 amino acids that are also well conserved between GTs links the highly conserved C-terminal domains to the N-terminal stretch of ∼800 amino acids that comprise the least conserved part of the molecules. The reason behind the strong divergence at the N terminus is unclear, but is reminiscent of what occurs in the Hsp70 chaperone family, where the N termini, containing the ATPase domain, have a high degree of homology, whereas the C-terminal portions, containing the hydrophobic peptide binding sites, are much less conserved (44). The KRE5 gene is the closest GT homolog in S. cerevisiae but displays less similarity than other GT cDNAs (20% as opposed to the 30–40% identity seen between bona fide GT cDNAs) (42). Kre5p is also a soluble protein of the ER whose size and domain organization are similar to those of GTs, the highest similarity also being found in the glycosyltransferase-like C-terminal portion. However, Kre5p lacks three of the four Cys residues conserved in the glucosyltransferase domains of all known active GTs (Fig. 4) (33, 39, 40, 42). In addition, some of the key residues that have been shown to be critical for glycoprotein reglucosylation are absent from Kre5p (Fig. 4) (40, 41, 42, 44). The −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ FIG. 4. Comparison of the C-terminal portions of GTs and related sequences from Schizosaccharomyces pombe (pom), Homo sapiens HUGT1 (hum), Rattus norvegicus (rat), Drosophila melanogaster (dro), and Saccharomyces cerevisiae (kre). The residues conserved with some bacterial glycosyltransferases that utilize UDP-Glc or UDP-Gal as sugar donors are highlighted with an asterisk. The conserved Cys residues are indicated with a dot. Residues that abolished GT activity when mutated are circled and correspond to D1334A, D1336A, Q1429A, and N1433A (rat); D1452A, Q1453A, D1454A, and L1455A (human HUGT1), and Y1312A plus D1352A (S. pombe) (D1428 in the rat enzyme corresponds to D1452 in the human and D1352 in the yeast GTs) (Fig. 4) (40, 41, 43 ).

314

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

divergence of its C-terminal domain (∼22% identity as opposed to the >65% identity between GTs) probably explains the evidence showing that Kre5p is not a functional GT. Kre5p is not essential in certain backgrounds, but deleted strains grow very poorly (42). Expression of S. pombe GT in Kre5p-minus mutants did not correct the slow growth rate phenotype, thus conﬁrming that GT and Kre5p have different functions (45). The function of Kre5p is currently unknown, but indirect evidence suggests it maybe involved in β-1,6 glucan biosynthesis (42). Drosophila melanogaster contains a single GT-encoding gene, but in other multicellular organisms more than one GT homolog can be identiﬁed (33). A human GT homolog (HUGT2) shows similarity along the entire protein with other GTs, but differs from the second human GT (HUGT1), being less conserved (53% identical to rat GT, as opposed to ∼90% identity between rat GT and HUGT1). Also, unlike HUGT1, it is not induced under ER stress conditions (41). The C-terminus glucosyltransferase domain in HUGT2 has only two key deviations from canonical GTs (G1409R and H1461Y; the numbers correspond to the HUGT1 sequence. See Ref. 41), but sequence variations accumulate in its N-terminal domain. Despite its highly conserved glucosyltransferase domain, GT activity could not be detected in lysates from HUGT2 overexpressing cells (41). Two GT homologs have also been found in C. elegans, but none of these genes has yet been characterized (Table I). The C-terminal glucosyltransferase domain in both genes is extremely conserved, but they are also highly dissimilar in their N terminus. Even though the evidence for the lack of GT activity in HUGT2 is less compelling than in the case of Kre5p, these examples may represent a group GT-like ER proteins of unknown function. B. Recognition of Nonnative Glycoproteins The most signiﬁcant ﬁnding in the isolation of GT was that the puriﬁed enzyme maintained its selectivity for nonnative conformations in its substrates. While it may still cooperate with other factors to recognize nonnative conformations in vivo, the ﬁnding that pure GT exclusively reglucosylates nonnative glycoproteins strongly suggests that the discrimination between native and nonnative conformations is performed primarily by the GT itself. No cofactors were found to modify the activity of isolated GT in vitro, and expression of GT cDNA is sufﬁcient for glycoprotein reglucosylation in cells devoid of such activity (45). Isolated GT from rat liver, S. pombe, and D. melanogaster, as well as recombinant GT from rat and human liver glucosylate only nonnative glycoproteins; however, owing to its availability, the enzyme from rat liver was employed in the studies on substrate speciﬁcity discussed below.

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

315

One portion of the nonnative glycoproteins substrates recognized by GT is the innermost GlcNAc unit that links the oligosaccharide to Asn residues on glycoproteins (residue a in Fig. 1). Denatured glycoproteins bearing only the GlcNAc residues attached to Asn (obtained by digestion with endo-β-N-acetylglucosaminidase H, Endo H) were potent inhibitors of GT-mediated glucosylation, and the GlcNAc–Asn structures inhibited GT activity only when presented by nonnative conformations (46). Identical denatured proteins lacking the GlcNAc residues had no effect on GT. Since the innermost GlcNAc is often intimately associated with the surrounding polypeptide backbone in native glycoproteins, the recognition of the GlcNAc–Asn bond could be an interesting way of interfacing oligosaccharide and protein requirements (47 ). Inhibition experiments with different high-mannose–type oligosaccharides (Man4–6GlcNAc2, ManGlcNAc2, or GlcNAc) bound to the same nonnative protein backbone showed that all forms were capable of inhibiting GT, but their inhibitory potency was not compared (48). However, a model in which GT evaluates the conformation of its substrates solely by the accessibility or the ﬂexibility of their GlcNAc–Asn region seems unlikely. Such a structure is fully exposed in isolated glycopeptides containing only one or a few amino acids, which are very poor GT substrates (27 ). The involvement of GlcNAc–Asn bonds bound to protein structures was also studied by coupling high-mannose–type glycopeptides via a chemical crosslinker to a nonglycosylated protein backbone with either folded or incompletely folded conformations (46). These neoglycoproteins contain an artiﬁcial oligosaccharide–protein bond, but display a fully accessible GlcNAc–Asn group. Only the nonnative conformers of these chimeras were glucosylated, demonstrating that while the GlcNAc– Asn group must be accessible, it must also be present on an improperly folded protein structure. A similar behavior was observed with RNase B, a glycoprotein with a single N-glycan whose innermost GlcNAc interacts with the surrounding polypeptide in the native conformation, rendering it inaccessible to endoglycosidases (48). Enhancing the exposure of the GlcNAc–Asn bond in an altered but largely native conformer of RNase B (see below) to the point of making it fully susceptible to endoglycosidases was not sufﬁcient to turn it into a substrate for GT. All these observations indicate that although exposure of the GlcNAc–Asn bond is a critical element for GT recognition, it is not sufﬁcient condition for reglucosylation. The information available on misfolded glycoprotein–GT interactions emphasizes that the oligosaccharide and the protein part are recognized simultaneously by GT. In fact, it was demonstrated that the oligosaccharide must be covalently bound to a nonnative protein backbone to be a substrate for GT; and since the products of at least two

316

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

different artiﬁcial bonds between the glycopeptides and the proteins were efﬁciently glucosylated [the amino group of Man9GlcNAc2–Asn linked either to the amino groups in amino acid side chains by glutaraldehyde or to sulfhydryl groups in Cys residues by N-succinimidyl 3-(2-pyridyldithio)propionate], a close proximity of the protein and the carbohydrate appears to be required rather than a speciﬁc chemical structure (46). At the same time, the amino acids forming the N-glycosylation consensus sequence do not appear to be absolutely required, as the proteins to which the glycans were covalently attached (β-lactoglobulin and staphylococcal nuclease) lacked such sequences. However, no direct comparison between similar nonnative neoglycoprotein conformers with different GlcNAc–protein couplings has been conducted. Although portions of the acceptor oligosaccharide are recognized by GT, the most distinctive feature of GT is its ability to discriminate between native and nonnative conformations in its substrates. This property is typically exhibited by molecular chaperones, but GT is unique in that it covalently modiﬁes its targets. Chaperones of the Hsp70 family recognize short peptide fragments rich in hydrophobic amino acids, which are normally buried in native conformation and become exposed in nonnative structures (49). GT is able to bind hydrophobic peptides under physiological pH and salt concentrations, although it is unclear whether this property is an intrinsic part of the mechanism by which it senses protein conformations (46). If portions of the polypeptide chain buried in native but exposed in misfolded conformations were the structural elements recognized by GT, it is unlikely that the enzyme would exhibit a signiﬁcant primary sequence speciﬁcity. Soybean agglutinin was subjected to extensive chemical modiﬁcation without affecting its Glc acceptor capacity (50). While some critical primary sequence elements may not have been modiﬁed in those experiments, the wide range of substrates analyzed in cell-free assays, including even bacterial proteins, makes the recognition of primary sequence elements rather unlikely. The experimental evidence reviewed above has established that a proper N-glycan is glucosylated by GT only when it is covalently bound to a protein with a nonnative conformation, but it remains unclear what feature(s) in the protein conformation are recognized by GT. How does GT distinguish between native and nonnative conformations? The answer to this question remains elusive at the moment. This is due in part to the experimental limitations derived from the nature of GT substrates, since nonnative proteins are heterogeneous and tend to behave in illdeﬁned ways that are hard to control and characterize. Many of the GT substrates used in vitro are in aggregates or are prone to aggregation,

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

317

limiting the experiments and the conclusions that can be drawn with them. Some also have more than one oligosaccharide, making the interpretations even more cumbersome. Ideally, the use of small, welldeﬁned monomeric proteins carrying a single oligosaccharide should allow the design and interpretation of experiments to explore the way GT recognizes nonnative structures. However, there are very few naturally occurring glycoproteins with these characteristics; consequently, neoglycoproteins have proved a useful tool. In one example of such an approach, staphylococcal nuclease, a wellcharacterized small bacterial protein, was converted into a neoglycoprotein by attaching a glycopeptide to single Cys residues engineered in selected portions of the molecule (46). This system demonstrated that GT is an exquisite sensor of subtle differences in protein conformation. A native full-length version of the neoglycoprotein having the glycan attached to Cys-70 was a poor substrate for GT, but the addition to the reaction mixture of diphosphothymidine (pdTp), an inhibitor of the nuclease activity known to stabilize the native conformation, converted the neoglycoprotein into an even poorer substrate. A truncated version of the nuclease lacking the last 14 amino acids at its C terminus is known to be in a compact but disordered structure. Proper folding can be favored, however, by addition of pdTp or the nuclease substrate (DNA). The truncated neoglycoproteins having the glycan attached to Cys-70 or Cys-124 were GT substrates without denaturation, and were recognized nearly as well as the denatured full-length neoglycoprotein. Addition of pdTp to incubation mixtures containing the truncated neoglycoproteins decreased but did not abolish their glucose acceptor capacities, suggesting that the glycans prevented neoglycoproteins from adopting fully native structures. Nevertheless, the CD spectra of the truncated neoglycoproteins in the presence of pdTp were superimposable with that of the full-length nuclease, indicating that the three species had similar secondary structures. Moreover, the nuclease activities of the truncated neoglycoproteins were about one-third to one-fourth of that of the full-length species. In the presence of DNA or pdTp, the truncated neoglycoproteins could be in equilibrium between two conformations, a native one with nuclease activity and a nonnative one that could be glucosylated by GT. As both conformations may not differ signiﬁcantly, these results demonstrated that GT is able to glucosylate glycoproteins with highly structured nativelike conformations. Cys-70 and Cys-124 were in opposite locations of the molecules and both truncated neoglycoproteins were similarly glucosylated by GT, indicating that the precise position of the oligosaccharide has little inﬂuence on recognition by GT, provided the glycan is linked to a protein domain

318

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

with a nonnative conformation (46). A similar pattern was found with RNase B, in which the oligosaccharide was moved around the molecule and appeared to be glucosylated as long as it was linked to a nonnative protein backbone (A. Helenius, personal communication). However, only N-glycans on nonnative domains were glucosylated by GT (51). The involvement of the innermost GlcNAc residue in the recognition of nonnative conformers by GT ensures that nonglycosylated proteins devoid of their native conformations would not interfere with GT function. At the same time, the requirement that the N-glycan must be covalently linked to a nonnative protein backbone suggests that GT senses protein conformations mostly locally by modifying oligosaccharides present on misfolded domains. This requirement indicates that GT could evaluate the folding status of discrete portions of large glycoproteins containing independently folding domains. This ability might underlie the conservation of some N-glycosylation sites required for efﬁcient assembly but not for biological activity of certain glycoproteins. In an attempt to evaluate the conformations recognized by GT, three different conformers of RNase B were used to directly compare the interaction of GT with different conformations of the same glycoprotein (48). Using this deﬁned set of conformers, which remained soluble and monomeric, a preference of GT for partially structured nonnative conformations rather than unstructured polypeptides was noticed. The ﬁrst one, designated RNase BS, contained a peptide bond cleaved between residues 20 and 21, yielding a perturbed but largely native conformation that preserved full ribonuclease activity, but was sensitive to proteases that do not attack native RNase B. The GlcNAc–Asn bond was fully accessible to endoglycosidases as opposed to what happened in the native conformation. The loose but largely native structure of RNase BS was not recognized by GT. A second conformer of RNase B was obtained by removal of the N-terminal peptide (S-peptide, residues 1–20), yielding a fragment (residues 21–124, RNase BSProt) containing the remainder of the RNase B molecule with its four disulﬁde bonds and the oligosaccharide. This form, devoid of ribonuclease activity, became a good substrate for GT. It was also very sensitive to endoglycosidases and proteases, but was probably largely structured owing to the constraints of having all four disulﬁde bonds intact. This conformation could be reversed by readdition of the S-peptide to regenerate the nativelike RNase BS′ , with recovery of full ribonuclease activity and concomitant loss of recognition by GT. Interestingly, a fully unfolded conformer of RNase B obtained by reduction and alkylation of the disulﬁde bonds was markedly less recognized by GT than the partially folded RNase BSProt. A similar lack of reglucosylation of fully unfolded species was observed

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

319

for reduced glycosylated forms of RNase and bovine trypsin inhibitor, which were not reglucosylated when translated and translocated into microsomes in vitro (52). Both model substrates described above (staphylococcal nuclease and RNase B) revealed that highly structured nonnative conformations not grossly departing from the native state can be efﬁciently glucosylated by GT. The RNase B system showed, in addition, that GT can discern between different conformers of the same glycoprotein, with an apparent preference for partially structured conformations over fully unfolded forms. However, since all the in vitro studies on the speciﬁcity of GT were done using GT puriﬁed from rat liver, it is not known whether other GTs recognize nonnative proteins in similar ways. The high conservation at the C terminus probably reﬂects the requirements of the sugar transfer reaction (recognition of the sugar donor and the acceptor N-glycan), but the variability seen in the large N-terminal domain may imply functional variation. While the experimental evidence that GT prefers partially structured conformations is still based on only a few examples, and consequently may reﬂect a bias toward the substrates and the conditions used in vitro, it seems to agree with the observations obtained from experiments in living cells. A chimeric glycoprotein consisting of growth hormone fused to a piece of inﬂuenza virus hemagglutinin (HA) failed to leave the ER and was found to accumulate intracellularly and to contain monoglucosylated oligosaccharides (Glc1Man7–9GlcNAc2) (53). A more detailed study later revealed that at a nonpermissive temperature, the G protein of the mutant vesicular stomatitis virus strain ts045 (VSVts045-G) was reglucosylated posttranslationaly. This mutant failed to fold properly at 41◦ C and was retained in the ER with ∼50% of its oligosaccharides dynamically alternating between the monoglucosylated (Glc1Man9GlcNAc2) and the deglucosylated state (Man9GlcNAc2). On shifting to the permissive temperature (30◦ C) VSVts045-G protein was properly folded and was no longer reglucosylated (54). This was the ﬁrst conclusive demonstration of selective reglucosylation of a glycoprotein that failed to fold properly in vivo. When folded at the permissive temperature, VSVts045-G could also be reglucosylated on shifting to the nonpermissive temperature (55); this indicates that GT is able to glucosylate glycoproteins at advanced stages of folding in vivo (56), since VSVts045-G synthesized at the permissive temperature contained most or all of the disulﬁde bonds. Furthermore, the fact that formation of monoglucosylated oligosaccharides in Trypanosoma cruzi depends exclusively on GT activity (Fig. 3A) made it possible to establish that cruzipain, a lysosomal proteinase having two to three N-oligosaccharides and six to seven disulﬁde bridges,

320

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

was glucosylated by GT only when most or all of its disulﬁde bridges had been formed (57), indicating again a requirement of at least a partially structured molecule for GT-mediated reglucosylation in vivo. A preference of GT for distinct nonnative conformations is somewhat supported by the ﬁnding that protein misfolding induced by dithiothreitol (DTT) in a S. pombe mutant that transferred Man9GlcNAc2 to nascent polypeptides (thus synthesizing underglycoylated glycoproteins) did not result in a generalized increase in glycoprotein reglucosylation (58). This could be due to the fact that DTT prevented glycoproteins from reaching the hypothetical advanced folding stages required for GT-mediated glucosylation and/or that in the mutant cells employed, interaction of folding glycoproteins was displaced toward interaction with chaperone systems other than GT-CNX/CRT (see below). It is worth mentioning that glycoprotein underglycosylation leads to a generalized upregulation of the unfolded protein response with concomitant induction of various chaperones and folding factors, and as will be further described below, different conditions may determine whether glycoproteins interact with one or another of the available ER chaperone systems such as BiP or GT-CNX/CRT. IV. CHAPERONES AND PROTEIN FOLDING IN THE ENDOPLASMIC RETICULUM Proteins enter the secretory pathway in the ER, where they are covalently modiﬁed (cleavage of the signal peptide, N-glycosylation, formation of disulﬁde bonds) and acquire their proper tertiary and, in some cases, quaternary structures. Proteins that fail to fold correctly are initially retained in the ER and eventually transported back to the cytosol, where they are degraded by the proteasome (59–61). The mechanisms whereby cells assist and monitor the proper folding and assembly of newly synthesized proteins have been referred to as quality control (62, 63). Protein folding in the ER is facilitated by a battery of molecular chaperones (BiP/GRP68, GRP94, GRP170, etc.) and by proteins that facilitate Pro cis–trans isomerization or disulﬁde bond formation (protein disulﬁde isomerase PDI or ERp59, ERp72, ERp57, etc.) (64–70). The lumen of the ER is an oxidizing environment that promotes formation of disulﬁde bonds and is the main cellular Ca2+ reservoir. The function of several chaperones and folding factors is dependent on Ca2+ concentration (71). Calnexin (CNX) and calreticulin (CRT) represent a novel class of molecular chaperones resident in the ER (72–74 ). Calnexin (also termed p88 and IP90) was described by three different groups as a protein that was transiently associated in the ER with class I major

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

321

histocompatibility complex (MHC) molecules, with membrane-bound immunoglobulins, and T-cell antigen receptor molecules or that could be phosphorylated in an isolated microsomal fraction (72, 73). CNX is a type I, 572–amino acid transmembrane protein. The 89–amino acid cytosolic tail has an ER-localization sequence at its C terminus (RKPRRE). The transmembrane domain is composed of a 21–amino acid stretch followed by the lumenal portion. The middle region of this portion, called the P domain, contains two motifs: motif 1, IXDP(D/E)(A/D)XKP(E/D)DWD(D/E), repeated tandemly four times, followed by motif 2, GXWXXPXIXNPXY, again repeated tandemly four times. There is a high-afﬁnity, low-capacity Ca2+ binding site in the region encompassing the four motif 1 repeats. In addition, Ca2+ appears to bind with low afﬁnity to the acidic cytosolic and N-terminal portions (75). A Pro-rich region within the P domain also has lectinlike properties able to bind monoglucosylated high-mannose–type oligosaccharides (see below) (76). Calreticulin is a 400–amino acid soluble protein that has an ER retrieval signal at its C terminus (74). The middle or P domain is highly similar to the respective CNX domain. Instead of four, it has only three tandemly repeated motifs 1 and 2. It has a high-afﬁnity, low-capacity and a low-afﬁnity, high-capacity Ca2+ binding site at the P domain and at the extremely acidic C-terminal domain, respectively. As will be discussed below, the lectin properties of the P domain are indistinguishable from those of CNX (76). Chemical crosslinking studies conducted on intact CHO cells showed that CNX and CRT form part of a large, weakly associated, heterogeneous protein network including BiP, GRP94, and other ER-resident proteins (77). V. INTERACTION OF GLYCOPROTEINS WITH CALNEXIN AND CALRETICULIN Several glycoproteins (α 1-antitrypsin, α 1-antichymotrypsin, α-fetoprotein, complement 3, transferrin, and apoB-100) were coimmunoprecipitated with anti-CNX antibodies from metabolically labeled HepG2 cells. No interaction between CNX and glycoproteins was detected when N-glycosylation was prevented by tunicamycin (this antibiotic prevents N-glycosylation by inhibiting the ﬁrst step in the formation of dolicholP-P derivatives) (78). Albumin, the major nonglycosylated protein secreted by HepG2 cells, did not associate with CNX. Interaction of CNX with glycoproteins was transient, correlated with their half-times of secretion, and could be prolonged when amino acid analogs that interfered with proper folding were introduced into glycoproteins. It was also described that CNX associated transiently with incompletely assembled

322

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

class I MHC molecules before they were fully folded and transported to the Golgi (79, 80). Mutant cells unable to form the complete complexes (because of lack of β 2m and/or defective peptide loading) showed prolonged class I MHC complex–CNX association, and again, the rates of exit from the ER of different isotypic complexes and the association of CNX with the complexes varied in a similar fashion. These results strongly suggested that CNX could be involved in quality control, regulating exit of class I MHC molecules from the ER. From these and similar experiments it was concluded that CNX (and also CRT) interacted with newly synthesized glycoproteins only transiently during their folding, behaving in vivo as bona fide molecular chaperones. A seminal observation was that glycans had to be partially deglucosylated for glycoprotein–CNX/CRT interaction to occur (81, 82). It was noticed that HA and VSV-G protein associated with CNX during folding, and not only tunicamycin, but also glucosidase inhibitors prevented the interaction. Inhibitors of ER α-mannosidase I had no effect, indicating that CNX binding depended primarily on the presence of partially deglucosylated N-glycans. That the N-glycans had to be monoglucosylated was suggested by the interaction between CNX and VSVts045-G protein, known to be folding-deﬁcient and to possess monoglucosylated oligosaccharides added posttranslationaly at the nonpermissive temperature (54, 81). Further work performed in vitro with a rabbit reticulocyte lysate translation system supplemented with dog pancreas microsomes conﬁrmed that indeed CNX-bound HA had monoglucosylated oligosaccharides that were generated either by partial deglucosylation of the transferred oligosaccharide or by posttranslational reglucosylation of glucose-free compounds (82). Moreover, release of the glycoprotein from CNX required removal of the remaining glucose by GII. Incompletely folded glycoproteins were also found to interact transiently and in a monoglucosylated-dependent manner with CRT, the CNX soluble homolog (83–85). From the long list of glycoproteins reported to associate with CNX/CRT, it may be concluded that most glycoproteins, irrespective of their ﬁnal intracellular destination or soluble or membrane-bound status, transiently interact with CNX and CRT in mammalian cells (83–100). In all cases tested, addition of GI/GII inhibitors (such as castanospermine and 1-deoxynojirimycin or its N-methyl or N-butyl derivatives) prevented not only interaction of CNX/CRT with folding glycoproteins in mammalian cells but also dissociation of already formed glycoprotein–CNX/CRT complexes (55, 82–96, 101 ). Furthermore, no glycoprotein–CNX/CRT interaction was observed in GI- or GII-deﬁcient cell lines(89, 102).Asexpectedfromdifferencesobservedinmammalian

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

323

and T. cruzi N-oligosaccharide processing pathways (Figs. 2 and 3A), addition of 1-deoxynojirimycin to the protozoan cells actually enhanced glycoprotein–CRT interaction (trypanosomatid cells lack CNX) (57). VI. CALNEXIN AND CALRETICULIN ARE LECTINS SPECIFIC FOR MONOGLUCOSYLATED OLIGOSACCHARIDES That both CNX and CRT behaved as lectins was later shown by in vitro binding studies with isolated oligosaccharides. Only monoglucosylated oligosaccharides were retained by immobilized CNX or CRT from a mixture of labeled high-mannose–type compounds (57, 76, 103, 104). Binding was optimal for Glc1Man9GlcNAc2, and compounds having a lower Man content showed a diminished binding capacity, with Glc1Man5GlcNAc2 (residues a–g and l in Fig. 1) still showing about 65% of the binding seen for Glc1Man9GlcNAc2. On the other hand, Glc1Man4GlcNAc2 (residues a–d, f, g, and l in Fig. 1) was not retained by CNX or CRT. This indicated that the α(1-6) branch is important for recognition (76, 104). Both Glcα(1-3)Man and Glcα(1-3)Glc inhibited binding of Glc1Man9GlcNAc2 to the lectins to a similar extent (76). The former disaccharide is present at the nonreducing end of the oligosaccharide probe, whereas the latter one is present in Glc2Man9GlcNAc2, which is not retained by the lectins, suggesting a role for the polymannose core in CNX/CRT–glycan interaction. No differences between the oligosaccharide binding properties of CNX and CRT were found. CNX/CRT–glycan binding required Ca2+, and the presence of adenosine nucleotides or peptides linked to the glycan moiety had little effect on it (76). Studies of monoglucosylated chicken IgG–CRT interaction by surface plasmon resonance revealed a micromolar association constant and indicated a single oligosaccharide binding site (105). Whereas free Glc1Man9GlcNAc2 inhibited the interaction, no effect of Man9GlcNAc2 was observed, attesting to its exquisite speciﬁcity. Although interaction of glycoproteins with CNX/CRT in vivo is restricted to species that do not display their native structure, experiments performed in vitro with puriﬁed glycoproteins and lectins showed that, independent of the protein conformation, the presence of monoglucosylated saccharides is not only a necessary but also a sufﬁcient condition for glycoprotein–CNX/CRT interaction (52, 57, 105, 106). In addition, optimal coprecipitation with CNX/CRT was observed with glycoproteins having at least two monoglucosylated glycans. Creation of an additional N-glycosylation site in glycoproteins having a single one (as in RNase B, the human erythrocyte anion exchanger or human class I MHC heavy chain) or elimination of one site in glycoproteins having two of

324

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

them (VSV-G protein) resulted in a respective increase or decrease in CNX/CRT binding (52, 107, 108). Similarly, the presence of at least two of the canonical four N-glycans was reported to be required for tyrosinase–CNX interaction (109). A. Differential Binding to Calnexin and Calreticulin In Vivo While assays performed so far with isolated components showed no differences in the binding properties of both lectins, glycoprotein– CNX/CRT interactions are not equivalent in vivo. The pattern of glycoproteins coprecipitated with CNX or CRT from mammalian cells only partially overlapped (83, 99). This difference might be related to the soluble or membrane-bound status of CRT and CNX, and to the relative position of glycans in membrane glycoproteins. Oligosaccharides located closer to the ER membrane would interact more easily with CNX, whereas those more lumenally oriented would preferentially associate with CRT. In fact, similar patterns of glycoproteins were found to interact with CRT and a truncated, soluble fragment of CNX, or with the full-length version of CNX and CRT artiﬁcially anchored to the ER membrane (86, 110). Moreover, oligosaccharides located in the more lumenally oriented top/hinge domain of HA preferentially affected binding to CRT, whereas CNX was less discriminating but mainly bound glycans close to the ER membrane (111). Also, interaction of CNX with the human erythrocyte anion exchanger, a protein having 12–14 transmembrane segments, was not signiﬁcantly modiﬁed when the single N-glycan present in wild-type species was moved to other short extracytosolic loops (97). Interestingly, human class I MHC heavy chain (a membrane glycoprotein) interacts ﬁrst with CNX during its biosynthesis, but on association with β 2m, CRT binds to the heavy chain/β 2m complex until the assembly process is complete (88, 89, 112–116). This sequential interaction would suggest that the oligosaccharide on the heavy chain is ﬁrst close to the ER membrane exposed to CNX, but a change in the heavy chain conformation resulting from β 2m binding would make it more accessible to the ER lumen resulting in increased binding to CRT. B. Consequences of Glycoprotein–Calnexin/Calreticulin Interactions The selective association of CNX/CRT with incompletely assembled glycoproteins and the close correlation of such interactions with their residence in the ER indicated that the lectins participated in the retention of incompletely folded glycoproteins (117 ). Together with the selective reglucosylation of incompletely folded glycoproteins, these

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

325

FIG. 5. Model proposed for the quality control of glycoprotein folding. Protein-linked Glc3Man9GlcNAc2 is partially deglucosylated to the monoglucosylated derivative by GI and GII, and this structure is recognized by CNX/CRT. Man9GlcNAc2 is glucosylated by GT if complete deglucosylation occurs before lectin binding. The glycoprotein is liberated from the CNX/CRT anchor by GII and reglucosylated by GT only if not properly folded. On adoption of the native tertiary structure, the glycoprotein is released from CNX/CRT by GII and not reglucosylated by GT.

observations were conceptualized by A. Helenius and co-workers in a model for quality control of glycoprotein folding depicted in Figure 5 (81). According to this model, monoglucosylated glycoproteins generated either by partial deglucosylation of the transferred glycan or by GT-mediated glucosylation interact with CNX/CRT. This interaction would be followed by a shuttle between glucosylated (CNX/CRT-bound) and deglucosylated (CNX/CRT-free) forms catalyzed by the opposing activities of GT and GII. After acquiring their proper tertiary structure, glycoproteins would be deglucosylated by GII but not reglucosylated by GT and would no longer bind to CNX/CRT. Permanently misfolded molecules would continue to interact with the lectins, which would not be sensing the conformation of its glycoprotein ligands. This task is reserved to the reglucosylating enzyme, and cycles of binding and release to the lectins would be mediated by independently acting enzymes that covalently modify oligosaccharides. Cells have other, alternative and potentially overlapping mechanisms applicable to both glycosylated and nonglycosylated proteins for the ER retention of misfolded conformers, such as binding to other chaperones or formation of reversible disulﬁde bonds with matrix proteins of the ER (63, 118, 119).

326

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

Glycoprotein–CNX/CRT interaction not only prevents ER exit of misfolded conformers but also decreases the folding rate, increasing the folding efﬁciency of glycoproteins by preventing their premature oligomerization and degradation and by facilitating formation of native disulﬁde bonds. Simultaneous expression of CNX increased assembly of class I MHC heavy chains molecules with β 2m in D. melanogaster cells (87). A reduction in the level of aggregates and the use of conformational monoclonal antibodies revealed that this was due to a higher efﬁciency of heavy chain folding. Similarly, preventing glycoprotein– CNX/CRT interactions using GI/GII inhibitors increased the folding rate of HA expressed in an in vitro translation/tanslocation system, although the overall efﬁciency decreased due to aggregation, premature trimerization, and degradation as well as to formation of nonnative disulﬁde bridges (120). Moreover, glucosidase inhibitors produced similar effects on folding, disulﬁde bond formation, and productive homodimerization of the human insulin receptor expressed in CHO cells (94). Similar results were observed in S. pombe GII minus mutants, in which carboxypeptidase Y arrived at a higher rate but in decreased amounts to the vacuole compared to wild-type cells (13). How do glycoprotein–lectin interactions have such effects on glycoprotein folding? CNX/CRT could help to maintain glycoproteins in solution and prevent aggregation by binding to its oligosaccharides at particular stages of folding. The lectins also function to recruit other components of the ER folding machinery. A key observation in this regard was the discovery that CNX and CRT recruit ERp57, a thiol oxidoreductase in the ER lumen (121–123). ERp57 exclusively interacted with monoglucosylated glycoproteins (121–123) and was found together with CNX and CRT during the assembly of the class I MHC complex (124–127). Functional evidence for the role of ERp57 in the oxidative refolding of monoglucosylated RNase B in concert with CNX or CRT was shown in a cell-free system (128); moreover, mixed disulﬁde-bonded species between ERp57 and monoglucosylated glycoproteins could be demonstrated in vivo (129). CNX and CRT thus appear to bring together ERp57 and glycoproteins to facilitate formation of native disulﬁde bonds. However, additional mechanisms may be involved since ERp57, through its association with CNX/CRT, also interacts with glycoproteins lacking Cys residues (121, 122, 130). While cycles of deglucosylation and reglucosylation have been shown to mediate cycles of binding and release to CNX/CRT, the ﬁrst round of binding could be the most critical step. As mentioned above, VSVts045-G associated with CNX/CRT at the nonpermissive temperature, and while a higher folding rate was observed on lowering the temperature in the

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

327

absence of GI/GII inhibitors (a condition that allowed the shuttle to occur), the overall folding efﬁciency was similar whether or not the inhibitors were added (55). A more drastic effect on VSV-G folding was observed if access to CNX/CRT was precluded in the ﬁrst step (55 ). Similarly, in a S. pombe alg6 mutant, GT-mediated reglucosylation but not GII-mediated deglucosylation was essential for viability at high temperatures (see below) (43). It is now well established that the initial interaction between glycoproteins and CNX/CRT is mediated by the recognition of monoglucosylated glycans; in all cases tested, the interaction is precluded by the addition of GI/GII inhibitors. There are several reports, however, suggesting that the initial recognition is followed by a direct protein–protein interaction between CNX/CRT and their ligands. According to this alternative model, dissociation of the glycoprotein–CNX/CRT complexes would follow a change in conformation in the glycoprotein, in CNX/CRT, or in both. The existence of direct protein–protein interactions in the complexes has been largely inferred from coimmunoprecipitation studies with CNX/CRT, the technique most commonly used for studying CNX/CRT–ligand interactions in living cells. The persistence of the glycoprotein under study in immunoprecipitates treated with glycosidases that disrupt the lectin–glycoprotein association (Endo H, N-glycanase, GII) has been taken as an indication of direct protein–protein interactions. Despite its convenience, coimmunoprecipitation is not the best technique for studying interactions in which at least one of the species involved has not yet attained its proper tertiary structure. Such proteins are structurally unstable and may tend to precipitate nonspeciﬁcally together with immunocomplexes. Detergents pose an additional drawback, as CNX and its ligands could remain in the same detergent micelle after the glycosidase treatment and thus coprecipitate even when not directly interacting. Furthermore, in certain cases removal of N-oligosaccharides from glycoprotein–CNX/CRT complexes prior to immunoprecipitation precluded coprecipitation, whereas the same enzymatic treatment performed on preformed immunocomplexes failed to displace the glycoprotein from the protein A–sepharose beads, suggesting that a protein–bead interaction might be responsible, in some cases, for the apparent protein–protein interactions. No coimmunoprecipitation after glycosidase treatment was observed when the interactions of CNX/CRT with bovine pancreas RNase B translated in a rabbit reticulocyte–dog pancreas microsome system, of RNase B with isolated soluble truncated CNX, or of cruzipain with CRT in T. cruzi cells lysates were studied (52, 57, 106). On the contrary, several membrane-bound glycoproteins (T cell receptor α chain, murine and human MHC class I

328

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

heavy chains, etc.) and a soluble one (α 1-antitrypsin) were reported to display protein–protein interactions with CNX (88, 103, 131, 132). While the type of interaction could depend on the substrate glycoprotein, methods other than coimmunoprecipitation are required to validate a model involving a lectin–glycan interaction followed by a protein– protein association of the glycoprotein with CNX/CRT. CNX has also been reported to interact with naturally nonglycosylated proteins or with glycoproteins in which N-glycosylation had been prevented either by mutation of the consensus glycosylation sites or by the addition of tunicamycin (108, 131, 133 ). Prevention of glycosylation often leads to permanent misfolding and aggregation, and reported cases of CNX-glycan–free glycoprotein interactions may reﬂect nonspeciﬁc coprecipitations, as has been observed with other misfolded proteins and ER chaperones (134, 135). Nonspeciﬁc coprecipitation was also probably the cause of the reported interaction between CNX and the ε subunit of the T-cell receptor, (a nonglycosylated protein) overexpressed in COS cells (136). Moreover, no evidence has been provided for a productive role for the interactions of monoglucosylated proteins and CNX. There have been no reports on in vivo interaction of CRT with nonglycosylated proteins. Recently, it was noticed using in vitro studies that both CRT and a soluble form of CNX prevented aggregation not only of denatured monoglucosylated glycoproteins (Glc1Man9GlcNAc2–soybean agglutinin) but also of nonglycosylated proteins. The lectins also suppressed the thermal denaturation of nonglycosylated proteins such as malate dehydrogenase and citrate synthase (137, 138). CNX and CRT formed stable complexes with unfolded conformers but not with the native molecules. It was concluded that CRT and truncated CNX behaved in vitro as classical chaperones and ATP was proposed to modulate the chaperone functions. Whether the lectins perform a similar role in vivo is an open question, as monoglucosylated glycans have always been required for glycoprotein–CNX/CRT interaction. The conditions in the ER could favor a hypothetical conformational change in CNX/CRT induced by their interaction with monoglucosylated glycans, which would in turn be required to enable protein–protein interactions. C. The Role of Glycoprotein–Calnexin/Calreticulin Interactions Folding facilitation and ER retention of misfolded species mediated by glycoprotein–CNX/CRT interaction are not required for cell viability under normal growth conditions. Mammalian and yeast cells deﬁcient in GI or GII activities, in which monoglucosylated glycans cannot be

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

329

formed, are viable (13, 139, 140), indicating that a substantial proportion of many glycoproteins may still fold properly in the absence of interaction with CNX/CRT, or that other quality-control and folding factors may compensate for the lack of CNX/CRT function (63, 141, 142). For instance, the amount of carboxypeptidase Y that reached the vacuole in GII-deﬁcient S. pombe cells after a short pulse with [35S]Met decreased by only 50% with respect to wild-type cells (13). Similarly, about half of HA molecules folded properly when translated in a rabbit reticulocyte–dog pancreas system in the presence of GI/GII inhibitors (120). Furthermore, the upregulation of chaperones and other folding-assisting proteins (unfolded protein response) may compensate for or ameliorate the consequences of a total or reduced ability of glycoproteins to interact with CNX/CRT. This has been observed in mammalian and S. pombe cells grown in the presence of GI/GII inhibitors or in mutants lacking those activities as well as in S. pombe mutants lacking GT (13, 89, 143). The nonessential character of glycoprotein–CNX/CRT interaction for cell viability may be highlighted by the fact that this folding facilitating mechanism is probably not operative in S. cerevisiae, as this yeast does not reglucosylate glycoproteins and lacks several components of the CNX/CRT cycle (22, 23 ). Nonetheless, glycoprotein–CNX/CRT interaction is essential for viability under conditions of intense ER stress such as those caused by underglycosylation of glycoproteins and high temperature: alg6/gpt1 S. pombe double mutants which transfer Man9GlcNAc2 and are devoid of GT activity grew poorly and with altered morphology at 28◦ C, but could not grow at 37◦ C. Growth at 37◦ C was rescued on reintroduction of GT by cDNA transfection and also by culture in 1 M sorbitol (43), suggesting that the affected glycoprotein(s) might be involved in cell wall formation. Leukemic cells lacking CNX are viable and express normal levels class I MHC on their cell surface (144), probably reﬂecting the compensation of CNX absence by CRT. Mice embryos homozygous-null for CRT died after 14–18 days of development due to severe heart defects (145), and mice homozygous-null for calmegin, a testis-speciﬁc CNX homolog, were quasi-sterile, although they displayed normal spermatogenesis and mating. As their sperm cells did not adhere to egg extracellular matrix, it was speculated that calmegin-mediated folding facilitation was essential for sperm cell surface expression of a glycoprotein with a key role in fertility (146). GI/GII inhibitors prevented VSV maturation by interfering with G protein folding, as well as formation of infectious human immunodeﬁciency virus (HIV) type I particles, probably due to misfolding of loop V1–V2 in gp120 (141, 147, 148). In addition, glucosidase inhibitors

330

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

prevented folding of tyrosinase in melanoma cells and assembly of hepatitis B virus particles by blocking the correct folding of M glycoprotein (149, 150). These results indicate that although not required for cell growth under normal conditions, glycoprotein–CNX/CRT interaction is indeed required under certain conditions or for proper folding of particular proteins. VII. N-GLYCAN PROCESSING AND GLYCOPROTEIN DEGRADATION The 26S proteasome is the main degradation site of soluble and transmembrane misfolded glycoproteins retained in the ER (59–61). Retrograde transport from the lumen of the ER to the cytosol involves the Sec61p complex. After retrotranslocation, glycoproteins are deglycosylated by a neutral cytosolic N-glycanase activity prior to proteasome degradation (151–153 ). This activity generates an N,N ′ diacetylchitobiose structure at the reducing end of the oligosaccharides and a conversion of the Asn to Asp in the protein moiety. The high-mannose–type oligosaccharides thus released are then degraded in the mammalian cell cytosol by Endo H-like and α-mannosidase activities that yield Man5GlcNAc. This oligosaccharide is then transported into the lysosomes in an ATP-dependent process and further degraded (154–155). It is not clear how misfolded proteins are targeted for degradation, but in many cases preventing CNX/CRT–glycoprotein interactions using glucosidase inhibitors accelerated their disposal (91, 156–159). This would suggest that impairing the normal folding pathway by interfering with the lectin-based chaperone mechanism results in an increased proportion of incorrectly folded species and their consequent degradation. There are examples, however, where the opposite was seen (160), perhaps reﬂecting that the process varies for different glycoproteins and that it may be difﬁcult to generalize the consequences of interfering with the normal folding pathway of different substrates. From studies using inhibitors of mannose trimming in the ER, evidence is accumulating favoring a role for ER-mannosidases in targeting glycoproteins for degradation. An increasing number of reports describe how the use of α-mannosidase I inhibitors (kifunensin, 1deoxymannojirimycin) delayed proteasome degradation of several glycoproteins (misfolded α 1-antitrypsin variants, TCR CD3δ subunit, MHC class I heavy chain, ribophorin I, yeast prepro-α-factor, asialoglycoprotein receptor subunits, mutant α 2-plasmin inhibitors) (157–159, 161–164). However, no effect was observed on the proteasome-mediated degradation of the TCR-α subunit and the MHC class I heavy chain

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

331

(162, 165), indicating that the effect of mannose removal could be protein- and cell type-speciﬁc. Since demannosylation is a slower process than deglucosylation, ERretained glycoproteins are expected to bear more demannosylated oligosaccharides, so that the mannose content could act as a timer for residence in the ER. Although removal of Man residues yields poorer substrates for both GII and GT, the inﬂuence of the Man content on both activities is probably not identical (27, 38 ). Therefore, depending on the relative amounts of those opposite enzymatic activities (apparently only GT synthesis is induced under conditions that leads to ER accumulation of misfolded proteins; see Refs. 13 and 41) and their differential speciﬁcity with respect to saccharides of varying Man content, inhibition of Man removal may or may not favor the production of monoglucosylated glycans. It may be speculated that only when production of those species was signiﬁcantly enhanced by ER α-mannosidase I inhibitors would CNX/CRT binding favor folding and thus delay translocation of glycoproteins to the cytosol. The presence of Man8GlcNAc2 isomer B (lacking residue i in Fig. 1) was proposed to be a signal for degradation (166). This isomer is the only one produced in S. cerevisiae and the main one produced in the mammalian cell ER, in this case by α-mannosidase I. N-Glycans were shown to inﬂuence the degradation of a folding-incompetent carboxypeptidase Y (CPY∗ ) (167). Analysis of the degradation of CPY∗ expressed in S. cerevisiae cells bearing mutations that resulted in the accumulation of Glc2Man8GlcNAc2, Glc1Man8GlcNAc2, Man9GlcNAc2, or Man8GlcNAc2 showed that the glycoforms having the Man8GlcNAc2 structure were degraded at a higher rate than the others (166). The existence of lectin that recognized the Man8GlcNAc2 structure was suggested, although the apparent lack of ER α-mannosidase in S. pombe restricts the general role of Man8GlcNAc2 as a degradation signal in yeasts (17). Furthermore, a Man8GlcNAc2 isomer with a structure different from that of structure B (Fig. 1) and Man7GlcNAc2 can be generated in the ER in mammalian cells, since at least two α-mannosidases with different speciﬁcities (I and II) have been described in higher eukaryotes (14). Also, there are trypanosomatid species in which Man7GlcNAc2 or Man6GlcNAc2 are transferred to protein and in which, therefore, no Man8GlcNAc2 may be formed (4). The action of mannosidase II has also been implicated in targeting glycoproteins for degradation (168). Some misfolded proteins are not retrotranslocated to the cytosol for proteasomal degradation. In S. cerevisiae, some incompletely folded proteins escaped the ER and, after reaching the Golgi, were routed to the vacuole for degradation. Genetic evidence indicated that targeting to the

332

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

vacuole was mediated by a Golgi-located receptor (Vps10p) (169, 170), but a role for N-glycans in this process has not been evaluated. The significance of the dual pathway for disposing of improperly folded proteins in S. cerevisiae is unclear at the moment, but the vacuolar pathway may target proteins not showing gross folding defects that can be transported out of the ER. Proteasome-independent degradation of incompletely folded species has also been observed in mammalian cells; whereas unfolded human thyoperoxidase molecules were degraded in the proteasome, partially folded species appeared to be degraded in the ER (171). Similarly, degradation of α 1-antitrypsin inhibitor variant PI Z (168 ) and posttranslational degradation of apolipoprotein B (172) seem to be proteasome-independent. Finally, in some cases incompletely folded glycoproteins may not be degraded, but sorted to their ﬁnal destination; an inactive tyrosinase was found in the melanosomes of B16 melanoma cells incubated with the GI/GII inhibitor N-butyldeoxynojirimycin. Lack of activity was caused not by retention of the Glc units per se but by the absence of CNX-mediated folding facilitation, a fact that yielded a conformer unable to bind two essential Cu2+ atoms (149, 173). VIII. SUMMARY AND FUTURE PERSPECTIVES Proteins entering the secretory pathway acquire their proper tertiary and in some cases also quaternary structures in the ER. Incompletely folded species are prevented from transit to the Golgi apparatus and eventually degraded by the proteasome. This chapter describes the emerging principles by which N-glycan processing in the ER participates in the quality-control process. Monoglucosylated glycans formed by partial deglucosylation by GI and GII of oligosaccharides transferred from lipid derivatives to proteins (Glc3Man9GlcNAc2) mediate the binding of glycoproteins to two ER resident lectins, calnexin (CNX) and calreticulin (CRT). Further deglucosylation of glycans by GII liberates glycoproteins from CNX/CRT. Glycans may then be reglucosylated when linked to incompletely folded protein moieties and recognized again by CNX/CRT. Deglucosylation–reglucosylation cycles catalyzed by the opposing activities of GII and GT stop when proper folding is achieved as glycoproteins become substrates for GII but not for GT. The CNX/CRT– monoglucosylated glycan interaction is one of the mechanisms by which cells retain incompletely folded glycoproteins in the ER, in addition, it enhances folding efﬁciency by preventing protein aggregation and allowing intervention of additional ER chaperones and folding facilitating proteins. Increasing evidence suggests that Man removal by ER mannosidases might act as a timer mechanism for the disposal of incompletely folded glycoproteins bound for proteasome degradation.

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

333

The reactions described constitute a novel system, relying on the concerted action of carbohydrate-processing enzymes, for retaining nonnative conformers and facilitating protein folding and oligomerization. CNX and CRT are somewhat unconventional chaperones that apparently do not directly sense the folding status of the substrate proteins as classical chaperones do. This task is reserved to an enzyme (GT) that senses glycoprotein conformation and introduces a covalent modiﬁcation on those lacking their native conformations. This covalent carbohydrate modiﬁcation is the element recognized by this new kind of chaperone. Although some features of this system are increasingly clear, several aspects remain obscure and will be the object of future research. Detailed biochemical and structural studies are needed to understand the recognition of nonnative structures by GT, as well as ERp57 and CNX/CRT–ligand interaction. The controversial protein–protein interaction between CNX/CRT and folding glycoproteins needs to be explored, and the role of Man removal in the disposal of misfolded species has to be further substantiated. Understanding the interplay between the CNX/CRT pathway with other folding factors in the ER will advance our understanding of how N-glycosylation participates in the quality-control mechanisms and protein biogenesis in the secretory pathway. ACKNOWLEDGMENTS Work in the author’s laboratory was supported by grants from the United Nations Development Program/World Bank/World Health Organization Program for Research and Training in Tropical Diseases, the National Institutes of Health (USA), and the Howard Hughes Medical Institute.

REFERENCES 1. Parodi, A. J., and Leloir, L. F. (1979). The role of lipid intermediates in the glycosylation of proteins in the eukaryotic cell. Biochim. Biophys. Acta 559, 1–37. 2. Hubbard, S. C., and Ivatt, R. J. (1981). Synthesis and processing of asparaginelinked oligosaccharides. Annu. Rev. Biochem. 50, 555–583. 3. Kornfeld, R., and Kornfeld, S. (1985). Assembly of asparagine-linked oligosaccharides. Annu. Rev. Biochem. 54, 631–664. 4. Parodi, A. J. (1993). N-Glycosylation in trypanosomatid protozoa. Glycobiology 3, 193–199. 5. Yagodnik, C., de la Canal, C., and Parodi, A. J. (1987). Tetrahymena pyriformis cells are deﬁcient in all mannose-P-dolichol-dependent mannosyltransferases but not in mannose-P-dolichol synthesis. Biochemistry 26, 5937–5943. 6. de la Canal, L., and Parodi, A. J. (1985). Glycosylation of proteins in the protozoon Euglena gracilis. Comp. Biochem. Physiol. 81B, 803–805. 7. Nilsson, I. M., and von Heijne, G. (1993). Determination of the distance between the oligosaccharyltransferase active site and the endoplasmic reticulum membrane. J. Biol. Chem. 268, 5798–5801.

334

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

8. Holst, B., Bruun, A. W., Kielland-Brandt, M. C., and Winther, J. R. (1996). Competition between folding and glycosylation in the endoplasmic reticulum. EMBO J. 15, 3538–3546. 9. Turco, S. J., Stetson, B., and Robbins, P. W. (1977). Comparative rates of transfer of lipid-linked oligosaccharides to endogenous glycoprotein acceptors in vitro. Proc. Natl. Acad. Sci. USA 74, 4411–4414. 10. Ballou, L., Gopal, P., Krummel, B., Tammi, M., and Ballou, C. E. (1986). A mutation that prevents glucosylation of the lipid-linked oligosaccharide precursor leads to underglycosylation of secreted yeast invertase. Proc. Natl. Acad. Sci. USA 83, 3081– 3085. 11. Shailubhai, K., Pukazhenthi, B., Saxena, E. S., Varma, G., and Vijay, A. K. (1991). Glucosidase I, a transmembrane endoplasmic reticular glycoprotein with a luminal catalytic domain. J. Biol. Chem. 266, 16587–16593. 12. Trombetta, E. S., Simons, J. F., and Helenius, A. (1996). Endoplasmic reticulum glucosidase II is composed of a catalytic subunit, conserved from yeast to mammals, and a tightly bound noncatalytic HDEL-containing subunit. J. Biol. Chem. 271, 27509–27516. 13. D’Alessio, C., Fern´andez, F., Trombetta, E. S., and Parodi, A. J. (1999). Genetic evidence for the heterodimeric structure of glucosidase II. The effect of disrupting the subunit-encoding genes on glycoprotein folding. J. Biol. Chem. 274, 25899– 25905. 14. Weng, S., and Spiro, R. G. (1993). Demonstration that a kifunensin-resistant α-mannosidase with a processing action on N-linked oligosaccharides occurs in rat liver endoplasmic reticulum and various cultured cells. J. Biol. Chem. 268, 25656– 25663. 15. Jelinek-Kelly, S., and Herscovics, A. (1988). Glycoprotein biosynthesis in Saccharomyces cerevisiae. Puriﬁcation of the α-mannosidase which removes one speciﬁc mannose residue from Man9GlcNAc. J. Biol. Chem. 263, 14757–14763. 16. Romero, P. A., Vall´e, F., Howell, P. L., and Herscovics, A. (2000). Mutation of Arg273 to Leu alters the speciﬁcity of the yeast N-glycan processing class I α1,2mannosidase. J. Biol. Chem. 275, 11071–11074. 17. Ziegler, F. D., Gemmill, T. R., and Trimble, R. B. (1994). Glycoprotein synthesis in yeast. Early events in N-oligosaccharide processing in Schizosaccharomyces pombe. J. Biol. Chem. 269, 12527–12535. 18. Parodi, A. J., Lederkremer, G. Z., and Mendelzon, D. H. (1983). Protein glycosylation in Trypanosoma cruzi. The mechanism of glycosylation and structure of protein-bound oligosaccharides. J. Biol. Chem. 258, 5589–5595. 19. Parodi, A. J., Mendelzon, D. H., and Lederkremer, G. Z. (1983). Transient glucosylation of protein-bound Man9GlcNAc2, Man8GlcNAc2 and Man7GlcNAc2 in calf thyroid slices: A possible recognition signal in the processing of glycoproteins. J. Biol. Chem. 258, 8260–8265. 20. Parodi, A. J., Mendelzon, D. H., Lederkremer, G. Z., and, Mart´ın-Barrientos, J. (1984). Evidence that transient glucosylation of protein-linked Man9GlcNAc2, Man8GlcNAc2 and Man7GlcNAc2 occurs in rat liver and Phaseolus vulgaris cells. J. Biol. Chem. 259, 6351–6357. 21. Trombetta, S., Bosch, M., and Parodi, A. J. (1989). Glucosylation of glycoproteins by mammalian, plant, fungal and trypanosomatid protozoa microsomal proteins. Biochemistry 28, 8108–8116. 22. Fern´andez, F., Trombetta, S. E., Hellman, U., and Parodi, A. J. (1994). Puriﬁcation to homogeneity of UDP-Glc:glycoprotein glucosyltransferase from

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

23.

24.

25.

26.

27.

28.

29.

30. 31.

32.

33.

34. 35.

36. 37.

38.

335

Schizosaccharomyces pombe and apparent absence of the enzyme from Saccharomyces cerevisiae. J. Biol. Chem. 269, 30701–30706. Jakob, C. A., Burda, P., te Heesen, S., Aebi, M., and Roth, J. (1998). Genetic tailoring of N-linked oligosaccharides: The role of glucose residues in glycoprotein processing in Saccharomyces cerevisiae in vivo. Glycobiology 8, 155–164. Bosch, M., Trombetta, S. E., Engstrom, U., and Parodi, A. J. (1988). Characterization of dolichol diphosphate oligosaccharide:protein oligosaccharyltransferase and glycoprotein-processing glucosidases occurring in trypanosomatid protozoa. J. Biol. Chem. 263, 17360–17365. Dairaku, K., and Spiro, R. G. (1997). Phylogenetic survey of endomannosidase indicates late evolutionary appearance of this N-linked oligosaccharide processing enzyme. Glycobiology 7, 579–586. Cazzulo, J. J., and Parodi, A. J. (1982). Protein glycosylation in Trypanosoma cruzi. II. Partial characterization of protein-bound oligosaccharides labeled “in vivo.” J. Biol. Chem. 257, 7641–7645. Sousa, M., Ferrero-Garc´ıa, M., and Parodi, A. J. (1992). Recognition of the oligosaccharide and protein moieties of glycoproteins by the UDP-Glc:glycoprotein glucosyltransferase. Biochemistry 31, 97–105. Ga˜ na´ n, S., Cazzulo, J. J., and Parodi, A. J. (1991). A major proportion of N-glycoproteins are transiently glucosylated in the endoplasmic reticulum. Biochemistry 30, 3098–3104. Lang, L., Reitman, M., Tang, J., Roberts, R. M., and Kornfeld, S. (1984). Lysosomal enzyme phosphorylation. Recognition of a protein-dependent determinant allows speciﬁc phosphorylation of oligosaccharides present on lysosomal enzymes. J. Biol. Chem. 259, 14663–14671. Smith, P. L., and Baenziger, J. U. (1988). A pituitary N-acetylgalactosamine transferase that speciﬁcally recognizes glycoprotein hormones. Science 242, 930–933. Trombetta, S. E., Ga˜ na´ n, S., and Parodi, A. J. (1991). The UDP-Glc:glycoprotein glucosyltransferase is a soluble protein of the endoplasmic reticulum. Glycobiology 1, 155–161. Trombetta, S., and Parodi, A. J. (1992). Puriﬁcation to apparent homogeneity and partial characterization of rat liver UDP-Glc:glycoprotein glucosyltransferase. J. Biol. Chem. 267, 9236–9240. Parker, C. G., Fessler, L. I., Nelson, R. E., and Fessler, J. H. (1995). Drosophila UDP-Glc:glycoprotein glucosyltransferase: Sequence and characterization of an enzyme that distinguishes between denatured and native proteins. EMBO J. 14, 1294–1303. P´erez, M., and Hirschberg, C. (1986). Topography of glycosylation reactions in the rough endoplasmic reticulum membrane. J. Biol. Chem. 261, 6822–6830. Trombetta, E. S., and Helenius, A. (1999). Glycoprotein reglucosylation and nucleotide sugar utilization in the secretory pathway: Identiﬁcation of a nucleoside diphosphatase in the endoplasmic reticulum. EMBO J. 18, 3282–3292. Abeij´on, C., and Hirschberg, C. (1992). Topography of glycosylation reactions in the endoplasmic reticulum. Trends Biochem. Sci. 17, 32–36. Petrescu, A. J., Butters, T. D., Reinkensmeier, G., Petrescu, S., Platt, F. M., Dwek, R. A., and Wormald, M. R. (1997). The solution NMR structure of glucosylated N-glycans involved in the early stages of glycoprotein biosynthesis and folding. EMBO J. 16, 4302–4310. Grinna, L. S., and Robbins, P. W. (1980). Substrate speciﬁcities of rat liver microsomal glucosidases which process glycoproteins. J. Biol. Chem. 255, 2255–2258.

336

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

39. Fern´andez, F., Jannatipour, M., Hellman, U., Rokeach, L., and Parodi, A. J. (1996). A new stress protein: Synthesis of Schizosaccharomyces pombe UDP-Glc:glycoprotein glucosyltransferase mRNA is induced under stress conditions but the enzyme is not essential for cell viability. EMBO J. 15, 705–713. 40. Tessier, D. C., Dignard, D., Zapun, A., Radominska-Pandya, A., Parodi, A. J., Begeron, J. J. M., and Thomas, D. Y. (2000). Cloning and characterization of mammalian UDP-glucose:glycoprotein glucosyltransferase and the development of a speciﬁc substrate for this enzyme. Glycobiology 10, 403–412. 41. Arnold, S. M., Fessler, L. I., Fessler, J. H., and Kaufman, R. J. (2000). Two homologues encoding human UDP-glucose:glycoprotein glucosyltransferase differ in mRNA expression and enzymatic activity. Biochemistry 39, 2149–2163. 42. Meaden, P., Hill, K., Wagner, J., Slipetz, D., Sommer, S. S., and Bussey, H. (1990). The yeast KRE5 gene encodes a probable endoplasmic reticulum protein required for (1-6)-β-glucan synthesis and normal growth. Mol. Cell. Biol. 10, 3013– 3019. 43. Fanchiotti, S., Fern´andez, F., D’Alessio, C., and Parodi, A. J. (1998). The UDP-Glc:glycoprotein glucosyltransferase is essential for Schizosaccharomyces pombe viability under conditions of extreme endoplasmic reticulum stress. J. Cell Biol. 143, 625–635. 44. Hendrick, J. P., and Hartl, F. U. (1993). Molecular chaperone functions of heatshock proteins. Annu. Rev. Biochem. 62, 349–384. 45. Castro, O., Chen, L. Y., Parodi, A. J., and Abeij´on, C. (1999). Uridine diphosphateglucose transport into the endoplasmic reticulum of Saccharomyces cerevisiae: In vivo and in vitro evidence. Mol. Biol. Cell. 10, 1019–1030. 46. Sousa, M., and Parodi, A. J. (1995). The molecular basis for the recognition of misfolded glycoproteins by the UDP-Glc:glycoprotein glucosyltransferase. EMBO J. 15, 4196–4203. 47. Dessen, A., Gupta, D., Sabewsan, S., Brewer, C. F., and Sacchettini, J. C. (1995). X-Ray structure of the soybean agglutinin cross-linked with a biantennary analog of the blood group I carbohydrate antigen. Biochemistry 34, 4933–4942. 48. Trombetta, E. S., and Helenius, A. (2000). Conformational requirements for glycoprotein reglucosylation in the endoplasmic reticulum. J. Cell. Biol. 148, 1123–1129. 49. Bukau, B., and Horwich, A. L. (1998). The Hsp70 and Hsp60 chaperone machines. Cell. 92, 351–366. 50. Sousa, M. C., and Parodi, A. J. (1996). The interaction of the UDP-Glc:glycoprotein glucosyltransferase with the acceptor glycoprotein. Cell. Mol. Biol. 42, 609–616. 51. Ritter, C., and Helenius, A. (2000). Recognition of local glycoprotein misfolding by the ER folding sensor UDP-glucose:glycoprotein glucosyltransferase. Nature Struct. Biol. 7, 278–280. 52. Rodan, A. R., Simons, J. F., Trombetta, E. S., and Helenius, A. (1996). N-Linked oligosaccharides are necessary and sufﬁcient for association of glycosylated forms of bovine RNase with calnexin and calreticulin. EMBO J. 15, 6921–6930. 53. Rizzolo, L. J., and Kornfeld, R. (1988). Post-translational protein modiﬁcation in the endoplasmic reticulum. Demonstration of fatty acylase and deoxymannojirimycin-sensitive alpha-mannosidase activities. J. Biol. Chem. 263, 9520–9525. 54. Suh, P., Bergmann, J. E., and Gabel, C. A. (1989). Selective retention of monoglucosylated high mannose oligosaccharides by a class of mutant vesicular stomatis virus G proteins. J. Cell Biol. 108, 811–819.

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

337

55. Cannon, K., and Helenius, A. (1999). Trimming and readdition of glucose to N-linked oligosaccharides determines calnexin association of a substrate glycoprotein in living cells. J. Biol. Chem. 274, 7537–7544. 56. de Silva, A. M., Balch, W. E., and Helenius, A. (1990). Quality control in the endoplasmic reticulum: Folding and misfolding of vesicular stomatitis virus G protein in cells and in vitro. J. Cell Biol. 111, 857–866. 57. Labriola, C., Cazzulo, J. J., and Parodi, A. J. (1999). Trypanosoma cruzi calreticulin is a lectin that binds monoglucosylated oligosaccharides but not protein moieties of glycoproteins. Mol. Biol. Cell 10, 1381–1394. 58. Fern´andez, F., D’Alessio, C., Fanchiotti, S., and Parodi, A. J. (1998). A misfolded protein conformation is not a sufﬁcient condition for in vivo glucosylation by the UDP-Glc:glycoprotein glucosyltransferase. EMBO J. 17, 5877–5886. 59. Bodsky, J. L., and McCracken, A. A. (1997). ER-associated and proteasomemediated protein degradation: How two topologically restricted events came together. Trends Cell Biol. 7, 151–156. 60. Plemper, R. K., and Wolf, D. H. (1999). Retrograde protein translocation: ERADication of secretory proteins in health and disease. Trends Biochem. Sci. 24, 266– 270. 61. Bonifacino, J. S., and Weissman, A. M. (1998). Ubiquitin and the control of protein fate in the secretory and endocytic pathways. Annu. Rev. Cell Dev. Biol. 14, 19–57. 62. Hurtley, S. M., and Helenius, A. (1989). Protein oligomerization in the endoplasmic reticulum. Annu. Rev. Cell. Biol. 5, 277–307. 63. Ellgaard, L., Molinari, M., and Helenius, A. (1999). Setting the standards: Quality control in the secretory pathway. Science 286, 1882–1888. 64. Mazzarella, R. A., Srinivasan, M., Haugejorden, S. M., and Green, M. (1990). ERp72, an abundant luminal endoplasmic reticulum protein, contains three copies of the active site sequences of protein disulﬁde isomerase. J. Biol. Chem. 265, 1094– 1101. 65. Chaudhuri, M. M., Tonin, P. N., Lewis, W. H., and Srinivasan, P. R. (1992). The gene for a novel protein, a member of the protein disulﬁde isomerase/form I phophoinositide-speciﬁc phospholipase C family, is ampliﬁed in hydroxyurearesistant cells. Biochem. J. 281, 645–650. 66. Hirano, N., Shibasaki, F., Kato, H., Sakai, R., Tanaka, T., Nishida, J., Yazaki, Y., Takenawa, T., and Hirai, H. (1994). Molecular cloning and characterization of a cDNA for bovine phospholipase C-α: Proposal of redesignation of phopholipase C-α. Biochem. Biophys. Res. Commun. 204, 375–382. 67. Haas, I. G., and Wabl, M. (1983). Immunoglobulin heavy chain binding protein. Nature 306, 387–389. 68. Sorger, P. K., and Pelham, H. R. (1987). The glucose-regulated protein grp94 is related to heat shock protein hsp90. J. Mol. Biol. 194, 341–344. 69. Lin, H. Y., Masso-Welch, P., Di, Y. P., Cai, J. W., Shen, J. W., and Subjeck, J. R. (1993). The 170-kDa glucose-regulated stress protein is an endoplasmic reticulum protein that binds immunoglobulin. Mol. Biol. Cell 4, 1109–1119. 70. Helenius, A., Marquardt, T., and Braakman, I. (1992). The endoplasmic reticulum as a protein folding compartment. Trends Cell Biol. 2, 227–231. 71. Meldolesi, J., and Pozzan, T. (1998). The endoplasmic reticulum Ca2+ store: A view from the lumen. Trends Biochem. Sci. 23, 10–14. 72. Williams, D. B. (1995). Calnexin: A molecular chaperone with a taste for carbohydrate. Biochem. Cell Biol. 73, 123–132.

338

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

73. Bergeron, J. J. M., Brenner, M. B., Thomas, D. Y., and Williams, D. B. (1994). Calnexin: A membrane-bound chaperone of the endoplasmic reticulum. Trends Biochem. Sci. 19, 124–128. 74. Michalak, M., Corbett, E. F., Mesaeli, N., Nakamura, K., and Opas, M. (1999). Calreticulin: One protein, one gene, many functions. Biochem. J. 344, 281–292. 75. Tjoelker, L. W., Seyfried, C. E., Eddy, R. L., Byers, M. G., Shows, T. B., Calderon, J., Schreiber, R. B., and Gray, P. W. (1994). Human, mouse, and rat calnexin cDNA cloning: Identiﬁcation of potential calcium binding motifs and gene localization to human chromosome 5. Biochemistry 33, 3229–3236. 76. Vassilakos, A., Michalak, M., Lehrman, M. A., and Williams, D. B. (1998). Oligosaccharide binding characteristics of the molecular chaperones calnexin and calreticulin. Biochemistry 37, 3480–3490. 77. Tatu, U., and Helenius, A. (1997). Interactions between newly synthesized glycoproteins, calnexin and a network of resident chaperones in the endoplasmic reticulum. J. Cell Biol. 136, 555–565. 78. Ou, W.- J., Cameron, P. H., Thomas, D. Y., and Bergeron, J. J. M. (1993). Association of folding intermediates of glycoproteins with calnexin during protein maturation. Nature 364, 771–776. 79. Degen, E., and Williams, D. B. (1991). Participation of a novel 88-kDa protein in the biogenesis of murine class I histocompatibility molecules. J. Cell Biol. 112, 1099–1115. 80. Hochstenbach, F., David, V., Watkins, S, and Brenner, M. B. (1992). Endoplasmic reticulum resident protein of 90 kilodaltons associates with the T- and B-cell antigen receptors and major histocompatibility complex antigens during their assembly. Proc. Natl. Acad. Sci. USA 89, 4734–4738. 81. Hammond, C., Braakman, I., and Helenius, A. (1994). Role of N-linked oligosaccharide recognition, glucose trimming, and calnexin in glycoprotein folding and quality control. Proc. Natl. Acad. Sci. USA. 91, 913–917. 82. Hebert, D. N., Foellmer, B., and Helenius, A. (1995). Glucose trimming and reglucosylation determine glycoprotein association with calnexin in the endoplasmic reticulum. Cell 81, 425–433. 83. Peterson, J. R., Ora, A., Nguyen Van, P., and Helenius, A. (1995). Calreticulin is a lectin-like molecular chaperone for glycoproteins of the endoplasmic reticulum. Mol. Biol. Cell 6, 1173–1184. 84. Nauseef, W. M., McCormick, S. J., and Clark, R. A. (1995). Calreticulin functions as a molecular chaperone in the biosynthesis of myeloperoxidase. J. Biol. Chem. 270, 4741–4747. 85. Otteken, A., and Moss, B. (1996). Calreticulin interacts with newly synthesized human immunodeﬁciency virus Type 1 envelope glycoprotein, suggesting a chaperone function similar to that of calnexin. J. Biol. Chem. 271, 97–103. 86. Wada, I., Imai, S., Kai, M., Sakane, F., and Kanoh, H. (1995). Chaperone function of calreticulin when expressed in the endoplasmic reticulum as the membraneanchored and soluble forms. J. Biol. Chem. 270, 20298–20304. 87. Vassilakos, A., Cohen-Doyle, M. F., Peterson, P. A., Jackson, M. R., and Williams, D. B. (1996). The molecular chaperone calnexin facilitates folding and assembly of class I histocompatibility molecules. EMBO J. 15, 1495–1506. 88. Zhang, Q., Tector, M., and Salter, R. D. (1995). Calnexin recognizes carbohydrate and protein determinants of class I major histocompatibility complex molecules. J. Biol. Chem. 270, 3944–3948.

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

339

89. Balow, J. P., Weissman, J. D., and Kearse, K. P. (1995). Unique expression of major histocompatibility complex class I proteins in the absence of glucose trimming and calnexin association. J. Biol. Chem. 270, 29025–29029. 90. Van Leeuwen, J. E. M., and Kearse, K. P. (1996). Calnexin associates exclusively with individual CD3δ and T cell antigen receptor (TCR)α proteins containing incompletely trimmed glycans that are not assembled into multisubunit TCR complexes. J. Biol. Chem. 271, 9660–9665. 91. Kearse, K. P., Williams, D. B., and Singer, A. (1994). Persistence of glucose residues on core oligosaccharides prevents association of TCRα and TCRβ proteins with calnexin and results speciﬁcally in accelerated degradation of nascent TCRα proteins within the endoplasmic reticulum. EMBO J. 13, 3678–3686. 92. Chang, W., Gelman, M. S., and Prives, J. M. (1997). Calnexin-dependent enhancement of nicotinic acetylcholine receptor assembly and surface expression. J. Biol. Chem. 272, 28925–28932. 93. Keller, S. H., Lindstrom, J., and Taylor, P. (1998). Inhibition of glucose trimming with castanospermine reduces calnexin association and promotes proteasome degradation of the α-subunit of the nicotinic acetylcholine receptor. J. Biol. Chem. 273, 17064–17072. 94. Bass, J., Chiu, G., Argon, Y., and Steiner, D. F. (1998). Folding of insulin receptor monomers is facilitated by the molecular chaperones calnexin and calreticulin and impaired by rapid dimerization. J. Cell Biol. 141, 637–646. 95. Pipe, S. W., Morris, J. A., Shah, J., and Kaufman, R. J. (1998). Differential interaction of coagulation factor VIII and factor V with protein chaperones calnexin and calreticulin. J. Biol. Chem. 273, 8537–8544. 96. Toyofuku, K., Wada, I., Hirosaki, K., Park, J. S., Hori, Y., and Jimbow, K. (1999). Promotion of tyrosinase folding in COS 7 cells by calnexin. J. Biochem. 125, 82–89. 97. Popov, M., and Reithmeier, R. A. F. (1999). Calnexin interaction with N-glycosylation mutants of a polytopic membrane glycoprotein, the human erythrocyte anion exchanger 1 (band 3). J. Biol. Chem. 274, 17635–17642. 98. Ermonval, M., Duvet, S., Zonnevald, D., Cacan, R., Buttin, G., and Braakman, I. (2000). Truncated N-glycans affect protein folding in the ER of CHO-derived mutant cell lines without preventing calnexin binding. Glycobiology 10, 77–87. 99. Helenius, A. et al. (1997). Calnexin, calreticulin and the folding of glycoproteins. Trends Cell Biol. 7, 193–199. 100. Trombetta, E. S., and Helenius, A. (1998). Lectins as chaperones in glycoprotein folding. Curr. Opin. Struct. Biol. 8, 587–592. 101. Wada, I., Kai, M., Imai, S., Sakane, F., and Kanoh, H. (1997). Promotion of transferrin folding by cyclic interactions with calnexin and calreticulin. EMBO J. 16, 5420–5432. 102. Ora, A., and Helenius, A. (1995). Calnexin fails to associate with substrate proteins in glucosidase-deﬁcient cell lines. J. Biol. Chem. 270, 26060–26062. 103. Ware, F. E., Vassilakos, A., Peterson, P. A., Jackson, M. R., Lehrman, M. A., and Williams, D. B. (1995). The molecular chaperone calnexin binds Glc1Man9GlcNAc2 oligosaccharide as an initial step in recognizing unfolded glycoproteins. J. Biol. Chem. 270, 4697–4704. 104. Spiro, R. G., Zhu, Q., Bhoyroo, V., and S¨oling, H.-D. (1996). Deﬁnition of the lectin-like properties of the molecular chaperone, calreticulin, and demonstration of its copuriﬁcation with endomannosidase from rat liver Golgi. J. Biol. Chem. 271, 11588–11594.

340

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

105. Patil, A. R., Thomas, C. J., and Surolia, A. (2000). Kinetics and the mechanism of interaction of the endoplasmic reticulum chaperone, calreticulin, with monoglucosylated (Glc1Man9GlcNAc2) substrate. J. Biol. Chem. 275, 24348– 24356. 106. Zapun, A., Petrescu, S., Rudd, P. M., Dwek, R. A., Thomas, D. Y., and Bergeron, J. J. (1997). Conformation independent binding of monoglucosylated ribonuclease B to calnexin. Cell 88, 29–38. 107. Zhang, Q., and Salter, R. D. (1998). Distinct patterns of folding and interactions with calnexin and calreticulin in human class I MHC proteins with altered N-glycosylation. J. Immunol. 160, 831–837. 108. Cannon, K. S., Hebert, D. N., and Helenius, A. (1996). Glycan-dependent and -independent association of vesicular stomatitis virus G protein with calnexin. J. Biol. Chem. 271, 14280–14284. 109. Branza-Nichita, N., Negroiu, G., Petrescu, A. J., Garman, E. F., Platt, F. M., Wormald, M. R., Dwek, R. A., and Petrescu, S. M. (2000). Mutations at critical N-glycosylation sites reduce tyrosinase activity by altering folding and quality control. J. Biol. Chem. 275, 8169–8175. 110. Damilczyk, U. G., Cohen-Doyle, M. F., and Williams, D. B. (2000). Functional relationship between calreticulin, calnexin and the ER luminal domain of calnexin. J. Biol. Chem. 275, 13089–13097. 111. Hebert, D. N., Zhang, J. X., Chen, W., Foellmer, B., and Helenius, A. (1997). The number and location of glycans on inﬂuenza hemagglutinin determine folding and association with calnexin and calreticulin. J. Cell Biol. 139, 613–623. 112. Sadasivan, B., Lehner, P. J., Ortmann, B., Spies, T., and Cresswell, P. (1996). Roles for calreticulin and a novel glycoprotein, tapasin, in the interaction of MHC class I molecules with TAP. Immunity 5, 103–114. 113. Harris, M. R., Yu, Y. Y. L., Kindle, C. S., Hansen, T. H., and Solheim, J. C. (1998). Calreticulin and calnexin interact with different protein and glycan determinants during the assembly of MHC class I. J. Immunol. 160, 5404–5409. 114. Van Leeuwen, J. E. M., and Kearse, K. P. (1996). Deglucosylation of N-linked glycans is an important step in the dissociation of calreticulin-class I-TAP complexes. Proc. Natl. Acad. Sci. USA. 93, 13997–14001. 115. Lewis, J. W., and Elliott, T. (1998). Evidence for successive peptide binding and quality control stages during MHC class I assembly. Curr. Biol. 8, 717–720. 116. Van Leeuwen, J. E. M., and Kearse, K. P. (1996). The related molecular chaperones calnexin and calreticulin differentially associate with nascent T cell antigen receptor proteins within the endoplasmic reticulum. J. Biol. Chem. 271, 25345–25349. 117. Labriola, C., Cazzulo, J. J., and Parodi, A. J. (1995). Retention of glucose residues added by the UDP-Glc:glycoprotein glucosyltransferase delays exit of glycoproteins from the endoplasmic reticulum. J. Cell Biol. 130, 771–779. 118. Zhang, J.-X., Braakman, I., Matlack, K. E. S., and Helenius, A. (1997). Quality control in the secretory pathway: The role of calreticulin, calnexin and BiP in the retention of glycoproteins with C-terminal truncations. Mol. Biol. Cell 8, 1943–1954. 119. Reddy, P. S., and Corley, R. B. (1998). Assembly, sorting and exit of oligomeric proteins from the endoplasmic reticulum. Bioessays 20, 546–554. 120. Hebert, D. N., Foellmer, B., and Helenius, A. (1996). Calnexin and calreticulin promote folding, delay oligomerization and suppress degradation of inﬂuenza hemagglutinin in microsomes. EMBO J. 15, 2961–2968. 121. Oliver, J. D., Van der Wal, F., Bulleid, N. J., and High, S. (1997). Interaction of the thiol-dependent reductase ERp57 with nascent glycoproteins. Science 275, 86–88.

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

341

122. Elliott, J. G., Oliver, J. D., and High, S. (1997). The thiol-dependent reductase ERp57 interacts speciﬁcally with N-glycosylated integral membrane proteins. J. Biol. Chem. 272, 13849–13855. 123. Oliver, J. D., Llewelyn Roderick, H., Llewelyn, D. H., and High, S. (1999). ERp57 functions as a subunit of speciﬁc complexes formed with the ER lectins calreticulin and calnexin. Mol. Biol. Cell 10, 2573–2582. 124. Farmery, M. R., Allen, S., Allen, A. J., and Bulleid, N. J. (2000). The role of ERp57 in disulﬁde bond formation during the assembly of major histocompatibility complex class I in a synchronized semipermeabilized cell translation system. J. Biol. Chem. 275, 14933–14938. 125. Hughes, E. A., and Cresswell, P. (1998). The thiol oxidoreductase ERp57 is a component of the MHC class I peptide-loading complex. Curr. Biol. 8, 709–712. 126. Lindquist, J. A., Jensen, O. N., Mann, M., and Hammerling, G. J. (1998). ER60, a chaperone with thiol-dependent reductase activity involved in MHC class I assembly. EMBO J. 15, 2186–2195. 127. Morrice, N. A., and Powis, S. J. (1998). A role for the thiol-dependent reductase ERp57 in the assembly of MHC class I molecules. Curr. Biol. 4, 713–716. 128. Zapun, A., Darby, N. J., Tessier, D. C., Michalak, M., Bergeron, J. J. M., and Thomas, D. Y. (1998). Enhanced catalysis of ribonuclease B folding by the interaction of calnexin or calreticulin with ERp57. J. Biol. Chem. 273, 6009–6012. 129. Molinari, M., and Helenius, A. (1999). Glycoproteins form mixed disulphides with oxidoreductases during folding in living cells. Nature 402, 90–93. 130. High, S., Lecomte, F., Russell, S., Abell, B., and Oliver, J. (2000). Glycoprotein folding in the endoplasmic reticulum: A tale of three chaperones? FEBS Lett. 476, 38–41. 131. Arunachalam, B., and Cresswell, P. (1995). Molecular requirements for the interaction of class II major histocompatibility complex and invariant chain with calnexin. J. Biol. Chem. 270, 2784–2790. 132. Bennett, M. J., Van Leeuwen, J. E. M., and Kearse, K. P. (1998). Calnexin association is not sufﬁcient to protect TCRα proteins from rapid degradation in CD4+CD8+ thymocytes. J. Biol. Chem. 273, 23674–23680. 133. Carre˜ no, B. M., Schreiber, K. L., McKean, D. J., Stroynowski, Y., and Hansen, T. H. (1995). Aglycosylated and phosphatidylinositol-anchored MHC class I molecules are associated with calnexin. Evidence implicating the class I-connecting peptide segment in calnexin association. J. Immunol. 154, 5173–5180. 134. Gibson, R., Schlesinger, S., and Kornfeld, S. (1979). The nonglycosylated glycoprotein of vesicular stomatitis virus is temperature sensitive and undergoes intracellular aggregation at elevated temperatures. J. Biol. Chem. 254, 3600– 3605. 135. Hurtley, S. M., Bole, D. G., Hoover-Litty, H., Helenius, A., and Copeland, C. S. (1989). Interactions of misfolded inﬂuenza virus hemagglutinin with binding protein (BiP). J. Cell Biol. 108, 2117–2126. 136. Rajagopalan, S., Xu, Y., and Brenner, M. B. (1994). Retention of unassembled components of integral membrane proteins by calnexin. Science 263, 387–390. 137. Ihara, Y., Cohen-Doyle, M. F., Saito, Y., and Williams, D. B. (1999). Calnexin discriminates between protein conformational states and functions as a molecular chaperone in vitro. Mol. Cell 4, 331–341. 138. Saito, Y., Ihara, Y., Leach, M. R, Cohen-Doyle, M. F., and Williams, D. B. (1999). Calreticulin functions in vitro as a molecular chaperone for both glycosylated and nonglycosylated proteins. EMBO J. 18, 6718–6729.

342

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

139. Ray, M. K., Young, J., Sundaram, S., and Stanley, P. (1991). A novel glycosylation phenotype expressed by Lec23, a Chinese hamster ovary mutant deﬁcient in α-glucosidase I. J. Biol. Chem. 266, 22818–22825. 140. Reitman, M. L., Trowbridge, L. S., and Kornfeld, S. (1982). A lectin-resistant mouse lymphoma cell line is deﬁcient in glucosidase II, a glycoprotein processing enzyme. J. Biol. Chem. 257, 10357–10363. 141. Hammond, C., and Helenius, A. (1994). Folding of VSV-G protein: sequential interaction with BiP and calnexin. Science 266, 456–458. 142. Molinari, M., and Helenius, A. (2000). Chaperone selection during glycoprotein translocation into the endoplasmic reticulum. Science 288, 331–333. 143. Pahl, H. L., and Baeuerle, P. A. (1995). A novel signal transduction pathway from the endoplasmic reticulum to the nucleus is mediated by transcription factor NFkappa B. EMBO J. 14, 2580–2588. 144. Scott, J. E., and Dawson, J. R. (1995). MHC Class I expression and transport in a calnexin-deﬁcient cell line. J. Immunol. 155, 143–148. 145. Mesaeli, N., Nakamura, K., Zvaritch, E., Dickie, P., Dziak, E., Krause, K. H., Opas, M., MacLennan, D. H., and Michalak, M. (1999). Calreticulin is essential for cardiac development. J. Cell Biol. 144, 857–868. 146. Ikawa, M., Wada, I., Kominami, K., Watanabe, D., Toshimori, K., Nishimune, Y., and Okabe, M. (1997). The putative chaperone calmegin is required for sperm fertility. Nature 387, 607–611. 147. Gruters, R. A., Neefjes, J. J., Tersmette, M., Tulp, A., Huisman, H. G., Miedema, F., and Ploegh, H. L. (1987). Interference with HIV-induced syncytium formation and viral infectivity by inhibitors of trimming glucosidases. Nature 330, 74–77. 148. Fischer, P. B., Karlsson, G. B., Butters, T. D., Dwek, R. A., and Platt, F. M. (1996). N-Butyldeoxynojirimycin-mediated inhibition of human immunodeﬁciency virus entry correlates with changes in antibody recognition of the V1-V2 region of gp 120. J. Virol. 70, 7143–7152. 149. Petrescu, S. M., Petrescu, A. J., Titu, H. N., Dwek, R. A., and Platt, F. M. (1997). Inhibition of N-glycan processing in B16 melanoma cells results in inactivation of tyrosinase but does not prevent its transport to the melanosome. J. Biol. Chem. 272, 15796–15803. 150. Mehta, A., Lu, X., Block, T. M., Blumberg, B. S., and Dwek, R. A. (1997). Hepatitis B virus (HBV) envelope glycoproteins vary drastically in their sensitivity to glycan processing: Evidence that alteration of a single N-linked glycosylation site can regulate HBV secretion. Proc. Natl. Acad. Sci. USA 94, 1822–1827. 151. Suzuki, T., Park, H., Kitajima, K., and Lennarz, W. J. (1998). Peptides glycosylated in the endoplasmic reticulum of yeast are subsequently deglycosylated by a soluble peptide:N-glycanase activity. J. Biol. Chem. 273, 21526–21530. 152. Wiertz, E. J. H. J., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J., and Ploegh, H. L. (1996). The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84, 769–779. 153. Beb¨ok, Z., Mazzochi, C., King, S. A., Hong, J. S., and Sorscher, E. J. (1998). The mechanism underlying cystic ﬁbrosis transmembrane conductance regulator transport from the endoplasmic reticulum to the proteasome includes Sec61α and a cytosolic, deglycosylated intermediary. J. Biol. Chem. 273, 29873–29878. 154. Saint-Pol, A., Bauvy, C., Codogno, P., and Moore, S. E. H. (1997). Transfer of free polymannose-type oligosaccharides from the cytosol to lysosomes in cultured human hepatocellular carcinoma HEPG2 cells. J. Cell Biol. 136, 45–59.

N -GLYCAN PROCESSING AND GLYCOPROTEIN FOLDING

343

155. Saint-Pol, A., Codogno, P., and Moore, S. E. H. (1999). Cytosol-to-lysosome transport of free polymannose-type oligosaccharides. Kinetic and speciﬁcity studies using rat liver lysosomes. J. Biol. Chem. 274, 13547–13555. 156. Romagnoli, P., and Germain, R. (1995). Inhibition of invariant chain (li)-calnexin interaction results in enhanced degradation of li but does not prevent the assembly of alfa beta li complexes. J. Exp. Med. 182, 2027–2036. 157. Ayalon-Soffer, M., Shenkman, M., and Lederkremer, G. (1999). Differential role of mannose and glucose trimming in the ER degradation of asialoglycoprotein receptor subunits. J. Cell Sci. 112, 3309–3318. 158. Su, K., Stoller, T., Rocco, J., Zemsky, J., and Green, R. (1993). Pre-Golgi degradation of yeast prepro- -factor expressed in a mammalian cell. Inﬂuence of cell type-speciﬁc oligosaccharide processing on intracellular fate. J. Biol. Chem. 268, 14301–14309. 159. de Virgilio, M., Kitzm¨uller, C., Schwaiger, E., Klein, M., Kreibich, G., and Ivessa, N. E. (1999). Degradation of a short-lived glycoprotein from the lumen of the endoplasmic reticulum: The role of N-linked glycans and the unfolded protein response. Mol. Biol. Cell 10, 4059–4073. 160. Marcus, N. Y., and Perlmutter, D. H. (2000). Glucosidase and mannosidase inhibitors mediate increased secretion of mutant 1-antitrypsin Z. J. Biol. Chem. 275, 1987–1992. 161. Liu, Y., Choudhury, P., Cabral, C. M., and Sifers, R. N. (1999). Oligosaccharide modiﬁcation in the early secretory pathway directs the selection of a misfolded glycoprotein for degradation by the proteasome. J. Biol. Chem. 274, 5861– 5867. 162. Yang, M., Omura, S., Bonifacino, J. S., and Weissman, A. M. (1998). Novel aspects of degradation of T Cell receptor subunits from the endoplasmic reticulum (ER) in T cells: Importance of oligosaccharide processing, ubiquitination, and proteasomedependent removal from ER membranes. J. Exp. Med. 187, 835–846. 163. Wilson, C. M., Farmery, M. R., and Bulleid, N. J. (2000). Pivotal role of calnexin and mannose trimming in regulating the endoplasmic reticulium-associated degradation of major histocompatibility complex class I heavy chain. J. Biol. Chem. 275, 21224–21232. 164. Chung, D. H., Ohashi, K., Watanabe, M., Miyasaka, N., and Hirosawa, S. (2000). Mannose trimming targets mutant α2-plasmin inhibitor for degradation by the proteasome. J. Biol. Chem. 275, 4981–4987. 165. Moore, S. E., and Spiro, R. G. (1993). Inhibition of glucose trimming by castanospermine results in rapid degradation of unassembled major histocompatibility complex class I molecules. J. Biol. Chem. 268, 3809–3812. 166. Jakob, C. A., Burda, P., Roth, J., and Aebi, M. (1998). Degradation of misfolded endoplasmic reticulum glycoproteins in Saccharomyces cerevisiae is determined by a speciﬁc oligosaccharide structure. J. Cell Biol. 142, 1223–1233. 167. Knop, M., Hauser, N., and Wolf, D. (1996). N-Glycosylation affects endoplasmic reticulum degradation of a mutated derivative of carboxypeptidase Yscy in yeast. Yeast 12, 1229–1238. 168. Cabral, C., Choudhury, P., Liu, Y., and Sifers, R. (2000). Processing by endoplasmic reticulum mannosidases partitions a secretion-impaired glycoprotein into distinct disposal pathways. J. Biol. Chem. 275, 25015–25022. 169. Hong, E., Davidson, A. R., and Kaiser, C. A. (1996). A pathway for targeting soluble misfolded proteins to the yeast vacuole. J. Cell Biol. 135, 623–633.

344

E. SERGIO TROMBETTA AND ARMANDO J. PARODI

170. J¨orgensen, M. U., Emr, S. D., and Winther, J. R. (1999). Ligand recognition and domain structure of Vps10p, a vacuolar protein sorting receptor in Saccharomyces cerevisiae. Eur. J. Biochem. 260, 461–469. 171. Fayadat, L., Siffroi-Fernandez, S., Lanet, J., and Franc, J. L. (2000). Degradation of human thyoperoxidase in the endoplasmic reticulum involves two different pathways depending on the folding state of the protein. J. Biol. Chem. 275, 15948– 15954. 172. Cavallo, D., Rudy, D., Mohammadi, A., Macri, J., and Adeli, K. (1999). Studies on degradative mechanisms mediating post-translational fragmentation of apolipoprotein B and the generation of the 70-kDa fragment. J. Biol. Chem. 274, 23135–23143. 173. Branza-Nichita, N., Petrescu, A. J., Dwek, R. A., Wormald, M. R., Platt, F. M., and Petrescu, S. M. (1999). Tyrosinase folding and copper loading in vivo: A crucial role for calnexin and -glucosidase II. Biochem. Biophys. Res. Commun. 261, 720–725.

FUNCTIONAL GENOMIC APPROACHES TO UNDERSTANDING MOLECULAR CHAPERONES AND STRESS RESPONSES By KEVIN J. TRAVERS,*,† CHRISTOPHER K. PATIL,† and JONATHAN S. WEISSMAN*,† *,†Howard Hughes Medical Institute; *Department of Cellular and Molecular Pharmacology; and † Department of Biochemistry and Biophysics, University of California–San Francisco, San Francisco, California 94143

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Identiﬁcation of a Eukaryotic Chaperonin . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Hsp104/Clp Family: A Connection between Proteolytic and Folding Machinery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Hsp33: A Redox-Regulated Chaperone . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. The Heat Shock Response and Protein Translation . . . . . . . . . . . . . . . . . . III. Functional and Genomic Analysis of the Unfolded Protein Response . . . . . . . A. Protein Folding in the Secretory Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Targets of the Unfolded Protein Response . . . . . . . . . . . . . . . . . . . . . . . . . . C. Coregulation of ERAD and the UPR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Strategy for Identifying UPR Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Experimental Strategy for Identifying Direct UPR Targets . . . . . . . . . . . . B. Computational Strategy for Identifying Direct UPR Targets . . . . . . . . . . . V. An Overview of Unsupervised Search Strategies . . . . . . . . . . . . . . . . . . . . . . . . . A. Deﬁnition of Unsupervised Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Hierarchical Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Singular Value Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Self-Organizing Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. UPR as a Case Study in the Comparison of Supervised versus Unsupervised Searches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Distribution of UPR Targets within a Clustered Data Set Focused on Cell Cycle Changes and Stress Treatments . . . . . . . . . . . . . . . . . . . . . . . B. Distribution of UPR Targets within a Clustered Data Set Obtained from a Diverse Set of Knockout Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

345 346 346 349 350 351 353 354 356 360 363 363 366 370 370 371 374 375 376 377 378 380

I. INTRODUCTION Historically, efforts to characterize the cell’s basal protein folding and degradation machinery have been spearheaded by the analysis of proteins speciﬁcally induced by the classic heat shock response. Yet it is clear that cellular responses to heat and other stresses that compromise protein folding encompass a range of biochemical activities far broader than molecular chaperones and proteases. The availability of complete genomic sequences from a range of eukaryotes, bacteria, and archae, 345 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

346

KEVIN J. TRAVERS ET AL.

coupled with genomic array technology, has now made it possible to rapidly identify all the targets of a speciﬁc stress response. Such data have the potential to provide a wealth of information on the requirements for proper folding as well as guide the discovery of new molecular chaperones, proteases, and other proteins important for maintaining the integrity of protein structures in vivo. How best to collect and interpret this avalanche of information, however, presents an enormous challenge with which molecular biologists are now grappling. In this chapter we start (Section II) with a few key examples that underscore how characterization of proteins identiﬁed solely on the basis of their status as conserved targets of the heat shock response has led to insights into the cell biology of protein synthesis, folding, and degradation. In particular, this strategy has allowed the discovery of new general molecular chaperones, a connection between the cellular chaperone and degradation machinery, and, more recently, an unexpected relationship between ribosome metabolism and the heat shock response. We next concentrate on lessons learned from the application of genomic array technology to a speciﬁc folding stress response, the unfolded protein response (UPR), which is critical for maintaining the integrity of protein folding in the endoplasmic reticulum (ER). In Section III we focus on characterization of the different cellular processes induced by the UPR and speculate on how these various processes might act in a coherent way to aid ER folding. Section IV looks at the analytical and experimental approaches employed to speciﬁcally dissect the UPR from other cellular stresses, thereby making it possible to identify the speciﬁc targets of the UPR. Finally, in Sections V and VI we compare the highly directed approach used to identify UPR targets to more generic “unsupervised” clustering strategies. In those approaches, a broad range of genomic expression experiments is analyzed to identify patterns of relationships between genes (e.g., molecular chaperones) and conditions (e.g., cellular stresses) without biasing the analysis with any a priori preconception of the underlying biological processes. II. HISTORICAL PERSPECTIVE A. Identification of a Eukaryotic Chaperonin By the early 1990s, the basic paradigm that efﬁcient protein folding in vivo depends on a set of highly conserved proteins termed molecular chaperones was well established (Cheng et al., 1989; Gething and

FUNCTIONAL GENOMIC APPROACHES

347

Sambrook, 1992; Hemmingsen et al., 1988; Ostermann et al., 1989; Pelham, 1986; Rothman, 1989). In addition, the relationship between the heat shock proteins and molecular chaperones was so intimate that these terms had become nearly synonymous. Three major classes of conserved chaperones/heat shock proteins emerged from these studies: the chaperonin family (including Hsp60 and the Escherichia coli protein GroEL; reviewed in Sigler et al., 1998), which is characterized by the formation of ring-shaped oligomers; Hsp70, which functions together with Hsp40 and (in bacteria) GrpE (reviewed in Bukau and Horwich, 1998); and Hsp90 (for reviews, see Buchner, 1999; Mayer and Bukau, 1999). Among these three classes, the evidence that GroEL and its homologs play an essential and general role in assisting de novo protein folding is arguably the most compelling. For example, in vivo overexpression of the E. coli chaperonin GroEL and its co-chaperonin GroES assists the folding of heterologously expressed proteins (Goloubinoff et al., 1989b) as well as suppresses temperature-sensitive mutations in a variety of endogenous proteins (Van Dyk et al., 1989). Moreover, Horwich, Hartl, and co-workers had shown that mutations in the Saccharomyces cerevisiae mitochondrial Hsp60 lead to a folding defect in a number of different newly translocated proteins including the monomeric protein DHFR, which in vitro can fold without chaperone assistance (Cheng et al., 1989; Ostermann et al., 1989). Similarly, Lorimer and co-workers had succeeded in reconstituting the in vitro refolding of ribulose-1,5bisphosphate carboxylase/oxygenase (RuBisCo) in a reaction that was absolutely dependent on GroEL, GroES, and ATP hydrolysis (Goloubinoff et al., 1989a; Viitanen et al., 1990). The Hsp60 chaperonin family, however, appeared to pose a challenge to the generality of the requirement for molecular chaperones. In contrast with the Hsp70 and Hsp90 families, there was no obvious GroEL homolog in the eukaryotic cytosol. Indeed, the only homologs to GroEL outside of bacteria were localized in endosymbiotically derived organelles: mitochondria and chloroplasts. The identiﬁcation of the true eukaryotic homolog of bacterial Hsp60 began in neither of these kingdoms. Instead, acting on emerging evidence indicating that archae were more closely related to eukaryotes than to bacteria, Horwich and co-workers studied the heat shock response of the archaebacteria Sulfolobus shibatae (Trent et al., 1991). This organism normally grows at 70◦ C; but on exposure to even more extreme temperatures (85◦ C), S. shibatae produces a single heat shock protein, termed TF55 (Trent et al., 1990), which, as its name implies, has a molecular weight similar to that of the 60-kDa GroEL (Fig. 1). Like

348

KEVIN J. TRAVERS ET AL.

FIG. 1. Heat shock of Sulfolobus shibatae leads to the production of a single protein. Cells were pulse-labeled with [35S]methionine at 70◦ C (left lane) or 85◦ C (right lane) and newly synthesized proteins were analyzed by SDS-PAGE. Only a single polypeptide, named TF55, incorporates label under heat shock conditions. (Modiﬁed from Trent et al., 1990.)

GroEL, TF55 forms a ring structure, hydrolyzes ATP, and binds unfolded proteins (Trent et al., 1991). However, the sequence of TF55 shows little similarity to GroEL and instead is closely related to the ubiquitous eukaryotic protein t-complex polypeptide 1 (TCP1). Subsequent analyses demonstrated that TCP1, also known as TRiC and CCT, is indeed a ring complex that is required for the folding of at least a few essential proteins, such as actin and tubulin, both in vivo and in vitro (Frydman et al., 1992; Gao et al., 1992; Lewis et al., 1992; Yaffe et al., 1992). Moreover, a recent microarray-based analysis revealed that in human cells TCP1 is also induced in response to heat (Schena et al., 1996). Thus, TCP1 in eukaryotes and TF55 in archae form a related family of ubiquitous, heatinducible molecular chaperones, termed the class II chaperonins, that are distantly related to bacterial GroEL (Horwich and Willison, 1993; Kim et al., 1994). The discovery that TCP1 is the eukaryotic chaperonin immediately raised the question of whether it also required a GroES-like co-chaperonin. While GroEL can assist the folding of some substrates in vitro without GroES, this is the case only for proteins that are also able to fold without GroEL’s assistance (Martin et al., 1991; Schmidt et al., 1994); in vivo, GroES is required for life under all conditions (Fayet

FUNCTIONAL GENOMIC APPROACHES

349

et al., 1989; Hohfeld and Hartl, 1994). A convergence of structural and mechanistic studies had established the role for GroES in chaperonin folding: GroES acts as a lid which both displaces polypeptides into the central cavity of GroEL and prevents them from diffusing away as they fold in the protected environment provided by the chaperone (Chen et al., 1994; Mayhew et al., 1996; Weissman et al., 1995, 1996; Xu et al., 1997). Given this central role for GroES, it was difﬁcult to imagine how TCP1 could act as a chaperone for the folding of such notoriously recalcitrant proteins as actin and tubulin without a similar lid. In principle, of course, it was always possible that the eukaryotic co-chaperonin was present but not detected; however, the fact that only TF55 was induced during heat shock (Trent et al., 1990) and the subsequent failure to detect a GroES homolog in the completed yeast genome (Goffeau et al., 1996) all but ruled out this possibility. The crystal structures of archae chaperonins (Ditzel et al., 1998; Klumpp et al., 1997) suggested an elegant solution to how the class II chaperonins can function without a co-chaperonin partner: The archae protein and its eukaryotic homologs contain a helical protrusion not present in GroEL that can itself form a lid that mimics GroES. This allows TCP1 to act as an integrated GroEL– GroES complex. B. The Hsp104/Clp Family: A Connection between Proteolytic and Folding Machinery The heat shock response encompasses many classes of protein that fall outside the three major classes mentioned previously (Hsp60, Hsp70, and Hsp90) (Richmond et al., 1999; Schena et al., 1996). Cloning and characterization of the major large heat shock protein in S. cerevisiae, Hsp104, has revealed a direct link between the cellular proteolytic and refolding machineries (for reviews, see Gottesman et al., 1997; Schirmer et al., 1996). In yeast, Hsp104 plays a critical role in the acquired tolerance to a variety of stresses including exposure to high temperature (Sanchez and Lindquist, 1990; Sanchez et al., 1992). Cloning of Hsp104 by Lindquist and co-workers revealed that it was homologous to the ClpA/B proteins of E. coli (Parsell et al., 1991). Like Hsp104, ClpA/B are heat-inducible proteins that play an important role in thermotolerance (for reviews, see Gottesman et al., 1997; Wickner et al., 1999). ClpA functions in protein degradation by unfolding substrates in an ATPdependent manner and feeding them into the ClpP proteolytic chamber in a manner thought to be analogous to that in which the 19S complex assists protein degradation by the 20S proteasome (Baumeister et al., 1998; Bochtler et al., 1999; Hoskins et al., 1998; Kim et al., 2000; Thompson et al.,

350

KEVIN J. TRAVERS ET AL.

1994; Weber-Ban et al., 1999; Wickner et al., 1994); however, Hsp104 does not appear to have any signiﬁcant role in protein degradation (Parsell et al., 1991). Rather, Hsp104 is a bona fide molecular chaperone that specializes in the ATP-dependent reactivation of protein aggregates formed at high temperatures (Parsell et al., 1994). This activity can be reproduced in vitro, where Hsp104 can act in concert with Hsp70/40 chaperones to allow refolding of otherwise unrecoverable protein aggregates (Glover and Lindquist, 1998). In addition, Hsp104 also has an intriguing role in the maintenance of the yeast prion [PSI+], which is mediated by self-propagating aggregates of the endogenous translation termination factor Sup35p (Chernoff et al., 1995). The functional relationship between Hsp104 and ClpA/B has now come full circle with the realization that, depending on the substrate and circumstance, ClpA and/or ClpB can promote the degradation, unfolding, or even refolding of proteins (Diamant et al., 2000; Goloubinoff et al., 1999; Pak et al., 1999; WeberBan et al., 1999; Wickner et al., 1994; Wickner et al., 1999). The unifying feature of degradation and unfolding/refolding activities is the ability of these family members to use energy derived from ATP hydrolysis to break up inappropriate protein structures, whether the ultimate goal is recovery or destruction of the substrate. At ﬁrst glance it seems odd that proteases, which are involved in protein destruction, and chaperones dedicated to protein refolding should share the same components, as in the case of bacterial ClpA/B proteins, but a good case can be made that it makes both functional and mechanistic sense. First, misfolded proteins must undergo “triage” (Gottesman et al., 1997) in which the protein is either handed over to the chaperone machinery for another attempt at refolding or sent to be recycled by proteases. Although little is known about how such decisions are made, sharing certain components might, in principle, facilitate the coordination of these processes. Second, misfolded proteins, especially highly aggregated ones, pose a problem similar to efforts at both refolding and degradation: The incorrect structure must be dissolved before either event can occur efﬁciently. Thus, the basic requirements for the regulatory subunits of a protease and a chaperone specializing in refolding are the same—the ability to recognize and reverse misfolded structures—even if their ends are diametrically opposed. C. Hsp33: A Redox-Regulated Chaperone The heat shock response in E. coli is arguably the best studied stress response in terms of both the scope and function of its target genes (reviewed in Gross, 1996). Whole genome analysis began in earnest with

FUNCTIONAL GENOMIC APPROACHES

351

the studies of Blattner and co-workers, who used lambda clone spot blots to monitor heat-induced products at ten-gene resolution (Chuang and Blattner, 1993). Subsequently, these studies were extended to singlegene resolution (Richmond et al., 1999) using both glass- (Schena et al., 1995) and nylon-based high-density arrays of PCR products. These studies found 77 genes that were induced at least 5-fold at the RNA level on temperature upshift. Although many of the known targets were chaperones and proteases, the function of roughly a third of these genes was unknown. Taking advantage of this wealth of data, Bardwell, Jakob, and co-workers have recently begun to systematically analyze the function and structure of the conserved but previously uncharacterized components of the E. coli heat shock response (Bugl et al., 2000; Jakob et al., 1999; Korber et al., 2000, 1999; Staker et al., 2000). The ﬁrst protein analyzed, Hsp33, turned out to be a molecular chaperone, but with a twist ( Jakob et al., 1999, 2000). Unlike traditional chaperones such as Hsp70 and GroEL, where substrate binding is regulated by ATP binding and hydrolysis, Hsp33’s ability to interact with unfolded proteins is determined by its redox state. In vitro, oxidized Hsp33 efﬁciently interacts with denatured citrate synthase and inhibits its aggregation. Addition of a reducing agent, however, causes reduction of an internal disulﬁde bond in Hsp33 and completely inhibits its chaperone activity. Subsequent addition of hydrogen peroxide oxidizes the internal disulﬁde bond and fully restores chaperone function. In the cytosol, where Hsp33 resides, protein disulﬁde bonds are in general quite rare and only ∼20% of Hsp33 is found in the oxidized (active) form. By contrast, deletion of the thioredoxin and glutathione reductase genes, which is known to increase the level of oxidation in the cytosol (Derman et al., 1993; Prinz et al., 1997), results in disulﬁde bond formation in ∼60% of Hsp33 molecules. Moreover, deletion of Hsp33 in a strain missing thioredoxin dramatically increases the cell’s sensitivity to hydrogen peroxide. Taken together, these data argue that Hsp33 is a molecular chaperone that is activated by oxidation and appears to play a role in protecting against oxidative damage, implying an as-yet uncharacterized relationship between heat and oxidative stress and raising the question of why a redox-responsive chaperone would be signiﬁcantly upregulated by heat stress. D. The Heat Shock Response and Protein Translation Adjacent to Hsp33 on the chromosome is another highly conserved protein, Hsp15, which shows a 43-fold induction at high temperature, a level that is considerably greater than nearly all of the well-studied

352

KEVIN J. TRAVERS ET AL.

heat shock proteins, including GroEL and DnaK (Hsp70) (Korber et al., 2000). Hsp15 is not a molecular chaperone, but rather a speciﬁc RNA binding protein with structural similarities to RNA binding motifs found in a ribosomal protein (S4) and some tRNA synthetases (Staker et al., 2000). Analysis of the in vivo substrates of Hsp15 revealed that it recognized the 50S ribosome with high afﬁnity (KD 5 nM ). By contrast, Hsp15 does not recognize the 70S form of the ribosome and binding to the 50S subunit appears to require the presence of unreleased nascent chains. These data raise the intriguing possibility that an important defect during heat stress is the premature dissociation of 70S ribosomes, leading to accumulation of free 50S ribosomes that have failed to release their nascent chains, and that Hsp15 plays a role in recognizing and resolving such complexes. Studies of additional stress-induced proteins, FtsJ (Bugl et al., 2000) and MiaA (Tsui et al., 1996), further strengthen the connection between protein translation and the heat shock response. FtsJ is a highly conserved methytransferase that methylates the 23S ribosomal RNA. Deletion of FtsJ caused a signiﬁcant growth defect even under nonstress conditions, an increase in temperature sensitivity, and a signiﬁcantly altered polysome proﬁle (Bugl et al., 2000). Intriguingly, an additional protein involved in RNA metabolism, the tRNA dimethylallyl diphosphate transferase MiaA, is also under heat shock regulation. The exact nature and extent of the defect in RNA metabolism under heat stress are not fully understood, and how stress proteins repair these defects remains an open question. Nonetheless, it is now clear that the protein translation machinery, along with the folding and degradation system, is a major target of the heat shock response. Recently, Brown and co-workers completed a comprehensive analysis of the response of Saccharomyces cerevisiae to numerous stress conditions in addition to heat shock (Gasch et al., 2000). In this study, the authors used high-density spotted microarrays (Schena et al., 1995) to analyze the transcriptional proﬁles resulting from a variety of temperature changes, oxidative damage, hyper- and hypoosmotic shock, and nutrient deprivation, among other conditions. The transcriptional proﬁles resulting from each of these treatments were examined over a time course, resulting in a total of 142 measurements for each gene. Using a method of analyzing the resulting data that groups genes based on similarity of expression under each condition (termed hierarchical clustering; see Section V), Gasch and colleagues identiﬁed a set of genes that responds to all of these conditions, a set they refer to as the environmental stress response, or ESR. Approximately 600 genes are repressed during the ESR, including genes involved in a number of growth-related

FUNCTIONAL GENOMIC APPROACHES

353

processes, RNA metabolism, nucleotide biosynthesis, secretion, and genes encoding ribosomal components. Approximately 300 genes are ESR-induced, including genes in such functional categories as carbohydrate and fatty acid metabolism, protection from oxidative stress, redox regulation, cell wall modiﬁcation, protein folding and degradation, and DNA damage repair. More detailed analysis of the ESR suggests that cellular responses to stress are not under the control of a single regulatory pathway, but are more likely to represent a summation of individual stress response pathways. In summary, we have seen a number of striking examples in which analysis of proteins identiﬁed as conserved targets of the heat shock response in bacteria, archae, or eukaryotes has led to fundamental insights into the relationship between protein synthesis, folding, and degradation. The rapid advances in genomics and expression analysis are now providing a wealth of new data to mine. Functional characterization of these targets will naturally lag behind. Nonetheless, the availability of genetically manipulable model organisms such as E. coli, S. cerevisiae, and even mice, makes it possible to directly test function in vivo. In addition, advances in X-ray crystallography, including semiautomated phase determination using multiple anomalous dispersion (MAD) and strong, tunable synchrotron radiation sources, as well as improvements in multidimensional nuclear magnetic resonance (NMR) have dramatically increased the speed with which new structures can be determined (Abola et al., 2000; Burley, 2000; Montelione et al., 2000). It is now reasonable to start with structure determination as an initial step in the functional characterization of a protein, and, indeed, similarities at the level of tertiary structure can often provide insights that would not emerge from similarities in the primary sequence alone (Thornton et al., 2000). Given these powerful tools, it is clear that such “functional genomics” approaches will become increasingly important in the analysis of protein folding and stress responses. III. FUNCTIONAL AND GENOMIC ANALYSIS OF THE UNFOLDED PROTEIN RESPONSE The previous section looked at examples in which detailed analysis of speciﬁc proteins identiﬁed as targets of the heat shock response has led to the identiﬁcation of new molecular chaperones as well as an expanded view of the requirements for efﬁcient protein production and folding. In this section we focus on how the use of DNA microarrays has provided a global view of the spectrum of targets of a speciﬁc stress response, the unfolded protein response (UPR), which plays a critical

354

KEVIN J. TRAVERS ET AL.

role in allowing the cell to survive ER folding stress and to correct folding defects in the ER. The methodology used to distinguish the direct targets of the UPR from other genes whose transcription might be indirectly affected by conditions that disrupt ER folding is described in the following section. A. Protein Folding in the Secretory Pathway All proteins of the plasma membrane, secretory organelles, and extracellular space are ﬁrst translocated into the ER in an unfolded state (for reviews, see Brodsky and McCracken, 1999; Ellgaard et al., 1999; Helenius et al., 1992). The ER is then responsible for the proper folding and maturation of these proteins prior to their leaving for their ﬁnal destinations. The output of the ER as a protein-folding vesicle can be enormous—in some cells the daily ﬂux exceeds the cell’s entire mass (Helenius et al., 1992). In contrast to the relatively simple requirement for folding in the cytosol, ER folding is typically dependent on covalent modiﬁcations such as glycosylation and disulﬁde bond formation, and even the correct orientation of sequences that span the lipid bilayer. Furthermore, the ER must target mature proteins to their ﬁnal destination; and in the case of improperly folded or immature proteins waiting to undergo oligomerization, it must prevent them from moving through the secretory pathway and eventually target them for destruction. The ER presents an environment highly enriched in factors that distinguish between correct and incorrect structures and that assist the folding of nascent polypeptides, making studies of the ER particularly fruitful for efforts to understand how efﬁcient folding is achieved in vivo. Numerous studies have therefore focused on the particular folding factors found within the ER. Not surprisingly, many homologs of cytoplasmic heat shock proteins are found within the ER, including a number of Hsp70 homologs [e.g., Kar2p (Rose et al., 1989) and Lhs1p (Craven et al., 1996) in S. cerevisiae], peptidylprolyl isomerases [e.g., Fkb2p (Partaledis et al., 1992)], and Hsp90 in higher eukaryotes. The entire machinery for synthesis of N-linked sugars resides within the ER or in its membrane, where carbohydrates are attached to proteins as they are synthesized. In addition, the ER contains a number of proteins that promote the correct formation of disulﬁde bonds [e.g., protein disulﬁde isomerase (PDI) and Ero1p]. The ER also has a robust “quality control” system responsible for recognizing misfolded proteins and preventing their progress through the secretory pathway (for reviews, see Brodsky and McCracken, 1999; Ellgaard et al., 1999; Hurtley and Helenius, 1989). This quality control system provides the dual beneﬁt of keeping folded proteins within

FUNCTIONAL GENOMIC APPROACHES

355

the chaperone-rich folding environment of the ER and preventing misfolded proteins from passing to the cell surface where they could potentially be toxic. The exact mechanism by which the quality control system works remains unclear in many cases, but it seems likely that multiple different mechanisms contribute to quality control. These may act at the level of active retention in the ER by immobilization or aggregation, inhibiting recognition by the ER export machinery, or promoting continuous retrieval from post-ER compartments (Brodsky and McCracken, 1999; Ellgaard et al., 1999; Nehls et al., 2000). Despite the ER’s highly optimized folding environment, an inevitable consequence of the large ﬂux of proteins through the ER is that the folding process will occasionally fail, resulting in the production of unrecoverably misfolded proteins. Individually, irrevocably misfolded ER proteins are blocked from further progress through the secretory pathway and are retrotranslocated into the cytosol, where they undergo ubiquitin- and proteasome-dependent degradation, a set of constitutive multistep processes collectively known as ER-associated degradation (ERAD) (Hampton and Bhakta, 1997; Hampton et al., 1996; Werner et al., 1996; Wiertz et al., 1996). This well-studied process involves ER membrane proteins, chaperones, and cytosolic components of the ubiquitination and proteolytic machinery (reviewed in Brodsky and McCracken, 1999). On a global level, however, when unfolded proteins accumulate in the ER, another pathway, the UPR, is activated. During the UPR, information regarding the folding state of the ER is communicated to the nucleus, where transcription of target genes is upregulated (Cox et al., 1993; Cox and Walter, 1996; Mori et al., 1993, 1996). Speciﬁcally, accumulation of unfolded ER-resident proteins activates the transmembrane kinase/nuclease Ire1p (Cox et al., 1993; Mori et al., 1993; Shamu and Walter, 1996), which initiates the nonconventional splicing of HAC1 mRNA, enabling its efﬁcient translation (Cox and Walter, 1996; Gonzalez et al., 1999; Kawahara et al., 1997, 1998; Mori et al., 1996; Sidrauski et al., 1996; Sidrauski and Walter, 1997). Hac1p is a bZIP transcription factor, which is thought to directly induce transcription of UPR target genes. The response is dispensable for normal growth but absolutely required for survival of folding stress, suggesting that the target genes act to increase folding capacity and/or alleviate the toxic consequences of diminished folding competence. Prior to global gene expression analysis using DNA microarrays, only a handful of UPR target genes had been identiﬁed, and all but one of these encoded an ER-resident chaperone (reviewed in Chapman et al., 1998). The goal of expression analysis was to determine comprehensively the scope of UPR transcriptional output and to deduce from the functions of target genes the mechanisms by which the UPR allows the cell to survive conditions resulting in folding stress.

356

KEVIN J. TRAVERS ET AL.

B. Targets of the Unfolded Protein Response Using high-density microarrays in conjunction with UPR-defective strains as described in Section IV, the complete spectrum of transcriptional targets of the UPR has recently been deﬁned (Casagrande et al., 2000; Travers et al., 2000). The list of speciﬁc UPR targets identiﬁed by these studies is remarkably large, consisting of nearly 400 genes (more than 5 percent of the yeast genome). Despite the large number of targets, the UPR appears to be focused in large part on the secretory pathway (Fig. 2; detailed in Table I). At the time of our studies, more than half (208) of the 381 UPR targets had either been previously characterized

FIG. 2. Many aspects of secretory pathway function are transcriptionally induced by the UPR. A schematic diagram of the secretory pathway is shown. The number of genes whose function is either known or can be inferred from homology to characterized genes is indicated underneath each functional category. See Table I for a complete listing of the genes in each category. (Reproduced from Travers et al., 2000.)

FUNCTIONAL GENOMIC APPROACHES

357

TABLE I Secretory Pathway Genes Upregulated by the UPR a Functional categories

Genes

Translocation Translocon Posttranslational translocation Signal peptidase

SEC61, SBH1, SLS1 SEC62, SEC71, SEC72 SPC2

Glycosylation/modification Core oligosaccharide synthesis Oligosaccharyltransferase Glycoprotein processing GPI anchoring Golgi/O-linked glycosylation

DPM1, PM140, RHK1, SEC59 OST2, OST3, SWP1, WBP1 ALG6, ALG7, MNS1, RAM2, STE24 GAA1, GPI12, LAS21, MCD4 KTR1, MNN11, PMT1, PMT2, PMT3, PMT5

Protein folding Chaperones Disulﬁde bond formation

FKB2, JEM1, LHS1, SCJ1, YFR041C b ERO1, EUG1, MPD1, MPD2, PDI1

Protein degradation ER-associated degradation (ERAD) DER1, HRD1/DER3, HRD3, UBC7 Ubiquitin/proteasome DOA4, PEX4 Vesicle trafficking/transport Budding (ER-Golgi) Fusion (ER-Golgi) Retrieval (Golgi-ER) Distal secretion Lipid/inositol metabolism Fatty acid metabolism Heme biosynthesis Phospholipid biosynthesis Sphingolipid biosynthesis Sterol metabolism Vacuolar protein sorting Cell wall biogenesis

ERV25, SEC12, SEC13, SEC16, SEC24, SED4, SFB2, SFB3, YMR040W c BOS1, TRS120 ERD2, RER2, RET2, SEC26, SEC27 APL3, ARL3, BFR1, MYO5, SEC6, TUS1, YPT10 ACB1, HAP1, MGA2, YJR107W d DFR1, HEM12, HEM13, HEM15, RIB1 EPT1, INP51, LPP1, OP13, SCS3, SLC1 LCB1 ARE1, HMG2, YHR073W e LUV1, STP22, VPS17, VPS35 CHS7, CSR1, ECM3, ECM8, ECM31, EXG2, GAS5, PKC1, SPF1, YKR027W f YOR239W g

a Bold headings correspond to functional categories illustrated in Figure 2. Genes are assigned to subcategories according to summaries of published data listed in the Saccharomyces Genome Database (SGD) and Yeast Protein Database (YPD). b DnaJ homolog with predicted signal sequence. c Homolog of BAP-29 and BAP-31, sorting proteins that control anterograde transport of certain membrane proteins from the ER to the Golgi complex (Ng et al., 1997). d Homolog of acylglycerol lipase. e Homolog of human oxysterol-binding protein (OSBP). f Homolog of Chs6p, involved in localization of Chs3p. g Homolog of Chs5p, required for protease activation of Chs3p.

358

KEVIN J. TRAVERS ET AL.

or were homologous to previously characterized genes. Of these 208 characterized targets, approximately half are either known or suspected on the basis of homology to play roles in secretion. Moreover, classes of molecular chaperones found exclusively outside the ER in yeast, such as Hsp60, Hsp90, and Hsp104, are not induced by the UPR; even within a single chaperone family, only those members that are ER-localized are UPR targets. For example, there are 15 DnaJ homologs in yeast, of which only the three ER-localized proteins JEM1 (Nishikawa and Endo, 1997), SCJ1 (Blumberg and Silver, 1991), and YFR041c are induced by the UPR (Fig. 3). Similarly, the ER-localized Hsp70 homologs, KAR2 and LHS1, are UPR targets, whereas no mitochondrial or cytosolic Hsp70s are induced.

FIG. 3. UPR induction is speciﬁc to the secretory pathway. The inductions of three DnaJ homologs under each of the nine indicated experimental conditions (see Section IV) are shown. Each induction is indicated as a bar representing the log2 fold induction relative to an untreated strain. The three homologs shown here are JEM1 (localized to the secretory pathway), MDJ1 (localized to the mitochondria), and YDJ1 (localized to the cytoplasm). Only JEM1 shows a UPR-dependent induction following DTT and tunicamycin treatment. (Reproduced from Travers et al., 2000.)

FUNCTIONAL GENOMIC APPROACHES

359

The targets of the UPR are not limited to the ER. Rather, the UPR affects virtually every stage of the secretory pathway (Fig. 2; Table I). Target genes that play a role in secretion can be grouped into functional categories: r Translocation, including components of the translocon [SEC61

(Sanders et al., 1992) and SBH1 (Finke et al., 1996)], the signal peptidase [SPC2 (Mullins et al., 1996)], and the posttranslational translocation complex [SEC62/71/72 (Panzner et al., 1995)], but no subunits of the signal recognition particle (SRP) or SRP receptor. r Protein glycosylation, including synthesis of the core N-linked oligosaccharide [DPM1 (Orlean et al., 1988), PMI40 (Smith et al., 1992), RHK1 (Kimura et al., 1997), and SEC59 (Bernstein et al., 1989)], the oligosaccharyltransferase itself [OST2, OST3, SWP1, and WBP1 (Silberstein and Gilmore, 1996)], covalent modiﬁcation of glycoproteins [ALG6 (Runge et al., 1984), ALG7 (Barnes et al., 1984), MNS1 (Camirand et al., 1991), RAM2 (Mayer et al., 1993), and STE24 (Fujimura-Kamada et al., 1997)], O-linked glycosylation [KTR1 (Lussier et al., 1996); MNN11 ( Jungmann et al., 1999), PMT1, PMT2, PMT3, and PMT5 (Gentzsch and Tanner, 1996)], and synthesis and attachment of glycosylphosphatidylinositol (GPI) anchors to plasma membrane–bound glycoproteins [GAA1 (Hamburger et al., 1995), GPI12 (Watanabe et al., 1999), LAS21 (Benachour et al., 1999), and MCD4 (Gaynor et al., 1999)]. r Vesicular transport, including the COPII vesicle coat [SEC13, SEC16 (Kaiser and Schekman, 1990), SEC24 (Salama et al., 1993), and ERV25 (Belden and Barlowe, 1996)], vesicle budding [SEC12 (Barlowe and Schekman, 1993), SED4 (Gimeno et al., 1995), SFB2, SFB3 (Peng et al., 2000), and YMR040W (Ng et al., 1997)], vesicle fusion with Golgi membranes [BOS1 (Newman et al., 1990) and TRS120 (Sacher et al., 1998)], and the retrieval of ER proteins from the Golgi bodies [ERD2 (Semenza et al., 1990), RER2 (Sato et al., 1999), RET2 (Cosson et al., 1996), SEC26, and SEC27 (Duden et al., 1994)], as well as several genes encoding proteins engaged in trafﬁc between the Golgi bodies, distal secretory organelles, and cell surface [APL3 (Cowles et al., 1997), ARL3 (Huang et al., 1999), BFR1 (Jackson and K´ep`es, 1994), MYO5 (Geli and Riezman, 1996), SEC6 (TerBush and Novick, 1995), TUS1, and YPT10 (Louvet et al., 1999)]. r Cell wall biosynthesis and maintenance, including structural components of the cell wall [ECM3, ECM8, ECM31 (Lussier et al., 1997), EXG2 (Larriba et al., 1993), and GAS5], the chitin synthase pathway [CHS7 (Trilla et al., 1999), CSR1 (Santos and Snyder, 2000),

360

KEVIN J. TRAVERS ET AL.

YKR027W, and YOR239W ], and regulatory factors [PKC1 (Iguala et al., 1996) and SPF1 (Suzuki and Shimma, 1999)]. r Vacuolar protein targeting [LUV1 (Smith et al., 1998), STP22 (Li et al., 1999), VPS17 (Kohrer and Emr, 1993), and VPS35 (Paravicini et al., 1992)]. r Protein degradation, including ER-resident proteins of the ERAD system [DER1 (Knop et al., 1996), HRD1, HRD3 (Hampton et al., 1996), and UBC7 (Hiller et al., 1996)] and cytosolic ubiquitinconjugating enzymes [DOA4 (Papa and Hochstrasser, 1993) and PEX4 (Wiebel and Kunau, 1992)]. Given that half of the UPR-induced ORFs for which functional data are available have a known role in the secretory pathway, the 173 transcriptional targets of the UPR with no currently known function are particularly strong candidates for genes with important secretory functions. For example, expression analysis suggested that the previously uncharacterized S. cerevisiae homolog of the Yarrowia lipolytica gene SLS1 [ORF YOL031c; later named PER100/SIL1 (Tyson and Stirling, 2000)] merited further investigation in the context of ER or secretory function, a suggestion borne out by our ﬁnding that PER100/SLS1 is necessary for efﬁcient ERAD (Travers et al., 2000). Despite its breadth, the UPR does not result in an indiscriminate induction of all ER or secretory components. For example, many components of the COPII vesicle coat involved in ER–Golgi trafﬁcking are targets, including particularly strong induction of two homologs of SEC24 which have been suggested to play a role in export of speciﬁc cargo (Roberg et al., 1999). In contrast, components of the COPI/“coatomer” involved in retrograde transport are less well represented, and those that are regulated by the UPR typically show only modest induction. Similarly, of the ER-resident chaperones involved in folding or secretion of speciﬁc substrates (Ellgaard et al., 1999; Lau et al., 2000), only CHS7, required for maturation of chitin synthase III (Trilla et al., 1999), is upregulated. This speciﬁcity suggests that rather than merely increasing the capacity of the secretory pathway, the UPR results in a remodeling of the secretory pathway, with speciﬁc induction of those activities essential under folding stress.

C. Coregulation of ERAD and the UPR The strong and speciﬁc induction of ERAD components was particularly intriguing, as it suggested a connection between the UPR and another pathway related to protein misfolding in the ER. A convergence of

FUNCTIONAL GENOMIC APPROACHES

361

mechanistic studies from a number of different labs has now conﬁrmed an intimate coordination between these responses. First, efﬁcient ERAD requires an intact UPR. In particular, deletion of IRE1 decreased the rate of degradation of CPY∗ (Travers et al., 2000), a well-characterized ERAD substrate that is a constitutively misfolded form of the vacuolar peptidase CPY (Knop et al., 1996). Similarly, deletion of IRE1 slowed the ERAD-mediated degradation of the MHC class I heavy chain (H-2Kb) in yeast (Casagrande et al., 2000). Second, loss of ERAD function leads to chronic UPR induction. Studies with the constitutively misfolded protein CPY∗ led to the identiﬁcation of a number of factors whose activity was required for ERAD (Knop et al., 1996). These mutants showed a small but signiﬁcant induction of the UPR (Friedlander et al., 2000; Knop et al., 1996; Travers et al., 2000). Alleles of SEC61 with speciﬁc defects in ERAD, as well as deletions of several other ERAD components, also caused constitutive UPR induction (Zhou and Schekman, 1999). Thus the chronic accumulation of misfolded proteins in the ER appears to be a general consequence of loss of ERAD. Third, simultaneous loss of ERAD and UPR function greatly decreases cell viability. Ng and co-workers conducted a screen to identify genomic mutations synthetically lethal with loss of the UPR (Ng et al., 2000). This screen resulted in the isolation of a large panel of mutants which were then further classiﬁed based on functional analysis. Analysis of the rate of degradation of CPY∗ indicated that one-third of their mutants had defects in ERAD. The functional signiﬁcance of this genetic interaction is emphasized by the ﬁnding that UPR defects are synthetically lethal with defects in a large number of ERAD components [SON1, UBC1, UBC7, HRD1, HRD3, and DER1 (Friedlander et al., 2000; Ng et al., 2000; Travers et al., 2000)] which act at multiple steps in the ERAD pathway. The coordination between the UPR and ERAD revealed in the above studies argues that the ERAD system can no longer be thought of as solely a “constitutive” response. Rather, through the action of the UPR, a cell can modulate the functional level of ERAD. This is most clearly demonstrated by the increased rate of degradation of CPY∗ seen when the UPR is induced without increasing the amount of unfolded protein (Travers et al., 2000). Likewise, the UPR cannot be considered simply “dispensable under normal growth conditions,” but rather must be thought of as one aspect of a redundant system which serves an essential role. Specifically, the identiﬁcation of synthetic lethality between UPR and ERAD components argues that the UPR plays an important role even in the absence of acute stress.

362

KEVIN J. TRAVERS ET AL.

FIG. 4. Schematic model illustrating the coordinated action of the UPR with other cellular systems. As misfolded proteins enter the endoplasmic reticulum, they either fold to their native states and are then moved through the rest of the secretory pathway, or fail to properly fold. In the event that a protein unrecoverably fails to fold, it can be degraded by the ERAD machinery. During cellular stress, an accumulation of misfolded proteins leads to activation of the UPR (gray arrow). In the model illustrated here, in addition to upregulation of chaperones to directly assist the folding of proteins, the UPR enhances the rate of secretion to the distal secretory pathway and the degradation of misfolded species. (Reproduced from Travers et al., 2000.)

Taken together, these considerations suggest that the ER is constantly generating irrevocably misfolded proteins and that removal of such proteins is an essential process under all growth conditions (Fig. 4). In the absence of a functional UPR, the ERAD capacity is sufﬁcient to dispose of this ﬂux, provided the cell does not face an unusual stress. Conversely, when the UPR is available in a cell with diminished ERAD capacity, accumulating misfolded proteins can still be handled by multiple mechanisms, including refolding by chaperones, clearance from the ER by anterograde vesicular transport, or an alternative means of degradation (e.g., in the vacuole or by the action of other UPR-induced ERAD genes). Thus, the UPR and ERAD represent partially overlapping or compensatory means to the same essential end: elimination of misfolded secretory proteins, which are inevitably generated during the course of normal growth. The above studies establish a critical functional link between the ERAD system and the UPR. How the other activities induced by the UPR improve the state of folding within the ER is far less clear. Nevertheless, it is tempting to speculate that these various activities induced by the UPR act in concert to reduce the lumenal concentration of misfolded protein,

FUNCTIONAL GENOMIC APPROACHES

363

either by directly refolding proteins or by removing them from the ER (Fig. 4). In this “ﬁx-or-clear” model, abundant chaperones would bind to misfolded species, prevent aggregation, and promote folding. Similarly, glycosylation enzymes would assist in the folding of proteins that require carbohydrate modiﬁcation to attain their proper conformation. Consistent with this suggestion, mutations that compromise either addition of GPI anchors or protein glycosylation are lethal in the absence of UPR function (Ng et al., 2000). Moreover, UPR induction in mammalian cells was recently shown to accelerate synthesis of the dolichol oligosaccharides employed in asparagine-linked glycosylation (Doerrler and Lehrman, 1999). In the event that direct attempts to increase the efﬁciency of folding fail, induction of speciﬁc COPII components might enable efﬁcient packaging of cargo proteins (possibly including unfolded proteins) into anterograde vesicles, or simply increase the overall capacity of anterograde transport. Such an increase in secretory capacity might facilitate targeting of misfolded species to the vacuole for degradation (Hong et al., 1996), consistent with our observation that several genes involved in vacuolar targeting are also UPR targets. Similarly, induction of inositol and lipid biosynthetic enzymes would generate new membranes, thereby increasing the volume of the ER, simultaneously diluting unfolded proteins and preparing the compartment to receive an inﬂux of newly synthesized folding factors. Finally, induction of ERAD components directly enhances the clearance of misfolded proteins from the ER to the cytosol, perhaps allowing them to be targeted for degradation in the vacuole. IV. STRATEGY FOR IDENTIFYING UPR TARGETS The UPR studies began with the objective of identifying the speciﬁc transcriptional targets of the UPR, with the ultimate goal of establishing the physiological role of this stress response by analysis of the functions of the target genes. A major challenge facing any effort to identify the speciﬁc transcriptional targets of a given cellular response such as the UPR is that virtually any treatment of a cell is likely to have myriad effects on that cell’s physiology. In this section we outline the strategies employed—drug treatments, genetic knockout strains, and analytical approaches—to determine the speciﬁc targets of the UPR. A. Experimental Strategy for Identifying Direct UPR Targets In the laboratory, the UPR is experimentally induced by the addition of chemical agents known to interfere with protein folding in the ER (Chapman et al., 1998). In yeast, reagents commonly employed are

364

KEVIN J. TRAVERS ET AL.

the reducing agent dithiothreitol (DTT), which prevents disulﬁde bond formation, and tunicamycin (Tm), which inhibits N-linked glycosylation of nascent secretory proteins. These agents are thought to speciﬁcally challenge protein folding in the ER, as disulﬁde bonds and N-linked glycosylation are vanishingly rare outside the secretory pathway. Nonetheless, treatment with either drug is likely to trigger responses unrelated to folding of secretory proteins. For example, treatment of cells with DTT induces the UPR but also slows cell growth, thus potentially affecting expression of the nearly 800 cell cycle–regulated genes (Spellman et al., 1998), few of which have anything to do with protein folding. In addition, the action of these drugs might activate bona fide responses to ER protein folding stress that act in parallel to the UPR. For example, Gasch and colleagues recently identiﬁed a set of genes that respond to numerous forms of stress, including DTT treatment, which they term the environmental stress response (see Section II; Gasch et al., 2000). In either case, a broad spectrum of genes, which includes but is not limited to direct UPR targets, would be affected by DTT or tunicamycin treatment. Therefore, it is critical that some means of distinguishing these contributions be established if the goal is to study the speciﬁc targets of the UPR rather than the gross transcriptional response to drug treatment. The strategy for identifying the direct targets of the UPR relied critically on the ability to “genetically dissect” this pathway from all other cellular responses: in essence, analyzing the transcriptional response to the inducing stimulus in the wild-type cell and in mutant cells deﬁcient in the pathway, and focusing on those changes occurring only in the wildtype (UPR-competent) cell. In order to make such a comparison, we took advantage of the wealth of knowledge already generated by genetic characterization of the pathway. The mechanism by which the accumulation of unfolded protein in the ER results in the speciﬁc and rapid induction of the UPR targets has been the subject of intensive studies. These efforts have revealed a linear signal transduction pathway between the ER and nucleus composed of two primary players, the ER-resident transmembrane kinase-nuclease Ire1p and the bZIP transcription factor Hac1p (see Section III). When unfolded proteins accumulate in the ER lumen, Ire1p becomes an active nuclease, initiating a nonspliceosomal splicing reaction in which a translation-inhibiting “intron” is removed from HAC1 mRNA (Gonzalez et al., 1999; Sidrauski and Walter, 1997); spliced HAC1 mRNA can then be translated into functional Hac1p, which activates transcription of UPR target genes (Cox and Walter, 1996; Mori et al., 1996; Sidrauski et al., 1996). Cells lacking either the IRE1 or HAC1 gene have identical phenotypes in that neither mutant is able to upregulate transcription of these canonical UPR target genes under conditions of

FUNCTIONAL GENOMIC APPROACHES

365

ER folding stress (Cox and Walter, 1996; Mori et al., 1996). While treatment of growing cells with either DTT or Tm might result in a host of transcriptional changes, similar treatment of ire1 or hac1 mutants would result only in changes unrelated to the UPR. Therefore, we operationally deﬁned the UPR target genes as those genes which respond to treatment with ER unfolding agents in a wild-type cell but not in a ire1 or hac1 mutant. We collected data on the transcriptional response to ER protein unfolding by genomic array hybridization. Brieﬂy, cultures of wild-type or UPR mutant cells were grown in parallel and harvested for mRNA puriﬁcation either immediately before or after speciﬁed times of DTT or Tm treatment. All strains were derived from the same parental stock, grown under identical conditions, and treated with the same concentration of drug, so that the sole variable distinguishing the cultures treated with a given drug was the presence (wild-type) or absence (ire1 or hac1) of an intact UPR pathway. Wild-type cells were treated with DTT over a time course (15, 30, 60, and 120 minutes) or with Tm for 60 minutes, and mutants were treated with either DTT or Tm for 0 or 60 minutes. Thus in total ﬁve measurements were taken from the wild-type strain and four from UPR-deﬁcient mutant strains. Relative mRNA levels were measured using high-density synthetic oligonucleotide array chips produced by Affymetrix (Wodicka et al., 1997). On these chips, each gene is represented by 40 “sequence elements,” subdivisions of the glass substrate on which 20-base oligonucleotides have been lithographically printed. Twenty of the oligonucleotides are perfect matches to various regions of the gene of interest, and the other 20 serve as controls for differential hybridization efﬁciency, identical to the perfect matches but containing a single mismatch at the center position. The measured expression level of each gene is a function of the relative intensities of the perfect match and mismatch sequence elements, and is essentially a weighted average of the differences between the corresponding pairs of perfect match and mismatch intensities (Chee et al., 1996; Lipshutz et al., 1999; Lockhart et al., 1996; Wodicka et al., 1997). An alternative strategy developed by Brown and co-workers utilizes high-density arrays of full-length cDNAs spotted onto glass slides (Schena et al., 1995). In this approach, RNA samples from two different conditions being compared are labeled with different-color ﬂuorophores and measured simultaneously. The degree of induction of given genes is then determined by measuring the ratio of hybridization of the two ﬂuorophores to the particular cDNA. Overall, it is clear that both approaches can yield high-quality and robust data. Indeed, using spotted

366

KEVIN J. TRAVERS ET AL.

cDNA arrays, Ploegh and co-workers obtained a qualitatively similar picture of the UPR as obtained by us using Affymetrix arrays (Casagrande et al., 2000). Thus the choice of which array to use is perhaps best made on the basis of more pragmatic concerns such as cost and availability. In order to calculate the fold change in mRNA level of each gene, hybridization intensities for each drug-treated sample were compared to the intensity in an untreated control sample of the same genotype. The data for each gene therefore include nine points: fold change relative to untreated controls for ﬁve treated wild-type samples and four treated mutant samples. Importantly, each of these nine points relates the expression level of a given gene in a particular strain after timed drug treatment to the expression level immediately prior to drug treatment. This minimizes any artifacts that might arise due to incidental mutations in the strains which accumulate during their prolonged propagation, or constitutive gene expression changes in the mutant strains which are unrelated to the acute stress response itself—difﬁculties that have plagued other gene expression studies (Hughes et al., 2000). B. Computational Strategy for Identifying Direct UPR Targets From the genome-wide picture of the comprehensive transcriptional response to treatment with agents that challenge ER folding, we stringently screened for genes whose expression patterns under these nine conditions ﬁt the principal criterion for consideration as speciﬁc targets of the UPR: namely, that the gene in question change consistently under both drug treatments in the wild-type but not in UPR-deﬁcient mutant cells. Three different and complementary tests were applied to identify genuine UPR targets: similarity of expression proﬁle to the proﬁles of previously characterized UPR targets, signiﬁcance of difference between fold changes in the wild-type and mutant strains, and a high absolute value of average fold change across the samples derived from wild-type cells. In the ﬁrst stage of the screen, we identiﬁed genes whose expression patterns closely matched those of seven known UPR targets. Naively, one might demand that fold change be positive in the wild-type and zero in the mutants. But previous work had shown that some of the known target genes of the UPR retained some residual regulation even in a ire1 mutant [e.g., KAR2 (Cox and Walter, 1996)]; therefore, we did not demand that upregulation of target candidates be completely absent in the UPR mutant samples. Instead, we compiled a “canonical response,” representative of a typical UPR target gene, by geometrically averaging the log-fold change in each condition for the seven known UPR targets [KAR2, PDI1, EUG1, FKB2, LHS1, ERO1, and INO1 (Chapman et al.,

FUNCTIONAL GENOMIC APPROACHES

367

FIG. 5. The canonical UPR transcriptional response. Transcriptional changes were measured in cells treated with nine different conditions as indicated (wild-type cells treated with DTT for 15, 30, 60, or 120 minutes; wild-type cells treated with tunicamycin (Tm) for 60 minutes; ire1 cells treated with DTT or Tm for 60 minutes; and hac1 cells treated with DTT or Tm for 60 minutes). The canonical response was calculated as the average of the log2 fold inductions for each of the seven known targets of the UPR (KAR2, LHS1, EUG1, PDI1, FKB2, ERO1, and INO1). (Reproduced from Travers et al., 2000.)

1998)]. This average shows a rapid response which is largely but not entirely complete after 15 minutes, steady or slowly increasing over the two-hour DTT time course, and minimal in UPR mutant samples (Fig. 5). The data for each gene were then compared against this canonical response and evaluated for similarity to the pattern of expression in the nine conditions represented. The metric used for comparison was essentially one of correlation (Eisen et al., 1998): Proﬁles for two genes under comparison are normalized relative to one another, and each element of one proﬁle is compared against the corresponding element of the other. This metric is independent of scale and is therefore able to recognize as similar genes those which display similar patterns across the conditions (time or genotype) but are of different magnitude. Also, rather than forcing residual regulation in the mutant to be below some particular arbitrarily determined value, the metric allows passage of genes whose residual regulation is simply small relative to the fold change in the wild-type. The correlation is sensitive to kinetic differences, i.e., the pattern across time points in the DTT time course. As such, it was suitable for eliminating those genes which are upregulated in late but not early time points and whose average induction in the wild-type is still high relative to induction in the UPR mutants. We wished to exclude these genes from consideration of the primary effects of the UPR, as they might represent the action of divergent transcriptional regulatory events downstream of HAC1 or of parallel pathways activated by cellular morbidity under long-term exposure to toxic drugs. Genes that passed the primary (correlation) screen for similarity to the known UPR target genes were secondarily screened by a statistical

368

KEVIN J. TRAVERS ET AL.

method (Z-test) to establish that the observed induction was not due to chance ﬂuctuation in the data. Speciﬁcally, the distribution (mean and standard deviation) of expression levels for a given gene in the wildtype samples was compared to the distribution of its expression levels in the mutant samples. A gene was “passed” if the wild-type distribution differed signiﬁcantly from the mutant distribution. Technically, we demanded that the differences in the mean induction in the wild-type and UPR-deﬁcient strains differed by at least 3.6 standard deviations, which corresponds to a chance occurrence of approximately 1 in 6000 (the number of genes in the yeast genome). Because the level of scatter in the data is determined empirically for each gene, this metric makes it possible to speciﬁcally disregard genes with noisy data while retaining genes that undergo subtle regulatory changes, provided these differences are statistically signiﬁcant. This gene-speciﬁc analysis of the scatter is critical, as the degree of accuracy of the expression data can vary dramatically from one gene to another depending on a speciﬁc gene’s expression level as well as the quality of the gene speciﬁc probes on the microarrays. The primary and secondary metrics are independently quite stringent (either one alone eliminates more than 90% of the genome) and are complementary when applied in serial: The correlation metric determines whether the expression pattern of a gene is similar to genes known to be speciﬁc targets of the response of interest, and the Z-test statistic determines whether this similarity (speciﬁcally, the qualitative difference between regulation in the wild-type and mutant cells) is quantitatively significant enough to merit attention. As both of these screens were independent of the magnitude of regulatory change, as has been done in other studies (Galitski et al., 1999), our ﬁnal criterion for inclusion as a UPR target was that the average induction in the wild-type strain be at least 1.5-fold. This ﬁnal metric serves to limit the scope of genes to be analyzed to those whose regulation by the UPR are most likely to result in a meaningful change in abundance of the encoded protein and might therefore be expected to affect cellular physiology. In some sense this ﬁnal criterion is more arbitrary than the ﬁrst two, as it is possible that even small changes in expression levels would be physiologically important. Fortunately, the number of genes eliminated by this test was relatively small and these consisted primarily of genes that passed the ﬁrst two screens because of repression in the UPR-deﬁcient strains. From a genome of 6300 genes and a transcriptional data set in which nearly 1000 genes underwent a greater than twofold change under at least one of the experimental conditions, we arrived at a more manageable set of 381 genes. The criteria outlined above were designed to provide a conservative estimate of the number of UPR targets. Indeed, two of the known target genes used to compile the canonical UPR target,

FUNCTIONAL GENOMIC APPROACHES

369

KAR2 and INO1, did not pass all three criteria. Thus, while the UPR certainly has transcriptional effects on genes not within our target set, we can claim with a high degree of conﬁdence that every gene within the target set represents a genuine and speciﬁc target of the UPR. The UPR is in many ways nearly ideally suited for the above analysis. The mechanism of the UPR has been studied in detail, and the genes required for pathway activation identiﬁed. The response is not essential, and therefore the genes involved can be readily eliminated in yeast, allowing study of knockout strains. Additionally, there are two mechanistically unrelated drugs (DTT and tunicamycin) that cause speciﬁc defects in folding in the ER. Without such an impressive arsenal for speciﬁcally dissecting the UPR, it would have been possible to delineate some, but not all, of the features of the response. To illustrate this point, we can now go back and reanalyze our data in a simpler, less “preinformed” way, comparing the resulting set of UPR targets to the set we deﬁned in our genetic analysis. For example, a straightforward form of analysis is to select those genes for which both DTT and tunicamycin treatment result in a certain level of induction. To perform this analysis, we averaged the inductions at each time point of DTT treatment to generate a single value to follow, then examined the numbers of genes induced by both drug treatments at various levels of induction. Not surprisingly, as more and more stringent criteria are used, fewer and fewer total genes are included, and the number of false positives decreases. By demanding that each potential UPR target be induced two-fold, we ﬁnd that we would select approximately 440 genes, a number comparable to the number of targets found to be a part of the response. However, we also found that at this level of induction, approximately half of the targets identiﬁed as bona fide UPR transcriptional targets are not identiﬁed. Instead, approximately 250 genes that we excluded from the list of UPR targets are included. This analysis suggests that, in fact, a large part of the UPR’s transcriptional output would have been identiﬁed by an analysis that focused on the magnitude of induction due to drug treatments of wild-type strains. However, we would have missed approximately half of the UPR, including such genes as the protein disulﬁde isomerase homolog MPD2 (Tachikawa et al., 1997) and the translocon components SEC66 (Feldheim et al., 1993) and SEC72 (Feldheim and Schekman, 1994). In addition, without the use of strains unable to induce the UPR, we would have identiﬁed a number of genes as UPR targets that clearly should not be, such as the cytosolic Hsp70 and Hsp40 homologs, SSA4 (Werner-Washburne et al,, 1987) and SCJ1 (Blumberg and Silver, 1991), respectively. Such false positives and false negatives would have confounded the qualitative categorization that led us to conclude that the UPR has signiﬁcant roles in many

370

KEVIN J. TRAVERS ET AL.

aspects of secretory biology, but is without a coherent role in the cytosolic compartment. V. AN OVERVIEW OF UNSUPERVISED SEARCH STRATEGIES A. Definition of Unsupervised Search The analysis of the ER response to folding stress described above is based on the assumption that the speciﬁc targets of the UPR would share a limited set of related functions that can act in a coherent way to aid protein folding in the ER. Both the panel of expression experiments and the computational analysis of these data were thus focused on identifying only those genes that are direct targets of the UPR. The validity of this approach is strongly supported by the nature of the UPR targets identiﬁed. Nonetheless, as a general strategy for characterizing a cellular response to a physiologic condition, such a directed approach has two important potential pitfalls. First, it is inherently laborious. For each physiological condition, a variety of different cell strains, time points, and drug treatments must be examined, and these different treatments have to be tailored to the speciﬁc problem at hand. Analysis of the UPR in S. cerevisiae is almost ideally suited for such studies: The mechanism of the UPR is very well characterized, and the response is not essential, thereby making it possible to generate control strains that are unable to mount an unfolded protein response but are otherwise healthy. In general these sorts of manipulations will be more difﬁcult in higher eukaryotes, requiring, for example, either the generation of transgenic deletion animals or the expression of dominant negative inhibitors. A second and perhaps more fundamental drawback to such directed approaches is that they are based on a preconceived notion about the underlying biology of a process. Even in the case of a response to ER folding stress it is clear that a number of genes are induced in a manner that does not depend on the canonical UPR (i.e., genes that are induced in ire1 and hac1 strain) (Travers et al., 2000). A priori, there is no reason to exclude the possibility that such genes are not as important as the direct UPR targets inaiding ER folding. In less well-understood or more complex responses the potential to be misled by such biases is clearly greater. The above limitations have motivated the development of automated approaches for identifying intrinsic relationships between different genes and physiologic conditions. In the most basic form such “unsupervised clustering” approaches (Kohonen, 1997) take a broad set of expression experiments designed in aggregate to elicit most if not all of

FUNCTIONAL GENOMIC APPROACHES

371

the intrinsic transcriptional responses that an organism has been programmed to have. The data are then analyzed by an objective algorithm that identiﬁes patterns within such data without regard to any preconceived notions about the underlying biology. Only after such intrinsic patterns have been identiﬁed and catalogued is there subjective intervention to try to assign meaning to such patterns. For example, in the case of the ER, we might guess that genes that cluster near known ER chaperones like protein disulﬁde isomerase (PDI) and BiP might be involved with alleviating stress and/or aiding protein folding in the ER. The problem of identifying underlying patterns in complex multidimensional data and assigning meanings to such patterns is profound— one that is at the heart of efforts to create artiﬁcial intelligence. A variety of sophisticated mathematical algorithms have been developed for this general problem (Gordon, 1981; Hartigan, 1975; Jobson, 1992; Kohonen, 1997), several of which are now beginning to be applied to expression data in earnest (Alter et al., 2000; Butte et al., 2000; Eisen et al., 1998; Getz et al., 2000; Holter et al., 2000; Hughes et al., 2000; Tamayo et al., 1999). A detailed analysis of these efforts, let alone of the underlying mathematical approaches, is beyond the scope of this review. Instead we offer an overview of some of the basic methods and results from recent analysis of expression data using unsupervised clustering. B. Hierarchical Clustering The most highly developed and widely applied strategy for clustering expression data involves grouping of expression data into hierarchical clusters (Eisen et al., 1998; Hughes et al., 2000). This approach is related to the methods used to generate phylogenetic trees or to analyze relationships among DNA or protein sequences. In this case “similarity” is judged not on the basis of evolutionary relationship or sequence similarity, but rather on how closely correlated the expression patterns of two genes or experimental conditions are. For the expression data analysis, one starts by constructing an N × M matrix in which each of the N rows represents a different gene and each of the M columns represents a different expression experiment. For example, the value of position (2,3) in the matrix represents the expression level of the second gene under the third experimental condition. Initially, the order of the rows and the columns is arbitrary. They are then sorted so that rows (columns) that are most similar to each other (most highly correlated) are placed next to each other. When the clustering algorithm is applied to a matrix in which the expression data are color-coded (green for repressed and red for induced genes), the relationship between genes becomes apparent on

372

KEVIN J. TRAVERS ET AL.

visual inspection (Figs. 6 and 7; see color insert). Thus a major advantage of this approach is that the combination of statistical analysis and visual display makes it possible to intuitively analyze massive amounts of data. Hierarchical clustering of expression data has had a number of impressive successes. For example, in an analysis of a large yeast data set encompassing a broad range of conditions there was a striking clustering of 126 genes that were strongly downregulated in response to a variety of stresses. Of these genes, nearly 90 percent (112) encoded ribosomal components (Eisen et al., 1998). Similarly, components of other large protein complexes were tightly clustered, including the proteasome, the minichromosome mantainance complex, and the F1F0 ATPase complex. In addition, related proteins not in a single complex are regulated, including glycolytic enzymes, genes of the tricarboxylic acid cycle, and oxidative phosphorylation (Eisen et al., 1998). More recently, a group from Rosetta Inpharmatics published a hierarchical cluster analysis of a large data set of expression experiments from 300 diverse mutations and chemical treatments (Hughes et al., 2000). These studies focused in large part on the clustering of expression conditions rather than genes. By looking at expression data from deletion strains of unknown genes and examining which known genes they clustered near, it was possible to infer the function of the gene. For example, deletion of a previously uncharacterized ORF (YER044c/ERG28) results in an expression pattern that clusters closely to that of deletion of known components of the ergosterol biosythesis pathway. Subsequent analysis revealed that Erg28p is indeed a conserved component of the sterol FIG. 6. Example of hierarchical clustering of genes on the basis of expression proﬁles. Changes in levels of transcription were measured in yeast cells grown under a variety of cell cycle arrest conditions and cellular stress conditions. The fold changes from these experiments are indicated in the left panel, in which reductions in mRNA levels are indicated with a green spot and increases, with a red spot. The data were analyzed by hierarchical clustering, resulting in the arrangement of genes indicated in the middle panel. Finally, the rows and columns were randomized prior to clustering (right panel), indicating that the pattern observed in the middle panel is not due to an artifact of the clustering algorithm. (Modiﬁed from Eisen et al., 1998.) FIG. 7. Example of hierarchical clustering of genes and experimental conditions on the basis of expression proﬁles. After collecting expression data from a panel of 276 deletion strains, 11 strains with repressible alleles of essential genes, and 13 drug inductions, genes that changed more than threefold in at least two experiments and experiments with at least two genes showing at least a threefold change were selected, resulting in a set of 568 genes with measurements from 127 conditions. In addition to hierarchical clustering of the genes (clustered transcript response index), experiments were hierarchically clustered based on the transcriptional proﬁles generated in each experiment (clustered proﬁle index). Large clusters are labeled on both axes. (For further details, see Hughes et al., 2000, from which this ﬁgure was reproduced.)

FUNCTIONAL GENOMIC APPROACHES

373

biosynthetic pathway. The Rosetta group was also able to use clustering to identify unknown drug targets. Treatment of yeast with the common topical anesthetic dicyclonine results in an expression pattern that most closely resembled proﬁles resulting from perturbation of the ergosterol pathway, and indeed, detailed analysis argued strongly that dicyclonine inhibits Erg2p. A potential limitation of the hierarchical clustering approach is that each gene is assigned to a single cluster. For genes invoked in only a single response this is adequate. But many genes have multiple functions or are controlled by more than one regulatory pathway, and this is difﬁcult to capture in a single hierarchical tree. In such cases genuine relationships might be missed, and the order of the hierarchical tree will tend to depend strongly on the speciﬁc expression experiments being analyzed. For example, the ER Hsp70 homolog BiP (Kar2p) is involved in protein folding (Bukau and Horwich, 1998), ER degradation, protein translocation (Vogel et al., 1990), and karyogamy (Rose et al., 1989). BiP might be expected to cluster near any of these classes of gene, and exactly where it clusters is likely to be highly dependent on the set of conditions used to generate the array of expression data. A related limitation is that strict phylogenetic trees are most appropriate for situations in which the FIG. 8. Distribution of UPR target genes in the clustered data from Figure 6. All characterized ORFs were hierarchically clustered on the basis of 80 experiments that examined transcriptional changes during the cell cycle and during a variety of treatments designed to elicit stress responses (Eisen et al., 1998). This clustering generated a linear ordering of all genes. From this list, only the characterized ORFs are shown. Every gene in the list was examined for the number of UPR targets within a window of 100 genes on either side. A plot of this count (upper panel) as a function of position in the ordered list illustrates the tendency of UPR targets to cluster near each other under this ordering. The positions of a number of strongly induced UPR targets are indicated (black), as are the positions of a number of non-UPR targets (red) that cluster very near UPR targets. The bottom panel demonstrates the result of a random ordering of data. The list of genes was randomly ordered and the above analysis performed a total of ﬁve times. The results were then averaged to generate this plot. FIG. 9. Distribution of UPR target genes in the clustered data from Figure 7. A list of all ORFs was hierarchically clustered on the basis of 300 experiments that examined transcriptional changes in strains bearing deletion of a particular gene (276 experiments), carrying a tetracycline-repressible allele of an essential gene (11 experiments), or strains treated with a well-characterized compound (13 experiments) (Hughes et al., 2000). As above, a linear ordering of genes was generated from the clustering, and every gene in the list was then examined for the number of UPR targets within a window of 100 genes on either side. The positions of a number of strongly induced UPR targets are indicated (black), as are the positions of a number of non-UPR targets (red) that cluster very near UPR targets. As in Figure 8, the bottom panel demonstrates the result of a random ordering of data in which the list of genes was randomized and the above analysis performed. The plot shown here is the average of ﬁve such calculations.

374

KEVIN J. TRAVERS ET AL.

data result from a true hierarchical descent (such as in organismal evolution) and are not designed to capture the multiple distinct ways in which expression patterns might be similar (Tamayo et al., 1999). For example, DTT treatment will cause misfolding of ER proteins as well as slow growth, but this does not mean that the UPR and cell cycle–regulated genes are in any functionally meaningful sense related to each other. Recently, Getz and colleagues (2000) have described a strategy termed “coupled two-way clustering” that can aid the analysis of gene microarray data where the contributions of a variety of biological mechanisms to the gene expression levels are entangled in a large body of experimental data. This approach starts by identifying relevant subsets of the data. By focusing on these subsets, it is possible to recognize partitions and correlations that are masked when the full dataset is examined. C. Singular Value Decomposition A second approach that helps overcome these limitations of hierarchical clustering is a technique termed singular value decomposition (SVD), also known in statistics as principal-component analysis (Alter et al., 2000; Holter et al., 2000). SVD identiﬁes a set of “characteristic modes” which are “somewhat analogous to the characteristic vibration modes of a tuned violin string” (Holter et al., 2000). Just as the sound of the instrument can be speciﬁed by the contribution of the individual vibrational modes, the complex expression patterns are the result of a limited number of underlying responses. As with hierarchical clustering, the data are initially arranged in a matrix with columns representing expression array experiments and rows corresponding to individual genes. Instead of simply rearranging the order of the columns and the rows, a transformation of the data is applied that creates new “eigengenes” and “eigenarrays” that are linear combinations of the original genes and conditions, respectively. Extending the analogy to musical notes on a string, plucking a string will typically excite a mixture of vibrational modes, with the composition being determined by where and how hard the string is struck. However, the right linear combination of concerted pushes along the length of the string would result in a single pure note. Similarly, an eigengene composed of a linear combination of actual genes excites only one of the primary modes of the expression data, with the strength of the mode corresponding to its relative signiﬁcance. Recently, Alter and coworkers and Holter and coworkers independently applied SVD analysis to previously described data on the yeast cell cycle (Alter et al., 2000; Holter et al., 2000). In those experiments mRNA levels as a function of cell cycle were determined by synchronizing

FUNCTIONAL GENOMIC APPROACHES

375

S. cerevisiae in various stages of the cell cycle, releasing the block and collecting expression data as a function of time as the cell cycle progresses (Spellman et al., 1998). These analyses revealed two primary characteristic modes, both of which vary sinusoidally with time but are out of phase with each other. Similarly, the main features of the sporulation upregulated genes identiﬁed by Chu (Chu et al., 1998) and the response of human ﬁbroblasts to serum (Iyer et al., 1999) could be captured by the two dominant characteristic modes. Thus, complex gene expression patterns from widely disparate organisms and responses can be reduced to a small number of characteristic modes. All of the above experiments involved relatively simple time courses. It remains to be seen whether more complex data sets involving multiple different stimuli will also be amenable to SVD analysis. D. Self-Organizing Maps Recently, Tamayo and colleagues have applied a rather different approach to clustering expression data, termed self-organizing maps (SOM) (Tamayo et al., 1999). Instead of arranging the data as a twodimensional matrix, here the quantitative expression data from N genes and M expression experiments deﬁnes N points in an M-dimensional space. Sets of genes that respond similarly to the various expression conditions will thus tend to be mapped near each other, thereby forming clusters of related genes. Importantly, here proximity is measured as a distance in Euclidean space rather than a correlation coefﬁcient as was done in hierarchical clustering. In a low-dimensional space (3 or less), identifying such clusters could be accomplished by simple visual inspection, but when hundreds of experiments are involved, ﬁnding clusters becomes far more challenging. SOM analysis is a strategy particularly well suited for identifying features in complex, multidimensional data. In this approach, one chooses a geometry of “nodes,” such as a 3 × 2 grid, that are initially placed in the M-dimensional space without regard to the expression data points (in general, both the number and the geometry of the initial nodes are important input parameters to this analysis, which should be explored empirically). The nodes are then iteratively adjusted by randomly selecting a data point and moving each node closer to the data point in such a way that the closest node is moved the most and further nodes are moved progressively less. After many iterations, the nodes move in proximity to gene clusters, with neighboring nodes deﬁning related clusters. Tamayo and colleagues provide a more complete description of the SOM strategy, as well as a software package for applying and displaying SOMs to expression data (Tamayo et al., 1999).

376

KEVIN J. TRAVERS ET AL.

As a test case, SOMs were applied to the yeast cell cycle expression data of Cho and colleagues using a 6 × 5 rectangular grid, yielding a total of 30 nodes (Cho et al., 1998). The SOM analysis readily identiﬁed the major features of the cell cycle data. For example, 91 of the 105 previously identiﬁed late G1-peaking genes were contained within three neighboring nodes, whereas distal nodes corresponded to G1, S, G2, and M phases. SOM analysis was also applied to expression data during the course of hematopoietic differentiation in four different well-characterized cell lines. In a single set of experiments, the SOM analysis found the predominant patterns of gene regulation during macrophage differentiation that had previously been obtained by years of detailed studies, and it identiﬁed several new genes involved in macrophage differentiation. In addition, the SOM analysis highlighted a set of genes and pathways associated with “differentiation therapy” in which retinoic acid is used to induce neutrophil maturation in promyelocytic leukemia. In summary, there are now a number of increasingly sophisticated mathematical strategies for identifying patterns within complex data and new approaches are being developed and tested at a rapid pace. Rather than being mutually exclusive, the information from these different analyses will likely be complementary. For example a new algorithm, termed “relevance networks,” recently described by Butte and colleagues (2000) makes it possible to readily combine expression data with other large-scale quantitative data such as drug sensitivity. It is clear that the clustering approaches are able to extract meaningful data from generic data sets. In the near future large public data sets of clustered expression data will take their place next to sequence and structure databases as invaluable tools for providing functional insights. VI. UPR AS A CASE STUDY IN THE COMPARISON OF SUPERVISED VERSUS UNSUPERVISED SEARCHES Owing to the growing interest in collecting expression data, large quantities of data have now been published and can be readily accessed through the World Wide Web. The previous section described a variety of algorithms, for which convenient software implementations now exist, that facilitate the identiﬁcation of biologically meaningful patterns contained in such data. In the present section, we extend this discussion, using the UPR as a test case, to examine the ability of such algorithms to identify the targets of a speciﬁc transcriptional response from a broad collection of expression data obtained using a large number of disparate experimental conditions. In particular, we will look at the ability of hierarchical clustering algorithms to identify bona fide UPR targets and

FUNCTIONAL GENOMIC APPROACHES

377

to discriminate such targets from functionally related non-UPR targets when applied to two previously published collections of yeast expression experiments. A. Distribution of UPR Targets within a Clustered Data Set Focused on Cell Cycle Changes and Stress Treatments The ﬁrst data set to be examined is one analyzed by Eisen and coworkers; it consists of 80 expression experiments designed to examine a spectrum of unrelated aspects of yeast biology (Eisen et al., 1998). These experiments include time points following release from several different forms of cell cycle arrest and time points from a variety of stress conditions, such as heat shock, cold shock, and DTT treatment. Because these data contain a variety of different stress conditions including DTT treatment, which causes a speciﬁc defect in ER folding, this data set provides a good test case for examining the ability of hierarchical clustering to distinguish UPR genes from generic stress response targets. To test this, we have compared the previously described list of genes [which includes only the characterized genes as of the time of publication (Eisen et al., 1998)] ordered on the basis of a hierarchical clustering of expression experiments to our own list of UPR targets (Fig. 8, see color insert). The metric of comparison was simply the density of UPR targets in that list calculated as a count of UPR targets within 200 genes on either side of each gene in the list. The density of UPR targets is then plotted as a function of position. If the hierarchical clustering perfectly segregated the UPR targets from non-UPR targets, then the plot would contain a single peak, whereas a complete failure to identify UPR targets would result in a random distribution, which is illustrated in the second panel of the ﬁgure. The height of each peak in this plot relative to the average height of the random distribution indicates the relative probability of ﬁnding a UPR target clustered with other such targets. Examination of the data of Eisen and colleagues plotted in this manner reveals that most UPR targets are contained within a limited number of peaks, demonstrating that hierarchical clustering does in fact create a reasonable grouping of genes with respect to this stress response. Indeed, there are a number of remarkable successes; for example, the known ERAD components HRD1 and HRD3 (Hampton et al., 1996) cluster in close proximity with PDI1 (Mizunaga et al., 1990) and ERO1 (Pollard et al., 1998; Frand and Kaiser, 1998), two genes involved in promoting the folding of disulﬁde-containing proteins. However, one of the clear drawbacks to this clustering method as a strategy for identifying UPR targets is that there are numerous instances of false positives

378

KEVIN J. TRAVERS ET AL.

in a list of UPR genes. For example, we ﬁnd that PDI1, ERO1, and KAR2, all previously characterized as ER folding enzymes, cluster very near one another. However, these genes are found in a peak that contains four cytosolic Hsp70 homologs [SSA1-4 (Werner-Washburne et al., 1987)] as well as other cytosolic chaperones such as HSP82 (Nathan et al., 1997) and HSP104 (Glover and Lindquist, 1998) and the mitochondrial chaperones MDJ1 (Rowley et al., 1994) and HSP60 (Cheng et al., 1989). While this clustering is functionally reasonable in that all of these genes are involved in protein folding, it would certainly not be expected that cytosolic or mitochondrial chaperones would be induced in response to ER folding stress, and our own UPR data set conﬁrms this expectation. The data set examined by Eisen and co-workers might be biased toward UPR targets in that conditions known to induce stress responses represent nearly half of these expression experiments. For dissection of the broadest spectrum of transcriptional responses, it would be highly useful to start with a set of data representing expression changes under a broader variety of growth conditions designed ideally to elicit as many responses from an organism as possible. B. Distribution of UPR Targets within a Clustered Data Set Obtained from a Diverse Set of Knockout Strains A data set collected by Rosetta Inpharmatics and published recently (Hughes et al., 2000) provides a useful test case for such analysis. This data set includes 300 different experiments, the majority of which involve growth of a strain deleted for a particular gene, but also includes several drug treatment conditions. Although this data set includes growth in tunicamycin, which would induce the UPR, this represents only 1 of the 300 conditions and would not be expected to greatly bias the clustering toward UPR targets. As with the set of data published by Eisen and colleagues the Rosetta group used a version of hierarchical clustering to order the genes and experiments in their data. Using a list of ∼75% of the genes hierarchically clustered (T. R. Hughes, personal communication), we performed the same analysis as described above, analyzing each gene in the list for its “nearness” to other bona fide UPR targets (Fig. 9, see color insert). As with the data of Eisen and colleagues even using this broader data set, there is an impressive clustering of UPR targets. Indeed, examination of the plot reveals a single major peak that encompasses many of the UPR targets. Again, however, we ﬁnd that peaks containing UPR targets also contain quite a few false positives of functionally related genes that

FUNCTIONAL GENOMIC APPROACHES

379

do not act in the ER and are not UPR targets. For example, the ER chaperones MPD1 (Tachikawa et al., 1995), LHS1 (Craven et al., 1996), and EUG1 (Tachibana and Stevens, 1992) are contained within the single major cluster along with other UPR targets such as the ERAD component DER1 (Knop et al., 1996). Also contained within this peak, however, are the cytosolic chaperones CCT1 (Frydman et al., 1992; Gao et al., 1992; Lewis et al., 1992; Yaffe et al., 1992), SSA1, and SSA3 (Werner-Washburne et al., 1987) as well as the mitochondrial chaperones HSP60 (Cheng et al., 1989) and MDJ1 (Rowley et al., 1994). It is worth noting that the composition of the clusters in these two instances is not the same. For instance, whereas LHS1, DER1, and PDI1 cluster very near one another in the Rosetta Inpharmatics data set, these three genes are separated from one another in the Eisen data set. This supports the notion that the exact outcome of clustering a given data set is highly dependent on the nature of the experiments from which the data set is composed. In summary, the above analysis clearly illustrates the utility of hierarchical clustering in identifying features of a complex response such as the UPR, even when analyzing data sets encompassing a broad spectrum of experimental conditions. In addition to ER chaperones, these analyses also grouped together different classes of functionally unrelated targets of the UPR, such as those involved in folding and degradation of misfolded ER proteins. Nonetheless, clustering alone cannot either deﬁnitively identify UPR targets or reveal the full spectrum of activities induced by the UPR. First, the various UPR targets are distributed over several distant peaks. Moreover, the content of these subclusters differs depending on the data set being analyzed. By analyzing the experiments in these data sets that cause separation of subclusters of UPR targets, it might be possible to gain some insight into the different regulatory processes that separate groups of targets. Second, even within clusters highly enriched for UPR targets, there are many non-UPR targets. In part, this results from functionally related genes such as cytosolic or mitochondrial chaperones that act outside of the UPR. While such groupings might confound efforts to speciﬁcally identify UPR targets, they could provide clues to the function of unidentiﬁed genes. Development and application of further computational approaches together with larger experimental data sets should make it possible to more speciﬁcally identify UPR targets while also revealing these other relationships. In any case, it is clear that with the present tools, data clustering provide valuable insights into even as complex and multifaceted a response as the UPR and that such data complement rather than supplant information from more targeted approaches.

380

KEVIN J. TRAVERS ET AL.

ACKNOWLEDGMENTS The authors thank Drs. M. Eisen, T. Hughes, and J. Trent for allowing us to reproduce ﬁgures from their works, and Dr. T. Hughes for kindly supplying data for our analysis. C.P. is supported by a fellowship from the Howard Hughes Medical Institute. This work was supported with grants from the Searle Scholar Program and the David and Lucille Packard Foundation to J.S.W. J.S.W. is an investigator of the Howard Hughes Medical Institute.

REFERENCES Abola, E., Kuhn, P., Earnest, T., and Stevens, R. C. (2000). Automation of X-ray crystallography. Nat. Struct. Biol. 7(suppl.) 973–977. Alter, O., Brown, P. O., and Botstein, D. (2000). Singular value decomposition for genomewide expression data processing and modeling. Proc. Natl. Acad. Sci. USA 97, 10101– 10106. Barlowe, C., and Schekman, R. (1993). SEC12 encodes a guanine-nucleotide-exchange factor essential for transport vesicle budding from the ER. Nature 365, 347–349. Barnes, C., Hansen, W. J., Holcomb, C. L., and Rine, J. (1984). Asparagine-linked glycosylation in Saccharomyces cerevisiae: Genetic analysis of an early step. Mol. Cell Biol. 4, 2381–2388. Baumeister, W., Walz, J., Zuhl, F., and Seemuller, E. (1998). The proteasome: Paradigm of a self-compartmentalizing protease. Cell 92, 367–380. Belden, W. J., and Barlowe, C. (1996). Erv25p, a component of COPII-coated vesicles, forms a complex with Emp24p that is required for efﬁcient endoplasmic reticulum to Golgi transport. J. Biol. Chem. 271, 26939–26946. Benachour, A., Sipos, G., Flury, I., Reggiori, F., Canivenc-Gansel, E., Vionnet, C., Conzelmann, A., and Benghezal, M. (1999). Deletion of GPI7, a yeast gene required for addition of a side chain to the glycosylphosphatidylinositol (GPI) core structure, affects GPI protein transport, remodeling, and cell wall integrity. J. Biol. Chem. 274, 15251–15261. Bernstein, M., Kepes, F., and Schekman, R. (1989). Sec59 encodes a membrane protein required for core glycosylation in Saccharomyces cerevisiae. Mol. Cell Biol. 9, 1191–1199. Blumberg, H., and Silver, P. A. (1991). A homologue of the bacterial heat-shock gene DnaJ that alters protein sorting in yeast. Nature 349, 627–630. Bochtler, M., Ditzel, L., Groll, M., Hartmann, C., and Huber, R. (1999). The proteasome. Annu. Rev. Biophys. Biomol. Struct. 28, 295–317. Brodsky, J. L., and McCracken, A. A. (1999). ER protein quality control and proteasomemediated protein degradation. Sem. Cell Dev. Biol. 10, 507–513. Buchner, J. (1999). Hsp90 & Co.—a holding for folding. Trends Biochem. Sci. 24, 136–141. Bugl, H., Fauman, E. B., Staker, B. L., Zheng, F., Kushner, S. R., Saper, M. A., Bardwell, J. C., and Jakob, U. (2000). RNA methylation under heat shock control. Mol. Cell 6, 349–360. Bukau, B., and Horwich, A. L. (1998). The Hsp70 and Hsp60 chaperone machines. Cell 92, 351–366. Burley, S. K. (2000). An overview of structural genomics. Nat. Struct. Biol. 7(suppl.) 932– 934. Butte, A. J., Tamayo, P., Slonim, D., Golub, T. R., and Kohane, I. S. (2000). Discovering functional relationships between RNA expression and chemotherapeutic susceptibility using relevance networks. Proc. Natl. Acad. Sci. USA 97, 12182–12186.

FUNCTIONAL GENOMIC APPROACHES

381

Camirand, A., Heysen, A., Grondin, B., and Herscovics, A. (1991). Glycoprotein biosynthesis in Saccharomyces cerevisiae. Isolation and characterization of the gene encoding a speciﬁc processing alpha-mannosidase. J. Biol. Chem. 266, 15120–15127. Casagrande, R., Stern, P., Diehn, M., Shamu, C., Osario, M., Zuniga, M., Brown, P. O., and Ploegh, H. (2000). Degradation of proteins from the ER of S. cerevisiae requires an intact unfolded protein response pathway. Mol. Cell 5, 729–735. Chapman, R., Sidrauski, C., and Walter, P. (1998). Intracellular signaling from the endoplasmic reticulum to the nucleus. Annu. Rev. Cell Dev. Biol. 14, 459–485. Chee, M., Yang, R., Hubbell, E., Berno, A., Huang, X. C., Stern, D., Winkler, J., Lockhart, D. J., Morris, M. S., and Fodor, S. P. (1996). Accessing genetic information with highdensity DNA arrays. Science 274, 610–614. Chen, S., Roseman, A. M., Hunter, A. S., Wood, S. P., Burston, S. G., Ranson, N. A., Clarke, A. R., and Saibil, H. R. (1994). Location of a folding protein and shape changes in GroEL–GroES complexes imaged by cryo-electron microscopy. Nature 371, 261– 264. Cheng, M. Y., Hartl, F. U., Martin, J., Pollock, R. A., Kalousek, F., Neupert, W., Hallberg, E. M., Hallberg, R. L., and Horwich, A. L. (1989). Mitochondrial heat-shock protein hsp60 is essential for assembly of proteins imported into yeast mitochondria. Nature 337, 620–625. Chernoff, Y. O., Lindquist, S. L., Ono, B., Inge-Vechtomov, S. G., and Liebman, S. W. (1995). Role of the chaperone protein Hsp104 in propagation of the yeast prion-like factor [psi+]. Science 268, 880–884. Cho, R. J., Campbell, M. J., Winzeler, E. A., Steinmetz, L., Conway, A., Wodicka, L., Wolfsberg, T. G., Gabrielian, A. E., Landsman, D., Lockhart, D. J., and Davis, R. W. (1998). A genome-wide transcriptional analysis of the mitotic cell cycle. Mol. Cell 2, 65–73. Chu, S., DeRisi, J., Eisen, M., Mulholland, J., Botstein, D., Brown, P. O., and Herskowitz, I. (1998). The transcriptional program of sporulation in budding yeast. Science 282, 699–705. Chuang, S. E., and Blattner, F. R. (1993). Characterization of twenty-six new heat shock genes of Escherichia coli. J. Bacteriol. 175, 5242–5252. Cosson, P., Demolliere, C., Hennecke, S., Duden, R., and Letourneur, F. (1996). Deltaand zeta-COP, two coatomer subunits homologous to clathrin-associated proteins, are involved in ER retrieval. EMBO J. 15, 1792–1798. Cowles, C. R., Odorizzi, G., Payne, G. S., and Emr, S. D. (1997). The AP-3 adaptor complex is essential for cargo-selective transport to the yeast vacuole. Cell 91, 109–118. Cox, J. S., Shamu, C. E., and Walter, P. (1993). Transcriptional induction of genes encoding endoplasmic reticulum resident proteins requires a transmembrane protein kinase. Cell 73, 1197–1206. Cox, J. S., and Walter, P. (1996). A novel mechanism for regulating activity of a transcription factor that controls the unfolded protein response. Cell 87, 391–404. Craven, R. A., Egerton, M., and Stirling, C. J. (1996). A novel Hsp70 of the yeast ER lumen is required for the efﬁcient translocation of a number of protein precursors. EMBO J. 15, 2640–2650. Derman, A. I., Prinz, W. A., Belin, D., and Beckwith, J. (1993). Mutations that allow disulﬁde bond formation in the cytoplasm of Escherichia coli. Science 262, 1744– 1747. Diamant, S., Ben-Zvi, A. P., Bukau, B., and Goloubinoff, P. (2000). Size-dependent disaggregation of stable protein aggregates by the DnaK chaperone machinery. J. Biol. Chem. 275, 21107–21113.

382

KEVIN J. TRAVERS ET AL.

Ditzel, L., Lowe, J., Stock, D., Stetter, K. O., Huber, H., Huber, R., and Steinbacher, S. (1998). Crystal structure of the thermosome, the archaeal chaperonin and homolog of CCT. Cell 93, 125–138. Doerrler, W. T., and Lehrman, M. A. (1999). Regulation of the dolichol pathway in human ﬁbroblasts by the endoplasmic reticulum unfolded protein response. Proc. Natl. Acad. Sci. USA 96, 13050–13055. Duden, R., Hosobuchi, M., Hamamoto, S., Winey, M., Byers, B., and Schekman, R. (1994). Yeast beta- and beta′ -coat proteins (COP). Two coatomer subunits essential for endoplasmic reticulum-to-Golgi protein trafﬁc. J. Biol. Chem. 269, 24486–24495. Eisen, M. B., Spellman, P. T., Brown, P. O., and Botstein, D. (1998). Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. USA 95, 14863– 14868. Ellgaard, L., Molinari, M., and Helenius, A. (1999). Setting the standards: Quality control in the secretory pathway. Science 286, 1882–1888. Fayet, O., Ziegelhoffer, T., and Georgopoulos, C. (1989). The groES and groEL heat shock gene products of Escherichia coli are essential for bacterial growth at all temperatures. J. Bacteriol. 171, 1379–1385. Feldheim, D., and Schekman, R. (1994). Sec72p contributes to the selective recognition of signal peptides by the secretory polypeptide translocation complex. J. Cell Biol. 126, 935–943. Feldheim, D., Yoshimura, K., Admon, A., and Schekman, R. (1993). Structural and functional characterization of Sec66p, a new subunit of the polypeptide translocation apparatus in the yeast endoplasmic reticulum. Mol. Biol. Cell 4, 931–939. Finke, K., Plath, K., Panzner, S., Prehn, S., Rapoport, T. A., Hartmann, E., and Sommer, T. (1996). A second trimeric complex containing homologs of the Sec61p complex functions in protein transport across the ER membrane of S. cerevisiae. EMBO J. 15, 1482–1494. Frand, A. R., and Kaiser, C. A. (1998). The ERO1 gene of yeast is required for oxidation of protein dithiols in the endoplasmic reticulum. Mol. Cell 1, 161–170. Friedlander, R., Jarosch, E., Urban, J., Volkwein, C., and Sommer, T. (2000). A regulatory link between ER-associated protein degradation and the unfolded-protein response. Nat. Cell Biol. 2, 379–384. Frydman, J., Nimmesgern, E., Erdjument-Bromage, H., Wall, J. S., Tempst, P., and Hartl, F. U. (1992). Function in protein folding of TRiC, a cytosolic ring complex containing TCP-1 and structurally related subunits. EMBO J. 11, 4767–4778. Fujimura-Kamada, K., Nouvet, F. J., and Michaelis, S. (1997). A novel membraneassociated metalloprotease, Ste24p, is required for the ﬁrst step of NH2-terminal processing of the yeast a-factor precursor. J. Cell Biol. 136, 271–285. Galitski, T., Saldanha, A. J., Styles, C. A., Lander, E. S., and Fink, G. R. (1999). Ploidy regulation of gene expression. Science 285, 251–254. Gao, Y., Thomas, J. O., Chow, R. L., Lee, G. H., and Cowan, N. J. (1992). A cytoplasmic chaperonin that catalyzes beta-actin folding. Cell 69, 1043–1050. Gasch, A. P., Spellman, P. T., Kao, C. M., Carmel-Harel, O., Eisen, M. B., Storz, G., Botstein, D., and Brown, P. O. (2000). Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell 11, 4241–4257. Gaynor, E. C., Mondesert, G., Grimme, S. J., Reed, S. I., Orlean, P., and Emr, S. D. (1999). MCD4 encodes a conserved endoplasmic reticulum membrane protein essential for glycosylphosphatidylinositol anchor synthesis in yeast. Mol. Biol. Cell 10, 627–648. Geli, M. I., and Riezman, H. (1996). Role of type I myosins in receptor-mediated endocytosis in yeast. Science 272, 533–535.

FUNCTIONAL GENOMIC APPROACHES

383

Gentzsch, M., and Tanner, W. (1996). The PMT gene family: Protein O-glycosylation in Saccharomyces cerevisiae is vital. EMBO J. 15, 5752–5759. Gething, M. J., and Sambrook, J. (1992). Protein folding in the cell. Nature 355, 33–45. Getz, G., Levine, E., and Domany, E. (2000). Coupled two-way clustering analysis of gene microarray data. Proc. Natl. Acad. Sci. USA 97, 12079–12084. Gimeno, R. E., Espenshade, P., and Kaiser, C. A. (1995). SED4 encodes a yeast endoplasmic reticulum protein that binds Sec16p and participates in vesicle formation. J. Cell Biol. 131, 325–338. Glover, J. R., and Lindquist, S. (1998). Hsp104, Hsp70, and Hsp40: A novel chaperone system that rescues previously aggregated proteins. Cell 94, 73–82. Goffeau, A., Barrell, B. G., Bussey, H., Davis, R. W., Dujon, B., Feldmann, H., Galibert, F., Hoheisel, J. D., Jacq, C., Johnston, M., Louis, E. J., Mewes, H. W., Murakami, Y., Philippsen, P., Tettelin, H., and Oliver, S. G. (1996). Life with 6000 genes. Science 274, 546, 563–567. Goloubinoff, P., Christeller, J., Gatenby, A., and Lorimer, G. (1989a). Reconstitution of active dimeric ribulose bisphosphate carboxylase from an unfolded state depends on two chaperonin proteins and Mg-ATP. Nature 342, 884–889. Goloubinoff, P., Gatenby, A. A., and Lorimer, G. H. (1989b). GroE heat-shock proteins promote assembly of foreign prokaryotic ribulose bisphosphate carboxylase oligomers in Escherichia coli. Nature 337, 44–47. Goloubinoff, P., Mogk, A., Zvi, A. P., Tomoyasu, T., and Bukau, B. (1999). Sequential mechanism of solubilization and refolding of stable protein aggregates by a bichaperone network. Proc. Natl. Acad. Sci. USA 96, 13732–13737. Gonzalez, T. N., Sidrauski, C., D¨orﬂer, S., and Walter, P. (1999). Mechanism of nonspliceosomal mRNA splicing in the unfolded protein response pathway. EMBO J. 18, 3119–3132. Gordon, A. E. (1981). Classiﬁcation Methods for Exploratory Analysis of Multivariate Data. Chapman and Hall, New York. Gottesman, S., Wickner, S., and Maurizi, M. R. (1997). Protein quality control: Triage by chaperones and proteases. Genes Dev. 11, 815–823. Gross, C. A. (1996). In E. coli and Salmonella Cellular and Molecular Biology. (F. C. Neidhardt, ed.), pp. 1382–1399. ASM Press, Washington D.C. Hamburger, D., Egerton, M., and Riezman, H. (1995). Yeast Gaa1p is required for attachment of a completed GPI anchor onto proteins. J. Cell Biol. 129, 629–639. Hampton, R. Y., and Bhakta, H. (1997). Ubiquitin-mediated regulation of 3-hydroxy-3methylglutaryl-CoA reductase. Proc. Natl. Acad. Sci. USA 94, 12944–12948. Hampton, R. Y., Gardner, R. G., and Rine, J. (1996). Role of 26S proteasome and HRD genes in the degradation of 3-hydroxy-3-methylglutaryl-CoA reductase, an integral endoplasmic reticulum membrane protein. Mol. Biol. Cell 7, 2029–2044. Hartigan, J. (1975). Clustering Algorithms Wiley, New York. Helenius, A., Marquardt, T., and Braakman, I. (1992). The endoplasmic reticulum as a protein-folding compartment. Trends Cell Biol. 2, 227–231. Hemmingsen, S. M., Woolford, C., van der Vies, S. M., Tilly, K., Dennis, D. T., Georgopoulos, C. P., Hendrix, R. W., and Ellis, R. J. (1988). Homologous plant and bacterial proteins chaperone oligomeric protein assembly. Nature 333, 330–334. Hiller, M. M., Finger, A., Schweiger, M., and Wolf, D. H. (1996). ER degradation of a misfolded luminal protein by the cytosolic ubiquitin–proteasome pathway. Science 273, 1725–1728. Hohfeld, J., and Hartl, F. U. (1994). Role of the chaperonin cofactor Hsp10 in protein folding and sorting in yeast mitochondria. J. Cell Biol. 126, 305–315.

384

KEVIN J. TRAVERS ET AL.

Holter, N. S., Mitra, M., Maritan, A., Cieplak, M., Banavar, J. R., and Fedoroff, N. V. (2000). Fundamental patterns underlying gene expression proﬁles: Simplicity from complexity. Proc. Natl. Acad. Sci. USA 97, 8409–8414. Hong, E., Davidson, A. R., and Kaiser, C. A. (1996). A pathway for targeting soluble misfolded proteins to the yeast vacuole. J. Cell Biol. 135, 623–633. Horwich, A. L., and Willison, K. R. (1993). Protein folding in the cell: Functions of two families of molecular chaperone, hsp 60 and TF55-TCP1. Phil. Trans. R. Soc. Lond., Ser. B: Biol. Sci. 339, 313–325; discussion 325–326. Hoskins, J. R., Pak, M., Maurizi, M. R., and Wickner, S. (1998). The role of the ClpA chaperone in proteolysis by ClpAP. Proc. Natl. Acad. Sci. USA 95, 12135–12140. Huang, C. F., Buu, L. M., Yu, W. L., and Lee, F. J. (1999). Characterization of a novel ADP-ribosylation factor-like protein (yARL3) in Saccharomyces cerevisiae. J. Biol. Chem. 274, 3819–3827. Hughes, T. R., Marton, M. J., Jones, A. R., Roberts, C. J., Stoughton, R., Armour, C. D., Bennett, H. A., Coffey, E., Dai, H., He, Y. D., Kidd, M. J., King, A. M., Meyer, M. R., Slade, D., Lum, P. Y., Stepaniants, S. B., Shoemaker, D. D., Gachotte, D., Chakraburtty, K., Simon, J., Bard, M., and Friend, S. H. (2000). Functional discovery via a compendium of expression proﬁles. Cell 102, 109–126. Hughes, T. R., Roberts, C. J., Dai, H., Jones, A. R., Meyer, M. R., Slade, D., Burchard, J., Dow, S., Ward, T. R., Kidd, M. J., Friend, S. H., and Marton, M. J. (2000). Widespread aneuploidy revealed by DNA microarray expression proﬁling. Nat. Genet. 25, 333– 337. Hurtley, S. M., and Helenius, A. (1989). Protein oligomerization in the endoplasmic reticulum. Annu. Rev. Cell Biol. 5, 277–307. Iguala, J. C., Johnson, A. L., and Johnston, L. H. (1996). Coordinated regulation of gene expression by the cell cycle transcription factor Swi4 and the protein kinase C MAP kinase pathway for yeast cell integrity. EMBO J. 15, 5001–5013. Iyer, V. R., Eisen, M. B., Ross, D. T., Schuler, G., Moore, T., Lee, J. C. F., Trent, J. M., Staudt, L. M., Hudson, J., Jr., Boguski, M. S., Lashkari, D., Shalon, D., Botstein, D., and Brown, P. O. (1999). The transcriptional program in the response of human ﬁbroblasts to serum. Science 283, 83–87. Jackson, C. L., and K´ep`es, F. (1994). BFR1, a multicopy suppressor of brefeldin A-induced lethality, is implicated in secretion and nuclear segregation in Saccharomyces cerevisiae. Genetics 137, 423–437. Jakob, U., Eser, M., and Bardwell, J. C. (2000). Redox switch of Hsp33 has a novel zincbinding motif. J. Biol. Chem. 275, 38302–38310. Jakob, U., Muse, W., Eser, M., and Bardwell, J. C. (1999). Chaperone activity with a redox switch. Cell 96, 341–352. Jobson, J. (1992). Applied Multivariate Data Analysis: Categorical and Multivariate Methods. Springer, New York. Jungmann, J., Rayner, J. C., and Munro, S. (1999). The Saccharomyces cerevisiae protein Mnn10p/Bed1p is a subunit of a Golgi mannosyltransferase complex. J. Biol. Chem. 274, 6579–6585. Kaiser, C. A., and Schekman, R. (1990). Distinct sets of SEC genes govern transport vesicle formation and fusion early in the secretory pathway. Cell 61, 723–733. Kawahara, T., Yanagi, H., Yura, T., and Mori, K. (1997). Endoplasmic reticulum stressinduced mRNA splicing permits synthesis of transcription factor Hac1p/Ern4p that activates the unfolded protein response. Mol. Biol. Cell 8, 1845–1862. Kawahara, T., Yanagi, H., Yura, T., and Mori, K. (1998). Unconventional splicing of HAC1/ERN4 mRNA required for the unfolded protein response. Sequence-speciﬁc and non-sequential cleavage of the splice sites. J. Biol. Chem. 273, 1802–1807.

FUNCTIONAL GENOMIC APPROACHES

385

Kim, S., Willison, K. R., and Horwich, A. L. (1994). Cystosolic chaperonin subunits have a conserved ATPase domain but diverged polypeptide-binding domains. Trends Biochem. Sci. 19, 543–548. Kim, Y. I., Burton, R. E., Burton, B. M., Sauer, R. T., and Baker, T. A. (2000). Dynamics of substrate denaturation and translocation by the ClpXP degradation machine. Mol. Cell 5, 639–648. Kimura, T., Kitamoto, N., Kito, Y., Iimura, Y., Shirai, T., Komiyama, T., Furuichi, Y., Sakka, K., and Ohmiya, K. (1997). A novel yeast gene, RHK1, is involved in the synthesis of the cell wall receptor for the HM-1 killer toxin that inhibits beta-1,3-glucan synthesis. Mol. Gen. Genet. 254, 139–147. Klumpp, M., Baumeister, W., and Essen, L. O. (1997). Structure of the substrate binding domain of the thermosome, an archaeal group II chaperonin. Cell 91, 263– 270. Knop, M., Finger, A., Braun, T., Hellmuth, K., and Wolf, D. H. (1996). Der1, a novel protein speciﬁcally required for endoplasmic reticulum degradation in yeast. EMBO J. 15, 753–763. Kohonen, T. (1997). Self-Organizing Maps. Springer, New York. Kohrer, K., and Emr, S. D. (1993). The yeast VPS17 gene encodes a membrane-associated protein required for the sorting of soluble vacuolar hydrolases. J. Biol. Chem. 268, 559– 569. Korber, P., Stahl, J. M., Nierhaus, K. H., and Bardwell, J. C. (2000). Hsp15: A ribosomeassociated heat shock protein. EMBO J. 19, 741–748. Korber, P., Zander, T., Herschlag, D., and Bardwell, J. C. (1999). A new heat shock protein that binds nucleic acids. J. Biol. Chem. 274, 249–256. Larriba, G., Basco, R. D., Andaluz, E., and Luna-Arias, J. P. (1993). Yeast exoglucanases. Where redundancy implies necessity. Arch. Med. Res. 24, 293–299. Lau, W. W., Howson, R. W., Malkus, P., Schekman, R., and O’Shea, E. K. (2000). Pho86p, an endoplasmic reticulum (ER) resident protein in Saccharomyces cerevisiae, is required for ER exit of the high-afﬁnity phosphate transporter Pho84p. Proc. Natl. Acad. Sci. USA 97, 1107–1112. Lewis, V. A., Hynes, G. M., Zheng, D., Saibil, H., and Willison, K. (1992). T-Complex polypeptide-1 is a subunit of a heteromeric particle in the eukaryotic cytosol. Nature 358, 249–252. Li, Y., Kane, T., Tipper, C., Spatrick, P., and Jenness, D. D. (1999). Yeast mutants affecting possible quality control of plasma membrane proteins. Mol. Cell Biol. 19, 3588– 3599. Lipshutz, R. J., Fodor, S. P., Gingeras, T. R., and Lockhart, D. J. (1999). High density synthetic oligonucleotide arrays. Nat. Genet. 21, 20–24. Lockhart, D. J., Dong, H., Byrne, M. C., Follettie, M. T., Gallo, M. V., Chee, M. S., Mittmann, M., Wang, C., Kobayashi, M., Horton, H., and Brown, E. L. (1996). Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat. Biotech. 14, 1675– 1680. Louvet, O., Roumanie, O., Barthe, C., Peypouquet, M. F., Schaeffer, J., Doignon, F., and Crouzet, M. (1999). Characterization of the ORF YBR264c in Saccharomyces cerevisiae, which encodes a new yeast Ypt that is degraded by a proteasome-dependent mechanism. Mol. Gen. Genet. 261, 589–600. Lussier, M., Sdicu, A. M., Camirand, A., and Bussey, H. (1996). Functional characterization of the YUR1, KTR1, and KTR2 genes as members of the yeast KRE2/MNT1 mannosyltransferase gene family. J. Biol. Chem. 271, 11001–11008. Lussier, M., White, A. M., Sheraton, J., di Paolo, T., Treadwell, J., Southard, S. B., Horenstein, C. I., Chen-Weiner, J., Ram, A. F., Kapteyn, J. C., Roemer, T. W., Vo, D. H.,

386

KEVIN J. TRAVERS ET AL.

Bondoc, D. C., Hall, J., Zhong, W. W., Sdicu, A. M., Davies, J., Klis, F. M., Robbins, P. W., and Bussey, H. (1997). Large scale identiﬁcation of genes involved in cell surface biosynthesis and architecture in Saccharomyces cerevisiae. Genetics 147, 435– 450. Martin, J., Langer, T., Boteva, R., Schramel, A., Horwich, A. L., and Hartl, F. U. (1991). Chaperonin-mediated protein folding at the surface of groEL through a ‘molten globule’-like intermediate. Nature 352, 36–42. Mayer, M. P., and Bukau, B. (1999). Molecular chaperones: The busy life of Hsp90. Curr. Biol. 9, R322–R325. Mayer, M. P., Prestwich, G. D., Dolence, J. M., Bond, P. D., Wu, H. Y., and Poulter, C. D. (1993). Protein farnesyltransferase: Production in Escherichia coli and immunoafﬁnity puriﬁcation of the heterodimer from Saccharomyces cerevisiae. Gene 132, 41–47. Mayhew, M., da Silva, A. C., Martin, J., Erdjument-Bromage, H., Tempst, P., and Hartl, F. U. (1996). Protein folding in the central cavity of the GroEL–GroES chaperonin complex. Nature 379, 420–426. Mizunaga, T., Katakura, Y., Miura, T., and Maruyama, Y. (1990). Puriﬁcation and characterization of yeast protein disulﬁde isomerase. J. Biochem., (Tokyo) 108, 846–851. Montelione, G. T., Zheng, D., Huang, Y. J., Gunsalus, K. C., and Szyperski, T. (2000). Protein NMR spectroscopy in structural genomics. Nat. Struct. Biol. 7(suppl.), 982– 985. Mori, K., Kawahara, T., Yoshida, H., Yanagi, H., and Yura, T. (1996). Signalling from endoplasmic reticulum to nucleus: Transcription factor with a basic-leucine zipper motif is required for the unfolded protein-response pathway. Genes to Cells 1, 803–817. Mori, K., Ma, W., Gething, M., and Sambrook, J. (1993). A transmembrane protein with a cdc2+/CDC28-related kinase activity is required for signaling from the ER to the nucleus. Cell 74, 743–756. Mullins, C., Meyer, H. A., Hartmann, E., Green, N., and Fang, H. (1996). Structurally related Spc1p and Spc2p of yeast signal peptidase complex are functionally distinct. J. Biol. Chem. 271, 29094–29099. Nathan, D. F., Vos, M. H., and Lindquist, S. (1997). In vivo functions of the Saccharomyces cerevisiae Hsp90 chaperone. Proc. Natl. Acad. Sci. USA 94, 12949–12956. Nehls, S., Snapp, E. L., Cole, N. B., Zaal, K. J., Kenworthy, A. K., Roberts, T. H., Ellenberg, J., Presley, J. F., Siggia, E., and Lippincott-Schwartz, J. (2000). Dynamics and retention of misfolded proteins in native ER membranes. Nat. Cell Biol. 2, 288– 295. Newman, A. P., Shim, J., and Ferro-Novick, S. (1990). BET1, BOS1, and SEC22 are members of a group of interacting yeast genes required for transport from the endoplasmic reticulum to the Golgi complex. Mol. Cell. Biol. 10, 3405–3414. Ng, D. T., Spear, E. D., and Walter, P. (2000). The unfolded protein response regulates multiple aspects of secretory and membrane protein biogenesis and endoplasmic reticulum quality control. J. Cell Biol. 150, 77–88. Ng, F. W., Nguyen, M., Kwan, T., Branton, P. E., Nicholson, D. W., Cromlish, J. A., and Shore, G. C. (1997). p28 Bap31, a Bcl-2/Bcl-XL- and procaspase-8-associated protein in the endoplasmic reticulum. J. Cell Biol. 139, 327–338. Nishikawa, S., and Endo, T. (1997). The yeast JEM1p is a DnaJ-like protein of the endoplasmic reticulum membrane required for nuclear fusion. J. Biol. Chem. 272, 12889– 12892. Orlean, P., Albright, C., and Robbins, P. W. (1988). Cloning and sequencing of the yeast gene for dolichol phosphate mannose synthase, an essential protein. J. Biol. Chem. 263, 17499–17507.

FUNCTIONAL GENOMIC APPROACHES

387

Ostermann, J., Horwich, A. L., Neupert, W., and Hartl, F. U. (1989). Protein folding in mitochondria requires complex formation with hsp60 and ATP hydrolysis. Nature 341, 125–130. Pak, M., Hoskins, J. R., Singh, S. K., Maurizi, M. R., and Wickner, S. (1999). Concurrent chaperone and protease activities of ClpAP and the requirement for the N-terminal ClpA ATP binding site for chaperone activity. J. Biol. Chem. 274, 19316–19322. Panzner, S., Dreier, L., Hartmann, E., Kostka, S., and Rapoport, T. A. (1995). Posttranslational protein transport in yeast reconstituted with a puriﬁed complex of Sec proteins and Kar2p. Cell 81, 561–570. Papa, F. R., and Hochstrasser, M. (1993). The yeast DOA4 gene encodes a deubiquitinating enzyme related to a product of the human tre-2 oncogene. Nature 366, 313–319. Paravicini, G., Horazdovsky, B. F., and Emr, S. D. (1992). Alternative pathways for the sorting of soluble vacuolar proteins in yeast: A vps35 null mutant missorts and secretes only a subset of vacuolar hydrolases. Mol. Biol. Cell 3, 415–427. Parsell, D. A., Kowal, A. S., Singer, M. A., and Lindquist, S. (1994). Protein disaggregation mediated by heat-shock protein Hsp104. Nature 372, 475–478. Parsell, D. A., Sanchez, Y., Stitzel, J. D., and Lindquist, S. (1991). Hsp104 is a highly conserved protein with two essential nucleotide-binding sites. Nature 353, 270–273. Partaledis, J. A., Fleming, M. A., Harding, M. W., and Berlin, V. (1992). Saccharomyces cerevisiae contains a homolog of human FKBP-13, a membrane-associated FK506/ rapamycin binding protein. Yeast 8, 673–680. Pelham, H. R. (1986). Speculations on the functions of the major heat shock and glucoseregulated proteins. Cell 46, 959–961. Peng, R., De Antoni, A., and Gallwitz, D. (2000). Evidence for overlapping and distinct functions in protein transport of coat protein Sec24p family members. J. Biol. Chem. 275, 11521–11528. Pollard, M. G., Travers, K. J., and Weissman, J. S. (1998). Ero1p: A novel and ubiquitous protein with an essential role in oxidative protein folding in the endoplasmic reticulum. Mol. Cell 1, 171–182. ˚ Prinz, W. A., Aslund, F., Holmgren, A., and Beckwith, J. (1997). The role of the thioredoxin and glutaredoxin pathways in reducing protein disulﬁde bonds in the Escherichia coli cytoplasm. J. Biol. Chem. 272, 15661–15667. Richmond, C. S., Glasner, J. D., Mau, R., Jin, H., and Blattner, F. R. (1999). Genome-wide expression proﬁling in Escherichia coli K-12. Nuc. Acids Res. 27, 3821–3835. Roberg, K. J., Crotwell, M., Espenshade, P., Gimeno, R., and Kaiser, C. A. (1999). LST1 is a SEC24 homologue used for selective export of the plasma membrane ATPase from the endoplasmic reticulum. J. Cell Biol. 145, 659–672. Rose, M. D., Misra, L. M., and Vogel, J. P. (1989). KAR2, a karyogamy gene, is the yeast homolog of the mammalian BiP/GRP78 gene. Cell 57, 1211–1221. Rothman, J. E. (1989). Polypeptide chain binding proteins: Catalysts of protein folding and related processes in cells. Cell 59, 591–601. Rowley, N., Prip-Buus, C., Westermann, B., Brown, C., Schwarz, E., Barrell, B., and Neupert, W. (1994). Mdj1p, a novel chaperone of the DnaJ family, is involved in mitochondrial biogenesis and protein folding. Cell 77, 249–259. Runge, K. W., Huffaker, T. C., and Robbins, P. W. (1984). Two yeast mutations in glucosylation steps of the asparagine glycosylation pathway. J. Biol. Chem. 259, 412–417. Sacher, M., Jiang, Y., Barrowman, J., Scarpa, A., Burston, J., Zhang, L., Schieltz, D., Yates, J. R., 3rd, Abeliovich, H., and Ferro-Novick, S. (1998). TRAPP, a highly conserved novel complex on the cis-Golgi that mediates vesicle docking and fusion. EMBO J. 17, 2494–2503.

388

KEVIN J. TRAVERS ET AL.

Salama, N. R., Yeung, T., and Schekman, R. W. (1993). The Sec13p complex and reconstitution of vesicle budding from the ER with puriﬁed cytosolic proteins. EMBO J. 12, 4073–4082. Sanchez, Y., and Lindquist, S. L. (1990). HSP104 required for induced thermotolerance. Science 248, 1112–1115. Sanchez, Y., Taulien, J., Borkovich, K. A., and Lindquist, S. (1992). Hsp104 is required for tolerance to many forms of stress. EMBO J. 11, 2357–2364. Sanders, S. L., Whitﬁeld, K. M., Vogel, J. P., Rose, M. D., and Schekman, R. W. (1992). Sec61p and BiP directly facilitate polypeptide translocation into the ER. Cell 69, 353– 365. Santos, B., and Snyder, M. (2000). Sbe2p and sbe22p, two homologous Golgi proteins involved in yeast cell wall formation. Mol. Biol. Cell 11, 435–452. Sato, M., Sato, K., Nishikawa, S., Hirata, A., Kato, J., and Nakano, A. (1999). The yeast RER2 gene, identiﬁed by endoplasmic reticulum protein localization mutations, encodes cis-prenyltransferase, a key enzyme in dolichol synthesis. Mol. Cell Biol. 19, 471– 483. Schena, M., Shalon, D., Davis, R. W., and Brown, P. O. (1995). Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467– 470. Schena, M., Shalon, D., Heller, R., Chai, A., Brown, P. O., and Davis, R. W. (1996). Parallel human genome analysis: Microarray-based expression monitoring of 1000 genes. Proc. Natl. Acad. Sci. USA 93, 10614–10619. Schrimer, E. C., Glover, J. R., Singer, M. A., and Lindquist, S. (1996). HSP100/Clp proteins: A common mechanism explains diverse functions. Trends Biochem. Sci. 21, 289– 296. Schmidt, M., Buchner, J., Todd, M. J., Lorimer, G. H., and Viitanen, P. V. (1994). On the role of groES in the chaperonin-assisted folding reaction. Three case studies. J. Biol. Chem. 269, 10304–10311. Semenza, J. C., Hardwick, K. G., Dean, N., and Pelham, H. R. (1990). ERD2, a yeast gene required for the receptor-mediated retrieval of luminal ER proteins from the secretory pathway. Cell 61, 1349–1357. Shamu, C. E., and Walter, P. (1996). Oligomerization and phosphorylation of the Ire1p kinase during interacellular signaling from the endoplasmic reticulum to the nucleus. EMBO J. 15, 3028–3039. Sidrauski, C., Cox, J. S., and Walter, P. (1996). tRNA ligase is required for regulated mRNA splicing in the unfolded protein response. Cell 87, 405–413. Sidrauski, C., and Walter, P. (1997). The transmembrane kinase Ire1p is a site-speciﬁc endonuclease that initiates mRNA splicing in the unfolded protein response. Cell 90, 1031–1039. Sigler, P. B., Xu, Z., Rye, H. S., Burston, S. G., Fenton, W. A., and Horwich, A. L. (1998). Structure and function in GroEL-mediated protein folding. Annu. Rev. Biochem. 67, 581–608. Silberstein, S., and Gilmore, R. (1996). Biochemistry, molecular biology, and genetics of the oligosaccharyltransferase. FASEB J. 10, 849–858. Smith, A. M., Archer, J. E., and Solomon, F. (1998). Regulation of tubulin polypeptides and microtubule function: Luv1p [correction of Rki1p] interacts with the beta-tubulin binding protein Rbl2p. Chromosoma 107, 471–478. Smith, D. J., Proudfoot, A., Friedli, L., Klig, L. S., Paravicini, G., and Payton, M. A. (1992). PMI40, an intron-containing gene required for early steps in yeast mannosylation. Mol. Cell. Biol. 12, 2924–2930.

FUNCTIONAL GENOMIC APPROACHES

389

Spellman, P. T., Sherlock, G., Zhang, M. Q., Iyer, V. R., Anders, K., Eisen, M. B., Brown, P. O., Botstein, D., and Futcher, B. (1998). Comprehensive identiﬁcation of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Mol. Biol. Cell 9, 3273–3297. Staker, B. L., Korber, P., Bardwell, J. C., and Saper, M. A. (2000). Structure of Hsp15 reveals a novel RNA-binding motif. EMBO J. 19, 749–757. Suzuki, C., and Shimma, Y. I. (1999). P-Type ATPase spf1 mutants show a novel resistance mechanism for the killer toxin SMKT. Mol. Microbiol. 32, 813–823. Tachibana, C., and Stevens, T. H. (1992). The yeast EUG1 gene encodes an endoplasmic reticulum protein that is functionally related to protein disulﬁde isomerase. Mol. Cell Biol. 12, 4601–4611. Tachikawa, H., Funahashi, W., Takeuchi, Y., Nakanishi, H., Nishihara, R., Katoh, S., Gao, X. D., Mizunaga, T., and Fujimoto, D. (1997). Overproduction of Mpd2p suppresses the lethality of protein disulﬁde isomerase depletion in a CXXC sequence dependent manner. Biochem. Biophys. Res. Commun. 239, 710–714. Tachikawa, H., Takeuchi, Y., Funahashi, W., Miura, T., Gao, X. D., Fujimoto, D., Mizunaga, T., and Onodera, K. (1995). Isolation and characterization of a yeast gene, MPD1, the overexpression of which suppresses inviability caused by protein disulﬁde isomerase depletion. FEBS Lett. 369, 212–216. Tamayo, P., Slonim, D., Mesirov, J., Zhu, Q., Kitareewan, S., Dmitrovsky, E., Lander, E. S., and Golub, T. R. (1999). Interpreting patterns of gene expression with self-organizing maps: Methods and application to hematopoietic differentiation. Proc. Natl. Acad. Sci. USA 96, 2907–2912. TerBush, D. R., and Novick, P. (1995). Sec6, Sec8, and Sec15 are components of a multisubunit complex which localizes to small bud tips in Saccharomyces cerevisiae. J. Cell Biol. 130, 299–312. Thompson, M. W., Singh, S. K., and Maurizi, M. R. (1994). Processive degradation of proteins by the ATP-dependent Clp protease from Escherichia coli. Requirement for the multiple array of active sites in ClpP but not ATP hydrolysis. J. Biol. Chem. 269, 18209–18215. Thornton, J. M., Todd, A. E., Milburn, D., Borkakoti, N., and Orengo, C. A. (2000). From structure to function: Approaches and limitations. Nat. Struct. Biol. 7(suppl.) 991– 994. Travers, K. J., Patil, C. K., Wodicka, L., Lockhart, D. J., Weissman, J. S., and Walter, P. (2000). Functional and genomic analyses reveal an essential coordination between the unfolded protein response and ER-associated degradation. Cell 101, 249–258. Trent, J. D., Nimmesgern, E., Wall, J. S., Hartl, F. U., and Horwich, A. L. (1991). A molecular chaperone from a thermophilic archaebacterium is related to the eukaryotic protein t-complex polypeptide-1. Nature 354, 490–493. Trent, J. D., Osipiuk, J., and Pinkau, T. (1990). Acquired thermotolerance and heat shock in the extremely thermophilic archaebacterium Sulfolobus sp. strain B12. J. Bacteriol. 172, 1478–1484. Trilla, J. A., Dur´an, A., and Roncero, C. (1999). Chs7p, a new protein involved in the control of protein export from the endoplasmic reticulum that is speciﬁcally engaged in the regulation of chitin synthesis in Saccharomyces cerevisiae. J. Cell Biol. 145, 1153– 1163. Tsui, H. C., Feng, G., and Winkler, M. E. (1996). Transcription of the mutL repair, miaA tRNA modiﬁcation, hfq pleiotropic regulator, and hﬂA region protease genes of Escherichia coli K-12 from clustered Esigma32-speciﬁc promoters during heat shock. J. Bacteriol. 178, 5719–5731.

390

KEVIN J. TRAVERS ET AL.

Tyson, J. R., and Stirling, C. J. (2000). LHS1 and SIL1 provide a lumenal function that is essential for protein translocation into the endoplasmic reticulum. EMBO J. 19, 6440–6452. Van Dyk, T. K., Gatenby, A. A., and LaRossa, R. A. (1989). Demonstration by genetic suppression of interaction of GroE products with many proteins. Nature 342, 451– 453. Viitanen, P. V., Lubben, T. H., Reed, J., Goloubinoff, P., O’Keefe, D. P., and Lorimer, G. H. (1990). Chaperonin-facilitated refolding of ribulosebisphosphate carboxylase and ATP hydrolysis by chaperonin 60 (groEL) are K+ dependent. Biochemistry 29, 5665–5671. Vogel, J. P., Misra, L. M., and Rose, M. D. (1990). Loss of BiP/GRP78 function blocks translocation of secretory proteins in yeast. J. Cell Biol. 110, 1885–1895. Watanable, R., Ohishi, K., Maeda, Y., Nakamura, N., and Kinoshita, T. (1999). Mammalian PIG-L and its yeast homologue Gpi12p are N-acetylglucosaminylphosphatidylinositol de-N-acetylases essential in glycosylphosphatidylinositol biosynthesis. Biochem. J. 339, 185–192. Weber-Ban, E. U., Reid, B. G., Miranker, A. D., and Horwich, A. L. (1999). Global unfolding of a substrate protein by the Hsp100 chaperone ClpA. Nature 401, 90–93. Weissman, J. S., Hohl, C. M., Kovalenko, O., Kashi, Y., Chen, S., Braig, K., Saibil, H. R., Fenton, W. A., and Horwich, A. L. (1995). Mechanism of GroEL action: Productive release of polypeptide from a sequestered position under GroES. Cell 83, 577– 587. Weissman, J. S., Rye, H. S., Fenton, W. A., Beechem, J. M., and Horwich, A. L. (1996). Characterization of the active intermediate of a GroEL–GroES-mediated protein folding reaction. Cell 84, 481–490. Werner, E. D., Brodsky, J. L., and McCracken, A. A. (1996). Proteasome-dependent endoplasmic reticulum-associated protein degradation: An unconventional route to a familiar fate. Proc. Natl. Acad. Sci. USA 93, 13797–13801. Werner-Washburne, M., Stone, D. E., and Craig, E. A. (1987). Complex interactions among members of an essential subfamily of hsp70 genes in Saccharomyces cerevisiae. Mol. Cell Biol. 7, 2568–2577. Wickner, S., Gottesman, S., Skowyra, D., Hoskins, J., McKenney, K., and Maurizi, M. R. (1994). A molecular chaperone, ClpA, functions like DnaK and DnaJ. Proc. Natl. Acad. Sci. USA 91, 12218–12222. Wickner, S., Maurizi, M. R., and Gottesman, S. (1999). Posttranslational quality control: Folding, refolding, and degrading proteins. Science 286, 1888–1893. Wiebel, F. F., and Kunau, W. H. (1992). The Pas2 protein essential for peroxisome biogenesis is related to ubiquitin-conjugating enzymes. Nature 359, 73–76. Wiertz, E. J., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J., and Ploegh, H. L. (1996). The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84, 769–779. Wodicka, L., Dong, H., Mittmann, M., and Ho, M. H. (1997). Genome-wide expression monitoring in Saccharomyces cerevisiae. Nat. Biotech. 15, 1359–1367. Xu, Z., Horwich, A. L., and Sigler, P. B. (1997). The crystal structure of the asymmetric GroEL–GroES–(ADP)7 chaperonin complex. Nature 388, 741–750. Yaffe, M. B., Farr, G. W., Miklos, D., Horwich, A. L., Sternlicht, M. L., and Sternlicht, H. (1992). TCP1 complex is a molecular chaperone in tubulin biogenesis. Nature 358, 245–248. Zhou, M., and Schekman, R. (1999). The engagement of Sec61p in the ER dislocation process. Mol. Cell 4, 925–934.

THE YEAST PRION [PSI +]: MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES By TRICIA R. SERIO* and SUSAN L. LINDQUIST†

†

*Department of Molecular Genetics and Cell Biology and The Howard Hughes Medical Institute, The University of Chicago, Chicago, Illinois 60637

I. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. [PSI + ] Phenotype and Inheritance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Sup35: The [PSI + ] Determinant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Molecular Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Reversible Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. De Novo Induction of [PSI + ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. [PSI + ] Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Separating Prion Initiation and Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Sup45 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. [PIN + ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. GdnHCl and Excess Hsp104 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conformational Replication In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Self-Assembly and Conformational Change . . . . . . . . . . . . . . . . . . . . . . . . . B. Propagation of Alternate Conformations . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Functional Consequences of the [PS I + ] State . . . . . . . . . . . . . . . . . . . . . . . . . . A. Known Open-Reading Frame Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Global Physiologic Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

391 392 392 394 395 395 396 400 400 401 401 402 403 405 406 406 407 409 409

I. OVERVIEW The central dogma of molecular biology has traditionally posited that nucleic acids alone function as genetic determinants because these macromolecules can template their own replication. The prion or protein-only hypothesis expands the central dogma to include proteins as etiologic agents for disease transmission (Grifﬁth, 1967; Prusiner 1982) and elements of inheritance for phenotypic traits (Wickner, 1994). Prions are unique proteins that can exist in more than one stable conformation, and at least one of these states can be transmitted to newly synthesized protein as a form of templated replication. Since each physical state is associated with a distinct phenotypic state, the trait becomes heritable. The idea of a protein-based genetic element has profound implications not only for our understanding of human diseases thought to arise from similar self-propagating protein misfolding events but also for an 391 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

392

TRICIA R . SERIO AND SUSAN L. LINDQUIST

organism’s genetic diversity and evolvability. A single prion protein can adopt alternate conformations, and hence functional states, expanding the phenotypic breadth of a single genome. Because prion inheritance is reversible (i.e., epigenetic), associated phenotypes can be sampled by the organism prior to ﬁxation (True and Lindquist, 2000). This chapter focuses on recent advances in our understanding of the molecular processes underlying the inheritance and functional consequences of the Saccharomyces cerevisiae prion [PSI +]. An expanded discussion of the history of [PSI +] including a summary of evidence supporting its classiﬁcation as a yeast prion can be found in Volume 57 of Advances in Protein Chemistry (Serio and Lindquist, 2001). A. [PSI +] Phenotype and Inheritance In 1965, Cox reported the ﬁrst phenotypic and genetic characterization of a modulator of translation termination efﬁciency, which he named [PSI +] (Cox, 1965). In strains classiﬁed as [PSI +], nonsense mutations were suppressed (or read-through) at low efﬁciency, while in [psi −] strains, translation terminated faithfully at all stop codons (Serio and Lindquist, 1999). Remarkably, the magnitude of [PSI +]’s effects on read-through efﬁciency varied not only among yeast strains but also within a single genetic background (Cox, 1965; Derkatch et al., 1997; Zhou et al., 1999). [PSI +] is a dominant trait that is inherited through the cytoplasm (Fink and Conde, 1976; Cox et al., 1980) in a non-Mendelian manner: when [PSI +] and [psi −] haploid strains are crossed, the resulting diploid is [PSI +], and only [PSI +] progeny are produced following meiosis (Cox, 1965). The [PSI +] and [psi −] states are not absolute, however. [PSI +] strains convert to the [psi −] state at rates approximating normal mutation rates (∼10−6), and new [PSI +] elements spontaneously appear in [psi −] strains at a similar frequency (Cox, 1965; Lund and Cox, 1981). Moreover, different strengths of [PSI +] within a single genetic background (i.e., “strains” of [PSI +]) can similarly interconvert (Cox, 1965; Derkatch et al., 1997; Zhou et al., 1999), and the efﬁciency of both mitotic and meiotic transmission is directly dependent on the strength of [PSI +]. Strong [PSI +] is inherited by nearly 100% of progeny, while weak [PSI +] can be transmitted to less than 70% of progeny (All-Robyn and Liebman, 1983). B. Sup35: The [PSI +] Determinant The [PSI +] genetic determinant is the protein product of the nuclearencoded gene SUP35 (Chernoff et al., 1992b; Ter-Avanesyan et al., 1994;

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

393

Wickner, 1994; Chernoff et al., 1995; Derkatch et al., 1996; Patino et al., 1996). Sup35 is a 685-residue essential protein (Gerlach, 1975; Wilson and Culbertson, 1988; Ter-Avanesyan et al., 1989; Kushnirov et al., 1990) that is a component of the translation termination complex, known in other organisms as eukaryotic release factor 3 (eRF3) (Stansﬁeld et al., 1995; Zhouravleva et al., 1995). The primary sequence of Sup35 has been divided into three regions—N (the N-terminal region; aa 1–123), M (the middle region; aa 124–253), and C (the C-terminal region; aa 254–685)—based on amino acid distribution, homology to other proteins, and function in [PSI +] propagation and translation termination (Kikuchi et al., 1988; Kushnirov et al., 1988; Wilson and Culbertson, 1988). C is responsible for the protein’s translation termination activity. It is essential for viability but dispensable for the maintenance of [PSI +] (Kushnirov et al., 1990; Ter-Avanesyan et al., 1993, 1994; Derkatch et al., 1996). M is highly charged (18% glutamic acid, 19% lysine, 5% aspartic acid) and, in the context of an N deletion, is not required for either viability or [PSI +] induction (Kushnirov et al., 1990; Ter-Avanesyan et al., 1993; Derkatch et al., 1996). N is the prion-determining domain of Sup35; residues 1–114 are necessary for [PSI +] propagation (Ter-Avanesyan et al., 1994), but this region is not required for either translation termination or viability (Kushnirov et al., 1990; Ter-Avanesyan et al., 1993). Over 75% of N’s sequence is composed of just four amino acids: glycine (17%), tyrosine (16%), asparagine (16%), and glutamine (28%) (Kikuchi et al., 1988; Kushnirov et al., 1988; Wilson and Culbertson, 1988). N can be further divided into two regions: an N-terminal glutamine-rich sequence (aa 8–24) and a proximal region containing ﬁve imperfect repeats of the nonapeptide QGGYQ(Q)QYNP (Liu and Lindquist, 1999). The region encompassing residues 8–24 of Sup35 impacts multiple aspects of [PSI +] biology (DePace et al., 1998; Santoso et al., 2000). The glutamine residues in this region maintain a speciﬁc polarity necessary for [PSI +] propagation; mutations that change glutamines or asparagines to charged amino acids such as serine and arginine do not support [PSI +], while replacing residues 8–24 with polyglutamine (polyQSup35) does (DePace et al., 1998). The nonglutamine residues in this region are important for the species speciﬁcity of prion propagation (Chernoff et al., 2000; Kushnirov et al., 2000a; Santoso et al., 2000). The homologous glutamine-rich regions of Sup35s from different yeast species all exhibit the ability to function as prions; but in strains expressing more than one form, the prions are propagated independently. Coreplication can occur, however, if the glutamine-rich regions are identical (Santoso et al., 2000), suggesting that the nonglutamine residues contribute to

394

TRICIA R . SERIO AND SUSAN L. LINDQUIST

self-association speciﬁcity. In support of this idea, polyQSup35 coreplicates with wild-type S. cerevisiae Sup35. It is currently unclear, however, if polyQSup35 is capable of similarly interacting with the Sup35s from other species as a promiscuous prion. The integrity of the nonapeptide repeat region (aa 42–94) is also crucial for [PSI +] propagation. A single amino acid substitution in the second repeat G58E disrupts [PSI +] propagation in some strains (Young and Cox, 1971; Doel et al., 1994), as does deletion of even a few of the repeats (Liu and Lindquist, 1999; Parham et al., 2001). More remarkably, expanding the total number of repeats to eight has the opposite effect; the spontaneous rate of [psi −] to [PSI +] conversion increases by three orders of magnitude (Liu and Lindquist, 1999). These observations together indicate that the nonapeptide repeats directly inﬂuence the frequency with which the protein undergoes the conformational conversion necessary to establish the prion state. C. Molecular Model The prion cycle directly inﬂuences both the physical and functional state of the Sup35 protein. In [PSI +] strains, Sup35 has increased resistance to proteolytic digestion and is mostly found in large, sedimentable complexes, while in [psi −] strains, Sup35p remains sensitive to proteolysis and is mostly soluble (Patino et al., 1996; Paushkin et al., 1996). These distinct physical states can be directly visualized using Sup35p fusions to the green ﬂuorescent protein (GFP) (Patino et al., 1996). In [psi −] cells, Sup35–GFP ﬂuorescence is diffusely distributed throughout the cell; however, in [PSI +] cells, ﬂuorescence from the same fusion coalesces into foci, which colocalize with endogenous epitope-tagged Sup35 as detected by immunoﬂuorescence ( J. Liu and S. Lindquist, unpublished observation). Notably, when brieﬂy expressed, Sup35–GFP is present in cytoplasmic complexes as soon as its ﬂuorescence can be detected, indicating that the [PSI +] state is quickly imparted to newly synthesized Sup35p (Patino et al., 1996). These [PSI +]-dependent changes in the physical state of Sup35 provide a molecular basis for both the prion’s phenotype and its pattern of inheritance (Patino et al., 1996; Paushkin et al., 1996). Sequestration of Sup35 into large complexes in the [PSI +] cytoplasm precludes its function in translation termination, leading to an increase in nonsense codon read-through. Indeed, partial loss-of-function mutations in Sup35, while Mendelian, have the same phenotype as [PSI +] (Hawthorne and Leupold, 1974; Inge-Vechtomov and Andrianova, 1975; Chernoff et al., + 1992b). Since Sup35[PS I ] complexes are cytoplasmic and capable of

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

395

imparting their character to newly synthesized protein, the cytoplasmic inheritance of [PSI +] can also be explained by this model. In support of this idea, weaker variants of [PSI +] have a lower percentage of their total Sup35 protein in cytoplasmic complexes and pass [PSI +] to their daughters with lower efﬁciency (Zhou et al., 1999). II. REVERSIBLE CURING For nearly 30 years, one of the most bafﬂing characteristics of [PSI +] was its metastability, and this trait contributed to early arguments for [PSI +] as a yeast prion (Wickner, 1994). [PSI +] disappeared and reappeared spontaneously at a low frequency and reproducibly following crossing in certain genetic backgrounds. While this behavior is indeed unusual for most nucleic acid determinants, some factors, such as transposons and lysogenic viruses, do display similar genetics. [PSI +], however, can be ultimately distinguished from such agents by unique treatments, which inﬂuence its metastability. A. De Novo Induction of [PSI +] [PSI +] was ﬁrst linked to Sup35 with the observation that SUP35containing plasmids induced a [PSI +]-like phenotype in yeast (Chernoff et al., 1992a). It was not the presence of the SUP35 gene per se that was responsible for this effect, but rather, the expression of Sup35 protein (Derkatch et al., 1996). Overexpression of any fragment of Sup35 containing the N region can induce [PSI +] de novo (Derkatch et al., 1996). Most notably, [PSI +] remains stable and heritable once protein levels are returned to wild-type and the extra episomal gene is lost, indicating that a transient increase in Sup35 levels is sufﬁcient (Chernoff et al., 1992a; Derkatch et al., 1996). This phenotypic change is accompanied by a heritable change in the physical state of Sup35. While the precise molecular mechanism remains a mystery, it has been suggested that excess Sup35 has an increased propensity to adopt the [PSI +] state by creating an increased likelihood of Sup35:Sup35 interactions, perhaps by causing an imbalance with partner proteins and/or chaperones (Derkatch et al., 1996; Patino et al., 1996; Derkatch et al., 1998). 1. [PIN +] The reintroduction of [PSI +], either spontaneously or by overexpression of Sup35, is inﬂuenced by another yeast prion, [PIN +] (Lund and Cox, 1981; Derkatch et al., 1997, 2000). Like [PSI +], [PIN +] is inherited in a dominant, non-Mendelian pattern (Derkatch et al., 1997) and

396

TRICIA R . SERIO AND SUSAN L. LINDQUIST

has the hallmarks of a prion, including reversible curing (Derkatch et al., 2000). Although the [PIN +] determinant has not yet been isolated, [PIN +] propagation is independent of [PSI +] (Derkatch et al., 1997; 2000). A series of elegant genetic experiments has also demonstrated that [PIN +] can be propagated in the absence of Sup35 (Derkatch et al., 1997, 2000) and is independent of the Sup35 partner protein, Sup45 (Derkatch et al., 1998). B. [PSI +] Curing Conversion from the [PSI +] to the [psi −] state is known as curing. While this phenomenon occurs spontaneously at a low frequency (Cox, 1965), both chemical and biological factors that increase curing rates have been identiﬁed. New information is emerging on the molecular mechanisms by which these treatments act, and these studies provide new insights into the prion replication process in vivo. 1. Chemical Agents The spontaneous rate of [PSI +] curing is dramatically enhanced in the presence of low concentrations of certain chemical agents such as guanidine-HCl (GdnHCl), methanol, and ethylene glycol (Singh, 1979; Lund and Cox, 1981; Tuite et al., 1981; All-Robyn and Liebman, 1983; Liebman and All-Robyn, 1984; Cox et al., 1988). The fact that these treatments are nonmutagenic to nucleic acids was one of the early lines of evidence that the switch from [PSI +] to [psi −] was not due to a change in nucleic acid content or sequence. The molecular mechanisms of action for these treatments remain largely unknown, but the effects of GdnHCl have been the most extensively studied. GdnHCl treatment induces nearly 100% conversion from [PSI +] to [psi −] (Cox et al., 1988; Tuite et al., 1988) and is similarly effective in eliminating two other yeast prions, [URE3] and [PIN +] (Wickner, 1994; Derkatch et al., 1997). In the case of [PSI +], GdnHCl treatment eliminates [PSI +] from cultures progressively over time and requires continuous cell division (Tuite et al., 1981). If GdnHCl is removed from the media prior to complete curing, a culture will stably maintain a mixture of both [PSI +] and [psi −] cells (Cox, 1993; Eaglestone et al., 1999). These observations suggest that GdnHCl interferes with replication of the [PSI +] physi+ cal state: new Sup35[PS I ] is not produced in the presence of GdnHCl, + but preexisting Sup35[PS I ] is not affected. [psi −] cells appear only + when cell division dilutes out the pool of preexisting Sup35[PS I ] . If GdnHCl is removed before this occurs, newly synthesized Sup35 becomes conversion-competent again, and prion replication resumes in

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

397

+

cells retaining Sup35[PS I ] . While the molecular mechanism of GdnHCl curing is unknown, the heat shock protein (hsp) Hsp104 is sensitive to low concentrations of GdnHCl in vitro (Glover and Lindquist, 1998) and is a potential target. 2. Hsp104 Hsp104 is a highly conserved member of the Clp/Hsp100 family of molecular chaperones (Parsell et al., 1991), a subfamily of the AAA family of proteins that function in a wide array of protein remodeling events in vivo. In yeast, Hsp104 is required both for induced thermotolerance (Sanchez and Lindquist, 1990) and for tolerance to many different types of chemical and growth stresses (Sanchez et al., 1992). Hsp104’s protective activity both in vivo and in vitro is accomplished through a novel mode of action: the rescue of damaged proteins that have already aggregated (Parsell et al., 1994; Glover and Lindquist, 1998). Disaggregating activity is unique among Hsps, which typically act to prevent aggregation; but this function appears to be evolutionarily conserved among members of the Clp/Hsp100 family, highlighting a special niche for these factors in the response to extreme stress (Mogk et al., 1999; Weber-Ban et al., 1999; Zolkiewski, 1999). In addition to its protective role, HSP104 also genetically interacts with [PSI +]. Hsp104 was originally isolated as an extra-copy modiﬁer of [PSI +], heritably and efﬁciently eliminating [PSI +] from yeast strains when present in excess (Chernoff et al., 1995). Even a transient increase in Hsp104 levels is sufﬁcient for complete [PSI +] curing. Surprisingly, elimination of Hsp104 function, either by deletion of the HSP104 gene or mutation of both Walker-type nucleotide binding sites in Hsp104, has the same effect (Chernoff et al., 1995; Patino et al., 1996). Thus, intermediate levels of Hsp104 are required for faithful [PSI +] propagation. Genetic links between HSP104 and two other yeast prions, [URE3] and [PIN +], have also been described; in these cases, disruption of Hsp104, but not overexpression, induces prion loss (Derkatch et al., 1997; Moriyama et al., 2000). Two distinct roles have been proposed to explain the genetic interaction between Hsp104 and [PSI +]. Although both models posit a direct interaction between Hsp104 and Sup35, they differ in the point of action: replication or inheritance. One model suggests that Hsp104 actually promotes the conformational conversion of Sup35 to the [PSI +] state, whether directly or indirectly through the formation of a conformational and/or oligomeric intermediate (Patino et al., 1996; Lindquist and Schirmer, 1999). Following from this, the balance between Sup35 and Hsp104 needed for conversion would be disrupted when Hsp104 is in

398

TRICIA R . SERIO AND SUSAN L. LINDQUIST

excess. The other model proposes that Hsp104’s function is solely + to disaggregate preexisting Sup35[PS I ] . A low level of the protein would therefore enhance partitioning to daughter cells on cell division (Paushkin et al., 1996; Kushnirov and Ter-Avanesyan, 1998). However, with an excess for Hsp104, the rate of disaggregation would exceed that of conversion leading to curing. While these models are not mutually exclusive (curing by Hsp104 deﬁciency and excess may occur by different mechanisms, for example) (Patino et al., 1996), they are experimentally distinguishable. In the ﬁrst model, for example, [PSI +] curing by excess Hsp104 would be depen+ dent on cell division to dilute preexisting SUP35[PS I ] , as is the case for GdnHCl curing; for the second model, excess Hsp104 should cure [PSI +] in nondividing cells. Unfortunately, our current understanding of this aspect of [PSI] biology is limited. In vitro, stable complexes of Hsp104 and its protein substrates have never been isolated from solution, as is the case for the other members of the Clp/Hsp100 family. Several observations, however, indirectly suggest a transient physical interaction. For example, the presence of Sup35 alters the ATPase activity of Hsp104, and the circular dichroism spectra and light scattering intensities of solutions of Sup35 and Hsp104 are distinct from the arithmetic sum of these parameters from solutions of each protein alone (Schirmer and Lindquist, 1997). In addition, Sup35 aggregates characteristic of [PSI +] induction cannot be detected in Hsp104-deﬁcient cells, suggesting an Hsp104 requirement for conversion (Glover et al., 1997; Zhou et al., 2001). Finally, nonsectored [psi −] colonies arise when Hsp104 levels are elevated, suggesting that curing by this means is an early event (Chernoff et al., 1995). Technical difﬁculties have hampered a rigorous analysis of the molecular mechanism of Hsp104 curing. For example, [PSI +] status is monitored by reversion to prototrophy, a colony-based phenotype; thus, the read-out is temporally distant from the experimental manipulations that effect it. For assessing [PSI +] curing following Hsp104 loss, new transcription of HSP104 must be preferentially blocked and preexisting Hsp104 must be degraded. Unfortunately, a suitable tightly regulated promoter has not been described, and the half-life of Hsp104 far exceeds the generation time of yeast (T. R. Serio and S. L. Lindquist, unpublished observation). In the case of [PSI +] curing by excess Hsp104, the issue of a tightly regulated promoter resurfaces; leaky expression of Hsp104 is sufﬁcient to cure [PSI +] before a controlled experiment can be initiated (T. R. Serio and S. L. Lindquist, unpublished observation). The GAL1 promoter is tightly regulated and inducible, but the expression switch is a carbon-source shift, which greatly alters the chaperone complement

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

399

of cells and complicates interpretation of curing effects (see below). In + addition, analyses of Hsp104 effects on preexisting SUP35[PS I ] requires dual regulation of two different genes (i.e., SUP35 and HSP104 ), posing yet another technical difﬁculty. Thus, a true molecular understanding of the genetic interaction between Hsp104 and [PSI +] awaits the necessary technical advances. 3. Combined Chaperone Effects Deconvoluting the combined effects of chaperones on prion propagation in vivo is a complicated task. For example, altering the levels of one chaperone often induces compensatory changes in the expression of other chaperones and profoundly alters the levels of many other proteins in the cell (Werner-Washburne et al., 1987; 1989; Werner-Washburne and Craig 1989). [Hsp104 is unique in this regard; changes in Hsp104 levels do not appreciably alter the synthesis or turnover of any other protein (Parsell et al., 1993, 1994).] Moreover, it is extremely important to consider the normal expression patterns of chaperones in relation to one another. The complement of these proteins present in the cell varies widely under different physiologic conditions, and the misexpression of one factor, while potentially altering [PSI +] inheritance, may not accurately represent a state naturally accessible to cells in the environment or during normal growth and development. Despite these difﬁculties, several experiments demonstrate a clear but complex role for the Hsp70 family in [PSI ] metabolism. Hsp104 functionally interacts with Hsp70 (Ssa1) and Hsp40 (Ydj1) in the rescue of aggregated proteins both in vivo and in vitro (Sanchez et al., 1993; Glover and Lindquist, 1998), and both Ssa and Ssb proteins, two major subfamilies of Hsp70 in yeast, modulate prion regulation by Hsp104. For example, overexpression of Ssa1 interferes with [PSI +] curing by excess Hsp104 (Chernoff et al., 1995; Newman et al., 1999) and strengthens the nonsense suppression phenotype of [PSI +] cells (Newman et al., 1999). Both effects are consistent with the observation that soluble Sup35 levels are decreased in cells overexpression Ssa1 (Newman et al., 1999). Overexpression of other members of the Hsp70 family, Ssb1 and Ssb2, also alters [PSI +] curing by excess Hsp104, but in this case, the effect is stimulatory (Chernoff et al., 1999). The Hsp70 and Hsp40 chaperones may exert an effect on [PSI ] metabolism that is independent of Hsp104, as well. Overexpression of Ssb1, Ssa1, and Ydj1 alone has been reported to cure weak [PSI +] “strains” (Kushnirov et al., 2000b). Unlike [PSI +] curing by excess Hsp104, elevation of these factors leads to inefﬁcient curing that manifests as mixed cultures of [PSI +] and [psi −] cells as well as sectored

400

TRICIA R . SERIO AND SUSAN L. LINDQUIST

colonies. In addition, a mutant form of Ssa1, containing a single amino acid substitution L483W, cures [PSI +] in ssa1ssa2 but not wildtype strains ( Jung et al., 2000). This conundrum can be explained if Ssa1L483W is both a loss-of-function mutant for Ssa1/2 activity and a dominant negative for Ssa3/4, which become elevated in ssa1ssa2 strains ( Jung et al., 2000). Although the relationships between chaperone biology and prion maintenance are complex, they may reﬂect a very important aspect of [PSI +] biology. Changes in cell physiology accompanied by a massive induction of Hsp104, such as heat shock, only modestly cure [PSI +] (∼1–2%), while others, such as meiosis, have no effect (Singh, 1979; Tuite et al., 1981; Sanchez and Lindquist, 1990; Sanchez et al., 1992). Notably, when expressed to the same levels on its own, Hsp104 efﬁciently cures [PSI +] (Chernoff et al., 1995). During heat shock or meiosis, [PSI +] may be protected by concomitant increases in the levels of other chaperones that interfere with Hsp104’s ability to cure [PSI +] when overexpressed. Perhaps these chaperones divert Hsp104 to other targets and/or directly alter Sup35 folding.

III. SEPARATING PRION INITIATION AND PROPAGATION In the simplest model of [PSI] metabolism, de novo conversion from the [psi −] to the [PSI +] state involves two discrete steps: initiation and propagation. During initiation, the ﬁrst replicon (e.g., seed or nucleus) is formed. During propagation, newly synthesized protein is inﬂuenced by the replicon to convert to the [PSI +] state. Since the [PSI +] read-out is a colony-based assay (reversion to prototrophy), these steps are nearly impossible to separate. However, recent studies of [PSI +] interaction with other proteins, chemical agents, and prions have provided our ﬁrst glimpses into the multistep process of prion induction (Fig. 1).

A. Sup45 Sup45 is the yeast homolog of the eukaryotic release factor 1 (Frolova et al., 1994; Stansﬁeld et al., 1995). Presumably, Sup45 interacts directly with Sup35 to form a functional translation termination complex (Paushkin et al., 1997; Eurwilaichitr et al., 1999). When overexpressed, Sup45 blocks [PSI +] induction by excess Sup35 but has no effect on either the phenotype or the propagation of [PSI +] in already converted strains (Derkatch et al., 1998). These observations suggest that Sup45 may inhibit [PSI +] initiation but not propagation. How might this

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

401

FIG. 1. Discrete events in the Sup35 prion cycle. Nascent Sup35 (gray) converts indirectly through an intermediate form (gray, Sup35∗ ) or directly to the prion form + (Sup35[PS I ] ) during the initiation phase (gray arrows). Sup45, [PIN +], and Hsp104 + have been postulated to act at this point. Once acquired, Sup35[PS I ] replicates either through conversion of nascent Sup35 (black) directly or indirectly through an intermediate (black, Sup35∗ ). GdnHCl treatment has been demonstrated to act to block replication, while Hsp104 has been postulated to act here as well.

distinction arise? One possibility is that in [psi −] cells, Sup45 binding can block de novo conversion of Sup35, but in [PSI +] cells, preexisting + SUP35[PS I ] outcompetes Sup45 for interaction with newly synthesized Sup35. B. [PIN +] Induction of [PSI +] by excess Sup35 depends on the presence of the [PIN +] prion (Derkatch et al., 1997). In a recent study, [PSI +] and [PIN +] curing by GdnHCl were separately assessed; although the tendency was for both prions to be either retained or lost together, some [PSI +] [pin−] isolates were found (Derkatch et al., 2000). The inheritance of [PSI +] in the absence of [PIN +] clearly demonstrates that the [PIN +]-dependent component of de novo [PSI +] induction is initiation and not propagation. The molecular mechanism by which [PIN +] confers [PSI +] inducibility to yeast strains remains unclear, but future studies aimed at characterizing this process will undoubtedly provide insight into [PSI +] initiation. C. GdnHCl and Excess Hsp104 As detailed above, GdnHCl treatment appears to inhibit propagation of [PSI +] by blocking newly synthesized Sup35 from converting to the + SUP35[PS I ] state (Cox, 1993; Eaglestone et al., 2000). A recent study has revealed that GdnHCl treatment has no effect on [PSI +] induction

402

TRICIA R . SERIO AND SUSAN L. LINDQUIST

by excess Sup35 (Zhou et al., 2001). Since conversion to [PSI +] was scored in the absence of GdnHCl, it is unclear if induction occurred during the GdnHCl treatment or after the agent was removed. Sup35 has a half-life that exceeds the yeast generation time (T. R. Serio and S. L. Lindquist, unpublished observation), and Sup35 levels would be expected to remain elevated for at least another generation after GdnHCl treatment was suspended. Similarly, excess Hsp104 does not interfere with the de novo induction of [PSI +] by overexpression of Sup35 (Zhou et al., 2001). Again, these results must be cautiously interpreted. While the turnover rates of Sup35 and Hsp104 are similar (T. R. Serio and S. L. Lindquist, unpublished observation), the relative levels of induction were not quantiﬁed, and the result may be explained if Sup35 levels remained elevated longer than those of Hsp104. Further studies are needed to address these issues, but these observations have intriguing implications for a molecular distinction between the physical state alterations that Sup35 must undergo in initiation and propagation. IV. CONFORMATIONAL REPLICATION In Vitro Prion replication for Sup35 has been modeled in vitro as the assembly of ﬁbers from denatured protein (Glover et al., 1997; King et al., 1997; Serio et al., 2000). Fibers formed from Sup35 have all of the characteristics of the amyloid ﬁbers assembled from the Alzheimer’s disease–associated Aβ, lysozyme, and the transmissible spongiform encephalopathy (TSE)–linked PrP (Serpell et al., 1997). For example, all of these ﬁbers bind to the diagnostic dye Congo red, inducing a shift in its absorption spectrum, and exhibit green birefringence when bound to Congo red and viewed under polarized light (Glover et al., 1997; King et al., 1997; A. G. Cashikar and S. L. Lindquist, unpublished observation). By x-ray diffraction, these ﬁbers also display a common set of reﬂections characteristic of the cross β-pleated sheet structure (Sunde et al., 1997; Serio et al., 2000). + Sup35 ﬁbrillization in vitro has been linked to Sup35[PS I ] replication in vivo by several observations. First, the N region is required both for [PSI +] inheritance in vivo and for ﬁbrillization in vitro (Ter-Avanesyan et al., 1994; Glover et al., 1997; King et al., 1997). Mutations in N that disrupt [PSI +] inheritance in vivo block ﬁber formation in vitro (Ter-Avanesyan et al., 1994; Glover et al., 1997; DePace et al., 1998; Liu and Lindquist, 1999), and mutations that enhance the spontaneous appearance of [PSI +] in vivo accelerate the rate of ﬁber formation in vitro (Liu and Lindquist, 1999). Strikingly, lysates from [PSI +] but not [psi −] cells can accelerate the assembly of puriﬁed recombinant Sup35 in vitro

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

403

(Glover et al., 1997). Thus, an understanding of the ﬁbrillization process in vitro will likely have direct implications for the molecular mechanism + of Sup35[PS I ] replication in vivo and perhaps the molecular basis for disease progression in the mammalian amyloidoses. A. Self-Assembly and Conformational Change When the NM fragment of Sup35 (aa 1–253) is diluted from denaturant into physiologic buffers, ﬁbers form slowly, and the kinetics of ﬁbrillization can be divided into two discrete steps: a lag phase, during which no ﬁbers are detected, and an assembly phase, during which the bulk of protein rapidly converts to the ﬁber state (Glover et al., 1997). The transition from the monomeric to the oligomeric state is also accompanied by a change in secondary structure from random coil to β-rich (Glover et al., 1997). The addition of a small quantity of preformed ﬁbers or [PSI +] lysate accelerates not only self-assembly but also conformational change, suggesting that the two transitions are linked processes (Glover et al., 1997). In a systematic study, the relationship between structure acquisition and self-assembly in NM ﬁbrillization reactions was monitored through a series of biochemical and microscopic probes (Fig. 2) (Serio et al., 2000). Under these conditions, the conformational transition could not be temporally separated from self-assembly of NM. That is, the protein adopts the β-rich structure as it assembles into ﬁbers. The concentration dependences of the lag and assembly phases were also assessed in this study as an indirect measure of the relationship between conformational conversion and self-association. The results were consistent with a β-rich structure’s being conferred to protein concomitant with assembly (Serio et al., 2000; DePace et al., 1998). The transition from soluble, unstructured NM to highly ordered ﬁbers is not direct, however. By both transmission electron microscopy and atomic force microscopy, oligomeric complexes of NM were observed after dilution into physiologic buffer, and these complexes were no longer present once the ﬁbrillization reaction had reached completion (Serio et al., 2000). A large intermediate was also detected by light scattering immediately after dilution into buffer, and conditions that either enhanced or diminished the formation of this intermediate had a correlative effect on the overall rate of ﬁbrillization. Notably, formation of this intermediate could not be detected by changes in CD spectrum or protease mapping, suggesting that its formation was not associated with the transition to the ﬁber conformation. Kinetic experiments placed this intermediate on pathway for conformational conversion during the assembly phase.

404

TRICIA R . SERIO AND SUSAN L. LINDQUIST

FIG. 2. Biochemical probes for NM conversion in vitro. The N and M regions of Sup35 have distinct sensitivities to the proteases chymotrypsin (CHY) and endoproteinase Glu-C (V8); CHY sites are exclusively found in N, while V8 sites are limited to M. V8: Anti-Sup35 Western blot of NM alone (− −) or following digestion with V8 (arrow). NM was removed from an assembly reaction over a time course, and each sample was then treated with protease for the same amount of time. CHY: Anti-Sup35 Western blot of NM alone (− −) or following digestion with CHY (arrow) as for V8. SDS: Anti-Sup35 Western blot of NM following incubation in 2% SDS at either 100◦ C or 25◦ C for 10 min. Samples were taken over the same time course as for V8 and CHY.

When freshly diluted protein or protein that had been incubated in physiologic buffer for a period of time was added to a small amount of preformed ﬁbers, preincubated NM assembled at a rate 40-fold faster than freshly diluted NM. These results together indicate that an intermediate on pathway for both nucleation and assembly forms during the lag phase (Serio et al., 2000). These observations were inconsistent with some aspect of each of the models previously proposed for amyloid ﬁbrillogenesis (Serio et al., 2000). Thus, a new model, “nucleated conformational conversion,” was proposed (Serio et al., 2000). According to this model, ﬁbers arise de novo from nuclei formed during the lag phase. These nuclei are structured, stable complexes of NM formed when structurally molten, oligomeric complexes undergo a conformational rearrangement during the lag

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

405

phase. These ﬂuid intermediates are also competent for rapid assembly onto preformed ﬁbers, acquiring the β-rich structure through a templating or induced-ﬁt mechanism at ﬁber ends. Such structurally ﬂuid oligomeric complexes may act to partially restrict the conformational space available to a random coil peptide (Dill, 1999), providing an environment that channels sampling toward a stable, cross-β structure and overcomes kinetic folding barriers. While the disease-associated amyloid proteins of humans and the protein-based genetic elements of yeast are subject to distinct evolutionary pressures (Lansbury, 1999), micelle-like structures have been described for other amyloids and appear to contribute at least to nucleation (Lundberg et al., 1997), and other partially structured intermediate complexes seem to undergo additional rearrangements when they convert to ﬁbers (Harper et al., 1997; Walsh et al., 1997; Chiti et al., 1999; Rochet and Lansbury, 2000). It will be of great interest to determine how evolutionary pressures have honed the nature of the proteins’ assembly processes. In some cases, these may lead to separate pathways of amyloid formation; in others, to similar pathways with different rate-limiting steps. B. Propagation of Alternate Conformations Using a chimeric protein containing the S. cerevisiae glutamine-rich region (aa 1–39), the repeat region from Candida albicans (aa 40–140), and the S. cerevisiae C region (aa 254–685), Chien and Weissman have recently linked prion-associated differences in suppression strength ([PSI +] “strains”) to differences in amyloid assembly in vitro (Chien and Weissman, 2001). In these studies, a range of heritable suppression strengths was induced in cells expressing the chimeric Sup35 by transient overexpression of S. cerevisiae NM. In contrast, only a single strong [PSI +]-like phenotype was obtained when C. albicans NM was transiently induced. In vitro, chimeric NM adopts distinct protease digestion patterns when seeded with preformed ﬁbers from either S. cerevisiae NM or C. albicans NM; and once formed, chimeric NM ﬁbers can seed only the ﬁbrillization of NM from the organism originally used as seed. Thus, heterotypic proteins can induce distinct physical changes in a single protein which can then be propagated in vitro. Currently, it is unclear if this distinction is due to unique conformations or to a difference in accessibility owing to the assembly of ﬁber tangles that arise from limiting points of nucleation. These studies will undoubtedly form the basis for future work correlating in vivo phenotypic variation with conformational distinctions. For example, overexpression of S. cerevisiae NM in vivo produces a range

406

TRICIA R . SERIO AND SUSAN L. LINDQUIST

of [PSI +]-like phenotypes for chimeric NM, but only a single one is produced in vitro. Can the phenotypic diversity in vivo be recapitulated in vitro? The clearest and most direct evidence that NM can adopt distinct self-perpetuating states comes from visualization of ﬁbers by transmission electron microscopy (Glover et al., 1997). Both straight and curly morphologies have been observed; and, remarkably, once the protein adopts one of these states, it is perpetuated throughout the length of the ﬁber. While this observation would seem to explain the existence of strains in vivo, no evidence to date ties the distinct ﬁber morphologies to the in vivo phenomenon. In addition, the in vitro work indicates that chimeric NM’s ability to seed heterotypic NM becomes limited by the identity of the original seed. Does this restriction hold true in vivo if [PSI +] is passed from chimeric NM strains back to [psi −] strains expressing only S. cerevisiae or C. albicans NM? V. FUNCTIONAL CONSEQUENCES OF THE [PS I + ] STATE Since [PSI +] is an informational suppressor, the [PSI ] status of a strain will undoubtedly have phenotypic consequences for the cell, whether direct or indirect (True and Lindquist, 2000). Both speciﬁc targets and uncharacterized phenotypes arising from [PSI +]-dependent changes in physiology have recently been reported and are summarized below.

A. Known Open-Reading Frame Effects [PSI +]’s

ability to suppress nonsense mutations in open-reading frames such as ADE2, ADE1, TRP5, and CYC1 have been known for more than 35 years and have facilitated the genetic characterization of the prion by providing selectable phenotypes to distinguish the alternate physical states of Sup35 (Inge-Vechtomov, 1964; Cox, 1965; Liebman, 1975; Singh, 1979). Undoubtedly, the genetic diversity present in commonly used yeast strains includes distinct complements of genes containing nonsense mutations on which [PSI +] can act and alter the ﬁtness of the strain. One example is a yeast strain containing a nonsense mutation in the gene encoding heat shock factor (HSF ), a protein essential for viability and induction of the heat shock response (Lindquist and Kim, 1996). [PSI +] is required for expression of HSF, and hence viability, in this strain. Similarly, nonsense alleles of other essential genes have been identiﬁed in strains that have become obligately [PSI +]; some examples include POL3 (Kokoska et al., 2000), KAR1, GCN4, and NUP1 (Styles, 1991).

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

407

[PSI +] has been reported to interact genetically with a missense mutation in the NAM9 gene (Chacinska et al., 2000). NAM9 is a nuclear gene that encodes a mitochondrial small ribosomal subunit protein (Boguta et al., 1992; Dmochowska et al., 1995; Boguta et al., 1997; Biswas and Getz, 1999; Chacinska et al., 2000), and the nam9-1 allele encodes a protein with a single amino acid substitution S82L which acts as a Mendelian nonsense suppressor (Boguta et al., 1992; Biswas and Getz, 1999). In the presence of [PSI +], nam9-1 strains are respiratory-deﬁcient, but in a [psi −] background, nam9-1 strains are respiratory-competent (Chacinska et al., 2000). While the molecular target of the read-through event(s) leading to this phenotype is not yet characterized, this observation provides an extremely useful tool for further [PSI +] studies: a positive selection for the [psi −] state. B. Global Physiologic Changes On rich media, the growth of [PSI +]/[psi −] isogenic pairs in several diverse genetic backgrounds is indistinguishable. However, when growth of the same strains is assessed under limiting nutrient conditions, in diverse carbon or nitrogen sources, in the presence of speciﬁc metals, salts, or inhibitors, or following exposure to chemical or thermal stresses, strain- and [PSI +]-dependent constellations of phenotypes are observed (Eaglestone et al., 1999; True and Lindquist 2000). The effects of [PSI +] on ﬁtness under diverse growth conditions are not uniform; in some strains, [PSI +] is advantageous, while in others [PSI +] is disadvantageous (True and Lindquist, 2000). However, the capacity of [PSI +] to profoundly alter yeast physiology is clear; in 150 phenotypic tests, ∼50% of traits were altered by the presence of [PSI +] in at least one genetic background. Two of these [PSI +]-dependent traits (e.g., zinc metabolism and cell-wall biosynthesis) were affected similarly in every genetic background tested, suggesting that these pathways are a common target of [PSI +]. The genetic diversity underlying these phenotypes likely arises from several sources (Fig. 3). As described above, the complement of nonsense mutations in a given strain background will be unique. [PSI +] could also, at low efﬁciency, read through naturally placed stop codons to sample the genetic diversity present in the 3′ untranslated regions of these genes (Levitt, 1991). In addition, there are several open-reading frames interrupted by a single stop codon in the published yeast genome; read-through at these sites has the potential to add another domain or functionality to the 5′ uninterrupted reading frame (True and Lindquist, 2000). [PSI +] could also indirectly alter expression by modulating mRNA

408

TRICIA R . SERIO AND SUSAN L. LINDQUIST

FIG. 3. Potential sources of genetic variation acted upon by [PSI +]-mediated nonsense suppression. Strain genotypes (rectangles represent independent genes) are diagrammed on the left with natural stop codons (white octagons), stop codon mutations or interruptions (gray octagons), or in-frame downstream stop codons (hatched octagon) shown. [PSI +]-mediated nonsense suppression can alter strain phenotype (compare [PSI +] to [psi −]) by acting at nonsense mutations to restore expression of a functional product, by adding a domain or functionality to interrupted ORFs, or by sampling the genetic diversity present in the 3′ UTRs of most genes (hatched rectangle).

turnover by the nonsense-mediated mRNA decay pathway (Culbertson, 1999; Czaplinkski et al., 1999). The multiple possible direct and indirect targets of [PSI +]-mediated nonsense suppression, along with its epigenetic mode of inheritance and reversibility, have profound implications for the ﬁtness and evolvability of yeast (True and Lindquist, 2000). By sampling the genetic diversity already present in strains or that newly acquired by mutation, [PSI +] provides a mechanism to test new phenotypes without committing to their ﬁxation, an outcome that could lead to the accumulation of multiple singly disadvantageous traits that are combinatorily advantageous. Examples of this idea may already be present in the literature for strains containing nonsense mutations in essential genes, but it is unclear if

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

409

the wild-type parent strains were themselves [PSI +] or if the prion state was selected for once the mutation was acquired (Styles, 1991; Lindquist and Kim, 1996; Kokoska et al., 2000). Undoubtedly, future studies will systematically test the hypothesis that [PSI +] can facilitate a strain’s evolvability. While any factor that uncovers preexisting genetic variation on a genome-wide scale seems likely to have an impact on evolutionary processes, it is important to note that this does not imply that [PSI +] evolved for this purpose. It might be an accident of nature or serve some as yet unknown function. VI. CONCLUSION [PSI +]

Recent work on the prion has advanced our understanding of the replication of conformational genetic information and of the physiologic consequences of the Sup35 prion cycle. These studies have opened new potential avenues for research, such as the universality of conversion mechanisms for yeast proteins and the mammalian amyloidogenic factors, the implications of [PSI +] informational suppression for organismal evolvability and ﬁtness, and how a single protein can alternately choose between a host of conformational states available to it. Nevertheless, a fundamental understanding of the regulation of the [PSI +] prion cycle in vivo remains elusive. What triggers spontaneous conversion? How are initiation and propagation similar and/or distinct? How do the regions of Sup35 essential for [PSI +] propagation contribute to protein–protein interaction and conformational conversion? What role do chaperones play in these processes? How do they recognize Sup35? Future work will continue to expand our mechanistic understanding of this unique process and the functional consequences that accompany it. Moreover, it is likely that they will have a broader impact, improving our understanding of other self-perpetuating changes in protein conformation and their roles in general biology and disease. REFERENCES All-Robyn, J., and S. Liebman (1983). Genetics 104, s2. Biswas, T. K., and G. S. Getz (1999). Biochemistry 38, 13042–13054. Boguta, M., A. Chacinska, M. Murawski, and B. Szczesniak (1997). Acta Biochim. Pol. 44, 251–258. Boguta, M., A. Dmochowska, P. Borsuk, K. Wrobel, A. Gargouri, J. Lazowska, P. P. Slonimski, B. Szczesniak, and A. Kruszewska (1992). Mol. Cell Biol. 12, 402–412. Chacinska, A., M. Boguta, J. Krzewska, and S. Rospert (2000). Mol. Cell Biol. 20, 7220–7229. Chernoff, Y., A. Galkin, E. Lewitin, T. Chernova, G. Newnam, and S. Belenkiy (2000). Mol. Microbiol. 35, 865–876.

410

TRICIA R . SERIO AND SUSAN L. LINDQUIST

Chernoff, Y. O., S. G. Inge-Vechtomov, I. L. Derkach, M. V. Ptyushkina, O. V. Tarunina, A. R. Dagkesamanskaya, and M. D. Ter-Avanesyan (1992a). Yeast 8, 489–499. Chernoff, Y. O., S. L. Lindquist, B. Ono, S. G. Inge-Vechtomov, and S. W. Liebman (1995). Science 268, 880–884. Chernoff, Y. O., G. P. Newnam, J. Kumar, K. Allen, and A. D. Zink (1999). Mol. Cell. Biol. 19, 8103–8112. Chernoff, Y. O., M. V. Ptyushkina, M. G. Samsonova, G. I. Sizonencko, Y. I. Pavlov, M. D. Ter-Avanesyan, and S. G. Inge-Vechtomov (1992b). Biochimie 74, 455–461. Chien, P., and J. S. Weissman (2001). Nature 410, 223–227. Chiti, F., P. Webster, N. Taddei, A. Clark, M. Stefani, G. Ramponi, and C. M. Dobson (1999). Proc. Natl. Acad. Sci. USA 96, 3590–3594. Cox, B. (1965). Heredity 20, 505–521. Cox, B. (1993). The Early Days of Yeast Genetics (M. N. Hall and E. Linder, eds.), pp. 219–239. Cold Spring Harbor Laboratory Press. Cox, B. S., M. F. Tuite, and C. S. McLaughlin (1988). Yeast 4, 159–178. Cox, B. S., M. F. Tuite, and C. J. Mundy (1980). Genetics 95, 589–609. Culbertson, M. R. (1999). Trends Genet. 15, 74–80. Czaplinkski, K., M. J. Ruiz-Echevarria, C. I. Gonzalez, and S. W. Peltz (1999). BioEssays 21, 685–696. DePace, A. H., A. Santoso, P. Hillner, and J. S. Weissman (1998). Cell 93, 1241–1252. Derkatch, I. L., M. E. Bradley, and S. W. Liebman (1998). Proc. Natl. Acad. Sci. USA 95, 2400–2405. Derkatch, I. L., M. E. Bradley, S. V. Masse, S. P. Zadorsky, G. V. Polozkov, S. G. Inge-Vechtomov, and S. W. Liebman (2000). EMBO J. 19, 1942–1952. Derkatch, I. L., M. E. Bradley, P. Zhou, Y. O. Chernoff, and S. W. Liebman (1997). Genetics 147, 507–519. Derkatch, I. L., Y. O. Chernoff, V. V. Kushnirov, S. G. Inge-Vechtomov, and S. W. Liebman (1996). Genetics 144, 1375–1386. Dill, K. A. (1999). Protein Sci. 8, 1166–1180. Dmochowska, A., A. Konopinska, M. Krzymowska, B. Szczesniak, and M. Boguta (1995). Gene 162, 81–85. Doel, S. M., S. J. McCready, C. R. Nierras, and B. S. Cox (1994). Genetics 137, 659–670. Eaglestone, S. S., B. S. Cox, and M. F. Tuite (1999). EMBO J. 18, 1974–1981. Eaglestone, S. S., L. W. Ruddock, B. S. Cox, and M. F. Tuite (2000). Proc. Natl. Acad. Sci. USA 97, 240–244. Eurwilaichitr, L., F. M. Graves, I. Stansﬁeld, and M. F. Tuite (1999). Mol. Microbiol. 32, 485–496. Fink, G. R., and J. Conde (1976). Studies on KAR, a gene required for nuclear fusion in yeast. In International Cell Biology 1976–77: Papers Presented at the First International Congress on Cell Biology (B. R. Brinkley and K. R. Porter, eds.), pp. 414–419. Rockefeller University Press, New York. Frolova, L., X. Le Goff, H. H. Rasmussen, S. Cheperegin, G. Drugeon, M. Kress, I. Arman, A. L. Haenni, J. E. Celis, and M. Philippe, et al. (1994). Nature 372, 701–703. Gerlach, W. L. (1975). Mol. Gen. Genetics 138, 53–63. Glover, J. R., A. S. Kowal, E. C. Schirmer, M. M. Patino, J. J. Liu, and S. Lindquist (1997). Cell 89, 811–819. Glover, J. R., and S. Lindquist (1998). Cell 94, 73–82. Grifﬁth, J. (1967). Nature 215, 1043–1044. Harper, J. D., C. M. Lieber, and P. T. Lansbury, Jr. (1997). Chem. Biol. 4, 951–959. Hawthorne, D. C., and U. Leupold (1974). Curr. Topics Microbiol. Immunol. 64, 1–47.

MOLECULAR INSIGHTS AND FUNCTIONAL CONSEQUENCES

411

Inge-Vechtomov, S. (1964). Vestnick LGU 9, 112–117. Inge-Vechtomov, S., and V. Andrianova (1975). A new type of supersuppressor in yeast. In Molecular Mechanisms of Genetic Processes (N. Dubinin and D. Gol’dfarb, eds.), pp. 181–186. John Wiley and Sons, New York. Jung, G., G. Jones, R. D. Wegrzyn, and D. C. Masison (2000). Genetics 156, 559–570. Kikuchi, Y., H. Shimatake, and A. Kikuchi (1988). EMBO J. 7, 1175–1182. King, C. Y., P. Tittmann, H. Gross, R. Gebert, M. Aebi, and K. Wuthrich (1997). Proc. Natl. Acad. Sci. USA 94, 6618–6622. Kokoska, R. J., L. Stefanovic, J. DeMai, and T. D. Petes (2000). Mol. Cell. Biol. 20, 7490– 7504. Kushnirov, V. V., N. V. Kochneva-Pervukhova, M. B. Chechenova, N. S. Frolova, and M. D. Ter-Avanesyan (2000a). EMBO J. 19, 324–331. Kushnirov, V. V., D. S. Kryndushkin, M. Boguta, V. N. Smirnov, and M. D. Ter-Avanesyan (2000b). Curr. Biol. 10, 1443–1446. Kushnirov, V. V., M. D. Ter-Avanesyan, A. R. Dagkesamanskaia, O. Chernov Iu, S. G. Inge-Vechtomov, and V. N. Smirnov (1990). Molek. Biol. 24, 1037–1041. Kushnirov, V. V., and M. D. Ter-Avanesyan (1998). Cell 94, 13–16. Kushnirov, V. V., M. D. Ter-Avanesyan, M. V. Telckov, A. P. Surguchov, V. N. Smirnov, and S. G. Inge-Vechtomov (1988). Gene 66, 45–54. Lansbury, P. T., Jr. (1999). Proc. Natl. Acad. Sci. USA 96, 3342–3344. Levitt, R. C. (1991). Genomics 11, 484–489. Liebman, S. W. (1975). J. Mol. Biol. 94, 595–610. Liebman, S. W., and J. A. All-Robyn (1984). Curr. Genetics 8, 567–573. Lindquist, S., and G. Kim (1996). Proc. Natl. Acad. Sci. USA 93, 5301–5306. Lindquist, S., and E. C. Schirmer (1999). The role of Hsp104 in stress tolerance and prion maintenance. In Molecular Chaperones and Folding Catalysts. Regulation, Cellular Function and Mechanisms (B. Bukau, ed.), pp. 347–380. Harwood Academic Publishers, Amsterdam. Liu, J. J., and S. Lindquist (1999). Nature 400, 573–576. Lund, P. M., and B. S. Cox (1981). Genet. Res. 37, 173–182. Lundberg, K. M., C. J. Stenland, F. E. Cohen, S. B. Prusiner, and G. L. Millhauser (1997). Chem. Biol. 4, 345–355. Mogk, A., T. Tomoyasu, P. Goloubinoff, S. Rudiger, D. Roder, H. Langen, and B. Bukau (1999). EMBO J. 18, 6934–6949. Moriyama, H., H. K. Edskes, and R. B. Wickner (2000). Mol. Cell. Biol. 20, 8916–8922. Newman, G. P., R. D. Wegrzyn, S. L. Lindquist, and Y. O. Chernoff (1999). Mol. Cell. Biol. 19, 1325–1333. Parham, S. N., C. G. Resende, and M. F. Tuite (2001). EMBO J. 20, 2111–2119. Parsell, D. A., A. S. Kowal, M. A. Singer, and S. Lindquist (1994). Nature 372, 475–478. Parsell, D. A., Y. Sanchez, J. D. Stitzel, and S. Lindquist (1991). Nature 353, 270–273. Parsell, D. A., J. Taulien, and S. Lindquist (1993). Phil. Trans. R. Soc. Lond. B, Biol. Sci. 339, 279–285. Patino, M. M., J. J. Liu, J. R. Glover, and S. Lindquist (1996). Science 273, 622–626. Paushkin, S. V., V. V. Kushnirov, V. N. Smirnov, and M. D. Ter-Avanesyan (1996). EMBO J. 15, 3127–3134. Paushkin, S. V., V. V. Kushnirov, V. N. Smirnov, and M. D. Ter-Avanesyan (1997). Mol. Cell. Biol. 17, 2798–2805. Prusiner, S. B. (1982). Science 216, 136–144. Rochet, J. C., and P. T. Lansbury (2000). Curr. Opin. Struct. Biol. 10, 60–68. Sanchez, Y., and S. L. Lindquist (1990). Science 248, 1112–1115.

412

TRICIA R . SERIO AND SUSAN L. LINDQUIST

Sanchez, Y., D. A. Parsell, J. Taulien, J. L. Vogel, E. A. Craig, and S. Lindquist (1993). J. Bacteriol. 175, 6484–6491. Sanchez, Y., J. Taulien, K. A. Borkovich, and S. Lindquist (1992). EMBO J. 11, 2357–2364. Santoso, A., P. Chien, L. Z. Osherovich, and J. S. Weissman (2000). Cell 100, 277–288. Schirmer, E. C., and S. Lindquist (1997). Proc. Natl. Acad. Sci. USA 94, 13932–13937. Serio, T. R., A. G. Cashikar, A. S. Kowal, G. J. Sawicki, J. J. Moslehi, L. Serpell, M. F. Arnsdorf, and S. L. Lindquist (2000). Science 289, 1317–1321. Serio, T. R., and S. L. Lindquist (1999). Annu. Rev. Cell Dev. Biol. 15, 661–703. Serio, T., and S. Lindquist (2001). Adv. Prot. Chem. 57, 335–366. Serpell, L. C., M. Sunde, and C. C. Blake (1997). Cell. Mol. Life Sci. 53, 871–887. Singh, A. (1979). Proc. Natl. Acad. Sci. USA 76, 1952–1956. Stansﬁeld, I., K. M. Jones, V. V. Kushnirov, A. R. Dagkesamanskaya, A. I. Poznyakovski, S. V. Paushkin, C. R. Nierras, B. S. Cox, M. D. Ter-Avanesyan, and M. F. Tuite (1995). EMBO J. 14, 4365–4373. Styles, C. (1991). http://www.tiac.net/users/cstyles/Idios.htm Sunde, M., L. C. Serpell, M. Bartlam, P. E. Fraser, M. B. Pepys, and C. C. Blake (1997). J. Mol. Biol. 273, 729–739. Ter-Avanesyan, M. D., A. R. Dagkesamanskaya, and V. N. Smirnov (1989). Dokl. Akad. Nauk SSSR 308, 1472–1475. Ter-Avanesyan, M. D., A. R. Dagkesamanskaya, V. V. Kushnirov, and V. N. Smirnov (1994). Genetics 137, 671–676. Ter-Avanesyan, M. D., V. V. Kushnirov, A. R. Dagkesamanskaya, S. A. Didichenko, Y. O. Chernoff, S. G. Inge-Vechtomov, and V. N. Smirnov (1993). Mol. Microbiol. 7, 683–692. True, H. L., and S. L. Lindquist (2000). Nature 407, 477–483. Tuite, M., F. Izgu, C. Grant, and M. Crouzet (1988). Genetic control of tRNA suppression in Saccharomyces cerevisiae: Allosuppressors. In Genetics of Translation, New Approaches (M. F. Tuite, M. Picard, and M. Bolotin-Fukuhara, eds.), Vol. 14, pp. 393–402. Springer-Verlag, Berlin. Tuite, M. F., C. R. Mundy, and B. S. Cox (1981). Genetics 98, 691–711. Walsh, D. M., A. Lomakin, G. B. Benedek, M. M. Condron, and D. B. Teplow (1997). J. Biol. Chem. 272, 22364–22372. Weber-Ban, E. U., B. G. Reid, A. D. Miranker, and A. L. Horwich (1999). Nature 401, 90–93. Werner-Washburne, M., J. Becker, J. Kosic-Smithers, and E. A. Craig (1989). J. Bacteriol. 171, 2680–2688. Werner-Washburne, M., and E. A. Craig (1989). Genome 31, 684–689. Werner-Washburne, M., D. E. Stone, and E. A. Craig (1987). Mol. Cell Biol. 7, 2568–2577. Wickner, R. B. (1994). Science 264, 566–569. Wilson, P. G., and M. R. Culbertson (1988). J. Mol. Biol. 199, 559–573. Young, C., and B. Cox (1971). Heredity 26, 413–422. Zhou, P., I. L. Derkatch, and S. W. Liebman (2001). Mol. Microbiol. 39, 37–46. Zhou, P., I. L. Derkatch, S. M. Uptain, M. M. Patino, S. Lindquist, and S. W. Liebman (1999). EMBO J. 18, 1182–1191. Zhouravleva, G., L. Frolova, X. Le Goff, R. Le Guellec, S. Inge-Vechtomov, L. Kisselev, and M. Philippe (1995). EMBO J. 14, 4065–4072. Zolkiewski, M. (1999). J. Biol. Chem. 274, 28083–28086.

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION By JOEL R. HOSKINS, SUVEENA SHARMA, B. K. SATHYANARAYANA, and SUE WICKNER Laboratory of Molecular Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892

I. II. III. IV. V.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clp ATPase Family of Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chaperone Activity of Clp ATPases and Their Participation in Proteolysis . . Structure of Clp ATPases: Alone and with Partner Proteases . . . . . . . . . . . . . . Mechanism of Action of Clp ATPases as Chaperones and as Components of Degradation Machinery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Substrate Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Substrate Unfolding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Substrate Release or Translocation to the Protease . . . . . . . . . . . . . . . . . . VI. Clp ATPase Specificity Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

413 414 415 416 419 419 422 424 425 426 426

I. INTRODUCTION Clp ATPases form a large family of homologous ATPases that participate in many cellular functions including DNA replication, tolerance to heat stress, control of gene expression, protein degradation, and inheritance of prionlike factors. Protein unfolding/refolding has been implicated in each of these processes, and biochemical studies reveal that the Clp ATPases are classical molecular chaperones. They act by catalyzing ATP-dependent protein unfolding, often catalyzing reactions involving disassembly of protein complexes and aggregates. Additionally, a close relationship between protein unfolding and degradation has emerged, demonstrating that Clp chaperones not only assist in pathways of protein remodeling, but also act in intimate complexes with proteolytic components to unfold and present proteins for degradation. In this chapter, we review the structure and mechanism of action of Clp ATPases, citing as examples mainly the extensively studied prokaryotic members, but also keeping in mind that homologous proteins exist in all organisms and very likely carry out similar reactions (Gottesman et al., 1997a,b; Horwich, 1995; Horwich et al., 1999; Porankiewicz et al., 1999; Schirmer et al., 1996; Wawrzynow et al., 1996; Wickner et al., 1999). 413 ADVANCES IN PROTEIN CHEMISTRY, Vol. 59

C 2002 by Academic Press. Copyright All rights of reproduction in any form reserved. 0065-3233/02 $35.00

414

JOEL R . HOSKINS ET AL.

II. Clp ATPase FAMILY OF PROTEINS Clp ATPases are highly conserved and they are present in all organisms examined so far (Gottesman et al., 1990b). They have been identiﬁed in bacteria, some archaea, yeast, plants, insects, and humans. Moreover, many organisms have multiple family members. For example, Escherichia coli has four Clp ATPases: ClpA (Gottesman et al., 1990a), ClpB (Squires et al., 1991), ClpX (Gottesman et al., 1993; Wojtkowiak et al., 1993), and HslU (also referred to as ClpY) (Chuang et al., 1993; Missiakas et al., 1996; Rohrwild et al., 1996). The Clp ATPase family can be broadly divided into two subfamilies: Class 1, containing two nucleotide-binding domains, such as ClpA (Gottesman et al., 1990a) and ClpB (Kitagawa et al., 1991) in E. coli, Hsp104 in Saccharomyces cerevisiae (Parsell et al., 1991), ClpC in plants (Shanklin et al., 1995), and ClpE in Bacillus subtilis (Derre et al., 1999); and Class 2, containing one nucleotide-binding domain and including ClpX in E. coli, yeast, plants, mice, and humans (Corydon et al., 2000; Gottesman et al., 1993; Halperin et al., 2001; van Dyck et al., 1998) and HslU (Chuang et al., 1993) [see Fig. 1 for salient features of the sequence alignments; reviewed by Schirmer et al. (1996)]. While the two ATP-binding domains differ from one another, they are highly conserved

FIG. 1. Schematic representation of the domain structures of different classes of Clp ATPases.

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

415

in all members of Class 1. The single ATP-binding domain of Class 2 Clp ATPases shares signiﬁcant homology to the second ATP binding domain of Class 1. Sequence analysis suggested that the Clp ATPases are members of the AAA+ superfamily of ATPases (ATPases associated with a variety of cellular activities) (Confalonieri and Duguet, 1995; Neuwald et al., 1999; Patel and Latterich, 1998; Swafﬁeld et al., 1995). The key feature of the family is a highly conserved ATP-binding module. Another feature is a protein recognition domain located C-terminally to the ATPase domain, referred to as the sensor-2 (Neuwald et al., 1999) or SSD (sensor and substrate discrimination) (Smith et al., 1999) domain. Many AAA ATPases, including many Clp ATPases, are components of molecular machines, functioning in a large number of diverse cellular processes. III. CHAPERONE ACTIVITY OF Clp ATPases AND THEIR PARTICIPATION IN PROTEOLYSIS The founding member of the Clp ATPase family is ClpA. It was ﬁrst identiﬁed as the ATPase component of a two-component ATPdependent protease (referred to as ClpAP or protease Ti) (Hwang et al., 1988; Katayama et al., 1998; for reviews, see Gottesman et al., 1997b; Wickner et al., 1999). ClpP, the proteolytic component, is unrelated to the Clp ATPases, but is a member of a large family of serine proteases. By itself ClpP can degrade only short peptides, but when present as a complex with ClpA or ClpX, it can degrade large speciﬁc proteins. ClpP homologs have been found in the chloroplasts and plastids of higher plants and in the mitochondrial matrix of higher plants and animals, including humans (Corydon et al., 2000; Larsen and Finley, 1997; Porankiewicz et al., 1999). Interestingly, ClpP homologs have not been detected in S. cerevisiae or Mycoplasma genitalium. Some of the Clp ATPases act as classical ATP-dependent chaperones in the absence of a partner protease. For example, ClpX participates in DNA transposition and replication of bacteriophage Mu in vivo and is essential for Mu growth (Mhammedi-Alaoui et al., 1994). In vitro Mu transposition is promoted by an extremely stable complex containing a tetramer of the transposase (MuA) bound to the recombining DNA. ClpX mediates the disassembly of MuA transposase tetramers from Mu DNA after recombination. MuA is not modiﬁed or degraded and is released as monomers which are able to perform multiple rounds of recombination in vitro ( Jones et al., 1998; Levchenko et al., 1995). ClpA, independent of ClpP, performs ATP-dependent chaperone functions in vitro which mimic those of the DnaK/Hsp70 chaperone system (Wickner et al., 1994). It catalyzes the remodeling of inactive dimers of

416

JOEL R . HOSKINS ET AL.

the plasmid P1 replication initiator protein, RepA, into active monomers that bind with high afﬁnity to DNA sites in the P1 origin of replication. In an analogous reaction, ClpX dissociates the dimeric initiator protein, TrfA, of plasmid RK2 into monomers in vitro (Konieczny and Helinski, 1997). Like other classical chaperones, both ClpA and ClpX prevent heat inactivation of several proteins in vitro (Wickner et al., 1994; Wawrzynow et al., 1995). In every case, the chaperone substrates for ClpA and ClpX are also substrates for degradation by the respective ClpAP or ClpXP protease. For example, ClpXP degrades MuA, and ClpAP degrades RepA (Levchenko et al., 1995; Wickner et al., 1994). Thus the speciﬁcity of protein recognition resides on the ATPase component. Some Clp ATPases, for example ClpB in E. coli and Hsp104 in S. cerevisiae, act solely in the absence of a proteolytic component (see chapter by Seris and Lindquist in this volume). Both clpB and hsp104 mutants are defective in thermotolerance (Parsell et al., 1991; Sanchez et al., 1992; Squires et al., 1991) and are unable to reactivate and resolubilize aggregates formed during exposure to high temperature (Parsell et al., 1994; Vogel et al., 1995). In vitro Hsp104 and ClpB participate in reactivating proteins that have been denatured and allowed to aggregate (Diamant et al., 2000; Glover and Lindquist, 1998; Mogk et al., 1999; Motohashi et al., 1999; Zolkiewski, 1999). Neither ClpB nor Hsp104 alone carries out these disaggregation reactions; rather, they act in conjunction with DnaK/Hsp70 chaperone systems. The observation that Hsp104 requires the participation of the eukaryotic Hsp70 chaperone system and ClpB requires the E. coli DnaK chaperone system suggests that the two classes of chaperones interact. Many of the Clp ATPases, including ClpB, Hsp104, ClpX, and HslU, as well as the ClpP and Hs1V proteases, are induced by heat shock, suggesting that they participate in renaturation of heatinactivatedproteins anddegradationofunfolded andabnormalproteins. In view of the fact that Clp chaperones and proteases are not generally essential for growth, there is very likely an overlap in the functions carried out by the various cellular chaperones and proteases. IV. STRUCTURE OF Clp ATPases: ALONE AND WITH PARTNER PROTEASES Clp ATPases self-assemble into oligomeric rings in the presence of ATP or nonhydrolyzable ATP analogs. It has been demonstrated both by biochemical techniques such as gel ﬁltration and by electron microscopy that ClpA, ClpX, ClpB, Hsp104, and HslU form hexamers (Grimaud et al., 1998; Kessel et al., 1995, 1996; Kim et al., 2000a; Parsell et al., 1991; Rohrwild et al., 1997; Zolkiewski et al., 1999). The Clp ATPases known to

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

417

FIG. 2. Structure of E. coli ClpAP. (A) Side view of ClpA12P14 reconstructed from electron micrographic data (Beuron et al., 1998). (B) Top view of the ClpA6 model (Beuron et al., 1998). (C) Cutaway view of the model of ClpA12P14 (Beuron et al., 1998). (D) Structure of ClpP14 shown in a ribbon diagram (Wang et al., 1997). (E) Cutaway view of model of ClpP14 (Reprinted from Larsen and Finley, 1997, with permission of Elsevier Science).

be components of degradation machinery form stable complexes with their corresponding proteolytic component. By electron microscopy the ClpA, ClpX, and HslU hexameric rings are seen bound at one or both ends of two stacked oligomeric rings of their respective proteolytic component (Grimaud et al., 1998; Kessel et al., 1995, 1996; Rohrwild et al., 1996, 1997) (Fig. 2). The structures are remarkably similar in appearance to the archaebacterial and eukaryotic 26S proteosomes, suggesting a common mechanism of action despite little sequence similarity between the analogous components (Grimaud et al., 1998; Kessel et al., 1995). Interestingly, there is a symmetry mismatch between the stacked seven-membered rings of ClpP and the six-membered rings of ClpA or ClpX. This symmetry mismatch is not currently believed to be critical for the mechanism of degradation since hexameric rings of HslU associate with the two stacked hexameric rings of HslV. The crystal structure of one Clp ATPase, HslU, has been solved alone and in a complex with its proteolytic component, HslV (Bochtler et al., 2000; Sousa et al., 2000) (Fig. 3, see color insert). The E. coli HslU

418

JOEL R . HOSKINS ET AL.

structure conﬁrmed that Clp ATPases are structurally similar to the ATP-binding domains of AAA ATPases, including the N-ethylmaleimidesensitive fusion protein, Sec18, VAT, and p97 (Cdc48p) (Babor and Fass, 1999; Coles et al., 1999; Lenzen et al., 1998; May et al., 1999; Yu et al., 1998, 1999; Zhang et al., 2000). The H. influenzae HslUV structure solved by McKay and colleagues showed that a hexameric ring of HslU binds intimately to each axial end of HslV (Sousa et al., 2000). In the complex, the carboxy-terminal domain of HslU binds between adjacent subunits of HslV. HslU, unlike other Clp ATPases and AAA ATPases, has an intermediate domain, referred to as the “I domain,” that interrupts the polypeptide sequence of the nucleotide-binding domain. The I domains extend outward from the molecule and resemble tentacles. Because of their location on the molecule, it has been suggested that they are responsible for the initial recognition of substrates. An axial channel can be seen in the HslU structure, suggesting that substrates can be channeled through HslU and presented to HslV. The crystal structure of E. coli HslUV solved by Huber and collaborators showed HslU in contact with HslV through the opposite axial face of HslU (Bochtler et al., 2000). In that orientation HslU appears to interact very loosely with HslV through the I domain. The fact that E. coli HslU is about 80% identical in amino acid sequence to Haemophilus influenzae HslUV would suggest that the structures of the two proteins should be similar. The issue of the orientation is not resolved; however, several lines of biochemical and electron micrographic evidence suggest that the structure solved by McKay and co-workers is most likely the biologically active form (Ishikawa et al., 2000; Kim et al., 2001; Song et al., 2000; Wang et al., 2001). The crystal structures of ClpP and HslV reveal that the proteolytic active sites are located within the chamber formed by the junction of the two stacked rings (Bochtler et al., 1997; Wang et al., 1997) (Figs. 2 and 3), forming a structure similar to that of the 20S proteolytic core of the proteasome (Groll et al., 1997; Lowe et al., 1995; Seemuller et al., 1995). With this molecular architecture, the active sites of the protease are sequestered from the cytoplasm. The cavities formed by ClpP or HslV are large enough to accommodate unfolded proteins of roughly 30– 40 kDa. Both electron microscopy and X-ray crystallography show that there are small pores at either end of the proteolytic component that are only large enough to allow short polypeptides or unfolded proteins entry into the proteolytic chamber without major conformational changes. Several mechanistic aspects regarding the chaperone activity of Clp ATPases and their role in proteolysis have emerged from these structural studies: (1) The active sites of the protease are sequestered from the cytoplasm and thus are not in a position to degrade most cellular

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

419

proteins; (2) the portals to the proteolytic chambers are not large enough to allow entry of globular proteins without undergoing major conformational changes; and (3) the Clp ATPase components are positioned over the pores in an excellent position to regulate the entry of a substrate into the chamber.

V. MECHANISM OF ACTION OF Clp ATPases AS CHAPERONES AND AS COMPONENTS OF DEGRADATION MACHINERY Our current model (see Fig. 4A and B, see color insert) for the mechanism of action of Clp chaperones is that Clp ATPases recognize and bind speciﬁc substrates through speciﬁc motifs. They also recognize misfolded, aggregated proteins, presumably through exposed hydrophobic regions. In reactions requiring ATP hydrolysis, they catalyze unfolding of speciﬁcally tagged substrates and either (1) release the unfolded protein, allowing it the opportunity to refold in a native conformation, (2) deliver it to another cellular chaperone system for refolding; or (3) when associated with proteases, Clp ATPases translocate the bound unfolded substrate to the proteolytic component for degradation. Thus, the mechanism of protein remodeling by Clp ATPases is envisioned as being the same as the early steps in the pathway of degradation. Many data support this general mechanism. A. Substrate Recognition The initial step in both protein remodeling and degradation by Clp chaperones and proteases is substrate recognition and binding by the ATPase component (step 1 in Fig. 4A and B). In general, Clp ATPases must be in their assembled multimeric form to bind substrates. For the speciﬁc substrates studied to date, Clp ATPases recognize and bind with high afﬁnity to a small 10–20 amino acid region of the substrate residing near an end, either the amino or carboxy terminus. For several substrates, including MuA (Levchenko et al., 1995), Mu vir repressor (Laachouch et al., 1996), and SsrA-tagged polypeptides (Keiler et al., 1996), sites in the carboxy-terminal sequences of the substrate are primarily responsible for recognition by ClpX and/or ClpA. For others, including certain β-galactosidase fusion proteins bearing hydrophobic amino-terminal amino acids (Shrader et al., 1993; Tobias et al., 1991), RepA (Wickner et al., 1994), HemA (Wang et al., 1999) UmuD′ (Frank et al., 1996), and λO (Gonciarz-Swiatek et al., 1999), recognition is through sites near the amino terminus.

420

JOEL R . HOSKINS ET AL.

Several speciﬁc recognition tags have been studied in detail. The Sauer laboratory showed that proteins which become stalled on ribosomes due to the lack of an in-frame stop codon are tagged for degradation by the addition of an 11–amino acid peptide encoded by a small RNA, ssrA (Karzai et al., 1999; Karzai and Sauer, 2001; Keiler et al., 1996). The tag, AANDENYALAA, is added to the carboxy terminus of a nascent polypeptide chain by cotranslational switching of the ribosome from the damaged messenger RNA to ssrA RNA (Keiler et al., 1996). This tag then targets the abnormal protein covalently linked to it for degradation. In E. coli both ClpAP and ClpXP as well as two other proteases, FtsH and Tsp, degrade SsrA-tagged proteins (Gottesman et al., 1998; Herman et al., 1998; Keiler et al., 1996). The C-terminal two amino acids of SsrA, Ala– Ala, are critical for recognition by ClpA and ClpX; the substrates are no longer degraded by ClpAP and ClpXP when those residues are changed to Asp–Asp (Gottesman et al., 1998). A study comparing SsrA-tagged model proteins of different known stabilities showed that the presence of the tag, but not the intrinsic stability of the attached protein, determines the efﬁciency of degradation by ClpXP (Kim et al., 2000b). The requirements for recognition of RepA by ClpA and λO by ClpX have also been explored. RepA associates with ClpA with a ratio of one RepA dimer per ClpA hexamer (Pak and Wickner, 1997). In contrast to SsrA, binding by RepA to ClpA is through a signal in RepA that is near but not at the N terminus, in the vicinity of amino acids 10–15 (Hoskins et al., 2000a). This result shows that tags need not be located at the precise ends of substrates to be recognized. When the RepA tag, the ﬁrst 15 amino acids of RepA, was fused to an otherwise stable protein, the fusion protein was targeted for degradation by ClpAP. However, when more of RepA (∼50 amino acids) was included in the targeting signal, the fusion proteins had higher afﬁnity for ClpA, suggesting the possibility of additional secondary recognition sites. A similar conclusion with respect to λO protein was reached by the Zylicz laboratory based on the observation that ClpX was able to interact, albeit poorly, with λO lacking the N-terminal recognition signal (Gonciarz-Swiatek et al., 1999). Thus, these recognition signals are in some way different from the SsrA tag FIG. 3. Crystal structure of H. influenzae HslU12V12 (Sousa et al., 2000). (A) Ribbon diagram of HslUV complex in side view. (B) Top view of the same HslUV complex. HslU is colored gold except for a single protomer in each hexameric ring which is colored white. HslV is colored blue except for one protomer in each ring, where one is colored pink and the other red. FIG. 4. (A) Model of the mechanism of action of Clp chaperones (see text for discussion). (B) Model of the mechanism of action of Clp ATPases in complexes with proteases (see text for discussion).

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

421

where the tag is the sole factor in determining the strength of the interaction between the substrate and ClpX (Kim et al., 2000b). ClpA also recognizes a set of proteins that follow the N-end rule for degradation in bacteria (Shrader et al., 1993; Tobias et al., 1991). When arginine, lysine, leucine, phenylalanine, tyrosine, or tryptophan were fused to the N terminus of the test protein, β-galactosidase, the halflife of the protein in vivo was about two minutes; the half-lives of fusion proteins with other amino-terminal amino acids were about ten hours. Amino-terminal arginine and lysine are secondary destabilizing residues because their activity depends on their conjugation to the primary destabilizing residues, leucine or phenylalanine, by leucine, phenylalanine– transfer RNA–protein transferase. Interestingly, the unstable proteins were stabilized in E. coli strains with deletions in clpA, implicating a role of ClpA in the degradation reaction. Up to now these observations have not been reproduced in vitro with puriﬁed ClpAP, and it is possible other cellular components participate in this reaction. There are interesting examples of the targeting of a substrate, which is not recognized as a substrate itself, through its complex with a protein that is recognized by the Clp ATPase. For example, UmuD′ , a subunit of the error-prone mutagenesis DNA polymerase in E. coli, is not recognized by ClpXP in its homodimeric form. However, it is degraded by ClpXP when it is in a heterodimeric complex with UmuD (Frank et al., 1996; Gonzalez et al., 2000). UmuD is the unprocessed and inactive form of UmuD′ , containing an additional 24 amino acids at its amino-terminal end. This extra N-terminal portion is essential for recognition by ClpXP. Interestingly, only the UmuD′ subunit is degraded, allowing UmuD to target an excess of UmuD′ for degradation. A peptide corresponding to the 24 amino-terminal amino acids of UmuD is also able to promote degradation of UmuD′ by ClpXP. However, this same peptide did not stimulate degradation of other substrates by ClpXP, and its mechanism of activation is currently unknown. In a similar fashion, the Mu repressor is degraded by ClpXP much more efﬁciently when it is in heterocomplexes with mutant Mu vir repressor (Welty et al., 1997). Taken together, the substrate recognition data do not allow a consensus binding motif to be formulated for any of the Clp ATPases. This may reﬂect the fact that the list of known speciﬁc Clp ATPase substrates is rather short. Two things are clear: (1) The rules governing substrate selection are not simple and not fully understood yet, and (2) there is great diversity in the sequences recognized by the various Clp ATPases and also in the sequences recognized by a single Clp ATPase. Speciﬁc substrates bound to Clp ATPases have been visualized by electron microscopy. ClpAP and ClpXP bind RepA and λO, respectively, at

422

JOEL R . HOSKINS ET AL.

their distal ends, away from ClpP (Ishikawa et al., 2001; Ortega et al., 2000). This suggests that substrates enter the Clp ATPase for unfolding by way of the axial pore. Domains, referred to as “sensor- and substratediscrimination domains,” that may be important in substrate recognition have been identiﬁed in the C-terminal domains of ClpA, ClpX, HslU, and ClpB (Smith et al., 1999). The isolated domains bind some substrates with the appropriate speciﬁcity. However, some of the domains cross-reacted with substrates normally degraded by other proteases. Therefore, it is not clear if these substrate-binding domains are the only substrate interaction sites. Moreover, they are located in the C-terminal regions of the Clp ATPases, and the crystal structure of HslU solved by McKay and co-workers would suggest that the substrate docking site should be in the N-terminal region (Sousa et al., 2000). B. Substrate Unfolding The crystal structures of ClpP and HslV showed that the only access to the proteolytic chamber is through small axial pores, large enough to allow the passage of no more than one or two unfolded polypeptide chains. From these observations it was predicted that the ATPase components acted by unfolding the substrate in preparation for its translocation to ClpP, a reaction that could use energy derived from ATP hydrolysis (Gottesman et al., 1997a; Larsen and Finley, 1997; Wang et al., 1997) (step 2 in Fig. 4A and B). Importantly, Horwich and collaborators showed that ClpA carries out ATP-dependent protein unfolding (Weber-Ban et al., 1999). An SsrA tag was fused to green ﬂuorescent protein (GFP), a protein that ﬂuoresces green in its native, folded state. Unfolding of the GFP was monitored by a decrease in ﬂuorescence. When the tagged GFP was incubated with ClpA and ATP (in the absence of ClpP), a small, temporary drop in ﬂuorescence was observed. The small, brief effect could be explained by unfolding by ClpA followed by rapid refolding of the released GFP. To better observe the unfolding event, a mutant form of GroEL that binds unfolded proteins but is unable to release them was used. In the presence of the trap molecule, nearly all of the GFP ﬂuorescence was lost due to unfolding by ClpA and subsequent binding by the mutant GroEL. Deuterium–hydrogen exchange experiments further showed that unfolding by ClpA exposes practically all of the polypeptide chain to water, allowing nearly complete hydrogen exchange. Thus, ClpA has the capacity to unfold substrate proteins globally. ClpX also catalyzes substrate unfolding, and kinetic experiments revealed that unfolding is the rate-determining step in degradation (Kim et al., 2000b; Singh et al., 2000). Thus, global unfolding may be

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

423

a general mechanism for preparing substrates for translocation from Clp ATPases to their proteolytic partners. Knowing that Clp ATPases catalyze protein unfolding and that the pores leading into the proteolytic cavities are small, it seemed likely that substrate proteins are processively unfolded from the attachment point of the degradation signal and threaded in a unidirectional manner from the chaperone into the protease. Matouschek and colleagues tested this by studying degradation of fusion proteins of two well-studied proteins, barnase and dihydrofolate reductase (DHFR) (Lee et al., 2001). Both barnase and DHFR are stabilized against unfolding by their tight binding ligands, barstar and methotrexate, respectively. It was found that when substrates were constructed such that a DHFR domain was inserted after a degradation tag and before barnase, barnase was protected from degradation by ClpAP in the presence of methotrexate. Alternatively, when barnase was inserted after the tag and before DHFR, the DHFR could no longer protect barnase from degradation in the presence of methotrexate. Corresponding results were obtained with ClpXP and with proteasomes. Thus, it appears that substrates are unfolded unidirectionally, starting with the domain nearest the recognition tag. ClpA is able to bypass the initial speciﬁc binding and ATP-dependent unfolding reactions by directly binding unfolded proteins (Hoskins et al., 2000b). This binding does not require a recognition signal, and proteins that are not bound in their native state are bound when unfolded. This suggests that other recognition sites, perhaps simply hydrophobic regions, become exposed following the initial binding reaction, contributing to the stability of the complex between ClpA and the substrate. Interestingly, ClpX is unable to bind unfolded proteins that lack speciﬁc recognition signals (Singh et al., 2000). The same or overlapping sites on ClpA bind unfolded proteins and native tagged proteins (Hoskins et al., 2000b). Unfolded untagged proteins that are bound by ClpA are degraded by ClpAP. This raises the question of the in vivo relevance of this reaction. ClpA mutants are slightly defective in degrading abnormal proteins and ClpA levels increase during heat stress, although not in a σ 32-dependent manner (Katayama et al., 1988). In addition, abnormal proteins synthesized during starvation are stabilized in ClpA mutants (Damerau and St. John, 1993). Thus, ClpA may play at least a limited role in vivo by interacting with unfolded proteins that escape surveillance by the predominant chaperones. Protein unfolding is very likely a general mechanism used by Clp ATPases to carry out their other biological roles in addition to translocation of substrates to proteases. Although it is not a prerequisite that a protein be globally unfolded to be remodeled, very likely RepA dimers,

424

JOEL R . HOSKINS ET AL.

TrfA dimers and MuA tetramers pass through an unfolded intermediate in the pathway of remodeling by Clp ATPases. C. Substrate Release or Translocation to the Protease The ﬁnal step in protein remodeling by Clp ATPases is the release of the substrate (step 3, Fig. 4A). There is no evidence to suggest that proteins refold while associated with Clp chaperones; most likely, in remodeling reactions, Clp ATPases release substrates in an unfolded form [as has been shown for GFP-SsrA (Weber-Ban et al., 1999) and RepA-GFP (Hoskins et al., 2000b)], allowing the unfolded protein the opportunity to fold correctly. However, as mentioned earlier, ClpB participates in refolding reactions in combination with DnaK and Hsp104 acts with the eukaryotic Hsp70 chaperone system. Because of this speciﬁcity, it is likely that some of the Clp ATPases transfer unfolded substrates directly to DnaK/Hsp70 refolding machines. When associated with a protease component, unfolded substrates are translocated from the ClpATPase to the protease without release from ClpA and recapture by the protease (Hoskins et al., 1998) (step 3, Fig. 4B). Substrates “trapped” in a complex with ClpA during steady-state unfolding conditions (by the exchange of ATP with a nonhydrolyzable ATP analog) require additional ATP for translocation to ClpP or for release (Hoskins et al., 2000b). Thus, it appears that ATP hydrolysis may be critical for conformational changes in ClpA that enable both unfolding and translocation. Because the entrances to the proteolytic chambers of ClpP and HslV are small, it seemed likely that the unfolded substrates are threaded into the proteolytic chamber unidirectionally. The Horwich laboratory addressed this question with studies of ClpAP. They constructed substrates containing the carboxy-terminal SsrA tag and a ﬂuorescent probe near either the amino or the carboxy terminus (Reid et al., 2001). They observed a more rapid rise (by 2–4 sec) in ﬂuorescence anisotropy when the probe was adjacent to the carboxy-terminal end than when it was near the amino-terminal end. They also measured the kinetics of ﬂuorescence resonance energy transfer between engineered donor ﬂuorophores in the ClpP cavity and the ﬂuorescein-labeled substrates. Taken together, the results suggest that translocation is directional, with the SsrA-tagged carboxy terminus of the substrate entering ClpP ﬁrst. This ﬁnding is consistent with the observation that unfolding occurs ﬁrst near the tagged end of the substrate (Lee et al., 2001). Degradation is a consequence of translocation (step 4, Fig. 4B) once substrates are in the cavity of ClpP, they are degraded. It is presumed,

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

425

but not shown, that the small oligopeptide products simply diffuse out of the proteolytic chamber. VI. Clp ATPase SPECIFICITY FACTORS Additional cellular factors have been identiﬁed that serve to modulate the activity of Clp ATPases. For example, ClpXP degrades the stationary phase sigma factor of E. coli, σ s, very poorly. Efﬁcient degradation requires an additional protein called RssB (regulator of sigma S) (Bearson et al., 1996; Mufﬂer et al., 1996; Pratt and Silhavy, 1996). (RssB is also referred to as SprE in E. coli, MviA in Salmonella, and ExpM in Erwinia.) Interestingly, RssB is homologous to response regulator proteins and, like many other response regulators, is phosphorylated by acetyl phosphate in vitro (Bouche et al., 1998). However, thus far no cognate sensor protein has been identiﬁed. A direct role of RssB in degradation was established by the demonstration that σ s degradation by ClpXP requires RssB in vitro (Zhou et al., 2001). Biochemical dissection of the reaction showed that RssB acts catalytically to deliver σ s to ClpXP (Zhou et al., 2001). RssB appears to increase the afﬁnity of σ s for ClpX. Separately, σ s and RssB bind ClpX weakly, but together they form a stable ternary complex with ClpX. When ClpP is present, σ s is degraded and RssB is released and recycled. Both in vivo and in vitro RssB has no effect on degradation of other ClpXP substrates (Zhou and Gottesman, 1998; Zhou et al., 2001) or degradation by ClpAP. Thus it appears that this unique targeting protein is regulated through speciﬁc signaling pathways involving phosphate transfer reactions. In turn it regulates proteolysis by catalyzing the presentation of a speciﬁc substrate to a speciﬁc protease. Another example of a Clp speciﬁcity-enhancing factor is SspB (stringent starvation protein B) (Levchenko et al., 2000). During the puriﬁcation of ClpX, a factor associated with ribosomes was observed to stimulate degradation of SsrA-tagged substrates. The stimulatory factor was identiﬁed as SspB, a protein induced by starvation but with no known function. It does not stimulate degradation of other ClpXP substrates, nor does it stimulate degradation by ClpAP. Like RssB, it is not consumed during the degradation of SsrA-tagged substrates. It acts by speciﬁcally binding to the SsrA portion of SsrA-tagged substrates, thereby increasing the efﬁciency with which they are recognized and degraded ClpXP. SspB acts similarly in vivo, stimulating the rate of degradation of SsrA-tagged substrates more than tenfold. These examples of speciﬁcity factors involved in modulating the recognition speciﬁcity of Clp ATPases point out the existence of yet another level of regulation of degradation to assure accurate degradation of

426

JOEL R . HOSKINS ET AL.

speciﬁc substrates by speciﬁc proteases. It would not be surprising if more examples of speciﬁcity factors emerge. One could imagine that these factors may be involved in substrate selection for chaperone activity of Clp ATPases, although examples have not been identiﬁed yet. VII. SUMMARY Although much has been learned about the structure and function of Clp chaperones and their role in proteolysis, the mechanism of protein unfolding catalyzed by Clp ATPases and the mechanism of translocation of the unfolded proteins from Clp ATPases to partner proteases remain unsolved puzzles. However, models in which mechanical force is used to destabilize the structure of the substrate in a processive and directional manner are probable. It also seems likely that when ClpA ATPases are associated with proteases, unfolding is coupled to extrusion of the unfolded protein into the proteolytic cavity. In summary, it is anticipated that the large family of Clp ATPases will accomplish their many important cellular functions by similar mechanisms and what has been learned by studying the prokaryotic members reviewed here will shed a great deal of light on all members of the family. REFERENCES Babor, S. M., and Fass, D. (1999). Proc. Natl. Acad. Sci. USA 96, 14759–14764. Bearson, S. M., Benjamin, W. H., Jr., Swords, W. E., and Foster, J. W. (1996). J. Bacteriol. 178, 2572–2579. Beuron, F., Maurizi, M. R., Belnap, D. M., Kocsis, E., Booy, F. P., Kessel, M., and Steven, A. C. (1998). J. Struct. Biol. 123, 248–259. Bochtler, M., Ditzel, L., Groll, M., and Huber, R. (1997). Proc. Natl. Acad. Sci. USA 94, 6070 – 6074. Bochtler, M., Hartmann, C., Song, H. K., Bourenkov, G. P., Bartunik, H. D., and Huber, R. (2000). Nature 403, 800 – 805. Bouche, S., Klauck, E., Fischer, D., Lucassen, M., Jung, K., and Hengge-Aronis, R. (1998). Mol. Microbiol. 27, 787–795. Chuang, S. E., Burland, V., Plunkett, G., 3rd, Daniels, D. L., and Blattner, F. R. (1993). Gene 134, 1–6. Coles, M., Diercks, T., Liermann, J., Groger, A., Rockel, B., Baumeister, W., Koretke, K. K., Lupas, A., Peters, J., and Kessler, H. (1999). Curr. Biol. 9, 1158 –1168. Confalonieri, F., and Duguet, M. (1995). BioEssays 17, 639– 650. Corydon, T. J., Wilsbech, M., Jespersgaard, C., Andresen, B. S., Borglum, A. D., Pedersen, S., Bolund, L., Gregersen, N., and Bross, P. (2000). Mamm. Genome 11, 899–905. Damerau, K., and St. John, A. C. (1993). J. Bacteriol. 175, 53 –63. Derre, I., Rapoport, G., Devine, K., Rose, M., and Msadek, T. (1999). Mol. Microbiol. 32, 581–593. Diamant, S., Ben-Zvi, A. P., Bukau, B., and Goloubinoff, P. (2000). J. Biol. Chem. 275, 21107–21113.

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

427

Frank, E. G., Ennis, D. G., Gonzalez, M., Levine, A. S., and Woodgate, R. (1996). Proc. Natl. Acad. Sci. USA 93, 10291–10296. Glover, J. R., and Lindquist, S. (1998). Cell 94, 73 – 82. Gonciarz-Swiatek, M., Wawrzynow, A., Um, S. J., Learn, B. A., McMacken, R., Kelley, W. L., Georgopoulos, C., Sliekers, O., and Zylicz, M. (1999). J. Biol. Chem. 274, 13999 –14005. Gonzalez, M., Rasulova, F., Maurizi, M. R., and Woodgate, R. (2000). EMBO J. 19, 5251– 5258. Gottesman, S., Clark, W. P., de Crecy-Lagard, V., and Maurizi, M. R. (1993). J. Biol. Chem. 268, 22618 –22626. Gottesman, S., Clark, W. P., and Maurizi, M. R. (1990a). J. Biol. Chem. 265, 7886–7893. Gottesman, S., Maurizi, M. R., and Wickner, S. (1997a). Cell 91, 435– 438. Gottesman, S., Roche, E., Zhou, Y., and Sauer, R. T. (1998). Genes Dev. 12, 1338 –1347. Gottesman, S., Squires, C., Pichersky, E., Carrington, M., Hobbs, M., Mattick, J. S., Dalrymple, B., Kuramitsu, H., Shiroza, T., Foster, T., et al. (1990b). Proc. Natl. Acad. Sci. USA 87, 3513–3517. Gottesman, S., Wickner, S., and Maurizi, M. R. (1997b). Genes Dev. 11, 815 – 823. Grimaud, R., Kessel, M., Beuron, F., Steven, A. C., and Maurizi, M. R. (1998). J. Biol. Chem. 273, 12476 –12481. Groll, M., Ditzel, L., Lowe, J., Stock, D., Bochtler, M., Bartunik, H. D., and Huber, R. (1997). Nature 386, 463 – 471. Halperin, T., Zheng, B., Itzhaki, H., Clarke, A. K., and Adam, Z. (2001). Plant Mol. Biol. 45, 461 – 468. Herman, C., Thevenet, D., Bouloc, P., Walker, G. C., and D’Ari, R. (1998). Genes Dev. 12, 1348 –1355. Horwich, A. L. (1995). Curr. Biol. 5, 455 – 458. Horwich, A. L., Weber-Ban, E. U., and Finley, D. (1999). Proc. Natl. Acad. Sci. USA 96, 11033 –11040. Hoskins, J. R., Kim, S. Y., and Wickner, S. (2000a). J. Biol. Chem. 275, 35361–35367. Hoskins, J. R., Pak, M., Maurizi, M. R., and Wickner, S. (1998). Proc. Natl. Acad. Sci. USA 95, 12135–12140. Hoskins, J. R., Singh, S. K., Maurizi, M. R., and Wickner, S. (2000b). Proc. Natl. Acad. Sci. USA 97, 8892– 8897. Hwang, B. J., Woo, K. M., Goldberg, A. L., and Chung, C. H. (1988). J. Biol. Chem. 263, 8727– 8734. Ishikawa, T., Beuron, F., Kessel, M., Wickner, S., Maurizi, M. R., and Steven, A. C. (2001). Proc. Natl. Acad. Sci. USA 98, 4328 – 4333. Ishikawa, T., Maurizi, M. R., Belnap, D., and Steven, A. C. (2000). Nature 408, 667– 668. Jones, J. M., Welty, D. J., and Nakai, H. (1998). J. Biol. Chem. 273, 459 – 465. Karzai, A. W., and Sauer, R. T. (2001). Proc. Natl. Acad. Sci. USA 98, 3040 – 3044. Karzai, A. W., Susskind, M. M., and Sauer, R. T. (1999). EMBO J. 18, 3793–3799. Katayama, Y., Gottesman, S., Pumphrey, J., Rudikoff, S., Clark, W. P., and Maurizi, M. R. (1988). J. Biol. Chem. 263, 15226 –15236. Keiler, K. C., Waller, P. R., and Sauer, R. T. (1996). Science 271, 990 –993. Kessel, M., Maurizi, M. R., Kim, B., Kocsis, E., Trus, B. L., Singh, S. K., and Steven, A. C. (1995). J. Mol. Biol. 250, 587–594. Kessel, M., Wu, W., Gottesman, S., Kocsis, E., Steven, A. C., and Maurizi, M. R. (1996). FEBS Lett. 398, 274 –278. Kim, K. I., Cheong, G. W., Park, S. C., Ha, J. S., Woo, K. M., Choi, S. J., and Chung, C. H. (2000a). J. Mol. Biol. 303, 655 – 666.

428

JOEL R . HOSKINS ET AL.

Kim, Y. I., Burton, R. E., Burton, B. M., Sauer, R. T., and Baker, T. A. (2000b). Mol. Cell 5, 639 – 648. Kim, Y. I., Levchenko, I., Fraczkowska, K., Woodruff, R. V., Sauer, R. T., and Baker, T. A. (2001). Nat. Struct. Biol. 8, 230 –233. Kitagawa, M., Wada, C., Yoshioka, S., and Yura, T. (1991). J. Bacteriol. 173, 4247– 4253. Konieczny, I., and Helinski, D. R. (1997). Proc. Natl. Acad. Sci. USA 94, 14378–14382. Laachouch, J. E., Desmet, L., Geuskens, V., Grimaud, R., and Toussaint, A. (1996). EMBO J. 15, 437– 444. Larsen, C. N., and Finley, D. (1997). Cell 91, 431– 434. Lee, C., Schwartz, M. P., Prakash, S., Iwakura, M., and Matouschek, A. (2001). Mol. Cell 7, 627– 637. Lenzen, C. U., Steinmann, D., Whiteheart, S. W., and Weis, W. I. (1998). Cell 94, 525 –536. Leonhardt, S. A., Fearson, K., Danese, P. N., and Mason, T. L. (1993). Mol. Cell Biol. 13, 6304 – 6313. Levchenko, I., Luo, L., and Baker, T. A. (1995). Genes Dev. 9, 2399 –2408. Levchenko, I., Seidel, M., Sauer, R. T., and Baker, T. A. (2000). Science 289, 2354 –2356. Lowe, J., Stock, D., Jap, B., Zwickl, P., Baumeister, W., and Huber, R. (1995). Science 268, 533–539. May, A. P., Misura, K. M., Whiteheart, S. W., and Weis, W. I. (1999). Nat. Cell Biol. 1, 175 –182. Mhammedi-Alaoui, A., Pato, M., Gama, M. J., and Toussaint, A. (1994). Mol. Microbiol. 11, 1109 –1116. Missiakas, D., Schwager, F., Betton, J. M., Georgopoulos, C., and Raina, S. (1996). EMBO J. 15, 6899 –6909. Mogk, A., Tomoyasu, T., Goloubinoff, P., Rudiger, S., Roder, D., Langen, H., and Bukau, B. (1999). EMBO J. 18, 6934 – 6949. Motohashi, K., Watanabe, Y., Yohda, M., and Yoshida, M. (1999). Proc. Natl. Acad. Sci. USA 96, 7184 –7189. Mufﬂer, A., Fischer, D., Altuvia, S., Storz, G., and Hengge-Aronis, R. (1996). EMBO J. 15, 1333 –1339. Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. (1999). Genome Res. 9, 27– 43. Ortega, J., Singh, S. K., Ishikawa, T., Maurizi, M. R., and Steven, A. C. (2000). Mol. Cell 6, 1515 –1521. Pak, M., and Wickner, S. (1997). Proc. Natl. Acad. Sci. USA 94, 4901– 4906. Parsell, D. A., Kowal, A. S., Singer, M. A., and Lindquist, S. (1994). Nature 372, 475 –478. Parsell, D. A., Sanchez, Y., Stitzel, J. D., and Lindquist, S. (1991). Nature 353, 270–273. Patel, S., and Latterich, M. (1998). Trends Cell Biol. 8, 65 –71. Perier, F., Radeke, C. M., Raab-Graham, K. F., and Vandenberg, C. A. (1995). Gene 152, 157–163. Porankiewicz, J., Wang, J., and Clarke, A. K. (1999). Mol. Microbiol. 32, 449 – 458. Pratt, L. A., and Silhavy, T. J. (1996). Proc. Natl. Acad. Sci. USA 93, 2488 – 2492. Reid, B. G., Fenton, W. A., Horwich, A. L., and Weber-Ban, E. U. (2001). Proc. Natl. Acad. Sci. USA 98, 3768 –3772. Rohrwild, M., Coux, O., Huang, H. C., Moerschell, R. P., Yoo, S. J., Seol, J. H., Chung, C. H., and Goldberg, A. L. (1996). Proc. Natl. Acad. Sci. USA 93, 5808 –5813. Rohrwild, M., Pfeifer, G., Santarius, U., Muller, S. A., Huang, H. C., Engel, A., Baumeister, W., and Goldberg, A. L. (1997). Nat. Struct. Biol. 4, 133 –139. Sanchez, Y., Taulien, J., Borkovich, K. A., and Lindquist, S. (1992). EMBO J. 11, 2357–2364. Schirmer, E. C., Glover, J. R., Singer, M. A., and Lindquist, S. (1996). Trends Biochem. Sci. 21, 289–296.

Clp ATPases AND THEIR ROLE IN PROTEIN UNFOLDING AND DEGRADATION

429

Seemuller, E., Lupas, A., Stock, D., Lowe, J., Huber, R., and Baumeister, W. (1995). Science 268, 579 –582. Shanklin, J., DeWitt, N. D., and Flanagan, J. M. (1995). Plant Cell 7, 1713 –1722. Shrader, T. E., Tobias, J. W., and Varshavsky, A. (1993). J. Bacteriol. 175, 4364 – 4374. Singh, S. K., Grimaud, R., Hoskins, J. R., Wickner, S., and Maurizi, M. R. (2000). Proc. Natl. Acad. Sci. USA 97, 8898 – 8903. Smith, C. K., Baker, T. A., and Sauer, R. T. (1999). Proc. Natl. Acad. Sci. USA 96, 6678 – 6682. Song, H. K., Hartmann, C., Ramachandran, R., Bochtler, M., Behrendt, R., Moroder, L., and Huber, R. (2000). Proc. Natl. Acad. Sci. USA 97, 14103 – 14108. Sousa, M. C., Trame, C. B., Tsuruta, H., Wilbanks, S. M., Reddy, V. S., and McKay, D. B. (2000). Cell 103, 633 – 643. Squires, C., and Squires, C. L. (1992). J. Bacteriol. 174, 1081–1085. Squires, C. L., Pedersen, S., Ross, B. M., and Squires, C. (1991). J. Bacteriol. 173, 4254 – 4262. Swafﬁeld, J. C., Melcher, K., and Johnston, S. A. (1995). Nature 374, 88 –91. Tobias, J. W., Shrader, T. E., Rocap, G., and Varshavsky, A. (1991). Science 254, 1374 –1377. Vogel, J. L., Parsell, D. A., and Lindquist, S. (1995). Curr. Biol. 5, 306 –317. van Dyck, L., Dembowski, M., Neupert, W., and Langer, T. (1998). FEBS Lett. 438, 250–254. Wang, J., Hartling, J. A., and Flanagan, J. M. (1997). Cell 91, 447– 456. Wang, J., Song, J. J., Franklin, M. C., Kamtekar, S., Im, Y. J., Rho, S. H., Seong, I. S., Lee, C. S., Chung, C. H., and Eom, S. H. (2001). Structure 9, 177–184. Wang, L., Elliott, M., and Elliott, T. (1999). J. Bacteriol. 181, 1211–1219. Wawrzynow, A., Banecki, B., and Zylicz, M. (1996). Mol. Microbiol. 21, 895 – 899. Wawrzynow, A., Wojtkowiak, D., Marszalek, J., Banecki, B., Jonsen, M., Graves, B., Georgopoulos, C., and Zylicz, M. (1995). EMBO J. 14, 1867–1877. Weber-Ban, E. U., Reid, B. G., Miranker, A. D., and Horwich, A. L. (1999). Nature 401, 90 –93. Welty, D. J., Jones, J. M., and Nakai, H. (1997). J. Mol. Biol. 272, 31– 41. Wickner, S., Gottesman, S., Skowyra, D., Hoskins, J., McKenney, K., and Maurizi, M. R. (1994). Proc. Natl. Acad. Sci. USA 91, 12218 –12222. Wickner, S., Maurizi, M. R., and Gottesman, S. (1999). Science 286, 1888 –1893. Wojtkowiak, D., Georgopoulos, C., and Zylicz, M. (1993). J. Biol. Chem. 268, 22609 –22617. Yu, R. C., Hanson, P. I., Jahn, R., and Brunger, A. T. (1998). Nat. Struct. Biol. 5, 803 – 811. Yu, R. C., Jahn, R., and Brunger, A. T. (1999). Mol. Cell 4, 97–107. Zhang, X., Shaw, A., Bates, P. A., Newman, R. H., Gowen, B., Orlova, E., Gorman, M. A., Kondo, H., Dokurno, P., Lally, J., Leonard, G., Meyer, H., van Heel, M., and Freemont, P. S. (2000). Mol. Cell 6, 1473 –1484. Zhou, Y., and Gottesman, S. (1998). J. Bacteriol. 180, 1154 –1158. Zhou, Y., Gottesman, S., Hoskins, J. R., Maurizi, M. R., and Wickner, S. (2001). Genes Dev. 15, 627– 637. Zolkiewski, M. (1999). J. Biol. Chem. 274, 28083 –28086. Zolkiewski, M., Kessel, M., Ginsburg, A., and Maurizi, M. R. (1999). Protein Sci. 8, 1899 – 1903.

This . Page Intentionally Left Blank

AUTHOR INDEX

A Abbas-Terki, T., 176, 180, 182 Abegg, A. L., 255, 276 Abeij´on, C., 310, 335 Abeliovich, H., 359, 387 Abell, B., 326, 341 Abola, E., 353, 379 Achsel, T., 258, 279 Adachi, H., 254, 281 Adam, Z., 414, 427 Adams, G. M., 207, 213 Adams, J., 207, 213 Adams, M. D., 28, 43, 202, 214 Adeli, K., 332, 344 Adjadj, E., 255, 256, 272 Adman, E. T., 132, 142, 153 Admon, A., 369, 381 Aebi, M., 307, 331, 335, 343 Aebi, U., 198, 215 Afendis, S. J., 204, 206, 215 Agard, D. A., 197, 214 Agatusuma, T., 172, 184 Aghdasi, B., 269, 280 Ahn, J. Y., 213, 213 Ahn, K., 198, 221 Aime-Sempe, C., 32, 43 Akasaka, K., 172, 184 Akey, C. W., 223, 240 Akhavannik, A., 157, 176, 180 Akhtar, N. J., 124, 154 Akinaga, S., 171, 172, 184 Akioka, H., 208, 221 Akiyama, K., 213, 213 Akiyama, Y., 294, 301 Akopian, T. N., 193, 199, 200, 203, 213, 214, 217 Aksoy, I. A., 29, 37 Akutsu, R., 114, 154 Albers, M. W., 178, 184, 185

Alberto, D., 32, 37 Albright, C., 359, 386 Aldape, R. A., 270, 275 Aldrich, H. C., 189, 193, 196, 197, 202, 218, 221 Alex, L. A., 158, 180 Alexandrescu, A. T., 255, 272 Alfano, C., 4, 17, 37 Alfano, M., 270, 280 Ali, A., 115, 150 Ali, J. A., 163, 168, 172, 180 Aligue, R., 157, 176, 180 Allan, V. J., 193, 196, 218 Allen, A. J., 326, 341 Allen, G. V., 115, 147 Allen, S., 326, 341 Allona, I., 142, 154 Alm, E. J., 45, 71 Almoguera, C., 115, 147, 148, 154 Alnemri, E. S., 167, 185 Alonso-Llamazares, A., 126, 154 Aloria, K., 95, 96, 99, 102 Alter, O., 371, 373, 374, 379 Altuvia, S., 424, 428 Amano, N., 202, 217 Amano, O., 114, 156 Amerik, A. Y., 206, 219 Amin, J., 115, 147 Amons, R., 128, 152 Ampe, C., 82, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 100, 102, 103, 104 Andaluz, E., 359, 384 Anders, K., 364, 374, 388 Andersen, J. B., 259, 277 Anderson, C. L., 295, 299 Anderson, C. W., 160, 185 Anderson, K. M., 108, 114, 143, 151 Anderson, P., 127, 151 Anderson, S. K., 261, 272 Andersson, O., 212, 218 431

432

AUTHOR INDEX

Andley, U. P., 122, 128, 148 Andreeva, L., 244, 272, 272 Andres, A.-C., 114, 151 Andresen, B. S., 414, 415, 426 Andrews, P. R., 255, 277 Anﬁnsen, C. B., 283, 284, 299, 300, 301 Ang, D., 3, 4, 12, 17, 18, 37, 40, 43, 44, 231, 233, 240, 242, 267, 281 Angulo, A., 212, 219 Ansari, H., 270, 275 Anton, L. C., 203, 220, 221 Aoyama, A., 114, 151 Aquilina, J. A., 134, 148 Aragoncillo, C., 128, 142, 149, 154 Arahata, K., 108, 144, 155 Aramaki, H., 202, 217 Aravind, L., 207, 208, 209, 213, 219, 415, 428 Archer, J. E., 96, 98, 360, 388 Arendt, C. S., 190, 196, 198, 199, 213, 215 Ar´evalo-Rodr´ıguez, M., 272, 272, 282 Argon, Y., 158, 182, 322, 326, 339 Armon, T., 206, 214 Armour, C. D., 366, 371, 372, 377, 383 Armstrong, D. L., 269, 280 Armstrong, J. N., 115, 154 Arnold, S. M., 310, 311, 312, 314, 331, 336 Arrigo, A.-P., 106, 114, 115, 121, 126, 127, 139, 143, 144, 148, 150, 153, 154 Arunachalam, B., 328, 341 Asano, T., 114, 151 Ashcroft, A. E., 66, 70, 71 Ashford, V. A., 182 Ashworth, A., 76, 77, 101 ˚ Aslund, F., 351, 386 Asmal, M., 260, 275 Atenico, D. P., 228, 239 Atkinson, T., 53, 71 Augusteyn, R. C., 145, 156 Augustine, J. G., 248, 281 Avalon-Soffer, M., 330, 344 Aviel, S., 158, 182 Avila, J., 93, 95, 96, 99, 100, 104 Ayling, A., 5, 37 Azem, A., 235, 236, 240

B Babity, J. M., 115, 154 Babor, S. M., 418, 426

Baboshina, O. V., 205, 208, 214 B¨achinger, H.-P., 261, 262, 275, 282 Bader, M. W., 291, 292, 296, 299, 300, 301 Baenziger, J. U., 309, 335 Baeuerle, P. A., 329, 332, 342 Baggenstoss, B. A., 178, 184 Baici, A., 9, 10, 12, 14, 25, 41, 42 Baines, A. J., 78, 103 Bajorek, M., 192, 207, 216 Bajramovic, J. J., 115, 155 Baker, D., 45, 71, 197, 214 Baker, E. N., 286, 293, 295, 301 Baker, T. A., 349, 384, 415, 416, 418, 419, 421, 425, 428, 429 Balbach, J., 244, 252, 253, 275 Balch, W. E., 319, 320, 336 Baldwin, R. L., 246, 247, 248, 250, 251, 275, 276, 277, 279, 280, 281 Baldwin, T. O., 106, 152 Ballinger, C. A., 36, 37, 38, 206, 214 Ballou, C. E., 304, 334 Ballou, D. P., 291, 292, 299 Ballou, L., 304, 334 Balogh, G., 125, 127, 130, 135, 144, 145, 150, 155 Balow, J. P., 322, 324, 329, 330, 339 Ban, C., 159, 168, 180 Banach, M., 7, 38 Banavar, J. R., 371, 373, 374, 383 Banecki, B., 12, 25, 30, 37, 43, 413, 416, 429 Banerji, U., 161, 171, 177, 181 Baneyx, F., 5, 37, 112, 125, 128, 142, 154, 155 Bang, H., 256, 260, 266, 268, 276, 280 Bann, J. G., 261, 282 Bao, S. D., 269, 280 Barany, G., 171, 177, 181 Barbacci, E. G., 171, 174, 182, 184 Bard, M., 366, 371, 372, 377, 383 Bardelli, A., 32, 37 Bardwell, J. C., 14, 31, 37, 285, 286, 287, 288, 289, 290, 291, 292, 296, 299, 300, 301, 351, 352, 379, 383, 384, 388 Barent, R. L., 261, 275 Barford, D., 161, 181 Barlowe, C., 359, 379 Barnes, C., 359, 379 Barnett, B. J., 251, 279 Barnum, S. R., 142, 145, 152 Barouch, W., 12, 15, 28, 37, 43 Barrell, B., 84, 102, 349, 377, 378, 382, 386

AUTHOR INDEX

Barrowman, J., 359, 387 Barry, C. E. III, 143, 156 Bartha, B., 167, 185 Barthe, C., 359, 384 Barthelmess, I. B., 267, 279 Bartolucci, S., 80, 100 Bartunik, H. D., 35, 42, 161, 173, 184, 189, 190, 191, 192, 214, 216, 417, 418, 426, 427 Basco, R. D., 359, 384 Basha, E., 107, 113, 116, 120, 155 Bass, J., 322, 326, 339 Bassing, C. H., 270, 275 Basu, S., 158, 185 Batelier, G., 79, 86, 102, 113, 122, 123, 128, 134, 152 Bateman, O. A., 124, 154 Bates, A. D., 170, 182 Bates, D. J., 10, 38 Bates, P. A., 203, 222, 418, 429 Bauer, M. F., 225, 230, 231, 236, 238, 239, 241 Bauer, M. W., 189, 214 Bauer, S. H., 189, 214 Bauer, V. J., 178, 184 Baulieu, E. E., 178, 180, 183 Baum, J., 262, 275 Baumeister, W., 75, 76, 78, 88, 100, 101, 104, 187, 188, 189, 190, 191, 192, 193, 196, 197, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 211, 212, 214, 215, 216, 217, 218, 219, 220, 221, 222, 349, 379, 384, 416, 417, 418, 426, 428, 429 Bauvy, C., 330, 342 Bayer, P., 259, 280 Bayney, R. M., 261, 277 Beach, D., 157, 181 Beal, R. E., 208, 222 Beall, B. W., 115, 116, 150 Bearson, S. M., 424, 426 Beb¨ok, Z., 330, 342 Bech-Otschir, D., 205, 209, 211, 216, 217, 220 Becker, K., 224, 242 Beckerich, J.-M., 18, 39 Beckwith, J., 285, 287, 289, 290, 291, 292, 293, 295, 296, 297, 298, 300, 301, 351, 380, 386 Beechem, J. M., 49, 68, 70, 72, 349, 389 Beever, J. E., 189, 190, 214 Beggs, S., 115, 152 Behlke, J., 111, 112, 134, 139, 148, 152, 154

433

Behnke, M., 207, 213 Behrendt, R., 189, 220, 418, 429 Behrens, S., 269, 275 Beinert, H., 2, 43 Beintema, J. J., 250, 280 Beisel, C., 10, 12, 14, 27, 40 Beito, T. G., 178, 182 Belden, W. J., 359, 379 Belin, D., 290, 292, 293, 295, 296, 298, 300, 301, 351, 380 Belnap, D. M., 189, 211, 214, 217, 417, 426 Benachour, A., 359, 379 Benaroudj, N., 202, 205, 214 Bender, A. T., 7, 37 Benedetti, E. L., 145, 153 Benghezel, M., 359, 379 Benjamin, T. L., 122, 152 Benjamin, W. H., Jr., 424, 426 Benkovic, S., 127, 145, 150 Benndorf, R., 139, 148 Bennett, H. A., 366, 371, 372, 377, 383 Bennett, M. J., 328, 341 Bennink, J. R., 203, 220, 221 Bensaude, O., 3, 4, 40, 41 Ben-Zvi, A. P., 350, 381, 414, 427 Bercovich, B., 206, 214 Berengian, A. R., 119, 134, 148, 151, 153 Berg, H. C., 285, 300 Bergenhem, N., 262, 276 Berger, A., 246, 281 Bergerat, A., 163, 180 Bergeron, J. J., 323, 324, 327, 338 Bergeron, J. J. M., 309, 310, 311, 312, 321, 326, 336, 337, 338, 341 Bergmann, J. E., 319, 336 Berjanskii, M. V., 30, 37 Berlin, V., 354, 386 Bernhardt, A., 271, 280 Bernier, A. F., 121, 126, 152 Berno, A., 365, 380 Bernstein, M., 359, 379 Berrevoets, C. A., 178, 185 Berriman, M., 258, 279 Bertelsen, E. B., 21, 37 Berthold, J., 230, 231, 236, 238, 239, 241 Bertoli, E., 30, 37 Berube, P., 160, 181 Bessette, P. H., 295, 300 Bessinger, M., 284, 300 Betton, J. M., 268, 279, 414, 428

434

AUTHOR INDEX

Betzel, C., 254, 281 Beuron, F., 211, 214, 416, 417, 421, 426, 427 Beyer, A., 207, 214 Bhakta, H., 355, 382 Bhamidipati, A., 94, 95, 96, 97, 98, 98, 103 Bharadwaj, R., 41 Bharti, A., 115, 143, 153 Bhoyroo, V., 323, 339 Bielka, H., 112, 148 Bies, C., 238, 242 Biessmann, H., 127, 140, 152 Bijlmakers, M. J., 177, 180 Billett, M. A., 207, 216 Bilwes, A. M., 158, 180 Bimston, D. N., 32, 37, 43 Binart, N., 157, 178, 180, 182 Binder, R. J., 158, 185 Bindokas, V. P., 272, 278 Bishop, J. M., 157, 183 Bitton, R., 158, 171, 180 Blagosklonny, M. V., 163, 171, 174, 184 Blake, J. A., 202, 214 Blaschek, H., 248, 250, 280 Blatch, G. L., 29, 39 Blattner, F. R., 349, 351, 380, 386, 414, 426 Blobel, G., 224, 227, 228, 239, 241, 242 Block, T. M., 330, 342 Bloemendal, H., 106, 112, 114, 119, 121, 122, 123, 125, 126, 128, 130, 135, 145, 150, 153, 155, 198, 215 Blom, J., 231, 239, 240 Blond-Elguindi, S., 2, 37 Blum, P., 3, 43, 267, 281 Blumberg, B. S., 330, 342 Blumberg, H., 358, 370, 379 Blumenfeld, N., 206, 214 Bochkareva, E. S., 74, 99 Bochtler, M., 189, 190, 191, 192, 196, 199, 214, 216, 220, 349, 379, 417, 418, 426, 427, 429 Bodsky, J. L., 320, 330, 337 Boehm, U., 199, 214 Boelens, W. C., 107, 108, 113, 123, 134, 144, 148, 151, 198, 215 Boespﬂug-Tanguay, O., 141, 148 Boguski, M. S., 374, 383 Bogyo, M., 188, 189, 200, 202, 219, 220, 221, 330, 342, 355, 389 Bohen, S. P., 173, 175, 180 Boisvert, D. C., 53, 71

Bol, S., 114, 156 Bole, D. G., 328, 341 Bolliger, L., 230, 231, 240 Bolund, L., 414, 415, 426 B¨omer, U., 236, 238, 239 Bond, P. D., 349, 385 Bondoc, D. C., 359, 385 Bonifacino, J. S., 204, 214, 320, 330, 337, 343 Bonner, J. J., 127, 140, 152 Bonniaud, P., 115, 143, 144, 148 Boorstein, W. R., 108, 148 Booth, R. J., 116, 148 Booy, F. P., 211, 214, 417, 426 Borglum, A. D., 414, 415, 426 Borkakoti, N., 353, 388 Borkovich, K. A., 157, 180, 349, 387, 416, 429 Borodovsky, A., 189, 202, 221 Bosch, M., 306, 307, 309, 310, 334, 335 Bose, S., 163, 164, 182, 261, 275 Bostian, K. A., 259, 271, 276 Boswell, B. A., 261, 262, 275, 282 Boteva, R., 348, 385 Bothmann, H., 268, 269, 275 Botstein, D., 96, 101, 103, 352, 364, 371, 373, 374, 379, 380, 381, 383, 388 Bouchard, J. P., 160, 181 Bouche, S., 425, 426 Bouchez, D., 190, 219 Bouckaert, J., 273, 275 Bouhouche, I., 172, 182 Bouloc, P., 420, 427 Bourenkov, G. P., 35, 42, 161, 173, 184, 189, 214, 417, 418, 426 Bouzat, J. L., 189, 190, 214 Bouzyk, M. M., 108, 114, 143, 151 Bova, M. P., 112, 116, 122, 123, 124, 128, 134, 141, 148, 150 Bovey, F. A., 245, 275 Bowman, M. E., 272, 281 Braakman, I., 239, 240, 320, 322, 325, 329, 337, 338, 339, 340, 354, 382 Braaten, D., 270, 275 Bradﬁeld, C. A., 174, 178, 180, 181 Bradley, M., 163, 182 Brady, J. P., 143, 148 Brady, S. T., 30, 43 Braig, K., 48, 49, 54, 72, 349, 389 Brambl, R., 138, 142, 153 Brandts, J. F., 247, 250, 262, 275, 278

AUTHOR INDEX

Brannigan, J. A., 192, 214 Branton, P. E., 357, 359, 385 Branza-Nichita, N., 324, 332, 340, 343 Braun, B., 196, 219 Braun, T., 360, 361, 378, 384 Braun, W., 258, 280 Brazil, B. T., 57, 58, 71 Brehmer, D., 16, 17, 18, 19, 20, 34, 37 Brems, D. N., 250, 275 Brennan, M., 247, 250, 275 Brenner, M. B., 141, 148, 321, 322, 328, 337, 338, 341 Bresnick, E. H., 178, 180 Brewer, C. F., 315, 336 Briknarova, K., 18, 20, 37 Bril, A., 108, 114, 143, 151 Brinker, A., 35, 42, 161, 173, 184 Brinkmann, A. O., 178, 185 Brive, L., 18, 20, 37 Brockway, M. J., 178, 183 Brodin, P., 255, 275 Brodsky, B., 262, 275 Brodsky, J. L., 230, 237, 238, 239, 354, 355, 379, 389 Bross, P., 414, 415, 426 Brown, C., 84, 102, 377, 378, 386 Brown, E. L., 365, 384 Brown, J. R., 89, 101 Brown, P. O., 348, 349, 351, 352, 356, 361, 364, 365, 366, 371, 373, 374, 379, 380, 381, 383, 387, 388 Brown, R. C., 178, 181 Bruan, B. C., 204, 214 Bruckner, P., 262, 267, 275, 281 Bruey, J. M., 115, 143, 144, 148, 150 Brugge, J., 178, 184 Brument-Larignon, N., 108, 114, 143, 151 Brunger, A. T., 418, 429 Brunke, M., 267, 278 Brunner, M., 225, 230, 231, 235, 236, 238, 239, 240, 241 Brunsting, J. F., 32, 41, 127, 151 Bruskin, A., 171, 182 Bruun, A. W., 304, 334 Bucala, R., 270, 280 Buchberger, A., 3, 4, 5, 10, 12, 14, 15, 16, 18, 21, 22, 27, 37, 38, 40, 41, 42, 233, 239 Bucher, P., 209, 217 Buchner, E., 7, 30, 37

435

Buchner, J., 4, 38, 43, 61, 71, 106, 112, 113, 121, 123, 125, 126, 128, 130, 132, 134, 135, 144, 148, 149, 150, 151, 152, 154, 155, 157, 158, 162, 163, 164, 167, 168, 169, 170, 176, 180, 182, 184, 185, 248, 261, 262, 275, 278, 281, 284, 300, 347, 348, 379, 387 Buchou, T., 157, 178, 182 Buckle, A. M., 55, 56, 57, 71 Buevich, A. V., 262, 275 Bugl, ¨ H., 261, 275, 351, 352, 379 Bujard, H., 14, 25, 38 Bukau, B., 3, 4, 5, 6, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 34, 37, 38, 39, 40, 41, 42, 43, 81, 83, 99, 127, 139, 148, 153, 228, 233, 239, 242, 265, 267, 275, 277, 316, 336, 347, 349, 350, 373, 379, 381, 382, 385, 414, 416, 427, 428 Bukovsky, A., 270, 281 Bukrinsky, M., 270, 280 Bulleid, N. J., 326, 330, 340, 341, 343 Bullock, G., 115, 140, 152 Bult, C. J., 202, 214 Bundersen, C. B., 30, 37 Burchard, J., 366, 371, 372, 377, 383 Burda, P., 307, 331, 335, 343 Burgman, P. W. J. J., 127, 151 Burkhoff, D., 269, 278 Burkholder, W. F., 2, 15, 20, 21, 25, 39, 42, 44 Burland, V., 414, 426 Burley, S. K., 353, 380 Burlingame, A. L., 198, 213, 221 Burn, P., 171, 177, 181 Burri, L., 199, 214 Burston, J., 359, 387 Burston, S. G., 46, 48, 49, 50, 52, 53, 54, 67, 68, 71, 72, 347, 349, 380, 387 Burton, R. E., 349, 384 Burtscher, H., 264, 280 Bussey, H., 311, 312, 314, 336, 349, 359, 382, 385 Butler, P. C., 144, 151 Butt, E., 126, 148 Butte, A. J., 371, 375, 380 Butters, T. D., 310, 327, 329, 335, 342 Buttin, G., 322, 339 Buttner, ¨ M., 18, 27, 37 Buu, L. M., 359, 383 Bycroft, M., 262, 278

436

AUTHOR INDEX

Byers, B., 359, 381 Byers, M. G., 321, 337 Byrne, M. C., 365, 384

C Cabral, C. M., 330, 331, 343, 344 Cacan, R., 322, 339 Cai, J. W., 320, 337 Cai, Z. L., 212, 219, 355, 380 Cairns, L., 140, 153 Calderon, J., 321, 337 Callebaut, I., 265, 275 Cameron, A. M., 270, 275 Cameron, P. H., 321, 338 Camirand, A., 359, 380, 385 Campbell, K. S., 29, 37, 282 Campbell, M. J., 375, 380 Candido, E. P. M., 108, 113, 115, 123, 128, 134, 149, 151, 152 Candido, M., 113, 122, 152 Canivenc-Gansel, E., 359, 379 Cannon, K. S., 324, 328, 338 Canon, K., 319, 327, 336 Cantley, L. C., 260, 271, 282 Caparon, M. H., 255, 276 Caplan, A. J., 29, 37, 167, 173, 174, 181, 228, 239 Caplan, S., 84, 102 Carden, M. J., 78, 103 Cardenas, M. E., 173, 174, 181, 257, 272, 272, 275, 282 Cardillo, T. S., 83, 101 Cardozo, C., 199, 214, 219 Carlsson, U., 249, 262, 276 Carmel-Harel, O., 352, 364, 381 Carne, A., 76, 77, 101 Caron, A., 115, 140, 155 Carr, D. C., 261, 275 Carranco, R., 115, 148 Carrascosa, J. L., 84, 85, 86, 87, 98, 102 Carratu, ` L., 145, 148 Carrello, A., 161, 173, 180, 183 Carre˜ no, B. M., 328, 341 Carrington, M., 414, 427 Carver, J. A., 106, 112, 122, 123, 134, 135, 148, 152, 155 Carver, L. A., 174, 178, 180 Casado, R., 128, 142, 149, 154

Casagrande, R., 356, 361, 366, 380 Caspers, G.-J., 106, 107, 108, 109, 111, 114, 115, 119, 120, 121, 122, 149 Caspers, P., 263, 281 Casteels, P., 259, 272, 278 Castillo, V., 200, 217 Castro, O., 314, 336 Catelli, M.-G., 157, 172, 178, 180, 182 Cato, A. C., 7, 32, 34, 38, 42 Caubit, X., 261, 280 Caucutt, G., 167, 185 Cavallo, D., 332, 344 Cazzulo, J. J., 308, 309, 320, 323, 324, 330, 335, 335, 337, 340 Cegielska, A., 157, 185 Cejka, Z., 189, 201, 203, 211, 215, 219, 220, 221 Cenedella, R. J., 145, 149 Cerami, A., 270, 280 Cerchia, L., 80, 100 Cerundolo, V., 198, 214 Chadli, A., 172, 182 Chai, A., 348, 349, 387 Chakrabarti, B., 125, 155 Chakraburtty, K., 366, 371, 372, 377, 383 Chamberlin, M. J., 426, 426 Chamness, G. C., 106, 115, 149 Chan, W. K., 178, 181 Chandrasekhar, G. M., 18, 37, 145, 149 Chang, H. C., 173, 178, 181, 185 Chang, H. C. J., 261, 275 Chang, P. F. L., 142, 156 Chang, W. P., 32, 38, 322, 339 Chang, Y., 114, 152 Chang, Z. Y., 112, 123, 125, 128, 132, 149, 156 Chapman, R., 355, 363, 367, 380 Chapon, F., 115, 140, 155 Chappell, T. G., 12, 24, 38 Chardin, P., 18, 37 Charette, S. J., 115, 121, 126, 144, 149, 152, 155 Charlesworth, M. C., 175, 182 Charng, M. J., 269, 280 Chase, T., 115, 147 Chateau, D., 115, 140, 155 Chatellier, J., 57, 71 Chatterjee, M. T., 145, 149 Chaudhuri, M. M., 320, 337 Chavany, C., 158, 171, 174, 180, 182 Chazin, W. J., 255, 275, 278

AUTHOR INDEX

Chee, M. S., 365, 380, 384 Cheetham, M. E., 29, 39 Cheetham, M. I. E., 29, 37 Chen, C. F., 157, 158, 181 Chen, D., 7, 41 Chen, E. Y., 203, 214 Chen, J., 69, 71, 296, 300 Chen, J.-J., 7, 8, 43 Chen, L., 158, 182, 204, 205, 209, 218, 219 Chen, P., 157, 158, 181, 193, 196, 198, 199, 214 Chen, S., 35, 41, 48, 49, 50, 54, 55, 56, 64, 71, 72, 175, 178, 179, 183, 184, 188, 205, 206, 209, 221, 349, 380, 389 Chen, S. W., 207, 213 Chen, S. X., 106, 123, 125, 130, 132, 144, 150 Chen, S. Y., 161, 173, 175, 181 Chen, T., 297, 300 Chen, W., 324, 340 Chen, W.-J., 227, 239 Chen, X., 79, 83, 84, 99, 138, 155 Chen, Y. G., 269, 270, 275, 277 Chen, Y.-M., 114, 128, 142, 144, 151, 156, 157, 158, 181 Chen, Y. Q., 270, 282 Cheng, H. N., 245, 275 Cheng, M. Y., 74, 99, 346, 347, 377, 378, 380 Cheng, Y., 2, 43 Chen-Weiner, J., 359, 385 Cheong, G. W., 416, 420, 422, 428 Cherﬁls, J., 18, 37 Chernenko, G., 34, 44 Chernoff, Y. O., 350, 380 Chevray, P. M., 207, 221 Chiba, T., 213, 218 Chiesi, M., 115, 140, 152 Chinkers, M., 161, 173, 177, 183, 184 Chirico, W. J., 227, 228, 239 Chiu, G., 322, 326, 339 Cho, R. J., 375, 380 Choe, H.-W., 251, 278 Chohan, S., 162, 164, 167, 168, 169, 174, 175, 179, 183 Choi, J., 259, 278 Choi, S. J., 196, 215, 416, 420, 422, 428 Choudhury, P., 330, 331, 343, 344 Chow, R. L., 75, 76, 93, 100, 348, 378, 381 Chow, Y.-H., 157, 163, 185 Chretien, P., 115, 152 Christeller, J. T., 74, 100, 347, 382

437

Christen, P., 9, 10, 12, 14, 22, 25, 38, 41, 42 Chu, S., 374, 380 Chuang, S. E., 351, 380, 414, 426 Chui, W., 112, 123, 128, 149 Chuman, L., 260, 281 Chung, C. H., 189, 207, 213, 213, 219, 221, 414, 415, 416, 417, 418, 420, 422, 427, 428, 429 Chung, D. H., 330, 343 Chung, J., 297, 300 Chu-Ping, M., 204, 206, 215 Church, G. M., 204, 209, 217 Ciechanover, A., 187, 204, 206, 214, 216 Cieplak, M., 371, 373, 374, 383 Ciocca, D. R., 106, 115, 149 Cirino, G., 7, 38 Cjeka, Z., 188, 204, 205, 208, 209, 216 Clardy, J., 258, 259, 278, 281 Clark, J. I., 106, 130, 132, 134, 141, 142, 149, 153, 164, 165, 171, 181 Clark, K., 203, 219 Clark, R. A., 322, 338 Clark, W. P., 414, 415, 423, 427 Clarke, A. K., 413, 414, 415, 427, 428 Clarke, A. R., 48, 52, 53, 54, 67, 71, 72, 349, 380 Clarke, D. J., 30, 38 Clarke, P. A., 161, 171, 177, 181 Clawson, A., 213, 222 Clayton, R. A., 202, 214 Clegg, J. S., 115, 128, 150, 152 Cleland, J. L., 57, 58, 71 Cleveland, D. W., 74, 93, 99 Clevenger, C. V., 32, 38 Cliff, M. J., 52, 53, 71 Clore, G. M., 290, 301 Cobb, B. A., 141, 145, 149 Cochran, B. H., 161, 173, 177, 181, 184 Codogno, P., 330, 342 Coffey, E., 366, 371, 372, 377, 383 Cofﬁno, P., 198, 221 Coggeshall, R. E., 115, 152 Cohen, P., 126, 154, 161, 181 Cohen, R. E., 206, 218 Cohen-Doyle, M. F., 322, 324, 326, 328, 338, 340, 341 Cole, J. A., 297, 300, 301 Cole, N. B., 355, 385 Coleman, C. S., 125, 149 Coleman, J. S., 142, 145, 150

438

AUTHOR INDEX

Coles, M., 203, 214, 418, 426 Colgan, J., 260, 275 Collada, C., 128, 142, 149, 154 Collier, N. C., 127, 149 Colonbonilla, E., 256, 278 Comoglio, P. M., 32, 37 Confalonieri, F., 207, 214, 415, 426 Connell, P., 36, 37, 38, 206, 214 Conway, A., 375, 380 Conzelmann, A., 359, 379 Cook, K. H., 248, 250, 275 Cooper, B. A., 171, 184 Cooper, P. G., 134, 148 Copeland, C. S., 328, 341 Corbett, E. F., 321, 337 Cordell, J., 86, 87, 98, 102 Corey, E. J., 196, 215 Corley, R. B., 325, 340 Corman, M. L., 171, 184 Corsi, A. K., 30, 38, 230, 239 Corydon, T. J., 414, 415, 426 Cosson, P., 359, 380 Costigan, M., 115, 152 Cotto, J. J., 295, 300 Couette, B., 178, 183 Coumailleau, P., 178, 182 Couprie, J., 290, 300 Coux, O., 187, 188, 189, 204, 205, 208, 209, 214, 216, 219, 221, 414, 417, 428 Cowan, N. J., 74, 75, 76, 79, 80, 81, 82, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 98, 99, 100, 101, 102, 103, 104, 348, 378, 381 Cowles, C. R., 359, 380 Cox, J. S., 355, 364, 365, 366, 380, 387 Coyle, J. E., 59, 66, 70, 71 Crabbe, M. J., 134, 149 Craescu, C. T., 255, 256, 272 Craig, E. A., 2, 3, 4, 7, 11, 14, 28, 29, 31, 35, 37, 39, 40, 41, 43, 44, 83, 99, 108, 145, 148, 150, 224, 227, 228, 229, 230, 231, 235, 236, 237, 238, 239, 240, 241, 242, 370, 377, 378, 389 Craik, D. J., 255, 277 Craiu, A., 203, 214 Crane, B. R., 158, 180 Crane, D. D., 143, 156 Cravatt, B. F., 189, 202, 221 Craven, R. A., 354, 378, 380 Creighton, T. E., 106, 150, 249, 275, 287, 288, 289, 295, 300, 301

Crenshaw, D. G., 271, 275 Cresswell, P., 324, 326, 328, 340, 341 Crews, C. M., 197, 201, 216, 218 Crocoll, A., 34, 38 Croes, Y., 144, 148 Cromlish, J. A., 357, 359, 385 Crooke, E., 264, 275, 278 Crooke, H., 297, 300 Crotchett, B., 207, 213 Crotwell, M., 360, 386 Crouzet, M., 359, 384 Crowe, J. H., 125, 130, 135, 144, 145, 155 Cruickshank, A. A., 207, 213 Cruikshank, A. A., 199, 218 Cry, D. M., 206, 218 Csernely, P., 158, 181 Cuesta, R., 127, 144, 149 Cui, D. F., 296, 300 Culbertson, M. R., 5, 43, 79, 83, 84, 104, 208, 215 Cullen, W., 171, 182 Cumsky, M. G., 235, 239 Cunningham, A. F., 116, 149 Curran, B. P., 145, 149 Currie, R. W., 114, 115, 154 Cutforth, T., 157, 181 Cuthill, S., 157, 185 Cwirla, S. E., 2, 37 Cyr, D. M., 13, 28, 30, 31, 36, 40, 42, 226, 228, 235, 236, 239, 242 Czar, M. J., 7, 39, 178, 179, 182

D Dahlmann, B., 190, 204, 214, 218, 222 Dahlquist, F. W., 21, 37 Dahlqvist, K.-I., 245, 275 Dai, H. Z., 125, 132, 156, 366, 371, 372, 377, 383 Dai, K., 157, 158, 181 Dai, Q. H., 262, 275 Dai, R. M., 203, 214 Dailey, F. E., 285, 300 Dairaku, K., 308, 335 D’Alessio, C., 305, 312, 320, 326, 327, 329, 330, 331, 334, 336, 337 Dallinger, P., 288, 300 Dalman, F. C., 178, 180 Dalrymple, B., 414, 427

AUTHOR INDEX

Damerau, K., 423, 426 Damilczyk, U. G., 324, 340 Daniels, D. L., 414, 426 Danielson, P. D., 269, 282 Danishefsky, S., 170, 171, 174, 184 Dantuma, N. P., 201, 218 Darby, N. J., 289, 295, 300, 326, 341 D’Ari, R., 420, 427 Dartigalongue, C., 268, 275 Das, A. K., 161, 163, 181, 182 Das, B. K., 112, 119, 128, 154 Das, K. P., 125, 132, 149 Dasgupta, G., 157, 184 da Silva, A. C., 349, 385 Datta, S. A., 132, 149, 154 Dautz, R. A., 255, 276 David, V., 322, 338 Davidson, A. R., 332, 343, 363, 383 Davies, G. J., 159, 163, 170, 185 Davies, J., 359, 385 Davies, P., 244, 272, 278 Davis, J. E., 231, 240 Davis, J. M., 262, 275 Davis, R. W., 348, 349, 351, 352, 365, 375, 380, 382, 387 Davis, T. A., 3, 39 Dawson, J. R., 329, 342 Dawson, S. P., 207, 216 Dawson, T. M., 271, 282 Dean, N., 359, 387 De Antoni, A., 359, 386 Debarbieux, L., 289, 300 De Costa, B., 171, 185 de Crecy-Lagard, V., 414, 427 Dee, M. F., 171, 184 Deeg, M., 199, 200, 215 Degen, E., 322, 338 de Giuli, R., 200, 220 De Grip, W. J., 132, 155 de Haard-Hoekman, W. A., 106, 112, 114, 121, 122, 123, 126, 128, 130, 153, 155 Deibel, M. R. J., 161, 173, 183, 261, 279 Deibel, M. R., Jr., 178, 184 Deisenhofer, J., 117, 149 de Jong, W. W., 106, 107, 108, 109, 111, 112, 113, 114, 115, 119, 120, 121, 122, 123, 125, 126, 127, 128, 130, 132, 133, 134, 135, 144, 145, 148, 149, 150, 151, 153, 154, 155

439

Dekker, P. J. T., 225, 231, 238, 239, 241 de la Canal, C., 304, 333 De Lange, F., 132, 155 Deleo, A. B., 158, 185 del Mazo, J., 96, 99 DeLuca-Flaherty, C., 11, 19, 38, 164, 181 Demady, D. R., 7, 37 Demand, J., 19, 32, 34, 40, 206, 218 DeMarini, D. J., 208, 215 DeMartino, G. N., 187, 204, 206, 207, 208, 212, 213, 215, 216, 218, 220, 221 de Massay, B., 163, 180 Dembowski, M., 414, 429 Demolliere, C., 359, 380 De Mot, R., 189, 203, 214, 218, 220, 221 de N´echaud, B., 127, 140, 149 Denesyuk, A. I., 119, 149 De Neve, R., 87, 103 Denis, M., 157, 185 Dennis, D. T., 74, 100, 347, 382 Densmore, V., 271, 276 Der, P., 141, 149 Derham, B. K., 106, 122, 134, 149 DeRisi, J., 374, 380 Derman, A. I., 351, 380 Derre, I., 414, 426 Deshaies, R. J., 188, 204, 205, 206, 209, 221, 227, 228, 239 de Silva, A. M., 319, 320, 336 Desmet, L., 419, 428 Dessen, A., 315, 336 Deuerling, C., 83, 99 Deuerling, E., 3, 38, 265, 267, 275, 277 Devasahayam, G., 272, 282 Deveraux, Q., 208, 211, 215, 222 De Veylder, L., 259, 272, 278 Devine, K., 414, 426 de Virgilio, M., 330, 343 Dewallef, Y., 273, 275 DeWitt, N. D., 414, 429 Dhruvakumar, S., 208, 219 Di, Y. P., 320, 337 Diamant, S., 4, 6, 38, 43, 112, 128, 132, 135, 155, 350, 381, 414, 427 Diamond, D. L., 124, 155 Dias, P., 198, 221 Diaz-Latoud, C., 115, 143, 144, 148 Dick, L. R., 199, 203, 207, 213, 218, 219 Dick, T. P., 199, 200, 212, 215, 216, 219 Dickie, P., 329, 342

440

AUTHOR INDEX

Diehn, M., 356, 361, 366, 380 Diercks, T., 203, 214, 418, 426 Dietmeier, K., 223, 230, 231, 236, 238, 239, 240, 241 Dietz, K., 200, 219 Dignard, D., 309, 310, 311, 312, 336 Di Liberto, M., 18, 19, 20, 39 Dillmann, W. H., 106, 115, 149, 152 Dillon, P., 207, 216 Ding, L. L., 108, 115, 124, 128, 134, 141, 148, 149, 150 Ding, Z., 34, 44 DiOrio, C. I., 171, 182, 184 di Paolo, T., 359, 385 Dirnbach, E., 249, 275 Dirr, H. W., 29, 39 Distefano, M. D., 125, 149 Distel, B., 239, 240 Dittmar, K. D., 7, 32, 35, 38, 39, 41, 167, 173, 175, 179, 181, 182, 183, 185 Ditzel, L., 49, 71, 75, 88, 99, 189, 190, 191, 192, 193, 197, 214, 215, 216, 349, 379, 381, 418, 426, 427 Djabali, K., 127, 140, 149 Dmitrovsky, E., 371, 373, 374, 375, 388 Do, H., 80, 92, 102 Dobberstein, B., 223, 241 Dobrzynski, J. K., 86, 99 Dobson, C. M., 89, 101, 255, 276 Dodge, R. W., 251, 275 Dodson, E. J., 159, 163, 170, 185 Dodson, G., 159, 163, 170, 185, 192, 193, 214, 215 Doelling, J. H., 190, 208, 215 Doerrler, W. T., 363, 381 Dohmen, R. J., 199, 203, 214, 216, 219 Dohmen, T., 188, 205, 206, 209, 221 Doignon, F., 359, 384 Dokurno, P., 203, 222, 418, 429 Dolenc, I., 199, 200, 215 Dolence, J. M., 349, 385 Dolinski, K., 173, 174, 181, 257, 275 Domany, E., 371, 373, 382 Domdey, H., 258, 271, 276 Donahoe, P. K., 269, 270, 282 Donaldson, G. K., 74, 102 Dong, H., 365, 384, 389 Donnenberg, M. S., 286, 300 Donner, D. B., 158, 165, 181, 185 Donze, O., 8, 38, 176, 180, 181

Dore, C., 160, 181 Dore, J. J., Jr., 270, 282 D¨orﬂer, S., 355, 364, 382 Doty, J. L., 171, 184 Dougherty, B. A., 202, 214 Douglas, M. G., 227, 228, 239 Dow, S., 366, 371, 372, 377, 383 Dower, W. J., 2, 37 Downing, A. K., 255, 282 Downing, K. H., 96, 99 Downs, C. A., 142, 145, 149, 150 Drakenberg, T., 245, 249, 255, 257, 265, 275, 277, 278 Dreier, L., 224, 241, 359, 386 Drewello, M., 245, 270, 279, 280 Driessen, A. J. M., 4, 40 Driscoll, J., 203, 215 Drubin, D. G., 79, 83, 84, 104 Drugge, R. J., 257, 276 Du, F. Y., 187, 221 Duan, W., 209, 218 Dubendorff, J. W., 75, 103 Dubiel, W., 7, 38, 188, 204, 205, 207, 208, 209, 211, 212, 215, 216, 217, 220, 221 Dubrovsky, L., 270, 280 Ducasse, C., 115, 121, 126, 143, 144, 148, 154 Dudek, J., 238, 242 Duden, R., 359, 380, 381 Dudich, I. V., 119, 149 Dugave, C., 290, 300 Duggleby, H. J., 192, 214 Duglas-Tabor, Y., 143, 148 Duguest, M., 207, 214 Duguet, M., 415, 426 Duhring, J. L., 178, 185 Duina, A. A., 173, 181, 261, 275 Dujon, B., 349, 382 Dumdey, R., 205, 209, 211, 217, 220 Dunbar, J. D., 158, 185 Dunn, J. J., 75, 103 Dunster, N. J., 67, 72 Dupert, J. M., 115, 140, 155 Dur´an, A., 359, 360, 388 Dutta, R., 159, 185 Duvet, S., 322, 339 Dwek, R. A., 310, 323, 324, 327, 329, 330, 332, 335, 338, 340, 342, 343 Dyson, H. J., 13, 30, 31, 40 Dziak, E., 329, 342

441

AUTHOR INDEX

E Earnest, T., 353, 379 Easton, D. P., 9, 38 Eaves, D. J., 297, 301 Eberhardt, N. L., 144, 151 Echols, H., 5, 17, 40, 44 Eckerskorn, C., 61, 71, 139, 150, 188, 199, 200, 215, 219, 221 Eddison, M., 96, 100 Eddy, R. L., 321, 337 Egerton, M., 354, 359, 378, 380, 382 Eggers, D. K., 92, 99 Eggers, M., 212, 216 Ehring, B., 199, 215 Ehrnsperger, M., 4, 38, 106, 112, 113, 121, 123, 125, 126, 130, 132, 134, 135, 144, 148, 149, 150, 152, 154 Eichmann, K., 202, 215 Eilers, M., 223, 227, 235, 239 Eisen, M. B., 352, 364, 367, 371, 372, 374, 376, 380, 381, 383, 388 Eisenberg, D., 189, 218 Eisenberg, E., 3, 4, 11, 12, 15, 28, 35, 37, 38, 39, 42, 43 Eisenstark, A., 259, 276 Ellenberg, J., 355, 385 Ellgaard, L., 325, 329, 337 Elliott, J. G., 326, 340 Elliott, M., 419, 429 Elliott, T., 198, 214, 324, 340, 419, 429 Ellis, R. J., 70, 71, 74, 100, 284, 300, 347, 382 El Ouarmiri, M., 115, 155 Emch, S., 200, 201, 220 Emerman, M., 143, 150 Emmerich, N. P. N., 199, 215 Emr, S. D., 332, 343, 359, 360, 380, 381, 384, 386 Endicott, S., 212, 213, 222 Endo, T., 358, 386 Enenkel, C., 199, 215, 217 Engel, A., 74, 76, 101, 104, 188, 189, 203, 204, 206, 209, 217, 219, 221, 416, 417, 428 Engel, J., 262, 267, 275, 276 Engel, K., 106, 126, 128, 130, 132, 151 Engert, J. C., 160, 181 Englander, S. W., 5, 42, 54, 66, 67, 68, 72 Engman, D. M., 163, 182

Engstrom, U., 307, 335 Ennis, D. G., 419, 421, 427 Enzlin, J., 198, 215 Eom, S. H., 418, 429 Epstein, C. J., 284, 300 Erdjument-Bromage, H., 5, 38, 75, 100, 348, 349, 378, 381, 385 Erdmann, A., 211, 221 Ermonval, M., 322, 339 Ernst, J. F., 76, 103 Escher, A., 267, 278 Eser, M., 351, 383 Espenshade, P., 359, 360, 382, 386 Esposito, G., 134, 152 Essen, L. O., 75, 76, 78, 88, 100, 101, 349, 384 Esser, K., 83, 99 Estep, P., 204, 209, 217 Estrada, L., 35, 41, 179, 182 Evans, P. A., 255, 276 Evans, T. C., Jr., 273, 276 Ewalt, K. L., 61, 71 Eyheralde, I., 207, 216 Eytan, E., 203, 215

F Faber, H. R., 10, 39, 314, 336 Faber, L. E., 7, 42, 178, 184, 185 Fabianek, R. A., 297, 300 Fadden, P., 164, 165, 171, 181 Fairlamb, A. H., 258, 279 Faivre, J.-F., 108, 114, 143, 151 Fan, L. Y., 114, 156 Fan, R., 7, 38 Fanarraga, M. L., 95, 96, 99, 102 Fanchiotti, S., 312, 320, 327, 329, 330, 336, 337 Fandrich, M., 89, 101 Fang, F., 142, 152 Fang, H., 359, 385 Fang, H. Q., 261, 279 Fang, Y., 167, 173, 174, 181 Fangh¨anel, J., 259, 280 Farber, R., 115, 143, 153 Fardeau, M., 115, 140, 155 Farmer, J. D., 260, 278 Farmery, M. R., 326, 330, 341, 343 Farnsworth, P. N., 132, 153

442

AUTHOR INDEX

Farr, G. W., 5, 44, 53, 62, 63, 64, 69, 71, 72, 74, 79, 80, 82, 83, 86, 99, 104, 348, 378, 390 Farrell, A., 176, 181 Farrelly, F. W., 157, 180 Fass, D., 418, 426 Fauman, E. B., 351, 352, 379 Faure, A., 115, 140, 155 Fayadat, L., 332, 343 Fayet, O., 4, 44, 348, 381 Fedoroff, N. V., 371, 373, 374, 383 Feierbach, B., 93, 94, 95, 96, 99, 103 Feifel, B., 14, 22, 38, 231, 240 Feil, I. K., 121, 149 Feldheim, D., 15, 38, 369, 381 Feldman, D. E., 80, 99 Feldman, R., 188, 205, 206, 209, 221 Feldmann, H., 209, 218, 349, 382 Felts, S. J., 158, 165, 181 Feng, G., 352, 389 Feng, H., 22, 23, 26, 27, 41, 57, 58, 59, 72 Feng, Y., 255, 276 Fenteany, G., 196, 203, 214, 215 Fenton, W. A., 45, 46, 49, 50, 54, 55, 56, 59, 61, 64, 65, 68, 70, 71, 72, 347, 349, 387, 389, 424, 428 Feramisco, J. R., 157, 178, 180, 185 Fern´andez, F., 305, 307, 309, 310, 311, 312, 320, 326, 327, 329, 330, 331, 334, 336, 337 Fernandez, J. J., 86, 87, 98, 102 Fernando, P., 115, 128, 142, 149, 152 Ferreira, L. R., 158, 181 Ferrell, K., 7, 38, 188, 204, 205, 207, 208, 209, 211, 212, 215, 216, 220, 221 Ferrero-Garc´ıa, M., 308, 309, 310, 315, 331, 335, 335 Ferreyra, R. G., 80, 99 Ferro-Novick, S., 359, 385, 387 Ferry, J. G., 193, 196, 197, 218 Fersht, A. R., 55, 56, 57, 66, 71, 72, 262, 264, 266, 276, 277, 278 Fesik, S. W., 256, 278 Fessler, J. H., 309, 310, 311, 312, 314, 331, 335, 336 Fessler, L. I., 309, 310, 311, 312, 314, 331, 335, 336 Ficner, R., 258, 279 Finger, A., 360, 361, 378, 382, 384 Fink, A. L., 9, 41, 262, 281

Fink, G. R., 368, 381 Finke, K., 359, 381 Finkelstein, D. B., 157, 180 Finkelstein, R. R., 115, 155 Finley, D., 187, 188, 192, 204, 205, 206, 207, 208, 209, 214, 215, 216, 217, 218, 219, 220, 221, 413, 415, 417, 422, 427, 428 Fischer, D., 424, 425, 426, 428 Fischer, G., 3, 43, 244, 245, 246, 249, 253, 256, 257, 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 276, 277, 278, 279, 280, 281, 282 Fischer, M., 188, 193, 196, 197, 198, 199, 216 Fischer, P. B., 329, 342 FitzGerald, L. M., 202, 214 Fitzgibbon, J., 258, 279 Fitz-Gibbon, S., 189, 218 Flaherty, K. M., 11, 19, 38, 164, 181 Flanagan, J., 4, 5, 17, 30, 40, 42 Flanagan, J. M., 187, 218, 414, 417, 418, 422, 429 Fleck, J., 190, 219 Fleischmann, R. D., 202, 214 Fleming, M. A., 354, 386 Fletterick, R. J., 10, 38 Fliss, A. E., 167, 173, 174, 181 Flocco, M. T., 2, 38 Floth, C., 4, 44 Flury, I., 359, 379 Flynn, G. C., 2, 12, 21, 23, 24, 26, 37, 38, 41, 112, 128, 138, 153 Fodor, S. P., 365, 380, 384 Foellmer, B., 322, 324, 326, 329, 338, 340 Folk, W. R., 30, 37 Follettie, M. T., 365, 384 Fong, Y., 7, 41 Fontalba, A., 93, 95, 100, 104 Forgey, R. W., 255, 276 Forreiter, C., 136, 149 Fors´en, S., 245, 255, 257, 275, 277, 278 Forterre, P., 163, 180 Fortin, M. G., 7, 41 Foster, J. W., 424, 426 Foster, T., 414, 427 Fourie, A. M., 212, 219 Fox, R. O., 255, 265, 276, 277 Fraczkowska, K., 418, 428 Franc, J. L., 332, 343 Franceschelli, S., 145, 148 Franciolini, F., 24, 38

443

AUTHOR INDEX

Frand, A. R., 299, 300, 377, 381 Frank, E. G., 419, 421, 427 Franke, E. K., 270, 275, 276 Franke, W. W., 203, 206, 219 Franklin, M. C., 418, 429 Fransson, C., 249, 276 Frech, C., 289, 300 Frederick, C. A., 248, 281 Freedman, J. H., 115, 142, 143, 152 Freeman, B. C., 4, 21, 32, 38, 43, 261, 276 Freeman, K., 108, 142, 156 Freemont, P. S., 203, 222, 418, 429 Frere, J.-M., 254, 281 Freskg˚ard, P. O., 249, 262, 276 Freund, S. M., 57, 71 Frey, J., 261, 272 Fried, V. A., 188, 204, 205, 208, 209, 216 Friedberg, E. C., 204, 209, 219 Friedlander, R., 361, 381 Friedli, L., 359, 388 Friedman, J., 260, 278 Frien, M., 19, 32, 34, 40 Friend, S. H., 366, 371, 372, 377, 383 Frishman, D., 61, 71, 139, 150, 188, 202, 203, 219 Fr¨ohli, E., 114, 151 Fromentin, A., 115, 143, 150 Frommel, C., 193, 196, 197, 198, 220 Fruchart, J. S., 259, 272, 278 Fru¨ h, K., 198, 212, 215, 219, 221 Frydman, J., 3, 5, 7, 35, 38, 75, 80, 82, 92, 99, 100, 102, 103, 348, 378, 381 Fu, H. Y., 190, 208, 209, 215, 216, 221 Fugua, S. A. W., 106, 115, 149 Fujimoto, D., 369, 378, 388 Fujimura-Kamada, K., 359, 381 Fujimuro, M., 208, 209, 215, 220, 221 Fujinami, K., 211, 215 Fujita, N., 161, 177, 184 Fujiwara, T., 207, 208, 221 Fukazawa, H., 171, 181 Fukui, M., 115, 150 Fulton, J., 261, 275 Funahashi, W., 369, 378, 388 Fu¨ nfschilling, U., 9, 38 Fung, B. K. K., 122, 124, 134, 148 Fung, E. T., 270, 275 Fung-Leung, W. P., 212, 219 Furtak, K., 50, 54, 55, 56, 61, 62, 63, 64, 71, 72

Furuichi, Y., 359, 384 Furusaka, A., 207, 211, 222 Furutani, M., 80, 100 Futcher, B., 364, 374, 388

G Gabel, C. A., 319, 336 Gaber, R. F., 173, 174, 181, 182, 261, 275, 278 Gabrielian, A. E., 375, 380 Gaburjakova, M., 269, 279 Gachotte, D., 366, 371, 372, 377, 383 Gaczynska, M., 203, 214 Gadelle, D., 163, 180 Gaestel, M., 4, 38, 106, 112, 113, 119, 121, 125, 126, 128, 130, 132, 134, 135, 144, 148, 149, 150, 151, 152, 153, 154, 156 Gafni, A., 249, 275 Gaillardin, C., 18, 39 Gaitanaris, G., 4, 39 Galat, A., 244, 258, 260, 261, 276 Galibert, F., 349, 382 Galigniana, M. D., 7, 32, 38, 39, 41 Galitski, T., 368, 381 Gallaschun, R. J., 171, 184 Gallinger, S., 261, 272 Gallo, M. V., 365, 384 Gallwitz, D., 359, 386 Gama, M. J., 415, 428 Gambee, J. E., 261, 282 Gambill, D. B., 229, 231, 235, 236, 237, 238, 239, 242 Gamblin, S. J., 122, 154 Gamer, J., 14, 25, 38 Ga˜ na´ n, S., 308, 309, 310, 335, 335 Ganoth, D., 203, 206, 214, 215 Gantz, P., 170, 184 Gao, B., 11, 12, 38 Gao, X. D., 369, 378, 388 Gao, Y., 75, 76, 82, 93, 96, 100, 348, 378, 381 Garabedian, M. J., 7, 41, 173, 182 Garcia, P. D., 224, 227, 240 Garcia-Echeverria, C., 257, 276 Garc´ıa-Gardena, G., 7, 38 Gardner, N., 142, 153 Gardner, R. G., 208, 216, 355, 360, 376, 382 Garel, J.-R., 246, 250, 276, 279 Garland, D., 143, 148

444

AUTHOR INDEX

Garman, E. F., 324, 340 Garrido, C., 115, 143, 144, 148, 150 Gasch, A. P., 352, 364, 381 Gaskins, H. R., 189, 190, 214 G¨assler, C. S., 15, 16, 17, 18, 19, 20, 27, 34, 37, 38 Gassner, C., 291, 292, 299 Gatenby, A. A., 74, 100, 102, 347, 382, 389 Gaume, B., 235, 236, 240 Gautschi, M., 9, 38 Gay, N. J., 207, 221 Gaynor, E. C., 359, 381 Ge, B., 160, 181 Gebauer, M., 32, 34, 38, 44 Gehrig, P., 60, 71 Gehring, H., 9, 10, 25, 42 Gehring, U., 32, 34, 38, 44 Gehring, W., 115, 151 Geier, E., 202, 215 Geiger, B., 127, 139, 153 Geiser, J. R., 96, 101 Geissler, A., 236, 239 Geissler, S., 88, 100 Geleziunas, R., 204, 217 Geli, M. I., 359, 382 Gelman, M. S., 322, 339 Gemmecker, G., 256, 278 Gemmill, T. R., 306, 331, 334 Genevaux, P., 28, 38 Genevi`ere, M., 108, 154 Genschik, P., 190, 219 Gentzsch, M., 359, 382 Geoghagen, N. S. M., 202, 214 Georgiou, G., 295, 300 Georgopoulos, C., 3, 4, 12, 13, 14, 15, 17, 18, 28, 29, 30, 31, 37, 38, 39, 40, 41, 43, 44, 74, 100, 157, 185, 233, 240, 267, 281, 285, 290, 292, 301, 347, 348, 381, 382, 414, 416, 427, 428, 429 Gerards, W. L., 198, 215 Gerisch, G., 190, 220 Gerlinger, U. M., 211, 216 Germain, R., 330, 344 Germeroth, L., 3, 22, 42 Gernold, M.-J., 106, 150 Gervasoni, P., 60, 71 Gething, M.-J., 2, 37, 106, 127, 150, 229, 230, 240, 241, 346, 347, 355, 364, 382, 385 Getz, G., 371, 373, 382

Geuskens, V., 419, 428 Geuze, H. J., 330, 342, 355, 389 Ghadge, G. D., 272, 278 Ghazal, P., 212, 219 Ghislain, M., 203, 207, 216 Ghosh, D., 132, 154 Giannoukos, G., 179, 181, 185 Gibbons, D. L., 53, 71 Gibbs, J., 203, 220 Gibson, R., 328, 341 Giddings, T. H., 176, 184 Gierasch, L. M., 3, 22, 23, 26, 27, 40, 41, 57, 58, 59, 71, 72 Gijsen, M. L. J., 112, 122, 155 Gilbert, H. F., 112, 123, 128, 149, 295, 300 Gilmore, R., 359, 387 Gimeno, R. E., 359, 360, 382, 386 Gingeras, T. R., 365, 384 Gingras-Breton, G., 139, 152 Ginsburg, A., 416, 429 Ginsburg, D. B., 198, 218 Girod, P. A., 208, 216 Girshovich, A. S., 74, 99 Gisler, S. M., 9, 12, 14, 25, 38, 41 Gitler, C., 171, 179, 185 Glas, R., 189, 202, 221 Glaser, R. L., 115, 150 Glasner, J. D., 349, 351, 386 Glatz, A., 125, 127, 130, 135, 144, 145, 150, 155 Glick, B. S., 227, 230, 231, 233, 235, 236, 240, 242, 267, 278 Glickman, M. H., 188, 192, 204, 205, 207, 208, 209, 214, 215, 216, 219, 221 Glockshuber, R., 287, 288, 289, 300, 301 Glodek, A., 202, 214 Glover, J. R., 3, 5, 6, 38, 349, 350, 377, 382, 387, 413, 414, 416, 427, 429 Gocayne, J. D., 202, 214 Goeckeler, J., 237, 239 Goffeau, A., 349, 382 Gogol, E. P., 207, 213 Golbik, R., 197, 203, 216, 222, 262, 276 Gold, B. G., 271, 276 Goldberg, A. L., 187, 188, 189, 190, 193, 199, 200, 201, 202, 203, 204, 205, 206, 213, 214, 215, 217, 218, 219, 221, 222, 414, 415, 416, 417, 427, 428, 429 Goldberger, R. F., 284, 300 Goldman, J. E., 115, 140, 141, 148, 150, 156

AUTHOR INDEX

Goloubinoff, P., 3, 4, 6, 38, 39, 41, 43, 74, 100, 104, 112, 125, 128, 130, 132, 135, 139, 144, 145, 153, 155, 347, 350, 381, 382, 389, 414, 416, 427, 428 Golub, T. R., 371, 373, 374, 375, 380, 388 Gomez, L., 128, 142, 149, 154 Gonciarz-Swiatek, M., 416, 427 Gong, Y. D., 125, 132, 156 Gonzalez, M., 419, 421, 427 Gonzalez, T. N., 355, 364, 382 Goodwill, K. E., 189, 218 Goody, R. S., 10, 41 Gopal, P., 304, 334 Gorbea, C., 187, 203, 207, 208, 211, 214, 216, 219 Gordon, A. E., 371, 382 Gordon, B., 113, 123, 128, 134, 152, 208, 209, 220 Gordon, C., 7, 38, 188, 207, 208, 209, 215, 216, 220, 221 Gordon, H. S., 271, 276 Gorman, M. A., 203, 222, 418, 429 Gorovits, B. M., 53, 71 Goruppi, S., 32, 37 G¨othel, S. F., 244, 267, 276 Goto, S., 114, 126, 151 Gottesman, M., 2, 4, 15, 20, 21, 25, 39, 42, 44, 425, 429 Gottesman, S., 127, 151, 188, 189, 206, 217, 221, 349, 350, 382, 389, 413, 414, 415, 416, 417, 419, 420, 422, 423, 427, 429 Gottlicher, M., 178, 183, 185 G¨ottlinger, H. G., 270, 281 Gowen, B., 53, 69, 72, 203, 222, 418, 429 Gr¨aber, S., 4, 38, 130, 135, 144, 149 Graﬂ, R., 250, 280 Gragerov, A., 2, 4, 20, 21, 25, 39, 44 Grallert, H., 61, 71, 128, 135, 148 Graml, W., 188, 202, 203, 219 Gramm, C. F., 203, 214, 219 Grammatikakis, A., 173, 177, 181 Grammatikakis, N., 161, 173, 177, 181, 184 Grandison, P. M., 116, 148 Grantcharova, V., 45, 71 Grantham, J., 78, 86, 87, 98, 102, 103 Grathwohl, C., 245, 276 Grauschopf, U., 288, 300 Graves, B., 416, 429 Gray, C. W., 212, 216

445

Gray, P. W., 321, 337 Green, C. J., 244, 272, 272 Green, M., 158, 184, 320, 337 Green, N., 359, 385 Green, R., 330, 343 Greene, L. E., 3, 4, 11, 12, 15, 28, 35, 37, 38, 39, 42, 43 Greene, M. K., 15, 29, 39 Greene, W. C., 204, 217 Greenleaf, A. L., 272, 279 Gregersen, N., 414, 415, 426 Grein, K., 83, 90, 92, 103 Greller, L. D., 108, 114, 143, 151 Grenert, J. P., 164, 165, 171, 175, 181 Grenier, L., 199, 207, 213, 218 Griest, T. A., 122, 128, 148 Grimaud, R., 416, 417, 419, 422, 423, 427, 428, 429 Grimm, R., 201, 221 Grimme, S. J., 359, 381 Grimsley, G. R., 251, 279 Grinna, L. S., 310, 331, 335 Grishchuk, E. L., 96, 100 Grivell, L. A., 231, 240 Groemping, Y., 6, 16, 39, 42 Groenen, P. J. T. A., 106, 114, 125, 128, 130, 135, 150, 153 Groettrup, M., 198, 200, 201, 212, 215, 216, 220 Groger, A., 418, 426 Gr¨oger, A., 203, 214 Groll, M., 189, 190, 191, 192, 193, 196, 197, 198, 199, 200, 207, 214, 215, 216, 217, 219, 220, 349, 379, 418, 426, 427 Grondin, B., 359, 380 Gronenborn, A. M., 290, 301 Groner, B., 171, 179, 185 Groome, A., 143, 148 Grootegoed, J. A., 178, 185 Groß, M., 66, 67, 71 Gross, C. A., 15, 42, 259, 268, 279, 282, 350, 382 Gross, M., 35, 39 Grosse, M., 59, 71 Groth-Vasselli, B., 132, 153 Gruhler, A., 199, 216, 217, 227, 242 Grundstr¨om, T., 255, 275 Grunert, H. P., 251, 253, 277 Gruters, R. A., 329, 342 Grziwa, A., 190, 222

446

AUTHOR INDEX

Gu, J. R., 143, 150 Gu, Y. Z., 178, 181 Guargliardi, A., 80, 100 Gu¨ ckel, R., 199, 211, 215, 216 Guddat, L. W., 288, 289, 300 Guerreiro, N., 134, 148 Guerriero, V., Jr., 36, 42 Guevara, M.-A., 142, 154 Guex, N., 39 Guiard, B., 224, 226, 227, 230, 231, 235, 236, 237, 238, 239, 240, 241, 242 Guicheney, P., 115, 140, 155 Guilhot, C., 290, 300 Guillaume, G., 254, 281 Guillet, D., 126, 153 Guimond, A., 121, 126, 139, 150, 152 Gunsalus, K. C., 353, 385 Guo, D., 158, 185 Guo, Q. X., 269, 280 Gupta, D., 315, 336 Gupta, M., 127, 151 Gupta, R. S., 74, 100, 104 Gurbuxani, S., 115, 143, 144, 148 Gustafsson, J. A., 157, 178, 182, 183, 185 Guthrie, B., 264, 265, 276, 278 Gutsche, I., 76, 78, 88, 100

H Ha, J.-H., 11, 12, 16, 21, 22, 26, 39 Ha, J. S., 416, 420, 422, 428 Haandrikman, A. J., 259, 276 Haas, A. L., 205, 208, 214 Haas, I. G., 238, 242, 320, 337 Haber, E., 283, 299 Hacker, J., 259, 260, 276, 278, 279, 282 Hackett, R. L., 272, 282 Haebel, P. W., 286, 293, 295, 301 Hagerman, P. J., 250, 276 Hahn, U., 251, 252, 253, 277, 279 Hahnel, R., 178, 183 Hakala, K., 204, 220 Haley, D. A., 112, 116, 122, 123, 141, 148, 150 Hall, D. R., 258, 279 Hall, J., 359, 385 Hall, J. G., 255, 277 Hallberg, E. M., 74, 99, 346, 347, 377, 378, 380

Hallberg, R. L., 74, 99, 346, 347, 377, 378, 380 Halperin, T., 414, 427 Halsall, D. J., 53, 71 Halvorson, H. R., 247, 250, 275 Hamada, M., 171, 181 Hamamoto, S., 359, 381 Hamburger, D., 359, 382 Hamilton, G. S., 270, 280 Hamilton, S. L., 269, 280 Hamman, B. D., 238, 240 Hammann, A., 115, 143, 150 Hammerling, G. J., 326, 341 Hammond, C., 322, 325, 329, 338, 340 Hampton, R. Y., 208, 216, 355, 360, 376, 382 Han, Y., 122, 124, 148 Handford, P. A., 255, 282 Handschumacher, R. E., 161, 173, 181, 183, 257, 261, 276, 277, 279 Hane, S. D., 259, 271, 276 Hanein, D., 223, 240 Haner, M., 198, 215 Hanes, S. D., 259, 271, 272, 272, 278, 282 Hani, J., 258, 271, 276 Hankinson, O., 178, 181 Hannavy, K., 230, 240 Hansen, T. H., 324, 328, 340, 341 Hansen, W. J., 3, 39, 91, 92, 99, 100, 224, 227, 240, 359, 379 Hanson, P. I., 418, 429 Hao, Y., 34, 44 Hara, M., 171, 184 Haracska, L., 208, 209, 216 Harding, J. J., 106, 122, 134, 149 Harding, M. W., 257, 258, 276, 354, 386 Hardwick, K. G., 359, 387 Hardy, S. J. S., 124, 155 Harfull, G., 255, 276 H¨arndahl, U., 142, 145, 150 Harper, J. W., 163, 173, 177, 185 Harrington, W. F., 246, 281 Harris, G. L., 260, 281 Harris, J. R., 211, 219 Harris, M. R., 324, 340 Harris, S., 29, 42 Harrison, C. J., 18, 19, 20, 39 Harrison, R. K., 259, 260, 276 Harrison, S. C., 122, 152, 154 Harrison-Lavoie, K., 84, 102 Hartigan, J., 371, 382

AUTHOR INDEX

Hartl, F. U., 3, 4, 5, 15, 17, 18, 19, 20, 30, 34, 35, 38, 39, 40, 42, 43, 61, 66, 67, 70, 71, 74, 75, 76, 82, 83, 89, 90, 92, 99, 100, 101, 102, 103, 104, 139, 150, 159, 160, 161, 162, 163, 164, 165, 170, 171, 173, 174, 175, 176, 183, 184, 185, 235, 236, 241, 267, 281, 284, 300, 312, 327, 329, 336, 346, 347, 348, 349, 377, 378, 380, 381, 383, 385, 386, 388 Hartling, J. A., 417, 418, 422, 429 Hartmann, C., 189, 214, 220, 349, 379, 417, 418, 426, 429 Hartmann, E., 224, 241, 359, 381, 385, 386 Hartmann-Petersen, R., 207, 211, 216 Hartskeerl, R. A., 116, 155 Hartson, S., 7, 8, 43 Hartson, S. D., 7, 39, 171, 177, 181 Harwood, J. L., 144, 145, 155 Hasegawa, K., 114, 126, 151 Haslbeck, M., 106, 123, 125, 130, 132, 144, 150 Hastie, N. D., 207, 216 Hastings, R. A., 207, 216 Hasumi, H., 262, 278 Hata, M., 29, 41 Hatayama, T., 9, 44 Haugejorden, S. M., 320, 337 Hauser, N., 331, 344 Hauser, S., 265, 277 Hawkins, A. R., 254, 282 Hay, N., 52, 53, 71 Hayano, T., 248, 257, 258, 262, 264, 268, 277, 278, 281 Hayashi, S., 204, 218 Hayashi, Y., 108, 144, 155 Hayer, M. K., 5, 17, 40 Hayer-Hartl, M., 18, 19, 20, 39 Hayes, N. V., 78, 103 Hayess, K., 139, 148 Hays, L. G., 132, 153 Haystead, T. A. J., 164, 165, 171, 181 He, Y. D., 366, 371, 372, 377, 383 Head, M. W., 141, 150 Heads, R., 244, 272, 272 Hebert, C., 158, 181 Hebert, D. N., 322, 324, 326, 328, 329, 338, 340 Heckathorn, S. A., 142, 145, 149, 150 Hegerl, R., 201, 205, 211, 217, 220

447

Heikkila, J. J., 115, 128, 142, 149, 150, 152, 153 Heim, N., 238, 242 Heinemann, U., 251, 278 Heinemeyer, W., 188, 193, 196, 197, 198, 199, 200, 215, 216, 217, 219 Heitman, J., 173, 174, 181, 244, 257, 261, 270, 272, 272, 275, 277, 282 Helenius, A., 230, 241, 305, 310, 315, 318, 319, 320, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 334, 335, 336, 337, 338, 339, 340, 341, 354, 355, 360, 367, 371, 372, 376, 381, 382, 383 Helinski, D. R., 416, 428 Heller, R., 348, 349, 387 Hellman, U., 310, 311, 312, 336 Hellmuth, K., 360, 361, 378, 384 Helm, K. W., 114, 150 Hemmingsen, S. M., 74, 100, 347, 382 Hendershot, L. M., 238, 240 Henderson, S. J., 48, 72 Hendil, K. B., 190, 207, 208, 211, 212, 216, 218, 220, 221 Hendle, J., 121, 149 Hendrick, J. P., 3, 39, 61, 71, 312, 327, 329, 336 Hendrickson, W. A., 20, 21, 25, 44 Hendriks, I. L., 198, 215 Hendrix, R. W., 74, 100, 347, 382 Hengge-Aronis, R., 424, 425, 426, 428 Henke, W., 205, 209, 216 Henklein, P., 193, 196, 197, 198, 220 Hennecke, H., 297, 300 Hennecke, S., 359, 380 Hennessy, F., 29, 39 Henninger, H., 231, 233, 242 Henriques, A. O., 115, 116, 150 Henzel, W. J., 7, 42 Hepburne-Scott, H. W., 134, 149 Herbertsson, H., 249, 276 Herde, P., 6, 42 Herman, C., 420, 427 Hernandez, L. D., 115, 155 Herrmann, C., 16, 39 Herrmann, J., 2, 40 Herrmann, J. M., 224, 231, 240 Herschlag, D., 351, 384 Herscovics, A., 306, 334, 359, 380 Hersh, L. B., 201, 220 Hershko, A., 187, 203, 204, 206, 214, 215, 216

448

AUTHOR INDEX

Herskowitz, I., 374, 380 Herzberg, O., 254, 277 Hessefort, S., 35, 39 Hesterkamp, T., 4, 5, 18, 37, 83, 99, 265, 277 Hettema, E. H., 239, 240 Heusch, M., 204, 217 Heuser, J., 127, 149 Heysen, A., 359, 380 Hicke, L., 204, 217 Hickey, E., 115, 126, 127, 139, 152, 154 Hickey, M. J., 143, 156 Hiddinga, H. J., 144, 151 Higgins, K. A., 255, 277 High, S., 265, 281, 326, 340, 341 Higuchi, S., 202, 217 Hill, C. P., 192, 207, 212, 213, 217, 221, 222, 270, 281, 282 Hill, K., 223, 240, 311, 312, 314, 336 Hiller, M. M., 360, 382 Hilt, W., 199, 211, 215, 216, 217 Himmel, K. L., 84, 104 Hinck, A. P., 255, 272 Hirai, H., 320, 337 Hirano, H., 189, 198, 217, 220 Hirano, N., 320, 337 Hirata, A., 359, 387 Hirata, D., 96, 100, 103 Hirosaki, K., 322, 339 Hirosawa, S., 330, 343 Hirose, T., 114, 154 Hirschberg, C., 310, 335 Hirschﬁeld, I. N., 61, 71 Hisamatsu, H., 213, 213 Hisamatsu, Y., 269, 278 Hitotsumatsu, T., 115, 150 Hixson, J. D., 53, 71 Hiyama, T., 128, 154 Hizuta, A., 158, 182 Hlaing, J., 178, 183 Ho, M. M., 365, 389 Hobbs, M., 414, 427 Hochstenbach, F., 322, 338 Hochstrasser, M., 190, 193, 196, 198, 199, 206, 208, 213, 214, 215, 219, 360, 386 H¨ockendorf, J., 199, 214 H¨ockendorff, J., 199, 219 Hodel, A., 265, 277 Hoenders, H. J., 124, 155 Hoff, K. G., 2, 17, 28, 39, 42, 43 Hoffman, K., 161, 173, 183

Hoffman, L., 205, 206, 217, 220 Hoffmann, K., 173, 181, 261, 279 Hoffmann, M., 211, 216 Hofmann, K., 208, 209, 217, 218 Hofmann, S., 225, 239 Hoheisel, J. D., 349, 382 H¨ohfeld, J., 7, 18, 19, 20, 32, 34, 35, 36, 38, 39, 40, 42, 206, 214, 218, 349, 383 Hohl, C. M., 349, 389 Hohn, B., 74, 101 Hohn, T., 74, 101 Holbrook, J. J., 53, 71 Holcomb, C. L., 359, 379 Holland, I. B., 30, 38 Holmes, F. E., 78, 103 Holmes, K. C., 10, 39 Holmes, S., 229, 239 Holmgren, A., 351, 386 Holst, B., 304, 334 Holstein, S. E., 26, 28, 43 Holter, N. S., 371, 373, 374, 383 Holzhu¨ tter, H. G., 200, 201, 220 Holzman, T. F., 264, 281 Hom-Booher, N., 9, 41 Homuth, G., 2, 40 Hong, E., 332, 343, 363, 383 Hong, J. S., 330, 342 Hong, S.-W., 133, 150 Hongenesch, J. B., 178, 181 Honig, B., 41 H¨onlinger, A., 28, 41, 224, 241 Honore, B., 173, 175, 181 Hood, W. F., 255, 276 Hoover-Litty, H., 328, 341 Hopkins, D. A., 114, 154 Hoppe, T., 203, 204, 217, 218 Horazdovsky, B. F., 360, 386 Horenstein, C. I., 359, 385 Hori, Y., 322, 339 Horne, S. M., 261, 268, 277, 279 Horovitz, A., 48, 52, 53, 71, 72 Horowitz, P. M., 53, 57, 58, 71 Horst, M., 230, 231, 240, 241 Horton, H., 365, 384 Horvath, I., 127, 145, 148, 150 Horwich, A. L., 4, 5, 14, 37, 40, 44, 45, 46, 49, 50, 53, 54, 55, 56, 59, 61, 62, 63, 64, 65, 68, 69, 70, 71, 72, 73, 74, 75, 77, 79, 80, 81, 82, 83, 99, 101, 102, 104, 127, 148, 188, 217, 224, 242, 346, 347, 348, 349,

449

AUTHOR INDEX

350, 373, 377, 378, 379, 380, 383, 384, 385, 386, 387, 388, 389, 390, 413, 422, 423, 424, 427, 428, 429 Horwitz, J., 106, 112, 114, 119, 123, 124, 128, 130, 134, 135, 141, 148, 149, 150, 153, 154 Hoskins, J. R., 4, 44, 349, 350, 383, 386, 389, 415, 416, 419, 420, 422, 423, 424, 425, 427, 429 Hosobuchi, M., 359, 381 Hostein, I., 161, 171, 177, 181 Hottenrott, S., 261, 277 H¨otzl, H., 188, 204, 206, 209, 217 Hough, R., 203, 217 Houle, F., 139, 150 Houry, W. A., 3, 43, 61, 71, 139, 150, 250, 251, 267, 277, 280, 281 Houseweart, M., 270, 281 Howell, P. L., 306, 334 Howells, A. J., 163, 172, 180 Howson, R. W., 360, 384 Hoyt, M. A., 96, 101, 103 Hrdy, I., 128, 132, 154 Hruska, K. A., Jr., 158, 184 Hu, W., 23, 26, 41 Hu, Z., 36, 37 Hu, Z. Y., 207, 211, 222 Huang, C. F., 359, 383 Huang, G. C., 266, 277 Huang, H. C., 189, 219, 414, 416, 417, 428 Huang, L., 198, 213, 221 Huang, Q. L., 112, 116, 122, 123, 128, 134, 141, 148, 150 Huang, S. F., 125, 132, 156, 237, 240 Huang, W., 7, 39, 171, 177, 181, 204, 209, 219 Huang, X. C., 365, 380 Huang, Y. J., 93, 94, 96, 103, 353, 385 Hubbard, S. C., 304, 333 Hubbell, E., 365, 380 Huber, H., 49, 71, 75, 88, 99, 349, 381 Huber, R., 49, 71, 75, 88, 99, 189, 190, 191, 192, 193, 196, 197, 198, 199, 200, 207, 214, 215, 216, 217, 218, 219, 220, 349, 379, 381, 417, 418, 426, 427, 428, 429 Huber-Wunderlich, M., 288, 289, 300 H¨ubner, D., 257, 277 Hubner, S., 7, 32, 34, 38, 42 H¨ubscher, U., 4, 44 Hudson, J., Jr., 374, 383 Hudson, T. J., 160, 181

Huffaker, T. C., 79, 83, 84, 99, 359, 387 Hughes, A. L., 189, 190, 217 Hughes, E. A., 326, 341 Hughes, T. R., 366, 371, 372, 377, 383 Huisman, H. G., 329, 342 Hultgren, S. J., 286, 300 Humphreys, D. P., 285, 301 Hung, S. H. Y., 258, 280 Hunt, T., 126, 154, 160, 185 Hunter, A. S., 48, 54, 71, 349, 380 Hunter, T., 7, 39, 177, 182, 259, 271, 272, 277, 278, 279, 281 Hunter, W. N., 258, 279 Huot, J., 139, 150 Hurley, J. H., 10, 39 Hurtley, S. M., 320, 328, 337, 341, 354, 383 Huse, M., 269, 277 Hutcheson, A. M., 140, 141, 149, 153 Hutchinson, J. P., 57, 58, 59, 71 Hutchinson, K. A., 157, 161, 173, 183, 185 Hutchison, K. A., 7, 39, 178, 179, 182, 261, 279 Huth, J. R., 290, 301 Hwang, B. J., 415, 427 Hwang, D., 203, 219 Hwang, S. T., 227, 229, 240, 241 Hynes, G., 74, 76, 77, 79, 84, 86, 87, 98, 101, 102, 348, 378, 384

I Ichihara, A., 211, 215, 221 Ifeanyi, F., 145, 150 Iguala, J. C., 360, 383 Ihara, Y., 328, 341 Iida, T., 80, 100 Iimura, Y., 359, 384 Ikai, A., 211, 221 Ikawa, M., 329, 342 Ikuina, Y., 172, 184 Ikura, M., 159, 185 Ikura, T., 248, 262, 264, 277, 278 Im, Y., 418, 429 Imai, S., 322, 324, 338, 339 Imamoto, N., 228, 242 Immler, D., 126, 148 Inaguma, Y., 108, 114, 115, 126, 127, 144, 150, 151, 153 Inbar, E., 71

450

AUTHOR INDEX

Ingelﬁnger, D., 258, 279 Inge-Vechtomov, S. G., 350, 380 Ingley, E., 161, 180 Ingui, C., 163, 171, 174, 184 Inokuchi, H., 291, 292, 301 Inoue, T., 189, 220 Inouye, M., 159, 185 Inze, D., 259, 272, 278 Irie, K., 171, 184 Irie, S., 32, 34, 43 Irwin, A. D., 7, 39 Iseki, S., 114, 156 Ishibashi, T., 212, 217 Ishii, T., 158, 182 Ishikawa, T., 189, 217, 418, 421, 427, 428 Ishima, R., 159, 185 Ishiura, S., 108, 144, 155 Isobe, T., 126, 150 Ito, H., 126, 127, 150, 151 Ito, K., 290, 291, 292, 294, 300, 301 Ito, M., 258, 278 Itzhaki, H., 414, 427 Ivatt, R. J., 304, 333 Ivessa, N. E., 330, 343 Iwaki, A., 115, 150 Iwaki, T., 115, 150 Iwakura, M., 422, 424, 428 Iwamoto, I., 126, 127, 150, 151 Iyer, V. R., 364, 374, 383, 388

J J¨aa¨ ttel¨a, M., 7, 42 Jackiw, V. H., 178, 180, 181 Jackson, A. P., 163, 172, 180 Jackson, C. L., 359, 383 Jackson, G. S., 53, 71 Jackson, M. R., 322, 323, 324, 326, 328, 338, 339 Jackson, R. J., 160, 185 Jackson, S. A., 115, 150 Jackson, S. E., 264, 277 Jacob-Dubuisson, F., 286, 300 Jacobs, M., 259, 277 Jacq, A., 30, 38 Jacq, C., 349, 382 Jaeger, J., 59, 71 Jaenicke, R., 106, 112, 122, 150, 155, 170, 184, 197, 216

J¨ager, M., 259, 277 J¨ager, S., 196, 198, 199, 216, 217 Jahn, R., 418, 429 Jakana, J., 112, 123, 128, 149 Jakob, C. A., 307, 331, 335, 343 Jakob, M., 270, 280 Jakob, U., 106, 126, 128, 130, 132, 135, 148, 150, 151, 157, 163, 164, 176, 182, 185, 296, 301, 351, 352, 379, 383 James, M. N., 254, 281 James, P., 60, 71 Jamsa, E., 224, 241 Jander, G., 290, 291, 300 Jannatipour, M., 310, 311, 312, 336 Janowski, B., 257, 277 Jap, B., 190, 191, 192, 193, 200, 218, 418, 428 Jarosch, E., 361, 381 J¨aschke, A., 272, 277 Jascur, T., 231, 240 Jayaraman, T., 269, 278 Jelinek-Kelly, S., 306, 334 Jelinsky, S. A., 204, 209, 217 Jenness, D. D., 360, 384 Jen¨o, P., 231, 240, 241 Jensen, C., 211, 215 Jensen, C. C., 213, 222 Jensen, O. N., 326, 341 Jensen, R. E., 231, 240 Jensen, S. E., 254, 281 Jentsch, S., 19, 34, 39, 203, 204, 217, 218 Jespersgaard, C., 414, 415, 426 Jiang, H., 213, 217 Jiang, J., 36, 38 Jiang, J. H., 206, 214 Jiang, Y., 359, 387 Jimbow, K., 322, 339 Jimenez, J., 176, 182 Jin, H., 349, 351, 386 Jin, Z. J., 115, 142, 143, 152 Jinn, T.-L., 114, 128, 144, 151 Joab, I., 157, 178, 182 Joachimiak, A., 48, 72 Joanisse, D. R., 115, 144, 151 Jobson, J., 371, 383 Jocobs, M. D., 265, 277 Johansson, A., 249, 276 Johns, K., 254, 281 Johnson, A. B., 141, 148 Johnson, A. E., 80, 92, 102, 224, 231, 238, 240, 242

AUTHOR INDEX

Johnson, A. L., 360, 383 Johnson, B. D., 175, 181 Johnson, C., 4, 12, 40 Johnson, E. R., 12, 22, 26, 39 Johnson, E. S., 199, 203, 209, 217, 219 Johnson, J. L., 7, 39, 167, 174, 178, 182 Johnson, L. N., 177, 182 Johnston, L. H., 360, 383 Johnston, M., 349, 382 Johnston, S. A., 204, 207, 209, 219, 415, 429 Johnston, S. C., 212, 217 Jonasson, P., 249, 276 Jonda, S., 289, 300 Jones, A. R., 366, 371, 372, 377, 383 Jones, J. M., 259, 276, 415, 421, 427, 429 Jones, T. R., 330, 342, 355, 389 Jonsen, M., 416, 429 Jonsson, B. H., 249, 262, 276 Joo, B., 127, 145, 150 Jordan, R., 3, 10, 11, 12, 16, 39, 40, 42 Jordano, J., 115, 147, 148, 154 J¨orgensen, M. U., 332, 343 Jove, R., 157, 163, 185 Judson, I., 161, 171, 177, 181 Julie, M., 108, 154 Jung, K., 425, 426 Jungblut, P., 205, 209, 216 Jungmann, J., 359, 383 Jungnickel, B., 30, 42, 223, 240 Jung-Testas, I., 178, 180 Junop, M., 159, 168, 180

K Kabani, M., 18, 39 Kabsch, W., 10, 39 Kaczowska, S. J., 188, 189, 218 Kad, N. M., 52, 53, 71 Kagawa, S., 208, 221 Kai, M., 322, 324, 338, 339 Kainosho, M., 159, 185 Kairies, N., 197, 200, 216, 220 Kaiser, C. A., 299, 300, 332, 343, 359, 360, 363, 377, 381, 382, 383, 386 Kakeda, M., 28, 43 Kalbacher, H., 201, 220 Kalies, K.-U., 223, 240 Kallen, J., 258, 277, 279, 280 Kalousek, F., 74, 99, 346, 347, 377, 378, 380

451

Kalton, H. M., 174, 182, 261, 278 Kamei, K., 126, 127, 150, 151 Kameyama, K., 211, 221 Kampinga, H. H., 4, 32, 40, 41, 127, 151 Kampranis, S. C., 170, 182 Kamtekar, S., 418, 429 Kanaya, E., 290, 300 Kanazawa, M., 28, 43, 228, 242 Kandil, E., 212, 217 Kane, T., 360, 384 Kanehori, K., 202, 217 Kaneko, Y., 9, 38 Kanelakis, K. C., 7, 32, 35, 37, 39, 41, 179, 182 Kanemori, M., 266, 279 Kang, P. J., 229, 230, 236, 239, 240, 241 Kania, M., 211, 221 Kanoh, H., 322, 324, 338, 339 Kanon, B., 4, 40 Kantorow, M., 125, 151 Kao, C. M., 352, 364, 381 Kapelari, B., 188, 204, 205, 206, 209, 211, 217 Kapp´e, G., 107, 108, 115, 151 Kapteyn, J. C., 359, 385 Kapur, A., 134, 152 Karlsson, G. B., 329, 342 Karlsson, R., 98, 103 Karplus, M., 258, 279 Karplus, P. A., 258, 259, 281 Karpov, V., 209, 218 Karzai, A. W., 12, 14, 39, 202, 217, 420, 427 Kasahara, M., 212, 213, 217, 218, 220 Kashi, Y., 54, 55, 56, 71, 84, 102, 349, 389 Kasi, V. S., 144, 151 Katakura, Y., 377, 385 Katayama, Y., 415, 423, 427 Katayama-Fujimura, Y., 127, 151 Katchalski, E., 246, 281 Kato, H., 115, 151, 320, 337 Kato, J., 359, 387 Kato, K., 108, 114, 115, 126, 127, 150, 151, 153 Kato, S., 26, 43, 208, 220, 268, 277 Katoh, S., 369, 388 Katzen, F., 297, 300, 301 Kaufman, R. J., 310, 311, 312, 314, 322, 331, 336, 339 Kaur, H., 111, 130, 133, 154 Kautz, R. A., 265, 277 Kawahara, H., 213, 218 Kawahara, T., 355, 359, 365, 383, 384, 385

452

AUTHOR INDEX

Kawamoto, T., 202, 217 Kawamura-Watabe, A., 26, 43 Kawashima, T., 202, 217 Kawashima-Ohya, Y., 202, 217 Kazim, L., 138, 155 Kazlauskas, A., 174, 182 Ke, H. M., 258, 270, 277, 282 Kearse, K. P., 322, 324, 328, 329, 330, 339, 340, 341 Kedersha, N. L., 127, 151 Kehl, M., 190, 220 Kehren, V., 76, 103 Keiholz, W., 200, 219 Keiler, K. C., 202, 217, 419, 420, 427 Keilholz, W., 199, 200, 215 Keller, S. H., 322, 339 Kellermann, J., 76, 104, 188, 204, 206, 209, 217 Kelley, W. L., 28, 29, 30, 38, 39, 40, 416, 427 Kellis, J. T., 262, 278 Kelly, A., 198, 214 Kelly, R. M., 189, 214 Kendall, D. A., 265, 281 Kennedy, W. M., 290, 301 Kenworthy, A. K., 355, 385 Kepes, F., 359, 379, 383 Kerlavage, A. R., 202, 214 Kern, D., 249, 257, 262, 265, 277 Kern, G., 249, 257, 262, 264, 265, 277, 280 Kerstein, M., 119, 153 Kessel, M., 189, 211, 214, 217, 416, 417, 421, 426, 427, 429 Kessler, B. M., 189, 202, 221 Kessler, H., 203, 214, 418, 426 Kester, K., 111, 130, 133, 154 Khalawan, S. A., 145, 149 Khan, S., 212, 216 Khandijian, E. W., 115, 155 Kharbanda, S., 115, 143, 153 Khurseed, B., 7, 41 Kidd, M. J., 366, 371, 372, 377, 383 Kieffer, L. J., 261, 277 Kiefhaber, T., 251, 252, 253, 257, 276, 277 Kielland-Brandt, M. C., 304, 334 Kikuchi, M., 290, 300 Kim, B., 416, 417, 427 Kim, K. B., 197, 201, 216, 218 Kim, K. I., 416, 420, 422, 428 Kim, K. K., 107, 112, 116, 128, 151 Kim, L., 2, 40

Kim, P. S., 246, 250, 277 Kim, R., 107, 112, 116, 128, 151 Kim, S., 4, 40, 73, 77, 101, 348, 384 Kim, S. H., 107, 112, 116, 128, 151 Kim, S. Y., 420, 427 Kim, Y. I., 349, 384, 418, 428 Kimura, T., 359, 384 Kimura, Y., 198, 217 Kindle, C. S., 324, 340 King, A. M., 366, 371, 372, 377, 383 King, J., 14, 31, 37 King, S. A., 330, 342 Kinoshita, T., 359, 389 Kipper, J., 199, 215 Kirchhausen, T., 62, 63, 64, 71 Kirkness, E. F., 202, 214 Kirschner, M., 187, 204, 217 Kirschner, M. W., 74, 93, 99, 102, 127, 136, 149, 151, 260, 271, 280, 282 Kishigami, S., 290, 291, 292, 300, 301 Kishikawa, M., 114, 154 Kishore, V., 256, 257, 276, 278 Kisselev, A. F., 193, 199, 200, 213, 217 Kitagawa, M., 414, 428 Kitajima, K., 330, 342 Kitamoto, N., 359, 384 Kitareewan, S., 371, 373, 374, 375, 388 Kito, Y., 359, 384 Kittmar, K. D., 7, 39 Kitzm¨uller, C., 330, 343 Klamp, T., 199, 214 Klauck, E., 425, 426 Klaus, C., 231, 235, 236, 239, 240 Klein, C., 115, 153 Klein, H. L., 88, 89, 90, 91, 92, 104 Klein, J., 212, 218 Klein, M., 330, 343 Kleinschmidt, J. A., 203, 206, 211, 219 Kleinz, J., 197, 222 Klemenz, R., 114, 115, 151 Klenk, H.-P., 202, 222 Klig, L. S., 359, 388 Klinefelter, G. R., 271, 282 Klis, F. M., 359, 385 Kloetzel, P.-M., 106, 150, 188, 193, 196, 197, 198, 199, 204, 212, 214, 215, 216, 219, 220, 221 Klostermeier, D., 10, 11, 12, 14, 16, 17, 18, 19, 20, 27, 34, 37, 39, 40 Klumpp, M., 75, 88, 101, 349, 384

AUTHOR INDEX

Klunder, J. M., 207, 213 Klunker, D., 89, 101 Knauf, U., 106, 126, 150, 151 Knighton, D. R., 182 Knipfer, N., 188, 217 Knoblauch, N. T. M., 4, 40 Knoblauch, R., 173, 182 Knop, M., 331, 344, 360, 361, 378, 384 Knott, V., 255, 282 Knowlton, J. R., 192, 207, 212, 213, 217, 221, 222 Kobayashi, G. S., 145, 148 Kobayashi, M., 365, 384 Kobayashi, N., 57, 71 Kobayashi, R., 157, 181 Kobayashi, T., 291, 292, 300, 301 Koch, B. D., 227, 228, 239 Kochel, K., 32, 34, 43 Kocsis, E., 189, 211, 214, 217, 416, 417, 426, 427 Koegel, M., 203, 218 Koerkamp, M. G., 239, 240 Kofron, J. L., 256, 257, 276, 278 Kogure, K., 115, 151 Kohane, I. S., 371, 375, 380 Kohda, K., 212, 217 Kohler, A., 192, 207, 216 Kohno, K., 229, 241 Kohonen, T., 371, 384 Kohrer, K., 360, 384 Kohtz, D. S., 170, 178, 184 Koide, T., 211, 221 Koidl, S., 267, 279 Koike, H., 202, 217 Kok, J., 259, 276 Kokke, B. P. A., 113, 123, 134, 151 Kolk, A. H. J., 116, 155 Koll, H., 227, 242 Kolter, R., 259, 268, 278 Kominami, K., 329, 342 Komissarova, N., 4, 39 Komiya, T., 229, 240 Komiyama, T., 359, 384 Kondo, E., 270, 281 Kondo, H., 203, 222, 418, 429 Konieczny, I., 416, 428 Koning, F., 158, 183 Konings, A. W. T., 4, 40, 127, 151 Konings, W. N., 259, 276 Konno, M., 258, 278

453

Konop´asek, I., 128, 132, 154 Kontinen, V., 259, 277 Koonin, E. V., 207, 219, 415, 428 Kopp, F., 190, 218 Korber, P., 288, 300, 351, 352, 384, 388 K¨ordel, J., 255, 275, 278 Koretke, K. K., 203, 214, 216, 418, 426 Koretle, K. K., 188, 202, 203, 219 Kornbluth, S., 19, 33, 43, 271, 275, 282 Kornfeld, R., 304, 310, 319, 322, 333, 336 Kornfeld, S., 304, 309, 310, 328, 329, 333, 335, 341 Korpela, T., 119, 149 Korzun, R., 4, 30, 42 Kosano, H., 175, 182 Kosic-Smithers, J., 229, 239 Kost, S. L., 178, 182 Koster, A. J., 187, 201, 204, 206, 207, 211, 212, 218, 221 Kostka, S., 188, 193, 196, 197, 198, 220, 224, 241, 359, 386 Koszinowski, U. H., 212, 215, 216 Koteiche, H. A., 119, 121, 134, 141, 151, 153 Kotlyarov, A., 121, 126, 134, 148, 152, 154 Kotzbauer, P. T., 9, 43 Kovacs, E., 127, 145, 150 Kovalenko, O., 349, 389 Kowal, A. S., 350, 386, 416, 428 Kozawa, T., 171, 184 Kozutsumi, Y., 229, 241 Kraft, R., 188, 193, 196, 197, 198, 199, 204, 209, 214, 220, 221 Krainer, E., 230, 231, 238, 241 Krajewski, M., 32, 34, 43 Kral, S., 196, 219 Kramer, J., 229, 239 Kramer, L., 192, 207, 213, 221 Kramer, M. L., 245, 253, 279 Kramer, P., 140, 152 Kraulis, P. J., 117, 151 Krause, K. H., 329, 342 Krauss, S., 261, 280 Krco, C. J., 178, 182 Kreibich, G., 330, 343 Krief, S., 108, 114, 143, 151 Krieg, U. C., 229, 241 Krimmer, T., 188, 193, 196, 197, 198, 199, 216, 226, 231, 240, 242 Krishna, P., 161, 183 Kristensen, P., 190, 218

454

AUTHOR INDEX

Kroemer, G., 115, 143, 144, 148 Krofron, J. L., 256, 278 Kronidou, N. G., 230, 231, 240 Kruber, S., 245, 279 Kru¨ ger, E., 199, 221 Kruisbeek, A. M., 158, 183 Krummel, B., 304, 334 Kruse, M., 267, 278 Krutchinsky, A., 212, 213, 221, 222 Krutzsch, H., 164, 165, 171, 181 Krzewska, J., 6, 40 Kuang, J., 260, 271, 282 Kubota, H., 76, 77, 101, 104 Ku¨ brich, M., 230, 231, 238, 239, 241 Kudva, Y. C., 144, 151 Kuehn, L., 212, 215, 216 Kufe, D., 115, 143, 153 Kuhn, P., 353, 379 Ku¨ llertz, G., 269, 270, 278 Kumar, G. S., 111, 130, 133, 154 Kumar, L. V. S., 134, 141, 151, 152 Kumar, S., 115, 143, 153 Kunau, W.-H., 231, 240, 360, 389 Kuppuswamy, D., 144, 151 Kuramitsu, H., 414, 427 Kuriyan, J., 18, 19, 20, 39, 269, 277, 286, 301 Kurochkin, A. V., 23, 25, 26, 28, 43 Kurtz, R. B., 261, 275 Kurtz, S., 115, 152 Kurzik-Dumke, U., 115, 143, 152 Kushner, S. R., 351, 352, 379 Kusters, R., 265, 281 Kutay, U., 30, 42 Kuwajima, K., 248, 262, 264, 277, 278 Kuzmic, P., 256, 257, 276, 278 Kuznetsov, G., 158, 182 Kwan, T., 357, 359, 385

L Laachouch, J. E., 419, 428 Laan, H., 259, 276 Laass, K., 126, 148 Labriola, C., 320, 323, 324, 330, 337, 340 Ladbury, J. E., 161, 162, 163, 164, 165, 166, 167, 168, 169, 171, 172, 174, 175, 176, 179, 183 Ladjimi, M. M., 34, 43 Lai, M. M., 270, 280

Laity, J. H., 251, 275 Lally, J., 203, 222, 418, 429 Laloraya, S., 231, 242 Lam, Y. A., 206, 218 Lambert, H., 115, 121, 126, 127, 139, 144, 149, 150, 152, 155 Lambertson, D., 204, 219 LaMorticello, D. M., 140, 152 Lander, E. S., 160, 181, 368, 371, 373, 374, 375, 381, 388 Landon, F., 127, 140, 149 Landon, M., 140, 152 Landrieu, I., 259, 272, 278 Landry, J., 106, 114, 115, 121, 126, 127, 139, 144, 148, 149, 150, 152, 154, 155 Landry, S. J., 3, 15, 29, 39, 40, 57, 58, 59, 71, 72 Landsman, D., 375, 380 Landt, O., 251, 279 Lane, W. S., 196, 215, 260, 278 Lanet, J., 332, 343 Lang, K., 257, 262, 264, 276, 278, 280 Lang, L., 115, 152, 309, 335 Langen, H., 3, 6, 41, 416, 428 Langer, G., 7, 42 Langer, T., 3, 4, 5, 6, 17, 39, 40, 42, 229, 235, 236, 239, 241, 348, 385, 414, 429 Langren, H., 139, 153 Laroia, G., 127, 144, 149 LaRossa, R. A., 347, 389 Larriba, G., 359, 384 Larronen, S., 115, 140, 152 Larsen, C. N., 187, 206, 208, 218, 219, 415, 417, 422, 428 Lashkari, D., 374, 383 Lassmann, H., 115, 155 Laszlo, A., 115, 152, 206, 214 Lathigra, R., 116, 155 Latterich, M., 415, 428 Lau, W. W., 360, 384 Laufen, T., 9, 10, 12, 14, 15, 22, 23, 24, 25, 26, 27, 38, 40 Lavoie, J. N., 126, 127, 139, 150, 152 Law, D. T. S., 230, 241 Lazar, S. W., 259, 268, 278 Leach, K. L., 178, 184 Leach, M. R., 328, 341 Learn, B. A., 416, 427 Lechleider, R. J., 269, 282 Lecomte, F., 326, 341

AUTHOR INDEX

Ledeboer, A. M., 259, 276 Lederkremer, G. Z., 306, 308, 310, 330, 334, 344 le Douarin, B., 108, 114, 143, 151 Lee, B.-H., 209, 218 Lee, C., 418, 422, 424, 428, 429 Lee, D. H., 206, 218 Lee, F. J., 359, 383 Lee, G. H., 75, 76, 93, 100, 348, 378, 381 Lee, G. J., 106, 108, 111, 113, 114, 115, 125, 127, 128, 130, 131, 133, 135, 136, 137, 138, 150, 152, 155 Lee, I. H., 112, 123, 128, 149 Lee, J. C., 374, 383 Lee, J. D., 212, 219 Lee, J. O., 290, 300 Lee, J. P., 272, 278 Lee, M.-K., 209, 218 Lee, S., 13, 31, 42, 142, 145, 152 Lee, W. H., 157, 158, 181 Lefr`ere, I., 108, 114, 143, 151 Legge, G. B., 13, 30, 31, 40 Leguy, R., 96, 102 Lehmann, H., 199, 215 Lehner, P. J., 324, 340 Lehrman, M. A., 321, 323, 324, 328, 337, 339, 363, 381 Leicht, B. G., 127, 140, 152 Leimgruber, R. M., 255, 276 Lejeune, A., 254, 281 Leloir, L. F., 304, 333 Leng, X. H., 163, 173, 177, 185 Lennarz, W. J., 330, 342 Lenzen, C. U., 418, 428 Leof, E. B., 270, 282 Leonard, G., 203, 222, 258, 279, 418, 429 Lepage, P., 160, 181 Lerner, M., 238, 242 Leroux, M. R., 83, 89, 90, 92, 101, 103, 113, 122, 123, 128, 134, 151, 152 Leshinsky, E., 203, 215 Letourneur, F., 359, 380 Leunissen, J. A. M., 106, 107, 108, 109, 111, 114, 115, 119, 120, 121, 122, 145, 149, 151, 153 Levchenko, I., 415, 416, 418, 419, 425, 428 Leven, S., 259, 276 Levin, E. G., 139, 153 Levine, A. S., 419, 421, 427 Levine, E., 371, 373, 382

455

Levinson, W., 157, 183 Levoie, J. N., 115, 149 Levy, F., 199, 203, 214, 216 Levy, M. A., 127, 149 Lewin, H. A., 189, 190, 214 Lewis, J. W., 324, 340 Lewis, M., 245, 279 Lewis, S. A., 79, 80, 81, 82, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 98, 99, 100, 101, 103, 104 Lewis, S. E., 115, 152 Lewis, V. A., 74, 77, 101, 348, 378, 384 Lewis, W. H., 320, 337 Li, B., 269, 282 Li, C. C. H., 203, 214 Li, D.-P., 7, 41 Li, L., 22, 26, 39 Li, P., 171, 181 Li, T., 209, 218 Li, W., 127, 151, 261, 277 Li, X.-Y., 145, 152 Li, Y., 360, 384 Li, Z., 115, 155 Li, Z. L., 115, 140, 155 Li, Z. Y., 266, 277 Lian, J. B., 115, 154 Liang, J., 259, 278 Liang, J. J. N., 112, 119, 124, 128, 132, 142, 145, 152, 153, 154 Liang, P., 128, 152 Liang, T. J., 207, 211, 222 Liberek, K., 4, 6, 12, 14, 17, 30, 37, 40, 43, 44, 233, 240 Liddington, R. C., 122, 152 Liebman, S. W., 350, 380 Liermann, J., 203, 214, 418, 426 Lilie, H., 9, 38, 112, 113, 121, 149, 170, 184, 248, 262, 278, 281 Lill, R., 29, 42, 142, 153, 236, 240, 264, 278 Lim, J. H., 235, 236, 237, 240, 263, 281 Lin, C. S., 258, 280 Lin, C.-Y., 114, 128, 142, 144, 151, 156 Lin, H. Y., 320, 337 Lin, J.-H., 173, 177, 181 Lin, L., 204, 217 Lin, L.-N., 262, 278 Lin, P., 83, 101 Lin, W. C., 142, 156 Lindberg, U., 98, 103 Linder, B., 115, 142, 143, 152

456

AUTHOR INDEX

Linderoth, N., 158, 184 Lindner, R. A., 28, 43, 106, 112, 122, 123, 134, 135, 148, 152, 155 Lindquist, J. A., 326, 341 Lindquist, S., 3, 5, 6, 7, 38, 41, 44, 108, 115, 142, 152, 153, 154, 157, 158, 161, 173, 174, 175, 176, 180, 181, 183, 261, 275, 349, 350, 377, 380, 382, 385, 386, 387, 413, 414, 416, 427, 428, 429 Lindstrom, J., 322, 339 Lingappa, V. R., 3, 39 Linsten, K., 201, 218 Liou, A. K., 77, 78, 102 Lippens, G., 259, 272, 278 Lippincott-Swartz, J., 355, 385 Lipscomb, W. N., 245, 279 Lipshutz, R. J., 2, 37, 365, 384 Lis, J. T., 115, 150 Lissin, N. M., 74, 99 Lithgow, T., 9, 38 Litt, M., 140, 152 Litwack, G., 167, 185 Liu, D. J., 159, 185 Liu, F., 270, 275 Liu, J., 258, 260, 268, 277, 278 Liu, X. Y., 262, 275 Liu, Y., 115, 151, 330, 331, 343, 344 Liu, Y. Q., 111, 133, 154 Livingston, D. J., 270, 275 Llewelyn, D. H., 326, 341 Llewelyn Roderick, H., 326, 341 Llorca, O., 84, 85, 86, 87, 98, 102 Lockhart, D. J., 356, 358, 360, 362, 365, 367, 370, 375, 380, 384, 388, 631 Loh, S. N., 255, 278 Lohmann, F., 115, 143, 152 Lombardero, J., 272, 282 Lombes, M., 178, 183 Lommel, L., 205, 209, 218 Longati, P., 32, 37 Longo, D. L., 203, 214 Longoni, S., 115, 140, 152 Loones, M.-T., 114, 152 Lopez-Buesa, P., 2, 3, 11, 40, 41 Lopez-Hoyo, N., 2, 3, 41 Lopez-Ilasaca, M., 269, 270, 278 Lorimer, G. H., 5, 42, 54, 66, 67, 68, 72, 74, 79, 80, 100, 102, 104, 106, 152, 347, 348, 382, 387, 389 Loris, R., 273, 275

Los, D. A., 125, 130, 135, 144, 145, 155 Lottspeich, F., 61, 71, 139, 150, 189, 190, 199, 200, 201, 215, 220, 221, 222 Louis, E. J., 349, 382 Louvain, J.-F., 160, 176, 182 Louvet, O., 359, 384 Lovren, E. W., 140, 152 Low, K. B., 61, 71 Lowe, J., 49, 71, 75, 88, 99, 140, 152, 190, 191, 192, 193, 196, 197, 198, 200, 216, 218, 220, 349, 381, 418, 427, 428, 429 Lowman, S., 134, 150 Lowry, D. F., 21, 37 Lu, C., 5, 15, 17, 40, 42 Lu, D. P., 244, 272, 278 Lu, K. P., 259, 260, 271, 272, 278, 279, 280, 281, 282 Lu, P. J., 244, 272, 278 Lu, X., 228, 239, 330, 342 Lu, Z., 28, 30, 40 Luban, J., 260, 270, 275, 276 Lubben, T. H., 74, 102, 104, 347, 389 Lucassen, M., 425, 426 Lucchiari-Hartz, M., 202, 215 L¨uders, J., 19, 32, 34, 40, 206, 218 Ludwig, B., 259, 260, 276, 278, 279, 282 Luhrmann, R., 258, 279 Luirink, J., 83, 99, 265, 281 Lum, P. Y., 366, 371, 372, 377, 383 Luna-Arias, J. P., 359, 384 Lund, P. A., 52, 53, 71, 285, 301 Luo, L., 415, 416, 419, 428 Luo, W., 144, 156 Lupas, A., 89, 101, 187, 188, 189, 193, 196, 202, 203, 204, 206, 207, 208, 212, 214, 216, 218, 219, 220, 221, 418, 426, 429 Lussier, M., 359, 385 L¨utcke, H., 265, 277 Lutsch, G., 112, 121, 126, 134, 139, 148, 152, 154 Lutz, T., 2, 40 Lyman, S. K., 230, 240

M Ma, B. J., 113, 122, 152 Ma, C. P., 212, 218 Ma, L., 272, 278 Ma, P. C., 203, 209, 217

AUTHOR INDEX

Ma, Q., 174, 182 Ma, W., 127, 155, 355, 364, 385 Ma, Y. T., 207, 213 Maack, S., 197, 216 Maarse, A., 231, 236, 238, 239, 240 Maarse, S. A., 230, 231, 238, 241 Maasdam, D., 114, 156 Macarthur, M. W., 245, 278 Macdonald, J. R., 261, 282 Macke, J. P., 96, 101 MacKinnon, C., 262, 281 MacLennan, D. H., 329, 342 MacRae, T. H., 106, 128, 134, 152 Macri, J., 332, 344 Madden, B., 7, 35, 42, 178, 184 Madura, K., 204, 205, 209, 218, 219 Maeda, Y., 178, 185, 359, 389 Magendantz, M., 96, 98 Mahe, Y., 199, 216 Mahlke, K., 225, 226, 240 Makarow, M., 224, 241 Maki, K., 248, 262, 278 Maki, N., 268, 277 Makino, K., 202, 217 Makino, S., 202, 217 Malancon, S. B., 160, 181 Malayaman, N., 207, 211, 222 Malfois, M., 121, 149 Malkus, P., 360, 384 Mallick, P., 189, 218 Mally, A., 19, 40 Maloney, A., 161, 171, 177, 181 Manganaro, T., 269, 282 Maniatis, T., 204, 218, 219 Mann, C. J., 207, 216, 262, 278 Mann, K., 193, 197, 215, 259, 260, 276, 278, 279 Mann, M., 326, 341 Mannhaupt, G., 209, 218 Mannherz, H. G., 10, 39 Manning-Krieg, U. C., 231, 240, 241 Mannion, R. J., 115, 152 Manteuffel, R., 127, 153 Marahiel, M. A., 244, 267, 276 Maratea, D., 261, 278 Marco, S., 84, 86, 102 Marcu, M. G., 172, 182 Marcus, N. Y., 330, 343 Maresca, B., 144, 145, 148, 155 Marin, R., 115, 152

457

Marion, T. N., 7, 35, 42, 178, 184 Maritan, A., 371, 373, 374, 383 Markley, J. L., 255, 272, 278, 281 Marks, A. R., 269, 278, 279 Marquardt, T., 354, 382 Marsh, J. A., 173, 174, 181, 182, 261, 275, 278 Marsh, M., 177, 180 Marszalek, J., 2, 4, 12, 40, 43, 233, 240 Martensson, L. G., 249, 276 Martin, B., 28, 43 Martin, E., 7, 32, 34, 38, 42 Martin, F., 223, 235, 236, 237, 238, 239, 240, 263, 281 Martin, H., 226, 242 Martin, J. L., 115, 152, 225, 226, 240, 269, 282, 286, 288, 289, 291, 300, 301, 346, 347, 348, 349, 377, 378, 380, 385 Martin, L., 95, 102 Martin, N., 290, 291, 292, 293, 295, 296, 298, 300, 301 Martin, R. A., 74, 99 Mart´ın-Barrientos, J., 306, 308, 334 Martinez-Oyanedel, J., 251, 278 Martinez-Yamout, M., 13, 30, 31, 40 Marton, M. J., 366, 371, 372, 377, 383 Maruya, M., 169, 182 Maruyama, T., 80, 100 Maruyama, Y., 377, 385 Marx, S. O., 269, 279 Marz, S. O., 269, 278 Masin, J., 128, 132, 154 Maskos, K., 15, 29, 39 Massague, J., 269, 270, 275, 277 Masselos, D., 66, 70, 71 Masson, P., 212, 218 Masso-Welch, P., 320, 337 Masters, E. I., 192, 207, 213, 221 Masuda, H., 96, 100 Mathews, L. M., 269, 280 Mathieu, J., 160, 181 Mathur, S., 122, 128, 148 Matlack, K. E. S., 3, 15, 40, 223, 230, 234, 240, 241, 325, 329, 340 Matouschek, A., 223, 231, 235, 236, 237, 240, 242, 262, 267, 278, 422, 424, 428 Matquardt, T., 320, 337 Matsufuji, S., 204, 218 Matsumoto, K., 171, 184 Matsumoto, S., 77, 101

458

AUTHOR INDEX

Matsuyama, M., 171, 185 Matsuzawa, S.-i., 32, 43 Matszalek, J., 416, 429 Matthews, C. R., 253, 262, 278, 281 Matthey-Dupraz, A., 295, 299 Mattick, J. S., 414, 427 Matts, R. L., 7, 8, 39, 43, 171, 177, 181 Matuschewski, K., 203, 204, 217 Matzuk, M. M., 269, 270, 271, 275, 276, 280 Mau, R., 349, 351, 386 Maupin-Furlow, J. A., 188, 189, 193, 196, 197, 202, 218, 221 Maurizi, M. R., 127, 151, 188, 189, 206, 211, 214, 217, 221, 349, 350, 382, 383, 386, 388, 389, 413, 414, 415, 416, 417, 418, 419, 421, 422, 423, 424, 425, 426, 427, 428, 429 Maxwell, A., 159, 163, 168, 170, 172, 180, 182, 185 Maxwell, J., 57, 58, 59, 72 May, A. P., 418, 428 Mayer, A., 206, 214, 236, 240 Mayer, J., 248, 281 Mayer, M. P., 9, 10, 12, 14, 15, 16, 17, 18, 19, 20, 22, 23, 24, 25, 26, 27, 34, 37, 38, 40, 42, 347, 349, 385 Mayer, R. J., 140, 152, 207, 216 Mayer, S., 258, 281 Mayer, T. U., 203, 218 Mayhew, M., 66, 67, 71, 349, 385 Mayr, J., 193, 196, 218 Mayr, L. M., 244, 246, 248, 251, 252, 254, 264, 278, 279, 280 Mazzarella, R. A., 320, 337 Mazzochi, C., 330, 342 McCallum, C. D., 80, 92, 102 McCarthy, A. A., 286, 293, 295, 301 McCarty, J. S., 9, 12, 14, 16, 21, 25, 37, 38, 40, 233, 239 McColl, D. J., 15, 30, 42 McCormack, E. A., 78, 86, 87, 98, 102 McCormack, T. A., 199, 218 McCormick, D. J., 7, 35, 42, 178, 184 McCormick, S. J., 322, 338 McCourt, D. W., 158, 184 McCracken, A. A., 320, 330, 337, 354, 355, 379, 389 McDermott, H., 140, 152 McDowell, R. S., 57, 58, 71 Mcgee, W. A., 262, 281

McGovern, K., 285, 300 McGuire, J., 178, 182 McGuire, W. L., 106, 115, 149 McGurk, G., 207, 216 Mchaourab, H. S., 112, 116, 119, 121, 122, 123, 124, 134, 141, 148, 150, 151, 153 McIntosh, J. R., 96, 100 McKay, D. B., 11, 12, 16, 19, 21, 22, 25, 26, 38, 39, 41, 42, 43, 164, 181, 189, 220, 417, 418, 422, 429 McKean, D. J., 328, 341 McKee, J. J., 116, 148 McKenney, K., 4, 44, 350, 389, 415, 416, 419, 429 McMacken, R., 3, 4, 10, 11, 12, 14, 16, 17, 37, 39, 40, 42, 416, 427 McMahon, N., 167, 175, 182, 185 McMurray, M., 270, 282 McNeil, L. K., 189, 190, 214 Mcnemar, C. W., 255, 278 McQuade, K. L., 158, 185 McWherter, C. A., 255, 276 Meacham, G. C., 36, 40, 206, 218 Meaden, P., 311, 312, 314, 336 Meadow, N. D., 10, 39 Means, A. R., 271, 275, 282 Mech, C., 256, 268, 276 Medalia, O., 188, 204, 206, 209, 217 Mehl, A. F., 10, 11, 14, 42 Mehlen, A., 126, 153 Mehlen, P., 126, 153 Mehta, A., 330, 342 Meijer, M., 225, 230, 231, 233, 236, 238, 239, 240, 241 Melandri, F. D., 199, 218 Melcher, K., 415, 429 Meldolsei, J., 320, 321, 337 Melki, R., 79, 86, 87, 96, 102, 103, 113, 122, 123, 128, 134, 152 Melnick, J., 158, 182 Mendelzon, D. H., 306, 308, 310, 334 Menoret, A., 158, 185 Mercier, J., 160, 181 Merck, K. B., 106, 112, 114, 119, 121, 122, 123, 125, 128, 130, 135, 150, 153 Merlin, A., 231, 236, 239, 240 Merrick, J. M., 202, 214 Mertens, D., 84, 102 Mesaeli, N., 321, 329, 337, 342 Mesirov, J., 371, 373, 374, 375, 388

AUTHOR INDEX

Messing, A., 141, 148 Mester, J., 157, 178, 182 Mestril, R., 115, 147 Metcalf, P., 286, 293, 295, 301 Metcalfe, S. M., 244, 260, 261, 276 Mewes, H. W., 188, 202, 203, 219, 349, 382 Meyer, B. K., 174, 182 Meyer, D. I., 224, 227, 241 Meyer, H. A., 359, 385, 418, 429 Meyer, H. E., 126, 148, 155, 203, 222 Meyer, M. R., 366, 371, 372, 377, 383 Meyer, T. H., 199, 215 Mhammedi-Alaoui, A., 415, 428 Mi, H., 272, 277 Miao, B., 229, 231, 236, 239, 240 Michaelis, G., 83, 99 Michaelis, S., 359, 381 Michalak, M., 321, 323, 326, 329, 337, 341, 342 Michaud, C., 199, 219 Michaud, S., 115, 151 Michelini, A. T., 112, 128, 138, 153 Michels, A. A., 4, 40 Michnick, S. W., 258, 279 Miedema, F., 329, 342 Mihara, K., 229, 231, 240 Mikereit, P., 233, 239 Miklos, D., 5, 44, 74, 79, 84, 102, 104, 348, 378, 390 Mikol, V., 258, 279 Milburn, D., 353, 388 Milkereit, P., 21, 37 Millan, J. A., 32, 34, 43 Miller, A. D., 57, 58, 59, 65, 71 Miller, B., 235, 239 Miller, C., 259, 276 Miller, I., 127, 151 Miller, J. H., 189, 218 Miller, K. R., 223, 240 Miller, P., 158, 171, 174, 180, 182, 184 Miller, R. J., 272, 278 Miller, S. C., 255, 281 Mimnaugh, E. G., 158, 164, 165, 171, 174, 176, 180, 181, 182, 185 Minami, Y., 34, 39, 212, 220 Minchin, R. F., 161, 180 Minowada, G., 115, 139, 153 Miranker, A. D., 350, 389, 422, 423, 429 Miron, T., 127, 139, 153 Miskovic, D., 115, 152

459

Misra, L. M., 229, 241, 242, 354, 373, 386, 389 Misselwitz, B., 3, 12, 15, 22, 24, 25, 26, 40, 230, 233, 234, 240, 241 Missiakas, D., 259, 268, 279, 285, 290, 292, 295, 296, 297, 298, 299, 300, 301, 414, 428 Misura, K. M., 418, 428 Mitchell, R., 270, 280 Mitchison, T. J., 93, 102 Mitra, B., 285, 301 Mitra, M., 371, 373, 374, 383 Mittmann, M., 365, 384, 389 Miura, T., 377, 378, 385, 388 Miyasaka, N., 330, 343 Miyawaki, M., 211, 215 Mizukami, T., 171, 184 Mizunaga, T., 369, 377, 378, 385, 388 Mizuno, S., 171, 181 Mizuno, T., 28, 43 Mo, X. Y., 201, 221 Moareﬁ, I., 18, 20, 34, 35, 42, 161, 173, 184 Moczko, M., 235, 238, 241 Model, K., 223, 240 Moens, U., 261, 280 Moerman, A. M., 115, 153 Moerschell, R. P., 189, 219, 414, 417, 428 Mogi, T., 291, 292, 301 Mogk, A., 2, 3, 6, 38, 39, 40, 41, 139, 153, 267, 275, 350, 382, 416, 428 Mohammadi, A., 332, 344 Mohrle, V., 199, 216 Mohrs, K., 267, 279 Molinari, M., 230, 241, 325, 326, 329, 337, 340, 341, 354, 355, 360, 381 Momand, J., 157, 184 Monaco, J. J., 198, 213, 217, 218 Moncharmont, B., 178, 183 Mondesert, G., 359, 381 Monne-van Muijen, M., 158, 183 Montelione, G. T., 353, 385 Montgomery, D. L., 22, 23, 26, 27, 41 Moody, P. C., 192, 214 Moomaw, C. R., 204, 206, 215 Moore, J. M., 258, 279 Moore, M. J., 125, 149 Moore, S. E. H., 330, 342, 343 Moore, T., 374, 383 Moran, C. P., Jr., 115, 116, 150 Morange, M., 3, 41, 114, 126, 152, 154

460

AUTHOR INDEX

Morano, K. A., 176, 182 Morﬁni, G., 30, 43 Morgan, D. O., 176, 181 Morgan, K., 160, 181 Mori, K., 355, 359, 364, 365, 383, 384, 385 Mori, M., 28, 43, 228, 242 Morimoto, R. I., 3, 4, 9, 19, 21, 23, 27, 32, 33, 37, 38, 41, 43, 261, 276 Morishima, Y., 7, 32, 35, 39, 41, 179, 182 Morishita, R., 114, 151 Mornon, J. P., 265, 275 Moro, F., 231, 235, 241 Moroder, L., 35, 42, 161, 173, 184, 189, 192, 207, 216, 220, 418, 429 Morrice, N. A., 326, 341 Morris, D. P., 272, 279 Morris, J. A., 322, 339 Morris, M. S., 365, 380 Morrow, G., 108, 153 Morshauser, R. C., 23, 26, 41 Moss, B., 322, 338 Mossner, E., 289, 300 Mothes, W., 224, 240 Motohashi, K., 416, 428 Mott, J. D., 213, 213 Moudgil, K. D., 116, 148 Moulai, J., 122, 152 Moutiez, M., 290, 300 Moyer, J. D., 158, 171, 174, 180, 182, 184 Moyer, M. P., 171, 174, 182 Msadek, T., 414, 426 Muchowski, P. J., 106, 130, 132, 134, 141, 142, 149, 153 Mucke, ¨ M., 251, 263, 266, 279, 280 Mueske, C. S., 144, 151 Mufﬂer, A., 424, 428 Muir, S., 173, 181, 257, 270, 275 Mulder, E., 178, 185 Mulders, J. W., 145, 153 Mulholland, J., 374, 380 Mullane, K. P., 29, 37 Muller, ¨ G., 224, 241 Muller, ¨ H., 230, 231, 236, 238, 239, 241, 242 Muller, ¨ J., 259, 280 Muller, S., 76, 104 Muller, ¨ S. A., 188, 189, 203, 204, 206, 209, 217, 219, 221, 416, 417, 428 Muller, ¨ W. E. G., 106, 114, 139, 148

Muller-Taubenberger, ¨ A., 190, 220 Mullins, C., 359, 385 Multhaup, G., 14, 25, 38 Mun, A., 9, 38 Munchbach, M., 107, 114, 153 Munoz, M. J., 176, 182 Munro, S., 359, 383 Murakami, H., 227, 241 Murakami, Y., 171, 181, 204, 212, 218, 220, 349, 382 Murakata, C., 172, 184 Murata, S., 213, 218 Murphey, W., 140, 152 Murphy, P. J., 7, 41 Murray, B. W., 212, 218 Murzin, A. G., 192, 214 Musch, ¨ A., 230, 241 Muschler, P., 167, 168, 169, 185 Muse, W., 291, 292, 299, 300, 351, 383 Mustafa, F. B., 209, 218 Muzzi, M. L., 171, 184 Myers, C. E., 171, 185 Myers, E. W., 28, 43 Myers, M. P., 21, 38 Myszka, D. G., 270, 282 Myung, J., 201, 218

N Nabeshima, Y., 213, 218 N´adasdi, L., 127, 145, 150 Nadeau, K., 163, 182 Nagata, K., 9, 44, 115, 155 Nagy, I., 189, 203, 214, 218, 220, 221 Nair, S., 35, 41 Nair, S. C., 175, 178, 179, 183, 184, 261, 275 Nakagawa, H., 211, 215 Nakai, A., 9, 44 Nakai, H., 415, 421, 427, 429 Nakamoto, H., 128, 142, 145, 153, 154 Nakamura, K., 321, 329, 337, 342 Nakamura, N., 359, 389 Nakanishi, H., 369, 388 Nakano, A., 359, 387 Nakano, H., 171, 172, 184 Nakao, K., 178, 185 Nakayama, E., 158, 182 Nakayama, H., 126, 150

AUTHOR INDEX

Nakazawa, A., 115, 143, 153 Nalin, C., 115, 143, 153 Nall, B. T., 250, 262, 264, 279, 281 Nandi, D., 198, 218 Naqvi, N. I., 209, 218 Narberhaus, F., 107, 114, 153, 154 Nardai, G., 158, 181 Narumi, H., 171, 172, 184 Nastainczyk, W., 238, 242 Nathan, D. F., 161, 174, 175, 176, 183, 377, 385 Nathans, D., 207, 221 Naudat, V., 255, 256, 272 Naumann, M., 205, 209, 216, 220 Nauseef, W. M., 322, 338 Navarro, D., 158, 183 Nebreda, A. R., 126, 154 Neckers, L. M., 158, 163, 164, 165, 171, 172, 174, 176, 180, 181, 182, 184, 185 Neefjes, J. J., 329, 342 Negroiu, G., 324, 340 Nehls, S., 355, 385 Neidhardt, F. E., 14, 41 Nelsestuen, G. L., 273, 276 Nelson, J. W., 287, 288, 301 Nelson, R. E., 309, 310, 311, 312, 314, 335 Nelson, R. J., 3, 41 Nemoto, T., 161, 169, 182, 183 Nesper, M., 190, 220 Neumann, D., 127, 153 Neupert, W., 2, 28, 40, 41, 74, 99, 102, 224, 225, 226, 227, 229, 230, 231, 235, 236, 238, 239, 240, 241, 242, 346, 347, 377, 378, 380, 386, 414, 429 Neuwald, A. F., 207, 219, 415, 428 Newman, A. P., 359, 385 Newman, R. H., 203, 222, 418, 429 Ng, D. T., 202, 222, 361, 363, 385 Ng, D. W. T., 209, 219 Ng, K. K. S., 273, 279 Ngo, K., 212, 219 Ngo, W., 32, 38 Nguyen, M., 357, 359, 385 Nguyen, P., 158, 165, 171, 180, 181 Nguyen, T. H., 230, 241 Nguyen, V. T., 3, 41 Nguyen Van, P., 322, 324, 338 Nicchitta, C. V., 158, 162, 185 Nicholl, I. D., 140, 153

461

Nicholls, A., 41 Nicholls, K. A., 134, 148 Nicholson, D. W., 143, 156, 357, 359, 385 Nicolas, A., 163, 180 Nicolet, C., 3, 35, 41 Nicolet, C. M., 229, 239 Niedermann, G., 202, 215 Nieland, T. J., 158, 183 Nierhaus, K., 265, 281 Nierhaus, K. H., 351, 352, 384 Nigam, S. K., 158, 182 Nikiforov, V., 4, 39 Nilsson, I. M., 304, 333 Nimmesgern, E., 3, 5, 38, 75, 76, 92, 100, 104, 170, 171, 174, 184, 347, 348, 378, 381, 388 Nishida, J., 320, 337 Nishihara, K., 266, 279 Nishihara, R., 369, 388 Nishikawa, S., 358, 359, 386, 387 Nishimune, Y., 329, 342 Nitsch, M., 76, 104 Nixon, J. E., 189, 190, 214 Niyaz, Y., 34, 44 Noble, M. E. M., 177, 182 Nocker, A., 107, 114, 153 Noda, C., 208, 213, 213, 221 Noel, J. P., 259, 271, 272, 279, 281 Nogales, E., 96, 99 Noguchi, S., 7, 37 Nohara, D., 127, 150 Nollen, E. A. A., 32, 41 Nonaka, I., 108, 144, 155 Noonan, L. C., 116, 148 Norbury, C. C., 203, 220 Norcum, M. T., 75, 102 Norman, D. G., 106, 126, 155 Normington, K., 229, 241 Norris, K., 158, 181 Nouvet, F. J., 359, 381 Nover, L., 127, 136, 149, 151, 153 Novick, P., 359, 388 Nucifora, F. C., 270, 275 Nudler, E., 4, 39 Nunes, J. A., 258, 279 Nunes, S. L., 199, 218 Nunoshiba, T., 202, 217 Nussbaum, A. K., 199, 200, 215, 219 Nyoumura, K., 211, 220 Nyu-i, N., 108, 144, 155

462

AUTHOR INDEX

O Oberdorf, A. M., 115, 154 Obermann, W. M. J., 160, 161, 165, 176, 183, 185 O’Brien, M. C., 11, 41, 161, 162, 163, 164, 165, 167, 168, 169, 174, 175, 179, 183 O’Brien, R., 164, 165, 166, 171, 172, 176, 183 Ochel, H.-J., 164, 165, 171, 181 Odaert, B., 259, 272, 278 Odefey, C., 252, 254, 278, 279 Odorizzi, G., 359, 380 Oesterreich, S., 106, 115, 149 Ogata, C. M., 20, 21, 25, 44 Oguchi, S., 189, 220 Ogura, T., 14, 43 Oh, H. J., 138, 155 Ohan, N., 115, 150, 153 Ohara-Nemoto, Y., 161, 183 Ohashi, K., 330, 343 Ohba, M., 208, 221 Ohishi, K., 359, 389 Ohlson, M., 2, 43 Ohmiya, K., 359, 384 Ohno, S., 108, 114, 144, 154, 155 Ohta, H., 158, 182 Ohtsuka, K., 3, 4, 29, 38, 40, 41, 92, 100, 228, 242 Oinonen, C., 193, 219 Okabe, M., 329, 342 Okamoto, K., 126, 150 Okamura, H. H., 230, 242 O’Keefe, B., 7, 41 O’Keefe, D. P., 74, 104, 347, 389 Old, L. J., 158, 185 Oliver, J. D., 326, 340, 341 Oliver, S. G., 349, 382 Olsen, G. J., 189, 190, 202, 214 Olsen, R., 261, 280 Omura, S., 330, 343 Omura, T., 231, 240 Ondrias, K., 269, 278, 279 Ono, A. M., 159, 185 Ono, B., 350, 380 Ono, T., 158, 182 Onodera, K., 378, 388 Oobatake, M., 251, 279 Ooi, T., 251, 279 Opas, M., 321, 329, 337, 342 Oppenheim, J., 74, 104

Opperman, H., 157, 183 Oppliger, W., 227, 230, 231, 239, 240 Ora, A., 322, 324, 338, 339 Orengo, C. A., 353, 388 Orlean, P., 359, 381, 386 Orlova, E., 203, 222, 418, 429 Orlowski, M., 199, 219 Orphanides, G., 168, 180 Ortaldo, J. R., 261, 272 Ortega, J., 421, 428 Ortmann, B., 324, 340 Ortolan, T. G., 204, 219 Osario, M., 356, 361, 366, 380 Osawa, Y., 7, 37 O’Shea, E. K., 360, 384 Osipiuk, J., 347, 348, 349, 388 Oster, G. F., 231, 242 Osterman, D. G., 261, 277 Ostermann, J., 74, 102, 229, 230, 240, 241, 347, 386 Osteryoung, K. W., 127, 153 Ota, I. M., 203, 209, 217 Ota, M., 161, 183 Otteken, A., 322, 338 Ottinger, E. A., 171, 177, 181 Otto, A., 289, 301 Otwinowski, A., 53, 71 Ou, M. S., 188, 189, 197, 202, 218, 221 Ou, W.-J., 321, 338 Oude Essink, B. B., 112, 121, 122, 123, 153 Oudega, B., 265, 281 Ouerfelli, O., 170, 171, 174, 184 Overbeek, R., 202, 214 Overkamp, P., 119, 153, 155 Owen, B. A., 158, 165, 181 Owen, D. J., 177, 182 Owen, H. A., 142, 145, 152 Owen, T. A., 115, 154 Owens-Grillo, J. K., 161, 163, 173, 183, 185, 261, 279

P Paal, K., 9, 10, 12, 22, 23, 24, 25, 26, 40 Pace, C. N., 251, 279 Paciucci, R., 93, 100 Packschies, L., 10, 16, 17, 18, 19, 20, 34, 37, 41 Padmanhaban, A., 286, 300

AUTHOR INDEX

Pahl, A., 266, 280 Pahl, H. L., 329, 332, 342 Pai, E. F., 10, 39 Pain, D., 227, 241 Pain, R. H., 254, 281, 282 Pak, M., 349, 350, 383, 386, 420, 424, 427, 428 Palfrey, H. C., 272, 278 Pali, T., 127, 145, 150 Palleros, D. R., 9, 41 Palmisano, D. V., 132, 153 Palombella, V. J., 204, 219 Palter, K. B., 127, 140, 152 Palzkill, T., 254, 281 Panaretou, B., 161, 162, 164, 165, 166, 167, 168, 169, 174, 175, 176, 179, 183 Pandey, P., 115, 143, 153 Pang, J., 212, 219 Pang, Y., 23, 25, 26, 28, 41, 43 Panzner, S., 224, 241, 359, 381, 386 Papa, F. R., 206, 208, 215, 219, 360, 386 Papapetropoulos, A., 7, 38 Papp, O., 19, 40 Paravicini, G., 359, 360, 386, 388 Pardini, C. L., 145, 148 Parfenova, M., 119, 148 Paris, K., 57, 58, 71 Park, H., 159, 185, 330, 342 Park, J. S., 322, 339 Park, S. C., 416, 420, 422, 428 Parker, C. G., 309, 310, 311, 312, 314, 335 Parker, S. B., 163, 173, 177, 185 Parmentier, Y., 190, 219 Parodi, A. J., 304, 305, 306, 307, 308, 309, 310, 311, 312, 315, 316, 317, 318, 320, 323, 324, 326, 327, 329, 330, 331, 333, 334, 335, 335, 336, 337, 340 Parraga, M., 96, 99 Parsell, D. A., 349, 350, 386, 414, 416, 428, 429 Partaledis, J. A., 354, 386 Patel, D., 15, 30, 42 Patel, S., 415, 428 Pater, M. M., 34, 44 Patil, A. R., 323, 324, 339 Patil, C. K., 356, 358, 360, 362, 367, 370, 388, 631 Pato, M., 415, 428 Patterson, C., 36, 37, 38, 40, 206, 214, 218 Paul, C., 121, 126, 154

463

Pauli, D., 115, 148 Paulin, D., 115, 140, 155 Paunola, E., 224, 241 Pavletich, N. P., 159, 162, 163, 164, 165, 171, 176, 183, 185 Payne, G. S., 359, 380 Payton, M. A., 359, 388 Paz, I. B., 157, 184 Pearce, D. A., 76, 103 Pearl, L. H., 159, 161, 162, 163, 164, 165, 166, 167, 168, 169, 171, 172, 174, 175, 176, 179, 183 Peattie, D. A., 258, 279 Pecht, A., 266, 280 Pedersen, S., 414, 415, 416, 426, 429 Peek, J. A., 286, 301 Pego, J., 125, 149 Pegoraro, S., 35, 42, 161, 173, 184 Peitsch, M. C., 39, 41 Pelham, H. R., 227, 241, 320, 337, 357, 359, 386, 387 Pellecchia, M., 13, 15, 22, 23, 26, 27, 30, 41, 43 Peng, I., 86, 99 Peng, R., 359, 386 Penney, M., 208, 221 Perdew, G. H., 7, 41, 157, 161, 173, 174, 178, 181, 182, 183, 185 Pereira, L., 158, 183 Peres Ben-Zvi, A., 3, 6, 38, 39 P´erez, M., 310, 335 Perlmutter, D. H., 330, 343 Perng, M. D., 140, 153 Perrett, S., 66, 72 Perry, K. M., 125, 153 Peskin, C. S., 231, 242 Peter, E., 113, 122, 152 Peters, J., 76, 104, 203, 206, 211, 214, 216, 219, 418, 426 Petersens, U. M., 212, 218 Peterson, J. R., 322, 324, 338 Peterson, M. R., 258, 279 Peterson, P. A., 198, 212, 219, 221, 322, 323, 324, 326, 328, 338, 339 Petko, L., 108, 115, 142, 152, 153 Petrash, J. M., 122, 128, 141, 145, 148, 149 Petrescu, A. J., 310, 324, 327, 330, 332, 335, 340, 342, 343 Petrescu, S., 310, 323, 324, 327, 335, 338 Petrescu, S. M., 324, 330, 332, 340, 342, 343

464

AUTHOR INDEX

Petri, R., 7, 42 Petris, A., 24, 38 Pettigrew, D. W., 10, 39 Peypouquet, M. F., 359, 384 Pfanner, N., 28, 41, 223, 224, 225, 226, 227, 229, 230, 231, 233, 235, 236, 237, 238, 239, 240, 241, 242, 263, 267, 279, 281 Pfeifer, G., 76, 104, 188, 189, 202, 203, 204, 205, 208, 209, 215, 216, 219, 221, 416, 417, 428 Pfeil, W., 119, 149 Pfund, C., 2, 3, 11, 40, 41, 83, 99 Phatnani, H. P., 272, 279 Philippsen, P., 349, 382 Philipsen, R. L. A., 107, 151 Piatigorsky, J., 125, 151 Picard, D., 7, 8, 38, 41, 160, 176, 180, 181, 182 Pichersky, E., 414, 427 Pickart, C. M., 188, 192, 204, 205, 206, 208, 215, 219, 220, 222 Pierce, D. W., 9, 41 Pierce, M. M., 262, 281 Pierpaoli, E. V., 9, 12, 14, 25, 38, 41 Pike, I., 140, 152 Pilon, M., 231, 233, 241 Pinkau, T., 347, 348, 349, 388 Pinkner, J., 286, 300 Piotrowicz, R. S., 139, 153 Pipe, S. W., 322, 339 Piper, P. W., 159, 161, 162, 163, 164, 165, 166, 167, 168, 169, 171, 172, 174, 175, 176, 179, 183 Pitluk, Z., 84, 102 Plamondon, L., 199, 207, 213, 218 Plath, K., 3, 40, 223, 228, 230, 234, 240, 241, 359, 381 Platt, F. M., 310, 324, 327, 329, 330, 332, 335, 340, 342, 343 Plemper, R. K., 204, 219, 320, 330, 337 Plesofsky, N., 142, 153 Plesofsky-Vig, N., 138, 142, 153 Ploegh, H., 356, 361, 366, 380 Ploegh, H. L., 189, 202, 221, 329, 330, 342, 355, 389 Pl¨uckthun, A., 60, 71, 259, 261, 268, 269, 272, 275, 277, 279, 282 Plumier, J. C. L., 114, 115, 154 Plunkett, G. III, 414, 426 Poe, M., 258, 280

Poellinger, L., 157, 174, 178, 182, 183, 185 Pohl, J., 2, 38 Pokala, N., 113, 128, 130, 152 Polevoda, B., 198, 217 Pollard, M. G., 299, 301, 377, 386 Pollock, R. A., 74, 99, 346, 347, 377, 378, 380 Polman, C. H., 115, 155 Pongratz, I., 174, 182 Ponting, C. P., 208, 209, 213 Ponzetto, C., 32, 37 Pookanjanatavip, M., 125, 153 Poon, R. Y. C., 7, 39, 177, 182 Poortmans, F., 273, 275 Popov, M., 322, 324, 339 Porankiewicz, J., 413, 415, 428 Portetelle, D., 259, 272, 278 Portier, M. M., 127, 140, 149 Poulter, C. D., 349, 385 Powis, S. J., 326, 341 Pozzan, T., 320, 321, 337 Prakash, S., 422, 424, 428 Prapapanich, V., 35, 41, 173, 175, 178, 179, 181, 183, 184 Prasad, K., 12, 15, 28, 37, 43 Pratt, G., 203, 212, 215, 217 Pratt, L. A., 424, 428 Pratt, W. B., 4, 7, 32, 35, 37, 38, 39, 41, 42, 157, 158, 161, 163, 167, 171, 173, 175, 177, 178, 179, 180, 181, 182, 183, 184, 185, 261, 279 Pray-Grant, M. G., 174, 178, 181, 182 Preckel, T., 212, 219 Prehn, S., 359, 381 Prehoda, K. E., 255, 281 Preissner, R., 188, 220 Prescott, A., 140, 153 Presley, J. F., 355, 385 Prestidge, R. L., 116, 148 Preuss, M., 57, 58, 59, 65, 71 Preville, X., 121, 126, 153, 154 Pr´evost, M. C., 115, 140, 155 Priemer, M., 199, 215 Prieto-Dapena, P., 115, 147, 154 Primm, T. P., 112, 123, 128, 149 Prinz, W. A., 351, 380, 386 Prip-Buus, C., 377, 378, 386 Privalsky, M. L., 178, 183 Prives, J. M., 322, 339 Prochaska, D. J., 142, 145, 152 Prockop, D. J., 262, 267, 275, 276

AUTHOR INDEX

Prodromou, C., 159, 161, 162, 163, 164, 165, 166, 167, 168, 169, 171, 172, 174, 175, 176, 179, 183 Prohaszka, Z., 158, 181 Proske, R. J., 204, 206, 215 Proudfoot, A., 359, 388 P¨uhler, G., 190, 197, 216, 222 Pukazhenthi, B., 305, 334 Pumphrey, J., 415, 423, 427 Purugganan, M. D., 207, 220

Q Qadri, I., 158, 183 Qian, Y. Q., 15, 30, 42 Qin, J., 290, 301 Qin, L., 159, 185 Quass, T., 251, 252, 277 Quemeneur, E., 290, 300 Querol, A., 235, 239 Quiniou, E., 255, 256, 272 Quinlan, R. A., 106, 126, 140, 141, 149, 153, 154, 155 Quinn, P. T., 111, 133, 154 Quiocho, F. A., 112, 123, 128, 149

R Radanyi, C., 157, 178, 182, 183 Radcliffe, P. A., 96, 103 Rademacher, F., 76, 103 Radermacher, M., 85, 102 Radford, S. E., 59, 66, 67, 70, 71 Radominska-Pandya, A., 309, 310, 311, 312, 336 Rafestin-Oblin, M. E., 178, 183 Rahfeld, J. U., 259, 260, 261, 263, 264, 265, 271, 277, 278, 279, 280, 281, 282 Raina, S., 259, 268, 275, 279, 285, 290, 292, 295, 296, 298, 299, 300, 301, 414, 428 Rajagopalan, S., 328, 341 Rajapandi, T., 35, 42 Rajaraman, K., 132, 154 Raleigh, D. P., 255, 282 Ram, A. F., 359, 385 Ramachandran, R., 189, 214, 220, 418, 429 Ramakrishna, T., 132, 135, 141, 151, 154 Raman, B., 125, 132, 135, 154

465

Ramm, K., 269, 279 Rammensee, H. G., 199, 200, 201, 212, 215, 216, 219, 220 Ramos, P. C., 199, 219 Ramsey, A. J., 161, 183 Randall, L. L., 124, 155 Rando, O. J., 204, 219 Ranganathan, R., 259, 271, 279 Ranson, N. A., 46, 48, 53, 54, 62, 63, 64, 67, 69, 71, 72, 349, 380 Rao, C. M., 125, 132, 134, 135, 141, 149, 151, 152, 154 Rao, J., 167, 173, 174, 181 Rao, P. C., 130, 154 Rape, M., 203, 204, 217, 266, 280 Rapoport, G., 414, 426 Rapoport, T. A., 3, 12, 15, 22, 24, 25, 26, 30, 40, 42, 223, 224, 228, 230, 233, 234, 240, 241, 359, 381, 386 Rapp, U. R., 23, 25, 26, 28, 43 Raquet, X., 254, 281 Rascher, C., 259, 280 Rasmussen, T. P., 208, 215 Rassow, J., 225, 227, 230, 231, 235, 236, 238, 239, 240, 241, 242, 263, 267, 279, 281 Rasulova, F., 421, 427 Ratajczak, T., 161, 173, 178, 180, 183 Ratliff, K. S., 235, 236, 237, 240 Ravagnan, L., 115, 143, 144, 148 Ravid, R., 115, 155 Ray, M. K., 329, 341 Rayner, J. C., 359, 383 Raynes, D. A., 36, 42 Realini, C., 189, 192, 208, 212, 213, 217, 219, 222 Rechsteiner, M., 187, 189, 192, 203, 204, 205, 206, 207, 208, 211, 212, 213, 215, 216, 217, 219, 220, 222 Reck, K., 261, 282 Reddy, M. C., 132, 153 Reddy, P. S., 325, 340 Reddy, V. S., 189, 220, 417, 418, 422, 429 Redeuilh, G., 178, 183 Reed, J., 7, 42, 74, 104, 347, 389 Reed, J. C., 9, 19, 23, 25, 26, 28, 32, 33, 34, 37, 38, 39, 43 Reed, S. H., 204, 209, 219 Reed, S. I., 80, 104, 359, 381 Rees, D. C., 245, 279 Regan, L., 161, 174, 175, 179, 183

466

AUTHOR INDEX

Reggiori, F., 359, 379 Reich, C. I., 202, 214 Reid, B. G., 350, 389, 422, 423, 424, 428, 429 Reid, E., 297, 301 Reid, G. A., 233, 235, 240 Reid, K. L., 9, 41 Reidt, U., 258, 279 Reiken, S., 269, 278 Reimer, U., 245, 270, 271, 279, 280 Reinkensmeier, G., 310, 327, 335 Reinstein, J., 6, 9, 10, 11, 12, 14, 16, 17, 18, 19, 20, 21, 27, 34, 37, 39, 40, 41, 42, 43, 163, 164, 167, 168, 169, 182, 185, 233, 239 Reithmeier, R. A. F., 322, 324, 339 Reitman, M., 309, 329, 335, 341 Remington, S. J., 10, 39 Renoir, J. M., 178, 180 Renoir, M., 157, 178, 182 Rensing, C., 285, 301 Repasky, E., 138, 155 Reuter, K., 258, 279 Reynolds, S. E., 207, 216 Rho, S. H., 418, 429 Rice, J., 257, 276 Rich, D. H., 256, 257, 270, 276, 278, 280 Richard, L. M., 83, 101 Richmond, C. S., 207, 208, 219, 349, 351, 386 Richter, A., 160, 181 Richter, K., 167, 168, 169, 185 Riehl, R. M., 178, 184 Rietsch, A., 287, 292, 293, 295, 296, 298, 301 Riezman, H., 359, 382 Riley, D. J., 157, 158, 181 Riley, M. I., 30, 37 Rimerman, R. A., 35, 41, 173, 175, 178, 179, 181, 183, 184, 261, 275 Rine, J., 208, 216, 355, 359, 360, 376, 379, 382 Risse, B., 29, 42 Ritter, C., 318, 336 Riviere, S., 244, 261, 276 Rizzolo, L. J., 319, 322, 336 Robbins, P. W., 304, 310, 331, 334, 335, 359, 385, 386, 387 Robert, K. J., 360, 386 Robert, P., 108, 114, 143, 151 Roberts, B. T., 96, 101 Roberts, C. J., 366, 371, 372, 377, 383 Roberts, R. M., 309, 335 Roberts, T. H., 355, 385

Robertson, H. A., 114, 115, 154 Robertson, H. M., 189, 190, 214 Robinson, C. V., 59, 66, 67, 70, 71 Robinson, W. G., Jr., 143, 148 Rocap, G., 419, 420, 429 Rocco, J., 330, 343 Roche, E. D., 202, 217, 420, 427 Rock, K. L., 189, 201, 203, 214, 219, 221 Rockabrand, D., 3, 43, 267, 281 Rockel, B., 203, 214, 418, 426 Rockwell, S., 269, 282 Rodan, A. R., 322, 323, 324, 327, 336 R¨oder, D., 3, 6, 41, 139, 153, 416, 428 Roder, J., 261, 272 Rodriguez, D., 141, 148 Rodriguez-Cerezo, E., 142, 154 Rodriguez-Gdiharpour, S., 262, 281 Roe, S. M., 159, 161, 162, 163, 164, 165, 166, 167, 168, 169, 171, 172, 174, 175, 176, 179, 183 Roemer, T. W., 359, 385 Roersma, E. S., 126, 155 Rogalla, T., 121, 126, 154 Rogers, S. W., 208, 219 Rohrwild, M., 189, 219, 414, 416, 417, 428 Rojas, A., 115, 154 Rokeach, L., 310, 311, 312, 336 Romagnoli, P., 330, 344 Romero, P. A., 306, 334 Rommelaere, H., 82, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 100, 102, 103, 104 Roncero, C., 359, 360, 388 Roobol, A., 78, 103 Roof, W. D., 261, 279 Roos, R. P., 272, 278 R¨osch, P., 251, 254, 278 Rose, M. D., 229, 230, 241, 242, 354, 359, 373, 386, 387, 389, 414, 426 Roseman, A. M., 48, 49, 50, 53, 54, 64, 69, 71, 72, 111, 125, 128, 130, 131, 133, 135, 152, 349, 380 Roseman, S., 10, 39 Rosemblit, N., 269, 278 Rosen, B. P., 285, 301 Rosen, C., 207, 216 Rosen, M. K., 258, 279 Rosen, N., 159, 162, 163, 164, 170, 171, 174, 184, 185 Rosenberg, A. H., 75, 103 Rosenstein, R., 253, 281

467

AUTHOR INDEX

Rosenwirth, B., 270, 281 Rospert, S., 9, 38, 267, 278 Ross, B. M., 414, 416, 429 Ross, C. A., 270, 275 Ross, D. T., 374, 383 Ross, S., 9, 38 Rossi, J., 115, 152 Rossi, M., 80, 100 Rossmeissl, P., 259, 282 Roth, J., 307, 331, 335, 343 Rothblatt, J., 15, 38 Rothblatt, J. A., 224, 227, 241 Rothman, J. E., 2, 12, 24, 38, 347, 386 Rothstein, L., 203, 219 Rothwarf, D. M., 250, 251, 275, 277, 280 Roumanie, O., 359, 384 Rouse, J., 126, 154 Rouviere, P. E., 259, 268, 279 Rouvinen, J., 193, 219 Rowland, M. B., 62, 63, 64, 71 Rowley, N., 377, 378, 386 Roy, S. K., 128, 142, 145, 153, 154 Ruan, Y., 261, 275 Rubin, C. S., 115, 142, 143, 152 Rubin, D. M., 188, 192, 204, 205, 207, 208, 209, 215, 216, 219, 221 Rubin, G., 157, 181 Rucknagel, ¨ K. P., 259, 260, 261, 265, 266, 279, 281 Rucknagel, P., 9, 38 Rudd, P. M., 323, 324, 327, 338 Rudiger, ¨ S., 3, 4, 6, 9, 10, 13, 14, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 31, 34, 37, 38, 40, 41, 42, 139, 153, 416, 428 Rudikoff, S., 415, 423, 427 Rudolph, R., 248, 262, 278 Rudy, D., 332, 344 Ruepp, A., 188, 202, 203, 219 Ruigrok, C. C. M., 239, 240 Rumswick, M. J., 207, 221 Runge, K. W., 359, 387 Ruppert, T., 212, 215 Russell, L. C., 161, 183 Russell, P. A., 157, 176, 180 Russell, R., 10, 11, 14, 16, 42 Russell, S. J., 204, 207, 209, 219, 326, 341 Russo, A. A., 159, 162, 163, 164, 165, 171, 176, 183, 185 Rutherford, S. L., 158, 183, 260, 279 Rutkat, K., 61, 71

Ryan, M. T., 223, 224, 240, 241 Ryan, S. L., 142, 149 Ryazantsev, S., 139, 148 Rybin, V., 286, 293, 295, 301 Rye, H. R., 46, 50, 72, 349, 389 Rye, H. S., 49, 50, 64, 68, 70, 72, 347, 387

S Sabatini, D. M., 270, 271, 280, 282 Sabewsan, S., 315, 336 Sacchettini, J. C., 315, 336 Sacher, M., 359, 387 Sadasivan, B., 324, 340 Sadis, S., 208, 209, 221 Saeki, M., 208, 220 Saenger, W., 251, 278 Saga, S., 126, 151 Saha, S. K., 159, 185 Sahil, R., 122, 152 Saibil, H. R., 46, 48, 49, 53, 54, 62, 63, 64, 69, 71, 72, 74, 77, 101, 106, 111, 123, 125, 128, 130, 131, 132, 133, 135, 144, 150, 152, 348, 349, 378, 380, 384, 389 Saint-Pol, A., 330, 342 Saito, A., 207, 221 Saito, Y., 328, 341 Sakaguchi, M., 229, 240 Sakai, R., 320, 337 Sakaki, Y., 115, 150 Sakane, F., 322, 324, 338, 339 Sakka, K., 359, 384 Salama, N. R., 359, 387 Saldanha, A. J., 368, 381 Salter, R. D., 322, 324, 328, 338 Salter-Cid, L., 212, 219 Salzmann, U., 196, 219 Sambrook, J. F., 2, 37, 127, 150, 229, 241, 346, 347, 355, 364, 382, 385 Sameshima, M., 169, 182 Samson, L. D., 204, 209, 217 Sanchez, E. R., 7, 41, 42, 178, 180, 184 Sanchez, Y., 349, 350, 386, 387, 414, 416, 428, 429 Sanders, S. L., 230, 241, 359, 387 Sandmeier, E., 9, 12, 14, 25, 41 Sandmeyer, S., 143, 150 Santarius, U., 189, 219, 416, 417, 428 Santi, D. V., 125, 153

468

AUTHOR INDEX

Santos, B., 359, 387 Santos-Martinez, M. L., 188, 202, 203, 219 Sanz, P., 227, 241 Saper, M. A., 351, 352, 379, 388 Saraste, M., 207, 221 Sarkar, A., 245, 281 Sarvas, M., 259, 277 Sasajima, H., 211, 220 Sasakawa, C., 286, 301 Sassa, H., 189, 198, 217, 220 Sastry, S., 158, 184 Sathyanarayana, U. G., 207, 219 Sato, K., 359, 387 Sato, M., 359, 387 Sato, S., 161, 177, 184 Sato, T., 32, 34, 43, 211, 215 Satoh, K., 211, 220 Sauer, R. T., 202, 217, 349, 384, 415, 418, 419, 420, 421, 425, 427, 428, 429 Sauk, J. J., 158, 181 Sausville, E., 164, 165, 171, 181 Sautiere, P., 255, 256, 272 Sawada, H., 211, 220 Saxena, E. S., 305, 334 Scarpa, A., 359, 387 Schachman, H. K., 125, 156 Schade, R., 205, 209, 211, 216, 217, 220 Schaeffer, J., 359, 384 Sch¨afer, R., 114, 151 Schaiff, W. T., 158, 184 Schalling, M., 160, 181 Scharf, K.-D., 108, 127, 153, 154 Scharl, E. C., 80, 82, 83, 99 Schatz, G., 223, 227, 229, 230, 231, 233, 235, 236, 239, 240, 241, 242, 267, 278 Schauber, C., 204, 219 Schauer, T. M., 190, 220 Scheibel, T., 162, 163, 164, 170, 182, 184 Scheider, C., 18, 20, 34, 42 Schekman, R., 15, 30, 38, 227, 228, 230, 231, 233, 237, 238, 239, 240, 241, 359, 360, 361, 369, 379, 381, 383, 384, 387, 390 Schelbert, B., 259, 260, 279 Schena, M., 348, 349, 351, 352, 365, 387 Scheraga, H. A., 250, 251, 275, 277, 280 Scherer, G., 245, 249, 253, 257, 264, 265, 270, 277, 279, 280 Scherer, P., 231, 240 Scherer, P. E., 229, 231, 241 Scherrer, L. C., 178, 179, 182

Scheuﬂer, C., 18, 20, 34, 35, 42, 161, 173, 184 Schiebel, E., 83, 88, 89, 90, 92, 100, 101, 103 Schieltz, D., 188, 205, 206, 209, 221, 359, 387 Schiene, C., 244, 269, 270, 278, 279 Schierhorn, A., 259, 260, 261, 270, 279, 280 Schild, H., 199, 200, 201, 212, 215, 216, 219, 220 Schilke, B., 2, 3, 4, 40, 41, 43 Schindler, T., 251, 253, 264, 275, 279, 280 Schirle, M., 199, 200, 201, 215, 219, 220 Schirmer, E. C., 413, 414, 429 Schirmer, R. E., 245, 280 Schirra, C., 14, 25, 38 Schlee, S., 6, 42 Schlenker, S., 203, 204, 217, 218 Schlenstedt, G., 29, 42 Schlesinger, M. J., 127, 149, 178, 184 Schlesinger, S., 328, 341 Schleyer, M., 224, 225, 241, 242 Schmid, D., 9, 10, 25, 42 Schmid, F. X., 2, 9, 40, 42, 244, 246, 247, 248, 250, 251, 252, 253, 254, 257, 258, 262, 263, 264, 265, 266, 267, 275, 276, 277, 278, 279, 280, 281, 282, 289, 300 Schmid, K., 235, 236, 240, 267, 278 Schmidt, B., 225, 241, 260, 278, 282 Schmidt, D., 18, 42 Schmidt, F. X., 66, 72 Schmidt, M., 188, 193, 196, 197, 198, 199, 204, 214, 219, 220, 221, 348, 387 Schmidtke, G., 193, 196, 197, 198, 200, 201, 220 Schmitt, M., 235, 241 Schnaider, T., 158, 181 Schnall, T., 209, 218 Schneider, C., 32, 37, 159, 162, 163, 164, 170, 171, 174, 184, 185 Schneider, H. C., 230, 231, 235, 236, 238, 239, 241 Schneider, M. D., 269, 280 Schneider, R. J., 127, 144, 149 Schneider-Mergener, J., 3, 4, 13, 14, 22, 31, 40, 42 Schneikert, J., 7, 32, 34, 38, 42 Schnur, R. C., 158, 171, 174, 180, 182, 184 Scholle, B., 83, 99 Scholz, C., 2, 40, 263, 264, 265, 266, 267, 276, 280, 282 Sch¨onbrunner, E. R., 258, 281

AUTHOR INDEX

Sch¨onfeld, H.-J., 4, 5, 9, 12, 14, 18, 19, 22, 25, 32, 34, 37, 38, 40, 41, 42, 111, 154, 263, 281 Sch¨onﬁsch, B., 235, 236, 238, 239, 241 Schoofs, G., 189, 220 Schramel, A., 348, 385 Schreiber, K. L., 328, 341 Schreiber, R. B., 321, 337 Schreiber, S. L., 178, 184, 185, 196, 215, 258, 259, 260, 276, 278, 279, 281 Schrimer, E. C., 349, 387 Schr¨oder, H., 4, 5, 9, 10, 15, 17, 18, 21, 22, 23, 24, 25, 26, 27, 37, 38, 40, 42, 233, 239 Schroter, C., 201, 220 Schubert, U., 203, 220 Schuh, S., 178, 184 Schuirema, R. A., 116, 155 Schuler, G., 374, 383 Schuler, H., 98, 103 Schulte, T. W., 163, 164, 165, 171, 172, 174, 181, 184 Schultz, D. A., 251, 280 Schulz, G. E., 245, 280 Schulze-Specking, A., 3, 38, 267, 275 Schumacher, R. J., 21, 38, 80, 82, 83, 99, 104 Schumann, T., 261, 277 Schumann, W., 2, 40 Schuster, H.-P., 9, 10, 11, 12, 16, 43 Schutkowski, M., 245, 252, 253, 257, 260, 270, 271, 277, 278, 279, 280, 282 Schutt, C. E., 98, 103 Schutz, A. R., 176, 184 Schwager, F., 414, 428 Schwaiger, E., 330, 343 Schwartz, A. L., 204, 206, 214 Schwartz, B. D., 158, 184 Schwartz, M. P., 223, 237, 240, 242, 422, 424, 428 Schwarz, E., 235, 241, 377, 378, 386 Schwarz, G., 201, 220 Schweiger, M., 360, 382 Scidmore, M. A., 230, 242 Scott, J. E., 329, 342 Scott, J. L., 202, 214 Scroggins, B. T., 7, 39 Sdicu, A. M., 359, 385 S´ebastein, M., 108, 154 Secco, C., 178, 183 Seckler, R., 289, 301 Sedbrook, J. C., 84, 104

469

Seeger, M., 205, 208, 209, 216, 220 Seemu¨ ller, E., 188, 193, 196, 199, 200, 204, 206, 209, 214, 215, 217, 220, 349, 379, 418, 429 Seetharam, R., 74, 104 Segel, G. B., 83, 101 Seidel, M., 425, 428 Seidel, R., 6, 11, 12, 16, 39, 40, 42 Sekerina, E., 259, 280 Sela, M., 246, 281, 283, 299, 301 Semenchenko, V., 30, 37 Semenza, J. C., 359, 387 Sen, A. C., 125, 155 Sendak, R. A., 250, 280 Seng, T. W., 261, 277 Seol, J. H., 189, 219, 414, 417, 428 Seong, I. S., 418, 429 Sepehrnia, B., 157, 184 Sepp-Lorenzino, L., 170, 171, 174, 184 Serrano, L., 262, 278 Serysheva, I., 112, 123, 128, 149 Sessa, W. C., 7, 38 Seternes, O. M., 261, 280 Seyfried, C. E., 321, 337 Seytter, T., 235, 241 Sha, B., 13, 31, 42 Shadidy, M., 261, 280 Shah, J., 322, 339 Shah, P. C., 269, 282 Shah, V., 7, 38 Shailubhai, K., 305, 334 Shaknovich, R., 170, 178, 184 Shakoori, A. R., 115, 154 Shalon, D., 348, 349, 351, 352, 365, 374, 383, 387 Shamu, C. E., 355, 356, 361, 366, 380, 387 Shank, P. R., 259, 271, 276 Shanklin, J., 414, 429 Shao, F., 296, 301 Shao, J., 7, 39 Shao, X., 262, 278 Shapiro, L., 202, 217 Sharkey, T. D., 142, 145, 150 Sharma, K. K., 111, 130, 133, 154 Sharma, S. V., 171, 172, 184 Shaw, A., 203, 222, 418, 429 Shearstone, J. R., 112, 125, 128, 154 Shen, J. W., 320, 337 Shen, M., 260, 271, 280, 282 Shenkman, M., 330, 344

470

AUTHOR INDEX

Sheraton, J., 359, 385 Sherlock, G., 364, 374, 388 Sherman, D. R., 143, 156 Sherman, F., 76, 83, 101, 103, 198, 217 Sherman, M. Y., 206, 218 Sherry, B., 270, 280 Shevchenko, A., 83, 90, 92, 103 Shewach, D. S., 7, 41 Shi, L., 9, 41 Shiau, K., 197, 214 Shibasaki, F., 320, 337 Shibmara, N., 213, 213 Shieh, B. H., 260, 281 Shilling, J., 229, 230, 239, 240 Shim, J., 359, 385 Shimbara, N., 212, 220 Shimizu, Y., 208, 220, 221 Shimma, Y. I., 360, 388 Shimotakahara, S., 251, 275 Shinohara, H., 114, 151 Shiotsu, Y., 172, 184 Shiozawa, T., 28, 43 Shirai, T., 359, 384 Shiroza, T., 414, 427 Shoemaker, D. D., 366, 371, 372, 377, 383 Shore, D., 108, 142, 156 Shore, G. C., 357, 359, 385 Shou, W., 269, 270, 271, 275, 276, 280 Shows, T. B., 321, 337 Shrader, T. E., 188, 217, 419, 420, 429 Shtilerman, M., 5, 42, 54, 66, 67, 68, 72 Shue, G., 170, 178, 184 Siddique, M., 108, 154 Sidrauski, C., 355, 363, 364, 367, 380, 382, 387 Siegers, K., 83, 88, 89, 90, 92, 100, 101, 103 Siegmund, H. I., 170, 184 Siekierka, J. J., 258, 280 Sielecki, A., 254, 281 Siezen, R. J., 124, 155 Sifers, R. N., 330, 331, 343, 344 Siffroi-Fernandez, S., 332, 343 Sigal, N. H., 258, 280 Siggia, E., 355, 385 Sigler, P. B., 46, 49, 50, 53, 55, 56, 68, 71, 72, 347, 349, 387, 390 Sijts, A., 196, 219 Silberg, J. J., 2, 11, 15, 16, 17, 28, 39, 42 Silberstein, S., 359, 387 Silhavy, T. J., 424, 428

Siligardi, G., 161, 162, 164, 167, 168, 169, 174, 175, 179, 183 Silver, P. A., 29, 42, 358, 370, 379 Silverstein, A. M., 7, 35, 37, 41, 161, 163, 171, 173, 177, 179, 181, 182, 184, 185 Simon, J., 366, 371, 372, 377, 383 Simon, M. I., 158, 180 Simon, S. M., 231, 242 Simons, J. F., 305, 322, 323, 324, 327, 334, 336 Simons, S. S., 179, 185 Simons, S. S., Jr., 179, 181 Simpson, R., 143, 156 Singer, A., 322, 330, 339 Singer, M. A., 7, 44, 349, 350, 386, 387, 413, 414, 416, 428, 429 Singer, T., 199, 217 Singh, S. K., 349, 350, 386, 388, 416, 417, 421, 422, 423, 424, 427, 428, 429 Sipos, G., 359, 379 Sirrenberg, C., 225, 231, 235, 239, 241 Sketon, N. J., 57, 58, 71 Skowronek, M. H., 238, 242 Skowyra, D., 4, 17, 42, 44, 350, 389, 415, 416, 419, 429 Slade, D., 366, 371, 372, 377, 383 Slaughter, C. A., 187, 204, 206, 207, 208, 212, 213, 215, 216, 218, 220, 221 Slepenkov, S. V., 10, 11, 16, 42 Sliekers, O., 416, 427 Slingsby, C., 107, 113, 116, 120, 124, 154, 155 Slipetz, D., 311, 312, 314, 336 Slonim, D., 371, 373, 374, 375, 380, 388 Smith, A. M., 360, 388 Smith, C. K., 415, 421, 429 Smith, D., 35, 41 Smith, D. F., 7, 8, 35, 42, 161, 173, 175, 178, 181, 182, 183, 184, 261, 275 Smith, D. J., 359, 388 Smith, D. L., 69, 71, 111, 133, 154 Smith, D. O., 175, 178, 179, 184 Smith, J. B., 111, 133, 154 Smith, J. L., 192, 214 Smith, P. L., 309, 335 Smith, P. R., 74, 101 Smith, T., 158, 181 Smulders, R. H. P. H., 112, 122, 133, 154, 155 Smykal, P., 128, 132, 154 Smyth, M. G., 84, 85, 102 Snapp, E. L., 355, 385

AUTHOR INDEX

Snavely, M., 18, 44 Snyder, M., 359, 387 Snyder, S. H., 270, 271, 275, 280, 282 Sobek, A., 190, 218 Sodroski, J., 270, 281 Soemitro, S., 259, 276 Soga, S., 171, 172, 184 Sogin, N. L., 189, 190, 214 Solary, E., 115, 143, 144, 148, 150 Solheim, J. C., 324, 340 S¨oling, H.-D., 323, 339 Soll, J., 225, 239, 241, 242 Solomon, F., 96, 98, 360, 388 Solter, L. F., 189, 190, 214 Sommer, S. S., 311, 312, 314, 336 Sommer, T., 359, 361, 381 Somoza, J. R., 255, 281 Sondek, S., 80, 82, 83, 99 Sondermann, H., 18, 20, 34, 42, 165, 176, 183 Sone, M., 291, 292, 294, 301 Song, H. K., 189, 214, 220, 417, 418, 426, 429 Song, H. Y., 158, 185 Song, J., 19, 32, 33, 37, 41, 43 Song, J. J., 418, 429 Song, J. L., 296, 300 Songyang, Z., 193, 217 Sorger, P. K., 320, 337 Sorrentino, R., 7, 38 Sorscher, E. J., 330, 342 Soti, C., 158, 181 Soto, A., 142, 154 Souchet, M., 108, 114, 143, 151 Soulie, S., 79, 86, 102 Sousa, M. C., 12, 22, 26, 39, 42, 189, 220, 308, 309, 310, 315, 316, 317, 318, 331, 335, 335, 336, 417, 418, 422, 429 Southard, S. B., 359, 385 Sowadski, J. M., 182 Soza, A., 212, 216 Spatrick, P., 360, 384 Spear, E. D., 209, 219, 361, 363, 385 Spector, A., 127, 136, 155 Spellman, P. T., 352, 364, 367, 371, 372, 374, 376, 381, 388 Spence, J., 157, 185 Spencer, D. B., 253, 281 Spendlove, I., 140, 152 Spenner, J. M., 237, 240 Sperling, J., 188, 204, 206, 209, 217 Spies, T., 324, 340

471

Spiro, R. G., 305, 308, 323, 330, 331, 334, 335, 339, 343 Spitzfaden, C., 258, 277, 280 Spouge, J. L., 207, 219, 415, 428 Sprang, S. R., 2, 37 Spreadbury, C. L., 116, 149 Squires, C., 414, 416, 427, 429 Srikakulam, R., 80, 103 Srinivasan, M., 320, 337 Srinivasan, P. R., 320, 337 Srivastava, P. K., 158, 182, 185 St. John, A. C., 423, 426 Staadluinen, A. A., 107, 151 Stachon, U., 188, 193, 196, 197, 198, 199, 216 Staeck, O., 12, 15, 22, 24, 25, 26, 40, 233, 234, 240, 241 Stafford, S. J., 285, 301 Stahl, J. M., 106, 150, 351, 352, 384 Staker, B. L., 351, 352, 379, 388 Stamnes, M. A., 260, 281 Stancato, L. F., 157, 161, 163, 171, 179, 183, 185 Stancovski, I., 206, 214 Standaert, R. F., 196, 215, 258, 259, 281 Standera, S., 196, 198, 216, 219 Standing, K. G., 212, 213, 221, 222 Staniforth, R. A., 53, 71 Stanley, B. A., 125, 149 Stanley, P., 329, 341 Staudenmann, W., 60, 71 Staudt, L. M., 374, 383 Stearns, T., 93, 94, 95, 96, 99, 101, 103 Stebbins, C. E., 159, 162, 163, 164, 171, 185 Steegborn, C., 253, 275 Steel, D. G., 249, 275 Stefano, F. D., 161, 171, 177, 181 Stege, G. J. J., 127, 151 Steger, K. A., 207, 219 Stehle, T., 122, 154 Steiger, R. H., 114, 151 Stein, G. S., 115, 154 Stein, J. L., 115, 154 Stein, R. L., 199, 203, 207, 213, 218, 219, 246, 259, 260, 276, 281 Steinbacher, S., 49, 71, 75, 85, 88, 96, 99, 102, 103, 349, 381 Steinberg, I. Z., 246, 281 Steiner, D. F., 322, 326, 339 Steiner, E., 176, 184 Steiner, J. P., 270, 271, 280, 282

472

AUTHOR INDEX

Steinert, M., 260, 282 Steinmann, B., 267, 281 Steinmann, D., 418, 428 Steinmetz, L., 375, 380 Steitz, T. A., 10, 38 Stenberg, G., 66, 72 Stensgard, B., 167, 175, 178, 182, 184, 185 Stepaniants, S. B., 366, 371, 372, 377, 383 Stepanova, L., 163, 173, 177, 185 Stern, D., 365, 380 Stern, P., 356, 361, 366, 380 Sternlicht, H., 5, 44, 74, 79, 86, 99, 104, 348, 378, 390 Sternlicht, M. L., 5, 44, 74, 79, 86, 99, 104, 348, 378, 390 Stetson, B., 304, 334 Stetter, K.-O., 49, 71, 75, 88, 99, 349, 381 Stevanovic, S., 199, 200, 201, 212, 215, 219, 220 Steven, A. C., 188, 189, 211, 214, 217, 222, 416, 417, 418, 421, 426, 427, 428 Stevens, R. C., 353, 379 Stevens, S. Y., 22, 23, 26, 27, 41 Stevens, T. H., 378, 388 Stevenson, S., 84, 102 Stewart, D. E., 245, 281 Stewart, E. J., 297, 301 Stewart, P. L., 112, 116, 122, 123, 141, 148, 150 Stirling, C. J., 354, 360, 378, 380, 389 Stitzel, J. D., 349, 350, 386, 414, 416, 428 Stock, D., 49, 71, 75, 88, 99, 190, 191, 192, 193, 200, 216, 218, 349, 381, 418, 427, 428, 429 Stocker, S., 188, 202, 203, 219 Stohwasser, R., 198, 216 Stokkermans, J., 145, 153 Stoldt, V., 76, 103 Stoller, G., 261, 265, 266, 279, 280, 281, 282 Stoller, T., 330, 343 Stoltze, L., 201, 220 Stone, D. E., 227, 242, 370, 377, 378, 389 Stone, K. D., 286, 300 Storz, G., 352, 364, 381, 424, 428 Stoughton, R., 366, 371, 372, 377, 383 Strickland, E., 204, 220 Striker, R., 286, 300 Stromer, T., 106, 123, 125, 130, 132, 144, 150 Stroud, R. M., 125, 153 Stroynowski, Y., 328, 341

Strynadka, N. C., 254, 281 Stuart, R. A., 227, 242 Studer, S., 114, 154 Studier, F. W., 75, 103 Stukenberg, P. T., 260, 271, 280, 282 Stumpf, G., 258, 271, 276 Styles, C. A., 368, 381 Su, K., 330, 343 Suarez, J. D., 271, 282 Subjeck, J. R., 9, 38, 138, 155, 320, 337 Subramaniam, S., 189, 190, 214 Suck, D., 10, 39 Sugiyama, Y., 108, 114, 144, 154, 155 Suh, P., 319, 336 Suh, W.-C., 15, 42 Suhan, J. P., 127, 148 Sukhaswami, M. B., 132, 154 Sullivan, D. S., 79, 83, 84, 99 Sullivan, M. A., 163, 182 Sullivan, W. P., 7, 35, 42, 161, 164, 165, 167, 170, 171, 173, 178, 181, 182, 184, 185 Sultmann, H., 212, 218 Sumegi, M., 188, 204, 206, 209, 217 Sun, J. L., 261, 279 Sun, L., 330, 342, 355, 389 Sun, T.-S., 112, 119, 124, 128, 154 Sun, T.-X., 124, 154 Sun, X. X., 295, 301 Sundaram, S., 329, 341 Sundby, C., 142, 145, 150 Sundquist, W. I., 270, 281, 282 Suntio, T., 224, 241 Supertifurga, A., 267, 281 Surewicz, W. K., 125, 132, 149 Surolia, A., 323, 324, 339 Susek, R. E., 142, 154 Susin, S. A., 115, 143, 144, 148 Susskind, M. M., 420, 427 Sutoh, K., 254, 281 Sutton, G. G., 202, 214 Sutton, S. U., 79, 101 Suzdak, P. D., 270, 280 Suzuki, A., 108, 114, 144, 154, 155 Suzuki, C., 360, 388 Suzuki, M., 202, 207, 217, 221, 257, 258, 268, 277, 281 Suzuki, N., 142, 145, 153 Suzuki, T., 330, 342 Svensson, M., 249, 262, 276 Svergun, D. I., 121, 149

AUTHOR INDEX

Swafﬁeld, J. C., 204, 206, 207, 215, 220, 415, 429 Swaminathan, S., 208, 215 Sweder, K., 205, 209, 218 Swenson, K. I., 271, 282 Swindells, M. B., 159, 185 Swords, W. E., 424, 426 Szabo, A., 4, 5, 30, 42 Szalay, A. A., 267, 278 Szebenyi, G., 30, 43 Szyperski, T., 13, 15, 30, 41, 43, 353, 385

T Tabak, H. E., 239, 240 Tachibana, C., 369, 378, 388 Tachikawa, H., 378, 388 Tai, P. K., 178, 185 Taillandier, D., 208, 216 Takagi, T., 161, 183, 211, 221 Takahashi, E., 207, 208, 221 Takahashi, M., 171, 185 Takahashi, N., 248, 257, 258, 262, 264, 268, 277, 278, 281 Takahashi, S., 251, 279 Takaoka, M., 198, 217 Takayama, S., 9, 18, 19, 20, 32, 33, 34, 37, 38, 39, 43 Takeda, S., 12, 22, 25, 26, 39, 43 Takenawa, T., 320, 337 Takeuchi, T., 171, 181 Takeuchi, Y., 369, 378, 388 Tam, Y., 115, 150 Tamaoki, T., 172, 184 Tamayo, P., 371, 373, 374, 375, 380, 388 Tammi, M., 304, 334 Tampe, R., 199, 215 Tamura, N., 189, 201, 202, 220, 221 Tamura, T., 187, 201, 202, 203, 211, 218, 220, 221 Tan, M. C., 158, 183 Tanahashi, N., 204, 207, 208, 211, 212, 213, 213, 215, 218, 220, 221 Tanaka, K., 187, 188, 189, 198, 204, 207, 208, 209, 211, 212, 213, 213, 214, 215, 216, 217, 218, 220, 221 Tanaka, N., 72, 158, 182 Tanaka, R., 26, 43 Tanaka, S., 198, 217

473

Tanaka, T., 159, 185, 320, 337 Tanfani, F., 30, 37 Tang, J., 309, 335 Tang, S. C., 34, 44 Tanguay, R. M., 108, 115, 139, 144, 151, 152, 153, 154, 155 Taniguchi, M., 171, 185 Tanner, W., 359, 382 Tanzer, M. L., 144, 156 Tap, W. D., 79, 80, 81, 82, 94, 98, 103, 104 Tateishi, J., 115, 150 Tatsuta, T., 14, 43 Tatu, U., 321, 338 Taulien, J., 157, 180, 349, 387, 416, 429 Taylor, P., 322, 339 Taylor, R. K., 286, 301 Taylor, S. S., 182 Tector, M., 322, 324, 328, 338 te Heesen, S., 307, 335 Tempst, P., 5, 38, 75, 100, 348, 349, 378, 381, 385 Ten Eyck, L. F., 182 Teo, T.-S., 209, 218 Terada, K., 28, 43, 228, 242 TerBush, D. R., 359, 388 Tersmette, M., 329, 342 Tessier, D. C., 309, 310, 311, 312, 326, 336, 341 Teter, S. A., 3, 43, 267, 281 Tettelin, H., 349, 382 Texter, F. L., 66, 70, 71, 253, 281 Thalhammer, T., 261, 277 Thali, M., 270, 281 Theodorakis, N. G., 9, 43 Th´eriault, J. R., 144, 155 Thevenet, D., 420, 427 Theyssen, H., 9, 10, 11, 12, 16, 21, 37, 41, 43, 233, 239 Thickman, K., 32, 38 Thiele, B. R., 255, 276 Thiele, D. J., 176, 182 Thierfelder, J. M., 127, 151 Thies, M. J. W., 248, 281 Thirumalai, D., 67, 72 Thiyagarajan, P., 48, 72 Thomas, C. J., 323, 324, 339 Thomas, D. Y., 309, 310, 311, 312, 321, 323, 324, 326, 327, 336, 337, 338, 341 Thomas, J. G., 142, 155 Thomas, J. O., 74, 104, 348, 378, 381

474

AUTHOR INDEX

Thomas, P. J., 204, 220 Thomas, R. L., 75, 76, 93, 100 Thompson, L. J., 36, 37, 38, 206, 214 Thompson, M. W., 201, 220, 349, 388 Thomson, J. A., 124, 155, 251, 258, 279 Thony-Meyer, L., 297, 300 Thornton, J. M., 245, 278, 353, 388 Thress, K., 19, 33, 43 Thrower, J. S., 205, 206, 220 Thulasiraman, V., 7, 8, 43, 80, 92, 99, 103 Thulin, E., 255, 275 Tian, G., 79, 80, 81, 82, 93, 94, 95, 96, 97, 98, 101, 103, 104 Tilly, K., 14, 31, 37, 74, 100, 347, 382 Timpl, R., 262, 275 Tipper, C., 360, 384 Tissieres, A., 3, 41 Titu, H. N., 330, 332, 342 Tjoeler, L. W., 321, 337 To, W. Y., 213, 220 Tobe, T., 286, 301 Tobias, J. W., 419, 420, 429 Tobind, F. L., 108, 114, 143, 151 Toda, T., 96, 100, 103 Todd, A. E., 353, 388 Todd, M. J., 67, 72, 348, 387 Toes, R. E. M., 199, 215 Toft, D. O., 4, 7, 35, 42, 43, 158, 161, 164, 165, 167, 170, 171, 173, 174, 175, 178, 181, 182, 184, 185, 261, 276 Tohe, A., 208, 209, 215, 220, 221 Tohma, S., 213, 218 Tokunaga, F., 211, 221 Tokunaga, H., 26, 43 Tokunaga, M., 26, 43 Tomb, J. F., 202, 214 Tomchick, D. R., 192, 214 Tom´e, F., 115, 140, 155 Tomomori, C., 159, 185 Tomoyasu, T., 3, 6, 14, 25, 38, 39, 41, 43, 139, 153, 267, 275, 350, 382, 416, 428 Tong, K. I., 159, 185 Tongaonkar, P., 204, 219 Tonin, P. N., 320, 337 Topping, T. B., 124, 155 Torii, N., 207, 211, 222 T¨or¨ok, Z., 125, 127, 130, 135, 144, 145, 150, 155 Torronen, A., 286, 293, 295, 301 Toshimori, K., 329, 342

Toth, C. R., 198, 221 Toussaint, A., 415, 419, 428 Townsend, A., 198, 214 Toyofuku, K., 322, 339 Tradler, T., 3, 43, 265, 266, 267, 269, 270, 278, 281, 282 Trame, C. B., 189, 220, 417, 418, 422, 429 Travers, K. J., 299, 301, 356, 358, 360, 362, 367, 370, 377, 386, 388, 631 Treadwell, J., 359, 385 Trent, J. D., 75, 104, 347, 348, 388 Trent, J. M., 374, 383 Trepel, J., 158, 165, 171, 180, 181 Trigon, S., 126, 154 Trilla, J. A., 359, 360, 388 Trimble, R. B., 306, 331, 334 Trivedi, V. D., 132, 154 Trombetta, E. S., 305, 309, 310, 311, 315, 318, 322, 323, 324, 326, 327, 329, 331, 334, 335, 336, 339 Trombetta, S., 306, 309, 310, 334 Trombetta, S. E., 307, 309, 310, 335 Tropschug, M., 224, 242, 258, 267, 272, 277, 278, 279, 281 Trowbridge, L. S., 329, 341 Trowsdale, J., 198, 214 Truckses, D. M., 255, 281 Trus, B. L., 416, 417, 427 Truscott, R. J. W., 134, 148 Tsai, M.-Y., 30, 43 Tsai, S., 161, 180 Tsichlis, P. N., 173, 177, 181 Tsui, H. C., 352, 389 Tsui, S. K., 114, 154 Tsurumi, C., 208, 220, 221 Tsuruo, T., 161, 177, 184 Tsuruta, H., 189, 220, 417, 418, 422, 429 Tsvetkova, N. M., 125, 130, 135, 144, 145, 155 Tulp, A., 329, 342 Turco, S. J., 304, 334 Turner, G., 187, 221 Tyedmers, J., 238, 242 Typke, D., 211, 221 Tyson, J. R., 360, 389

U Udgaonkar, J. B., 250, 281 Udono, H., 158, 182

AUTHOR INDEX

Udvardy, A., 188, 204, 206, 207, 208, 209, 216, 217, 220 Ueguchi, C., 28, 43 Uehara, Y., 171, 181, 185 Ueling, D. E., 258, 276 Uenaka, A., 158, 182 Uerkvitz, W., 190, 218 Ullrich, T., 196, 199, 216 Ulrich, H. D., 203, 204, 217, 218 Ulrich, P., 270, 280 Um, S. J., 416, 427 Uma, S., 7, 8, 43 Ungermann, C., 226, 235, 236, 240, 242 Ungewell, H., 26, 28, 43 Ungewickell, E., 3, 26, 28, 43 Urban, J., 361, 381 Ursic, D., 5, 43, 79, 83, 84, 104, 208, 215 Ustrell, V., 189, 192, 208, 212, 215, 219

V Vainberg, I. E., 79, 80, 81, 82, 88, 89, 90, 91, 92, 94, 98, 103, 104 Vainberg, R. L., 93, 100 Vajdos, F. F., 270, 281 Vale, R. D., 9, 41 Valencia, A., 15, 18, 27, 37, 38 Valent, Q. A., 265, 281 Vall´e, F., 306, 334 Valpuesta, J. M., 84, 85, 86, 87, 98, 102 van Bleek, G. M., 158, 183 van Boekel, M. A. M., 108, 132, 134, 148, 149, 155 Vancompernolle, K., 127, 139, 153 van dan Ijssel, P., 106, 126, 155 van de Boogaart, P., 107, 151 Vandekerckhove, J., 82, 87, 88, 89, 90, 91, 92, 93, 94, 96, 100, 102, 103, 104, 127, 139, 153 van de Klundert, F. A. J. M., 112, 122, 127, 155 VanDemark, A. P., 192, 219 van den Berg, M., 239, 240 Vanden Heuvel, J. P., 174, 182 Van den Ijssel, P., 140, 141, 149, 153 van den Oetelaar, P. J. M., 124, 155 Van der Eb, A., 114, 156 van der Klei, I., 227, 242 Vander Kooi, C. W., 22, 23, 26, 27, 41 Vanderleyden, J., 203, 218

475

van der Vies, S. M., 74, 100, 347, 382 Van der Wal, F., 326, 340 van der Zandt, H., 121, 149 Van Doren, S. R., 30, 37 Van Duyne, G. D., 258, 259, 281 van Dyck, L., 414, 429 Van Dyk, T. K., 347, 389 van Heel, M., 203, 222, 418, 429 Vanhove, M., 254, 281 Van Leeuwen, J. E. M., 322, 324, 328, 339, 340, 341 Van Montagu, M., 259, 272, 278 van Montfort, R. L. M., 107, 113, 116, 120, 155 van Nocker, S., 208, 209, 215, 221 van Noort, J. M., 115, 155 Van Sechel, A. C., 115, 155 van Someren, P. F. H. M., 124, 155 van Wilpe, S., 225, 241 Vardy, L., 96, 103 Varma, G., 305, 334 Varoutas, P. C., 163, 180 Varshavsky, A., 187, 199, 203, 205, 209, 216, 217, 219, 221, 419, 420, 429 Varvasovszki, V., 125, 127, 130, 135, 144, 145, 150, 155 Vasilikiotis, C., 17, 44 Vassilakos, A., 321, 322, 323, 324, 326, 328, 337, 338, 339 Veeraraghavan, S., 262, 264, 281 Vega, L. R., 96, 98 Veinger, L., 4, 43, 112, 128, 132, 135, 155 Veit, T., 16, 39, 167, 168, 169, 185 Veldscholte, J., 178, 185 Velten, M., 34, 43 Venema, G., 259, 276 Venter, J. C., 28, 43, 202, 214 Verbon, A., 116, 155 Verdecia, M. A., 272, 281 Verlaan-DeVries, M., 114, 156 Verma, R., 188, 204, 205, 206, 209, 221 Verschuure, P., 107, 151 Vestweber, D., 229, 241 Vetter, I., 209, 218 Vicart, P., 115, 140, 155 Vickery, L. E., 2, 11, 15, 16, 17, 28, 39, 42 Vierling, E., 106, 107, 108, 111, 113, 114, 115, 116, 120, 125, 127, 128, 130, 131, 133, 135, 136, 137, 138, 144, 145, 150, 152, 153, 154, 155

476

AUTHOR INDEX

Vierstra, R. D., 190, 208, 209, 215, 216, 221 Vigh, L., 125, 127, 130, 135, 144, 145, 148, 150, 155 Viitanen, P. V., 74, 102, 104, 347, 348, 387, 389 Vijay, A. K., 305, 334 Villoutreix, B. O., 34, 43 Vinci, F., 290, 300 Vinh, D. B., 79, 83, 84, 104 Vionnet, C., 359, 379 Virden, R., 254, 281, 282 Virgalitta, G., 233, 239 Virgallita, G., 21, 37 Viswanath, R., 125, 149 Vitiello, A., 212, 219 Vo, D. H., 359, 385 Voellmy, R., 115, 147 Vogel, J. L., 416, 429 Vogel, J. P., 229, 230, 241, 242, 354, 359, 373, 386, 387, 389 Voges, D., 187, 221 Voglino, L., 158, 162, 185 Voisine, C., 2, 43, 230, 231, 235, 237, 242 Volk, L., 7, 39 Volker, K. K., 188, 202, 203, 219 Volkmer, J., 238, 242 Volkwein, C., 361, 381 von Ahsen, O., 230, 231, 233, 235, 236, 237, 242, 263, 281 von Heijne, G., 304, 333 Von Herrath, M., 212, 219 Voorter, C. E. M., 106, 114, 126, 128, 130, 153, 155 Voos, W., 226, 229, 230, 231, 233, 235, 236, 237, 238, 239, 240, 241, 242 Vooter, C. E., 106, 149 Vos, M. H., 174, 175, 176, 183, 377, 385 Vothknecht, U. C., 239, 242

W Wabl, M., 320, 337 Wachter, C., 227, 233, 235, 240, 242 Wachter, E., 258, 281 Wada, C., 414, 428 Wada, I., 322, 324, 329, 338, 339, 342 Wagner, J., 311, 312, 314, 336 Wagner, R., 223, 240 Wakayama, T., 114, 156

Wakim, N. G., 178, 185 Walden, P. D., 82, 93, 96, 100 Waldmann, T., 76, 83, 90, 92, 103, 104 Walensky, L. D., 271, 282 Wali Karzai, A., 10, 11, 14, 42 Walke, S., 106, 123, 125, 130, 132, 135, 144, 148, 150 Walker, A. I., 160, 185 Walker, G., 9, 40, 207, 216 Walker, G. C., 420, 427 Walker, J. E., 207, 221 Walkinshaw, M. D., 258, 277, 279, 280 Wall, D., 12, 13, 14, 15, 30, 37, 40, 41, 43 Wall, J. S., 5, 38, 75, 100, 104, 347, 348, 378, 381, 388 Wallace, B. J., 292, 301 Wallace, M., 208, 221 Waller, P. R., 202, 217, 419, 420, 427 Walsh, C. T., 125, 149, 163, 182, 258, 268, 270, 277, 278, 281 Walsh, D., 115, 155 Walsh, M. T., 125, 155 Walter, G., 127, 153 Walter, P., 209, 219, 224, 227, 240, 242, 355, 356, 358, 360, 361, 362, 363, 364, 365, 366, 367, 370, 380, 382, 385, 387, 388, 631 Walter, S., 66, 69, 71, 72 Walter, W. A., 2, 3, 41, 259, 282 Walton, M., 161, 171, 177, 181 Walz, J., 188, 189, 201, 203, 211, 212, 214, 218, 221, 349, 379 Wampler, J. E., 245, 281 Wandless, T. J., 258, 279 Wang, C., 365, 384 Wang, C. C., 192, 198, 207, 213, 220, 221, 295, 296, 300, 301 Wang, E. W., 189, 202, 221 Wang, H., 23, 25, 26, 28, 41, 43 Wang, J., 53, 71, 413, 415, 417, 418, 422, 428, 429 Wang, K. Y., 127, 136, 155 Wang, L., 419, 429 Wang, T. W., 269, 270, 282 Wang, W., 207, 221 Wang, X. F., 270, 275 Wang, X. Y., 138, 155 Wang, Y., 296, 300 Wang, Z., 57, 58, 59, 72 Ward, T. R., 366, 371, 372, 377, 383

AUTHOR INDEX

Ware, F. E., 323, 324, 328, 339 Warth, R., 160, 182 Watanabe, D., 329, 342 Watanabe, K., 202, 217 Watanabe, M., 330, 343 Watanabe, T. K., 207, 221 Watanabe, Y., 416, 428 Watanable, R., 359, 389 Watarai, M., 286, 301 Waters, E., 106, 108, 115, 135, 155 Waters, M. G., 224, 227, 228, 239, 242 Watkins, S., 322, 338 Watson, J. D., 116, 148 Wawrousek, E. F., 143, 148 Wawrzynow, A., 12, 14, 25, 28, 30, 37, 38, 43, 413, 416, 427, 429 Waye, M. M., 114, 154 Wearsch, P. A., 158, 162, 185 Webb, M. R., 52, 53, 71 Weber, H. P., 258, 280 Weber, L. A., 115, 126, 127, 139, 152, 154 Weber-Ban, E. U., 188, 217, 350, 389, 413, 422, 423, 424, 427, 428, 429 Wedemeyer, W. J., 250, 280 Wefes, I., 188, 204, 205, 208, 209, 216, 221 Wehmeyer, N., 115, 155 Weichselbaum, R., 115, 143, 153 Weikl, T., 162, 167, 168, 169, 170, 184, 185, 261, 275 Weinstock, K. G., 202, 214 Weis, W. I., 273, 279, 418, 428 Weissman, A. M., 204, 214, 320, 330, 337, 343 Weissman, I., 260, 278 Weissman, J. D., 322, 324, 329, 330, 339 Weissman, J. S., 70, 72, 299, 301, 349, 356, 358, 360, 362, 367, 370, 377, 386, 388, 389, 631 Welch, W. J., 3, 9, 39, 41, 91, 92, 99, 100, 115, 127, 139, 148, 153, 157, 178, 180, 182, 184, 185 Weleber, R. G., 140, 152 Welsh, M. J., 106, 156 Welty, D. J., 415, 421, 427, 429 Weng, S., 305, 331, 334 Wente, S. R., 125, 156 Wenzel, T., 192, 199, 200, 221 Werner, E. D., 355, 389 Werner, J. M., 282 Werner-Washburne, M., 3, 41, 227, 228, 229, 239, 242, 370, 377, 378, 389

477

Wernstedt, C., 22, 26, 39 Westermann, B., 2, 40, 377, 378, 386 Wetzker, R., 269, 270, 278 Wheeler, K. A., 254, 282 Whitaker, H. C., 78, 103 Whitby, F. G., 192, 207, 212, 213, 217, 221 White, A. M., 359, 385 White, E. H., 46, 72 White, F. A., 115, 152 White, F. H., 283, 299, 301 White, H. E., 48, 49, 54, 72, 106, 123, 125, 130, 132, 144, 150 White, O., 202, 214 Whiteheart, S. W., 418, 428 Whitelaw, M. L., 7, 41, 178, 182, 185 Whitesell, L., 158, 171, 175, 176, 178, 179, 180, 182, 184, 185 Whitﬁeld, K. M., 230, 241, 359, 387 Whitlock, J. P., Jr., 174, 182 Whitt, S. R., 161, 183 Wickner, S., 4, 17, 42, 44, 188, 206, 221, 349, 350, 382, 383, 386, 389, 413, 415, 416, 419, 420, 421, 422, 423, 424, 425, 427, 428, 429 Wickner, W., 264, 265, 275, 276, 278 Wider, G., 258, 277, 280 Widmer, H., 258, 277, 280 Wiebel, F. F., 360, 389 Wiech, H., 157, 176, 185 Wiedmann, M., 2, 3, 39, 41, 230, 241 Wiegand, G., 197, 216 Wienhues, U., 224, 242 Wiertz, E. J., 330, 342, 355, 389 Wieske, M., 121, 126, 139, 148, 154 Wigley, D. B., 159, 163, 170, 185 Wikstrom, A. C., 157, 185 Wilbanks, S. M., 12, 22, 26, 39, 164, 181, 189, 220, 417, 418, 422, 429 Wilchek, M., 127, 139, 153 Wilcox, C. B., 272, 282 Wild, J., 259, 282 Wilhelmsson, A., 157, 185 Wilk, S., 199, 219 Wilkinson, C. R. M., 7, 38, 188, 207, 208, 209, 215, 221 Wilkinson, K. D., 188, 209, 221 Willbold, D., 251, 254, 278 Williams, D. B., 230, 241, 321, 322, 323, 324, 326, 328, 330, 337, 338, 339, 340, 341 Williams, D. L., 116, 148

478

AUTHOR INDEX

Williams, G. A., 134, 148 Williams, G. T., 9, 43 Williams, K. P., 202, 217 Williams, R. C., Jr., 79, 86, 102 Willison, K. R., 73, 74, 76, 77, 78, 79, 84, 85, 86, 87, 98, 101, 102, 348, 378, 383, 384 Willy, P. J., 213, 213 Wilm, M., 202, 215 Wilsbech, M., 414, 415, 426 Wilson, C. M., 330, 343 Wilson, H. L., 188, 189, 197, 202, 218, 221 Winchester, D., 32, 37 Winey, M., 176, 184, 359, 381 Winkelhaus, S., 127, 151 Winkelmann, D. A., 80, 103 Winkler, J., 365, 380 Winkler, K. E., 271, 282 Winkler, M. E., 352, 389 Winquvist, O., 212, 219 Wintermeyer, E., 260, 278, 282 Winther, J. R., 288, 300, 304, 332, 334, 343 Winzeler, E. A., 375, 380 Wisniewski, T., 140, 156 Wistow, G., 139, 156 Witt, E., 199, 221 Witt, S. N., 10, 11, 16, 19, 40, 42 Wittmann-Liebold, B., 257, 276 Wlodawer, A., 193, 215 Wodicka, L., 356, 358, 360, 362, 365, 367, 370, 375, 380, 388, 389, 631 Wojtkowiak, D., 414, 416, 429 Wolf, D. H., 188, 193, 196, 197, 198, 199, 200, 204, 211, 215, 216, 217, 219, 320, 330, 331, 337, 344, 360, 361, 378, 382, 384 Wolf, S., 203, 221 Wolfe, T. C., 212, 219 Wolfner, M. F., 115, 150 Wolfsberg, T. G., 375, 380 Wollner, S., 257, 270, 277, 280 Won, K. A., 80, 104 Wong, M. L., 198, 213, 221 Woo, K. M., 199, 200, 202, 217, 222, 415, 416, 420, 422, 427, 428 Wood, S. P., 48, 54, 71, 349, 380 Woodbury, R. L., 124, 155 Woodgate, R., 419, 421, 427 Woodman, P. G., 193, 196, 218 Woodruff, R. V., 418, 428 Woodward, E., 198, 218 Woolf, C. J., 115, 152

Woolford, C., 74, 100, 347, 382 Woolfson, D. N., 161, 174, 175, 179, 183 Wootton, D., 108, 142, 156 Workman, P., 161, 171, 177, 181 Worland, P. J., 176, 182 Wormald, M. R., 310, 324, 327, 332, 335, 340, 343 Worthylake, D., 10, 39 Wouters, D., 255, 256, 272 Wrba, A., 250, 280 Wright, P. E., 13, 30, 31, 40 Wu, B., 18, 44 Wu, G. J. S., 132, 141, 142, 149, 153 Wu, H. Y., 349, 385 Wu, J. W., 255, 282 Wu, W., 189, 217, 416, 417, 427 Wu, X. Y., 272, 272, 282 Wu, Y., 36, 37, 38, 115, 155, 206, 212, 214, 219 Wulf, G., 244, 272, 278 Wulﬁng, ¨ C., 272, 282 Wunderlich, M., 287, 289, 300, 301 Wurtz, M., 74, 101 Wuthrich, ¨ K., 13, 15, 30, 41, 43, 245, 258, 276, 277, 280 Wyman, C., 17, 44 Wyns, L., 273, 275

X Xanthoudakis, S., 143, 156 Xie, A., 30, 37 Xie, T., 291, 300 Xie, Y. M., 187, 205, 221 Xie, Z., 9, 19, 32, 33, 43 Xu, J., 260, 271, 282 Xu, M., 179, 185 Xu, W., 206, 218 Xu, Y., 7, 44, 328, 341 Xu, Z., 46, 49, 50, 55, 68, 72, 286, 300, 347, 349, 387, 390 Xuong, N.-H., 182

Y Yaffe, M. B., 5, 44, 74, 79, 104, 260, 271, 282, 348, 378, 390 Yaffe, M. P., 228, 239

479

AUTHOR INDEX

Yagodnik, C., 304, 333 Yahara, I., 169, 182 Yamada, H., 28, 43 Yamamoto, C., 171, 181 Yamamoto, K., 213, 218 Yamamoto, K. R., 7, 41 Yamamoto, M., 114, 156 Yamamoto, Y., 202, 217 Yamano, T., 158, 182 Yamasaki, M., 208, 220, 221 Yamazaki, M., 202, 217 Yamazaki, T., 159, 185 Yan, W., 28, 29, 44 Yan, Y., 122, 152, 154 Yanagi, H., 77, 101, 266, 279, 355, 359, 365, 383, 384, 385 Yanagi, S., 77, 104 Yang, C. F., 80, 92, 103 Yang, H., 209, 218 Yang, H. M., 125, 132, 156 Yang, H. Y., 209, 218 Yang, J., 271, 275 Yang, M., 330, 343 Yang, R., 365, 380 Yang, W., 159, 168, 180 Yang, X., 34, 44 Yang, Y., 198, 212, 215, 219, 221 Yao, D., 270, 282 Yao, Y., 192, 198, 207, 213, 221 Yaron, O., 134, 141, 148, 150 Yasuda, K., 9, 44 Yates, J., 188, 205, 206, 209, 221 Yates, J. R. III, 132, 153, 359, 387 Yazaki, Y., 320, 337 Ybarra, J., 53, 71 Yeager, M., 190, 217 Yeh, C.-H., 128, 142, 144, 156 Yeh, C.-y., 270, 282 Yeh, K. W., 142, 156 Yeh, Y. C., 128, 151 Yem, A. W., 161, 173, 183, 261, 279 Yeung, T., 359, 387 Yewdell, J. W., 203, 220, 221 Yifrach, O., 48, 52, 53, 72 Yin, L.-Y., 36, 37 Yohda, M., 416, 428 Yokosawa, H., 208, 209, 211, 215, 220, 221 Yokota, H., 128, 151 Yokota, K., 208, 221 Yokota, S., 77, 101, 104

Yokoyama, K., 161, 183 Yoneda, Y., 228, 242 Yonemoto, W., 178, 184 Yoo, S., 270, 281, 282 Yoo, S. J., 189, 219, 414, 417, 428 York, I. A., 201, 221 Yoshida, H., 355, 365, 385 Yoshida, M., 108, 144, 155, 416, 428 Yoshida, S., 114, 154 Yoshida, T., 80, 100 Yoshikawa, M., 286, 301 Yoshimura, K., 369, 381 Yoshimura, T., 211, 221 Yoshioka, S., 414, 428 Young, D. B., 116, 155 Young, H. A., 261, 272 Young, I. G., 292, 301 Young, J., 329, 341 Young, J. C., 160, 161, 162, 170, 175, 185 Young, K., 261, 278 Young, K. D., 261, 268, 277, 279 Young, L.-S., 128, 144, 156 Young, P., 208, 212, 218, 222 Young, R., 261, 278, 279 Younger, J. M., 36, 40, 206, 218 Yu, C. A., 291, 300 Yu, J., 286, 301 Yu, R. C., 418, 429 Yu, W. L., 359, 383 Yu, Y. Y. L., 324, 340 Yuan, H. E. H., 270, 276 Yuan, X., 255, 282 Yuan, Y., 143, 156 Yumiko, E., 12, 38 Yura, T., 77, 101, 104, 266, 279, 355, 359, 365, 383, 384, 385, 414, 428

Z Zaal, K. J., 355, 385 Zabala, J. C., 93, 95, 96, 99, 100, 102, 104 Zagnitko, O. P., 204, 206, 215 Zahn, R., 55, 56, 57, 66, 71, 72 Zaitsu, K., 7, 35, 42, 178, 184 Zamanillo, D., 126, 154 Zand, D. J., 9, 43 Zander, T., 288, 289, 291, 300, 351, 384 Zantema, A., 114, 156 Zantopf, D., 188, 199, 220, 221

480

AUTHOR INDEX

Zapun, A., 287, 301, 309, 310, 311, 312, 323, 324, 326, 327, 336, 338, 341 Zarnt, T., 264, 265, 266, 280, 282 Z´arsky, V., 128, 132, 154 Zav’yalov, G. A., 119, 149 Zav’yalov, V. P., 119, 149 Zeilstra-Ryalls, J., 4, 44 Zeiner, M., 32, 34, 38, 44 Zemsky, J., 330, 343 Zeng, B. F., 261, 282 Zeng, L., 2, 25, 39 Zervos, A. S., 269, 282 Zhang, H. Z., 286, 300 Zhang, J.-X., 324, 325, 329, 340 Zhang, L., 359, 387 Zhang, M. Q., 364, 374, 388 Zhang, O., 13, 30, 31, 40 Zhang, Q., 322, 324, 328, 338 Zhang, S., 296, 300 Zhang, W., 36, 40 Zhang, W. Y., 206, 218 Zhang, W. Z., 145, 156 Zhang, X. D., 203, 222, 418, 429 Zhang, Y. X., 158, 185, 261, 275 Zhang, Z. G., 207, 211, 212, 213, 217, 222 Zhao, X., 2, 15, 20, 21, 25, 39, 42, 44 Zhao, Y. D., 270, 282 Zheng, B., 414, 427 Zheng, C. X., 125, 132, 156 Zheng, D., 74, 77, 101, 348, 353, 378, 384, 385 Zheng, F., 351, 352, 379 Zheng, J., 144, 156, 182 Zhoa, J., 125, 153 Zhong, W. W., 359, 385 Zhou, H., 21, 37 Zhou, J. M., 266, 277

Zhou, L., 202, 214 Zhou, M., 361, 390 Zhou, Q., 7, 41 Zhou, S., 7, 41 Zhou, X. Z., 244, 260, 271, 272, 278, 280, 282 Zhou, Y., 420, 425, 427, 429 Zhu, Q., 323, 339, 371, 373, 374, 375, 388 Zhu, X., 20, 21, 25, 44 Zhu, Y. Q., 143, 156 Ziegelhoffer, T., 2, 3, 41, 108, 148, 348, 381 Ziegler, F. D., 306, 331, 334 Ziemienowicz, A., 4, 44 Zigler, J. S., Jr., 130, 154 Zimmerman, R., 19, 32, 34, 40, 224, 241 Zimmermann, R., 157, 176, 185, 238, 242, 267, 278 Zolkiewski, M., 416, 429 Zonnevald, D., 322, 339 Zryd, J. P., 208, 216 Zufall, N., 230, 231, 235, 237, 242 Zuhl, ¨ F., 188, 197, 214, 222, 349, 379 Zuiderweg, E. R. P., 22, 23, 26, 27, 41 Zuker, C. S., 260, 279, 281 Zuniga, M., 356, 361, 366, 380 Zurini, M. G. M., 258, 277 zur Nieden, U., 127, 153 Zvaritch, E., 329, 342 Zvi, A. P., 350, 382 Zwickl, P., 187, 188, 189, 190, 191, 192, 193, 197, 200, 202, 204, 206, 207, 218, 221, 222, 418, 428 Zybarth, G., 270, 280 Zydowsky, L. D., 258, 277 Zylicz, M., 4, 12, 14, 15, 17, 18, 25, 28, 30, 31, 37, 38, 40, 43, 44, 413, 414, 416, 427, 429

SUBJECT INDEX

A AAA ATPases, 20S proteasome activation, 202–203 α-Actin, CCT binary complex structure, 84–87 Aggregation, Hsp70 protein folding disaggregation, 5–6 overview, 4 Allostery GroEL polypeptide binding, 64 GroEL ring actions, 52 Amide bond, isomerization, 245–246 Antitumor drugs, Hsp90 ATPase inhibitors, 171–172 Apical domain, GroEL, peptide binding, 54–56 Archaeal AAA ATPases, 20S proteasome activation, 202–203 Assays, prolyl isomerases, 256–257 ATP GroEL binding, 53–54 Hsp90 binding, 175–176 Hsp90 binding in vivo, 164–166 hydrolysis, Hsp70 ATPase cycle, 12–16 protein translocation, 226–227 ATPases AAA ATPases, 20S proteasome activation, 202–203 Clp ATPase chaperone activity, 415–416 characteristics, 414–415 proteolysis role, 415–416 speciﬁcity factors, 424–425 structure, 416–419 substrate recognition, 419–422 substrate release, 423–424 substrate unfolding, 422–423 Hsp70 ATP hydrolysis, 12–16

nucleotide association, 10–11 nucleotide dissociation exchange factor stimulation, 17–20 nucleotide dissociation mechanism, 16–17 substrate-binding domain, 26–28 Hsp90 conformational changes, 166–171 inhibitors, 171–172

B Bacterial AAA ATPases, 20S proteasome activation, 202–203 Bag proteins Hsp70 ATPase cycle, 19–20 Hsp70 co-chaperones, 31–34 Barnase, protein folding in vitro, 262 Barstar, protein folding in vitro, 262 Base, 19S regulatory complex, 206–209 Biosynthesis, cell wall, UPR target, 359–360 BiP, preprotein translocation exits, 230–231 Bite–chew mechanism, proteasome, 200–201 Brownian ratchet model, lumenal Hsp70, 231–233

C Calcium binding, prolyl isomerases, 273 Calnexin differential binding in vivo, 324 endoplasmic reticulum, 320–321 glycoprotein interaction, 321–330 overview, 323–324 Calreticulin differential binding in vivo, 324 endoplasmic reticulum, 320–321 glycoprotein interaction, 321–330 overview, 323–324 481

482

SUBJECT INDEX

Carbonic anydrase, prolyl isomerization, 249 Catalysis proline-limited folding, trigger factor, 265 prolyl isomerization, de novo protein folding, 267–268 RNase T1 folding, 263–264 CCT, see Cytosolic chaperonin-containing TCP-1 Cell cycle, UPR target distribution data, 376–377 Cell wall, biosynthesis and maintenance, UPR target, 359–360 Chaperones Clp ATPase activity proteolysis participation, 415–416 substrate recognition, 419–422 substrate release, 423–424 substrate unfolding, 422–423 endoplasmic reticulum, 320–321 Hsp33, 350–351 Hsp60, 347 Hsp70, see Heat shock protein 70 Hsp70-binding protein 1, 36 Hsp90, see Heat shock protein 90 Hsp104 proteolytic and folding machinery, 349–350 [PSI +], 401–402 [PSI +] curing, 397–400 small Hsps bound substrate refolding, 135–138 model, 127–129 substrate binding site, 132–134 substrate complexes, 130–132, 138 substrate conformation, 134–135 trigger factor properties, 265–266 tubulin speciﬁc folding pathway, 94–96 overview, 93 postfolding functions, 97–98 yeast cofactors, 96–97 Chaperonins, type II CCT–α-actin binary complex structure, 84–87 CCT genetics, 83–84 CCT structure, 84–87 CCT subunits and assembly, 76–78 CCT subunits and assembly speciﬁcity, 79–80 CCT target protein cycling, 81–83

functional studies, 74–75 subunits and assembly overview, 75–76 thermosome, 78–79 thermosome structure, 88 Chemical agents, [PSI +] curing, 396–397 CHIP, Hsp70 co-chaperones, 35–36 Client proteins, Hsp90 activation pathway, 175–176 altered, interactions, 176–180 Clp ATPase chaperone activity, 415–416 characteristics, 414–415 proteolysis role, 415–416 speciﬁcity factors, 424–425 structure, 416–419 substrate recognition, 419–422 substrate release, 423–424 substrate unfolding, 422–423 Clp proteins, proteolytic and folding machinery, 349–350 Clustering, UPR hierarchical, unsupervised search, 371–373 knockout strain data, 377–379 target distribution data, 376–377 CNX, see Calnexin Co-chaperones Bag proteins, 31–34 CHIP, 35–36 DnaJ proteins, 28–31 Hip, 34–35 Hop, 34–35 Hsp70-binding protein 1, 36 Hsp90 interaction, 172–175 Collagen, protein folding in vitro, 262–263 Complementary DNA, UDP–Glc:glycoprotein glucosyltransferase, 310–312 Computational techniques, UPR target identiﬁcation, 366–370 Conformation Hsp70 substrate-binding cavity, 24–26 Hsp90 ATPase cycle, 166–171 sHsps chaperone substrate, 134–135 Conformational replication, [PSI +] in vitro alternate conformations, 405–406 overview, 402–403 self-assembly, 403–405

483

SUBJECT INDEX

Conserved sequence motifs, sHsps, X-ray structure, 120–121 CRT, see Calreticulin α-Crystallin, dominant mutations, 140–141 Crystal structure, see X-ray structure Curing, [PSI +] chemical agents, 396–397 combined chaperone effects, 399–400 Hsp104, 397–399 reversible, de novo induction, 395–396 Cyclophilins characteristics, 257–258 Escherichia coli periplasm, 268 HIV-1 capsid protein interaction, 270–271 prolyl isomerase as domains, 260–261 substrate speciﬁcities, 259–260 Cytochrome c, maturation, Dsb protein role, 297 Cytoskeleton, sHsp interaction, 139–140 Cytosolic chaperonin-containing TCP-1 genetics, 83–84 structure, 84–87 subunits and assembly, 76–78 target protein cycling, 81–83 target range and speciﬁcity, 79–80 Cytosolic chaperonin-containing TCP-1–α-actin binary complex prefoldin complex, 90–93 structure, 84–87 tubulin folding, 93–96 Cytosolic heat shock protein 70, protein translocation, 227–228

D Degradation Clp ATPase activity substrate recognition, 419–422 substrate release, 423–424 substrate unfolding, 422–423 ERAD, UPR coregulation, 360–363 glycoproteins, N-glycan processing, 330–332 20S proteasome products, 199–202 UPR target, 360 Disaggregation, protein folding, Hsp70 role, 5–6 Disulﬁde bonds DsbA role, 289

Escherichia coli, DsbA discovery, 284–286 formation vs. isomerization, 298 isomerization by DsbC, 292–296 Disulﬁde catalyst, DsbA, 286–290 Dithiothreitol, UPR target identiﬁcation, 364 DnaJ proteins Hsp70 ATP hydrolysis, 12–16 Hsp70 co-chaperones, 28–31 Drosophila melanogaster, GT cDNA, 311, 314 DsbA crystal structure, 286–287 cysteine 30, 288 DsbB pathway comparison, 290–291 DsbC comparison, 294–295 equilibrium constant, 287–288 Escherichia coli disulﬁde bonds, 284–286 redox properties, 288–289 substrate binding, 289–290 thioredoxin functional comparison, 287 translocated proteins, 289 DsbB, periplasm oxidizing power, 290–292 DsbC, disulﬁde bonds isomerization, 292–296 DsbD, oxidizing environment, 296–297 DsbG, identiﬁcation, 295–296 Dsb proteins, cytochrome c maturation, 297 DTT, see Dithiothreitol

E Endoplasmic reticulum chaperones, 320–321 N-glycan processing, 304–308, 330–332 glycoprotein–CNX–CRT interaction, 328–330 lumenal Hsp70 proteins, 229–230 protein folding, 320–321, 354–355 protein translocation, 224–227 UPR unsupervised search, 370–371 Endoplasmic reticulum-associated degradation, UPR coregulation, 360–363 ER, see Endoplasmic reticulum ERAD, see Endoplasmic reticulum-associated degradation Escherichia coli disulﬁde bonds DsbA discovery, 284–286 isomerization by DsbC, 292–296

484

SUBJECT INDEX

periplasm, prolyl isomerases, 268–269 Eukaryotic chaperonin, identiﬁcation, 346–349

F FK506 binding proteins characteristics, 258–259 Escherichia coli periplasm, 268–269 functions, 270–271 prolyl isomerase as domains, 260–261 substrate speciﬁcities, 259–260 transmembrane signaling, 269–270 FKBPs, see FK506 binding proteins Functional genomic analysis, UPR ERAD coregulation, 360–363 protein folding in secretory pathway, 354–355 targets, 356–360

G GdnHCl, [PSI +], 401–402 Genetics, CCT, 83–84 GIM, discovery, 88–89 Glucosyltransferase, GT competition, 310 N-Glycan processing endoplasmic reticulum, 304–308 glycoprotein degradation, 330–332 Glycoproteins calnexin interaction, 321–323 calreticulin interaction, 321–323 CNX–CRT interaction, 324–330 degradation, N-glycan processing, 330–332 nonnative, recognition, 314–320 reglucosylation nonnative glycoproteins, 314–320 overview, 308–309 transient type, 306–307 UDP–Glc:glycoprotein glucosyltransferase, 309–314 Glycosylation, UPR target, 359 Golgi apparatus, N-glycan processing, 307–308 GroEL model peptide studies, 56–59 polypeptide binding, 59–62 allostery, 64 mechanism, 62–64

polypeptide binding, structural changes during binding, 65–67 folding cavity role, 69–70 folding initiation changes, 67–69 overview, 64–65 structure allosteric effects, 52 apical domain peptide binding, 54–56 ATP and GroES binding, 53–54 folding cycle, 50–52 GroES binding, 48–49 nucleotide binding, 48–49 overview, 46–48 GroES, GroEL binding, 48–49, 53–54 GrpE, Hsp70 ATPase cycle, 18–19 GT, see UDP–Glc:glycoprotein glucosyltransferase GTP tubulin binding, 93 tubulin folding, 94–96 GTPase, tubulin postfolding functions, 97–98

H Heat shock protein 33, 350–351 Heat shock protein 60, 347 Heat shock protein 70 ATPase cycle ATP hydrolysis, 12–16 nucleotide association, 10–11 ATPase cycle, nucleotide dissociation basic mechanism, 16–17 nucleotide exchange factor stimulation, 17–20 Bag proteins, 31–34 CHIP, 35–36 co-chaperones Bag proteins, 31–34 CHIP, 35–36 DnaJ proteins, 28–31 Hip, 34–35 Hop, 34–35 Hsp70-binding protein 1, 36 cofactors, 228–229 cytosolic, protein translocation, 227–228 DnaJ proteins, 28–31

485

SUBJECT INDEX

general aspects, 8–10 Hip and Hop, 34–35 HspBP1, 36 lumenal, see Lumenal heat shock protein 70 mitochondrial, preprotein translocation exits, 230–231 overview, 1–2 protein folding aggregation prevention, 4 disaggregation, 5–6 native state, 4–5 overview, 3–4 regulatory protein control, 6–8 protein translocation, various roles, 237–238 [PSI +] curing, 399–400 substrate afﬁnity role, 23–24 substrate binding ATPase domain coupling, 26–28 coupling mechanism, 24–28 role in chaperone activity, 23–24 substrate-binding domain structure, 20–23 substrate speciﬁcity, 2–3 Heat shock protein 70-binding protein 1, Hsp70 co-chaperones, 36 Heat shock protein 90 altered client protein interactions, 176–180 ATPase cycle, conformational changes, 166–171 ATPase inhibitors, 171–172 ATP binding, 175–176 ATP binding in vivo, 164–166 chaperone interaction, 172–175 co-chaperone interaction, 172–175 domain structure and function N-terminal nucleotide-binding domains, 162–164 overview, 158–161 hydrolysis, 175–176 hydrolysis in vivo, 164–166 Heat shock protein 104 proteolytic and folding machinery, 349–350 [PSI +], 401–402 [PSI +] curing, 397–400 Heat shock proteins, see Small heat shock proteins

Heat shock response, protein translation, 351–353 Hierarchical clustering, UPR unsupervised search, 371–373 Hip, Hsp70 co-chaperones, 34–35 HIV-1 capsid protein, Cyp18 interaction, 270–271 Hop, Hsp70 co-chaperones, 34–35 Hsp33, see Heat shock protein 33 Hsp60, see Heat shock protein 60 Hsp70, see Heat shock protein 70 Hsp90, see Heat shock protein 90 Hsp104, see Heat shock protein 104 HspBP1, see Heat shock protein 70-binding protein 1 Hydrolysis ATP, Hsp70 ATPase cycle, 12–16 Hsp90, 175–176 Hsp90 in vivo, 164–166

I Immune system, proteasome, 201 Immunophilins, oxidative stress protection, 272 Inheritance, [PSI +], 392 Inner membrane protein, DsbB, 291–292 Isomerization amide bond, 245–246 cis–trans type, nonprolyl peptide bonds, 253–255 disulﬁde bonds, formation vs. isomerization, 298 DsbC by disulﬁde bonds, 292–296 peptide bonds, prolyl isomerases, 273 prolyl bonds catalysis, de novo protein folding, 267–268 folded proteins, 255–256 phosphorylation type, 271–272 Pin 1 type, 271–272 prolyl bonds, protein folding fast and slow species, 246–247 folding impact, 247–248 multiple isomerizations, 248–249 RNase A, 250–251 RNase T1, 251–253

486

SUBJECT INDEX

K Knockout strains, UPR target distribution data, 377–379 KRE5 gene, GT homolog, 312–314

L Lectins, speciﬁc for monoglucosylated oligosaccharides CNX–CRT binding in vivo, 324 glycoprotein interaction, 324–330 overview, 323–324 Lid, 19S regulatory complex, 209–211 Lumenal heat shock protein 70 Brownian ratchet model, 231–233 model experimental evidence, 233–236 protein translocation BiP concentration, 230–231 endoplasmic reticulum, 229–230 mitochondrial Hsp70, 229–230 mtHsp70 concentration, 230–231 Pulling–holding model, 236–237 Pulling model, 233

M Mannosidases, N-glycan processing, 305–306 Membranes, sHsp interactions, 144–145 Messenger RNA, UPR target identiﬁcation, 365–366 Mitochondria, protein translocation, 224–227 Mitochondrial heat shock protein 70, 230–231 Mitochondrial import stimulation factor, 228–229 Models GroEL polypeptide folding, 56–59 lumenal Hsp70 action Brownian ratchet model, 231–233 experimental evidence, 233–236 Pulling–holding model, 236–237 Pulling model, 233 [PSI +], 394–395 sHsp chaperone, 127–129 Molecular model, [PSI +], 394–395

Monoglucosylated oligosaccharides, speciﬁc lectins CNX–CRT binding in vivo, 324 glycoprotein interaction, 324–330 overview, 323–324 Monomers, sHsp, secondary and tertiary structure, 116–119 MSF, see Mitochondrial import stimulation factor mtHsp70, see Mitochondrial heat shock protein 70 Mutations α-crystallin, 140–141 loss-of-function, sHsps, 142–143

N Neoglycoproteins, GT glucosylation, 317–318 NM fragment alternate prion conformation, 405–406 Sup35, 403–404 Nonprolyl peptide bonds, cis–trans isomerization, 253–255 Nucleotide-binding domains, heat shock protein 90, 162–164 Nucleotide exchange factors, Hsp70 ATPase cycle Bag proteins, 19–20 GrpE, 18–19 overview, 17–18 Nucleotides dissociation, Hsp70 ATPase cycle, basic mechanism, 16–17 GroEL binding, 48–49 Hsp70 ATPase cycle, 10–11 Hsp70 substrate-binding cavity ATPase domain coupling, 26–28 conformational changes, 24–26

O Oligomers, sHsps insoluble aggregates in vivo, 127 interactions, 120 organization, 109–114 phosphorylation-induced structural changes, 125–127

SUBJECT INDEX

subunit exchange, 124–125 temperature-induced structural changes, 125 Open reading frame, [PSI +] effects, 406–407 Oxidation DsbD role, 296–297 periplasm, DsbB, 290–292 Oxidative stress, immunophilin protection, 272

P P39A–ribonuclease T1, nonprolyl peptide bonds, 254 PA28, 20S proteasome activator, 212–213 Parvulins characteristics, 259 prolyl isomerase as domains, 260–261 Peptides bond isomerization, prolyl isomerases, 273 DsbC complex, 295 GroEL apical domain binding, 54–56 Peptidylglutamylpeptide-hydrolyzing activity, proteasome, 199–202 Periplasm disulﬁde bonds isomerization by DsbC, 292–296 Escherichia coli, prolyl isomerases, 268–269 oxidizing power, DsbB, 290–292 PGPH, see Peptidylglutamylpeptide-hydrolyzing activity Phenotypes, [PSI +], 392 Phosphorylation dependent prolyl isomerization, 271–272 induced structural changes in sHsps oligomers, 125–127 [PIN +], 395–396, 401 Pin 1, dependent prolyl isomerization, 271–272 Polypeptides, GroEL allostery, 64 binding, 59–62 binding mechanism, 62–64 model studies, 56–59 structural changes during binding, 65–67 folding cavity role, 69–70

487

folding initiation changes, 67–69 overview, 64–65 Prefoldin biochemistry, 90 discovery, 88–89 role in vivo, 90–93 Preproteins, translocation channels, 230–231 Prion, see Yeast prion Proline limited protein folding, catalysis, trigger factor, 265 prolyl isomerase protein folding in vitro, 261–263 Prolyl bonds double bond character, 244–245 isomerization catalysis, de novo protein folding, 267–268 folded proteins, 255–256 phosphorylation type, 271–272 Pin 1 type, 271–272 isomerization, protein folding fast and slow species, 246–247 folding impact, 247–248 multiple isomerizations, 248–249 RNase A, 250–251 RNase T1, 251–253 Prolyl isomerases assay, 256–257 Ca2 + binding, 273 cyclophilins, 257–258 Cyp18–HIV-1 capsid protein interaction, 270 discovery, 256–257 Escherichia coli periplasm, 268–269 FKBP, 258–259, 270–271 larger protein domains, 260–261 nomenclature, 257 oxidative stress protection, 272 parvulins, 259 peptide bond isomerization, 273 phosphorylation-dependent prolyl isomerization, 271–272 Pin 1-dependent prolyl isomerization, 271–272 protein folding in vitro autocatalytic folding, 264 proline-limited steps, 261–263 RNase T1 folding, 263–264

488

SUBJECT INDEX

substrate speciﬁcities, 259–260 transmembrane signaling mediators, 269–270 β Propeptides, 20S proteasome assembly, 197–198 Proteases Clp ATPase translocation, 423–424 HsIV, composition, 189 20S Proteasomes AAA ATPases, 202–203 assembly, 197–199 catalytic mechanism, 193–196 degradation products, 199–202 homomeric rings, 190–192 PA28, 212–213 processing, 196–197 proteolytic activity, 199–202 19S regulatory complex base subunit, 206–209 lid subunits, 209–211 overview, 203–204 structural features, 211 subcomplexes, 204–206 substrate entry path, 192 subunit composition, 188–190 subunit fold, 192–193 26S Proteasomes, structural features, 211 Protein folding de novo, prolyl isomerization catalysis, 267–268 endoplasmic reticulum, 320–321 GroEL, 50–52 Hsp70 role aggregation prevention, 4 native state, 4–5 overview, 3–4 protein folding processes, 5–6 regulatory protein control, 6–8 Hsp104–Clp family, 349–350 proline-limited folding, catalysis, trigger factor, 265 prolyl isomerases in vitro autocatalytic folding, 264 proline-limited steps, 261–263 RNase T1 folding, 263–264 prolyl isomerization fast and slow species, 246–247 folding impact, 247–248 multiple isomerizations, 248–249

RNase A, 250–251 RNase T1, 251–253 secretary pathway, 354–355 sHsp-bound substrate refolding, 135–138 tubulin pathway, 94–96 tubulin-speciﬁc chaperones, postfolding function, 97–98 unfolding, Clp ATPase activity, 422–423 Protein glycosylation, UPR target, 359 Proteins cycling by CCT, 81–83 folded, prolyl bonds, 245, 255–256 interacting, sHsps, 143–144 larger, prolyl isomerase as domains, 260–261 translation, heat shock response, 351–353 Protein translocation Clp ATPase to protease, 423–424 cytosolic Hsp70s, 227–228 DsbA role, 289 endoplasmic reticulum BiP concentration, 230–231 lumenal Hsp70 proteins, 229–230 mtHsp70 concentration, 230–231 overview, 224–227 Hsp70, various roles, 237–238 mitochondria, 224–227 UPR target, 359 Proteolysis Clp ATPase role, 415–416 Hsp104–Clp family, 349–350 20S proteasome, 199–202 Pulling–holding model, lumenal Hsp70, 236–237 Pulling model, lumenal Hsp70, 233

R Reglucosylation glycoproteins overview, 308–309 UDP–Glc:glycoprotein glucosyltransferase, 309–314 UDP–Glc:glycoprotein glucosyltransferase neoglycoproteins, 317–318 RNase B, 318–319 staphylococcal nuclease, 317 vesicular stomatitis virus strain ts045-G, 319–320

SUBJECT INDEX

19S Regulatory complex, 20S proteasome activation base subunit, 206–209 cellular functions, 203–204 lid subunit, 209–211 structural features, 211 subcomplexes, 204–206 Regulatory proteins, Hsp70 role, 6–8 Replication, conformational, [PSI +] in vitro alternate conformations, 405–406 overview, 402–403 self-assembly, 403–405 Reversible curing, [PSI +] de novo induction, 395–396 Ribonuclease, protein folding in vitro, 262 Ribonuclease A, prolyl isomerization role, 250–251 Ribonuclease B, GT recognition, 318–319 Ribonuclease T1 folding, prolyl isomerase role, 263–264 nonprolyl peptide bonds, 254 prolyl isomerization role, 251–253 RNA, mRNA, UPR target identiﬁcation, 365–366 RNase, see Ribonuclease RNase A, see Ribonuclease A RNase B, see Ribonuclease B RNase T1, see Ribonuclease T1

S Saccharomyces cerevisiae, tubulin cofactors, 96–97 Saccharomyces pombe, tubulin cofactors, 96–97 Secondary structure, sHsp monomer, 116–119 Secretory pathway, protein folding, 354–355 Self-organizing maps, UPR unsupervised search, 374–375 sHsps, see Small heat shock proteins Signaling pathways, transmembrane, prolyl isomerase effects, 269–270 Singular value decomposition, UPR unsupervised search, 373–374 Small heat shock proteins bound substrate refolding, 135–138 expression pattern complexity, 114–116 model, 127–129 oligomer

489

insoluble aggregates in vivo, 127 organization, 109–114 phosphorylation-induced structural changes, 125–127 subunit exchange, 124–125 temperature-induced structural changes, 125 potential substrates α-crystallin dominant mutations, 140–141 cytoskeleton interaction, 139–140 interacting proteins, 143–144 loss-of-function mutants, 142–143 membrane interactions, 144–145 overview, 138–139 sequence relationships, 107–109 sHsp–substrate complexes, 130–132, 138 substrate binding site, 132–134 substrate complexes conformation, 134–135 features, 130–132 insoluble type, 138 X-ray structure conserved sequence motifs, 120–121 dimeric building block, 119–120 monomer secondary and tertiary structures, 116–119 oligomeric interactions, 120 other sHsp structural properties, 121–123 SOM, see Self-organizing maps Speciﬁcity factors, Clp ATPase, 424–425 Staphylococcal nuclease, GT sensing, 317 Stress oxidative, immunophilin protection, 272 UPR target distribution data, 376–377 Substrate afﬁnity, Hsp70, role in chaperone activity, 23–24 Substrate-binding cavity GroEL, 69–70 Hsp70, coupling mechanism ATPase domain coupling, 26–28 conformational changes, 24–26 Substrate-binding domain, Hsp70, 20–23 Substrate-binding site DsbA, 289–290 sHsp chaperone, 132–134

490

SUBJECT INDEX

Sup35 conformational replication in vitro overview, 402–403 self-assembly, 403–405 [PSI +], 392–394 Sup45, [PSI +], 400–401 Supervised search, UPR, unsupervised search comparison cell cycle and stress, 376–377 knockout strain data, 377–379 overview, 375–376 SVD, see Singular value decomposition

chaperone properties, 265–266 functions in vivo, 266–267 overview, 264–265 proline-limited folding catalysis, 265 Tubulin, speciﬁc chaperones cofactors in Saccharomyces cerevisiae and S. pombe, 96–97 folding pathway, 94–96 overview, 93 postfolding functions, 97–98

U T t-complex polypeptide 1, identiﬁcation, 348–349 TCP1, see t-complex polypeptide 1 TEM-1 β-lactamase, nonprolyl peptide bonds, 254 Temperature, induced structural changes in sHsps oligomers, 125 Tertiary structure, sHsp monomer, 116–119 Thermosome structure, 88 subunits and assembly, 78–79 Thioredoxin, DsbA functional comparison, 287 Threonine, N-terminal, 20S proteasome, 193, 196 TIM, see Translocase components of inner membrane TOM, see Translocase components of outer membrane Transforming growth factor β, transmembrane signaling, 269–270 Translation, heat shock response, 351–353 Translocase components of inner membrane, 224–225 Translocase components of outer membrane, 224 Translocation, see Protein translocation Transmembranes, signaling, prolyl isomerase effects, 269–270 Transport, vesicular, UPR target, 359 Trigger factor

UDP–Glc:glycoprotein glucosyltransferase cDNA, 310–312 competing glucosyltransferase, 310 homolog KRE5, 312–314 puriﬁcation, 309–310 reglucosylation neoglycoproteins, 317–318 RNase B, 318–319 staphylococcal nuclease, 317 vesicular stomatitis virus strain ts045-G, 319–320 Unfolded protein response functional genomic analysis ERAD coregulation, 360–363 protein folding in secretory pathway, 354–355 targets, 356–360 target identiﬁcation computational strategy, 366–370 experimental strategy, 363–366 unsupervised search deﬁnition, 370–371 hierarchical clustering, 371–373 self-organizing maps, 374–375 singular value decomposition, 373–374 unsupervised vs. supervised search cell cycle and stress, 376–377 knockout strain data, 377–379 overview, 375–376 Unsupervised search, UPR deﬁnition, 370–371 hierarchical clustering, 371–373 self-organizing maps, 374–375

491

SUBJECT INDEX

singular value decomposition, 373–374 supervised search comparison cell cycle and stress, 376–377 knockout strain data, 377–379 overview, 375–376 UPR, see Unfolded protein response

V Vacuolar protein targeting, UPR target, 360 Vesicular stomatitis virus strain ts045-G, GT reglucosylation, 319–320 Vesicular transport, UPR target, 359 VSVts045-G, see Vesicular stomatitis virus strain ts045-G

X X-ray structure Clp ATPase HsIU, 417–418 DsbA, 286–287 DsbC, 293 HsIV, 418 sHsps

conserved sequence motifs, 120–121 dimeric building block, 119–120 monomers, 116–119 oligomeric interactions, 120

Y Yeast, tubulin cofactors, 96–97 Yeast prion, [PSI +] conformational replication in vitro alternate conformations, 405–406 overview, 402–403 self-assembly, 403–405 curing chemical agents, 396–397 combined chaperone effects, 399–400 Hsp104, 397–399 de novo induction, 395–396 GdnHCl, 401–402 global physiologic changes, 407–409 Hsp104, 401–402 inheritance, 392 molecular model, 394–395 open reading frame effects, 406–407 phenotype, 392 [PIN +], 401 Sup35, 392–394 Sup45, 400–401

This . Page Intentionally Left Blank