Advances in Immunology Volume 80

ADVANCES IN

Immunology VOLUME 80

T . his Page Intentionally Left Blank

ADVANCES IN

Immunology EDITED BY FRANK J. DIXON The Scripps Research Institute La Jolla, California ASSOCIATE EDITORS

Frederick Alt K. Frank Austen Tadamitsu Kishimoto Fritz Melchers Jonathan W. Uhr Emil Raphael Unanue

VOLUME 80

Amsterdam

Boston London New York Oxford Paris San Francisco Singapore Sydney Tokyo

San Diego

This book is printed on acid-free paper.

⬁

Copyright 2002, Elsevier Science (USA). 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 ®rst 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 speci®c 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-2002 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. 0065-2776/02 $35.00 Explicit permission from Academic Press is not required to reproduce a maximum of two ®gures or tables from an Academic Press chapter in another scienti®c 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 imprint of Elsevier Science 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 International Standard Serial Number: 0065±2776 International Standard Book Number: 0±12±022480±1 PRINTED IN THE UNITED STATES OF AMERICA 02 03 04 05 06 07 MV 9 8 7 6 5 4 3

2 1

CONTENTS

Contributors

ix

Protein Degradation and the Generation of MHC Class I-Presented Peptides

Kenneth L. Rock, Ian A. York, Tomo Saric, and Alfred L. Goldberg I. II. III. IV.

Introduction Protein Degradation Antigen Presentation Conclusion References

1 2 21 51 52

Proteolysis and Antigen Presentation by MHC Class II Molecules

Paula Wolf Bryant, Ana-Maria Lennon-DumeÂnil, Edda Fiebiger, CeÂcile LagaudrieÁre-gesbert, and Hidde L. Ploegh I. II. III. IV.

Introduction Protease Activity in Antigen-Presenting Cells Proteolytic Digestion of Ii Proteolytic Control of Vesicle Biogenesis and Class II Traf®cking through the Endocytic Pathway V. The Role of Ii in Regulating the Proteolytic Activities of APCs VI. Antigen Processing VII. Concluding Remarks References

71 73 87 92 94 96 102 103

Cytokine Memore of T Helper Lymphocytes

Max LoÈhning, Anne Richter, and Andreas Radbruch I. Introduction II. Cytokine Signals in the Induction of T Cell Cytokine Memore v

115 117

vi

contents

III. Key Transcription Factors in the Induction and Maintenance of Cytokine Memory IV. Role of the Antigen-Presenting Cell in the Induction of Cytokine Memory V. T Cell Receptor Signals in the Induction of Cytokine Memory VI. Epigenetic Modi®cations of Cytokine Genes VII. Stability and Plasticity of Cytokine Memory VIII. Cytokine Memory as Part of T Cell Differentiation Programs IX. Cytokine Memory of Memory T Cells References

122 130 133 138 145 149 154 158

Ig Gene Hypermutation: A Mechanism Is Due

Jean-Claude Weill, Barbara Bertocci, Ahmad Faili, Said Aoufouchi, SteÂphane Frey, Annie de Smet, SeÂbastien Storck, Auriel Dahan, FreÂdeÂric Delbos, Sandra Weller, Eric Flatter, and Claude-AgneÁs Reynaud I. II. III. IV. V. VI. VII. VIII. IX. X.

Introduction Mismatch Repair The ``Nickase'' The Role of Transcription Hypermutation Is Generated by a DNA Polymerase Error-Prone DNA Polymerases More DNA Polymerases Targeting to Non-Ig Genes AID Conclusion References

183 183 186 187 188 191 194 195 195 196 197

Generalization of Single Immunological Experiences by Idiotypically Mediated Clonal Connections

Hilmar Lemke and Hans Lange I. Introduction II. Idiotypic Transformation of Single Immunological Experiences in the Adult III. Transfer of Maternal Immunological Experience to the Offspring IV. Conclusion References

203 204 223 228 229

The Aging of the Immune System

B. Grubeck-Loebenstein and G. Wick I. The Biological Aging Process II. The Aging of the Immune System

243 248

contents III. The Consequences of Immune Senescence IV. Modes of Intervention V. Conclusion References Index Contents of Recent Volumes

vii 263 268 271 273 285 293

This Page Intentionally Left Blank

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors' contributions begin.

Said Aoufouchi (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Barbara Bertocci (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Paula Wolf Bryant (71), Department of Microbiology, The Ohio State University, Columbus, Ohio 43210 Auriel Dahan (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Annie de Smet (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France FreÂdeÂric Delbos (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Ahmad Faili (183), INSERM Unite 373, Faculte de MeÂdecine Necker-Enfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Edda Fiebiger (71), Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115 Eric Flatter (183), INSERM Unite 373, Faculte de MeÂdecine Necker-Enfants Malades, Universite Paris V, 75730 Paris Cedex 15, France SteÂphane Frey (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Alfred L. Goldberg (1), Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115 B. Grubeck-Loebenstein (243), Institute for Biomedical Aging Research of the Austrian Academy of Sciences, A-6020 Innsbruck, Austria CeÂcile LagaudrieÁre-Gesbert (71), Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115 Hans Lange (203), Biochemical Institute of the Medical Faculty of the Christian-Albrechts-University, D-24118 Kiel, Germany Hilmar Lemke (203), Biochemical Institute of the Medical Faculty of the Christian-Albrechts-University, D-24118 Kiel, Germany ix

x

contributors

Ana-Maria Lennon-DumeÂnil (71), Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115 Max LoÈhning (115), Deutches Rheumaforschungszentrum, 10117 Berlin, Germany Hidde L. Ploegh (71), Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115 Andreas Radbruch (115), Deutches Rheumaforschungszentrum, 10117 Berlin, Germany Claude-AgneÁs Reynaud (183), INSERM Unite 373, Faculte de MeÂdecine Necker-Enfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Anne Richter (115), Deutches Rheumaforschungszentrum, 10117 Berlin, Germany Kenneth L. Rock (1), Department of Pathology, University of Massachusetts Medical School, Worcester, Massachusetts 01655 Tomo Saric (1), Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115 SeÂbastien Storck (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Jean-Claude Weill (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France Sandra Weller (183), INSERM Unite 373, Faculte de MeÂdecine NeckerEnfants Malades, Universite Paris V, 75730 Paris Cedex 15, France G. Wick (243), Institute for Biomedical Aging Research of the Austrian Academy of Sciences, and the Institute for Pathophysiology, University of Innsbruck Medical School, A-6020 Innsbruck, Austria Ian A. York (1), Department of Pathology, University of Massachusetts Medical School, Worcester, Massachusetts 01655

ADVANCES IN IMMUNOLOGY, VOL. 80

Protein Degradation and the Generation of MHC Class I-Presented Peptides KENNETH L. ROCK,* IAN A.YORK,* TOMO SARIC,À AND ALFRED L.GOLDBERGÀ *Department of Pathology,University of Massachusetts Medical School,

Worcester, Massachusetts 01655; and À Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115

Over the past decade there has been considerable progress in understanding how MHC class Ipresented peptides are generated. The emerging theme is that the immune system has not evolved its own specialized proteolytic mechanisms but instead utilizes the phylogenetically ancient catabolic pathways that continually turnover proteins in all cells. Three distinct proteolytic steps have now been de®ned in MHC class I antigen presentation. The ®rst step is the degradation of proteins by the ubiquitin±proteasome pathway into oligopeptides that either are of the correct size for presentation or are extended on their amino-termini. In the second step, aminopeptidases trim Nextended precursors into peptides of the correct length to be presented on class I molecules. The third step involves the destruction of peptides by endo- and exopeptidases, which limits antigen presentation, but is important for preventing the accumulation of peptides and recycling them back to amino acids for protein synthesis or production of energy. The immune system has evolved several components that modify the activity of these ancient pathways in ways that enhance the generation of class I-presented peptides. These include catalytically active subunits of the proteasome, the PA28 proteasome activator, and leucine aminopeptidase, all of which are upregulated by interferon-g. In addition to these pathways that operate in all cells, dendritic cells and macrophages can also generate class I-presented peptides from proteins internalized from the extracellular ¯uids by degrading them in endocytic compartments or transferring them to the cyotosol for degradation by proteasomes. # 2002, Elsevier Science (USA).

I. Introduction All proteins are continually turned over in cells (Goldberg and St. John, 1976). Although this process may appear wasteful, it subserves many important biological functions. Cells eliminate damaged or abnormal proteins whose accumulation would be toxic (Sherman and Goldberg, 2001). Furthermore, many biological processes are precisely regulated by degrading key enzymes (e.g., hydrolysis of cyclins as cells progress though the cell cycle; Murray, 1995) or regulatory proteins (e.g., destruction of the inhibitory protein IkB which releases the transcription factor NFkB; Chen et al., 1995). Moreover, the levels of all proteins are maintained by balancing their production and turnover and 1 Copyright 2002, Elsevier Science (USA) . All rights reserved. 0065-2776/02 $35.00

2

kenneth l. rock et al.

the destruction of cell proteins provides a source of amino acids for protein synthesis or production of energy. Another important but more recently evolved function of protein catabolism is to provide a source of peptides for immune surveillance. A fraction of the peptides produced from the degradation of cellular proteins are transported into the endoplasmic reticulum (ER) and bound by MHC class I molecules. These peptide±MHC complexes are then transported to the cell surface where they are displayed to the immune system. In this manner cells display on their surface a sampling of their expressed genes, which allows the immune system to monitor the health of individual cells in tissues and to identify ones that are infected with intracellular pathogens, such as viruses, or those expressing mutated proteins (e.g., tumors). Cytotoxic T lymphocytes (CTL) recognize these foreign peptides and are stimulated to kill the offending cells. This is the major mechanism through which the immune system eliminates virally infected and malignantly transformed cells. This review is divided into two parts. The ®rst section discusses the pathways that are involved in the continual and regulated degradation of cellular proteins. These pathways predate the evolution of the immune system and have shaped the way in which the antigen presentation pathway operates. In some cases these pathways favor antigen presentation by generating certain peptides, while in other cases they may constrain antigen presentation, in that certain peptides are not generated or are destroyed. The second section reviews how the immune system uses and has modi®ed these pathways to generate antigenic peptides. II. Protein Degradation A. Protein Degradation Pathways Cells degrade proteins through two major pathways, which are present in different subcellular compartments and degrade different cell components. One operates in endocytic compartments, particularly lysosomes and late endosomes, while the other is located in the cytosol and nucleus. The proteases in these two pathways and their regulation are very different from one another. They also subserve different roles in antigen presentation. The proteolytic pathway in late endocytic compartments is responsible for degrading many membrane proteins and extracellular proteins that enter these vesicles via endocytosis (Luzio et al., 2000; Tsao and von Zastrow, 2000; Waterman and Yarden, 2001). These vesicles contain a proton antiporter that makes them acidic. Members of the papain family of cysteine proteases, termed cathepsins (Turk et al., 2000), as well as other cysteine proteases (e.g., asparaginyl endopeptidase (Chen et al., 1997) (Manoury et al., 1998)), reside in this compartment and are maximally active in its acidic environment. Some lysoso-

generation of mhc class i-presented peptides

3

mal proteases, like cathepsins E, D, L, and S, have endoproteolytic activities, while others, such as cathepsins B, H, and X, predominantly exhibit exopeptidase activity (Bohley and Seglen, 1992; Therrien et al., 2001). Some of these proteolytic enzymes are responsible for generating the majority of peptides that are presented on MHC class II molecules (Medd and Chain, 2000; Villadangos et al., 1999) and in certain specialized antigen-presenting cells they may produce some of the peptides presented on class I molecules (discussed further below). However, the generation of MHC class I-presented peptides in all other cells does not involve the endocytic pathway. The proteolytic pathway in the cytosol and nucleus is responsible for degrading the bulk of cellular proteins and generating the majority of class I-presented peptides (Rock et al., 1994). Early studies of this pathway revealed that it was energy dependent (Etlinger and Goldberg, 1977; Gronostajski et al., 1985). Thus, cellular proteins that were normally degraded were stabilized when cells were depleted of ATP. This energy requirement was surprising because thermodynamically the cleavage of peptide bonds should not require energy and all proteases that were known at the time were ATP independent. This ATP dependence was an important clue in ®nding the responsible proteolytic system with novel biochemical properties. These studies led to the discovery of ATPdependent processes (ubiquitination) (Ciechanover et al., 1978; Hershko et al., 1980; Wilkinson et al., 1980) that target proteins for degradation and the ATPdependent proteases (26S proteasomes) that are involved in their hydrolysis (Driscoll and Goldberg, 1990; Hough et al., 1986). Although the consumption of energy to degrade proteins may appear wasteful, it turns out to be energy well spent because it allows the cell to carefully regulate the degradation of its proteins and it protects cell constituents from inappropriate proteolytic attack. B. Ubiquitin-Conjugation and Protein Degradation Many cellular proteins are targeted for destruction by covalent linkage to the cofactor ubiquitin (Ciechanover et al., 2000; Weissman, 2001). Ubiquitin is a small protein whose structure is highly conserved through evolution and is abundant in the cytoplasm and nucleus of all eukaryotic cells (Finley and Chau, 1991; Schlesinger et al., 1975). In this conjugation reaction, the C-terminus of ubiquitin is ®rst activated by an ATP-dependent linkage to a ubiquitinactivating enzyme, E1, via a thioester bond. The activated ubiquitin is then transferred by a trans-thiolation reaction to one of the cell's many ubiquitincarrier proteins, E2s. Finally, a speci®c ubiquitin±protein ligase, E3, facilitates the transfer of the activated ubiquitin to a speci®c protein substrate. Each ubiquitin molecule is conjugated via an isopeptide bond to the epsilon amino group of a lysine side chain on the substrate or to other previously bound ubiquitin moieties to form a long chain. Once the polyubiquitin chain is longer

4

kenneth l. rock et al.

than four residues, it targets the protein for degradation by 26S proteasomes (see below). The selectivity of this process is conferred by the E2 and E3 enzymes. Cells contain a single E1 enzyme, about 10±20 different E2 enzymes, and perhaps over a hundred distinct E3 enzymes with many new ones being discovered in the past few years (Hershko et al., 2000; Weissman, 2001). Each of these E3 enzymes speci®cally binds a speci®c group of proteins destined for degradation and confers the high degree of speci®city to the process of ubiquitination, either alone or in combination with an associated E2 enzyme or other cofactors. In the latter case, the E2 and E3 enzymes function in pairs and each particular combination has speci®city for a distinct set of substrates. E3 enzymes can recognize speci®c sequences (e.g., certain N-terminal residues or internal ``destruction boxes''), post-translational modi®cations (e.g., phosphorylation), and probably certain changes in protein conformation (e.g., denaturation). The need for ubiquitination prior to degradation evolved to provide a remarkable degree of selectivity and regulation to the degradative process and allows cells to selectively degrade some proteins, while sparing others. For example, damaged and mutant proteins tend to be quickly eliminated. Most recently, a novel mode of regulation of protein ubiquitination at the level of E3 ubiquitin± protein ligases is emerging. The post-translational modi®cation of some E3 ligases through covalent attachment of distinct ubiquitin-like molecules is required for targeted ubiquitination of their substrates. For example, in vitro attachment of a ubiquitin-like protein SUMO-1 to Mdm2 (an E3 ligase for the tumor supressor p53) abrogates its self-ubiquitination and increases its ubiquitin±ligase activity toward p53 (Buschmann et al., 2000). In addition, covalent modi®cation of cullin-1 (a component of an E3 enzyme in a complex structure called SCF ubiquitin±protein ligase) by another ubiquitin-like modi®er, NEDD8, plays an essential role in the function of SCF in ®ssion yeast and causes marked augmentation of the SCF-dependent polyubiquitination of target proteins (Hori et al., 1999; Lyapina et al., 2001; Osaka et al., 2000). It also appears that the modi®cation of substrates directly by some of these ubiqutin-like proteins, such as binding of SUMO-1 to IkB, could have a role in counteracting the function of ubiquitin. Thus, it seems that complex systems have evolved to precisely regulate timely ubiquitination of crucial target proteins and their degradation by the proteasome system to ensure proper cellular function. The involvement of ubiquitin in protein degradation was ®rst discovered in studies fractionating the ATP-dependent proteolytic pathway present in soluble cell extracts (Ciechanover et al., 1978). Subsequent studies demonstrated ubiquitin conjugation to substrates (Ciechanover et al., 1980; Hershko et al., 1980) and that several components, eventually de®ned as E1, E2, and E3, were needed to modify proteins with ubiquitin and target them for destruction

generation of mhc class i-presented peptides

5

(Hershko and Ciechanover, 1998; Hershko, 1996). In the absence of any of these components many cellular proteins were not degraded in the extracts. Therefore, ubiquitination was clearly essential for the turnover of many intracellular proteins in vitro. Subsequent studies have explored the role of the ubiquitin conjugation pathway in vivo. Ubiquitin and E1 in yeast and mammalian cells were shown to be essential for cell viability and for turnover of many regulatory proteins. For example, progression through the cell cycle requires the programmed destruction by this pathway of many regulatory molecules, such as cyclins (Glotzer et al., 1991) or cyclin-dependent kinase inhibitors (Sutterluty et al., 1999; Zachariae and Nasmyth, 1999). Since there are no inhibitors that can be used to block ubiquitination in vivo, studies examining its role in vivo have often had to utilize mutations that disrupt the pathway in a conditional manner or that disrupt only certain components of the pathway. For example, the loss of speci®c E2 or E3 enzymes allows viability but leads to the stabilization of speci®c groups of cell proteins (Kho et al., 1997; Gonen et al., 1999; Seol et al., 1999; Townsley et al., 1997). One very informative approach has been the use of mammalian cells that have a mutation in their E1 ubiquitin-activating enzyme that renders it thermolabile (Ciechanover et al., 1984). When these cells are shifted to nonpermissive temperature (typically 41±428C) ubiquitination is reduced. Under these conditions the degradation of short-lived normal and abnormal proteins is impaired, demonstrating that ubiquitination is important for protein degradation in vivo. It has been reported that the degradation of long-lived proteins, which make up the bulk of cell proteins, is not reduced in the TS mutants at the nonpermissive temperature (Ciechanover et al., 1984; Finley et al., 1984). However, it is uncertain whether this is because the degradation of these proteins is ubiquitin independent, because the residual E1 activity is suf®cient for their degradation in a ubiquitin-dependent manner, or because of technical dif®culties in measuring degradation of long-lived proteins in these cells. Therefore the role of ubiquitin in the degradation of long-lived proteins remains an important unresolved question. However, long-lived proteins are still degraded by the same proteolytic particle as the ubiquitin-conjugated proteins (see below). Besides its role in marking proteins for degradation by 26S proteasomes, ubiquitin also mediates other important biological reactions. For example, the conjugation of a single ubiquitin molecule to histones H2A and H2B in higher eukaryotes has been related to changes in chromatin organization and this modi®cation does not target these histones for degradation (Mimnaugh et al., 1997; Wu et al., 1981). In addition, mono- or polyubiquitination is required as a signal for endocytosis of certain integral plasma membrane proteins, such as uracil permease in yeast and growth hormone receptor in mammalian cells, and their sorting and subsequent degradation in the lysosome (Hicke, 1997). Some

kenneth l. rock et al.

6

of these cell surface proteins appear to be degraded also by the proteasome (Hicke, 1999). C. Proteasomes Further biochemical analysis of the ATP-dependent degradative system in cell extracts led to the discovery of an ATP-dependent proteolytic complex, the 26S proteasome. Like ubiquitin, proteasomes are phylogenetically old, and their structure has been highly conserved throughout evolution. In fact, these particles are almost indistinguishable in yeast and man (Voges et al., 1999). Unlike ubiquitin, which is only found in eukaryotes, proteasomes are also present in simpler forms in some prokaryotic organisms (e.g., archaea and actinomyces), and thus proteasomes are more ubiquitous and older than ubiquitin (Dahlmann et al., 1989). Proteasomes are a major cell constituent, composing up to 2% of cellular protein, and are essential for viability. They are found in the cytoplasm and nucleus of all eukaryotic cells (Brooks et al., 2000; Enenkel et al., 1999; Fabunmi et al., 2001; Palmer et al., 1996; Reits et al., 1997). The 26S proteasomes degrade in an ATP-dependent manner polyubiquitinated substrates and in some cases unmodi®ed proteins and are the forms active in vivo. The 26S proteasome is a complex composed of two very different components, the core 20S particle and one or two 19S regulatory particles (also called PA 700). Proteins are degraded within the 20S particles. D. The 20S Proteasome These structures were independently discovered in many contexts (e.g., as novel proteolytic complexes or cytosolic particles; Hough et al., 1987; Wilk and Orlowski, 1983) before their common identity and critical function in the ATPand ubiquitin-dependent proteolytic pathway was demonstrated (Driscoll and Goldberg, 1990). The 20S proteasome is a cylindrical, 720-kDa particle composed of 28 subunits arranged in four stacked rings (Groll et al., 1997; Voges et al., 1999). Each ring contains seven distinct but homologous subunits surrounding a central channel. The two identical inner rings (containing the bsubunits) contain the proteolytic active sites and together form a central chamber in which proteins are degraded. The two outer rings (containing the a subunits) together with b-rings form two antechambers found on the each side of the 20S proteasome through which substrates must pass in order to be degraded in the central chamber. The only opening into the 20S particle through which substrates can enter is at the center of the a-ring. This center is normally maintained in a closed position (Groll et al., 1997). Proteins or peptides cannot therefore freely enter the inner cavity of the 20S proteasome from the cytosol. Only three of the seven b-type subunits in each b-ring of the 20S proteasome possess catalytic activity (Baumeister et al., 1998; Groll et al., 1997). Their proteolytic active sites are oriented toward the central cavity, so

generation of mhc class i-presented peptides

7

that cleavage of proteins can only occur within the cavity. This architecture of the particle, in which proteolysis is isolated to a separate chamber and access is tightly regulated (see below), serves to prevent the unregulated and potentially harmful destruction of essential cell constituents. The proteasome functions through a unique catalytic mechanism, which differs from that of the four classical families of proteases (serine, cysteine, aspartyl, and metalloproteases). Each catalytic b-subunit of the proteasome contains an amino-terminal threonine residue, whose hydroxyl group serves as the nucleophile that attacks the peptide bond (Brannigan et al., 1995; Fenteany et al., 1995; Lowe et al., 1995; Seemuller et al., 1995a). Although proteasomes and their bacterial homolog, HslUV (Rohrwild et al., 1996), are the only proteases utilizing such a mechanism, other hydrolytic enzymes (e.g., penicillin acylase), belonging to the family of N-terminal hydrolases, utilize hydrolysis by the N-terminal residue and have similar tertiary conformations. This specialized conformation and activity of proteasomal subunits depend on extensive, tight interactions with speci®c neighboring subunits. Consequently, when expressed individually, the catalytic 20S b-subunits are enzymatically inactive. Since ribosomal translation of all proteins begins from an initial methionine residue, a special mechanism is also necessary to generate the catalytic Nterminal threonine (Kruger et al., 2001). This is achieved during the initial assembly of the 20S particle from two smaller precursor complexes (containing one a-ring and one b-ring). The immature catalytic b-subunits in a b-ring of this precursor complex undergo an autocatalytic cleavage of an N-terminal propeptide that generates the free N-terminal threonine and activates the proteasome (Jager et al., 1999; Mayr et al., 1998; Schmidtke et al., 1996; Seemuller et al., 1996). Because of this unique mechanism, it has proven possible to synthesize inhibitors that speci®cally affect proteasomes and no other cellular protease (see below). These inhibitors have proved to be valuable tools in clarifying the mechanism of the proteasome's action and in determining its role in cells. The 20S proteasome has been termed a multicatalytic protease. This name is justi®ed because each of the three proteolytically active b-subunits in a single b-ring contains a different type of active site (Wilk and Orlowski, 1983). These sites, which function together in degrading polypeptides, differ in speci®city for different amino acid sequences. This difference in speci®city has been most clearly shown using small ¯uorogenic peptide substrates (Cardozo, 1993; Orlowski et al., 1993). One site, commonly referred to as the ``chymotrypsin-like site'' cleaves preferentially on the carboxylic side of hydrophobic amino acids. The second or ``trypsin-like site'' cleaves preferentially after basic residues, and the third site cleaves primarily after acidic residues. This latter site was traditionally termed ``peptidylglutamyl peptide hydrolase'' (PGPH) (Orlowski et al., 1991) However, it has been found that it cleaves faster after aspartic residues than after glutamates, and therefore it has been suggested that these sites be

8

kenneth l. rock et al.

called ``post-acidic'' or ``caspase-like'' (Kisselev et al., 1999a). However, these preferences are far from absolute. For example, cleavages after branched chain amino acids (which had earlier been ascribed to some unidenti®ed site; Orlowski et al., 1993) actually are performed by the post-acidic site (Cardozo et al., 1996, 1999; Dick et al., 1998) as well as by the chymotrypsin-like sites (McCormack et al., 1998). The speci®cities of these sites also depend on more distant residues (Eisenlohr et al., 1992b; Nussbaum et al., 1998; Shimbara et al., 1998). However, these preferences have not yet been systematically de®ned, even though they can have major implications for immunodominance (see below). With the resolution of the three-dimensional structure of the yeast proteasome by Huber and coworkers (Groll et al., 1997), it has proven possible to localize these different active sites through binding of peptide inhibitors (Fenteany et al., 1995; Groll et al., 2000). A variety of different nomenclatures have been proposed to refer to these different subunits, usually based on arbitrary genetic names. However, a rational terminology has been proposed for the different aand b-subunits based upon the X-ray analysis, in which the chymotrypsin-like sites are located on the b5-subunits, the trypsin-like on b2-subunits, and the post-acidic ones on the b1-subunits (Baumeister et al., 1998; Finley et al., 1998). These different speci®cities allow the proteasome to cleave virtually all proteins into oligopeptides. As proteins thread through the proteasome's channel their degradation is highly processive, leading to the complete destruction of most polypeptides to small fragments of 3±22 residues in length (Kisselev et al., 1999b). There is recent evidence that peptide substrate binding to sites within the particle can activate other active sites through an allosteric mechanism. These ®ndings have led to the idea that these different sites function in protein breakdown in an ordered mechanism (e.g., bite±chew model) where a cleavage at one site stimulates additional cleavages until the products are small enough to diffuse out of the proteasome (Kisselev et al., 1999a). However, the location of these regulatory sites and their precise signi®cance remain unclear (Schmidtke et al., in press). The proteasome's two outer rings are not proteolytically active. They are essential for the assembly of the particle, serve as docking sites for additional regulatory subunits (see below), and control the entry of substrates into the particle as well as product release (Voges et al., 1999). The opening through the a-rings into the central channel of the proteasome appears to be gated, and entry is precisely regulated (Groll et al., 2001). When this gate is closed the proteasome is inactive because substrates cannot enter the particle. Proteasomes isolated from cells under gentle conditions are in this latent (inactive) state. This closed form of 20S particle was initially revealed by X-ray crystallography of yeast proteasomes (Groll et al., 1997). However, when these particles are incubated in vitro at 378C, with time, there is a spontaneous activation of the 20S proteasomal peptidase activities, most probably due to the opening of

generation of mhc class i-presented peptides

9

the gate into the particle. This spontaneous activation in vitro can be inhibited by the presence of physiological concentrations (10 mM) of K‡ or Na‡ ions (KoÈhler et al., 2001; K. M. Woo and A. L. Goldberg, in preparation). Therefore, it is likely that the high concentrations (100 mM) of K‡ within all cells help maintain free 20S particles in an inactive form. The gate in a-rings is composed of the N-termini of several subunits, but the tail of a3 appears to serve as a scaffold for interaction with conserved residues on other a-subunits (Kohler et al., 2001). Deletion or mutation of this Nterminal region leaves the gate in an open state allowing both substrates (peptide and denatured proteins) to enter, and the products to exit, more easily (KoÈhler et al., 2001). However, even in the opened state, this gate has a small diameter (13 AÊ) that can only accommodate unfolded polypeptides (Groll et al., 1997). Therefore protein substrates must ®rst be unfolded to enter into the 20S proteasome and be degraded. This gate in vivo is regulated by the complexes (such as the 19S particle) that associate with a-rings and activate the proteasome. Thus, ef®cient and selective protein breakdown also requires proteasome-associated regulatory components. So, for example, the 19S particle has a function to bind selectively ubiquitinated or other appropriate substrates and unfold them completely, open the gate in the a-ring of the 20S proteasome, and promote the translocation of the substrate into this particle for degradation. In vitro, puri®ed 20S particles can degrade only certain denatured proteins (e.g., carboxymethylated lactalbumin) and proteins with little structure, such as casein. However, the rates of hydrolysis are quite low unless the particle is activated, e.g., by treatment with low concentrations of sodium dodecyl sulfate (SDS). The 20S particles are not able to degrade ubiquitin-conjugated proteins. Their degradation of unconjugated proteins is ATP independent. Since protein degradation in vivo is ATP dependent, the 20S proteasome as an isolated particle does not appear to play a major role in degrading cellular proteins. In fact, conditions that favor activation of the 20S particle are likely to lead to nonspeci®c hydrolysis of important cell components, which would be highly toxic to the cell. Though probably inactive in vivo by itself, this particle does play an essential role in protein hydrolysis as the proteolytic core of the ATPdependent 26S proteasome. E. The 26S Proteasome The 26S proteasomes are 2.5-MDa complexes and are the only structures in the cells capable of degrading polyubiquitinated substrates (Voges et al., 1999). In vitro, they can also degrade certain proteins that are not modi®ed with ubiquitin (Murakami et al., 1992, 1999; Tokunaga et al., 1994), and it remains unclear to what extent they may function in a ubiquitin-independent mode in vivo (see below). The hydrolysis of both ubiquitinated and nonubiquitinated

10

kenneth l. rock et al.

proteins by these particles is linked to ATP hydrolysis. In fact, 26S proteasomes are the only ATP-dependent proteases present in the nucleus and cytosol of eukaryotic cells. By contrast, mitochondria lack ubiquitin or proteasomes, but contain several ATP-dependent proteases resembling those ®rst described in eubacteria (e.g., La (Kutejova et al., 1993) and FtsH (Guelin et al., 1994; Lupas et al., 1997) ). Since the bulk of protein hydrolysis is ATP dependent in vivo, the 26S proteasome must be the key protease responsible for the degradation of the great majority of intracellular proteins (Rock et al., 1994). The 26S proteasome is formed when one or two 19S complexes associate in an ATP-dependent process with the a-rings of the 20S particle (Voges et al., 1999). The 19S particle (890 kDa) is made up of at least 18 different subunits and consists of two different functional entities (Ferrell et al., 2000; Glickman et al., 1998). The base, which touches the a-rings of the 20S core particle, contains six different ATPases (a type of enzyme that hydrolyzes ATP and uses the energy released to drive biological processes) and two additional nonATPase subunits. These ATPases form a ring that uses ATP to unfold protein substrates (Braun et al., 1999). They also have the capacity to open the gate in the a-ring of the 20S proteasome. In doing so, they promote translocation of substrates into the inner chambers of the 20S particle, where they are degraded. The outer entity of the 19S particle, termed the lid, is attached on top of the base and provides speci®city to proteolysis. It contains the binding site for ubiquitin chains and also enzymes that break down the polyubiquitin chains so that ubiquitin can be recycled for use in subsequent rounds of proteolysis (Baumeister et al., 1998; Rubin et al., 1997). The lid and the base are highly conserved in sequence from yeast to mammals, suggesting an important biological function for these structures (Voges et al., 1999). A deletion of one of the base subunits in yeast led to the loss of the lid of the 19S complex and caused accumulation of ubiquitin-conjugated proteins (Glickman et al., 1998). The base, with its six ATPase subunits, stimulates the hydrolysis of peptides and nonubiquitinated proteins (e.g., casein) by the 20S core particle (Glickman et al., 1999). The ATPases play essential functions because targeted mutations of individual subunits were either lethal or interfered with cell growth (Rubin et al., 1998). Interestingly, point mutations in these different ATPases cause defects in the degradation of different types of proteins, suggesting a role of these ATPases in the recognition of different substrates. Moreover, certain mutations in the ATP-binding domain of one of these ATPases (Rpt2) also prevented peptide hydrolysis (Rubin et al., 1998), and genetic analysis indicated that this ATPase controls gating of the channel in the a-ring of the 20S particle (Kohler et al., 2001). However, the base without the lid complex is not suf®cient to allow the degradation of polyubiquitinated proteins by the 20S proteasome (Glickman et al., 1998).

generation of mhc class i-presented peptides

11

These six homologous ATPases belong to the large AAA family of ATPases, all of which are hexameric rings that serve diverse functions in cells and include components of the various ATP-dependent proteases in bacteria and mitochondria (Schmidt et al., 1999). Much has been learned about the function of these ATPases by studies of the close homolog, the proteasome activating nucleotidase (PAN), which promotes protein degradation by archaebacterial proteasomes (Ng et al., submitted for publication; Zwickl et al., 1999). For mechanistic studies, this simpler complex and the analogous ATPase-regulated proteases in bacteria (ClpAP or HslUV) offer many advantages (e.g., they do not require ubiquitin). The PAN complex is a substrate-activated ATPase that binds protein substrates and has been shown to catalyze both the ATP-dependent unfolding of globular substrate (e.g., the model substrate GFP) and its translocation into the 20S particle (Zwickl et al., 1999; Ng et al., submitted for publication; Benaroudj and Goldberd, 2000). Unfolding appears to occur on the surface of the ATPase ring and to precede the translocation step, in which the polypeptide passes through the center of this ring in a speci®c C- to N-terminal direction (A. Navon, in preparation). Thus, it behaves like a ``reverse molecular chaperone.'' The PAN complex as well as the base of the 19S complex has many chaperone-like properties, such as the ability to prevent protein aggregation and to refold a denatured enzyme. Thus, PAN appears to be the evolutionary precursor of the base of the 19S complex. The capacity of the base to use ATP to present substrates in an extended form to the 20S particle, to open the gate into this particle, and to inject the substrate into the proteasome for rapid hydrolysis clearly evolved in early prokaryotes before the linkage of ubiquitination to proteasome function. It is also clear that the lid portion evolved independently, since most of its subunits are homologous to a ubiquitous but poorly understood complex composed of eight subunits, termed the COP9 signalosome (Glickman et al., 1998; Henke et al., 1999; Wei and Deng, 1999). This multimeric structure has been implicated in signal transduction and various regulatory functions (especially in plants) (Glickman et al., 1998; Wei et al., 1998; Bech-Otschir et al., 2001). Recently, the COP9 signalosome has been also found to play an important role in mediating E3 ubiquitin ligase-mediated responses in plants (Schwechheimer et al., 2001) and yeast (Lyapina et al., 2001). The fusion of these two ancestral structures with the 20S proteasome has clearly provided eukaryotes with a highly selective and highly regulated proteolytic machine. Moreover, recent studies indicate that a number of other cellular proteins, including, among others, ubiquitin-conjugating enzymes, deubiquitinating enzymes, proteins possessing ubiquitin-like domains, and heat shock proteins, may associate temporarily with this degradative complex (Luders et al., 2000; Tongaonkar et al., 2000; Verma et al., 2000); so its true complexity and central role are still not fully understood.

12

kenneth l. rock et al.

F. PA28 and Hybrid Proteasomes In higher eukaryotes, there is an additional proteasome-regulatory complex, termed 11S REG or PA28, that appears to play an important role in MHC class I antigen presentation (Dubiel et al., 1992; Ma et al., 1992). PA28 is a 200-kDa heptameric ring structure composed of 28-kDa subunits that can associate with a-rings at one or both ends of the 20S particles (Rechsteiner et al., 2000). The subunits of PA28 have no homology to components of the 19S complex and appear to have arisen relatively recently in evolution since they are present in higher eukaryotes but not in yeast (Tanahashi et al., 1997). The mammalian PA28 complex is composed of three identical a-subunits and three or four identical beta-subunits, which are constitutively expressed in many cells and tissues (Rechsteiner et al., 2000). However, transcrition and cellular content of a- and b-subunits increases severalfold, and in some cells dramatically, after treatment with the cytokine interferon-g (Honore et al., 1993; Realini et al., 1997; Ahn et al., 1995). The PA28 ring has a central opening through which substrates may pass into the 20S particle or products may exit (Knowlton et al., 1997; KoÈhler et al., 2001; Whitby et al., 2000). A homologous complex composed of PA28 g-subunits also exists in the nucleus of mammalian cells (Realini et al., 1997), as well as in Drosophila (Masson et al., 2001) and even trypanosomes (PA26) (Whitby et al., 2000), but it is not regulated by interferon-g. Recently the complex of PA26 with the yeast 20S proteasome has been solved by X-ray diffraction, which has shown that binding of this complex opens the gate in the a-ring of the proteasome (Whitby et al., 2000). This effect clearly can explain the ability of PA28 to dramatically enhance the ability of the latent 20S proteasome to cleave small peptide substrates by its three active sites. Opening the channel in the a-ring may also facilitate the exit of peptide products generated within the proteasome, and several recent observations support such a possible role in vivo (KoÈhler et al., 2001; Stohwasser et al., 2000a; Whitby et al., 2000). However, a number of earlier studies had also argued that PA28 may allosterically activate the different active sites (Rechsteiner et al., 2000). It is noteworthy that binding of PA28 and opening of the peripheral gate, while stimulating hydrolysis of peptides up to 50-fold, does not increase the rate of degradation by proteasomes of proteins or ubiquitinated proteins (i.e., the physiological substrates), but can alter where the proteins are cleaved (see below). It remains unclear whether PA28 mediates these effects by promoting the premature release of peptides that would otherwise be cleaved further within the proteasome and/or by stimulating allosterically the activity of its proteolytic sites. The actual role of PA28 in vivo has remained highly controversial, despite appreciable evidence suggesting a role in stimulating antigen presentation. The well-studied PA28±20S proteasome complexes are unlikely to be important in protein breakdown since formation of these structures and

generation of mhc class i-presented peptides

13

their peptidase activity do not require ATP, unlike breakdown of proteins. PA28 has been reported to be essential for formation of specialized immunoproteasomes (see below) (Preckel et al., 1999), but this conclusion has been refuted (Murata et al., 2001). Moreover, PA28, unlike components of the 26S proteasome, is not essential for protein degradation or MHC class I presentation generally. In fact, mice lacking this complex are viable, grow at normal rates, and can present some viral antigens normally (T. Chiba, S. Murata, and K. Tanaka, ``Normal Immunoproteasome Assembly in Mice Lacking Both PA28a and b.'' Abstract from the meeting ``Proteolysis and Biological Control,'' Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, May 2±6, 2001). Recently, evidence has been obtained that PA28 in vivo may actually exist in larger stuctures that contain both 20S particles and the 19S regulatory complex (Hendil et al., 1998; Tanahashi et al., 2000). These ®ndings were obtained initially by immunoprecipitation (Hendil et al., 1998), and after interferon-g treatment, these hybrid structures compose a large fraction (up to 40%) of the proteasomes in the cell (Tanahashi et al., 2000). In this hybrid structure, PA28 appears to stimulate the hydrolysis of a nonubiquitinated protein, ornithine decarboxylase, by an ATP-dependent process (Tanahashi et al., 2000). Although puri®cation and characterization of these intracellular complexes has proven impossible due to their extreme lability, they have been reconstituted in vitro, and the PA28_ -ring has been shown to associate speci®cally with the asymetric form of the 26S proteasome containing one 19S cap (P. Cascio, in preparation). This association generates a structure that upon electron microscopy behaves as a 19S±20S±PA28 hybrid complex. Formation of this complex enhances hydrolysis of small peptides (but not proteins) apparently by opening the gate at the end of the 20S particle opposite to the 19S particle. Most likely, in such a complex, nonubiquitinated or ubiquitinated proteins are degraded in an ATPdependent manner, but cleavage patterns may change and products may now leave through the PA28-bound end. Accordingly, these hybrid complexes generate a set of peptide products that are different from those released by pure 26S proteasomes (P. Cascio, in preparation). This, presumably, may account for the ability of PA28 to enhance antigen presentation. G. The Role of Proteasomes in Protein Degradation in Vivo In yeast, the deletion of genes encoding any of the 20S subunits and most 19S components is lethal, indicating that these particles and most of their subunits play an essential function (Hilt and Wolf, 1995). Therefore it has been dif®cult to use mutational analysis to test for new roles of these particles in vivo. A major advance in the study of the biological role of proteasomes in mammalian cells was the development of selective inhibitors that can block proteasome functions in intact cells. The ®rst compounds that were identi®ed were small

14

kenneth l. rock et al.

peptide aldehydes that were substrate analogs that bound to active sites and functioned as transition state analogs (Vinitsky et al., 1992). A variety of such reversible inhibitors (e.g., Ac±Leu±Leu±nLeu±al and Z±Leu±Leu±Leu±al, commonly termed ALLN and MG132, respectively) that could rapidly enter living cells and reversible block protein breakdown (Lee and Goldberg, 1998b; A. Kisselev and A. L. Goldberg, submitted for publication) were identi®ed. The peptide aldehyde inhibitors, such as ALLN, did not affect metallo- or aspartyl proteases but could block the activity of certain cysteine proteases (e.g., calpains) (Tsubuki et al., 1996; Wang and Yuen, 1994). However, their effects on proteasomes and potential effects on cysteine proteases could be distinguished by using several related compounds that differed in their activity (Ki ) against the two sets of proteases or by comparisons with the effects of selective inhibitors of lysosomal proteases (e.g., E-64 or weak bases, such as chloroquine). Subsequently, several additional classes of proteasome inhibitors have been identi®ed that are highly speci®c. One very selective, irreversible inhibitor is lactacystin (Omura et al., 1991). This natural product undergoes spontaneous transition to an active form clastolactacystin±b-lactone, which reacts covalently with the N-terminal threonine on the catalytically active b-subunits of the proteasome (Fenteany et al., 1995; Dick et al., 1996). Although this interaction is covalent, this bond is slowly hydrolyzed (t1=2  20 h) with release of the bound inhibitor and regeneration of the active proteasomes. Other more recently described and highly speci®c inhibitors are the peptide vinyl sulfones, which covalently modify the active sites (Bogyo et al., 1997). Peptide boronic acids (e.g. MG262) (Adams et al., 1998) and natural epoxyketones epoxomycin (Meng et al., 1999b) and eponemycin (Meng et al., 1999a) are very potent inhibitors due to their unusually slow off-rates. Epoxyketones are the most selective inhibitors of the proteasome known (Groll et al., 2000). These inhibitors primarily affect the chymotrypsin-like site, but also can block other active sites of the proteasome. Therefore, some sites are blocked at lower concentrations of inhibitor than others. However, at high concentrations, lactacystin and the peptide aldehydes have been shown to block all three active sites. When cells are treated with the peptide aldehyde inhibitors or lactacystin, there is a dose-dependent inhibition of protein degradation (Rock et al., 1994; Craiu et al., 1997b). It is dif®cult or impossible to relate the inhibition of the different active sites to the reduction in protein breakdown, since inhibition of one site may lead to enhanced cleavages at others (Kisselev et al., 1999a; Schmidtke et al., in press). In addition, much of the eventual reduction in proteolysis may be secondary due to the accumulation of fragments within the particles and of undergraded polyubiquitinated proteins in the cells (Figueiredo-Pereira et al., 1994; Schubert et al., 2000). The degradation of normal and abnormal short-lived proteins was inhibited as expected from the earlier studies with ubiquitin conjugation mutants (Ciechanover et al., 1984) and cell-free

generation of mhc class i-presented peptides

15

biochemical studies. In addition, these agents were found in mammalian cells, but not in lower eukaryotes (Lee and Goldberg, 1996), to inhibit also the degradation of long-lived proteins, which constitute the bulk of cellular proteins (Craiu et al., 1997b; Rock et al., 1994). At high concentrations these inhibitors can block the degradation of all of these classes of proteins by more than 90%. Therefore, proteasomes are responsible for degrading the majority of cellular proteins under normal conditions. Many studies have studied individual short-lived regulatory proteins, such as the tumor suppressor protein p53 (Scheffner et al., 1993), the inhibitor of NFkB, IkB (Tanaka et al., 2001), various transcription factors (e.g., b-catenin, c-Jun) (Aberle et al., 1997) (Treier et al., 1994), the cyclin-dependent kinase inhibitor p27 (Pagano et al., 1995), and others. They have demonstrated that the proteasome inhibitors block their degradation and lead to the accumulation of these proteins in their active forms in cells. In other words, in the absence of degradation, there is an accumulation of ubiquitinated species, most of which are rapidly deubiquitinated by cellular isopeptidases, leaving the functional polypeptides. Cells tolerate a marked inhibition of protein degradation by proteasome inhibitors surprisingly well for at least 10 h (Craiu et al., 1997b). They continue to synthesize proteins at normal rates and exclude vital dyes, although cell composition clearly changes as a consequence of the accumulation of shortlived proteins (Figueiredo-Pereira et al., 1994; Schubert et al., 2000; Vinitsky et al., 1994). With time, there is also a buildup of misfolded or damaged proteins which triggers the expression of heat shock proteins and the ER stress response (Bush et al., 1997; Lee and Goldberg, 1998a; Zimmermann et al., 2000). Cells arrest in cell cycle, at least in part, due to the inhibition of cyclin degradation and, depending on the cell type and degree of inhibition, viability decreases. Apoptosis usually occurs 4± 48 h after inhibitor addition, presumably due to the disrupted regulation of essential cell functions (Drexler, 1997). A variety of mechanisms can contribute to the initiation of the cell death program, such as the increase in cellular levels of p53 (Lopes et al., 1997), the lack of inhibition of apoptosis by NFkB (Van Antwerp et al., 1998), or the induction of JNK±kinase by the accumulation of abnormal proteins (Meriin et al., 1998). Interestingly, this apoptosis is most marked in cancer cells (An et al., 1998), and as a consequence certain boronate proteasome inhibitors are now in advanced clinical trials for cancer therapy, including multiple myeloma (Adams et al., 1999; Teicher et al., 1999). H. Products of Proteasomal Degradation When isolated, proteasomes degrade proteins all the way to oligopeptides, but in vivo such products never accumulate in measurable amounts due to their rapid degradation by cytosolic or nuclear peptidases to single amino acids (Falk et al., 1990). The sizes of the products generated by puri®ed proteasomes have

16

kenneth l. rock et al.

been systematically analyzed and are very similar independent of the protein substrate. Mammalian (Emmerich et al., 2000; Kisselev et al., 1999b; Wang et al., 1999), yeast (Nussbaum et al., 1998), and even the simpler archaeal (Kisselev et al., 1998) 20S particles produce peptides from 3 to 30 residues in length whose amounts decrease as their length increases according to a lognormal distribution (Kisselev et al., 1999b). About two-thirds of these peptides are less than 7±8 residues in length and thus are too short to serve in antigen presentation (see below). Interestingly, 26S proteasomes produce a similar spectrum except that the average size is slightly shorter (Emmerich et al., 2000; Kisselev et al., 1999b). This difference presumably re¯ects the greater activities of the catalytic sites after association of the proteasome with the 19S particles or more rapid release of peptides from the SDS-activated 20S particles used in such studies (see below). What determines the size of the fragments generated by proteasomes? In the molecular ruler model it was originally proposed that fragment size would be determined by the distance between the adjacent active sites that were assumed to make the two cleavages which generate the peptides (Wenzel et al., 1994). In the symmetric proteasomes from archaebacteria, which contain seven identical sites, equidistant from each other, this distance would generate peptides of eight or nine residues, However, the archaeal proteasomes produce a broad, log-normal size distribution of peptides as do mammalian proteasomes whose sites are not evenly spaced (Kisselev et al., 1998, 1999b). Therefore, the distance between active sites is not a major factor in¯uencing the size of proteasome products. Sequence speci®city of the proteasome's active sites also cannot be an important determinant of the size distribution of products. As mentioned previously the archaeal proteasomes have 14 active sites that are all identical and of chymotrypsin-like type (Seemuller et al., 1995b; Zwickl et al., 1992). These proteasomes produce a size distribution of peptides similar to that of mammalian particles, which contain two chymotryptic, two tryptic and two caspase-like sites at variable distances (Kisselev et al., 1998, 1999b). Moreover, inactivation of the individual sites in mammalian (Schmidtke et al., 1998) or yeast (Dick et al., 1998; Nussbaum et al., 1998) proteasomes alters the speci®c peptides that are generated but not their size distribution. It has therefore been proposed that the size of peptide products is determined by two competing processes: further proteolytic cleavages and the ability of peptides to diffuse out of the particle (Kisselev et al., 1999b). In other words, the longer a polypeptide fragment resides in the proteasome lumen, the more it is likely to be cleaved further by the active sites to smaller peptides that are able to diffuse more easily out of the particle to escape further hydrolysis. In support of this model, it has been found that the mean size of proteasome products increases by 20±30% by a mutation in the a-ring of the 20S (a3DN) that leaves the central channel in an

generation of mhc class i-presented peptides

17

open position (KoÈhler et al., 2001). In fact, in such a mutant most products are longer than eight residues (KoÈhler et al., 2001). Thus although some very small peptides (two or three residues) may exit the proteasome through narrow slits in its lateral walls (Groll et al., 1997), the major route of exit seems to be through the gated channel in the a-ring, whose diameter may in¯uence product size. Moreover, this gate is regulated by proteasome activators PA28 and the 19S particle, which must facilitate product exit, as well as substrate entry (see below). I. Fate of Oligopeptides Produced by Proteasomes The cytosolic proteolytic pathway rapidly converts intracellular proteins to free amino acids without the buildup of polypeptide or small peptide intermediates. Initially, proteasomes, as principal components of this pathway, degrade these proteins into oligopeptides (see above), With the exception of proteasome products serving in MHC class I antigen presentation, the great majority of these peptides serve no known function. Despite their continual production, it has not been possible to detect such peptides free in the cytoplasm (Falk et al., 1990), because they are presumably rapidly hydrolyzed to single amino acids by diverse cytosolic endo- and exopeptidases (Botbol and Scornik, 1983; Tomkinson, 1999). In fact, their buildup would likely be deleterious to cells, since such fragments might interfere with essential intermolecular interactions. This process also subserves an important homeostatic function in cells, especially under fasting conditions, because it provides a continual source of amino acids for de novo protein synthesis and for the production of energy. The peptidases that catalyze this rapid hydrolysis of proteasome products have not been well characterized. In an attempt to identify these enzymes we found that the degradation of these peptides in cytosolic extracts could be inhibited by o-phenanthroline, indicating that metalloproteases are involved in this process (Saric et al., 2001). Cytosolic aminopeptidases, most of which are metallopeptidases, were identi®ed by others to be part of this catabolic pathway, because di- and tri-peptides, but not the longer oligopeptides, accumulated in the cytosol of cells treated with a general inhibitor of aminopeptidases, bestatin. (Botbol and Scornik, 1983, 1997). Therefore, while aminopeptidases appear to predominantly cleave di- and tripeptides to amino acids, a majority of longer peptides must be initially cleaved by other enzymes, presumably the cytosolic endopeptidases, to these smaller fragments before they are further degraded to amino acids (T. Saric et al., in preparation). One cytosolic enzyme that appears important in this process is thimet oligopeptidase (TOP, EC 3.4.24.15). TOP is a ubiquitous and evolutionarily conserved metalloendopeptidase that cleaves many different oligopeptide sequences (Barrett et al., 1995). The addition of highly selective inhibitors of TOP stabilized several peptides added to cell extracts as well as

18

kenneth l. rock et al.

oligopeptides produced by proteasomes (T. Saric et al., in preparation). However, other, still unidenti®ed endopeptidases also were contributing to this process. For archaebacteria it has been proposed that a large peptidase complex, termed tricorn protease, functions in association with aminopeptidases downstream of the proteasome in completing the protein catabolic pathway (Tamura et al., 1996, 1998). However, there is no evidence in eukaryotic cells for a similar large peptidase complex. Nevertheless, it is very likely that, similarly to the proposed pathway in archaea, the products of the proteasome in eukaryotic cells are hydrolyzed into amino acids by the concerted action of both endopeptidases and aminopeptidases. Presumably, the longer oligopeptides are ®rst cleaved into shorter ones (three to six residues) by the action of TOP and other endopeptidases. These fragments are then further hydrolyzed by aminopeptidases, dipeptidases, or tripeptidases to single amino acids (T. Saric et al., in preparation). Very short proteasome products are presumably degraded by aminopeptidases directly to amino acids. Mammalian cell extracts contain several potent aminopeptidases of broad substrate speci®city but no carboxypeptidase activity has been found (Beninga et al., 1998). Some well-characterized aminopeptidases include leucine aminopeptidase (Taylor, 1993), puromycin-sensitive aminopeptidase (Constam et al., 1995; McLellan et al., 1988), and bleomycin-hydrolyzing aminopeptidase (O'Farrell et al., 1999). When peptides are added to cell extracts, their amino-terminal residues can be progressively trimmed by exopeptidases or they can be destroyed by an endoproteolytic attack (Beninga et al., 1998; T. Saric et al., in preparation). Trimming by aminopeptidases can be partially blocked by bestatin or by acetylating the N-terminal amino group of the substrate, which prevents hydrolysis by aminopeptidases (Mo et al., 1999). Leucine aminopeptidase has been found to be particularly active in hydrolyzing certain longer peptides and to be regulated by cytokines (Beninga et al., 1998; Harris et al., 1992) (see below), but depending on the sequence of the peptide and its length other aminopeptidases appear more active (T. Saric et al., in preparation). Thus, it appears likely that the fate of peptides in the cytosol in vivo is determined by a kinetic competition between destruction by peptidases, especially TOP, and trimming by aminopeptidases. Since interferon-g induces the major trimming enzyme, leucine aminopeptidase, it presumably can favor peptide trimming over endoproteolytic cleavage and complete degradation. In vivo, the degradation of peptides obviously is not so effective as to prevent the development of immune responses (Rammensee et al., 1995). Suf®cient amounts of the peptides must survive attack by cytosolic peptidases and become presented by MHC class I molecules. However, presently it is unknown how some peptides, destined to serve in MHC class I presentation, escape from complete destruction in the cytosol. The competition between the

generation of mhc class i-presented peptides

19

destructive and the protecting forces has important consequences in controlling the extent of MHC class I antigen presentation (see below). J. Other Proteolytic Pathways in the Cytoplasm Other cytosolic proteases, such as calpains (Sorimachi et al., 1997) and caspases (Salvesen and Dixit, 1997), undoubtedly can play a role in degrading some cellular proteins. However, these enzymes do not appear to completely degrade proteins to small peptides, but rather provide speci®c, limited cleavage of substrates in order to irreversibly modulate their activities (Carillo et al., 1994; Choi et al., 1997; Kubbutat and Vousden, 1997). In addition, these proteases are all ATP independent whereas the degradation of most cellular proteins in vivo is ATP dependent (Craiu et al., 1997b; Rock et al., 1994). Moreover, if cell extracts are depleted of proteasomes, the remaining proteases are not suf®cient to degrade many substrates. Therefore, the nonproteasomal proteases are not thought to play a role in the degradation of the bulk cellular proteins. The calpains are a group of cytosolic ubiquitously expressed cysteine proteases that require calcium ions for activity. Their physiological functions and substrates are still not well understood. Recently, they have been implicated in proteolytic cleavage of several important regulatory proteins, such as p53 (Kubbutat and Vousden, 1997), cyclin D1 (Choi et al., 1997), and c-Fos and c-Jun (Carillo et al., 1994), and were reported to be involved in apoptosis (Johnson, 2000). The main components of the proteolytic pathway required for executing cell death are caspases, ubiquitous cysteine proteases with a narrow primary speci®city for cleaving at the carboxyl-terminus of aspartate residues (Stennicke and Salvesen, 1998). Thus, caspases are very selective and make only a small number of cuts, usually one, in a substrate molecule, which never results in their complete degradation. The possibility that there may be other important systems for overall protein degradation, besides the ubiquitin±proteasome system, was suggested by the observation that a murine cell line could be adapted to grow in high doses of the proteasome inhibitors vinyl sulfone or lactacystin (Glas et al., 1998). This result was very surprising because proteasomes are essential for viability in yeast and treatment of mammalian cells with high doses of proteasome inhibitors beyond 12 to 24 h resulted in loss of viability (see above). However, even in the presence of high doses of proteasome inhibitors adapted cells could proliferate and were able to generate MHC class I antigenic peptides and to degrade ubiquitinated proteins. Therefore, it was suggested that there was an alternate proteolytic pathway that could be induced in cells and substitute for proteasomes. However, subsequent studies revealed that proteasomes were only partially inhibited in the vinyl sulfone-treated cells and were presumably still involved in selective ubiquitin-dependent protein degradation (Princiotta et al., 2001).

20

kenneth l. rock et al.

Therefore, there is no ®rm evidence for an inducible alternate proteolytic pathway that can substitute for its functions. Subsequent studies suggested that tripeptidyl peptidase II (TPPII) might be important for the survival of cells adapted to grow in the presence of proteasome inhibitors (Geier et al., 1999). TPPII is a subtilisin-type serine peptidase that assembles into a very large (4 MDa) rod-like particle. It can remove tripeptides from the amino-terminus of oligopeptide substrates (Balow et al., 1986), although it may also have very weak endopeptidase activity (Geier et al., 1999). Compared to untreated controls, cells grown in the presence of proteasome inhibitors have increased levels of TPPII and are highly sensitive to inhibitors of TPPII. Moreover, transfection of TPPII into control cells rendered them more resistant to the toxic effects of proteasome inhibitors (Wang et al., 2000). Recently, it has been demonstrated that TPPII, among other proteins, is upregulated in Burkitt's lymphoma cells that are resistant to apoptosis and do not accumulate ubiquitin conjugates in response to toxic doses of proteasome inhibitors. In these cells too, apoptosis could be induced by treating cells with TPPII inhibitors, suggesting that this peptidase may be required for lymphoma cell survival (Gavioli et al., 2001). Precisely how TPPII helps these cells survive is unclear. Presumably, the partially inhibited proteasomes incompletely degrade certain cellular proteins and/or may produce abnormal oligopeptide products whose accumulation is toxic and which TPPII might help destroy. K. Degradation of Proteins in the Endoplasmic Reticulum Cells are able to selectively degrade abnormal or misfolded proteins from the secretory pathway (Plemper and Wolf, 1999b). For example, incompletely assembled complexes of ER membrane proteins (e.g., T cell receptor subunits, Yang et al., 1998), or misfolded integral membrane proteins in the ER (e.g., mutant cystic ®brosis transmembrane conductance regulator, CFTR; Xiong et al., 1999), are degraded before they reach the cell surface. It was originally thought that there must be a proteolytic system resident in the ER to hydrolzye these substrates. However, it was subsequently found that the degradation of many ER proteins was inhibited by proteasome inhibitors. In yeast the degradation of ER proteins was blocked in mutants of certain ubiquitin-conjugating enzymes, UBC6 and UBC7 (Biederer et al., 1997). Since the ubiquitinconjugating enzymes and proteasomes are cytosolic (or associated with the cytosolic surface of the ER membrane (Biederer et al., 1997), these ®ndings implied that ER proteins were being translocated into the cytoplasm and degraded by the proteasome pathway. Translocation of ER proteins into the cytoplasm has been demonstrated for a mutant soluble protein carboxypeptidase, ysc Y, in yeast (Hiller et al., 1996; Plemper and Wolf, 1999a) and for the integral membrane MHC class I molecules (Wiertz et al., 1996a,b) and mutant soluble protein a1-antitrypsin Z (Qu et al., 1996) in mammalian cells. Not only

generation of mhc class i-presented peptides

21

are mutated and improperly folded secretory proteins degraded by this pathway, termed also ERAD for ER-associated degradation, but so are normal membrane proteins, such as 3-hydroxy-3-methylglutaryl-CoA reductase in yeast (Hampton et al., 1996). Although ERAD is now thought to be the major pathway for degrading ER proteins, the ER itself also contains several proteolytic activities. One is the signal peptidase that is responsible for cleaving the leader peptide from newly synthesized proteins as they are cotranslocationally transported into the ER. Other proteolytic activities in the ER are not well studied. The ER clearly contains aminopeptidases although their molecular identi®es are unknown. These peptidases may play a role in degrading signal peptides into amino acids and in processing the N-terminally extended precursors of antigenic peptides to mature MHC class I ligands (see below). Like the cytoplasm, the ER appears to lack carboxypeptidase activity.

III. Antigen Presentation A. Structure and Assembly of MHC Class I Molecules MHC class I molecules have evolved to bind small fragments of most cellular proteins and display them on the surface of cells. Presentation of these peptides allows the immune system to monitor cells that are synthesizing viral or mutant proteins and to selectively eliminate them. The MHC class I molecule contains on one end of its heavy chain a transmembrane and a cytoplasmic domain for anchorage to the cell membrane and on its other end a peptide-binding groove. The peptide binding site on different MHC class I molecules is quite promiscuous so that a single class I molecule can bind a large number of different peptides. It has speci®city for binding the side chain of the C-terminal amino acid and one or two other residues of the peptide. However, all other positions can be occupied by any amino acid, which allows peptides of many different sequences to bind. Stable binding is achieved because additional binding energy comes from interactions with the amino group of the N-terminus, the carboxylate of the C-terminus, and main chain atoms of the peptide backbone which are features common to all peptides. Therefore a limited number of different class I molecules (three to six class I molecules in any individual) can bind sequences that are found within most cell proteins (reviewed in Rammensee, 1995). Although many different peptides can bind to a particular class I molecule, they all must be of the same number of amino acids, typically 8, 9, or 10 residues. This size restriction is a consequence of the peptide-binding groove having closed ends which bind the alpha amino group of the peptide's N-terminus and the carboxylate of its C-terminus (reviewed in Engelhard, 1994; Rammensee, 1995). Occasionally longer peptides bind within the groove

22

kenneth l. rock et al.

by bulging out of the center of the groove, but this mode of binding is not common (Chen et al., 1994). Therefore, the cellular machinery that generates the presented peptides must make peptides of the right size. The peptide is a key component of the MHC class I complex. In the absence of a bound peptide, class I molecules are unstable and dissociate at physiological temperature. Moreover in vivo they are not transported out of the ER to the plasma membrane. The class I heavy and light chains are cotranslationally transported into the ER where they fold and assemble into a heterodimer. The class I dimer associates with several ER proteins including the chaperones calnexin, calreticulin, and ERP57 and tapasin prior to arrival of the antigenic peptides. These molecules are thought to aid assembly of the class I heterodimer and stabilize it until it binds peptide (reviewed in Solheim, 1999). In addition, one of these associated molecules, tapasin, appears to play several roles in the regulating, the traf®cking, and the loading of class I molecules: it binds to the class I molecule and the peptide transporter in the ER, transporter associated with antigen presentation (TAP) (Androlewicz, 1999) (see below), it helps retain the empty class I molecule until it binds peptide (Bangia et al., 1999), and it helps promote the loading of peptides into the class I molecule (Grandea and Van Kaer, 2001). Once the class I molecule binds an appropriate peptide, it is released from this multisubunit complex and is rapidly transported via the default exocytic pathway to the plasma membrane. B. Origin of MHC Class I-Presented Peptides It is believed that the class I-presented peptides can originate from virtually any cellular protein. In support of this concept it has long been known that most viral or transfected genes can serve as a source of class I-derived peptides. Moreover, the peptides that are naturally bound to class I molecules from uninfected cells have been eluted, and a portion of them have been sequenced by mass spectroscopy (for example, Guimezanes et al., 2001; Hunt et al., 1992). These peptides derive from a large number of different cellular proteins. It was originally thought that MHC class I-presented peptides would derive from the hydrolysis of mature cellular proteins. It is clear that some and probably many class I-presented peptides derive from the turnover of mature proteins. For example, class I-presented peptides are readily generated and presented from mature proteins that are microinjected into the cytoplasm of cells (for example, Craiu et al., 1997a; Grant et al., 1995). However, recent evidence suggests that class I-presented peptides may preferentially derive from recently synthesized proteins (Reits et al., 2000; Schubert et al., 2000). It was initially proposed on theoretical grounds that a signi®cant number of class I-presented peptides might derive from mistakes in the translation process (defective ribosomal productsÐDRiPs) (Yewdell et al., 1996). These incom-

generation of mhc class i-presented peptides

23

plete proteins would be rapidly degraded by the ubiquitin±proteasome system, which is the principal pathway that eliminates abnormal polypeptides, and thereby be converted into class I-presented peptides. In support of the concept of DRiPs, it was found that proteasome inhibitors stabilized a signi®cant fraction (approximately 30%) of newly synthesized and polyubiquitinated proteins (Schubert et al., 2000). It was interpreted that this pool of rapidly degraded proteins represented DRiPs. It should be noted that while some of these rapidly degraded proteins may be DRiPs, some will also be normal proteins that naturally have very short half-lives. Also there is no evidence that these proteins contain translational errors in their primary sequence and are not simply defective in folding. As well, even with long half-lives the amount of degradation of the polypeptide (i.e., the number of molecules hydrolyzed per minute) is higher after synthesis due to the exponential nature of the decay curve. Inhibiting protein synthesis was found to reduce by two- to three fold the transport of class I molecules out of the ER. In contrast, the transport of the transferrin receptor was not affected by protein synthesis inhibitors. Since peptides are needed for class I molecules to exit the ER (see above), this selective reduction in the exocytosis of class I molecules could re¯ect a reduction in peptide supply. Similarly, blocking protein synthesis was found to increase the mobility of the peptide transporter TAP (see below) in the ER membrane, which may also re¯ect a reduction in overall peptide supply because TAP's mobility slows when it binds and transports peptides (Reits et al., 2000). These ®ndings are consistent with the hypothesis that newly synthesized proteins are the source of many class I-presented peptides. If this hypothesis is correct, the data do not yet distinguish whether the peptides are generated from newly synthesized DRiPs or normal short-lived proteins. However, these assays are indirect and do not measure the supply of antigenic peptides, and it is also possible that the effects of inhibiting protein synthesis are due to the rapid decay of some short-lived protein(s) that in¯uences the retention of class I molecules in the ER and the mobility of TAP. To understand what cellular processes generate the presented peptides from cellular proteins, it is important to know where these peptides are generated in the cell. The vast majority of class I-presented peptides originate in the cytosol. The strongest evidence for this conclusion comes from the analysis of cells that are de®cient in a peptide transporter, TAP. This ATP-dependent transporter is present in the ER membrane and translocates peptides of about 7±16 residues in length from the cytosol into the lumen of the ER. In cells that lack this transporter, MHC molecules lack peptides and are largely retained in the ER (reviewed in Ritz and Seliger, 2001). There appear to be a few peptides that are generated in the ER, e.g., from signal sequences, or that may translocate into the ER by some other mechanism and which can bind to certain MHC molecules. For example, in TAP-de®cient cells HLA-A2 is expressed at about

24

kenneth l. rock et al.

30% of wild type levels because it binds a peptide derived from signal sequences (Henderson et al., 1992). However, these cases are the exceptions rather than the rule. Therefore, most of the presented peptides must be generated from proteins residing in the cytoplasm or proteins in transit through the cytoplasm of cells. Based on the ®ndings with cells lacking TAP, it is likely that the majority of presented peptides are generated in the cytoplasm. However, some class I-presented peptides are generated from proteins that are resident in the nucleus or mitochondria (for example, Tjoa and Kranz, 1994; Yamazaki et al., 1997), and it is unclear whether they are generated in these subcellular compartments or in the cytoplasm prior to their localization in these compartments. The nucleus contains proteolytic pathways similar to those of the cytoplasm and would be expected to generate presentable peptides. There is some circumstantial evidence that viral proteins that accumulate in Promyelocytic leukemia protein (PML) bodies in the nucleus may be degraded at this site and generate presented peptides (Anton et al., 1999). Many of the class I-presented peptides are derived from cytosolic or nuclear proteins; however, some others derive from proteins, e.g., membrane proteins, that cotranslationally translocate into the ER. Surprisingly, the presentation of epitopes from membrane proteins is most often dependent on TAP and cytosolic proteolysis (Gallimore et al., 1998; Roelse et al., 1994) and therefore the peptides must have been generated in the cytoplasm. Two mechanisms may account for their production in the cytoplasm. In some cases the proteins might fail to translocate into the ER and be degraded in the cytoplasm perhaps due to mistakes in the translation process (DRiPs) or from proteins that fail to be sorted into the ER. In other cases, the proteins may actually enter the ER but be transported back to the cytosol. As discussed above, it is now well established that many resident ER proteins, other components of the membrane, and secreted polypeptides can be rapidly degraded. This process, often termed ER-associated degradation, actually involves their extraction from the ER and hydrolysis by cytosolic proteasomes. In fact the folding of certain molecules in the ER may be quite inef®cient and the misfolded proteins are degraded by this mechanism. For example, even in normal individuals 20% of the newly synthesized CFTR molecules do not fold correctly in the ER and are subsequently degraded by proteasomes in the cytosol (Jensen et al., 1995). One clear example of a class I-presented peptide generated in this way is an epitope derived from tyrosinase. This peptide epitope originally contains an Asn that is modi®ed with N-linked glycan when the protein is translocated into the ER. However, the presented peptide has had its Asn converted to an Asp, which is a consequence of the N-linked sugar being removed by peptide N-glycanase, a cytosolic enzyme (Mosse et al., 1998). This ®nding is clear evidence that the protein was originally in the ER and subsequently transferred back to the cytoplasm.

generation of mhc class i-presented peptides

25

In summary, these data indicate that the proteolytic pathways that generate most class I-presented peptides are operating in the cytoplasm and possibly the nucleus of cells. These are the same compartments where most cellular proteins are degraded by the ubiquitin±proteasome pathway (see above). In fact the turnover of cellular proteins and the generation of presented peptides are not independent processes. There is now abundant evidence that the ubiquitin±proteasome pathway plays a key role in the generation of many class Ipresented peptides. C. Ubiquitination in Antigen Presentation Several types of experiments have demonstrated an important role for ubiquitination in the presentation of certain antigens. In one set of experiments, a reduction in ubiquitin conjugation was found to inhibit antigen presentation. When ovalbumin protein was loaded into the cytosol of cells that had a thermolabile ubiquitin-activating enzyme (E1), the presentation of an ovalbumin peptide on class I was reduced (Michalek et al., 1993). This effect was due to a requirement for ubiquitin conjugation for degradation of the antigen into peptides and not other steps in the pathway. In another set of experiments, enhancing ubiquitination was found to accelerate antigen presentation. The protein antigens were genetically modi®ed at their N-termini so that they became good substrates for E3a, a ubiquitin protein ligase that preferentially ubiquitinates proteins that have bulky or charged residues at their aminotermini, according to the N-end rule (Kwon et al., 1998). Several antigens that were modi®ed to be rapidly degraded were presented more quickly on class I molecules in vitro (Grant et al., 1995) and could stimulate stronger CTL responses in vivo (Tobery and Siliciano, 1997, 1999). In one case such modi®cations did not stimulate antigen presentation presumably because ubiquitination was not rate limiting for degradation or was not involved in the generation of the particular peptide (Goth et al., 1996). In yet other experiments, polyubiquitination of ovalbumin was required to generate a class I-presented peptide in vitro in cell extracts (Ben-Shahar et al., 1997). However, it is clear that ubiquitination is not required for the generation of all class I-presented peptides. This has been clearly shown for a denatured form of ovalbumin in which all potential sites for polyubiquitination have been chemically blocked (i.e., by methylation of free amino groups) so that ubiquitin conjugation to the epsilon amino groups of lysines or the alpha amino group is no longer possible. When cells were injected with ovalbumin that had been denatured and exhaustively methylated, presentation of an ovalbumin peptide still occurred and still required proteasomes (Michalek et al., 1996). Similarly, when denatured and oxidized ovalbumin was incubated with puri®ed 26S proteasomes without ubiquitin or ubiquitin-conjugating enzymes, it was degraded and the appropriate antigenic peptide was generated (Ben-Shahar et al.,

26

kenneth l. rock et al.

1997; Cascio et al., 2001). Presumably denaturation has exposed sites on ovalbumin that allow it to bind to and be translocated into the proteasome without polyubiquitin. It is widely assumed that ubiquitin conjugation is essential for degradation by 26S particles; however, it is well established in vitro that 26S proteasomes degrade some substrates (e.g., ornithine decarboxylase, P21, calmodulin, and casein) without ubiquitin conjugation (reviewed in Pickart, 1997). It is presently unclear to what extent ubiquitination is required to generate class I-presented peptides from different proteins because it is unclear to what extent degradation of proteins generally requires ubiquitin (see above). At the nonpermissive temperature, the temperature-sensitive E1 mutant cells present some antigens normally and there is no reduction in the production of stable class I complexes (Cox et al., 1995), a process that requires peptides. However, the antigen presentation that occurs under these conditions could still be dependent on ubiquitin conjugation because at the nonpermissive temperature E1 activity is only partially reduced (Finley et al., 1984; Kulka et al., 1988). Also, these cells lose viability with time after E1 inhibition which prevents many types of studies. Therefore, experiments with tempreture-sensitive (TS) cells that give negative data are not de®nitive. Given the important role of ubiquitination in overall protein degradation, it seems likely that it will play a major role in generating many class I-presented peptides. However, de®nitively addressing this issue will require the development of more effective methods to block this process in intact cells. As discussed above, ubiquitination marks a protein for rapid degradation but these conjugates are actually hydrolyzed by proteasomes. Therefore, the ®nding that the presentation of some antigens requires ubiquitination implicated proteasomes in this process. Importantly, these ®ndings speci®cally implicate the 26S particle, since it is the only protease that can recognize and degrade ubiquitin-conjugated proteins. D. Role of Proteasome in Antigen Presentation The proteasome plays a critical role in the generation of the majority of MHC class I-presented peptides. The strongest evidence for this conclusion has come from studies analyzing the effects of proteasome inhibitors on antigen presentation. These agents have been shown to block the presentation of ovalbumin, bgalactosidase, in¯uenza nucleoprotein, sendai nucleoprotein, and many other antigens. In addition, when cells are treated with proteasome inhibitors they continue to synthesize class I heavy and light chains at normal rates but these polypeptides fail to assemble into stable complexes, which is precisely the result expected when peptide supply is limiting (see above). Therefore, proteasome inhibitors appear to block the generation of most class I-presented peptides. These effects are due to the inhibitors blocking the activity of proteasomes and not other molecular targets in cells. Although initial studies used peptide

generation of mhc class i-presented peptides

27

aldehyde inhibitors (Harding et al., 1995; Rock et al., 1994) several structurally unrelated extremely speci®c proteasome inhibitors, including lactacystin (Craiu et al., 1997b), peptide boronic acids (Adams et al., 1998; Gardner et al., 2000), peptide vinyl sulfones (Bogyo et al., 1997), and epoxomicin (Schwarz et al., 2000a), also reduce class I antigen presentation. In the case of lactacystin, which does not inhibit other known proteases (except for a very weak activity against TPPII) and in cells only covalently modi®es proteasomes subunits, as with peptide aldehydes, the degree of inhibition of class I presentation correlates nicely with the inhibition of protein degradation (Craiu et al., 1997a; Rock et al., 1994). Furthermore, for inhibitors that are chemically less speci®c, the degree of reduction in antigen presentation correlates with inhibition of the proteasome rather than other proteases (Rock et al., 1994). Moreover, this concern, while important, initially has proven less relevent for studies of antigen presentation because speci®c inhibition of other possible target enzymes (lysosomal or calpains) has no effect on class I presentation (Morrison et al., 1986; Rock et al., 1994). Cells tolerate treatment with proteasome inhibitors well for many hours. Over this period they continue to synthesize proteins normally and remain fully viable. Nevertheless, the inhibition of protein degradation will affect many cellular processes such as progression through the cell cycle (Dietrich et al., 1996) and the pattern of gene expression (through stabilization of short-lived transcription factors; Zimmermann et al., 2000). Cells will also build up damaged or abnormal proteins and this eventually triggers a heat shock response (Lee and Goldberg, 1998a; Meriin et al., 1998). However, the block in antigen presentation is due to a loss of peptide generation by proteasomes rather than to a disruption of other required steps in the pathway. In fact, when antigenic peptides of the correct size (eight or nine residues) have been injected into or expressed from minigenes in cells treated with proteasome inhibitors, the presentation of the minimal peptides, which don't require proteolysis, is not blocked by proteasome inhibitors in contrast to the marked antigens presented from proteins (Rock et al., 1994). Similarly, the loss of stable MHC class I complexes in cells treated with proteasome inhibitors can be reversed by the addition of antigenic peptides (Rock et al., 1994). These results have clearly demonstrated that in proteasome-treated cells the antigen presentation defect is due to an inhibition in the generation of presentable peptides (or longer precursors) while all other steps in the presentation pathway are unimpaired. It is clearly important to perform these kinds of controls in experiments using proteasome inhibitors to rule out pleotropic effects of these agents, especially in long-term exposures and in cells not extensively studied. Several other lines of evidence have independently demonstrated a role for the proteasome in antigen presentation. The ®nding that some antigens require ubiquitination for their presentation on class I molecules (Michalek et al., 1993)

28

kenneth l. rock et al.

has indicated a role for 26S proteasomes in this process because these particles are absolutely required for degrading ubiquitin conjugates. In addition, transfection or deletion of active-site subunits of the proteasome itself (LMP 2 or 7) or of regulatory protein complexes (PA28) has been shown to in¯uence the ef®ciency of presentation of speci®c antigens (see below). Furthermore, it has been demonstrated that puri®ed proteasomes can generate antigenic peptides or their precursors from protein or oligopeptide substrates (discussed below). Although many such model studies with synthetic peptides and 20S proteasomes may be highly arti®cial, recent studies with 26S proteasomes and immunoproteasomes have shown the generation from ovalbumin of antigenic peptides in a manner that appears highly physiological. E. Interferon-g-Induced Changes in Proteasomes and Antigen Presentation The ®rst evidence suggesting that the proteasome might be involved in antigen presentation came from studies that demonstrated that proteasomes from some antigen-presenting cells (e.g., dendritic cells) and other cells, if treated with interferon-g, contained two subunits, LMP2 and LMP7, that are encoded in the major histocompatibility complex (Glynne et al., 1991; Martinez and Monaco, 1991; Ortiz-Navarete et al., 1991). This genetic region, which contains many genes of immunological importance, suggested that these proteins might subserve an immune function, as has indeed turned out to be the case. Sequencing showed close homologies to proteasome b-subunits, in particular subunits now known to contain proteolytic active sites. Subsequent studies revealed that these two MHC-encoded subunits (LMP2 and LMP7) and a third one encoded outside of the MHC (MECL1) (Hisamatsu et al., 1996) are incorporated into proteasomes and alter their peptidase activity in ways that can enhance the generation of some antigenic peptides (Driscoll et al., 1993; Ehring et al., 1996; Gaczynska et al., 1993, 1994, 1996; Groettrup et al., 2001). Because of this role in antigen presentation, proteasomes containing these subunits have been referred to as ``immunoproteasomes.'' In this review we refer to the proteasomes containing LMP2, LMP7, and MECL-1 as immunoproteasomes and to the particles containing b1, b2, and b5 as ``constitutive'' proteasomes. The expression of proteasome b-subunits is induced coordinately by interferon-g, a potent stimulator of antigen presentation (see below). Once expressed they are incorporated into the b-rings of the 20S core particles in place of the homologous constitutively expressed b1, b2, and b5 subunits and thereby form immunoproteasomes. These six b-subunits (LMP2, LMP7, MECL1, b1, b2, and b5) are the only ones that have N-terminal threonine residues (the nucleophile of the proteasome's active sites) and these are catalytically active. The proteasome inhibitors lactacystin and vinyl sulfone bind

generation of mhc class i-presented peptides

29

covalently to these six subunits (and no other ones) (Bogyo et al., 1997; Craiu et al., 1997b) indicating that they are all catalytically active. LMP2, LMP7, and MECL1 incorporate into proteasomes during the assembly of the particle and are unable to insert into preexisting particles (Aki et al., 1994). Therefore, in cells treated with interferon-g, proteasomes often contain both constitutive and immunoproteasomes with the percentage of immunoproteasomes increasing over several days. The 20S core particle has six active sites, two that preferentially cleave after hydrophobic amino acids (chymotrypsin-like), two that cleave after basic residues (tryptin-like), and two that cleave after acidic residues (peptidyl glutamyl peptide hydrolyzing or caspase-like activity). It is now well established that the presence of the immune subunits alters the catalytic activity against peptide substrates, although there had been some discrepancies reported for the exact effects. In many (but not all) studies, increased expression of the immune subunits induced by interferon-g (LMP2, LMP7, and MECL1) stimulated cleavages of small ¯uorogenic peptides after hydrophobic, basic, and branched chain residues while suppressing ones after acidic residues (Aki et al., 1994; Driscoll et al., 1993; Gaczynska et al., 1993a). Similar ®ndings were also obtained when individual cDNAs for these subunits were transfected into the cells (Gaczynska et al., 1994; even though interferon-g induces the coordinate expression of all three together). Deletion of the immune subunits in knockout mice and cells lacking the MHC-encoded subunits (Stohwasser et al., 1996; Van Kaer et al., 1994) had the opposite effect of lowering the chymotryptic and tryptic activities while increasing the caspase-like activity. In some other studies the pattern of changes that was observed was different (e.g., a reduction in chymotryptic activity and no change in tryptic activity; Boes et al., 1994; Eleuteri et al., 1997) for reasons that are not clear. In any case, all of these studies indicate that when the immune subunits replace the constitutive ones they alter the activity of the proteasome's active sites and change where the proteasome cleaves its substrates. Therefore, immunoproteasomes generate a spectrum of oligopeptide products different than that of constitutive particles (described below). The new pattern of peptides that are produced by immunoproteasomes has been suggested to be more favorable for antigen presentation, because the TAP transporter and MHC molecules have speci®city for the P1 position, preferring hydrophobic and basic residues over acidic ones. Transfection of the immunob-subunits into cells has been shown to enhance the presentation of several antigenic peptides (Cerundolo et al., 1995; Sibille et al., 1995; van Hall et al., 2000). On the other hand deletion of LMP2 or LMP7 genes in mice suppresses the presentation of certain antigenic peptides (Fehling et al., 1994; Van Kaer et al., 1994). These changes in the ef®ciency of antigen presentation can often be correlated with an increased ability of puri®ed immunoproteasomes to

30

kenneth l. rock et al.

generate the class I-presented peptide from longer peptides (Stohwasser et al., 1996; Van Kaer et al., 1994). How these immune subunits may alter cleavages to enhance antigen presentation is discussed below. While in some cases the immunoproteasomes are essential for generating particular MHC class I-presented peptides, constitutive proteasomes almost certainly produce many other antigenic peptides. Cells from mice in which the LMP7 gene was disrupted have only a 50% reduction in surface class I expression (Fehling et al., 1994) and LMP2-de®cient cells have normal levels of class I expression (Van Kaer et al., 1994). Similarly, cell lines that have a homozygous deletion in the MHC that eliminates LMP2, LMP7, and the TAP transporter will express normal levels of class I molecules when they are reconstituted with TAP by transfection (Arnold et al., 1992; Cerundolo et al., 1995; Zweerink et al., 1993). Since class I molecules must bind peptides to be transported from the ER and stably expressed on the cell surface, many antigenic peptides must be generated in the LMP2- and LMP7-de®cient cells, although the diversity of the peptides being produced is unknown. There are also examples where interferon-g treatment reduces the presentation of certain antigens (Van den Eynde and Morel, 2001; reviewed in Morel et al., 2000). Puri®ed constitutive proteasomes can generate class I-presented peptides (see below). To fully understand how important immunoproteasomes are (relative to constitutive proteasomes) to overall peptide generation, it will be necessary to examine cells or animals that lack all three immune subunit genes, and these have not yet been generated. Also, interferon-g also induces the proteasome activator PA28, which may in¯uence the function of LMPs (see below). However, a mouse with a targeted disruption of its PA28b gene (see below) was reported by one group to have almost no immunoproteasomes (whether this is due to the de®ciency in PA28 is presently unclearÐsee below). These animals have marked defects in the presentation of multiple antigens. However, they still generate class Ipresented peptides, suggesting that constitutive proteasomes do supply class I-presented peptides (Preckel et al., 1999). F. The Role of PA28 in Antigen Presentation The proteasome activator complex PA28 is normally present in most tissues and its expression is increased by exposure to interferon-g, a potent stimulator of class I antigen presentation (Ahn et al., 1995; Groettrup et al., 1995b). When added to puri®ed 20S particles, it stimulates the cleavage of small peptides (see above) and altered where peptide substrates were cleaved (Dick et al., 1996; Li and Rechsteiner, 2001). This alteration results in the generation of a different set of peptides including, in some cases, more antigenic peptides. In some

generation of mhc class i-presented peptides

31

studies, PA28 stimulated the proteasome to make more antigenic peptides of the correct size for presentation (Dick et al., 1996; Shimbara et al., 1997). These ®ndings suggested that PA28 might stimulate antigen presentation in vivo. In vivo, PA28 appears to play a role in the generation of some, but not all, epitopes. In some studies, transfection of PA28a alone, PA28b alone, or PA28a and PA28b together stimulated the ability of cells to present a peptide generated from a mouse cytomegalovirus antigen (Groettrup et al., 1996) but not from lymphocytic choriomeningitis virus (Schwarz et al., 2000). Mice with targeted deletions in either PA28b alone, or in both PA28a and b, have been constructed. The mice lacking PA28b rapidly degrade their PA28a subunits and lack PA28 complexes altogether. Unexpectedly, these mice also lack immunoproteasomes, implying that PA28a and/or b is required for the assembly of immunoproteasomes. These mice show a marked defect in the presentation of epitopes from multiple antigens (Preckel et al., 1999), but it is unclear whether this is due to the loss of immunoproteasomes alone, or whether PA28 has an additional role. In contrast, mutant mice in which the genes for both the PA28a and b subunits were disrupted, assemble immunoproteasomes normally (Murata et al., 2001). Although the presentation of some epitopes (e.g., melanoma antigen TRP2-derived peptide) is reduced in these mice, other antigens are unaffected (e.g., epitopes from ovalbumin), and the immune response to in¯uenza seemed normal (Murata et al., 2001); thus, the defect in antigen presentation seems to be less profound in these mice, suggesting that much of the defect in the PA28b knockout mice could be due to their lack of immunoproteasomes. Since the PA28a subunit is not present in the PA28b-only knockout mice, these mice should be phenotypically similar to the double PA28a=b knockout mice. Therefore, the difference in the assembly of immunoproteasomes in these two strains of mice is dif®cult to explain. Nevertheless, these studies indicate that PA28 is involved in generation of some but not all class-I-presented peptides. PA28 can associate with both ends of the 20S particle. However, as discussed previously, a hybrid complex has recently been described that consists of the 20S proteasome capped on one end with the 19S complex and on the other end with PA28. Since this hybrid complex can degrade proteins in an ATPdependent manner, while the PA28±20S±PA28 complex cannot, it is likely that it is the form that participates in antigen presentation. PA28 may stimulate antigen presentation by altering the proteasome's pattern of cleavages in proteins or by promoting the release of longer peptides from the interior of proteasomes, as discussed below.

32

kenneth l. rock et al.

G. Are There Nonproteasomal Pathways That Generate Presented Peptides? The clearest example of a nonproteasomal pathway for generating class I-presented peptides has been demonstrated in the ER. HLA-A2 class I molecules bind and present a sequence found in leader peptides. This presentation of this peptide occurs in cells lacking a TAP transporter and is believed to be generated when the signal peptidase removes the signal peptide during the translocation into the ER (Henderson et al., 1992; Wei and Cresswell, 1992). However, most class I molecules do not bind signal sequence peptides generated in this manner. It is also clear that if class I-presentable peptides are generated in the cytoplasm by any mechanism they will be presented on class I molecules. For example, when short oligopeptides are generated from transfected minigenes or when synthetic peptides are injected into the cytoplasm they are presented on class I molecules. Therefore, if nonproteasomal proteases in the cytosol generate such peptides, they will contribute to antigen presentation. It is possible that such alternate pathways play a role in the presentation of at least some peptides. The primary evidence for the existence of nonproteasomal pathways in antigen presentation has come from examples where proteasome inhibitors fail to block antigen presentation. In some of these studies, the presentation of a speci®c antigen was not blocked by the inhibitors (Anton et al., 1998; Vinitsky et al., 1997), or the stable assembly of MHC class I molecules, which is an indirect measure of peptide production, was not completely blocked by proteasome inhibitors (Vinitsky et al., 1997). Unless the proteasomes were completely inhibited in the studies, then they do not provide strong evidence for a nonproteasomal pathway that generates presented peptides. Unfortunately, it is dif®cult to establish de®nitely whether proteasome inhibition was complete and in some studies that purported to demonstrate an alternate pathway (Glas et al., 1998; Wang et al., 2000) it was subsequently shown that the proteasomes were still degrading proteins (Princiotta et al., 2001). Interestingly, the degree to which proteasome inhibitors blocked class I assembly depends on which class I allele was examined. For example the assembly of Kb , Db , Ld , HLA-A1, HLAA68, HLA-A31, HLA-A33, HLA-B51, and HLA-A8 (Bai and Forman, 1997; Benham and Neefjes, 1997; Rock et al., 1994) was markedly inhibited while the assembly of HLA-A3, HLA-A11, and HLA-B35 (Benham et al., 1998) was much less affected. It is interesting that the class I alleles (HLA-A3 and HLA-A11), whose assembly is most resistant to proteasome inhibitors, are the ones that bind peptides with a basic C-terminal residue (Benham et al., 1998) while the sensitive ones bind peptides with hydrophobic ends (Rammensee et al., 1995). Most proteasome inhibitors bind with higher af®nity to the

generation of mhc class i-presented peptides

33

chymotryptic site, which may generate peptides with hydrophobic ends, than to the tryptic sites, which may generate the peptides with basic ends (Bogyo et al., 1997; Fenteany and Schreiber, 1996). Therefore, it is possible that there was incomplete inhibition of the proteasome's tryptic site with continued production of peptides with basic C-termini. Some studies found that the presentation of certain antigens was actually enhanced by proteasome inhibitors (Anton et al., 1998; Schwarz et al., 2000a). This ®nding has been interpreted to suggest that proteasomes are destroying the peptides that are generated by other proteases. In many cases this interpretation may be correct. However, in some cases low doses of proteasome inhibitors enhanced presentation while high doses blocked this process (Schwarz et al., 2000a). The results with high doses of inhibitor strongly suggest that proteasomes are required to present these antigens. The enhancement at low doses might be due to inhibition of one of the proteasome's active sites that is destroying the same peptide or due to the inhibitor binding one active site and allosterically affecting the activity at other sites. No other cytosolic protease has been identi®ed that de®nitively generates class I-presented peptides from whole proteins. Initial reports suggested that TPPII might be involved in generating MHC class I-presented peptides (Geier et al., 1999; Glas et al., 1998). TPPII has aminopeptidase and weak endopeptidase activity against oligopeptide substrates in vitro and its overexpression allows cells with partially inhibited proteasome to survive (and present antigens). However, it is unlikely that TPPII can generate class I-presented peptides from proteins in vivo (Princiotta et al., 2001). In summary, while it is possible that nonproteasomal proteases in the cytosol can generate some presented peptides, the identity of such proteases has not been identi®ed and their role in presentation has not been de®nitively demonstrated. H. In Vitro Analyses of Cleavages by 20S Proteasomes and Immunoproteasomes To bind tightly to MHC class I molecules and to be presented to the immune system, antigenic peptides must be of the correct size (typically 8, 9, or 10 residues) and sequence to bind to a class I molecule. Does the proteasome generate the presented peptide in its ®nal form or as a longer precursor? A number of experiments have addressed this question by characterizing the peptides generated by puri®ed proteasomes. The majority of these studies have incubated 20S particles with long synthetic oligopeptides (typically 24 or so residues) that contain an antigenic epitope. In these reactions, many cleavage products are generated including the presented peptide of the correct size (Eggers et al., 1995; Niedermann et al., 1999; Svensson et al., 1996). Cleavages also occur within the epitope itself, which destroy the antigenic peptide (Cerundolo et al., 1997; Luckey et al., 1998; Niedermann et al., 1995), and outside

34

kenneth l. rock et al.

the epitope, sometimes generating long peptides that perhaps serve as precursors for presentation (Niedermann et al., 1996, 1999; Stoltze et al., 1998). Similar studies have compared the products generated by proteasomes and immunoproteasomes. For some antigens the two particles make the same cleavages but at different rates, resulting in different amounts of the various cleavage products (Cardozo and Kohanski, 1998). For other antigens the proteasomes and immunoproteasomes make some of the same as well as some different cleavages resulting in a distinct pattern of peptide products (Gaczynska et al., 1993b; Groettrup et al., 1995a). For some antigens, immunoproteasomes make more of the correct antigenic peptide than proteasomes (Sijts et al., 2000a,b; reviewed in Van den Eynde and Morel, 2001). The addition of PA28 to these reactions has been reported to increase the number of cleavages that yield the correct antigenic peptide (Groettrup et al., 1995a). A few studies have analyzed the ability of 20S particles to generate presented peptides from large proteins. These studies are more dif®cult to perform because the rates of protein hydrolysis tend to be very low, and a very large number of different peptides are generated. Because of the complexity of the peptides produced from whole proteins, these studies haven't analyzed all of the cleavage products but have focused instead on asking whether the proper presented peptide is generated. The peptide products have been separated by HPLC and the fractions that should contain the correct peptide have been added to antigen-presenting cells and assayed for their ability to stimulate a speci®c T cell. These studies have detected the generation of some of correct antigenic peptide (Ben-Shahar et al., 1997, 1999; Dick et al., 1994; LucchiariHartz et al., 2000; Niedermann et al., 1995, 1996). For one antigen, ovalbumin, a more systematic and quantitative analysis of the antigenic peptides was performed by af®nity purifying these peptide from the reaction mixture with a speci®c antibody. Again, the antigenic peptide in ®nal form was detected. However, in addition even more N-terminally extended precursors were generated and a few that were extended on the C-terminus (Cascio et al., 2001). These results contrasted with some other studies that detected N-extended peptides only rarely when oligopeptide substrates of different antigens were hydrolyzed by 20S particles (Lucchiari-Hartz et al., 2000; Niedermann et al., 1996). Whether this re¯ects a difference between protein versus oligopeptide substrates, the speci®c sequences or technical differences in the experimental systems are presently unclear. While these studies have provided valuable information, most have signi®cant limitations: they have used the 20S particle which is easier to isolate and analyze than the 26S proteasomes. However, 26S particles are the form of the proteasome that degrades proteins in vivo (see above) and this larger complex cleaves proteins differently than the smaller 20S structure (Brown and Monaco, 1993; Emmerich et al., 2000; Kisselev et al., 1999b). Moreover, to obtain

generation of mhc class i-presented peptides

35

suf®cient amounts of cleavage products for analysis, the 20S particle has most often been activated with agents like SDS which may open the central channel and may alter the catalytic properties of the proteasome's active sites leading to changes where cleavages are made. Many studies have used very long nonlinear incubations in which the initial proteasome products can reenter the particles and undergo additional rounds of hydrolysis. In vivo such repeated actions are extremely unlikely because other peptidases are much more active against small peptides than proteasomes that preferentially degrade proteins. Finally many of the studies have not been performed under conditions where peptide yields and protein consumption are quantitated. Therefore, these studies have indicated that the core 20S particle has the potential to generate the presented peptide in its ®nal form as well as in longer forms, although how these ®ndings extrapolate to what happens with 26S particles in living cells is unclear. I. In Vitro Analyses of Cleavages by 26S Proteasomes and Immunoproteasomes As discussed above, the reasons that 26S proteasomes have not been used in most studies is that these larger complexes are more dif®cult to isolate and maintain in active form. Moreover, many of their substrates must be ubiquitinated and presently it is not technically possible to produce suf®cient amounts of ubiquitin-conjugated substrates for in vitro analysis. One study has added an antigenic protein, ovalbumin, to cell extracts that contain proteasomes and ubiquitin-conjugating enzymes (Ben-Shahar et al., 1997). The correct antigenic peptide was generated in these reactions although the other cleavage products were not analyzed. Other studies have used proteins that do not need to be ubiquitinated to be degraded. When b-casein or enolase was added to 26S proteasomes many of the products were found to be different than those generated from 20S particles (Emmerich et al., 2000; Kisselev et al., 1998). In fact, the average size of the peptides produced from 26S proteasomes was slightly smaller than that produced from 20S particles (Kisselev et al., 1999b). Class I-presented peptides have not yet been identi®ed in b-casein so these analyses couldn't characterize the cleavages involved in generating presented peptides Similar studies analyzed ornithine decarboxylase fused with an ovalbumin peptide and found that 26S proteasomes generated the correct eightresidue antigenic peptide as well as an N-extended version of this epitope (BenShahar et al., 1999). ODC is a special protein that is rapidly degraded without requiring ubiquitin conjugation when it is bound by the cofactor antizyme. Similar studies have been performed analyzing the generation of a class I-presented peptide from ovalbumin. In contrast to native proteins, if this molecule is denatured and chemically modi®ed, presentation of the immunodominant epitope SIIINFEKL is not dependent on ubiquitination in vivo (see above). Similarly, it was found that denatured and oxidized ovalbumin could be

36

kenneth l. rock et al.

degraded without ubiquitination by 26S proteasomes at rates suf®ciently high to analyze the cleavage products. Af®nity chromatograpy with a monoclonal antibody speci®c for the presented peptide or N-extended versions was used to isolate quantitatively all fragments generated (Cascio et al., 2001). The correct eight-residue peptide was produced as well as many N-extended precursors and a small amount of some C-terminally extended products (see below). In this same system, SDS-activated 20S proteasomes also generated the correct 8mer and N- and C-extended epitopes, although the proportions of the various products were different from those released by the 26S particle. However, it is noteworthy that the majority of the antigenic peptides were produced as Nextended precursors (see below). In summary, even though the proteolytic core of the 26S complex is the 20S proteasome, the 26S particle makes somewhat different cleavages in proteins than the 20S core by itself. The limited available data indicate that 26S proteasomes and 20S particles generate class I-presented peptides in the ®nal form as well as many N-extended precursors. It will clearly be important to extend these analyses to additional protein antigens. J. What Determines Where Proteasomes Cleave Substrates and Whether Antigenic Peptides Are Produced? It is clear that a typical protein contains many potential peptide sequences that if generated could bind and be presented on class I molecules, but only a minority of these epitopes actually get presented. There are many factors that in¯uence this process (Niedermann et al., 1995; see Yewdell and Bennink, 1999, for review); however, one is whether the peptide is actually generated or is destroyed by proteasomes or subsequently by peptidases in the cytosol. Understanding where proteasomes cleave would aid in understanding why certain speci®cities dominate immune responses and predicting what regions of a protein are likely to be presented to the immune system. Where proteasomes cleave substrates must depend on a number of factors including the speci®city of its three active sites, perhaps the distance between the active sites, cooperativity or allosteric effects in¯uencing catalytic activities, substrate concentration in the proteasome, mode of delivery of the protein in the central chamber, and probably the length of time the substrate dwells in the proteolytic chamber (see above). However, so far only the possible in¯uence of the cleavage speci®city of the proteasome for substrate sequences has been studied. The speci®city of the proteasome's active sites has been characterized to some extent using small peptide substrates and inhibitors. These sites show strong preferences for the P1 position, although several distant residues including the P4 as well as some other residues clearly can in¯uence the cleavages made. The best-characterized speci®city is for the P1 position and the active

generation of mhc class i-presented peptides

37

sites have been named on the basis of this preference. As discussed above, the chymotryptic site cleaves after hydrophobic amino acids, the tryptic site after basic residues, and the caspase-like site after acidic residues, although it also cleaves after large-branched chain groups. Whether these active sites have similar cleavage preferences with large substrates is unclear and somewhat controversial. Some studies have failed to document such preferences as proteasomes hydrolyze small proteins (e.g., Ehring et al., 1996). However in other studies with long oligopeptides or with protein substrates, increases in cleavages after branched chain, hydrophobic, or basic residues were observed (Cardozo and Kohanski, 1998; Cascio et al., 2001). Therefore, the residue at P1 probably in¯uences where cleavages occur, but is only one factor among several. In part this may be because amino acids at positions other than P1 can in¯uence the generation of presented peptides. There are several examples where the amino acid at the carboxy-terminal ¯anking residue (P10 ) in¯uences whether puri®ed proteasomes can generate the proper cleavage and how ef®ciently cells generate the presented peptide (Altuvia and Margalit, 2000; Beekman et al., 2000; Mo et al., 2000; Niedermann et al., 1995, 1996; Theobald et al., 1998). In addition, it is clear that other positions can in¯uence this process although these have not been as well characterized yet (Nazif and Bogyo, 2001). Of particular interest was the ®nding that 26S and 20S immunoproteasomes degrade the model substrate ovalbumin to the potential antigenic epitopes severalfold more rapidly than normal particles, even though the rate of substrate consumption, peptide bond cleavage, and mean product size were indistinguishable. Surprisingly the yield of the generated epitope, SIINFEKL, was not different but the immunoproteasomes were more effective speci®cally in generating N-extended versions. Some of these N-extended peptides were produced two to four times faster than normal. The cleavages made by the two forms of proteasomes could then be quantitiated and in all six cases could be explained by the tendency of immunoproteasome to cleave faster after hydrophobic residues and especially basic residues and slower after acidic groups (Cascio et al., 2001; Kisselev et al., 1999b). Another major factor may be the time a peptide dwells within the proteasome's central chamber and the tendency to diffuse away must depend inversely on the size of the peptide, its af®nity to the inner chamber (about which nothing is known), and the size of openings for diffusion out of the central 20S core. While very small peptides (two to three residues) may perhaps escape further hydrolysis through small cracks in the cylinder walls, the primary route for product release is the opening in the a-ring. As recent studies on yeast mutants indicate, the patency of this exit path determines the mean size of products and in particular the ability of longer peptides to leave the particle

38

kenneth l. rock et al.

(KoÈhler et al., 2001). It is therefore of appreciable interest that PA28, the interferon-g-induced proteasome activator, by associating with 20S particles leaves the gate in the a-ring in an open conformation (Whitby et al., 2000), which presumably would lead to the release of more peptides of suf®cient size to serve in antigen presentation (i.e., eight residues or longer). It is noteworthy that most products of pure 26S proteasomes (perhaps 70%) are too small to serve in antigen presentation (Kisselev et al., 1999b). In addition, perhaps 95% of the time that an ovalbumin is degraded by the 26S particle, the immunodominant peptide is destroyed by internal changes (Cascio et al., 2001). Any factor that reduces the likelihood of much further change (e.g., more rapid exit from the particle) could enhance the yield of longer peptides (> eight residues). However, direct evidence for such speculation is presently lacking. Moreover, this analysis raises many questions and the properties of the interferong-induced hybrid complex (19S±20S±PA28) are basically unknown. It is similarly unclear whether the function of PA28 is only gating the channel and if so it is unclear why this function would not also occur similarly in a doubly capped 19S±20S complex. Systematic studies of the products generated from different proteins by PA28-containing complexes have not been carried out and are technically dif®cult. Based on the cleavages that are observed in known substrates, algorithms are being developed to predict where proteasomes will cleave antigens. Some of these appear to be able to predict many antigenic peptides that get produced (Kuttler et al., 2000; Nussbaum et al., 2001). As we understand more about speci®city of the proteasome's active sites and other factors that in¯uence where cleavages are made, these models should continue to improve and will be very useful. K. How Often Does a Proteasome Generate a Presented Peptide or Its Precursor? If the proteasome were to always cleave a particular protein substrate in an identical site, then a series of nonoverlapping (contiguous) peptides would be generated. However, proteasomes actually produce a complex set of overlapping peptides. For example, when ovalbumin is degraded on average there are at least 40 to 50 cleavages in the substrate but the total number of peptides released is several hundred, such that the discrete product cannot even be distinguished by standard HPLC approaches (Kisselev et al., 1999b). The process thus appears to be a stochastic one in which a large number of different cleavages can occur and vary with each time a protein is taken into the 20S chamber. Therefore, an antigenic peptide will not necessarily be produced each time an antigen molecule is degraded. In fact, the majority of the time, the proteasome produces peptides that are too small to be presented on class I molecules. The 26S proteasome produces

generation of mhc class i-presented peptides

39

peptides of a wide spectrum of sizes that follows a log-normal distribution varying from 2 to > 24 residues. The 20S particle produces peptides in a similar size range. Therefore, the majority of peptide products (two-thirds) are too small (3±7 residues in length) to be presented by class I molecules (Kisselev et al., 1999b). A quantitative analysis of how often puri®ed proteasomes generate a class I-presented peptide has been performed for the antigen ovalbumin. The correct eight-residue peptide (SIINFEKL) or N-terminally extended versions of this peptide are generated by 26S proteasomes 6% of the time and by 26S immunoproteasomes 8% of the time (Cascio et al., 2001). Interestingly, all of these forms of the proteasome generate the N-extended peptides more frequently than the octamer (69% of the time for 20S proteasomes, 83% for 26S immunoproteasomes; Cascio et al., 2001). These results indicate that the class I-presented peptide and its precursors are generated relatively infrequently, perhaps 10% of the time. Similar results have been obtained with 20S proteasomes. Most times the proteasome must be cleaving within the epitope and destroying it or generating peptides with C-terminal extensions that cannot be trimmed and presented (see below). Presumably, the frequency with which a presented peptide or its precursors are generated varies for different antigens. The number of antigenic peptides bound to class I molecules generated from a de®ned number of antigen molecules has been quanti®ed in cells infected with the intracellular bacteria listeria. This number varies with different antigens and different epitopes. In the most ef®cient case, a presented peptide is generated from every three antigen molecules degraded. In other cases this ratio varies from 1 in 10 to 1 in 350 (Pamer et al., 1997). These numbers re¯ect not only the rate of production of these peptides by proteasomes, but also their subsequent trimming and destruction by peptidases (see below). L. In Vivo Analysis of Antigen Cleavage: Evidence for Two Proteolytic Steps in Antigen Processing One of the limitations of the in vitro systems that have been used to analyze how antigens are cleaved is that it is uncertain whether the experimental conditions faithfully reproduce those that occur in intact cells. To map where proteasomes cleave antigens in vivo, small oligopeptide substrates have been introduced into the cytoplasm of cells by electroporation or by transfection with minigenes. When the presented antigenic peptide (e.g., an 8-mer) is introduced into cells, it is transported ef®ciently and presented on class I molecules. Since these peptides do not require proteolysis, their presentation is not blocked by proteasome inhibitors. In contrast, presentation of the same peptides with one or more C-terminal ¯anking residues, when delivered into the cytosol, is blocked by proteasome inhibitors (Craiu et al., 1997a; Mo et al., 1999; Stoltze

40

kenneth l. rock et al.

et al., 1998). This result, which has been observed for ®ve different antigenic peptides, has led to the important conclusion that only the proteasome can make the proper C-terminal cleavages to generate a presented peptide and that cells contain no other proteolytic activities, i.e., no carboxypeptidases, that can remove these residues. This critical role of the proteasome in determining the C-termini of antigenic peptides ®ts well with suggestions made earlier based on ®ndings with immunoproteasomes, where altered peptidase speci®cities should enhance the yield of peptides with C-termini appropriate for transport by TAP and binding to MHC class I molecules (Gaczynska et al., 1994). Very different results, however, have been obtained with MHC-presented peptides that have N-terminal extensions. In this situation, although the extra residues are removed from the presented peptide and the normal peptide is presented, proteasome inhibitors fail to block this process (Craiu et al., 1997a; Mo et al., 1999). High concentrations of different classes of proteasome inhibitors (peptide aldehydes and lactones) failed to block trimming of N-extended precursors. Moreover, this applied to ®ve different peptides of unrelated sequences and with different N-terminal residues. These results indicate that N-terminal ¯anking residues can be ef®ciently removed from peptides by a proteolytic activity in cells that is distinct from the proteasome. These ®ndings also indicate that N-extended peptides, which may be the most frequent versions of antigenic peptides produced by proteasomes (see above) and are also produced in vivo (Paz et al., 1999), can serve as precursors for presentation. In fact such N-extended peptides were presented as ef®ciently as the mature peptide (Craiu et al., 1997a). Thus the proteolytic enzymes that remove these N-terminal extensions must be important in the process of antigen presentation (see below). M. Trimming of Peptides in the Cytoplasm To identify the proteases in cells that may trim N-extended peptides, synthetic versions of N-extended SIINFEKL antigenic peptide were added to cytosolic extracts of HeLa cells. The N-terminal extensions were removed and removal of proteasomes from the extracts did not affect this process. The 11-mer peptide was converted ®rst to a 10-mer, then to a 9-mer, and then an 8mer, indicating the involvement of an aminopeptidase. This process was signi®cantly inhibited by the aminopeptidase inhibitor bestatin. Furthermore, aminopeptidases only degrade peptides that have a free a amino group and it was found that acetylation of this group stabilized the N-extended peptide (Beninga et al., 1998). Similarly, when the N-extended precursors with acetylated Ntermini were injected into cells, trimming and presentation on MHC class I did not occur (Mo et al., 1999). Therefore, aminopeptidases must be playing a key role in this process in vivo. In extracts, there was no trimming of the peptides on

generation of mhc class i-presented peptides

41

the C-terminus, which is consistent with the ®nding that the proteasome is the only activity in cells that can generate the proper C-terminal end of a peptide (see above). Of particular interest was the ®nding that cytosol from cells treated with interferon-g trimmed N-extended epitopes faster than extracts from control cells (Beninga et al., 1998), suggesting that cells contained one or more interferon-inducible aminopeptidase(s). Fractionation of the cytosolic extracts revealed a single peak of aminopeptidase activity that was normally present, but increased severalfold after interferon-g treatment. This peak was identi®ed as the well-characterized enzyme leucine aminopeptidase (Beninga et al., 1998; Harris et al., 1992), an aminopeptidase that, despite its name, can remove virtually all amino acids from the N-termini of oligopeptides. It was also the major activity in HeLa extracts that trimmed an N-extended version of the ovalbumin peptide SIINFEKL. Similar experiments with an N-extended antigenic peptide from vesicular stomatitis virus also showed trimming in cytosolic extracts. Leucine aminopeptidase was not detected in these preparations, indicating that other aminopeptidases can also trim amino-terminal ¯anking residues. Fractionation of the extracts revealed two distinct peaks which were identi®ed as puromycinsensitive aminopeptidase and bleomycin-hydrolyzing aminopeptidase activity (Stoltze et al., 2000), two well-characterized aminopeptidases present in most cells. These potent aminopeptidases are not induced by interferon-g, unlike leucine aminopeptidase. Cells contain other aminopeptidases, dipeptidyl aminopeptidases and tripeptidyl peptidases, including TPPII, which remove tripeptides from the N-termini of polypeptides (Tomkinson, 1999). It remains to be determined whether these enzymes play a role in trimming antigenic peptides. These ®ndings are consistent with the results of experiments showing that the trimming of N-extended peptides was not blocked by proteasome inhibitors in vivo (see above). Three kinds of experiments strongly suggest that these aminopeptidases function in vivo. First, acetylation of the a amino group of Nextended peptides inhibits the presentation of these epitopes (Mo et al., 1999). This minimal modi®cation blocks hydrolysis of these peptides by aminopepidases but not by endopeptidases like the proteasome. Second, treatment of cells with an inhibitor of puromycin-sensitive aminopepidase and bleomycinhydrolyzing aminopeptidase (alanine±alanine±phenyalanine±chloromethylketone) can block the generation in vivo of a presented peptide from VSV (Stoltze et al., 2000). Although these inhibitors are not speci®c to these enzymes, this latter ®nding further suggests that this antigenic epitope is predominantly produced as N-extended precursor in vivo. Thus, the proteasome is primarily generating N-extended versions of this peptide in vivo. Finally, the induction by interferon-g of leucine aminopeptidase (Beninga et al., 1998),

42

kenneth l. rock et al.

together with other components of the antigen presentation pathway (e.g., TAP, LMP2, and LMP7), suggests a special role for this enzyme in the immune response. Bleomycin-hydrolyzing aminopeptidase is blocked by the cysteine protease inhibitor E64. Although alanine±alanine±phenyalanine±chloromethylketone (which inhibits both bleomycin-hydrolyzing aminopeptidase and puromycinsensitive aminopeptidase) blocked the presentation of the N-extended VSV peptides, E64 did not (Stoltze et al., 2000). Presumably, in the presence of E64, puromycin-sensitive aminopeptidase is suf®cient for trimming the VSV peptide in vivo. Interestingly, in another study both E64 and proteasome inhibitors were required to block the presentation of long oligopeptide constructs containing a cytomegalovirus antigenic peptide (Lopez and Del Val, 1997). Since E64 does not inhibit the proteasome, it is likely that E64 is inhibiting a second protease involved in antigen processing, possibly bleomycin-hydrolyzing aminopeptidase. These ®ndings indicate that several aminopeptidases are involved in antigen presentation. However, it remains unclear why multiple aminopeptidases are essential and how they differ in substrate preference, or why leucine aminopeptidase may be important in the process. Presumably, the different aminopeptidases may be more active against one substrate or another. In some cases, the function of these different aminopeptidases may be redundant. Certainly their primary role is not to trim antigenic peptides, but to complete the degradation of proteins to free amino acids. When peptides with very long N-terminal extensions (e.g., 25 residue extensions) are expressed in cells, they are still very ef®ciently trimmed and presented on class I molecules. This trimming is also resistant to proteasome inhibitors (Craiu et al., 1997a). It is possible in these situations that the long N-terminal extension will be removed through the concerted action of both endopeptidases and aminopeptidases. However, further trimming may cease when the proper 8- or 9-mer is generated, because at this length it binds tightly (effectively irreversibly) to the MHC class I molecule and is delivered to the cell surface. In contrast, longer peptides, even if transported by TAP, presumably associate weakly with MHC class I and may undergo ``back transport'' into the cytosol for further trimming, until the proper size to ®t the MHC class I groove is achieved. What prevents aminopeptidases from continuing to hydrolyze antigenic peptides until they are too short for antigen presentation? In vitro, these enzymes do not stop trimming when they reach the eight- or nine-residue antigenic peptide but continue to remove residues (Beninga et al., 1998; Stoltze et al., 2000). Therefore, it is likely that they both create and destroy antigenic peptides (see below). Aminopeptidases play an important role in helping to degrade the products of proteasomes to amino acids (see above). Therefore, it appears that the

generation of mhc class i-presented peptides

43

immune system has taken advantage of this phylogenetically old catabolic system in the cytosol to help it trim precursor peptides to their proper size for antigen presentation. It also seems to have evolved the capacity to increase the cell's capacity to trim longer peptides through the induction of leucine aminopeptidase by interferon-g. N. Trimming of Peptides in the Endoplasmic Reticulum Experimentally, peptides can be directed into the ER by attaching an Nterminal signal sequence that causes the peptide to be cotranslationally transported into the lumen of the ER, where it is removed from the peptide by ``signal peptidase.'' When a leader sequence is attached to an antigenic epitope with an N-terminal extension, the peptide is trimmed and presented on class I molecules. Presentation from these constructs still occurs in cells that lack the TAP transporter. In this situation, presentation can only occur if the peptide is trimmed in the ER, because any peptide trimmed in the cytosol would not be able to gain access to the class I molecules. These experiments indicate that Nextended peptides can be trimmed in the ER, and there is evidence that at least some N-extended peptides are transported from the cytosol into the ER (Paz et al., 1999). The molecular identify of the ER peptidase that trims such peptides is at present unknown. Very different results were obtained when a signal sequence was attached to peptides with a two-residue C-terminal extension. In this case, the C-terminally extended peptides were presented poorly if at all (Eisenlohr et al., 1992a). However, if cells were transfected with a dipeptide carboxypeptidase (anigotensin converting enzyme) that was targeted into the ER, then the C-terminally extended peptide was trimmed and presented (Eisenlohr et al., 1992a). These results indicate that, like the cytoplasm, the ER lacks a carboxypeptidase that can trim off the extra residues. The failure to trim C-terminally extended peptides in the ER has been con®rmed for another construct (Craiu et al., 1997a). These experiments indicate that peptidases in the ER lumen have the potential to trim long N-extended precursor peptides to their proper size. However, they don't indicate whether presented peptides that are transported through TAP are accessible to these luminal peptidases. This is an issue because prior to binding peptides, newly synthesized class I molecules are maintained in a complex in association with the TAP transporter (reviewed in Cresswell et al., 1999). This ®nding raised the possibility that TAP directly transports peptides onto the class I binding groove. In other words, class I molecules might be loaded preferentially with peptides from TAP before the peptides are exposed to the peptidases in the ER lumen. However, recent experiments indicate that peptides transported via TAP are fully accessible to the ER lumen before binding stably to class I molecules. Thus, certain mutations in class I molecules

44

kenneth l. rock et al.

(Beissbarth et al., 2000), or a truncation in the molecule (tapasin) that binds class I molecules to TAP (Bangia et al., 1999), appear to disrupt the association of class I with TAP, but do not interfere with antigen presentation (Beissbarth et al., 2000; Lehner et al., 1998). This implies that the association of class I with TAP is of no functional signi®cance. However, the association of class I molecules with TAP is easily disrupted during solubilization of cells, and it is possible that the class I mutant molecules or truncated tapasin still associates with class I molecules in intact cells. The strongest evidence that TAPtransported peptides are accessible to the ER lumen comes from experiments where an anti-SIINFEKL antibody transported into the ER lumen was shown to block the presentation of SIINFEKL but not other peptides transported through TAP. However, after the peptide was stably bound to class I molecules, it could not be ``stripped off'' by the anti-SIINFEKL antibody (Hilton et al., 2001). Therefore, before the TAP-transported peptide stably bound to class I molecules, it was accessible to molecules in the ER lumen. It is presently uncertain whether N-terminally extended peptides are primarily trimmed in the cytosol, the ER lumen, or both. A few ®ndings suggest that some peptides are likely to be trimmed in the cyotosol. Leucine aminopeptidase is cytosolic and its induction by interferon-g (Beninga et al., 1998) suggests it may play an important role in trimming. Similarly, the two aminopeptidases implicated in trimming N-extended VSV peptides (Stoltze et al., 2000) are both cytosolic, although the effects of the inhibitors used in the experiments on ER enzymes have not been reported. In addition, peptides with amino-terminal extensions that are too long to be transported by TAP are ef®ciently trimmed, presumably in the cytosol, and presented. On the other hand, certain long peptides may be more ef®ciently transported by TAP than their shorter versions (Lauvau et al., 1999). In these cases, long peptides may be preferentially delivered into the ER lumen. As noted above, some N-extended peptides have been detected in the ER in vivo (Paz et al., 1999). More experiments will need to be done to assess the relative importance of trimming in the cytoplasm versus in the ER. O. Destruction of Antigenic Peptides by PeptidasesÐa Third Proteolytic Step in Antigen Processing As reviewed above, nearly all the peptides produced by the proteasome undergo further hydrolysis by endopeptidases and aminopeptidases into amino acids. In fact recent quantitative studies indicate that the great majority of peptides produced (>98%) cannot serve in antigen presentation (Cascio et al., 2001). It was uncertain whether this degradative process may limit antigen presentation or whether antigenic peptides are intrinsically resistant to these peptidases or protected from them in some way, for example, by chaperones. Recently this issue has been investigated in vitro and in vivo.

generation of mhc class i-presented peptides

45

To test for susceptibility to cellular proteases, antigenic peptides or their Nextended precursors have been added to cytosolic extracts and their stability has been studied over time. Peptides representing six different epitopes were all hydrolyzed rapidly, but at different rates. The degradation of all these peptides was blocked by the heavy metal chelator o-phenanthroline, indicating that they were being hydrolyzed by metalloproteases. One of the important metalloproteases responsible for this process was identi®ed as TOP (see above). Puri®ed TOP degraded all six peptides and two highly speci®c inhibitors of TOP slowed the hydrolysis of ®ve of the six peptides (T. Saric et al., in preparation). Therefore TOP is a key enzyme for degradation for many or most peptides, although other peptidases clearly can also contribute to their destruction in extracts. These ®ndings contrasted with earlier publications that reported that antigenic peptides were resistant to hydrolysis by puri®ed TOP (Portaro et al., 1999), although treating cells with an inhibitor of TOP (Silva et al., 1999) could reduce antigen presentation. These ®ndings led the authors to propose that TOP might bind peptides without hydrolyzing them and protect them from degradation. The explanation for these opposite ®ndings is unclear. However, the more recent studies showed that the same source of TOP as used by Portaro et al., (1999), as well as other sources, could degrade most of the antigenic peptides studied (T. Saric et al., in preparation). Therefore, it seems clear now that TOP can rapidly destroy many antigenic peptides both in vitro and in vivo (see below). The one antigenic peptide studied whose hydrolysis in these extracts was not affected by TOP inhibitors was degraded very rapidly by another peptidase. This enzyme was identi®ed as puromycin-sensitive aminopeptidase, based on its sensitivity to puromycin and the aminopeptidase inhibitor bestatin and the particular sensitivity of this peptide to the puri®ed enzyme (T. Saric et al., 2001). These results indicate that antigenic peptides are not inherently resistant to cytosolic peptidases. On the contrary, like other proteasomal products, they are very susceptible to hydrolysis by both endopeptidases and aminopeptidases in the cytosol. To determine whether some potential antigenic peptides are destroyed by peptidases in vivo or may be protected in the cytosol, the effect of overexpressing TOP on antigen presentation in intact cells was investigated. Transfection of the TOP gene markedly inhibited the class I presentation of three different antigenic peptides whether they were expressed in the cytoplasm from minigenes or generated by proteasomes from the full-length protein. In contrast, when one of these peptides was introduced into the ER with a signal sequence, so that it bypassed TOP in the cytoplasm, its presentation was not reduced by TOP transfection. Therefore, TOP was inhibiting antigen presentation by destroying peptides in the cytoplasm and did not affect other steps in the pathway. This conclusion was further reinforced by the ®nding that TOP

46

kenneth l. rock et al.

transfection did not inhibit the presentation on class II molecules of a peptide generated in the endocytic compartment (X. Y. Mo et al., submitted for publication). Moreover, overexpresson of TOP decreased the overall supply of peptides to MHC class I molecules, indicated by a marked reduction in the expression of MHC class I molecules on the cell surface. This reduction in class I expression was reversed when an antigenic peptide was delivered into the ER with a signal sequence, so that it bypassed TOP in the cytosol. These experiments indicate that a majority of antigenic peptides are susceptible to hydrolysis by cytosolic peptidases, and speci®cally by TOP (X. Y. Mo et al., submitted for publication). In control cells (i.e., when TOP is not overexpressed), TOP is clearly important in degrading many proteasome products to small peptides (T. Saric et al., in preparation). To determine whether it also destroys antigenic peptides in normal cells, the effect on antigen presentation of treating cells with a highly speci®c inhibitor of TOP (Cpp±AAF±pAb) was investigated. The presentation of ovalbumin on class I molecules was markedly increased when cells were treated with this agent (X. Y. Mo et al., submitted for publication). Therefore, TOP (and presumably other peptidases) in all cells, by destroying peptides and their precursors, is limiting antigen presentation. These ®ndings indicate that antigenic peptides are not ef®ciently protected from destruction in the cytosol until they are transported by TAP into the ER. However, these experiments do not eliminate the possibility that some protection is afforded to peptides that bind to cytosolic proteins (such as heat shock proteins). If so, then this protection must be partial; and, as peptidases destroy the free peptide, the bound peptide must dissociate and also be degraded. In any case, these ®ndings clearly indicate that whether an antigenic peptide is presented on the surface depends on a kinetic competition between the rate of generation (by proteasome and cytosolic trimming enzymes) and the rate of destruction of the peptide and longer precursors (by cytosolic peptidases before it can reach the TAP transporter). For the purpose of immune surveillance, it may seem counterproductive, or grossly inappropriate, to have cells destroying many of their class I-presentable peptides. This inef®ciency may be an unavoidable consequence of the immune system utilizing for its source of presented peptides the cellular pathway whose major role is to break down proteins into amino acids. On the other hand, having peptides available to peptidases in the cytosol may also be useful for antigen presentation. Having peptide precursors free in the cytoplasm allows cellular enzymes to trim the many long peptides produced by proteasomes to the 8±10 mers needed for antigen presentation. This trimming process can clearly enhance the ef®ciency of immune surveillance. However, very high levels of antigen presentation may be counterproductive. It may be important to limit antigen presentation in the absence of infections to avoid the potential

generation of mhc class i-presented peptides

47

for autoimmunity or in¯ammation. Indeed, in the uninfected state, the level of class I molecules on most cells is relatively low. However, in infections, interferon-g is produced which will shift the balance away from the destruction of peptides to increase their net yield and favor antigen presentation (see below). P. Role of Interferon-g in MHC Class I Antigen Presentation Although most nonlymphoid tissues express very low levels of MHC class I molecules, they can be induced to express high levels of these molecules by proin¯ammatory cytokines, particularly interferon-g. Interferon-g increases the expression of almost every known component of the MHC class I antigen presentation pathway including MHC class I heavy and light chains, TAP, the LMP2 and LMP7 subunits of the immunoproteasome, PA28a, PA28b, and leucine aminopeptidase. The increased numbers of immunoproteasomes should generate more antigenic peptides and their N-extended precursors. Induction of PA28a and PA28b and the formation of mixed 19S±20S PA28 complexes also appears to increase the generation of potential precursors (Groettrup et al., 1995a, 1996; Stohwasser et al., 2000b; Tanahashi et al., 2000; van Hall et al., 2000), although the mechanism remains unclear. The increased level of leucine aminopeptidase is predicted to increase the trimming of the N-extended precursors to the correct size for presentation resulting in more presentable peptides. Furthermore, the increased numbers of TAP transporters should bind and transport more of the antigenic peptides before they are destroyed by cytosolic peptidases. In contrast, the levels of TOP and puromycin-sensitive aminopeptidase, which are involved in the destruction of peptides, are not altered by interferong. (Preliminary suggestions that interferon may reduce the cellular content of TOP (Rock and Goldberg, 1999) have not withstood more systematic analysis.) All these adaptions should shift the balance between the destruction and the production of peptides to favor antigen presentation. Presumably there is some advantage to the class I pathway being relatively quiescent in healthy parenchymal cells (i.e., cells that are not specialized for antigen presentation). It is possible that constitutive high-level expression of some of the antigen-presenting components might be deleterious to some cells, or to the organism. For example, the presence of high levels of immunoproteasomes, PA28 complexes, hybrid proteasomes, or leucine aminopeptidase might somehow affect the normal catabolic processes in some cells. Increased numbers of peptides in the ER might interfere with processes in the ER, e.g., by competing for binding with chaperones. It is also possible that constitutive high-level presentation of autologous (self ) might, under some conditions, predispose individuals to autoimmunity. However, in infections there is a clear advantage to stimulating antigen presentation to enable the immune system to

48

kenneth l. rock et al.

detect pathogens more ef®ciently, and this is achieved through the production of interferon-g early in infections. Q. Generation of Class I-Presented Peptides from Exogenous Antigens To detect infections, the immune system posts sentinal cells in all tissues. These cells are derived from the bone marrow and include dendritic cells and possibly macrophages. They survey the antigens present in the tissue and then migrate and report this information to T lymphocytes in the lymph node or spleen. In many circumstances, these bone marrow-derived cells are the only cells that provide all of the signals that are needed to stimulate immune responses from naive T cells and they are therefore essential for immune surveillance. In a viral infection, these sentinal cells may acquire viral antigens by direct infection. In this case, they will generate class I-presented peptides by the mechanisms described above. However, in infections with tissue-tropic viruses that don't infect the sentinal cells (such as poliovirus, Sigal and Rock, 2000) or when tumors arise (Huang et al., 1994), then the sentinel cells must acquire antigen after it is released from the affected cells into the extracellular ¯uids (reviewed in Heath and Carbone, 2001). Therefore, dendritic cells and macrophages have the specialized ability to internalize antigens from the environment and present their peptides on both MHC class I and class II molecules. This capacity is not shared by most other cells, which can only present on class I molecules peptides derived from proteins they have themselves synthesized. There are at least two different pathways by which macrophages and dendritic cells can generate class I-presented peptides from exogenous antigens. In one pathway, a fraction of the antigen that is internalized into endocytic compartments is transferred into the cytosol (Kovacsovics-Bankowski and Rock, 1995). Once the exogenous protein is in the cytosol, class I-presented peptides are generated by the same mechanisms that are involved in presenting endogenous cellular antigens. Consequently, the presentation of these antigens is blocked by proteasome inhibitors (Brossart and Bevan, 1997; Oh et al., 1997; Shen et al., 1997). In infections with polio virus (Signal and Rock, 2000) and probably several other viruses in vivo (reviewed in Heath and Carbone, 2001), this endosome to cytosol pathway is the only mechanism for presenting exogenous viral antigens. Similarly, this is the major route involved in immune surveillance for tumor (Huang et al., 1994) and tissue antigens (Kurts et al., 1996). In another pathway, class I-presented peptides or their precursors are generated in the endocytic compartment. In this case, agents that raise the pH and inhibit proteolysis in endosomes block class I antigen presentation (MartinezKinader et al., 1995), just as they affect presentation on MHC class II. Cysteine protease inhibitors similarly block this process (Huang et al., 1994; Yee et al.,

generation of mhc class i-presented peptides

49

1997), suggesting that cathepsins are involved in generating class I-presented peptides. The role of cathepsins has been demonstrated more de®nitively by analyzing cells from mice that have targeted mutations in various cathepsins. The generation of a class I-presented peptide is markedly reduced in dendritic cells and macrophages that lack cathepsin S, but is not affected by the absence of cathepsin B, L, or D (L. Shen and K. L. Rock, unpublished). What controls whether an exogenous antigen is presented primarily via the cytosolic (i.e., endosome to cytosol) or the endosomal degradation pathway is unclear. It is likely that this depends, at least in part, on whether the proteases in the endocytic compartment are suf®cient to generate the correct eight- or nineresidue peptide. It is likely that the cathepsins are generally less well suited to generate these exact peptide sequences. In fact this may be one of the reasons that the MHC class II-binding groove is open and binds class II-presented peptides that are very variable in their length and longer than the class I-presented peptides. The ®rst example where class I-presented peptides were generated in endosomes was with a construct in which the eight-residue peptide sequence was placed at the carboxy-terminus and only required a single cleavage to be generated (Song and Harding, 1996). However, internal peptide epitopes can also be generated by this pathway. In some cases whether an antigen is presented by the endosome to cytosol, or the endosomal degradation, pathway depends on its physical form for reasons that aren't clear. Thus ovalbumin bound to iron oxide microspheres is presented almost exclusively by the cytoplasmic pathway, whereas this same antigen in polylactide±polyglycolide microspheres is presented by both pathways (L. Shen et al., in preparation). Possibly this may be due to where the antigen localizes in the endocytic compartment and its accessibility to appropriate proteases or the cytosol (Berthiaume et al., 1995). Exogenous antigens can also be acquired by macrophages and dendritic cells in the form of peptides bound to heat shock proteins (HSPs) of the ER or cytoplasm that are released from infected cells or tumors (reviewed in Anderson and Srivastava, 2000). Heat shock proteins normally bind unfolded proteins and peptides in cells and presumably also some peptides produced by the ubiquitin±proteasome pathway. When released from dying cells, the HSPs bind to receptors on macrophages and dendritic cells and deliver their bound peptides for class I presentation by an as yet unde®ned mechanism (Basu et al., 2001; Binder et al., 2000). R. Pathogens That Interfere with the Generation of Presented Peptides The role of the immune system is to identify and eliminate infections. However, pathogens have evolved under selection pressure from the immune

50

kenneth l. rock et al.

system and several have developed countermeasures to reduce or avoid detection and increase their survival in the host. Some of these mechanisms interfere with the MHC class I antigen presentation. For example several viruses of the herpes family encode proteins that block the TAP transporter or stimulate the degradation of class I molecules (reviewed in Johnson and Hill, 1998). Here we will focus only on those mechanisms that may affect the proteolytic generation of MHC class I-presented peptides. Since the ubiquitin±proteasome pathway is essential for viability, there are only limited ways in which viruses can interfere with this process without killing the host cell. The best-characterized example of evading immune responses by blocking protein degradation is Epstein±Barr virus. This virus infects humans and can persist in a state of latency, in which it expresses a single protein, EBNA1, that is very poorly immunogenic. This is because EBNA1 contains a glycine±alanine repeat that prevents its degradation by proteasomes (Levitskaya et al., 1995, 1997). Consequently, class I-presented peptides are not generated. If this sequence is fused to several other proteins, it similarly interferes with their presentation on class I molecules (Dantuma et al., 2000; Sharipo et al., 1998). In this way latently infected cells escape immune destruction without interfering with the turnover of other cellular proteins. Another potential way in which viruses could reduce antigen presentation without killing the host cell would be to selectively block the interferon-ginduced components of the pathway (LMPs, PA28) or by affecting only certain proteasomal cleavages. There are no de®nitive examples of viruses using these mechanisms to evade immune detection. However, several viral proteins, including HIV-tat (Nakamura et al., 1998; Seeger et al., 1997), tax (Beraud and Greene, 1996), HBx (Hu et al., 1999; Huang et al., 1996; Zhang et al., 2000), and E1A (Turnell et al., 2000), have been found to bind to proteasome subunits and these interactions might affect proteasome function and antigen presentation. Tat has been shown to interfere with PA28 binding to isolated proteasomes (Seeger et al., 1997) and this might affect the generation of class I-presented peptides (see above). The hepatitis B virus X protein (HBx) decreases the activity of the 26S proteasome's chymotryptic and tryptic activities (Hu et al., 1999) and may also interfere with PA28 binding to 20S particles (Fischer et al., 1995), although such effects have not been described in vivo. Again these effects may function to reduce class I antigen presentation. However, it remains to be determined whether these viral proteins actually lower antigen presentation and enable the viruses to evade immune responses. Viral inhibition of immunoproteasomes or aminopeptidases or stimulation of cytosolic peptidase activities that destroy peptides have not yet been described.

generation of mhc class i-presented peptides

51

IV. Conclusion It is now clear that the immune system has used the phylogenetically old and highly conserved protein degradation pathways as a source of peptides for antigen presentation. The ubiquitin±proteasome pathway provides a convenient and relatively comprehensive pool of peptides from almost all cellular proteins and consequently is the major source of class I-presented peptides. In this process the proteasome generates the proper C-terminus of most presented peptides and either the correct N-terminus or longer precursor fragments that have N-¯anking residues. The peptidases that normally hydrolyze these peptides into amino acids also in¯uence antigen presentation. Aminopeptidases, such as leucine aminopeptidase, can trim the N-extended peptides to the proper size for antigen presentation. There may be additional proteases that can generate class I-presented peptides but these are as yet not de®ned. Aminopeptidases and endopeptidase like thimet oligopeptidase can destroy antigenic peptides and limit antigen presentation. The balance between the production and destruction of antigenic peptides determines the level of antigen presentation. This process allows the immune system to monitor the genes expressed by cells and to detect viral or mutant sequences. The immune system has also evolved components that modify the basic housekeeping pathways of proteolysis to make more and often better peptides for antigen presentation. Several of these components alter the activity and speci®city of proteasomes so that this particle makes a different set of peptides or a higher proportion of peptides appropriate for antigen presentation. The LMP and MECL-1 b-subunits of the proteasome and the PA28 complex that associates with proteasomes appear to serve these functions. The expression of these molecules is increased in response to interferon-g that is made during infections. In addition, interferon-g also increases the expression of leucine aminopeptidase, which can trim N-extended peptides to their proper size for presentation. These components probably function synergistically to increase the production of class I-presented peptides and thereby shift the balance between the production and destruction of peptides toward antigen presentation. There is also an important specialized pathway of class I antigen presentation in macrophages and dendritic cells that allows them to monitor the antigens in peripheral tissues and report this information to T cells in lymphoid tissues. This route is the major mechanism for detecting and initiating T cell responses to viral infections and tumors in peripheral tissues. In this pathway the class Ipresented peptides may be generated by the ubiquitin±proteasome pathway in the cytosol or by cathepsins in endosomes.

52

kenneth l. rock et al. References

Aberle, H., Bauer, A., Stappert, J., Kispert, A., and Kemler, R. (1997). Beta-catenin is a target for the ubiquitin±proteasome pathway. EMBO J. 16, 3797±3804. Adams, J., Behnke, M., Chen, S., Cruickshank, A. A., Dick, L. R., Grenier, L., Klunder, J. M., Ma, Y. T., Plamondon, L., and Stein, R. L. (1998). Potent and selective inhibitors of the proteasome: Dipeptidyl boronic acids. Bioorg. Med. Chem. Lett. 8, 333±338. Adams, J., Palombella, V. J., Sausville, E. A., Johnson, J., Destree, A., Lazarus, D. D., Maas, J., Pien, C. S., Prakash, S., and Elliott, P. J. (1999). Proteasome inhibitors: A novel class of potent and effective antitumor agents. Cancer Res. 59, 2615±2622. Ahn, J. Y., Tanahashi, N., Akiyama, K., Hisamatsu, H., Noda, C., Tanaka, K., Chung, C. H., Shibmara, N., Willy, P. J., and Mott, J. D. (1995). Primary structures of two homologous subunits of PA28, a g-interferon-inducible protein activator of the 20S proteasome. FEBS Lett. 366, 37±42. Aki, M., Shimbara, N., Takashina, M., Akiyama, K., Kagawa, S., Tamura, T., Tanahashi, N., Yoshimura, T., Tanaka, K., and Ichihara, A. (1994). Interferon-g induces different subunit organizations and functional diversity of proteasomes. J. Biochem. 115, 257±269. Altuvia, Y., and Margalit, H. (2000). Sequence signals for generation of antigenic peptides by the proteasome: Implications for proteasomal cleavage mechanism. J. Mol. Biol. 295, 879±890. An, B., Goldfarb, R. H., Siman, R., and Dou, Q. P. (1998). Novel dipeptidyl proteasome inhibitors overcome bcl-2 protective function and selectively accumulate the cyclin-dependent kinase inhibitor p27 and induce apoptosis in transformed, but not normal, human ®broblasts. Cell Death Differ. 5, 1062±1075. Anderson, K. M., and Srivastava, P. K. (2000). Heat, heat shock, heat shock proteins and death: A central link in innate and adaptive immune responses. Immunol. Lett. 74, 35±39. Androlewicz, M. J. (1999). The role of tapasin in MHC class I antigen assembly. Immunol. Res. 20, 79±88. Anton, L. C., Schubert, U., Bacik, I., Princiotta, M. F., Wearsch, P. A., Gibbs, J., Day, P. M., Realini, C., Rechsteiner, M. C., Bennink, J. R., and Yewdell, J. W. (1999). Intracellular localization of proteasomal degradation of a viral antigen. J. Cell Biol. 146, 113±124. Anton, L. C., Snyder, H. L., Bennink, J. R., Vinitsky, A., Orlowski, M., Porgador, A., and Yewdell, J. W. (1998). Dissociation of proteasomal degradation of biosynthesized viral proteins from generation of MHC class I-associated antigenic peptides. J. Immunol. 160, 4859±4868. Arnold, D., Driscoll, J., Androlewicz, M., Hughes, E., Cresswell, P., and Spies, T. (1992). Proteasome subunits encoded in the MHC are not generally required for the processing of peptides bound by MHC class I molecules. Nature 360, 171±174. Bai, A., and Forman, J. (1997). The effect of the proteasome inhibitor lactacystin on the presentation of transporter associated with antigen processing (TAP)-dependent and TAP-independent peptide epitopes by class i molecules. J. Immunol. 159, 2139±2146. Balow, R. M., Tomkinson, B., Ragnarsson, U., and Zetterqvist, O. (1986). Puri®cation, substrate speci®city, and classi®cation of tripeptidyl peptidase II. J. Biol. Chem. 261, 2409±2417. Bangia, N., Lehner, P. J., Hughes, E. A., Surman, M., and Cresswell, P. (1999). The N-terminal region of tapasin is required to stabilize the MHC class I loading complex. Eur. J. Immunol. 29, 1858±1870. Barrett, A. J., Brown, M. A., Dando, P. M., Knight, C. G., McKie, N., Rawlings, N. D., and Serizawa, A. (1995). Thimet oligopeptidase and oligopeptidase m or neurolysin. Methods Enzymol. 248, 529±556. Basu, S., Binder, R. J., Ramalingam, T., and Srivastava, P. K. (2001). Cd91 is a common receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin. Immunity 14, 303±313. Baumeister, W., Walz, J., Zuhl, F., and Seemuller, E. (1998). The proteasome: Paradigm of a selfcompartmentalizing protease. Cell 92, 367±380.

generation of mhc class i-presented peptides

53

Bech-Otschir, D., Kraft, R., Huang, X. H., Henklein, P., Kapelari, B., Pollmann, C., and Dubiel, W. (2001). COP9 signalosome-speci®c phosphorylation targets p53 to degradation by the ubiquitin system. EMBO J. 20, 1630±1639. Beekman, N. J., van Veelen, P. A., van Hall, T., Neisig, A., Sijts, A., Camps, M., Kloetzel, P. M., Neefjes, J. J., Melief, C. J., and Ossendorp, F. (2000). Abrogation of ctl epitope processing by single amino acid substitution ¯anking the C-terminal proteasome cleavage site. J. Immunol. 164, 1898±1905. Beissbarth, T., Sun, J., Kavathas, P. B., and Ortmann, B. (2000). Increased ef®ciency of folding and peptide loading of mutant mhc class i molecules. Eur. J. Immunol. 30, 1203±1213. Benaroudj, N., and Goldberg, A. L. (2000). PAN, the Proteasome Activating Nucleotidase from archaebacteria, is a molecular chaperone which unfolds protein substrate. Nature Cell Biol. 2, 833±839. Benham, A. M., Gromme, M., and Neefjes, J. (1998). Allelic differences in the relationship between proteasome activity and MHC class I peptide loading. J. Immunol. 161, 83±89. Benham, A. M., and Neefjes, J. J. (1997). Proteasome activity limits the assembly of mhc class i molecules after IFN-g stimulation. J. Immunol. 159, 5896±5904. Beninga, J., Rock, K. L., and Goldberg, A. L. (1998). Interferon-g can stimulate post-proteasomal trimming of the N terminus of an antigenic peptide by inducing leucine aminopeptidase. J. Biol. Chem. 273, 18734±18742. Ben-Shahar, S., Cassouto, B., Novak, L., Porgador, A., and Reiss, Y. (1997). Production of a speci®c major histocompatibility complex class i-restricted epitope by ubiquitin-dependent degradation of modi®ed ovalbumin in lymphocyte lysate. J. Biol. Chem. 272, 21060±21066. Ben-Shahar, S., Komlosh, A., Nadav, E., Shaked, I., Ziv, T., Admon, A., DeMartino, G. N., and Reiss, Y. (1999). 26S proteasome-mediated production of an authentic major histocompatibility class I-restricted epitope from an intact protein substrate. J. Biol. Chem. 274, 21963±21972. Beraud, C., and Greene, W. C. (1996). Interaction of HTLV-I tax with the human proteasome: Implications for NFkB induction. J. Acquir. Immune De®c. Syndr. Hum. Retrovirol. 13, S76±84. Berthiaume, E. P., Medina, C., and Swanson, J. A. (1995). Molecular size-fractionation during endocytosis in macrophages. J. Cell Biol. 129, 989±998. Biederer, T., Volkwein, C., and Sommer, T. (1997). Role of cue1p in ubiquitination and degradation at the er surface. Science 278, 1806±1809. Binder, R. J., Han, D. K., and Srivastava, P. K. (2000). CD91: A receptor for heat shock protein gp96. Nat. Immunol. 1, 151±155. Boes, B., Hengel, H., Ruppert, T., Multhaup, G., Koszinowski, U. H., and Kloetzel, P. M. (1994). Interferon g stimulation modulates the proteolytic activity and cleavage site preference of 20S mouse proteasomes. J. Exp. Med. 179, 901±909. Bogyo, M., McMaster, J. S., Gaczynska, M., Tortorella, D., Goldberg, A. L., and Ploegh, H. (1997). Covalent modi®cation of the active site threonine of proteasomal beta subunits and the Escherichia coli homolog HSLV by a new class of inhibitors. Proc. Natl. Acad. Sci. USA 94, 6629±6634. Bohley, P., and Seglen, P. O. (1992). Proteases and proteolysis in the lysosome. Experientia 48, 151±157. Botbol, V., and Scornik, O. A. (1983). Peptide intermediates in the degradation of cellular proteins: Bestatin permits their accumulation in mouse liver in vivo. J. Biol. Chem. 258, 1942±1949. Botbol, V., and Scornik, O. A. (1997). Measurement of muscle protein degradation in live mice by accumulation of bestatin-induced peptides. Am. J. Physiol. Endocrinol. Metab. 36, E1149± E1157. Brannigan, J. A., Dodson, G., Duggleby, H. J., Moody, P. C., Smith, J. L., Tomchick, D. R., and Murzin, A. G. (1995). A protein catalytic framework with an N-terminal nucleophile is capable of self-activation. Nature 378, 416±419.

54

kenneth l. rock et al.

Braun, B. C., Glickman, M., Kraft, R., Dahlmann, D., Kloetzel, P.-M., Finley, D., and Schmidt, M. (1999). The base of the proteasome regulatory particle exhibits chaperone-like activity. Nat. Cell Biol. 1, 221±226. Brooks, P., Fuertes, G., Murray, R. Z., Bose, S., Knecht, E., Rechsteiner, M. C., Hendil, K. B., Tanaka, K., Dyson, J., and Rivett, J. (2000). Subcellular localization of proteasomes and their regulatory complexes in mammalian cells. Biochem. J. 346(1), 155±161. Brossart, P., and Bevan, M. J. (1997). Presentation of exogenous protein antigens on major histocompatibility complex class I molecules by dendritic cells: Pathway of presentation and regulation by cytokines. Blood 90, 1594±1599. Brown, M. G., and Monaco, J. J. (1993). Biochemical puri®cation of distinct proteasome subsets. Enzymes Proteins 47, 343±353. Buschmann, T., Fuchs, S. Y., Lee, C. G., Pan, Z. Q., and Ronai, Z. (2000). Sumo-1 modi®cation of mdm2 prevents its self-ubiquitination and increases mdm2 ability to ubiquitinate p53. Cell 101, 753±762. Bush, K. T., Goldberg, A. L., and Nigam, S. K. (1997). Proteasome inhibition leads to a heat-shock response, induction of endoplasmic reticulum chaperones, and thermotolerance. J. Biol. Chem. 272, 9086±9092. Cardozo, C. (1993). Catalytic components of the bovine pituitary multicatalytic proteinase complex (proteasome). Enzymes Proteins 47, 296±305. Cardozo, C., Chen, W. E., and Wilk, S. (1996). Cleavage of pro-x and glu-x bonds catalyzed by the branched chain amino acid preferring activity of the bovine pituitary multicatalytic proteinase complex (20S proteasome). Arch. Biochem. Biophys. 334, 113±120. Cardozo, C., and Kohanski, R. A. (1998). Altered properties of the branched chain amino acidpreferring activity contribute to increased cleavages after branched chain residues by the ``immunoproteasome.'' J. Biol. Chem. 273, 16764±16770. Cardozo, C., Michaud, C., and Orlowski, M. (1999). Components of the bovine pituitary multicatalytic proteinase complex (proteasome) cleaving bonds after hydrophobic residues. Biochemistry 38, 9768±9777. Carillo, S., Pariat, M., Steff, A. M., Roux, P., Etienne-Julan, M., Lorca, T., and Piechaczyk, M. (1994). Differential sensitivity of fos and jun family members to calpains. Oncogene 9, 1679± 1689. Cascio, P., Hilton, C., Kisselev, A. F., Rock, K. L., and Goldberg, A. L. (2001). 26S proteasomes and immunoproteasomes produce mainly N-extended versions of an antigenic peptide. EMBO J. 20, 2357±2366. Cerundolo, V., Benham, A., Braud, V., Mukherjee, S., Gould, K., Macino, B., Neefjes, J., and Townsend, A. (1997). The proteasome-speci®c inhibitor lactacystin blocks presentation of cytotoxic T lymphocyte epitopes in human and murine cells. Eur. J. Immunol. 27, 336±341. Cerundolo, V., Kelly, A., Elliott, T., Trowsdale, J., and Townsend, A. (1995). Genes encoded in the major histocompatibility complex affecting the generation of peptides for tap transport. Eur. J. Immunol. 25, 554±562. Chen, J. M., Dando, P. M., Rawlings, N. D., Brown, M. A., Young, N. E., Stevens, R. A., Hewitt, E., Watts, C., and Barrett, A. J. (1997). Cloning, isolation, and characterization of mammalian legumain, an asparaginyl endopeptidase. J. Biol. Chem. 272, 8090±8098. Chen, Y., Sidney, J., Southwood, S., Cox, A. L., Sakaguchi, K., Henderson, R. A., Apella, E., Hunt, D. F., Sette, A., and Engelhard, V. (1994). Naturally processed peptides longer than nine amino acid residues bind to the class i mhc molecule hla-a2.1 with high af®nity and with different conformations. J. Immunol. 152, 2874±2881. Chen, Z., Hagler, J., Palombella, V. J., Melandri, F., Scherer, D., Ballard, D., and Maniatis, T. (1995). Signal-induced site-speci®c phosphorylation targets IkB alpha to the ubiquitin±proteasome pathway. Genes Dev. 9, 1586±1597.

generation of mhc class i-presented peptides

55

Choi, Y. H., Lee, S. J., Nguyen, P., Jang, J. S., Lee, J., Wu, M. L., Takano, E., Maki, M., Henkart, P. A., and Trepel, J. B. (1997). Regulation of cyclin d1 by calpain protease. J. Biol. Chem. 272, 28479±28484. Ciechanover, A., Finley, D., and Varshavsky, A. (1984). Ubiquitin dependence of selective protein degradation demonstrated in the mammalian cell cycle mutant ts85. Cell 37, 57±66. Ciechanover, A., Heller, H., Elias, S., Haas, A. L., and Hershko, A. (1980). Atp-dependent conjugation of reticulocyte proteins with the polypeptide required for protein degradation. Proc. Natl. Acad. Sci. USA 77, 1365±1368. Ciechanover, A., Hod, Y., and Hershko, A. (1978). A heat-stable polypeptide component of an ATPdependent proteolytic system from reticulocytes. Biochem. Biophys. Res. Commun. 81, 1100±1105. Ciechanover, A., Orian, A., and Schwartz, A. L. (2000). Ubiquitin-mediated proteolysis: Biological regulation via destruction. Bioessays 22, 442±451. Constam, D. B., Tobler, A. R., Rensing-Ehl, A., Kemler, I., Hersh, L. B., and Fontana, A. (1995). Puromycin-sensitive aminopeptidase. Sequence analysis, expression, and functional characterization. J. Biol. Chem. 270, 26931±26939. Cox, J. H., Galardy, P., Bennink, J. R., and Yewdell, J. W. (1995). Presentation of endogenous and exogenous antigens is not affected by inactivation of E1 ubiquitin-activating enzyme in temperature-sensitive cell lines. J. Immunol. 154, 511±519. Craiu, A., Akopian, T., Goldberg, A., and Rock, K. L. (1997a). Two distinct proteolytic processes in the generation of a major histocompatibility complex class i-presented peptide. Proc. Natl. Acad. Sci. USA 94, 10850±10855. Craiu, A., Gaczynska, M., Akopian, T., Gramm, C. F., Fenteany, G., Goldberg, A. L., and Rock, K. L. (1997b). Lactacystin and clasto-lactacystin beta-lactone modify multiple proteasome beta-subunits and inhibit intracellular protein degradation and major histocompatibility complex class I antigen presentation. J. Biol. Chem. 272, 13437±13445. Cresswell, P., Bangia, N., Dick, T., and Diedrich, G. (1999). The nature of the MHC class I peptide loading complex. Immunol. Rev. 172, 21±28. Dahlmann, B., Kopp, F., Kuehn, L., Niedel, B., Pfeifer, G., Hegerl, R., and Baumeister, W. (1989). The multicatalytic proteinase (prosome) is ubiquitous from eukaryotes to archaebacteria. FEBS Lett. 251, 125±131. Dantuma, N. P., Heessen, S., Lindsten, K., Jellne, M., and Masucci, M. G. (2000). Inhibition of proteasomal degradation by the gly±ala repeat of Epstein±Barr virus is in¯uenced by the length of the repeat and the strength of the degradation signal. Proc. Natl. Acad. Sci. USA 97, 8381±8385. Dick, L. R., Aldrich, C., Jameson, S. C., Moomaw, C. R., Pramanik, B. C., Doyle, C. K., Demartino, G. N., Bevan, M. J., Forman, J. M., and Slaughter, C. A. (1994). Proteolytic processing of ovalbumin and beta-galactosidase by the proteasome to yield antigenic peptides. J. Immunol. 152, 3884±3894. 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). Contribution of proteasomal beta-subunits to the cleavage of peptide substrates analyzed with yeast mutants. J. Biol. Chem. 273, 25637±25646. Dick, T. P., Ruppert, T., Groettrup, M., Kloetzel, P. M., Kuehn, L., Koszinowski, U. H., Stefanovic, S., Schild, H., and Ramensee, H.-G. (1996). Coordinated dual cleavages induced by the proteasome regulator PA28 lead to dominant MHC ligands. Cell 86, 253±262. Dietrich, C., Bartsch, T., Schanz, F., Oesch, F., and Wieser, R. J. (1996). P53-dependent cell cycle arrest induced by n-acetyl±1-leucinyl±1-leucinyl±1-norleucinal in platelet-derived growth factorstimulated human ®broblasts. Proc. Natl. Acad. Sci. USA 93, 10815±10819. Drexler, H. C. A. (1997). Activation of the cell death program by inhibition of proteasome function. Proc. Natl. Acad. Sci. USA 94, 855±860.

56

kenneth l. rock et al.

Driscoll, J., Brown, M. G., Finley, D., and Monaco, J. J. (1993). MHC-linked LMP gene products speci®cally alter peptidase activities of the proteasome. Nature 365, 262±264. Driscoll, J., and Goldberg, A. L. (1990). The proteasome (multicatalytic protease) is a component of the 1500-kda proteolytic complex which degrades ubiquitin-conjugated proteins. J. Biol. Chem. 265, 4789±4792. Dubiel, W., Pratt, G., Ferrell, K., and Rechsteiner, M. (1992). Puri®cation of an 11 S regulator of the multicatalytic protease. J. Biol. Chem. 267, 22369±22377. Eggers, M., Boes-Fabian, B., Ruppert, T., Kloetzel, P. M., and Koszinowski, U. H. (1995). The cleavage preference of the proteasome governs the yield of antigenic peptides. J. Exp. Med. 182, 1865±1870. Ehring, B., Meyer, T. H., Eckerskorn, C., Lottspeich, F., and Tampe, R. (1996). Effects of majorhistocompatibility-complex-encoded subunits on the peptidase and proteolytic activities of human 20S proteasomes: Cleavage of proteins and antigenic peptides. Eur. J. Biochem. 235, 404±415. Eisenlohr, L. C., Bacik, I., Bennink, J. R., Bernstein, K., and Yewdell, J. W. (1992a). Expression of a membrane protease enhances presentation of endogenous antigens to MHC class I-restricted T lymphocytes. Cell 71, 963±972. Eisenlohr, L. C., Yewdell, J. W., and Bennink, J. R. (1992b). Flanking sequences in¯uence the presentation of an endogenously synthesized peptide to cytotoxic T lymphocytes. J. Exp. Med. 175, 481±487. Eleuteri, A. M., Kohanski, R. A., Cardozo, C., and Orlowski, M. (1997). Bovine spleen multicatalytic proteinase complex (proteasome)ÐReplacement of x, y, and z subunits by LMP7, LMP2, and mec11 and changes in properties and speci®city. J. Biol. Chem. 272, 11824±11831. Emmerich, N. P., Nussbaum, A. K., Stevanovic, S., Priemer, M., Toes, R. E., Rammensee, H. G., and Schild, H. (2000). The human 26S and 20S proteasomes generate overlapping but different sets of peptide fragments from a model protein substrate. J. Biol. Chem. 275, 21140± 21148. Enenkel, C., Lehmann, A., and Kloetzel, P. M. (1999). GFP-labelling of 26S proteasomes in living yeast: Insight into proteasomal functions at the nuclear envelope/rough er. Mol. Biol. Rep. 26, 131±135. Engelhard, V. H. (1994). Structure of peptides associated with MHC class I molecules. Curr. Opin. Immunol. 6, 13±23. Etlinger, J. D., and Goldberg, A. L. (1977). A soluble ATP-dependent proteolytic system responsible for the degradation of abnormal proteins in reticulocytes. Proc. Natl. Acad. Sci. USA 74, 54±58. Fabunmi, R. P., Wigley, W. C., Thomas, P. J., and DeMartino, G. N. (2001). Interferon g regulates accumulation of the proteasome activator PA28 and immunoproteasomes at nuclear pml bodies. J. Cell Sci. 114, 29±36. Falk, K., RoÈtzschke, O., and Rammensee, H.-G. (1990). Cellular peptide composition governed by major histocompatibility complex class I molecules. Nature 348, 248±251. Fehling, H. J., Swat, W., Laplace, C., Kuhn, R., Rajewsky, K., Muller, U., and von Boehmer, H. (1994). Mhc class I expression in mice lacking the proteasome subunit LMP-7. Science 265, 1234±1237. Fenteany, G., and Schreiber, S. L. (1996). Speci®c inhibition of the chymotrypsin-like activity of the proteasome induces a bipolar morphology in neuroblastoma cells. Chem. Biol. 3, 905±912. Fenteany, G., Standaert, R. F., Lane, W. S., Choi, S., Corey, E. J., and Schreiber, S. L. (1995). Inhibition of proteasome activities and subunit-speci®c amino-terminal threonine modi®cation by lactacystin. Science 268, 726±731. Ferrell, K., Wilkinson, C. R. M., Dubiel, W., and Gordon, C. (2000). Regulatory subunit interactions of the 26S proteasome, a complex problem. Trends Biochem. Sci. 25, 83±88.

generation of mhc class i-presented peptides

57

Figueiredo-Pereira, M. E., Berg, K. A., and Wilk, S. (1994). A new inhibitor of the chymotrypsinlike activity of the multicatalytic proteinase complex (20S proteasome) induces accumulation of ubiquitin±protein conjugates in a neuronal cell. J. Neurochem. 63, 1578±1581. Finley, D., and Chau, V. (1991). Ubiquitination. Annu. Rev. Cell Biol. 7, 25±69. Finley, D., Ciechanover, A., and Varshavsky, A. (1984). Thermolability of ubiquitin-activating enzyme from the mammalian cell cycle mutant ts85. Cell 37, 43±55. Finley, D., Tanaka, K., Mann, C., Feldmann, H., Hochstrasser, M., Vierstra, R., Johnston, S., Hampton, R., Haber, J., McCusker, J., Silver, P., Frontali, L., Thorsness, P., Varshavsky, A., Byers, B., Madura, K., Reed, S. I., Wolf, D., Jentsch, S., Sommer, T., Baumeister, W., Goldberg, A., Fried, V., Rubin, D. M., and Toh-e, A. (1998). Uni®ed nomenclature for subunits of the Saccharomyces cerevisiae proteasome regulatory particle. Trends Biochem. Sci. 23, 244±245. Fischer, M., Runkel, L., and Schaller, H. (1995). Hbx protein of hepatitis b virus interacts with the cterminal portion of a novel human proteasome alpha-subunit. Virus Genes 10, 99±102. Gaczynska, M., Goldberg, A. L., Tanaka, K., Hendil, K. B., and Rock, K. L. (1996). Proteasome subunits x and y alter peptidase activities in opposite ways to the interferon-g-induced subunits LMP2 and LMP7. J. Biol. Chem. 271, 17275±17280. Gaczynska, M., Rock, K. L., and Goldberg, A. L. (1993a). G-interferon and expression of MHC genes regulate peptide hydrolysis by proteasomes. Nature 365, 264±267. Gaczynska, M., Rock, K. L., Spies, T., and Goldberg, A. L. (1994). Peptidase activities of proteasomes are differentially regulated by the major histocompatibility complex-encoded genes for LMP2 and LMP7. Proc. Natl. Acad. Sci. USA 91, 9213±9217. Gallimore, A., Schwarz, K., van den Broek, M., Hengartner, H., and Groettrup, M. (1998). The proteasome inhibitor lactacystin prevents the generation of an endoplasmic reticulum leaderderived T cell epitope. Mol. Immunol. 35, 581±591. Gardner, R. C., Assinder, S. J., Christie, G., Mason, G. G., Markwell, R., Wadsworth, H., McLaughlin, M., King, R., Chabot-Fletcher, M. C., Breton, J. J., Allsop, D., and Rivett, A. J. (2000). Characterization of peptidyl boronic acid inhibitors of mammalian 20S and 26S proteasomes and their inhibition of proteasomes in cultured cells. Biochem J 346(2), 447±454. Gavioli, R., Frisan, T., Vertuani, S., Bornkamm, G. W., and Masucci, M. G. (2001). c-myc overexpression activates alternative pathways for intracellular proteolysis in lymphoma cells. Nat. Cell Biol. 3, 283±288. Geier, E., Pfeifer, G., Wilm, M., Lucchiari-Hartz, M., Baumeister, W., Eichmann, K., and Niedermann, G. (1999). A giant protease with potential to substitute for some functions of the proteasome. Science 283, 978±981. Glas, R., Bogyo, M., McMaster, J. S., Gaczynska, M., and Ploegh, H. L. (1998). A proteolytic system that compensates for loss of proteasome function. Nature 392, 618±622. Glickman, M. H., Rubin, D. M., Coux, O., Wefes, I., Pfeifer, G., Cjeka, Z., Baumeister, W., Fried, V. A., and Finley, D. (1998). A subcomplex of the proteasome regulatory particle required for ubiquitin±conjugate degradation and related to the COP9-signalosome and EIF3. Cell 94, 615±623. Glickman, M. H., Rubin, D. M., Fu, H., Larsen, C. N., Coux, O., Wefes, I., Pfeifer, G., Cjeka, Z., Vierstra, R., Baumeister, W., Fried, V., and Finley, D. (1999). Functional analysis of the proteasome regulatory particle. Mol. Biol. Rep. 26, 21±28. Glotzer, M., Murray, A. W., and Kirschner, M. W. (1991). Cyclin is degraded by the ubiquitin pathway. Nature 349, 132±138. Glynne, R., Powis, S. H., Beck, S., Kelly, A., Kerr, L.-A., and Trowsdale, J. (1991). A proteasomerelated gene between the two abc transporter loci in the class II region of the human MHC. Nature 353, 357±360. Goldberg, A. L., and St. John, A. C. (1976). Intracellular protein degradation in mammalian and bacterial cells: Part 2. Annu. Rev. Biochem. 45, 747±803.

58

kenneth l. rock et al.

Gonen, H., Bercovich, B., Orian, A., Carrano, A., Takizawa, C., Yamanaka, K., Pagano, M., Iwai, K., and Ciechanover, A. (1999). Identi®cation of the ubiquitin carrier proteins, E2s, involved in signal-induced conjugation and subsequent degradation of IkappaBalpha. J. Biol. Chem. 274, 14823±14830. Goth, S., Nguyen, V., and Shastri, N. (1996). Generation of naturally processed peptide/MHC class I complexes is independent of the stability of endogenously synthesized precursors. J. Immunol. 157, 1894±1904. Grandea, A. G., and Van Kaer, L. (2001). Tapasin: An ER chaperone that controls MHC class I assembly with peptide. Trends Immunol. 22, 194±199. Grant, E. P., Michalek, M. T., Goldberg, A. L., and Rock, K. L. (1995). Rate of antigen degradation by the ubiquitin±proteasome pathway in¯uences MHC class I presentation. J. Immunol. 155, 3750±3758. Groettrup, M., Khan, S., Schwarz, K., and Schmidtke, G. (2001). Interferon-g inducible exchanges of 20S proteasome active site subunits: Why? Biochimie 83, 367±372. Groettrup, M., Ruppert, T., Kuehn, L., Seeger, M., Standera, S., Koszinowski, U., and Kloetzel, P. M. (1995a). The interferon-g-inducible 11S regulator (PA28) and the LMP2/LMP7 subunits govern the peptide production by the 20S proteasome in vitro. J. Biol. Chem. 270, 23808±23815. Groettrup, M., Ruppert, T., Kuehn, L., Seeger, M., Standera, S., Koszinowski, U., and Kloetzel, P. M. (1995b). The interferon-g-inducible 11S regulator (PA28) and the LMP2/LMP7 subunits govern the peptide production by the 20S proteasome in vitro. J. Biol. Chem. 270, 23808±23815. Groettrup, M., Soza, A., Eggers, M., Kuehn, L., Dick, T. P., Schild, H., Rammensee, H. G., Koszinowski, U. H., and Kloetzel, P. M. (1996). A role for the proteasome regulator PA28a in antigen presentation. Nature 381, 166±168. Groll, M., Bajorke, M., Kohler, A., Moroder, L., Rubin, D. M., Huber, R., Glickman, M., and Finley, D. (2000). A gated channel into the proteasome core particle. Nat. Struct. Biol. 7, 1062±1067. Groll, M., Ditzel, L., Lowe, J., Stock, D., Bochtler, M., Bartunik, H. D., and Huber, R. (1997). Structure of 20S proteasome from yeast at 2.4-angstrom resolution. Nature 386, 463±471. Groll, M., Kim, K. B., Kairies, N., Huber, R., and Crews, C. M. (2000). Crystal structure of epoxomicin: 20S proteasome reveals a molecular basis for selectivity of alpha, beta-epoxyketone proteasome inhibitors. J. Am. Chem. Soc. 122, 1237±1238. Gronostajski, R. M., Pardee, A. B., and Goldberg, A. L. (1985). The ATP dependence of the degradation of short-and long-lived proteins in growing ®broblasts. J. Biol. Chem. 260, 3344± 3349. Guelin, E., Rep, M., and Grivell, L. A. (1994). Sequence of the AFG3 gene encoding a new member of the FTSH/YME1/TMA subfamily of the AAA-protein family. Yeast 10, 1389±1394. Guimezanes, A., Barrett-Wilt, G. A., Gulden-Thompson, P., Shabanowitz, J., Engelhardt, V. H., Hunt, D. F., and Schmitt-Verhulst, A. M. (2001). Identi®cation of endogenous peptides recognized by in vivo or in vitro generated alloreactive cytotoxic T lymphocytes: Distinct characteristics correlated with CD8 dependence. Eur. J. Immunol. 31, 421±432. 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. Harding, C. V., France, J., Song, R., Farah, J. M., Chatterjee, S., Iqbal, M., and Siman, R. (1995). Novel dipeptide aldehydes are proteasome inhibitors and block the MHC-I antigen-processing pathway. J. Immunol. 155, 1767±1775. Harris, C. A., Hunte, B., Krauss, M. R., Taylor, A., and Epstein, L. B. (1992). Induction of leucine aminopeptidase by interferon-g: Identi®cation by protein microsequencing after puri®cation by preparative two-dimensional gel electrophoresis. J. Biol. Chem. 267, 6865±6869. Heath, W. R., and Carbone, F. R. (2001). Cross-presentation, dendritic cells, tolerance and immunity. Annu. Rev. Immunol. 19, 47±64.

generation of mhc class i-presented peptides

59

Henderson, R. A., Michel, H., Sakaguchi, K., Shabanowitz, J., Appella, E., Hunt, D. F., and Engelhard, V. H. (1992). HLA-A2.1-associated peptides from a mutant cell line: A second pathway of antigen presentation. Science 255, 1264±1266. Hendil, K. B., Khan, S., and Tanaka, K. (1998). Simultaneous binding of PA28 and PA700 activators to 20S proteasomes. Biochem. J. 332, 749±754. Henke, W., Ferrell, K., Bech-Otschir, D., Seeger, M., Schade, R., Jungblut, P., Naumann, M., and Dubiel, W. (1999). Comparison of human COP9 signalsome and 26S proteasome lid. Mol. Biol. Rep. 26, 29±34. Hershko, A. (1996). Lessons from the discovery of the ubiquitin system. Trends Biochem. Sci. 21, 445±449. Hershko, A., and Ciechanover, A. (1998). The ubiquitin system. Annu. Rev. Biochem. 67, 425±479. Hershko, A., Ciechanover, A., Heller, H., Haas, A. L., and Rose, I. A. (1980). Proposed role of ATP in protein breakdown: Conjugation of protein with multiple chains of the polypeptide of ATPdependent proteolysis. Proc. Natl. Acad. Sci. USA 77, 1783±1786. Hershko, A., Ciechanover, A., and Varshavsky, A. (2000). The ubiquitin system. Nat. Med. 6, 1073±1081. Hicke, L. (1997). Ubiquitin-dependent internalization and down-regulation of plasma membrane proteins. FASEB J. 11, 1215±1226. Hicke, L. (1999). Gettin down with ubiquitin: Turning off cell-surface receptors, transporters and channels. Trends Cell Biol. 9, 107±112. 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. Hilt, W., and Wolf, D. H. (1995). Proteasomes of the yeast S. cerevisiae: Genes, structure and functions. Mol. Biol. Rep. 1, 3±10. Hilton, C. J., Dahl, A. M., and Rock, K. L. (2001). Anti-peptide antibody blocks peptide binding to MHC class I molecules in the endoplasmic reticulum. J. Immunol. 166, 3952±3956. Hisamatsu, H., Shimbara, N., Saito, Y., Kristensen, P., Hendil, K. B., Fujiwara, T., Takahashi, E., Tanahashi, N., Tamura, T., Ichihara, A., and Tanaka, K. (1996). Newly identi®ed pair of proteasomal subunits regulated reciprocally by interferon g. J. Exp. Med. 183, 1807±1816. Honore, B., Leffers, H., Madsen, P., and Celis, J. E. (1993). Interferon-gamma up-regulates a unique set of proteins in human keratinocytes. Molecular cloning and expression of the cDNA encoding the RGD-sequence-containing protein IGUP I-5111. Eur. J. Biochem. 218, 421±430. Hori, T., Osaka, F., Chiba, T., Miyamoto, C., Okabayashi, K., Shimbara, N., Kato, S., and Tanaka, K. (1999). Covalent modi®cation of all members of human cullin family proteins by nedd8. Oncogene 18, 6829±6834. Hough, R., Pratt, G., and Rechsteiner, M. (1986). Ubiquitin±lysozyme conjugates: Identi®cation and characterization of an ATP-dependent protease from rabbit reticulocyte lysate. J. Biol. Chem. 261, 2400±2408. Hough, R., Pratt, G., and Rechsteiner, M. (1987). Puri®cation of two high molecular weight proteases from rabbit reticulocyte lysate. J. Biol. Chem. 262, 8303±8313. Hu, Z., Zhang, Z., Doo, E., Coux,O., Goldberg, A. L., and Liang, T. J. (1999). HepatitisB virus X protein is both a substrate and a potential inhibitor of the proteasome complex. J. Virol. 73, 7231±7240. Huang, A. Y. C., Golumbek, P., Ahmadzadeh, M., Jaffee, E., Pardoll, D., and Levitsky, H., (1994). Role of bone marrow-derived cells in presenting MHC class I-restricted tumor antigens. Science 264, 961±965. Huang, J. K., Kwong, J., Sun, E. C. Y., and Liang, T. J. (1996). Proteasome complex as a potential cellular target of hepatitis B virus X protein. J. Virol. 70, 5582±5591. Hunt, D. F., Henderson, R. A., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., Cox, A. L., Appella, E., and Engelhard, V. H. (1992). Characterization of peptides bound to the class I MHC molecule HLA-A2.1 by mass spectrometry. Science 255, 1261±1263.

60

kenneth l. rock et al.

Jager, S., Groll, M., Huber, R., Wolf, D. H., and Heinemeyer, W. (1999). Proteasome beta-type subunits: Unequal roles of propeptides in core particle maturation and a hierarchy of active site function. J. Mol. Biol. 291, 997±1013. Jensen, T. J., Loo, M. A., Pind, S., Williams, D. B., Goldberg, A. L., and Riordan, J. R. (1995). Multiple proteolytic systems, including the proteasome, contribute to CFTR processing. Cell 83, 129±135. Johnson, D. C., and Hill, A. B. (1998). Herpesvirus evasion of the immune system. Curr. Top. Microbiol. Immunol. 232, 149±177. Johnson, D. E. (2000). Noncaspase proteases in apoptosis. Leukemia 14, 1695±1703. Kho, C. J., Huggins, G. S., Endege, W. O., Hsieh, C. M., Lee, M. E., and Haber, E. (1997). Degradation of E2A proteins through a ubiquitin-conjugating enzyme, UbcE2A. J. Biol. Chem. 272, 3845±3851. Kisselev, A. F., Akopian, T. N., Castillo, V., and Goldberg, A. L. (1999a). Proteasome active sites allosterically regulate each other, suggesting a cyclical bite±chew mechanism for protein breakdown. Mol. Cell 4, 395±402. Kisselev, A. F., Akopian, T. N., and Goldberg, A. L. (1998). Range of sizes of peptide products generated during degradation of different proteins by archaeal proteasomes. J. Biol. Chem. 273, 1982±1989. Kisselev, A. F., Akopian, T. N., Woo, K. M., and Goldberg, A. L. (1999b). The sizes of peptides generated from protein by mammalian 26 and 20S proteasomes: Implications for understanding the degradative mechanism and antigen presentation. J. Biol. Chem. 274, 3363±3371. Kisselev, A. F., and Goldberg, A. L. (2001). Proteasome inhibitors: from research tools to drug candidates. Chem. Biol. 8, 739±758. Knowlton, J. R., Johnston, S. C., Whitby, F. G., Realini, C., Zhang, Z., Rechsteiner, M., and Hill, C. P. (1997). Structure of the proteasome activator regalpha (PA28a). Nature 390, 639±643. KoÈhler, A., Bajorek, M., Groll, M., Moroder, L., Rubin, D. M., Huber, R., Glickman, M. H., and Finley, D. (2001a). The substrate translocation channel of the proteasome. Biochimie 83, 325±332. KoÈhler, A., Cascio, P., Leggett, D. S., Woo, K. M., Goldberg, A. L., and Finley, D. (2001b). The axial channel of the proteasome core particle is gated by the rpt2 ATPase and controls both substrate entry and product release. Mol. Cell 7, 1143±1152. Kovacsovics-Bankowski, M., and Rock, K. L. (1995). A phagosome-to-cytosol pathway for exogenous antigens presented on MHC class I molecules. Science 267, 243±246. Kruger, E., Kloetzel, P. M., and Enenkel, C. (2001). 20S proteasome biogenesis. Biochimie 83, 289± 293. Kubbutat, M. H., and Vousden, K. H. (1997). Proteolytic cleavage of human p53 by calpain: A potential regulator of protein stability. Mol. Cell. Biol. 17, 460±468. Kulka, R. G., Raboy, B., Schuster, R., Parag, H. A., Diamond, G., Ciechanover, A., and Marcus, M. (1988). A Chinese hamster cell cycle mutant arrested at G2 phase has a temperature-sensitive ubiquitin-activating enzyme, E1. J. Biol. Chem. 263, 15726±15731. Kurts, C., Heath, W. R., Carbone, F. R., Allison, J., Miller, J. F., and Kosaka, H. (1996). Constitutive class I-restricted exogenous presentation of self antigens in vivo. J. Exp. Med. 184, 923±930. Kutejova, E., Durcova, G., Surovkova, E., and Kuzela, S. (1993). Yeast mitochondrial ATP-dependent protease: Puri®cation and comparison with the homologous rat enzyme and the bacterial ATP-dependent protease la. FEBS Lett. 329, 47±50. Kuttler, C., Nussbaum, A. K., Dick, T. P., Rammensee, H. G., Schild, H., and Hadeler, K. P. (2000). An algorithm for the prediction of proteasomal cleavages. J. Mol. Biol. 298, 417±429. Kwon, Y. T., Reiss, Y., Fried, V. A., Hershko, A., Yoon, J. K., Gonda, D. K., Sangan, P., Copeland, N. G., Jenkins, N. A., and Varshavsky, A. (1998). The mouse and human genes encoding the recognition component of the N-end rule pathway. Proc. Natl. Acad. Sci. USA 95, 7898±7903.

generation of mhc class i-presented peptides

61

Lauvau, G., Kakimi, K., Niedermann, G., Ostankovitch, M., Yotnda, P., Firat, H., Chisari, F. V., and van Endert, P. M. (1999). Human transporters associated with antigen processing (TAPs) select epitope precursor peptides for processing in the endoplasmic reticulum and presentation to t cells. J. Exp. Med. 190, 1227±1239. Lee, D. H., and Goldberg, A. L. (1996). Selective inhibitors of the proteasome-dependent and vacuolar pathways of protein degradation in Saccharomyces cerevisiae. J. Biol. Chem. 271, 27280±27284. Lee, D. H., and Goldberg, A. L. (1998a). Proteasome inhibitors cause induction of heat shock proteins and trehalose, which together confer thermotolerance in Saccharomyces cerevisiae. Mol. Cell. Biol. 18, 30±38. Lee, D. H., and Goldberg, A. L. (1998b). Proteasome inhibitors: Valuable new tools for cell biologists. Trends Cell Biol. 8, 397±403. Lehner, P. J., Surman, M. J., and Cresswell, P. (1998). Soluble tapasin restores MHC class I expression and function in the tapasin-negative cell line 220. Immunity 8, 221±231. Levitskaya, J., Coram, M., Levitsky, V., Imreh, S., Steigerwald-Mullen, P. M., Klein, G., Kurilla, M. G., and Masucci, M. G. (1995). Inhibition of antigen processing by the internal repeat region of the Epstein±Barr virus nuclear antigen-1. Nature 375, 685±688. Levitskaya, J., Sharipo, A., Leonchiks, A., Ciechanover, A., and Masucci, M. G. (1997). Inhibition of ubiquitin/proteasome-dependent protein degradation by the gly±ala repeat domain of the Epstein±Barr virus nuclear antigen 1. Proc. Natl. Acad. Sci. USA 94, 12616±12621. Li, J., and Rechsteiner, M. (2001). Molecular dissection of the 11S reg (PA28) proteasome activators. Biochimie 83, 373±383. Lopes, U. G., Erhardt, P., Yao, R. J., and Cooper, G. M. (1997). p53-dependent induction of apoptosis by proteasome inhibitors. J. Biol. Chem. 272, 12893±12896. Lopez, D., and Del Val, M. (1997). Selective involvement of proteasomes and cysteine proteases in MHC class I antigen presentation. J. Immunol. 159, 5769±5772. Lowe, J., Stock, D., Jap, B., Zwickl, P., Baumeister, W., and Huber, R. (1995). Crystal structure of the 20S proteasome from the archaeon t. acidophilum at 3.4 a resolution. Science 268, 533±539. Lucchiari-Hartz, M., van Endert, P. M., Lauvau, G., Maier, R., Meyerhans, A., Mann, D., Eichmann, K., and Niedermann, G. (2000). Cytotoxic T lymphocyte epitopes of hiv-1 nef: Generation of multiple de®nitive major histocompatibility complex class I ligands by proteasomes. J. Exp. Med. 191, 239±252. Luckey, C. J., King, G. M., Marto, J. A., Venketeswaran, S., Maier, B. F., Crotzer, V. L., Colella, T. A., Shabanowitz, J., Hunt, D. F., and Engelhard, V. H. (1998). Proteasomes can either generate or destroy MHC class I epitopes: Evidence for nonproteasomal epitope generation in the cytosol. J. Immunol. 161, 112±121. Luders, J., Demand, J., and Hohfeld, J. (2000). The ubiquitin-related BAG-1 provides a link between the molecular chaperones Hsc70/Hsp70 and the proteasome. J. Biol. Chem. 275, 4613±4617. Lupas, A., Zuhl, F., Tamura, T., Wolf, S., Nagy, I., De Mot, R., and Baumeister, W. (1997). Eubacterial proteasomes. Mol. Biol. Rep. 24, 125±131. Luzio, J. P., Rous, B. A., Bright, N. A., Pryor, P. R., Mullock, B. M., and Piper, R. C. (2000). Lysosome±endosome fusion and lysosome biogenesis. J. Cell Sci. 113, 1515±1524. Lyapina, S., Cope, G., Shevchenko, A., Serino, G., Tsuge, T., Zhou, C., Wolf, D. A., Wei, N., Shevchenko, A., and Deshaies, R. J. (2001). Promotion of nedd8±cul1 conjugate cleavage by COP9 signalosome. Science 292, 1382±1385. Ma, C. P., Slaughter, C. A., and DeMartino, G. N. (1992). Identi®cation, puri®cation, and characterization of a protein activator (PA28) of the 20 S proteasome (macropain). J. Biol. Chem. 267, 10515±10523.

62

kenneth l. rock et al.

Manoury, B., Hewitt, E. W., Morrice, N., Dando, P. M., Barrett, A. J., and Watts, C. (1998). An asparaginyl endopeptidase processes a microbial antigen for class Ii MHC presentation. Nature 396, 695±699. Martinez, C. K., and Monaco, J. J. (1991). Homology of proteasome subunits to major histocompatibility complex-linked LMP gene. Nature 353, 664±667. Martinez-Kinader, B., Lipford, G. B., Wagner, H., and Heeg, K. (1995). Sensitization of MHC class I-restricted T cells to exogenous proteins: Evidence for an alternative class I-restricted antigen presentation pathway. Immunology 86, 287±295. Masson, P., Andersson, O., Petersens, U. M., and Young, P. (2001). Identi®cation and characterization of a Drosophila nuclear proteasome regulatorÐa homolog of human 11S reg g (PA28 g). J. Biol. Chem. 276, 1383±1390. Mayr, J., Seemuller, E., Muller, S. A., Engel, A., and Baumeister, W. (1998). Late events in the assembly of 20S proteasomes. J. Struct. Biol. 124, 179±188. McCormack, T. A., Cruikshank, A. A., Grenier, L., Melandri, F. D., Nunes, S. L., Plamondon, L., Stein, R. L., and Dick, L. R. (1998). Kinetic studies of the branched chain amino acid preferring peptidase activity of the 20S proteasome: Development of a continuous assay and inhibition by tripeptide aldehydes and clasto-lactacystin beta-lactone. Biochemistry 37, 7792±7800. McLellan, S., Dyer, S. H., Rodriguez, G., and Hersh, L. B. (1988). Studies on the tissue distribution of the puromycin-sensitive enkephalin-degrading aminopeptidases. J. Neurochem. 51, 1552± 1559. Medd, P. G., and Chain, B. M. (2000). Protein degradation in MHC class II antigen presentation: Opportunities for immunomodulation. Semin. Cell Dev. Biol. 11, 203±210. Meng, L., Kwok, B. H., Sin, N., and Crews, C. M. (1999a). Eponemycin exerts its antitumor effect through the inhibition of proteasome function. Cancer Res. 59, 2798±2801. Meng, L. H., Mohan, R., Kwok, B. H. B., Elofsson, M., Sin, N., and Crews, C. M. (1999b). Epoxomicin, a potent and selective proteasome inhibitor, exhibits in vivo antiin¯ammatory activity. Proc. Natl. Acad. Sci. USA 96, 10403±10408. Meriin, A. B., Gabai, V. L., Yaglom, J., Shifrin, V. I., and Sherman, M. Y. (1998). Proteasome inhibitors activate stress kinases and induce hsp72: Diverse effects on apoptosis. J. Biol. Chem. 273, 6373±6379. Michalek, M. T., Grant, E. P., Gramm, C., Goldberg, A. L., and Rock, K. L. (1993). A role for the ubiquitin-dependent proteolytic pathway in MHC class I-restricted antigen presentation. Nature 363, 552±554. Michalek, M. T., Grant, E. P., and Rock, K. L. (1996). Chemical denaturation and modi®cation of ovalbumin alters its dependence on ubiquitin conjugation for class I antigen presentation. J. Immunol. 157, 617±624. Mimnaugh, E. G., Chen, H. Y., Davie, J. R., Celis, J. E., and Neckers, L. (1997). Rapid deubiquitination of nucleosomal histones in human tumor cells caused by proteasome inhibitors and stress response inducers: Effects on replication, transcription, translation, and the cellular stress response. Biochemistry 36, 14418 ±14429. Mo, A. X., van Lelyveld, S. F., Craiu, A., and Rock, K. L. (2000). Sequences that ¯ank subdominant and cryptic epitopes in¯uence the proteolytic generation of MHC class I-presented peptides. J. Immunol. 164, 4003± 4010. Mo, X. Y., Cascio, P., Lemerise, K., Goldberg, A. L., and Rock, K. (1999). Distinct proteolytic processes generate the C and N termini of MHC class I-binding peptides. J. Immunol. 163, 5851±5859. Morel, S., Levy, F., Burlet-Schiltz, O., Brasseur, F., Probst-Kepper, M., Peitrequin, A. L., Monsarrat, B., Van Velthoven, R., Cerottini, J. C., Boon, T., Gairin, J. E., and Van den Eynde, B. J. (2000). Processing of some antigens by the standard proteasome but not by the immunoproteasome results in poor presentation by dendritic cells. Immunity 12, 107±117.

generation of mhc class i-presented peptides

63

Morrison, L. A., Lukacher, A. E., Braciale, V. L., Fan, D. P., and Braciale, T. J. (1986). Difference in antigen presentation to MHC class I- and class II-restricted in¯uenza virus-speci®c cytolytic T lymphocyte clones. J. Exp. Med. 163, 903±921. Mosse, C. A., Meadows, L., Luckey, C. J., Kittlesen, D. J., Huczko, E. L., Slingluff, C. L., Shabanowitz, J., Hunt, D. F., and Engelhard, V. H. (1998). The class I antigen-processing pathway for the membrane protein tyrosinase involves translation in the endoplasmic reticulum and processing in the cytosol. J. Exp. Med. 187, 37±48. Murakami, Y., Matsufuji, S., Hayashi, S. I., Tanahashi, N., and Tanaka, K. (1999). ATP-dependent inactivation and sequestration of ornithine decarboxylase by the 26S proteasome are prerequisites for degradation. Mol. Cell. Biol. 19, 7216 ±7227. Murakami, Y., Matsufuji, S., Kameji, T., Hayashi, S., Igarashi, K., Tamura, T., Tanaka, K., and Ichihara, A. (1992). Ornithine decarboxylase is degraded by the 26S proteasome without ubiquitination. Nature 360, 597±599. Murata, S., Udono, H., Tanahashi, N., Hamada, N., Watanabe, K., Adachi, K., Yamano, T., Yui, K., Kobayashi, N., Kasahara, M., Tanaka, K., and T., C. (2001). Immunoproteasome assembly and antigen presentation in mice lacking both PA28alpha and PA28beta. EMBO J. 20, 5898±5907. Murray, A. (1995). Cyclin ubiquitination: The destructive end of mitosis. Cell 81, 149±152. Nakamura, T., Tanaka, T., Takagi, H., and Sato, M. (1998). Cloning and heterogeneous in vivo expression of tat binding protein-1 (TBP-1) in the mouse. Biochim. Biophys. Acta 1399, 93±100. Nazif, T., and Bogyo, M. (2001). Global analysis of proteasomal substrate speci®city using positional-scanning libraries of covalent inhibitors. Proc. Natl. Acad. Sci. USA 98, 2967±2972. Ng, D., Zwickl, P., and Goldberg, A. L. The activation by the pan±ATPase complex of protein degradation by archaebacterial 20S proteasomes. Submitted for publication. Niedermann, G., Butz, S., Ihlenfeldt, H. G., Grimm, R., Lucchiari, M., Hoschutzky, H., Jung, G., Maier, B., and Eichmann, K. (1995). Contribution of proteasome-mediated proteolysis to the hierarchy of epitopes presented by major histocompatibility complex class I molecules. Immunity 2, 289±299. Niedermann, G., Geier, E., Lucchiari-Hartz, M., Hitziger, N., Ramsperger, A., and Eichmann, K. (1999). The speci®city of proteasomes: Impact on MHC class I processing and presentation of antigens. Immunol. Rev. 172, 29± 48. Niedermann, G., King, G., Butz, S., Birsner, U., Grimm, R., Shabanowitz, J., Hunt, D. F., and Eichmann, K. (1996). The proteolytic fragments generated by vertebrate proteasomes: Structural relationships to major histocompatibility complex class I binding peptides. Proc. Natl. Acad. Sci. USA 93, 8572±8577. Nussbaum, A. K., Dick, T. P., Keilholz, W., Schirle, M., Stevanovic, S., Dietz, K., Heinemeyer, W., Groll, M., Wolf, D. H., Huber, R., Rammensee, H. G., and Schild, H. (1998). Cleavage motifs of the yeast 20S proteasome beta subunits deduced from digests of enolase 1. Proc. Natl. Acad. Sci. USA 95, 12504±12509. Nussbaum, A. K., Kuttler, C., Hadeler, K. P., Rammensee, H. G., and Schild, H. (2001). Paproc: A prediction algorithm for proteasomal cleavages available on the www. Immunogenetics 53, 87±94. O'Farrell, P. A., Gonzalez, F., Zheng, W. J., Johnston, S. A., and Joshua-Tor, L. (1999). Crystal structure of human bleomycin hydrolase, a self-compartmentalizing cysteine protease. Structure 7, 619±627. Oh, Y. K., Harding, C. V., and Swanson, J. A. (1997). The ef®ciency of antigen delivery from macrophage phagosomes into cytoplasm for MHC class I-restricted antigen presentation. Vaccine 15, 511±518.

64

kenneth l. rock et al.

Omura, S., Fujimoto, T., Otoguro, K., Matsuzaki, K., Moriguchi, R., Tanaka, H., and Sasaki, Y. (1991). Lactacystin, a novel microbial metabolite, induces neuritogenesis of neuroblastoma cells. J. Antibiot. 44, 113±116. Orlowski, M., Cardozo, C., Hidalgo, M. C., and Michaud, C. (1991). Regulation of the peptidylglutamyl-peptide hydrolyzing activity of the pituitary multicatalytic proteinase complex. Biochemistry 30, 5999±6005. Orlowski, M., Cardozo, C., and Michaud, C. (1993). Evidence for the presence of ®ve distinct proteolytic components in the pituitary multicatalytic proteinase complex: Properties of two components cleaving bonds on the carboxyl side of branched chain and small neutral amino acids. Biochemistry 32, 1563±1572. Ortiz-Navarete, V., Seelig, A., Gernold, M., and HaÈmmerling, G. J. (1991). Subunit of the 20S proteasome (multicatalytic proteinase) encoded by the major histocompatibility complex. Nature 353, 662±664. Osaka, F., Saeki, M., Katayama, S., Aida, N., Toh-e, A., Kominami, K., Toda, T., Suzuki, T., Chiba, T., Tanaka, K., and Kato, S. (2000). Covalent modi®er nedd8 is essential for scf ubiquitin±ligase in ®ssion yeast. EMBO J. 19, 3475±3484. Pagano, M., Tam, S. W., Theodoras, A. M., Beerromero, P., Delsal, G., Chau, V., Yew, P. R., Draetta, G. F., and Rolfe, M. (1995). Role of the ubiquitin±proteasome pathway in regulating abundance of the cyclin-dependent kinase inhibitor p27. Science 269, 682±685. Palmer, A., Rivett, A. J., Thomson, S., Hendil, K. B., Butcher, G. W., Fuertes, G., and Knecht, E. (1996). Subpopulations of proteasomes in rat liver nuclei, microsomes and cytosol. Biochem. J. 316, 401± 407. Pamer, E. G., Sijts, A. J., Villanueva, M. S., Busch, D. H., and Vijh, S. (1997). Mhc class I antigen processing of listeria monocytogenes proteins: Implications for dominant and subdominant CTL responses. Immunol. Rev. 158, 129±136. Paz, P., Brouwenstijn, N., Perry, R., and Shastri, N. (1999). Discrete proteolytic intermediates in the MHC class I antigen processing pathway and MHC i-dependent peptide trimming in the ER. Immunity 11, 241±251. Pickart, C. M. (1997). Targeting of substrates to the 26S proteasome. FASEB J. 11, 1055±1066. Plemper, R. K., and Wolf, D. H. (1999a). Endoplasmic reticulum degradation: Reverse protein transport and its end in the proteasome. Mol. Biol. Rep. 26, 125±130. Plemper, R. K., and Wolf, D. H. (1999b). Retrograde protein translocation: Eradication of secretory proteins in health and disease. Trends Biochem. Sci. 24, 266±270. Portaro, F. C., Gomes, M. D., Cabrera, A., Fernandes, B. L., Silva, C. L., Ferro, E. S., Juliano, L., and de Camargo, A. C. (1999). Thimet oligopeptidase and the stability of MHC class I epitopes in macrophage cytosol. Biochem. Biophys. Res. Commun. 255, 596±601. Preckel, T., Fung-Leung, W. P., Cai, Z., Vitiello, A., Salter-Cid, L., Winqvist, O., Wolfe, T. G., 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). Impaired immunoproteasome assembly and immune responses in PA28-/-mice. Science 286, 2162±2165. Princiotta, M. F., Schubert, U., Chen, W. S., Bennink, J. R., Myung, J., Crews, C. M., and Yewdell, J. W. (2001). Cells adapted to the proteasome inhibitor 4-hydroxy5-iodo-3-nitrophenylacetyl-leuleu-leucinal-vinyl sulfone require enzymatically active proteasomes for continued survival. Proc. Natl. Acad. Sci. USA 98, 513±518. Qu, D., Teckman, J. H., Omura, S., and Perlmutter, D. H. (1996). Degradation of a mutant secretory protein, alpha1-antitrypsin z, in the endoplasmic reticulum requires proteasome activity. J. Biol. Chem. 271, 22791±22795. Rammensee, H. G. (1995). Chemistry of peptides associated with MHC class I and class II molecules. Curr. Opin. Immunol. 7, 85±96.

generation of mhc class i-presented peptides

65

Rammensee, H. G., Friede, T., and Stevanoviic, S. (1995). MHC ligands and peptide motifs: First listing. Immunogenetics 41, 178±228. Realini, C., Dubiel, W., Pratt, G., Ferrell, K., and Rechsteiner, M. (1994). Molecular cloning and expression of a gamma- interferon-inducible activator of the multicatalytic protease. J. Biol. Chem. 269, 20727±20732. Realini, C., Jensen, C. C., Zhang, Z., Johnston, S. C., Knowlton, J. R., Hill, C. P., and Rechsteiner, M. (1997). Characterization of recombinant REGa, REGb, and REGg proteasome activators. J. Biol. Chem. 272, 25483±25492. Rechsteiner, M., Realini, C., and Ustrell, V. (2000). The proteasome activator 11S reg (PA28) and class I antigen presentation. Biochem. J. 345(1), 1±15. Reits, E. A. J., Benham, A. M., Plougastel, B., Neefjes, J., and Trowsdale, J. (1997). Dynamics of proteasome distribution in living cells. EMBO J. 16, 6087±6094. Reits, E. A., Vos, J. C., Gromme, M., and Neefjes, J. (2000). The major substrates for TAP in vivo are derived from newly synthesized proteins. Nature 404, 774 ±778. Ritz, U., and Seliger, B. (2001). The transporter associated with antigen processing (TAP): Structural integrity, expression, function, and its clinical relevance. Mol. Med. 7, 149 ±158. Rock, K. L., and Goldberg, A. L. (1999). Degradation of cell proteins and the generation of MHC class I-presented peptides. 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). Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class I molecules. Cell 78, 761±771. Roelse, J., Gromme, M., Momburg, F., Hammerling, G., and Neefjes, J. (1994). Trimming of TAPtranslocated peptides in the endoplasmic reticulum and in the cytosol during recycling. J. Exp. Med. 180, 1591±1597. Rohrwild, M., Coux, O., Huang, H. C., Moerschell, R. P., Yoo, S. J., Seol, J. H., Chung, C. H., and Goldberg, A. L. (1996). HSLV-HSLU: A novel ATP-dependent protease complex in Escherichia coli related to the eukaryotic proteasome. Proc. Natl. Acad. Sci. USA 93, 5808±5813. Rubin, D. M., Glickman, M. H., Larsen, C. N., Dhruvakumar, S., and Finley, D. (1998). Active site mutants in the six regulatory particle ATPases reveal multiple roles for ATP in the proteasome. EMBO J. 17, 4909± 4919. Rubin, D. M., van Nocker, S., Glickman, M., Coux, O., Wefes, I., Sadis, S., Fu, H., Goldberg, A., Vierstra, R., and Finley, D. (1997). ATPase and ubiquitin-binding proteins of the yeast proteasome. Mol. Biol. Rep. 24, 17±26. Salvesen, G. S., and Dixit, V. M. (1997). Caspases: Intracellular signaling by proteolysis. Cell 91, 443± 446. Saric, T., Beninga, J., Graef, C. I., Akopian, T. N., Rock, K. L., and Goldberg, A. L. (2001). Major histocompatibility complex class I-presented antigenic peptides are degraded in cytosolic extracts primarily by thimet oligopeptidase. J. Biol. Chem. 276, 36474±36481. Scheffner, M., Huibregtse, J. M., Vierstra, R. D., and Howley, P. M. (1993). The HPV-16 E6 and E6AP complex functions as a ubiquitin±protein ligase in the ubiquitination of p53. Cell 75, 495±505. Schlesinger, D. H., Goldstein, G., and Niall, H. D. (1975). The complete amino acid sequence of ubiquitin, an adenylate cyclase stimulating polypeptide probably universal in living cells. Biochemistry 14, 2214±2218. Schmidt, M., Lupas, A. N., and Finley, D. (1999). Structure and mechanism of ATP-dependent proteases. Curr. Opin. Chem. Biol. 3, 584±591. Schmidtke, G., Eggers, M., Ruppert, T., Groettrup, M., Koszinowski, U. H., and Kloetzel, P. M. (1998). Inactivation of a de®ned active site in the mouse 20S proteasome complex enhances major histocompatibility complex class I antigen presentation of a murine cytomegalovirus protein. J. Exp. Med. 187, 1641±1646.

66

kenneth l. rock et al.

Schmidtke, G., Emch, S., Groettrup, M., and Holzhutter, H.-G. Evidence for the existence of a noncatalytic modi®er site of peptide hydrolysis by the 20S proteasome. J. Biol. Chem. 275, 22056± 22063. Schmidtke, G., Kraft, R., Kostka, S., Henklein, P., Frommel, C., Lowe, J., Huber, R., Kloetzel, P. M., and Schmidt, M. (1996). Analysis of mammalian 20S proteasome biogenesisÐthe maturation of beta-subunits is an ordered two-step mechanism involving autocatalysis. EMBO J. 15, 6887±6898. Schubert, U., Anton, L. C., Gibbs, J., Norbury, C. C., Yewdell, J. W., and Bennink, J. R. (2000). Rapid degradation of a large fraction of newly synthesized proteins by proteasomes. Nature 404, 770±774. Schwarz, K., de Giuli, R., Schmidtke, G., Kostka, S., van den Broek, M., Kim, K. P., Crews, C. M., Kraft, R., and Groettrup, M. (2000a). The selective proteasome inhibitors lactacystin and epoxomicin can be used to either up- or down-regulate antigen presentation at nontoxic doses. J. Immunol. 164, 6147±6157. Schwarz, K., van Den Broek, M., Kostka, S., Kraft, R., Soza, A., Schmidtke, G., Kloetzel, P. M., and Groettrup, M. (2000b). Overexpression of the proteasome subunits LMP2, LMP7, and MECL-1, but not PA28 a/b, enhances the presentation of an immunodominant lymphocytic choriomeningitis virus T cell epitope. J. Immunol. 165, 768±778. Schwechheimer, C., Serino, G., Callis, J., Crosby, W. L., Lyapina, S., Deshaies, R. J., Gray, W. M., Estelle, M., and Deng, X.-W. (2001). Interactions of the COP9 signalosome with the E3 ubiquitin ligase scftir1 in mediating auxin response. Science 292, 1379±1382. Seeger, M., Ferrell, K., Frank, R., and Dubiel, W. (1997). HIV-1 tat inhibits the 20S proteasome and its 11S regulator-mediated activation. J. Biol. Chem. 272, 8145±8148. Seemuller, E., Lupas, A., and Baumeister, W. (1996). Autocatalytic processing of the 20S proteasome. Nature 382, 468± 471. Seemuller, E., Lupas, A., Stock, D., Lowe, J., Huber, R., and Baumeister, W. (1995a). Proteasome from thermoplasma acidophilumÐA threonine protease. Science 268, 579±582. Seemuller, E., Lupas, A., Stock, D., Lowe, J., Huber, R., and Baumeister, W. (1995b). Proteasome from thermoplasma acidophilum: A threonine protease. Science 268, 579±582. Seol, J. H., Feldman, R. M., Zachariae, W., Shevchenko, A., Correll, C. C., Lyapina, S., Chi, Y., Galova, M., Claypool, J., Sandmeyer, S., Nasmyth, K., and Deshaies, R. J. (1999). Cdc53/cullin and the essential Hrt1 RING-H2 subunit of SCF de®ne a ubiquitin ligase module that activates the E2 enzyme Cdc34. Genes & Dev. 13, 1614±1626. Sharipo, A., Imreh, M., Leonchiks, A., Imreh, S., and Masucci, M. G. (1998). A minimal glycine± alanine repeat prevents the interaction of ubiquitinated IkB alpha with the proteasome: A new mechanism for selective inhibition of proteolysis. Nat. Med. 4, 939±944. Shen, Z. H., Reznikoff, G., Dranoff, G., and Rock, K. L. (1997). Cloned dendritic cells can present exogenous antigens on both MHC class I and class II molecules. J. Immunol. 158, 2723±2730. Sherman, M. Y., and Goldberg, A. L. (2001). Cellular defenses against unfolded proteins: A cell biologist thinks about neurodegenerative diseases. Neuron 29, 15±32. Shimbara, N., Nakajima, H., Tanahashi, N., Ogawa, K., Niwa, S., Uenaka, A., Nakayama, E., and Tanaka, K. (1997). Double-cleavage production of the CTL epitope by proteasomes and PA28: Role of the ¯anking region. Genes Cells 2, 785±800. Shimbara, N., Ogawa, K., Hidaka, Y., Nakajima, H., Yamasaki, N., Niwa, S., Tanahashi, N., and Tanaka, K. (1998). Contribution of proline residue for ef®cient production of MHC class I ligands by proteasomes. J. Biol. Chem. 273, 23062±23071. Sibille, C., Gould, K. G., Willard-Gallo, K., Thomson, S., Rivett, A. J., Powis, S., Butcher, G. W., and De Baetselier, P. (1995). LMP2‡ proteasomes are required for the presentation of speci®c antigens to cytotoxic T lymphocytes. Curr. Biol. 5, 923±930. Sigal, L. J., and Rock, K. L. (2000). Bone marrow-derived antigen-presenting cells are required for the generation of cytotoxic T lymphocyte responses to viruses and use transporter associated with

generation of mhc class i-presented peptides

67

antigen presentation (TAP)-dependent and -independent pathways of antigen presentation. J. Exp. Med. 192, 1143±1150. Sijts, A. J., Ruppert, T., Rehermann, B., Schmidt, M., Koszinowski, U., and Kloetzel, P. M. (2000a). Ef®cient generation of a hepatitis B virus cytotoxic T lymphocyte epitope requires the structural features of immunoproteasomes. J. Exp. Med. 191, 503±514. Sijts, A. J., Standera, S., Toes, R. E., Ruppert, T., Beekman, N. J., van Veelen, P. A., Ossendorp, F. A., Melief, C. J., and Kloetzel, P. M. (2000b). MHC class I antigen processing of an adenovirus CTL epitope is linked to the levels of immunoproteasomes in infected cells. J. Immunol. 164, 4500±4506. Silva, C. L., Portaro, F. C., Bonato, V. L., de Camargo, A. C., and Ferro, E. S. (1999). Thimet oligopeptidase (EC 3.4.24.15), a novel protein on the route of MHC class I antigen presentation. Biochem. Biophys. Res. Commun. 255, 591±595. Solheim, J. C. (1999). Class I MHC molecules: Assembly and antigen presentation. Immunol. Rev. 172, 11±19. Song, R., and Harding, C. V. (1996). Roles of proteasomes, transporter for antigen presentation (TAP), and b2 -microglobulin in the processing of bacterial or particulate antigens via an alternate class I MHC processing pathway. J. Immunol. 156, 4182± 4190. Sorimachi, H., Ishiura, S., and Suzuki, K. (1997). Structure and physiological function of calpains. Biochem. J. 328, 721±732. Stennicke, H. R., and Salvesen, G. S. (1998). Properties of the caspases. Biochim. Biophys. Acta 1387, 17±31. Stohwasser, R., Kuckelkorn, U., Kraft, R., Kostka, S., and Kloetzel, P. M. (1996). 20S proteasome from LMP7 knock out mice reveals altered proteolytic activities and cleavage site preferences. FEBS Lett. 383, 109±113. Stohwasser, R., Salzmann, U., Giesebrecht, J., Kloetzel, P. M., and Holzhutter, H. G. (2000a). Kinetic evidences for facilitation of peptide channelling by the proteasome activator PA28. Eur. J. Biochem. 267, 6221±6230. Stohwasser, R., Soza, A., Eggers, M., Koszinowski, U. H., and Kloetzel, P. M. (2000b). Pa28ab double and PA28b single transfectant mouse B8 cell lines reveal enhanced presentation of a mouse cytomegalovirus (MCMV) pp89 MHC class I epitope. Mol. Immunol. 37, 13±19. Stoltze, L., Dick, T. P., Deeg, M., Pommerl, B., Rammensee, H. G., and Schild, H. (1998). Generation of the vesicular stomatitis virus nucleoprotein cytotoxic T lymphocyte epitope requires proteasome-dependent and -independent proteolytic activities. Eur. J. Immunol. 28, 4029± 4036. Stoltze, L., Schirle, M., Schwarz, G., Schroter, C., Thompson, M. W., Hersh, L., Kalbacher, H., Stevanovich, S., Rammensee, H. G., and Schild, H. (2000). Two new proteases in the MHC class I processing pathway. Nat. Immunol. 1, 413± 418. Sutterluty, H., Chatelain, E., Marti, A., Wirbelauer, C., Senften, M., Muller, U., and Krek, W. (1999). P45(skp2) promotes p27(kip1) degradation and induces S phase in quiescent cells. Nat. Cell Biol. 1, 207±214. Svensson, K., Levy, F., Sundberg, U., Boman, H. G., Hendil, K. B., and Kvist, S. (1996). Proteasomes generate in vitro a natural peptide of in¯uenza-A nucleoprotein functional in HAL-B27 antigen assembly. Int. Immunol. 8, 467± 478. Tamura, N., Lottspeich, F., Baumeister, W., and Tamura, T. (1998). The role of tricorn protease and its aminopeptidase-interacting factors in cellular protein degradation. Cell 95, 637±648. Tamura, T., Tamura, N., Cejka, Z., Hegerl, R., Lottspeich, F., and Baumeister, W. (1996). Tricorn proteaseÐThe core of a modular proteolytic system. Science 274, 1385±1389. Tanahashi, N., Murakami, Y., Minami, Y., Shimbara, N., Hendil, K. B., and Tanaka, K. (2000). Hybrid proteasomes: Induction by interferon-g and contribution to ATP-dependent proteolysis. J. Biol. Chem. 275, 14336±14345.

68

kenneth l. rock et al.

Tanahashi, N., Yokota, K., Ahn, J. Y., Chung, C. H., Fujiwara, T., Takahashi, E., DeMartino, G. N., Slaughter, C. A., Toyonaga, T., Yamamura, K., Shimbara, N., and Tanaka, K. (1997). Molecular properties of the proteasome activator PA28 family proteins and g-interferon regulation. Genes Cells 2, 195±211. Tanaka, K., Kawakami, T., Tateishi, K., Yashiroda, H., and Chiba, T. (2001). Control of i kappa b alpha proteolysis by the ubiquitin±proteasome pathway. Biochimie 83, 351±356. Taylor, A. (1993). Aminopeptidases: Towards a mechanism of action. Trends Biochem. Sci. 18, 167±171. Teicher, B. A., Ara, G., Herbst, R., Palombella, V. J., and Adams, J. (1999). The proteasome inhibitor PS-341 in cancer therapy. Clin. Cancer Res. 5, 2638±2645. Theobald, M., Ruppert, T., Kuckelkorn, U., Hernandez, J., Haussler, A., Ferreira, E. A., Liewer, U., Biggs, J., Levine, A. J., Huber, C., Koszinowski, U. H., Kloetzel, P. M., and Sherman, L. A. (1998). The sequence alteration associated with a mutational hotspot in p53 protects cells from lysis by cytotoxic T lymphocytes speci®c for a ¯anking peptide epitope. J. Exp. Med. 188, 1017±1028. Therrien, C., Lachance, P., Sulea, T., Purisima, E. O., Qi, H. T., Ziomek, E., Alvarez-Hernandez, A., Roush, W. R., and Menard, R. (2001). Cathepsins x and b can be differentiated through their respective mono- and dipeptidyl carboxypeptidase activities. Biochemistry 40, 2702±2711. Tjoa, B. A., and Kranz, D. M. (1994). Generation of cytotoxic T-lymphocytes to a self-peptide/class I complex: A model for peptide-mediated tumor rejection. Cancer Res. 54, 204±208. Tobery, T., and Siliciano, R. F. (1999). Cutting edge: Induction of enhanced CTL-dependent protective immunity in vivo by n-end rule targeting of a model tumor antigen. J. Immunol. 162, 639±642. Tobery, T. W., and Siliciano, R. F. (1997). Targeting of hiv-1 antigens for rapid intracellular degradation enhances cytotoxic T lymphocyte (CTL) recognition and the induction of de novo CTL responses in vivo after immunization. J. Exp. Med. 185, 909±920. Tokunaga, F., Goto, T., Koide, T., Murakami, Y., Hayashi, S., Tamura, T., Tanaka, K., and Ichihara, A. (1994). Atp- and antizyme-dependent endoproteolysis of ornithine decarboxylase to oligopeptides by the 26S proteasome. J. Biol. Chem. 269, 17382±17385. Tomkinson, B. (1999). Tripeptidyl peptidases: Enzymes that count. Trends Biochem. Sci. 24, 355±359. Tongaonkar, P., Chen, L., Lambertson, D., Ko, B., and Madura, K. (2000). Evidence for an interaction between ubiquitin-conjugating enzymes and the 26S proteasome. Mol. Cell. Biol. 20, 4691±4698. Townsley, F. M., Aristarkhov, A., Beck, S., Hershko, A., and Ruderman, J. V. (1997). Dominantnegative cyclin-selective ubiquitin carrier protein E2-C/UbcH10 blocks cells in metaphase. Proc. Natl. Acad. Sci. USA 94, 2362±2367. Treier, M., Staszewski, L. M., and Bohmann, D. (1994). Ubiquitin-dependent c-jun degradation in vivo is mediated by the delta domain. Cell 78, 787±798. Tsao, P., and von Zastrow, M. (2000). Downregulation of g protein-coupled receptors. Curr. Opin. Neurobiol. 10, 365±369. Tsubuki, S., Saito, Y., Tomioka, M., Ito, H., and Kawashima, S. (1996). Differential inhibition of calpain and proteasome activities by peptidyl aldehydes of di-leucine and tri-leucine. J. Biochem. 119, 572±576. Turk, B., Turk, D., and Turk, V. (2000). Lysosomal cysteine proteases: More than scavengers. BiochemÐ. Biophys. ActaÐProtein Struct. Mol. Enzymol. 1477, 98±111. Turnell, A. S., Grand, R. J., Gorbea, C., Zhang, X., Wang, W., Mymryk, J. S., and Gallimore, P. H. (2000). Regulation of the 26S proteasome by adenovirus E1A. EMBO J. 19, 4759±4773. Van Antwerp, D. J., Martin, S. J., Verma, I. M., and Green, D. R. (1998). Inhibition of TNF-induced apoptosis by NFkB. Trends Cell Biol. 8, 107±111.

generation of mhc class i-presented peptides

69

Van den Eynde, B. J., and Morel, S. (2001). Differential processing of class-I-restricted epitopes by the standard proteasome and the immunoproteasome. Curr. Opin. Immunol. 13, 147±153. van Hall, T., Sijts, A., Camps, M., Offringa, R., Melief, C., Kloetzel, P. M., and Ossendorp, F. (2000). Differential in¯uence on cytotoxic T lymphocyte epitope presentation by controlled expression of either proteasome immunosubunits or PA28. J. Exp. Med. 192, 483± 494. Van Kaer, L., Ashton-Rickardt, P. G., Eichelberger, M., Gaczynska, M., Nagashima, K., Rock, K. L., Goldberg, A. L., Doherty, P. C., and Tonegawa, S. (1994). Altered peptidase and viral-speci®c T cell response in LMP2 mutant mice. Immunity 1, 533±541. Verma, R., Chen, S., Feldman, R., Schieltz, D., Yates, J., Dohmen, J., and Deshaies, R. J. (2000). Proteasomal proteomics: identi®cation of nucleotide-sensitive proteasome-interacting proteins by mass spectrometric analysis of af®nity-puri®ed proteasomes. Mol. Biol. Cell 11, 3425±3439. Villadangos, J. A., Bryant, R. A. R., Deussing, J., Driessen, C., Lennon-Dumenil, A. M., Riese, R. J., Roth, W., Saftig, P., Shi, G. P., Chapman, H. A., Peters, C., and Ploegh, H. L. (1999). Proteases involved in MHC class II antigen presentation. Immunol. Rev. 172, 109±120. Vinitsky, A., Anton, L. C., Snyder, H. L., Orlowski, M., Bennink, J. R., and Yewdell, J. W. (1997). The generation of MHC class I associated peptides is only partially inhibited by proteasome inhibitorsÐInvolvement of nonproteasomal cytosolic proteases in antigen processing. J. Immunol. 159, 554±564. Vinitsky, A., Cardozo, C., Sepp-Lorenzino, L., Michaud, C., and Orlowski, M. (1994). Inhibition of the proteolytic activity of the multicatalytic proteinase complex (proteasome) by substrate-related peptidyl aldehydes. J. Biol. Chem. 269, 29860±29866. Vinitsky, A., Michaud, C., Powers, J. C., and Orlowski, M. (1992). Inhibition of the chymotrypsinlike activity of the pituitary multicatalytic proteinase complex. Biochemistry 31, 9421±9428. Voges, D., Zwickl, P., and Baumeister, W. (1999). The 26S proteasome: A molecular machine designed for controlled proteolysis. Annu. Rev. Biochem. 68, 1015±1068. Wang, E. W., Kessler, B. M., Borodovsky, A., Cravatt, B. F., Bogyo, M., Ploegh, H. L., and Glas, R. (2000). Integration of the ubiquitin±proteasome pathway with a cytosolic oligopeptidase activity. Proc. Natl. Acad. Sci. USA 97, 9990±9995. Wang, K. K. W., and Yuen, P. W. (1994). Calpain inhibitionÐAn overview of its therapeutic potential. Trends Pharmacol. Sci. 15, 412± 419. Wang, R., Chait, B. T., Wolf, I., Kohanski, R. A., and Cardozo, C. (1999). Lysozyme degradation by the bovine multicatalytic proteinase complex (proteasome): Evidence for a nonprocessive mode of degradation. Biochemistry 38, 14573±14581. Waterman, H., and Yarden, Y. (2001). Molecular mechanisms underlying endocytosis and sorting of erbb receptor tyrosine kinases. FEBS Lett. 490, 142±152. Wei, M. L., and Cresswell, P. (1992). HLA-A2 molecules in an antigen-processing mutant cell contain signal sequence-derived peptides. Nature 356, 443± 446. Wei, N., and Deng, X. W. (1999). Making sense of the COP9 signalosome: A regulatory protein complex conserved from arabidopsis to human. Trends Genet. 15, 98±103. Wei, N., Tsuge, T., Serino, G., Dohmae, N., Takio, K., Matsui, M., and Deng, X. W. (1998). The COP9 complex is conserved between plants and mammals and is related to the 26S proteasome regulatory complex. Curr. Biol. 8, 919±922. Weissman, A. M. (2001). Themes and variations on ubiquitylation. Nat. Rev. Mol. Cell Biol. 2, 169±178. Wenzel, T., Eckerskorn, C., Lottspeich, F., and Baumeister, W. (1994). Existence of a molecular ruler in proteasomes suggested by analysis of degradation products. 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). Structural basis for the activation of 20S proteasomes by 11S regulators. Nature 408, 115±120.

70

kenneth l. rock et al.

Wiertz, E. J., Tortorella, D., Bogyo, M., Yu, J., Mothes, W., Jones, T. R., Rapoport, T. A., and Ploegh, H. L. (1996a). Sec61-mediated transfer of a membrane protein from the endoplasmic reticulum to the proteasome for destruction. Nature 384, 432± 438. Wiertz, E. J. H. J., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J., and Ploegh, H. L. (1996b). The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84, 769±779. Wilk, S., and Orlowski, M. (1983). Evidence that pituitary cation-sensitive neutral endopeptidase is a multicatalytic protease complex. J. Neurochem. 40, 842±849. Wilkinson, K. D., Urban, M. K., and Haas, A. L. (1980). Ubiquitin is the ATP-dependent proteolysis factor I of rabbit reticulocytes. J. Biol. Chem. 255, 7529±7532. Wu, R. S., Kohn, K. W., and Bonner, W. M. (1981). Metabolism of ubiquitinated histones. J. Biol. Chem. 256, 5916±5920. Xiong, X., Chong, E., and Skach, W. R. (1999). Evidence that endoplasmic reticulum (ER)associated degradation of cystic ®brosis transmembrane conductance regulator is linked to retrograde translocation from the ER membrane. J. Biol. Chem. 274, 2616±2624. Yamazaki, H., Tanaka, M., Nagoya, M., Fujimaki, H., Sato, K., Yago, T., Nagata, T., and Minami, M. (1997). Epitope selection in major histocompatibility complex class I-mediated pathway is affected by the intracellular localization of an antigen. Eur. J. Immunol. 27, 347±353. 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 proteasome-dependent removal from ER membranes. J. Exp. Med. 187, 835±846. Yee, S., Okada, Y., Ogasawara, K., Omura, S., Takatsuki, A., Kakiuchi, T., Muno, D., Kominami, E., and Mizuochi, T. (1997). MHC class I presentation of an exogenous polypeptide antigen encoded by the murine aids defective virus. Microbiol. Immunol. 41, 563±570. Yewdell, J. T., and Bennink, J. R. (1999). Immunodominance in major histocompatibility complex class I-restricted T lymphocyte responses. Annu. Rev. Immunol. 17, 51±88. Yewdell, J. W., Aton, L. C., and Benink, J. R. (1996). Defective ribosomal products (DRiPs): A major source of antigenic peptides for MHC class I molecules? J. Immunol. 157, 1823±1826. Zachariae, W., and Nasmyth, K. (1999). Whose end is destruction: Cell division and the anaphasepromoting complex. Genes Dev. 13, 2039±2058. Zhang, Z., Torii, N., Furusaka, A., Malayaman, N., Hu, Z., and Liang, T. J. (2000). Structural and functional characterization of interaction between hepatitis B virus X protein and the proteasome complex. J. Biol. Chem. 275, 15157±15165. Zimmermann, J., Erdmann, D., Lalande, I., Grossenbacher, R., Noorani, M., and Furst, P. (2000). Proteasome inhibitor induced gene expression pro®les reveal overexpression of transcriptional regulators atf3, gadd153 and mad1. Oncogene 19, 2913±2920. Zweerink, H. J., Gammon, M. C., Utz, U., Sauma, S. Y., Harrer, T., Hawkins, J. C., Johnson, R. P., Sirotina, A., Hermes, J. D., Walker, B. D., and Biddison, W. E. (1993). Presentation of endogenous peptides to MHC class I-restricted cytotoxic T lymphocytes in transport deletion mutant T2 cells. J. Immunol. 150, 1763 ±1771. Zwickl, P., Grziwa, A., Puhler, G., Dahlmann, B., Lottspeich, F., and Baumeister, W. (1992). Primary structure of the Thermoplasma proteasome and its implications for the structure, function, and evolution of the multicatalytic proteinase. Biochemistry 31, 964 ±972. Zwickl, P., Ng, D., Woo, K. M., Klenk, H. P., and Goldberg, A. L. (1999). An archaebacterial ATPase, homologous to ATPases in the eukaryotic 26S proteasome, activates protein breakdown by 20S proteasomes. J. Biol. Chem. 274, 26008±26014.

ADVANCES IN IMMUNOLOGY, VOL. 80

Proteolysis and Antigen Presentation by MHC Class II Molecules PAULA WOLF BRYANT,* ANA-MARIA LENNON-DUME¨NIL,À EDDA FIEBIGER,À CE¨CILE LAGAUDRIE©RE-GESBERT,À AND HIDDE L. PLOEGHÀ *Department of Microbiology,The Ohio State University, À

Columbus,Ohio 43210; and Department of Pathology, Harvard Medical School, Boston, Mossachusetts 02115

I. Introduction Proteolysis is the primary mechanism used by all cells not only to dispose of unwanted proteins but also to regulate protein function and maintain cellular homeostasis. Proteases that reside in the endocytic pathway are the principal actors of terminal protein degradation. Consequently, the number of endocytic proteases must be immense, with wide tissue distributions coupled to broad and overlapping enzymatic speci®city. As inferred from their intracellular location, these enzymes usually function at low pH (Chapman et al., 1997). The assignment of speci®c functions to the individual endocytic proteases was once thought unrealistic due to their apparent redundancy in higher organisms. However, recent technological advances, such as the generation of knockout mice and active site-directed probes, have led to a far better characterization of intracellular proteases with more unique features (Chapman et al., 1997; McGrath, 1999; Turk et al., 2000). Indeed, proteases are detected all along the endosomal/lysosomal route, and, accordingly, the optimal pH range for some is rather broad. In addition, the expression of certain endocytic proteases is regulated either at the level of gene transcription or enzyme maturation, and their activity is controlled by the presence of endogenous protease inhibitors. Some endocytic proteases exhibit tissue-speci®c expression and enzymes with distinct, rather than broad, substrate speci®cities have been identi®ed. Hence, some of the ``housekeeping'' proteases are now recognized as having precise biological functions rather than merely carrying out bulk proteolysis (Chapman et al., 1997; McGrath, 1999; Turk et al., 2000). The presentation of peptide antigens by MHC class II molecules is strictly dependent on the action of proteases (Nakagawa and Rudensky, 1999; Riese and Chapman, 2000; Villadangos and Ploegh, 2000; Watts, 2001). Class II molecules scour the endocytic pathway for antigenic peptides to bind and present at the cell surface for recognition by CD4‡ T cells (Watts, 1997; Wolf and Ploegh, 1995). The specialized cell types that support antigen presentation by class II molecules are commonly referred to as professional antigen± presenting cells (APCs), which include bone marrow-derived B lymphocytes, 71 Copyright 2002, Elsevier Science (USA). All rights reserved. 0065-2776/02 $35.00

72

paula wolf bryant et al.

dendritic cells (DCs), and macrophages. In addition, class II molecules are expressed in ``nonprofessional'' APCs either constitutively (i.e., thymic cortical and medullary epithelial cells) or upon induction with interferon-g (IFN-g) (i.e., endothelial cells, intestinal epithelial cells, mast cells, and ®broblasts). Before MHC class II molecules meet their intended T-cell receptor (TCR) at the cell surface, a complex series of biosynthetic and proteolytic events ensures correct assembly, traf®cking, and peptide loading of the class II molecule (Watts, 1997; Wolf and Ploegh, 1995). Accordingly, the endocytic proteases required for the function of class II molecules are found in the distinct APC types and, in some cases, are themselves IFN-g inducible (Chapman, 1998). MHC class II products consist of an a chain and a b chain, which dimerize shortly after synthesis in the endoplasmic reticulum (ER) to create a cleft in which antigenic peptides are captured (Watts, 1997; Wolf and Ploegh, 1995) (Fig. 1A). Newly synthesized ab-dimers assemble onto a scaffold of a homotrimeric, type II membrane protein, the invariant chain (Ii). There are two isoforms of Ii in mice, p31 and p41, distinguished by alternative splicing of Ii transcripts (Koch, 1988; O'Sullivan et al., 1987; Strubin et al., 1986). The resulting p41 isoform contains an extra exon (exon 6b), C-terminal to Ii's trimerization domain. As we shall see below, the additional domain of p41 plays a key role in the control of protease expression and activity in APCs. In humans, these two splice variants each yield two protein products, resulting in four Ii isoforms: p31(p33), p33(p35), p41, and p43 (Cresswell, 1996). Ii prevents premature association of the class II heterodimer with peptides or unfolded segments of polypeptide chains in the ER by means of a short segment of its lumenal region, designated CLIP (Class II-associated Ii derived Peptides), which ®lls the peptide-binding groove of class II (Ghosh et al., 1995). Interaction with Ii also facilitates folding of the ab-heterodimer. The cytoplasmic tail of Ii contains targeting signals that deliver the majority of (ab±Ii)3 complexes from the trans-Golgi network (TGN) directly into the endocytic pathway (Bakke and Dobberstein, 1990; Benaroch et al., 1995; Lotteau et al., 1990). Only 5±10% of newly synthesized class II±Ii complexes fail to be sorted and are instead delivered to the plasma membrane, followed by rapid internalization into endocytic vesicles (Castellino and Germain, 1995; Pinet and Long, 1998; Roche et al., 1993; Zhong et al., 1997). Thus, all elements of the endocytic pathway receive ab±Ii. Likewise, the same endocytic compartments that contain class II±Ii complexes at different stages of maturation compose much of the pathway traveled by internalized antigens (Watts, 1997). Antigens that enter the endocytic pathway are processed into class II-presentable peptides by resident proteases (Chapman, 1998; Harding, 1996) (Fig. 1A). Successful MHC class II-restricted antigen presentation not only depends on the availability of processed peptides, but also on the rate of Ii breakdown and removal. Ii must be proteolytically destroyed to free the class II peptide-

antigen presentation by mhc class ii molecules

73

binding groove for loading of antigenic peptides. Upon arrival of ab±Ii complexes in endocytic compartments, Ii is degraded in a stage-speci®c manner by a combination of aspartyl and cysteine proteases (Amigorena et al., 1995; Maric et al., 1994; Villadangos et al., 1997) (Figs. 1A and 1B). As class II molecules themselves are poor substrates for these proteases, the peptide-binding cleft affords protection against proteolysis of its contents. To complete peptide binding, most class II alleles require interaction with the nonclassical class II dimer, H-2M/HLA-DM (referred to hereafter as DM). DM induces release of the Ii remnant CLIP and facilitates loading with antigenic peptides (Alfonso and Karlsson, 2000; Busch et al., 2000; Kropshofer et al., 1999). Those peptides that can no longer be dislodged by DM generate the pool of peptide-loaded class II molecules at the cell surface (Fig. 1A). Protease attack of Ii yields class II±Ii intermediates of distinct conformations required for proper traf®cking, interaction with the accessory molecule DM, and acquisition of peptides (Fig. 1B). Ii proteolysis is therefore not ``haphazard'' but instead follows a wellde®ned path. The constancy of the Ii intermediates produced by proteolysis has facilitated the identi®cation of speci®c proteases capable of performing the necessary cleavages to yield an ab-dimer free to bind peptides. Cathepsins S (Nakagawa et al., 1999; Riese et al., 1996; Shi et al., 1999), L (Nakagawa et al., 1998), and F (Shi et al., 2000) were all shown capable of performing a critical cleavage late in the staged destruction of Ii (see Section III). On the other hand, assigning speci®c functions to individual proteases in the generation of class IIrestricted epitopes has been more challenging (see Section VI). Rather than relying on the function of a single protease, the generation of peptides from a given antigen may require the combination of several proteases with distinct and overlapping speci®cities (Villadangos and Ploegh, 2000). Nonetheless, it is likely that the processing of both antigens and Ii are executed by similar, if not the same, endocytic proteases. Thus, the antigen presentation pathway of class II molecules has evolved to utilize the cell's proteolytic ``housekeeping'' machinery for its own success. Here we review those proteases required for Ii breakdown and removal from the class II ab-dimer and those likely to be important in processing antigens into class II-presentable peptides. II. Protease Activity in Antigen-Presenting Cells A. Endocytic Proteases The proteases contained in the endocytic pathway are classi®ed into four major groups based on the active-site amino acid used by the enzyme to hydrolyze amide bonds of proteins: cysteine, aspartyl, serine, and metalloproteases. More than 16 de®ned intracellular acidic proteasesÐreferred to as the cathepsinsÐspan these four groups. By virtue of their ``profession,'' APCs

A DM

endocytic pathway GILT AEP, Cat B,S,E

Ag

endocytic pathway

TGN

Aspartic

golgi

(abIi)3

ER

ER DM

ER

a +b ab-dimer MHC class II

(Ii)3

(abIi)3

Cathepsins

ab Iip22

to cell surface

? Cat S,L,F

Cysteine ab Iip10

ab CLIP

ab peptide

antigen presentation by mhc class ii molecules

75

Fig. 1(A). The generation of peptide-loaded MHC class II molecules in the endocytic route. MHC class II ab-dimers are synthesized in the endoplasmic reticulum (ER) where they quickly associate with homotrimers of Ii to form nonameric complexes (ab±Ii)3 . The CLIP region of Ii lies within the peptide-binding groove of the ab-dimer. The nonameric complexes exit the ER and traverse the Golgi stacks. Once in the TGN, the majority of (ab±Ii)3 complexes are delivered directly to compartments in the endocytic pathway by sorting signals in the cytoplasmic tail of Ii. A few of these complexes escape to the cell where they are then quickly internalized. In the endocytic pathway, Ii is removed in a stepwise fashion by a combination of aspartic and cysteine proteases until the CLIP region of Ii is all that remains in the class II peptide-binding groove (ab± CLIP). The accessory molecule, DM, interacts with ab±CLIP (black arrows)Ðor some intermediate between ab±Iip10 and ab±CLIP (gray arrows)Ðinducing the release of these Ii remnants. The continued interaction of DM with the now empty class II molecule preserves its structure and facilitates loading of antigenic peptides. The peptides loaded onto class II are generated from endocytosed antigens. These antigens must ®rst be partially unfolded, possibly by the action of the thiol reductase GILT, to render the protein accessible to proteases. To date, asparaginyl endopeptidase (AEP), Cat B, Cat S, and Cat E are the only proteases shown to be important in vivo for the generation of some class II presentable peptides. The resulting ab±peptide complexes are then deposited on the cell surface for recognition by CD4‡ T cells. (B) Schematic view of Ii and its breakdown intermediates. Both the p31 and p41 isoforms of Ii are borken down in an indistinguishable manner. The nonameric class II±Ii complex (ab±Ii)3 is disrupted by COOH-terminal degradation of Ii by aspartyl proteases to yield ab-dimers attached to an Ii degradation intermediate of 22 residues (Iip22). All subsequent COOH-terminal processing is performed by cysteine proteases, yielding the ab-associated intermediates Iip18 and Iip10. The speci®c proteases absolutely required for these processing steps are not known. An NH2 -terminal cleavage converts Iip10 into CLIP. This cleavage is performed by Cat S in B cells, DCs, and macrophages and by Cat L in cTECs. Cat F can digest Iip10 into CLIP in vitro and may function as a ``backup'' enzyme in macrophages that lack Cat S.

76

paula wolf bryant et al.

harbor many of these enzymes. Cathepsins B, C, D, E, F, H, K (Punturieri et al., 2000), L, O, S, V, and Z (X) have all been identi®ed in various professional APC types, and this list is certainly not complete (Riese and Chapman, 2000; Turk et al., 2000; Villadangos and Ploegh, 2000). Of the enzymes identi®ed in professional APCs, cathepsins D and E are aspartyl proteases, while the remaining cathepsins (Cat) are cysteine proteases and belong to the papain family of enzymes (McGrath, 1999). Based on chromosomal location and sequence homology, these cysteine proteases are subdivided into Cat B-like and Cat Llike subgroups (Cygler and Mort, 1997; McGrath, 1999). This classi®cation, however, does not appear to be an accurate means of predicting which cathepsins could play a role in Ii breakdown and/or antigen processing. For instance, cathepsins S, L, and F were all shown to be capable of performing similar Ii cleavages (see below). Although both Cat S and Cat L belong to the same L-like subgroup, Cat F is distinct (Wex et al., 2000). In addition to the cathepsins, a member of the legumain family of cysteine proteases, asparaginyl endopeptidase (AEP), is also present in APCs (i.e., B cells) (Manoury et al., 1998) and may well be one of the more speci®c enzymes in terms of preferred cleavage sites. The most abundant cysteine and aspartyl proteases expressed in APCs are cathepsins B and D, respectively (Chapman et al., 1997; McGrath, 1999). The structures of cathepsins B (Musil et al., 1991), L (Fujishima et al., 1997; Guncar et al., 1999), K (McGrath et al., 1997; Zhao et al., 1997), H (Guncar et al., 1998), and S (McGrath et al., 1998) have been determined, as well as the structures of the procathepsins B (Podobnik et al., 1997), L (Coulombe et al., 1996), and K (McGrath, 1999). The mature cathepsins range between 20 and 30 kDa (the proenzymes are 60±100 amino acids larger) and are composed of two equivalent sized domains, stabilized by the presence of disul®de bonds. The two domains are separated by the catalytic center that contains the active site nucleophile used to cleave bound substrates. The speci®city of an enzyme is determined by the architecture of its active site. Cathepsins K, L, S, and F are endopeptidases: they hydrolyze internal amide bonds of proteins (Chapman et al., 1997; McGrath, 1999; Turk et al., 1997). Catalysis by endoproteases relies on a triad composed of a cysteine, a histidine, and an asparagine residue. The histidine and asparagine residues polarize the cysteine residue to generate the nucleophile that attacks the carbonyl carbon of the targeted amide bond of the substrate (Chapman et al., 1997; McGrath, 1999; Turk et al., 1997). Enzymatic activity is associated with an open active site approximately 15 AÊ in length to which substrates bind (McGrath, 1999; Turk et al., 1997; Turk et al., 1998). In contrast, the amino-exopeptidase activities of Cat H and Cat C, and the carboxylexopeptidase activities of Cat B and Cat Z, are generated by obstructing part of the active site with side chains that stabilize either the N terminus or the C terminus of the bound substrate (McGrath, 1999; Turk et al., 2000). The result is hydrolysis of one or two amino acids from either peptide termini. At lower pH,

antigen presentation by mhc class ii molecules

77

the occluding loop of Cat B is ¯exible, allowing this enzyme to also function as an endopeptidase (Illy et al., 1997; Nagler et al., 1997). A limited endoproteolytic activity has also been described for Cat H (Illy et al., 1997; McGrath, 1999). However, the major endoproteases present in the endocytic pathway are the aspartic protease Cat D, and the cysteine proteases Cat S and Cat L. With the exception of AEP, the lysosomal proteases show a rather relaxed cleavage speci®city, which makes the prediction of cleavage sites in antigens dif®cult. B. Biosynthesis of the Active Enzyme Lysosomal proteases are synthesized in the ER as inactive zymogens or proenzymes (McGrath, 1999; Turk et al., 2000). The propiece can vary in size and is located at the N terminus of the mature enzyme. The role of the propiece in the maturation of the precursor enzyme into an active protease resembles that of the chaperone Ii in MHC class II maturation. In the ER, the inactive protease requires its propiece for ef®cient folding, just as the assembly and folding of the class II ab-dimer are facilitated by Ii (Tao et al., 1994; Vernet et al., 1995). Furthermore just as the CLIP region of Ii lies in the peptide-binding cleft of the class II ab-dimer, the propeptide, if of appropriate length, occupies the active site of the enzyme. Occupation of the active site prevents premature activation of the enzyme as it travels through the Golgi on its way to the endocytic pathway (Coulombe et al., 1996; LaLonde et al., 1999; Turk et al., 1996b). The enzymatic functions of the protease and the peptide-binding function of the class II ab-dimer require that the active site of each be liberated of its occupants (the propiece and CLIP). Both of these events occur in the endocytic pathway. The cathepsins exit the ER and travel through the Golgi, where they acquire post-translational modi®cations including glycosylation and phosphorylation, each in¯uenced by the presence of the propiece (Gieselmann et al., 1983; Lingeman et al., 1998). These modi®cations are required for proper sorting of proteases in the TGN. Phosphomannosyl residues of the protease bind to two types of mannose 6-phosphate receptors that shuttle between the Golgi and endocytic compartments (Kasper et al., 1996; Koster et al., 1993; Pohlmann et al., 1995). Once deposited in an endocytic compartment, the propiece is removed spontaneously by an autocatalytic mechanism, or by the action of another protease, to yield the mature, active form of the enzyme (one chain or two chain depending on the enzyme; LaLonde et al., 1999; Riese and Chapman, 2000). Thus, both the cathepsins and MHC class II molecules are rendered functional in the endocytic pathway through proteolytic events. C. Regulation of Protease Activity 1. Intracellular pH The inhibition of class II peptide loading with agents that raise the pH of acidic intracellular organelles was the ®rst evidence that successful class

78

paula wolf bryant et al.

II-restricted antigen presentation relies on the function of endocytic proteases (Neefjes et al., 1990; Wolf and Ploegh, 1995). As mentioned, the propiece of a lysosomal protease must be removed to activate the enzyme. Whether or not cleavage of the propeptide occurs by autocatalysis or by the action of another protease, an acidic pH is necessary. It is thought that the protonation of groups adjacent to the propeptide at low pH may increase the susceptibility of the propeptide to cleavage. Alternatively, the tertiary structure of the proenzyme may change with decreasing pH, rendering the propeptide vulnerable to cleavage (Riese and Chapman, 2000). The pH in each vacuolar organelle of eukaryotes, which include endosomes, lysosomes, and the TGN, is maintained by ATP-dependent proton pumps known as vacuolar H‡-ATPases (Nelson, 1992). Neutralization of these compartments with acidotropic drugs, such as chloroquine and ammonium chloride (Thorens and Vassalli, 1986), and carboxylic ionophores such as monensin, nigericin, and X537A (Tartakoff, 1983), prevents lysosomal protease function and, in turn, class II-restricted antigen presentation. Selective inhibitors of the vacuolar H‡-ATPase include macrolide antibiotics such as ba®lomycin A1 and concanamycin B (Bowman et al., 1988; Woo et al., 1992; Yoshimori et al., 1991). Inactivation of the vacuolar H‡-ATPase of mammalian cells with these agents affects not only pH but also the formation of endosomal carrier vesicles as well as sorting events in the TGN (Benaroch et al., 1995; Mellman et al., 1986). Hence, an acidic pH and/or protease function in the endocytic pathway is important for the intracellular transport of proteins, including their sorting and recycling in endosomes. Although an acidic pH is required to remove the propiece and thereby activate the enzyme, cathepsin S is unique in that it maintains its enzymatic function at neutral pH (Bromme et al., 1989; Shi et al., 1992). This feature of Cat S be®ts its critical role in Ii processing (discussed below), a prerequisite for loading of class II molecules with antigenic peptides. Class II±Ii complexes and Cat S are present in a variety of endocytic structures, from early endosomes of relatively high pH to the acidic lysosomes (Driessen et al., 1999). Earlier experiments conducted largely in B cells placed considerable emphasis on late endosomal compartments for proteolysis of Ii and maturation of class II molecules, but it is now clear that early endosomes are an essential part of the class II antigen presentation pathway (Brachet et al., 1999; Pond and Watts, 1999). Hence, the pH ¯exibility of Cat S, in principle, enables class II peptide loading in a variety of endocytic compartments. 2. Cytokines and Transcriptional Control The expression of MHC class II molecules and key accessory components of the class II antigen presentation pathway are induced by IFN-g. Accordingly, IFN-g regulates the expression of many of the cathepsins with critical roles in

antigen presentation by mhc class ii molecules

79

antigen presentation, such as cathepsins S, L, and F (Chapman, 1998; Shi et al., 2000). The expression of Cat B (Li and Bever, 1998), Cat H (Lafuse et al., 1995), and Cat W (Chapman, 1998) is also induced by IFN-g. The cytokine interleukin6 (IL-6) was shown to raise the pH of early endosomes, which could in¯uence the pattern of Ii processing in this compartment (Fuchs et al., 1989), although this was not tested. Moreover, IL-6 treatment of DCs appeared to in¯uence the types of epitopes derived from hen egg lysosome (HEL) that were presented by class II molecules (Drakesmith et al., 1998). Although the mechanism by which IL-6 in¯uences antigen presentation is not known, its alteration of endosomal pH could affect the types of peptides generated from intact antigens and/or could in¯uence the peptide-editing functions of the accessory molecule DM. The development of immature DCs, incapable of antigen presentation, into APCs that can support antigen presentation by class II molecules, requires the action of proin¯ammatory cytokines. In turn, the same cytokines that induce DC maturation also regulate the activity of proteases that participate in Ii and/ or antigen processing. The proin¯ammatory cytokines tumor necrosis factor a (TNF-a) and IL-1b rapidly increase the activity of Cat S and Cat B in human DCs and promote class II peptide loading and presentation on the cell surface (Fiebiger et al., 2001). In contrast, the antiin¯ammatory cytokine IL-10 attenuates the levels of both Cat B and Cat S in DCs (Fiebiger et al., 2001). Like IL-6, IL-10 seems to inhibit protease activities in DCs by elevating the pH of the endocytic pathway (Fiebiger et al., 2001). An earlier study demonstrated that peptide loading of the murine class II molecule I-Ak in immature DCs could be induced upon exposure to TNF-a (Inaba et al., 2000). Hence, the balance of pro- and antiin¯ammatory cytokines in DCs, and consequently the cellular proteolytic activities, strictly correlates with antigen presentation. Certain cathepsins exhibit tissue-speci®c expression that favors their participation in antigen presentation. Active Cat S is present predominantly in bone marrow-derived APCs (B cells, DCs, and macrophages). Accordingly, Cat S is the key enzyme required during late stages of Ii proteolysis in these APC types (Nakagawa et al., 1999; Shi et al., 1999). In contrast, active Cat L is not found in naive B cells or DCs, but rather in macrophages and cortical thymic epithelial cells (Shi et al., 2000). Cat L plays a central role in Ii degradation in the thymus necessary for class II-mediated selection of maturing T cells at this location (Nakagawa et al., 1998). Additionally, the cysteine protease Cat F, capable of performing the same cleavages of Ii as Cat S in vitro, is preferentially active in IFN-g-induced alveolar macrophages (Shi et al., 2000). The expression levels of cathepsins in macrophages can be regulated at the level of transcription and/or of enzyme maturation. Activators of macrophages can either increase or decrease intracellular levels of cathepsins, as well as induce cathepsin secretion into the extracellular space, thereby regulating the intracellular and extracellular proteolytic potential of these APCs (Liuzzo et al., 1999; Wang et al.,

80

paula wolf bryant et al.

2000). Regulation of proteolysis in macrophages or DCs not only is relevant for antigen presentation but also affects the microenvironment of these APCs. Indeed, the release of cathepsins by macrophages can serve to remodel the extracellular matrix and guide cell motility. 3. Endogenous Inhibitors There are three types of endogenous cysteine protease inhibitors present in APCs that bind tightly and reversibly to the enzyme's active site: the enzyme's propiece, the cystatin superfamily of inhibitors, and a fragment of the p41 (exon 6b) isoform of Ii (discussed in Section VI; Chapman et al., 1997). Both the propeptides and the cystatins bind the enzyme's active site in reverse orientation to that of natural substrates (Stubbs et al., 1990). This prevents their digestion by the enzyme while occupying its active site. The propiece appears to function as an inhibitor only until the enzyme reaches an acidic compartment, upon which the propiece self-dissociates and is cleaved. There is no evidence that the released propieces reside in the endocytic pathway to regulate the activity of resident proteases. The cystatins are perhaps the most studied endogenous cysteine protease inhibitors (Turk and Bode, 1991). They play an essential role in protecting cells and tissues from inappropriate proteolysis by enzymes that are overexpressed or escape the endocytic pathway (Chapman et al., 1997; Turk et al., 1997). Appropriately, the cystatins are expressed throughout the body in a tissuespeci®c manner and greatly outnumber cysteine proteases in the cytoplasm and extracellular spaces (Chapman et al., 1997). Mutations in some cystatins or alterations in the balance of these with their cognate cysteine proteases have been implicated in several diseases (Huh et al., 1999). The cystatins also inactivate cysteine proteases made by pathogens and thereby may serve as a defense mechanism against infection (Collins and Grubb, 1998; Korant et al., 1998; Stoka et al., 1995). On the other hand, cystatin-like protease inhibitors made by pathogens can also modulate the protease content and the processing of antigens in APCs (Manoury et al., 2001). The cystatin superfamily of cysteine protease inhibitors includes the ste®n subfamily (class I ˆ cystatins A and B), the cystatin subfamily (class II ˆ cystatins C, S, and D), and the kininogen subfamily (class III) (Brown and Dziegielewska, 1997). Two new class II cystatins have been de®ned, cystatin E(M) and cystatin F (leukocystatin), whose expression is restricted to hematopoietic cells (Halfon et al., 1998; Sotiropoulou et al., 1997). Although all of the cystatins mentioned target the cysteine proteases cathepsins B, F, K, L, S, and H, cystatin C has been shown to also inhibit the asparaginyl endopeptidase (Alvarez-Fernandez et al., 1999). Intriguing is the unusually long propiece of Cat F. In addition to the normal Cat L-like propiece domain, Cat F is unique in that its propiece has an additional amino-terminal domain that resembles cystatin (Nagler et al., 1999). Thus, the

antigen presentation by mhc class ii molecules

81

cystatin-like domain of Cat F may not only serve to inhibit the zymogen form of the enzyme but also, upon its removal in the endocytic pathway, may function as an inhibitor of other cysteine proteases. The mechanism by which cystatins inhibit cysteine proteases was revealed by determination of the three-dimensional structure of the complex of human cystatin B and papain (Stubbs et al., 1990). Three highly conserved regions of these inhibitorsÐtwo rigid hairpin loops and the ¯exible N-terminal region of the moleculeÐform a wedge that binds to the active site of cysteine proteases (Turk et al., 1997). For the majority of proteases exhibiting an ``open'' active site, the cystatins bind and inhibit the enzyme in a one-step reaction mechanism (Bjork and Ylinenjarvi, 1990; Lindahl et al., 1992; Pol et al., 1995; Turk et al., 1994; Turk et al., 1995a; Turk et al., 1996a). A protease with a partially obstructed active site, such as Cat B or Cat H, is not as accessible to these endogenous inhibitors (Illy et al., 1997; Quraishi et al., 1999). In this case, the cystatins inhibit via a two-step reaction mechanism, consisting of an initial weak binding followed by a conformational change that displaces the occluding loop of the active site so that a tight complex forms between the inhibitor and the enzyme (Musil et al., 1991; Nycander et al., 1998; Pavlova et al., 2000). The recent surge of attention given to cysteine proteases by immunologists has led to the discovery that the cystatins, once thought to primarily function in the extracellular space, may also play a role in regulating the proteolytic environment of the endocytic pathway and thus, in turn, antigen presentation. Dendritic cells are perhaps the most potent APCs in the body. They are distinct in that their ability to present antigen via class II molecules is tightly coordinated with their own maturation (Thery and Amigorena, 2001). Although immature dendritic cells have the ability to take up and process antigens, the peptides generated fail to be presented in a stable fashion on the cell surface by class II molecules. Results of one study indicated that Ii degradation by Cat S is inhibited in immature DCs (from the H-2b haplotype) by increased expression of cystatin C (Pierre and Mellman, 1998). Additional studies have shown a defect in the loading of I-Ak molecules with an HEL epitope in immature DCs, which is remedied upon induction of DC maturation with in¯ammatory mediators (Inaba et al., 2000). It is unclear whether the peptide-loading defect seen for I-Ak in immature DCs is the result of aberrant processing of HEL caused by cystatin C inhibition of Cat S (Inaba et al., 2000). It is likely that other mechanisms, besides Cat S inhibition by cystatin C, in¯uence the traf®cking of class II molecules in DCs. In Cat S = , ¯t3-I-induced DCs (mature), ab± Iip10 complexes accumulate in lysosomes, but eventually are deposited at the cell surface (Driessen et al., 1999). This is in marked contrast to the fate of ab± Iip10 complexes in immature DCs, in which the complexes are retained in lysosomes and then degraded, suggesting that other factors contribute to the retention of class II molecules in immature DCs (Villadangos and Ploegh,

82

paula wolf bryant et al.

2000). A study recently reported by Villadangos et al., demonstrated that class II-peptide complexes are in fact generated in immature DCs and delivered to the cell surface. However, in contrast to mature DCs, the class II-peptide complexes deposited on the surface of immature DCs are rapidly endocytosed and then degraded. Thus, the rate of endocytosisÐrather than Ii proteolysis as regulated by cystatin CÐmay dictate the antigen presentation capabilities of immature versus mature DCs (Villadangos et al., 2001). Finally, cathepsins likely regulate the activity of one another. Recently, Honey et al. (2001) demonstrated that in the absence of Cat S, the levels of active Cat L protein increased. Thus, Cat S could play a role in degrading mature Cat L. We have shown that competitive inhibitors can also exhibit a chaperone function in vivo: the p41 isoform of Ii protects a mature form of Cat L (see below). Whether this can be generalized to other protease-inhibitor complexes in vivo remains to be established. D. Targeting the Enzyme: Small-Molecule Active SiteDirected Probes Experiments in the past have relied on inhibitors of broad substrate speci®city to demonstrate the involvement of cysteine proteases in antigen presentation or any other physiological process. These inhibitors include the synthetic compounds leupeptin, E64, and benzyloxycarbonyl±Phe±Ala±diazomethane (Z±Phe±Ala±CHN2 ). Although these are potent inhibitors of cysteine proteases, they do not serve to distinguish between the active site similarities of closely related cathepsins, such as Cat L and Cat S. Much effort has been devoted to developing analogs of these inhibitors to visualize the enzymatic activity of a particular cathepsin in distinct cell types. The enzymatic mechanism utilized by cysteine proteases is conserved and well de®ned. This has enabled the development of a wide range of electrophilic substrate analogs that are reactive only with the conserved active site of cysteine proteases (Bogyo et al., 2000; Greenbaum et al., 2000). These inhibitors consist of a peptide speci®city determinant attached to an electrophile that becomes irreversibly alkylated when bound in close proximity to an attacking nucleophile (Bogyo et al., 2000). Such mechanism-based inhibitors include vinyl sulfones (Palmer et al., 1995), epoxysuccinic derivatives (Barrett et al., 1982), and dimethyl ketone derivatives (Pliura et al., 1992; Shaw, 1994; Shaw et al., 1986). By engineering these inhibitors to contain af®nity labels, such as radioactive iodine or biotin, they can be used as probes with which to measure the protease activity in distinct tissues (Bogyo et al., 2000; Greenbaum et al., 2000) (summarized in Table I). Such active site-directed probes not only measure the function of known cathepsins but also serve to identify new active proteases in a given pathway. Af®nity-labeled versions of the active site-directed inhibitors N-morpholinurea±leucine±homophenylalanine±vinyl sulfone±phenyl (LHVS) and JPM-

antigen presentation by mhc class ii molecules

83

TABLE I Active Site-Directed Probes and Affinity Labels Inhibitor

Speci®city

Af®nity label

Membrane permeable

A. Epoxysuccinates 1. E-64 NH H2N

O

H N

N H

H

N H H

O

OH

General

None

Yes

General

125

No

Cat B

None

Yes/no

Cat B

125

I

Yes

Cat L

None

Yes

General

125

No

O O

2. JPM-565 O

H N

125

O

HO

H

N H H

OH

I

O O

3. CA-074 O N H H

H H N

O O

OMe

N

O O

4. MB-074-OMe HO

O

125

N H H

O O

H H N

OMe

N

O O

5. CLIK148 O O

H H N

N H H

N

N

O O

6. DCG-04 O

O

H N

H2N

N H

O HO O

NH

O

H N O

N H H

H

OEt

O O

I, biotin

H S

NH H

N H

O

(continues)

paula wolf bryant et al.

84

TABLE I (continued) Inhibitor

Membrane permeable

Speci®city

Af®nity label

Cat S (>>>>B>L)

None

Yes

Cat S (>>>>B>L)

125

I

Yes

Cat S (>>>>B>L)

Biotin

Yes

General

125

I

Yes

General

125

I

Yes

B. Vinyl sulfones 7. LHVS O N

H N

N H

O

O

S

O

O

8. LHVS-PhOH O N

H N

N H

O

O

S

O

125

O

OH

9. Biotin-LHVS O

H N

N H HN

O

S

O

O

O

H HN O

S N H H

C. Diazomethyl ketones 10. ZYA-CH2 N2 125 OH O N H

O

H N

N2

O

O

11. FmocYA-CH2 N2 125 OH O N H

H N

O N2

O

O

Note: The authors thank Dr. H. S. Overkleeft for his assistance in preparing Table I.

antigen presentation by mhc class ii molecules

85

565 (both iodinated and biotinylated) have been used successfully to obtain a pro®le of the cathepsins that are active in distinct APC types (Driessen et al., 1999, No. 108; Shi et al., 2000, No. 95; Lennon-Dumenil, submitted for publication). Both inhibitors covalently bind the active sites of cysteine proteases. The membrane-permeable peptide vinyl sulfone, LHVS, speci®cally inhibits Cat S at nanomolar concentrations and Cat B to a lesser extent (Palmer et al., 1995). At higher concentrations, the speci®city of LHVS extends to additional cysteine proteases, including Cat L (Felbor et al., 2000; Lennon-Dumenil et al., submitted for publication). JPM-565 is an analog of the natural product E-64, which itself is a peptide epoxide that inhibits most of the known lysosomal cysteine proteases by covalently modifying the active-site nucleophile (Barrett et al., 1982). JPM-565 is not membrane permeable and exhibits the same general speci®city for cysteine proteases as E-64 (Bogyo et al., 2000; Greenbaum et al., 2000). Active cysteine proteases present in cell lysates are covalently modi®ed by the JPM-565 or LHVS af®nity-labeled probes. The labeled enzymes can be separated by SDS±PAGE and visualized either by autoradiography (when labeled with 125 I-JPM-565 or 125 I-LHVS; Bogyo et al., 2000; Driessen et al., 1999; Shi et al., 2000) or by blotting with streptavidin±horseradish peroxidase followed by chemiluminescense (when labeled with LHVSbiotin or JPM-565biotin ) (Greenbaum et al., 2000; Lennon-Dumenil et al., 2001; and P.W.B., unpublished data). Since the covalent modi®cations by LHVS and JPM-565 are mechanism based, labeling with these probes is proportional to the enzymatic activity of the protease targeted (Palmer et al., 1995). More recent developments include the synthesis of ¯uorescent probes that afford multiplexing and visualization of proteolytic activities in living cells (Bogyo, personal communication). In addition to the E-64 analog JPM-565, other peptide±epoxide-based inhibitors include the Cat B-speci®c inhibitors CA030 and CA074. These compounds were designed to mimic a carboxy-terminal dipeptide by replacing the terminal guanidinobutylamine of E-64 with a proline residue and terminating the chain with a free carboxylic group (Murata et al., 1991; Towatari et al., 1991). CA030 and CA074 bind with their peptide portion to Cat B's active site in inverse orientation to that observed for other epoxide-containing inhibitors of the E-64 family (Turk et al., 1995b). The speci®city of CA030 and CA074 is achieved through a tight binding composed of two H bonds between the Cterminal proline of CA030 and two highly conserved histidine residues from the occluding loop of Cat B (Turk et al., 1995b). To increase membrane permeability of these Cat B-speci®c inhibitors, an analog in which the proline free-acid group of CA-074 was converted to its methyl ester was generated and used in vivo to measure Cat B activity (Buttle et al., 1992). Recently, a new peptide epoxide, MB-074, was developed by replacing the isopropyl moiety of CA-074 with a phenol group as a site for attachment of radioactive iodine, yielding an af®nity-labeled probe with which to study Cat B activity (Bogyo et al., 2000).

86

paula wolf bryant et al.

A series of epoxysuccinate-based Cat L- and Cat S-speci®c inhibitors was also developed, designated CLIK (Katunuma et al., 1999). The representative inhibitor of this series, CLIK148, was shown to be the ®rst E-64 derivative using both prime-site and non-prime-site interactions along the enzyme's active-site cleft (Tsuge et al., 1999). CLIK148 inhibits Cat L exclusively. The speci®city for Cat L comes from the N-terminal pyridine of CLIK148 (Tsuge et al., 1999). Finally, using E-64 as a model, analogs were synthesized by varying the core peptide recognition sequence while adding af®nity tags at distal sites (Greenbaum et al., 2000). The resulting probes, DCG-03 and DCG-04, maintained the same general cysteine protease speci®city as the parental E-64 compound. Moreover, a library of DCG-04 derivatives was constructed in which the P2 position (leucine), the main speci®city determinant for many cysteine proteases, was replaced with different amino acids. This library was used to obtain the activity pro®les for multiple proteases in crude cellular extracts (Greenbaum et al., 2000). A ®nal group of well-characterized, active site-directed inhibitors are the membrane permeable, diazomethyl ketones, such as Cbz±Tyr±Ala±CN2 . As with the inhibitors above, Cbz±Tyr±Ala±CN2 irreversibly binds to the active site of cysteine proteases in proportion to their activity. An iodinated version of this compound, Cbz±125 ITyr±Ala±CN2 , has been used to demonstrate the activity of cysteine proteases in living cells (Mason et al., 1989a,b). Another analog of the peptidyl diazomethane probe, ¯uoren-9-ylmethoxycarbonyl (Fmoc)±[I2 ]Tyr±Ala±CHN2 , was produced by blocking the N terminus with a Fmoc group to permit further modi®cations of the probe without damaging the diazomethane group (Crawford et al., 1988; Xing et al., 1998). An iodinated form of Fmoc±[I2 ]Tyr±Ala±CHN2 reacted speci®cally with Cat B and Cat L, but not with Cat S. Use of this af®nity-labeled inhibitor revealed that unlike Cat B and Cat L, the active site of Cat S was restricted in that it could not accommodate the bulky di-iodotyrosine of the inhibitor (Xing et al., 1998). The advances in inhibitor and probe design summarized above provide an exciting prospect for the assessment of protease pro®les in cell populations available in limited numbers, such as subsets of professional APCs, and should allow the construction of a complete catalog of the proteases relevant for antigen presentation. Further re®nement of inhibitors may ultimately yield compounds suf®ciently selective to block a single protease in living cells. This approach nicely complements the genetic tools that are now available. E. Cathepsin-Deficient Mice Perhaps the most precise tool for determing which cysteine proteases are required for antigen presentation by class II molecules is a mouse in which an individual cathepsin gene has been ``knocked out.'' Mice de®cient in Cat B

antigen presentation by mhc class ii molecules

87

(Deussing et al., 1998), Cat D (Saftig et al., 1995), Cat S (Nakagawa et al., 1999; Shi et al., 1999), Cat L (Roth et al., 2000), Cat S/L (Shi et al., 2000), and Cat S/B (Driessen et al., in press) have been generated. The class II antigen presentation pathway in these mutant mice, including intracellular traf®cking, Ii processing, antigen processing, and peptide loading and presentation, have been studied in detail. These in vivo models for protease function quickly exposed the limitations of in vitro studies utilizing cultured cells and cysteine protease inhibitors as a means to assign function to individual cathepsins. Although cathepsins B, D, and L were all implicated by in vitro studies to be important in the degradation of protein antigens into class II presentable peptides, none appear indispensable in vivo (Deussing et al., 1998). Likewise, the ability of Cat B or Cat D to digest Ii in vitro is irrelevant for Ii processing in vivo. We now know that Cat S in bone marrow-derived APCs (Nakagawa et al., 1999; Shi et al., 1999) and Cat L in cortical thymic epithelial cells (Nakagawa et al., 1998) are required for a rate-limiting cleavage step late in the staged breakdown of Ii in vivo. The phenotypes of these mice with regard to antigen presentation are discussed in detail in the following sections. III. Proteolytic Digestion of Ii The primary function of Ii is to ensure that peptide-receptive class II molecules end up in the endocytic pathway. That class II molecules depend on Ii in order to function is somewhat paradoxical: the chaperone function of Ii in the formation of class II ab-dimers is required to create the peptidereceptive cleft, yet the physical complex formed between Ii and the ab-dimer prevents class II molecules from binding peptides. Thus, the function of proteases in the destruction of Ii, once ab±Ii complexes reach the endocytic pathway, is indispensable. The requirement for Ii does not abruptly end once class II molecules are delivered to their peptide-binding destination. Ii contains signals that retain class II molecules in endocytic compartments until peptideloading events ensue (Cresswell, 1996). In addition, occupation of the class II peptide-binding groove with the CLIP fragment of Ii preserves the peptidereceptive status of the class II ab-dimer, thereby preventing degradation of an empty class II molecule, until antigenic peptides are loaded with (or without) the assistance of DM (Alfonso and Karlsson, 2000). Hence, proteases must coordinate their digestion of Ii in such a way as to preserve the peptidereceptive state of the class II ab-dimer. Which are the key endocytic proteases that participate in the stepwise degradation of Ii? A. Cat B and Cat D Are Not Required for Ii Proteolysis Early experiments employing protease inhibitors of broad speci®city revealed the importance of both cysteine and aspartyl proteases in Ii

88

paula wolf bryant et al.

breakdown (Amigorena et al., 1995; Blum and Cresswell, 1988; Maric et al., 1994; Neefjes and Ploegh, 1992). As Cat B and Cat D are the most abundant endocytic proteases, they were predicted to be the major enzymes involved in Ii proteolysis. Indeed, in vitro, both Cat B and Cat D could digest Ii away from ab±Ii complexes either in cell lysates or in puri®ed preparations (Avva and Cresswell, 1994; Mizuochi et al., 1994; Reyes et al., 1991; Roche and Cresswell, 1991). However, the commercially available enzyme preparations used for these initial in vitro studies proved impure, as they were contaminated with other, less abundant cathepsins. Experiments using pure preparations of both Cat B and Cat D failed to reproduce the original resultsÐneither enzyme was capable of removing Ii from ab±Ii complexes synthesized in vitro (Riese et al., 1996). Still, these in vitro experiments were ¯awed because they did not emulate the in vivo levels and/or ratios of enzymes to substrates, nor could the role of other proteases in initiating Ii breakdown or in rendering Ii cleavage sites accessible to key enzymes be assessed. Which enzymes participate in Ii proteolysis in vivo? To correctly identify the proteases involved in Ii breakdown in vivo, mutant mice that are de®cient in distinct cathepsins were analyzed. The proteolytic digestion of Ii in APCs isolated from mice that lacked either Cat B (Deussing et al., 1998) or Cat D (Villadangos et al., 1997) was unaffected. In fact, the phenotype of these mice with regard to the rate of Ii digestion, and the types of proteolytic intermediates generated, was indistinguishable from that of wildtype mice. Hence, despite their copious levels in APCs, neither Cat B nor Cat D is essential for normal Ii removal in vivo. B. Conversion of ab±Ii to ab±Iip10 The proteolytic conversion of ab±Ii molecules into ab±peptide complexes was largely dissected by performing pulse±chase experiments on APCs followed by immunoprecipitation of class II±Ii complexes after different times of chase. To reveal the proteolytic steps that remove Ii from class II molecules during their maturation, the APCs analyzed in these biochemical experiments were treated with speci®c protease inhibitors, resulting in the accumulation of Ii breakdown intermediates (Morton et al., 1995; Neefjes and Ploegh, 1992; Villadangos et al., 1997). These represent ``normal'' intermediates of Ii processing (i.e., occur in uninhibited cells): stalling Ii processing by inhibition of speci®c proteases merely allows their capture and detection. The Ii fragments that remain associated with class II molecules are coprecipitated with the ab-dimer and can be visualized by SDS±PAGE. The identity of these Ii processing intermediates was determined by their molecular weights and their reactivity with antibodies speci®c for distinct regions of intact Ii. A unique feature of certain class II alleles (i.e., murine I-Ab , human DR1) when bound to more terminal Ii breakdown intermediates as well as peptides is the stability of

antigen presentation by mhc class ii molecules

89

the complex in SDS. The intact complex can be visualized by SDS±PAGE if the immunoprecipitated complexes are not boiled prior to separation (SadeghNasseri and Germain, 1991). Such experiments conducted on both mouse and human APCs have led to a model for the stepwise degradation of intact Ii into the ®nal end product, CLIP, as depicted in Fig. 1B (Driessen et al., 1999; Riese and Chapman, 2000; Villadangos et al., 1999). The ®rst step in Ii processing involves unraveling the class II±Ii nonameric complex that is delivered to early endosomes. The framework of this complex is the Ii homotrimer, held together via trimerization domains located C-terminal to CLIP (Cresswell, 1996). Each individual Ii in the homotrimer is attached to one class II ab-dimer via its CLIP region and the dimer's peptide-binding groove. Proteolytic cleavage of the carboxy-terminal region of Ii results in disruption of the homotrimerÐand thus the nonameric complexÐinto monomeric ab±Ii complexes. Noncysteine proteases are responsible for these early Ii processing events, as the cysteine protease inhibitor leupeptin does not prevent elimination of the C-terminal trimerization region (Blum and Cresswell, 1988; Maric et al., 1994; Villadangos et al., 1997). However, the identity of the protease(s) involved remains unknown. Although Cat D = mice exhibited no defect in Ii processing (Villadangos et al., 1997), inhibitors of aspartic proteases affect Ii breakdown (Maric et al., 1994; Zhang et al., 2000). Cat E, or possibly AEP, may participate in these early cleavage events. All steps subsequent to the ®rst carboxy-terminal cleavage are leupeptin sensitive and thus rely on cysteine proteases. Several class II-associated Ii breakdown intermediates accumulate in APCs treated with leupeptin (Blum and Cresswell, 1988; Neefjes and Ploegh, 1992). The ®rst predominant intermediate is a 22- to 24-kDa (Iip22) N-terminal fragment resulting from removal of the COOH-terminal trimerization domain by an aspartic protease. This fragment spans residues 1 through 160 of intact Ii and is detected more readily in human APCs and in H-2d haplotype mice (Villadangos et al., 1997). The second major Ii intermediate that accumulates in the presence of leupeptin is approximately 10 kDa (Iip10). The Iip10 fragment begins at the very NH2 terminus of Ii and extends just through to the COOH-terminus of CLIP (100 residues; Villadangos et al., 1997). A processing intermediate of 18 kDa also accumulates in bone marrow-derived mouse APCs (Villadangos et al., 1997). The data support a precursor/product relationship between these Ii processing intermediates, with Iip22 preceding Iip18, which proceeds Iip10. The cysteine proteases absolutely essential for conversion of Iip22 into Iip10 in vivo are still not known. C. Generation of ab±CLIP Complexes Treatment of APCs with leupeptin uncovers two key Ii processing steps that require cysteine proteases: the conversion of ab±Iip22 into ab±Iip10,

90

paula wolf bryant et al.

and further digestion of ab±Iip10 into the degradative end product ab±CLIP. ab±CLIP is not generated in the presence of leupeptin. Since the cysteine protease Cat S is expressed primarily in APC, and its expression is induced by INF-g (which upregulates class II expression), Cat S was considered as a likely participant in these late stages of Ii breakdown. Indeed, treatment of B cells and DCs with the Cat S-speci®c inhibitor LHVS (Bromme and McGrath, 1996) resulted in the accumulation of ab±Iip10 complexes (Riese et al., 1996; Villadangos et al., 1997). Moreover, ab±Iip10 complexes isolated from LHVS-treated APCs could be converted in vitro into ab±CLIP by digestion with recombinant Cat S (Riese et al., 1996). Hence, ab±Iip10 (not ab±Iip22) serves as an in vivo substrate for Cat S to complete Ii proteolysis. Unlike Cat B and Cat D, the ability of Cat S to cleave ab±Iip10 complexes into ab±CLIP in vitro held true in vivo. The bulk of Ii processing was stalled at the Iip10 stage in bone marrow-derived APCs isolated from Cat S = (H±2b ) mice (Nakagawa et al., 1999; Shi et al., 1999). To a lesser extent, ab±p22 and ab±p18 also accumulated in the absence of Cat S, indicating that, although not absolutely required, Cat S can mediate processing of these intermediates (Driessen et al., 1999). Whereas ab±p22 accumulated preferentially in early and late endosomes, ab±p18 accumulated in lysosomal compartments (Driessen et al., 1999). Despite the accumulation of ab±p10 (Ab ±Iip10) complexes in Cat S = APCs, ab±peptide complexes were eventually detected at the cell surface, albeit with considerable delay (Shi et al., 1999; Wolf Bryant et al., submitted for publication). The defect in peptide-loading extended beyond kinetics to include the types of epitopes that could be presented by class II molecules in these mutant mice (Nakagawa et al., 1999; Shi et al., 1999). Despite the defect observed in Ii proteolysis in Cat S-de®cient B cells, T-cell selection in Cat S = mice was unaffected by the mutation, as normal numbers of CD4‡ T cells were found in peripheral lymphoid compartments (Nakagawa et al., 1999; Shi et al., 1999). CD4‡ T cells are positively selected as they migrate through the thymus by recognition of class II±peptide (self) complexes expressed on cortical thymic epithelial cells. The proteases required for the late stages of Ii processing are tissue speci®c. Whereas Cat S is required to convert ab±Iip10 into ab±CLIP in BM-derived B cells and DCs, Cat L is required for this cleavage in cTECs (Nakagawa et al., 1998), thus explaining why T-cell selection is normal in Cat S = mice. Cat L is also expressed in macrophages along with Cat S. Indeed, the defect in Ii breakdown seen in Cat S = B cells was not as severe as that in Cat S = macrophages (Shi et al., 2000). However, the Ii processing defect was no more pronounced in macrophages (alveolar and peritoneal) isolated from Cat S/L double-de®cient mice than it was in the single, Cat S = macrophages (Shi et al., 2000). Cat S = macrophages treated

antigen presentation by mhc class ii molecules

91

with IFN-g express Cat F, which can digest ab±Iip10 into ab±CLIP, at least in vitro (Shi et al., 2000). If the rates of ab±p10 conversion into ab±peptide re¯ect the responsible enzymes' ef®ciency in generating ab±CLIP (Wolf Bryant et al., submitted for publication), then Cat S is still by far the most ef®cient and the preferred protease for generating ab±CLIP even in macrophages, with Cat F running a close second. The requirement for Cat S in generating ab±CLIP complexes is not only tissue and cell-type speci®c, but also varies widely among class II alleles. The class II alleles I-Aq , I-Ak , I-Au , and I-As are less dependent on Cat S for ef®cient Ii removal and subsequent peptide loading than is I-Ab (Nakagawa et al., 1999; Villadangos et al., 1997). D. Proteolytic Requirements for DM-Mediated Peptide Loading The removal of CLIP from the peptide-binding groove of most class II alleles and subsequent loading with antigenic peptides is catalyzed by the accessory molecule DM (Alfonso and Karlsson, 2000; Busch et al., 2000). In DM-de®cient animals, 99.8% of I-Ab molecules are occupied with CLIP (Fung-Leung et al., 1996; Martin et al., 1996; Miyazaki et al., 1996). The dependency on DM for CLIP±peptide exchange is allele speci®c. A physical interaction between DM and class II molecules is required for exchange of CLIP for antigenic peptides. The lateral aspects of DM bind to a region near the N terminus of the class II peptide-binding cleft (Doebele et al., 2000; Guerra et al., 1998; Mosyak et al., 1998). Kinetic analysis of DM-mediated CLIP±peptide exchange (Vogt et al., 1996) suggests that interaction with DM imparts an ``open-transition'' state to the class II peptide-binding cleft that favors CLIP release. The crystal structure of I-Ad ±CLIP suggests that the breakage of a few hydrogen bonds between the N terminus of the peptide-binding groove and CLIP would suf®ce to mediate CLIP's release (Mosyak et al., 1998). Loss of a single H bond at position 81 of the I-Ad b-chain, or of two H bonds at b82, is suf®cient to render I-Ad incapable of stable interaction with CLIP (and the antigenic peptide OVA) (Bryant et al., 1999). The changes in class II structure imposed by the b81 and b82 mutations may exemplify the types of alterations mediated by DM interaction that enable CLIP±peptide exchange. What are the minimal conformational requirements for DM's in vivo substrate(s)? No study to date has identi®ed the in vivo substrate for DM. DM does not interact with class II molecules associated with intact Ii, and thus Ii proteolysis is a prerequisite for DM function. The identity of DM's substrate cannot be found in the DM = mice alone. DM = mice contain a functionally complete repertoire of endocytic proteases, and thus the degradation of I-Ab -associated Ii proceeds to completion regardless of the DM mutation. Therefore, the accumulation of Ab ±CLIP complexes in DM = APCs merely emphasizes the

92

paula wolf bryant et al.

requirement for DM in liberating the peptide-binding groove of I-Ab at some stage during Ii removal. It is not known whether DM can interact with Ii breakdown intermediates upstream of CLIP in vivo. In vitro, DM could exchange the Ii breakdown intermediates p22 and p10 in complex with the human class II molecule, HLA-DR1, in addition to CLIP (Denzin and Cresswell, 1995; Denzin et al., 1996). However, the accumulation of ab±Iip10 complexes in Cat S = APCs together with the delayed appearance of ab± peptide complexes on the cell surface (Driessen et al., 1999; Nakagawa et al., 1999; Shi et al., 1999) suggests that DM does not prefer either p22 or p10 as its substrate, but rather a more terminal Ii processing intermediate, such as CLIP. Thus, the appearance of ab±peptide complexes on the surface of Cat S = APCs, if DM dependent, could result from one of two mechanisms. First, a yet unknown enzyme distinct from Cat S that can generate the ab±CLIP substrate used by DM may be present in B cells. Alternatively, DM may exchange Iip10 for antigenic peptides or some processing intermediate between Iip10 and CLIP, albeit considerably less ef®ciently.

IV. Proteolytic Control of Vesicle Biogenesis and Class II Trafficking through the Endocytic Pathway A. Targeting Class II Molecules to Endocytic Compartments Successful delivery of MHC class II molecules to compartments of the endosomal/lysosomal pathway relies mainly on their prior association with Ii (Bakke and Dobberstein, 1990; Lotteau et al., 1990). Ii contains two leucinebased signals in its cytoplasmic tail that confer this speci®c targeting function (Bremnes et al., 1994; Odorizzi et al., 1994). The cytosolic factors that interact with Ii's tail and thus direct these traf®cking events have yet to be identi®ed. The m-chain of the adaptor protein complexes AP1 and AP2, which provide speci®city for clathrin assembly and recognize sorting motifs in the cytoplasmic domains of membrane proteins, can interact with Ii's cytoplasmic tail in vitro (Hofmann et al., 1999). Whether such interactions occur in living cells remains to be demonstrated. Ii also affects the ``architecture'' of the endocytic pathway, an attribute that may play a role in antigen processing and/or presentation. The expression of high levels of Ii induces the formation of large endocytic compartments called macrosomes (Pieters et al., 1993; Romagnoli et al., 1993). Macrosome formation correlates with a delay in transport of proteins from early endosomes to late endosomes/lysosomes (Gorvel et al., 1995; Romagnoli et al., 1993). Langerhans cells are the only APCs found to contain such large acidic vesicles. The presence of macrosomes in these APCs correlates with a high rate of Ii synthesis (Kampgen et al., 1991). Moreover, the formation of macrosomes is depend-

antigen presentation by mhc class ii molecules

93

ent on the trimerization of Ii (Gedde-Dahl et al., 1997), a property that is required for proper delivery of ab±Ii complexes to endosomal compartments (Arneson and Miller, 1995). Thus, it is likely that the ``trimerized'' cytoplasmic tail of Ii interacts with components of the molecular machinery involved in membrane traf®c. To identify these components, we recently synthesized an af®nity matrix in which the trimerization of Ii's cytoplasmic tail is forced chemically. This approach allowed us to identify hsc70 as one the cytosolic partners capable of interacting with Ii (LagaudrieÁre-Gesbert et al., submitted for publication). Hsc70 was initially identi®ed as the uncoating ATPase that dissociates clathrin triskelions from clathrin coated-vesicles (Schlossman et al., 1984). Analysis of the clathrin-coated vesicle cycle in cells expressing ATPasede®cient hsc70 mutants suggested that this protein may also be required to chaperone cytosolic clathrin triskelions to allow their recruitement to coated pits (Newmyer and Schmid, 2001). The ability of Ii to recruit hsc70 activity might cause uncoating and thereby inhibit subsequent ®ssion reactions essential for retrieval of vesicular constituents. This would shift the balance toward fusion and lead to an increase in size of the target organelle carrying Ii, as seen in COS cells expressing Ii. Indeed, expression of a dominant-negative version of hsc70 in Ii-transfected COS cells counteracted the ability of Ii to modify the endocytic pathway, demonstrating an interaction in vivo of Ii with hsc70 as part of the machinery of vesicular transport (LagaudrieÁre-Gesbert et al., submitted for publication). While earlier studies postulated a role for hsc70 in peptide delivery to class II molecules (Panjwani et al., 1999) by a yet unidenti®ed mechanism, our recent data suggest that hsc70's role extends to maintenance of a properly organized endocytic pathway and, in this manner, contributes to MHC class II-restricted antigen presentation. B. Is There a Connection between the Processing of Ii and the Architecture of the Endocytic Pathway? Dendritic cells isolated from Cat S = mice have been exploited as a model to dissect the role of proteases in the traf®cking and maturation of class II molecules along the endocytic route. Immuno¯uorescence and subcellular fractionation experiments demonstrated that in the absence of Cat S, class II molecules complexed to N-terminal fragments of Ii (mainly Iip10) accumulate in late-endocytic compartments of DCs (Driessen et al., 1999). These ®ndings were reminiscent of earlier studies conducted in leupeptin-treated B cells and immature BM-derived DCs, in which the retention of class II molecules in late-endocytic compartments correlated with the low activity of cysteine proteases in general (Brachet et al., 1997) or of Cat S (Pierre and Mellman, 1998). Hence, Ii proteolysis must proceed in an ordered fashion to the end stage of CLIP to ensure proper traf®cking of class II molecules through the endocytic route.

94

paula wolf bryant et al.

Furthermore, the cell biological studies referred to above proved that active Cat S is not restricted to late-endocytic compartments, but instead can be found all along the endocytic route of DCs, including early endosomes (Driessen et al., 1999). The intracellular distribution of Cat S agrees with its unique ability to remain active at neutral pH. Thus, the conversion of ab±Iip10 into ab±CLIP could occur in a variety of endocytic compartments, enabling class II molecules to sample distinct pools of differentially compartmentalized antigenic peptides. Since, in addition to Iip10, the Ii intermediates p22 and p18 also accumulate in late compartments of Cat S = DC (Driessen et al., 1999), those fragments of Ii may represent additional substrates for Cat S. Cat S not only governs MHC class II traf®cking via its role in Ii degradation, but also affects the architecture of endocytic vesicles themselves, independent of class II±Ii. DCs isolated from Cat S = mice exhibit increased levels of some late-endocytic markers, such as LAMP-1, when analyzed by confocal microscopy (Driessen et al., 1999). This correlated with an increase in b-hexosaminidase activity in fractionated Cat S = DC (Driessen et al., 1999). A more thorough morphological examination of endocytic vesicles in Cat S = cells was conducted using electron microscopy with striking results: compartments that stained for late-endocytic markers were considerably enlarged in Cat S = cells as compared to those in wildtype cells (P. Peters, personal communication). Since the cytoplasmic tail of Ii, responsible for binding hsc70, remains intact in Cat S = DCs (note that Iip18, Iip22 and Iip10 fragments retain this element), it is tempting to propose that the recruitment of the uncoating ATPase accounts for this abnormal morphology. The induction of enlarged late-endocytic vesicles in the absence of Cat S may parallel the formation of macrosomes observed in Ii-transfected COS cells, whose formation relies on the interaction of the Ii with hsc70. Either Cat S plays a direct role in controlling protein traf®cking along the endocytic pathway or the enlarged endocytic compartments observed in Cat S = cells are merely the result of the accumulation of class II Iip10 complexes and recruitment of hsc70. The answer awaits the generation and characterization of Cat S  Ii double knockout mice. V. The Role of Ii in Regulating the Proteolytic Activities of APCs So far we have emphasized the degree to which the class II antigen presentation pathway exploits the normal housekeeping functions of a cellÐspeci®cally proteolysisÐfor its own purpose and success. From proper traf®cking of class II complexes through endocytic vesicles to Ii breakdown and peptide acquisition, the functions of a cell's proteases are indispensable. However, the immune system may not be doing all the ``taking'' and no ``giving.'' The proteolytic machinery of the cell may itself exploit components of the class II antigen presentation pathway for its own bene®t. We now discuss recent data that

antigen presentation by mhc class ii molecules

95

demonstrate a role for Ii in the regulation of the proteolytic environment of an individual APC. A. The p41 Fragment Binds the Active Site of Cat L In addition to their indistinguishable functions as chaperones for class II folding and intracellular traf®cking (Peterson and Miller, 1992; Takaesu et al., 1995, 1997), both the p31 and p41 isoforms of murine Ii can be converted into CLIP (Fineschi et al., 1995; Takaesu et al., 1995, 1997), presumably by Cat S. p31 and p41 are expressed in different ratios in the various types of APCs. Whereas p41 represents no more than 10% of the total pool of Ii in splenocytes, its expression levels are considerably higher in macrophages, DC, and Langerhans cells (Kampgen et al., 1991; Koch and Harris, 1984; Pierre and Mellman, 1998; Pure et al., 1990). The p41-speci®c 64-aa segment resembles a thyroglobulin type-1 domain, rich in cysteine residues (O'Sullivan et al., 1987). This p41 segment was found noncovalently bound to the mature form of Cat L puri®ed from human kidney (Ogrinc et al., 1993). The crystal structure of Cat L in a complex with the p41 segment complex shows that this fragment occupies the Cat L active site (Guncar et al., 1999). Moreover, in vitro studies demonstrate that the 65-aa segment of p41 inhibits Cat L enzymatic activity (Bevec et al., 1996; Fineschi et al., 1996; Turk et al., 1999). What is the functional relevance of Cat L±p41 association in vivo? B. p41 Is Required for Full Cat L Activity As most lysosomal hydrolases, Cat L is synthesized in the ER as a proenzyme (Erickson, 1989; McGrath, 1999). The Cat L proregion consists of 96 amino acids that occupy its active-site cleft, maintaining the enzyme in an inactive state (Coulombe et al., 1996). During export along the endocytic pathway, proCat L undergoes several proteolytic cleavages to generate the Cat L singlechain (30 kDa) and two-chain mature forms, composed of a 25-kDa heavy chain linked to a 5-kDa light chain by disul®de bonds (Erickson, 1989; Ishidoh and Kominami, 1998; McGrath, 1999; Reilly et al., 1989). Mutant mouse strains de®cient for Ii, or expressing either p31 or p41 Ii, were developed to study the function of each isoform in antigen presentation (Takaesu et al., 1995, 1997). We have used these animals to study the functional signi®cance of the Cat L± p41 interaction in vivo. Contrary to expectations, Cat L expression and activity are strongly reduced in APCs isolated from Ii-de®cient mice, as measured by immunoblotting and active-site labeling experiments (Lennon-Dumenil et al., 2001). In agreement with the described Cat L±p41 interaction, the mature two-chain forms of Cat L depend on p41 (not p31) to be expressed at wildtype levels: levels of mature Cat L are considerably decreased in the endocytic compartments of cells that lack only the p41 isoform of Ii (Lennon-Dumenil et al., 2001).

96

paula wolf bryant et al.

C. p41 Protects the Mature Forms of Cat L from Degradation by Surrounding Cysteine Proteases Why are the levels of mature Cat L decreased in the absence of p41? This question was addressed by performing pulse±chase experiments in the presence of different protease inhibitors. These analyses showed that in the absence of p41, Cat L is degraded in acidic compartments by cysteine proteases (Lennon-Dumenil et al. submitted for publication). Thus, p41 appears to protect Cat L from premature destruction. p41 is therefore not merely an inhibitor of Cat L enzymatic activity, but serves as a chaperone to help maintain a pool of mature enzyme in late-endocytic compartments of antigen-presenting cells. Prevention of Cat L destruction by p41 might be linked to the ability of p41 to bind the active site of Cat L (Guncar et al., 1999). This interaction is reminiscent of the interaction of Cat L with its propeptide, which not only maintains the enzyme in an inactive state but also assists its folding and stabilizes its conformation (Jerala et al., 1998; McGrath, 1999). As pro-Cat L traverses endocytic compartments, the attendant drop in pH induces conformational changes in the propeptide, which then detaches from the Cat L active site and is cleaved to generate the 30-kDa single-chain form of the enzyme (Jerala et al., 1998). Even though removal of the Cat L propeptide is necessary for enzymatic activation, it may destabilize the tertiary structure of the enzyme suf®ciently to allow partial unfolding and subsequent degradation. In lateendocytic compartments p41 may function as a Cat L chaperone, not unlike the propeptide in compartments with neutral pH. This would allow the cell to maintain a pool of (latent) mature Cat L in late-endocytic compartments that is protected from destruction by hydrolases. Our data demonstrate that leupeptin-sensitive cysteine proteases control the turnover of Cat L in late-endocytic compartments by partially degrading its mature active forms (Lennon-Dumenil et al., 2001). Which enzymes are involved in this process? The levels of 24-kDa Cat L are increased in BM macrophages from Cat S and Cat B knockout mice (AML, unpublished data). In agreement with this observation, recent data published by Honey et al. (2000) showed increased Cat L activity in cells lacking Cat S. In addition to Cat S and Cat B, Cat L could regulate its own levels of activity by self-degradation. In this context, p41 would exert a protective effect by preventing self-destruction of mature Cat L. VI. Antigen Processing The presentation of antigen requires that it ®rst be fragmented by proteases into smaller peptides that can ®t easily into the peptide-binding groove of an MHC molecule. MHC class II molecules bind and present peptides derived

antigen presentation by mhc class ii molecules

97

from antigens that gain access to the endocytic pathway by endocytosis, phagocytosis, or both (Watts, 1997; Wolf and Ploegh, 1995). Antigens taken up by APCs traverse the endocytic route via vesicles of decreasing pH. The gradual exposure of antigen to increased protease activity ensures that the antigen is not completely destroyed but is instead broken down into class II-presentable peptides. Many if not all of the endocytic compartments may host the loading of peptides onto class II ab-dimers. Given the fact that each endocytic vesicle is likely to vary in its proteolytic content coupled to the different population of antigens housed in each, class II molecules are ensured access to a diverse pool of antigenic peptides. Moreover, the location at which class II molecules become receptive to peptide loading may determine the types of epitopes presented to T cells. In contrast to our detailed understanding of Ii processing, the proteolytic events that lead to the generation of class II-presentable peptides from intact antigen remain somewhat enigmatic. Nonetheless, the generation of these T-cell epitopes will no doubt require speci®c cleavages during the course of antigen destruction. Whereas the proteolytic cleavages required for ef®cient Ii removal from class II molecules occur during the late stages of Ii breakdown (i.e., Iip10 into CLIP), the speci®c cleavages required to ensure the generation of immunogenic epitopes from intact antigens may instead be needed during the early stages of antigen processing (see below). A. Antigen Acquisition Antigen acquisition is the ®rst step in MHC class II-restricted antigen presentation. The mechanisms by which acquisition occurs are diverse. These mechanisms are regulated and can be highly speci®c. The mode of antigen uptake by an APC depends on the source of the antigen, the type of APC, and the activation state of the APC. The method of internalization utilized dictates the endocytic compartment in which the antigen will be targeted. The mechanisms of antigen uptake can be either speci®c or nonspeci®c (Bakke and Nordeng, 1999). Nonspeci®c uptake involves ¯uid-phase endocytosis of extracellular ¯uid. The primary, nonspeci®c mechanism used by DCs to internalize exogenous antigens is macropinocytosis, in which large volumes of extracellular ¯uid are engulfed and the macrosolutes captured are concentrated in class II-positive endocytic compartments (Cella et al., 1997; Sallusto et al., 1995). Macropinocytosis is linked to membrane-ruf¯ing activity rather than uptake mediated by clathrin-coated pits. Only immature DCs can internalize antigen whereas mature DCs rather function in the presentation of the resulting peptides on the cell surface via class II molecules (Thery and Amigorena, 2001). Speci®c mechanisms of antigen uptake involve endocytosis and phagocytosis mediated by receptors expressed on the surface of APCs (Bakke and Nordeng,

98

paula wolf bryant et al.

1999; van Bergen et al., 1999; Watts, 1997). Each type of professional APCÐB cells, macrophages, and DCsÐexpresses distinct receptors that mediate the internalization of exogenous antigens (summarized in Table II). These receptors include the B-cell receptor (BCR, B cells only), Fc receptors (B cells, macrophages, and DCs), and members of the macrophage mannose receptor (MMR) C-type lectin family (macrophages and DCs). These receptors serve as scaffolds to concentrate the antigen they internalize in peptide-loading compartments. The result is an immune response with enhanced sensitivity. In fact, receptor-mediated endocytosis of antigen enables the immune system to respond to 103 to 104 lower antigen concentrations as compared to ¯uid-phase uptake of antigen. Indeed, allergen presentation to T cells is 100- to 1000-fold more effective if the allergen has been targeted via allergen-speci®c IgE and FceRI on APCs (Maurer et al., 1995, 1998). As atopic individuals have signi®cantly higher levels of FceRI on the cell surface of several APCs (Maurer et al., 1994), this is a good example of a mechanism that lowers the individual's TABLE II Modes of Antigen Uptake by District APC Types APC type

Ag-uptake mechanism

Receptors

B cell

Receptor-mediated endocytosis

BCR FcRs

Dendritic cell

Macropinocytosis Phagocytosis (nonspeci®c) Receptor-mediated phagocytosis

None None FcgRs Complement receptors Complement receptors FcgRs, FceRs ILT3 C-type lectins (DEC-205,MMR) Heat shock protein receptors (CD91) Integrins (avb5 for apoptotic corps) None

Receptor-mediated endocytosis

Restored macropinocytosis on activated DCs Macrophage

Receptor-mediated phagocytosis

Receptor-mediated endocytosis

Macropinocytosis

FcgRs Complement receptors CD14 Toll-like receptors FcgRs C-type lectins (MMR) ILT3 Integrins (avb3) None

antigen presentation by mhc class ii molecules

99

threshold to develop allergen-speci®c T-cell responses. In addition, phagocytosis of microorganisms and other particles by macrophages and DCs is most ef®cient when mediated by speci®c receptors, such as Fcg receptors, C3R, or CD14 (Gregory, 2000). Interaction between these receptors and their ligands leads to actin polymerization and phagosome formation. In most instances, the phagosome develops into a phagolysosome by fusing with early/ late endosomes and lysosomes of the host cell, and thus its contents become part of the ``normal'' endocytic route traveled by class II molecules. The phagosomes of the intracellular bacteria Legionella and Mycobacterium, however, exhibit limited fusion with their host cells' lysosomes (Ojcius et al., 1996). This may represent a means by which these pathogens evade the immune response. Unlike macrophages and DCs, B cells exhibit low rates of pinocytosis and phagocytosis. To compensate, they possess an antigen-speci®c receptor known as the BCR (Siemasko and Clark, 2001). The speci®city of the BCR allows it to recognize rare and low-af®nity antigens. Antigens capable of crosslinking the BCR are preferentially captured, as BCR aggregation enhances its internalization. The BCR is composed of membrane Ig, noncovalently associated with an Iga±Igb heterodimer. The membrane portion recognizes and binds speci®c antigen, while the heterodimer contains the immunoreceptor tyrosine-based activation motif that initiates the signaling cascades necessary for internalization and targeting of the antigen to endocytic compartments. BCR±antigen complexes are internalized via clathrin-coated pits into early endosomes and then sorted into various endocytic compartments containing class II molecules. The sorting of Ag to early endosomes depends on signals located in the cytoplasmic tail of Igb, whereas delivery of antigen to lateendocytic vesicles enriched for class II molecules requires the phosphorylation of tyrosine residues in the tail of Iga (Bakke and Nordeng, 1999; Siemasko and Clark, 2001). The FcgR receptor is utilized by most professional APCs to facilitate the enodyctosis or phagocytosis of immune complexes (Amigorena and Bonnerot, 199a,b). Several types of Fc receptors have been characterized that can be distinguished in part by their cytoplasmic tail sorting signals. These signals determine to which intracellular compartment the ligand bound by the Fc receptor will be delivered (Bakke and Nordeng, 1999). Macrophages and DCs take up mannosylated and/or fucosylated antigens via mannose receptors. Upon endocytosis, the manose receptor±ligand complex quickly uncouples and the mannose receptor is recycled back to the cell surface intact for round after round of antigen capture and internalization. The macrophage mannose receptor primarily recycles through peripheral, early endosomes (Engering et al., 1997; Tan et al., 1997), although some entry into late endosomes has been observed (Prigozy et al., 1997). Bone marrow-derived

100

paula wolf bryant et al.

DCs express another receptor that belongs to the MMR family, designated DEC-205 (Jiang et al., 1995; Mahnke et al., 2000). In contrast to MMR, DEC205 delivers antigens to late endosomes and lysosomes rich in MHC class II molecules (Mahnke et al., 2000). Acidic clusters in the cytoplasmic tail of DEC205 are responsible for its intracellular sorting functions. The targeting functions of DEC-205 ensure that antigen and class II molecules intersect. B. Marking the Antigen for Degradation Antigen degradation starts with the exposure of the native protein to low pH in a reducing environment to dissociate the antigen from its receptor and destabilize its native structure. Disul®de bounds must be reduced to unfold the protein and improve access of proteolytic enzymes (Fig. 1A). Reduction of disul®de bonds can be a rate-limiting step in antigen degradation in endocytic compartments (Collins et al., 1991; Jensen, 1991, 1993). The lysosomal thiol reductase (GILT) is an enzyme active at low pH that is capable of catalyzing disul®de bond reduction both in vivo and in vitro. GILT is expressed constitutively in antigen-presenting cells and is induced by IFN-g in other cell types, suggesting a role in antigen processing (Arunachalam et al., 2000; Phan et al., 2000). C. Antigen Degradation by Endocytic Proteases The redundancy in cleavage speci®city between the endocytic proteases has made identifying the enzymes required to generate speci®c T-cell epitopes dif®cult. The MHC class II peptide-binding cleft is open at both ends and can therefore accommodate peptides of various lengths, ranging from 12 residues to as many as 30 residues, with a preferred size of 15 residues. The characteristic peptide bound by class II consists of a core sequence with ragged N and C termini. This suggests that more than one enzyme is responsible for the proteolysis of antigen, involving an initial cleavage by endopeptidases followed by sequential trimming of the ends by amino- and carboxy-peptidases (Watts, 1997). Whether a speci®c protease dedicated to the generation of class II-presentable peptides from a single antigen is required is still a matter of debate. Recently, AEP was shown to initiate the proteolytic degradation of the carboxy-terminal domain of the tetanus toxin antigen (TTCF) in B lymphoblasts (Antoniou et al., 2000; Manoury et al., 1998). Cleavage of TTCF at a single site by AEP was shown to be critical for the generation and presentation of a variety of T-cell epitopes, even those distant from the cleavage site (Antoniou et al., 2000). Thus, a single cleavage by AEP was proposed to ``unlock'' the native TTCF structure, facilitating its subsequent degradation into presentable peptides by unde®ned proteases (Antoniou et al., 2000). Given the strict speci®city of AEP, it is unlikely that the bulk of antigens endure the same proteolytic fate as TTCF.

antigen presentation by mhc class ii molecules

101

Most of what we know concerning antigen degradation comes from studies performed in vitro, which, as shown for Ii processing, do not necessarily mirror what takes place in vivo. In vitro, cathepsins B, D, and E can generate T-cell epitopes when incubated with intact antigens (Bennett et al., 1992; Rodriguez and Diment, 1992, 1995; van Noort and Jacobs, 1994; Vidard et al., 1992). In vivo, a modest shift in the ef®ciency of presentation of some antigenic determinants was seen in APCs isolated from Cat B- and Cat D-de®cient animals (Deussing et al., 1998). Nonetheless, the overall capacity of Cat B = or Cat D = APCs to process and present antigens via class II molecules was unaffected. Even Cat S = APC were still capable of normal antigen presentation for most epitopes tested (Nakagawa et al., 1999; Shi et al., 1999). It is still possible that the presence or absence of one of these major cathepsins might be pivotal for the generation of immunogenic peptides from an invading pathogen or in other disease states. Cat B is important for the degradation of peptides, proteins, toxins, and even cell surface receptors that enter the cell via endocytosis or phagocytosis (Authier et al., 1999; Mort and Buttle, 1997; Zhang et al., 2000). In addition, the activity of Cat S was shown to be critical for the presentation of a pathogenic, arthritis-inducing, collagen-derived epitope in I-Aq mice (Nakagawa et al., 1999). Functional studies using a speci®c inhibitor of Cat E showed that this enzyme is essential for the processing of ovalbumin in murine A20 cells (Bennett et al., 1992). Furthermore, the regulation of Cat E during human B-cell activation indicates Cat E plays an important role in antigen degradation in these cells (Sealy et al., 1996). Little or no evidence exists for a role of Cat D and Cat L in antigen degradation in vivo. It is possible that Cat L's ability to degrade extracellular matrix proteins might be important in pathologic conditions where peptides from such proteins are presented. Indeed, the detection of a small population of ``empty'' class II complexes at the cell surface of DCs raises the possibility that processing outside the endocytic pathway could also play a role in class IIrestricted antigen (Santambrogio et al., 1999a, b). In addition, extracellular proteases may further trim antigens that occupy the cleft of class II molecules already at the cell surface and thus assist in generating the ®nal epitope seen by T cells. Such a role has been postulated for aminopeptidase N (APN, CD13) (Larsen et al., 1996). APN was capable of digesting the NH2 -terminal end of a long peptide bound in the cleft of a class II molecule. Perhaps more pronounced defects in antigen degradation would be observed in APCs where more than one of the essential cathepsins is absent. Studies with human DCs and high concentrations of the active-site inhibitor LHVS (at concentrations that inhibit Cat S, Cat B, and Cat L) indicate that Cat S and Cat B act concertedly to degrade antigen (Fiebiger et al., 2001). To avoid the limitations inherent in using pharmacological inhibitors and cell lysates, Driessen et al. conducted a study in which the contributions of cathepsins S, B, L,

102

paula wolf bryant et al.

and D in the degradation of antigen could be assessed in vivo (Driessen et al., 2001). The fate of a radiolabeled immune complex, 125 I-labeled F(ab0 )2 , internalized by the FcgR was followed in bone marrow-derived APCs isolated from the different cathepsin mutant mice. Nonspeci®cally radiolabeled 125 I-labeledF(ab0 )2 was degraded rather slowly by proteases into relatively stable peptide intermediates that could be resolved and visualized by SDS± PAGE and autoradiography. The results showed that both Cat B and Cat S mediated the bulk of 125 I-labeled F(ab0 )2 degradation, via independent pathways, while Cat L and Cat D were dispensable (Driessen et al., 2001). Regardless of the proteases involved, antigen unfolding and subsequent proteolysis must be balanced in such a way that the T-cell epitopes destined for presentation by class II molecules are not themselves destroyed in the process. The open ends of the MHC class II binding groove do allow binding of longer protein fragments, with the T-cell epitope lying in the core of the groove. Thus, the antigen may initially be digested into fragments somewhat longer than the ®nal immunogenic peptide. Once in the peptide-binding groove of class II, that which is tightly enclosed by the a-helices of the cleft would be protected from further degradation, while the ends of the fragment hanging outside the cleft would be accessible to further trimming by endo- and exopeptidases (i.e., Cat H and Cat B). Thus, the ®nal antigenic peptide seen by T cells may not be fully processed until after it is loaded onto class II. VII. Concluding Remarks The processing events that lead to peptide-loaded class II molecules can be de®ned in molecular terms because of decades of work in the ®eld of lysosomal biology. The classi®cation of essential proteases, the de®nition of their speci®city and the design of speci®c inhibitors predate the immunologist's appreciation of these efforts. As class II-restricted antigen presentation is completely dependent on lysosomal proteolysis, it is a subject that offers itself naturally to the blending of these two ®elds of study. For example, by drawing on the research of lysosomal biologists, immunologists were able to determine the key events of Ii proteolysis. It is clear from this review that the picture is still not complete. The proteases involved in the generation of the Iip10 intermediate are still unknown. Moreover, the speci®c proteolytic requirements for the generation of T-cell epitopes from intact antigens remain unclear. It is possible that the proteolytic digestion of antigen requires the action(s) of other components of the endocytic pathway, such as chaperones and the thiol reductase GILT. A more complete picture of the molecular and cellular requirements for the initiation of an antigen-speci®c immune response will depend on continued interactions between the ®elds of lysosomal biology and antigen presentation.

antigen presentation by mhc class ii molecules

103

References Alfonso, C., and Karlsson, L. (2000). Nonclassical MHC class II molecules. Annu. Rev. Immunol. 18, 113±142. Alvarez-Fernandez, M., Barrett, A. J., Gerhartz, B., Dando, P. M., Ni, J., and Abrahamson, M. (1999). Inhibition of mammalian legumain by some cystatins is due to a novel second reactive site. J. Biol. Chem. 274, 19195±19203. Amigorena, S., and Bonnerot, C. (1999a). Fc receptor signaling and traf®cking: A connection for antigen processing. Immunol. Rev. 172, 279±284. Amigorena, S., and Bonnerot, C. (1999b). Fc receptors for IgG and antigen presentation on MHC class I and class II molecules. Semin. Immunol. 11, 385±390. Amigorena, S., Webster, P., Drake, J., Newcomb, J., Cresswell, P., and Mellman, I. (1995). Invariant chain cleavage and peptide loading in major histocompatibility complex class II vesicles. J. Exp. Med. 181, 1729±1741. Antoniou, A. N., Blackwood, S. L., Mazzeo, D., and Watts, C. (2000). Control of antigen presentation by a single protease cleavage site. Immunity 12, 391±398. Arneson, L. S., and Miller, J. (1995). Ef®cient endosomal localization of major histocompatibility complex class II-invariant chain complexes requires multimerization of the invariant chain targeting sequence. J. Cell. Biol. 129, 1217±1228. Arunachalam, B., Phan, U. T., Geuze, H. J., and Cresswell, P. (2000). Enzymatic reduction of disul®de bonds in lysosomes: Characterization of a gamma-interferon-inducible lysosomal thiol reductase (GILT). Proc. Natl. Acad. Sci. USA 97, 745±750. Authier, F., Metioui, M., Bell, A. W., and Mort, J. S. (1999). Negative regulation of epidermal growth factor signaling by selective proteolytic mechanisms in the endosome mediated by cathepsin B. J. Biol. Chem. 274, 33723±33731. Avva, R. R., and Cresswell, P. (1994). In vivo and in vitro formation and dissociation of HLA±DR complexes with invariant chain-derived peptides. Immunity 1, 763±774. Bakke, O., and Dobberstein, B. (1990). MHC class II-associated invariant chain contains a sorting signal for endosomal compartments. Cell 63, 707±716. Bakke, O., and Nordeng, T. W. (1999). Intracellular traf®c to compartments for MHC class II peptide loading: Signals for endosomal and polarized sorting. Immunol. Rev. 172, 171±187. Barrett, A. J., Kembhavi, A. A., Brown, M. A., Kirschke, H., Knight, C. G., Tamai, M., et al. (1982). l-trans-Epoxysuccinyl-leucylamido(4-guanidino)butane (E-64) and its analogues as inhibitors of cysteine proteinases including cathepsins B, H and L. Biochem. J. 201, 189±198. Benaroch, P., Yilla, M., Raposo, G., Ito, K., Miwa, K., Geuze, H. J., et al. (1995). How MHC class II molecules reach the endocytic pathway. EMBO J. 14, 37±49. Bennett, K., Levine, T., Ellis, J. S., Peanasky, R. J., Samloff, I. M., Kay, J., et al. (1992). Antigen processing for presentation by class II major histocompatibility complex requires cleavage by cathepsin E. Eur. J. Immunol. 22, 1519 ±1524. Bevec, T., Stoka, V., Pungercic, G., Dolenc, I., and Turk, V. (1996). Major histocompatibility complex class II-associated p41 invariant chain fragment is a strong inhibitor of lysosomal cathepsin L. J. Exp. Med. 183, 1331±1338. Bjork, I., and Ylinenjarvi, K. (1990). Interaction between chicken cystatin and the cysteine proteinases actinidin, chymopapain A, and ®cin. Biochemistry 29, 1770±1776. Blum, J. S., and Cresswell, P. (1988). Role for intracellular proteases in the processing and transport of class II HLA antigens. Proc. Natl. Acad. Sci. USA 85, 3975±3979. Bogyo, M., Verhelst, S., Bellingard-Dubouchaud, V., Toba, S., and Greenbaum, D. (2000). Selective targeting of lysosomal cysteine proteases with radiolabeled electrophilic substrate analogs. Chem. Biol. 7, 27±38.

104

paula wolf bryant et al.

Bowman, E. J., Siebers, A., and Altendorf, K. (1988). Ba®lomycins: A class of inhibitors of membrane ATPases from microorganisms, animal cells, and plant cells. Proc. Natl. Acad. Sci. USA 85, 7972±7976. Brachet, V., Pehau-Arnaudet, G., Desaymard, C., Raposo, G., and Amigorena, S. (1999). Early endosomes are required for major histocompatibility complex class II transport to peptideloading compartments. Mol. Biol. Cell 10, 2891±2904. Brachet, V., Raposo, G., Amigorena, S., and Mellman, I. (1997). Ii chain controls the transport of major histocompatibility complex class II molecules to and from lysosomes. J. Cell Biol. 137, 51±65. Bremnes, B., Madsen, T., Gedde-Dahl, M., and Bakke, O. (1994). An LI and ML motif in the cytoplasmic tail of the MHC-associated invariant chain mediate rapid internalization. J. Cell Sci. 107, 2021±2032. Bromme, D., and McGrath, M. E. (1996). High level expression and crystallization of recombinant human cathepsin S. Protein Sci. 5, 789±791. Bromme, D., Steinert, A., Friebe, S., Fittkau, S., Wiederanders, B., and Kirschke, H. (1989). The speci®city of bovine spleen cathepsin S: A comparison with rat liver cathepsins L and B. Biochem. J. 264, 475±481. Brown, W. M., and Dziegielewska, K. M. (1997). Friends and relations of the cystatin superfamilyÐ New members and their evolution. Protein Sci. 6, 5±12. Bryant, P. W., Roos, P., Ploegh, H. L., and Sant, A. J. (1999). Deviant traf®cking of I-Ad mutant molecules is re¯ected in their peptide binding properties. Eur. J. Immunol. 29, 2729±2739. Busch, R., Doebele, R. C., Patil, N. S., Pashine, A., and Mellins, E. D. (2000). Accessory molecules for MHC class II peptide loading [see Comments]. Curr. Opin. Immunol. 12, 99±106. Buttle, D. J., Murata, M., Knight, C. G., and Barrett, A. J. (1992). CA074 methyl ester: A proinhibitor for intracellular cathepsin B. Arch. Biochem. Biophys. 299, 377±380. Castellino, F., and Germain, R. N. (1995). Extensive traf®cking of MHC class II-invariant chain complexes in the endocytic pathway and appearance of peptide-loaded class II in multiple compartments. Immunity 2, 73±88. Cella, M., Sallusto, F., and Lanzavecchia, A. (1997). Origin, maturation and antigen presenting function of dendritic cells. Curr. Opin. Immunol. 9, 10±16. Chapman, H. A. (1998). Endosomal proteolysis and MHC class II function. Curr. Opin. Immunol. 10, 93±102. Chapman, H. A., Riese, R. J., and Shi, G. -P. (1997). Emerging roles for cysteine proteases in human biology. Annu. Rev. Physiol. 59, 63±88. Collins, A. R., and Grubb, A. (1998). Cystatin D, a natural salivary cysteine protease inhibitor, inhibits coronavirus replication at its physiologic concentration. Oral Microbiol. Immunol. 13, 59±61. Collins, D. S., Unanue, E. R., and Harding, C. V. (1991). Reduction of disul®de bonds within lysosomes is a key step in antigen processing. J. Immunol. 147, 4054±4059. Coulombe, R., Grochulski, P., Sivaraman, J., Menard, R., Mort, J. S., and Cygler, M. (1996). Structure of human procathepsin L reveals the molecular basis of inhibition by the prosegment. EMBO J. 15, 5492±5503. Crawford, C., Mason, R. W., Wikstrom, P., and Shaw, E. (1988). The design of peptidyldiazomethane inhibitors to distinguish between the cysteine proteinases calpain II, cathepsin L and cathepsin B. Biochem. J. 253, 751±758. Cresswell, P. (1996). Invariant chain structure and MHC class II function. Cell 84, 505±507. Cygler, M., and Mort, J. S. (1997). Proregion structure of members of the papain superfamily: Mode of inhibition of enzymatic activity. Biochimie 79, 645±652. de Baey, A., and Lanzavecchia, A. (2000). The role of aquaporins in dendritic cell macropinocytosis. J. Exp. Med. 191, 743±748.

antigen presentation by mhc class ii molecules

105

Denzin, L. K., and Cresswell, P. (1995). HLA-DM induces CLIP dissociation from MHC class II alpha beta dimers and facilitates peptide loading. Cell 82, 155±165. Denzin, L. K., Hammond, C., and Cresswell, P. (1996). HLA±DM interactions with intermediates in HLA±DR maturation and a role for HLA±DM in stabilizing empty HLA±DR molecules. J. Exp. Med. 184, 2153±2165. Deussing, J., Roth, W., Saftig, P., Peters, C., Ploegh, H. L., and Villadangos, J. A. (1998). Cathepsins B and D are dispensable for major histocompatibility complex class II-mediated antigen presentation. Proc. Natl. Acad. Sci. USA 95, 4516±4521. Doebele, R. C., Busch, R., Scott, H. M., Pashine, A., and Mellins, E. D. (2000). Determination of the HLA±DM interaction site on HLA±DR molecules. Immunity 13, 517±527. Drakesmith, H., O'Neil, D., Schneider, S. C., Binks, M., Medd, P., Sercarz, E., et al. (1998). In vivo priming of T cells against cryptic determinants by dendritic cells exposed to interleukin 6 and native antigen. Proc. Natl. Acad. Sci. USA 95, 14903±14908. Driessen, C., Bryant, R. A., Lennon-Dumenil, A. M., Villadangos, J. A., Bryant, P. W., Shi, G. P., et al. (1999). Cathepsin S controls the traf®cking and maturation of MHC class II molecules in dendritic cells. J. Cell Biol. 147, 775±790. Driessen, C., Lennon-Dumenil, A. M., and Ploegh, H. L. (2001). Individual cathepsins degrade immune complexes internalized by antigen-presenting cells via Fcgamma receptors. Eur. J. Immunol. 31, 1592±1601. Engering, A. J., Cella, M., Fluitsma, D., Brockhaus, M., Hoefsmit, E. C., Lanzavecchia, A., et al. (1997). The mannose receptor functions as a high capacity and broad speci®city antigen receptor in human dendritic cells. Eur. J. Immunol. 27, 2417±2425. Erickson, A. H. (1989). Biosynthesis of lysosomal endopeptidases. J. Cell. Biochem. 40, 31±41. Felbor, U., Dreier, L., Bryant, R. A., Ploegh, H. L., Olsen, B. R., and Mothes, W. (2000). Secreted cathepsin L generates endostatin from collagen XVIII. EMBO J. 19, 1187±1194. Fiebiger, E., Meraner, P., Weber, E., Fang, I.-F., Stingl, G., Ploegh, H., et al. (2001). Cytokines regulate proteolysis in major histocompatibility complex class II-dependent antigen presentation by dendritic cells. J. Exp. Med. 193, 881±892. Fineschi, B., Arneson, L. S., Naujokas, M. F., and Miller, J. (1995). Proteolysis of major histocompatibility complex class II-associated invariant chain is regulated by the alternatively spliced gene product, p41. Proc. Natl. Acad. Sci. USA 92, 10257±10261. Fineschi, B., Sakaguchi, K., Appella, E., and Miller, J. (1996). The proteolytic environment involved in MHC class II-restricted antigen presentation can be modulated by the p41 form of invariant chain. J. Immunol. 157, 3211±3215. Fuchs, R., Schmid, S., and Mellman, I. (1989). A possible role for Na‡, K‡-ATPase in regulating ATP-dependent endosome acidi®cation. Proc. Natl. Acad. Sci. USA 86, 539±543. Fujishima, A., Imai, Y., Nomura, T., Fujisawa, Y., Yamamoto, Y., and Sugawara, T. (1997). The crystal structure of human cathepsin L complexed with E-64. FEBS Lett. 407, 47±50. Fung-Leung, W. P., Surh, C. D., Liljedahl, M., Pang, J., Leturcq, D., Peterson, P. A., et al. (1996). Antigen presentation and T cell development in H2-M-de®cient mice. Science 271, 1278±1281. Gedde-Dahl, M., Freisewinkel, I., Staschewski, M., Schenck, K., Koch, N., and Bakke, O. (1997). Exon 6 is essential for invariant chain trimerization and induction of large endosomal structures. J. Biol. Chem. 272, 8281±8287. Ghosh, P., Amaya, M., Mellins, E., and Wiley, D. C. (1995). The structure of an intermediate in class II MHC maturation: CLIP bound to HLA±DR3. Nature 378, 457±462. Gieselmann, V., Pohlmann, R., Hasilik, A., and Von Figura, K. (1983). Biosynthesis and transport of cathepsin D in cultured human ®broblasts. J. Cell Biol. 97, 1±5. Gorvel, J. P., Escola, J. M., Stang, E., and Bakke, O. (1995). Invariant chain induces a delayed transport from early to late endosomes. J. Biol. Chem. 270, 2741±2746.

106

paula wolf bryant et al.

Greenbaum, D., Medzihradszky, K. F., Burlingame, A., and Bogyo, M. (2000). Epoxide electrophiles as activity-dependent cysteine protease pro®ling and discovery tools. Chem. Biol. 7, 569±581. Gregory, C. D. (2000). CD14-dependent clearance of apoptotic cells: Relevance to the immune system. Curr. Opin. Immunol. 12, 27±34. Guerra, C. B., Busch, R., Doebele, R. C., Liu, W., Sawada, T., Kwok, W. W., et al. (1998). Novel glycosylation of HLA-DRalpha disrupts antigen presentation without altering endosomal localization. J. Immunol. 160, 4289±4297. Guncar, G., Podobnik, M., Pungercar, J., Strukelj, B., Turk, V., and Turk, D. (1998). Crystal structure of porcine cathepsin H determined at 2.1 AÊ resolution: Location of the minichain C-terminal carboxyl group de®nes cathepsin H aminopeptidase function. Structure 6, 51±61. Guncar, G., Pungercic, G., Klemencic, I., Turk, V., and Turk, D. (1999). Crystal structure of MHC class II-associated p41 Ii fragment bound to cathepsin L reveals the structural basis for differentiation between cathepsins L and S. EMBO J. 18, 793±803. Halfon, S., Ford, J., Foster, J., Dowling, L., Lucian, L., Sterling, M., et al. (1998). Leukocystatin, a new class II cystatin expressed selectively by hematopoietic cells. J. Biol. Chem. 273, 16400± 16408. Harding, C. V. (1996). Class II antigen processing: Analysis of compartments and functions. Crit. Rev. Immunol. 16, 13±29. Hofmann, M. W., Honing, S., Rodionov, D., Dobberstein, B., von Figura, K., and Bakke, O. (1999). The leucine-based sorting motifs in the cytoplasmic domain of the invariant chain are recognized by the clathrin adaptors AP1 and AP2 and their medium chains. J. Biol. Chem. 274, 36153±36158. Honey, K., Duff, M., Beers, C., Brissette, W. H., Elliott, E. A., Peters, C., et al. (2001). Cathepsin S regulates the expression of cathepsin L and the turnover of GILT in B lymphocytes. J. Biol. Chem. 16, 16. Huh, C. G., Hakansson, K., Nathanson, C. M., Thorgeirsson, U. P., Jonsson, N., Grubb, A., et al. (1999). Decreased metastatic spread in mice homozygous for a null allele of the cystatin C protease inhibitor gene. Mol. Pathol. 52, 332±340. Illy, C., Quraishi, O., Wang, J., Purisima, E., Vernet, T., and Mort, J. S. (1997). Role of the occluding loop in cathepsin B activity. J. Biol. Chem. 272, 1197±1202. Inaba, K., Turley, S., Iyoda, T., Yamaide, F., Shimoyama, S., Reis e Sousa, C., et al. (2000). The formation of immunogenic major histocompatibility complex class II±peptide ligands in lysosomal compartments of dendritic cells is regulated by in¯ammatory stimuli. J. Exp. Med. 191, 927±936. Ishidoh, K., and Kominami, E. (1998). Gene regulation and extracellular functions of procathepsin L. Biol. Chem. 379, 131±135. Jensen, P. E. (1991). Reduction of disul®de bonds during antigen processing: Evidence from a thioldependent insulin determinant. J. Exp. Med. 174, 1121±1130. Jensen, P. E. (1993). Acidi®cation and disul®de reduction can be suf®cient to allow intact proteins to bind class II MHC. J. Immunol. 150, 3347±3356. Jerala, R., Zerovnik, E., Kidric, J., and Turk, V. (1998). pH-induced conformational transitions of the propeptide of human cathepsin L: A role for a molten globule state in zymogen activation. J. Biol. Chem. 273, 11498±11504. Jiang, W., Swiggard, W. J., Heu¯er, C., Peng, M., Mirza, A., Steinman, R. M., et al. (1995). The receptor DEC-205 expressed by dendritic cells and thymic epithelial cells is involved in antigen processing. Nature 375, 151±155. Kampgen, E., Koch, N., Koch, F., Stoger, P., Heu¯er, C., Schuler, G., et al. (1991). Class II major histocompatibility complex molecules of murine dendritic cells: Synthesis, sialylation of invariant

antigen presentation by mhc class ii molecules

107

chain, and antigen processing capacity are down-regulated upon culture. Proc. Natl. Acad. Sci. USA 88, 3014±3018. Kasper, D., Dittmer, F., von Figura, K., and Pohlmann, R. (1996). Neither type of mannose 6phosphate receptor is suf®cient for targeting of lysosomal enzymes along intracellular routes. J. Cell. Biol. 134, 615±623. Katunuma, N., Murata, E., Kakegawa, H., Matsui, A., Tsuzuki, H., Tsuge, H., et al. (1999). Structure based development of novel speci®c inhibitors for cathepsin L and cathepsin S in vitro and in vivo. FEBS Lett. 458, 6±10. Koch, N. (1988). Posttranslational modi®cations of the Ia-associated invariant protein p41 after gene transfer. Biochemistry 27, 4097±4102. Koch, N., and Harris, A. W. (1984). Differential expression of the invariant chain in mouse tumor cells: Relationship to B lymphoid development. J. Immunol. 132, 12±15. Korant, B. D., Strack, P., Frey, M. W., and Rizzo, C. J. (1998). A cellular anti-apoptosis protein is cleaved by the HIV-1 protease. Adv. Exp. Med. Biol. 436, 27±29. Koster, A., Saftig, P., Matzner, U., von Figura, K., Peters, C., and Pohlmann, R. (1993). Targeted disruption of the M(r) 46,000 mannose 6-phosphate receptor gene in mice results in misrouting of lysosomal proteins. EMBO J. 12, 5219±5223. Kropshofer, H., Hammerling, G. J., and Vogt, A. B. (1999). The impact of the non-classical MHC proteins HLA-DM and HLA-DO on loading of MHC class II molecules. Immunol. Rev. 172, 267±278. Lafuse, W. P., Brown, D., Castle, L., and Zwilling, B. S. (1995). IFN-gamma increases cathepsin H mRNA levels in mouse macrophages. J. Leukoc. Biol. 57, 663±669. LaLonde, J. M., Zhao, B., Janson, C. A., D'Alessio, K. J., McQueney, M. S., Orsini, M. J., et al. (1999). The crystal structure of human procathepsin K. Biochemistry 38, 862±869. Larsen, S. L., Pedersen, L. O., Buus, S., and Stryhn, A. (1996). T cell responses affected by aminopeptidase N (CD13)-mediated trimming of major histocompatibility complex class IIbound peptides. J. Exp. Med. 184, 183±189. Lennon-Dumenil, A. M., Roberts, R. A., Valentijn, K., Driesseu, C., Overkleeft, H. S., Erickson, A., Peters, P. J., Bikoff, E., Ploegh, H. L., and Wolf Bryant, P. (2001). The p41 isoform of invariant chain is a chaperon for cathepsin L. EMBO J. 20, 4055±4064. Li, Q., and Bever, C. T., Jr. (1998). Modulation of interferon gamma induced increases in cathepsin B in THP-1 cells by adrenergic agonists and antagonists. Cell Biol. Int. 22, 13±20. Lindahl, P., Abrahamson, M., and Bjork, I. (1992). Interaction of recombinant human cystatin C with the cysteine proteinases papain and actinidin. Biochem. J. 281, 49±55. Lingeman, R. G., Joy, D. S., Sherman, M. A., and Kane, S. E. (1998). Effect of carbohydrate position on lysosomal transport of procathepsin L. Mol. Biol. Cell 9, 1135±1147. Liuzzo, J. P., Petanceska, S. S., Moscatelli, D., and Devi, L. A. (1999). In¯ammatory mediators regulate cathepsin S in macrophages and microglia: A role in attenuating heparan sulfate interactions. Mol. Med. 5, 320±333. Lotteau, V., Teyton, L., Peleraux, A., Nilsson, T., Karlsson, L., Schmid, S. L., et al. (1990). Intracellular transport of class II MHC molecules directed by invariant chain. Nature 348, 600±605. Mahnke, K., Guo, M., Lee, S., Sepulveda, H., Swain, S. L., Nussenzweig, M., et al. (2000). The dendritic cell receptor for endocytosis, DEC-205, can recycle and enhance antigen presentation via major histocompatibility complex class II-positive lysosomal compartments. J. Cell Biol. 151, 673±684. Manoury, B., Gregory, W. F., Maizels, R. M., and Watts, C. (2001). Bm-CPI-2, a cystatin homolog secreted by the ®larial parasite Brugia malayi, inhibits class II MHC-restricted antigen processing. Curr. Biol. 11, 447±451.

108

paula wolf bryant et al.

Manoury, B., Hewitt, E. W., Morrice, N., Dando, P. M., Barrett, A. J., and Watts, C. (1998). An asparaginyl endopeptidase processes a microbial antigen for class II MHC presentation. Nature 396, 695±699. Maric, M. A., Taylor, M. D., and Blum, J. S. (1994). Endosomal aspartic proteinases are required for invariant-chain processing. Proc. Natl. Acad. Sci. USA 91, 2171±2175. Martin, W. D., Hicks, G. G., Mendiratta, S. K., Leva, H. I., Ruley, H. E., and Van Kaer, L. (1996). H2-M mutant mice are defective in the peptide loading of class II molecules, antigen presentation, and T cell repertoire selection. Cell 84, 543±550. Mason, R. W., Bartholomew, L. T., and Hardwick, B. S. (1989a). The use of benzyloxycarbonyl [125 I]iodotyrosylalanyldiazomethane as a probe for active cysteine proteinases in human tissues. Biochem. J. 263, 945±949. Mason, R. W., Wilcox, D., Wikstrom, P., and Shaw, E. N. (1989b). The identi®cation of active forms of cysteine proteinases in Kirsten-virus-transformed mouse ®broblasts by use of a speci®c radiolabelled inhibitor. Biochem. J. 257, 125±129. Maurer, D., Ebner, C., Reininger, B., Fiebiger, E., Kraft, D., Kinet, J. P., et al. (1995). The high af®nity IgE receptor (Fc epsilon RI) mediates IgE-dependent allergen presentation. J. Immunol. 154, 6285±6290. Maurer, D., Fiebiger, E., Reininger, B., Ebner, C., Petzelbauer, P., Shi, G. P., et al. (1998). Fc epsilon receptor I on dendritic cells delivers IgE-bound multivalent antigens into a cathepsin S-dependent pathway of MHC class II presentation. J. Immunol. 161, 2731±2739. Maurer, D., Fiebiger, E., Reininger, B., Wolff-Winiski, B., Jouvin, M. H., Kilgus, O., et al. (1994). Expression of functional high af®nity immunoglobulin E receptors (Fc epsilon RI) on monocytes of atopic individuals. J. Exp. Med. 179, 745±750. McGrath, M. E. (1999). The lysosomal cysteine proteases. Annu. Rev. Biophys. Biomol. Struct. 28, 181±204. McGrath, M. E., Klaus, J. L., Barnes, M. G., and Bromme, D. (1997). Crystal structure of human cathepsin K complexed with a potent inhibitor. Nat. Struct. Biol. 4, 105±109. McGrath, M. E., Palmer, J. T., Bromme, D., and Somoza, J. R. (1998). Crystal structure of human cathepsin S. Protein Sci. 7, 1294±1302. Mellman, I., Fuchs, R., and Helenius, A. (1986). Acidi®cation of the endocytic and exocytic pathways. Annu. Rev. Biochem. 55, 663±700. Miyazaki, T., Wolf, P., Tourne, S., Waltzinger, C., Dierich, A., Barois, N., et al. (1996). Mice lacking H2-M complexes, enigmatic elements of the MHC class II peptide-loading pathway. Cell 84, 531±541. Mizuochi, T., Yee, S. T., Kasai, M., Kakiuchi, T., Muno, D., and Kominami, E. (1994). Both cathepsin B and cathepsin D are necessary for processing of ovalbumin as well as for degradation of class II MHC invariant chain. Immunol. Lett. 43, 189±193. Mort, J. S., and Buttle, D. J. (1997). Cathepsin B. Int. J. Biochem. Cell Biol. 29, 715±720. Morton, P. A., Zacheis, M. L., Giacoletto, K. S., Manning, J. A., and Schwartz, B. D. (1995). Delivery of nascent MHC class II-invariant chain complexes to lysosomal compartments and proteolysis of invariant chain by cysteine proteases precedes peptide binding in B-lymphoblastoid cells. J. Immunol. 154, 137±150. Mosyak, L., Zaller, D. M., and Wiley, D. C. (1998). The structure of HLA±DM, the peptide exchange catalyst that loads antigen onto class II MHC molecules during antigen presentation. Immunity 9, 377±383. Murata, M., Miyashita, S., Yokoo, C., Tamai, M., Hanada, K., Hatayama, K., et al. (1991). Novel epoxysuccinyl peptides: Selective inhibitors of cathepsin B, in vitro. FEBS Lett. 280, 307±310. Musil, D., Zucic, D., Turk, D., Engh, R. A., Mayr, I., Huber, R., et al. (1991). The re®ned 2.15 AÊ X-ray crystal structure of human liver cathepsin B: the structural basis for its speci®city. EMBO J. 10, 2321±2330.

antigen presentation by mhc class ii molecules

109

Nagler, D. K., Storer, A. C., Portaro, F. C., Carmona, E., Juliano, L., and Menard, R. (1997). Major increase in endopeptidase activity of human cathepsin B upon removal of occluding loop contacts. Biochemistry 36, 12608±12615. Nagler, D. K., Sulea, T., and Menard, R. (1999). Full-length cDNA of human cathepsin F predicts the presence of a cystatin domain at the N-terminus of the cysteine protease zymogen. Biochem. Biophys. Res. Commun. 257, 313±318. Nakagawa, T., Roth, W., Wong, P., Nelson, A., Farr, A., Deussing, J., et al. (1998). Cathepsin L: critical role in Ii degradation and CD4 T cell selection in the thymus. Science 280, 450±453. Nakagawa, T. Y., Brissette, W. H., Lira, P. D., Grif®ths, R. J., Petrushova, N., Stock, J., et al. (1999). Impaired invariant chain degradation and antigen presentation and diminished collagen-induced arthritis in cathepsin S null mice. Immunity 10, 207±217. Nakagawa, T. Y., and Rudensky, A. Y. (1999). The role of lysosomal proteinases in MHC class IImediated antigen processing and presentation. Immunol. Rev. 172, 121±129. Neefjes, J. J., and Ploegh, H. L. (1992). Inhibition of endosomal proteolytic activity by leupeptin blocks surface expression of MHC class II molecules and their conversion to SDS resistance alpha beta heterodimers in endosomes. EMBO J. 11, 411±416. Neefjes, J. J., Stollorz, V., Peters, P. J., Geuze, H. J., and Ploegh, H. L. (1990). The biosynthetic pathway of MHC class II but not class I molecules intersects the endocytic route. Cell 61, 171±183. Nelson, N. (1992). The vacuolar H(‡)-ATP-aseÐOne of the most fundamental ion pumps in nature. J. Exp. Biol. 172, 19±27. Newmyer, S. L., and Schmid, S. L. (2001). Dominant-interfering hsc70 mutants disrupt multiple stages of the clathrin coated vesicles cycle in vivo. J. Cell Biol., in press. Nycander, M., Estrada, S., Mort, J. S., Abrahamson, M., and Bjork, I. (1998). Two-step mechanism of inhibition of cathepsin B by cystatin C due to displacement of the proteinase occluding loop. FEBS Lett. 422, 61±64. Odorizzi, C. G., Trowbridge, I. S., Xue, L., Hopkins, C. R., Davis, C. D., and Collawn, J. F. (1994). Sorting signals in the MHC class II invariant chain cytoplasmic tail and transmembrane region determine traf®cking to an endocytic processing compartment. J. Cell Biol. 126, 317±330. Ogrinc, T., Dolenc, I., Ritonja, A., and Turk, V. (1993). Puri®cation of the complex of cathepsin L and the MHC class II-associated invariant chain fragment from human kidney. FEBS Lett. 336, 555±559. Ojcius, D. M., Gachelin, G., and Dautry-Varsat, A. (1996). Presentation of antigens derived from microorganisms residing in host-cell vacuoles. Trends Microbiol. 4, 53±59. O'Sullivan, D. M., Noonan, D., and Quaranta, V. (1987). Four Ia invariant chain forms derive from a single gene by alternate splicing and alternate initiation of transcription/translation. J. Exp. Med. 166, 444±460. Palmer, J. T., Rasnick, D., Klaus, J. L., and Bromme, D. (1995). Vinyl sulfones as mechanism-based cysteine protease inhibitors. J. Med. Chem. 38, 3193±3196. Panjwani, N., Akbari, O., Garcia, S., Brazil, M., and Stockinger, B. (1999). The HSC73 molecular chaperone: Involvement in MHC class II antigen presentation. J. Immunol. 163, 1936±1942. Pavlova, A., Krupa, J. C., Mort, J. S., Abrahamson, M., and Bjork, I. (2000). Cystatin inhibition of cathepsin B requires dislocation of the proteinase occluding loop: Demonstration by release of loop anchoring through mutation of his110. FEBS Lett. 487, 156±160. Peterson, M., and Miller, J. (1992). Antigen presentation enhanced by the alternatively spliced invariant chain gene product p41. Nature 357, 596±598. Phan, U. T., Arunachalam, B., and Cresswell, P. (2000). Gamma-interferon-inducible lysosomal thiol reductase (GILT). Maturation, activity, and mechanism of action. J. Biol. Chem. 275, 25907±25914.

110

paula wolf bryant et al.

Pierre, P., and Mellman, I. (1998). Developmental regulation of invariant chain proteolysis controls MHC class II traf®cking in mouse dendritic cells. Cell 93, 1135±1145. Pieters, J., Bakke, O., and Dobberstein, B. (1993). The MHC class II-associated invariant chain contains two endosomal targeting signals within its cytoplasmic tail. J. Cell Sci. 106, 831±846. Pinet, V. M., and Long, E. O. (1998). Peptide loading onto recycling HLA±DR molecules occurs in early endosomes. Eur. J. Immunol. 28, 799±804. Pliura, D. H., Bonaventura, B. J., Smith, R. A., Coles, P. J., and Krantz, A. (1992). Comparative behaviour of calpain and cathepsin B toward peptidyl acyloxymethyl ketones, sulphonium methyl ketones and other potential inhibitors of cysteine proteinases. Biochem. J. 288, 759±762. Podobnik, M., Kuhelj, R., Turk, V., and Turk, D. (1997). Crystal structure of the wild-type human procathepsin B at 2.5 AÊ resolution reveals the native active site of a papain-like cysteine protease zymogen. J. Mol. Biol. 271, 774±788. Pohlmann, R., Boeker, M. W., and von Figura, K. (1995). The two mannose 6-phosphate receptors transport distinct complements of lysosomal proteins. J. Biol. Chem. 270, 27311±27318. Pol, E., Olsson, S. L., Estrada, S., Prasthofer, T. W., and Bjork, I. (1995). Characterization by spectroscopic, kinetic and equilibrium methods of the interaction between recombinant human cystatin A (ste®n A) and cysteine proteinases. Biochem. J. 311, 275±282. Pond, L., and Watts, C. (1999). Functional early endosomes are required for maturation of major histocompatibility complex class II molecules in human B lymphoblastoid cells. J. Biol. Chem. 274, 18049±18054. Prigozy, T. I., Sieling, P. A., Clemens, D., Stewart, P. L., Behar, S. M., Porcelli, S. A., et al. (1997). The mannose receptor delivers lipoglycan antigens to endosomes for presentation to T cells by CD1b molecules. Immunity 6, 187±197. Punturieri, A., Filippov, S., Allen, E., Caras, I., Murray, R., Reddy, V., et al. (2000). Regulation of elastinolytic cysteine proteinase activity in normal and cathepsin K-de®cient human macrophages. J. Exp. Med. 192, 789±799. Pure, E., Inaba, K., Crowley, M. T., Tardelli, L., Witmer-Pack, M. D., Ruberti, G., et al. (1990). Antigen processing by epidermal Langerhans cells correlates with the level of biosynthesis of major histocompatibility complex class II molecules and expression of invariant chain. J. Exp. Med. 172, 1459±1469. Quraishi, O., Nagler, D. K., Fox, T., Sivaraman, J., Cygler, M., Mort, J. S., et al. (1999). The occluding loop in cathepsin B de®nes the pH dependence of inhibition by its propeptide. Biochemistry 38, 5017±5023. Reilly, J. J., Jr., Mason, R. W., Chen, P., Joseph, L. J., Sukhatme, V. P., Yee, R., et al. (1989). Synthesis and processing of cathepsin L, an elastase, by human alveolar macrophages. Biochem. J. 257, 493±498. Reyes, V. E., Lu, S., and Humphreys, R. E. (1991). Cathepsin B cleavage of Ii from class II MHC alpha- and beta-chains. J. Immunol. 146, 3877±3880. Riese, R. J., and Chapman, H. A. (2000). Cathepsins and compartmentalization in antigen presentation. Curr. Opin. Immunol. 12, 107±113. Riese, R. J., Wolf, P. R., Bromme, D., Natkin, L. R., Villadangos, J. A., Ploegh, H. L., et al. (1996). Essential role for cathepsin S in MHC class II-associated invariant chain processing and peptide loading. Immunity 4, 357±366. Roche, P. A., and Cresswell, P. (1991). Proteolysis of the class II-associated invariant chain generates a peptide binding site in intracellular HLA-DR molecules. Proc. Natl. Acad. Sci. USA 88, 3150± 3154. Roche, P. A., Teletski, C. L., Stang, E., Bakke, O., and Long, E. O. (1993). Cell surface HLA-DRinvariant chain complexes are targeted to endosomes by rapid internalization. Proc. Natl. Acad. Sci. USA 90, 8581±8585.

antigen presentation by mhc class ii molecules

111

Rodriguez, G. M., and Diment, S. (1992). Role of cathepsin D in antigen presentation of ovalbumin. J. Immunol. 149, 2894±2898. Rodriguez, G. M., and Diment, S. (1995). Destructive proteolysis by cysteine proteases in antigen presentation of ovalbumin. Eur. J. Immunol. 25, 1823±1827. Romagnoli, P., Layet, C., Yewdell, J., Bakke, O., and Germain, R. N. (1993). Relationship between invariant chain expression and major histocompatibility complex class II transport into early and late endocytic compartments. J. Exp. Med. 177, 583±596. Roth, W., Deussing, J., Botchkarev, V. A., Pauly-Evers, M., Saftig, P., Hafner, A., et al. (2000). Cathepsin L de®ciency as molecular defect of furless: Hyperproliferation of keratinocytes and pertubation of hair follicle cycling. Faseb J. 14, 2075±2086. Sadegh-Nasseri, S., and Germain, R. N. (1991). A role for peptide in determining MHC class II structure. Nature 353, 167±170. Saftig, P., Hetman, M., Schmahl, W., Weber, K., Heine, L., Mossmann, H., et al. (1995). Mice de®cient for the lysosomal proteinase cathepsin D exhibit progressive atrophy of the intestinal mucosa and profound destruction of lymphoid cells. EMBO J. 14, 3599±3608. Sallusto, F., Cella, M., Danieli, C., and Lanzavecchia, A. (1995). Dendritic cells use macropinocytosis and the mannose receptor to concentrate macromolecules in the major histocompatibility complex class II compartment: Downregulation by cytokines and bacterial products. J. Exp. Med. 182, 389±400. Santambrogio, L., Sato, A. K., Carven, G. J., Belyanskaya, S. L., Strominger, J. L., and Stern, L. J. (1999a). Extracellular antigen processing and presentation by immature dendritic cells. Proc. Natl. Acad. Sci. USA 96, 15056±15061. Santambrogio, L., Sato, A. K., Fischer, F. R., Dorf, M. E., and Stern, L. J. (1999b). Abundant empty class II MHC molecules on the surface of immature dendritic cells. Proc. Natl. Acad. Sci. USA 96, 15050±15055. Schlossman, D. M., Schmid, S. L., Braell, W. A., and Rothman, J. E. (1984). An enzyme that removes clathrin coats: Puri®cation of an uncoating ATPase. J. Cell Biol. 99, 723±733. Sealy, L., Mota, F., Rayment, N., Tatnell, P., Kay, J., and Chain, B. (1996). Regulation of cathepsin E expression during human B cell differentiation in vitro. Eur. J. Immunol. 26, 1838±1843. Shaw, E. (1994). Peptidyl diazomethanes as inhibitors of cysteine and serine proteinases. Methods Enzymol. 244, 649±656. Shaw, E., Angliker, H., Rauber, P., Walker, B., and Wikstrom, P. (1986). Peptidyl ¯uoromethyl ketones as thiol protease inhibitors. Biomed. Biochim. Acta 45, 1397±1403. Shi, G. P., Bryant, R. A., Riese, R., Verhelst, S., Driessen, C., Li, Z., et al. (2000). Role for cathepsin F in invariant chain processing and major histocompatibility complex class II peptide loading by macrophages. J. Exp. Med. 191, 1177±1186. Shi, G. P., Munger, J. S., Meara, J. P., Rich, D. H., and Chapman, H. A. (1992). Molecular cloning and expression of human alveolar macrophage cathepsin S, an elastinolytic cysteine protease. J. Biol. Chem. 267, 7258±7262. Shi, G. P., Villadangos, J. A., Dranoff, G., Small, C., Gu, L., Haley, K. J., et al. (1999). Cathepsin S required for normal MHC class II peptide loading and germinal center development. Immunity 10, 197±206. Siemasko, K., and Clark, M. R. (2001). The control and facilitation of MHC class II antigen processing by the BCR. Curr. Opin. Immunol. 13, 32±36. Sotiropoulou, G., Anisowicz, A., and Sager, R. (1997). Identi®cation, cloning, and characterization of cystatin M, a novel cysteine proteinase inhibitor, down-regulated in breast cancer. J. Biol. Chem. 272, 903±910. Stoka, V., Nycander, M., Lenarcic, B., Labriola, C., Cazzulo, J. J., Bjork, I., et al. (1995). Inhibition of cruzipain, the major cysteine proteinase of the protozoan parasite, Trypanosoma cruzi, by proteinase inhibitors of the cystatin superfamily. FEBS Lett. 370, 101±104.

112

paula wolf bryant et al.

Strubin, M., Long, E. O., and Mach, B. (1986). Two forms of the Ia antigen-associated invariant chain result from alternative initiations at two in-phase AUGs. Cell 47, 619±625. Stubbs, M. T., Laber, B., Bode, W., Huber, R., Jerala, R., Lenarcic, B., et al. (1990). The re®ned 2.4 AÊ X-ray crystal structure of recombinant human ste®n B in complex with the cysteine proteinase papain: A novel type of proteinase inhibitor interaction. Embo J. 9, 1939±1947. Takaesu, N. T., Lower, J. A., Robertson, E. J., and Bikoff, E. K. (1995). Major histocompatibility class II peptide occupancy, antigen presentation, and CD4‡ T cell function in mice lacking the p41 isoform of invariant chain. Immunity 3, 385±396. Takaesu, N. T., Lower, J. A., Yelon, D., Robertson, E. J., and Bikoff, E. K. (1997). In vivo functions mediated by the p41 isoform of the MHC class II-associated invariant chain. J. Immunol. 158, 187±199. Tan, M. C., Mommaas, A. M., Drijfhout, J. W., Jordens, R., Onderwater, J. J., Verwoerd, D., et al. (1997). Mannose receptor-mediated uptake of antigens strongly enhances HLA class II-restricted antigen presentation by cultured dendritic cells. Eur. J. Immunol. 27, 2426±2435. Tao, K., Stearns, N. A., Dong, J., Wu, Q. L., and Sahagian, G. G. (1994). The proregion of cathepsin L is required for proper folding, stability, and ER exit. Arch. Biochem. Biophys. 311, 19±27. Tartakoff, A. M. (1983). Perturbation of vesicular traf®c with the carboxylic ionophore monensin. Cell 32, 1026±1028. Thery, C., and Amigorena, S. (2001). The cell biology of antigen presentation in dendritic cells. Curr. Opin. Immunol. 13, 45±51. Thorens, B., and Vassalli, P. (1986). Chloroquine and ammonium chloride prevent terminal glycosylation of immunoglobulins in plasma cells without affecting secretion. Nature 321, 618±620. Towatari, T., Nikawa, T., Murata, M., Yokoo, C., Tamai, M., Hanada, K., et al. (1991). Novel epoxysuccinyl peptides. A selective inhibitor of cathepsin B, in vivo. FEBS Lett. 280, 311±315. Tsuge, H., Nishimura, T., Tada, Y., Asao, T., Turk, D., Turk, V., et al. (1999). Inhibition mechanism of cathepsin L-speci®c inhibitors based on the crystal structure of papain±CLIK148 complex. Biochem. Biophys. Res. Commun. 266, 411±416. Turk, B., Dolenc, I., Zerovnik, E., Turk, D., Gubensek, F., and Turk, V. (1994). Human cathepsin B is a metastable enzyme stabilized by speci®c ionic interactions associated with the active site. Biochemistry 33, 14800±14806. Turk, B., Stoka, V., Bjork, I., Boudier, C., Johansson, G., Dolenc, I., et al. (1995a). High-af®nity binding of two molecules of cysteine proteinases to low-molecular-weight kininogen. Protein Sci. 4, 1874±1880. Turk, B., Stoka, V., Turk, V., Johansson, G., Cazzulo, J. J., and Bjork, I. (1996a). High-molecularweight kininogen binds two molecules of cysteine proteinases with different rate constants. FEBS Lett. 391, 109±112. Turk, B., Turk, D., and Turk, V. (2000). Lysosomal cysteine proteases: More than scavengers. Biochim. Biophys. Acta 1477, 98±111. Turk, B., Turk, V., and Turk, D. (1997). Structural and functional aspects of papain-like cysteine proteinases and their protein inhibitors. Biol. Chem. 378, 141±150. Turk, D., Guncar, G., Podobnik, M., and Turk, B. (1998). Revised de®nition of substrate binding sites of papain-like cysteine proteases. Biol. Chem. 379, 137±147. Turk, D., Guncar, G., and Turk, V. (1999). The p41 fragment story. IUBMB Life 48, 7±12. Turk, D., Podobnik, M., Kuhelj, R., Dolinar, M., and Turk, V. (1996b). Crystal structures of human procathepsin B at 3.2 and 3.3 angstroms resolution reveal an interaction motif between a papainlike cysteine protease and its propeptide. FEBS Lett. 384, 211±214. Turk, D., Podobnik, M., Popovic, T., Katunuma, N., Bode, W., Huber, R., et al. (1995b). Crystal structure of cathepsin B inhibited with CA030 at 2.0-AÊ resolution: A basis for the design of speci®c epoxysuccinyl inhibitors. Biochemistry 34, 4791±4797.

antigen presentation by mhc class ii molecules

113

Turk, V., and Bode, W. (1991). The cystatins: Protein inhibitors of cysteine proteinases. FEBS Lett. 285, 213±219. van Bergen, J., Ossendorp, F., Jordens, R., Mommaas, A. M., Drijfhout, J. W., and Koning, F. (1999). Get into the groove! Targeting antigens to MHC class II. Immunol. Rev. 172, 87±96. van Noort, J. M., and Jacobs, M. J. (1994). Cathepsin D, but not cathepsin B, releases T cell stimulatory fragments from lysozyme that are functional in the context of multiple murine class II MHC molecules. Eur. J. Immunol. 24, 2175±2180. Vernet, T., Berti, P. J., de Montigny, C., Musil, R., Tessier, D. C., Menard, R., et al. (1995). Processing of the papain precursor: The ionization state of a conserved amino acid motif within the Pro region participates in the regulation of intramolecular processing. J. Biol. Chem. 270, 10838±10846. Vidard, L., Rock, K. L., and Benacerraf, B. (1992). Diversity in MHC class II ovalbumin T cell epitopes generated by distinct proteases. J. Immunol. 149, 498±504. Villadangos, J. A., Bryant, R. A., Deussing, J., Driessen, C., Lennon-Dumenil, A. M., Riese, R. J., et al. (1999). Proteases involved in MHC class II antigen presentation. Immunol. Rev. 172, 109±120. Villadangos, J. A., and Ploegh, H. L. (2000). Proteolysis in MHC class II antigen presentation: Who's in charge? Immunity 12, 233±239. Villadangos, J. A., Riese, R. J., Peters, C., Chapman, H. A., and Ploegh, H. L. (1997). Degradation of mouse invariant chain: Roles of cathepsins S and D and the in¯uence of major histocompatibility complex polymorphism. J. Exp. Med. 186, 549±560. [Published erratum appears in J. Exp. Med. 1997, 186(11), 5945. Villadangos, J. A., Cardoso, M., Steptoe, R. J., van Berkel, D., Pooley, J., Carbone, F. R., and Shortman, K. (2001). MHC class II expression is regulated in dendritic cells independently of invariant chain degradation. Immunity 14, 739±749. Vogt, A. B., Kropshofer, H., Moldenhauer, G., and Hammerling, G. J. (1996). Kinetic analysis of peptide loading onto HLA±DR molecules mediated by HLA±DM. Proc. Natl. Acad. Sci. USA 93, 9724±9729. Wang, Z., Zheng, T., Zhu, Z., Homer, R. J., Riese, R. J., Chapman, H. A., Jr., et al. (2000). Interferon gamma induction of pulmonary emphysema in the adult murine lung. J. Exp. Med. 192, 1587±1600. Watts, C. (1997). Capture and processing of exogenous antigens for presentation on MHC molecules. Annu. Rev. Immunol. 15, 821±850. Watts, C. (2001). Antigen processing in the endocytic compartment. Curr. Opin. Immunol. 13, 26±31. Wex, T., Levy, B., Wex, H., and Bromme, D. (2000). Human cathepsins W and F form a new subgroup of cathepsins that is evolutionary separated from the cathepsin B- and L-like cysteine proteases. Adv. Exp. Med. Biol. 477, 271±280. Wolf, P. R., and Ploegh, H. L. (1995). How MHC class II molecules acquire peptide cargo: Biosynthesis and traf®cking through the endocytic pathway. Annu. Rev. Cell. Dev. Biol. 11, 267±306. Woo, J. T., Shinohara, C., Sakai, K., Hasumi, K., and Endo, A. (1992). Inhibition of the acidi®cation of endosomes and lysosomes by the antibiotic concanamycin B in macrophage J774. Eur. J. Biochem. 207, 383±389. Xing, R., Addington, A. K., and Mason, R. W. (1998). Quanti®cation of cathepsins B and L in cells. Biochem. J. 332, 499±505. Yoshimori, T., Yamamoto, A., Moriyama, Y., Futai, M., and Tashiro, Y. (1991). Ba®lomycin A1, a speci®c inhibitor of vacuolar-type H(‡)-ATPase, inhibits acidi®cation and protein degradation in lysosomes of cultured cells. J. Biol. Chem. 266, 17707±17712. Zhang, T., Maekawa, Y., Hanba, J., Dainichi, T., Nashed, B. F., Hisaeda, H., et al. (2000). Lysosomal cathepsin B plays an important role in antigen processing, while cathepsin D is

114

paula wolf bryant et al.

involved in degradation of the invariant chain inovalbumin-immunized mice. Immunology 100, 13±20. Zhao, B., Janson, C. A., Amegadzie, B. Y., D'Alessio, K., Grif®n, C., Hanning, C. R., et al. (1997). Crystal structure of human osteoclast cathepsin K complex with E-64. Nat. Struct. Biol. 4, 109±111. Zhong, G., Romagnoli, P., and Germain, R. N. (1997). Related leucine-based cytoplasmic targeting signals in invariant chain and major histocompatibility complex class II molecules control endocytic presentation of distinct determinants in a single protein. J. Exp. Med. 185, 429±438.

ADVANCES IN IMMUNOLOGY, VOL. 80

Cytokine Memory of T Helper Lymphocytes MAX LO«HNING, ANNE RICHTER, AND ANDREAS RADBRUCH Deutsches Rheumaforschungszentrum,10117 Berlin, Germany

Memory is one of the key features of the adaptive immune system. Speci®c T and B lymphocytes are primed for a particular antigen and upon challenge with it will react faster than naive lymphocytes. They also memorize the expression of key effector molecules, in particular cytokines, which determine the type and scale of an immune reaction. While in primary activations differential expression of cytokine genes is dependent on antigen±receptor signaling and differentiation signals, in later activations the expression is triggered by antigen±receptor signaling and dependent on the cytokine memory. The molecular basis of the cytokine memory implies differential expression of transcription factors and epigenetic modi®cations of cytokine genes and gene loci. GATA-3 for Th2 and T-bet for Th1 cells expressing interleukin-4 or interferon-g, respectively, are prime candidates for key transcription factors of cytokine memory. The essential role of epigenetic modi®cations is suggested by the requirement of DNA synthesis for the establishment of a cytokine memory in Th lymphocytes. At present the molecular link between transcription factors and epigenetic modi®cations of cytokine genes in the establishment and maintenance of cytokine memory is not clear. The initial cytokine memory is not stable against adverse differentiation signals, while in repeatedly stimulated lymphocytes it is stabilized by a variety of mechanisms. # 2002, Elsevier Science (USA)

I. Introduction Cytokine memory in T helper (Th) lymphocytes is established and maintained in the context of cellular differentiation programs, which are induced upon activation of a T cell via its antigen receptor. Th1 and Th2 programs have been described as two polar developmental programs of mature Th cells (Mosmann and Coffman, 1989). They are characterized by the expression of distinct sets of cytokine genes, but also by the differential expression of surface molecules (e.g., cytokine and chemokine receptors) and their downstream signaling molecules and transcription factors (reviewed in Glimcher and Murphy, 2000). As such, these cells comprise ef®cient units of concerted effector functions in de®ned types of immune reactions. Th1 cells predominantly induce in¯ammatory immune reactions, which are effective in the control of intracellular pathogens. They are also associated with certain autoimmune diseases. Th2 cells support ef®cient responses to extracellular pathogens. They are also central mediators of asthma and allergy (reviewed in Abbas et al., 1996; Romagnani, 1994). 115 Copyright 2002, Elsevier Science (USA). All rights reserved. 0065-2776/02 $35.00

116

max loÈhning et al.

Cytokines of the Th1 program include interferon-g (IFN-g) and tumor necrosis factor-b (TNF-b), while interleukin-4 (IL-4), IL-5, and IL-13 are canonical Th2 cytokines (Mosmann and Coffman, 1989; Mosmann et al., 1986). In a given restimulation with antigen, however, individual T cells can produce cytokines of one or the other subset in diverse combinations (Assenmacher et al., 1994; Bucy et al., 1994; Openshaw et al., 1995), and also T cells expressing cytokines of both subsets have been described (Chen et al., 1994; Firestein et al., 1989; Groux et al., 1997). Cytokine secretion by T cells is transient and requires activation of the cell (Slifka et al., 1999). T cells regulate the production of cytokines predominantly on the level of cytokine gene transcription. In general, cytokine proteins and mRNA are not stored intracellularly but rather synthetized de novo upon stimulation of a T cell via its antigen receptor. An exception to this rule is TNF-a, for which intracellular storage and activation-induced splicing of a TNF-a pre-mRNA have been described (Yang et al., 1998b). Upon primary stimulation it takes days until a Th cell starts to express cytokine genes like IL-2, IFN-g, IL-4, or IL-10, and the kinetics of expression differ between cytokines and extend over days (Fig. 1) (Assenmacher et al., 1994, 1998a; Cardell and Sander, 1990; Cardell et al., 1993; Lederer et al., 1996). After secondary stimulation, cytokine expression is reinitiated within hours with similar and rapid kinetics for various cytokines (Cardell and Sander, 1990; Openshaw et al., 1995). Remarkably, in secondary stimulations, T cells will express selectively those effector cytokines, e.g., IFN-g, IL-4, and IL-10, that they already had expressed upon primary activation, a phenomenon termed cytokine memory.

Fig. 1. Differential kinetics of cytokine production in primary versus secondary T cell responses. After primary activation of naive T cells, cytokines are secreted transiently over days, with different kinetics for different cytokines. After reactivation, T cells secrete cytokines within hours and with similar kinetics for different cytokines. Preactivated T cells can memorize the expression of those effector cytokines, e.g., IFN-g, IL-4, and IL-10, that they had been instructed to express during primary activation. APC, antigen-presenting cell; TCR, T cell receptor.

cytokine memory of t helper lymphocytes

117

Here we discuss the induction, maintenance, reversibility, and molecular basis of cytokine memory. II. Cytokine Signals in the Induction of T Cell Cytokine Memory Cytokines present in the microenvironment of a Th cell during its primary activation are potent inducers of the various differentiation programs. In addition, other factors like the strength and duration of T cell receptor (TCR) triggering, the type and activation status of the antigen-presenting cell (APC), and the differential addressing of costimulatory receptors shape T cell differentiation (see below and reviewed in Constant and Bottomly, 1997; Lanzavecchia and Sallusto, 2000b; O'Garra, 1998; Seder and Paul, 1994). A prominent part of T cell differentiation is the induction to express distinct cytokines. Upon primary activation with antigen and costimulation via CD28, a naive Th cell is induced to express IL-2. The effector cytokines IFN-g, IL-4, IL-5, and IL-10, however, seem to require additional induction signals (see below and reviewed in Murphy et al., 2000). A. Instructive Cytokines for Th1 Differentiation 1. IL-12 The most prominent differentiation factor of the Th1 developmental program is IL-12, which can be produced by activated macrophages (Hsieh et al., 1993; Manetti et al., 1993; Seder et al., 1993) or dendritic cells (Macatonia et al., 1995; Scheicher et al., 1995). The transcription factors interferon regulatory factor (IRF)-1 and IRF-2 have been shown to play a central role in the IFN-ginduced IL-12 production by macrophages (Lohoff et al., 1997, 2000; Taki et al., 1997). The IL-12 receptor consists of an IL-12Rb1 and an IL-12Rb2 chain (Presky et al., 1996). In contrast to the b1 subunit, the b2 chain is not expressed by naive Th cells. Its expression is induced as a consequence of T cell activation. Signals mediated by IFN-g in mice (Szabo et al., 1997a), by IFN-a=b in humans (Gollob et al., 1997; Rogge et al., 1997), or by IL-12 itself (Galbiati et al., 1998) later on support further expression of the b2 subunit on Th1- but not on Th2polarized cells (see also Section VII). Instead, the key transcription factor of the Th2 developmental program, GATA-3, which is upregulated by IL-4 signals, antagonizes IL-12Rb2 expression (Ouyang et al., 1998). The binding of IL-12 to its receptor induces tyrosine phosphorylation of the Janus kinases Jak2 and Tyk2 (Bacon et al., 1995a). In T cells from Tyk2-de®cient mice, a partial reduction in IL-12-induced IFN-g production has been observed (Karaghiosoff et al., 2000; Shimoda et al., 2000). Upon Jak2 and Tyk2 activation, signal transducer and activator of transcription 4 (Stat4) is phosphorylated and dimerizes and translocates into the nucleus where it participates in the regulation of certain target genes (Bacon et al., 1995b; Jacobson et al.,

118

max loÈhning et al.

1995). The important but not absolutely essential roles of IL-12 signals in Th1 differentiation and IFN-g induction have been demonstrated by the analysis of IL-12-de®cient mice (Magram et al., 1996) or Stat4-de®cient mice (Kaplan et al., 1996b; Thierfelder et al., 1996), which show strongly reduced but detectable levels of IFN-g produced by Th cells. Th cells from mice that lack not only Stat4 but also Stat6, the central signal transduction molecule downstream of IL-4 in the Th2 developmental program, secrete considerable amounts of IFN-g but not IL-4 upon TCR stimulation in vitro (Kaplan et al., 1998). This points to the existence of IL-12/Stat4-independent pathways for IFN-g induction. Alternatively, Th1 differentiation could be the default pathway that may be taken in the absence of differentiating signals by IL-12 or IL-4. In CD8‡ T cells, TCR-mediated IFN-g production is largely Stat4 independent, as evidenced by the comparison of Stat4-de®cient and wild-type mice, whereas IL-12 plus IL-18 stimulation requires Stat4 to induce IFN-g independent of TCR triggering, in both CD4‡ and CD8‡ T cells (Carter and Murphy, 1999; Yang et al., 1999). IL-18, originally designated IFN-g-inducing factor (IGIF), can be secreted by activated macrophages and dendritic cells (Okamura et al., 1995; Stoll et al., 1998). Acting on its own, it does not induce IFN-g expression and Th1 development in primarily stimulated naive T cells. However, it enhances IL-12induced Th1 differentiation and IFN-g secretion (Robinson et al., 1997). 2. IL-23 IL-12 is a heterodimer consisting of a p35 and a p40 subunit. Recently, a novel p19 protein has been identi®ed, which is distantly related to p35 and also combines with p40 to form IL-23. IL-23 is secreted by stimulated dendritic cells and macrophages. It binds to IL-12Rb1 but not to IL-12Rb2. Like IL-12, it induces the phophorylation of Stat4 in human T cells, although apparently to an extent lower than that of IL-12. However unlike IL-12, it does not induce signi®cant IFN-g production in naive human Th cells activated with anti-CD3 and anti-CD28 (Oppmann et al., 2000). 3. TCCR-L A novel member of the type I cytokine receptor family with homology to the IL-12Rb2 chain has recently been cloned and named T cell cytokine receptor (TCCR). TCCR expression is highest in unpolarized T cells and decreases within 3 days of Th1 or Th2 differentiation. TCCR-de®cient mice show reduced IFN-g responses and IgG2a levels after immunization with protein antigens and enhanced susceptibility to the intracellular bacterium Listeria monocytogenes. TCCR does not bind to IL-12 or associate with the IL-12R chains. Interestingly, IFN-g induction in Th1 cultures of TCCR-de®cient T cells is reduced even when the Th cells are stimulated with anti-CD3 and anti-CD28

cytokine memory of t helper lymphocytes

119

in the absence of APCs. This suggests that activated T cells may secrete TCCR ligand, which could serve to optimize their IFN-g production in an autocrine fashion (Chen et al., 2000b). 4. IFN-a In human but not in murine T cells, IFN-a, like IL-12, leads to Stat4 phosphorylation, IFN-g induction, and Th1 development (Brinkmann et al., 1993; Cho et al., 1996; Parronchi et al., 1992; Rogge et al., 1998). A molecular explanation for this species difference has recently been proposed (Farrar et al., 2000a). In human and murine T cells, stimulation of the IFN-a receptor induces recruitment and activation of Stat2 (Yan et al., 1996). In humans, the carboxy-terminus of phosphorylated Stat2 serves as an adapter for Stat4 (Farrar et al., 2000b). Stat4 then is phosphorylated, which eventually leads to Th1 differentiation. In contrast to humans, in mice the carboxy-terminal region of Stat2 lacks the domain required for Stat4 recruitment, due to a minisatellite insertion in the murine Stat2 gene, which prevents Stat4 activation in response to IFN-a (Farrar et al., 2000a). Phosphorylation of tyrosine and serine residues in Stat molecules in response to receptor activation is commonly accepted to regulate the activity of these transcription factors (reviewed in Darnell, 1997; Hoey and Grusby, 1999). However, recently also methylation of an amino-terminal arginine residue, which is completely conserved in all seven Stat molecules, has been implicated in the regulation of the activation of Stat1 in response to IFN-a=b signals (Mowen et al., 2001). The arginine methylation is mediated by the protein arginine methyltransferase 1 (PRMT1), which associates with Stat1 in HeLa cells independent of IFN-a stimulation. Stat1 methylation is detectable in cell lines even without IFN-a treatment and it appears to be upregulated by IFN-a. The methyltransferase inhibitor 50-methylthioadenosine reduces the methylation status and also the DNA-binding capacity of Stat1 possibly due to an enhanced interaction of hypomethylated Stat1 with the protein inhibitor of activated Stat1 (PIAS1), which has been shown to interfere with the DNA binding of Stat1 (Liu et al., 1998a; Mowen et al., 2001). Whether methylation of the completely conserved amino-terminal arginine residue also occurs in the other Stat molecules and what in¯uence this may have on the activity of these factors are not clear. B. Instructive Cytokines for Th2 Differentiation 1. IL-4 IL-4 initiates the Th2 developmental program (Hsieh et al., 1992; Le Gros et al., 1990; Seder et al., 1992; Swain et al., 1990). In contrast to IL-12R, a functional receptor for IL-4 is expressed already on naive Th cells. The signaling mechanism of the IL-4R complex, which consists of an

120

max loÈhning et al.

IL-4-receptor-speci®c a chain (IL-4Ra) and a common g chain (gc), has recently been reviewed in detail (Nelms et al., 1999). Stat6 is the central signal transduction factor of the differentiative IL-4 signals (Hou et al., 1994). After triggering of the IL-4R, the Janus kinases Jak1 and Jak3 become activated and phosphorylate certain tyrosine residues in the cytoplasmic domain of the IL-4Ra chain (Miyazaki et al., 1994; Smerz-Bertling and Duschl, 1995; Witthuhn et al., 1994). This permits the recruitment of Stat6 to the IL-4Ra subunit where it is phosphorylated. Activated Stat6 dimerizes, translocates into the nucleus, and activates or enhances transcription of several IL-4-responsive genes by binding to speci®c DNA motifs (Darnell, 1997; Hoey and Grusby, 1999; Mikita et al., 1996). Stat6 target genes include IL-4Ra chain, CD23, MHC class II, and immunoglobulin e and g1 switch transcript promoters (Kaplan et al., 1996a; Linehan et al., 1998; Shimoda et al., 1996; Takeda et al., 1996). Whether Stat6 is directly involved in the activation of the IL-4 gene is unclear (reviewed in Szabo et al., 1997b; Wurster et al., 2000). Activated Stat6 binds to a region of the proximal IL-4 promoter that contains a putative Stat6 binding site (Curiel et al., 1997; Lederer et al., 1996). When multimerized in gene expression reporter constructs, this region enhances the expression of a reporter gene in response to IL-4. However, the contribution of this Stat site to either acute IL-4 transcription or Th2 development is not fully understood. Differentiated Th2 cells produce IL-4 independent of further IL-4/Stat6 signals (Huang et al., 1997). It has been suggested that phosphorylated Stat6 could bind to a silencer element 30 of the IL-4 gene, inactivating it and thus activating the IL-4 promoter/enhancer indirectly (Kubo et al., 1997). In addition, a putative Stat binding site has been identi®ed in a recently described IL-4 enhancer downstream of the IL-4 gene, which is reported to act in an activation-inducible, Th2-speci®c manner (Agarwal et al., 2000). However, the binding of activated Stat molecules to this site has not yet been demonstrated. At present, Stat6 appears to control IL-4 transcription rather indirectly, via the inactivation of a putative silencer element (Kubo et al., 1997) and induction of Th2-speci®c transcription factors like GATA-3 and c-Maf (Ho et al., 1996; Zhang et al., 1997a; Zheng and Flavell, 1997; and see below). The importance of IL-4 signals and Stat6 in Th2 development is evident from the analysis of IL-4-de®cient mice (Kopf et al., 1993; KuÈhn et al., 1991) or Stat6-de®cient mice (Kaplan et al., 1996a; Shimoda et al., 1996; Takeda et al., 1996). Both de®ciencies lead to a severe impairment of type 2 immune responses associated with markedly reduced production of Th2 cytokines and of IgE. Likewise, ectopic expression of inducible, activated Stat6 in developing Th1 cells results in IL-4-independent induction of the Th2-speci®c transcription factors GATA-3 and c-Maf and expression of type 2 cytokines. IL-12Rb2 expression and IFN-g production are reduced in such cells (Kurata et al., 1999; see also Section VII).

cytokine memory of t helper lymphocytes

121

Despite the pivotal role of IL-4 signals in Th2 differentiation, IL-4/Stat6independent Th2 development has been observed. Upon activation of Stat6de®cient Th cells by antigen, IL-4-independent Th2 development and IL-4 memory expression can take place in a relatively small population of CD4‡ T cells in vitro and ex vivo (Finkelman et al., 2000; Jankovic et al., 2000; Ouyang et al., 2000). In vitro, the Th2 differentiation of these Stat6-independent IL-4 producers is stable, presumably due to an IL-4/Stat6-independent GATA-3 autoactivation loop (Ouyang et al., 2000). BCL-6 is a zinc-®nger transcriptional repressor that can inhibit Stat6-mediated transcription by competing with Stat6 for binding to its DNA recognition motif (Dent et al., 1997). BCL-6 recruits a histone deacetylase-containing corepressor complex (Dhordain et al., 1998). BCL-6-de®cient mice suffer from in¯ammations of multiple organs and show enhanced Th2 cytokine production and overshooting Th2 immune responses (Dent et al., 1997; Ye et al., 1997). In mice that are de®cient for BCL-6 and for IL-4 or Stat6, the in¯ammatory Th2 immune reactions persist (Dent et al., 1998). This suggests an additional role for BCL-6 as a repressor of IL-4/Stat6-independent Th2 development, possibly by repressing other factors involved in Th2 differentiation, e.g., GATA-3. Also for natural killer (NK) T cells, IL-4 production independent of IL-4 signals has been shown in IL-4Ra- or Stat6-de®cient mice (Kaplan et al., 1999; Noben-Trauth et al., 1997). 2. IL-6 IL-6 has been proposed to be involved in Th2 differentiation in an IL-4dependent manner (Rincon et al., 1997). However, there are con¯icting data as to whether the reported IL-6-mediated enhancement of IL-4 production in T cell cultures in the presence of APCs was due to a direct Th2-differentiating effect of IL-6 on naive Th cells, since in another study only a moderate IFN-g reduction but no IL-4 induction in the presence of IL-6 was observed (Joseph et al., 1998). More recently, IL-6 has been reported to inhibit directly the IL12-driven development of IFN-g producers, independent of IL-4, via the enhancement of suppressor of cytokine signaling-1 (SOCS-1) expression in activated Th cells (Diehl et al., 2000). IL-6 has been shown earlier to induce expression of SOCS-1, also called Jak-binding protein (JAB) or Stat-induced Stat inhibitor-1 (SSI-1), in a murine myeloid leukemia cell line and in liver cells (Endo et al., 1997; Naka et al., 1997; Starr et al., 1997). SOCS-1 inhibits cytokine receptor signaling by interacting directly with Janus kinases (Jak1/2/3, Tyk2) to interfere with their kinase activity, which leads to reduced phosphorylation of cytokine receptor components, Jaks, and Stat molecules in response to various cytokines (Endo et al., 1997; Naka et al., 1997; Starr et al., 1997). In activated Th cells, IL-6-enhanced SOCS-1 expression is suggested to inhibit activation of Stat1 in response to IFN-g (Diehl et al., 2000), thereby

122

max loÈhning et al.

interfering with a proposed direct enhancing effect of IFN-g signals on IFN-g production (Bradley et al., 1996). SOCS-1-mediated Stat1 inhibition may diminish the IFN-g-induced enhancement of IL-12Rb2 expression, which could result in reduced IL-12 responsiveness of the Th cells (Szabo et al., 1997a). This has not been tested yet. 3. MCP-1 (CCL2) Recently, the chemokine monocyte chemoattractant protein-1 (MCP-1), now called chemokine ligand 2 (CCL2), has been suggested to play a critical role in the generation of Th2 immune responses in vivo (Gu et al., 2000) and to enhance IL-4 production in an IL-4-dependent manner in vitro (Karpus et al., 1997). After immunization of MCP-1-de®cient mice with a haptenized protein antigen, lymph node cells produced reduced levels of IL-4, IL-5, and IL-10 but normal amounts of IFN-g and IL-2 in response to the protein antigen. Also the serum concentration of hapten-speci®c IgG1 was reduced. Infections with Leishmania major were ameliorated in MCP-1-de®cient BALB/c mice compared with wild-type BALB/c (Gu et al., 2000). Whether IL-4 is still able to drive Th2 development in MCP-1-de®cient T cells has not been tested so far, nor has a molecular mechanism been proposed for a putative MCP-1-enhanced Th2 development. More recently, MCP-1-de®cient mice have been demonstrated to have defects in mounting Th1-dependent experimental autoimmune encephalomyelitis (EAE). The severity of EAE and the frequency of relapses were reduced in the absence of MCP-1, which was accompanied by fewer in¯ammatory in®ltrates, lower frequencies of in®ltrating macrophages in the central nervous system, and reduced serum levels of IFN-g. The reason for the reduced EAE susceptibility of MCP-1-de®cient mice appears not to be T cellintrinsic but rather due to MCP-1-de®cient non-T cells as indicated by the inability of primed wild-type T cells to adoptively transfer the disease to MCP1-de®cient hosts. Primed MCP-1-de®cient T cells were able to transfer EAE to wild-type recipients (Huang et al., 2001). Thus, it seems likely that MCP-1 is critically involved in the recruitment and activation of APCs and effector cells like macrophages and NK cells (Allavena et al., 1994; Fuentes et al., 1995; Maghazachi et al., 1994), and also activated T cells (Carr et al., 1994; Loetscher et al., 1996) in response to in¯ammatory stimuli. MCP-1 de®ciency may thus result in defects in the generation of both type 1 and type 2 immune responses (for a recent, detailed review, see Luther and Cyster, 2001). III. Key Transcription Factors in the Induction and Maintenance of Cytokine Memory In the induction and maintenance of cytokine memory in individual T cells, information on the initial differentiating cytokine signals can be stored at

cytokine memory of t helper lymphocytes

123

several levels. Selective and stable activation or induction of speci®c transcription factors, in concert with stable epigenetic modi®cations of selected cytokine genes, today is discussed as the most likely molecular basis for the cytokine memory of an individual cell. Stat molecules themselves are unlikely to serve as long-term information storage because they are ubiquitously expressed and their activation is transient. Phosphorylation of Stat molecules occurs within minutes after engagement of cytokine receptors, but inactivation, mainly via dephosphorylation by protein tyrosine phosphatases, is also rapid. The entire activation and inactivation cycle of a Stat molecule is estimated to take approximately 20 min, such that, following a cytokine stimulation, already after a few hours activated Stat molecules are no longer detectable (Darnell, 1997; Haspel et al., 1996; Shuai et al., 1992). Thus, Stats rather appear to function as transducers of differentiating cytokine signals to activate expression of selective target genes, the products of which may then initiate and stabilize the respective differentiation program and cytokine memory. Differential induction of distinct transcription factors by Stats is likely to play a key role in the induction and maintenance of cytokine memory in individual T cells. Expression of GATA-3, c-maf, and JunB is enhanced in the Th2 developmental lineage, while T-bet expression is upregulated in the course of Th1 differentiation. These factors have been shown to induce and cooperate with additional transcription factors and coactivators, like NFAT, NIP45, NF-kB, and p300/CBP, to either directly transactivate cytokine genes and/or to modify their chromatin structure such that speci®c sets of cytokine genes become readily accessible for transcription in future T cell stimulations (for recent reviews, see Dong and Flavell, 2000; Glimcher and Murphy, 2000; Murphy et al., 2000; O'Garra and Arai, 2000). A. GATA-3 as a Key Transcription Factor for Th2 Cytokine Memory The zinc-®nger transcription factor GATA-3 controls developmental processes at distinct stages. In humans, GATA-3 haplo-insuf®ciency leads to the HDR syndrome characterized by hypoparathyroidism, sensorineural deafness, and renal anomaly (Van Esch et al., 2000). In murine embryogenesis, GATA-3 is essential for the development of the central nervous system and for fetal hematopoiesis (Pandol® et al., 1995). In T cell development, GATA-3 de®ciency causes an arrest before the earliest CD4/CD8 double-negative thymocyte stage (Hendriks et al., 1999; Ting et al., 1996). In mature T cells, GATA-3 expression is selectively enhanced in Th2 cells (Zhang et al., 1997a; Zheng and Flavell, 1997). Transgenic overexpression of GATA-3 in CD4 T cells leads to type 2 cytokine expression whereas transfection of antisense GATA-3 in Th2 clones reduces the production of type 2 cytokines (Zheng and Flavell, 1997). In mice expressing a dominant-negative mutant of GATA-3, reduced Th2 cytokine production and attenuated Th2-dependent allergic airway in¯ammation have

124

max loÈhning et al.

been observed (Zhang et al., 1999a). GATA-3 transactivates the IL-5 promoter directly (Lee et al., 1998; Siegel et al., 1995; Zhang et al., 1997a) but has only a limited capacity to activate the IL-4 promoter (Ranganath et al., 1998; Zhang et al., 1998). Several intergenic regions in the IL-4/IL-13 locus with GATA-3-dependent enhancer activity have been described (Ranganath et al., 1998). Based on an interspecies homology search for conserved noncoding sequences (CNS), a 401-bp element (CNS-1) located in the intergenic region between the IL-4 gene and the IL-13 gene has been identi®ed (Loots et al., 2000). In mice transgenic for a human yeast arti®cial chromosome containing CNS-1, inducible deletion of CNS-1 in Th2-polarized cells leads to a reduction in the frequencies of Th cells producing human IL-4 or IL-13 without affecting the amount of these cytokines produced per cell by the remaining human IL-4 or IL-13 producers. Thus, CNS-1 is assumed to act as a coordinate regulator of the chromatin structure in the IL-4/IL-13/IL-5 gene cluster. The localization of CNS-1 between the IL-4 gene and the IL-13 gene overlaps with two Th2speci®c DNase I-hypersensitive sites (Loots et al., 2000; Takemoto et al., 1998). Remarkably, in the region of one of these DNase I-hypersensitive sites, CNS1 contains a GATA binding site, and in vitro, assembly of a stimulation-inducible, Th2-speci®c DNA±protein complex containing GATA-3 has been described (Takemoto et al., 2000). However, assembly of this complex in vivo remains to be demonstrated. In mice transgenic for an IL-4 promoter reporter construct fused to a DNA fragment containing the intergenic DNase I-hypersensitive sites, including CNS-1, IL-4 promoter activity is strongly enhanced compared with that of mice carrying the IL-4 promoter construct alone (Lee et al., 2001). This intergenic enhancer element is not Th2 cell speci®c. For Th2 speci®city, the IL-4 promoter requires an enhancer from the second intron of the IL-4 gene. This intragenic enhancer has originally been described as a mast-cell-speci®c enhancer of IL-4 transcription (Henkel et al., 1992). In mobility shift assays of mast cell nuclear extracts, this DNA region has been shown to bind to GATA-1, GATA-2, Stat5, and PU.1, a member of the Ets transcription factor family. In addition, this region contains several putative GATA-3 binding sites, one of which binds to a protein, putatively GATA-3, in the T cell line EL-4 (Henkel and Brown, 1994; Hural et al., 2000). In a transgenic situation, ectopic GATA-3 expression in polarized Th1 cells leads to a strong enhancement of the activity of an IL-4 promoter only when linked to both the intergenic and the intragenic enhancers (Lee et al., 2001). In addition, GATA-3 has been shown to bind to a recently identi®ed third, distal IL-4 enhancer in cells of a Th2 clone. This enhancer is located in a stimulation-induced, Th2-speci®c DNase I-hypersensitive site downstream of the IL-4 locus (Agarwal et al., 2000). On its own, this distal enhancer element

cytokine memory of t helper lymphocytes

125

enhances the activity of an IL-4 promoter construct only weakly, like the intragenic enhancer, but in contrast to the intragenic enhancer the distal enhancer confers no Th2 speci®city (Lee et al., 2001). The central role of GATA-3 not only in the induction of the Th2 but also in the suppression of the Th1 developmental program is evident from the analysis of IL-4- or Stat6-de®cient T cells with ectopic GATA-3 expression. In Th cells activated under Th1-polarizing conditions, retroviral GATA-3 transduction is suf®cient to induce expression of type 2 cytokines and to inhibit type 1 cytokine and IL-12Rb2 expression independent of IL-4 or Stat6 (Ferber et al., 1999; Ouyang et al., 1998, 2000). Moreover, additional features of the Th2 developmental program like induction of c-Maf and of Th2-speci®c DNase I-hypersensitive sites in the IL-4 locus are mediated by ectopic GATA-3 expression without the requirement of Stat6 (Ouyang et al., 2000). Even in Th cells de®cient for Stat6 or IL-4, ectopic GATA-3 expression not only initiates the Th2 developmental program but also induces expression of the endogenous GATA-3 gene, pointing to the existence of an autoregulatory loop in the regulation of GATA-3 expression (Ouyang et al., 2000). This GATA-3 autoactivation could provide the molecular basis for the observed IL-4/Stat6 independence of Th2 commitment and memory (Dent et al., 1998; Finkelman et al., 2000; Huang et al., 1997; Jankovic et al., 2000; Ouyang et al., 2000; and see Section II.B.1). Thus, the initial role of Stat6 may be to augment the GATA-3 expression level above a certain threshold. Later, GATA-3 may drive the Th2 developmental program independent of Stat6. B. Regulation of GATA-3 Expression and Activity Our understanding of the regulation of the GATA-3 gene is limited, particularly with regard to its expression in Th2 cells. To what extent T-cell-speci®c DNase I-hypersensitive sites that have been identi®ed up to 10 kb upstream of and also within the GATA-3 locus are involved in the control of GATA-3 expression is not yet clear (Gregoire and Romeo, 1999; Lieuw et al., 1997). A silencer element, which is located in the region of a DNase I-hypersensitive site 7.5 kb upstream of the human GATA-3 locus, suppresses the activity of the human GATA-3 promoter in nonhematopoietic and erythrocytic cell lines but not in a T cell line, suggesting that this silencer may confer T-cell-speci®c expression of the human GATA-3 gene. A transcriptional activator enhancing the activity of the human GATA-3 promoter in all cell types tested has been identi®ed in the ®rst intron of the human GATA-3 gene (Gregoire and Romeo, 1999). The ®rst infron of the murine GATA-3 gene contains a positive transcriptional regulatory element that is position- and orientation-dependent, as deduced from transient transfections of GATA-3 promoter reporter constructs in murine cell lines expressing GATA-3. Site-speci®c mutation of a double GATA binding site located in this ®rst infron has been reported not to affect

126

max loÈhning et al.

GATA-3 transcriptional activity in those transient transfection assays (George et al., 1994). In Th2 development, GATA-3 expression is enhanced by IL-4/Stat6 signals (Kurata et al., 1999; Ouyang et al., 1998), whereas in Th1 differentiation IL-12/ Stat4 mediates GATA-3 suppression (Ouyang et al., 1998). Whether Stats regulate GATA-3 expression directly or only indirectly is unknown. Upon an initial enhancement of GATA-3 expression, which is usually mediated by activated Stat6, GATA-3 can autoactivate its own gene expression independent of further IL-4/Stat6 signaling (Ouyang et al., 2000). However, whether this GATA-3 autoactivation is due to a direct interaction of GATA-3 with regulatory elements of the GATA-3 gene or occurs via intermediate, GATA-3-induced factors remains to be determined. Recently, by retroviral gene transduction the GATA family members GATA-1, GATA-2, and GATA-4 have been shown to be able to induce endogenous GATA-3 expression and Th2 development in Stat6de®cient Th cells (Ranganath and Murphy, 2001). Stimulation of rat T cells with a mitogenic anti-CD28 antibody alone has been described to enhance GATA-3 expression (Rodriguez-Palmero et al., 1999), probably at least in part due to an enhancement of sensitivity of the activated cell to IL-4R signaling (Davis et al., 1999; Kubo et al., 1999; Skapenko et al., 2001). However, CD28 signals are not absolutely required for Th2 development, as is obvious from the analysis of CD28-de®cient mice (Brown et al., 1996). Activation of NF-kB is required for enhanced expression of GATA-3 and type 2 cytokines in early Th2 differentiation, as deduced from the analysis of mice de®cient for the p50 subunit of NF-kB. This information provides a link between TCR stimulation and cytokine signaling in the induction of Th2 development. NF-kB p50 de®ciency appears not to affect T-bet and IFN-g expression under Th1-polarizing conditions (Das et al., 2001). Transforming growth factor-b (TGF-b) has recently been reported to negatively in¯uence Th2 development via inhibition of GATA-3 expression (Gorelik et al., 2000; Heath et al., 2000a). However, TGF-b also negatively affects Th1 development (Hoehn et al., 1995; Schmitt et al., 1994), probably by inhibition of IL-12Rb2 chain expression, with the consequence of reduced sensitivity to IL-12 (Bright and Sriram, 1998; Gorham et al., 1998). Several factors appear to contribute to the post-translational regulation of the transcriptional activity of GATA-3. In vivo, GATA-1, a differentiation factor for the erythroid lineage, can associate with the transcriptional coactivator p300/ CBP, which confers intrinsic acetyltransferase activity (Blobel et al., 1998; Boyes et al., 1998). In vitro, p300/CBP-mediated acetylation of GATA-1 has been shown. This has been suggested to induce conformational changes in GATA-1 leading to an improvement in its DNA binding capacity and thus to increased transcription of GATA-1-dependent genes (Boyes et al., 1998). Acetylation of GATA-3 in T cells and a correlation of this acetylation state

cytokine memory of t helper lymphocytes

127

with the transactivating capacity of GATA-3 have been reported (Yamagata et al., 2000). Phosphorylation of GATA-3 may participate in the regulation of GATA-3 activity, perhaps by altering the ability of GATA-3 to interact with other regulatory factors (Chen et al., 2000a). In a Th2 clone, cAMP enhances the phosphorylation of GATA-3 and of p38 mitogen-activated protein (MAP) kinase, which is able to phosphorylate GATA-3 in vitro. This may explain the Th1 cytokineinhibiting and Th2 cytokine-enhancing effects of cAMP (Betz and Fox, 1991; Chen et al., 2000a; Lee et al., 2000). The recently cloned repressor of GATA (ROG), which is transiently induced upon T cell stimulation in both Th1 and Th2 cells, binds to GATA-3 and inhibits its binding to DNA. ROG reduces the production of type 2 cytokines in a Th2 clone but it also downmodulates IFN-g production in a Th1 clone, pointing to the existence of targets other than GATA-3 for ROG (Miaw et al., 2000). Stimulatory cofactors for GATA proteins have been identi®ed as well. Friend of GATA-1 (FOG-1) binds to GATA-1, which leads to synergistic activation of gene transcription in erythroid and megakaryocytic cell differentiation (Tsang et al., 1997, 1998). FOG-2 interacts with GATA-4, -5, and -6 (Lu et al., 1999; Svensson et al., 1999; Tevosian et al., 1999) and it is essential for heart development and coronary vascularization (Tevosian et al., 2000). Interaction of a FOG with GATA-3 has not been reported so far. C. Further Th2 Transcription Factors 1. c-Maf c-Maf is a Th2-speci®c, basic-region/leucine-zipper transcription factor that binds to and transactivates the IL-4 promoter (Ho et al., 1996). Synergies among c-Maf and NFATp (Ho et al., 1996), NFAT-interacting protein 45 (NIP45) (Hodge et al., 1996a), and JunB (Li et al., 1999) in the transactivation of the IL-4 promoter have been observed. Transgenic overexpression of c-maf in Th cells leads to an increase in the production of IL-4 and other type 2 cytokines, to enhanced IL-4-dependent antibody class switching to IgG1 and IgE, and to reduced IFN-g production (Ho et al., 1998). In T cells activated under Th1polarizing conditions or in Stat6-de®cient Th cells, c-maf overexpression neither induces IL-4 nor inhibits IFN-g expression (Ho et al., 1998; Ranganath and Murphy, 2001). Mice lacking c-maf are characterized by a defect in lens development (Kim et al., 1999c) and c-maf-de®cient Th2-polarized cells selectively produce reduced amounts of IL-4 but not of other type 2 cytokines, suggesting a selective role for c-Maf in the regulation of IL-4 transcription (Kim et al., 1999b). Ectopic expression of GATA-3 in Stat6-de®cient Th cells induces c-maf expression (Ouyang et al., 2000). However, whether this is a direct or indirect effect of GATA-3 remains to be shown. Whether IL-4 and Stat6 are able

128

max loÈhning et al.

to induce c-maf expression independent of GATA-3 is not clear either. Interestingly, in rats the c-maf promoter contains two Maf recognition elements (MARE), and Pax6, a central transcription factor for lens development, as well as c-Maf itself, activates the c-maf promoter (Sakai et al., 2001). Thus, a positive autoregulatory loop, which has already been described for GATA-3 gene expression (Ouyang et al., 2000), may also exist for the c-maf gene. 2. JunB Expression of JunB, a member of the AP-1 family, is selectively enhanced in the early course of Th2 differentiation whereas it remains on a lower level in Th1 cells (Li et al., 1999). JunB binds directly to the IL-4 promoter and activates it in synergy with c-Maf, probably due to cooperation of the two factors in DNA binding. The activity of JunB is regulated at least in part via its phosphorylation by the MAP kinase c-Jun NH2 -terminal kinase (JNK), which links IL-4 gene transcription to TCR triggering (see Section V). Transgenic JunB overexpression leads to enhanced production of type 2 cytokines without affecting IFN-g production in Th cells activated under Th1-polarizing conditions (Li et al., 1999). However, in Stat6-de®cient Th cells ectopic JunB expression by retroviral gene transduction does not induce IL-4 or suppress IFN-g production (Ranganath and Murphy, 2001). D. T-bet as a Key Transcription Factor for Th1 Cytokine Memory In symmetry to the presumptive master role for GATA-3 in the control of Th2 cytokine memory, T-bet may be the key transcription factor regulating cytokine memory in the Th1 developmental lineage. T-bet (for T-box expressed in T cells) is rapidly and selectively induced in naive Th cells upon stimulation under Th1-polarizing conditions but not under Th2-polarizing conditions (Szabo et al., 2000). In a T cell lymphoma, T-bet transactivates the IFN-g promoter and represses IL-2 promoter activity whereas it does not affect the activity of an IL-4 promoter construct. Ectopic expression of T-bet induces IFN-g and strongly reduces IL-5 expression. To a lesser extent IL-4 production is suppressed, not only in naive Th cells activated under Th2-inducing conditions, but also in type 2-polarized CD8‡ T cells and in Th2 cells polarized for up to 3 weeks. In Th2 clones, forced T-bet expression still is able to reduce IL-4 and IL-5 memory expression, although IFN-g induction in those cells is rather inef®cient. The T-bet-driven conversion of cytokine memory in polarized Th2 cells is independent of IFN-g signals since it also occurs in IFN-gR-de®cient Th2 cells (Szabo et al., 2000). Thus, the simultaneous upregulation of type 1 and downregulation of type 2 cytokine expression, even in committed Th2 cells, supports the idea of T-bet as a key regulator of the Th1 developmental program. The induction of T-bet expression itself is less clear. So far it has not been clari®ed whether T-bet is induced by IL-12/Stat4 signals or, to name just one

cytokine memory of t helper lymphocytes

129

alternative, by IFN-g/Stat1 signals. Regarding the more pronounced repression of IL-5 compared with IL-4 expression by T-bet, it is tempting to speculate on a direct negative effect of T-bet on GATA-3, which would result in exactly such a phenotype. All in all, the available evidence suggests a scenario of T cell polarization with the key transcription factors T-bet and GATA-3 inhibiting each other's expression, and thereby the adverse developmental programs, stabilizing the converse developmental programs by direct and indirect autoregulatory loops and thus committing an individual Th cell stably to either the Th1 or the Th2 lineage. E. Further Th1 Transcription Factors 1. ERM Expression of ERM, a member of the Ets transcription factor family, is Th1 speci®c and requires IL-12/Stat4 signals. In restimulation of Th1 cells without IL-12, ERM is not expressed, although the cells express IFN-g. ERM enhances the activity of an IFN-g promoter construct modestly but it does not induce IFN-g production in Stat4-de®cient Th cells (Ouyang et al., 1999). Thus, it remains to be clari®ed at which steps of Th1 differentiation ERM is involved. 2. Txk Txk is a member of the Tec family of tyrosine kinases. It is constitutively expressed in human Th1 and Th0 clones but not in Th2 clones. Short-term stimulation of peripheral blood CD4 T cells with IL-12 enhances Txk expression and IL-4 abrogates Txk expression. Txk antisense RNA reduces IFN-g secretion by human T cells, but does not affect IL-2 and IL-4 production. In Jurkat T cells, Txk expression leads to moderate transactivation of an IFN-g promoter construct but not of IL-2 or IL-4 promoter constructs (Kashiwakura et al., 1999). Neither the molecular basis for selective Txk regulation nor the mechanism of its contribution to IFN-g transcription have been elucidated yet. 3. CIITA MHC class II transactivator (CIITA) is a coactivator for MHC class II gene expression in APCs. In murine T cells, CIITA expression is restricted to Th1 cells and depends on IFN-g signaling. CIITA reduces IL-4 expression in a Th2 clone. CIITA de®ciency leads to IL-4 production in IL-12-stimulated Th cells without affecting IFN-g production (Gourley et al., 1999). A molecular mechanism for this apparent negative effect of CIITA on IL-4 gene transcription has recently been proposed (Sisk et al., 2000). The transcriptional coactivator p300/ CBP binds to NFATp and NFATc and enhances their transactivation activity (Avots et al., 1999; Garcia-Rodriguez and Rao, 1998). Cotransfection of NFATs and p300/CBP augments the activity of an IL-4 promoter construct (Sisk et al.,

130

max loÈhning et al.

2000). CIITA also binds to p300/CBP and competes with NFATp for binding to p300/CBP. It is likely that in Th1-polarized IFN-g-producing T cells, CIITA competes with NFATs for binding to p300/CBP and thus negatively affects NFAT-mediated transcription of the IL-4 gene. This competition effect may also affect other genes under control of NFATs, e.g., Fas ligand (Gourley and Chang, 2001). GATA-3 gene expression has been reported to be unaffected by CIITA (Gourley et al., 1999). IV. Role of the Antigen-Presenting Cell in the Induction of Cytokine Memory The in¯uence of the type of APC on the polarization of an emerging T cell response and the selective induction of cytokine memory has been a matter of debate for years (e.g., Schmitz et al., 1993; reviewed in Banchereau and Steinman, 1998; Seder and Paul, 1994). Recently several groups have de®ned and analyzed discrete types of APCs in detail for their capacity to polarize Th cell differentiation (for recent reviews, see Lanzavecchia and Sallusto, 2000a; Moser and Murphy, 2000; Reid et al., 2000). In mice, antigen-pulsed CD8a‡ dendritic cells (DCs), supposed to be derived from lymphoid precursor cells, skew the antigen-speci®c T cell response toward production of Th1 cytokines, like IFN-g and IL-2, whereas CD8a DCs, probably derived from myeloid precursor cells, induce a Th2-biased T cell response, with production of IL-4, IL-5, and IL-10. CD8a‡ DC-derived IL-12 seems to be critically involved in the induction of the type 1 cytokine memory (Maldonado-Lopez et al., 1999). By and large, similar ®ndings have been reported in another study using CD11b expression as a criterion for the discrimination of DC subsets. Antigen-pulsed, lymphoid lineage CD11c‡ CD11blo DCs induce mainly IFN-g memory, some IL-10 memory, but no IL-4 memory in reactive Th cells. The T cell response induced by CD11c‡ CD11bhi myeloid DCs is characterized by a similar capacity to secrete IFN-g, but also IL-4 and IL-10 (Pulendran et al., 1999). Neither study addresses the question whether the observed IL-4 memory induction is dependent on IL-4 or other signals. In humans, CD40 ligand (CD40L)-matured, myeloid DCs in vitro induce naive Th cells to become IFN-g producers, which is largely blockable by the addition of anti-IL-12. In contrast, CD40L-matured lymphoid DCs, derived from peripheral CD4‡ CD3 CD11c plasmacytoid cells, neither secrete IL-12 IL-4, IL-6, IL-10, nor IL-13. The induce a Th2 cytokine memory in an apparently IL-4-independent manner (Rissoan et al., 1999). In another report, plasmacytoid DCs from peripheral blood induced a strong Th1 cytokine memory upon viral or CD40L-mediated maturation (Cella et al., 2000). According to those experiments, the immature precursors of plasmacytoid or myeloid DCs support the generation of IL-4 and IL-10 and inhibit the gener-

cytokine memory of t helper lymphocytes

131

ation of IFN-g memory. The Th1 development observed with the matured DCs is largely dependent on IL-12 and IFN-a/b, which can be produced by mature plasmacytoid and myeloid DCs (Cella et al., 1999a,b). A possible explanation of the apparently divergent results comes from analysis of the kinetics of DC activation and maturation in vitro (Kalinski et al., 1999; Langenkamp et al., 2000). After lipopolysaceharide (LPS) stimulation for a few hours, DCs secrete IL-12, IL-6, and some IL-10 and thus can induce strong IFN-g responses but weak IL-4 responses. After 1 to 2 days of LPS activation, however, production of IL-12, IL-6, and IL-10 is exhausted and a larger fraction of the stimulated Th cells starts to express IL-4, while less cells express IFN-g (Langenkamp et al., 2000). Thus, the timing of interaction of an activated DC with a Th cell in the course of DC activation may be critical for the polarization of cytokine memory in the T cell. The relevance of the maturation status of the DC at the time of interaction with the T cell is also evident from recent reports comparing immature monocyte-derived DCs before and after maturation induced by in¯ammatory cytokines. In contrast to mature DCs, immature DCs induce only weak proliferation of alloreactive Th cells. After repetitive stimulations with immature DCs, the remaining T cells secrete almost exclusively IL-10, but no IL-4 or IFN-g. T cells repeatedly stimulated with mature DCs secrete large amounts of IFN-g and IL-2. In cocultures, Th cells generated with immature DCs inhibit the proliferation of the Th1 cells generated with mature DCs in an IL-10-, TGFb-, or CTLA-4-independent but cell contact- and T cell activation-dependent manner (Jonuleit et al., 2000). The repeated exposure to immature DCs thus probably results in the acquisition of regulatory capacities by the responding Th cells in vitro, while mature DCs preferentially induce effector T cells. In vivo, injection of immature human DCs pulsed with a viral antigen leads to a long-lasting reduction of the antigen-speci®c lytic activity of CD8 T cells, to reduced numbers of IFN-g-secreting cells, and to the appearance of antigenspeci®c IL-10 producers (Dhodapkar et al., 2001). Thus, antigen presentation by immature DCs, which may occur in vivo in the absence of in¯ammation, may be an option for the generation of regulatory T cells. In¯ammatory signals would induce maturation of the immature DCs and the rapid activation of effector T cells capable of mediating an acute immune response, both Th1 or Th2 (commented in Roncarolo et al., 2001). The conditions of DC maturation seem to be critical for their cytokine memories. Immature monocyte-derived DCs matured with LPS and IFN-g secrete high levels of IL-12 upon later stimulation with CD40L, inducing a Th1 cytokine memory in naive human Th cells. In contrast, maturation with LPS and prostaglandin E2 results in DCs secreting IL-10 but very little IL-12, which enhances Th2 cytokine production by the activated T cells (Kalinski et al., 1997; Vieira et al., 2000).

132

max loÈhning et al.

In the studies mentioned so far, Th2 differentiation appeared to be the default pathway taken in the absence of IL-12 signals from the priming DC. However, recently a proposal for active induction of Th2 development by DCs has been made. This study reports the differential responsiveness of a line of immature murine myeloid DCs and of murine splenic DCs toward yeasts or hyphae of Candida albicans. Exposure to yeast activates DCs to produce IL-12. Exposure to hyphae results in secretion of IL-4 but not IL-12 by a subpopulation of DCs. Both the Th1-skewing effect of the yeast-exposed DCs and the Th2-skewing effect of the hyphae-exposed DCs depend on the capacity of the DCs to produce IL-12 or IL-4, respectively, as has been demonstrated by the use of DCs de®cient for IL-12 or IL-4 (d'Ostiani et al., 2000). The mechanism of differential cytokine expression by the DCs, in response to the two forms of the fungus, is not yet understood. A differential effect of LPS and ES-62, a phosphorylcholine-containing glycoprotein secreted by the ®larial nematode Acanthocheilonema viteae, on cytokine secretion by murine DCs has been described. LPS stimulation of bone marrow-derived DCs leads to IL-12 secretion. ES-62-stimulated DCs secrete little IL-12 and the stimulated Th cells preferentially express IL-4 (Whelan et al., 2000). Whether ES-62stimulated DCs actively drive Th2 differentiation, e.g., via secretion of IL-4, is not clear. Compared with unstimulated DCs, ES-62-stimulated DCs enhance secretion of IL-4 by Th cells only moderately. DCs from murine Peyer's patches activate Th cells preferentially for the production of IL-4, IL-10, and some IFN-g, while splenic DCs induce IFN-g and only a little IL-4. In addition, DCs from Peyer's patches but not from spleen have been reported to respond to CD40L stimulation with secretion of IL-10 (Iwasaki and Kelsall, 1999). IL-10 appears to exert multiple effects on DCs like reduction of surface expression of MHC and costimulatory molecules, reduction of IL-12 secretion, and a diminished capacity to prime naive T cells (Buelens et al., 1995, 1997; Corinti et al., 2001; De Smedt et al., 1997). IL-4 modi®es the expression level of costimulatory molecules on antigenpresenting DCs and B cells. In a murine model of diabetes, IL-4 is reported to selectively enhance B7.2 (CD86) but not B7.1 (CD80) expression on DCs, resulting in a reduced cytotoxic T lymphocyte (CTL) activity in that model and the prevention of diabetes (King et al., 2001). In contrast, in murine mixed allogeneic lymphocyte cultures IL-4 is required for the upregulation of both B7.1 and B7.2 on DCs and B cells, effective activation of alloreactive CD4 T cells, and rapid rejection of allogeneic skin grafts (Bagley et al., 2000). IL-4 is required in the priming but not in the effector phase of the CTL response for ef®cient tumor rejection by CD8 T cells (SchuÈler et al., 1999). Not only subsets of DCs but also B cell subpopulations have recently been described to selectively drive Th1 and Th2 differentiation. Stimulation of murine B cells with anti-CD40 antibodies in the presence of IL-12 and IL-18

cytokine memory of t helper lymphocytes

133

activates B cells to produce IFN-g (Yoshimoto et al., 1997). CD40L stimulation of human B cells induces expression of IL-12 by them, which could be enhanced by the addition of IFN-g and abrogated by IL-10. The addition of IL-4 augmented the secretion of IL-6 (Schultze et al., 1999). Coculture of naive B cells with polarized Th1 cells will generate B effector 1 (Be1) cells capable of producing IL-12, IFN-g, IL-2, and some IL-6 and IL-10 upon mitogenic restimulation. Coculture with polarized Th2 cells results in Be2 cells expressing IL-4, IL6, IL-10, and IL-2. Activation of naive Th cells together with Be1 cells leads to the generation of Th1 cells secreting IFN-g and IL-2 upon restimulation. Th2 cells are generated from naive Th cells in an IL-4-dependent manner after coculture with Be2 cells. In IL-4-de®cient Th cells instructed by Be2 cells, induction of the Th2 cytokine memory is abrogated by a neutralizing anti-IL-4 antibody, indicating that IL-4 secreted by the Be2 cells is required for the observed Th2 differentiation (Harris et al., 2000). Thus, due to their responsiveness to and apparent memory of the cytokine environment during their primary activation and interaction with polarized Th cells, B memory cells as antigenpresenting cells may contribute to the maintenance of a Th cytokine memory by conservative instruction of newly recruited naive Th cells. V. T Cell Receptor Signals in the Induction of Cytokine Memory The impact of the duration and strength of TCR stimulation, altered peptide ligands, antigen structure, antigen dose, and various costimulatory molecules on T cell differentiation and cytokine induction have been reviewed recently (Constant and Bottomly, 1997; Dong and Flavell, 2000; Hunter and Reiner, 2000; Lanzavecchia and Sallusto, 2000b; Watts and DeBenedette, 1999). Here we focus on recent ®ndings on the relevance of TCR and accessory signal transduction molecules and transcription factors on T cell polarization and cytokine memory. Stimulation of a T cell via its TCR and costimulatory signals, leads to the activation of various members of the mitogen-activated protein kinase (MAPK) family, some of which are involved in the transcriptional activation of cytokine genes. Activated MAPKs translocate into the nucleus where they can phosphorylate various transcription factors. The three major MAPK pathways in mammals are the extracellular signal-regulated kinase (ERK), the p38, and the JNK pathways (for recent reviews, see Chang and Karin, 2001; Dong and Flavell, 2000). A. ERK TCR triggering activates the Ras/ERK MAPK pathway. In mice transgenic for a dominant-negative form of H-Ras, defective Th2 but normal Th1 development and normal proliferation have been observed. The defect in IL-4

134

max loÈhning et al.

induction is probably due to a reduction in the phosphorylation of IL-4Ra, Jak1, Jak3, and Stat6 in response to IL-4. Thus, similar to costimulatory signals via CD28 (Kubo et al., 1999), the TCR-activated Ras/MAPK pathway also appears to favor Th2 differentiation, via the enhancement of IL-4R signaling, probably by augmenting the activity of Janus kinases (Yamashita et al., 1999). Stimulation of CD4 T cells with anti-CD3 and anti-CD28 induces a transient inhibition of IL-4R signaling for about 8 h, which correlates with the kinetics of MAPK activity and can be blocked by a MAPK kinase inhibitor. This transient inhibition also extends to signaling in response to IL-2, IL-6, and IFN-a (Zhu et al., 2000). Thus, in the early course of T cell activation, the MAPK pathway may negatively affect signaling of various inductive cytokines. B. p38 MAPK The p38 MAPK pathway is activated in response to proin¯ammatory and stress signals but also in response to mitogenic stimuli, and it mediates activation of several transcription factors, like ATF-2. Inhibition of p38 MAP kinase during restimulation of murine Th1 cells has been reported to reduce IFN-g expression. IL-4 production by Th2 cells is not affected (Rincon et al., 1998). However, in other studies with human and murine Th2 cells a reduction of IL-4 expression by inhibition of p38 MAPK has been observed (Chen et al., 2000a; Schafer et al., 1999). Recently, in murine Th1 cells inhibition of p38 MAPK or of MAPKextracellular signal-regulated kinase kinase 4 (MEKK4) have both been described to selectively abrogate IL-12- plus IL-18-induced but not TCR-induced IFN-g memory expression. IL-18, in synergy with IL-12, induces GADD45b expression, which in turn activates MEKK4 and p38 MAPK resulting in cytokineinduced IFN-g expression (Yang et al., 2001). p38 MAPK is probably important for recall transcription of IFN-g and other cytokine genes. It seems to be part of various signaling cascades, but it probably does not induce or maintain a selective cytokine memory (Yamashita et al., 1999; Zhang et al., 1999b). C. JNK Similar to the p38 MAPK pathway, JNKs are also activated in response to in¯ammatory mediators, stress, and mitogens. Naive Th cells express only very low levels of JNKs but JNK expression is upregulated within days after T cell activation (Weiss et al., 2000). JNKs phosphorylate components of the AP-1 transcription factor complex, like c-Jun, and ATF-2. An inhibitory role for JNK1 in Th2 differentiation has been described based on the analysis of JNK1de®cient mice. These mice are characterized by enhanced T cell proliferation, moderately reduced activation-induced cell death, and enhanced Th2 development, possibly caused by an increased nuclear accumulation of NFATc (Dong et al., 1998). The contribution of NFATc to optimal Th2 induction had been shown earlier (Ranger et al., 1998a; Yoshida et al., 1998; and see below).

cytokine memory of t helper lymphocytes

135

Recently a mechanism linking JNK1 to NFATc activity has been proposed. JNK1 phosphorylates the calcineurin target domain of NFATc1-a, the major NFATc isoform expressed in T cells. This phosphorylation interferes with the binding of calcineurin to NFATc1-a and thereby inhibits dephosphorylation and nuclear translocation of NFATc1-a (Chow et al., 2000). JNK2 de®ciency leads to a reduction in early IL-12-induced IFN-g expression of Th cells, accompanied by a reduced IL-12Rb2 expression, while IL-4 induction and proliferation are not affected. The reduced IFN-g expression in JNK2-de®cient Th cells does not re¯ect a direct requirement of JNK2 for IFN-g transcription, since addition of exogenous IFN-g, which is known to enhance IL-12Rb2 expression (Gollob et al., 1997; Rogge et al., 1997; Szabo et al., 1997a), fully restored IL-12-induced expression of IFN-g upon restimulation (Yang et al., 1998a). Others have observed reduced production of IFN-g, IL-4, and IL-2 upon primary stimulation of JNK2-de®cient Th cells, suggesting that JNK2 may be relevant for ef®cient T cell activation in general (Sabapathy et al., 1999). Enhanced proliferative activity and production of elevated levels of type 2 cytokines, but normal levels of IFN-g, have been described for CD4 T cells from JNK2-de®cient mice expressing a dominant-negative JNK1 transgene (Dong et al., 2000). This phenotype is similar to what had already been observed in JNK1-de®cient T cells (Dong et al., 1998), con®rming the inhibitory in¯uence of JNK1 on Th2 differentiation. D. Rac2 The small guanosine triphosphatase Rac2 is expressed at elevated levels in murine Th1 but not Th2 cells. Constitutively active Rac enhances JNK, p38 MAP kinase, and NF-kB activity, and it leads to the transactivation of an IFN-g promoter construct in a p38- and NF-kB-dependent but JNK-independent manner. T cells transgenic for constitutively active Rac2 produce elevated levels of IFN-g but similar amounts of IL-4 compared with levels produced by nontransgenic T cells. Th1-polarized Rac2-de®cient Th cells show moderately reduced but still substantial IFN-g production (Li et al., 2000). Thus, in Th cells Rac2 is not indispensable for induction of IFN-g expression. It rather appears to amplify the expression of IFN-g, probably by enhancing the activity of p38 MAPK and NF-kB. Activated Rac1 binds to Stat3 and contributes to Stat3 phosphorylation, probably via Rac1-enhanced Jak2 activity, in response to epidermal growth factor stimulation (Simon et al., 2000). E. NF-kB Whether NF-kB is preferentially involved in the regulation of either type 1 or type 2 cytokines is not clear (reviewed in Akira and Kishimoto, 1997). Inhibition of the NF-kB pathway in transgenic mice impairs a Th1-mediated delayed-type hypersensitivity (DTH) response. A Th2-dependent allergic lung in¯ammation

136

max loÈhning et al.

is affected only partially (Aronica et al., 1999). In a Th2-dependent murine model of airway in¯ammation, however, de®ciency of the p50 subunit of NFkB strikingly reduces type 2 cytokine production, probably due to the requirement of NF-kB for the enhancement of GATA-3 and the induction of type 2 cytokine expression in the early course of Th2 differentiation. In contrast, NFkB inhibition reduces neither established GATA-3 expression nor type 2 memory cytokine expression. T-bet and IFN-g expression in primary Th1polarizing stimulations are not affected (Das et al., 2001). All in all, NF-kB seems to play a differentially pronounced role in the induction of both type 1 and type 2 cytokine memory. F. NFAT The nuclear factor of activated T cells (NFAT) transcription factor family consists of NFATc (NFATc1 or NFAT2), NFATp (NFATc2 or NFAT1), NFAT4 (NFATc3 or NFATx), and NFAT3 (NFATc4). All four NFATs contain a calcineurin binding domain and a DNA binding domain. T cell activation leads to an increase in intracellular Ca2‡ levels, which activates the protein phosphatase calcineurin. Active calcineurin can associate with and dephosphorylate NFATs resulting in nuclear translocation of the NFAT±calcineurin complex (Shibasaki et al., 1996). In human T cells, an involvement of Ca2‡ in¯ux and calcineurin activity in IL-4 and IL-10 gene expression has been shown (Feske et al., 2001). While NFATp and NFAT4 are constitutively expressed in T cells, NFATc expression is rapidly upregulated upon T cell stimulation. The impact of the NFAT family members on T cell activation has recently been reviewed in detail (Rao et al., 1997; Ser¯ing et al., 2000). Here we focus on recent ®ndings regarding the regulation of cytokine gene expression by the different NFATs. G. NFATc (NFATc1 or NFAT2) Upon stimulation of naive Th cells via TCR and CD28, the expression of two long isoforms of NFATc is rapidly upregulated. Differentiation into Th1 or Th2 effector cells is accompanied by a switch to the expression of a short isoform via the inducible usage of an alternative proximal polyadenylation site (Chuvpilo et al., 1999). NFATc-de®cient lymphocytes are defective in proliferation and induction of expression of type 2 but not type 1 cytokines (Ranger et al., 1998a; Yoshida et al., 1998). Defects of Th2 development have also been observed in mice de®cient for either the CD4 molecule (Brown et al., 1997a; Fowell et al., 1997) or the T-cell-associated Tec kinase Itk (Fowell et al., 1999), which is activated by the p56(lck) tyrosine kinase upon CD4 triggering (Heyeck et al., 1997; Turner et al., 1990). Mice expressing a dominant-negative p56(lck) kinase or a dominant-negative form of calcineurin show reduced Th2 development (Yamashita et al., 1998, 2000). The molecular basis for the apparent contribution but not necessity of CD4 signaling for Th2 differentiation may be the

cytokine memory of t helper lymphocytes

137

augmentation of the intracellular Ca2‡ concentration in response to TCR stimulation, mediated by ltk (Liu et al., 1998b). Defective Ca2‡ in¯ux in the absence of CD4/Lck/ltk activity is likely to result in reduced activation of the Ca2‡ -dependent phosphatase calcineurin. Reduced dephosphorylation and nuclear translocation of NFATc is the consequence, which may lead to reduced transcription of the IL-4 gene (Feske et al., 2001; Fowell et al., 1999). The transcriptional coactivator p300/CBP binds to the transactivation domain of NFATp and NFATc and enhances their transactivation activity (Avots et al., 1999; Garcia-Rodriguez and Rao, 1998). Notably, the intrinsic histone acetyltransferase activity of p300/CBP appears not to be required for the enhancement of the transactivation activity of NFATc by p300/CBP (Avots et al., 1999). Cotransfection of NFATc or NFATp and p300/CBP augments transcription of an IL-4 promoter construct (Sisk et al., 2000). Interestingly, NFATc has recently been reported to bind to GATA-3 as demonstrated by coimmunoprecipitation of the two factors, and NFATc and GATA-3 together transactivate the IL-5 promoter synergistically (Ser¯ing et al., 2000). At least some of the effects of NFATc on Th2 cytokine expression could be mediated through cooperation with GATA-3. The ability of NFATc to form a complex with the coactivator p300/CBP would then recruit an (histone) acetyltransferase function to the putative GATA-3/NFATc/(p300/CBP) complex, which could open the chromatin structure of GATA-3/NFATc target genes by local histone acetylation and also enhance GATA-3 activity by acetylation of GATA-3 itself (Boyes et al., 1998; Yamagata et al., 2000; see also Section III.B). H. NFATp (NFATc2 or NFAT1) and NFAT4 (NFATc3 or NFATx) In T cells, NFATp is the most abundant NFAT protein (Xanthoudakis et al., 1996). It is constitutively expressed and its nuclear localization is enhanced upon T cell stimulation (Cron et al., 1999). c-Maf acts in synergy with NFATp and NFAT-interacting protein (NIP) 45 in transactivation of the IL-4 promoter (Ho et al., 1996; Hodge et al., 1996a). Likewise, NFATp cooperates with AP-1 (Fos and Jun) in the activation of promoter constructs of several cytokines, like IL-2, IL-3, IL-4, and granulocyte-macrophage colony-stimulating factor (GM-CSF), and of Fas ligand, while NFATp can induce the TNF-a and IL-13 promoters without AP-1 (Macian et al., 2000). In vivo, NFATp binds to the IL-4 promoter and to the Th2-speci®c distal IL-4 enhancer to which also GATA-3 binds (Agarwal et al., 2000). It is not clear whether NFATc binds to this IL-4, enhancer as well. NFATp-de®cient mice are characterized by modest splenomegaly and hyperproliferation of T and B cells, reduced Fas ligand expression, and moderately enhanced type 2 cytokine production and Th2 immune responses (Hodge et al., 1996b; Kiani et al., 1997; Xanthoudakis et al., 1996). This phenotype is reminiscent of the hyperproliferative and preferentially Th2-biased phenotype of JNK1-de®cient mice, which is associated with a nuclear accumulation of

138

max loÈhning et al.

NFATc (Dong et al., 1998). Based on the observation of prolonged IL-4 gene expression in NFATp-de®cient T cells, the physiological role of NFATp in vivo seems to be to switch off IL-4 transcription and to antagonize the activating NFATc (Kiani et al., 1997). T cells de®cient for both NFATp and NFATc proliferate normally but they have defects in the expression of Th1 and Th2 cytokines, in Fas ligand induction, and in cytotoxic activity (Peng et al., 2001). Mice de®cient in NFAT4, which is preferentially expressed in CD4 and CD8 double-positive but also single-positive thymocytes and peripheral T cells, have reduced numbers of double-positive cells in the thymus and of mature CD4 and CD8 T cells in the periphery, accompanied by increased Fas ligand expression and higher sensitivity to apoptosis in response to TCR triggering. This suggests that NFAT4 may downmodulate signals from the TCR. However, TCR-dependent cytokine memory expression is normal in these mice (Oukka et al., 1998). Simultaneous de®ciency for both NFATp and NFAT4 leads to drastic splenomegaly, T cell hyperproliferation, and increased sensitivity to TCR engagement, indicating that both NFATp and NFAT4 act together or can back up each other in increasing the activation threshold of TCR signaling. T cells from NFATp/NFAT4 double-de®cient mice produce enhanced levels of Th2 cytokines and reduced amounts of Th1 cytokines, possibly due to constitutive nuclear localization of NFATc in the absence of its antagonists (Ranger et al., 1998b). Like in NFATp single-de®cient T cells (Hodge et al., 1996b), induction of Fas ligand (FasL) expression after TCR stimulation and activationinduced cell death are reduced in NFATp/NFAT4 double-de®cient T cells (Ranger et al., 1998b). Expression of the FasL-inducing transcription factor Egr3 is impaired in these cells. Egr3 expression is induced by NFATp and NIP45. Notably, Egr3 is expressed preferentially in Th1 but not in Th2 cells (Rengarajan et al., 2000) offering a molecular explanation for enhanced FasL expression and preferential Fas/FasL-mediated apoptosis of Th1 cells (Varadhachary et al., 1997; Zhang et al., 1997b). Taken together, TCR and some costimulatory signals are critical and suf®cient for memory expression of cytokine genes (Slifka et al., 1999). In the induction of cytokine memory, TCR signals can selectively enhance or inhibit both Th1 and Th2 development, and the particular effect in the individual situation may depend on the presence or absence of other differentiation factors, like T-bet or GATA-3. In any case, factors of the TCR signaling cascade seem to be essential members of multifactorial complexes in the induction of cytokine memory (lezzi et al., 1999; Richter et al., 1999). VI. Epigenetic Modifications of Cytokine Genes Apart from a selective setup of key transcription factors, like GATA-3 and T-bet, which control the induction and maintenance of Th cell cytokine

cytokine memory of t helper lymphocytes

139

memory, epigenetic modi®cation of cytokine genes imprints the original instruction for a particular cytokine memory in the genome of preactivated Th cells (see below and reviewed in Avni and Rao, 2000). Transcriptionally inactive gene loci are frequently characterized by condensed chromatin, in which DNA is tightly packed around nucleosomes with deacetylated histones. CpG nucleotides of transcriptional control regions are often methylated. In this state, transcription is inhibited due to reduced accessibility of the DNA for transcription factors and due to methylation-dependent recruitment of transcriptional repressors, like the methylated-CpG binding protein 2 (MeCP2) (Nan et al., 1997). Methyl-CpG binding proteins are able to recruit histone deacetylases to methylated DNA, which will stabilize the dense DNA±nucleosome structure (Jones et al., 1998; Nan et al., 1998). Transcriptionally ``open'' chromatin correlates with acetylated histones and DNA hypomethylation (reviewed in Bird and Wolffe, 1999; Hsieh, 2000; Robertson and Wolffe, 2000). Within minutes after TCR stimulation of naive T cells, the chromatin-remodeling BAF complex associates with their chromatin, contributing to activationinduced chromatin decondensation in T cells (Zhao et al., 1998). Thirty minutes after stimulation of naive Th cells via TCR and CD28, transient transcription of both the IL-4 and the IFN-g genes is initiated, independent of exogenous inductive cytokines and Stat6 or Stat4 signaling. Apparently, these effector cytokine loci are accessible in naive CD4 T cells (Grogan et al., 2001). For sustained transcription of the IL-4 and IFN-g genes and for the establishment of a cytokine memory, Stat6 or Stat4 signaling and induction of GATA-3 and T-bet are required. Early transient transcription of the IL-4 gene under Th1-polarizing conditions has also been demonstrated for naive Th cells from mice with a targeted insertion of a bicistronic IL-4 reporter gene (Grogan et al., 2001). These observations of early low-level transcriptional activity at the IL-4 locus irrespective of cytokine memory-inducing conditions are in line with a previous report on the ablation of Th cells transgenic for a toxic gene under the control of the IL-4 promoter. Transgenic cells were eliminated during IL-4- or IL-12polarized activations (Kamogawa et al., 1993). The low level of expression of the IL-4 gene in freshly activated Th cells is re¯ected in the inability to demonstrate IL-4 transcripts at 6 or 24 h after onset of activation but only at 48 h after onset of activation. IFN-g transcription has been detected already at 6 h and exclusively in IL-12-polarized cells (Lederer et al., 1996). Moreover, only IL-2 and not IL-4 protein is detectable by intracellular cytokine staining in naive Th cells after stimulation for 5 h in cells which later developed a memory for IL-4 expression within about 40 h of primary activation (Richter et al., 1999). The hypothesis that in naive Th cells effector cytokine gene loci are maintained in an at least partially accessible conformation is supported by the ®nding that in naive cells the cytokine genes are not located in heterochromatin. In Th1 cell lines the IL-4 gene locus and in Th2 cell lines the IFN-g gene

140

max loÈhning et al.

locus are frequently repositioned to heterochromatin (Grogan et al., 2001). Additional epigenetic changes at effector cytokine gene loci and the appearance of DNase I-hypersensitive sites, indicating binding of speci®c factors to the DNA, have been observed in polarized Th cells. It remains to be shown to what extent Ikaros proteins contribute to the repositioning (Brown et al., 1997b, 1999; Kim et al., 1999a). A. DNase I Hypersensitivity and DNA Demethylation A relaxed chromatin structure, i.e., a stretch of DNA with a low density of nucleosomes as indicated by increased sensitivity to DNase I digestion, is observed within the ®rst days of Th1 and Th2 differentiation for the respective effector cytokine genes, the IFN-g and the IL-4, IL-5, and IL-13 genes (Agarwal and Rao, 1998; Agarwal et al., 2000; Takemoto et al., 1998, 2000). For the IFN-g, IL-3, IL-4, IL-5, and IL-13 genes demethylation of particular CpGs has been shown in differentiated T cells (Agarwal and Rao, 1998; Bird et al., 1998; Fitzpatrick et al., 1998, 1999; Melvin et al., 1995; Young et al., 1994). The genes coding for the type 2 cytokines IL-4, IL-5, and IL-13 are located in close proximity on human chromosome 5 and mouse chromosome 11. Coordinate induction of DNase I-hypersensitive sites at the IL-4 and IL-13 genes has been observed in the course of Th2 polarization, and even the basal transcriptional activity of the murine RAD50 gene, which is located between the IL-5 and IL13 genes, is strongly increased in a Th2 clone compared with a Th1 clone (Agarwal and Rao, 1998). Between the IL-4 and IL-13 genes, a conserved noncoding sequence (CNS-1) has been identi®ed that is suggested to act as a coordinate regulator of the chromatin structure of the IL-4/IL-13/IL-5 gene cluster and is marked by two Th2-speci®c DNase I-hypersensitive sites (Loots et al., 2000; see also Section III.A). The molecular basis of induction of DNase I-hypersensitive sites and its connection to local DNA demethylation are not clear. DNA demethylation could be achieved ``passively'' in the course of DNA replication, maintaining the newly synthesized DNA strand in a demethylated state via local inhibition of DNA methyltransferases. Alternatively, active demethylation might occur, although the bare existence of DNA demethylases is still a matter of debate (reviewed in Hsieh, 2000). Sequence- or site-speci®c demethylating enzymes have not been demonstrated. Targeted recruitment of ``unspeci®c'' demethylation machineries to particular DNA regions via interaction with sequencespeci®c proteins appears more likely. Regarding histone modi®cations, binding of the TCR- and/or cytokine signal-activated transcription factors NFATp, NFATc, and Stat6 to the transcriptional coactivator p300/CBP, which contains histone acetyltransferase activity, has been reported (Avots et al., 1999; GarciaRodriguez and Rao, 1998; Gingras et al., 1999). This may focus histone acetylation and thus chromatin relaxation selectively to the region of cytokine pro-

cytokine memory of t helper lymphocytes

141

moters and regulatory elements with binding sites for these transcription factors. NFATc can bind to GATA-3 (Ser¯ing et al., 2000). A hypothetical GATA-3/NFATc/(p300/CBP) complex would then recruit histone acetyltransferase activity to GATA-3/NFATc target sequences. This seems to be the case for GATA-1, which is important for the development of the erythroid lineage and which can directly interact with p300/CBP (Blobel et al., 1998; Boyes et al., 1998). GATA-1 is involved in the formation of DNase I-hypersensitive sites in the b-globin locus control region (Boyes and Felsenfeld, 1996; Stamatoyannopoulos et al., 1995). GATA-3 binds to CNS-1 in the IL-4/IL-13/IL-5 gene cluster (Takemoto et al., 2000) and to an IL-4 enhancer located in a stimulation-induced, Th2speci®c DNase I-hypersensitive site downstream of the IL-4 locus (Agarwal et al., 2000). Ectopic expression of GATA-3 in Stat6-de®cient Th cells is suf®cient to induce the Th2-speci®c pattern of DNase I hypersensitivity of the IL-4 locus (Lee et al., 2000; Ouyang et al., 2000). Thus, the contribution of Stat6 to chromatin remodeling of the IL-4 locus is probably only indirect, via the induction of GATA-3. As demonstrated in c-maf-de®cient CD4 T cells, c-Maf is not indispensable for the formation of DNase I-hypersensitive sites in the IL4 locus (Agarwal et al., 2000). Thus, GATA-3 itself remains a likely candidate to mediate the recruitment of further chromatin-remodeling factors, e.g., histone acetyltransferases, to the IL-4 gene cluster and to prevent the repositioning of this cluster to heterochromatin. Whether T-bet might play a similar role in chromatin modi®cation of the IFN-g locus during Th1 development is not clear. B. Cell Cycle Dependence A close link between epigenetic modi®cation and the capacity to remember expression of modi®ed cytokine genes is strongly suggested by the observation that upon primary activation of naive Th cells under Th1- or Th2-polarizing conditions such cells must progress into the S phase of the ®rst cell cycle after the onset of activation to establish a memory for IFN-g, IL-4, or IL-10. The initial transition into the S phase, i.e., the start of DNA replication, obviously is required and suf®cient to establish the cytokine memory. Several groups have shown this by using inhibitors that arrest progression of the cells at various stages of the cell cycle (Bird et al., 1998; Richter et al., 1999). The initial DNA synthesis of the ®rst S phase may be a window of opportunity for chromatin remodeling by nucleosome disassembly and DNA hemidemethylation. While induction of the cytokine memory for IFN-g, IL-4, and IL-10 is blocked by mycophenolic acid or l-mimosine, i.e., is dependent on entry into the S phase, the memory of Th cells to express IL-2 is independent of entry into the cell cycle (Bird et al., 1998; Richter et al., 1999; A. Richter, unpublished data), perhaps due to acquisition of a memory for IL-2 during thymic maturation (Rothenberg and Ward, 1996).

142

max loÈhning et al.

Although completion of cell division is not required for the induction of a cytokine memory for IFN-g (Bird et al., 1998) or IL-4 and IL-10 (Richter et al., 1999), the frequency of cells with such cytokine memories correlates positively with the proliferation of activated Th cells (Ben-Sasson et al., 2001; Bird et al., 1998; Gett and Hodgkin, 1998; Richter et al., 1999). It does not, however, correlate with the number of cell cycles, as has been claimed, suggesting an intrinsic cell cycle counting mechanism of Th cells that would differentially regulate the accessibility of certain effector cytokine loci (Bird et al., 1998; Gett and Hodgkin, 1998; reviewed in Reiner and Seder, 1999). However, the frequencies of cells with an established cytokine memory correlate with the time elapsed since the onset of T cell activation and not with the number of cell cycles the cells have performed in this time (Ben-Sasson et al., 2001; Richter et al., 1999). It seems plausible that DNA synthesis during cell division offers a periodical opportunity for chromatin remodeling of cytokine gene loci, but apparently other factors, which are dependent on time rather than cell division, are decisive in the establishment of cytokine memory. Concomitant signaling from the TCR and the IL-4R is required for ef®cient Th2 differentiation of naive Th cells (lezzi et al., 1999; Richter et al., 1999). Notably, even when the entry into the initial S phase is blocked by l-mimosine, a pulse of IL-4 signal can still induce a commitment to IL-4 memory, which is established later, when the cell is allowed to progress through the initial S phase by release from l-mimosine, within a day (Richter et al., 1999). Thus, simultaneous triggering of the TCR and the IL-4R apparently activates one or several Th2 differentiation factors, independent of DNA synthesis, that are stably active for at least 1 day. C. Allelic Probabilities of Cytokine Memory The proposal that epigenetic modi®cation of cytokine gene loci is a critical molecular parameter of cytokine memory leads to the question whether memory for expression is a stable characteristic of individual cytokine gene loci. For murine Th cells heterozygous for IL-2 or IL-4 alleles, monoallelic expression of these cytokines has been reported (Bix and Locksley, 1998; HollaÈnder et al., 1998), as well as for human CD4 T cells expressing IL-2 (Matesanz et al., 2000). These observations prompted speculations about allelic exclusion of cytokine gene expression (Chess, 1998), although there still remains some confusion on the extent of monoallelic expression. The number of cells analyzed in these original analyses by single-cell RT-PCR was rather low. Of individual murine Th2 clones, approximately half showed monoallelic IL-4 expression and the other half biallelic IL-4 expression, with the allelic expression pattern of the individual clones remaining constant over time (Bix and Locksley, 1998). Based on that result, it was suggested that a stochastic process may stably activate or inactivate individual IL-4 alleles. Recently, in

cytokine memory of t helper lymphocytes

143

Th2 clones with monoallelic expression of IL-4, a preference for coexpression of IL-5 or IL-13 from the same allelic locus has been described, a ®nding in support of the idea of coordinate regulation of the accessibility of individual IL-4/IL-13/IL-5 gene loci (Kelly and Locksley, 2000). Another line of evidence for monoallelic expression of cytokine genes comes from targeted insertion of reporter genes downstream of the IL-2 or IL-4 promoters into the murine germline and cytometric analysis of large numbers of individual Th cells heterozygous for these insertions, con®rming the existence of monoallelic cytokine expression. Biallelic versus monoallelic expression occurred at stochastic or even slightly positively correlated frequencies, favoring a model of independent regulation of both alleles in individual T cells (Hu-Li et al., 2001; Naramura et al., 1998; Riviere et al., 1998). The probability of biallelic cytokine expression has been suggested to be dependent on the strength of the TCR signal (Chiodetti et al., 2000; Riviere et al., 1998). While monoallelic expression of cytokine genes as such has clearly been demonstrated, it is less clear whether the memory for cytokine expression is also allele speci®c. In the Th2 clones analyzed by Bix and Locksley (1998) allelic preference was memorized. Murine Th2-polarized Th cells heterozygous for wildtype IL-4 and an IL-4 gene with insertion of a ¯uorescent reporter gene, when isolated according to expression of one or the other, both, or none of the IL-4 alleles, could not memorize their allelic preference at the time of isolation in later restimulations (Hu-Li et al., 2001). Already after 3 days of Th2 polarization, essentially all cells were competent to express each of the IL-4 alleles, but in a given restimulation they did reexpress each IL-4 allele only with a certain probability. The early establishment of memory for both alleles of a cytokine gene of a cell is ef®cient and probably occurs fast and simultaneously, but memory expression upon restimulation is limited by as yet unknown mechanisms. Both IL-4 producers or nonproducers of such polarized Th2 populations express comparable levels of GATA-3 (Hu-Li et al., 2001). However, in the offspring of cloned Th2 cells, isolated according to mono- or biallelic expression of IL-4, the ratios of cells expressing one or the other allele upon restimulation are biased and this bias is a stable property of the individual cells of a given clone (Hu-Li et al., 2001; Riviere et al., 1998). Thus, in Th2 cells polarized for several weeks and restimulated several times, individual alleles of the IL-4 gene are memorized with characteristic probabilities, re¯ecting different accessibilities. It remains to be shown to what extent this accessibility is due to epigenetic modi®cation of the IL-4 locus. D. Scenario for the Induction and Maintenance of Cytokine Memory Summarizing the experimental evidence, a scenario for the induction and maintenance of the memory for IL-4 may be proposed (Fig. 2). In naive

144

max loÈhning et al.

Fig. 2. Scenario for the induction and maintenance of cytokine memory. Cytokine memory at the IL-4 gene locus has been characterized in some detail and is depicted here to exemplify general mechanisms, such as epigenetic modi®cation of cytokine genes and speci®c expression of transcription factors. For details, see text.

Th cells, the accessibility of the IL-4 locus appears to be limited due to a relatively compact chromatin structure and local DNA methylation. Speci®c transcription factors for expression of the IL-4 gene, like GATA-3 and c-Maf, are absent or expressed at low levels (Fig. 2A). Upon activation of the naive Th cell, simultaneous signaling via the TCR and the IL-4R induces transient

cytokine memory of t helper lymphocytes

145

activation of Stat6 and of several nonspeci®c, TCR-dependent transcription factors. Activated Stat6 enhances expression of GATA-3. GATA-3 stabilizes and enhances its own gene expression by autoactivation. GATA-3 also induces expression of c-maf (Ouyang et al., 2000). At this stage of activation, the IL-4 gene locus is probably still poorly accessible due to epigenetic constraints (Fig. 2B). Transient expression of the IL-4 gene at low level can occur at this stage, independent of Stat6 and GATA-3 (Grogan et al., 2001). Upon entry into the S phase of the ®rst cell cycle of activation-induced proliferation, nucleosome disassembly and DNA hemimethylation may occur and provide a window of opportunity for binding of speci®c transcription factors, e.g., GATA-3, to regulatory regions of the IL-4 gene locus (Fig. 2C). Such transcription factors would recruit chromatin-modifying enzymes to their target sequences, e.g., histone acetyltransferases and DNA demethylases, or prevent the deacetylation of histones and methylation of DNA by blocking the access of the respective enzymes to the chromatin. As a result, the chromatin and DNA of regulatory sequences of the IL-4 gene locus would become accessible for TCR-induced transcription factors. This accessibility would make cytokine memory expression dependent on TCR signaling and independent of IL-4R signaling (Fig. 2D). Since TCR signaling is transient, cytokine memory expression of IL-4 would also be transient. In naive Th cells activated by TCR signaling in the absence of the inductive signals required for chromatin remodeling and epigenetic modi®cation, the cytokine gene loci would remain in their naive state. Such memory Th cells would still be naive with respect to cytokine memory and might acquire a cytokine memory upon later restimulations in the presence of memory-inducing signals. In naive Th cells activated by TCR signaling in the presence of inductive signals for particular cytokines, other cytokine genes may be actively silenced, as seems to be the case for the IL-4 gene locus in polarized Th1 cells, which can be sequestered to transcriptionally inactive heterochromatin (Grogan et al., 2001). VII. Stability and Plasticity of Cytokine Memory Knowledge about the stability and plasticity, i.e., potential reversibility of cytokine memory, is critical for the design of therapeutic strategies in immune disorders with immunopathology caused by chronic or repeated T cell activation, e.g., autoimmunity and allergy. Murine Th cells polarized in vitro by IL-12 or IL-4 to differentiate into Th1 or Th2 cells show different degrees of stability of their cytokine memory, depending on the duration of polarization. After approximately 1 week of Th1 polarization of murine CD4‡ T cells, reactivation of the cell population in the presence of IL-4 resulted in the induction of memory for IL-4 and the reduction of memory for IFN-g (Murphy et al., 1996;

146

max loÈhning et al.

Perez et al., 1995; Szabo et al., 1995). After 2 to 3 weeks of Th1 polarization by IL-12, the cytokine pattern of the Th1 population remained largely unaffected by subsequent restimulation in the presence of IL-4. A moderate reduction of IFN-g memory was observed (Hu-Li et al., 1997; Murphy et al., 1996; Sornasse et al., 1996). The memory for IL-4, at least in the mouse, seems to be stabilized somewhat earlier, such that already after 1 week of culture in IL-4, treatment with IL-12 would hardly suppress IL-4 or induce IFN-g memory (Murphy et al., 1996; Perez et al., 1995; Szabo et al., 1995). Recently, a relationship between cell proliferation and the loss of plasticity of cytokine memory has been suggested. In cells that had divided one to three times during primary activation under Th1- or Th2-polarizing conditions, adversely polarizing cytokines could still modulate the memory for IFN-g and IL-4, while in cells that had divided four or more times they could not (Grogan et al., 2001). In vivo, the cytokine memory of in vitro-generated, adoptively transferred Th1 and Th2 populations can persist for months (Hu-Li et al., 1997; Mocci and Coffman, 1995; Swain, 1994). The available evidence suggests that, while some plasticity in the cytokine memory particularly of early Th1 populations may exist, the potential to revert a given cytokine memory is lost gradually after repeated restimulations under either Th1- or Th2-polarizing conditions. When analyzed on the level of individual cells, the question of stability and plasticity of cytokine memory is more complex, mostly due to the heterogeneity of polarized T cells with regard to their cytokine production (Assenmacher et al., 1994; Bucy et al., 1994; Openshaw et al., 1995). Since not all cells of a polarized Th1 or Th2 population actually memorize IFN-g or IL-4 in recall stimulations, on the level of populations it is not clear whether T cells responding to adverse instruction signals had already acquired a cytokine memory and reversed it or were derived from uncommitted cells. Isolated, individual IFN-g- but not IL-4producing murine Th cells, polarized for 1 week with IL-12 (Assenmacher et al., 1996; Scheffold et al., 2000), when restimulated for 5 days in the presence of IL-4, almost completely lost their memory for IFN-g and acquired memory for IL-4, IL-5, and IL-10 at frequencies similar to those of naive cells activated in the presence of IL-4. When polarized for longer time periods with IL-12, the IFN-g cytokine memory became increasingly resistant to IL-4. No IL-4 memory could be induced after 3 weeks and no IL-10 memory after 4 weeks. Even after 4 weeks the Th1 cells could be induced to memorize IL-5, a clear indication that they could still respond to IL-4 and be able to express a gene of the IL-4/IL-13/IL-5 gene cluster. After about 4 weeks, the memory for IFN-g became resistant to IL-4 (Assenmacher et al., 1998b). Thus, the memories for the various cytokine genes are independent of each other and of different stabilities at least with respect to the converting signals analyzed. The memory for IL-4 is refractory to IL-12 after 1±2 weeks of Th2 polarization, while the memory for IFN-g becomes refractory to IL-4 only after 3±4 weeks.

cytokine memory of t helper lymphocytes

147

The relatively early stabilization of the IL-4 memory in Th2 cells, which become resistant to conversion by IL-12 after 1 week, at least in part may be due to downregulation of IL-12Rb2 chain expression in the course of Th2 polarization (Rogge et al., 1997; Szabo et al., 1995, 1997a; see also Section II.A.1), mediated by GATA-3 (Ouyang et al., 1998). Thus, it has been speculated that the loss of IL-12 responsiveness may be the molecular basis for stable commitment to the Th2 lineage (Guler et al., 1996; Szabo et al., 1995). Functional IL-12 responsiveness of 1-week-polarized murine Th2 populations could be restored by the addition of IFN-g, which prevents IL-12Rb2 downregulation and, in conjunction with IL-12, also induces memory for IFN-g (Hu-Li et al., 1997; Szabo et al., 1997a). In contrast, transgenic or ectopic expression of the IL-12Rb2 chain in developing murine Th2 cells resulted in activation of the IL12 signaling molecule Stat4 and in enhanced proliferation, but not in induction of IFN-g memory (Heath et al., 2000b; Nishikomori et al., 2000). Thus the forced maintenance of IL-12 signaling in murine Th2 cells is not suf®cient as such to make Th2 cells respond to IL-12 by IFN-g expression, while IL-12 plus IFN-g signaling may induce IFN-g memory in such cells but does not extinguish IL-4 memory. This points to a role for IFN-g signaling in IFN-g memory induction in addition to its IL-12Rb2-maintaining function. In human Th2 cells, expression of the IL-12Rb2 chain is also downregulated (Rogge et al., 1997). Nevertheless, TCR restimulation of human Th2 cell populations polarized for 2±3 weeks with IL-4 and of human Th2 clones in the presence of IL-12 still resulted in substantial induction of IFN-g expression as well as in upregulation of IL-12Rb2 and T-bet and downregulation of GATA3 (Smits et al., 2001; Sornasse et al., 1996; Yssel et al., 1994). Individual Th2 cells polarized for 3 weeks with IL-4 and isolated according to secretion of IL-4 but not IFN-g by the cellular af®nity matrix technology (Assenmacher et al., 1998a; Brosterhus et al., 1999; Manz et al., 1995), frequently responded to restimulation in the presence of IL-12 with coexpression of IFN-g and IL-4; i.e., in many cells the IL-4 memory is maintained and a memory for IFN-g is acquired independently (Smits et al., 2001). Since in the human studies no neutralizing anti-IFN antibodies had been added during Th2 polarization, endogenous production of IFNs may account for the observed responsiveness to IL-12, maintaining low-level expression of IL-12Rb2 and providing IFN-g signaling (Gollob et al., 1997; Rogge et al., 1997; see also Section II.A.1). In contrast to the selective downregulation of IL-12Rb2 expression on Th2 cells but not on Th1 cells (Rogge et al., 1997; Szabo et al., 1997a), expression of a functional IFN-aR is maintained on both Th lineages. Since in human but not in murine Th1 and Th2 cells IFN-a stimulation results in Stat4 phosphorylation (Cho et al., 1996; Rogge et al., 1998; see also Section II.A.4), IFN-a-mediated Stat4 activation in human Th2 cells theoretically would be a way to induce IFN-g expression in Th2 cells. However, neither human Th2 clones nor Th2

148

max loÈhning et al.

cells polarized for 2 weeks expressed IFN-g in response to TCR and IFN-a stimulation (Rogge et al., 1998). Thus, activated Stat4 as such, and as induced by IL-12 or IFN-a, is not suf®cient to induce IFN-g memory and silence IL-4 memory in human Th2 cells. Similar to the defect of Th2 cells in IL-12 responsiveness (Rogge et al., 1997; Szabo et al., 1995, 1997a), murine Th1 clones and Th1 cells polarized for 2 weeks with IL-12 show an impairment in IL-4 signaling, as evidenced by reduced phosphorylation of Jak3 and Stat6 in response to IL-4 stimulation, although IL4R densities on the surfaces of Th1 and Th2 cells are comparable (Huang and Paul, 1998; Kubo et al., 1997). The reduced phosphorylation of signal transduction molecules downstream of the IL-4R may be due to activity of SOCS-1, expression of which has been reported to be induced by IFNs in human monocytes (Dickensheets et al., 1999; Losman et al., 1999; see also Section II.B.2). Transient inhibition of signaling in response to IL-4, IL-2, IL-6, and IFN-a could also be due to TCR and/or CD28 signaling, possibly mediated by activities of the MAP kinase and calcineurin pathways (Zhu et al., 2000; see also Section V.A). In attempts to analyze the molecular basis of the stabilization of Th1 cytokine memory in more detail, several groups have expressed activated Stat6 or GATA3 ectopically in short- and long-term polarized murine Th1 cells and Th1 clones (Farrer et al., 2001; Ferber et al., 1999; Kurata et al., 1999; Lee et al., 2000; Ouyang et al., 1998, 2000; Zheng and Flavell, 1997). Forced expression of an inducible, activated Stat6 in developing Th1 cells results in induction of expression of GATA-3, c-maf, and Th2 cytokines and inhibition of IL-12Rb2 and IFNg expression. This capacity of Stat6 to deviate Th1 differentiation is lost after 2 to 3 weeks of IL-12 polarization (Kurata et al., 1999). Retroviral transduction of GATA-3 expression leads to strong induction of IL-4 in polarized Th1 cells during the initial phase of T cell activation. Already after 1 week of IL-12 polarization, the ability of GATA-3 to induce IL-4 and inhibit IFN-g memory is nearly lost, and it is almost absent in Th1 clones (Lee et al., 2000; Ouyang et al., 1998). GATA-3 autoactivation of transcription of its own gene has been observed in IL-4- or Stat6-de®cient, IL-12-polarized Th cells transduced with GATA-3 early after activation, like in Th2 cells (Ouyang et al., 2000; see also Section III.A and B). In an established Th1 clone, however, ectopic GATA-3 expression does not seem to activate the endogenous GATA-3 gene (Lee et al., 2000). Thus, in the course of advanced Th differentiation, mechanisms precluding the expression of transcription factors are also likely to contribute to the stability of cytokine memory and the irreversibility of Th cell commitment. There is an ongoing debate whether cytokine memory in T cells is based on instruction or selection by cytokine signals (Coffman and Reiner, 1999). The instructive model implies that instructive cytokine signals, like IL-12, IFN-a, or IL-4, directly induce the memory for a certain cytokine. In the selective model, commitment to a certain cytokine memory occurs at random upon activation of

cytokine memory of t helper lymphocytes

149

the T cell. Cytokines then selectively expand those T cells that have acquired particular cytokine memories. To clarify this issue, Th cells activated for 1 week in vitro in the presence of antibodies neutralizing IL-12, IFN-g, and IL-4, have been sorted according to secretion of IL-4 by the cellular af®nity matrix technology (Manz et al., 1995; Ouyang et al., 2000). Stimulation of sorted IL-4-negative cells with IL-4 induced IL-4 memory in these cells in a Stat6dependent manner (Farrar et al., 2001). Even more convincingly, in sorted individual Th1 cells expressing IFN-g but not IL-4, a memory for IL-4 and IL10 could be induced by IL-4 (Assenmacher et al., 1998b). Both studies argue in favor of instruction of Th cells by IL-4 for an IL-4 memory. Along the same lines, ectopic GATA-3 expression in Stat6-de®cient Th cells increased the frequency of IL-4 producers progressively with time, with no selective outgrowth of the GATA-3-transduced T cells (Farrar et al., 2001). Taken together, IL-4 and GATA-3 appear to have the capacity to instruct a T cell for IL-4 cytokine memory, and this effect does not require selective expansion of subpopulations of Th cells, arguing against a purely selective mechanism in the generation of Th2 cytokine memory. In analogy to the potential of GATA-3 to modulate the cytokine memory of developing Th1 cells, the Th2 cytokine memory can be reversed by the candidate key transcription factor of the Th1 developmental program, T-bet. In murine Th2 cells polarized for up to 3 weeks, and in type 2-polarized CD8‡ cytotoxic T cells, ectopic T-bet expression is able to induce high frequencies of IFN-g producers while the percentages preferentially of IL-5 but also of IL-4 producers are reduced. In Th2 clones, marked repression of IL-5 and IL-4 memory expression but only little induction of IFN-g expression could be achieved by ectopic T-bet (Szabo et al., 2000; see also Section III.D). Thus, T-bet is able to extinguish the memory for IL-5 and to some extent also for IL-4, by mechanisms not clear so far and probably involving downregulation of GATA-3, and to induce the memory for IFN-g in established Th2 cells. In summary, in the progress of Th cell differentiation, Th cells are instructed to memorize particular cytokines. This cytokine memory is stabilized on several levels, including the selective expression of transcription factors, cytokine receptors, epigenetic modi®cation of the cytokine genes to be memorized, and epigenetic inactivation of those not to be memorized. Some of these molecular events obviously make the cells refractory to instructive signals of adverse T cell developmental programs. VIII. Cytokine Memory as Part of T Cell Differentiation Programs The differences between the Th1 and Th2 developmental programs are not restricted to different cytokine memories. A variety of other molecules are differentially expressed as well.

150

max loÈhning et al.

A. Differential T Cell Recruitment to Target Tissues: Selectin Ligands and Chemokine Receptors 1. Selectin Ligands Type 1 cytokine-producing Th cells contribute to the control of in¯ammatory immune reactions. The local encounter of Th1 cells with other proin¯ammatory effector cells may be promoted by the preferential recruitment of Th1 cells into in¯amed tissue, due to preferential expression of the ligands for P- and Eselectin by Th1 cells (Austrup et al., 1997). Expression of P- and E-selectin (CD62P and CD62E) is induced on vascular endothelial cells by in¯ammatory mediators, like TNF-a, IL-1, histamine, thrombin, or bacterial endotoxin (reviewed in Kansas, 1996). Fucosylation of certain carbohydrates by the enzyme a1,3-fucosyltransferase VII (FucT-VII) is required for the formation of functional selectin ligands (Maly et al., 1996). A second fucosyltransferase, FucT-IV, is expressed by leucocytes and may contribute to optimal generation of active P- and E-selectin ligands (Weninger et al., 2000). Th1 but not Th2 cells generated in vitro bind to P-selectin (Austrup et al., 1997; Borges et al., 1997) via the functionally activated form of P-selectin glycoprotein ligand-1 (PSGL-1) (Moore et al., 1992; Norgard et al., 1993; Sako et al., 1993). Although PSGL1 expression as such is comparable on Th1 and Th2 cells, the P-selectin-binding form generated by FucT-VII is expressed preferentially on Th1 cells (Borges et al., 1997). In vitro, expression of FucT-VII is induced upon activation of CD4‡ T cells, strongly enhanced by the addition of IL-12 and may be somewhat reduced by IL-4 signaling (Lim et al., 1999; Wagers et al., 1998). Although IFNg and FucT-VII are both induced by IL-12, they are not mandatorily coexpressed in individual cells, which is particulary true for IFN-g producers generated in vivo (Syrbe et al., 1999; Thoma et al., 1998; Tietz et al., 1998). 2. Chemokine Receptors Chemokine signals control the attraction of lymphocytes to distinct tissues and cells (for recent reviews, see Loetscher et al., 2000; Moser and Loetscher, 2001). Selective recruitment of functionally distinct T cells is achieved by regulated expression of particular sets of chemokine receptors. In general, the expression of many chemokine receptors and, consequently, chemotactic migration are modulated by activation of the T cells via the TCR and often depend also on the presence of cytokines like IL-2 or IL-4 (Bleul et al., 1997; D'Ambrosio et al., 1998; Jourdan et al., 1998; Loetscher et al., 1996; Patterson et al., 1999; Sallusto et al., 1999a; Zingoni et al., 1998). CD4‡ and CD8‡ T cells, polarized in vitro with IL-12, and T cells identi®ed in vivo in in¯amed tissue preferentially express the chemokine receptors CCR5 and CXCR3, the ligands of which are proin¯ammatory chemokines (Bonecchi et al., 1998; D'Ambrosio et al., 1998; Loetscher et al., 1998; Qin et al., 1998;

cytokine memory of t helper lymphocytes

151

Sallusto et al., 1998). Th2 cells also express CCR5 and CXCR3 but less than Th1 cells (Bonecchi et al., 1998; Loetscher et al., 1998; Sallusto et al., 1998). Type 2polarized CD4‡ and CD8‡ T cells show preferential but not exclusive expression of CCR3, CCR4, and CCR8 (Andrew et al., 2001; Bonecchi et al., 1998; D'Ambrosio et al., 1998; Sallusto et al., 1997, 1998; Zingoni et al., 1998). The ligand for CCR3, eotaxin (CCL11), expression of which is enhanced in atopic asthma (Ying et al., 1997), and one of the two CCR4 ligands, macrophagederived chemokine (MDC) (CCL22), have been described to coordinately and sequentially contribute to the recruitment of Th2 cells to the lung in a murine model of allergic airway in¯ammation (Lloyd et al., 2000). In mice de®cient for CCR4, allergic airway in¯ammation is not impaired and no defects in cellular recruitment, bronchial hyperreactivity, and allergen-speci®c antibody production are observed, pointing to the redundancy of CCR4 (Chvatchko et al., 2000). The interaction of CCR4 with its second ligand, thymus and activation-regulated chemokine (TARC) (CCL17), seems to be involved in the selective recruitment of a subset of human memory T cells to in¯amed skin in psoriasis, but not to in¯amed intestine in Crohn's disease (Campbell et al., 1999). In skin from allergic patients, expression of CCR8 and its ligand I-309 (CCL1) is enhanced after exposure to allergen (Sebastiani et al., 2001). In CCR8-de®cient mice, a reduced accumulation of eosinophils in the lung in allergic airway in¯ammation and diminished local type 2 cytokine production have been described (Chensue et al., 2001). However, neither CCR8 nor CCR4 de®ciency affect IL-4-driven Th2 development in vitro, indicating that CCR8 is not directly required for the differentiation of Th2 cells but probably is required for the appropriate migration of Th2 cells (Chensue et al., 2001; Chvatchko et al., 2000). Similar to the expression pattern of CCR3, the seven-transmembrane receptor chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2) is also expressed on human eosinophils, on basophils, and on subsets of CD4‡ and CD8‡ T cells from peripheral blood. CRTH2-expressing T cells show a strong preference for production of type 2 but not type 1 cytokines (Cosmi et al., 2000; Nagata et al., 1999a,b). Among peripheral blood T cells, CCR3 expression is by far less frequent than CRTH2 or CCR4 expression (Cosmi et al., 2000). CRTH2 binds to prostaglandin D2, which is secreted by activated mast cells and is considered to be a mediator of allergic in¯ammation. By binding to CRTH2, prostaglandin D2 induces chemotaxis of Th2 cells, eosinophils, and basophils (Hirai et al., 2001). This points to a function of CRTH2 in the coordinate recruitment of Th2 cells, eosinophils, and basophils to sites of mast cell activity and local type 2 immune responses, like in IgEmediated allergy. Taken together, the apparent ®ne-tuning in the expression of chemokine receptors and prostanoid receptors, which in part seems to be coordinately regulated together with other features of T cell differentiation

152

max loÈhning et al.

programs, may contribute to the ef®cient recruitment of specialized subsets of T cells to their target tissues. B. Differential Regulation of T Cell Effector Functions: Cytokine Receptors and Costimulators Selective expression of cytokine receptors as part of T cell differentiation programs may serve several functions. One may be to stabilize T cell differentiation programs by switching off the susceptibility to adverse differentiation signals. The preferential maintenance of expression of the IL-12Rb2 chain on differentiated Th1 cells and its downregulation by GATA-3 in the course of Th2 development have already been discussed (Ouyang et al., 1998; Rogge et al., 1997; Szabo et al., 1997a; and see Sections II.A.1 and VII). Differentially expressed cytokine receptors may also provide an opportunity to modulate the activity of selected T effector cell subsets in situ. Selected examples are discussed here. 1. IL-18R IL-18R, a member of the type I IL-1R family previously called IL-1R-related protein (IL-1Rrp) (Parnet et al., 1996; Torigoe et al., 1997), is preferentially expressed on murine Th1 but not Th2 cells (Xu et al., 1998b). IL-18R expression is upregulated by IL-12. In turn, IL-18 signaling enhances expression of IL-12Rb2 (Ahn et al., 1997; Xu et al., 1998b; Yoshimoto et al., 1998). Treatment with anti-IL-18R antibodies reduces local in¯ammation and LPS-induced shock in association with decreased production of proin¯ammatory cytokines (Xu et al., 1998b). In IL-18-de®cient mice, IFN-g induction is defective in vivo, and cytotoxic activity of NK cells is reduced (Takeda et al., 1998). The IL-18mediated enhancement of IL-12-induced Th1 differentiation and the TCRindependent expression of the IFN-g memory in both CD4‡ and CD8‡ T cells, induced by stimulation of the cells with IL-12 plus IL-18, have already been discussed (Carter and Murphy, 1999; Robinson et al., 1997; Yang et al., 2001; see Sections II.A.1 and V.B). 2. T1/ST2 Another member of the type I IL-1R family, T1/ST2 (Klemenz et al., 1989; Tominaga, 1989), also shows selective expression on T cell subsets. In contrast to the IL-18R, T1/ST2 is preferentially expressed on murine Th2 cells (LoÈhning et al., 1998, 1999; Xu et al., 1998a). T1/ST2 is also expressed on mast cells, and the promoter used in mast cells contains two putative GATA binding sites pointing to a possible coordinate regulation of Th2 development and T1/ST2 expression by GATA-3 (GaÈchter et al., 1996; Moritz et al., 1998). Treatment of BALB/c mice with depleting antibodies against T1/ST2 results in resistance to L. major infection (Xu et al., 1998a). Type 2 cytokine-(co)expressing CD4‡ T

cytokine memory of t helper lymphocytes

153

cells from granulomas induced by Schistosoma mansoni preferentially upregulate T1/ST2, and enhanced T1/ST2 expression appears to colocalize with effector activity in the Th2-dominated immune response (LoÈhning et al., 1999). T1/ST2-de®cient mice show defects in primary granuloma formation and type 2 cytokine secretion in response to S. mansoni (Townsend et al., 2000). However, the Th2-biased immune reaction after infection with Nippostrongylus brasiliensis and also IL-4-driven Th2 differentiation in vitro are not impaired in the absence of T1/ST2 (Hoshino et al., 1999; Senn et al., 2000; Townsend et al., 2000). T1/ST2 does not bind to IL-1a, IL-1b, or IL-1 receptor antagonist and, to date, no ligand for T1/ST2 has been identi®ed unambigously. However, in the cytoplasmic domain of T1/ST2 the amino acid positions relevant for signaling in the IL-1R are conserved, suggesting that T1/ST2 signaling may be relevant too (Kumar et al., 1995; Mitcham et al., 1996; Reikerstorfer et al., 1995; RoÈssler et al., 1995). Blocking the interaction of T1/ST2 with its putative ligand by injection of soluble T1/ST2 attenuates eosinophilic lung in®ltration and pulmonary IL-4 and IL-5 secretion in a Th2-dependent model of allergic airway in¯ammation, suggesting that the ligand is expressed in the lung (Coyle et al., 1999; Lambrecht et al., 2000; LoÈhning et al., 1998). Crosslinking of T1/ST2 by an anti-T1/ST2 antibody stimulates proliferation and IL-4 and IL-5 memory expression by Th2 but not Th1 cells in a dose-dependent manner, even in the absence of TCR engagement (Meisel et al., 2001). Taken together, T1/ST2 not only is preferentially expressed on Th2 cells but also appears to play an important role in Th2 immune reactions. 3. ICOS The CD28-related inducible costimulator ICOS (Hutloff et al., 1999) is preferentially expressed on Th2 clones and Th2 cells after repeated polarizing stimulation (Coyle et al., 2000; McAdam et al., 2000). However, this Th2 association is less clear than that of T1/ST2. Upon primary stimulation, transient ICOS expression is induced on most CD4‡ and CD8‡ T cells (Hutloff et al., 1999; Mages et al., 2000; Yoshinaga et al., 1999) and, even after several rounds of Th cell polarization, the reactivated Th1 cells continue to express ICOS at a level lower than that of Th2 cells (McAdam et al., 2000). Costimulation of primary T cells by CD28 increases TCR-induced ICOS expression (Beier et al., 2000; McAdam et al., 2000). Engagement of ICOS costimulates T cell proliferation independent of CD28. It enhances secretion of various type 1 and type 2 cytokines with the exception of IL-2, which appears to be speci®cally costimulated by CD28 (Hutloff et al., 1999; Swallow et al., 1999; Yoshinaga et al., 1999). The ICOS ligand, termed B7 homolog (B7h) (Mages et al., 2000; Swallow et al., 1999), B7-related protein-1 (B7RP-1) (Yoshinaga et al., 1999), or GL50 (Ling et al., 2000), belongs to the B7 superfamily, several novel members of

154

max loÈhning et al.

which have recently been identi®ed and, in part, have been attributed to specialized functions (for a recent review, see Coyle and Gutierrez-Ramos, 2001). ICOS ligand is expressed on B cells and expression is induced in ®broblasts by treatment with TNF-a (Ling et al., 2000; Swallow et al., 1999; Yoshinaga et al., 1999). Inhibition of the interaction of ICOS with its ligand by soluble ICOS±Ig fusion protein decreases type 1 and type 2 cytokine production in an infection with N. brasiliensis, but antiviral CTL responses are not impaired (Kopf et al., 2000). Further, it reduces IL-5 secretion and eosinophilic in®ltration of the lung in a Th2-dependent model of airway in¯ammation (Coyle et al., 2000). In ICOS-de®cient mice immunized with T-cell-dependent protein antigens, lower production of IL-4 but not IFN-g, defects in the formation of germinal centers, and reduced levels of IgG1, IgG2a, and IgE have been observed (Dong et al., 2001; Tafuri et al., 2001). IL-4-driven Th2 differentiation in vitro is largely unaffected. The defective IgG production could in part be restored by stimulation of the B cells via CD40 (McAdam et al., 2001), which ®ts the previous observation that ICOS costimulation enhances expression of CD40 ligand on CD4‡ T cells (Hutloff et al., 1999). Thus, the interaction of Th cells with B cells via ICOS and CD40 ligand on the T cell side and ICOS ligand and CD40 on the B cell side appears to play an important role not only in T cell activation and cytokine production but also in antibody class switching and B cell differentiation to antibody-secreting plasma cells. In the context of T cell differentiation, distinct programs of gene expression seem to be induced in concert, providing the activated Th cells with the tools to function as coordinators of distinct immune reactions (Fig. 3). Such differentially expressed genes provide interesting targets for therapeutic modulation of distinct types of immune responses, although none of these molecules described so far show a direct linkage with the expression of IL-4 or IFN-g, i.e., none are under exactly the same molecular control. It remains to be shown which of these genes are induced by the same inductive signals and memorized in the same way as the effector cytokines IL-4, IL-10, and IFN-g. IX. Cytokine Memory of Memory T Cells A. Distinct Localizations and Cytokine Memories of Central and Effector Memory T Cells Memory T cells provide for rapid and ef®cient effector responses upon encountering their antigen (reviewed in Ahmed and Gray, 1996; Dutton et al., 1998). Recent evidence indicates functional heterogeneity of memory T cells, apart from their cytokine memory, with respect to selective homing, effector function, and activation requirements.

Fig. 3. Features of Th differentiation programs. Like GATA-3, the key transcription factor of the Th2 developmental program, T-bet is a prime candidate for key transcription factor of Th1 polarization. The predominance of GATA-3 or T-bet in an individual cell determines its differentiation to Th2 or Th1 gene expression. Not every Th1 or Th2 cell expresses all genes of the respective Th developmental program at all times or in all restimulations. PSGL-1, P-selectin glycoprotein ligand-1.

156

max loÈhning et al.

The memory for cytokine expression may serve as a general indicator for progression of memory cell differentiation toward a dedicated effector cell memory. In polyclonally stimulated T cell populations, those Th cells expressing IL-2 early, later will be the ones that express IFN-g and, even later, also IL-10 (Assenmacher et al., 1998a). Likewise, in mice transgenic for a ¯uorescent reporter gene under control of the IL-2 promoter, those Th cells expressing the IL-2 reporter gene in primary stimulation show enhanced effector cytokine secretion upon secondary challenges (Saparov et al., 1999). Thus, IL-2 production in the primary response may indicate that the T cell has exceeded an activation threshold predisposing it for later acquisition of effector functions. The duration of TCR stimulation may be one factor in¯uencing the developmental fate of a T cell (reviewed in Lanzavecchia and Sallusto, 2000a). Short engagement of TCR in the absence of polarizing cytokines generates a Th cell population with limited capacity to produce effector cytokines when compared with Th1 and Th2 cells. Nevertheless, these nonpolarized cells show a lower activation threshold for induction of proliferation than naive cells. Upon adoptive transfer, they migrate preferentially to the lymph nodes and less ef®ciently to the spleen as compared with Th1 and Th2 cells (lezzi et al., 2001). This preferential migration to lymph nodes is probably due to the maintenance of considerable expression of the lymph node homing receptor CD62L (L-selectin) (Gallatin et al., 1983) and the chemokine receptor CCR7, both of which are downregulated on polarized Th cells. By the analysis of CCR7-de®cient mice, the critical involvement of CCR7 in the migration of T cells and DCs to lymph nodes has been demonstrated (FoÈrster et al., 1999). Expression of CCR7 on preactivated CD45RA CD4‡ and CD8‡ T cells from human blood has been used to discriminate between central and effector memory populations (Sallusto et al., 1999b). The central memory cells expressing CCR7 are characterized by high expression of CD62L and production of IL-2. Central memory cells have no memory for effector cytokines, like naive cells. Unlike naive CD45RA‡ T cells, central memory cells proliferate in response to weak TCR stimulation. Effector memory T cells express lower levels of CCR7 and CD62L and, in the case of CD8‡ cells, frequently contain perforin. TCR stimulation leads to rapid production of effector cytokines but only little IL-2 secretion (Sallusto et al., 1999b). Thus, effector memory cells are the memory cells with a memory for effector cytokines. Similar observations regarding the memory for effector cytokines and IL-2, and the sensitivity toward antigen, have been made for murine memory Th cells subdivided according to high and low expression of CD62L (Ahmadzadeh et al., 2001). Another study in the mouse describes preferential expression of CCR7 on in vitro-differentiated Th1 cells but not Th2 cells resulting in different localization patterns in the spleen and lymph nodes upon adoptive transfer and immunization with the T-cell-speci®c antigen. Notably, retroviral expression of CCR7 in

cytokine memory of t helper lymphocytes

157

Th2 cells changes their splenic localization pattern to a Th1-like distribution. The ability of the Th2 cells to provide help to B cells for antibody production in vivo is impaired (Randolph et al., 1999). Obviously, the differential migration behavior of Th cell subsets is critical for their selective support of distinct immune responses. Preferential localization of central and effector memory CD4‡ and CD8‡ T cells has recently been demonstrated in mice. Central memory T cells did home preferentially to secondary lymphoid organs. Effector memory T cells did reside in nonlymphoid tissues such as lung or liver (Masopust et al., 2001; Reinhardt et al., 2001). In secondary lymphoid organs, central memory T cells may support proliferative expansion of the memory pool by secretion of IL-2. In nonlymphoid tissues, effector memory T cells may control immune reactions against local antigen, by secretion of effector cytokines and cytolytic effector molecules, in the case of CD8‡ effector memory T cells. B. Cytokine Signals in the Persistence of Memory T Cells and of Cytokine Memory Cytokines play a critical role in the control of T cell homeostasis (reviewed in Van Parijs and Abbas, 1998). Activation of a CD4 T cell by engagement of TCR and costimulatory receptor induces expression of IL-2 and of the high-af®nity IL-2R complex. Signaling from the IL-2R induces the T cell to enter the cell cycle. At this stage it is highly susceptible to activation-induced cell death (AICD) triggered by continued stimulation with IL-2, which now leads to enhanced expression of Fas ligand and reduced expression of the Fas signaling inhibitor FLIP (Refaeli et al., 1998). IL-15 may be able to block this IL-2induced AICD (Marks-Konczalik et al., 2000). The balance between the opposing effects of IL-15 and IL-2 may control homeostasis of CD8‡ memory T cells in vivo (Ku et al., 2000; Waldmann et al., 2001). A requirement for a continuous supply of IL-12 has been reported in the maintenance of protective Th1 responses to infections with the intracellular parasites L. major and Toxoplasma gondii. Transient administration of recombinant IL-12 to IL-12-de®cient mice during infection with these parasites at ®rst results in the generation of a bene®cial IFN-g-dominated cytokine response. However, after cessation of IL-12 treatment the mice revert to a susceptible phenotype associated with the loss of the initial Th1 immunity (Park et al., 2000; Yap et al., 2000). In line with this, a prolonged supply of IL-12 by vaccination with a leishmanial protein together with IL-12 DNA but not IL-12 protein leads to long-term protective IFN-g responses in BALB/c mice (Gurunathan et al., 1998; Stobie et al., 2000). This evidence has been suggested to argue against the existence of a stable long-lasting cytokine memory for IFN-g, although it is not clear whether in these models a Th1 effector memory has been established a priori. In a Th1dominated murine model of chronic psoriasis, injection of neutralizing antiIL-12 antibodies ameliorates the disease even in late and chronic stages. After

158

max loÈhning et al.

discontinuation of the anti-IL-12 treatment, in¯amed skin lesions return, most likely due to the persistence of IFN-g effector memory cells (Hong et al., 2001). Apart from these anecdotal observations, little is known about the contribution of effector memory cells with distinct cytokine memories to protective and pathological immune reactions in vivo. Acknowledgments We thank Mario Assenmacher, Hyun-Dong Chang, Osman SoÈzeri, Torsten Stamm, Lars Tykocinski, Alexander Scheffold, and Alf Hamann for helpful discussions. Work cited from this laboratory was supported by the Deutsche Forschungsgemeinschaft through Sonderforschungsbereich 421. Max LoÈhning was supported by a fellowship from the Studienstiftung des deutschen Volkes. This chapter is based on work published until April 2001.

References Abbas, A. K., Murphy, K. M., and Sher, A. (1996). Functional diversity of helper T lymphocytes. Nature 383, 787±793. Agarwal, S., Avni, O., and Rao, A. (2000). Cell-type-restricted binding of the transcription factor NFAT to a distal IL-4 enhancer in vivo. Immunity 12, 643±652. Agarwal, S., and Rao, A. (1998). Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 9, 765±775. Ahmadzadeh, M., Hussain, S. F., and Farber, D. L. (2001). Heterogeneity of the memory CD4 T cell response: Persisting effectors and resting memory T cells. J. Immunol. 166, 926±935. Ahmed, R., and Gray, D. (1996). Immunological memory and protective immunity: Understanding their relation. Science 272, 54±60. Ahn, H. J., Maruo, S., Tomura, M., Mu, J., Hamaoka, T., Nakanishi, K., Clark, S., Kurimoto, M., Okamura, H., and Fujiwara, H. (1997). A mechanism underlying synergy between IL-12 and IFN-g-inducing factor in enhanced production of IFN-g. J. Immunol. 159, 2125±2131. Akira, S., and Kishimoto, T. (1997). NF-IL6 and NF-kB in cytokine gene regulation. Adv. Immunol. 65, 1±46. Allavena, P., Bianchi, G., Zhou, D., van Damme, J., Jilek, P., Sozzani, S., and Mantovani, A. (1994). Induction of natural killer cell migration by monocyte chemotactic protein-1, -2 and -3. Eur. J. Immunol. 24, 3233±3236. Andrew, D. P., Ruf®ng, N., Kim, C. H., Miao, W., Heath, H., Li, Y., Murphy, K., Campbell, J. J., Butcher, E. C., and Wu, L. (2001). C-C chemokine receptor 4 expression de®nes a major subset of circulating nonintestinal memory Tcells of both Th1 and Th2 potential. J. Immunol. 166, 103±111. Aronica, M. A., Mora, A. L., Mitchell, D. B., Finn, P. W., Johnson, J. E., Sheller, J. R., and Boothby, M. R. (1999). Preferential role for NF-kB/Rel signaling in the type 1 but not type 2 T celldependent immune response in vivo. J. Immunol. 163, 5116±5124. Assenmacher, M., LoÈhning, M., Scheffold, A., Manz, R. A., Schmitz, J., and Radbruch, A. (1998a). Sequential production of IL-2, IFN-g and IL-10 by individual staphylococcal enterotoxin Bactivated T helper lymphocytes. Eur. J. Immunol. 28, 1534±1543. Assenmacher, M., LoÈhning, M., Scheffold, A., Richter, A., Miltenyi, S., Schmitz, J., and Radbruch, A. (1998b). Commitment of individual Th1-like lymphocytes to expression of IFN-g versus IL-4 and IL-10: Selective induction of IL-10 by sequential stimulation of naive Th cells with IL-12 and IL-4. J. Immunol. 161, 2825±2832. Assenmacher, M., Scheffold, A., Schmitz, J., Segura Checa, J. A., Miltenyi, S., and Radbruch, A. (1996). Speci®c expression of surface interferon-g on interferon-g producing T cells from mouse and man. Eur. J. Immunol. 26, 263±267.

cytokine memory of t helper lymphocytes

159

Assenmacher, M., Schmitz, J., and Radbruch, A. (1994). Flow cytometric determination of cytokines in activated murine T helper lymphocytes: Expression of interleukin-10 in interferon-g and in interleukin-4-expressing cells. Eur. J. Immunol. 24, 1097±1101. Austrup, F., Vestweber, D., Borges, E., LoÈhning, M., BraÈuer, R., Herz, U., Renz, H., Hallmann, R., Scheffold, A., Radbruch, A., and Hamann, A. (1997). P- and E-selectin mediate recruitment of T-helper-1 but not T-helper-2 cells into in¯amed tissues. Nature 385, 81±83. Avni, O., and Rao, A. (2000). T cell differentiation: A mechanistic view. Curr. Opin. Immunol. 12, 654±659. Avots, A., Buttmann, M., Chuvpilo, S., Escher, C., Smola, U., Bannister, A. J., Rapp, U. R., Kouzarides, T., and Ser¯ing, E. (1999). CBP/p300 integrates Raf/Rac-signaling pathways in the transcriptional induction of NF-ATc during T cell activation. Immunity 10, 515±524. Bacon, C. M., McVicar, D. W., Ortaldo, J. R., Rees, R. C., O'Shea, J. J., and Johnston, J. A. (1995a). Interleukin 12 (IL-12) induces tyrosine phosphorylation of JAK2 and TYK2: Differential use of Janus family tyrosine kinases by IL-2 and IL-12. J. Exp. Med. 181, 399±404. Bacon, C. M., Petricoin, E. F., 3rd, Ortaldo, J. R., Rees, R. C., Larner, A. C., Johnston, J. A., and O'Shea, J. J. (1995b). Interleukin 12 induces tyrosine phosphorylation and activation of STAT4 in human lymphocytes. Proc. Natl. Acad. Sci. USA 92, 7307±7311. Bagley, J., Sawada, T., Wu, Y., and lacomini, J. (2000). A critical role for interleukin 4 in activating alloreactive CD4 T cells. Nat. Immunol. 1, 257±261. Banchereau, J., and Steinman, R. M. (1998). Dendritic cells and the control of immunity. Nature 392, 245±252. Beier, K. C., Hutloff, A., Dittrich, A. M., Heuck, C., Rauch, A., Buchner, K., Ludewig, B., Ochs, H. D., Mages, H. W., and Kroczek, R. A. (2000). Induction, binding speci®city and function of human ICOS. Eur. J. Immunol. 30, 3707±3717. Ben-Sasson, S. Z., Gerstel, R., Hu-Li, J., and Paul, W. E. (2001). Cell division is not a ``clock'' measuring acquisition of competence to produce IFN-g or IL-4. J. Immunol. 166, 112±120. Betz, M., and Fox, B. S. (1991). Prostaglandin E2 inhibits production of Th1 lymphokines but not of Th2 lymphokines. J. Immunol. 146, 108±113. Bird, A. P., and Wolffe, A. P. (1999). Methylation-induced repressionÐBelts, braces, and chromatin. Cell 99, 451±454. Bird, J. J., Brown, D. R., Mullen, A. C., Moskowitz, N. H., Mahowald, M. A., Sider, J. R., Gajewski, T. F., Wang, C. R., and Reiner, S. L. (1998). Helper T cell differentiation is controlled by the cell cycle. Immunity 9, 229±237. Bix, M., and Locksley, R. M. (1998). Independent and epigenetic regulation of the interleukin-4 alleles in CD4‡ T cells. Science 281, 1352±1354. Bleul, C. C., Wu, L., Hoxie, J. A., Springer, T. A., and Mackay, C. R. (1997). The HIV coreceptors CXCR4 and CCR5 are differentially expressed and regulated on human T lymphocytes. Proc. Natl. Acad. Sci. USA 94, 1925±1930. Blobel, G. A., Nakajima, T., Eckner, R., Montminy, M., and Orkin, S. H. (1998). CREB-binding protein cooperates with transcription factor GATA-1 and is required for erythroid differentiation. Proc. Natl. Acad. Sci. USA 95, 2061±2066. Bonecchi, R., Bianchi, G., Bordignon, P. P., D'Ambrosio, D., Lang, R., Borsatti, A., Sozzani, S., Allavena, P., Gray, P. A., Mantovani, A., and Sinigaglia, F. (1998). Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s. J. Exp. Med. 187, 129±134. Borges, E., Tietz, W., Steegmaier, M., Moll, T., Hallmann, R., Hamann, A., and Vestweber, D. (1997). P-selectin glycoprotein ligand-1 (PSGL-1) on T helper 1 but not on T helper 2 cells binds to P-selectin and supports migration into in¯amed skin. J. Exp. Med. 185, 573±578. Boyes, J., By®eld, P., Nakatani, Y., and Ogryzko, V. (1998). Regulation of activity of the transcription factor GATA-1 by acetylation. Nature 396, 594±598.

160

max loÈhning et al.

Boyes, J., and Felsenfeld, G. (1996). Tissue-speci®c factors additively increase the probability of the all-or-none formation of a hypersensitive site. EMBO J. 15, 2496±2507. Bradley, L. M., Dalton, D. K., and Croft, M. (1996). A direct role for IFN-g in regulation of Th1 cell development. J. Immunol. 157, 1350±1358. Bright, J. J., and Sriram, S. (1998). TGF-b inhibits IL-12-induced activation of Jak-STAT pathway in T lymphocytes. J. Immunol. 161, 1772±1777. Brinkmann, V., Geiger, T., Alkan, S., and Heusser, C. H. (1993). Interferon a increases the frequency of interferon g-producing human CD4‡ T cells. J. Exp. Med. 178, 1655±1663. Brosterhus, H., Brings, S., Leyendeckers, H., Manz, R. A., Miltenyi, S., Radbruch, A., Assenmacher, M., and Schmitz, J. (1999). Enrichment and detection of live antigenspeci®c CD4(‡) and CD8(‡) T cells based on cytokine secretion. Eur. J. Immunol. 29, 4053± 4059. Brown, D. R., Green, J. M., Moskowitz, N. H., Davis, M., Thompson, C. B., and Reiner, S. L. (1996). Limited role of CD28-mediated signals in T helper subset differentiation. J. Exp. Med. 184, 803±810. Brown, D. R., Moskowitz, N. H., Killeen, N., and Reiner, S. L. (1997a). A role for CD4 in peripheral T cell differentiation. J. Exp. Med. 186, 101±107. Brown, K. E., Baxter, J., Graf, D., Merkenschlager, M., and Fisher, A. G. (1999). Dynamic repositioning of genes in the nucleus of lymphocytes preparing for cell division. Mol. Cell. 3, 207±217. Brown, K. E., Guest, S. S., Smale, S. T., Hahm, K., Merkenschlager, M., and Fisher, A. G. (1997b). Association of transcriptionally silent genes with lkaros complexes at centromeric heterochromatin. Cell 91, 845±854. Bucy, R. P., Panoskaltsis-Mortari, A., Huang, G. Q., Li, J., Karr, L., Ross, M., Russell, J. H., Murphy, K. M., and Weaver, C. T. (1994). Heterogeneity of single cell cytokine gene expression in clonal T cell populations. J. Exp. Med. 180, 1251±1262. Buelens, C., Verhasselt, V., De Groote, D., Thielemans, K., Goldman, M., and Willems, F. (1997). Human dendritic cell responses to lipopolysaccharide and CD40 ligation are differentially regulated by interleukin-10. Eur. J. Immunol. 27, 1848±1852. Buelens, C., Willems, F., Delvaux, A., Pierard, G., Delville, J. P., Velu, T., and Goldman, M. (1995). Interleukin-10 differentially regulates B7-1 (CD80) and B7-2 (CD86) expression on human peripheral blood dendritic cells. Eur. J. Immunol. 25, 2668±2672. Campbell, J. J., Haraldsen, G., Pan, J., Rottman, J., Qin, S., Ponath, P., Andrew, D. P., Warnke, R., Ruf®ng, N., Kassam, N., Wu, L., and Butcher, E. C. (1999). The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memory T cells. Nature 400, 776±780. Cardell, S., HoÈiden, I., and MoÈller, G. (1993). Manipulation of the superantigen-induced lymphokine response: Selective induction of interleukin-10 or interferon-g synthesis in small resting CD4‡ T cells. Eur. J. Immunol. 23, 523±529. Cardell, S., and Sander, B. (1990). Interleukin 2, 4 and 5 are sequentially produced in mitogenstimulated murine spleen cell cultures. Eur. J. Immunol. 20, 389±395. Carr, M. W., Roth, S. J., Luther, E., Rose, S. S., and Springer, T. A. (1994). Monocyte chemoattractant protein 1 acts as a T-lymphocyte chemoattractant. Proc. Natl. Acad. Sci. USA 91, 3652±3656. Carter, L. L., and Murphy, K. M. (1999). Lineage-speci®c requirement for signal transducer and activator of transcription (Stat)4 in interferon g production from CD4(‡) versus CD8(‡) T cells. J. Exp. Med. 189, 1355±1360. Cella, M., Facchetti, F., Lanzavecchia, A., and Colonna, M. (2000). Plasmacytoid dendritic cells activated by in¯uenza virus and CD40L drive a potent TH1 polarization. Nat. Immunol. 1, 305±310.

cytokine memory of t helper lymphocytes

161

Cella, M., Jarrossay, D., Facchetti, F., Alebardi, O., Nakajima, H., Lanzavecchia, A., and Colonna, M. (1999a). Plasmacytoid monocytes migrate to in¯amed lymph nodes and produce large amounts of type I interferon. Nat. Med. 5, 919±923. Cella, M., Salio, M., Sakakibara, Y., Langen, H., Julkunen, I., and Lanzavecchia, A. (1999b). Maturation, activation, and protection of dendritic cells induced by double-stranded RNA. J. Exp. Med. 189, 821±829. Chang, L., and Karin, M. (2001). Mammalian MAP kinase signalling cascades. Nature 410, 37±40. Chen, C. H., Zhang, D. H., LaPorte, J. M., and Ray, A. (2000a). Cyclic AMP activates p38 mitogenactivated protein kinase in Th2 cells: Phosphorylation of GATA-3 and stimulation of Th2 cytokine gene expression. J. Immunol. 165, 5597±5605. Chen, Q., Ghilardi, N., Wang, H., Baker, T., Xie, M. H., Gurney, A., Grewal, I. S., and de Sauvage, F. J. (2000b). Development of Th1-type immune responses requires the type I cytokine receptor TCCR. Nature 407, 916±920. Chen, Y., Kuchroo, V. K., Inobe, J. -L., Ha¯er, D. A., and Weiner, H. L. (1994). Regulatory T cell clones induced by oral tolerance: Suppression of autoimmune encephalomyelitis. Science 265, 1237±1240. Chensue, S. W., Lukacs, N. W., Yang, T. Y., Shang, X., Frait, K. A., Kunkel, S. L., Kung, T., Wiekowski, M. T., Hedrick, J. A., Cook, D. N., Zingoni, A., Narula, S. K., Zlotnik, A., Barrat, F. J., O'Garra, A., Napolitano, M., and Lira, S. A. (2001). Aberrant in vivo T helper type 2 cell response and impaired eosinophil recruitment in CC chemokine receptor 8 knockout mice. J. Exp. Med. 193, 573±584. Chess, A. (1998). Expansion of the allelic exclusion principle? Science 279, 2067±2068. Chiodetti, L., Barber, D. L., and Schwartz, R. H. (2000). Biallelic expression of the IL-2 locus under optimal stimulation conditions. Eur. J. Immunol. 30, 2157±2163. Cho, S. S., Bacon, C. M., Sudarshan, C., Rees, R. C., Finbloom, D., Pine, R., and O'Shea, J. J. (1996). Activation of STAT4 by IL-12 and IFN-a: Evidence for the involvement of ligand-induced tyrosine and serine phosphorylation. J. Immunol. 157, 4781±4789. Chow, C. W., Dong, C., Flavell, R. A., and Davis, R. J. (2000). c-Jun NH(2)-terminal kinase inhibits targeting of the protein phosphatase calcineurin to NFATc1. Mol. Cell. Biol. 20, 5227±5234. Chuvpilo, S., Zimmer, M., Kerstan, A., Glockner, J., Avots, A., Escher, C., Fischer, C., Inashkina, I., Jankevics, E., Berberich-Siebelt, F., Schmitt, E., and Ser¯ing, E. (1999). Alternative polyadenylation events contribute to the induction of NF-ATc in effector T cells. Immunity 10, 261±269. Chvatchko, Y., Hoogewerf, A. J., Meyer, A., Alouani, S., Juillard, P., Buser, R., Conquet, F., Proudfoot, A. E., Wells, T. N., and Power, C. A. (2000). A key role for CC chemokine receptor 4 in lipopolysaccharide-induced endotoxic shock. J. Exp. Med. 191, 1755±1764. Coffman, R. L., and Reiner, S. L. (1999). Instruction, selection, or tampering with the odds? Science 284, 1283±1285. Constant, S. L., and Bottomly, K. (1997). Induction of Th1 and Th2 CD4‡ T cell responses: The alternative approaches. Annu. Rev. Immunol. 15, 297±322. Corinti, S., Albanesi, C., la Sala, A., Pastore, S., and Girolomoni, G. (2001). Regulatory activity of autocrine IL-10 on dendritic cell functions. J. Immunol. 166, 4312±4318. Cosmi, L., Annunziato, F., Galli, M. I. G., Maggi, R. M. E., Nagata, K., and Romagnani, S. (2000). CRTH2 is the most reliable marker for the detection of circulating human type 2 Th and type 2 T cytotoxic cells in health and disease. Eur. J. Immunol. 30, 2972±2979. Coyle, A. J., and Gutierrez-Ramos, J. C. (2001). The expanding B7 superfamily: Increasing complexity in costimulatory signals regulating T cell function. Nat. Immunol. 2, 203±209. Coyle, A. J., Lehar, S., Lloyd, C., Tian, J., Delaney, T., Manning, S., Nguyen, T., Burwell, T., Schneider, H., Gonzalo, J. A., Gosselin, M., Owen, L. R., Rudd, C. E., and GutierrezRamos, J. C. (2000). The CD28-related molecule ICOS is required for effective T cell-dependent immune responses. Immunity 13, 95±105.

162

max loÈhning et al.

Coyle, A. J., Lloyd, C., Tian, J., Nguyen, T., Erikkson, C., Wang, L., Ottoson, P., Persson, P., Delaney, T., Lehar, S., Lin, S., Poisson, L., Meisel, C., Kamradt, T., Bjerke, T., Levinson, D., and GutierrezRamos, J. C. (1999). Crucial role of the interleukin 1 receptor family member T1/ST2 in T helper cell type 2-mediated lung mucosal immune responses. J. Exp. Med. 190, 895±902. Cron, R. Q., Bort, S. J., Wang, Y., Brunvand, M. W., and Lewis, D. B. (1999). T cell priming enhances IL-4 gene expression by increasing nuclear factor of activated T cells. J. Immunol. 162, 860±870. Curiel, R. E., Lahesmaa, R., Subleski, J., Cippitelli, M., Kirken, R. A., Young, H. A., and Ghosh, P. (1997). Identi®cation of a Stat-6-responsive element in the promoter of the human interleukin-4 gene. Eur. J. Immunol. 27, 1982±1987. D'Ambrosio, D., Iellem, A., Bonecchi, R., Mazzeo, D., Sozzani, S., Mantovani, A., and Sinigaglia, F. (1998). Selective up-regulation of chemokine receptors CCR4 and CCR8 upon activation of polarized human type 2 Th cells. J. Immunol. 161, 5111±5115. d'Ostiani, C. F., Del Sero, G., Bacci, A., Montagnoli, C., Spreca, A., Mencacci, A., RicciardiCastagnoli, P., and Romani, L. (2000). Dendritic cells discriminate between yeasts and hyphae of the fungus Candida albicans: Implications for initiation of T helper cell immunity in vitro and in vivo. J. Exp. Med. 191, 1661±1674. Darnell, J. E., Jr. (1997). STATs and gene regulation. Science 277, 1630±1635. Das, J., Chen, C. H., Yang, L., Cohn, L., Ray, P., and Ray, A. (2001). A critical role for NF-kB in GATA3 expression and TH2 differentiation in allergic airway in¯ammation. Nat. Immunol. 2, 45±50. Davis, L. S., Schulze-Koops, H., and Lipsky, P. E. (1999). Human CD4‡ T cell differentiation and effector function: Implications for autoimmunity. Immunol. Res. 19, 25±34. De Smedt, T., Van Mechelen, M., De Becker, G., Urbain, J., Leo, O., and Moser, M. (1997). Effect of interleukin-10 on dendritic cell maturation and function. Eur. J. Immunol. 27, 1229±1235. Dent, A. L., Hu-Li, J., Paul, W. E., and Staudt, L. M. (1998). T helper type 2 in¯ammatory disease in the absence of interleukin 4 and transcription factor STAT6. Proc. Natl. Acad. Sci. USA 95, 13823±13828. Dent, A. L., Shaffer, A. L., Yu, X., Allman, D., and Staudt, L. M. (1997). Control of in¯ammation, cytokine expression, and germinal center formation by BCL-6. Science 276, 589±592. Dhodapkar, M. V., Steinman, R. M., Krasovsky, J., Munz, C., and Bhardwaj, N. (2001). Antigenspeci®c inhibition of effector T cell function in humans after injection of immature dendritic cells. J. Exp. Med. 193, 233±238. Dhordain, P., Lin, R. J., Quief, S., Lantoine, D., Kerckaert, J. P., Evans, R. M., and Albagli, O. (1998). The LAZ3(BCL-6) oncoprotein recruits a SMRT/mSIN3A/histone deacetylase containing complex to mediate transcriptional repression. Nucleic Acids Res. 26, 4645±4651. Dickensheets, H. L., Venkataraman, C., Schindler, U., and Donnelly, R. P. (1999). Interferons inhibit activation of STAT6 by interleukin 4 in human monocytes by inducing SOCS-1 gene expression. Proc. Natl. Acad. Sci. USA 96, 10800±10805. Diehl, S., Anguita, J., Hoffmeyer, A., Zapton, T., Ihle, J. N., Fikrig, E., and Rincon, M. (2000). Inhibition of Th1 differentiation by IL-6 is mediated by SOCS1. Immunity 13, 805±815. Dong, C., and Flavell, R. A. (2000). Cell fate decision: T-helper 1 and 2 subsets in immune responses. Arthritis Res. 2, 179±188. Dong, C., Juedes, A. E., Temann, U. A., Shresta, S., Allison, J. P., Ruddle, N. H., and Flavell, R. A. (2001). ICOS co-stimulatory receptor is essential for T-cell activation and function. Nature 409, 97±101. Dong, C., Yang, D. D., Tournier, C., Whitmarsh, A. J., Xu, J., Davis, R. J., and Flavell, R. A. (2000). JNK is required for effector T-cell function but not for T-cell activation. Nature 405, 91±94. Dong, C., Yang, D. D., Wysk, M., Whitmarsh, A. J., Davis, R. J., and Flavell, R. A. (1998). Defective T cell differentiation in the absence of Jnk1. Science 282, 2092±2095.

cytokine memory of t helper lymphocytes

163

Dutton, R. W., Bradley, L. M., and Swain, S. L. (1998). T cell memory. Annu. Rev. Immunol. 16, 201±223. Endo, T. A., Masuhara, M., Yokouchi, M., Suzuki, R., Sakamoto, H., Mitsui, K., Matsumoto, A., Tanimura, S., Ohtsubo, M., Misawa, H., Miyazaki, T., Leonor, N., Taniguchi, T., Fujita, T., Kanakura, Y., Komiya, S., and Yoshimura, A. (1997). A new protein containing an SH2 domain that inhibits JAK kinases. Nature 387, 921±924. Farrar, J. D., Ouyang, W., LoÈhning, M., Assenmacher, M., Radbruch, A., Kanagawa, O., and Murphy, K. M. (2001). An instructive component in T helper cell type 2 (Th2) development mediated by GATA-3. J. Exp. Med. 193, 643±650. Farrar, J. D., Smith, J. D., Murphy, T. L., Leung, S., Stark, G. R., and Murphy, K. M. (2000a). Selective loss of type I interferon-induced STAT4 activation caused by a minisatellite insertion in mouse Stat2. Nat. Immunol. 1, 65±69. Farrar, J. D., Smith, J. D., Murphy, T. L., and Murphy, K. M. (2000b). Recruitment of Stat4 to the human interferon-a=b receptor requires activated Stat2. J. Biol. Chem. 275, 2693±2697. Ferber, I. A., Lee, H. J., Zonin, F., Heath, V., Mui, A., Arai, N., and O'Garra, A. (1999). GATA-3 signi®cantly downregulates IFN-g production from developing Th1 cells in addition to inducing IL-4 and IL-5 levels. Clin. Immunol. 91, 134±144. Feske, S., Giltnane, J., Dolmetsch, R., Staudt, L. M., and Rao, A. (2001). Gene regulation mediated by calcium signals in T lymphocytes. Nat. Immunol. 2, 316±324. Finkelman, F. D., Morris, S. C., Orekhova, T., Mori, M., Donaldson, D., Reiner, S. L., Reilly, N. L., Schopf, L., and Urban, J. F., Jr. (2000). Stat6 regulation of in vivo IL-4 responses. J. Immunol. 164, 2303±2310. Firestein, G. S., Roeder, W. D., Laxer, J. A., Townsend, K. S., Weaver, C. T., Hom, J. T., Linton, J., Torbett, B. E., and Glasebrook, A. L. (1989). A new murine CD4‡ T cell subset with an unrestricted cytokine pro®le. J. Immunol. 143, 518±525. Fitzpatrick, D. R., Shirley, K. M., and Kelso, A. (1999). Cutting edge: Stable epigenetic inheritance of regional IFN-g promoter demethylation in CD44highCD8‡ T lymphocytes. J. Immunol. 162, 5053±5057. Fitzpatrick, D. R., Shirley, K. M., McDonald, L. E., Bielefeldt-Ohmann, H., Kay, G. F., and Kelso, A. (1998). Distinct methylation of the interferon g (IFN-g) and interleukin 3 (IL-3) genes in newly activated primary CD8‡ T lymphocytes: Regional IFN-g promoter demethylation and mRNA expression are heritable in CD44(high)CD8‡ T cells. J. Exp. Med. 188, 103±117. FoÈrster, R., Schubel, A., Breitfeld, D., Kremmer, E., Renner-Muller, I., Wolf, E., and Lipp, M. (1999). CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 99, 23±33. Fowell, D. J., Magram, J., Turck, C. W., Killeen, N., and Locksley, R. M. (1997). Impaired Th2 subset development in the absence of CD4. Immunity 6, 559±569. Fowell, D. J., Shinkai, K., Liao, X. C., Beebe, A. M., Coffman, R. L., Littman, D. R., and Locksley, R. M. (1999). Impaired NFATc translocation and failure of Th2 development in ltk-de®cient CD4‡ T cells. Immunity 11, 399±409. Fuentes, M. E., Durham, S. K., Swerdel, M. R., Lewin, A. C., Barton, D. S., Megill, J. R., Bravo, R., and Lira, S. A. (1995). Controlled recruitment of monocytes and macrophages to speci®c organs through transgenic expression of monocyte chemoattractant protein-1. J. Immunol. 155, 5769±5776. GaÈchter, T., Werenskiold, A. K., and Klemenz, R. (1996). Transcription of the interleukin-1 receptorrelated T1 gene is initiated at different promoters in mast cells and ®broblasts. J. Biol. Chem. 271, 124±129. Galbiati, F., Rogge, L., Guery, J. C., Smiroldo, S., and Adorini, L. (1998). Regulation of the IL-12 receptor b2 subunit by soluble antigen and IL-12 in vivo. Eur. J. Immunol. 28, 209±220.

164

max loÈhning et al.

Gallatin, W. M., Weissman, I. L., and Butcher, E. C. (1983). A cell-surface molecule involved in organ-speci®c homing of lymphocytes. Nature 304, 30±34. Garcia-Rodriguez, C., and Rao, A. (1998). Nuclear factor of activated T cell (NFAT)-dependent transactivation regulated by the coactivators p300/CREB-binding protein (CBP). J. Exp. Med. 187, 2031±2036. George, K. M., Leonard, M. W., Roth, M. E., Lieuw, K. H., Kioussis, D., Grosveld, F., and Engel, J. D. (1994). Embryonic expression and cloning of the murine GATA-3 gene. Devlopment 120, 2673±2686 Gett, A. V., and Hodgkin, P. D. (1998). Cell division regulates the T cell cytokine repertoire, revealing a mechanism underlying immune class regulation. Proc. Natl. Acad. Sci. USA 95, 9488±9493. Gingras, S., Simard, J., Groner, B., and P®tzner, E. (1999). p300/CBP is required for transcriptional induction by interleukin-4 and interacts with Stat6. Nucleic Acids Res. 27, 2722±2729. Glimcher, L. H., and Murphy, K. M. (2000). Lineage commitment in the immune system: The T helper lymphocyte grows up. Genes Dev. 14, 1693±1711. Gollob, J. A., Kawasaki, H., and Ritz, J. (1997). Interferon-g and interleukin-4 regulate T cell interleukin-12 responsiveness through the differential modulation of high-af®nity interleukin12 receptor expression. Eur. J. Immunol. 27, 647±652. Gorelik, L., Fields, P. E., and Flavell, R. A. (2000). Cutting edge: TGF-b inhibits Th type 2 development through inhibition of GATA-3 expression. J. Immunol. 165, 4773±4777. Gorham, J. D., Guler, M. L., Fenoglio, D., Gubler, U., and Murphy, K. M. (1998). Low dose TGF-b attenuates IL-12 responsiveness in murine Th cells. J. Immunol. 161, 1664±1670. Gourley, T. S., and Chang, C. H. (2001). Cutting edge: The class II transactivator prevents activation-induced cell death by inhibiting Fas ligand gene expression. J. Immunol. 166, 2917±2921. Gourley, T., Roys, S., Lukacs, N. W., Kunkel, S. L., Flavell, R. A., and Chang, C. H. (1999). A novel role for the major histocompatibility complex class II transactivator CIITA in the repression of IL-4 production. Immunity 10, 377±386. Gregoire, J. M., and Romeo, P. H. (1999). T-cell expression of the human GATA-3 gene is regulated by a non-lineage-speci®c silencer. J. Biol. Chem. 274, 6567±6578. Grogan, J. L., Mohrs, M., Harmon, B., Lacy, D. A., Sedat, J. W., and Locksley, R. M. (2001). Early transcription and silencing of cytokine genes underlie polarization of T helper cell subsets. Immunity 14, 205±215. Groux, H., O'Garra, A., Bigler, M., Rouleau, M., Antonenko, S., de Vries, J. E., and Roncarolo, M. G. (1997). A CD4‡ T-cell subset inhibits antigen-speci®c T-cell responses and prevents colitis. Nature 389, 737±742. Gu, L., Tseng, S., Horner, R. M., Tam, C., Loda, M., and Rollins, B. J. (2000). Control of TH2 polarization by the chemokine monocyte chemoattractant protein-1. Nature 404, 407±411. Guler, M. L., Gorham, J. D., Hsieh, C. S., Mackey, A. J., Steen, R. G., Dietrich, W. F., and Murphy, K. M. (1996). Genetic susceptibility to Leishmania: IL-12 responsiveness in TH1 cell development. Science 271, 984±987. Gurunathan, S., Prussin, C., Sacks, D. L., and Seder, R. A. (1998). Vaccine requirements for sustained cellular immunity to an intracellular parasitic infection. Nat. Med. 4, 1409±1415. Harris, D. P., Haynes, L., Sayles, P. C., Duso, D. K., Eaton, S. M., Lepak, N. M., Johnson, L. L., Swain, S. L., and Lund, F. E. (2000). Reciprocal regulation of polarized cytokine production by effector B and T cells. Nat. Immunol. 1, 475±482. Haspel, R. L., Salditt-Georgieff, M., and Darnell, J. E., Jr. (1996). The rapid inactivation of nuclear tyrosine phosphorylated Stat1 depends upon a protein tyrosine phosphatase. EMBO J. 15, 6262±6268.

cytokine memory of t helper lymphocytes

165

Heath, V. L., Murphy, E. E., Crain, C., Tomlinson, M. G., and O'Garra, A. (2000a). TGF-b1 downregulates Th2 development and results in decreased IL-4-induced STAT6 activation and GATA-3 expression. Eur. J. Immunol. 30, 2639±2649. Heath, V. L., Showe, L., Crain, C., Barrat, F. J., Trinchieri, G., and O'Garra, A. (2000b). Cutting edge: Ectopic expression of the IL-12 receptor-b2 in developing and committed Th2 cells does not affect the production of IL-4 or induce the production of IFN-g. J. Immunol. 164, 2861±2865. Hendriks, R. W., Nawijn, M. C., Engel, J. D., van Doorninck, H., Grosveld, F., and Karis, A. (1999). Expression of the transcription factor GATA-3 is required for the development of the earliest T cell progenitors and correlates with stages of cellular proliferation in the thymus. Eur. J. Immunol. 29, 1912±1918. Henkel, G., and Brown, M. A. (1994). PU.1 and GATA: Components of a mast cell-speci®c interleukin 4 intronic enhancer. Proc. Natl. Acad. Sci. USA 91, 7737±7741. Henkel, G., Weiss, D. L., McCoy, R., Deloughery, T., Tara, D., and Brown, M. A. (1992). A DNase I-hypersensitive site in the second intron of the murine IL-4 gene de®nes a mast cell-speci®c enhancer. J. Immunol. 149, 3239±3246. Heyeck, S. D., Wilcox, H. M., Bunnell, S. C., and Berg, L. J. (1997). Lck phosphorylates the activation loop tyrosine of the ltk kinase domain and activates ltk kinase activity. J. Biol. Chem. 272, 25401±25408. Hirai, H., Tanaka, K., Yoshie, O., Ogawa, K., Kenmotsu, K., Takamori, Y., Ichimasa, M., Sugamura, K., Nakamura, M., Takano, S., and Nagata, K. (2001). Prostaglandin D2 selectively induces chemotaxis in T helper type 2 cells, eosinophils, and basophils via seven-transmembrane receptor CRTH2. J. Exp. Med. 193, 255±261. Ho, I. C., Hodge, M. R., Rooney, J. W., and Glimcher, L. H. (1996). The proto-oncogene c-maf is responsible for tissue-speci®c expression of interleukin-4. Cell 85, 973±983. Ho, I. C., Lo, D., and Glimcher, L. H. (1998). c-maf promotes T helper cell type 2 (Th2) and attenuates Th1 differentiation by both interleukin 4-dependent and -independent mechanisms. J. Exp. Med. 188, 1859±1866. Hodge, M. R., Chun, H. J., Rengarajan, J., Alt, A., Lieberson, R., and Glimcher, L. H. (1996a). NFAT-driven interleukin-4 transcription potentiated by NIP45. Science 274, 1903±1905. Hodge, M. R., Ranger, A. M., Charles de la Brousse, F., Hoey, T., Grusby, M. J., and Glimcher, L. H. (1996b). Hyperproliferation and dysregulation of IL-4 expression in NF-ATp-de®cient mice. Immunity 4, 397±405. Hoehn, P., Goedert, S., Germann, T., Koelsch, S., Jin, S., Palm, N., Ruede, E., and Schmitt, E. (1995). Opposing effects of TGF-b2 on the Th1 cell development of naive CD4‡ T cells isolated from different mouse strains. J. Immunol. 155, 3788±3793. Hoey, T., and Grusby, M. J. (1999). STATs as mediators of cytokine-induced responses. Adv. Immunol. 71, 145±162. HollaÈnder, G. A., Zuklys, S., Morel, C., Mizoguchi, E., Mobisson, K., Simpson, S., Terhorst, C., Wishart, W., Golan, D. E., Bhan, A. K., and Burakoff, S. J. (1998). Monoallelic expression of the interleukin-2 locus. Science 279, 2118±2121. Hong, K., Berg, E. L., and Ehrhardt, R. O. (2001). Persistence of pathogenic CD4(‡) Th1-like cells in vivo in the absence of IL-12 but in the presence of autoantigen. J. Immunol. 166, 4765±4772. Hoshino, K., Kashiwamura, S., Kuribayashi, K., Kodama, T., Tsujimura, T., Nakanishi, K., Matsuyama, T., Takeda, K., and Akira, S. (1999). The absence of interleukin 1 receptor-related T1/ST2 does not affect T helper cell type 2 development and its effector function. J. Exp. Med. 190, 1541±1548. Hou, J., Schindler, U., Henzel, W. J., Ho, T. C., Brasseur, M., and McKnight, S. L. (1994). An interleukin-4-induced transcription factor: IL-4 Stat. Science 265, 1701±1706. Hsieh, C. L. (2000). Dynamics of DNA methylation pattern. Curr. Opin. Genet. Dev. 10, 224±228.

166

max loÈhning et al.

Hsieh, C. S., Heimberger, A. B., Gold, J. S., O'Garra, A., and Murphy, K. M. (1992). Differential regulation of T helper phenotype development by interleukins 4 and 10 in an ab T-cell-receptor transgenic system. Proc. Natl. Acad. Sci. USA 89, 6065±6069. Hsieh, C. S., Macatonia, S. E., Tripp, C. S., Wolf, S. F., O'Garra, A., and Murphy, K. M. (1993). Development of Th1 CD4‡ T cells through IL-12 produced by Listeria-induced macrophages. Science 260, 547±549. Hu-Li, J., Huang, H., Ryan, J., and Paul, W. E. (1997). In differentiated CD4‡ T cells, interleukin 4 production is cytokine-autonomous, whereas interferon g production is cytokine-dependent. Proc. Natl. Acad. Sci. USA 94, 3189±3194. Hu-Li, J., Pannetier, C., Guo, L., LoÈhning, M., Gu, H., Watson, C., Assenmacher, M., Radbruch, A., and Paul, W. E. (2001). Regulation of expression of IL-4 alleles: Analysis using a chimeric GFP/ IL-4 gene. Immunity 14, 1±11. Huang, D., Wang, J., Kivisakk, P., Rollins, B. J., and Ransohoff, R. M. (2001). Absence of monocyte chemoattractant protein 1 in mice leads to decreased local macrophage recruitment and antigenspeci®c T helper cell type 1 immune response in experimental autoimmune encephalomyelitis. J. Exp. Med. 193, 713±726. Huang, H., Hu-Li, J., Chen, H., Ben-Sasson, S. Z., and Paul, W. E. (1997). IL-4 and IL-13 production in differentiated T helper type 2 cells is not IL-4 dependent. J. Immunol. 159, 3731±3738. Huang, H., and Paul, W. E. (1998). Impaired interleukin 4 signaling in T helper type 1 cells. J. Exp. Med. 187, 1305±1313. Hunter, C. A., and Reiner, S. L. (2000). Cytokines and T cells in host defense. Curr. Opin. Immunol. 12, 413±418. Hural, J. A., Kwan, M., Henkel, G., Hock, M. B., and Brown, M. A. (2000). An intron transcriptional enhancer element regulates IL-4 gene locus accessibility in mast cells. J. Immunol. 165, 3239±3249. Hutloff, A., Dittrich, A. M., Beier, K. C., Eljaschewitsch, B., Kraft, R., Anagnostopoulos, I., and Kroczek, R. A. (1999). ICOS is an inducible T-cell co-stimulator structurally and functionally related to CD28. Nature 397, 263±266. Iezzi, G., Scheidegger, D., and Lanzavecchia, A. (2001). Migration and function of antigen-primed nonpolarized T lymphocytes in vivo. J. Exp. Med. 193, 987±994. Iezzi, G., Scotet, E., Scheidegger, D., and Lanzavecchia, A. (1999). The interplay between the duration of TCR and cytokine signaling determines T cell polarization. Eur. J. Immunol. 29, 4092±4101. Iwasaki, A., and Kelsall, B. L. (1999). Freshly isolated Peyer's patch, but not spleen, dendritic cells produce interleukin 10 and induce the differentiation of T helper type 2 cells. J. Exp. Med. 190, 229±239. Jacobson, N. G., Szabo, S. J., Weber-Nordt, R. M., Zhong, Z., Schreiber, R. D., Darnell, J. E., Jr., and Murphy, K. M. (1995). Interleukin 12 signaling in T helper type 1 (Th1) cells involves tyrosine phosphorylation of signal transducer and activator of transcription (Stat)3 and Stat4. J. Exp. Med. 181, 1755±1762. Jankovic, D., Kullberg, M. C., Noben-Trauth, N., Caspar, P., Paul, W. E., and Sher, A. (2000). Single cell analysis reveals that IL-4 receptor/Stat6 signaling is not required for the in vivo or in vitro development of CD4‡ lymphocytes with a Th2 cytokine pro®le. J. Immunol. 164, 3047±3055. Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N., Strouboulis, J., and Wolffe, A. P. (1998). Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat. Genet. 19, 187±191. Jonuleit, H., Schmitt, E., Schuler, G., Knop, J., and Enk, A. H. (2000). Induction of interleukin 10producing, nonproliferating CD4(‡) T cells with regulatory properties by repetitive stimulation with allogeneic immature human dendritic cells. J. Exp. Med. 192, 1213±1222.

cytokine memory of t helper lymphocytes

167

Joseph, S. B., Miner, K. T., and Croft, M. (1998). Augmentation of naive, Th1 and Th2 effector CD4 responses by IL-6, IL-1 and TNF. Eur. J. Immunol. 28, 277±289. Jourdan, P., Abbal, C., Noraz, N., Hori, T., Uchiyama, T., Vendrell, J. P., Bousquet, J., Taylor, N., Pene, J., Yssel, H., and Nora, N. (1998). IL-4 induces functional cell-surface expression of CXCR4 on human T cells. J. Immunol. 160, 4153±4157. Kalinski, P., Hilkens, C. M., Snijders, A., Snijdewint, F. G., and Kapsenberg, M. L. (1997). IL-12de®cient dendritic cells, generated in the presence of prostaglandin E2, promote type 2 cytokine production in maturing human naive T helper cells. J. Immunol. 159, 28±35. Kalinski, P., Schuitemaker, J. H., Hilkens, C. M., Wierenga, E. A., and Kapsenberg, M. L. (1999). Final maturation of dendritic cells is associated with impaired responsiveness to IFN-g and to bacterial IL-12 inducers: Decreased ability of mature dendritic cells to produce IL-12 during the interaction with Th cells. J. Immunol. 162, 3231±3236. Kamogawa, Y., Minasi, L. A., Carding, S. R., Bottomly, K., and Flavell, R. A. (1993). The relationship of IL-4- and IFN-g-producing T cells studied by lineage ablation of IL-4-producing cells. Cell 75, 985±995. Kansas, G. S. (1996). Selectins and their ligands: Current concepts and controversies. Blood 88, 3259±3287. Kaplan, M. H., Schindler, U., Smiley, S. T., and Grusby, M. J. (1996a). Stat6 is required for mediating responses to IL-4 and for development of Th2 cells. Immunity 4, 313±319. Kaplan, M. H., Sun, Y. L., Hoey, T., and Grusby, M. J. (1996b). Impaired IL-12 responses and enhanced development of Th2 cells in Stat4-de®cient mice. Nature 382, 174±177. Kaplan, M. H., Wurster, A. L., and Grusby, M. J. (1998). A signal transducer and activator of transcription (Stat)4-independent pathway for the development of T helper type 1 cells. J. Exp. Med. 188, 1191±1196. Kaplan, M. H., Wurster, A. L., Smiley, S. T., and Grusby, M. J. (1999). Stat6-dependent and -independent pathways for IL-4 production. J. Immunol. 163, 6536±6540. Karaghiosoff, M., Neubauer, H., Lassnig, C., Kovarik, P., Schindler, H., Pircher, H., McCoy, B., Bogdan, C., Decker, T., Brem, G., Pfeffer, K., and Muller, M. (2000). Partial impairment of cytokine responses in Tyk2-de®cient mice. Immunity 13, 549±560. Karpus, W. J., Lukacs, N. W., Kennedy, K. J., Smith, W. S., Hurst, S. D., and Barrett, T. A. (1997). Differential CC chemokine-induced enhancement of T helper cell cytokine production. J. Immunol. 158, 4129±4136. Kashiwakura, J., Suzuki, N., Nagafuchi, H., Takeno, M., Takeba, Y., Shimoyama, Y., and Sakane, T. (1999). Txk, a nonreceptor tyrosine kinase of the Tec family, is expressed in T helper type 1 cells and regulates interferon g production in human T lymphocytes. J. Exp. Med. 190, 1147±1154. Kelly, B. L., and Locksley, R. M. (2000). Coordinate regulation of the IL-4, IL-13, and IL-5 cytokine cluster in Th2 clones revealed by allelic expression patterns. J. Immunol. 165, 2982±2986. Kiani, A., Viola, J. P., Lichtman, A. H., and Rao, A. (1997). Down-regulation of IL-4 gene transcription and control of Th2 cell differentiation by a mechanism involving NFAT1. Immunity 7, 849±860. Kim, J., Sif, S., Jones, B., Jackson, A., Koipally, J., Heller, E., Winandy, S., Viel, A., Sawyer, A., Ikeda, T., Kingston, R., and Georgopoulos, K. (1999a). Ikaros DNA-binding proteins direct formation of chromatin remodeling complexes in lymphocytes. Immunity 10, 345±355. Kim, J. I., Ho, I. C., Grusby, M. J., and Glimcher, L. H. (1999b). The transcription factor cMaf controls the production of interleukin-4 but not other Th2 cytokines. Immunity 10, 745±751. Kim, J. I., Li, T., Ho, I. C., Grusby, M. J., and Glimcher, L. H. (1999c). Requirement for the c-Maf transcription factor in crystallin gene regulation and lens development. Proc. Natl. Acad. Sci. USA 96, 3781±3785.

168

max loÈhning et al.

King, C., Mueller Hoenger, R., Malo Cleary, M., Murali-Krishna, K., Ahmed, R., King, E., and Sarvetnick, N. (2001). Interleukin-4 acts at the locus of the antigen-presenting dendritic cell to counter-regulate cytotoxic CD8‡ T-cell responses. Nat. Med. 7, 206±214. Klemenz, R., Hoffmann, S., and Werenskiold, A. K. (1989). Serum- and oncoprotein-mediated induction of a gene with sequence similarity to the gene encoding carcinoembryonic antigen. Proc. Natl. Acad. Sci. USA 86, 5708±5712. Kopf, M., Coyle, A. J., Schmitz, N., Barner, M., Oxenius, A., Gallimore, A., Gutierrez-Ramos, J. C., and Bachmann, M. F. (2000). Inducible costimulator protein (ICOS) controls T helper cell subset polarization after virus and parasite infection. J. Exp. Med. 192, 53±61. Kopf, M., Le Gros, G., Bachmann, M., Lamers, M. C., Bluethmann, H., and Kohler, G. (1993). Disruption of the murine IL-4 gene blocks Th2 cytokine responses. Nature 362, 245±248. Ku, C. C., Murakami, M., Sakamoto, A., Kappler, J., and Marrack, P. (2000). Control of homeostasis of CD8‡ memory T cells by opposing cytokines. Science 288, 675±678. Kubo, M., Ransom, J., Webb, D., Hashimoto, Y., Tada, T., and Nakayama, T. (1997). T-cell subsetspeci®c expression of the IL-4 gene is regulated by a silencer element and STAT6. EMBO J. 16, 4007±4020. Kubo, M., Yamashita, M., Abe, R., Tada, T., Okumura, K., Ransom, J. T., and Nakayama, T. (1999). CD28 costimulation accelerates IL-4 receptor sensitivity and IL-4-mediated Th2 differentiation. J. Immunol. 163, 2432±2442. KuÈhn, R., Rajewsky, K., and MuÈller, W. (1991). Generation and analysis of interleukin-4 de®cient mice. Science 254, 707±710. Kumar, S., Minnich, M. D., and Young, P. R. (1995). ST2/T1 protein functionally binds to two secreted proteins from Balb/c 3T3 and human umbilical vein endothelial cells but does not bind interleukin 1. J. Biol. Chem. 270, 27905±27913. Kurata, H., Lee, H. J., O'Garra, A., and Arai, N. (1999). Ectopic expression of activated Stat6 induces the expression of Th2-speci®c cytokines and transcription factors in developing Th1 cells. Immunity 11, 677±688. Lambrecht, B. N., De Veerman, M., Coyle, A. J., Gutierrez-Ramos, J. C., Thielemans, K., and Pauwels, R. A. (2000). Myeloid dendritic cells induce Th2 responses to inhaled antigen, leading to eosinophilic airway in¯ammation. J. Clin. Invest. 106, 551±559. Langenkamp, A., Messi, M., Lanzavecchia, A., and Sallusto, F. (2000). Kinetics of dendritic cell activation: Impact on priming of TH1, TH2 and nonpolarized T cells. Nat. Immunol. 1, 311±316. Lanzavecchia, A., and Sallusto, F. (2000a). Dynamics of T lymphocyte responses: Intermediates, effectors, and memory cells. Science 290, 92±97. Lanzavecchia, A., and Sallusto, F. (2000b). From synapses to immunological memory: The role of sustained T cell stimulation. Curr. Opin. Immunol. 12, 92±98. Le Gros, G., Ben-Sasson, S. Z., Seder, R., Finkelman, F. D., and Paul, W. E. (1990). Generation of interleukin 4 (IL-4)-producing cells in vivo and in vitro: IL-2 and IL-4 are required for in vitro generation of IL-4-producing cells. J. Exp. Med. 172, 921±929. Lederer, J. A., Perez, V. L., DesRoches, L., Kim, S. M., Abbas, A. K., and Lichtman, A. H. (1996). Cytokine transcriptional events during helper T cell subset differentiation. J. Exp. Med. 184, 397±406. Lee, G. R., Fields, P. E., and Flavell, R. A. (2001). Regulation of IL-4 gene expression by distal regulatory elements and GATA-3 at the chromatin level. Immunity 14, 447±459. Lee, H. J., O'Garra, A., Arai, K., and Arai, N. (1998). Characterization of cis-regulatory elements and nuclear factors conferring Th2-speci®c expression of the IL-5 gene: A role for a GATAbinding protein. J. Immunol. 160, 2343±2352. Lee, H. J., Takemoto, N., Kurata, H., Kamogawa, Y., Miyatake, S., O'Garra, A., and Arai, N. (2000). GATA-3 induces T helper cell type 2 (Th2) cytokine expression and chromatin remodeling in committed Th1 cells. J. Exp. Med. 192, 105±115.

cytokine memory of t helper lymphocytes

169

Li, B., Tournier, C., Davis, R. J., and Flavell, R. A. (1999). Regulation of IL-4 expression by the transcription factor JunB during T helper cell differentiation. EMBO J. 18, 420±432. Li, B., Yu, H., Zheng, W., Voll, R., Na, S., Roberts, A. W., Williams, D. A., Davis, R. J., Ghosh, S., and Flavell, R. A. (2000). Role of the guanosine triphosphatase Rac2 in T helper 1 cell differentiation. Science 288, 2219±2222. Lieuw, K. H., Li, G., Zhou, Y., Grosveld, F., and Engel, J. D. (1997). Temporal and spatial control of murine GATA-3 transcription by promoter-proximal regulatory elements. Dev. Biol. 188, 1±16. Lim, Y. C., Henault, L., Wagers, A. J., Kansas, G. S., Luscinskas, F. W., and Lichtman, A. H. (1999). Expression of functional selectin ligands on Th cells is differentially regulated by IL-12 and IL-4. J. Immunol. 162, 3193±3201. Linehan, L. A., Warren, W. D., Thompson, P. A., Grusby, M. J., and Berton, M. T. (1998). STAT6 is required for IL-4-induced germline Ig gene transcription and switch recombination. J. Immunol. 161, 302±310. Ling, V., Wu, P. W., Finnerty, H. F., Bean, K. M., Spaulding, V., Fouser, L. A., Leonard, J. P., Hunter, S. E., Zollner, R., Thomas, J. L., Miyashiro, J. S., Jacobs, K. A., and Collins, M. (2000). Cutting edge: Identi®cation of GL50, a novel B7-like protein that functionally binds to ICOS receptor. J. Immunol. 164, 1653±1657. Liu, B., Liao, J., Rao, X., Kushner, S. A., Chung, C. D., Chang, D. D., and Shuai, K. (1998a). Inhibition of Stat1-mediated gene activation by PIAS1. Proc. Natl. Acad. Sci. USA 95, 10626±10631. Liu, K. Q., Bunnell, S. C., Gurniak, C. B., and Berg, L. J. (1998b). T cell receptor-initiated calcium release is uncoupled from capacitative calcium entry in Itk-de®cient T cells. J. Exp. Med. 187, 1721±1727. Lloyd, C. M., Delaney, T., Nguyen, T., Tian, J., Martinez, A. C., Coyle, A. J., and GutierrezRamos, J. C. (2000). CC chemokine receptor (CCR)3/eotaxin is followed by CCR4/monocytederived chemokine in mediating pulmonary T helper lymphocyte type 2 recruitment after serial antigen challenge in vivo. J. Exp. Med. 191, 265±274. Loetscher, P., Moser, B., and Baggiolini, M. (2000). Chemokines and their receptors in lymphocyte traf®c and HIV infection. Adv. Immunol. 74, 127±180. Loetscher, P., Seitz, M., Baggiolini, M., and Moser, B. (1996). Interleukin-2 regulates CC chemokine receptor expression and chemotactic responsiveness in T lymphocytes. J. Exp. Med. 184, 569±577. Loetscher, P., Uguccioni, M., Bordoli, L., Baggiolini, M., Moser, B., Chizzolini, C., and Dayer, J. M. (1998). CCR5 is characteristic of Th1 lymphocytes. Nature 391, 344±345. LoÈhning, M., Grogan, J. L., Coyle, A. J., Yazdanbakhsh, M., Meisel, C., Gutierrez-Ramos, J. C., Radbruch, A., and Kamradt, T. (1999). T1/ST2 expression is enhanced on CD4‡ T cells from schistosome egg-induced granulomas: Analysis of Th cell cytokine coexpression ex vivo. J. Immunol. 162, 3882±3889. LoÈhning, M., Stroehmann, A., Coyle, A. J., Grogan, J. L., Lin, S., Gutierrez-Ramos, J. C., Levinson, D., Radbruch, A., and Kamradt, T. (1998). T1/ST2 is preferentially expressed on murine Th2 cells, independent of interleukin 4, interleukin 5, and interleukin 10, and important for Th2 effector function. Proc. Natl. Acad. Sci. USA 95, 6930±6935. Lohoff, M., Duncan, G. S., Ferrick, D., MittruÈcker, H. W., Bischof, S., Prechtl, S., Rollinghoff, M., Schmitt, E., Pahl, A., and Mak, T. W. (2000). De®ciency in the transcription factor interferon regulatory factor (IRF)-2 leads to severely compromised development of natural killer and T helper type 1 cells. J. Exp. Med. 192, 325±336. Lohoff, M., Ferrick, D., MittruÈcker, H. W., Duncan, G. S., Bischof, S., Rollinghoff, M., and Mak, T. W. (1997). Interferon regulatory factor-1 is required for a T helper 1 immune response in vivo. Immunity 6, 681±689.

170

max loÈhning et al.

Loots, G. G., Locksley, R. M., Blankespoor, C. M., Wang, Z. E., Miller, W., Rubin, E. M., and Frazer, K. A. (2000). Identi®cation of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288, 136±140. Losman, J. A., Chen, X. P., Hilton, D., and Rothman, P. (1999). Cutting edge: SOCS-1 is a potent inhibitor of IL-4 signal transduction. J. Immunol. 162, 3770±3774. Lu, J. R., McKinsey, T. A., Xu, H., Wang, D. Z., Richardson, J. A., and Olson, E. N. (1999). FOG-2, a heart- and brain-enriched cofactor for GATA transcription factors. Mol. Cell. Biol. 19, 4495±4502. Luther, S. A., and Cyster, J. G. (2001). Chemokines as regulators of T cell differentiation. Nat. Immunol. 2, 102±107. Macatonia, S. E., Hosken, N. A., Litton, M., Vieira, P., Hsieh, C. S., Culpepper, J. A., Wysocka, M., Trinchieri, G., Murphy, K. M., and O'Garra, A. (1995). Dendritic cells produce IL-12 and direct the development of Th1 cells from naive CD4‡ T cells. J. Immunol. 154, 5071±5079. Macian, F., Garcia-Rodriguez, C., and Rao, A. (2000). Gene expression elicited by NFAT in the presence or absence of cooperative recruitment of Fos and Jun. EMBO J. 19, 4783±4795. Mages, H. W., Hutloff, A., Heuck, C., Buchner, K., Himmelbauer, H., Oliveri, F., and Kroczek, R. A. (2000). Molecular cloning and characterization of murine ICOS and identi®cation of B7h as ICOS ligand. Eur. J. Immunol. 30, 1040±1047. Maghazachi, A. A., Al-Aoukaty, A., and Schall, T. J. (1994). C-C chemokines induce the chemotaxis of NK and IL-2-activated NK cells. Role for G proteins. J. Immunol. 153, 4969±4977. Magram, J., Connaughton, S. E., Warrier, R. R., Carvajal, D. M., Wu, C. Y., Ferrante, J., Stewart, C., Sarmiento, U., Faherty, D. A., and Gately, M. K. (1996). IL-12 de®cient mice are defective in IFN-g production and type 1 cytokine responses. Immunity 4, 471±481. Maldonado-Lopez, R., De Smedt, T., Michel, P., Godfroid, J., Pajak, B., Heirman, C., Thielemans, K., Leo, O., Urbain, J., and Moser, M. (1999). CD8a‡ and CD8a subclasses of dendritic cells direct the development of distinct T helper cells in vivo. J. Exp. Med. 189, 587±592. Maly, P., Thall, A., Petryniak, B., Rogers, C. E., Smith, P. L., Marks, R. M., Kelly, R. J., Gersten, K. M., Cheng, G., Saunders, T. L., Camper, S. A., Camphausen, R. T., Sullivan, F. X., Isogai, Y., Hindsgaul, O., von Andrian, U. H., and Lowe, J. B. (1996). The a(1,3)fucosyltransferase FucTVII controls leukocyte traf®cking through an essential role in L-, E-, and P-selectin ligand biosynthesis. Cell 86, 643±653. Manetti, R., Parronchi, P., Giudizi, M. G., Piccinni, M.-P., Maggi, E., Trinchieri, G., and Romagnani, S. (1993). Natural killer cell stimulatory factor (interleukin 12 [IL-12]) induces T helper type 1 (Th1)-speci®c immune responses and inhibits the development of IL-4-producing Th cells. J. Exp. Med. 177, 1199±1204. Manz, R., Assenmacher, M., P¯uger, E., Miltenyi, S., and Radbruch, A. (1995). Analysis and sorting of live cells according to secreted molecules, relocated to a cell-surface af®nity matrix. Proc. Natl. Acad. Sci. USA 92, 1921±1925. Marks-Konczalik, J., Dubois, S., Losi, J. M., Sabzevari, H., Yamada, N., Feigenbaum, L., Waldmann, T. A., and Tagaya, Y. (2000). IL-2-induced activation-induced cell death is inhibited in IL15 transgenic mice. Proc. Natl. Acad. Sci. USA 97, 11445±11450. Masopust, D., Vezys, V., Marzo, A. L. and Lefrancois, L. (2001). Preferential localization of effector memory cells in nonlymphoid tissue. Science 291, 2413±2417. Matesanz, F., Delgado, C., Fresno, M., and Alcina, A. (2000). Allelic selection of human IL-2 gene. Eur. J. Immunol. 30, 3516±3521. McAdam, A. J., Chang, T. T., Lumelsky, A. E., Green®eld, E. A., Boussiotis, V. A., Duke-Cohan, J. S., Chernova, T., Malenkovich, N., Jabs, C., Kuchroo, V. K., Ling, V., Collins, M., Sharpe, A. H., and Freeman, G. J. (2000). Mouse inducible costimulatory molecule (ICOS) expression is enhanced by CD28 costimulation and regulates differentiation of CD4‡ T cells. J. Immunol. 165, 5035±5040.

cytokine memory of t helper lymphocytes

171

McAdam, A. J., Greenwald, R. J., Levin, M. A., Chernova, T., Malenkovich, N., Ling, V., Freeman, G. J., and Sharpe, A. H. (2001). ICOS is critical for CD40-mediated antibody class switching. Nature 409, 102±105. Meisel, C., Bonhagen, K., LoÈhning, M., Coyle, A. J., Gutierrez-Ramos, J. C., Radbruch, A., and Kamradt, T. (2001). Regulation and function of T1/ST2 expression on CD4‡ T cells: Induction of type 2 cytokine production by T1/ST2 cross-linking. J. Immunol. 166, 3143±3150. Melvin, A. J., McGurn, M. E., Bort, S. J., Gibson, C., and Lewis, D. B. (1995). Hypomethylation of the interferon-g gene correlates with its expression by primary T-lineage cells. Eur. J. Immunol. 25, 426±430. Miaw, S. C., Choi, A., Yu, E., Kishikawa, H., and Ho, I. C. (2000). ROG, repressor of GATA, regulates the expression of cytokine genes. Immunity 12, 323±333. Mikita, T., Campbell, D., Wu, P., Williamson, K., and Schindler, U. (1996). Requirements for interleukin-4-induced gene expression and functional characterization of Stat6. Mol. Cell. Biol. 16, 5811±5820. Mitcham, J. L., Parnet, P., Bonnert, T. P., Garka, K. E., Gerhart, M. J., Slack, J. L., Gayle, M. A., Dower, S. K., and Sims, J. E. (1996). T1/ST2 signaling establishes it as a member of an expanding interleukin-1 receptor family. J. Biol. Chem. 271, 5777±5783. Miyazaki, T., Kawahara, A., Fujii, H., Nakagawa, Y., Minami, Y., Liu, Z. J., Oishi, I., Silvennoinen, O., Witthuhn, B. A., Ihle, J. N., et al. (1994). Functional activation of Jak1 and Jak3 by selective association with IL-2 receptor subunits. Science 266, 1045±1047. Mocci, S., and Coffman, R. L. (1995). Induction of a Th2 population from a polarized Leishmaniaspeci®c Th1 population by in vitro culture with IL-4. J. Immunol. 154, 3779±3787. Moore, K. L., Stults, N. L., Diaz, S., Smith, D. F., Cummings, R. D., Varki, A., and McEver, R. P. (1992). Identi®cation of a speci®c glycoprotein ligand for P-selectin (CD62) on myeloid cells. J. Cell Biol. 118, 445±456. Moritz, D. R., Rodewald, H. R., Gheyselinck, J., and Klemenz, R. (1998). The IL-1 receptor-related T1 antigen is expressed on immature and mature mast cells and on fetal blood mast cell progenitors. J. Immunol. 161, 4866±4874. Moser, B., and Loetscher, P. (2001). Lymphocyte traf®c control by chemokines. Nat. Immunol. 2, 123±128. Moser, M., and Murphy, K. M. (2000). Dendritic cell regulation of TH1±TH2 development. Nat. Immunol. 1, 199±205. Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L. (1986). Two types of murine helper T cell clone. I. De®nition according to pro®les of lymphokine activities and secreted proteins. J. Immunol. 136, 2348±2357. Mosmann, T. R., and Coffman, R. L. (1989). TH1 and TH2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7, 145±173. Mowen, K. A., Tang, J., Zhu, W., Schurter, B. T., Shuai, K., Herschman, H. R., and David, M. (2001). Arginine methylation of STAT1 modulates IFN-a=b-induced transcription. Cell 104, 731±741. Murphy, E., Shibuya, K., Hosken, N., Openshaw, P., Maino, V., Davis, K., Murphy, K., and O'Garra, A. (1996). Reversibility of T helper 1 and 2 populations is lost after long-term stimulation. J. Exp. Med. 183, 901±913. Murphy, K. M., Ouyang, W., Farrar, J. D., Yang, J., Ranganath, S., Asnagli, H., Afkarian, M., and Murphy, T. L. (2000). Signaling and transcription in T helper development. Annu. Rev. Immunol. 18, 451±494. Nagata, K., Hirai, H., Tanaka, K., Ogawa, K., Aso, T., Sugamura, K., Nakamura, M., and Takano, S. (1999a). CRTH2, an orphan receptor of T-helper-2-cells, is expressed on basophils and eosinophils and responds to mast cell-derived factor(s). FEBS Lett. 459, 195±199.

172

max loÈhning et al.

Nagata, K., Tanaka, K., Ogawa, K., Kemmotsu, K., Imai, T., Yoshie, O., Abe, H., Tada, K., Nakamura, M., Sugamura, K., and Takano, S. (1999b). Selective expression of a novel surface molecule by human Th2 cells in vivo. J. Immunol. 162, 1278±1286. Naka, T., Narazaki, M., Hirata, M., Matsumoto, T., Minamoto, S., Aono, A., Nishimoto, N., Kajita, T., Taga, T., Yoshizaki, K., Akira, S., and Kishimoto, T. (1997). Structure and function of a new STAT-induced STAT inhibitor. Nature 387, 924±929. Nan, X., Campoy, F. J., and Bird, A. (1997). MeCP2 is a transcriptional repressor with abundant binding sites in genomic chromatin. Cell 88, 471±481. Nan, X., Ng, H. H., Johnson, C. A., Laherty, C. D., Turner, B. M., Eisenman, R. N., and Bird, A. (1998). Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393, 386±389. Naramura, M., Hu, R. J., and Gu, H. (1998). Mice with a ¯uorescent marker for interleukin 2 gene activation. Immunity 9, 209±216. Nelms, K., Keegan, A. D., Zamorano, J., Ryan, J. J., and Paul, W. E. (1999). The IL-4 receptor: Signaling mechanisms and biologic functions. Annu. Rev. Immunol. 17, 701±738. Nishikomori, R., Ehrhardt, R. O., and Strober, W. (2000). T helper type 2 cell differentiation occurs in the presence of interleukin 12 receptor b2 chain expression and signaling. J. Exp. Med. 191, 847±858. Noben-Trauth, N., Shultz, L. D., Brombacher, F., Urban, J. F., Jr., Gu, H., and Paul, W. E. (1997). An interleukin 4 (IL-4)-independent pathway for CD4‡ T cell IL-4 production is revealed in IL4 receptor-de®cient mice. Proc. Natl. Acad. Sci. USA 94, 10838±10843. Norgard, K. E., Moore, K. L., Diaz, S., Stults, N. L., Ushiyama, S., McEver, R. P., Cummings, R. D., and Varki, A. (1993). Characterization of a speci®c ligand for P-selectin on myeloid cells: A minor glycoprotein with sialylated O-linked oligosaccharides. J. Biol. Chem. 268, 12764±12774. O'Garra, A. (1998). Cytokines induce the development of functionally heterogeneous T helper cell subsets. Immunity 8, 275±283. O'Garra, A., and Arai, N. (2000). The molecular basis of T helper 1 and T helper 2 cell differentiation. Trends Cell Biol. 10, 542±550. Okamura, H., Tsutsi, H., Komatsu, T., Yutsudo, M., Hakura, A., Tanimoto, T., Torigoe, K., Okura, T., Nukada, Y., Hattori, K., et al. (1995). Cloning of a new cytokine that induces IFN-g production by T cells. Nature 378, 88±91. Openshaw, P., Murphy, E. E., Hosken, N. A., Maino, V., Davis, K., Murphy, K., and O'Garra, A. (1995). Heterogeneity of intracellular cytokine synthesis at the single-cell level in polarized T helper 1 and T helper 2 populations. J. Exp. Med. 182, 1357±1367. Oppmann, B., Lesley, R., Blom, B., Timans, J. C., Xu, Y., Hunte, B., Vega, F., Yu, N., Wang, J., Singh, K., Zonin, F., Vaisberg, E., Churakova, T., Liu, M., Gorman, D., Wagner, J., Zurawski, S., Liu, Y., Abrams, J. S., Moore, K. W., Rennick, D., de Waal-Malefyt, R., Hannum, C., Bazan, J. F., and Kastelein, R. A. (2000). Novel p19 protein engages IL-12p40 to form a cytokine, IL-23, with biological activities similar as well as distinct from IL-12. Immunity 13, 715±725. Oukka, M., Ho, I. C., de la Brousse, F. C., Hoey, T., Grusby, M. J., and Glimcher, L. H. (1998). The transcription factor NFAT4 is involved in the generation and survival of T cells. Immunity 9, 295±304. Ouyang, W., Jacobson, N. G., Bhattacharya, D., Gorham, J. D., Fenoglio, D., Sha, W. C., Murphy, T. L., and Murphy, K. M. (1999). The Ets transcription factor ERM is Th1-speci®c and induced by IL-12 through a Stat4-dependent pathway. Proc. Natl. Acad. Sci. USA 96, 3888±3893. Ouyang, W., LoÈhning, M., Gao, Z., Assenmacher, M., Ranganath, S., Radbruch, A., and Murphy, K. M. (2000). Stat6-independent GATA-3 autoactivation directs IL-4-independent Th2 development and commitment. Immunity 12, 27±37.

cytokine memory of t helper lymphocytes

173

Ouyang, W., Ranganath, S. H., Weindel, K., Bhattacharya, D., Murphy, T. L., Sha, W. C., and Murphy, K. M. (1998). Inhibition of Th1 development mediated by GATA-3 through an IL4-independent mechanism. Immunity 9, 745±755. Pandol®, P. P., Roth, M. E., Karis, A., Leonard, M. W., Dzierzak, E., Grosveld, F. G., Engel, J. D., and Lindenbaum, M. H. (1995). Targeted disruption of the GATA3 gene causes severe abnormalities in the nervous system and in fetal liver haematopoiesis. Nat. Genet. 11, 40±44. Park, A. Y., Hondowicz, B. D., and Scott, P. (2000). IL-12 is required to maintain a Th1 response during Leishmania major infection. J. Immunol. 165, 896±902. Parnet, P., Garka, K. E., Bonnert, T. P., Dower, S. K., and Sims, J. E. (1996). IL-1Rrp is a novel receptor-like molecule similar to the type I interleukin-1 receptor and its homologues T1/ST2 and IL-1R AcP. J. Biol. Chem. 271, 3967±3970. Parronchi, P., De Carli, M., Manetti, R., Simonelli, C., Sampognaro, S., Piccinni, M. P., Macchia, D., Maggi, E., Del Prete, G., and Romagnani, S. (1992). IL-4 and IFN a and g exert opposite regulatory effects on the development of cytolytic potential by Th1 or Th2 human T cell clones. J. Immunol. 149, 2977±2983. Patterson, B. K., Czerniewski, M., Andersson, J., Sullivan, Y., Su, F., Jiyamapa, D., Burki, Z., and Landay, A. (1999). Regulation of CCR5 and CXCR4 expression by type 1 and type 2 cytokines: CCR5 expression is downregulated by IL-10 in CD4-positive lymphocytes. Clin. Immunol. 91, 254±262. Peng, S. L., Gerth, A. J., Ranger, A. M., and Glimcher, L. H. (2001). NFATc1 and NFATc2 together control both T and B cell activation and differentiation. Immunity 14, 13±20. Perez, V. L., Lederer, J. A., Lichtman, A. H., and Abbas, A. K. (1995). Stability of Th1 and Th2 populations. Int. Immunol. 7, 869±875. Presky, D. H., Yang, H., Minetti, L. J., Chua, A. O., Nabavi, N., Wu, C. Y., Gately, M. K., and Gubler, U. (1996). A functional interleukin 12 receptor complex is composed of two b-type cytokine receptor subunits. Proc. Natl. Acad. Sci. USA 93, 14002±14007. Pulendran, B., Smith, J. L., Caspary, G., Brasel, K., Pettit, D., Maraskovsky, E., and Maliszewski, C. R. (1999). Distinct dendritic cell subsets differentially regulate the class of immune response in vivo. Proc. Natl. Acad. Sci. USA 96, 1036±1041. Qin, S., Rottman, J. B., Myers, P., Kassam, N., Weinblatt, M., Loetscher, M., Koch, A. E., Moser, B., and Mackay, C. R. (1998). The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain in¯ammatory reactions. J. Clin. Invest. 101, 746±754. Randolph, D. A., Huang, G., Carruthers, C. J., Bromley, L. E., and Chaplin, D. D. (1999). The role of CCR7 in TH1 and TH2 cell localization and delivery of B cell help in vivo. Science 286, 2159±2162. Ranganath, S., and Murphy, K. M. (2001). Structure and speci®city of GATA proteins in Th2 development. Mol. Cell. Biol. 21, 2716±2725. Ranganath, S., Ouyang, W., Bhattarcharya, D., Sha, W. C., Grupe, A., Peltz, G., and Murphy, K. M. (1998). GATA-3-dependent enhancer activity in IL-4 gene regulation. J. Immunol. 161, 3822±3826. Ranger, A. M., Hodge, M. R., Gravallese, E. M., Oukka, M., Davidson, L., Alt, F. W., de la Brousse, F. C., Hoey, T., Grusby, M., and Glimcher, L. H. (1998a). Delayed lymphoid repopulation with defects in IL-4-driven responses produced by inactivation of NF-ATc. Immunity 8, 125±134. Ranger, A. M., Oukka, M., Rengarajan, J., and Glimcher, L. H. (1998b). Inhibitory function of two NFAT family members in lymphoid homeostasis and Th2 development. Immunity 9, 627±635. Rao, A., Luo, C., and Hogan, P. G. (1997). Transcription factors of the NFAT family: Regulation and function. Annu. Rev. Immunol. 15, 707±747. Refaeli, Y., Van Parijs, L., London, C. A., Tschopp, J., and Abbas, A. K. (1998). Biochemical mechanisms of IL-2-regulated Fas-mediated T cell apoptosis. Immunity 8, 615±623.

174

max loÈhning et al.

Reid, S. D., Penna, G., and Adorini, L. (2000). The control of T cell responses by dendritic cell subsets. Curr. Opin. Immunol. 12, 114±121. Reikerstorfer, A., Holz, H., Stunnenberg, H. G., and Busslinger, M. (1995). Low af®nity binding of interleukin-1 b and intracellular signaling via NF-kB identify Fit-1 as a distant member of the interleukin-1 receptor family. J. Biol. Chem. 270, 17645±17648. Reiner, S. L., and Seder, R. A. (1999). Dealing from the evolutionary pawnshop: How lymphocytes make decisions. Immunity 11, 1±10. Reinhardt, R. L., Khoruts, A., Merica, R., Zell, T., and Jenkins, M. K. (2001). Visualizing the generation of memory CD4 T cells in the whole body. Nature 410, 101±105. Rengarajan, J., Mittelstadt, P. R., Mages, H. W., Gerth, A. J., Kroczek, R. A., Ashwell, J. D., and Glimcher, L. H. (2000). Sequential involvement of NFAT and Egr transcription factors in FasL regulation. Immunity 12, 293±300. Richter, A., LoÈhning, M., and Radbruch, A. (1999). Instruction for cytokine expression in T helper lymphocytes in relation to proliferation and cell cycle progression. J. Exp. Med. 190, 1439±1450. Rincon, M., Anguita, J., Nakamura, T., Fikrig, E., and Flavell, R. A. (1997). Interleukin (IL)-6 directs the differentiation of IL-4-producing CD4‡ T cells. J. Exp. Med. 185, 461±469. Rincon, M., Enslen, H., Raingeaud, J., Recht, M., Zapton, T., Su, M. S., Penix, L. A., Davis, R. J., and Flavell, R. A. (1998). Interferon-g expression by Th1 effector T cells mediated by the p38 MAP kinase signaling pathway. EMBO J. 17, 2817±2829. Rissoan, M. C., Soumelis, V., Kadowaki, N., Grouard, G., Briere, F., de Waal Malefyt, R., and Liu, Y. J. (1999). Reciprocal control of T helper cell and dendritic cell differentiation. Science 283, 1183±1186. Riviere, I., Sunshine, M. J., and Littman, D. R. (1998). Regulation of IL-4 expression by activation of individual alleles. Immunity 9, 217±228. Robertson, K. D., and Wolffe, A. P. (2000). DNA methylation in health and disease. Nat. Rev. Genet. 1, 11±19. Robinson, D., Shibuya, K., Mui, A., Zonin, F., Murphy, E., Sana, T., Hartley, S. B., Menon, S., Kastelein, R., Bazan, F., and O'Garra, A. (1997). IGIF does not drive Th1 development but synergizes with IL-12 for interferon-g production and activates IRAK and NFkB. Immunity 7, 571±581. Rodriguez-Palmero, M., Hara, T., Thumbs, A., and Hunig, T. (1999). Triggering of T cell proliferation through CD28 induces GATA-3 and promotes T helper type 2 differentiation in vitro and in vivo. Eur. J. Immunol. 29, 3914±3924. Rogge, L., Barberis-Maino, L., Bif®, M., Passini, N., Presky, D. H., Gubler, U., and Sinigaglia, F. (1997). Selective expression of an interleukin-12 receptor component by human T helper 1 cells. J. Exp. Med. 185, 825±831. Rogge, L., D'Ambrosio, D., Bif®, M., Penna, G., Minetti, L. J., Presky, D. H., Adorini, L., and Sinigaglia, F. (1998). The role of Stat4 in species-speci®c regulation of Th cell development by type I IFNs. J. Immunol. 161, 6567±6574. Romagnani, S. (1994). Lymphokine production by human T cells in disease states. Annu. Rev. Immunol. 12, 227±257. Roncarolo, M. G., Levings, M. K., and Traversari, C. (2001). Differentiation of T regulatory cells by immature dendritic cells. J. Exp. Med. 193, F5±F9. RoÈssler, U., Thomassen, E., Hultner, L., Baier, S., Danescu, J., and Werenskiold, A. K. (1995). Secreted and membrane-bound isoforms of T1, an orphan receptor related to IL-1-binding proteins, are differently expressed in vivo. Dev. Biol. 168, 86±97. Rothenberg, E. V., and Ward, S. B. (1996). A dynamic assembly of diverse transcription factors integrates activation and cell-type information for interleukin 2 gene regulation. Proc. Natl. Acad. Sci. USA 93, 9358±9365.

cytokine memory of t helper lymphocytes

175

Sabapathy, K., Hu, Y., Kallunki, T., Schreiber, M., David, J. P., Jochum, W., Wagner, E. F., and Karin, M. (1999). JNK2 is required for ef®cient T-cell activation and apoptosis but not for normal lymphocyte development. Curr. Biol. 9, 116±125. Sakai, M., Serria, M. S., Ikeda, H., Yoshida, K., Imaki, J., and Nishi, S. (2001). Regulation of c-maf gene expression by Pax6 in cultured cells. Nucleic Acids Res. 29, 1228±1237. Sako, D., Chang, X. J., Barone, K. M., Vachino, G., White, H. M., Shaw, G., Veldman, G. M., Bean, K. M., Ahern, T. J., Furie, B., et al. (1993). Expression cloning of a functional glycoprotein ligand for P-selectin. Cell 75, 1179±1186. Sallusto, F., Kremmer, E., Palermo, B., Hoy, A., Ponath, P., Qin, S., FoÈrster, R., Lipp, M., and Lanzavecchia, A. (1999a). Switch in chemokine receptor expression upon TCR stimulation reveals novel homing potential for recently activated T cells. Eur. J. Immunol. 29, 2037±2045. Sallusto, F., Lenig, D., FoÈrster, R., Lipp, M., and Lanzavecchia, A. (1999b). Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401, 708±712. Sallusto, F., Lenig, D., Mackay, C. R., and Lanzavecchia, A. (1998). Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J. Exp. Med. 187, 875±883. Sallusto, F., Mackay, C. R., and Lanzavecchia, A. (1997). Selective expression of the eotaxin receptor CCR3 by human T helper 2 cells. Science 277, 2005±2007. Saparov, A., Wagner, F. H., Zheng, R., Oliver, J. R., Maeda, H., Hockett, R. D., and Weaver, C. T. (1999). Interleukin-2 expression by a subpopulation of primary T cells is linked to enhanced memory/effector function. Immunity 11, 271±280. Schafer, P. H., Wadsworth, S. A., Wang, L., and Siekierka, J. J. (1999). p38a mitogen-activated protein kinase is activated by CD28-mediated signaling and is required for IL-4 production by human CD4‡CD45RO‡ T cells and Th2 effector cells. J. Immunol. 162, 7110±7119. Scheffold, A., Assenmacher, M., Reiners-Schramm, L., Lauster, R., and Radbruch, A. (2000). Highsensitivity immuno¯uorescence for detection of the pro- and anti-in¯ammatory cytokines g interferon and interleukin-10 on the surface of cytokine-secreting cells. Nat. Med. 6, 107±110. Scheicher, C., Mehlig, M., Dienes, H. P., and Reske, K. (1995). Uptake of microparticle-adsorbed protein antigen by bone marrow-derived dendritic cells results in up-regulation of interleukin1a and interleukin-12 p40/p35 and triggers prolonged, ef®cient antigen presentation. Eur. J. Immunol. 25, 1566±1572. Schmitt, E., Hoehn, P., Germann, T., and RuÈde, E. (1994). Differential effects of interleukin-12 on the development of naive mouse CD4‡ T cells. Eur. J. Immunol. 24, 343±347. Schmitz, J., Assenmacher, M., and Radbruch, A. (1993). Regulation of T helper cell cytokine expression: Functional dichotomy of antigen-presenting cells. Eur. J. Immunol. 23, 191±199. SchuÈler, T., Qin, Z., Ibe, S., Noben-Trauth, N., and Blankenstein, T. (1999). T helper cell type 1associated and cytotoxic T lymphocyte-mediated tumor immunity is impaired in interleukin 4de®cient mice. J. Exp. Med. 189, 803±810. Schultze, J. L., Michalak, S., Lowne, J., Wong, A., Gilleece, M. H., Gribben, J. G., and Nadler, L. M. (1999). Human non-germinal center B cell interleukin (IL)-12 production is primarily regulated by T cell signals CD40 ligand, interferon g and IL-10: Role of B cells in the maintenance of T cell responses. J. Exp. Med. 189, 1±12. Sebastiani, S., Allavena, P., Albanesi, C., Nasorri, F., Bianchi, G., Traidl, C., Sozzani, S., Girolomoni, G., and Cavani, A. (2001). Chemokine receptor expression and function in CD4‡ T lymphocytes with regulatory activity. J. Immunol. 166, 996±1002. Seder, R. A., Gazzinelli, R., Sher, A., and Paul, W. E. (1993). Interleukin 12 acts directly on CD4‡ T cells to enhance priming for interferon g production and diminishes interleukin 4 inhibition of such priming. Proc. Natl. Acad. Sci. USA 90, 10188±10192. Seder, R. A., and Paul, W. E. (1994). Acquisition of lymphokine-producing phenotype by CD4‡ T cells. Annu. Rev. Immunol. 12, 635±673.

176

max loÈhning et al.

Seder, R. A., Paul, W. E., Davis, M. M., and Fazekas de St. Groth, B. (1992). The presence of interleukin 4 during in vitro priming determines the lymphokine-producing potential of CD4‡ T cells from T cell receptor transgenic mice. J. Exp. Med. 176, 1091±1098. Senn, K. A., McCoy, K. D., Maloy, K. J., Stark, G., Frohli, E., Rulicke, T., and Klemenz, R. (2000). T1-de®cient and T1-Fc-transgenic mice develop a normal protective Th2-type immune response following infection with Nippostrongylus brasiliensis. Eur. J. Immunol. 30, 1929±1938. Ser¯ing, E., Berberich-Siebelt, F., Chuvpilo, S., Jankevics, E., Klein-Hessling, S., Twardzik, T., and Avots, A. (2000). The role of NF-AT transcription factors in T cell activation and differentiation. Biochim. Biophys. Acta 1498, 1±18. Shibasaki, F., Price, E. R., Milan, D., and McKeon, F. (1996). Role of kinases and the phosphatase calcineurin in the nuclear shuttling of transcription factor NF-AT4. Nature 382, 370±373. Shimoda, K., Kato, K., Aoki, K., Matsuda, T., Miyamoto, A., Shibamori, M., Yamashita, M., Numata, A., Takase, K., Kobayashi, S., Shibata, S., Asano, Y., Gondo, H., Sekiguchi, K., Nakayama, K., Nakayama, T., Okamura, T., Okamura, S., and Niho, Y. (2000). Tyk2 plays a restricted role in IFN-a signaling, although it is required for IL-12-mediated T cell function. Immunity 13, 561±571. Shimoda, K., van Deursen, J., Sangster, M. Y., Sarawar, S. R., Carson, R. T., Tripp, R. A., Chu, C., Quelle, F. W., Nosaka, T., Vignali, D. A., Doherty, P. C., Grosveld, G., Paul, W. E., and Ihle, J. N. (1996). Lack of IL-4-induced Th2 response and IgE class switching in mice with disrupted Stat6 gene. Nature 380, 630±633. Shuai, K., Schindler, C., Prezioso, V. R., and Darnell, J. E., Jr. (1992). Activation of transcription by IFN-g: Tyrosine phosphorylation of a 91-kD DNA binding protein. Science 258, 1808±1812. Siegel, M. D., Zhang, D. H., Ray, P., and Ray, A. (1995). Activation of the interleukin-5 promoter by cAMP in murine EL-4 cells requires the GATA-3 and CLE0 elements. J. Biol. Chem. 270, 24548±24555. Simon, A. R., Vikis, H. G., Stewart, S., Fanburg, B. L., Cochran, B. H., and Guan, K. L. (2000). Regulation of STAT3 by direct binding to the Rac1 GTPase. Science 290, 144±147. Sisk, T. J., Gourley, T., Roys, S., and Chang, C. H. (2000). MHC class II transactivator inhibits IL-4 gene transcription by competing with NF-AT to bind the coactivator CREB binding protein (CBP)/p300. J. Immunol. 165, 2511±2517. Skapenko, A., Lipsky, P. E., Kraetsch, H. G., Kalden, J. R., and Schulze-Koops, H. (2001). Antigenindependent Th2 cell differentiation by stimulation of CD28: Regulation via IL-4 gene expression and mitogen-activated protein kinase activation. J. Immunol. 166, 4283±4292. Slifka, M. K., Rodriguez, F., and Whitton, J. L. (1999). Rapid on/off cycling of cytokine production by virus-speci®c CD8‡ T cells. Nature 401, 76±79. Smerz-Bertling, C., and Duschl, A. (1995). Both interleukin 4 and interleukin 13 induce tyrosine phosphorylation of the 140-kDa subunit of the interleukin 4 receptor. J. Biol. Chem. 270, 966±970. Smits, H. H., Rietschoten, J. G., Hilkens, C. M., Sayilir, R., Stiekema, F., Kapsenberg, M. L., and Wierenga, E. A. (2001). IL-12-induced reversal of human Th2 cells is accompanied by full restoration of IL-12 responsiveness and loss of GATA-3 expression. Eur. J. Immunol. 31, 1055±1065. Sornasse, T., Larenas, P. V., Davis, K. A., de Vries, J. E., and Yssel, H. (1996). Differentiation and stability of T helper 1 and 2 cells derived from naive human neonatal CD4‡ T cells, analyzed at the single-cell level. J. Exp. Med. 184, 473±483. Stamatoyannopoulos, J. A., Goodwin, A., Joyce, T., and Lowrey, C. H. (1995). NF-E2 and GATA binding motifs are required for the formation of DNase I hypersensitive site 4 of the human bglobin locus control region. EMBO J. 14, 106±116. Starr, R., Willson, T. A., Viney, E. M., Murray, L. J., Rayner, J. R., Jenkins, B. J., Gonda, T. J., Alexander, W. S., Metcalf, D., Nicola, N. A., and Hilton, D. J. (1997). A family of cytokineinducible inhibitors of signalling. Nature 387, 917±921.

cytokine memory of t helper lymphocytes

177

Stobie, L., Gurunathan, S., Prussin, C., Sacks, D. L., Glaichenhaus, N., Wu, C. Y., and Seder, R. A. (2000). The role of antigen and IL-12 in sustaining Th1 memory cells in vivo: IL-12 is required to maintain memory/effector Th1 cells suf®cient to mediate protection to an infectious parasite challenge. Proc. Natl. Acad. Sci. USA 97, 8427±8432. Stoll, S., Jonuleit, H., Schmitt, E., Muller, G., Yamauchi, H., Kurimoto, M., Knop, J., and Enk, A. H. (1998). Production of functional IL-18 by different subtypes of murine and human dendritic cells (DC): DC-derived IL-18 enhances IL-12-dependent Th1 development. Eur. J. Immunol. 28, 3231±3239. Svensson, E. C., Tufts, R. L., Polk, C. E., and Leiden, J. M. (1999). Molecular cloning of FOG-2: A modulator of transcription factor GATA-4 in cardiomyocytes. Proc. Natl. Acad. Sci. USA 96, 956±961. Swain, S. L. (1994). Generation and in vivo persistence of polarized Th1 and Th2 memory cells. Immunity 1, 543±552. Swain, S. L., Weinberg, A. D., English, M., and Huston, G. (1990). IL-4 directs the development of Th2-like helper effectors. J. Immunol. 145, 3796±3806. Swallow, M. M., Wallin, J. J., and Sha, W. C. (1999). B7h, a novel costimulatory homolog of B7.1 and B7.2, is induced by TNF-a. Immunity 11, 423±432. Syrbe, U., Siveke, J., and Hamann, A. (1999). Th1/Th2 subsets: Distinct differences in homing and chemokine receptor expression? Springer Semin. Immunopathol. 21, 263±285. Szabo, S. J., Dighe, A. S., Gubler, U., and Murphy, K. M. (1997a). Regulation of the interleukin (IL)-12R b2 subunit expression in developing T helper 1 (Th1) and Th2 cells. J. Exp. Med. 185, 817±824. Szabo, S. J., Glimcher, L. H., and Ho, I. C. (1997b). Genes that regulate interleukin-4 expression in T cells. Curr. Opin. Immunol. 9, 776±781. Szabo, S. J., Jacobson, N. G., Dighe, A. S., Gubler, U., and Murphy, K. M. (1995). Developmental commitment to the Th2 lineage by extinction of IL-12 signaling. Immunity 2, 665±675. Szabo, S. J., Kim, S. T., Costa, G. L., Zhang, X., Fathman, C. G., and Glimcher, L. H. (2000). A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100, 655±669. Tafuri, A., Shahinian, A., Bladt, F., Yoshinaga, S. K., Jordana, M., Wakeham, A., Boucher, L. M., Bouchard, D., Chan, V. S., Duncan, G., Odermatt, B., Ho, A., Itie, A., Horan, T., Whoriskey, J. S., Pawson, T., Penninger, J. M., Ohashi, P. S., and Mak, T. W. (2001). ICOS is essential for effective T-helper-cell responses. Nature 409, 105±109. Takeda, K., Tanaka, T., Shi, W., Matsumoto, M., Minami, M., Kashiwamura, S., Nakanishi, K., Yoshida, N., Kishimoto, T., and Akira, S. (1996). Essential role of Stat6 in IL-4 signalling. Nature 380, 627±630. Takeda, K., Tsutsui, H., Yoshimoto, T., Adachi, O., Yoshida, N., Kishimoto, T., Okamura, H., Nakanishi, K., and Akira, S. (1998). Defective NK cell activity and Th1 response in IL-18de®cient mice. Immunity 8, 383±390. Takemoto, N., Kamogawa, Y., Jun Lee, H., Kurata, H., Arai, K. I., O'Garra, A., Arai, N., and Miyatake, S. (2000). Cutting edge: Chromatin remodeling at the IL-4/IL-13 intergenic regulatory region for Th2-speci®c cytokine gene cluster. J. Immunol. 165, 6687±6691. Takemoto, N., Koyano-Nakagawa, N., Yokota, T., Arai, N., Miyatake, S., and Arai, K. (1998). Th2speci®c DNase I-hypersensitive sites in the murine IL-13 and IL-4 intergenic region. Int. Immunol. 10, 1981±1985. Taki, S., Sato, T., Ogasawara, K., Fukuda, T., Sato, M., Hida, S., Suzuki, G., Mitsuyama, M., Shin, E. H., Kojima, S., Taniguchi, T., and Asano, Y. (1997). Multistage regulation of Th1-type immune responses by the transcription factor IRF-1. Immunity 6, 673±679. Tevosian, S. G., Deconinck, A. E., Cantor, A. B., Rieff, H. I., Fujiwara, Y., Corfas, G., and Orkin, S. H. (1999). FOG-2: A novel GATA-family cofactor related to multitype zinc-®nger proteins Friend of GATA-1 and U-shaped. Proc. Natl. Acad. Sci. USA 96, 950±955.

178

max loÈhning et al.

Tevosian, S. G., Deconinck, A. E., Tanaka, M., Schinke, M., Litovsky, S. H., Izumo, S., Fujiwara, Y., and Orkin, S. H. (2000). FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell 101, 729±739. Thierfelder, W. E., van Deursen, J. M., Yamamoto, K., Tripp, R. A., Sarawar, S. R., Carson, R. T., Sangster, M. Y., Vignali, D. A., Doherty, P. C., Grosveld, G. C., and Ihle, J. N. (1996). Requirement for Stat4 in interleukin-12-mediated responses of natural killer and T cells. Nature 382, 171±174. Thoma, S., Bonhagen, K., Vestweber, D., Hamann, A., and Reimann, J. (1998). Expression of selectin-binding epitopes and cytokines by CD4‡ T cells repopulating scid mice with colitis. Eur. J. Immunol. 28, 1785±1797. Tietz, W., Allemand, Y., Borges, E., von Laer, D., Hallmann, R., Vestweber, D., and Hamann, A. (1998). CD4‡ T cells migrate into in¯amed skin only if they express ligands for E- and Pselection. J. Immunol. 161, 963±970. Ting, C. N., Olson, M. C., Barton, K. P., and Leiden, J. M. (1996). Transcription factor GATA-3 is required for development of the T-cell lineage. Nature 384, 474±478. Tominaga, S. (1989). A putative protein of a growth speci®c cDNA from BALB/c-3T3 cells is highly similar to the extracellular portion of mouse interleukin 1 receptor. FEBS Lett. 258, 301±304. Torigoe, K., Ushio, S., Okura, T., Kobayashi, S., Taniai, M., Kunikata, T., Murakami, T., Sanou, O., Kojima, H., Fujii, M., Ohta, T., Ikeda, M., Ikegami, H., and Kurimoto, M. (1997). Puri®cation and characterization of the human interleukin-18 receptor. J. Biol. Chem. 272, 25737±25742. Townsend, M. J., Fallon, P. G., Matthews, D. J., Jolin, H. E., and McKenzie, A. N. (2000). T1/ST2de®cient mice demonstrate the importance of T1/ST2 in developing primary T helper cell type 2 responses. J. Exp. Med. 191, 1069±1076. Tsang, A. P., Fujiwara, Y., Hom, D. B., and Orkin, S. H. (1998). Failure of megakaryopoiesis and arrested erythropoiesis in mice lacking the GATA-1 transcriptional cofactor FOG. Genes Dev. 12, 1176±1188. Tsang, A. P., Visvader, J. E., Turner, C. A., Fujiwara, Y., Yu, C., Weiss, M. J., Crossley, M., and Orkin, S. H. (1997). FOG, a multitype zinc ®nger protein, acts as a cofactor for transcription factor GATA-1 in erythroid and megakaryocytic differentiation. Cell 90, 109±119. Turner, J. M., Brodsky, M. H., Irving, B. A., Levin, S. D., Perlmutter, R. M., and Littman, D. R. (1990). Interaction of the unique N-terminal region of tyrosine kinase p56lck with cytoplasmic domains of CD4 and CD8 is mediated by cysteine motifs. Cell 60, 755±765. Van Esch, H., Groenen, P., Nesbit, M. A., Schuffenhauer, S., Lichtner, P., Vanderlinden, G., Harding, B., Beetz, R., Bilous, R. W., Holdaway, I., Shaw, N. J., Fryns, J. P., Van de Ven, W., Thakker, R. V., and Devriendt, K. (2000). GATA3 haplo-insuf®ciency causes human HDR syndrome. Nature 406, 419±422. Van Parijs, L., and Abbas, A. K. (1998). Homeostasis and self-tolerance in the immune system: Turning lymphocytes off. Science 280, 243±248. Varadhachary, A. S., Perdow, S. N., Hu, C., Ramanarayanan, M., and Salgame, P. (1997). Differential ability of T cell subsets to undergo activation-induced cell death. Proc. Natl. Acad. Sci. USA 94, 5778±5783. Vieira, P. L., de Jong, E. C., Wierenga, E. A., Kapsenberg, M. L., and Kalinski, P. (2000). Development of Th1-inducing capacity in myeloid dendritic cells requires environmental instruction. J. Immunol. 164, 4507±4512. Wagers, A. J., Waters, C. M., Stoolman, L. M., and Kansas, G. S. (1998). Interleukin 12 and interleukin 4 control T cell adhesion to endothelial selectins through opposite effects on a1,3fucosyltransferase VII gene expression. J. Exp. Med. 188, 2225±2231. Waldmann, T. A., Dubois, S., and Tagaya, Y. (2001). Contrasting roles of IL-2 and IL-15 in the life and death of lymphocytes: Implications for immunotherapy. Immunity 14, 105±110. Watts, T. H., and DeBenedette, M. A. (1999). T cell co-stimulatory molecules other than CD28. Curr. Opin. Immunol. 11, 286±293.

cytokine memory of t helper lymphocytes

179

Weiss, L., Whitmarsh, A. J., Yang, D. D., Rincon, M., Davis, R. J., and Flavell, R. A. (2000). Regulation of c-Jun NH(2)-terminal kinase (Jnk) gene expression during T cell activation. J. Exp. Med. 191, 139±146. Weninger, W., Ulfman, L. H., Cheng, G., Souchkova, N., Quackenbush, E. J., Lowe, J. B., and von Andrian, U. H. (2000). Specialized contributions by a(1,3)-fucosyltransferase-IV and FucT-VII during leukocyte rolling in dermal microvessels. Immunity 12, 665±676. Whelan, M., Harnett, M. M., Houston, K. M., Patel, V., Harnett, W., and Rigley, K. P. (2000). A ®larial nematode-secreted product signals dendritic cells to acquire a phenotype that drives development of Th2 cells. J. Immunol. 164, 6453±6460. Witthuhn, B. A., Silvennoinen, O., Miura, O., Lai, K. S., Cwik, C., Liu, E. T., and Ihle, J. N. (1994). Involvement of the Jak-3 Janus kinase in signalling by interleukins 2 and 4 in lymphoid and myeloid cells. Nature 370, 153±157. Wurster, A. L., Tanaka, T., and Grusby, M. J. (2000). The biology of Stat4 and Stat6. Oncogene 19, 2577±2584. Xanthoudakis, S., Viola, J. P., Shaw, K. T., Luo, C., Wallace, J. D., Bozza, P. T., Luk, D. C., Curran, T., and Rao, A. (1996). An enhanced immune response in mice lacking the transcription factor NFAT1. Science 272, 892±895. Xu, D., Chan, W. L., Leung, B. P., Huang, F., Wheeler, R., Piedra®ta, D., Robinson, J. H., and Liew, F. Y. (1998a). Selective expression of a stable cell surface molecule on type 2 but not type 1 helper T cells. J. Exp. Med. 187, 787±794. Xu, D., Chan, W. L., Leung, B. P., Hunter, D., Schulz, K., Carter, R. W., McInnes, I. B., Robinson, J. H., and Liew, F. Y. (1998b). Selective expression and functions of interleukin 18 receptor on T helper (Th) type 1 but not Th2 cells. J. Exp. Med. 188, 1485±1492. Yamagata, T., Mitani, K., Oda, H., Suzuki, T., Honda, H., Asai, T., Maki, K., Nakamoto, T., and Hirai, H. (2000). Acetylation of GATA-3 affects T-cell survival and homing to secondary lymphoid organs. EMBO J. 19, 4676±4687. Yamashita, M., Hashimoto, K., Kimura, M., Kubo, M., Tada, T., and Nakayama, T. (1998). Requirement for p56(lck) tyrosine kinase activation in Th subset differentiation. Int. Immunol. 10, 577±591. Yamashita, M., Katsumata, M., Iwashima, M., Kimura, M., Shimizu, C., Kamata, T., Shin, T., Seki, N., Suzuki, S., Taniguchi, M., and Nakayama, T. (2000). T cell receptor-induced calcineurin activation regulates T helper type 2 cell development by modifying the interleukin 4 receptor signaling complex. J. Exp. Med. 191, 1869±1879. Yamashita, M., Kimura, M., Kubo, M., Shimizu, C., Tada, T., Perlmutter, R. M., and Nakayama, T. (1999). T cell antigen receptor-mediated activation of the Ras/mitogen-activated protein kinase pathway controls interleukin 4 receptor function and type-2 helper T cell differentiation. Proc. Natl. Acad. Sci. USA 96, 1024±1029. Yan, H., Krishnan, K., Greenlund, A. C., Gupta, S., Lim, J. T., Schreiber, R. D., Schindler, C. W., and Krolewski, J. J. (1996). Phosphorylated interferon-a receptor 1 subunit (IFNaR1) acts as a docking site for the latent form of the 113 kDa STAT2 protein. EMBO J. 15, 1064±1074. Yang, D. D., Conze, D., Whitmarsh, A. J., Barrett, T., Davis, R. J., Rincon, M., and Flavell, R. A. (1998a). Differentiation of CD4‡ T cells to Th1 cells requires MAP kinase JNK2. Immunity 9, 575±585. Yang, J., Murphy, T. L., Ouyang, W., and Murphy, K. M. (1999). Induction of interferon-g production in Th1 CD4‡ T cells: Evidence for two distinct pathways for promoter activation. Eur. J. Immunol. 29, 548±555. Yang, J., Zhu, H., Murphy, T. L., Ouyang, W., and Murphy, K. M. (2001). IL-18-stimulated GADD45b required in cytokine-induced, but not TCR-induced, IFN-g production. Nat. Immunol. 2, 157±164.

180

max loÈhning et al.

Yang, Y., Chang, J. F., Parnes, J. R., and Fathman, C. G. (1998b). T cell receptor (TCR) engagement leads to activation-induced splicing of tumor necrosis factor (TNF) nuclear pre-mRNA. J. Exp. Med. 188, 247±254. Yap, G., Pesin, M., and Sher, A. (2000). Cutting edge: IL-12 is required for the maintenance of IFNg production in T cells mediating chronic resistance to the intracellular pathogen Toxoplasma gondii. J. Immunol. 165, 628±631. Ye, B. H., Cattoretti, G., Shen, Q., Zhang, J., Hawe, N., de Waard, R., Leung, C., Nouri-Shirazi, M., Orazi, A., Chaganti, R. S., Rothman, P., Stall, A. M., Pandol®, P. P., and Dalla-Favera, R. (1997). The BCL-6 proto-oncogene controls germinal-centre formation and Th2-type in¯ammation. Nat. Genet. 16, 161±170. Ying, S., Robinson, D. S., Meng, Q., Rottman, J., Kennedy, R., Ringler, D. J., Mackay, C. R., Daugherty, B. L., Springer, M. S., Durham, S. R., Williams, T. J., and Kay, A. B. (1997). Enhanced expression of eotaxin and CCR3 mRNA and protein in atopic asthma: Association with airway hyperresponsiveness and predominant co-localization of eotaxin mRNA to bronchial epithelial and endothelial cells. Eur. J. Immunol. 27, 3507±3516. Yoshida, H., Nishina, H., Takimoto, H., Marengere, L. E., Wakeham, A. C., Bouchard, D., Kong, Y. Y., Ohteki, T., Shahinian, A., Bachmann, M., Ohashi, P. S., Penninger, J. M., Crabtree, G. R., and Mak, T. W. (1998). The transcription factor NF-ATc1 regulates lymphocyte proliferation and Th2 cytokine production. Immunity 8, 115±124. Yoshimoto, T., Okamura, H., Tagawa, Y. I., Iwakura, Y., and Nakanishi, K. (1997). Interleukin 18 together with interleukin 12 inhibits IgE production by induction of interferon-g production from activated B cells. Proc. Natl. Acad. Sci. USA 94, 3948±3953. Yoshimoto, T., Takeda, K., Tanaka, T., Ohkusu, K., Kashiwamura, S., Okamura, H., Akira, S., and Nakanishi, K. (1998). IL-12 up-regulates IL-18 receptor expression on T cells, Th1 cells, and B cells: Synergism with IL-18 for IFN-g production. J. Immunol. 161, 3400±3407. Yoshinaga, S. K., Whoriskey, J. S., Khare, S. D., Sarmiento, U., Guo, J., Horan, T., Shih, G., Zhang, M., Coccia, M. A., Kohno, T., Tafuri-Bladt, A., Brankow, D., Campbell, P., Chang, D., Chiu, L., Dai, T., Duncan, G., Elliott, G. S., Hui, A., McCabe, S. M., Scully, S., Shahinian, A., Shaklee, C. L., Van, G., Mak, T. W., et al. (1999). T-cell co-stimulation through B7RP-1 and ICOS. Nature 402, 827±832. Young, H. A., Ghosh, P., Ye, J., Lederer, J., Lichtman, A., Gerard, J. R., Penix, L., Wilson, C. B., Melvin, A. J., McGurn, M. E., et al. (1994). Differentiation of the T helper phenotypes by analysis of the methylation state of the IFN-g gene. J. Immunol. 153, 3603±3610. Yssel, H., Fasler, S., de Vries, J. E., and de Waal Malefyt, R. (1994). IL-12 transiently induces IFN-g transcription and protein synthesis in human CD4‡ allergen-speci®c Th2 T cell clones. Int. Immunol. 6, 1091±1096. Zhang, D. H., Cohn, L., Ray, P., Bottomly, K., and Ray, A. (1997a). Transcription factor GATA-3 is differentially expressed in murine Th1 and Th2 cells and controls Th2-speci®c expression of the interleukin-5 gene. J. Biol. Chem. 272, 21597±21603. Zhang, D. H., Yang, L., Cohn, L., Parkyn, L., Homer, R., Ray, P., and Ray, A. (1999a). Inhibition of allergic in¯ammation in a murine model of asthma by expression of a dominant-negative mutant of GATA-3. Immunity 11, 473±482. Zhang, D. H., Yang, L., and Ray, A. (1998). Differential responsiveness of the IL-5 and IL-4 genes to transcription factor GATA-3. J. Immunol. 161, 3817±3821. Zhang, J., Salojin, K. V., Gao, J. X., Cameron, M. J., Bergerot, I., and Delovitch, T. L. (1999b). p38 mitogen-activated protein kinase mediates signal integration of TCR/CD28 costimulation in primary murine T cells. J. Immunol. 162, 3819±3829. Zhang, X., Brunner, T., Carter, L., Dutton, R. W., Rogers, P., Bradley, L., Sato, T., Reed, J. C., Green, D., and Swain, S. L. (1997b). Unequal death in T helper cell (Th)1 and Th2 effectors: Th1, but not Th2, effectors undergo rapid Fas/FasL-mediated apoptosis. J. Exp. Med. 185, 1837±1849.

cytokine memory of t helper lymphocytes

181

Zhao, K., Wang, W., Rando, O. J., Xue, Y., Swiderek, K., Kuo, A., and Crabtree, G. R. (1998). Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 95, 625±636. Zheng, W., and Flavell, R. A. (1997). The transcription factor GATA-3 is necessary and suf®cient for Th2 cytokine gene expression in CD4 T cells. Cell 89, 587±596. Zhu, J., Huang, H., Guo, L., Stonehouse, T., Watson, C. J., Hu-Li, J., and Paul, W. E. (2000). Transient inhibition of interleukin 4 signaling by T cell receptor ligation. J. Exp. Med. 192, 1125±1134. Zingoni, A., Soto, H., Hedrick, J. A., Stoppacciaro, A., Storlazzi, C. T., Sinigaglia, F., D'Ambrosio, D., O'Garra, A., Robinson, D., Rocchi, M., Santoni, A., Zlotnik, A., and Napolitano, M. (1998). The chemokine receptor CCR8 is preferentially expressed in Th2 but not Th1 cells. J. Immunol. 161, 547±551.

This Page Intentionally Left Blank

ADVANCES IN IMMUNOLOGY, VOL. 80

Ig Gene Hypermutation: A Mechanism Is Due JEAN-CLAUDE WEILL, BARBARA BERTOCCI, AHMAD FAILI, SAID AOUFOUCHI, STE¨PHANE FREY, ANNIE DE SMET, SE¨BASTIEN STORCK, AURIEL DAHAN, FRE¨DE¨RIC DELBOS, SANDRA WELLER, ERIC FLATTER, AND CLAUDE-AGNE©S REYNAUD INSERM Unite¨ 373, Faculte¨ de Me¨decine Necker-Enfants Malades, Universite¨ Paris V, 75730 Paris Cedex15, France

I. Introduction Hypermutation of immunoglobulin genes is a unique adaptive process in biology. The encounter of a receptor (the antibody at the surface of a B lymphocyte) with its speci®c ligand (the antigen) triggers a series of cellular and molecular events resulting in the introduction of mutations, mainly base substitutions, within the gene coding for this receptor. The cells having performed the most adapted mutations, i.e., expressing the receptor with the best af®nity for the ligand, are subsequently selected and maintained in the organism for several years, being able to respond ef®ciently to a reencounter with the original ligand. Although the precise molecular mechanism responsible for this gene-speci®c mutagenesis is still unknown, several advances have been made recently concerning possible factors or molecular events involved. In this review, we discuss: (a) the possible participation of some mismatch repair (MMR) components, the mammalian MutS-homologs, (b) the role of Ig gene transcription in the targeting of the process, (c) the occurrence of DNA breaks as speci®c priming events, (d) the involvement of an error-prone DNA polymerase and the emergence of several new candidate enzymes, and (e) a new partner, activation-induced cytidine deaminase, whose molecular contribution is the subject of intense speculation. II. Mismatch Repair Since mismatch repair-de®cient mice have been available, they have been studied by several groups including ours and the different results obtained and models proposed have been discussed in several reviews (Wood, 1998; Reynaud et al, 1999). In summary, what seems to be agreed on by most investigators is that postreplicative mismatch repair is not directly involved in the process, as opposed to the initial proposition of Cascalho et al. (1998), the reduction of mutation observed in chronically stimulated B cells from Peyer's patches being 183 Copyright 2002, Elsevier Science (USA). All rights reserved. 0065-2776/02 $35.00

jean-claude weill et al.

184

most probably due to their selective elimination. This proposition has been supported by the observation of a very high level of instability of general microsatellite DNA sequences in Msh2 = and Pms2 = PNAhigh B cells (> 50%), most likely caused by the high mitotic rate of these cells (every 6±8 h; Frey et al., 1998). Such an alteration in B cell survival has also been observed after immunization in germinal centers from MSH2-de®cient mice (Vora et al., 1999). What came as a surprise is the speci®c pattern of mutation observed in mice inactivated for the MutS homologs. In the Msh2 = genetic background, mutations are largely targeted to the G position of RGYW hot spots (Rogozin and Kolchanov, 1992). Outside these hot spots, mutations are targeted mainly on G/C pairs and very rarely on A/T pairs. This observation has been extended to Msh6 = mice, whereas somatic mutations in Msh3 = mice appear normal, both qualitatively and quantitatively (Wiesendanger et al., 2000). In summary, mismatch repair-de®cient mice have shown a normal frequency of mutations in primary immune responses, probably caused by the intense selection exerted by the immunizing antigen, and, in contrast, a reduced mutation frequency in chronic responses, most likely related to a diminished survival of memory B cells. Both types of response nevertheless display a modi®ed pattern of mutations, with an increased focusing on a few hot spot positions and a strong bias for G/C pairs in both Msh2 = and Msh6 = mice (Table I). In order to explain this surprising effect only observed in mice de®cient for the MutS homologs (MSH2 and MSH6) and not the MutL homologs (PMS2 and MLH1), several propositions have been formulated (reviewed in Reynaud et al., 1999). We favor a model in which the mutator enters at ®rst into DNA preferentially on G/C pairs. Displacement of one DNA strand for a short distance (see below) would then occur, the mutase being able to ®ll in the gap in an error-prone TABLE I Incidence of MMR Deficiency on Hypermutation Quantitative alterations Mutation frequency in Genetic background

Qualitative alterations

Primary responses

Chronic responses

Hot spot targeting

G/C bias

Msh2

=

Normal

Reduced

Yes

Yes

Msh6 Msh3 Pms2

=

Normal Normal Normal

Reduced Normal Reduced

Yes No No

Yes No No

= =

References Phung et al. (1998), Frey et al. (1998), Rada et al. (1998), Jacobs et al. (1998), Bertocci et al. (1998) Wiesendanger et al. (2000) Wiesendanger et al. (2000) Winter et al. (1998), Frey et al. (1998) Bertocci et al. (1998)

ig gene hypermutation: a mechanism is due

185

fashion on all four bases. Ligation would terminate the process and the error introduced by the mutase would be ®xed by replication. The role of the MSH2/ MSH6 complex would be to assist the strand displacement event and to allow the ¯ap structure to be cut by an endonuclease, likely candidates for the latter event being XP-F and ERCC1 endonucleases involved in nucleotide excision repair (Fig. 1). In the absence of the MSH2/MSH6 components, the mutation would be centered mainly on the G/C pairs. This proposition is supported by the observation that MSH2 and MSH3 can bind to a branched structure containing 30 nonhomologous ends allowing their removal during genetic recombination in yeast (Sugawara et al., 1997; Evans et al., 2000). Our model is close to the one formulated by Rada et al. (1998) but it does not involve, as in their proposition, a new as yet unknown mismatch repair-driven mutagenic function in the MutS-dependent step of the reaction. Such a two-step model is validated by several situations where only the ®rst part of the mechanism involving single-base modi®cations targeted predominantly at G/C pairs takes place, as in the Burkitt lymphoma cell lines that undergo the hypermutation process in culture (DeneÂpoux et al., 1997; Sale and Neuberger, 1998) or in lower species such as Xenopus (Wilson et al., 1992). The introduction of a dominant negative mutant of MSH2 in mouse B cells undergoing hypermutation that would impair mismatch repair without affecting DNA binding should allow us to test this proposition (Studamire et al., 1999).

Fig. 1. A modi®ed two-step model for hypermutation. See text for details.

186

jean-claude weill et al.

As a striking analogy, it has been proposed that MSH2 could play a similar role during another molecular event taking place in germinal center B cells: switch recombination (SR) (Kenter, 1999). This proposition was based on the ®nding that the entry points of switch recombination were localized on Rogozin motifs. In the absence of MSH2, SR could take place at these breakpoints, whereas, in the normal situation, it could occur very often at internal microhomology sites between the two S regions involved (Ehrenstein and Neuberger, 1999). Nevertheless, there are some caveats to this proposition. At ®rst, the concentration of breakpoints on Rogozin motifs was mainly found in the Sm region and not in the Sg3 or Sa regions. Moreover, the drastic reduction in SR observed in the Msh2 = background was also found in MLH1- and PMS2de®cient B cells. This was observed in an in vitro assay which showed that the proliferation of B cells was not altered (Schrader et al., 1999). Despite some analogies, it is therefore dif®cult at this stage to conclude that the entry points of the two processes, hypermutation and switch recombination, are identical and that the mismatch repair complex plays a similar role in both of them. III. The ``Nickase'' In our simpli®ed model as well as in previously proposed models (Brenner and Milstein, 1966), the process must start with a single-strand or a doublestrand nick into the V gene region. In an original approach, Neuberger and colleagues have addressed this precise question in the constitutively hypermutating Burkitt cell line Ramos by expressing the enzyme terminal deoxynucleotidyl transferase (Tdt), which can add nucleotides to 30 ends of double- or single-strand DNA (Sale and Neuberger, 1998). They have observed such additions in the rearranged mutating V gene and noted that 2/3 of the insertions occurred within or abutted to a Rogozin hot spot. Moreover, in most of the cases, Tdt had inserted nucleotides in V gene sites where a deletion or duplication had occurred. When looking at transgenic mice for Tdt under the control of the N-myc promoter, which show a high expression of Tdt in Peyer's patch centroblasts, we have found rare cases of nucleotide insertions, all of which were embedded in a sequence deletion event (our unpublished data, in collaboration with N. Doyen). Deletions and duplications are part of the hypermutation process, but, in this speci®c setting, addition of nucleotides by Tdt could occur on nicks and thereafter induce strand displacements leading to DNA modi®cations at the site of insertion. Such DNA displacements are generated by Tdt in vitro on a primer-template DNA substrate (Kunkel et al., 1986). Alternatively, structures with large gaps may represent more accessible targets for Tdt. These data suggest that nicks are present during hypermutation.

ig gene hypermutation: a mechanism is due

187

More recently, two groups, using LM-PCR, have reported double-strand breaks (DSB) in V genes undergoing hypermutation (Papavasiliou and Schatz, 2000; Bross et al., 2000). These DSBs could be mainly detected on the 50 of the breaks, the 30 ends being probably less accessible. The group of D. Schatz shows that these DSBs are essentially present in the G2/M phase of the cell cycle and suggests that an error-prone DSB repair implying the sister chromatid may be the molecular basis of the process. Such a model is rendered attractive by the fact that a similar mode of mutagenesis has been described in stress-induced quiescent Escherichia coli cells (reviewed in Bridges, 2001). In the latter model, pol IV seems to be responsible for introducing the mutations. This makes a strong case for its eucaryotic equivalent (DinB1) if a G2/M DSB repair is effectively con®rmed as the basis of Ig gene hypermutation. Jacobs and colleagues (Bross et al., 2000), on the other hand, propose that these DSBs occur during the process of transcription and are thereafter repaired by an error-prone nonhomologous end-joining pathway. Both of these attractive propositions rely on the presence of DSBs in the domain covered by the hypermutation process. Another interpretation could be that these DSBs arise as the polymerase involved in semiconservative replication reaches nicks generated during hypermutation and left unrepaired before entering the S phase. This could explain why these DSBs are essentially found in the G2/M phase and why only one side of these breaks is accessible to the LM-PCR primer ligation. If replication proceeds 50 to 30 along the Ig locus, leading strand synthesis reaching a single-strand nick would generate a bluntended DNA, whereas nicks occurring on the DNA strand subjected to lagging strand synthesis would be, on the other hand, less likely to be converted into ligatable ends before the replication fork collapses, in agreement with the quantitative difference in detection of 50 versus 30 DSB ends. Moreover, whether it is homologous recombination or nonhomologous endjoining events that are responsible for the mechanism, one should keep in mind that classical partners involved in either process, such as RAD54 for the ®rst one and DNA-PKcs for the latter, have been shown to be dispensable, their inactivation having no incidence on hypermutation (Jacobs et al., 1998; Bemark et al., 2000b). IV. The Role of Transcription Several reports have shown that transcription and hypermutation are quantitatively correlated starting with the earlier observation that the hypermutation domain at the heavy chain locus extends over 1 kb downstream from the Ig promoter (Lebecque and Gearhart, 1990). From these observations, many different models have been proposed in which the transcription complex, as it progresses through the V gene, induces an error-prone repair process. In

188

jean-claude weill et al.

such propositions, the frequency of mutation is directly correlated to the rate of transcription elongation (Betz et al., 1994; Peters and Storb, 1996; Goyenechea et al., 1997; Fukita et al., 1998). We have designed an Ig transgene in which a transcription stop is inserted in the leader intron and have observed that transcription elongation can be reduced up to 100-fold without affecting the frequency of mutation (Reynaud et al., 2001). These results therefore do not support a direct mechanistic link between transcription elongation and hypermutation, but rather favor an accessibility role for transcription whereby loading to the promoter and enhancer regions with transcription factors is suf®cient to open speci®cally the V domain and give access to the mutator complex. These data do not contradict previous results taking into account that transcription rate may be proportional to the amount of transcription factors and, therefore, also to the level of accessibility. Moreover, one cannot exclude from our results that a minimal amount of transcription may play a role in the process. In fact, we have observed that the targeting of hot spots on the V transgene was different in the copy containing the transcription stop when compared to the control copy, which suggests that the transcript per se may participate in the opening of the DNA double helix as it becomes accessible to the nickase/mutase complex. V. Hypermutation Is Generated by a DNA Polymerase How are mutations introduced during hypermutation? Since the proposition that Ig gene could mutate either to generate the repertoire or to mature the af®nity against a speci®c antigen, models involving almost every possible molecular mechanism have been proposed to explain this phenomenon. Whereas the analysis of selected Ig sequences showed mainly base substitutions, noncoding Ig sequences displayed deletions and duplications of DNA segments (Wu and Kaartinen, 1995; Goossens et al., 1998; Wilson et al., 1998; Levy et al., 1998), which evoked the signature of a DNA polymerase and argued against base-modifying enzymes or chemical DNA modi®cation reactions as the basis of the process. We wanted to ®nd out by another approach whether it was a polymerase that was involved and whether proofreading and/or MMR was still active during the process. To this end, we analyzed the behavior of simple repetitive sequences during hypermutation. DNA polymerases slip when they copy simple repetitive sequences made of tandem repeats of one or a small number of bases that are scattered as microsatellites in the genome of eukaryotes (Streisinger et al., 1966). The slippage that can result from addition or deletion of repeat units is corrected for small repeat numbers (<8) by the proofreading activity of DNA polymerases, but most of the alterations occurring on longer runs are repaired by postreplicative mismatch repair processes (Tran et al., 1997), as shown by instability of microsatellites occurring in yeast

ig gene hypermutation: a mechanism is due

189

and mammalian cells in which mismatch repair is inactive (Fishel et al., 1993; Leach et al., 1993; Strand et al., 1993). Such repeats are naturally present in noncoding sequences at the Ig locus and one can follow their behavior in memory B cells or in cells undergoing hypermutation (Fig. 2). Nevertheless, they are usually short, containing on average 6 bases, and we have designed an Ig transgene in which a mono- or a dinucleotide run, displaying 18 mono- and 11 direpeats respectively, were inserted in the V region. Frameshifts and mutations occurring in the transgene were analyzed in normal mice and in mice with a mismatch repair-de®cient genetic background (Msh2 = or Pms2 = ). The results can be summarized as follows (Bertocci et al., 1998; Frey et al., 1998): (1) Natural microsatellites present in the genome outside Ig loci are not altered in Peyer's patch PNAhigh B cells. (2) A high frequency of alterations, including expansions, contractions, and transient misalignments, are observed at short natural repeats of either 6 Gs in the JH intron ¯anking rearranged VH genes from human memory B cells (Fig. 2a). (3) In contrast, at the 11 GT or 18 T repeats in the mutated Ig transgene V region, few frameshifts (and no transient misalignments) are found compared to the number of base substitutions accumulated on the transgene (1 frameshift for 70 mutations). Some G are mutated within the (GT)11 repeat, but almost no mutations are observed within the (T)18 stretch (Figs. 2b and 2c). (4) In a MMR-de®cient background, the level of mutations drops three- to four fold while the frequency of frameshifts on the repeats remains the same. (5) A higher ratio of deletions and insertions over base substitutions are found in either transgenic or endogenous Ig loci of MMR-de®cient mice (1 event per 20 base substitutions, compared to 1 per 100 in mismatch repair pro®cient mice). We propose from these experiments that, during the hypermutation process, the following occur: (1) The general MMR is not repressed in cells undergoing hypermutation. (2) A DNA polymerase without proofreading generates frameshifts and transient misalignments, as observed on small natural repeats. (3) The low level of frameshifts observed on longer mono- or dinucleotides suggests that this polymerase proceeds by short segments, thus never extending through a 22- or 18-bp run, even though the (GT)11 repeat, but not the (T)18 repeat, provides entry points for the enzyme. (4) The frequency of frameshift per mutation on the microsatellite sequence inserted in the V transgene could increase on mismatch repair de®cient background if hypermutation occurs outside replication, these frameshifts being then generated by the absence of postreplicative MMR.

Fig. 2. Behavior of the mutase on short mono- and dinucleotide repeats. (a) On a natural (G)6 repeat located in the human JH Cm intron, various types of events are observed: targeting of Rogozin hot spots (consensus RGYW sequence on either DNA strand, with R for purine, Y for pyrimidine, and W for A or T), insertion of one or several bases, deletions, and transient misalignment mutations (marked with an asterisk), always on the 50 side of the repeat. (b) On a stretch of 11 GT inserted in an Ig transgene, few frameshifts (deletion/insertion events) are observed on the repeat (monitored by singlecell PCR, Bertocci et al., 1998) compared to the frequency of mutations on the transgene. No case of transient misalignment is found despite a strong targeting of the two ¯anking Rogozin hot spots. The repeat carries mutations targeted to G nucleotides. (c) On a (T)18 microsatellite inserted in an Ig transgene, few frameshifts, no transient misalignments, and almost no mutations are observed in the repeat, while targeting of the two ¯anking hot spots occurs

ig gene hypermutation: a mechanism is due

191

(5) The increase in insertion and duplication events on Ig sequences in mismatch repair-de®cient mice suggests an abortive role for the mismatch repair components on the most prominent DNA alterations. In conclusion, these experiments suggested that hypermutation is generated by a DNA polymerase with low ®delity, acting on short DNA segments outside semiconservative replication. Some MMR components (MSH2/MSH6) would still be present at this stage in order to permit the progression of the mutase, the whole mismatch repair complex being able to abort part of the large deletions and duplications created during the process. VI. Error-Prone DNA Polymerases V gene hypermutation probably involves the copying of a normal DNA template with a high rate of base misincorporations. This implies a DNA polymerase that does not discriminate base pairs stringently and that can also elongate beyond a mispair. Classical DNA polymerases involved in global semiconservative DNA replication, which on the average incorporate a wrong base every 10,000 events (10 3 to 10 5 ;a , d, or "; Kunkel, 1992), could theoretically be in charge of the hypermutation process, but this would necessarily imply the local inhibition of proofreading and mismatch repair activities targeted at the replicating locus undergoing hypermutation. For this reason, an obvious analogy comes from the prokaryotic world in which error-prone DNA synthesis has been demonstrated in stress-induced mutations (reviewed in Radman, 1999; Friedberg and Gerlach, 1999; Woodgate, 1999; Goodman and Tippin, 2000a; Bridges, 2001). During SOS repair in E. coli, which is usually triggered by a block of the replication fork in front of a lesion, a cascade of molecular events may lead to the error-prone bypass of the lesion by speci®c DNA polymerases, pol II, pol IV, and polV (UmuD0 C2). Pol V will do so by introducing mutations, for example, a G in front of the 30 Tof 6 ±4 T±T photoproduct, thus creating a T±C transition. Another recently discovered polymerase in E. coli, DinB (Pol IV), will bypass lesions with ®delity but it will easily produce frameshifts that may liberate replication forks that are blocked at misaligned primer-template structures on repetitive nucleotide runs (Tang et al., 2000). Moreover DinB may also play a role in adaptive mutation taking place in stress-induced quiescent E. coli cells, thereby involving an error-prone homologous recombination event (Bridges, 2001). DNA polymerases phylogenetically related to pol IV and pol V have been isolated in yeast: Rad30 and Rev1 (Table II). Rev1, which is a cytidyl transferase, acts in concert with the Rev3±Rev7 complex (pol z) in order to add C in front of abasic sites and to progress through the lesion, the whole complex being also able to carry mutagenic replication of UV-induced lesions (reviewed in Lawrence and Maher, 2001). On the other hand, yeast Rad30 catalyzes error-free translesion synthesis on thymidine dimers (Washington et al., 1999). Very recently, homologs of E. coli DinB

192

jean-claude weill et al.

(pol k), of yeast pol z, and of yeast Rad30 (Rad30A, pol Z; and Rad30B, pol i) have been found in humans and their functions have been investigated. Pol Z is a very nice example of what a translesion bypass polymerase in higher eukaryotes could be. Effectively, pol Z can bypass a cys±sin T±T dimer by adding two As in front of the Ts, allowing the replication fork to proceed error-free across this lesion (Johnson et al., 1999b; Masutani et al., 2000). Accordingly, patients carrying a null mutation of pol Z (XP-V syndrome) present UV-induced cancers, which shows that, when this enzyme is absent, another one may take its place and introduce mutations in the genome (Johnson et al., 1999a; Masutani et al., 1999). What could seem paradoxical is that pol Z is, in fact, very unfaithful when it copies normal DNA (10 2 ), making on average 25 times more mistakes than the least stringent of all previously known polymerases, pol b (Matsuda et al., 2000). A similar observation has been made with pol i, which can insert an A in front of the 30 T of 6±4 T±T photoproduct with a 30-fold better ef®ciency than opposite an undamaged T, while it cannot bypass a cis±syn T±T dimer (McDonald et al., 1999; Johnson et al., 2000b). All these results are still preliminary and, for some of them, quite controversial (Tissier et al., 2000). Nevertheless, it makes sense for most authors that a DNA polymerase that can incorporate a correct base in front of the distortion provided by a T±T dimer will not be as stringent on normal DNA as the polymerases involved in semiconservative replication. As for the regulated hypermutation process, pol Z or any new DNA polymerase with low ®delity (i or k) would be a good candidate to enter the V gene and copy normal DNA in an error-prone fashion. For pol Z, the analysis of Ig sequences from memory B cells of XPV patients has shown a normal frequency of Ig gene mutation albeit with a de®cit at A / T bases. Whether the other translesion bypass DNA polymerases known so far, pol k, pol i, or pol z, are involved in Ig hypermutation is under intense investigation in several laboratories (see note added in proof ). Pol z itself represents a slightly different case. First, it is related to the family of replicative polymerases (a, d, and also ") and not to the DinB/UmuC/Rad30 family (Table II). Second, whereas it is involved in error-prone bypass of UV damages in yeast, it is not particularly error-prone when copying normal DNA (Johnson et al., 2000b). Finally, the human protein is about twice the size of the yeast homolog (353 vs 173 ka), suggesting that it might be involved in other types of DNA transactions. Accordingly, inactivation of this gene results in early embryonic death and prevents the in vitro culture of any cell type (Bemark et al., 2000b; Wittschieben et al., 2000; Esposito et al., 2000a). Recently, it has been reported that partial inhibition of Rev3 expression by an antisense transgene resulted in a diminution of somatic mutation, particularly on hot spots selected during an immune response (Diaz et al., 2001). Such data have been

ig gene hypermutation: a mechanism is due

193

TABLE II An Immunologist's Survey of Mammalian DNA Polymerases and Nucleotidyl Transferases Type A family (includes E. coli pol I and pol g) Pol theta (W) Unknown function (identified by sequence homology) Homolog of Drosophila mus308 gene: proposed to be involved in interstrand crosslinks repair (Sharief et al., 1999) Type B family (includes pol a, d, and e) Pol zeta (z) Composed of Rev7 and Rev3, its catalytic subunit Rev3 k.o.: early embryonic lethality (Day 9.5) Unknown function in mammals In yeast: nonessential enzyme, involved in error-prone translesion synthesis, capacity to extend a mispair, reasonable fidelity on undamaged DNA (10-4±10-5) DinB/UmuC/Rad30 family (includes E. coli pol IV and pol V) Pol eta (Z) Skin cancer suceptibility XP-V gene in humans (RAD30A) Error-free translesion synthesis of UV-induced DNA damages Error-prone on undamaged DNA (1 mistake every 18±380 bases) Pol kappa (k) Homolog of E. coli polIV (stress-induced DNA polymerase) (Ogi et al., 1999; (DinB1) Gerlach et al., 1999) Unknown function in mammals (no equivalent in Saccharomyces cerevisiae) High expression in testis (also expressed in ovary, prostate, spleen) Conflicting data on its capacity to bypass specific lesions (abasic sites, AAFadducts) and on its fidelity on undamaged DNA (Ohashi et al., 2000b; Johnson et al., 2000a) Generates frameshifts on nonreiterated nucleotides (Ohashi et al., 2000a) Pol iota (i) Unknown function (translesion bypass?) (RAD30B) Error-prone on undamaged DNA, especially in front of thymidines Controversial data on fidelity of lesion bypass Possesses dRP±lyase activity (role in base excision repair?) (Bebenek et al., 2001) Rev1 Deoxycytidyl transferase (dCTP) activity Incorporates dCTP in front of an abasic site Polymerase X family Pol beta (b) No proofreading activity Involved in base excision repair K.o., neonatal lethality due to abnormal neurogenesis (Sugo et al., 2000) Normal Ig hypermutation after adoptive transfer of pol b k.o. hematopoietic cells (Esposito et al., 2000b) Ubiquitous expression, strong in testis Tdt Template-independent polymerization (biased toward G/C) Involved in N additions at TCR and Ig V±D±J junctions (CDR3 diversi®cation) Expressed only in tissues where V±D±J rearrangement occurs (thymus, bone marrow) Pol lambda (l) Unknown function No proofreading activity Ubiquitous, with high expression in testis and fetal liver Pol mu (m) No proofreading activity, error-prone on undamaged DNA Some terminal transferase activity Ubiquitous, with higher expression in lymphoid tissues (highest in thymus, germinal centers, and Burkitt lymphomas) Note. See other references in text.

194

jean-claude weill et al.

interpreted as strong support for a direct implication of Rev3 in the generation of mutations. VII. More DNA Polymerases Data obtained on microsatellites at normal and transgenic Ig loci suggested the involvement in the hypermutation process of an error-prone polymerase acting on small patches, therefore pointing toward a pol b-like polymerase. We searched databases for new DNA polymerases of this family, which also includes Tdt, a lymphoid-speci®c polymerase that contributes to the generation of diversity of Ig and TCR molecules by adding untemplated bases during V±D±J recombination (Table II). Two new genes were found, pol l and pol m (Aoufouchi et al., 2000) These two enzymes were also identi®ed in a parallel investigation by the group of L. Blanco (Dominguez et al., 2000; Garcia-Diaz et al., 2000). Pol m was cloned and sequenced in mouse and human. It shows a genomic organization identical to that of Tdt, with the same exon/intron organization. It contains a pol X domain, diagnostic of this polymerase family, two helix±hairpin±helix (HhH) domains involved in nonspeci®c DNA binding, and a BRCT domain: such a domain has been described in a family of proteins, collectively characterized as DNA-damage responsive cell cycle checkpoint proteins, and is supposed to mediate protein±protein interactions (Bork et al., 1997). At the amino acid level, pol m shows a homology of 60% with Tdt, with 40% of complete identity. After puri®cation, the enzyme can catalyze a template-dependent polymerization reaction with low processivity. It does not possess exonucleolytic activity, but shows terminal transferase activity. Its expression pattern is lymphoid, with a strong expression in germinal center B cells, but also in thymus. The two Burkitt lymphomas Ramos and BL2 express pol m at a rather high level. In Ramos, hypermutation is constitutive whereas in BL2 it must be triggered by anti-IgM crosslinking in the presence of T cells in order for its rearranged V genes to mutate. Complete functional pol m cDNA represents only 10% of the various forms of mRNAs obtained in the BL2 cell line, all the other forms being nonfunctional splice variants. These splice variants can also be isolated from lymph nodes or thymus. Whether this multitude of variants represents a speci®c mode of regulation remains an open issue. Pol l is closer to pol b, although its domain organization with a BCRT domain is similar to that of pol m and Tdt. Its expression is highest in testis and fetal liver, with a very low expression in peripheral lymphoid tissue, which does not make it an obvious candidate for hypermutation. Neither pol m nor pol l is induced by DNA damage. We have recently obtained knockout (k.o.) mice for pol m and pol l. Both k.o. are viable. Preliminary results indicate that these polymerases are dispensable for hypermutation. Whereas pol l = mice have no obvious phenotype, pol

ig gene hypermutation: a mechanism is due

195

m = animals show a partial depletion of the B lymphoid compartments. The origin of this de®ciency is currently under investigation. VIII. Targeting to Non-Ig Genes The strict speci®city of mutation targeting to the rearranged V gene has recently been addressed, the obvious question being to identify putative genes whose alteration could participate in a multistep lymphomagenesis process. In a ®rst attempt, two oncogenes, c-myc and bcl-6, identi®ed as mutated after translocation into the Ig locus have been studied. The untranslocated bcl-6 gene was found targeted by the hypermutation mechanism in human memory B cells over its promoter-proximal region, whereas the c-myc gene was not altered (Shen et al., 1998; Pasqualucci et al., 1998). The mutation frequency observed is, however, 50- to 100-fold lower than that for Ig genes, and none of these mutations have been reported to induce a deregulation of bcl-6 expresion. Similar studies performed in the mouse failed to detect mutations on bcl-6: a likely explanation could be the lower mutation load observed in mouse memory B cells. The second case described is the Fas gene (CD95), and, considering the role of this gene in promoting apoptosis of germinal center centroblasts, inactivating mutations could indeed be involved in B cell transformation (MuÈschen et al., 2000). It has been proposed that some gene translocation events in the Ig locus might be driven by DNA breaks within the V gene accompanying the mutation process (Goossens et al., 1998). In such a view, loci targeted for mutation, even at a reduced rate, may then represent ``active partners'' promoting translocation to a variety of genomic sites: accordingly, translocation of bcl-6 to many non-Ig loci has been observed (Yoshida et al., 1999; Akasaka et al., 2000). Other genes for which mutations/translocations have been described in B cell lymphomas remain attractive candidates to add to the list of possible bystander targets of hypermutation: bcl-10 (marginal zone B cell lymphomas, follicular lymphomas, diffuse large cell lymphomas), IkBa (Hodgkin disease), or B29 (chronic lymphocytic leukemia), even though the mutations described so far in these genes do not always fall into a promoter-proximal domain (Du et al., 2000; Jungnickel et al., 2000; Thompson et al., 1997). IX. AID The group of Honjo has recently described an enzyme, activation-induced cytidine deaminase (AID), whose inactivation suppresses switch recombination and hypermutation (Muramatsu et al., 2000). This protein shows homology with a cytidine deaminase involved in the site-speci®c editing of the mRNA coding for apolipoprotein B. The possible role of AID has been discussed in

196

jean-claude weill et al.

several reviews (Lieber, 2000; Jacobs and Bross, 2001). Three main possibilities exist for its function: It could be involved in signaling processes upstream of both switch recombination and hypermutation events by editing a key molecule in the differentiation pathway of germinal center B cells. In favor of this proposition would be the observation of enlarged germinal centers in both knockout mice and patients with type II hyper-IgM syndrome, which corresponds to the inactivation of both AID alleles (Revy et al., 2000). Alternatively, it could participate actively in both mechanisms, either by editing an mRNA coding for a key factor or by direct alteration of the target V gene, at the level of the RNA transcript (whose presence is strictly required for switch recombination) or even at the DNA level. In this latter case, the editing activity would have a rather loose speci®city of sequence recognition, allowing the targeting of the various switch DNA regions, as well as the whole repertoire of rearranged VH, Vk, and Vl genes and of their ¯anking sequences, even after extensive diversi®cation by previous rounds of hypermutation. Moreover, since both mechanisms, switch and hypermutation, can occur independently, in particular in humans for which mutated IgM‡ B cells represent a large fraction of the memory cell pool, this would require a differential triggering of both mechanisms by the same enzymatic activity. X. Conclusion In conclusion, the ®eld of hypermutation is crowded with experiments and models but a precise molecular description of the process is due. It looks as if we should be ready for surprises. The ®rst molecule that seems to be a major player in the process has been discovered after a cDNA subtraction designed to elucidate switch recombination, and its putative properties generate many more questions than they bring straightforward explanations. The growing number of ``sloppy'' polymerases obviously attracts attention (Goodman and Tippin, 2000b), but each one may have a precise function in order to allow the replication fork to bypass faithfully a speci®c lesion. Although it is a great source of adaptation to be able to mutate DNA in a cell, such activity may have to be used with care, and it remains to be seen whether the immune system has evolved a unique molecular design to do this or whether it has diverted from its function one of the ``sloppy'' enzymes around. Note Added in Proof Several relevant studies have been published since the submission of this chapter. These include: Diaz, M., Verkoczy, L. K., Flajnik, M. F., and Klinman, N. R. (2001). Decreased frequency of somatic hypermutation and impaired af®nity maturation but intact germinal center fornation in mice expressing antisense RNA to DNA polymerase zeta. J. Immunol. 167, 327±335.

ig gene hypermutation: a mechanism is due

197

Frank, E. G., Tissier, A., McDonald, J. P., Rapic-Otrin, V., Zeng, X., Gearhart, P. J., and Woodgate, R. (2001). Altered nucleotide misinsertion ®delity associated with pol iota-dependent replication at the end of a DNA template. EMBO J. 20, 2914±2922. Matsuda, T., Bebenek, K., Masutani, C., Rogozin, I. B., Hanaoka, F., and Kunkel, T. A. (2001). Error rate and speci®city of human and murine DNA polymerase eta. J. Mol. Biol. 312, 335±346. Pasqualucci, L., Neumeister, P., Goossens, T., Nanjangud, G., Chaganti, R. S., Kuppers, R., and Dalla-Favera, R. (2001). Hypermutation of multiple proto-oncogenes in B-cell diffuse large-cell lymphomas. Nature 412, 341±346. Poltoratsky, V., Woo, C. J., Tippin, B., Martin, A., Goodman, M. F., and Scharff, M. D. (2001). Expression of error-prone polymerases in BI.2 cells activated for Ig somatic hypermutation. Proc. Natl. Acad. Sci. USA 98, 7976±7981. Rogozin, I. B., Pavlov, Y. I., Bebenek, K., Matsuda, T., and Kunkel, T. A. (2001). Somatic mutation hotspots correlate with DNA polymerase eta error spectrum. Nat. Immunol. 2, 530±536. Sale, J. E., Calandrini, D. M., Takata, M., Takeda, S., and Neuberger, M. S. (2001). Ablation of XRCC2/3 transforms immunoglobulin V gene conversion into somatic hypermutation. Nature 412, 921±926. Zan, H., Komori, A., Li, Z., Cerutti, A., Schaffer, A., Flajnik, M. F., Diaz, M., and Casali, P. (2001). The translesion DNA polymerase zeta plays a major role in lg and bcl-6 somatic hypermutation. Immunity 14, 643±653. Zeng, X., Winter, D. B., Kasmer, C., Kracmer, K. H., Lehmann, A. R., and Gearhart, P. J. (2001). DNA polymerase eta is an A-T mutator in somatic hypermutation of immunoglobulin variable genes Nat. Immunol. 2, 537±541.

References Akasaka, H., Akasaka, T., Kurata, M., Ueda, C., Shimizu, A., Uchiyama, T., and Ohno, H. (2000). Molecular anatomy of BCL6 translocations revealed by long-distance polymerase chain reactionbased assays. Cancer Res. 60, 2335±2341. Aoufouchi, S., Flatter, E., Dahan, A., Faili, A., Bertocci, B., Storck, S., Delbos, F., Cocea, L., Gupta, N., Weill, J. C., and Reynaud, C. A. (2000). Two novel human and mouse DNA polymerases of the polX family. Nucleic Acids Res. 28, 3684±3693. Bebenek, K., Tissier, A., Frank, E. G., McDonald, J. P., Prasad, R., Wilson, S. H., Woodgate, R., and Kunkel, T. A. (2001). 50 -Deoxyribose phosphate lyase activity of human DNA polymerase  in vitro. Science 291, 2156±2159. Bemark, M., Khamlichi, A. A., Davies, S. L., and Neuberger, M. S. (2000a). Disruption of mouse polymerase z (Rev3) leads to embryonic lethality and impairs blastocyst development in vitro. Curr. Biol. 10, 1213±1216. Bemark, M., Sale, J. E., Kim, H. J., Berek, C., Cosgrove, R. A., and Neuberger, M. S. (2000b). Somatic hypermutation in the absence of DNA-PKcs or recombination-activating gene (RAG). J. Exp. Med. 192, 1509±1514. Bertocci, B., Quint, L., Delbos, F., Garcia, C., Reynaud, C. A., and Weill, J. C. (1998). Probing immunoglobulin gene hypermutation with microsatellites suggests a nonreplicative short patch DNA synthesis process. Immunity 9, 257±265. Betz, A. G., Milstein, C., Gonzalez-Fernandez, A., Pannell, R., Larson, T., and Neuberger, M. (1994). Elements regulating somatic hypermutation of an immunoglobulin k gene: Critical role for the intron enhancer/matrix attachment region. Cell 77, 239±248. Bork, P., Hofmann, K., Bucher, P., Neuwald, A. F., Altschul, S. F., and Koonin, E. V. (1997). A superfamily of conserved domains in DNA damage-responsive cell cycle checkpoint proteins. FASEB J. 11, 68±76. Brenner, S., and Milstein, C. (1966). Origin of antibody variation. Nature 211, 242±243.

198

jean-claude weill et al.

Bridges, B. A. (2001). Hypermutation in bacteria and other cellular systems. Philos. Trans. R. Soc. London B 356, 29±39. Bross, L., Fukita, Y., McBlane, F., DeÂmollieÁre, C., Rajewsky, K., and Jacobs, H. (2000). DNA double-strand breaks in immunoglobulin genes undergoing somatic hypermutation. Immunity 13, 589±597. Cascalho, M., Wong, J., Steinberg, C., and Wabl, M. (1998). Mismatch repair co-opted by hypermutation. Science 279, 1207±1210. DeneÂpoux, S., Razanajaona, D., Blanchard, D. Meffre, G., Capra, J. D., Banchereau, K., and Lebecque, S. (1997). Induction of somatic mutation in a human B cell line in vitro. Immunity 6, 35±46. Diaz, M., Flajnik, M. F., and Klinman, N. (2001). Evolution and the molecular basis of somatic hypermutation of antigen receptor genes. Philos. Trans. R. Soc. London B 356, 67±72. Dominguez, O., Ruiz, J. F., Lain de Lera, T., Garcia-Diaz, M., Gonzalez, M. A., Kirchhoff, T., Martinez-A, C., Bernad, A., and Blanco, L. (2000). DNA polymerase mu (Pol m) homologous to TdT, could act as a DNA mutator in eukaryotic cells. EMBO J. 19, 1731±1742. Du, M. Q., Peng, H., Liu, Hongxiang, Hamoudi, R. A., Diss, T. C., Willis, T. G., Ye, H., Dogan, A., Wotherspoon, A. C., Dyer, M. J. S., and Isaacson, P. G. (2000). BCL10 gene mutation in lymphoma. Blood 95, 3885±3890. Ehrenstein, M. R., and Neuberger, M. S. (1999). De®ciency in Msh2 affects the ef®ciency and local sequence speci®city of immunoglobulin class-switch recombination: Parallels with somatic hypermutation. EMBO J. 18, 3484±3490. Esposito, G., Godin, I., Klein, U., Yaspo, M. L., Cumano, A., and Rajewsky, K. (2000a). Disruption of the Rev3I-encoded catalytic subunit of polymerase z in mice results in early embryonic lethality. Curr. Biol. 10, 1221±1224. Esposito, G., Texido, G., Betz, U. A. K., Gu, H., MuÈller, W., Klein, U., and Rajewsky, K. (2000b). Mice reconstituted with DNA polymerase b-de®cient fetal liver cells are able to mount a T cell-dependent immune response and mutate their Ig genes normally. Proc. Natl. Acad. Sci. USA 97, 1166±1171. Evans, E., Sugawara, N., Haber, J. E., and Alani, E. (2000). The Saccharomyces cerevisiae Msh2 mismatch repair protein localizes to recombination intermediates in vivo. Mol. Cell. 5, 789±799. Fishel, R., Lescoe, M. K., Rao, M. R. S., Copeland, N. G., Jenkins, N. A., Gaber, J., Kane, M., and Kolodner, R. (1993). The human mutator gene homolog Msh2 and its association with hereditary nonpolyposis colon cancer. Cell 75, 1027±1038. Frey, S., Bertocci, B., Delbos, F., Quint, L., Weill, J. C., and Reynaud, C. A. (1998). Mismatch repair de®ciency interferes with the accumulation of mutations in chronically stimulated B cells and not with the hypermutation process. Immunity 9, 127±134. Friedberg, E. C., and Gerlach, V. L. (1999). Novel DNA polymerases offer clues to the molecular basis of mutagenesis. Cell 98, 413±416. Fukita, Y., Jacobs, H., and Rajewsky, K. (1998). Somatic hypermutation in the heavy chain locus correlates with transcription. Immunity 9, 105±114. Garcia-Diaz, M., Dominguez, O., Lopez-Fernandez, L. A., Teresa de Lera, L., Saniger, M. L., Ruiz, J. F., Parraga, M., Garcia-Ortiz, M. J., Kirchhoff, T., del Mazo, J., Bernad, A., and Blanco, L. (2000). DNA polymerase lambda (Pol l), a novel eukaryotic DNA polymerase with a potential role in meiosis. J. Mol. Biol. 301, 851±867. Gerlach, V. L., Aravind, L., Gotway, G., Schultz, R. A., Koonin, E. V., and Friedberg, E. C. (1999). Human and mouse homologs of Escherichia coli DinB (DNA polymerase IV), members of the UmuC/DinB superfamily. Proc. Natl. Acad. Sci. USA 96, 11922±11927. Goodman, M. F., and Tippin, B. (2000a). The expanding polymerase universe. Nat. Rev. Mol. Cell. Biol. 1, 101±109.

ig gene hypermutation: a mechanism is due

199

Goodman, M. F., and Tippin, B. (2000b). Sloppier copier DNA polymerases involved in genome repair. Curr. Opin. Genet. Dev. 10, 162±168. Goossens, T., Klein, U., and KuÈppers, R. (1998). Frequent occurrence of deletions and duplications during somatic hypermutation: Implications for oncogene translocations and heavy chain disease. Proc. Natl. Acad. Sci. USA 95, 2463±2468. Goyenechea, B., Klix, N., Yelamos, J., Williams, G. T., Riddell, A. Neuberger, M. S., and Milstein, C. (1997). Cells strongly expressing Igk transgenes show clonal recruitment of hypermutation: A role for both MAR and the enhancers. EMBO J. 16, 3987±3994. Jacobs, H., and Bross, L. (2001). Towards an understanding of somatic hypermutation. Curr. Opin. Immunol. 13, 208±218. Jacobs, H., Fukita, Y., van der Horst, G. T. J., de Boer, J., Weeda, G., Essers, J., de Wind, N., Engelward, B. P., Samson, L., Verbeek, S., MeÂnissier de Murcia, J., de Murcia, G., te Riele, H., and Rajewsky, K. (1998). Hypermutation of immunoglobulin genes in memory B cells of DNA repair-de®cient mice. J. Exp. Med. 187, 1735±1743. Johnson, R. E., Kondratick, C. M., Prakash, S., and Prakash, L. (1999a). hRAD30 mutations in the variant form of xeroderma pigmentosum. Science 285, 263±265. Johnson, R. E., Prakash, S., and Prakash, L. (1999b). Ef®cient by-pass of a thymine±thymine dimer by yeast DNA polymerase, PolZ. Science 283, 1001±1004. Johnson, R. E., Prakash, S., and Prakash, L. (2000a). The human DINB1 gene encodes the DNA polymerase PolW. Proc. Natl. Acad. Sci. USA 97, 3838±3843. Johnson, R. E., Washington, M. T., Haracska, L., Prakash, S., and Prakash, L. (2000b). Eukaryotic polymerases i and z act sequentially to bypass DNA lesions. Nature 406, 1015±1019. Jungnickel, B., Staratschek-Jox, A., BraÈuninger, A., Spieker, T., Wolf, J., Diehl, V. Hansmann, M. L., Rajewsky, K., and KuÈppers R. (2000). Clonal deleterious mutations in the IkBa gene in the malignant cells in Hodgkin's lymphoma. J. Exp. Med. 191, 395±401. Kenter, A. L. (1999). The liaison of isotype class switch and mismatch repair: An illegitimate affair. J. Exp. Med. 190, 307±310. Kunkel, T. A. (1992). DNA replication ®delity. J. Biol. Chem. 267, 18251±18254. Kunkel, T. A., Gopinathan, K. P., Dube, D. K., Snow, E. T., and Loeb, L. A. (1986). Rearrangements of DNA mediated by terminal transferase. Proc. Natl. Acad. Sci. USA 83, 1867±1871. Lawrence, C. W., and Maher, V. M. (2001). Mutagenesis in eukaryotes dependent on DNA polymerase zeta and Rev1p. Philos. Trans. R. Soc. London B 356, 41±46. Leach, F. S., Nicolaides, N. C., Papadopoulos, N., Liu, B. Jen, J., Parsons, R., Peltomaki, P., Sistonen, P., Aaltonen, L. A., Nystrom-Lahti, M., et al. (1993). Mutations of a mutS homolog in hereditary nonpolyposis colorectal cancer. Cell 75, 1215±1225. Lebecque, S. G., and Gearhart, P. J. (1990). Boundaries of somatic mutation in rearranged immunoglobulin genes: 50 boundary is near the promoter, and 30 boundary is 1kb from V(D)J gene. J. Exp. Med. 172, 1717±1727. Levy, Y., Gupta, N., Le Deist, F., Garcia, C., Fischer, A., Weill, J. C., and Reynaud, C. A. (1998). Defect in IgV gene somatic hypermutation in common variable immunode®ciency syndrome. Proc. Natl. Acad. Sci. USA 95, 13135±13140. Lieber, M. (2000). Antibody diversity: A link between switching and hypermutation. Curr. Biol. 10, 798±800. Masutani, C., Kusumoto, R., Iwai, S., and Hanaoka, F. (2000). Accurate translesion synthesis by human DNA polymerase Z. EMBO J. 19, 3100±3109. Masutani, C., Kusumoto, R., Yamada, A., Dohmae, N., Yokoi, M. Yuasa, M., Araki, M., Iwai, S., Takio, K., and Hanaoka, F. (1999). The human XPV (xeroderma pigmentosum variant) gene encodes human polymerase . Nature 399, 700±704. Matsuda, T., Bebenek, K., Masutani, C., Hanaoka, F., and Kunkel, T. A. (2000). Low ®delity DNA synthesis by human DNA polymerase Z. Nature 404, 1011±1013.

200

jean-claude weill et al.

McDonald, J. P., Rapic-Otrin, V., Epstein, J. A., Broughton, B. C., Wang, X., Lehmann, A. R., Wolgemuth, D. J., and Woodgate, R. (1999). Novel human and mouse homologs of Saccharomyces cerevisiae DNA polymerase Z. Genomics 60, 20±30. Muramatsu, M., Kinoshita, K., Faragasan, S., Yamada, S., Shinkai, Y., and Honjo, T. (2000). Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102, 553±563. MuÈschen, M., Re, D., Jungnickel, B., Diehl, V., Rajewsky, K., and KuÈppers, R. (2000). Somatic mutation of the CD95 gene in human B cells as a side-effect of the germinal center reaction. J. Exp. Med. 192, 1833±1839. Ogi, T., Kato, T., Kato, T., and Ohmori, H. (1999). Mutation enhancement by DINB1, a mammalian homologue of the Escherichia coli mutagenesis protein dinB. Genes Cell 4, 607±618. Ohashi, E., Bebenek, K., Matsuda, T., Feaver, W. J., Gerlach, V. L., Friedberg, E. C., Ohmori, H., and Kunkel, T. A. (2000a). Fidelity and processivity of DNA synthesis by DNA polymerase kappa, the product of the human DINB1 gene. J. Biol. Chem. 275, 39678±39684. Ohashi, E., Ogi, T., Kusumoto, Iwai, S., Masutani, C., Hanaoka, F., and Ohmori, H. (2000b). Errorprone bypass of certain DNA lesions by the human DNA polymerase k. Genes Dev. 14, 1589±1594. Papavasiliou, F. N., and Schatz, D. G. (2000). Cell-cycle-regulated DNA double-strand breaks in somatic hypermutation of immunoglobulin genes. Nature 408, 216±221. Pasqualucci, L., Migliazza, A., Fracchiolla, N., William, C., Neri, A., Baldini, L., Chaganti, R. S., Klein, U., KuÈppers, R., Rajewsky, K., and Dalla-Favera, R. (1998). BCL-6 mutations in normal germinal center B cells: Evidence of somatic hypermutation acting outside Ig loci. Proc. Natl. Acad. Sci. USA 95, 11816±11821. Peters, A., and Storb, U. (1996). Somatic hypermutation of immunoglobulin genes is linked to transcription initiation. Immunity 4, 57±65. Phung, Q. H., Winter, D. B., Granston, A., Tarone, R. E., Bohr, V. A., Fishel, R., and Gearhart, P. J. (1998). Increased hypermutation at G and C nucleotides in immunoglobulin variable genes from mice de®cient in the MSH2 mismatch repair protein. J. Exp. Med. 187, 1745±1751. Rada, C., Ehrenstein, M. R., Neuberger, M. S., and Milstein, C. (1998). Somatic hypermutation in MSH2-de®cient mice is more focused on intrinsic hot spots suggesting hypermutation targeting to be a two stage process. Immunity 9, 135±141. Radman, M. (1999). Enzymes of evolutionary change. Nature 401, 866±869. Revy, P., Muto, T., Levy, Y., Geissmann, F., Plebani, A., Sanal, O., Catalan, N., Forveille, M., Dufourcq-Lagelouse, R., Gennery, A., Tezcan, I., Ersoy, F., Kayserili, H., Ugazio, A. G., Brousse, N., Muramatsu, M., Notarangelo, L. D., Kinoshita, K., Honjo, T., Fischer, A., and Durandy, A. (2000). Activation-induced cytidine deaminase (AID) de®ciency causes the autosomal recessive form of the hyper-IgM syndrome (HIGM2). Cell 102, 565±575. Reynaud, C. A., Bertocci, B., Frey, S., Delbos, F., Quint, L., and Weill, J. C. (1999). Mismatch repair and immunoglobulin gene hypermutation: Did we learn something? Immunol. Today 20, 522±527. Reynaud, C. A., Frey, S., Aoufouchi, S., Faili, A., Bertocci, B., Dahan, A., Flatter, E., Delbos, F., Storck, S., Zober, C., and Weill, J. C. (2001). Transcription, b-like DNA polymerases and hypermutation. Philos. Trans. R. Soc. London B 356, 91±97. Rogozin, I. B., and Kolchanov, N. A. (1992). Somatic hypermutagenesis in immunoglobulin genes. II. In¯uence of neighbouring base sequence on mutagenesis. Biochim. Biophys. Acta 1171, 11±18. Sale, J. E., and Neuberger, M. S. (1998). TdT-accessible breaks are scattered over the immunoglobulin V domain in a constitutively hypermutating B cell line. Immunity 9, 859±869. Sharief, F. S., Vojta, P. J., Ropp, P. A., and Copeland, W. C. (1999). Cloning and chromosomal mapping of the human DNA polymerase y (PolQ), the eighth human DNA polymerase. Genomics 59, 90±96.

ig gene hypermutation: a mechanism is due

201

Schrader, C. E., Edelmann, W., Kucherlapati, R., and Stavnezer, J. (1999). Reduced isotype switching in splenic B cells from mice de®cient in mismatch repair enzymes. J. Exp. Med. 190, 323±330. Shen, H. M., Peters, A., Baron, B., Zhu, X., and Storb, U. (1998). Mutation of the bcl-6 gene in normal B cells by the process of somatic hypermutation of Ig genes. Science 280, 1750±1752. Strand, M., Prolla, T. A., Liskay, R. M., and Petes, T. D. (1993). Destabilization of tracts of simple repetitive DNA in yeast by mutations affecting DNA mismatch repair. Nature 365, 274±276. Streisinger, G., Okada, Y., Emrich, J., Newton, J., Tsugita, A. Terzaghi, E., and Inouye, M. (1966). Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol. 31, 77±84. Studamire, B., Price, G., Sugawara, N., Haber, J. E., and Alani, E. (1999). Separation-of-function mutations in Saccharomyces cerevisiae MSH2 that confer mismatch repair defects but do not affect nonhomologous-tail removal during recombination. Mol. Cell. Biol. 19, 7558±7567. Sugawara, N., PaÃques, F., Colaiacovo, M., and Haber, J. E. (1997). Role of Saccharomyces cerevisiae Msh2 and Msh3 repair proteins in double-strand break-induced recombination. Proc. Natl. Acad. Sci. USA 94, 9214±9219. Sugo, N., Aratani, Y., Nagashima, Y., Kubota, Y., and Koyama, H. (2000). Neonatal lethality with abnormal neurogenesis in mice de®cient in DNA polymerase beta. EMBO J. 19, 1397±1404. Tang, M., Pham, P., Shen, X., Taylor, J. S., O'Donnell, M., Woodgate, R., and Goodman, M. F. (2000). Roles of E. coli DNA polymerases IV and V in lesion-targeted and untargeted SOS mutagenesis. Nature 404, 1014±1018. Thompson, A. A., Talley, J. A., Do, H. N., Kagan, H. L., Kunkel, L., Berenson, J., Cooper, M. D., Saxon, A., and Wall, R. (1997). Aberrations of the B-cell receptor B29 (CD79b) gene in chronic lymphocytic leukemia. Blood 90, 1387±1394. Tissier, A., Frank, E. G., McDonald, J. P., Iwai, S., Hanaoka, F., and Woodgate, R. (2000). Misinsertion and bypass of thymine±thymine dimers by human DNA polymerase i. EMBO J. 19, 5259±5266. Tran, H. T., Keen, J. D., Kricker, M., Resnick, M. A., and Gordenin, D. A. (1997). Hypermutability of homonucleotide runs in mismatch repair and DNA polymerase proofreading yeast mutants. Mol. Cell. Biol. 17, 2859±2865. Vora, K. A., Tumas-Brundage, K. M., Lentz, V. M., Cranston, A., Fishel, R., and Manser, T. (1999). Severe attenuation of the B cell immune response in Msh2-de®cient mice. J. Exp. Med. 189, 471±482. Washington, M. T., Johnson, R. E., Prakash, S., and Prakash, L. (1999). Fidelity and processivity of Saccharomyces cerevisiae DNA polymerase Z. J. Biol. Chem. 274, 36835±36838. Wiesendanger, M., Kneitz, B., Edelmann, W., and Scharff, M. D. (2000). Somatic hypermutation in MutS homologue (MSH)3-, MSH6- and MSH3/MSH6-de®cient mice reveals a role for the MSH2±MSH6 heterodimer in modulating the base substitution pattern. J. Exp. Med. 191, 579±584. Wilson, M., Hsu, E., Marcuz, A., Courtet, M., Du Pasquier, L., and Steinberg, C. (1992). What limits af®nity maturation of antibodies in XenopusÐthe rate of somatic mutation or the ability to select mutants? EMBO J. 11, 4337±4347? Wilson, P. C., de Bouteiller, O., Liu, Y. J., Potter, K., Banchereau, J., Capra, J. D., and Pascual, V. (1998). Somatic hypermutation introduces insertions and deletions into immunoglobulin V genes. J. Exp. Med. 187, 59±70. Winter, D. B., Phung, Q. H., Umar, A., Baker, S. M., Tarone, R. E., Tanaka, K., Liskay, R. M., Kunkel, T. A., Bohr, V. A., and Gearhart, P. J. (1998). Altered spectra of hypermutation in

202

jean-claude weill et al.

antibodies from mice de®cient for the DNA mismatch repair protein PMS2. Proc. Natl. Acad. Sci. USA 95, 6953±6958. Wittschieben, J., Shivji, M. K. K., Lalani, E., Facobs, M. A., Marini, F., Gearhart, P. J., Rosewell, I., Stamp, G., and Wood, R. D. (2000). Disruption of the developmentally regulated Rev3I/gene causes embryonic lethality. Curr. Biol. 10, 1217±1220. Wood, R. D. (1998). DNA repair: Knockouts still mutating after ®rst round. Curr. Biol. 8, 757±760. Woodgate, R. (1999). A plethora of lesion-replicating DNA polymerases. Genes Dev. 13, 2191±2195. Wu, H. Y., and Kaartinen, M. (1995). The somatic hypermutation activity of a follicular lymphoma links to large insertions and deletions of immunoglobulin genes. Scand. J. Immunol. 42, 52±59. Yoshida, S., Kaneita, Y., Aoki, Y., Seto, M., Mori, S., and Moriyama, M. (1999). Identi®cation of heterologous translocation partner genes fused to the BCL6 gene in diffuse large B-cell lymphomas: 50 -RACE and LA-PCR analyses of biopsy samples. Oncogene 18, 7994±7999.

ADVANCES IN IMMUNOLOGY, VOL. 80

Generalization of Single Immunological Experiences by Idiotypically Mediated Clonal Connections HILMAR LEMKE AND HANS LANGE Biochemical Institute of the Medical Faculty of the Christian-Albrechts-University, D-24118 Kiel, Germany

Clonal interactions of B cells by idiotope-speci®c mutual recognition of their antigen receptors with the participation of T cells were assumed to form a web of unknown density, referred to as the idiotypic network. Although these clonal connections were proposed to ful®ll important internal regulatory functions, their biological signi®cance, especially in relation to antigen-induced immune responses, remained a mystery. In view of this, we postulate that the basic function of the idiotypic internal connection between B and T cell antigen receptors is to transform antigen-induced cellular activations, by idiotypic crossreactivity, into the regulation of cell clones with different antigen speci®cities. This process leads not only to the suppression of major clones but also to the activation of minor ones. The latter activating property may allow the generalization of single antigenic experiences, so that the immune system in its entirety bene®ts in its battle against environmental microbes. Such idiotypic clonal interactions are particularly effective in early ontogeny. During a short neonatal imprinting period, maternal immunological knowledge in the form of somatically mutated, high-af®nity IgG antibodies, acquired through a continuous encounter with external antigens, guides the initial ontogenetic development of the immune system and so exerts longlasting transgenerational advantageous effects in the offspring. #2002, Elsevier Science (USA).

I. Introduction A quarter of a century ago, the fascinating discovery that the immune system is able to mount antibody responses not only to external antigens but also to internal antigenic determinants located in the variable regions of immunoglobulin H and L chains (Kunkel et al., 1963; Oudin and Michel, 1963) led to the formulation of the idiotypic network theory (Jerne, 1974). According to this theory, cell-bound and secreted antigen receptors of complementary clones of B lymphocytes may recognize each other as idiotypes and anti-idiotypes and thereby form a web of interacting variable immunoglobulin domains. These multiple interactions were supposed to enable an equilibrium among the involved cell clones leading to the concept that the immune system as such is in a state of balance which, after disturbance by stimulation with external antigens, can be reestablished by idiotypic±anti-idiotypic reactions. It was assumed that 203 Copyright 2002, Elsevier Science (USA). All rights reserved. 0065-2776/02 $35.00

204

hilmar lemke and hans lange

external antigens ``incidentally disturb the dynamic harmony of the system'' (Jerne, 1985). After the ®rst delineation of the idiotypic network as an integrating system of B lymphocytes, manifold idiotypic interactions between B and T cells were discovered which demonstrated an intimate connection and interdependent development of the B and T cell compartments (Janeway et al., 1990; Marcos et al., 1988; Martinez et al., 1988; Rajewsky and Takemori, 1983). Hence, the functional signi®cance and physiological regulatory role of the idiotypic network could be expanded to the entire lymphocyte-mediated adaptive immune system and was regarded as an established fact (Bona et al., 1981; Rajewsky and Takemori, 1983). Indeed, idiotypic interactions are still supposed to play a de®nite although nonde®ned role during the development of infectious, autoimmune, and malignant diseases (Cerny and Hiernaux, 1990; Shoenfeld et al., 1997). Nevertheless, it was possible to inform the scienti®c community with a high impact tailwind that ``Nobody todayÐor 99% of immunologists don't believe the network theory'' (Marshall, 1996; see also Cohn, 1994, p. 39±40). How is it possible that scienti®c facts can become a matter of belief ? An understandable reason may be the fact that, despite all the well-documented idiotypically induced clonal alterations in various experimental and clinical settings, the biological signi®cance of the idiotypic networkÐfor what is it good?Ðremained a mystery. In the present communication, we propose the generalization of single immunological experiences with external antigens as the basic function of idiotypic interactions. This ability can be deduced from immunological reactivities of adult individuals and is particularly obvious in early ontogeny, allowing the transfer of the collective maternal immunological experience to the offspring and thus enabling a maternally mediated immunological education before the ®rst encounter with external antigens. Hence, the immune system is endowed with not only the capacity of self-recognition, i.e., ``to identify and establish the uniqueness of the individual along a life-long ontogeny'' (Coutinho, 1995), but also the postulated function, namely, the ability to generalize and make use of particular single events in a different antigenic context as well as to transfer maternal experience to the next generation, giving the individual, beyond any clonal alteration of unknown relevance, an advantage in the battle with environmental microbes and their antigenic products. Moreover, we would like to stress that, in a general sense, this function of the immune system is analogous to a respective ability of the nervous system. II. Idiotypic Transformation of Single Immunological Experiences in the Adult A. Dichotomy between Binding of Antigen and Expression of Idiotopes The combined variable domains of immunoglobulin heavy and light chains (FV fragment) form the antigen-binding region (paratope) that recognizes

idiotypic network functions

205

antigenic determinants. In addition, the FV fragment itself contains multiple immunogenic epitopes (idiotopes; Jerne, 1985) that may be recognized not only by xenogeneic anti-idiotypic antibodies (Eichmann, 1972), but also by allogeneic (Kunkel et al., 1963; Oudin and Michel, 1963), syngeneic (Rajewsky et al., 1981), and even autologous anti-idiotypic antibodies (Cosenza, 1976; Rodkey et al., 1983). It has been estimated that an antibody FV fragment can express several (Bona et al., 1981; Strickland et al., 1987) or up to 15±20 idiotopes (Novotny et al., 1986), the collection of which characterizes a particular idiotype. Hence, an idiotype may be de®ned either by polyclonal antisera through recognition of different sets of idiotopes or by monoclonal antibodies through recognition of particular single idiotopes. These may be located at variable distances from or may overlap with the paratope, i.e., they may be binding-site related (Brient and Nisonoff, 1970) and nonrelated (Kelus and Gell, 1968). Moreover, the assumption that certain idiotopes may coincide with the paratope led to two conclusions: (a) that anti-idiotypic antibodies recognizing such idiotopes may carry an internal image of the corresponding epitope and (b) that, therefore, the immune system probably contains such internal images of all epitopes of external antigens (Jerne, 1974; Jerne et al., 1982). In addition, this concept led to the proposal that internal image antiidiotypic antibodies could be used as surrogate antigens for therapeutic vaccinations (Nisonoff and Lamoyi, 1981) and, indeed, in numerous investigations (Magliani et al., 1998; MoÈller, 1986; Ruiz et al., 1998; Sacks et al., 1982; Shoenfeld et al., 1997; Stein and SoÈderstroÈm, 1984; Su et al., 1992; Whittum-Hudson et al., 1996), such anti-Id have been successfully employed to induce immunity to infectious diseases or against tumor cell growth. However, even when an anti-Id has almost identical contact sites with the Id as the antigen itself, this mimicry is functional and cannot be understood in the sense of identical threedimensional structures (Fields et al., 1995). This assessment is in line with earlier observations suggesting that an anti-Id can hardly be de®ned as an internal image antibody by in vitro tests (Chen and KoÈhler, 1990), but only by its ability to induce antigen-reactive antibodies in vivo since this therapeutic effectiveness also depends on preexisting conditions in the network (Chen and KoÈhler, 1990; Kohler et al., 1989). Moreover, anti-idiotypic antibodies that are directed at individual idiotopes and therefore certainly do not represent internal image antibodies may nevertheless be able to function as surrogate antigen and induce neutralizing antiviral antibodies (Osterhaus et al., 1989; Yang and Thanavala, 1995). From these considerations the following conclusions can be made: 1. A particular antibody expresses a paratope that is speci®c for a given epitope and, at the same time, the combined VH =VL regions of this antibody contain multiple idiotopes that can be recognized by anti-idiotypic antibodies which

206

hilmar lemke and hans lange

display different antigen speci®cities. Hence, even when an antibody is monospeci®c for an antigen it is multireactive with internal network components. 2. Particular anti-idiotypic antibodies as internal network antigens can be recognized by their ability to induce immunity to external antigens, for instance, against an antigen-carrying microbe, even though they do not perfectly mimic the epitope in its three-dimensional structure. This already exquisitely demonstrates the validity of the idiotypic network theory, at least in the sense that the immune system itself contains the relevant speci®cities and that these can be used to induce immunity to external antigens. B. Physiological Idiotypic Interactions Occur Exclusively within an Individual Immune System The term idiotype was originally introduced for V region epitopes of rabbit immunoglobulins recognized by non-inbred, i.e., allogeneic rabbit antisera (Oudin and Michel, 1963). Later on, idiotopes as phenotypic markers of the complementarity-determining regions (Capra and Kehoe, 1975; Carson and Weigert, 1973; Jeske et al., 1986) could be de®ned by anti-idiotypic antibodies from various genetically different donors, ranging from xenogeneic to autologous (see above). Since anti-idiotypic antibodies were mainly used as a means to study the repertoire development and regulation of particular single or idiotypically crossreactive B cell clones, it is understandable that genetic differences between the donors of idiotype and anti-idiotype were not regarded as being important and consequently ``the term `idiotype' is now used irrespective of the source of the antiserum'' (Nisonoff, 1997). This view is also acceptable when anti-idiotypes are used as surrogate antigens for vaccinations (Nisonoff, 1997), since even xenogeneic anti-Id have been shown to affect physiologically important idiotypic networks (Bluestone et al., 1982). However, when considering the biological function of anti-idiotypic antibodies and thus of a hypothetical idiotypic network, it is essential to remember that any immune system is confronted exclusively with intrinsic self components, but not with non-self antigen receptors from other immune systems, with the only exception that early ontogeny is under maternal in¯uence (see below). Therefore, the real function of an idiotypic network can a priori only be unraveled with autologous reactants Ab1, Ab2, Ab3, and so forth. Since this is not practicable, at least syngeneic anti-idiotypic antibodies seem to be mandatory. This position is corroborated by several observations which have demonstrated that Ab2 of allo- or syngeneic origin may exhibit quite different speci®cities: (a) A crossreactive idiotype of a-(1,3)-dextran-binding BALB/c myeloma proteins could only be de®ned by allogeneic antisera from A / J mice, while syngeneic anti-idiotypic antisera recognized individual idiotypes of the respective myeloma proteins (Schuler et al., 1981).

idiotypic network functions

207

(b) Differences between allo- and syngeneic anti-idiotypic antibodies have also been described for anti-arsonate antibodies in as much as the stimulation of allogeneic anti-Id is comparable to antibody responses to foreign antigens and not to internal network antigens (Meek and Capra, 1990; Meek et al., 1989, 1987). (c) In rabbits, it has also been observed that autologous anti-Id consistently showed a reactivity pattern more restricted than that of isologous ones (Rodkey et al., 1983). The different reactivities of allo- and xenogeneic anti-idiotypic antibodies on the one hand and syngeneic and autologous anti-idiotypic antibodies on the other are depicted in Fig. 1 and the previous observations lead to the following conclusions: 3. The greater the genetic difference between Id and anti-Id donors, the broader the crossreactivity of the corresponding anti-Id will be and the opposite seems to hold true when anti-Id are raised in a syngeneic or even an autologous host. 4. The real biological function of the idiotypic network can only be deduced from the reactivities of real internal, namely autologous, or at least syngeneic network components. C. Idiotypic Interactions Deal with D Region Variability of Antigen-Induced Antibodies In the majority of investigations, the idiotypic network has been studied with xenogeneic and allogeneic anti-idiotypic antibodies (Cerny and Cronkhite, 1983; Kearney et al., 1983), and even these could interfere with physiological idiotypic interactions (Bluestone et al., 1982; Moser et al., 1983; Rajewsky and Takemori, 1983; Urbain et al., 1983). This internal reactivity of external antiidiotypic antibodies can easily be explained by the fact that idiotopes which are recognized by syngeneic or autologous anti-idiotypes may be included in the broader reactivity pattern of allo- and xenogeneic Ab2 (see above and Rodkey et al., 1983). However, since syngeneic differ from allogeneic anti-idiotypic responses, which more closely resemble those to external antigens (Meek et al., 1987), the question arises as to which idiotypic speci®cities of antigen-induced antibodies (Ab1) are recognized by real internal, i.e., autologous or syngeneic Ab2 antibodies. This question also results from the following observations. The immune response of BALB/c mice to the random terpolymer of (Glu60 ±Ala30 ±Tyr10 )n (GAT) was originally analyzed with xenogeneic rabbit anti-idiotypic antiserum recognizing a collection of public idiotopes termed p.GAT which are determined by VH CDR2 and VL CDR3 (Fougereau et al., 1990a). However, when BALB/c mice were immunized with particular GAT-speci®c syngeneic

A

B

Allogeneic or xenogeneic BALB/c Id = Ab1 induction

1 reaction

2 reaction

3

4

5

A/J anti-Id = Ab2

reaction

reaction

1'

Syngeneic or autologous BALB/c Id = Ab1 1

BALB/c anti-Id = Ab2 induction

1'

a crossreactive 2' idiotope (CRI) IdX

2

anti-Id1 defines an individual IdI 2' idiotope

3'

3

3'

4'

4

4'

5'

5

5'

anti-Id1 defines

reaction

anti-IdI2

anti-IdI3

anti-IdI4

anti-IdI5

Fig. 1. Distinct reactivities of anti-idiotypic antibodies of allogeneic/xenogeneic or syngeneic/autologous origin. (A) Anti-idiotypic antibodies (anti-Id ˆ Ab2) that are raised in an allogeneic or xenogeneic recipient exhibit additional reactions with other than the inducing idiotype (Id ˆ Ab1). Therefore, such Ab2 can de®ne a crossreactive idiotope (CRI or IdX). (B) In contrast, when anti-idiotypic antibodies are induced in a syngeneic or even the autologous host they react almost exclusively with the inducing Ab1 and therefore de®ne an individual idiotope (IdI).

idiotypic network functions

209

monoclonal antibodies, the resulting monoclonal anti-Id never recognized public idiotopes of the p.GAT collection, but only private or individual idiotopes (Leclercq et al., 1982; Somme et al., 1983), a ®nding which is in line with the reactivity pattern depicted in Fig. 1. The authors concluded that an individual is normally tolerant to its germline-encoded gene products, i.e., the p.GAT idiotopes, and that it is very dif®cult to overcome this suppression, whereas an anti-idiotypic immune response to individual idiotopes is not suppressed in the syngeneic host (Somme et al., 1983). This view has been con®rmed in the rat by showing that in a fully syngeneic system only private, conformational idiotopes of the H chain, but not the L chain, were immunogenic and that each Ab1 expressed its own individual idiotopes (Leger and Dean, 1991). Such an expression of idiotopes by the heavy chain variable region has also been described before (Black et al., 1976). Hence, more precisely it must be asked to which parts of the immunoglobulin variable domains, in particular of the H chain, an individual is not tolerant so that an autologous or a syngeneic anti-idiotypic immune response to private idiotopes can be induced. Apart from the generation of somatically mutated antibodies during T-dependent immune responses, which can be recognized by idiotypically speci®c T cells (see below), the following investigations prove an exceedingly strong association of private idiotopes with the CDR3 of the heavy chain: (a) In the primary response to the hapten NP in C57BL/6 mice a major idiotype (NPb ) is expressed that can be characterized by D region-speci®c syngeneic anti-idiotypes (Rajewsky et al., 1981) or by allogeneic anti-idiotypes derived from CBA mice immunized with the NPb idiotype-expressing mAb B18 (m, l) (Reth et al., 1979). These allo-de®ned idiotopes were further analyzed and could be localized to the VH ±D boundary, thus allowing the conclusion that ``the VH to D boundary serves as a target of idiotypic selection'' (Cumano and Rajewsky, 1985). Strikingly, the exchange of a single D region-encoded amino acid destroyed all idiotopes recognized by allogeneic anti-idiotypic antibodies (Radbruch et al., 1985). (b) The immune response of A/J mice to the hapten p-azophenylarsonate is characterized by the recurrent crossreactive idiotype A (CRIA ), which is de®ned by allogeneic monoclonal anti-Id of BALB/c origin that react with idiotopes of the VH CDR2, the D region, and VL CDR1 and 2 (Jeske et al., 1986). Although the CRIA -encoding genes are probably absent from the genome of BALB/c mice, these can be manipulated with xenogeneic rabbit anti-CRIA antibodies to express CRIA -like antibodies. The CRIA and CRIA -like antibodies are encoded by different VH genes, but they comprise highly homologous D gene segments (Moser et al., 1983), thereby demonstrating that the idiotypic manipulation was directed at the D region. The importance of the D region for

210

hilmar lemke and hans lange

expression of the dominant CRIA has also been stressed by others who showed that CRIA -positive and CRIA -negative anti-Ars antibodies differ markedly in their D regions while being very homologous in their VH and VL sequences (Gridley et al., 1985). (c) Although the idiotypic analysis in the galactan system by Rudikoff and coworkers (Rudikoff et al., 1983) was performed with allogeneic anti-Id, they reached the conclusion that ``apparently all of the idiotypic determinants recognized by homologous antisera reside in the most variable part of the immunoglobulin molecule, CDR-3,'' thus con®rming earlier observations that individual idiotopes depend on variations in the D region (Clevinger et al., 1980). (d) The immune response of BALB/c mice to phosphorylcholine (PC) is dominated by a major idiotype (T15) which is identical to that of the PCbinding myeloma protein TEPC15 (Cosenza, 1976). Analyses with syngeneic as well as allogeneic anti-idiotypic antibodies revealed that the T15 idiotype is mainly determined by the D region, although conformational in¯uences from other parts of the VH domain were also detectable (Berek, 1984). This association between D region and expression of private idiotopes has also been strengthened by Kearney and Pollok (1984). An autologous anti-T15 antibody could suppress T15‡ anti-PC antibody production in vivo, whereas anti-PC B cells expressing variant D region sequences were not suppressed. These variants had undergone an altered D±JH joining that caused the insertion of one additional amino acid into the heavy chain CDR3 (Pollok et al., 1984). The authors concluded that ``a function of the heavy chain D region may be to contribute to the formation of molecular target sites for idiotype-directed regulatory cells and/or antibodies'' (Pollok and Kearney, 1984). (e) A xenogeneic rabbit antiserum raised against the VH CDR3 peptide of a monoclonal anti-CD4 antibody displayed anti-idiotypic activity for that particular whole antibody, but not for a panel of other anti-CD4 antibodies, demonstrating that the stretch of amino acids of the VH CDR3 represents a private idiotope (Attanasio et al., 1990). Immunization with synthetic peptides of the third hypervariable region of anti-dextran antibodies gave also rise to idiotypespeci®c antisera (McMillan et al., 1983). (f) Recently, we reported the production of 19 syngeneic anti-idiotypic antibodies to the 2-phenyloxazolone-(phOx)-binding antibody H11.5 (m, k) that expresses the major Ox1 idiotype (IdOx1 ) of the primary anti-phOx response in BALB/c mice. None of these anti-IdOx1 reacted with anti-phOx antibodies containing an altered D region in which arginine as the middle of three amino acids was replaced by glycine (Lange et al., 1996). The ®ndings of these various reports lead to the following suggestions, which are depicted in Fig. 2.

idiotypic network functions

211

Idiotype Anti-Idiotype D region-associated individual idiotopes

IdI

D region-associated individual idiotopes

IdI Fig. 2. Dominant recognition of D region-associated individual idiotopes by syngeneic or autologous anti-idiotypic antibodies (Ab2). If this rule is applicable, the recognition of Ab2 by Ab3, Ab3 by Ab4, and so forth, it seems logical that Ab2 binds Ab1 not with its D region, but possibly with paratope-associated structures. However, this view is hypothetical and presently under investigation. Moreover, this ®gure supposes that Ab2 can function as a surrogate antigen if the corresponding Ab1 coexpresses the D region-associated idiotopes plus the antigen-reactive paratope. This view would explain the observation that even individual anti-idiotypes can seemingly function as ``internal image'' antibodies.

5. In the syngeneic or autologous host, only private antibody idiotopes are immunogenic. 6. Depending on the antigen system, private idiotopes can mostly be recognized by syngeneic and autologous anti-idiotypic antibodies and in some experimental systems also by allogeneic anti-idiotypic antibodies. 7. These private idiotopes are at least mainly determined by and are probably located within or close to the CDR3 of the heavy chain, but as conformational epitopes they may also be in¯uenced by other parts of the variable domains. D. Idiotypic Activation or Suppression of B and T Cells Occurs under Physiological Conditions The crucial point is whether the principle ability to mount syngeneic antiidiotypic immune responses (Sakato and Eisen, 1975) represents a real physiological property of the immune system. Despite the huge amount of experimental work on the idiotype network, the old reservations against its functionality still predominate largely because the signi®cance of a seemingly endless number of idiotypic suppressions is not conceivable. This is probably the reason why most immunologists do not believe the network theory (Marshall, 1996). Even former proponents have turned into opponents since they

212

hilmar lemke and hans lange

(i) now regard idiotypic interactions as epiphenomena, (ii) argue that a network with integral idiotypic T/B cell interactions cannot exist because of the now well-established antigen recognition by T cells, or (iii) simply regard the network as being out of date (personal communications). Given this general bias against the idiotypic network theory, a brief summary of the data that support its assumed physiological function is required in order to redress the balance. 1. Second-Step Idiotypic Responses Reveal a Physiological Function of Idiotypic Cellular Interactions First, it seemed possible to argue against the network since anti-idiotypic responses can, in most cases, only be induced after coupling of the idiotype to a foreign carrier and after prolonged immunizations in the presence of adjuvants. If this argument is valid, one must accept that every induction of the second step in the idiotypic cascade, e.g., induction of anti-idiotypic antibodies (Ab2) after immunization with antigen or induction of Ab3 after immunization with Ab1, represents real physiological properties of the immune system. In fact, numerous investigations have proven its functionality: (a) Shortly after the propagation of the idiotypic network theory it could be shown that the peak immune response to PC with the dominant T15 idiotype was followed by the appearance of anti-T15-secreting plasma cells that were assumed to be involved in the regulation of the anti-PC response (Cosenza, 1976). Such antigen-induced auto-anti-idiotypic (Ab2) responses could be detected in mice (Bonilla et al., 1990; Brown and Rodkey, 1979; Forsyth and Hoffmann, 1990; Kim et al., 1989; Kluskens and Kohler, 1974; Nicoletti and Cerny, 1992) after immunization with thymus-dependent or -independent antigens (Fernandez and Moller, 1979; Goidl et al., 1980) as well as in chickens (Forsyth and Hoffmann, 1990) and humans (Arreaza et al., 1993; Geha, 1983; Saint-Remy et al., 1988). On the other hand, a clear physiological function of idiotypic interactions can also be inferred from the observation that the transient activation of major Ox1 idiotypic antibodies during an immune response can suppress the corresponding Ab2 B cells for long periods of time, the duration of which probably depends on the amount of induced Ab1 antibodies (Lange and Lemke, 1996). (b) Immunization with T-dependent antigens, but not with T-independent antigens, not only induces an antigen-speci®c response, but leads to an increase in Ig-secreting cells as well as serum immunoglobulins which display no reactivity for the antigen used for immunization (Antoine et al., 1977; Rosenberg and Chiller, 1979; Urbain-Vansanten, 1970). Strikingly, these non-antigen-speci®c responses are far stronger than the response to the antigen itself and affect, as a T cell-dependent phenomenon, mainly IgG isotypes (Rosenberg and Chiller, 1979). A proportion of these ``nonspeci®c'' immunoglobulins are idiotypically

idiotypic network functions

213

similar to the simultaneously induced antigen-speci®c antibodies (Avrameas et al., 1976; Cazenave et al., 1974) and such idiotype crossreactive Ig can be enhanced by prior stimulation with antigen or anti-idiotype (Eichmann et al., 1977). (c) In rabbits, it could be shown that anti-idiotype (Ab2) induced Ab3 which, however, displayed no reactivity with the original antigen (Urbain et al., 1977). Despite this, after immunization with the corresponding antigen, the animals produced antibodies with idiotypic characteristics similar to those of the original Ab1, demonstrating that an activation of idiotypic interactions can modulate the clonal composition of an antigen-induced response (Urbain et al., 1977). Moreover, in a fully syngeneic system in mice, immunization with an Ab1 myeloma protein induced not only Ab2 but also often auto-Ab3 and possibly auto-Ab4, and, after immunization with the antigen, these mice produced considerable amounts of idiotypes resembling the starting Ab1 (Bona et al., 1981). Strikingly, even the inoculation of an unmodi®ed Ab1 in the absence of adjuvant induced in syngeneic recipient BALB/c mice not only a thymus-dependent Ab2 but also an autologous Ab3 response (Baskin et al., 1990). The production of auto-Ab3 antibodies after immunization with an Ab1 could be veri®ed by the isolation of Ab2- and Ab3-producing hybridomas (Hebert et al., 1990) and has frequently been described in mice (Baskin et al., 1995, 1990; Powell et al., 1988), rabbits (Comacchio et al., 1996), and tumor patients (Fagerberg et al., 1996; Frodin et al., 1991; Ma et al., 1998; Wettendorff et al., 1989). These data unequivocally demonstrate ``that the immune system of a single animal (or human), after producing speci®c antibodies to an antigen (or antibody), continues to produce antibodies to the idiotopes of the antibodies that it has itself made'' (Jerne, 1985). Hence, this production of autologous second-step antibodies clearly represents a physiological property of the immune system. (d) A further argument for the functionality of the idiotypic network is the frequently successful application of intravenous immunoglobulin (IVIg) in the therapy of autoimmune diseases (Achiron et al., 2000; Macias et al., 1999; Mouthon and Kazatchkine, 1996; Shoenfeld et al., 1997). It is assumed that anti-idiotypic antibodies which are contained in the IVIg preparations suppress the autoantibody-secreting cell clones. 2. Physiological Idiotypic Interactions between B and T Cells and among T Cells Second, the argument that B and T lymphocytes cannot idiotypically be connected via their antigen receptors because of the now well-known mechanism of antigen recognition by T cells is incompatible with the experimental data. (a) Antibodies as internal network protein antigens induce autoanti-idiotypic antibodies that are of the IgM or the IgG isotype and result

214

hilmar lemke and hans lange

from T cell-dependent immune responses (Baskin et al., 1990, 1991; Forsyth and Hoffmann, 1990; Goidl et al., 1980; Jackson and Mestecky, 1979; Kim et al., 1989; Nicoletti and Cerny, 1992; Schrater et al., 1979), forcing the conclusion that T helper cells have to participate in this auto-anti-idiotypic response. Such autologous idiotype-speci®c T helper cells have been demonstrated in different experimental systems and adoptive transfer experiments (Janeway et al., 1987, 1975; Jorgensen and Hannestad, 1977; Julius et al., 1977a, b; Pierce et al., 1981). (b) Furthermore, these normal Id-speci®c TH cells or the corresponding T cell lines derived from them can speci®cally be targeted by anti-idiotypic antibodies (Black et al., 1976; Eichmann and Rajewsky, 1975; Janeway et al., 1987; Kuribayashi et al., 1988; Martinez et al., 1986a; UytdeHaag et al., 1987), thus demonstrating that antigen receptors of B and T cells may express common idiotopes (Ertl et al., 1982; Hammerling et al., 1976; Julius et al., 1980; Kuribayashi et al., 1988; Martinez et al., 1986a, 1987a, b; Roubaty et al., 1990; Sharpe et al., 1985; Tite et al., 1986; Zhou and Whitaker, 1993). Therefore, it is not surprising that anti-idiotypic antibodies raised against immunoglobulin idiotopes could successfully be employed for the isolation of T cell antigen receptors (Kuribayashi et al., 1988; Martinez et al., 1986a; Zhou and Whitaker, 1993). (c) In contrast to idiotypic interaction of antibodies or B cells where recognizing and being recognized cannot be differentiated (Jerne, 1985), there are two different types of idiotypic interactions between T and B cells. As just mentioned, one possibility is that B cell antigen receptors can directly recognize epitopes ˆ idiotopes of T cell antigen receptors (TCR) which, so to speak, are recognized. This interaction is, in contrast to the second type, nonMHC-restricted. On the other hand, the real physiological recognition of protein antigens (including idiotypes) by T cell antigen receptors requires a processing by antigen-presenting cells (APC) followed by the presentation of small peptide fragments through MHC molecules on the surface of these cells. B cells are particularly ef®cient in presenting processed peptides since they can bind antigen speci®cally through their antigen receptor (BCR). Therefore, antigen-speci®c B cells can stimulate T cells at antigen concentrations several orders of magnitude lower than those required for activation by nonspeci®c APC (Pierce et al., 1988; Yi et al., 1996; Yi and Osterborg, 1996). B cells, however, do not only process BCR-bound antigen and present its fragments, but at the same time they present peptides of their intrinsic immunoglobulin H and L chains that can then be recognized by T cells (Fig. 3). Depending on the experimental system, a B cell presentation of intrinsic immunoglobulin peptides can be detected by allotype-speci®c (Rudensky and Yurin, 1989; Rueff-Juy et al., 1995; Yurin et al., 1989) or idiotype-speci®c normal T cells or T cell lines (Bogen et al., 1986; Bogen and Weiss, 1988; Jorgensen et al., 1983;

idiotypic network functions

215

antigenspecific T cells

TCR MHC-II

idiotypespecific

B cells TCR

T cells

allotypespecific

MHC-II

TCR

MHC-II

T cells anti-Ig 2

Ag

anti-Id

BCR

1

BCR

Fig. 3. Idiotypic interaction between B and T cells. Intrinsic antigen receptors are processed to a small degree in resting B cells (arrow 1). Engagement of the BCR by anti-immunoglobulin or antiidiotypic antibody has been shown to enhance this processing (arrow 2). We assume that this is also the case when B cells are activated by antigen. The processing generates not only antigen-peptides (black squares) but also allotypic peptides (gray squares derived from gray areas of the constant region of the BCR) and idiotypic peptides (white squares derived from the white areas of the BCR). All three kinds of peptides can be presented by MHC class II molecules on the cell membrane and can then be recognized by antigen-, allotype-, or idiotype-speci®c T cells. Idiotype-speci®c T cells mainly recognize D region-associated peptides or somatically mutated peptides of the antibody V regions. In addition to this MHC-restricted idiotypic interaction of B and T cells, it has been demonstrated that the T cell antigen receptor may be recognized in an MHC-independent fashion by anti-idiotypic antibodies (not shown here).

Jorgensen and Hannestad, 1977, 1979, 1982; King et al., 1993; Mazel et al., 1990; Rudensky et al., 1990; Weiss and Bogen, 1989; Wilson et al., 1990). These Id-speci®c T cells recognize their corresponding idiotypic peptides in association with MHC class II molecules also when presented by nonspeci®c APC (Dembic et al., 2001, 2000; Rees et al., 1987; Saeki et al., 1990; Yi et al., 1996).

216

hilmar lemke and hans lange

Id-speci®c T cells may belong to the MHC II-restricted CD4‡ cells of the TH 1 as well as the TH 2 type (Bogen and Weiss, 1993; Lauritzsen and Bogen, 1991; Yi and Osterborg, 1996; Zang et al., 2000b) and either exert a helper function or may be cytotoxic for B lymphoma or myeloma cells (Bogen et al., 1995; Dembic et al., 2000; Lauritzsen et al., 1993, 1994). However, Id-speci®c, CD8‡ T cells that recognize the respective peptides in the context of MHC class I molecules have also been described (Chakrabarti and Ghosh, 1992a, b; see also below). (d) Hence, the question arises again as to which idiotypic peptides are recognized by Id-speci®c T cells and where these peptides are located in the antibody V regions. Although Id-speci®c T cells or T cell clones may recognize somatically mutated V region idiotypic peptides of the light chain (Bogen and Weiss, 1993; Eyerman and Wysocki, 1994; Jorgensen et al., 1983; Jorgensen and Hannestad, 1979; Kristoffersen et al., 1987; Rudensky et al., 1990; Sakato et al., 1990; Weiss and Bogen, 1989), there seems to be more recognition of VH region-encoded idiotypic peptides (Fagerberg et al., 1999; Gleason and Kohler, 1982; McNamara and Kohler, 1984; Saeki et al., 1990; Wen and Lim, 1999). Moreover, it is striking that Id-speci®c T cells in particular recognize individual idiotypic peptides of the VH CDR3 (Clemens et al., 1998; Masaki et al., 1996; Wen and Lim, 1997, 1998; Wen et al., 1998) and that this T cell stimulatory capacity of peptides of the VH CDR3 loop can be used arti®cially to stimulate T cells by replacing the VH CDR3 by foreign peptides (Zaghouani et al., 1993a, b). Moreover, it has been shown that recognition of a VH CDR3 peptide by an Idspeci®c T cell clone is of exquisite speci®city in that ``not even minimal sequence deviations of the J558 Id peptide are allowed'' (Clemens et al., 1998). This is reminiscent of several reports which demonstrated that the exchange of a single amino acid of the D region destroyed all idiotopes recognized by whole sets of anti-idiotypic antibodies (see above; e.g., Lange et al., 1996; Radbruch et al., 1985). It is of particular interest that resting small B cells are also able to present their idiotypic peptides to Id-speci®c T cells in vivo (Munthe et al., 1999) and that this presentation of intrinsic idiopeptides can be enhanced by engagement of the BCR either by anti-immunoglobulin (Bartnes and Hannestad, 1997) or by anti-idiotype (Liu et al., 1994). These reactivities are shown in Fig. 3. Taken together, the above reports prove that idiotopes of B cell antigen receptors are either recognized as conformational epitopes by anti-idiotypic antibodies or may be perceived as idiotypic peptides by T lymphocytes. Hence, ``Id processing by APC, including B cells, destroys the ®rst and reveals the second category'' (Hannestad et al., 1986). Strikingly, autologous or syngeneic idiotope-speci®c B as well as T cells mainly recognize non-germline-encoded VH CDR3-associated variability or somatically generated mutations within the variable regions of the H and L chains. (e) Finally, the involvement of idiotope-speci®c T cells in physiological cellular interactions has been particularly well documented in animal experi-

idiotypic network functions

217

mental systems of various diseases, such as experimental SLE (Blank et al., 1991; Fricke et al., 1991), experimental autoimmune encephalomyelitis (Yamamura et al., 1994), and autoimmune diabetes (Tikochinski et al., 1999), or in tumor models like B cell lymphoma (Kwak et al., 1990, 1996a, b) or myeloma (Bogen et al., 1995; Dembic et al., 2000; Lauritzsen et al., 1994). Experimental autoimmune encephalomyelitis (EAE), which is an animal model for multiple sclerosis (MS; Jiang et al., 1992), is induced by CD4‡ T cells which react with myelin basic protein (MBP; Wekerle, 1991). Consequently, attenuated MBPreactive CD4‡ T cell clones can be employed for the induction of anti-EAE immunity (Ben-Nun et al., 1981a, b) that depends on cytolytic T cells of the CD8‡ phenotype which are idiotypically speci®c for the inducing encephalitogenic CD4‡ T cells (Jiang et al., 1998, 1992; Lider et al., 1988; Sun et al., 1988). Since the CD8‡ T cells are speci®c for Vb chain idiotopes of the encephalitogenic T cells (Jiang et al., 1998), immunizations with TCR idiotypic peptides of the Vb chain, either of the CDR2 (Vandenbark et al., 1989) or the VDJ joining sequences (Howell et al., 1989), were also able to prevent EAE. In concurrence with the idea that T and B cells are intimately connected idiotypically, an ef®cient suppression of active EAE by an intravenous administration of IVIg has also been reported (Achiron et al., 2000). Identical cellular interactions seem to exist in humans. A vaccination of MS patients with irradiated MBP-reactive T cells depleted the MBP-reactive cells from the circulation and cytolytic CD8‡ T cell clones could be isolated from these patients (Zhang et al., 1993). Such an idiotypic T cell vaccination was also possible with peptides of the MBP-reactive clones and the resulting T cells were cytolytic for the auto ˆ MBP-reactive T cells, expressed the CD8 membrane marker, and preferentially recognized VbCDR3-related idiotopes (Zang et al., 2000a). Anti-idiotypic CD4‡ T cells of the TH 2 type were also induced which only reacted with activated target T cells and contributed to the inhibition of the MBP-reactive clones (Zang et al., 2000b). Furthermore, evidence was presented that anti-idiotypic T cell vaccination also induced an anti-idiotypic antibody response that contributed to the suppression of the MBP-reactive T cells in those patients (Hong et al., 2000). Autoimmune diabetes in NOD mice is also induced by CD4‡ T cells that react with a particular peptide of the 60-kDa heat shock protein (hsp60; Tikochinski et al., 1999). Since the autoreactive T cells express a common CDR3 region idiotope that can be directly presented to anti-idiotypic T cells, an idiotypic T cell vaccination with the irradiated autoreactive T cell clone or the CDR3 peptide arrests the disease (Elias et al., 1999). Moreover, the wellknown fact that a shift from one autoimmune disease to another may occur could be documented in NOD mice in which the idiotypic induction of SLE prevented the appearance of spontaneous insulin-dependent diabetes mellitus

218

hilmar lemke and hans lange

(Krause et al., 1999). The successful treatment of this experimental SLE, as well as primary autoimmune anti-phospholipid syndrome (APS), with intravenous immunoglobulin IVIg (see above) points again to an intimate idiotypic connection between the B and the T cell systems (Krause et al., 1995). In other clinical studies it has been shown that autologous anti-idiotypic antibodies to T cell receptors may be generated (Nishimura et al., 1992) and T cells have been observed which react with autologous anti-rabies virus antibodies in an idiotype-speci®c manner (UytdeHaag et al., 1987). Similarly, joint-derived T cells from patients suffering from rheumatoid arthritis may react with self-immunoglobulin or anti-Ig antibody (van Schooten et al., 1994). Moreover, autoantibody-reactive idiotype-speci®c T cells develop in patients with myasthenia gravis (Yi and Lefvert, 1992) and SLE (Williams et al., 1995), and in tumor patients, Id-speci®c T cells have been observed which may even confer protection against B lymphoma (Bendandi et al., 1999; Kwak et al., 1999; Levitsky et al., 1996; Nelson et al., 1996; Osterroth et al., 2000; Wen and Lim, 1997, 1998) or may be directed against myeloma proteins (Cull et al., 1999; Fagerberg et al., 1999; Osterborg et al., 1995; Yi et al., 1997). Hence, in addition to the idiotypic interaction between B cells and the recognition of TCR idiotopes by anti-idiotypic antibodies, antibody idiotopes can be recognized in an MHC-restricted fashion by T cells and there is ample evidence that (i) idiotypic interactions among T cells also occur and (ii) this T±T interaction is only initiated when one partner is activated, as was the case for the presentation of idiotopes of the intrinsic BCR to idiotype-speci®c T cells. These various idiotypic interactions within the adaptive immune system are schematically summarized in Fig. 4. This view is compatible with a recently published receptor presentation hypothesis (Zhang et al., 2001). Third, interest in the regulation of the idiotypic network in the 1980s has been compared with a tidal wave which, after it has gone, ``leaves behind an empty beach'' (Cohn, 1994). This assessment may to a small extent apply to experimental immunologists, but is certainly wrong as far as clinical immunologists are concerned (cited literature herein and Shoenfeld et al., 1997). As evidenced by a short computer-based literature search, the early fundamental investigations of the idiotypic interactions started to bear fruit and there is still a lively interest in idiotypic network regulation. And, after all, how can an unsolved scienti®c problem become out of date? The above-mentioned ®ndings prompt the following conclusions: 8. Autologous anti-idiotypic responses at various stages of the idiotypic cascade depend on T helper cell activity. 9. The reactivity of anti-idiotypic antibodies has uncovered idiotypic crossreactivity between antigen receptors of B and T cells.

idiotypic network functions

219

(a)

BCR

(b)

(c)

(d)

TCR

Fig. 4. Idiotypic interactions within the adaptive immune system. Physiological idiotypic clonal interactions principally occur (a) between antigen receptors of B cells (BCR) and (b), in addition, anti-idiotypic antibodies may recognize antigen-receptors of T cells (TCR) in a non-MHC-restricted way, thus indicating a BCR±TCR crossreactivity. Both reactivities are indicated by black arrows. (c) On the other hand, T cells are able to recognize BCR idiopeptides that are presented by B cells themselves and this recognition is MHC restricted. (d) T cells themselves process and present their TCR idiopeptides via MHC molecules and these presented idiopeptides can be recognized by other T cells. This idiotypic T±T interaction is initiated when the activation of one of the cell clones leads to an enhanced presentation of TCR idiopeptides. The MHC-restricted recognition is indicated by white arrows.

10. Physiological induction of autologous anti-idiotypic antibodies during immune responses to external antigens can in turn lead to activation of idiotypically crossreactive T cells. 11. T cells can recognize idiopeptides of individual idiotopes of T as well as of B cell antigen receptors that are mainly associated with D region variability or created during antigen-induced immune responses by somatic hypermutations; these idiopeptides are most effectively presented by MHC molecules on the surface of T and B cells, but also by nonspeci®c antigen-presenting cells. 12. The presentation of idiopeptides by B and T cells is enhanced by activation of the cells or engagement of their antigen receptor, i.e., by treatment with anti-Ig or anti-idiotype, and, as we assume, probably during the course of an antigen-induced immune response. 13. In various autoimmune diseases and tumors in experimental animals as well as in patients, idiotype-speci®c T cells may be the causative agents or may develop during the disease and may confer protection. 14. The investigations mentioned above reaf®rm the former assessment that the B and T cell systems form an integrated idiotypic network.

220

hilmar lemke and hans lange

E. The Idiotypic Network Transforms Clonal Selection by Antigen into Stimulatory or Suppressive Intrasystemic Clonal Connections The function of the idiotypic network has mainly been seen in the mutual suppression of idiotype, anti-idiotype, and further steps in the idiotypic cascade thus establishing an equilibrium among the cell clones involved, and antigeninduced immune responses were regarded to ``incidentally disturb the dynamic harmony of the system'' (Jerne, 1985). In addition, the assumption that the ontogenetic development of the immune system starts from an incipient network (Jerne, 1984) concurs with the concept that the idiotypic network should be germline encoded (Coutinho, 1995; Fougereau et al., 1990b; Meek and Capra, 1990; Varela and Coutinho, 1991). Indeed, such a genetically programmed network with a high degree of idiotypic connectivity between its members has been found which, however, is restricted to the repertoire of natural antibodies (Avrameas, 1991; Coutinho, 1995; Holmberg and Coutinho, 1985; Varela and Coutinho, 1991). These are to a great extent autoreactive and multispeci®c, they are secreted by B-1 B cells without need for prior antigenic stimulation and they show extensive idiotypic crossreactivity (Avrameas, 1991; Coutinho et al., 1995; Holmberg and Coutinho, 1985). In contrast, the majority of B cells in the adult belong to the population of B-2 cells that are responsible for immune responses to external antigens and were proposed to be idiotypically disconnected (Varela and Coutinho, 1991). This bisection of the immune system has led to the concept of second generation immune networks proposing that only one part forms an idiotypic network through high connectivity and can, because of its internal regulatory function, be regarded as a self-related internal activity set (Holmberg and Coutinho, 1985) or as the central immune system (Varela and Coutinho, 1991). The other population of B-2 cells represents the peripheral immune system, which enables responses to external antigens and therefore shows few idiotypic interactions (Coutinho, 1995; Varela and Coutinho, 1991). This view, however, is not quite compatible with the data presented above for the following reasons: (a) A whole set of data has proven that antigen-induced immune responses also lead to the activation of second and possibly further steps of the idiotypic cascade, con®rming the original conclusion ``that the immune system of a single animal (or human), after producing speci®c antibodies to an antigen (or antibody), continues to produce antibodies to the idiotopes of the antibodies that it has itself made'' (Jerne, 1985). Hence, idiotypic speci®cities reactive with populations of Ab1, Ab2, Ab3, etc., preexist in the system and can indeed by activated under physiological conditions by antigen or the respective preceding internal network component.

idiotypic network functions

221

(b) Furthermore, as outlined above, those speci®cities of anti-idiotypic antibodies as well as idiotype-speci®c T cells that are related to antigen-induced immune responses, are mainly directed at non-germline-encoded D region variability and V region epitopes that are generated by the somatic hypermutation process induced by external antigens. Hence, in contrast to idiotypic connectivity in the central immune system, the idiotypic interactions in the peripheral immune system predominantly depend on and deal with speci®cities that are rendered immunogenic by an activation of B or T cell clones or are created by the immune system's interaction with the environment. This raises the question as to the origin of these anti-idiotypic B and T cell speci®cities. Sequence analyses have demonstrated that anti-idiotypic antibodies related to different antigenic systems generally exhibit somatic mutations, thus allowing the conclusion that the idiotypic network is not strictly germline encoded and that ``somatic events (i.e., mutation, junctional diversity, etc.) are probably important'' in its generation (Meek and Capra, 1990). However, this does not explain the origin of antigen-related Ab2. One possibility, not envisaged so far, would be that the observed somatic mutations were introduced during the prolonged immunization procedures necessary for the isolation of these monoclonal anti-idiotypic antibodies, since none of them could be produced after early primary immunization. The other possibility would be that antigenrelated Ab2 antibodies of the peripheral immune system are the products of unknown antigen-induced immune responses and their quali®cation as antiidiotypic is merely incidental or ``operational'' (Rajewsky, 1983). This second explanation implies (i) that idiotypic interactions within the peripheral immune system may depend on somatic mutation-inducing, namely T-dependent, immune responses and (ii) that this idiotypic network would then indeed start from a germline-encoded basis, but develop during ontogeny, and its generation would then be driven by the continuous encounter with environmental antigens. A determination between both possibilities, which do not have be mutually exclusive, cannot be made at present. Irrespective of this, in the present context it is crucial that idiotypic speci®cities to antigen-induced antibody responses of B2 cells are not only contained in the system but also can be activated under physiological conditions. The analysis of idiotypic network interactions has been hampered by the major drawback that all anti-idiotypic speci®cities within the idiotypic cascade had to be related to the antigen-reactive idiotype (Ab1) as the only ®xed point with a known antigen speci®city, whereas antigen speci®cities of anti-idiotypic Ab2, Ab3, or Ab4 antibodies induced by the Ab1 antibody were unknown. Consequently, antibodies of the idiotypic cascade (Ab2, etc.) were prepared as the ®rst step and then tested for their capacity to regulate the expression of that idiotype or intermediate steps during the antigen- or idiotype-induced

222

hilmar lemke and hans lange

immune responses. The primary concept of an idiotypic network regarded the immune system as being in a state of dynamic harmony (whenever this is possible), which is incidentally disturbed by stimulation with antigen. The outlined experimental procedure, however, as well as its underlying theoretical interpretation represent, with regard to the sequentially activated steps in the idiotypic cascade, a backward directed view. In contrast, we propose the following hypothesis, which is outlined in Fig. 5. (i) As the initial step, external antigen enters the system by clonal selection through speci®c antigen receptors. This event may tolerogize or activate the reactive clones and leads, with respect to the antigen-speci®c paratopes, to the activation of secondary effector functions. (ii) Further on, this contact with environmental antigens and the primary activation of antigen-speci®c idiotypic are transformed into intrasystemic cellular interactions. This process is mediated by a population of anti-idiotypic Ab2 cells and antibodies that recognize non-germline-encoded Ab1 idiotypes, i.e., those determined mainly by D region variability and somatic mutations. (iii) Since Ab2 antigen receptors clonal selection

Ag1

Ag2

Ag3

Ag4

known

unknown

unknown

unknown

Id1.1-.n

Id2.1-.n

Id3.1-.n

Id4.1-.n

clonal connections

Ab1

Ab2

Ab3

Ab4

Fig. 5. Idiotypic crossreactivity allows a generalization of single immunological experiences with external antigens. Antigen activation, through the process of clonal selection, increases the concentration of reactive antibodies, i.e., idiotypes or Ab1. Each of these antibodies displays a set of idiotopes (Id1.1-.n) that are D region-associated when the antibodies stem from the early primary immune response and are composed of germline-encoded V region gene segments or that may represent somatically mutated V region epitopes. The autologous immune system contains the relevant anti-idiotypic antibodies (Ab2) for recognition of Ab1 individual idiotopes. Under appropriate conditions, which depend on the concentration of each Ab1 clonal product and unknown conditions in the network itself, these anti-idiotypic clones may be activated and give rise to elevated concentrations of Ab2 in the circulation. Again, each clonal Ab2 displays a set of idiotopes (Id2.1-.n) that can be recognized by intrinsic anti-(anti-idiotypic) antibodies (Ab3) and so forth. These clonal connections are the basis for an idiotypic cascade, in which the only known antigen is that recognized by Ab1, whereas those antigens recognized by antibody populations of Ab2, Ab3, and Ab4 are entirely unknown. Hence, the autologous activation of the idiotypic cascade by the initial stimulation with antigen leads to the activation of B cell clones with receptors displaying numerous antigen speci®cities. It can be concluded that in this way the idiotypic network allows the generalization of single antigen-induced immune responses.

idiotypic network functions

223

exhibit multiple antigen speci®cities that are different from those of Ab1, the primary stimulation and clonal selection by antigen is followed by clonal connection which may not only suppress but may also activate idiotypically reactive clones. Thereby, an antigen-speci®c activation is transformed into an idiotypespeci®c internal regulatory process. We propose that following antigen activation, onward-directed intrasystemic idiotypic responses represent a basic function of the idiotypic network in an adult individual. These interactions enable the immune system to make use of single immunological experiences with external antigens in a different antigenic context, thus allowing a generalization of such single antigenic stimuli. Hence, in addition to idiotypically induced clonal alteration of unknown relevance, such idiotypic network interactions support the immune system in its battle against environmental antigens. (iv) This idea does not deny the possible suppressive function of anti-idiotypic antibodies. However, we wish to direct attention to the stimulatory potential of the idiotypic network, which has so far not been viewed in this way, and thus avoid the vicious circle of idiotypic suppression that has been overemphasized and has not helped to explain satisfactorily the signi®cance of an extrapolated idiotypic network. From these considerations we draw the following conclusions: 15. The population of antigen-inducible B-2 B cells contains idiotype- as well as anti-idiotype-expressing clones that coexist in the immune system until one of them is stimulated by antigen. 16. Antibodies as antigen receptors of these cells exhibit a dual function in that they recognize antigen with the paratope while internal idiotypic interactions are mainly dependent on non-germline-encoded D region variability and somatic mutations. 17. The dichotomy of the B cell antigen receptor enables a transformation of the paratope-mediated recognition of antigen into an idiotype-speci®c internal regulatory process that leads to the increased production of antibodies with different antigen speci®cities. 18. We propose that this antigen-activated anti-idiotypic internal stimulation represents a basic function of the idiotypic network. In this way, immune responses to single antigens mediated by clonal selection can be transmitted, by idiotypic clonal connections, to other antigen recognition systems. Hence, the idiotypic network provides the mechanism for a generalization of single immunological experiences and it can be expected that the immune system in its entirety gains a bene®t in its battle against environmental microbes. III. Transfer of Maternal Immunological Experience to the Offspring Early ontogeny is a decisive developmental period in which idiotypic clonal interactions exert obvious and important regulatory functions as schematically

224

hilmar lemke and hans lange

knowledge

depicted in Fig. 6. Since this subject has been reviewed recently (Lemke and Lange, 1999) it is discussed only brie¯y here. The peculiarity of the neonatal period in terms of the general development of the immune system and in particular for idiotypic clonal interactions can be deduced from the following observations:

Parental Generation

F1 Generation

Immunological Learning

Immunological Learning Birth

Birth

1

1

2

2

prenatal

postnatal

maternally mediated

environmental

environmental

germlineencoded

germlineencoded

Age

prenatal

postnatal

Age

maternally mediated

Fig. 6. Transfer of immunological knowledge from the mother to the offspring. As a consequence of antigen activation, the immune system gains knowledge of the external world. This immunological knowledge consists of those components that can be used for effector functions, namely, (a) the number of activated B and T cell clones, (b) the amount of each clonal antigen receptor in cell-bound and free form, and (c) their af®nity for antigen. Hence, in each generation, the potential immunological knowledge in the form of germline-encoded VH =VL region gene combinations becomes effective when the system is activated by external antigens (solid black line). Stimulation with T-dependent antigens leads to a potentiation of this genetic background knowledge through the process of somatic hypermutation, giving rise to immune-maturated, highaf®nity IgG antibodies (dotted line). This environmentally induced immunological knowledge can further be enhanced through idiotypic interactions in quantitative (and qualitative?) terms. A collection of high-af®nity IgG antibodies is transferred from the mother before and/or after birth to the offspring (marked with little lines). Early ontogeny is an extremely sensitive period for the development of the immune system and various data indicate that during the neonatal period an immunological imprinting (hatched area, marked with circle 2) takes place which is brought about by idiotypic interactions between maternal antibodies and B and T cell antigen receptors of the nascent immune system of the newborn. (Most likely, this process is guided by various cellular growth factors that are also transferred from the mother.) Since maternal high-af®nity IgG antibodies not only function as passive protective mediators but also exert various stimulatory effects in the immune system of the offspring (see text), it can be assumed that the overall immunological knowledge of the newborn's immune system increases considerably during the immunological imprinting period (circle 2). This function of the idiotypic network thus allows the newborn to gain a bene®t from the immunological knowledge of the mother that has been acquired through a lifelong continuous encounter with external antigens. The development of this enhanced knowledge is indicated by the upper line (circle 1) and depends on the immunological imprinting (circle 2).

idiotypic network functions

225

(a) Whereas idiotypic suppression by anti-idiotypic antibodies in the adult is only of short duration, idiotypic responses in the neonatal period induce longterm effects that can even persist until adulthood (Augustin and Cosenza, 1976; Hiernaux et al., 1981; Rothstein and Vastola, 1984; Strayer et al., 1975) or the reactive idiotype may even be permanently lost (Kearney et al., 1983; Vakil et al., 1986). Such anti-idiotypic suppression can be induced experimentally not only by exogenous antibodies but also by maternally derived anti-idiotypic antibodies (Victor et al., 1983; Weiler et al., 1977). Moreover, if anti-idiotypic antibodies either induced by direct immunization of the neonate or obtained from the mother are directed toward multispeci®c, crossreactive IgM antibodies, the resulting effects are not only related to the idiotype speci®c clones but also may give rise to a long-lasting severe disturbance of a large part of the entire immune system (Bernabe et al., 1981; Vakil et al., 1986). Evidently, these experiments have demonstrated that antibodies of the mother have a transgenerational in¯uence and guide the generation of the antibody repertoire (Andrade et al., 1990; Kearney and Vakil, 1986; Martinez et al., 1986b; Stein, 1985; Vakil et al., 1986; Victor et al., 1983; Weiler et al., 1977; Wikler et al., 1980). In fact it can be concluded that, in a natural situation, the development of the newborn's immune system depends on these earliest ontogenetic idiotypic experiences. Since the transferred maternal antibodies are mainly the products of thymus-dependent immune responses that have undergone immune maturation, it is clear that the earliest ontogenetic development of the immune system depends on and re¯ects the immunological history of the mother. (b) This raises the ®rst question as to how long after birth such long-term idiotypic alternations can be induced and when the transition to the adult situation occurs. In the mouse, some investigations suggest that this sensitive period is restricted to the ®rst 3 weeks of life: (i) The induction of long-term effects of idiotypic±anti-idiotypic interactions is only possible around birth (see above). (ii) Particular idiotypic interactions were only operative during the ®rst weeks in life, but exert a determinative in¯uence on the composition of the adult repertoire, allowing the assumption of developmental windows (Kearney and Vakil, 1986; Vakil and Kearney, 1988) and the conclusion that the ``idiotypedirected selection of the adult B cell repertoire appears to be limited to fetal± neonatal stages of development'' (Elliott and Kearney, 1992). (iii) Similarly, an immunoglobulin-dependent selection of the T cell repertoire in the neonate is stable until adulthood (Martinez et al., 1986b) and ``only operates for the ®rst 3 weeks of life'' and could be passed on to the F2 generation (Martinez et al., 1985). Strikingly, latest results have also shown that the ®rst 3 weeks of life of a mouse are a ``crucial period during which the T-cell compartment is conditioned to permit normal B cell development'' (Baumgarth et al., 2000). (iv) Newborn mice can mount an immune response to syngeneic IgE during the ®rst 2±3 weeks of life that causes a long-term suppression of IgE synthesis (Haba and

226

hilmar lemke and hans lange

Nisonoff, 1988) and the suppressive IgG±anti-IgE antibodies can only be induced during a limited short period from 2 to 12 days of life (Haba and Nisonoff, 1990, 1994, 1995). These results concur with the idea that immunological experiences in early life contribute to the appearance of allergic diseases in later life (de Weck et al., 1995; Holt, 1995) and that the transmission of atopy is only detectable through the maternal line (Cookson et al., 1992). Collectively, these data emphasize the determinative in¯uence of idiotypic interactions between maternal antibodies and the nascent immune system of newborns and show that early ontogeny is an exceptionally sensitive phase. It has been argued that during this sensitive period maternal antibodies cause an immunological imprinting aÁ la Konrad Lorenz (Lemke and Lange, 1999). (c) The second question is whether those clonal alterations which are induced during that sensitive phase of immunological imprinting are of any biological signi®cance and are not only, so to speak, just overtones that do not change the melody. Again several reports have proven that in early ontogeny, idiotypic suppression by maternal antibodies only represents one side of the coin while the other side offers idiotypic activation that can clearly have advantageous effects with regard to protection against environmental microbes, tumor cell growth, or autoimmune diseases. (i) Neonatal treatment with minute amounts of idiotypic or anti-idiotypic antibodies can select clones that are not expressed during an immune response in normal mice (Hiernaux et al., 1981; Rubinstein et al., 1982) and such an early activation may have long-lasting effects (Rubinstein et al., 1983). (ii) Similarly, the injection into newborn mice of a monoclonal antiidiotypic antibody directed to a myeloma antibody can induce protection against myeloma tumor cell growth that lasts until adulthood and is mediated by idiotype-speci®c T cells (Gorczynski et al., 1986). (iii) F1 mice born to immunized dams developedÐeven without immunizationÐan IgM titer to the maternally experienced antigen (Ismaili et al., 1995; Lemke et al., 1994). (iv) When female mice were exposed to collagen II, their offspring had a signi®cantly reduced incidence of collagen II-induced arthritis that was proposed to result from an altered immune response in those animals (Van Vollenhoven et al., 1990). (v) Moreover, a neonatal injection of idiotype to the Schistosoma soluble egg antigen not only induced B and T cell responsiveness at an adult age but also improved the pathology and prolonged survival when these animals were infected with Schistosoma mansoni (Montesano et al., 1999a,b). (vi) The postnatal transfer of a neutralizing monoclonal antibody that reacted with the F-glycoprotein (F-gp) of respiratory syncytial virus sensitized the offspring in such a way that a primary immunization after weaning with the puri®ed RSV F-gp resulted in a secondary type of immune response with the formation of virus-neutralizing antibodies (Okamoto et al., 1989). Similarly, when F2 mice received, via their F1 mothers, a small

idiotypic network functions

227

amount of antibodies from the tertiary response of their grandmothers, half of them developed a secondary anti-hapten as well as an anti-carrier response after primary immunization with the hapten-carrier complex phOxcoupled chicken serum albumin (Lemke et al., 1994). These ®ndings unequivocally answered the initial question of our investigations: whether the molecular characteristics of the primary anti-phOx immune response of animals that had received IgG±anti-phOx antibodies from their mothers are identical to those of ``normal'' mice. This was obviously not the case and the experiments demonstrated that maternal antibodies are able to exert an active stimulatory in¯uence on the nascent immune system of the newborns. (vii) The early primary anti-phOx antibody response in normal BALB/c mice is characterized by the dominant Ox1 idiotype (IdOx1 ), which is of the highest af®nity (Berek and Milstein, 1987). By contrast, in the offspring of tertiary immunized dams, half of the non-IdOx1 antibodies among the early Day 7 primary antibodies exhibited a strong increase in af®nity, being either identical (60%) or even 7±25 times higher (40%) than that of IdOx1 antibodies (Lange et al., 1999). Collectively, these various observations demonstrate that maternally derived idiotype or anti-idiotype, without participation of antigen, can induce profound, advantageous, and long-lasting in¯uences on immune responses to antigens experienced by the mother. Maternal IgG antibodies that are transferred to the offspring are mainly derived from thymus-dependent immune responses which had undergone immune maturation, i.e., they had acquired a higher af®nity through somatic mutations and, therefore, represent the mother's collective immunological experience with external antigens. It may be called into question whether the observed advantageous effects of maternal antibodies may also result from early immunological experiences of the mother and not only from immune responses that occur shortly before or during pregnancy. However, since the memory of the B cell compartment depends not only on memory cells but also on long-lived plasma cells in the bone marrow (Manz et al., 1997) whose survival and antibody-secretion is independent of antigen (Manz et al., 1998), it can be expected that a large proportion of serum IgG antibodies result from these bone marrow plasma cells and can be transferred to the offspring for a long period of time. This favors the idea that maternal antibodies represent a large proportion of the entire immunological experience of the mother and have a transgenerational potential to guide the education of the neonatal immune system. These considerations lead to our ®nal conclusion: 19. The proposed basic function of idiotypic clonal interactions is that maternal immunological knowledge in the form of somatically mutated, highaf®nity IgG antibodies, which have been acquired through a continuous

228

hilmar lemke and hans lange

encounter with external antigens, exert transgenerational advantageous effects on the immune system of the offspring and guide its early ontogenetic development. IV. Conclusion Clonal interactions of B cells by idiotype-speci®c mutual recognition of their antigen receptors with the participation of T cells form a web of unknown density, referred to as the idiotypic network. These clonal connections were proposed to ful®ll important internal regulatory functions, e.g., to establish the individual's molecular identity at the start of ontogeny by a germline-encoded network organization of the central immune system (Varela and Coutinho, 1991). However, the functional signi®cance of this hypothetical idiotypic network, especially in relation to antigen-induced immune responses, remained a mystery. In view of this, we postulate that the basic function of the internal idiotypic connection between B and T cell antigen receptors is to transform antigen-induced cellular activations, by idiotypic crossreactivity, into the regulation of cell clones with different antigen speci®cities. This process leads not only to the suppression of major clones but also to the activation of minor ones. This supports the diversi®cation of the T and B cell repertoire and ensures that the interaction with the world of external antigens can harmonically be integrated into the whole system. The activating property induces clones with different antigenic speci®cities and thus allows the generalization of single immunological experiences by idiotypic clonal connections. In this way, the immune system in its entirety bene®ts in its battle against environmental microbes. Such idiotypic clonal connections are particularly effective in early ontogeny. During a short neonatal imprinting period, maternal immunological knowledge in the form of somatically mutated, high-af®nity IgG antibodies, which are acquired through a continuous encounter with external thymusdependent antigens, guides the initial ontogenetic development of the immune system. This enables the transfer of maternal immunological experiences to the offspring and so exerts long-lasting transgenerational advantageous effects in the offspring. We hope that these considerations will help to stimulate a renewed interest in idiotypic network interactions for the purpose of answering numerous unsolved questions. Acknowledgments We wish to dedicate this paper to the memory of Niels Kaj Jerne who recognized the biological signi®cance of idiotypically mediated clonal interactions within the immune system. We apologize to all those who contributed so very much to the elaboration of the idiotypic network as an integrating system of B and T cells and whose work could not be quoted herein. We are indebted to our colleague Dr. Lee Shaw for critical reading of the manuscript. This work was supported by the

idiotypic network functions

229

Deutsche Forschungsgemeinschaft through Grants Le 328/3-4 and Le 328/7-1 and by the HenselStiftung of the Christian-Albrechts-University at Kiel.

References Achiron, A., Mor, F., Margalit, R., Cohen, I. R., Lider, O., and Miron, S. (2000). Suppression of experimental autoimmune encephalomyelitis by intravenously administered polyclonal immunoglobulins. J. Autoimmun. 15, 323±330. Andrade, L., Martinez, A. C., and Coutinho, A. (1990). In ``The Immunology of the Fetus'' G. Chaouat, Ed.), pp. 187±192. CRC Press, Boca Raton, FL. Antoine, J. C., Petit, C., Bach, M. A., Bach, J. F., Salomon, J. C., and Avrameas, S. (1977). T cell dependence of cells synthesizing immunoglobulin without detectable antibody function induced after an antigenic stimulation. Eur. J. Immunol. 7, 336±341. Arreaza, E. E., Gibbons, J. J., Jr., Siskind, G. W., and Weksler, M. E. (1993). Lower antibody response to tetanus toxoid associated with higher auto-anti-idiotypic antibody in old compared with young humans. Clin. Exp. Immunol. 92, 169±173. Attanasio, R., Kennedy, R. C., Allan, J. S., Maino, V. C., Buck, D., and Kanda, P. (1990). Antiidiotypic antibodies of a prede®ned speci®city generated against CDR3VH synthetic peptides de®ne a private anti-CD4 idiotype. Mol. Immunol. 27, 513±522. Augustin, A., and Cosenza, H. (1976). Expression of new idiotypes following neonatal idiotypic suppression of a dominant clone. Eur. J. Immunol. 6, 497±501. Avrameas, S. (1991). Natural autoantibodies: from `horror autotoxicus' to `gnothi seauton.' Immunol. Today 12, 154±159. Avrameas, S., Antoine, J. C., Ternynck, T., and Petit, C. (1976). Development of immunoglobulin and antibody-forming cells in different stages of the immune response. Ann. Immunol. (Paris) 127, 551±571. Bartnes, K., and Hannestad, K. (1997). Engagement of the B lymphocyte antigen receptor induces presentation of intrinsic immunoglobulin peptides on major histocompatibility complex class II molecules. Eur. J. Immunol. 27, 1124±1130. Baskin, J. G., Powell, T. J., Jr., Srinivas, R. V., Elliott, M., and Lamon, E. W. (1995). Moloney leukemia virus-induced cell surface antigen mimicry by monoclonal antibodies. Immunol. Res. 14, 292±316. Baskin, J. G., Ryan, T. M., Vakil, M., Kearney, J. F., and Lamon, E. W. (1990). Thymus-dependent antiidiotype and anti-antiidiotype responses to a dinitrophenyl-speci®c monoclonal antibody. J. Immunol. 145, 202±208. Baskin, J. G., Vakil, M., Kearney, J. F., Ryan, T., and Lamon, E. W. (1991). Thymus-dependent network responses to a monoclonal cross-reactive antiidiotypic antibody. J. Immunol. 147, 1849±1855. Baumgarth, N., Jager, G. C., Herman, O. C., and Herzenberg, L. A. (2000). CD4‡ T cells derived from B cell-de®cient mice inhibit the establishment of peripheral B cell pools. Proc. Natl. Acad. Sci. USA 97, 4766±4771. Bendandi, M., Gocke, C. D., Kobrin, C. B., Benko, F. A., Sternas, L. A., Pennington, R., Watson, T. M., Reynolds, C. W., Gause, B. L., Duffey, P. L., Jaffe, E. S., Creekmore, S. P., Longo, D. L., and Kwak, L. W. (1999). Complete molecular remissions induced by patient-speci®c vaccination plus granulocyte±monocyte colony-stimulating factor against lymphoma [see Comments]. Nat. Med. 5, 1171±1177. Ben-Nun, A., Wekerle, H., and Cohen, I. R. (1981a). The rapid isolation of clonable antigen-speci®c T lymphocyte lines capable of mediating autoimmune encephalomyelitis. Eur. J. Immunol. 11, 195±199. Ben-Nun, A., Wekerle, H., and Cohen, I. R. (1981b). Vaccination against autoimmune encephalomyelitis with T-lymphocyte line cells reactive against myelin basic protein. Nature 292, 60±61.

230

hilmar lemke and hans lange

Berek, C. (1984). The D segment de®nes the T15 idiotype: The immunoresponse of A/J mice to Pneumococcus pneumoniae. Eur. J. Immunol. 14, 1043±1048. Berek, C., and Milstein, C. (1987). Mutation drift and repertoire shift in the maturation of the immune response. Immunol. Rev, 96, 23±41. Bernabe, R. R., Coutinho, A., Cazenave, P. A., and Forni, L. (1981). Suppression of a ``recurrent'' idiotype results in profound alterations of the whole B-cell compartment. Proc. Natl. Acad. Sci. USA 78, 6416±6420. Black, S. J., Hammerling, G. J., Berek, C., Rajewsky, K., and Eichmann, K. (1976). Idiotypic analysis of lymphocytes in vitro. I. Speci®city and heterogeneity of B and T lymphocytes reactive with anti-idiotypic antibody. J. Exp. Med. 143, 846±860. Blank, M., Ben-Bassat, M., and Shoenfeld, Y. (1991). Modulation of SLE induction in naive mice by speci®c T cells with suppressor activity to pathogenic anti-DNA idiotype. Cell. Immunol. 137, 474±486. Bluestone, J. A., Auchincloss, H., Jr., Cazenave, P. A., Ozato, K., and Sachs, D. H. (1982). Antiidiotypes to monoclonal anti-H-2 antibodies. III. Syngeneic anti-idiotypes on antibodies induced by in vivo administration of xenogeneic anti-idiotypes. J. Immunol. 129, 2066±2068. Bogen, B., Malissen, B., and Haas, W. (1986). Idiotope-speci®c T cell clones that recognize syngeneic immunoglobulin fragments in the context of class II molecules. Eur. J. Immunol. 16, 1373±1378. Bogen, B., Munthe, L., Sollien, A., Hofgaard, P., Omholt, H., Dagnaes, F., Dembic, Z., and Lauritzsen, G. F. (1995). Naive CD4‡ T cells confer idiotype-speci®c tumor resistance in the absence of antibodies. Eur. J. Immunol. 25, 3079±3086. Bogen, B., and Weiss, S. (1988). B lymphoma cells process and present their endogenous Ig: Implications for network theory. Adv. Exp. Med. Biol. 237, 877±882. Bogen, B., and Weiss, S. (1993). Processing and presentation of idiotypes to MHC-restricted T cells. Int. Rev. Immunol. 10, 337±355. Bona, C. A., Heber-Katz, E., and Paul, W. E. (1981). Idiotype-anti-idiotype regulation. I. Immunization with a levan-binding myeloma protein leads to the appearance of auto-anti-(anti-idiotype) antibodies and to the activation of silent clones. J. Exp. Med. 153, 951±967. Bonilla, F. A., Zaghouani, H., and Bona, C. (1990). Patterns of idiotypic similarity and their structural bases among antibodies speci®c for foreign or self antigens. Chem. Immunol. 48, 49±62. Brient, B. W., and Nisonoff, A. (1970). Quantitative investigations of idiotypic antibodies. IV. Inhibition by speci®c haptens of the reaction of anti-hapten antibody with its anti-idiotypic antibody. J. Exp. Med. 132, 951±962. Brown, J. C., and Rodkey, L. S. (1979). Autoregulation of an antibody response via network-induced auto-anti-idiotype. J. Exp. Med. 150, 67±85. Capra, J. D., and Kehoe, J. M. (1975). Hypervariable regions, idiotypy, and the antibody-combining site. Adv. Immunol. 20, 1±40. Carson, D., and Weigert, M. (1973). Immunochemical analysis of the cross-reacting idiotypes of mouse myeloma proteins with anti-dextran activity and normal anti-dextran antibody. Proc. Natl. Acad. Sci. USA 70, 235±239. Cazenave, P. A., Ternynck, T., and Avrameas, S. (1974). Similar idiotypes in antibody-forming cells and in cells synthesizing immunoglobulins without detectable antibody function. Proc. Natl. Acad. Sci. USA 71, 4500±4502. Cerny, J., and Cronkhite, R. (1983). An independent regulation of distinct idiotopes of the T15 idiotype by autologous T cells. Ann. N. Y. Acad. Sci. 418, 31±39. Cerny, J., and Hiernaux, J. (Eds.) (1990). ``Idiotypic Network and Diseases.''d Am. Soc. Microbiol., Washington, DC. Chakrabarti, D., and Ghosh, S. K. (1992a). Induction of syngeneic cytotoxic T lymphocytes against a B cell tumor. I. Role of idiotypic immunoglobulin. Cell. Immunol. 142, 54±66.

idiotypic network functions

231

Chakrabarti, D., and Ghosh, S. K. (1992b). Induction of syngeneic cytotoxic T lymphocytes against a B cell tumor. II. Characterization of anti-idiotypic CTL lines and clones. Cell. Immunol. 144, 443±454. Chen, J. J., and KoÈhler, H. (1990). In ``Idiotypic Network in Biology and Medicine'' (A. Osterhaus and F. Uytdehaag, Eds.), pp. 141±147. Excerpta Medica, Amsterdam. Clemens, A., Rademaekers, A., Specht, C., and Kolsch, E. (1998). The J558 VH CDR3 region contributes little to antibody avidity: however, it is the recognition element for cognate T cell control of the a(1 ! 3) dextran-speci®c antibody response. Int. Immunol. 10, 1931±1942. Clevinger, B., Schilling, J., Hood, L., and Davie, J. M. (1980). Structural correlates of cross-reactive and individual idiotypic determinants on murine antibodies to a(1 ! 3) dextran. J. Exp. Med. 151, 1059±1070. Cohn, M. (1994). The wisdom of hindsight. Annu. Rev. Immunol. 12, 1±62. Comacchio, R. M., Bradley, J., and Hohmann, A. W. (1996). A comparison of the anti-idiotypic responses generated by antibodies to a protein and a hapten: A common interspecies idiotype on antibodies against human albumin induces an idiotypic network in rabbits. Immunol. Cell Biol. 74, 72±80. Cookson, W. O., Young, R. P., Sandford, A. J., Moffatt, M. F., Shirakawa, T., Sharp, P. A., Faux, J. A., Julier, C., Nakumuura, Y., Nakumura, Y., et al. (1992). Maternal inheritance of atopic IgE responsiveness on chromosome 11q [see Comments]. Lancet 340, 381±384. Published erratum appears in Lancet 1992, 340, 1110. Cosenza, H. (1976). Detection of anti-idiotype reactive cells in the response to phosphorylcholine. Eur. J. Immunol. 6, 114±116. Coutinho, A. (1995). The network theory: 21 years later [editorial]. Scand. J. Immunol. 42, 3±8. Coutinho, A., Kazatchkine, M. D., and Avrameas, S. (1995). Natural autoantibodies. Curr. Opin. Immunol. 7, 812±817. Cull, G., Durrant, L., Stainer, C., Haynes, A., and Russell, N. (1999). Generation of anti-idiotype immune responses following vaccination with idiotype-protein pulsed dendritic cells in myeloma. Br. J. Haematol. 107, 648±655. Cumano, A., and Rajewsky, K. (1985). Structure of primary anti-(4-hydroxy-3-nitrophenyl) acetyl (NP) antibodies in normal and idiotypically suppressed C57BL/6 mice. Eur. J. Immunol. 15, 512±520. de Weck, A., Derer, M., Mayer, P., and Stumper, B. (1995). Allergy starts in early life: Clinical and experimental evidence. In ``Prediction and Prevention of Childhood Allergy'' (S. Sasaki, T. Miyamoto, and J. Hopkin, Eds.), pp. 181±197. Churchill±Livingstone, Osaka. Dembic, Z., Rottingen, J. A., Dellacasagrande, J., Schenck, K., and Bogen, B. (2001). Phagocytic dendritic cells from myelomas activate tumor-speci®c T cells at a single cell level. Blood 97, 2808±2814. Dembic, Z., Schenck, K., and Bogen, B. (2000). Dendritic cells puri®ed from myeloma are primed with tumor-speci®c antigen (idiotype) and activate CD4‡ T cells. Proc. Natl. Acad. Sci. USA. Eichmann, K. (1972). Idiotypic identity of antibodies to streptococcal carbohydrate in inbred mice. Eur. J. Immunol. 2, 301±307. Eichmann, K., Coutinho, A., and Melchers, F. (1977). Absolute frequencies of lipopolysaccharidereactive B cells producing A5A idiotype in unprimed, streptococcal A carbohydrate-primed, antiA5A idiotype-sensitized and anti-A5A idiotype-suppressed A/J mice. J. Exp. Med. 146, 1436±1449. Eichmann, K., and Rajewsky, K. (1975). Induction of T and B cell immunity by anti-idiotypic antibody. Eur. J. Immunol. 5, 661±666. Elias, D., Tikochinski, Y., Frankel, G., and Cohen, I. R. (1999). Regulation of NOD mouse autoimmune diabetes by T cells that recognize a TCR CDR3 peptide. Int. Immunol. 11, 957±966. Elliott, M., and Kearney, J. F. (1992). Idiotypic regulation of development of the B-cell repertoire. Ann. N. Y. Acad. Sci. 651, 336±345.

232

hilmar lemke and hans lange

Ertl, H. C., Greene, M. I., Noseworthy, J. H., Fields, B. N., Nepom, J. T., Spriggs, D. R., and Finberg, R. W. (1982). Identi®cation of idiotypic receptors on reovirus-speci®c cytolytic T cells. Proc. Natl. Acad. Sci. USA 79, 7479±7483. Eyerman, M. C., and Wysocki, L. (1994). T cell recognition of somatically-generated Ab diversity. J. Immunol. 152, 1569±1577. Fagerberg, J., Ragnhammar, P., Liljefors, M., Hjelm, A. L., Mellstedt, H., and Frodin, J. E. (1996). Humoral anti-idiotypic and anti-anti-idiotypic immune response in cancer patients treated with monoclonal antibody 17-1A. Cancer Immunol. Immunother. 42, 81±87. Fagerberg, J., Yi, Q., Gigliotti, D., Harmenberg, U., Ruden, U., Persson, B., Osterborg, A., and Mellstedt, H. (1999). T-cell-epitope mapping of the idiotypic monoclonal IgG heavy and light chains in multiple myeloma. Int. J. Cancer 80, 671±680. Fernandez, C., and Moller, G. (1979). Antigen-induced strain-speci®c autoantiidiotypic antibodies modulate the immune response to dextran B 512. Proc. Natl. Acad. Sci. USA 76, 5944±7. Fields, B. A., Goldbaum, F. A., Ysern, X., Poljak, R. J., and Mariuzza, R. A. (1995). Molecular basis of antigen mimicry by an anti-idiotope. Nature 374, 739±742. Forsyth, R. B., and Hoffmann, G. W. (1990). A study of auto-anti-idiotypes to BSA. J. Immunol. 145, 215±223. Fougereau, M., Corbet, S., Mazza, G., Nahmias, C., and Schiff, C. (1990a). Genetic and structural basis of idiotype network: The `GAT' model. Chem. Immunol. 48, 14±29. Fougereau, M., Mazza, G., Nahmias, C., Gonzales, R., Avtameas, S., Strosberg, A. D., and Schiff, C. (1990b). In ``Idiotype Networks in Biology and Medicine'' (A. Osterhaus and F. Uytdehaag, Eds.), pp. 23±26. Excerpta Medica, Amsterdam. Fricke, H., Mendlovic, S., Blank, M., Shoenfeld, Y., Ben-Bassat, M., and Mozes, E. (1991). Idiotype speci®c T-cell lines inducing experimental systemic lupus erythematosus in mice. Immunology 73, 421±427. Frodin, J. E., Faxas, M. E., Hagstrom, B., Lefvert, A. K., Masucci, G., Nilsson, B., Steinitz, M., Unger, P., and Mellstedt, H. (1991). Induction of anti-idiotypic (ab2) and anti-anti-idiotypic (ab3) antibodies in patients treated with the mouse monoclonal antibody 17-1A (ab1): Relation to the clinical outcomeÐAn important antitumoral effector function? Hybridoma 10, 421±431. Geha, R. S. (1983). Presence of circulating anti-idiotype-bearing cells after booster immunization with tetanus toxoid (TT) and inhibition of anti-TT antibody synthesis by auto-antiidiotypic antibody. J. Immunol. 130, 1634±1639. Gleason, K., and Kohler, H. (1982). Regulatory idiotypes: T helper cells recognize a shared VH idiotope on phosphorylcholine-speci®c antibodies. J. Exp. Med. 156, 539±549. Goidl, E. A., Schrater, A. F., Thorbecke, G. H., and Siskind, G. W. (1980). Production of auto-antiidiotypic antibody during the normal immune response. IV. Studies of the primary and secondary responses to thymus-dependent and -independent antigens. Eur. J. Immunol. 10, 810±814. Gorczynski, R., Baumal, R., Boulanger, M., and Marks, A. (1986). Protection of adult BALB/c mice against MOPC 315 myeloma by neonatal administration of monoclonal anti-idiotypic antibody to MOPC 315 IgA. J. Natl. Cancer Inst. 77, 801±807. Gridley, T., Margolies, M. N., and Gefter, M. L. (1985). The association of various D elements with a single-immunoglobulin VH gene segment: In¯uence on the expression of a major cross-reactive idiotype. J. Immunol. 134, 1236±1244. Haba, S., and Nisonoff, A. (1988). Immunological responsiveness of neonatal A/J mice to isotypic determinants of syngeneic IgE. J. Exp. Med. 168, 713±724. Haba, S., and Nisonoff, A. (1990). Inhibition of IgE synthesis by anti-IgE: Role in long-term inhibition of IgE synthesis by neonatally administered soluble IgE. Proc. Natl. Acad. Sci. USA 87, 3363±3367. Haba, S., and Nisonoff, A. (1994). Effects of syngeneic anti-IgE antibodies on the development of IgE memory and on the secondary IgE response. J. Immunol. 152, 51±57.

idiotypic network functions

233

Haba, S., and Nisonoff, A. (1995). Prolongation of the responsiveness of newborn mice to syngeneic IgE by inhibition of IgE synthesis. Immunol. Lett. 47, 205±208. Hammerling, G. J., Black, S. J., Berek, C., Eichmann, K., and Rajewsky, K. (1976). Idiotypic analysis of lymphocytes in vitro. II. Genetic control of T-helper cell responsiveness to anti-idiotypic antibody. J. Exp. Med. 143, 861±869. Hannestad, K., Kristoffersen, G., and Briand, J. P. (1986). The T lymphocyte response to syngeneic lambda 2 light chain idiotopes: Signi®cance of individual amino acids revealed by variant lambda 2 chains and idiotope-mimicking chemically synthesized peptides Published erratum appears in Eur. J. Immunol. 1986, 6, 1470. Eur. J. Immunol. 16, 889±893. Hebert, J., Bernier, D., Boutin, Y., Jobin, M., and Mourad, W. (1990). Generation of anti-idiotypic and anti-anti-idiotypic monoclonal antibodies in the same fusion: Support of Jerne's Network Theory. J. Immunol. 144, 4256±4261. Hiernaux, J., Bona, C., and Baker, P. J. (1981). Neonatal treatment with low doses of anti-idiotypic antibody leads to the expression of a silent clone. J. Exp. Med. 153, 1004±1008. Holmberg, D., and Coutinho, A. (1985). Natural antibodies and autoimmunity. Immunol. Today 6, 356±357. Holt, P. G. (1995). Early acquisition of sensitization in childhood asthma. Pediatr. Pulmonol. Suppl. 11, 44±46. Hong, J., Zang, Y. C., Tejada-Simon, M. V., Li, S., Rivera, V. M., Killian, J., and Zhang, J. Z. (2000). Reactivity and regulatory properties of human anti-idiotypic antibodies induced by T cell vaccination J. Immunol. 165, 6858±6864. Howell, M. D., Winters, S. T., Olee, T., Powell, H. C., Carlo, D. J., and Brostoff, S. W. (1989). Vaccination against experimental allergic encephalomyelitis with T cell receptor peptides. Science 246, 668±670. Ismaili, J., Brait, M., Leo, O., and Urbain, J. (1995). Assessment of a functional role of autoanti-idiotypes in idiotype dominance. Eur. J. Immunol. 25, 830±837. Jackson, S., and Mestecky, J. (1979). Presence of plasma cells binding autologous antibody during an immune response. J. Exp. Med. 150, 1265±127. Janeway, C. A., Jr., Broughton, B., Smith, L. A., Marion, T. N., and Bottomly, K. (1987). Direct receptor:receptor interactions between T and B lymphocytes: Idiotypic restriction in the antibody response to a cloned helper T cell receptor. J. Mol. Cell. Immunol. 3, 83±94. Janeway, C. A., Jr., Sakato, N., and Eisen, H. N. (1975). Recognition of immunoglobulin idiotypes by thymus-derived lymphocytes. Proc. Natl. Acad. Sci. USA 72, 2357±2360. Janeway, C. A., Rojo, J., Portoles, P., Dunn, E., and Bottomly, K. (1990). Idiotypic interactions between B- and T-lymphocyte receptors. In ``Idiotype Networks in Biology and Medicine'' (A. Osterhaus and F. UytedeHaag, Eds.), pp. 125±131. Excerpta Medica, Amsterdam. Jerne, N. K. (1974). Towards a network theory of the immune system. Ann. Immunol. (Paris) 125C, 373±389. Jerne, N. K. (1984). Idiotypic networks and other preconceived ideas. Immunol. Rev. 79, 5±24. Jerne, N. K. (1985). The generative grammar of the immune system. Science 229, 1057±1059. Jerne, N. K., Roland, J., and Cazenave, P. A. (1982). Recurrent idiotopes and internal images. EMBO J. 1, 243±247. Jeske, D., Milner, E. C., Leo, O., Moser, M., Marvel, J., Urbain, J., and Capra, J. D. (1986). Molecular mapping of idiotopes of anti-arsonate antibodies. J. Immunol. 136, 2568±2574. Jiang, H., Kashleva, H., Xu, L. X., Forman, J., Flaherty, L., Pernis, B., Braunstein, N. S., and Chess, L. (1998). T cell vaccination induces T cell receptor Vbeta-speci®c Qa-1-restricted regulatory CD8(‡) T cells. Proc. Natl. Acad. Sci. USA 95, 4533±4537. Jiang, H., Zhang, S. I., and Pernis, B. (1992). Role of CD8‡ T cells in murine experimental allergic encephalomyelitis. Science 256, 1213±1215.

234

hilmar lemke and hans lange

Jorgensen, T., Bogen, B., and Hannestad, K. (1983). T helper cells recognize an idiotope located on peptide 88-114/117 of the light chain variable domain of an isologous myeloma protein (315). J. Exp. Med. 158, 2183±2188. Jorgensen, T., and Hannestad, K. (1977). Speci®city of T and B lymphocytes for myeloma protein 315. Eur. J. Immunol. 7, 426±431. Jorgensen, T., and Hannestad, K. (1979). T helper lymphocytes recognize the VL domain of the isologous mouse myeloma protein 315. Scand. J. Immunol. 10, 317±323. Jorgensen, T., and Hannestad, K. (1982). Helper T cell recognition of the variable domains of a mouse myeloma protein (315): Effect of the major histocompatibility complex and domain conformation. J. Exp. Med. 155, 1587±1596. Julius, M. H., Augustin, A. A., and Cosenza, H. (1977a). Recognition of a naturally occurring idiotype by autologous T cells. Nature 265, 251±253. Julius, M. H., Cosenza, H., and Augustin, A. A. (1977b). Parallel expression of new idiotypes on T and B cells. Nature 267, 437±439. Julius, M. H., Cosenza, H., and Augustin, A. A. (1980). Enrichment of hapten-speci®c helper T cells using anti-immunoglobulin combining site antibodies. Eur. J. Immunol. 10, 112±116. Kearney, J. F., Pollok, B. A., and Stohrer, R. (1983). Analysis of idiotypic heterogeneity in the anti-a 1-3 dextran and anti-phosphorylcholine responses using monoclonal anti-idiotype antibodies. Ann. N. Y. Acad. Sci. 418, 151±170. Kearney, J. F., and Vakil, M. (1986). Idiotype-directed interactions during ontogeny play a major role in the establishment of the adult B cell repertoire. Immunol. Rev. 94, 39±50. Kelus, A. S., and Gell, P. G. H. (1968). Immunological analysis of rabbit anti-antibody systems. J. Exp. Med. 127, 215. Kim, Y. T., Deblasio, T., Thorbecke, G. J., Weksler, M. E., and Siskind, G. W. (1989). Production of auto-anti-idiotype antibody during the normal immune response. XIV. Evidence for the antigenindependent operation of the idiotype network. Immunology, 67, 191±196. King, C. A., Wills, M. R., Hamblin, T. J., and Stevenson, F. K. (1993). Idiotypic IgM on a B-cell surface requires processing for recognition by anti-idiotypic T cells. Cell. Immunol. 147, 411±424. Kluskens, L., and Kohler, H. (1974). Regulation of immune response by autogenous antibody against receptor. Proc. Natl. Acad. Sci. USA 71, 5083±5087. Kohler, H., Kieber-Emmons, T., Srinivasan, S., Kaveri, S., Morrow, W. J., Muller, S., Kang, C. Y., and Raychaudhuri, S. (1989). Revised immune network concepts. Clin. Immunol. Immunopathol. 52, 104±116. Krause, I., Blank, M., Kopolovic, J., Afek, A., Goldberg, I., Tomer, Y., and Shoenfeld, Y. (1995). Abrogation of experimental systemic lupus erythematosus and primary antiphospholipid syndrome with intravenous g globulin. J. Rheumatol. 22, 1068±1074. Krause, I., Tomer, Y., Elias, D., Blank, M., Gilburd, B., Cohen, I. R., and Shoenfeld, Y. (1999). Inhibition of diabetes in NOD mice by idiotypic induction of SLE. J. Autoimmun. 13, 49±55. Kristoffersen, G., Hannestad, K., and Hansen, T. (1987). Two M315 idiotopes de®ned by isologous monoclonal antibodies: One depends on germline and the other on mutated murine lambda 2 light chain sequences. Scand. J. Immunol. 26, 535±546. Kunkel, H. G., Mannik, M., and Williams, R. C. (1963). Individual antigenic speci®cities of isolated antibodies. Science 140, 1218±1219. Kuribayashi, K., Tanaka, C., Matsubayashi, Y., Masuda, T., Udono, H., Abe, M., Nakayama, E., and Shiku, H. (1988). Anti-idiotypic antibodies against UV-induced tumor-speci®c CTL clones: Preparation in syngeneic combination. J. Immunol. 141, 4074±4080. Kwak, L. W., Campbell, M. J., Zelenetz, A. D., and Levy, R. (1990). Combined syngeneic bone marrow transplantation and immunotherapy of a murine B-cell lymphoma: active immunization with tumor-derived idiotypic immunoglobulin. Blood 76, 2411±2417.

idiotypic network functions

235

Kwak, L. W., Pennington, R., and Longo, D. L. (1996a). Active immunization of murine allogeneic bone marrow transplant donors with B-cell tumor-derived idiotype: A strategy for enhancing the speci®c antitumor effect of marrow grafts. Blood 87, 3053±3060. Kwak, L. W., Thielemans, K., and Massaia, M. (1999). Idiotypic vaccination as therapy for multiple myeloma. Semin. Hematol. 36, 34±37. Kwak, L. W., Young, H. A., Pennington, R. W., and Weeks, S. D. (1996b). Vaccination with syngeneic, lymphoma-derived immunoglobulin idiotype combined with granulocyte/macrophage colony-stimulating factor primes mice for a protective T-cell response. Proc. Natl. Acad. Sci. USA 93, 10972±10977. Lange, H., Kobarg, J., Yazynin, S., Solterbeck, M., Henningsen, M., Hansen, H., and Lemke, H. (1999). Genetic analysis of the maternally induced af®nity enhancement in the non-Ox1 idiotypic antibody repertoire of the primary immune response to 2-phenyloxazolone. Scand. J. Immunol. 49, 55±66. Lange, H., and Lemke, H. (1996). Induction of a non-oscillating, long-lasting humoral immune response to an internal network antigen. Int. Immunol. 8, 683±688. Lange, H., Solterbeck, M., Berek, C., and Lemke, H. (1996). Correlation between immune maturation and idiotypic network recognition. Eur. J. Immunol. 26, 2234±2242. Lauritzsen, G. F., and Bogen, B. (1991). Idiotype-speci®c, major histocompatibility complex restricted T cells are of both Th1 and Th2 type. Scand. J. Immunol. 33, 647±656. Lauritzsen, G. F., Weiss, S., and Bogen, B. (1993). Anti-tumour activity of idiotype-speci®c, MHCrestricted Th1 and Th2 clones in vitro and in vivo. Scand. J. Immunol. 37, 77±85. Lauritzsen, G. F., Weiss, S., Dembic, Z., and Bogen, B. (1994). Naive idiotype-speci®c CD4‡ T cells and immunosurveillance of B-cell tumors. Proc. Natl. Acad. Sci. USA 91, 5700±5704. Leclercq, L., Mazie, J. C., Somme, G., and Theze, J. (1982). Monoclonal anti-GAT antibodies with different ®ne speci®cities express the same public idiotype. Mol. Immunol. 19, 1001±1009. Leger, O., and Dean, C. J. (1991). Characterization of syngeneic rat monoclonal antibodies to the HSN tumor using syngeneic monoclonal anti-idiotopic antibodies. J. Immunol. 146, 2879±2886. Lemke, H., and Lange, H. (1999). Is there a maternally induced immunological imprinting phase a la Konrad Lorenz? Scand. J. Immunol. 50, 348±354. Lemke, H., Lange, H., and Berek, C. (1994). Maternal immunization modulates the primary immune response to 2-phenyl-oxazolone in BALB/c mice. Eur. J. Immunol. 24, 3025±3030. Levitsky, H. I., Montgomery, J., Ahmadzadeh, M., Staveley-O'Carroll, K., Guarnieri, F., Longo, D. L., and Kwak, L. W. (1996). Immunization with granulocyte-macrophage colonystimulating factor-transduced, but not B7-1-transduced, lymphoma cells primes idiotype-speci®c T cells and generates potent systemic antitumor immunity. J. Immunol. 156, 3858±3865. Lider, O., Reshef, T., Beraud, E., Ben-Nun, A., and Cohen, I. R. (1988). Anti-idiotypic network induced by T cell vaccination against experimental autoimmune encephalomyelitis. Science 239, 181±183. Liu, K. J., Parikh, V. S., Tucker, P. W., and Kim, B. S. (1994). Surface immunoglobulins mediate ef®cient transport of antigen to lysosomal compartments resulting in enhanced speci®c antigen presentation by B cells. Eur. J. Immunol. 24, 2755±2760. Ma, J., Samuel, J., Kwon, G. S., Noujaim, A. A., and Madiyalakan, R. (1998). Induction of antiidiotypic humoral and cellular immune responses by a murine monoclonal antibody recognizing the ovarian carcinoma antigen CA125 encapsulated in biodegradable microspheres. Cancer Immunol. Immunother. 47, 13±20. Macias, A., Arce, S., Leon, J., Mustelier, G., Bombino, G., Domarco, A., Perez, R., and Lage, A. (1999). Novel cross-reactive anti-idiotype antibodies with properties close to the human intravenous immunoglobulin (IVIg). Hybridoma 18, 263±272.

236

hilmar lemke and hans lange

Magliani, W., Polonelli, L., Conti, S., Salati, A., Rocca, P. F., Cusumano, V., Mancuso, G., and Teti, G. (1998). Neonatal mouse immunity against group B streptococcal infection by maternal vaccination with recombinant anti-idiotypes. Nat. Med. 4, 705±709. Manz, R. A., Lohning, M., Cassese, G., Thiel, A., and Radbruch, A. (1998). Survival of long-lived plasma cells is independent of antigen. Int. Immunol. 10, 1703±1711. Manz, R. A., Thiel, A., and Radbruch, A. (1997). Lifetime of plasma cells in the bone marrow. Nature 388, 133±134. Marcos, M. A., De la Hera, A., Pereira, P., Marquez, C., Toribio, M., Coutinho, A., and Martinez, C. (1988). B cell participation in the recursive selection of T cell repertoires. Eur. J. Immunol. 18, 1015±1020. Marshall, E. (1996). Disputed results now just a footnote [news]. Science 273, 174±175. Martinez, C., Bernabe, R. R., de la Hera, A., Pereira, P., Cazenave, P. A., and Coutinho, A., (1985). Establishment of idiotypic helper T-cell repertoires early in life. Nature 317, 721±723. Martinez, C., Bragado, R., de la Hera, A., Toribio, M. L., Marcos, M. A., Bandeira, A., Pereira, P., and Coutinho, A. (1986a). Functional and biochemical evidence for the recognition of T cell receptors by monoclonal antibodies to an immunoglobulin idiotype. J. Mol. Cell. Immunol. 2, 307±313. Martinez, C., Coutinho, A., Bandeira, A., de la Hera, A., Toribio, M., Marcos, M. A., and Pereira, P. (1987a). Sporadic idiotypic cross-reactivities between antibodies and T helper cells: One example of aleatory expression of T cell idiotypes. J. Mol. Cell. Immunol. 3, 21±28. Martinez, C., Pereira, P., De la Hera, A., Toribio, M., Marcos, M. R., Marquez, C., and Coutinho, A. (1987b). A common idiotope on T cell receptors and antibodies expressed in the absence of network selection. Eur. J. Immunol. 17, 1391±1394. Martinez, C., Pereira, P., Toribio, M. L., Marcos, M. A., Bandeira, A., de la Hera, A., Marquez, C., Cazenave, P. A., and Coutinho, A. (1988). The participation of B cells and antibodies in the selection and maintenance of T cell repertoires. Immunol. Rev. 101, 191±215. Martinez, C., Toribio, M. L., De la Hera, A., Cazenave, P. A., and Coutinho, A. (1986b). Maternal transmission of idiotypic network interactions selecting available T cell repertoires. Eur. J. Immunol. 16, 1445±1447. Masaki, H., Irimajiri, K., Horiuchi, A., Yamaguchi, J., Toyosaki, T., Suzuki, R., and Kurane, I. (1996). Establishment and characterization of an anti-idiotypic CD4‡ CD8 T cell line to murine anti-a (1 ! 3) dextran antibody. Cell. Immunol. 174, 180±189. Mazel, S. M., Rudensky, A. Y., and Yurin, V. L. (1990). Immunoglobulin-speci®c T-B cell interaction. III. B cell activation by immunoglobulin-recognizing T cell clones. Eur. J. Immunol. 20, 833±839. McMillan, S., Seiden, M. V., Houghten, R. A., Clevinger, B., Davie, J. M., and Lerner, R. A. (1983). Synthetic idiotypes: The third hypervariable region of murine anti-dextran antibodies. Cell 35, 859±863. McNamara, M., and Kohler, H. (1984). Regulatory idiotopes: Induction of idiotype-recognizing helper T cells by free light and heavy chains. J. Exp. Med. 159, 623±628. Meek, K., and Capra, J. D. (1990). Structural implications of AB2s: Are novel D segments involved with anti-idiotypic speci®city? Chem. Immunol. 48, 30±48. Meek, K., Hasemann, C., Pollok, B., Alkan, S. S., Brait, M., Slaoui, M., Urbain, J., and Capra, J. D. (1989). Structural characterization of antiidiotypic antibodies: Evidence that Ab2s are derived from the germline differently than Abls. J. Exp. Med. 169, 519±533. Meek, K., Jeske, D., Alkan, S., Urbain, J., and Capra, J. D. (1987). Structural characterization of syngeneic and allogeneic anti-idiotypic antibodies in the anti-arsonate system. Monogr. Allergy 22, 109±119. MoÈller, G. (Ed.) (1986). ``Anti-Idiotypic Antibodies as Immunogens.'' Munksgaard, Copenhagen.

idiotypic network functions

237

Montesano, M. A., Colley, D. G., Eloi-Santos, S., Freeman, G., and Secor, W. E. (1999a). Neonatal idiotypic exposure alters subsequent cytokine, pathology, and survival patterns in experimental Schistosoma mansoni infections. J. Exp. Med. 189, 637±645. Montesano, M. A., Colley, D. G., Freeman, G. L., Jr., and Secor, W. E. (1999b). Neonatal exposure to idiotype induces Schistosoma mansoni egg antigen-speci®c cellular and humoral immune responses. J. Immunol. 163, 898±905. Moser, M., Leo, O., Hiernaux, J., and Urbain, J. (1983). Idiotypic manipulation in mice: BALB/c mice can express the crossreactive idiotype of A/J mice. Proc. Natl. Acad. Sci. USA 80, 4474±4478. Mouthon, L., and Kazatchkine, M. D. (1996). Autoantibodies in therapeutic preparations of human IgG (IVIg). In ``Autoantibodies'' (J. B. Peter, and Y. Shoenfeld, Eds.), pp. 91±95. Elsevier, Amsterdam. Munthe, L. A., Kyte, J. A., and Bogen, B. (1999). Resting small B cells present endogeneous immunoglobulin variable-region determinants to idiotope-speci®c CD4(‡) T cells in vivo. Eur. J. Immunol. 29, 4043±4052. Nelson, E. L., Li, X., Hsu, F. J., Kwak, L. W., Levy, R., Clayberger, C., and Krensky, A. M. (1996). Tumor-speci®c, cytotoxic T-lymphocyte response after idiotype vaccination for B-cell, nonHodgkin's lymphoma. Blood 88, 580±589. Nicoletti, C., and Cerny, J. (1992). A study of autologous anti-idiotypic antibody-forming cells in mice of different ages and genetic backgrounds. Cell. Immunol. 144, 332±346. Nishimura, M., Sakai, K., Akaza, T., Mitomi, Y., Nieda, M., Minami, M., and Juji, T. (1992). Antiidiotypic antibody to T-cell receptor in multiply transfused patients may play a role in resistance to graft-versus-host disease. Transfusion 32, 719±728. Nisonoff, A. (1997). In ``Idiotypes In Medicine: Autoimmunity, Infection and Cancer.'' (Y. Shoenfeld, R. C. Kennedy, and S. Ferrone, Eds.). Elsevier, Amsterdam. Nisonoff, A., and Lamoyi, E. (1981). Implications of the presence of an internal image of the antigen in anti-idiotypic antibodies: Possible application to vaccine production. Clin. Immunol. Immunopathol. 21, 397±406. Novotny, J., Handschumacher, M., and Haber, E. (1986). Location of antigenic epitopes on antibody molecules. J. Mol. Biol. 189, 715±721. Okamoto, Y., Tsutsumi, H., Kumar, N. S., and Ogra, P. L. (1989). Effect of breast feeding on the development of anti-idiotype antibody response to F glycoprotein of respiratory syncytial virus in infant mice after post-partum maternal immunization. J. Immunol. 142, 2507±2012. Osterborg, A., Yi, Q., Bergenbrant, S., Holm, G., Lefvert, A. K., and Mellstedt, H. (1995). Idiotypespeci®c T cells in multiple myeloma stage I: An evaluation by four different functional tests. Br. J. Haematol. 89, 110±116. Osterhaus, A. D., Rimmelzwaan, G. F., Weijer, K., and Uytdehaag, A. G. (1989). Induction of antiviral immune response by immunization with monoclonal anti-idiotype antibodies directed to private idiotopes. Viral Immunol. 2, 255±262. Osterroth, F., Garbe, A., Fisch, P., and Veelken, H. (2000). Stimulation of cytotoxic T cells against idiotype immunoglobulin of malignant lymphoma with protein-pulsed or idiotype-transduced dendritic cells. Blood 95, 1342±1349. Oudin, J., and Michel, M. (1963). Une nuovelle forme d'allotypie des globulines du serum de lapin, apparement lieÂe aÁ la function et aÁ la speÂci®te des anticorps. C. R. Acad. Sci. (Paris) 257, 805±808. Pierce, S. K., Morris, J. F., Grusby, M. J., Kaumaya, P., van Buskirk, A., Srinivasan, M., Crump, B., and Smolenski, L. A. (1988). Antigen-presenting function of B lymphocytes. Immunol. Rev. 106, 149±180. Pierce, S. K., Speck, N. A., Gleason, K., Gearhart, P. J., and Kohler, H. (1981). Balb/c T cells have the potential to recognize the TEPC 15 prototype antibody and its somatic variants. J. Exp. Med. 154, 1178±1187.

238

hilmar lemke and hans lange

Pollok, B. A., and Kearney, J. F. (1984). Identi®cation and characterization of an apparent germline set of auto-anti-idiotypic regulatory B lymphocytes. J. Immunol. 132, 114±121. Pollok, B. A., Kearney, J. F., Vakil, M., and Perry, R. P. (1984). A biological consequence of variation in the site of D-JH gene rearrangement. Nature 311, 376±379. Powell, T. J., Jr., Spann, R., Vakil, M., Kearney, J. F., and Lamon, E. W. (1988). Activation of a functional idiotype network response by monoclonal antibody speci®c for a virus (M-MuLV)induced tumor antigen. J. Immunol. 140, 3266±3272. Radbruch, A., Zaiss, S., Kappen, C., Bruggemann, M., Beyreuther, K., and Rajewsky, K. (1985). Drastic change in idiotypic but not antigen-binding speci®city of an antibody by a single aminoacid substitution. Nature 315, 506±508. Rajewsky, K. (1983). Symmetry and asymmetry in idiotypic interactions. Ann. Immunol. (Paris), 134D, 133±141. Rajewsky, K., and Takemori, T. (1983). Genetics, expression, and function of idiotypes. Annu. Rev. Immunol. 1, 569±607. Rajewsky, K., Takemori, T., and Reth, M. (1981). Analysis and regulation of V gene expression by monoclonal antibodies. In ``Monoclonal Antibodies and T-Cell Hybridomas: Perspectives and Technical Advances'' (G. J., HaÈmmerling, U. HaÈmmerling, and J. F. Kearney, Eds.), pp. 399±409. Elsevier, Amsterdam. Rees, A. D., Praputpittaya, K., Scoging, A., Dobson, N., Ivanyi, J., Young, D., and Lamb, J. R. (1987). T-cell activation by anti-idiotypic antibody: Evidence for the internal image. Immunology 60, 389±393. Reth, M., Imanishi-Kari, T., and Rajewsky, K. (1979). Analysis of the repertoire of anti-(4-hydroxy3-nitrophenyl)acetyl (NP) antibodies in C57BL/6 mice by cell fusion. II. Characterization of idiotopes by monoclonal anti-idiotope antibodies. Eur. J. Immunol. 9, 1004±1013. Rodkey, L. S., Binion, S. B., Brown, J. C., and Adler, F. L. (1983). Natural regulation of antibody clones by auto-anti-idiotypic antibodies in rabbits. Ann. N. Y. Acad. Sci. 418, 16±25. Rosenberg, Y. J., and Chiller, J. M. (1979). Ability of antigen-speci®c helper cells to effect a classrestricted increase in total Ig-secreting cells in spleens after immunization with the antigen. J. Exp. Med. 150, 517±530. Rothstein, T. L., and Vastola, A. P. (1984). Homologous monoclonal antibodies induce idiotopespeci®c suppression in neonates through maternal in¯uence and in adults exposed during fetal and neonatal life. J. Immunol. 133, 1151±1154. Roubaty, C., Bedin, C., and Charreire, J. (1990). Prevention of experimental autoimmune thyroiditis through the anti-idiotypic network. J. Immunol. 144, 2167±2172. Rubinstein, L. J., Goldberg, B., Hiernaux, J., Stein, K. E., and Bona, C. A. (1983). Idiotypeantiidiotype regulation. V. The requirement for immunization with antigen or monoclonal antiidiotypic antibodies for the activation of b2 > 6 and b2 > 1 polyfructosan-reactive clones in BALB/c mice treated at birth with minute amounts of anti-A48 idiotype antibodies. J. Exp. Med. 158, 1129±1144. Rubinstein, L. J., Yeh, M., and Bona, C. A. (1982). Idiotype-anti-idiotype network. II. Activation of silent clones by treatment at birth with idiotypes is associated with the expansion of idiotypespeci®c helper T cells. J. Exp. Med. 156, 506±521. Rudensky, A. Y., Mazel, S. M., Blechman, J. M., and Yurin, V. L. (1990). Immunoglobulin-speci®c T±B cell interaction. IV. B cell presentation of idiotypic determinant(s) of monoclonal antisurface immunoglobulin antibody to idiotope-recognizing helper T clones. Eur. J. Immunol. 20, 1691±1696. Rudensky, A. Y., and Yurin, V. L. (1989). Immunoglobulin-speci®c T±B cell interaction. I. Presentation of self immunoglobulin determinants by B lymphocytes. Eur. J. Immunol. 19, 1677±1683. Rudikoff, S., Pawlita, M., Pumphrey, J., Mushinski, E., and Potter, M. (1983). Galactan-binding antibodies: Diversity and structure of idiotypes. J. Exp. Med. 158, 1385±400.

idiotypic network functions

239

Rueff-Juy, D., Sanchez, P., Faure, M., Drapier, A. M., and Cazenave, P. A. (1995). Emergence in C kappa knockout mice of a diverse cytotoxic T lymphocyte repertoire that recognizes a single peptide from the immunoglobulin constant kappa light chain region. Eur. J. Immunol. 25, 2752±2756. Ruiz, P. J., Wolkowicz, R., Waisman, A., Hirschberg, D. L., Carmi, P., Erez, N., Garren, H., Herkel, J., Karpuj, M., Steinman, L., Rotter, V., and Cohen, I. R. (1998). Idiotypic immunization induces immunity to mutated p53 and tumor rejection. Nat. Med. 4, 710±712. Sacks, D. L., Esser, K. M., and Sher, A. (1982). Immunization of mice against African trypanosomiasis using anti-idiotypic antibodies. J. Exp. Med. 155, 1108±1119. Saeki, Y., Chen, J. J., Shi, L. F., Okuda, Y., and Kohler, H. (1990). Idiotype-speci®c T helper clones recognize a variable H chain determinant. J. Immunol. 144, 1625±1628. Saint-Remy, J. M., Lebecque, S. J., Lebrun, P. M., and Jacquemin, M. G. (1988). Human immune response to allergens of house dust mite, Dermatophagoides pteronyssinus. IV. Occurrence of natural autologous anti-idiotypic antibodies. Eur. J. Immunol. 18, 83±87. Sakato, N., and Eisen, H. N. (1975). Antibodies to idiotypes of isologous immunoglobulins. J. Exp. Med. 141, 1411±1426. Sakato, N., Rugdech, P., Yoda, T., Ota, A., Zhao, Y., Semma, M., Suzuki, Y., and Fujio, H. (1990). Activation of idiotype-speci®c CD4‡ T-cell line: Cellular processing of exogeneous self-immunoglobulin. Immunology 71, 153±157. Schrater, A. F., Goidl, E. A., Thorbecke, G. J., and Siskind, G. W. (1979). Production of auto-antiidiotypic antibody during the normal immune response to TNP-®coll. III. Absence in nu/nu mice: Evidence for T-cell dependence of the anti-idiotypic-antibody response. J. Exp. Med. 150, 808±817. Schuler, W., Weiler, E., and Weiler, I. J. (1981). Biological and serological comparison of syngeneic and allogeneic anti-idiotypic antibodies. Mol. Immunol. 18, 1095±1105. Sharpe, A. H., Gaulton, G. N., Ertl, H. C., Finberg, R. W., McDade, K. K., Fields, B. N., and Greene, M. I. (1985). Cell receptors for the mammalian reovirus. IV. Reovirus-speci®c cytolytic T cell lines that have idiotypic receptors recognize anti-idiotypic B cell hybridomas. J. Immunol. 134, 2702±2706. Shoenfeld, Y., Kennedy, R. C., and Ferrone, S. (Eds.) (1997). Idiotypes in Medicine: Autoimmunity, Infection and Cancer. Elsevier, Amsterdam. Somme, G., Roth, C., Mazie, J. C., Salem, P., and Theze, J. (1983). Public and individual idiotopes in the anti-poly(Glu60, Ala30, Tyr10) response: Analysis by monoclonal antibodies. Eur. J. Immunol. 13, 1023±1030. Stein, K. E. (1985). Network regulation of the immune response to bacterial polysaccharide antigens. Curr. Top. Microbiol. Immunol. 119, 57±74. Stein, K. E., and SoÈderstroÈm, T. (1984). Neonatal administration of idiotype or antiidiotype primes for protection against Escherichia coli K13 infection in mice. J. Exp. Med. 160, 1001±1011. Strayer, D. S., Lee, W. M., Rowley, D. A., and Kohler, H. (1975). Anti-receptor antibody. II. Induction of long-term unresponsiveness in neonatal mice. J. Immunol. 114, 728±733. Strickland, F. M., Gleason, J. T., and Cerny, J. (1987). Serologic and molecular characterization of the T15 idiotype. I. Topologic mapping of idiotopes on TEPC15. Mol. Immunol. 24, 631±635. Su, S., Ward, M. M., Apicella, M. A., and Ward, R. E. (1992). A nontoxic, idiotope vaccine against gram-negative bacterial infections. J. Immunol. 148, 234±238. Sun, D., Ben-Nun, A., and Wekerle, H. (1988). Regulatory circuits in autoimmunity: Recruitment of counter-regulatory CD8‡ T cells by encephalitogenic CD4‡ T line cells. Eur. J. Immunol. 18, 1993±1999. Tikochinski, Y., Elias, D., Steeg, C., Marcus, H., Kantorowitz, M., Reshef, T., Ablamunits, V., Cohen, I. R., and Friedmann, A. (1999). A shared TCR CDR3 sequence in NOD mouse autoimmune diabetes. Int. Immunol. 11, 951±956.

240

hilmar lemke and hans lange

Tite, J. P., Kaye, J., Saizawa, K. M., Ming, J., Katz, M. E., Smith, L. A., and Janeway, C. A., Jr. (1986). Direct interactions between B and T lymphocytes bearing complementary receptors. J. Exp. Med. 163, 189±202. Urbain, J., Francotte, M., Franssen, J. D., Hiernaux, J., Leo, O., Moser, M., Slaoui, M., UrbainVansanten, G., Van Acker, A., and Wikler, M. (1983). From clonal selection to immune networks: Induction of silent idiotypes. Ann. N. Y. Acad. Sci. 418, 1±8. Urbain, J., Wikler, M., Franssen, J. D., and Collignon, C. (1977). Idiotypic regulation of the immune system by the induction of antibodies against anti-idiotypic antibodies. Proc. Natl. Acad. Sci. USA 74, 5126±5130. Urbain-Vansanten, G. (1970). Concomitant synthesis, in separate cells, of non-reactive immunoglobulins and speci®c antibodies after immunization with tobacco mosaic virus. Immunology 19, 783±797. UytdeHaag, F., Claassen, I., Bunschoten, H., Loggen, H., Ottenhoff, T., Teeuwsen, V., and Osterhaus, A. (1987). Human anti-idiotypic T lymphocyte clones are activated by autologous antirabies virus antibodies presented in association with HLA-DQ molecules. J. Mol. Cell. Immunol. 3, 145±155. Vakil, M., and Kearney, J. F. (1988). Regulatory in¯uences of neonatal multispeci®c antibodies on the developing B cell repertoire. Int. Rev. Immunol. 3, 117±131. Vakil, M., Sauter, H., Paige, C., and Kearney, J. F. (1986). In vivo suppression of perinatal multispeci®c B cells results in a distortion of the adult B cell repertoire. Eur. J. Immunol. 16, 1159±1165. Vandenbark, A. A., Hashim, G., and Offner, H. (1989). Immunization with a synthetic T-cell receptor V-region peptide protects against experimental autoimmune encephalomyelitis. Nature 341, 541±544. van Schooten, W. C., Devereux, D., Ho, C. H., Quan, J., Aguilar, B. A., and Rust, C. J. (1994). Jointderived T cells in rheumatoid arthritis react with self-immunoglobulin heavy chains or immunoglobulin-binding proteins that copurify with immunoglobulin. Eur. J. Immunol. 24, 93±98. Van Vollenhoven, R. F., Thorbecke, G. J., and Siskind, G. W. (1990). Exposure of female mice to type II collagen reduces susceptibility to collagen-induced arthritis in offspring. Eur. J. Immunol. 20, 2149±2152. Varela, F. J., and Coutinho, A. (1991). Second generation immune networks. Immunol. Today 12, 159±166. Victor, C., Bona, C., and Pernis, B. (1983). Idiotypes on B lymphocytes: Association with immunoglobulins. J. Immunol. 130, 1819±1825. Weiler, I. J., Weiler, E., Sprenger, R., and Cosenza, H. (1977). Idiotype suppression by maternal in¯uence. Eur. J. Immunol. 7, 591±597. Weiss, S., and Bogen, B. (1989). B-lymphoma cells process and present their endogenous immunoglobulin to major histocompatibility complex-restricted T cells. Proc. Natl. Acad. Sci. USA 86, 282±286. Wekerle, H. (1991). Immunopathogenesis of multiple sclerosis. Acta Neurol. (Napoli) 13, 197±204. Wen, Y., and Lim, S. H. (1999). Different properties of T-cell epitopes within complementaritydetermining regions 1 and 2 of idiotypic VH in B-lymphoma. Scand. J. Immunol. 50, 296±301. Wen, Y. J., and Lim, S. H. (1997). T cells recognize the VH complementarity-determining region 3 of the idiotypic protein of B cell non-Hodgkin's lymphoma. Eur. J. Immunol. 27, 1043±1047. Wen, Y. J., and Lim, S. H. (1998). In-vivo immune responses to idiotypic VH complementaritydetermining region 3 peptide vaccination in B-cell non-Hodgkin's lymphoma. Br. J. Haematol. 103, 663±668. Wen, Y. J., Ling, M., and Lim, S. H. (1998). Immunogenicity and cross-reactivity with idiotypic IgA of VH CDR3 peptide in multiple myeloma. Br. J. Haematol. 100, 464±468.

idiotypic network functions

241

Wettendorff, M., Iliopoulos, D., Tempero, M., Kay, D., DeFreitas, E., Koprowski, H., and Herlyn, D. (1989). Idiotypic cascades in cancer patients treated with monoclonal antibody CO17-1A. Proc. Natl. Acad. Sci. USA 86, 3787±3791. Whittum-Hudson, J. A., An, L. L., Saltzman, W. M., Prendergast, R. A., and MacDonald, A. B. (1996). Oral immunization with an anti-idiotypic antibody to the exoglycolipid antigen protects against experimental Chlamydia trachomatis infection. Nat. Med. 2, 1116±1121. Wikler, M., Demeur, C., Dewasme, G., and Urbain, J. (1980). Immunoregulatory role of maternal idiotypes: Ontogeny of immune networks. J. Exp. Med. 152, 1024±1035. Williams, W. M., Staines, N. A., Muller, S., and Isenberg, D. A. (1995). Human T cell responses to autoantibody variable region peptides. Lupus 4, 464±471. Wilson, A., George, A. J., King, C. A., and Stevenson, F. K. (1990). Recognition of a B cell lymphoma by anti-idiotypic T cells. J. Immunol. 145, 3937±3943. Yamamura, T., Kondo, T., Sakanaka, S., Kozovska, M., Geng, T. C., Takahashi, K., and Tabira, T. (1994). Analysis of T cell antigen receptors of myelin basic protein speci®c T cells in SJL/J mice demonstrates an alpha chain CDR3 motif associated with encephalitogenic T cells. Int. Immunol. 6, 947±954. Yang, Y. F., and Thanavala, Y. (1995). A comparison of the antibody and T cell response elicited by internal image and noninternal image anti-idiotypes. Clin. Immunol. Immunopathol. 75, 154±158. Yi, Q., Eriksson, I., He, W., Holm, G., Mellstedt, H., and Osterborg, A. (1997). Idiotype-speci®c T lymphocytes in monoclonal gammopathies: Evidence for the presence of CD4‡ and CD8‡ subsets. Br. J. Haematol. 96, 338±345. Yi, Q., Holm, G., and Lefvert, A. K. (1996). Idiotype-induced T cell stimulation requires antigen presentation in association with HLA-DR molecules. Clin. Exp. Immunol. 104, 359±365. Yi, Q., and Lefvert, A. K. (1992). Idiotypic and antiidiotypic T and B lymphocytes in myasthenia gravis. J. Immunol. 149, 3423±3426. Yi, Q., and Osterborg, A. (1996). Idiotype-speci®c T cells in multiple myeloma: Targets for an immunotherapeutic intervention? Med. Oncol. 13, 1±7. Yurin, V. L., Rudensky, A. Y., Mazel, S. M., and Blechman, J. M. (1989). Immunoglobulin-speci®c T±B cell interaction. II. T cell clones recognize the processed form of B cells' own surface immunoglobulin in the context of the major histocompatibility complex class II molecule. Eur. J. Immunol. 19, 1685±1691. Zaghouani, H., Kuzo, Y., Kuzo, H., Mann, N., Daian, C., and Bona, C. (1993a). Engineered immunoglobulin molecules as vehicles for T cell epitopes. Int. Rev. Immunol. 10, 265±278. Zaghouani, H., Steinman, R., Nonacs, R., Shah, H., Gerhard, W., and Bona, C. (1993b) Presentation of a viral T cell epitope expressed in the CDR3 region of a self immunoglobulin molecule. Science 259, 224±227. Zang, Y. C., Hong, J., Rivera, V. M., Killian, J., and Zhang, J. Z. (2000a). Preferential recognition of TCR hypervariable regions by human anti-idiotypic T cells induced by T cell vaccination. J. Immunol. 164, 4011±4017. Zang, Y. C., Hong, J., Tejada-Simon, M. V., Li, S., Rivera, V. M., Killian, J. M., and Zhang, J. Z. (2000b). Th2 immune regulation induced by T cell vaccination in patients with multiple sclerosis. Eur. J. Immunol. 30, 908±913. Zhang, J., Medaer, R., Stinissen, P., Ha¯er, D., and Raus, J. (1993). MHC-restricted depletion of human myelin basic protein-reactive T cells by T cell vaccination. Science 261, 1451±1454. Zhang, X., Smith, D. S., Guth, A., and Wysocki, L. J. (2001). A receptor presentation hypothesis for T cell help that recruits autoreactive B cells. J. Immunol. 166, 1562±1571. Zhou, S. R., and Whitaker, J. N. (1993). Speci®c modulation of T cells and murine experimental allergic encephalomyelitis by monoclonal anti-idiotypic antibodies. J. Immunol. 150, 1629±1642.

This Page Intentionally Left Blank

ADVANCES IN IMMUNOLOGY, VOL. 80

The Aging of the Immune System B.GRUBECK-LOEBENSTEIN* AND G.WICK*,À *Institute for Biomedical Aging Research of the Austrian Academy of Sciences, À

and the Institute for Pathophysiology, University of Innsbruck Medical School, A-6020 Innsbruck, Austria

I. The Biological Aging Process A. General Introduction In spite of the tragic AIDS epidemic, the deadly effects of cardiovascular diseases and of tumors, and the devastating results of drug abuse, accidents, and other plagues of modern time, it is important to be aware that there is only one phenomenon that affects every human being as well as all members of the animal and plant kingdomsÐthe aging process. As will be discussed in more detail below, this is even true for children and young adults, since it appears that age-related diseases are often the price for biological characteristics, e.g., innate genetically determined traits, that are advantageous up to the age of reproductive maturity, but deleterious thereafter. Aging represents the most incisive personal, medical, and socioeconomic problem of our society. Therefore, studying age-related biological phenomena is not only a fascinating theoretical subject, but one with tremendous practical implications. Many years ago, the World Health Organization (WHO), together with the National Institute of Aging (NIA), identi®ed four major areas of gerontology (the science of research on aging) that its member states were urged to concentrate their scienti®c efforts on. These were: 1. nutrition (worldwide undernutrition, paradoxically atherosclerosis in developed countries) 2. the nervous system (e.g., Alzheimer's and Parkinson's diseases) 3. the immune system (infections, tumors, autoimmune diseases) 4. the endocrine system (e.g., osteoporosis). While it seems that all these priorities are still valid, studying the aging process of the immune system is especially rewarding. Before describing the characteristic age-dependent alterations of immune reactivity in old experimental animals (mice) and humans, we will brie¯y re¯ect on the aging process in general with special emphasis on the nature of age-related diseases. It can clearly be stated that the phenomenon of aging has not been conceived by nature. Human and animal organisms are constructed to function optimally up to the time of reproduction, beyond which represents life past the ``warranty 243 Copyright 2002 , Elsevier Science (USA). All rights reserved. 0065-2776/02 $35.00

244

b. grubeck-loebenstein and g. wick

period.'' In nature, old animals will, therefore, be seen very rarely, if ever, except in cases of domestic animals or animals kept in zoos or as pets, viz. animals not living under wild conditions. Signi®cantly aged men or menopausal women were certainly unknown at times of the ``Tyrolean iceman,'' i.e., approximately 5000 years ago. Indeed, to the 19th century, reaching an old age was a very rare event in human society. Human aging can thus be considered as an artifact of civilization. It is, however, beyond the scope of this review to dwell in depth on these general gerontological aspects. Aging is a basic biological process that is not restricted to higher organisms, but can, of course, also be observed in plants and even protozoa, such as yeast (Saccharomyces cervisiae) ( Jazwinski, 1998). Aging is a multifactorial process for which no simple explanation or cause can be identi®ed, and for which no ``treatment'' can be offered. In the literature, many more or less scienti®cally founded theories for the aging process are given. In the present context, only the two main theories will brie¯y be mentioned, i.e., the stochastic and the deterministic concepts for aging, respectively. It is important to distinguish between aging of a single individual and that of a whole population. One could bring forward the argument that bacteria, under favorable cultural conditions, can divide inde®nitively, and that such bacterial cultures are therefore immortal. With the same reasoning, one could pretend that each human being is also immortal, since his/her genetic material has been propagated forever. In reality, of course, mankind considered as a whole is immortal, similar to bacteria and unicellular animals and plants, while this is not the case individually. B. Theories of Aging 1. Stochastic Theories In essence, these theories see the aging process as a result of damage by stochastic processes that, foremost in their totality, entail the aging of a single cell and the whole organism, respectively. The main of these theories, which are not mutually conclusive, but partly even supplementing (for which only representative key references are given here), are as follows. a. Free Radicals. Oxygen radicals can damage many cellular molecules, such as proteins, as well as DNA. Living organisms have developed many protective mechanisms against the effect of free radicals, such as the superoxide dismutase (SOD), catalase, and other enzymes (Finkel and Holbrook, 2000). b. Glycosylation. Glucose can associate with various types of proteins and ®nally lead to the formation of so-called advanced glycosylation end products (AGEs), resulting in cross-linking of the affected molecules (e.g., collagen) and thus functional impairment (Brownlee et al., 1986).

the aging of the immune system

245

c. DNA Aging. DNA can be damaged by a great number of intrinsic and extrinsic factors, and each cell has highly ef®cient repair mechanisms against these alterations. However, this repair capacity shows a signi®cant decline during aging, thus leading to an accumulation of cellular alterations as a consequence of DNA malfunction (Hart and Setlow, 1974). 2. Deterministic Theories These theories are based on the assumption that aging is genetically determined, i.e., that the existence of ``biological clocks'' in single cells or the whole organism determine their life span. a. The Hay¯ick Phenomenon. Leonard Hay¯ick, a pioneer of gerontology, made the interesting observation that normal cells can go through only a limited number of divisions and that this number is dependent on the age of the donor (Hay¯ick and Moorehead, 1961). b. Telomeres. Telomeres are repetitive DNA sequences at the chromosomal ends that are essential for the start of DNA replication via interaction with DNA polymerase. These sequences are synthesized onto the ends of the DNA strand by the enzyme telomerase that, after birth, is only active in germ cells, pluripotent stem cells, and tumor cells. In somatic cells, the telomerase genes are silenced, and telomere shortening can be observed after each division, ®nally leading to the cessation of replication in senescent cells. When senescent somatic cells are transfected with telomerase genes, they are again able to synthesize telomeres and thus survive for a longer period of time (Harley et al., 1990). c. Cell Cycle Control. The cell cycle is subject to the control of a complex interaction of factors that either promote division (cyclins, cyclin-dependent kinases (CDKs)) or inhibit it (CDK inhibitors, anti-oncogenes). In normal cells, increasing age leads to a preponderance of inhibitory effects and culminates in cell cycle arrest and apoptosis (cellular senescence). Certain tumor viruses, e.g., the human papilloma virus-16 (HPV-16) encode proteins that associate with CDK-inhibitors, thus inactivating them and leading to transformation and immortalization of the respective target cell (Jansen-DuÈrr, 1998). Thus oncogene products are important target structures for new types of tumor therapy, and also represent interesting molecules with respect to the reactivation of senescent normal cells. Cell cycle control and the ®nal stage of apoptosis are not only crucial during aging, but also play an important role in youth, e.g., during negative T cell selection. d. Gerontogenes. In the gerontological literature, the term ``gerontogenes'' has been coined for genetic elements that play a role in cellular senescence

246

b. grubeck-loebenstein and g. wick

(Sohal, 1988). First, such genes were identi®ed in simple organisms, such as the small nematode Caenorhabditis elegans, but later homologs of such genes were also found in higher animals and in humans. However, it turned out that gerontogenes can be de®ned only in the context of the function of other genes, meaning that they were not conceived as ``aging genes'' by nature. Such gerontogene candidates may cooperate in the preservation or repair of somatic components of each organism, including the life-prolonging effect of caloric restriction that unfolds its activity via various metabolic pathways (Masoro, 2000). It also includes a high ef®ciency of DNA repair, a higher resistance against damage by free radicals and oxidation, and better immune reactivity. Knowledge of such genes is important for all future gerontological research as well as for formulating ``anti-aging'' strategies. In this respect, three classes of genes turned out to be of special importance, i.e., those encoding for stress responses, those responsible for production of proin¯ammatory cytokines, and those associated with metabolism (Lee et al., 1999). Over the years, many attempts have been made to approach the phenomenon of aging from an evolutionary viewpoint. These included the longabandoned theory of ``killer genes,'' which were considered to serve the purpose of removing parents and grandparents in order to provide ``space'' for younger generations. Another concept based on more solid evidence postulated that an accumulation of mutations was a major mechanism underlying cellular aging, postulating that no selective pressure must exist against such changes beyond the reproductive period. Such mutations could accumulate uninhibited over many cellular generations with deleterious consequences that would individually become manifest only with increased age. For our own gerontological work, the most interesting concept is that of antagonistic pleiotropism (Nesse and Williams, 1995; Carnes and Olshansky, 2001). As mentioned above, certain genes are important to guarantee optimal survival of the organism until the reproductive period. Due to their pleiotropic effects, however, these genes may play a crucial role later in life by contributing to the development of complex age-related diseases. With respect to the immune system, an example of this type of effect will be given when we discuss our new ``autoimmune concept'' for the pathogenesis of atherosclerosis. C. General Cellular Age-Dependent Morphological and Functional Changes Most organs display signs of atrophy in older age, such as the central nervous system and striated muscles. Organ systems with high regenerative potential (e.g., for hematopoiesis, spermatogenesis) show a slower aging process. In older age, atrophied tissue is replaced by functionally inferior tissue (fat or

the aging of the immune system

247

connective tissue), leading to the restriction of speci®c functions in the involved organs. However, some organs show hyperplasia rather than atrophy with increasing age. Examples for this phenomenon are the nests of proliferating lymphoid cells in the aging bone marrow, leading to monoclonal or oligoclonal immunoglobulin production, benign prostatic hyperplasia, and proliferation of intima cells in the course of the development of atherosclerosis. The generally impaired function of senescent cells can ®rst be described as a disturbed communication with their microenvironment. The structure that presents and mediates the connectivity of a cell to its environment is the plasma membrane with its inserted and more or less mobile receptor molecules. The binding of a ligand to a receptor, the signal transduction pathways, and the liberation of intracellular signal molecules, such as cAMP and cGMP, respectively, can all be disturbed with increasing age. Thus, decreased expression of receptors or impairment of their mobility due to the known age-related increase of plasma membrane viscosity (e.g., via an increased molar ratio of free cholesterol/phospholipids (C/Pl) ) can impair the subsequent signal transduction. This latter mechanism has been studied in our laboratory in lymphoid cells. Using complex and sensitive analytic methods on the living single-cell level, we were able to show that the plasma membrane of lymphocytes and other cells in elderly people (> 65 years) become more viscous due to the increased incorporation of cholesterol compared to lymphoid cells of young donors (< 35 years) (JuÈrgens et al., 1989). This parameter shows an excellent correlation with the age-dependent decrease of the response to mitogens or antigens (Schwarz et al., 1987). All cells require cholesterol for the composition of intracellular and plasma membranes and, in special cases, e.g., the adrenals, the production of steroid hormones as well. A nucleated cell can, on one hand, produce its own cholesterol, and on the other take up cholesterol from the environment in the form of low-density lipoproteins (LDLs, the transport particles for the water-insoluble cholesterol) via LDL receptors. A ®nely regulated balance exists between these two processes. We were able to show that lymphocytes of healthy elderly donors do not respond to an increase of environmental LDL concentration by appropriate downregulation of LDL receptors and thus are overloaded with cholesterol, resulting in increased membrane viscosity with the functional consequence of a lowered mitogen response (Huber et al., 1991). If cholesterol is removed from the plasma membrane of mononuclear cells from elderly donors, resulting in its ``¯uidization,'' these cells respond to mitogens again as to those of young donors (Traill et al., 1990). This observation may form the basis for future preventive and therapeutic anti-immunosenescence strategies.

248

b. grubeck-loebenstein and g. wick II. The Aging of the Immune System

A. General Considerations In the elderly, the incidence of severe infections is high (Gravenstein et al., 1998) and the protective effect of vaccination is low (Grubeck-Loebenstein et al., 1998), both due to the fact that the function of the immune system declines with age. Although this is generally accepted, the exact nature of the underlying defects is far from clear. Intensive research over the past decades has attempted to clarify the basic mechanisms of age-related immune dysfunctions, but it has turned out to be a dif®cult task. Many published reports con¯ict in theoretically and practically important points. The production of the TH type-1 cytokine interferon-g (IFNg) has, for instance, been reported to be high, low, or unchanged during aging, and many similarly contradictory examples could be cited (Miller, 1996b). Controversial observations may result from variations among species, strains, organs, and culture systems, but may also be due to interindividual differences in the course of the biological aging process. The health status of elderly persons may also play a role in this respect. To identify ``healthy'' old probands and appropriate young controls, a working party of the European Economic Community ``Concerted Action on Aging'' (EURAGE) established the so-called SENIEUR protocol of selection criteria for immunogerontological studies (Lighthart, 1984). This rather elaborate protocol takes into account (a) the living conditions of healthy volunteers, (b) laboratory values, and (c) drug intake. Still, even using the SENIEUR protocol, a high degree of controversy on substantial issues could not be avoided. However, due to increasing worldwide efforts by many laboratories, a more convincing and clear-cut picture has recently started to emerge. It is the goal of this review to summarize recent reliable data on the cellular and molecular mechanisms of immunosenescence and to try to interpret how basic dysfunctions may eventually lead to disability and disease in the elderly. B. Bone Marrow Progenitor Cells To determine whether abnormalities observed in the aged peripheral immune cell pool are due to de®ciencies in progenitor cells or the hematopoietic microenvironment, studies were performed in aged individuals, including centenarians and old mice (Globerson, 1999). These studies indicate that numbers of myeloid and erythroid colonies developing from peripheral blood CD34‡ cells of old donors are normal under optimal culture conditions (Bagnara et al., 2000), and no major changes in numbers of peripheral leukocytes, platelets, or red cells were observed in the healthy aged (Stulnig et al., 1995a). Whether there is a decrease in absolute numbers of bone marrow progenitor cells is not fully understood. However, reimplantation experiments in mice (Morrison et al., 1996) and the poor results of bone marrow transplantation in

the aging of the immune system

249

elderly individuals (Hakim et al., 1997) suggest that the aged bone marrow microenvironment has a signi®cantly reduced ability to support hematopoietic regeneration. C. T Lymphocyte MaturationÐThe Involution of the Thymus The thymus, the central lymphoid organ, is almost fully developed at birth, but its involution starts soon after puberty and continues throughout life (Steinmann, 1986). In the human, thymic tissue is almost completely replaced by fat by the age of 60 (George and Ritter, 1996), and the size of the organ decreases progressively with age (Aspinall and Andrew, 2000a,b). Few intact tissue islands retaining the cortical and medullary architecture seen in the young remain. The phenotypes of all thymic T-cell intermediates are present and there is still ongoing T-cell receptor gene rearrangement (Jamieson et al., 1999), but the rate of naõÈve T-cell output of the thymus dramatically declines (Aspinall and Andrew, 2000a; Mackall et al., 1995; Mackall and Gress, 1997; Kampinga et al., 1997). Interestingly, experiments in both humans and mice suggest that there are no age-related changes in the total numbers of lymphocytes or in the number of CD4‡ or CD8‡ T lymphocytes within the peripheral lymphoid pool (Stulnig et al., 1995a), which may be due to the fact that memory T cells proliferate to compensate for the loss of thymic output, leading to a reduction of the naõÈve T-cell pool (Aspinall and Andrew, 2000a,b). The development of a method for detecting recent thymic emigrants by PCR ampli®cation of the DNA circles formed during T-cell receptor rearrangement (TCR), called T-cell rearrangement exision circles (TRECs), allows for a rigorous analysis of thymic function in old age (Douek et al., 1998; Douek and Koup, 2000). This new technology has provided mounting evidence that the agerelated changes in the thymus are quantitative rather than qualitative, and that even the adult thymus can be reconstituted to some extent (Poulin et al., 1999; Fig. 1). TRECs also rise after bone marrow transplantation in humans, even in recipients over 50 years of age (Douek and Koup, 2000; Toubert et al., 2000). These recent data indicate that some capacity to produce naõÈve T cells is preserved even in old age. Thus, restoration of thymic function may be a target for immunological intervention in the elderly. Certain types of immune interventions, e.g., affecting the endocrine system, can lead to thymic ``rejuvenation,'' as discussed later in this review. D. NaiÈve T Cells NaõÈve T-cell activation requires T-cell receptor (TCR) stimulation with peptide antigen bound to MHC as well as ligation of T cell coreceptors by costimulatory ligands on the antigen-presenting cells (APCs) (Swain et al., 1996; Vella et al., 1997). Activation of naõÈve T cells stimulates interleukin (IL)-2 production. IL-2 can upregulate its own speci®c receptor by induction of

250

b. grubeck-loebenstein and g. wick

Fig. 1. Thymus cellularity/thymocyte subpopulations. Age-dependent decrease of murine CD8‡ thymus cellularity (top), however, with preserved proportions of CD4 CD8 , CD4‡ CD8 , CD4‡ CD8 , and CD4 CD8‡ subpopulations, respectively. Reprinted with permission from Dr. Richard Boyd.

IL-2 receptor a-chain, and it drives proliferation and extension of activated T cells and differentiation of T cells into effectors. Relatively little information is available on naõÈve T cells in old age, but the responsiveness to neoantigens may be particularly compromised in the elderly due to the early involution of the thymus, which leads to a shrinkage of the naõÈve T-cell repertoire. A recent study in which naõÈve T cells were de®ned as CD95 , CD45RA‡ , CD62L‡ cells demonstrates a sharp decrease of naõÈve T cells with age, from 800 cells/mL in young adults to 177 cells/mL in centenarians (Fagnoni et al., 2000). The naõÈve T-cell count was lower in the CD8‡ than the CD4‡ subsets at any age, and the older individuals were almost completely depleted of circulating naõÈve CD8‡ T cells (13 cells/mL). Additionally, intrinsic defects of naõÈve T cells have been described in aged mice (Linton et al., 1996; Haynes et al., 1999). Thus, it has been shown that

the aging of the immune system

251

aged naõÈve CD4‡ T cells produce low levels of IL-2, leading to inef®cient generation of effectors. The cells expand poorly, giving rise to few effectors with less activated phenotypes and reduced ability to produce cytokines. The aged cells also respond less vigorously in vivo. Addition of exogenous IL-2 or other g c-receptor signaling cytokines restores expansion. Only effectors generated in the presence of IL-2 are able to produce IL-2 in normal amounts and become polarized to secrete TH-2 cytokines. However, the defect reappears when IL-2-treated cells are reimplanted into irradiated aged or young mice. This suggests that defects in naõÈve CD4‡ T cells might be overcome by strategies to induce greater IL-2 availability, leading to more vigorous primary responses. In a more recent study, two further defects in the very early stages of activation induced by peptide and APC complexes were detected in naõÈve CD4 T cells from aged transgenic mice, whose T cells express a TCR speci®c for a peptide derived from pigeon cytochrome C, at the single-cell level (Garcia and Miller, 2001). First, a decrease in the proportion of T cells/APC conjugates that could relocalize signal proteins to the immune synapse was observed. Second, aging diminished the frequency of cytoplasmic NF-AT migration to the T-cell nucleus among cells that could generate immune synapses containing LAT, c-Cbl, or PLC-g. Little is known on the functional characteristics of CD8‡ naõÈve T cells in aged mice, and only controversial information is available on the function of naõÈve T cells in the elderly. Predictions of longevity have been made based on the numbers of naõÈve T cells. Thus, mice characterized by relatively low levels of CD4 and CD8 memory cells and high levels of CD4 naõÈve cells lived longer than conventional controls, indicating that naõÈve T-cell numbers may be used as a biomarker of aging that can predict longevity at middle age (Miller, 2001). E. Memory/Effector T Cells 1. Phenotypic Changes In the almost complete absence of the thymus, the aging immune system is characterized by continuous reactivations, clonal expansions, and elimination of memory/effector T cells of various speci®cities. This eventually leads to changes in the T-cell repertoire. Various lymphocyte subpopulations change during aging, the most striking change being a shift in the expression of CD45 isoforms from the CD45RA‡ CD45RO to the CD45RA CD45RO‡ subsets in humans (Stulnig et al., 1995a; Cossarizza et al., 1996). This change occurs within both the CD4 and the CD8 pools. The proportion of mouse T cells expressing the memory cell marker CD44 also increases with age in the blood, spleen, and peripheral lymph nodes in both CD4‡ and CD8‡ cells (Grossmann et al., 1991; Lerner et al., 1990; Ernst et al., 1990). Studies in aging mice have also documented an increased proportion of T cells expressing glycoprotein P, better

252

b. grubeck-loebenstein and g. wick

known for its role endowing neoplastic cells with multiple drug resistance (Witkowski and Miller, 1993). The proportion of T cells expressing glycoprotein P also increases with age in humans, although much of the change seems to be accomplished in early adulthood (Pilarski et al., 1995). Interestingly, aged antigen-experienced T lymphocytes lose important costimulatory molecules, such as CD40L (Lio et al., 1996; Fernandez-Gutierrez et al., 1999) and CD28 (Effros, 1997; Boucher et al., 1998); the loss of CD28 expression has been particularly well documented. Above all, the CD8 subset shows progressively decreasing CD28 expression (Chamberlain et al., 2000). CD4‡ T cells, almost all of which are CD28‡ in young adults, also show an increasing CD28 fraction in the elderly (Weyand et al., 1998). Telomere shortening in the CD28 cells is more pronounced than in CD28‡ cells from the same donor, indicating that the former have undergone more rounds of cell division than the latter (Monteiro et al., 1996; Batliwalla et al., 1996, 2000). CD28 cells frequently also reaquire a CD45RA‡ phenotype (Nociari et al., 1999; Bandres et al., 2000). Since these cells also express molecules such as CD11a and CD95 at high concentrations and have a very polarized cytokine production pro®le (see below), the phenotype is consistent with a state of terminal effector cell differentiation (Champapne et al., 2001). Further characteristics, such as loss of growth potential (Bandres et al., 2000) and an increased resistance to apoptosisinducing stimuli (Schirmer et al., 1998; Spaulding et al., 1999), are reminiscent of the well-de®ned phenotype of senescent ®broblasts or keratinocytes (Toussaint et al., 2000; Campisi, 1998). Terminally differentiated CD8‡ T cells also frequently carry the natural killer (NK) cell marker CD56 (Tarazona et al., 2000). Interestingly, the fraction of NK1:1‡ T (NKT) cells, a recently described population of T cells that shares some characteristics with NK cells and is restricted by CD1d (Godfrey et al., 2000), has also been described to be enlarged in the elderly (Miyaji et al., 2000). 2. Repertoire With advancing age, healthy humans frequently demonstrate large clonal expansions of CD8‡ T cells in the peripheral blood that persist for long periods of time (Chamberlain et al., 2000; Posnett et al., 1994; Schwab et al., 1997; Wack et al., 1998). These expansions are often remarkable large and may comprise more than 50% of the circulating CD8‡ lymphocyte population. CD8‡ expansions are essentially undetectable in cord blood and infants and increase in prevalence with age. The etiology of most expansions in healthy individuals is not known, but studies of T cell receptor usage have strongly suggested that these cells arise in response to continued antigenic stimulation (Wack et al., 1998; Fitzgerald et al., 1995). CD4‡ T-cell expansions are more rare, but have also been observed in the elderly (Schwab et al., 1997). The surface phenotype of clonal T-cell expansion in the elderly is also consistent

the aging of the immune system

253

with their prior activation by antigen. They are usually CD45RO‡, but sometimes also CD45RA‡ (see above), CD28 , CD95‡ , CD11abright , CD57‡ , but CD62 (Posnett et al., 1994; Hingorani et al., 1993; Hamann et al., 1997). They also lack markers of acute activation, such as HLADR, CD25, and CD69 (Azuma et al., 1993). Clonal CD28 T cells proliferate poorly in vitro, but have been shown to be autoreactive, which might explain why they are maintained in vivo at a high frequency for long periods of time (Fitzgerald et al., 1995; Azuma et al., 1993). Similar clonal expansions have also been detected in aged mice (Callahan et al., 1993). One can summarize that although absolute and relative numbers of memory T cells increase with aging, their clonal diversity becomes increasingly restricted. Whether the accumulation of large expanded, not fully functioning clones has implications for the development of immunosenescence has not yet been clari®ed. Recent data from our laboratory indicate that elderly persons who fail to mount a humoral response to in¯uenza vaccination have an increased frequency of expanded clones in their CD8 pool that produce large amounts of IFNg, but no other cytokines (unpublished observation). These observations strongly favor the possibility that the presence of expanded clonotypes indeed affects the function of the immune system in old age. 3. Cytokine Production It is safe to say that cytokine production in old age is one of the most controversial topics in immunological research (Miller, 1996a; Bernstein and Murasko, 1998). Increased, decreased, or unchanged concentrations have been reported for each imaginable cytokine. The reasons for this long-lasting controversy are not entirely clear, but the publication of a relatively large number of well-documented and convincing reports has allowed a more clear-cut picture to recently emerge. a. Interleukin 2. There is now almost general agreement that total IL-2 production decreases with age (Miller, 1996a), for more than one reason. First, and probably most important, the number of naõÈve T cells, which are extremely good IL-2 producers, is greatly diminished (Fagnoni et al., 2000). Second, low IL-2 production, low expression of IL-2 receptors, and a poor response to IL-2 have also been observed in memory/effector T cells from older animals and humans (Flurkey et al., 1992; Miller, 1984). In a model of immunosenescence in which monoclonal T-cell populations can be studied longitudinally throughout their proliferative life span in culture, ability to secrete IL-2 declines as the cultures age (Pawelec et al., 1997). Reduced IL-2 production by memory/ effector T cells in the elderly are presumably due to alterations in signal transduction pathways, but this needs to be further clari®ed (Gorgas et al., 1997; Whisler et al., 1996).

254

b. grubeck-loebenstein and g. wick

b. Interferon g. Although many in vitro studies on IFNg production by T cells have shown a decline with age (Ouyang et al., 2000), there are now many contradictory indirect indications that interferon-g production, in fact, increases in the elderly. Early studies consistently reported that neopterin, a macrophage product, the production of which can exclusively be triggered by interferon-g, is elevated in old age (Stulnig et al., 1993; Diamondstone et al., 1994; Catania et al., 1997). In accordance with this result, the production of other interferon-g-inducible mediators, such as tumor necrosis factor (TNF)-a, have also been shown to be increased in many elderly cohorts (Bruunsgaard et al., 2000, 2001; Saurwein-Teissl et al., 2000). There is thus little doubt that aging must be associated with an increase in the whole body load of interferon-g. It is, however, less clear which T-cell types are responsible for this increased production. CD28 terminally differentiated T effector cells may be good candidates, since they produce large amounts of interferon-g and, as mentioned above, increase profoundly with aging. Many laboratories, including our own, have shown a good correlation between the numbers of CD28 cells and interferon-g production (Bandres et al., 2000). Since CD28 cells may be autoreactive (Nociari et al., 1999), interferon-g production could be continuously stimulated, which would lead to chronic elevation of basal interferon-g levels in lymphoid or other tissues and might initiate the overall cytokine imbalance and chronic in¯ammatory state (``in¯ammage'') recently associated with aging (Franceschi et al., 2000a). Interferon-g produced by CD28 cells could stimulate monocytes and dendritic cells to produce interleukin-12, which would, in turn, engage interferon-g production by CD28‡ T cells and thus induce a vicious circle. On the other hand, interferon-g inhibits the production and some effects of TH 2 cytokines. An imbalance between pro- and antiin¯ammatory cytokines would be the result, which could supposedly support the generation of TH 1 rather than TH 2 cells following priming. Recent data from our laboratory do demonstrate that puri®ed CD4‡ T cells from elderly persons, who failed to produce protective anti-in¯uenza antibodies following immunization, do not produce IL-4 and IL-5 upon in vitro stimulation (unpublished observation). Interferon-g production by NK cells and NKT cells could further contribute to the preponderance of TH 1 activity. c. Interleukin-4, Interleukin-5, and Interleukin-10. It has been long believed that there are age-associated changes in immune responsiveness characterized by a type 2 cytokine phenotype in neonates developing into a predominant type 1 phenotype in adults and reverting to a type 2 phenotype in the elderly (Shearer, 1997). This hypothesis was primarily based on studies in the mouse, but has turned out to be oversimpli®ed and is probably wrong for humans. Previous studies were mostly performed on unfractioned PBMC in which the high proportion of memory cells vs low naõÈve T-cell numbers, a

the aging of the immune system

255

typical feature in the elderly, was frequently neglected. Another problem may be that in a subgroup of elderly persons IL-4 seems to be predominantly produced by CD8‡ rather than by CD4‡ T cells. This surprising ®nding was recently published (Yen et al., 2000) and has been con®rmed by our laboratory in a study that additionally demonstrated IL-4 production by CD8‡ cells as a speci®c characteristic of elderly persons who are still capable of raising a protective humoral immune response following immunization (Fig. 2). Type 2 cytokine production by CD8 T cells might thus either be a compensatory mechanism that occurs in old age or, alternatively, a genetic feature that allows some circumvention of the functional consequences of immunosenescence. In this context, it is of interest that genetic loci associated with the variation of stem cell cycling activity and longevity have now been mapped. One such locus maps to chromosome 11 in an area containing the IL-3, -4, -5, -13, GM-CSF cytokine gene cluster (DeHaan and VanZant, 1999). IL-4/IL-5 production by

Fig. 2. CD8‡ T lymphocytes produce increased amounts of type 2 cytokines in elderly persons. An increased production of IL-4 and IL-5 by CD45RO‡ T cells is frequently observed in elderly persons. In contrast, CD8‡ T cells from young persons hardly produce type 2 cytokines. PBMCs were stained with Cytochrome-anti-CD8, PE-anti-CD45RA, and FITC-anti-IL-4 or anti-IL-5 following a 5-h stimulation with PMA (0:5 mg=mL) and ionomycin (0:5 mg=mL). The CD8‡ cells within the lymphocyte gate were analyzed. The percentages of cytokine-producing cells within the CD45RA population are shown.

256

b. grubeck-loebenstein and g. wick

CD8 cells from selected elderlies could be under a different regulatory control than in CD4 T cells and could hamper the interpretation of data from unselected cohorts. Decreased IL-4/IL-5 production by CD4‡ T cells in elderly persons who fail to produce speci®c antibodies after vaccination has already been pointed out above. Further work is needed to clarify signal transduction pathways responsible for the production of IL-4 and IL-5 by the different T-cell subsets in old age. Interleukin-10, an anti-in¯ammatory cytokine produced by TH 2 cell as well as by other cell types, has also been studied in aged humans and mice (Saurwein-Teissl et al., 2000; Pioli et al., 1998; Sadeghi et al., 1999). Although the literature is, again, relatively controversial, there are strong indications that T cells produce less interleukin-10 in the elderly than in the young (Saurwein-Teissl et al., 2000), which would again support the concept of a cytokine imbalance in favor of TH 1 cytokines and a decreased production of TH 2 cytokines in the elderly. 4. Helper Function The signi®cance of T cell help for antibody production by B cells and the effects of aging on this process will be discussed in Section II.G. In the context of memory/effector T cells, it should only be mentioned that the cytokine microenvironment, as well as molecules responsible for cell to cell interactions between T and B cells, are major determinants for intact antibody production in old age. Decreased numbers of CD28‡ and CD40L‡ T cells, and of a lack of type 2 cytokines, are both likely to endanger normal T-cell/B-cell communication, B-cell growth, differentiation, and antibody production in the elderly. 5. Cytotoxicity A decline with age in the in vitro production of antigen-speci®c cytotoxic T cells has been described (Mbawuike et al., 1993). Thus, resting CD8‡ cells had a reduced ability to generate clones of cytotoxic effectors when exposed to activating targets plus IL-2 (Miller, 1984; Saxena et al., 1988). The question of whether cytotoxic effectors, once generated, show an age-dependent decline in lytic function is controversial, with data in favor as well as against this possibility (Gottesman et al., 1988; Bloom et al., 1988). The obvious loss of antigen-speci®c MHC class I-restricted cytotoxicity may be partially counterbalanced by less speci®c killing mechanisms. Thus, increased numbers and/or increased functional activity of NKT cells, NK cells, and macrophages has been described in old age (Miyaji et al., 2000). These cell types will be referred to in detail later. 6. Proliferation Once naõÈve T cells have been successfully stimulated, costimulated, and cytokine availability and utility assured, the T-cell response requires clonal expansion. The entry of the cell into the cell cycle is the ®nal consequence of

the aging of the immune system

257

T-cell activation. The expression of protooncogenes such as c-myc and c-myb and the production of IL-2 and its receptor are prerequisites for this process. Decreased expression of these parameters, as frequently observed in old age (Nagel et al., 1988; Schwab et al., 1985; Gamble et al., 1990), may inhibit cell progression from G0 to G1 and from G1 to S (Perillo et al., 1993). T cells capable of entering the cell cycle clonally expand upon antigenic stimulation until a substantial population of one speci®city is generated. There is general consensus that the proliferation of T lymphocytes from humans or mice following stimulation with nonspeci®c stimuli, such as PHA, decreases with aging (Globerson, 1995; Pawelec et al., 1997). This may be due to several defects: Low IL-2 production, low expression of IL-2 receptors, signaling defects, loss of important costimulatory molecules such as CD28, or loss of replicative capacity due to clonal senescence. The latter feature has also been observed in a T-cell long-term culture system (Pawelec et al., 2000). Using this system, some investigators (Grubeck-Loebenstein et al., 1994; McCarron et al., 1987), but not others (Pawelec et al., 2000), observed lower numbers of population doublings in T cells obtained from old persons. This discrepancy is not surprising, as a residual proliferative capacity may be retained in some T cells until late in life, but may be lost early in others. Consistent with this idea, limiting dilution experiments suggest that murine naõÈve T cells are much more likely than memory T cells to produce IL-2 in culture and more likely to proliferate in response to IL-2 after activation by mitogens or alloantigens (Miller, 1984; Nordin et al., 1983). Poor proliferation of CD28 T cells (see above) also suggests that, depending on previous antigenic exposure, certain T-cell populations may proliferate better than others. Loss of the capacity to express telomerase upon activation and short telomeres as a result may be responsible for the loss of replicative capacity in old memory/effector cells (Effros, 2000). The demonstration of very short telomeres in CD28 T cells supports this theory (Monteiro et al., 1996; Batliwalla et al., 1996, 2000). F. B Cell Aging and Antibody Production Primary antibody responses in aged humans are often weak and short-lived, and the antibodies produced bind the lower af®nity antigens less well than those produced in young adults (Goidl et al., 1976). Many old individuals also exhibitÐmostly benignÐmonoclonal serum immunoglobulin peaks (Radl, 1990; LeMaoult et al., 1999) and there is a strongly increased occurrence of autoantibodies, as discussed below (Rowley et al., 1968; Weksler, 2000). This leads to a decline in speci®c antibodies with maintenance of total immunoglobulin levels. Studies in mice have also revealed the absence of a signi®cant increase in the af®nity of primary antibody responses in old mice (> 22 months). Reduced germinal center reactions are correlated to declining humoral responses in aging in that there is a severe reduction (60±95%) in

258

b. grubeck-loebenstein and g. wick

germinal center reactions in mice of > 22 months of age (Szakal et al., 1990; Kosco et al., 1989). The reduction of germinal center reactions is a gradual course in life during which both the number and volume of germinal center diminishes progressively (Fig. 3). The loss of robust germinal center responses reduces the conditions for antibody af®nity maturation and may also prevent establishment of antibody-forming cell populations in the bone marrow, which act as long-term antibody producers. Parallel to the loss of germinal centers, a reduction in the frequency and size of plasmacytic foci in the periarteriolar lymphoid sheath in the spleen is observed (Zheng et al., 1997). Lymphoid tissue from aged mice also contains atrophic follicular dendritic cells (FDCs), indicating that antigen may be bound and trapped less well in the form of 24

20 18

20 16

Number of GC per 1 ⫻ 10 mm2

18 14

16 12

14 12

10

10

8

8 6

6 4

4 2

2 0

GL-7+ cells in live-gated splenic lymphonocytes (%)

GC GC cells

22

3

6

8

10

17

0

Age (months)

Fig. 3. Progressive decline in germinal center formation during aging. Spleens of C57BL /6 mice immunized 14 days earlier with NP-CGG precipitated in alum were prepared for immunohistologic or ¯ow cytometric analysis. For enumeration of GCs (black bars), splenic sections were stained with PNA-HRP and GCs were counted under a light microscope. For enumeration of GC cells (grey bars), single-cell suspension of splenocytes was stained with anti-B220 and GL-7 antibodies and analyzed by ¯ow cytometry. Reprinted with permission from Zheng et al. (1997).

the aging of the immune system

259

unprocessed molecules in old age, which would mean that antigen deposits could last for much shorter periods of time than in the young (Szakal et al., 1990). This mechanism may be linked to the short-term maintenance of serum antibody titer and the failure to establish fully functioning memory B cells in old age. This is independent of age-related T-cell dysfunction, since FDC develop normally in athymic nude mice (Mitchell et al., 1972; Mjaaland and Fossum, 1987). Similar to humans, aged mice also develop clonal B cell expansions, the majority of which derive from the CD5‡ subset (Weksler, 2000). Although some of the described phenomena may be the result of intrinsic age-related B-cell defects, the low-level and brief production period of speci®c antibodies to foreign antigens in the elderly is mainly due to lack of regulatory control of T cells on B cells. In addition to signals mediated by the T-cell antigen receptor and the B-cell receptor, the germinal center reaction and T/B cooperation generally have been found to be dependent on the interaction between costimulatory molecules and their ligands on the T and B cell, respectively, as well as on the presence of a certain cytokine microenvironment. It is well established that CD4‡ T help is required for germinal center formation and the activation of the IgG somatic hypermutation machinery (Miller C. et al., 1995). Indeed, the rate of IgG hypermutation in germinal center cells is directly proportional to the number of available T helper cells. It has also been demonstrated that T cells may support the expansion of B cells expressing a particular germ-line-encoded VH rearrangement in response to a speci®c antigen. CD4‡ T lymphocytes from aged animals were unable to sustain this preferential expression (Yang et al., 1996). Three pairs of cell-to-cell interactions are considered particularly important in this context: CD40 and CD40L, CD80/CD86 and CD28, and OX40 and OX40L. While the CD40±CD40L interaction enables B cells to respond to activated T cells, the interaction between CD80/ CD86 and CD28 allows T cells to respond to activated B cells by proliferation and the production of cytokines (Shahinian et al., 1993; Tan et al., 1993). OX40± OX40L interaction plays an important role in B cell activation and plasma cell development (Walker et al., 2000). CD40L, CD28, and OX40L have all been demonstrated to be expressed at reduced levels on T cells from aged mice and humans (see also section II.E). Cytokines also play an important role. Although type 1 as well as type 2 cytokines can support antibody production, it is generally accepted that IL-4 (in humans and mice) and IL-5 (in mice) are of particular relevance for enhancing B-cell proliferation of anti-Ig-stimulated B cells and for inducing antibody production. Age-related changes in the production of cytokines have been discussed at length in section II.E, the main message being that aging is accompanied by a change in the polarization of the immune response due to a relative overproduction of IFNg accompanied by decreased levels of TH2 cytokines.

260

b. grubeck-loebenstein and g. wick

G. Macrophages and Dendritic Cells Recently, there has been major consensus that aging is associated with a chronic activation of innate immunity as the ®rst line of defense (Franceschi et al., 2000a). This claim has been made on the basis of many studies demonstrating increased production of cytokines and other nonspeci®c effector molecules by cells of the innate immune system in the elderly (Bruunsgaard et al., 2000, 2001; Saurwein-Teissl et al., 2000; Ershler and Keller, 2000). Thus, increased production of IL1a and -b, of TNFa, IL-6, complement, prostaglandin and other monocyte/dendritic cell (DC) products has been observed in aged humans as well as experimental animals. Many reports in which antigen-presenting cells from aged donors have been tested for their ability to support T-cell activation also show no age-related decrease of this speci®c function (Miller, 1996a), although no studies exist thus far in which antigen uptake, processing, and presentation have been analyzed in detail. The reason for the unchanged or even enhanced function of cells of the innate immune system in the elderly is not entirely clear. Most likely, there are several operative stimuli: Excess IFNg production by lymphocytes and NK cells, degenerative processes, clinical and subclinical infection, continuous surface exposure to bacteria or allergens, e.g., in the skin, respiratory tract, or gut. It is also not yet clear how much, and at which anatomic sites, macrophages and DCs contribute to the generation of a proin¯ammatory status. In spite of the fact that DCs are known to be the most ef®cient antigenpresenting cell type to activate T cells and are key modulators of the immune response, very few groups have addressed the topic of DC and aging. DC maturation and aging has been studied in the human system using DC derived from monocytes by in vitro stimulation with IL-4 and GM-CSF (Steger et al., 1996b; Lung et al., 2000). The responsiveness of these cells to maturationinducing stimuli was tested and found not to be impaired in old age (SaurweinTeissl et al., 1998). A recent study additionally demonstrates that, in vitro, the migratory capacity of monocyte-derived DC is not impaired (Pietschmann et al., 2000). DC differentiated from monocytes in vitro were also shown to have an unimpaired capacity to stimulate T cells (Steger et al., 1996b, 1997b; Lung et al., 2000). These results are important because they show that DC can still represent useful tools for immunotherapy in old age, particularly as carriers of tumor vaccines, but there is still a paucity of information about the in vivo situation. Decreased numbers of Langerhans cells (LCs) were described in the epidermis and mucosa from aged animals and humans (Sprecher et al., 1990; Gilchrest et al., 1982). The extent of dermal LC reduction was about one third. These results together with those of another study indicating that the capacity of LC to migrate from the skin to the draining lymph node was reduced in aged rodents (Sprecher et al., 1990), may indicate impaired DC migration in vivo. In

the aging of the immune system

261

view of the above-mentioned in vitro results, this may not re¯ect an intrinsic DC defect, but rather be due to factors such as decreased permeability of tissue barriers. The cytokine microenvironment found in aged peripheral tissues, such as the skin, is also still very poorly de®ned and may not support optimal DC maturation and migration, but a de®nitive picture on these important issues has not yet emerged. In contrast to DC, age-related constitutive defects have been described in macrophages. A recent study indicates that despite an unchanged degree of differentiation, bone-marrow-derived macrophages from aged mice present low levels of MHC class II gene induction by IFNg because of impaired transcription (Herrero et al., 2001). The consequences of overactivity of the aged innate immune system have recently been a matter of great interest. On one hand, a continuous proin¯ammatory status is now well documented as a factor in the development of agerelated disorders not previously associated with immune reactivity, such as atherosclerosis or Alzheimer's disease (see also below). On the other hand, chronic activation of innate immunity is also compatible with extreme longevity in good health, and proin¯ammatory characteristics have been documented in healthy centenarians, which may at least partially compensate for declining T-cell function (Franceschi et al., 2000a,b; Baggio et al., 1998). Further work is needed to clarify whether healthy centenarians have a speci®c genetic background, which protects them from potential detrimental effects of in¯ammation or whether a certain threshold of in¯ammation is the prerequisite for the development of diseases. H. Natural Killer Cells NK cells play a critical role in the innate immune response against infections and tumors. In addition to their cytotoxic function, NK cells produce cytokines and chemokines and also express the corresponding receptors. Due to their high IFNg production, they primarily promote TH1-mediated immune responses. Age-associated alterations in the number and function of NK cells have been reported by several groups, and there is now general consensus that a progressive increase in the percentage of NK cells with the mature phenotype occurs in elderly donors (Miyaji et al., 2000; Solana and Mariani, 2000). The lytic capacity of NK cells considered on a single cell basis, however, is decreased (Faccini et al., 1987; Mariani et al., 1990), as is the proliferative response of NK cells to IL-2 or other cytokines (Borrego et al., 1999). In contrast, IL-2triggered TNFa and perforin production are not impaired (Borrgeo et al., 1999). Taken together, the presently available results indicate that age-related functional defects are presumably compensated for by an increased number of mature NK cells; total NK activity thus seems to remain intact even in very old persons.

262

b. grubeck-loebenstein and g. wick

I. Activation and Signal Transduction in Immune Cells in Old Age Very few data are available on the effect of aging on activation signals in APCs or B cells, and these aspects will therefore be excluded from this review. However, substantial amount of work has been performed on the effects of aging on T-cell activation. A reduced calcium in¯ux indicates that early events in T-cell activation are compromised in the elderly (Gupta, 1989; Whisler et al., 1991). The earliest surface alterations affected by activation such as CD69 and CD71 are also decreased (Lio et al., 1996), which may, in the ®rst instance, be caused by disturbed signal transduction. Disturbances may be due to alterations in the expression of the TCR or in the signaling cascade through either TCR components or costimulatory molecules. Thus, it has been shown that antibodies against the signal transducing TCR associated CD3z chain precipitate a series of tyrosine phosphorylated proteins in activated T cells. Although the level of these proteins (Whisler et al., 1998) as well as the association of ZAP-70 with the z chain (Garcia and Miller, 1995) are retained, their degree of phosphorylation after T-cell activation declines with age in mice and humans (Whisler et al., 1998; Garcia and Miller, 1997). In the same way, the expression level of ZAP-70 is unchanged, but its activity decreases with aging (Whisler et al., 1999). PKC-mediated signaling seems also affected by the aging process. In human T cells, a selective reduction of one isoform of protein kinase C (PKC) has been suggested (Whisler et al., 1995). In mice, assays for PKC distribution show a decline with age in the proportion of T-cell/APC conjugates that displayed relocalization of PKCf (Yang and Miller, 1999; Miller, 2000), indicating that a key checkpoint in the activation process may be disturbed. Further data from the literature indicate that T-cell receptor-associated p59fyn enzymatic activity, which is essential for signalling via CD2, but not p56lck activity was reduced in a high proportion of T cells from the elderly compared to the young, although protein levels were the same (Whisler et al., 1995). The p56lck pathway may still be affected by the aging process, as the usual association between CD4 and p56lck may be compromised in T cells from old persons (Tinkle et al., 1998). These data indicate that the very early signal transduction pathways required for T-cell activation are compromised in old age. However, downstream signaling pathways are also affected. Thus, a decreased production of second messengers, such as IP3 and DAG, has been reported (Utsuyama et al., 1997; Hirokawa, 1999). The family of mitogenactivated protein kinases (MAPKs), which are considered essential for normal cell growth and function, also seems to be compromised. In rats, MAPK/RAS activities are decreased in old T cells, and in man, CD3-stimulated T cells from 50% of old subjects showed reductions in MAPK activation (Miller, 2000; Hirokawa, 1999; Pahlavani et al., 1998; Whisler et al., 1996). The ERK and

the aging of the immune system

263

JNK kinases have been reported to be diminished in CD3/PMA-stimulated T cells from elderly humans, accompanied by decreased Raf-1 kinase activation. ERK2 activation may represent the rate-limiting step for IL-2 production by old T cells. Although the issue of age-related changes in signal transduction is of paramount importance for the understanding of immunoscenece in general and immunointervention in particular, our knowledge in this area is obviously still far too fragmented to draw clear-cut conclusions with practical relevance. III. The Consequences of Immune Senescence A. Infectious Diseases The increased frequency and severity of infectious diseases are the most striking clinical features in the elderly (Gravenstein et al., 1998). An example is the high incidence of pneumonia, which is mostly caused either by in¯uenza or by pneumococcal infection. A recent epidemiological study on invasive Streptococcus pneumoniae infections in the United States, for instance, demonstrates that the incidence of disease was highest among children under 2 years of age and adults aged 65 years or older (Robinson et al., 2001); 28.6% of all patients were at least 65 years old. In¯uenza is associated with a signi®cantly higher morbidity and mortality in the elderly (Nichol et al., 1994). Each year, tens of thousands of deaths are attributed to in¯uenza and related complications, such as pneumonia, and 80% of in¯uenza deaths occur in individuals over 65 years of age. Other infections that frequently occur in the elderly are urinary tract and skin infections and an increase in hospital-acquired and nursing home infections. Impairments in primary host defenses also contribute to the increased rates of infections, such as reduced cough re¯ex leading to aspiration pneumonia, urinary and fecal incontinence predisposing to urinary tract infections, and immobility predisposing to wound infections. Old individuals may also fail to respond normally to therapy for infection and may present with infections secondary to unusual organisms (opportunistic infections), recurrent infections with the same pathogen or reactivation of quiescent diseases, such as those caused by M. tuberculosis and herpes zoster virus. The latter disorder is caused by Varicella zoster virus and is increasingly prevalent with advancing age, as are its severity and complications. A relatively new and exciting facet of infectious disease-associated agerelated complications concerns atherosclerosis, one of the main killers in the developed world. It has long been known that immunologic/in¯ammatory processes take place during the development of atherosclerosis, but it was not clear whether these events were of a primary or secondary nature. During the past decade, we have developed a new ``autoimmune'' hypothesis of atherogenesis that describes a common denominator for all types of infections that have

264

b. grubeck-loebenstein and g. wick

been incriminated as possible initiating or perpetuating factors in atherosclerosis (Wick et al., 1992, 1995; Wick and Xu, 1999). We have shown that humoral antibodies and T cells reactive with the stress protein heat shock protein 60 (HSP 60) seem to be the effector mechanisms that lead to the damage of stressed arterial endothelial cells (Schett et al., 1995) and subsequent mononuclear cell in®ltration of the intima and, when classic atherosclerosis risk factors (such as high serum cholesterol levels, etc.) persist, progress into fatty streaks and ®nally severe lesions, i.e., atherosclerotic plaques. HSPs are phylogenetically highly conserved constituents of prokaryontic and eukaryontic cells classi®ed into various families according to their molecular weight (Welch, 1992; Morimoto, 1993). Here, the 60-kDa HSP family is of special interest, members of which constitute a considerable number of the total proteins of viral envelopes, bacteria, and parasites. Among other physiological properties, HSP 60 act as chaperones, i.e., are expressed after mild stress, associate with other proteins of the organism and protect them from denaturation. Similar to the situation in microbial organisms, stressed human cells also express HSP 60 that exert a chaperoning function. Classical risk factors for atherosclerosis (i.e., high blood pressure, smoking, overweight, high blood cholesterol levels, diabetes, etc.) lead to the expression of HSP 60 by arterial endothelial cells, especially at those areas of the vascular tree that are subjected to major hemodynamic (turbulent) stress, notorious predilection sites for the later development of atherosclerosis. Due to the life-long exposure of arterial endothelial cells to higher blood pressure as compared to venous endothelial cells, the former show a lower threshold for HSP 60 expression after being subjected to various stress factors (Amberger et al., 1997). We have shown that atherosclerosis may be the price exacted for our protective immune reactivity against microbial HSP 60 by a cross-reactivity with autologous HSP 60 expressed on the surface of arterial endothelial cells when these are stressed by atherosclerosis risk factors. A second possibility, of course, is the emergence of bona ®de autoimmune reactions as a response to biochemically altered autologous HSP 60 released in the case of various types of cellular damage or protein modi®cation after appearance in the circulation (Xu et al., 2000). Thus atherosclerosis is a perfect example of the Darwinian±evolutionary concept for the development of age-related diseases: As mentioned in the Introduction, genetic traits that are bene®cial during youth, such as the ability to mount a protective anti-microbial HSP 60 response, may become deleterious in the postreproductive phase of our lives when selective pressure is no longer effective. This concept is also in line with our recent observation that the total lifelong infectious load shows a signi®cant increase of the odds ratio to develop atherosclerosis. Again, infectious load (mainly re¯ected by respiratory infections) is correlated signi®cantly with the presence of antibodies to bacterial

the aging of the immune system

265

HSP 60 that cross-reacts with human HSP 60 (Mayr et al., 2000; Kiechl et al., 2001). B. The Low Protective Effect of Vaccination in the Elderly Infectious diseases in the elderly might be principally prevented by vaccination, but the protective effect of vaccination is poor in old age (GrubeckLoebenstein et al., 1998). Decreased antibody production and shortened duration of humoral protective immunity following immunizations are characteristic in persons of more than 60 years of age (Stein, 1994). At best, incomplete protection of elderly adults is afforded by the present in¯uenza and pneumococcal vaccines (Christenson et al., 2001), and vaccines known to achieve satisfactory immunization results in children and younger adults also leave much to be desired in the elderly (Fig. 4; Steger et al., 1996a, 1997a). A cohort of 300 elderly (> 60 years) and 300 young (< 35 years) persons was recently analyzed in our laboratory for the presence of speci®c antibodies against tetanus, diphtheria, tick-borne encephalitis, and in¯uenza. Lack of protection was obsevered in many elderly persons despite regular vaccinations in both frail and healthy elderlies. Another problem resulting from the early involution of the thymus (see Section II.C) is the vaccination of aged persons with new antigens, such as those for yellow fever virus, rabies, or new in¯uenza virus strains which they have not come in contact with before. These vaccinations are becoming increasingly popular due to the great mobility of today's elderly population. Determinations of antibody titers should be made before an elderly person is considered to be satisfactorily protected against such pathogens. Antibody determinations also seem to be indicated to assess the duration Young

Old

Fig. 4. Percentage of persons with (&) and without (&) protective antibody titers to tetanus. Young (< 30 years, n ˆ 30) and old (> 65 years, n ˆ 32) healthy volunteers selected according to the criteria of the SENIEUR protocol for immunogerontological studies of the European Community's Concerted Action Program on Aging were analyzed. Reprinted with permission from Grubeck-Loebenstein et al. (1998).

266

b. grubeck-loebenstein and g. wick

of protection after vaccination. It can be concluded that general vaccination strategies cannot be uncritically applied in the elderly. C. Disorders Due to Excess Immunoreactivity In classical textbooks, the three characteristics of the immune system are (a) the ability to distinguish self from nonself, (b) speci®city, and (c) memory. As discussed earlier, the restricted T-cell repertoire and narrowing of memory is one reason why elderly people are more susceptible to infections and do not appropriately respond to vaccinations. In general, the immune response against exogenous antigen in the elderly is decreased, and its speci®cty ``broadened''; i.e., the ability to discriminate self from nonself is impaired. This was shown in our laboratory by experiments where cytotoxic T lymphocyte (CTL) responses against allogeneic target cells were signi®cantly decreased in old (> 24 months) vs young ( 3 months) mice. However, the CTLs of these old mice also showed signi®cant cytotoxic activity against syngeneic control targets, while pure allogeneic speci®city was preserved in CTLs from young donors (Wick et al., 1989). In humans, cellular and humoral autoimmunity increase with age even in completely healthy, SENIEUR-protocol-compatible individuals, i.e., not due to underlying disease. Figure 5 depicts the age-dependent rise of the incidence of anti-nuclear autoantibodies (ANA). However, the occurrence of such

Fig. 5. ANA in old age. Age-dependent increase of the incidence of anti-nuclear autoantibodies (ANA) in a large number of female and male randomly selected, clinically healthy volunteers participating in an atherosclerosis prevention program. Unpublished data obtained with sera from the Bruneck-Study (Xu et al., 1993).

the aging of the immune system

267

antibodies is not accompanied by concomitant disease, in this case connective tissue diseases. It must, therefore, be emphasized that humoral and cellular autoimmunity do not equal autoimmune disease. As a matter of fact, autoimmune diseases start much earlier in life, but sequalae such as that seen in rheumatoid arthritis and, as brie¯y discussed above, atherosclerosis, become clinically manifest later. Several explanations have been put forward for the age-dependent increase of autoimmunity, including a physiological response to autoantigens that have been biochemically altered due to lifelong exposure to various environmental factors, and/or known pre- and posttranslational changes (protein aging). Furthermore, heightened autoreactivity may also be due to age-related alterations of immunoregulation caused by the mechanisms mentioned above and re¯ected by an altered ratio of naõÈve to memory/effector T cells, a restricted T-cell repertoire, a shift to a proin¯ammatory cytokine pro®le, oligo- and polyclonal activation of B cells and, currently controversial, a decrease in suppressor cell activity (Globerson, 1999). As mentioned in Section II.G, a continuous proin¯ammatory status is now well documented to play a role in the development of age-related disorders. Alzheimer's disease is a typical example. Large epidemiological studies demonstrated that Alzheimer's disease was less frequent in patients with rheumatoid arthritis and other disorders chronically treated with anti-in¯ammatory drugs than in untreated control groups (McGeer et al., 1990, 1996). This striking observation may be explained by the fact that nonspeci®c in¯ammatory factors are involved in the maturation of amyloid plaques in the brain and the propagation of pathology in the surrounding tissue. Thus, high numbers of microglial cells are found around the mature plaques of AD (Itagaki et al., 1989; Paresce et al., 1996) brains. These cells are activated, since they express increased levels of Fc receptors, MHC class I and class II molecules, as well as b2 integrin and the vitronectin receptor (McGeer et al., 1993; Eikelenboom et al., 1994). They may also produce cytokines such as IL-1, IL-6, and TNFa (Walker et al., 1995; Meda et al., 1995). Cytokine production can be augmented by aggregated Ab, particularly in the presence of other stimulatory agents, such as IL-1b or IFNg (Meda et al., 1995; Gitter et al., 1995). TNFa has been demonstrated to trigger the production of toxic free radicals in glial cells and to augment the cytotoxicity of the aggregated Alzheimer Ab fragment 25±35 on neuronal cells (Blasko et al., 1997). In¯ammatory cytokines can also alter the metabolism of the Ab precursor protein (bAPP), since combinations of TNFa or IL-1b and IFNg have been shown to trigger the production of Ab peptides and inhibit the secretion of the soluble neuroprotective APPs by human neuronal cells and astrocytes (Blasko et al., 1999, 2000). In¯ammatory compounds known to be produced at increased concentrations in old age may thus support the development of brain pathology in AD, a typical age-related disorder.

268

b. grubeck-loebenstein and g. wick

Osteoporosis is a condition characterized by rare®cation of the bone mass in a localized form or affecting the whole skeleton, allowing for distinction of primary and secondary forms. Secondary osteoporosis develops as a consequence of different causes, including infectious diseases, metabolic disorders, malnutrition, drugs, long immobilization, etc. The most common primary form is senile and postmenopausal osteoporosis that develops in association with agerelated changes of the sex-hormone status. In the past few years, a previously unknown association between osteoclast function and T-cell development emerged that may have profound implications on prevention and therapy of osteoporosis. A molecule belonging to the TNF family called osteoprotegerin ligand (OPGL; also known as TNF-related activation-cytokine (TRACE), or receptor activator of NFkB ligand (RANKL), or osteoclast differentiation factor (ODF) (Kong and Penninger, 2000; Lacey et al., 1998; Nakagawa et al., 1998) was identi®ed as an osteoclast differentiation factor that also regulates the interaction between T cells and DCs in vitro. OPGL is able to activate mature osteoclasts and stimulate osteoclastogenesis together with CSF-1. OPGL is expressed on osteoblasts, and its expression can be upregulated by 1.25 (OH)2 D3, IL-11, PGE2, or parathyroid hormone (Burgess et al., 1999). Thus, OPGL seems to be a key regulator of osteoclastogenesis and OPGLknockout mice develop an increased bone mass, so-called osteopetrosis, in response to a lack of osteoclasts (Bucay et al., 1998). DCs appear normal in OPGL-de®cient mice, but these animals show defects in the early differentiation of T and B lymphocytes. OPGL-de®cient mice are devoid of lymph nodes, but show normal spleen architecture and possess Peyer's patches (Kong et al., 1999a,b). OPGL, therefore, emerged both as a new regulator of lymph node organogenesis and development, and as an essential osteoclast differentiation factor. Recently, OPG, a decoy receptor for OPGL, has been discovered (Simonet et al., 1997). Interestingly, the OPG gene was shown to be estrogenresponsive, corroborating its effectiveness with the above-mentioned known association of the development of osteoporosis with the postmenopausal decline of sex hormones. These discoveries not only open new avenues for treatment of osteoporosis, but also show exciting new interconnections between the immune and the skeletal systems. IV. Modes of Intervention As mentioned in Section II.C, there is mounting evidence that even the adult thymus can, at least to some extent, contribute to T-cell reconstitution. Recent results reveal no age-related decline in either the number or the function of T-cell progenitors in the thymus, but changes in the thymic microenvironment in terms of the cytokines produced. As intact thymic function has been demonstrated to be crucial for the full recovery of immune reactivity, particularly

the aging of the immune system

269

reconstitution of the naõÈve T-cell pool after T-cell-depleted bone marrow transplantation, reconstitution of the thymic microenvironment may also be a critical factor for the success of strategies aiming at reversing immunosenescence. Possible means by which thymic functions could be reconstituted in old age have, therefore, been a matter of great interest. Several concepts as to how this goal might be achieved have been put forth, as follows. A. Hormones As mentioned, the ®rst event re¯ecting a decline of immune function is the involution of the thymus, which commences after puberty and reaches completion around the ®fth decade, when most of the thymic tissue has been replaced by fat. Hormonal in¯uences are of paramount importance, both in thymic maturation (including the T-cell selection process) and in thymic involution. Thus, we have shown that the entire glucocorticoid metabolism, starting with cholesterol and resulting in the production of the end products corticosterone and cortisol, respectively, takes place in the murine and avian thymus (Lechner et al., 2000, 2001). We have also shown that different subpopulations of thymocytes express different levels of glucocorticoid receptors (Wiegers et al., 2001). The group of Ashwell (Ashwell et al., 2000) has forwarded a concept suggesting a mutual antagonism between glucocorticoid- and T-cell receptor ligand-induced thymocyte apoptosis. However, age-dependent alterations of these mechanisms have not yet been addressed. In cooperation with R. L. Boyd (unpublished observation), we have studied the role of sex hormones in thymic involution and have shown that castration of both male and female old (>2 years) mice brought about a complete regrowth of the thymus to rates found in young animals. This process was completed within a few weeks and affected all T-cell populations, i.e., resulting in thymocyte subset ratios equivalent to those observed in young mice. Since castration is, of course, ethically unacceptable for ``immunologic rejuvenation,'' Boyd et al. now approach this issue by administering synthetic inhibitors of lutinizing-hormone-releasing hormone (LHRH) that has the same effect. The molecular mechanisms underlying this promising approach are still being elucidated. B. Cytokines A hypothesis that has recently been forwarded is that age-associated thymic atrophy results from defects in the thymic environment, which lead to a reduction in the amount of available IL-7 (Aspinall and Andrew, 2000a,b). This theory is based on several studies, foremost among them being those showing that treatment of mice with anti-IL-7 antibodies resulted in a form of thymic atrophy resembling the one observed with aging (Bhatia et al., 1995). IL-7 has also been shown by several studies to be an essential cytokine for T-cell development, since IL-7 is believed to be needed to induce TCR b gene

270

b. grubeck-loebenstein and g. wick

rearrangement in T-cell progenitors, presumably by supporting the survival of these cells (Oosterwegel et al., 1997). A recent study also suggests that the amount of intrathymically available IL-7 decreases with age (Aspinall and Andrew, 2000a,b). To test whether IL-7 was a limiting factor, experiments were carried out to provide exogenous IL-7 to old mice and follow the effects in the thymus and later in the blood (Aspinall and Andrew, 2000a,b). Groups of animals were injected subcutaneously twice a day with 500 ng of recombinant murine IL-7 over a 5-day period and then analyzed at two time points after the cessation of treatment. IL-7 had a major effect on both the thymic weight and total number of thymocytes compared with the control group. These changes were, however, no longer detectable 6 weeks after the cessation of treatment, suggesting that the boost of thymopoiesis seen following IL-7 treatment requires the continuous presence of elevated levels of this cytokine. Analysis of the blood of animals at this time revealed a signi®cant increase in the percentage of cells with the naõÈve cell phenotype compared with untreated controls. Summarizing these experiments, one may assume that treatment with IL-7 can reverse thymic atrophy and induce increased thymopoiesis which, in turn, leads to the appearance of cells with the naõÈve cell phenotype in the blood. Future work will, however, have to address important questions, such as which dose of IL-7 should be used to achieve maximum effects and which formulation of the cytokine could change its half life in a way that allows it to persist in the tissues for longer periods. C. Tissue Engineering According to another concept, a three-dimensional framework of biocompatible inorganic matrices might support the extracorporal reconstitution of tissues. The capacity of a tantalum-coated carbon matrix to support reconstitution of functioning thymic tissue was, therefore, tested (Poznansky et al., 2000). A thymic organoid was engineered by seeding matrices with murine thymic stroma. Coculture of human bone marrow-derived hematopoietic progenitor cells in this xenogeneic environment generated mature functional Tcells within 14 days. The proportional T cell yield from this system was highly reproducible, generating over 70% CD3 T cells from either AC133‡ or CD34‡ progenitor cells. Cultured T cells expressed high levels of TCR excision circles (TRECs), demonstrating de novo T-lymphopoiesis and function of fully mature T cells. This system permits the ex vivo genesis of naõÈve T cells from stem cells that could be reimplanted in the elderly to form a ``new'' naõÈve T-cell pool. Of course, it must be kept in mind that reconstitution of the thymus and/or the generation of a new naõÈve T-cell pool in old age would raise a great number of further questions. Would there be enough space for naõÈve T cells when the aged T-cell repertoire is dominated by large expanded clones of exhausted, not fully functioning memory cells? How could antigen be targeted to the new

the aging of the immune system

271

naõÈve T-cell population without being trapped by these memory/effector cells? Would a depletion of these senescent memory/effector cells be a necessary prerequisite for the successful regeneration of a naõÈve T-cell pool? Would the generation of a rejuvenated naõÈve T-cell pool entail the necessity to revaccinate old persons? Which vaccination strategies should then be used? Although cautious optimism may be allowed that immune senescence will eventually be reversible, it is obvious that it will be far from easy to achieve this task. Much knowledge about the molecular nature of factors responsible for the functional defects within the aged immune system will be needed to de®ne how the ``utopian'' situation of a young immune system in an old organism can be attained. V. Conclusion Aging represents the most incisive personal, medical, and socioeconomic problem in developed countries. In the present review, we ®rst discuss molecular and cellular aspects of aging followed by a brief summary of the most popular examples for stochastic and deterministic theories of aging, respectively, and a description of general cellular age-dependent morphological and functional changes leading to the proper topic, i.e., the aging of the immune system. In the elderly, the incidence of severe infections is high and the protective effect of vaccination low. Furthermore, autoimmune reactivity increases with age. However, the mechanisms underlying age-related immune dysfunctions are far from being clear. As a matter of fact, there are few ®elds in immunology that are less controversial than immunosenescence. Reasons for these controversial observations are discussed with special emphasis on the dif®culty to differentiate primary from secondary age-dependent alterations of immune reactivity, i.e., the latter developing due to underlying disease. Although bone marrow progenitor cells seem to be little affected in old individuals, there is a signi®cantly reduced ability of the microenvironment to support hematopoietic regeneration. The ®rst indication of immunosenescence is the involution of the thymus entailing a loss of naõÈve T cells in the periphery. The absence of the thymus is accompanied by continuous reactivation, clonal expansion, and elimination of memory effector T cells of various speci®cities leading to changes in the T-cell repertoire re¯ected by characteristic changes of lymphocyte subpopulations, e.g., a shift from the CD45RA‡ CD45RO to the CD45RA CD45RO‡ subset in elderly humans. Telomere shortening is more pronounced in CD28 than in CD28‡ cells from a donor, indicating that the former have undergone more rounds of cell division than the latter consistent with a state of terminal effector cell differentiation.

272

b. grubeck-loebenstein and g. wick

With advancing age, healthy humans often display large clonal expansion of the CD8‡ T cells in the periphery persisting for a long period of time. It has been suggested that this phenomenon is due to lifelong continued antigenic stimulation. Recent data from our laboratory indicate that elderly donors who fail to mount a humoral immune response to in¯uenza vaccines show an increased frequency of expanded CD8‡ clones that produce large amounts of IFNg strongly favoring the hypothesis that the presence of these clonotypes does play a role in immunosenescence. With respect to cytokine production in older age, the literature is also most controversial but in recent years a more clear-cut general picture has emerged showing an imbalance in favor of Th 1 cytokines and decreased production of Th 2 cytokines. These facts, together with decreased numbers of CD28‡ and CD40L‡ T cells, are likely to endanger normal T-cell/B-cell communication, B-cell growth differentiation, and antibody production in the elderly. Antigen-speci®c MHC class I restricted cytotoxicity is decreased in the elderly but this process seems to be counterbalanced by less speci®c killing mechanisms, such as increased numbers and functional activity of NKT cells, NK cells, and macrophages. The function of dendritic cells seems to be affected only to a minor degree in older age while the situation is less clear with respect to macrophages. In old mice, despite an unchanged degree of differentiation, bone-marrow-derived macrophages present only low levels of MHC class II gene expression upon induction by IFNg due to impaired transcription. In general, the consequences of an overactivity of the aged innate immune system has recently become a matter of major interest. On one hand, a continuous proin¯ammatory status has emerged as an important factor in the development of age-related disorders previously not yet associated with immune reactivity, such as Alzheimer's disease and atherosclerosis. On the other hand, studies in centenarians have documented that the chronic activation of innate immunity is compatible with extreme longevity and good health. Alterations of signal transduction in cells of the immune systems of older age have, surprisingly, received very little attention in spite of the importance of this issue for the understanding of immunosenescence. At this moment, we do not yet dare to draw clear-cut conclusions from the data that are available so far. In the ®nal part of this review, we discuss the clinical consequences of immunosenescence, the most important being the decreased resistance of the elderly to infections and the role of the proin¯ammatory state of the defense system for the development of widespread diseases. Of paramount practical importance in this respect is the problem of overcoming the low ef®ciency of vaccination in the elderly by devising new vaccine formulations and vaccination strategies. Furthermore, atherosclerosis seems to be a perfect example for the antagonistic pleiotropism of genes, i.e., that we have to ``pay'' by age-related diseases for genetic traits that are bene®cial in younger years up to the age of

the aging of the immune system

273

reproduction, but become deleterious later in live. In the case of atherosclerosis, we have shown that immune reactivity against microbial HSP 60 may result in a cross-reactivity with eukaryotic HSP 60 expressed by endothelial cells of arteries that have been stressed by classical atherosclerosis risk factors, thus leading to the initiation of this disease. Disorders due to excess of speci®c immunity are rare in the elderly in spite of the fact that the incidence of autoantibodies is increased. It is emphasized that the presence of humoral and cellular autoimmune reactions does not equal the occurrence of autoimmune disease. As a matter of fact, autoimmune diseases characteristically begin to develop at younger age but the consequences are often experienced only in later years. However, the already mentioned hyperactivity of the nonspeci®c innate defense system seems to be associated with variety of age-related diseases, such as Alzheimer's disease and osteoporosis. Finally, a short paragraph of this review is devoted to possible ways of intervention with age-related problems of the immune reactivity. This includes treatment with hormones, cytokines, and by tissue engineering. Acknowledgments The authors wish to acknowledge the long standing cooperation and friendship with all collaborators mentioned in this review. We also want to thank Anita Hohenegger and Anita Ender for secretarial help in preparing this manuscript. Our own work was supported by grants from the Austrian Science Fund (Projects: G. Wick: P12213-Med, P14741-Path; B. Grubeck-Loebenstein: P12440), from the Austrian Federal Ministry for Education, Science, and Culture (B. GrubeckLoebenstein: GZ 70.062/2-pr14/99; GZ 70.060/2-Pr/4/99) as well as from the European Union (B. Grubeck-Loebenstein: Contracts #QLK6-CT-1999-02031; #GLK6-CT-1999-02004).

References Amberger, A., Maczek, C., JuÈrgens, G., Michaelis, D., Schett, G., Trieb, K., Eberl, T., Jindal, S., Xu, Q., and Wick, G. (1997). Co-expression of ICAM-1, VCAM-1, ELAM-1, and Hsp60 in human arterial and venous endothelial cells in response to cytokines and oxidized low-density lipoproteins. Cell Stress Chaperones 2(2), 94±103. Ashwell, J. D., Lu, F. W., and Vacchio, M. S. (2000). Glucocorticoids in T cell development and function. Annu. Rev. Immunol. 18, 309±345. Aspinall, R., and Andrew, D. (2000a). Thymic atrophy in the mouse is a soluble problem of the thymic environment. Vaccine 18, 1629±1637. Aspinall, R., and Andrew, D. (2000b). Thymic involution in aging. J. Clin. Immunol. 20(4), 250±256. Azuma, M., Phillips, J. H., and Lanier, L. L. (1993). CD28-T lymphocytes: Antigenic and functional properties. J. Immunol. 150(4), 1147±1159. Baggio, G., Donazzan, S., Monti, D., Mari, D., Martini, S., Gabelli, C., Dalla Vestra, M., Previato, L., Guido, M., Pigozzo, S., Cortella, I., Crepaldi, G., and Franceschi, C. (1998). Lipoprotein(a) and lipoprotein pro®le in healthy centenarians: A reappraisal of vascular risk factors. FASEB J. 12(6), 433±437. Bagnara, G. P., Bonsi, L., Strippoli, P., Bonifazi, F., Tonelli, R., D'Addato, S., Paganelli, R., Scala, E., Fagiolo, U., Monti, D., Cossarizza, A., Bonafe, M., and Franceschi, C. (2000). Hemopoiesis

274

b. grubeck-loebenstein and g. wick

in healthy old people and centenarians: Well-maintained responsiveness of CD34‡ cells to hemopoietic growth factors and remodeling of cytokine network. J. Gerontol. A: Biol. Sci. Med. Sci. 55(2), 61±70. Bandres, E., Merino, J., Vazquez, B., Inoges, S., Moreno, C., Subira, M. L., and Sanchez-Ibarrola, A. (2000). The increase of IFN-g production through aging correlates with the expanded CD8‡ CD28 CD57‡ subpopulation. Clin. Immunol. 96(3), 230±235. Batliwalla, F., Monteiro, J., Serrano, D., and Gregersen, P. K. (1996). Oligoclonality of CD8 ‡ T cells in health and disease: Aging, infection, or immune regulation? Hum. Immunol. 48(1±2), 68±76. Batliwalla, F. M., Rufer, N., Lansdorp, P. M., and Gregersen, P. K. (2000). Oligoclonal expansions in the CD8( ‡ )CD28( ) T cells largely explain the shorter telomeres detected in this subset: Analysis by ¯ow FISH. Hum. Immunol. 61(10), 951±958. Bernstein, E. D., and Murasko, D. M. (1998). Effect of age on cytokine production in humans. Age 21, 137±151. Bhatia, S. K., Tygrett, L. T., Grabstein, K. H., and Waldschmidt, T. J. (1995). The effect of in vivo IL7 deprivation on T cell maturation. J. Exp. Med. 181(4), 1399±1409. Blasko, I., Marx, F., Steiner, E., Hartmann, T., and Grubeck-Loebenstein, B. (1999). TNFa plus IFNg induce the production of Alzheimer b-amyloid peptides and decrease the secretion of APPs. FASEB J. 13(1), 63±68. Blasko, I., Schmitt, T. L., Steiner, E., Trieb, K., and Grubeck-Loebenstein, B. (1997). Tumor necrosis factor a augments amyloid b protein (25±35) induced apoptosis in human cells. Neurosci. Lett. 238(1±2), 17±20. Blasko, I., Veerhuis, R., Stampfer-Kountchev, M., Saurwein-Teissl, M., Eikelenboom, P., and Grubeck-Loebenstein, B. (2000). Costimulatory effects of interferon-g and interleukin-1b or tumor necrosis factor a on the synthesis of Ab1±40 and Ab1±42 by human astrocytes. Neurobiol. Dis. 7(6), 682±689. Bloom, E. T., Kubota, L. F., and Kawakami, K. (1988). Age-related decline in the lethal hit but not the binding stage of cytotoxic T-cell activity in mice. Cell. Immunol. 114(2), 440±446. Borrego, F., Alonso, M. C., Galiani, M. D., Carracedo, J., Ramirez, R., Ostos, B., Pena, J., and Solana, R. (1999). NK phenotypic markers and IL2 response in NK cells from elderly people. Exp. Gerontol. 34(2), 253±265. Boucher, N., Dufeu-Duchesne, T., Vicaut, E., Farge, D., Effros, R. B., and Schachter, F. (1998). CD28 expression in T cell aging and human longevity. Exp. Gerontol. 33(3), 267±282. Brownlee, M., Vlassara, H., Kooney, A., Ulrich, P., and Cerami, A. (1986). Aminoguanidine prevents diabetes-induced arterial wall protein cross-linking. Science 232(4758), 1629±1632. Bruunsgaard, H., Pedersen, A. N., Schroll, M., Skinhoj, P., and Pedersen, B. K. (2000). TNF a, leptin, and lymphocyte function in human aging. Life Sci. 67(22), 2721±2731. Bruunsgaard, H., Pedersen, M., and Pedersen, B. K. (2001). Aging and proin¯ammatory cytokines. Curr. Opin. Hematol. 8(3), 131±136. Bucay, N., Sarosi, I., Dunstan, C. R., Morony, S., Tarpley, J., Capparelli, C., Scully, S., Tan, H. L., Xu, W., Lacey, D. L., Boyle, W. J., and Simonet, W. S. (1998). Osteoprotegerin-de®cient mice develop early onset osteoporosis and arterial calci®cation. Genes Dev. 12(9), 1260±1268. Burgess, T. L., Qian, Y., Kaufman, S., Ring, B. D., Van, G., Capparelli, C., Kelley, M., Hsu, H., Boyle, W. J., Dunstan, C. R., Hu, S., and Lacey, D. L. (1999). The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145(3), 527±538. Callahan, J. E., Kappler, J. W., and Marrack, P. (1993). Unexpected expansions of CD8-bearing cells in old mice. J. Immunol. 151(12), 6657±6669. Campisi, J. (1998). The role of cellular senescence in skin aging. J. Invest. Dermatol. Symp. Proc. 3(1), 1±5. Carnes, B. A., and Olshansky, S. J. (2001). Heterogeneity and its biodemographic implications for longevity and mortality. Exp. Gerontol. 36(3), 419±430.

the aging of the immune system

275

Catania, A., Airaghi, L., Motta, P., Manfredi, M. G., Annoni, G., Pettenati, C., Brambilla, F., and Lipton, J. M. (1997). Cytokine antagonists in aged subjects and their relation with cellular immunity. J. Gerontol. A: Biol. Sci. Med. Sci. 52(2), 93±97. Chamberlain, W. D., Falta, M. T., and Kotzin, B. L. (2000). Functional subsets within clonally expanded CD8(‡) memory T cells in elderly humans. Clin. Immunol. 94(3), 160±172. Champagne, P., Ogg G. S., King A. S., Knabenhans C., Ellefsen, K., Nobile M., Appay V., Rizzardi G. P., Fleury S., Lipp M., FoÈrster R., Rowland-Jones S., SeÂkaly R. P., McMichael A. J., and Pantaleo G. (2001). Skewed maturation of memory HIV-speci®c CD8 T lymphocytes. Nature 410, 106±111. Christenson, B., Lundbergh, P., Hedlund, J., and Ortqvist, A. (2001). Effects of a large-scale intervention with in¯uenza and 23-valent pneumococcal vaccines in adults aged 65 years or older: A prospective study. Lancet 357(9261), 1008±1011. Cossarizza, A., Ortolani, C., Paganelli, R., Barbieri, D., Monti, D., Sansoni, P., Fagiolo, U., Castellani, G., Bersani, F., Londei, M., and Franceschi, C. (1996). CD45 isoforms expression on CD4‡ and CD8‡ T cells throughout life, from newborns to centenarians: Implications for T cell memory. Mech. Ageing. Dev. 86, 173±195. DeHaan, G., and VanZant, G. (1999). Genetic analysis of hemopoietic cell cycling in mice suggests its involvement in organismal life span. FASEB J. 13, 707±713. Diamondstone, L. S., Tollerud, D. J., Fuchs, D., Wachter, H., Brown, L. M., Maloney, E., Kurman, C. C., Nelson, D. L., and Blattner, W. A. (1994). Factors in¯uencing serum neopterin and b2 microglobulin levels in a healthy diverse population. J. Clin. Immunol. 14(6), 368±374. Douek, D. C., and Koup, R. A. (2000). Evidence for thymic function in the elderly. Vaccine 18, 1638±1641. Douek, D. C., McFarland, R. D., Keiser, P. H., Gage, E. A., Massey, J. M., Haynes, B. F., Polis, M. A. Haase, A. T., Feinberg, M. B., Sullivan, J. L., Jamieson, B. D., Zack, J. A., Picker, L. J., and Koup, R. A. (1998). Changes in thymic function with age and during the treatment of HIV infection. Nature 396, 690±695. Effros, R. B. (1997). Loss of CD28 expression on T lymphocytes: A marker of replicative senescence. Dev. Comp. Immunol. 21(6), 471±478. Effros, R. B. (2000). Costimulatory mechanisms in the elderly. Vaccine 8, 1661±1663. Eikelenboom, P., Zhan, S. S., Kamphorst, W., van der Valk, P., and Rozemuller, J. M. (1994). Cellular and substrate adhesion molecules (integrins) and their ligands in cerebral amyloid plaques in Alzheimer's disease. Virchows Arch. 424(4), 421±427. Ernst, D. N. Hobbs, M. V., Torbett, B. E., Glasebrook, A. L., Rehse, M. A., Bottomly, K., Hayakawa, K., Hardy, R. R., and Weigle, W. O. (1990). Differences in the expression pro®les of CD45RB, Pgp-1, and 3G11 membrane antigens and in the patterns of lymphokine secretion by splenic CD4‡ T cells from young and aged mice. J. Immunol. 145, 1295±1302. Ershler, W. B., and Keller, E. T. (2000). Age-associated increased interleukin-6 gene expression, late-life diseases, and frailty. Annu. Rev. Med. 51, 245±270. Faccini, A., Mariani, E., Mariani, A. R., Papa, S., Vitale, M., and Manzoli, F. A. (1987). Increased number of circulating Leu 11‡ (CD16) large granular lymphocytes and decreased NK activity during human ageing. Clin. Exp. Immunol. 68, 340±370. Fagnoni, F. F., Vescovini, R., Passeri, G., Bologna, G., Pedrazzoni, M., Lavagetto, G., Casti, A., Franceschi, C., Passeri, M., and Sansoni, P. (2000). Shortage of circulating naõÈve CD8(‡) T cells provides new insights on immunode®ciency in aging. Blood 95(9), 2860±2868. Fernandez-Gutierrez, B., Jover, J. A., De Miguel, S., Hernandez-Garcia, C., Vidan, M. T., Ribera, J. M., Banares, A., and Serra, J. A. (1999). Early lymphocyte activation in elderly humans: Impaired T and T-dependent B cell responses. Exp. Gerontol. 34(2), 217±229. Finkel, T., and Holbrook, N. J. (2000). Oxidants, oxidative stress and the biology of ageing. Nature 408(6809), 239±247.

276

b. grubeck-loebenstein and g. wick

Fitzgerald, J. E., Ricalton, N. S., Meyer, A. C., West, S. G., Kaplan, H., Behrendt, C., and Kotzin, B. L. (1995). Analysis of clonal CD8‡ T cell expansions in normal individuals and patients with rheumatoid arthritis. J. Immunol. 154(7), 3538±3547. Flurkey, K., Stadecker, M., and Miller, R. A. (1992). Memory T lymphocyte hyporesponsiveness to non-cognate stimuli: A key factor in age-related immunode®ciency. Eur. J. Immunol. 22(4), 931±935. Franceschi, C., Bonafe, M., Valensin, S., Olivieri, F., De Luca, M., Ottaviani, E., and De Benedictis, G. (2000a). In¯amm-aging: An evolutionary perspective on immunosenescence. Ann. N. Y. Acad. Sci. 908, 244±254. Franceschi, C., Valensin, S., Bonafe, M., Paolisso, G., Yashin, A. I., Monti, D., and De Benedictis, G. (2000b). The network and the remodeling theories of aging: Historical background and new perspectives. Exp. Gerontol. 35(6±7), 879±896. Gamble, D. A., Schwab, R., Weksler, M. E., and Szabo, P. (1990). Decreased steady state c-myc mRNA in activated T cell cultures from old humans is caused by a smaller proportion of T cells that transcribe the c-myc gene. J. Immunol. 144(9), 3569±3573. Garcia, G. G., and Miller, R. A. (1995). Increased Zap-70 association with CD3z in CD4 T cells from old mice. Cell. Immunol. 190(2), 91±100. Garcia, G. G., and Miller, R. A. (1997). Differential tyrosine phosphorylation of z chain dimers in mouse CD4 T lymphocytes: Effect of age. Cell. Immunol. 175(1), 51±57. Garcia, G. G., and Miller, R. A. (2001). Single-cell analyses reveal two defects in peptide-speci®c activation of naõÈve T cells from aged mice. J. Immunol. 166(5), 3151±3157. George, A. J., and Ritter, M. A. (1996). Thymic involution with ageing: Obsolescence or good housekeeping? Immunol. Today 17(6), 267±272. Gilchrest, B. A., Murphy, G. F., and Soter, N. A. (1982). Effect of chronologic aging and ultraviolet irradiation on Langerhans cells in human epidermis. J. Invest. Dermatol. 79(2), 85±88. Gitter, B. D., Cox, L. M. Rydel, R. E., and May, P. C. (1995). Amyloid b peptide potentiates cytokine secretion by interleukin-1 b-activated human astrocytoma cells. Proc. Natl. Acad. Sci. USA 92(23), 10738±10741. Globerson, A. (1995). T lymphocytes and aging. Int. Arch. Allergy Immunol. 107(4), 491±497. Globerson, A. (1999). Hematopoietic stem cells and aging. Exp. Gerontol. 34(2), 137±146. Godfrey, D. I., Hammond, K. J., Poulton, L. D., Smyth, M. J., and Baxter, A. G. (2000). NKT cells: Facts, functions and fallacies. Immunol. Today 21(11), 573±583. Goidl, E. A., Innes, J. B., and Weksler, M. E. (1976). Immunological studies of aging. II. Loss of IgG and high avidity plaque-forming cells and increased suppressor cell activity in aging mice. J. Exp. Med. 144(4), 1037±1048. Gorgas, G., Butch, E. R., Guan, K. L., and Miller, R. A. (1997). Diminished activation of the MAP kinase pathway in CD3-stimulated T lymphocytes from old mice. Mech. Ageing Dev. 94(1±3), 71±83. Gottesman, S. R., Edington, J. M., and Thorbecke, G. J. (1988). Proliferative and cytotoxic immune functions in aging mice. IV. Effects of suppressor cell populations from aged and young mice. J. Immunol. 140(6), 1783±1790. Gravenstein, S., Fillit, H., and Ershler, W. B. (1998). Clinical Immunology of Aging. In ``Geriatric Medicine and Gerontology'' (R. Tallis, H. Fillit, and J. C. Brocklehurst, Eds.), pp. 109±121. Grossmann, A., Maggio-Price, L., Jinneman, J. C., and Rabinovitch, P. S. (1991). In¯uence of aging on intracellular free calcium and proliferation of mouse T-cell subsets from various lymphoid organs. Cell. Immunol. 135, 118±131. Grubeck-Leobenstein, B., Berger, P., Saurwein-Teissl, M., Zisterer, K., and Wick, G. (1998). No immunity for the elderly. Nat. Med. 4, 870.

the aging of the immune system

277

Grubeck-Loebenstein, B., Lechner, H., and Trieb, K. (1994). Long-term in vitro growth of human T cell clones: Can postmitotic `senescent' cell populations be de®ned? Int. Arch. Allergy Immunol. 104(3), 232±239. Gupta, S. (1989). Membrane signal transduction in T cells in aging humans. Ann. N. Y. Acad. Sci. 568, 277±282. Hakim, F. T., Cepeda, R., and Kaimei, S. (1997). Constraints on CD4 recovery post chemotherapy in adults: Thymic insuf®ciency and apoptotic decline of expanded peripheral T cells. Blood 90, 3789±3798. Hamann, D., Baars, P. A., Rep, M. H., Hooibrink, B., Kerkhof-Garde, S. R., Klein, M. R., and van Lier, R. A. (1997). Phenotypic and functional separation of memory and effector human CD8‡ T cells. J. Exp. Med. 186(9), 1407±1418. Harley, C. B., Futcher, A. B., and Greider, C. W. (1990). Telomeres shorten during ageing of human ®broblasts. Nature 345(6274), 458±460. Hart, R. W., and Setlow, R. B. (1974). Correlation between deoxyribonucleic acid excision-repair and life-span in a number of mammalian species. Proc. Natl. Acad. Sci. USA 71(6), 2169±2173. Hay¯ick, L., and Moorehead, P. S. (1961). The serial cultivation of human diploid cell strains. Exp. Cell Res. 25, 585±621. Haynes, L., Eaton, S. M., and Swain, S. L. (2000). The defects in effector generation associated with aging can be reversed by addition of IL-2 but not other related gc-receptor binding cytokines. Vaccine 18, 1649±1653. Haynes, L., Linton, P. J., Eaton, S. M., Tonkonogy, S. L., and Swain, S. L. (1999). Interleukin 2, but not other common g chain-building cytokines, can reverse the defect in generation of CD4 effector T cells from naõÈve T cells of aged mice. J. Exp. Med. 190(7), 1013±1024. Herrero, C., Marques, L., Lloberas, J., and Celada, A. (2001). IFN-g-dependent transcription of MHC class II IA is impaired in macrophages from aged mice. J. Clin. Invest. 107(4), 485±493. Hingorani, R., Choi, I. H., Akolkar, P., Gulwani-Akolkar, B., Pergolizzi, R., Silver, J., and Gregersen, P. K. (1993). Clonal predominance of T cell receptors within the CD8‡ CD45RO‡ subset in normal human subjects. J. Immunol. 151(10), 5762±5769. Hirokawa, K. (1999). Age-related changes of signal transduction in T cells. Exp. Gerontol. 34(1), 7±18. Huber, L. A., Xu, Q. B., JuÈrgens, G., BoÈck, G., Buhler, E., Gey, K. F., SchoÈnitzer, D., Traill, K. N., and Wick, G. (1991). Correlation of lymphocyte lipid composition membrane microviscosity and mitogen response in the aged. Eur. J. Immunol. 21, 2761±2765. Itagaki, S., McGeer, P. L., Akiyama, H., Zhu, S., and Selkoe, D. (1989). Relationship of microglia and astrocytes to amyloid deposits of Alzheimer disease. J. Neuroimmunol. 24(3), 173±182. Jamieson, B. D., Douek, D. C., Killian, S., Hultin, L. E., Scripture-Adams, D. D., Giorgi, J. V., Marelli, D., Koup, R. A., and Zack, J. A. (1999). Generation of functional thymocytes in the human adult. Immunity 10(5), 569±575. Jansen-DuÈrr, P. (1998). The making and the breaking for senescence: Changes of gene expression during cellular aging and immortalization. Exp. Gerontol. 33, 291±301. Jazwinski, S. M. (1998). Genetics of longevity. Exp. Gerontol. 33(7±8), 773±783. JuÈrgens, G., Xu, Q., Huber, L. A., BoÈck, G., Howanietz, H., Wick, G., and Traill, K. N. (1989). Promotion of lymphocyte growth by high density lipoproteins (HDL): Physiological signi®cance of the HDL binding site. J. Biol. Chem. 264(15), 8549±8556. Kampinga, J., Groen, H., Klatter, F. A., Pater, J. M., van Petersen, A. S., Roser, B., Nieuwenhuis, P., and Aspinall, R. (1997). Post-thymic T-cell development in the rat. Thymus 24(3), 173±200. Kiechl, S., Egger, G., Mayr, M., Wiedermann, C. J., Bonora, E., Oberhollenzer, F., Muggeo, M., Xu, Q., Wick, G., Poewe, W., and Willeit, J. (2001). Chronic infections and the risk of carotid

278

b. grubeck-loebenstein and g. wick

atherosclerosis: Prospective results from a large population study. Circulation 103(8), 1064±1070. Kong, Y. Y., Feige, U., Sarosi, I., Bolon, B., Tafuri, A., Morony, S., Capparelli, C., Li, J., Elliott, R., McCabe, S., Wong, T., Campagnuolo, G., Moran, E., Bogoch, E. R., Van, G., Nguyen, L. T., Ohashi, P. S., Lacey, D. L., Fish, E., Boyle, W. J., and Penninger, J. M. (1999a). Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402(6759), 304±309. Kong, Y. Y., and Penninger, J. M. (2000). Molecular control of bone remodeling and osteoporosis. Exp. Gerontol. 35(8), 947±956. Kong, Y. Y., Yoshida, H., Sarosi, I., Tan, H. L., Timms, E., Capparelli, C., Morony, S., OliveiradosSantos, A. J., Van, G., Itie, A., Khoo, W., Wakeham, A., Dunstan, C. R., Lacey, D. L., Mak, T. W., Boyle, W. J., and Penninger, J. M. (1999b). OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397(6717), 315±323. Kosco, M. H., Burton, G. F., Kapasi, Z. F., Szakal, A. K., and Tew, J. G. (1989). Antibody-forming cell induction during an early phase of germinal centre development and its delay with ageing. Immunology 68(3), 312±318. Lacey, D. L., Timms, E., Tan, H. L., Kelley, M. J., Dunstan, C. R., Burgess, T., Elliott, R., Colombero, A., Elliott, G., Scully, S., Hsu, H., Sullivan, J., Hawkins, N., Davy, E., Capparelli, C., Eli, A., Qian, Y. X., Kaufman, S., Sarosi, I., Shalhoub, V., Senaldi, G., Guo, J., Delaney, J., and Boyle, W. J. (1998). Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93(2), 165±176. Lechner, O., Dietrich, H., Wiegers, G. J., Vacchio, M., and Wick, G. (2001). Glucocorticoid production in the chicken bursa and thymus. Int. Immunol. 13(6), 769±776. Lechner, O., Wiegers, G. J., Oliveira-Dos-Santos, A. J., Dietrich, H., Recheis, H., Waterman, M., Boyd, R., and Wick, G. (2000). Glucocorticoid production in the murine thymus. Eur. J. Immunol. 30(2), 337±346. Lee, C. K., Klopp, R. G., Weindruch, R., and Prolla, T. A. (1999). Gene expression pro®le of aging and its retardation by caloric restriction. Science 285(5432), 1390±1393. LeMaoult, J., Manavalan, J. S., Dyall, R., Szabo, P., Nikolic-Zugic, J., and Weksler, M. E. (1999). Cellular basis of B cell clonal populations in old mice. J. Immunol. 162(11), 6384±6391. Lerner, A., Yamada, T., and Miller, R. A. (1990). Pgp-1hi T lymphocytes accumulate with age in mice and respond poorly to concanavalin A. J. Immunol. 145, 1295±1302. Lighthart, G. J. (1984). Admission criteria for immunogerontological studies in man: The Senieur Protocol. Mech. Ageing Dev. 28, 47±55. Linton, P. J., Haynes, L., Klinman, N. R., and Swain, S. L. (1996). Antigen independent changes in CD4 T cells with aging. J. Exp. Med. 184, 1891±1900. Lio, D., Candore, G., Cigna, D., D'Anna, C., Di Lorenzo, G., Giordano, C., Lucania, G., Mansueto, P., Melluso, M., Modica, M. A., and Caruso, C. (1996). In vitro T cell activation in elderly individuals: Failure in CD69 and CD71 expression. Mech. Ageing Dev. 89(1), 51±58. Lung, T. L., Saurwein-Teissl, M., Parson, W., SchoÈnitzer, D., and Grubeck-Loebenstein, B. (2000). Unimpaired dendritic cells can be derived from monocytes in old age and can mobilize residual function in senescent T cells. Vaccine 18, 1606±1612. Mackall, C. L., Fleisher, T. A., Brown, M. R., Andrich, M. P., Chen, C. C., Feuerstein, I. M., Horowitz, M. E., Magrath, I. T., Shad, A. T., and Steinberg, S. M. (1995). Age, thymopoiesis, and CD4‡ T-lymphocyte regeneration after intensive chemotherapy. N. Engl. J. Med. 332(3), 143±149. Mackall, C. L., and Gress, R. E. (1997). Thymic aging and T-cell regeneration. Immunol. Rev. 160, 91±102. Mariani, E., Roda, P., Mariani, A. R., Vitale, M., Degrassi, A., Papa, S., and Facchini, A. (1990). Ageassociated changes in CD8‡ and CD16‡ cell reactivity: Clonal analysis. Clin. Exp. Immunol. 81(3), 479±484.

the aging of the immune system

279

Masoro, E. J. (2000). Caloric restriction and aging: An update. Exp. Gerontol. 35, 299±305. Mayr, M., Kiechl, S., Willeit, J., Wick, G., and Xu, Q. (2000). Infections, immunity, and atherosclerosis: Associations of antibodies to Chlamydia pneumoniae, Helicobacter pylori, and cytomegalovirus with immune reactions to heat-shock protein 60 and carotid or femoral atherosclerosis. Circulation 102(8), 833±839. Mbawuike, I. N., Lange, A. R., and Couch, R. B. (1993). Diminished in¯uenza A virus-speci®c MHC class I-restricted cytotoxic T lymphocyte activity among elderly persons. Viral Immunol. 6(1), 55±64. McCarron, M., Osborne, Y., Story, C. J., Dempsey, J. L., Turner, D. R., amd Morley, A. A. (1987). Effect of age on lymphocyte proliferation. Mech Ageing Dev. 41, 211±218. McGeer, P. L., Kawamata, T., Walker, D. G., Akiyama, H., Tooyama, I., and McGeer, E. G. (1993). Microglia in degenerative neurological disease. Glia 7(1), 84±92. McGeer, P. L., McGeer, E., Rogers, J., and Sibley, J. (1990). Anti-in¯ammatory drugs and Alzheimer disease. Lancet 335(8696), 1037. McGeer, P. L., Schulzer, M., and McGeer, E. G. (1996). Arthritis and anti-in¯ammatory agents as possible protective factors for Alzheimer's disease: A review of 17 epidemiologic studies. Neurology 47(2), 425±432. Meda, L., Cassatella, M. A., Szendrei, G. I., Otvos, L., Jr, Baron, P., Villalba, M., Ferrari, D., and Rossi, F. (1995). Activation of microglial cells by b-amyloid protein and interferon-g. Nature 374(6523), 647±650. Miller, C., Stedra, J., Kelsoe, G., and Cerny, J. (1995). Facultative role of germinal centers and T cells in the somatic diversi®cation of IgVH genes. J. Exp. Med. 181(4), 1319±1331. Miller, R. A. (2000). Effect of aging on T lymphocyte activation. Vaccine 18, 1654±1660. Miller, R. A. (1984). Age-associated decline in precursor frequency for different T cell-mediated reactions, with preservation of helper or cytotoxic effect per precursor cell. J. Immunol. 132(1), 63±68. Miller, R. A. (1995). Immune system. In ``Handbook of Physiology'' (E. Marsoro, Ed.), pp. 555±590. Oxford Univ. Press, London. Miller, R. A. (1996a). Aging and the Immune response. In ``Handbook of the Biology of Aging'' (E. L. Schneider and J. W. Rowe, Eds.), pp. 355±392. Academic Press, San Diego. Miller, R. A. (1996b). The aging immune system: Primer and prospectus. Science 273, 70±74. Miller, R. A. (2001). Biomarkers of aging: Prediction of longevity by using age-sensitive T-cell subset determinations in a middle-aged, genetically heterogeneous mouse population. J. Gerontol. A: Biol. Sci. Med. Sci. 56(4), 180±186. Mitchell, J., Pye, J., Holmes, M. C., and Nossal, G. J. (1972). Antigens in immunity: Antigen localization in congenitally athymic ``nude'' mice. Aust. J. Exp. Biol. Med. Sci. 50(5), 637±650. Miyaji, C., Watanabe, H., Toma, H., Akisaka, M., Tomiyama, K., Sato, Y., and Abo, T. (2000). Functional alteration of granulocytes, NK cells, and natural killer T cells in centenarians. Hum. Immunol. 61(9), 908±916. Mjaaland, S., and Fossum, S. (1987). The localization of antigen in lymph node follicles of congenitally athymic nude rats. Scand. J. Immunol. 26(2), 141±147. Monteiro, J., Batliwalla, F., Ostrer, H., and Gregersen, P. K. (1996). Shortened telomeres in clonally expanded CD28 CD8‡ T cells imply a replicative history that is distinct from their CD28 ‡ CD8‡ counterparts. J. Immunol. 156(10), 3587±3590. Monteiro, J., Hingorani, R., Choi, I. H., Silver, J., Pergolizzi, R., and Gregersen, P. K. (1995). Oligoclonality in the human CD8‡ T cell repertoire in normal subjects and monozygotic twins: Implications for studies of infectious and autoimmune diseases. Mol. Med. 1(6), 614±624. Morimoto, R. I. (1993). Cells in stress: Transcriptional activation of heat shock genes. Science 259, 1409±1410.

280

b. grubeck-loebenstein and g. wick

Morrison, S. J., Wandycz, A. M., Akashi, K., Globerson, A., and Weissmann, I. L. (1996). The aging of hematopoietic stem cells. Nat. Med. 2(9), 1011±1016. Nagel, J. E., Chopra, R. K., Chrest, F. J., McCoy, M. T., Schneider, E. L., Holbrook, N. J., and Adler, W. H. (1988). Decreased proliferation, interleukin 2 synthesis, and interleukin 2 receptor expression are accompanied by decreased mRNA expression in phytohemagglutinin-stimulated cells from elderly donors. J. Clin. Invest. 81(4), 1096±1102. Nakagawa, N., Kinosaki, M., Yamaguchi, K., Shima, N., Yasuda, H., Yano, K., Morinaga, T., and Higashio, K. (1998). RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253(2), 395±400. Nesse, R. M., and Williams, G. C. (1995). ``Evolution and Healing. The New Science of Darwinian Medicine.'' Phoenix, Orion Books, London. Nichol, K. L., Margolis, K. L., Wuorenma, J., and Von Sternberg, T. (1994). The ef®cacy and cost effectiveness of vaccination against in¯uenza among elderly persons living in the community. N. Engl. J. Med. 331(12), 778±784. Nociari, M. M., Telford, W., and Russo, C. (1999). Postthymic development of CD28 CD8‡ T cell subset: Age-associated expansion and shift from memory to naõÈve phenotype. J. Immunol. 162(6), 3327±3335. Nordin, A. A., and Collins, G. D. (1983). Limiting dilution analysis of alloreactive cytotoxic precursor cells in aging mice. J. Immunol. 131(5), 2215±2218. Oosterwegel, M. A., Haks, M. C., Jeffry, U., Murray, R., and Kruisbeek, A. M. (1997). Induction of TCR gene rearrangements in uncommitted stem cells by a subset of IL-7 producing, MHC classII-expressing thymic stromal cells. Immunity 6(3), 351±360. Ouyang, Q., Cicek, G., Westendorp, R. G., Cools, H. J., van der Klis, R. J., and Remarque, E. J. (2000). Reduced IFN-g production in elderly people following in vitro stimulation with in¯uenza vaccine and endotoxin. Mech. Ageing Dev. 121(1±3), 131±137. Pahlavani, M. A., Harris, M. D., and Richardson, A. (1998). Activation of p21ras/MAPK signal transduction molecules decreases with age in mitogen-stimulated T cells from rats. Cell. Immunol. 185(1), 39±48. Paresce, D. M., Ghosh, R. N., and Max®eld, F. R. (1996). Microglial cells internalize aggregates of the Alzheimer's disease amyloid b-protein via a scavenger receptor. Neuron 17(3), 553±565. Pawelec, G., Adibzadeh, M., Rehbein, A., Hahnel, K., Wagner, W., and Engel, A. (2000). In vitro senescence models for human T lymphocytes. Vaccine 18(16), 1666±1674. Pawelec, G., Rehbein, A., Haehnel, K., Merl, A., and Adibzadeh, M. (1997). Human T-cell clones in long-term culture as a model of immunosenescence. Immunol. Rev. 160, 31±42. Perillo, N. L., Naeim, F., Walford, R. L., and Effros, R. B. (1993). In vitro cellular aging in Tlymphocyte cultures: Analysis of DNA content and cell size. Exp. Cell Res. 207(1), 131±135. Pietschmann, P., Hahn, P., Kudlacek, S., Thomas, R., and Peterlik, M. (2000). Surface markers and transendothelial migration of dendritic cells from elderly subjects. Exp. Gerontol. 35(2), 213±224. Pilarski, L. M., Paine, D., McElhaney, J. E., Cass, C. E., and Belch, A. R. (1995). Multidrug transporter P-glycoprotein 170 as a differentiation antigen on normal human lymphocytes and thymocytes: Modulation with differentiation stage and during aging. Am. J. Hematol. 49(4), 323±335. Pioli, C., Pucci, S., Barile, S., Frasca, D., and Doria, G. (1998). Role of mRNA stability in the different patterns of cytokine production by CD4‡ cells from young and old mice. Immunology 94(3), 380±387. Posnett, D. N., Sinha, R., Kabak, S., and Russo, C. (1994). Clonal populations of T cells in normal elderly humans: The T cell equivalent to ``benign monoclonal gammapathy.'' J. Exp. Med. 179(2), 609±618.

the aging of the immune system

281

Poulin, J. F., Viswanathan, M. N., Harris, J. M., Komanduri, K. V., Wieder, E., Ringuette, N., Jenkins, M., McCune, J. M., and Sekaly, R. P. (1999). Direct evidence for thymic function in adult humans. J. Exp. Med. 190(4), 479±486. Poznansky, M. C., Evans, R. H., Foxall, R. B., Olszak, I. T., Piascik, A. H., Hartman, K. E., Brander, C., Meyer, T. H., Pykett, M. J., Chabner, K. T., Kalams, S. A., Rosenzweig, M., and Scadden, D. T. (2000). Ef®cient generation of human T cells from a tissue-engineered thymic organoid. Nat. Biotechnol. 18(7), 729±734. Radl, J. (1990). Age-related monoclonal gammapathies: Clinical lessons from the aging C57BL mouse. Immunol. Today 11(7), 234±236. Robinson, K. A., Baughman, W., Rothrock, G., Barrett, N. L., Pass, M., Lexau, C., Damaske, B., Stefonek, K., Barnes, B., Patterson, J., Zell, E. R., Schuchat, A., and Whitney, C. G. (2001). Epidemiology of invasive Streptococcus pneumoniae infections in the United States, 1995±1998: Opportunities for prevention in the conjugate vaccine era. JAMA 285(13), 1729±1735. Rowley, M. J., Buchanan, H., and Mackay, I. R. (1968). Reciprocal change with age in antibody to extrinsic and intrinsic antigens. Lancet 2, 24±26. Sadeghi, H. M., Schnelle, J. F., Thoma, J. K., Nishanian, P., and Fahey, J. L. (1999). Phenotypic and functional characteristics of circulating monocytes of elderly persons. Exp. Gerontol. 34(8), 959±970. Saurwein-Teissl, M., Blasko, I., Zisterer, K., Neuman, B., Lang, B., and Grubeck-Loebenstein, B. (2000). An imbalance between pro- and anti-in¯ammatory cytokines, a characteristic feature of old age. Cytokine 12(7), 1160±1161. Saurwein-Teissl, M., SchoÈnitzer, D., and Grubeck-Loebenstein, B. (1998). Dendritic cell responsiveness to stimulation with in¯uenza vaccine is unimpaired in old age. Exp. Gerontol. 33, 625±631. Saxena, R. K., Saxena, Q. B., and Adler, W. H. (1998). Lectin-induced cytotoxic activity in spleen cells from young and old mice: Age-related changes in types of effector cells, lymphokine production and response. Immunology 64(3), 457±461. Schett, G., Xu, Q., Amberger, A., Van der Zee, R., Recheis, H., Willeit, J., and Wick, G. (1995). Autoantibodies against heat shock protein 60 mediate endothelial cytotoxicity. J. Clin. Invest. 96(6), 2569±2577. Schirmer, M., Vallejo, A. N., Weyand, C. M., and Goronzy, J. J. (1998). Resistance to apoptosis and elevated expression of Bcl-2 in clonally expanded CD4 ‡ CD28 T cells from rheumatoid arthritis patients. J. Immunol. 161(2), 1018±1025. Schwab, R., Hausman, P. B., Rinnooy-Kann, E., and Weksler, M. E. (1985). Immunological studies of ageing. X. Impaired T lymphocytes and normal monocyte response from elderly humans to the mitogenic antibodies OKT3 and Leu 4. Immunology 55(4), 677±684. Schwab, R., Szabo, P., Manavalan, J. S., Weksler, M. E., Posnett, D. N., Pannetier, C., Kourilsky, P., and Even, J. (1997). Expanded CD4‡ and CD8‡ T cell clones in elderly humans. J. Immunol. 158(9), 4493±4499. Schwarz, S., Berger, P., Frisch, H., Moncayo, R., Phillips, J. A., 3rd, and Wick, G. (1987). Growth hormone blocking antibodies in a patient with deletion of the GH-N gene. Clin. Endocrinol. 27(2), 213±224. Shahinian, A., Pfeffer, K., Lee, K. P., Kundig, T. M., Kishihara, K., Wakeham, A., Kawai, K., Ohashi, P. S., Thompson, C. B., and Mak, T. W. (1993). Differential T cell costimulatory requirements in CD28-de®cient mice. Science 261(5121), 609±612. Shearer, G. M. (1997). Th1/Th2 changes in aging. Mech. Ageing Dev. 94, 1±5. Simonet, W. S., Lacey, D. L., Dunstan, C. R., Kelley, M., Chang, M. S., Luthy, R., Nguyen, H. Q., Wooden, S., Bennett, L., Boone, T., Shimamoto, G., DeRose, M., Elliott, R., Colombero, A., Tan, H. L., Trail, G., Sullivan, J., Davy, E., Bucay, N., Renshaw-Gegg, L., Hughes, T. M., Hill, D.,

282

b. grubeck-loebenstein and g. wick

Pattison, W., Campbell, P., and Boyle, W. J. (1997). Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 89(2), 159±161. Sohal, R. S. (1988). Effect of hydrogen peroxide administration on life span, superoxide dismutase, catalase, and glutathione in the adult house¯y, Musca domestica. Exp. Gerontol. 23(3), 211±216. Solana, R., and Mariani, E. (2000). NK and NK/T cells in human senescence. Vaccine 18(16), 1613±1620. Spaulding, C., Guo, W., and Effros, R. B. (1999). Resistance to apoptosis in human CD8‡ T cells that reach replicative senescence after multiple rounds of antigen-speci®c proliferation. Exp. Gerontol. 34(5), 633±644. Sprecher, E., Becker, Y., Kraal, G., Hall, E., Harrison, D., and Shultz, L. D. (1990). Effect of aging on epidermal dendritic cell populations in C57BL/6J mice. J. Invest. Dermatol. 94(2), 247±253. Steger, M. M., Maczek, C., Berger, P., and Grubeck-Loebenstein, B. (1996a). Vaccination against tetanus in the elderly: Do recommended vaccination strategies give suf®cient protection. Lancet 348(9029), 762. Steger, M. M., Maczek, C., Berger, P., and Grubeck-Loebenstein, B. (1997a). The immune response to tetanus in elderly people: Shortened immunological memory affects the duration of protection. Wien. Klin. Wochenschr. 109(19), 767±770. Steger, M. M., Maczek, C., and Grubeck-Loebenstein, B. (1996b). Morphologically and functionally intact dendritic cells can be derived from the peripheral blood of aged individuals. Clin. Exp. Immunol. 105(3), 544±550. Steger, M. M., Maczek, C., and Grubeck-Loebenstein, B. (1997b). Peripheral blood dendritic cells reinduce proliferation in in vitro aged T cell populations. Mech. Ageing Dev. 93(1±3), 125±130. Stein, B. E. (1994). Vaccinating elderly people. Protecting from avoidable disease. Drugs Aging 5(4), 242±253. Steinmann, G. (1986). Changes in the human thymus during aging. Curr. Top. Pathol. 75, 43±80. Stulnig, T., Maczek, C., BoÈck, G., Majdic, O., and Wick, G. (1995a). Reference intervals for human peripheral blood lymphocyte subpopulations from ``healthy'' young and aged subjects. Int. Arch. Allergy Immunol. 108, 205±210. Stulnig, T., Mair, A., Jarosch, E., Schober, M., Schonitzer, D., Wick, G., and Huber, L. A. (1993). Estimation of reference intervals from a SENIEUR protocol compatible aged population for immunogerontological studies. Mech. Ageing Dev. 68(1±3), 105±115. Stulnig, T. M., Klocker, H., Harwood, H. J., Jr., Jurgens, G., Schonitzer, D., Jarosch, E., Huber, L. A., Amberger, A., and Wick, G. (1995b). In vivo LDL receptor and HMG-CoA reductase regulation in human lymphocytes and its alterations during aging. Arterioscler. Thromb. Vasc. Biol. 15(7), 872±878. Swain, S. L., Croft, M., and Dubey, C. (1996). From naõÈve to memory T cells. Immunol. Rev. 150, 143±167. Szakal, A. K., Taylor, J. K., Smith, J. P., Kosco, M. H., Burton, G. F., and Tew, J. J. (1990). Kinetics of germinal center development in lymph nodes of young and aging immune mice. Anat. Rec. 227(4), 475±485. Tan, P., Anasetti, C., Hansen, J. A., Melrose, J., Brunvand, M., Bradshaw, J., Ledbetter, J. A., and Linsley, P. S. (1993). Induction of alloantigen-speci®c hyporesponsiveness in human T lymphocytes by blocking interaction of CD28 with its natural ligand B7/BB1. J. Exp. Med. 177(1), 165±173. Tinkle, C. W., Lipschitz, D., and Ponnappan, U. (1998). Decreased association of p56lck with CD4 may account for lowered tyrosine kinase activity in mitogen-activated human T lymphocytes during aging. Cell. Immunol. 186(2), 154±160. Toubert, A., Clave, E., Talvensaari, K., Douay, C., and Charron, D. (2000). New tools in assessing immune reconstitution after hematopoietic stem cell transplantation. Vox Sang 78(29), 29±31.

the aging of the immune system

283

Toussaint, O., Medrano, E. E., and von Zglinicki, T. (2000). Cellular and molecular mechanisms of stress-induced premature senescence (SIPS) of human diploid ®broblasts and melanocytes. Exp. Gerontol. 35(8), 927±945. Traill, K. N., Huber, L. A., Wick, G., and JuÈrgens, G. (1990). Lipoprotein interactions with T cells: An update. Immunol. Today 11(11), 411±417. Utsuyama, M., Wakikawa, A., Tamura, T., Nariuchi, H., and Hirokawa, K. Impairment of signal transduction in T cells from old mice. (1997). Mech. Ageing Dev. 93, 131±144. Vella, A. T., Mitchell, T., and Groth, B. (1997). CD28 engagement and proin¯ammatory cytokines contribute to T cell expansion and long-term survival in vivo. J. Immunol. 158, 4714±4720. Wack, A., Cossarizza, A., Heltai, S., Barbieri, D., D'Addato, S., Fransceschi, C., Dellabona, P., and Casorati, G. (1998). Age-related modi®cations of the human alphabeta T cell repertoire due to different clonal expansions in the CD4‡ and CD8‡ subsets. Int. Immunol. 10(9), 1281±1288. Walker, D. G., Kim, S. U., and McGeer, P. L. (1995). Complement and cytokine gene expression in cultured microglial derived from postmortem human brains. J. Neurosci. Res. 40, 478±491. Walker, L. S., Gulbranson-Judge, A., Flynn, S., Brocker, T., and Lane, P. J. (2000). Co-stimulation and selection for T-cell help for germinal centres: The role of CD28 and OX40. Immunol. Today 21(7), 333±337. Weksler, M. E. (2000). Changes in the B-cell repertoire with age. Vaccine 18, 1624±1628. Welch, W. J. (1992). Mammalian stress response: Cell physiology, structure/function of stress proteins, and implications for medicine and disease. Physiol. Res. 72, 1063±1081. Weyand, C. M., Brandes, J. C., Schmidt, D., Fulbright, J. W., and Goronzy, J. J. (1998). Functional properties of CD4‡ CD28 T cells in the aging immune system. Mech. Ageing Dev. 102(2±3), 131±147. Whisler, R. L., Bagenstose, S. E., Newhouse, Y. G., and Carle, K. W. (1997). Expression and catalytic activities of protein tyrosine kinases (PTKs) Fyn and Lck in peripheral blood T cells from elderly humans stimulated through the T cell receptor (TCR)/CD3 complex. Mech. Ageing Dev. 98(1), 57±73. Whisler, R. L., Chen, M., Liu, B., and Newhouse, Y. G. (1999). Age-related impairments in TCR/ CD3 activation of ZAP-70 are associated with reduced tyrosine phosphorylations of z-chains and p59fyn/p56lck in human T cells. Mech. Ageing Dev. 111(1), 49±66. Whisler, R. L., Karan®lov, C. I., Newhouse, Y. G., Fox, C. C., Lakshmanan, R. R., and Liu, B. (1998). Phosphorylation and coupling of z-chains to activated T-cell receptor (TCR)/CD3 complexes from peripheral blood T-cells of elderly humans. Mech. Ageing Dev. 105(1±2), 115±135. Whisler, R. L., Newhouse, Y. G., and Bagenstose, S. E. (1996). Age-related reductions in the activation of mitogen-activated protein kinases p44mapk/ERK1 and p42mapk/ERK2 in human T cells stimulated via ligation of the T cell receptor complex. Cell. Immunol. 168(2), 201±210. Whisler, R. L., Newhouse, Y. G., Donnerberg, R. L., and Tobin, C. M. (1991). Characterization of intracellular ionized calcium responsiveness and inisitol phophate production among resting and stimulated peripheral blood T cells from elderly humans. Aging: Immunol. Infect. Dis. 3, 27±36. Whisler, R. L., Newhouse, Y. G., Grants, I. S., and Hackshaw, K. V. (1995). Differential expression of the a- and b-isoforms of protein kinase C in peripheral blood T and B cells from young and elderly adults. Mech. Ageing Dev. 77(3), 197±211. Wick, G., Huber, L. A., Offner, F., Winter, U., Bock, G., Schauenstein, K., JuÈrgens, G., and Traill, K. N. (1989). Immunode®ciency in old age. Curr. Probl. Dermatol. 18, 120±130. Wick, G., Kleindienst, R., Dietrich, H., and Xu, Q. (1992). Is atherosclerosis an autoimmune disease? Trends Food Sci. Technol. 3, 114±119. Wick, G., Schett, G., Amberger, A., Kleindienst, R., and Xu, Q. (1995). Is atherosclerosis an immunologically mediated disease? Immunol. Today 16(1), 27±33. Wick, G., and Xu, Q. (1999). AtherosclerosisÐAn autoimmune disease. Exp. Gerontol. 34, 559±566.

284

b. grubeck-loebenstein and g. wick

Wiegers, G. J., Kno¯ach, M., BoÈck, G., Dietrich, H., Falus, A., Boyd, R., Wick, G. (2001). CD4‡ CD8‡ TCRlow thymocytes express low levels of glucocorticoid receptors while being sensitive to glucocorticoid-induced apoptosis. Eur. J. Immunol. 31, in press. Witkowski, J. M., and Miller, R. A. (1993). Increased function of P-glycoprotein in T lymphocyte subsets of aging mice. J. Immunol. 150, 1296±1306. Xu, Q., Kleindienst, R., Waitz, W., Dietrich, H., and Wick G. (1993). Increased expression of heat shock protein 65 coincides with a population of in®ltrating T lymphocytes in atherosclerosic lesions of rabbits speci®cally responding to heat shock protein 65. J. Clin. Invest. 91, 2693±2702. Xu, Q., Schett, G., Perschinka, H., Mayr, M., Egger, G., Oberhollenzer, F., Willeit, J., Kiechl, S., and Wick, G. (2000). Serum soluble heat shock protein 60 is elevated in subjects with atherosclerosis in a general population. Circulation 102(1), 14±20. Yang, D. L., and Miller, R. A. (1999). Cluster formation by protein kinase C theta during murine T cell activation: Effect of age. Cell. Immunol. 195, 28±36. Yang, X., Stedra, J., Cerny, J. (1996). Relative contribution of T and B cells to hypermutation and selection of the antibody repertoire in germinal centers of aged mice. J. Exp. Med. 183(3), 959±970. Yen, C. J., Lin, S. L., Huang, K. T., and Lin, R. H. (2000). Age-associated changes in interferon-g and interleukin-4 secretion by puri®ed human CD4‡ and CD8‡ T cells. J. Biomed. Sci. 7(4), 317±321. Zheng, B., Han, S., Takahashi, Y., and Kelsoe, G. (1997). Immunosenescence and germinal center reaction. Immunol. Rev. 160, 63±77.

INDEX

A

Ac±Leu±Leu±nLeu±al, effect on proteasomes, 14 Activation-induced cytidine deaminase, role in hypermutation, 195±196 Active site-directed probe, targeting of cysteine protease, 82±86 Aging activation and signal transduction in immune cells, 262±263 areas of scienti®c research, 243 B cells, and antibody production, 257±259 bone marrow progenitor cells, 248±249 cellular changes in, 246±247 deterministic theories, 245±246 excess immunoreactivity in, disorders due to, 266±268 macrophages and dendritic cells in, 260±261 memory/effector T cells, 251±257 naive T cells, 249±251 natural killer cells in, 261 stochastic theories, 244±245 thymic function in, 249 and thymic reconstitution, 268±271 Allelic probabilities, cytokine memory, 142±143 Alzheimer's disease, 267 Aminopeptidases in catabolic pathway, 17±18 interferon-inducible, 41±43 puromycin-sensitive, 45 trimming of N-extended peptides, 51 Antagonistic pleiotropism, 246 Antibodies antigen-induced, idiotypic speci®cities of, 207±211 anti-SIINFEKL, 44 combined VH =VL regions, 205±206 monoclonal, neutralizing, 226

presence in elderly, 265±266 production, and B cell aging, 257±259 Antigen presentation class II-restricted, 102 by dendritic cells, 81±82, 131 IFN-g-induced changes, 28±30 MHC class II molecules, 71±73 MHC class I molecules, 21±22 IFN-g role, 47±48 nonproteasomal pathways, 32±33 PA28 role, 30±31 process, 96±102 proteasome role, 26±28 proteolytic step in, 44±47 ubiquitination in, 25±26 virus effects, 50 Antigen-presenting cells protease activity in, 73±87 proteolytic activity, regulatory role of Ii, 94±96 role in induction of cytokine memory, 130±133 Antigens acquisition step, 97±100 binding, and idiotype expression: dichotomy, 204±206 cleavage of: proteolytic steps in, 39±40 clonal selection by, transformed by idiotypic network, 220±223 degradation, by endocytic proteases, 100±102 exogenous, class I-presented peptides generated from, 48±49 induced antibodies, idiotypic speci®cities of, 207±211 marking for degradation, 100

285

286

index

APC, see Antigen-presenting cells Apoptosis induced by tripeptidyl peptidase II inhibitors, 20 proteasome inhibitors and, 15 Asparaginyl endopeptidase antigen degradation by, 100 present in APCs, 76 Atherosclerosis autoimmune hypothesis, 263±265 HSP 60 and, 273 ATP, protein degradation dependent on, 3 ATPase, in 26S proteasome base, 10±11 Atrophy organ, 246±247 thymus, cytokines and, 269±270

B

B cells aging, and antibody production, 257±259 antigen-speci®c, 214 antigen-speci®c receptor BCR, 99 B-2, antigen-inducible, 223 ICOS ligand expression, 154 idiotypic activation or suppression, 211±219 memory, diminished survival, 184 switch recombination, 186 BCL-6, competition with Stat6, 121 B effector 1 cells, generation, 132±133 Bleomycin, aminopeptidase hydrolyzing, 42 Bone marrow progenitor cells, aging and, 248±249

C

CA074, cathepsin B-speci®c inhibitor, 85 Cathepsins Cat L, p41 and, 95±96 identi®ed in APC types, 76±77 regulation of each other, 82 role in antigen degradation, 101±102 Ii proteolysis, 87±88 tissue-speci®c expression, 79 Cell cycle control, aging and, 245 cytokine memory dependent on, 141±142 Cells, see also speci®c types cytosolic trimming of N-extended epitopes, 41 immune, in old age, 262±263

MBP-reactive, 217 proteasome inhibitor effect, 27 senescent, impaired function, 247 sentinel, 48 Cellular autoimmunity, increase with age, 266±267 Cellular interactions idiotypic, 212±213 role of idiotype-speci®c T cells, 216±218 Chemokine ligand 2, see MCP-1 Chemokine receptors CCR7, in T cell migration, 155 T cell recruitment and, 150±152 Cholesterol, in elderly, 247 CIITA transcription factor, Th1, 129±130 CLICK148, cathepsin L-speci®c inhibitor, 86 ab±CLIP complexes, 89±90 Clonal expansion, CD8‡ T cells, 252±253, 272 Clonal selection, transformed by idiotypic network, 220±223 c-Maf, Th2 transcription factor, 127 Conserved noncoding sequence-1, GATA-3 binding site, 124 COP9 signalosome, 11 Crossreactive idiotype A, 209±210 Cullin-1, 4 Cystatins, 80±81 Cysteine protease inhibitors present in APCs, 80±82 targeting of, 82±86 Cytidine deaminase, activation-induced, role in hypermutation, 195±196 Cytokine memory allelic probabilities, 142±143 APC role in induction of, 130±133 induction and maintenance, 122±130 scenario for, 143±145 of memory T cells, 154±158 part of T cell differentiation programs, 149±154 persistence, 156±158 stability and plasticity, 145±149 TCR signals in induction of, 133±138 Cytokine receptors, regulation of T cell effector functions, 152±154 Cytokines genes, epigenetic modi®cations, 138±145 in¯ammatory, 267 instructive, for Th1 and Th2 differentiation, 117±122

index production in old age, 253±256, 271 regulation of protease activity, 78±79 secretion by T cells, 116 signals, in memory persistence, 156±158 and thymic atrophy, 269±270 Cytoplasm peptide generation in, 24±25 proteolytic pathways in, 19±20 trimming of peptides in, 40±43 Cytotoxicity antigen-speci®c MHC class I restricted, 272 memory/effector T cells, 256

D

Defective ribosomal products, 22±23 Demethylation, DNA, connection to DNase I hypersensitivity, 140±141 Dendritic cells activation and maturation, 130±131 atrophic follicular, 258±259 Cat S = , 94 development into APCs, 79 generation of class I-presented peptides, 48±49 immature, 81±82 increased production in elderly, 260±261 treated with LHVS, 90 Destruction antigenic peptides by peptidase, 44±47 Cat L, preventive role of p41, 96 Deterministic theories, aging, 245±246 Disease, age-related, 263±265 DM, see H-2M/HLA-DM DNA aging of, 245 demethylation, connection to DNase I hypersensitivity, 140±141 DNA polymerase error-prone, 191±192 generation of Ig hypermutation, 188±191 pol, 194 sloppy, 196 DNase I, hypersensitivity, connection to DNA demethylation, 140±141 Double-strand breaks, repair, 187

E

E3 enzymes, 4 Endocytic compartment

287

class I-presented peptides generated in, 48±49 targeting class II molecules to, 92±93 Endocytic proteases activity in APCs, 73, 76±77 antigen degradation by, 100±102 regulation by cytokines, 78±79 endogenous inhibitors, 80±82 intracellular pH, 77±78 Endoplasmic reticulum MHC class I molecules in, 23±24 peptide trimming in, 43±44 protein degradation in, 20±21 Epigenetic modi®cations, cytokine genes, 138±145 Epoxyketones, proteasome inhibitors, 14 Epstein±Barr virus, latency, 50 ERK, Ras/ERK MAPK pathway, 133±134 ERM transcription factor, Th1, 129 ES-62, dendritic cells stimulated by, 132 Extracellular signal-regulated kinase, see ERK

F

Fc receptors, expressed by APCs, 98±99 Fragments FV , immunogenic epitopes, 205 generated by proteasomes, 16±17 Free radicals, aging and, 244 Fucosyltransferases, in selectin ligand generation, 150 FV fragment, immunogenic epitopes, 205

G

GATA-3 expression and activity, regulation, 125±127 induction of IL-4 memory, 148 recruitment of chromatin-remodeling factors, 141 for Th2 cytokine memory, 123±125 Gated opening, 20S proteasome a-rings, 8±9, 12, 37±38 Germinal center response, in aging, 258±259 Gerontogenes, 245±246 Glycosylation, aging and, 244

H

Haptens, immune responses to, 209±210 Hay¯ick phenomenon, 245

288

index

Heat shock proteins HSP 60, atherosclerosis and, 264±265, 273 peptides bound to, 49 H-2M/HLA-DM nonclassical class II dimer, 73 peptide loading mediated by, 91±92 Hormones, and thymic involution, 269 hsc70, interaction with Ii, 93 Hybrid proteasomes, 12±13 Hydrolysis bleomycin, by aminopeptidase, 42 protein and ATP, 9±10 Hypermutation activation-induced cytidine deaminase role, 195±196 IgG, in germinal center cells, 259 Ig genes DNA polymerase-generated, 188±191 mismatch repair role, 183±186 nickase role, 186±187 transcription role, 187±188 Hyperplasia, organ, 247

I

ICOS, expression on Th2 cells, 153±154 Idiotype activation or suppression of B and T cells, 211±219 expression, and antigen binding: dichotomy, 204±206 Idiotypic interactions between B and T cells, 213±219 and D region variability, 207±211 in individual immune system, 206±207 within peripheral immune system, 221 Idiotypic network theory, 203±204, 211±212 transformation of clonal selection, 220±223 Ii, see Invariant chain Immune response antigen-induced, 220±222 anti-phOx, 227 thymus-dependent, 225 Immune surveillance, antigen presentation levels and, 46±47 Immune system aged innate: overactivity of, 261 aging of, 248±263 individual, idiotypic interactions in, 206±207

senescence consequences of, 263±268, 272±273 controversies about, 271 Immunoglobulin gene hypermutation generated by DNA polymerase, 188±191 mismatch repair role, 183±186 nickase role, 186±187 transcription role, 187±188 IgG, hypermutation in germinal center cells, 259 syngeneic IgE, immune response to, 225±226 Immunological experiences maternal, transfer to offspring, 223±228 single, idiotypic transformation, 204±223 Immunoproteasomes cleavages by, 33±36 degradation of substrate ovalbumin, 37 peptides produced by, 29±30 viral inhibition, 50 Immunoreactivity, excess, disorders due to, 266±268 Infectious disease, consequence of immune senescence, 263±265 Interferon-a, for Th1 differentiation, 119 Interferon-g induced changes in proteasomes, 28±30 memory for, 146±149 production in old age, 254 role in MHC class I antigen presentation, 47±48 Interleukin-2, production in old age, 253 Interleukin-4 GATA-3±dependent enhancer activity, 124 memory in Th2 cells, stabilization, 147 selective enhancement of B7.2, 132 sustained gene transcription, 139 for Th2 differentiation, 119±121 Interleukin-6, for Th2 differentiation, 121±122 Interleukin-7, intrathymically available, 269±270 Interleukin-10, production in old age, 254±256 Interleukin-12, for Th1 differentiation, 117±118 Interleukin-23, for Th1 differentiation, 118 Interleukin-4 receptor, and TCR, concomitant signaling, 142 Interleukin-18 receptor, expression on Th1 cells, 152

index Invariant chain ab±Ii conversion to ab±Iip10, 88±89 complex with class II, 72, 78 interaction with hsc70, 93 processing, relation to endocytic pathway, 93±94 proteolysis, 73 cathepsin role, 87±88 regulatory role in APCs proteolytic activity, 94±96

J

Jak-binding protein, see SOCS-1 JNK, inhibitory role in Th2 differentiation, 134±135 JPM-565, inhibitor of cysteine proteases, 85 JunB, Th2 transcription factor, 128 Knockout mice cathepsin, 86±87 PA28b, 31 pol m, 194

K

L

Lactacystin, proteasome inhibitor, 14 Langerhans cells, decreased numbers in elderly, 260 Late endocytic compartments, proteolytic pathway in, 2±3 LHVS, see Vinyl sulfones LMP2, proteasomes containing, 28±30 Lymphocytes, cholesterol overload, 247 Lysosomal proteases antigen presentation dependent on, 102 propiece, 77 Lysosomal thiol reductase (GILT), 100

M

Macrophages generation of class I-presented peptides, 48±49 mannosylated antigen uptake, 99 production in elderly, 260±261 Major histocompatibility complex, see MHC Maturation amyloid plaques, 267 dendritic cells, 130±131 T cells, 249 MCP-1, for Th2 differentiation, 122

289

Memory B cells, diminished survival, 184 cytokine, see Cytokine memory for IFN-g, 146±149 IL-4, GATA-3 induction of, 148 Memory/effector T cells cytokine production, 253±256 phenotypic changes, 251±252 proliferation, 256±257 repertoire, 252±253 Memory T cells central and effector, 154±156 persistence, 156±158 Metallopeptidases, in catabolic pathway, 17 Methylated-CpG binding protein 2, 139 MHC class I antigen presentation, IFN-g role, 47±48 molecules, structure and assembly, 21±22 peptides presented by generation frequency, 38±39 generation from exogenous antigens, 48±49 nonproteasomal pathways, 32±33 origin of, 22±25 restricted cytotoxicity, antigen-speci®c, 272 MHC class II molecules antigen presentation by, 71±73 IFN-g-induced expression, 78±79 targeting to endocytic compartment, 92±93 Mismatch repair, role in Ig gene hypermutation, 183±186, 189 Monocyte chemoattractant protein-1, see MCP-1 MSH2/MSH6 complex, mismatch repair and, 185 Mutation in E1 ubiquitin-activating enzyme, 5 Ig genes, mismatch repair role, 183±186 targeting to non-Ig genes, 195 Myelin basic protein, reactive cells, 217

N

Natural killer cells, in old age, 261 NFATc, regulatory role in cytokine gene expression, 136±137 NFATp, T cells de®cient for, 137±138 NF-kB p50 de®ciency, 126 regulatory role in Th cytokines, 135±136 Nicks, presence during Ig hypermutation, 186±187

290

index

N-terminal extensions aminopeptidase-trimmed, 51 MHC-presented peptides with, 40±42 Nuclear factor of activated T cells, see NFAT

O

Offspring, transfer of maternal immunological experience to, 223±228 Oligopeptides, produced by proteasomes, fate of, 17±19 Organs, aging effects, 246±247 Osteoporosis, postmenopausal, 268 Osteoprotegerin ligand, 268 Ovalbumin, class I-presented peptide generated from, 35±36

P

p41, and Cat L, 95±96 PA28 complex with 20S proteasome, 12±13 role in antigen presentation, 30±31 Pathogens, interfering with generation of presented peptides, 49±50 Peptides antigenic, destruction by peptidases, 44±47 class II-associated Ii derived (CLIP), 72, 89±90 idiotypic, recognized by Id-speci®c T cells, 216, 219 loading, DM-mediated, 91±92 MHC class I-presented generation frequency, 38±39 generation from exogenous antigens, 48±49 nonproteasomal pathways, 32±33 origin of, 22±25 pathogen interference, 49±50 trimming in cytoplasm, 40±43 in endoplasmic reticulum, 43±44 Phenotypic changes, memory/effector T cells, 251±252 pHi , regulation of protease activity, 77±78 Phosphorylcholine, immune response to, 210 Plasma membrane, ¯uidization, 247 Plasticity, cytokine memory, 145±149 p38 MAPK pathway, 134 pol e, in hypermutation process, 192 pol m, homology with Tdt, 194 Post-acidic site, 20S proteasome, 8

P1 position, preferred by proteasome active site, 36±37 Progenitor cells, bone marrow: aging and, 248±249 Proin¯ammatory status, Alzheimer's disease and, 267 Proliferation, memory/effector T cells, 256±257 Proteasome activating nucleotidase complex, 11 Proteasomes, see also Immunoproteasomes cleavage of antigens, 39±40 degradation of proteins, products of, 15±17 generation of presented peptide: frequency, 38±39 hybrid, 12±13 IFN-g-induced changes, 28±30 oligopeptides produced by, fate of, 17±19 role in antigen presentation, 26±28 protein degradation in vivo, 13±15 20S, see 20S proteasome 26S, see 26S proteasome substrate cleavage sites, 36±38 Protein degradation blocking, 50 in cytoplasm, 19±20 in endoplasmic reticulum, 20±21 by proteasomes, products of, 15±17 proteolytic pathway, 2±3 role of proteasomes in vivo, 13±15 20S proteasome, 6±9 ubiquitin conjugation and, 3±6 Protein turnover, balancing act, 1±2 Proteolysis and DM-mediated peptide loading, 91±92 Ii, 73 cathepsin role, 87±88 Proteolytic pathway alternate, 19±20 protein degradation, 2±3 Puromycin, aminopeptidase sensitive to, 45, 47

R

Rac2, expression in Th1 cells, 135 Reconstitution, thymus, aging and, 268±271 ROG, repressor of GATA, 127

S

SCF, polyubiquitination dependent on, 4 19S complex, in formation of 26S proteasome, 10

index Selectin ligands, T cell recruitment and, 150 Self±nonself discrimination, immune response in elderly, 266 Senescence cellular, impaired function, 247 immune system consequences of, 263±268, 272±273 controversies about, 271 Signal transduction, immune cells in old age, 262±263 SOCS-1, expression in activated Th cells, 121±122 20S proteasome catalytic subunits, 7±8 cleavages by, 33±35 complex with PA28, 12±13 immune subunits, 29 ring con®guration, 6 26S proteasome ATP-dependent degradation, 6 cleavages by, 35±36 degradation of polyubiquitinated substrates, 9±10 formation, 10 11S REG, see PA28 Stability, cytokine memory, 145±149 Stat6 BCL-6 competition with, 121 role in activation of IL-4 gene, 120 Stochastic theories, aging, 244±245 Suppressor of cytokine signaling-1, see SOCS-1 Switch recombination, 186

T

TAP mobility, 23 peptides transported via, 43±44 Target tissues, differential T cell recruitment to, 150±152 T-bet reversal of Th2 cytokine memory, 149 for Th1 cytokine memory, 128±129 T cell cytokine receptor, for Th1 differentiation, 118±119 T cell receptors anti-idiotypic antibodies to, 218 and IL-4R, concomitant signaling, 142 naive T-cell activation requiring, 249±251 restimulation of polarized Th2 cells, 147

291

signals, in induction of cytokine memory, 133±138 stimulation duration, 155 T cells cytokine memory in, 148±149 cytokine secretion, 116 de®cient for NFATp, 137±138 differentiation programs, cytokine memory as part of, 149±154 effector functions, regulation, 152±154 idiotypic activation or suppression, 211±219 maturation, 249 memory, see Memory T cells naive, 249±251 polarized, and cytokine memory plasticity, 146 TCR, see T cell receptors Telomeres aging and, 245 shortening, 271 Terminal deoxynucleotidyl transferase, 186 Th1 cytokine memory stabilization, 148±149 T-bet for, 128±129 differentiation, instructive cytokines for, 117±119 program description, 115±116 transcription factors, CIITA, 129±130 Th2 cytokine memory GATA-3 for, 123±125 reversal, 149 differentiation, instructive cytokines for, 119±122 IL-4 memory in, 147 induction of development by dendritic cells, 131±132 program description, 115±116 transcription factors c-Maf, 127 JunB, 128 Thimet oligopeptidase antigen presentation inhibited by, 45±46 in catabolic pathway, 17±18 Thymus atrophy, cytokines and, 269±270 intact function, and immune reactivity, 268±269

292

index

Thymus (continued) involution, T cell maturation and, 249 tissue engineering, 270±271 Tissue engineering, thymic organoid, 270±271 Transcription, role in Ig hypermutation, 187±188 Trimming by aminopeptidases, 18 peptides in cytoplasm, 40±43 in endoplasmic reticulum, 43±44 Tripeptidyl peptidase II inhibitors, apoptosis induced by, 20 T1/ST2, selective expression on T cells, 152±153

Txk transcription factor, Th1, 129

U

Ubiquitin, conjugation, and protein degradation, 3±6 Ubiquitination, in antigen presentation, 25±26

V

Vaccination, in elderly, 265±266, 272 Vinyl sulfones dendritic cells treated with, 90 inhibitors of cysteine proteases, 82±85 Viruses, reduction of antigen presentation, 50

CONTENTS OF RECENT VOLUMES

Volume 74 Biochemical Basis of Antigen-Speci®c Suppressor T Cell Factors: Controversies and Possible Answers Kimishice Isihzaka, Yasuyuki Ishii, Tatsumi Nakano, and Katsuji Sugik The Role of Complement in B Cell Activation and Tolerance Michael C. Carroll Receptor Editing in B Cells David Nemazee Chemokines and Their Receptors in Lymphocyte Traf®c and HIV Infection Pius Loetscher, Bernhard Moser, and Marco Bacciolini Escape of Human Solid Tumors from T-Cell Recognition: Molecular Mechanisms and Functional Signi®cance Francesco M. Marincola, Elizabeth M. Jaffee, Daniel J. Hicklin, and Soldano Ferrone The Host Response to Leishmania Infection Werner Solbacii and Tamas Laskay Index

Volume 75 Exploiting the Immune System: Toward New Vaccines against Intracellular Bacteria JuÈrgen Hess, Ulrich Schaible,

BaÈrbel Raupach, and Stefan H. E. Kaufmann The Cytoskeleton in Lymphocyte Signaling A. Bauch, F. W. Alt, G. R. Crabtree, and S. B. Snapper TGF-b Signaling by Smad Proteins Kohei Miyazono, Peter ten Dijke, and Carl-Henrik Heldin MHC Class II-Restricted Antigen Processing and Presentation Jean Pieters T-Cell Receptor Crossreactivity and Autoimmune Disease Harvey Cantor Strategies for Immunotherapy of Cancer Cornelis J. M. Meliey, Rene E. M. Toes, Jan Paul Medema, Sjoerd H. van der Burg, Ferry Ossendorp, and Rienk Offringa Tyrosine Kinase Activation in the Decision between Growth, Differentiation, and Death Responses Initiated from the B Cell Antigen Receptor Robert C. Hsueh and Richard H. Scheuermann The 30 IgH Regulatory Region: A Complex Structure in a Search for a Function Ahmed Amine Khamlichi, Eric Pinaud, Catherine Decourt, Christine Chauveau, and Michel Cogne Index

293

294

contents of recent volumes

Volume 76 MIC Genes: From Genetics tok Biology Seiamak Bahram CD40-Mediated Regulation of Immune Responses by TRAF-Dependent and TRAF-Independent Signaling Mechanisms Amrif C. Grammer and Peter E. Lipsky Cell Death Control in Lymphocytes Kim Newton and Andreas Strassen Systemic Lupus Erythematosus, Complement De®ciency, and Apoptosis M. C. Pickering, M. Botto, P. R. Taylor, P. J. Lachmann, and M. J. Walport Signal Transduction by the High-Af®nity Immunoglobulin E Receptor FceRI: Coupling Form to Function Monica J. S. Nadler, Sharon A. Matthews, Helen Tuhner, and Jean-Pierre Kinet Index

Volume 77 The Actin Cytoskeleton, Membrane Lipid Microdomains, and T Cell Signal Transduction S. Celeste Posey Morley and Barbara E. Bierer Raft Membrane Domains and Immunoreceptor Functions Thomas Harder Human Basophils: Mediator Release and Cytokine Production John T. Schroeder, Donald W. MacGlashan, Jr., and Lawrence M. Lichtenstein Btk and BLNK in B Cell Development Satoshi Tsukada, Yoshihiro Baba, and Dai Watanabe

Diversity and Regulatory Functions of Mammalian Secretory Phospholipase A2 s Makoto Murakami and Ichiro Kudo The Antiviral Activity of Antibodies in Vitro and in Vivo Paul W. H. I. Parren and Dennis R. Burton Mouse Models of Allergic Airway Disease Clare M. Lloyd, Jose-Angel Gonzalo, Anthony J. Coyle, and Jose-Carlos Gutierrez-Ramos Selected Comparison of Immune and Nervous System Development Jerold Chun Index

Volume 78 Toll-like Receptors and Innate Immunity Shizuo Akira Chemokines in Immunity Osamu Yoshie, Toshio Imai, and Hisayuki Nomiyama Attractions and Migrations of Lymphoid Cells in the Organization of Humoral Immune Responses Christoph Schaniel, Antonius G. Rolink, and Fritz Melchers Factors and Forces Controlling V(D)J Recombination David G. T. Hesslein and David G. Schatz T Cell Effector Subsets: Extending the Th1/ Th2 Paradigm Tatyana Chtanova and Charles R. Mackay MHC-Restricted T Cell Responses against Posttranslationally Modi®ed Peptide Antigens Ingelise Bjerring Kastrup, Mads Hald Andersen, Tim Elliot, and John S. Haurum

contents of recent volumes

295

Gastrointestinal Eosinophils in Health and Disease Marc E. Rothenberg, Anil Mishra, Eric B. Brandt, and Simon P. Hogan

76 and LAT and Their B Cell Counterpart, BLNK/SLP-65 Deborah Yablonski and Arthur Weiss

Index

Xenotransplantation David H. Sachs, Megan Sykes, Simon C. Robson, and David K. C. Cooper

Volume 79 Neutralizing Antiviral Antibody Responses Rolf M. Zinkernagel, Alain Lamarre, Adrian Ciurea, Lukas Hunziker, Adrian F. Ochsenbein, Kathy D. McCoy, Thomas Fehr, Martin F. Bachmann, Ulrich Kalinke, and Hans Hengartner Regulation of Interleukin-12 Production in Antigen-Presenting Cells Xiaojing Ma and Giorgio Trinchieri Mechanisms of Signaling by the Hematopoietic-Speci®c Adaptor Proteins, SLP-

Regulation of Antibacterial and Antifungal Innate Immunity in Fruit¯ies and Humans Michael J. Williams Functional Heavy-Chain Antibodies in Camelidae Viet Khong Nguyen, Aline Desmyter, and Serge Muyldermans Uterine Natural Killer Cells in the Pregnant Uterus Chau-Ching Liu and John Ding-E Young Index

This Page Intentionally Left Blank