ADVANCES IN
Immunology VOLUME 70
This Page Intentionally Left Blank
ADVANCES IN
Immunology EDITED BY FRANK J. DlXON The Scripps Research Institute La Jolla, California ASSOCIATE EDITORS
Frederick Alt K. Frank Austen Tadamitsu Kishimoto Fritz Melchers Jonathan W. Uhr
VOLUME 70
ACADEMIC PRESS San Diego London Boston New York Sydney Tokyo Toronto
This book is printed on acid-free paper.
@
Copyright 0 1998 by ACADEMIC PRESS All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher’s consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923), for copying beyond that permitted by Sections 107 or 108 of the US. 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-1998 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. 0065-2776/98 $25.00
Academic Press u division of Hurcourt Brace & Compuny
525 B Street, Suite 1900, San Diego, California 92101-4495, USA http:l/www.apnet.com Academic Press 24-28 Oval Road, London NW 1 7DX, UK http://www.hbuk.co.uk/ap/ International Standard Book Number: 0-12-022470-4 PRINTED IN THE UNITED STATES OF AMERICA 98 99 0 0 0 1 02 0 3 E B 9 8 7 6
5
4
3 2 1
CONTENTS
ix
CONTRIBUTORS
Biology of the Interleukin-2 Receptor
BRADH. NELSONAN11 DENNIS M. WILLERFORI) I. Introdiiction 11. The IL-2 Receptor Complex 111. IL-2 Receptor Expression IV. Cellular Responses to IL-2 Receptor Signals V. Mechani9in of IL-2 Receptor Activation VI. Intracellular Signaling by the IL-2 Receptor VII. In Viuo Studies of IL-2 Receptor Function in Lymphocyte Development VIII. In Viuo Studies of IL-2 Receptor Function in Peripheral Lymphocytes IX. Summary and Conclusions References
1
-
3 1
10 19 21
42 53 64 66
Interleukin-12: A Cytokine at the Interface of Inflammation and Immunity
GIORC:IO TRINCHIEHI I. 11. 111. IV. V. VI. VII. VIII. IX. X.
Introduction IL-12 Molecule and Its Genes IL-12 Receptor and Signal Transduction Production of IL-12 Molecular Control of IL-12 Gene Expression IL-12 Effects on Heinatopoietic Stem Cells Induction of IFN--y and Other Cytokines by IL-12 Mitogenic Activity of IL-12 Activation of Cytotoxic Lyinphocytes by IL-12 Effect of IL-12 on the Differentiation of T Helper Cells v
83 86 95 101 114 119 122 127 129 134
vi
CONTENTS
XI. Effects of IL-12 on B-Cell Responses and Vaccination XII. IL-12 in Delayed-Type Hypersensitivity, & w a y Hyperresponsiveness, and Graft Rejection XIII. IL-12 in Organ-Specific Autoimmune Diseases XIV. IL-12 in the Inflammatory Response XV. IL-12 in Infectious Diseases XVI. Antitumor Effects of IL-12 XVII. Concluding Remarks References
148
152 157 166 168 187 193 194
Recent Progress on the Regulation of Apoptosis by Bcl-2 Family Members
ANDYJ. MINN,RACHEL E. SWAIN,AVERILMA, AND CRAIGB. THOMPSON
I. Significance of Programmed Cell Death 11. The Genetics of Programmed Cell Death 111. The Bcl-2 Family IV. Mitochondria Can Control A optosis V. Structure/Function Studies o Bcl-xL VI. How Do Bcl-2 Family Members Regulate Cell Sunivd? VII. Conclusion References
T!
245 247 250 257 261 266 269 271
Interleukin-18:A Novel Cytokine That Augments Both Innate and Acquired Immunity
HARUKI OKAMURA, HIROKO TSUTSUI, SHIN-ICHIRO KASHIWAMURA, TOMOHIRO YOSHIMOTO, A N D KENJINAKANISHI I. 11. 111. IV. V. VI. VII. VIII.
Introduction Molecular Structure of IL-18 and Its Gene Producing Cells Re uirement of Caspase-1 for Processing of IL-18 Bio ogical Function Receptors for IL-18 Role of IL-18 in Host Defenses PatholOgiCdl Roles O f IL-18 IX. Perspective References
7
281 282 285 286 287 294 298 301 304 305
CD4' T-cell Induction and Effector Functions: A Comparison of Immunity against Soluble Antigens and Viral Infections
ANNETTEOXENIUS, ROLFM. ZINKERNAGEL, A N D HANSHENGARTNER
I. Introduction 11. Activation of CD4' T Cells 111. CD4+ T-cell Effector Functions
313 314 330
vii
CONTENTS
IV. Conclusions References
351 352
Current Views in lntracellular Transport: Insights from Studies in Immunology
VICTOR W. Hsri
AND
PETERJ. PETERS
I. Introduction 11. A General Mechanism of Intracellular Transport 111. Complexities of Transport in V i m IV. Secretory Pathways V. Endocytic Pathways VI. Transport in Polarized Cells VII. Perspective References
369 373 383 385 392 398 402 402
Phylogenetic Emergence and Molecular Evolution of the Immunoglobulin Family
J. MARCHALONIS, SAMUEL F. SCHLUTEH, RALPIIM . BEKNSTEIN, SIIANXIANG SHEN.AND ALLENB. EDMUNDWN
JOHN
I. Introduction Evolutionary Emergence of the Combinatorial Immune System Ancient Foundations of the Coinbinatorial Irninune System Emergence of Bonn Fide Iininunoglobulins Iininunoglobulins and T-cell Receptors of Jawed Vertebrates Framework 4 of the Variable Doniain Encoded by the Joining Segment VII. Evolutioniry Comparisons of T-cell Receptors VIII. Evolution of Li lit Chains IX. Origin and Evo ution of Iininunoglobulin Heavy Chains X. Segmental Gene Organization in Evolution XI. Molecular Events Underlying the Ex h i v e Emergence of Immunoglobulins and Their Initial P ases of Evolution XI. Conclusions References 11. 111. IV. V. VI.
K
1
417 418 418 425 430 439 44 1 45 1 460 475
485 491 492
Current Insights into the “Antiphospholipid” Syndrome: Clinical, Immunological, and Molecular Aspects
DAVIDA. KAXDIAH,ANDREJSALI,YONGHIJA SHENG,EDWARD J. VICTORIA, DAVID M . MAKQLIIS, STEPHENM. COU’ITS,A N D STEVENA. KRI1.IS
I. Introduction 11. “Antiphospholipid” Antibodies 111. Clinical Features of the “Antiphospholipid’ Syndrome IV. P2-Glycoprotein I V. Iinmunogenicity and Animal Models
507 508 512 520 531
viii
CONTENTS
VI. Prothroinbin
VII. Lupus Anticoagulant Antibodies and Protein C Activation VIII. IX. X. XI. XII. XIII.
Lupus Anticoagulant Antibodies and Phos hatic~ylethanolamine Antiphospholipid Antibodies and Endothe ial Cells Pathogenesis of the Aiitiphos holipid Syndrome Laboratory Investigations of t ie Antiphos holipid Syndrome Antiphospholipid Syndrome and Future T ierapies Suminary and Conclusions References
INDEX CONTENTS OF RECENTVOLUMES
P
P
I;
533 535 536 536 537 543 545 8547 548
565 573
CONTRIBUTORS
Ralph M. Bernstein (417 ) , FDA/clm-/H FM-Fj4 I, Betliesda, Maryland 20892 Stephen M. Coutts (507, L,a Jolla Pharniaceutical Coinpiny, San Diego, California 92121 Allen B. Edmunson (417), Oklahoma Medical Reseracli Foundation, Oklahorna City, Oklalionia 6 3 104-.5046 Hans Hengartner (313), Departnient of Pathology, Institute of Experimental Immunology, Unicxmity of Ziirich, 8091 Zurich, Switzerland Victor W. Hsu (3691,Division of Rheumatology, Immuno1ogy,and Allergy, Brighani and Women’s Hospital, Haivarcl Medical School, Boston, Massachusetts 21005 David A. Kandiah (507),Department of Immiinology, Allergy, and Infectious Disease, University of Nccv South Wales School of Medicine, Saint George Hospital, Kogarah 22 17, Australia Shin-ichiro Kashiwamura (281), Laboratoiy of Host Defenses, Institute for Advances Medical Sciences, Hyogo College of Medicine, Hyogo 663, Japan Steven A. Krilis (507), Department of Immunology, University of New South Wales School of Medicine, Saint George Hospital, Kogarah 2217, Australia Averil Ma (245). Coininittee on Imiiiunology and the Department of Medicine, Universitv of Chicago. Chicago, Illinois 60637-5420 John J. Marchalonis (417), Department of Microbiology and Immu~~, nology, College of Medicine, University of Arizona, T L K S OArizona 85724 David M. Marquis (507), La J o h Pharmaceutical Company, San Diego, California 92 121 Andy J. Minn (245), Gwen Kiiapp Center for Lupus and Iniinuriology Research and the Committee on Inimiuiology, University of Chicago, Chicago, Illinois 60637-5420 I\
X
CONTRIBUTORS
Kenji Nakanishi (281),Laboratory of Host Defenses, Institute for Advances Medical Sciences, and Department of Immunology and Medical Zoology, Hyogo College of Medicine, Hyogo 663,Japan Brad H.Nelson (l),The Virginia Mason Research Center, Seattle, Washington 98101;and Department of Immunology, University of Washington School of Medicine, Seattle, Washington 98195 Haruki Okamura (281),Laboratory of Host Defenses, Institute for Advanced Medical Sciences, Hyogo College of Medicine, Hyogo 663, Japan Annette Oxenius (313),Department of Pathology, Institute of Experimental Immunology, University of Zurich, 8091 Zurich, Switzerland Peter J. Peters (369),Department of Cell Biology, Faculty of Medicine and Institute of Biomembranes, Utrecht University, 3584 CX Utrecht, The Netherlands Andrej Sali (507),Rockefeller University, New York, New York 10021 Samuel F.Schluter (417),Department of Microbiology and Immunology, College of Medicine, University of Arizona Health Sciences Center, Tucson, Arizona 85724-5049 Shanxiang Shen (417),National Institutes of Health, Bethesda, Maryland 20892 Yonghua Sheng (507),Department of Immunology, University of New South Wales School of Medicine, Saint George Hospital, Kogarah 2217,Australia Rachel E. Swain (245),Committee on Cancer Biology, University of Chicago, Chicago, Illinois 60637-5420 Craig B.Thompson (245),Gwen Knapp Center for Lupus and Immunology Research and Committees on Immunology and Cancer Biology, Department of Medicine, University of Chicago, Chicago, Illinois 60637-5420 Giorgio Trinchieri (83),Immunology Program, Wistar Institute of Anatomy and Biology, Philadelphia, Pennsylvania 19104-4268 Hiroko Tsutsui (281),Department of Immunology and Medical Zoology, Hyogo College of Medicine, Hyogo 663,Japan Edward J. Victoria (507),La Jolla Pharmaceutical Company, San Diego, California 92121 Dennis M. Willerford (11, Departments of Immunology and Medicine, University of Washington School of Medicine, Seattle, Washington 98195;and the Puget Sound Blood Center, Seattle, Washington 98104 Tomohiro Yoshimoto (281),Laboratory of Host Defenses, Institute for Advances Medical Sciences, and Department of Immunology and Medical Zoology, Hyogo College of Medicine, Hyogo 663,Japan Rolf M. Zinkernagel(313),Department of Pathology, Institute of Experimental Immunology, University of Zurich, 8091 Zurich, Switzerland
ADV4NCE5 IN 1MMUNOI.OCX VOL 70
Biology of the Interleukin-2 Receptor BRAD H. NELSON'st AND DENNIS M. WILLERFORDtr~r§ 'The Virginia Mason Research Center, Seattle, Washington, 98 101; Departments of flmmunology and #Medicine, University of Washington School of Medicine, Seattle, Washington, 98 195; and §The Puget Sound 6load Center, Seattle, Washington 98 104
I. Introduction
Homeostatic regulation of' the immune system requires extensive communication among its cellular constituents, which include lymphocytes, macrophages, dendritic cells, and stromal elements. These cell-cell interactions typically occur at close range and include activation of a wide variety of cell surface signaling molecules by direct contact, as well as signaling through secretion of an array of soluble mediators. Among the latter, interleukin-2 ( IL-2) is perhaps the most extensively studied. First identified as a growth factor for T cells in vitro (Gillis and Smith, 1977a,b; Morgan et nl., 1976), IL-2 has also been implicated in the functional differentiation of T cells, as well as in the growth and effector function of B and natural killer (NK) cells. Receptors for IL-2 are also expressed on developing T and B cells. The remarkable in vitro properties of IL-2 suggested that this lymphokine acted at the heart of the immune response -by mediating antigen-triggered T-cell expansion and promoting effector cell differentiation. This view has grown more complicated in recent years, based on a greater appreciation for the redundancy of in wivo growth and regulatory signals, as well as on unexpected observations regarding the function of IL-2. These include the studies of Lenardo (1991) demonstrating that IL-2 may promote T-cell apoptosis under some circumstances and the observation that mice lacking IL-2 have phenotypically normal lymphoid development and only moderate defects in immune responses (Kundg et al., 1993; Schorle et al., 1991). To understand the complex biologic properties of IL-2 and related cytokines, it is essential to consider the signals generated by its receptor. The IL-2R is multimeric, consisting of two obligate signaling subunits, IL-2RP (CD122) and yc (CD132), and a variably expressed IL-ZRa subunit (CD25), which regulates affinity for IL-2 (Kondo et al., 1994a; Nakainura et nE., 1994; Nelson et nl., 1994; Siege1 et al., 1987; Waldmann, 1989; Wang and Smith, 1987). As is the case with several other cytokine receptor systems, IL-2 receptor components are shared with receptors for other lymphokines, including IL-4, 7, 9, and 1Fi (reviewed in Leonard et al., 1994). This sharing of subunits explains some of the overlapping 1
Capvnght 0 I998 In Aucirmic Pirs, All ngllt, of ~ p ~ ~ d IIIt any ~ ~6 tm m rercrvrd lK)fi5-?iififlR 525 ()(I
2
BRAD H NELSON AND D E N N I S M. WILLERFORD
properties of these lymphokines, and in many instances it is difficult to sort out which cytokine/receptor interaction is responsible in vivo for the generation of a given cellular signal. Like other cytokine receptors, the IL-2R contains no intrinsic enzymatic activity. Rather, the intracellular portions of the receptor chains associate with a variety of cytoplasmic proteins, including tyrosine kinases of the Jak family. Oligomerization of receptor subunits brings these regulatory enzymes into close proximity, activating the signaling complex by phosphorylation of regulatory tyrosines on the kinases themselves, as well as on the cytoplasmic domains of IL-2RP and yL.The phosphorylated cytoplasmic domain of IL-2RP plays a critical role in attracting downstream signaling molecules into the activated receptor complex, where they may serve as substrates for receptor-associated enzymes. A diversity of downstream signals are activated by the IL-BR, including those which promote cell growth and division, as well as signals that influence cell survival and differentiation. Our understanding of these downstream signals is far from complete, particularly with regard to the regulation of cellular responses other than proliferation. The biologic function of the IL-2 receptor has also been examined at a genetic level in studies of humans with immune deficiencies attributable to mutations in receptor subunits, as well as in mice generated using genetargeting techniques. The phenotypes of these mutations fall into two broad categories. Mutations in the y' gene lead to severe defects in lymphopoiesis in knockout mice and, in humans, are responsible for X-linked severe combined iinmunodeficiency (XSCID) (for review see Leonard et al., 1994). The developmental and functional defects in T and B cells associated with disruption of yc signaling reflect the participation of this subunit in multiple cytokine receptor complexes. In contrast, mutations that selectively affect signals delivered by the high-affinity IL-2R do not block T- and B-cell development, but lead to an inability to regulate the overall size of the lymphoid tissues, fimctiond defects in immune responses, and autoimmunity. This constellation of defects demonstrates that I I d - 2 ~ signals are essential for the proper regulation of immune respollses 2ind underscores their negative regulatory role in homeostasis of the peripheral lymphoid tissues. This review attempts to connect what is known regarding receptor activation from biochemical and cellular studies with the now considerable literature regarding immune system developrnent and function when various components of the IL-2R signal are disrupted. The premise is that the diverse functions that are revealed by such genetic studies can guide future investigations aimed at understanding the intracellular signals generated by the IL-2R.
HIO1,OGY OF TtlE I N T E R L E U I I N - 2 RECEPTOR
3
II. The 11-2 Receptor Complex
A. BINDIN(:OF IL-2 BY
THE
IL-2R
Depending on which of the three 1L-2R subunits are expressed by a given cell, several different binding affinities for IL-2 are observed. The a subunit alone binds IL-2 with low affinity (& 10 nM), whereas the p subunit binds IL-2 with veiy low affinity (K,, 100 nM). When coexpressed, IL-2Ra and p form a pseiido-liigli-affinity receptor complex (I& 30 pM) (Anderson et al., 1995; Arinia e f al., 1992). Unlike IL-2Ra and /3, the yc subunit has no measurable affinity for IL-2, but when coexpressed with IL-2RP forins an intermediate affinity receptor complex (& 1 nM). Finally, coexpression of the a, 0, and y( subunits results in 10 pM) that is thought to the formation of a high-affinity IL-2R (& consist of a triineric a:P:y, complex (Takeshita et al., 199213). Of all these possible receptor combinations, only two are competent to ~ ) , is expressed signal. Thus, the intermediate affinity receptor ( p : ~ which by NK cells, macrophages, and resting T cells, can signal in the presence of high concentrations of IL-2, whereas the high-affinity receptor complex (a:p:yc)is expressed on activated lymphocytes and signals even at low concentrations of IL-2 (reviewed in Smith, 1988). Although signaling by both interinediate- and high-affinity IL-2R complexes can be demonstrated readily in uitro, mice with a targeted disruption of the IL-2Ra gene show a very similar phenotype to mice lacking the IL-2 gene itself (Schorle et nl., 1991; Willerford et al., 199S), suggesting that most of the biologic effects of IL-2 in o are mediated by signals delivered through the highaffinity ( a : P : y cIL-2R ) (see Section VIII). Thus, tlie expression of the IL2Ra subunit by cells, although not required for intracellular signaling per se, is nonetheless a critical determinant of 1L-2 responsiveness. The a,0, and ycchains of the high-affinity IL-2R bind to distinct sites on IL-2, and these associations a p p e a to occur in a stepwise manner (Imler, et d., 1992; Moreau c t d . , 199.5; Sauve et al., 1991; Voss et al., 1993; Zurawski, et al., 1990).Available data suggest a model for IL-B/IL2R binding whereby IL-2 first binds to the IL-2Ra and IL-2RP chains and, subsequently, the yc chain is recruited into the complex. In support of this model, evidence shows that IL-2Ra and IL-2RP preassociate as a heterodimer in the absence of ligand. First, the affinity of the IL-2RP chain for IL-2 is very low unless IL-2RP is coexpressed with IL-2Ra. This synergistic effect does not require prior binding of ligand to IL-ZRa, as a mutant form of IL-2 (F42A) that fails to bind to the isolated IL-2Ra subunit nevertheless binds to the IL-2RdIL-2RP complex with higher affinity than to IL-2RP done (Grant et al., 1992; Roessler et nl., 1994). Second, precise ineasurenients of receptodligand interactions using tlie
-
-
-
-
-
4
BRAD H. NELSON AND DENNIS M . WILLERFORD
technique of surface plasmon resonance demonstrate that the extracellular domains of IL-2Ra and IL-2RP bind IL-2 with kinetics that are indicative of simultaneous contributions from both subunits (Balasubramanian et al., 1995). Once a complex is formed between IL-2 and the IL-2Ra:IL-2RP heterodimer, the yc chain is recruited to the receptor complex, an interaction that is driven by a 10-fold increase in receptor affinity, via both an increase in the association rate constant and a decrease in the dissociation rate constant (Matsuoka et al., 1993). The recruitment of yc is required for intracellular signaling, as experiments with chimeric receptor chains have shown that signaling is initiated by ligand-induced heterodimerization of the cytoplasmic domains of IL-2RP and yc (Nakamura et al., 1994; Nelson, et aE., 1994) (see Section V,B). Thus, the association and dissociation of yewith the receptor complex serve to switch the receptor between inactive and active states. One implication of this model is that the magnitude of the IL-2R signal could be influenced by the presence of other cytokines that use yc as part of their receptors (see later), although such competitive interactions have not been demonstrated directly.
B. MOLECULAR CLONING AND STRUCTURAL FEATURES OF THE a, P, AND yc SUBUNITS The IL-2Ra chain was the first IL-2R component to be cloned, this being accomplished in 1984 by three groups (Cosman et al., 1984; Leonard et al., 1984; Nikaido et al., 1984). IL-2Ra, also known as Tac antigen or CD25, is a -55-kDa polypeptide with an extracellular domain containing 219 amino acid residues, a transmembrane domain of 19 residues, and a cytoplasmic domain containing only 13 residues. Although the cytoplasmic domain is very short, it is nonetheless highly conserved between mice and humans, suggesting that it may play an important functional role, although to date none has been identified. Unlike IL-2Rp and ye,IL-2Ra does not share the typical features of the hematopoietin receptor superfamily. The a chain of the IL-15 receptor has been cloned and found to have structural homology to IL-2Ra (Giri et nl., 1995). In particular, the two chains share an extracellular protein-binding motif known as the “sushi domain,” which is also found in complement receptor proteins. The genes encoding IL2Ra and IL-15Ra are located close to one another on chromosome 10 in humans and chromosome 2 in mice and have a similar exodintron structure, suggesting a close evolutionary relationship (Anderson et al., 1995). Based on binding studies with cloned IL-2Ra, it became clear that this protein constituted the low-affinityreceptor and that one or more additional subunits must be required to form the high-affinity IL-2R complex. A monoclonal antibody that inhibited binding of IL-2 to the putative IL2Rp chain (Tsudo et al., 1989) was utilized by Taniguchi and colleagues
BIOLOGY OF THE INTERLEUKIN-2 KECEPTOH
5
to clone the corresponding cDNA for this receptor subunit (Hatakeyaina et al., 1989b).The -75-kDa IL-BRP chain is composed of an extracellular domain of 214 residues, a transmembrane domain of 25 residues, and a long cytoplasmic domain of 286 residues. IL-2RP displays the characteristic structural features shared by members of the hematopoietin receptor superfamily, including a set of four conserved cysteine residues in the extracellular domain and a ineinbraiie-proximal WSXWS motif. The intracellular domain of IL-BRP, wliich lacks any intrinsic enzymatic function, displays the box 1 and box 2 motifs characteristic of this receptor superfanily (Hatakeyama et al., 198913).The gene encoding hunian IL-2Rp is located on chromosome 22q12-13 (Shibuya et a l , 1990) and, thus far, has not been associated with any genetic diseases in humans. Scatchard analysis of IL-2 binding to lymphoid cells expressing IL-2RP but not IL-2Ra showed the presence of an intermediate affinity receptor with & values in the nanomolar range (Hatakeyarna et al., 1989b). However, when analogous studies were performed in nonlymphoid cells, IL-2 failed to bind IL-2RP in the absence of IL-2Ra. To explain this Ascrepancy, it was proposed that the IL-2R might contain a third subunit that was expressed in lymphoid but not nonlymphoid cells (Hatakeyama et al., 1989b). This hypothesis was supported by studies showing that a -64kDa polypeptide coprecipitated with IL-2RP in the presence of IL-2 (Takeshita et al., 1990). A functional role for the 64-kDa chain in signaling by both intermediate- and high-affinity IL-2R complexes was also suggested based on correlations between expression of this protein and IL-2 responsiveness (Arima et d.,1992; Takeshita et al., 1992a; Voss et al., 1992; Zurawski et al., 1990).The 64-kDa chain was purified by coimmunoprecipitation with IL-2R0, followed by two-dimensional gel electrophoresis. Peptide sequence data obtained from the purified protein were used to isolate a full-length cDNA encoding the IL-2Ry subunit (Takeshita et al., 1992a). The newly identified IL-2Ry chain (now called yJ was found to contain 232 residues in the extracellular domain, 29 in the transmembrane domain, and 86 in the cytoplasmic domain. Furthermore, yc was a novel inember of the hematopoietic receptor superfamily by virtue of conserved cysteine residues and a WSXWS motif in the extraceIlular domain, as well as a box 1 motif in the cytoplasmic domain (Takeshita et al., 1992a). The gene encoding ycis located at Xq13 in huinans (Noguchi et al., 1993a,c; Puck et al., 1993) and region 40 of the X chromosome in mice (Cao et al., 1993; DiSanto et nl., 1994). C. SHARING OF THE IL-2RP A N D -yc CHAINS WITH OTIHEI-I CYTOKINE RECEPTORS The location of the gene encoding the ycchain proved to be extremely significant, as it corresponded to the locus for XSCID (Noguchi et al.,
6
BRAD 11. NELSON AND DENNIS M U‘ILLEKFOKD
1993c; Puck et nl., 1993). Males with this genetic disorder lack T cells and, as a result, are severely iinmunocomproinised. The developmental block in the T-cell lineage in XSCID is in marked contrast to the phenotype of IL-2 -/- mice (Schorle et nl., 1991), which produce normal numbers of functional lymphocytes and are able to mount iinrnune responses (see Section VII,E,l). This discrepancy between XSCID and IL-2 4- phenotypes suggested that ycwas a component of another receptor complex that was essential for normal T-cell development (Noguchi et al., 1993~). A number of studies soon demonstrated that in addition to the IL-2R, the yc chain was a functional component of the receptors for IL-4, IL-7, IL-9, and IL-15 (reviewed in Leonard et nl., 1994). Chemical cross-linking experiineiits demonstrated that yL associated with the IL-4, IL-7, and IL-9 receptors in the presence of ligand. Moreover, antibodies to the extracellular domain of yc were found to block proliferative responses to these cytokines (Kimura et aE., 1995; Kondo et al., 1993, 1994b; Noguchi et al., 199313; Russell et al., 1993b). Furthermore, a dominant-interfering mutant of yL,which lacked the cytoplasmic domain, was found to inhibit the proliferation of BAF3 cells in response to either IL-2 or IL-7 (Kawahara,et al., 1994). Finally, receptor reconstitution experiments showed that yL was required for intracellular signaling in response to IL-15 (Gin et al., 1994). Due to its involvement in at least five different receptor complexes, the IL-2Ry chain is now referred to as the “common y chain,” or yr (Noguchi et al., 199317). The XSCID phenotype, as well as the severe lyinphopoietic defect in mice with a targeted disruption of the yL gene (Cao et al., 1995; DiSanto et al., 1995), is now thought to result from defective signaling by multiple cytokine receptors, of which the IL-7 receptor is probably the most important (Peschon et al., 1994; von Freeden-Jeffryet nl., 1995) (see Section VII). The IL-BRP chain is also not exclusive to the IL-BR, as it has been shown to be a component of the IL-15 receptor (Bainford et al., 1994; Carson et al., 1994; Gin et nl., 1995; Grabstein et al., 1994). Indeed, the IL-2 and IL-15 receptors are very similar, in that both contain IL-2RP and yc in combination with either IL-2Ra or IL-l5Ra, respectively (Gin et al., 1994, 1995). The similarities extend further, as the IL-2Ra and IL15Ra chains are themselves structurally homologous and both are able to bind ligand independent of the IL-2RP or yr chains. Indeed, the IL-15Ra subunit alone binds IL-15 with an affinity in the 10 pM range (Gin et nl., 1995). As might be expected, IL-2 and IL-15 appear to activate the same intracellular signaling pathways in lymphocytes (Grabstein et al., 1994; Johnston et al., 1995a; Lin et al., 1995) Notably, however, a second receptor complex for IL-15 has been reported in mast cells that operates indepen-
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
7
dent of IL-2RP and ycand generates a distinct intracellular signal involving the activation of Jak2, rather than Jakl and Jak3 (Tagaya et al., 1996a,b). 111. 11-2 Receptor Expression
A. EXPRESSION OF THE IL-2Ra CHAIN Of tlie three IL-2R subunits, the IL-2Ra chain demonstrates tlie most tightly regulated expression. In the thymus, IL-2Ra is expressed on primitive CD4-CD8-CD3- triple-negative (TN) cells, a stage that marks the first steps in T-cell receptor (TCR) rearrangement and irreversible conirnitrnent to the T-cell lineage (reviewed in Godfrey and Zlotnick, 1993). Expression of IL-2Ra is lost abruptly following successful TCRP rearranginent and expression and remains off for the remainder of thyrnocyte development (Godfrey and Zlotnick, 1993; Rothenberg, 1992). Mature, resting T cells in the periphery fail to express the IL-2Ra chain and therefore are unresponsive to IL-2. However, following TCR stimulation, IL-2Ra transcription is induced, leading to cell surface expression (reviewed in Crabtree, 1989; Rothenberg, 1992). Expression of IL-2Ra may also be induced in T cells by non-TCR stimuli, including IL-1 or tumor necrosis factor (TNF)-a (Plaetinck et nl., 1990). IL-2Ra is also expressed in tlie B-cell lineage. In mice, bone marrow pre-B cells express IL-2Ra, although the functional significance of this is uncertain, as IL-2 does not effectively support proliferation of such cells, and a similar expression pattern has not been described in humans (Chen al., 1994; Rolink et al., 1994). On mature B cells, IL-2Ra is induced following activation through the B-cell receptor (BCR), resulting in the expression of a functional, highaffinity IL-2R (Jung et al., 1984; Tsudo et al., 1984; Waldmann et al., 1984; Zubler et al., 1984). The molecular rnechanisni underlying the exquisite regulation of IL2Ra gene expression by tlie TCR has been under investigation for over a decade. A potent enhancer has been identified between positions -299 and -228 relative to the major transcription start site and is now termed PRRI (for positive regulatory region I ) (Lin et nl., 1990; Plaetinck et nl., 1990). This region contains binding sites for both NF-KB and serum response factor (SRF) and is responsive to diverse activation signals, including those resulting from TCR ligation or stimulation with phorbol esters, IL-1, and TNF-a. More recently, a second regulatory region termed PRRII has been identified between nucleotides -137 and -64 of the IL-2Ra gene (John et al., 1995). This region can function as a phorbol ester-inducible enhancer element and contains binding sites for both the Ets family protein Elf-1 and the cliromatin-associated protein HMG-I(Y). As Elf-1 has been found to associate with NF-KB p50 in vitro, it has been proposed that physical
8
BRAD 1-1, NELSON AND DENNIS M. WILLERFORD
interactions between these proteins may allow functional cooperativity between PRRI and PRRII (John et nl., 1995). IL-2 itself can also upregulate expression of the IL-2Ra chain and does so through a region of the IL-2Ra promoter/enhancer termed PRRIII, which lies distal to PRRI and PRRII (John et al., 1996; L'ecine et al., 1996; Plaetinck et al., 1990; Serdobova et al., 1997; Soldaini et al., 1995; Sperisen et al., 1995). PRRIII resembles PRRII in that it contains binding sites for Elf-1 and HMG-I(Y), but differs by also having a binding site for the transcription factor Stat5. As described in detail in Section VI,B,2, Stat5 activity is induced rapidly in lymphocytes in response to IL-2 stimulation and plays an essential role in IL-2-mediated expression of the IL-2Ra chain (John et al., 1996; L'ecine et al., 1996).
B. EXPRESSION OF THE IL-2RP CHAIN In contrast to IL-2Ra, the IL-2RP chain is constitutively expressed at low to moderate levels on resting T cells, B cells, NK cells, monocytes/ macrophages, and neutrophils (Begley et al., 1990; Djeu et al., 1993; Dukovich et al., 1987; Espinoza-Delgado et al., 1990; Sharon et al., 1990; Siegel et al., 1987; Wei et al., 1993 ;Zola et al., 1991). Expression of this subunit is modulated during T-cell development, as it is expressed at low levels on a small fraction of TN thymocytes, at very low levels on CD4+8+ double-positive (DP) thymocytes, and at moderate levels on the CD4-8' single-positive (SP) subset (Kondo et al., 1994a). Although IL-2RP differs from IL-2Ra in being constitutively expressed on cells, like IL-2Ra it is upregulated on mature T cells by stimulation with antigen or a variety of other agents (Casey et al., 1992; Cerdan et al., 1995; Hatakeyama et al., 1989b; Siegel et al., 1987). Similarly, IL-BRP expression is upregulated on B cells activated by BCR stimulation, which is enhanced by IL-4 and IL5 (Loughnan and Nossal, 1989). Studies of the 5' flanking portion of the human IL-2RP gene have led to the identification of three regulatory elements within the first 363 bp upstream of the major transcription start site (Lin et al., 1993; Lin and Leonard, 1997). One of these regions (-56 to -34) contains a binding site for Ets family proteins (specifically, Ets1- and GA-binding protein) and is involved in both basal and inducible promoter/enhancer activity. This site may also contribute to the tissuespecific expression of IL-2RP, as it was found to be active in lymphoid but not nonlymphoid cell lines. A second region (- 170 to - 139) binds the factors Spl and Sp3 as well as the immediate-early factor Egr-1, the latter being expressed in T cells on treatment with phorbol esters (Lin and Leonard, 1997). As with Ets-binding sites, mutation of the Spl- and Egr-l-binding sites severely disrupts the basal and inducible activity of the IL-2RP promoter/enhancer. Therefore, consistent with its complex pattern
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
9
of expression in cells, transcription of the IL-2RP gene appears to be regulated through interactions between both constitutive and inducible DNA-binding proteins.
C. EXPRESSION OF THE yLCHAIN Consistent with it being a component of at least five hematopoietic cytokine receptors, the yc chain is expressed constitutively in multiple hematopoietic lineages, including CD4+ and CD8+ T cells, B cells, N K cells, monocytes/macrophages, granulocytes, and neutrophils (Bosco et al., 1994; Epling-Burnette et al., 1995; He et al., 1995; Kondo et al., 1994a; Liu et al., 1994; Sugainura et al., 1996).The ycchain is expressed throughout thymocyte development, from the TN to SP stages. In the spleen, y,is expressed by both mature T cells and B220' B cells (Kondo et nl., 1994a). In contrast to the IL-2Ra and /3 chains, surface expression of y,decreases in response to T-cell activation rather than increasing (Kondo et a1 , 19944. The decline in expression is rapid, beginning within 4 hr of stimulation, and transient, being reversed by 24 hr poststimulation. This appears to be a T-cell-specific phenomenon, as activation of B220' B cells with LPS enhances, rather than suppresses, expression of yc(Kondo et al., 1994a). One explanation for the decreased expression of yc in activated T cells may come from work demonstrating that ycis a target of the calciumactivated neutral protease calpain, due to a so-called PEST sequence in the cytoplasmic domain of yL that serves as a recognition site for the enzyme (Noguchi et al., 1997). Activation of thymocytes by treatment with anti-CD3 was shown to induce the proteolytic degradation of the ycsubunit, presumably due to a rise in intracellular calcium and the consequent activation of calpain. The addition of A h M , an inhibitor of m-calpain, not only inhibited the anti-CD3-triggered degradation of yLin thymocytes, but also enhanced the proliferative response of CD4' thymocytes to anti-CD3 stimulation (Noguchi et al., 1997). The 5' flanhng region of the human gene encoding yc includes cis elements that appear to be involved in the transcriptional regulation of yc (Noguchi et al., 1993a; Ohbo et al., 1995). A 633-bp region upstream of the transcription start site is sufficient to drive expression of a reporter gene in hematopoietic cell lines. Within this region lie sequences characteristic for the binding of several transcription factors, including PU.l and PEA-3. Another binding site characteristic of ETS proteins is found within a short segment harboring basal promoter activity. Reporter gene expression was enhanced by treatment of cells with either phorbol esters or phytohemagglutinin (PHA), despite the fact that surface expression of the ycchain decreases on activation of T cells with similar stimuli. Furthermore, IL-2 was found to decrease expression of this reporter gene as well as endogenous yc
10
BRAD H . NELSON AND DEKNIS M . WILLEKFORD
inRNA. Therefore, it appears that the regulation of yr expression in resting and activated lymphocytes may be quite complex, involving at a minimum both transcriptional and posttranslational mechanisms. IV. Cellular Responses to 11-2 Receptor Signals
A. IL-2 AS A GHOWTII FACTOR 1. T-cell Prol$eration. IL-2 was initially identified as T-cell growth factor, a substance present in supernatants of PHA-stimulated lymphoblasts that promoted the growth of T cells in vitro (Gillis et al.,1978; Gillis and Smith, 197713; Morgan et al.,1976). This discovery led to the development of T-cell lines that could be maintained indefinitely in uitro, reagents that have contributed iinmensely to our current understanding of T-cell biology (reviewed in Smith, 1988).T cells require two sequential signals to engage in active proliferation (reviewed in Crabtree, 1989; Schwartz, 1990; Weiss and Imboden, 1987). The first is provided by TCR engagement, which induces expression of a high-affinity IL-2R complex (Cantrell and Smith, 1983; Leonard et al., 1982; Meuer et al., 1984; Robb et al., 1981; Waldmann, 1989), a process that involves de novo expression of IL-2Ra, as well as enhanced expression of the IL-2RP chain (Kondo et ul., 1994a; Leonard et al., 1982; Siege1 et al., 1987; Waldmann, 1989, 1991). TCR engagement also increases the expression of Jak3, a critical component of proliferative signaling (Kawamura et al., 1994). Once T cells are sensitized by TCR signals, IL-2 acts as a cell cycle progression factor, thus supporting proliferation. T-cell proliferation in vitro is suppressed when the IL-2/IL-2R interaction is blocked by antibodies to IL-2 or IL-2Ra (Leonard et al., 1982; Smith et nl., 1983). Thus, when standard in uitro culture conditions are used, IL2 appears to be the major factor promoting the growth of T cells. In the context of appropriate costimulatory conditions, TCR signaling also induces synthesis of IL-2; hence, T-cell proliferation is actuated by an autocrine and/or paracrine hormonal circuit (Meuer et al., 1984; Smith, 1988). One consequence for iinmunoregulation is that activation of both IL-2 and its receptor by the TCR restricts the local IL-2 response to cells with appropriate antigen specificity, thereby minimizing bystander effects. However, the intracellular signaling pathways activating IL-2 and IL-2Ra expression diverge downstream of the TCWCDS complex (Crabtree, 1989;Rothenberg, 1992), and each pathway may therefore be influenced by separate sets of conditions within the cell. For example, TCR stimulation is sufficient to induce IL-2Ra in the absence of costimulation provided by accessory cells, whereas IL-2 secretion requires an additional signal,
13IOI.OGY 01.' T l I E IKTERLEUKIN-2 RECEPTOR
11
such as that supplied by CD28 ligation (Crabtree, 1989; Schwartz, 1990). In addition, IL-2Ra expression may be induced independent of TCR signals, e.g., by IL-1 or TNF (Plaetinck et a1 , 1990). The distinct requirements for IL-2 and IL-2Ra induction are also reflected in distinct patterns of expression in T-cell sublineages: although IL-2Ra expression is common to both CD4' and CD8+T cells, IL-2 production is largely restricted to the CD4' subset, particularly, naive cells and those with a T h l profile (Schwartz, 1990). This partial dissociation of IL-2 and IL-2Ra expression underscores the paracrine nature of IL-%mediated signals and reflects the specialized functions of lymphocyte subsets. 2. Cell Cycle Regulation hi) IL-2 TCR signaling in resting, GO T cells activates metabolic pathways and generates characteristic changes in cell size and RNA content indicative of progression to the G1 phase of the cell cycle However, in the absence of IL-2, S-phase entry does not occur. Moreover, T cells deprived of IL2 undergo cell cycle arrest in G1 (Cantrell and Smith, 1984).IL-2 promotes the characteristic niorpliologic changes and RNA accumulation associated wit11 progression through G 1, induces the expression of growth-associated protooncogenes, including c-myc and c-myh (Sliibuyaet nl., 1992; Stern and Smith, 1986),and promotes entiy into S phase. The molecular mechanisms responsible for the effect of IL-2 on cell cycle progression have been investigated. In eukaiyotes, the cell cycle is coordinated by the sequential activation of cyclin-dependent hnases (CDK), which are regulated by interaction with cyclin proteins and by inhibitors of cyclin/CDK function (for reviews see Hartwell and Kastan, 1994; Hunter and Pines, 1994; Sherr, 1994; Sherr and Roberts, 199.5). TCR signals stimulate the synthesis of CDKs and G1 cyclin proteins, but this is not sufficient to generate the active CDK complexes required for progression to S phase (Ajchenbaum et nl., 1993; Firpo et nl., 1994; Karnitz and Abraham, 1996). In addition to further promoting the expression of cyclin and CDK proteins, IL-2R signals give rise to active cycliidCDK complexes (Firpo et nl., 1994). A specific mechanism for these properties of IL-2 on cell cycle progression effect has emerged based on the identification of G1 cyclidCDK inhibitors (Sherr and Roberts, 1995), the most relevant in this regard being p27"'1". Resting T cells express high levels of p27"1", which are unaffected by TCR ligation (Firpo et nl., 1994);hence the G1 cylins and CDKs that are induced by TCR signals may associate, but reinain functionally inactive. In addition to promoting the synthesis of additional cyclin and CDK proteins, IL-2R signals lead to a rapid decline in p27K'I"levels, and these changes are accompanied by the appearance of active cycliii E/CDK2 complexes and cell cycle progression (Fiipo et nl., 1994; Nourse et nl., 1994). The effects
12
BRAD H. NELSON AND DENNIS M . WILLERFOHD
of IL-2R signals on cell cycle progression, particularly the decline in ~ 2 7 ~ ' p ' levels, are inhibited by rapamycin, implying that these effects are mediated by activation of the mammalian target of rapamycin (mTOR) protein by the IL-2R (see Section VI,C,3). These studies provide a molecular explanation for the requirement for two sequential signals for T-cell proliferation and identi? a very specific role played by IL-2R signals in regulating the cell cycle in T cells. The effects of IL-2 in the cell cycle have been interpreted to indicate an in vivo role for IL-2 in postantigenic clonal expansion of T cells. This conclusion assumes that the outcome of transiting the cell cycle is the reproduction of T cells. However, it is important to consider that several alternative cell fates are also determined by processes linked to traverse of the cell cycle, notably cell death and differentiation into memory and effector cells. Importantly, these fates tend to limit or even negate the effects of cell reproduction. Thus, the IL-2R, through its effects on cell cycle progression, may promote several different fates in T cells. The specific fate adopted by a given cell is therefore likely to depend on the context provided by other signals.
3. T-Cell ProZ~erat~~n in the Absence uf IL-2R Function Proliferation of T cells in response to mitogens in vitro is largely dependent on the production of IL-2 and subsequent IL-2R signaling (see earlier discussion). These experiments have been revisited more recently utilizing T cells from mice with targeted intactivation of the genes for IL-2 (Schorle et al., 1991) or for IL-2Ra (Willerford et al., 1995). In both these strains, T-cell proliferation in vitro to lectins, antibody cross-linking of CD3, or antigen is markedly suppressed in homozygous-deficient mice (Kramer et al., 1994; Schirnpl et al., 1992, 1994; Schorle et al., 1991; Van Parijs et al., 1997; Willerford et al., 1995). However, it is important to note that in most experiments that have interrupted IL-2/IL-2R signaling, a degree of T-cell proliferation is still observed, indicating the presence of other Tcell growth factors. Indeed, IL-4, IL-7, and IL-15 all promote T-cell proliferation in a manner similar to IL-2, a property that reflects the shared utilization of yc in the receptors for these lymphokines as well as common signals generated by their specific subunits (Bamford et al., 1994; Chazen et a!., 1989; Gin et al., 1995; Grabstein et al., 1994; Keegan et al., 1994; Morrisey et al., 1989; Paul, 1991). Given that peripheral T cells from ycdeficient mice also display low-level proliferation in response to TCR signals, there must be additional, ?,-independent signals that contribute to T-cell cycle progression (Cao et al., 1995).Thus, while IL-2 is the major growth factor used by T cells stimulated through the TCR under typical cell culture conditions, this role can clearly be subserved by other factors.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
13
In uiuo studies that address the role of IL-2R signals in T-cell clonal expansion are discussed fiirther in Section VIII. B. ROLEOF IL-2R SIGNALS I N LYMPHOCYTE EFFECTOR FUNCTION In addition to its role in the expansion of activated lymphocytes, IL2R signals support the functional differentiation of mature lymphocytes, including cytotoxic T lymplioyctes (CTL),natural killer cells, and B cells. As with other cytokine receptors that promote cellular differentiation, it has been difficult to separate general effects of IL-2 on the growth and survival of cells that are developing along a given pathway from receptormediated induction of lineage-specificdifferentiation events. Nevertheless, the complexity of intracellular signaIing pathways that are activated by the IL-2R leaves much room for the latter possibility. In historical terms, the concept of IL-2 as a T-cell growth factor has perhaps channeled investigative energy toward understanding proliferative responses at the expense of effects on differentiation; at a practical level, the latter are also more difficult to assay. Either way, one of the main challenges at present in understanding IL-2R signaling is to understand the connection between receptor activation and lymphocyte differentiation events. As early experiments utilized IL-2 to grow CD8' T cells in uitro, it was also determined that these IL-2-dependent cell lines retained CTL activity. Moreover, IL-2 enhances CTL activity in activated primary T cells, and this effect is inhibited by antibodies that block IL-WIL-2R interaction (Depper et al., 1983; Gillis and Watson, 1981; Maraskovsky et al., 1989). That this activity represents more than trophic effects of IL-2 is suggested by the fact that IL-2R signals induce or upregulate mRNA for FasL, perforin, and granzyme B, all of which are involved in the mechanism of CTL-mediated killing (Liu et nl., 1990; Makrigiannis and Hoskin, 1997; Smytli et al., 1990; Suda et al., 1995). NK cells also proliferate and upregulate their cytolytic activity in response to IL-2 (reviewed in Trinchieri, 1989). However, such cells do not generally express IL-2Ra! (Tsudo et al., 1987) and, as a result, these effects require relatively high doses of IL-2. In contrast, IL-15 delivers similar signals to NK cells and is effective at picomolar concentrations (Carson et al., 1997; DiSanto, 1997), suggesting that the physiologic signals promoting N K effector function are likely delivered by IL-2RP and ycin the context of the IL-15R. Host defenses medated by the huinoral arm of the immune response involve the activation of B cells by antigen-mediated BCR stimulation, as well as interactions with activated T cells. T-cell heIp for B-cell effector function includes promoting proliferation as well as differentiation into antibody-secreting cells. T-cell signals are delivered by dxect interactions, as well as by soluble mediators, including IL-2 and IL-4 (Howard, et nl.,
14
BRAD II. NELSON A N D DENNIS M. WILLERFORD
1984). Following activation by BCR signals, induction of IL-2Ra (as well as upregulation of IL-2RP) results in expression of the high-affinity IL2R (Jung et nl., 1984; Loughnan and Nossal, 1989; Tsudo et al., 1984; Waldmann et nl., 1984; Zubler et al., 1984). Analogous to its effects on T cells, IL-2 promotes the proliferation of activated B cells (Tsudo et al., 1984; Zubler et al., 1984). In addition, IL-2 promotes the secretion of IgM in primary B-cell cultures (Nakanishi et al., 1984a,b; Waldmann et al., 1984). This property reflects specific effects of IL-2R signals on gene expression required for B-cell differentiation, notably on chromatin reinodeling at the J chain locus, followed by transcriptional activation of this gene (Blackman et al., 1986). Thus, the effects of IL-2R signaling in B cells include both events in common with those in T cells, such as promoting cell cycle progression, as well as distinct patterns of downstream gene activation. OF CELLSURVIVAL AND CELLDEATII BY THE IL-2R C. REGULATION
Physiologic cell death is now recognized as a fundamental mechanism for development and tissue homeostasis in multicellular organisms (for reviews see Raff, 1992; Vaux, 1993; White, 1996; Yang and Korsmeyer, 1996). Cell death is regulated by many factors, including a long list of developmental and hormonal signals, as well as by toxic and infectious insults. Ultimately, these regulatory signals converge on a central cascade of cysteine proteases, known as caspases, which, on activation, affect the characteristic membrane and nuclear changes that are collectively recognized as apoptosis (reviewed in Martin and Green, 199s; Takahashi and Earnshaw, 1996). Cell death is a particularly important regulatory process in the lymphoid lineage. During T- and B-cell development, a large number of cells must be eliminated because the random nature of V(D)J recombination frequently results in the generation of cells bearing antigen receptors, which are either reactive with self proteins or, in the case of T cells, may lack the proper determinants to interact with MHC molecules on antigenpresenting cells. In the peripheral lymphoid compartment, the overall number of lymphocytes remains constant over time, despite a continuous input of cells from the primary lymphoid organs and the expansion of reactive clones during immune responses. Here, cell death is required both to regulate the overall size of the secondary lymphoid tissues and to adjust their composition so as to maintain a balance between naive cells with a diverse repertoire of antigen specificities and activated or memory lymphocytes with repertoires biased toward recognition of previously encountered pathogens (Tanchot and Rocha, 1997). In considering the regulatory role of cell death in the immune system, it is useful to make a distinction between death that occurs in resting
BIOLOGY OF THE INTEKLEUKIN-2 RECEPTOR
15
lyinphoctyes and that which occurs as a consequence of cellular activation, the latter being termed activation-induced cell death (AICD). Death in resting lymphocytes is generally viewed as a default pathway, which is opposed by positive survival signals. Such signals may be provided by extracellular stimuli, and many examples of trophic factors regulating cell survival have been described (Raff, 1992; White, 1996). Susceptibility of resting cells to apoptosis correlates inversely with expression of cytoprotective members of the bcl-2 family (for review see Yang and Korsmeyer, 1996); in the lymphoid system, bcl-2 and bcl-x have been characterized most extensively . Bcl-2 is overexpressed in the context of chroinosome translocations involving the immunoglobulin loci, and these appear to underlie the majority of follicular lymphomas in humans (Yang and Korsmeyer, 1996). When hcl-2 is overexpressed in peripheral lymphocytes in mice, such cells are relatively resistant to resting cell death in vitro and appear to have an extended lifespan in v i m . In particular, the accumulation of ineinory B cells is enhanced (Sentman et nl., 1991; Strasser et nl., 1990; Strasser et nl., 1991a,b). Peripheral lymphocytes require expression of bcl-2 for optimal viability, as targeted inactivation of the bcl-2 gene results in resting, mature T and B cells with heightened sensitivity to apoptosis, and involution of the peripheral lymphoid tissues in adult inice (Veis et nl., 1993). Bcl-x appears to share inany functional properties with bcl-2 and is inducible in T cells following stiinulation through TCR and CD28 (Boise et nl., 1993,1995b;Chao et nl., 1995).Bcl-x is required for thyinocyte survival at the immature DN stage (Ma et nl., 1995b). Taken together, these studies indicate a critical role of bcl-2 Family members in mediating resting cell survival in the immune system. The tendency of resting lymphocytes to undergo cell death appears to be present throughout development. At the earliest thymic stage, cells that have initiated V( D)J recombination will die unless a productive antigen receptor rearrangement occurs, leading to the assembly of a signalingcompetent pre-TCR (Rothenberg, 1992;Willerford et nl., 1996). Most DN tliyinocytes die by a default pathway in the absence of positive selection, which amounts to a survival (and differentiation?) signal delivered following interaction between TCR and MHC molecules expressed on thymic epithelial cells (Kisielow and von Boehmer, 1995) (see Section VI1,A). It has been determined that the default pathway of cell death also operates in the peripheral lymphoid compartment, as survival of peripheral T cells is dependent on the expression of MHC molecules (Kirberg et nl., 1997; Takeda et nl., 199613; Tanchot et nl., 1997). This requirement is most notable for naive T cells and occurs in the absence of cognate high-affinity antigen. MHC-dependent, antigen-independent peripheral T-cell survival is reminiscent of positive selection in the thymus and could potentially
16
BRAD H. NELSON AND DENNIS M. WILLERFORD
involve the recognition of MHC/self-peptide complexes by the TCR. B cells also require an intact antigen receptor in order to survive in the periphery, based on the observation that ablation of surface Ig expression on mature cells using conditional gene knockout technology results in disappearance of the peripheral B-cell population (Lam et al., 1997).These stuhes suggest that activity of the antigen receptor signaling complex in peripheral lymphocytes is required for survival. Thus, lymphocytes have an inherent propensity to undergo apoptosis in the absence of survival signals,which appear to be delivered primarily through the antigen receptor signaling complex. This propensity to undergo cell death reflects the need for ongoing cellular turnover in the immune system and is presumably an important mechanism by which homeostasis is achieved. It has long been noted that IL-2 is required for the sui-vival of T cells in culture (Smith, 1988). This effect can even be seen on resting T cells, where apoptosis is slowed via signals delivered through the intennediateaffinity (IL-2RP:yJ IL-2R (Boise et al., 1995a; Gonzalez-Garcia et al., 1997).It is not clear how viability is enhanced in such cells, as the cytoprotective effect of IL-2 in resting cells is not accompanied by the upregulation of bcl-2, whereas bct-XL was induced in one report (Gonzalez-Garcia et al., 1997) but not in another (Boise et al., 1995a). The in vim relevance of this pathway is not known. In contrast to apoptosis in resting cells, AICD occurs in the circumstance of cellular activation induced by antigen receptor stimulation. Rather that regulating survival against a default death pathway, AICD delivers a positive death stimulus to the cell. This distinction presents difficulties, as antigen receptors may deliver survival signals under some Circumstances and death signals under others. The difference between these outcomes is commonly understood to reflect a “weak’ signal in the case of survival and a “strong” signal in the case of AICD. A firm understanding of the parameters that govern antigen receptor signal strength is elusive at present; however, examples at either end of this spectrum are fairly clear. During T- and B-cell development, AICD is termed negative selection and serves to eliminate T and B cells that express antigen receptors recognizing self-antigens with high affinity, a process that is fundamental to self-tolerance (for reviews see Goodnow et al., 1995; Green and Scott, 1994; Kisielow and von Boehmer, 1995; Rothenberg, 1992).Activation of mature T cells by antigen or bacterial superantigens also induces substantial cell death under some circumstances (Jones et al., 1990; Kawabe and Ochi, 1991; Liu and Janeway, 1990 ;MacDonald et al., 1991; Rocha and von Boehmer, 1991; Russell et nl., 1991; Shi et at., 1989; Webb et al., 1990).AICD is also one potential outcome following ligation of the B-cell receptor on mature B cells (Goodnow et al., 1995; Green and Scott, 1994).The pattern of AICD in response
HIOL0C;Y OF THE INTERLEUKIN-2 RECEPTOR
17
to superantigen immunization in vivo correlates with cell cycle progression. T-cell subsets bearing V/3 segments that react with particular superantigen expand over a period of several days; typically, deletion then occcurs such that after 10 days the reactive V/3 subsets are reduced to half their original levels. Cells undergoing apoptosis under these condition are those that have undergone DNA replication in vivo (MacDonald et al., 1991; Renno et al., 1995). Other data indicate that AICD in vitro is closely tied to cell cycle progression (Boehnie and Lenardo, 1993; Zhu and Anasetti, 1995). AICD may act as a check on clonal expansion, as well as participating in the termination of immune responses in vivo. Moreover, exuberant activation of the immune system, either in response to a genuine pathogen or by autoantigen, can damage host tissues. AICD likely plays an important role in preventing such host injury (Abbas, 1996; Critchfield et al., 1994; Green and Scott, 1994). The mechanism for the induction of AICD in T cells involves death signals delivered by Fas and other members of the TNF receptor family. Fas (CD95) contains a characteristic intracelluar death domain, a protein interaction motif that is shared by a number of proteins that regulate apoptosis. Upon receptor ligation, the death domain mediates clustering of the death adapter protein FADD/MORT-1, which in turn recruits and activates caspase 8. This leads to the characteristic cascade of caspase activation that serves as a common pathway for apoptosis induced by several pathways (reviewed in Chinnaiyan and Dixit. 1997; Nagata, 1994). Fas is expressed constitutively on T cells and is further upregulated folllowing activation. The membrane-bound ligand for Fas (FasL) is induced on T cells following TCR stimulation and can activate Fas, either by interaction on the same cell or on neighboring cells, to induce cell death in an autocrine or paracrine manner (Alderson et n l , 1995; Brunner et aZ., 1995; Dhein et nt., 1995; Ju et nl., 1995). AICD in T cells can be blocked by inhibition of the Fas/FasL interaction. In addition, T cells derived from either Fasdeficient Ipr mice (Russell et nl., 1993a) or FasL-deficient gld mice (Russell and Wang, 1993) also exhibit deficient AICD responses to TCR signals, both in vitro and in vivo (Singer and Abbas, 1994). The phenotype of Zpr and gld mice underscores the physiologic importance of Fas-mediated AICD. These mice are susceptible to lymphoid expansion and autoimmunity, indicating that peripheral AICD is an important mechanism for maintaining self-tolerance (Singer et al., 1994). In additon, mutations in Fas have been identified in humans with inherited autoimmune disorders accompanied by lymphadenopathy (Drappa et al., 1996; Fisher et al., 1995; Rieux-Laucat et al., 1995). It has been suggested that the role of Fas in lymphoid cell death may overlap with other apoptosis-iiiducing members of the TNF receptor family. In this regard, TNF-a! also medates AICD
18
BRAD H. NELSON AND DENNIS M . WILLERFORD
in T cells (Speiser et nl., 1996; Sytwu et al., 1996; Zheng et al., 1995), whereas a potential role for the newly identifed receptors for TRAIL require further definition (Schneider et nl., 1997).Thus, physiologic AICD may involve a complex set of interactions among receptors and ligands, each with potentially distinct patterns of regulation. There appear to be at least two control points for Fas-mediated cell death signals. Because Fas is expressed consitutively, AICD is regulated in part at the level of FasL expression (Suda et al., 1995). However, induction of FasL is not itself sufficient to induce apoptosis efficiently, as upregulation of FasL occurs rapidly following TCR stimulation, whereas maximal sensitivity to AICD requires repetitive stimulation (Boehme and Lenardo, 1993;Wonget nl., 1997). Moreover, resting T cells are resistant to Fas stimulation when coculturedwith cells expressing FasL and undergoing AICD (Hornung et nl., 1997). Thus, a second control point for AICD is in the signaling machinery downstream of the Fas receptor. In T cells, an inhibitor of caspase 8 activation, which goes by many names, including cFLIP, is expressed contstitutively in the resting state and may contribute to the resistance of these cells to Fas-mediated apoptosis (Irmler et al., 1997). Regulation of cFLIP and other proteins interacting with the death signaling machinery may consitute this second control mechanism. One connection that has emerged is with the protooncogene c-myc, expression of which is associated with both cellular proliferation and apoptosis. Cell death induced by c-myc appears to depend on Fas/FasL interaction and correlates with increased cellular sensitivity to Fas signaling (Huber et al., 1997). The straightforward view of IL-2 as a T-cell growth factor was complicated by the observations of Lenardo (1991j, who found that T-cell clones cultured with exogenous IL-2 exhibited a high degree of apoptosis following restiinulation through the antigen receptor. IL-2 also promotes AICD in primary T cells that are first activated by concanavalin A in order to induce high-affinity IL-2 receptors, then exposed to IL-2 for 2 days prior to restimulation by CD3 cross-linkmg (Lenardo, 1991).The major correlate of IL-2-induced susceptibility to AICD is movement through the S phase of the cell cycle. Thus, cell death after restiinulation of T-cell clones correlates with the degree of T-cell proliferation, as measured by the incorporation of [3H]thymidine,and is inhibited when cells are blocked at the G1 phase of the cell cycle, but not when blocked in S phase (Boehme and Lenardo, 1993). The importance of cell cycle effects in explaining the IL-2-mediated susceptibility to AICD is underscored by the fact that other lymphokines that promote T-cell proliferation, including IL-4, IL-7, and IL-15, also promote AICD (Boehme and Lenardo, 1993; Van Parijs et nl., 1997). However, data of Russell and colleagues (Wang et al., 1996) suggest
BIOLOGY OF TIHE INTEHLKUKIN-2 RECEPTOR
19
that IL-2 may be more effective than other cytokines in this regard. Thus, a major consequence of the cell cycle progression signals delivered by the IL-BR is that T cells are primed for apoptosis following repeated or continuous TCR stimulation. Therefore, an important question regarding IL-2R signaling is whether susceptibility to AICD is mediated entirely by the signals that promote cell cycle progression or also includes more direct influences on cell death pathways. The requirement of IL-2R signals for the negative regulation of the immune system suggests that sensitizing cells to AICD may be an important function for these signals in uiuo (see Section VIII). V. Mechanism of 11-2 Receptor Activation
A. A GENERAL MODELFOR GIiownr FACTOH RECEPTOR ACTIVATION Receptors for soluble growth factors typically consist of an extracellular ligand binding domain connected via a short, hydrophobic membranespanning segment to a cytoplasmic domain that mediates the activation of intracellular signaling pathways. In the case of receptor tyrosine kinases (RTK, e.g., EGFR, PDGFR, insulin receptor), which have intrinsic tyrosine kinase domains within their cytoplasmic regions, signaling is initiated by ligand-induced oligomerization of the extracellular domains of receptor subunits (reviewed in Ullrich and Schlessinger, 1990). Because receptor diffusion is essentially limited to two dimensions by residence in the membrane, this clustering brings the cytoplasmic domains of the receptor components into close proximity. In the case of RTKs, low-level constitutive kinase activity permits cross-phosphoiylation of the clustered cytoplasmic domains at the site of regulatory tyrosines, resulting in the upregulation of kinase activity (Ullrich and Schlessinger, 1990). Aggregation of the extracellular domains may occur by the interaction of two monomeric receptor subunits with dimeric ligand or by the binding of a single divalent ligand to two receptor subunits, which are the either the same (homooligomerization) or different (heterooligomerization). Alternatively, the ligandreceptor interaction may be monomeric but induce a conformational change in the extracelluar domain that favors receptor oligomerization. Examples of each of these mechanisms have been described (Ullrich and Schlessinger, 1990). Although differing from RTKs by a lack of intrinsic kinase domains, members of the heinatopoietin receptor superfamily are similarly activated by ligand-induced oligomerization, as illustrated by the crystal structure of the growth hormone extracelluar domain complexed with ligand (Cunningliam et al., 1991; deVos et al., 1992). One growth hormone molecule binds to two identical receptor chains, with two dstinct facets of the ligand
20
BRAD H. NELSON AND DENNIS M. WILLERFORD
contacting the same binding pocket on the receptor molecules. Stabilization of the receptor dimer by interaction with ligand is supplemented by contacts between the two extracellular domains C-terminal to the ligand binding site. The concept of ligand-induced dimerization as a mechanism for cytokine receptor activation extends to include heteromeric receptor complexes. For example, the IL-6 receptor consists of two chains, IL-6Ra and gp130. The a chain is required for binding of IL-6, but not for receptor signaling (Hibi et al., 1990). The binding of IL-6 to IL-6Ra leads to formation of a complex with gp130. These IL-6:IL-GRa:gp130 complexes dimerize, which brings two gp130 intracytoplasmic domains into close proximity, leading to receptor activation (Murakami et al., 1993).Therefore, the IL-6 receptor resembles the growth hormone receptor in that it is activated through ligand-induced homodimerization, but differs in its 1igand:receptor stoichiometry and by its requirement for an additional receptor subunit for ligand binding.
B. ACTIVATION OF THE IL-2 RECEPTOR Prior to the cloning of the yechain, the known components of the IL2R superficially resembled the IL-6 receptor, consisting of a ligand-binding subunit (IL-2Ra) and an essential signaling subunit (IL-2RP) with sequence homology to gp130. To test whether the IL-2R might signal through homodimerization of IL-BRP, a chimeric receptor chain was constructed to contain the transinembrane and cytoplasmic domains of IL-ZRP fused to the extracellular domain of c-kit, a receptor tyrosine kinase that homodimerizes on binding the divalent ligand stem cell factor. Ligation of this chimeric htIIL-2RP chain was sufficient for proliferative signaling in the pro-B-cell line BAF3 (Nelson et al., 1994). However, this result did not reflect a complete IL-2R signal, as induction of the IL-2Ra gene did not occur. Furthermore, when introduced into T cells, this chimeric receptor was not sufficient for proliferative signaling (Nelson et al., 1994).With the cloning of ye(Takeshita et al., 1992a),it seemed likely that the cytoplasmic domain of this receptor chain might also be required for signaling. Two groups subsequently used a chimeric receptor strategy to investigate the role of ycin IL-2R activation (Nakamura et al., 1994; Nelson et al., 1994). In both studies, dimerization of the IL-2RP and yecytoplasmic domains was accomplished by attaching these regions to the extracellular domains of heterologous receptors. In one case, the extracellular domains were derived from either c-kit or the granulocyte/macrophage colony-stimulating factor (GM-CSF) receptor, and dimerization of the chimeric chains was induced by the addition of stem cell factor or GM-CSF, respectively (Nelson et al., 1994). In the other case, the chimeric receptor chains were constructed using the extracellular domain of IL-2Ra, and dimerization
BIOLOGY OF THE INTERLEUKIK-2 RECEPTOR
21
was induced by the addition of anti-IL-2Ra antibodies (Nakaniura et al., 1994). In both studies, the dinierizing agents (i.e., stern cell factor, GMCSF, or anti-IL-2Ra antibody) were able to induce proliferative signals, provided both IL-2RP- and y,-derived chimeric chains were expressed. Therefore, it was concluded that ligand-induced heterodimerization of the cytoplasmic domains of IL-2RP and -yc is both necessary and sufficient for IL-2R-mediated proliferative signaling. Thus, the growth hormone receptor, the IL-6 receptor, and the IL-2 receptor provide three distinct examples of how ligand-induced oligomerizationcan lead to receptor activation within the hematopoietic receptor superfamily. Moreover, these studies underscore the mechanistic parallels between signaling by these receptors and members of the RTK family. VI. lntracellular Signaling by the 11-2 Receptor
One of the most intriguing aspects of hematopoietic cytokine receptor signaling is that the simple act of receptor chain dimerization is sufficient to initiate the multitude of intracellular events underlying the proliferative and differentiative responses of blood cells (Heldin, 1995). In the case of the interleukin-2 receptor, lieterodiinerization of the cytoplasmic domains of the P and ycchains can trigger such diverse physiological responses in T cells as proliferation, activation of effector function, and sensitization to death signals. Like other members of the heinatopoietin receptor superfamily, the IL-2R has no intrinsic catalytic function but instead signals through receptor-associated kinases, in particular, the Janus tyrosine kinases ( Jaks). Receptor ligation induces the catalytic activation of Jaks and other kinases, which initiate a cascade of intracellular phosphorylation events and the formation of a inultiineric signaling complex at the inner surface of the cell membrane. A number of downstream effector molecules are recruited to the IL-2R signaling complex, including the transcription factors Stat3 and Stat5, the adaptor protein Shc, and the lipid kinase PI3 kinase, each of which transmits a unique signal to target genes in the nucleus. Although much progress has been made in defining the biochemical components of several IL-2R signaling pathways, a current challenge is to link these biochemical events to the regulation of specific genes and cellular responses in lymphocytes. This section summarizes current knowledge of IL-2R signaling by first describing the signaling domains of the receptor chains themselves, followed by a brief account of some of the key molecules and pathways that are activated by the receptor. Finally, the molecular pathways that regulate cell proliferation and viability in response to IL-2 are discussed. This is not intended as an exhaustive review of IL-2R signaling mechanisms, which were covered extensively by a review article in this
22
BRAD 13. NELSON A N D D E N N I S M . WILLEKFOKU
series (Karnitz and Abraham, 1996), but instead is an attempt to highlight recent discoveries in the field particularily as they pertain to in vivo functions of the IL-2R. A. CYTOPLASMIC SUBDOMAINS OF THE P
AND
ycSUBUNITS
Following oligornerization by IL-2 binding, a multirneric signaling coinplex forms rapidly around the cytoplasmic domains of IL-2RP and yc. Formation of this complex is dependent on the catalytic activity of intracelMar kinases, particularly the Jaks, which phosphorylate tyrosine residues on both IL-2RP and yc(Asao et al., 1990; Sugamura et al., 1990) (Fig. 1). These phosphotyrosine motifs form docking sites for cytoplasmic-signaling proteins that have SH2 or phosphotyrosine-binding (PTB) domains (Kavanaugh and Williams, 1994; Pawson, 1995). Once recruited to the activated receptor, these proteins become substrates for kinases, resulting in their activation and further propagation of the intracellular signal. Therefore, to understand how the diverse cellular responses to IL-2R signaling are
FIG.1. Signaling domains of IL8R. (A) The three chains of IL-2R, with emphasis on the cytoplasmic subdomains of IL-2Rfi and y c . Box 1 and box 2 motifs on IL-2RP and yc are shown as black boxes in the membrane-proximal region. Tyrosine residues are designated “Y” and are numbered according to the human receptor sequences (Hatakeyama et nZ., 198913; Takeshita et nZ., 1992a). Also shown are the A and H regions of IL-2RP and the PROX region of (B) Tyrosine kinases associated with IL-2RP and yc chains. Jakl and Jak3 bind to box 1 and box 2 motifs in the membrane-proximal regions of IL-2RP and ye, respectively. In addition, Syk associates with the membrane-proximal region of IL-2RP, whereas the Src family kinases Lck, F p , and Lyn associate with the A region.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
23
activated, a detailed account is required of the interactions between the cytoplasmic domains of IL-2RP and yoand the signaling proteins that are brought into the receptor complex. The cytoplasmic domain of the IL-BRP subunit was initially divided into three regions by Taniguchi and colleagues on the basis of amino acid sequence and restriction enzyme sites (Hatakeyama et al., 1989a). Despite their somewhat arbitrary origm, these regions have since proven to perform discrete signaling functions and therefore remain convenient designations for describing the signaling properties of IL-BRP (Fig. 1).The membraneproximal region of IL-2RP contains box 1 and box 2 motifs common to most hematopoietic cytokine receptors. These motifs generally constitute the binding sites for Janus tyrosine kinases, specifically Jakl in the case of IL-2RP (Boussiotis et al., 1994; Miyazaki et al., 1994; Russell et al., 1994; Tanaka et al., 1994). In addition, the tyrosine kinase Syk has been reported to associate with the membrane-proximal region of IL-2RP (Minami et al., 1995; Qin et al., 1994). The region encompassing box 2 was initially dubbed the S region due to the presence of a large number of serine residues; deletion of this region abrogates the activation of Jakl and Jak3 and all known signaling events downstream of the IL-2R (Hatakeyama et al., 1989a; Merida et al., 1993; Minami et al., 1995; Miyazaki et ul., 1995; Satoh et al., 1992; Shibuya et al., 1992; Witthuhn et al., 1994).The middle portion of IL-2RP, extending from amino acids 313 to 382, is rich in acidic residues, hence the name “A region.” This region of human IL2RP contains four tyrosine residues (Y338, Y355, Y358, and Y361) and contains binding sites for src-family tyrosine kinases (Lck, Fyn, and Lyn) (Hatakeyama et nl., 1991; Kobayashi et al., 1993; Minami et al., 1993), the adaptor molecule Shc (Evans et al., 1995; Friedmann et al., 1996; Ravichandran and Burakoff, 1994; Ravichandran et al., 1995, 1996), and, in some cells, Stat transcription factors (Gaffen et al., 1996) ( Lord et al., 1998).This region can be deleted from IL-2RP without compromising Jak activation or proliferative signaling in lymphoid cell lines, provided that the H region is intact (Hatakeyama et al., 1989a; Witthuhn et al., 1994). The distal portion of IL-2RP was named the H region because it constitutes one-“half” of the cytoplasmic domain (T. Miyazaki, personal communication). It contains two tyrosine residues (Y392 and Y510) that are involved in the activation of Stat transcription factors (Friedmann et al., 1996; Fujii et al., 1995; Gaffen et al., 1995, 1996). Like the A region, the H region can be deleted from IL-2RP without compromising Jak activation or proliferative signaling (Hatakeyama et al., 1989a; Lord et a[., 1998). However, as described in detail in Section C, deletion of both the A and the H regions abrogates the proliferative response, suggesting that these cytoplasmic
24
BRAD H. NELSON A N D DENNIS M . WILLEHFORD
domains are involved in the activation of parallel and partially redundant signaling pathways that regulate cell growth and division. As with IL-ZRP, the cytoplasmic domain of the ycchain has been divided into three regions based on amino acid sequence (Fig. 1).When yc was first cloned, Sugamura and colleagues noted sequence homology between the membrane-proximal region of yc and the SH2 motif found in many signaling proteins (Takeshita et al., 1992a). However, the SH2-like region of yc is short and lacks most of the residues required for binding to phosphotyrosine (Koch et al., 1991). Thus, this region of yc is unlikely to function as an SH2 domain and will be referred to simply as the membraneproximal or “PROX” region (extending from residues 284 to 321). This region contains a box 1 motif that is essential for association with, and activation of, the tyrosine kinase Jak3, as well as for IL-2R-induced cell proliferation (Asao et al., 1993, 1994; Boussiotis et al., 1994; Goldsmith et al., 1995; Miyazaki et al., 1994; Nelson et al., 1996, 1997; Russell et al., 1994).The PROX region may also play a Jak3-independent role in signaling, as it shares homology with the membrane-proximal regions of the (Y subunits of the IL-3, IL-5, and GM-CSF receptors, none of which appear to bind Jak molecules, yet are nevertheless required for receptor activation (Cornelis et al., 1995; Poloskaya et al., 1994; Quelle et al., 1994; Takaki et al., 1994). Just C-terminal to the PROX region lies 14 residues with weak homology to the box 2 motif found in many hematopoietic cytokine receptors (residues 322 to 335). This region performs a function common to other box 2 motifs, as it is essential for the association and activation of Jak3 (Asao et al., 1994; Miyazaki et al., 1994; Nelson et al., 1996; Russell et al., 1994). In fact, the PROX and box 2 regions of yc are necessary and sufficient for Jak3 activation and proliferative signaling (Goldsmith et al., 1995; Nelson, et al., 1996).The remaining C-terminal region of yc(residues 336 to 369) is highly conserved across species (Cao et al., 1993), but is not homologous to other hematopoietic cytokine receptors. The function of this region remains obscure, as it can be deleted without any apparent effect on Jak3 activation or proliferative signaling (Asao et al., 1994; Goldsmith et al., 1995; Nelson, et al., 1996). B. SIGNALING PATHWAYS DOWNSTREAM OF THE IL-2R 1. The Janus Tyrosine Kinases Jakl and Jak3 It has long been recognized that stimulation of hematopoietic cytokine receptors induces the rapid tyrosine phosphorylation of multiple cytoplasmic substrates, despite the fact that these receptors lack intrinsic kinase activity. The general mechanism by which this occurs was resolved with the discovery that members of the Janus tyrosine kinase (or Jak) family
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
25
associate with the cytoplasmic domains of hematopoietic cytokine receptors and are activated rapidly in response to receptor oligomerization. There are four known Janus kinases in mammals: Jakl, Jak2, Jak3, and Tyk2 (reviewed in Ihle, 1996a). In the case of the IL-2R, the catalyhc activation of two Jailus kinases, Jakl and Jak3, is induced within minutes of IL-2 binding (Asao et al., 1994; Brunn et al., 1995; Johnston et al., 1994, 1995a; Musso et al., 1995; Tanaka et al., 1994; Tortolani et al., 1995; Wakao et al., 1995; Witthuhn et al., 1994). By coprecipitation experiments, Jakl has been shown to associate constitutively with the IL-2RP chain and Jak3 with the yc chain (Boussiotis et al., 1994; Miyazaki et al., 1994; Russell et al., 1994; Tanaka et al., 1994) (Fig. 1).When the IL-2R is ligated, JakS appears to also associate with IL-gRP, perhaps by transfer from 7‘. In both cases, the Jak molecule interacts with the membrane-proximal cytoplasmic region of the receptor chain, specifically the box 1 and box 2 motifs (Asao et al., 1994; Higuchi et al., 1996; Miyazaki et al., 1994; Nelson, et al., 1996; Russell et al., 1994; Witthuhn et al., 1994). Conversely, the N-terminal region of Jak3 appears to mediate association with yc (Chen et al., 1997a; Kawahara et al., 1995), and a similar model may apply to the association of Jakl with IL-2RP. Catalytic activation of Jakl and Jak3 upon ligation of the IL-2R is associated with the phosphorylation of multiple tyrosine residues that are thought to perform a regulatory function. The conventional model for this process is that receptor dimerization brings the associated Jak molecules in close proximity, which promotes cross-phosphorylation and transactivation of the kinases (Darnell, 1997; Ihle, 1995; Ullrich and Schlessinger, 1990). This model is supported by the finding that Jakl fails to become activated in the absence of Jak3 catalytic activity (Kawahara et al., 1995; Nelson, et al., 1996; Oakes et al., 1996; B. H. Nelson, unpublished observations). However, the converse does not appear to be true, as Sugmura and colleagues have introduced point mutations into the box 1 motif of IL-2RP (P257S and P259S) that abrogate the ability of the receptor chain to bind or activate Jakl; surprisingly, these mutant IL-2RP chains can still interact with ycto promote normal activation of Jak3 in response to ligand (Higuchi et al., 1996). The mechanism of Jak activation by the IL-2R was also addressed in a study in which the cytoplasmic domain of yc was replaced by a covalently attached Jak3polypeptide on the assumption that this would allow ligand-induced interaction of Jak3 with Jakl and other components on the IL-2RP chain (Nelson et al., 1997). Contrary to the result expected under the “proximity” model, the attached Jak3 molecule failed to become activated in response to ligand unless the PROX regon of yc was retained as part of the receptor chain. This requirement for the PROX region might reflect a structural role in orienting Jak3 with respect to the IL-BRP
26
BRAD H. NELSON AND DENNIS M. WILLERFORD
chain. Alternatively, the PROX region could facilitate activation of Jak3
by generating an additional activation signal. Indeed, it was found that the PROX region alone, although insufficient to activate Jakl or Jak3, nonetheless induced low-level tyrosine phosphorylation of IL-2RP and the associated tyrosine phosphatase SHP-2 in response to ligand. Furthermore, upon stimulation of the wild-type IL-2R in T cells, tyrosine phosphorylation of IL-2RP was shown to precede the activation of Jak3. Taken together, these data suggest an alternate model for early activation of IL-2R signaling, in which the PROX region of yc interacts with another as yet unidentified tyrosine kinase. Upon receptor dimerization, this second kinase would induce the phosphorylation of Jak3 as well as the IL-2RP chain. Once activated, Jak3 may then induce phosphorylation and activation of Jakl (Nelson et al., 1997). It appears that Jakl is dispensible for the mitogenic response to IL-2R signaling, as mutations in IL-2RD that abrogate Jakl binding can still mediate robust proliferation in serum-starved MOLT-4 cells. These mutant IL-BRP chains appear to interact normally with ycto induce the activation of Jak3 and Stat5 (Higuchi et al., 1996). One note of caution, however, is that these experiments were performed in a transformed T-cell line that can grow independent of IL-2 unless deprived of serum. Therefore, it remains possible that Jakl is required for proliferative signaling in normal, nontransformed lymphocytes. In contrast, it is clear that the catalytx activation of Jak3 is a critical event in IL-2R signaling. Taniguchi and colleagues overexpressed a kinase-deficient form of Jak3 in the IL-3-dependent proB cell line BAF3 and evaluated the effect on signaling through a reconstituted IL-2R complex (Kawaharaet al., 1995).This dominant-interfering form of Jak3 abrogated the activation of Jakl and the induction of c-fos in these cells and markedly inhibited cellular proliferation and the induction of c-myc. Intriguingly, induction of the antiapoptotic gene bcl-2 occurred normally in response to IL-2. This latter event was blocked by overexpression of a truncated yc chain lacking the cytoplasmic domain, suggesting that yc induces expression of bcl-2 by a Jak3-independent pathway. In addition to these experiments, studies of IL-2R signaling in a human B-cell line lacking Jak3 have shown that Jak3 is required for tyrosine phosphorylation of IL-2RP, Jakl, and Stat5 (Oakes et al., 1996). Other studies using site-directed mutants of the cytoplasmic domain of yc have demonstrated a strict correlation between the ability of mutants to activate Jak3 and their ability to induce multiple downstream events, including activation of Jakl, expression of c-myc and c-fos, and cell proliferation (Nelson, et al., 1996). Conversely, mutants of yc that retained the ability to activate Jak3 remained competent to induce these events. Regarding the induction of the c-myc and c-fos genes, there appear to be differences
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
27
between the results obtained in nontransformed T-cell lines and the results of other studies using the fibroblast cell line L929 or the transformed Tcell line ED40515(-). In the latter studies, a truncation mutant of ye (EDE30-4) that was competent to induce catalytic activation of Jak3 and expression of c-myc failed to induce the c-fos gene (Asao et al., 1993, 1994), suggesting that the C terminus of yc provides a signal for c-fos induction that is required in addition to the activation of Jak3. The discrepancy between these and the former studies (Nelson et al., 1996) has not been resolved. The in vivo substrates of Jakl and Jak3 have not been definitively identified; however, several proteins demonstrate diminished tyrosine phosphorylation in cells in which the catalytic activity of Jak3 has been disrupted. In the case of IL-2R signaling, these include Jakl, the IL-BRP chain, the adaptor molecule Shc, and the transcription factor Stat5 (Kawahara et al., 1995) (Oakes et al., 1996) (B. H. Nelson, unpublished results). The only enzymes reported to undergo normal tyrosine phosphorylation or activation in the absence of Jak3 catalytic activity are the tyrosine kinase Lck (Gonzdez-Garcia et al., 1997) and the tyrosine phosphatase SHP-2 (Adachi et al., 1997). Thus, Jak3 appears to play a major role in the proximal activation of the IL-2R-signaling complex. Other cytokine receptors that utilize the yc chain, including the IL-4, IL-7, IL-9, and IL-15 receptors, also activate both Jakl and Jak3, which raises the issue of how the specificity of intracellular signals is achieved. One potential mechanism is at the level of the substrates of Jakl and Jak3 that are recruited to the receptor complex through binding interactions with either the receptor chains themselves or with various adaptor molecules in the complex. Because the individual receptor chains differ in the bindmg sites contained within their cytoplasmic domains, each may present a distinct set of substrates to Jakl and Jak3, thereby imparting unique biochemical characteristics to cytokine signals. It should also be noted that cytokine receptors that share yc mediate a number of overlapping functions in lymphocytes, particularly with respect to cell proliferation. Accordingly, Jakl and Jak3 appear to also mediate a common set of biochemical events for this group of receptors. 2. Stat Transcription Factors Jak kinase activity is generally associated with the tyrosine phosphorylation and activation of one or more Stat proteins, a family of molecules that was first discovered through studies of interferon signaling (reviewed in Darnell, 1997). Stat molecules are latent transcription factors that, in unstimulated cells, are localized to the cytoplasm. Cytohne stimulation induces tyrosine phosphorylation of receptor chains, thereby creating docking sites for Stats, which have an SH2 domain near the C terminus. The
28
BRAD H. NELSON AND DENNIS M. WILLERFORD
recruited Stats undergo rapid tyrosine phosphorylation, which induces them to form homo- or heterotypic dimers that subsequently translocate to the nucleus, where they bind to characteristic DNA sequence motifs within specific target genes. The IL-2R activates both the A and the B isoforms of Stat5 and, in several cell types, including primary T cells, can also activate Stat3 (Brunn et al., 1995; Frank et al., 1995; Fujii et nl., 1995; Gaffen et al., 1995; Gilmour et al., 1995; Hou et al., 1995; Johnston et al., 1995a; Lin et al., 1995; Pernis et al., 1995; Wakao et al., 1995). Stat activation by the IL-2R requires the tyrosine hnase activity of Jak3 (Fujii et al., 1995; Oakes et al., 1996; B. H. Nelson, unpublished results) but not Jakl (Higuchiet a!.,1996).In addition, Stat activation requires the presence of either Y392 or Y510 on the IL-SRP chain (Friedmann et al., 1996; Fujii et al., 1995), although in the T-cell lines CTLL2 and HT-2 Stats can also be activated via Y338 of IL-2RP (Gaffen et al., 1996; Lord et al., 1998) Site-directed mutants of IL-2RP that lack all cytoplasmic tyrosine residues fail to activate Stat factors, despite normal activation of Jakl and Jak3 (Friedmann et al., 1996; Gaffen et al., 1996; Lord et al., 1998). Thus, as is the case with other cytokine receptors, the IL-2R appears to induce nuclear signals through Stat proteins by the following series of events: (1) Jak-dependent phosphorylation of one or more tyrosine residues on the receptor chain, (2) recruitment of Stats through their SH2 domains to phosphorylated sites on the receptor, (3) tyrosine phosphorylation of the recruited Stat molecules, either by Jak3 or another receptor-associated kinase, (4) homo- or heterodimerization of Stat molecules via reciprocal interactions between their SH2 domains, (5) translocation of Stat dimers to the nucleus, and (6) binding to specific regulatory sequences within target genes. In addition to tyrosine phosphorylation, Stat factors also require serine phosphorylation for optimal transcriptional activation (reviewed in Darnell, 1997). With the IL-2R, this is carried out by a serine/ threonine kinase that has not been identified but is distinct from p42/44 MAP kinase, mTOR (the mammalian target of rapamycin), or p70 S6 kinase (Beadling et al., 1996; Kirken et al., 1997). Among the target genes of IL-2 that are Stat5 dependent are (1)cis (Matsumoto et al., 1997), a member of a novel family of proteins that are negative regulators of signaling by other cytokine receptors (Endo et al., 1997; Naka et al., 1997; Starr et al., 1997; Yoshimura et al., 1995);although neither cis nor other members of this protein family have yet been shown to associate with the IL-2R complex or modulate IL-2R signaling, data from other receptors suggest that such interactions are likely to be described in the near future; (2) osm, which encodes the cytokine oncostatin M (Yoshimura et al., 1996); and ( 3 ) the IL-2Ra gene (see later). So far, no role
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
29
for Stat3 or Statrj has been defined in the regulation of growth-related genes such as c-myc, c-fos, bcl-2, or bcl-x in the context of the IL-2R. Of the just-mentioned Stat5-dependent genes, IL-2Ra has been the most extensively characterized in terms of its mechanism of regulation. As described in Section III,A, a transient wave of IL-2Ra expression is induced in T cells by TCR signals, whereas maximal and sustained IL-2Ra transcription requires additional stimulation with IL-2 (reviewed in Nabholz et al., 1995). Experiments with transgenic mice harboring a reporter gene containing the 5’ flanking region of the IL-2Ra gene, together with DNase hypersensitivityanalyses, led to the identification of a 78 nucleotide element that is responsive to the IL-2 signal and is located 1.3 kb upstream of the major transcription start site (Soldaini et nl., 1995). This segment, which has since been named PRRIII (for positive regulatory region 111), was found to contain two potential binding sites for Stat proteins in addition to potential binding sites for Ets and GATA proteins (Sperisen et al., 1995). A homologous segment was subsequently identified in the human ZL-2Ra gene -4 kb upstream of the transcription start site (John et al., 1996; L’ecine et al., 1996). PRRIII of the human gene has been shown to bind both Stat5A and Stat5B, apparently via the more distal of the two potential Stat-binding sites (Lecine et al., 1996). Additional transcription factors, including Elf-1, GATA-1, and HMG-I(Y), also bind to sites in PRRIII but, in contrast to Stat5A and B, may not require IL-2 stimulation for their activation or expression (John et nl., 1996; L‘ecine et al., 1996; Serdobova et al., 1997). Experiments with reporter genes have shown that none of the DNA-binding sites in PRRIII is sufficient for IL-2-inducible transcription, whereas all three sites are necessary for optimal activity (John et al., 1996; L’ecine et al., 1996; Serdobova et al., 1997). Thus, it appears that the three known DNA-binding sites within the PRRIII element interact cooperatively to induce transcription of the IL-2Ra gene in response to IL-2 and that Stat5 is a critical inducible factor that switches the element to an active state. The essential role of Stat5 in regulation of the IL-2Ra gene is highlighted by the fact that splenocytes from mice with a targeted disruption of the Stat5A gene demonstrate defective induction of IL-2Ra expression in response to IL-2 and consequently require high doses of IL-2 for proliferation (Nakajima et al., 1997a). Notably, however, there appears to be at least one other inducible activity required for induction of the IL-2Ra gene in response to IL-2, as the IL-3 receptor can also activate Stat5A and StatB yet fails to induce expression of IL-2Ra (Ascherman et al., 1997).It is not known whether this additional activity represents a posttranslational modification of one or more proteins already known to bind PRRIII or the involvement of additional regulatory factors.
30
BRAD H. NELSON AND DENNIS M. WILLERFORD
Although the physiological role of Stat factors in IL-2R signaling is not fully defined, the just-mentioned list of target genes suggests that Statmediated signals may be involved in both positive and negative regulation of IL-2R activity via induction of the ZL-2Ra and cis genes, respectively. The requirement for Stat5 for expression of the cytokine oncostatin M raises the additional possibility that Stats may be involved in immunomodulatory aspects of IL-2R signaling. Finally, the essential role of Stat4 and Stat6 in the differentiation of CD4' T cells to the T h l and Th2 subclasses, respectively (Kaplan et al., 1996a,b; Shimoda et al., 1996; Takeda et al., 1996a; Thierfelder et al., 1996),suggests that Stat5 may also play a role in lymphoid differentiation. In this regard, Stat5 might be expected to function in the development of NK cells and T C R y 8 T cells, as these lymphoid classes are severely reduced in IL-2RP-l-mice (Suzuki et al., 1997a) (see Section VII). 3. MAP Kinase Pathways The IL-2R activates Ras and downstream components of the MAP kinase pathway, including Raf, MEK, and p42/44 MAP kinase (Duronio et al., 1992; Graves et al., 1992; Izquirdo et al., 1992; Karnitz et al., 1995; Perkins and Collins, 1993; Satoh et al., 1991; Turner et al., 1991; Welham et al., 1994a,b),which are associated with mitogenic signaling by a wide variety of growth factor receptors (reviewed in Marais and Marshall, 1996; Marshall, 1994).Activation of this pathway involves recruitment of the adaptor molecules Shc and Grb2 and the guanine nucleotide exchange factor mSOS to the IL-2R complex (Burns et al., 1993; Cutler et al., 1993; Karnitz et al., 1995; Liu et al., 1994; Ravichandran and Burakoff, 1994; Zhu et aE., 1994). Downstream of this pathway lies the protooncogene cfos, which is induced through the semm-response element in the promoter-proximal region, as well as fra-1, c-jun, and junB (Hatakeyama et al., 1992; Shibuya et al., 1992). Much has been learned about the components of the IL-ZR complex required for activation of the Ras/MAP kinase pathway in lymphocytes. First, the catalytic activity of Jak3 is essential, as tyrosine phosphorylation of Shc and induction of cfos fail to occur with a kinase-deficient form of Jak3 (Kawahara et al., 1995) (B. H. Nelson, unpublished results). Second, the A region of IL-2RP is required, as deletion of this region has been shown to abrogate the phosphorylation or activation of Shc, Ras, and p42/44 MAP kinase and the induction of cfos, c-jun, and other related protooncogenes (Evans et al., 1995; Hatakeyama et al., 1992; Ravichandran et al., 1996; Satoh et al., 1992; Shibuya et al., 1992). Third, within the A region, Y338 is required, as point mutation of this residue to phenylalanine abrogates the phosphorylation of Shc and p42/44 MAP kinase and the induction of c-fos (Evans et al., 1995; Friedmann et al., 1996; Gaffen et al., 1996) (Lord et al., 1998). Fourth, recruitment and tyrosine phosphoryla-
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
31
tion of Shc by the IL-2R appear sufficient to initiate the activation of this pathway. This was shown by replacing the entire A and H regions of IL2RP with a covalently linked Shc molecule (Lord et nl., 1998). Specifically, the N terminus of Shc was attached just C-terminal to the box 1 and box 2 regions of IL-2RP. As expected, the resulting IL-2RPIShc fusion protein mediated normal activation of Jakl and Jak3 and underwent tyrosine phosphorylation in response to ligand. More importantly, the fusion protein mediated normal phosphorylation of ~ 4 9 4 4MAP kinase and induction of c-fos in CTLLB and BAF3 cells, despite lacking Y338 and all other residues in the A region. Collectively, these and other results suggest a linear sequence of events by which the IL-2R induces activation of the Ras/MAP kinase pathway: (1)receptor dirnerization activates Jakl and Jak3, (2)Y338 of IL-2RP is phosphorylated, ( 3 ) Shc binds to this site through its PTB domain (Friedmann et aE., 1996; Ravichandran et al., 1996), (4) Shc undergoes tyrosine phosphorylation, ( 5 ) Grb2 binds to phosphorylated tyrosine residues in the CH domain of Shc (Harmer and DeFranco, 1997; Ravichandran and Burakoff, 1994; Zhu et al., 1994), and (6) mSOS is recruited to the receptor complex (Ravichandran and Burakoff, 1994; Ravichandran et al., 1995) and induces the activation of Ras. Subsequent events downstream of Ras leading to activation of p42/44 MAP kinase and the induction of c-fos are thought to occur by the standard Ras/Raf/MEK/MAP kinase pathway defined for other growth factor receptors. PI3 kinase also appears to be involved in IL-2R-mediated MAP kinase activation, as wortmannin (a pharmacologic inhibitor of PI3 kinase) significantly lminishes the catalpc activity of both MEK and MAP kinase in IL-2-stimulated T cells without affecting Raf activity (Karnitz et al., 1995). Although much is known about the mechanism by which IL-2R activates the Ras/MAP kinase pathway, the physiological function of this pathway in the cellular response to IL-2 remains undefined. Importantly, the A region of IL-2RP, although absolutely required for activation of the Ras/ MAP kinase pathway and induction of c-fos, is redundant with the H region for IL-2R-induced proliferation and viability of lymphoid cell lines (see Section V1,C). It remains possible, however, that the dispensibility of the A region merely reflects the permissive growth properties of cultured cell lines and that the proliferative response of primary lymphocytes in vivo requires activation of the Ras/MAP hnase pathway by the IL-2R. In neuronal cells, activation of the p42/44 MAP kinase pathway has been associated with the delivery of cell survival signals (Xia et al., 1995), raising the additional possibility that the role of this pathway in the biologc response to IL-2 may include the regulation of lymphocyte viability. Furthermore, evidence shows that this pathway may affect lymphocyte differentiation, as overexpression of a dominant p42/44 MAP kinase mutation in the thymus
32
BRAD H. NELSON AND DENNIS M. WILLERFORD
leads to preferential differentiation of the CD4+ T-cell lineage at the expense of the CD8' lineage (Sharp et al., 1997). In addition to the conventional p42/44 MAP kinases, alternative MAP kinase pathways have been described that involve proteins of the c-jun amino-terminal kinase (JNK) and p38 kinase pathways. Generally asssociated with inflammatory cytokines or stress responses, JNK and p38 pathways are also activated by IL-2R signals (Crawley et al., 1997). It was reported that the activity of p38 MAP kinase is essential for IL-2R-mediated proliferation, as a pharmacological inhibitor of p38 was found to suppress T-cell proliferation in a dose-dependent manner (Crawley et al., 1997). However, this group has since found that the activation of p38 MAP kinase in BAF3 cells appears to require the A region of IL-BRP, which itself is dispensible for proliferation, therefore an essential role for p38 in the proliferative response seems unlikely (B. M. J. Foxwell, personal communication). In other receptor systems, these kinases have been asssociated with a number of cellular responses, including cellular activation, differentiation, and apoptosis (Fanger et nl., 1997; Kyriakis and Avruch, 1996; Verheij et at., 1996). Further investigation into the functional consequences of JNK and p38 activation by the IL-2R will thus be of great interest. 4. Other Components of the IL-2R Signaling Complex a. Src Family Kinases. The first tyrosine kinase shown to be activated by the IL-2R was the Src family member Lck. Lck was shown to physically associate with the acidic domain of IL-2RP through an unconventional interaction involving the kinase domain of Lck (Hatakeyama et al., 1991) (Fig. 1).Moreover, IL-2 was found to induce the catalyhc activity of Lck within minutes of stimulation, an event that is dependent on the presence of the A region (Hatakeyama et al., 1991; Horak et al., 1991; Kim, et al., 1993; Minami et al., 1993). In cells lacking Lck, the related kinases Lyn and Fyn have been shown to undergo catalytic activation in response to IL-2 (Kobayashi et al., 1993; Torigoe et al., 1992), therefore the IL-2R appears to be somewhat permissive in its interactions with Src family kinases. Activation of Src family kinases by the IL-2R is associated with the catalytic activation of PI3 kinase (Taichman et al., 1993). In T-cell lysates, PI3 kinase activity coprecipitates with Fyn, and the amount of coprecipitating activity is enhanced severalfold by stimulation with IL-2 (Augustine et al., 1991). The interaction between these two molecules is mediated by the SH3 domain of Fyn and two proline-rich regions in the p85 subunit of PI3 kinase (Karnitz and Abraham, 1996; Karnitz, et al., 1994;Pleiman et al., 1994; Prasadet al., 1993).It should be noted, however, that Src family kinases are likely to represent only one of several means by which the IL-2R activates PI3 kinase. Given the fact that the acidic
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
33
region of IL-2RP is redundant for proliferative signaling in lymphocytes, the physiological role of Lck and Fyn in IL-2R signaling remains unclear. A correlation between IL-2R-mediated activation of Lck and the prolongation of T-cell viability has been described, which is discussed in Section D.
b. ZRS-1 and IRS-2. In addition to Shc and Grb2, two other adaptor proteins have been reported to participate in IL-2R signaling. Insulin receptor substrate 1 (IRS-1) and its homolog IRS-2 are large cytoplasmic proteins that were initially identified as downstream components of the insulin and IL-4 receptors, respectively (Sun et al., 1995; Wang et al., 1992; White and Jahn, 1994). These molecules contain multiple tyrosine residues that are phosphorylated in response to receptor stimulation and serve as docking sites for several cytoplasmic signal-transducing molecules containing SH2 domains, including Crk, Grb-2, Nck, PI3 kinase, and SHPTP2 (Gustafson et al., 1995; Myers et al., 1996; Skolnik et al., 1993; Sun et al., 1993). Consistent with their central role in recruiting substrates to activated receptor complexes, IRS-1 and IRS-2 are required for optimal cell proliferation in response to insulin and IL-4. IL-2 has been shown to induce the tyrosine phosphorylation of IRS-1 and IRS-2 in PHA-activated human peripheral blood T cells, as well as human NK cells and B cells (Johnston et al., 1995b). Moreover, IRS-1 was shown to physically associate with Jakl and Jak3, and IRS-2 with Jakl, in IL-%stimulated T cells. One functional consequence of IRS-1 phosphorylation in response to IL-2 may be activation of PI3 kinase, as the p85 regulatory subunit of PI3 kinase coprecipitates with IRS-1 after IL-2 stimulation (Johnston et al., 1995b). c. SHP-2. Two SH2 domain-containing tyrosine phosphatases, SHP-1 and SHP-2, have been implicated in signaling by hematopoietic cytokine receptors (reviewed in Ihle, 1996b). Tyrosine phosphorylation of SHP2, but not SHP-1, occurs in response to IL-2R activation; however, no corresponding change in the catalytic activity of this enzyme has been reported (Adachi et al., 1997; Nelson et aZ., 1997). Both the membraneproximal and A regions of IL-2RP are required for phosphorylation of SHP-2, whereas the catalytic activity of Jak3 is not (Adachi et al., 1997). Indeed, a truncation mutant of 'ye that fails to bind or activate Jak3 can nevertheless interact with IL-BRP to induce modest tyrosine phosphorylation of SHP-2 (Nelson et al., 1997). One clue to the physiological role of SHP-2 in IL-2R signaling is suggested by studies of receptor tyrosine kinases, where it may contribute to the activation of Ras by serving as an adaptor molecule (Bennett et al., 1994; Li et al., 1994; Noguchi et al., 1994).A second clue may come from studies of the IL-6 receptor subunit gp130, which is thought to promote cell proliferation in part through the
34
BRAD H. NELSON AND DENNIS M. WILLERFORD
recruitment of SHP-2 to a cytoplasmic tyrosine residue (Fukada et al., 1996). Notably, however, similar roles for SHP-2 in IL-2R signaling have yet to be established.
C. MITOGENICSIGNALING BY THE IL-2R Of all the cellular responses regulated by the IL-2R, the role of the receptor in controlling lymphocyte proliferation is by far the best characterized at the molecular level. Therefore, the following section summarizes current knowledge and hypotheses concerning the molecular mechanism of proliferative signaling by the IL-BR, and many of the key factors are illustrated in Fig. 2. Although some of the following signaling events have been mentioned in previous sections, they are included here as well to provide a comprehensive overview of the process of mitogenic signaling. 1. Role of Jak Molecules and Other Tyrosine Kinases Signals that promote cell cycle progression are initiated by IL-2-mediated heterodimerization of the IL-BRP and ycchains, which results in the rapid activation of several tyrosine kinases, including Jakl, Jak3, Syk, and Lck
FIG.2. A model for mitogenic signaling by IL-2R. Depicted are the components of the IL-2R complex known or hypothesized to be involved in transduction of the proliferative signal, including the cytoplasmic domains of IL-2RP and ye, the tyrosine kinase Jak3, the STAM protein, three tyrosine residues on IL-BRP, the adaptor molecule Shc, and the transcription factor Stat5 Also shown are examples of target genes involved in the regulation of cell proliferation and viability. Other factors involved in mitogenic signaling, such as PI3 kinase and mTOR, are not shown, as they have not yet been localized to specific pathways or sites in the receptor complex.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
35
or Fyn (Fig. 1).Of these, Jak3 has been shown to be essential for mitogenic signaling. Evidence supporting this conclusion includes: ( 1) fibroblasts expressing a reconstituted IL-2R complex will not proliferate in response to IL-2 unless Jak3 is coexpressed (Miyazah et al., 1994); (2)overexpression of a catalytically inactive version of Jak3 markedly inhibits IL-2R-mediated proliferation of the pro-B cell line BAF3, presumably by a dominantnegative mechanism (Kawahara et al., 1995); (3) deletion of either the S region of IL-2RP or the PROX or box 2 regions of yc abrogates the activation of Jak3 by IL-2 and similarily ablates the proliferative response (Asao et al., 1993, 1994; Goldsmith et al., 1995; Hatakeyama et al., 1989a; Mori et al., 1991; Nelson et al., 1996, 1997; Witthuhn et al., 1994); (4) T cells from Jak3-null mice fail to proliferate in response to IL-2 (Nosaka et al., 1995; Park et al., 1995; Thomis et al., 1995); and (5) signaling can be restored to nonfunctional truncation mutants of yr by covalent attachment of the Jak3 molecule (Nelson et at., 1997), but this requires that the Jak3 molecule have a functional catalytic domain (B. H. Nelson, unpublished results). In contrast to the critical role of Jak3, Jakl is apparently dispensible for proliferative signaling, as point mutants of IL-BRP that eliminate the binding and activation of Jakl do not affect proliferative response to IL-2R activation (Higuchi et al., 1996). Importantly, however, this result has yet to be confirmed in a nontransformed lymphocyte. As described earlier in Section VI,B,4, the Src family kinases Lck, Fyn and Lyn are also activated by the IL-2R, however, deletion of the A region of IL-SRP abrogates the binding and activation of these kinases without disrupting proliferative signaling (Hatakeyama et al., 1989a, 1991; Minami et at., 1993), suggesting that these kinases are either not involved in the mitogenic signal or are redundant with other factors. Sykis activated rapidly by the IL-2R and associates with the S region of IL-BRP, which itself is essential for mitogenic signaling (Minami et al., 1995; Qin et al., 1994). Moreover, when artificially clustered at the cell surface in the form of a chimeric CD16/Syk receptor, Syk can mediate induction of the c-myc gene in BAF3 cells (Minami et al., 1995).Nevertheless, Syk has not been directly demonstrated to be essential for IL-2-induced proliferation of lymphocytes, and in fact may not be, as it is expressed at very low levels in IL-2responsive peripheral T cells (Chan et al., 1994). Thus, of the tyrosine kinases known to be activated by the IL-2R, Jak3 appears to play the predominant role in mitogenic signaling (Fig. 2). This raises the critical issue of which factors downstream of Jak3 transmit the proliferative signal to the nucleus.
2. Role of Tyrosine Residues on the IL-2RP Chain There is strong evidence from mutational studies that tyrosine residues within the cytoplasmic region of the IL-2RP chain play an essential role in
36
BRAD H. NELSON AND DENNIS M. WILLERFORD
proliferative signaling downstream of Jak3. By mutating the six cytoplasmic tyrosine residues of the human IL-2RP chain to phenylalanine, Leonard and colleagues completely abrogated IL-2R-mediated proliferation of the myeloid cell line 32D (Friedmann et al., 1996). Add-back experiments in which one or more of these residues were retained as tyrosine revealed that at least one of three specific tyrosine residues (Y338, Y392, or Y510) is required for the mitogenic signal (Fig. 2). Intriguingly, any one of these three residues was sufficient for a proliferative response, whereas simultaneous elimination of all three tyrosines completely abrogated mitogenesis. Similar conclusions were drawn from studies in the T-cell lines HT-2 (Gaffen et al., 1996) and CTLL-2 (Lord et al., 1998). In the experiments with CTLL-2 cells, it was established that tyrosine residues on IL2RP are required for proliferation even in the face of normal Jak activity. The requirement for specific tyrosines on IL-2RP is in contrast to analogous experiments with the erythropoietin receptor, where mitogenesis can occur in the absence of receptor tyrosine residues, apparently through the activation of Jak2 alone (Joneja and Wojchowski, 1997; Nakamura et al., 1996). It also differs from studies of the yc chain, where cytoplasmic tyrosine residues have been found to be dispensible for proliferative signaling (Gaffen et al., 1996; Nelson et al., 1996). The apparent functional redundancy of Y338, Y392, and Y510 with respect to proliferative signaling is entirely consistent with studies involving large deletions of the IL-2RP chain. Taniguchi and colleagues showed that mitogenesis is largely unaffected by deletion of the acidic region of IL2RP (which contains Y338) or of the H region (which contains Y392 and Y510) (Hatakeyama et al., 1989a). However, if both the A and the H regions are deleted simultaneously, the proliferative response is abrogated, despite the retention of normal Jak activity (Lord et al., 1998). Thus, Jak3 catalytic activity is not sufficient for proliferative signaling by the IL-2R. Rather, mitogenic signaling proceeds from the catalytic activation of Jak3 to phosphorylation of one or more tyrosine residues in the A or H regions of IL-2RP. Studies in CTLL-2 and BAF3 cells have shown that the shared ability of the A and H regions of IL-2RP to promote cell proliferation also applies to the induction of certain protooncogenes. Thus, induction of the c-myc, bcl-2, and bcl-x genes proceeds normally, provided either the A or the H region of IL-2RP is present (Miyazaki et al., 1995; Shibuya et al., 1992) (J. D. Lord and B. H. Nelson, unpublished results). Further paralleling the proliferative response, induction of these genes is dependent on the presence of one or more tyrosine residues in these two regions of IL2RP ( J . D. Lord and B. H. Nelson, unpublished results). In contrast to these major protooncogenes, other gene induction events are fully or
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
37
partially dependent on either the A region (e.g., c-fos) or the H region as described in previous sections. (e.g.,Stat targets such as cis and IL-~RcY), These studies are consistent with a model in which the cytoplasmic region of IL-BRP, after being phosphorylated on specific tyrosine residues by Jak3, serves as a protein interaction domain to recruit cytoplasmic signaling molecules to the activated receptor complex, where, in turn, they are activated by receptor-associated kinases. Experimental observations suggest that different sections of the IL-2RP chain mediate interaction with distinct, but partially overlapping, sets of cytoplasmic signaling proteins, thereby giving rise to discernible patterns of downstream biochemical events.
3. Role of Factors Downstream of IL-2Rp Phosphorylated tyrosine residues on growth factor receptors generally serve as binding sites for cytoplasmic signaling molecules containing SH2 or PTB domains (Pawson, 1995), serving to recruit these proteins into the receptor complex where they may be activated, principally by phosphorylation. As described in the preceding section, three tyrosine residues on IL2RP (Y338, Y392, and Y510) have been implicated in transduction of the proliferative signal. Therefore, there is currently great interest in identifying the cytoplasmic factors that bind to these tyrosine-containing motifs and transmit the mitogenic signal. As illustrated in Fig. 2, two candidates for such factors are the adaptor molecule Shc, which binds to phosphorylated Y338 and is involved in activation of the RasIMAPK pathway (Burns et al., 1993; Cutler et al., 1993; Friedmann et al., 1996; Karnitz et al., 1995; Liu et al., 1994; Ravichandran and Burakoff, 1994; Zhu et al., 1994), and the transcription factor Stat5, which can be activated through Y392, Y510 (Friedmann et al., 1996; Fujii et al., 1995), or, in some cells, Y338 (Gaffen et nl., 1996) (Lord et al., 1998). The role of Shc in mitogenic signaling was investigated by linking the Shc protein covalently to the IL2RP chain in place of the A and H regions (Lord et al., 1998). In this experiment, the signals induced by Shc could be studied in isolation from other A or H region-dependent events. As expected, the IL-BRPI Shc fusion protein induced Jak activation in response to ligand and also demonstrated normal aspects of Shc function, including phosphorylation of p42/44 MAP kinase and induction of c-fos. Moreover, the the IL-2RPI Shc fusion protein was able to promote a robust proliferative response in both the T-cell line CTLL2 and the pro-B cell line BAF3, which corresponded with normal induction of the c-myc, bcl-2, and bcl-x genes. Thus, the proliferative function contributed by the A and H regions of the IL2RP can be replaced by a covalently associated Shc protein. Intriguingly, the IL-2RPIShc fusion protein was unable to support the long-term
38
BRAD H. NELSON AND DENNIS M. WILLERFORD
viability of transfected cells, despite the fact that induction of bcl-2 and bcl-x was normal. Thus, it appears that Shc is able to mediate some, but not all, of the signaling events normally induced through Y338, suggesting that Y338 may mediate cell expansion through both Shc and a second downstream molecule that has yet to be identified. If this is the case, it would imply that individual phosphotyrosine motifs on the cytoplasmic domain of the IL-2RP can interact with multiple downstream factors. The evidence implicating Stat5 in proliferative signaling is so far strictly correlative, in that Y392 and Y510 in the H region of IL-2RP can promote both Stat5 activation and cell proliferation (Friedmann et al., 1996; Fujii et al., 1995; Gaffen et al., 1996); thus, the possibility that a second, unidentified factor binds to Y392 and Y510 and is responsible for mitogenic signaling is not precluded. Indeed, if such a factor existed and were able to also interact with Y338, it would provide a simple explanation for the apparent functional redundancy of the A and H regions (and their respective tyrosine residues) with respect to proliferative signaling. To detect the activity of such a factor, it will be necessary to simultaneously abrogate the activity of Shc and Stat5 in cells while retaining one or more phosphorylated tyrosine residues on the IL-2RB chain. Although such an experiment has not been reported for the IL-2R, a mutant form of the EPO receptor has been described that can induce the proliferation of BAF3 cells independent of the Shc and Stat5 pathways (Klingmuller et al., 1997), indicating that alternative proliferative pathways do in fact exist in lymphocytes. At this time, all that can be concluded for the IL-2R is that the factors responsible for transmitting the mitogenic signal downstream of Y338, Y392, and Y510 of IL-2RP have yet to be definitively characterized; however, the available evidence suggests Shc and Stat5 as the best candidates. 4. Role of PI3 Kinase and Akt
Ligation of the IL-2R induces the catalytic activation of PI3 kinase, a lipid and serinelthreonine kinase that has been implicated in mitogenic signaling by a number of growth factor receptors (Merida, et al., 1991, 1993; Monfar et al., 1995; Reif et al., 1997; Remillard et al., 1991) (for reviews see Karnitz and Abraham, 1996; Vanhaesebroeck et al., 1997). Several biochemical events have been reported to lie downstream of PI3 kinase in the context of IL-2R signaling, including activation of the serine/ threonine kinases Akt and p70 S6 kinase (Ahmed et al., 1997; Karnitz et al., 1995; Karnitz and Abraham, 1996; Monfar et al., 1995; Reif et al., 1997) and the threoninehyrosine kinase MEK (Karnitz et al., 1995).Wortmannin,
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
39
which is a potent inhibitor of PI3 kinase [and, at high concentrations, of mTOR as well (Brunn et al., 1996)], has been shown to cause a modest reduction in IL-2-induced proliferation of the T-cell line CTLL2 (Karnitz et al., 1995). Gain-of-function experiments involving the PI3 kinase pathway have yet to provide a consistent picture of its role in lymphocyte proliferation. Overexpression of constitutively active forms of Akt in the pro-B-cell line BAF3 resulted in constitutive expression of the protooncogenes c-myc and bct-2 and inhibited cell cycle arrest and apoptosis on cytokine withdraw1 (Ahmed et al., 1997; Songyang et al., 1997). In contrast, expression of a constitutively active form of PI3 kinase in the T-cell line Kit225 was not sufficient to induce DNA synthesis or GUS phase transition (Brennan et al., 1997). Thus, PI3 kinase may contribute to lymphocyte proliferation, however, it appears to be neither necessary nor sufficient for this response. Cantrell and colleagues reported that IL-2 increases the activity of the proinitogenic transcription factor E2F in T cells through a pathway involving PI3 kinase and Akt (Brennan et al., 1997). This was shown using a transient transfection assay in which expression of a reporter gene linked to two tandem E2F-binding sites was induced by IL-2R activation. This inducible E2F activity was blocked by LY294002, a pharinacologic inhibitor of PI3 kinase, or by a dominant-negative version of PI3 kinase. Conversely, constitutively active forms of PI3 kinase or Akt were sufficient to induce reporter gene activity. LY294002 was shown to also inhibit IL-2-mediated induction of the cyclin D3 gene and downregulation of p27kipl expression. These two events are thought to promote phosphorylation of Rb, which results in the release of free and active E2F (Brennan et al., 1997). Given the general involvement of E2F in cell cycle regulation, these results suggest a mechanism by which PI3 kinase and Akt could contribute to proliferative signaling by the IL-2R. The PI3K pathway has also been implicated in the maintenance of cell viability, as described in Section D. 5. Role of mTOR
An important clue to how the IL-SR transduces signals affecting cell cycle progression is provided by the antibiotic rapamycin, a lipophilic inacrolide derived from the bacterium Streptonigces hygroscopicus (for reviews see Abraham and Wiederrecht, 1996; Karnitz and Abraham, 1996). Rapainycin inhibits T-cell proliferation in late G1 in response to IL-2 (Bierer et al., 1990; Duinont et al., 1990; Morice et al., 1993a) and indeed blocks cell cycle progression in response to a number of growth factors, both in inainmals and in yeast. This conserved activity suggests that rapamycin acts at a key control point for cell growth regulation, and hence, there has been great interest in defining the biochemical pathways that
40
BRAD H. NELSON AND DENNIS M . WILLERFORD
are affected by the agent. In mammalian cells, the effects of rapamycin appear to be mediated through mTOR (the mammalian target of rapamycin) (reviewed in Karnitz and Abraham, 1996). The structure of mTOR initially suggested that it was a lipid kinase, although evidence shows that it may in fact function as a serinehhreonine kinase. Indeed, Abraham and colleagues reported that the PHAS-1 protein, which is involved in the regulation of protein translation, is directly phosphorylated on serine and threonine residues by mTOR (Brunn et al., 1997). Another molecule that is activated through mTOR, although not thought to be a direct substrate, is p70 S6 kinase (Brown et al., 1995; Calvo et al., 1992; Chung et al., 1992; Ferrari et al., 1993; Kuo et al., 1992; Price et al., 1992; reviewed in Chou and Blenis, 1995). This enzyme, which is essential for G1 progression in fibroblasts (Lane et al., 1993), phosphorylates the ribosomal protein S6 and thereby regulates the translation of a distinct subset of cellular mRNAs (Terada et al., 1994). These findings suggest that rapamycin may inhibit cell cycle progression by suppressing the translation of critical proteins involved in DNA replication or cell growth. Rapamycin has been shown to block the ability of IL-2R to downregulate expression of the cyclindependent kinase inhibitor ~ 2 7 ~ Iin ' ' T cells (Nourse et al., 1994), with a concomitant block in cyclin-dependent kinase activation (Firpo et al., 1994; Morice et al., 1993a,b),thus demonstrating one clear mechanism by which this agent inhibits cell growth. However, this is not the only pathway through which rapamycin antogonizes IL-2R signaling: in mice lacking p27, IL-2R-mediated cellular proliferation is intact and is still sensitive to inhibition by rapamycin (Nakayama et al., 1996). The precise mechanism by which mTOR is activated by the IL-2R has yet to be defined. There is evidence, however, in IL-3-dependent cell lines that one pathway to mTOR activation may operate through PI3 kinase (R. Abraham, personal cominunication). 6. Role of STAM
An addition to the list of factors that participate in mitogenic signaling by the IL-2R is the STAM protein (Takeshita et al., 1996, 1997). STAM may function as an adaptor protein, as it contains both an SH3 domain and an immunoreceptor tyrosine-based activation motif (ITAM) region, the latter being characteristic of signaling proteins associated with TCR and BCR. Overexpression of a putative dominant-interfering mutant of STAM, consisting of a deletion of the SH3 domain, was shown to inhibit both IL-2- and GM-CSF-induced proliferation of the pro-B-cell line BAF3 in a transient assay (Takeshita et al., 1997). It is unclear at present how STAM functions in the IL-2R mitogenic pathway. The ITAM region of STAM is reported to bind directly to Jak3 (or Jak2 in the case of the GMCSF receptor); however, it is not known whether STAM influences the
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
41
catalytic activity of these kinases (Takeshita et al., 1997). Similarly, it is not known whether STAM has an effect on known signaling pathways for IL-2R-mediated cell cycle progression, such as those involving Shc, Stat5, mTOR, or PI3 kinase.
7. Target Genes of tlze IL-2R In addition to signal transducing proteins, the IL-2R activates a number of target genes involved in cell growth and division. These include protooncogenes such as c-myc, cfos, and c-jun, as well as G1 cyclins (Shibuya et al., 1992). Use of a differential hybridization technique has led to the identification of several new genes that are induced in an iminediate-early manner by IL-2 (Beadling et al., 1993), and this number is likely to grow with the application of newer techniques, such as high-density DNA array hybridization, to the study of cytokine signaling. Of the target genes induced by IL-2, the pathways leading to cfos and c-jun are relatively well defined (see Section VI,3). In contrast, little is known about the pathway(s) that regulates expression of the c-myc and cyclin genes, despite the ubiquitous and central role they play in the control of cell growth. The importance of c-myc expression to mitogenic signaling is underscored by the fact that, in all the mutational analyses that have been performed to date with IL-2R and other cytokine receptors, a perfect correlation has been observed between the expression of c-myc and the induction of cell proliferation. Regulation of the c-myc gene has been studied for over a decade in a wide variety of cells under many different stimulatory conditions, and three general control mechanisms have been identified, including modulation of transcriptional initiation, transcriptional elongation, and inRNA stability (reviewed in Spencer and Groudine, 1991). For most growth factors, including IL-2, it is not known which of these mechanisms is primarily responsible for regulating expression of c-nzyc. However, one clue may come from the observation that IL-2 can induce expression of a reporter gene containing a 2.3-kb fragment froin the 5' region of the c-myc gene, an effect that is enhanced by overexpression of STAM (Takeshita et al., 1997). This suggests that the IL-2R induces cmyc expression through cis regulatory elements located in the upstream 2.3-kb region of the gene. Studies with the IL-3 receptor further indicate that cytokine regulation of c-nzyc may involve the transcription factor E2F (Watanabe et al., 1995). Although these results are intriguing, more work is clearly required to establish whether such a mechanism is relevant to regulation of the endogenous c-rnyc gene by the IL-2R.
D. SIGNALINGPATHWAYS FOR CELLVIABILITY An important element of responses to signals generated by cytokine receptors, including the IL-2R, is the promotion of cell viability. In prolifer-
42
BRAD 11. NELSON AND DENNIS M . WILLERFORD
ating cells, this effect is somewhat difficult to separate from cellular reproduction. However, withdraw1 of cytokines, e.g., from proliferating T-cell lines, leads to apoptosis, indicating that positive survival signals are one component of the cytokine response. As discussed in Section V, IL-2R signals can promote the survival of T cells in the absence of proliferation (Akbar et al., 1996; Boise et al., 1995a; Gonzalez-Garcia et al., 1997), although the high concentrations of IL-2 required for this effect raise the issue of physiological relevance. The biochemical signal for T-cell viability appears to differ from that for proliferation in that it can occur without prior antigen stimulation and with only trace expression of Jak3 or phosphorylation of Jakl. The effects of IL-2 on the viability of resting T cells correlates with the catalytic activation of the tyrosine kinase Lck and is reduced markedly in the presence of LY294002, an inhibitor of PI3 kinase. Cell survival was also associated with induction of bcl-x expression in one study (Gonzalez-Garcia et nl., 1997), but not another (Boise et al., 1995b). The importance of the PI3 kinase pathway in signaling viability is supported by studies of Akt (also known as protein kinase B), a serinehhreonine kinase that is activated by IL-2R signals in a PI3 kinase-dependent manner (Ahmed et al., 1997; Brennan et al., 1997; Franke et al., 1995; Reif et al., 1997). Expression of a constitutively active Akt kinase is sufficient to induce bcl-2 expression and prolong the viability of BAF3 and 32D cells after cytokine withdrawal (Ahmed et nl., 1997; Songyang et al., 1997). Taken together, these studies suggest that the IL-2R may promote T-cell viability through a Jak-independent pathway involving PI3 kinase and Akt and the target genes bcl-x and/or bcl-2. This inodel is consistent with results from Taniguchi’s group demonstrating that IL-2R-mediated induction of the bcl-2 gene in the pro-B-cell line BAF3 does not require the cataytic activity of Jak3 (Kawahara et al., 1995). Although signals for cell viability and expression of bcl-2 family genes may be generated by a Jak-independent mechanism, they resemble the proliferative signal in requiring the presence of one or more tyrosine residues on the IL-2RP chain (Lord et al., 1998); perhaps these sites on IL-2RP are phosphorylated by Lck and serve as docking sites for the p85 regulatory subunit of PI3 kinase. VII. In Vivo Siudies of IL-2 Receptor Function in Lymphocyte Development
IL-2R components are expressed by iininature T and B cells and may therefore influence lymphocyte development at one or more stages. As IL-2RP and yc are shared with other cytokine receptors, investigations into the specific role of IL-2 and the high-affinity IL-2R in lymphocyte development must be considered together with experiments that address the contribution of signals generated by individual receptor components.
43
BIOLOGY OF THE INTERLEUKIN-2 KECEPTOH
Studies to date show that neither IL-2 nor its high-affinity receptor is required for the development of T, B, or NK cells; nevertheless, the immune deficiencies and inflammatory disorders associated with mutations of IL-2 or IL-2R components raise the possibility that the lymphocytes that develop in these animals have undergone abnormal selection processes. Because of its shared use in multiple lymphokine receptors, the yc chain is required for effective lymphopoiesis, principally due to the activities of the IL-7R in promoting the expansion and/or survival of early T- and Bcell progentiors. Mutations in the gene encoding yc are responsible for XSCID in humans and engender a severe lyinphophoietic defect in knockout mice. Moreover, the IL-2RP chain, which is shared by the IL-2 and IL-15 receptors, is required for the development of NK cells. The effects on lymphoid development of induced mutations in mice involving IL-2R components and related genes are summarized in Table I, and are discussed in detail below.
A. OVERVIEW OF T- AND B-CELLDEVELOPMENT The development of the imtnune system has two major components: the generation of lymphocytes with a large repertoire of antigen receptor specificities and the shaping of this repertoire by selection processes that exlude autoreactive cells and ensure optimal recognition under conditions of antigen presentation by accessory cells. The capacity of the immune system to respond to a nearly boundless array of different antigens is based TABLE I T- A Y D &CELL.DEVELOPMENT I N M l C E WITI-I MUT.4TIONS AFFECTINC: I L 9 R SI(:NALINC: Gene IL-2
T-cell Development
TCR Transgene Effects
B-Cell Development
Normal
Normal antigen-iuduced deletion in thymus Normal antigen-induced deletion in thymus Norinal positive and negative selection
Normal
I L - ~ R c Y Normal IL-2RP
x Jak3 IL-7 IL-7Ra IL-4
Normal; small thynus due to systemic disease Leaky block at DN stage Leaky block at DN stage Leaky block at DN stage Severe block at DN stage Normal
Incomplete rescue of T-cell development
Incoinplete rescue of T-cell development
Normal Nonnal
Leaky block at Stage Leaky block at stage Leaky block at stage Leaky block at stage Normal
pro-B pro-B pro-B pro-B
44
BRAD H. NELSON AND DENNIS M . WILLERFORD
on the ability of developing T and B cells to generate antigen receptors with a vast diversity of binding specificities. This is accomplished by the unique process of somatic gene rearrangement, involving the variable (V), diversity (D), and joining ( J ) gene segments of T-cell receptor and immunoglobulin genes (reviewed in Bogue and Roth, 1996; Gellert, 1997). Enormous combinatorid possibilities result from the large number of recombining segments, particularly V segments. Imprecision in joining these segments, which is an inherent property of the recombinase apparatus, further contributes to diversity of the receptor repertoire. Antigen receptor rearrangement is lineage specific and follows an ordered developmental pattern. In B cells, Ig heavy chains are rearranged prior to light chains, whereas in the major T-cell lineage, the TCRP locus rearranges before TCRa. Antigen receptor rearrangement is not merely a consequence of lymphocyte development, it also drives it: the stepwise expression of antigen receptor proteins triggers the major developmental events in both T- and B-cell lineages (reviewed in Kisielow and von Boehmer, 1995; Willerford et al., 1996).In the thymus, early TN cells have germline antigen receptor genes or have begun to rearrange TCRP. In-frame rearrangement of TCRP results in the expression of a functional pre-T-cell receptor, which includes the nascent TCRP chain, the pre-Ta chain, and components of the CD3 signaling apparatus, This complex signals a major developmental transition in the thymus, termed “beta selection,” involving a burst of cell division and expression of CD4 and CD8 to populate the DP thymocyte compartment. This transition also results in the activation of V(D)J recombination at the TCRa locus, and expression of the TCRa chain creates a complete ap T-cell receptor. The TCR drives the next set of developmental choices, based on binding to self-peptides presented by MHC molecules on thymic epithelial cells. The characteristics of this interaction determine negative and positive selection events, which result in the maturation of a population of exportable CD4’ or CD8’ single-positive (SP) T cells. In the B-cell compartment, in-frame rearrangement of IgH results in the expression of p chain, which associates with V-preB and h5 to form the pre-B-cell receptor. This receptor triggers cellular expansion and loss of CD43 expression. As in the DP thymocyte compartment, pre-B cells rearrange their light chains (IgK or Igh) and become immature IgM’AgD’ B cells. Thus, the major steps of T- and B-cell development are driven by successive signals generated following the stepwise rearrangement and expression of antigen receptor proteins.
B.
INTERLEUKIN-8 RECEPTOR FUNCTION DURING T- A N D B-CELLDEVELOPMENT Primitive thymocytes, comprising the CD4- CDB-CD3- TN population, contain a heterogeneous mix of cells, which span the developmental spec-
BIOLOGY OF T H E INTERLEUKIN-2 RECEPTOR
45
trum from multipotent progenitor cells to committed T cells having undergone productive rearrangement of one antigen receptor gene (for reviews see Godfrey and Zlotnick, 1993; Rothenberg, 1992; Shortnian and Wu, 1996; Willerford et al., 1996).The expression of the IL-2Ra chain (CD25) correlates with the major developmental events at the TN stage. Thus, cells within the CD25' subpopulation express Rag-1, Rag-2, and germline transcripts from the TCRP locus. Divergence of cells constituting the y6 and crp T-cell lineages occurs during this period, and rearrangement of TCRP is accomplished. Productive rearrangement and expression of the TCRP chain completes the pre-TCR, resulting in downregulation of CD25, and initiates a burst of proliferation and differentiation, which results in the filling of the CD4TD8' DP thymocyte compartment. Appropriate expansion and/or survival of the TN thymocyte population is critically dependent on yL,a property that is largely due to signaling in the context of the IL-7R. JL-7 stimulates the proliferation of this cellular subset (Conlon et al., 1989; Okazaki et nl., 1989; Watson et al., 1989). In additon, TCRP rearrangement in TN thymocytes in vitro is largely dependent on IL-7 (Muegge et al., 1993), an effect that likely reflects a trophic requirement for IL-7 at this stage, but could also indicate a more direct influence on the efficiency of V(D)J recombination (Cand'eias et al., 1997; Corcoran et al., 1996). Does CD25 expression in TN thymocytes indicate that high-affinity IL-2 receptors are functional at this stage? Experiments addressing this question have provided contradictory results. Human TN thymocyte subpopulations expressing either intermediate (i.e., CD25-) or high-affinity (i.e., CD25+) IL-2 receptors proliferate in response to IL-2 (Toribio et al., 1989). In mice, conflicting results have been obtained, with the clearest evidence for proliferative responses to IL-2 being found in fetal thymocytes (Raulet, 1985; Zuniga-Pflucker et al., 1990), with less convincing responses in cells derived from adult animals, which internalize the ligand-bound receptor inefficiently (Lowenthal et al., 1986; Raulet, 1985).In these studies the concentrations of IL-2 required for proliferation were often higher than expected for high affinity IL-2R interactions; however, high-affinity IL-2R binding has been demonstrated in the case of fetal thymocytes (Zuniga-Pfluckeret al., 1990). Nevertheless, it is possible that proliferation is not the only relevant response to IL-2R signals in TN thymocytes. Indeed, there is a suggestion that IL-2 also promotes differentiation of human immature thymocytes (Toribio et al., 1988) and antibodies to CD25, which block IL-2 signaling and inhibit thymocyte development in 14-day mouse fetal thymic organ culture (Jenkinson et al., 1987). The role of IL-2R signals in thymic development in vivo has been similarly examined using antibody inhibition studies. Administration of antibodies to CD25 in pregnant mice resulted in a profound block in
46
BRAD H. NELSON AND DENNIS M . WILLERFORD
T-cell development in neonates (Tentori et al., 1988). Following sublethal irradiation of mice, reconstitution of the thymus was slowed in animals receiving anti-CD25, but not an irrelevant antibody (Zuniga-Pflucker and Kruisbeek, 1990; Zuniga-Pflucker et al., 1990). These studies suggest that IL-2R signals could potentially play a role either in the early expansion of T-cell progenitors or in promoting efficient differentiation. These effects may be most evident under special conditions of kinetic stress, which were present in these experiments. As such, these results could be reconciled with data from IL-2 or IL-2Ra knockout mice that show no obvious defect in T-cell development under steady-state conditions. Alternatively, the observations made using antibody inhibition could reflect toxic effects of the reagent, rather than a specific block in IL-R signaling. As is the case for TN thymocytes, B-cell progenitors are importantly influenced by ye signals delivered by the IL-7R. IL-7 is secreted by bone marrow stromal cells and supports the proliferation of pro- and pre-B cells (Namen et al., 1988). IL-7R signals also promote B-cell antigen receptor rearrangement. Using a retroviral gene transfer system, cultured B-cell progenitors from IL-7Ra-deficient mice were reconstituted with wild-type or mutant IL-7Ra chains. Distinct mutations affected the proliferation of B-cell progenitors and the induction of antigen receptor rearrangement, suggesting either that IL-7R signals directly stimulate V( D)J recombination in B cells or that the efficiency of this process is enhanced by IL-7Rmediated survival signals (Corcoran et al., 1996). In developing mouse B cells, CD25 is expressed at the CD45RtIgM-CD43- pre-B-cell stage (Chen et al., 1994; Rolink et al., 1994), but is apparently not expressed in normal human B-cell precursors. The functional importance of CD25 expression in mouse pre-B cells is not clear, as such cells are apparently not responsive to IL-2. NK cells perform a cytolytic function in vivo that is thought to enhance host defenses against viruses, and possibly transformed cells as well. Unlike CTL, which recognize antigen via TCR interaction with MHC class I/ peptide complexes, NK activity is greatest against cells with low or absent class I expression, due to expression by NK cells of inhibitory receptors recognizing class I molecules (for reviews see Gumperz and Parham, 1995; Raulet, 1996; Spits et al., 1995). NK cells share some common features with T cells and appear to develop from a common progenitor (Carlyle et al., 1997; Rodewald et al., 1992; Sanchez et al., 1994). NK cells can develop in the thymus; however, extrathymic venues, particularly the bone marrow, may be the major sites of NK cell production. Although IL-2 stimulates the proliferation and activation of NK cells, this generally requires much higher doses than are required for T-cell proliferation, as IL-2Ra is not expressed on this subset (Trinchieri, 1989). Recent interest has focused
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
47
on IL-15, which promotes the development of NK cells from bone marrow progenitors in vitro, and supports their survival at concentrations consistent with signaling through the high-affinity IL-15R (Mr'ozek et al., 1996; Williams et d., 1997). Thus, IL-15 may represent the physiologic cytokine for NK cell development in vivo, whereas IL-2 may simply mimic this activity in vitro.
C. GENETIC STUDIESOF IL-2 AND IL-2Ra FUNCTION IN LYMPHOCYTE DEVELOPMENT 1. IL-2 and IL-2Ra Transgenic Mice Overexpression studies in mice have utilized transgenes encoding human IL-2 and IL-2Ra in order to distinguish transgenic from endogenous proteins. In one transgenic line in which human IL-2 was driven by the H-2K' promoter, expression of human IL-2 was detected in the thymus, but no resulting developmental abnormalities were reported (Ishida et aZ., 1989). A second study reported a human IL-2 transgene expressed under the control of the metallothionin promoter, which caused upregulation of IL-2Ra on thymocytes, but no other developmental consequences (Kromer et nl., 1991). Overexpression of human IL-2Ra using the H-2Kd promoter resulted in the acquisition of high-affinity receptors on the majority of thymocytes and conferred the ability to proliferate in response to IL-2, but again no developmental abnormalities were reported (Nishi et nl., 1988). However, a second report of mice bearing a human IL-2Ra transgene expressed under the SV40 promoter and enhancer indicated a modest decrease in CD4+CD8+thymocytes, with a concomitant increase in the DN subpopulation (Gutierrez-Ramos et al., 1989), suggesting a partial inhibition of T-cell development. Mice expressing a combination of IL-2 and IL-2Ra transgenes under the control of the H-2Kdpromoter developed a greatly expanded thymic population of large granular lymphocytes with NK activity and unrearranged T-cell receptor genes. The mice were small and died by 4 weeks of age, exhibiting infiltration of the cerebellum by NK cells and neuronal loss. This dramatic phenotype likely reflects the abnormally broad expression of high-affinity IL-2R and does not necessarily reflect a physiologic function of IL-WIL-2R interactions in the thymus. Given the probable role of IL-15R signals in NK cell development, this transgenic sytem probably generates an overly active IL-2RP and signal, and perhaps can be interpreted as equivalent to an overly active IL15R signal. 2. Interleukin-2- and IL-2Ra-Deficient Mice The functional role of IL-2 and the high-affinity IL-2R in lymphocyte development have been examined in mice with targeted disruption of
48
BRAD H. NELSON AND DENNIS M. WILLERFORD
either IL-2 or IL-2Ra genes (Schorle et al., 1991; Willerford et nl., 1995). No apparent defects in T- or B-cell development were identified in young adults of either strain, indicating that steady-state production of phenotypically normal lymphocytes does not require IL-2 or high-affinity IL-2R. Detailed analysis of CD4- CD8- CD3- TN thymocyte subsets in either IL-2Ra-deficient or IL-2Ra-/- X Rag-24- mice did not reveal any differences in either the phenotype or the distribution of cells in this subpopulation other than absence of CD25 expression (D. M. Willerford, unpublished observation). The development of NK cells also appears normal in mice lacking IL-2 (Schimpl and Hunig, 1994). In IL-2-deficient mice expressing a class I MHC-restricted TCR transgene, antigen-induced deletion of DP thymocytes was normal, both in vitro and in vivo (Kramer et al., 1994). Similar results were observed in IL-2Ra-deficient mice bearing a class I1 MHC-restricted TCR transgene (D. T. M. Leung and D. M. Willerford, manuscript in preparation). Thus, these gene targeting experiments do not support an essential role for IL-2R signaling in T-or B-cell development. However, a more subtle or redundant role for such signals, e.g., in promoting efficient progression through the TN thymocyte stage, is not ruled out by these studies. 3. Human IL-2 and IL-2Ra Deficiencies
Cases of human patients with severe combined immune deficiency have been reported with deficiencies in IL-2 secretion (Chatila et al., 1990; Pahwa et al., 1989; Weinberg and Parkman, 1990) or IL-2Ra expression (Roifman, 1997; Sharfe et al., 1997). T and B cells were present in these patients, consistent with the finding of phenotypically normal development of mature lymphocytes in mice deficient in IL-2 or IL-2Ra. In the report by Roifman and colleagues, a frameshift mutation was identified near the 5' end of the IL-2Ra gene, which abolished expression of the IL-2Ra protein (Sharfe et al., 1997). Detailed examination of the thymus revealed normal size and immunohistochemical staining patterns for CD4, CD8, and class I and I1 MHC. A striking absence of staining for CDla on cortical thymocytes was observed, whereas CDla expression could be normally upregulated on activated monocytes, suggesting that the defect was specific to thymic development. CD1 molecules are structurally related to class I MHC and include five genes in humans, designated CDla-e (Calabi and Milstein, 1986; Martin et al., 1986), and two in mice (CD1.l and CD1.2) (Bradbury et al., 1988), both of which are homologous to human CDld. Human CDlb appears to be involved in antigen presentation for glycolipids (Beckman et al., 1994), and the mouse CD1.l gene is required for efficient thymic selection of the NK1.l-bearing T-cell subset (Chen et al., 199%; Mendiratta et al., 1997). The finding that CDla expression is absent in
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
49
IL-2Ra-deficient human thymocytes cannot be investigated adequately in IL-2Ra-deficient mice, as no clear homolog to CDla has been identified. Nevertheless, abnormal thymus composition in this patient with IL-2Ra deficiency raises that possibility that abnormalities in T-cell development might as yet be unrecognized in knockout mice and may contribute to the immune and homeostatic abnormalities associated with this mutation. I N IL-2RP CHAIN-DEFICIENT MICE D. LYMPHOCYTE DEVELOPMENT
The role of IL-2RP-dependent signals (including, at a minimum, those induced by IL-2 and IL-15) in lymphocyte development has been characterized in mice with a targeted null mutation of the IL-2RP gene (Suzuki et al., 1995). As in mice deficient in IL-2 and IL-2Ra, the development of T and B lymphocytes is phenotypically normal in young animals laclang IL-2RP, although the thymus is small. This observation is thought to be a secondary consequence of the inflammatory disease that affects these animals, resulting in corticosteriod-mediated thymic involution, as T-cell development in fetal thymus organ cultures is normal. In contrast, the development of several specialized T-cell subsets appears to be impaired, including NKl' T cells and the the population of gut intraepithelial lymphoctyes utilizing yFfCR (Ohteki et al., 1997; Suzuki et al., 1997a). NK cell development is also abrogated in IL-ZRP-deficient mice. IL-2RP-deficient mice are susceptible to a severe and complex inflammatory disorder, beginning at 4weeks of age, that is dependent on T cells (see Section VIII,A,2). One potential explanation for this disorder is abnormal thymic repertoire selection leading to the maturation of autoreactive T cells. The role of IL-2RP in positive and negative selection processes in the thymus was investigated by introducing a TCR transgene specific for the male antigen, HY, into the IL-2RP-deficient background (Suzuki et al., 199713). Positive and negative selection were found to be normal in thymi from, respectively, female and male IL-2RP-deficient mice. However, given that the HY transgene is selected on class I MHC, and several aspects of the autoimmune disease in IL-2RP-deficient mice are dependent on CD4+ cells, a defect in class I1 MHC-dependent thymic selection remains a formal possibility.
E. ROLEOF yc IN LYMPHOCYTE DEVELOPMENT 1. Mutations ($yc in XSCZD XSCID represents the most common form of severe combined immunodeficiency in children, accounting for approximately half of patients in several series (Buckley et al., 1997; Fischer et al., 1997; Stephan et al., 1993). The disease gene (termed SCIDX1) was mapped to the vicinity of Xq12-13.1 (de Saint-Bade et al., 1987; Puck et al., 1989). ShortIy after
50
BRAD H. NELSON AND DENNIS M. WILLERFORD
the molecular cloning of yo it was determined that the ye gene localized to Xq13 and mapped near markers associated with SCIDX1. Moreover, mutations in the the ye coding sequence were identified from XSCID patients (Noguchi et al., 1993c; Puck et al., 1993). Subsequent analysis has identified yc mutations in most XSCID patients (Buckley et al., 1997; Fischer et al., 1997). Boys with XSCID have severe thymic hypoplasia and lack peripheral T cells, demonstrating a critical role for ye in the early stages of human T-cell development (reviewed in Conley, 1992; Fischer et al., 1997). B-cell numbers are generally preserved, indicating that, in contrast to the situation in mice (described later), yc is not required for human B-cell development. However, XSCID B-cells have intrinsic functional abnormalities, including defective mitogen-induced proliferative responses (Gougeon et al., 1990). These observations are supported by X chromosome inactivation studies in female carriers of XSCID mutations, which demonstrate a nonrandom pattern of X inactivation in T cells (i.e., only the wild-type X chromosome is active). A random X inactivation pattern is seen in immature B cells, suggesting that no competitive disadvantage is engendered by the mutant yc during development, whereas mature B cells are biased toward the wild-type X chromosome, indicating that optimal survival and/or expansion of peripheral B cells requires yc (Conley et al., 1988). XSCID is a lethal condition, for which the only longterm treatment is bone marrow transplantation
2. Mice Lacking ye Detailed studies of the role of ycin lymphocyte development and function have been facilitated by the generation of mice with null mutations in the ye gene (Cao et al., 1995; DiSanto et al., 1995; Ohbo et al., 1996). The development of T cells in these mice differs somewhat compared to human XSCID patients in that T cells, while markedly reduced in number, are present in the periphery and accumulate with age. Unlike the phenotype in humans, B-cell development is markedly impaired in ye-deficient mice. The developmental block in T- and B-cell lineages occurs at the transition from the TN to DP thymocyte stage and the pro- to pre-B-cell stage, respectively. This suggests that early lymphocytes do not efficiently achieve complete formation of the pre-TCR and pre-BCR in ye-deficient mice. At this point in development, yc could potentially mediate signals for the expansion or survival of early lymphoid progenitors. In addition, such signals could facilitate efficient V( D )J recombination, although the fact that the impairment in lymphoid development is leaky indicates that this process does not absolutely require ye. Furthermore, developmental defects due to disrupton of V(D)J recombination in T cells are rescued by expression of an appropriately selecting T-cell receptor ap transgene
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
51
(Shinkai et al., 1993). However, expression of either MHC class I- or class II-restricted TCR transgenes in the 7,-deficient background produces only partial rescue of thymic cellularity (DiSanto et al., 1996; Nakajima et al., 1997b), indicating that 7, has a role in early thymic development that is independent of the antigen receptor rearrangement process. A subtle role of 'ye in thymic negative selection has been suggested based on the partial imparrment in 7,-deficient mice of the deletion of Vflll' T cells, which normally occurs in the presence of the Mtv-9 provirus (Nakajima and Leonard, 1997), and a similar result was reported in inice lacking Jak3, which resemble 7,-deficient mice (Saijo et al., 1997, see later). However, negative seletion of T cells bearing VPS was normal in ?,-deficient mice, as was deletion of VP6' T cells in the presence of the self-superantigen Mls-ld (Nakajima and Leonard, 1997). Negative selection of transgenic T cells recognizing the H-Y antigen was also normal in male 7,-deficient mice (DiSanto et al., 1996). Development of the 76 T-cell sublineage is markedy impaired in mice lacking yc. Furthermore, NK cells are also markedly diminished, in parallel with the N K defect observed in mice lacking IL-2RP (Suzuki et al., 1995). Taken together, these data indicate that the production of NK cells is critically dependent on signals delivered by the IL-2RP and 7, chains, but not IL-2Ra, a description that is consistent with the IL-15 receptor. The generation of mice deficient in IL-15 or the IL-15Ra chain will assist in resolving this issue. Of the cytokine receptors that utilize yc,the IL-7R appears to have the most profound effects on early lymphocyte proliferation and differentiation (see Section B). These observations suggest that the developmental phenotype of 7,-deficient mice is in large part due to deficient IL-7R signaling at early stages. This view is supported by studies of mice rendered IL-7 deficient by antibody injection (Grabstein et al., 1993) or by gene targeting (von Freeden-Jeffry et al., 1995) or of mice lacking the IL-7Ra chain (Peschon et al., 1994). These animals resemble 7,-deficient mice in that they all display a marked reduction in both T- and B-cell production. For T cells , expression of a transgenic TCR in mice lacking IL-7Ra results in only partial rescue of thymic cellularity, similar to what is observed in 7,-deficient mice, suggesting that the defects in this T-cell sublineage are primarily mediated by IL-7R in the early stages of thymic development (Crompton et al., 1997). However, careful comparison of thymic development in mice lacking y,, IL-7, or IL-7Ra reveals subtle differences. In mice lacking IL-7Ra, the generation of mature T cells appears more severely limited than in mice lacking either IL-7 or ye,suggesting that IL7Ra may generate developmental signals in addition to those delivered in the context of the IL-7Rdyc heterodirner. Indeed, the novel lyinphokine TSLP, which is derived from thymic stroma, also supports the growth of
52
BRAD H. NELSON AND DENNIS M. WILLERFORD
early T- and B-cell progenitors in a manner similar to IL-7 (Friend et al., 1994). The receptor for TSLP includes IL-7Ra (Peschon et al., 1994), as well as a novel TSLP receptor, but is independent of yc. In studies that utilized a pre-B-cell line responsive to both IL-7 and TSLP, the TSLP response was inhibited either by antibodies to IL-7Ra or TSLP receptor, but not by antibodies to 7, (S. Levin and A. Farr, personal communication). Thus, the more severe defect in T-cell development in IL-7Ra-deficient mice likely reflects impairment of signals delivered by both IL-7 and TSLP. Although disrupted IL-7R signaling explains many of the defects seen in 7,-deficient mice, there are also additional IL-7R-independent signals generated by yc during lymphocyte development. First among these are signals required for the development of NK cells, which are absent in mice lacking ye,but intact in IL-7Ra-deficient mice (He and Malek, 1996). As discussed earlier, this difference likely reflects the participation of yc in a cytokine receptor distinct from IL-7 and IL-2, most likely the IL15R. There may also be IL-7R-independent, 7,-dependent signals for ap T-cell development. In bone marrow reconstitution studies in which recipients were treated with antibodies to y,, defective T- and B-cell development similar to that seen in 7,-deficient mice was observed. With donor bone marrow derived from IL-7Ra-deficient mice, an additional inhibition of T-cell development was observed with antibody treatment, suggesting that yc delivers additional, IL-7Ra-independent signals to early thymic progentors (He et al., 1997). One possibility is that this IL-7independent signal is provided by the high-affinity IL-2R, which is redundant at the TN thymocyte stage in the presence of intact IL-7R signaling. Alternatively, these observations could suggest a role for another cytokine that utilizes yc in its receptor. 3. Deficiency of y, Is Mimicked by Defects in Jak3 Signaling One crucial function of yc in the activation of the IL-2R is to recruit the tyrosine kinase Jak3 to the receptor complex, an event that is required for most of the identified events downstream of the IL-2R (see Section VI).The importance of Jak3 in IL-2R function in viwo is underscored by developmental defects seen in mice made deficient in Jak3 by gene targeting (Nosaka et al., 1995; Park et al., 1995; Thomis et al., 1995). These mice have a severe defect in both T- and B-cell development, as well as in T-cell activation, which parallels the defects seen in ?,-deficient mice. Similarly, a subset of patients with an XSCID-like phenotype (i.e., SCID with circulating B cells) but a normal 7, gene have been found to have mutations in Jak3 (Macchi et al., 1995; Russell et al., 1995). Jak3 activity also appears to correlate with the degree of immunodeficiency: a patient with combined T- and B-cell deficiency, but partial immune function, was
BIOLOGY OF THE INTERLEUKIN-2 RECEFTOR
53
described with a mutation in the cytoplasmicdomain of ycthat only partially interfered with Jak3 binding (Russell et al., 1994). Therefore, similar to in vitro studies of IL-2R signal transduction, the phenotypic correlation between Jak3 activity and yc function in both mice and humans indicates that Jak3 activation is a required proximal event in signaling by cytokine receptors that utilize yc. VIII. In Vivo Studies of IL-2 Receptor Function in Peripheral Lymphocytes
The concept of IL-2 as a T-cell growth factor suggested that interruption of IL-2R signals should impair the amplification of immune responses and result in immune deficiency. Thus, the initial desciiption of mice laclang IL-2, which exhibited a lymphoid system with a relatively normal appearance and function, was somewhat surprising (Schorle et al., 1991). One approach to reconciling this in vivo finding with the well-established role of IL-2 as a T-cell growth factor in vitro is to invoke redundancy in cytokmes mediating T-cell growth, a concept that has gained support with a better appreciation of the large family of cytokine receptors that share yc and have overlapping cellular effects. However, such explanations do not address the two major abnormalities in mice deficient in IL-2 or IL2R components: an inability to appropriately control the size of secondary lymphoid tissues and the emergence of inflammatory disorders, at least some of which represent autoimmunity. Thus, these studies indicate that the clearest in vivo function for IL-2R signals is to negatively regulate the size and functional activity of the peripheral lymphoid compartment. In addition, such studies should encourage further attempts to reframe questions regarding IL-2R signaling mechanisms in the context of negative regulation of iinmune functions. The peripheral consequences of mutations in mice affecting IL-2R signaling are summarized in Table 11. A. IL-2R SIGNALS REGULATETHE SIZEA N D CONTENT OF THE SECONDARY LYMPHOID TISSUES 1. Peripheral Consequences of lL-2- and IL-2Ra Mutations in Mice
Mice lacking the capacity to make either IL-2 (Schorle et al., 1991) or the IL-2Ra chain (Willerford et al., 1995) have a similar phenotype. Young adult mice develop polyclonal expansion of the peripheral lymphoid compartment, with all the major cellular subsets represented at levels 5- to 10-fold higher than in wild-type littermates, suggesting a global defect in lymphoid homeostasis. T cells in these animals are characterized by a memory cell phenotype, with high expression of CD44 and low expression of CD62L, suggestive of previous activation. Depending on the age of the animals, avarying increase in the proportion of peripheral T cells expressing
TABLE I1 PERIPHERAL PHENOTYPES OF MICEWITH MUTATIONSAFFECTING I L B R SIGNALING Gene
T Cells
IL-2
Expanded, activated, memory phenotype
IL-2Ra
Expanded, activated, memory phenotype
IL-2RP Expanded, activated blastic
x
B Cells
Cellular Immunity
Humoral Immunity
Idammatory Disorders
Hemolyhc anemia Expanded initially, decline Expansion to Near-normal with age superantigens intact antiviral antibodies Inflammatory bowel disease Partly reduced CD4 Increased Ig levels and CD8 responses to virus Hemolyt~canemia Expanded initially, decline Expansion to Inflammatory bowel with age superantigens intact disease Increased Ig levels Partial reduction of T-cell expansion in oioo to antigen Wasting, granulocytosis Decreased initially decline Expansion to Absent with organ infiltration further with age superantigens intact Death by 12 weeks Increased Ig levels Absent antiviral responses Proliferative typhlitis Decreased Expansion to described in a subset superantigens delayed
Decreased; activated, memory phenotype, accumulate with age Jak3 Decreased; activated, Decreased memory phenotype, accumulate with age IL-7 Decreased Decreased IL-7Ra Nearly absent Normal IL-4 Normal
CD4+ T-cell subset alterations
Impaired IgE production
None reported
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
55
the activation marker CD69 is also observed. In addition to increased numbers, B-cell activity is increased, as evidenced by marked elevations in serum immunoglobulins. Older adult mice experience a progressive loss of B cells as a consequence of impaired bone marrow production; the etiology of this process is not clear. IL-2- and IL-2Ra-deficient mice are subject to two types of pathologic autoreactivity (Sadlack et al., 1993, 1994; Willerford et al., 1995). The first is a fatal, severe anemia, accompanied by markedly increased erythropoiesis in the spleen. This is likely a hemolytic anemia due to red cell autoantibodies. Only a minority of adult animals manifest anemia, but the hemolyhc process is likely present in a compensated form in the majority, since even animals with a normal red cell mass usually have increased splenic erythropoiesis and high numbers of circulating reticulocytes. Other autoantibodies, including those specific for colon tissue or DNA, can also be demonstrated (Sadlack et al., 1993; D. T. M. Leung and D. M. Willerford, unpublished observations). Older mice uniformly develop an inflammatory bowel disease, with diarrhea and wasting, that is ultimately fatal. The inflammatory process is restricted to the colon and includes infiltration with activated lymphocytes and neutrophils, mucosal ulceration, absecces of intestinal crypts, and signs of abnormal crypt regeneration. Histologically, these features resemble human ulcerative colitis. Bowel inflammation requires intestinal flora, as it is abrogated in IL-2-deficient mice raised under gnotobiotic conditions (Sadlack et al., 1993). Most or all of the regulatory abnormalities observed in IL-2- or IL-2Radeficient mice can be traced to T cells. Thus, enlarged lymph nodes and spleen, loss of B cells, and inflammatory bowel disease were abrogated in IL-24- X nude mice (Kramer et al., 199.5). Similarly, inflammatory bowel disease did not develop in IL-24- X Rag-24- mice, which lacked T and B cells, but did develop in IL-24- X JH-/- mice which lacked B cells (Ma et al., 1995a). In the B-cell lineage, the hypersecretion of immunoglobulins includes isotypes characteristically T-cell dependent: IgG1, IgG2a, IgGzb, IgA, and IgE, but not IgM or IgG3. Moreover, this abnormality is abrogated in IL-24- X n d n u mice (Kramer et al., 1995). Regulation of the peripheral lymphoid compartment by IL-NL-2R interactions is autonomous to hematopoietic cells, as the phenotype of enlarged lymph nodes and spleen occurred in Rag-24- mice receiving IL-2-deficient bone marrow. The critical regulatory function of IL-2 appears to involve prominent paracrine effects, since the lymphoid expansion in IL-24- bone marrow recipients is prevented by admixture of a 30% fraction of IL-2+/+ bone marrow (Kramer et al., 1995). The most straightforward conclusion from the phenotypes of IL-2- and IL-2Ra-deficient mice is that the major role of IL-UIL-2R interactions in
56
BRAD H. NELSON AND DENNIS M. WILLERFORD
vivo is to negatively regulate the peripheral lymphoid compartment. These negative effects include maintaining homeostasis in terms of the overall size of the secondary lymphoid tissues, as well as controlling the emergence of autoreactive clones in the periphery. These functions may involve separate mechanisms or represent progressive steps in a unitary process, as defects in homeostasis are manifest as early as 4 weeks of age in IL-2Radeficient mice, whereas autoimmune disorders usually occur at a later age. The apparent similarity in the phenotypes of IL-2- and IL-2Ra-deficient strains suggests that, for the most part, the biologic function of IL-2 is mediated through the high-affinity form of the IL-2R.
2. Peripheral Phenotype of IL-2RP-Deficient Mice Mice homozygous for a null mutation of IL-2Rp exhibit a complex phenotype involving both lymphoid and nonlymphoid hematopoietic lineages (Suzuki et al., 1995, 199713). Beginning at approximately 4 weeks of age, mice display growth retardation and signs of ill health. The lymph nodes and spleen are enlarged, with an expansion of activated T cells that are blastic and express CD69. Although analogous to what is seen in IL2- and IL-2Ra-deficient mice, the T-cell activation phenotype appears to be more dramatic in mice lacking IL-2RP. B-cell numbers are decreased in young adult mice, but display evidence of increased function, as serum levels of IgGl and IgE (but not other isotypes) are markedly elevated. Autoantibodies reacting with red blood cells, nuclear antigens, and DNA are also detected. As with IL-2- and IL-2Ra-deficient mice, B-cell numbers decline further with time. A generalized increase in granulopoiesis in the bone marrow and spleen is also observed in IL-2RP-I- mice, with granulocyte infiltration of liver and lymph nodes. IL-2RP-deficient mice die by 12 weeks of age, although a specific cause of death has not been identified. Similar to the situation in IL-2- or IL-2Ra-deficient mice, many of the phenotypic abnormalities in IL-2RP-I- mice can be traced to T cells. Depletion of CD4' T cells following the injection of anti-CD4 antibodies improves the overall health of IL-2RP-I- mice and, specifically, prevents the development of autoantibodies and depletion of B cells. Transfer of IL-2RP-I- T cells into nude mice also recapitulates the B-cell abnormalities. The disorder resulting in the accumulation of granulocytes could represent an autonomous defect of myeloid cells, as IL-2RP is expressed in the myeloid lineage, or could be secondary to aberrant regulation of T cells, which can secrete myelopoietic cytokines. The granulopoietic defect was not abrogated by depletion of CD4' T cells; however, this abnormality did develop in nude mice receiving IL-2RP-I- T cells, suggesting that this disorder may be a result of deregulated T-cell function. Inflammatory bowel disease is not described in mice lacking IL-2RP; however, these
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
57
animals typically die at about the age that this disorder begins to manifest in IL-2- or IL-2Ra-deficient mice. Thus, taken with the results of targeted deletion of IL-2 or IL-2Ra, the phenotype of mice lacking IL-2Rp supports the general notion that one of the major physiologic functions of IL-2R signals is to negatively regulate the peripheral lymphoid compartment. The similarities between these two groups of mice, particularly the B-ceI1 abnormalities, presumably reflect the same defect in intracellular signals normally delivered by the highaffinity IL-2R. The apparently more severe abnormalities in T-cell regulation in IL-2RP-I- inice suggest that negative regulation of the peripheral lymphoid compartment also involves additional signals delivered by IL2RP independent of the IL-2R. Thus, T-cell homeostasis appears to require at least two nonredundant signals: one delivered by the high-affinity IL2R and the other by a second receptor utilizing IL-2RP, such as the IL15R. It will be of great interest to see whether mice with null mutations in IL-15 or IL-15Ra also manifest deregulation of peripheral T cells. 3. Peripherul Consequences of yc and Jak3 De$ciency in Knockout Mice
As noted earlier, mice lacking yc or Jak3 have a severe but incomplete impairment in T- and B-cell development, indicating that while yJJak3 signals are not essential for lymphocyte differentiation, they are critical for expansion of early progenitors and/or efficient transition through ratelimiting developmental steps (Cao et d.,1995; DiSanto et d.,1995; Nosaka et aE., 1995; Ohbo et al., 1996; Park et al., 1995; Thomis et al., 1995). Peripheral lymph nodes are essentially absent in 7'- and Jak3-deficient mice; only the mesenteric lymph node is readily identified (Cao et al., 1995; Ohbo et al., 1996; Park et al., 1995; Thomis et al., 1995).The absence of peripheral lymph nodes is rather striking, as lymph nodes are found readily in Rag-2-deficient mice, which have no mature lymphocytes. This apparent difference may indicate that yJJak3 signals are required for the development of the peripheral lymph node structure, a property that could be mediated by a nonantigen receptor-bearing cell. With age, lymphocytes accumulate abnormally in the spIeen and mesenteric lymph node (but apparently not peripheral lymph nodes) of yc- and Jak3-deficient mice. Expansion of T cells is accompanied by an activated phenotype in the CD4' subset, including an increased proportion of larger cells, upregulation of CD44 and CD69, and downregulation of CD62L. Bowel lesions were reported in one strain of ?,-deficient mice, but were initially described as a proliferative typhlitis seen in association with helicobacter-like organisms and apparently distinct histologically from the condition resembling ulcerative colitis that occurs in IL-2- and IL-2Ra-deficient mice (Cao et al., 1995).Peripheral T cells in ycmice exhibit increased uptake of bromodeoxy-
58
BRAD H. NELSON AND DENNIS M. WILLERFORD
uridine, which, considering the markedly hypoplastic thymus, indicates that the increase in mature T cells is most likely due to expansion in the periphery. In Jak3-deficient mice reconstituted with a Jak3 transgene expressed in the thymus but not in peripheral T cells, T-cell development was restored; however, the peripheral phenotype of activated T cells and lymphoid expansion was preserved, indicating that these abnormalities reflect the peripheral regulatory function of Jak3 (Thomis and Berg, 1997). Aside from the developmental abnormalities that reflect the participation of yc in multiple cytokine receptors, the peripheral T-cell phenotype of y,-deficient mice can be at least partly explained by the lack of signals delivered by the high-affinity IL-2R, underscoring the role of the IL-2R in peripheral lymphoid homeostasis. Mutation of ycor Jak3 is not, however, associated with the autoreactivity characteristic of mice lacking IL-2, IL2Ra, or IL-2RP. This could reflect the developmental abnormalities present in yc- and Jak3-deficient mice, e.g., the fact that there are relatively fewer T cells than in mice deficient in other IL-2R components. However, this explanation is undermined by the the fact that restoration of T-cell development in Jak3-deficient mice by thymic expression of a Jak3 transgene does not lead to autoimmunity (Thomis and Berg, 1997). An alternative explanation is that while IL-2R function in controlling the size of the peripheral T-cell compartment consists of negative regulatory signals, the emergence of autoimmunity (or, for that matter, the generation of normal immune responses) additionally requires positive signals delivered by other cytokine receptors that utilize yc. One candidate for such a positive effect is IL-4, which is also a growth factor for T cells. However, the combined deficiency of IL-2 and IL-4 does not reverse the autoimmune manifestations characteristic of IL-2 deficiency, demonstrating that IL-4 is not required for autoimmunity to develop (Sadlack et al., 1994). Whether other cytokines that utilize yc positively influence autoimmunity remains to be tested.
B. ROLEOF IL-2R SIGNALSIN IMMUNE FUNCTION 1. Immune Responses in IL-2- and IL-2R-Dejicient Mice Immune responses have been studied in detail in IL-2-deficient mice (Kundig et al., 1993; Schimpl and Hunig, 1994; Schimpl et al., 1992, 1994) and, to a lesser extent, in IL-2Ra-deficient mice (Van Parijs et al., 1997). Secondary antibody responses to the model T-cell-dependent antigen TNPkeyhole limpet hemocyanin were normal or enhanced in IL-2-deficient mice (Schimpl et al., 1992). Antibody responses were also assessed following vesicular stomatitis virus infection of IL-24- mice (Kundig et al., 1993), where the induction of neutralizing IgM antibodies was normal, but the kinetics of the delayed IgG response were slowed, but not dampened. This
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
59
suggests that T-cell help for Ig class switching is dependent on IL-2 for the fastest response. CTL responses to vaccinia virus infection in IL-2deficient were found to be indistinguishable from wild-type mice, although the frequency of specific CTL was reduced threefold in IL-24- mice after infection with LCMV. The functional consequences of the latter observation were minimal, however, as local swelling at the site of inoculation was normal and LCMV-primed IL-24- mice were resistant to rechallenge with virus. However, a second group found that clearance of LCMV was delayed following infection in IL-24- mice and that expansion of CD8+ CTL in the spleen was inhibited markedly (Cousens et al., 1995). Induction of NK activity on day 3 following LCMV infection was seen in IL-24mice; however, it was reduced in magnitude three- to ninefold (Kundig et al., 1993).These studies demonstrate that humoral and cellular immune responses in IL-%deficient mice are largely intact, suggesting either that the primary role of IL-2 is not in mediating the obligatory expansion of antigen-reactive T cells during immune responses or that this function is largely redundant with other cytokines. These experiments also indicate that mice lacking IL-2 are not especially immunocompromised. How, then, can these observations be reconciled with the reports of human SCID in patients with defective IL-2 production (Weinberg and Parkman, 1990) or a null mutation in IL-2Ra (Shade et aE., 1997)?One possible explanation is suggested by the observation that older IL-24- mice fail to generate CTL responses to viral infection in vivo (Kundig et al., 1993). It could be then that immune deficiency in the absence of IL-WIL-2R signals does not directly reflect the participation of these signals in immune responses, but rather is a secondary consequence of the abnormal, expanded peripheral lymphoid compartment that develops over time and may result in the inability to mount a properly organized immune response. In contrast to the situation in IL-24- mice, immune responses to viral challenge are essentially absent in IL-2RP-deficient mice, including T-celldependent and -independent phases of the humoral response to vesicular stomatitis virus, as well as both CD4+ and CD8+ T-cell-dependent phases of the response to LCMV. These results have at least two possible interpretations. One is that amplification of immune responses in vivo is absolutely dependent on cytokine receptor signals utilizing IL-BRP, but that either IL-2R or IL-15R signals can subserve this function. Alternatively, the severe illness that manifests early in life in IL-2RP-I- mice may itself contribute to the observed immune deficiency, much as appears to be the case in older IL-2-deficient mice. 2. Immunodeficiency in Patients with Defective IL-2 or IL-2R Function Severe combined immunodeficiency in humans is a clinical syndrome involving loss of function in both cellular and humoral arms of the immune
60
BRAD H. NELSON A N D DENNIS M. WILLERFORD
response, resulting in susceptibility to infection (Rosen et al., 1995). The clinical definition of SCID encompasses individuals with developmental defects in the gerierative lymphoid organs, as well as functional defects in settings where T and B cells are present. Among the latter group of patients, several have been described with impaired secretion of IL-2 by activated T cells (Chatila et al., 1990; Doi et al., 1988; Paliwa et al., 1989; Weinberg and Parkman, 1990). Although none of these cases described a defect in the IL-2 gene, in one instance (Pahwa et al., 1989) a clinically significant improvement in immune function was reported following therapy with IL2. Roifman and colleagues described a child with clinical SCID and a lack of IL-2Ra expression due to a mutation in the IL-2Ra gene (Roifman, 1997; Sharfe et al., 1997). This patient had several features reminiscent of the phenotype of IL-2Ra-deficient mice, including lymphadenopathy, splenomegaly, and autoimmune manifestations involving the skin and gut. Thymic abnormalities were also described in this patient (see Section VI,C,3). The SCID phenotype associated with yc or Jak3 deficiency in humans is readily explained by the characteristic absence of peripheral T cells. However, a mutation in the extracellular domain of yc has been described in two patients with SCID who had normal T-cell numbers (Sharfe et al., 1997).Expression of ycwas not affected, but reduced receptor affinity for IL-2 was observed, along with deficient responses to IL-2. This mutation therefore apparently differentially affects yc participation in different cytokine receptors, permitting ye to function in the context of receptors required for lymphoid development, but with reduced activity with respect to IL-2R function. These clinical observations illustrate an important point that is perhaps underappreciated in studies of IL-2- and IL-2Ra-deficient mice: that peripheral IL-2/IL-2R interactions are required for effective host defenses. C. ROLEOF IL-2R
SIGNALS IN
T CELL GROWTH A N D SURVIVAL
IN
VWO
1. T-cell Expansion in Response to Antigen
In order to test the long-standing hypothesis that IL-2R signals mediate expansion of T cells in vivo during immune responses, the behavior of T cells after activation in vivo was examined in mice with mutations in IL-WIL-2R signaling. In the first 3-5 days following immunization with bacterial superantigens, the expansion of T cells expressing the corresponding VP segments was intact in mice lacking IL-2Ra (Willerford et al., 1995), IL-2 (Kneitz et al., 1995), and IL-2RP (Suzuki et al., 199%). In y,-deficient mice, expansion of SEB-reactive T cells was less than wildtype mice after 3 days, but reached an equivalent peak at 5 days. Using a class I MHC-restricted TCR transgene bred into the IL-%deficient
BIOLOGY OF THE INTERLEUKIN-2 RECEE'TOR
61
background, Schiinpl and colleagues (Kramer et al., 1994) demonstrated that antigen-stimulated blast transformation and transition through the cell cycle in vivo did not require IL-2. These studies demonstrate that T cells can progress through the cell cycle and expand in vivo in the absence of IL-2/IL-2R signals, and indeed in the absence of signals for all lymphokines using yc, raising the question as to what other T-cell growth factors are utilized in uiuo. It is possible that a role for IL-2R is masked by the singular potency of superantigen signals in these experiments. Expansion and differentiation of T cells following exposure to antigen in vivo can be followed using the system reported by Jenkins and colleagues (Kearney et al., 1994), wherein T cells bearing a transgenic TCR are adoptively transferred into a syngeneic, nontransgenic host. Under these conditions, the TCRtransgenic T cells represent only a few percent of the normal T-cell complement, but can be followed accurately in viuo using a monoclonal antibody specific for the TCR clonotype. In addition to measuring antigen responses, this system has the advantage of examining mutant T cells in the context of a normal lymph node structure. D. T. M. Leung and D. M. Willerford (manuscript in preparation) bred the DO1l.10 TCR transgene (Murphy et al., 1990),which is class I1 MHC restricted and specific for an ovalbuminderived peptide, onto the IL-2Ra-deficient background and examined Tcell expansion in response to the antigenic peptide following adoptive transfer into normal BALB/c mice. A modest (approximately 50%) reduction in the degree of T-cell expansion was observed for IL-2Ra-deficient T cells compared with wild-type mice, suggesting that T-cell expansion following encounter with antigen may be at least partially dependent on high-affinity IL-ZR signals. Similar experiments have been performed using DO1l.10 X IL-24- T cells. In these experiments, following immunization with high doses of antigen subcutaneously, T-cell expansion was equivalent in IL-24- and IL-2+ transgenic T cells (Khoruts et al., 1998). However, when lower doses of antigen are used, IL-24- T cells exhibit impaired expansion in vivo (M. Jenkins, personal communication). Hence, the longstanding view that IL-2 amplifies immune responses by promoting T-cell expansion in vivo finds some support in these experiments. Further work with this system may help further clarify the role of IL-2R signals in immune responses and perhaps contribute to understanding the defect in host defenses in humans with abnormalities in IL-2 or IL-2Ra expression.
2. Role of IL-2R Signals in Antigen-Mediated T-cell Deletion in Vivo Lymphoid homeostasis requires that the majority of cells generated during iinniune responses undergo cell death (Sprent and Tough, 1994). Given the studies of Lenardo (1991), indicating that IL-2R signals may promote death as an outcome of TCR stimulation, it has been postulated
62
BRAD H. NELSON AND DENNIS M. WILLERFORD
that the defect in lymphoid homeostasis in IL-2R-deficient mice reflects a relative decrease in the proportion of cells undergoing AICD following antigenic encounter (Kneitz et al., 1995; Willerford et al., 1995). The role of IL-2 and IL-2R interactions in activation-induced T-cell death has been investigated in vivo by examination of superantigen (SEB)-mediated peripheral T-cell deletion in IL-2Ra-deficient mice (Willerford et al., 1995). In these experiments, deletion of VPS' T cells 10 days following immunization with SEB was partially impaired. Similar results have been obtained for the CD4+ T-cell subset in mice deficient in IL-2 (Kneitz et al., 1995) and for T cells in mice lacking yc (Nakajima and Leonard, 1997). These results would appear to support the notion that lymphoid accumulation in IL-2R-deficient mice could result from defective AICD following antigen exposure. However, one important caveat in interpreting these experiments is that the subset distribution of T cells in the mutant mice was biased toward a memory phenotype prior to SEB immunization, which could affect sensitivity to AICD in vivo. More recent work has provided data that conflict with the hypothesis that IL-2R signals are required for efficient AICD in vim. In contrast to the foregoing results, SEB-mediated T-cell deletion is intact in young mice lacking IL-2RP (Suzuki et al., 1997b). The role of IL-2R in antigenmediated AICD in vivo has also been investigated using IL-2Ra-deficient mice bearing the class I1 MHC-specific DO1l.10 TCR transgene (Murphy et al., 1990), where subcutaneous administration of an antigenic peptide resulted in the efficient deletion of peripheral T cells after 8 days, with no differences compared with wild-type littermates. Antigen-induced peripheral deletion was also normal in D011.10+, IL-2Ra-/- T cells that were transferred adoptively into normal BALB/c mice (D. T. M. Leung and D. M. Willerford, manuscript in preparation). These studies suggest that the IL-2R may not have an indispensable role in regulating AICD following the activation of T cells by high-affinity antigens in vivo. It is probably fair to say that experiments done to date have not exhausted the physiologic diversity of antigen dose, receptor affinity, and site of encounter with the immune system, so the precise in vivo role of IL-2R signals in AICD following T-cell stimulation with antigen remains incompletely defined. AICD in T cells is mediated through activation of Fas, via interaction with FasL, although data suggest that TNFa, and perhaps other TNF family members, may have overlapping roles in this process (see Section IV,C). Therefore, an important question regarding the lymphoproliferative defects in IL-2 and IL-2R-deficient mice is to what extent these are mediated by abnormalities in Fas/FasL regulation. Expression of Fas and its upregulation following TCR stimulation is normal in mice deficient in IL-
BIOLOCY OF THE INTERLEUKIN-2 RECEPTOR
63
2 or IL-2Ra (Kneitz et al., 1995; Van Parijs et al., 1997). IL-2 has also been reported to contribute to the upregulation of FasL on activated T cells (Suda et al., 1995). Moreover, because T cells are differentially sensitive to cell death mediated by Fas ligation, depending on their state of activation, IL-2R signals could also affect the Fas pathway by modifying the signaling apparatus downstream of the Fas receptor. One possible mechanism for such an effect is the upregulation of c-myc by IL-2R signals, which may increase the cellular sensitivity of Fas-mediated apoptosis (Huber et al., 1997). In support of this, activated T cells derived from IL-2 and IL-2Radeficient T cells are resistant to killing by antibody-mediated cross-linking of Fas (Kneitz et al., 1995; Van Parijs et al., 1997).Whether these observations reflect a specific effect of IL-2 receptor signals on the Fas signaling pathway or simply reflect differences in T-cell proliferation in vitro in the absence of 1L-2R signals is not yet clear (Boehme and Lenardo, 1993; Zhu and Anasetti, 1995). In contrast to the preceding results, Fas sensitivity was found to be normal in T cells derived from IL-2RP-deficient mice (Suzuk~et al., 1997b). One approach to evaluating the potential role of the Fas pathway in the disordered T-cell homeostasis in mice lacking IL-2 or IL-2Ra is to compare the phenotypes of these mutations with mice deficient in Fas or FasL. Superficially, both types of mutations are associated with enlarged lymph nodes and autoimmunity. However, the peripheral lymphoid expansion in IL-2- and IL-2Ra-deficient mice is general, involving more or less normal proportions of all lymphoid subsets, indicating a global defect in lymphoid homeostasis. In contrast, mice with genetic defects in the Fas pathway acquire large lymph nodes specifically because of the accumulation of an unusual CD4-CD8- T-cell subset (Cohen and Eisenberg, 1991),which is not increased in mice lacking IL-2 or IL-2Ra. Although exibiting strainspecific variation, the spectrum of autoimmune manifestations in Fas pathway-defective mice predominantly consists of autoantibody and immune complex &orders (Cohen and Eisenberg, 1991). While IL-2- and IL-2Ra-deficient mice develop a hemolytic anemia, which is presumably antibody mediated, other antibody or immune complex disorders seen in Fas/FalsL-deficient mice have not been described. Finally, the inflammatory bowel disease, which is a major component of the phenotype in IL2- and IL-2Ra-deficient mice, is not seen in Fas pathway-defective mice. Thus, a phenotypic cornparison indicates that the disrupted regulation of the peripheral lymphoid compartment in IL-2- and IL-2Ra-deficient mice is not due primarily to a defect in the Fas pathway, although a contribution of impaired Fas-mediated AICD to some of the abnormalites in these mice cannot be excluded.
64
BRAD H. NELSON AND DENNIS M. WILLERFORD
IX. Summary and Conclusions
Studies of the biology of the IL-2 receptor have played a major part in establishing several of the fundamental principles that govern our current understanding of immunology. Chief among these is the contribution made by lymphokines to regulation of the interactions among vast numbers of lymphocytes, comprising a number of functionally distinct lineages. These soluble mediators likely act locally, within the context of the microanatomic organization of the primary and secondary lymphoid organs, where, in combination with signals generated by direct membrane-membrane interactions, a wide spectrum of cell fate decisions is influenced. The properties of IL-2 as a T-cell growth factor spawned the view that IL-2 worked in vivo to promote clonal T-cell expansion during immmune responses. Over time, this singular view has suffered from increasing appreciation that the biologic effects of IL-2R signals are much more complex than simply mediating T-cell growth: depending on the set of conditions, IL-2R signals may also promote cell survival, effector function, and apoptosis. These sometimes contradictory effects underscore the fact that a diversity of intracellular signaling pathways are potentially activated by IL-2R. Furthermore, cell fate decisions are based on the integration of multiple signals received by a lymphocyte from the environment; IL-2R signals can thus be regarded as one input to this integration process. In part because IL-2 was first identified as a T-cell growth factor, the major focus of investigation in IL-2R signaling has been on the mechanism of mitogenic effects in cultured cell lines. Three critical events have been identified in the generation of the IL-2R signal for cell cycle progression, including heterodimerization of the cytoplasmic domains of the IL-2RP and yc chains, activation of the tyrosine kinase Jak3, and phosphorylation of tyrosine residues on the IL-2Rp chain. These proximal events led to the creation of an activated receptor complex, to which various cytoplasmic signaling molecules are recruited and become substrates for regulatory enzymes (especiallytyrosine kinases) that are associated with the receptor. One intriguing outcome of the IL-2R signaling studies performed in cell lines is the apparent functional redundancy of the A and H regions of IL2RP, and their corresponding downstream pathways, with respect to the proliferative response. Why should the receptor complex induce cell proliferation through more than one mechanism or pathway? One possibility is that this redundancy is an unusual property of cultured cell lines and that primary lymphocytes require signals from both the A and the H regions of IL-2RP for optimal proliferative responses in vivo. An alternative possibility is that the A and H regions of IL-2RP are only redundant with respect to proliferation and that each region plays a unique and essential
BIOLOGY OF THE INTERLEUKIK-2 RECEPTOR
65
role in regulating other aspects of lymphocyte physiology. As examples, the A or H region could prove to be important for regulating the sensitivity of lymphocytes to AICD or for promoting the development of NK cells. These issues may be resolved by reconstituting IL-2RP -/- mice with Aand H-deleted forms of the receptor chain and analyzing the effect on lymphocyte development and function in uivo. In addition to the redundant nature of the A and H regions, there remains a large number of biochemical activities mediated by the IL-2R for which no clear physiological role has been identified. Therefore, the circumstances are ripe for discovering new connections between molecular signaling events activated by the IL-2R and the regulation of immune physiology. Translating biochemical studies of IL-2R function into an understanding of how these signals regulate the immune system has been facilitated by the identification of natural mutations in IL-2R components in humans with immunodeficiency and by the generation of mice with targeted mutations in these genes. Efficient lymphopoiesis requires the yc chain, and mutations in yc are responsible for human XSCID, an important cause of congenital immunodeficiency. Boys with XSCID and mice with targeted disruption of the ycgene share defects in early T-cell development due to the participation of yL in multipte cytokine receptors. Signals from IL-7R are probably most important at this stage, but other 7,-containing cytokine receptors may also be involved. In the peripheral lymphoid compartment the most striking abnormality in mice lacking IL-2, IL-2Ra, or IL-BRP is an impaired ability to control the overall size of the secondary lymphoid tissues. The immune system is not encompassed by a physical capsule, as are other solid organs, and is subject to cellular fluxes from continuous production of cells by the primary lymphoid organs and the expansion of lymphocytes in the periphery during immune responses. Yet, under normal circumstances, the overall size of the peripheral lymphoid compartment is strictly controlled over time. Thus, one of the most important physiologic functions of the IL-2R in vivo appears to be the homeostatic regulation of lymphoid tissues. IL-2R signals may exert a negative regulatory influence by promoting AICD following encounter with antigen, thus limiting clonal expansion and participating in the termination of immune responses. However, data indicate that MHC (and presumably TCR-)-dependent signals may also control T-cell survival independent of encounter with antigen. It is worth considering whether such survival signals could be counterbalanced by IL-2R-dependent negative regulatory influences. IL-2R defects in mice also lead to the development of fatal inflammatory disorders, which are accompanied by autoantibodies, suggesting that IL-2R signals play a role in maintaining peripheral immune tolerance. One proposed mechanism for this effect is in preventing
66
BRAD 13. NELSON AND DENNIS M . WILLERFORD
the emergence of autoreactive clones during immune responses through IL-2R-mediated promotion of AICD, although the in vivo evidence for such an effect is conflicting. Thus, the pathogenic mechanism of inflammatory disorders in IL-2R-deficient mice is not well understood. Nevertheless, the fact that IL-2R signals are now known to be required to suppress these conditions may help illuminate the causes of similar conditions in humans. Efforts by many investigators to define the biologic role of the IL-2R now span more than 20 years and represent a substantial fraction of the scientific output of the immunology community. The identification of IL2R mutations in humans and the creation of such mutations in mice have provided some surprising and puzzling insights into the function of IL2R signals in vivo. These phenotypes provoke a necessary check on the assumptions that underlie experiments performed in cell systems and provide a more complete context in which to interpret results. More importantly, these complex phenotypes in vivo have stimulated a broader consideration of signals delivered by the IL-2R and will hopefully lead to a fuller understanding of the diversity of intracellular pathways that are utilized by this model cytokine receptor.
ACKNOWLEDGEMENTS The authors thank Phil Creenberg, Ken Kaushansky, Robert Abraham, Averil Ma, Raymond Doty, and Chaim Roifman for reading the manuscript and making many helpful suggestions. We are grateful to Drs. Roifinan and Abraham, along with Warren Leonard, Marc Jenkins, Steve Levin, and Andy Farr for sharing unpublished data. D.M.W. is supported by NIH Grants AI-01173 and AI-41051.
REFERENCES Abbas, A. K. (1996). Cell 84,655-657. Abraham, R. T., and Wiederrecht, C. J. (1996). Annu .Rev. Imniunol. 14, 483-510. Adachi, M., Ishino, M., Torigoe, T., Minami, Y., Matozaki, T., Miyazaki, T., Taniguchi, T., Hinoda, Y., and Iinai, K. (1997). Oncogene 14, 1629-1633. Ahmed, N. N., Crimes, H. L., Bellacosa, A,, Chan, T. O., and Tsichlis, P. N. (1997). Proc. Natl. Acad. Sci. USA 94, 3627-3632. Ajchenbaum, F., Ando, K., DeCaprio, J. A,, and Griffin, J. D. (1993). J. B i d . Chenz. 268, 4113-4119. Akbar, A. N., Borthwick, J. J., Wickremasinghe, R. G., Panayiotidis, P., Pilling, D., Bofill, M., Krajewski, S., Reed, J. C., and Salmon, M. (1996). Eur. J. Immztnol. 26, 294-299. Alderson, M. R., Tough, T. W., Davis-Smith, T., Braddy, S., Falk, B., Schooley, K. A,, Goodwin, R. G., Smith, C. A., Ramsdell, F., and Lynch, D. H. (1995).J . E x p Med. 181,71-77. Anderson, D. M., Kumaki, S., Ahdieh, M., Bertles, J., Tometsko, M., Loomis, A., Gin, J., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., et al. (1995).J. Biol. Chem. 270,2986229869. Arima, N., Kamio, M., Imada, K., Hori, T., Hattori, T., Tsudo, M., Okuma, M., and Uchiyama, T. (1992).J. Exp. Med. 176, 1265-1272.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
67
Asao, H., Takeshita, T., Ishii, N., Kuniaki, S., Nakamura, M., and Sugamura, K. (1993). Proc. Nutl. Acad Sci. U S A 90, 4127-4131. Asao, H., Takesliita, T., Nakamum, M., Nagata, K., and Sugamura, K. (1990).J . Exp. Merl. 171,637-644. Asao, H., Tanaka, N.. Isliii, N., Higuchi, M., Takeshita, T., Nakaniura, M., Shirasawa, T., and Sugamura, K. (1994). FEBS Lett. 351, 201-206. Aschernian, D. P., Migone, T.-S., Friedmann, M. C., and Leonard, W. J. (1997).J. B i d . Chein. 272, 8704-8709. Angustine, J. A., Sutor, S. L., and Abraham, R. T. (1991). Mol. Cell. B i d . 11, 4431-4440. Balasubramanian, S., Chernov-Rogan, T., Davis, A. M., Whitehorn, E., Tate, E., Bell, M. P., Znrawski, G., and Barrett, R. (1995). Znt. Zmmzit~ol.7, 1839-1849. Bamford, R. N., Grant, A. J.. Burton, J. D., Peters, C., Kurys, G., Goldman, C. K., Brennan, J., Roessler, E., and Waldmann, T. A. (1994). Proc. Nutl. Acad. Sci. U S A 91,4940-4944. Beadling, C., Johnson, K. W.. and Smith, K. A. (1993). Proc. Natl. A c [ d Sci. USA 90, 2719-
2723. Beadling, C., Ng, J., Babbage, J. W., and Cantrell, D. (1996). E M B O J . 15, 1902-1913. Beckman, E. M.. Porcelli, S. A,, Morita, C. T., Behar. S. M.. Furlong, S. T., and Brenner, M. B. (1994). Nature 372, 691-694. Begley, C., Burton, J., Tsudo, M., Brownstein, B., Ainbrus, J.. and Waldmann, T. (1990). Leukocyte Res. 14, 263-271. Bennett, A. M., Tang. T. L., Sugimoto, S., Walsh, C. T.. and Neel, B. G. (1994). Proc. Ncitl. Acad. Sci. USA 91, 7335-7339. Bierer, B., Mattila, P., Standaert, R., Herzenberg, L., Burakoff, S., Crabtree, G., and Sclireiber, S. (1990). Proc. Notl. Accirl. Sci. USA 87, 9231-9235. Blackman, M. A,, Tigges, M. A,, Minie, M. E., and Koshland, M. E. (1986).Cell 47,609-617. Boehme, S. A,, and Lenardo, M. J. (1993). E w /. Itr~minnol,23, 1552-1560. Bogue, M., and Roth, D. B. (1996). Cicrr. Opiti. Itnttuitio1. 8, 175-180. Boise, L. H., Gonzalez-Garcia, M.. Postema, C. E., Ding, L., Lindsten, T., Turka. L. A.. Mao, X., Nunez, G., and Thompson, C. B. (1993). Cell 74, 597-608. Boise, L. H., Minn, A. J., June, C. H., Lindsten, T.. and Thompson, C. B. (199%). Proc. Natl. Acnrl. Sci. USA 92, 5491-5495. Boise, L. H., Minn, A. J., Noel, P. J., June, C. H., Accavitti, M. A,, Lindsten, T., and Thompson, C. B. (1995b). Znimurtity 3, 87-98. Bosco, M. C., Espinoza-Delgado, I., Schwahe, M., Russell, S. M., Leonard, W. J., Longo, D. L., and Varesio, L. (1994). Blood 83, 3462-3467. Boussiotis, V. A,, Barber, D. L., Nakarai, T., Freeman, G. J., Gribben, J. G., Bernstein, G. M., D'Andrea, A. D., Ritz, J., and Nadler, L. M. (1994). Science 266, 1039-1042. Bradbury, A,, Belt, K. T., Neri, T. M., Milstein, C., and Calabi, F. (1988). E M B O /. 7, 3081-3086. Brennan, P., Babbage, J. W., Burgering, B., Croner, B., Reif, K.. and Cantrell, D. A. (1997). Inzmziriity 7, 679-689. Brown, E. J., Bed, P. A,, Keith. C. T.. Chen, J., Shin, T. B., and Schreiber, S. L. (1995). Nature 377, 441-446. Bninn, G., Hudson, C., Sehilic, A,, Williams, J., Hosoi, H., Houghton, P., Lawrence, J. J., and Abraham, R. (1997). Scierice 277, 99-101. Brunn, G. J., Falls, E. L., Nilson, A. E., and Abraham, R. T. (1995). J . Biol. Chenl. 270, 11628-11635. Brunn, G. J., Williams, J., Sabers, C., Wiederrecht, G., Lawrence, J. C., and Abraham, R. T. (1996). E M B O 1. 15,5256-5267.
68
BRAD 13. NELSON AND DENNIS M . WILLERFORD
Brunner, T., Mogil, R. J., LaFace, D., Yoo, N . J., Mahboubi, A,, Echeverri, F., Martin, S. J., Force, W. R., Lynch, D. H., Ware, C. F.,and Green, D. R. (1995).Nature 373,441-4. BucMey, R. H., Schiff, R. I., Schiff, S. E., Markert, M. L., Williams, L. W., Harville, T. O., Roberts, J. L., and Puck, J. M. (1997).J. Pediat. 130, 378-387. Burns, L. A., Karnitz, L. M., Sutor, S. L., and Abrahham, R. T. (1993). J. Bid. Chem. 268, 17659-17661. Calabi, F., and Milstein, C. (1986). Nature 323, 540-543. Calvo, V., Crews, C. M., Vik, T. A., and Bierer, B. E. (1992). Proc. Nutl. Acad. Sci. USA 89, 7571-7575. Candeias, S., Muegge, K., and Dunun, S. K. (1997). lmmunity 6, 501-508. Cantrell, D. A,, and Smith, K. A. (1983).J. Exp. Med. 158, 1895-1911. Cantrell, D. A,, and Smith, K. A. (1984). Science 224, 1312-1316. Cao, X., Kozak, C. A,, Liu, Y.-J., Noguchi, M., O’ConneIl, E., and Leonard, W. J. (1993). Proc. Nutl. Acud. Sci. USA 90, 8464-8468. Cao, X., Shores, E. W., Hu-Li, J., Anver, M. R., Kelsals, B. L., Russell, S. M., Drago, J., Noguchi, M., Grinberg, A,, Bloom, E. T., Paul, W. E., Katz, S. I., Love, P. E., and Leonard, W. J. (1995). Immunity 2, 223-238. Carlyle, J. R., Michie, A. M., Furlonger, C., Nakano, T., Lenardo, M. J., Paige, C. J., and Z’uniga-Pflucker,J. C. (1997).J . Exp. Med. 186, 173-182. Carson, W. E., Fehniger, T. A., Haldar, S., Eckhert, K., Lindemann, M. J., Lai, C. F., Croce, C. M., Baumann, H., and Caligiuri, M. A. (1997).J. C h . Invest. 99, 937-943. Carson, W. E., Gin, J. G., Lindemann, M. J.. Linett, M. L., Ahdieh, M., Paxton, R., Anderson, D., Eisenmann, J., Grabstein, K., and Caligiuri, M. A. (1994).J.Exp. Med. 180,1395-1403. Casey, L., Lichtman, A., and Boothby, M. (1992).J. lmmunol. 148, 3418-3426. Cerdan, C., Martin, Y., Courcoul, M., Mawas, C., Birg, F., and Olive, D. (1995).J. lmmunol. 154, 1007-1013. Chan, A. C., van Oers, N. S. C., Tran, A,, Turka, L., Law, C.-L., Ryan, J. C., Clark, E. A., and Weiss, A. (1994).J. Immunol. 152, 4758-4766. Chao, D. T., Linette, G. P., Boise, L. H., White, L. S., Thompson, C. B., and Korsmeyer, S. J. (1995).J . Exp. Med. 182, 821-828. Chatila, T., Castigli, E., Pahwa, R., Pahwa, S., Chirmule, N., Oyaizu, N., Good, R. A,, and Geha, R. S. (1990). Proc. Nutl. Acud. Sci. USA 87, 10033-10037. Chazen, G. D., Pereira, G. M. B., Le Gros, G., Gillis, S., and Shevach, E. M. (1989).Proc. Nat. Acud. Sci. USA 86,5923-5927. Chen, J., Ma, A,, Young, F., and Alt, F. W. (1994). lnt. lmnzunol. 6, 1265-8. Chen, M., Cheng, A., Chein, Y.-C., Hymel, A,, Hanson, E. P., Kimmel, L., Minami, Y., Taniguchi, T., Changelian, P. S., and O’Shea, J. J. (1997a). Proc. Natl. Acud. Sci. USA 94, 6910-6915. Chen, Y.-H., Chiu, N. M., Mandal, M., Wang, N., and Wang, C.-R. (1997b). Immunity 6,459-457. Chinnaiyan, A. M., and Dixit, V. M. (1997). Semin. lmmunol. 9, 69-76. Chou, M., and Blenis, J. (1995). Cum: Opin. Cell Bid. 7, 806-814. Chung, J., Kuo, C., Crabtree, G. R., and Blenis, J. (1992). Cell 69, 1227-1236. Cohen, P. L., and Eisenberg, R. A. (1991). Annu. Rev. Immunol. 9, 243-269. Conley, M. E. (1992).Annu. Rev. lmmunol. 10, 215-138. Conley, M . E., Lavoie, A,, Briggs, C., Brown, P., Guerra, C., and Puck, J. M. (1988). Proc. Nutl. Acad. Sci. USA 85, 3090-3094. Conlon, P. J., Morrissey, P. J., Nordan, R. P., Grabstein, K. H., Prickett, K. S., Reed, S . G., Goodwin, R., Cosman, D., and Namen, A. E. (1989). Blood 74, 1368-1373.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
69
Corcoran, A. E., Smart, F. M., Cowling, R. J., Croinpton, T., Owen, M. J., and Venkitaraman, A. R. (1996). EMBO J. 15, 1924-1932. Cornelis, S., Fache, I., van der Heyden, J.. Guisez, Y., Tavernier, J., Devos, R., Fiers, W., and Plaetinck, G. (1995). Eur. J . Immunol. 25, 1857-1864. Cosman, D., Ceretti, D. P., Larsen, A., Park, L., March, C., Dower, S., Gillis, S., and Urdal, D. (1984). Nature 312, 768-771. Cousens, L. P., Orange, J. S., and Biron, C. B. (1995).J Zmniunol. 155, 5690-5699. Crabtree, G. R. (1989). Science 243, 355-361. Crawley, J. B., Rawlinson, L., Lali, F. V., Page, T. H., Saklatxala, I., and Foxwell, B. M. (1997).1.Biol. Chem. 272, 15023-15027. Critchfield, J. M., Racke, M. K., Zuniga-Pflucker,J. C., Cannella, B., Raine, C. S., Goverrnan, J., and Lenardo, M. J. (1994). Science 263, 1139-43. Crompton, T., Outram, S. V., Buckland, J., and Owen, M. J. (1997). Eur. J . Immunol. 27, 100-104. Cunningham, B. C., Ultsch, M., devos, A. M., Mulkerrin, M. G., Clauser, K. R., and Wells, J. A. (1991). Science 254, 821-825. Cutler, R. L., Liu, L., Darnen, J. E., and Krystal, G. ( 1 9 9 3 ) .Bid. ~ Chein. 268,21463-21465. Darnell, J. E., Jr. (1997). Science 277, 1630-1635. Depper, J. M., Leonard, W. J., Robb, R. J., Waldmann, T. A,, and Greene, W. C. (1983). J . Zintnunol. 131, 600-606. de Saint-Bade, G . D., Arveiler, N., Oberle, I., Malcolm, S., Levinsky, R. J., Lau, Y. L., Hofker, M., Debre, M., Fischer, A,, Greicelli, C., and Mandel, J. L. (1987). Proc. Natl. Acad. Sci. USA 84, 7576-7579. deVos, A. M., Ultsch, M., and Kossiakoff, A. A. (1992). Science 255, 306-312. Dhein, J., Walczak, H., Baumler, C., Debatin, K. M., and Krammer, P. H. (1995).Nature 373,438-441. DiSanto, J. P. (1997). Curr. Biol. 7, R424-R426. DiSanto, J. P., Certain, S., Wilson, A,, MacDonald, H. R., Avner, P., Fischer, A,, and de Saint Bade, G. (1994). Eur. J . Ztnm~unol.24, 3014-3018. DiSanto, J. P., Guy-Grand, D., Fisher, A., and Tarakhovsky, A. (1996). J Exp. Med. 183, 1111-1118. DiSanto, J. P., Muller, W., Guy-Grand, D., Fischer, A., and Rajewsky, K. (1995). Proc. Natl. Acad. Sci. USA 92, 377-381. Djeu, J. Y., Liu, J. H., Wei, S., Rui, H., Pearson, C. A,, Leonard, W. J., and Blanchard, D. K. (1993).J . Zmmunol. 150, 960-970. Doi, S., Saiki, O., Tanaka, T., Ha-Kawa, K., Igarashi, T., Fujita, T., Taniguchi, T., and Kishimoto, S. (1988). Clin. Zmmimol. Zmmunopathol. 46, 24-36. Drappa, J., Vaishnaw, A. K., Sullivan, K. E., Chu, J. L., and Elkon, K. B. (1996). N . Engl. J Med. 335, 1643-1649. Dnkovich, M., Wano, Y., thi Bich Thuy, L., Katz, P., Cullen, B., Kehri, J., and Green, W. (1987). Nature 327, 518-522. Dumont, F.,J, Stanch, M. J., Koprak, S. L., Melino, M. R., and Sigal, N. (1990).J.Zmmunol. 144, 251-258. Duronio, V.. Welham, M. J., Abraham, S., Dryden, P., and Schrader, J. W. (1992). Proc. Natl. Acad. Sci. USA 89, 1587-1,591. Endo, T., Masuhara, M., Yokouchi, M., Suzuki, R., Sakamoto, H., Mitsui, K., Matsurnoto, A., Tanimura, S., Ohtsubo, M., Misawa, H., Miyazaki, T., Leonor, N., Taniguchi, T., Fujita, T., Kanakura, Y., Komiya, S., and Yoshirnura, A. (1997). Nature 387, 921-924. Epling-Biirnette, P. K., Wei, S., Liu, J., H., Pericle, F., Ussery, D., Russell, S. M., Leonard, W. J., and Djeu, J. Y. (1995). Ear. J . Zmmunol. 25, 291-294.
70
BRAD H. NELSON AND DENNIS M. WILLERFORD
Espinoza-Delgado, I., Ortaldo, J.,Winkler-Pickett, R., Sugamura, K., Varesio, L., and Longo, D. (1990).]. Exp. Med. 171, 1821-1826. Evans, G. A., Goldsmith, M. A,, Johnston, J. A,, Xu, W., Weiler, S. R., Erwin, R., Howard, 0. M. Z., Abraham, R. T., O’Shea, J. J., Greene, W. C., and Farrar, W. L. (1995).J. B i d . Chem. 270,28858-28863. Fanger, G., Gerwins, P., Widmann, C., Jarpe, M. B., and Johnson, G. L. (1997). Cum. O p i r i . Genet. Dev. 7, 67-74. Ferrari, S., Pearson, R., Siegmann, M., Kozma, S., and Thornas, G. (1993).1. B i d . Chem. 268, 16091-16094. Firpo, E. J., KofF, A,, Solomon, M. J., and J. M., R. (1994).Mol. Cell. B i d . 14,4889-4901. Fischer, A,, Cavazzana-Calvo, M., De Saint-Basile, G., DeVillartay, J. P., DiSanto, J. P., Hivroz, C., Rieux-Laucat, R., and Le Deist, F. (1997).Annu. Reu. Immunol. 15,93-124. Fisher, G. H., Rosenberg, F. J., Straus, S. E., Dale, J. K., Middleton, L. A,, Lin, A. Y., Strober, W., Lenardo, M. J., and Puck, J. M. (1995). Cell 81, 935-946. Frank, D. A,, Robertson, M. J,, Bonni, A,, Ritz, J., and Greenberg, M. E. (1995). Proc. Natl. Acad. Sci. USA 92, 7779-7783. Franke, T. F., Yang, %I., Chan, T. O., Datta, K., Kazlauskas, A,, and Tsichlis. P. N. (1995). Cell 81, 727-736. Friedinam, M. C., Migone, T. S., Russell, S. M., and Leonard, W. J. (1996).Proc. Natl. Acad. Sci. USA 93, 2077-2082. Friend, S. L., Hosier, S., Nelson, A,, Foxworthe, D., Williams, D. E., and Farr, A. (1994). Exp. Heinntol. 22, 321-328. Fujii, H., Nakagawa, Y., Schindler, U., Kawahara, A., Mori, H., Gouilleux, F., Groner, B., Ihle, J. N., Minami, Y., Miyazaki, T., et al. (1995). Proc. Natl. Acad. Sci. USA 92,5482-5486. Fukada, T., Hibi, M., Yamanaka, Y., Takahashi-Tezuka, M ., Fujitani, Y., Yamaguchi, T., Nakajirna, K., and Hirano, T. (1996). Immunity 5, 449-460. Gaffen, S. L., Lai, S. Y., Ha, M., Liu, X., Hennighausen, L., Greene, W. C., and Goldsmith, M. A. (1996).J. B i d . Chem. 271, 21381-21390. Gaffen, S. L., Lai, S. Y., Xu, W., Gouilleux, F., Groner, B., Goldsmith, M. A., and Greene, W. C. (1995). Proc. Natl. Acad. Sci. USA 92, 7192-7196. Gellert, M. (1997).Adv. Immunol. 64, 39-64. Gillis, S., Ferin, M. M., Ou, W., and Smith, K. A. (1978).J. Immzinol. 120, 2027-2032. Gillis, S., and Smith, K. A. (1977a).J. Exp. Med. 146, 468-482. Gillis, S., and Smith, K. A. (1977b). Nature 268, 164-166. Gillis, S., and Watson, J. (1981).J Immunol. 126, 1245-1248. Gilmour, K., Pine, R., and Reich, N. C. (1995).Proc. Natl. Acnd. Sci. USA 92,10772-10776. Gin, J., Kumaki, S., Ahdieh, M., Friend, D. J., Loomis, A., Shanebeck, K., DuBose, R., Cosman, D., Park, L. S., and Anderson, D. M. (1995). E M B O J. 14, 3654-3663. Gin, J. G., Ahdieh, M., Eisenman, J., Shaneheck, K., Grabstein, K., Kumaki, S., Narnen, A., Park, L. S., Cosman, D., and Anderson, D. (1994). E M B O J. 13, 2822-2830. Godfrey, D. I., and Zlotnick, A. (1993). Immunol. Today 14, 547-552. Goldsmith, M. A., Lai, S. Y., Xu, W., Amaral, M. C., Kuczek, E. J., Parent, L. J., Mills, G. B., Tarr, K. L., Longmore, G. D., and Greene, W. C. (1995).J.Bid. Chem. 270,21729-21737. Gonzalez-Garcia, A., Merida, I., Martinez-A,, C., and Carrera, A. C. (1997).J. Biol. Chem. 272, 10220-10226. Coodnow, C. C., Cyster, J. G., Hartley, S. B., Bell, S. E., Cooke, M. P., Healy, J. I., Akkaraju, S., Rathmell, J. C., Pogue, S. L., and Shokat, K. P. (1995). Adv. Immunol. 59, 279-358. Gougeon, M. L., Drean, G . , Le Deist, F., Dousseau, M., Fevrier, M., Diu, A,, Theze, J., Griscelli, C., and Fischer, A. (1990).J. Immunol. 145, 2873-2879.
BIOLOGY OF THE INTERLEIJKIN-2 RECEPTOR
71
Grabstein, K. H., Eisenman, J., Shanebeck, K., Rauch, C., Srinivasan, S., Fung, V., Beers, C.. Richardson, J., Schoenbom, M. A,, Ahdieh, M., Johnson, L., Alderson. M. R., Watson. J. D., Anderson. D. M., and Giri, J. G. (1994). Science 264, 965-968. Grabstein, K. H.. Waldschmidt, T. J., Finkdinan, F. D., Hess, B. W., Alpert, A. R., Boiani, N. E., Nainen, A. E., and Morrissey, P. J. (1993).J . Exp. Med. 178, 257-264. Grant, A. J., Roessler, E., JLI. G., Tsudo, M., Sugainura, K.. and Waldmann, T. A. (1992). Proc. N d Acad. Sci. U S A 89, 2165-2169. Graves, J. D., Downward, J., Izquierdo-Pastor, M., Rayter, S., Wanme, P. H., and Cantrell, D. A. (1992).J.Zntnmnol. 148, 2417-2422. Green, D. R., arid Scott, D. W. (1994). Czirr. Opin. Iirimi~nol.6, 476-487. Gumperz, J. E., and Parhain, P. (1995). Nature 378. Gustafson, T. A,, He, W., Craparo, A,, Schaub, C. D., and O’Neill, T. J. (1995).Mol. Cell. B i d . 15, 2500-2508. Gutierrez-Rainos. J. C.. Martinez-A,, C., Kohler, G., and Iglesias, A. (1989). Res. ImirLunol. 140, 661. Harmer, S. L., and DeFranco, A. L. (1997). Mol. Cell B i d . 17, 4087-4095. Hartwell. L. H., and Kastan, M . B. (1994). Science 266, 1821-1828. Hatakeyama, H., Kawahara, A,, Mori, H., Shibuya, H., and Taniguchi, T. (1992). Proc. N d . Acad. Sci. USA 89, 2022-2026. Hatakeyama, M., Kono, T., Kobayashi, N., Kawahara, A,, Leviu, S. D., Perlmutter, R. M., and Taniguchi, T. (1991). Science 252, 1523-1528. Hatakeyama, M., Mori, H., Doi, T., and Tanigiichi, T. (1989a). Cell 59, 837-845. Hatakeyaina, M., Tsudo, M., Minamoto, S., Kono, T., Doi, T., Miyata, T.. Miyasaka, M., and Taniguchi, T. (1989b). Science 244, 551-556. He, Y.-W., and Malek, T. R. (1996).].Exp.Med 184, 289-293. J., Adkins, B., and Malek, T. R. (1997).J . Immicnol. He, Y.-W., Nakajinia, H., Leonard. U’. 158, 2592-2599. He, Y. W., Adkins, B., Furse, R. K., and Malek, T. K. (1995). J. Iminnnol. 154, 1,596-1605. Heldin, C.-H. (1995). Cell 80, 213-22:3. Hibi, M., Murakami, M., Saito, M., IIirano, T., and Taga, T. (1990). Cell 63, 1149-1157. Higuchi, M., Asao, H., Tanaka, N., Oda, K., Takexlmita, T., Nakamura, M., Van Snick, J., and Sugainura, K . (1996). Eur. J. ~itimunol.26, 1322-1327. Horak, I. D., Cress, R. E., Lucas, P. L., Horiik, E. M., Wddmann, T. A., and Bolen, J. €3. (1991). Proc. Natl. Acad. Sci. USA 88, 1991-2000. Hoinung, F., Xheng, L., and Lenardo, M. J. (1997).J. hrrumol. 159, 3816-3822. Hou, J., Schindler. U., I-Ienzel, W. J.. Wnng, S. C., and McKnight, S. L. (1995). ltnintcnity 2, 321-329. Howard, M., Nakanishi. K., and Paul, W. E. (1984). I n m u n d . Reu. 78, 185-210. Huber, A.-O., Zornig, M., Lyon, D., Suda, T., Nagata, S., and Evan, G. I. (1997). Science 278, 1305-1308. Hunter, T., and Pines. J. (1994). Cell 79, 573-582. Ihle, J. (1996a). Plzilos. Trms. Royal Soc. 351, 159-166. Ihle. J. N . (1995). Nature 377, 591-594. Ihle, J. N. (l996b).Cancer Res. 68, 23-65. Imier, 1.-L., Miyajima, A,, and Zurawsk, G. (1992). EMBO J. 11, 2047-2053. Irnmler, M., Thoine, M., Hahne, M.. Schneider, P., Hofinann, K., Steiner. V., Rodiner, J. L., Sclrroter, M., Burns, K., Mattlmxinn. C., Rimoldi, D.. French, L. E., and Tschopp, J. (1997). Nature 388, 190-195. Ishida, Y., Nishi, M., Taguchi, O., Inaba, K., Minato, M., Kawaichi, M.. and Honjo, T. (1989).h t . In~t?i?tnol.1, 113-120.
72
BRAD H. NELSON AND DENNIS M. WILLERFORD
Izquirdo, M., Downward, J., Otani, H., Leonard, W. J., and Cantrell, D. A. (1992). Eur. J. Zmmunol. 22, 817-821. Jenkinson, E. J., Kingston, R. T., and Owen, J. J. T. (1987). Nature 329. John, S., Reeves, R. B., Lin, J. X., Child, R., Leiden, J. M., Thompson, C. B., and Leonard, W. J. (1995). Mol. Cell. Biol. 15, 1786-1796. John, S., Robbins, C. M., and Leonard, W. J. (1996). E M B O J . 15, 5627-5635. Johnston, J. A,, Bacon, C. M., Finbloom, D. S., Rees, R. C., Kaplan, D., Shibuya, K., Ortaldo, J. R., Gupta, S., Chen, Y. Q., Gin, J. D., and O’Shea, J. J. (1995a). Proc. Natl. Acad. Sci. USA 92, 8705-8709. Johnston, J. A,, Kawamura, M., Kirken, R. A., et at. (1994). Nature 370, 151-153. Johnston, J. A., Wang, L.-M., Hanson, W. P., Sun, X.-J., White, M. F., Oakes, S. A., Pierce, J. H., and O’Shea, J. J. (1995b). J. Biol. Chem. 270, 28527-28530. Joneja, N., and Wojchowski, D. M. (1997). J. Biol. Chem. 2772, 11176-11184. Jones, L.A., Chin, L. T., Longo, D. L., and Kruisbeek, A. M. (1990).Science250,1726-1729. Ju. S. T., Panka, D. J., Cui, H., Ettinger, R., el-Khatib, M., Sherr, D. H., Stanger, B. Z., and Marshak-Rothstein, A. (1995). Nature 373, 444-8. Jung, L. K. L., Hara, T., and Fu, S. M. (1984). J. Exp. MecL 160, 1597-1602. Kaplan, M. H., Schindler, U., Smiley, S. T., and Grusby, M. (1996a). Immunity 4,313-319. Kaplan, M. H., Sun, Y.-L., Hoey, T., and Grusby, M. J. (1996b). Nature 382, 174-177. Karnitz, L., Bums, L. A,, Sutor, S. L., Blenis, J., and Abraham, R. T. (1995). Mol. Cell. Biol. 15,3049-3057. Kamitz, L. M., and Abraham, R. T. (1996). Adv. Zmmunol. 61, 147-199. Kamitz, L. M., Sutor, S. L., and Abraham, R. T. (1994).J. Exp. Med 179, 1799-808. Kavanaugh, M. M., and Williams, L. T. (1994). Science 266, 1862-1865. Kawabe, Y., and Ochi, A. (1991). Nature 349, 245-248. Kawahara, A,, Minami, Y., Miyazaki, T., Ihle, J. N., and Taniguchi, T. (1995). Proc. Natl. Acad. Sci. USA 92, 8724-8728. Kawahara, A,, Minami, Y., and Taniguchi, T. (1994). Mol. Cell. B i d . 14, 5433-5440. Kawamura, M., McVicar, D. W., Johnston, J. A,, Blake, T. B., Chen, Y,-Q., Lal, B. K., Lloyd, A. R., Kelvin, K. J., Staples, J. E., ortaldo, J. R., and OShea, J. J. (1994). Proc. Natl. Acad. Sci. USA 91, 6374-6378. Kearney, E. R., Pape, K. A., Loh, D. Y., and Jenkins, M. K. (1994). Immunity 1,327-39. Keegan, A. D., Nelms, K., White, M., Wang, L.-M., Pierce, J. H., and Paul, W. E. (1994). Cell 76, 811-820. Khoruts, A., Mondino, A., Pape, K. A., Reiner, S. L., and Jenkins, M. K. (1998). J. Exp. Med. 187,225-236. Kim, Y. H., Buchholz, M. J., and Nordin, A. A. (1993). Zmmunology 90,3187-3191. Kimura, Y., Takeshita, T., Kondo, M., Ishii, N., Nakamura, M., Van Snick, J., and Sugamura, K. (1995). Znt. Znamunol. 7, 115-120. Kirberg, J., Bems, A., and von Boehmer, H. (1997). 1. Exp. Med. 186, 1269-1275. Kirken, R. A., Malabarba, M. G., Xu, J., DaSilva, L., Ewin, R. A,, Liu, X., Hennighausen, L., Rui, H., and Farrar, W. L. (1997).J. Biol. Chem. 272, 15459-15465. Kisielow, P., and von Boehmer, H. (1995). Adv. Zmmunol. 58, 87-209. Klingmuller, U., Wu, H., Hsiao, J,, Toker, A., Duckworth, B. C., Cantley, L. C., and Lodish, H. (1997). Proc. Natl. Acad. Sci. USA 94, 3016-3021. Kneitz, B., Herrmann, T., Yonehara, S., and Schimpl, A. (1995).Eur.J. Zmmunol. 25,25722577. Kobayashi, N., Kono, T., Hatakeyama, M., et al. (1993). Biochemistry 90, 4201-4205. Koch, C. A., Anderson, D., Moran, M. F., Ellis, C., and Pawson, T. (1991). Science 252, 668-674.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
73
Kondo, M., Ohashi, Y., Tada, K., Nakamnra, M., and Sugamura, K. (1994a).Eur. f. I?nnzunol. 24,2026-2030. Kondo, M., Takeshita, T., Higuchi, M., Nakamura, M., Sudo, T., Nishikawa, S., and Sugamura, K. (1994b). Science 263, 1453-1454. Kondo, M., Takeshita, T., Ishii, N., Nakamura, M., Watanabe, S., Arai, K., and Sugamura, K. (1993). Science 262, 1874-1877. Kramer. S., Mamalaki, C., Horak, I., Schimpl, A,, Kioussis. D., and Hunig, T. (1994).Eur. f. Immn.mol. 24, 2317-2322. Kramer, S., Schimpl, A.. and Hunig, T. (1995).f . Exp. Med. 182, 1769-1786. Kromer, G.. de Cid, R., Moreno de Ahoran, I., Conzalo, J.-A., Iglesias, A., Marinez-A,, C., and Gutierrez-Ramos, J. C. (1991). I t r m i u r w ~ .Rev. 122, 173-204. Kundlg, T. M., Schorle, H., Bachmann, M. F., Hengartner, H., Zinkernagel, R. M., and Horak, I. (1993). Science 262, 1059-1061. Kuo, C. J., Chung, J., Fiorentino, D. F., Flanagan, W. M., Blenis, J., and Crabtree, G. R. (1992). Natirre 358, 70-73. Kyriakis, J. M., and Avruch, J. (1996).f. B i d . Chern. 271, 24313-23316. Lecine, P., Algart’e, M., Rameil, P., Beadling, C., Bucher, P., Nabholz, M., and Imbert, J. (1996). Mol. Cell. B i d . 16, 6829-6840. Lam, K. P., Kuhn, R., and Rajewsky, K. (1997). Cell 90, 1073-108.3. Lane, H. A,, Fernandez, A,, Lamb, N. J. C., and Thomas, G. (1993).Nature 363, 170-172. Lenardo, M. J. (1991). Nature 353, 858-61. Leonard, W. J., Depper, J. M., Crabtree, G. R., Rudilaff, S . , Pumphrey, J., Robb, R. J.. Kronke, M., Svetlik, P. B., Peffer, N. J., Waldmann, T. A,, and Greene, W. C. (1984). Nature 311, 626-631. Leonard, W. J., Depper, J. M., Uchiyaina, T., Smith, K. A,, Waldmann, T. A,, and Greene, W. C. (1982). Nature 300, 267-269. Leonard, W. J., Noguchi, M., Russell, S. M., and McBride, 0. W. (1994). Zm?nnnol. Rec. 138,61-86. Li, W., Nishimura, R., Kashishian, A,, Batzer, A. G., Kim, W. J. H., Cooper, J. A,, and Schlessinger, J. (1994).Mol. Cell. B i d . 14, 509-517. Lin, B. B., Cross, S. L., Halden, N. F., Roman. D. G., Toledano, M. B., and Leonard, W.-J. (1990). Mul. Cell. B i d . 10, 850-853. Lin, J. X., Bhat, N. K., John, S., Queale, W. S . , and Leonard, W. J. (1993). Mol. Cell. Biol. 13, 6201-6210. Lin, J. X., and Leonard, W. J. (1997). Mol. Cell. Bid. 17, 3714-3722. Lin, J. X., Migone, T. S., Tsang, M.,Friedmann, M., Weatherbee, J. A,, Zhou, L., Yamauchi, A., Bloom, E. T., Mietz, J., John, S., and Leonard, W. J. (1995).Immunity 2, 331-339. Liu, C. C., Joag, S. V., Kwon, B. S . , and Young, J. D. (IUUO).]. Znzniunol. 144, 1196-1201. Liu. J. H., Wei, S., Ussery, D., Epling-Bnrnette. P. K., Leonard, W. J., and Djiu, J. Y. (1994). Bbod 84, 3870-3875. Liu. Y., and Janeway, C. A,, Jr. (1990).J . Exp. Med. 172, 1735-1739. Lord, J. D., McIntosh, B. C., Greenberg, P. D., and Nelson, B. H. (1998). Submitted for publication. Loughnan, M. S., and Nossal, G. J. (1989). Nature 340, 76-79. Lowenthal, J. W., Howe, R. C., Ceredig, R., and MacDonald. H. R. (1986).f. Znmunol. 137, 2579-2584. Ma, A,, Datta, M., Margosian, E., Chen, J., and Horak, I. (1995a).J. E x p Med. 182,15671572. Ma. A,, Pena, J. C., Chang, B., Margosian, E., Davidson, L., Alt, F. W., and Thompson, C. B. (1995b). Proc. N d . Aecirl. Sci. USA 92, 4763-4767.
74
BRAD H. NELSON AND DENNIS M. WILLERFORD
Macchi, P., VIlla, A,, Giliani, S., Sacco, M. G., Frattini, A., Porta, F., Ugazio, A. G . ,Johnston, J. A,, Candotti, F., O’Shea, J. J., Vezzoni, P., and Notarangelo, L. D. (1995). Nature 377,65-68. MacDonald, H. R., Baschieri, S., and Lees, R. K. (1991). Eur. 1.Immunol. 21, 1963-1966. Makrigiannis, A. P., and Hoskin, D. W. (1997).J. Immunol. 159, 4700-4707. Marais, R., and Marshall, C. (1996). Cancer Sum. 27, 101-125. Maraskovsky, E., Chen, W. F., and Shortman, K. (1989). J. bnmunol. 143, 1210-1214. Marshall, C. J. (1994). Cuw. Opin. Genet. Deu. 4, 82-89. Martin, L. H., Calabi, F., and Milstein, C. (1986).Proc. Natl. Acad. Sci. USA 83,9154-9158. Martin, S. J., and Green, D. R. (1995). CelE 82, 349-352. Matsumoto, A,, Masuhara, M., Mitsui, K., Yokouchi, M., Ohtsubo.M., Misawa, H., Miyajima, A,, and Yoshimura, A. (1997). Blood 89, 3148-3154. Matsuoka, M., Takeshita, T., Ishii, N., Nakamura, M., Ohkubo, T., and Sugamura, K. (1993). Eur. 1.Immunol. 23, 2472-2476. Mendiratta, S. K., Martin, W. D., Hong, S., Boesteanu, A., Hoyce, S., and Kaer, L. V. (1997). Immunity 6, 469-477. Merida, I., Diez, E., and Gaulton, G. N. (1991). J. Immunol, 147, 2202-2207. Merida, I., Williamson, P., Kuziel, W. A,. Greene, W. C., and Gaulton, G. N. (199.3). J . Biol. Cheni. 268, 6765-6770. Meuer, S. C., Hussey, R. E., Cantrell, D. A,, Hodgdon, J. C., Schlossman, S. F., Smith, K. A,, and Reinherz, E. L. (1984). Proc. Natl. Acad. Sci. USA 81, 1509-1513. Minami, Y., Kono, T., Yamada, K., et nl. (1993). EMBOJ. 12, 759-768. Minami, Y., Nakagawa, Y., Kawahara, A,, Miyazaki, T., Sada, K., Yamamura, H., and Taniguchi, T. (1995). Zmmunity 2, 89-100. Miyazaki, T., Kawahara, A,, Fujii, H., Nakagawa, Y., Minami, Y., Liu, Z. J., Oishi, I., Sibennoinen, O., Witthuhn, B. A,, Ihle, J. N., et nl. (1994). Science 266, 1045-1047. Miyazaki, T., Liu, Z. J., Kawahara, A,, Minami, Y., Yamada, K., Tsujimoto, Y., Barsoumian, E. L., Permutter, R. M., and Taniguchi, T. (1995). Cell 81, 223-31. Monfar, M., Lemon, K. P., Crammer, T. C., et al. (1995). Mol. Cell. Biol. 15, 326-337. Moreau, J.-L., Bossus, M., De Groote, D., Francois, C., Jacques, Y., Tartar, A., and Theze, J. (1995). Mol. Inzmunol. 32, 1047-1056. Morgan, D. A., Ruscetti, F., and Gallo, R. (1976). Science 193, 1007-1008. Mori, j., Barsoumaian, E. L., Hatakeama, M., and Taniguchi, T. (1991). Int. Inimunol. 3, 149-156. Morice, W. G., Brunn, G. J., Wiederrecht, G., Siekierka, J. J., and Abraham, R. T. (19934. J Biol Chein 268, 3734-3738. Morice, W. G., Wiederrecht, G., Brunn, G. J.. Siekierka, J. J., and Abraham, R. T. (1993bj. J Biol Chem 268,22737-22745. Monisey, P. J., Goodwin, R. G., Nordan, R. P., Anderson, D., Grabstein, K. H., Cosman, D., Sims, J., Lupton, S., Acres, B., Reed, S. G., Mochizuki, D., Eisenman, J., Conlon, P. J., and Namen, A. E. (1989).1.Exp. Med. 169, 707-716. Mr’ozek, E., Anderson, P., and Caligiuri, M. A. (1996). Blood 87, 2362-2340. Muegge, K., Vila, M. P., and Durum, S. K. (1993). Science 261, 93-95. Murakami, M., Hibi, M., Nakagawa, N., Nakagawa, T., Yasukawa, K., Yamanishi, K., Taga, T., and Kishimoto, T. (1993). Science 260, 1808-1810. Murphy, K. M., Heimberger, A. B., and Loh, D. Y. (1990). Science 250, 1720-1723. Musso, T., Johnston, J. A., Linnekin, D., Varesio, L., Rowe, T. K., O’Shea, J. J., and McVicar, D. W. (1995).].Exp,Med. 181, 1425-1431. Myers, M. G. J., Zhang, Y., Aldaz, G. A,, Crammer, T., Glasheen, E., Yenush, L., Wang, L. M., Sun, X. J.. Blenis, J., Pierce, J. H., and White, M. F. (1996). Mol. Cell. Biol. 16,4147-4155.
BIOLOGY OF THE INTERLEUKIN-2 RECEPTOR
75
Nabholz. M., Soldiani. E., Sperisen, P., Pla, M., Wang, S. M., MacDonald, H. R., Heichenbach, P., Beennann, F., and Buclier, P. (1995). I?nttittnobiology 193, 259-262. Nagata, S. (1994).Ado. I ? n t w n d . 57, 129-144. Naka, T., Narazaki, M., Hirata, M., Matsumoto, T., Minamoto, S., Aono, A,, Nishimoto, N., Kajita, T., Taga, T., Yoshizaki, K., Akira, S., and Kishinioto, T. (1997). Nuture 387, 924-928. Nakajima, H., and Leonard, W. J. (1997).J h ~ t i u t i u l ,159, , 4737-4744. Nakajima, H., Lie, X.-W., Wynshaw-Boris. A,. Rosenthal, L. A., Imada, K., Finblooin, D. S., Hennighausen. L., and Leonard, W. J. (1997a). Z ~ i ~ ~ ~7,f 691-701. f~i~y Nakajii-na, H., Shores, E. W., Noguchi, M., and Leonard, W. J. (1997b). J. Exp. Med. 185, 189-195. Nakamura, N., Chin, H., Miyasaka, N., and Miura, 0. (1996)./. B i d . Chem. 271, 1948319488. Nakamura, Y., Russell, S. M., Mess, S. A,, Fiiedniann, M., Erdos, M., Francois, C., Jacques, Y., Adelstein, S., a i d Leonard, W. J. (1994). Nature 369, 330-333. Nakanishi, K., Cohen, D. I., Blackman, M., Nielsen, E., Ohara, J., Hamaoka, T., Koshland, M. E., and Paul, W. E. (1984a).J. Exp. Med. 160, 1736-1751. Nakanishi, K., Malek, T. R., Smith, K. A,, Hamaoka, T., Shevach, E. M., and Paul, W. E. (1984b).J. Exp. Med. 160, 1605-1621. Nakayania, K., Ishida, N., Shirane, M., Inomata, A,, Inoue, T., Shishido, M., Horii, I., Loh, D. Y., and Nakayama, K.-I. (1996). Cell 85, 707-720. Namen, A. E., Lupton, S., Hjerrild, K., Wignall, J., Mochizuki, D. Y., Schmierer, A., Mosley, B., March, C. J., Urdal, D., and Gillis, S . (1988). Nature 333, 571-573. Nelson, B. H., Lord, J. D., and Greenberg, P. D. (1994). Nature 369, 333-336. Nelson, B. H., Lord, J. D., and Greenberg, P. D. (1996). M d . Cd1. B i d . 16, 309-317. Nelson, B. H., McIntosh, B. C., Rosencrans, L. L., and Greenberg, P. D. (1997). Proc. Natl. Acad. Sci. USA 94, 1878-1883. Nikaido, T., Shimizu, A., Ishida, N., Sabe, H., Teshigawara, K., Maeda, M., Uchiyama, T., Yodoi, J., and Honjo, T. (1984). Nature 311, 631-635. Nishi, M., Ishida, Y., and Honjo, T. (1988). Nature 331, 267-269. Noguchi, M., Adelstein, S., Cao, X., and Leonard, W. J. (1993a)./. B i d . C h i . 268, 1360113608. Noguchi, M., Nakamura, Y., Russell, S. M., et nl. (1993b). Science 262, 1877-1880. Noguchi, M., Sarin, A,, Aman, M. J.. Nakajima, H., Shores, E. W., Henkart, P. A,, and Leonard, W. J. (1997). Proc. Natl. Acad. Sci. USA 94, 11534-11539. Noguchi, M., Yi, H., Rosenblatt, H. M., Filipocich, A . H., Adelstein, S., Modi, W. S., Mcbride, 0. W., and Leonard, W. J. (1993~).Cell 73, 147-157. Noguchi, T., Matozaki, T., Horita, K., Fujioka, Y., and Kasuga, M. (1994). M d . Cell. B i d . 14, 6674-6682. Nosaka, T., van Deursen, J. M. A,, Tripp, R. A,, Thierfelder, W. E., Witthuhn, B. A,, McMickle, A. P., Doherty, P. C., Grosveld, G. C., and Ihle, J. N. (1995). Science 270, 800-802. Nourse, J., Firpo, E., Flanagan, W. M., Coats, S., Polyak, K., Lee, M.-H., Massague, J., Crabtree, G. R., and Roberts, J. M. (1994). Nature 372, 570-573. Oakes, S., Candotti, F., Johnston, D. A,, Chen, Y.-Q., Ryan, J. J., Taylor, N., Lie, S., Hennighausen, L., Notarangelo, L. D., Paul, W. E., Blaese, R. M., and O’Shea, J. J. (1996). Immunity 5, 605-61s. Ohbo, K., Suda, T., Hashiyama, M., Mantani, A,, Ikebe, M., Miyakawa, K., Moriyama, M., Nakamura, M., Katsuki, M., Takahashi, K., Yamaura, K.-I., and Sugamura, K. (1996). B h t / 87, 956-967.
76
BRAD H . NELSON AND DENNIS M. WILLERFORD
Ohbo, K., Takasawa, N., Ishii, N., Tanaka, N., Nakamura, M., and Sugamura, K. (1995). J. Biol. Chem. 270, 7479-7486. Ohteki, T., Ito, S., Suzuki, H., Mak, T. W., and Ohashi, P. S. (1997).J.Immunol. 159,59315935. Okazaki, H., Ito, M., Sudo, T., Hattori, M., Kano, S., Katsura, Y., and Minato, N. (1989). J. Immunol. 143, 2917-2922. Pahwa, R., Chatila, T., Pahwa, S., Paradise, C., Day, N. K., Geha, R., Schwartz, S. A., Slade, H., Oyaizy, N., and Good, R. A. (1989). Proc. Nutl. Acud. Sci. USA 86, 5069-5073. Park, S. Y., Saijo, K., Takahashi, T., Osawa, M., Arase, H., Hirayama, N., Miyake, K., Nakauchi, H., Shirasawa, T., and Saito, T. (1995). Imnmunity 3, 771-782. Paul, W. E. (1991). Blood 77, 1859-1870. Pawson, T. (1995). Nature 373, 573-579. Perkins, G. R., J., M., and Collins, M. K. L. (1993). J. Exp. Med. 178, 1429-1434. Pernis, A,, Gupta, S., Yopp, J., Garfein, E., Kashleva, H., Schindler, C., and Rothmaii, P. (1995).J. B i d . Cheni. 270, 14517-14522. Peschon, J. J., Morrissey, P. J., Grabstein, K. H., Ramsdell, F. J., Maraskovsky, E., Gliniak, B. C., Park, L. S., Ziegler, S. F., Williams, D. E., Ware, C. B., Meyer, J. D., and Davison, B. L. (1994).J. Exp. Med. 180, 1955-1960. Plaetinck, G., Combe, M. C., Corthesy, P., Sperisen, P., Kanamori, H., Honjo, T., and Nabholz, M. (1990).J. lmmunol. 145, 3340-3347. Pleiman, C. M., Hertz, W. M., and Cambier, J. C. (1994). Science 263, 1609-1612. Poloskaya,A., Zhao, Y., Lilly, M. B., and Kraft, A. S . (1994).J. Biol. Chem. 269,14607-14613. Prasad, K. V., Janssen, O., Kapeller, R., Raab, M., Cantley, L. C., and Rudd, C. E. (1993). Proc. Natl. Acad. Sci. USA 90, 7366-7370. Price, D. J., Grove,J. R., Calvo,V., Avruch, J., and Bierer, B. E. (1992).Science 257,973-977. Puck, J. M., Deschenes, S. M., Porter, J. C., Dutra, A. S., Brown, C. J., Willard, H. F., and Henthom, P. S. (1993). Hum. Mol. Gen. 2, 1099-1104. Puck, J. M., Nussbaum, R. L., Smead, D. L., and Conley, M. E. (1989).Am. J. Hum. Genet. 44, 724-730. Qin, S., Inazu, T., Yang, C., Sada, K., Taniguchi, T., and Yamamura, H. (1994). F E B S Lett. 345, 233-236. Quelle, F. W., Sato, N., Witthuhn, B. A,, Inhorn, R. C., Eder, M., Miyajima, A., Griffin, J. D., and Ihle, J. N. (1994). Mol. Cell. Biol. 14, 4335-4341. Raff, M. C. (1992). Nature 356, 397-400. Raulet, D. H. (1985). Nature 314, 101-103. Raulet, D. H. (1996). C u m Opin. Immunol. 8, 372-377. Ravichandran, K. S., and Burakoff, S . J. (1994).J.Biol.Chenz. 269, 1599-1602. Ravichandran, K. S., Igras, V., Shoelson, S. E., Fesik, S. W., and Burakoff, S. J. (1996). Proc. Natl. Acad. Sci. USA 93, 5275-5280. Ravichandran, K. S., Lorenz, U., Shoelson, S. E., and Burakoff, S. J. (1995). Mol. Cell. Biol. 15,593-600. Reif, K., Burgering, B. M. T., and Cantrell, D. A. (1997).J. Biol. Chem. 272,14426-14433. Remillard, B., Petrillo, R., and Maslinski, W. (1991). J. Biol. Chem. 266, 14167-14170. Renno, T., Haline, M., and MacDonald, H. R. (1995). J. Exp. Med. 181, 2283-2287. Rieux-Laucat, F., Le Diest, F., Hivroz, C., Roberts, I. A. G., Debatin, K. M., Fischer, A,, and de Villartay, J. P. (1995). Science 268, 1347-1349. Robb, R. J., Munck, A., and Smith, K. A. (1981). J. Exp. Med. 154, 1455-1474. Rocha, B., and von Boehmer, H. (1991). Science 251, 1225-1228. Rodewald, H.-R., Moignon, P., Lucich, J. L., Dosiou, C., Lopez, P., and Reinherz, E. L. (1992). Cell 69, 139-150.
BIOLOGY OF THE INTEHLEUKIN-2 RECEPTOR
77
Roessler, E., Grant, A., Ju, G., Tsudo, M., Sugamura, K., and Waldrnann, T. A. (1994). Proc. Natl. Acad. Sci. USA 91, 3344-3347. Roifman, C. M. (1997). Can. J. Allergy C h . Zrnmunol. 2, 60-62. Rolink, A., Crawunder, U., WinMer, T. H., Karasuyarna, H., and Melchers, F. (1994). Znt. Immnunol. 6, 1257-64. Rosen, F. S., Cooper, M. D., and Wedgwood, R. J. (1995). N . Engl. J. Med. 333,431-440. Rothenberg, E. V. (1992). Adv. Zmmunol. 51, 85-214. Russell, J. H., Rush, J. B., Weaver, C., and Wang, R. (1993a). Proc. Natl. Acad. Sci. USA 90, 4409-4413. Russell, J. H., and Wang, R. (1993). Enr. J. Zinmunol. 23, 2379-2382. Russell, J. H., White, C. L., Loh, D. Y., and Meleecly-Rey,P. (1991). Proc. Nntl. Acnd. Sci. USA 88, 215-2155. Russell, S. M., Johnston, J. A., Noguchi, M., Kawamura, M., Bacon, C. M., Friedmann, M., Berg, M., McVicar, D. W., Witthuhn, B. A,, Silvennoinen, O., et al. (1994). Science 266, 1042-1045. Russell, S. M., Keegan, A. D., Harada, N., Nakamura, Y., Noguchi, M., Leland, P., Friedmann, M. C., Miyajima, A., Pun, R. K., Pad, W. E., and Leonard, W. J. (1993b). Science 262, 1880-1883. Russell, S. M., Tayebi, N., Naajima, H., Riedy, M. C., Roberts, J. L., Aman, M. J., Migone. T.-S., Nopchi, M., Markert, M. L., Buckley, R. H., O’Shea, J. J., and Leonard, W. J. (1995). Science 270, 797-800. Sadlack, B., Kuhn, R., Schorle, H., Rajewsky, K., Muller, W., and Horak, I. (1994). Eur. J. Iininunol. 24,281-284. Sadlack, B., Merz, H., Schorle, H., Schimpl, A,, Feller, A. C., and Horak, I. (1993). Cell 75, 253-261. Saijo, K., Park, S., Ishida, Y., Arase, H., and Saito, T. (1997).J . Erp. Med. 185, 351-356. S’anchez, M.-J., Muench, M. O., Roncarolo, M. C., Lanier, L., and Phillips, J. H. (1994). J. Erp. Med. 180, 569-580. Satoh, T., Minami, Y., Kono, T., Yarnada, K., Kawahara, A,, Taniguchi, T., and Kaziro, Y. (1992).J. Biol. Clzein. 267,25423-25427. Satoh, T., Nakafnku, M., Miyajima, A., and Kaziro, Y. (1991). Proc. Natl. Acad. Sci. USA 88,3314-3318. Sauve, K., Nachrnan, M., Spence, C., Bailon, P., Campbell, E., Tsien, W.-H., Kondas, J. A,, Hakimi, J., and Ju, G . (1991). Proc. Natl. Acad. Sci. USA 88, 4636-4640. Schimpl, A,, and Hunig, T. (1994). Zn “Overexpression and Knockout of Cytokines in Transgenic Mice.” Academic Press, New York. Schimpl, A,, Hunig, T., Elbe, A., Berberich, I.. Kramer, S., Merz, H., Feller, A. C., Sadlack, B., Schorle, H., and Horak, I. (1994). In “Transgenesis and Targeted Mutagenesis in Immunology.” Academic Press, New York. Schimpl, A., Schorle, H., Hunig, T., Berberich, I., and Horak, I. (1992). In “New Advances on Cytokines” (S. Roinagnani, T. R. Mosmann, and A. K. Abbas, eds.). Raven Press. New York. Schneider, P., Thome, M., Burns, K., Bodmer, J,-L., Hofrnann, K., Kataoka, T., Holler, N., and Tschopp, J. (1997). Iminunity 7, 831-836. Schorle, H., Holtschke, T., Hunig, T., Schimpl, A,, and Horak, I. (1991).Nature 352,621-624. Schwartz, R. H. (1990). Science 248, 1349-1356. Sentman, C.L., Shutter, J. R., Hockenbey, D., Kanagawa, O., and Korsmeyer, S. J. (1991). Cell 67,879-888. Serdobova, I., Pla, M., Reichenbach, P., Sperisen, P., Ghysdael, J., Wilson, A., Freeman, J., and Nabholz, M. (1997).J. Exp. Med. 185, 1211-1221.
78
BHAD H. NELSON AND DENNIS M. WILLEHFOHD
Sharfe, N., Dadi, H. K., Shahar, M., and Roifinan, C. M. (1997). Proc. Natl. Acad. Sci. USA 94,3168-3171. Sharfe, N., Shahar, M., and Roifman, C. M. (1997). J. Clin. Invest. 100,3036-3043. Sharon, M., Gnarra, J., and Lonard, W. (1990). Proc. Natl. Acad .Sci. USA 87,4869-4873. Sharp, L. L., Schwartz, D. A,, Bott, C. M., Marhall, C. J., and Hedrick, S. M. (1997). lmnirtnity 7, 609-618. Sherr, C. J. (1994). Cell 79, 551-555. Sherr, C. J., and Roberts, J. M. (1995). Genes Dev. 9, 1149-1163. Shi, Y., Sahai, B. M., and Green, D. R. (1989). Nature 339, 625-626. Shibuya, H., Yoneyama, M., Nakamura, Y., Harada, H., Hatakeyarna, M., Minamoto, S., Kono, T., Doi, T., White, R., and Taniguchi, T. (1990).Nucleic Acids Res. 18,3697-3703. Shibuya, H., Yoneyama, M., Ninorniya-Tsuji, J., Matsumoto, K., and Taniguchi, T. (1992). Cell 70, 57-67. Shimoda, K., van Deursen, J., Sangster, M. Y., Sarawar, S. R., Carson, R. T., Tripp, R. A., Chus, C., Quelle, F. W., Nosaka, T., Vignali, D. A. A,, Doherty, P., Grosveld, G., Paul, W., and Ihle, J. (1996). Nature 380, 630-633. Shinkai, Y., Koyasu, S., Nakayarna, K., Murphy, K. M., Loh, D. Y., Reinherz, E. L., and Alt, F. W. (1993). Science 259, 822-5. Shortman, K., and Wu, L. (1996). Annu. Rev. Immunol. 14, 29-47. Siegel, J. P., Sharon, M., Smith, P. L., and Leonard, W. J. (1987). Science 238, 75-78. Singer, G. G., and Abbas, A. K. (1994). Immunity 1, 365-371. Singer, G. G., Carrera, A. C., Marshak-Rothstein, A,, Martinez, C., and Abbas, A. K. (1994). Curr. Opin. bnmunol. 6, 913-20. Skolnik, E. Y., Lee, C. K., Batzer, A., Vicente, L. M., Zhou, M., Daly, R., Myers, M. J. J,, Backer, J. M., Ullrich, A., White, M. F., and Schlessinger, J. (1993).EMBO J. 12, 19291936. Smith, K. A. (1988). Science 240, 1169-1176. Smith, K. A,, Favata, M. F., and Oroszlan, S. (1983).1. Znimunol. 131, 1808-1815. Smyth, M. J., Ortaldo, J. R., Shinkai, Y., Yagita, H., Nakata, M., Okumura, K., and Young, H. A. (1990).]. Exp. Med. 171, 1269-1281. Soldaini, E., Pla, M., Beermann, F., Espel, E., Corthesy, P., Barange, S., Waanders, G., MacDonald, H., and Nabholz, M. (1995).J. B i d . Chem. 270, 10733-10742. Songyang, Z., Baltimore, D., Cantley, L. C., Kaplan, D. R., and Franke, T. F. (1997). Proc. Natl. Acad Sci. USA 94, 11345-11350. Speiser, D. E., Sebzda, E., Ohteki, T.. Bachniann, M. F., Pfeffer, K., Mak, T. W., and Ohashi, P. (1996). Eur. 1~ Imrnunol. 26, 3055-3060. Spencer, C. A,, and Groudine, M. (1991). Adu. Cancer Res. 56, 1-48. Sperisen, P., W a g , S., Soldaini, E., Pla, M., Rusterholz, C., Bucher, P., Corthesy, P., Reichenbach, P., and Nabholz, M. (1995). J. B i d . Chem. 270, 10743-10753. Spits, H., Lanier, L. L., and Phillips, J. H. (1995). Blood 85, 2654-2670. Sprent, J., and Tough, D. F. (1994). Science 265, 1395-1400. Starr, R., Willson, T. A., Viney, E. M., Murray, L. J. L., Raper, J. R., Jenkns, B. J., Gonda, T. J., Alexander, W. S., Metcalf, D., Nicola, N. A., and Hilton, D. J. (1997). Nature 387,917-921. Stephan, J. L., Vlekova, V., Le Deist, F., Blanche, S., Donadieu, J., De Saint-Basile, G., Durandy, A,, Griscelli, C., and Fischer, A. (1993).J . Pediat. 123, 564-572. Stern, J. B., and Smith, K. A. (1986). Science 233, 203-206. Strasser, A,, Harris, A. W., Bath, M. L., and Cory, S. (1990). Nature 348, 331-333. Strasser, A,, Harris, A. W., and Cory, S. (1991a). Cell 67, 889-899.
BIOLOCY OF THE INTERLEUKIN-2 RECEPTOH
79
Strasser, A,, Whittinghain. S., Vaux, D. L., Bath, M. L., Adam, J. M., Cory, S., aiid Harris, A. Mi.(1991b). Proc. Nutl. Acad. Sci. USA 88, 8661-8665. Suda, T.. Okazakr, T., Naito. Y., Yokota, T.. Arai. N., Ozaki, S., Nakao, K., and Nagata, S. (199%5). J . I t n i t w r d 154, 3806-3813. Sugamura, K., Asao, H., Kondo, M., Tanaka, N., Isliii, N., Ohbo, K., Nakamura, M., and Takeshita, T. (1996). Annu Rec. Z ~ t ~ n i u u o14, / . 179-205. Sugamiira, K.. Takeshita, T., Asao, H., et ol. (1990).Ly,ry,hokine Am. 9,,539442. Sun, X. J., Crimmins, D. L., Myers, M. G. J,, Mirapeix, M., and White, M. F. (1993).M d . Cell. B i d . 13, 7418-7428. Sun, X. J., Wang. L. M., Zhang, Y., Yenush, L., Myers, M. G. J,, Glasheen, E., Lane, W. S., Pierce, J. H., and White, M. F. (1995).N & w s 377, 173-177. Suzuki, H., Duncan, G. S.. Takimoto, H., and Mak, T. W. (1997a).J.Exp.Med. 185,499-505. Suzuki, H., Hayakawa, A., Bouchard, D., Nakashima, I., and Mak, T. W. (199711). Znt. Ztnttzutid 9, 1367-1374. Suzuki, H., Kiirtdig, T. M., Furlonger, C., Wakehain, A,, Tirnms, E., Matsuyama,T., Schinits, R., Siinard, J. J. L., Ohashi, P. S., Criesser, H., Taniguchi, T., Paige, C., and Mak, T. W. (1995).Science 268, 1472- 1476. Sytwu, H.-K., Libau, R. S., and McDevitt, H. 0. (1996).Itrimunity 5, 17-30. Tagaya, Y., Bamford, R. N., DeFilippis, A. P., and Waklinann, T. A. (1996a).bnn~u?izty 4,329-336. Tagaya, Y., B~irtoii,J. D., Miyamoto, Y., and Waldinann, T. A. (1996b). E M B O J . 15,49284939. Taichman, R., Meiida, I., Torigoe, T., Gaulton, G. N., arid Reed, J. C.(1993).]. B i d . Chsni. 268,20031-20036. Takahashi, A,, and Earnshaw, W. C. (1996).Curr. Opiri. Genet. Deu. 6, 50-55. Takaki, S., Kanazawa, H., Shiiba, M., and Takatsu, K. (1994). Mol.Cell.Bio1. 14, 7404-1413. Takeda, K., Tanaka, T., Shi, W., Matsumoto, M., Minami, M., Kashiwainura,S.-I., Nakanishi, K.. Yoshida, N., Kishiiiioto, T., and Akira, S. (1996a).Nature 380,627-630. Takeda, S., Rodewdd, H. R., Arakawa, H., Bhethinann, €I., and Shirnizu, T. (199611). Itnniutiitiy 5, 217-228. Takeshita, T., Arita, T., Asao, H., Tanaka, M., Higuchi, M., Kuroda, H., Kaneko, K.. Munakata, H., Endo, Y., Fujita. T., a i d Siigamura, K. (1996).Biochem. Biophys. Res. Conanun. 225, 1035-1039. Takeshita, T., Arita, T., Higuchi, M., Asao, H., Endo, K., Kuroda, H., Tanaka, N., Murata, K., Ishii, N., and Suganiiira, K. (1997).Itmumit!/ 6, 449-455. Takeshita, T., Asao, H., Ohtani, K., Ishii, N., Kumaki, S., Tanaka, N., Munakata, H., Nakamura, M., and Sugamura, K. (1992a).Science 257, 379-382. Takeshita, T., Asao, H., Suzuki, J., and Sugainura, K. (1990).Znt. Zniinund. 2, 477-480. Takeshita, T., Ohtani, K., Asao, H., Kumaki, S., Nakamura, M., and Sugamura, K. (1992b). J . Inmuno/. 148, 215442158, Tanaka, N., Asao, H., Ohbo. K., Ishii, N., Takeshita, T., Nakamura, M., Sasaki, H., aiid Sugamura, K. (1994).Proc Nat/ .Aced. Sci. USA 91, 7271-7275. Tanchot, C., Lemonnier. F. A,, P’erariiau, B., Freitas, A. A,, ad Rocha, B. (1997).Science 276, 2057-2062. Tanchot. C., and Rocha. B. (1997). J . Exp. M i d 186, 1099-1106. Tentori, L., Longo, D. L.. Znniga-Pfliicker. J. C., Wing, C., and Kruisbeek, A. M. (1988). J. Exp. &led 168, 1741-1747. Terada, N.. Patel, H. R., Takase, K., Kohno, K., Nairn, A,, and Gelfanti. E. W. (1994). Proc. Natl. Acad. Sci. USA 91,11477-11481.
80
BRAD H. NELSON AND DENNIS M. WILLERFORD
Thierfelder, W. E., van Deursen, J. M., Yamamoto, K., Tiipp, 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). Nature 382, 171-174. Thomis, D. C., and Berg, L. J. (1997).J.Exp.Med. 185, 197-206. Thomis, D. C., Gurniak, C. B., Tivol, E., Sharpe, A. H., and Berg, L. J. (1995). Science 270, 794-797. Toribio, M. L., de la Hem, A,, Borst, J., Marcos, M. A., M’arquez, C., Alonso, J. M., B’arcena, A,, and Mart’inez, C. (1988).1.Exp. M e d 168, 2231-2249. Toribio, M. L., Guti’errez-Ramos,J. C., Pezzi, L., Marcos, M. A,, and Mart’inez, C. (1989). Nuture 342, 82-85. Torigoe, T., Saragovi, H. U., and Reed, J. C. (1992). Proc. Natl. Acad. Sci. USA 89,26742678. Tortolani, P. J., Lal, B. K., Riva, A,, Johnston, J. A., Chen, Y.-Q., Reaman, G. H., Beckwith, M., Longo, D., Ortaldo, J. R., Bhatia, K., McGrath, I., Kehrl, J., Tuscano, J., McVicar, D. W., and O’Shea, J. J. (1995).J. Immunol. 155, 5220-5226. Trinchieri, G. (1989).Adv. Immunol. 47, 187-376. Tsudo, M., Goldman, C. K., Bongiovanni, K. F., Chan, W. C., Winton, E. F., Yagita, M., Grimm, E. A,, and Waldmann, T. A. (1987). Proc. Nutl. Acad. Sci. USA 84, 5394-5398. Tsndo, M., Kitamnra, F., and Miyasaka, M. (1989).Proc. Nutl. Acad. Sci. USA 86,1982-1986. Tsudo, M., Uchiyama, T., and Uchino, H. (1984).J. Exp. Med. 160, 612-617. Turner, B., Rapp, U., App, H., Greene, M., Dobashi, K., and Reed, J. (1991). Proc. Nutl. Acud. Sci. USA 99, 1227-1231. Ullrich, A., and Schlessinger, J. (1990). Cell 61, 203-212. Vanhaesebroeck, B., Leevers, S. J., Panayotou, G., and Waterfield, M. D. (1997). TIBS 22, 267-272. Van Parijs, L., Biuckians, A., Ibragimov, A,, Ak, F. W., and Abbas, A. K. (1997).J. Immunol. 158,3738-3745. Vaw, D. L. (1993). Proc. Nutl. Acad. Sci. USA 90, 786-789. Veis, D. J.. Sorenson, C. M., Shutter, J. R., and Korsmeyer, S. J. (1993). Cell 75, 229-240. Verheij, M., Bose, R., Lin, X. H., Yao, B., JaMs, W. D., Grant, S., Birrer, M. J., Szabo, E., Zon, L. I., Kyriakis, J. M., Haimovitz-Friedman, A., Fuks, Z., and Kolesnick, R. N. (1996). Nature 380, 75-79. von Freeden-Jeffry, U., Vieira, P., Lucian, L. A,, McNeil, T., Burdach, S. E. G., and Murray, R. (1995).J. Exp. Med. 181, 1519-1526. Voss, S. D., Leary, T. P., Sondel, P. M., and Robbs, R. J. (1993). Proc. Nuti. Acad. Sci. U S A 90, 2428-2432. Voss, S. D., Sondel, P. M., and Robb, R. J. (1992).J. Exp. Med. 176,531-541. Wakao, H., Harada, N., Kitamura, T., Mui, A. L., and Miyajiina, A. (1995). E M B O J . 14,2527-2535. Waldmann, T. A. (1989).Annu. Rev. Biochem 58, 875-911. Waldmann, T. A. (1991).J. B i d . Chem. 266, 2681-2684. Waldmann, T. A., Goldman, C. K., Robb, R. J., Depper, J. M., Leonard, W. J., Sharrow, S. O., Bongiovanni, K. F., Korsmeyer, S. J., and Greene, W. G. (1984).J. Exp. Med. 160, 1450-1466. Wang, H. M., and Smith, K. A. (1987).J. Exp. Med. 166, 1055-69. Wang, L. M., Keegan, A. D., Paul, W. E., Heidaran, M. A., Gutkind, J. S., and Pierce, J. H. (1992). EMBO J. 11,4899-4908. Wang, R., Rogers, A. M., Rush, B. J., and Russell, J. H. (1996).Eur.J. lnimunol. 26,22632330. Watanabe, S., Ishida, S., Koike, K., and Arai, K. (1995). Mol.Biol.Cel1 6, 627-636.
BIOLOGY OF T I l E INTERLEUKIN-2 RECEPTOR
81
Watson, J. D , Morrissey, P. J., Namen, A. E., Conlon, P. J., and Widmer, M. B. (1989). J Inmunol 143, 1215-1222. Webb, S., Morns, C., and Sprent, 1. (1990). Crdl 63, 1249-1256. l . 1979Wei, S., Blanchard, D., Liu,&J..Leo!lard, W., and Djeu, J. (1993).J I n ~ m u t ~ o150, 1987. Weinberg, K., and Parkman, R. (1990). N . Engl. J . Mecl. 322, 1718-1723. Weiss, A,, and Imboden, J. (1987). Arlo. Itntnunol. 41, 1-38. Welham, M. J., Duronio, V., Leslie, K. B., Botwell, D., and Schrader, J. W. (199421).J. Biol. Chem. 269, 21165-21176. Weham, M. J., Duronio, V., and Schrader, J. W, (1994b).J . Biol. Cheni. 269, 586555873, White, E. (1996). Genes Deo. 10, 1-15. White, M. R., and Jahn, C. R. (1994).J . B i d . C h e m 269, 1-4. Willerford, D. M., Chen, J., Ferry, J. A,, Davidson, L., Ma, A., and Aft, F. W. (1995). btiinunity 3, 521-530. Willerford, D. M., Swat, W., and Alt, F. W. (1996). Cum. Opin. Genet. Deo. 6, 603-609. Williams, N . S., Moore, T. A,, Schatzle, J. D., Puzanov, I. J., Sivakumar, P. V., Zlotnik, A,, Bennett, M., and Kumar, V. (1997).J. Exp. Med. 186, 1609-1614. Witthuhn, B. A,, Silvennoinen, O., Miura, 0..et cil. (1994). Nature 370, 153-157. Wong, B., Arron, J., and Choi, Y. (1997).J , Exp, Med. 186, 1939-1944. Xia, Z., Dickens, M., Raingeaud, J., Davis, R. J., and Creenberg, M. E. (1995).Science, 13261331. Yang, E., and Korsmeyer, S. J. (1996). Blood 88, 386-401. Yoshimura, A,, Ichihara, M., Kinjyo, I., Moriyama, M., Copeland, N. G., Gilbert, D. J., Jenkins, N . A,, Hara, T., and Miyajima, A. (1996). EMBO J. 15, 1055-1063. Yoshimura, A,, Ohkubo, T., Kiguchi. T., Jenldns, N. A,, Gilbert, D. J., Copeland, N. G., Hara, T., and Miyajima, A. (1995). EMBO J , 14, 2816-2826. Zheng, L., Fisher, G., Miller, R. G., Peschon, J., Lynch, D. H., and Lenardo, M. J. (1995). Nature 377, 348-351. Zhu, L., and Atlasetti, C. (1995).J . Zmt~unol.154, 192-200. Zhu, X., Suen, K.-L., Barbacid. M., Bolen, J. B.. and Fargnoli, J. (1994). J.Biol.Chern. 269, 5518-5522. Zoh, H., Weedon, H., Thompson, C., Fung, M., Ingley, E., and Hapel, A. (1991). Ittiniunology 1991, 167-173. Zubler, R. H., Lowenthal, J. W.. Erard, F., Hashimoto, N., Devos, R., and Macdonald, H. R. (1984).J. Exp. Med. 160, 1170-1183. Zuniga-Pflucker, J. C., and Kruisbeek, A. M . (1990)./. Ioimunol. 144, 3736-3740. Zuniga-Pflucker,J. C., Smith, K. A,, Tentori, L., Pardoll, D. M., Longo, D. L., and Kruisbeek, A. M. (1990). Dea. Zrnmtinol. 1, 59-66. Zurawski, S. M., Imler, J.-L., and Zurauaki, C. (1990). EMBO J. 9, 3899-3905. This article was accepted for publication on February 4, 1998.
This Page Intentionally Left Blank
kDY9hlCE5 Ih IMMIIN0I.OC.I
\'01
70
interleukin-12: A Cytokine at the interface of inflammation and Immunity GlORGlO TRlNCHlERl The Wshr fflstiluk of Anubmy und Biokgy, Phi/ude$hio, Pennsybuniu
I9 f 04
I. Introduction
Adaptive immunity in higher organisms is an extremely specialized mechanism of resistance that is characterized by fine recognition of specific antigens and immunological memory. Thus, it represents a much more effective and specialized mechanism of defense against infections or pathologicd alteration than innate and natural resistance, which are present even in lower organisms and are characterized by a lack of memory and specificity for antigen. However, because adaptive immunity is based on the clonal expansion of B or T lymphocytes carrying the receptor for a specific antigen, it becomes effective only a few days after exposure to the antigen, especially in primary infections. Thus, even in higher organisms with a developed adaptive immune system, the first line of defense against primary infections is provided by the effector cells and mechanism of innate resistance. For example, production of antiviral substances such as interferon and activation of macrophages and natural killer ( N K ) celIs is observed within 1 or 2 days of viral infection and efficiently contributes in controlling the infection, although complete eradication of the infection usually requires the production of specific antibodies and the generation of cytotoxic T cells, which occur about 1 week after initial infection. The relationship between innate resistance and adaptive immunity is not only temporal, but also profoundly interactive via a complex crosstalk between inflammatory cells [including phagocpc cells, other antigenpresenting cells (APC), and N K cells] and the antigen-specific T and B lymphocytes (1). This regulatory cross-talk underlies the phenomenon of inflammation and is mediated by both direct cellular interactions and soluble factors, including cytokines and other pharmacological mediators. The role of cytokines produced by lymphocytes, i.e., lymphokines, in inflammation and in the immune response is well characterized, and the physiologic significance of lymphokines such as interleukin-2 ( IL-2), interferon-? (IFN-.)I), and IL-4 has been analyzed in depth. Phagocytes act as APC and as accessory cells for lymphocyte responses by providing membrane-bound costimulatory molecules (e.g., B7 antigen binding to the CD28 receptor on T cells) (2) and by secreting iminunoregulatory cyto83
C opvnght Q 1998 hy A r x l r m i c Prrs, !\I1 nghtf of rrpmdu~tionIn .m\ firriii rrsrrwd 0065 Zi7hiYY 825 011
84
GIORGIO TRINCHIERI
kines. Although cytokines such as tumor necrosis factor (TNF-a), IL-1, IL-6, and IL-8 that are derived in part from phagocytic cells have been studied extensively with respect to their role in inflammation, shock, and tissue damage, much less is known about their ability to regulate lymphocyte functions. Thus, many of the regulatory functions of accessory phagocytic cells on the immune response mediated by T and B lymphocytes remain unexplained. However, the primary role of phagocytw cells as a first line of defense against bacterial and parasitic infections implies that the ability of these cells to direct the generation of subsequent immune response is instrumental in determining the success or failure of the organism’s mechanisms. The interaction between phagocytic cells and lymphocytes is not unidirectional, and lymphocytes produce factors that are potent regulators of phagocytic cell functions, with both enhancing and suppressing effects. These factors include IFN-?, (the most potent enhancer), granulocyte/macrophage colony-stimulating factor (GM-CSF), macrophageCSF (M-CSF), IL-4, IL-10 (the most effective inhibitor), and others (3-5). The ability of bacterial products to activate phagocyhc cells has been utilized widely in the attempt to boost the immune system, by inducing an immune or inflammatory response against tumors or by providing an adjuvant effect in vaccination. Pure antigens are notoriously poor immunogens unless they have a complex and repetitive structure, and for the generation of efficient humoral or cellular immunity, they should be mixed with adjuvants to facilitate presentation of antigen to the immune system. Although adjuvants differ chemically, they are often irritants that induce an inflammatory response, and bacterial preparations have often been used effectively for this purpose. Furthermore, there have been sporadic observations since the 1700s of certain cancer patients undergoing bacterial infections and a concomitant remission of their malignant growth (6). In 1893 Coley compiled several observations of tumor regression associated with bacterial infections, primarily with streptococcus-induced erysipelas (7). He initiated treatment of cancer patients, first unsuccessfully, with Streptococcus pyogenes and then, in a large number of patients, with S. pyogenes in association with the gram-negative bacterium Bacillus prodigi osus (Serratia marcexens) (7,s).Although his results are difficult to evaluate by modern standards for clinical trials, Coley observed a high proportion of partial and even complete remissions or cures, especially in the case of soft tissue sarcomas. Among the most plausible explanations for those therapeutic benefits are the antitumor effects of infection-induced hyperthermia and the secretion of cytokines, of which TNF is thought to play a major role (8,9). However, despite the dramatic curative effects of TNF, alone or in association with IFN-7, on sarcomas when used at extremely high concentrations in isolated limbs, it is difficult to envision that such a
INTERLEUKIN-12
85
highly toxic molecule, acting alone by a systemic effect, could induce the antitumor response associated with Coley's toxin treatments or with spontaneous infection in certain cancer patients (8). The immune response to infectious agents and to nominal antigens is often characterized by a dominance of either cell-mediated or liumoraltype effector mechanisms ( l o ) ,which has been attributed to a dichotomy in the cytokine production pattern of Th CD4+cells (11).T h l cells produce IL-2, IFN-y, and lyinphotoxiii (LT) and favor cell-mediated immunity, delayed-type hypersensitivity (DTH), niacrophage activation, and production of opsonizing antibodies. Th2 cells produce IL-4, IL-5, IL-6, and IL10 and favor humoral responses, production of IgE and IgA, and activation of eosinophils and basophils. The differentiation of Th cells toward a Th1 or Th2 phenotype occurs early during an immune response and is influenced by many interrelated factors, including the nature and the concentration of the antigen, the anatomical localization of the immune response, the nature of the APC, and the inflammatory cytokine milieu at the site of the immune response (12). The dichotomy in the cytokine production pattern is not limited to CD4+ Th cells, but is also clearly demonstrated for CD8' T cells (13-16), for T cells with y6 T-cell receptor (TCR) (17), and, to a certain extent, for NK cells (18). The discovery of natural killer cell stimulatory factor (NKSF) or IL-12 (19-22) has, at least in part, provided an explanation for some of the immunoregulatory functions of phagocytic cells and APC, including their ability to direct Th cell differentiation, and may represent the missing link in the cross-talk between phagocytic cells and lymphocytes. IL-12 is produced by phagocybc cells, B cells, dendritic cells, and possibly other accessory cells and acts on T cells and NK cells by inducing proliferation and production of cytokines, especially IFN-y, and by enhancing generation and activity of cytotoxic lymphocytes. Acting directly or indirectly through generation of other cytokines or a cascade of cellular interactions, IL-12 is a key factor in the induction of T-cell-dependent and -independent activation of macrophages, generation of T helper type 1 (Thl) cells, generation of cytotoxic T lymphocytes, suppression of IgGl and IgE production, induction of organ-specificautoimmunity, and resistance to bacterial and parasitic infections as well as to tumors. Treatment of animals with recombinant IL-12 has been shown to potentiate the immune response against a large variety of infectious agents, to have a potent antitumor effect, and to act as a potent vaccine adjuvant inducing botli cellular and humoral immunity. These activities of endogenous and exogenous IL- 12 have generated much interest in its study, with rapid progress in our understanding of the immunobiology of this cytokine and in pursuing its use in clinical trials for a variety of pathological conditions.
86
GIORGIO TRINCHIERI
II. 11-12 Molecule and Its Genes
A. DISCOVERY, PURIFICATION, A N D CLONING OF IL-12
NK cell stimulatory factor, later named IL-12, was originally identified as a factor secreted by Epstein-Barr virus (EBV)-transformed human Bcell lines (BCL) that mediates several biological activities on human T and NK cells, including induction of IFN-7 production, enhancement of NK cell-mediated cytotoxicity, and cornitogenic effects on resting T cells (19). Human BCL can facilitate the growth and expansion of NK and T cells (23-25), and coculture of human thymocytes with BCL was reported to induce the production of IFN-7 (26). Because BCL produce a variety of cytokines, Kobayashi et nl. (19), while purifying lymphotoxin and other cytokines affecting the formation of hematopoietic colonies (27) from the EBV-transformed BCL supernatant fluid, investigated whether the effects of BCL on NK and T cells were mediated by a soluble factor(s) and identified a cytokine, then termed NKSF, which mediated induction of IFN-7, mitogenesis, and enhancement of cytotoxicity in T and NK cells. NKSF was purified to homogeneity from the conditioned medium of the phorbol diester-stimulated RPMI-8866 EBV-transformed BCL. Unlike any other cytokine, NKSF was shown to have a heterodimeric structure, composed of covalently linked chains designated p40 and p35 based on their apparent molecular weight (19). Purified NKSF (9200-fold enriched) exhibited its biological activities at concentrations in the range of 0.1 to 10 pM (19). The genes encoding the two polypeptide chains of NKSF were cloned on the basis of partial amino acid sequences of several peptides obtained from the purified proteins, and biologically active recombinant IL-12 was produced in eukaryotic cells transfected with the cDNA for both NKSF chains (21). The cytotoxic lymphocyte maturation factor (CLMF) was later identified in the conditioned medium of the NC37 cell line, an EBV-transformed BCL, on the basis of its ability to synergize with IL-2 in inducing the generation of lymphokine-activated killer (LAK) cells and to induce proliferation of human phytohernagglutin (PHA)-activated T-cell blasts (20). Purification and cloning of the genes encoding CLMF showed that NKSF and CLMF were the same cytokine (19-22), and the unifying term of IL-12 is now universally accepted.
B. IL-12 ~ 4 S0U B U N I T The gene encoding the IL-12 p40 subunit of the 70-kDa heterodimer has been mapped to human chromosome 5q31-q33 (28) and to a syntenic region in mouse chromosonie 11 (29-31), in a region containing genes encoding several cytokines and cytokine receptors, including IL-4. However, the mouse IL-12 p40 gene has been positioned 9.3 cM proximal to
INTERLEUKIN-I!?
87
the IL-4 gene and thus the IL-12 gene appears to be outside the region of closely linked cytokine genes (29). The human p40 gene is composed of eight exons and seven introns (32; S. Wolf, personal communication), similar to the mouse p40 gene (30, 31). It contains a single long-open reading frame encoding a 328 amino acid polypeptide with a 22 amino acid long hydrophobic signal peptide and a characteristic cleavage site immediately preceding the N-terminal sequence of the natural p40 protein (21, 22). Tlie p40 cDNA also contains a relatively long (1.32 kb) 3’untranslated sequence that includes one copy of the Alu repetitive sequence element and multiple copies of ATTTA mRNA destabilizing sequence common to many cytokine and protooncogene inRNAs (21). Tlie mature protein (calculated M , 34,700; pZ 5.4) contains 10 cysteine residues, four consensus sequences for asparagine-linked glycosylation, and one theoretical heparin-binding site (21,22). Immunoprecipitation or Western blotting of both natural and recombinant IL-12 p40 reveals heterogeneity from -36 to more than 40 kDa, with at least two dominant bands (33; G. Carra, personal communication). Treatment with N-glycosidase or chemical deglycosylationwith trifluorometlianesulfonic acid eliminated the heterogeneity and reduced the molecular inass to yield and single band of 36 kDa. Overall, these results indicate that the p40 subunit is composed of -10% N-linked carbohydrates, with a heterogeneity in the extent of glycosylation and no evidence of O-linked oligosaccharides (33). Of the four possible N-glycosylation sites, two were analyzed in sequenced peptides from the natural human IL-12, and Asnzoobut not AsnIo3was found to be N-glycosylated (33). The intracliain disulfide pairing of cysteine residues has been determined (34) and is shown in Fig. 1. Eight cysteine residues are involved in intrachain binding, whereas Cysli7is involved in is not paired the interchain binding with Cysi, of the p35 subunit. with any other cysteine in IL-12, but it was determined to be cysteinylated (disulfide bonded with a cysteine) and to contain thioglycolate (paired with the sulfur in thioglycolic acid) (34). Cysteinylation was demonstrated previously in other proteins, but IL-12 is the first protein in which thioglycolation is known to be involved. C. IL-12 ~ 3 SUBUNIT 5 The gene encoding IL-12 p35 was mapped on human chromosome 3p123q13.2 (28) and in the mouse, to a syntenic region on murine chromosome 6 in one study utilizing fluorescence in situ hybridization (31),but on mouse chromosome 3 using backcross analysis in another study (30). The gene structure of the human IL-12 p35 gene has not been reported, but in the mouse it consists of either eight (30) or seven (31)exons. This discrepancy appears to rest in the multiple transcription initiation sites on the mouse
88
GIORGIO TRINCHIERI
FIG.1. Intrachain and interchain disulfide bonds in the human IL-12 p70 heterohmer. The chromosome mapping of the two genes encoding the p40 and p35 chains is also indicated. Cysln on the p40 subunit is involved in the interchain binding with Cys7, of the p35 subunit; Cyseseis not paired with any other cysteine in IL-12, but it is cysteinylated and contains thioglycolate (34).
p35 gene, as determined by primer extension analysis (31) and by rapid amplification of cDNA ends (30), and as further suggested by the isolation of murine p35 cDNA with alternative 5’-untranslated regions (30, 35). The nucleotide sequence of human p35 cDNA isolated from human BCL contains a single long open reading frame encoding a 253 amino acid polypeptide. This sequence contains two potential translation initiation codons (residues 1 and 35) 5’ to the N-terminal sequence determined from the natural p35 protein (21). The upstream translation initiation codon is maintained in p35 genes in nonhuman primates (36) and in pigs (GenBank accession number SSVO8317), but not in the murine p35 (22) or in several other mammals. The sequence initiated from the second methionine encodes a typical hydrophobic signal peptide (residues 35-56) with a consensus cleavage site immediately adjacent to the N-terminal sequence of the mature p35 protein. The hypothetical 34 amino acid sequence beginning with the methionine at residue 1 is less hydrophobic and includes several basic residues (21). Based on sequence data, together with information from other species lacking the upstream initiation site and transfection data using cDNA lacking the methionine at residue 1, it was concluded that the second methionine is sufficient for expression of the functional IL-12 p35 subunit (21). However, other sequences similar to the sequence 1-34 in the IL-12 p35 are found linked to signal peptides of membrane-associated proteins, raising the possibility that this sequence may be involved in generating a membrane form of p35; the possible
INTERLEUKIN-12
89
expression of IL-12 as a membrane-bound forin in both human and murine macrophage cell lines has been suggested in an isolated study using cytofluoriinetric staining with anti-IL-12 antibodies (37). Immediately upstream of the first methionine of the p35 cDNA isolated from BCL, there is a possible TATA box that in the mouse gene has been suggested to be part of an ancestral IL-12 p35 promoter (31).However, the picture now emergmg is that although the major transcription initiation site in human BCL is upstream of that TATA sequence and is reflected in the published cDNA sequences (21, 22), other physiological IL-12-producing cells, e.g., IFN-.)I- and LPS-activated human nionocytes, predominantly use a transcription site downstream of the TATA box, generating a shorter mRNA that includes only the second methionine (M. Hayes, personal cominunication; X. Ma and G. Trinchieri, unpublished observation). The functional significance of these alternative transcripts in different cell types remains unknown. The mature IL-12 p35 peptide is 197 amino acids long (calculated M , 22,500; pZ 6.5) and contains seven cysteine residues and three consensus N-linked glycosylation sites. The 3’-untranslated 450-bp cDNA contains multiple copies of the ATTTA mRNA-destabilizing sequences. Like p40, p35 subunits appear on gels as a heterogeneous band, possibly due to differential glycosylation. Results from chemical and enzymatic deglycosylation suggest that the p35 subunit comprises -20% carbohydrates, 40% of which are 0 linked (33).
D. IL-12 FROM NONHUMAN SPECIES The two IL-12 genes have been cloned in several other mammalian species. The IL-12 p40 and p35 genes from nonhuman primates (rhesus macaque, pigtailed macque, and sooty mangabey) showed -96% homologywith the human genes (36). However, a Val-Ser-Leu at position 27-29 of mature human IL-12 p35 is replaced by a Gln-Pro-Pro sequence in the p35 inolecules of nonhuman primates (36). Such a change probably generated conformation and immunogenicity differences that might underlie the production of antihuman IL-12 antibodies in primates injected for 15-20 days with human recombinant IL-12 (38,39). Comparison of the sequence of the murine IL-12 subunits with their human counterparts revealed that die p40 subunits are more highly conserved than the p35 subunits (70 and 60% homology, respectively) (35).Although human IL12 binds to the mouse IL-12 receptor, it is not active on murine cells, whereas inurine IL-12 has biological activity on both human and mouse lymphocytes (35).A hybrid heterodimer consisting of murine p35 and huinan p40 was also biologically active on both human and mouse cells, whereas the combination of human p35 and mouse p40 was completely
90
GIOKCIO TRINCHIERI
inactive on murine cells, indicating that the inability of human IL-12 to act on murine cells is largely determined by the ~ 3 subunit ~ 5 (35). Five residue changes in three discontinuous sites of the murine p35 molecule eliminate bioactivity on mouse but not on human cells, suggesting that these residues are important elements in determining the species specificity of IL-12 (40). Only a partial sequence is available for rat IL-12 (41). Complete sequences for both chains have been reported for cat (42, 43; GenBank accession numbers Y07762, Y07761), red deer (g1223907, g1223905),woodchuck (g1262373,g1262371),cow (44, g555795, g555917), and pig (45, g984510, g927204), and sequences for the p40 chain are available for goat (g2253433) and sheep (g2199555) IL-12.
E. IL-12 HETERODIMERS AND ~ 4 MONOMERS 0 AND HOMODIMERS Transient transfection of COS cells or stable transfection of CHO cells with either p40 or p35 cDNA induces secretion of the respective IL-12 chains; cotransfection with both cDNAs in the same cells is required for secretion of the biologically active p70 form of IL-12 (21, 22). Unlike the cells transfected with p35 cDNA, primary cells or cell lines have never demonstrated production of the free p35 chain (46), raising the possibility that p35, in the absence of p40, is not secreted spontaneously from cells and that the apparent secretion from transfected cells is due to a release subsequent to cell death and lysis. Consistent with this interpretation is the fact that the majority of p35 protein is found to be cell associated in p35 cDNA-transfected cells (S. Wolf, personal communication). However, p40 cDNA-transfected cells produce large amounts of p40 protein not associated with p35, and BCL as well as other physiological producer cells of IL-12 produce the p40 chain in excess from severalfoldto more than 100fold over the production of the biologically active p70 heterodimer (20,46). The report by Mattner et al. (47) that supernatant fluid from murine IL-12 p40-transfected cells inhibited the biological activity of the murine IL-12 p70 heterodimer raised much interest in determining whether the free p40 chain might be a physiologic antagonist of IL-12 activity. Analysis of the p40 protein produced by p40 cDNA-transfected COS cells revealed that about 70% of the subunit was secreted as monomers and the remaining 30% as 80-kDa homodimers which, under reducing conditions, dissociated into 40-kDa species, suggesting that the human p40 homodimers are covalently linked (48). Peptide analysis confirmed that the 80-kDa molecules were dimers of the 40-kDa forms (48),but because analysis was performed under reducing conditions, the site of the hypothesized covalent linkage was not determined. The p40 hoinodimers competed for the binding of '251-labeledIL-12 to KIT225/K6 human T cells expressing the IL-12 receptor (IL-l2R), with ICs0values 5- to 10-fold higher than that of the hetero-
INTEHLEUKIN-12
91
dimers. Unlike the homodimers, the p40 monomers inhibited '251-labeled IL-12 binding only at concentrations at least 100 times higher than that of the heterodimer and never reached complete inhibition (48).The human p40 homodimers inhibited the proliferation of human PHA-activated Tcell blasts induced by 0.2 n g h l of IL-12 with an ICjo of -50 ng/ml, i.e., a -250-fold excess of homodimer was required for 50% inhibition (48). Chinese hamster ovary (CHO) cells stably transfectecl with murine p40 cDNA also produce monomeiic and homodimeric forms of p40 protein (49). Peptide analysis demonstrated that the murine homodimer arises from formation of a single intermolecular bond at CYS,,~, the same residue used in the p70 heterodimer for the interchain disulfide bond with the p35 chain, whereas in the monomeric p40, this cysteine is capped by cysteinylation (49). As with human p40, the murine p40 homodimer was 100-fold more efficient than the monomers in inhibiting IL-12 receptor binding or biological functions; however, unlike the poor inhibition of IL-12 function observed with human p40 homodimers, the mouse homodimer was equivalent to the p70 heterodimer in competing for IL-12 binding to the mouse cellular receptor and induced a 50% reduction in biological functions (proliferation, IFN-.)Iinduction, and NK cell activation) at concentrations approximately equivalent to those of IL-12 in the assay (47). A subsequent study (50) confirmed the highly inhibitory activity of the murine p40 homodiiner, but only with an IC5{,requiring a 33-fold molar excess of the homodimer over the heterodimer. Because the mouse p40 homodimer was able to inhibit IL-12 biological function at very low concentrations without mediating any detectable biological activity, it was of much interest that approximately 20% of the circulating mouse p40 in vivo, produced either constitutivelyor induced by lipopolysaccharide (LPS) injection, was present as a homodimer, as demonstrated by gel filtration and by SDS-PAGE analysis under reducing or nonreducing conditions (50). Moreover, injection of mice with 40 pg of recombinant murine p40 homodimer significantly inhibited the endotoxin-induced 1FN-y production (50).The ability of the IL-12 p40 homodirner to block IL-12 action in vivu has been extended to suppression of Thl responses and DTH (51). Overall, data argue in favor of the role of the p40 homodimer as a physiological antagonist of IL-12 action in vivo in the mouse, perhaps being produced at later times during an inflammatory response and thus contributing to the extinction of IL-12 biological effects (52). However, this is most likely not the case in humans. The human IL-12 p40 homodiiner has only a modest ability to compete for IL-12 biological activity (48), even considering that human IL-12 p40 may be secreted in vitru and in vivo at concentrations up to 100 times higher than those of the IL-12 p70 heterodiiner. Furthermore, careful analysis of the stoichiometry of the p40
-
92
GIORGIO TRINCHIERI
and p35 chains in the -40- and -70-kDa peaks of IL-12 produced by IFN-y- and LPS-stimulated human monocytes revealed no significant amounts of covalently linked p40 homodimer (G. Carra, F. Gerosa, and G. Trinchieri, unpublished results). Thus, the role of p40 homodimers as a physiological antagonist of IL-12 may represent a species difference between mice and humans. F. IL-12 STRUCTURE The primary amino acid sequence of the IL-12 p35 chain indicates an a-helix bundle structure, similar to most cytokines (40, 53). Comparison of the p35 amino acid sequence with those of IL-6 and G-CSF showed that many of the amino acid positions conserved between these two cytokines are also conserved in IL-12 p35 (54). Interestingly, three leucine zipper motifs located near the N terminus of the human p35 molecule are conserved in other species (42); whether these motifs are involved in binding of p35 to molecules other than p40 remains to be determined. The p40 sequence is not homologous with any other known cytokine, but rather belongs to the hematopoietic cytokine receptor family, which is characterized by four cysteines and one tryptophan in conserved positions in the extracellular portions and by a WSXWS motif (55).The p40 sequence has significant sequence homology with the extracellular portions of the IL-6 receptor and the ciliary neurotrophic factor (CNTF) receptor (35, 55). IL-6R, CNTF-R, and IL-12 p40 have an N-terminal immunoglobulinlike domain followed by the sequence characteristics of the receptor family; the WSXWS motif (which in the p40 sequence is modified by the insertion of an alanine) is near the C terminus in the p40 molecule. The intrachain disulfide pairing of the p40 molecule confirms the homology with the cytokine receptor family (34, 49). The second fibronectin-like domain of human IL-12 p40 contains an RGD sequence and in the mouse, the sequence QEDV, both halImarks of adhesion molecules binding to the integrin receptor family, and also the sequence VTCG similar to adhesion sequences in thrombospondin and properdin (53).Thus, it appears that the second domain of p40 resembles the active sites of adhesion molecules. Furthermore, the third domain of IL-12 p40 and other members of the cytokine receptor family has similarities with members of the gastrointestinal peptide family of hormones, particularly secretin and glucagon prohormones (53).These findings suggest that the different regions of the IL12 molecule may compose a functional mosaic relevant for the action of IL-12 in various tissue, e.g., at the mucosal surface. Most cytokine receptors can be released by cells in soluble forms, which usually have a C terminus immediately following the WSXWS motif and are produced either by proteolytic digestion of the transmembrane form or
INTERLEUKIN-12
93
by alternative splicing of the message with elimination of the exons encoding the transmembrane and cytoplasmic portions (56).The binding of IL6 to the IL-6R is a low-affinity interaction, but a high-affinity hexameric complex forms on association of a diiner of gp130 (a nonligand-binding, signal-transducing transmembrane protein), and signa1 transduction through gp130 is triggered (57). The soluble form of the IL-6R, unlike most other soluble receptors, does not compete for binding of IL-6 to the cellular receptors; rather, it binds in solution with IL-6 and this complex can bind to gp130 on the cell surface, mediating signal transduction and IL-6 biological activities (56,57).The CNTF-R is composed of three chains: gp130, sharedwith the IL-6R, the leukemia inhibitory factor (LIF)receptor p chain, and a CNTF-Ra chain. Like the IL-GR, the CNTF-Ra chain is released as a soluble protein that binds to CNTF, and the complex mediates signal transduction on cell types expressing gp130 and LIF-RP chain (58). IL-11, another member of the IL-6 family of cytokines which share the use of the gp130 chain as part of their receptor (59), also associates in solution to the soluble form of the a chain of its receptor and binds in this coinplexed form to the transmembrane gp130 monomer, inducing signal transduction and biological responses (60). Thus, it is possible that a primordial cytokine (the p35 equivalent), which, like IL-6, CNTF, and IL-11, had a multichain receptor, gave rise to the heterodimeric IL-12 during evolution. The transmembrane form of one chain of the receptor (the p40 equivalent) was lost, but an efficient association of the primitive cytokine and the primitive soluble receptor was maintained by the presence of a covalent linkage between the two chains. The heterodimeric complex, like the soluble IL-6WIL-6, CNTF-Ra/CNTF, and IL-11RdIL-11 complexes, would still be able to bind with high affinity to the one or inore remaining transmembrane chains of the receptor, inducing signal transduction and biological activity. If this hypothesis on the evolutionary origin of IL-12 is correct, one would assume that, analogous to the interaction of IL-6, CTNF, and IL-11 with the a chain of their receptor, the p35 and p40 chains of IL-12 have maintained a ligand-receptorlike affinity for each other, even in the absence of covalent linkage between the two chains. Indeed, when monomeric recombinant IL-12 p40 and p35 are added together to responsive cells, all the biological activities of IL-12 can be demonstrated (M. Rengaraju, A. D’Andrea, and G. Trinchieri, unpublished results), although at concentrations from two to five orders of magnitude higher than those effective for the covalently linked heterodimer. The homology of IL-12 p40 with cytokine receptors and the human growth hormone receptor, as well as that of p35 with various cytokines and with the growth hormone itself, has led to a three-dimensional
94
GIORGIO THINCHIERI
model of IL-12 heterodimers, with the growth hormone and its receptors as reference proteins (53). This model strengthens the hypothesis of the evolutionary origin of IL-12 from a cytokine and its receptor and shows that the regions of hoinology with adhesion molecules and gastrointestinal peptides are fully exposed on the protein surface and thus are of possible functional significance. It has also been shown that a single cDNA encoding a single polypeptide formed by the p40 chain C terminus joined to the p35 N terminus by a 15 amino acid flexible linker ([Gly-Gly-Gly-Gly-SerI3), but not the joining of the p35 C terminus to the p40 N terminus, allows the refolding of a biologically active IL-12 fusion protein (61, 62). Constructs of this type may be useful not only in understanding the structure of IL-12, but also in gene therapy by overcoming the need to express two different genes in order to obtain biologically active IL-12. G. EBV-INDUCED PROTEIN 3 (EBI3):A P40-RELATED PROTEIN Devergne et al. (63) isolated a novel cDNA encoding a hematopoietic receptor family member related to the IL-12 p40 and CNTF-Ra subunits, whose expression is induced in B lymphocytes by EBV infection and is constitutive in human placenta. M. Aste, G. Gri, and G. Trinchieri (unpublished results) have shown that the EBIS gene is induced in human monocytes under similar stimulation conditions as IL-12, although with slower and more prolonged kinetics. Most newly synthesized EBIS protein is retained in the endoplasmic reticulum associated with calnexin and a novel 60-kDa protein (63). The protein encoded by the EB13 gene has a predicted M , -25,000, the characteristic amino acids at conserved positions and motifs of the hemopoietin receptor family, 30% homology with CNTFR a , and 27% homology with IL-12 p40 with conserved amino acid substitutions at many of the nonidentical residues (63). Cotransfection of cells with EBIS and IL-12 p35 cDNA results in secretion of the complex of both chains (64), although the chains are not covalently linked because EB13 lacks the cysteine involved in interchain bonding of the Cys177of p40 with p35. Similarly, a large fraction of p35 in extracts of the trophoblast component of a human placenta specifically coimmunoprecipitated with EBIS, indicating that EBIS is a heterodimer with p35, in vivo (64).Although no cytokine function has yet been defined for EBIS and it does not interact with the IL-12R (64), the ability of EBIS to associate with p35 and its expression in IL-12 producing cells at later times of stimulation raise the possibility that it competes with IL-12 p40 for association with IL-12 p35. Thus EBIS might act as an antagonist in the formation of the p70 heterodimer at later times of phagocpc cell activation.
INTERLEUKIN-18
95
111. 11-12 Receptor and Signal Transduction
A. CELLULAR-BINDING SITESFOR IL-12 The binding of IL-12 to cellular receptors has been analyzed primarily in human cells, T lymphocytes, and NK cells (65, 66). Resting peripheral blood lymphocytes (PBL) did not express IL-12-binding sites, but after activation with PHA for 2 to 4 days they expressed 500-3000 binding sites/ cell with an affinity of 100 to 600 pM; the kinetics of expression of the receptor reflected the ability of the cells to proliferate in response to IL12 (65,67).The low affinity of the IL-12-binding sites originally identified could not explain the biological activity of IL-12, often observed at concentrations of a few pM or lower (19).More recent studies using '251-labeled IL-12 at higher specific activity identified three classes of IL-12-binding sites on PHA-activated blasts, with & values of 5-50 pM, 50-200 pM, and 2-6 nM (68).
B. IL-12 RECEPTOR Pl SUBUNIT A monoclonal antibody, 2-4E6, that immunoprecipitated the complex formed between radiolabeled IL-12 and IL-12-binding protein(s) on activated human T cells was produced and used in the cloning of a cDNA that encodes an IL-12R component of -100 kDa (68), corresponding to the molecular mass previously determined by cross-linking experiments for the IL-12-binding protein (67). This component of the IL-12 receptor was designated IL-RP1 (68), based on the sequence analogy of this chain with P chains of other receptors of the hemopoietin receptor family, and on the assumption that the p40 subunit of IL-12 replaced the transmembrane a chain of the receptor of other homologous cytokines, such as IL6 or CNTF, during evolution. IL-12RPl cDNA expression in COS cells led to the binding of both IL-12 and the 2-4E6 antibody (68). This cDNA encodes a mature 638 amino acid protein (calculated M , 70,426) preceded by a 24 amino acid signal peptide; a second hydrophobic area in the molecule, located between amino acids 541 and 571, represents the likely transmembrane region. Thus, the extracellular portion is 516 amino acids long and contains six predicted N-glycosylation sites; the cytoplasmic position is 91 amino acids long and contains three potential phosphorylation sites for casein kinase 11. A second cDNA clone obtained has a 13 amino acid deletion right before the stop codon, probably generated by alternate splicing, which results in a protein 2 amino acids shorter. IL-12RP1 belongs to the hemopoietin receptor family, with the two conserved pairs of cysteine residues and the WSXWS motif, and has the highest homology with gp130, followed by the G-CSF-R and LIF-R; however, unlike gp130. IL-12RP1 lacks an N-terminal immunoglobulin domain (68). The cytoplasmic do-
96
GIORGIO TRINCHIERI
main of IL-12Rp1 contains the box 1and box 2 motifs typical of members of the hemopoietin receptor families, but no tyrosine residues (68).The IL-12R composition and its signal transduction pathways are schematically represented in Fig. 2. Northern analysis of mRNA for IL-12p1 in several cell types has identified two species, the smaller one possibly corresponding to a yet uncloned alternatively spliced transcript lacking the intracellular domain (68). Expression of the IL-12Rp1 cDNA in COS cells results in the expression of binding sites with a 2-6 nM affinity corresponding to the low-affinity binding sites in PHA-activated blasts. In transfected COS cells, IL-12RP1 was present either as monomers or as covalently linked dimers or oligomers; IL-12 binds only to the preformed complexes and not to the monomers (68). Interestingly, IL-12RP1 cDNA-transfected CTLL cells express only monomers and fail to bind IL-12. Murine IL-12Rp1 was also cloned and found to encode a molecule with 54% amino acid homology with the human IL-12Rp1. When expressed in COS cells or in murine BdF3 pro-B cells, the murine IL-lBRPl also forins surface dimers/oligomers with IL-12-binding ability (69).The cytoplasmic portion of the murine /3l chain contains a tyrosine residue and has 55 additional amino acids compared to the human pl. Murine IL12Rpl-transfected B f l 3 cells bind IL-12 with affinities of 50 and 470 pM, corresponding to the medium- and high-affinity IL-12-binding
FIG.2. Schematic representation of the IL-12 receptor and of the signal transduction pathways activated by IL-12 binding.
INTERLEUKIN-18
97
sites on murine lymphoblasts; however, this receptor alone is not sufficient to transduce a signal (69), suggesting the existence of another subunit of the murine IL-12R. Four different cDNAs have been identified for mouse IL-12RP1 that correspond to transcripts expressed by activated mouse T cells (69, 70). In addition to the full transcripts encoding the transmembrane region, two types of transcripts have a 97-bp deletion that eliminates the transmembrane region and results in a frameshift, modifymg the sequence of the C terminus; one of these two transcripts also has a 42-bp deletion in the extracellular portion. The products of the 97-bp deletion transcripts can bind IL-12 and appear to remain cell associated even in the absence of a transmembrane domain (69). A third transcript has an intron insert of more than 3 kb near the transmembrane domain and encodes an IL-12-binding truncated protein (70).The possibility that these transcripts reflected the production of soluble IL-12-binding proteins and their physiological significance remains to be determined. C. IL-12 RECEPTOR 02 SUBUNIT Although the IL-12RP1 chain per se is not sufficient for IL-12 signal transduction, the ability of polyclonal and monoclonal antibodies against human IL-12RP1 to specifically inhibit IL-12 functions (68, 71) and the inability of IL-lBRPl knockout mice to respond to IL-12 (72) confirmed that this chain is an essential subunit of the functional IL-12R. An additional @-typeIL-12 receptor protein, IL-12Rp2, was identified by expression cloning techniques both in mice and in humans. When coexpressed with IL-12RP1, it confers both high-affinity IL-12 binding and IL-12 responsiveness (73). The cloned cDNA contain a very long (640 bp) G + C-rich 5’noncoding region and one long open reading frame encoding a 862 amino acid class I transmembrane protein that consists of a 27 amino acid signal peptide, a 595 amino acid extracellular domain, a 24 amino acid transmembrane region, and a 216 amino acid cytoplasmic tail containing three tyrosine residues (predicted M , of the mature protein is 94,059). The mouse IL-12RP2 is very similar to the human receptor with an overall 68% amino acid sequence homology (73). IL-12RP2 is a member of the cytokine receptor superfamily with one N-terminal immunoglobulin motif and is even more homologous than IL-12RP1 to gp130 and also to G-CSF and LIF receptors (73).The cytoplasmic domain of IL-12Rj32 contains a type 1 and possibly type 2 cytokine box motifs. Two different mRNA species have been identified for IL-12RP2, but no evidence of a transcript for a possible soluble receptor lacking the transmembrane region has been obtained (73). Transfection of IL-12RP2 in COS cells generates single class of IL-12 binding sites with a & of about 5 nM; like IL-12RP1, the IL-12RP2 protein is expressed as a dimer or an oligomer, regardless of
98
GIORGIO TRINCHIERI
the presence of the ligand (73). Coexpression of both IL-12RPl and P2 chains in COS cells results in the formation of a high number of lowaffinity binding sites for IL-12 (7.5 nM) and only a low number of highaffinity (55 pM) bindmg sites, possibly due to the competition between homo- and heteroassociation of the Pl and 8 2 chains; coimmunoprecipitation of the heterocomplexes, even using chemical cross-linking, could not be demonstrated (73).Ba/F3 cells cotransfected with both Pl and P2 proliferate readily in the presence of IL-12, with maximal proliferation observed at 100 pg/ml, whereas BdF3 cells transfected with IL-12RP1 alone do not respond to IL-12; however, high doses of IL-12 (10-100 ng/ ml) induce proliferation of Ba/F3 cells transfected with IL-12RP2, suggesting that the human PZ chain may be at least partidy capable of transducing proliferation signals and that the Pl chains contribute primarily to an increased affinity of IL-12 binding (73). However, no IL-12 response, even at high concentrations, was observed in mice lacking the IL-lSRPl chain (72). The p40 homodimers bind and compete for IL-12 binding in cells transfected with the IL-lBRPl chain, but not in those transfected with the P2 chain, suggesting that Pl binds the p40 subunit of IL-12, whereas P2 binds either the p35 subunit or a structure expressed on the heterodimer (51).
D. REGULATION OF IL-12 RECEPTOR EXPRESSION Resting human T and NK cells express low but variable levels of ILl2RPl as detected by antibody staining, although no IL-12 binding can be detected. Stimulation with mitogens, anti-CD3, and anti-CD3 plus antiCD28 upregulates both the Pl and the P2 chains of IL-12R within a few days of stimulation, with a peak at days 3-4 (74,75). IL-12RPl mRNA is found in activated T and NK cells and cell lines, as well as in B cells; however, B cells and BCL are usually unable to bind IL-12, in part due to their lack of expression of the ICl2RP2 chain (71,72,76,77). In human T cells, IFN-.)Iupregulates and IL-4 downregulates the expression of highaffinity binding sites, without affecting the expression of the 01 chain, thus probably acting at the level of the P2 chain (75). If the & of the identified receptor is more than 50 pmol/liter, it must be assumed that either signal transduction occurs at minimal occupancy of the receptors or additional unidentified receptor chains are required in determining a low number of high-affinity binding sites. The other discrepancy with the functional data is that receptors cannot be identified on resting T and NK cells, whereas certain biological activities of IL-12, e.g., enhancement of cell-mediated cytotoxicity or induction of IFN-.)I production, are mediated with a similar dose-response curve on both resting and activated NK and T cells (78). This discrepancy might be
INTERLEUKIN-12,
99
explained in part by the observations, using analysis of IL-12 binding by cytofluorimetry or staining with anti-IL-12Rpl antibodies, that resting NK cells express low levels of IL-12R (66,71). However, in situ hybridization experiments have shown that more than 10-2096 of resting peripheral blood T cells accumulate INF-y mRNA a few hours after exposure to IL12, suggesting that a significant proportion of peripheral blood T cells also expresses functional IL-12 receptors (78). In addition to the ability of IL-12 to direct differentiation of T h l cells (79, 80), it was shown that Thl, but not Th2, clones are responsive to IL12 (81, 82). Extinction of the IL-12 signaling pathway in early mouse Th2 cells results from the selective loss of IL-12RP2 subunit expression, without alteration in pl subunit expression (83). IL-12RP2 is not expressed by mouse naive resting CD4' T cells, but is induced upon antigen activation through the T-cell receptor; IL-4 inhibits expression of IL-12RP2, leading to the loss of IL-12 signaling and thus contributing to commitment to the Th2 pathway (83).IFN-y, however, prevents the loss ofp2 chain expression and restores the ability of early Th2 cells to respond to IL-12 (83). The T-cell genetic background also contributes to the maintenance of IL-12 signaling and IL-12RP2 subunit expression: when T cells from B10.D2 mice are activated without the addition of exogenous cytokines (neutral condition), they remain responsive to IL-12, whereas activated T cells from BALB/c mice cease IL-12Rp2 expression and become unresponsive to IL12 (84),consistent with their default differentiation along the Th2 pathway (85).Similarly, the IL-12Rp2 subunit is expressed on human T h l but not Th2 clones and is induced during differentiation of human naive cells along the Thl but not the Th2 pathway. However, in contrast with the mouse system, type I ( d p ) ,but not type I1 ( y ) ,interferons induce expression of the p2 chain during T-cell differentiation following T-cell receptor triggering (86). These results are difficult to reconcile with a previous report that IFN-y induces high-affinity IL-12-binding sites on human T cells (75). E. IL-12-INDUCED SIGNAL TRANSDUCTION Early reports indicated that IL-12 induces phosphorylation of the src family lck tyrosine kinase in human NK cells (87, 88), and the ability of IL-12 to induce the expression of the CD69 activation membrane molecule on NK cells was shown to be blocked by tyrosine kinase inhibitors (89). In activated human T cells, but not in N K cells, IL-12 induces tyrosine phosphorylation of a 44-kDa protein identified as an isoform of mitogenactivated serine-threonine kinase (MAPK) (90). IL-12 treatment of both human T and NK cells induces rapid tyrosine phosphorylation of both JAK2 and Tyk2 kinases, implicating these kinases
100
GIORGIO TRINCHIERI
in the immediate biochemical response to IL-12 (91, 92). Chimeric receptors composed of the transmembrane and cytoplasmic region of IL-12Rp1 and P2 chain fused to the extracellular domain of the epidermal growth factor (EGF) have been used to show that JAKZ is phosphorylated in response to EGF in cells expressing the 61 and/or P2 cytoplasmic domains, whereas phosphorylation of Tyk2 is observed only in cells expressing the P l cytoplasmic domain; however, direct physical association has only been demonstrated between JAKZ and p 2 and between Tyk2 and 01 (93). Following activation of the two Janus family kinases, three components of the STAT family of transcription factors are phosphorylated and activated: STAT1, STAT3, and STAT4 (92,94,95). STAT1 can dimerize with either STAT3 or STAT4, and STAT4 forms either homo- or heterodimer with STAT3 (92, 94). STAT4 is both tyrosine and serine phosphorylated in response to IL-12 (96). The serine phosphorylation is not required for DNA binding, but is required for transactivation, based on the finding that the serine kinase inhibitor H7 blocks the ability of IL-12 to induce STATdependent activation of a heterologous promoter, but not its ability to bind to DNA in vitro (97). Serine phosphorylation of STAT4 involves a Pro-Met-Ser-Pro sequence in its C terminus that resembles the consensus recognition sequence for the MAPK (96) and thus may be dependent on the reported activation of the 44-kDa MAPK in T cells (90). However, the mechanism of activation of the MAPK is not clear, as the RAS activation pathway of MAPK is not involved in IL-12 signaling (97). The activation of STAT4 by IL-12 is of particular interest because this transcription factor is not activated by other cytokines (98), with the exception of IFN-a in the human system (96, 99), raising the possibility that STAT4 activation underlies the specific biological responses to IL-12. Indeed, T and NK cells of STAT4 knockout mice are unresponsive to IL-12, and these mice have a phenotype equivalent to that of IL-12 or IL-12R knockout mice (100,101).T and NK cells from STAT4 knockout mice also do not produce IFN-y in response to IL-12, suggesting that STAT4 is directly involved in IFN-y gene transcription. Two adjacent binding sites for STAT4 have been identified in the first intron of the IFN-y gene: these binding sites are variants of the consensus sequence and, alone, they bind STAT4 inefficiently. Recognition of the variations of the consensus site and STAT4 binding require cooperative interactions (102). The conserved aminoterminal domain of STAT proteins was required for cooperative DNA binding, although this domain was not necessary for dimerization or binding to a single consensus site (102). Cotransfection of the full-length STAT4, but not of an amino-terminal-deleted STAT4, induced a fourfold induction of transcription of a reporter gene construct containing the IFN-y first intron (102).
INTERLEUKIN-12
101
Unlike other studies in the mouse (94, 98) and in human NK cells (92) that failed to detect activation of STAT4 by IFN-a, both tyrosine and serine phosphorylation of STAT4 in response to IFN-a was observed in human-activated T cells and the human NK cell line NK3.3 (96, 99). Although the reason for these species differences is not clear, the ability of IFN-a to mimic some of the signal transduction mechanisms of IL-12 may explain the reported effect of this cytokine on the T h l differentiation pathway. In BdF3 cells transfected with chimeric EGFIIL-12Rpl and p 2 receptors (93),phosphorylation of STAT3 in response to EGF was demonstrated in cells transfected with either the ,f3l or the /32 cytoplasmic domain, indicating that the presence and phosphorylation of a tyrosine residue (missing in the pl domain) are not absolute requirements for the activation of STAT factors. These results also show that the individual pl and p2 cytoplasmic domains are capable of signal transduction, similar to the observations with the complete receptors (73). However, proliferation of the transfected BdF3 cells in response to EGF was observed only when the p2 domain, alone or together with the pl domain, was present (93).
N. Production of 11-12 A. MEASUREMENT OF IL-12 PRODUCTION The existence of two separate genes controlling IL-12 production, the need for their simultaneous expression in the same cell type in order to produce biologically active IL-12 (21), and the production of a large excess of the p40 chain over the biologically active heterodimer have made the analysis of IL-12 production particularly complex. Transcripts for the p35 gene have been detected at very low abundance [lessthan 1fglpg total RNA in peripheral blood mononuclear cells (PBMC) or polymorphonuclear leukocytes (PMN) (32, 103)] in almost any cell type tested, including hematopoietic and solid tumor cell lines, and are upregulated on activation (46). Transcripts for the p40 gene have been detected only in cell types producing biologdly active IL-12 and their expression is highly regulated (46). Because expression of both genes is required for biologically active IL-12 expression, detection of mRNA for the tissue-specific and highly regulated p40 gene is a better indicator of IL-12 production than detection of the more ubiquitous p35 mRNA, especially when a mixture of different cell types is analyzed. However, the use of low sensitivity methods for mRNA detection, e.g., in situ hybridization on tissue sections, has led to a reported apparent dissociation of p35 and p40 mRNA expression in different cell types (104). Furthermore, upregulation and expression of the p35 gene are obviously needed for production of the biologically active
102
GIORGIO TRINCHIERI
p70 heterodimer, and p40 gene expression or p40 chain production alone can be considered only indicative for simultaneous production of the biologically active heterodimer in the absence of a direct immunological or biological demonstration of p70 secretion. Because of the extremely variable ratio between the free p40 chain and the p70 heterodimer, the level of p40 gene or protein expression cannot be extrapolated to obtain even an approximate level of heterodimer produced. The production of monoclonal and polyclonal antibodies to the p40 and p35 chains of IL-12 has facilitated analyses of IL-12 production (46, 105). Radioimmunoassays (RIA) or enzyme-linked immunosorbent assays (ELISA) assays that detect the p40 chain (either as a single chain or as a single chain and a complex in the p70 heterodimer) and the single chain p35 have been established (46). Measurement of the p70 heterodimer proved to be more difficult. The initial RIA for detection of human IL12 p70 was cumbersome and affected by contaminant p40 (46).Now several assays are available, some commercial, that allow a sensitive detection of human or mouse p70. These assays utilize a pair of antibodies, one reacting with the p40 chain and the other with either the p35 chain or a determinant specific for the heterodimer (46, 106). An antibody capture assay using an anti-p40 antibody to capture IL-12 and quantitating either proliferation (107) or IFN-.)Iproduction (108) by IL-12-responsive indicator cells as a measure of IL-12 biological activity represents an optimal method for the determination of biologically active p70. The use of the IL-12-dependent murine T cell clone 2D6 as indicator cells in this antibody capture assay has especially contributed to the reproducibility and sensitivity of mouse IL-12 detection (109). However, because the sensitivity of the antibody capture or immunological assays for either human or mouse IL-12 p70 is on the order of a few picograms and the p40 chain is often produced at a 100-fold excess over the p70 heterodimer, it is often difficult to quantitate p70 in supernatant and biological fluids that contain less than 1 nglml of IL-12 p40.
B. IL-12 PRODUCTION BY €3 LYMPHOCYTES IL-12 was originally discovered and characterized as a product of the lymphoblastoid B-cell lines RPMI-8866, ADP, and NC37 (19, 20). EBVtransformed BCL were all found to produce at least low levels of IL-12 constitutively,and its production was enhanced by stimulation with phorbol diesters (46). Most African Burkitt’s lymphoma cell lines produced no or negligible amounts of IL-12 (46, 76). However, most EBV(+) cell lines derived from AIDS-associated B-cell lymphomas (AABCL) constitutively produce very high levels of IL-12, which are enhanced by phorbol diester stimulation to levels much higher than those observed with BCL from
INTERLEUKIN-12
103
HIV-negative donors (76). In the case of AIDS-associated Burkitt’s Iymphoma (76), as well as in Hodgkin’s lymphoma (110), only EBV(+) cells were shown to produce IL-12, suggesting the EBV has a transactivating effect on IL-12 production. High levels of IL-12 were detected in the serum of SCID mice injected with human lymphocytes and in which EBV(+ ) human B-cell lymphomas were growing (R. Baiocchi, M. Caligiuri, G. Gri, and G. Trinchieri, unpublished results), suggesting that a similar production of IL-12 during the initial proliferation of EBF-transformed B cells in patients may affect the reactivity of the patient’s immune cells against the transformed cells. In addition to lymphomas, chronic B lymphocyte leukemia cells have also been shown to produce low levels of IL-12 (V. Pistoia and G. Trinchieri, unpublished results). Thus, IL-12 expression and/or secretion is not confined to in uitro-transformed normal B cells but can also be detected in malignant B-cell precursors “frozen” at early stages of B-cell differentiation pathway (early or intermediate B cells). Analysis of IL-12 mRNA accumulation has indicated that almost all cell types tested, including B cells, T cells, NK cells, and leukemic cell lines derived from these cells, as well as different malignant tumor cell lines, express transcripts of the p35 gene (46, 111).In the author’s experience, IL-12 p35 was expressed in the entire panel of the B-cell lines studied, whereas IL-12 p40 was detected only in the EBV(+) B-cell lines. The constitutive secretion of high levels of IL-12 by the majority of the EBV( +) AABCL, but only one of the non-AABCL obtained from patients with Burkitt’s lymphoma, raises the question whether IL-12 secretion by AABCL is related to HIV-1 and/or EBV. Modulation of lymphokine expression by virally encoded genes has been documented by Hsu et al. (112), who demonstrated that IL-10 has extensive homology to BCRF-1, an open reading frame in the EBV genome. In studying the B-cell lines for lymphokine expression, it was found that AABCL constitutively secrete large amounts of IL-10 (113),IL-7, and TNF (114-116). The secretion of large quantities of these lymphokines by AABCL compared with nonAABCL suggests that HIV-1 triggers B cells to secrete large amounts of IL-10, IL-7, TNF, and IL-12. However, HIV-1 does not infect B cells, and polymerase chain reaction (PCR) analysis revealed no HIV-1 transcripts in the AABCL. Moreover, lack of IL-12 expression in two EBV( -) cell lines derived from patients with AIDS and Burkitt’s lymphoma further suggests that in vivo exposure of B cells to HIV-1 alone may not induce lymphokine secretion and that both HIV-1 and EBV are required to trigger IL-12 secretion in tumor B cells. The observation that the AABCL secrete large amounts of IL-12 contrasts with data of several recent reports, which indicate that IL-12 production by monocytes from HIV-l-infected patients is impaired (117-119).
104
GIORGIO TRINCHIERI
Although any extrapolation to the in vivo situation from data obtained with established cell lines warrants caution, it seems possible that, in healthy individuals, IL-12-producing EBV-transformed cells are easily rejected by the immune response, leaving only cells that have lost the ability to produce IL-12 to give rise to Burktt’s lymphomas. In immunodeficient AIDS patients, these protective mechanisms may be inefficient, and IL-12producing EBV-transformed cells could give rise to Burkitt’s lymphomas in a relatively high proportion of patients. Although malignant or EBV-transformed cell lines produce IL-12, the physiological relevance of IL-12 production from normal B cells remains to be established. In situ detection of IL-12 mRNA in mice injected with LPS suggested that IL-12 was produced by B cells in addition to macrophages (104). In humans, only very low levels of IL-12 were found to be produced by peripheral blood B cells and by germinal center ( IgD -, CD38+) and naive (IgD+, CD38-), but not memory (IgD-, CD38-), tonsillar B lymphocytes (V. Pistoia, personal communication). Of the lowlevel IL-12 produced, most is secreted as p70 rather than the p40 chain, unlike the observations for B-cell lines and other cell types. Moreover, various stimuli that activate other functions of B cells have only a minimal effect on their ability to produce IL-12. Unlike in human B cells, no IL12 production has been demonstrated for murine B cells from immune lymph nodes (120). Thus, the physiological role of B-cell-produced IL-12 and the possible immunological effect of IL-12 produced by neoplastic B lymphocytes remain to be investigated. BY PHAGOCYTIC CELLS C. IL-12 PRODUCTION Phagocytic cells, not B cells, are probably the major physiologicalproducers of IL-12, a conclusion suggested by many in vitro and in vivo studies in infectious disease models (121). PBMC or purified monocytes produce high levels of IL-12 p40 and p70 when stimulated by bacteria, such as heat-fixed Staphylococcus aureus or Streptococcus extracts, or by bacterial products such as LPS (46). The producer cells within PBMC are mostly monocytes and other MHC class 11-positive cells, possibly dendritic cells (46). However, purified monocytes often produce lower levels of IL-12 than total PBMC, suggesting that other cells, e.g., T cells may contribute to the stimulation of monocytes to produce IL-12. In addition to monocytes, PMN also respond to LPS stimulation with the production of IL-12 p40 protein and, to a lesser extent, of the biologically active heterodimer (103). On a per cell basis, PMN produce less IL-12 p40 or p70 than monocytes (103). However, because of the large number of PMN present in the blood or in the inflammatorytissues, it is likely that IL-12 produced by PMN plays an important physiological role in the inflammatory response to infection.
INTERLEUKIN-12
105
Several studies have shown that both human and mouse central nervous system (CNS) microglial cells produce both IL-12 p40 and p70 in response to LPS or IFN-7 plus LPS (122-125). The ability of astrocytes to produce IL-12 is, however, controversial: two studies indicated that astrocytes produce even higher levels of IL-12 p40 and p70 than microglia (123, 124), whereas another study demonstrated no IL-12 production from astrocytes and reported that IL-12 production by microglia is inhibited by astrocytes (125). The production of IL-12 by phagocytic cells and, at least in part, by dendritic cells is induced by a variety of mechanisms that reflect the role played by IL-12 in inflammation and immunity. These mechanisms are either T-cell independent or dependent. The T-cell-independent mechanisms are important in the proinflammatory and inimunoregulatory role of IL-12 at the interface of innate resistance and adaptive immunity. Among these mechanisms is the induction of IL-12 by infectious agents and their products, of which LPS and bacterial DNA are the most typical examples (46, 126). During inflammation, however, an important mechanism of infection-independent induction of IL-12 and other cytokines is represented by the interaction of adhesion molecules with substrates of inflammatory origin, exemplified by the interaction of CD44 adhesion molecules with low molecular weight hyaluronan (127). Thus, although infection is probably the most effective mechanism of acute induction of IL-12 production, the CD44-mediated mechanism of IL-12 induction may be operative in aseptic inflammation, contributing to macrophage activation via the proinflammatory function of IL-12 and IFN-7 induction. The Tcell-dependent mechanism of IL-12 production is dependent on the ability of the CD40 ligand (CD40L) expressed on activated T cells to induce IL12 production in monocyte/macrophages and dendritic cells by interacting with the CD40 receptor on the surface of these cells (128, 129). The Tcell-dependent mechanisms of IL-12 induction play an important role in the T-cell immunoregulatory role of IL-12, particularly in the maintenance of Thl responses; however, it should be noted that nonantigen-specific T cells also participate in the innate resistance response, alongside or independently of antigen-specific T cells. In particular, functional CD40L expression on T lymphocytes in the absence of T-cell receptor engagement has been demonstrated to be involved in IL-12 and IFN-y production induced by IL-2 (130). To fully understand the regulation of IL-12 production, it is important to remember that the ability of various stimuli to induce its production is strictly regulated by powerful positive and negative feedback mechanisms. Such mechanisms are mediated by other cytokines, by pharmacologically active mediators, and by ligands for various cellular receptors, among
106
GIORGIO TRINCHIERI
which complement components and immunoglobulins assume particular relevance in the cross-regulation between innate and adaptive immunity compartments. 1. IL-12 Induction by lnfectious Pathogens Infectious pathogens that are able to induce production of IL-12 include bacteria, both protozoan and metazoan parasites, fungi, and viruses. IL12 is often involved in the regulation of the host response to these infections and many of these pathogens will be discussed in detail. Induction of IL-12 production with bacteria is observed with avirulent or heat-killed organisms, but, in uiwo, it is more efficient when live bacteria are used (46, 131).Fixed bacteria (e.g., S. aureus) or crude bacterial extracts (e.g., OK432 from Streptococcus pyogenes) are efficient inducers of IL-12 production in monocyte/macrophages (46, 132). The question of whether phagocytosis is either necessary or sufficient for the induction of IL-12 production is a complex one. Soluble microbial products such as purified protein derivative (PPD) or heat-shock proteins from mycobacteria are poorer inducers of IL-12 than the whole organisms, although they readily induce IL-1 or TNF (131).Chitin particles but not soluble chitin (polymers of N-acetyl-D-glucosamine) induce IL-12 production via mannose receptor-mediated phagocytosis and this induction is inhibited by soluble mannan (133).Cytochalasin D blocks both phagocytosis and IL-12 induction by chitin particles or Mycobacterium tuberculosis (131,133).However, discordant results have been reported for the production of IL-12 following latex bead phagocytosis (131, 134). The reason for these contrasting results might rest in the size of particles used: beads <1 p m in diameter, which are internalized by pinocytosis, do not induce IL-12 production, whereas beads >2 pm, which are phagocytosed, readily induce IL-12 production, probably by activating kinases associated with the cytoskeletal proteins, without triggering specific surface receptors ( 134). Several soluble microbial products have been identified that induce IL12 production. LPS from gram-negative bacteria is a potent stimulus for IL-12, although maximal activity requires priming of the producer cells, e.g., by IFN-y (32, 46, 135).The response to low-level LPS is mediated through CD14 receptors: dendritic cells, which do not express membrane CD14, produce IL-12 in response to LPS by utilizing serum-derived soluble CD14 (136). Lipoteichoic acid (LTA), a member of a class of surface glycolipids similar to LPS but present on gram-positive bacteria, is a potent inducer of IL-12 production and, like LPS, acts through a CD14-dependent pathway (137). Other microbial IL-12 inducers include trehalose dimycolate, a glycolipid from Mycobacteriurn spp. (138), glycoprotein fractions from Toxoplasma gondii (139) and Typanosomu cruzi (140), and the
INTERL,EUKIN-lZ
107
recombinant Leishmania antigen LeIF (141). Bacterial superantigens such as Staphylococcus or Streptococcus spp.-derived enterotoxins induce production of IL-12 both in vitro and in vim (142). The major mechanism of IL-12 induction by superantigens, however, appears to be induction of CD40L expression on activated T cells and not direct stimulation of macrophages (C. Son and G. Trinchieri, unpublished results). Bacterial DNA has also been shown to be an efficient inducer of IL-12 and other monokines, including IFN-a, due to the presence of immunostimulatory sequences composed of an unmethylated CpG dinucleotide flanked by two 5’ purines and two 3‘ pyrimidines (143,144).Interestingly, the presence of these immunostimulatory sequences in the plasmids has been shown to be essential for effective intradermal gene immunization (126). Another powerful inducer of IFN-a production, i.e., synthetic double-stranded RNA such as poly I-C, also induces IL-12 production (131, 145).
2. IL-12 Induction by Interaction with Injam?mtoy Extracellular Matrix The second T-cell-independent pathway of IL-12 induction is that induced by interaction of macrophages with components of the extracellular matrix selectively expressed during inflammation ( 127). Low molecular weight fragments (<5 X 10’) of hyaluronan, an extracellular matrix glycosaminoglycan that accumulates in inflammation, are potent inducers of IL12, IL-1, and various chemokines, whereas the ubiquitous high molecular weight form (>lo6)is ineffective (127).The induction of IL-12 production was inhibited by antibodies against the CD44 receptor that block hyaluronan binding or by competition with the high molecular weight form (127). Production of IL-12 was observed in thioglycolate-elicited macrophages or in adhesion-primed but not freshly expIanted resident macrophages and was strongly upregulated by treatment of the macrophages with IFN-y (127). The CD44-dependent mechanism of IL-12 induction might play an important role in the sterile inflamniatory response. For example, in patients with closed bone fractures and soft tissue hematomas following blunt trauma with no evidence of infection, high levels of IL-12 (>2 ng/ml) were present in the plasma and in the fracture serum from days 3 to 7 after trauma (146). Such levels of IL-12, together with other inflammatory cytokines, likely contribute to the alteration of systemic immunity observed in trauma patients. 3. T-Cell-Dependent IL-12 Induction Evidence for a T-cell-dependent pathway of induction of IL-12 production can be traced back to the work of Germann et al. (147),who reported that T-cell stimulatory factor (TSF), a cytokine required for the growth
108
GIORGIO TRINCHIERI
of T h l lymphocyte clones and later identified as IL-12 (81),was produced by accessory cells cocultured with activated T cells. The ability of murine dendritic cells to direct the development of T h l cells was demonstrated to be secondary to their ability to produce IL-12, which was induced in the dendritic cells only when exposed to T cells activated by specific antigen (148). Shu et al. (128) clarified the mechanism of T-cell-dependent IL-12 production by demonstrating that activated T cells induce IL-12 production by monocytes via CD40/CD40L interaction. Anti-CD40L antibodies inhibit IL-12 production in vivo (149),possibly explaining the ability of antiCD40L antibodies to suppress many in vivo models of pathogenesis involving T cells and also the T-cell abnormality observed in patients with X-linked hyper-IgM syndrome due to defects in the CD40L gene (150,151).However, CD40/CD40L stimulation is most likely bidirectional, with stimulation of both T cells and APC; APC produce IL-12 and upregulate B7, whereas T cells, in response to IL-12 and B7 stimulation, produce IFN-y and GM-CSF which, in turn, enhance APC ability to produce IL12 (152,153). Only one study (154) has analyzed the regulation of the two IL-12 transcripts by CD40/CD40L interaction and reported that p40, but not p35, transcripts are upregulated; however, the observation in many studies that the production of the p70 heterodimer is induced by CD40/ CD40L interaction as or more efficiently than that of the p40 chain (129, 155)points to an effective upregulation of the production of both polypeptide chains. B cells interfere with the ability of activated T cells to stimulate IL-12 production by APC, possibly because they express CD40 and thus compete for binding to CD40L on T cells (155). The ability of infectious agents to induce IL-12 production from APC is independent of CD40/CD40L interaction and of T cells, as clearly shown by the finding that LPS, S. aureus, and Listeria monocytogeneses can all induce IL-12 production in CD40 knockout mice (155)or in systems where CD40/CD40L interaction is blocked by soluble CD40L or anti-CD40L antibodies (156).However, it is likely that during an infection in viuo, the expression of CD40L on activated T cells contributes to maintaining the production of IL-12 initiated by the T-cell-independent mechanisms. Because CD40L is induced within a few hours of T-cell stimulation (157) and because stimulation of T cells by specific antigens or by other stimuli, e.g., IL-2 even in the absence of T-cell receptor engagement (130)upregulates functional CD40L expression, both antigen-specific and bystanderactivated T cells in an inflammatory response can efficiently participate in the induction of IL-12 production.
D. POSITIVE AND NEGATIVE MODULATION OF IL-12 PRODUCTION IFN-y has a powerful enhancing effect on the ability of monocytes and PMN to produce IL-12 (103,135).This observation is of particular interest
INTEHLEUKIN-18
109
because IL-12 is a potent inducer of IFN-y production by T and N K cells (78).Thus, IL-12-induced IFN-y acts as a potent positive feedback mechanism in inflammation by enhancing IL-12 production. Also, because IL-12 is the major cytokine responsible for the differentiation of Thl cells, which are potent producers of IFN-y (79), the enhancing effect of IFNy on IL-12 production may represent a mechanism by which T h l responses are maintained in vivo. In both monocytes and PMN, the enhancing effect of IFN-y on IL-12 production is observed when IFN-y is added simultaneously to the stimulus (e.g., LPS), but it is more effective when the producer cells are primed for several hours in the presence of IFNy (32,158).In addition to IFN-y, GM-CSF has a modest enhancing effect on IL-12 production by phagocytic cells, acting primarily at the level of the p40 gene (135, 158, 159), whereas M-CSF has no priming activity for IL-12 production (158).The ability of IFN-y to enhance IL-12 production is particularly evident in the case of infectious agents, e.g., certain mycobacteria, which are rather poor inducers of IL-12 production. In in vitro or in vivo infections with these microorganisms, IFN-y production precedes and is required for maximal IL-12 production (160). However, with many other inducers, such as LPS, toxoplasma, and S. aureus, IL-12 production in vivo and in vitro both precedes and is required for IFN-y production. For example, following injection of LPS, IL-12 is induced at 2-3 hr and IFN-y at 5-7 hr (161, 162). Neutralizing anti-IL-12 antibodies inhibit IFN-y production, but anti-IFN-y antibodies do not inhibit IL-12 production. With these potent inducers, IL-12 production is also observed in mice not expressing the IFN-y or IFN-y-receptor genes (163, 164). The role of TNF-a in the production of IL-12 is less well defined; in some experimental systems, it has been reported to enhance the ability of IFNy to prime phagocytic cells for IL-12 production (122, 1Fi6, 160). The positive feedback amplification of IL-12 production mediated by IFN-y obviously represents a potentially dangerous mechanism leading to uncontrolled cytokine production and possibly shock. There are, however, potent mechanisms that downregulate IL-12 production and the responsiveness of T and NK cells to IL-12. Some of the major positive and negative regulatory mechanisms of IL-12 production and function are schematically represented in Fig. 3. The Th2 cytokine IL-10 is a potent inhibitor of IL-12 production by phagocytic cells; the ability of IL-10 to suppress production of IFN-y and other T h l cytokines is due primarily to its inhibition of IL-12 production from APC and to inhibition of expression of other costimulatory surface molecules (e.g., B7) and soluble cytokines (e.g.,TNF-a, IL-lP) (108,165,166). However, IL-12 is able to induce IL-10 production and to prime T-cell clones for high IL-10 production both in vivo (167) and in vitro (168), indicating that IL-12 can induce factors
110
GIORGIO TRINCHIERI
FIG. 3. Major positive and negative regulatory mechanisms of IL-12 production and functions. Three major pathways of IL-12 production are effective in infection, inflammation, and immune response. Positive feedback (e.g., IFN-y) and negative feedback (e.g., IL10) mechanisms regulate IL-12 production. IL-12 production is also downregulated by pharmacological mediators such as PGE2, desensitization of the producer cells by LPS, and cross-linking of various complement and Fc receptors on the producer cells. The competition of the P40-homolog protein EBI3 with p40 for binding to p35 might represent a mechanism for downregulation of IL-12 heterodimer formation. For full biological activity on T and NK cells, IL-12 synergizes with various costimulators, including cytokines such as IL-2 and IL-18, and surface costimulatory molecules such as B7 binding to CD28 and LFA-3 binding to CD2. Competition of the p40 homodimers with the IL-12 heterodimer for receptor binding and downregulation of the IL- l2RP2 chain represent mechanisms of downregulation of the biological response of T and NK cells to IL-12. @, positive or synergistic stimulation; 8, antagonistic or inhibitory stimulation.
that enhance (e.g., IFN-y) or suppress (e.g., IL-10) IL-12 production. Another powerful inhibitor of IL-12 production is TGF-P (169). IL-4 and IL-13 can also partially inhibit IL-12 production (169, 170), suggesting the hypothesis that Th2 cells, by producing cytokines such as IL-10, IL-4, and IL-13, suppress IL-12 production and prevent the emergence of a T h l response (171,172).However, if monocytes are primed with IL-4 or IL-13 for 24 hr or longer, IL-12 production is not inhibited and instead is significantly enhanced (169, 173). The mechanism of enhancement of IL-12 production by IL-4 and IL-13 is not due to increased production of IL-10 or prostaglandin Ez (PGEJ (173, 174) and may be secondary to a differentiation effect on monocytes, which requires prolonged incubation and exposure to the cytokine, unlike the inhibitory
INTERLEUKIN-12
111
effect observed when the cytokines are added simultaneously to the IL12 inducers. The priming effect of IL-4 and IL-13 is at the transcriptional level for both p40 and p35 genes. At the protein level, the effect is observed for secretion of the IL-12 p40 chain and, even more efficiently, of the p70 heterodimer (174). IL-4 and IL-13 are particularly effective in enhancing IL-12 production in response to particulate inducers (S. aureus and L. major), whereas IFN-y is most active when LPS is used as an inducer (159, 169); however, IL-4/13 and IFN-y have an additive and, in some cases, a synergistic effect on IL-12 production (159, 173). The enhancing effect of IL-4/13 on IL-12 production is accompanied by a similar increase of TNF-a, MCP-1, and, to a lesser extent, IL-6, whereas production of IL-lP, IL-8, and IL-10 is still inhibited (169, 173). Furthermore, IL-4and IL-13-treated monocytes have significantly increased expression of MHC class I1 antigens, B7.2, and CD40, and decreased CD14 (174). Some of these surface antigen changes are reminiscent of those observed in longer cultures (7 days) of monocytes in the presence of GM-CSF and IL-4 or IL-13, which generate dendritic-like cells with much enhanced antigen-presenting activity and the ability to produce elevated levels of IL-12 (129,175). Interestingly, IL-4, which inhibits IL-12 induction when added simultaneously with a bacterial stimulus, specifically upregulates IL-12 p70 production induced by CD40L through the upregulation of IL12 p35 gene expression (176). Because IL-13 receptors, unlike those for IL-4, are not expressed on T cells, and, unlike IL-4, IL-13 is not an inducer of Th2 differentiation, the ability of IL-13 to enhance IL-12 production makes it a potential inducer of T h l responses. Indeed, it was shown that injection of IL-13 in mice infected with L. rnonocytogeneses increases IL12 production, decreases IL-4 production, and enhances resistance to the infection (177). Interestingly, treatment of PBMC from HIV( +) patients with IL-4 or IL-13 almost completely corrects their inability to produce IL-12 in response to S. aureus (174) and, in part by enhancing the deficient antigen-presenting ability of the patients’ monocytes, corrects the defective proliferative responses of their T cells to recall antigens (L. J. Montaner, personal communication). IFN-y, IL-10, and other cytokines have a profound modulatory effect on the production of IL-12 by phagocytic cells, but they have no significant effect on IL-12 production by B-cell lines. It is also of interest that phorbol diesters, which enhance IL-12 production from EBV-transformed cell lines, do not induce IL-12 production in phagocytic cells, although these compounds are potent inducers of other cytokines such as TNF-a in those cells (46, 135) Other cytokines that downregulate IL-12 production are MCP-1 (178) and IL-11 (179, 180). In addition to IL-11, another cytokine of the group
112
GIORGIO TRINCHIERI
sharing the use of the gp130 subunit in their receptor, IL-6, has been shown to inhibit both T-cell independent and -dependent induction of IL-12 production (176). Among the substances with pharmacological activity that inhibit IL-12 production, and of particular interest because they may be involved in physiologic immune regulation, are PGEz (181), corticosteroids, and catecholamines (135,182- 184). The presence of high levels of PGEz in the seminal plasma may be responsible for its immunosuppressive effect (185, 186). The immunosuppressive hormone, 1,25dihydroxyvitamin D3 also inhibits IL-12 production and T h l functions (187), whereas a deficiency of vitamin A in vivo leads to constitutive IL12 expression with excessive Thl cell activation and insufficient Th2 cell development, which is corrected by dietary retinoic acid supplementation (188). The immunosuppressive calcitonin gene-related peptide released from peripheral nerves acts, in part, by suppressing production of IL-12 (189, 190); interestingly, IL-12 itself has been detected in the free nerve endings of skin, and it has been suggested that modulation of IL-12 production may be one mechanism whereby the nervous system modifies cutaneous immune responses (191). Pentoxifylline and thalidomide are inhibitors of IL-12 production, suggesting that these substances might be used pharmacologically in clinical trials of immunological disorders characterized by an inappropriate type 1 immune response (192, 193). An important mechanism of cross-talk between innate resistance and adaptive immunity is represented in the ability of complement components and immunoglobulins to react with complement or Fc receptors on the effector cells of innate resistance, modulating their functions and cytokine production. The first indication that IL-12 production is also regulated by those interactions was provided by the finding that measles virus induces a profound depression of Thl responses following infection in part by interacting with its receptor on phagocytic cells, the CD46 molecule, to inhibit IL-12 production (194). CD46, or membrane cofactor protein, is a binding site for C3b and C4b. By binding CD46, both polymeric C3b and anti-CD46 antibodies can inhibit IL-12 production, suggesting that measles virus may utilize a physiological mechanism of IL-12 downregulation induced by activated complement in its induction of immunosuppresion (194). Note that this suppressive mechanism is quite selective, as the production of many other cytokines tested is only minimally or not at all affected (194). The ability to trigger monocyte/macrophage receptors and selectively suppress IL-12 production was extended to the interaction of iC3b with complement receptor 3 ( C D l l b ) and to the interaction of immunocomplexes with Fc receptors (195,196). The mechanism of action is not clear but may depend on [Ca2+Iiflux activation by receptor crosslinking (195) and, at least in part, on inhibition of IFN-y-induced tyrosine
INTERLEUKIN- 12
113
phosphorylation (196). Antibodies against C D l l b in a inurine model of IL-12-dependent septic shock lead to suppression of IL-12 and IFN-y production in vizjo (196),suggesting that previous results showing suppression of DTH and response against various infectious agents by anti-CDllb in vivo might be explained in part by the inhibition of IL-12 production. C D l l b , in addition to binding iC3b, is a receptor for a variety of ligands and mehates the binding of several bacteria and intracellular parasites, either opsonized or not: thus, its ability to selectively downregulate IL12 production may have physiological importance in many infections or inflammatory situations (195, 196).
E. PRODUCTION OF IL-12 BY DENDRITIC CELLSAND OTHER CELLTYPES In addition to phagocytic cells and B lymphocytes, other cell types have been shown to produce IL-12. Mast cells derived in vitro from mouse bone marrow in the presence of mast cell growth factors and considered representative of connective-type mast cells produce IL-12, whereas IL3-derived mucosal-type mast cells produce IL-4 (197). These data, which suggest the existence of different types of mast cells that favor Thl or Th2 differentiation, await confirmation with data from in vivo-differentiated mast cells. Other cell types reported to express IL-12 mRNAs and possibly secrete minute levels of IL-12 protein are keratinocytes and epidermoid carcinoma cell lines (198, 199). Exposure to hapten or UV irradiation induces IL-12 in keratinocytes (199,200),and anti-IL-12 antibodies induce a 50% inhibition of proliferation of human allogeneic T cells stiinulated using human epidermal cells containing Langerhans cells as APC (199). However, the physiological significance of this production is doubtful; when used as APC, keratinocytes induce no stimulation of IFN-y production even when CD28 costimulation is provided by anti-CD28 antibodies unless exogenous IL-12 is added to the cultures (201). Also, analysis of IL-12 production by skin cells suggests that Langerhaiis cells rather than keratinocytes are the major IL-12 producers (202). The production of IL-12 by Langerhans cells (202) raises the issue of the ability of professional APC such as dendritic cells to produce IL-12 and the role of this production during antigen presentation and T-cell activation. Earlier data on the production of IL-12 by PBMC showed that in addition to inonocytes and B cells, other MHC class II-positive cells were also responsible for IL-12 production (46). Kanangat et al. (203) showed that LPS treatment of highly purified mouse dendritic cells induces expression of IL-12 p40 inRNA. Definitive evidence that dendritic cells are producers of functional IL-12 came from studies demonstrating that these cells, when used as APC, induce a T h l response if endogenous
114
GIORGIO TRINCHIERI
IL-4 production is blocked, and that this Thl response is prevented by neutralizing anti-IL-12 antibodies (148). The production of IL-12 by dendritic cells was directly confirmed by immunocytochemistry, and stimulation of IL-12 production was found to require exposure to T cells in the presence of the specific antigen (148), possibly via CD40/CD40L interaction. Extensive studies with both human and mouse dendritic cells have now confirmed that dendritic cells are efficient producers of the IL12 that acts in inducing Thl responses upon antigen presentation by these APC (129,148,204). Interaction of CD40L on activated T cells with CD40 ligand on dendritic cells appears to be an important mechanism of IL-12 induction during antigen presentation (129, 205, 206); in addition, ligation of MHC class I1 molecules on the dendritic cells also results in some induction of IL12, suggesting a secondary mechanism of IL-12 induction that could play a role during antigen presentation or superantigen stimulation (206). In addition to the CD40/CD40L pathway, IL-12 production in dendritic cells is also activated by LPS, with involvement of a soluble CD14-dependent pathway (136), and by phagocytosis of bacteria (175) and microparticles (207). Interestingly, uptake of microparticle-adsorbed protein antigen by mouse bone marrow-derived dendritic cells results in de nuvu synthesis of transcripts for IL-12 and MHC class I1 and triggers prolonged, efficient antigen presentation (207). Both in the mouse (208) and in humans (209), IL-10 prevents the maturatioddifferentiation of dendritic cells, inducing the generation of cells with decreased ability to produce IL-12 and to induce a Thl-type of response, leading to the development of Th2 lymphocytes. IL-10 also downregulates the production of IL-12 by purified mouse spleen dendritic cells (206). Like IL-10, the presence of PGEz in cultures induces the generation of dendritic cells defective in IL-12 production and promoting Th2-type responses in T cells (210). The production of IL-12 appears to be maximal in certain stages of maturatiodactivation of dendritic cells. In mouse spleen dendritic cells, cultured in the presence of GM-CSF, IL12 production was observed only in the mature form of dendritic cells, obtained by treatment of the culture with TNF-a, IL-1, or LPS (211). Studies in FLT3 ligand-treated mice, which have dramatically increased dendritic cell numbers of both the myeloid type (CDllb bright, F4/80+, and LyG-C+) and the lymphoid type (CDllb dull or negative, CD8a+, CDld+, CD23+), have detected IL-12 production in response to IFNy plus S. aureus in the lymphoid dendritic cells only (212). V. Molecular Control of 11-12 Gene Expression
As with immunological and biological detection studies, analysis of the molecular control of IL-12 production is complicated by the need to analyze
INTERLEUKIN-12
115
the expression of two genes. At present, much more information is available on the p40 gene, which is highly inducible and expressed only in IL-12 producing cells, compared with that of the more ubiquitously expressed p35 gene. Upon activation of phagoc$c cells with LPS or S. aweus, accumulation of IL-12 p40 mRNA is observed within 2-4 hr, i.e., slightly delayed compared with that of other proinflammatory cytokines such TNFa,and subsides after several hours (46). Expression of the p35 gene is also upregulated on activation of phagocytic cells, although its ubiquitous constitutive expression has complicated analyses of its expression using nonpurified cell preparations (108, 111, 166) . In some studies (32, 103), but not in others (158),IFN-y has been shown to directly induce transcription and mRNA accumulation of the p35 gene, unlike the p40 gene on which IFN-y has a priming, but not a directly inducing effect. In activated phagocytic cells (both monocytes/macrophages and PMN) and in B-cell lines, p40 mRNA is approximately 10-fold more abundant than p35 mRNA, explaining the overproduction and secretion of the free p40 chain over the p35-containing biologically active heterodimer (21, 32, 103). The lower abundance of p35 mRNA would lead to the obvious conclusion that p35 gene expression is limiting for the production of the p70 heterodimer (213). However, a dramatic upregulation of the production of IL12 p70, e.g., in response to IFN-y and LPS, is often observed in conditions in which only a modest upregulation of p35 mRNA is evident, suggesting that translational and posttranslational mechanisms regulating the assembly of the heterodimer may play a role (32). A potential difficulty in the interpretation of data on mRNA accumulation is that the kinetics of p35 and p40 mRNA accumulation may differ; in particular, optimal priming for p35 mRNA accumulation requires longer preincubation with IFN-y (8-24 hr) than does p40 (2-8 hr) (158).Although upregulation of both p40 and p35 mRNA is observed and is probably necessary in most of the conditions in which increased production of p70 heterodimer takes place, upregulation of p40 mRNA alone, as observed in cells primed with GMCSF, is not sufficient for increased production of the p70 heterodimer (32,158,159).The enhancing effects of IFN-y and IL-13 on IL-12 production are much more relevant for the bioactive p70 heterodimer than for the p40 subunit (32,158,159,174,213), suggesting a physiological relevance in viva of the effect of these modulating cytokines. Because of the possibility that, at least in the mouse, the p40 subunit represents an antagonist of IL-12 bioactivity, the ability of IFN-y, IL-4, and IL-13 to modify the ratio between p40 and p70 is of particular interest; analogously, the inhibitory effect of IL-10 on IL-12 production was reported to be more powerful on the production of the p70 heterodimer than on the p40 subunit (213). A detailed molecular analysis that simultaneously examined nuclear transcription, steady-state mRNA, and secreted protein levels of IL-12 estab-
116
GIORGIO TRINCHIERI
lished that the human IL-12 p40 gene is primarily regulated by IFN-7 and LPS at the transcriptional level in monocpc cells (32). Both the human and the mouse IL-12 p40 gene promoters have been cloned (32, 214). The 3.3-kb human p40 promoter, linked to a luciferase reporter gene and transfected transiently into various IL-12-producing and nonproducing cell lines, largely recapitulated the cell specificity of the endogenous p40 gene, i.e., it is constitutively active in EBV-transformed B cell lines (e.g., RPMI-8866, CESS) and inducible in myeloid cell lines (e.g., THP-1 and RAW264.7), but inactive in T-cell lines (e.g., Molt-13 and Jurkat) (32). Moreover, this promoter construct responds to IFN-7 priming in monocytic cells, much like the endogenous p40 gene transcription, suggesting that it contains sufficient sequence elements to reconstitute the in vivo response. Comparison of the human and mouse IL-12 p40 promoters revealed several interesting features. A schematic representation of the IL-12 p40 promoter and of the elements involved in its regulation is depicted in Fig. 4. The promoters are well conserved up to -400 with respect to the transcription start site, where the homology breaks down with large gaps between them. Within the -400 proximal promoter region, several putative
FIG.4. cis and trans elements involved in the regulation of the IL-12 p40 promoter in quiescent phagocytx cells (top) and after activation with IFN-y and LPS (bottom).
INTERLEUKIN-12
117
transcription factor-binding motifs are very well conserved: ets at -211/206, PU.1 and NFKB between -124 and -105. Functional characterization of the human p40 promoter in myeloid cell lines has identified the ets element, TTTCCT (AGGAAA for the complement), as a major response region. This element interacts with a large nuclear complex named F1 that binds to a region between -196 and -292. By electrophoretic mobility shift assay (EMSA) and DNase I footprintlmethylation interference assays, it has been established that F1 (1) is induced by LPS or IFN-y in RAW264.7; (2) interacts with the ets-2 element within the -211/-206 region in a complex way, i.e., the interaction requires substantial flanking “anchoring” space; (3) may function as a transcription activator in response to IFN-y and LPS stimulation, as loss of binding results in dramatically decreased promoter activity; (4) is composed of multiple factors including ets-2, IRF-1, c-Rel, and a novel 109-kDa protein that is highly induced by either IFN-y or LPA; and (5) closely correlates with IL-12 p40 gene expression in various cell lines and primary human monocytes (215). A second factor that also interacts with this region but requires less physical space is a complex formed with a fragment derived from -196 to -243 of the p40 promoter, named F2. F2 appears to be more responsive to IFNy stimulation than to LPS, yet its identity remains to be established. Interestingly, the ets element in unstimulated RAW264.7 cells is occupied by PU.l, which becomes displaced by F1 on IFN-y or LPS stimulation (215). The regulation of IL-12 p40 gene transcription in the EBV-transformed B cell line RPMI-8866 appears to be somewhat divergent from that of monocytic cells. The transfected p40 promoter is constitutively active, paralleling the endogenous gene. The nuclear complex F1 is also constitutively present, but its role in the regulation of p40 gene transcription does not seem to be as prominent as in monocytic cells since elimination of the F1-binding element results in only a 30-50% decrease of the promoter activity (G. Gri, G. Trinchieri, and X. Ma, unpublished results). The composition of F1 in RPMI cells also differs from that of monocytic cells in that IRF-2, instead of IRF-1, is present. The implication of the differing composition of F1 is not clear at present. Another region of potential transcriptional regulation is the NFKB half site located between -116 and -106, TGAAATTCCCC (or GGGGAATTTCA for the complement). This site in the mouse IL-12 p40 promoter has been reported to bind NFKB (p50/p65 and p50/c-Rel) in macrophages activated by a number of IL-12-activating pathogens, includmg LPS and S. aureus (214). In EBV-transformed B cells, NFKB constitutively binds to this site (G. Gri, G. Trinchieri, and X. Ma, unpublished results). The NFKBcomplex is composed of c-Re1 and p50 heterodimers. Base substitu-
118
GIORGIO TRINCHIERI
tions at this site, which abolish the NFKB binding, result in a -80% decrease in the constitutive promoter activity in B-cell lines. Cotransfection experiments with various combinations of expression vectors containing cDNAs for NFKBp65, p50, c-Rel, and ets-2 demonstrated that ets-2 and c-Re1 synergistically activate the transfected p40 promoter in both IL-12 p40-expressing cells (RPMI-8866) and nonexpressing cells such as Bjab [EBV(-)B cell line] and Jurkat (T cell line) (G. Gri, G. Trinchieri, and X. Ma, unpublished results), strongly suggesting that c-Re1 and ets-2 are the transcription factors necessary and sufficient to determine the cell typespecific expression of the p40 gene. Plevy et al. (216) identified a third element located between -96 and -88 of the murine p40 promoter that is conserved in humans (between -81 and -73) and interacts with members of the C/EBP family of transcription factors that are inducible in the murine macrophage RAW264.7 cell line by heat-killed L. mnocytogenes. The C/EBP element exhibits functional synergy with the NFKBsite upstream (216).These data are consistent with the deficient production of IL-12 in C/EBP P-deficient mice in response to Candida albicans infection (217). Mice with a disrupted gene encoding the interferon consensus-binding protein ( ICSBP) are highly susceptible to infection with intracellular pathogens such as L. mnocytogenes and T. gondii and have defective constitutive and inducible IL-12 expression due to a selective deficiency in IL-12 p40 gene expression (218, 219). ICSBP is a transcription factor of the IRF family, which is expressed exclusively in cells of the immune system, including macrophages, unlike other more ubiquitous IRF family members (220). An ICSBP consensus site is upstream of the Ets site in the IL-12 p40 promoter, and cotransfection of ICSBP and an IL-12 promoter luciferase reporter in RAW264.7 cells dramatically increases both constitutive and IFN-.)I plus LPS-induced expression of IL-12 promoter activity (221). Although evidence of IRF-1 binding to the IL-12 p40 promoter has been reported (215), cotransfection of IRF-1 with the promoter is not sufficient to enhance promoter activity (221). A role for IRF-1 in the control of IL12 is further suggested by the deficient production of IL-12 in IRF-1deficient mice, although the complex immunological alterations in those animals, including the impaired response to IFN-.)I, make it difficult to establish whether IRF-1 acts directly on the IL-12 gene promoter or is indirectly required for the action of other factors that regulate the IL-12 promoter, such as IFN-.)I itself (222, 223). Analysis of IL-12 p35 gene expression is hindered not only by its ubiquity, but also by its low activity and inducibility. Comparative studies with cycloheximide (CHX) demonstrated fundamental differences in mRNA regulation of IL-12 p40 and p35 genes. The increase in S. aureus- or LPSinduced IL-12 p40 mRNA levels was abrogated when cells were pretreated
INTERLEUKIN-12
119
with CHX, suggesting that the regulation of the IL-12 p40 gene requires the induction of a CHX-sensitive, transcription activators(s). Conversely, IL-12 p35 mRNA was further upregulated by CHX, indicating that the activation of IL-12 p35 mRNAs requires only a presynthesized activator(s) that can be activated by either S. nureus or LPS at the posttranslational level. Superinduction of cytokine genes such as TNF-a, IL1-P (224), IFN7 , and IL-2 (225, 226) was observed when cells were induced in the absence of CHX for about 2 hr followed by the addition of CHX. IL-12 p40 and p35 steady-state mRNA levels also underwent superinduction when CHX was added 2 hr after S. aureus stimulation (227). The promoter of the mouse p35 gene contains putative elements, including Spl, AP1, ISRE, ICSBP, NFKB, GATA-1, and GAS (30, 31). Unlike the p40 gene, the p35 gene appears to initiate its transcription from multiple sites. The nature of these alternatively initiated transcripts with respect to their cell-type distribution and response to different stimuli remains to be investigated. Based on increasing numbers of observations, it appears that under certain conditions the p35 chain may be a rate-limiting and critical factor in determining the level of IL-12 p70 production by altering either its level of expression or its posttranslational modification in response to specific inducers of IL-12-producing cells, which would affect its association with the p40 chain (213). The Th2 cytokine IL-10 is a potent inhibitor of IL-12 production by phagocytic cells. Its ability to suppress production of IFN-.)Iand other T h l cytokines is due primarily to its inhibition of IL-12 production from APC and to inhibition of expression of other costimulatory surface molecules (e.g., B7) and soluble cytokines (e.g., TNF-a, IL-P) (108,165,166).Studies (227) on the effect of IL-10 on S. nureus- or LPS-induced IL-12 p40 and p35 gene expression in PBMCs and monocytes demonstrate that IL-10 inhibition of IL-12 production is accompanied by reduced steady-state mRNA levels of the p40 and p35 components of the heterodimeric cytokine. The mechanism(s) of IL-10 suppression of IL-12 p40 appears to be exerted mainly at the level of transcription, without significant modulation of mRNA stability. The transcriptional activity of IL-12 p35, primed by IFN-y and induced by LPS, was also substantially inhibited by IL-10. The tlp2 of S. aureus-induced IL-12 p40 was -4 hr and not altered by IL-10. It was also obsewed that CHX abolished the inhibitory effect of IL-10 on the induction of IL-12 p40, IL-12 p35, and TNF-a mRNAs. These findings, together with other reports (228-230), suggest that IL-10 may exert its negative effect through a newly synthesized repressor protein(s). VI. 11-12 Effects on Hematopietic Stem Cells
IL-12 by itself has not been described to affect the growth or differentiation of hematopoietic stem cells. However, in several experimental condi-
120
GIORCIO TRINCHIERI
tions, IL-12 reportedly synergizes with other growth factors, in particular IL-3 and stem cell factors (SCF),in supporting the formation of hematopoietic colonies. Thus, IL-12, together with IL-1, IL-6, LIF, and SCF, belongs to the group of synergistic hematopoietic growth factors that alone have little or no in vitro effect on proliferation and act predominantly to enhance the growth-promoting activity of other colony-stimulating factors (231). IL-12 at a concentration of several nanograms per milliliter with maximal activity at 10-100 ng/ml enhances both colony number and size of purified murine hematopoietic Lin- Sca-1' progenitor cells when added in culture together with other growth factors, particularly IL-3, SCF, and FLT3 ligand but also IL-4, G-CSF, and M-CSF (232-236). Single cell cloning experiments suggested that the stimulatory effects of IL-12 on the LinSca-1' cells were directly on the progenitor cells and not indirectly through cytokine production of potentially contaminant accessory cells (232). The cells that formed in response to IL-12 plus SCF were predominantly granulocytes and macrophages, although blasts were also present (232). In the presence of erythropoietin (EPO) and either SCF or IL-4, IL-12 enhances the growth of erythropoietic colonies (235). Progenitor cells enriched from bone marrow of mice treated with 5-fluorouracil in liquid cultures were used to show that IL-12 and SCF or IL-3 synergize in regulating survival and growth of myeloid stem cells and progenitor cells, including primitive long-term culture initiating cells that probably correspond to multipotent stem cells with long-term in vivo repopulating activity; lymphohematopoietic progenitor cells able to yield pre-B-cell colonies were also detected on replating in secondary culture containing SFC and IL-7 (234, 237). Very similar results were reported in the human system, where IL-12 was observed to synergize with IL-3 and SCF in enhancing the growth and colony-formingability of immature progenitor cells to differentiate into granulocytes, macrophages, or erythroid cells (238-241). The stimulatory effect of IL-12 was not observed on hematopoietic cells that were secondarily plated after a 48-hr liquid culture in the presence of IL-3, suggesting that IL-12 is active only on the most primitive progenitor cells (240). A requirement for serum or accessory cells in order to detect the hematopoietic effect of IL-12 suggests that an additional factor(s) present in the serum or produced by the accessory cells may be required for IL-12 action (240, 241). It is not known at present whether the hematopoietic progenitor cells express the IL-12 receptor. The fact that IL-12 is effective on single cells or on cultures with a very low number of cells suggests that IL-12 acts directly on the progenitor cells and not on contaminant cells, thus arguing for the presence of the receptor on these cells. However, whether the
INTERLEUKIN-12
121
receptors are constitutively expressed on the progenitor cells or induced
by the synergistic growth factor required for detecting the effect of IL-12 remains to be determined. When hematopoietic colony formation is analyzed on preparations of progenitor cells containing NK cells, an inhibitory effect of IL-12 on colony formation is observed due to the secretion by the NK cells of the hematopoietic inhibitory factors IFN-y and TNF-a, which synergize in blocking colony formation (238).Elimination of NK cells from the progenitor cell preparations or neutralization of IFN-y and TNF-a eliminates this inhibitory effect and reveals the costimulatory effect of IL-12 on colony formation (238). Thus, in vitro, IL-12 has a direct stimulatory effect on hematopoietic progenitor cells and an indirect inhibitory effect mediated by the induction of suppressive factors from contaminating cells (238). Similarly, it was shown that the recovery of hematopoietic stem and progenitor cells from liquid cultures of bone marrow cells was increased by IL12; however, if the cultures were stimulated by IL-12 plus IL-2, which synergize in the induction of IFN-y and other inhibitory cytokines, an increased number of hematopoietic precursor cells was observed only at day 1, folIowed by a marked decrease in precursor number on day 7 or 14 (242). A similar dual stimulatory and inhibitory effect of IL-12 on hematopoiesis was observed in vivo. Administration of recombinant IL-12 in mice suppresses hematopoiesis in the bone marrow, with a drastic but reversible decrease in the number of colony-forming cells, especially the erythroid cells, with a concomitant mobilization of the hematopoietic progenitor cells in the circulation and enhanced peripheral hematopoiesis in the spleen (243-245). Spleen size and cellularity were increased severalfold due largely to infiltration of macrophages and NK cells, but also to an increased number of colony-forming cells (243-245). The negative effects of IL-12 on hematopoiesis in vivo are mediated mostly by IFN-y, as the decrease in bone marrow cellularity and the infiltration in the spleen of NK cells and macrophages were not observed in IFN-yR genetically deficient mice, in which IL-12 administration only promotes hematopoiesis both in bone marrow and in the spleen (245).This ability of IL-12 to promote hematopoiesis in vivo results in a powerful protective effect in mice from lethal ionizing radiation; however, IL-12, while protecting the mice from the heinatopoietic damage of the radiation, sensitized their gastrointestinal tract to radiation, inducing a severe gastrointestinal syndrome that caused their death in 4 to 6 days (246). This gastrointestinal pathology was largely abolished by anti-IFN-y antibodies, indicating a role for secondary induced cytokines in the in vivo pathological effect of IL-12 (246).
122
GIORGIO TRINCHIERI
VII. Induction of IFN-7 and Other Cytokines by 11-12
A. IL-12 INDUCED IFN-7 PRODUCTION rtv VITRO
IL-12 induces IFN-y production from resting and activated NK and T cells, with a similar dose-response relationship and a half-maximal activity at 3.5 pM (19, 78). Within T cells, both CD4' and CD8', cells with an a@ TCR, and T cells with a y6 TCR are induced to produce IFN-y (78). However, it has been reported that IL-12 induces IFN-7 through the preferential activation of CD30' T cells (247). CD30 is a transmembrane glycoprotein of the nerve growth factor receptor family, which includes CD40, Fas antigen, and the two TNF-R, and is expressed on 15-20% of anti-CD3-activated T cells, derived from the CD45 RO' memory T-cell subset (248). Not only do CD30' T cells preferentially produce IFN-y in response to IL-12, but they proliferate and expand in response to IL-12 and their generation in anti-CD3-stimulated culture is dependent on both endogenous IL-2 and IL-12, as shown by the ability of antibodies neutralizing these two cytokines to prevent the expansion of the CD30' T-cell subset (247). The induction of IFN-y by IL-12 is characterized by a powerful synergistic effect with other IFN-y inducers, in particular IL-2 and phorbol diesters (19, 78). On T cells, IL-12 also synergizes with mitogenic lectins, with stimulation of the TCR-CD3 complex by anti-CD3 antibodies or alloantigens (78), and with stimulation of the CD28 receptor by anti-CD28 antibodies or its ligand B7 (165, 166). On NK cells, IL-12 synergizes with stimulation by ligands of the CD16 receptor for IgG-Fc (anti-CD16 antibodies or immunocomplexes) and by target cells (249). IL-12 rapidly increases the transcriptional rate of the IFN-y gene; however, when IL12 and IL-2 are added together to cells, most of the synergistic effects of the two inducers are observed at the posttranscriptional level, with an increase of more than two-fold in the half-life of the IFN-y mRNA in the treated cells (250-252). Both resting and activated NK and T cells are induced by IL-12 to produce IFN-y, although maximal IFN-y mRNA accumulation is reached in 2-4 hr in activated T or NK cells and in 18-24 hr in resting PBL (78). Within PBL, IL-12 induces IFN-y mRNA accumulation, as detected by in situ hybridization, in a proportion of both NK and T cells (78); however, NK cells might be a major contributor to the early production of IFN-y in response to IL-12 or IL-2 (253). Although the production of IFN-y is usually attributed to T and NK cells, other cell types have been reported to be possible producers of IFN-y, and in particular IL-12 has been described to induce IFN-y production in B cells (254) and in peritoneal macrophages (255).These results pose interesting questions on the presence of IL-12 receptors on these two cell types and on
INTEHLEUKIN-12
123
the physiological relevance of their ability to produce IFN-y, all questions that have not been fully addressed yet.
B. REQUIREMENT FOR COSTIMULATION FOR IL-12-INDUCED IFN-y PRODUCTION Although N K and T cells are the IFN-y producers in PBL preparations stimulated by IL-12, an accessory cell type (MHC class II-positive, nonmonocyte, non-B cell) is required for optimal IFN-y production by resting PBL (78). These accessory cells might provide costimulatory molecules for IFN-y production. In murine spleen cells, it has been shown that IL-12 synergizes with TNF-a and IL-1 in inducing IFN-y prodiiction (256-258). This synergistic effect of TNF-a was not demonstrated with human lymphocytes, but antibodies to TNF-a or IL-IP efficiently inhibited IL-12-induced IFN-y production, suggesting that these two cytokines, endogenously produced in the PBL cultures, possibly by the class IIpositive accessory cells, act as costimulatory molecules for IFN-y production together with IL-12 (108). Another costimulatory signal possibly provided by the assessory cells is the B7 molecule, the ligand for the CD28 receptor on T cells. Stimulation of T cells with B-transfected cells or with anti-CD28 antibodies strongly synergized with IL-12 for the induction of IFN-y production (165, 166), and blocking of B7/CD28 interaction with the hybrid recombinant molecule CTLA4-Ig significantly inhibited the ability of PBL to produce IFNy in response to IL-12 (165). On human lymphocytes, CD28 is expressed on CD4' T cells and on a subset of CD8' T cells, but not on resting or activated NK cells. It is of particular interest that B7/CD28 stimulation of resting T cells, when combined with IL-12, is a strong stimulus for IFNy production, even in the absence of signaling through the TCR. Unlike human NK cells, CD28 is expressed on activated murine NK cells, on which IL-12 synergizes with B7/CD28 costimulation in inducing IFN-y production (259). Anti-CD28 antibodies have been shown to enhance the expression of functional IL-12R and of the IL-12R PI chain induced by anti-CD3 stimulation of human T cells (74). However, the effect of CD28 stimulation is not limited to the enhanced expression of IL-12R, as the combination of IL-12 and anti-CD28 results in a dramatic increase in IFNy mRNA stability, suggesting that IL-12 and CD28 synergize by affecting IFN-y gene expression through different mechanisms (S. Robertson and G. Trinchieri, unpublished results). In addition to CD28, stimulation of the surface antigen CD2 by antiCD2 antibodies or by its ligand CD58 can regulate the responsiveness of activated T cells to IL-12 (260, 261). The combination of monoclonal antibodies directed against two different CD2 determinants (T11.1 and
124
GIOHGIO TRINCHIERI
T11.3) and able to deliver an activating signal to T cells (262) synergizes with IL-12 in inducing IFN-y and proliferation in T cells, whereas antibodies against T1l.l, which block the interaction of CD2 with CD58, only inhibit IFN-y induction (260). The synergistic effect of the CDgCD58 stimulation with IL-12 does not involve the regulation of IL-12R expression (260, 261). These results suggest that TNF-a, IL-lP, B7, and CD58, possibly provided to some extent by the class II-positive accessory cells (78, 261), are important costimulators for IFN-y production in response to IL-12. The ability of IL-10 to inhibit IFN-y production in T and NK cells is due primarily to its ability to suppression IL-12 production, but also, in part, to its ability to suppress expression of these costimulatory molecules on accessory cells (108, 165, 166). The presence of IL-10 during the stimulation of T cells with anti-CD3 prevents the upregulation of IL-12R, possibly by acting directly on the T cells or by preventing the production of IL12 or of costimulatory molecules required for IL-12R upregulation (74). However, on activated T and NK cells, which already express the IL-l2R, IL-10 is unable to block IFN-y production in response to IL-12 (108). TGF-6 is a potent suppressor of IFN-y production in response to IL-12. Like IL-10, TGF-/3 inhibits upregulation of IL-12R in response to antiCD3 antibodies (74),but, unlike IL-10, it also inhibits the ability of both resting and activated NK cells to respond to IL-12 (263, 264). C. IFN-y INDUCING FACTOR (IGIF, IL-18, IL-ly) A novel costimulator factor for IFN-y induction identified in the liver of mice undergoing endotoxic shock was named IFN-y inducing factor (IGIF) and was cloned both in mice and in humans (265-267). Murine and human IGIF are 157 amino acid proteins (266, 267) produced by activated macrophages and possibly other cell types such as keratinocytes (268), with structure and sequence analogy to the IL-1 cytokine family (269).The names IL-18 (267) and, based on its structure, IL-1y have been proposed for IGIF (269).Like IL-16, IGIF has an unusual leader sequence or predomain that must be cleaved by action of the protease caspase 1 or IL-lP converting enzyme for optimal secretion and biological activity (270, 271). It was originally reported that IGIF by itself was a more potent inducer of IFN-y and IL-12, and that it was effective in the absence of IL12 (266).However, it soon became clear that IGIF requires costimulation to induce IFN-y, e.g., by T-cell mitogens, IL-2, or anti-CD3 (267), and that it strongly synergizes with IL-12 in inducing IFN-y production (272). Studies with murine T-cell lines have shown an absolute requirement for coexposure or preexposure of the cells to IL-12 in inducing IGIF responsiveness and suggested that IL-12 upregulates expression of the
INTERLEUKIN-19
125
IGIF receptor (273),a still unidentified molecule different from the known receptors of the IL-1 family (274). IGIF is a costiinulatory factor for the activation of T h l but not Th2 cells (275); IGIF and IL-12 synergize on Thl clones to induce IFN-y production, but only IGIF induces IL-2 production (275). However, IGIF does not promote differentiation of naive murine T cells to Thl cells, and IL-12 is required for priming T h l cells for high IFN-y production and IGIF responsiveness (276). Although IGIF does not bind to the known IL-1R (274), it activates the IL-1R associated kinase and, like IL-1, activates NFKB; however, unlike IL-12, it does not activate STAT4 (276, 277). IGIF, but not IL-1, activates NFKB on Thl clones, whereas the reverse situation is true in Th2 clones (276).The exact role of IGIF and its synergism with IL-12 remain to be defined in the in vitro and in vivo situations in which IFN-y production is induced. IGIF may be an almost absolute requirement for IFN-y production, but, as already reported for TNF-a and IL-10, at least in the human system, low concentrations of endogenous IGIF may be present in in vitro culture or in vivo to support IFN-y when other stimuli, e.g., TCR ligands or IL-12, are present. Indeed, M. Aste and G. Trinchieri (unpublished observation) have observed that, although IGIF mRNA is rapidly induced in human monocytes by the same stimuli that induce IL-12 production, a constitutive level of expression of IGIF mRNA is observed in freshly isolated monocytes, unlike the case of IL-12 p40 mRNA. Thus, the limiting roles of IL12 and IGIF in determining either acute or chronic production of IFNy remain to be established.
D. IL-12 REQUIREMENT FOR IFN-y PRODUCTION IN VITRO AND I N VIVO
Not only is IL-12 a potent inducer of IFN-y production, but it is also most likely a required factor for efficient IFN-y production, depending on accessory cells. When human PBMC were treated in vitro with stimuli, e.g., S. aureus, that induce IL-12 production, they rapidly produced large amounts of IFN-y. This production of IFN-y was almost completely inhibited by neutralizing antibodies against IL-12 (46). Even when IFN-y inducers that are not known to stimulate IL-12 production were used (e.g., IL-2, mitogens, or anti-CD3 antibodies), IFN-y production from PBMC was inhibited up to 80%, indicating that endogenously produced IL-12 is required for optimal IFN-y production (46). In these cultures, IL-12 is most likely induced by CD40L-expressing T cells, activated by stimuli with or without TCR engagement (130). However, if purified T or N K cells lacking IL-12-producing accessory cells were stimulated to produce IFNy , e.g., by IL-2 or anti-CD3 antibodies, no inhibitory effect of anti-IL-12 antibodies could be demonstrated (46).
126
CIORGIO TRINCHIERI
Injection of mice with daily intraperitoneal injections of l p g of recombinant IL-12 induced high levels of serum IFN-y, but not until 48 hr after the first injection (278). This delayed response was probably due to the lack of appropriate costimulatory signals when only recombinant IL-12 was injected. The ability of endogenous IL-12 to induce rapid production of IFN-y in wivo has been clearly shown in several experimental models of infectious diseases. A very informative experimental model for the understanding of the role of IL-12 in inducing IFN-y in vivo is provided by endotoxin-induced shock (161,162). Several cytokines, particularly TNFa and IFN-y, have been shown to be responsible for pathologic reactions that may lead to shock and death observed in infection with gram-negative bacteria and in response to endotoxins. Mice injected with LPS produced IL-12, which induced IFN-y production, as demonstrated by the ability of neutralizing anti-IL-12 antibodies to almost completely suppress IFNy production (161, 162). Studies with animals genetically deficient for the subunit of IL-12 (279,280), for the IL-12Rfll subunit (72), or for STAT4 (100, 101) have fully confirmed that IL-12 is necessary in wiwo for optimal production of IFN-y in response to bacterial products.
E. INDUCTION OF OTHERCYTOKINES BY IL-12 Although IL-12 is particularly efficient in inducing the production of IFN-y and in synergizing with other inducers, e.g., IL-2, in this effect, IL-12 also induces other cytokines and potentiates the effect of other cytokine inducers. In particular, IL-12 induces or potentiates the induction of TNF-a, GM-CSF, IL-8, IL-3, and, in certain conditions, IL-10 and IL-4 (168, 249, 281-285). IL-12 induces production of TNF-a and GMCSF from human T or NK cells at only minimal levels and is much less efficient in this effect than IL-2 or, especially in the case of GM-CSF, IL7 (249,283). When the ability of IL-12 and IL-2 together to induce IFNy and GM-CSF is compared, at both the mRNA and protein levels, a strong synergistic effect is observed for IFN-y, but only an additive effect for GM-CSF (249). However, when IL-12 is added to T or NK cells induced by other stimuli, e.g., anti-CD3, anti-CD16, anti-CD28, or phorbol diesters, IL-12 induces a strong and significant enhancement in the production of both TNF-a and GM-CSF (165, 249). The ability of IL-12 to regulate the production of type 2 cytokines is complex. Although in most instances IL-12 has an inhibitory effect on the production of IL-4 (79,286),it has also been shown to cooperate with IL4 in the generation of IL-4-producing Th2 cells (168,287,288). The effect of IL-12 on the production of IL-10 is also complex, because on the one hand, it inhibits the generation of IL-lO-producing Th2 cells and downregulates the ability of T cells from atopic patients to produce IL-
IiYTERLEUKIN'-I5
127
I0 (286); on the other hand, it can directly stimulate T cells to produce IL-10 and can synergize in this effect with other stimuli such as anti-CD2, anti-CD3, IL-2, and B7/CD28 interaction (284, 289-291). IL-12 can also prime T cells for high IL-10 production, inducing the generation of clones that produce both IFN-y and IL-10 (168, 285). Indeed, IL-12 injection in vivo was shown to induce expression of mRNA for both IFN-y and IL10 (167), and although these data were orignally interpreted assuming that IL-10 was produced by macrophages rather than T cells, they should now be reinterpreted in light of in vitro data demonstrating the induction of IL-10 production by T cells in response to IL-12. Because IL-10 is a potent inhibitor of production of IL-12 and Thl-type cytokines, the ability of IL-12 to induce IL-10 production has been interpreted as a negative feedback mechanism (284). The likely importance of such a feedback mechanism is supported by findings that IL-10 knockout mice infected with T. gondii or T. cnrzi produce uncontrolled levels of proinflammatory cytokmes such as IL-12, TNF-a, and IFN-y, resulting in the death of the animals (292, 293). VIII. Mitogenic Activity of 11-12
IL-12 was originally described to have a comitogenic effect on human T cells (19). Although IL-12 was unable to directly induce proliferation of resting peripheral blood T cells, it induced a significant enhancement of proliferation that was more evident at later times of culture when added together with the mitogeii PHA or phorbol diesters (19, 281). In particular, when IL-12 was added together with PHA, it did not increase ['HITdR uptake at the peak of proliferation at day 3, but prevented the decline in proliferation and loss in viability obseived in the PHA-stimulated cultures at day 6 (19, 281). Similarly, IL-12 enhanced the proliferation of allostimulated T cells in mixed lyinphocyte cuItures (MLC) (19). Interestingly, low levels of IL-12 are produced by accessory cells in MLC, and anti-IL-I2 antibodies partially inhibit proliferation in the culture, indicating a role for endogenous IL-12 in the proliferative response of T cells to allostimulation (294). The inability of IL-12 by itself to induce proliferation of resting T ceIls is most likely due to the lack of functional IL-12R. Studies have clearly shown that T cells activated for a few days with anti-CD3 antibodies or mitogens proliferate in response to IL-12 (281, 29S, 296) and express high-affinity IL-12R (67, 74). It remains unclear why the proliferative effect of IL-12 requires the upregulation of high-affinity IL-12R on T cells, while the induction of cytokine production is observed in a large proportion of resting T cells (78), which do not express any detectable IL-12R. IL-12 induces
128
GIORGIO TRINCHIERI
proliferation of both CD4' and CD8' cells, but the maximal levels of proliferation are always lower than those induced by IL-2, even though IL-12 is active at lower molar doses than IL-2 (281, 295). The T-cell proliferation induced by IL-12 is not mediated by endogenous IL-2, as suggested by the inability of anti-IL-2AL-2R antibodies to block it and by the lack of inhibition of IL-12-induced proliferation by compounds such as cyclosporin A, which in most stimulation conditions block IL2 production (281, 295, 297). However, the mechanism of IL-2- and IL- 12-induced proliferation may have some common pathways because both are blocked by rapamycin (297). Like T cells, preactivated NK cells proliferate in response to IL-12 (281). The preferential proliferation of NK cells observed in cultures of human PBL stimulated with EBV(+) BCL is enhanced by the addition of low levels of exogenous IL-12 and is in part dependent on the production of endogenous IL-12 by the BCL, although the production of IL-2 by allostimulated T cells plays a major role in the proliferation of NK cells in these cultures (298). When IL-12 and IL-2 are added together to preactivated T cells, especially CD4+T cells, an additive effect is observed, especially at low concentrations of both cytokines, but usually not synergism (281,295). The ability of the two cytokines to affect each other's ability to induce proliferation is in part due to the induction of IL-12R by IL-2 (67, 74) and to the induction of IL-2RdCD25 by IL-12 (299, 300). However, on purified activated NK cells, Ty8 cells, and CD8+ T cells, low concentrations of both IL-2 and IL-12 had an additive effect on proliferation, but IL-12 inhibited in a dose-dependent manner the proliferation induced by high doses of IL-2 (281, 301, 302). The inhibitory effect of IL-12 on IL-2induced proliferation is dependent on the level of NK cell activation, possibly reflecting differences in the expression of the IL-12 and IL-2 receptor complex, and is blocked by anti-TNF-a antibodies (281).Although these results suggest a role for endogenous TNF-a in IL-12-mediated inhibition, TNF-a alone has no inhibitory activity when added to the cultures (281), and in one report it was shown to enhance IL-12-induced IL-2RdCD25 expression (300). Among other cytokines that have been shown to synergize with IL-12 in inducing proliferation are IL-7 on human T cells (303) and IL-4 on human N K cells (66) and on the IL-Pdependent mouse cell line CT4S (304). The synergistic proliferative effect of IL-4 and IL-12 is somewhat difficult to reconcile with the observations that IL-4 downregulates the expression of the high-affinity IL-12R and of the IL-12RP2 chain on human T cells (74, 75, 86). TGF-P also prevents expression of the high-affinity IL-12R, probably acting at the level of the IL-12RP2 chain (74,305),thus inhibiting IL-12 responsiveness in alloactivatedT cells (305).Costimulation
INTERLEUKIN-12
129
of either resting or preactivated human T cells with CD28 ligands (B7 or antibodies) and IL-12 results in a powerful, IL-12-independent proliferation of T cells (165, 166). Similarly, costiinulation of preactivated human T cells with appropriate CD2 ligands (anti-CD2 antibodies or CD58) results in efficient proliferation (260). Several studies have shown that T h l but not Th2 cells are responsive to IL-12 (81, 86, 104), and this phenomenon has now been clarified with the finding that Th2 clones permanently downregulate the expression of the IL-12RP2 chain (83, 86). Interestingly, long-term established T h l clones may lose their ability to produce IL-2, and when stimulated in vitro by specific antigen and APC, their proliferative response is partially or totally dependent on the IL-12 produced by the APC (166, 306). Murine T h l clones tolerized in witro with anti-CD3 antibodies as well as anergic CD4+ T cells isolated from mice tolerized to the Mls-1" antigen demonstrated defective induction of proliferation in response to IL-12 on restimulation with antigen (307). Although IL-12 did not prevent the induction of T-cell anergy in these murine models (307), IL-12 was reported to prevent anergy in melanoma-specific CD4' T-cell clones using a melanoma cell line as APC (308). IL-12 has been reported to restore, in part, the responsiveness to recall antigens in the anergic CD4' T cells of HIV+ patients (119, 309), mostly by preventing activation-induced apoptosis in the patients' CD4+ T cells (310-313). These studies raise the possibility that, like other growth factors, IL12 may facilitate T and NK cell proliferation by preventing cell death; however, IL-12-mediated activation of NK and T cells, unlike that mediated by IL-2 and IL-7, does not upregulate the expression of the apoptosis protective gene BCL2 (314, 315). IX. Activation of Cyiotoxic Lymphocytes by 11-12
A. ACTIVATION OF NK CELLS One of the activities by which IL-12 was originally identified was its ability to enhance NK cell cytotoxic activity (19), and the two original definitions of IL-12, NKSF or CLMF, refer to the ability of IL-12 to activate NK cells and lymphokine-activated killer (LAK) cells (19, 20). Treatment of PBL with IL-12 for several hours results in an enhancement of NK-cell-mediated cytotoxic activity that can be detected against NKsensitive target cells (e.g., K562), but also against NK-resistant target cells or virus-infected cells (19, 117, 316). Mice with disrupted IL-12 p40 genes have only a modest decrease in NK cell-mediated cytotoxicity, indicating that IL-12 is not essential for differentiation of NK cells (279). However, an immature human NK cell subset expressing the
130
GIOHGIO TRINCHIERI
NKR-P1A antigen, but neither CD56 nor CD16, is dependent on IL12 for differentiation in vitro to acquire the CD56' phenotype of mature NK cells (317). The enhancement of cytotoxic activity of NK cells preincubated with IL-12 is usually lower than that observed with IL2 and is comparable with that of IFN-a (19, 318); however, IL-12 enhances human NK cell cytotoxicity at concentrations between 0.1 and 10 pM, whereas IL-2 and IFN-a reached their maxiinal activity at concentrations higher than 1 nM. This sensitivity of resting NK cells to the enhancing effect of minimal concentrations of IL-12 is surprising, as no evidence of high-affinity IL-l2R has been demonstrated on these cells, although cytofluorirnetric analysis with IL-12 and IL-12 antibodies has suggested that resting NK cells, unlike most T cells, express low levels of IL-12R (66). It is also of interest that, unlike IL-12 induction of IFN-y production, the IL-12-mediated enhancement of cytotoxicity does not require tlie presence of accessory cells, which could provide costimulatory signals for upregulation of the IL-12R (316, 318). The simultaneous treatment of NK cells with IL-12 and IL-2 or IFN-a results in only an additive effect on cytotoxicity, very different from the strong synergism observed when IFN-y was induced by a combination of IL-2 and IL-12 in either NK or T cells (316). The effect of IL-12 on NK cytotoxicity is not blocked by antibodies to either IFN-dP or IL-2, suggesting that IL-12 acts directly on the cytotoxic cells and is not acting through the induction of other cytokines (318, 319). An inhibitory effect of anti-TNF-a antibody was observed when NK cells were cultured with IL-12 for 3 days (318) but not for 1 day (316), suggesting a possible requirement for this cytokine in longer term cultures only. When human PBMC are stimulated in vitro with S. aureus, a significant increase in NK cytotoxic activity is observed, which is partially inhibited by antibodies to IL-12, IFN-a, and IL-12, but is almost completely inhibited by a mixture of all three antibodies, indicating that the three cytokines may cooperate during infection in inducing NK cell activation (46). Indeed, IL-12 has been shown to be required for NK cell activation and migration in the regional lymph node in L. major infections (320). In vivo daily injections of exogenous IL-12 in mice result in an enhancement of NK cytotoxic activity in both liver and spleen: however, similar to the phenomenon described for IFN-a (321), NK activation is maximum at 2 days, declining thereafter (278). In vitro, IL-12 is chemotactic for human NK cells and stimulates their interaction with vascular endothelium via LFA-1/1CAM-1 and VLAWCAM-1 pathways (322). IL-12 activation of NK cells is accompanied by an increased expression of CD69 antigen, TNF-R p75, IL-2RaKD25, and the P2 integrin CDlla; however, IL-12, unlike IL-2, does not
INTERLEUKIN-12
131
upregulate the expression of pl integrins (89, 323, 324). IL-12 treatment of human NK cells results in an increased number of granules (316) and a change in their morphology from a prevalently electron-dense inside core to a more polymorphic morphology with the formation of vesicles, indicating cellular activation (C. Grossi and V. Pistoia, personal communication). Previous exposure of N K cells to IL-12, as well as to IL-2, enhances the release of granule-derived proteins on triggering by stimuli that activate the Ca2+ and/or a protein kinase C-dependent intracellular pathway, suggesting that these cytokines confer a level of activation to the NK cells that allows them to respond with maximal granule exocytosis and cytotoxic activity to stimulation (325). Several studies (249, 326-328) have demonstrated that IL-12 enhances transcription and accumulation of m R N A for several granule-associated molecules, particularly perforin and granzyrne B, during its activation of N K cells or cytotoxic T cells. A synergistic effect of IL-12 and IL-2 in inducing the perforin and granzyme R genes is marginal and is not observed consistently (249, 327), reflecting the effect of these two cytokines on cytotoxic activity and contrasting with the powerful synergism observed for the induction of IFN-y gene expression.
B. IL-12 ACTIVITY O N CYTOTOXIC T LYMPHOCYTES IL-12 was originally reported to synergize with IL-2 in inducing the generation of LAK cells (cytotoxic cells comprising activated N K cells and non-MHC-restricted CTL) from human-purified blood lymphocytes in the presence of hydrocortisone to block endogenous cytokine production (20). In the absence of hydrocortisone, IL-12 alone induces LAK cell generation with a mechanism that requires endogenous TNF-a, (329). The ability of IL-12 to upregulate the cytotoxic mechanism of resting and noncytotoxic peripheral blood T cells has been demonstrated by its ability to endow these cells with the ability to mediate antibodyredirected lysis of anti-CD3-coated target cells on an 18-hr culture (316). IL-12 also enhances the generation of hurnan-allospecific CTL (329, 330). Notably, IL-12 has been shown to induce both proliferation and generation of CTL activity in primary mixed cultures of lymphocytes from siblings identical for MHC class I1 antigens and displaying class I disparity, suggesting that IL-12 may play a role in helper T-cellindependent CTL generation (294). In this latter system (294), but not in the generation of CTL against fully allogeneic cells (329), the effect of IL-12 on CTL generation is IL-2 independent. As with NK cells, the ability of IL-12 to enhance CTL activity is, in part, based on induction-of the expression of genes encoding cytotoxic granule-associated proteins, e.g., perforin (294, 327, 330), suggesting that IL-12 may effect
132
GIORGIO TRINCHIERI
the generation of CTL by inducing both their expansion and their differentiation in cytolytic effector cells. In vivu intraperitoneal (ip) injection of high doses of IL-12 (1 pgl mouse/day for 4-5 days) in mice immunized in the footpad with allogeneic lymphocytes induces a strong increase in the CTL activity in the draining lymph nodes, although the number of cells recovered in the lymph node is decreased by 25-50% (278). However, endogenous IL-12 is not required for the generation of allospecific CTL, as demonstrated by studies in IL12 p40 genetically deficient mice (279). In most experimental systems, no major effect of either exogenous or endogenous IL-12 in the generation of anti-viral CTL has been demonstrated (331),with a modest decrease of CTL activity in anti-IL-12-treated mice observed only in influenza virus infection (332). However, the in vitru addition of exogenous IL-12 to influenza virus-infected human dendritic cells strongly enhanced the ability of these cells to induce CD8 Tcell proliferation and generation of CTL responses, especially with lymphocytes of donors with weak reactivity to influenza virus or at a low APC : T cell ratio (333);this CTL-enhancing effect of IL-12 is not mediated by IL-12-induced IFN-y. The ability of IL-12 to facilitate CTL generation against influenza virus nucleoprotein was shown in experiments in which mice were immunized with a class I peptide in incomplete Freund adjuvant: a CTL response was observed in only a minority of mice not treated with IL-12, whereas a single 1-pg IL-12 dose at the time of immunization induced a vigorous CTL response in all animals (334). Relatively few studies have investigated the role of IL-12 in the generation of antitumor CTL. Multiple ip injections of IL-12 were shown to induce CTL activity against target cells expressing two different mutations of p53 in mice immunized either with peptides mixed with the adjuvant QS21 (335) or with peptide-pulsed dendritic cells (336). In both systems, rejection of established tumors was observed, but the optimal IL-12 doses for generation of CTL activity in the QS21-peptide experiments were found to be very low (1 nglmouse), with higher doses found to be inefficient or suppressive (335), whereas the experiments with peptide-pulsed dendritic cells used 10 injections of 300 nglmouse every other day (336). The schedule and dosage of IL-12 administration is of primary concern in possible immunotherapeutic approaches, not only in order to avoid systemic toxicity, but also because treatment with high doses of IL-12 for 2 weeks, a typical protocol for obtaining a direct antitumor effect, has a profound suppressive effect on both antitumor and allospecific CTL activity (H. Kurzawa, W. Lee, and G. Trinchieri, unpublished results).
INTERLEUKIN-12
133
C. ACTIVATIONBY IL-12 OF MURINENK1 T CELLSA N D THEIRHUMAN EQUIVALENT NK1 T cells are a specialized population of T a p cells that coexpress receptors of the N K lineage and have the unique potential to very rapidly secrete huge amounts of cytokines, both T h l and Th2 type (337). They express a restricted TCR repertoire made by an invariant TCR a chain, Va14-Ja281, associated with oligoclonal TCR 0 chains (338). NK1 T cells recognize the products of the conserved family of CD1 genes, MHC class I-like molecules with a large hydrophobic-binding groove (339), consistent with their ability to present lipid antigens, including inycobacterial cell wall antigens (340). A comparable population of human T cells is, like the murine counterpart, either CD4-lCD8- or CD4+ and expresses NK cell markers, such as NKR-PIA, CD94, and CD69 (341). Injection of 0.5 pg IL-12 ip in mice enhanced the cytotoxic functions and NK1 antigen expression in hepatic NK1 T a b cells after 24 hr (342344). These IL-12-activated NK1 T cells are the major effector cells in the IL-12-mediated inhibition of experimental liver tumor metastasis (343). LPS treatment similarly induces NK1 T cells with potent cytotoxic activity and in vivo antimetastatic effects via production of IL-12, presumably from liver Kupffer cells (345). NK1 T cells are also major producers of IFN-y in response to IL-12, alone or in association with IL-2 and antiCD3 (346, 347). Although all NK1 T cells are cytotoxic, only the subset expressing the Ly-6C antigens can produce IFN-y in response to IL-12 alone or IL-2 plus IL-12 (346). It has been reported that IL-12 or IL-12 inducers such as L. monocytogenes and Bacille-Calmette-Gu6rin (BCG) reduce the pool of NK1 T cells in the liver and their cytotoxic activity after 3 days of treatment, suggesting that these spontaneously cytotoxic cells may contribute to immunosurveillance of the inflammatory process in the liver and are downregulated by IL-12 (348, 349). The reason for this major discrepancy with previous studies showing increased numbers and activation of NK1 T cells in response to IL-12 has been tentatively attributed to the different time of analysis (3 days rather than 1 day after treatment) and to the possibility that the increase in cytotoxic activity reported previously was due not to NK1 T cells but to the presence of large NK1-positive conventional NK cells in the liver (348).Ongoing studies using mice genetically deficient or transgenic for Val4 are, however, confirming that NK1 T cells are the primary responders to IL-12 in v i m ( M . Taniguchi, personal communication). Interestingly, in vitro culture of human PBL with IL-12 and IL-2 induces a selective expansion of cytotoxic CD4+ CD56' T cells, which are present in the liver of humans and may correspond to murine liver N K 1 T cells (350). This findicg raises the
134
GlORGlO TRINCHIERI
possibility that this subset of T cells may be a major responder cell type to IL-12 in humans as well. X. Effect of 11-12 on the Differentiation of T Helper Cells
A. IL-12 Is THE MAJORCYTOKINE RESPONSIBLE FOR THE GENERATION OFT^^ CELLS The requirement for IL-4 in the generation of IL-4-producing Th2 cells has been well established (351-353). More recently, the role of IL-12 for the efficient generation of IFN-y production by T h l cells has become evident, and it has been proposed that the balance between the levels of IL-4 and IL-12 early during an immune response may be responsible for the bias in the generation of Th2 and Th1 cells, respectively (354), although the presence of other cytokines and various other factors regulating the immune response also play a major role. Furthermore, the synergistic/ antagonistic interaction between IL-4 and IL-12 in regulating such responses is complex and not yet fully understood (168, 355). Stimulation in witru of PBL from atopic patients with allergens such as Dermatuphagoides pterunyssinus group 1(Der p. 1)results in the generation ofT-celllines and clones with the high IL-4 andlow IFN-y production typical of Th2 cells, whereas PBL stimulation with bacterial products [e.g.,purified protein derivative (PPD)] generates Thl-type T-cell lines and clones that produce IFN-y but not IL-4, When PBL were stimulated with Der p.1 in the presence of IL-12, T-cell lines and clones were generated that exhibited a reduced ability to produce IL-4 and an increased ability to produce IFNy (79).This Thl-inducing effect of IL-12 is not inhibited by anti-IFN-y, but is reduced by removal of NK cells from the PBL preparation. PPD-specific T-cell lines generated in the presence of anti-IL-12 antibodies during the initial antigenic stimulation produced significant levels of IL-4, unlike the cell lines generated in the absence of antibodies, and gave rise to PPDspecific CD4" cell clones showing a ThO/Th2 phenotype rather than a Thl phenotype (79).These results indicate that IL-12 not only facilitatesproliferation and activation of Thl cells in a memory response in vitru, but also that, as shown by the effect of anti-IL-12 antibodies, endogenously produced IL12 is important for T h l generation. The ability of IL-12 to directly initiate Thl cell development in native murine T cells was shown by Hsieh et al. (8O),who reported that naive CD4+ T cells derived from mice transgenic for an antiovalbumin TCR are induced by ovalbumin to develop into T h l cells in the presence of IL-12, whereas they develop into Th2 cells in the presence of IL-4. The effect of IL-4 was dominant over that of IL-12 (80).Neutralization of endogenous IFN-y also inhibited IL-12-induced generation, suggesting that the effect
LNTERLEUKIN-12
135
of IL-12 to induce Thl cell generation was demonstrated in inany models, in humans and in experimental animals, both in vitro and in vivo (357). The originally reported dominance of IL-4 action over IL-12 (80) was observed to be a much more complex interaction, with the two cytokines antagonizing or synergizing with each other, depending on the function analyzed (168, 355, 358).
B. MURINESTUDIESO N THE ROLE OF IL-12 I N T h l DIFFERENTIATION 1. Identity of IL-12 and T-cell Stirnulufonj Factor While the human and murine IL-12 were being characterized and cloned, the group of Gerinann and Rude identified a inurine factor (Tcell stimulatory factor or TSF) produced by accessory cells in response to antigen-activated T cells and able to induce proliferation and IFN-y production in Thl but not in TI12 clones by a mechanism independent of TCR stimulation and Caz+flux and resistant to cyclosporin A inhibition (147, 359). As the biological activities of IL-12 on Thl cell generation and function became better understood, it was clear that TSF was indeed the same factor as IL-12 (81, 287, 360). 2. Requirenzent for IFN-y in lL-12-Induced Thl Responses
I n several in vitro experimental models in which the mechanism of IL12-induced generation of the Thl cells was studied, the major mechanism of action of IL-12 was found to be that of priming T cells for production of IFN-y and of enhancing the ability of differentiated Th1 clones to produce IFN-y (81,287,358,359).The requirement for endogenous IFNy in the mechanism of Thl generation induced by IL-12 was an almost universal finding in rnurine studies (356, 358, 360, 361), although one of these studies (358) reported that differentiation of Thl in an accessory cell-dependent system, but not in an independent one, was blocked by neutralizing anti-IFN-y antibodies. One interpretation of these results is that IFN-y acts at the level of the APC, possibly enhancing their ability to maintain a high level of production of endogenous IL-12. This hypothesis is indeed supported by studies in mice expressing a dominant negative IFN-y receptor that, depending on the promoter used for the expression of the transgene, display tissuespecific unresponsiveness in the macrophages or in the T cells; experiments with these mice identified the inacrophage as the critical responsive cell in manifesting the effect of IFN-y in regulating Th1 subset development (362). However, another study (363) demonstrated that the phenotype of CD4+ T cells was also important and that naive CD4t T cells required endogenous IFN-y for IL-12-induced Th1 generation, whereas memory CD4' T cell did not. These results are compatible with the finding that IFN-
136
GIORGIO TRINCHIERI
y upregulates the expression of the IL-12RP2 chain (83).The presence of IFN-y may be required for IL-12 activity on naive or Th2-committed CD4' cells, which do not express or have downregulated the IL-12RP2 chain, but less so on memory CD4+ T cells with an activated phenotype. A third level on which IFN-y modulates the activity of IL-12 is the production of IL-2 by T h l cells. Studies by Bradley et al. (361, 364) have shown that while the presence of IL-12 during stimulation of either naive or memory CD4+ T cells induces high level IFN-y production, it inhibits the ability of these cells to produce IL-2 compared to cells stimulated in the absence of IL-12. This negative effect of IL-12 on IL-2 production is drastically diminished by neutralization of endogenous IFN-7. 3. Relative Role of IL-12 and IL-4 in Thl Responses Although early studies indicated an almost complete dominance of IL4 over IL-12 in inducing Th2 differentiation when both cytokines were present during primary stimulation (80, 358), it became clear that the relative concentration of the two cytokines is important and that (a) IL12 does not inhibit IL-4 mediated priming for IL-4 production, but IL-4 only partially inhibits the IL-12-mediated priming for IFN-y production; (b) high amounts of IL-12 in combination with relatively low levels of IL4 give rise to a T-cell population with at ThO-like cytokine profile; and (c) in the presence of relatively high amounts of IL-4, IL-12 enhances the development of Th2 cells (287, 358). One of the major mechanisms by which IL-4 prevents IL-12 signaling is by downregulating the expression of the IL-12RP2 chain (83),an effect that is counterbalanced by IFN-y (83), thus making the balance between IL-4 and IFN-y (and its inducer, IL-12) a key factor in determining the outcome of the CD4 effector Tcell responses (365).It is also most likely that this interplay between IL12IIFN-y and IL-4 is also controlled by the genetic background of the mice; the finding (84) that in neutral conditions (i.e., with no exogenous cytokines added) T cells from B10.D2 mice maintain expression of IL12RP2, whereas those from BALB/c mice rapidly lose it, may reflect different levels of endogenous IL-4, IL-12, and IFN-y in these strains (320, 366) or different responsiveness of the IL-12RP2 gene to IFN-y and IL-4. A locus on murine chromosome 11 controlling the differential maintenance of IL-12 responsiveness has been mapped (367), and the identification of the specific gene involved should shed some light on the mechanisms of regulation of Thl/Th2 responses. 4. Stability of the ThllTh2 Phenotype
Thl and Th2 clones are quite stable in their phenotype, indicating that extremely differentiated T cells lose their plasticity to respond to polarizing
INTERLEUKIN-11
137
stimuli, which are instead effective on naive or even memory T cells (361). Some of the original studies on the stability of polarized T h l or TI12 T cell populations suggested that the IL-12-induced Thl populations, after 1 week of culture, can still be induced by secondary stimulation in the presence of IL-4 to differentiate into IL-4-producing Th2 cells, whereas the Th2 population induced in the presence of IL-4 and anti-IL-12 could not be induced by IL-12 to differentiate into IFN-y-producing T h l cells (82, 368). This loss of plasticity of Th2 cells was attributed to a defined phenotypic change, i.e., the loss of the expression of the IL-12RP2 subunit (82, 83). However, the interpretation of these studies was made difficult by two factors. First, the use of polyclonal cell populations, although more reflective of the physiological in vivo conditions than the use of clones, &d not allow the investigators to distinguish between the expansion of nonterminally committed cells and the change in phenotype of individual cells. Second, the participation of endogenous cytokines, in particular IL4, IFN-y, and IL-12, could not be accurately evaluated. Indeed, Murphy et nl. (369), using single-cell analysis of cytokine expression by intracellular staining (370), showed that both T h l and TI12 polarized populations after 1 week culture can be skewed to the opposite phenotype by switching the culture conditions, most likely by expanding undifferentiated or incompletely differentiated cells, but that, after long-term stimulation for 3 weeks, this reversibility is lost. Nakamura et al. (371) addressed the same question using different approaches, i.e., analysis of IL-4 and IFN-y transcripts during stimulation and, particularly, the elimination of IL-Cproducing cells at different days of culture, utilizing the antiviral drug ganciclovir and T cells from mice transgenic for the thymidine kinase gene under the control of the IL-4 promoter (372). Those studies suggested that by day 2 of culture, a large proportion of T cells stimulated in the presence of IL-4 or IL-12 is irreversibly committed to the Th2 or T h l phenotype, respectively. The inability of Th2 cells to respond to IL-12 is due to downregulation of the IL-12RP2, but this subunit can be reexpressed if the cells are treated with IFN-y (83), an effect that is antagonized by IL4. Indeed, Nakamura et al. (365) demonstrated that the Th2 populations normally show a stable phenotype and fail to respond to IL-12 because of endogenous IL-4 production. The use of anti-IL-4 antibodies does not completely avoid this effect because IL-4 may be utilized intracellularly by the IL-4-producing cells (371). IFN-y abrogates the antagonistic effect of IL-4 and permits the conversion of Th2 populations into IFN-y producing cells. In the complete absence of IL-4, using cells from IL-4 genetically deficient mice, IFN-y is not required for this conversion and IL-12 can convert The populations in Thl cells that lose responsiveness to IL-4 due to the lack of the IL-4-mediated upregulation of IL-4R (365). The reported
138
CIORCIO TRINCHIERI
ability (82, 368) of IL-4 to revert the polarized Thl populations to a Th2 phenotype most likely reflected the presence of endogenous IL-4 during the primary stimulation, which prevented complete differentiation of the Thl cells induced by IL-12. The finding that T h l polarized populations from IL-4 knockout mice cannot be reverted by IL-4 to produce Th2 cytokines (e.g., IL-5) supports this explanation (365). Thus, in the absence of IL-4, IL-12 rapidly and irreversibly commits CD4 T cells to a Thl phenotype.
5. Is IL-12 Essential for Thl T-cell Generation? Many studies in oitro and in vizjo have clearly shown that IL-12 is a potent inducer of T h l responses. However, in IL-12 p40 genetically deficient mice, Thl responses are severely depressed, especially regarding IFN-y production, which is reduced to -10% of that of wild-type mice, but not completely absent, e.g., IL-2 production is almost normal (279, 280). In addition, the regulation of Thl response to bacterial or nominal antigen in the presence of adjuvant may be different from that observed in other experimental models: for example, the T h l response and priming for IFN-y in cardiac allograft recipients appear to be little affected in IL-12 p40 or p35 deficient mice (373), and IFN-.)Iproduction in several virus infections may be relatively independent of IL-12 (331). Although IL-12 was required in vitro for priming for optimal IFN-y production, primary antigen stimulation of TCR-transgenic CD4+ T cells, in the absence of IL-12, generated T cells with a Thl-type phenotype and able to produce significant amounts of IFN-y when challenged in a secondary stimulation in the presence of IL-12 and IGIF (276). However, IGIF, which strongly synergizes with IL-12 in inducing IFN-y production from polarized Thl cells, was unable to replace IL-12 in inducing T h l differentiation and did not represent a possible substitute for IL- 12 in supporting the modest T h l responses in IL-12 knockout mice (276). High doses of IL-2 and IL-15 are also able to induce some priming of T cells for IFN-y production, and at least the effect of IL-2 appears to be independent of endogenous I L- 12, indicating possible alternative pathways of induction of Th1 responses (374). Although Thl cells by definition are high producers of IL-2 and IFNy, most studies on the effect of IL-12 focus on IFN-y production. In one in which IL-2 production was analyzed, the addition of IL-12 to the primary stimulation of CD4 T cells did not significantly affect the ability of the cells to produce IL-2, while increasing severalfold their ability to produce IFN-y (368). In another study, IL-12 significantly inhibited the ability of the cells to produce IL-2, an effect that was mediated by IFN-y (364). Thus, IL-12 appears to be essential for the development of high IFN-y-
INTEKLEUIIN-12
139
producing Thl cells, but in its absence, moderate production of IFN-y and norinal production of IL-2 are still observed. 6. Ir IL-12 Necessnnj f o r Maintaining T h l Responses?
Thl cells maintain the expression of IL-12R and IL-12 responsiveness (82, 83), and IL-12, in synergy or not with costiinulatory molecules such as B7 on APC, can greatly augment cytokine production and proliferation of polarized Thl populations or clones (166, 299, 375). IGIF also strongly synergizes with IL-12 in inducing IFN-y production and proliferation of Thl cells (272,273,276).The IL-12 mediated enhancement of proliferation requires IL-2 production froin the antigen-stimulated T h l cells in most experimental conditions, and one possible mechanism of action of IL-12 is through its ability to enhance the expression of the IL-12Ra subunit, and thus to increase IL-2 responsiveness (299).However, certain terminally dfferentiated T-cell clones have lost their ability to produce IL-2 and are completely dependent on IL-12 produced by APC for proliferation when stimulated with antigen (306). Some of the studies reviewed earlier showed that polarized Th1 cells completely lose the ability to produce IL-4, whereas culture in the presence of IL-12, with or without IFN-y, can, under certain conditions, induce Th2 cells to produce IFN-7, generating cells that produce both IL-4 and IFN-7 (83, 369, 375, 376). Thus, in differentiated CD4+ T cells, IL-4 production is cytokine autonomous (although IL-4 is absolutely necessary for priming for its own production), whereas IFN-y production is cytokine dependent. This finding has led to the suggestion by IIu-Li et nl. (375) that the ability or lack thereof to produce IL-4 should be considered the defining property of Th2 and Th1 cells, respectively, rather than the ability to produce IFN-y, which either cell type can express depending on the recent exposure to a cytokine environment. Although in zjitro studies have shown that IL-12 plays a role in niaintaining and activating T h l cell function, results since the original report in T. godii-infected mice (377) in many in zjizjo infectious disease models have shown that once a powerful Thl response is established, endogenous IL12 can be neutralized without decreasing the protective response or IFNy production in animals. However, with less vigorous Th1 responses, e.g., in autoimmunity, IL-12 may be necessary for maintaining the response (378).
7. N K Cells Participate in the IL-12 Iiiduction of Thl Respniise NK cells have powerful cytotoxic activity and are efficient producers of' cytokines, particularly IFN-y (253, 379). In response to stimuli such as IL-2 (253) and IL-12 (78), not only are NK cells efficient producers of IFN-y, but they can respond at a much earlier time than antigen-specific
140
GIORGIO TRINCHIERI
T cells. An important in vivo role for NK cells as producers of IFN-y in a T-cell-independent mechanism of macrophage activation has been clearly demonstrated in L. monocytogenes (380) and T. gondii (256) infections and has been shown to depend on induction of IL-12 (257, 381). Migration of NK cells in draining lymph nodes following L. rnujor infection and their production of IFN-.)I have been demonstrated to be important for the T h l response in L. major-resistant mouse strains (382). In vivo depletion of NK cells blocks the Thl response in the resistant mouse strains (382) and also prevents the induction of a T h l response in susceptible BALBlc mice when IL-12 is used as an adjuvant in vaccination with soluble leishmania1 antigens (383). Neutralization of IL-12 in vivo during L. major infection of resistant mouse strains prevents the early migration of NK cells in the draining lymph nodes and their production of IFNy (320). NK cells also appear to play a role in human Thl response in vitro (384), and depletion of NK cells from the cultures significantly decreases the ability of IL-12 to induce a T h l response (79). Overall, data implicating proinflammatory early cytokines such as IL12, TNF-a, and IL-lP, and NK cell activation in setting the stage early in an immune response for the ensuing antigen-specific Thl-type immune response indicate a clear influence of innate resistance in directing the response of adaptive immunity. The role of NK cells might be mediated through their production of IFN-y, insofar as NK cells are the earliest IFN-y producers in an immune response. However, the requirement for NK cells in the IL-12-induced T h l response of human cells in vitro, an experimental system in which no effect of neutralization of IFN-y was observed, suggests that NK cells may favor the T h l response by mechanisms in addition to IFN-y secretion, possibly the production of other cytokines. 8. Polarization of CD8' T Cells and TyS T Cells CD8+ T cells (13, 14,385,386)and Ty6 T cells (17) can also be primed to produce Th2 cytokines and, thus, like CD4' T cells, can have a polarized Thl-type or Th2-type phenotype often referred to as TC1 and TC2. The conditions that induce polarization of CD8' T cells are similar to those inducing CD4' T cells, i.e., IL-12 is the major factor required for TC1, although differentiation of these cells is also observed in the presence of IFN-y plus anti-IL-4 or of TGF-P, whereas IL-4 is the major factor required for TC2 differentiation. TC2 cells produce nearly the same levels of IL-5 as Th2 cells, but much reduced levels of IL-4, and have much lower but not absent production of IFN-y compared with TC1 cells (387). The presence of IL-12 in addition to IL-2 during priming of TC1 cells
INTEHLEUKIN-12
141
results in a five-fold increase in IFN-y production, but, similar to the observation with Thl cells, a three-fold decrease in IL-2 production (385). Both TC1 and TC2 subsets are cytotoxic (386), and TC2 clones expressing CD40L may provide some B-cell help, although not efficiently because of their ability to lyse APC and B cells (386, 388). Interestingly, the IL-12 p40 homodimer, which blocks the ability of IL-12 to induce T h l differentiation in CD4' T cells, was reported to significantly enhance the priming for IFN-y production in alloreactive CD8' T cells (389). Although no conipelling evidence for a direct stimulation of the p40 homodimer on CD8' cells is provided in this study, the challenging possibility exists that the homodiiner on CD8' cells signals by binding to the IL-12Rp1 chain. Indeed, the use of chimeric receptors has shown that the cytoplasmic region of IL-12Rp1 alone can signal, transduce, and phosphorylate STAT3, although no biological activity was demonstrated in those experiments (93). Although IL-12 p40 and IL-12 p35 genetically deficient mice have a similar phenotype (280), only p35 knockout mice are resistant to Cyptococcus neoformans, and the susceptibility to infection in the p40-deficient mice can be overcome by treatment with p40 homodimer (J. Magram, personal communication). Thus, the p40 homodimer, perhaps by acting selectively on CD8' T cells, may mediate some of the functions of the p70 heterodinier. C. ROLEOF IL-12 IN DIFFERENTIATION OF HUMAN T h l CELLS Analysis of the role of IL-12-induced differentiation of human T h l cells has been obviously lianipered by the limitation of using human material, even though the original description of this activity of IL-12 was reported at the same time in human (79) and mouse (80) studies. One of the limitations of the human studies is that the recall response of memory T cells, rather than the primary response of naive T cells, is used as an indicator of antigen-specific responses. Yet, similar to observations in the murine system (361),IL-12 exerts aprofound altering effect on the cytokme production pattern of Th2-biased allergen-specificT cells by inhibiting IL4 and IL-10 production and boosting IFN-y production, but at polyclonal and clonal levels (79, 286). Furthermore, it was demonstrated that the ability of Thl-inducing recall antigens such as PPD to induce the generation of Thl-type clones is dependent on the ability of this type of antigen to induce endogenous IL-12 production from APC (79). Unlike in the murine system, IFN-y has not been generally found to play a necessary role in the IL-12-induced differentiation of human Thl cells (15, 79, 168), with the exception of one study in which anti-IFN-y antibodies were shown to partially prevent the IL-12 priming for IFN-y production in cord blood T cells (390). This species difference may be
142
GIORGIO TRINCHIEHI
explained by the fact that IFN-y induces IL-12RP2 subunit upregulation on mouse T cells (83),whereas on human T cells, IFN-a, but not IFN7 , has this function (86).Thus, IFN-(r is able to induce generation of T h l cells from allergen-specific T cells (391, 392) and is responsible, together with IL-12, for the Thl differentiation effect induced by poly-1:poly-C ( 145). As in the mouse, the polarization of human T cells is regulated by IL4 produced by T cells and by IL-12 produced by dendritic cells and other APC (393). The expression of B7 costimulatory molecules on APC is required for IL-$-induced differentiation of human Th2 cells, but not for the IL-&mediated differentiation of T h l cells (394), even though IL-12 and B7/CD28 stimulation strongly synergize for transient expression of IFN-y (165). An altered TCR ligand, i.e., an analog peptide derived from an allergen and able to boost IFN-y production in a Tho clone, induces APC production of IL-12 during antigen presentation (395). Thus, the interaction between T cells and APC mediated through the TCR and an altered T-cell ligand is bidirectional, enabling APC to deliver signals to the T cells, particularly IL-12 (395). Many studies in the human system have used cord blood or thymus cells to analyze the response of naive T cells. Many of the mechanisms studied with cord blood cells can be extrapolated to adult T cells, but some important differences are present. In particular, both cord blood and thymus CD4+ T cells in default conditions differentiate to Th2-type cells (396, 397). Furthermore, IL-12, which on naive T cells from adult donors induces differentiation of Thl cells and inhibits IL-4 production (79), acts on neonatal and cord blood CD4+ T cells to promote the differentiation of cells that produce high levels of both IFN-y and IL-4 (288, 398). However, on cord blood CD8' T cells, IL-12 completely inhibits the IL4-induced capacity of CD8+ T cells to produce IL-4 (399). Analysis at the single cell level (400) and using clonal limiting dilution (15, 168) has shown that the induction of Thl differentiation by IL-12 is rapid and requires only about 4 days of priming to become stable and irreversible, whereas IL-4-induced Th2 cell differentiation requires longer time and repeated stimulation, and Th2 cells can be reverted to IFN-yproducing Thl cells by a single restimulation in the presence of IL-12. Although human Th2 clones do not express detectable IL-12RP2 chain and have not been shown to signal in response to IL-12 (86, 401), the majority of them still respond to IL-12 with a low and transient IFN-y production (15, 402). The expression of ILlO in mouse T cells is restricted to Th2 cells, with few exceptions. However, in human T cells, IL-12 induces a priming for the production of both IFN-y and IL-10, thus resulting in the generation
In'TEHLEUKIN-12
143
of T-cell clones that produce both activators and inhibitors of macrophage activation (168, 285, 397).
DIFFERENTIAT~ON D. IL-12 AND Thl CELLGENERATION: OR SELECTION? The experimental systems used in studynig the Th response have not permitted determination of whether the different cytolanes that affect Th cell development induce differentiation of bipotential Th precursors or rather a selective priming and/or expansion of already committed Th1 and Th2 precursor cells (390, 403-405). This question is particularly relevant in human studies that have analyzed clonal expansion of memory Th cells (79, 406). However, once a Thl or Th2 response has been established, it appears to be relatively stable, and no factors capable of inducing qualitative changes in the cytokine profile of established murine or human T-cell clones have been reported. In the analysis of cytokine production froin human T cells stimulated with recall antigens (PPD) or allergens (Der p.l), the expansion of the sinall proportion of memory T cells was first obtained in polyclonal T-cell cultures, froin which single antigen-specific clones were obtained only after several weeks of culture of the polyclonal cell line (79, 406). During this culture period, emergence of Th cell subsets with characteristic cytokine production profiles could reflect differentiatioil of precursor Th cells, as well as positive selection (growth advantage) of certain Th subsets or negative selection (apoptosis, cytotoxicity, antiproliferative effects) of other subsets. Human T-cell cultures exposed to a polyclonal stiinulus that affects all T cells, such as PHA or anti-CD3, in the presence or absence of IL-12, displayed the enhanced IFN-y production and ahnost complete abrogation of IL-4 production observed in antigen-stimulated cultures (407). However, very different results were obtained when freshly isolated human peripheral blood T cells were immediately cloned by limiting dilution in cultures stimulated by PHA and IL-2, in the presence or absence of IL12 (15). When restimulated with anti-CD3 and phorbol diesters after 5 weeks of culture, the clones generated in the presence of IL-12 produced 5- to 20-fold higher levels of IFN--y than the clones generated in the absence of IL-12. This priming for IFN-y production required the addition of IL-12 within the first week, but its presence for inaxirnal priming was required only for 1 or 2 weeks (168). Once the clones were established for 2 or 3 weeks, removal or addition of IL-12 from the culture inediuin &d not significantly affect their ability to produce IFN-y (15).Because the clonal efficiency in these experiments was close to 100%,the priming effect of IL-12 was not due to selection of high IFN-y-producing clones,
144
GIORGIO TRINCHIERI
but was exerted on each single T cell, naive or memory. Furthermore, this effect was observed on both CD4+ and CD8+ cells, suggesting that IL-12 affects the differentiation of Thl-type clones from both subsets. Thus, the presence of IL-12 during the initial clonal proliferation of T cells induces an irreversible priming for high IFN-y production, which is maintained even when the clones are cultured for several weeks in the absence of IL12. However, unlike what is consistently observed in vivo and in polyclonal cultures and their derivative clones, the clones originated by limiting dilution in the presence or absence of IL-12 showed no significant difference in their average ability to produce IL-4 (15, 407). These results suggest that the ability of IL-12 to prime CD4' cells for high IFN-y production is due to a differentiation effect acting at the level of CD4+ T-cell clone precursors. The ability of IL-12 to downregulate IL-4 production, however, was not observed at the clonal level and is likely due to selective processes operative on polyclonal cultures and not to a direct effect on single clonal progenitors (15). The nature of these mechanisms remains to be investigated, although a possible selective proliferative effect of IL-12 on T h l clones or an IFN-y-mediated negative selection against IL-4-producing clones can be postulated. Alternatively, downregulation of IL-4 production might be a differentiation effect that requires cellular interaction or cell crowding (e.g.,for the production of Th2-suppressing factors such as IFNy ) during the initial phase of proliferation of the T cells; such interactions are not obtained in limiting dilution cultures, even in the presence of irradiated feeder cells. The ability of IL-12 to prime human T cells for IFN-y production has also been suggested by experiments showing that naive human cord blood T cells are unable to produce IFN-y, but acquire this ability after a few days of culture in the presence of IL-12 (390). However, selective effects and/or preferential proliferation of T-cell subsets could not be excluded in the polyclonal cultures of cord blood T cells. That the mechanisms underlying the enhancing effect of IL-12 on IFN-y and the inhibition of IL-4 production might be different is also suggested by data in the murine system, showing that the enhancement of IFN-y production is a direct effect of IL-12 on T cells, whereas the inhibition of IL-4 production is due to an indirect effect on APC or on other cell types present in the APC preparations (408). CD4' and CD8+ clones obtained by limiting dilution in the presence of IL-12 produced significantly more IL-10 than clones generated in the absence of IL-12 ( 168). However, in allergen-stimulated polyclonal T-cell cultures IL-12 was shown to downregulate both IL-10 and IL-4 (286). These apparently contradictory results are, however, consistent with the conclusion that IL-12 directly upregulates T h l cytokine production, but
INTERLEUKIN-12
145
suppresses Th2 cytokine production by an indirect, possibly selective mechanism. The high variability in the production of cytokines observed in human CD4’ T-cell clones expanded in the absence of exogenous IL-12 and IL4 or in the presence of neutralizing antibodies against these two cytokines suggests that some of the cells are already primed in vivo for cytokine production. The cloning of sorted CD45RO- “naive” CD4’ cells and CD45RO’ “memory” CD4’ cells supports this interpretation (168).A high proportion of the clones generated froin CD45RO’ CD4+ cells in the presence of neutralizing antibodes to IL-12 and IL-4 produced one or a combination of IFN-7, IL-4, and IL-10, with a pattern of production that was not always consistent with the classical paradigm of T h l and Th2 cells. When CD45RO- cells were cloned in the same conditions, the clones produced only negligible amounts of the three cytokines. However, in both populations, the presence of IL-12 during cloning endowed virtually all clones with the ability to produce high levels of IFN-7 and IL-10. IL-4, either endogenously produced or exogenously added, was necessary in the limiting dilution cultures to prime T-cell clones generated from CD45RO- cells for IL-4 production, whereas approximately half of the clones generated from CD45RO’ cells produced IL-4, even when expanded in the absence of IL-4. Thus, the requirement for IL-4 in the generation of IL-4-producing cells was difficult to evaluate when total PBL were cloned. Although IL-12 is a major inducer of a T h l response, it was also shown to potentiate IL-4 production and the development of Th2 cells from naive CD4+ murine T cells (287) and froin neonatal CD4’ human T cells (288), and to potentiate a Th2 response to Schistosoim mnnsoni in IFN-7 knockout mice (409). IL-12 does not prevent IL-4 production from CD4+ clones derived from limiting dilutions of “naive” adult peripheral blood CD45RO- cells and, in fact, significantly enhances the ability of IL-4 to prime the clones for high IL-4 production (168),thus extendmg previous results (287, 288, 409) by demonstrating that IL-12 can enhance IL-4 production at the single clonal level via a differentiation effect. Furthermore, when T cells were cloned in the simultaneous presence of IL-12 and IL-4, the I F N - 7 priming effect of IL-12 was only partially and often not significantly inhibited by IL-4, whereas the priming for IL-10 production was reproducibly and almost completely blocked by IL-4 (168). Thus, paradoxically, IL-4 is more potent in inhibiting priming of Th cells for production of IL-10, a Th2-type cytokine, than for the typical Thl-type cytokine, IFN-y. Figure 5 illustrates the cytokme production pattern of clones derived from human CD4’ CD45RO- T cells in the presence of IL-12 and/or IL-4.
146
GIOHGIO THINCHIEHI
'oool I* Q
I'
0
1
IL-10, pglml
IL-4, pglml
Fic. 5 . Cytokine production patterns froiu clones derived by lintiting dilution froin human CD4' CD45RO- T cells in the presence of feeder cells, PHA, and IL-2. Clones were grown for approximately 5 weeks in the presence of' the cytokines or neutralizing anticytokine antibodies shown in the legend and were then restiinulated with anti-CD3 antibody and phorbol &ester for measurement of cytokine production. Symbols represent the average cytokine production of a large number of clones (tSE): 0, IL-4 (anti-IL-12); 0, IL-12 + IL-4; A, IL-12; A,IL-12 (anti-IL-4); 0 , anti-IL-12; W, anti-IL-12 + anti-IL4. Modified froin Gerosa et ul. (1997).J. Exp. Med. 183, 2259-2269.
E. ACUTEINDUCTION VERSUS PRIMING FOR CYTOKINE PRODUCTION Froin the studies of the generation of Thl and TI12 cells, it is becoming clear that production of lymphokines, both type 1 and type 2, can be regulated through two different mechanisms. The first mechanism, observed particularly in preactivated lymphocytes, but also in resting T cells and NK cells, is the ability of various stimuli, including TCR and cytokine stimulation, to rapidly induce gene expression and cytokine production. For example, IL-12, alone or in synergywith other stimuli, induces accumulation of inRNA for IFN-y within a few hours of treatment of either resting or activated T or NK cells, followed by secretion of IFN-y. This acute induction of IFN-7 subsides within -2 days [or, in viva, even within less than 12 hr (161)]and does not induce a permanent alteration in the ability of the cells to produce IFN-y in response to IL-12 or other stimuli. The second mechanism, priming of cytokine genes, is quite different from acute induction. When T cells (and NK cells) are clonally expanded in the presence of IL-12 during the first few days of expansion, the clones are primed for high production of IFN-y and IL-10, even when cultured for several weeks in the absence of IL-12 and stimulated in the absence of IL-12; conversely, exposure of T cells to IL-4 during clonal expansion induces priming for IL-4 production and generation of IL-4-producing cells. IL-12 is particularly potent in mediating both acute induction of the
INTERLEUKIN-12
147
IFN-y gene and its stable priming for response to other stimuli. IL-12 siinilarly primes the IL-10 gene, but its ability to acutely induce this gene is modest and difficult to demonstrate. Analogously, the ability of IL-4 to acutely indiice the expression of the IL-4 gene has not been demonstrated, although IL-4 is necessary and extremely potent for the priming of the IL-4 gene and the generation of IL-4 producing cells. The effect of IL-2 011 lymphokine production is different from that of IL-12 or IL-4: IL-2, alone or in synergy with other stimuli, is a potent inducer of acute expression of the several lymphokmes, including IFN-7, IL-4, and IL-10. However, altliough the presence of IL-2 may be required in the priming phenomena of all three genes, IL-2 by itself does not determine the specificity of the priming, which is instead directed by IL-12 and IL-4 (78, 253, 410). As mentioned earlier, IL-4 and IL-12 both antagonize and synergize in inducing priming of lyinphokine genes. IL-4 almost coinpletely abolishes the IL-12 priming for IL-10 production, but only partially decreases the priming for IFN-7; IL-12 potentiates rather than inhibits the IL-4 priming of T cells for high IL-4 production. The priming of lymphokine genes represents a stable modification of the inducibility of the genes, which is analogous to the stable phenotype in the pattern of cytokine production typical of Th subsets. Thus, it is likely that this priming mechanism plays a role in the determination of the Th phenotype of actitated T cells. However, certain effects of IL-12 and IL-4 on Th generation are not observed when the ability of these cytokines to induce differentiation is analyzed at the single clonal level (e.g., in clonal analyses, the powerful ability of IL-12 to block IL-4 production is not reproduced and, paradoxically, IL- 12 induces T-cell priming for production of IL-10, a prevalently type 2 cytokine). Thus, although priming of lymphoknie genes is most likely the predominant mechanism by which IL-1.2 and IL-4 induce differentiation of Th cells, the final generation of cells with Th1 and Th2 phenotype, both in vivo and in oitt-0,also depends on coinplex indrect effects of the cytohnes, including selective mechanisms. The molecular mechaniwis of both the acute induction and the priming effects remain mostly undetermined. The major signal transduction mechanisnis for IL-4 and IL-12 have been elucidated, with the former cytokine inducing activation of STAT-6 (411) and the latter of STAT 1, 3, and 4 (94, 95); however, the role of these transcription factors in the induction of expression of the IL-4, IFN-7, and IL-10 genes remains to be elucidated. The priming effects may depend on the induction of a constitutive or facilitated expression of the transcription factors responsible for the expression of the lymphokine genes or on a stalile alteration of the genes in a transcriptionally prone conformation. For example, the IFN-y gene has
148
GIORGIO TRINCHIERI
been shown to be differentially methylated in T h l and Th2 clones (412). Posttranscriptional mechanisms could also underlie the priming effect. XI. Effects of 11-12 on B-Cell Responses and Vaccination
A. POSSIBLE DIRECTEFFECTOF IL-12 ON B CELLS As discussed earlier, human BCL constitutively expresses the IL-12RP1 subunit, but only in a few cases have they been shown to bind IL-12, possibly because of lack of the IL-12RP2 chain (74,76,77).Resting murine spleen and human peripheral blood spleen cells fail to bind IL-12, but they do so after stimulation with LPS and S. aureus, respectively (77). However, peritoneal B cells, both CD5+ B1 cells and conventional B cells, bind IL-12 in the absence of stimulation (77). Expression of IL-12RP1 mRNA can be readily detected in both spleen and peritoneal B cells (77), but little information is available on the expression of IL-12R02. Early studies showed that IL-12 in vitro suppressed the synthesis of IgE by human lymphocytes stimulated by IL-4 or IL-4 plus anti-CD40 antibodies through an IFN-dependent or -independent mechanism (413). The inhibitory effect of IL-12 on IgE synthesis was most likely indirect and mediated by T or NK cells (413).However, another study (414) showed that IL-12 enhanced the growth of S. auras-stimulated human B cells and, in the presence of IL-2, potently enhanced B-cell differentiation with an increased production of both IgG and IgM (414). These results suggest a direct effect on B cells. The ability of IL-12 to enhance the growth of S. auras-stimulated human B cells was inhibited by anti-IFN-y antibodies; PCR analysis at the single cell level indicated that IL-12 induces IFN-y production in B cells and that the endogenously produced IFN-.)Iis responsible for the effect of IL-12 on B-cell growth (254). In addition, it was shown that IL-12 induces IL-10 production in CD5' B splenic B cells and IL-6 production in both CD5+ and CD5- B cells (415). IL-12 in vivo treatment of mice immunized with phosphorylcholine conjugated to keyhole limpet hemocyanin (KLH) or with S. mansoni soluble antigen resulted in a loss of peritoneal CD5' B1 cells (416,417). However, in vitro studies showed that IL-12 is required for the growth of S . aureus-stimulated peripheral blood CD5+ Bla B cells, but not of CD5- B cells (418). Overall, information on the mechanisms of action of IL-12 on B cells is still scarce, and several unanswered questions remain about the expression of functional IL-12 receptors on B cells, although increasing evidence indeed points to a direct effect of IL-12 on B-cell functions. The reported effects of IL-12 on B1 cells are also difficult to interpret because in vitro they suggest a direct effect of IL-12 on B1 cells, with increased proliferation and secretion of IL-10, which is an autocrine growth factor for these cells,
INTERLEUKIN-12
149
whereas in vivo evidence suggests a depletion of peritoneal B1 cells in IL-12-treated animals. Whether this latter effect is due to actual death of the B 1 cells or to migration to different anatomical sites remains to be clarified.
B. EFFECTOF IL-12 ON ISOTYPE SELECTION AND USE AS A VACCINE ADJUVANT The ability of IL-12 to direct an immune response to the generation of Thl cells, involving a change in isotype selection, has generated interest in using it as an adjuvant in vaccination to iinprove the immune response to those pathogens for which a Th1 response is protective. The first example of such use was in the L. major infection model in the mouse (383). BALB/c mice are susceptible to L. major because they respond to the infection with a Th2 rather than a TI11 response, and different protocols of vaccination have been mostly ineffective because of their inability to shift the response from Th2 to Thl; however, vaccination of BALBlc animals with soluble Leishmania antigen plus IL-12 induced a strong and protective Thl memory response (383).Those studies raised much interest in the possible use of IL-12 as an adjuvant, but despite encouraging results, they have often been surprising and not easily interpretable within the paradigm of exclusive T h l induction by IL-12. Originally, IL-12 was shown to suppress IgG1, IgE, IgA, and, less efficiently, IgG2a and IgG3 in mice treated with anti-IgD antibodies (419). IL-12 also induced IFN-y in these animals and the role of this cytokine was complex: the suppression of IgE by IL-12 in anti-IgD-treated animals was IFN-y independent, whereas in animals treated with neutralizing antiIFN-y antibodies, IgGl was not suppressed (419). Furthermore, when IFN-y was neutralized in anti-IgD-treated animals, the production of IgG2a or IgG3 was strikingly enhanced by IL-12 (419). IL-12 did not inhibit IgE production in response to anti-IgE treatment, suggesting that IL-12 inhibits switching of B cells to cells that express IgE rather than inhibiting the differentiation of switched cells to IgE-secreting cells (419). Several later reports showed that in response to immunization with antigens such as KLH, TNP-KLH, hen egg white lysozyme (HEL), phospholipase A2 (PLA,), and alloantigens, treatment with IL-12 induced the inhibition of IgGl and IgE antibody production and upregulation of cytotoxic and complement-fixing antibody production of the IgG2a, IgGzb, and IgG3 subclasses (420-424). The timing and modality of treatment with IL-12 and antigens are obviously important, and the various studies have used different protocols, from coinjection of the IL-12 and the antigen to continuous daily treatment with high doses of IL-12. However, comparative studies to establish optimal
150
GIORCIO TRINCHIERI
protocols are missing. Nonetheless, it was of interest that an ovalbumin-IL12 fusion protein was found to be more efficient than ovalbumin plus free IL-12 in inducing a T h l response, inhibiting IgE production, and enhancing IgG2a and, less efficiently, IgGl (425).Also, absorption of both antigen (HIV gp120) and IL-12 on alum resulted in a much more efficient response in terms of Thl cytokine production and increase not only of IgG2 and IgG3, but also production (426). However, the in vitro proliferative response to gpl20 of the spleen cells of animals immunized with antigen and IL- 12 absorbed on alum was almost completely suppressed compared to animals immunized with gp12O and alum or a 1 2 0 and IL-12 not absorbed to alum (426). This immunosuppressive effect of IL-12 was inhibited by anti-IFN-y antibodies and correlated with the level of nitric oxide produced in the cultures (426). Many of the bacterial preparations used as adjuvant for vaccination, e.g., BCG or Corynebacteriurnp a m m , are good inducers of IL-12 production, but the role of adjuvant-induced IL-12 in the immune response has been analyzed in only a few cases. Iscoms formed by physical integration of Quillaria saponaria adjuvant and antigens induce IL-12 production in the serum of treated mice, and the neutralization of IL-12 in animals immunized with iscoms results in a decrease of total antigen-specific antibody response as well as IgG1, IgGBa, and IgG2b, suggesting a major role for endogenous IL-12 in the adjuvant capability of the iscoms (427). In animals immunized with ovalbumin, TNF and IL-1 were shown to mimic the adjuvant effect of LPS on accuinulation and follicular tnigration of antigenactivated T cells, whereas IL-12 mimicked the generation of Thl cells and help for IgG2a production (428). Overall, it is clear that conventional adjuvants induce several cytokines with various effects; however, IL-12 has a central role in favoring a T h l response and the production of antibodies of all subclasses, particularly the complement-fixing and cytotoxic ones. When the adjuvant effect of IL-12 was analyzed beyond the first couple of weeks after vaccination and primary response, the results became more complex. For example, in animals immunized with HEL, IgG2a was increased and IgGl decreased at 7 days, whereas both subclasses were enhanced at day 28 after immunization (421). Overall, it was observed that IL-12 as an adjuvant promoted a Thl-type response, but did not suppress a Th2-type recall response. Both IFN-y and IL-2 production were enhanced in IL-12-treated animals so that both IgGl and IgG2a were boosted following a secondary vaccination, either associated with IL-12 treatment or not, in animals primed by antigen and IL-12 (429). Administration of IL-12 during an ongoing immune response failed to permanently suppress and even enhanced antigen-specific IgE production (430), a result not completely surprising considering the ability of IL-12 in certain experimen-
INTERLEUKIN-12
1.51
tal conditions, both in vitro (168, 287, 288) and in vivo (429), to enhance IL-4 production and The-type responses. It was also shown that administration of IL-12 together with ovalbumin antigen profoundly, but transiently, inhibited antigen-specific IgE synthesis while enhancing IgG2a: these effects were much decreased on secondary challenge, and almost absent after tertiary challenge, although the increase in IgG2a was still significant in the latter condition (431). The presence of uninethylated CpG motifs in the bacterial DNA used for vaccination has been shown to be required for induction of proinflammatory cytolanes, including IL-12, and for effective intraderind gene immunization (126). Much interest has been generated by the possibility of using the IL-12 genes in combination with DNA vaccination. Plasmid DNA immunogens encoding IL-12 and either influenza virus or HIV antigens have proven very efficient in inducing a CTL response and IFN-7 production (432,433). When either GM-CSF or IL-12 was used in DNA immunization against various HIV antigens, the former enhanced antibody formation, whereas the latter decreased antibody formation and boosted CTL activity (432). However, when vectors encoding both GM-CSF and IL-12 were used, the two cytokines d ~ not d prevent each other’s effect on boosting antibody formation and CTL generation, but rather potentiated each other (338). The use of oral vaccination holds considerable promise in inducing protective iminunity against pathogens or in inducing tolerance, e.g., in autoimmunity. Oral immunization of mice with tetanus toxoid together with the adjuvant cholera toxin induced Th2-type responses with systemic IgG1, IgE, and IgA antibodies; if IL-12 was given ip to mice immunized orally, the response shifted to production of Thl-type cytokines, increased DTH, and increased serum IgG2a and IgG3, whereas IgG1, IgE, and IgA were markedly decreased (434). Interestingly, almost identical results, except for lack of effect on serum IgA, were- observed when IL-12 was given orally complexed to liposornes, a formulation that does not result in a significant level of serum IL-12 (434). Not only did IL-12 boost DTH responses during oral immunization, but when given at the site of attempted sensitization, it also induced powerful and long-lasting DTH reactivity in mice with already fully established, orally induced tolerance to ovalbumin (435). Two of the main mechanisms by which oral tolerance develops are the production of TGF-P by T cells and clonal deletion via apoptosis; systemic administration of anti-IL-12 antibodies to animals fed high doses of ovalbumin resulted in increased TGF-,f3 production and apoptosis, suggesting that the ability of IL-12 to prevent or revert oral tolerance may be due in part to the suppression of these two mechanisms of tolerance (436).
152
GIORGIO TRINCHIERI
XII. IL- 12 in Delayed-Type Hypersensitivity, Airway Hyperresponsiveness, and Graft Rejection
A. ROLEOF IL-12 IN DTH The ability of epidermal cells, keratinocytes, and Langerhans cells (198, 199,202) to produce IL-12, constitutively or on exposure to hapten, suggests a direct role for IL-12 in the development of IFN-.), producing cells in the skin that are important for DTH and contact hypersensitivity (CHS). There is good evidence that 1L-12 acts in the induction phase of CHS. Although IL-12 induction is not detected in the skin of mice on sensitization by topical application of haptens, unlike human skin (199), significant upregulation of IL-12 is detected in dendritic cells and macrophages of the regional draining lymph nodes (437,438). The critical functional role of IL-12 during cutaneous sensitization was clearly proven by the finding that ip injection of anti-IL-12 antibodies around the time of hapten sensitization resulted in failure to induce sensitization (437, 438). Injection of anti-IL- 12 not only prevented sensitization, but also induced tolerance: animals sensitized with hapten and treated with anti-IL-12 antibodies could not be sensitized to the same hapten after a 2-week rest, but were readily sensitized to an unrelated hapten, indicating the establishment of hapten-specific tolerance (438). Treatment of mice with IL-12 was also shown to enhance the acquisition of CHS, resulting in a response of greater magnitude and duration (439, 440). IL-12 injection around the time of sensitization increased the CHS response at the challenge phase with typical participation of effector CD8' cells, possibly in part by shifting the negatively regulatory CD4' cells with a Th2 phenotype to an effector T h l phenotype; in IL-12-treated CD8-depleted mice, challenge after cutaneous sensitization resulted in a CD4-mediated response with minimal edema and acute mononuclear cell infiltration, more typical of DTH than of CHS (440). In addition to its role in the sensitization phase of CHS, IL-12 also plays an important role in the elicitation phase. IL-12 can be detected in the skin at the site of challenge and is upregulated by neutralization of IL-4 (441), suggesting that focal production of Th2-type cytokines limits the inflammatory reaction. Furthermore, neutralization of IL-12 in the elicitation phase significantly suppresses the ear-swelling response (438), indicating that IL-12 is involved in the effector phase of CHS. Irradiation of animals with high doses of UV irradiation prevents the induction of DTH or CHS to hapten applied to distant unirradiated skin areas, inducing a systemic form of immunosuppression, whereas low doses of UV irradiation block CHS sensitization in the same skin area, possibly by depleting skin Langerhans cells (442). Interestingly, UV irradiation of
INTERLEUKIN-15
153
human inonocytes was shown to reduce IL-12 production, thereby limiting the activation of T h l cells (443). IL-12 treatment of the animals at the time of sensitization prevents both systemic and locally induced UV immunosuppression (437, 444, 445). The mechanism of prevention of systemic suppression may be due to the ability of IL-12 to act as an antagonist of IL-10, a likely mediator of this type of suppression (444). The ability of IL-12 to prevent local suppression is more obscure because UV irradiation induces depletion of the skin Langerhans cells, leaving open the question of which cells, in the presence of IL-12, become able to present the hapten (445). IL-12 not only prevents the induction of immunosuppression, but when administered at the time of resensitization, can overcome an established immunosuppressed state and revert it (444, 445). These results are reminiscent of the ability of IL-12, when injected at the site of sensitization, to reverse the tolerance to ovalbumin induced by oral immunization (435). Furthermore, the transfer of UV-induced tolerance by T cells to naive animals is prevented by IL-12 treatment of the recipients (444, 445). The transfer of tolerance is dependent on the presence of functional Fas and Fas ligand in the recipient, and IL-12 may act directly or indirectly on the recipients’ dendritic cells by preventing their apoptic death when presenting antigen to tolerogenic T cells (T. Schwarz, personal communication). A similar requirement for IL-12 in the induction of DTH, possibly acting directly on dendritic cells, has been reported for sensitization obtained by the transfer of dendritic cells, cells pulsed with a class I-restricted immunogenic peptide from the P815AB tumor rejection antigen of the murine mastocytoma cell line P815 (446,447).In vivo or in vitro treatment of the dendritic cells with IL-12 induces sensitization for DTH, but in the absence of IL-12, the dendritic cell transfer induces an anergic state (446, 447).
B. ROLE OF IL-12 IN GRAFT-VERSUS-HOST DISEASE The role of IL-12 in graft-versus-host disease (GVHD) has been studied primarily using mouse models, and very little information is available from human studies. Two distinct forms of GVHD have been analyzed: acute GVHD, typically observed in (C57BU6 X DBN2)Fl (BDF1) mice given parental C57B1/6 lymphocytes, and chronic GVHD observed in BDFl mice given parental DBN2 lymphocytes. Acute GVH is characterized by early lymphoid hyperplasia and increased NK cell activity, followed by generation of anti-host CTL, immunodepression, weight loss, and ultimately death. Chronic GVHD is an imrnunostiinulatory syndrome with Bcell hyperplasia, production of autoantibody, and immune complexinduced glomerulonephritis. Acute and chronic GVHD are associated with production of type 1 and type 2 cytokines, respectively.
154
GIORGIO TRINCHIERI
As expected, IL-12 stimulated the development of acute GVHD in mice that would develop chronic GVHD, an effect that was accompanied by the suppression of autoantibody formation, decreased serum immunoglobulin, generation of anti-host CTL, immunosuppression, weight loss, and death (448,449). However, results on the role of IL-12 in acute GVH are more complex. Acute GVHD is characterized by expression of type 1 cytokines, and IL-12 expression was detected in macrophages and target organs of mice (450),as well as in PBMC of patients undergoing acute GVHD (451), although serum levels of IL-12 were not associated with the development of acute GVHD in patients (452). Two studies observed that neutralization of IL-12 during induction of acute GVHD resulted in the expected polarization of the cytokine profile toward a Th2-type alloimmune response and conferred long-term protection from the disease, preventing generation of CTL, immunosuppression, weight loss, and death (449,453). Even when an enhanced Th2-type response was induced by neutralization of IL- 12, only modest activation of B cells was observed, with only moderate levels of antibodies and no gloinerulonephritis (453); thus, anti-IL-12 antibodies protect from acute GVHD without inducing chronic GVHD. However, one study (454) showed that administration of a single high dose of IL12 at the time of acute GVHD induction also significantly protected from the disease, reducing mortality and weight loss. Administration of IL-12 induced an early (days 2-3) increase in IFN-y production, which at this time derives from NK cells or NK1 T cells, followed by an inhibition of IFN-y production, prevalently from CD4' T cells, at day 4 (454). This surprising effect of IL-12 might rest in an activation of host NK cells or NK1 T cells, resulting in the suppression of iminunocompetent cells in the transplant or in the ability of a single injection of IL-12 to induce a partial unresponsiveness to IL-12 itself, as suggested by clinical trials (455) and experiments in mice (456). Another paradoxical result comes from animals in which acute GVHD was reduced by anti-B7.1 and -B7.2 antibodies; administration of anti-IL-12 antibody reversed the beneficial effect of the anti-B7 antibodies, possibly due to an impairment of natural immunity and hematopoiesis in anti-IL-&treated animals (457). Although these contradictory results remain difficult to fully interpret, the overall evidence points to the role of IL-12 as a central mediator of acute GVHD in mice. C. AIRWAYHYPERRESPONSIVENESS AND ASTHMA Asthma is characterized by increased ainvay responsiveness, elevated IgE, and chronic inflammation of the lung with infiltration of eosinophils and mast cells. This pathology is promoted by cytokines produced by Th2 CD4+ cells, particularly IL-5, which affects eosinophil infiltrates and IL4 and other cytokines, which affect mast cells. Mice of certain strains sensitized by ip immunization of antigens such as ovalbumin, sheep erythro-
INTERLEUKIK-12
155
cytes, or ragweed and then challenged intratraclieally or by aerosol administration of the antigen develop eosinophilia, increased IgE, and airway hyperresponsiveness in response to cholinergic agonists. IL-12 administered at high doses around the time of either sensitization or challenge suppresses eosinophilia, IgE levels, and hyperresponsiveness (458-461). At low doses of IL-12 (0.1 p g per injection), eosinophilia but not hyperresponsiveness was completely inhibited, whereas at high doses (1 pg per injection) both phenomena were inhibited (458).However, this dissociation between eosinophilia and ainvay hyperresponsiveness remains enigmatic because both IL-4 and IL-5 production were inhibited at either dose of IL- 12 (458). Inhibition of eosinophilia and hyperresponsiveness was observed even when IL-12 was administered at the time of second antigen challenge, reflecting the ability of IL-12 to inhibit responses associated with ongoing antigen-induced pulmonary inflammation (458). However, in other studies, IL-12 administered at the time of aerosol or intratracheal challenge was sliown to inhibit eosinophilia without affecting the production of specific IgE and to have a variable effect on airway hyperresponsiveness (459, 461). These effects of IL-12 were at least in part prevented by anti-IFN-y antibodies (458, 460). A possible role of a deficient expression of IL-12 in the generation of chronic lung inflammation and asthma is supported by data showing decreased numbers of IL-12-producing cells in bronchial biopsies from asthma patients compared with healthy controls (462) and a reduced production of IL-12 in vitro from S. nurezu-stimulated blood cultures of patients with allergic asthma compared with nonatopic donors (463). The differential airway antigen-specific immune response among different mouse strains was found to correlate with the ability of the strain to produce IL-12; strains susceptible to antigen-induced ainvay hyperresponsiveness produced lower levels of IL-12 (464). The critical role of IL-12 in the regulation of airway responses to allergen is supported by the finding that the treatment of C3H mice, a strain normally resistant to the induction of airway hyperresponsiveness, with the anti-IL-12 antibody at the time of ovalbumin airway exposure results in a three-fold increase in responsiveness, concomitant with significant increases in Tl12-type cytokines and a decrease in IFN-y at the pulmonary level (M. Wills-Karp, M. Wysocka, and G. Trinchieri, unpublished results). Thus, endogenous IL-12 appears to play a central role in preventing the induction of chronic bronchial inflanirnation and asthma, and treatment with recombinant IL-12 might reverse established pulmonary inflaininatory conditions.
D. ROLE OF IL-12
IN
ALLOGRAFTREJECTION 4
The reciprocal role of Thl- and Th2-type responses and cytokines in allograft rejection is a complex question, and assessing the role of IL-12
1S6
GIOKGIO TRINCI-IIERI
in such mechanisms has been a difficult task that has hitherto received only modest attention (373). Whereas T h l responses would be expected to play a determining role in graft rejection because of their ability to favor cell-mediated cytotoxicity and CTL generation, Th2-type cytokines have not been consistently shown to suppress graft rejection by antagonizing Thl response, and alternative mechanisms of graft rejection, e.g., those mediated by IL-5 and infiltrating eosinophils, may be supported by Th2 cells (373). Experiments to identify IL-12 production during graft rejection in both humans and mice have not been conclusive. Whereas the presence of IL12 p40 mRNA correlated with either rejection or viral hepatitis in human liver allografts (465), expression of IFN-y mRNA but not that of IL-12 p40 mRNA in renal allografts correlated with acute rejection (40). Rats in which donor-specific tolerance was induced by blood transfusions showed decreased expression of IFN-y mRNA in tolerized heart allografts compared to rejected allografts, but no significant difference was observed in IL-12 mRNA expression for either the p40 or the p35 subunit (466). In some experimental models, administration or blockage of IL-12 was indeed observed to correlate with graft rejection or prolongation, respectively. In the mouse model of skin graft prolongation by pretransplantation portal venous immunization with allogeneic cells, which results in preferential activation of Th2 cytokines, IL-12 in combination with anti-IL-10 antibodies reverse the graft prolongation (467), whereas anti-IL-12 in combination with IL-13 prolonged it (468). Although these results suggest an important role for IL-12 in regulating graft rejection, administering or antagonizing IL-12 by itself had only a modest effect on graft survival, and more consistent effects were observed only when the other regulatory pathways affecting Thl/Th2 balance were also modified, i.e., by blocking IL-10 or injecting IL-13. Another interesting model of IL-12 and graft rejection is the induction of high-level expression of the IL-12 p40 subunit in myoblasts by gene transfer with a retroviral vector (469);local production of excess p40 was expected to antagonize endogenous IL-12, and indeed the survival of IL-12 p40-transfected myoblasts transplanted in allogeneic recipients was substantiallyprolonged in association with impaired production of Thl cytokines, CTL generation, DTH responses, and, interestingly, a decrease in all IgG subclass antibodies (469). Unlike the previous results, Piccotti et al. (470) reported that treatment of mouse cardiac allograft recipients with either anti-IL-12 antibodies or IL-12 p40 homodimers resulted in an exacerbated graft rejection compared with control animals. Although Th2 cytokines were induced in the grafts of treated animals, IFN-y expression was not decreased and IL-12 p35 and p40 mRNA, undetectable in the control graft, became expressed (470).
INTERLEUKIN-11
157
The number of activated CTL in the graft was decreased by in v i m blocking of IL-12, although the number of CTL precursors was not affected (470). These results point to much of the complexity of ThlR112 regulation and the role of IL-12. The induced expression of IL-12 in these experimental conditions in which both Thl and Th2 cytokines are expressed may reflect the cooperation between IFN-.)I and IL-UIL-13 in priming for IL-12 production ( 159, 169). The differentiation of IFN-y-producing T h l cells reveals the existence of IL-12-independent mechanisms of Thl differentiation in certain experimental models, but not others. The results with IL12 antagonists were reproduced using IL-12 p40 genetically deficient mice, in which the duration of graft survival was also decreased in association with an apparently unaltered Thl response (470). Interestingly, Thl-type alloresponses were observed in both p35 and p40-deficient allograft recipients, although Th1 development was enhanced in p35-deficient recipients compared with their p40-deficient counterparts (373). The possibility that endogenous p40 stimulates a Thl response in p35-deficient inice was supported by the finding that treatment of these mice with anti-p40 monoclonal antibody decreased Thl functions to the level seen in p40-deficient recipients (373). These results are consistent with previous experiments (389) suggesting that IL-12 p40 homodimers may induce T h l differentiation in CD8+ T cells, while antagonizing IL-12 action on CD4+ T cells (389).In those experiments, p40 homodimers markedly prolonged allograft survival in mice depleted of CD8' T cells while inducing accelerated cardiac allograft rejection in unmodified recipients (389). Thus, although therapeutic manipulations of IL-12 activity and of the ThltTh2 balance may have the potential to affect graft survival, the complex and redundant regulatory mechanisms and the possible role of both Th1 and Th2 cells in allograft rejection are not well understood and thus make it difficult to predict the outcome of any therapeutic manoeuvre. XIII. 11-12 in Organ-Specific Autoimmune Diseases
A. ROLE OF ThltTh.2 RESPONSESI N AUTOIMMUNITY A role for Thl cells has been demonstrated or suggested in inany organspecific autoimmune diseases, both in humans and in mice, although exactly how these cells mediate their action, and in particular the requirement for IFN-y production and the relative importance of regulation of iinmunoglobulin production, varies from disease to disease. Also, the interpretation of the pathological mechanisms of human diseases and the possible role of IL-12 have been extrapolated in many cases from data in animal models, which often poorly reproduce the clinical situation. A common observation in models of autoimmune diseases is the apparent requirement for continu-
158
GIORGIO TRINCHIERI
ous expression of IL-12 for maintenance of a pathogenic T h l response (378),whereas in infectious diseases IL-12 is usually required only in the very first few days of infection and iininunological response (377). This difference may be due to a different nature or strength of the T h l response in autoimmunity in general, but it may also reflect particular characteristics of each disease situation. For example, the continuous requirement for IL12 expression in experimental allergic encephalitis may be due to epitope spreading, with continuous activation of new T-cell clones; in experimental colitis, the role of IL-12 may be to suppress TGF-&producing downregulatory T cells or to prevent tolerance induction by apoptosis; in collageninduced arthritis, the upregulation of Th-l-dependent antibody isotypes, mediated by continuous production of IFN-y, appears to play a major role. Information on the role of IL-12 in autoimmune diseases remains very incomplete, and analyses to date have provided seemingly inconsistent results. However, improved knowledge of these regulatory mechanisms should further our understanding of the pathogenic mechanisms in experimental animals and possible therapy of human autoimmune diseases.
B. INSULIN-DEPENDENT DIABETES MELLITUS Nonobese diabetic (NOD) mice and diabetes-prone BioBreeding (DPBB) rats spontaneously develop a diabetic syndrome that resembles human type I diabetes. Islet P-cell destruction involves the participation of infiltrating mononuclear cells, with a role for both CD4' and CD8+ T cells. The initiating event precipitating insulin-dependent diabetes mellitus (IDDM)is thought to be the recognition by T h l cells of islet cell antigens, many of which have been identified (471). A role for IL-12 in the spontaneous development of IDDM in female NOD mice is suggested by the increasing expression of IL-12 p40 and p35 mRNA in mononuclear cells from islets of these animals from age 5 weeks to the onset of diabetes (472). Treatment with a single dose of cyclophosphainide,which synchronizes and accelerates the disease in NOD mice, resulted in substantially increased expression of IL-12 p40 mRNA in both pancreas and spleen of NOD mice, an effect that was not observed in cyclophosphamide-treated BALB/c mice (473). Conversely treatment of NOD mice with complete Freunds adjuvant in early life, which protects from IDDM, resulted in decreased IL-12 p40 mRNA expression in the islets (472). Daily IL-12 administration for 30 days to female NOD mice resulted in the rapid onset of IDDM, with 100% incidence by 4 weeks and massive infiltration of lymphoid cells in the pancreatic islets (474). Only 60% of control untreated female NOD mice developed IDDM and at much Iater times, whereas IL-12-injected BALB/c mice did not develop
INTERLEUKIN-12
1Fj9
IDDM and their islets had a normal appearance (474). Treatment of cyclophosphamide-treated NOD mice with either IL-12 p40 hoinodimers (474) or neutralizing anti-IL-12 antibodies (L. Harrison and G. Trinchieri, unpublished results) to neutralize endogenous IL-12 resulted in a significant suppression of IDDM, associated with a reduction of islet infiltration by mononuclear cells (475). Mice deficient in IL-12 by targeted disruption of the p40 gene and backcrossed to the NOD background showed a reduced incidence of IDDM, confirming that endogenous IL-12 is required for IDDM development (476). IGIF/IL-18 was also detected in pancreas and spleen of diabetes-prone mice on treatment with cyclophosphamide and preceding the appearance of IFN-y, but not in nondiabetes-prone strains (477, 478). The IGIF gene maps within the IDD2 interval, one of the dmbetes-susceptibility loci, on mouse chromosome 9 and therefore it is a candidate for IDD2 susceptibility gene (477). Overall, these data suggest that expression of endogenous IL-12 in the islets of NOD mice parallels the progression of the inflammation and islet destruction and that IL-12 neutralization prevents the activation of the dabetogenic T cells, whereas the daily administration of excess exogenous IL- 12 hastens islet destruction. However, using a different protocol of IL-12 administration, i.e., five high-dose treatments in 2 weeks or weekly higli-dose injections, O’Hard et d.(479) showed a significant protection of IDDM development in female NOD mice. The weekly IL-12 treatment was particularly efficient in preventing IDDM, but no histological differences were observed in the islet infiltration in mice treated with IL-12 or not (479). In a model of adoptive transfer of IDDM into male NOD mice using spleen cells from diabetic female NOD mice, IL-12 treatment twice per week of the recipient mice did not prevent the transfer of the disease, suggesting that this schedule of IL-12 treatment prevents the development of diabetogenic T cells, but not their effect (479).These results are difficult to interpret and show that exposure of T cells to IL-12 at different times during their development and differentiation may lead to different outcomes. However, it is also possible that the weekly, high-dose treatments result in desensitization of the T cells to endogenous IL-12, similar to the findings of the effect of IL-12 predosing in cancer patients or in animal tumor models (455, 456). Information on the role of IL-12 in IDDM of DP-BB rats is much more limited. However, like the NOD mouse model, the pancreatic islets and thyroid of DP-BB rats showed an increase in IL-12 p40 mRNA with age of the animal and progression to disease (480). No equivalent increase in IL-12 p40 mRNA was observed in diabetes-resistant BB rats, but both IDDM development and IL-12 inRNA expression could be induced in these rats either by deletion of regulatory RT6’ T cells (480)or by infection
160
GIORGIO TRINCHIERI
with Kilham rat virus (481). Thus, in both the mouse and the rat, development of IDDM appears to correlate with IL-12 expression in the target organ. ALLERGICENCEPHALOMYELITIS C. EXPERIMENTAL AND MULTIPLESCLEROSIS Experimental allergic encephalomyelitis (EAE) is an autoimmune disease of the CNS and the most commonly used model for human multiple sclerosis (MS). Most of the EAE models used are monophasic and demyelination is minimal, but other models in which the disease is severe, protracting, and relapsing with extensive demyelination more closely resemble the clinical disease. IL-12 was shown to greatly affect a model of adoptively transferred EAE induced by injecting naive SJL/J mice with lymph node cells from mice primed with proteolipid protein (PLP) and restimulated in vitro with PLP (482). Addition of IL-12 during the in vitro restimulation resulted in a much more severe and prolonged disease after in vivo transfer; IFN-y and TNF-a were increased in the supernatant of the cells restimulated in the presence of IL-12, but neutralization of either cytokine did not decrease the severity of the transferred disease, suggesting that IL-12 directly affects the activation of the encephalitogenic T cells (482). The extent of perivascular infiltration and cytokine production was similar in animals receiving lymph node cells restimulated in the absence or the presence of IL-12, but in the later group, there was a profound increase in inducible nitric oxide synthase (iNOS) in macrophages (483), suggesting that the treatment with IL-12 enhances the ability of the transferred cells to induce the production of nitric oxide within the inflammatory foci, with a possible direct cytotoxic effect on oligodendrocytes. Unlike SJL mice, lymph node cells of BlOS mice restimulated in vitro with myelin basic protein (MBP) do not transfer EAE, unless IL-12 or IL-12 inducing microbial products such as LPS or bacterial DNA are added during the in vitro restimulation, indicating that IL-12 can unmask latent autoimmune disease in resistant mice (484, 485). In vivo treatment of the recipient SJL mice with IL-12 also increased the severity of disease, whereas treatment with anti-IL-12 antibodies for 12 days completely prevented the paralysis, with only a proportion of animals developing mild disease (482). Interestingly, if recipient mice were treated with anti-IL-12 antibodies for only 6 days, the disease was delayed, but its severity was not affected (482). Similar to the inability of anti-IFN-y to prevent the effet of IL-12 during the in vitro restimulation of lymph node cells, in vivo administration of anti-IFN-y antibodies to IL-12-treated recipient mice did not prevent
INTERLEUKIN-12
161
the exacerbating effect of IL-12, but rather resulted in even more severe disease (486). These observations are consistent with previous results indicating a protective role for IFN-y in EAE (487, 488) and indicate that the ability of IL-12 to induce encephalitogenic cells or to enhance their activity is independent of its ability to enhance IFN-y production. In a model of relapsing EAE induced by MBP immunization in (SJL X PL/J) F1 mice, IL-12 treatment in vim during the remission phase induced disease relapse and strongly enhanced the severity of spontaneous relapses, whereas anti-IL-12 antibodies prevented spontaneous relapse as well as the severe relapse induced by treatment of the mice with bacterial superantigens (489). Overall, the results in transfer or relapsing models of EAE indicate that, unlike the observations in infectious disease (377), IL12 in EAE is not only needed for the initiation of the T-cell response, but also for its maintenance and for the encephalitogenic effector phase. However, it remains unknown whether IL-12 affects already differentiated T cells or contributes to the recruitment of new naive T cells, possibly specific for different epitopes of the antigen. In rats, immunization with spinal cord emulsified in complete Freunds adjuvant induces a monophasic disease in the Lewis strain, but a severe and relapsing disease in the DA strain (490). The expression of IL-12 inRNA in the spinal cord of the two strains after immunization is quite similar, whereas in DA rats the expression of other proinflammatory cytokines is prolonged and production of immunodownmodulating cytokines such as TGF-P and IL-10 was almost absent (490-492). Similar to observations in mice, administration of IL-12 to MBP-immunized Lewis rats induces EAE relapse. However, although IL-12 is effective when administered up to 1week after recovery from the primary bout of disease, it is not effective when administered at later times; the relapses were characterized histologically by greater perivascular inflammation in the CNS and the induction of iNOS-positive cells (493). In clinical MS, both IL-12 and B7.1 expression has been detected in acute MS plaques in the CNS, particularly in early disease cases (494). Elevated production of IL-12 in response to anti-CD3 stimulation was observed in PBMC of MS patients with progressive disease, whereas patients with remitting-relapsing disease produced IL-12 at levels similar to those of healthy controls (49S).The induction of elevated IL-12 production mediated by the T cells from these patients was due to an increased expression of CD40L on the anti-CD3-stimulated T cells, which induced IL-12 production from either autologous or allogeneic non-T cells (495). MS patients with progressive disease have slightly higher levels of serum IL-12 than do healthy subjects or patients with other neurological dneases
162
ClORGIO TRINCHIERI
(496); however, no correlation was observed between the presence of MS or disease severity and the level of IL-12 in cerebrospinal fluid (497). Overall, these findings raise the possibility that an early event in the initiation of MS involves upregulation of B7.1 and IL-12, resulting in conditions that synergistically stimulate T-cell activation and the effector phase of a Thl-type immune response. In addition to EAE, analysis of experimental uveoretinitis (EAU) in mice has also implicated a role for IL-12. Anti-IL-12 antibodies in v i m prevented the development of EAU induced by immunization of B1O.A mice with interphotoreceptor retinoid-binding protein and conferred resistance to subsequent antigen challenge; however, anti-IL-4 antibodies at the time of rechallenge reversed the protection induced by IL-12 (498).This finding strongly suggests the importance of the Thl/Th2 equilibrium in induction/ protection of EAU (498), a conclusion strengthened by the observations that EAU can be adoptively transferred with T cells from mice primed in v i m and restiinulated in vitru with a peptide from the retinoid-binding protein only if the in vitru restimulation is perfonned in the presence of IL-12, and that in general the ability of the primed cells to transfer EAU correlates with their level of IFN-y production (499). D. COLLAGEN-INDUCED ARTHRITIS Collagen-induced arthritis (CIA) is a murine model for human rheumatoid arthritis, characterized by a severe swelling in the joints with a massive inflammatory infiltrate, which leads to joint destruction and deformities. CIA is induced in DBA/1 mice approximately 4 weeks after immunization with type I1 collagen emulsified with Mycobncterium tuberculosis in oil (complete Freund’s adjuvant). Immunization in the absence of M . tuberculosis or immunization of other strains (e.g., C57B1/6 or Bl0.Q) does not induce CIA. Treatment with IL-12 for 5 days at doses from 50 ng to 1 pg at the time of immunization or at time of onset of the disease was shown to induce severe disease even when the mice were immunized with collagen in oil only (incomplete Freund’s adjuvant) (500, 501). The induction of CIA was associated with enhanced IFN-y synthesis and strong upregulation of anticollagen antibodies, especially of the Thl-dependent IgG2a and IgG2b isotypes (500). Neutralization of IFN-y in vivu prevented the development of arthritis, but not the priming of Thl cells by IL-12 (500,502). IL-12 treatment of collagen-immunized mice of the resistant C57B1/6 or Bl0.Q strain failed to induce CIA; in these animals, IL-12 potentiated the IFN-y production by collagen-specific T cells, but did not enhance and instead decreased the titer of anti-collagen IgG2a or IgG2b antibodies (503).
INTERLEUKIN-18
163
Although those results suggested a clear role for IL-12 induced T h l cells in CIA, likely mediated through their effect on antibody formation, it was surprising to find that treatment with high doses of IL-12 (1 pg/ treatment) in complete Freund’s adjuvant for 1 or more weeks immediately after the immunization of DBAA mice efficiently protected against development of CIA (504). This protective effect was paralleled by a strong inhibition of collagen-specific IgGl antibodies, a modest decrease of IgG2b, and no significant change in IgG2a (504). These contrasting results led to the conclusion that IL-12 can both suppress and enhance CIA in DBA/1 mice, depending on the adjuvant used (504). However, in the two sets of experiments, different schedules and dosages of IL-12 administration were used, and, as in other models, it is possible that the suppressive effects of high doses reflect a desensitization to endogenous IL-12 (45,5, 456) or a reactive production of anti-inflamniatory mediators such as IL-10 or glucocorticoids (168, 486, 505). Inhibition of endogenous IL-12 by antibodies to IL-12 or byp40 homodimers only inconsistently ameliorated or delayed the onset of CIA, and in some cases, short duration anti-IL-12 treatment even enhanced the disease (M. Feldmann, personal communication). These results may be due to different requirements of IL-12 during the various stages of CIA induction, but the inconsistency of the results suggests that anti-IL-12 reagents were used that do not reproducibly and completely block IL-12 in uitro, as in other studies the use of higher affinity monoclonal antibodes or polyclonal antisera to IL-12 more consistently abrogated CIA ( M . Feldinann, personal communication, 501). Furthermore, IL-12 p40 genetically deficient mice, backcrossed into a DBN1 background, have a much reduced severity and incidence of CIA associated with a decreased antibody titer and IFN-7 production (506). Thus, these results confirm an important role for IL-12 in CIA onset, although in established disease IL-12 may have a suppressive role, as indicated in mice at late stages of CIA by an impressive exacerbation of arthritis shortly after cessation of anti-IL-12 treatment and by the ameliorating effect of IL-12 treatment of these stages (501). This antiinflammatory role of IL-12 on established CIA was associated with a 10-fold enhanced level of IL-10 in the serum of treated animals and was reversed by coadministration of anti-IL-10 antibodies (501),indicating that the ability of IL-12 to induce IL-10 may represent an important negative feedback mechanism in this elcperimental system to prevent excessive activation and tissue pathology. Very little information is available on the role of IL-12 in clinical rheumatoid arthritis, aside from the detection of IL-12 p40 inRNA and production of the IL-12 heterodimer in synovial samples from patients with this disease (507, 508).
164
GIORGIO TRINCHIERI
E. EXPERIMENTAL COLITISI N MICE AND CROHN’S DISEASE Crohn’s disease is a chronic inflammatory bowel disease that most commonly affects the terminal ileum and ascending colon. Much evidence suggests that immunological mechanisms are responsible for the disease, specificallythat Thl-type T cells possibly activated in response to microbial insults play a dominant role. In the mouse, intrarectal administration of 2,4,6-trinitrobenzene sulfonic acid (TNBS), which haptenates autologous colonic protein with trinitrophenyl (TNP), induces an intense immune response and a massive infiltration resembling the human Crohn’s disease (509). In this model of experimental colitis, administration of antibodies to IL-12 both early (at 5 days) and late (at 20 days) after induction of colitis leads to a striking improvement in both the clinical and the histopathological aspect of the established disease, often with complete regression of the pathology (509). These results are paralleled by a failure of the CD4’ T cells from the lamina propria of the anti-IL-12 treated animals to secrete IFN-.)I on in vitro stimulation (509). Injection of anti-CD40L antibodies at the time of TNBS administration also protected the mice from colitis and inhibited IL-12 production; IL-12 treatment reversed this protection effect of anti-CD40L antibodies (149). Interestingly, antiCD40L antibodies protected the animals from colitis only during the inducing phase of the disease, whereas anti-IL-12 antibodies also abrogated established disease, suggesting that the T-cell-dependent mechanism of IL-12 production is involved only at the early phases of the response and that IL-12 production is then maintained by a mechanism that depends on microbial or inflammatory products. Injection of TNBS-treated mice with anti-IL-12 antibodies also restored the T-cell tolerance against resident intestinal flora that, as in Crohn’s disease, is abrogated in experimental colitis (510). Explanation of these results may be provided by data in a different experimental model using mice transgenic for antiovalbumin TCR, in which IL-12 was shown to negatively regulate the two main mechanisms of mucosal tolerance: TGFp production and clonal deletion via apoptosis (436, 511, 512). IL-2 genetically deficient mice provide yet another experimental model of colitis; most of these animals develop colitis spontaneously, but the disease can also be experimentally induced in a time-controlled way by immunization with irrelevant exogenous antigens (e.g., TNP-KLH or TNPOVA) (513). IL-12 is abnormally expressed in the colon of immunized mice, and anti-IL-12 antibodies prevent the development of the colitis (514). In these mice, thymocyte maturation in the absence of IL-2 is abnormally directed by IL-12 toward the generation of single-positive, mature-activated Thl-type thymocytes capable of mediating colitis; these
INTERLEUKIN-12
165
defects in thyinocyte maturation are prevented by treatment with anti-IL12 antibodies (513). The important role of IL-12 and Thl cells in colitis is also supported by results in Crohn’s disease patients; IL-12 production was detected from lamina propria mononuclear cells, where a predominance of T h l cells was demonstrable (515, 516), and culture of these T cells in vitro in the presence of anti-IL-12 antibodies downregulated the development of IFNy-producing CD4’ T cells (516).
F. SPONTANEOUS AUTOIMMUNE DISEASE IN MRL/lpr MICE MRL/lpr mice develop a spontaneous autoimmune disease characterized by lymphadenopathy, autoantibody production, and inflammatory manifestations, such as nephritis, vasculitis, and arthritis, and have been used extensively as a model for clinical systemic lupus erythematosus (SLE). Spleen and peritoneal cells from MRL/lpr mice produce significantly more IL-12 than do cells from MRL/+ or BALB/c mice, and this production results in increased release of IFN-y and nitric oxide (517). Sera from MRL/lpr also contain increased levels of IL-12 compared to the control mice (517), and IL-12 is upregulated in the tubular epithelium of MRL/ lpr mice with nephritis (518).Daily injections of IL-12 into MRL/lpr mice led to increased serum levels of IFN-y and nitric oxide, and to accelerated and more severe glomerulonephritis (517). Strikingly, the pyelonephritis with extensive vasculitis and infiltration of mononuclear cells at the kidney medullary region was prominent in the control animals and almost completely abrogated in IL-12 treated mice (517).This paradoxical observation likely rests in the fact that glomerulonephritis is autoimmune, whereas pyelonephritis is reactive to local infection: thus IL-12 exacerbates the former while, by increasing antimicrobial resistance, improves the latter (517). In another autoimmune manifestation of MRL/lpr mice, i.e., the Sjogren-like syndrome characterized by lymphoid infiltration in salivary and lacrimal glands, local upregulation of IL-12 and IFN-.)Iexpression was observed (519, 520). Although MRL/lpr mice are often considered a model for human SLE, the autoimmune diseases in the two situations may be very different, and the predominance of Thl-type responses observed in MRL/lpr is unlikely to be a characteristic of human SLE. Indeed, in vitro IL-12 efficiently inhibited the elevated immunoglobulin production by SLE patients’ PBMC through a mechanism independent of the observed increased IFN-y and decreased IL-10 production (521). Furthermore, PBMC from newly diagnosed SLE patients produced in vitro less IL-12 and more IL-10 than PBMC from healthy controls (D. Honvitz, personal communication).
166
GIORGIO TRINCIIIERI
XIV. 11-12 in the Inflammatory Response
The ability of IL-12 to induce and regulate the production of IFN-7 and other cytokines such as GM-CSF and TNF-a, which are major stimulators of phagocytic cell activation, makes it a powerful proinflammatory cytokine, with a major role in local inflammation during infections and autoimmune responses. In addition, IL-12 plays a predominant role in the systemic inflammatory response syndrome and could play a direct or indirect role in the multiorgan failure or dysfunction observed in systemic inflammation. IL-12 also appears to be particularly sensitive to the downregulatory mechanisms (or compensatory antiresponse syndrome) activated in response to the inflammatory response. Thus, the immunosuppressive state following major infection or trauma may be due in part to IL-12 deficiency and corrected, as discussed later, by IL-12 treatment. Nakamuraet al. (522) reported that endotoxin treatment of mice induced a serum factor with a molecular mass of about 70,000 that stimulated IFN-y production. Later, these authors partially purified the factor and showed its identity with IL-12 (523). The in vivo production of IL-12 was then shown to be an absolute requirement for IFN-y production in mice treated with high (300 p g ) or low (1p g ) doses of LPS after priming with BCG infection (161, 162). IL-12 p70 is produced around 3 hr after LPS injection and reaches concentrations of up to several nanograms per milliliter in the serum, although the concentration of IL-12 p40 (both monomers and homodimers) is always manyfold higher (50, 161,162). At 2-3 hr after LPS injection, increased accumulation of niRNA for both IL-lBRPl and P2 in spleen cells is observed (L. Showe, M. Wysocka, and G. Trinchieri, unpublished results), suggesting that both IL-12 and IL-12R are expressed at this time. IFN-y is produced with a peak around 6-7 hr and its production is inhibited by neutralizing antibodies to IL-12 (161, 162), which also protect the animals from death from endotoxic shock (161). The requirement for IL-12 in IFN-7 production was confirmed in IL-12 p40, IL-12 p35, and IL-12RP1 genetically deficient mice, all of which produced 10-fold lower levels of IFN-.)Iin response to LPS than did wild-type animals (72, 279, 280). However, IL-12 alone is not sufficient for optimal IFN-y induction, and important cofactors are represented by TNF-cu (161) and IL-18 (265, 266). Although IL-12 is required for IFN-7 production, the production of IL-12 is IFN-y independent and observed in IFN-y-deficient mice (163). In baboons, IL-12 production in vivo was observed during experimental Escherichia coli-induced septic shock, but IL-12 production, unlike IFN-y, did not correlate with the dose of bacteria or with severity of shock (524). Thus, IL-12 production is likely under the control of downregulatory mechanisms, e.g., IL-10 production, and other cofactors,
INTERLEUKIN-12
167
together with IL-12, are responsible for determining the level of IFN-y produced. The Shwartzman reaction in mice is elicited by two injections of LPS: the priming injection given in the footpad and the lethal LPS challenge given intravenously 24 hr later. IL-12 or IFN-y can replace LPS in the priming injection, and anti-IL-12 prevents the priming effect of LPS, suggesting that IFN-y indiiced by IL-12 following LPS injection is responsible for the priming (525). IL-12 was also shown to account for the ability of BCG to sensitize mice to the lethal effect of TNF-a and to be able to replace BCG in this sensitizing effect (526). LPS can cause tolerance to its own action both in vivo and in vitro; LPS-induced tolerance is characterized by a decreased synthesis of various cytokines, particularly TNF-a on LPS rechallenge (527). In vitro, LPS desensitization for TNF-a production can be prevented and reversed by treatment of monocytes with IFN-y or, if IFN-y-producing nonmonocytic cells are present, by IL-12 via IFN-y induction (527). IL-12, both p40 and p70, are even more sensitive than TNF-a to LPS desensitization and, unlike TNF-a, the inability of LPS-pretreated inonocytes to produce IL-12 is not readily reversed by IFN-y treatment or by using a different inducer such as S. azcreus (C. Karp, M. Wysocka, X. Ma, and G. Trinchieri, unpublished results). In chronic multisystem inflammatory disorders such as Behqet’s disease (528) or sarcoidosis (529), disease activity correlates with the expression of elevated levels of IL-12 in the plasma and in the bronchoalveolar lavage fluid, respectively. These data suggest that these syndromes are Thlmediated diseases that may be driven by chronic, dysregulated production of IL-12 at the site of disease (529). Interestingly, an elevated expression of IL-12 was also found in human atherosclerotic plaques, and IL-12 was inducible in monocytes by highly oxidized low-density lipoprotein (LDL), but not by minimally modified LDL (530). These data clearly point to the possibility that production of IL-12 and the proinflammatory network plays a major role in the pathogenesis of the atherosclerotic plaque. A state comparable to LPS tolerance is observed after severe injury by trauma or bum, manifested in part by a severely depressed ability of mononuclear cells from the patients to produce IL-12 (531, 532). This endotoxin tolerance assumes clinical relevance because it likely underlies the great susceptibility of major trauma victims to infections. In a mouse model of burn injury, in which decreased production of IL-12 was also observed (53l),treatment with low-dose IL-12 (25 ng daily for 5 days) increased survival of the burn-injured mice after cecal ligation and puncture to the same level as the sham-burn control group (533).Although IL-12 acts through IFN-y, and IFN-.)I treatment was partially effective, IL-12
168
GIORGIO TRINCHIERI
was determined to be the most effective therapy tested so far in this model of burn-induced immune unresponsiveness (533).In a similar model, IL-12 was also effective in protecting bum-injured mice from herpes simplex type 1 infection, to which they are very susceptible (534). XV. 11-12 in Infectious Diseases
A. VIRUSES
1. Experimental Viral Infections Because of its ability to induce production of IFN-y, to favor T h l responses, and to enhance CTL generation, IL-12 was expected to play an important role in viral infections. Although some of this expectation proved correct, IL-12 does not appear to play as important a role in viral infections as it does, for example, in many bacterial and intracellular parasite infections, and IL-12-independent mechanisms of IFN--y production and generation of CTL are operative in antiviral immunity (331). Early attempts to treat viral infections with IL-12 showed that low doses of IL-12 (1-10 ng daily per injection) had some protective effect on lymphocytic choriomeningitis virus (LCMV) and murine cytomegalovirus (MCMV) infections, whereas higher doses of IL-12 (10-1000 ng daily per injection) were detrimental to resistance against LCMV infection (535). These high doses of IL-12 (1)inhibited CTL activity, (2) inhibited virusinduced CD8' T-cell expansion, (3)induced necrotic lesions in splenic white pulp, (4)resulted in >2 log increase in splenic and renal viral replication, and (5)decreased body weight and thymus mass (505,535). These IL-12 toxicities were prevented by treatment with neutralizing antibodies to TNF-a: LCMV infection was shown both to synergize with IL-12 in inducing in vim TNF-a production and to sensitize target tissues, particularly CD8+ T cells, to the effect of TNF-a (505).IL-12 treatment of LCMV-infected animals also induced an increase in circulating glucocorticoid levels, which were secondary intermediaries in the dramatic thymus atrophy induced by IL-12 (505). Unlike LCMV infection, the effective defense against MCMV infection requires NK cell-produced IFN-y, and IL-12 enhances this defense pathway (536).In particular, IL-12 treatment of MCMV-infected animals increased NK cytotoxicityand IFN-y production and resulted in an improved antiviral status; virus-induced hepatitis was decreased up to 50-fold and viral burden decreased below the level of detection (536).These protective effects of IL-12 were prevented by depletion of either NK cells or IFN-y (536).Similarly, a single injection of 20 ng IL-12 18 hr before a lethal challenge with encephalomyocarditis virus (EMCV) protected all the animals from death with a mechanism mediated by IFN-y (537).
INTERLEUKIN-12
169
Early expression in uivo of IL-12 inRNA or protein was observed after infection with several different viruses, including lactate dehydrogenase elevating virus, mouse hepatitis virus (MHV), mouse adenovirus (538), herpes simplex virus 1 (HSV-1) (539),MCMV (540), and influenza virus (541). The production of IL-12 in response to virus infection peaks in the first 1to 3 days after infection and is usually transient. In MCMV infection, the early IL-12 production is responsible for NK cell production of IFN-y; treatment of mice at day 3 of infection with anti-IL-12 antibodies decreased IFN-y production and resulted in a 1 log increase in virus titer (540). However, treatment with anti-IL-12 antibodies at days 7-9 after infection had no effect on IFN-y production from T cells or on virus clearance; at no time did anti-IL-12 treatment decrease NK cell cytotoxicity or T-cell functions (540). Almost identical results were obtained with infection of mice with the PR8 strain of influenza A virus: IL-12 was induced with a peak at 2-3 days after infection and anti-IL-12 antibody treatment at day 3 inhibited IFN-y production, mostly from NK cells, and resulted in a 1 log increase in lung virus titer, whereas treatment at day 7 did not affect IFN-y production or virus titer (541).Thus, in both MCMV and PR8 virus infection, the production of IL-12 and NK cell-derived IFN-y are important in early resistance to virus infection; however, IL-12 has little if any effect on the antigen-specific T-cell response (CTL activity and IFN-y production) and on the eventual virus clearance mediated by T cells. The observation that IL-12 is differentially regulated during various virus challenges, in particular the lack of production of IL-12 in LCMV infection compared to the early production in MCMV infections, moved Cousens et al. (542) to investigate the role of IFN-a/P in these infections. They reported that I F N - d P inhibited IL-12 and IFN-y production from S. aureus-stimulated mouse splenic cells, whereas TNF-(r and IL-6 were not inhibited (542).In uivo neutralization of IFN-d/3 expressed endogenously during MCMV infection enhanced early IL-12 and IFN-y protein production and, interestingly, revealed an early production of IL-12 and IFN-7 in LCMV infection that was completely undetectable in untreated LCMVinfected animals (542). These results suggest a new interplay between IFN-dP and IL-12: resistance to infection with viruses that induce high and efficient early expression of IFN-(r/P is independent of early IL-121 IFN-y production, whereas early production of IL-12 and IFN-y from NK cells is observed in infections with viruses that are poor inducers of I F N - d P production, and this early IL-12 production is important in controlling virus infection until an IL-12 independent, antigen-specific Tcell response is established.
170
GIORGIO TRINCHIERI
Infection of the central nervous system with vesicular stomatitis virus (VSV) is a particularly good example of the possibility of using IL-12 to treat viral infections. IL-12 administered intraperitoneally strongly enhanced immunity to VSV infection in the CNS, as indicated by (1)decreased viral titer, (2) increased expression of iNOS, ( 3 ) enhanced expression of class I and class I1 MHC antigens, (4)increased T-cell infiltration, and (5)decreased VSV-induced apoptosis (543).This antiviral effect of IL-12 treatment of VSV infection in the CNS was also observed in IFN-y genetically deficient mice and is therefore IFN-y independent (544). There is much clinical interest in the possibility that antiviral cytokines are effective in the therapy of chronic viral hepatitis. IL-12 was shown to be particularly effective in protection against M HV-induced hepatitis, an effect dependent on IFN-y (545).In mice transgenic for the hepatitis B antigen, IL-12 suppresses autoantibody formation by shifting the Th2 response to T h l predominance (546),and in hepatitis B virus transgenic mice, IL-12 treatment inhibits virus replication in liver and kidneys and induces clearance of the cytoplasmic hepatitis B core antigen from both tissues via IFN-.)Iinduction (547).These data suggest that IL-12 treatment suppresses the Th2-type cells that are probably involved in maintaining the infection in chronic hepatitis while inducing a Thl response that, through IFN-y production and noncytolyhc mechanisms, efficiently clears the virus. These models suggest that IL-12 may have therapeutic value for the treatment of chronic hepatitis virus infection. The ability of IL-12 to act as an adjuvant in vaccination and to enhance a Thl-type response is of obvious interest in preventing viral infections. However, to date, results have been reported only for respiratory syncyt~al virus (RSV). Because many of the clinical manifestations of RSV infection are due to the nature of the immune response to the virus, with a Th2-type response considered to be responsible for the severe lung inflammation observed in vaccinated children, a vaccination protocol that preferentially induces a Thl-type response was expected to be optimal. Indeed, immunization of mice with formalin-inactivated, alum-precipitated RSV in conjunction with IL-12 at the time of immunization resulted in inhibition of virus replication, increased IFN-7 production, and increased IgG2a antibodies on challenge (548). Thus, IL-12proved to be apowerful adjuvant for RSV vaccination, inducing a strong Thl-type response; however, no significant effect was observed on the clinical outcome or on CTL generation (548,549).In another model in which mice were vaccinated with RSV glycoprotein G as part of a recombinant vaccinia virus and then challenged with RSV, the presence of IL-12 during vaccination also determined a change in the nature but not the severity of the inflammatory lung disease observed following the RSV challenge (550).IL-12 treatment
INTERLEUKIN-12
171
during vaccination determined an almost complete disappearance of eosinophils and B cells from the lung, but an increase in the lymphoid infiltration, particularly of IFN-y-producing CD4' T cells, which resulted in similar or even more severe illness (550).Thus, in this model, reversal of the Th2associated pathology does not necessarily benefit the host (550).Interestingly, the use of anti-IL-4 instead of IL-12 during vaccination with inactivated virus resulted in the predominance of IFN-y-producing CD8' T cells in the lung, which was associated with less severe illness: based on these results, it was proposed that the phenotype of effector cells involved in the iininune response to virus challenge is a more important determinant of disease than the patterns of cflokine expression classically assigned to Thl and Th2 lymphocytes (549): A particular use of IL-12 in inducing deviation of the immune response during gene therapy has been reported by Yang et al. (551).The efficacy of gene therapy in the lung using a replication-defective adenovirus vector is limited because the infected cells expressing the transgene are rejected by the host iininune system within several weeks, and reinfection is impossible because of neutralizing IgA antibodies; however, the administration of IL12 to mice at the time of initial gene therapy treatment suppresses the generation of the neutralizing IgA and allows successful reinfection of the lung and repeated gene therapy (551).
2. Measles Vinis Infection with measles virus still kills 1-2 million children annually, mostly because the infection is accompanied by a marked and prolonged abnorinality of cell-mediated immunity that contributes to increased susceptibility to secondary infections, the major cause of death. Measles infection and vaccination are characterized by a predominant expression of Th2-type cytokines, and this iininune deviation is probably responsible for the observed cellular immunodeficiency. Karp et al. (194) observed that in vitro infection of human nionocytes with measles virus resulted in a profound inhibition of the ability of the cells to produce IL-12, whereas the production of inany other proinflammatory cytokines was almost unaffected. This suppression was observed in response to various bacterial stimuli or CD40 stimulation. The effect was at the transcriptional level and, interestingly, although measles virus infection selectively affects the production of IL-12, the induced transcription of both the p40 and the p35 genes was suppressed in virus-infected inonocytes (C. Karp, X, Ma, and G. Trinchieri, unpublished results). Measles vinis inhibits IL-12 production by binding to its receptor, membrane cofactor protein or CD46, which is a natural receptor for complement factors C3b and C4b. Inhibition of IL-12 production in human inonocytes is observed on binding of CD46
172
CIORGIO TRINCHIERI
to its various ligands, i.e., measles virus, polymerized C3b, or anti-CD46 monoclonal antibodies (194). Thus, measles virus downregulates IL-12 production most likely by utilizing a physiologic mechanism of immunoregulation through complement and one of its receptors. The ability of measles virus to inhibit production of IL-12 is one of several mechanisms of induced immunosuppression (552). Measles virus replicates poorly in dendritic cells, but its replication is maximally induced when CD40 on the dendritic cells is stimulated; the infected dendritic cells are unable to stimulate T-ceI1 proliferation and are defective in their ability to produce IL-12 (553).However, a role for decreased IL-12 production was not demonstrated in other experimental systems in which immunosuppression in vitro was induced by interaction of measles virus glycoproteins with the surface of uninfected lymphocytes or by infection of resting human dendritic cells (554, 555). 3. Human Immunodeficiency Virus a. Defective Production of IL-12 by HIV-Infected Patients. As compared with PBMC from uninfected control donors, PBMC from HIVinfected individuals were found to produce very similar levels of TNF-a and IL-10, 3- to 4-fold more IL-6 (118), and 10- to 20-fold less IL-12 free p40 chain and 5-fold less biologically active p70 heterodimer when challenged in vitro with S. aureus (118,556-558). A similar defect in IL12 production was also reported by Gazzinelli et al. (559) in response to the opportunistic pathogen T. gondii. Although alveolar macrophages from HIV patients spontaneously produced low levels of IL-12, their ability to produce IL-12 in response to S. aureus was depressed (560). Lower accumulation of both p40 and particularly p35 mRNA paralleled the decreased ability of patients’ PBMC to produce IL-12 (558). Although these results need to be extended to other pathogens or stimuli able to induce IL-12 production, the specific deficiency in the production of IL-12, while other inflammatory cytokines are produced normally or at increased levels by HIV-infected patients, suggests a possible role for IL-12 deficiency in HIV disease pathogenesis. The mechanism underlying the decreased IL-12 production by PBMC from HIV-infected patients remains elusive. In vitro incubation of human monocytes with HIV for 1week resulted in decreased production of IL12 (118, 561). However, in both in vivo and in vitro HIV infection, only a small proportion of monocytes is actually infected with HIV, so that the suppression of IL-12 production is likely to be an indirect result of HIV infection. The HIV gp120 envelope protein was reported to directly stimulate production of low levels of IL-12 (562), but also to render monocytes defective in responding to S. aureus with high levels of IL-12 production,
INTERLEUKIK-12
173
possibly as a function of IL-10 overproduction and secondary inhibition of IL-12 production (563). HIV-infected patients, at least in some phases of disease progression, have been described to overproduce certain Th-2 type cytokines, in particular IL-4 and IL-10 (564-566). Because Th2 cytokines and in particular IL-10 are able to suppress IL-12 production (108,169),it could be hypothesized that exposure of PBMC in vivo or in vitro to these cytokines is responsible for the decreased IL-12 production. However, these cytokines have a suppressive effect on all inflammatory cytokines, making it difficult to explain the selective IL-12 deficiency (108, 169). Furthermore, when PBMC are exposed to IL-4 or IL-13 for an extended period, their ability to produce IL-12 is boosted rather than inhibited (169) and, in PBMC from HIV patients, the deficient IL-12 production is corrected (174). Similarly, IL-12 production in patients’ PBMC was partially restored by treatment with IFN-y or IL-15 (174, 567, 568). In short-term S. aureus-stimulated cultures of PBMC, production of IL-10 but not IL-4 is observed (118, 558). Chougnet et al. (558)reported that IL-10 production is increased in PBMC cultures from HIV patients compared to healthy controls, unlike earlier studies (118) in which no significant dfference was observed. Because endogenously produced IL10 in culture is known to limit IL-12 production (108), the ability of neutralizing anti-IL-10 to correct deficient IL-12 production from HIV patients was tested (118).In the presence of antibodies, a similar increase in IL-12 production was observed in both HIV patients and healthy controls, making it unlikely that IL-10 is responsible for the differential ability of patient and control PBMC preparations to produce IL-12 (118).A significant role for PGEz in the deficient production of IL-12 in blood cells of HIV patients has also been excluded (174,557), despite the known activity of PGEz as a potent and selective inhibitor of IL-12 production (181)and the reports of enhanced PGE2production in HIV-infected monocytes (569) and of increased CAMP levels in PBMC from patients (570). Another aspect of IL-12 production that might be involved in the IL12 deficiency in HIV-infected patients and that deserves investigation is the role of T-cell signaling in IL-12 induction. Several lines of evidence suggest that, at least during antigen-specific stimulation, T cells are needed for the induction of IL-12 production in antigen-presenting cells (148, 306, 359). Thus, the IL-12 deficiency in the patients may rest in part on deficient signaling by activated T cells to the antigen-presenting cells. Because of the role of IL-12 in T-cell activation, this mechanism of immunodeficiency would be self-amplifying. Although the in vitro T-cell response and the ability of PBMC to respond to stimulation with IL-12 production are deficient, analysis of expression
174
CXOHGIO TRINCHIERI
of cytokine mRNA in peripheral blood or lymph nodes of HIV patients revealed increased expression of T h l cytokines, including mRNA for IFNy. a cytokine dependent on IL-12 for optimal induction (571,572).Whereas these observations might argue against the hypothesis of a predominant Th2 response in HIV-infected patients, they are not incompatible with the deficiency of IL-12 production. Because of infections and other immunological stimulations, patients’ lymphocytes and monocyte/macrophages are thought to be in an activated state. In these conditions, a constitutive expression of mRNA for activated lymphocyte products, including IFN-y, and of proinflaminatory cytokines in macrophages and antigen-presenting cells, including IL-12, is to be expected. However, the in vivo observation of constitutive cytokine gene expression, reflecting a chronic inflammatory situation, may not be associated with an efficient acute response to stimulation or may even be responsible for a deficient production of the cytokines required for an effective immune response, e.g., IFN-y and IL-12. If this is the case, a progressive failure of antigen-specific responses over a background of chronic activation in both macrophages and lymphocytes might result which, because of homeostatic mechanisms regulating lymphocyte activation, may even lead to Th cell deletion.
b. HZV-Znfected Patients Are Responsive to ZL-12. Although HIVinfected patients are deficient in their ability to produce IL-12, their T and NK cells in vitro have been shown to respond normally to IL-12. Patients with advanced HIV disease often show a very reduced cytotoxic activity in their peripheral blood NK cells, although the number of NK cells is not decreased (573, 574). IL-12 treatment in vitro enhances the NK-mediated cytotoxic activity of peripheral blood lymphocytes from HIVinfected patients, similar to the observed effect on those of healthy donors; this NK-enhancing effect of IL-12 is particularly evident on lymphocytes from patients with advanced disease and nearly absent NK cytotoxic activity, in which IL-12 restores cytotoxic activity to levels close to those of healthy donors (117, 575, ,576). IL-12 enhances the NK cytotoxic activity of the patients’ lymphocytes against both tumor target cells and virusinfected target cells (117); interestingly, IL-12 can also boost the NK cytotoxic activity of healthy donors’ lymphocytes against HIV-infected target cells (316). On freshly purified peripheral blood lymphocytes from HIV-infected patients, IL-12 alone or in synergy with IL-2 induces IFNy production, although, in advanced patients, to levels somewhat lower than those observed with lymphocytes from healthy donors (117,577). IL12 also enhances the PHA-induced IFN-y production in lymphocytes from HIV-infected patients, almost completely correcting the low PHA-induced IFN-y production observed in some patients (568).
INTERLEUKIK-I2
175
The induction of IFN-7 production and the enhancement of N K cytotoxicity represent rapid and short-lived effects of IL- 12, which may not lend themselves easily to a permanent in vivo therapeutic effect. It was therefore important to deinonstrate whether IL-12 can prime T-cell clones from HIV-infected patients for high IFN-7 production, a long-lasting and possibly irreversible effect (15).Peripheral blood T cells froni 10 HIV-infected patients at different stages of disease were cloned by limiting dilution in the presence or absence of IL-12 for the first 2 weeks of culture (578). A very high efficiency of clonal expansion was obtained by culture in the presence of irradiated feeder cells, PHA, and IL-2. On average, CD4+ clones cultured in the presence of IL-12 produced lo-fold more and CD8+ clones 5-fold inore IFN-y than clones originated in the absence of IL-12 when restimulated by anti-CD3 and phorbol diester after 1 month expansion (578). This pi-iining effect, which is analogous to that observed with T cells from healthy donors (13,was observed with patients at any stage of the disease (578). IL-12 was also shown to enhance the depressed proliferation of HIV patients to recall antigens, including HIV peptide, influenza virus, Candida, tetanus toxoid, streptokinase, and alloantigens to levels close to those of healthy donors (119, 309, 568, 579). These results are consistent with the ability of IL- 12 to enhance the proliferative response to antigens, alloantigens, and initogens observed with T cells from healthy individuals (19, 281) and with the primary and, in some cases, obligatory role of IL12 in antigen-induced proliferation of ineinory T cells and differentiated Thl cells (119, 166, 294, 306). The in uitro-enhancing effect of IL-12 on HIV patient T-cell proliferation could be due to the activation of unresponsive T cells or to replacement of insufficient IL-12 produced in vitro by patients' antigen-presenting cells. However, only ininimal effects were observed in patients with less than 200 CD4' T cells/min3 (309, 579). Another important function of IL-12, shared with other Thl cytokines such as IL-2 and IFN-7, is its ability to prevent mitogen-, anti-CD3-, or CD95 (Fas)-mediatedprogramiiied cell death in T cells from HIV+ donors (311, 312). IL-12 also inhibits apoptosis induced by gp120 or CD4 crosslinking and CD3/TCR activation in a human Th1 clone (310), an in vitro mechanism of induction of apoptosis that may mimic one of the pathogenic processes in HIV infection. Because death by apoptosis is one of the mechanisms proposed for CD4' cell depletion in AIDS, which could be favored by reduced production of Th1 cytokines, including IL-12, the ability of IL-12 to prevent T-cell receptor-induced apoptosis in patients' T cells represents a potentially iinportant therapeutic function of this cytokine.
176
GIORGIO TRINCHIERI
c. IL-12 as an Adjuvant in HIV Vaccination. Several investigators have investigated the usefulness of IL-12 as an adjuvant in vaccination against HIV in experimental animals. In those studies, IL-12 was used as protein or, most often, as an expression plasmid, together with a peptide vaccine or with plasmids expressing the antigen (338, 432, 580-582). In all these studies, IL-12 proved particularly powerful in inducing specific CTL generation and DTH, whereas the effect on immunoglobulin production was less marked and usually limited to enhancement of IgG2a. However, the simultaneous use of IL-12 and GM-CSF was shown to have a cooperative effect in inducing maximum CTL and antibody generation (338,432,580). Of particular clinical interest is the finding that intranasal immunization of mice with a DNA vaccine of IL-12- and GM-CSF-expressing plasmids in liposomes induced strong mucosal and cell-mediated immune responses against HIV antigens (338).
d. Effect of IL-12 on HIV Replication. IL-12 does not enhance HIV replication in resting PBMC, but it has been reported to do so in mitogenactivated lymphocytes or in cultures depleted of CD8' T cells, especially those from infected asymptomatic donors (583-585). However, in another study (568),IL-12 was shown to have little affect by itself on HIV replication in pHA-activated human T cells and to significantly decrease HIV expression in ACH-2 cells in the presence of suboptimal concentrations of phorbol diester. Furthermore, IL-12 has been shown to inhibit IL-2-induced HIV replication in PHA-activated human T cells (568) and in CD8-depleted PBMC (584). IL-12 has been reported to be less efficient than IL-2 in inducing CD8' T-cell-mediated suppression of HIV-1 replication in CD4' T cells (586).However, in a different experimental system, IL-12 decreased HIV-1 replication in human inacrophages cocultured with autologous peripheral blood mononuclear cells (587);in this latter system, the IL-12 antiviral effect was mediated by IFN-y produced by NK cells, making it possible that IL-2 and IL-12 exert an antiviral effect acting through different mechanisms and effector cell types, CD8' T cells for IL-2 and primarily NK cells or possibly CD4' T cells for IL-12. 4. Murine Retroviruses
Murine AIDS (MAIDS) is a syndrome of progressive lymphoproliferation and increasingly severe immunodeficiency that develops in mice of certain strains, e.g., C57BU6, following infection with a retrovirus mixture containing a replication-defective pathogenic murine leukemia virus (MuLV) designated BM5def and a helper MuLV (588).IL-12 and IFNy are produced in the first week of infection of C57BW6 mice, and IFNhas an important, although not essential, role in inducing proliferation
INTERLEUKIN-12
177
of the MuLV-infected B cells (589). The IFN-y mRNA level was maintained or continued to increase at a later time of infection, but production of IFN-y protein declined with the progression of the immunodeficient state and of T-cell anergy (589). At 4 weeks of infection, IFN-y could still be induced in T cells by LPS, IL-12, and/or anti-CD28 antibodies, but at later times the CD4+ T-cell anergy became irreversible (589, 590). Although IL-12 and IFN-y appear to have a pathogenetic role in the lymphoproliferation and immunodeficiency of MAIDS, treatment of the animals with 100-250 ng of IL-12 per mouse, 5 times a week, starting at the time of infection or up to 9 weeks after infection, markedly inhibited the development of splenoinegaly and lymphadenopathy, B-cell activation, Ig secretion, and T-cell immunodeficiency (591). The expression of the pathogenic BM5def was almost completely suppressed, whereas the expression of the helper MuLV was reduced to one-third (591). This effect of IL-12 is dependent on IFN-y production and is likely due to the inhibitory effect of high doses of IFN-y on viral replication in B cells; indeed, the use of lower concentrations of IL-12 resulted in exacerbation of the disease, probably because lower levels of IFN-y enhance the proliferation of B cells, whereas higher doses resulted in systemic toxicity, similar to what is observed in LCMV infection (588). Rauscher leukemia virus infection in mice is also very irnmunosuppressive, causing a block in dendritic cells but not T-cell functions (592). Treatment with five daily doses of 100 ng of IL-12 per mouse at the time of exposure resulted in an improveinent in the capacity of lymph node dendritic cells to stimulate allogeneic responses and in the restoration of DTH (592).Whether these data reflect a dxect effect of IL-12 on dendritic cells or an indirect effect mediated by IFN-y or other cytokines produced by T and NK cells remains to be determined.
B. BACTERIA 1. Listeria Monocytogenes
After the role of IL-12 as a mediator of IFN-y production in bacterial infection was demonstrated in human peripheral blood mononuclear cells (46), L. nzonocytogenes (257, 593) and 7'. gondii (256, 377) infections were the first two experimental murine models in which such a role was confirmed first in vitro and then in tjivo. Heat-killed L. rrwnocytogenes (HKLM) in vitro induces production of IFN-y from SCID splenocytes via a mechanism that involves IL-12 production, and IL-12 synergizes with TNF-a, and IL-2 in inducing IFN-y production (257). In vim, L. monocytogenes infection induces mRNA and protein production of both IL-12 and IFN-.)I begnning around 24 hr from infection, and anti-IL-12
178
GIOHCIO TRINCHIERI
antibodies efficiently block IFN-y production and resistance to infection in both SCID and wild-type aniinals (593, 594). A single dose of IL-12 at the time of infection significantly enhanced resistance of mice to L. rnoriocytogenes infection (595). Although NK cells appear to be the major producers of IFN-y in the primary response to L. monocytogenes, their ability to produce IFN-y is defective in mice lacking Ty6 cells, perhaps due to deficient TNF-a, production (596); production of IFN-y by Ty6 cells was shown to be induced by IL-12 in synergy with IL-1(597).Although the primary response to L. wwnoqtogenes required IL-12 in uivo for IFNy production, resistance to secondary challenge was blocked by anti-IFNy antibodies, but not by anti-IL-12 antibodies (598), indicating that, as previously shown in T. gondii infection (377), an established Th1 response is independent of IL-12 for IFN-y production. However, in vitro production of IFN-y by spleen cells from immune mice in response to HKLM, but not in response to anti-CD3 antibodies, was at least partially dependent on IL-12 (598). The requirement for IFN-7 in the resistance to listeriosis was clearly shown by the rapid death of Listeria-infected IFN-yR-deficient mice (599). IL- 10-deficient mice, conversely, have an increased resistance to listeriosis, with a dramatically enhanced proinflainmatory cytokine production and Thl response that was protective and did not result, as in the case of 7'. gondii or T. cruxi infection (292,293), in systemic toxicity (600). Interestingly, IL-13 treatment of mice resulted in an enhanced resistance to listeriosis, most likely by acting indirectly through stimulation of IL-12 production (177). Whereas in tiitro heat-killed L. rnonocytogenes was a potent inducer of IL-12 production, only the live bacteria were efficient in vivo (131, 601). However, HKLM, soluble listerial antigen preparations, or a synthetic peptide corresponding to a dominant MHC class II-restricted listerial determinant, which by themselves are inefficient vaccines, when coinjected with IL-12, elicited a potent antigen-specific iinmune response that conferred protective iininunity against L. monocytogenes (602-604). 2. Mycobacteriu Mycobucteriurn tuberculosis induces IL-12 production in both human and murine phagocytic and dendritic cells (46, 605, 606). Administration of IL-12 to inice enhances their resistance to M . tuberculosis infections, especially in the susceptible BALB/c strain (606,607),whereas the requirement for IL-12 in the resistance to the infection was shown by neutralization of endogenous IL-12 with monoclonal antibodies (606) or, more convincingly, in IL-12 p40 genetically deficient inice (608). The ability of M. tuberculosis to induce IL-12 production was in part dependent on phagocytosis because phagocytosis of large latex beads induced similar
INTERLEUKIN-I2
179
production of IL-12, but not, however, TNF-a (134, 609). Furthermore, arabinofuranosyl-terminated lipoarabinornannan derived from rapidly growing Mycohacteriuni sp., but not the extensively mannosylated form, is capable of inducing IL-12 (610). IL-12 used as an adjuvant for an experimental subunit vaccine based on secreted antigens from M. tuberculosis was able to accelerate the development of an efficient immune response, but not to change the final outcome of a full vaccination regime (611). In humans, IL-12 was produced by pleural fluid cells of patients with tuberculosis pleuritis, and anti-IL-12 antibodies suppressed the proliferation of these cells in response to M . tuberculosis (612). Using in situ hybridization, it was found that the percentage of bronchoalveolar lavage cells expressing IL-12 p40 mRNA was much higher in patients with active compared to inactive tuberculosis (613) and by Elispot, the number of IL12-producing cells in the peripheral blood was found to be increased in patients with tuberculosis compared to healthy donors (614). In patients infected with M. leprae, it was observed that the tuberculoid lesions, characterized by a CD4' type 1 response, express 10 times more IL-12 mRNA and protein coinpared to leproniatous lesions, characterized by a CD8' type 2 response (FilFj).Anti-IL-12 antibodies blocked M. lepraeinduced T-cell proliferation in tuberculoid patients and IL-12 induced proliferation of CD4+ type 1 T cell clones from tuberculoid patients but not in CD8' type 2 T-cell clones from lepromatous patients (615). However, IL-12, in synergy with IL-2, restores both IFN-.)I production and proliferation in response to M . leprae in T cells from nonresponder patients almost to the level of responder patients (616). In M. aviuna infection of mice, endogenous IL-12 is required for resistance, and neutralization of IL-12 resulted in a several hundredfold increase in bacterial load (617),whereas treatment with recombinant IL-12 induced protection of susceptible BALB/c mice, which have a decreased IL- 12 production in response to infection (618,619). In a family in which several members have disseminated M. avium complex infection, it was observed that adherent cells from patients and their unaffected mothers produced abnorrnally low levels of IL-12 following stimulation with S. aureus, although they produced normal levels in response to S. aureus plus IFN-.)I (620). Thus, members of this fainily have a defect in IL-12 production that closely resembles that observed in HIV-infected patients (118) who also are susceptible to M . nviurn complex infection.
3. Other Bacterial Species Salmonella dublin infection of inice or of inacrophages in vitro induces the production of IL-12 (621-623), and the important role of IL-12 in the
180
GIORGIO TRINCHIERI
resistance to the infection is indicated by the increased salmonellosis and reduced survival time in mice orally challenged with S. dublin and treated with anti-IL-12 antibodies (624). Similarly, in S. thyphimurium infection, neutralization of IL-12 not only decreased the ability of mice to resist the infection, but also prevented the immunosuppression that accompanies the acute stages of the disease (625) showing a dual role of IL-12 in enhancing resistance and inducing immunosuppression. The role of IL-12 in murine Lyme borreliosis is complex; anti-IL-12treated mice infected with Borrelia burgdorferi had an increased number of spirocheta, but a significant decrease in peak arthritis severity, accompanied by a reduction in Thl response (626).Thus, the IL-12-induced innate immunity and T h l response is effective in preventing spirocheta growth, but it is also involved in the generation of the arthritis. Vitamin A deficiency is associated with higher IL-12 and IFN-.)I production and exacerbates murine Lyme arthritis (627). In humans, dendritic cells isolated from the dermis or from peripheral blood phagocytose B. burgdorferi and produce IL-12, thus probably playing a role in the initial T h l response to the bacteria, whereas Langerhans cells from the epidermis are unable to do so (628). Endogenous IL-12 is required for resistance to Brucella abortus infection in mice: neutralization of IL-12 at the time of infection prevents the generation of a protective T h l response, and the effects of this treatment in terms of exacerbation of the infection and inhibition of splenomegaly and granuloma formation in the liver are still evident 6 weeks after the treatment (629, 630). The ability of B. abortus to efficiently induce IL-12 production makes it a potential vaccine candidate (631). IL-12 is produced in vivo by animals immunized with live or killed Bordetella pertussis; however, an acellular vaccine constituted of various bacterial components adsorbed to alum induced a Th2 response and was unable to induce protective immunity (632). The same level of protection obtained with the IL-12-inducing whole cell vaccine can, however, be obtained by adding recombinant IL-12 to the acellular vaccine (632). The requirement for endogenous IL-12 in resistance to bacterial infections has also been shown with several other species, including Yersenia enterocolitica (633),Chlamydia tracomatis (634),Helicobacterpylori (635), and group B streptococci (636). Exogenous IL-12 was shown to be protective in mice for infections with Klebsiella pneumonia (637), group A and B streptococci (636, 638). IL-12 was also shown to be effective as an adjuvant in a vaccine against Y. enterocolitica (639). In humans, macrophages from patients with Whipple’s disease, a systemic infection in which the causative bacteria, Tropheryma tohippelii, accumulate within macrophages, have been found to be severely defective
INTERLEUKIN-12
181
in IL-12 production in response to various stimuli (640). The production of other cytokmes, with the exception of IFN-y, was within the normal values (640). Because a similar, although less severe, deficiency in IL-12 production was observed in two relatives of the patients, it is possible that a genetic defect in IL-12 production is responsible for susceptibility to Wliipple’s disease (640).
PARASITES C. PROTOZOAN 1. Leishmania Species
Treatment of the susceptible strain BALB/c with IL-12 during the first week following cutaneous infection with Leishmania nujor induces resistance to the infection with a shift from a Th2 to a T h l response (641, 642). The cured animals are resistant to a subsequent challenge (641), but if the IL-12 treatment is delayed more than 1 week, it is ineffective in curing the animals (642). Vaccination of BALB/c with soluble leishmania antigens (383) or the recombinant leishmania LACK antigen (643, 644), together with IL-12, also induces a protective T h l immunization. In addition to IL-12, anti-IL-4 antibodies or DNA immunization with the LACK gene in a bacterial plasmid able to induce IL-12 also generates protective immunity (644,645).Although IL-12 alone cannot cure BALB/c mice when administered 1 week after L. major infection, the combined treatment of BALB/c mice 3 weeks after infection with antimony-based leishmanicidal drugs and IL-12 or IFN-y can effectively cure them, shifting the response from Th2 to T h l (646, 647); the effectiveness of the therapy based on IFN-y is dependent on endogenous IL-12 and its effect is prevented by anti-IL-12 antibodies (647). The switch from a Th2 to a T h l response most likely does not involve a phenotypic change in already differentiated Th2 cells, but the IL-12-induced generation of T h l cells from yet uncommitted T cells or de n o m thymus emigrants (648). The role of endogenous IL-12 early in L. major infection has been an element of controversy. Reiner et al. (649) could not demonstrate IL-12 mRNA until 7 day after infection and proposed that amastigotes, which do not mature in vivo until about a week from infection, and not the infective metacyclic promastigotes, are able to induce IL-12 production. Indeed, proinastigotes have been shown to be able to efficiently suppress IL-12 production induced by various stimuli (159,650),although promastigotes, especially the immature procyclic forms, have been shown to have some ability to induce IL-12 in vitro in human monocytes (159). However, other studies have shown that IL-12 is already produced at 24 hr from infection in the lymph nodes of infected animals (or in peritoneal macrophages, followingip injection of the parasite) (320,651,652).The important
182
GIORGIO TRINCHIERI
role of this early IL-12 production was clearly proven by the observation that anti-IL-12 antibodies injected at the time of infection blocked IFNy production and NK cell activation in the popliteal lymph node 2 days after infection (320). The essential role of endogenous IL-12 in resistance to L. major infection has been confirmed by the inability of both IL-12 p40 and p35 genetically deficient mice to resist the infection (652, 653). Mice lacking the expression of CD40L are also unable to produce IL-12 and to resist infection, in part because of lack of IL-12 production at a late time of infection (654, 655). The early production of IL-12 is comparable between resistant (C3H) and susceptible strains (BALB/c), although a delay in IL-12 production was observed in the resistant C57BU6 strain (320). However, whereas T cells from C3H mice rapidly upregulated both the IL-12RP1 and the p2 chain within 1 or 2 days of infection, this upregulation was minimal in BALB/c mice (D. Jones, M. M. Elloso, L. Showe, D. Williams, G. Trinchieri, and P. Scott, unpublished results), possibly due to the downregulatory effect of IL-4 on IL-12 responsiveness (330, 656). In experimental visceral leishmaniasis induced by L. donooani infection, IL-12 treatment both before infection or 2 weeks after challenge improves resistance: in this model, however, unlike in L. major infection in BALB/c mice, the disease is characterized not by a Th2 response, but by an ineffective Thl response (657),which was boosted by IL-12 treatment. Vaccination with heat-killed L. major in BALB/c mice has been used to obtain a cross-reactive Th2 response to L. donovnni: the animals so vaccinated were unable to resist a challenge with L. donooani, but treatment with IL-12 successfully induced antileishmanial activity (658) by inducing a switch from a Th2 to a Thl response. Although these effects of IL-12 are mostly mediated by induction of a Thl response and production of IFN-y in IFN-y gene-disrupted mice, IL-12 still enhances resistance to L. donovani by a mechanism that involves production of TNF-a and activation of iNOS (659). The essential role of IL-12 both early and late in the resistance to L. donooani infection was shown by the ability of antiIL-12 antibodies to exacerbate infection when administered the first week of infection or from the second to the fourth weeks (660). In humans, IL-12 expression was detected in most lesions of individuals with cutaneous leishmaniasis and correlated strongly with the level of IFNy expression (661). The addition of IL-12 to culture of lymphocytes of patients with active visceral leishmaniasis restored their ability to proliferate and produce IFN-y in response to leishmanial antigens (662, 663). A gene from L. braziliensis was cloned and termed LeIF because of its homology with the eukaryotic ribosomal protein eIF4A: LeIF is a potent inducer of
INTEHLEUKIN-12
183
IL-12 production and possibly one of the important antigens recognized by the host immune response to Leishrnnniu infection (141).
2. Toroplusmu gonclii IL-12 is required for the T-cell-independent induction of IFN-y in SCID mice in response to T. gondii infection (256).The optimal production of IFN-y by NK cells requires not only IL-12, but also TNF-a, IL-1, and B7/CD28 interaction (256, 258, 259, 664). The expression of CD28 on NK cells is probably upregulated by IL-15 (259). IL-10 and TGF-/3 are negative regulators of IL-12 production and responsiveness (263,377,664, 665). IL-10 genetically deficient mice rapidly die of T. gonclii infection, although they control parasite growth more efficiently than wild-type mice (377, 66.5): the reason for death is an uncontrolled production of proinflainmatory cytokines, including IL-12 and IFN-y, and a pathology reminiscent of a systemic inflammatory response syndrome (377). However, the toxic response is in part dependent on T lymphocytes, and IL-10 genetically deficient SCID mice survive T. gondii infection longer than wild-type SCID mice (665). IL-12 is produced within the first few days of toxoplasma infection and its production is unimpaired in IFN-y-deficient mice (164). In these latter animals, neutralization of IL-12 blocks the cytotoxic NK cell response, but does not decrease survival, suggesting that IL-12-induced IFN-y is necessary for the control of parasite growth (164). In normal mice, treatment with either anti-IL-12 or anti-IFN-y antibodies at the time of infection blocks the induction of resistance to T. gondii and all animals die within 2 weeks (377).However, once a chronic infection controlled by a Thl response is established at 4 weeks, treatment with anti-IFN-y induces the death of the animals, but the treatment with anti-IL-12 antibodies is ineffective, indicating that whereas IFN-y is still required for the antiparasite inacrophage activity, the maintenance of the established T h l response is IL-12 independent (377). The requirement for IL-12 in the resistance to T. gondiii infection has been confirmed using IL-12 p40 genetically deficient mice (666). 3. Tnjpnnosomn cruzi T. cmai, a hemoflagellate protozoan parasite that is the causative agent of human Chagas’ disease, is a potent inducer of IL-12 production in macrophages (667, 668). Only live, UV-, or gamma-irradiated trypomastigote forms are able to induce IL-12, but not the heat-killed parasites or lysates, or the epimastigote forms (667, 668). A glycosylphosphatidylinositol-anchored mucin-like glycoprotein isolated from T. cruzi has been shown to be able to initiate in macrophages the synthesis of proinflammatory
184
GIORGIO TRINCHIERI
cytokines, including IL-12 (140). Endogenous IL-12 is required for both innate and acquired immunity to T. cmzi infection, and its effect requires the participation of TNF-a and IFN-y (669,670).Administration of recombinant IL-12 enhances the resistance of the mice to T. cmzi (670).Endogenous IL-10 downregulates IL-12 production and limits the resistance to the infection; however, although IL-10-deficient SCID mice survive a T. cruzi infection longer than wild-type SCID mice, IL-10-deficient C57BL/ 6 mice died earlier of infection than wild-type C57BLJ6 mice (293, 669). The IL-10-deficient mice had a lower parasite burden, but much higher serum levels of IL-12, TNF-a, and IFN-y, and mortality was prevented by neutralizing anti-IL-12 antibodies, suggesting that there is a critical requirement for IL-10 to prevent the development of a systemic immune inflammatory response associated with activation of CD4' T cells and overproduction of IL-12 (293). In humans, IL-12 has been shown to potentiate both proliferative response and cytotoxicityin response to T. cmzi antigen stimulation in PBMC from patients with the different forms of Chagas' disease (671, 672). 4. Plasmodium Species
Because of the high worldwide incidence of malaria and the emergence of plasmodium strains resistant to many antimalarial drugs, there is much interest in new methods of prevention that could destroy the plasmodium during the intrahepatic cycle. Intraperitoneal injection of a single dose of 150 ng of IL-12 2 days before challenge of mice with Plasmodium yoelii protected 100% of mice against hepatic malaria via a mechanism that required IFN-y and iNOS (673). Similarly, a single subcutaneous injection of 10 pg/kg of recombinant human IL-12 in seven rhesus monkeys 2 days before challenge with P. cynomolgi sporozoites induced an increase in plasma levels of IFN-y and protected the monkeys against malaria (674). IL-12 treatment is effective not only against the hepatic stage of malaria infection, but it also induces protection against blood-stage P. chubaudi AS with a mechanism that requires IFN-y, TNF-a, and iNOS activation (675). PATHOGENS D. FUNGAL 1. Candida albicans The outcome of systemic challenge of mice with the fungus C. albicans is determined in part by immunological events occurring shortly after infection and leading, in resistant strains of mice challenged with live vaccine strains of the yeast, to a protective T h l response, whereas the mice challenged with a virulent strain have an exacerbative Th2 response (676). IL-12 is readily induced in mice infected with the vaccine strains and its level, rather than the production of IFN-y, correlatedwith induction
INTERLEUKIN-12
185
of T h l response (677).Treatment with anti-IL-12 antibodies of mice undergoing a healing infection with a vaccine strain ablates the development of acquired anticandidal resistance and switches the response from T h l to Th2 (678). Surprisingly, however, IL-12 treatment does not improve, but rather worsens the resistance to virulent C. albicans infection in mice and high doses of IL-12 have a toxic effect, possibly by inducing a fungal-type systemic inflammatory syndrome, on mice infected with the vaccine strains (676, 678). The failure of IL-12 to protect mice from C. albicans infection is, at least in part, explained by the particular physiological role played by neutrophils in C. albicans infection (679). Neutrophils have a major role in providing a first line of defense against Candida. In vitro neutrophils in response to C. albicans produced both IL-12 and IL-10: IL-12 was predominantly produced in response to the vaccine strains of C. albicans, whereas IL-10 was produced in response to the virulent strains (680). Similarly, in vivo IL-12 production from neutrophils was found to be associated with healing infection, whereas IL-10 production with progressive infection, indicating that neutrophils have not only an effector role in C. albicans infection, but also an immunomodulatory one, regulating T h l and Th2 differentiation (679, 680). IL-12 treatment of mice with progressive infection increased the production of IL-10 from neutrophils, and it was therefore unable to prevent the exacerbating Th2 response; however, if mice were made neutropenic, IL-12 treatment induced IFN-.)Iproduction and the generation of a T h l response that resulted in protection of the animals against the C. albicans infection (680).
2. Crzjptococcus neoformans High-dose IL-12 treatment, combined or not with the antifungal agent fluconazole,of mice infected intravenouslywith C. n e o f o m n s dramatically reduced the level of infection in the brain and liver, but had no effect on spleens or lungs (681). However, in a model of intratracheal infection, IL-12 induced a strong response against the pulmonary infection and completely prevented the dissemination to the brain (682).The protective effect of IL-12 was associated with an increase in Thl-type cytokines and iNOS in the lungs of infected mice (683) and was almost abolished by neutralization of TNF-a (684). Because C. neoformans is an opportunistic pathogen that causes serious life-threatening disease in both healthy and immunocompromised persons, the strong activity of IL-12 against this fungal pathogen and its cooperativitywith antifungal drugs have an interesting therapeutic potential. 3. Histoplasma capsulaturn H . cap.sulatum is another fungal pathogen that is emerging as a serious opportunistic infection in immunocompromised hosts. In mice, IL-12 has
186
GIORGIO TRINCHIERI
been shown to lower the mortality rate of both norinal and SCID mice infected with H . capsulatum through the induction of IFN-.)/ (685-687). Inhibition of endogenous IL-12 with anti-IL-12 antibodies induced accelerated mortality, an effect that is, however, reversed if the animals were treated with anti-IL-4 (685, 687). In reinfection histoplasmosis, treatment with anti-IL-12 antibodies did not alter survival (687). Thus, endogenous IL-12 is necessary for the establishment of a protective T h l response to H . capsulatum and treatment with exogenous IL-12 enhances the resistance to the infection. 4 . Coccidioides immitis
Coccidioidomycosis is a mycotic disease endemic to the southwestern United States and Central and South America that can cause severe and fatal pulmonary and disseminated infections. In mice, the DBM2 strain is resistant to the infection with a prevalent T h l response, whereas BALBk mice are susceptible with a prevalent Th2 response. IL-12 treatment around the time of infection protects BALBk inice by inducing a Th2 to Thl switch, whereas anti-IL-12 treatment prevents the ability of DBA/2 mice to resist the infection, inducing a Th2 response (688).
E. HELMINTHIC PARASITES I. Schistosoma mansoni Injection of S. munsoni eggs intravenously results in the formation of pulmonary granulomas that are initially characterized by a ThO/Thl type of response, but that rapidly switch to the expression of prevalently Th2
type cytokines (689). Anti-IL-12 or anti-IFN-y antibody treatment of the animals enhances the Th2 response and the granuloma formation, whereas treatment with IL-12 prevents the switch to a Th2-type response and decreases the granulomas (689). In IFN-y genetically deficient mice, IL12 treatment exacerbates the Th2-dependent pathology by failing to suppress the production of Th2 cytokines and boosting IgE levels, while enhancing lymphocyte proliferation (409). However, in B-cell-deficient mice, higher levels of IL-12 are produced with maintenance of a T h l response, but, even in the absence of a Th2 response, the size and the number of egg granulomas in the liver of infected animals are unchanged (450). Not only can IL-12 treatment prevent the formation of lung granulomas, but vaccination with eggs in combination with IL-12 commits the mice toward a T h l response, such that they develop only minimal granulomas and lung fibrosis on subsequent egg challenge (689, 690). In a prophylactic vaccine model, IL-12 has been shown to enhance the protective effect of immunization with either a soluble lung-stage larval antigen preparation or irradiated cercariae (691, 692). It is of interest that
INTEHLEUKIN-12
187
inultiple vaccinations with irradiated cercariae without IL-12 resulted in a Th2-dominant response, whereas mice vaccinated in the presence of IL12, in addition to a strongly polarized Thl response, also showed a significant increase in parasite-specific IgG antibodies that were able to protect naive recipients in transfer experiinents (692).
2. Filarial H e h i n t s The filarial helinint Bnrgia nzalayi induces Th2 response both in mice (693) and in hurnans (694). In mice, IL-12 treatment in vivo or in vitro induces a Thl response to B. inalayi infection and, even if administered after a Th2 response has already been established, profoundly inhibits the production of Th2 cytokines (693). However, the elimination of blood bone microfilariae was not altered by IL-12 treatment (693). 3. Intestinal Nematodes
Resistance and expulsion of intestinal nematodes are primarily mediated by a Th2-type response, and particularly by the production of IL-4 (695). Treatment with IL-12 before or during the infection has been shown to downregulate Th2 responses while promoting Th1 responses and to eliininate or decrease the protective immune response to Nyppostrongylus brasiliensis (167,696), Stroiigyloides stereoralis (697),and Trichuris muris (698). In general, IL-12 administered during the initiation of an iinrnune response to nematodes can, through the induction of IFN-y, change the predominant response froin Th2 to Thl, whereas IL-12 treatment has less effect once the production of Th2 cytokines has become established (167). These effects of IL-12 are consistent with the hypothesis that Th2associated responses protect against and Thl responses exacerbate nematode infections (167, 695).
XVI. Antitumor Effects of 11-12 A. TOXICITY OF SYSTEMIC IL-12 TREATMENT In order to obtain an efficient antituinor effect of IL-12 in v i m , high doses of recombinant IL-12 need to be injected for an extended period of time. The most commonly used protocol in inice involved five daily injections of IL-12, inost usually ip, followed by a 2-day rest: although this protocol was initiated for empirical reasons, it proved necessary to avoid the formation of pulinonary edema, a toxicity observed when mice are treated without interruption even with low concentrations of IL-12 (P. Bouchard, personal communication). The half-life of injected recombinant IL-12 is approximately 3.5 hr in inice (161), 18 hr in rhesus inonkeys (38), and between 5 and 10 hr in hurnans (699). Most mouse strains tolerate
188
GIORGIO TRINCHIERI
repeated injections of up to 1 pglmouselday, but some mouse strains, e.g., C3H and A/J, show lethal toxicity at such doses and need to be treated at IL-12 doses 5-10 lower in tumor treatment experiments (456). The primary toxicities observed in normal mice treated with IL-12 were hematological alterations, hepatotoxicity,and skeletal muscle degeneration. Daily administration of IL-12 for 7 days led to severe anemia with red blood cells dropping to about half of the normal values (245, 700). Both lymphopenia and neutropenia, possibly due to emargination of leukocytes onto vascular endothelium and liver, were also observed (700).Splenomegaly was an early and constant finding, largely caused by extramedullar hematopoiesis involving the erythroid, myeloid, and megakaryotic lineages (245, 700). The bone marrow, however, was hypoplastic, with a loss of mature neutrophils and precursor cells of all lineages (243, 244). The anemia in IL-12-treated mice appears too early to be uniquely due to decreased generation of erythrocytes, and it has been hypothesized that increased tissue erythrophagocytosis, as suggested by the marked activation of Kupffer cells in the liver, may play a role (700). The hematological toxicities in IL-12-treated mice are mediated by IFN-y: in IFN-yRdeficient mice, spleen cellularity was less increased, there were fewer infiltrating NK cells, but a strong extramedullary hematopoiesis was still induced, showing that, in the absence of IFN-7, IL-12 promotes hematopoiesis, consistent with its in vitro activities (245). Significant elevation in transaminases and mildly increased liver weights were observed in IL-12-treated mice (701). Intense macrophage infiltrates as well as of NK and CD8' T cells were localized around central veins and terminal portal vessels, associated with occasional necrosis of isolated hepatocytes observed after 4- 10 days of treatment, followed by progression with continued administration of IL-12 to areas of coagulative necrosis with marked elevation of serum transaminases (700, 701). In IFN-yR genetically deficient mice the activation of macrophages was still observed, often associated with an increased number of eosinophils (701). Skeletal muscle toxicity was seen in mice treated with doses of IL-12 of more than 1 p g and consisted in visibly white muscle at necropsy, muscle necrosis and calcification, and elevation of serum muscle enzymes beginning after about 5 days of treatment (700). Normal mice given high doses of IL-12 displayed ascites and pleural effusion, but pulmonary edema was observed only in IFN-yR-deficient mice or in mice treated with uninterrupted continuous treatment for more than 10 days (700, 701). IL-12 damage to the intestinal tract was particularly evident from its ability to sensitize it to radiation: in mice treated with IL-12 doses as low as 40 ng for 3 or 4 days and then given 1200 cGy radiation, the lumen of small intestine was distended with fluid ingest and displayed severe mucosal damage with
INTERLEUKIN-12
189
marked shortening of villi or villi fusion and loss of epithelial lining cells (246). This sensitization of the intestinal tract to ionizing radiation, which resulted in the death of' the animals within 4 to 6 days after irradiation, was prevented by neutralization of IFN-y (246). A heinatological toxicity similar to that described in the mouse was also observed in nonhuman primates treated with IL-12 (39, 702); however, although signs of extramedullary hernatopoiesis were evident in these animals, the bone marrow was characterized by hypercellularity with increase of all three lineages rather than by the hypocellularity observed in mice (39). Squirrel monkeys (Sciureus suimiri) receiving daily subcutaneous injections of human rIL-12 in doses ranging from 0.1 to 50 pg/kg/day for 14 days showed dose-related fever, mild to moderate anemia, leukocytosis, hypoproteinemia, hypoalbuininernia, hypophosphatemia, hypocalcemia, generalized lymph node enlargement, splenomegaly, and thymic cortical atrophy (39).Two of the six high-dose animals developed pulmonary edema and ascites; neither hepatoxicity, besides evidence of Kupffer cell hypertrophy and hyperplasia, nor muscle degeneration was observed in IL-12treated squirrel monkeys (39). Phase I clinical trials were initiated in 1994 in cancer patients and identified a maximum tolerated dose of 500 ng/kg with reversible side effects consisting of fever, mild anemia, neutropenia, thrombocytopenia, fatigue, and myalgia. Oral stomatitis of unknown origin and elevation of transaminases were observed at the 1000-ng/kg dose (699). Biological effects included elevations in IFN-7 in the serum that peaked in the first 3 to 4 days, decreasing thereafter despite continuing dosing (699). Phase I1 trials were initiated in 17 patients with daily 500-ng/kg iv injections, but it was interrupted after a few injections because of profound neurastenia requiring admission for 9 patients and associated with two treatmentrelated deaths with multiorgan toxicity, including intestinal bleeding (455, 699). This difference in toxicity between the phase I and the phase I1 trials was attributed to the fact that in phase I a single predose was given and then the daily treatment was initiated after 2 weeks, whereas in the phase I1 trials daily injections were immediately started (455, 699). Indeed, in mice, it was shown that a single predose 1 week before initiation of daily treatment effectively protected the animals from IL-12 toxicity, allowing treatment with higher doses (J. Leonard, personal communication; 455, 456).
B. ANTITUMOREFFECTS OF IL-12 I N EXPEHIMENTAL ANIMALS IL-12 was shown to have a powerful antitumor and antimetastatic activity against many murine tumors (703-707). With several tumors, IL-12 sys-
190
C;IORGIO TRINCHIERI
temic treatment was effective even when initiated up to 2 to 7 weeks after tumor transfer (703, 705). Often, IL-12 induced only a temporary tumor regression and tumor growth resumed when the treatment was interrupted; in some cases, e.g., following peritumord IL-2 inoculation of the RenCd tumor or systemic treatment of the CSAlM fibrosarcoma, complete regression was observed, and the animals that rejected the tumor were specifically resistant to challenge with the same tumor type (703, 705). The depletion of CD8' T cells or both CD8' and CD4' T cells was most often required to suppress the antitumor effect of IL-12, whereas depletion of CD4' T cells or NK cells had only a marginal effect (703, 704). In all tumor models analyzed, the antitumor effect of IL-12 was at least in part mediated by IFN-y (704,705,708,709).However, IFN-y production is clearly not the only antitumor mechanism mediated by IL-12 because the antitumor effect of IL-12 is reduced in nude mice compared to euthymic mice, but a -10fold higher level of IFN-y was induced in nude mice compared to euthymic mice (708). Endogenous IL-12 is induced in response to tumor cell transplantation (710, 711), and it was demonstrated to be necessary in vivo for rejection of spontaneously regressing P815 tumor variants (711). IL-12 treatment also delayed tumor appearance and decreased tumor incidence induced by the carcinogen 3-methylcholanthrene (712); interestingly, in contrast to the characteristically round, hard, well-circumscribed, and protruding tumors normally induced by the carcinogen, most tumors induced in IL-12-treated mice were atypical with flat, soft, and invasive characteristics (712). The ability of IL-12-induced IFN-y to affect tumor growth may be due to direct effects on the tumor or indirect ones on the host cells. Many of the tumors affected by IL-12 in vim are sensitivein vitro to the antiproliferative effect of IFN-7, suggesting a direct effect on the tumor cells (708). In addition, in tumor cells and/or in host cells IFN-y induces activation of iNOS and production of nitric oxide, which inhibits tumor growth (713715). Another mechanism by which IL-12 induced-IFN-y can prevent tumor growth is its ability to inhibit angiogenesis in vivo (716, 717). The antiangiogenic effect of IL-12 is not direct, but mediated, through IFNy, by the production of the chemokine interferon-inducible protein- 10 ( IP-lo), which, in addition to being a chemoattractant for lymphocytes, has a powerful inhibitory effect on proliferation and hfferentiation of endothelial cells (718). Because IL-12 can effectively inhibit angiogenesis induced in vivo by basic fibroblast growth factor mixed in a gel pellet (718, 719), the antiangiogenic effect of IL-12 on tumors couId be attributed to induction of IP-10 from host cells; however, production of IP-10 was also observed in tumor tissues from IL-12 treated mice (709). By rendering tumor cells unresponsive to IFN-y by overexpression of a dominant nega-
INTERLEUKIN-15
191
tive truncated IFN-yR construct (720),it was demonstrated that the ability of tumors cells to respond to IFN-y was necessary for both the antitumor effect of IL-12 and its ability to inhibit angiogenesis induced by the tumor cells (C. Couglilan, W. Lee, and G. Trinchieri, unpublished results). Thus, the ability of IFN-y to induce the production of antiangiogenic factors in the tumor cells themselves appears to play a major role in the antitumor effect of IL-12. In order to itnprove the local delivery of IL-12 to the tumor sites, several viral vectors have been prepared. Because IL-12, unlike other cytokines, is encoded by two separated genes, bicistronic vectors (721) or vectors encoding fusion proteins (61, 62) have been utilized. Different types of viral vectors, including retroviral vectors (721-723), adenovirus vectors (724, 725), and vaccinia virus vectors (726, 727), have been prepared for expression of IL-12, and overall promising antitumor activity has been obtained by injecting them peri- or intratuinorally. Interestingly, the use of adenoviral vectors producing the IL-12 p40 homodimers inhibited the antitumor effect of injection of a bicistronic vector encoding IL-12 heterodiiners in a inurine bladder carcinoma model (723). In addition to the use of viral vectors, injection of cDNA encodmg IL-12 at distant sites from the tumor injection site or gene gun-mediated transfection with IL-12 genes on the skin overlaying intradermal tumors were shown to result in regression of established primary and metastatic tumors (728, 729). IL-12-transfected fibroblasts were also used to deliver IL-12 locally in the proximity of tumors with good antitumor activity against different tumors (730, 731). The regression of the murine BL-6 melanoma was IL12 dose dependent, with better regression obtained by injecting the highest proportion of IL-12-producing fibroblasts, whereas the most efficient antitumor immunity to subsequent challenges was observed at intermediate doses of IL-12 (730). The first use of transfected tumor cells producing IL-12 was reported for the C-26 murine colon carcinoma cells (722); these transfected cells produced only 30-80 pg/inl of IL-12 and were able to induce tumor rejection, which was mediated by CD8+T cells, only when inhibitory CD4' T cells were eliminated by antibody treatment (722).However, C-26 tumor cells engineered to express much higher levels of IL-12 induced a significant infiltration of CD8' T cells and NK cells and induced an efficient tumor rejection, even when CD4' cells were not depleted (732). Effective eradication of established murine tumors and induction of an efficient antitumor immunity were obtained using other IL-12-producing tumor cells of different origin (721, 733). The toxicity of IL-12 treatment alone makes particularly attractive the possibility to combine its use with other antitumor agents with which IL-12
192
GIORGIO TRINCHIERI
may synergize for antitumor activity without increased toxicity. Combined treatment of Lewis lung tumors with IL-12 and fractionated radiation therapy was shown to result in a synergistic antitumor response (734). Administration of IL-12 with pulse IL-2 was reported to induce rapid and complete eradication of murine renal cell carcinoma with a greater effect than with each cytokine separately and with a tolerable toxicity (735). Furthermore, systemic or local IL-2 was shown to enhance the antitumor effect of IL-12 gene therapy (736) and IL-12 to potentiate the curative effect of a vaccine based on IL-2-transduced tumor cells (737). A strong synergistic antitumor effect against poorly immunogenic tumors and induction of antitumor immunity was observed by utilizing systemic IL-12 therapy and induction of expression on the tumor cells by gene therapy of the costimulatory molecule B7.1 (738, 739). A strong synergistic antitumor effect on contralateral wild-type tumor growth was observed when tumor cells transfected with IL-12 and with the gene encoding IGIFAL-18 were coinjected in mice (C. Coughlan, W. Lee, and G. Trinchieri, unpublished results). The combined use of IL-12 and IL-18-transfected tumor cells induced a potent antiangiogenic effect that was mediated by IFN--y and required IFN-7 responsiveness in the tumor cells. An effective therapeutical vaccination against Meth-A tumor, able to induce regression of established tumors, was obtained by immunizing mice with a mutated p53 peptide and IL-12 in the presence of the QS2l adjuvant (335);interestingly, a good CTL response and antitumor effect were observed only with low doses of IL-12, whereas the high doses normally utilized for obtaining an in vivo antitumor effect were immunosuppressive (335).IL-12 has also been shown to enhance the specific immunotherapy of cancer mediated by dendritic cells pulsed with a mutated p53 peptide (336)or a class I-restricted tumor peptide derived from the P815AB antigen of P815 tumor cells (446). Interestingly, injected dendritic cells pulsed with the P815 peptide induced T-cell anergy in v i m , but IL-12 could prevent or revert the anergic state (447). Because in vitro treatment of the dendritic cells with IL-12 rendered them effective in inducing an antitumor response, these data raise the possibility of a direct effect of IL-12 on dendritic cells (446, 447). Tumor cells transfected with IL-2 or IL-12 were compared for their ability to induce vaccination against established tumors: although both cell types induced similar CTL generation, the IL-12-transfected ones were more efficient in inducing tumor regression because of their ability to induce complement-hng antibodies and systemic activation of T h l rather than Th2 cells (740).
c. ANTITUMOR EFFECTSOF IL-12 O N HUMAN TUMORS The majority of data on the antitumor effects of IL-12 in humans are obviously based on in vitro experiments. After the original report by Lieber-
INTERLEUKIN-12
193
inan et al. (741)that IL-12 increases NK and antibody-dependent cytotoxicity of human PBL against tumor cells, many papers have reported an activity of IL-12 on the NK or CTL activity of cancer patients PBL or tumorinfiltrating lymphocytes against autologous or allogeneic tumors (742-750). IL-12 was shown in vivo to enhance the antitumor activity of human NK cells or PBL coinjectedwith human tumor cells into SCID mice (751,752). Malignant CD4' T cells from patients with Sezary syndrome, a forin of cutaneous T-cell lymphomas that produce prevalently Th2-type cytokines and PBMC from patients, were reported to produce lower levels of IL12 than those from healthy donors (753, 754). IL-12 treatment of the malignant lymphocytes in vitro resulted in a significant decrease in the production of IL-4 (753).Because of the poor prognosis and high degree of fatality associated with Sezary syndrome, on the basis of these results, trials with IL-12 have been initiated (A. Rook, personal communication). It is too early to evaluate the results of the clinical trials of IL-12 in cancer patients; however, a stable partial response in a patient with renal cell carcinoma, a transient complete remission in a melanoma patient, and four patients with stable disease have been reported in the first phase I trial (699). XVII. Concluding Remarks
Although IL-12 has been discovered only relatively recently, a remarkable level of knowledge has been reached for this cytokine, and many possible therapeutic applications are being tested or have been proposed. The biological interest for this cytokine stems from its dual function as a proinflammatory cytokme and as an iminunomodulatory molecule. Being produced at the early times of inflammation in response to infection or other stimuli, it contributes via the production of IFN-y and other cytokines to the inflammatory process itself and particularly to the activation of macrophages, while setting the stage for the ensuing antigen-specific adaptive immune response by directing CD4' and CD8' T cells to differentiate into Thl or TC1 effector and memory cells. Thus, IL-12 represents a bridge between innate resistance and adaptive immunity, with a central role in the regulation of the response to infection and tumor cells, as well as in autoimmunity and allergy. Because of these activities, its clinical use has been proposed in infections, both as an immunopotentiating agent or as an adjuvant in vaccination, against tumors, and for the prevention of severe allergic syndromes. Conversely, the use of IL-12 antagonists has been proposed in autoimmunity and in systemic inflammatory responses. The early reports of toxicities in the initial clinical trials, although not unexpected and directly mediated by the physiological functions of this molecule, have dampened some of the early enthusiasm and forced a more
194
GlORGlO TRINCHIERI
thoughtful approach to its planned use. Undoubtedly, the study of IL-12 has taught us much about the physiology of innate and adaptive immunity: whether we will be able to harness its potent biological activities for efficient therapeutical applications is now the challenge before us.
ACKNOWLEDGMENTS The author thanks Drs. Ellen PurC, Louise Showe, and Jihed Chehimi for critical review of the manuscript, Mrs. Marion Sacks for typing, and Ms. Marina Hoffman for editing. The author was supported in part by NIH Grants CA 10815,CA 20833, CA 32898, and A1 34412.
REFERENCES 1 . Trinchieri, G., Kubin, M., Bellone, G., and Cassatella, M. C. (1993). Cytokine crosstalk between phagocytic cells and lymphocytes: Relevance for differentiatiodactivation of phagocytic cells and regulation of adaptive immunity.J. Cell. Biochem. 53,301-308. 2. Damle, N. K., Klussman, K., Linsley, P. S., and Aruffo, A. (1992).Differential costimdatory effects of adhesion molecules B7, ICAM-1, LFA-3, and VCAM-1 on resting and antigen primed CD4+ T lymphocytes.J. Immunol. 148,1985-1992. 3. Nathan, C. F., Murray, H. W., Wiebe, M. E., and Rubin, B. Y. (1983). Identification of interferon-y as the lymphokine that activates human macrophage oxidative metabolism and antimicrobial activity. J . Exp. Med. 158, 670-689. 4. Fiorentino, D. F., Bond, M. W., and Mosmann, T. R. (1989). Two types of mouse helper T cell. IV. Th2 clones secrete a factor that inhibits cytokine production by Thl clones. J. E x p Med. 170, 2081-2095. 5. Bogdan, C., Vodovotz, Y., and Nathan, C. (1991). Macrophage deactivation by interleukin l0.J. Exp. Med. 174, 1549-1555. 6. de Lisle, D. (1774). Trait6 du Vice CancCreux. Couturier Fils, Paris, France. 7. Coley,W. B. (1893).Treatment of malignant tumors by repeated inoculation of erysipelas, with a report of 10 cases. Am. J. Med. Sci. 105, 487-564. 8. Stames, C. 0. (1991). Coley’s toxins in perspective. Nature 357, 11-12. 9. Carswell, E. A,, Old, L. J., Kassel, R. L., Green, S., Fiore, N., and Williamson, B. (1975). An endotoxin-induced serum factor that causes necrosis of tumors. Proc. Nutl. Acad. Sci. USA 72, 3666-3668. 10. Parish, C. R., and Liew, F. Y. (1972). Immune response to chemically modified flagellin. 3. Enhanced cell-mediated immunity during high and low zone antibody tolerance to flagellin. J. Exp. Med. 135, 298-311. 1 1 . Mosmann, T. R., and Coffinan, R. L. (1989). TH1 and TH2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu. Reu. Immunol. 7, 145-173. 12. Seder, R. A,, and Paul, W. E. (1994). Acquisition of lymphokine-producing phenotype by CD4+ T cells. Annu. Rev. lnimunol. 12, 635-673. 13. Erard, F., Wild, M., Garcia-Sanz, J. A,, and LeGros, G. (1993). Switch of CD8 T cells to noncytolytic CD8 cells that make TI12 cytokines and help B cells. Science 260, 1802-1803. 14. Seder, R. A,, Boulay, J. L., Finkelman, F., Barbier, S., Ben Sasson, S. Z., Le Gros, G., and Paul, W. E. (1992). CD8+ T cells can be primed in vitro to produce IL-4. J. Immunol. 148, 1652-1656. 15. Manetti, R., Gerosa, F., Giudizi, M. G., Biagiotti, R., Parronchi, P., Piccinni, M., Sampognaro, S., Maggi, E., Romagnani, S., and Trinchieri, G. (1994). Interleukin-I2
INTERLEUKIN-12
19s
induces stable priming for interferon-y (IFN-y) production during differentiation of human T helper (Th) cells and transient IFN-y production in established Th2 cells clones. J . Exp. Med. 179, 1273-1283. 16. Magi, E., Giudizi, M. G . , Biagiotti, R., Annunziato, F., Manetti, R., Piccinni, M., Parronchi, P., Sampognaro, S., Giannarini, L., Zuccati, G., and Romagnani, S. (1994). Th2-like CD8+ T cells showing B cell helper function and reduced cytolytic activity in human inimunodeficiency virus type 1 infection. J . Exp. Med. 180, 489-495. 17. Ferrick, D.A,, Schrenzel, M. D., Mulvania, T., Hsieh, B., Ferlin, W. G., and Lepper, H. (1995). Differential production of interferon-gamma and interleukin-4 in response to Thl- and TIi2-stimulating pathogens by gamma delta T cells in vivo. Nature 373, 255-257. 18. Warren, H. S., Kinnear, B. F., Phillips, J. H., and Lanier, L. L. (1995). Production of IL-5 by human NK cells and regulation of' 1L-5 secretion by IL-4, IL-10, and IL12.1. In~viunol.154, 5144-5152. 19. Kobayashi, M., Fitz, L., Ryan, M., Hewick, R. M., Clark, S. C., Chan, S., Loudon, R., Sherman, F., Perussia, B., and Trinchieri, G. (1989). Identification and purification of Natural Killer cell stiniulatory factor (NKSF), a cytokine with multiple biologic effects on human lymphocytes. J , ESP. Med. 170, 827-846. 20. Stern, A. S., Podlaski, F. J., Hulmes, J. D., Pan, Y. E., Quinn, P. M., Wolitzky, A. G., Familletti, P. C., Stremlo, D. L., Truitt, T., Chizzonite, R., and Gately, M. K. (1990). Purification to homogeneity and partial characterization of cytotoxic lymphocyte maturation factor from hinnaii B-lymphohlastoidcells. Proc. Natl. Acad. Sci. USA 87,68086812. 21. Wolf, S. F., Temple, P. A,, Kobayashi, M., Young, D., Dicig, M,, Lowe, L., Dzialo, R., Fitz, L., Ferenz. C., Hewick, R. M.. Kelleher, K., Herrmann, S. H., Clark, S. C., Azzoni, L., Chan, S. H., Trinchieri, C., and Perussia, B. (1991). Cloning of cDNA for natural killer cell stimulatory factor, a heterorliineric cytokine with multiple biologic . 3074-3081. effects on T and natural killer cells. J . h n u t ~ o l 146, 22. Gubler, U., Chua, A. O., Schoenhant, D. S., Dwyer, C. M., McCoinas, W., Motyka, R., Nabavi, N., Wolitzky, A. C., Quinn, P. M., Familletti, P. C., and Gately, M. K. (1991). Coexpression of two distinct genes is required to generate secreted bioactive cytotoxic lymphocyte maturation factor. Proc. Nntl. Acrid. Sci. USA 88, 4143-4147. 23. Van de Griend, R. J., Krimpen, B. A., Ranfeltap, C. P. M., and Bolhuis, R. H. (1984). Rapidly expanded activated human killer clones have strong antitumor cell activity and have the surface phenotype of either T, non-T, or null cells.]. Imnzuuol. 132,31853191. 24. London, L., Penissia, B., and Trinchieri, G. (1985). Induction of proliferation in uitro of resting human natural killer cells: Expression of surface activation antigens. J. 1mnruru)l. 134, 718-727. 25. Perussia, B., Ramoni, C., Anegbn, I., Cnturi, M. C., Faust, J., and Trinchieri, G. (1987). Preferential proliferation of natural killer cells anlong peripheral blood mononuclear cells cocultured with B lymphoblastoitl crll lines. Nut. Irnmnurt. Cell Growth Re&. 6, 171-188. 26. Reeni, G. H., Cook, L. A,, Henriksen, D. M., and Vilcek. J. (1982). Gainina interferon induction in human thymocytes activated by lectins and B cell lines. Infect. Inrmrcn. 37, 216-221. 27. Murphy, M., Loudon, R., Kobayashi, M., and Trincliieri, C . (1986). Gamma interferon and lyniphotoxin, released by activated T cells, synergize to inhibit granulocytenionocyte colony formation. J . Esp. Med. 164, 263-279.
196
GIORGIO TRINCHIERI
28. Sieburth, D., Jabs, E. W., Warrington, J. A,, Li, X., Lasota, J., LaForgia, S., Kelleher, K., Huebner, K., Wasmuth, J. J., and Wolf, S. F. (1992).Assignment of genes encoding a unique cytokine (1L-2) composed of two unrelated subunits to chromosomes 3 and 5. Genomics 14, 59-62. 29. Noben-Trauth, N., Schweitzer, P. A., Johnson, K. R., Wolf, S. F., Knowles, B. B., and Shultz, L. D. (1996).The interleukin-12 beta subunit (p40) maps to mouse chromosome 11. Mamm. Genome. 7,392. 30. Tone, Y., Thompson, S. A,, Babik, J. M., Nolan, K. F., Tone, M., Raven, C., and Waldman, H. (1996). Structure and chromosomal location of the mouse interleukin12 p35 and p40 subunit genes. Eur. J , lmmunol. 26, 1222-1227. 31. Yoshimoto, T., Kojima, K., Funakoshi, T., Endo, Y., Fujita, T., and Nariuchi, H. (1996). Molecular cloning and characterization of murine IL-12 genes. J. lmmunol. 156, 1082-1088. 32. Ma, X . , Chow, J. M., Gri, G., Carra, G., Gerosa, F., Wolf, S. F., Dzialo, R., and Trinchieri, G. (1996). The interleukin-12 p40 gene promoter is primed by interferony in inonocytic cells. ]. Exp. Med. 183, 147-157. 33. Podlaski, F. J., Nanduri, V. B., Hulmes, J. D., Pan, Y.-C. E., Levin, W., Danho, W., Chizzonite, R., Gately, M. K., and Stem, A. S. (1992). Molecular characterization of interleukin 12. Arch. Biochem. Biophys. 294, 230-237. 34. Tangarone, B. S., Vath, J. E., Nickbarg, E. B., Yu, W., Harris, A. S., and Scoble, H. A. (1996). The disulfide bond structure of recombinant human interleukin-12. In “Techniques in Protein Chemistry VII: Abstracts of the 9th Symposium of the Protein Society” (D. R. Marshak, ed.), p. 150. Academic Press, Boston. 35. Schoenhaut, D. S., Chua, A. O., Wolitzky,A. G., Quinn, P. M., D y e r , C. M., McComas, W., Familletti, P. C., Gately, M. K., and Gubler, U. (1992). Cloning and expression of murine IL-12. J. lmmunol. 148, 3433-3440. 36. ViUinger, F., Brar, S. S., Mayne, A,, Chikkala, N ., and Ansari, A. A. (1995).Comparative sequence analysis of cytokine genes from human and nonhuman primates.]. Immunol. 155,3946-3954. 37. Fan, X., Sibalic, V., Niederer, E., and Wuthrich, R. P. (1996). The proinflammatory cytokine interleukin-12 occurs as a cell membrane-bound form on macrophages. Biochem. Biophys. Res. Commun. 225, 1063-1067. 38. Nadeau, R. R., Ostrowski, C., Ni-Wu, G., and Liberato, D. J. (1995).Pharmacokinetics and pharrnacodyiamics of recombinant human IL-12 in male rhesus monkeys. /. Pharmacol. Exp. Therapeut. 274, 78-83. 39. Sarmiento, U . M., Riley, J. H., Knaack, P. A., Lipman, J. M., Becker, J. M., Gately, M. K., Chizzonite, R., and Anderson, T. D. (1994). Biologic effects of recombinant human interleukin-12 in squirrel monkeys (Sciureussaimiri).Lob. lnuest. 71,862-873. 40. Zuo, X. J., Jordan, S. C., Wilkinson, A,, Danovitch, G. M., Barba, L., Schwieger, J., and Nast, C. C. (1995). Interleukin-12 mRNA levels in renal allograft fine-needle aspirates do not correlate with acute transplant rejection. Tmnsphtation 60, 13601362. 41. Mathieson, P. W., and Gillespie, K. M. (1996). Cloning of a partial cDNA for rat interleukin-12 (IL-12) and analysis of IL-12 expression in vivo. Scnnd. J. Immunol. 44, 11-14. 42. Bush, K., Day, N. K., Kraus, L. A,, Good, R. A., and Bradley, W. G. (1994). Molecular cloning of feline interleukin 12 p35 reveals the conservation of leucine-zipper motifs present in human and murine IL-12 p35. Mol. Immunol. 31, 1373-1374. 43. Schijns, V. E., Wierda, C. M., Vahlenkamp, T. W., and Horzinek, M. C. (1997). Molecular cloning of cat interleukin-12. lmmunogenetics 45, 462-463.
INTERLEUKIN-12
197
44. Zarlenga, D. S., Canals, A., Aschenbrenner, R. A., and Gasbarre, L. C. (1995). Enzy-
matic amplification and molecular cloning of cDNA encoding the small and large subunits of bovine interleukin 12. Biochim. Biophys. Acta 1270, 215-217. 45. Keefe, R. G., Choi, Y., Femck, D. A., and Stott, J. L. (1997).Bovine cytokine expression during different phases of bovine leukemia virus infection. Vet. Im~nunol.Immunopathol. 56, 39-51. 46. D'Andrea, A,, Rengaraju, M., Valiante, N. M., Chehiini, J., Kubin, M., Aste-Amezaga, M., Chan, S. H., Kobayashi, M., Young, D., Nickbarg, E., Chizzonite, R., Wolf, S. F., and Trinchieri, G. (1992). Production of natural killer cell stimulatory Factor (NKSF/ IL-12) by peripheral blood mononuclear cells. j . Exp. Med. 176, 1387-1398. 47. Mattner, F., Fischer, S., Guckes, S., Jin, S., Kaulen, H., Schmitt, E., Rude, E., and Germann, T. (1993).The interleukin-12 subunit p40 specifically inhibits effects of the interleukin-12 heterodimer. Eur. J. Immunol. 23,2202-2208. 48. Ling, P., Gately, M. K., Gubler, U., Stern, A. A,, Lin, P., Hollfelder, K., Su, C., Pan, Y. C., and Hakimi. J. (1995).Huinan IL-12 p40 homodirnerbinds to the IL-12 receptor but does not mediate biologic activity. J. lminunol. 154, 116-127. 49. Nickbarg, E. B., Vath, J. E., Pittinan, D. D., Leonard, J. E., Walburger, K. E., and Bond, M. D. (1995). Structural characterization of the recombinant P40 heavy chain subunit monomer and homodimer and murine IL-12. Bioorgan. Chem. 23,380-396. 50. Heinzel, F. P., Hujer, A. M., Ahmed, F. N., and Rerko, R. M. (1997).In vivo production and function of IL-12 p40 hoinoclmers. j . Immunol. 158, 4381-4388. 51. Gately, M. K., Carvajal, D. M., Connaughton, S. E., Gillessen, S., Warner, R. R., Kolinsky, K. D., Wdkinson, V. L., D y e r , C. M., Higgins, G. F., Jr., Podlaski, F. J., Faherty, D. A,, Familletti, P. C., Stem, A. S., and Presky, D. H. (1996). Interleukin12 antagonist activity of mouse interleukin-12 p40 homodiiner in vitro and in vivo. Ann. N.Y. Acad. Sci. 795, 1-12. 52. Gerinann, T., Rude, E., Mattner, F., and Gately, M. K. (1995).The IL-12 p40 homodimer as a specific antagonist of the IL-12 heterodimer. Immunol. Today 16,500-501. 53. D y e r , D. S. (1996). Molecular model of interleukin 12 that highlights amino acid sequence homologies with adhesion domains and gastrointestinal peptides. j . Mol. Graphics 14, 148-157. 54. Merberg, D. M., Wolf, S. F., and Clark, S. C. (1992). Sequence similarity between NKSF and the IL-G/G-CSF family. Immunol. Toduy 13, 77-78. 55. Gearing, D. P., and Cosman, D. (1991). Homology of the p40 subunit of natural killer cell stimulatory factor (NKSF) with the extracellular domain of the interleukin-6 receptor. Cell. 66, 9-10, 56. Taga, T., and Kishimoto,T. (1993). Cytokine receptors and signal transduction. FASEB J. 7, 3387-3396. 57. Taga, T., Hibi, M., Hirata, Y., Yainasaki, K., Yasukawa, K., Matsuda, T., Hirano, T., and Kishimoto, T. (1989). Interleukin-6 triggers the association of its receptor with a possible signal transducer gp130. Cell 58, 573-581. (58. Davis, S., Aldrich, T. H., Stahl, N., Pan, L., Taga, T., Kishimoto, T., Ip, N . Y., and Yancopoulos, G. D. (1993). LIFR beta and gp130 as heterodimerizing signal transducers of the tripartite CNTF receptor. Science 260, 1805-1805. 59. T a p , T., and Kishimoto, T. (1997). Gp130 and the interleukin-6 family of cytokines. Annu. Rev. Imnwnol. 15, 797-819. 60. Neddermann, P., Graziani, R., Ciliberto, G. and Paonessa, G. (1996). Functional expression of soluble human interleukin-11 (IL-11) receptor alpha and stoichiometry of in vitro IL-11 receptor complexes with gp130. J. Biol. Chem. 271, 30986-30991.
198
GIORGIO THINCHIERI
61. Anderson, R., Macdonald, I., Corbett, T., Hacking, G., Lowdell, M. W., and Prentice, H. G. (1997). Construction and biological of an interleukin-12 fusion protein (Flexi12): Delivery to acute rnyeloid Ieukemic blasts using adeno-associated virus. Hum. Gene. Ther. 8, 1125-1135. 62. Lieschke, G. J., Rao, P. K., Gately, M. K., and Mulligan, R. C. (1997).Bioactive murine and human interleukin-12 fusion proteins which retain antitumor activity in vivo. Nuture Biotech. 15, 35-40. 63. Devergne, O., Hummel, M., Koeppen, H., LeBeau, M. M., Nathanson, E. C., Kieff, E., and Birkenbach, M. (1996). A novel interleukin-12 p40-related protein induced by latent Epstein-Barr virus infection in B lymphocytes.J. Virol. 70, 1143-1153. 64. Devergne, O., Birkenbach, M., and Kieff, E. (1997). Epstein-Barr virus-induced gene 3 and the p35 subunit of interleukin 12 form a novel heterodimeric hematopoietin. Proc. Nntl. Acud. Sci. USA 94, 12041-12046. 65. Desai, B. B., Quinn, P. M., Wolitzky, A. G., Mongini, P. K. A., Chizzonite, R., and Gately, M. K. (1992). The IL-12 receptor. 11. Distribution and regulation of receptor expression. I. Immunol. 148, 3125-3132. 66. Naume, B., Gately, M. K., Desai, B. B., Sundan, A., and Espevik, T. (1993). Synergistic effects of interleukin 4 and interleukin 12 on NK cell proliferation. Cytokine 5,38-46. 67. Chizzonite, R., Truitt, T., Desai, B. B., Nunes, P., Podlaski, F. J., Stem, A. S., and Gately, M. K. (1992). IL-12 receptor. I. Characterization of the receptor on PHAactivated human lymphoblasts.1. Immunol. 148, 3117-3124. 68. Chua, A. O., Chizzonite, R., Desai, B. B., Truitt, T. P., Nunes, P., Minetti, L. J., Warner, R. R., Presky, D. H., Levine, S. F., Gately, M. K., and Gubler, U. (1994). Expression cloning of a human IL-12 receptor component: A new member of the cytokine receptor superfainily with strong homology to gp130. J . Immunol. 153, 128-136. 69. Chua, A. O., Wilkinson, V. L., Presky, D. H., and Gubler, U. (1995). Cloning and characterization of a mouse IL-12 receptor-0 component.]. Immunol. 155,4286-4294. 70. Showe, L. C., Wysocka, M., Wang, B., Lineman-Williams, D., Peritt, D., Showe, M. K., andTnnchieri, G. (1996).Structure ofthe mouse IL-12R01 chain and regulation of its expression in BCGLPS treated mice. Ann. N.Y. Acud. Sci. 795, 413-415. 71. Wu, C. Y., Warner, R. R., Carvajal, D. M., Chua, A. O., Minetti, L. J., Chizzonite, R., Mongini, P. K., Stem, A. S., Gubler, U., Presky, D. H., and Gately, M. K. (1996). Biological function and distribution of human interleukin- 12 receptor beta chain. Eur. J. Immunol. 26,345-350. 72. Wu, C., Ferrante, J., Gately, M. K., and Margram, J. (1997). Characterization of IL12 receptor pl chain (IL-lBR@)-deficientmice. J. Immunol. 159, 1658-1665. 73. 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 beta-type cytokine receptor subunits. Proc. Nntl. Acud. Sci. USA 93, 1400214007. 74. Wu, C., Warner, R. R., Wang, X., Presky, D. H., and Gately, M. K. (1997). Regulation of interleukin-12 receptor beta 1chain expression and interleukin-12 binding by human peripheral blood mononuclear cells. E w .1.Immunol. 27, 147-154. 75. Gollob, J. A., Kawasaki, H., and Ritz, J. (1997).Interferon-y and interleukin-4 regulate T cell interleukin-12 responsivenessthrough the differential modulation of high-affinity interleukin-12 receptor expression. Eur. 1.Immunol. 27, 647-652. 76. Benjamin, D., Sharma, V., Kubin, M., Klein, J. L., Sartori, A., Holiday, J., and Trinchieri, G. (1996).IL-12 expression in AIDS-related lymphoma B cell lines.]. Immunol. 156, 1626- 1637.
INTERLEUKIN-12
199
77. Vogel, L. A., Showe, L. C., Lester, T. L., McNutt, R. M., Van Cleave, V. H., and Metzger, D. W. (1996). Direct binding of IL-12 to human and murine 13 lymphocytes. Int. Iinrriunol. 8, 1955-1962. 78. Chan, S. H., Perussia, B., Gupta, J. W., Kobayashi, M., Pospisil, M., Young, H. A., Wolf, S. F., Young, D., Clark, S. C., and Trinchieri, G. (1991). Induction of IFN-y production by N K cell stimulatory factor (NKSF): Characterization of‘ the responder cells and synergy with other inducers. J . Exp. Med. 173, 869-879. 79. Manetti, R., Parronchi, P., Giudizi, M. G., Piccinni, M.-P., Ma@, E., Trinchieri, G., and Romagnani, S. (1993). Natural killer cell stimulatory factor (NKSFIIL-12) induces Thl-type specific immune responses and inhibits the development of IL-4 producing Th cells. J. Erp. Med. 177, 1199-1204. 80. Hsieh, C., Macatonia, S. E., Tripp. C. S., Wolf; S. F., OGarra, A,, and Murphy, K. M. (1993). Listerbindwed Thl development in &TCR transgenic CD4+ cells occurs through macrophage production of IL-12. Science 260, 547-549. 81. Germann, T., Gately, M. K., Schoenhaut, D. S., Lohoff, M., Mattner, F., Fischer, S., Jin, S., Schmitt, E., and Rude, E. (1993). Interleukin-l2/T cell stimulating factor, a cytokine with multiple effects 011 T helper type 1 (Thl) but not on Th2 cells. Eur. 1. Imiiiunol. 23, 1762-1770. 82. Szabo, S. J., Jacobson, N. G., Dighe, A. S., Gubler, U., and Murphy, K. M. (1995). Developmental commitment to the The lineage by extinction of IL-12 signaling. r?nnttlnity 2, 665-675. 83. Szabo, S. J., Dighe, A. S., Gubler, U., and Murphy, K. M. (1997). Regulation of the interleukin (IL)-12R /32 subunit expression in developing T helper (Thl) and TI12 cells. J . Exp. Med. 185, 817-824. 84. Guler, M. L., Jacobson, N. G., Guhler, U., and Murphy, K. M. (1997). T cell genetic background determines maintenance of IL- 12 signaling.J , Immunol. 159,1767- 1774. 85. Hsieh, C. S., Macatonia, S. E., O’Garra, A,, and Murphy, K. M. (1995). T cell genetic background determines default T helper phenotype development in vitro. /. Erp. Med. 181, 713-721. 86. Rogge, L., Barberis-Maino, L., Biffi, 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. 87. Pignata, C., Prasad, K. V. S., Hallek, M., Druker, B., Rudd, C. E., Robertson, M. J., and Ritz, J. (1995). Phosphorylation of src family lck tyrosine kinase following IL-12 activation of human N K cells. Cell. Imnunol. 165, 211-216. 88. Pignata, C., Prasad, K. V. S., Robertson, M. J., Levine, H., Rudd. C. E., and Ritz, J. (1993). FcyRIIIA-mediated signaling involves src-family lck in human natural killer celIs. J. I,nnzunoE. 151, 6794-6800. 89. Gerosa, F., Tommasi, M., Renati, C., Gandini, G., Libonati, M., Tridente, G., Carra, G., and Trinchieri, G. (1993). Differential effects of tyrosine kinase inhibition in CD69 antigen expression and lyhc activity induced by rIL-2, rIL-12 and rIFN-a! in human NK cells. Cell. linnzz~iiol.150, 382-390. 90. Pignata, C.. Sangherd, J. S., Cossette, L., Pelech, S., and Ritz, J. (1994).Interleukin12 induces tyrosine phosphorylation and activation of 44-kD mitogen-activated protein kinase in human T cells. Blood 83, 184-190. 91. Bacon, C. M., McVicar, D. W., Ortaldo, J. R., Rees, R. C., O’Shea, J. J., and Johnston, J. A. (1995). Interleukin 12 (IL-12) induces tyrosine phosphorylation of JAK2 and TYK2: Differential use of janus family tyrosine kinases by IL-2 and IL-12. /. E x p Med. 181, 399-404.
200
GIORGIO TRINCHIERI
92. Yu, C., Lin, J., Fink, D. W., Akira, S., Bloom, E. T., and Yamauchi, A. (1996). Differential utilization of Janus kinase-signal transducer and activator of transcription signaling pathways in the stimulation of human natural killer cells by IL-2, IL-12, and IFN-a. J. lmmunol. 157, 126-137. 93. Zou, J., Presky, D. H., Wu, C. Y., and Gubler, U. (1997). Differential associations between the cytoplasmic regions of the interleukin-12 receptor subunits beta 1 and beta 2 and JAK kinases. J. Biol. Chem. 272, 6073-6077. 94. Jacobson, N. G., Szabo, S. J., Weber-Nordt, R. M., Zhong, Z., Schreiber, R. D., Damell, J. E., Jr., and Murphy, K. M. (1995). Interleukin 12 signaling in T helper type 1 (Thl) cells involves tyrosine phosphorylation of signal transducer and activator of transcription (Stat)3 and Stat4. J. Exp. Med. 181, 1755-1762. 95. Bacon, C. M., Petricoin, E. F., 111, Ortaldo, J. R., Rees, R. C., Lamer, A. C., Johnston, J. A,, and O’Shea, J. J. (1995). IL-12 induces tyrosine phosphorylation and activation of STAT4 in human lymphocytes. Proc. Natl. Acad. Sci. USA 92, 7307-7311. 96. 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. 97. Bacon, C. M., Cho, S. S., and O’Shea, J. S. (1996). Signal transduction by interleukin12 and interleukin-2: A comparison and contrast. Ann. N.Y. Acad. Sci. 795, 41-59. 98. Yamamoto, K., Quelle, F. W., Thierfelder, W. E., Kreider, B. L., Gilbert, D. S., Jenkins, N. A., Copeland, N. G., Silvennoinen, O., and Ihle, J. N. (1994). Stat4, a novel gamma interferon activation site-binding protein expressed in early myeloid differentiation. Mol. Cell. Biol. 14, 4342-4349. 99. Murphy, K. M., Murphy, T. L., Szabo, S. J., Jacobson, N. G., Guler, M. L., Gorham, J. D., and Gubler, U. (1997). Regulation of IL-12 receptor expression in early Thelper responses implies two phases of Thl differentiation: Capacitance and development. Chem. Immunol. 68, 54-59. 100. Thierfelder, W. E., van Deursen, J. M., Yamamoto, K., Tripp, R. A., Sarawar, S. R., Carson, R. T., Sangster, M. Y., Vignali, D. A. A., Doherty, P. C., Grosveld, G. C., and Ihle, J. N. (1996). Requirement for Stat 4 in interleukin-12-mediated responses of natural killer and T cells. Nature 382, 171-174. 101. Kaplan, M. H., Sun, Y., Hoey, T., and Grusby, M. J. (1996). Impaired IL-12 responses and enhanced development of Th2 cells in Stat4-deficient mice. Nature 382,174-177. 102. Xu, X., Sun, Y. L., and Hoey, T. (1996). Cooperative DNA binding and sequenceselective recognition conferred by the STAT amino-terminal domain. Science 273, 794-797. 103. Cassatella, M. A., Meda, L., Gasperini, S., D’Andrea, A., Ma, X., and Trinchieri, G. (1995). Interleukin-12 production by human polymorphonuclear leukocytes. Eur. J. lmmunol. 25, 1-5. 104. Bette, M., Jin, S. C., Germann, T., Schafer, M. K., Weihe, E., Rude, E., and Fleischer, B. (1994). Differential expression of mRNA encoding interleukin-12 p35 and p40 subunits in situ. Eur. J. Immunol. 24, 2435-2440. 105. Chizzonite, R., Truitt, T., Podlaski, F. J., Wolitzky, A. G., Quinn, P. M., Nunes, P., Stern, A. S., and Gately, M. K. (1991). IL-12: Monoclonal antibodies specific for the 40-kDa subunit block receptor binding and biologic activity on activated human lymphoblasts.J. Immunol. 147, 1548-1556. 106. Wilkinson, V. L., Warner, R. R., Truitt, T. P., Nunes, P., Gately, M. K., and Presky, D. H. (1996). Characterization of anti-mouse IL-12 monoclonal antibodies and measurement of mouse IL-12 by ELISA. J. lmmunol. Methods 189, 15-24.
INTERLEUKIN-12
20 1
107. Gately, M. K., a i d Chizzonite, R. (1992). Measurement of human and mouse interleukin 12. In “Current Protocols in Immunology” (J. E. Coligan, A. M. Kruisbeek, D. H. Margulies, E. M. Shevach and W. Strober eds.),Vol. 1,pp. 1-8. Wiley, New York. 108. D’Andrea, A., Aste-Amezaga, M., Valiante, N. M., Ma, X., Kubin, M., and Trinchieri, G. (1993).Interleukin-10 inhibits human lymphocyte IFN--yproduction by suppressing natural killer cell stimulatory factor/interleukin-l2 synthesis in accessory cells. J . Exp. Med. 178, 1041-1048. 109. Mamo, S., Ahn, H., Yu, W., Tomnra, M., Wysocka, M., Yamamoto, N., Kobayashi, M., Hamaoka, T., Trinchieri, G., and Fujiwara, H. (1997). Establishment of an IL12-responsive T cell clone: Its characterization and utilization in the quantitation of IL-12 activity. J . Leukocyte Biol. 61, 346-352. 110. Schwaller, J., Tobler, A,, Niklaus, G., Hunvitz, N., Hennig, I., Fey, M. F., and Borisch, B. (1995). Interleukin-12 expression in human lymphomas and nonneoplastic lymphoid disorders. Blood 85, 2182-2188. 111. Wolf, S., Seiburth, D., Pernssia, B., Yetz-Adalpe, J., D’Andrea, A,, and Trinchieri, G. (1992). Cell sources of natural killer cell stimulatory factor (NKSF/IL-12) transcripts and subunit e-wpression.FASEB J . 6, A1335. 112. Hsu, D. H., de Waal Malefyt, R., Fiorentino, D. F., Dang, M. N., Vieira, P., de Vries, J., Spits, H., Mosman, T. R., and Moore, K. W. (1990). Expression of interleukin 10 activity by Epstein-Barr virus protein BCRF1. Science 250, 830-832. 113. Benjamin, D., Knobloch, T., and Dayton, M. (1992). Human B-cell interleukin-10: B-cell lines derived from patients with acquired immunodeficiency syndrome and Burkitt’s lymphoma constitutively secrete large quantities of interleukin-10. Blood 80, 1289-1298. 114. Benjamin, D., Hooker, S., and Miller, J. (1990). Differential effects of teleocidin on TNFa receptor regulation in human B cell lines: Relationship to coexpression of IL2 and IL-1 receptors and to lymphokine secretion. Cell. Zmmunol. 125, 480-497. 115. Benjamin, D., Kofler, G., and Tschacler, E. (1992). Human B-cell TNFP microheterogeneity. Lyniphokine Cytokine Res. 11, 45-54. 116. Benjamin, D., Venkatanarayanan, S., Knobloch, T. J., Armitage, R. J., Dayton, M A., and Goodwin, R. G. (1994). B cell IL-7: Human B cell lines constitutive secrete IL7 and express IL-7 receptors. J . lmmiinol. 152, 4749-4757. 117. Chehimi, J., Starr, S., Frank, I., Rengaraju, M., Jackson, S. J., Llanes, C., Kobayashi, M., Pernssia, B., Young, D., Nickharg, E., Wolf, S. F., and Trinchieri, G. (1992) Natural killer cell stimulatory factor (NKSF) increases the cytotoxic activity of NK cells from both healthy donors and HIV-infected patients. J . Exp. Med. 175,789-796. 118. Chehimi, J., Starr, S., Frank, I., D’Andrea, A., Ma, X., MacGregor, R. R., Sennelier, J., and Trinchieri, G. (1994).Impaired interleukin-12 production in human immunodeficiency virus-infected patients. J. E x p Med. 179, 1361-1366. 119. Clerici, M., Lucey, D. R., Berzofsky, J. A., Pinto, L. A., Wynn, T. A,, Blatt, S. P., D o h , M. J,, Hendrix, C. W., Wolf, S. F., and Shearer, G. M. (1993). Restoration of HIV-specific cell-mediated immune responses by interleukin-12 in vitro. Science 262, 1721-1724. 120. Gn&y, J., Ria, F., Galbiati, F., and Adorini, L. (1997). Normal B cells fail to secrete interleukin-12. Eur. J. ~ ? ~ L ? ? ~ l l I27, l O ~ .1632-1639. 121. Trinchieri, G. (1994). Interleukin 12: A cytokine produced by antigen-presenting cells with imrnunoregulatory functions in the generation ofT helper cells type 1and cytotoxic lymphocytes. Blood 84, 4008-4027. 122. Becher, B., Dodelet, V., Fedorowicz, V., and Antel, J. P. (1996).Soluble tuiiior necrosis factor receptor inhibits interleukin 12 production by stimulated human adult microglial cells in vitro. J. Clin. Invest. 98, 1539-1543.
202
C,IORC.IO TRINCHIERI
123. Stalder, A. K., Pagenstecker, A., Yu, N. C., Kincaid, C., Chiang, C., Hobbs, M. V., Bloom, F. E., and Campbell, I. L. (1997). Lipopolysaccharide-inducedIL-12 expression in the central nervous system and cultured astrncytes and microglia. J. Immunol. 159, 1344-1351. 124. Constantinescu, C, S., Frei, K., Wysocka, M., Trinchieri, G., Malipiero, U., Rostami, A,, and Fontana, A. (1996). Astrocytes and microglia produce interleukin-12 p40. Ann. N.Y. Acad. Sci. 795, 328-333. 125. Aloisi, F., Penna, G., Cerase, J., Iglesias, B. M., and Adorini, L. (1997).IL-12production by central nervous system microglia is inhibited by astr0cytes.J. Immunol. 159, 16041612. 126. Sato, Y., Roman, M., Tighe, H., Lee, D., Corr, M., Nguyen, M., Silverman, G. J., Lotz, M., Carson, D. A., and Raz, E. (1996). Immunostimulatory DNA sequences necessary for effective intradermal gene immunization. Science 273, 352-354. 127. Hodge-Dufour, J., Noble, P. W., Horton, M. R., Bao, C., Wysocka, M., Burdick, M. D., Strieter, R. M., Trinchieri, G., and Pur6, E. (1997). Induction of IL-12 and chemokines by hyaluronan requires adhesion-dependent priming of resident but not elicited macrophages. J. Immunol. 159, 2492-2500. 128. Shu, U., Kiniwa, M., Wu, C. Y., Maliszewski, C., Vezzio, N., Hakimi, J., Gately, M., and Delespesse, G . (1995). Activated T cell induce interleukin-12 production by monocytes via CD40-CD40 ligand interaction. Eur. J. Immunol. 25, 1125-1128. 129. Cella, M., Scheidegger, D., Plamer-Lehmann, K., Lane, P., Lanzavecchia, A,, and Alber, G . (1996). Ligation of CD40 on dendritic cells triggers production of high levels of interleukin-12 and enhances T cell stimulatory capacity:T-T help via APC activation. J. Exp. Med. 184, 747-752. 130. Armant, M., Armitage, R., Boiani, N., Delespesse, G., and Sarfati, M. (1996).Functional CD40 ligand expression on Tlymphocytes in the absence ofT cell receptor engagement: Involvement in interleukin-2-induced interleukin- 12 and interferon-gamma production. Eur. J. Immunol. 26, 1430-1434. 131. Skeen, M. J., Miller, M. A., Shinnick, T. M., and Ziegler, H. K. (1996). Regulation of rnurine macrophage IL- 12 production: Activation of macrophages in vivo, restimulation in vitro, and modulation by other cytokines.J. lmmunol. 156, 1196-1206. 132. Fujimoto, T., Duda, R. B., Szilvasi, A,, Chen, X., Mai, M., and ODonnell, M. A. (1997). Streptococcal preparation OK-432 is a potent inducer of IL-12 and a T helper ceI1 1 dominant state. J. ImmunoL 158,5619-5626. 133. Shibata, Y., Metzger, W. J., and Myrvik, Q. N. (1997). Chitin particle-induced cellmediated immunity is inhibited by soluble mannan. J. Immnnul. 159, 2462-2467. 134. Fulton, S. A,, Johnsen, J. M., Wolf, S. F., Sieburth, D. S., and Boom, W. H. (1996). Interleukin-12 production by human monocytes infected with Mycobacterlum tuberculosis: Role of phagocytosis. Infect. Imniun. 64, 2523-2531. 135. Kubin, M., Chow, J. M., andTrinchieri, G . (1994).Differential regulation ofinterleukin12 (IL-12),tumor necrosis factor-@,and IL-lB production in human myeloid leukemia cell lines and peripheral blood mononuclear cells. Blood 83, 1847-1855. 136. Verhasselt, V., Buelens, C., Willems, F., De Groote, D., Haeffner-Cavaillon, N., and Goldman, M. (1997). Bacterial lipopolysaccharide stimulates the production of cytokines and the expression of costimulatory molecules by human peripheral blood dendritic cells: Evidence for a soluble CD14-dependent pathway J. Immunol. 158,29192925. 137. Cleveland, M. G . , Gorham, J. D., Murphy, T. L., Tuomanen, E., and Murphy, K. M. (1996). Lipoteichoic acid preparations of gram-positive bacteria induce interleukin12 through a CD14-dependent pathway. Infect. Immun. 64, 1906-1912.
INTERLEUKIN-12
203
138. Oswald, I. P., Dozois, C. M., Petit, J., and Lemaire, G. (1997). IL-12 synthesis is an obligatory step in trehalose dimycolate-induced activation of mouse peritoneal . 1364-1369. macrophages. Zigect. Z i ~ i ~ i i t m65, 139. Gmnvald, E., Cl~iarainonte.M.. Hieny, S., Wysocka, M., Trinchieri, G., Vogel, S. N., Gazzinelli, R. T., and Slier, A. (1996). Biochemical characterization and protein hnase C dependency of monokine-inducing activities of Toroplusniu gondii. Znfect. h i n u n . 64,2010-2018. 140. Camargo, M. M., Altneida, I. C., Pereira, M. E. S., Ferguson, M. A. J,, Travassos, L. R., and Gazzinelli, R. T. (1997). Glycosylpliospliatidylinositol-anchoredinucin-like glycoproteins isolated from Trypmmwrtiu criizi trypomastigotes initiate the synthesis of proinflainrnatory cytokines by macrophages. J , Zn~munol.158, 5890-5901. 141. Skeiky, Y. A. W., Guderian. J. A., Benson, D. R., Bacelar, O., Carvaho, E. M., Kubin, M., Badaro, R., Trinchieii, G., and Reed, S. G. (1995). A recombinant Lt.ishianin antigen that stimulates human peripheral blood inononuclear cells to express a Thltype cytokine profile and to produce interleukin 12.1. Exp. Med. 181, 1527-1537. 142. Leimg, D. Y. M., Gately, M., Trumble, A,, Ferguson-Darnell, B., Schlievert, P. M., and Picker, L. J. (1995). Bacterial superantigens induce T cell expression of the skinselective homing receptor, the cutaneous lymphocyte-associatedantigen, via stimulation of interleukin 12 production. 1.Exp. Merl. 181, 747-753. 143. Halpern, M. D., Kurlander. R. J., and Pisetsky, D. S. (1996). Bacterial DNA induces murine interferon-garnlna production by stiinulation of interleukin-12 and tumor necrosis factor-alpha. Cell. Ztnnninol. 167, 72-78. 144. Klininan, D. M., Yi, A. K., Beaucage, S. L., Conover, J., and Krieg, A. M. (1996). CpG motifs present in bacteria DNA rapidly induce lymphocytes to secrete interleukin 6, interleukin 12, and interferon gamma. Proc. Nntl. A c d . Sci. USA 93, 2879-2883. 14.5. Manetti, R., Annunziato, F., Tomasevic, L., Gianno, V., Parronchi, P., Romagnani, S., and Maggi, E. (1995). Polyinosinic acid: Polycytidylic acid promotes T helper type 1specific immune responses by stimulating macrophage production of IFN-a and IL12. Eur. J . ~ r i ~ i ~ i u 25, i i o ~2656-2660. . 146. Hauser, C. J., Zhou, X., Joshi, P., Cuchens, M. A., Kregor, P., Devidas, M., Kennedy, R. J., Poole, G. V., and Hughes, J. L. (1997).The immune microenvironment of human fracturekoft tissue heinatomas and its relationship to systemic immunity. 1. T?-nmut 42, 895-903. 147. Gerinann, T., Paitenheimer, A., and Riide, E. (1990). Requirements for the growtli of TH1lymphocyte clones. Eur. J . Zrnniunol. 20, 2035-2040. 148. Macatonia, S. E., Hosken. N. A., Litton, M., Vieira, P., Hsieh, C., 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 Thl cells from naive CD4+ T cells. J . Ztimiunol. 154, 5071-5079. 149. Stiiber, E., Strober, W., and Neurath, M. (1996). Blocking the CD40L-CD40 interaction in vivo specifically prevents the priming of T helper 1 cells through the inhibition of interleukin 12 secretion. 1. Erp. Med. 183, 693-698. 150. Men, R. C., Arniitage, R. J., Conley, M. E., Rosenblatt, H., Jenkins, N. A., Copeland, N. G,, Bedell, M. A., Edelhoff, S., Disteche, C. M., Simoneaiuc, D. K., Fanslow, W. C., Belmont, J., and Spriggs, M. K. (1993). CD40 ligand gene defects responsible for X-linked hyper-IgM syndrome. Science 259, 990-993. 151. DiSanto, J. P., Bonnefoy, J. Y., Gauchat, J. F., Fischer, A., and de Saint Basile, G. (1993).CD40 ligaid mutations in X-linked immunodeficiencywith hyper-IgM. Nature 361, 541-543.
204
GIORCIO TRINCHIERI
152. Kennedy, M. K., Picha, K. S., Fanslow, W. C., Grabstein, K. H., Alderson, M. R., Clifford, K. N., Chin, W. A,, and Mohler, K. M. (1996).CD40/CD40 ligand interactions are required for T cell-dependent production of interleukin-12 by mouse macrophages. Eur. J. Immunol. 26,370-378. 153. Peng, X., Kasran, A,, Warmerdam, P. A,, de Boer, M., and Ceuppens, J. L. (1996). Accessory signaling by CD40 for T cell activation: Induction of Thl and Th2 cytokines and synergy with interleukin-12 for interferon-gamma production. Eur. J. Immunol. 26, 1621-1627. 154. Kato, T., Hakamada, R., Yamane, H., and Nariuchi, H. (1996). Induction of IL-12 p40 messenger RNA expression and IL-12 production of macrophages via CD40CD40 ligand interaction. J. Immunol. 156, 3932-3938. 155. Maruo, S., Oh-Hora, M., Ahn, H., Ono, S., Wysocka, M., Kaneko, Y., Yagita, H., Okumura, K., Kikutani, H., Kishimoto, T., Kobayashi, M., Hamaoka, T., Trinchieri, G., and Fujiwara, H. (1997). B cells regulate CD40 ligand-induced IL-12 production in antigen-presenting cells (APC) during T celVAPC interactions. J. Immunol. 158, 120-126. 156. DeKruyff, R. H., Gieni, R. S., and Umetsu, D. T. (1997). Antigen-driven but not lipopolysaccharide-driven IL-12 production in macrophages requires triggering of CD40. J. Zmmunol. 158, 359-366. 157. Roy, M., Waldschmidt, T., Aruffo, A,, Ledbetter, J. A., and Noelle, R. J. (1993). The regulation of the expression of gp39, the CD40 ligand, on normal and cloned CD4+ T cells. J. Immunol. 151, 2497-2510. 158. Hayes, M. P., Wang, J., and Norcross, M. A. (1995). Regulation ofinterleukin-12 expression in human monocytes: Selective priming by IFN-y of LPS-inducible p35 and p40 genes. Blood 86,646-650. 159. Sartori, A., Oliveria, M. A. P., Scott, P., and Trinchieri, G. (1997). Metacyclogenesis modulates the ability of Leishmania promastigotes to induce IL-12 production in human mononuclear cells. J. Zmmunol. 159, 2849-2857. 160. Flesch, I. E. A., Hess, J. H., Huang, S., Aguet, M., Rothe, J., Bluethmann, H., and Kaufmann, S. H. E. (1995).Earlyinterleukin 12 production by macrophages in response to mycobactend infection depends on interferon y and tumor necrosis factor a. J. Exp. Med. 181, 1615-1621. 161. Wysocka, M., Kubin, M., Vieira, L. Q., Ozmen, L., Garotta, G., Scott, P., and Trinchieri, G. (1995). Interleukin-12 is required for interferon-? production and lethality in lipopolysaccharide-induced shock in mice. Eur. 1.Immunol. 25, 672-676. 162. Heinzel, F. P., Rerko, R. M., Ling, P., Hakimi, J., and Schoenhaut, D. S. (1994). Interleukin 12 is produced in vivo during endotoxemia and stimulates synthesis of gainma interferon. Infect, lmmun. 62,4244-4249. 163. Heinzel, F. P., Rerko, R. M., Ahmed, F., and Hujer, A. M. (1996). IFN-y-dependent production of IL-12 during murine endotoxemia. 1.Immunol. 157,4521-4528. 164. Scharton-Kersten, T. M., Wynn, T. A., Denkers, E. Y., Bala, S., Grunvald, E., Hieny, S., Gazzinelli, R. T., and Sher, A. (1996). In the absence of endogenous IFN-y, mice develop unimpaired IL-12 responses to T o x ~ l a s m agondii while failing to control acute infection. J. Immunol. 157, 4045-4054. 165. Kubin, M., Kamoun, M., and Trinchieri, G. (1994). Interleukin-12 synergizes with B7/CD28 interaction in inducing efficient proliferation and cytokine production of human T cells. 1.Exp. Med. 180, 211-222. 166. Murphy, E. E., Terres, G., Macatonia, S. E., Hsieh, C., Mattson, J., Lanier, L., Wysocka, M., Trinchieri, G., Murphy, K., and O’Garra, A. (1994). B7 and IL-12 cooperate for
INTERLEUKIN-18
205
proliferation and IFN-y production by mouse T helper clones that are unresponsive to B7 costimulation. J. Exp. Med. 180, 223-231. 167. Finkelman, F. D., Madden, K. B., Cheever, A. W., Katona, I. M., Morris, S. C., Gately, M. K., Hubbard, B. R., Cause, W. C., and Urban, J. F., Jr. (1994).Effects of interleukin 12 on immune responses and host protection in mice infected with intestinal nematode parasites. J. Exp. Med. 179, 1563-1572. 168. Gerosa, F., Paganin, C., Peritt, D., Paiola, F., Scupoli, M. T., Aste-Amezaga, M., Frank, I., and Trinchieri, G. (1996). Interleukin-12 primes human CD4 and CD8 T cell clones for high production of both interferon-y and interleukin-10. J. Exp. Men. 183, 2559-2569. 169. D’Andrea, A., Ma, X., Aste-Ainezaga, M., Paganin, C., and Trinchieri, G. (1995). Stimulatory and inhibitory effects of IL-4 and IL-13 on production of cytokines by human peripheral blood mononuclear cells: Priming for IL-12 and TNF-ry production. J. Exp. Med. 181,537-546. 170. de Wad Malefyt, R., Figdor, C. G., Huijbens, R., Mohan-Peterson, S., Bennett, B., Culpepper, J., Dang, W., Zurawski, G., and de Vries, J. E. (1993). Effects of IL-13 on phenotype, cytokine production, and cytotoxic function of human monocytes. J. Zmmunol. 151, 6370-6381. 171. Hino, A,, and Nariuchi, H. (1996). Negative feedback mechanism suppresses interleukin-12 production by antigen-presenting cells interacting with T helper 2 cells. Eur. J lmmunol. 26, 623-628. 172. Vezzio, N., Sarfati, M., Yang, L. P., Demeure, C. E., and Delespesse, G. (1996). Human The-like cell clones induce IL-12 production by dendritic cells and inay express several cytokine profiles. Znt. Zmniunol. 8, 1963-1970. 173. Minty, A., Ferrara, P., and Caput, D. (1997). Interleukin-13 effects on activated monocytes lead to novel secretion profiles intermediated between those induced by interleukin-10 and by interferon-y. Eur. Cytokine Netw. 8, 189-201. 174. Marshall, J. D., Robertson, S. E., Trinchieri, G., and Chehimi, J. (1997). Priming with IL-4 and IL-3 during HIV-1 infection restores in vitro IL-12 production by mononuclear cells of HIV-infected patients. J. Imaiunol. 159, 5705-5714. 175. Ma, X., D’Andrea, A,, Kubin, M., Aste-Ainezaga, M., Sartori, A., Monteiro, J., Showe, L., Wysocka, M., and Trinchieri, G. (1995).Production of interleukin-12. Res. Immunol. 146,432-438. 176. Takenaka, H., Maruo, S., Yamamoto, N., Wysocka, M., Ono, S., Kobayashi, M., Yagita, H., Okumura, K., Hamaoka, T., Tiinchieri, G., and Fujiwara, H. (1997). Regulation of T cell-dependent and -independent IL-12 production by three Th2-type cytokines IL-10, IL-6, and IL-4. J . Leuk. B i d . 61, 80-87. 177. Flesch, I. E., Wandersee, A,, and Kaufinann, S. H. (1997). Effects of IL-13 on inurine listeriosis. lnt. Zmmunol. 9, 467-474. 178. Chensue, S. W., Warmington, K. S., Ruth, J. H., Sanghi, P. S., Lincoln, P., and Kunkel, S. L. (1996). Role of monocyte chemoattractant protein-1 (MCP-1) in Thl (mycobacterial) and Th2 (schistosomal) antigen-induced granuloma formation: Relationship to local inflammation, Th cell suspension, and IL-12 production. J. lmmunol. 157,4602-4608. 179. Trepicchio, W. L., Bozza, M., Pedneault, G., and Dorner, A. J. (1996). Recoinbinant human IL-11 attenuates the inflammatory response through down-regulation of proinflaniinatory cytokine release and nitric oxide production.]. Zmmunol. 157,3627-3634. 180. Leng, S. X., and Elias, J. A. (1997). Interleukin-11 inhibits macrophage interleukin12 production. J. lmniunol. 159, 2161-2168.
206
GIORCIO TRINCHIERI
181. van der Pouw Kraan, T. C. T. M., Boeije, L. C. M., Smeenk, R. J. T., Wijdenes, J., and Aarden, L. A. (1995). Prostaglandin-E is a potent inhibitor of human interleukin 12 production. J. Exp. Med. 181, 775-779. 182. Baron, P., Constantin, G., D’Andrea, A,, Ponzin, D., Scarpini, E., Scarlato, G., Trinchieri, G., Rossi, P., and Cassatella, M. A. (1993). Production of tumor necrosis factor and other proinflammatory cytokines by human mononuclear phagocytes stimdated with myelin P2 protein. Proc. Nutl. Acud. Sci. USA 90, 4414-4418. 183. Elenkov, I. J., Papanicolaou, D. A,, Wilder, R. L., and Chrousos, G. P. (1996). Modulatory effects of glucocorticoids and catecholamines on human interleukin-12 and interleukin-10 production: Clinical implications. Proc. Assoc. Am. Phys. 108,374-381. 184. Blotta, M. H., DeKruyff, R. H., and Umetsu, D. T. (1997). Corticosteroids inhibit IL-12 production in human monocytes and enhance their capacity to induce IL-4 synthesis in CD4+ lymphocytes. I. I-mmunol. 158, 5589-5595. 185. Kelly, R. W., Carr, G. G., and Critchley, H. 0. (1997). A cytokine switched induced by human seminal plasma: An immune modulation with implications for sexually transmitted disease. Hum. Repro. 12, 677-681. 186. Kelly, R. W. (1997). Prostaglandins in primate semen: Biasing the immune system to benefit spermatozoa and virus? Prostugland. Leukot. Essent. Fatty Acids 57,113-118. 187. Lemire, J. M., Archer, D. C., Beck, L., and Spiegelberg, H. L. (1995).Immunosuppressive actions of 1,25-dihydroxyvitamin D3: Preferential inhibition of Thl functions. J. Nutr. 125, 1704s-17085. 188. Cantorna, M. T., Nashold, F. E., and Hayes, C. E. (1995).Vitamin A deficiency results in a priming environment conductive for Thl cell development. Eur. J. Zmmnunol. 25, 1673-1679. 189. Fox, F. E., Kubin, M., Cassin, M., Niu, Z., Hosoi, J., Toni, H., Granstein, R. D., Trinchieri, G., and Rook, A. H. (1997).Calcitonin gene-related peptide inhibits proliferation and antigen presentation by human peripheral blood mononuclear cells: Effects on B7, interleukin-10, and interleukin-12. J. Invest. D e m t o l . 108, 43-48. 190. Toni, H., Hosoi, J., Beisert, S.,Xu, S.,Fox, F. E., Asahina, A., Takashima, A., Rook, A. H., and Granstein, R. D. (1997). Regulation of cytokine expression in macrophages and the Langerhans cell-like line XS52 by calcitonin gene-related peptide. J. L e d . B i d . 61, 216-223. 191. Turka, L. A,, Goodman, R. E., Rutkowski, J. L., Sima, A. A., Merry, A., Mitra, R. S., Wrone-Smith, T., Toews, G., Strieter, R. M., and Nickoloff, B. J. (1995). Interleukin 12: A potential link between nerve cells and the immune response in inflammatory disorders. Mol. Med. 1, 690-699. 192. Moller, D. R., Wysocka, M., Greenlee, B. M., Ma, X., Wahl, L., Trinchieri, G., and Karp, C. L. (1997). Inhibition of human interleukin-12 production by pentoxifylline. Im?nulaologg 91, 197-203. 193. Moller, D. R., Wysocka, M., Greenlee, B. M., Ma, X., Wahl, L., Flockhart, D. A,, Trinchieri, G., and Karp, C. L. (1997). Inhibition of interleukin 12 production by thalidomide. Submitted for publication. 194. Karp, C. L., Wysocka, M., Wahl, L. M., Ahearn, J. M., Cuomo, P. J., Sherry, B., Trinchieri, G., and Griffin, D. E. (1996). Mechanism of suppression of cell-mediated immunity by measles virus. Science 273, 228-231. 195. Suttenvala, F. S., Noel, G. J., Clynes, R., and Mosser, D. M. (1997). Selective suppression of interleukin-12 induction after macrophage receptor ligation. J. Exp. Med. 185, 1977-1985. 196. Marth, T., and Kelsdl, B. L. (1997). Regulation of interleukin-12 by complement receptor 3 signaling. J. En?. Med. 185, 1987-1995.
INTERLEUKIN-12
207
197. Smith, T. J., Ducharme, L. A,, and Weis, J. 13. (1994). Preferential expression of interleulan-12 or interleukm-4 by mririne bone niarrow mast cells derived in mast cell growth factor or interleukin-3. E14r.J . Zmnunol. 24, 822-826. 198. Aragane, Y., Riemann, H., Barhdwaj, R. S., Schwarz, A,, Sawada, Y., Yainada, H., Luger, T. A,, Kubin, M., Trinchieri, G., and Schwarz, T. (1994). IL-12 is eqressetl and released by human keratinocytes and epiderinoid carcinoma cell lines. j . Inmrunol. 153, 5366-5372. 299. Muller, G., Saloga, J., Gennann, T., Bellinghairsen, I., Mohamadzadeh, M., Knop, J., and Enk, A. € I . (1994). Identification and induction of human kewtinocyte-derived IL-12. J. Clin. Inoest. 94, 1799-1805. 200. Enk, C. D., Mahanty, S., Rlauvelt, A., and Katz, S. I. (1996). UVB induces IL12 transcription in human keratinocytes in vivo and in vitro. Photochsm. Photobiol. 63, 854-859. 201. Goodman, R. E . , Nestle, F., Naidu, Y. M., Green, J. M., Thompson, C . H., Nickoloff, B. J., and Turka, L. A. (1994). Keratinocyte-derived T cell costimulation induces preferrential production of IL-2 and IL-4 but not IFN-y.J. Imnzt4nol. 152,5189-5198. 202. Kang, K., Kubin, M., Cooper, K. D., Lessin, S. R., Trinchieri, G., and Rook, A. 14. (1996). IL-12 synthesis by human Langerlians cells. J. Inzmtinol. 156, 1402- 1407. 203. Kanangat, S., Nair, S., Babu, J. S., and Rouse, B. T. (1995). Exp-ession of cytokine mRNA in niurine splenic dendritic cells and better induction of T cell-derived cytokines by dendritic cells than by macrophages during in vitro costinidation assay using specific antigens. J. Leuk. B i d . 57, 310-316. 204. Heufler, C., Koch, F., Stanzl, U., Topar, G., Wysocka, M., Trinchieri, G., Enk, A,, Steinman, R. M., Romani, N., and Schuler, G. (1996). Interleulan-12 is produced by dendritic cells and mediates Thl development as well as IFN-y production by Thl cells. Eur. J. bnmrinol. 26, 659-668. 205. Kelsall, B. L., Stuber, E.. Neurath, M., and Strober, W. (1996). lnterleukin-12 production by dendritic cells: The role ofCD40-CD40L interactions in T h l T-cell responses. Ann. N.Y. Actid. Sci. 795, 116-126. 206. Koch, F., Stanzl, U., Jennewein, P., Janke, K., Iieufler, C., Klmpgen, E., Romani, N., and Schuler, G. (1996). High level IL-12 production by inurine dendritic cells: Upregulation via MHC class I1 and CD40 molecules and downregulation by IL-4 and IL-10. J. Exp. Mecl. 184, 741-746. 207. Scheicher, C., Mehlig, M., Dienes, H. P., and Keske, K. (1995).Uptake ofmicroparticleadsorbed protein antigen by bone marrow-derived dendritic cells results in upregulation of IL-1-a and IL-12 p40/p35 and triggers prolonged, efficient antigen presentation. Eur. 1. Inimtinol. 25, 1566-1572. 208. De Smedt, T . , Van Mechelen, M., De Becker, G., Urbain, J.. Leo, O., and Moser, M. (1997). Effect of interleukin-10 on dentlritic cell maturation and function. Eur. J . Zmniunol. 27, 1229-1235. 209. Buelens, C., Verhasselt, V., DeGroote, D., Thielenians, K., Goldinan, M., and Willems, F. (1997).Interleukin-10 prevents the generation of dendritic cells from human peripheral blood mononuclear cells cultured with interleukin-4 and granulocyte/macrophagecolony-stimulating factor. Eur. J . Imrnutlol. 27, 756-762. 210. Kalinska, P., Hilkens, C. M. U.. Snijders, A,, Snijdewint, F. G. M., aid Kapsenberg, M. L. (1997). IL-12-deficient dendritic cells, generated in the presence of prostaglandin EB,promote type 2 cytokine production in maturing human naive T helper cells. J. l t t l t n t i t d . 159, 28-35. 211. Winzler, C., Rovere, P., Rescigno, M., Granucci, F., Penna, G., Adorini, L., Ziinrnernianii, V. S., Davoust, J., and Ricciardi-Castagnoli, P. (1997). Maturation stages of
208
212.
213.
214.
215.
216.
21 7.
218.
219.
220.
221.
222.
223.
224.
GIORCIO TRINCHIERI
mouse dendritic cells in growth factor-dependent long-term cultures. J. E x p Med. 185, 317-328. Pulendran, B., Lingappa, J., Kennedy, M. K., Smith, J., Teepe, M., Rudensky, A., MaIiszewski, C. R., and Maraskovsky, E. (1997). Developmental pathways of dendritic cells in vivo. J . lmmunol. 159, 2222-2231. Snidjers, A,, Hilkens, C. M., van der Pouw Kraan, T. C., Engel, M., Aarden, L. A., and Kapsenberg, M. L. (1996).Regulation of bioactive IL-12 production in lipopolysaccharide-stimulated human monocytes is determined by the expression of the p35 subunit. J. lmmunol. 156, 1207-1212. Murphy, T. L., Cleveland, M. G., Kulesza, P., Magram, J., and Murphy, K. M. (1995). Regulation of interleukin 12 p40 expression through an NF-kB half-site. Mol. Cell. B i d . 15, 5258-5267. Ma, X., Neurath, M., Gii, G., and Trinchieri, G. (1997). Identification and characterization of a novel ets-%related nuclear complex implicated in the activation of the human IL-12 p40 gene promoter. J . Biol. Chem. 272, 10389-10395. Plevy, S. E., Gemberling, J. H. M., Hsu, S., Dorner, A. J., and Smale, S. T. (1997). Multiple control elements mediate activation of the murine and human interleukin 12 p40 promoters: Evidence of functional synergy between C/EBP and Re1 proteins. Mol. Cell. Biol. 17, 4572-4588. Screpanti, I . , Romani, L., Musiani, P., Modesti, A,, Fattori, E., Lazzaro, D., Sellitto, C., Scarpa, S., Bellavia, D., Lattanzio, G., Bistoni, F., Frati, L., Cortese, R., Gulino, A., Ciliberto, G., Costantini, F., and Poli, V. (1995). Lymphoproliferative disorder and imbalanced T-helper response in CIEBPP-deficient mice. EMBO J. 14, 1932-1941. Holtschke, T., Lohler, J., Kanno, Y., Fehr, T., Giese, N., Rosenbauer, F., Lou, J., Knobeloch, K. P., Gabriele, L., Waring, J. F., Bachmann, M. F., Zinkernagel, R. M., Morse, H. C., 111, Ozato, K., and Horak, I. (1996). Immunodeficiency and chronic myelogenous leukemia-like syndrome in mice with a targeted mutation of the ICSBP gene. Cell 87, 307-311. Scharton-Kersten, T., Contursi, C., Masumi, A,, Sher, A,, and Ozato, K. (1997).ICSBPdeficient mice display impaired resistance to intracellular infection due to a primary defect in IL-12 p40 induction. J. E x p Med. 186, 1523-1534. Nelson, N., Kanno, Y., Hong, C., Contursi, C., Fujita, T., Fowlkes, B. J., O’Connell, E., Hu-Li, J.. Paul, W. E., Jankovic, D., Sher, A. F., Coligan, J. E., Thornton, A,, Appella, E., Yang, Y., and Ozato, K. (1996). Expression of IFN regulatory factor family proteins in lymphocytes: Induction of Stat-1 and IFN consensus sequence binding protein expression by T cell activation. J. lmmunol. 156, 3711-3720. Wang, I., Ma, X., Contursi, C., Masumi, A,, Lo, J., Trinchieri, G., and Ozato, K. (1997). A critical role of interferon consensus sequence binding protein (ICSBP) in the activation of IL-12 p40 gene promoter. J. lnterf Cykokine Res. 17, 576. 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 ofTh1-type immune responses by the transcription factor IRF-1. Immunity 6,673-679. Lohoff, M., Ferrick, D., Mittrucker, 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. Osipovich, O., Fegeding, K., Misuno, N., Kolesnikova, T., Savostin, I., Sudarikov, A,, and Voitenok, N. (1993). Differential action of cycloheximide and activation stimuli on transcription of tumor necrosis factor-&, IL-1P, IL-8, and p53 genes in human monocytes. J. lmmunol. 150,4958-4965.
INTERLEUKIN-I2
209
225. Shaw, J., Meerovitch, K., Elliot, J. F., Bleackley, R. C., and Paetkau, V. (1987). Induction, suppression and superinduction of lymphokine mRNA in T lymphocytes. Mol. Inamirnol. 24, 409-419. 226. Zuhiaga, A,, MuAoz, E., and Huber. B. (1991). Superinduction of IL-2 gene transcription in the presence of cycloheximide. J. Zmmunol. 146, 3857-3863. 227. Aste-Amezaga, M., Ma,X., Sartori, A,, andTrinchieri, G. (1998). Molecular mechanisms of the induction of interleukin-12 and its inhibition by IL-10. J. imnmnol., in press. 228. Bogdan, C., Paik, J., Vodovotz, Y., and Nathan, C. (1992). Contrasting mechanisms for suppression of macrophage cytokine release by transforming growth factor-beta and interleukin-lo. J . Biol. Chenz. 267, 23301-23308. 229. Wang, P., Wu, P., Siegel, M. I., Egan, R. W., and Billah, M. M. (1994). IL-10 inhibits transcription of cytokine genes in human peripheral blood mononuclear cells. J. Immunol. 153, 811-816. 230. Kasama, T., Strieter, R. M., Lukacks, N. W., Burdick, M. D., and Kunkel, S. L. (1994). Regulation of neutrophil-derived chemokine expression by IL-10. J. Imntunol. 152, 3559-3569. 231. Ogawa, M. (1993). Differentiation and proliferation of hematopoietic stem cells. Blood 81,2844-2853. 232. Jacobsen, 8. E., Veiby, 0. P., and Smeland, E. B. (1993). Cytotoxic lymphocyte maturation factor (interleukin 12) is a synergistic growth factor for hematopoietic stem cells. J. Exp. Med. 178, 413-418. 233. Ploemacher, R. E., van Soest, P. L., Boudewijn, A., and Neben, S. (1993). Interleukin12 enhances interleukin-3 dependent niutllineage hematopoietic colony formation stimulated by interleukin-1 1 or steel factor. Leukemiu 7, 1374-1380. 234. Ploemacher, R. E., van Soest, P. L., Voonvinden, H., and Boudewijn, A. (1993). Interleukin-12 synergizes with interleukin-3 and steel factor to enhance recovery of murine hemopoietic stem cells in liquid culture. Leukemia 7, 1381-1388. 235. Dyhedal, I., Larsen, S., and Jacohsen, S. E. W. (1995). IL-12 directly enhances in vitro murine erythopoiesis in combination with IL-4 and stem cell factor. J. Immunol. 154, 4950-4955. 236. Jacobsen, S. E., Okkenhaug, C., Myklebust, J., Veihy, 0. P., and Lyman, S. D. (1995). The FLT3 ligand potently and directly stimulates the growth and expansion of primitive murine bone marrow progenitor cells in vitro: Synergistic interactions with interleukin (IL) 11,IL-12, and other heinatopoietic growth factors.]. Exp. Med. 181,1357-1363. 237. Hirayama, F., Katayama, N., Neben, S., Donaldson, D., Nickbarg, E. B., Clark, S. C., and Ogawa, M. (1993). Synergistic interaction between interleukin-12 and steel factor in support of proliferation of murine lymphohematopoietic progenitors in culture. Blood 83, 92-98. 238. Bellone, G., and Trinchieri, G. (1994). Dual Stimulatory and inhibitory effect of NK cell stimulatory factor/IL-12 on human heinatopoiesis. J. Intnzunol. 153, 930-937. 239. Fardoun-Joalland, D., Teixeira-Lebrun, G., Lenormand, B., Dzondo-Gadet, M., and Vannier, J. P. (1995). Synergxsm of interleukin-12 and interleukin-3 on development of hematopoietic progenitors. Eur. J. Huematol. 54, 172-175. 240. Hirao, A,, Takane, Y., Kawano, Y., Sato, J., Suzue, T., Abe, T., Saito, S., Kawahito, M., Okamoto, Y., and Makimoto, A. (1995). Synergism of interleukin 12, interleukin 3 and serum factor on primitive human hematopoietic progenitor cells. Stern Cells 13,47-53. 241. Bertolini, F., Soligo, D., Lazzari, L., Corsini, C., Servida, F., and Sirchia, G. (1995). The effect of interleukin-12 in ex-vivo expansion of human haemopoietic progenitors. Brit. J. Haenmtol. 90, 935-938.
210
GlORGlO TRINCHIERI
242. Verma, U. N., and Mazumder, A. (1995). Interleukin-12 (IL-12)alone or in synergistic combination with IL-2 for in vitro activation of human bone marrow: Differential effects at different time points. Bone Marrow Transplan. 16, 365-372. 243. Jackson, J. D., Yan, Y., Brunda, M. J., Kelsey, L. S., and Talmadge, J. E. (1995). Interleukin-12 enhances peripheral hematopoiesis in vivo. Blood 85, 2371-2376. 244. Tare, N. S . , Bowen, S., Warner, R. R., Carvajal, D. M., Benjamin, W. R., Riley, J. H., Anderson, T. D., and Gately, M. K. (1995). Administration of recombinant IL12 to mice suppresses hematopoiesis in the bone marrow but enhances heinatopoiesis in the spleen. 1. Znterferon Cytokine Res. 15, 377-383. 245. Eng, V. M., Car, B. D., Schnyder, B., Lorenz, M., Lugli, S., Aguet, M., Anderson, T. D., Ryffel, B., and Quesniaux, V. F. J. (1995). The stirnulatory effects of interleukin (IL)-12on hematopoiesis are antagonized by IL-12-induced interferon gamma in ~;ivo. 1.Exp. Med. 181, 1893-1898. 246. Neta, R., Stiefel, S . M., Finkelman, F., Herrmann, S . , and Ali, N. (1994).IL-12protects bone marrow from and sensitizes intestinal tract to ionizing radiation. I. Zmmnnol. 153, 4230-4237. 247. Alzona, M., Jack, H. M., Fisher, R. I., and E k , T. (1995). IL-12 activates IFN-gamma production through the preferential activation of CD30+ T cells.]. Zmmunol. 154,9-16 248. Alzona, M., Jack, H. M., Fisher, R. I., arid Ellis, T. M. (1994). CD30 defines a subset of activated human T cells that produce IFN-gamma and IL-5 and exhibit enhanced B cell helper activity. J. Zmmunol. 153, 2861-2867. 249. Aste-Amezaga, M., D’Andrea, A,, Kubin, M., and Trinchieri, G. (1994). Cooperation of natural killer cell stimulatory f~ctor/interleukin-12with other stimuli in the induction of cytokines and cytotoxic cell-associated molecules in human T and NK cells. Cell. Zmmunol. 156, 480-492. 250. Chan, S. H., Kobayashi, M., Santoli, D., Perussia, B., and Trinchieri, G. (1992). Mechanisms of IFN-y induction by natural killer cell stimulatory factor (NKSFOL12): Role of transcription and mRNA stability in the synergistic interaction between NKSF and IL-2. I. Immunol. 184,92-98. 251. Nagy, E., Buhlinann, J. E., Henics, T., Waugh, M., and Rigby, W. (1994). Selective modulation of IFN-gamma mRNA stability by IL-12/NKSF. Cell. bnmunol. 159, 140-151. 252. Ye, J., Ortaldo, J. R., Conlon, K., Winkler-Pickett, R., and Young, H. A. (1995). Cellular and molecular mechanisms of IFN-y production induced by IL-2 and IL-12 in a human NK cell line. J . Leuk. Bid. 58, 225-233. 253. Trinchieri, G., Matsumoto-Kobayashi, M., Clark, S. C., Sheehra, J., London, L., and Perussia, B. (1984). Response of resting human peripheral blood natural killer cells to interleukin-2. I. Exp. Med. 160, 1147-1169. 254. Li, L., Young, D., Wolf, S. F., and Choi, Y. S. (1996). Interleukin-12 stimulates B cell growth by inducing IFN-y. Cell. Z~nmunol168, 133-140. 255. Puddu, P., Fantuzzi, L., Borghi, P., Varano, B., Rainaldi, G., Guillemard, E., Malomi, E., Nicaise, P., Wolf, S. F., Belardelli, F., and Gessani, S. (1997). IL-12 induces IFNy expression and secretion in mouse peritoneal macrophages.]. Zmmunol. 159,34903497. 256. Gazzinelli, R. T., Hieny, S., Wynn, T. A,, Wolf, S., and Sher, A. (1993). Interleukin12 is required for the T-lymphocyte independent induction of interferon-y by an intracellular parasite and induces resistance in T-deficient hosts. Proc. Natl. Acad. Sci. USA 90, 6115-6119. 257. Tripp, C. S., Wolf, S. F., and Unanue, E. R. (1993). Interleukin 12 and tumor necrosis factor alpha are costimulators of interferon gamma production by natural killer cells
INTERLEUKIN-1%
258.
259.
260.
261.
262.
263.
264. 265.
266.
267.
268.
269.
270.
211
in severe combined iinmunodeficiency mice with listeriosis, and interleukin 10 is a physiologic^ antagonist. Proc. Nntl. Acnd. Sci. USA 90, 3725-3729. Hunter, C. A,, Chizzonite, R., and Remington. J. S. (199.5).IL-lP is required for IL12 to induce production of IFN-y by N K cells: A role for IL-lP in the T cellindependent mechanism of resistance against intracellular pathogens. J. Irnrnzmol. 155,4347-4354. Hunter, C. A,, Ellis-Neyer, L., Gabriel, K. E., Kennedy, M. K., Grahstein, K. H., Linsley, P. S., and Remington. J. S. (1997). The role of the CD28A37 interaction in the regulation of NK cell responses during infection with Toxoplmnla gondii. J. Imn2i~nol. 158, 2285-2293. Gollob, J. A., Li, J., Reinherz, E. L., and Ritz, J. (1995). CD2 regulates responsiveness of activated T cells to interleukin 12.1. Esp. Med. 182, 721-731. Gollob, J. A,, Li, J., Kawasaki, II., Daley, J. F., Groves, C., Reinherz, E. L., and Ritz. J. (1996). Molecular interaction between CD58 and CD2 counter-receptors mediates the ability of inonocytes to augment T cell activation by IL-12.J. Imnicinol. 157, 18861893. Meuer, S. C . , Hussey, R. E., Fabli. M.. Fox. D., Acuto, O., Fitzgerald, K. A,, Hodgdon, J. C., Protentis, J. P., Schlossnian, S. F., and Reinherz, E. L. (1984). An alternative pathway of' T-cell activation: A finictional role for the 50 Kd T11 sheep erythrocyte receptor protein. Cell. 36, 897-906. Hunter, C. A,, Berinudez. L, Beernick, H., Waegell, W., and Reniington, J. S. (1995). Transforming growth factor-beta inhibits interleukin-12-induced production of interferon-gamma by natural killer cells: A role for transfonning growth factor-beta in the regulation of T cell-indepenclent resistance to Torop/nsmn gondii. Ezir. 1.I n m z t nol. 25, 994-1000. Bellone, G., Aste-Amezaga, M., Tiinchieri, G., and Rodeck, U. (1995). Replation of natiiral killer cell functions by TGF-P. J . Imniinol. 155, 1066- 1073. Okamura, H., Nagata, K., Koniatsu, T., Tanirnoto, T., Nitkata, Y., Tanabe, F., Akita, K., Torigoe, K., Okura, T., and Fukuda, S. (1995). A novel costimuhtory factor for gainma interferon induction found in the livers of mice causes endotoxic shock. Infect. fmnLt4n. 63, 3966-3972. Okamura, H., Tsntsi, H.. Koinatsu, T., Yutsudo, M., Hakura, A,, Tanirnoto, T., Torigoe, K., Okura, T., Nukada, Y.. and Hattori, K. (1995). Cloning of a new cytokine that induces IFN-y production by T cells. Nature 378, 88-91. Ushio, S., Namba, M., Okura, T.. Hattori, K.. Nukada, Y., Akita, K., Tanabe, F., Konishi, K., Micallef, M., Fiijii, M., Torigoe, K., Tluiiinoto, T., Fukuda, S., Ideda, M., Okamura, H., and Kurimoto. M. (1996).Cloning of the cDNA for hi.iirian IFN-gammaindiicing factor, expression in Eschericltia coli, and studies on the biologic activities of the protein. J . Imniunol. 156, 4274-4279. Stoll, S., Miiller, G., Kurimoto, M., Saloga, J., Tanimoto, T., Yaniauchi, H., Okamura, H., Knop, J., and Enk, A. H. (1997). Production of IL-18 (IFN-y-inducing factor) messenger RNA and functional protein by inurine kemtinocytes. J . Zmn~zmol.159, 298-302. Bazan, J. F., Timans, J. C., and Kastelein, R. A. (1997). A newly defined intedenkinl? Nature 379, 591. Gn, Y., Knida, K., Tsutsui, H., KLI, G., Hsiao, K., Flemming, M. A,. Hayashi, N., Higashino, K., Okainura, H., Nakanishi, K., Kurimoto, M., Tanimoto, T., Flavell, R. A,, Sato, V., Harding, hl. W.. Livingston. D. J., and Su, M. S. (1997). Activation of interferon-y inducing factor mediated by interleukin- 1P converting enzyme..Science 275,206-209.
2 12
GIORGIO TRINCHIERI
271. Chayur, T., Banerjee, S., Hugunin, M., Butler, D., Herzog, L., Carter, A,, Quintal, L., Sekut, L., Talanian, R., Paskind, M., Wong, W., Kamen, R., Tracey, D., and Allen, H. (1997). Caspase-1 processes IFN-gamma-inducing factor and regulates LPSinduced IFN-gamma production. Nature 386, 619-623. 272. Micallef, M. J., Ohtsuki, T., Kohno, K., Tanabe, F., Ushio, S., Namba, M., Tanimoto, T., Torigoe, K., Fujii, M., Ikeda, M., Fukuda, S., and Kurimoto, M. (1996). Interferongamma-inducing factor enhances T helper 1cytokine production by stimulated human T cells: Synergism with interleukin-12 for interferon-gamma production. Eur J. Zmmunol. 26, 1647-1651. 273. Ahn, H., 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-y-inducing factor in enhanced production of IFN-7. J . Inmunot. 159,2125-2131. 274. Hunter, C. A., Timans, J., Pisacane, P., Menon, S., Cai, C., Walker, W., Aste-Amezaga, M., Chizzonite, R., Bazar, J. F., and Kastelein, R. A. (1997). Comparison of the effect of interleukin-la (IL-la), IL-lP and IGIF (IL-ly) on the production of interferongamma by NK cells. Eur. /. ZmwnoZ. 27, 2787-2792. 275. Kohno, K., Kataoka, J., Ohtsuki, T., Suemoto, Y., Okamoto, I., Usui, M., Ideda, M., and Kurimoto, M. (1997). IFN-gamma-inducing factor (IGIF)is a costimulatory factor on the activation of Thl but not Th2 cells and exerts is effect independently of IL12. J. Zwimunol. 158, 1541-1550. 276. Robinson, D., Shibuya, K., Mui, A,, Zonin, F., Murphy, E., Sana, T., Hartley, S., Menon, S., Kastelein, R., Bazan, F., and O’Carra, A. (1997). ICIF does not drive Thl development, but synergizes with IL-12 for interferon-? production and activates IRAK and NF-kB. Zmmunity 7, 571-581. 277. Matsumoto, S., Tsuji-Takayama, K., Aizawa, Y., Koide, K., Takeuchi, M., Ohta, T., and Kurimoto, M. (1997). Interleukin-18 activates NF-kappaB in murine T helper type 1 cells. Biochem. Biophys. Res. Cornniun. 234, 454-457. 278. Gately, M. K., Warner, R. R., Honasoge, S., Carvajal, D. M., Faherty, D. A,, Connaughton, S. E., Anderson, T. D., Sarmiento, U., Hubbard, B. R., and Murphy, M. (1994). Administration of recombinant IL-12 to normal mice enhances cytolytic lymphocyte activityand induces production of IFN-y in vivo. lnt. Zmnmnol.6,157-167. 279. Magram, J., Connaughton, S. E., Warner, R. R., Carvajal, D. M., Wu, C. Y., Ferrante, J., Stewart, C., Sarmiento, U., Faherty, D. A., and Cately, M. K. (1996).IL-12-deficient mice are defective in IFN gamma production and type 1cytokine responses. Zmmunity 4,471-481. 280. Magram, J., Sfarra, J., Connaughton, S., Faherty, D., Warner, R., Carvajal, D., Wu, C. Y., Stewart, C., Sanniento, U., and Gately, M. K. (1996). IL-12-deficient mice are defective but not devoid of type 1 cytokine responses. Ann. N.Y. Acad. Sci. 795, 60-70. 281. Perussia, B., Chan, S., D’Andrea, A,, Tsuji, K., Santoli, D., Pospisil, M., Young, D., Wolf, S., and Trinchieri, C . (1992). Natural killer cell stimulatory factor or IL-12 has differential effects on the proliferation of TCRcup+,TCRyS+ T lymphocytes and NK cells. 1. Zmniunol. 149, 3495-3502. 282. Valiante, N. M., and Trinchieri, C. (1993). Identification of a novel signal transduction surface molecule on human cytotoxic lymphocytes.J. Exp. Med. 178, 1397-1406. 283. Naume, B., Johnsen, A., Espevik, T., and Sundan, A. (1993). Gene expression and secretion of cytokines and cytokine receptors from highly purified CD56+natural killer cells stimulated with interleukin-2, interleukin-7 and interleukin-12. Eur. J. Immunol. 23, 1831-1838.
INTEKLEUKIN-12
2 13
284. Meyaard, L., Hovenkamp, E., Otto, S. A,, and Miedema, F. (1996). IL-12-induced IL-10 production by human T cells as a negative feedback for IL-12-induced immune responses. J. Inziiitmol. 156, 2776-2782. 285. Windhagen, A., Anderson, D. E., Carrizosa, A,, Williams, R. E., and Hafler, D. A. (1996). IL-12 induces human T cells secreting IL-10 with IFN-7. J. Immunol. 157 1127-1131. 286. Marshall, J., Secrist, H., DeKruyff, R. H., Wolf, S . F., and Umetsu, D. T. (1995). IL12 inhibits the production of IL-4 and IL-10 in allergen-specific human CD4+ T lymphocytes. J. Immuncd 155, 111- 117. 287. Schmitt, E., Hoehn, P., Germanti, T., and Rude, E. (1994). Differential effects of interleukin-12 on the development of naive mouse CD4' T cells. E u r J. fmmunol. 24,343-347. 288. Wu, C. Y.. Demeure, C. E.. Gately, M., Podlaski, F., Yssel, H., Kiniwa, M., and Delespesse, G. (1994). In vitro maturation of human neonatal CD4 T lymphocytes. I. Induction of IL-4-producing cells after long-term culture in the presence of IL-4 plus either IL-2 or IL-12. J, Immunol. 152, 1141-1153. 289. Jeannin, P.. Delneste, Y., Seveso, M., Life, P., and Bonnefoy, J. (1996).IL-12 synergizes with IL-2 and other stimuli in inducing IL-10 production by human T cells.J . Immunol. 156, 3159-3165. 290. Daftarian, P. M., Kumar, A,, Kryworuchko, M., and Diaz-Mitoma, F. (1996). IL-10 production is enhanced in human T cells by IL-12 and IL-6 and in inonocytes by tumor necrosis factor-a. J. lmmz~nol.157, 12-20. 291. Peng, X., Kasran, A,, and Ceuppens, J. L. (1997). Interleukin 12 and B7/CD28 interaction synergistically upregulate interleukin 10 production by human T cells. Cytokine 9,499-506. 292. Gazzinelli, R. T., Wysocka, M., Hieny, S., Scharton-Kersten, T., Cheever, A., Kiihn, R., Muller, W., Trinchieri, G., and Sher, A. (1996). In the absence of endogenous IL10, mice acutely infected with Toxophmn gondiii succumb to a lethal immune response dependent on CD4+ T cells and accompanied by overproduction of IL-12, IFN-7 and TNF-a. J. lnimunol. 157, 798-805. 293. Hunter, C. A., Ellis-Neyes, I,. A,, Slifer, T., Kanaly, S., Grunig, G., Fort, M., Rennick, D., and Araujo, F. G. (1997).IL-10 is required to prevent immune hyperactivity during infection with Trypanosonla crnti. J. Immttnol. 158, 3311-3316. 294. Chouaib, S., Chehimi, J., Bani, L., Genetet, N., Tursz, T., Gay, F., Trinchieri, G., and Mami-Chouaib, F. (1994). Interleukin 12 induces the differentiation of major histocompatibility complex class I-primed cytotoxic T-lymphocyte precursors into allospecific cytotoxic effectors. Proc. Nntl. Acnd. Sci. USA 91, 12659-12663. 295. Gately, M. K., Desai, B. B., Wolitzky, A. G., Quinn, P. M., D y e r , C. M., Podlaski, F. J,, Familletti, P. C., Sinigaglia, F., Chizzonite, R., Gubler, U., and Stern, A. S. (1991). Regulation of human lymphocyte proliferation by a heterodimeric cytokine, IL-12 (cytotoxic lymphocyte maturation factor). /. ImmunoZ. 147, 874-882. 296. Bertagnolli, M. M., Lin, B-Y., Young, D., and Herrmann, S. H. (1992). IL-12 augments antigen-dependent proliferation of activated T lymphocytes. J. lmnzunol. 149, 37783783. 297. Bertagnolli, M. M., Yang, L., Herrmann, S. H., and Kirkinan, R. L. (1994). Evidence that rapamycin inhibits interleukin- 12-induced proliferation of activated T lymphocytes. Trnizsplnntntion 58, 1091-1096. 298. Valiante, N. M., Rengaraju, M., and Trinchieri, G. (1992). Role of the production of natural killer cell stimulatory factor (NKSF/IL-12) in the ability of B cell lines to stimulate T and NK cell proliferation. Cell. Irnmzrnol. 145 187-198.
214
GIORGIO TRINCHIERI
299. Yanagida, T., Kato, T., Igarashi, O., Inoue, T., and Nariuchi, H. (1994). Second signal activity of IL-12 on the proliferation and IL-2R expression of T helper cell-1 clone. J. Inimunol. 152, 4919-4928. 300. Ueta, C., Kawasumi, Ha., Fujiwara, H., Miyagawa, T., Kida, H., Ohmoto, Y., Kishimoto, S., and Tsuyuguchi, I. (1996). Interleukin-12 activates human gamma delta T cells: Synergistic effect of tumor necrosis factor-alpha. Eur. J. Imrnunol. 26, 3066-3073. 301. Robertson, M. J., Soiffer, R. J., Wolf', S. F., Manley, T. J., Donahue, C., Young, D., Herrmann, S. H., and Ritz, J. (1992). Response of human natural killer (NK) cells to NK cell stimulatory factor (NKSF): Cytolytic activity and proliferation of NK cells are differentially regulated by NKSF. J. Exp. Med. 175, 779-788. 302. Mehrotra, P. T., Wu, D., Crim, J. A,, Mostowski, H. S., and Siegel, J. P. (1993). Effects of IL-12 on the generation of cytotoxic activity in human CD8+ T lymphocytes. J. Immunol. 151,2444-2452. 303. Mehrotra, P. T., Grant, A. J., and Siegel, J. P. (1995). Synergistic effects of IL-7 and IL-12 on human T cell activation. J. Zrnmunol. 154, 5093-5102. 304. Xu, H., Rizzo, L. V., and Caspi, R. R. (1995). IL-12 induces growth of the IL-4dependent CT4S line and has a synergistic effect on IL-4-induced CT4S proliferation. J. Imrnunol. Methods 181, 245-251. 305. Pardoux, C., Asselin-Paturel, C., Chehimi, J., Gay, F., Mami-Chouaib, F., and Chouaib, S. (1997). Functional interaction between TGF-/3 and IL-12 in human primary allogeneic cytotoxicity and proliferative response. J. Zmniunol. 158, 136-143. 306. Mama, S., Toyo-aka, K., Oh-hora, M., Tai, X., Iwata, H., Takenaka, H., Yamada, S., Ono, S., Hamaoka, T., Kobayashi, M., Wysocka, M., Trinchieri, G., and Fujiwara, H. (1996). IL-12 produced by antigen-presenting cells induces IL-2-independent proliferation of T helper cell clones. J, Zrnviunol. 156, 1748-1755. 307. Quill, H., Bhandoola, A,, Trinchieri, G., Haluskey, J., and Peritt, D. (1994). Induction of IL-12 responsiveness is impaired in anergic T lymphocytes.J. Exp.Med. 179,10651070. 308. Becker, J. C., and Brocker, E. B. (1994). Prevention of anergy induction in cloned T cells by interleukin 12. Exp. Demmtol. 3, 283-289. 309. Landay, A. L., Clerici, M., Hashemi, F., Kessler, H., Berzofsky, J. A., and Shearer, G. M. (1996). In vitro restoration of T cell immune function in human immunodeficiency virus-positive persons: Effects of interleukin (IL)-12 or anti-IL-10. J. Infect. Dis. 173, 1085-1091. 310. Radrizzani, M., Accornero, P., Amidei, A., Aiello, A,, Delia, D., Kurrle, R., and Colombo, M. P. (1995). IL-12 inhibits apoptosis induced in a human Thl clone by gp120/ CD4 cross-linking and CD3TCR activation or by IL-2 deprivation. Cell. Immun. 161, 14-21. 311. Clerici, M., Sarin, A., Coffrnan, R. L., Wynn, T. A., Blatt, S. P., Hendrix, C. W., Wolf, S. F., Shearer, G. M., and Henkart, P. A. (1994). Type l/type 2 cytokine modulation of T cell programmed cell death as a model for HIV pathogenesis. Proc. Nutl. Acud. Sci. USA 91, 11811-11815. 312. Estaquier, J., Idziorek, T., Zou, W., Emilie, D., Farber, C. M., Bourez, J. M., and Ameisen, J. C. (1995). T helper type 1T helper type 2 cytokines and T cell death: Preventive effect of interleukin 12 on activation-induced and CD95 (FAS/APO-1)mediated apoptosis of CD4f T cells from human immunodeficiency virus-infected persons. J. Exp. Med. 182, 1759-1767. 313. Clerici, M., Fusi, M. L., Ruzzante, S., Piconi, S., Biasin, M., Arienti, D., Trabattoni, D., and Villa, M. L. (1997). Type 1 and type 2 cytokines in HIV infection: A possible role in apoptosis and disease progression. Ann. Med. 29, 185-188.
INTERLEUKIN-12
21s
314. Amiant, M., Delespesse, G., and Sarfati, M. (1995). IL-2 and IL-7 but not IL-12 protect natural killer cells from death by apoptosis and up-regulate bcl-2 expression. Inzr~iunology85, 331-337. 315. Azzoni, L., Kanakaraj, P., Zatsepina, O., and Penissia, B. (1996).IL-12-induced activation of NK and T cells occurs in the absence of immediate-early activation gene expression. J . Immtrnol. 157, 3235-3241. 316. Chehimi, J., Valiante, N. M., D'Andrea, A,, Rengaraju, M., Rosado, Z., Kobayashi, M., Perussia, B., Wolf, S., Starr, S . E., and Trinchieri, G. (1993). Enhancing effect of natural killer cell stirnulatory factor (NKSF/IL-12) on cell-mediated cytotoxicityagainst tunlor-derived and virus-infected cells. Eur. J. Immunol. 23, 1826- 1830. 317. Bennett, I. M., Zatsepina, O., Zamai, L., Azzoni, L., Mikheeva, T., and Penissia, B. (1996). Definition of a natural killer NKR-PlA'/CD~56~/CD16functionally immature lruinan N K cell subset that differentiates in vitro in the presence of interleukin 12. J . Exp. Med. 184, 1845-18.56. 318. Nauine, B., Gately, M., and Espevik, T. (1992).A comparative study of IL-12 (cytotoxic lymphocyte maturation factor)-, IL-2-, and IL-7-induced effects on immunomagnetically purified CD56+ NK cells. J . Im.mnmol. 148, 2429-2436. 319. Kennedy, R. C., Zhon, E.-M., Lanford, R. E., Chanh, T. C., and Bona, C. A. (1987). Possible role of anti-idiotypic antibodies in the induction of tumor immunity. J. Clin. Znoe,st. 80, 1217-1224. 320. Scharton-Kersten, T., Afonso, L. C. C., Wysocka, M., Trinchieri, G., and Scott, P. (1995). IL-12 is required for N K cell activation and subsequent Thl cell development in experimental leishmaniasis./. lrnniunol. 154, 5320-5330. 321, Brunda, M. J.. Taramelli, D., Holden, H. T., and Varesio, L. (1983). Suppression of in oitro maintenance and interferon-mediated augmentation of natural killer cell activity by adherent peritoneal cells from normal mice. J . Zmnzunol. 130, 1974-1979. 322. Allavena, P., Paganin, C., Zhou, D., Bianchi, G., Sozzani, S., and Mantovani, A. (1994).IL-12 is chemotactic for N K cells and stiniulates their interaction with vascular endothelium. Blood 84, 2261-2268. 323. Rabinowich, H., Herberman, R. B., and Whiteside, T. L. (1993). Differential effects of IL-12 and IL-2 on expression and function of cellular adhesion molecules on purified human natural killer cells. Cell. Znzinztnol. 152, 481-498. 324. Jewett, A,, and Bonavida, B. (1994). Activation of the human immature natural killer cell subset by ILL2 and its regulation by endogenous TNF-alpha and IFN-gamma secretion. Cell. Itmnunol. 154, 273-286. 32,5. Bonnema, J. D., Rivlin. K. A,, Ting, A. T., Schoon, R. A,, Abraham, R. T., and Leibson, P. J. (1994). Cytokine-enhance N K cell-mediated cytotoxicity: Positive modulatory effects of IL-2 and IL-12 on stimulus-dependent granule exocytosis. J . Imntunol. 152, 2098-2104. 326. Salcedo, T. W., Azzoni, L., Wolf: S. F., and Perussia, B. (1993). Modulation of perforin and granzyme messenger RNA expression in human natural killer cells. J . Zmmztnol. 151, 2511-2520. 327. Cesano, A,, Visonneau, S., Clark, S . C., and Santoli, D. (1993). Cellular and molecular mechanisms of activation of MHC nonrestricted cytotoxic cells by IL-12. J . Imnmnol. 151,2943-2957. 328. DeBlaker-Hohe, D. F., Yamanchi, A,, Yu, C. R., Horvath-Arcidiacono, J. A., and Bloom, E. T. (1995). IL-12 synergizes with 11,-2 to induce lymphokine-activated cytotoxicity and perforin and granzyine gene expression in fresh human N K cells. Cell. Iinmunol. 165, 33-43.
216
GlORGlO TRINCHIERI
329. Gately, M. K., Wolitzky, A. G., Quinn, P. M.. and Chizzonite, R. (1992). Regulation of human cytolyt~clymphocyte responses by interleukin-12. Cell. Im~nunol.143,127-142. 330. Bloom, E. T., and Horvath, J. A. (1994). Cellular and molecular mechanisms of the IL-12-induced increase in allospecific murine cytolytic T cell activity: Implications for the age-related decline in CTL. 1.Immunol. 152, 4242-4254. 331. Biron, C. A,, and Gazzinelli, R. T. (1995) Effects of IL-12 on immune responses to microbial infections: A key mediator in regulating disease outcome. Cum. Opin. Immunol. 7, 48.5-496. 332. Monteiro, J., and Trinchieri, G . (1996). Interleukin 12 and its role in viral infection. In “Options for the Control of Influenza HI”,(L. E. Brown, A. W. Hsmpson, and R . G . Webster, eds.), pp. 180-182. Elsevier, Amsterdam, The Netherlands. 333. Bhardwaj, N., Seder, R. A,, Reddy, A., and Feldman, M. V. (1996).IL-12 in conjunction with dendritic cells enhances antiviral CD8+ CTL responses in vitro. I. Clin. Invest. 98, 715-722. 334. O’Toole. M., Wooters. J., Brown, E.. Swiniarski,H., Cull. G . , Leger, L., and Hemnann, S. (1996). Interleukin-12 as an adjuvant in peptide vaccines. Ann. N.Y. Acad. Sci. 795, 379-38 1. 335. Noguchi, Y., Richards, E. C., Chen, Y., and Old, L. J. (1995). Influence of IL-12 on p53 peptide vaccination against established Meth A sarcoma. Proc. Natl. Acad. Sci. USA 92,2219-2223. 336. Gabrilovich, D. I., Cunningham, H. T., and Carbone, D. P. (1996). IL-12 and mutant P53 peptiide-pulsed dendritic cells for specific irnrnunotherapy of cancer. 1. Immunothsr. Emph. Tumor lmmitnol. 19, 414-418. 337. Bendelac, A,. Rivera, M. N., Park, S. H., and Roark, J. H. (1997). Mouse CD1-specific N K 1 T cells: development, specificity. and function. Annu. Reu. Irnmunol. 15,535-562. 338. Okada, E., Sasaki, S., Ishii, N., Aoki, I., Yasuda, T., Nishioka, K., Fukushima, J., Miyamaki, J., Wahren, B., and Okuda, K. (1997). Intranasal immunization of a DNA vaccine with IL-12- and granulocyte-macrophagecolony-stimulatingfactor (GM-CSF)expressing plasmids in liposomes induces strong mumsal and cell-mediated immune responses against HIV-1 antigens. J. Imniunol. 159, 3638-3647. 339. Zeng, Z., Castafio, A. R.,Segelke, B. W., Stura, E. A., Peterson, P. A., and Wilson, I. A. (1997).Crystal structure of mouse CD1: An MHC-like fold with a large hydrophobic binding groove. Science 277, 339-345. 340. Beckrnan. E. M., Porcelli, S. A,, Morita, C. T., Behar, S. M., Furlong, S. T., and Brenner, M. B. (1994).Recognition of a lipid antigen by CD1-restricted alpha beta+ T cells. Nature 372, 691-694. 341. Davodeau, F., Peyrat, M. A,, Necker, A,, Dorninici. R., Blanchard, F., Leget, C., Gaschet, J., Costa, P., Jacques, Y., Godard, A,, Vie, H., Poggi, A,, Romagne, F., and Bonneville, M. (1997). Close phenotypic and functional similarities between human and murine alpha-beta T cells expressing invariant TCR alpha-chains. 1. Immunol. 158 5W3-5611. 342. Hashimoto, W., Takeda, K., Anzai, R., Ogasawara, K., Sakihara, H., Sugiura, K.. Seki. S., and Kumagai, K. (1995). Cytotoxic NK1.l Ag’ cup T cells with intermediate TCR induced in the liver of mice by IL-12.1. Immunol. 154,433-4340. 343. Takeda, L., Seki, S., Ogasawara, K., Anzai, R., Hashirnoto, W., Sugiura, K., Takahashi, M., Satoh, M., and Kumagai, K. (1996). Liver NK1.1’ CD4+ cup cell activated by IL12 as a major effector in inhibition of experimental tumor metastasis. J. Immunol. 156, 3366-3373. 344. Anzai, R., Seki, S., Ogasawara, K., Hashimoto, W.,Sugiura, K., Sato, M., Kumagai. K.. and Takeda, K. (1996) Interleukin-12 induces cytotoxic NK1+ alpha beta T cells in the lungs of euthymic and thymic mice. Immunology 88, 82-89.
INTERLEUKIN-12
217
345. Takahashi, M., Ogasawara, K., Takeda, K., Hashimoto, W., Sakihara, H., Kumagai, K., Anzai, R., Satoh, M., and Seki, S. (1996). LPS induces NK1.1+ cup T cells with potent cytotoxicity in the liver of mice via production of IL-12 from Kupffer cells. J . hnlnunol. 156, 2436-2442. 346. Sato, N., Yahata, T., Santa, K., Ohta, A., Ohmi, Y., Habu, S., and Nishimura, T. (1996). Functional characterization of NKl.l + Ly-GC+ cells. Immunol. Lett. 54, 5-9. 347. Chen, H., and Paul, W. E. (1997). Cultured NK1.1' CD4+ T cells produce large amounts of IL-4 and IFN-y upon activation by anti-CD3 or CD1. J. Inainunol. 159,2240-2249. 348. Emoto, M., Emoto, Y., and Kaufniann, S. H. (1997). TCR-mediated target cell lysis by CD4+NK1+ liver T lymphocytes. Int. Iitatnunol. 9, 563-571. 349. Emoto, M., Emoto, Y., and Kaufinann, S. H. (1997). Bacille Calmette Gukrin and interleukin-12 down-modulate interleulan-4-producing CD4+ NK1+ T lymphocytes. Eur. J . Imnmunol. 27, 183-188. 350. Satoh, M., Seki, S., Hashimoto, W., Ogasawara, K., Kobayashi, T., Kumagai, K., Matsuno, S., and Takeda, K. (1996) Cytotoxic y6 or cup T cells with natural killer cell marker, CD56, induced from human peripheral blood lymphocytes by a combination of IL-12 and IL-2. ]. Immunol. 157, 3886-3892. 351. 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 viva and in vitro: IL-2 and IL-4 are required for in vitro generation of IL-4-producing ce1ls.J. Exp. Med. 172,921-929. 3,52. Swain, S. L., Weinberg, A. D., English, M., and Huston, G. (1990). IL-4 directs the development of Th2-like helper effectors. ], Inanaunol. 145, 3796-3806. 353. Seder, R. A,, Paul, W. E., Davis, M. M., and Fezekas de St Groth, B. (1992). The presence of interleukin 4 during in vitro priming determines the lymphokine-producing potentialofCD4+ Tcells from Tcell receptor transgenic mice.]. Exp. Med. 176,10911098. 354. Trinchieri, G. (1993). Interleukin-12 and its role in the generation of Till cells. Imnn~unol. Toduy 14, 335-338. 355. Trinchieri, G., and Scott, P. (1995). Interleukin-12: A proinflammatory cytokine with iinmunoregulatory functions. Res. Irnnzunol. 146, 423-431. 356. Macatonia, S. E., Hsieh, C.. Murphy, K. M., and O'Garra, A. (1993). Dendritic cells and macrophages are required for Thl development of CD4+ T cells from cuP-TCR transgenic mice: IL-12 substitution for macrophages to stimulate IFN-y production is IFN-y-dependent. Int. Imnnauunol. 5, 1119-1 128. 357. Trinchieri, G. (1995). Interleukin- 12: A proinflammatory cytokine with immunoregulatory functions that bridge innate resistance and antigen-specific adaptive iIrimuIiity. Annu. Rev. I ~ J L ~ u 13, ~ o 251-276. ~. 358. Seder, R., A,, Gazzinek, R., Slier, A,, and Paul, W. E. (1993). IL-12 acts directly on CD4' T cells to enhance priming for IFN-y production and diminishes IL-4 inhibition of such priming. Proc. Ned. Acad. Sci. USA 90, 10188-10192. 359. Germann, T., Jin, S., Mattner, F., and Rude, E. (1991). Components of an antigen-/ T cell receptor-independent pathway of lyinphokine production. Eur. J. Imrnunol. 21, 1857-1861. 360. Schinitt, E., Hoehn, P., Huels, C., Goedert, S., Palm, N., Rude, E., and Germann, T. (1994). T helper type 1 developinent of naive CD4+ T cells requires the coordinate action of interleukin-12 and interferon-y and is inhibited by transforming growth factor-p. Eur. J , Imrnunol. 24, 793-798. 361. Bradley, L. M., Yoshimoto, K., and Swain, S. L. (1995). The cytokines IL-4, IFN-y, and IL-12 regulate the development of subsets of memory effector helper T cells in vitro. J. Immntcnol. 155, 1713-1724.
218
GIORGIO TRINCHIERI
362. Dighe, A. S., Campbell, D., Hsieh, C. S., Clarke, S., Greaves, D. R., Gordon, S., Murphy, K. M., and Schreiber, R. D. (1995). Tissue-specific targeting of cytokine unresponsiveness in transgenic mice. Immunity 3, 657-666. 363. Wenner, C. A,, Guler, M. L., Macatonia, S. E., O’Garra, A., and Murphy, K. M. (1996). Role of IFN-y and IFN-a in IL-12-induced T helper cell-1 development. J. Immunol. 156, 1442-1447. 364. Bradley, L. M., Dalton, D. K., and Croft, M. (1996). A direct role for IFN-y in regulation of Thl cell development. J. lmniunol. 157, 1350-1358. 365. Nakamura, T., Lee, R. K., Nam, S. Y., Podack, E. R., Bottomly, K., and Flavell, R. A. (1997). Roles of IL-4 and IFN-gamma in stabilizing the T helper cell type 1 and 2 phenotype. J. Immunol. 158, 2648-2653. 366. Gieni, R. S., Fang, Y., Trinchieri, G., Umetsu, D. T., and DeKrnyff, R. H. (1996). Differential production of IL-12 in BALB/c and DBM2 mice controls IL-4 versus IFN-y synthesis in primed CD4 lymphocytes. Int. Irnmunol. 8, 1511-1520. 367. Gorham, J. D., Guler, J. L., Steen, R. G., Mackey, A. J., Ddy, M. J., Frederick, K., Dietrich, W. F., and Murphy, K. M. (1996). Genetic mapping of a murine locus controlling development of T helper 1/T helper 2 type responses. Proc. Nutl. Acad. Sci. USA 93, 12467-12472. 368. Perez, V. L., Lederer, J. A., Lichtman, A. H., and Abbas, A. K. (1995). Stability of Thl and Th2 populations. lnt. Immunol. 7, 869-875. 369. 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. 370. 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 popu1ations.J. Exp. Med. 182, 1357-1367. 371. Nakamura, T., Kamogawa, Y., Bottomly, K., and Flavell, R. A. (1997). Polarization of IL-4- and IFN-gamma-producing CD4+ T cells following activation of naive CD4+ T cells. J. Immunol. 158, 1085-1094. 372. Kamogawa, Y., Minasi, L. E., Carding, S. R., Bottomly, K., and Flavell, R. A. (1993). the relationship of IL-4- and IFN-y-producing T cells studied by lineage ablation of IL-4-producing cells. Cell 75, 985-995. 373. Piccotti, J. R., Chan, S. Y., VanBuskirk, A. M., Eichwald, E. J., and Bishop, D. K. (1997).Are Th2 helper T lymphocytesbeneficial, deleterious, or irrelevant in promoting allograft survival? Transplantation 63, 619-624. 374. Seder, R. A. (1996). High-dose IL-2 and IL-15 enhance the in vitro priming of naive CD4+ cells for IFN-gamma but have differential effects on priming for IL-4. J. Immunol. 156, 2413-2422. 375. 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 gamma production is cytokine-dependent. Proc. Natl. Acad. Sci. USA 94, 3189-3194. 376. Palm, N., Germann, T., Goedert, S., Hoehn, P., Koelsch, S., Rude, E., and Schmitt, E. (1997). Co-development of naive CD4+ cells towards T helper type 1or T helper type 2 cells induced by a combination of IL-12 and IL-4. Immunobiology 196,475-484. 377. Gazzinelli, R. T., Wysocka, M., Hayashi, S., Denkers, E. Y., Hieny, S., Caspar, P., Trinchieri, G., and Sher, A. (1994). Parasite induced IL-12 stimulates early IFN-y synthesis and resistance during acute infection with Toxoplusniu gondii. J. lmniunol. 153,2533-2543. 378. Seder, R. A., Kelsall, B. L., and Jankovic, D. (1996). Differential roles for IL-12 in the maintenance of immune responses in infectious versus autoimmune disease. J. Immunol. 157, 2745-2748.
INTEHLE UKIN- 12
219
379. Trinchieri, G. (1989). BioloQ of natural killer cells. A h Zmniuriol. 47, 187-376. 380. Bancroft, G. J., Schreiber, R. D., Bosma, G. C., Bosma, M. J., and Unanue, E. R. (1987). A T cell-independent mechanism of niacrophage activation by interferongamma. ]. Inzmunol. 139, 1104-1107. 381. Not In, U., Wysocka, M., Hayashi, S., Denkers, E. Y., Hieny, S., Caspar, P., Trinchier, G., and Slier, A. (1994). Parasite induced IL-12 stimuIates early IFN-.)Isynthesis and resistance during acute infection with Toroplasma goiadii.]. Zmiiwnol. 153,2533-2543. 382. Scharton, T. M., and Scott, P. (1993). Natural killer cells are a source of interferon gamma that drives differentiation of CD4+ T cell subsets and induces early resistance to Leishmania major of mice. ]. Exp. Merl. 178, 567-577. 383. Afonso, L. C . C., Scharton, T. M., Vieira, L. Q., Wysocka, M., Trinchieri, G., and Scott, P. (1994). The adjuvant effect of interleukin-12 in a vaccine against Lebhinnnin nuqor. Science 263, 235-2:37. 384. Romagnani, S. (1992). Induction of T H l and TH2 responses: A key role for the ‘natural’ immune response? Znainztnol. Todny 13, 379-381. 385. Croft, M., Carter, L., Swain, S. L., and Dutton, R. W. (1994). Generation of polarized antigen-specific CD8 effector populations: Reciprocal action of interleukin ( IL)-4 and IL-12 in promoting type 2 versus type 1cytokine profiles.]. Erp. Med. 180,1715-1728. 386. Sad, S., Marcotte, R., and Mosmann, T. R. (1995). Cytokine-induced differentiation of precursor mouse CD8+ T cells into cytotoxic CD8+ T ceIIs secreting Thl or Th2 cytokines. lnainiitzity 2, 271-279. 387. Del Prete, G. F., De Carli, M., Ricci, M., and Romagnani, S. (1991). Helper activity for immunoglobulin synthesis of T helper type 1 (Thl) and Th2 human T cell clones: The help of Thl clones is limited by their cytolytic capacity.]. Exp. Med. 174,809-813. 388. Cronin, D. C., 11, Stack, R., and Fitch, F. W. (1995). IL-4-producing CD8+ T cell clones can provide B cell help. ]. Zmniunol. 154, 3118-3127. 389. Piccotti, J. R., Chan, S. Y., Li, K., Eichwald, E. J., and Bishop, D. K. (1997).Differential effects of IL-12 receptor blockade with IL-12 p40 hoinodimer on the induction of CD4+ and CD8’ IFN-y-producing cells. J . Inimunol. 158, 643-648. 390. Wu, C. Y., Demeure, C., Kiniwa, M., Gately. M., and Delespesse, G. (1993). IL-12 induces the production of IFN-gamma by neonatal human CD4 T cells. J. Ztrununol. 151, 1938-1949. 391. Parronchi, P., Mohapatra, S., Sampognaro, S., Giannaiini, L., Wahn, U., Chong, P., Mohapatra, S., Ma@, E., Renz, H., and Romagiani, S. (1996). Effects of interferonalpha on cytokine profile, T cell receptor repertoire and peptide reactivity on human allergen-specific T cells. Eur. 1. ltnmunol. 26, 697-703. 392. 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 (alpha and gamma) exert opposite regnlatory effects on the development of cytolytic potential by Thl or TI12 huinan T cell clones.]. Inimuiaol. 149, 2977-2983. 393. Ohshima, Y., and Delespesse, G. (1997).T cell-derived IL-4 and dendritic cell-derived IL- 12 regulate the lymphokine-producing phenotype of alloantigen-primed naive human CD4 T cells. J . Irnmnunol. 158, 629-636. 394, Palmer, E. M., and van Seventer, G. A. (1997). Human T helper cell differentiation is regulated by the combined action of cytokines and accessory cell-dependent costimulatory signals. 1. lmniunol. 158, 2654-2662. 395. Matsuoka, T., Kohrogi, H., Ando, M., Nishimura, Y., and Matsushita, S. (1996).Altered TCR ligands affect antigen-presenting cell responses: Up-regulation of IL-12 by an analogne peptide. ]. Imninunol. 157, 4837-4843.
220
GIORGIO TRINCHIERI
396. Yang, L., Byun, D., Demeure, C. E., Vezzio, N., and Delespesse, G. (1995). Default development of cloned human naive CD4 T cells into interleukin-4- and interleukin5-producing effector cells. Eur. J. Immunol. 25, 3517-3520. 397. Mingari, M. C., Maggi, E., Cambiaggi, A,, Annunziato, F., Schiavetti, F., Anetti, R., Moretta, L., and Romagnani, S. (1996). Development in vitro of human CD4+ thymocytes into functionally mature Th2 cells: Exogenous interleukin-12 is required for priming thymocytes to produce both Thl cytokines and interleukin-10. Eur. J. Iminunol. 26, 1083-1087. 398. Shu, U., Demeure, C. E., Byun, D. G., Podlaski, F., Stem, A. S., and Delespesse, G. (1994). Interleukin 12 exerts a differential effect on the maturation of neonatal and adult human CD45RO- CD4 T cells. J. Clin. Invest. 94, 1352-1358. 399. Byun, D. G., Demeure, C. E., Yang, L. P., Shu, U., Ishihara, H., Vezzio, N., Gately, M. K., and Delespesse, G. (1994). In vitro maturation of neonatal human CD8 T lymphocytes into IL-4- and IL-5-producing cells. J. Immunol. 153,4862-4871. 400. 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-level. J. Exp. Med. 184, 473-483. 401. Hilkens, C. M. U., Messer, G., Tesselaar, K., van Rietschoten, A. G. I., Kapsenberg, M. L., and Wierenga, E. A. (1996). Lack of IL-12 signaling in human allergen-specific Th2 cells. J. Immunol. 157, 4316-4321. 402. Yssel, H., Fader, S., de Vries, J. E., and de Waal Malefyt, R. (1994). IL-12 transiently induces IFN-gamma transcription and protein synthesis in human CD4+ allergenspecific Th2 T cell clones. Int. Immunol. 6, 1091-1096. 403. Swain, S. L., McKenzie, D. T., Weinberg, A. D., and Hancock, W. (1988).Characterization of T helper 1 and 2 cell subsets in normal mice: Helper T cells responsible for IL-4 and IL-5 production are present as precursors that require priming before they develop into lymphokine-secreting cells. J. Immunol. 141, 3445-3455. 404. Street, N. E., Schumacher, J. H., Fong, A. T., Bass, H., Fiorentino, D. F.., Leverah, J. A., and Mosmann, T. R. (1993). Heterogeneity of mouse helper T cells: Evidence from bulk cultures and limiting dilution cloning for precursors of Thl and Th2 cells. J. Immunol. 144, 1629-1639. 405. Gajewski, T. F., Joyce, J., and Fitch, F. W. (1989). Antiproliferative effect of IFN-y in immune regulation. 111. Differential selection of Thl and Th2 murine helper T lymphocyte clones using recombinant IL-2 and recombinant IFN-y. J. Immunol. 143, 15-22. 406. Del Prete, G. F., De Carli, M., Mastromauro, C., Biagiotti, R., Macchia, D., Falagiani, P., Ricci, M., and Romagnani, S. (1991). Purified protein derivative of Mycobucterium tuberculosis and excretory-secretoryantigen(s)of Toxocuru cunis expand in vitro human T cells with stable and opposite (type 1T helper or type 2 T helper) profile of cytokine production. J. Clin. Invest. 88, 346-350. 407. Gerosa, F., and Trinchieri, G. (1994). Mechanisms of T helper cell differentiation induced by interleukin-12. In “Cytokines: Basic Principles and Practical Applications” (S. Romagnani, G. Del Prete, and A. K. Abbas, eds.), pp. 251-263, Ares-Serono Symposia Publications, Rome, Italy. 408. DeKruyff, R. H., Fang, Y., Wolf, S. F., and Umetsu, D. T. (1995). IL-12 inhibits IL4 synthesis in keyhole limpet hemocyanin-primed CD4+ T ceIIs through an effect on antigen-presenting cells. J . Immunol. 154, 2578-2587. 409. Wynn, T. A., Jankovic, D., Hieny, S., Zioncheck, K., Jardieu, P., Cheever, A. W., and Sher, A. (1995). IL-12 exacerbates rather than suppresses T helper 2-dependent pathology in the absence of endogenous IFN-7.J. Immunol. 154,3999-4009.
INTERLEUKIN-12
22 1
410. Van der Pouw-Kraan, T., Van Kooten, C., Rensuik. I., and Aarden, L. (1992). Interleu-
411.
412.
413.
414.
415.
416.
417.
418.
419.
420.
421.
422.
423.
424. 425.
kin (1L)-4production by human T cells: Differential regulation of IL-4 vs IL-2 production. Eur. 1. Immunol. 22, 1237-1241. Quelle, F. W., Shimoda, K., Thierfelder, W., Fischer, C., Kim, A,, Ruben, S. M., Cleveland, J. L., Pierce, J. H., Keegan, A. D., and Nelms, K. (1995).Cloning of inurine Stat6 and hninan Stat6, Stat proteins that are tyrosine phosphorylated in responses to IL-4 and IL-3 but are not required for mitogenesis. Mol. Cell. Biol. 15, 3336-3343. Young, H. A,, Ghosh, P., Ye, J., Lederer, J., Idchtnian, A,, Gerard, J. R., Penix, L., Wilson, C. B., Melvin, A. J., McGurn, M. E., Lewis, D. B., and Taub, D. D. (1994). Differentiation of the T helper phenotypes by analysis of the methylation state of the IFN--y gene. J. Immunol. 153, 3603-3610. Kiniwa, M., Gately, M., Gubler, U., Chizzonite, R., Fargeas, C., and Delespesse, G. (1992). Recoinbinant interleukin-12 suppresses the synthesis of iininniioglobuliii E by interleukin-4 stimulated human lymphocytes. J. Clin. Invest. 90, 262-266. Jelinek, D. F., and Braaten, J. K. (1995). Role of IL-12 in human B lymphocyte proliferation and differentiation. J . Inimunol. 154, 1606-1613. Spencer, N. F., and Daynes. R. A. (1997). IL-12 directly stimulates expression of IL10 by CD5+ B cells and IL-6 by both CDS+ and CD5- B cells: Possible involvement ~ . 745-754. in age-associated cytolane dysregulation. Int. ~ n m l u n o 9, Vogel, L. A,, Lester, T. L., Van Cleave, V. H., and Metzger, D. W. (1996). Inhibition of murine B1 lymphocytes by interleukin-12. Eur. J , Immunol. 26, 219-223. Velupillai, P., Sypek, J., and Ham, D. A. (1996). Interleukin-12 and -10 and gamma interferon regulate polyclonal and ligand-specificexpansion of inurine B-1 cells. Infect. Immun. 564,4557-4560. Jones, B. M. (1996). Effect of 12 neutralizing anti-cytokine antibodies on in vitro activation of B-cells. Interleukin-12 is required by Bla but not B2 cells. Scand. J. Immunol. 43, 64-72. Morris, S. C., Madden, K. B., Adainovicz, J. J., Cause, W. C., Hubbard, B. R., Gately, M. K., and Finkelman, F. D. (1994).Effects of IL-l2on invivo cytokine gene expression and Ig isotype selection. J . lmnwizol. 152, 1047-1056. McKnight, A. J., Zimmer, G. J., Fogelinan, I., Wolf, S. F., and Abbas, A. K. (1994). Effects of IL-12 on helper T cell-dependent immune responses in uiuo. J . Immunol. 152, 2172-2179. Buchanan, J. M., Vogel, L. A., Van Cleave, V. H., and Metzger, D. W. (1995). Interleukin 12 alters the isotype-restricted antibody response of mice to hen eggwhite lysozyme. Int. Immunol. 7, 1519-1528. Germann, T., Bongartz, M., Dlugonska, H., Hess, H., Schmitt, E., Kobe, L., Kolsch, E., Podlaski, F. J., Gately, M. K., and Rude, E. (1995). Interleukin-12 profoundly upregulates the synthesis of antigen-specific complement-fixing IgG2a, IgG2b and IgG3 antibody subclasses in vivo. Eur. 1.Iinmunol. 25, 823-829. Van Cleave, V., Wolf, S. F., Murray, K., Wiencis, A,, Ketchum, M., Bliss, J., Haire, T., Resmini, C., Maylor, R., and Alderman, E. (1995).Immunoglobulin isotype modulation after administration of IL-12. Ado. Exp. Mecl. B i d 383, 43-52. Gracie, J. A., and Bradley, J. A. (1996). Interleukin-12 induced interferon-gannnadependent switching of dloantibody subclass. Eur. J . Immunol. 26, 1217-1221. Kim, T. S., DeKmyff, R. H., Rupper, R., Maecker, H. T., Levy, S., and Umetsu, D. T. (1997). An ovalbumin-IL-12 fusion protein is more effective than ovalbumin plus free recombinant IL-12 inducing a T helper cell type-1-dominated immune response and inhibiting antigen-specific IgE production. J. Inalnlrnol. 158,4137-4144.
222
CIORGIO TRINCHIERI
426. Jankovic, D., Caspar, P., Zweig, M., Garcia-Moll, M., Showalter, S. D., Vogel, F. R., and Sher, A. (1997). Adsorption to aluminum hydroxide promotes the activity of IL12 as an adjuvant for antibody as well as type 1 cytokine responses to HIV-1 gp120. J . Zmmunol. 159, 2409-2417. 427. Villacres-Eriksson, M., Behboudi, S., Morgan, A. J., Trinchieri, G., and Morein, B. (1997).Immunomodulation by Quilluja saponuria adjuvant formulations: Zn vivo stimulation of interleukiu-12 and its effects on the antibody response. Cytukine 9, 73-82. 428. Pape, K. A., Khoruts, A,, Mondino, A,, and Jenkins, M. K. (1997). Inflammatory cytokines enhance the in vivo clonal expansion and differentiation of antigen-activated CD4+ T cells. J . Immunol. 159, 591-598. 429. Bliss, J., Van Cleave, V., Murray, K., Wiencis, A., Ketchum, M., Maylor, R., Haire, T., Resmini, C., Abbas, A. K., and Wolf, S. F. (1996). IL-12, as an adjuvant, promotes a T helper 1 cell, but does not suppress a T helper 2 cell recall response.]. Zmmunol. 156,887-894. 430. Germann, T., Guckes, S., Bongartz, M., Dlugonska, H., Schmitt, E., Kolbe, L., Kolsch, E., Podlaski, F. J., Gately, M. K., and Rude, E. (1995). Administration of IL-12 during ongoing immune responses fail to permanently suppress and can even enhance the synthesis of antigen-specific IgE. Znt. Zmmunol. 7 , 1649- 1657. 431. Rempel, J. D., Wang, M., and HayGIass, K. T. (1997). In vivo IL-12 administration induces profound but transient commitment to T helper cell type 1-associatedpatterns of cytokine and antibody production. J . Zmmunol. 159, 1490-1496. 432. Kim, J. J., Ayyavoo, V., Bagarazzi, M. L., Chattergoon, M. A,, Dang, K., Wang, B., Boyer, J. D., and Weiner, D. B. (1997). In vivo engineering of a cellular immune response by coadministration of IL-12 expression vector with a DNA immunogen. J. Zmmunol. 158, 816-826. 433. Iwasaki, A., Stiernholm, B. J., Chan, A. K., Berinstein, N. L., and Barber, B. H. (1997). Enhanced CTL responses mediated by plasmid DNA immunogens encoding costimulatory molecules and cytokines. J . Zmmunol. 158, 4591-4601. 434. Marinaro, M., Boyaka, P. N., Finkelman, F. D., Kiyono, H., Jackson, R. J,, Jirillo, E., and McGhee, J. R. (1997).Oral but not parental interleukin (IL)-12 redirects T helper 2 (Th2)-type responses to an oral vaccine without altering mucosal IgA. J. Exp. Med. 185,415-427. 435. Claessen, A. M., von Blomberg, B. M., De Groot, J., Wohers, D. A,, Kraal, G., and Scheper, R. J. (1996). Reversal of mucosal tolerance by subcutaneous administration of interleukin-12 at the site of attempted sensitization. Zmmunology 88, 363-367. 436. Maith, T., Strober, W., and Kelsall, B. L. (1996). High dose oral tolerance in avalbumin TCR-transgenic mice: Systemic neutralization of IL-12 augments TGF-beta secretion and T cell apoptosis. J. Zmmunol. 157, 2348-2357. 437. Muller, G., Saloga, J., Germann, T., Schuler, G., Knop, J., and Enk, A. H. (1995). IL12 as mediator and adjuvant for the induction of contact sensitivity in vivo. J. Zmmunol. 155,4661-4668. 438. Riemann, H., Schwarz, A., Grabbe, S., Aragane, Y., Luger, T. A,, Wysocka, M., Kubin, M., Trinchieri, G., and Schwarz, T. (1996). Neutralization of interleukin 12 in v i m prevents induction of contact hypersensitivity and induces hapten-specific tolerance. J . Zmnzunol. 156, 1799-1803. 439. Maguire, H. C., Jr. (1995). Murine recombinant interleukin-I2 increases the acquisition of allergic contact dermatitis in the mouse. Znt. Arch. Allergy Immunol. 106, 166-168. 440. Dilulio, N. A,, Xu, H., and Fairchild, R. L. (1996). Diversion of C D 4 f T cell development from regulatory T helper to effector T helper cells alters the contact hypersensitivity response. Eur. J. Immunol. 26, 2606-2612.
INTERLEUKIN-12
223
441. Asada, H., Linton, J., and Katz, S. I. (1997). Cytokine gene expression during the elicitation phase of contact sensitivity: regulation by endogenous 1L-4.1.bioest. Dentintol. 108, 406-411. 442. Kripte, M. L. (1990). Pliotoiininuiiolo~.Photoc.hrm. Yhotobiol. 52, 919-924. 443. Krenier, I. B., Hilkens, C. M. U.. Sylva-Steenland. R. M. R., Koomen, C. W., Kapsenberg, M. L., Bos, J. D.. aiid Teunissen, M. B. M. (1996). Reduced IL-12 production by nionocytes upon ultraviolet-B irradiation selectively limites activation of T helper1 cells. /. Zttsmunol. 157, 1913-1918. 444. Schinitt, D. A,, Owen-Schaub, L.. and Ullrich, S. E. (1995). Effect of IL-12 on immune suppression and suppressor cell induction by ultraviolet radiation./. Zmmnunol. 154, 5114-5120. 445. Schwarz, A,, Grabbe, S., Aragane, Y., Sandkuhl. K., Riemann, H., Luger, T. A., Kubin, M., Trinchieri, G., and Schwarz, T. (1996). IL-12 prevents ultraviolet-B induced local iiiiiiiuiiosuppression and breaks uhdviolet-B induced tolerance. J. Irioest. Dennatol. 106, 1187-1191. 446. Bianchi, R., Grohmann, U., Bellaclonna, M. L.. Silla, S., Fdlarino, F., Ayroldi, E., Fioretti, M. C . , and Puccetti, P. (1996). IL-12 is both required and sufficient for initiating T cell reactivity to a class I-restricted tumor peptide (P815AB) following transfer of P815AB-pulsed dendritic cells. J . I t n n ~ i i n o l . 157, 1589-1597. 447. Grohmann, U., Bianchi, H.. Ayroldi, E., Belladonnna. M. L., Surace, D., Fioretti, M. A,, and Puccetti, P. (1997).A tumor-associated and self antigen peptide presented by dendritic cells may induce T cell anergy in oioo, bnt IL-12 can prevent or revert the anergic state. J. Ztimunol. 158, 3593-3609. 4 M Via, C. S., Kus, L'., Gately, M. K., and Finkelman, F. D. (1994). IL-12 stiinulates the development of acute graft-versus-host disease. J . Ztt~rtlzcriol. 153, 4040-4047. 449. Williamson, E., Garside, P., Bradley, J. A,, and Mowat, A. M. (1996). IL-12 is a central mediator of acute graft-versus-host disease in mice. /. Immrrnol. 157, 689-699. 450. Hernandez, H. J., Wang, Y.. and Stadecker, M. J. (1997). 111 infection with Schistosonui tnnnsoni, B cells are required for T helper type 2 responses but not for granuloma formation. J . Znitiiirnol. 158, 4832-4837. 451. Tanaka, J., Imaniura, M., Kasai, M., Hashino, S., Kobayashi, S., Noto, S., Higa, T., Sakurada, K., and Asaka, M. (1997). The iinportant balance between cytokines derived from type 1 and type 2 helper T cells in the control of graft-versus-host disease. Bone Marrow Transpl. 19, 571-576. 452. Bonnotte, B., Burdiles, A. M., Chehiini, 3.. Carayol, G., Pardoux, C., Dietrich, P. Y., Kuhin, M., Blay. J. Y., Caignard, A,, Ibrahim. A., Robinet, E., Hayat M., Pico, J. L., Bourhis, J. H., and Chouaib, S. (1996).Seruii~interleukin-12 levels in patients undergoing allogeneic or autologousbone inarrow transplantation. Eur. Cyto. Netw. 7,389-394. 453. Williamson, E., Garside, P., Bradley, J. A,, More, I. A. R., and Mowat, A. M. (1997). Nelitralidng IL-12 during induction of niuiine acute graft-versus-host disease polarizes the cytokine profile toward a Th2-type alloimmune response and confers long term protection from disease. /. Zn~mirnol.159, 1208-1215. 454. Sykes, M.. Szot. G. L., Nguyen, P. L.. and Pearson, D. A. (1995). Interleukin-12 iiihibits mniine graft-versus-host disease. Blood 86, 2429-2438. 455. Cohen, J. (199s). IL-12 deaths: Explanation and a puzzle. Science 270, 908. 456. Conghlin, C. M., Wysocka, M., Trinchieri, G., and Lee, W. M. F. (1997). The effect of IL-12 desensitization on the anti-tumor efficacy of recombinant IL-12. Cancer Res. 57, 2460-2467. 457. Saito, K., Yagita, H., Hashiinoto, H., Oknmura, K., and Azuina, M. (1996). Effect of CD80 and CD86 blockade and anti-interleukin-12 treatment on monse acute graftversus-host disease. E u r . J. Znzmuriol. 26, 3098-3106.
224
GIORGIO TRINCHIERI
458. Gavett, S. H., O’Hearn, D. J., Li, X., Huang, S. K., Finkelman, F. D., and WillsKarp, M , (1995). Interleukin 12 inhibits antigen-induced airway hyperresponsiveness, inflammation, and TI12 cytokine expression in mice. J. Exp. Med. 182, 1527-1536. 459. Kips, J. C., Bmsselle, G. J,, Joos, G. F., Peleman, R. A., Tavernier, J. H., Devos, R. R., and Pauwels, R. A. (1996).Interleukin- 12 inhibits antigen-induced airway hyperresponsiveness in mice. Am. J. Resp. Crit. Care Med. 153, 535-539. 460. Iwamoto, I., Kumano, K., Kasai, M., Kurasawa, K., and Nakao, A. (1996). Interleukin12 prevents antigen-induced eosinophil recruitment into mouse airways. Am. J. Resp. Crit. Care Med. 154, 1257-1260. 461. Sur, S., Lam, J., Bouchard, P., Sigounas, A., Holbert, D., and Metzger, W. J. (1996). Iinmunomodulatory effects of IL-12 on allergic lung inflammation depend on timing of doses. J. Immunol. 157, 4173-4180. 462. Naseer, T., MinshalI, E. M., Leung, D. Y., Laberge, S., Emst, P., Martin, R. J,, and Hamid, Q. (1997). Expression of IL-12 and IL-13 mRNA in asthma and their modulation in response to steroid therapy. Am. J. Rev. Crit Care Med. 155, 845-851. 463. van der Pouw Kraan, T. C. T. M., Boeije, L. C. M., de Croot, E. R., Stapel, S. O., Snijders, A., Kapsenberg, M. L., van der Zee, J. S., and Aarden, L. A. (1997). Reduced production of IL-12 and IL-12-dependent IFN-y release in patients with allergic asthma. 1. Immunol. 158, 5560-5565. 464. Lin, J. Y., Wang. L. F., and Lin, R. H. (1996). The association between lung innate immunity and differential airway antigen-specific immune responses. Int. Immunol. 8,499-507. 465. Cish, R. G., Krams, S. M., and Martinez, 0. M. (1995). Interleukin-12: A possible cytotoxic T-lymphocyte differentiation factor in allograft recipients. Transpl. Proc. 27,459-460. 466. Cuturi, M. C., Heslan, J. M., Josien, R., Douillard, P., and Soulillou, J. P. (1997). High interleukin- 12 p40 mRNA expression in tolerant heart allografts in recipient rats treated by donor-specific transfusion. Transplant. Proc. 29, 1170. 467. Gorczynski, R. M., Hozumi, N., Wolf, S., and Chen, Z. (1995). Interleukin 12 in combination with anti-interleukin 10 reverses graft prolongation after portal venous immunization. Transplantation 60, 1337- 1341. 468. Gorczynski, R. M., Cohen, Z., Fu, X. M., Hua, Z., Sun, Y., and Chen, Z. (1996). Interleukin-13, in combination with anti-interleukin-12, increases graft prolongation after portal venous immunization with cultured allogeneic bone marrow-derived dendritic cells. Transplantation 62, 1592- 1600. 469. Kato, K., Shimozato, O., Hoshi, K., Wakinioto, H., Hamada, H., Yagita, H., and Okumura, K. (1996). Local production of the p40 subunit of interleukin 12 suppresses T-helper 1-mediated immune responses and prevents allogeneic myohlast rejection. Proc. Nut. Acad. Sci. USA 93, 9085-9089. 470. Piccotti, J. R.., Chan, S. U., Goodman, R. E., Magram, J., Eichwald, E. J., and Bishop, D. K. (1996). IL-12 antagonism induces T helper 2 responses, yet exacerbates cardiac allograft rejection: Evidence against a dominant protective role for T helper 2 cytoldnes in doimmunity. J. Immunol. 157, 1951-1957. 471. Trembleau, S., Germann, T., Gately, M. K., and Adorini, L. (1995). The role of IL-12 in the induction of organ-specificautoimmune diseases. Immunol. Today 16,383-386. 472. Rabinovitch, A., Suarez-Pinzon, W. L., and Sorenson, 0.(1996).Interleukin 12 mRNA expression in islets correlates with beta-cell destmction in NOD mice. J. Autoimmun. 9, 645-651. 473. Rothe, H., Burkart, V., Faust, A., and Kolh, H. (1996). Interleukin-12 gene expression is associated with rapid development of diabetes mellitus in non-obese diabetic mice. Diabetobgia 39, 119-122.
INTERLEUKIN-12
225
474. Tremblean, S., Penna, G., Bosi, E., Mortara, A., Gately, M. K., and Adorini, L. (1995). Interleukm 12 administration induces T helper type 1cells and accelerates autoiinmune diabetes in NOD mice. J . Exp. Med. 181, 817-821. 475. Rothe, H., O’Harra, R. M., Jr., Martin, S., and Kolb, H. (1997). Suppression of cyclophosphamide induced diabetes development and pancreatic Thl reactivity in NOD inice treated with the interlenkin (IL)-12 antagonist IL-l2(p40)2. Diabetologia 40, 641-646. 476. Adorini, L., Aloisi, F., Galbiati, F., Gately, M. K., Gregory, S., Penna, G., Ria, F., Smiroldo, S., andTrenibleau, S. (1997).The key cytokine driving Thl-mediated autoimmune diseases. Chem. Irmnunol. 68, 175- 179. 477. Rothe, H., Jenkins, N. A., Copeland, N. G., and Kolb, H. (1997). Active stage of autoimmune diabetes is associated with the expression of a novel cytokine, IGIF, which is located near Idd2. J. Clin. Inve.rt. 99, 469-474. 478. Rothe, H., Hibino, T., Itoh, Y., Kolb, H., and Margin, S. (1997). Systemic production of interferon-gamma inducing factor (IGIF) versus local IFN-gamma expression involved in the development of Thl insulitis in NOD mice. J , Autoimnzun. 10,251-256. 479. O’Hara, R. M., Jr., Henderson, S. L., and Nagelin, A. (1996). Prevention of a Th1 lsease by a Thl cytokine: IL-12 and diabetes in NOD mice. Ann. N.Y. Acad. Sci. 795, 241-249. 480. Zipris, D., Greiner, D. L., Malkani, S., Whalen, B., Mordes, J. P., and Rossini, A. A. (1996). Cytokiiie gene expression in islets and thyroids of BR rats: IFN-gamma and IL-12 p40 mRNA increase with age in both diabetic and insulin-treated nondiabetic BB rats. J . Immunol. 156, 1315-1321. 481. Chung, Y. H., Jun, H. S., Kang, Y., Hirasawa, K., Lee, B. R., Van Rooijen, N., and Yoon, J. W. (1997). Role of macrophages and macrophage-derived cytokines in the pathogenesis of Kilham rat virus-induced autoimmune diabetes in diabetes-resistant BioBreeding rats. J. Imrnunol. 159, 466-471. 482. Leonard, J. P., Waldburger, K. E., and Goldman, S. J. (1995).Prevention of experiniental autoimmune encephalomyelitis by antibodies against interleukin 12. J. Exp. Med. 181,381-386. 483. Waldburger, K. E., Hastings, R. C., Schaub, R. G., Goldnian, S. J., and Leonard, J. P. (1896). Adoptive transfer of experimental allergic encephalomyelitis after in vitro treatment with recombinant murine IL-12: Preferential expansion of interferongamma-producing cells and increased expression of macrophage-associated inducible nitric oxide synthase as immunomodulatory mechanisms. Ant. J . Pathol. 148,375-382. 484. Segal, B. M., and Shevach, E. M. (1996). IL-12 unmasks latent autoimmune disease in resistant mice. J . Exp. Med. 184, 771-775. 485. Segal, B. M., Klininan, D. M., and Shevach, E. M. (1997). Microbial products induce autoimmune disease by an IL-12-dependent pathway. J. Znimunol. 158, 5087-5090. 486. Leonard, J. P., Waldburger, K. E., and Goldnian, S. J. (1996).Regulation of experimental autoimmune encephalomyelitis by interleukin-12. Ann. N.Y. Acad. Sci. 795, 216-226. 487. Billiau, A,, Heremans, H., Vandekerckhove, F., Dijkmans, R., Sobis, H., Meulepas, E., and Carton, H. (1988). Enhancement of experimental allergic encephalomyelitis in mice by antibodies against IFN-y. J. Inirr~unol.140, 1506-1514. 488. Duong, T. T., St. Louis, J., Gilbert, J. J., Finkelman, F. D., and Strejan, G. H. (1992). Effect of anti-interferon-ganima and anti-interleukin-2 monoclonal antibody treatment on the development of actively and passively induced experimental allergic encephalomyelitis. J. Neuroimmun. 36, 105-114.
226
GIORGlO TRINCHIERI
489. Constantinescu, C . S., Hillard, B. A,, Wysocka, M., Lavi, E., Ventura, E., Trinchieri, G., and Rostaini, A. (1997). Interleukin-12 is involved in spontaneous and superantigeninduced relapses in experimental allergic encephalomyelitis. Neurology 48, 2070. 490. Issazadeh, S., Lorentzen, J. C., Must&, M. I., Hojeberg, B., Mussener, A,, and Olsson, T. (1996). Cytokines in relapsing experimental autoimmune encephalomyelitis in DA rats: Persistent mRNA expression of proinflammatory cytokines and absent expression of interleukin-10 and transforming growth factor-beta.J. Neuroirnrnimol. 69,103-115. 491, Issazadeh, S., Ljungdahl, A,, Hojeberg, B., Mustafa, M., and Olsson,T. (1995).Cytokine production in the central nervous system of Lewis rats with experimental autoimmune encephalomyelitis: Dynamics of inRNA expression for interleukin-10, interleukin-12, cytolysin, turnor necrosis factor a and tumor necrosis factor /3. J. Neuroirtarnunol. 61, 205-212. 492. Diab, A., Zhu, J., Xiao, B. G., Must&, M., and Link, H. (1997). High IL-6 and low IL-10 in the central nervous system are associated with protracted relapsing EAE in DA rats. J. Neuropthol. Exp. Neural. 56, 641-650. 493. Smith, T., Hewson, K., C. I., Leonard, J. P., and Cuzner. M. L. (1997). Interleukin12 induces relapse in experimental allergic encephalomyelitis in the Lewis rat. Am. J . Pothol. 150, 1909-1917. 494. Windhagen, A,, Hewcombe, J., Dangond, F., Strand, C., Woodroofe, M. N., Cuzner, M. L., and H d e r , D. A. (1995). Expression of costiinulatory molecules B7-1 (CD80), B7-2 (CD86), and interleukin 12 cytokine in multiple sclerosis lesions. J. E x p Med. 182, 1985-1996. 495. Bdashov, K. E., Smith, D. R., Khoury, S. J., Hafler, D. A., and Weiner, H. L. (1997). Increased interleukin 12 production in progressive multiple sclerosis: Induction by activated CD4+ T cells via CD40 ligand. Proc. Nntl. Acad. Sci. USA 94, 599-603. 496. Nicoletti, F., Patti, F., Cocuzza, C., Zaccone, P., Nicoletti, A., Di Marco, R., and Reggio, A. (1996). Elevated serum levels of interleukin-12 in chronic progressive multiple sclerosis. f. Neur(dmnunol, 70, 87-90. 497. Drulovic, J., Mostarica-Stojkovic, M., Levic, Z., Stojsavljevic, N., Pravica, V., and Mesaros, S. (1997). Interleukin-12 and tumor necrosis factor-alpha levels in cerebrospinal fluid of multiple sclerosis patients. J . Neurol. Sci. 147, 145-150. 498. Yokoi, H., Kato, K., Kezuka, T., Sakai, J., Usui, M., Yagita, H., and Ohimura, K. (1997). Prevention of experimental autoimmune uveoretinitis by inonoclonal antibody to interleukin-12. Eur. J. Inimnunol. 27, 641-646. 499. Xu, H., Rizzo, L. V., Silver, P. B., and Caspi, R. R. (1997). Uveitogenicity is associated with a Thl-like lymphokine profile: Cytokine-dependent modulation of early and committed effector T cells in experimental autoiininune uveitis. Cell. Immunol. 178, 69-78. ,500. Germann, T., Szeliga, J., Hess, H., Storkel, S., Podlaski, F. J,, Gately, M. K., Schmitt, E., and Rude, E. (1995). Administration of interleukin 12 in combination with type I1 collagen induces severe arthritis in D B N l mice. Proc. Notl. Acad. Sci. USA 9,48234827. 501. Joosten, L. A. B., Lubberts, E., Helsen, M. M. A., and van der Berg, W. B. (1997). Dud role of IL-12 in early and Iate stages of inurine collagen type I1 arthritis. J. I m n i u d . 159, 4094-4102. 502. Germann, T., Hess, H., Szeliga, J., and Rude, E. (1996). Characterization of the adjuvant effect of IL-12 and efficacy of IL-12 inhibitors in type 11 collagen-induced arthritis. Ann. N.Y. Acud. Sci. 795, 227-240. 503. Szeliga, J., Hess, H., Rude, E., Schmitt, E., and Germann, T. (1996). LL-12 promotes cellular but not lnnnoral type I1 collagen-specific Thl-type responses in C57BU6 and BlO.Q mice and fails to induce arthritis. Int. Immunol. 8, 1221-1227.
INTERLEUKIN-I2
227
504. Hess, H., Cately, M. K., Rude, E., Schmitt, E., Szeliga, J., and Gerinann, T. (1996). High doses o f interleukin-12 inhihit the development of joint disease in DBM1 inice immunized with type I1 collagen in complete Frennd's adjuvant. Etir. J. frriniunol. 26, 187-191. 505. Orange, J. S., Salazar-Mather, T. P., Opal. S. M., Spcnccr, R. L., Miller, A. H., McEwen, B. S., and Biron, C. A. (19%). Mechanism of interleukin 12-mediated toxicities during experimental viral infections: Role of tumor necrosis factor and glucocorticoids. J. Exp. Med. 181, 901 -914. 506. McIntyre, K. W.. Shuster, D. J., Gillooly. K. M., Warner, R. R., Connaughton, S. E.. Hall, L. B., Arp, L. H., Gately, M. K., and Magram, J. (1996). Reduced incidence and severity of collagen-induced arthritis in interleukin-12-deficient mice. Eur. J. fnimunol. 26, 2933-2938. 507. Buclit, A., Larsson, P., Weisbrot, L., Thonie, C., Pisa, P., Smedegard, G., Keystone. E. C., and Gronberg, A. (1996). Expression of interferon-gamma (IFN-y),IL-10, IL12 and transforming growth factor-beta (TCF-beta)niRNA in synovial fluid cells from patients in the early and late phases of rheurnatoid arthritis (RA). C h . E.rp. fmtnunol. 103,357-367. 508. Kotake, S., Schumacher. H. R., Jr.. Yarboro. C. H., Arayssi, T. K., Pando, J. A,, and Kanik, K. S. (1997). I n vivo gene expression of type 1 and type 2 cytokines in synovial tissues from patients in early stages of rheumatoid, reactive, and undifferentiated arthritis. Proc. Assoc. Am. Phys. 109, 286-301. 509. Neurath, M. F., Fuss, I., Kelsall, B. L., Stiiber. E., and Strober, W. (1995).Antibodes to interleukin 12 abrogate esta1)lishe.d experimental colitis in mice. /. Exp. Metl. 182, 1281-1290. 510. Duchmann, R., Schmitt, E., Knolle, P., Meyer Zuni Biischenfelde, K. H., and Neurath, M. (1996).Tolerance towards resident intestinal flora in mice is abrogated in experimental colitis and restored by treatment with interleukin-10 or antibodies to interleukin12. Bir. J. fniiniinol. 26, 934-938. 511. Marth, T.. Strober, W., Seder, R. A,, and Kelsall, B. L. (1997).Regulation of transforming growth factor-beta production by interleukin-12. Eur. /. fmrniinol. 27, 1213-1220. 512. Strober, W., Kelsall, B., FLISS, I., Marth, T., Ludviksson, B., Ehrhardt, R., and Neurath, M. (1997). Reciprocal IFN-gamma and TGF-beta responses regulate the occurrence l. 18, 61-64. of mucosal inflammation. I ~ n n ~ n oToc/ay. 513. Ludviksson, B. R., Gray, B., Strober, W., and Ehrhardt. R. 0. (1997). Dysregulated intrathymic development in the IL-2-deficient mouse leads to colitis-inducing thymocytes. J. fmnunol. 158, 104- 1 11. 514. Ehrhardt, R. O., Ludviksson, B. R., Gray, B., Neurath, M., and Strober W. (1997). Induction and prevention of colonic inflainination in IL-2-deficient mice. J. fmmunol. 158,566-573. 515. Monteleone, G., Biancone, L., Marasco, R., Morrone, G.. Marasco, O., Luzza, F., and Pallone, F. (1997). Interleukin 12 is expressed and actively released by Crohn's disease intestinal lamina propria niononuclear cells. Gastroenterology 112, 1169-1 178. 516. Parronchi, P., Romagnani, P., Annunziato, F., Sampognaro, S., Becchio. A., Giannarini, L., Maggi, E., Pupilh, C., Tonelli, F., and Romagnani, S. (1997). Type 1 T-helper cell predominance and interleukin-12 expression in the gut of patients with Crohn's disease. Ant. J. Patlid. 150, 82:3-832. 517. Huang, F. P., Feng, G. J., Lindop, G., Stott, D. I.. and Liew, F. Y. (1996). The role of interleukin 12 and nitric oxide in the development of spontaneous autoimmune disease in MRL:MP-1pr:lpr mice. J. Exp. Med. 183, 1447-1459.
228
GIORGIO TRINCHIERI
518. Fan X., Oertli, B., and Wuthrich, R. P. (1997). Up-regulation of tubular epithelial interleukin-12 in autoimmune MRL-Fas(lpr) mice with renal injury. Kidney Int. 51, 79-86. 519. Yanagi, K., Haneji, N., Hamano, H., Takahashi, M., Higashiyama, H., and Hayashi, Y. (1996). In viva role of IL-10 and IL-12 during development of Sjogren’s syndrome in MRWLpr mice. Cell. Zinmunol. 168, 243-250. 520. Hayashi, Y., Haneji, N., and Hamano, H. (1996).Cytokinegene expression andautoantibody production in Sjogren’s syndrome of MRUpr mice. Autoimmunity 23,269-277. 521. Houssiau, F. A,, Mascart-Lemone, F., Stevens, M., Libin, M., Devogelaer, J. P., Goldman, M., and Renauld, J. C. (1997). IL-12 inhibits in vitro immunoglobulin production by human lupus peripheral blood mononuclear cells (PBMC). Clin. Erp. Immunol. 108,375-380. 522. Nakamura, K., Okamura, H., Wada, M. Nagata, K., and Tamura, T. (1989). Endotoxininduced serum factor that stimulates gamma interferon production. Infect. Immun. 57,590-595. 523. Nakamura, K., Okamura, H., Nagata, K., Komatsu, T., and Tamura, T. (1993). Purification of a factor which provides a costimulatory signal for gamma interferon production. Znfect Zmmun. 61, 64-70. 524. Jansen, P. M., van der Pouw Kraan, T. C., de Jong, I. W., van Mierlo, C., Wijdenes, J., Chang, A. A,, Aarden, L. A,, Taylor, F. B., Jr., and Hack, C. E. (1996). Release of interleukin-12 in experimental Escherichia coli septic shock in baboons: Relation to plasma levels of interleukin-10 and interferon-gamma. Blood 87, 5144-5151. 525. Ozmen, L., Pericin, M., Hakimi, J., Chizzonite, R. A,, Wysocka, M., Trinchieri, G., Gately, M., and Carotta, G. (1994). IL-12, IFN-y and TNF-a are the key cytokines of the generalized Shwartzman reaction. ]. Exp. Med. 180, 907-916. 526. Cauwels, A,, Fiers, W., and Brouckaert, P. (1996). Murine IL-12 is involved in Calmette-Guerin bacillus-induced sensitization and is by itself sufficient to sensitize mice to the lethal effect of human TNF. J. Zmniunol. 156, 4686-4690. ,527. Randow, R., Docke, W., Bundschuh, D. S., Hartung, T., Wendel, A., and Volk, H. (1997). In vitro prevention and reversal of lipopolysaccharidedesensitization by IFNy, IL-12, and granulocyte-macrophage colony-stimulating factor. ]. Zmrnunol. 158, 2911-2918. 528. Turan, B., Gallati, H., Erdi, H., Gurler, A., Michel, B. A., and Villiger, P. M. (1997). Systemic levels of the T cell regulatory cytokines IL-10 and IL-12 in Bechcet’s disease; soluble TNFR-75 as a biological marker of disease activity.]. Rheumatol. 24,128-132. 529. Moller, D. R., Forman, J. D., Liu, M. C., Noble, P. W., Greenlee, B. M., Vyas, P., Holden, D., Forrester, J. M., Lazarus, A,, Wysocka, M., Trinchieri, G., and Karp, C. (1996). Enhanced expression of IL-12 associated with Thl cytokine profiles in active pulmonary sarcoidosis.]. Zniinunol. 156, 4952-4960. 530. Uyemura, K., Demer, L. L., Castle, S. C., Jullien, D., Berliner, J. A,, Cately, M. K., Warner, R. R., Pham, N., Fogelman, A. M., and Modlin, R. L. (1996).Cross-regulatory roles of interleukin (1L)-12 and IL-10 in atherosclerosis.]. Clin. Znuest. 97,2130-2138. 531. O’Sullivan, S. T., Lederer, J. A,, Horgan, A. F., Chin, D. H., Mannick, J. A., and Rodrick, M. L. (1995).Major injury leads to predominance of the T helper-2 lymphocyte phenotype and diminished interleukin-12 production associated with decreased resistance to infection. Ann. Surg. 222, 482-490. 532. Keel, M., Schregenberger, N., Steckholzer, U., Ungethum, U., Kenney, J., Trentz, O., and Ertel, W. (1996). Endotoxin tolerance after severe injury and its regulatory mechanisms. ]. Trauma 41, 430-437.
INTERLEUKIN-12
229
533. O’Suilleabhain, C., O’Sullivan, S. T., Kelly, J. L., Lederer, J., Mannick, J. A,, and Rodrick, M. L. (1996). Interleukin-12 treatment restores normal resistance to bacterial challenge after bum injury. Surgery 120, 290-296. 534. Matsuo, R., Kobayashi, M., Herndon, D. N., Pollard, R. B., and Suzuki, F. (1996). Interleukin-12 protects thermally injured mice from herpes simplex virus type 1 infection. J. Leiik Biol. 59, 623-630. 535. Orange, J. S., Wolf, S. F., and Biron, C. A. (1994). Effects of IL-12 on the response and susceptibility to experimental viral infections. J. Immunol. 152, 1253-1264. 536. Orange, J. S., Wang, B., Terhorst, C., and Biron, C. A. (1995).Requirement for natural killer cell-produced interferon gamma in defense against murine cytomegalovirus infection and enhancement of this defense pathway by interleukin 12 administration. J . Exp. Med. 182, 1045-1056. 537. Ozmen, L., Aguet, M., Trinchieri, G., and Garotta, G. (1995). The in vivo antiviral activity of IL-12 is mediated by IFN-y. J. Virol. 69, 8147-8150. 53s. Coutelier, J. P., Van Broeck, J., and Wolf, S. F. (1995). Interleukin-12 gene expression after viral infection in the mouse. 1.Virol. 69, 1955-1958. 539. Kanangat, S., Thomas, J., Gangappa, S., Babu, J. S., and Rouse, B. T. (1996). Herpes simplex virus type 1-mediated up-regulation of IL-12 (p40) mRNA expression: Implications in immunopathogenesis and protection. 1.Immunol. 156, 1110-1116. 540. Orange, J. S., and Biron, C. A. (1996). An absolute and restricted requirement for IL-12 in natural killer cell IFN-gamma production and antiviral defense: Studies of natural killer and T cell responses in contrasting viral infections. 1. lmmunol. 156, 1138-1142. 541. Monteiro, J., and Trinchieri, G. (1996). Does IL-12 play a role in the viral immune response? Ann. N.Y. Acad. Sci. 795, 366-367. 542. Cousens, L. P., Orange, J. S., Su, H. C., and Biron, C. A. (1997). Interferon-alplidbeta inhibition of interleukin 12 and interferon-gamma production in vitro and endogenously during viral infection. Proc. Natl. Acad. Sci. U S A 94, 634-639. 543. Bi, Z., Quandt, P., Komatsu, T., Barna, M., and Reiss, C. S. (1995). IL-12 promotes enhanced recovery from vesicular stomatitis virus infection of the central nervous system. 1.Imnwnol.155, 5684-5689. 544. Komatsu, T., and Reiss, C. S. (1997). IFN-y is not required in the IL-12 response to vesicular stomatitis virus infection of the olfactory bulb. /. Immunol. 159,3444-3452. ,545. Schijns, V. E. C. J., Wierda, C. M. H., van Hoeij, M., and Horzinek, M. C. (1996). Exacerbated viral hepatitis in IFN-y receptor-deficient mice is not suppressed by IL12.1. hnmfinol. 157, 815-821. 546. Milich, D. R.. Wolf, S. F., Hughes, J. L., and Jones, J. E. (1995). IL-12 suppresses autoantibody production by reversing helper T-cell phenotype in hepatitis B e antigen transgenic mice. Proc. Natl. Acnd. Sci. USA 92, 6847-6851. 547. Cavanaugh, V. J., Guidotti, L. G., and Chisari, F. V. (1997). Interleukin-12 inhibits hepatitis B virus replication in transgenic mice. J . Virol. 71, 3236-3243. 548. Tang, Y. W., and Graham, B. S. (1995). Interleukin-I2 treatment during immunization elicits a T helper cell type 1-like immune response in mice challenged with respiratory syncytial virus and improves vaccine immunogenicity. J. Infect. Dis. 172, 734-738. 549. Tang, Y. W., and Graham, B. S. (1997). T cell source of type 1 cytokines determines illness patterns in respiratory syncytial virus-infected mice. J. Clin. Invest. 99, 21832191. 550. Hussell, T., Khan, U., and Openshaw, P. (1997). IL-12 treatment attenuates T helper cell type 2 and B cell responses but does not improve vaccine-enhanced lung illness. J. ltnmunol. 159, 328-334.
230
GIOHGIO TRINCHIEHl
551. Yang, Y., Trinchieri, G., and Wilson, J. M. (1995).Recombinant IL-12 prevents formation of blocking IgA antibodies to recombinant adenovirus and allows repeated gene therapy to mouse lung. Nature Med. 1, 890-893. 552. Oldstone, M. B. A. (1996). Virus-lymphoid cell interactions. Proc. N d . Acud. Sci. USA 93, 12756-12758. 553. Fugier-Vivier, I., Servet-Delprat, C., Rivailler, P., Rissoan, M. C., Lui, Y. J.. and Rabourdin-Combe, C. (1997). Measles virus suppresses cell-mediated immunity by interfering with the survival and functions of dendritic and T cells. j . Exp. Med. 186, 813-823. 554. Schlender, J., Schnorr, J., Spielhofer, P., Cathomen, T., Cattaneo, R., Billeter, M. A,, Ter Meulen, V., and Schneider-Schaulies. S. (1996). Interaction of measles virus glycoproteins with the surface of uninfected peripheral blood lymphocytes induces immunosuppression in vitro. Proc. Natl. Acad. Sci. USA 93, 13194-13199. ,55555. Schnorr, J. J., Xanthakos, S., Keikavoussi, P., Kampgen, E., Ter Meulen, V., and Schneider-Schaulis, S. (1997). Induction of maturation of human blood dendritic cell precursors by measles v i m is associated with immunosuppression. Proc. Nntl. Acad. Sci. USA 94, 5326-5331. 556. Harrison, T. S., and Levitz, S. M. (1996).Role of IL-12 in peripheral blood mononuclear cell responses to fungi in persons with and without HIV infection. 1. Immunol. 156, 4492-4497. 557. Meyaard, L., Hovenkamp, E., Pakker, N., van der Pouw Kraan, T. C., and Miedema, F. (1997). Interleukin-12 (IL-12) production in whole blood cultures from human immunodeficiency virus-infected individuals studied in relation to IL-10 and prostaglandin E2 production. Blood 89, 570-576. 558. Chougnet, C., Wynn, T. A,, Clerici, M., Landay, A. L., Kessler, H. A,, Rusnak, J., Melcher, G. P., Sher, A,, and Shearer, G. M. (1996). Molecular analysis of decreased interleukin-12 production in persons infected with human immunodeficiency virus. J. Infect. Dis. 174, 46-53. 559. Gazzinelli, R. T., Bda, S., Stevens, R., Baseler, M., Wahl, L., Kovacs, J., and Sher, A. (1995).HIV infection suppresses type 1lymphokine and IL-12 responses to Toxoplmmu gondiii but fails to inhibit the synthesis of other parasite-induced monokines.j . Immtrnol. 155, 1565-1574. 560. Denis, M., and Ghadirian, E. (1994). Dysregulation of interleukin 8, interleukin 10, and interleukin 12 release by alveolar macrophages from HIV type 1-infected subjects. AIDS Res. Hum. Retrov. 10, 1619-1627. 561. Yoo, J., Chen, H., Kraus, T., Hirsch, D., Polyak, S., George, I., and Sperber, K. (1996). Altered cytokine production and accessory cell function after HIV-1 infection. 1.Immunol. 157, 1313-1320. 562. Fantuzzi, L., Gessani, S., Borghi, P., Varano, B., Conti, L., Puddu, P., and BelardeUi, F. (1996). Induction of interleukin-12 (IL-12) by recombinant glycoprotein gp120 of human immunodeficiencyvirus type 1in human monocytes/macrophages:requirement of gamma-interferon for IL-12 secretion. j . Virol. 70,4121-4124. 563. Taoufik, Y., Lantz, O., Wallon, C., Charles, A., Dussaix, E., and Delfraissy, J. F. (1997). Human immunodeficiency virus gp120 inhibits interleukin-12 secretion by human monocytes: An indirect interleukin-lo-mediated effect. Blood 89, 2842-2848. 564. Clerici, M., Herkin, F. T., Venzon, D. J., Blatt, S., Hendrix, G. W., Wynn, T. A,, and Shearer, G. (1993). Changes in interleukin-2 and interleukin-4 production in asymptomatic, human immunodeficiencyvirus-seropositiveindividuals.j . Clin. Inoest. 91, 789-795.
INTERLEUKIN-12
23 1
565. Clerici, M., and Shearer, G. M. (1993). A Thl-Th2 switch is a critical step in the etiology of HIV infection. l n i m w d Today 14, 107-111. 566. Clerici, M., and Shearer, G. M. (1994). The Thl-Th2 hypothesis of HIV infection: I ~ 575-581. New insights. lmnwnol. T o ~ ( J15, 567. Hamson, T. S., and Levitz, S. M. (1997). Priming with IFN-y restores deficient IL12 production by peripheral blood mononuclear cells from HIV-seropositive donors. J . lmmunol. 158, 459-463. 568. Chehimi, J., Marshall, J. D., Salvucci,0..Frank, I., Chehimi, S.,Kawecki, S., Bacheller, D., Rifat, S., and Chouaib, S. (1997). IL-1.5 enhances immune functions during HIV infection. J. Zmimmol. 158, 5978-5987. 569, Mastino, A,, Grelli, S., Piacentini, M., Oliverio, S., Favalli, C., Perno, C. F., and Garaci, E. ( 1993). Correlation between induction of lymphocyte apoptosis and prostaglandin E2 production by inacrophages infected with HIV. Cell. Immtinol. 152, 120-130. ,570. Hoffinan, B., Nishanian, P.. Nguyen, T., Liu, M., and Fahey, J. L. (1993). Restoration of T cell fuuction in HIV infection by reduction of intracellular CAMP levels with adenosine analogues. AlDS 7, 659-664. 571. Graziosi, C., Pantdeo, G., Gantt, K. R.. Fortin, J.. Demarest, J. F., Cohen, 0. J.. Sekaly, R. P., and Fanci, A. S. (1994). Lack of evidence for the dichotomy of T h l and Th2 predominance in HIV-infected indivitlids. Science 265, 248-252. 572. Fan. J., Bass, H.Z., and Fahey. J. L. (1993). Elevated IFN-y and decreased IL-2 gene expression are associated with HIV infection. J . Immzcnol. 154, 5031-5040. 573. Poli, G., Introna, M., Zaiiaboni, F., Pen, G., Carbonan. M., Aiuti, F., Lazzarin, A,, Moroni, M., and Mantovani, A. (1985). Natural killer cells in intravenous drug abusers with lymphadenopathy syndrome. Clin. EX!),Immunol. 62, 128-135. 574. Sirianni, M. C., Tagliaferri, F., and Aiuti, F. (1990). Pathogenesis of the natural killer cell deficiency in AIDS. Immunol. Today 11, 81-82. ,575. Sirianni, M. C., Ansotegui, I. J., Aiuti, F., and Wigzell, H. (1994). Natural killer cell stimulatory (NKSF)IIL-12 and cytolytic activities of PBUNK cells from human immunodeficiency virus type-I infected patients. Scnnd. J . I r n n ~ u t d40, 83-86. 576. Lin, S. J,, Roberts, R. L., Ank, B. J., Nguyen, Q. H., Thomas, E. K., and Stiehm, E. R. (1997). Human immunodeficiency vinis (HIV) type-1 GPl20-specific cellmediated cytotoxicity (CMC) and natural killer ( N K ) activity in HIV-infected (HIV+) subjects: Enhancement with interleukin-2 (IL-2). IL-12, and IL-15. Clin.lmimnol. 82, 163- 173. ~ml~lIi71(~~th0l. 577. Seder, R. A,, Grabstein, K. H., Berzofsky, J. A,, iuid McDyer, J. F. (1995). Cytokine interactions in human iinmiinodecifiencyvirus-infected individuals: roles of interleukin (1L)-2, IL-12, and IL-15.1. Exp. Med. 182, 1067-1077. 578. Paganin, C., Frank, I., and Trinchieri, G. (1995).Priming for high interferon-? production induced by interleukin-12 in both CD4' and CDB' T eel1 clones from BIVinfected patients. J . Clin.Znoccst. 96, 1677-1682. 579. Uherova, P., Connick, E., MaWhinney, S., Schlichtemeier, R., Schooley, R. T., and Kuritzkes, D. R. (1996).In vitru efyect of interleukin-12 on antigen-specificlymphocyte proliferative responses from persons infected with human immunodeficiency virus type 1.1. lnfect. Dis. 174, 483-489. ,580. Ahlers, J. D., Dunlop, N., Alling, D. W., Nara. P. L., and Berzofsky, J. A. (1997). Cytokine-in-adjnvat steering of the immune response phenotype to HIV-1 vaccine constructs: Granulocyte-macrophage colony-stimulating factor and TFN-a synergize with IL-12 to enhance induction of cytotoxic T 1ylnphocytes.J. lmmunol. 158,39473958.
232
GIORCIO TRINCHIERI
581. Tsuji, T., Hamajima, K., Fukushima, J., Xin, K., Ishii, N., Aoki, I., Ishigatsubo, Y., Tani, K., Kawamoto, S., Nitta, Y., Miyazaki, J,, Koff, W. C., Okubo, T., and Okuda, K. (1997).Enhancement of cell-mediated immunity against HIV-1 induced by coinoculation of plasmid-encoded HIV-1 antigen with plasmid expressing IL-12. J. Immunol. 158,4008-4013. 582. Hamajima, K., Fukushima, J., Bukawa, H., Kaneko, T., Tsuji, T., Asakura, Y., Sasaki, S., Xin, K. Q., and Okuda, K. (1997). Strong augment effect of IL-12 expression plasmid on the induction of HIV-specific cytotoxic T lymphocyte activity by a peptide vaccine candidate. Clin. lnimunol. lmniunopathol. 83, 179-184. 583. FoIi, A,, Saville, M.W., Baseler, M. W., and Yarchoan, R. (1995). Effects of the Thl and Th2 stimulatory cytokines interleukin-12 and interleukin-4 on human immunodeficiency virus replication. Blood 85, 2114-2123. 584. Perales, M. A,, Skolnik, P. R., and Lieberman, J. (1996). Effect of interleukin 12 on in vitro HIV type 1 replication depends on clinical stage. AIDS Res. Hum. Retrovir. 12, 659-668. 585. Bayard-McNeeley, M., Doo, H., He, S., Hafner, A., Johnson, W. D., Jr., and Ho, J. L. (1996). Differential effects of interleukin-12, interleukin-15, and interleukin-2 on human immunodeficiencyvirus type 1replication in vitro. Clin. Ding. Lab. Immunol. 3, 547-553. 586. Kinter, A. L., Bende, S. M., Hardy, E. C., Jackson, R., and Fauci, A. S. (1995). Interleukin 2 induces CD8f T cell-mediated suppression of human immunodeficiency virus replication in CD4+ T cells and this effect overrides its ability to stimulate virus expression. Proc. Natl. Acad. Sci USA 93, 10985-10989. 587. Akridge, R. E., and Reed, S. G. (1996). Interleukin-12 decreases human immunodeficiency virus type 1 replication in human macrophage cultures reconstituted with autologous peripheral blood mononuclear cells. J. Infect. Dis. 173, 559-564. 588. Ciese, N. A,, Gazzinelli, R. T., Morawetz, R. A,, and Morse, H. C., I11 (1995). Role of IL-12 in MAIDS. Res. Immunol. 146, 600-604. 589. Ciese, N. A., Gazzinelli, R. T., Actor, J. K., Morawetz, R. A,, Sarzotti, M., and Morse, H. C., 111 (1996). Retrovirus-elicited interleukin-12 and tumour necrosis factor-alpha as inducers of interferon-gamma-mediated pathology in mouse AIDS. Immunology 87,467-474. 590. Andrews, C., Swain, S. L., and Muralidhar, G. (1997). CD4 T cell anergy in murine AIDS. J. Immunol. 159, 2132-2138. 591. Gazzinelli, R. T., Giese, N. A., and Morse, H. C., I11 (1994). In vivo treatment with interleukin 12 protects mice from immune abnormalities observed during murine acquired immunodeficiency syndrome (MAIDS). J. Exp. Med. 180,2199-2208. 592. Knight, S. C., and Patterson, S. (1997). Bone marrow-derived dendritic cells, infection with human immunodeficiency virus, and imrnunopathology. Annu. Reu. Inimanol. 15, 593-615. 593. Liu, W., and Kurlander, R. J. (1995). Analysis of the interrelationship between IL-12, TNF-a, and IFN-7 production during murine listeriosis. Cell. Immunol. 163,260-267. ,594. Tripp, C. S., Gately, M. K., Hakimi, J., Ling, P., and Unanue, E. R. (1994). Neutralization of IL-12 decreases resistance to Listeria in SCID and CB-17 mice. 1.Immunol. 152, 1883-1887. 595. Wagner, R. D., Steinberg, H., Brown, J. F., and Czuprynski, C. J. (1994). Recombinant interleukin-12 enhances resistance of mice to Listeria monocytogenes infection. Microb. Pathogen. 17, 175-186. 596. Ladel, C. H., Blum, C., and Kaufmann, S. H. (1996). Control of natural killer cellmediated innate resistance against the intracellular pathogen Listeria monocytogenes by gammddelta T lymphocytes. Infect. lmmun. 64, 1744-1749.
INTERLEUKIN-12
233
597. Skeen, M. J.. and Ziegler, H. K. (1995). Activation of y8 T cells for production of IFN-y is mediated by bacteria via macrophage-derived cytokines IL-1 and IL-12. 1. Inllnu710~.154, 5832-5841. 598. Tripp, C. S., Kanagawa, O., and Unanue, E. R. (1995). Secondary response to Listeria infection requires IFN-gamma but is partially independent of IL-12. J . Imrnicnol. 155,3427-3432. 599. Dai, W. J., Bartens, W., Kohler, G., Hufiiagel, M.. Kopf, M., and Brombacher, F. (1997). Impaired macrophage listericidal and cytokine activities are responsible for the rapid death of Listeria monocytogenes-infected IFN-gamma receptor-deficient mice. J . Immunol. 158, 5297-5304. 600. Dai, W. J., Kohler, G., and Brombacher, F. (1997).Both innate and acquired immunity to Listeria monocytogenes infectioii are increased in IL-10-deficient mice. J . Immunol. 158, 2259-2267. 601. Song, F., Matsuzaki, G., Mitsuyania, M., and Nomoto, K. (1996). Differential effects of viable and killed bacteria on IL-12 expression of macrophages. /. Zrrirnunol. 156, 2979-2984. 602. Miller, M. A,, Skeen, M. J., and Ziegler, H. K. (1995). Nonviable bacterial antigens administered with IL-12 generate antigen-specific T cell responses and protective immunity against Listeria monocytogenes.J. Trnnzunol. 155, 4817-4828. 603. Miller, M. A., Skeen, M. J., and Ziegler, H. K. (1996). Protective immunity to Listeria monocytogenes elicited by immunization with heat-killed Listeria and IL-12: Potential mechanism of IL-12 adjuvanticity. Ann. New York Acnd. Sci. 797, 207-227. 604. Miller, M. A,, Skeen, M. J., and Ziegler, 13. K. (1997).A synthetic peptide administered with IL-12 elicits immunity to Lbteria nwnocytogens. 1.Immunol. 159, 3675-3679. 605. Henderson, R. A., Watkins, S. C., and Flynn, J. L. (1997).Activation ofhuinan dendritic cells following infection with Mycobacterium tuberculosis. J. Imrnunol. 159, 635-643. 606. Cooper, A. M., Roberts, A. D., Rhoades, E. R., Callahan, J. E., Getzy, D. M., and Orme, I. M. (1995).The role of interleukin-12 in acquired immunity to Mycobncterium tuberculosis infection. Irnmunolog!y 84, 423-432. 607. Flynn, J. L., Goldstein, M. M., Triebold, K. J., Sypek, J., Wolf, S., and Bloom, B. R. (1995). IL-12 increases resistance of BALB/c mice to Mycobncterium tuhercukxis infection. /. Iinmunol. 155, 2515-2524. 608. Cooper, A. M., Magram, J.. Ferrante, J., and Orme, I. M. (1997). Interleukin 12 (IL12) is crucial to the development of protective immunity in mice intravenously infected with mycobacterium tuberculosis. /. Exp. Med. 186, 39-45. 609. Ladel, C. H., Szalay, G., Riedel, D., and Kaufmann, S. H. (1997). Interleukin-12 secretion by Mycobncterium tnherciLlosis-infected macrophages Infect. Immun. 65, 1936-1938. 610. Yoshida, A,, and Koide, Y. (1997). Arabinofuranosyl-terminated and mannosylated lipoarabinomannans from Mycobncteriuni tuberculosis induce different levels of interleukin-12 expression in inurine macrophages. Infect. Inmzun. 65, 1953-1955. 611. Lindblad, E. B., Elhay, M. 5.. Silva, R., Appelberg, R., and Andersen, P. (1997). Adjuvant modulation of iinninne responses to tuberculosis subunit vaccines. Infect. Itnrnun. 65, 623-629. 612. Zhang, M., Gately, M. K., Wang, E., Gong, J., Wolf, S. F., Lu, S., Modlin, R. L., and Barnes, P. F. (1994). Interleukin 12 at the site of disease in tuberculosis. /. C h . Invest. 93, 1733-1739. 613. Taha, R . A,, Kotsimbos, T. C., Song, Y. L., Menzies, D., and Hamid, Q (1997). IFNgannna and IL-12 are increased in active compared with inactive tuberculosis. Am. /. Rmp. Crit. Cure Med. 155, 1135-1139.
234
GIORC.10 TRINCHIERI
614. Munk, M. E., Mayer, P., Anding P., Feldman, K., and Kaufmann, S. H. E. (1996). Increased nnmbers of interleitkin-12-producing cells in human tuberculosis. Infect. Zntmun. 64, 1078-1080. 615. Sieling, P. A,, Wang, X. H., Gately, M. K., Oliveros, J. L., McHugh, T., Barnes, P. F., Wolf, S. F., Golkar, L., Yaniamura, M., and Yo@, Y. (1994). IL-12 regulates T helper type 1cytokine responses in human infectious disease.]. Immnnol. 153,36393647. 616. de Jong, R., Janson, A. A. M., Faller, W. R., Nzafs, B., and Ottenhoff, T. H. M. (1997). IL-2 and IL-12 act in synergy to overcome antigen-specific T cell unresponsiveness in mycobactend disease. J. Irnmund. 159, 786-793. 617. Saunders, B. M., Zhan, Y., and Cheers, C. (1995). Endogenous interleukin-12 is involved in resistance of mice of Mycohacterizim nvium complex infection. Infect. Immnun. 63, 4011-4015. 618. Kobayashi, K., Kasama, T.. Yamazaki, J., Hosaka, M . , Katsura, T., Mochizuki, T., Soejima, K., and Nakamura, R. M. (1995). Protection of mice from Mycobacterium nvium infection by recombinant interleukin-12. Antimicro. Agents Chernother. 39, 1369-1371. 619. Kobayashi, K., Yamazaki, J., Kasama, T., Katsura, T., Kasahara, K., Wolf, S. F., and Shimainura, T. (1996). Interleukin (1L)-12 deficiency in susceptible mice infected with Mycobacterium auium and amelioration of established infection by IL-12 replacement therapy. 1.Infect. Dis. 174, ,564-573. 620. Frucht, D. M., and Holland, S. M. (1996). Defective inonocyte costiinulation for IFN-y production in familial disseminated Mycobncteriicrn auiiim complex infection: Abnorinal IL-12 regulation. J . Inzrnunol. 157, 411-416. 621. Chong, C., Bost, K. L., and Clements, J. D. (1996). Differential production of interleukin-12 mRNA by rnurine macrophages in response to viable or killed Salmonella spp. Infect. Immun. 64, 1154-1160. 622. Bost, K. L., and Clements, J. D. (1997). Iiitracellular Salmonella dublin induces substantial secretion of the 40-kilodaton subunit of interleukin-12 (IL-12) but minimal scretion of IL-12 as a 70-kilodalton protein in murine macrophages. Infect. Irnmun. 65, 3186-3192. 623. Bost, K. L., and Clements, J. D. (1995). In vivo induction of interleukin-12 mRNA expression after oral immunization with Salmonella dublin or the B subunit of Escherichia coli heat-labile enterotoxin. hfect. bnniun. 63, 1076-1083. 624. Kincy-Cain, T., Clements, J. D., and Bost, K. L. (1996). Endogenous and exogenous interleukin-12 augment the protective immune response in mice orally challenged with Sabnunelln dsiblirt. Infect. lmniun. 64, 1437-1440. 625. Mastroeni, P., Harrison, J. A., Chabalgoity, J. A,, and Hormaeche, C. E. (1996). Effect of interleukin 12 neutralization on host resistance and gamma interferon production in mouse typhoid. Infect. Immnuti. 64, 189-196. 626. Anguita, J., Persing, D. H., Rinch, M., Barthold, S. W., and Fikrig, E. (1996).Effect of anti-interleukin 12 treatment of murine lyme borreliosis. J. Clin. Invest. 97,1028-1034. 627. Cantorna, M. T., and Hayes, C. E. (1996). Vitamin A deficiency exacerbates murine Lyme arthritis. J . Infect. Dis. 174, 747-751. 628. Filgueira, L., NestlC, F. O., Rittig, M., Joller, H. I., and Groscurth, P. (1996). Human dendritic cells phagocytose and process Borrelia hzirgrlorferi. J. Onmunol. 157,29983005. 629. Zhan, Y., and Cheers, C. (1995). Endogenous interleukin-12 is involved in resistance to Brucella abortus infection. Infect. Immun. 63, 1387-1390.
INTERLEUKIN-12
235
630. Zhaii, Y., Liu, Z., and Cheers, C. (1996).Tuinor necrosis factor alpha and interleukin12 contribute to resistaiice to the intracelhilar bacteiium Brucella nhortu.~by different mechanisms. Iifecf. I r i i i r ~ t t r i ,64, 2782-2786. 631. Zaitseva, M., Golding, H., Manischewitz, J., Wehb, D., and Golding, B. (1996).Bmcellri C L I I O ~ ~ T as ~ S a potential vaccine candidate: Induction of interleukin-12 secretion and enhanced B7.1 and B7.2 and intercellular adhesion inolecule 1 surface expression in ehitriatetl hiinian monocytes stinialated by heat-inactivated B. nhortu.u. Iilfect. Ziniituri. 64,3109-3117. 632. Malion, R. P., Ryan, M. S., Griffin, F., and Mills, K. H. (1996). Interleukin-12 is produced by tnacrophages in response to live or killed Bordetelln pertussis and enhances the efficacy of an acellular peitu vaccine by promoting induction of Thl cells. Infect. Itiitnun. 64, 5295-5301. 633. Bohn, E., and Autenrieth, 1. B. (1996). IL-12 is essential for resistance against Yesinin enterocoliticn by triggering IFN-y production in NK cells and CD4+T ce1ls.J. Itmnunol. 156, 1458-1468. 634. Perry, L. L., Feilzer, K., and Chldwell, H. D. (1997).Inllnunitj to Chlaniydiin trtlclzotiintis is inediated by T helper 1 cells through IFN-gannna -dependent and -independent pathways. /, I I I L ~ I L Z L158, ~ W ~ .3344-3352. 63.5. Karttunen, R. A,, Karttuiien. T. J., Yousfi, M. M., El-Zimaity. H. M., Graham, D. Y., and El-Zaatari, F. A. (1997). Expression of inRNA for- interferon-gamma, interleukin10. and interleulan-12 (p40) in normal gastric inucosa and in inucosa infected with Helicobacter pyliiri. Scanrl. J. Gn.striieiitero/. 32, 22-27. 636. Mancuso, G., Cusumano, V., Genovese, F., Gambuzza, M., Beninati, C., and Teti, C . (1997). Role of interleuhi1 12 in experiinental iieonatal sepsis caused by group B streptococci. Infect. I ~ ~ L I I L W65, I . 3731-3735. 637. Creenberger, M. J., Kunkel, S. L., Strieter, R. M., Lukacs, N. W., Brainson, J., Caddie, J., Graham, F. L., Hitt, M., Danforth, J. M., and Standiford, T. J. (1996). IL-12 gene therapy protects mice in lethal Klebsielln pneumonia. J. Iiniiiutiol. 157, 3006-3012. 638. Metzger, D. W., Raeder, R., Van Cleave, V. II., and Royle, M. D. P. (1995). Protection of mice from group A streptococcal skin infection by interleukin-12. 1. Itfect. Dis. 171, 1643-1645. 639. Noll, A., and Autenrieth, I. B. (1996). Iininiinity against Yersinia enterocolitiua by vaccination with Yersinia HSP60 iminunostimulating coinplexes or Yersinia HSPGO plus interleukin-12. Iifect. Z w i ~ t m64, 2955-2961. 640. Marth, T., Neurath. M., Cuccherini, B. A,, and Strolier, W. (1997).Defects ofnionocyte interleulaii 12 production and humoral immunity in \VIiipple’s disease. Grrstroetiterology 113,442-448. 641. Ijeinzel, F. P., Schoenhaut, D. S., Rerko, R. M., Rosser, L. E., and Gately, M. K. (1993). Reconihiiiaiit interleukin 12 cures mice infected with Leishmania major. /. Exp. Me(!. 177, 1,505-1509. 642. Sypek, J. P., Chung, C. L., Major S. E. H., Subramanyain. J. M., Goldlnan, S. J., Sieburth, D. S., Wolf, S. F., and Schaub, R. G. (1993). Resolution of cutaneous leishmaniasis: Interleukin-12 initiates a protective T helper type 1 inmune response. J . Exp. Med. 177, 179771802, 643. Mougneaii, E., Altare, F., Wakil, A. E., Zheng. S., Chppola, T., Wang. Z.. Wddiiiann, R., Locksley, R. M., and Glaichenhaus. N. (199s). Expression cloning of a protective Leishrimnia antigen. Scierice 268, 563-566. 644. N’akil, A. E., Wang, Z. E., and Locksley, R. M. (1996). Leislziiu~niaiiwjor: Targeting IL-4 in successfulimmunomodulation ofmiirine infection. Exp. Prirmital. 84,214-222.
236
CIORGIO TRINCHIERI
645. Gumnathm, S., Sacks, D. L., Brown, D. R., Reiner, S. L., Charest, H., Glaichenhaus, N., and Seder, R. A. (1997). Vaccination with DNA encoding the irninunodominant LACK parasite antigen confers protective immunity to mice infected with Leishnlania major. J. Exp. Med. 186, 1137-1147. 646. Nabors, G. S., Afonso, L. C., Farrell, J. P., and Scott, P. (1995). Switch from a type 2 to a type 1T helper cell response and cure of established Leishinunia major infection in mice is induced by combined therapy with interleukin 12 and Pentostam. Proc. Natt!. Acad. Sci. USA 92, 3142-3146. 647. Li, J., Suttenvala, S., and Farrell, J. P. (1997).Successful therapy ofchronic, nonhealing murine cutaneous leishmaniasis with sodium stibogluconate and gamma interferon depends on continued interleukin-12 production. Infect. Inmiun. 65, 3225-3230. 648. Mocci, S., and Coffman, R. L. (1997). The mechanism of in uitro T helper cell type 1 to T helper cell type 2 switching in highly polarized Leishmania major-specific T cell populations. J. Immunol. 158, 1559-1564. 649. Reiner, L. S., Zheng, S., Wang, Z., Stowring,L., and Locksley,R. M. (1994).Leishmania promastigotes evade interleukin 12 (IL-12) induction by macrophages and stimulate a broad range of cytokines from CD4+ T cells during initiation of infection. J. Exp. Med. 179,447-456. 650. Carrera, L., Gazzinelli, R. T., Badolato, R., Hieny, S., Muller, W., Kuhn, R., and Sacks, D. L. (1996). Leishmania promastigotes selectively inhibit interleukin 12 induction in macrophages from susceptible and resistant mice. J . Exp. Med. 183, 515-526. 651. Vieira, L. Q., Hondowicz, B. D., Afonso, L. C. C., Wysocka, M., Trinchieri, G., and Scott, P. (1994). Infection with Leishinunia mujor induces interleukin-12 production in vivo. Iinmunol. Lett. 40, 157-161. 652. Mattner, F., Alber, G., Magram, J., and Kopf, M. (1997). The role of IL-12 and IL4 in Leishmania major infection. Chem. Immunol. 68, 86-109. 653. Mattner, F., Magram, J., Ferrante, J., Launois, P., Di Padova, K., Behin, R., Gately, M. K., Louis, J. A,, and Alber, G. (1996). Genetically resistant mice lacldiig interleukin~ mount a polarized Th2 cell 12 are susceptible to infection with Leishmania V X Z ~ C Jand response. Eur. J. Immunol. 26, 1553-1559. 654. Campbell, K. A., Ovendale, P. J., Kennedy, M. K., Fanslow, W. C., Reed, S. G., and Maliszewski, C. R. (1996). CD40 ligand is required for protective cell-mediated immunity to Leishmania nmjor. Iminunity 4, 283-289. 655. Kamanaka, M., Yu, P., Yasui, T., Yoshida, K., Kawabe, T., Horii, T., Kishimoto, T., and Kikutani, H. (1996). Protective role of CD40 in Leishmuiiia major infection at two distinct phases of cell-mediated immunity. Immunity 4, 275-281. 656. Launois, P., Swihart, K. G., MiIon, G., and Louis, J. A. (1997). Early production of IL-4 in susceptible mice infected with Leishmania major rapidly induces IL-12 unresponsiveness. J. Immunol. 158, 3317-3324. 657. Murray, H. W., and Hariprashad, J. (1995). Interleukin 12 is effective treatment for an established systemic intracellular infection: Experimental visceral leishmaniasis. J. Exp. Med. 181,387-391. 658. Murray, H. W., Hariprashad, J., and Coffman, R. L. (1997). Behavior of visceral Leishmania donouani in an experimentally induced T helper cell 2 (Th2)-associated response model. J. Exp. Med. 185, 867-874. 659. Taylor, A. P., and Murray, H. W. (1997). Intracellular antimicrobial activity in the absence of interferon-y: Effect of interleukin-12 experimental visceral leishrnaniasis in interferon-? gene-disrupted mice. J. Exp. Med. 185, 1231-1239. 660. Murray, H. W. (1997). Endogenous interleukin-12 regulates required resistance in experimental visceral leishmaniasis. J. Infect. Dis. 175, 1477-1479.
INTERLEUKIN-12
237
661. Melby, P. C., Andrade-Narvaez, F., Darnell, B. I., and Valencia-Pacheco, G. (1996). In situ expression of interleukin-10 and interleukin-12 in active human cutaneous leishmaniasis. FEMS Iinmunol. Mecl. Microbial. 15, 101-107. 662. Bacellar, O., Brodskyn, C., Gerreiro, J., Barral-Netto, M., Costa, C. H., Coffman, R. L., Johnson, W. D., and Carvalho, E. M. (1996). Interleukin-12 restores interferongamma production and cytotoxic responses in visceral leishmaniasis. J. Iifect. Dis. 173, 1515-1518. 663. Ghalib, H. W., Whittle, J. A,, Kubin, M., Hashim, F. A,, El-Hassai, A. M., Grabstein, K. H., Trinchieri, G., and Reed, S. G. (1995). IL-12 enhances Thl-type responses in human Leishrnanio donouani infections. f. lnm2~moE.54, 4623-4629. 664. Hunter, C. A,, Subauste, C. S., van Cleave, V. H., and Reinington, J. S. (1994). Production of gamma interferon by natural killer cells from Toxoplusnln gondii-infected SCID mice: Regulation by interleukin-10, interleukin-12, and tumor necrosis factor alpha. Infecf. Iinnzun. 62, 2818-2824. 665. Neyer, L. E., Grunig, G., Fort, M., Remington, J. S., Rennick, D., and Hunter, C. A. (1997).Role of interleukin-10 in regulation ofT-cell-dependent andT-cell-independent mechanisms of resistance to Toxoplusina gondii. Irfect. lminun. 65, 1675-1682. 666. Scharton-Kersten, T. M., Yap, G., Magram, J., and Sher, A. (1997). Inducible nitric oxide is essential for host control of persistent but not acute infection with the intracellular pathogen Toxoplusmn gondii, J . Exp. Med. 185, 1261-1273. 667. Aliberti, J. C., Cardoso, M. A,, Martins, G. A,, Gazzinelli, R. T., Vieira, L. Q., and Silva, J. S. (1996). Interleukin-12 inediates resistance to Typanosomu cruzi in mice and is produced by inurine macrophages in response to live trypomastigotes. Infect. I I ~ W L U64, ~ . 1961-1967. 668. Frosch, S., Krans, S., and Fleischer, B. (1996). Trr~pnnosomacruzi is a potent inducer of interleukin-12 production in macrophages. Med. Microbiol. Imniunol. 185,189-193. 669. Abraliamsohn, I. A,, and Coffinan, R. L. (1996). Tnjpanosoia cruzi: IL-10, TNF, IFN-gamma, and IL-12 regulate innate and acquired immunity to infection. Exp. Parasitol. 84, 231-234. 670. Hunter, C. A,, Slifer, T., and Araujo, F. (1996). Interleuldn-12-mediated resistance to Trypanosonm cmzi is dependent on tumor necrosis factor alpha and gamma interferon. fn$e&. fmmun. 64, 2381-2386. 671. Brodskyn, C. I., Barral, A,, Bulhoes, M. A., Souto, T., Machado, W. C., and BarralNetto, M. (1996). Cytotoxicity in patients with different clinical forms of Chagas’ disease. Clin. Exp. Inimunol. 105, 450-455. 672. de Barros-Muon, S., Guariento, M. E., and Abrahamsolin, I. (1997). IL-12 enhances proliferation of peripheral blood mononuclear cells from Chagas’ disease patients to Typanosoinn cruzi antigen. Imrrrunol. Lett. 57, 39-45. 673. Sedegah, M., Finkelman, F., and Hoffman, S. L. (1994). Interleukin 12 induction of interferon gamma-dependent protection against malaria. Proc. Natl. Acad. Sci. USA 91, 10700-10702. 674. Hoffman, S. L., Crutcher. J. M., Pun, S. K., Ansari, A. A., Villinger. F., Franke, E. D., Singh, P. P., Finkelman, F., Gately, M. K., Dutta, G. P., and Sedegah, M. (1997).Sterile protection of monkeys against malaria after administration of interleukin12. Nature Med. 3, 80-83. 675 Stevenson, M. M., Tam, M. F., Wolf, S. F., and Sher, A. (1995). IL-12-induced protection against blood-stage Plnsmodiuin chabaudi AS requires IFN-y and TNF-a and occurs via a nitric oxide-dependent mechanism. J . Imniunol. 155, 2545-2556. 676. Roniani, L., Bistoni, F., and Puccetti, P. (1997). Initiation of T-helper cell immunity to Candirla albicans by IL-12: The role of neutrophils. Chenr. Imniunol. 68,110-135.
238
GIORGIO TRINCHIERI
677. Romani, L., Mencacci, A., Tonnetti, L., Spaccapelo, R., Cenci, E., Wolf, S., Puccetti, P., and Bistoni, F. (1994). Interleukin-12 but not interferon-y production correlates with induction of T helper type-1 phenotype in murine candidiasis. Eur. J. Irnnwnol. 24, 909-915. 678. Romani, L., Mencacci, A,, Tonnetti, L., Spaccapelo, R., Cenci, E., Puccetti, P., Wolf, S. F., and Bistoni, F. (1994). Interieukin-12 is both required and prognostic in vivo for T helper type 1 differentiation in murine candidiasis.]. h n u n o l . 153,5167-5157. 679. Romani, L., Mencacci, A,, Cenci, E., Del Sero, G., Bistoni, F., and Puccetti, P. (1997). An immunoregulatory role for neutrophils in CD4+ T helper subset selection in mice with canddiasis. ]. Imrriunol. 158,2356-2362. 680. Romani, L., Meneacci, A,, Cenci, E., SpaccapeIo, R., Del Sero, G., Nicoletti, I., Trinchieri, G., Bistoni, F., and Puccetti, P. (1997). Neutrophil production of IL-12 and IL-10 in candidiasisand efficacy of IL-12 therapy in neutropenic mice.]. Ininwnol. 158, 5349-5356. 681. Clemons, K. V., Brummer, E., and Stevens, D. A. (1994). Cytokine treatment of central nervous system infection: Efficacy of interleukin-12 alone and synergy with conventional antifungal therapy in experimental cryptococcosis. Antimicrob. Agents. Chevnother. 38, 460-464. 682. Kawakami, K., Tohyama, M., Xie, Q., and Saito, A. (1996). IL-12 protects mice against pulmonary and disseminated infection caused by Cnjptococcus neofonnans. C h i . Exp. Inzmunol. 104, 208-214. 683. Kawakami, K., Tohyama, M., Qifeng, X., and Saito, A. (1997). Expression of cytokines and inducible nitric oxide synthase mRNA in the lungs of mice infected with Cyptococcus neofonnans: Effects of interleukin-12. Infect. Imrnun. 65, 1307-1312. 684. Kawakami, K., Qifeng, T., Tohyama, M., Qureshi, M. H., and Saito,A. (1996). Contribution of tumor necrosis factor-alpha (TNF-alpha) in host defence mechanism against CnjptococnLs neoformans. Clin. Exp. Imniunol. 106, 468-474. 685. Zhou, P., Sieve, M. C., Bennett, J,, Kwon-Chung, K. J., Tewari, R. P., Gazzinelli, R. T., Sher, A,, and Seder, R. A. (1995). IL-12 prevents mortality in mice infected with Histoplama capszllatum through induction of IFN-y.]. Inununol. 155,785-795. 686. Zhou, P., Sieve, M. C., Tewari, R. P., and Seder, R. A. (1997).Interleukin-12 modulates the protective immune response in SCID mice infected with Histoplasnw capsulatum. Infect. Immun. 65, 936-942. 687. Allendoerfer, R., Biovin, G. P., and Deepe, G. S., Jr. (1997). Modulation of immune responses in murine pulmonary histoplasmosis. 1.Infect. Dis. 175, 905-914. 688. Magee, D. M., and Cox, R. A. (1996). Interleukin-12 regulation of host defenses against Coccirlivides iwintitis. Infect. bnniun. 64, 3609-3613. 689. Wynn, T. A., Eltoum, I., Oswald, 1. P., Cheever, A. W., and Sher, A. (1994). Endogenous interleukin 12 (IL-12) regulates granuloma formation induced by eggs of Schistosonui inansoni and exogenous IL-12 both inhibits and prophylactically immunizes against Exp. Med. 179, 1551-1561. egg pathology. I. 690. Wynn, T. A., Cheever, A. W., Jankovic, D., Poindexter, R. W., Caspar, P., Lewis, F. A., and Sher, A. (1995). An IL-12-based vaccination method for preventing fibrosis induced by schistosome infection. Nature 376, 594-596. 691. Mountford, A. P., Anderson, S., and Wilson, R. A. (1996). Induction of Thl cellmediated protective immunity to Schistosomnu mansoni by co-administration of larval antigens and IL-12 as an adjuvant. J. Imnzunol. 156, 4739-4745. 692. Wynn, T. A., Reynolds, A., James, S., Cheever, A. W., Caspar, P., Hieny, S., Jankovic, D., Strand, M., and Sher, A. (1996). IL-12 enhances vaccine-induced immunity to
INTERLEUKIN-12
693.
694.
69.5. 696.
697.
698. 699.
700. 701.
702.
703.
704
705.
706.
707.
239
scliistosonies by augmenting both liuniord and cell-mediated imniune responses against the parasite. J . Z t t i t n i i t d . 157, 4068-4078. Pearlman, E., Heinzel, F. P., Hazlett, F. E., Jr., and Kazura, J. W. (1995). 1L-12 modillation of T helper responses to the filarial helmintli, Bnigici itwlnyi. J . Z t t m i i n d 154,4658-4664. Mahanty, S., Ravicliandran, M., Ranian, U., Jayarainan, K.. Kuniaraswaini. V.. and Nutnian, T. B. (1997). Regulation of parasite antigen-driven iinniune responses by interleukin-10 (IL-10)and IL-12 in lyrnphatic filariasis.Irfect. I n ~ t m i n65, . 1742-1747. Locksley, R. M. (1994). Th2 cells: Help for helminths. j . Exp, Mcd. 179, 1405-1407. Urban, J. F., Jr., Fayer, R.. Sullivan, C., Goldhill, J., Shea-Donohue, T., Madden, K., Morris, S. C., Katona, I., Cause, W., Ruff, M., Mansfield, L. S., and Finkelinan, F. 11. (1996). Local T H l and TH2 responses to parasitic infection in the intestine: Regulation by IFN-gainina and 11,-4. Vet. I t t i t r u m d . Zmttiuiiopnthol. 54, 337-344. Rotinan, H. L., Sclinyder-Candrian, S., Scott, P.. Nolan. T. J., Schad, G. A,, and Abrahani, 13.(1997).IL-12 eliminates the Th-2 dependent protective inunune response of inice to larval Strongyloicks .stereoralis. Parosite Irrmunol. 19, 29-39. Bancroft, A. J., Else. K. J., Sypek, J. P., and Grencis, R. K. (1997). Interleukin-12 proniotes a chronic iutestind nematode infection. Eur. J. Znurruiiol. 27, 866-870. Lotze, M. T., Zitvogel, L., Cainpht.11, R., Robbiiis, P. D., Elder, E., Haluszczak, C., Martin, D., Wliiteside, T. L., Storkus, W. J., and Tahara, H. (1997). Cytokine gene therapy of caiicer using interleukin-12: Murine and clinical trials. Ann. N.Y. Acud. Sci. 795,440-454. Brunda, M. J., and Gately, M. K. (1995). Interleukin-12: Potential role in cancer therapy. Import. Arlo. Oncol. 1, 3-18. Car, B. D., Eng, V. M., Schnyder, B., LeHir, M., Shakhov, A. N., \Voerly, G., Huang, S., Agriet, M., Anderson, T. D.. and Ryffel, B. (1995). Role of interferon-y in IL-12induced pathology in mice. Am. /. Pcithol. 147, 1693- 1707. Bree, A. G . .Schlerman, F. J., Kaviani, M. D., Hastings, R. C., Hitz, S. L.. a i d Goltlman, S. 1. (1994). Multiple effects on peripheral hematology following administration of recombinant human interleukin 12 to nonhuman primates. Biochett~Biophys. Kes. C ( J I ~ I ~204, ~ ~ U1150-1 I I . 157. Brnnda, M. J,, Luistro, L., Warner, R. R., Wright. I<. B., Hubbard, B. R., Murphy. M., B’olf, S. F., and Gately, M. K. (1993). Antitunlor and antimetastatic activity of interleukin 12 against murine tumors. 1.Exp. Met!. 178, 1223-1230. Nastala, C. L., Edington, H. D., McKinney, T. G., Tahara, €I., Nalesnik, M. A., Brunda, M. J., Gately, M. K., U’olf, S. F.. Schreiber, R. D., Storkus, W. J., and Lotze, M. T. (1994). Recombinant IL-12 administration induces tumor regression in association with IFN-y production. /. Imtnunol. 153, 1697-1706. Zou, J., Yainamoto, N.. Fujii, T., Takenaka, H., Kohayashi, M., Herrmanu. S. H., Wolf, S. F., Fujiwara, H., and Hamaoka, T. (1995). Systemic administration of rIL-12 induces complete tumor regression and protective irnmunity: Responsp is correlated with a striking reversal of suppressed interferon-y production by anti-tumor T cells. Znt. ~1711?21it10l.7, 1135- 1145. Mu, J., Zou, J.. Yainamoto, N., Tsutsui, T., Tai, X., Kobayashi, M., Hernnann, S., Fujiwara, H., and Hainaoka, T. (1995). Administration of rIL-12 prevents outgrowtl~ of tumor cells metastasizing spontaneously to liing and lymph nodes. Ceincer Res. 55, 4404-4408. Verbik, D. J., Stinson, W.\V,, Brunda, M. J., Kessinger, A,, and Josh, S. S. (1996). In vivo therapeutic effects of interleiikin-12 against highly nietastatic residual lymt . 219-229. pho111a. C h i , ESP. M e t ~ t ~14,
240
GlOHGlO TRINCHIEXI
708. Biunda, M. J., Luistro, L., Hendrzak, J. A,, Fountoulakis, M., Garotta, G., and Gately, M. K. (1995). Role of interferon-y in mediating the antitumor efficacy of interleukin12. J. Inmunother. Emph. Turnor Intrnunol. 17, 71-77. 709. Tannenbauin, C. S., Wicker, N., Armstrong, D., Tubbs, R., Finke, J., Bukowski, R. M., and Hamilton, T. A. (1996). Cytokine and chemokine expression in tumors of mice receiving systemic therapy with IL-12. J. Zmmunol. 156, 693-699. 710. Hegde, S. P., Bright, J. J., Kausalya, S., and Khar, A. (1995). AK-5 tumor-induced expression of interleukin-12: Role of IL-12 in NK-mediated AK-5 regression. Cell. rmn7nunoi.162,241-247. 711. Fallarino, F., Uyttenhove, C., Boon, T., and Gajewski, T. F. (1996). Endogenous IL12is necessaryfor rejection of PA15 tumor valiants inviv0.J. Zmmunol. 156,1095-1100. 712. Noguchi, Y., Jungbluth, A., Richards, E. C., and Old, L. J. (1996). Effect of interleukin 12 on tumor induction by 3-methylcholanthrene. Proc. Natl. Acad. Sci. USA 93,1179811801. 713. Wigginton, J. M., Kuhns, D. B., Back, T. C., Brunda, M. J.. Wiltrout, R. H., and Cox, G. W. (1996). Interleukin 12 primes macrophages for nitric oxide production in vivo and restores depressed nitric oxide production by macrophages from tumor-bearing mice: Iinplications for the antitumor activity of interleukin 12 an&or interleukin 2. Cancer Res. 56, 1131-1136. 714. Yu, W. G., Yamainoto, N., Takenaka, H., Mu, J., Tai, X. G., Zou, J, P., Ogawa, M., Tsutsui, T., Wijesuriya, R., Yoshida, R., Herrmann, S., Fujiwara, H., and Hamaoka, T. (1996). Molecular mechanisms underlying IFN-gamma-mediated tumor growth inhibition induced during tumor irnmunotherapy with rIL-12. Znt. Zmmunol. 8, 855-865. 715. Tsung, K., Meko, J. B., Peplinski, 6. R., Tsung, Y. L., and Norton, J. A. (1997). IL12 induces T helper 1-directed antitumor response. J. Immunol. 158, 3359-3365. 716. Voest, E. E., Kenyon, B. M., O’Reilly, M. S., Truitt, G., D’Amato, R. J., and Folkman, J. (199Fj). Inhibition of angiogenesis in vivo by interleukin 12. JNCI 87, 581-586. 717. Majewski, S., Marczak, M., Szmurlo, A., Jablonska, S., and Bollag, W. (1996). Interleukin-12 inhibits angiogenesis induced by human tumor cell lines in vivo. J. Invest. D e m t o l . 106, 1114-1118. 718. Sgadari, C., Angiolillo, A. L., and Tosato, G. (1996). Inhibition of angiogenesis by interleukin-12 is mediated by the interferon-inducible protein 10. Blood 87, 38773882. 719. Watanabe, M., McCormick, K. L., Volker, K., Ortaldo, J. R., Wigginton, J. M., Brunda, M. J., Wiltrout, R. H., and Fogler, W. E. (1997). Regulation of local host-mediated anti-tumor mechanisms by cytokines: Direct and indirect effects on leukocyte recruitment and angiogenesis. Ant. J. Pathol. 150, 1869-1880. 720. Dighe, A. S., Richards, E., Old, L. J., and Schreiber, R. D. (1994). Enhanced in vivo growth and resistance to rejection of tumor cells expressing dominant negative IFNy receptors. Immunity 1, 447-554. 721. Tahara, H., Zitvogel, L., Storkus, W. J., Zeh, M. J., 111, McKinney, T. G., Schreiber, R. D., Gubler, U.,Robbins, P. D., and Lotze, M. T. (1995). Effective eradication of established murine tumors with IL-12 gene therapy using a polycistronic retroviral vector. J . Immunol. 154, 6466-6474. 722. Martinotti, A., Stoppacciaro, A., Vagliani, M., Melani, C., Spreafico, F., Wysocka, M., Parmiani, G., Trinchieri, G., and Colombo, M. (1995). CD4 T cells inhibit i7i zjivo CD8-mediated immune response against a inurine colon carcinoma transduced with IL-12 genes. Eur. J . Zmmunol. 25, 137-146.
INTERLEUKIN- 12
24 1
723. Chen, L., Chen. D., Block, E., O’Donnell, M., Knfe, D. W., and Clinton, S. K. (1997). Eradication of murine bladder carcinoma by intratumor injection of a bicistronic adenoviral vector carrying cDNAs for the IL-12 heterodimer and its inhibition hy the n ~351-359. l . IL-12 p40 subunit homodirner. ]. r n ~ ~ ~159, 724. Bramson, J., Hitt, M., Gallichan, W. S., Rosenthal, K. L., Gauldie, J., and Graham, F. L. (1996). Construction of a double recombinant adenovirus vector expressing a heterodimeric cytokine: In vitro and in viva production of biologicallyactive interleukin12. Hum. Gene Ther. 7, 333-342. 725. Bramson, J. L., Hitt, M., Addison, C. L., Muller, W. J., Gaul&, J., and Graliam, F. L. (1996). Direct intratumoral injection of an adenovirus expressing interleukin-12 induces regression and long-lasting immunity that is associated with highly localized expression of interleukin-12. Hum Gene Tlier. 7, 1995-2002. 726. Meko, J. B., Yim, J. H.. Tsung, K., and Norton, J. A. (1995). High cytokine production and effective antitumor activity of a recombinant vaccinia virus encoding muiine interleukin 12. Cancer Res. 55, 4765-4770. 727. Meko, J. B., Tsung, K., and Norton, J. A. (1996). Cytolane production and antitumor effect of a nonreplicating, noncytopathic recombinant vaccinia virus expressing interleukin-12. Siirgery 120, 274-282. 728. Tan, J., Newton, C. A., Djeu, J. Y., Cutsch, D. E., Chang, A. E., Yang, N. S., Klein, T. W., and Hua, Y. (1996). Injection of complementary DNA encoding interleukin12 inhibits tumor establishment at a distant site in a murine renal carcinoma model. Cancer Res. 56,3399-3403. 729. Rakhmilevich, A. L., Turner, J., Ford, M. J., McCabe, D., Sun, W. H., Sondel, P. M., Crota, K., and Yang, N. S. (1996).Gene gun-mediated skin transfection with interleukin 12 gene results in regression of established primary and metastatic murine tumors. Proc. Not. Acad. Sci. USA 93, 6291-6296. 7,330. Tahara, H., Zeh, H. J., 111, Storkus, W. J,, Pappo, I., Watkins, S. C., Gubler, U., Wolf, S. F., Rohbins, P. D.. and Lotze, M. T. (1994). Fibroblasts genetically engineered to secrete interleukm 12 can suppress tumor growth and induce antitumor immunity to a murine melanoma in ciao. Cancer Hes. 54, 182-189. 731. Zitvogel, L., Tahara. H., Kobbins, P. D., Storkus, W. J., Clarke, M. R., Nalesnik, M. A,, and Lotze, M. T. (1995). Cancer immunotherapy of established tumors with IL-12. Effective delivery by genetically engineered fibroblasts.]. Imnitnol. 155, 13931403. 732. Colombo, M. P., Vagliani, M., Spreafico, F., Parenza, M., Chiodoni, C., Melani, C., and Stoppacciaro, A. (1996). Amount of interleukin-12 available at the tumor site is critical for tumor regression. Cancer Res. 56, 2531-2534. 733. Obana, S., Miyazawa, H., Hara, E., Tamura, T., Nariuchi, H., T,akata, M., Fijimoto, S., and Yamamoto, H. (1995). Induction of anti-tumor immunity by mouse tumor cells transfected with mouse interleukin-I2 gene. Jap. J. Med. Sci. B i d . 48, 221-236. 734. Teicher, B. A., Ara, C., Menon, K., and Schaub, R. G. (1996). In viva studies with interleukin-12 alone and in combination with monocyte colony-stimulating factor an& or fractionated radiation treatment. Znt. 1, Concer. 65, 80-84. 735. Wigginton, J. M., Komschlies, K. L., Hack, T. C., Franco, J. L., Bmnda, M. J., and Wiltrout, R. H. (1996).Administration of interleukin 12 with pulse interleukin 2 and the rapid and complete eradication of murine rend carcinoma. J. Natl. Cancer h s t . 88, 38-43. 736. Pappo, I., Tahara, H., Robbins, P. D., Gately, M. K., Wolf, S. F., Barnea, A., and Lotze, M. T. (1995). Administration of systemic or local interleukin-2 enhances the anti-tumor effects of‘ interleukin-12 gene therapy. J. Surg. Rex 58, 218-226.
242
GIORGIO TRINCIIIERI
737. Vagliani, M . , Rodolfo, M., Cavallo, F., Parenza, M., Melani, C., Parmiani, G., Forni, G., and Colombo, M. P. (1996). Interleukin 12 potentiates the curative effect of a vaccine based on interleukin 2-transduced tumor cells. Cancer Res. 56, 467-470. 738. Coughlin, C. M., Wysocka, M., Kurzawa, H. L., Lee, W. M. F., Trinchieri, G., and Eck, S. L. (1995). B7-1 and IL-12 synergistically induce effective antitumor immunity. Cancer Res. 55,4980-4987. 739. Zitvogel, L., Robbins, P. D., Storkus, W. J., Clarke, M. R., Maeurer, M. J., Campbell, R. L., Davis, C. G., Talvara, H., Schreiber, R. D., and Lotze, M. T. (1996). Interleukin12 and B7.1 co-stimulation cooperate in the induction of effective antitumor immunity and therapy of established tumors. Eur. J. lntmzrnol. 26, 1335-1341. 740. Rodolfo, M., Zilocchi, C., Melani, C., Cappetti, B., Arioli, I., Parmiani, G., and Colombo, M. P. (1996). Iininunotherapy of experimental metastases by vaccination with interleukin gene-transduced adenocarcinoma cells sharing tumor-associated antigens: Comparison between IL-12 and IL-2 gene-transduced tumor cell vaccines. J. bnmunol. 157,5536-5542. 741. Lieberman, M., Sigal, R., Williams, N., aiid Daly, J. (1991). Natural killer cell stimulatory factor (NKSF) augments natural killer cell and antibody-dependent tumoricidal response against colon carcinoma cell lines. /. Surg. Res. 50, 410-415. 742. Andrews. J. V., Schoof, D. D., Bertagnolh, M. M., Peoples, G. E., Goedegebuure, P. S., and Eberlein, T. J. (1993). Iinmunomodulatory effects of interleukin-12 on human tumor-infiltrating lymphocytes.]. lmntunother. 14, 1-10. 743, Bigda, J., Mysliwska, J., Dziadziuszko, R., Bigda, J,, Mysliwski, A., and Hellmann, A. (1993). Interleukin-12 augments natural killer-cell mediated cytotoxicity in haiiy cell leukemia. Leuk. Lymph. 10, 121-125. 744. Rossi, A. R.. Pericle, F., Rashleigh, S., Jauiec, J., and Djeu, J. Y. (1994). Lysis of neuroblastoma cell lines by human natural killer cells activated by interleukin-2 and interleukin-12. Blood 83, 1323-1328. 745. Kuge, S., Watanabe, K., Makino, K., Tokuda, Y., Mitomi, T., Kawainura, N., Habu, S., and Nishimura, T. (1995). Interleukin-12 augments the generation of autologous tumor-reactive CD8+ cytotoxic T lymphocytes from tumor-infiltrating lymphocytes. ]up. ]. Cancer Res. 86, 135-139. 746. Rashleigh, S. P., Kusher, D. I., Endicott, J. N., Rossi, A. R., and Djeu, J. Y. (1996). Interleukins 2 and 12 activate natural killer cytolytic responses of peripheral blood mononuclear cells from patients with head and neck squamous cell carcinoma. Arch. Otolayn. Head Neck Surg. 122, 541-547. 747. DeCesare, S. L., Michelini-Norris, B.. Blanchard, D. K.. Barton, D. P., Cavanagh, D., Roberts, W. S., Fioiica, J. V., Hoffnian, M. S., and Djeu, J. Y. (1995). Interleukin-12mediated tuinoricidal activity of patient lymphocytes in an autologous in vitro ovarian cancer assay system. Gyriecul Oncul. 57, 86-95. 748. Hauagiri, T., Takenoyama, M., Yoshimoto, T., Hirashiina, C., Yoshino, I., Nakanishi, K., Nagashirna, A., Nomoto, K., and Yasumoto, K. (1996). Effects of interleukin-12 on in vitro culture with interleukin-2 of regional lymph node lymphocytes froin lung cancer patients. Cancer Itnrnunol. brmunother. 43, 87-93. 749. Uharek, L., Zeis, M., Glass, B., Steininann, J., Dreger, P., Gassmann, W., Schmitz, N., aiid Muller-Ruchholtz, W. (1996). High lybc activity against huinan leukemia cells after activation of allogeneic NK cells by IL-12 and IL-2. Leukemia 10, 1758-1764. 7.50. Steger, G. G., Gnant, M. F., Djavanmard, M. P., Mader, R. M., Jakesz, R., Pierce, W., DeKeniion, J. B., Figlin, R., and Belldegrun, A. (1997). The in vitro effects of interleukin-12 upon tumor-infiltrating lymphocytes derived from renal cell carcinoma. J. Can. Res. Clin Oncol. 123, 317-324.
INTEHI,EUKIN-12
243
7S1. Hill, L. L., Penissia, B., McCue, P. A., and Konigold, R. (1994). Effect of human natural killer cells on the metastatic growth of human melanoma xeiiografts in mice with severe combined iirirnunodeficiency. Cancer Re.s. 54, 763-770. 7.52. Iwanuma, Y., Chen, F. A.. Egilinez, N. K., Takita, H., arid Banked, R. B. (1997). Antitumor immune response of human peripheral blood lymphocytes coengrafted with tumor into severe combined iniinunodeficic~ntmice. Cnncer Res. 57, 2937-2942. 753. Rook, A. H., Kahin, M., Cassin, M., Vonderheid, E., Vowels, B. R., Wolfe. J. T., Wolf, S. F.. Sin&, A,, Trinchieri, G., and Lessin, S. R. (1995). Interleukiii 12 reverses cytokine and immune abnorinalities in Sezaiy syndrome./. Zmniunol. 154, 1491-1498. 754. Rook, A. H., Kubin, M., Fox, F. E.. Niu, Z..Gassin, M., Vowels, B. R., Gottleib, S. L., Vonderheid, E. C.. Lessin, S. R., and Trinchieri. G., (1996). The potential therapeutic role of interleiikin-12 in cutaneous T-cell lymphoma. Aiin. N.Y. Acnd Sci. 795, 310-318. This article was accepted for publication on Deceinber 15, 1997.
This Page Intentionally Left Blank
AI)\'AYCE5
Ih I M M U N O L O C \ V O L 70
Recent Progress on the Regulation of Apoptosis by Bcl-2
Family Members ANDY 1. MINN,',f RACHEL E. SWAIN,* AVERIL MA,?,§ AND CRAIG 6. THOMPSON',t,*,§,(,II 'Gwen Knapp Center for Lupus ond Immunology Research, tCommi&e on Immunology, #Committee on Cancer Biology, heparhnent of Medicine, THoword Hughes, Medical Instihk, I I D e p h e n t of Molecular Genetics and Cell Biology, University of Chicago, Chicago, Illinois 60637
1. Significance of Programmed Cell Death
The development of multicellular organisms and the renewal of differentiated cell types in adult organisms rely on an expansion in cell numbers, making the control of cell division an important process in metazoan development and homeostasis. However, properly regulated development and homeostasis also require cell death (Ellis et al., 1991; Jacobson et al., 1997). Developmentally regulated cell death has been studied in both invertebrate and vertebrate animals and has been referred to as programmed cell death (PCD). Originally, this term was used to emphasize that this type of death is part of developmental programs. With the finding that regulated forms of cell death also occur in adult multicellular organisms, the term PCD has been adopted to describe all forms of cell death that are mediated by an intracelliilar program as opposed to those that occur through necrosis. Despite the long-standing knowledge that cells can undergo PCD, its importance in metazoan development and homeostasis has only recentIy been widely appreciated. PCD has many important roles in development. For example, PCD is utilized for the many morphologcal changes that take place during embryogenesis. This can be seen early in the embryo when PCD sculpts the digits by eliminating the unwanted cells between them (Saunders, 1966) or when PCD causes the Mullerian duct to regress in the male in order to establish sexual dimorphism (Jost, 1971). PCD is also important for the elimination of excess cells or cells that have developed improperly. In the developing vertebrate nervous system, many neuronal cells are produced in excess. The number of neurons that is needed is dictated by the size of the organ that the neurons enervate (Hamburger, 1975, 1992). The neurons that survive are the ones that have successfully competed for trophic factors andlor synaptic connections. Therefore, the developmental program successfully uses mechanisms of PCD in order to ensure that the supply of cells is exactly equivalent to the demand that is established by the developing organ. 245
< opymglit D I998 by A ~ a d e i n iPrcss ~ All rrglits of reproduction in any form rrserved 0065 2776rJ8 $25 00
246
ANDY J. MINN et nl
In addition to the developing nervous system, the vertebrate immune system also illustrates the importance of cell death in eliminating cells that have developed improperly (Abbas, 1996; Nossal, 1994; Surli and Sprent, 1994). As with neurond cells, lymphocyte development begns with the generation of a large excess of cells. The purpose of such overproliferation is to allow for lymphocytes to stochastically generate antigen receptors of unique specificity. The minority of lymphocytes that produce a functional antigen receptor through V( D)J recombination are signaled to survive through a process of positive selection while the rest are eliminated by PCD. However, the random generation of antigen receptors also produces lymphocytes that are self-reactive. Therefore, a mechanism must be in place to eliminate those cells that pose a threat to the welfare of the organism. Potentially autoreactive lymphocytes are eliniinated by PCD through a process of negative selection. The developmental processes of positive and negative selection in the T lymphocyte immune system lead to the removal of greater than 95% of all thymocytes by PCD, an extreme but illustrative example of the role that PCD plays in the development of inulticellular organism. The important roles that PCD play in the maintenance of inulticellular organisms are not limited to development. Once development has been completed, PCD reinains an active process to ensure proper organismal homeostasis. It has been proposed that nearly all cells in an adult organism are programmed to die and that this fate must be actively suppressed by environmental survival cues (Raff, 1992; Weil et al., 1996). For example, a tissue-specific cell dies if it is not residing in the proper tissue. Such a cell is committed to die because it is not receiving appropriate environmental survivd cues. In this way, the requirement for proper environmentd survival signals not only serves to dictate the tissue specificity of various cells but also creates a niche of limited size to support them. The homeostasis of a particular organ system can be maintained not only by constant survival signals that inhibit cell death but also by signals that activate the PCD machinery. During an immune response, lymphocytes that are reactive to a particular foreign antigen undergo a dramatic expansion in order to combat the foreign antigen. Once the pathogen is contained, the expanded cells must be removed in order to return cell numbers to a homeostatic level (Osborne, 1996). PCD rids the host of cells that have completed their function through a series of “death receptors.” The failure of PCD mechanisms to remove lymphocytes after the elimination of antigen can lead to disorders in which lymphocytes accumulate inappropriately. However, the triggering of excessive PCD by viruses such as human immunodeficiency virus (HIV)can lead to the tremendous
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
247
loss of lymphocytes and result in acquired immune deficiency syndrome known as AIDS. Since PCD plays a pivotal role in the development and in the maintenance of homeostasis within inulticellular organisms, disruptions in the pathways by which the cell death program operates are thought to contribute to the pathology of various human diseases (Thompson, 1995). Mutations in genes known to control PCD have been associated with pathologies such as cancer, autoimmunity, and neurodegenerative disorders. Thus, much attention has focused on trying to define the molecular mechanism by which PCD is regulated. This review concentrates on a family of critical cell death regulators known as the Bcl-2 family and summarizes some of the most recent advances in our understanding of their properties, structure, and function. II. The Genetics of Programmed Cell Death
The components of the programmed cell death pathway are highly conserved throughout metazoan evolution. It is likely that all signals that initiate programmed cell death ultimately trigger a central execution mechanism. Some of these signals include growth factor withdrawal, viral infection, inappropriate oncogene activation, cytotoxic T-cell killing, drug treatment, DNA damage, and other events that perturb normal cellular function (Yang and Korsmeyer, 1996). Based on an analysis of several examples of cell death in mammals, Wyllie and colleagues grossly characterized the morphological changes that result from this central execution mechanism and named it apoptosis, a Greek term meaning “fallingoff’ which describes leaves or petals (Kerr et al., 1972). Morphologically, apoptosis results in the loss of cell volume, membrane blebbing, and chromatin condensation (Cohen, 1993). DNA fragmentation into oligonucleosomd size fragments occurs through a calcium-dependent endonuclease, and the cell fragments into apoptosis bodies to facilitate consumption by phagocytes. Because this process does not result in the leakage of cellular contents, dead cells can be cleanly removed. Although some biochemical and morphological differences are apparent between certain cells undergoing programmed cell death and the classical description of apoptosis (Ellis et al., 1991; Schwartz et al., 1993), apoptosis is generally equated with PCD. It is difficult to ascertain whether a phenomenon associated with a dying cell is critically regulated by the central cell death machinery or is merely a consequence of the activation of this machinery. The morphological change that were found to be associated with apoptosis guided investigations into the nature of the central cell death machinery. The prominence and deleterious consequences of the nuclear changes that occur during
248
mm
J. M I N N et ni
apoptosis led to the early hypothesis that the central death machinery focused around the control of DNA condensation and DNA fragmentation. However, cells with their nuclei removed are still able to undergo apoptosis (Jacobsonet al., 1994).Thus, “nuclear death” is a consequence of apoptosis, and the irreversible step controlled by the cell death machinery is likely to involve cytoplasmic factors. Insight into the nature of these cytoplasmic factors has been obtained with the identification of genes that control programmed cell death. During the development of the nematode Caenorhabditis elegans, 1090 somatic cells are generated, and of these cells, 131 are programmed to die, Genetic analysis revealed that three genes, CED-3, CED-4, and CED9, control all 131 of these developmental programmed cell deaths (Hengartner and Horvitz, 1994b). PCD requires CED-3 and CED-4, as loss of function mutations in either of these genes prevents the death of nearly all cells that normally are programmed to die (Ellis and Horvitz, 1986). Genetic mosaic experiments demonstrate that these gene products function cell autonomously (Yuan and Horvitz, 1990). Dominant gain of function CED-9 mutations prevent all cells from dying and result in viable but functionally compromised worms. Loss of function CED-9 mutations are lethal, presumably because cells that normally should survive undergo programmed cell death (Hengartner et al., 1992).Extended genetic analysis of these genes in C. elegans show that CED-9 acts as an inhibitor of CED4 and prevents CED-4 from activating the death-inducing properties of CED-3 (Shaham and Horvitz, 1996). Thus, the programmed cell death occurs during the development of C. elegans is genetically encoded and functions in a fashion in which CED-9 inhibits CED-4, which in turn activates CED-3. A mammalian homolog of CED-9 was identified as bcl-2, a gene that was initially cloned from human B-cell lymphomas with a characteristic t(14;18) chromosomal translocation (Bakhshi et al., 1985; Cleary et al., 1986; Hengartner and Horvitz, 1994a). Studies revealed that Bcl-2 is a novel protooncogene that does not promote cell proliferation but rather promotes cell survival. When cells are subjected to a wide variety of cytotoxic stimuli that induce mammalian cell death, such as growth factor withdrawal, chemotherapeutic drugs, metabolic toxins, viral infections, and inappropriate oncogene expression, expression of Bcl-2 is able to inhibit death in each case. Mammalian Bcl-2 is able to complement CED-9 in C. elegans (Hengartner and Horvitz, 1994a; Vaux et al., 1992). Thus, in mammalian cells, a tremendous variety of cell death signals all converge on a central cell death machinery that is controlled by Bcl-2. Furthermore, this central cell death machinery is conserved from the developmental cell death program of C. elegans.
CED-3 was cloned and shown to be ho~nologousto a previously cloned maininalian gene interleubn-lp ( IL-lp) converting enzyme, or ICE (MiLira et d., 1993). ICE was initially identified as a cysteine protease that converts the 31-kDa pro form of IL-lp into a 17.5-kDa mature form. A role for this protein in cell death was confirmed when ICE was transiently introduced into a mammalian cell line and found to cause agoptosis. Many more ICE-like proteases have been cloned from inaininalian systems, and these proteases have been renamed caspases (Faucheu et nl., 1995; Kainens et d., 1995; Kumar et nl., 1994; Wang et nl., 1994). Transfection of many of these caspases into mammalian cells causes apoptosis, and elimination of some of these genes in mice by gene targeting causes defects in apoptosis (Kuida et d , 1995, 1996). All of these caspases are synthesized in a pro form and must be activated by proteolytic cleavage (Kuinar, 1995). The proteolytic activity of these proteases is directed to rare cleavage sites containing Asp-X, where X is any ainino acid. Some targets for caspases include poly(ADP)-ribose polymerase (Lazebnik et nl., 1994), nuclear lainins (Takahashi et nl., 1996),fodrin (Cryns et nl., 1996),pel-activated kinase 2 (PAK2) (Rude1and Bokoch, 1997),and DNA fragmentation factor (DFF) (Liu et d ,1997). Cleavage of some of these substrates has been proposed to irreversibly cominit a cell to PCD and induce at least some of the morphological changes associated with apoptosis. It has also been proposed that one reason for the exi5tence of multiple caspases is that some are involved in the ainplification of the response through cleavage of the pro forins of other caspases. Such early acting caspases inay include caspase 1 (ICE) (Enari et nl., 1995) and caspase 2 (Nedd2) (Harvey et nl., 1997). Other caspases nre inore distally involved in the execution of the cell death program, such as caspase 3 (CPP32) (Faleiro et a1 , 1997). Thus, much evidence points toward inembers of the caspase family as tlie executioners of tlie central cell death machinery. The critical role of caspases in the cell death program has been further highlighted by studies on various maininalian cell death receptors (Yuan, 1997). The best characterized of the death receptors belong to the TNF family of receptors and include TNF-Rl and Fas/APO- 1. These receptors seem to tie involved in preserving lymphocyte homeostasis by eliminating previously activated or inappropriately activated lymphocytes during an immune response (Abbas, 1996). The cytoplasmic domain of these receptors contains an amino acid region known a s the death domain. This domain is involved in the recruitment of caspase 8 (FLICE, Mach, Mch5), which is activated after ligation of the Fas receptor (Boldin et a1 , 1996; Muzio et nl., 1996).This finding suggests that death receptors can function through a direct link to the protease machinery that is responsible for causing cell death.
250
ANDY J MINN
et
ni
Although the mechanism that death receptors utilize to cause apoptosis seems elegantly simple, the pathways that most apoptotic signals use do not seem to be so direct. Similar to the way that CED-9 acts upstream of CED-3 in C. elegans, most apoptotic stimuli in mammalian cells eventually converge on the Bcl-2 family members to regulate a step prior to caspase activation (Chinnaiyan et al., 1996). What is the mechanism by which Bcl2 functions to prevent caspase activation and thus apoptosis? The difficulty in answering this question is partly due to the fact that many cytoplasmic changes occur in cells undergoing apoptosis, including alterations in the cellular redox state (Hockenberry et al., 1993; Kane et nl., 1993), changes in the subcellular distribution of ions (Barry et al., 1993; Lam et al., 1994), and the disruption of mitochondrial function (Zamzami et al., 1995a). By preventing apoptosis, BcI-2 may indirectly prevent all of these changes. Many proposals regarding the mechanism by which Bcl-2 functions, such as by regulating calcium fluxes or by acting as an antioxidant, have been shown likely to be incorrect, illustrating the difficulty in distinguishing between cause and effect (Jacobson and Raff, 1995; Reynolds and Eastman, 1996; Shimizu et nl., 1995). Moreover, these proposals have proved inadequate in explaining the fundamental question of how Bcl-2 inhibits caspase activation. Thus, the biochemical mechanisms utilized by Bcl-2 to regulate cell survival have remained elusive. An outline of the current understanding of the PCD pathway is presented in Fig. 1. 111. The &I-2 Family
CED-9 appears to be the only protein of its land in C. elegans. In contrast, several mammalian proteins exist that are homologs of Bcl-2 perhaps because of the complexity of mammalian organisms as compared to the invertebrate nematode. Furthermore, some of these homologs promotes rather than inhibit cell death. These death agonists and antagonists comprise the Bcl-2 family and share homology in as many as four amino acid regions denoted as Bcl-2 homology (BH)1, BH2, BH3, or BH4 (Boyd et al., 1995; Yin et al., 1994; Zha et al., 1996a). The BH1 and BH2 domains are approximately 21 and 16 amino acids long, respectively, and are generally separated by 30-40 amino acids. The BH4 domain, which is located at the N terminus, is generally found in the protein that inhibit cell death. The BH3 domain is found in all family members, but appears to be especially important in family members that promote cell death because some death agonists lack all of the homology domains except the BH3. Figure 2 diagrams the architecture of some of the Bcl-2 family members discussed in this review.
25 1
REGULATION OF i\POPTOSIS BY 13~1-2 FAMILY MEMBERS
A
The Genetics of Programmed Cell Death in C. elegans
CED-9
B
-1
CED-4 4 CED-3 4 Cell Death
The Pathway of Programmed Cell Death in Mammals
Death Receptors Growth Factor Withdrawal Cell Cycle Perturbations DNNMetabolie Damage
Central Cell Death Signal
- Apaf-1 (CED-4) Activation
Caspase (CED-3) Activation
Endonuelease Activation
-
f
cellSurface Alterstions
-
Cell Death and PhagoWosis
Cytoskeietal Reorganization
BcI-2 BcI-x,
Bax Bak
1 Bcl-2 Family Members (CED-9) FIG.1. Suniinary of the pathways of prograinmed cell death in C elegans and in maini d s . See text for details.
A. PROTEINS THATINHIBITAPOPTOSIS The first hoinolog of Bcl-2 was cloned using low stringency hybridization from chicken libraries. This approach yielded a homologous gene named bcl-x (Boise et al., 1993), bcl-x was subsequently cloned from human libraries where it was discovered that bcl-x has two alternatively spliced forms, bcl-xL and bcZ-xs. Bcl-xLhas the largest open reading frame of 233
252
ANDY
1. M I N N
et al.
FIG.2. Schematic representation of the general architecture of Bcl-2-family members. Bcl-2 homology (BH) domains and the carboxy-terminal transmembrane (TM) domain are labeled. On the right of each structure are representative Bcl-2 families that share the depicted structure. Diagram is not drawn to scale.
amino acids and, like Bcl-2, protects cells against multiple apoptotic stimuli. Although the predicted molecular mass of Bcl-xL is 26 kDa, the protein migrates aberrantly slowly on an SDS-PAGE gel at 30 kDa. In Bcl-xs, 63 amino acids containing the BH1 and BH2 regions are removed by alternative splicing and, in contrast to Bcl-xL,result in a protein that promotes cell death. Bcl-xs will be discussed in more detail later. Bcl-x also has a third form called Bcl-xs, which is a 209 amino acid protein that results from an unspliced transcript (Gonzalez-Garcia et al., 1994). Bcl-xpcontains a unique stretch of 21 amino acids at the C terminus and lacks the 19 hydrophobic amino acids that anchor BcI-xL to membranes. A similar product that results from an unspliced transcript has also been described for Bcl-2 (Tsujimoto and Croce, 1986). Other members of the Bcl-2 family that have been shown to inhibit apoptosis include Mcl-1 (Zhou et al., 1997), A1 (Lin et al., 1996), and Bcl-w (Gibson et al., 1996). The existence of multiple Bcl-2 family members in mammals is likely due to the need to regulate cell death in a tissue-specific manner or to allow cells to rapidly alter their apoptotic threshold in response to changing environmental signals. Consistent with this, it has been found that Bcl2, BcI-xL, Bcl-w, Mcl-1, and A1 exhibit differences in tissue expression, developmental expression, and inducibility in response to extrinsic stimuli (Boise et al., 1995; Gonzalez-Garcia et al., 1995; Krajewski et al., 1994, 1995; Lin et al.,1993; Ma et d., 1995; Yang et al.,1996). Germline elimination of bcl-x: results in lethality at embryonic day 13 associated with massive neuronal and hematapoietic cell death (Motoyama et al., 1995).
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY M E M B E R S
253
bcl-x knockout chimeric mice also show survival defects for immature but not for peripheral lymphocytes (Maet al., 1995).In contrast, bcl-2 knockout mice complete embryonic development but demonstrate several abnormalities postnatally (Nakayainaet al., 1994; Veis et at., 1993).These abnormalities include the death of resting peripheral lymphocytes but not of immature or activated lymphocytes. The different results from the bcl-x and bcl-2 knockout mice are consistent with the developmental expression patterns of each gene in the nervous system and in lymphocytes. Although elimination of bcl-x results in the loss of both Bcl-xLand Bcl-xs, the phenotype is thought to result predominantly from the absence of BcI-xL because only defects in cell survival are observed and because Bcl-x, expression is extremely low in mouse tissue (Gonzalez-Garcia et al., 1994). Finally, homologs of the death-inhibiting Bcl-2 family have also been found in proteins encoded by DNA viruses. The best characterized is the weakly homologous adenovirus protein E1B-19k (Han et al., 1996). Other examples include BHRFl from EBV (Henderson et al., 1993), LMWFHL from the African swine fever virus (Neilan et al., 1993), and KSbcl-2 from Kaposi sarcoma-associated herpesvirus (HHV8) (Cheng et nl., 1997). The purpose of these Bcl-2 homologs may be to counter the propensity of host cells to undergo apoptosis while infection is being established. APOPTOSIS B. PROTEINS THATPROMOTE One of the first pro-apoptotic Bcl-2 family members to be characterized was Bax (Oltvai et al., 1993). Bax is a 21-kDa protein that was cloned based on its ability to coirninunoprecipitate with Bcl-2. Bax, in harboring a BH1, BH2, and BH3 domain, shares extensive amino acid homology with other Bcl-2 family members. Despite this extensive homology, Bax either accelerates apoptosis in response to a death stimulus or is able to kill cells directly when transfected into mammalian cells. When Bcl-2 is coexpressed with Bax, cells become resistant to this pro-apoptotic effect. In the absence of Bcl-2, Bax forins homodimers, but in the presence of Bcl-2, Bax and Bcl2 forin heterodimers with each other. Bax also interacts with and counters the protective properties of Bcl-xL. When mutations were introduced into either the BH1 or the BH2 domain of Bcl-2 or BcI-xL, heterodiiner formation was prevented, suggesting that Bcl-2 and BcI-xL both interact with Bax through the BH1 and BH2 regions (Sedlak et al., 1995; Yin et al., 1994). One model, referred to as the rheostat model, to explain the role of heterodimerization proposes that the amount of Bax that is not heterodiinerized to Bcl-2 (or Bcl-xL)sets the apoptotic threshold of a cell (Yang and Korsmeyer, 1996). The greater the amount of uncoinplexed Bax, the lower the apoptotic threshold.
254
ANDY J , MINN af ol
Interaction cloning with death antagonists and degenerate polymerase chain reaction cloning to Bcl-2 homology domains resulted in the isolation of Bak, another Bcl-2 family member that promotes apoptosis (Chittenden et al., 1995b; Farrow et al., 1995; Kiefer et al., 1995). Bak is a 26-kDa protein that also contain BH1, BH2, and BH3 domains. In general, it is functionally similar to Bax. Bak interacts with death antagonists such as Bcl-2, Bcl-xl,, and E1B-19k and inhibits their survival properties. In addition, Bak is able to kill cells directly when overexpressed in certain systems. Interestingly, Bak is also able to inhibit cell death under some circumstances, suggesting that its effect on cell survival is context dependent. Although Bax and Bak share the BH1, BH2, and BH3 doinains with death antagonists, the BH3 region has been suggested to be critical in conferring pro-apoptotic properties. This is because the BH3 domain of Bak was found to be sufficient to both cause apoptosis and heterodimerize with Bc~-xL(Chittenden et al., 1995a). In addition, when the BH3 domain of Bax was added to Bcl-2 by mutagenesis, the resulting hybrid molecule promoted apoptosis (Hunter and Parslow, 1996). Consistent with the importance of the BH3 domain in the death-promoting Bcl-2 family members, several pro-apoptotic Bcl-2 family members have been cloned that share no amino acid homology to other Bcl-2 proteins except for the BH3 domain. These include Bik (Boyd et al., 1995), Bid (Wang et al., 1996b), and Hrk (Inohara et al., 1997). Bad is a death agonist that was originally identified to have weakly homologous BH1 and BH2 domains (Yang et aE., 1995), but recent data demonstrate that it is likely also a BH3-only protein (Kelekar et al., 1997; Zha et al., 1997). The proapoptotic function of these proteins generally correlates with the ability to heterodimerize with Bcl-2 and/or BcI-xL. Thus, heterodimerization with the death agonists seems to involve their BH3 domains. However, it is unclear how this domain promotes cell death. One idea is that BH3 domains of death agonists can act as competitive diinerization substrates with Baxl Bak and bind to BcI-~/Bc~-xL to displace prebound Bax/Bak. According to the rheostat model, this would result in an overall increase in free Baxl Bak, leading to cell death. In support of this idea, Bad is only able to promote cell death if it can interact with Bcl-xL. The alternatively spliced form of Bcl-x, Bcl-xs, was originally characterized as being a pro-apoptotic protein (Boise et al., 1993). When cointroduced with Bcl-2, Bcl-xs was able to inhibit the antiapoptotic effects of Bcl-2 and BcI-xL. Bcl-xs lacks the BH1 and BH2 domains due to alternative splicing; however, it retains the BH3 domain and is the only pro-apoptotic Bcl-2 protein with a BH4 domain. Yeast two-hybrid studies showed that Bcl-xs is able to heterodimerize with Bcl-2 or BcI-xL (Sato et al., 1994;
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
25s
Sedlak et al., 1995). However, these results may be artifactual as this interaction is not seen in mammalian cells (Minn et al., 1996).
C SUBCELLULAR LOCALIZATIOK OF Bcl-2 FAMILY MEMBERS Bcl-2 and Bcl-xL,localize to similar subcellular compartments. Electron microscopy has revealed that these proteins can be found in the outer membrane of the mitochondria, the outer nuclear envelope, and the endoplasmic reticulum (Akao et al., 1994; Krajewsh et al., 1993; Lithgow et al., 1994). Furthermore, the staining pattern of Bcl-2 on mitochondrial membranes is patchy, suggesting that it can localize to contact sites, which are regions where the mitochondrial outer and inner membranes meet. Targeting to the outer mitochondrial membrane seems to depend on the carboxy-terminaltransmembrane tail. In Bcl-2, this 19 amino acid sequence is necessary and sufficient to direct the protein to outer membranes of isolated mitochondria through an ATP-dependent, temperature-sensitive process (Nguyen et al., 1993). The C-terminal transmembrane sequence of Bcl-2 can be replaced by a similar sequence from the yeast outer mitochondrial membrane protein Mas70p. This chimeric protein retains full antiapoptotic function and suggests that the C terminus of Bcl-2 may simply function to target Bcl-2 to correct intracellular locations. Deletion of the C terminus of Bcl-2 either abrogates or diminishes its antiapoptotic properties, depending on the experimental system (Hockenberry et al., 1993; Nguyen et al., 1994). Eliniination of the C terminus of Bcl-xL also interferes with its antiapoptotic effects (A. J. Minn, B. S. Chang, and C. B. Thompson, unpublished data). These results argue that targeting of Bcl-2 and Bcl-xL to the mitochondria is important for their abilities to regulate programmed cell death. Other reports suggest that there may not be an absolute requirement for membrane attachment of Bcl-2 and Bcl-xLfor antiapoptotic function. Deletion of a C-terminal region from Bcl-2 that includes the membrane anchor domain was reported to have no effect on survival function when assayed using cell death induced by nerve growth factor (NGF)withdrawal or tumor necrosis factor (TNF) treatment (Borner et al., 1994). Similarly, microinjection of Bcl-x, into symphathetic neurons followed by NGF deprivation resulted in protection against cell death that was comparable to that exhibited by Bcl-x,. (Gonzalez-Garcia et al., 1995). When Bcl-2 was retargeted to the endoplasinic reticular (ER) membrane using the ActA insertion sequence it was still able to prevent some forms of cell death but not others (Zhu et al., 1996). However, it is difficult to interpret experiments that attempt to target Bcl-2 away from the mitochondria because Bcl-2 family members heterodimerize with each other. Therefore, mutants or variants of Bcl-2 that lack the transmembrane tail may still
256
ANDY J. MINN et al.
retain significant mitochondrial association through interaction with other mitochondria-associated Bcl-2 proteins. Alternatively, nonmitochondrially localized Bcl-2 may be able to inhibit cell death by sequestering deathpromoting Bcl-2 family members away from the mitochondria. Yet another possibility is the Bcl-2 has antiapoptotic properties that are independent of mitochondrial association. Bcl-2 has been reported to influence calcium flux from the endoplasmic reticulum and the nucleus, as will be discussed later. In general, however, most studies indicate that mitochondrial localization is important for the full antiapopotic effect of the protein. The death agonists that contain BH1, BH2, and BH3 domains, such as Bax and Bak, also contain a carboxy terminus transmembrane domain that targets these proteins to the outer mitochondrial, outer nuclear, and ER membranes (Krajewski et al., 1996; Yang and Korsmeyer, 1996). However, unlike Bcl-2 and Bc~-xL,which seem to reside predominantly at these intracellular membranes, evidence shows that Bax can be a predominantly cytosolic protein and is capable of being targeted to the intracellular membranes after an apoptotic signal (Hsu et al., 1997). How this process is controlled is unclear. As in the case of Bcl-2 and BcI-xL, the absence of a transmembrane tail from Bax compromises but does not completely eliminate function (Antonsson et al., 1997; Zha et al., 1996b). Many of the death agonists, particularly the BH3-only versions, do not have a transmembrane anchor and are thought to be cytosolic. These proteins include Bid and Bad. However, the localization of these proteins, and thus their pro-apoptotic function, may be under the regulation of cell survival signals. Bad has been shown to undergo phosphorylation in the presence of growth factors such as IL-3 (Zha et al., 1996~). The phosphorylated form of Bad is subsequently sequestered in the cytosol through binding to 14-3-3 proteins, molecules that recognize phosphoserine residues. Upon removal of growth factor, Bad is dephosphorylated and changes partners from 14-3-3 to Bcl-xL. Binding to BcI-xL not only redistributes Bad to the mitochondrial membrane, but is also inactivates the cell survival function of Bcl-xL and/or displaces another death agonist such as Bax, leading to the rapid death of the cell. The proto-oncogene Akt, a serinethreonine kinase that phosphorylates PI 3-kinase and is involved in the cell survival signal delivered by growtWsurvival factor receptors (Franke et al., 1997), has been shown to phosphorylate Bad and may be one explanation for the antiapoptotic effects of PI 3-kinase/Akt (Datta et al., 1997; del Peso et al., 1997).
D. WHICHIs THE EFFECTOR AND WHICHIs THE REGULATOR? Although the extent of heterodimerization has been proposed to be important in setting the cellular apoptotic threshold, the question remains how Bcl-2 family members function to regulate cell survival. Does Bax
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
257
have a unique biochemical property that allows it to kill cells, and does Bcl-xLinhibit cell death simply by acting as a regulatory protein that ties up Bax? Another equally possible, although not mutually exclusive, model is that Bcl-x, has a unique biochemical property that directly promotes cell survival, and Bax simply inactivates this property through heterodimerization. Unfortunately, not only are the critical regulatory steps of cell death difficult to study due to the pleiotropic changes that occur in dying cells, but also the precise biochemical function of Bcl-2 family members in controlling this ill-defined step is elusive due to the extensive dimerization that occurs with family members and nonfamily members. Despite this, recent data may be unraveling the biochemical mechanism(s) by which Bcl-2 family members control apoptosis.
N. Mitochondria Can Control Apoptosis The localization of Bcl-2 to the outer mitochondria membrane suggests that Bcl-2 can regulate cell death through a mitochondria-dependent mechanism. While initial findings that cells without mitochondrial DNA are able to undergo apoptosis and are protected by Bcl-2 led to the interpretation that mitochondria are not essential in the control of PCD (Jacobson et al., 1993), more recent data, however, suggest that this organelle may play a central role in the cell death pathway. Although many changes are observed in cells undergoing apoptosis, such as DNA fragmentation, generation of reactive oxygen species, and alterations in plasma membrane lipid composition, one of the earliest identifiable events in the loss of mitochondrial transmembrane potential (Kroemer, 1997).This electrochemical grahent across the inner mitochondrial membrane results mainly from the pumping of protons out of the matrix. This transmembrane potential is essential for mitochondrial bioenergetics, such as the production of ATP by oxidative phosphorylation. Lipophilic cationic dyes partition into the mitochondrial matrix according to the Nernst equation and, therefore, can be used as a measurement of the transmembrane potential. Cells undergoing apoptosis show very early signs of a reduced mitochondrial transmembrane potential (Zamzamiet al., 199513). Cells that have lost their transmembrane potential are irreversibly committed to an apoptotic demise, which has led to the suggestion that apoptosis-induced alterations to the mitochondria involve the “point of no return.” One way mitochondria lose their transmembrane potential is through a phenomenon known as permeability transition (PT) (Zoratti and Szabo, 1995). This process involves the sudden opening of a large permeability pathway across the inner membrane, leading to the loss of the transmem-
258
ANDY J . M I N N et
(I/.
brane potential and swelling of the mitochondria. The permeability pathway results from the opening of the permeability transition pore. The exact composition of this “megachannel” is unclear; however, it is thought to exist at contact sites, where the inner and outer mitochondrial membranes, adjoin, and to consist of both inner membrane channels, such as the adenine nucleotide translocator, and outer membrane channels, such as the voltage-dependent anion channel (VDAC) and the peripheral benzodiazapine receptor (Kroemer, 1997). Much data suggest that the opening of the PT pore is sensitive to the redox status of certain proteins, the redox status of pyridine nucleotides, the presence of calcium, the matrix pH, and the concentration of adenine nucleotides (Zoratti and Szabo, 1995). Thus, the PT pathway has been suggested to be a wide-spectrum sensor for the cellular environment, perhaps making it a suitable monitor for apoptotic signals. Additionally, PT is a process that exhibits positive feedback since the disruption of mitochondrial function leads to the generation of many of the signals that contribute to PT in the first place, such as reactive oxygen species and increased cytosolic calcium. The potential role that PT plays in programmed cell death was demonstrated by the use of a cell-free system that recapitulates the essential features of apoptosis. When isolated mitochondria are induced to undergo PT, added nuclei undergo morphological changes characteristic of apoptosis (Zamzami et al., 1996). Isolated mitochondria containing Bcl-2 are inhibited from undergoing PT induced by some (but not all) agents, and known PT inhibitors, such as cyclosporin A and bongkrekic acid, can inhibit apoptosis when added to whole cells. One mechanism that explains how PT leads to apoptosis involves the release of a mitochondrial protein called apoptosis initiating factor (AIF) (Susin et al., 1996).AIF is able to activate the downstream caspase, caspase 3 (CPP32), and cause apoptotic changes to isolated nuclei. The inhibition of PT by pharmacological PT inhibitors or by Bcl-2 prevents the release of AIF. In addition, AIF may be another Gactor contributing to the self-amplification of PT, as AIF is also a potent PT inducer. Thus, in response to apoptotic signals, PT may result in the release of AIF and lead to the activation of downstream apoptotic effector functions. Bcl-2 may act directly, or indirectly, to prevent PT. Consistent with a role for Bcl-2 family members in controlling PT, overexpression of Bax also seems to cause PT in vivo (Xiang et al., 1996). Further support that disruptions in mitochondrial homeostasis are critical to the initiation of apoptosis derive from the demonstration that the release of cytochrome c by mitochondria activates caspase 3 and results in subsequent apoptotic changes to nuclei (Liu et al., 1996).Cytochrome c normally resides between the outer and the inner mitochondrial membranes to carry electrons between complex I11 and complex IV of the electron transport
KEGULATION OF .4POPTOSIS BY Bcl-2 FAMILY MEMBERS
259
chain. Cytochrome c is imported into the intermembrane space as apocytochrome c, but is then subsequently modified to the heme-containing holocytochrome c, a form that normally does not traverse back to the cytosol (Dumont et al., 1991).Unlike AIF, cytosolic cytochroine c seems to require additional cofactors, like dATP, in order to activate downstream caspases and cause apoptotic clianges. Bcl-2 can act in situ on the mitochondrial membrane to prevent the release of cytochrome c (Kluck et al., 1997; Yang et al., 1997). Furthermore, the addition of exogenous cytochrome c bypassed the inhibitory effects of Bcl-2. Thus, these data suggest that the release of cytochrome c from the mitochondria is another mechanism to activate the downstream apoptotic machinery, and Bcl-2 inhibits this activation by preventing the release of cytochrome c. What mechanism is responsible for the release of cytochroine c? In contrast to the release of AIF, which seems to result after the PT-induced loss of mitochondrial transmembrane potential, cytoclirome c can be released from mitochondria that retain a transmembrane potential as measured by the uptake of cationic lipophilic dyes. Cells undergoing apoptosis exhibit swelling of the mitochondria prior to the loss of mitochondria1 transmembrane potential (Vander Heiden et al., 1997). This swelling, as determined by electron microscopy, outer membrane permeability studies, and direct measurement of mitochondrial size, leads to an increase in the matrix volume and a physical disruption of the outer mitochondrial membrane. The release of cytochrome c from the intermembrane space is also observed coincidental with the alterations in initochondrial structure. Mitochondria1 swelling and the release of cytochrome c are followed by the loss of initochondrial transmembrane potential. In cells expressing Bclxl. mitochondria do not swell and cytochrorne c is not redistributed to the cytosol. These data suggest that apoptotic stimuli lead to the loss of' mitochondrial volume control which, in turn leads to the rupture of the outer mitochondrial membrane and the release of cytochrome c froin the intermembrane space. Presently, it is unclear how these mitochondrial events leading to cytochrome c release relate to PT and AIF. The ability of pharmacological PT inhibitors to inhibit apoptosis in cell culture argues for the primary importance of at least the PT pore; therefore, it would be interesting to determine if PT inhibitors, in addition to preventing the release of AIF, are able to prevent the release of cytochrome c. One possibility is that the release of cytochrorne c and the release of AIF are on a sequential pathway whereby AIF is released after cytochrome c. Regardless, a consistent observation is that Bcl-2 and Bcl-x,, are able to prevent alterations in mitochondrial function that would otherwise result in the release of mitochondria1 proteins that activate downstream caspases.
260
ANDY J. M I N N et nl.
A. IDENTIFICATION OF A MAMMALIAN CED-4 HOMOLOG Genetic evidence from C. elegans indicates that CED-9 is an upstream inhibitor of a pathway in which CED-4 activates CED-3 to cause PCD. The ability of Bcl-2 (CED-9 homolog) to prevent the release of mitochondria1 proteins that activate caspase 3 (CED-3homolog) provides one mechanism to explain the genetic evidence from C. elegans. Several studies have provided insight on how CED-4 might act between CED-9 and CED-3 as a positive regulator of PCD (Chinnaiyan et al., 1997a,b; Seshagiri and Miller, 1997; Wu et al., 1997a). CED-3, like its mammalian homolog, is a cysteine protease that is synthesized in a pro form that can be activated proteolytically (Xue et al., 1996).When CED-3 is transfected into mammalian or insect cells, it can promote apoptosis, albeit weakly, and although one group reported that CED-4 is able to cause apoptosis when transfected alone, the consensus is that CED-4 does not cause apoptosis when introduced by itself. However, when CED-4 and CED-3 are coexpressed, a synergistic effect on cell death is observed. These results can be explained by the findings that CED-4 is able to interact with and enhance the proteolytic activation of CED-3, resulting in an increase in CED-3mediated apoptosis. The ability of CED-4 to enhance CED-3 proteolysis and activation requires a nucleotide-binding niotif in CED-4. Furthermore, CED-4 complex formation with CED-3 involves the CED-3 prodomain. Mutations or deletions of the CED-3 prodomain prevent interaction with CED-4 and inhibit the effects of CED-4 on CED-3. CED-9 has also been shown to interact with CED-4 (Chinnaiyan et al., 1997b; James et al., 1997; Spector et nl., 1997; Wu et al., 1997a,b).When CED-4 is coexpressed with CED-9 and CED-3, it forms a trimolecular complex with CED-9 and CED-3, but is unabIe to enhance CED-3 proteolytic activation, suggesting that CED-9 binding interferes with the activity of CED-4 and CED-3 (Seshagiri and Miller, 1997; Wu et al., 1997a). Furthermore, when CED-9 is expressed together with CED-4 and CED3 in either insect cells or mammalian cells, CED-9 inhibits the apoptosis that results from the synergism between CED-4 and CED-3. The ability of CED-4 to regulate the proteolytic activation of CED-3 and to synergize with CED-3 to cause PCD indicates that a mainmalian CED-4 homolog might be involved in the cytochroine c-mediated activation of caspase 3. This prediction was confirmed with the biochemical purification of Apaf-1 and the demonstration that it is one of the additional cofactors required by cytochrome c to activate caspase 3 and to cause apoptosis in a cell-free system (Zou et al., 1997). Apaf-1 is a 130-kDa protein that contains a 320 amino acid region that is homologous to CED4. Interestingly, the amino-terminal 85 amino acids also show homology
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
261
to the prodomain of CED-3. The carboy terminus contains 12WD repeats, which are thought to mediate protein-protein interactions. When transfected into mammalian cells, Apaf-1 does not cause cell death; however, when extracts are made from these cells, the addition of dATP, another cofactor required for cytochrome c-dependent caspase 3 activation, leads to the rapid activation of caspase 3. Apaf-1 binds directly to cytochrome c and also contains a nucleotide-binding motif, perhaps explaining the requirement for dATP. It has yet to be determined whether Apaf-1 interacts with any of the death-inhibiting Bcl-2 family members. OF Bcl-2 ON CELLPHYSIOLOGY B. OTHEREFFECTS
One explanation for how Bcl-2 is able to regulate mitochondrial function and the distribution of mitochondrial proteins is through the regulation of membrane permeability. This putative function is also suggested by studies in which Bcl-2 was able to decrease the efflux of calcium from the endoplasmic reticulum after treatment with thapsigargin, a specific inhibitor of the ER calcium pump (Lam et al., 1994). Because Bcl-2 is also localized on the outer nuclear envelope and the nucleus is another site for calcium storage, the effect of Bcl-2 on nuclear calcium flux was also examined. In response to various stimuli, both nuclei in .situ and isolated nuclei were found to exclude calcium in a Bcl-2-dependent manner (Marin et al., 1996). Together, these results suggest that Bcl-2 is able to influence the membrane permeability of the intracellular membranes to which it distributes. V. Structure/Function Studies of &I-&
To begin to understand how Bcl-2 family members interact and what their potential biochemical functions might be, the three-dimensional structure of BcI-XL was solved by a combination of nuclear magnetic resonance (NMR) and X-ray crystallography techniques (Muchmore et d., 1996). BcI-xL is a predominantly (Y helical protein that consists of two central hydrophobic helices surrounded by five amphipathic hFlices (Fig. 3A). The central hydrophobic helices are approximately 30 A long and correspond to parts of the BH1 and BH2 domains. BH3, BH4, and parts of BH1 and BH2 contribute to the five surrounding amphipathic helices. An elongated hydrophobic cleft is formed on the surface of Bcl-xL by the juxtapositioning of BH1, BH2, and BH3 (Fig. 3B). The structure of BclxL also reveals a flexible, unstructured region between BH4 and BH3 that is dispensable for antiapoptotic function and seems to act as a negative regulatory domain (Chang et al., 1997; Uhlmann et al., 1996). Due to the extensive amino acid homology between Bcl-xL and other members of the
KEGULATION OF APOF'TOSIS BY Bcl-2 FAMILY MEMBERS
263
Bcl-2 family, these other family members are also predicted to have a similar tertiary structure. A THEBH3 DOMAIN OF DEATH AGONISTSINTERACTS WITH A HYDROPHOBIC CLEFTFORMED BY THE BH1, BH2 A N D BH3 DOMAINS OF BcI-xL The hydrophobic cleft formed by the BH1, BH2, and BH3 domains of Bcl-x, is the site for interaction with other Bcl-2 family members. Using NMR, it was determined that the BH3 domain of Bak forms an Q helix and binds to the BcI-xL hydrophobic cleft (Sattler et al., 1997).This complex is stabilized through various hydrophobic and electrostatic interactions involving highly conserved residues from the BH1, BH2, and BH3 domains of BcI-xL along with highly conserved amino acids from the BH3 domain of Bak. The nature of this interaction between BcI-xL and various BH3-containing death agonists was confirmed by peptide competition experiments and mutations in either full-length Bcl-x, or full-length BH3containing death agonists. Because Bax and Bak are predicted to have the same overall structures as BcI-XL in the absence of heterodimerization, the BH3 domain of the death agonists likely forms intramolecular interactions and participates in the formation of a Bax/Bak hydrophobic cleft. Thus, these intramolecular interactions formed by the BH3 domain of Bax and Bak would need to be disrupted in order for the BH3 domain to form new interactions with the hydrophobic cleft of BcI-xL. Binding of the BH3 domain to Bcl-x, would also require the rotation of the BH3 domain along its helical axis. Details concerning these structural requirements and what might induce these conformational changes are unclear; however, the potential need for Bcl-2 family members to undergo changes is discussed later. Nonetheless, heterodimerization between pro-apoptotic and antiapoptotic Bcl-2 Family members has been suggested to be an important property for Bcl-2 family proteins to regulate cell survival,and these structural studies provide important insight into this property.
B. Bcl-2 FAMILY MEMBERSSHARESTRUCTURAL HOMOLOGY TO BACTERIAL PORE-FORMING DOMAINS The structure of Bcl-xLbears significant similarity to the pore-forming domain found in various bacterial toxins, particularly diphtheria toxin and FIG.3. The structure of Bcl-xL.(A) Rihbon depiction of the three-dimensional structure of Bcl-xL showing that the protein consists of two central hydrophobic helices (a5and a6) surrounded by five amphipathic helices. Each alpha helix is labeled, as are the regions corresponding to the BH1, BH2, and BH3 domains. The unstructured loop between amino acids 26 and 76 is not sliown. (B) A space-filled depiction of Bcl-x,~showing the hydrophobic cleft (light gray) farmed by the BIIl, BII2, and BH3 domains.
264
ANDY J. M I N N et al.
the colicins. These pore-forming domains contain two central hydrophobic helices encased by five to eight amphipathic helices (Parker and Pattus, 1993).These domains can be induced to undergo a conformational change from a water-soluble to a membrane-inserted configuration. Bacterial colicins are plasmid-encoded proteins that are secreted by bacteria to kill other bacteria (Cramer et al., 1995). Colicins bind bacterial receptors in the outer membrane to translocate to the inner membrane. Through mechanisms that are not well understood, colicin pore-forming domains are induced to undergo a water-soluble to membrane-bound alteration in structure that results from the insertion of its two central hydrophobic helices. The membrane-bound molecule is then thought to form an ion channel in response to the transmembrane potential of the inner membrane by inserting two other helices to create a four-helix bundle. The creation of an ion channel leads to the depolarization of the bacteria, deenergization, and cell death. Diphtheria toxin in another molecule that contains a pore-forming domain (London, 1992). This molecule enters cells through receptor-mediated internalization of endosomes. Once in the endosome, the low pH is thought to induce the conformational change that allows the pore-forming domain to create an ion channel. This pore is thought to facilitate the translocation of the diphtheria toxin A fragment into the cytosol where it can ADP-ribosylate protein synthesis elongation factor 2 (EF-2) and inhibit protein synthesis. Unlike the colicins, the formation of a pore by the diphtheria toxin pore-forming domain may not occur through the membrane insertion of four helices from an individual molecule because a peptide that corresponds to the two central hydrophobic helices can mimic the ion channel properties of the entire pore-forming domain (Silverman et al., 1994).This finding, along with studies that analyzed the conductance of the channel in relation to protein concentration, argues that more than one molecule is needed to form an ion channel. However, studies suggesting that the diphtheria toxin ion channel may be monomeric leave this issue unsettled (Huynh et al., 1997). In summary, structural studies indicate that BcI-xL consists of two central hydrophobic helices that are surrounded by five amphipathic helices. This structure closely resembles the pore-forming domain found in several bacterial toxins and suggests that BcI-xL may also possess the ability to form pores in biological membranes. The ability of Bcl-xLto heterodimerize with BH3-containing death agonists is the result of interactions with a hydrophobic cleft formed by the BH1, BH2, and BH3 domains of BcI-xL. Finally, a large unstructured region is found between BH4 and BH3 domains of Bcl-xLand may serve as a negative regulatory domain.
REGULATION OF APOFTOSIS BY BcI-2 FAMILY MEMBERS
265
C Bc~-xL, Bcl-2, AND Bax CANFORM IONCHANNELS Permeability transition, the release of AIF, and the release of cytochrome c are all indications that mitochondria undergo alterations in membrane permeability as a result of apoptotic stimuli. The ability of Bcl-2 to prevent these alterations suggests that it might directly or indirectly function to control membrane permeability. One way that membrane permeability can be controlled is through ion channels. Therefore, the structural similarity between BcI-XL and bacterial pore-forming domains prompted functional studies aimed at determining whether Bcl-xL and other Bcl-2 family members could also form ion channels. Using planar lipid bilayers and synthetic lipid vesicles, BcI-xL, Bcl-2, and Bax were all found to form ion channels that are voltage and/or pH sensitive (Antonssonet al., 1997; Minn et al., 1997; Schendel et nl., 1997; Schlesinger et nl., 1997). Low pH was shown to promote ion channel formation by these proteins, similar to what is observed with the bacterial pore-forming domains. Low pH is thought to facilitate ion channel formation by inducing the conforinational change that exposes the two central helices to allow for membrane insertion. In addition to insertion, the selectivity of the ion channels formed by Bcl-2, BcI-xL, or Bax is influenced by pH. At more neutral pH values, the ion channels formed by BcI-xL and Bcl-2 are more selective for cations than for anions, whereas Bax is reported to be either mildly cation selective or anion selective. At lower pH values, Bax is reported to have enhanced selectivity, whereas Bcl-2 and Bcl-xL seem to lose the ability to discriminate between cations and anions. These results may be due to the protonation of different amino acid residues that participate in the ion selectivity filter in these molecules. The ion channels formed by the pro-apoptotic and antiapoptotic proteins also show differences with regard to their voltage dependence. The Bcl-2 and B c ~ - x L ion channels behave in an Ohmic fashion, whereby current responds linearly with voltage. In contrast, Bax displays slight rectification in response to positive voltages. The ion channels formed by Bcl-xL,Bcl-2, and Bax all display multiple conductance states with complex opening kinetics. In general, these proteins form ion channels that can each range in conductance from a few picosieinens to over a nanosiemen. These ion channels can be predominantly open, predominantly closed, or exhibit flickering behavior. Additional studles with BcI-xL indicate that it integrates stably into lipid membranes and allows the passage of large organic cations (A. J. Minn, M. F. Fill, and C. B. Thompson, unpublished data). One explanation for this collective behavior is that these Bcl-2 family members are able to oligomerize in the membrane to form larger and smaller ion-conducting units.
266
ANDY J. MINN et al.
However, it cannot be ruled out that individual molecules of the Bcl-2 family are adapting distinct conformations to give rise to the complex properties of the ion channel. In summary, Bcl-2, BcI-xL, and Bax form ion channels in lipid membranes that have distinct properties, which include ion selectivity, conductance, and voltage dependence, It is presently unclear whether the ability to form ion channels with different characteristics is related to the ability of these proteins to differentially regulate cell survival. VI. How Do Bcl-2 Family Members Regulate Cell Survival?
The propensity for heterodimerization between pro-apoptotic and antiapoptotic family members has led to the suggestion that heterodimerization is an important mechanism for regulating the cell survival function of Bcl2 f'amily members. Mutants of Bcl-2 and Bcl-xLhave been described that alter amino acid residues in the BH1 and/or BH2 regions and fail to bind to Bax. When these mutants were assayed for their ability to protect cells from apoptosis, a correlation was found between the inability to heterodimerize with the failure to protect (Sedlak et al., 1995; Yin et al., 1994). Mutants of Bax also have been constructed that fail to hmerize with Bcl-xl, but retain the capacity to antagonize the protective effects of BcI-xL (Simonian et al., 1996). Furthermore, Bax and Bak are cytotoxic to yeast cells, an organisin that contains no known Bcl-2 proteins (Ink et al., 1997; Jurgensmeier et al., 1997; Zha et al., 199613). Cotransformation of Bax or Bak with wild-type Bcl-2 or BcI-xL, but not mutants that fail to bind Bax/Bak, suppresses this yeast toxicity. These data argue that Bax is a direct effector that kills cells and that Bcl-xl/Bcl-2 function to bind and inactivate Bax. It has been suggested, however, that Bcl-2 can function independently of heterodimerization. Additional mutants of BcI-xL have been described that fail to bind to Bax or Bak but still retain a majority of their protective properties (Cheng et al., 1996). Likewise, KSbcl-2, a viral inhibitor of apoptosis, fails to interact with Bax or Bak (Cheng et al., 1997). Heterodimerization-independent effects were also seen in yeast cells transformed with wild-type Bcl-2 (Longo et al., 1997). It was reported that expression of Bcl-2 can protect yeast that normally die when cultured in spent media for several days or when placed under conditions of oxidative stress. One explanation for these different findings concerning the importance of heterodimerization is provided by structural data on the Bcl-xJBak BH3 complex. Nearly all of the described Bcl-2 and BcI-XLmutants that disrupt heterodimerization were engineered in residues that are not predicted to
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBEHS
267
be directly involved in interactions with the BH3 domain of death agonists. Thus, unanticipated structural alterations could have been introduced, explaining why some nonbinding mutants are protective while others are not. This concern is supported by findings that show that some of these mutants of Bcl-2 and Bcl-xl, actually do bind to a truncated Bax protein that retains its BH3 domain (Ottilie et al., 1997a). The defect in these mutants may be multifaceted, leading to secondary defects such as the inability to heterodinierize with Bax. With the acquisition of structural information on the heterodiinerization of Bcl-xLwith the Bak BH3 domain, new mutants that disrupt critical interacting residues may help to better define the role of heterodherization. In general, however, present data would argue that heterodimerization contributes, at least partly, to the ability of Bcl-2 family members to regulate cell death. If heterodiinerization is a method to regulate the activity of Bcl-2 family members, what is the biochemical nature of this activity? The finding that Bcl-2 family members can form ion channels suggests that this biochemical property may be important in the ability of Bcl-2 family members to control apoptosis. Bcl-2 and Bcl-x12,which localize to the outer mitochondrial membranes, can prevent swelling and other alterations in mitochondrial properties both in isolated mitochondria, cell-free systems, and in viuo. One possibility is that many of these apoptosis-associated alterations of the mitochondria result in swelling due to perturbations in the mechanism that mitochondria utilize to maintain volume homeostasis. The permeability transition pore, or components thereof, has been proposed to regulate mitochondrial volume. Opening of this pore leads to large amplitude swelling of isolated mitochondria (Zoratti and Szabo, 1995). Because this pore is thought to be composed of both outer and inner initochondrial membrane proteins, one possibility is that Bcl-2 family members act in conjunction with these proteins to prevent opening of the pore. This hypothesis is supported by electron microscopy and cell fractionation studies that suggest that Bcl-2 may be localized preferentially to initochondrial contact sites, areas where the inner and outer mitochondrial membrane meet (de Jong et al., 1994; Hockenberry et al., 1990).Alternatively, Bcl-2 family members may be able to act as stand-alone channels at contact sites and/or the outer mitochondrial membrane to influence mitochondrial homeostasis. Although it is thought that the outer mitochondrial membrane is generally permeable to large molecules, evidence suggests that the permeability of this membrane is regulated by resident channels (Mannella, 1992). The formation of another permeability pathway by Bcl-2 family members on the outer mitochondrial membrane may be important under certain conditions, such as those that occur during apoptosis. Consistent with this idea, evidence suggests that expression of Bcl-x,~in cells that have received an
268
ANDY J . MINN et 01.
apoptotic stimulus may prevent the accumulation of protons in the mitochondrial intermembrane space (Vander Heiden et al., 1997). Assuming the ion channel properties of Bcl-2 family members do control apoptosis, there are several possible mechanisms that may account for how anti- and pro-apoptotic family members are able to have opposite effects on cell survival. For example, the Bax ion channel may promote mitochondrial dysfunction, whereas the Bcl-2 ion channel may retard this process. Bax was shown to cause the release of cytochrome c when expressed in yeast (Manon et al., 1997), suggesting that Bax may directly promote mitochondrial changes that, in a mammalian cell, would lead to apoptosis. The role of heterodimerization may be to regulate the formation of ion channels by controlling insertion into membranes. Alternatively, since it is possible that BcI-XL regulates apoptosis independently of heterodimerization, the formation of separate ion channeIs by Bax and Bcl-xL may give rise to different channels that are able to neutralize each other’s effects. Another possibility is that the formation of hybrid or nonfunctional channels through intermembrane interactions between opposing Bcl-2 family members, which are distinct from those that occur in solution, may prevent apoptosis. Consistent with some of these possibilities, Bcl-2 was shown to inhibit Bax ion channel activity (Antonsson et al., 1997). If Bcl-2 family members are able to form ion channels that regulate mitochondrial homeostasis, they could influence the propensity for cytochrome c and AIF to redistribute to the cytosol during apoptosis. In turn, caspase activation may be either prevented or promoted. Although some of the data discussed so far would support such a model, another model to regulate caspase activity has also been suggested based on the finding that Bcl-xL,as well as CED-9, interacts with CED-4 (Chinnaiyan et al., 199713; Wu et al., 1997b). Both CED-9 and Bcl-xL mutants that were previously found to be defective in inhibiting apoptosis also failed to bind to CED-4. Therefore, because CED-9 can inhibit the ability of CED-4 to promote the activation and processing of CED-3, Bcl-2 and BcI-XL may, in a similar fashion, prevent caspase activation by binding to a mammalian homolog of CED-4, such as Apaf-1, to form an inactive complex. Apaf-1 is, in fact, involved in the cytochrome-c dependent activation of caspase 3, as previously discussed. Heterodimerization between Bcl-2 family members may serve to act as competitive dimerization substrates for Apaf-1 binding. For example, Bax can act to cause cell death by preventing Bcl2 and BcI-XL from binding to proteins Iike Apaf-1. This is supported by experiments demonstrating that Bax and a Bak BH3 peptide are able to inhibit Bcl-xL from interacting with CED-4 (Chinnaiyan et al., 1997b; Ottilie et al., 1997b).
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
269
This model, whereby Bcl-2 family members regulate cell death by binding to CED-4-like proteins, is not incompatible with a model in which Bcl-2 family inembers regulate mitochondrial function through ion channel formation. Bcl-2 family members may act at multiple steps to regulate apoptosis (Fig.4).On one level, these proteins may influence the redistribution of cytochrome c and indirectly regulate caspase activation by forming ion channels in mitochondrial membranes. On another level, Bcl-2 family members may be capable of reinforcing their influence on cell fate by interacting with CED-4-like proteins. Because a cell has numerous mitochondria, even if a minority release cytochrome c, Bcl-xLand Bcl-2 can still prevent apoptosis by binding to Apaf-1 and preventing the cytochrome c-dependent activation of caspases. In contrast, Bax can function not only to promote the release of cytochrome c from mitochondria, but also prevent BcI-xLfrom inhibiting Apaf-1. VIi. Conclusion
The biological significance of apoptosis is now widely appreciated, although the mechanism that controls apoptosis is still poorly understood. As seems to be true with many fundamental biological processes, disruptions in pathways that control these processes likely contribute to many human diseases, making the elucidation and understanding of these pathways of potential importance. Although apoptosis is a complex process, much data suggest that Bcl-2 family members regulate a focal point’where a motley of upstream apoptotic signals first converge. From this point on, however, the processes that execute the cell death program seem to diverge. This phenomenon seems to result from the ability of Bcl-2 family members to control the activation of caspases. Data suggest that Bcl-2 family members may regulate caspase activation by preventing alterations in mitochondrial homeostasis that would otherwise lead to the release of mitochondrial proteins that are capable of activating downstream caspases. The formation of ion channels by Bcl-2 family members may be one mechanism by which these proteins control mitochondrial processes. A direct link between Bcl2 family members and downstream caspases has also been discovered by studies that demonstrate that both the former and the latter proteins are able to interact with CED-4. The control of CED-Pmediated processing of caspases through interactions with Bcl-2 family members may be yet another way to regulate apoptosis. In addition to the multitude of proteins that associated with Bcl-2 family members mentioned in this review, many other interesting proteins can also associate and, in some cases, influence cell survival. Some additions to this list include Raf-1 (Wang et al., 1996a), calcineurin (Shibasaki et al., 1997), p28 Bap31 (Ng et al., 1997), and
FIG.4. Bcl-2 family members may regulate apoptosis at multiple steps in the cell death pathway. One shared feature that results from multiple apoptotic stimuli is an early alteration in mitochondrial function that manifests as organelle swelling, rupture of the outer mitochondrial membrane, eventual loss of transmembrane potential, and the release of mitochondrial proteins, such as cytochrome c and AIF. Antiapoptotic Bcl-2 family members, such as BclxL, are able to prevent these characteristic mitochondrial changes, whereas proapoptotic Bcl-2 family members, such as Bax, are able to promote these changes. Bcl-x,, and Bax may independently regulate mitochondrial physiology through the formation of ion channels and/or through the regulation of each others function by heterodimerization. The release of mitochondrial proteins, such as cytochrome c, along with other factors, such as dATP, results in the activation of Apaf-1. Apaf-1 binds to downstream caspases and processes them into proteolytically active forms. Bcl-xLmay bind to Apaf-1 and prevent this conversion, whereas Bax may complex with BcI-xLto prevent it from binding to Apaf-1.
REC:ULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
271
cytochroine c (Kliarbanda et al., 1997). Future studies will need to test the importance of the biochemical activities identified for members of the Bcl-2 family, determine the relative importance of the interactions between Bcl-2 family members and the proteins they interact with, and how these protein interactions regulate function.
ACKNOWLEDGMENT The authors thank Aineeta Kelekar tor her tliouglitful discussions and critique of this manuscript.
REFERENCES Abbas, A. K. (1996). Die and let live: Eliminating dangerous lymphocytes. Cell 84,655-657. Akao, Y., Otsuki, Y., Kataoka, S., Ito. Y., and Tsujinioto, Y. (1994). Multiple subcellular localization of bcl-2: Detection in nuclear outer membrane, endoplasniic reticulum mem54, . 2468-2471. brane, and initochondrial membranes. Cancer R ~ . Y Antonsson, B., Conti, F., Ciavatta, A., Montessuit, S., Lewis, S., Martinou, I., Bernasconi, L., Bernard, A,, Mermod, J. J., Mazzei, G., Maundrell, K., Gainbale, F., Sadoul, R., and Martinou, J. C. (1997). Inhibition of Bax channel-forming activity by BcI-2. Science 277,370-372. BaHishi, A,, Jensen, J. P., Goldinan, P., Wiight, J. J., McBritle, 0. W., Epstein, A. L., and Korsineyer, S. J. (1985). Cloning the chromosomal breakpoint of t(14;18) human lyniphomas: Clustering around JH on chromosome 14 and near a transcriptional unit on 18. Cell 41,899-906. Barry, M. A., Reynolds, J. E., rind Eastnian, A. (1993). Etoposide-induced apoptosis in human HL-60 cells is associated with intracelh~laracidification. Cancer Res. 53, 2349-2357. Boise, L. H., Gonzalez-Garcia, M., Postenia, C. E., Ding, L., Lindsten, T., Turka, L. A,, Mao, X., Nnnez, G., and Thompson, C. B. (1993).1x1-x, a bcl-&-relatedgene that functions as a dominant regulator of apoptotic cell death. Cell 74, 597-608. Boise, L. H., Minn, A. J., Noel, P. J.. June, C. H., Accavitti, M. A,, Lindsten, T., and Thompson, C. B. (1995). CD28 costimulation can promote T cell survival by enhancing Imniutrity 3,87-98. the expression of B~1-x~. Boldin, M. P., Goncharov, T. M.. Goltsev, Y. V., and Wallach, D. (1996). Involvement of MACH, a novel MORTlRADD-interacting protease, in FasiAPO-1- and TNF receptorinduced cell death. Cell 85, 803-815. Borner, C., Martinou, I., Mattmann, C., Irniler, M., Schaerer, E., Martinou, J. C., and Tschopp, J. (1994). The protein bcl-2 alpha does not require niernbrane attachment, but two conserved doinains to suppress apoptosis. J , Cell Bid. 126, 1059-1068. Boyd, J. M., Gallo, G. J., Elangovan, B., Houghton, A. B., Malstrom. S., Avery, B. J., Ebb, R. G., Subramanian, T., Chittenden, T., Lutz, R. J,, et (11. (1995). Bik, a novel deathinducing protein, shares a distinct sequence motif with Bcl-2 family proteins and interacts with viral and cellular survival-promoting proteins. Oncogene 11, 1921-1928. Chang, B. S., Minn, A. J., Muchniore, S. W., Fesik, S. W., and Thompson, C. B. (1997). Identification of a novel regulatory domain in Bcl-xt,and Bc1-2. EMBO 1. 16, 968-977. Cheng, E. H., Levine, B., Boise, I,. H., Thompson. C. B., and Hardwick, J. M. (1996).R u independent inhibition of apoptosis by Bcl-x,,. Nature 379, 554-556. Cheng, E. H., Nicholas, J., Bellows, D. S.. Hayard, G. S., Guo, H. G., Reitz, M. S., and Hardwick, J. M. (1997). A Bcl-2 homolog encoded by Kaposi sarcoma-associated virus.
272
ANDY J. MINN et a1
human herpesvirus 8, inhibits apoptosis but does not heterodimerize with Bax or Bak. Proc. Natl. Acad. Sci. USA 94, 690-694. Chinnaiyan, A. M., Chaudhary, D., O’Rourke, K., Koonin, E. V., and Dixit, V. M. (1997a). Role of CED-4 in the activation of CED-3. Nature 388, 728-729. Chinnaiyan, A. M., O’Rourke, K., Lane, B. R., and Dixit, V. M. (199%). Interaction of CED4 with CED-3 and CED-9: A molecular framework for cell death. Science 275,1122-1126. Chinnaiyan, A. M., Orth, K., O’Rourke, K., Duan, H., Poirier, G. G., and Dixit, V. M. (1996). Molecular ordering of the cell death pathway: Bcl-2 and Bcl-xLfunction upstream of the CED-3-like apoptotic proteases. J. Bid. Chem. 271, 4573-4576. Chittenden, T., Flemington, C., Houghton, A. B., Ebb, R. G., Gallo, G. J., Elangovan, B., Chinnadurai, G., and Lutz, R. J. (1995a). A conserved domain in Bak, distinct from BH1 and BH2, mediates cell death and protein binding functions. EMBO J. 14, 5589-5596. Chittenden, T., Harrington, E. A,, OConnor, R., Flemington, C., Lutz, R. J., Evan, G. I., and Guild, B. C. (1995b). Induction of apoptosis by the Bcl-2 homologue Bak. Nature 374, 733-736. Cleary, M. L., Smith, S. D., and Sklar, J. (1986). Cloning and structural analysis of cDNAs for bcl-2 and a hybrid bc1-2hmmunoglobulin transcript resulting from the t( 14;18)translocation. Cell 47, 19-28. Cohen, J. J. (1993). Apoptosis. lmmunol. Today 14, 126-130. Cramer, W. A., Heymann, J. B., Schendel, S. L., Deny, B. N., Cohen, F. S., Elkins, P. A,, and Stauffacher, C. V. (1995). Structure-function of the channel-forming colicins. Annu. Rev. Biophys. Biomol. Struct. 24, 611-641. Cryns, V. L., Bergeron, L., Zhu, H., Li, H., and Yuan, J. (1996). Specific cleavage of alpha-fodrin during Fas- and tumor necrosis factor-induced apoptosis is mediated by an interleukin-1beta-converting enzymelced-3 protease distinct from the poly(ADP-ribose) polymerase protease. J. Biol. Chem. 271,31277-31282. Datta, S. R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y., and Greenberg, M. E. (1997). Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91, 231-241. de Jong, D., Prins, F. A., Mason, D. Y., Reed, J. C., van Ommen, G. B., and Kluin, P. M. (1994). Subcellular localization of the bcl-2 protein in malignant and normal lymphoid cells. Cancer Res. 54, 256-260. del Peso, L., Gonzalez-Garcia, M., Page, C., Herrera, R., and Nunez, G. (1997).Interleukin3-induced phosphorylation of BAD through the protein kinase Akt. Science 278,687-689. Dumont, M. E., Cardillo, T. S., Hayes, M. K., and Sherman, F. (1991). Role of cytochrome c heme lyase in mitochondria1import and accumulation of cytochrome c in Saccharomyces cerevisiae. Mol. Cell. Biol. 11, 5487-5496. Ellis, H. M., and Horvitz, H. R. (1986). Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817-829. Ellis, R. E., Yuan, J. Y., and Horvitz, H. R. (1991). Mechanisms and functions of cell death. Annu. Reu. Cell Biol. 7, 663-698. Enari, M., Hase, A., and Nagata, S. (1995). Apoptosis by a cytosolic extract from Fasactivated cells. EMBO J. 14, 5201-5208. Faleiro, L., Kobayashi, R., Fearnhead, H., and Lazebnik, Y. (1997). Multiple species of CPP32 and Mch2 are the major active caspases present in apoptotic cells. EMBO J, 16,2271-2281. Farrow, S. N., White, J. H., Martinou, I., Raven, T., Pun, K. T., Grinham, C. J., Martinou, J. C., and Brown, R. (1995). Cloning of a bcl-2 homologue by interaction with adenovirus E1B 19K. Nature 374, 731-733.
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
273
Faucheu, C., Diu, A., Chan, A. W., Blanchet, A. M., Miossec, C., Herve, F., CollardDutilleul, V., Gu, Y., Aldape, R. A,, Lippke, J. A,, et nl. (1995). A novel human protease similar to the interleulan-1 beta converting enzyme induces apoptosis in transfected cells. EMBO J . 14, 1914-1922. Franke, T. F., Kaplan, D. R., and Cantley, L. C. (1997). PI3K: Downstream AKTion blocks apoptosis. Cell 88, 435-437. Gibson, L., Holmgreen, S. P., Huang, D. C., Bernard, O., Copeland, N. G., Jenkins, N. A., Sutherland, G. R., Baker, E., Adanis, J. M., and Cory, S. (1996). bcl-w, a novel member of the bcl-2 family, promotes cell survival. Oncogene 13, 665-675. Gonzalez-Garcia, M., Garcia, I., Ding, L., O’Shea, S., Boise, L. H., Thompson, C. B., and Nunez, G. (1995). bcl-x is expressed in einbryonic and postnatal neural tissues and functions to prevent neuronal cell death. Proc. Notl. Acad. Sci. USA 92, 4304-4308. Gonzalez-Garcia, M., Perez-Ballestero, R., Ding, L., Duan, L., Boise, L. H., Thompson, C. B., and Nunez, C. (1994). Bcl-xl,is the major bcl-x inRNA form expressed during innrine development and its product localizes to mitochondria. Dezjelopinent 120, 3033-3042. Hamburger, V. (1975). Cell death in the development of the lateral motor column of the chick embryo. J . Comnp. Nerirol. 160, Hamburger, V. (1992). History of the discovery of neuronal death in embryos. /. Neurobiol. 23, 1116-1123. Han, J., Sabbatini, P., Perez, D., Rao, L., Modha, D., and White, E. (1996). The E1B 19K protein blocks apoptosis by interacting with and inhibiting the $3-inducible and deathpromoting Bax protein. Genes Deu. 10, 461-477. , A. J., and Kurnar, S. (1997). Functional activation of Nedd2ACH-I Harvey, N. ~ L .Butt, (caspase-2) is an early process in apoptosis. J. B i d . Chem. 272, 13134-13139. Henderson, S.. Huen, D., Rowe, M., Dawson, C., Johnson, G., and Rickinson, A. (1993). Epstein-Barr virus-coded BHRFl protein, a viral homologue of Bcl-2, protects human B cells from programmed cell death. Proc. Natl. Acad. Sci. USA 90, 8479-8483. Hengartner, M. 0.. Ellis, R. E., and Horvitz, H. R. (1992). Cnenorhabditis eleguns gene ced-9 protects cells from programmed cell death. Nature 356, 494-499. Hengartner, M. 0..and Horvitz, H. R. (1994a).C. elegnns cell survival gene ced-9 encodes a frrnctional hornolog of the mammalian proto-oncogene bcl-2. Cell 76, 66,5-676. Hengartner, M. O., and Horvitz, H. R. (1994b). Programmed cell death in Caenorhabditis elegnns. Curr. Opin. Genet. Dez;. 4, 581-586. Hockenberry, D., Nunez, G., Millinian, C., Schreiber, R. D., and Korsmeyer, S. J. (1990). Bcl-2 is an inner mitochondria1 membrane protein that blocks programmed cell death. Nature 348, 334-336. Hockenberry, D. M., Oltvai, Z. N., Yin, X. M., Milliman, C. L., and Korsmeyer, S. J. (1993). Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 75, 241-251. Hsu, Y. T., WoIter, K. G., and Youle. R. J. (1997). Cytosol-to-membrane redistribution of Bax and Bcl-X(L) during apoptosis. Proc. Nntl. Acad. Sci. USA 94, 3668-3672. Hunter, J. J,, and Parslow, T. G. (1996). A peptide sequence from Bax that converts Bcl2 into an activator of apoptosis. J. B i d . Chem. 271, 8521-8524. Huynh, P. D., Cui, C., Zhan, H., Oh, K. J., Collier, R. J., and Finkelstein, A. (1997). Probing the structure of the diphtheria toxin channel: Reactivity in planar lipid bilayer membranes of cysteine-substituted mutant channels with inethanethiosulfonate derivatives. /. Gen. Physiol. 110, 229-242. Ink, B., Zornig, M., Baum, B., Hajibaglieri, N., James, C., Chittenden, T., and Evan, G. d ~e cwith e . s morphological (1997). Human Bak induces cell death in S c h ~ ~ ~ , ~ n c c h n r op? o~ ~ changes similar to those with apoptosis in inainmalian cells. Mol. Cell. Bid. 17,2468-2474.
274
ANDY J. M l N N et al.
Inohara, N., Ding, L., Chen, S., and Nunez, G. (1997). Harakiri, a novel regulator of cell death, encodes a protein that activates apoptosis and interacts selectively with survivalpromoting proteins Bcl-2 and Bcl-xL.EMBO ]. 16, 1686-1694. Jacobson, M. D., Burne, J. F., King, M. P., Miyashita, T., Reed, J. C., and Raff, M. C. (1993). Bcl-2 blocks apoptosis in cells lacking mitochondrial DNA. Nature 361,365-369. Jacobson, M. D., Burne, J. F., and Raff, M. C. (1994). programmed cell death and Bcl-2 protection in the absence of a nucleus. EMBO]. 13, 1899-1910. Jacobson, M. D., and Raff, M. C. (1995). Prograinmed cell death and Bcl-2 protection in very low oxygen. Nature 374, 814-816. Jacobson, M. D., Weil, M., and Raff, M. C. (1997). Programmed cell death in animal development. Cell 88, 347-354. James, C., Gschmeissner, S., Fraser, A., and Evan, G. I. (1997). CED-4 induces chromatin condensation in Schizosaccharoiriyces ponibe and is inhibited by direct physical association with CED-9. Curr. B i d . 7 , 246-252. Jost, A. (1971). Use of androgen antagonists and antiandrogens in studies on sex differentiation. Gynecol. Inuest. 2, 180-201. Jurgensmeier, J. M., Krajewski, S., Armstrong, R. C., Wilson, G. M., Okersdorf, T., Fritz, L. C., Reed, J. C., and Ottilie, S. (1997). Baw- and Bak-induced cell death in the fission yeast Scliizosaccharoniyces pombe. Mol. Biol. Cell 8, 325-339. Karnens, J., Paskind, M., Hugunin, M., Talanian, R. V., Allen, H., Banach, D., Bump, N., Hackett, M., Johnston, C. G., Li, P., et al. (1995). Identification and characterization of ICH-2, a novel member of the interleukin-1 beta-converting enzyme family of cysteine proteases. J. B i d . Chem. 270, 15250-15256. Kane, D. J., Sarafian, T. A,, Anton, R., Hahn, H., Gralla, E. B., Valentine, J. S., Ord, T., and Bredesen, D. E. (1993). Bcl-2 inhibition of neural death: Decreased generation of reactive oxygen species. Science 262, 1274-1277. Kelekar, A., Chang, B. S., Harlan, J. E., Fesik, S. W., and Thompson, C. B. (1997). Bad is a BH3 domain-containing protein that forms an inactivating dimer with Bcl-x,,. MoE. Cell. Biol.17, 7040-7046. Kerr, J. F., Wyllie, A. H., and Currie, A. R. (1972).Apoptosis: A basic biological phenomenon with wide-ranging implications in tissue kinetics. Br. J. Cancer 26, 239-257. Kharbanda, S., Pandey, P., Schofield, L., Israels, S., Roncinske, R., Yoshida, K., Bharti, A., Yuan, Z. M., Saxena, S., Weichselbaum, R., Nalin, C., and Kufe, D. (1997). Role for Bclxi. as an inhibitor of cytosolic cytochroine C accumulation in DNA damage-induced apoptosis. Proc. Natl. Acad. Sci. USA 94, 6939-6942. Kiefer, M. C., Brauer, M. J., Powers, V. C., Wu, J. J., Urnansky, S. R., Tomei, L. D., and Barr, P. J. (1995). Modulation of apoptosis by the widely distributed Bcl-2 hornologue Bak. Nature 374, 736-739. Kluck, R. M., Bossy-Wetzel, E., Green, D. R., and Newmeyer, D. D. (1997). The release of cytochroine c from mitochondria: A primary site for Bcl-2 regulation of apoptosis. Science 275, 1132-1136. Krajewski, S., Bodrug, S., Krajewska, M., Shabaik, A,, Gascoyne, R., Berean, K., and Reed, J. C. (1995).Imniunohistochemical analysisof Mcl-1 protein in human tissues: Differential regulation of Mcl-1 and Bcl-2 protein production suggests a unique role for Mcl-1 in control of programmed cell death in vivo. Am. J. Pathol. 146, 1309-1319. Krajewski, S., Krajewska, M., and Reed, J. C. (1996). Iinmunohistocheniical analysis of in vivo patterns of Bak expression, a proapoptotic member of the BcI-2 protein family. Caiicer Res. 56, 2849-2855. Krajewski, S., Krajewska, M., Shabaik, A,, Wang, H. G., Irie, S., Fong, L., and Reed, J. C. (1994). Iininunohistocheinical analysis of in vivo patterns of Bcl-X expression. Cancer Res. 54,5501-5507.
HECUI.ATION OF APOPTOSIS BY Rcl-2 FAMILY M E M B E R S
275
Krajewsk, S.. Tanaka. S., Takayama. S., Schibler, M. J., Fenton, W., and Reed, J. C. (1993). Investigation of the suhcellnlar distribution of the bcl-2 oncoprotein: Residence in the nuclear envelope, endoplasmic reticuluni, and outer initochonclrial membranes. Cancer Res. 53, 4701-4714. Kroeiner, G. (1997). The proto-oncogene BcI-2 and its role in regulating apoptosis. Nnt. ~ e c i 3,614-620. . Kuida, K., Lippke, J. A., Ku, G., Harding, M. W., Livingston, D. J., Su, M. S., and Flavell, R. A. (1995). Altered cytokine export and apoptosis in mice deficient in interleukin-1 beta converting enzyme. Science 267, 2000-2003. Knida, K., Zheng, T. S., Na, S., Kuan, C., Yang, D., Karasuyama, H., Rakic, P., and Flavell, R. A. (1996).Decreased apoptosis i n the brain and preinatnre lethality in CPP32-deficient mice. Natrire 384, 368-372. Kuinar, S. (1995). ICE-like proteases in apoptosis. Trends Biochern. Sci. 20, 198-202. Kuinar, S., Kinoshita, M., Noda, M., Copeland, N.G., and Jenkins, N. A. (1994). Induction of apoptosis by the mouse Nedd2 gene, wliicli encodes a protein similar to the product of the Cnenorlmbditis ekgnns cell death gene ced-3 and the manimalian IL-1 hetaconverting enzyme. Genes Deli. 8, 1613-1626. Lam. M.. Dubyak, G.. Chen, L., Nunez. G.. Miesfeld, R. L., and Distelhorst, C. W. (1994). Evidence that BCL-2 represses apoptosis by regulating endoplasmic reticulum-associated Ca” flnxes. Proc. Nutl. Arod. Sci. U S A 91, 6568-6573. Lazebnik, Y. A,, Kaufinann. S. H., Desnoyers, S., Poirier, G. G., and Earnshaw, W. C . (1994). Cleavage of poly(ADP-ribose) polymerase by a proteinase with properties like ICE. Nntiire 371, 346-347. Lin, E. Y., Orlofsky, A,, Berger, M. S., and Prystowsky, M. B. (1993). Characterization of A l , a novel heiiiopoietic-specifc early-response gene with sequence similarity to bcl-2. /. Znununol. 151, 1979-1988. Lin, E. Y., Orlofsky, A., Wang, H. G., Reed, J. C., and Prystowsky. M. B. (1996). A l , a BcI-2 faniily member, prolongs cell survival and permits rnyeloid differentiation. Blood 87, 983-992. Lithgow, T., van Diiel, R., Beitrani, J. F., and Strasser. A. (1994). The protein product of the oncogene bcl-2 is a component of the nnclear envelope, the endoplasmic reticulum. and the outer mitochondria1 meinl)rane. C d l Groicth. Difer. 5, 41 1-417. Liu, X., Kiln, C. N., Yang, J., Jeniiiierson. R., and Wang, X. (1996). Induction of apoptotic program in cell-free extracts: Requirement for dATP and cytochroine c. Cell86, 147-157. Liu. X., Zon, H., Slaughter, C., and Wang, X. (1997). DFF, a heterodiineiic protein that functions clownstream of caspase-3 to trigger DNA fragmentation during apoptosis. Cell 89, 175-184. London, E. (1992). Diphtheria toxin: Membrane interaction and membrane translocation. Biochiin. Biophys. Act0 1113, 25-51 Longo, V. D., Ellerby, L. M., Bretlesrn, D. E., Valentine. J. S., and GralIa, E. B. (1997). Hinnaii Bcl-2 reverses survival defects in yeast lachng superoxide dismutase and delays death of wild-type yeast. /. Cell Biol. 137, 1581-1588. Ma, A,, Pena, J. C., Chang, B., Margosian, E., Davidson, L., Alt, F. W., and Thompson, C. B. (1995). Bclx regulates the survivnl of donble-positive thyinocytes. Proc. Nntl. Acnd. Sci. USA 92, 4763-4767. Mannella, C. A. (1892). The ‘ins’ and ‘outs‘ of mitochondrial membrane channels. Trends Biochem. Sci. 17, 315-320. Manon, S., Chandhuri. B., and Guerin, M. (1997). Release of cytochrome c and decrease of cytochrome c oxidase in Bax-expressing yeast cells, and prevention of these effects by coexpression of Bcl-x,,.F E B S Lett. 415, 29-32.
276
ANDY J. M I N N rt al.
Marin, M. C., Fernandez, A,, Bick, R. J., Brisbay, S., Buja, L. M., Snugs, M., McConkey, D. J,, von Eschenbach, A. C., Keating, M. J.. and McDonnell, T. J. (1996). Apoptosis suppression by bcl-2 is correlated with the regulation of nuclear and cytosolic Ca". Oncogene 12, 2259-2266. Minn, A. J., Boise, L. H., and Thompson, C. B. (1996). Bcl-xs anatagonizes the protective effects of Bcl-xL.J. B i d . Chem. 271, 6306-6312. Minn, A. J., Velez, P., Schendel, S. L., Liang, H., Muchmore, S. W., Fesik, S. W., Fill, M., and Thompson, C. B. (1997). Bcl-xLforms an ion channel in synthetic lipid membranes. Nature 385, 353-357. Miura, M., Zhu, H., Rotello, R., Harhvieg, E. A., and Yuan, J. (1993).Induction of apoptosis in fibroblasts by IL-1 beta-converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell 75, 653-660. Motoyama, N., Wang, F., Roth, K. A,, Sawa, H., Nakayarna, K., Nakayama, K., Negishi, I., Senju, S., Zhang, Q., Fujii, S . , et nl. (1995). Massive cell death of immature hematopoietic cells and neurons in Bcl-x-deficient mice. Science 267, 1506-1510. Muchmore, S. W., Sattler, M., Liang, H., Meadows, R. P., Harlan, J. E., Yoon, H. S., Nettesheim, D., Chang, B. S., Thompson, C. B., Wong, S. L., Ng, S. L., and Fesik, S. W. (1996). X-ray and NMR structure of human Bcl-x,,, an inhibitor of programmed cell death. Nature 381, 335-341. Muzio, M., Chinnaiyan,A. M., Kischkel, F. C., O'Rourke, K., Shevchenko,A., Ni, J., Scaffidi, C., Bretz, J. D., Zhang, M., Gentz, R., Mann, M., Krammer, P. H., Peter, M. E., and Dixit, V. M. (1996). FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/APO-1) death-inducing signaling complex. Cell 85,817-827. Nakayama, K., Nakayama, K., Negishi, I., Kuida, K., Sawa, H., and Loh, D. Y. (1994). Targeted disruption of Bcl-2 alpha beta in mice: Occurrence of gray hair, polycystic kidney disease, and lymphocytopenia. Proc. Natl. Acad. Sci. USA 91, 3700-3704. Neilan, J. G., Lu, Z., Afonso, C. L., Kutish, G. F., Sussman, M. D., and Rock, D. L. (1993). An African swine fever virus gene with similarity to the proto-oncogene bcl-2 and the Epstein-Barr virus gene BHRF1.J. Virol. 67, 4391-4394. Ng, F. W. H., Nguyen, M., Kwan, T., Branton, P. E., Nicholson, D. W., Cromlish, J. A,, and Shore, G. C. (1997). p28 Bap31, a Bcl-2/Bcl-xld-and procaspase-8-associatedprotein in the endoplasinic reticulum. I. Cell B i d . 139, 327-338. Nguyen, M., Branton, P. E., Walton, P. A,, Oltvai, Z. N., Korsmeyer, S. J.. and Shore, G. C. (1994). Role of membrane anchor domain of Bcl-2 in suppression of apoptosis caused by E1B-defective adenovirus. J. Biol. Chem. 269, 16521-16524. Nguyen, M., Millar, D. G., Yong, V. W., Korsmeyer, S. J., and Shore, G. C. (1993).Targeting of Bcl-2 to the mitochondrial outer membrane by a COOH-terminal signal anchor sequence. J. Biol. Chem. 268, 25265-25268. N o d , G. J. (1994). Negative selection of lymphocytes. Cell 76, 229-239. Oltvai, 2. N., Milliman, C . L., and Korsmeyer, S. J. (1993). Bcl-2 heterodinierizes in vivo with a conserved homolog, Bax, that accelerates programmed ce11death. CelZ 74,609-619. Osborne, B. A. (1996). Apoptosis and the maintenance of homoeostasis in the immune system. Cum. Opin. Zniinunol. 8, 245-254. Ottilie, S., Diaz, J. L., Chang, J,, Wilson, G., Tuffo, K. M., Weeks, S., McConnell, M., Wang, Y., Oltersdorf, T., and Fritz, L. C. (1997a). Structural and functional complementation of an inactive Bcl-2 mutant by Bax truncation. J. B i d . Chem. 272, 16955-16961, Ottilie, S., Wang, Y., Banks, S., Chang, J., Vigna, N. J., Weeks, S., Armstrong, R. C., Fritz, L. C., and Oltersdorf, T. (199%). Mutational analysis of the interactingcell death regulators CED-9 and CED-4. Cell Death Differ. 4, 526-533.
REGULATION OF APOPTOSIS BY Bcl-2 FAMILY MEMBERS
277
Parker, M. W., and Pattus, F. (1993). Rendering a membrane protein soluble in water: A common packing motif in bacterial protein toxins. Trends Biochem. Sci. 18, 391-395. Raff, M. C. (1992). Social controls on cell survival and cell death. Nature 356, 397-400. Reynolds, J. E., and Eastman, A. (1996). Intracellular calcium stores are not required for Bcl-2-mediated protection from apoptosis. 1.Biol. Chem. 271, 27739-27743. Rudel, T., and Bokoch, G. M. (1997). Membrane and morphological changes in apoptotic cells regulated by caspase-mediated activation of PAK2. Science 276, 1571-1574. Sato, T., Hanada, M., Bodrug, S., Irie, S., Iwama, N., Boise, L. H., Thompson, C. B., Golemis, E., Fong, L., Wang, H. G., et al. (1994). Interactions among members of the BcI-2 protein family analyzed with a yeast two-hybrid system. Proc. Natl. Acad. Sci. USA 91, 9238-9242. Sattler, M., Liang, H., Nettesheim, D., Meadows, R. P., Harlan, J. E., Eberstadt, M., Yoon, H. S., Shuker, S. B., Chang, B. S., Minn, A. J., Thompson, C. B., and Fesik, S. W. (1997). Structure of Bcl-xL-Bak peptide complex: Recognition between regulators of apoptosis. Science 275, 983-986. Saunders, J. W., Jr. (1966). Death in embryonic systems. Science 154, 604-612. Schendel, S. L., Xie, Z., Montal, M. O., Matsuyama, S., Montal, M., and Reed, J. C. (1997). Channel formation by antiapoptotic protein Bcl-2. Proc. Nntl. Acad. Sci. USA 94, 5113-5118. Schlesinger, P. H., Gross, A,, Yin. X. M., Yamamoto, K., Saito, M., Waksman, G., and Korsmeyer, S. J. (1997). Comparison of the ion channel characteristics of proapoptotic BAX and antiapoptotic BCL-2. Proc. Natl. Acnd. Sci. USA 94, 11357-11362. Schwartz, L. M., Smith, S. W., Jones, M. E., and Osborne, B. A. (1993). Do all programmed cell deaths occur via apoptosis? Proc. Nutl. Acad. Sci. USA 90, 980-984. Sedlak, T. W., Oltvai, Z. N., Yang, E., Wang, K., Boise, L. H., Thompson, C. B., and Korsmeyer, S. J. (1995). Multiple Bcl-2 family members demonstrate selective dimerizations with Bax. Proc. Natl. Acacl. Sci. USA 92, 7834-7838. Seshagiri, S., and Miller, L. K. (1997). Caenorhnbditis elegnns CED-4 stimulates CED-3 processing and CED-3-induced apoptosis. Cum. Biol. 7, 455-460. Shaham, S., and Horvitz, H. R. (1996). Developing Cnenorhabditis elegans neurons may contain both cell-death protective and killer activities. Genes Deu. 10, 578-591. Shibasaki, F., Kondo, E., Akagi, T., and McKeon, F. (1997). Suppression of signalling through transcription factor NF-AT by interactions between cdcineurin and Bcl-2. Nature 386, 728-731. Shimizu, S., Eguchi, Y., Kosaka, H., Kamiike, W., Matsuda, H., and Tsujimoto, Y. (1995). Prevention of hypoxia-induced cell death by Bcl-2 and Bcl-x,~.Nature 374, 811-813. Silverman, J. A,, Mindell, J. A,, Finkelstein, A,, Shen, W. H., and Collier, R. J. (1994). Mutational analysis of the helical hairpin region of diphtheria toxin transmembrane domain. ]. Biol. Chem. 269,22524-22532. Simonian, P. L., Grillot, D. A. M., Merino, R., and Nunez, G. (1996). Bax can antagonize Bcl-x,~during etoposide and cisplatin-induced cell death independently of its heterodimerization with Bcl-xL.1.Biol. Chem 271, 22764-228772. Spector, M. S., Desnoyers, S., Hoeppner, D. J., and Hengartner, M. 0. (1997). Interaction between the C. elegans cell-death regulators CED-9 and CED-4. Nature 385, 653-656. Surh, C. D., and Sprent, J. (1994). T-cell apoptosis detected in situ during positive and negative selection in the thymus. Nature 372, 100-103. Susin, S. A,, Zamzami, N., Castedo, M., Hirsch, T., Marchetti, P., Macho, A., Daugas, E., Geuskens, M., and Kroemer, G. (1996). Bcl-2 inhibits the niitochondrial release of an apoptogenic protease. /. Exp. Med. 184, 1331-1341.
278
ANDY J , MINN et ol
Takahashi, A,, Alnemri, E. S., Lazebnik, Y. A,, Fernandes-Alnemri, T., Litwack, G., Moir, R. D., Goldman, R. D.. Poirier, G. G., Kaufinann, S. H., and Earnshaw, W. C. (1996). Cleavage of lamin A by Mch2 alpha hut not CPP32: Multiple interleukin 1beta-converting enzyme-related proteases with distinct substrate recognition properties are active in apoptosis. €'roc. Natl. Acad. Sci. USA 93, 8395-8400. Thompson, C. B. (1995). Apoptosis in the pathogenesis and treatment of disease. Science 267, 1456-1462. Tsujimoto, Y., and Croce. C. M. (1986). Analysis of the structure, transcripts, and protein products of bcl-2, the gene involved in human follicular lymphoma. Proc. Nutl. Acud. Sci. USA 83, 5214-5218. Uhlmann, E. J., D'Sa-Eipper, C., Subramanian, T., Wagner, A. J., Hay, N., and Chinnadurai, G. (1996). Deletion of a nonconserved region of Bcl-2 confers a novel gain of function: Suppression of apoptosis with concomitant cell proliferation. Cancer Res. 56,2506-2509. Vander Heiden, M. G., Chandel, N. S.,Williamson, E. K., Schumacker, P. T., andThompson, C. B. (1997). Bcl-xLregulates the membrane potential and volume homeostasis of mitochondria. Cell 91, 627-637. Vaux, D. L., Weissman, I. L., and Kim, S. K. (1992). Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 258, 1955-1957. Veis, D. J., Sorenson, C. M., Shutter, J. R., and Korsmeyer, S. J. (1993). Bcl-%deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75, 229-240. Wang, H. G., Rapp, U. R., and Reed, J. C. (1996a). Bcl-2 targets the protein kinase Raf1 to mitochondria. Cell 87, 629-638. Wang, K., Yin, X. M., Chao, D. T., Milliman, C. L., and Korsmeyer, S. J. (1996b). BID: A novel BH3 domain-only death agonist. Genes Den 10, 2859-2869. Wang, L., Minra, M., Bergeron, L., Zhu, H., and Yuan, J. (1994). Ich-I, an Iceked-3related gene, encodes both positive and negative regulators of programmed cell death. Cell 78, 739-750. Weil, M., Jacobson, M. D., Coles, H. S., Davies, T. J., Gardner, R. L., Raff, K. D., and Raff, M. C. (1996). Constitutive expression of the machinery for programmed cell death. 1. Cell Bid. 133, 1053-1059. Wu, D., Wallen, H. D., Inohara, N., and Nunez, G. (1997a). Interaction and regulation of the Caenorhabditis elegans death protease CED-3 by CED-4 and CED-9. J. Biol. Chern. 272, 21449-21454. Wu, D., Wallen, H. D., and Nunez, G. (1997b). Interaction and regulation of subcellular localization of CED-4 by CED-9. Scieme 275, 1126-1129. Xiang, J., Chao, D. T., and Korsmeyer, S. J. (1996). BAX-induced cell death may not require interleukin 1 beta-converting enzyme-like proteases. Proc. Nutl. Acud. Sci. U S A 93, 14559-14563. Xue, D., Shaham, S., and Horvitz, H. R. (1996). The Caenorhabditis elegans cell-death protein CED-3 is a cysteine protease with substrate specificities similar to those of the human CPP32 protease. Genes Deu. 10, 1073-1083. Yang, E., and Korsmeyer, S. J. (1996). Molecular thanatopsis; A discourse on the BCL2 family and cell death. Blood 88, 386-401. Yang, E., Zha, J., Jockel, J., Boise, L. H., Thompson, C. B., and Korsmeyer, S. J. (1995). Bad, a heterodimeric partner for Bcl-xL and Bcl-2, displaces Bax and promotes cell death. Cell 80, 285-291. Yang, J., Liu, X., Bhalla, K., Kim, C. N., Ibrado, A. M., Cai, J., Peng, T. I., Jones, D. P., and Wang, X. (1997). Prevention of apoptosis by Bcl-2: Release of cytochrome c from mitochondria blocked. Science 275, 1129-1132.
REC;LJIATIOK OF APOPTOSIS BY BcI-2 FAMILY MEMBERS
279
Yang, T., Buchan, H. L., Townseud, K. J., and Craig, R. W. (1996). MCL-1, a member of the BLC-2 family, is induced rapidly in response to signals fix cell differentiation or death, but not to signals for cell proliferation. J . Cell. Physiol. 166, 523-536. Yin, X. M., Oltval, Z. N., and Korsmeyer, S. J. (1994). BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterotlinierization with Bax. Nature 369, 321-323. Yuan, J. (1997). Transducing signals of life and death. C u m Opin. Cell B i d . 9, 247-251. Yuan, J. Y., and Horvitz, H. R. (1990). The Cnenorhnhditis elegms genes ced-3 and ced4 act cell autonomously to cause programmed cell death. Dec. Biol. 138, 33-41. Zamzanii, N., Marchetti, P., Castedo, M., Decaudin, D., Macho, A., Hirsch, T., Susiii, S. A., Petit, P. X., Mignotte, B., and Kroerner, G. (19954. Sequential reduction of mitochondria1 transmembrane potential and generation of reactive oxygen species in early programmed cell death. J . Exp. Mecl. 182, 367-377. Zamzami, N., Marchetti, P., Castedo, M., Zanin, C., Vayssiere, J. L., Petit, P. X., and Kroerner, G. (199511).Reduction in mitochondria1potential constitutes an early irreversible step of progranirned lymphocyte death in vivo. J . Exp. Med. 181, 1661-1672. Zainzami, N . , Susin, S. A., Marchetti, P., Hirsch, T., Gornez-Monterrey, I., Castedo, M., and Guido, K. (1996). Mitochondria1control of nuclear apoptosis.]. Exp. Merl. 183,15331,544. Zha, H., Aime-Sempe, C., Sato, T., and Reed, J. C . (199fia). Prodpoptotic protein Bay lieterodinierizes with Bcl-2 and homotlimerizes with Bax via a novel domain (BH3)distinct from BH1 and BH2. J. Biol. Cheni. 271, 7440-7444. Zha, H., Fisk, H. A,. Yaffe, M. P., Mahajan, N., Herman, B., and Reed, J. C. (1996b). Stmcture-function cotnparisons of the proapoptotic protein Bax in yeast and mammalian cells. Mol. Cell. Biol. 16, 6494-6508. Zha, J., Harada, H., Osipov, K., Jockel, J., Waksnian, G., and Korsmeyer, S. J. (1997). BI-I3 domain o f BAD is required for heterodimerization with BCL-XI,and pro-apoptotic activity. J . B i d . Clzerri. 272, 24101-24104. Serine phosphorylation Zha, J., Harada, H., Yang, E., Jockel, J., and Korsmeyer, S. J. (1996~). of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCLXL. Cell 87, 619-628. Zhou, P., Qian, L., Kozopas, K. M., and Craig, R. U'. (1997). Mcl-1, a Bcl-2 family member, delays the death of hematopoietic cells under a variety of apoptosis-inducirigconditions. Blood 89, 630-643. Zhu, W., Cowie. A., Wasfy, G. W., Penn, L. Z., Leber, B., and Andrews, D. W. (1996). Bcl-2 niutaiits with restricted subcellular location reveal spatially distinct pathways for apoptosis in different cell types. EMBO J. 15, 4130-4141. Zoratti, M., and Szabo, I. ( 1995). The rnitochondrial permeability transition. Biochirn. Biophys. Actrr 1241, 139-176. Zou, H., Henzel, W. J., Liu, X., Lutschg, A,, and Wang, X. (1997). Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochroine c-dependent activation of caspase-3. Cell 90, 405-413. This article was accepted for publication on December 15, 1997.
This Page Intentionally Left Blank
ADVANCES IN IMMUNOLOGY, VOL 70
Interleukin-18: A Novel Cytokine That Augments Both Innate and Acquired Immunity HARUKI OKAMURA,' HIROKO TSUTSUl,t SHIN-ICHIRO KASHIWAMURA,' TOMOHIRO YOSHIMOTO,*,t AND KENJI NAKANISHl'it 'loboratory of Host Dekqses, Instifute for Advanced Medical Sciences, beparfmeni of Immunology and Medical Zoology, Hvogo College of Medicine, Nishinomiya, Hvogo 663, Japan
1. Introduction
In the last three decades, along with the progress in cell culture technique, protein biochemistry and molecular biology, numerous surface antigens, cytokines, and molecules involved in intracellular signal transduction have been identified and their genes cloned. Intensive studies also revealed the presence of sophisticated antigen recognition systems that initiate antigen-specific T- and B-cell responses. Interferons are the most well characterized of the cytokines that play a central role in the host defenses to viral, bacterial, and parasitic infection. Type 1interferon ( IFNa, IFNP) is secreted by virus-infected cells, whereas type I1 interferon ( IFNy) is secreted by T and natural killer (NK) cells under certain conditions of activation of these immune cells. Although IFNy was originally identified as a factor with antiviral activity (Wheelock, 1965), this cytokine regulates many aspects of immune responses to infectious agents. T helper cells can be divided into two distinct subsets of effector cells based on the profile of cytokines they produce (Mosmann and Coffman, 1989). T helper type1 (Thl) cells produce interleukin (1L)-2, IFNy, and TNFP, which are associated with inflammation and tissue injury and lead to the development of cell-mediated immune responses. Th2 cells produce IL-4, IL-5, and IL-10, which induce B cells to proliferate and differentiate and evoke humoral immune responses. Macrophages infected with intracellular microbes produce IL-12 that stimulates naive T cells to develop into T h l cells and activates NK cells to produce IFNy (Hsieh et al., 1993; Trinchieri, 1995) that induces IL-12R on naive T cells, which accelerates the polarization to TI11 (Szabo et al., 1997).On the contrary, IL-12 prevents the outgrowth of Th2 cells by induction of IFNy from Thl cells and NK cells (Mosmann and Coffman, 1989; Hsieh et al., 1993; Trinchieri, 1995). Thus, macrophages infected with intracellular microbes can preferentially induce TI11 cells without affecting TI12 development in the orchestration with IFNy. T h l cells in turn produce IFNy, which activates macrophages to promote their lalling activity of microbes Thus, macrophages and NK cells represent a first line of defense against intracellular microbes, and 281 All n$t\
Copvriglit 0 l99h hy Acadrtna Prr% o! reproditchon 111 diiy form reserved OOfij 277WYX $25 00
282
HARUKI OKAMURA et nl
acquired immune response amplifies the action of innate immunity, particularly on repeated exposures to the same pathogens. Interleukin-18 (IL-18), originally called IFNy-inducing factor (IGIF), is a recently cloned cytokine of approximately 18 kDa synthesized by Kupffer cells and activated macrophages (Okamura et al., 1995a).Its major activity is the induction of IFNy production from anti-CD3-activated T h l cells, especially in the presence of IL-12. IL-18 also acts as a costimulant for T h l clones augments IL-2 production and IL-2Ra and Fas L expression (Kohno et nl., 1997; Dao et al., 1996), and activates NK cells to produce IFNy and to express increased level of Fas ligand (Tsutsui et al., 1996). A cDNA for human IL-18 has been cloned (Ushio et al., 1996). Human IL-18 has the capacity to stimulate cytotoxic NK cell activity and stimulates T cells to produce IFNy, IL-2, and granulocyte/macrophage colony-stimulating factor (GM-CSF) (Ushio et al., 1996; Micallef et al., 1996). Although IL-18 may play a critical role in host defense against intracellular microbes, the excess production of IL-18 may induce local or systemic injury in the host (Okamura et al., 1995a). Thus, IL-18 may be a double-edged sword and may require tight regulation to be secreted. This review summarizes the IL-18/IL-l8R system and its pathophysiological roles in immune response. II. Molecular Structure of 11-18 and Its Gene
A. HISTORY OF IL-18
IFNy, initially identified by Wheelock (1965) in the culture supernatant of mitogen-stimulated T cells, is now one of the most well-characterized cytokines contributing to the host defense against various microbes (Farrar and Schreiber, 1993; Billiau, 1996; Boehm et aE., 1997). IFNy is secreted from most, if not all, CD8+ T cells, Thl cells, and NK cells. Secretion of IFNy from T cells and NK cells is induced by stimulation with antigens, anti-CD3, or mitogens and with IL-2, anti-CD16, or activated macrophages, respectively (Farrar and Schreiber, 1993). The action of IFNy is transduced into the nucleus of target cell via several intracellular molecules, i.e., IRFs, JAKUJAK2, and STAT1 (Billiau,1996; Boehmet al., 1997). Activated macrophages-derived cytokines stimulate T cells and NK cells to produce IFNy. IL-12 has been shown to induce the production of IFNy by T cells and NK cells (Trinchieri, 1995). However, the level of IFNy production induced by IL-12 is far less compared to that induced by soluble products from activated macrophages, suggesting the presence of other IGIFs. The authors have been interested in elucidating the regulatory mechanism of IFNy production in viva High levels of IFNy production in oivo can be induced by injection of anti-CD3, PPD, or Staphylococcus
INTERLEUKIN-I8
283
enterotoxin A (SEA) to mice that have been infected with Bacillus Calmette-Guerin (BCG) (Wadaet al., 1985).Similar results were also obtained by injection of anti-CD3 or SEA into mice that had been treated with heat-killed Propionibncteriuni acnes (Okamura et al., 1982).To the authors’ surprise, injection of lipopdysaccharide (LPS) into mice treated with BCG or P. aciw ‘also induced high serum level of IFNy. This induction of IFNy production by LPS stimulation was unexpected, as it is well-established evidence that LPS stimulates macrophages and B cells. The authors could not find any reports that LPS drectly stimulates T cells and NK cells to produce IFNy. Then, two possibilities were predcted: (1) LPS induces high-level production of IFNy by causing macrophages to produce IL-12 or (2) some unidentified mediator induces high-level production of IFNy in uivo. The second possibility was examined (Nakainura et al., 1989,1993). Based on the action, this unidentified factor was tentatively designated as an IFNy-inducing factor. IL-12 stimulates nylon wool-purified T cells in the presence of antiCD3 antibody to produce IFNy. It was found that addition of sera from P. ncnes-primed and LPS-challenged mice also induces IFNy production from T cells (Okamura et al., 1995b). Although these sera contained high levels of IL-12, added sera could induce IFNy production more strikingly than an excess dose of IL-12, substantiating further the presence of IGIF in the sera. The liver tissue extract of P. ncnes-primed and LPS-challenged inice had activity similar to that of sera of mice challenged with LPS (Okamura et nl., 1995b). The authors tried to purify the factor with IFNyinducing activity from the liver extract because liver tissue extract provides rich sources for physicochemical study of this molecule.
B. MOLECULAR STRUCTURE OF IL-18 A N D ITSDESIGNATION Okamura et al., (1995b) purified the protein froin the extracts of liver tissues from P. acnes-primed and LPS-challenged inice by combination of several different types of column chromatography. First, proteins in the liver extracts were precipitated with 45% saturated aininonium sulfate. The precipitated proteins were redissolved and fractionated by a hydrophobic phenyl-Sepharose C L 4 B colunin and an anion-exchange DEAESepharose CL-6B column. The biologically active fraction was separated by chromatofocusing through a Mono P column, ultrafiltrated through a Superdex-75 column, and reverse-phase chroinatographed through a protein C4 column. After these separations, the purified protein was analyzed by preparative SDS-PAGE. Each fraction was tested for its activity to induce IFNy products from anti-CD3-stimulated T cells. After these extensive purification studies, a single protein was obtained with the strong activity. This hoinogeneously purified protein was designated as IGIF.
284
HARUKI OKAMURA et al.
IGIF is a single peptide chain with a molecular weight of 18,000 and a
p l of 4.8. A partial amino acid sequence from N-terminal showed no identity with any other reported proteins. Therefore, the authors tried to determine the amino acid sequence of this purified protein. Purified IGIF was cleaved with trypsin and rechromatographed to yield several fragments. Each fragment was analyzed for their amino acid sequence. Based on the obtained amino acid sequences of two fragments, the authors synthesized mixtures of DNA oligomers (20 based) for the probes to screen the cDNA library. After screening of the mouse liver cDNA library, the sequence of this cytokine was cloned and.determined. Because this new factor is a pleiotropic cytokine and shows redundancy to other cytokines, this novel cytokine was designated interleukin-18. Cloned murine and human IL-18 cDNA encode the novel proteins consisting of 192 and 193 amino acids, respectively (Okamura et al., 1995a; Ushio et al., 1996). From these predicted amino acid sequences, it became clear that IL-18 contains unusual leader sequence, including 35 amino acids, necessary for the secretion of IL-18 from the cell membrane. IL18 was considered to be translated as the precursor form. The 24-kDa precursor was then enzymatically cleaved to produce the 18-kDa mature form IL-18 (Fig. 1) (Bazan et aZ., 1996; Gu et al., 1997; Ghayur et al., 1997). Although IL-18 peptide has no apparent similarity to sequences in the data bases, a homology in the structure by 12% to IL-la and by 19% to IL-lfl is detectable by the fold recognition method (Bazan et al., 1996).
FIG. 1. ICE processing of proIL-18. IL-18 is synthesized as a 24-kDa polypeptide precursor (pro-IL-18) devoid of a conventional signal sequence. This pro-IL-18 is cleaved by ICE at the authentic processing site ( A ~ p ~ ~ - A with s n ~ ~high ) efficacy to generate the biologically active 18-kDa polypeptide IL-18. P. ucnes-primed and LPS-challenged wildtype Kupffer cells produce active form IL-18, whereas ICE-deficient Kupffer cells synthesize the IL-18 precursor but fail to process it into the active form.
INTERLEUKIN-I8
285
These conserved protein structures observed in the positions of P sheets were considered to be important for IL-la and IL-lP to bind IL-1 receptors. These data suggested the possibility that IL-18 shares the common ancestor ligand-receptors. Moreover, the unusual leader sequence may be analogous to the IL-1P prodomain, which is recognized and cleaved by a converting enzyme such as caspase 1. Also, the possibility that IL18 utilizes NF-KB signaling strategy was presumed by the IL-1-binding paradigm. Indeed, IL-18 activates NF-KB (Matsumoto et al., 1997).Thus, IL-18 was proposed to be designated as IL-ly (Bazan et al., 1996). Most of these assumptions were proven to be the case, as will be described later. However, the biological activities of IL-18 are different from those of IL-1, whereas IL-la and IL-1P share almost the same activities (Dinarello, 1989). In addition, IL-la and IL-10 utilize common receptors, but IL-18 does not share the receptor with IL-1 because IL-1 failed to inhibit the binding activity of IL-18 to the cells (Torigoe et al., 1997). Therefore, the authors propose to designate IGIF as IL-18 to emphasize the differences of IL-1 and IL-18. 111. Producing Cells
Activated macrophages were first shown to express high levels of IL-18 (Okamura et al., 1995a). Investigations have revealed that IL-18 is expressed by various types of cells other than macrophages. Udagawa et nl. (1997) have shown that some osteoblasts produce IL-18 that suppresses the differentiation of osteoclasts via the induction of GM-CSF but not IFNy. Thus, IL-18 might be involved in the regulation of bone homeostasis. In addition to constitutive expression of IL-18 mRNA, its expression is also induced in many organs in various diseases. Stoll et aZ. (1997) observed that IL-18 mRNA is expressed constitutively in epidermal keratinocytes and that this expression is promptly upregulated after stimulation with contact sensitizers in vivo. Keratinocytes produced functional IL-18 after stimulation with allergen. It can be assumed that IL-18 enhances Thlmediated skin diseases, although the mechanism of activation of caspase 1 in the skin remains to be elucidated. Interestingly, IL-18 mRNA is induced in the adrenal cortex, particularly in the zona reticularis and fasciculata that produce glucocorticoid by acute cold stress (Conti et aZ., 1997). IL-18 mRNA expression is also upregulated in pancreas in nonobese diabetic mice after the onset of diabetes mellitus (Rothe et al., 1997). These reports suggest that IL-18, like IL-1, may play an important role in connecting the immune system to other biosystems, such as endocrine and nervous systems.
286
HARUKI OKAMURA et nl.
It has been demonstrated that the secretion of mature IL-18 requires the cleavage of the precursor IL-18 by caspase-1 in Kupffer celIs (Gu et al., 1997;Ghayur et al., 1997) (described later), indicating that the expression of IL-18 in mRNA does not always mean the production of functional IL18 at the protein level. Thus, mature IL-18 could not be detected by reverse transcriptase polymerase chain reaction (RT-PCR) or Northern blotting analysis. To identify functional IL-18, a bioassay such as an IFNy production inducing assay needs to be perfonned (Okamura et al., 1995a). In fact, unstimulated Kupffer cells from nontreated mice expressing IL18 in mRNA level do not produce functional IL-18, but appropriately activated Kupffer cells secrete it (Tsutsui et al., 1997). It is again noted that careful experimental determination is required for the detection of functional IL-18. The authors now describe the typical case of macrophages. Because resident peritoneal adherent cells did not express detectable levels of IL18 mRNA and because P. acnes-induced peritoneal exudate cells express a high level of IL-18 mRNA (Okamura et al., 1995a), IL-18 seemed to be induced in association with the activation of marcophages. In fact, mature IL-18 is secreted from LPS-activated P. acnes-elicited Kupffer cells (Okamura et al., 1995a; Tsutsui et al., 1997). When Kupffer cells derived from P. acnes-primed mice are incubated by themselves, they do not produce mature IL-18, despite their expression of IL-18 mRNA. Even though Kupffer cells from nontreated mice are stimulated with LPS, they do not produce mature IL-18. To date, only the sequential stimulation with P. acnes and LPS induces mature IL-18 production from Kupffer cells. As for peritoneal macrophages, the combination of these stimuli does not induce the production of bioactive IL-18 from them. Thus, although macrophages are the producer cells of IL-18, there seems to be differences in the potency of production of mature IL-18 among their subsets or differential stages. IV. Requirement of Caspase-1 for Processing of 11-18
Caspases, originally termed IL-lP converting enzyme (ICE) or CED3 protease family with a substrate specificity for aspartic acid, play pivotal roles in inflammation and apoptotic cell death (Alnemri et al., 1996; Henkart, 1996). Caspase-3 (CPP32) is involved in apoptosis (Henkart, 1996; Kuida et al., 1996). Caspase-1 (ICE) is essentially required to induce production of the bioactive mature IL-1P, a potent proinflammatory cytokine, from its bioinactive precursor form, proIL-lP (Thornberry et al., 1992; Kuida et al., 1995; Li et al., 1995). Caspase-1 is synthesized as an inactive 45-kDa proenzyme, processed to an active tetramer of two
INTEHLEUKIN-18
287
10-kDa and two 20-kDa subunits and localized at tlie plasma membrane (Walker et al., 1994; Singer et al., 1995). Mice deficient in caspase-1 have a defect in the export of mature IL-1fi after stimulation with LPS (Kuida et nl., 1995; Ki et al., 1995). Caspase-1 cleaves the IL-18 precursor polypeptide into a bioactive IL18 (Gu et al., 1997; Ghayur et al., 1997). Murine IL-18 is synthesized as a precursor polypeptide of 192 amino acids (proIL-18)laclang a conventional leader sequence (Okamura et al., 1995a), which suggests that proIL-18 may be analogous to proIL-lfi (Bazan et al., 1996). ProIL-18 of 24 kDa is cleaved at the site of A~p’~-Asn’‘into the bioactive polypeptide of 18 kDa (Gu et al., 1997; Ghayur et al., 1997), suggesting that aspartic acid-specific protease, including tlie caspases, may be involved. Indeed, coexpression studies of proIL-18 with various caspases revealed that only the coexpression with caspase-1 resulted in the production of 18 kDa mature IL-18 (Gu et al., 1997). Caspase-4 also cleaved proIL-18, but in much less yield. This was also tlie case for an in vitro cleavage assay using both recombinant human and murine proIL-18 as substrates. In addition, the cleavage of proIL-18 by caspase-1 was inhibited by a specific inhibitor or by using inactive mutant caspase-1 instead of wild-type caspase-1. Murine proIL-18 was cleaved by human recombinant caspase-1 into functional IL-18, which had the ability to induce IFNy production by murine splenocytes or cloned T cells (Gu et al., 1997; Ghayur et al., 1997). The essential role of caspase-1 in the production of bioactive mature IL-18 was also proven by mice deficient in caspase-1. LPS-activated Kupffer cells from wild-type mice, that had been treated with P. acnes, secreted mature IL-18, whereas Kupffer cells from P. acnes-primed caspase-1 -/- mice did not secrete it (Gu et al., 1997). When caspase-1 +/+ mice were sequentially treated with heat-killed P. acnes and a small amount of LPS or with a large amount of LPS alone, the serum level of IL-18 and IFNy was elevated rapidly. In contrast, caspase-1 -/- mice did not respond to such treatment with the facilitated serum level of IL-18 and IFNy (Gu et al., 1997; Ghayur et al., 1997).This abnormality was improved completely by treatment with recoinbinant IL-18 but was not affected by IL-la or IL-1fi treatment (Ghayur et al., 1997). Thus, caspase-1 plays a critical role in the processing of proIL-18. V. Biological Function
A. EFFECTS O N T CELLS IL-18 has been described originally as a factor that enhances IFNy production from anti-CD3-stimulated T h l cells, particularly in the presence of IL-12 (Okamura et nE., 1995a). Naive T cells do not respond or
288
HARUKI O K A M U M et al.
respond meagerly to IL-12 or IL-18 on anti-CD3-coated plates. However, the combination of IL-12 and IL-18 induces them to proliferate and to produce IFNy in a synergistic manner (Okamura et al., 1995a; Yoshimoto et al., 1997; Ahn et al., 1997). In general, IFNy production by T cells is induced following stimulation with anti-TCR (anti-CD3) and anti-CD28 (Lindstein et al., 1989). Therefore, IL-18 was described originally as acting as a costimulatory factor for the production of IFNy by anti-CD3stimulated T cells in the presence of IL-12 (Okamura et al., 1995a; Yoshimot0 et al., 1997; Kohno et al., 1997; Ahn et al., 1997). However, it turned out to be the case that the level of IFNy in the culture supernatants of T cells stimulated with IL-12 plus IL-18 in the absence of anti-CD3 extends to the same level, indicating that T cells require no anti-CD3 stimulation for the IFNy production by IL-12 and IL-18 stimulation (T. Yoshimoto, unpublished observation). To substantiate this conclusion further, T cells were stimulated with IL-12 and IL-18 in the presence of various doses of cyclosporin A (1-100 ng/ml). This addition of cyclosporin A did not affect the action of IL-12 and IL-18 to induce IFNy production from T cells. These results indicate that the TCR-mediated induction of the nuclear factor of activated T cells (NFAT) is not required for IFNy production by T cells stimulated with IL-12 and IL-18 (T. Yoshimoto, unpublished observation). Because IL-12 and IL-18 induce T cells to produce IFNy in a synergistic manner, the mechanism of synergy between IL-12 and IL-18 was studied. T cells stimulated with IL-12 and IL-18 for 3 days strikingly produce IFNy. T cells stimulated with IL-12 for 3 days and recultured with IL-18 also produce high levels of IFNy, whereas T cells stimulated with IL-18 for 3 days and subsequently stimulated with IL-12 produce IFNy meagerly (T. Yoshimoto, unpublished observation), indicating that the induction of IFNy production by T cells is regulated by the ordered action of IL12 and IL-18. Like murine T cells, human T cells also require being simultaneously stimulated with IL-12 and IL-18 or sequentially stimulated with IL-12 and IL-18 to produce significant amounts of IFNy (Tominaga, unpublished observation). Highly purified human T cells from tonsils produce a very small amount of IFNy when stimulated with IL-12 alone for 6 days, whereas T cells cultured with IL-12 and IL-18 can produce over 20 times more IFNy. Measurement of IL-18R on IL-12-stimulated or nonstimulated T cells using '251-IL-18reveal that IL-12-stimulated T cells express IL-18R in a dose-dependent manner, and IL-18 dose dependently induced such IL-12-stimulated T cells to produce IFNy. In addition to acting as a costimulatory factor for IFNy production, IL-18 enhances the production of GM-CSF, IL-2, and IL-2Ra by T cells and augments Tcell proliferation (Ushio et al., 1996; Micallef et al., 1996; Kohno et al.,
INTERLEUKIN-I8
289
1997). Thus, IL-12 and IL-18 induce T cells to proliferate and to produce a large amount of IFNy that amplifies both innate and acquired immunity and may provide us with strong defense system.
B. EFFECTS ON B CELLS The activation, proliferation, and differentiation of B cells are highly regulated events in which the action of T cells and their soluble products play a major role (Nakanishi et al., 1984a,b; Howard et al., 1984; Kishimoto, 1985). However, several reports have demonstrated that activated B cells also produce cytokines such as IL-1 (Pistoia et al., 1986), IL-6 (Smeland et al., 1989), and IL-10 (Gerard et al., 1993; Velupillai and Harn, 1994). It has been shown that a coinbination of IL-12 and IL-18 induces antiCD40-activated highly purified murine B cells to produce IFNy, which inhibits IL-4-dependent IgE and IgGl production and enhances IgG2a production without inhibiting the B-cell proliferative response (Yoshirnoto et al., 1997).It has also been demonstrated that anti-CD40 is not a prerequisite for inducing IFNy-producing B cells, although stimulation with antiCD40 enhances IFNy production from B cells (Yoshimoto et al., 1997). Thus IL-12- and IL-18-stimulated B cells act as regulatory cells that differentially regulate IgGlAgE and IgG2a responses by production of IFNy in zjitro. As T cells require costimulation with IL-12 to respond to IL-18 by marked production of IFNy, B cells also require stimuIation with IL-12 to become responsive to IL-18. To reveal whether IFNy production by B cells is also regulated orderly by IL-12 and IL-18, B cells stimulated with IL-12 for 3 days were examined for their capacity to proliferate and to produce IFNy in response to IL-18. These B cells responded to IL-18 by their marked proliferation and production of IFNy (T. Yoshimoto, unpublished observation). The capacity of human B cells to produce IFNy in response to IL-12 and IL-18 was also examined. Purified human B cells from tonsils weakly produce IFNy when stimulated with IL-12 alone for 5 days, whereas B cells cultured with IL-12 and IL-18 produce IFNy strongly (Tominaga, unpublished observation). The potentiality that B cells produce IFNy may have enormous implications concerning understanding the regulation of immune response against infection as well as the traditional division of cell-mediated and humoral immunity. C. EFFECTS ON T h l
AND
Th2 RESPONSES
The differentiation of naive CD4+ T cells into IL-4-producing Th2 cells or IFNy-producing Thl cells depends on their mode of “priming” (Swain et al., 1991; Seder and Paul, 1994). IL-4, itself present at the time of priming, plays a critical role in the development of naive T cells into IL-
290
HARUKI OKAMURA et
Nr.
4 producers (Swain et al., 1990; Seder et al., 1992). Yoshimoto and Paul (1994) and Yoshimoto et al., (1995a,b) revealed that a potential source of IL-4 that could affect the priming of naive CD4' T cells is a set of CD4+NK1.1t splenic T cells that can produce IL-4 within 30 to 90 min of in vivo challenge with anti-CD3 antibodies or Staphylococcus enterotoxin B (SEB). IL-12, produced by activated macrophages and dendritic cells, has been shown to induce Thl-associated responses by stimulating T cell and NK cell production of IFNy and by inhibiting T-cell production of IL-4 (Hsieh et al., 1993; Seder et al., 1993; Trinchieri, 1995). Kohno et al. (1997) demonstrated that IL-18 is a costimulatory factor on the activation of T h l but not Th2 cells. The reactivity of IL-18 on T h l and Th2 cells was examined using Thl and Th2 clones. The results were essentially the same within clones of the same type. All T h l clones responded to IL-18 by the augmented IFNy and IL-2 production and cell proliferation, whereas little or no substantial augmentation of IL-4 production and proliferation in response to IL-18 was observed in any Th2 clone. They further demonstrated that IL-18, as well as IL-12, was released through interaction between Thl cells and APC in the presence of specific Ag (Kohno et al., 1997).These results may indicate a differential responsiveness of Thl and Th2 clones to IL-18, although further study using other T h l or Th2 clones with different antigen specificities would be required to confirm this observation. As mentioned earlier, it is well known that IL-12 induces naive CD4+ T cells to differentiate into Thl cells (Swain et aE., 1991; Hsieh et al., 1993; Seder et al., 1993; Seder and Paul, 1994; Trinchieri, 1995). It is, therefore, interesting to examine whether IL-18 can polarize naive CD4' T cells to become T h l cells. It has been demonstrated that liver lymphocytes have a substantial proportion of IL-4 producing CD4'NKl.l' T cells. Injection of P. acnes downregulated IL-4 production by diminishing the number of CD4'NKl.l' T cells via production of IL-12 and IL-18 from Kupffer cells (Matsui et al., 1997). Furthermore, this treatment upregulates IFNy production by increasing CD4-NK1.1- T cells among CD4- liver lymphocytes. Like P. acnes treatment, injection of IL-12 downregulated IL-4 production but upregulated IFNy production by liver lymphocytes. In contrast, injection of IL-18 fails to replace the action of P. acnes-elicited Kupffer cells. However, injected IL-12 and IL-18 act in a synergistic manner for the diminution of CD4'NK1.1+ T cell and for induction of IFNy-producing cells. These results indicated that IL-18 itself had no ability to induce naive CD4' T cells to differentiate into T h l cells, but it promoted the action of IL-12 by the activation of IL-12-stimulated T cells. Furthermore, in an in vitro system, plate-bound anti-CD3 plus IL-12 can polarize naive CD4+ T cells to Th1 cells, whereas plate-bound anti-CD3 plus IL-18 fails to polarize (T. Yoshimoto, unpublished observation).
1NTERL.EUKIN-I8
29 1
Although a more detailed study concerning the function of IL-18 for the polarization of naive CD4+ T cells into T h l or Th2 in vitro and in zjioo is required at present, IL-18 is considered as a costimulatory factor that induces IFNy production from T cells, B cells, and NK cells. Figure 2 summarizes the current interpretation of the action of IL-18 in development and activation of Thl, NK, and B cells.
D EFFECTS ON NK CELLS IL-18 increases cytotoxic activity of spleen cells against YAC-1 cell (Hyodo, unpublished observation). The effect of IL-18 was not inhibitable by the treatment with antibodies against IL-2, a potent NK cell stirnulatory factor, indicating that IL-18 directly, or at least not via IL-2, upregulates the cytotoxicityof lymphocytes.This action was inhibitable by concanamycin A, an inhibitor of exocytosis (Muroi et al., 1993; Kataoka et al., 1996),indicating that the facilitated cytotoxicity is dependent on the exocytosis of cyto-
FIG.2. IFNy production by T and B cells is sequentially regulated by IL-12 and IL18. IL-12 is produced by macrophages infected with intracellular microbes (P ncnes). IL12 induces NK cells to produce IFNy, which in turn activates macrophages. Furthermore, IL-12 induces Tho and B cells to develop into IL-18R+ Th1 and B cells, respectively. Therefore, NK, Thl, and B cells produce IFNy strongly in response to IL-18, particularly in the presence of IL-12. Thus, IL-18-induced IFNy production can further activate infected inacrophages and enhance their killing action. IL-18 effectively stiinulates NK cells and IL-12-stimulated T and B cells to produce IFNy. However, such stirnulatory action of IL18 on Th2 cells is uncertain at the present time.
292
HARUKI OKAMURA et nl
killing substances such as perforin and granzyme A (Kagi, 1996). Because IL-18 had no such effect on the spleen cells derived from NK cell-depleted mice, IL-18 facilitates the cytotoxicity of NK cells. IL-12, originally designated as a natural killer cell stimulatory factor (Kobayashi et al., 1989; Tanaka et al., 1996), is a potent growth as well as activating factor for NK cells. Interestingly, IL-12 and IL-18 additively augmented the NK activity of spleen cells. This orchestration of IL-12 and IL-18 in enhancing NK cell activity is demonstrated in vivo. When IL-12 and IL-18 were administered to mice, the early exclusion of live Listeria monocytogenes, which is mainly dependent on the NK cell activity of the host, was augmented in the host liver and spleen (Hyodo, unpublished observation). IL-18 enhances the action of a second killing apparatus of NK cells, Fas ligand ( FasL) (Tsutsui et al., 1996). The Fas/FasL system plays a critical role in the maintenance of immune homeostasis (Nagata and Golstein, 1995). In general, activated lymphocytes, through their antigen or cytokine receptor, are required to be eliminated soon after their roles are over. Apoptotic cell death system is employed in the destruction of postactivated lymphocytes, presumably because apoptosis, death without rupture of the cell membrane (Smith et al., 1989), does not prevail in an additional inflammatory reaction. Both lpr/lpr mice, naturally occurring mutant mice lacking functional Fas, and mice artificially deficient in Fas have the peripheral immune tissue accumulated with abnormal lymphocytes (Wu et al., 1994; Adachi et al., 1996).This was also observed in mutant mice for FasL, gld/gld (Takahashi et al., 1994). Fas is constitutively expressed on many organs, whereas FasL is an inducible molecule (Nagata, 1994; Nagata and Golstein, 1995). Although FasL is induced on T cells via the activation of TCR, the precise mechanism underlying this induction has yet to be elucidated. To investigate how FasL is regulated in NK cells, a cloned NK cell line was established from the liver of BALB/c nude mice (Tsutsui et al., 1996). The cloned NK cells, negative for TCR or BCR but positive for the p chain of IL-2R, did not kill conventional NK target cells, such as YAC-1 and B16, a melanoma cell line (Tsutsui et al., 1996). However, they expressed a small amount of functional FasL. IL-18 upregulated functional FasL expression on the cloned hepatic NK cells, which was not inhibitable by the treatment with anti-IFNy, suggesting that IL-18 directly augments FasL expression (Tsutsui et al., 1996). In contrast, IL-lp, TNFa, or IFNy had no FasL-enhancing activity (Tsutsui et al., 1996). Thus, one of the physiological roles of IL18 is the upregulation of FasL on lymphocytes. E. EFFECT ON IgE RESPONSE Helminths such as Nippostrongylus brasiliensis (Nb), Heligmosomoides polygyrms, Ascaris, and schistosomes induce increases in the levels of IgE
INTERLEUKIN-18
293
and in the number of eosinophil in the host (Finkelman et al., 1990). Injections of anti-IL-4 (Finkelman et al., 1986) and anti-IL-5 (Finkelman et al., 1990) inhibited this Nb-induced IgE response and eosinophilia, respectively, indicating that these responses are IL-4 and IL-5 dependent, respectively. IL-4 is a pleiotropic cytokine that induces B cells to develop into IgE-producing cells (Coffman et al., 1986). The production of IgE in vitro and in vivo can be inhibited by injection of IFNy (Snapper and Paul, 1987) or IL-2 (Nakanishi et al., 199.5).CD4+ T cells differ in the patterns of cytolanes they express (Mosmann and Coffman, 1989).T h l cells secrete IFNy and IL-2, and Th2 cells secrete IL-4, IL-5, and IL-10. Therefore, it has been suggested that the balance of Th2 and T h l cells largely determines the levels of IgE produced during an immune response. IL-12 polarizes naive T cells to develop into T h l cells (Hsieh et al., 1993; Seder et al., 1993; Trinchieri, 1995), suggesting that IL-12 may provide a unique therapeutic way for the treatment of allergic disorders. Indeed, Finkelman et al. (1994) reported that IL-12 inhibits IgE production in vivo via its action to promote endogenous IFNy production. Because IL-18 induces IFNy production from Thl cells more strikingly than IL-12 (Okamura et al., 1995a), the authors examined whether IL-12 andlor IL-18 inhibit this IL-4-dependent IgE response in Nb-inoculated mice. IFNy+” and IFNy-’- C57BLl6 mice were used. Injection of IL-12 into Nb-inoculated- or anti-IgD-injected IFNy+’+mice dose dependently inhibited IgE production, whereas injection of IL-18 only modestly inhibited IgE production in these mice. However, daily ip injection of IL-18 with a suboptimal dose of IL-12 into Nb-inoculated- or anti-IgD-injectedIFNy+’+mice almost completely inhibited IgE production but enhanced IgG2a production markedly (Yoshimoto et al., 1997). Like anti-IgDinjected-IFNy+’+mice, IFNy-’- mice produced IgE in response to antiIgD. However, injection of a mixture of IL-12 and IL-18 failed to inhibit IgE production in such treated IFNy-I- mice. Furthermore, injection of the anti-IFNy antibody reversed the IL-12- and IL-18-induced IgE inhibition in normal mice (Yoshimotoet al., 1997). These results indicated that endogenous IFNy from IL-12- and IL-18-stimulated cells can suppress IgE production. IL-18 acts on Thl cells and in combination with IL-12 strongly induces them to produce IFNy (Okamura et al., 1995a). Indeed, T cells from anti-IgD-, IL-12-, and IL-18-treated mice produced high levels of IFNy (Yoshimoto et al., 1997). To the authors’ surprise, B cells from these mice also produce IFNy (Yoshimotoet al., 1997).Furthermore, B cells obtained from normal mice could develop into IFNy-producing cells in IFNy-’host mice in response to in vivo treatment with IL-12 and IL-18 (Yoshimoto et al., 1997). In an in vitro system, it was also demonstrated that IL-12
294
HAHUKI OKAMURA et al.
and IL-18 induce anti-CD40-activated B cells to produce IFNy, which inhibits IL-4-dependent IgE and IgGl production and enhances IgG2a production (Yoshimoto et nl., 1997). Li et al. (1996) have reported that IL-12 stimulates human B cells to produce IFNy. These results indicate that B cells can act as regulatory cells in the immune response. As IgE responses are an important factor in allergic responses, the authors suggest that IL-12/IL-18 treatment could present a unique approach for the treatment of allergic diseases.
F. EFFECT ON OSTEOCLASTS Udagawa et al. (1997) have shown that osteoblasts produce IL-18 that suppresses the differentiation of osteoclasts though induction of GM-CSF but not IFNy. They established two types of osteoblastic stoma1 cells. Recombinant IL-18 inhibited osteoclast formation in cocultures of osteoblasts and hemopoietic cells in the presence of osteoclastogenic agents. This inhibition was shown to be due to GM-CSF and not to IFNy, both of which are induced by IL-18. Antibodies against IFNy failed to prevent the effect of IL-18. Moreover, IL-18 inhibited the formation of osteoclasts from hemopoietic cells from IFNyR-deficient mice. On the contrary, neutralizing antibodies against GM-CSF prevented the effect of IL-18 on osteoclast formation. These results imply that IL-18 participates in the local control of osteoclastogenesis. IL-1 has been shown to augment the formation of osteoclast (Dinarello, 1989),whereas IL-18 inhibits this formation. Therefore, it will be necessary to compare the function of IL-18 and IL-1 in an osteogenesis system. IL-18 may regulate bone homeostasis by counteracting IL-1. VI. Receptors for 11-18
To determine the molecular structure of IL-l8R, monoclonal antibodies against human IL-18R were established (Torigoe et al., 1997).These monoclonal antibodies allowed molecular cloning of cDNA encoding human IL-18R. Therefore, the regulation and molecular nature of human IL-18R were demonstrated. Subsequently, progresses in studies concerning the structure, regulation, and function of mIL-18R on T and B celIs were shown. A. EXPRESSION OF IL-18R ON HUMAN T CELLS To reveal the molecular structure, function, and cell distribution of IL18R, establishment of monoclonal antibodies against IL-18R was essential. Torigoe et aZ. (1997) first screened for cell lines that had high-level expression of IL-18R in 18 human leukemic cell lines of different origins using
INTERLEUKIN-18
295
12sI-hIL-18.They selected the Hodgkin’s disease cell line, L428, as the most strong hIL-18R-expressing cell line. L428 cells express low-affinity IL-18R (18,00O/cell, & of 18.5 nM) and no high-affinity IL-18R. Torigoe et al. (1997) first tested whether this IL-18-binding molecule could bind IL-1p because the tertiary structures of IL-18 and IL-lP have weak similarity. Thus, they examined the capacity of IL-18 and IL-1p to inhibit the binding of 12s1-hIL-18to L428 cells. Results clearly indicated that the binding of lZ5I-hIL-18to L428 cells was specificallyinhibited by the addition of unlabeled hIL-18, indicating that IL-18R is distinct from IL-1R (Torigoe et al., 1997). They then tried to establish monoclonal antibodies against IL-18R by immunizing BALB/c mice with L428. They selected monoclonal antibodies against IL-18R based on their capacity to inhibit 1251-IL-18 binding to LA28 cells. Through these procedures, they established the anti-hIL-18R mAb (No. 117-1OC)that inhibits 12,51-IL-18 binding to L428 cells and as described later, using this monoclonal antibody, they purified IL-18R and determined the amino acid sequence and finally cloned cDNA encoding for human IL-18R. B. REGULATION OF IL-18R BY IL-12 Freshly purified T cells from human tonsils cannot respond to IL-18 by their proliferation and production of IFNy, whereas T cells stimulated with IL-12 showed marked responsiveness, suggesting that IL-12 may induce an increase in the expression of IL-18R. To determine the period required for rendering T cells most responsive to IL-18, T cells were incubated with IL-12 for up to 6 days. Human T cells stimulated with IL12 for 5-6 days have the highest responsive property to IL-18. Binding of IZ51-h1L-18revealed that human T cells cultured with medium alone meagerly express low-affinity IL-18R (& 30 nM), whereas T cells cultured with IL-12 for 6 days express it strongly (Tominaga, unpublished observation). Flow cytometric analysis using anti-hIL-18R mAb further revealed that human T cells cultured with medium alone or IL-18 do not have molecules recognized by this monoclonal antibody, whereas T cells cultured with IL12 or IL-12 plus IL-18 do. This anti-hIL-18R monoclonal antibody inhibits 1251-hIL-18binding to IL-18R and completely inhibits the capacity of IL18 to induce IL-12-pretreated T cells to proliferate and to produce IFNy. Taken together, these results indicate that IL-12 induces IL-18R and that such IL-12-stimulated T cells are responsive to IL-18.
C. MOLECULAR CLONING OF IL-18R Molecular cloning of IL-18R has been accomplished using the hIL-18Rexpressing cell line L428 (Torigoe et al., 1997).Human IL-18R was purified from LA28 cells followed by wheat germ agglutinin-Sepharose 4B chroma-
296
HARUKI OKAMURA et nl.
tography and anti-hIL-18R monoclonal antibody-Sepharose chromatography. The N-terminal and internal amino acid sequences of hIL-18R were determined, and internal amino acid sequence matched those of the human IL-1R-related protein (hIL-1Rrp). IL-1Rrp cDNA has been cloned by RTPCR using degenerated primers for conserved amino acid sequences in the IL-1R type 1 sequence (Parnet et al., 1996). IL-1 does not bind to IL-lRrp, for which the ligand has been unknown (Parnet et al., 1996). COS-1 cells transfected with hIL-1Rrp cDNA express low-affinity hIL18R (49,00O/cell,& 46 nM) by Scatchard plot analysis using lS1-hIL-18, indicating that IL-18R is composed of IL-1Rrp (Torigoe et al., 1997). Furthermore, these COS-1 cells transfected with cDNA encoding for hILlRrp were stained specifically with FITC-anti-18R monoclonal antibody. The capacity of expressed hIL-1Rrp to drive the IL-18 signal was examined by measuring the NF-KB DNA binding (Parnet et al., 1996). Results revealed that IL-18R (IL-1Rrp) is capable of inducing NF-KB DNA binding ability in response to stimulation with rhIL-18 (Torigoe et al., 1997), suggesting that the functional IL-18R component is IL-1Rrp. As noted earlier, because IL-12 stimulation also induces high-affinity IL-18R on T cells, further studies are needed to identify the functional role of this receptor in IL-18 signal transduction. OF IL-18R ON MURINE T AND B CELLS D. EXPRESSION As reported previously, IL-18, especially in combination with IL-12, induces Thl clones to produce IFNy (Okamura et al., 1995a; Kohno et al., 1997). Using an IL-12-responsive T-cell clone, 2D6 (Ahn et al., 1997), the authors investigated how IL-12 and IL-18 collaborate for IFNy production. 2D6 that had been fed with the culture medium containing IL-12 produced IFNy in response to IL-18, whereas 2D6 that had been fed with IL-2 did not. Furthermore, 2D6 obtained from cultures containing IL-12 could bind IL-18, indicating that IL-12 stimulation induces IL-18R on 2D6. It was further investigated whether IL-12 and IL-18 collaborate for IFNy production from freshly prepared naive T cells. Consistent with results using human T cells, murine T cells that had been stimulated with IL-12 for 3 days exhibited dose-dependent proliferation and IFNy production in response to IL-18, suggesting that IL-12 renders murine T cells responsive to IL-18 by induction of IL-18R. The number and affinity of IL-18R on T cells before and after stimulation with IL-12 for 3 days were examined. Nonstimulated or IL-18-stimulated T cells did not specifically bind lz51-IL-18.In contrast, T cells stimulated with IL-12 had the capacity to specifically bind lSI-IL-18. The shape of the Scatchard plot obtained from the initial binding study is consistent of the presence of both high-affinity and low affinity IL-18 binding sites. Measurement of
INTERLEUKIN-18
297
binding of lZ5I-IL-l8on IL-12-stimulated T cells revealed that they express 405 high-affinity IL-18R (& 430 pM) and 5500 low-affinity IL-18R (& 31.4 nM) (T. Yoshimoto, unpublished observation). Murine B cells stimulated with IL-12 also respond to IL-18 by their growth and significant production of IFNy. Therefore, the presence of high and low affinity IL-18R on B cells stimulated with IL-12 for 3 days was examined. Like IL-l2-stimulated T cells, IL-12-stimulated B cells express both high-affinity and low-affinity IL-18R. The initial Scatchard plot analysis revealed that there are 160 high-affinity IL-18R (& 457 pM) and 2400 low-affinity IL-18R (& 93.6 nM) on IL-12-stimulated B cells. In contrast, nonstimulated B cells did not express these two types of IL18R (T. Yoshimoto, unpublished observation). cDNA encoding for human IL-18 receptor (IL-18R) has been cloned. From this sequence, the authors cloned cDNA encoding for murine IL18R by PCR. Using this cDNA as a probe, the capacity of IL-12 and/or IL-18 to induce an increase in the expression of IL-18R-mRNA in murine splenic T and B cells was examined by Northern blot analysis. Murine splenic T cells cultured by themselves or with IL-18 express no IL-1Rrp mRNA, whereas T cells cultured with IL-12 or IL-12 plus 1L-18 clearly express IL-1Rrp mRNA. In contrast, murine B cells cultured with IL-12 or IL-12 plus IL-18 failed to express IL-1Rrp mRNA, although these B cells express both high- and low-affinity IL-18 R when examined by Scatchard plot analysis (T. Yoshimoto, unpublished observation). Thus, it was needed to detect IL-1 Rrp rnRNA in IL-12-stimulated B cells using the RT-PCR method. E. IL-ISR-MEDIATED SIGNALING PATHWAY
Many cytokines signal through different cell surface receptors to activate the transcription factor NF-KB in the nucleus, where it binds to specific regulatory DNA sequences in the promoters of several cytokine-inducible genes (Baeuerle and Henkel, 1994). It has been reported that COS cells transfected with the chimeric receptor, in which the IL-1Rrp cytoplasmic domain is fused to the extracellular and transmembrane region of type 1 IL-IR, responds to IL-1 by activation of NF-KB (Parnet et al., 1996). Torigoe et al. (1997) revealed that COS-1 cells transfected with hIL-1Rrp cDNA express low-affinity hIL-18R and respond to hIL-18 by activation of NF-KB, further indicating that a functional IL-18R is composed of IL1Rrp. Because the IL-12 signal is transduced by STAT4 (Szabo et al., 1995), a synergistic action seen with IL-12 and IL-18 may be explained by the IL-l2-dependent induction of IL-18R and by binding of NF-KB and STAT4 to the promoter region of the IFNy gene.
298
HARLJKI OKAMURA et ol.
Members of the TRAF protein family have been implicated in activation of NF-KB by the TNF receptor superfamily (Hsu et al., 1995). The new TRAF protein (TRAFG)has provided insight into the mechanism of NFKB activation through the IL-l/IL-lR (Cao et al., 1996). As mentioned earlier, the amino acid sequences of the type 1IL-1R cytoplasmic domain are highly conserved with that of IL-1Rrp (IL-18R) (Torigoe et al., 1997). Therefore, it is postulated that the receptor proximal signaling events leading to activation of NF-KB by IL-18 are mediated by TRAFG, which is under investigation. VII. Role of 11-18 in Host Defenses
Immunity is necessary for host survival, but also has the potential to cause injury to the host. Mice are relatively resistant to LPS. However, P. acnes pretreatment can sensitize them to LPS-induced lethal shock (Tsutsui et al., 1992; Yoshimoto et al., 1992). The authors investigated whether T cells are required for this sensitization and found that T-cell-deficient mice are highly resistant to this sequential treatment with P. acnes and LPS but after receiving T cells they become susceptible to this sequential treatment. Thus, T-cell-mediated immunity in some circumstances can cause severe injury to the host. Protection against microbes is mediated by both innate and acquired immunity. Phagocytic cells and NK cells constitute the innate immune system, and T cells and B cells constitute the acquired immune system. The principal protective immune response against extracellular bacteria consists of specific antibodies that opsonize the bacteria for phagocytosis and acivate the complement system. Intracellular microbes cannot be monitored by specific antibodies because of their intracellular nature. Defense against intracellular microbes is principally mediated by phagocytic cells. However, intracellular microbes can survive and replicate within host cells because they have developed mechanisms for resisting lysosomal degradation in phagocytic cells. Thus, acquired immunity that amplifies innate immunity is required for killing of intracellular microbes. Specific immune responses to intracellular bacteria consist of two type of reactions: (1)killing of phagocyted microbes by activation of macrophages with IFNy (Sher and Coffman, 1992) and (2) lysis of infected cells by CD8+ T cells. The best documented example of the first reaction is infection of mice with Leishnulnia major, a protozoa that survives within macrophages. An example of the second reaction may be infection caused by protozoa of the genus Plasmodium. Thus, the role of IL-18 in these infections was examined.
INTEKLEUKIN-18
299
A LEISHMANIA MA]OR Cutaneous inoculation of most inbred strains with L. major induces a localized lesion that heals spontaneously and has an acquired resistance to reinfection. However, some strains are susceptible to this infection and suffer from fatal visceral disease. Resistance and susceptibility of inbred strain of mice to L. major are associated with IFNy and IL-4 production by T cells, respectively (Slier and Coffinan 1992; Reiner and Locksley, 1995). CD4' T cells from regonal lymph nodes of infected resistant C57BU6 mice contain high levels of IFNy but little IL-4, whereas CD4' T cells from infected susceptible BALB/c mice contain high levels of IL4 but little IFNy (Heinzel et al., 1991). Healing and nonhealing of the infected host can be convertible. Host resistance can be transferred by the administration of anti-IL-4 (Sadick et al., 1990) or IL-12 (Heinzel et al., 1993) into susceptible mice. In contrast, host susceptibility can be transferred by alternative treatments (Belosevic et al., 1989; Heinzel et al., 1995). Injection of IL-12 into nonhealing mice such as BALB/c can induce Thl cells that produce IFNy and protect the infected host froin suffering from visceral leishmaniasis (Heinzel, 1993), indicating that IL-12 is a powerful factor modulating host immunity. Although IL-12 is a major deterniinant of naive T cells into IFNy-producing Th1 cells, administration of higher doses of IL-12 has been reported to be very toxic to the host (Hall, 1995). IL-18 acts as a strong cofactor for IFNy production froin T cells (Okamura et al., 1995a; Yoshiinoto et al., 1997; Matsui, 1997).Thus, injection of IL-12 and IL-18 may provide a very powerful way for the treatment of susceptible mice to L. major. Suboptimal doses of IL-12 were used to minimize its toxicity.This dose could not cure susceptible mice. Administration of various doses of IL-18 poorly modulated host immunity. However, a combination of IL-12 and IL-18 completely cured susceptible mice, providing us with a rationale to use this combination for treatment (T. Yoshimoto, unpublished observation). Mice receiving both IL- 12 and IL18 have the highest expression of IFNy and no expression of IL-4 mRNA in lymphocytes of regonal lymph nodes. Furthermore, these lymphocytes produced large amounts of IFNy in response to Con A stimulation.
B MALARIA Malaria is a inosyuite-borne infection caused by protozoa of the genus Plasmodium. Despite extensive efforts to eradicate this disease, malaria remains one of the most prevalent infectious diseases in tropical and subtropical regions (Nussennveig and Nussenzweig, 1989). Infection is initiated when sporozoites are infected into the host by the bite of infected
300
HARUKI OKAMURA et al.
mosquitoes (Anopheles).The sporozoites enter the bloodstream, are rapidly phagocyted by Kupffer cells, and are then passed into hepatocytes, where sporozoites develop into merozoites by asexual division (Meis et aE., 1983). One or 2 weeks after infection, the hepatocytes burst, resulting in releasing thousands of merozoites, thereby initiating the erythrocyte stage of the life cycles (Hollingdale, 1983). The type of T-cell response varies with the stage of infection (Schofield et al., 1987; Podoba and Stevenson, 1991). CD4' T cells mediate immunity against blood stage, whereas CD8+ T cells protect against the liver stage (Schofield et al., 1987; Hoffman et al., 1989; Mellouk et al., 1991; TaylorRobinson and Philips, 1993). The immune response to extraerythrocytes stages of infection has been studied extensively using the muiine model system. The protective effects of CD8+T cells may be mediated by a direct lysis of infected hepatocytes or indirectly by activation of macrophages to produce nitric oxide and TNFa (Stevenson et al., 1995; Doolan et al., 1996) that kills parasites. The resistance of P. berghei sporozoite-infected mice to challenge with infection is abrogated by treatment with anti-IFNy (Schofield et al., 1987; Mellouk et al., 1991), indicating that T-cell-mediated immunity plays a critical role in host defense against infection with Plasmodiurn. Mice can be infected by inoculation of the erythrocyte stage of this protozoa into normal mice. Merozoites develop sequentially into ring forms, trophozoites and schizonts. The erythrocyte cycle continues when schizont-infected erythrocytes burst and release merozoits that invade erythrocytes in inoculated mice. The authors examined the liver of inoculated mice and found that they suffered from severe liver injury. Histological investigation revealed massive cell necrosis in the liver. Hepatocytes in these necrotic areas were stained with the tunnel method, indicating that they died from cell apoptosis (S. Kashiwamura, unpublished observation). IL-18 is originally found in Kupffer cells from inice sequentially treated with P. acnes and LPS (Okamura et al., 1995a). Much higher levels of IL-18 were found in P. berghei-inoculated mice. To determine whether this IL-18 plays a role in defending against this infection, anti-IL-18 was administered. This treatment clearly shortened the survival times of these infected mice (S. Kashiwamura, unpublished observation). To substantiate this observation further, IL-12 and/or IL-18 was injected into mice inoculated with P. berghei. In the case of IL-12, it has already been reported that injection of IL-12 in sporozoite-infected mice prevented them from suffering from parasitemia (Sedegah et al., 1994). Although infected erythrocytes were administered, the authors found that a combination of low doses of IL-12 used to minimize its toxicity and relatively high doses of IL-18 markedly
INTERLEUKIN-18
30 1
diminished mortality of mice inoculated with P. berghei (S. Kashiwamura, unpublished observation). VIII. Pathological Roles of 11-18
A. ENDOTOXIN-INDUCED LIVERINJURY The improvement of nutrition and hygiene and the availability of the appropriate medical care have resulted in the drastic reduction of bacterial infectious diseases. Lethal septic syndrome, however, still remain a threat, partly because the bacterial pathogens have adapted to these situations by developing new pathogenic weapons and partly because intensive medical treatments often suppress the immune system of the patients, who become susceptible to opportunistic infections. Endotoxin lipopolysaccharide from gram-negative bacteria is one of the causal molecules of septic syndrome. Septic syndrome is a systemic inflammatory responsive disease mediated by multiple cytokines, such as IFNy, IL-1, and TNF-a (Dinarello et al., 1993).Acute liver injury was accompanied frequently with septic syndrome (Ghosh et al., 1993), but the how liver injury occurs in the endotoxin syndrome is still unclear. Because the rodent is generally resistant to LPS, many experimental rodent models representing human endotoxin syndrome have been established (Vogel et al., 1980). Heat-killed P. acnes is one of the sensitizing substances to LPS (Tsutsui et al., 1992; Yoshimoto et al., 1992). When a sinall amount of LPS is administered to mice that have been injected with P. acnes 7 days previously, most of the mice died of lethal shock, and the mice that sunrived subsequently suffered from acute liver injury. The authors have investigated the mechanism of how P. acnes treatment sensitizes mice to LPS and how this liver injury occurs. P. acnes treatment reduced the hepatic CD4'NKl.l' T-cell population (Matsui et al., 1997). The liver has Kupffer cells, tissue macrophages, hepatic NK cells, originally termed pit cells, and resident T cells, which consist of CD4+NK1.1-, CD4+NK1.1+,and CD4-NK1.1' T cells (Matsui et al., 1997). Under normal conditions, CD4+NK1.1+T cells promptly produce IL-4 but no IFNy after the appropriate stimulation, which is observed on splenic CD4+NK1.1+T cells, one of Th2 driving cell populations (Yoshimoto, 1995; Emoto, 1995). In contrast, the other two populations, CD,+NKl.l- and CD,-NKl.l+, produced little or no IL-4 but a certain amount of IFNy. P. acnes treatment decreased the proportion of hepatic CD4+NK1.1+T cells, resulting in no IL-4 production by hepatic T cells. The reduction of this population by P. acnes was inhibited mainly by the administration of the anti-IL-12 antibody and partly by the antiIL-18 antibody (Matsui et al., 1997). In fact, Kupffer cells from P. acnes-
302
HARUKI OKAMURA et nl.
primed mice spontaneously produced a small amount of IL-18 and IL-12. One of the actions of P. acnes is, at least partly, attributabIe to IL-18, as well as IL-12 induced in the liver (Matsui et al., 1997; Okainura et al., 1995a). P. acnes treatment also drives hepatic CD4-T cells to produce a large amoung of IFNy, which is inhibitable by the anti-IL-12 antibody. Taken together, the treatment with P. acnes polarizes the liver T cells to type 1through the induction of IL-12 and IL-18 production from Kupffer cells (Matsui et al., 1997). Another aspect of P. acnes treatment is reported to increase the sensitivity to attack by FasL (Tanaka et al., 1997). Fas, one of the receptors of apoptosis-inducing molecules, is constitutively expressed on many tissues, including the liver (Ogasawara et al., 1993). The systemic injection of the agonistic anti-Fas antibody to nontreated mice leads to massive liver injury (Ogasawaraet al., 1993), whereas FasL in soluble form does not cause the liver disorder. P. acnes-primed mice, however, are much more sensitive to FasL (Tanaka et al., 1997).This may be due to Fas-enhancing cytokines such as IFNy and TNF-a, which may be induced by P. acnes treatment (Tanaka et al., 1997). The hepatotoxic cytokine cascade is evoked after an LPS challenge (Tsutsui et al., 1997). Cytokine mRNA expression in the liver was determined by RT-PCR. After the LPS challenge, IL-12 mRNA was first enhanced and then IFNy was induced. IL-18 was expressed in the mRNA level at the constant level before and after the LPS challenge. Interestingly, the TNF-a expression curve has two peaks, the first of which was direct reaction to LPS and the second of which was induced. The administration of the anti-IL-18 antibody at the time of LPS challenge abolished the induction of IFNy- and the second TNF-a-mRNA expressions in the liver. The administration of the anti-IFNy antibody reduced the second TNFa-mRNA enhancement, but &d not affect IL-18 or IL-12 expressions. Thus, IL-18 and IL-12 promptly produced by LPS challenge seem to induce IFNy, and this IFNy enhances TNF-a production from Kupffer cells. In addition, the administration of antibody against IL-18, IFNy, or TNF-a protected the mice from acute liver injury induced by LPS challenge. Thus, IL-18 evoked an inflammatory cytokine reaction with the production of IFNy and then TNF-a in the liver, followed by developing acute liver injury. LPS challenge also induced FasL in the liver. IL-18 has functional FasLaugmenting activity on cloned hepatic NK cells (Tsutsui et al., 1996), suggesting that FasL may be involved in this liver injury. In fact, after a LPS challenge, FasL was expressed in the liver later after the activation of cytokine cascade. This was also inhibitable by treatment with the antiIL-18 antibody, indicating that IL-18 accounts for both TNF-a- and FasLmediated hepatotoxic pathways in endotoxin-induced liver injury in mice
INTERLEUKIN-18
303
(Tsutsui et al., 1997). Preliminary data reveal that IL-18-deficient mice are resistant to liver injury induced by the sequential administration of P. acnes and LPS, although they are sensitive to systemic lethal shock induced by the same treatment (Takeda, unpublished observation), suggesting that the mechanism affecting systemic shock may differ froin that by which liver injury occurs.
B FAILURES OF MULTIPLEORGANS I N MICE Several lines of evidence suggest that IFNy is critically involved in the initiation and progression of autoimmune diseases, such as insulindependent diabetes inellitus (IDDM). The modulation of endogenous IFNy expression correlates well with the onset of IDDM in nonobese diabetes (NOD) mice (Shehedeh et nl., 1993). It has also been shown that IL-12 is expressed prior to the onset of the disease and that injection of IL-12 accelerates the development of the disease (Trembleau et nl., 1995; Kothe et al., 1996). Kothe et al. (1997) examined whether IL-18 mKNA is also upregulated in correlation with the onset of IDDM in NOD mice. They showed that treatment of NOD mice with cyclophosphamide, an accelerating and synchronizing agent of the disease onset, rapidly augmented the expression of IL-18 in the pancreas in advance of IFNy mRNA expression. It was of interest that this treatment with cyclophosphamide did not induce IL-18 expression in normal, nondiabeted mice. At present, it is uncertain why only NOD mice are sensitive to this treatment. As mentioned earlier, IL-12 induces the expression of IL-18R on T cells and IL-18 acts as a cofactor for IFNy production and FasL expression by T or NK cells (Dao et al., 1996; Ahn et al., 1997; Tsutsui et al., 1996). Thus, dysregulated expression of both IL-12 and IL-18 may be crucially involved in the initiation and progression of IDDM in NOD mice. The administration of IL-18 not only failed to accelerate the onset of IDDM in these mice, but also interfered in the accelerating action of IL-12. Thus, the resultant effect of IL-18 may be dependent on the stage of the hsease or the level of IL-12 produced. As described earlier, IL-18 plays crucial roles in inducing liver injury in an animal model. In addition to this liver injury-inducing action, administration of IL-18 and a sinall amount of IL-12 cause serious tissue injuries in a variety of organs (Chikano, unpublished observation). When either IL18 (1pg/day) or IL-12 (0.1 pg/day) is administered daily to normal BALB/ C mice, no pathological changes occur. However, when a mixture of IL18 and IL-12 is injected, the authors observed remarkable pathological changes or manifestations, such as severe diarrhea, rapid weight loss, and a marked atrophy in the thymus. In such treated mice, blood glucose is
304
HARUKI OKAMURA et
d.
transiently elevated at first and then lowered to less than half of normal levels. Histological analysis revealed the presence of hemorrhagic erosion in the intestine and colon, notable fatty degeneration in the liver, and necrotic change in the secretary gland of the pancreas. Marked thymus atrophy was observed in mice treated with a mixture of IL-12 and IL-18. The cortex zone became thin, the border between the medulla and the cortex became obscure, the population of CD4+CD8+(double positive) T cells was eliminated, and many apototic cells were found. Similar changes were observed in IprApr mice deficient in functional Fas after treatment with IL-12 and IL-18, suggesting that Fas-independent tissue injury may be principally responsible for inducing these changes. Further studies are needed to clarifythe underlying mechanisms how and why IL-18 is involved as a causative agent in various disorders with unknown etiology. IX. Perspective
IL-18, originally designated as IGIF, is a multifunctional cytokine, the roles of which extend over both acquired and innate immunity. Its major role has been thought to be the strong ability to induce IFNy production from T or NK cells (Okamura et al., 1995a; Ushio et al., 1996). However, pleiotropy and redundancy, the common characteristics of many cytokines are also applicable to IL-18. It also induces GM-CSF and IL-6, augments NK activity, and stimulates the expression of FasL (Ushio et al., 1996; Tsutsui et al., 1996). As the supply of recombinant IL-18 spreads, other functions will be added to these actions. Moreover, although IL-18-producing cells were thought to be activated macrophages at first (Okamura et al., 1995a), recent investigations reveal that various types of cells, including keratinocytes (Stoll et al., 1997),osteoclasts (Udagawa et al., 1997), and adrenal cortex cells (Conti et al., 1997), express IL-18. This suggests that IL-18 plays important physiological roles in fields other than the immune system. From the time of its discovery, it was noticed that IL-18 has similarities to IL-1 in its amino acid sequence (Okamura et al., 1995a). Both cytokines have an unusual leader sequence that is not homologous to that of ordinary secretory cytokines. The amino acid sequences of both IL-lP and IL-18 have common sequences in the cleavage site of precursor molecule that are recognizable by caspase-1 (Okamura et al., 1995a). In fact, IL-18 has been proven to be processed by caspase 1 and released as an active form (Gu et al., 1997; Ghayur et al., 1997). This indicates that the regulation of this protease also determines the production of mature IL-18, suggesting the possibility of new strategies for the development of anti-inflammatory
INTERLEUKIN-18
305
reagents. Much has to be done to fully understand the pathophysiological roles of IL-18. Preliminary experiments indicate that IL-18 is involved in the defense mechanisms against infection and in tumor rejection. However, unregulated production of bioactive IL-18 seems to be pathogenic, because IL-18 is involved in the tissue injury accompanying inflammatory reactions. To the contrary, IL-18 is involved in the suppressions of allergy via suppression of IgE production (Yoshimotoet al., 1997).Thus, it can be considered that IL-18 is involved in the various diseases in diverse manners. IL-18 augments the production of IFNy in T or NK cells. It is well confirmed that IL-12 and IL-18 exert a marked synergism on IFNy induction (Okamura et al., 1995a; Yoshirnoto et al., 1997; Matsui et al., 1997; Ahn et al., 1997). Furthermore, IL-18 accelerates the development of T h l cells, although IL-18 does not induce T h l cells directly. More investigation is needed to answer the following questions. How is IL-18 production regulated? How is its processing regulated? What roles does IL-18 play in the defense against pathogens? How is IL-18 involved in the pathogenesis of various diseases? How do the signals through IL18R transmit into the nuclei? Nevertheless, IL-18 promises us the fruitful scientific tools to build a bridge between basic and clinical sciences. ACKNOWLEDGMENTS These studies are supported by a grant-in-aid for scientific research on priority areas (351,366), a Hitech Research Center grant from the Ministry of Education, Science, and Culture of Japan, and by CREST (Core Research for Evolution4 Science and Technology) of Japan Science and Technology Corporation. The authors thank Drs. William E. Paul and Tadamitsu Kishimoto for their helpful discussions during the writing of this review. We also acknowledge the suggestions and help of the following persons: Drs. K. Matsui, N. T. Komatsu, S . Shinka, K. Higashino, T. Tamura, S. Akira (Hyogo College of Medicine), H. Fujiwara, T. Harnaoka (Osaka University), M. Kurimoto, S. Fukuda, T. Tanimoto, K. Toigoe, and M. Ikeda (Hayashibara Biochemical Laboratories).
REFERENCES Adachi, M., Suematsu, S., Suda, T., Watanabe, D., Fukuyama, H., Ogasawara, J,, Tanaka, T., Yoshida, N., and Nagata, S. (1996). Enhanced and accelerated lymphoproliferations in Fas-null mice, Proc. Natl. Acad. Sci. USA 93, 2131-2136. 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-y-inducing factor in enhanced production of IFN-y. J. Zniniunol. Alnemri, E. S., Livingston, D. J., Nicholson. D. W., Salvesen, G., Thornberry, N. A., Wong, W. W., and Yuan, J. (1996). Hurnan ICE/CED-3 protease nomenclature. Cell 87, 171. Baeuerle, P. A,, and Henkel, T. (1994). Function and activation of NF-KB in the immune system. Annu. Reu. bnmunol. 12, 141-179. Bazan, J. F., Timans, J. C., and Kastelein, R. A. (1996). A newly defined interleuhn-l? Nature 379, 591.
306
HARUKI OKAMURA et nl.
Belosevic, M., Finbloom, D. S., and van der Meide, P. H. (1989). Administration of monoclonal anti-IFN y antibodies in uiuo abrogates natural resistance of C3H/HeN mice to infection with Leishinanin mujur. J . Exp. Med. 177, 1505-1509. Billiau,A. (1996).Interferon-y: Biology and role in pathogenesis. Adu. Inwnunol. 62,Gl-130. Boehm, U., Klamp, T., Groot, M., and Howard, J. C. (1997).Cellular responses to interferony. Annu. Rer;. hmniunol. 15, 749-795. Cao, Z., Xiong, J., Takeuchi, M., Kurama, T., and Goeddel, D. V. (1996). TRAF6 is a signal transducer for Interleukin-1. Nature 383, 443-446. Coffman, R. L., Ohara, J., Bond, M. W., Carty, J., Zlotnik, A., and Paul, W. E. (1986). B cell stimulatory factor-1 enhances the IgE response of lipopolysaccharide-activated B cells. J. Inimunol. 136, 4538-4541. Conti, B., Jding, J., Tinti, C., Son, J., and Joh, T. (1997).Induction of interferion-y inducing factor in the adrenal cortex. J. B i d . Chern. 264, 2035-2037. Dao, T., Ohashi, K., and Kayano, T. (1996). Inteferon-y-inducing factor, a novel cytokine, enhances fas ligand mediated cytotoxity of inurine T helper 1 cells. Cell. Inmzunol. 173,230-235. Dinarello, C. (1989). Interleukin-1 and its biologically related cytokines. Adv. Imniunol. 44, 153-205. Dinarello, C. A,, Gelfand, J. A,, and Wolff, S. M. (1993). Anticytokine strategies in the treatment of the systemic inflammatory response syndrome. JAMA 269, 1829-1835. Doolan, D., Sedegah, M., Medstrom, R., Hobart, P., Charoenvit, Y., and Hoffman, S. (1996). Circumventing genetic restriction of protection against malaria with inultigene DNA immunization: CD8+T cell-, interferon y-, and nitric oxide-dependent immunity. J. Exp. Med. 183, 1739-1746. liver Emoto, M., Emoto, Y., and Kaufmann, S. H. (1995). IL-4 producing CD4+ TCR (YV lymphocytes:Influence of thymus, /32-mifroglobulinand NKI.1 expression. Int. Irriiriuitol. 7, 1729-1739. Farrar, M., and Schreiber, R. (1993). The molecular cell biology of interferon-g and its receptor. Annu. Reu. biin~unol.11, 571-611. Finkelman, F. D., Katona, I. M., Urban, J. I., Snapper, C. M., Ohara, J., and Paul, W. E. (1986). Suppression of in vivo polyclonal IgE responses by monoclonal antibody to the lymphokine B-cell stiinulatory Factor 1. Proc. Natl. Acnd. Sci. USA 83, 9675-9678. Finkelman, F. D., Holmes, J., Katona, I. M., Urban, J. J., Beckmann, M. P., Park, L. S., Schooley, K. A., Coffman, R. L., Mosmann, T. R., and Paul, W. E. (1990). Lymphokine control of in vivo immunoglobulin isotype selection. Arinu. Reu. 1??mlZtnOl.8, 303-333. Finkelman, F. D., Madden, K. B., Cheever, A. W., Katona, I. M., Morris, S. C., Gately, M. K., Hubbard, B. R., Cause, W. C., and Urban, J. F. J. (1994). Effects of interleukin 12 on immune responses and host protection in mice infected with intestinal nematode parasites. J. Exp. Med. 179, 1563-1572. Gerard, C., Bruyns, C., Marchant, A., Abramowicz, D., Vanderabeele, P., Delvaux, A,, Fiers, W., Goldmann, M., and Veln, T. (1993). Interleukin 10 reduces the release of tumor necrosis factor and prevents lethality in experimental endotoxemia. J. Exp. Med. 177, 547-550. Ghayur, T., Banerjee, S., Hugunin, M., Butler, D., Herzog, L., Carter, A,, Quinta, L., Sekut, L., Tdanian, R., Paskind, M., Wong, W., Kamen, R., Tracey, D., and Allen, H. (1997). Caspase-1 processed IFN-y-inducing factor and regulates LPS-induced IFN-y production. Nnture 386, 619-623. Ghosh, S., Latimer, R. D., Gray, B. M., Hanvood, R. J., and Oduro, A. (1993). Endotoxininduced organ injury. Crit. Care Med. 21, S19-S23.
INTERLEUKIN-18
307
Gu, Y., Kuida, K., Tsutsui, H., Ku, G.. Hsiao, K., Fleming, M., Hayashi, N., Higashino, K., Okamura, H., Nakanishi, K., Kurimoto, M., Tanimoto, T., Flavell, R.. Sato, V., Harding, M., Livingston. D.. and Su, M. (1997). Activation of interferon-gamma inducing factor mediated by interleukin-lbeta converting enzyme. Science 275, 206-209. Hall, S. S. (1995). IL-12 at the crossroads. Science 268, 1432-1434. Heinzel. F. P., Sadick, M. D., Mutha, S., and Locksley, R. M. (1991). Production of interferon-y, IL-2. IL-4,and IL-I0 by CD4' lymphocytes in vivo during healing and progression of inurine leishmaniasis. Proc. Natl. Acad Sci. USA 88, 7011-7015. Heinzel, F. P., Schoenhaut, D. S., Rerko, R. M., Rosser, L. E., and Gately, M. K. (1993). Recoinbinant interleukin 12 cnrcs inice infected with Leishrtuznia mnjor. I. Exyi. Merl. 177, 1505-1509. Heinzel, F. P., Rerko, R. M., and Pe;ulinan, E. (1995). Endogenous IL-12 is required for control of Th2 cytokine responses capable of exacerbating leishinaniasis in normally resistant mice. J. Irnmunol, 155, 7.30-739. Henkart, P. A. (1996).ICE family proteases: Mediators of all apoptotic cell death? Immunity 4, 195-201. Hoffman, S., Isenbarger,D., Long, C;., Sedegah, M., Szarfman, A,, Waters, L., Hollingdale, M., van der Meide, P., Finbloom, D.. and Bdlou, W. (1989). Sporozoite vaccine induces genetically restricted T cell elimination of nidaria from hepatocjes. Science 244, 10781081. Hollingdale, M. R. (1983). Malaria and the liver. Hepatology 5, 327-335. Howard, M., Nakanishi, K., and P a d . W. E. (1984). B cell growth and differentiation factors. Inr9nunol. Reu. 78, 185-210. 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. Hsu, H., Shu, H. B., Pan, M. G., and Goeddel, D. V. (1995).TRADD-TRAF 2 and TRADDFADD interactions define two distinct TNF receptor 1 signal transduction pathways. Cell 84, 299-308. Kataoka, T., Shinohara, N., Takayama, H., Takakn. K., Kondo, S., Yonehara, S., and Nagai, K. (1996).Conanamycin A, a powerful tool for characterization and estimation ofcontribution of perforin-and Fas-based lytic pathways in cell-mediated cytotoxicity.J. Inimunol. 156,3678-3686. Kishimoto, T. (1985). Factors affecting B cell growth and differentiation. Annu. Reu. Immunol. 3, 133-157. Kobayashi, M., Fitz, L.. Ryan. M., Hewick, R. M., Clark, S. C., Chan, S., Loudon, R., Sherman, F., Perussia, B., and Trinchieri, G. (1989). Identification and purification of natural killer cell stiinulatory factor (NKSF). a cytokine with multiple biologic effects on human Iyinpliocytes./. Exp. Med 170, 827-845. Kolino, K., Kataoka, J., Ohtsuki, T., Suernoto, Y., Okarnoto, I., Usni, M., Ikeda, M., arid Kurimoto, M. (1997). IFN-y-inducing factor (IGIF) is a costimulatory factor on the activation of Tlil but not Th2 cells and exerts its effect independently of IL-12.1. Immunol. 158, 1541-1550. Kuida, K., Lippke, J. A,, Ku, G., Harding, M. W., Livingston, D. J., Su, M. S.-S., and Flavell, R. A. (1995). Altered cytokine export and apoptosis in inice deficient in interleukin-lp converting enzyme. Science 267, 2000-2003. Kuida, K., Zheng, T. S., Na, S., Knan. C., Yang. D., Karasuyama, H., Rakic, P., and Flavell, R. A. (1996). Decreased apoptosis in the brain and premature lethality in CPP32-deficient mice. Nature 384, 368-373.
308
HARUKI OKAMURA et al.
Kagi, D., Ledermann, B., Biirki, K., Zinkernagel, R. M., and Hengartner, H. (1996).Molecular mechanisms of lymphocyte-mediated cytotoxicity and their role in immunological protection and pathogenesis in vivo. Annu. Reu. Zmmunol. 14, 207-232. Li, L., Young, D., Wolf, S. F., and Choi, Y. S. (1996). Interleukin-12 stimulates B cell growth by inducing IFN-y. Cell. Zinmunol. 168, 133-140. Li, P., Allen, H., Banerjee, S., Franklin, S., Herzog, L., Johnston, C., McDowell, J., Paskind, M., Rodman, L., Salfeld, J.. Towne, E., Tracyey, D., Wardwell, S., F.-Y, W., Wong, W., Kamen, R., and Seshadri, T. (1995). Mice deficient in IL-lp-converting enzyme are defective in production of mature IL-1/3 and resistant to endotoxin shock. Cell 80, 401-411. Lindstein, T., June, C., Ledbetter, J., Stella, G., and Thompson, C. (1989). Regulation of lymphokine messenger RNA stability by a surface mediated T cell activation pathway. Science 244, 399-343. Matsui, K., Yoshimoto, T., Tsutsui, H., Hyodo, Y., Hayashi, N., Hiroishi, K., Kawada, N., Okamura, H., Nakanishi, K., and Higashino, K. (1997).Propionibacteriumacnes treatment diminished CD4+NK1.1+T cells but induces type 1 T cells in the liver by induction of IL-12 and IL-18 production from Kupffer cells. J. Zmmunol. 159, 97-106. Matsumoto, S., Tsuji-Takayama, K., Aizawa, Y., Koide, K., Takeuchi, M., Ohta, T., and Kurimoto, M. (1997). Interleukin-18 activates NF-KB in murine T helper type 1 cells. Biochem. Biophys. Rex Commun. 234, 454-457. Meis, J. E. G. M., Verhave, J. P., Jap, P. H. K., and Meuwissen, J. H. E. (1983). An ultrastructual study on the role of Kuppfer cells in the process of infection by PlmmorEium berghei in rats. Parasitology 86, 231-242. Mellouk S., Green, S. J., Nacy, C. A,, and Hoffman, S. L. (1991). IFN-y inhibits development of Plmnwdium berghei exoerythrocyticstages in liepatocytes by an L-arginine-dependent effector mechanism. J. Immunol. 146, 3971-3976. Micallef, M., Ohtsuki, T., Kohno, K., Tanabe, F., Ushio, S., Namba, M., Tanimoto, T., Torigoe, M., Fujii, M., Ikeda, M., and Kurimoto, M. (1996). IFN-y-inducing factor enhances T helper 1 cytokine production by stimulated human T cells: Synergism with interleukin-12 for interferon-y production. Eur. J. Zmmunol. 26, 1647-1651. Mosmann, T. R., and Coffman, R. L. (1989). TH1 and TH2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu. Reu. Immunol. 7, 145-173. Muroi, M., N. Shiragami, Nagao, K., Yaamasaki, M., and Takatsuki, A. (1993). Folimycin (concanamycin A), a specific inhibitor of V-ATPase, blocks intracellular translocation of the glycoprotein of vesicular stomatitis virus before arrival to the Golgi apparatus. Cell Struct. Funct. 18, 139-149. Nagata, S. (1994). Fas and Fas ligand: A death factor and its receptor. Ado. Zmmunol. 57, 129-144. Nagata, S., and Golstein, P. (1995). The Fas death factor. Science 267, 1449-1456. Nakamura, K., Okamura, H., Nagata, K., Komatsu, T., and Tamura, T. (1993). Purification of a factor which provides a costimulatory signal for gamma interferon production. Infect. Zmmun. 61, 64-70. Nakamura, K., Okamura, H., Wada, K., Nagata, K., Komatsu, T., and Tamura, T. (1989). Endotoxin-induced serum factor that stimulates gamma interferon production. Infect. Immun. 38,590-595. Nakanishi, K., Cohen, D. I., Blackman, M., Nielsen, E., Ohara, J., Hamaoka, T., Koshland, M. E., and Paul, W. E. (1984a). Ig RNA expression in normal B cells stimulated with anti-IgM antibody and T cell-derived growth and differentiation factors. J. Exp. Med. 160, 1736-1751.
INTERLEUKIN-18
309
Nakanishi, K., Malek, T., Smith, K., Hamaoka, T., Shevach, E., and Paul, W. (1984b). Both interleukin 2 and a second T cell-derived factor in EL-4 supernatant have activity as differentiation factor in IgM synthesis. J. Exp. Med. 160, 1605-1621. Nakanishi, K., Yoshimoto, T., Chu, C . C., Matsumoto, H., Hase, K., Nagai, N., Tanaka, T., Miyasaka, M., Paul, W. E., and Shinka, S. (1995). IL-2 inhibits IL-4-dependent IgE and IgGl production in vitro and in vivo. Int. Immunol. 7, 259-268. Nussennveig, V., and Nussenzweig, R. (1989). Rationale for the development of an engineered oporozoite malaria vaccine. Adv. Iniiiaunol. 45, 283-334. Ogasawara, I., Watanabe-Fukunaga, R., Adachi, M., Matsuzawa, A., Kasuga, T., Kitaniura, Y., and Nagata, S. (1993).Lethal effect of the anti-Fas antibody in mice. Nature 364,806-809. Okamura, H., Kawaguchi, K., Shoji, K., and Kawade, Y. (1982). High-level induction of gamma interferon with various mitogens in mice pretreated with Propionibacteriumacnes. Infect. Immun. 38, 440-443. Okamura, H., Tsutsi, H., Komatsu, T., Yutsudo, M., Hakura, A,, Tanimoto, T., Torigoe, K., Okura, T., Nakuda, Y., Hattori, K., Akita, K., Namba, M., Tanabe, F., Konishi, K., Fykuda, S., and Kurimoto, M. (1995a). Cloning of a new cytokine that induces IFN-7 production by T cells. Nature 378, 88-91. Okamura, H., Nagata, K., Komatsu, T., Tanimoto, T., Nukata, Y., Tanabe, F., Akita, K., Torigoe, K., Okura, T., Fukuda, S., and Kurimoto, M. (1995b). A novel costimulatory factor for gamma interferon induction found in the liver of mice causes endotodc shock. Infect. Immun. 63, 3966-3972. Parnet, P., Garka, K. E., Bonnet, T. P., Dower, S. K., and Sim, J. E. (1996). IL-1Rrp is a novel receptor-like molecule similar to the type I interleuldn-1 receptor andits homologues Tl/STS and IL-1R AcP. J. Biol. Chem 271, 3967-3970. Pistoia, V., Cozzolino, F., Rubartelli, A,, Torcia, M., Roncella, S., and Ferrarini, M. (1986). In vitro production of interleukin 1 by normal and malignant human B lymphocytes. J. Immimol. 136, 1688-1692. Podoba, J. E., and Stevenson, M. M. (1991). CD4+ and CD8' T cells both contribute to acquired immunity to blood-stage Plasmodium chabaudi AS. Infect. Immun. 59,51-58. Reiner, S. L., and Locksley, R. M. (1995).The regulation of immunity to Leishmania major. Annu. Rev. Imnmunol. 13, 151-177. Rothe, H., Jenkins, N., Copeland, N., and Kolb, H. (1997). Active stage of autoimmune diabetes is associated with the expression of a novel cytokine, IGIF, which is located near Id&. J. Clin. Inve.Tt. 99, 469-474. Rothe, M., Burkart, V., Faust, A., and Kolb, H. (1996). Interleukin-12 gene expression is associatedwith development of diabetes mellitus in non-obese diabetes mice. Diabetotogia 39, 119-122. Sadick, M. D., Heinzel, F. P., Holaday, B. J., Pn, R. T., Dawkins, R. S., and Locksley, R. M. (1990). Cure of murine leishrnaniasis with anti-interleukin 4 monoclonal antibody: Evidence for a T cell-dependent, interferon gamma-independent mechanism. J. Exp. Med. 171, 115-127. Schofield, L., Villaquiran, J., Ferreira, A,, Schellekens, H., and Nussenzweig, R. (1987). Ganima interferon, CD8' T cells and antibodies are required for immunity to malaria sporozoites. Nature 330, 664-666. Sedegah, M. S., Finkelman, F., and Hoffman, S. L. (1994). Interleukin 12 induction of interferon y-dependent protection against malaria. Proc. Nut!. Acad. Sci. USA 91,1070010702. Seder, R. A,, Gazzinelli, R., Slier, A,, and Paul, W. E. (1993). Interleukin 12 acts directly on CD4+ T cells to enhance priming for interferon gamma production and &minishes interleukin 4 inhibition of such priming. Proc. Natl. Acad. Sci. USA 90, 10188-10192.
310
IIARUKI OKAMURA et ol
Seder, R. A,, and Paul, W. E. (1994). Acqu on of lymphokine-producing phenotype by CD4+ T cells. Annu. Rev. Immttnol. 12, 635-673. Seder, R. A,, Paul, W. E., Davis, M. M., Fazekas, D. S., and 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.]. Exp. Med. 176, 10911098. Shehedeh, N. N., LaRosa, F., and Lafferty, K. J. (1993).Altered cytokine activity in adjuvant inhibition of autoimmune diabetes. J Autoimmune. 6, 291-300. Sher, A,, and Coffman, R. (1992). Regulation of immunity to parasites by T cells and Tcell-derived cytokines. Annu. Rev. Immunol. 10, 385-409. Singer, I. I., Scott, S., Chin, J,, Bayne, E. K., Limjuco, G., Widner, J.. Miller, D. K., Chapmann, K., and Kostura, M. J. (1995). The interleukin-lp-converting enzyme (ICE) is localized on the external cell surface membranes and in the cytoplasmic ground substance of human monocytes by immuno-electron microscopy. J. Exp. Med. 182, 1447-1459. Smeland, E. B., Blomhoff, H. K., Funderud, S., Shalaby, M. R., and Espevik, T. (1989). Interleukin 4 induces selective production of interleukin 6 from normal human B lymphocytes.]. Exp. Med. 170, 1463-1468. Smith, C. A,, Williams, G. T., Kinston, R., Jenkinson, E. J., and Owen J. J. (1989).Antibodies to CD3fl-cell receptor complex induce death by apoptosis in immature T cells in thymic cultures. Nature 337, 181-184. Snapper, C. M., and Paul, W. E. (1987). Interferon-y and B cell stimulatory factor-1 reciprocally regulate Ig isotype production. Science 236, 944-947. Stevenson, M. M., Tam, M. F., Wolf, S. F., and Sher, A. (1995). IL-12-induced protection against blood-stage Plasmodium chabnudi AS requires IFN-y and TNF-a and occurs via a nitric oxide-dependent mechanism. J. Immnunol. 155, 2545-2556. Stoll, S., Mnller, G., Kurimoto, M., Saloga, J., Tanimoto, T., Yamauchi, H., Okamura, H., Knop, J., and Enk A. H. (1997). Production of IL-18 (IFN-y-inducing factor) messenger RNA and functional protein by rnurine keratinocytes. J Immunol. 159, 298-302. Swain, S. L., Bradley, L. M., Croft, M., Tonkowagy, S., Atkins, G., Weinberg, A. D., Duncan, D. D., Hedrick, S. M., Dutton, R. W., and Huston, G. (1991). Helper T-cell subsets: Phenotype, function and the role of lymphokines in regulating their development. Iinmunol. Rev. 123, 115-144. Swain S. L., Weinberg, A. D., English, M., and Huston, G. (1990). IL-4 directs the development of Th2-like helper effectors. ]. Immunol. 145, 3796-3806. Szabo, S. J., Dighe, A. S., Gubler, U., and Murphy, K. M. (1997). Regulation of the interleukin (IL)-12R p2 subunit expression in developing T helper (Thl) and Th2 cells. 1.Exp. Med. 185, 817-824. 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. Takahashi, T., Tanaka, M., Brannan, C. I., Jenkins, N. A,, Copeland, N. G., Suda, T., and Nagata S. (1994). Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76, 969-976. Tanaka, M., Suda, T., Yatomi, T., Nakamura, N., and Nagata S. (1997). Lethal effect of recombinant human Fas ligand in mice pretreated with Propionibacterium acnes. 1.Immunol. 158, 2302-2309. Tanaka, Y., Takahashi, A,, Watanabe, K., Takayama, K., Yahata, T., Habu, S., and Nishimura T. (1996). A pivotal role off IL-12 in Th-1 dependent mouse liver injury. Int. Immunol. 8, 569-576.
INTERLEUKIN-18
311
Taylor-Robinson, A. W., and Philips, R. S. (1993). Protective CD4' T cell lines raised against Plnsmo(liurn chaboudi show characteristics of either Thl or Th2 cells. Parmite Inmunol. 15, 30 1-3 10. Thornberry, N. A., Bull, H. G., Calaycay, J. R., Chapman, K. T., Howard, A. D., Kostura, M. J., Miller, D. K., Molineaux, S. M., Weidner, J. R., Aunins, J., Elliston, K. O., Ayala, J. M., Casano, F. J., Chin, J., Ding, G. F.-F., Egger, L. A,, Gaffney, E. P., Limjuco, G., Palyha, 0. D., Raju, S. M., Rolando, A. M., Salley, J. P., Yamin, T. T., Lee, T. D., Shively, J. E., MacCross, M., Mumford, R. A,, Schmidt, J. A,, and Tocci, M. J. (1992). A novel heterodimeric cysteine protease is required for interleukin-lp processing in monocytes. Nature 356, 768-774. Torigoe, K., Ushio, S., Okura, T., Kobayashi, S., Taniai, M., Kinikata, T., Murakami, T., Sanou, O., Kojinia, H., Fuji, M., Ohta, T., Ikeda, M., Ikegami, H., and Kurimoto, M. (1997). Functional and characterization of the human Interleukin-18 (hIL-18) receptor. J . Biol. Chem. 272, 25737-25742. Trembleau, S., Penna, G., Bosi, E., Mortara, A,, Gately M., and Adorini, L. (1995). Interleukin-12 administration induces T helper type 1 cells and accelerates autoimmune diabetes in NOD mice. J . Exp. Med. 181, 817-821. Trinchieri, G. (1995). Interleukin-12: A proinflammatory cytokine with iminunoregulatory functions that bridge innate resistance and antigen-specificand adaptive immunity. Annu. Reu. Znimnunol. 13, 251-276. Tsutsui, H., Matsui, K., Kawada, N., Hyodo, Y., Hayashi, N., Okamura, H., Higashino, K., and Nakanishi, K. (1997).IL-18 accounts for both TNF-a- and FasL-mediated hepatotoxic pathways in endotoxin-induced liver injury. J . Zmmunol. 159, 3961-3967. Tsutsui, H., Mizoguchi, Y., and Morisawa S. (1992). Importance of direct hepatocytolysis by liver macrophages in experimental fulminant hepatitis. Hepato-Gnstroenterol. 39, 553-559. Tsutsui, H., Nakanishi, K., Matsui, K.. Higashino, K., Okamura, H., Miyazawa, Y., and Kaneda K. (1996). Interferon-y-inducing factor up-regulates Fas ligand-mediated cytotoxic activity of rnurine natural killer cell clones. J. Zmmnunol.157, 3967-3973. Udagawa, N., Honvood, N. J., Elliott, J., Mackay, A,, Owens, J., Okamura, H., Kurimoto, M., Chambers, T. J., Martin, T. J., and Gillespie, M. T. (1997).Interleukin-18 (interferony-inducing factor) is produced by osteoblasts and acts via granulocyte/inacrophage colonystimnlating Factor and not via interferon-g to inhibit osteoclast formation. J . Exp. Med. 185, 1005-1012. Ushio, S.. Namba, M., Okura, T., Hattori, K., Nukada, Y., Akita, K., Tanabe, F., Konislii, K., Micallef, M., Fujii, M., Torigoe, K., Tanimoto, T., Fukuda, S., Ikeda, M., Okamura, H., and Kurimoto, M. (1996). Cloning of the cDNA for human IFN-y-inducing factor, expression in Escherichia coli, and studies on the biologic activities of the protein. J. Inimunol. 156, 4274-4279. Velnpillai, P., and Ham, D. A. (1994). Oligosaccharide-specificinduction of interleukin 10 production by B220t cells from schistosoma-infected mice: A mechanism for replation of CD4t T-cell subsets. Proc. Nntl. Acnd. Sci. USA 91, 18-22. Vogel, S. N., Moore, R. N., Sipe, J. D., and Rosenstereich, D. L. (1980). BCG-induced enhancement of endotoxin sensitivity in C3H/HeJ mice. I. In vivo studies. J. Zrnmunnl. 124, 2004-2009. Wada, M., Okamura, H., Nagata, K., Shimoyarna, T., and Kawada, Y. (1985). Cellular mechanisms in in vivo production of gamma interferon induced by lipopolysaccharide in mice infected with Mycobacterium boais BCG. J. Zntet@ron Res. 5, 431-443. Walker, N. P. C., Talanian, R. V., Brady, K. D., Dang, L. C., Bump, N. J.. Ferenz, C. R., Franklin, S., Ghayur, T., Nackett, M. C., Hammill, L. D., Herzog, L., Hugunin, M. HOLIY,
312
HARUKI OKAMURA et al.
W. Mankovich, J. A., McGuiness, L., Orienwicz, E., Paskind, M., Pratt, C. A., Reis, P., Summani, A,, Terranova, M., Wich, J. P., Xiong, L., Muler, A., Tracey, D. E., Kamen, R., and Wong W. W. (1994). Crystal structure of the cysteine protese interleukin-lpconverting enzyme: A (p20/p10)2 homodimer. Cell 78, 343-352. Wheelock, E. (1965). Interferon-like virus-inhibitor induced in human leukocytes by phytohemagglutinin. Science 149, 310-311. Wu, J., Zhou, T., Zahng, J., He, J.. Gause, W. C., and Mountz, J. D. (1994). Correction of accelerated autoimmune disease by early replacement of the mutated Ipr gene with the normal Fas apoptosis gene in the T cells of trangenic MRL-lpr/lpr mice. Proc. Natl. Acad. Sci. USA 91, 2344-2348. Yoshimoto, T., Nakanishi, K., Hirose, K., Hiroishi, K., Okamura, H., Takemoto, Y., Kanamaru, A,, Hada, T., Tamura, T., Kakishita, E., and Higashino, K. (1992). High serum IL-6 level reflects susceptible status of the host to endotoxin and I L - l h m o r necrosis factor. J. lmmunol. 148, 3596-3603. Yoshiinoto,T., and Paul, W. E. (1994). CD4P"', N K l . l p 5T cells promptly produce interleukin 4 in response to in vivo challenge with anti-CD3. J. Exp. Med. 179, 1285-1295. Yoshimoto, T., Bendelac, A,, Hu, L. J., and Paul, W. E. (1995a). Defective IgE production by SJL inice is linked to the absence of CD4+, NK1.P T cells that promptly produce interleukin 4. Proc. Natl. Acad. Sci. USA 92, 11931-11934. Yoshimoto,T., Bendelac, A., Watson, C., Hu-Li, J., and Paul, W. E. (1995b).Role of NK1.lt T cells in TH2response and immunoglobulin E production. Science 270, 1845-1847. Yoshimoto, T., Okamura, H., Tagawa, Y., Iwakura, Y., and Nakanishi, K. (1997). Interleukin 18 together with interleukin 12 inhibits IgE production by induction of interferon-y production from activated B cells. Proc. Nutl. Acad. Sci. USA 94, 3948-3953.
This article was accepted for publication on December 15, 1997
ADVANCE5 IN IMMUNOLOGY. VO1. i l l
CD4+ T-cell Induction and Effector Functions: A Comparison of Immunity against Soluble Antigens and Viral Infections ANNETTE OXENIUS, ROLF M. ZINKERNAGEL, AND HANS HENGARTNER Depariment of Paihdogy-, Institute of Experimental Immunology, University of Zurich, 809 1 Zurich, Swifzedand
I. Introduction The central task of the immune system is to defend the host against a wealth of pathogens in the environment. Studying immune responses to infectious agents such as viruses, bacteria, protozoa, and multicellular parasites analyzes immunology with respect to its biological purpose and evolution, which was driven by the coexistence of pathogens and hosts. Initial immunological research focused on infectious diseases, which was followed by a period concentrating on antigen-specific immunity, which was based on chemically defined antigens such as soluble proteins (y-globulin, ovalbumin, lysozyme, etc.) or hapten-carrier conjugates. These studies provided insight into the specificity of immune recognition and in the mechanisms of cellular interactions involved in this process; furthermore, these experimental model systems led to the definition of rules governing antigen-specific activation or tolerization of immune responses based on antigen dose, route of application, and inolecular nature of the antigen. However, generalization on the basis of some of these results may be difficult because the immune responses elicited by soluble protein model antigens are usually not relevant for the survival of the host (with the exception of toxins and vaccines); kinetics and efficiency of such immune responses therefore do not underly the pressure imposed by infectious agents. Nevertheless, based on the conceptual knowledge obtained with purified antigens and thanks to enormous methodological progress in biochemistry, molecular biology, embryology, and cell culture, the key parameters orchestrating an immune response on infection can and must be reevaluated in infectious disease models. Certainly, many immunological aspects may be equivalent for soluble model antigens and infectious agents whereas other aspects are unique to infectious disease models: replication kinetics, cell tropism, and cytopathogenicity of infectious agents influence time-dependent amplification of the antigen load, anatomical localization of the antigen, and pathological aspects. In addition, pathogens efficiently trigger responses of the innate immune system, affecting the generation of specific immune responses. The immune response has to cope with the pathogen-induced dynamics by its own dynamics, coordinat313
Capfight 0 1998 by Academic Prrsn. All rights of reproduction in any forni reserved. 00G-ZTifi/98 925 00
314
ANNETTE OXENIUS et al
ing its effector functions in the most efficient way to finally resolve infection, keep the infection at a minimal level, or even tolerate infection in the case of certain noncytopathic pathogens. Although the effector mechanisms and effector cells involved in protective immune responses against different pathogens are usually manyfold and differ in relevance for each pathogen, this review focuses mainly on the role of CD4’ T cells for antiviral protection. The first section discusses parameters influencing Th cell activation after immunization with soluble model antigens in comparison to viral infections, and the second section addresses the role of virus-specific Th cell effector functions in several selected viral disease models. II. Activation of CD4+ T Cells
A. ANTIGEN ADMINISTRATION, ANTIGEN DOSE, AND
GEOGRAPHICAL LOCATION
1 . Soluble Antigens Soluble antigens able to load major histocompatibility complex (MHC) class I1 molecules can be used to activate antigen-specific CD4’ T cells and are usually either whole proteins requiring processing for the presentation of the relevant antigenic peptides or, alternatively, directly MHC class II-binding antigenic peptides not requiring processing. The routes of administration as well as the dose of soluble antigen have been shown to be crucial parameters influencing the induction of Th cell activation versus Th cell tolerance. The first studies performed in this line demonstrated that very high concentrations of soluble protein administered repeatedly and systemically induced a state of specific immune paralysis (high zone tolerance) that was abrogated after a critical time threshold, probably when newly generated T and B cells emerged from the thymus or the bone marrow and when the antigen had been eliminated from circulation (Mitchison, 1964; Weigle, 1973). This phenomenon is dependent on very high doses of antigen, probably leading to a very dense antigen presentation on MHC class I1 molecules such that virtually every specific Th cell is induced and subsequently repeatedly restimulated, eventually leading to Th cell unresponsiveness and/or deletion. This process of Th cell “overactivation”has also been demonstrated on the clonal level in vitro using Th cell clones or in viva using T-cell-receptor- (TCR-) transgenic animals immunized with high doses of protein antigen or peptide antigen (Critchfield et al., 1994; Kearnery et al., 1994; Liblau et al., 1995).Intermediate doses of soluble antigen applied systemically, leading most probably to a more restricted antigen presentation pattern, favor proper Th cell activation (Mitchison, 1964). The observation that administration of very
CD4* T-CELL INDUCTION AND EFFECTOR FUNCTIONS
315
low doses of soluble antigen induces Th cell unresponsiveness (Mitchison, 1964) might be explained by the poor antigen presentation of the low concentrations of soluble protein, which is most probably restricted to antigen-specific B cells, which are the only antigen-presenting cells (APCs) capable of concentrating antigen via their surface Ig receptors (Lanzavecchia, 1987). Naive B cells express very little costimulatory molecules and are thus believed to be able to induce Th cell unresponsiveness (Eynon and Parker, 1992; Fuchs and Matzinger, 1992; Gilbert and Weigle, 1994). However, tolerance induction by administration of soluble protein antigen could also be observed in B-cell-deficient mice, suggesting that mechanisms other than antigen presentation by B cells might also be involved in antigenspecific tolerance induction (Phillips et al., 1996). Thus, Th cell activation seeins to critically depend on an appropriate dose of the soluble antigen that preferentially targets the antigen to professional APCs and is generally not easily achieved with truly soluble antigens. Protein aggregates have been shown to induce specific Th cell responses, whereas soluble, nonaggregated protein was shown to preferentially induce Th cell tolerance (Weigle, 1973). It is interesting to note that soluble proteins administered systemically may induce tolerance even independently of the antigen dose. Many proteins used for the just-mentioned studies were presumably contaminated with lipopolysaccaride (LPS), which has the ability to activate macrophages and B cells. At very low and at very high concentrations, LPS would either be too diluted or its effect overwhelmed by the protein and Th cells would be tolerized, whereas at intermediate doses, LPS would unspecifically activate macrophages and B cells and thus prevent Th cell tolerance induction. In this scenario, Th cell priming at intermediate protein doses would, in fact, represent a LPS-mediated side effect (W. 0. Weigle, personal communication). A powerful method to improve the antigenicity of soluble proteins or peptides is to administer them in association with adjuvants. Commonly used adjuvants in the murine system are complete Freunds adjuvant (CFA), incomplete Freunds adjuvant (IFA),alum, liposomes, immune-stimulating complexes (ISCOM), and others. A twofold effect is achieved by administering the soluble antigen in adjuvant: (1)The soluble protein is retained within a locally applied antigen depot, releasing the antigen slowly and over a prolonged time period, thus inhibiting rapid consumption and proteolysis of the antigen and enabling a prolonged presentation of the relevant antigen on MHC class I1 inolecules on APCs. (2) The addition of heatkilled inycobacteria (such as in CFA) and LPS induces local inflammation, which leads to nonantigen-medated upregulation of costimulatory molecules on the APC surface enhancing the quality of the APC for activation of naive Th cells.
316
A N N E m E OXENIUS et a1
The localization of the administered antigen depends on the route of application and/or the use of adjuvants retaining the antigen at a certain location; systemic immunization with soluble antigens leads to rapid dissemination over the blood and thereby to dilution of the antigen; in contrast, local application in adjuvants directs the antigen primarily to the local draining lymph node. Systemicallyapplied soluble antigen is most probably only poorly filtered out of the circulation of the spleen by the phagocytosing marginal zone macrophages, which are responsible for trapping particulate antigens; thus, antigen capture is most probably performed by antigenspecific B cells that are able to concentrate antigen via their surface IG receptors. However, a locally administered antigen can be trapped efficiently by immature dendritic cells (DC) (such as the Langerhans cells in the skin) via macropinocytosis or mannose receptor-mediated uptake and is then transported within the DC to the local draining lymph node where it is presented to the pool of recirculating naive Th cells (Sallusto and Lanzavecchia, 1994; Steinmann, 1991). However, a characteristic common to all soluble antigens, independent of their route of administration, is their quantitatively defined concentration, which never increases after application but rather diminishes over time due to proteolysis and consumption. Taken together, soluble antigens such as proteins or peptides are usually poor activators of naive Th cells unless administered together with adjuvants (Warren et al., 1986). In fact, systemically administered soluble antigens usually even lead to the induction of Th cell unresponsiveness.
2. lnfectious Agents The major differences between soluble nonreplicating antigens and infectious replicating agents are (1)the dynamics of antigen concentration due to amplification of the antigen and (2) the antigen localization, which is not only directed by the route of infection but also by the cell tropism of the virus, replication kinetics, and spread within the host. Thus, the antigen dose actually present at a given site of the organism at a given time point after infection is very difficult to measure and thus only little quantitative control can be exerted. The limiting and augmenting factors concerning antigen concentration are manyfold and interrelated: virus replication kinetics, cytopathogenicity leading to the liberation of newly synthesized antigens, or immunopathology caused by immune effector cells all contribute to an increasing antigen load. However, immune mechanisms such as innate immunity and specific T- and B-cell responses reduce the antigen load by clearing virus. In addition, the number of antigen-presenting cells is influenced by viral cytopathogenicity or immu-
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
317
nopathology. Thus, antigen concentrations usually start at a defined concentration of the inoculum, amplify to certain levels, and drop later on, when infection is controlled by the immune system. A possible crucial difference between viral antigens and soluble antigens is that the former are usually always locally very highly concentrated. The local antigen concentrations derived from infectious agents may in fact exceed by far what is usually achieved on immunization with exogenous soluble proteins (Battegay et al., 1996; Oxenius et al., 1997). Depending on the cell tropism and the route of infection, virus-derived antigens first appear in distinct secondary lymphoid organs: systemic application of viruses leads to capture of the virus by the marginal zone macrophages of the spleen; local administration of the virus leads to local replication in suitable host cells and thus finally to the liberation of viral antigens, which can be either trapped by locally resident immature DC or macrophages or transported directly to the draining lymph node via the lymph. Infection via mucosal surfaces usually leads to antigen presentation in Peyers patches or in the tonsils (Mims, 1987; Paul, 1993). As for soluble model antigens, activation of virus-specific naive Th cells occurs in secondary lymphoid organs where epitopes of virus-derived proteins are presented on MHC class I1 molecules. Despite the difficult quantification of antigen load on viral infections, the initial virus dose in the inoculum as well as the replication kinetics of the virus have been shown to influence the functional outcome of T-cell activation. In the case of CD8' T cells it bas been demonstrated, for example, that infection with high doses of widely replicating lymphocytic choriomeningitis virus (LCMV)-Docileleads to exhaustion of virus-specific CTLs and thus to the establishment of a virus carrier status (Moskophidis et al., 1993). Such an exhaustive activation of virus-specific CD4' T cells has not yet been described after overwhelming virus infection and it will be interesting to reveal whether similar rules govern the fate of CD8+ and CD4+ T cells during persistent viral infection.
B. ANTIGENPRESENTATION ON MHC CLASS I1 MOLECULES 1. Soluble Antigens In contrast to MHC class I molecules, which are (or can be) expressed on virtually all nucleated cells within an organism, MHC class I1 molecules are usually only expressed on antigen-presenting cells such as dendritic cells, macrophages, and B cells. APCs are, via different mechanisms, capable of trapping soluble antigens, antigen complexes, or particulate antigens, which are subsequently degraded in specialized intracellular compartments to peptide fragments that are potentially able to load MHC class I1 mole-
318
ANNETTE OXENIUS et al.
cules. The complex of MHC class I1 and peptide antigen then reaches the cell surface where it can be recognized by antigen-specific Th cells (Germain and Margulies, 1993). The different mechanisms of antigen uptake involve (1) phagocytosis of particulate antigen or antigen complexes by macrophages, (2) macropinocytosis or mannose receptor-mediated uptake of antigen by immature dendritic cells, (3) Ig receptor-mediated uptake by antigen-specific B cells, and (4) Fc-receptor-mediated uptake of antibody-antigen complexes mainly by macrophages. In all cases the antigen is trapped after uptake in endosomal vesicles, which are progressively acidified, thus enabling acidic proteases to continously degrade antigen into peptide fragments (Neefjes and Momburg, 1993). Heterodimeric MHC class I1 molecules, consisting of an a and /3 chain, are newly synthesized in the endoplasmic reticulum (ER) where they associate with the invariant chain (Ii) (Cresswell, 1992). The association with Ii is thought to (1)prevent (at least to some extent) loading of newly synthesized MHC class I1 molecules with ER-resident peptides, (2)traffic newly synthesized MHC class I1 molecules to endosoinal compartments due to an Ii-inherent endosomal targeting signal (Bakke and Dobberstein, 1990; Lamb et al., 1991), (3) thermodynamically stabilize the structure of the MHC class I1 cr/3 heterodimer (Sant and Miller, 1994; Stocking et al., 1989). MHC Class I1 molecules are then transported to endosomal compartments where they are loaded primarily, but not exclusively (Castellino and Germain, 1995),with antigenic peptides within the MIIC compartment (Amigorena et al., 1994; Tulp et al., 1994). Having reached the endosomal compartments, the still associated Ii is subjected to proteolysis (predominantly by cathepsin S) (Riese et al., 1996), resulting finally in MHC class I1 crp heterodimers being loaded with a residual 20-25 mer Ii fragment (CLIP, class I1 associated invariant chain peptide) (Ghosh et nl., 1995). The CLIP fragment is exchanged with the antigenic peptides by the catalytic action of the MHC class I1 locus-encoded H2-M molecule (HLA-DM is the corresponding molecule in humans) (Denzin and Cresswell, 1995; Kropshofer et al., 1997; Miyazaki et al., 1996; Sherman et al., 1995). Ultimately, the MHC class 11-peptide complexes are shuttled to the cell surface when they are recognized by CD4' T cells. The functional importance of Ii and H2-M molecules in the just-described MHC cIass I1 loading pathway for soluble antigens has been documented by the virtual absent capacity of Ii- or H2-M-deficient APCs to utilize soluble protein antigens for MHC class I1 presentation in uitro. MHC Class I1 molecules from Ii-deficient mice were shown to be mostly devoid of antigenic peptides (SDS instable) (Miller and Germain, 1986; Viville et al., 1993), whereas MHC class I1 molecules from H2-M-deficient mice were shown to be
C:1>4’ T-CELL INDUCTION AND EFFECTOR FUNCTIOKS
319
almost exclusively loaded with the CLIP peptide ( Fung-Leung et al., 1996; 1997). Miyazaki et al., 1996; Tourne et d., In contrast to soluble antigens, which require antigen processing MHC class I1 binding peptides do not require processing and are thus able to load any MHC class I1 expressing APC, including the previously mentioned Ii- andsH2-M-deficient APCs. MHC class I1 binding peptides either may bind to empty MHC class I1 molecules, which are newly synthesized or are recycling from the cell surface to endosoinal compartments, or may be taken up during the process of endocytosis/macropinocytosis to be loaded onto MHC class I1 molecules in endosomal compartments. 2. Infectious Agents In addition to the just-described “classical” MHC class I1 presentation pathway, viruses may use additional pathways of MHC class I1 loading. One major difference to exogenous, soluble proteins is that intracellular pathogens such as viruses synthesize viral proteins within the infected cell. These intracellularly synthesized proteins are usually the sources of MHC class I binding peptides, but may also, in certain situations, be able to load MHC class I1 molecules. Intracellularly synthesized viral proteins are a priori not different from any intracellularly synthesized host cell protein and thus studies, investigating the mechanism(s) of MHC class I1 presentation of endogenously synthesized host cell proteins, may therefore be equally inforinative for MHC class I1 presentation of endogenously synthesized viral antigens. MHC class I1 presentation studies performed with intracellularly synthesized self or neo-self antigens in professional antigenpresenting cells almost exclusively demonstrated that certain proteins naturally residing in or passing through the EK during their biosynthesis were able to load MHC class I1 molecules (Bikoff et al., 1995; Bodmer et al., 1994; Chen et al., 1990; Moreno et al., 1991; Newcomb and Cresswell, 1993; Rudensky et al., 1991a,b; Weiss and Bogen, 1991). In contrast, endogenously synthesized proteins that do not have natural access to the ER are generally not presented on MHC class I1 molecules. The pathways used by the intracellularly synthesized proteins are apparently not identical in all cases, as in some studies the presence of Ii proved to be inhibitory for MHC class I1 loading (Bodiner et al., 1994; Newcoinb and Cresswell, 1993), whereas in other studies the presence of Ii seemed to be irrelevant for MHC class I1 loading (Mdnati et al., 1992; Nuchtern et d., 1989; Sweetseret al., 1989).Endogenously synthesized membrane proteins might alternatively load MHC class I1 molecules during a membrane-endosome recycling process that would predctably involve membrane protein processing in endosomal compartments. Although this might be true for some membrane proteins, it has also been demonstrated that certain membrane
320
ANNETTE OXENIUS et ~l
proteins do not require acidic proteolysis, thus favoring a more direct MHC class I1 loading mechanism (Chen et al., 1990). Several studies on MHC class I1 presentation of infectious virus-derived endogenously synthesized antigens are not easily generalizable. The difficulties in comparing these results are based on one hand on the fact that nonprofessional APCs were often used for these studies. The relevance of the nature of the APC has been revealed, for example, by studies where identical intracellularly synthesized antigens were differentially capable of loading MHC class I1 molecules in fibroblasts as compared to myeloma cells (Bikoff, 1991, 1992; Moreno et al., 1991). On the other hand, and most importantly, the characteristics of a viral infection in terms of cytopathogenicity, influence on host cell membrane structure and host cell protein synthesis, and processing may lead to apparently contradictory experimental findings. Several studies performed with recombinant vaccinia viruses revealed that MHC class I1 loading by intracellularly synthesized viral proteins was generally chloroquine sensitive, regardless of the subcellular localization of the recombinant proteins, and thus indicating processing in endosomes (Jaraquemadaet al., 1990; Jin et al., 1988;Malnati et al., 1992). Vaccinia infection leads to a major shutdown in cellular host protein synthesis resembling treatment with brefeldin A (BFA) or cycloheximide (Fields, 19901, which may deprive the infected APC of newly synthesized MHC class I1 molecules that could potentially be loaded outside endosomal compartments. Thus, only recycled MHC class I1 molecules from the surface would be available for loading virus-derived membrane proteins, a process that should be inhibitable by chloroquine (Polydefkis et al., 1990). In addition, cytoplasmic proteins derived from recombinant vaccinia viruses appear to gain access to endosomal compartments, possibly be degradation-promoting pathways (Ciechanover, 1994), by association with heart shock proteins (Schirmbeck and Reimann, 1994), or as a result of the large intracellular aggregates formed by vaccinia that might fuse with endosomal compartments (Fields, 1990).Because vaccinia virus and influenza virus are both cytolytic viruses, it is conceivable that they disrupt intracellular compartments. This renders a comparison of the processing and antigen presentation pathways of viral proteins in different subcellular localizations difficult, if not impossible, as the proteins may finally be uniformly distributed within the virus-infected cells. Thus, infectious viral systems using noncytopathic viruses such as LCMV that only minimally affect host cell functions may represent a more convenient system to study antigen presentation of endogenously synthesized viral proteins (Oxenius et al., 1995).Analysis of MHC Class I1 presentation of endogenously synthesized viral proteins after LCMV infection of professional APCs revealed that the subcellular localization of the protein crucially
CD4+ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
32 1
influences whether MHC class I1 loading can occur independently of acidic Compartments. In analogy to data obtained with transfected cell lines and analysis of eluted natural self-peptides bound to MHC class I1 molecules, this study demonstrated that only the intracellularly synthesized membrane-associated LCMV-glycoprotein was able to load MHC class I1 molecules independently of processing in acidic compartments, whereas the intracellularly synthesized cytosolic LCMV-nucleoprotein was shown to be incapable of loading MHC class I1 molecules in professional APCs. In addition, this process was shown to be TAP independent, suggesting that the intracellularly synthesized LCMV-GP, reaching the ER during its natural biosynthetic pathway, is most probably able to bind to newly synthesized MHC class I1 molecules by efficiently competing with Ii. In general, these data indicate that alternative, albeit less efficient, processing pathways exist, allowing MHC class I1 molecules to be loaded with endogenously synthesized proteins. In the case of viral infections with noncytopathic viruses, virus-infected professional APCs may be able to activate virus-specific Th cells even before the infected cell is lysed and before viral antigens are liberated to enter the “classical” MHC class I1 presentation pathway. Under normal circumstances the assessment of such alternative MHC class I1 presentation pathways in vivo has proven to be difficult, mainly because of the dominating efficiency of the classical presentation pathway. Nevertheless, under circumstances where classical antigen presentation is severely hampered, functional MHC class I1 presentation by endogenously synthesized viral proteins could be demonstrated after LCMV infection of Ii-deficient mice. At early time points after infection, only endogenously synthesized membrane-associated LCMV-glycoprotein and not cytosolic LCMV nucleoprotein was presented on professional APCs, leading to early, selective activation of LCMV-glycoprotein-specificTh cells (Oxenius et al., 1997).This suggests that noncytopathic viruses infecting APCs and being capable of presenting Th cell epitopes derived from endogenously synthesized proteins are able to functionally induce Th cells more readily than only after CD8’ T-cell-mediated cell lysis, which would liberate sufficiently enough viral antigens that could enter the “classical” MHC class I1 presentation pathway. C. APC TARGETING 1. Soluble Antigens
Administration of soluble antigen in v i m leading to MHC class I1 associated antigen presentation has revealed apparently conflicting results in terms of the APC subset primarily targeted by a certain antigen (Constant
322
ANNETTE OXENIUS
et
(il
et al., 1995b,c; Gu6ry et al., 1996; Levin et al., 1993).The form in which an antigen is delivered (soluble protein antigen in saline, soluble protein antigen in adjuvant, peptide antigens) may critically favor the targeting of one APC subset or the other. It has been demonstrated in several studies that administration of peptide antigen in adjuvant requires the presence of dendritic cells for priming of naive CD4' T cells, whereas the protein form of the same antigen is presented only poorly by dendritic cells (Constant et al., 1995b,c; Levinetal., 1993);instead, the presence of B cells seems to be aprerequisite for priming with protein antigen (Constant et al., 1995b), especially if the antigen concentration is limiting. Contradictory results were obtained by Gu6ry et al. (1996),who showed that dendritic cells but not B cells present antigenic complexes to CD4+T cells followingthe administration of protein antigen in adjuvant. The requirement of B cells to present protein antigen to CD4' T cells apparently vanes from one protein to another; whereas hen egg lysozyme (HEL),keyhole limpet hemocyanine (KLH), and ovalbumin (OVA)do not seem to absolutely require B cells for activation of naive CD4' T cells, cytochrome c, chicken y-globulin, conalbumin, and human collagen IV seem to require B cells as M C s for priming CD4+T-cell responses (Constant etal., 199%). Several considerations may help to explain, at least partly, these contradictory observations: (1)the dose of antigen effectively present on in vivo administration (influenced by the stability of the protein); preferential targeting of the protein to antigen-specific B cells might occur if the antigen concentrations are limiting. (2)The precursor frequency of antigenspecific B cells may influence their involvement in protein antigen presentation, i.e., B cells at a very low precursor frequency or exhibiting low avidities for the protein antigen might not be able to reach a critical threshold concentration of APC density or antigen density to induce priming of naive CD4' T cells. (3) The form of the antigen effectively present on administration might play a role: mixing of protein antigens with adjuvant such as CFA or IFA leads to the denaturation of the proteins. Denatured proteins often have the tendency to aggregate especially if natively hydrophobic core domains are exposed to each other on denaturation. Thus, protein aggregates might be preferentially phagocytosed by macrophages or macropinocytosedby immature dendritic cells. (4)Most readout systems defining the APC primarily involved in antigen presentation were indirect assays, i.e., the capacity of a certain APC subset to activate naive CD4' T cells in vivo. The activation of naive CD4' T cells requires distinct costimulatory signals provided by the APC in addition to MHC class I1 presentation of the relevant epitope. These costimulatory requirements might thus play an important, if not a crucial, role in the differential APC requirements of certain soluble antigens for priming of naive CD4' T cells. This issue is addressed in more detail in the next section.
CD4' T - C E L L INDUCTION AND EFFECTOR F U N C T I O N S
323
2. lnfectious Agents The targeting of different subsets of APCs by antigens derived from infectious viruses may be divided into two categories: (1)The cell tropism of the virus, replication kinetics, cytopathogenicity, and iinmunopathology influence geographical antigen distribution, antigen load, and the integrity of antigen-presenting cell compartments in secondary lymphoid organs, whereas (2) virus-derived antigens liberated froin infected host cells may be physically regarded as equivalent to soluble protein antigens, and thus soine of the previously mentioned parameters for APC targeting might as well be applied to these viral antigens, although viral antigens most probably never appear in truly soluble form but are usually aggregated and/or meinbrane associated. The cell tropism of a virus may influence in either a positive or a negative fashion the kinetics and the efficiency by which viral determinants are presented on MHC class I1 molecules on professional APCs. On one side, viruses directly infecting professional APCs such as LCMV, EBV, and HIV potentially can very rapidly present viral determinants on MHC class I1 molecules by class I1 loading with intracellularly synthesized viral proteins (Oxenius et al., 1995) and thus rapidly activate naive CD4' T cells. However, efficient and prolonged presentation is only achieved if the virus is not cytopathic, thus destroying the APC within a short time period on infection or if the infected APC is not readily lysed by cytotoxic effector T cells, which can result in severe immunopathology accompanied with immunosuppression (Odermatt et nl., 1991). However, viruses not infecting professional APCs or viruses that cannot load MHC class I1 molecules by intracellularly synthesized proteins require host cell destruction and liberation of viral antigens that can be taken up by professional APCs. Depending on the cytopathic or noncytopathic character of the viral infection, infected cells are lysed either by the virus or by CD8' effector T cells. The thus liberated antigens most probably reach high local concentrations (Battegay et al., 1996),which are even sufficient to override some APC deficiencies (such as Ii deficiency) and thus are able to load MHC class I1 molecules on local APCs. The aggregated or membraneassociated physical appearance of viral protein antigens derived from lysed virus-infected cells is prone to phagocytosis. In addition, because these protein aggregates usually exhibit native protein structures, antigen-specific B cells might also, at least to some extent, be involved in antigen uptake and processing. Due to the postulated rnultimeric appearance of these antigens, the Ig receptors on the surface of virus-specific B cells are likely to be cross-linked, resulting in upregulation of costiinulatory moIecules (Cassel and Schwartz, 1994), and thus most probably enabling priming of naive CD4' T cells and facilitating T-B collaboration (Bachmann and Zinkernagel, 1997).
324
ANNETTE OXENIUS et al
D. THCELLACTIVATION 1. Soluble Antigens CD4' T cells recognize antigenic peptides bound to MHC class I1 molecules on the surface of APCs via their T-cell receptor. This interaction confers the specificity component for the interaction of a given Th cell with peptide-loaded MHC class I1 molecules and is, in terms of Th cell activation, generally described as signal 1. This signal 1 alone has been shown in many instances to be insufficient for activation of naive Th cells (Dubey et al., 1995; Jenkins, 1994). More precisely, it has been shown to induce Th cell unresponsiveness (anergy) rather than Th cell activation (Celis and Saibara, 1992; Harding et al., 1992; Schwartz, 1990). Thus, activation of antigen-specificnaive CD4' T cells requires a second signaling event that must be equally provided by the APC (Mueller et al., 1989). This second signal is provided by costimulatory molecules located on the APC surface, the best characterized ones being the B7 molecules (B7-1 and B7-2), which have been shown to act as ligands for CD28 and CTLA4 on CD4' T cells (Linsley et al., 1991; Linsley and Ledbetter, 1993). The interaction of CD40 on the APC (especially on B cells) and CD40L (expressed on activated CD4' T cells) has been suggested to be of costimulatory character for T cells (Grewal et al., 1995; van Essen et al., 1995). However, because CD40L is only expressed on activated CD4' T cells, this CD40-CD40L interaction is apparently not a costimulatory signal for the activation of naive CD4+ T cells, but may be, in certain experimental systems, important for the clonal expansion of antigen-specific Th cells (van Essen et al., 1995). This, however, remains to be shown, and the apparent requirement of the CD40-CD40L interaction for the clonal expansion of pigeon cytochrome c-specific CD4' T cells probably reflects the requirement of CD40-competent B cells acting as APCs for Th cell priming (Constant et al., 1995c) rather than a general requirement of the CD40-CD40L interaction for Th cell induction. The question of whether CD40-CD40L interactions are crucially involved in Th cell activation following viral infection is discussed in the next section. The requirement for the two types of signals to successfully activate naive CD4+ T cells has led to much controversy regarding the putative role of different classes of professional APCs in CD4' T cell priming in vivo. Based on their constitutive expression of costimulatory molecules, as well as their abundance of surface MHC class I1 molecules, it has been postulated that the dendritic cell is the only subset of APCs having the capacity to prime naive CD4+ T cells (Cassel and Schwartz, 1994; Steinmann, 1991). Indeed, the potency of dendritic cells to activate resting CD4+ T cells either in vitro (Cassel and Schwartz, 1994; Ellis et al., 1991;
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
325
Inaba and Steinmann, 1985) or in vivo (Inaba et al., 1990; Levin et al., 1993; Sornasse et al., 1992) is well documented. Macrophages and B cells, however, have been shown to require activation signals themselves to upregulate costimulatory molecules and to become competent APCs for resting T cells (Janewayand Bottomly, 1994;Jenkins, 1994). Such activation signals include soluble factors such as interferon (IFN)y, IFNa, bacterial carbohydrates, or other microbial constituents in the case of macrophages and IFN-y, microbial polysaccharides, or surface Ig cross-linking antigens in the case of B cells. Nevertheless, mature dendritic cells also have passed through an activation phase: immature dendritic cells located at potential entry sites of pathogens such as the Langerhans cells in the skin are very efficient in antigen uptake, and it is only on granulocyte/macrophage colony-stimulating factor (GM-CSF) and interleukin (IL)-4-induced activation that they upregulate MHC and costimulatory molecule expression to become fully competent dendritic cells after migration into local secondary lymphoid organs (Austyn, 1996; SaUusto and Lanzavecchia, 1994). However, an unresolved controversy still exists on the issue of whether B cells play a central role in the initiation of Th cell responses. Several groups have reported that mice deprived of B cells were unable to be primed by soluble antigens for proliferative T-cell responses, suggesting that B cells are essential for CD4+ T cell priming (Constant et al., 199513; Janeway et al., 1987; Kurt-Jones et al., 1988; Ron et al., 1981; Ron and Sprent, 1987). This failure to prime was restored when purified antigenspecific B cells were transferred into B-cell-deficient mice (Kurt-Jones et al., 1988; Ron and Sprent, 1987). In contrast, other studies indicated that B cells were not necessary for the priming of CD4+ T-cell responses (Bottomly et al., 1980; Gotoff, 1968; Lassila et al., 1988; Ron and Sprent, 1987; Ronchese and Hausmann, 1993; Sunshine et al., 1991).Even studies using the same experimental antigen such as KLH revealed different results: Epstein et al. (1995) reported successful T-cell priming in B-cell-deficient mice, whereas Liu et al. (1995) failed to obtain T-cell priming. In addition, several groups reported that antigen presentation by B cells is rather tolerogenic than immunogenic for naive T cells (Eynon and Parker, 1992; Fuchs and Matzinger, 1992; Gilbert and Weigle, 1994). Taken together, the relevance of B cells in the activatiodtolerization of naive Th cells may depend on several factors, including (1)the ability of an antigen to induce costimulatory molecule expression on B cells (e.g., by cross-linking of the Ig receptors) or via a certain cytokine milieu, (2) the actual concentration of the antigen (low concentrations requiring antigen concentration via the Ig receptors of B cells), (3) the physical appearance of the antigen (truly soluble proteins versus aggregates), and (4) the precursor frequency of antigen-specific B
326
ANNETTE OXENIUS et rd.
cells. In addition, although B cells may not prime native T cells, they may amplify specific T cells previously activated by dendritic cells. Another aspect of Th cell activation that has become a focus of interest is the description of apparently discrete Th cell phenotypes as defined by distinct cytokine secretion patterns. A cellular basis of this distinctive pattern of lymphokine production was provided by the seminal observation of Mosman and Coffman (1987, 1989) that long-term clones of mouse CD4+ T cells could be subdivided into those producing IL-2, IFNy, and TNFa (Thl clones) and those producing IL-4, IL-5, IL-6, IL-10, and IL13 (Th2 clones). In many studies the selective generation of Thl and Th2 CD4+T-cell clones was performed by the addition of either IL-12 or IFNy (Thl clones) or IL-4 (Th2 clones) into the culture medium (Mosmann and Coffman, 1989; O'Garra et al., 1993; O'Garra and Murphy, 1994; Swain et al., 1990).Long-term Thl or Th2 clones have lost their reversibility and seem to be definitively polarized (Murphy et al., 1996),and the production of the discrete cytokine patterns has been analyzed on a single cell level using intracellular FACS staining (Assenmacher et al., 1994). Functional aspects of the discrete Th cell populations are discussed in Section 111. The possible factors governing the acquisition of a certain Th cell phenotype have been the subject of many investigations. The parameters suggested to influence Thl/Th2 development involve (1) the subset of the professional APC (Duncan and Swain, 1994),(2)the cytokine milieu during the process of Th cell priming (Swain et al., 1991), (3) the antigen dose (Hosken et al., 1995), (4) inherent characteristics of an antigen (Julia et al., 1996), (5) the genetic background of the animal (Hsieh et al., 1995), and (6) the degree of TCR cross-linking (Constant et al., 1995a). The relevance of these parameters for the selective induction of Thl or Th2 phenotypes in murine Th cells in v i m is still a matter of debate and most probably is the sum of different parameters that will determine the phenotype of the activated Th cell. A second open question is whether activation of Th cells in vivo leads to the generation of such distinct polarized Th cell populations as demonstrated for Th cell clones in vitro or whether the situation in vivn might rather move within a continuum between the two extremes. In general, appropriate Th cell activation leads to a series of intracellular signal transduction events that originate at the TCR-CD3 complex and finally lead to the induction of transcription of certain genes; the entry of the cell into the cell cycle, resulting in clonal expansion, up- or downregulation of surface molecules such as CD69, CD40L, IL-2R, or L-selectin; and the acquisition of effector functions such as the capacity to secrete cytokines and the ability to provide cognate help for B-cell activation. Effector functions of Th cells are described in more detail in Section 111.
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
327
2. fnfectious Agents Because MHC class I1 presentation usually occurs after uptake of soluble proteins/protein complexes, an implication of this mode of antigen presentation is that the direct interaction of CD4' T cells with virally infected cells is not required for CD4' T-cell activation, as opposed to CDBt T cells that are preferentially activated by vinis-infected cells. However, some viruses such as measles virus, human immunodeficiency virus (HIV)-l, and LCMV have tropisms for cells bearing MHC class I1 molecules. In this circumstance, CD4' T cells may recognize and be directly activated by the infected cell due to the ability of some endogenously synthesized viral proteins to directly load MHC class I1 molecues (Long, 1992; Malnati et nl., 1992; Oxenius et nl., 199S, 1997; Pinet et nl., 1994). In addition to the previously discussed requirements and mechanisms leading to activation of naive CD4t T cells (such as the necessity of costimulatory molecule expression on APCs and the parameters influencing the induction of Th1 versus Th2 Th cell phenotypes), several additional aspects influencing and governing Th cell activation on infection have to be considered: viral infection of the host organisin leads to the induction of multiple immune effector functions aimed at defending the organism against the virus. Innate defense mechanisms involve the activation of IFNy-producing natural killer (NK) cells, which leads to macrophage activation and thus the upregulation of costimulatory molecule expression. Several microbial constituents have been shown to directly induce costimulatory molecule expression on macrophages, probably by the binding to specific receptors. It seems likely that these receptors originally evolved to allow the phagocytic cells in primitive organisms to recognize microorganisms by binding to structures such as bacterial carbohydrates or lipopolysaccharides that were not found in eukaryotes. These receptors still serve this function in innate immunity, as well as playmg an important part in the initiation of adaptive immune responses. The induction of costimulatory molecule expression by common microbial constituents is believed to allow the immuiie system to discriminate between antigens born by infectious agents and antigens associated with innocuous proteins, including self-proteins (if not administered in adjuvant) (Janeway, 1989; Liu and Janeway, 1991; Razi-Wolf et al., 1992). Viral infections usually lead to local inflammatory reactions at the site of infection, attracting several proinflammatory cells such as polymorphonuclear leukocytes, monocytes, and macrophages. On inflammation, a local cytokine milieu is generated, which is again favorable for the activation of APCs such as, for example, for locally resident immature dendritic cells. In addition to dendritic cells arid macrophages, B cells are also usually
328
ANNE’ITE OXENIUS et al.
rapidly activated to express costimulatory molecules, as viral particles usually express surface antigens in a more or less highly repetitive form and thus induce cross-linking of the surface Ig receptors on a virus-specific B cell. The cross-linking is sufficient to upregulate costimulatory molecule expression on B cells (June et al., 1994) and may allow for Th cell activation by B cells. These considerations, together with the dynamic kinetics of the viral antigenic burden and the fact that viral infections usually induce a complex immune response exhibiting innate immune effector functions, cytotoxic T-cell activation, and antibody production, might all be responsible for less strictly defined Th cell activation pathways after infection as compared to the probably more subtle pathways described for protein antigens. The CD40-CD40L interaction between APCs and activated Th cells has been shown to (1) exhibit costimulatory function in experimental systems using soluble proteins (Grewal et al., 1995; van Essen et al., 1995) and (2) govern protective Thl phenotype Th cell development in certain infectious experimental systems, such as infection with Leishmania (Campbelletal., 1996;Kamanakaetal., 1996; Soongetal., 1996;Stuberetal., 1996). The primary defect in CD40- or CD40L-deficient mice on Leishmania infection was demonstrated to be a lack of IL-12 production by macrophages, which is apparently instrumental in resolvingLeishmania infection. This lack of IL-12 production was apparently causing an impairment of macrophages to reach a leishmaniocidalactivation state. In contrast, Th cell induction and effector functions in viral infections seemed to be less dependent on a functional CD40-CD40L interaction: vesicular stumatitis virus (VSV)infection in the absence of a CD40-CD40L interaction induced Th cell effectors, which were able to protect against a challenge infection with a recombinant vacciniavirusexpressingVSV-derivedproteins. In addition, LCMVinfection of CD40-deficient mice induced a Thl-type Th cell response that was able to mediate cognate help to CD40-competent B cells upon adoptive transfer. Furthermore, the induction of LCMV-specificTh cell responses was shown to occur independently of B cells potentially serving as APCs (offering the possibility to analyze involvement of CD40-CD40L interaction in Th cell priming apart from B-cell APC involvement), thus rendering dendritic cells or macrophages the prominent APC subsets in Th cell induction on LCMV infection (Oxenius et al., 1996). These observations strongly suggest that activation of at least certain antiviral Th cell responses is not critically dependent on CD40-CD40L costimulatory interaction, whereas protein antigens are more dependent and anti-Leishrnnia protective immune responses seem to critically depend on a certain activation status of “effector” macrophages that is apparently not achieved in the absence of Thl-type effector Th cells, as observed
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
329
in CD40- or CD40L-deficient mice. However, not only were some viral infections shown to functionally induce virus-specific Th cells in the absence of a CD40-CD40L interaction, but immunization with sheep red blood cells (SRBC) induced normal Th cell responses in the absence of the CD40-CD40L interaction (Foy et al., 1993). As discussed earlier, the observed impairment of Th cell responses in the absence of a functional CD40-CD40L interaction may be due to the absence of activated B cells or, more generally, to the absence of costimulatory competent APCs in the case of iminunizations with soluble antigens. This view is supported by the findings (1)that adoptive transfer of B7.1-transfected APCs into CD40L-deficient mice can rescue the Th cell response (Grewal et al., 1996) and (2) that CD40-dependent upregulation of B7.2 in vivo led to full reconstitution of cellular and humoral immune responses in CD40Ldeficient mice (Yang and Wilson, 1996). The phenotype of Th cells induced on viral infections seems to be generally skewed into the Thl phenotype direction. This is reflected by (1)the predominance of IL-2- and IFNy-producing virus-specific Th cells and (2) the predominant IgG2a isotype of antiviral antibodies (Coutelier et al., 1987; Zinkernagel, 1993). In addition, intracellular bacteria such as listeria and mycobacteria induce Thl-dominated Th cell responses (Abbas et nl., 1996; Hsieh et al., 1993). Whether this biased T h l phenotype development is influenced by innate iminune mechanisms associated with a certain cytokine milieu induced after viral infection, by the antigen concentrations reached after virus infection, by the antigen-targeted APCs, or by some inherent characteristics of viral epitopes involved in this phenotype commitment is still unclear. However, protozoa infections such as Leishnzunia have been shown to be able to induce both Thl- and Th2-type Th cell responses depending on the genetic background of the animal (Boom et al., 1990; Heinzel et al., 1989) and on the Leishmania T-cell epitope controlling initial Th cell induction. It has been demonstrated that the normally observed nonprotective Th2 Th cell induction in BALB/c mice on Leishmania infection was crucially controlled by the iinmunodoininant LACK epitope (Leishmania hoinolog of receptors for activated C kinase) during initial Th cell activation. If this epitope was functionally deleted or if the respective Th cells were tolerized, T h l Th cell development was observed Leishmania infection in BALB/c mice (Julia et ul., 1996). Taken together, the parameters governing T h l and Th2 phenotype development on infection are still poorly defined, and it seems likely that in addition to the genetic background of a host, the infectious agent may selectively influence the phenotype of Th cell responses. It will be of great interest and value to define these parameters in more detail to understand
330
ANNE'ITE OXENIUS et ol
on one hand the mechanisms pathogens use to drive Th cell phenotype development and on the other hand to be able to efficiently interfere with these mechanisms in order to manipulate Th cell responses in favor of the host. 111. CD4+ T-cell Effector Functions
Effector functions mediated by activated CD4+ T cells include (1)cognate help for B cells promoting B-cell activation and Ig isotype switch; (2) secretion of cytokines exhibiting multiple functions, such as optimal activation of different cell subsets (macrophages, dendritic cells, cytotoxic T cells, etc.), direct effects on pathogen replicatiodsurvival, and governing isotype production by activated B cells; (3) delayed-type hypersensitivity (DTH) reactions; and (4) certain CD4' T cells have been shown to exhibit cytolytic potential. In general, CD4' T cells are believed to play a central role in the regulation of the cooperation of the different arms of an immune response in acute and especially chronic infections. The relevance of CD4' effector T-cell functions in viral infections differ from one virus to another and therefore have to be characterized for each viral infection separately. Initial studies characterizing Th cell effector fuctions in vivo were performed using soluble and chemically defined antigens for immunization. In vitro parameters characterizing antigen-specific Th cell activation are recall Th cell proliferations and assessment of concomitant cytokine release. In vivo parameters, which are perhaps biologically more relevant for the analysis of Th cell effector functions, include the generation of (isotypeswitched) antigen-specific antibody responses as well as antigen-specific DTH reactions on (re)immunization with soluble antigens. Because both antibody production and DTH reaction are also Th cell effector functions in viral infections, this section focuses on the different Th cell effector functions and their relevance in the control of different viraI infections. The first part gives a more general overview of different Th cell effector functions in viral infections, and the second part briefly discusses several selected disease models in terms of CD4' effector T-cell functions involved in or dispensable for effective resolution of infection. One of the most important and thoroughly studied effector mechanisms of virus-specific Th cell is their capacity to drive the production of virusspecific antibodies. Neutralizing antiviral antibodies (with specificity for viral surface antigens) have been shown to be crucially involved in the resolution of certain viral infections (e.g., VSV, rabies) (Lefrancois, 1984; Murphy, 1977; Wagner, 1987). Antibodies represent the most efficient defense against secondary infections with the same virus (Mims, 1987; Thomsen and Marker, 1988). Certain viruses such as VSV, as well as some
CDJ’ T-CELL INDUCTION AND EFFECTOH FUNCTIONS
33 1
repetitive soluble antigens, are able to activate specific B cells independently of T help (TI antigens), and they are usually characterized by a highly repetitive organization of the relevant B-cell antigen either on the surface of a virus or as an inherent structure of the antigen TI antigenmediated activation of antigen-specific B cells results in the secretion of IgM isotype antibodies, but induction of isotype switch and long-term secretion of antibodies are largely dependent on specific T help (Bachmann and Zinkernagel, 1996). Several studies have established that TI1 cells can support antigen-specific B-cell responses through two distinct pathways (reviewed in Melchers and Andersson, 1984;Julius, 1982).One, referred to as cognate help, is thought to result from direct T-B interaction, whereas the other, referred to as noncognate or bystander help, is mediated through factors released from activated T cells that act in a noncognate fashion on antigen-activated B cells. The former is probably more relevant for antigen-specific B-cell responses. Classic cognate help appears to require covalently linked T and B-cell determinants, whereas special rules may apply, however, for T-B interactions in antiviral B-cell responses. For example, mice primed with influenza virus cores [devoid of the surface antigens hemagglutinin (HA) and neuraminidase (NA)] or with purified matrix (M) protein exhibited an enhanced anti-HA antibody response when challenged with intact virus, but not when challenged with a mixture of virus cores and purified HA (Russell, 1979; Russell and Liew, 1980). Because viral internal determinants are not covalently linked to HA, the observed “cognate” help was termed interinolecularhntrastructurd help (Lake and Mitchison, 1976). Comparable findings were reported using Th cell clones specific for internal influenza virus proteins that could mediate help for anti-HA antibody production (Scherle and Gerhard, 1986) or VSV-induced autoantibody production in transgenic inice expressing the VSV-G (Zinkernagel et al., 1990). However, Th cells with specificity for viral surface antigens are unable to mediate help for B cells specific for virus internal proteins, demonstrating that intramolecular help is required to promote antibody responses against virus internal proteins. Apparently, virus internal proteins are at the time point of binding to sIgs on B cells no longer physically linked to viral surface antigens (Oxenius et al., 1998). Surface molecules critically involved in successful cognate T-B interaction leading to B-cell activation involve TCR-MHC-peptide interaction and interaction between CD40L on the activated TI1 cell and CD40 on the virus-specific B cell. The relevance of a functional CD40-CD40L interaction in terms of antibody production has been demonstrated for both soluble antigens (Banchereau et nl., 1994; Parker, 1993) and virus infections (Borrow et al., 1996; Oxenius et nl., 1996).
332
ANNE'ITE OXENIUS et a/
The kinetics of antiviral antibody production usually seems to depend more on the availability and precursor frequency of B cells exhibiting antiviral specificity than on the presence of activated T help as Th cell priming before virus infection often only marginally affects the kinetics of antiviral antibody production. This has, for example, been demonstrated for the kinetics of VSV-neutralizing antibody production (Charan et al., 1986), for LCMV-binding antibody production (unpublished results), for anti-influenza antibody responses (Liang et al., 1994), and for the late appearance of HIV-neutralizing antibodies and of LCMV-neutralizingantibodies where virus-specific T help is primed at early time points after infection (Oxenius et al., 1995; Pantaleo and Fauci, 1995). Another important feature of Th cell effector functions is the secretion of various cytokines that can exert their function locally and systemically such as direct inhibition of virus replication and the amplification and regulation of other components of the host immune response. The cytokine pattern secreted by certain Th cells has often been correlated with distinct Thl or Th2 phenotypes of activated Th cells. Although this strict classification into T h l and Th2 subsets has been demonstrated clearly many times in in vitro studies and also in a few in vivo disease models, such as in Leishmania infection, it is still a matter of debate whether such clearly defined Th cell phenotypes are generally induced upon infection or whether they represent a rather extreme situation. Nevertheless, distinct effector functions can be attributed to different T-cell-secreted cytokines. The Thl-type cytokines include IL-2, IFNy, and TNF, which are all involved preferentially in cell-mediated immune responses such as DTH reactions and cytotoxic T-cell responses. IL-2 is the major growth factor for antigen-specific CD8+ and CD4' T cells, it activates NK cells, and it promotes B-cell differentiation (Farrar et al., 1982). IFNy and TNF activate macrophages and NK cells, inhibit replication of some viruses directly, and promote resistance of uninfected cells to viral infection (Staeheli, 1990; Wong and Goeddel, 1986). It should be noted though that IFNy is also produced by NK cells (Handa et al., 1983). In addition to IFNy produced by T and NK cells, IFNP is produced by fibroblasts and IFNa by leukocytes at the site of infection. Virus-derived double-stranded RNA, for example, is a potent inducer of interferons. The type of interferon interfering with vird replication is dependent on the virus and the infected cell type. For example, vaccinia virus was shown to be most sensitive to inhibition by IFNy, less by IFNa, and least by IFNP (Meshkova and Mentkevich, 1987; Metz and Esteban, 1972). Inhibition of translation and/or degradation of viral mRNAs can account for many aspects of interferon-mediated inhibition of virus replication, but does not
CD4+ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
333
rule out other mechanisms (Kerr and Brown, 1978; Paez and Esteban, 1984; Rice and Kerr, 1984). Th2-type cytokines include IL-4, IL-5, and IL-10, which are involved in promoting humoral responses and differentiation of eosinophils and mast cells (Chen and Zlotnik, 1991; Lopez et al., 1988; Mosmann et al., 1986; Paul, 1991). The paracrine effects of cytokines secreted by CD4' T cells can be focused locally by cell-cell interactions: cognate interaction of T and B cells, for example, leads to polar release of the cytokines at the site of cell contact, thus specifically promoting activation and hfferentiation of the virus-specific B cell involved in cognate interaction (Kupfer et nl., 1991; Lanzavecchia, 1990). The distinctive cytokine profiles attributed to Thl and Th2 Th cell subsets may play a critical role in the control of pathogens if CD4' Th cell responses become strongly skewed into one or the other subset during infection. As mentioned earlier, this was initially appreciated in vivo for parasitic infections such as leishmaniasis: acquisition of a T h l phenotype by CD4' T cells leads to resolution of infection, whereas acquisition of a Th2 phenotype leads to disease progression (Heinzel et al., 1989). The opposite was shown for infection with Trichuris muris (Else et al., 1994). These examples demonstrate that distinct Th cell phenotypes may be critically involved in several infectious model systems. It has been suggested that Thl-like Th cell responses are especially beneficial for the resolution of infections with obligate intracellular pathogens such as viruses, as they have been shown to promote cell-mediated responses and local inflammatory responses. Indeed, most virus infections induce Thl-type cytokinesecreting Th cells, at least during acute phases of infection. Cytokine profiles secreted by murine CD4' T cells obtained from lymph nodes and bronchoalveolar lavage fluid on acute nonfatal infection with influenza virus were strongly biased toward the Thl pattern (Carding et al., 1993; Sarawar and Doherty, 1994). Furthermore, Th cell responses after acute infection with LCMV were shown to be dominated by IFNy- and IL-2producing LCMV-specific Th cells (Oxenius et al., 1996). Human CD4' T cells specific for HIV-1 have also been shown, at least early on infection, to exhibit a Thl-type cytokine secretion pattern. In some reports, however, it was shown to change to a Th2-type pattern at later stages of infection (Clerici et al., 1993; Clerici and Shearer, 1994). In addition, adoptive transfer experiments with polarized virus-specific Th cell clones have provided information about antiviral and biological activities of individual Th ceIl subsets. For example, adoptive transfer of polarized Th1 cells specific for influenza virus mediated protection against
334
ANNETTE OXENIUS et nl
influenza virus, whereas adoptive transfer of polarized Th2 cells failed to promote recovery from infection (Graham et al., 1994). A preponderance of Thl or Th2 cytokine profiles during infection might result from an initially “uncommited” CD4’ T cell that acquires a certain phenotype during the course of the infection and/or Th cells exhibiting a certain phenotype may selectively overgrow others. In addition, a positive feedback mechanism exerted by the phenotype of commited Thl or Th2 cells may act to skew the phenotype of the Th cell response further into Thl or TH2 character. Thus, early events of the immune response, such as innate immunity dominated by IFNy or early preferential IL-4 production, might be crucial in directing the phenotype of the Th cell response induced. Selective expression or neutralization of Thl-type or Th2-type cytokines at the sites of viral infection have also provided information about their immunoregulatory character and their direct antiviral effects, e.g., infections with recombinant vaccinia viruses expressing IL-2, IFNy, TNF, or IL-4 have clearly demonstrated that the Thl-type cytokines IL-2, IFNy, and TNF were able to resolve the infection, even in T-cell-deficient mice, whereas IL-4 expression at the site of infection inhibited resolution of infection (Ramsay et al., 1993; Rainshaw et al., 1992). Delayed-type hypersensitivity reactions are a further effect of certain Thl-type cytokines secreted by CD4’ T cells at the site of infection: CD4’ T cells secrete intercrines (MIF, MCF); IFNy, which activates local macrophages, thus increasing the release of inflammatory mediators; TNFp, which causes local tissue destruction and leads to increased expression of adhesion molecules on local blood vessels; and IL-3/GM-CSF, which activates monocyte production by bone marrow stem cells (Cher and Mosmann, 1987). A fourth effector mechanism of CD4’ T cells has been claimed to be a direct cytolytic activity against antigen-loaded target cells; some virusspecific CD4t T cells have been shown to specifically lyse MHC class II+ virus-infected target cells in vitro and thus potentially could directly eliminate virus-infected cells in vim (Fleischer, 1984; Koelle et al., 1994; Muller et al., 1992). However, the relevance of these in vitro findings for in vivo situations still remains unclear, especially (1)since most viruses infect a much broader range of target cells in vivo than only MHC class 11+cells and (2)because virus-infected MHC class 11’ cells always concomitantly express MHC class I molecules, they should be targets for virusspecific CD8+ effector CTLs. The large body of evidence for functional cytotoxicity of CD8’ T cells in vivo suggests that this population is much more important for the lysis of virus-infected cells and thus the direct cytotoxic activity of CD4’ T cells is rather unlikely to be an obligate in vivo effector function (Kagi et al., 1996; Oxenius et al., 1998).
CD4‘ T-CELL IKDUCTIOK AND EFFECTOR FUNCTIONS
335
A fifth effector mechanism that may be viewed as a regulatory mechanism of CD4’ T cells consists of T help for optimal CTL induction. Although it has been demonstrated in various viral infectious model systems that protective CD8’ effector CTLs can be induced in the absence of CD4’ T cells such as, for example, for LCMV (Aimed et nl., 1988; Leist et nl., 1989), ectromelia (Buller et al., 1987),CMV (Koszinowsky, 1991), vaccinia (Binder and Kundig, 1991), HBV (Ando et al., 1993), and influenza (Tripp et al., 1995), the magnitude of the CTL response was usually reduced in the absence of T help (Leist et nl., 1989). Nevertheless, in these situations CD8’ T cells were induced efficiently enough to control acute viral infection, although different viruses may more or less depend on the presence of CD4’ T cells for the induction of efficient CTL responses. The cell tropism of a virus (infection of professional APCs or not) may play an important role for the requirement of T help for efficient CTL induction: Viruses infecting professional APCs (e.g., LCMV or ectromelia) and thus readly and optimally activating naive CD8’ T cells to become effector CTLs are much less dependent or even independent on T help compared to viruses infecting other cell subsets. It is interesting to note, however, that sustained CTL responsiveness might in both cases be dependent on T help, as CTL responses in chronic viral infections (Matloubian et nl., 1994) seem to depend on T help as well as CTL responses induced on infection with large doses of widely replicating viruses that are prone to induce exhaustion of CD8+ T cells (Battegay et nl., 1994; Moskophidis et al., 1993). The mechanism by which CD4’ T cells mediate help for CTLs is still not clearly defined, but it most probably involves IL-2 secretion by Th cells. In contrast to cognate T-B interactions, T-T interactions are not of cognate nature (at least in the mouse), as inurine T cells do not express MHC class I1 molecules. The relevance of virus-specific Th cell responses for the resolution of different viral infections cannot be generally formulated but has to be investigated for each virus independently. Using adoptive transfer experiments of purified lymphocyte populations or in vivo depletion techniques with monoclonal “therapeutic” antibodies has provided information about the protective role and the relevance of different cell subsets in dfferent virus infections. More recently, thanks to the development of transgene and gene “knockout” technologies in the murine system, mice exhibiting permanent deficiencies of selected cell populations can be generated (e.g., mature CD4’ T cells are absent in MHC class II-deficient mice and mature CD8’ T cells are absent in P2M-deficient mice) or usually extremely heterogeneous cell populations (such as T and B cells) can be selectively enriched in desired (virus-specific) specificities by the expression of an appropriate receptor as a transgene (e.g., TCR or immunoglobulin). The
336
ANNE'lTE OXENIUS et al
latter also allows for the observation of virus-specific cells upon infection in vivo on a cellular level. Results concerning the protective involvement of CD4' T cells in several viral infections are apparently controversial, but the different parameters often applied in these studies, such as virus dose and/or virulence of a virus strain as well as short-term versus long-term protective readouts, might explain some of these discrepancies. In addition, in the case of genetically modified mice, the immune system may develop some degree of flexibility to compensate for a deficiency that is present during ontogeny, thus sometimes rendering comparisons with normal mice difficult. The following paragraphs briefly summarize the involvement of virusspecific CD4' T-cell responses in protective immunity against some selected viral infections (Table I). A. LYMPHOCYTIC CHORIOMENINGITIS VIRUS
LCMV is a noncytopathic arenavirus whose natural host is the mouse. A large body of evidence demonstrates that protective immunity is primarily mediated by LCMV-specific CD8' effector T cells after acute systemic infection (Buchmeier et al., 1980; Kagi et al., 1994; Moskophidis et al., 1987; Zinkernagel and Althage, 1977). Depletion of CD4' T cells by monoclonal antibodies only reduces CTL activity by a factor of 5-15, which does not influence the resolution of acute infection (Ahmed et al., 1988; Leist et al., 1989). Similarly, cytotoxic T-cell responses were efficiently induced in MHC class II-deficient mice capable of eliminating LCMV after acute infection (Battegay et al., 1996). Nevertheless, CD4' T cells were shown to be required to sustain CD8+T-cell responses during chronic LCMV infection: Whereas normal mice were able to clear certain LCMV isolates within 3 months after infection, mice even transiently depleted of CD4' T cells became life-long carriers, which was correlated with a disappearing CTL response (Matloubian et al., 1994). Thus, CD4' T-cell function may become critical if CTL activity has to be maintained for several weeks or months. This may be relevant for HIV infection where antiviral CD8' Tcell responses are ultimately lost, perhaps due to the loss of T help. Also in this line, Battegay et al. (1994) demonstrated that adult CD4' T-celldeficient mice exhibited enhanced establishment of a virus carrier status on infection with widely replicating LCMV-Docile. CD4' T cells have also been demonstrated to be involved in the establishment of optimal CTL memory in terms of CTL precursor frequencies: CDCdepleted mice showed reduced precursor frequencies of memory CTLs and were less protected against challenge infection than untreated control mice (von Herrath et al., 1996).A direct protective role of a LCMV-
TABLE I PROTECTIVE CAPACITY OF VIRUS-SPECIFIC CD4+ T CELLSI N SELECTEDMURINEVIRUSINFECTIONS Vinis
Cytopathic
Adoptive Transfer of Virus-Specific CD4' T Cells
LCMV
No
Usually not protective (Oxenius et d , 1998) exception: La Posta et al. (1993)
Influenza n r u s
Yes
Sendai wrus
Yes
Thl transfer protective (via ah production) (Craham et a/., lW4; Scherle et a / , 19921 n.p.'
Ectrnmelia virus
Yes
r1.p
Yes
Th I transfer protective (Vaenius et d..1998) n.p.
CD4-Deficient Mice" Short term: nonnal protection (Battegay et a!., 1996). CD4' T cells required for long-term protection (Thomsen et nl., 1996) Clearance froin lungs (Allan rt al., 1990)
Protected (Hoii et ol., 1995)
CD8-Deficient M i d Slow or absent w u s clearance (Fung-hung et o l , 1991)
Protection in niost cases (Bender el a / , 1992: Eichelherger et a/., 1991; Epstein ?i nl., 1993) Delxyed clearance (Hou et al.. 1992)
Yes Yes
nw
YCP
Polio vinis
Yes
Theiler'F WNS
Yes
Measles vims
Yes
Limited virus spread (Jonjic c't d , 1990) and inhibition of interstitial pneumonia (Kadima-Nzuijand Craighead. 1990) Tlr 1 transfer protective against HSV-1 challenge (Manickan uf a l , 1995) Thl transfer protective (via neutralizing ab produetionl (Mahou et d..1995) Thl transfer: accelerated disease (Gerety et a/., 1994) Control of ic M\' infection (Reich el a/.. 199%
Protected (BuUer et nl., 19x71 but s incomplete v i ~ clearance (Karupiah et o l , 1996) Protected (Battegav ct ul., 19961
k t h a l outcome 01infection (Kanipiah et nl., 19%)
Protected (if' B-cell competent1 (Battegay et a l , 1996) Delayed and inappropriate clearance (Jonjic et a / . , 19891
Protected (Browninget 01, 1990)
Increased susceptibility to HS\' infection of skin and newons system (Nash et a l , 1987) n.p.
Protected (Manickan and House,
Not protected (lethal infection) (Borrow rt nl., 1993: Kurtz et al., 19951 BALBlc mice become susceptible to infection (Finke and Liebert, 1994)
' Genetically engineered mice deficient for either CD4 or MHC class I1 or normal mice depleted in uiljo of CD4' T cells. ' Genetically engineered mice deficient for either CD8 or P2-M or normal mice depleted in oit;o of CD8' T cells. ' Not published.
Protected (Sprigs et
a/.,
1902)
Protected (Jmjic ~t nl., 19901
1995)
n.p.
Disease-inducing CD4- T-cell response enhanced (Pullen et or., 1993) BALB/c mice remain resistant to infectioii (Finke and Liebert, 1994)
338
ANNETTE OXENIUS et al
specific Th cell clone against LCMV infection has been demonstrated: adoptive transfer of a LCMV-glycoprotein-specific Th cell clone mediated protection against low doses of intrdcerebrally inoculated LCMVArmstrong virus (La Posta et al., 1993). Cytokines such as IFNy secreted by Th cell clones are most likely able to control or at least delay lowdose LCMV infection in secluded compartments such as the brain, where diffusion of cytokines is minimal due to the blood-brain barrier (Kundig et al., 199313). The generation of a LCMV-specific antibody response is strictly dependent on T help (Ahmedet d., 1988).Although LCMV-neutralizing antibodies only appear very late after initial infection and in moderate titers (Battegay et al., 1993), antiviral antibodies seem to play a crucial role in protection against reinfection (Thomsen and Marker, 1988).Thus preexisting neutralizing anti-LCMV antibody titers efficiently inhibited infection with LCMV (Seiler et al., 1997), and even LCMV immune serum with little or no neutralizing activity was shown to markedly reduce both virus spread and replication as well as LCMV-specificcellular immune responses (Thomsen and Marker, 1988). However, in certain situations, neutralizing antibodies have also been demonstrated to enhance LCMV-induced disease in the absence of primed cytotoxic T-cell responses (Battegay et al., 1993). Preexisting neutralizing antibodies showed little effect on local virus spread in peripheral tissues, but they reduced hernatogenic spread and infection of antigen-presenting cells, thus delaymg the generation of primary cytotoxic T-cell responses and indirectly modulating the extent of iminunopathology in peripheral organs. Thus, neutralizing antibody responses may modulate the CTL response in a beneficial or harmful way depending on (1)preexisting antibody titers, (2) the site of infection, and (3) the replicative kinetics and the cell tropism of the virus strain. In addition, studies show that CD4-mediated LCMV-specific antibody production might play an important role in the long-term control of LCMV infection. MHC class II-deficient mice as well as B cell-deficient mice were perfectly able to control acute LCMV infection; however, viral titers spontaneously reappeared within 2-3 months after infection and no CTL memory could be demonstrated (Thomsen et al., 1996).This indicates that long-term control of LCMV infection is dependent on all three major components of the immune system, whereas control of acute infection is primarily mediated by CD8+ effector T cells. As described for some vaccinia and influenza virus-specific CD4+ T cells clones in uitro, Muller et al. (1992) reported on the appearance of LCMVspecific CD4+ “cytotoxic” T cells on iininunization of P2M-deficient mice with LCMV. Although the cytolytic potential of LCMV-specific CD4’ T cells could be shown in vitro, the question of whether this cytolytic function
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
339
plays any role in vivo remains at least doubtful as, despite the presence of these cytolytic CD4' T cells, LCMV could not be efficiently controlled in p2M-deficient mice after infection nor in CD8-deficient mice (FungLeung et al., 1991; Muller et al., 1992). In addition, it is conceivable that these cytolytic CD4+ T cells are only generated and observable in mice that are permanently deficient of CD8' cytotoxic T cells and may thus be the result of the plasticity of the immune system to compensate for lacking certain effector functions, especially as cytolytic CD4t T cells were never detected on LCMV infection of normal mice. In addition, LCMV infection of MHC class II-restricted TCR transgenic mice with specificity for a LCMV GP epitope revealed that LCMV-specific CD4' T cells are incapable of controlling LCMV infection, even if present at extremely high frequencies (Oxenius et al., 1998). In contrast to being protective, it has been reported that CD4+ T cells can induce Fas-dependent pathogenesis after infection of P2M-deficient inice with LCMV. Transfer of immune splenocytes from LCMV-infected p2M-deficient mice into irradiated infected /EM-deficient mice resulted in death of the recipient animals. This demonstrated a role for CD4+ T-cell-mediated cytotoxicity in virus-induced immunopathology (Zajak et nl., 1996).
B. HUMAN IMMUNODEFICIENCY VIRUS The pathological changes of the immune system observed after HIV infection, especially the associated decline of CD4' T-cell numbers and loss of CD4+T-cell function, have offered direct evidence for the important role of CD4' T cells in successful host immune responses to viral (and other) infections in humans. HIV-specific CD4' T cells are usually difficult to recover from HIV-infected individuals. This is probably due to (1)the infection of large numbers of HIV-specific CD4' T cells in lymph nodes, as infectious HIV particles were shown to be trapped in the form of immune complexes on follicular dendritic cells and thus activation of HIVspecific Th cells by HIV-presenting APCs in the lymph node is likely to result in intimate contact with infectious virus (Pantaleo et al., 1994).These infected HIV-specific Th cells are targets for CTL-mediated lysis as well as being impaired in their function and survival as the half-life of an HIV-infected Th cell is about 1.5 days (Ho, 1995; Perelson et al., 1996). (2) Chronic (over)stimulation of HIV-specific CD4+ T cells in secondary lymphoid organs (due to the antigen reservoir in follicular structures) might lead to exhaustive activation a s has been described for CD8' T cells during chronic LCMV infection (Moskopllidis et al., 1993). In a first phase, HIVinfected individuals commonly exhibit a loss of HIV-specific CD4' T cell responses before a more general dysfunction of CD4' T cells is observed
340
ANNETTE OXENIUS et a1
(Clerici et al., 1994). This may contribute to the eventual inability of the host to control HIV infection in a manner comparable to the failure of CD4-deficient mice to control infection with intermediate doses of widely replicating LCMV-Docile (Battegay et aE.,1994).Thus, CD4' T cells might be important for the sustained maintenance of CTL activity over a long time period, Another important parameter possibly leading to the disappearance of CD4+ T cell responses, as well as finally to the disappearance of CD8' T-cell responses, might be the complete destruction of lymphoid tissue as disease progresses (Pantaleo et al., 1993). Some investigators suggested a critical role of the Thl/Th2 Th cell phenotype balance for the course of HIV infection. Thl-type Th cell responsiveness was primarily found in long-term nonprogressors and was correlated with an efficient cellular immune response. However, Th2-type Th cell responsiveness was suggested as contributing to the disease progression of AIDS (Clerici and Shearer, 1994). It has been demonstrated in vitro that Thl clones are less efficiently infected with HIV as compared to Th2 clones (Romagnani et al., 1994). Th2-type cytokines increase activation-induced programmed cell death in peripheral blood mononuclear cells (PBMCs) of HIV-infected individuals, whereas Thltype cytokines can prevent such activation-induced programmed cell death, which may also partly account for the decline of CD4' T cell counts at late stages of disease (Clerici and Shearer, 1994). In addition to contributions of CD4' T cells in the control of HIV infection, patients with reduced CD4' T cell counts andlor functions show enhanced susceptibility to opportunistic infections and tumors, demonstrating that CD4+ T-cell functions are critically involved in many viral, bacterial, mycobacterial, fungal, and protozoal infections. C. HEPATITIS B VIRUS
A variety of possible outcomes are associated with the infection of noncytopathic, hepatotropic hepatitis B virus. The infection may be acute, associated with hepatitis of a few weeks of duration, or persistent, either without disease or associated with chronic liver disease and hepatoma formation. The immune response to HBV-encoded antigens is responsible for both viral clearance and disease pathogenesis during HBV infection. The humoral response against viral env-antigens contributes to the elimination of circulating virus particles; the cellular response, which is directed against env-, nucleocapsid-, and pol-antigens, eliminates infected cells. Both MHC class I- and MHC class II-restricted responses are vigorous in acutely infected patients who clear the virus, but these responses are relatively weak in chronically infected patients (Chisari, 1995).
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
34 1
During acute infection, usually a strong antinucleocapsid antigen-specific Th cell response is observed and only a weak anti-env antigen-specific Th cell response is observed. Because anti-env responses can clearly be induced after vaccination with plasma-derived or recombinant HBsAg (Celis et al., 1984, 1988) and because anti-env Th cell responses may occur in preclinical incubation periods of disease (Vento et al., 1987), it has been suggested that HBV-env-specific Th cells may be exhaustively activated by high doses of virus and/or by high concentrations of circulating HBSAg that may occur if the early immune response is incapable of rapidly and completely eradicating infection. The nucleocapsid-specific Th cell response temporally coincides with viral clearance in patients with acute hepatitis and thus has been attributed a critical role in viral clearance (Ferrari et al., 1990), especially since chronically infected patients usually only show weak nucleocapsid-specific Th cell responses (Ferrari et al., 1990). The functional significance of a vigorous nucleocapsid-specific Th cell response may include a contribution for HBV-specific CTL activation (Missaleet al., 1993;Nayersinaet al., 1993;Penna et al., 1991) andintermolecular help for antibody production against HBV surface antigens (i.e., HBV-neutralizing antibodies) (Milich and McLachlan, 1986; Milich et al., 1987). Intrahepatic macrophages express MHC class I1 inolecules and can efficiently process and present HBV antigens. In chronic disease, mostly Thl-type Th cells are found in the liver (Barnaba et aZ., 1994), which are able to release inflammatory cytolanes to recruite antigen-nonspecific inflammatory cells that amplify the pathologic effect mediated by CD8' effector T cells. However, cytokines such as IFNy, TNFa, and IL-2 have been shown to act beneficially by noncytolytically inhibiting HBV gene expression and viral replication, thus interfering with the viral life cycle and possibly contributing to viral clearance without destruction of the infected cell (Chisari, 1995).
D. INFLUENZA Influenza virus is commonly used to study host-virus relationships. Influenza virus is a negative-stranded RNA virus whose genome consists of eight single-stranded segments that encode at least 10 polypeptides, including hemagglutinin (HA), neuraminidase (NA), nucleoprotein (NP), and matrix protein (Choppin and Compans, 1975). Both cellular and humoral immune responses are involved in the successful clearance of infection (Virelizier et al., 1979). HA-specific antibodies have been shown to neutralize influenza virus and play a major role for protection against subsequent infection (Virelizier, 1975).
342
ANNETTE OXENIUS et nl
However, cytotoxic T cells have been shown to be able to clear the virus during ongoing infection (Ennis et al., 1978; Yap et al., 1978). Both the humoral and, to a lesser extent, the cytotoxic T-cell response appear to be Th cell dependent (Askonas et al., 1981; Bums et al., 1975). Hence, the activation of influenza virus-specific Th cells may influence the magnitude and efficiency of the anti-influenza immune response. Concerning the prominent phenotype of Th cells induced after influenza infection, different observationswere reported: in situ mRNA hybridization studies during primary and secondary murine influenza pneumonia revealed that IFNy and TNFP were predominantly produced by CD8+ T cells, whereas IL-4 and IL-10 were predominantly produced by CD4+ T cells. Interestingly, the frequency of cytokine-producing lymphocytes early in mesenterial lymph nodes and late in the lung was much higher than the assumed precursor frequency of virus-specific effectors. If this can be generalized, induction of cytokine gene expression for T cells that are not responding directly to the invading pathogen may be a prominent feature of virus infections (Carding et al., 1993; Tough et al., 1996). However, focusing on the protective potential of influenza-specific Th cells, it has been reported that adoptive transfer of influenza-specific Th2 clones (in vitro noncytolytic) did not promote recovery from experimental influenza infection but rather exacerbated pulmonary pathology with concomitant eosinophilia. Adoptive transfer of influenza-specific Thl clones (in vitro cytolytic) was protective in vivo after virus challenge (Graham et al., 1994). A more indirect protective capacity of influenza-specific Th cell clones was reported by Scherle et al. (1992),who demonstrated that mice can recover from pulmonary influenza virus infection in the absence of class I-restricted CTLs. Intranasal exposure of T-cell-deficient n d n u mice resulted in persistent infection followed by death after 3-4 weeks. Adoptive transfer of CD4” T cells clones revealed reduced mortality and led to reduced or absent virus titers in the lungs. The Th cell response was antigen specific and no CTL response could be detected. Only influenza viruses containing the relevant Th cell epitope were cleared successfully, indicating that protection was specific. The transferred clonal Th cells did not, however, mediate protection directly but rather induced protective antibody responses as transfer of Th cell clones into SCID mice did not mediate protection (Scherle et al., 1992). Furthermore, functional analysis of influenza virus-specific Th cell clones in viwo showed that Th cells specific for internal viral proteins provided “cognate” help for B-cell responses to HA, thus supporting intermolecularhntrastructural T-B interaction in the antiinfluenza-specific antibody response (Scherle and Gerhard, 1986). Another approach in studying protective CD4+T-cell functions in influenza infection is the use of mice that are deficient in CD8+ CTLs (e.g.,
CD4’ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
343
P2M-deficient mice). CD8-depleted mice or P2M-deficient mice eliminated influenza virus from the respiratory tract, thus indicating that influenza infection can be terminated by CD4” effector cells, but simultaneous removal of both subsets led to a fatal outcome of the infection. However, the protective immune response may naturally be skewed to the development of CD8’ T-cell-mediated effector functions (Eichelberger et al., 1991). In addition, it was demonstrated that P2M-deficient mice exhibited delayed virus clearence in the case of less virulent influenza strains and increased mortality after challenge with more virulent influenza strains (Bender et al., 1992). Administration of anti-IFNy antibodies to P2M-deficient mice caused a further delay in virus clearance but no switch to Th2-type Th cell responses, supporting again the idea that CD4+ T cells can control infection of certain strains of influenza (at least less virulent strains) either directly or more probably via induction of neutralizing antibody responses (Sarawar et nl., 1994). In line with this, it was demonstrated that P2M-deficient mice can be protected against influenza A virus infection by vaccination with recombinant vaccinia viruses expressing either HA or N but not by recombinant vaccinia viruses expressing influenza virus internal proteins. These findings suggest again that CD4+ T cells are involved in protection against influenza infection via induction of antibodies recognizing viral surfice antigens (Epstein et nl., 1993). The fact that influenza virus generates serotypes may also suggest that the primary mechanism for protection from influenza infection is mediated by antibodies (Bachinann and Zinkernagel, 1996). Similar findings were observed after Sendai virus (a parainfluenza virus) infections. Analysis of the host response to Sendai virus infection in MHC class II-deficient mice revealed no impaired CD8’ CTLp generation (Hou et nl., 1995). Similar to influenza virus infections, delayed clearance of Sendai virus occurred in mice lacking MHC class I-restricted CD8” T cells (Hou et nl., 1992). Th cells might not be strictly required for the induction of influenzaspecific protective CTL responses (Tripp et al., 1995); nevertheless, it has been shown that influenza-specific CD4” T-cell clones enhance the generation of influenza-specific secondary CTL responses in vitro : restimulation of purified CD8’ T cells required “he1p”either by the presence of Th cells or by the addition of the corresponding Th cell supernatants, whereby T h l clones promoted significantly stronger memory CTL response inductions than Th2 clones. This could be correlated with IL-2 secretion being important for CD8’ T-cell restimulation (Pallaclinoet al., 1991). Taken together, conditions could be defined in influenza and parainfluenza virus infections by modifymg dose, route of infection, and/or virulence
344
ANNETTE OXENIUS et 01.
of the virus strain in which either CD4' (via induction of antibody responses) or CD8+ T-cell responses were sufficient to induce protective immunity and to promote resolution from acute infection (Kast et al., 1986; Lightman et al., 1987; Mackenzie et al., 1989).
E. MEASLES VIRUS Measles virus (MV) is a paramyxovirus and the causative agent of highly contagious acute measles and its complications. Acute measles is usually not a severe disease and is controlled by cell-mediated immunity. Complications of acute measles include (1)acute measles encephalitis, probably as a consequence of measles virus-induced autoimmunity against myelin antigens (Johnson and Griffin, 1986), and (2) chronic central nervous system (CNS) disorders, namely subacute sclerosing panencephalitis or measles inclusion body encephalitis, which develop on the basis of persistent MV infection of brain cells (Schneider-Schaulies and Liebert, 1991). Investigations of the mechanisms that may control MV infection revealed apparently contradictory results, depending on the nature of the host. In humans recovering from acute measles infection, stimulation of PBMC with autologous MV-infected lymphoblastoid cell lines resulted primarily in the expansion of MV-specific CD8' T cells, suggesting that the MHC class I-restricted CTL response may be an important factor in the control and elimination of MV infection (Van Binnendijk et al., 1990). However, experimental systems using either rats or mice revealed that CD4' T cells are essential in the control of measles virus infection. In v i m depletion of CD4+or CD8' T cells in the murine model for experimental acute measles encephditis demonstrated that removal of CD8+ T celIs in measles virus-resistant BALB/c mice did not interfere with virus clearance from the brain. In contrast, depletion of CD4' T cells rendered BALB/c mice susceptible to infection. Also, in measles virus-susceptible C3H mice, CD4' T cells played a role in recovery from measles infection but seemed less effective as CD4' T cells from resistant BALB/c mice (Finke and Liebert, 1994). Similar results were obtained when the mechanisms controlling MV infection in the rat brain were investigated. Adoptive transfer of individual MV-specific CD4' T-cell lines was sufficient to control an intracerebral MV infection without participation of CD8' T cells or neutralizing antibodies (Reich et al., 1992).
F. POXVIRUSES (ECTROMELIA AND VACCINIA) Host responses to poxvirus infection are varied and quite dependent on the host and virus species. For example, recovery from experimental ectromelia virus infection in the mouse is strongly dependent on the generation of ectromelia-specific CTL responses. However, vaccinia virus infec-
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
345
tion in humans or rhesus monkeys did not lead to classical primary CTL responses and thus cell-mediated cytotoxicity appears to be associated with antibody-dependent cellular cytotoxicity (ADCC) and NK cells (Buller and Palumbo, 1991). Early appearing nonspecific innate immune responses, including IFNy activation, inflammation, and NK cell activation, are beneficial but do not suffice to eliminate infection. Specific immunity, including antibody production, CTL activation, and DTH responses, is responsible for clearance of infection. Although antibody production does not appear to play a central role in controlling primary infection, it appears to be important in preventing reinfection (Buller and Palumbo, 1991). Vaccinia virus infection in mice leads to the generation of primary CTL responses that are effective in viral clearance (Blanden, 1974; Hirsch et al., 1968). Nonetheless, several situations, CD4+ T cells are also able to confer protection against vaccinia virus infection (Kundig et al., 1993a; Oxenius et al., 1998; Spriggs et al., 1992). A comparison of the antivaccinia protective capacity mediated by either CD4' or CD8' T cells revealed that CD8' effector T cells are more efficient in protection against vaccinia virus infection than CD4' effector T cells (Binder and Kundig, 1991). Primed cytotoxic T cells could protect against intracerebral challenge infection with lo3 to lo4 times more recombinant vaccinia virus particles as compared to primed CD4' T cells. Another study using recombinant vaccinia viruses expressing the VSV-derived G or N protein showed that protection against recombinant vaccinia virus challenge of VSV-primed mice could be mediated by either CD4' or CD8+ effector T cells, depending on the haplotype of the mouse strain and the recombinant protein and thus most probably depending on the availability of CD4' or CD8' T cell epitopes within the recombinantly expressed protein. In general, if CD8' T-cell epitopes were available, they seemed to be dominant for protection (Kundig et al., 1993a). in line with this, PZMdeficient mice were shown to survive infection with high doses of vaccinia virus, suggesting also a CD4' T-cell-mediated protective effector mechanism (Spriggs et al., 1992). In addition, MHC class 11-restricted TCR transgenic mice with specificity for the LCMV GP were protected against infection with recombinant vaccinia virus expressing the LCMV GP but not against infection with recombinant vaccinia virus expressing an irrelevant protein (Oxenius et al., 1998). A direct role of cytokines such as IFNy, TNFa, and IL-2 in recovery from vaccinia virus infection was demonstrated using recombinant vaccinia viruses expressing these cytokines at high levels at the site of infection. Immunodeficient mice recovered from infection with recombinant vaccinia virus expressing either IFNy or IL-2 (Karupiah et al., 1990; KohonenCorish et al., 1990).
346
ANNEITF: OXENIUS ef al
Another experimental approach providing evidence for a direct role of IFNy in protection against vaccinia virus infection investigated the antiviral relevance of soluble mediators that may operate in the vicinity of virusspecific effector T cells. VSV-immune mice were challenged with a mixture of VSV-N- and LCMV-GP-expressing vaccinia viruses. Protection was observed against the VSV-N-expressing vaccinia virus in ovaries or testes after systemic challenge (but not against the coadministered LCMV-GPexpressing vaccinia virus). However, protection against both recombinant vaccinia viruses was observed in the brain after intracerebral challenge, indicating that 'bystander' protection by soluble mediators secreted by either CD4+ or CD8+ immune effector T cells is functional in organs where their diffusion is limited (e.g., in the brain) (Kundig et al., 1993b). Additional evidence that vaccinia virus clearance in mice is not dependent on direct perforin-mediated cytotoxic effector mechanisms of CTLs came from infection of perforin-deficient mice with vaccinia virus. These mice were still able to control infection with vaccinia virus whereas they were unable to control LCMV infection (Kiigi et al., 1995). Infection with ectromelia virus causing generalized mousepox leads to the induction of a vigorous CTL response, also in the absence of CD4+ T cells (Buller et al., 1987). Nonetheless, analysis of the different roles of CD4' and CD8' T cell subsets in the control of generalized ectromelia infection revealed that infection in resistant C57BLJ6 and AKR mice was limited by CD8+ effector functions (0'Neil1 and Brenan, 1987),but CD4' T cells were found to be required for the induction of optimal CTL responses. Thus, CD4 deficiency was associated with incomplete virus clearance and nonfatal persistence of ectromelia virus at low levels in most organs and at high levels in the skin (Karupiah et al., 1996). In addition, IFNy has been attributed a central role in the recovery from mousepox infection: anti-IFNy treatment (and to a much lesser degree anti-IFNa or anti-IFNP treatment) resulted in virus persistence in all tissues tested except the primary site of infection, where virus clearance appears to be delayed. Anti-IFNy-treated mice finally succumbed to infection (Karupiah et al., 1993). G. VESICULARSTOMATITIS VIRUS VSV belongs to the family of rhabdoviridae and is a relative of rabies (Wagner, 1987). VSV replicates only abortively in mice unless it reaches the central nervous system where it causes lethal encephalitis (Wagner, 1987). VSV infection in mice induces a rapid Th-independent VSVneutralizing IgM response, which is followed by a largely Th-dependent VSV-neutralizing IgG response. VSV is not a polyclonal B-cell activator but a very potent activator of VSV-specific B cells due to the efficient
CD4’ T-CELL 1KDUCTIOK AND EFFECTOR FIJNCTIOKS
347
cross-linking capacity of highly repetitive organization of the VSV-G (representing the determinant recognized by VSV-neutralizing antibodies) on the virion surface (Bachinann and Zinkernagel, 1996). VSV-neutralizing antibodies are mainly of the IgG2a isotype (Coutelier et al., 1987) and have been shown to effectively protect mice from lethal intravenous infection (Lefrancois, 1984).Thus, induction of VSV-specific Th cell responses plays a crucial role in protection against VSV infection via induction of Ig classswitched neutralizing antibodies. CD4 deficiency or deficiency in CD40 or CD40L inolecules impairs Ig class switch and induction of B-cell memory (Battegay et al., 1996; Oxenius et al., 1996);B-cell deficiency results in a lethal outcome of VSV infection. Most probably due to the efficient T-independent activation of VSVspecific B cells, only limited T help involved in cognate T-B interaction is required to induce Ig class switch. Transgenic mice expressing a MHC class II-restricted TCR specific for a VSV-G epitope only show marginally enhanced kinetics of VSV-neutralizing antibody production on infection with VSV (C. Burkhart and K. Maloy, personal communication). In addition, VSV-specific Th cell priming prior to VSV infection resulted in a secondary type antibody response of ELISA-binding VSV-specific antibodies, but did not influence the kinetics of the protective VSV-neutralizing antibody response, suggesting that the “bottleneck’ of the neutralizing antibody response is on the B-cell side and not critically dependent on the kinetics of VSV-specific Th cell induction (Charan et al., 1986).
H. CYTOMEGALOVIRUS Cytoinegalovirus (CMV) belongs to the family of the herpesviruses, and infection can usually be controlled by an immunocompetent host. However, CMV infection of iininunocompromised hosts leads to a fatal outcome. For example, interstitial CMV pneumonia can cause mortality in irradiated, bone marrow-transplanted leukemia patients, and HIV-infected individuals often show manifestations of CMV disease as symptoms of AIDS (notably retinitis and colitis). CMV infection in iinrnunocompetent hosts generally leads to viral latency in leukocytes (Jordan, 1978). In mice, CMV infections serve as a model in investigating the contributions of different immune effector cell subsets in the resolution of acute infection. Adoptive cell transfer experiments demonstrated that CD8’ effector T cells are responsible for the control of acute MCMV infection, whereas CD4’ T cells were dispensable for this short-term protection (Reddehase et al., 1985, 1987, 1988). Importantly, transfer of CD8’ T effector T cells prevented the lethal bone inarrow aplasia that is caused by CMV-induced failure of hematopoietic stem cell generation (Mutter et al., 1988) and resulted in limited virus spread and tissue lesions (Reddehase et al., 1987).
348
A N N E l T E OXENIUS et (11.
With respect to long-term virus control, CD4+ T-cell depletion revealed delayed clearance of replicating virus in host tissues; nevertheless, protective CD8+ effector T cells were generated, which finally restricted virus replication to acinar glandular epithelial cells of the salivary glands (Jonjic et al., 1989). Thus, CD4' T cells seem to influence the quality of CTL responses to MCMV, which apparently has to be optimal for efficient clearance of MCMV from all host tissues. However, under circumstances where CD8' T cells were absent (CD8-deficient mice), recovery from CMV infection could be observed. Adoptive transfer studies revealed that the limitation of virus spread required the cooperation of CMV-primed CD4' T cells with other cells. Although the CD4' T-cell subset was essential for the antibody response in cell recipients, the capacity of the transferred cells to generate antiviral antibodies did not correlate with virus clearance. This compensatory protective capacity gained by CD4' T cells in CDSdeficient mice was absent in normal mice recovering from CMV infection (Jonjic et al., 1990). CMV infection of the lungs and subsequently developing interstitial pneumonia after intraperitoneal CMV infection was shown to be inhibited by adoptive transfer of CD4' T cells whereas this did not influence virus replication in the salivary glands, the preferential site of CMV infection in the mouse (Kadima-Nzuij and Craighead, 1990). I. HERPESSIMPLEXVIRUS Acute infection with herpes simplex virus (HSV) induces both cellular and humoral immune responses, which are important in the resolution of acute disease but which do not prevent latent infection generating a lifelong reservoir of HSV, which periodically replicates in selected tissues (e.g., HSV-1 latently infects neurons and can remain in a quiescent state for years). Because antibodies cannot penetrate cells, because infected neurons do not present HSV-1 glycoprotein-derived peptides on their surface, as they usually do not express MHC class I molecules, and because neurons are usually not accessible by naive T cells, CD8' T cells do not recognize and thus lyse HSV-harboring neurons. Actually, neither arm of the immune system is capable of protection from persistent infection. Control of acute HSV infection is generally mediated by CD8' T cells (Sethi et al., 1983),but conditions could be identified where CD4' T cells also play an important role for viral clearance (Nash et al., 1987). For example, mice deficient of CD4' T cells showed a markedly increased susceptibility to herpes simplex virus infection of the skin and the nervous system leading to zosteriform lesions, whereas P2M-deficient mice were as resistant to challenge as were immunocompetent mice of the same genetic background (Manickan and Rouse, 1995).Vaccination with recom-
CD4' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
349
binant vaccinia viruses expressing herpes simplex virus immediate-early proteins induced protective iininunity in BALB/c mice against herpes simplex virus challenge through CD4' T cells of the T h l phenotype without the involvement of CD8' effector T cells (Manickan et al., 1995).
J. PICORNAVIRUSES (POLIOVIRUSAND TIIEILER.~ VIRUS) Poliomyelitis is an acute paralytic disease caused by the infection of the central nervous system by virulent strains of poliovirus. Induction of virusspecific neutralizing antibodies has been considered the major mechanism of protection against poliovirus infection. At least four neutralizing antibody epitopes have been identified and located on the three-dimensional structure of the virus particle (Hogle et al., 1985). If immunity against poliovirus is mediated by antibody responses and, in particular, if secretory IgA in the gut plays a role in protection, one might have predicted that poliovirusspecific CD4+ T cells of the Th2 phenotype would be involved in the protective mechanism. However, using transgenic mice bearing the human poliovirus receptor (PVR) (Koike et al., 1991; Ren and Racaniello, 1992) and adoptive transfer of poliovirus-specific Th cell clones, Mahon et al. (1992) demonstrated that the predominant cytokines secreted by poliovirus-stimulated spleen cells were IL-2 and IFNy, which are characteristic of a Thl phenotype. Although poliovirus-specific T h l clones displayed cytotoxic activity in vitro, adoptive transfer experiments demonstrated that poliovirus-specific Th cells (specific for virus internal proteins) could confer protective immunity in oivo through the stimulation of neutralizing antibody production (Mahon et al., 1995). Theiler's murine encephalomyelitis virus (TMEV) is a picornavirus that causes biphasic central nervous system dsease on intracranial inoculation into susceptible mouse strains (Lehrich et al., 1976; Lipton, 1995). The early phase is an acute, encephalitic disease occurring in the first month after infection and it results from lytic virus replication predominantly within neurons of the CNS. Surviving animals usually develop a chronic progressive deinyelinating disease of the CNS (Friedmann and Lorch, 1985).The mechanisms involved in the development of demyelating lesions are complex and both direct viral damage and T-cell-mediated immunopathology (predominantly by CD4' T cells of the Thl phenotype) are involved (Lipton and Dal Canto, 1976; Welsh et al., 1987). Examination of the role of CD4' and CD8+ T cells in protection from TMEV revealed that while CD8+ T cells do make some contribution to protection at early time points after infection, the CD4+ T-cell subset is of major importance as CD4+ T-cell-depleted mice die within 3-5 weeks after TMEV infection. The key protective role of CD4' T cells in TMEV-infected mice was shown to be the ability to provide help to B cells for antibody production at early
350
ANNETTE OXENIUS et nl
time points after infection in addition to other CD4’ T-cell-dependent mechanisms that contribute to the control of virus replication and are thus of importance in determining the levels of virus subsequently persisting in the CNS, which in turn influences the severity of the chronic deinyelinating disease (Borrow et al., 1993).
K. MOUSEMAMMARY TUMOR Viws Mouse mammary tumor virus (MMTV) is a milk-transmitted type B retrovirus that encodes a superantigen in its 3‘ long terminal repeat. MMTV infection is correlated with a strong superantigen-induced Vp elementspecific CD4’ T-cell response. Different MMTV viruses have been described that exhibit superantigen activity for different Vp elements on CD4’ T cells, thus 5-40% of the CD4’ T-cell repertoire can possibly react with the encoded superantigen of a given MMTV strain (Acha-Orbea and MacDonald, 1995). The strong superantigen-induced CD4’ T-cell expansion can be measured within 3-4 days after infection and peaks 6 days after infection (Waanders et al., 1993; Webb et al., 1990). A few days after this initial expansion, unresponsiveness of the superantigen-reactive T cells is observed (Rammensee et al., 1989; Webb et al., 1990) and finally they disappear from the repertoire. This deletion is life-long and occurs in the presence or absence of the thymus (Acha-Orbea and MacDonald, 1995). Parallel to the expansion of superantigen-reactive CD4+ T cells, a strong polyclonal B-cell activation leads to cell proliferation and differentiation into Ig-secreting, antibody-forming cells (Held et al., 1993a).An antiviral antibody response can be detected (predominantly of the IgG2a and IgG2b isotypes), but these MMTV-specific antibodies do not seem to drastically interfere with an established infection (Acha-Orbea and MacDonald, 1995). It has been repeatedly suggested that B ceIls are preferentially infected by MMTVwith concomitant integration of the viral DNA into the B-cell genome and thus are the prominent APCs during superantigeninduced T-cell activation. This has also been proposed to be a cause for the observed superantigen-reactive CD4+ T-cell unresponsiveness (Eynon and Parker, 1992; Fuchs and Matzinger, 1992). However, the tremendous activation of superantigen-reactive CD4+ T cells leads to “cognate” help for MMTV-infected B cells, which finally induces the observed polyclonal B-cell activation and expansion (Held et al., 1993b). This polyclonal B-cell activation is clearly dependent on the presence and activation of superantigen-reactive CD4’ T cells and, vice versa, no productive infection in the absence of superantigen-reactive CD4’ cells was observed (Held et al., 1993b). Although MMTV seems to be a T-independent, B-cell activator, MMTV-infected B cells have a short survival time in the absence of additional factors or functions provided by superantigen-reactive CD4’ T cells
CD.1' T-CELL INDUCTION AND EFFECTOR FUNCTIONS
351
( Acha-Orbea and MacDonald, 1995). After stable MMTV infection has been achieved, infection ultimatively spreads to the mammary gland and leads to virus transmission via the milk. Thus, in contrast to many viruses that aim to evade efficient immune responses for their survival, MMTV deliberately uses the host immune response for its own purposes. In particular, the superantigen-induced activation and deletion of Vpx-expressing CD4' T cells and the expansion of infected B cells is an instrumental step for successful persistent infection and stable integration of the viral genome.
IV. Conclusions
The coevolution of pathogens and the immune system in their respective hosts has established a subtle balance between the parasitic usurpation of the host cellular functions by the virus with the aim to survive and, most importantly, to be transmitted, whereas the defense mechanisms of the host aim at resolving infection without causing fatal immunopathology. Both sides have developed multiple mechanisms to achieve their respective goals, which have become increasingly apparent over the recent years. Focusing on the role of virus-specific CD4+ effector T cells in viral infections, it might not be surprising that generalizations are almost impossible to make. Depending on the characteristics of a virus (cytopathogenicity, cell tropism, virulence, persistence, etc.), CD4' T cells may play a more or less crucial role in the resolution of infection. Generally, acute infections with noncytopathic viruses such as LCMV or HBV are controlled by cytotoxic CD8' effector T cells, whereas infections with cytopathic viruses may be controlled either by virus-neutralizing antibodies (e.g., VSV, influenza, and polio) or by variable contributions of CD4' T cells, antibodies, and CD8' T cells (e.g., vaccinia virus, ectromelia, MCMV, and HSV). In the latter cases, the virus dose, the route of infection, and the virulence of the virus strain determines, at least partly, whether CD4' T cells (and/or antibodies) can be sufficient for viral clearence. The diverse effector mechanisms of Th cells (secretion of cytokines, cognate help for B cells, help for CTL induction, DTH reaction) are of different importance in different viral infections. Viruses that are sensitive to antivirally active cytokines (e.g., vaccinia) may be entirely controlled by T-cell-dependent cytokine secretion. Cognate help for antiviral antibody production is not only important for the resolution of primary infections of some cytopathic viruses and especially of those viruses hidden in iminunoprivileged sites, but also for the induction of memory antibody titers, which represent the most potent protection against reinfection.
352
ANNETTE OXENIUS et al.
Usually the generation of cytotoxic T-cell responses on acute infection is not, or only partly, dependent on CD4' T-cell functions. However, the situation may be different for the long-term control of infections with viruses that have the tendency to persist, even at levels below the threshold of detection by normal assays. Optimal CD4' T-cell-dependent induction of CTL activity, sustainment of prolonged CTL activity, optimal memory CTL function, and/or the induction of (neutralizing) antibody responses may be crucial for the long-term protection against many virus infections. Viral infections generally induce Th cells exhibiting a T h l phenotype, which contribute to the induction of cellular immune responses that are of crucial importance in the defense against intracellular pathogens. The mechanism(s) governing this skewing of the Th cell phenotype is still unclear, but it may depend on the nature of viruses to induce innate immune responses dominated by IFNy secretion, NK cell, and macrophage activation. The APC requirements for Th cell induction on viral infections are not as clearly defined as for soluble antigens but may depend on the viral host cell tropism, replication kinetics, and the cytopathic or noncytopathic nature of the virus. Some viruses exhibit additional pathways for antigen presentation on MHC class I1 molecules whereby intracellularly synthesized viral antigens are able to directly load MHC class I1 molecules. Taken together, although virus-specific CD4' T cells have been shown to promote recovery from virus infections in some selected infectious systems,virus-specific cytotoxic CD8+effector T cells are apparently crucial for the control of acute infections with noncytopathic viruses and important for the resolution of several infections by highly pathogenic (or virulent) cytopathic viruses. However, in the latter case, neutralizing antibody responses or T-cell-secreted cytokines may additionally be of crucial relevance for clearance of the virus. In general, the cooperations of the different arms of the immune system, the CD4' T cells representing one central regulator of these arms particularly important for generation of protective antiviral antibody responses, offer the optimal defense against viral infections, and such optimal cooperation seems to become particularly important in persisting viral infections.
ACKNOWLEDGMENTS Annette Oxenius thanks Martin F. Bachmann and Paul Kleneiman for the critical discussions and the many helpful suggestions for this review.
REFERENCES Abbas, A. K., Murphy, K. M., and Sher, A. (1996).Functional diversity of helper T lymphocytes. Nature 383, 787-793.
CD4+ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
353
Acha-Orbea, H., and MacDonald, H. R. (1995). Superantigens of mouse mammary tumor virus. Annu. Reo. lmmunol. 13,459-486. Ahmed, R., Butler, L. D., and Bhatti, L. (1988).T4+T helper cell function in vivo: Differential requirement for induction of antiviral cytotoxic T-cell and antibody responses. J. Virol. 62,2102-2106. Allan, W., Tabi, Z., Cleary, A., and Doherty, P. C. (1990). Cellular events in the lymph node and lung of mice with influenza. J. Imnwnol. 144, 3980-3986. Amigorena, S., Drake, J. R., Webster, P., and Mellman, I. (1994). Transient accumulation of new class I1 MHC molecules in a novel endocytic compartment in B lymphocytes. Nature 369, 113-120. Ando, K., Moriyama, T., Guidotti, L. G., Wirth, S., Schreiber, R. D., Schlicht, H. J , , Huang, S., and Chisari, F. V. (1993). Mechanisms of class I restricted immunopathology: A transgenic mouse model of fulminant hepatitis. J. Exp. Med. 178, 1541-1554. Askonas, B. A., Mullbacher, A., and Ashman, R. B. (1981). Cytotoxic T memory cells in virus infection and the specificity of helper T cells. Immunology 45, 79. 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-gamma and in interleuhn-4-expressing cells. Eur. 1.Imnmunol. 24,1097-1101. Austyn, J. M. (1996). New insights into the mobilization and phagocytic activity of dendritic cells. /. Exp. Med. 183, 1287-1292. Bachmann, M. F., and Zinkernagel, R. M. (1996). The influence of virus structure on antibody responses and virus serotype formation. Immunol. Today 17, 553-558. Bachmann, M. F., and Zinkernagel, R. M. (1997). Neutralizing antiviral B cell responses. Annu. Rev. Irnmunol. 15, 235-270. Bakke, O., and Dobberstein, B. (1990). MHC class 11-associated invariant chain contains a sorting signal for endosomal compartments. Cell 63, 707-716. Banchereau, J., Bazan, F., Blanchard, D., BriBre, F., Galizzi, J. P., van Kooten, C., Liu, Y. J., Rousset, F., and Saeland, S. (1994). The CD40 antigen and its ligand. Annu. Reo. Immunol. 12,881-922. Barnaba, V., Franco, A., Paroli, M., Benvenuto, R., De, P. G., Burgio, V. L., Santilio, I., Balsano, C., Bonavita, M. S., and Capelli, G. (1994). Selective expansion of cytotoxic T lymphocytes with a CD4+CD56+surface phenotype and a T helper type 1 profile of cytokine secretion in the liver of patients chronically infected with Hepatitis B virus. J . Iminunol. 152,3074-3087. Battegay, M., Bachmann, M. F., Burkhart, C., Viville, S., Benoist, C., Mathis, D., Hengartner, H., and Zinkernagel, R. M. (1996). Antiviral immune response of mice lacking MHC class I1 or its associated invariant chain. Cell. Iwmunol. 167, 115-121. Battegay, M., Kyburz, D., Hengartner, H., and Zinkemagel, R. M. (1993). Enhancement of disease by neutralizing antiviral antibodies in the absence of primed antiviral cytotoxic T cells. Eur. J , Immunol. 23, 3236-3241. Battegay, M., Moskophidis, D., Rahemtulla, A,, Hengartner, H., Mak, T. W., and Zinkernagel, R. M. (1994). Enhanced establisliment of a virus carrier state in adult CD4+ T cell deficient mice. J . Virol. 68, 4700-4704. Battegay, M., Moskophidis, D., Waldner, H., Briindler, M. A,, Fung, Leung, W. P., Mak, T. W., Hengartner, H., and Zinkemagel, R. M. (1993). Impairment and delay of neutralizing antiviral antibody responses by v i m specific cytotoxic T cell. J. Immunol. 151,54085415. Bender, B. S., Croghan, T., Zhang, L., and Small, P. A. (1992). Transgenic mice lacking class I major histocompatibility complex-restricted T cells have delayed viral clearance and increased mortality after influenza virus challenge. J. Exp. Med. 175, 1143-1145.
354
ANNETTE OXENIUS et nl
Bikoff, E. K. (1991). Class 11-restricted IgC2ab-specific T cells recognize a signal-minus form of the V-CH3b antigen. E w . J. linmunol. 21, 1411-1417. Bikoff, E. K. (1992). Formation of complexes between self-peptides and MHC class I1 molecules in cells defective for presentation of exogenous protein antigens. J. Imniunol. 149, 1-8. Bikoff, E. K., Huang, L. Y., Episkopou, V., van Meenvikj, J., Germain, R. N., and Robertson, E. J. (1995). Defective major histocompatibility complex class I1 assembly, transport, peptide acquisition, and CD4' T cell selection in inice lacking invariant chain expression. J. Exp. Med. 177, 1699-1712. Binder, D., and Kiindig, T. M. (1991). Antiviral protection by CDBt versus CD4' T cells: CD8+ T cells correlating with cytotoxic activity in vitro are more efficient in antivaccinia virus protection than CD4-dependent interleukins. J . Irnmunol. 146, 4301-4307. Blanden, R. V. (1974). T cell response to viral and bacterial infection. Transplant. Rev. 19, 56-84. Bodmer, H.. Viville, S., Benoist, C., and Mathis, D. (1994). Diversity of endogenous epitopes bound to MHC class I1 molecules limited by invariant chain. Science 263, 1284-1286. Boom, W. H., Liebster, L., Abbas, A. K., and Titus, R. A. (1990). Patterns of cytokine secretion in murine leishmaniasis: Correlation with disease progression or resolution. Infect. Inimun. 58, 3863-3870. Borrow, P., Tishon, A,, Lee, S., Xu, J., Grewal, I. S., Oldstone, M. B. A., and Flavell, R. A. (1996). CD40L-deficient mice show deficits in antiviral immunity and have impaired memory CD8+ CTL response. J. Exp. Med. 183, 2129-2142. Borrow, P., Welsh, C. J. R., a i d Nash, A. A. (1993). Study ofthe mechanisms by which CD4+ T cell contribute to protection in Theiler's inurine encephalomyelitis. ~ n i m ~ n o ~ o g y 80, 502-506. Bottomly, K., Janeway, C. A,, Jr., Mathieson, 8. J., and Mosier, D. E. (1980). Absence of an antigen-specific helper T cell required for the expression of the T15 idiotype in mice treated with anti-y antibody. Em-. J . Intmunol. 10, 159. Browning, M. J., Huang, A. S., and Reiss, C. S. (1990). Cytolytx T lymphocytes from the BALBh-H-2dm2 mouse recognize the vesicular stomatitis virus glycoprotein and are restricted by class I1 MHC antigens. J. lnimunol. 145, 985-994. Buchmeier, M. J., Welsh, R. M., Dutko, F. J., and Oldstone, M. B. A. (1980). The virology and immunobiology of lyinpliocytic choriomeningitis virus infection. Adu. Imniunol. 30,275-331. Buller, R. M., Holmes, K. L., Hugin, A,, Frederickson, T. N., and Morse, H. C., 111 (1987). Induction of cytotoxic T-cell responses in vivo in the absence of CD4 helper cells. Nature 328, 77-79. Buller, R. M., and Palumbo, G. J. (1991).Poxvirus pathogenesis. Microbiol. Rezj. 55,80-122. Burns, W. H., Billups, L. C., and Notkns, A. L. (1975). Thymus dependence of viral antigens. Nature 256, 654-656. Campbell, K. A,, Ovendale, P. J., Kennedy, M. K., Fanslow, W. C., Reed, S. G., and Maliszewski, C. R. (1996). CD40 ligand is required for protective cell-mediated immunity to Leishinunia nmjor. Iivimunittj 4, 283-289. Carding, S. R., Allan, W., McMickle, A,, and Doherty, P. C. (1993). Activation of cytokine genes in T cells during primary and secondary munne influenza pneumonia. J. E x i . Med. 177,475-482. Cassel, D. J., and Schwartz, R. H. (1994). A quantitative analysis of antigen-presenting cell function: Activated B cells stimulate naive CD4 T cells but are inferior to dendritic cells in providing costimulation. J . Exp. Med. 180, 1829-1840.
CD4’ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
355
Castellino, F., and Gerniain, R. N. (1995). Extensive trafficking of MHC class 11-invariant chain complexes in the endocytic pathway aid appearance of peptide-loaded class I1 in niultiple compartments. Immunity 2, 73-88. Celis, E., Kung, P. C., and Chang, T. W. (1984). Hepatitis B virus-reactive human T lymphocyte clones: Antigen specificity and helper function for antibody synthesis. J. Initnztnol. 132, 1511-1516. Celis, E., Ou, D., and Otvos, L. (1988). Recognition of hepatitis B virus surface antigen by human T lymphocytes: Proliferative and cytotoxic responses to a minor antigenic determinant defined by synthetic peptides. J . Zmniunol. 140, 1808-1815. Celis, E., and Saibara, T. (1992). Binding of T cell receptor to major histocoinpatibility complex class 11-peptide complexes at the single-cell level results in the indnction of antigen unresponsiveness (anergy). Etrr. J. Immunol. 22, 3127-3134. Charan, S., Hengartner, H., and Zinkemagel. R. M. (1986). Primary antibody responses to a well-defined and unique hapten are not enhanced by preiininunization with carrier: Analysis in a vird model. Proc. N d . Acad. Sci. USA 83, 2604-2608. Chen, B. P., Madrigal, A., and Parham, P. (1990). Cytotoxic T cell recognition of an endogenous class I HLA peptide presented by a class 11 HLA molecule. J. Exp. Med. 172, 779-788. Chen, W . F., and Zlotnik, A. (1991). ILIO: A novel cytotoxic T cell differentiation factor. J. Zmmunol. 147, 528-534. Cher, D. J., and Mosmann, T. R. (1987). Two types of murine helper T cell clone. 11. Delayed-type hypersensitivity is mediated by TH1 clones. J. Znimitnol. 138,3688-3694. Chisari, F. V. ( 1995). Hepatitis B virus iininunopatliogeiiesis. Annrr. Rer;. Zmmiinol. 13, 29-60, Choppin, P. W., and Coinpans, R. W. (1975). The structure of influenza vinis. “Influenza Virus and Influenza” (E. D. Kilboume, ed.), pp. 15-51. Academic Press, New York. Ciechanover, A. (1994). The ubiquitin-proteasome proteolyhc pathway. Cell 79, 13-21. Clerici, M., Hakim, F. T., Venzon, D. J., Blatt, S., Hendrix, C. W., Wynn, T. A., and Shearer, G. M. (1993). Changes in interleukin-2 and interleukin-4 production in asymptomatic, human immunodeficiency virus-seropositive individuals.J. Clin. Inoest. 91, 759-765. Clerici, M., and Shearer, G. M. (1994). The Thl-Th2 hypothesis of HIV infection: New insights. ZminunoE. Today 15, 575-581. Clerici, M., Wynn, T. A,, Berzofsky, J. A., Blatt, S. P., Hendrix, C. W., Slier, A,, Coffman, R. L., and Shearer, G. M. (1994). Role of interleukin-10 in T helper cell dysfunction in asymptomatic individuals with the human iinmunodeficiency vinis. J. Clinic. huest. 93, 768-775. Constant, S . , Pfeiffer, C., Woodard, A,, Pasqualini, T., and Bottomly, K. (1995a). Extent of T cell receptor ligation can determine the functional differentiation of naive CD4’ T cells. J. Exp. Med. 182, 1591-1596. Constant, S., Sant’Angelo,D., Pasqualini, T., Taylor, T., Levin, D., Flavell, R., and Bottomly, K. (199513).Peptide and protein antigens require distinct antigen-presenting cell subsets for the priming of CD4+ T cells. J. Iriimunol. 154, 4915-4923. B lymphocytes Constant, S., Schweitzer, N., West, J., Ranney, P., and Bottomly, K. (1995~). can be competent antigen-presenting cells for priming CD4’ T cells to protein antigens in vivo. J. Zmmiinol. 155, 3734-3741. Coutelier, J. P., Van der h g t , J. T. M.. Heessen, F. W. A., Waniier, G., and Van Snick, J. V. (1987). IgG2a restriction of murine antibodies elicited by viral infections. 1. Exp. Med. 165, 64-69. Cresswell, P. (1992).Chemistry and functional role of the invariant chain. C u r . Biol. 4,87.
356
ANNElTE OXENIUS et
d.
Critchfield, J. M., Racke, M. K., Zuniga, P. J., Cannella, B., Raine, C. S., Goverman, J., and Lenardo, M. J. (1994). T cell deletion in high antigen dose therapy of autoimmune encephalomyelitis. Science 263, 1139-1 143. Denzin, L. K., and Cresswell, P. (1995). HLA-DM induces CLIP dissociation from MHC class I1 alpha beta dimers and facilitates peptide loading. Cell 82, 155-165. Dubey, C., Croft, M., and Swain, S. L. (1995). Costimulatory requirements of naive CD4+ T cells. J. Immunol. 155, 45-57. Duncan, D. D., and Swain, S. L. (1994). Role of antigen-presenting cells in the polarized development of helper T cell subsets: Evidence for differential cytokine production by Tho cells in response to antigen presentation by B cells and macrophtages. Eur. ].lm.mwnol. 24,2506-2514. Eichelberger, M., Allan, W., Zijlstra, M., Jaenisch, R., and Doherty, P. C. (1991).Clearance of influenza virus respiratory infection in mice lacking class I major histocompatibility complex-restricted CD8+ T cells. J. Exp. Med. 174, 875-880. Ellis, J., Chain, B. M., Davies, D. H., Ibrahim, M. A., Katz, D. R., Kaye, P. M., and Lightstone, E. (1991).Antigen presentation by dendritic cells provides optimal stimulation for the production of interleukin (IL)2, IL4 and interferon-y by allogeneic T cells. Eur. J. lmmunol. 21, 2803-2809. Else, K. J., Finkelman, F. D., Maliszewski, C. R., and Grencis, R. K. (1994). Cytokinemediated regulation of chronic intestinal helminth infection. J. Exp. Med. 179,347-351. Ennis, F. A., Wells, M. A,, Butchko, G. M., and Albrecht, P. (1978).Evidence that cytotoxicT cells are part of the host’s response to influenza pneumonia.]. Exp. Med. 148,1241-1250. Epstein, M. M., Di Rosa, F., Jankovic, D., Sher, A,, and Matzinger, P. (1995). Successful T cell priming in B cell-deficient mice. J. Exp. Med. 182, 915-922. Epstein, S. L., Misplon, J. A., Lawson, C. M., Subbarao, E. K., Connors, M., and Murphy, B. R. (1993).P2-Microglobulin-deficientmice can be protected against influenza A infection by vaccination with vaccinia-influenza recombinants expressing hemagglutinin and neuraminidase. J. Immunol. 150, 5484-5493. Eynon, E. E., and Parker, D. C. (1992). Small B cells as antigen-presenting cells in the induction of tolerance to soluble protein antigens. J. Exp. Med. 175, 131-138. Farrar, J. J., Benjamin, W. R., Hilfiker, M. L., Howard, M., Farrar, W. L., and FullerFarrar, J. (1982). The biochemistry, biology, and role of interleukin 2 in the induction of cytotoxic T cell and antibody-forming B cell responses. Immunol. Rev. 63, 129-166. Ferrari, C., Penna, A., Bertoletti, A., Valli, A., Antoni, A. D., Giuberti, T., C a d i , A,, Petit, M. A., and Fiaccadori, F. (1990).Cellular immune response to hepatitis B virus-encoded antigens in acute and chronic hepatitis B virus infection. J Immunol. 145, 3442-3449. Fields, B. N. (1990). “Fields Virology,” Vol. 2, pp. 1-1267. Raven Press, New York. Finke, D., and Liebert, U. G. (1994).CD4+ T cells are essential in overcoming experimental murine measles encephalitis. Immunology 83, 184-189. Fleischer, B. (1984). Acquisition of specific cytotoxic activity by human T4+T lymphocytes in culture. Nature 308, 365-367. Foy, T. M., Shepherd, D. M., Dune, F. H., Aruffo, A., Ledbetter, J. A,, and Noelle, R. J. (1993). In vivo CD40-gp39 interactions are essential for thymus-dependent humoral immunity. 11. Prolonged suppression of the humoral immune response by an antibody to the ligand for CD40, gp39. J. Exp. Med. 178, 1567-1575. Friedmann, A., and Lorch, Y. (1985).Theiler’svirus infection: A model for multiple sclerosis. Prog. Med. Virol. 31, 43. Fuchs, E. J., and Matzinger, P. (1992). B cells turn off‘virgin but not memory T cells. Science 258, 1156-1159.
CD4+ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
357
Fung-Leung, W. P., Kiindig, T. M., Zinkernagel, R. M., and Mak, T. W. (1991). Immune response against Iynphocytic choriomeningitisvims infectiou in mice without CD8 expression.]. Exp. Med. 174, 1425-1429. Fung-Leung, W. P., Surh, C. D., Liljedahl, M., Pang, J., Leturcy, D., Peterson, P. A,, Webb, S., and Karlsson, L. (1996). Antigen presentation and T cell development in H2M-deficient mice. Science 271, 1278-1281. Gerety, S. J., Rundell, M. K., DdCanto, M. C., and Miller, S. D. (1994). Class 11-restricted T cell responses in Theiler’s murine encephalomyelitisvirus-induced demyelating disease. VI. Potentiation of demyelination with and characterization of an immunopathologic CD4+T cell line specific for an immunodominant VP2 epitope.]. Immunol. 152,919-929. Germain, R. N., and Margulies, D. H. (1993). The biochemistiy and cell biology of antigen processing and presentation. Annu. Reo. Inurucnol. 11, 403-450. Ghosh, P., Amaya, M., Mellins, E., and Wiley, D. C. (1995).The structure ofan intermediate in class I1 MHC maturation: CLIP bound to HLA-DR3. Nature 378, 457-462. Gilbert, K. M., and Weigle, W. 0. (1994). Tolerogenicity of resting and activated B cells. 1.Exp. Med. 179,249-258. Gotoff, S. P. (1968). Lymphocytes in congenital innnunological deficiency diseases. Clin. Exp. Inimunol. 3, 843. Graham, M. B., Braciale, V. L., and Braciale, T. J. (1994). Influenza virus-specific CD4+ T helper type 2 T lymphocytes do not promote recovery froin experimental virus infection. ]. Exp. Med. 180, 1273-1282. Grewal, I. S., Foellmer, H. G., Grewal, K. D., Xu, J., Hardottir, F., Baron, J. L., Janeway, C. A., Jr., and Flavell, R. A. (1996). Requirement for CD40 ligand in costimulation induction, T cell activation, and experimental allergic encephalomyelitis. Science 273, 1864-1867. Grewal, I. S., Xu, J., and Flavell, R. A. (1995).Impairment of antigen-specificT-cell priming in mice lacking CD40 ligand. Nature 378, 617-620. Gu&y, J.C., Ria, F., and Adorini, L. (1996). Dendritic cells bnt not B cells predent antigenic complexes to class 11-restrictedT cells after immunization of protein in adjuvant. ]. Exp. Med. 183, 751-757. Handa, K.. Suzuki, R., Matsui, H., Shirnizu, Y., and Kumagai, K. (1983). Natural killer (NK) cells as responder to interleukin 2 (IL-2):11. IL-%induced interferon gamma production. 1. Immunol. 130, 988-992. Harding, F. A,, McArthur, J. G.. Gross, J. A,, Raulet, D. H., and Allison, J. P. (1992). CD28mediated signalling co-stimulates murine T cells and prevents induction of anergy in T-cell clones. Nature 356, 607-609. Heinzel, F. P., Sadick, M. D., Holaday, B. J., Coffinan, R. L., and Locksley, R. M. (1989). Reciprocal expression of interferon y or interleukin 4 during the resolution or progression of murine leishmaniasis: Evidence for expansion of distinct helper T cell subsets. 1.Exp. Med. 169,59-72. Held, W., Shakhov, A. N., Izui, S., Waanders, C . A,, Scarpellino, L., MacDonald, H. R., and Acha-Orbea, H. (19934. Superantigen-reactive CD4+T cells are required to stimulate B cells after infection with mouse mammary tumor virus. ]. Exp. Med. 177, 359-366. Held, W., Waanders, G. A., Shakhov, A. N., Scarpellino,L., Acha-Orbea, H., and MacDonald, H. R. (1993b). Superantigen-induced immune stimulation amplifies mouse mammary tumor virus infection and allows virus transmission. Cell 74, 184-190. Hirsch, M. S., Nahmias, A. J., Murphy, F. A,, and Kramer, J. H. (1968). Cellular immunity in vaccinia infection of mice: Anti-thymocyte serum effects on primary and secondary responsiveness. 1.Exp. Med. 128, 121-131.
358
ANNETTE OXENIUS et al
Ho, D. D. (199535).Rapid turnover ofplasmavirions and CD4 lymphocytes in HIV-1 infection. Nature 373, 123-126. Hogle, J. M., Chow, M., and Filman, D. J. (1985). The three dimensional structure of poliovirus at 2.9 A resolution. Science 229, 1385-1365. Hosken, N. A,, Shibuya. K., Heath, A. W., Murphy, K. M., and O’Garra, A. (1995). The effect of antigen dose on CD4+T helper cell phenotype development in a T cell receptor&transgenic model. J. Exp. Med. 182, 1579-1584. Hou, S., Doherty, P. C., Zijlstra, M., Jaenisch, R., and Katz, J. M. (1992). Delayed clearance of Sendai virus in mice lacking class I MHC-restricted CD8’ T cells. J. Zmiriicnol. 149, 1319-1325. Hou, S., Mo, X. Y., Hyland, L., and Doherty, P. C. (1995). Host response to sendai virus in mice lacking class I1 major histocompatibility complex glycoproteins. J . Virol. 69, 1429-1434. Hsieh, C. S., Macatonia, S. E., O’Garra, A,, and Murphy, K. (1995).T cell genetic background determines default T helper cell phenotype development in vitro. 1. Exp. Med. 181, 713-721. 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. Inaba, K., Metlay, J. P., Crowley, M. T., and Steinmann, R. M. (1990). Dendritic cells pulsed with protein antigens in vitro can prime antigen-specific, MHC-restricted T cells in s i b . J. Exp. Med. 175, 15. Inaba, K., and Steinmann, R. M. (1985). Protein-specific helper T-lymphocyte formation initiated by dendritic cells. Science 229, 475-479. Janeway,C. A,, Jr. (1989).Approachingthe asymptote? Evolution and revolution in immunology. Cold Spring Harbor Lab. Press 0, 1-13. Janeway, C., and Bottomly, K. (1994). Signals and signs for lymphocyte responses. Cell 76,275-285. Janeway, C. A,, Ron, Y., and Katz, K. E. (1987).The B cell is the initiating antigen-presenting cells in the peripheral lymph nodes. 1. Zmmunol. 138, 1051-1055. Jaraquemada, D., Marti, M., and Long, E. 0. (1990). An endogenous processing pathway in vaccinia virus-infected cells for presentation of cytoplasmicantigens to class 11-restricted T cells. J . Exp. Med. 172, 947-954. Jenkins, M. K. (1994). The ups and downs of T cell costimulation. Zmmunity 1, 443-446. Jin, Y., Shih, W. K., and Berkower, I. (1988). Human T cell response to the surface antigen of hepatitis B virus (HBsAg): Endosomal and nonendosomal processing pathways are accessible to both endogenous and exogenous antigen. J . Exp. Med. 168, 293-306. Johnson, R., and Griffin, D. E. (1986). Virus-induced autoimmune demyelating disease of the central nervous system. “Concepts in Viral Pathogenesis I1 (A. L. Notkins and M. B. A. Oldstone eds.), p. 203. Springer-Verlag, New York. Jonjic, S., Mutter, W., Weiland, F., Reddehase, M. J., and Koszinowski, U. H. (1989). Siterestricted persistent cytomegalovirusinfection after selective long-term depletion of CD4+ T lymphocytes.J. Exp. Med. 169, 1199-1212. Jonjic, S., Pavic, I., Lucin, P., Rukavina, D., and Koszinowski, U. H. (1990). Efficacious control of cytomegalovirus infection after long-term depletion of CD8’ T lymphocytes. I. Vird. 64, 5457-5464. Jordan, M. C. (1978). Murine models of cytomegalovirus latency. “Persistent Viruses” (J. G. Stevens, G. J. Todaro, and C. F. Fox, eds.),pp. 805-810. Academic Press, New York. Julia, V., Rassoulzagedan, M., and Glaichenhaus, N. (1996). Resistance to leishmania major induced by tolerance to a single antigen. Science 274, 421-423.
CD4’ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
359
Julius, M. H. (1982). Celluliw interactions involved in T-dependent B cell activation. Immnunol. Today 3, 295. June, C. H., Bluestone, J. A., Nadler, L. M., and Thompson, C. B. (1994). The B7 and CD28 receptor families. Iminunol. Today 15, 321-331. Kadima-Nzuij, M., and Craighead, J. E. (1990).T-lymphocyte effects on murine cytomegalovirus pulmonary infection. Am. J. Pathol. 137, 907-912. Kagi, D., Lederman, B., Burki, K., Zinkernagel, R. M., and Hengartner, H. (1996). Molecular mechanisms of lymphocyte-mediated cytotoxicity and their role in immunological protection and pathogenesis in vivo. Ann,n. Rea. Iminunol. 14, 207-232. Kagi, D., Ledermann, B., Burki, K., Seiler, P., Odermatt, B., Olsen, K. J., Podack E., Zinkernagel, R. M., and Hengartner, H. (1994). Cytotoxicity mediated by T cells and natural killer cells is greatly imparied in perforin-deficient mice. Nature 369, 31-37. Kagi, D., Seiler, P., Pavlovic, J., Ledermann, B., Burki, K., Zinkernagel, R. M., and Hengartner, H. (1995). The roles of perforin- and fas-dependent cytotoxicity in protection against cytopathic and noncytopathic viruses. Eur. ]. Imnznnol. 25, 2356-2362. Kamanaka, M., Yu, P., Yasui, T., Yoshida, K., Kawabe, T., Horii, T., Kishimoto, T., and Kikutani, H. (1996). Protective role of CD40 in leishmania major infection at two distinct phases of cell-mediated immunity. Inmunity 4, 275-282. Karupiah, G., Blanden, R. V., and Rainshaw, I. A. (1990). Interferon gamma is involved in the recovery of athyrnic nude mice from recombinant vaccinia virushterleukin 2 infection. 1. Exp. Med. 172, 1495-1503. Kampiali, G., Buller, R. M. L., van Roojen, N., Duarte, C. J., and Chen, J. (1996). Different roles for CD4’ and CD8+ T lymphocytes and macrophage subsets in the control of generalized virus infection. ]. Virol. 70, 8301-8309. Karupiah, G., Frederickson, T. N., Holmes. K. I., Kliairallah, L. H., and Buller, R. M. L. (1993). Importance of interferons in the recovery from mousepox.]. Virol.67,4214-4226. Kast, W. M., Bronkhorst, A. M., DeWaal, L. P., and Melief, C. J. M. (1986). Cooperation between cytotoxic and helper T lymphocytes in protection against lethal Sendai virus infection: Protection by T cells is MHC-restricted and MHC-regulated; a model for MHC-disease associations. ]. Exp. Med. 164, 723-738. Keamey, E. R., Pape, K. A,, Loh, D. Y., and Jenkins, M. K. (1994).Visualization of peptidespecific T cell immunity and peripheral tolerance induction in vivo. Inmwzity 1,327-339. Kerr, I. M., and Brown. R. E. (1978). pppA2’p5’A2’p5’A: An inhibitor of protein synthesis synthesized with an enzyme fraction from interferon-treated cells. Proc. Natl. Acad. Sci. U S A 75,256-260. Koelle, D. M., Corey, L., Burke. R. L., Eisenberg. R. L., Cohen, G. H., Pichyangkura, R., and Triezenberg, S. 1. (1994).Antigen-specificitiesof human CD4+T cell clones recovered from recurrent genital herpes simplex virus type 2 lesions. ]. Virol. 68, 2803-2810. Kohonen-Corish, M. R. J., King, N. J. C., Woodhams, C. E., and Ramshaw, I. A. (1990). Immiinodeficient mice recover from infection with vaccinia virus expressing interferony . Eur. ]. Imrnunol. 20, 157-161. Koike, S., Taya, C., Kurata, T., Abe, S., Ise, I., Yonekawa, H., and Nomoto, A. (1991). Transgenic mice susceptible to poliovirus. Proc. Natl. Acad. Sci. U S A 88, 951-955. Koszinowsky, U. H. (1991). Molecular aspects of immune recognition of cytomegalovirus. Transplant. Proc. 23, 70-74. Kropshofer, H., Hainnierling, G . J. and Vogt, A. B. (1997). How HLA-DM edits the MHC class 11 peptide repertoire: survival of the fittest? Itnrnunol. Today 18, 77-82. Kiindig, T., Castelmur, I., Bachniann, M. F., Abraham, D., Binder, D., Hengartner, H. and Zinkernagel, R. M. (1993a). Fewer protective cytotxic T cell epitopes than T helper cell epitopes on vesicular stomatitis vints. ]. Virol. 67, 3680-3683.
360
ANNETTE OXENIUS et al
Kundig, T. M., Hengartner, H., and Zinkemagel, R. M. (1993b).T cell dependent interferon gamma exerts an antiviral effect in the central nervous system but not in peripheral sold organs. J. Immunol. 150, 2316-2321. Kupfer, A., Mosmann, T. R., and Kupfer, H. (1991). Polarized expression of cytokines in cell conjugates of helper T cells and splenic 8 cells. Proc. Natl. Acad. Sci. USA 88,775-779. Kurt-Jones, E. A,, Liano, E. D., Hayglass, K. A,, Benacerraf, B., Sym, M. S., and Abbas, A. K. (1988). The role of antigen-presenting B cells in uivo: Studies of B deficient mice. J. Immunol. 140, 3773-3778. Kurtz, C. I., Sun, X. M., and Fujinami, R. S. (1995). B-lymphocyte requirement for vaccinemediated protection from Theiler's murine encephalomyelitis virus-induced central nervous system disease. J. Virol. 69, 5152-5155. Lake, P., and Mitchison, N. A. (1976). Regulatory mechanisms in the immune response to cell surface antigens. Cold Spring H a d . Symp. Quant. Biol. 41, 589. Lamb, C. A,, Yewdell, J. W., Bennink, J. R., and Cresswell, P. (1991). Invariant chain targets HLA class I1 molecules to acidic endosomes containing internalized influenza virus. Proc. Natl. Acad. Sci. USA 88, 5998-6002. Lanzavecchia, A. (1987). Antigen uptake and accumulation in antigen-specific B cells. Immunol. Reu. 99, 39-51. Lanzavecchia, A. (1990). Receptor-mediated antigen uptake and its effect on antigen presentation to class 11-restricted T lymphocytes. Annu. Reu. Immunol. 8, 773-793. La Posta, V. J., Auperin, D. D., Kamin-Lewis, R., and Cole, G. A. (1993). Cross-protection against lymphocytic choriomeningitis virus mediated by a CD4+ T cell clone specific for an envelope glycoprotein epitope of Lassa virus. J. Virol. 67, 3497-3506. Lassila, O., Vainio, O., and Matzinger, P. (1988). Can B cells turn on virgin T cells? Nature 334,253-255. Lefrancois, L. (1984). Protection against lethal viral infection by neutralizing and nonneutralizing monoclonal antibodies: Distinct mechanisms of action in vivo. J. Viral. 51, 208-214. Lehrich, J. R., Amason, B. G. W., and Hochberg, F. H. (1976). Demyelinative myelopathy in mice induced by DA virus. Science 29, 149. Leist, T. P., Kohler, M., and Zinkernagel, R. M. (1989). Impaired generation of antiviral cytotoxicity against lymphocytic choriomeningitis and vaccinia virus in mice treated with CD4-specific monoclonal antibody. Scand. J. Immunol. 30, 679-686. Levin, D., Constant, S., Pasqualini, T., Flavell, R., and Bottomly, K. (1993). Role of dendritic cells in the priming of CD4+ T lymphocytes to peptide antigen in vivo. J. Immunol. 151,6742-6750. Liang, S., Mozdzanowska, K., Palladino, G., and Gerhard, W. (1994). Heterosubtypic immunity to influenza type A virus in mice. J. Immunol. 152, 1653-1661. Liblau, R. S., Tisch, R., Shokat, K., Yang, X. D., Dumont, N., Goodnow, C. C., and McDevitt, H. 0. (1995). Intravenous injection of soluble antigen induces thymic and peripheral T-cell apoptosis. Proc. Natl. Acad. Sci. USA 93, 3031-3036. Lightman, S., Cobbold, S., Waldmann, H., and Askonas, B. A. (1987). Do L3T4' T cells act as effector cells in protection against influenza virus infection. Immunology 62,139-144. Linsley, P. S., Brady, W., Urnes, M., Grosmaire, L. S., Damle, N. K., and Ledbetter, J. A. (1991). CTLA-4 is a second receptor for the B cell activation antigen B7.J Exp. Med. 174,561-569. Linsley, P. S., and Ledbetter, J. A. (1993). The role of the CD28 receptor during T cell responses to antigen. Annu. Rev. Immunol. 11, 191-212. Lipton, H. L. (1995). Theiler's virus infection in mice: An unusual biphasic disease process leading to demyelination. Infect. Immun. 11, 1147.
CD4’ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
36 1
Lipton, H. L., and Dal Canto, M. C. (1976). Theiler’s virns-induced demyelination: Prevention by immunosuppression. Science 192, 62. Liu, Y., and Janeway, C. A., Jr. (1991). Microbial induction of costimulatory activity for CD4’ T cell growth. Intl. Immunol. 3, 323-332. Liu, Y., Ramarathinam, L., Guo, Y., Huszar, D., Trounstine, M., and Zhao, M. (1995). Gene-targeted B-deficient mice reveal a critical role for B cells in the CD4 T cells response. Intl. Imnmuno~.7, 1353-1362. Long, E. 0. (1992). Antigen processing for presentation to CD4+ T cells. New Biol. 4, 274-284. Lopez, A. F., Sanderson, C. J., Gamble. J. R., Campbell, H. D., Young, I. G., and Vadas, M. A. (1988).Recombinant human interleukin 5 is a selective activator of human eosinophil function. J. Exp. Med. 167, 219-224. Mackenzie, C. D., Taylor, P. M., and Askonas, B. A. (1989). Rapid recovery of lung histology correlates with clearance of influenza virus by specific CD8+cytotoxic T cells. lnimunology 67, 375-381. Mahon, B. P., Katrak, K., and Mills, K. H. G. (1992). Antigenic sequences of poliovirus recognized by T cells: Serotype-specific epitopes on W1 and W3 and cross-reactive epitopes on VP4 defined using CD4’ T-cell clones. J. Virol. 66, 7012-7020. Mahon, B. P., Katrak, K., Noinoto, A,, Macadam, A. J., Minor, P. D., and Mills, K. H. G. (1995).Poliovirus-specificCD4+Thl clones with both cytotoxicand helper activity mediate protective humordl immunity against lethal pohovirus infection in transgenic mice expressing the huinan poliovirus receptor. 1. Exp. Med. 181, 1285-1292. Malnati, M. S., Marti, M., LaVaute, T., Jaraqueinada, D., Biddison, W., DeMars, R., and Long, E. 0. (1992). Processing pathways for presentation of cytosolic antigen to MHC class 11-restricted T cells. Nature 357, 702-704. Manickan, E., Francotte, M., Kuklin, N., Dewerchin, M., Molitor, C., Gheysen, D., Slaoui, M., and Rouse, B. T. (1995). Vaccination with recombinant vaccinia virus expressing ICP27 induces protective immunity against herpes simplex virus through CD4+ Thl cells. /. Virol. 69, 4711-4716. Manickan, E., and Rouse, B. T. (1995). Roles of different T cell subsets in control of herpes simplex virus infection determined by using T cell-deficient mouse models. J. Virol. 69,8178-8179. Matloubian, M., Conception, R. J., and Ahmed, R. (1994). CD4’ T cells are required to sustain CD8’ cytotoxic T cell responses during chronic viral infection. J. Virol. 68, 80568063. Melcliers, F., and Anderson, J. (1984).B cell activation: Three steps and their variations. Cell 37, 715. Meshkova, E. N., and Mentkevich, L. M., (1987). Potentiation of the antiviral activity of human interferons by virosol in in vitro experiments. Vopr. Virosol. 32, 570-573. Metz, D. H., and Esteban, M. (1972). Interferon inhibits viral protein synthesis in L cells infected with vaccinia virus. Nature 238, 385-388. Milich, D. R., and MacLachlan, A. (1986). The nucleocapsid of hepatitis B virus is both a T-cell-independent and a T-cell dependent antigen. Science 234, 1398-1401. Milich, D. R., McLachlan,A., Thornton, G. B., and Hughes, J. L. (1987).Antibody production to the nucleocapsid and envelope of the hepatitis B virus primed by a single synthetic T cell site. Nature 329, 547-549. Miller, J., and Germain, R. N. (1986). Efficient cell surface expression of class I1 MHC molecules in the absence of associated invariant chain. /. Exp. Med. 164, 1478-1489. Minis, C. A. (1987).“The pathogenesis of Infectious Disease,” Vol. 3, pp. 254-269. Academic Press, London.
362
ANNETTE OXENIUS et nl.
Missale, G., Redeker, A,, Person, J,, Fowler, P., Guilhot, S., SchIicht, H. J., Ferrari, C., and Chisari, F. V. (1993). HLA-A31- and HLA-Aw68-restrictedcytotoxicT cell responses to a single hepatitis B virus iiucleocapsid epitope during acute viral hepatitis. J. Exp. Med. 177, 751-762. Mitchison, N. A. (1964). Induction of immunological paralysis in two zones of dosage. Roy. Soc. Proc. 161, 275-292. Miyazaki, T., Wolf, P., Toume, S., Waltzinger, D., Dierich, A., Barois, N., Ploegh, H. L. Benoist, C., and Mathis, D. (1996). Mice lacking H2-M complexes, enigmatic elements of the MHC class I1 peptide-loading pathway. Cell 84, 531-541. Moreno, J., Vignali, D. A. A,, Nadimi, F., Fuchs, S., Adorini, L., and Haemmerling, G. J. (1991). Processing of an endogenous protein can generate MHC class 11-restrictedT cell determinants distinct froin those derived from exogenous antigen.]. lmmunol. 147,33063313. Moskophidis, D., Cobbold, S. P., Waldmann, H., and Lehmann-Gruhe, F. (1987). Mechanism of recovery from acute virus infection: treatment of lymphocytic chorionieningitis virus-infected mice with monoclonal antibodies reveals that Lyt-2+T lymphocytes mediate clearance of virus and regulate the antiviral antibody response. J. Virol. 61, 1867-1874. Moskophidis, D., Lechner, F., Pircher, H. P., and Zinkemagel, R. M. (1993). Virus persistence in acutely infected iinmunocompetent mice by exhaustion of antiviral cytotoxic effector T cells. Nature 362, 758-761. Mosmann, T. R., Chenvinski, H., Bond, M. W., Giedlin, M. A., and Coffinan, R. L. (1986). Two types of inurine helper T cell clones. I. Definition according to profiles of lymphokine activities and secreted proteins. J. lmmunol. 136, 2348-2357. Mosmann, T. R., and Coffman, R. L. (1987). Two types of mouse helper T-cell clones; implications for immune regulation. lmmunol. Today 8, 223-227. Mosmann, T. R., and Coffman, R. L. (1989). TH1 and TH2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. lmmunol. 7, 145-173. Mueller, D. L., Jenkins, M. K., and Schwartz, R. H. (1989). Clonal expansion versus functional clonal inactivation: A costimulatory signalling pathway determines the outcome of T cell antigen receptor occupancy. Annu. Rev. lmtnunol. 7, 445-480. Muller, D., Koller, B. H., Whitton, J. L., LaPan, K. E., Brigman, K. K., and Frelinger, J. A. (1992). LCMV-specific, class 11-restrictedcytotoxic T cells in beta Z-microglobulindeficient mice. Science 255, 1576-1578. 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 longterm stimulation. J. Exp. Me[. 183, 901-913. Murphy, F. A. (1977). Rabies pathogenesis. Arch. Virol. 54, 279-297. Mutter, W., Reddehase, M. J., Busch, F. W., Buhring, H. J., and Koszinowski, U. H. (1988). Failure in generating hemopoietic stem cells is the primary cause of death from cytomegalovinis disease in the immunoconipromised host. J. E x p Med. 167, 1645-1658. Nash, A. A., Jayasuriya, A,, Phelan, J., Cobbold, S. P., Waldmann, H., and Prospero, T. (1987). Different roles for L3T4+ and Lyt2+ T cell subsets in the control of an acute herpes simplex virus infection of the skin and newous system.1.Gen. Virol. 68,825-833. Nayersina, R., Fowler, P., Guilhot, S., Missale, G., Cemy, A., Schlicht, H. J., Vitiello, A,, Chesnut, R., Person, J. L., and Redeker, A. G. (1993). HLA A2 restricted cytotoxic T lymphocyte responses to multiple hepatitis B surface antigen epitopes during hepatitis B virus infection. J. lrnmund. 150, 4659-4671. Neefjes, J. J., and Momburg, F. (1993). Cell biology of antigen presentation. Curr. q i n . ltnmunol. 5, 27-34.
CD4‘ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
363
Newcomb, J . R., and Cresswell, P. (1993). Characterization of peptides bound to purified HLA-DR niolecnles and their absence from invwiant chain-associated Crp dimers. J. I m t t ~ u i w / .150, 499-507. Nuchtern, J. G., Bonifacino, J. S., Biddison, W. E., and Klausner, R. D. (1989). Brefeldin A implicates egress from endoplasmic reticulum in class I restricted antigen presentation. Nature 339, 223-226. Odermatt, B., Eppler, M., Leist, T. P., Hengartner, H., and Zinkeniagel, R. M. (1991). Virus-triggered acquired immunodeficiency by cytotoxic T-cell dependent destruction of antigen-presenting cells and lynnph follicle stnicture. Proc. Nritl. Acud. Sci. USA 88, 82528256. O’Garra, A., Macatonia, S. E., Hsieh, C. S., and Muqhy, K. M. (1993). Regulatory role of IL4 and other cytokines in T helper cell development in an alpha beta TCR transgenic ntonse system. Res. lmtntltrol. 144, 620-625. O’Garra, A,, and M~irphy,K. (1994).Role ofcytokines in determining T-lymphocyte function. Curr. Ol~in.Imtiunol. 6, 458-466. O’Neill, H. C., and Brenan, M. (1987). A role for early cytotoxic T cells in resistance to ectromelia vims infection in inice. J. Gen. Virol. 68, 26-10. Oxenius, A,, Bachmann, M. F., Ashton-Rickardt, P. G., Tonegawa, S., Zinkeniagel, R. M., and Hengartner, H.(1995). Presentation of endogenous viral proteins in association with major histocompatibility complex 11: On the role of intracellular coinpartmentalization, invariant chain and the TAP tiansporter system. Eur. /. Itiztiiunol. 25, 3402-3411. Oxenins, A,, Bachinann, M. F., Mathis, D., Benoist, C., Zinkernagel, R. M., and Hengartner, H. (1997).Functional in vivo MHC class I1 loading by endogenously synthesized glycoprotein during viral infection. J . I t t m u t d 158, 5717-5726. Oxenius, A., Bachmann, M. F., Zinkernagel, R. M., and Hengartner, H. (1998). Viriisspecific MHC class 11-restricted TCR transgenic mice: Iinplications of elevated Th cell frequencies on huinoritl and c e h l a r iinrnune responses after viral infection. Eur. 1. I t r i i t w t i o l . 28, 390-400. Oxenius, A,, Campbell, K. A,, Maliszewski, C. R., Kishinioto, T., Kikntani, H., Hengartner, H., Zinkernagel, R. M., and Bachmann, M. F. (1996). CD40-CD40 ligand interactions are critical in T-B cooperation but not for other anti-viral CD4+ T cell functions. J. Exp. Mecl. 183, 2209-2218. Paez, E., and Esteban, M. (1984). Nature and inodc of action of vaccinia virus products that block activation of interferon-mediated ooo(A2’p)nA-synthetase.Virology 134, 29-39. Palladino, G., Scherle, P. A., and Gerhard, W. (1991). Activity of CD4+ T-cell clories of type 1 and type 2 in generation of influenza virns-specific cytotoxic responses in vitro. 1.W r d . 65, 6071-6076. Pantaleo, G., and Fanci, A. S. (1995). New concepts in the iminunopathogenesis of HIV infection. Atznu. Reo. Ztttntutiol. 13, 487-512. Pantaleo, G., Graziosi, C., Demarest, J. F., Cohen, 0. J., Vaccarezza, M., Gantt, K., MuroCacho, C., and Fauci, A. S. (1994). Role of lymphoid organs in the pathogenesis of human iinmunodeficiency virus (HIV) infection. I I W ~ Z U Reu. ~ J ~ 140, . 10,5-130. Pantaleo. C . , Graziosi, C., and Fauci, A. S. (1993). The irnmiiiia~~~thogeiiesis of hnnian immunodeficiency virus infection. N . Ettgl. J. Med. 328, 227-235. Parker, D. C. (1993).T cell-dependent B cell activation. Annu. Reu. ltniiiuttol. 11,331-360. Paul, W. E. (1991). Interleukin-4: A prototypic immunoregulatory lymphokine. B h o d 177, 1859-1870. Paul, W’.E. (1993).“Fundamental Innnnnology,” Vol. 3, pp. 1- 1490.Raven Press, New York. Penna, A,, Chisari, F. V., Bertoletti, A,. Missale, G., Fowler, P., Ginberti, T., Fiaccadori, F., and Ferraii, C. (1991). Cytotoxic T lymphocytes recognize an HLA-A2-restricted epitope within the hepatitis B kirris nncleocapsid autigen. J . E x p Med. 174, 1565-1570.
364
ANNETTE OXENIUS et al.
Perelson, A. S., Neumann, A. U., Markowitz, M., Leonard, J. M., and Ho, D. D. (1996). HW-1 dynamics in vivo: Virion clearance rate, infected cell life-span, and viral generation time. Science 271, 1582-1586. Phillips, J. A,, Romball, C. G., Hobbs, M. V., Emst, D. N., Shultz, L., and Weigle, W. 0. (1996). CD4+ T cell activation and tolerance induction in B cell knockout mice. J . Exp. Med. 183, 1339-1344. Pinet, V., Malnati, M. S., and Long, E. 0.(1994).Two processing pathways for the MHC class 11-restricted presentation of exogenous influenza virus antigen. J. Immunol. 152, 4852. Polydeflds, M., Koenig, S., Flexner, C., Obah, E., Gebo, K., Chakrabarti, S., Earl, P. L., Moss, B., and Siliciano, R. F. (1990).Anchor sequence-dependent endogenous processing of human immunodeficiencyvirus 1envelope glycoprotein gp160 for CD4' T cell recognition. J Exp. Med. 171, 875-887. Pullen, L. C., Miller, S. D., DalCanto, M. C., and Kim, B. S. (1993). Class I-deficient resistant mice intrdcerebrdy inoculated with Teiler's virus show an increased T cell response to viral antigens and susceptibilityto demyelination. Eur. J. Immunol. 23,22872293. Rammensee, H. G., Kroschewski, R., and Frangoulis, B. (1989). Clonal anergy induced in mature Vb6+ T lymphocytes on immunizing Mls-1 mice with Mls-1 expressing cells. Nature 339, 541-544. Ramsay, A. J., Ruby, J., and Rainshaw, I. A. (1993). The case for cytokines as effector molecules in the resolution of virus infection. Immunol. Today 14, 155-157. Ramshaw, I., Ruby, J., Ramsay, A,, Ada, G., and Karupiah, G. (1992).Expressionofcytokines by recombinant vaccinia viruses: A model for studying cytokiues in virus infections in vivo. Immunol. Rev. 127, 157-182. Razi-Wolf, Z., Freeman, G . J., Galvin, F., Benacerraf, B., Nadler, L., and Reiser, H. (1992). Expression and function of the inurine B7 antigen, the major costimulatory molecule expressed by peritoneal exudate cells. Proc. Natl. Acad. Sci. USA 89, 4210-4214. Reddehase, M. J., Jinjic, S., Weiland, F., Mutter, W., and Koszinowsky, U. H. (1988). Adaptive immunotherapy of murine cytomegalovirus adrenalitis in immunocompromised host: CD4-independent antiviral function of CD8-positive memory T lymphocytesderived from latently infected donors. J . Virol. 62, 1061. Reddehase, M. J., Mutter, W., Munch, K., Buhring, H. J., and Koszinowski, U. H. (1987). CD8-positive T lymphocytes specific for murine cytomegdovirus immediate-early antigens mediate protective immunity. J . Virol. 61, 3102-3108. Reddehase, M. J., Weiland, F., Mtinch, K., Jonjic, S., Luske, A., and Koszinowski, U. H. ( 1985). Interstitial murine cytomegdovirus pneumonia after irradiation: Characterization of cells that limit viral replication during infection of the lungs. 1.Virol.55, 264. Reich, A., Erlwein, O., Niewiesk, S., Ter Meulen, V., and Liebert, U. G. (1992).CD4+T cells control measles virus infection in the central nervous system. Immunology 76, 185-191. Ren, R., and Racaniello, V. R. (1992). Human poliovirus receptor gene expression and poliovirus tissue tropism in transgenic mice. J . Virol. 66, 296-302. Rice, A. P., and Kerr, I. M. (1984). Interferon-mediated, double-stranded RNA-dependent protein kinase is inhibited in extracts from vaccinia virus-infected cells. Nature 50, 229-236. Riese, R. J., Wolf, P. R., Bromine, D., Natkin, L. R., Villadangos, J. A., Ploegh, H. L., and Chapman, H. A. (1996). EssentiaI role for cathepsin S in MHC class 11-associatedinvariant chain processing and peptide loading. Irnmunity 4, 357-366. Romagnani, S., Del Prete, C., Manetti, R., Ravina, A., Annunzidto, F., De Carli, M., Mazzetti, M., Piccinni, M. P., Delios, M. M., Parronchi, P., Sampognaro, S., and Ma@, E. (1994). Role of Thl Th2 cytokines in HIV infection. Immunol. Rev. 140, 73-92.
CD4+ T-CELL INDUCTION AND EFFECTOR FUNCTIONS
365
Ron, Y., DeBaetselier, J., Gordon, J., Feldmarin, J., and Segal, S. (1981).Defective induction of antigen-reactive proliferating T cells in B-cell deprived mice. Eur. J. Immunol. 11, 964- 968. Ron, Y., and Sprent, J. (1987). T cell priming in vivo: A major role for B cells in presenting antigen to T cells in lymph nodes. J . Irnmunol. 138, 2848-2856. Ronchese, F., and Hausmann, B. (1993). B lynphocytes in vivo fail to prime naive T cells but can stimulate antigen-experienced T lymphocytes. J . Exp. Med. 177, 679-690. Rudensly, A., Rath, S., Preston-Hurlburt, P., Murphy, D. B., and Janeway, C. A. (1991a). On the complexity of self. Nature 353, 660-662. Rudensky, A. Y., Preston-Hurlburt, P., Hong, S. C., Barlow, A,, and Janeway, C. A. (1991b). Sequence analysis of peptides bound to MHC class I1 molecules. Nature 353, 622-627. Russell, S. M. (1979). T cells primed by influenza virion internal components can cooperate in the antibody response to haemagglutinin. Nature 280, 147-148. Russell, S. M., and Liew, F. Y. (1980). Cell cooperation in antibody responses to influenza virus. I. Priming of helper T cells by internal components of the virion. Eur. J. Immunol. 10, 791-796. Sallusto, F., and Lanzavecchia, A. (1994). Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colonystimulating factor plus interleukin 4 and downregulated by tumor necrosis factor alpha. J. Exp. Med. 179, 1109-1118. Sant, A. J., and Miller, J. (1994). MHC class I1 antigen processing: Biolo@ of invariant chain. Cum. Opin. Immunol. 6, 57-63. Sarawar, S. R., and Doherty, P. C. (1994). Concurrent production of interleukin-2, interleukin-10 and gamma interferon in the regional lymph nodes of mice with influenza pneumonia. J. Virol. 68, 3112-3119. Sarawar, S. R., Sangster, M., Coffman, R. L., and Doherty, P. C. (1994). Administration of anti-IFNy antibody to 02-microglohnlin-deficient mice deIays influenza virus clearance but does not switch to a T helper cell 2 phenotype. J. Irninunol. 153, 1246-1253. Scherle, P., and Gerhard, W. (1986). Functional analysis of influenza-specifichelper T cell clones in vivo. J. Exp. Med. 164, 1114-1128. Scherle, P. A,, Palladino, G., and Gerhard, W. (1992). Mice can recover from pulmonary influenza virus infection in the absence of class I-restricted cytotoxic T cells. J. Imrnunol. 148,212-217. Schirmbeck, R., and Reimann, J. (1994). Peptide transporter-independent, stress proteinmediated endosomal processing of endogenous protein antigens for major histocompatibility complex class I presentation. Eur. J. Immunol. 24, 1478-1486. Schneider-Schaulies, S., and Liebert, U.G. (1991).Pathogenetic aspects of persistent measles infections in brain tissue. Sem. Neurosci. 3, 139. Schwartz, R. H. (1990). A cell culture model for T lymphocyte clonal anergy. Science 248, 1349-1356. Seiler, P., Kalinke, U., Riilicke, T., Bucher, E. M., Bfise, C., Zinkernagel, R. M., and Hengartner, H. (1998). Eiihaiiced virus clearance by early inducible lymphocytic choriomeningitis virus neutralizing antibodies in iininunoglobulin transgenic mice. J . Virol. 72, 2253-2258. Sethi, K. K., Omata, Y., and Schneweis, K. E. (1983). Protection from mice by fatal herpes simplex virus infection by adoptive transfer of cloned virus-specific and H-Zrestricted cytotoxic T lymphocytes. J . Gen. Virol. 64, 443-447. Sherman, M. A,, Weber, D. A., and Jensen, P. E. (1995). DM enhances peptide loading to class I1 MHC by release of invariant chain-derived peptide. Immunity 3, 197-205.
366
ANNETTE OXENIUS et nl
Soong, L., Xu, J. C., Grewal, S., Kima, P., Sun, J., Longley Jr., B. J., Ruddle, N. H., McMahonPratt, D., and Flavell, R. A. (1996). Disruption of CD40-CD40 ligand interactions results in enhanced susceptibility to leishmania amazoniensis infection. Inimunity 4, 263-274. Sornasse, T., Flamand, V., De Becker, G., Bazin, H., Tielemans, F., Tielemans, K., Urbain, Leo, O., and Moser, M. (1992). Antigen-pulsed dendritic cells can efficiently induce an antibody response in vivo. 1.Exp. M e d 175, 15. Spriggs, M. K., Koller, B. H., Sato, T., Morrissey, P. J.. Fanslow, W. C., Smithies, O., Voice, R. F., Widmer, M. B., and Maliszewski, C. R. (1992). B2-microglobulin-, CD8+ T-celldeficient mice survive inoculation with high doses of vaccinia virus and exhibit altered IgG responses. Proc. Nutl. Acud. Sci. USA 89, 6070-6074. Staeheli, P. (1990). Interferon-induced proteins and the antiviral state. Adu. Vinis Res. 38, 147-200. Steinmann, R. M. (1991). The dendritic cell system and its role in immunogenicity. Anizu. Reu. Itmunot. 9, 271. Stockinger, B., Pessara, U., Lin, R. €I., Habicht, J., Grez, M., and Koch, N. (1989). A role of Ia-associated invariant chains in antigen processing and presentation. Cell 56,684-689. Stiiber, E., Strober, W., and Neurath, M. (1996). Blocking the CD40L-CD40 interaction in vivo specifically prevents the priming of T helper 1 cells through the inhibition of interleukin 12 secretion. J . Exp. Med. 183,693-698. Sunshine, G. H., Jimmo, B. L., Ianelli, C., and JaMs, L. (1991). Strong priming of T cells adoptively transferred into scid mice. J. Exp. Med. 174, 1653. Swain, S. L., Bradley, L. M., Croft, M., Tonkonogy, S., Atkins, G., Weinberg, A. D., Duncan, D. D., Iledrick, S. M., Dutton, R. W., and Huston, G. (1991). Helper T cell subsets: Phenotype, function and the role of lymphokines in regulating their development. Immunol. Reu. 123, 115-144. Swain, S. L., Weinberg, A. D., English, M., and Huston, G. (1990). IL-4 directs the development of Th-2-like helper effectors. J. Zmmunol. 145, 3796-3806. Sweetser, M. T., Morrison, L. A,, Braciale, V. L., and Braciale, T. J. (1989). Recognition of pre-processed endogenous antigen by class I but not class I1 mHC-restricted T cells. Nature 342, 180-182. Thomsen, A. R., Johansen, J., Marker, O., and Christensen, J. P. (1996). Exhaustion of CTL memory and recrudescence of viremia in lymphocytx choriomeningitis virus-infected MHC class 11-deficient and B cell deficient mice.]. Immunol. 157, 3074-3080. Thomsen, A. R., and Marker, 0. (1988). The complementary roles of cellular and humoral immunity in resistance to re-infection with LCM virus. Immunology 65, 9-15. Tough, D. F., Borrow, P., and Sprent, J. (1996). Induction of bystander T cell proliferation by viruses and type I interferon in vivo. Science 272, 1947-1950. Tounie, S., Miyazaki, T., Oxenius, A., Klein, L., Fehr, T., Kyewski, B., Benoist, C., and Mathis, D. (1997). Selection of a broad repertoire of CD4' T cells in H2-MaO/O mice. Immunity 7 , 187-196. Tripp, R. A., Sarawar, S., and Doherty, P. C. (1995). Characteristics of influenza virusspecific CD8+ T cell response in mice homozygous for disruption of the H-2IAb gene. 1.Immunol. 155, 2955-2959. Tulp, A., Venvoerd, D., Dobberstein, B., Ploegh, H. L., and Pieters, J. (1994). Isolation and characterization of the intracellular MHC class I1 compartment. Nuture 369,120-126. Van Binnendijk, R. S., Poelen, M. C. M., Kiupers, K. C., Osterhaus, A. D. M., and Uytdehaag, F. G. C. M. (1990).The predominance of CD8' T cells after infection with measles virus suggests a role for CD8+ class I MHC-restricted cytotoxic T lymphocytes (CTL) in recovery from measles. J. Immunol. 144, 2394-2399.
CD4’ T-CELL INDUCTION AND EFFECTOR FUhCTIONS
367
van Essen, D., Kikutani, H., and Gray, D. (1995). CD40 ligand-transduced costirnulation of T cells in the development of helper function. Nature 378, 620-623. Vento, S., Rondanelli, E. G., Ranieri, S., O’Brien, C. J., Williams, R., and Etldelston, A. L. (1987). Prospective study of cellular immunity to hepatitis-B-vinis antigens from tlie early incubation phase of acute hepatitis B. Lancet 2, 119-122. Virelizier, J. L. (1975). Host defence against influenza virus: Tlie role of antihemagglutinin antibodies. J . Zniniurid. 115, 434. Virelizier, J. L., Allison, A. C., arid Schild, G. C. (1979). Iniinr~neresponses to influenza vinis in the mouse, and their role in control of infection. Br. bled. Bull. 35, 65. Viville, S.. Neefjes, J., Lotteau, V., Dierich, A,. LeMeur, M., Ploeglr, H., Benoist, C., and Mathis, D. (1993). Mice lacking the MHC class 11-associated invariant chain. Cell 72,6335448. von Herrath, M. G., Yokoyama, M., Dockter, J., Oldstone, M. B. A,, aiid Whitton, J. L. (1996).CD4-deficient mice have reduced levels of ineinory cytotoxic T lymphocytes after immunization and show cliniinislied resistance to srhequent virus challenge. /. Virol. 70, 1072-1079. Waanders. G. A,, Shakov,A. N., Held. W., Karapetian, O., Acha-Orbea, H., and MacDonalcl, H. R. (1993). Peripheral T cell artivation and deletion inctuced by transfer of lymphocyte subsets expressing endogenous or exogenous niouse manimary tumor virus. /. Exp. Med. 177, 1359-1366. Wagner, R. R. (1987). “The Rhabdoviruses,” Vol. 1, pp. 1-544. Plenum Press, New York. Warren, H. S., Vogel, F. K., and Chedid, L. A. (1986). Current status of immunological adjuvants. Annu. Reo. Zninit~riol.4, 369-388. Webb, S., Morns, C., and Sprent, J. (1990).Extrathymic tolerance of mature T cells: Clonal elimination as a consequence of immunity. Cell 63, 1249-1256. Weigle, W. 0. (1973). Imniunological unresponsiveness. Ado. In~niunol.16, 61- 122. Weiss, S., aiid Bogen, B. (1991). M€IC class 11-restricted presentation of intracellular antigen. Cell 64, 767-776. Welsh, C. J. R., Tonks, P., Nash, A. A,, arid Blakemore, !A.’ F. (1987). Tlie effect of L3T4 T cell depletion on the pathogenesis of Theiler’s murine encephalomyelitis virus infectiou in CBA mice. J . Geii. Virol. 68, 1659-1667. Wong, G. H., and Goeddel, D. V. (1986). Tuinorir necrosis fktors alpha and beta inhibit vims replication and synergize with interferons. Nntttrr 323, 819-822. Yang, Y., and Wilson, J. M. (1996). CD40 ligand-dependent T cell activation: Requirement of B7-CD28 signaling through CD40. Scieiicc 273, 1862- 1864. Yap, K. L.. Ada, G. L., and McKenzie, I. F. C. (1978). Transfer of specific cytotoxic T lymphocytes protects mice inoculated with inHuenza virus. Nrrtuw 273, 238. Zajak, A. J.. Quinn, D. G., Cohen, P. I., and Freliiiger, J. A. (1996). Fas-dependent CD4’ cytotoxic T-cell-mediated pathogenesis during vinis infection. Proc. NatE. Acad. Sci. USA 93, 14730-14735. Zinkernagel, R. M. (1993). “Iniinunity to Viruses,” Vol. :3, pp. 1211-1250. Raven Press, New York. Zinkeniagel, R. M., and Althage, A. (1977). Anthird protection hy virus-immune cytotoxic T cells: Infected target cells are lysed before infectious virus progeny is assembled. J . Exp. Med. 145, 644-6351. Zinkernagel, R. M., Cooper, S.. Chambers, J., Lazzarini, R. A., Hengartner, H., and Arnheiter, H. (1990). Virus induced autoaiitibody response to a transgenic viral antigen. Notwe 344, 68-71. This article was accepted for publication on January 1 3 , 1998
This Page Intentionally Left Blank
ADVANLE.5 IN 1MMUhOLOC.Y VOL. 70
Current Views in lntracellular Transport: Insights from Studies in Immunology VICTOR W. HSU*f’ AND PETER J. PETERSt ‘Division of Rheumatology, Immunology, and Allergy, Brigham and Women‘s Hospital, Harvard Medical School, Boston, Massachusetts 02 1 15; and tThe Netherlands Cancer Institwe, Plesmanlaanlaan 12 1, 1066 CX Amsterdam, The Netherlands
1. Introduction
How proteins are transported within the cell is best understood for a general transport system that shuttles proteins within membrane-bound carriers. This review summarizes current concepts of how this transport occurs by examining a general mechanism describing this transport and then detailing variations in this mechanism as they pertain to specific intracellular transport pathways. In this context, examples of studies in immunology that have provided insights into our current understanding of these transport pathways will be highlighted. Intracellular membrane carriers travel along specific pathways that connect a series of membrane-bound compartments, termed organelles ( Fig. 1).These pathways are used by nearly all proteins that must be brought either out to or in from the extracellular environment (Palade, 1975). Pathways that bring materials out are termed exocytic or secretory pathways, whereas pathways that bring materials into the cell are termed endocytic pathways. A few examples exist where proteins may be secreted without traversing these pathways (Kuchler, 1993).However, the mechanistic detail of how this is accomplished is not well understood and thus will not be a focus of this review.
A. SECRETORY PATHWAYS The endoplasmic reticulum (ER)is the port of entry for all proteins that are transported through the secretory pathways. In the cytosol, nascent proteins that are still being synthesized on ribosomes are targeted to the ER membrane because they contain an amino-terminal stretch of hydrophobic residues, known as the signal peptide (Blobel and Dobberstein, 1975).A protein complex termed the signal recognition particle (SRP) binds to the signal peptide and targets the nascent protein to another complex on the ER membrane. The target complex on the ER contains the binding site for the SRP,the SRP receptor, and a channel that translocates the nascent
’
Correspondmg anthor: Brigham and Women’s IIospital, 1Jimmy Fund Way, Smith 538, Boston. MA 02115. 369
Copyright B 1998 hy Acddellric I’rrss. All rights of reproduction in an)- forin 11w3ved. O(lfi5-277fiI98 $25.00
370
VICTOR W. HSU AND PETER J. PETERS
PM
8
c
30 ER
F I ~ 1. . Intracellular coinpartments and transport pathways. ER, endoplasmic reticulum; IC, intermediate compartment; CGN, cis-Golgi network; TGN, truns-Colgi network; EE, early endosome; LE, late endosome; PM, plasma membrane. Arrows indicate welldocumented pathways.
protein across the ER membrane, the Sec61 complex. The entire assembled complex that includes the ribosome, SRP, and the translocation machinery is restricted to the portion of the ER termed the rough ER (Rapoport et al., 1996; Walter and Johnson, 1994). The ER membrane contains lipidlinked oligosaccharides that are transferred onto select asparagine residues
C U R R E N T VIEWS IN INTRACEL,I,UI,AH T R A N S P O R T
371
of translocating polypeptides. This process results in N-linked glycosylation of proteins. The initially formed glycoproteins have immature carbohydrates, termed high-mannose chains, that are subsequently modified by enzymes that trim and add sugar residues as glycoproteins proceed out of the ER and through the Golgi complex. A second type of glycosylation results in oligosaccharides added onto serine or threonine residues of proteins, and is termed O-linked glycosylation. This process is thought to be initiated when proteins are enroute from the ER to the Golgi complex (Hirschberg and Snider, 1987; Kornfeld and Kornfeld, 1985). In addition to its role in protein biosynthesis, the ER also monitors newly synthesized proteins for their proper folding and assembly before they are transported out of the ER. This latter function has been termed cellular quality control (Hurtley and Helenius, 1989; Rose and Doms, 1988). Proteins exit from discrete regions of the smooth ER termed transitional elements (Farquhar, 1985),where they transit to a membane compartment known as the intermediate compartment (IC) (Hauri and Schweizer, 1992) before reaching the Golgi complex. At the Golgi complex, proteins may be sorted into several different routes (Mellman and Simons, 1992). Some proteins are retained in the Golgi complex, whereas others are sorted for retrograde transport from the Golgi complex back to the ER. As sorting for retrograde transport appears to occur mainly on the cis side of the Golg complex, this side of the Golgi complex has been termed the cisGolgi network (CGN) (Hsu et al., 1991; Mellman and Simons, 1992). Other proteins continue in the anterograde direction, through the Golgi stacks, to reach the truns-Golgi network (TGN). At the TGN, proteins are either transported directly to the cell surface or diverted toward the endosomal/lysosomal system (Griffiths and Simons, 1986). In specialized cells that possess secretory granules, newly synthesized proteins are diverted from the TGN to these granules (Burgess and Kelly, 1987; Huttner and Tooze, 1989). In addition to being the major sorting station for newly synthesized proteins arising from the ER, the Golgi complex also modfies carbohydrates on glycoproteins. The Golgi glycosidases can also be used to divide the Golgi stacks functionally into cis, medid, and trans stacks. As glycoproteins are transported vectorially through the stacks of the Golgi complex, they are sequentially modified by several glycosidases and glycotransferases. Thus, the glycosylation status of a glycoprotein can be used to monitor biochemically its transport. For instance, transport past the medial-Golgi stack results in glycoproteins becoming modified by N-acetylglucosamine transferase at the medial-Golgi stacks. As a result, these glycoproteins acquire resistance to endoglycosidase H (endo H). Similarly, transport through the TGN results in glycoproteins being modified by sialic acid
372
VICTOR W. HSU AND PETER J. PETERS
additions that render these proteins sensitive to neuraminidase (Hirschberg and Snider, 1987).
B. ENDOCYTIC PATHWAYS Proteins at the cell surface are internalized into endosomes by a process termed endocytosis (Goldstein et al., 1985; Helenius et al., 1983; Pastan and Willingham, 1983).Like the Golgi complex, the endosomes are divided functionally into distinct compartments. Internalized proteins are transported first to the early endosome. The early endosome can be further divided functionallyinto the sorting endosome and the recycling endosome (Geuze et al., 1983; Griffiths et al., 1989). Internalized proteins encounter the sorting endosome first, where proteins in the fluid phase are thought to be carried passively further into the cell, whereas membrane-associated proteins can be sorted for either further transport into the cell or return to the cell surface via the recycling endosome. Proteins that are transported further into the cell reach the late endosome (Salzman and Maxfield, 1988; Schmid et al., 1988). Here, they are then transported either to the lysosome where proteins are degraded or toward the TGN to join the exocytic pathway. Similarly, proteins of the exocytic pathway can be diverted from the TGN to join the endosomal pathway at the late endosome (Geuze et al., 1988). In some cases, proteins can be endocytosed for transport all the way back to the ER (Sandvig et al., 1992). The endosomal system has a pH gradient that becomes progressively more acidic from early endosomes (pH 5.5-6.0) to the lysosomes (pH 4.5-5.0) (Mellman et al., 1986). The acidic pH of the endosomal system serves at least two major purposes. First, acidic pH plays a critical physiologic role in the dissociation of ligands from their receptors. Many receptors bind to their ligand at neutral pH on the cell surface and then release their ligand at an acidic pH at the early endosome. These unbound receptors are recycled back to the cell surface to perform another round of ligand uptake (Dautry-Varsat et al., 1983; Klausner et al., 1983). Second, acidic pH also activates hydrolases that reside in lysosome. These lysosomal hydrolases degrade proteins, a process that is essential for downregulating protein function and recycling amino acids for biosynthesis of new proteins. Lysosomal hydrolases are synthesized in the ER as inactive forms that are subsequently processed proteolytically to their active forms in the acidic environment of the lysosome. This conversion prevents lysosomal hydrolases from wreaking havoc on other proteins when hydrolases are transported through the secretory pathway to the lysosome. As lysosomal degradation represents a major cellular mechanism of downregulating protein function, pharmacologic agents that block lysosomal acidification have been used extensively to study protein regulation (Kornfeld and Mellman, 1989).
373
C U R R E N T VIEWS I N INTRACELLULAR TRANSPORT
II. A General Mechanism of lntracellular Transport
How transport occurs among the various intracelhlar membrane compartments is best understood for a mechanism termed vesicular transport. A general model of vesicular transport has emerged that is conserved in all eukaryotic organisms from yeast to man (Fig. 2). In this model, transport is initiated from one membrane compartment by the recruitment of cytosolic coat proteins onto membranes that result in the deformation of these membrane sites into coated buds. These buds then undergo scission to form mature coated transport vesicles. The membrane-bound coat proteins must then be released to the cytosol before a coated vesicle can fuse with its target compartment. Coated vesicles are directed to fuse with their target compartment by protein complexes that bridge an interaction between the membranes of the vesicle and the target compartment. These complexes are formed from components that reside in both the cytosol and membranes of the vesicle and the target compartment (Rothman and Wieland, 1996; Schekman and Orci, 1996). One critical set of components is a family of transmembrane proteins termed SNARES. Current paradigm for how vesicles fuse specifically with their target compartment is embodied by the SNARE hypothesis, which suggests that fusion is mediated by a specific SNARE on the vesicular membrane that interacts with its partner SNARE on the membrane of the target compartment (Sollner et al., 1993). Emerging evidence suggests that transport also occurs by nonvesicular means. Transport from the ER to the Golgi complex appears to involve membrane structures that have been named vesicular tubular clusters (VTCs) (Balch et al., 1994) or transport complexes (TCs) (Scales et at., 1997). Transport structures of even greater magnitude are suggested by a mechanism that invokes maturation of entire membrane compartments.
Buddincl
'
Formation I TaraLcig
coat protein recruitment
Docking
EmIQ!l
/
coat protein release
FIG.2. General mechanistic steps of vesicular transport.
374
VICTOR W. HSU AND PETER J. PETERS
In this case, transport is accomplished by one membrane compartment evolvinginto another (Salzmanand Maxfield, 1989; Stoorvogel et al., 1991). This process would have the same net result as having transport vesicles shuttling cargo proteins from one membrane compartment to another. However, the mechanistic detail of transport by compartmental maturation is not nearly as well characterized as vesicular transport. Depending on where this mechanism is invoked, it has been termed cisternal maturation (for transport through the Golgi stacks) (Saraste and Kuismanen, 1984) or endosomal maturation (for transport through the endosomes) (Salzman and Maxfield, 1989). These types of transport would require a counter transport mechanism to operate concurrently, as maturation into a different compartment would require proteins that are residents of the previous compartment be transported back to the previous compartment. Thus, counter transport of resident proteins could be mediated by transport vesicles. Another possibility is that transport by organelle maturation would not necessarily exclude concurrent vesicular transport in the same direction. Whatever mechanisms are operative within a given transport pathway, they all appear to use many key regulatory proteins defined from studies of vesicular transport (Scales et al., 1997). These include the coat proteins that form the transport structures. A. COATPROTEINS REGULATING TRANSPORT Thus far, three major families of coat proteins have been well characterized (Table I). The first one discovered is the clathrin coat protein complex, which is composed of six subunits. Two of them are termed clathrin heavy and light chains. They assemble in trimeric pairs that are joined together at one end of the heavy chains and result in an extended structure called the clathrin triskelion. Triskelions assemble into a higher order structure of cage-like hexagonal lattices. This lattice attaches to membranes through the other four subunits of the clathrin coat termed adaptor proteins (AP) or adaptins (Brodsky, 1997; Schmid, 1997). Differences in the adaptin subunits enable clathrin coat proteins to form coated vesicles for more than one transport pathway. Clathrin with AP-1 adaptins form vesicles that transport from the TGN to the late endosome. Clathrin with AP-2 adaptins form vesicles that transport from the cell surface to the early endosome (Kirchhausen et al., 1997; Trowbridge et al., 1993). A third adaptin-like complex (termed AP-3) has been identified (Dell’Angelica et al., 1997; Simpson et al., 1997).They mediate transport from the TGN to the lysosome (Cowles et al., 1997; Simpson et al., 1997).They are also localized on endocytic structures, although the exact transport pathways mediated by these AP-3 complexes remain to be determined (Dell’Angelicaet al., 1997).
CURRENT VIEWS I N INTRACELLULAR TRANSPORT
375
TABLE I COAT PROTEINS
F A M I L I E S OF
Name
Subunits (molecular mass)
Clathrin triskelion AP-1 adaptins
Heavy chain (192 kDa) Light chain (30 kDa) y (100 kDa) j3l (100 kDa) pl (47 kDa) crl (19 kDa) a (100 kDa) j32 (100 kDa) p2 (50 kDa) d2 (17 kDa) 6 (160 kDa) j33 (120 kDa) p3 (47 kDa) cx3 (25 kDa) a (135 kDa) j3 (107 kDa) j3’ (102 kDa) y (98 kDa) 6 (61 kDa) E (36 kDa) t (20 kDa) Sec3lp (150 kDa) Sec24p (105 kDa) Sec23p (85 kDa) Secl3p (34 kDa)
AP-2 adaptins
AP-3 adaptins
COPI
COP11
Transport pathway
TCN-to-late endosome
Plasma mernbrane-to-early endosome
TGN-to-lysosome Early endosome-to-?
ER-to-Golgi”,intra-Golgi, and Golgi-to-ER
ER-to-Golgi”
” Current evidence suggests that COPT mediates more specifically IC-to-Golgi, whereas COPII inediatrs ER-ta-IC.
The other two major coat protein families are called COPI and COPII (COP for Coat Proteins) (Rothman and Wieland, 1996; Schekman and Orci, 1996). COPS do not appear to form a cage-like lattice on membrane vesicles as the clathrin coat proteins. Although analysis by electron microscopy reveals clathrin-coated vesicles to be highly geometric, COP-coated vesicles appears as a dense fuzzy layer of proteins that coat a vesicle (Orci et al., 1986). COPI is composed of seven subunits, named (Y, p, p’, 7 , 6, E , and 5 (Serafini et al., 1991; Stenbeck et al., 1993). This protein complex forms coated vesicles that transport between the ER and the Golgi complex and among the Golgi stacks (Kreis et nl., 1995). Some subunits of COPI have been shown recently to assemble novel coat protein complexes on endosomal membranes (Aniento et al., 1996; Whitney et nl., 1995). However, the pathway that is regulated by these complexes remains to be de-
376
VICTOR W, HSU AND PETER J. PETERS
termined. COPII is composed of four subunits, termed Secl3p, Sec23p, Sec24p, and SecSlp (Sec for Secretory mutants, as COPII subunits were originally identified in yeast) ( Barlowe et al., 1994). COPII-coated vesicles transport from the ER to the Golgi complex (Bednarek et al., 1996). It should be noted, however, that the precise involvement of both COPI and COPII in transport between the ER and the Golgi complex is currently undergoing some revision (Aridoret al., 1995; Scales et al., 1997). However, regardless of the exact transport step that is regulated by coat proteins, it has become apparent that all intracellular transport pathways will likely require coat proteins. Thus, more families of coat proteins remain to be identified. Coat proteins recognize different sequence motifs on the cytoplasmic portion of transported (cargo) proteins. By coupling the recognition of sequence motifs with the formation of transport vesicles, coat proteins sort cargo proteins into distinct transport pathways that emanate from organelles. Different transport motifs have now been defined for several intracellular pathways (Table 11). An acidic motif that resides on cytoplasmic portions of transmembrane proteins appears to enhance transport of proteins with this motif out of the ER (Nishimura and Balch, 1997). It remains unclear which coat protein complex recognizes this motif. At the Golgi complex some subunits of COPI have been suggested to recognize a phenylalanine-based motif and thereby sort proteins with this motif for TABLE I1 INTRACELLULAR SORTING SIGNALS Sorting Signal
Consensus sequence"
Transport Pathway ~~~
Diacidic Diphenylalanine Dilysine Diarginine Tyrosine based
DXE F/I/L/Y)F KKXX KXKXX XXRR
mdJ NPXY
Dileucine Acidic clusters Ubiquitin addition
L/(I/L/V) (M/V)/L E(D/E)D S(X,)KSS
~~
Out of the ER Intra-Golgi (anterograde) Golgi-to-ER, intra-Golgi (retrograde) Plasma membrane-to-early endosome Golgi-to-ER TCN-to-endosomellysosome Plasma membrane-to-early endosome/TGN TCN-to-basolateral surface TGN-to-endosomellysosome Plasma membrane-to-early endosome Plasma membrane-to-TGN (? via endosomes) Plasma membrane-to-earlv endosome
a A single-letter amino acid code is used, where X is any amino acid and 4 is any bulky hydrophobic residue. Degenerate use of amino acids is indicated within parentheses. The dilysine motif is at the carboy1 terminus of type I transmembrane proteins, whereas the diarginine motif is at the amino terminus of type 11 transmembrane proteins.
CURRENT VIEWS IN INTRACELLULAR TRANSPORT
377
transport anterograde through the Golgi stacks (Fiedler et al., 1996). A mutually exclusive set of COPI subunits are thought to recognize a dilysinebased motif and sort type I transmembrane proteins with this motif back to the ER (Letourneur et al., 1994). Another basic residue-based motif is the diarginine motif that mediates retrograde transport of type I1 transmembrane proteins from the Golgi complex to the ER (Schultze et al., 1994). However, the coat protein complex that would recognize this motif remains uncharacterized. At the TGN, clathrin with AP-1 adaptins recognizes a tyrosine-based motif (Honing et al., 1996; Ohno et al., 1995) and sorts proteins with this motif from the TGN to the late endosome (Letourneur and Klausner, 1992; Lobel et al., 1989). A dileucine-based motif also mediates sorting from the TGN to the late endosome (Johnson and Kornfeld, 1992; Letourneur and Klausner, 1992), but the identity of the coat protein complex(es) that would be predicted to recognize this motif at the TGN remains unclear. Both the tyrosine-based (Glickman et al., 1989; Ohno et al., 1995) and the dileucine-based (Heilker et al., 1996) motifs are also recognized by clathrin with AP-2 adaptins that enables proteins with these motifs to be transported from the cell surface to the early endosome (Letourneur and Klausner, 1992; Lobel et al., 1989). A tyrosine-based motif (Humphrey et al., 1993) and an acidic cluster motif (Vorhees et al., 1995) mediate the sorting of proteins from the cell surface to the TGN. However, it remains unclear whether this sorting results in direct transport of proteins from the cell surface to the TGN or occurs via endosomes. Among the documented pathways, motifs that would dictate transport from the early endosome to the cell surface, and from the late endosome to the lysosome remain uncharacterized. B. THEARF FAMILY OF SMALL GTPASES REGULATING COATPROTEINS The recruitment of cytosolic coat proteins to membranes is regulated by a subfamily of Ras-like GTPases known as ADP-ribosylation factor (ARF) (Boman and Kahn, 1995; Donaldson and Klausner, 1994). The name ARF was derived originally from studies in vitro that identified ARFl as a cofactor for ADP ribosylation of the a! subunit of heterotrimeric GTPases by bacterial toxins (Kahn and Gilman, 1984). The physiologic relevance of this finding remains obscure, as compared to the well-defined role of A R F l in regulating coat proteins. At present there are at least 10 known ARF-like members. They include six ARF proteins, three ARL proteins (ARL for ARF-Like), and Sarlp (Boman and Kahn, 1995; Nakano and Muramatsu, 1989). A R F l recruits COPI (Donaldson et al., 1992; Palmer et al., 1993) and clathrin AP-1 (Stamnes and Rothman, 1993;Traub et nl., 1993), whereas Sarlp recruits COP11 (Barlowe et a[., 1994) onto membranes to form coated vesicles. The small GTPases predicted to recruit
378
VICTOR W. HSU AND PETER J. PETERS
AP-2 and AP-3 have yet to be identified. Other than ARFl and Sarlp, members of the ARF-like family of small GTPases have all been identified by homology cloning. Thus, coat protein complexes predicted to be regulated by the other ARFs remain to be identified. Evidence indicates possible functional redundancy, as ARF3 appears to function in many respects similar to A R F l (Morinaga et al., 1996; Taylor et aZ., 1992). However, much of the experimental data on ARFs come from in vitro studies, where the subtleties of ARF action in vivo may be obscured (Donaldson and Klausner, 1994).In this regard, the only other ARF that has been characterized appreciably in vivo is ARFG. It localizes to the plasma membrane and early endosome (D’Souza-Schorey et al., 1998; Peters et al., 1995). Functional analysis suggests that ARFG regulates transport between these two compartments (D’Souza-Schoreyet al., 1995; D’Souza-Schorey et al., 1998). A coat protein complex predicted to be regulated by ARFG remains to be characterized. How ARF-like GTPases regulate the recruitment of their respective coat proteins to membranes is probably best understood for how ARFl recruits COPI (Fig. 3). ARFl activation by its binding of GTP (ARF1GTP) recruits COPI from the cytosol to membranes (Donaldson et al., 1992; Palmer et aZ., 1993), and this recruitment results in coated buds that eventually mature into coated vesicles. Subsequently, ARFl inactivation by the hydrolysis of the bound GTP to GDP (ARF1-GDP) results in vesicle uncoating by the release of COPI from membranes back to the cytosol (Tanigawaet al., 1993). Like all small GTPases, the GTPase cycle of ARFl is regulated by a guanine nucleotide exchange factor (GEF) that catalyzes ARFl activation and a GTPase-activating protein (GAP) that catalyzes ARFl inactivation. Thus far, three GEFs have been identified for A R F l
f
GTP
I Cytosol
GDP FIG.3. Recruitment of cytosolic COPI to membranes is regulated by the GTPase cycle of ARF1.
C U R R E N T VIEWS IN I N T R A C E L I J J I A R T R A N S P O R T
379
(Chardin et al., 1996; Meacci et al., 1997; Morinaga et al., 1996). However, they have been identified based on in vitro guanine nucleotide exchange assays. Thus, it remains unclear which of these GEFs act truly as ARFl GEF in vivo. This problem has been highlighted for ARNO, which had been identified originally as an A R F l GEF in vitro (Chardin et al., 1996). A GEF would likely direct an ARF for its locahzation on membrane. Thus, the finding that ARNO is localized to the cell surface rather than the Golgi region questions whether ARNO would be a GEF for A R F l in vivo because ARFl is localized to the Golgi region and not to the cell surface (Frank et nE., 1998). An effector of activated A R F l is phospholipase D (PLD), which can alter the phospholipid composition of membranes (Brown et al., 1993; Cockcroft et al., 1994). The lipid bilayer has, as its major constituents, four forms of phospholipids that include phosphatidylcholine (PC). PLD converts PC into phosphatidic acid (PA).COPI coat proteins have increased affinity for membranes with increased proportion of PA (Ktistakis et al., 1996).Thus, it has been proposed that activation of PLD by A R F l converts PC to PA in the lipid bilayer. Such a conversion in localized regions of membranes would result in COP1 recruitment to form coated buds. Other factors would also modulate the affinity of COPI to membranes, such as its binding to cytoplasmicdilysine motifs on transmembrane proteins. Aside from PA, COPI also binds to phosphoinositides. Particularly, a-COP has preferential affinity for phospliatidylinositol trispliosphate ( PIPJ) (Chaudhary et al., 1998). This interaction may modulate the affinity of COPI for its transport motifs, similar to that suggested for phosplioinositides regulating the affinity of clathrin adaptins for the tyrosine-based motifs (Rapoport et al., 1997). Although a molecular description of how coat proteins are recruited to membranes by the activation of ARF is beginning to be appreciated, it remains less clear how the inactivation of ARF induces coat proteins to detach from membranes. The mechanistic steps of coat protein detachment is probably best understood for the COPII system. After the formation of COPII-coated vesicles, Sarlp becomes inactivated by its GAP (Sec23p) that results in the release of Sarlp to the cytosol. However, a vesicular intermediate exists where transport vesicles have COPII, but lack Sarlp. Only subsequently is COPII released from membranes of transport vesicles to allow these vesicles to fuse with their target compartment (Barlowe et al., 1994).Thus, it appears that the inactivation of Sarlp does not directly result in the uncoating of COPII-coated vesicles. One possibility is that targeting of the transport vesicle by its doclang with the target compartment provides the stimulus to uncoat the vesicle. This would suggest that compo-
380
VICTOR W. HSU AND PETER J. PETERS
nents of the docking complex might initiate the uncoating of vesicles. However, this possibility remains to be tested experimentally. C. REGULATION OF ARF GTPASES Evidence suggests that the GEFs and GAPS of ARF do not function autonomously in driving the GTPase cycle of ARF. Their catalytic activities are regulated by products of lipid metabolism. Specifically, the catalytic activity of ARNO is enhanced by phosphatidylinositol bisphosphate (PIP2) (Chardin et al., 1996), whereas the catalytic activity of ARFl GAP is enhanced by diacylglycerol (DAG) (Anntony et al., 1997). As DAG can be a downstream product of PLD action, a negative regulatory loop can exist for the action of ARFl that is based on lipid metabolites. For instance, activated ARFl that stimulates PLD may ultimately generate DAG, which activates ARFl GAP and results in ARFl inactivation. Additionally, the function of ARFl regulators may be modulated by select classes of proteins that are being transported by the intracellular pathways. This is suggested by studies on a class of soluble proteins, termed KDEL proteins, that cycle between the ER and the Golgi complex (Pelham, 1991). Their cycling is essential for cellular homeostasis, but the exact nature of this requirement had remained elusive (Townsley et al., 1994). It has been shown that retrieval of KDEL proteins by the KDEL receptor regulates the interaction of the KDEL receptor with an ARFl GAP (Aoe et al., 1997, 1998). This finding suggests that select cargo proteins, such as the KDEL proteins, provide a regulatory signal to the pathways that transport them. Such a regulatory mechanism bears fundamental resemblance to a general signal transduction process, as ligands (KDEL proteins) on the luminal side of the membrane bind to the transmembrane KDEL receptor that then transduces a signal to cytosolic transport regulators by recruiting one of these regulators (ARF1 GAP) to the membrane. Of particular significance is that such a finding may provide a mechanistic explanation for how transport and orgenelle structure are regulated coordinately. Studies have suggested that organelles are dynamic structures maintained in steady state by the balance of membranes provided by transport carriers that feed into and emanate from these organelles. In support of this view, studies that perturb the GTPase cycle of ARFl provide insight into how both transport and organelle identity can be regulated coordinately. For the Golgi complex, perturbations that inactivate ARFl induce the redistribution of the entire Golgi complex to the ER (Dascher and Balch, 1994; Doms et al., 1989; Lippincott-Schwartz et al., 1989; Peters et al., 1995). This result has been interpreted as due to perturbing the balance of anterograde and retrograde transport between the ER and the Golgi complex to favor retrograde transport. Thus, the Golgi com-
plex eventually loses all its membranes to the EK (Klausner et al., 1992; Lippincott-Schwartz, 1993).An intriguing possibility is that KDEL proteins that are transported out of the ER are used by the Golgi complex to sense the degree of incoming membrane transport from the ER through their binding and activation of the KDEL receptor. Consistent with this hypothesis, KDEL proteins are abundant and ubiquitous (Pelham, 1989). Thus, they are likely to be in many transport vesicles that emanate from the ER. Moreover, constitutive activation of the KDEL receptor also results in the redistribution o f the Golgi complex to the ER (Hsu et nl., 1992). Thus, the activation status of the KDEL receptor may provide a mechanism to sense anterograde transport from the ER to the Golgi complex that is then transduced through its interaction with A R F l GAP to regulate retrograde transport from the Golgi coniplex to tlie ER. D. SNARES MEDIATING VESLCLEFUSION Although coat proteins regulated by the ARF family of sinall GTPases determine the formation of transport vesicles, how these transport vesicles target properly to fuse with the correct compartment is regulated by other families of conserved proteins. One family consists of transmembrane proteins, termed SNARES. A specific vesicular-SNARE (V-SNARE)on a transport vesicle will recognize a specific target-SNARE (t-SNARE) on the target coinpartment. Thus, a vesicle fuses specifically with its target compartment because of a physical interaction between a V-SNAREand its corresponding t-SNARE (Pfeffer, 1996; Rothman, 1994). SNAREs were originally identified as a requisite component for transport vesicles to fuse with their target compartment. Reconstitution of transport in an in vitro transport assay has resulted in the isolation of many molecular components that are critical for the fusion process (Balch et al., 1984). Initially, a cytosolic protein was found to be critical for vesicle fusion, and the function of this protein was inliihited by a reducing agent named N-ethylmaleimide (NEM). Therefore, this protein was termed NEMsensitive factor ( N S F ) (Block et al., 1988). Other cytosolic factors were necessary for the function of NSF and thus were identified as soluble NSF associated proteins (SNAPs) (Clary et nl., 1990). For a transport vesicle to fuse with its target coinpartment, cytosolic NSF and SNAPs are recruited into a membrane-bound complex that bridges the membranes of the vesicle and the target compartment (Weidman et al., 1989). Unlike ARFs and coat proteins that have multiple family members that function similarly in different transport pathways, the same NSF and SNAPs are used universally as part of the fusion machinery (Wilson et al., 1989). Thus, a key question became whether there would he pathway-specific components that regulate fusion by N S F and SNAPs at distinct intracellular
382
VICTOR W. HSU AND PETER J. PETERS
sites. Subsequent purification of a molecular complex that contained both NSF and SNAPs revealed the presence of transmembrane SNAREs (for SNAP receptors) (Sollner et al., 1993).This has led to the SNARE hypothesis that suggests how the SNAREs in this complex would provide specificity to ensure that vesicles fuse with their proper target compartment. Pairing of the correct SNAREs during the docking of transport vesicles to their target compartment is accompanied by the recruitment of cytosolic NSF and SNAPs to form a large multimeric complex, termed the SNARE fusion complex. Cytosolic NSF is recruited into this complex by its binding of ATP. Upon fusion between membranes of the transport vesicle and target compartment, ATP that is bound to NSF is hydrolyzed to ADP. This hydrolysis results in both NSF and SNAPs being released back to the cytosol and the unpairing of the cognate SNAREs. However, for a target compartment to maintain its unique composition of t-SNARES in the face of V-SNARESthat are constantly supplied by incoming transport vesicles, V-SNARESare transported froin the target compartment back to their originating compartment. In the returning vesicles, these V-SNARES are inactivated so that they do not compete with another set of v-SNARES that are directing the fusion of the returning vesicles (Pfeffer, 1996; Rothman and Warren, 1994). SNAREs E. Rab SMALLGTPases REGULATING SNAREs are regulated by another family of sinall GTPases named Rabs (Brennwald et al., 1994; Lian et al., 1994; Sogaard et al., 1994). Homology cloning has identified more than 40 Rabs to date (Novick and Zerial, 1997). Distinct Rabs function in specific transport pathways to enable a particular set of transport vesicles to fuse with its target compartment. Rabs are cytosolic proteins that attach to membranes of a transport vesicle on their activation by binding of GTP. Activated Rab induces the ability of the correct pair of v-SNARE and t-SNARE to interact (Brennwald et al., 1994; Lian et al., 1994; Lupashin and Waters, 1997; Sogaard et al., 1994). Upon membrane fusion, the GTP-bound form of Rab is hydrolyzed to GDP that results in SNARE inactivation. The inactivated SNAREs that have become integral membrane proteins in the target compartment are then transported back to their point of origin to mediate another round of vesicle fusion. Rabs are recycled by a different mechanism. The membrane-associated, GDP-bound form of Rab is extracted out of the membrane to the cytosol by a quanine nucleotide dissociation inhibitor (GDI).The exchange of GDP for GTP on Rab causes its dissociation from cytosolic GDI. The activated Rab associates with membranes of newly formed transport vesicles to mediate another round of vesicle fusion (Soldati et al., 1994; Ullrich et al., 1994).
C U H K E N T \ W W S IN I N T R A C E L L U I A R T R A N S P O R T
383
The cognate pairs of v- and t-SNARES are originally thought to be sufficient for both vesicle targeting and fusion. However, it has become increasingly clear that SNARESmediate more specifically the fusion event and that Rabs regulate the SNARE-mediated fusion indirectly by modulating a more upstream process that docks the vesicle to its proper target coinpartinent (Barlowe, 1997; Lupashin et nl., 1996). In this regard, fainilies of proteins on both vesicles and target compartments have been identified that are suggested to mediate vesicle targeting (Hay et al., 1997; TerBusli et al., 1996). The interaction of these proteins, one on the vesicle and another on the target compartment, would target a vesicle to dock to its proper destination. Where vesicle uncoating fits in this scheme reinaim largely an unexplored question, but it needs to occur before vesicle fusion because only uncoated vesicles can fuse with their target compartment. 111. Complexities of Transport in Vivo
A niajor goal in studying transport has been to understand the roles of COP1 and COPII coat proteins in the early secretory system. In vitro reconstitution studies and yeast genetics originally suggested that COP1 mediates anterograde transport from the ER to the Go18 complex, transport through the Golg stacks, and retrograde transport froin the Golgi complex to the ER. However, COPII appears to mediate only anterograde transport from the ER to the Golgi complex. Does this mean that there is a redundant class of transport vesicles going from the ER to the Go@ complex? More recent evidence suggests that redundancy is not the answer. There are at least two competing models for how COPI and COPII might function in a mutually exclusive manner. One model explains transport mainly by vesicular means. Thus, anterograde transport from the ER to the Golgi complex is mediated niainly by COPII-coated vesicles, whereas anterograde and retrograde transport through the Golgi stacks and retrograde transport froin the Golgi complex back to the ER is mediated by COPI-coated vesicles. An alternate model that invokes both vesicular and nonvesicular transport has gained support from recent studies. They suggest that transport out of the ER is mediated by COPII-coated vesicles that coalesce to form larger inembrane structures, termed vesicular tubular clusters (VTCs) (Aridor et nl., 1995) or transport complexes (TCs) (Scales et al., 1997),that are thought to be equivalent to the intermediate compartment (IC). At this point, COPII becomes replaced with COPI, and these large membrane structures move along inicrotubules to the Golgi complex. Subsequent transport through the Golgi stacks would occur by cisternal
384
VICTOR W. HSU AND PETER J. PETERS
maturation. However, retrograde transport through the Golgi stacks and back to the ER would occur by COPI-coated vesicles. Thus, this model suggests that COP11 mediates anterograde vesicular transport from the ER to the IC, and COPI mediates retrograde vesicular transport from the Golgi complex and the IC back to the ER. In contrast, anterograde transport from the IC through the Golgi stacks would occur through large membrane structures that are coated by COPI. A. ROLEOF CYTOSKELETONS The transport of large membrane structures in the early secretory system seems to require microtubules (Presley et al., 1997; Scales et al., 1997). This finding may explain the enigmatic role of cytoskeletons in transport. In vitro reconstitution experiments had suggested that cytoskeletons were not essential for many pathways of vesicular transport (Rothman and Wieland, 1996; Schekman and Orci, 1996). These findings conflict with in vivo evidence that suggests cytoskeletons do play a role in intracellular transport (Lamaze et al., 1996, 1997). Moreover, studies of yeast mutants that are defective in endocytosis have uncovered several mutations in components of the actin cytoskeleton (Geli and Riezman, 1996; Kubler and Riezman, 1993). Similarly, although ARF GTPases are defined regulators of membrane traffic, ARF6 has also been shown to affect cytoskeletal organization (D’Souza-Schoreyet al., 1997; Radhakrishna et al., 1996). Also, although Rho GTPases are known to regulate cytoskeletal organization, they have been shown to affect membrane traffic (Lamaze et al., 1996; Murphy et al., 1996; Van Aelst and D’Souza-Schorey, 1997). Thus, one possibility is that while vesicular transport may not require cytoskeleton, a given transport pathway may be served by other forms of transport carriers that do require cytoskeleton. This possibility implies that more than one type of transport carrier can serve a given direction in a transport pathway. Multiple types of carriers may enhance the efficiency of transport in uivo. This possibility would explain effects on intracellular transport by disrupting cytoskeletal organization. For instance, when microtubules are disrupted by pharmacologic agents such as nocodazole, ER-to-Golgi transport is only temporarily delayed and protein targeting seems unaffected (Cole et al., 1996). This observation can be explained if transport to the Golgi complex is served by large membrane structures from distant ER sites that require microtubules and by vesicles from closer ER sites that do not require microtubules. Thus, because microtubules are not needed over short distances of transport by vesicles, vesicular transport can be increased over time to compensate for the loss of transport that requires microtubules. As nocodazole does not disrupt microtubules that hold together the Golgi stacks (Pavelka and Ellinger, 1983), transport through the Golgi stacks
CURRENT VIEWS IN INTRACELLULAK TRANSPORT
385
would not be affected by nocodazole. In this regard, a circumstance when microtubules are truly dispersed from holding the Golgi stacks together is during mitosis, when all transport ceases (Warren, 1993).
B. TRANSPORT RECULATINC ORGANELLE STRUCTURE Another emerging concept in the regulation of transport pathways is that what happens in one direction of a particular pathway often affects transport in the other direction of a bidirectional transport pathway (Cosson and Letourneur, 1997). In this regard, there is increased appreciation that most experimental readouts cannot distinguish a direct effect versus an indrect effect on one direction of a bidirectional transport pathway (Gaynor and Emr, 1997). However, we are beginning to appreciate at least why this coupling in transport regulation exists. As many essential regulatory components are transmembrane proteins, such as SNARES,these components need to be recycled back to their site of function by transport in the counter drection (Pfeffer, 1996). In addition, organelles, such as the Golgi complex, and endosornes are dynamic structures maintained in steady state by membranes that feed into and out of these organelles by bidirectional transport pathways (Klausner et al., 1992; Lippincott-Schwartz, 1993; Mellman and Simons, 1992). Thus, transport out of an organelle can only be maintained homeostatically if it is coupled to transport that feeds into the organelle. IV. Secretory Pathways
Although a protein may fold into several tertiary conformations, only one conformation will likely confer the biological activity of the protein. Current evidence suggests that the ER has several classes of proteins that facilitate the proper folding and assembly of newly synthesized proteins. Some ER proteins enhance the rate of protein folding and thus function as folding catalysts. Other ER proteins facilitate the fraction of newly synthesized proteins that reach their proper folding and assembly and thus function as molecular chaperones. These ER proteins belong to the family of heat-shock proteins, termed hsp70. They are so named originally because their levels are increased by various cellular stresses, such as heat or cold shock, and glucose starvation. The hsp70 proteins are abundant and ubiquitous. They have members in both the ER lumen and the cytosol (Gething and Sambrook, 1992; Hartl, 1996). A. PROTEIN ASSEMBLY BY MOI~ECULAK CHAPERONES: STUDIES O N THE IMMUNOGLOBULIN-BINDINC: PROTEIN A member of the lisp70 Camily is the immunoglobulin-binding protein (BiP). This protein was so named because it was found originally as an
386
VICTOR M’. HSU AND PETER J. PETERS
ER protein that regulated the secretion of immunoglobulin (Ig) heavy chains (Haas and Wabl, 1983).B lymphocytes (B cells) go through developmental stages that transform them into a mature B cell that can secrete Igs, either onto the cell surface for a mature B cell or to the extracellular environment for plasma cells (Rolink and Melcher, 1991). It was noted that at even earlier stages of B-cell development, a pre-B cell synthesized Ig heavy chains, but not the light chains. The heavy chains in these preB cells were retained in the ER, where they were bound to BiP (Haas and Wabl, 1983). They became released from BiP upon the developmentally regulated synthesis of Ig light chains that allowed the complete complex of heavy and light chains to be assembled and transported out of the ER (Bole et al., 1986; Mains and Sibley, 1982). Subsequent characterization of BiP suggests that it functions more generally as a chaperone in the folding and assembly of many newly synthesized proteins in the ER (Copeland et al., 1986; Gething et al., 1986; Kassenbrock et al., 1988).How BiP accomplishes this role is beginning to be understood. Some insights have been derived from studies of its homolog in bacteria, DnaK. It is an ATPase, whereby its binding of ATP regulates its interaction with folding polypeptides. When bound to ATP, DnaK binds and dissociates from its substrate polypeptides rapidly, whereas the ADP-bound form of DnaK binds and dissociates from polypeptides more slowly (Palleros et al., 1991; Schmid et al., 1994). The cycle of ATP binding by DnaK is catalyzed by factors that have mechanistic parallels to the role of GEF and GAP in catalyzing GTP binding by small GTPases. Thus, DnaJ accelerates the hydrolysis of ATP by DnaK, whereas GrpE enhances the dissociation of ADP from DnaK (McCarty et d., 1995). Current views suggest that proper protein folding accommodates two opposing forces. Although a nascent polypeptide chain progresses to its proper folding conformation, its folding intermediates are subject to aggregation that is characteristic of misfolded proteins. The lisp70 proteins assist in protein folding by binding to portions of a polypeptide that have not folded completely into a domain. This function prevents unfolded portions that are often hydrophobic from being exposed to the surrounding aqueous environment that would result in the irreversible aggregation of folding intermediates. Thus, the role of hsp70 proteins can be viewed as “solubilizing” a folding protein to prevent it from being aggregated irreversibly into an “insoluble” state (Hartl, 1996). Consistent with this mechanism, when a peptide library was used to screen for characteristics that governed the binding of BiP to its substrate polypeptides, the selected peptides had hydrophobic characteristics (Blond-Elguindi et al., 1993). Aside from its role in promoting the correct folding of a protein, the ATP binding cycle of BiP seems to serve another major role in the ER. BiP binds to newly
C U R R E N T \'IEWS IN INTHACELLULAH T H A N S P O R T
387
synthesized proteins that are translocating into the ER lumen. In this context, it has been suggested that the ATP cycle of BiP serves as a molecular ratchet that induces a progressively greater portion of a translocating protein being bound to BiP (Sanders et nl., 1992). In this manner, translocating proteins are transferred vectorially from the cytosol to the ER lumen.
B. ENDOPLASMIC RETICULUMRETENTIONA N D DEGRADATION: STUDIES O N THE T-CELLANTIGENRECEPTOR Much of our understanding of how proper protein assembly affects transport has been contributed by studies on the T-cell antigen receptor (TCR).T cells recognize specific antigens presented as antigenic peptides bound to major liistocoinpatibility complex (MHC) molecules on the surface of an antigen-presenting cell (APC). The specificity of this antigen recognition is provided by the TCR. The subunits of the TCR can be divided functionally into two types of complexes. One set consists of heterodimers that are integral membrane proteins. Their lurninal domains possess polymorphic sequences that enable the TCR to distinguish combinations of different antigens bound to polymorphic MHC molecules. Typically, the two main forms are called TCR-a//3 and TCR-y/6. The other subunits are invariant and are termed CD3-y, 6, E , and the family that consists of 7,and Fc-y. These CD3 subunits are also integral membrane proteins that serve mainly to transduce the signal of antigen binding by the polymorphic TCR heterodimers to intracellular metabolic pathways of the T cell. Studies on the intracellular fate of the TCR have provided insights into inechanisms that regulate protein assembly, transport, and degradat'ion. These mechanisms contribute to ensure that only properly assembled TCRs are transported out of the ER and underlie the concept of cellular quality control (Klausner et al., 1990). Analysis of which subunits of the TCR interact with each other to build the complete complex was first done by examining mutant T-cell lines that expressed only select subunits (Bonifacino et al., 1988). These results were confirmed subsequently by examining different combinations of TCR subunits of the dfl-containing TCRs that were expressed heterologously in fibroblasts (Manolios et al., 1990). Results of these studies suggest that dimeric complexes are formed first. Preferential pairings include virtually any combination among a, /3, 7, 6, E chains, with a notable exception of y with 6. There was a preference by which the diineric coniplexes were subsequently assembled into higher order complexes. Thus, a//3can associto forin C Y / / ~ / ~ / or I E with Y / E to form a//3/y/&. In either case, ate with 6 / ~ the remaining dimer then assembles to form a hexameric ~ / ~ / Y / E / S I E (Manolios et nl., 1991). Although individual subunits cannot exit the ER,
c,
c
388
VICTOR W HSU A N D PETER J PETERS
their assembly into a hexameric complex renders them competent to leave the ER (Bonifacino et al., 1989; Minaini et al., 1987). In the ER, the [ family of dimers, which consist of pairings among f; 7,Fc-y will only assemble as the final dimer to complete the TCR complex ( Sussman et al., 1988;Weissman et al., 1989).Significantly,y and S that have assembled into dimers still prevent these dimers from associating, such that $6 does not associate with Sle. More detailed examination of this finding revealed that S and y contain transmembrane positively charged residues that seem responsible for repulsing the pairing of these two chains. This repulsion appears to be neutralized only when assembly into a hexameric complex is achieved (Manolioset al., 1991). Thus, these studies suggest that individual chains of the TCR possess determinants that regulate the order by which a multimeric complex is assembled. Moreover, these chains possess determinants, or motifs, that normally retain the individual chains in the ER, but their assembly into oligomeric complexes hides these determinants and enibles an assembled complex to leave the ER. Studies on the TCR also contributed to understanding how proteins are degraded in the ER. ER degradation is a universal mechanism of disposing proteins that do not succeed in either folding or oligomerizing in the ER (Bonifacino and Lippincott-Schwartz, 1991). Degradation occurs by targeting those proteins from the ER lumen to the cytosol (Kopito, 1997). Other immunologically related proteins that have been documented to be disposed by ER degradation include CD4 (Willey et al., 1992), MHC class I (Moore and Spiro, 1993) and class I1 (Koppelman and Cresswell, 1990) molecules, Fc-y receptor (Lobell et al., 1993), and immunoglobulin chains (Sitia et al., 1987). Studies on the a chain of TCR revealed that distinct determinants target an ER protein for retention versus degradation. The TCR-a chain is a type I integral membrane glycoprotein with a luminal domain, a transmembrane domain, and a cytoplasmic domain. Chimeras of a chain were generated by swapping with the domains of a topologically similar protein, the interleukin-2 receptor a subunit (Tac antigen). This protein is normally transported out to the cell surface as a single chain, thus suggesting that Tac contains neither retention nor degradation determinants. Chimera studies suggested that the luminal domain of TCR-a had an ER retention determinant, whereas the transmembrane domain possessed an ER degradation determinant (Bonifacino et al., 1990). Consistent with this finding, a truncated form of the a chain that contained only its luminal domain was retained stably in the ER. Serial truncation suggested that a peptide determinant on the second immunoglobulin-like domain of a chain was recognized by BiP. Binding to BiP resulted in ER retention of the a chain (Suzuki et al., 1991).
C U R R E N T VIEWS I N INTRACELLULAR TRANSPORT
389
Studies on the transmembrane domain of the Q chain revealed that a charged residue in the middle of the transmembrane domain was responsible for the degradation of the a! chain. In support of this finding, mutating Tac, by introducing a charged residue in the transmembrane domain, conferred degradation onto mutated Tac (Bonifacino et aE., 1990). The position of the charged residue influenced the efficiency of degradation, as there seemed to be a window within the middle of the transmembrane domain where an introduced charged residue resulted in degradation (Bonifacino et al., 1991). The length of the transmembrane domain also influenced ER retention and degradation, as decreasing the length of the transmembrane domain enhanced both processes (Lankford et al., 1993). The molecular machinery that is targeted by determinants of ER degradation has been elucidated. A hint for how this mechanism occurs comes from the observation that a gene product of cytomegalovirus induces ER degradation of MHC class I molecules. Subcellular fractionation revealed that intermediate products of degradation were translocated from the lumen of the ER to the cytosol through the same translocation machinery used to import all newly synthesized proteins from the cytosol to the ER lumen (Wiertz et al., 1996a,b). Degradation proceeds by the covalent attachment of a peptide, termed ubiquitin, to arginine residues on ER proteins that have translocated out to the cytosol (Hiller et al., 1996; Ward et al., 1995; Wiertz et.al., 1996).These ubiquitinated proteins are targeted to a large multisubunit complex termed the proteasome (Ciechanover, 1994). The proteasome lies on the cytosolic face of ER membranes. Its cylindrical structure establishes a microenvironment inside that efficiently degrades proteins into peptides (Bauineister and Lupas, 1997). The Q chain of the TCR has also been shown to be translocated from the ER lumen to the cytosol for degradation (Huppa and Ploegh, 1997). Thus, it appears that the charged residue in the transmembrane domain of TCR subunits targets the proteins to the ER translocation machinery that then extrudes a protein out of the ER lumen for degradation by the proteasome. Currently, it remains unclear how the ER translocation machinery recognizes the charged degradation determinant, whether this recognition might be direct or other factors might recognize the charge determinant to bring a protein to the translocation machinery.
C. BIDIRECTIONAL TRANSPORT IN THE EARLY SECRETORY SYSTEM Initial studies of anterograde transport from the ER to the Golgi complex suggest that proteins are transported out of the ER without a sorting signal. Thus, it was proposed that transport out of the ER was by default. This was conceptualized by the bulk flow hypothesis, which suggests that proteins leaving the ER toward the cell surface are transported much like an
390
VICTOR W. HSU AND PETER J , PETERS
object floating downstream in a river. Proteins are actively diverted from this flow if they possess either a retention motif to stay at a specific organelle along the secretory pathway or a sorting motif to divert them elsewhere (Pfeffer and Rothman, 1987). However, more recent evidence suggests that active sorting also occurs for transport out of the ER. Analysis of a prototypic transmembrane protein that is transported to the cell surface suggests that it is concentrated into transport vesicles that emanate from the ER. This concentration appears to be mediated by a diacidic-based motif that resides on the cytoplasmic tail of many transmembrane proteins (Balch et al., 1994; Nishimura and Ralch, 1997). Early evidence for retrograde transport comes from studies of a class of abundant, soluble ER proteins that have a common carboxy-terminal sequence of KDEL. These proteins, dubbed KDEL proteins, help in the proper folding and assembly of newly synthesized proteins in the ER, such as that discussed for BiP (Pelham, 1989).They are thought to be retained in the ER in a calcium-dependent manner (Booth and Koch, 1989). However, this ER retention mechanism is not perfect. At any given time, a fraction of KDEL proteins is transported out of the ER and then retrieved from the Golgi complex to the ER (Munro and Pelham, 1987; Pelham, 1988). A transmembrane receptor at the Golgi complex recognizes the KDEL sequence to retrieve the KDEL proteins (Lewis and Pelham, 1992; Semenza et al., 1990). However, this mechanism then begs the logical next question: how is the transmembrane KDEL receptor recognized and sorted at the Golgi complex for retrograde transport? An answer to this question is suggested by studies on MHC class I molecules. I N THE EARLY SECRETORY SYSTEM: STUDIES ON THE MHC D. SORTING CLASS I MOLECULES
T cells recognize antigenic peptides that are bound to MHC molecules on the surface of APCs. Presentation of antigens can be divided into two major categories based on the meclianism of antigen acquisition by MHC molecules. The endogenous pathway of antigen presentation refers to acquisition of cytosolic antigens that are translocated into the ER lumen, where they bind to MHC class I molecules. The exogenous pathway of antigen presentation refers to the acquisition of antigens endocytosed from the extracellular environment. These antigens bind to MHC class I1 molecules in the endosomaMysosoma1 compartments. Details of the endogenous pathway of antigen presentation have been elucidated greatly over the past decade. Studies suggest that cytosolicproteins become ubiquitinated to be targeted for degradation by the proteasome. The generated cytosolic peptides are then translocated into the ER lumen by a transporter complex, termed transporter of antigenic peptides (TAP).This transporter
CURRENT VIEWS IN INTRACELLLJLARTRANSPORT
391
is distinct froin the general translocation machinery used by all newly synthesized proteins. MHC class I molecules in the lumen of the ER assemble into tripartite complexes that are composed of a type I transmembrane glycoprotein (termed heavy chain) that binds noncovalently to a soluble protein (termed light chain or j32-microglobulin) and a peptide of 8-10 residues (Gerinain and Margulies, 1993). I n some tumor cell lines, it was noted that both heavy and light chains of the class I molecules were synthesized, but they failed to appear on the cell surface. Closer examination of tlie fate of class I heavy chains revealed that they were not assembled into tripartite complexes with j32inicroglobulin and peptide. The unassembled heavy chains seemed to remain in the ER (Klar and Hanimerling, 1989). However, subsequent analysis suggested that these heavy chains exited the ER and reached the Golgi complex before being recycled back to the ER. As these class I heavy chains cycled mainly through the cis side of the Golgi complex, it was suggested that sorting for transport back to the ER occurred mainly on the cis side of the Golgi complex. In this context, the cis side of the Golgi complex became known as the cis-Golgi network (CGN) and was defined as the predominant Golgi region where proteins were transported back to the ER. Sorting at the CGN can distinguish unasseinbled class I heavy cliains froin other itinerant proteins that are transported beyond the Golgi complex. Thus, the Golgi complex imposes another level of cellular quality control beyond the ER to ensure that only properly assembled protein complexes are transported out of the early secretory system (Hsu et nl., 1991). Other studies on the intracellular fate of MHC class I molecules revealed a mechanism by which transmembrane proteins are sorted selectivelyat the Golgi complex for retrograde transport to the ER. A type I transmembrane protein that is a 19-kDa gene product of early adenovirus infection (hence named E19) was found to bind class I molecules and result in their apparent retention in the ER (Andersson et nl., 198s). Further analysis suggested that a cytoplasmic dilysine motif was responsible for maintaining E l 9 in the ER (Jackson et nl., 1990; Nilsson et nl., 1989). Like the cbstribution of unassehbled MHC class I molecules, the ER localization of E l 9 also represented a steady-state distribution because more detailed analysis revealed that El9 was transported out of the ER to the Golgi complex and then retrieved to the ER (Jackson et ul., 1993). The host machinery that recognized the dilysine motif on El9 turned out to be the COP1 coat protein complex (Cosson and Letouimeur, 1994; Letourneur et al., 1994). Thus, studies on El9 provide an example of not only how viruses subvert host mechanisms to their benefit, but also how studies of these viral mechanisms provide critical insight into functions within the cell. The dilysine
392
VICTOR W. HSU AND PETER J. PETERS
motif that is recognized by COPI appears in many type I transmembrane proteins that cycle between the ER and the Golgi complex (Teasdale and Jackson, 1996). A mechanism has been proposed for how two different classes of motifs can be recognized by COPI to mediate bidirectional transport in the early secretory system. A dilysine-based motif enables transmembrane proteins to head retrograde because a subset of COPI subunits, a,p', and E , recognizes this motif. A phenylalanine-based motif on the cytoplasmic tail of transmembrane proteins mediates their transport anterograde, because a mutually exclusive set of COPI subunits, namely b, y , and f; recognizes this motif (Fiedler et al., 1996). V. Endocytic Pathways
Endocytosis has been defined traditionally as a transport process that brings materials into the cell. Mechanistically, this can occur by vesicular transport or by another process termed phagocytosis. Although endocytosis was originally defined to include both types of transport, the usage of the term endocytosis now often implies transport other than phagocytosis. The difference between these two main types of transport is that phagocytosis can bring material into the cell that is much larger than that by vesicular transport. Thus, although transport vesicles are typically about 0.1 pm in diameter, phagocytic structures can be up to several microns in diameter. This difference in size results in important consequences for how these two transport processes differ mechanistically. The initial step of vesicular transport involves a localized inward curvature of membrane to form buds (such as clathrin-coated pits). In contrast, the initial step of phagocytosis involves the cell surface enveloping around a structure that is to be endocytosed. As this structure often approaches the size of the cell itself, a cell must reorganize its shape extensively so that its plasma membrane can wrap around the target structure (Swanson and Baer, 1995). Thus, this process has a critical requirement for the actin cytoskeleton. The completion of phagocytosis results in an endocytic compartment termed the phagosome. After the initial formation of a phagosome, proton-ATPases are also detected. As a phagosome evolves into a phagolysosome by the acquisition of lysosomal proteases, acidification by proton-ATPases is critical for activating many of these proteases (Sinai and Joiner, 1997). Two mechanistic possibilities can be envisioned for how phagosomes evolve into phagolysosomes. A phagosome can simply fuse directly with a lysosome to form a phagolysosome. Alternatively, vesicular transport that selectively transports components out and into a phagosome may cause it to mature into a phagolysosome. Evidence suggests that the latter mechanism is more likely (Desjardins et al., 1994). Some mycobacteria have evolved mechanisms to
C U R R E N T VIEWS IN I N T R A C E L L U U R TRANSPORT
393
selectively prevent proton-ATPases from appearing in the evolving phagosome (Sturgill-Koszycki et al., 1994). This result likely contributes to the ability of mycobacteria to survive in their host cell (Cleniens and Horwitz, 1995). A. MECHANISMS OF ENDOCYTOSIS Endocytosis has been studied the most extensively for transport by clathrin-coated vesicles. Activation of many cell surface receptors induces their association with clathrin AP-2 adaptins and results in the endocytosis of the activated receptors (Pearse and Robinson, 1990; Schmid, 1997). Emerging evidence suggests that other proteins can serve as functional homologs of AP-2. Studies on a large family of G protein-coupled seven transmembrane receptors that includes the adrenergic receptors suggest that cytosolic arrestins interact with the cytoplasmic tail of these receptors and promote recepltor endocytosis (Ferguson et al., 1996). It has been discovered that p-arrestin causes endocytosis of adrenergic receptors by physically coupling the receptors to clathrin triskelions. Thus, arrestins act as another class of adaptins (Goodman et al., 1996). In this regard, the HIV gene product Nef may cause CD4 to be internalized through a similar mechanism (Foti et al., 1997). The same stereotypic steps of coated vesicle formation, docking, and fusion that have been worked out for COP-mediated transport are thought to occur for clathrin-mediated transport (Schmid, 1993). In vitro reconstitution assays suggest that the recruitment of adaptins to membranes is affected by GTP. However, unlike COP and AP-1 recruitment to membranes that is enhanced by GTP, recruitment of AP-2 adaptins to the plasma membrane is inhibited by GTP (Carter et nE., 1993).Thus, it rernaiiis unclear whether an ARF-like small GTPase is required to recruit adaptins onto membranes. In vitro reconstitution assays revealed an additional requirement for GTP late in the formation of clathrin-coated vesicles. A GTP-binding protein named dynamin has been identified to fulfill this requirement (Herskovits et nl., 1993; van der Bliek et al., 1993). Isoforms of dynamin have been identified, and one possibility is that they subserve similar functions in other clathrin-mediated transport pathways (Urrutia et al., 1997). A dynamin that cannot hydrolyze GTP forms a collar-like structure around the neck of a coated bud as it is about to undergo scission to form a coated vesicle (Hinshaw and Schmid, 1995; Takei et al., 1995). However, the dynamin collar does not simply act mechanically in scission of the bud. Rather, dynamin recruits several cytosolic proteins to the bud, and these proteins are thought to mediate the scission process. Many of these recruited components have been identified from brain. Of particular interest is synaptojanin, which is an inositol-5-phosphatase (McPherson et
394
VICTOR 1%‘.IISU AND PETEH J . PETERS
al., 1996). In this regard, it is worthwhile to note that phosphoinositides have emerged as critical regulators of transport (De Camilli et al., 1996). Thus, inositol-5-phosphatasewould likely play an important role in regulating the formation of coated vesicles, as they can convert 5’-phosphorylated inositol forms of PIPz and PIP3 into PIP and PIP2, respectively. Although endocytosisby nonclathrin-mediated transport is less well characterized, it is being increasingly appreciated. It remains unclear how many different types of transport are embodied by this term. One class can be defined pharmacologically because its action is unaffected by cytosolic acidification and potassium depletion that characteristically affect clathrinmediated transport (Sandvig and van Deurs, 1994). Another class is implicated by the identification of morphologically distinct structures termed caveolae (Parton, 1996). They were first noted by electron microscopy as flask-shaped structures on surfaces of endothelial cells ( Palade, 1953). Caveolae can be pinched off from the cell surface in a GTP-dependent manner (Schnitzer et al., 1996) and they contain molecular components that mediate vesicle fusion, such as NSF, SNAP, and SNARE (Schnitzer et al., 1995). Moreover, in v i m evidence suggests that some molecules are transported through these structures (Parton et al., 1994; Schnitzer et al., 1994). Biochemically, caveolae have membranes enriched for cholesterol. These membrane domains on the cell surface also cluster certain proteins, such as glycosylphosphatidylinositol (GP1)-anchored proteins (Lisanti et nl., 1994). Moreover, many activated cell surface receptors also appear to cluster in cholesterol-rich domains of the plasma membrane (Liu et al., 1996).Reconstitution studies suggest that a predicted multispanning membrane protein, named caveolin, is required to organize the membrane into flask-shaped structures that are characteristic of caveolae (Fra et al., 1995). Isoforms of caveolin that have tissue-specific distribution have been identified (Scherer et al., 1996; Tang et nl., 1996).Their absence in lymphocytes and neuroblastomas has been suggested to account for the lack of caveolae in these cells (Fra et al., 1994; Gorodinsky and Harris, 1995). B. SORTINGI N THE ENDOCYTIC PATHWAYS: STUDIESON REGULATORS OF T-CELLACTIVATION
Activation of the TCR is thought to result in its endocytosis from the cell surface. Moreover, a partially assembled complex of TCR that lacks the C chain does not reach the cell surface, but is diverted to the lysosome (Lippincott-Schwartz et al., 1988; Minami et al., 1987). Studies to understand these itineraries of the TCR revealed that a tyrosine-based motif exists in the cytoplasmic domains of both CD3-y and CD3-6 chains of the TCR. This motif mediates the endocytosis of TCR from the cell surface and the diversion of a partially assembled TCR from the TGN to the
C U K K E N T ~ 7 1 I N~ INTRACELLULAH ~ s TRANSPORT
39s
lysosome. In chimeric studies that appended the cytoplasmic tail of CD3y to a reporter protein, the Tac antigen, a dileucine-based motif was discovered that behaved similarly to the tyrosine-based motif. In CD3 chains, the combination of both motifs results in a more efficient diversion of these protein from the TGN to the endosomal/lysosomal pathways ( Letourneur and Klausner, 1992). The tyrosine-based motif has been shown to be recognized by the y chains of the clathrin AP-1 aiid AP-2 complexes (Ohno et al., 1995). Significantly, the consensus sequence of the tyrosine-based motif is YXX+, where X is any amino acid aiid is any bulky hydrophobic residue (Kirchhausen et al., 1997; Trowbridge et al., 1993). This sequence is remarkably similar to the consensus sequence that is recognized by the SH2 domain of many intracellular ~nediatorsof signal transduction (Pawson, 1995).In this regard, studies on CTLA-4 have elucidated the context by which a tyrosine-based motif operates. CTLA-4 is a costimulation receptor that competes with CD28 to determine the activation status of T cells. Costimulation of CD28 and TCR results in a positive signal that promotes T-cell activation, whereas costiinulation of CTLA-4 and TCR delivers a negative sigiial that inhibits T-cell activation. Tyrosine phosphorylation in its cytoplasmic tail is required for CTLA-4 to deliver a negative signal (Jenkins, 1994). Most intriguingly, the same tyrosine, when unphosphorylated, is bound by clathrin AP-2 adaptins. Thus, it has been proposed that the phosphorylation status of this single tyrosine determines whether CTLA4 remains on the cell surface when it is activated or internalized when it is inactivated (Shiratori et al., 1997). This finding suggests a coordinated mechanism by which tyrosine phosphorylation regulates both signal transduction and protein transport. More recently, a number of other motifs have also been identified to mediate endocytosis. They include acidic clusters (Vorhees et al., 1995) and ubiquitin addition (Hicke arid Riezman, 1996). Furthermore, even the dilysine motif that mediates sorting of proteins retrograde from the Golgi to the ER can inedate endocytosis when a protein reaches the cell surface (Itin et al., 1995). Thus, a major focus has been to understand how protein transport is guided by a transport motif under physiologic conditions. A major explanation seems to be that the surroundng residues of a motif contribute to its recognition by the cellular sorting machinery. Thus, not all tyrosine-based motifs are recognized with equal avidity by both AP-1 and AP-2 adaptins. This difference would enable some proteins to be recognized preferentially at one intracellular site over the other (Kirchhausen et al., 1997). However, two other concepts have also emerged to explain how proteins with the same motif can be localized to different regions of the cell. First,
+
396
VICTOR W. HSU AND PETER J. PETERS
a motif recognized by more than one coat protein complex will result in a protein being sorted by the first coat protein complex that a cargo protein encounters as it is transported through the transport pathways. An example of this mechanism is the dilysine motif that mediates sorting from the Golgi complex to the ER. When a dilysine protein is overexpressed, it oversaturates the binding capacity of COPI coat proteins. Thus, some dilysine proteins reach the cell surface and can be sorted to the early endosome (Kappeler et al., 1994).However, at physiologic levels, all dilysine proteins would normally be recognized by COPI at the Golgi complex and retrieved to the ER. Second, a motif can be physically hidden during transport and be exposed at a later stage of transport for its function. An example of this mechanism is suggested from studies on the epidermal growth factor receptor (EGFR).A tyrosin-based motif in EGFR is hidden during its transport from the ER to the cell surface. At the cell surface, stimulation by its ligand results in receptor phosphorylation in its cytoplasmic tail. As a result, a conformational change exposes a tyrosine-based motif to clathrin AP-2 for the endocytosis of activated EGFR (Nesterov et nl., 1995). C. ENDOCYTIC COMPARTMENTS: STUDIESON MHC CLASS I1 MOLECULES Antigen presentation by MHC class I1 molecules represents one of the most complex, and fascinating, examples of coordinated protein assembly and transport. Mature MHC class I1 molecules on the cell surface consist of a and /3 heterodimers with their bound peptide. Upon their synthesis in the ER, class I1 heterodimers assemble with a third subunit, termed the invariant chain (Ii). Because Ii forms homotrimers, a fully assembled complex of MHC class I1 molecules in the ER is a nonameric complex that is composed of three subunits each of a, /3, and Ii. The Ii covers the peptide-binding cleft of class I1 d/3 heterodimers and prevents them from acquiring peptide antigens in the ER. During transport through the secretory pathway, the cytoplasmic tail of Ii possesses a dileucine-based motif that diverts the nonameric complexes from the TGN to the endosomal pathway. Hydrolases that are activated by the acidic environment of the endosomal pathway degrade Ii, except for a peptide portion, termed CLIP, that remains in the peptide-binding groove of the alp heterodimer. At the same time, protein antigens from the extracellular environment are endocytosed into the endocytic pathways, where they are degraded into peptides. An MHC-like molecule, termed HLA-DM in humans, catalyzes the release of the CLIP peptide from class I1 molecules so that exogenously derived peptides are transferred into the peptide-binding groove of class I1 molecules. The assembled class I1 molecules are then transported to
CURRENT VIEWS I N INTRACELLULAR TRANSPORT
397
the cell surface, where they present their bound peptides to T cells (Cresswell, 1994). A major goal in recent years has been to identify the exact intracellular compartment(s) where protein antigens are degraded into peptides and then loaded onto class I1 molecules. Ultrastructural analysis of a human B-cell line by immunogold electron microscopy first suggested that MHC class I1 molecules acquired their peptide antigens in a compartment, termed MHC class I1 compartment (MIIC). This compartment has lysosomal characteristics because it also contains proteins that are typically found in lysosomes, such as lysosomal associated membrane protein-1 (LAMP-l), but not markers of the early endosome, such as transferrin receptor (TfR), and the late endosome, such as mannose-6-phosphate receptor (M6PR) (Peters et al., 1991). Subsequently, biochemical purification of a compartment in a murine B-cell line identified a compartment termed class I1 vesicles (CIIV). Class I1 molecules were shown to acquire their antigenic peptides in CIIVs, but this compartment behaved more like endosomes, as they contain both TfR and M6PR (Amigorena et al., 1994).One explanation for these conflicting results is that CIIV may represent more than one compartment, as biochemical purification of a membrane compartment results in an enrichment, and rarely in the pure isolation, of a membrane compartment. Yet another possibility is that class I1 molecules may acquire their antigen peptide in more than one endosomaVlysosomalcompartment. This possibility is supported by a study that used electron microscopy with improved techniques of lipid preservation (Liou et al., 1996) to examine the membrane compartments of the endocytic pathways (Kleijmeer et al., 1997). Following an endocytic tracer transported from the cell surface to the lysosome, investigators identified two morphologically distinct compartments that contained TfR. Thus, these two compartments are defined as early endosomes. The third and fourth compartments encountered by the tracer were enriched for MGPR, suggesting that these are late endosomes. CIIVs have characteristics that best resemble the fourth compartment. The fifth and sixth compartments that were finally reached by the tracer possessed LAMP-1. Thus, these are lysosomes, of which MIICs resemble the sixth compartment. MHC class I1 complexes that contained Ii were first detected appreciably in the third compartment, and the associated Ii was largely degraded by the fourth compartment. HLA-DM that catalyzes peptide loading appeared mainly in the fourth, fifth, and sixth compartments. Thus, these results suggest that class I1 can load peptide antigens in multiple compartments that span the fourth, fifth, and sixth compartments. These compartments correlate functionally as late endosomes and lysosomes. Furthermore, as all six compartments are detected not only in
39s
VICXOR W. HSU AND PETER J. PETERS
professional APCs (such as B cells and niacrophages), but also in fibroblasts, it appears that the compartments where class I1 acquire its peptide may not be specialized compartments. Rather, they are proposed to exist in all cells. Thus, while the compartments of the endocytx pathways can be defined functionally as early endosome, late endosome, and lysosome, there may be a inore complicated array of membrane compartments that constitute these functional designations (Kleijmeer et al., 1997). Once class I1 molecules are assembled with the peptide antigen, they are tramported to the cell surface for presentation to T cells. However, little is known about the transport pathway taken by the peptide-bound class I1 molecules to the cell surface. One possibility is that compartments with assembled class I1 molecules are transported in entirety to the cell surface where they are then extruded into die extracellular environment (Raposo et al., 1996; Wubbolts et al., 1996). Precedence for this type of mechanism can be seen in the exocytosis of granules from cytotoxic T lymphocyte (CTL) and natural killer cell. These granules are thought to be specialized lysosoines that accumulate inany components that enable killing of target cells. For instance, when a CTL is activated by contact with a specific target cell, CTL granules are transported out to the extracellular space between the two cells. One component of these granules is perforin, which induces pores on the cell surface of the target cell. Another set of components are granzymes. They penetrate the target cell through the pores formed by perforin and induce apoptosis of the target cell (Tschopp and Nabholz, 1990). However, rather than the proteins of the CTL granules being extruded into the extracellular environment, it appears that the entire CTL granule is extruded out (Peters et al., 1991). Thus, a mechanistic parallel has been proposed for how a lysosomal-like CTL granule and a lysosomal-like MIIC compartment are transported to the cell surface. VI. Transport in Polarized Cells
Cells in vivo often have polarized cell surfaces that subserve their function. For instance, epithelial cells, which constitute the most diverse set of cell types in the body, have two types of cell surfaces termed apical and basolateral. These two surfaces are separated physically by tight junctions. In the intestines, the apical surface of epithelial cells faces the intestinal lumen, whereas the basolateral surfaces interface with the rest of the tissue, including the bloodstream. Thus, transport in one direction through the epithelial cells results in nutrients being delivered from the intestinal lumen to the bloodstream, whereas transport in the other direction results in antibodies being delivered to the intestinal lumen. Proteins can be targeted
~ : U K R E N T1 7 1 ~ I~N sINTRACELLULAR TRANSPORT
399
to either the basolateral or the apical cell surface by two major mechanisms. Upon transport out of the early secretory system, proteins can be sorted directly at the TGN to either the basolateral or the apical surface. Alternatively, proteins that reach one surface can be further sorted to the other surface by a transport process termed transcytosis (Rodriguez-Boulan and Powell, 1992). A. THANSPORT TO THE BASOLATERAL SURFACE It had been assumed that transport from the TGN to the basolateral surface is by default, whereas transport to the apical surface requires a sorting determinant. This assumption was based mainly on observations that proteins normally residing on the surface of nonpolarized cells are chstributed on the basolateral surface. Thus, as transport to the cell surface of nonpolarized cells had been assumed to occur by default, it was assumed that transport to the basolaterd surface would occur similarly (Matter and Mellnian, 1994). Other findings have suggested otherwise. Transport to the basolateral surface requires a sorting determinant that can be divided into two groups. Both reside on the cytoplasmic domain of transmembrane proteins (Casanova et nl., 1991; Hunziker et nl., 1991). One type of signal is very similar to those recognized by the clathrin coat proteins, and they can be either a tyrosine-based or a dileucine-based signal (Matter et al., 1992, 1994; PriIl et al., 1993). The other type, as exemplified by the LDL receptor, appears to require a glycine residue spaced a certain distance from negatively charged residues (Matter et al.. 1994). Polarized cells are complicated by having not only two functionally distinct surfaces, basolateral and apical, but also two functionally distinct endosoines that subserve their respective cell surfaces. Thus, proteins sorted to the basolaterd surfice by a basolateral targeting signal are maintained in the basolateral domain because their targeting determinant is also recognized by the basolateral endosome to resort internalized proteins back to the basolateral cell surface (Matter et nl., 1993). However, as recognition of the sorting determinants is recognized to a different extent by the TGN and the basolateral endosome, some proteins might be efficiently sorted from the TGN to the basolateral surface, but not be recognized as efficiently by the basolateral endosome for recycling to the basolateral surface. Thus, these proteins would escape the basolateral domain and become transported to the apical surface by transcytosis (Matter and Mellman, 1994). One such example is the polymeric immunoglobulin receptor (pIg-R), which is discussed in inore detail in the context of transcytosis.
B THANSPORT TO THE APICALSURFACE Studies of transport from the TCrN to the apical surface suggest that this pathway is fundamentally chfferent from most intracellular transport
400
VICTOR W. HSU AND PETER J. PETERS
pathways. As apical proteins leave the TGN, they are transported in membrane structures that are biochemically distinct from conventional transport vesicles. These membrane structures are enriched for glycosphingolipid and cholesterol that are insoluble to nonionic detergents, such as Triton X-100 (Brown and Rose, 1992), and have been termed sphingolipidcholesterol rafts (Simons and Ikonen, 1997). Furthermore, although transport from TGN to the basolateral surface requires molecular components needed for fusion, such as NSF, SNAPS, and SNARES, transport to the apical surface does not require these components (Ikonen et al., 1995). It was originally proposed that the difference between sorting to the basolateral versus the apical surface is that apical transport occurs by default in the absence of a cytoplasmic sorting determinant that is required for basolateral sorting. Consistent with this view, basolateral sorting determinants appear necessary and sufficient for targeting to the basolateral surface. Their removal from proteins that are targeted basolaterally results in these proteins being targeted apically (Casanova et al., 1991; Hunziker et al., 1991). However, evidence suggests that active mechanisms also exist to sort proteins to the apical surface. One line of evidence comes from studies of how the apical cell surface is enriched for a particular class of integral membrane proteins that attach themselves directly to phospholipids of the lipid bilayer through a GPI anchor. At the TGN, GPI-anchored proteins are sorted into transport carriers heading for the apical surface because the lipid moiety of the GPI-anchored proteins interacts preferentially with glycosphingolipid-rich carriers. Moreover, a lectin-like mechanism exists that recognizes luminal glycosylation of proteins and directs these glycoproteins to the apical surface (Scheiffele et al., 1995). Finally, one of the few known transmembrane proteins that is sorted to the apical surface, influenza neuraminidase, appears to use a transmembrane-sorting determinant (Kundu et al., 1996). C. TRANSCYTOSIS: STUDIES ON THE POLYMERIC IMMUNOGLOBULIN RECEPTOR Studies on the polymeric immunoglobulin receptor (pIg-R) have contributed much to understanding how transcytosis occurs. This receptor is a single transmembrane protein that binds to polymeric immunoglobulins, IgA and IgM, through the luminal domain of the receptor. The physiologic role of pIg-R is best understood for its transport of IgA through epithelial cells that line the intestine. BasoIateraI pIg-R that faces the basement membrane binds to dimeric IgA and then transports the bound IgA across the epithelial cell by transcytosis to the apical side. IgA on the apical surface is then released into the intestinal lumen by proteolysis that cleaves
CUHKENT VIEWS IN INTRACELLUIAR TRANSI'OHT
40 1
at the junction between the luminal and the transmembrane domain of pIg-R. Thus, the released IgA still associates with the luminal domain of the pIg-R. The identity of the responsible protease(s) that cleaves the pIg-R remains undetermined, but it appears to be a leupeptin-sensitive endoprotease (Casanova, 1996; Mostov et al., 1995). To accomplish the transcytosis of IgA, newly synthesized pIg-R is transported first through the biosynthetic pathway of an epithelial cell, where it is then sorted from the TGN to the basolateral surface to acquire IgA, and is subsequently transcytosed across to the apical surface to release IgA. Transcytosis from the basolateral surface to the apical surface appears to occur through intermediary sorting compartments. Upon internalization from the basolateral surface, pIg-R is first transported to the basolateral early endosome. From there, the receptor is transported across the cell to a recycling endosome and then sorted to the apical cell surface. The recycling endosome in epithelial cells appears to function as a common compartment that receives proteins that are being transcytosed from either the basolateral early endosome or the apical early endosome (Geuze et al., 1984; Sztul et al., 1985). The sorting signals for transcytosis of pIg-R have been localized to its cytoplasmic tail. A cytoplasmic motif of HRXXV (where X can be any amino acid) that lies near the membrane spanning domain of the receptor is required to sort pIg-R from the TGN to the basolateral surface (Areoti et al., 1993). In the first leg of transcytosis from the basolateral cell surface to basolateral early endosome, two other cytoplasmic motifs mediate internalization of the receptor. Both motifs are similar to the tyrosine-based motif defined for clathrin-mediated endocytosis in nonpolarized cells. Each can function independently to mediate endocytosis from the basolateral surface (Okamoto et nl., 1992). In the second leg of transcytosis from the basolateral early endosome to the apical surface, a serine-based motif is required to sort pIg-R into a transcytotic pathway. Identification of this sorting motif provides a different example of how transport motifs can be regulated by phosphorlyation. Unlike tyrosine phosphorylation that inactivates a tyrosine-based motif, serine phosphorylation activates the serine-based motif for transcytosis of the pIg-R (Casanova et nl., 1990). The role of phosphorylation appears to confer a negative charge to the serine residue, as mutational analysis that substitutes serine with a negatively charged amino acid results in transcytosis, whereas substitution with any other residue abrogates transcytosis (Apodaca and Mostov, 1993). The serine-based motif lies more distal to the membrane-spanning region of' pIg-R than the HRXXV motif that sorts the receptor basolaterally. Thus, it has been proposed that phosphorylation of the serine-based motif not only activates this motif, but also
402
VICTOR Mi.HSU AND PETER
I.
PETERS
inactivates the nearby basolateral sorting motif. At the basolateral early endosome, serine phosphorylation would serve as a trafficking switch to prevent pIg-R from being recycled back to the basolateral cell surface and to direct the receptor into a transcytotic pathway toward the apical surface (Casanova, 1996; Mostov et al., 1995). However, transcytosis induced by serine phosphorylation can be substituted by pIg-R binding to its ligand (Hirt et al., 1993). Thus, ligand binding may result in a conformational change in the cytoplasmic tail of pIg-R that exposes a yet to be defined transport motif for traiiscytosis that can also be accomplished by phosphorylation of the serine-based motif.
VII. Perspective The pathways that connect the intracellular compartments were revealed in the 1960s by pioneering morphologic studies using electron microscopy. Since then, our understanding of the mechanistic details by which proteins are transported through these pathways has progressed immensely and has resulted in many fundamental concepts. In particular, key regulatory components of transport exist as families of proteins that function in mechanistically similar ways at different intracellular sites. Because many transport pathways have been identified to be bidirectional, regulation of transport in one direction indirectly regulates transport in the other direction. In this regard, the coordinate regulation of bidirectional transport provides a mechanism by which organelle identity is maintained, as organelles are dynamic structures that are maintained in steady state by membranes that feed into and emanate from them. However, as highlighted in this review, many details remain to be resolved. Progress toward elucidating these details will undoubtedly continue to be contributed by studies in immunology. This better understanding of transport mechanisms will likely be reciprocated with a better understanding of the immune system. More generally, as protein function can be altered by its localization, further understanding of transport mechanisms will likely result in a greater understanding of many disease processes.
ACKNOWLEDGMENTS We thank members of our respective laboratories for inany stimulating discussions and James Casanova, Marianne Wessling-Resnick, and Peter van der Sluijs for their critical comments on this review. Inherent in our treatment of the literature has been some subjectivity in the topics that are selected. Thus, we apologize to our colleagues who feel their work fits the goal of this review, but is not specifically cited.
REFERENCES Amigorena, S., Drake, J. R., Webster, P., and Mellman, I. (1994).Transient accumulation of new class I1 MHC molecules in a novel endocytic compartment in B lymphocytes. Nature 369, 113-120.
CURRENT VIEWS IN INTHACEL1,ULAR TRANSPORT
403
Andersson, M., Paabo, S., Nilsson. T., and Peterson, P. A. (1985). Impaired intracellular transport of class I MHC antigens as a possible means for adenoviruses to evade immune snrveillance. Cell 43, 215-222. Aniento, F., Gu, F., Parton, R. G., and Gruenberg, J. (1996). An endosomal beta-COP is involved in the pH-dependent formation of transport vesicles destined for late endosomes. J , Cell B i d . 133, 29-41. Anntony, B., Huber, I., Paris, S., Chahre, M., and Cassel, D. (1997). Activation of ADPrihosylation factor 1 GTPase-activatingprotein by pliosphatidylcholine-derived diacylglycerols. /. Biol. Chern. 272, 30848-30851. Aoe, T., Ciiherman, E., Lee, A,, Cassel, D., Peters, 1’. J., and Ilsu, V. W. (1997). The KDEL receptor, ERD2, regulates intracehlar traffic by recruiting a GTPase-activating protein for ARF1. E M B O / 16, 7305-7316. Aoe. T., Lee, A. J., van Donselaar, E., Peters, P. J.. and Hsu, V. W. (1998). Modulation of intracellular transport by transported proteins: Insight from regulation of COPI-mediated transport. Proc. Natl. Acad. Sci. USA, 95, 1624-1629. Apodaca, G.. and Mostov, K. E. (1993). Traiiscytosis of placental alkaline phospliatasepolyineiic ini~nunoglobulinreceptor fiision proteins is regulated by mutation of ser664. 1. B i d . Cham. 268, 23712-23719. Areoti, B.. Kosen. P. A,, Knntz. I. D., Cohen, P. E., and Mostov, K. E. (1993). Mutational and secondary strnctiiral analysisofthc basolateral sorting signal of the polymeric immunoglobulin receptor. ]. Cell B i d 123, 1149-1160. Aridor, M., Bannykh, S. I., Rowe, T., and Balch, W. E. (1995).Sequential coupling between COPII and COP1 vesicle coats in endoplasmic reticulum of Golgi transport. /. Cell B i d 131, 875-893. Balch, W. E., Dunphy, W. G., Braell, W.A,, and Rothinan, J. E. (1984). Reconstitution of the transport of protein between siiccessive compartments of the Golgi measured by the coiipled incoiporation of N-acetylglucosamine.Cell 39, 405-416. Balch, W. E., McCaffery,J. M., Plutner, H., and Farcpihar, M. G. (1994).Vesicular stoinatitis virus glycoprotein is sorted and concentrated diiring exyort from the endoplasmic reticnluni. Cell 76, 841-852. Barlowe, C. (1997). Coupled ER to Cole transport reconstituted with purified cytosolic proteins. /, Cell Biol. 139, 1097-1108. Barlowe, C., Orci, L., Yeung, T., Hosobnchi, M., IIainamoto, S., Salama, N., Rexach, M. F., Ravazzola. M., Amlierdt, M., and Schekman, R. (1994). COPII: A membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77, 895-907. Baumeister, W., and Lnpas, A. (1997).The proteasome. Cum-. Opin. Stmct. B i d . 7,273-27H. Bednarek, S. Y., Orci, L., and Schekman, R. (1996). Traffic COPS and the forination of vesicle coats. Trunrls Cell B i d . 6, 468-473. Blobel. G., and Dobberstein, B. (197rj).Transfer of proteins across membranes.]. Cell B i d . 67, 835-851. Block, M. R.. Glick, B. S.. Wilcos, C. A., Wielantl. F. T.. and Rothinan, J. E. (1988). Purification of an N-ethiyln~aleiri~ide-sensitive protein catalyzing vesicular transport. €‘roc. Nntl. Acnd Sri. USA 85, 78Fj2-7856. Blond-Elgnindi, S., Cwirla, S. E., Dower, W. J., Lipshutz, R. J., Sprang, S. R., Sambrook, J. F., and Gething, M. J. (1993). Affinity panning of a library of peptides displayed on bacteriophages reveals the binding specificity of Rip. Cell 75, 717-728. Bole, D. G., Hendershot, L. M., and Keaniey, J. F. (1986). Posttranshtiond association of imniunoglobulin heavy chain binding protein with nascent heavy chains in nonsecretiiig and secreting hyhridoinas. /. Cell R i d . 102, 1558-1.566. Bonian, A. L., and Kahn, R. A. (199s).Arf proteins: The membrane traffic police? Trends Biocltem. Sci. 20, 147-150.
404
VICTOR W. HSU AND PETER J. PETERS
Bonifacino, J. S., Chen, C., Lippincott-Schwartz, J., Ashwell, J. D., and Klausner, R. D. (1988). Subunit interactions within the T-ceU antigen receptor: Clues from the study of partial complexes. Proc. Natl. Acad. Sci. USA 85, 6929-6933. Bonifacino, J. S., Cosson, P., and Klausner, R. D. (1990). Colocalized transmembrane determinants for ER degradation and subunit assembly explain the intracellular fate of TCR chains. Cell 63, 503-513. Bonifxino, J. S., Cosson, P., Shah, N., and Klausner, R. D. (1991). Role of potentially charged transmembrane residues in targeting proteins for retention and degradation within the endoplasmic reticulum. EMBO J. 10, 2783-2793. Bonifacino, J. S., and Lippincott-Schwartz, J. (1991). Degradation of proteins within the endoplasmic reticulum. C u m Opin. Cell Bid. 3, 592-600. Bonifacino, J. S., Snzuki, C. K., and K h m e r , R. D. (1990). A peptide sequence confers retention and rapid degradation in the endoplasmic reticulum. Science 247, 79-82. Bonifacino,J. S., Suzuki, C. K., Lippincott-Schwartz,J., Weissman, and A. M., and Klausner, R. D. (1989).Pre-Golgi degradation of newly synthesized T-cell antigen receptor chains: Intrinsic sensitivity and the role of subunit assembly. J , Cell B i d . 109, 73-83. Booth, C., and Koch, G. L. E. (1989). Perturbation of cellular calcium induces secretion of luminal ER proteins. Cell 59, 729-737. Brennwald, P., Kearns, B., Champion, K., Keranen, S., Bankaitis, V., and Novick, P. (1994). Sec9 is a SNAP-25-like component of a yeast SNARE complex that may be the effector of Sec4 function exocytosis. Cell 79, 245-258. Brodksy, F. M. (1997). New fashions in vesicle coats. Trends Cell B i d 7, 175-179. Brown, D. A,, and Rose, J. K. (1992).Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface. Cell 68, 533-544. Brown, H. A,, Gutowski, S., Moomaw, C. R., Slaughter, C., and Sternweis, P. C. (1993). ADP-ribosylation factor (ARF), a small GTPase-dependent regulatory protein stimulates phospholipase D activity. Cell 75, 1137-1144. Burgess, T. L., and Kelly, R. B. (1987). Constitutive and regulated secretion of proteins. Annu. Rev. Cell Biol. 3, 243-293. Carter, L., Redelmeier, T., Woollenweber, L., and Schmid, S. (1993). Multiple GTPbinding proteins participate in clathrin-coated vesicle-mediated endocytosis.J. Cell B i d . 120,37-45. Casanova, J, E. (1996). Structure and function of the polymeric immunoglobulin receptor. 1n “Essentials ofMucosa1 Immunity,” Chap. 13, pp. 151-166. (M. Kagnoffand H. Kiyono, Eds.) Academic Press, Sail Diego. Casanova, J. E., Apodaca, G., and Mostov, K. E. (1991).An autonomous signal for basolateral sorting in the cytoplasmic domain of the polymeric iinmunoglobulin receptor. Cell 66, 65-75. Casanova, J. E., Breitfeld, P. P., Ross, S. A,, and Mostov, K. E. (1990). Phosphorylation of the polymeric immunoglobulin receptor required for its efficient transcytosis. Science 248, 742-745. Chardin, P., Paris, S., Antonny, B., Robineau, S., Beraud-Dufour, S., Jackson, C. L., and Chabre, M. (1996). A human exchange factor for ARF contains Sec7- and pleckstrinhomology domains. Nature 384, 481-484. Chaudhary, A,, Gu, Q.-M., Thum, O., Profit, A. A., Jeyakumar, L., Fleischer, S., and Prestwich, G. D. (1998).Specificinteraction of Golgi coatomer alpha-COP with phosphatidylinositol (3, 4, 5) trisphosphate. 1.B i d . Cheni. 273, 8344-8350. Ciechanover, A. (1994). The ubiquitin-proteasome proteolytic pathway. Cell 79, 13-21. Clay, D. O., Griff, I. C., and Rothman, J. E. (1990). SNAPS,a family of NSF attachment proteins involved in intracellular membrane fusion in animals and yeast. Cell 61,709-721.
CURRENT VIEWS IN INTRACELLULAR TRANSPORT
405
Clemens, D. L., and Honvitz, M. A. (1995).Characterization of the mycobacterium tuberculosis phagosome and evidence that phagosomal maturation is inhibited. 1. Exp. Med. 181,257-270. Cockcroft, S., Thomas, G. M., Fensome, A., Geny, B., Cunningham, E., Gout, I., Hiles, I., Totty, N. F., Truong, O., and Hsuan, J. J. (1994). Phopholipase D: A downstream effector of ARF in grandocytes. Science 263,523-526. Cole, N. B., Siaky, N., Marotta, A., Song, J., and Lippincott-Schwartz, J. (1996). Golgi dispersal during microtubule disruption: Regeneration of Golgi stacks at peripheral endoplasmic reticulum exit sites. Mol. Biol. Cell 7, 631-650. Copeland, C. S., Doms, R., W., Bolzau, E. M., Webster, R. G., and Helenius, A. (1986). Assembly of influenza hemagglutinin timers and its role in intracellular transport. J. Cell Biol. 103, 1179-1191. Cosson, P., and Letourneur, F. (1997). Coatomer (COP1)-coated vesicles: Role in intracellular transport and protein sorting. Cum. @in. Cell Biol. 9, 484-487. Cosson, P., and Letourneur, F. (1994). Coatomer interaction with di-lysine endoplasmic reticulum retention motifs. Science 263, 1629-1631. Cowles, C. R., Odorizzi, G., Payne, G. S., and Emr, S. D. (1997).The AP-3 adaptor complex is essential for cargo-selective transport to the yeast vacuole. Cell 91, 109-118. Cresswell, P. (1994). Assembly, transport, and function of MHC class I1 molecules. Annu. Rev. Immunol. 12, 259-293. D’Souza-Schorey, C., Boshans, R. L., McDonough, M., Stahl, P. D., and Van Aelst, L. (1997). A role for POR1, a Racl-interacting protein, in ARF6-mediated cytoskeletal rearrangements. EMBO J 16, 5445-5454. D’Souza-Schorey, C., Li, G., Colombo, M. I., and Stahl, P. D. (1995). A regulatory role for ARFG in receptor-mediated endocytosis. Science 267, 1175-1178. D’Souza-Schorey, C., van Donselaar, E., Hsu, V. W., Yang, C., Stahl, P. D., and Peters, P. J. (1998). ARFG targets recycling vesicles to the plasma membrane: Insights from an ultrastructural investigation.J Cell B i d . 140, 603-616. Dascher, C., and Balch, W. E. (1994).Dominant inhibitory mutants ofARFl block endoplasmic reticulum to Golgi transport and trigger disassembly of the Golgi apparatus. /. Biol. Chem. 269, 1437-1448. Dautry-Varsat,A., Ciechanover, A,, and Lodish, H. (1983).pH and the recyclingoftransfemn during receptor-mediated endocytosis. Proc. Natl. Acad. Sci. USA 80, 2258-2262. De Camilli, P., Emr, S. D., McPherson, P. S., and Novick, P. (1996).Phosphoinositides as regulators in membrane traffic. Science 271, 1533-1539. Dell’Angelica, E. C., Ohno, H., Ooi, C. E., Rabinovich, E., Roche, K. W., and Bonifacino, J. S. (1997). AP-3: An adaptor-like protein complex with ubiquitous expression. EMBO J. 16,917-928. Desjardins, M., Huber, L. A,, Parton, R. G., and Griffiths, G. (1994).Biogenesis ofphagolysosomes proceeds through a sequential series of interactions with the endocytic apparatus. 1. Cell Biol. 124, 677-688. Doms, R. W., Russ, G., and Yewdell, J. W. (1989). Brefeldin A redistributes resident and itinerant Golgi proteins to the endoplasmic reticulum. J. Cell B i d . 109, 61-72. Donaldson, J. G., Cassel, D., Kahn, R. A., and Klausner, R. D. (1992). ADP-ribosylation factor, a small GTP-binding protein, is r6quired for binding of the coatomer protein betaCOP to Golgi membranes. Proc. Natl. Acad. Sci. USA 89, 6408-6412. Donaldson, J. G., and Klausner, R. D. (1994). ARF: A key regulatory switch in membrane traffic and organelle structure. C u m @in. Cell Bid. 6, 527-532. Farquhar, M. G. (1985). Progress in unraveling pathways of Golgi traffic. Annu. Rezj. Cell Biol. 1, 447-488.
406
VICTOR W. HSU AND PETER J. PETEHS
Ferguson, S. S. G., Downey, W. E., Colapietro, A.-M., Barak, L. S., Menard, L., and Caron, M. G. (1996). Role of beta-arrestin in mediating agonist-promoted G protein-coupled receptor internalization. Science 271, 363-366. Fiedler, K., Veit, M., Stamnes. M. A,. and Rothnian. J. E. (1996). Bimodal interaction of coatomer with the p24 family of putative cargo receptors. Science 273, 1396-1399. Foti, M., Mangasarian, A , , Pipet, V., Lew, D., Krause, K.-FI., Trono, D., and Carpentier, J.-L. ( 1997). Nef-mediated clathrin-coated pit formation. J . Cell B i d . 139, 37-47. Fra, A. M., Williamson, E., Siinons, K., and Parton, R. G. (1995). De novo forination of caveolae in lymphocytes hy expression of VIP2l-caveolin. Proc. Nntl. Acnd. Sci. USA 92, 86.55-8659. Fra, A. M., Williamson, E., Siinons, K., and Parton, R. G . (1994). Detergent-insoluble glycolipid microdomains i n lymphocytes in the absence of caveolae. J . B i d Chem. 269, 30745-30748. Frank, S., Upender, S., Hansen, S. H., and Casanova, J. E. (1998). ARNO is a guanine nucleotide exchange factor for ADP-ribosylation factor 6.1. Bid. Chem 273, 23-27. Gaynor, E. C., and Emr, S. D. (1997). COPI-independent anterograde transport: Cargoselective ER to Golgi protein transport in yeast COP1 mutants.]. Cell Biol. 136,789-802. Geli, M. I., and Riezman, H. (1996).Role of type I myosins in receptor-mediated endocytosis in yeast. Science 272, 533-535. Gerinain. R. N.. and Margulies, D. H. (1993). The biochemistry and cell biology of antigen processing and presentation. Antw. Reu. Immutzol. 11, 403-450. Gething, M.-J., McCamnion, K., and Sambrook, J. (1986). Expression of wild-type and mutant forms of influenza heinagglutinin: The role of folding in intrac!ellular transport. Cell 36, 939-9S0. Gething, M. J., and Sainbrook, J. (1992). Protein folding in the cell. Nature 355, 33-45. Geuze, H. J., Slot, J. W., Strous, G. J., Lodish, H. F., and Schwartz,A. L. (1983).Intracellular site of asialoglycoprotein receptor-ligand uncoupling: Double label immunoelectron microscopy during receptor mediated endocytosis. Cell 32, 277-287. Geuze, H. J.. Slot, J. W., Strous, G. J., Peppard, J., von Figura, K., Hasilik, A,, and Schwartz, A. L. (1984).lntracellular receptor sorting during endocytosis: Comparative inimunoelectron microscopy of multiple receptors in rat liver. Cell 37, 195-204. Geuze, H. J., Stoorvogel, W., Strous, G. J., Slot, J. W., Bleekemolen, J. E., and Mellman, I. (1988). Sorting of inannose 6-phosphate receptors and lysosomal ineinbrane proteins in endocytic vesicles. J . Cell B i d . 107, 2491-2501. Glickman, J. N., Conihear, E., and Pearse, B. M. (1989). Specificity of binding of clathrin adaptors to signals on the mannose-6-phosphate/insulin-like growth factor I1 receptor. E M B O J . 8, 1041-1047. Goldstein, J. L., Brown, M. S., Anderson, R. G., Russell, D. W., and Schneider, W. J. (1985). Receptor-mediated endocytosis:concepts emerging from the LDL receptor system. Annu. Reu. Cell Biol. 1, 1-35, Goodman, 0. B., Krupnick, J. G., Santini, F., Gurevich, V. V., Penn, R. B., Gagnon, A. W., Keen, J. H., and Benovic, J. L. (1996). Beta-arrestin acts as a clathrin adaptor in endocytosis of the bet&-adrenergic receptor. Nature 383,447-450. Gorodinsb, A,, and Harris, D. A. (1995). Glycolipid-anchored proteins in neuroblastotna cells form detergent-resistant complexes without caveolin. ]. Cell Biol. 129, 619-627. Griffiths, G., Back, R., and Marsh, M. (1989). A quantitative analysis of the endocytic pathway in hahy hamster kidney cells. I. Cefl Biol. 109, 2703-2720. Griffiths, G.. and Siinons, K. (1986). The trans Golgi network: Sorting at the exit site of the Golgi complex. Science 234, 438-443.
C U R R E N T VIEWS IN INTHACELLIJLAR T R A N S P O R T
407
Haas, I. G., and Wabl. M. (1983). Iininunoglobulin heavy chain binding protein. Nature 306, 387-389. Hartl, F. U . (1996). Moleciilar chaiyeroiies in cellular protein folding. Nature 381,571-580. Hauri, II.-P., and Schweizer, A. ( 1992).Relationship of the ER-Golgi intermediate compartinent to endoplasmic reticuhim and Golgi apparatus. Curr. Opin. Cell B i d . 4, 600-608. Hay, J. C., Chao, D. S., Kuo, C. S., and Scheller, H . H. (1997).Protein interactions regulating vesicle transport hatween the endnplasmic reticulum and Golgi apparatus in mammalian cells. Cell 89, 149-1.58. Hcilker, R., Maiming-Kreig, U.. Zuber, J. F.. u r i d Spiess, M. (1996). In vitro binding of clathrin adaptors to sorting signals correlates with endocytosis and basolateral sorting. E M B O J . 15, 2893-2899. Helenius, A,. Mellnian, I., Wall, D., and Hubhard, A. (1983). Endosomes. T r e n h Biocherti. Sci. 8, 245-250. Ilerskovits, J. S., Bnrgess, C. C . , Obar, R. A., and Vallee, R. B. (1993). Effects of mutant rat dynaniin on endocytosis.1.Cell Biol. 122, 565-578. Ilicke. L., and Riezinan, H. (1996). Ubic~iiitinatio~a of a yeast plasma inenibrane receptor signals its ligand-stimulated endocytosis. Cell 6, 1721-1742. Hiller. M. M., Finger, A,, Schweiger, M.. and Wolf. D. H. (1996). ER degradation of a inisfolded luininiil protein by the cytosolic ilbiqi.iitiia-proteaSOlllepathway. Science 273, 1725-1728. Hinshaw, J. E., and Schinid, S. L. (1995).Dyiaainin self-assembles into rings suggesting a mechanisni for coated vesicle budding. Nafurc 374, 190-192. Hirschberg, C. B., and Snider, M. D. (1987). Topography of glycosylation in the rough endoplasmic reticulum and Golgi apparatus. Annri. Rev. Biochem. 56, 63-87. IIirt, R. P., Hughes, G. J.. Frutiger, S., Miclietti, P., Perregaux, 0..Poulin-Godefroy, O., Jeanguenat, N.. Neutra. M. R., and Krachenl~uhl,I.-P. (1993). Transcytosis of the polymeric imiiiunoglobulin receptor requires phosphoiylatioii of serine 664 in the absence h i t not the presence of dimeiic IgA. Cell 74, 245. Honing, S., Griffith. J., Geuze, H. J., and Hunziker, W. (1996).The tyrosine-based lysosonial targeting signal in lamp-1 mediates sorting into Golgi-derived clatlirin-coated vesicles. E M B O ] . 15, 5230-5239. Hsu, 17. W., Shah, N., and Klausner, 11. D. (1992). A brefeldin A-like phenotype is induced by the overexpression of a huinaii ERD-2-like protein, ELP-1. Cell 69, 625-635. Hsu, V. W., Yiian, L. C., Nuchtern, J. G., Lippincott-Schwartz, J., Hammerling, G. J., antl Klausner, R. D. ( 1991).A recycling pathway between the endoplasmic reticulum and the Golgi apparatus for retention of unasseinbled MHC class I molecules. Nature 352, 441-444. Huinphrey, J. S., Peters, P. J., Yuan, L. C., and Bonifacino, J. S. (1993). Localization of TGN38 to the trans-Golgi network: Involvement of a cytoplasmic tyrosine-containing sequence. I. Cell Biol. 120, 1123-11:35. Hunziker, W., Harter, C.. Matter. K., antl Melhnan, I. (1991). Basolateral sorting in MDCK cells requires a distinct cytoplasmic tlomaiii deterininant. Cell 66, 907-920. Huppa. J. B., and Ploegh, H. L. (1997). The alpha chain of the T cell antigen receptor is degratletl in the cytosol. Inimtmity 7, 113-122. Hiirtley, S. M., and Helenius, A. (1989). Protein oligomerization in the endnplasmic reticnh i i n . h n u Reu. Cell Biol. 5, 277-307. Huttner. W. B.. and Tooze, S. A. (1989). Binsynthetic protein transport in the secretory pathway. Curr Opin. Cell Biol. 1, 648-6.54. Ikonen, E., Tagaya, M., Ullricla, O., Montecucco, C., antl Sinions, K. (1995). Different requirements for NSF, SNAP, and Rab proteins in apical and basolaterd transport in MDCK cells. Cell 81, 571-580.
408
VICTOR W. HSU AND PETER J. PETERS
Itin, C., Kappeler, F., Linstedt, A. D., and Hauri, H.-P. (1995). A novel endocytosis signal related to the KKXX ER-retrieval signal. EMBO J. 14, 2250-2256. Jackson, M. R., Nilsson, T., and Peterson, P. A. (1990). Identification of a consensus motif for retention of transmembrane proteins in the endoplasmic reticulum. EMBO J. 9, 3153-3162. Jackson, M. R., Nilsson, T., and Peterson, P. A. (1993). Retrieval of transmembrane proteins to the endoplasmic reticulum. J. Cell Biol. 121, 317-333. Jenkins, M. K. (1994). The ups and downs of T cell costimulation. Immunity 1, 443-446. Johnson, K. F., and Kornfeld, S. (1992).A €€is-Leu-Leusequence near the carbovl terminus of the cytoplasmic domain of the cation-dependent mannose 6-phosphate receptor is necessary for the lysosomal enzyme sorting function. J. Biol. Chem. 267, 17110-17115. Kahn, R. A,, and Gilman, A. G. (1984). Purification of a protein cofactor required for ADPribosylation of the stimulatory regulatory component of adenylate cyclase by cholera toxin. J. Biol. Chem. 259, 6228-6234. Kappeler, F., Itin, C., Schindler, R., and Hauri, H.-P. (1994). A dual role for COOHterminal lysine residues in pre-Golgi retention and endocytosis of ERGIC-53. J. Biol. Chem. 269, 6279-6281. Kassenbrock, C. K., Garcia, P. D., Walter, P., and Kelly, R. B. (1988). Heavy-chain binding protein recognizes aberrant polypeptides translocated in vitro. Nature 333, 90-93. Kirchhausen, T., Bonifacino, J. S., and Riezman, H. (1997). Linking cargo to vesicle formation: Receptor tail interactions with coat proteins. Cum Opin. CeZZ Biol. 9,488-495. Klar, D., and Hanimerhng, G. J. (1989).Induction of assembly of MHC class I heavy chains with beta2-microglobulin by interferon-gamma. EMBO J 8, 475-481. Klausner, R. D., Donaldson, J. G., and Lippincott-Schwartz, J. (1992). Brefeldin A: Insights into the control of membrane traffic and organelle structure.J. Cell Biol. 116,1071-1080. Klausner, R. D., Lippincott-Schwartz, J., and Bonifacino, J. S. (1990). The T cell antigen receptor: Insights into organelle biology. Annu. Rev. Cell Biol. 6, 403-431. Klausner, R. D., van Renswoude, J., Ashwell, G., Kempf, C., Schechter, A. N., Deang, A,, and Bridges, K. R. (1983). Receptor-mediated endocytosis of transferrin in K562 cells. J. Biol. Chem. 258, 4715-4724. Kleijmeer, M. J., Morkowski, S., Griffith, J. M., Rudensky, A. Y., and Geuze, H. J. (1997). Major histocompatibility complex class I1 compartments in human and mouse B lymphoblasts represent conventional endocytic compartments. 1.Cell Biol. 139, 639-649. Kopito, R. R. (1997). ER quality control: The cytoplasmic connection. Cell 88, 427-430. Koppehan, B., and Cresswell, P. (1990). Rapid nonlysosomal degradation of assembled class I1 glycoproteins incorporating a mutant DR alpha chain.]. Immunol. 145,2730-2736. Komfeld, R., and Kornfeld, S. (1985).Assembly of asparagine-linked oligosaccharides.Annu. Rev. Biochem. 54, 631-664. Kornfeld, S., and Mellman, I. (1989). The biogenesis of lysosomes. Annu. Rev. Cell Biol. 5, 483-525. Kreis, T. E., Lowe, M., and Pepperkok, R. (1995). COPS regulating membrane traffice. Annu. Reti. Cell Deti. Biol. 11, 677-706. Ktistakis, N. T., Brown, H. A., Waters, M. G., Sternweis, P. C., and Roth, M. G. (1996). Evidence that phospholipase D mediates ADP ribosylation factor-dependent formation of Golgi coated vesicles. J. Cell Biol. 134, 295-306. Kubler, E., and Riezrnan, H. (1993). Actin and fimbrin are required for the internalization step of' endocytosis in yeast. EMBO J. 12, 2855-2862. Kuchler, K. (1993). Unusual routes of protein secretion: The easy way out. Trends Cell Biol. 3,421-426.
C U R R E N T VIEWS IN INTHACELLULAR TRANSPORT
409
Kundu, A,, Avalos, R. T., Sanderson, C. M., and Nayak, D. P. (1996). Transmembrane domain of influenza vinrs neuraminiclase, a type 11 protein, possesses an apical sorting signal in polarized MDCK cells. J. Virol. 70, 6508-6515. Lainzue, C., Chnan, T.-H.. Terlecky, L. J.. Bokoch, G. M., and Schniid, S. L. (1996). Regulation of I-eceptor-mediated endocytosis by Rho and Rac. Nature 382, 177-179. Lamaze, C., Fujimoto, L. M., Yin, H. L., and Schmid, S. L. (1997). The actin cytoskeleton is required for receptor-mediated endocytosis in mammalian cells. J . B i d . Chem. 272, 20332-20335. Lankford, S. P., Cosson, P., Bonifacino, J. S., and Klausner, R. D. (1993). Transmembrane domain length affects charge-mediated retentioil and degradation of proteins within the endoplasmic reticulum. J. Biol. CI~eni.268, 4814-4820. Letouineur, F., Gaynor, E. C., Hennecke, S.. Deinolliere, C., Duden, R., Emr, S. D., Riezman, H., and Cosson, P. (1994). Coatomer is essential for retrieval of dilysine-tagged proteins to the endoplastnic reticulum. Cell 79, 1199-1207. Letourneur, F., and Klausner, R. D. (1992). A i ~ ~di-leucine e l motif and a tyrosine-based inotif independently mediate lysosomal targeting and endocytosis of CD3 chains. Cell 69, 1143-1157. Lewis, M. J., and Pelham, H. R. (1992). Ligand-induced redistribution of a human KDEL receptor from the Golgi complex to the endoplasmic reticulum. Cell 68, 353-364. Lian, J. P., Stone, S., Jiang, Y., Lyons, P., and Ferro-Novick, S. (1994). Yptlp iinplicated in v-SNARE activation. Nature 372, 698-701. Lou, W., Ceuze, 1-1. J., and Slot, J. W. (1996). Improving structural integrity of cryosections for iinmiinogold labeling. Hisfochem Cell Biol. 106, 41-58. Lippincott-Schwartz, J. (1993). Bidirectional membrane traffic between the endoplasmic . reticulum and Golgi apparatus. Tren,ds Cell Biol. 3, 81-88. Lippincott-Schwartz, J., Bonifacino, J. S., Yuan, L. C., and Klausner, R. D. (1988). Degradation from the endoplasmic reticuliuni: Disposing of newly synthesized proteins. Cell 54, 209-220. Lippincott-Schwartz, J., Yuan, L. C., Bonifacino, J. S., and Klausner, R. D. (1989). Rapid redistribution of Golgi proteins into the ER in cells treated with brefeldin A: Evidence for membrane cycling from Golgi to ER. Cell 56, 801-813. Lisanti, M. P., Scherer, P. E., Vidugiriene, J., Tang, Z., Hennanowski-Vosatka, A,, Tu, Y.-H., Cook, R. F., and Sargiacomo, M. (1994). Characterization of caveolin-rich membrane domains isolated from an endothelial-rich source: Implications for human disease. J. Cell Biol. 126, 111-126. Liu, P., Ying, Y., KO, Y. G., and Anderson, R. G. (1996). Localization of platelet-derived growth factor-stimulated phosphorylation cascade to caveolae.J. Biol. Chern. 271,1029910303. Lobel, P.. Fujimoto, K., Ye, R. D., Griffiths, G., and Kornfeld, S . (1989). Mutations in the cytoplasmic domain of the 275 kd mannose 6-phosphate receptor differentially alter lysosomal enzyme sorting and endocytosis. Ccll 57, 787-796. Lobell, R. B., Arm, J. P., Raizman, M. B., Austen, K. F., and Katz, H. R. (1993). Intracehlar degradation of Fc gamma RIII in iiiouse bone marrow cultured-derived progenitor mast cells prevents its surface expression and associated function. J. Biol. Chern. 268, 12071212. Lupashin, V. V., Hamamoto, S., and Schekman. R. W. (1996). Biochemical requirements for targeting and fusion of ER-derived transport vesicles with purified yeast Golgi membranes. J . Cell Biol. 132, 277-289. Lupashin, V. V., and Waters, M. G. (1997).t-SNARE activation through transient interaction with a Rab-like CTPase. Science 276, 1255-1258.
410
VICTOR W. MSU AND PETER J. PETERS
Mains, P. E., and Sibley, C. H. (1982). The requirement of light chain for the surface deposition of the heavy chain of immunoglobulin M. J. Biol. Chem. 258, 5027-5033. Manolios, N., Bonifacino, J. S., and Klausner, R. D. (1990).Transmeinbrane helical interactions and the assembly of the T cell receptor complex. Science 249, 274-277. Manolios, N., Letoumeur, F., Bonifacino, J. S., and Klausner, R. D. (1991). Painvise, cooperative and inhibitory interactions describe the assembly and probable structure of the T-cell antigen receptor. EMBO J. 10, 1643-1651. Matter, K., Hunziker, W., and Mellman, I. (1992). Basolateral sorting of LDL receptor in MDCK cells: The cytoplasmicdomain contains two tyrosine-dependent targeting determinants. Cell 71, 741-753. Matter, K., and Mellman, I. (1994). Mechanisms of cell polarity: Sorting and transport in epithelial cells. Czirr. Opin. Cell Bid. 6, 545-554. Matter, K., Whitney, J. A,, Yamamoto, E. M., and Mellinan, I. (1993). Common signals control low density lipoprotein receptor sorting in endosomes and the Golgi complex of MDCK cells. Cell 74, 1053-1064. Matter, K., Yamamoto, E. M., and Mellman, I. (1994).Structural requirements and sequence motifs for polarized sorting and endocytosis of LDL and Fc receptors in MDCK cells. J . Cell Biol. 126, 991-1004. McCarty, J. S., Buchberger, A,, Reinstein, J,, and Bukau, B. (1995). The role of ATP in the functional cycle of the DnaK chaperone system. J , Mol. Biol. 249, 126-137. McPherson, P. S., Garcia, E. P., Slepnev, V. I., David, C., Zhang, X., Grabs, D., Sossin, W. S., Bauerfeind, R., Nemoto, Y., and De Camilli, P. (1996). A presynaptic inositol-5phosphatase. Nature 379, 353-357. Meacci, E. Tsai, S.-C., Adamik, R., Moss, J., and Vaughan, M. (1997). Cytohesin-1, a cytosolic guanine nucleotide-exchange protein for ADP-ribosylation factor. Proc. Natl. Acad. Sci. U S A 94, 1745-1748. Mellman, I., Fuchs, R., and Helenius, A. (1986). Acidification of the endocytic and exocyhc pathways. Annu. Rezj. Biocherri. 55, 663-700. Mellman, I., and Simons, K. (1992).The Golgi complex: In vitro veritas? Cell 68,829-840. Minami, Y., Weissman, A. M., Samelson, L. E., and Klausner, R. D. (1987). Building H inultichain receptor: Synthesis, degradation, and assembly of the T-cell antigen receptor. Prnc. Natl. Acad. Sci. USA 84,2688-2692. Moore, S. E. H., and Spiro, R. C. (1993). Inhibition ofglucose trimming by castanospermine results in rapid degradation of nnassembled major histocompatibility complex cIass I niokxules. J. Bid. Chern. 268, 3809-3812. Morinaga, N., Tsai, S.-C., Moss, J., and Vaughan, M. (1996). Isolation of a brefeldin Ainhibited guanine nucleotide-exchange protein for ADP ribosylation factor (ARF) 1 and ARF3 that contains a Sec7-like domain. Proc. Nutl. Acad. Sci. USA 93, 12856-12860. Mostov, K. E., Altschuler, Y., Chapin, S. J., Enrich, C., Low, S. €I., Luton, F., RichmanEisenstat, J., Singer, K. L., Tang, K., and Weimbs, T. (1995). Regulation of protein traffic in polarized epithelial cells: The polymeric immunoglobulin receptor model. Cold Spring Harbor Symp. Quant. Bid. 60, 775-781. Munro, S., and Pelham, H. R. (1987). A C-terminal signal prevents secretion of luminal ER proteins. Cell 48, 899-907. Murphy, C., Saffrich, R., Grummt, M., Gouniier, H., Rybin, V., Rubino, M., Auvinen, P., Lutcke, A., Parton, R. B., and Zerial, M. (1996). Endosome dynamics regulated by a Rho protein. Nature 384, 427-432. Nakano, A., and Muramatsn, M. (1989). A novel GTP-binding protein, SARlp, is involved in transport from the endoplasmic reticulum to the Golgi apparatus. J. Cell Biol. 109, 2677-269 1.
C;URHENT VIEWS I N INTRACELLZJLAR TKANSPORT
411
Nesterov, A., Kurten, R. C., and Gill, G. N . (1995). Association of epidermal growth factor receptors with coated pit adaptins via a tyrosine pliospliorylatioii-regcilatedmechanisni. J. B i d . Chenz. 270, 6320-6327. Nilsson, T., Jackson, M., and Peterson, P. A. (1989).Short cytoplasmic sequences serve as retention signals for transmembrane proteins i n the endoplasinic reticnluin. Cell 58, 707-718. Nishimura, N., and Balch, W. E. (1997). A di-acidic signal required for selective export from the endoplasniic reticuluin. Science 277, 556-558. Novick, P., and Zerial, M. (1997). The diversity of Kab proteins in vesicle transport. C a w . Opin.Cell B i d . 9, 496-504. Ohno, H., Stewart, J., Fournier, M.-C., Bosshart, H., Rhee, I., Miyatake, S., Saito, T., Galluser, A,, Kirchhausen, T., and Bonifacino, J. S. (1995). Interaction of tyrosine-based signals with clathrin-associated proteins. Scieiict. 269, 1872-1875. Okainoto, C. T., Shia, S. P., Bird, C., Mostov, K. E., and Roth, M. G. (1992).The cytoplasmic domain of the polymeric immiiiioglobuliii receptor contains two internalization signals that are distinct froin its I)asolateral sorting signal. J . B i d . Chern. 267, 9925-9932. Orci, L., Glick, B. S., and Rothman, J. E. (1986). A new type of coated vesicular carrier that appears not to contain clathrin: Its possible role in protein transport within the Golgi stack. Cell 46, 171-184. Palade, G. (1975). 1ntrac.ellnlar aspects of the process of protein synthesis. Science 189, 347-358. Palade, G. E. (19Yj3). The fine structure of blood capillaries. J . Appl. Phys. 24, 1424. Palleros, D. R., Welch, W. J., and Fink, A. J. (1991). Interaction of lisp70 with unfolded proteins: Effects of temperature and nucleotides on the kinetics of binding. Proc. Nutl. Acnd. Sci. USA 88, 57194723. Palmer, D. J.. Helms, J. B., Beckers, C. J., Orci, L., and Rothman, J. E. (1993). Binding of coatoiner to Golgi menibranes requires ADP-ribosylation factor. /. Biol. Clzem. 268, 12083- 12089. Parton. R. G. (1996). Caveolae and ciiveolins. Cui-I-. Opin. Cell Biol. 8, 542-548. Parton, R. G., Joggerst, B., and Simons, K. (1994). Regulated internalization of caveolae. J . Cell B i d . 127, 1199-1215. Pastan, I., and Willingham, M. C. (1983). Receptor-mediated endocytosis: Coated pits, receptorsomes and the Golgi. Trends Biochem Sci. 8, 250-254. Pavelka, M., and Ellinger, A. (1983). Effect of colchicine on the Golgi complex of rat pancreatic aciiiar cells. J. Cell B i d . 97, 737-748. Pawson, T. (1995). Protein modules and signalling networks. Nature 373, 573-580. Pearse. B. M., and Robinson. M. S . (1990). Clathrin, adaptors, and sorting. Annn. Reu. Cell B i d . 6, 151-171. Pelham, H. R. (1989). Control of protein exit from the endoplasniic reticulum. Anno. Reu. Cell Biol. 5, 1-23. Pelhain, H. R. (1988). Evidence that lumird ER proteins are sorted from secreted proteins in a post-ER compartment. EMBO J. 7, 913-918. Pelhain, H. R. (1991). Recycling of proteins between the endoplasinic reticuluiii and Colgj complex. Curr. Opin. Cell B i d . 3, 585-591. Peters, P. J., Borst, J., Oorschot, V., Fukuda, M., Kiahenhuhl, O., Tschopp, J., Slot, J. W., and Geuze, H. J. (1991). Cytotoxic T lymphocyte granules are secretory lysosomes, containing both perforin and grarrzyines.J . Ey7. M e d 173, 1099-1109. Peters, P. J., Hsu, V. W., Ooi, C. F., Finazzi, D., Teal, S. B., Oorschot, V., Donakkon, J. G., and Klausner, R. D. (1993. Overexpression of wild-type and mutant ARFl and
412
VICTOR W. HSU AND PETER J. PETERS
ARF6: Distinct perturbations of nonoverlapping ineinbrane compartments. J . Cell B i d . 128, 1003-1017. Peters, P. J., Neefjes, J. J., Oorschot, V., Ploegh, H. L., and Genze, H. J. (1991).Segregation of MHC class I1 molecules from MHC class I niolecules in the Golgi complex for transport to lysosomal compartments. Nature 349, 669-676. Pfeffer, S. R. (1996). Transport vesicle docking: SNARES and associates. Annu. Rev. Cell Deu. Bid. 12, 441-461. Pfeffer, S. R., and Rothman, J. E. (1987). Biosynthetic protein transport and sorting by the endoplasinic reticulum and Golgi. Aiinzr. Aeu. Biochet7~.56, 829-852. Presley, J. F., Cole, N. B., Schroer, T. A., Hirschherg, K., Zall, K. J. M., and LippincottSchwartz, J. (1997). ER-to-Golgi transport visualized in living cells. Nature 389, 81-85. Prill, V., Lehmann, L., von Figura, K., and Peters, C. (1993). The cytoplasmic tail of lysosomal acid phosphatase contains overlapping but distinct signals for basolateral sorting and rapid internalization in polarized MDCK cells. E M B O J. 12, 2181-2193. Radhakrishna, H., Klausner, R. D., and Donaldson, J. G. (1996). Aluminum fluoride stimulates surface protrusions in cells overexpressing ARF6 GTPase. J. Cell B i d 134, 935-947. Rapoport, I., Miyazali, M., Boll, W., Duckworth, B., Cantley, L. C . , Shoelson, S., and Kirchhausen, T. (1997). Regulatory interactions in the recognition of endocytic sorting signals by AP-2 complexes. E M B O J. 16, 2240-2250. Rapoport, T. A,, Jungnickel, B., and Kutay, U. (1996).Protein transport across the eukaryotic endoplasmic reticulum and bacterial inner membranes. Annu. Rev. Biocheni. 65,271-303. Raposo, G., Nijman, H., Stoorvogel, W., Leijendekker, R., Harding, C . V., Melief, C. J. M., and Genuze, H. J. (1996). B lymphocytes secrete antigen-presenting vesicles. J . E x p . Med. 183, 1161-1172. Rodriguez-Boulan, E., and Powell, S. K. (1992). Polarity of epithelial and neurond cells. Annu. Rev. Cell Biol. 8, 395-427. Rolink, A,, and Melcher, F. (1991).Molecular and cellular origins of B lymphocyte diversity. Cell 66, 1081-1094. Rose, J. K., and Donis, R. W. (1988). Regulation of protein export from the endoplasinic reticuhim. Annu. Rev. Cell Bid. 4, 257-288. Rothman, J. E. (1994). Mechanisms of intracellular protein transport. Nature 372, 55-63. Rothman, J. E., and Warren, G. (1994).Iinplications of the SNARE hypothesis for intracellular membrane topology and dynamics. Curr. B i d . 4, 220-233. Rothman, J. E., and Wieland, F. T. (1996). Protein sorting by transport vesicles. Science 272,227-234. Salzman, N. H., and Maxfield, F. R. (1988). Intracellular fusion of seqnentially formed endocytic compartments. J. CeZl BioZ. 106, 1083-1091. Salzman, N. H., and Maxfield, F. R. (1989). Fusion accessibility of endocyhc compartments along the recycling and lysosoinal endocytic pathways in intact cells.J. Cell B i d . 109,20972104. Sanders, S. L., Whitfield, K. M., Vogel, J. P., Rose, M. D., and Schekman, R. W. (1992). SecGlp and BiP directly facilitate polypeptide translocation into the ER. Cell 69,353-365. Sandvig, K., Garred, O., Pry&, K., Kozlov, J., Hansen, S. H., and van Deurs, B. (1992). Retrograde transport of endocytosed Shga toxin to the endoplasrnic reticulum. Nature 358, 510-512. Sandvig, K., and van Deurs, B. (1994). Endocytosis without clathrin. Trevids Cell Biol. 4, 275-277. Saraste, J., and Kuismanen, E. (1984).Pre- and post-Golgi vacuoles operate in the transport of Sernliki Forest virus membrane glycoproteins to the cell surface. Cell 38, 535-549.
C U R R E N T VIEWS I N INTRACELLULAR TRAKSPORT
413
Scales, S. J., Pepperkok, R., and Kreis, T. E. (1997). Visualization of ER-to-Golgi transport in living cells reveals a sequential mode ofaction for COP11 and COPI. Cell 90,1137-1148. Scheiffele, P., Peranen, J., and Simons, K. (1995). N-glycans as apical sorting signals in epithelial cells. Nature 378, 96-98. Schekman, R., and Orci, L. (1996). Coat proteins and vesicle budding. Scietice 271, 15261532. Scherer, P. E., Okamoto, T., Chun. M., Lochsli, H. F., and Lisnati, M. P. (1996). Identification, sequence, and expression of caveolin-2 defines a caveolin gene family. Proc. Natl. Acarl. Sci. USA 93, 131-135. Schmid, D., Baici, A., Gehring, H., and Christen, P. (1994).Kinetics of molecular chaperone action. Science 263, 971-973. Schmid, S. L. (1993). Biochemical requirements for the formation of clathrin- and COPcoated transport vesicles. C u m +in. Cell B i d . 1993, 621-627. Schmid, S . L. (1997). Clathrin-coated vesicle formation and protein sorting: An integrated process. Annu. Rev. Biocheni. 66, 511-548. Schmid, S. L., Fuchs, R., Male, P., and Mellman, I. (1988). Two distinct subpopulations of endosomes involved in membrane recycling and transport to lysosornes.Cell 52,73-83. Schnitzer, J. E., Liu, J.. and Oh, P. (1995). Endothelial caveolae have the molecular transport machinery for vesicle budding, docking, and fusion including VAMP, NSF, SNAP, annexins, and GTPases. J . B i d Chetn. 270, 14399-14404. Schnitzer, J. E., Oh, P., and McIntosh, D. P. (1996). Role of GTP hydrolysis in fission of caveolae directly from plasma membranes. Science 274, 239-242. Schnitzer, J. E., Oh, P., Pinney, E., and Allard, J. (1994). Filipin-sensitive caveolae-mediated transport in endothelium: Reduced transcytosis, scavenger endocytosis, and capillary permeability of select macromolecules. /. Cell B i d . 127, 1217-1232. Schultze, M. P., Peterson, P. A,, and Jackson, M. R. (1994).An N-terminal double-arginine motif maintains type I1 membrane proteins in the endoplasmic reticulum. E M B O 1. 13, 1696-1705. Semenza, J. C., Hardwick, K. G., Dean, N., and Pelham, H. R. (1990). ERD2, a yeast gene required for the receptor-mediated retrieval of luminal ER proteins from the secretory pathway. Cell 61, 1349-1357. Serafini, T., Stenbeck, G., Brecht, A., Lottspeich, F., Orci, L., Rothman, J. E., and Wieland, F. T. (1991). A coat subunit of Golgi-derived non-clathriii-coated vesicles with homology to the clathrin-coated vesicle coat protein beta-adaptin. Nature 349, 215-220. Shiratori, T., Miyatake, S., Ohno, H., Nakaseko, C., Isono, K., Bonifacino, J. S., and Saito, T. (1997). Tyrosine phosphorylation controls internalization of CTLA-4 by regulating its interaction with clathrin-associated adaptor complex AP-2. Immunity 6, 583-589. Simons, K., and Ikonen, E. (1997).Functional rafts in cell membranes. Nature 387,569-572. Simpson, F., Bright, N. A,, West, M. A., Newman, L. S., Darnell, R. B., and Robinson, M. S. (1996). A novel adaptor-related protein complex. J , Cell Biol. 133, 749-760. Simpson, F., Peden, A. A,, Christopoulou, L., and Robinson, M. S . (1997). Characterization of the adaptor-related protein complex, AP-3. J . Cell Biol. 137, 835-845 Sinai. A. P., and Joiner, K. A. (1997). Safe haven: The cell biology of nonfusogenic pathogen vacuoles. Atmu. Reu. Microbial. 51, 415-462. Sitia, R., Neuberger, M. S., and Milstein, C. (1'387).Regulation of membrane IgM expression in secretory B cells: Translational and post-traiislatioiial events. EMBO J, 6, 3968-3977. Sogaard, M., Tani, K., Ye, R. R., Geromanos, S., Tempst, P., Kirchhausen, T., Rothman, J. E., and Sollner, T. (1994). A rab protein is required for the assembly of SNARE complexes in the docking of transport vesicles. Cell 78, 937-948.
414
VICTOR W. MSU AND PETER J. PETERS
Soldati, T., Shapiro, A. D., Dirac Svejstrup. A. B., and Pfeffer, S. R. (1994). Membrane targeting of the sinall GTPase Rab9 is accompanied by nucleotide exchange. Nature 369, 76-78. Sollner, T., Whiteheart, S. W., Brunner, M., Erdjuinent-Bromage, H., Geroinanos, S., Teinpst, P., aiid Rothinan, J. E. (1993). SNAP receptors implicated in vesicle targeting and fusion. Nature 362, 318-324. Stamnes, M. A,, aiid Rothinaii, J. E. (1993). The binding of AP-1 clathrin adaptor particles to Golgi ineinbranes requires ADP-ribosylation factor, a sinall GTP-binding protein. Cell 73,999-1005. Stenbeck, G., Harter, C., Brecht, A,, Herrinann, D., Lottspeich, F., Oric, L., and Wieland, F. T. (1993). Beta prime COP, a novel subunit of coatomer. EMBO J. 12, 2841-2845. Stoorvogel, W., Strous, G. J., Geuze, H. J., Oorschot, V., and Schwartz, A. L. (1991). Late endosomes derive from early endosomes by maturation. Cell 65, 417-427. Sturgill-Koszycki,S., Schlesinger, P. H., Chakraborty, P., Haddix, P. L., Collins, H. L., Fok, A. K., Allen, R. D., Gluck, S. L., Heuser, J., and Russell, D. G. (1994).Lack of acidification of mycobacterium phagosoines produced by exclusion of the vesicular proton-ATPase. Science 263, 678-681. Sussman,J. J., Bonifacino, J. S., Lippincott-Scliwartz,J., Weissinan, A. M., Saito,T., Klausner, R. D., and Ashwell, J. D. (1988).Failure to syuthesize the T cell CD3-zeta chain: Structure and function of a partial T cell receptor complex. Cell 52, 85-95. Suzuki, C. K., Bonifacino, J. S., Lin, A. Y., Davis, M. M., and Klausner, R. D. (1991). Regulating the retention of T-cell receptor alpha chain variants within the endoplasmic reticulum: Ca(2+)-dependent association with BiP. J. Cell Biol. 114, 189-205. Swanson, J. A,, and Baer, S. C. (1995). Phagocytosis by zippers and triggers. T r e d Cell Bid. 5, 89-93. Sztul, E. S., Howell, K. E., aiid Palade, G. E. (1985). Biogenesis of the polymeric IgA receptor in rat hepatocytes. I. Kinetic studies of its intracellular forms. J. Cell Biol. 100, 1248-1254. Takei, K., McPherson, P. S., Schmid, S. L., and De Cainilli, P. (1995). Tubular membrane invaginations coated by dynarnin rings are induced by GTP-gamma S in nerve terminals. Nature 374, 186-190. Tang, Z., Scherer. P. E., Okamoto, T., Song, K., Chu, C., Kohtz, D. S., Nishimoto, I., Lodish, H. F., and Lisanti, M. P. (1996). Molecular cloning of caveolin-3, a novel member of the caveolin gene family expressed predominantly in mu&les.J. B i d . Chern. 271,22552261. Tanigawa, G., Orci, L., Amherdt, M., Ravazzola, M., Helms, J. B., aiid Rothinan, J. E. (1993). Hydrolysis of bound GTP by ARF protein triggers uncoating of Golgi-derived COP-coated vesicles. J. Cell B i d . 123, 1365-1371. Taylor, T. C., Kahn, H. A., and Melancon, P. (1992). Two distinct members of the ADPribosylation factor family of GTP-binding proteins regulate cell-free intra-Golgi transport. Cell 70, 69-79. Teasdale, R. D., and Jackson, M. R. (1996). Signal-mediated sorting of membrane proteins between the endoplasmic reticulum and the Golgi apparatus. Annu. Reu. Cell Dev. B i d . 12,27-54. TerBush, D. R., Maurice, T., Roth, D., aiid Novick, P. (1996).The Exocyst is a multiprotein complex required for exocytosis in Saccharonryce,s cerevisiae. EMBO J. 15, G483-6494. Townsley, F. M., Frigerio, G., and PeIharn, H. R. (1994). Retrieval of HDEL proteins is required for growth of yeast cells. J . Cell Biol. 127, 21-28. Traub, L. M., Ostrom, J. A., and Kornfeld, S. (1993). Biochemical dissection of AP-1 recruitment onto Golgi membranes. J. Cell Biol. 123, 561-573.
C U R R E N T VIE\VS IN INTHACELLULAR TRANSPORT
415
Trowbridge, I. S., Collawn, J. F., and Hopkins, C. R. (1993). Signal-dependent membrane protein trafficking in the endocytic pathway. Annu. Rev. Cell Biol. 9, 129-161. Tschopp, J., and Nabholz, M. (1990). Perforin-mediated target cell lysis by cytolytic T lymphocytes. Annu. Reo. Z t t i m u d . 8, 279-302. Ullrich, O., Hoiiuchi, H., Bucci, C., and Zerial, M. (1994). Membrane association of Rab5 mediated by GDP-dissociation inhibitor and accoiripanied by GDPGTP exchange. Nature 368, 157-160. Urrutia, R., Henley, J. R., Cook, T., and McNiven, M. A. (1997). The dynainins: Redundant or &tinct functions for an expanding family of related GTPase? Proc. Natl. Acad. Sci. USA 94,377-384. Van Aelst, L., and D’Souza-Schorey,C. (1997). Rho GTPases and signaling networks. Genes Dea. 11, 2295-2322. van der Bliek, A. M., Redelmeier, T. E., Damke, H., Tisdale, E. J., Meyerowitz, E. M., and Schinid, S. L. (1993). Mutations in Inmian dyiainin block an intermediate stage in coated vesicle formation. J . Cell Biol. 122, 553-563. Vorhees, P., Deignan, E., vaii Donselaar, E., Huiriphrey, J., Marks, M. S., Peters, P. J., and Bonifacino, J. S. (1995). An acidic sequence within the cytoplasmic domaiii of furin fuiictioiis as a determinant of trans-Golgi network locali7;ation and internalization from the cell surface. EMBO J . 14, 4961-4975. Walter, P., and Jolrnson, A. E. (1994). Sigiid sequence recognition and protein targeting to, the endoplasniic reticulum melnbrane. Annu. Ren. Cell Biol. 10, 87-119. Ward, C. L., Omura, S., and Kopito, R. R. (1995). Degradation of CFTR by the ubiquitiiiproteasonie pathway. Cell 83, 121-127. Warren, G. (1993). Membrane partitioning during cell division. Antitc. Rev. Biochem. 62, 323-348. Weidinan, P. J., Melancoii, P., Block, M. R., and Rothman, J. E. (1989). Binding of an Netliyliiialeiinide-sensitive fusion protein to Golgi membranes requires both a soluble protein(s) and an integral membrane receptor. f . Cell Biol. 108, 1589-1596. Weissinan, A. M., Frank, S. J., Orloff; D. G., Mercep, M., Ashwell, J. D., and Klausner, R. D. (1989). Role of the zeta chain in the expression of the T cell antigen receptor: Genetic reconstitution studies. EMBO J. 8, 3651-3656. Whitney, J. A,, Gomez, M., Sheff, D.. Kreis, T. E., and Mellman, I. (1995). Cytoplasmic coat proteins involved in endosonie function. Cell 83, 703-713. Wiertz, E. J. H. J., Jones, T. R., Sun. L., Bogyo, M., Geuze, H. J., and Ploegh, H. L. (199th). The huirian cytomegalovinis US11 gene product dislocates MHC class I heavy chains from the endoplasinic reticulum to the cytosol. Cell 84, 769-779. Wiertz, E. J. H., J., Tortorella, D., Bogyo, M., Yu, J., Mothes, W., Jones, T. R., Rapport, T. A,, and Ploegh, H. L. (1996b). Secfil-mediated transfer of a membrane protein froin the endoplasniic reticulum to the proteasoine for destruction. Nature 384, 432-438. Willey, R. L., Maldarelli, F.. Martin, M. A,, and Strebel, K. (1992).Human immunodeficieiicy virus type 1 Vpu protein induces rapid degradation of CD4. f . Virol. 66, 7193-7200. Wilson, D. W., Wilcox, C. A,, Flynn, G. C., Chen, E., Kuang, W. J., Henzel, W. J., Block, M. R., Ullrich, A,, and Rothman, J. E. (1989). A fusion protein required for vesiclemediated transport in both mammalian cells and yeast. Nature 339, 355-359. Wubbolts, R., Fernandez-Boja, M., Ooinen, L., Verwoerd, D., Janssen, H., Cdafat, J., Tulp, A,, Dusseljee, S., and Neefies, J. (1996). Direct vesicular transport of MHC class I1 inolecules froin lysosomal structures to the cell surface. f . Cell B i d . 135, 611-622. This article was accepted for publication on February 4,1998.
This Page Intentionally Left Blank
Phylogenetic Emergence and Molecular Evolution of the ImmunoglobuIin Family JOHN J. MARCHALONIS,' SAMUEL F. SCHLUTER,' RALPH M. BERNSTEIN,t SHANXIANG SHE"+ AND ALLEN 8. EDMUNDSON~ 'Department of Microbiology and Immunology, College of Medicine, Univerdy of Arizona, Tucson, Arizona 85724; 'FDA/cher/HFM-54 1 ond #Nationallnstiiutes of Healh, khesda, Marylond 20892; ond §OklohomoMedical Research Foundation, Okkahoma City, Oklahoma 63 104
1. Introduction
There has been a long fascination with the evolutionary origins of the human immune system that began in the past century (Beard, 1894; Metchinkoff, 1884; Noguchi, 1903; Widal and Sicard, 1897), showed a burst of renewed interest following tlie elucidation of antibody structure and cellular immunology in the 1960s and 1970s (Good and Papermaster, 1964; Grey, 1969; Hildeinann, 1974;Marchalonis, 1977),and gained considerable recent impetus (Du Pasquier, 1993; Hsu and Steiner, 1992; Hughes and Yeager, 1997; Klein, 1989; Litinan et al., 1993; Mardialonis and Schluter, 1989; Thompson, 1995) because of the application of recombinant DNA technology that allowed the precise determination of the sequences of immunoglobulins ( Ig) leading to an estimation of their evolution. This review focuses on the appearance and molecular evolution of antibodies and T-cell receptors (TCR), which are the antigen-specific recognition elements of the coinbiiiatorial immune system. With cei-kain exceptions, such as tlie modified y heavy chain of the camel (Muydemians et al., 1994) and a variant heavy chain of the nurse shark (Greenberg et al., 1995) that form only liomodimers, the functional antigen recognition units of iinmunoglobulins and TCRs are heterodimers consisting in Igs of light chainheavy chain pairs and their multimers, and in TCRs of pairs of chains of approximately the same size; Cwlp or ylS. The hallmarks of the combinatorial immune response are the existence of N-terminal variable domains in each chain of a lieteroclirner, which allows the formation of potentially enormous numbers of combining sites for antigens, the mechanism underlying the generation of diversity by recombination of gene segments, and the existence of lymphocytes expressing cornbinatorial receptors ( B a d et al., 1994; Klein, 1989; Marchalonis et al., 1996; Marchalonis and Schluter, 1989). 417
418
JOHN J. MARCHALONIS et al.
II. Evolutionary Emergence of the Combinatorial Immune System
The origin of the cornbinatorial iinmune system appears to be an explosive event, resulting in the appearance of large arrays of canonical V-C domains and joining segments of Igs and TCRs coincident with the emergence of the recombination activating system (RAG) genes in a relatively brief evolutionary span of approximately 10 million years immediately prior to the phylogenetic emergence of jawed vertebrates. Therefore, the molecular evolution of the combinatorial immune system will be considered as two major stages. The first step is its rapid establishment approximately 450 million years ago as documented by the presence of RAG genes (Bernstein et al., 1994, 19962; Greenhalgh and Steiner, 1995), Ig light (Greenberg et al., 1993; Hohinan et al., 1992; Schluter et al., 198913) and heavy chains (Litman et nl., 1985b), and TCRa, p, 7 , and 6 (Rast et al., 1997; Rast and Litman, 1994) chains, as well as molecules of the major histocompatibilitycomplex (MHC) (Bart1and Weissman, 1994; Hashimoto et al., 1992; Kasahara et al., 1992,1993)in a wide variety of extant cartilaginous fishes, which are the most primitive living gnathanstomes. Surprisingly, all attempts to establish the existence of these genes in the more primitive agnathan vertebrates, lampreys and hagfish, have as yet proved unsuccessful. Molecular explorations in this area are continuing with vigor, but it is now generally accepted that lower deuterostoines such as starfish and sea urchins (Smith and Davidson, 1992), as well as the typical invertebrate protostomes such as insects and crustaceans and acelomates, e.g., flat worms, lack the definitive elements of the coinbinatorial immune response (Klein, 1989; Marchalonis and Schluter, 1990a,b). 111. Ancient Foundations of the Combinatorial immune System
The immune system of vertebrates has mechanistic underpinnings in ancient systems for cell recognition, activation, and differentiation (Beck et al., 1994; Klein, 1989; Marchalonis and Schluter, 1990a),but is definitively characterized by the presence of lymphocytes expressing bonafide immunoglobulins or T-cell receptors as recognition units and containing the gene segments and rearrangement machinery necessary for the generation of the combinatorial immune response. The term “combinatorial” is used in the orignal sense of formation of heterodimeric receptors with the number of variable domains of one type designated p and the number of V domains ofthe other designated q such that the total number of combinations is p X q (Gally and Edelman, 1972). In addition, the updated definition entails the presence of V, J, and sometimes D segments and the possibility of intragene cluster rearrangement. The simplified phylogenetic
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
419
tree in Fig. 1 is based on patterns of embryonic development (Davidson et al., 1995) that indicates the distribution of genes and molecules involved in ancient and inflaininatory mechanism as well as those characteristic of the vertebrate coinbinatorial iminune response. The divergence between deuterostomes and protostomes occurred approximately 670 million years ago (Carroll, 1988). Available evidence indicates that the coinbinatorial iminune response is restricted to deuterostoines where it is unequivocally present in all placoderm-derived vertebrates; the so-called gnathanstoines or jawed vertebrates, which include the group from cartilaginous fishes to mammals. The vertebrate-type iinmune response was opportunistic, however, in that it built on primordial recognition mechanisms of innate immunity and inflammation that occur widely throughout the animal kingdom where they are found in protostoines, deuterostomes, and even acelomates. For example, lectins related to C-reactive proteins of the pentraxin family are found in arthropods such as the horseshoe crab (Ying et al., 1992) and also in all vertebrates studied (Liu et nl., 1994) where inter alia they act as acute-phase proteins in response to pneuinococcal infection
ArIhr
P-
Araihnids
Nematodes Nemertean worms
qhates
?I=
m
Turbellarianflatworms Amlomates
Coelenterates
Plants
Primitive eukalyote
Y
c
M
Amoebae
R
Sponges
T 88
Flagellates Yeasts
#
0
N-CAM CRP Antibody Complement Major Histocompatibilitymolecules Recombinationactivating genes T Cell receptor '"BigBang" of gene duplication IgW
FIG.1. Phylogenetic tree summarizing evolution based on developmental mechanism and body plans and depicting the types of immune surveillance molecules and their distribution throughout phylogeny. The shaded region is that critical to the origins of the vertebrate immune response. The combinatorial immune system of jawed vertebrates apparently arose as a "big bang" after the ostracodedplacoderm split (BB3) or just before the origin of the chordates (BB1 or BB2).
420
JOHN J. MARCHALONIS et nl
(Volanakis et nl., 1990). The immunoglobulin superfamily includes the bona Fcle immunoglobulin variable and constant domains, as well as a diverse set of molecules that show little homology in amino acid sequence but are characterized by the expression of the “immunoglobulin fold” in three-dimensional structure (Bork et al., 1994; Doolittle, 1995; Williams and Barclay, 1988). Because the first members of the Ig superfamily characterized, including members of the major histocompatibility complex and adhesion molecules functioning in interactions among cells (Williams and Barclay, 1988),were restricted to cell surfaces, it was proposed that these constituted extracellular molecules involved in recognition. However, more recent studies have established that large internal molecules involved in muscle attachment, such as “twitchin” in nematodes and insects (Ayme-Southgate et al., 1991; Benian et al., 1989), also contain a recognizable Ig domain motif. Thus, the identification of a molecule as a member of the Ig superfamily is not instructive by itself with respect to either recognition function or relationship to the immunoglobulin family members that carry out the antigen-specific recognition role in the combinatorial immune response (Bork et al., 1994). This point is critical because many surface receptors, particularly those for cytokines, which are often members of the general Ig superfamily, do not bind ligand in the characteristic heterodimeric association of V1N2 as do Igs and TCRs. Doolittle (1995) estimates that approximately 40% of cell associated proteins can contain Ig domain motifs. Members of the superfamily are distributed widely in phylogeny, with “twitchin” type molecules occurring in both protostorne insects (Seeger et al., 1988; Sun et al., 1990) and acelomate nematodes (Benian et al., 1989). Furthermore, superfamily members described in other acelomates, including sponges (Schacke, 1994),and viral hemaglutinins (Jin et al., 1989) also show immunoglobulin domain motifs. An additional characteristic feature that distinguishes true Igs (Igs and TCRs) from other members of the superfamily is that bona$cZe Igs only form covalent peptide linkages with other Ig domains, whereas many distal superfamily members occur as parts of large molecules, which may also contain lectin domains, fibronectin domains, and activation molecules including tyrosine kinases (Bork et al., 1994; Schacke, 1994). It is of interest with respect to the phylogenetic distribution of inducible defense mechanisms that both insects (Sun et id., 1990) and molluscs (Hoek et al., 1996) produce or modulate levels of soluble molecules similar to neural cell adhesion molecules (termed hemolins) in response to bacterial infection or vaccination with various material. These molecules do not show variability in amino acid sequence, but act more like nonspecific acute-phase proteins rather than induced antibodies or TCR. Molecules directly involved in cell activation of the inflammatory
EMERGENCE A N D EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
421
type are also distributed widely, and some of these also Gall within the “immunoglobulin superfamily.” Activity comparable to that of the macrophage produced cytokine interleukin (1L)-1 is found in lower deuterostomes (Beck et al., 1989; Raftos et al., 1991) where it is associated with a molecule serologically cross-reactive with the mammalian cytokine. The receptor for IL-1 is a member of the Ig superfamily (Sims et al., 1988), and the activation system requires the transcription factor Nf-KB (Baeuerle and Henkel, 1994). Differentiation in insects entails transcription factors homologous to Nf-KB (Ip et al., 1993), and the “toll” surface molecule acts in a manner very similar to the IL-1 receptor (Hoffman and Reichhart, 1997; Ip et al., 1993; Lemaitre et al., 1996). Thus, differentiation and cell activation mechanisms in protostomes and in the vertebrate immune system can be based on fundamentally similar mechanisms. It must be stressed that protostomes exhibit inducible defense reactions in response to infection, but that these responses do not represent the combinatorial immune response characteristic of jawed vertebrates. Attempts to delineate relics of initial steps in the combinatorial immune response by investigating lower deuterostomes such as echinoderms, tunicates, and even the most primitive living vertebrates, the agnathan cyclostomes, as represented by lampreys and hagfish, have thus far been inconclusive. Although a number of workers have described antibody-like proteins in cyclostomes (Litman et al., 1970; Marchalonis and Edelman, 1968; Raison and Hildemann, 1984; Varner et al., 1991), definitive structural studies or gene characterizations have not yet been published. Lampreys and hagfish have, however, been shown to possess molecules clearly related to complement components of higher vertebrates (Hanley et al., 1992; Ishiguro et al., 1992; Nonaka et al., 1994) and to have cells with the morphological characteristics of lymphocytes and plasma cells (Miller and Ratcliffe, 1989; Raison and Hildemann, 1984; Zapata and Cooper, 1990). Currently, a number of investigators (Litman et nl., 1993), including the authors, have used recombinant DNA primers and probes applicable to all jawed vertebrates, but have been unable to obtain definitive evidence for genes specifjmg Igs, TCRs, and the enzymes necessary for Ig genetic recombination in hagfish and lampreys. To approach the phylogenetic emergence of the rearrangement mechanism, the gene specifying the recombinase activating gene 1 (RAG-1) of carcharhine sharks has been isolated (Bernstein et al., 1994) using fully degenerate polymerase chain reaction (PCR) primers derived from conserved peptide segments of RAG-1 from human, mouse, chicken, and Xenopus (Greenhalgh et al., 1993; Schatz et al., 1989, 1992). The same primers allowed the isolation of homologs from the paddlefish (a chondrostean fish), the axolotl (an amphibian), the goldfish (a teleost), and the pig
422
JOHN J. MARCHALONIS et a!.
(a mammal that had not been previously studied). The deduced amino acid sequence of all known RAG-1 gene products corresponding to residues 808-1105 of the clawed toad Xenopus showed remarkable homology considering the long evolutionary divergence of the distinct species. For example, the shark sequence showed 75% identity to the goldfish, 82%identity to the xenopus, and 87% identity to the human sequence. Thus, there is no question that the most primitive living jawed fishes, as represented by chondrichthyans, possess the genetic machinery essential for the expression of the combinatorial immune response. This is consistent with intracluster rearrangement of V, D, and J elements (Kokubu et al., 1988a). A cDNA clone specifying the complete RAG-1 homolog of the bull shark Garcharhinus leua~s(Bernstein et al., 1996a) has a translated amino acid sequence consisting of 1113 residues that can be arranged into six domains expressing degrees of identity to human varying from less than 20% to greater than 80%. Consistent with previous reports from studies of higher vertebrates (Schatz et al., 1989,1992),the shark RAG-1 molecule contains a “RING” Zn-finger motif that is most similar to that of the yeast DNA postreplicative repair protein RAD-18. Figure 2 presents a comparative alignment of the complete RAG-1 sequences of shark, human, and chicken. An extremely interesting homology between a segment of the shark RAG-1 and resolvase, a member of the bacterial integrase family of site-specific recombinases, and to topoisomerase 1 possibly provides a clue to the function of the gene product in recombination (Dik et al., Gellert, 1996). Because RAG-1 shows homology to the integrase (INT) family of site-specific recombinases (Bernstein et al., 1996a; Hughes and Yeager, 1997), and RAG-1 functions in association with RAG-2 in the recombination of gene segments specifylng immunoglobulins and TCRs, RAG-2 has been analyzed for homology to integration host factors (IHF),
FIG.2. (A) Alignment of the comple translated 1113 amino acid sequence of the bull shark (S) (CarcharhinusZeucrw) RAG-1 with chicken ((2) and human (H) RAG-1. Alignments were made using CLUSTALW (Thompson et al., 1994). Amino acid identities shared with the chicken and human sequences are shaded, whereas universal identities shared with shark, human, chicken, trout, rabbit, andxenopus RAG-1 are indicated by an asterix. Domain 4,which shows homology to the bacterially encoded integrase (INT) family of site-specific recombinases, is boxed. The 10 sequential repeating shark RAG-1 TILEDD motif hexapeptides are underlined. (B) A pylogenetic tree depicting the evolutionary relationship between members of the INT family of site-specific recombinases. The RAG-1 INT motif is shown to have diverged prior to 1.8 billions of years ago (BYA) from a common ancestor with the bacterial integrase family of recombinases. The alternative possibility that the RAG-1 INT motif was derived by lateral transfer of a microbial integrase family gene prior to 0.4 BYA is also shown. The date for the divergence of the eukaryotic and prokaryotic lineages, 1.8 BYA, is from Donelson (1995), Doohttle (1994), and Doolittle et al. (1990).
A QKP
** *
*****
moskc *i;i+ir*
** *
***
--*
* *** *
*******
ZT T * * * * * * f * * * i * * *
DDTILEMIQYLEMIQYLK
*
T F P T W ScpAKW.2 SCPAK&P
*
** *
*** ** **
* *,
+
. . . . .VPTQPL
* ** **
1MHRKFSSAP.C
*
* **
*
mEps
STWK
SllKBK
******** * * ** ****** ********** . . . . . . . . . . . . . . . . . . . . . . . . . .
~GRCASASAEII~QD~1~QSDTIEO;DDTI~TILG~DTMLEDDTMLENDTILGDDTILEDDTMLEDDTILG AX#KTCR----- - - -----SQ-GFALDPDDG SLPPKISLESNDsVF& -------------TSGFTMNPQASWPLGIEDSLESQDSMEF
424
JOHN J. MARCHALONIS et a!
B Billions of Years Ago 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0
marnalian RAG I
FIG.2-Continued
molecules known to act in association with integrases in microbes. There was sufficient identity between RAG-2 of higher vertebrates and IHF to suggest that vertebrate recombinase activating genes and bacterial sitespecific recombinases share a common ancestry and may have similar modes of action. Thus, even though the recombinase activating genes are necessary for the characteristic combinatorial immune response of vertebrates, their appearance in the vertebrate immune system may provide yet another example of the opportunistic nature of this system in coopting or incorporating elements from other systems. This conclusion is extremely tantalizing because the authors and others have used the same PCR primers for RAG-1 that allow successful isolation of the gene in all gnathostome vertebrates to search for homologs from agnathans and lower deuterostomes with no success to date. Thus, the apparently sudden appearance of the RAG gene system in the jawed vertebrates may have been due to a lateral transfer of microbial genes approximately 450 million years ago, as was first suggested by Schatz et nl. (1992). The apparent “big bang” or sudden evolutionary emergence of the essentially complete combinatorial immune response thus entails both the incorporation of a genetic recombination mechanism and the emergence of bona fide immunoglobulin domains.
EMERGER’CE A N D EVOLUTION O F TIiE IMMUNOGLOBULIN FAMILY
42s
IV. Emergence of Bona Fide Immunoglobulins
Based on initial iminunocheinical findings that sharks (Clem and Small, 1967; Marchalonis and Edelman, 1965, 196610) and amphibians (Marchalonis and Edelinan, 1966a) had circulating Igs consisting of light and heavy chains that showed electrophoretic mobilities consistent with charge diversity, it was not surprising that the application of recombinant DNA technology established the presence of genes homologous to those of human IgM in primitive elasmobranchs (Kokubu et al., 1988a; Litman et al., 1985a). Two remarkable conclusions were drawn, however. The first was that the arrangement of gene segments specifjmg the elasmobranch chain was markedly different froin the arrangement in mammals. The VH,DH,JH, and CH segments occurred in individual clusters or “cassettes” rather than in the translocon arrangement typical of mammalian immunoglobulin genes (Hinds and Litinan, 1986; Kokubu et al., 198th). This observation and its implication are both considered later. The second unexpected finding was the extreme diversity in types of the two domain VJC molecules corresponding to light chains and TCRs in chondrichthyan fishes. An extensive array of light chains including families homologous to K chains (Greenberg et al., 1993; Marchalonis et al., l988c), more than three types of y-like families (Hohman et nl., 1993, 1995; Schluter et al., 1989b) as well as other forms, possibly restricted to elasmobranchs (Rast et nl., 1994; Shamblott and Litman, 1989a), were found to occur in all representatives of the chondrichthytes that were investigated in detail. Furthermore, only a single type of heavy chain corresponding to mammalian p chain was expected (Marchalonis, 1977). However, at least two, and possibly three, heavy chain isotypes have been described (Bernstein et al., 1996b; Greenberg et d . ,1996; Harding et al., 1990).In analyses of cDNA specifying T-cell receptors, Litman and colleagues have identified sequences corresponding to TCR alp in the horned shark (Hawke et al., 1996; Rast et al., 1995; Rast and Litman, 1994) and d/3 and ylS in the distantly related clearnose skate (Rast et al., 1997). Thus, there was an apparently great evolutionary leap in a relatively short time corresponding to the separation of ancestral jawed vertebrates froin their agnathan ancestors in which the definitive elements of the coinbinatorial immune response emerged in functional form. In addition, bona Jide MHC class I and class I1 antigens also occur in elasmobranchs (Bart1 and Weissman, 1994; Kasahara et nl., 1992, 1993; Klein et nl., 1991), but are apparently lacking in the agnathans. These inolecules are required for the presentation of peptide epitopes to cell surface-associated TCR, and their presence suggests the existence of functional antigen-specific T cells in cartilaginous
426
JOHN J. MARCHALONIS et a[.
fishes. However, such cells have not yet been shown in functional or immunohistochemical studies (Marchalonis et al., 1993; McKinney, 1992), although TCR p chains have been detected serologically in detergent extracts of sandbar shark spleen cells (Marchalonis et nl., 1997). Both rays (Beard, 1894) and sharks (Luer et al., 1995)possess Characteristic thymuses that appear early in ontogeny and involute with aging, as well as characteristic spleens and gut-associated lymphoid tissue (Zapata and Cooper, 1990). The second major development in the combinatorial immune system occurred with the emergence of mammals involving the formation of defined germinal centers within the lymph nodes (Liu et al., 1992), the appearance of IgG immunoglobulin (Marchalonis, 1977), and the process of affinity maturation following from somatic mutation under antigen selection and requiring the reexpression of the products of RAG genes in the germinal centers (Han et al., 1996). The application of recombinant DNA technology was necessary to enable characterization of immunoglobulins of primitive vertebrates such as elasmobranchs because these animals possess circulating immunoglobulins having variable region diversity comparable to those of higher vertebrates . the case of humans and mice, it was possible (Marchalonis et al., 1 9 8 8 ~ )In to obtain monoclonal antibodies because of the occurrence of multiple myeloma, a disease in which monoclonal plasma cells are generated in large quantity. Because of the relative ease of cloning and characterizing genes, there has been an explosion of information regarding molecules possibly related to immunoglobulins, and it is necessary to comment on the criteria defining members of the immunoglobulin family (Marchalonis and Schluter, 1989) as opposed to the more general superfamily (Bork et nl., 1994; Doolittle, 1995; Williams and Barclay, 1988). The immunoglobulin superfamily is an extremely large grouping of molecules that share the immunoglobulin fold, but may be involved in many cell functions and lack a clear-cut ancestral homology (Bork et al., 1994; Doolittle, 1995; Edelman, 1987; Hultgren et al., 1992). In contrast, the immunoglobulin family thus far has been detected only in the jawed vertebrates with its members being immunoglobulin light and heavy chains, T-cell receptor a,/3, y , and S chains, and the IgC-like domains of MHC. Other workers have referred to “bona$cle” Ig domains as C1 type and concur that they are restricted (thus far) to gnathostomes (Du Pasquier, 1993). Superfamily members range from molecules such as superoxide dismutase, which exhibits the Ig fold but does not have significant homology to Ig molecules, to others that have less than 20% identity to conventional Igs but show stretches of sequence homology in conserved regions responsible for folding or dimerization, e.g., invertebrate hemolins (Sun et al., 1990) or vertebrate cell surface adhesions molecules (Edelman, 1987). In
EMERGENCE A N D EVOLUTION OF T I I E IMMUNOGLOBULIN FAMILY
427
contrast, family members exhibit greater than 30% identity (Marchalonis and Schluter, 1989), whether in coinparisons of orthologous niolecules among the set of jawed vertebrates or in coinparisons of distinct family members within a species. For example, CA domains of humans and sandbar sharks and human C K and CA domains are approximately 40% identical. Statistical coinparisons of individual domains are useful in establishing Ig family relationships, particularly if database searches identify Igs as the closest matches with high probability, e.g., the closest match for sandbar shark VL doinains was a human VAVI sequence (Hohman et al., 1992, 1993)and all of the other matches were light chain variable domains of higher vertebrates. Ig or TCR chains do not form covalent peptide bonds with unrelated molecules, with the exception of transmembrane and short cytoplasmic segments of membrane-bound TCR and Ig heavy chains. This property serves as another useful criterion for the identification of bona Fde iininunoglobulin domains. Statistical analyses using the programs ALIGN and RELATE established that the ineinbrane-proximal domains of MHC class I (a3)and class I1 (a2 and P2) molecuIes are clearly related to Ig family constant domains (Beainan et al., 1987; Marchalonis et al., 1987a,b). These doinains are referred to as immunoglobulin-like modules with the N-terminal domains, which are unrelated to irninunoglobulins and are involved in peptide binding, being termed peptide-binding modules (Klein et al., 1991).The same types of analysis indicated that the human TCR /3 chain constant domain was clearly hoinologous to the constant doinain of the A light chain (Beainan et al., 1987; Hedricket al., 1984; Marchalonis et al., 1987a,b).Furthermore, these domains show greater than 30% identity in sequence as well as the sharing of conserved characteristic sequences, particularly the sequence KATLVCL in /3 band 4-2(B), which is a characteristic inotif of both TCRP chains and A light chains (Schluter et al., 1989b). The dendogram given in Fig. 3 depicts the relationships among Ig constant doinains of light chains and TCRs compared with the most Vregion-distal domains of heavy chains (e.g., the third constant domain of the y chain and the fourth constant domain of the p chain) and the iinmunoglobulin-like modules of MHC products and the inducible Ig superfamily inolecule of the cecropia moth (hemolin). Because of the low identity between heinolin domains and those of classical IgC regions, this molecule represents an out group in this analysis. The MHC domains and the human PZ inicroglobulin form a separate cluster that shows the expected distinctions of class I and class 11 between a and P chains and also illustrates phylogenetic discrimination within the clusters. Heavy chains and light chains form separate groupings within the Ig constant domains, and the TCR chains distribute as an early off-shoot of the light chains.
428
JOHN J. MAHCHALONIS et ol.
FIG.3. Rooted phylogenetic tree depicting the relationships among immunoglobulin constant domains of light chains, heavy chains, T-cell receptors, and MHC products. The ultimate carboxyl-termind domain was aiidyzed for heavy chains (e.g., Cp4 for IgM, Cy3 for IgC). Only the constant domain-like segments of the MHC products (Marchdonis et al., 1984, 1994) and heinolin (Sun et al., 1990) were included in the analysis. The tree was constructed using the progressive alignment procedure of Feng and Doolittle (1990).
The sandbar shark C h clusters with the C A group, and K light chains of mammals and amphibians form their own grouping. In addition to the Alike light chains of sharks (Schluter et d., 1989a), the horned shark has a separate type apparently restricted to the chondrichthyans (Rast et ul., 1994) and the nurse shark has a K-like light chain (Greenberg et al., 1993). All of these individual light chain types are apparently found in each type of shark studied, but the relative proportions vary in different species. For example, the h-like light chains comprise the major group of families in carcharhine sharks, whereas the major light chain type in the nurse shark resembles K. In contrast, that of the homed shark (Heterodonti~sfranchescii) is apparently restricted to sharks and it has been termed type I by Rast et aZ. (1994). Overall, this is a simplified diagram meant to illustrate the overall pliylogenetic relationships of the constant domains. The Ig-like domains of MHC products form a cluster separate from that of immunoglobulin chains and the TCR p chains cluster with the light chains. It is noteworthy that representatives of all of these immunoglobulin types are present in a clearly recognizable fashion in the most primitive extant jawed vertebrates.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
429
Figure 4 gives a comparative alignment of Ch type domains including Ch and CP of humans and sharks to incorporate maximal evolutionary distances. The /3 bands defined by X-ray crystallography of the human Ch molecule (Edmundson et al., 1975), which are also defined in crystallographic studies of TCR (Bentley et al., 1995; Bentley and Mariuzza, 1996; Garcia et al., 1996), are identified in Fig. 4.Ig and TCR domains consist of two P-pleated sheets, one made of four bands and the other consisting of three bands. Both variable and constant domains have this basic structure, but variable doinains usually have two extra P bands defining the second complementarity determining region (CDR2).An immunoglobulin three-dimensional structure has been reviewed in many places and will not be considered in detail here, except to provide a background for the evolutionary comparisons. Substantial homology exists within the P bands of these four domains, particularly in the 4-2 (B) band that is involved in forming the contact surface with either CIIl or TCR Ca, respectively. In addition, the sentinel tryptoplian (W), which is 14 residues distal to the cysteine, is uniforinly conserved. The constant domains of human and shark h chains form extremely similar predicted three-dimensional structures because of the similarity in P band size and sequence and length of
SbSCh GN HUCB EDLNK HnSCB G E N D T
___-__-----____----
Frc. 4. Comparative alignment of light chain constant region type domains; human (Hu) Clt (Kabat et nE., 1991), sandbar shark (Hohman et nl., 1993), human C@(Yanagi et NE., 1984), and horned shark (HnS) C@(Rast and Litman, 1994) constant region sequences are shown. The position of @ bands identified by X-ray crystallographyof the human CA molecule (Edmundsonetnl., 1975)is indicated. Sequence identities are shaded, and alternate identities are boxed.
430
JOHN J. MARCHALONIS et ol
intervening loops (Schluter et nl., 198913). Figure 5 illustrates the basic structure of two domain (V-C) Igs in evolution by superimposing the predicted structure of the TCRP chain upon that of the human A light chain Mcg (Kaymaz et al., 1993). Both the three and the four chain 0 band-defined surfaces are extremely similar, but the CP has large loops projecting forward that restrict the mobility of the V/3 domain. The C a domain has a more compact structure that differs somewhat from that of other IgC domains (Bentley and Mariuzza, 1996; Garcia et al., 1996), but all can be modeled in this general sense (Marchalonis et nl., 1994). This structure, which represents one-half of an I4H or a/P heterodimer, presents the general features of the immunoglobulin fold and the location of the framework and hypervariable segments. Although functional antigenbinding Igs or TCRs consist of heterodimers, individual chain types such as light chains can form LA. diiners and mutated heavy chains of camels (Muydermans et al., 1994) and nurse sharks (Greenberg et nl., 1995),which lack appropriate contact residues for the formation of heterodimers with light chains, can form homodimers. V. Immunoglobulins and T-cell Receptors of Jawed Vertebrates
The most striking feature of the evolutionaryemergence of the combinatorial immune system is the apparent conclusion that it is either not present at all or occurs in full functional capacity. As described earlier, the system
r
4 (D)
r-3
3-3 (G)
B
FIG.5. Three-dimensional model of human TCRP immunoglobulin domains superimposed on the structure of the human h chain Mcg. P bands are labeled using crystallographic nomenclature (Edmundson et nZ., 1975).Excess length in CP loops relative to those in the C h are shaded. Based on Marchdonis et aE. (1997).
EMERGENCE A N D EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
431
has coopted ancient inechanisins for cell recognition and division, but the full system having bonnjde iminunoglobulin V and C domains as well as joining segments and the requisite enzymes for rearrangement is undeniably present in species representing the most primitive jawed vertebrates, i.e., a broad range of chondrichtyes including chimeras (the ratfish Hyclrologus), rays (including the clearnose skate), and ancient (Heterodontus)and more recently evolved (Carchnrhinus) sharks. A diagrammatic overview of b o n a j d e Igs of gnathanstomes is given in Fig. 6. TCR and Ig light chains contain one V and one C domain with the coinplete V structure of Ig light chains and TCRa and y chains derived from V and J segments, and those of Ig heavy chains and TCRP and S chains composed of V, D, and J segments. Some coinrnents regarding nomenclature are required here because isotypes and chains of individual species were named on the basis of historical discovery rather than on a rational appraisal of relationships. Light chains homologous to human K and A chains are found in all gnathanstomes, and other classes lacking clear similarities to these occur in elasmo-
Ftc:. 6. Schematic representations of the domain structures of TCRs and immunoglobulins of vertebrate species. TM. transmembrane segment; 11, hinge segment.
432
JOHN J. MARCHALONIS et al.
branchs (Rast et al., 1994),teleosts (Ghaffari and Lobb, 1992, 1993; Lundqvist et al., 1996; Marchalonis and Schluter, 1990a) and amphibians (Haire et al., 1996; Schwager et al., 1991). TCRs are glycosylated and contain Cterminal transmembrane segments allowing them to associate with the plasma membranes of T cells, whereas light chains lack carbohydrate moieties as well as transmembrane segments. The p heavy chain, which was probably the first to emerge in evolution and ontogeny (Marchalonis, 1977), consists of a V domain formed of V, D, and J segments and four constant domains, Antibodies of this isotype occur as soluble molecules in serum and as a B-cell membrane receptor for antigen in all species. The degree of polymerization of the (L2p2) monomers can vary within various taxa with pentamers in sharks and mammals, tetramers in teleosts, and hexamers in some amphibians. Furthermore, the membrane-associated forms can differ in that the IgMM of sharks and mammals are larger than the secreted forms because the transmembrane pieces are linked to the C terminus of the Cp4 domain, but these forms are usually smaller in teleosts (Warr et al., 1976, 1992; Wilson and Warr, 1992) because the transmembrane gene segments associate with the Cp3 and the Cp4 gene segment is lost. The IgW (Bernstein et al., 199613) isotype, which has also been called IgNARC (Greenberg et al., 1996) of sharks, contains extra constant domains relative to the p chain. It does not seem to be present in appreciable quantities in serum, but may function as a second B-cell receptor. An extremely large heavy chain found on some catfish B cells (Wilsonet al., 1997)contains the C p l domain and seven additional constant domains expressing homology to C6. The actual Ig 6 chain of primates and mice has three C domains and serves as a second antigen receptor on primary B cells. Heavy chains containing one VHand two CHdomains occur in lungfish (Marchalonis, 1977), turtles (Grey, 1969), and ducks (Warr et aE., 1995), where they were originally termed “Nu” chains of the IgN isotype. Heavy chains of the same size were reported in skates, where they were called IgX (Kobayashi and Tomonaga, 1988; Kobayashi et al., 1992). The low molecular weight serum Igs of chickens (Leslie and Clem, 1969) and amphibians (Atwell and Marchalonis, 1975) are larger than the IgG of mammals in overall size: 7.5s (180 kDa) as opposed to 6.7s (150 kDa) in sedimentation coefficient (and mass) and in the size of the heavy chain, 65-68 kDa as opposed to 50 kDa. Although the chicken molecule was originally thought to be equivalent to IgG, Leslie and Clem (1969) termed the molecule IgY to emphasize the obvious distinctions between these molecules. Antibodies with these properties were also termed IgRAA to denote their presence in reptiles, amphibians, and avians, but also to raise the hypothesis that the non-IgM heavy chains of distinct vertebrate classes might have arisen
EMERGENCE AND E V O L U T I O N OF THE IMMUNO(I:LOHUI.IN
FAMILY
433
independently of one another (Atwell and Marchalonis, 1975).Warr et al. (1995) proposed that IgN and IgX are forms of IgY and that IgY is itself the lineal precursor of IgE and IgG. The E chain of IgE contains four constant domains, and the 7 chain of IgG, the major immunoglobulin of mammals, contains three constant domains with the original CH2domain contracted to become a “hinge” region. All heavy chains are glycosylated and can be expressed in membrane form, with IgM and IgD on primary B cells and the other isotypes expressed on memory cells. A. IMMUNOGLOBULINS OF CI-IONDHICIITIIYTES
The degree of diversity in the set of immunoglobulins composed of two domains (one V and one C) in cartilaginous fishes was surprising. Light chains orthologous to K (Greenberg et al., 1993) and A (Hohman et al., 1992; Schluteret nl., 1989a,b),as well as others (Rast et nl., 1994; Sharnblott and Litman, 1989a),are widely distributed among extant species. Carcharhine sharks, which comprise approximately 60% of living sharks (Compagno, 1988), have h-like light chains as the dominant isotype. At least three distinct sets of h-like chains are found in the sandbar shark, for example (Hohman et nl., 1995), but light chains orthologous to the dominant type I light chain of the distantly related horned shark (Heterodontus franschesci) (Rast et al., 1994) dso occur. Although Heterodontus and Cnrchnrhinus had an ancestral divergence approximately 150 million years ago, the constant regions of their type I light chains are 87% identical. The dominant sandbar shark C h shows only 43% identity to the autologous type I C regions, which is the same level of sequence identity it shows to the constant region of the pig A light chain. The dominant sandbar shark Ch, however, can be considered orthologous to that of the even more distantly related ratfish, as 73% identity is shown with the constant regions of the light chain of that species. These quantitative relationships are illustrated in Table I. Data indicate that the genes specifying the major C h of the sandbar shark emerged in evolution prior to the ancestral divergence between the sandbar shark and the ratfish, which occurred approximately 300 million years ago, and that the emergence of the Heterodontus type I chain preceded the divergence of the Heterodontris and the sandbar shark approximately 150 million years ago. Otherwise, the degree of relatedness among these h-like chains is approximately 40-50%, a finding that holds generally for comparison of immunoglobulin doniains among vertebrates. Results of similar magnitude occur in parallel analysis of mammalian K and h light chains. The human C h domain is only 34% identical to the autologous C K domain. However, human CA is 76% identical to pig CA, 71% identical to rabbit CA, and 70% identical to niouse Ch. Possibly, the
434
JOHN J , MARCHALONIS st a/
TABLE I PERCENTAGE IDENTITIES AMONG CONSTANT DOMAINS OF ISOTYPES OF A-IJKE LICRTCHAINS OF CI-~ONDRICHTHITES
Sandbar CA Hydrolagus c-LII Pig CA Heterodontus c-LI Sandbar CA (LI) Sandbar CAX
Sandbar CA (LI)
Hydrolagus c-LII
Pig CA
42
45
-
45
52
44
-
43
52
44
87
-
43
53
41
45
45
Sandbar CA
Heterodontus C1,I
Sandbar CAx
-
73
-
rate of evolution of K chains within the marnmds was greater than that of A chains as human C K is 59% identical to murine C K and 45% identical to that of the rabbit. In any case, the parallel argument can be made that the respective genes emerged prior to the particular speciations involved. The predominant serum immunoglobulin isotype of sharks is homologous to human IgM and occurs in both pentameric (19s) and monomeric (7s) forms (Celm and Small, 1967; Marchalonis and Edelman, 1965). Analysis of the structure of the genoinic genes and cDNA copies of p chains establish that this immunoglobulin is universal and enabled study of the rate of evolution of the immunoglobulin domains (see later). Litman and colleagues found that the gene segments specifylng shark p chains were homologous to those of higher species, i.e., VH, DH, JH, and CH, but that these were arranged in individual clusters (cassettes) rather than in the translocons characteristic of human and murine immunoglobulin heavy chain gene segments (Hinds and Litman, 1986; Kokubu et al., 1988a). Although the gene segments showed fusion or rearrangement in the germline in approximately half of the heavy chain clusters studied, others had normal recombination signal sequences with the segments separated by a few hundred bases. Thus, somatic recombination must occur and the possibility of junctional diversification was present. The gene segments specifylng shark light chains, i.e., V,, JL, and CL,were likewise organized in individual clusters (Hohman et al., 1993; Schluter et al., 1989b; Shamblott and Litman, 198913). Hohinan et al. (1992, 1993) made the surprising fincling that all of the clusters specifylng sandbar shark light chains had the VL and J L segments fused in register in the germline.
E M E R G E N C E A N D EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
435
Subsequently, it was shown that all of the sandbar shark type light chains (type I1 in Litman’s nomenclature) are fused in register in the distinct species where they occur (Rast et al., 1994) and that the type I light chains are all unfused in Heterodontus, but may be fused in other species (Rast et al., 1994). Based on the statistical “sampling theory,” the number of individual A clusters in the sandbar shark is at least 200 (Hohman et al., 1993). Approximately 30 clones have been analyzed in detail, finding no two identical. Considerable variation is apparent in the complementarity determining regions, with the CDR3 being the most diverse, even though V/J rearrangement does not occur in this species. Furthermore, individual CA sequences are distinct from one another, although they are approximately 90% identical overall. Two groups have described a second (Bernstein et al., 199613; Greenberg et al., 1996) and possibly a third (Harding et al., 1990a,b; Kobayashi and Tomonaga, 1988; Kobayashi et al., 1992) type of heavy chain in sharks. Bernstein et al. (1996b) and Greenberg et al. (1996) reported the presence of a novel high molecular weight heavy chain containing one V domain and six constant domains in the sandbar shark. The molecule, termed o, is distinct in sequence from the p chain, although the first constant domain shows appreciable identity to C p l , probably indicative of constraints imposed by the necessity to associate with light chains. Furthermore, the set of VII domains used by the o chain is distinct from that used by the p chain (Bernstein et al., 199613; Shen et al., 1996). Flajnik and associates (Greenberg et al., 1996) reported an orthologous molecule from the nurse shark. In addition, these workers described a molecule which they termed NAR (new or nurse shark antigen receptor) in the nurse shark that lacks residues in the variable segment and the first constant domain necessary for the formation of heterodimeric pairs with light chains. They suggested that this is a novel type of antigen receptor because it shows a pattern of sequence diversity distinct from that of V,, associated with either the p or the o chains which occur in association with light chains. Identification of genes specifying TCR chains in Heterodontus (Rast and Litman, 1994) and in the more anciently arisen clearnose skate (Rast et al., 1997) disclosed the exciting result that cdp and y/S molecules were present in the most primitive extant gnathostomes. The finding of T-cell receptors was expected, particularly since the notion has long been held that T-cell-mediated type immunity most probably preceded antibody formation (Hein, 1994; Hildemann, 1974). What was unexpected was the extensive repertoire present (Hawke et a/., 1996; Rast et al., 1997). Studies with more advanced gnathostomes, including teleosts (Partula et al., 1995, 1996) and amphibians (Fellah et al., 1993a), establish that considerable diversity occurred early in the evolution of jawed vertebrates
436
JOHN J. MARCHALONIS et nl
within the T-cell receptor compartment, and that this degree of diversity even surpassed that of antibodies. In addition to the molecular biological studies, cellular immunological investigation in teleosts (Michard-Vanhee et al., 1994; Miller et nl., 1994; Ruben et al., 1977; Stuge et nl., 1997) and in amphibians, most notably in Xenopus (Du Pasquier, 1993; Du Pasquier et al., 1989), has established the presence of T and B type lymphocytes and their modes of differentiation and function. In the case of cartilaginous fishes, however, the recombinant DNA studies have vastly outstripped the biological investigations. At this point in time, cells that can be definitively identified as T and B have not been isolated or cultured from elasmobranchs. Although the suggestion has been made that somatic mutation contributes to diversification of shark immunoglobulins (Hinds-Frey et al., 1993), this has to be definitively shown using cloned cells.
B. IMMUNOGLOBULINS OF BONYFISHES The predominant class of Igs in bony fishes is IgM, but the multiineric form found in serum occurs as a tetramer of the form ( L 2 ~ 2and ) ~ the monomeric form does not usually occur (Bernstein et al., 1997; Marchalonis, 1977; Wilson and Warr, 1992).A second isotype has been discovered in the catfish (Wilson et al., 1997) that has a heavy chain containing one V domain and eight constant domains. This molecule most probably serves as a receptor on some B cells. Although it shows some similarity, particulady in large heavy chain size, to the IgW of sharks, this Ig has been proposed to be a precursor of IgD which serves as a second Ig receptor with IgM on B cells of humans and mice. The light chains of bony fish tend to resemble K chains overall (Daggfeldt et al., 1993) with a set occurring in the catfish showing sufficient homology to K chains to give confidence to this identification (Ghaffari and Lobb, 1997). Gene segments of light chains of teleost fish, including codfish, trout (Daggfeldt et al., 1993), and catfish (Ghaffari and Lobb, 1993, 1997), occur in cluster organizations, but the VH, DII, JH, and C p segments occur in translocon arrangements (Ghaffari and Lobb, 1992; Hayman et al., 1993) comparable to those of mammals and amphibians. In contrast, the VL, JL, and CLsegments of the sturgeon, a chondrostean, are arranged in an organization similar to that of K chains of mammals rather than the clusters characteristic of teleosts (Lundqvist et al., 1996). In phylogenetic analyses, the sturgeon VLdomains show homology to those of teleosts, but the CLdomains cluster with those of sharks. cDNA clones specifying TCRa (Partula et nl., 1996) and p chains (Partula et al., 1995) have been identified and characterized in the rainbow trout. The group at the University of Mississippi Medical Center (Zhou et al., 1997) has also reported the characterization of full-length cDNA sequences for channel
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
437
catfish TCRa and P chains. Although TCRy and 6 chains are expected because of their occurrence in their more primitive chondrichthytes, these have not yet been reported for teleosts. Bony fishes, including the goldfish (Ruben et al., 1977),exhibit collaboration between T and B cells in the generation of antibodies to haptens, and these lymphocytes have been directly identified and cloned in the catfish (Miller et al., 1994; Stuge et al., 1997).
c. IMMUNOGLOBULINS OF AMPHIBIANS Early studies showed that anuran amphibians such as the bull frog, Rana catesbeiam (Marchalonis and Edelman, 1966a; Uhr, Finkelstein and Franklin, 1962),and the Queensland cane toad, Bufu innrinus (Atwell and Marchalonis, 1975; Marchalonis and Edelman, 1966b; Marchalonis and Germain, 1980), generated both 19s (IgM) and 7s (IgY) antibodies to a variety of antigens, including bacteriophage Salmonella flagella and proteins. Both isotypes were produced to thymus-dependent antigens and specific tolerance could also be induced (Marchalonis and Germain, 1971). In contrast, thymus-independent flagellar antigens induced only IgM antibodies and were not tolergenic. Extensive studies have been carried out using the clawed toad, Xenopus laevis, as a model for differentiation of the immune system, including antibody production and T-cell immunity (Du Pasquier et al., 1989). Xenopus has three Ig isotypes, IgM, IgG, and IgX, which are characterized by distinct heavy chains (Du Pasquier et al., 1989; Schwager et al., 1988a,b). In addition, three isotypes of light chains, p, which is equivalent to K chain (Schwager et nl., 1991), Q , which is apparently unique, and a third type homologous to mammalian h chains (Haire et al., 1996), have been characterized. cDNA clones specifjmg TCRP chains have been characterized in this species (Chretien et d., 1997). The complete sequence of the p chain (Fellah et al., 1992, 199313) and TCRP chains (Fellah et al., 1993a) has been determined for another wellstudied amphibian, the urodele Ambystoma mexicanum (axolotl).The segmental organization of Ig genes in the Xenopus resembles the translocons of mammals. Despite extensive diversity in numbers of Ig segments and expression of distinct classes followingimmunization, the antibody response of amphibians is relatively restricted (Wilson and Warr, 1992) and affinity maturation apparently does not occur. D. IMMUNOGLOBULINS OF REPTILESA N D BIRDS With the exception of chickens, which have the distinct bursa of Fabricius which generates mature B cetls (McCormack et al., 1991; Szenberg and Warner, 1962) and a readily accessible thymus, reptiles and birds have
438
JOHN J. MARCHALONIS et d.
been relatively neglected in molecular and cellular studies of the immune response. An ancient reptile, the tuatara Sphenodon punctatum, possesses organized lymphoid tissue and has three distinguishable types of Igs, most probably resembling IgM, IgY, and IgN (Marchalonis et al., 1969).Turtles could be readily immunized with the major antibodies being IgM and a low molecular weight species characterized by a sedimentation coefficient of 5.7s (IgN) smaller than IgG or IgY (Grey, 1969). Turchin and Hsu (1996) analyzed VH diversity in the turtle Chysemys scripta and partially characterized the C p sequence as well as isolating putative transcripts of the 7s IgY and 5.7s IgX heavy chains. VH genes from the Caiman were characterized by Litman’s group (1985b). The arrangement of turtle heavy chain gene segments is comparable to that of mammals with one C p and many VH segments. TCR have not yet been characterized in reptiles. Chickens have IgM, IgY, and IgA immunoglobulins (McCormack et al., 19911, but lack IgD molecules. Molecular analysis of light chains showed that there was one functional VA, one JA, and one CA in association with 25 VA pseudogenes and that diversity was generated by templated hypermutation (gene conversion) (Reynaud et al., 1987). A parallel situation occurs for heavy chain gene segments where only one functional VHoccurs (Reynaud et at., 1989). T and B cells (Cooper et al., 1991) and the MHC of chickens have been studied extensively with TCRa and P genes identified (Gobel et al., 1994; Tjoekler et al., 1990). The major antibody isotype in the duck is the 5.7s IgN (or IgX), which has a short heavy chain consisting of only two constant domains and is ineffective in the removal of pathogens (Magor et al., 1994b).
E. IMMUNOGLOBULINS OF MAMMALS Although mammals possess the anciently emerged IgM isotype and the full range of TCRs, major developments occurred within the evolution of this vertebrate class. Representatives of all mammals, including monotremes such as the echidna (Atwell and Marchalonis, 1975; Atwell et al., 1973), marsupials such as the brush-tailed possum (Marchalonis and Atwell, 1972), and all eutherians studied, have IgG as their major serum immunoglobulin. Because IgG molecules are not found in species more primitive than mammals, it has been suggested that the genes specifying the mammalian type y chain arose in the earliest ancestral mammals following divergence from reptilian stock. Furthermore, eutherians such as rodents and primates show affinity maturation correlated with somatic mutation and antigenic selection (Liu et al., 1992)occurring in B cells in germinal centers of lymph nodes where RAG-1 and RAG-2 genes are expressed in secondary
E M E R G E N C E AND EVOLUTION OF T H E IMMUNOC,LOBULIN FAMILY
439
differentiation (Han et al., 1996). This dramatic event can be considered a “second big bang” in the evolution of immunoglobulins. Despite the clear respective homology among orthologous Ig chains and TCR chains of mammals, there is a inarked plasticity in gene segmental organization and in mechanisms for the generation of diversity. Notably, humans and mice have essentially similar organizations of Ig (Lai, Wilson and Hood, 1989) and TCRa and /3 genes (Hood et al., 1995), but rabbits preferentially use one VEIgene out of inany in most V( D)J rearrangements (Knight and Winstead, 1997) and depend on somatic gene conversion for diversification. Moreover, rabbits have only one y heavy chain gene but carry 13 genes for different (Y chains (Burnett et al., 1989; Honjo et al., 1989) in their heavy chain translocon. Sheep generate antibody diversity by a process of somatic hypermutation that occurs in the ileal Peyer’s patches (Reynaud et al., 1991),which serve as an equivalent of the chicken’s bursa of Fabricius. Somatic gene conversion also plays a major role in the generation of the antibody repertoire in swine and cattle where the Ig heavy chain locus contains only a single J H and no IgD (Butler et nE., 1996). The finding that IgD occurs only in primates and rodents raises doubts that the putative IgD hoinolog of teleost is directly ancestral to the mammalian molecule. Although Wilson et al. (1997) have proposed a long-range lineal descent of IgD within the vertebrates, the possibility also exists that genes specifjmg non-p heavy chains arose independently in the emergence of separate vertebrate classes (Atwell and Marchalonis, 1975,1976).Thus, even though the molecules of the combinatorial immune system of gnathostomes show remarkable conservation, substantial variation and organization of gene segments and mechanisms for diversification can occur even within a single class of vertebrates, which is particularly noticeable in mammals. VI. Framework 4 of the Variable Domain Encoded by the Joining Segment
The V and C domains had a common ancestor (Hill et nl., 1966) but diverged from one another prior to the appearance of light chains, heavy chains, or TCR. All Igs incorporate a peptide segment encoded by a joining segment gene (J),which can vary in length from 13to 17 residues (Toyonaga and Mak, 1987). This set of related peptides shows considerable diversity in its N-terminal portion, which contributes to the C-terminal region of CDR3. However, the C-terminal 11 residues of the J segment, which define the fourth framework (Kabat et al., 1991), are strongly conserved and distinguished between heavy chain and light chain/TCR sets (Chothia et nl., 1988; Davis and Bjorkman, 1988; Schiffer et al., 1986). Figure 7 demonstrates that this distinction was an ancient one with the FR4
440
JOHN J. MARCMALONIS et nl
Heavv Chains Mouse J,1 Human IgM Chicken IgM Turtle IgM Xenopus IgM Trout IgM Catfish IgM Sb-Sh IgM Sb-Sh IgW Liaht Chainsm-cell receotors Hu A Chicken A Xenopus A Sb-Sh A Catfish L Trout L Hu K Xenopus K Nurse Sh K Hu TCRP
Axolotl TCRP Trout TCRP Skate TCRP Hu TCRa Trout TCRa Skate TCRa Hu TCRy Skate TCRy Hu TCRd Skate TCRB
F F F F F F F F F F F
G G G G G G G G G G G
F F F F F
G G G G G
T A G T G S G G K S E A K S Q G S S K
G G G G G G G G G G G G G G G G
T T T T T T T T T T T T T T T T
K T Q K K R K K K R K K K R K K T R G T R
V L L L L L V L L L L L L L L L L L V
T T T N I L E I R T T T V S
V L G V L G V L T - L G V D L D V G I - & V - & L S E V V V V V V V L I R N S D V V T I V T T V E
FIG. 7 . Framework 4 sequences of vertebrate immunoglobulins and T-cell receptors. Hu, human; Sb-Sh, sandbar shark.
EMEHGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
441
segments of heavy chains, including the w chain, having the motif WGxGT (usuallyuncharged)V, and light chains and TCRs usually having the characteristic pattern FGxGT(RorK)L. The TCRy chain and the chicken J h are exceptions because residue No. 6 can be T or L, which is uncharged like that of the JF,, even though residue No. 1 is an F consistent with the light chain/TCR set. The characteristic patterns, however, were established in evolution by the phylogenetic level of chondrichthytes and continued throughout gnathostoine evolution. Although the nucleotide codons specifying the amino acid residues in this region were not particularly useful in the design of gene probes, the characteristic peptide sequence proved valuable as antigens for the production of antibodies detecting Igs and TCRs across species (Marchalonis et nl., 1988a,b, 1993; Schluter et al., 1987, 1989,). Antibodies produced to the JH1 sequence detected Ig heavy chains in mammals and in sharks (Rosenshein et al., 1985), as would be expected from the sequence homology. Antibodies produced against the human JPl.2 sequence (ANYGYTFGSGTRLTW) reacted with TCRP chains of humans, mice (Schluter and Marchalonis, 1986), and the sandbar shark (Marchalonis et al., 1997) and with light chains of various species, including humans, mice, and sharks (Schluter et al., 1987). Interestingly, light chains of chickens and turkeys that have a T at residue position No. 6 do not react strongly with rabbit antibody directed against the JP peptide. It proved possible to make antibodies in rabbits because the negatively charged residue glutainic acid (E) occurs at that position. Analysis of residues critical for antigenicity was performed using modified synthetic peptides (Marchalonis et al., 1988b) with the conclusion that this residue was critical for binding to the antibody and that the large, positively charged amino acids lysine (K) and arginine ( R ) were equivalent. Although the J segment is not homologous to the classical Ig domains, it is the most conserved or characteristic motif of Igs in evolution (Marchalonis et al., 1988a,b). The o r i p of the J segment is unknown, but minigenes specifying it must have become an integral part of the V-C gene cluster at or immediately following the primordial divergence of V and C gene segments. The following section proposes that this segment with associated recombination signal sequences was inserted into the genome of ancestral gnathostomes, or their precursors, by retroviral transfection. VII. Evolutionary Comparisons of T-cell Receptors
The vast preponderance of studies on TCR has been carried out on humans and laboratory rodents (Bell et al., 1995; Kronenberg et al., 1986; Marchalonis et al., 1997; Toyonaga and Mak, 1987). TCRs are associated exclusively with the surface of T cells. Their polypeptide chains are two
442
JOHN
1. MAHCHALONIS
et nl.
domain Igs comparable to light chains that consist of one V domain, a CDR3 formed by either incorporating J segments alone (TCRa and y ) or a combination of diversity (D) segments and J segments (TCRP and 6). The constant region of these molecules is composed of one immunoglobulin constant domain, a hinge segment of varying length, a transmembrane (TM) segment, and a short C-terminal segment extending into the cytoplasm (CYT). The V domains of TCRa, P, y , and S chains are clearly homologous to those of serum Igs, most particularly light chains, and contain recognizable framework and CDR segments, with the CDR3 segment showing the most diversity (Davis and Bjorkman, 1988; Schiffer et al., 1986). The FR3 segments of VP domains, like those of VH structures (Capra and Kehoe, 1974), contain a fourth hypervariable region involved in the recognition of microbial superantigens (Domiati-Saad et al., 1996; Silverman, 1992). This segment can also be considered a public idiotype inasmuch as it is encoded completely within the VP segment. Because of the recent application of recombinant DNA technology, sequence and partial gene organization data are currently available for a, 0, y , and S chains of artiodactyl mammals (Hein, 1994), as well as for rodents and primates (Arden et al., 1995a,b), a and P chains of carnivores (Hein, 1994), a, 0, and y chains of chickens (Gobel et al., 1994; Rast et al., 1995; Tjoekler et al., 1990), a and chains of two amphibians, the axolotl (Fellah et al., 1993a) and the clawed toad (Chretien et al., 1997), the a chain of the puffer fish (Rast et al., 1995),a and 6 chains of the rainbow trout (Partula et al., 1995,1996),p and 6 chains of the horned shark (Hawke et al., 1996; Rast et al., 1995; Rast and Litman, 1994), and a, P, y , and S chains of another elasmobranch, the clearnose skate (Rast et al., 1997). The TCR genes and families of humans and mice have been compared in great detail, with the conclusion that there is strong conservation in gene sequence and organization between a and P TCRs, respectively, of these two species (Hood et al., 1995), which diverged greater than 65 million years ago. The number of individual germline Vp and V a genes is large, but it has been possible to identify lineal relatives or orthologs in the two species (Clark et al., 1995; Marchalonis et al., 1994); e.g., murine Vp8 corresponds to human VP3 and murine Vpll is the ortholog of human VPS. A phylogenetic analysis of V domains of a, P, y, and 6 chains of vertebrate species ranging from the skate to humans is illustrated in Fig. 8. Marchalonis et al. (1996) have shown that TCR V segments tend to cluster more closely with those of immunoglobulin light chains than with the heavy chain cluster, which forms a separate grouping. The phylogenetic studies are consistent with the conclusion that these three groups separated early from one another in the emergence of vertebrates and were clearly distinct in gnathostome vertebrates, as were the joining segments and the
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
443
FIG.8. Unrooted tree of representative TCR V dorndns of diverse vertebrate species. The dendogram was calculated using the program CLUSTALW (Thompson et al., 1994) and displayed using the program TreeView (Page, 199G). The clusters are predominantly Vy ( I ) , Vp (11), and V d 6 (III), hut mixed subsets are designated IA, IIA, IIB, and IIIB. Human sequences are from Arden et al. (1995a) as follows: (1)Vp; 1.1 (BVISIAINI), 3.1 (BV351), 6.5 (BVGSASN2T), 20 (BV20SIAST); (2) Va; 1.2 (AV152AINIT),4.1 (AV451), 14, (ADV14Sl), 21 (ADV21SEAIN): (.3) VS; 1.9 (GVlS2A2T). 10 (GV2SIP). GenBank accession numbers for sequences not cited in text are chicken Vp, M81149; catfish, U62043, U58508; and axolotl, L33787, L33270, L33337, L332G9.
constant domains. The tree shown in Fig. 8 was constructed using the computer program CLUSTALW (Thompson et al., 1994), which uses the neighbor joining method (Saitoh and Nei, 1987) and DRAWTREE (Felsenstein, 1993). Three major clusters (1-111) with subsets showing overlap are disclosed. These correspond to the TCR Vy set ( I ) , the TCR Vp set (II), and the TCR V d S group (111). In addition, some of the designated V domains do not cluster on the basis of their nominal identification; e.g., IA contains two V a sequences, IIA contains a human VS in association with a shark
444
JOHN J. MARCHALONIS et
(il.
Vp7 sequence, and two human Vp sequences cluster with the Vd/6group in cluster 111. Thus, there is considerable diversity in the TCR V region gene sequence, and in individual cases it may be hard to define the gene product as belonging to a particular group. A more general study (S. F. Schluter and J. J. Marchalonis, unpublished) was carried out analyzing variable domains of all families of immunoglobulins and TCRs with the conclusion that there is some intermingling of classical Ig V sequences and those of TCRs. The most notable example was the finding that some TCR Vys associate with the iminunoglobulin V13clusters. It is possible that V domain orthologs have been conserved throughout vertebrate evolution, particularly if there was a strong selective pressure for their maintenance through infection with common types of bacteria or viruses, for example. Some interesting associations in Fig. 8 are consistent with this hypothesis. For example, in group I, there is an association between human Vy4 and skate Vy5, with another between chicken Vy and skate Vy2. In group 111, there is a clustering of Va domains of the skate, human, catfish, and trout. In cluster 11, there is apparently a close association between axolotl TCRVB5 and human Vp20.These sequences are presented in aligned form in Fig. 9. Because data were obtained by sequencing cDNA clones, it is not certain where the N termini of the molecules are. However, the universally conserved cystein (C) should lie at position No. 22 or No. 23. In addition, the beginning of the second framework is clearly marked by the WYXQ sequence, and the conserved leucine in framework 3 is readily apparent, as are the absolutely conserved tyrosine ( Y ) and cysteine (C)at the carboxyl end of the third framework. As illustrated in Figs. 8 and 9,TCR p genes show considerable V family multiplicity, and it has been suggested that they utilize combinatorial mechanisms in the generation of immunologic diversity in elasmobranchs (Hawke et al., 1996).The possibility of orthologous relationships between individual V/3 gene families of elasmobranches and higher vertebrates is supported by studies of the clearnose skate (Rast et al., 1997)where it was found that the highest identity scorings for one skate family were with the horned shark V/3, as would be expected. However, for five other families, the next most related sequences were mammalian V/3 segments. In this case, the overall amino acid identities were in the range of 36-44%. Investigation of individual Vp families of the axolotl are consistent with this suggestion (Fellah et al., 1993a)because one of the Vp families in this species showed 56% identity in the amino acid sequence with the murine Vp14 and human VplS families. A V/3 family from the rainbow trout showed sequence similarity to the human Vb20 family (Partula et
EMERGENCE AND EVOLUTION OF THE IMMUNOGLORUIJN FAMILY
445
Ax0 BV5 HuVPPO skate Vp3 shark Vp3 HuVal.2 skate Va3 shark V61 HuVa.21 HuV63 skate V65 skate Va2 HuVp3.1 HuVa4.1 chick V a skate V@ chick Vy HuVy4 CDR2
FR3
BV5 HuVp20 skate Vp3 shark Vp3 HuVal.2 skate Va3 shark V61 HuVa21 HuV63 skate V65 skate Va2 HuVp3.1 HuVa4.i chick Va skate V@ chick Vy HuVy4 Ax0
F'IG. 9. Comparative alignments of representative TCK V segments seIected froin Fig . 8. Highly conserved residues are shaded or boxed.
al., 1995), thereby indicating that TCR V region genes arose early in evolution and were maintained over 400 million years of evolutionary time. The degree of complexity in these systems i s also noteworthy. TCRa chains of an elasmobranch, the clearnose skate, and a teleost, the rainbow trout, contain, respectively, four Va! families, six Jasegments, and two C a domains (Rast et al., 1997) and six V a families, 32 J a segments, and one C a domain (Partula et al., 1996). The possession of a large number of Jagene segments is a characteristic of inainmalian a chain genes, which is apparently a general property of this TCR. With respect to the TCR /3 chains, the clearnose skate (Rast et al., 1997) has six VP families, four J/3 segments, and one DP minigene. The horned shark (Hawke et al., 1996) contains seven VP families, some of which are characterized as subgroup
446
J O H N J. MARCHALONIS et nl.
I or I1 as defined by Kabat et al. (1991),and there are two distinguishable subgroups of CP. The organization of TCR gene segments in elasmobranchs is apparently in cluster arrangement (Litman et nl., 1993), but this is not identical to the kind shown for light chains and heavy chains because more than one V can be associated with a cluster. The TCRy and S chains of the clearnose skate also show considerable diversity with 5 Vy and 2 Jy segments and 5 V6 families, 2 JS, and 2 or 3 DS minigenes (Rast et al., 1997).TCRP genes of the clawed toad, Xenopus, likewise show considerable diversity with 10 VP families, 2 D segments, 10 J segments, and a single constant domain having been defined so far (Chothia et al., 1992; Chretien et al.,1997). TCR occur as heterodiiners of the form alp or ylS that are stably associated with the plasma membrane of T cells via transmembrane segments that are part of the C-terminal sequence of each chain. These are specified by exons lying within the respective TCR gene cluster. Immunoglobulin light chains have never been found to contain transmembrane segments, but each type of heavy chain can occur in association with light chains as part of a B-cell surface membrane receptor, with IgM and IgD being the receptors on virgin B cells, and the other isotypes functioning in that role on memory cells. The association of gene segments specifylng transmembrane and cytoplasmic domains with TCR and immunoglobulin heavy chains was an ancient evolutionary development, which is found widespread in elasmobranchs as well as in higher vertebrates. Figure 10 aligns the sequences of transmembrane and cytoplasmic domains of widely divergent vertebrates, comparing those of immunoglobulin chains, TCRP, TCRy, TCRa, and TCRS chains. Each group is a recognizable family in evolution on its own, although clear-cut similarities are observed in comparisons among the different groups. For example, the leucine (L) at position No. 5, with the single exception of the chicken Ca, is universally conserved in all transmembrane domains for all receptors. Another widely conserved residue is the tyrosine (Y) at position No. 12, which is absolutely conserved between immunoglobulin and TCRP constant domains. It is also shared by the TCRys. The lysine (K) at position No. 8 clearly distinguishes the TCR domains from the immunoglobulins because this positively charged residue is universally conserved for the TCR, whereas this position is occupied by the equivalent leucine (L)or isoleucine (I) in immunoglobulins. An individual difference characteristic of the C a set is the uniform occurrence of phenylanine ( F )and asparagine (N) at positions No. 12 and No. 13. The asparagine (N) is also present in the 6 as would be expected, but the phenylanine is present only in the shark example. Because of both the overall conservation of sequence in evolution and the particular association within families, it is reasonable to assume that
-------------------------TM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CYTO-
-__I_____
Mouse Cp Mouse Cyl CatfishCbB Shark Cp
T F V S L F L L S L P Y S T T V T L - - - - - - F K V K
Human Cp Chicken Cp Shark Cp Ax01 cp
L V T A
Y Y Y Y
E I L L
I M I L
L L L L
L I I V
G F C S
Y H q K
Mouse Cy Human Cy Pig Cy Skate Cy
T M T T
Y Y Y Y
L L L T
L L L I
L L L L
L L L L
L L L M
X G V I Y L A I T S F S L L R - - - - R T S V C G N E K T S Y G V V Y F A I I T C C L L R - - - - R T A F C C N G E K S K G V A L F A I I T Y C L F R - - - - E T T V C S A W K R S ~ S O L Y C G I I S A I L Y K - - - - - - M E Y W G T K K P L
Mouse Ca Human C a Chicken C a Pufferfish C a Skate Ca
G G G G G
L F L L M
R R K R R
I I I V M
L L I L L
L L F L F
b L M f V
Human C8 Sheep Cb Shark Cb
G L R M L F A K T V A V g F L L T A - - - - - - - - K L F F L G L R M L F A Z S V A V E F L F T A - - - - - - - - K L F F F G L R V L F F Y S V A U V M M T A - - - - - - - - R V L L F
I F I S L F t L S V C Y S A A V T L - - - - - - - F K V K A F I I L F V I S L L Y G G S V T L - - - - - - - V K V K I A T F I I L F F L S I F Y S A A V T L - - - - - - - - K V K
Y K & K K
A S S S
V V A T G
T I I A
L L F A
A A v V V
Y Y Y Y
G G I V I
A G T G
D D ~ u U
V I I L
L L V I I
L F F F
L L L L L
V V I V
M M I M M
S A L V L M A M - - K R K D S R G M G M M W Y - - - - K K M Y S T I A W K T - - - K T S Y T I S M C - - - - - R V K L
T T T T S
L L M L V
L -
-
-
-
- - - - R L W S S - - - - R L W S S - - - - - M W K K N Q - - - - R L W I S Q - - - - R V W T S
FIG.10. Comparative alignments of transmembrane and cytoplasmic domains of in~munoglobulinheavy chains and TCR chains. Shaded and bold-faced residues are shared among individual groups. Underlined residues are conserved within the Ca! and C6 groups.
448
JOHN J. MARCHALONIS et al
the patterns shown here reflect those of the original appearance of the transmembrane segments in the emergence of combinatorial immunoglobulins. These diverged from a common ancestor and were maintained through 450 million years of vertebrate evolution. This process parallels that observed for the FR4 sequence of the joining segments, which likewise must have differentiated early and were maintained throughout the evolution of gnathostomes. Figures 3 and 4 show the phyIogenetic clustering of immunoglobulin, TCR, and MHC domains and illustrate in aligned sequence form the exact relationships between examples of these taken from humans and elasmobranchs to represent the most divergent poles of living gnathostomes. Because TCR are membrane associated, comparative alignments of complete C a and Cy domains are provided in Figs. 11 and 12. The immunoglobulin domain of C a is more compact than that of the others and shows some peculiarities in three-dimensional structure (Bentley et al., 1995; Garcia et al., 1996), particularly with respect to the p chain. It is also worthwhile to consider the TCRy chain because the yS TCR has been proposed to represent the primordial T-cell receptor which represents the earliest separation between specific cell-mediated and humoral immunity (Hein, 1994). The TCRdfi, which shows obligatory dependence on the MHC to present peptides, may have emerged later, although these events would still have occurred within the explosive time frame of the separation of higher deuterostomes, possibly including ancestral agnathans, from ancestral gnathostomes. Figure 11 shows the comparative alignment of the C a domains of two mammals, humans and mice, and a bird, the chicken, and an elasmobranch, the clearnose skate. The C a domains show a relative lack of conservation by comparison with light chain and heavy chain immunoglobulin domains because the human and mouse IgC sequences, not counting the transmembrane and cytoplasmic segments, are only 60% identical. However, the transmembrane segments of these two species are 92% identical. The percentage of identity between the human and the skate is only 19% in the immunoglobulin domain and that between the chicken and the skate is 21%. Comparisons reflecting the same evolutionary time span for Cp4, Ch, and CKwould be in the range of 4 5 4 0 % . It is also noteworthy that even though these domains have a relatively conserved stretch of CLFTD around the first cysteine forming the intrachain disulfide bond, the C a domains lack the conserved tryptophan (W), which otherwise occurs 14 residues C-terminal to this cysteine (C). The TCRy chain has the tryptophan (W) in the correct relative position to the cysteine (C).This region of immunoglobulin structure that connects the cysteine (C) lying in p band 4-2 (B)
Human Mouse Chicken Skate Human Mouse Chicken Skate
Y I T D K T V L D M R S M D F K S N S A V A W S N - - - - K S D F A C A N A F N N S - I T P E D T F I T D K T V L D M K A M D S K S N G A I A W S N - - - - Q T S F T C Q D I F K - - - - - E T N A E T V V E - V A T S E N K H E A S X L - S T W A K - - - - K D E M Q C G A K H E G F G I L K G D D
Human Mouse Chicken Skate
F F P S P ( E S S C D V L K V E K S F E T D T N L N F Q N L ) S V I G F R I L L L K V A G F N L L M T T Y P S S D V P C D A T L T E K S F E T D M N L N F Q N L S V M G L R I L L L K V A G F N L L M T
Human Mouse Chicken Skate
L R L W S S L R L W S S
S K T S S N N S A Q V S I K D R S X S L L S F I N G S T P Q S E I T C E L E P N - - - - T ~ N I A
“Hinge”
--------------TM-------------------
P E A G A S T V C ~ T G M S L L - F K T D E N L N M L T F S Q L G L K I I E M K A V I F N V L I T E T D V G Q M S C I P L E E T D - - E D G E Y M R G I S L T V F G M R M L ~ V ~ G V I F N I L M S
+ - - CYT- - M L M W K K N Q V R V W T S
FIG.11. Comparative alignments of TCR C a domains.
* *
0
0
0
Hucyl SheepCyl Chicken Cy Skate Cy
D K Q L D A D V S P K - P T I F L P S I A E T K L Q K A G T Y L C L L E K F F P D V I K I H W Q
Hucyl SheepCyl ChickenCy Skate Cy
- E K K - E K u D E A N - - I D
Hucyl Sheep C y l Chicken Cy Skate Cy
R K E N N K N G - - - - v D Q E I I F ~ ~ E K T - - - - - - - - - - - - - - - - - - - - - - - -
N R N L A T D L S P K - P I I F L P S I A E I N U K T G T Y L C L L E K F F P D I I K V Y W K D K G S S - - A P E N Y E - - I M Q E E H E K Q V V Y Y C L I E K Q V V Y E C L I E K S Y P E V I R V K W T
D E S K A P - T I R L F P P L T A E L E K S S K S F A K C L M T D F F P D V I E V Q W K 0
S N T I L G S - Q E G N T M A - - T N D T Y M N R A L P S - Q Q G N T M N - - T z D T Y M - KE V T Q W 6 S P g E D K Y S V K G D G K E Q K F ~ Q T D S g T & M - ~ N T
K K V Y
.
F S W L T V L S W L L V S T W L S V ~ L I S R
0
*
P - E K S L D K E H R C I V Z - E N S M D K E H I C V V P S K N K I Y S C T Y L V V ~ K ~ W E S K P I S C € I A
0
0 .
Q H E R N I R G - - - - I N Q E I L F P P S N E V Y S S I V g T T E S P S D C L N Q E S K V T G
E H E S G G N S & - - S T Q P L F P K P S g E R S P Q F Q D C G Z Q S G N S - - - - - - - - - K H E I D S P E L T I T K D S K Y V g P N T G N P Y G P T C E P S V n E L Q - - - - - - - - - -------_ _ _ - _ _ _ _ _ _ _ _ - D V I T M D P K D N C S K D A N D T L L L Q L T N T S A Y Y M Y L L L
Hucyl Sheep C y l Chicken Cy Skate Cv
T G S K K A C L K D E S E V T A D N N S T K V C L E D E S N T L Q L Q L M N T S A Y Y Z Y L L L _ _ - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ T V V N R D H M T H K A A Q L V Y V V L - - - - - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ E N S E E E I P V S S L Q V A T L T Y T I L
Hucyl SheepCyl ChickenCy SkateCy
L L L L L L L M
---TRANS MEMBRANE K K K K
S S S S
V V S V
V V M L
Y Y Y Y
F F Y C
A I V G
-------------+
I I I I
I I V I
T C T S L F S A
-----------CYT- _ - - - - .
C L L R - - - R T A C V E R - - - R A A F F FK Y R T R T A I L Y X - - - - M E
F M A Y
C C N G E K S S C D D Q R S K P S G 5 K T
W G T K E P L
FIG.12. Comparative alignments of complete constant domains of TCR y chains.
EMEHGENCE A N D EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
451
with the tryptophan (W) located in P band 3.1 (C) is of interest because of its strong conservation of position in constant domains and also its recognizable homology to the corresponding seginents of variable domains (Bernstein et al., 199th; Marchalonis et nl., 1996). Data of Fig. 12 indicate that the overall evolution of Cy, like that of C a , is rapid. The human and sheep immunoglobulin domains are 65% identical, whereas the comparisons with chicken and skate yield 26 and 24%, respectively. An alignment of the immunoglobulin domains of the constant regions of TCRP chains of vertebrate species ranging from elasmobranchs to humans is illustrated in Fig. 13. There are 17 invariant residues in this alignment, including the cysteines forming the intrachain &sulfide bond and the tryptophan that protects this internal linkage. TCRP chains have the sequence KATLVCL around the first cysteine, which is a motif shared with A light chains (Schluter et al., 1989b). Another motif similar to that of A light chains occurs near the end of the intrachain loop, with the sequence SSYL in A light chains and SSRL in TCRP chains. The human CPl and CP2 domains are virtually identical; the only lfference is a substitution for Y instead of F in the conseived motif F(F or Y)PD in the segment between the first intracliain cysteine and the sentinel tryptophan. Based on this alignment, the human C P l and the mouse C P l are 73% identical. The human sequence is 36% identical to the chicken, 37% to the axolotl, 29% to the trout, 30% to the horned shark, and 23% to the skate. If the identification of TCRs of lower species is correct with the CP domains of these representative gnathostomes being orthologs, it would appear that the rate of diversification in evolution of the C p domains is more rapid than that of the CA or C K molecules. VIII. Evolution of Light Chains
Andysis of light chains of shark iininunoglobulins by isoelectric focusing (Marchalonis et al., 1988c) established that the population of these molecules within a species showed considerable diversity and that amino acid sequence analysis (Schluter et ctl., 1989a) indicated an homology to K and A light chains (Marchalonis et al., 1988c; Schluter et al., 1989a; Sledge et al., 1974). Because B cells or plasma cells of sharks have not yet been cloned and these animals apparently do not suffer from plasma cell malignancies such as multiple myeloma, which generate monoclonal immunoglobulins, it was necessary to isolate cDNA clones specifying individual light chains in order to determine the relationships between these molecules and those of mammals. Chondrichtymn fishes were shown to have a diverse set of light chains, some of which showed recognizable homology to K chains (Greenberg et al., 1993) and others to A chains (Marchalonis et al.,
HuCPl Chick Ax01 Trout HnS Skate HuCP2 MuCPl
E D L K N V F P P E V A V F E P S E A E I S H - - - T Q K A T L V C L A T G F F P D H V E L S W W G K N S E I I E P D V V I F S P S K Q E I Q E - - - K K K A T L V C L A S G F F P D H L N L V W K
E D G D E E
E P E P D D
G N N K L L
L I D F K R
S K T K N N
V V I L V V
T T R R F T
Q E P P P P
P P A P P P
S R K Q E K
V V V V V V
V K T T A S
L V V I V L
F L F L F F
D A E Q E E
P S P Q E I K K - - - K G K A T L V C L A T N F Y P D H V T L R W S P S A K K C E D R N K K K K K T L V C V A T R F Y P D H V T V F W Q P S P E E I R E - - - K K K A T V V C L V S D F D N I K I H W L P S D R E I K N - - - K G K A T V V C L I T D F Y P D N I K I R W I P S E A E I S H - - - T Q K A T L V C L A T G F Y P D H V E L S W W P S K A E I A N - - - K Q K A T L V C L A R G F F P D H V E L S W W
HuCPl Chick Ax01 Trout HnS Skate HuCP2 MuCPl
V N G K E - - - V H S G V S T D P Q P L K E Q P A L N D S R Y C L S S R L R V S A T F W Q N P R N V N G V K - - - R T E G V G T D E I S T S N - - - - - G S T Y S L T S R L R I S A Q E W F N P L N
HuCP1 Chick
H F R C Q V Q F Y G L S E N D E W T Q D R A K P V T Q I V S A E A W G R A D - C G F R F E C I A N F F K N G - - - - - - - - - - - - - T Q Q S I Q I I Y Y G P T G C D I
Ax01
T R H H
Trout HnS
Skate HuCP2 MuCPl
V N D Q V - - - T T T G V K T D D Q S P I R G - - - S D R M Y S L S S R L R L T K M D W M N P H N
V V F V V
N D D N N
F F V I F F
N V N G K D V V G K E G K E
R T E V R R
C S C I C R C E C Q C Q
E Q -
V V V V V V
K D -
- R T E D A N K D S D - V H S - V H S
G D N G G
A G T D N R A L W D - - - - K D G L Y S I T S R L R V P A N E W H K P E N T N I H T D L N A I L ~ ~ - - - E N T ~ Y ~ I ~ ~ R L R F D A L D ~ A R ~ ~ I H T D A S S Q S E D - - - E G M T F S I S S R F R L D A R D Y A K T E K V S T D P Q P L K E Q P A L N D S R Y C L S S R L R V S A T F W Q N P R N V S T D P Q A Y K E S - - - N Y - S H C L S S R L R V S A T F W H N P R N
Y F D - - - - - - - - - - - - - - - P E N I T V S R E T K G R E G C G V S F Y D G T - - - - - - - - - D N I R V N D T I S G D L Q G Q S G G E I D L Y T N E - - - - - - - - - - - S V P T T S S S T L A V K A E M C G I D H Y R N G - - - - - - - - - - - S T P Q T E Q G T H Y I K K E T C G L Q F Y G L S E N D E W T Q D R A K P V T Q I V S A E A W G R A D - C G F Q F H G L S E E D K W P E G S P K P V T Q N I S A E A W G R A D - C G I
FIG.13. Comparative alignment of TCR Cp immunoglobulin domains. Shading denotes sharing with the human Cpl sequence.
E M E R G E N C E AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
453
1993), whereas the relationship of others to “standard’ light chains of higher vertebrates was less definitive (Rast et al., 1994). Figure 14 compares the framework sequences 1 thru 3 of Vh domains of the sandbar shark, the set of six human V h sequences, a murine Vh6 (Vh26) and Vhl (J558), and a VA sequence of the clawed toad, Xenopus Zaevis. Sandbar shark C1.5 and U11 are individual members of the same V h family and are 85% identical. The third sandbar shark example here, VAx, represents a separate family, as illustrated by only 49% identity to SbSC1.5. Overall, the sandbar shark VhC1.5 shows approximately 50% identity to the remainder of the set, with the lowest being 45% identity to mouse VhJ558 and the highest being 58% identity to human VA6. The best computer matches of this sandbar shark sequence to the database have been human Vh6 sequences. Even including the CDR sequences, the overall identity remains better than 50%. Overall, there is thus a suggestion that V h genes have been conserved in vertebrate evolution from sharks to humans. The Vh6 gene is not expressed in inbred mice, but has been found in outbred wild mice (Reidl et al., 1992). In addition to the composite sequence identity, the putative Vh6 orthologs of human, mouse, and sandbar shark also show characteristic residues in FR3. The human Vh6 has conserved GSDRSSN, and the mouse has GSIDSSSN in the same region. The corresponding shark sequence is GSVDSSSN. The other human and mouse Vh sequences have two gaps in this region. The Xenopus V h does not require gaps in this region, but shows only 47% identity to the VA6 set. Figure 14 indicates highly conserved residues as well as absolutely conserved. There are 13 universally conserved residues as follows: Q at position No. 6, G at position No. 15, and C at position No. 22 in FR1, W and Y at positions No. 1 and No. 2 in FR2, and the Q at position No. 4 is substantially conserved, as is the P at positions No. 6 and No. 10 in this FR. The FR3, with one exception, begins with glycine (G) and has absolutely conserved RF and G at positions 5, 6, and 8, respectively, with the next residue, serine, also highly conserved. The leucine (L) at position No. 18 is universally conserved and there is a strong motif of the form LTIT or S followed by an absolutely conserved D at position No. 27, a Y at position No. 31, and the C that closes the interchain disulfide loop at position No. 33. There is also a highly conserved motif of the form DEADYYC that is found in human A chains and also in the sandbar shark VAX. The LTQP motif near the N terminus of the F R 1 is likewise generally apparent. A unrooted phylogenetic tree illustrating the relationships among the light chain variable region sequences of elasmobranchs and bony fishes was constructed with CLUSTALW and plotted by PHYLlP DRAWTREE programs (Fig. 15).Three major clusters were formed. One, which contains
-FR2--0
Sb~C1.5 SbSU11 SbSVAX HuVAl HUVhll HUVAlll HUVhlV HUVhV HUVhVl MuVAJ558 MuVA26 ChickVA XenOVA
0 .O
0
0
0 0
0 0
-______I
0
0
D P V L T Q P G S I S S S P G K T V T I T C(CDR1)W Y W Q K P D S A P V F V W S D I V L T Q P G S I S T S P G K T V K I T C(CDR1)W Y W Q K P D S A P V L V W D
E Q Q
g
T L N Q P
g
V L T Q P
p S g S
I V Q G G N V E I V S A A P G Q K V T I
S
S S S
* * s
E L T Q D g A V S V A L A L T Q P g S A S G S L Q D F M L T Q P H S V S E S P Q A V V T Q E S A LT. T S P Q F V L T Q P S S V S T S L - - A L T Q P S S V S A N P A V S I T Q P V S E S V K L
A P R Y L L R A P K L
0
0 0
0.0
G Q T V R I T C(CDR1)W G Q S V T I G K T V T I
S S
C(CDR1)W C(CDR1)W
G E T V T L T C(CDR1)W G S T V K L 8 C(CDR1)W G G T V K I T C(CDR1)W G E T V R I
S
P N g Q L P G T H Q Q Q P H R Y Q Q K P G Q Y Q Q K P G Q Y Q Q H A G K Y Q L R P G S Y Q E K P D H Y Q Y M G R Y Q Q - P G S Y Q Q K A G g
C(CDR1)W Y Q Q K C (CDR1) W Y Q
C(CDR1)W A L T Q P A S V S G S P G Q S I T I F D L T Q P g S V S V S P G Q T A S I T C(CDR1)W
0
Sbs C1.5 SbS U11 SbS VAX HuVAl HuVAll HuVAllI HuVAlV HuVAV HuVAVl MuVAJ558 MuVA26 ChickVA XenOVh
0
0
.
C(CDR1)W
0
0.0
& I Y 8 F & M 8
S P V L V I Y
A P L L V I Y
A P K V I I Y A P T T V I Y
a
0 0
0
G P
L F T G
L
S P T N
M I Y
I G
A P V T V I Y R P R Y L L R 0 0 . 0 .
( C D R 2 ) G I P N R F A G S V D S S S N K M H L T I T N V Q S E D A T D Y Y C ( C D R 2 ) G g P N R F T G S V D S S N D K M H L T I T T V Q S E D A & D Y Y C ( C D R 2 )G T P G R F (CDR2)G 2 P D R F
g S
(CDR2)G I S S R F (CDR2)G I P E R F (CDR2)G g P D R F
G R Y D S S A N 2 R L T I S Q V Q G S * * K S G A S A S L A I G G L Q G S
g g
G S G S
( C D R 2 ) G g P D R F S G S (CDR2)G (CDR2)G (CDR2)G
2
P D R F
g g
P A R F S D R F
(CDR2)N I P S R F (CDR2)G
g
P D R F
S S S S g
G S
G S
* * *
R S G N T A S L T I
2 2
S
G L
T
2
E D
Q A E D
A T L T I G T Q A L D A S L T I T G A Q A E D
D Y Y C D Y Y C
U
D Y Y C
EA
D Y Y C D Y Y C
* *
N S G N S S G N
* *
K S G N T A S L T V S G L Q A E D E _ B D Y Y C
D R S S N S A S L T V S E L K
* *
L I G D K A A L T I T G A Q
*
K S G S
2
E D E D 1
g A T L T I T G V R A D D
G S K D S P N N I G
X
U
2
G S I D S S S N S A F L T I N N V Q A E D G S
EA
G D U
U
D Y Y C
EA
D Y Y C
&
V Y Y C
N
L T I K G A L L E D D
Y F C
D Y Y C
FIG.14. Comparative alignments of V h framework 1,2, and 3 sequences of sandbar sharks, humans, mice, chicken, and Xenopus. Residues shared with sandbar shark CI.5 are shaded; additional matches are underlined; 0 , invariant positions; 0, positions shared by at least nine sequences, including those of sharks; *, gaps. Positions of the CDRs are indicated within parentheses.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
455
FIG.15. Unrooted tree showing the light chain V regioii types present in cartilaginous and bony fishes. Types 1, 2, and 3 classification are from Rast et al. (1994).
VLsequences of the horned shark and little skate, corresponds to the type 1 designation as defined by Rast et nE. (1994). This group has some “Alike” characteristics, but forms H group separately from the h-type molecules, which include the sandbar shark VA6 homologs and related molecules of the little skate and the ratfish. The third cluster includes the nurse shark K chain and the human V ~ 3 sequence. b VLsequences of the catfish and the trout also fall within this cluster. Figures 14 and 18 show actual sequence relationships within the K and A families substantiating these identifications. Comparative alignments of Ch domains of diverse vertebrate species are illustrated in Fig. 16. An exact coinparison between Ch sequences of sharks and those of the other vertebrates is somewhat difficult because sharks contain hundreds of A light chain clusters, each of which is approximately 90% identical to one another (Hohman et nl., 1992, 1993, 1995). The variability in sandbar shark A chain constant regions occurs primarily in the first 13 N-terminal residues and in residues lying outside of the P bands. Twenty-five residues are invariant in this alignment with certain motifs obvious in each sequence, e.g., the universally conserved PPS se-
Human Pig Mouse Chicken Xeno SbS
Q Q Q Q D N
P P P P V P
K K K K K R
Human Pig Mouse Chicken Xeno
A T D D K E
W W W W W W
K A D G S P V K A G T N K Y A A S S Y L S L T P E Q W K s s A L S A S D W K K A G G T T V U E T T K P S K Q S N N K Y K q D G T P V Z Q G M E T T Q P S K Q S N N K Y E A S S T L X L T & R A W E V I D G S - - Z R S G E T T A P Q R Q S N S Q Y H A S S Y L S L S A S D W S L Y D G K D Q X D S V Q S s G L S K Q S Q N L Y E E S S Y L S L T A D Q W L T Y D G S V R G N G V E T S R I Q Q E A Q N T F S V S S Y L X L S J & E W N
SbS
Human Pig Mouse Chicken Xeno SbS
A A S V A -
N A S A S
P P P P P P
T T S T g T
V V V I V M
T N T T g S
L L L L I L
F P P S S E E L - Q A N K A T L V C L I S D F Y P G A V T V F P P S S E E L - G Z N K A T L V C L I S D F Y P G A V T V F P P S S E E L - E Z N K A T L V C T I T D F Y P G V V T V F P P S K E E L N E A T K A T L V C L I N D F Y P S P V T V F P P S V E E I - A ~ K K A T V V C S L S D F T P R G & T V L P P S D I T - A K N M A T L V C L V S G F N P G A A E I
S H R S V S C Q V T H E - - G S T V E K T V A P T E C S V T H E - - G X I V E K T V T P S E C A V T H E - - G H T V E K S H = A D C S S H ~ ~ X r C R V T H D - - G ~ S & T K T & K ~ E R H E T Y S C K V S H Q - - G K E & I Q T & K U E C V S H E L ~ S C L V K H E A L A N P L R T Q I S R S S C M
C
FIG.16. Comparative alignments of CA domains of diverse vertebrate species. Matches with the human CA sequence are shaded. Additional identities are underlined. Xeno, Xenopus; SbS, sandbar shark.
EMERGENCE AND EVOLUTION OF TIIE IMMUNOGLOBULIN FAMILY
457
quence and the readily apparent KATLVCL around the first cysteine forming the intradomain loop. The sandbar shark CA is an exception in that there is a methionine (M) instead of the lysine (K) residue, but the others match exactly. Schluter et al. (1989b) initially identified this segment as highly conserved in CA domains, C p domains, and some segments of the p chain. Overall, the degree of identity illustrated here is in exact accordance with the phylogenetic relationships among the species considered; viz. the identity between human and pig is 76%, with mouse 70%, with chicken 58%,with Xenopus 46%, and with the sandbar shark 41%. As noted in the K comparisons below that, although all light chains contain overall similarities, the K and A show sufficient variance that characteristic differences can be noted. This observation holds for the KATLVCL motif in A light chains and TCRP chains. The sequence VCL is present in CK in the corresponding position, but the other residues do not match the A motif. Despite the fact that it is difficult to establish definitive orthologous relationships over the great phylogenetic distances covered here, the sequence comparison of VA and CA domains indicates approximately 50% conservation of V region sequence and slightly less (approximately 41%) for the constant regions over a time period of 450 million years. Despite the fact that all of the A light chains of sandbar shark have the V and L gene segments fused in register in the germline, complementarity determining regions matching exactly in relative positions to the CDR1, CDR2, and CDR3, respectively, of human A chains are readily identifiable (Hohman et al., 1992, 1993; Marchdonis et al., 1993). A comparison of the CDR sequences of individual sandbar shark VAs and representatives of the orthologous human VA6 subgroup is shown in Fig. 17. In addition, CDRs of sandbar shark VAx and human VA5 are added to give comparisons with other families. The CDRl segments of shark VA and human VA6 show numerous identities, but these are infrequent in CDR2 and CDR3. CDRS of both species is the most variable, which is noteworthy in the shark because V and J are fused in register in the germline. In contrast, the V and J require rearrangement in the human. Furthermore, the diversity in the shark CDRS is greater than that in the human VA6 set, both in length and in sequence polymorphism. Light chains that have been identified as orthologs of maminalian K chains have been described in Xenopiis (Schwager et al., 1991),the channel catfish (Ghaffari and Lobb, 1997), and the nurse shark (Greenberg et al., 1993). Figure 18 illustrates comparative alignments of VK framework sequences of these vertebrate species, including a comparison between ~ human V K subgroups. ~ The FGXGT(Kor R)(Vor L) motif human V K and of the FR4 is that expected of light chains. However, the EI-K(or R ) Cterminal sequence is characteristic of K chains. If only frameworks 1, 2,
Sandbar Shark CDRl________-__ ---- CDR2W A * S(W----S)E S D R M A S(G----C)A T * S(W----D)E K S G T A S(G----C)Y (G----C)A T * S(W----D) (G----C)A A * S(W----Y) (G----C)A v * s(w----H)N T s T v s R(G----c)G
C) 3 ; A C) T A
U31 T46 VA X
&
3M
s
u11
T54
02
C)
C)
TM
C) T L
Human VAVl(TH0) VAVl(LBV) VhVlONLT) VAVl (AMY) Vhvl VAVl(NIG) VAV(MCg)
(sun
C) T R C) T G C) G C) T G C) T R C) R C) ’$ G T S S D V G G
5
CDR3----5 P Y R S I(F
-x__----
_I____
Shlc5.1
~
* Q(W----Y)E D N Q * Q(W----Y) N Q * Q(W----Y) N Q Q(W----Y)D D N Q
* Q(W----Y)8 D T Q Q(W----Y)D T N Q N Y V S(W----Y)Fs V N K
R R R R R R R
A A A S R
A A R T
* * * * * *
*(F
V A D S A A S A R R S vwLN
G I F T * * G I F T * * G A N T F I s E R L Y L
*(F
P S(G---- C)* P L(G----C)* P S(G---- C)* P S(G----C ) * P S(G---- C)* P Y(G---- C)* P $(G---- C)*
* * * *
*(F
*(F *(F
Q 9 Y D Q 9 F D Q S Y D N N N H V V (F Q 9 Y N S N H H V V (F Q S Y D H*WV(F Q S Y D N L W V (F S S Y E G S D N F V (F
FIG.17. Alignment of V h chain CDR segments of sandbar shark and human. Positions of the FRs are indicated with parentheses. The three-letter designations in parentheses are from Kabat et al. (1991).
FRI
F R 2e
HuVKIII E I V L T Q S P G T L S L ~ ~ ~ ~ D1 I1 V 1L T Q S P A S L A V Q I V M T Q S P D Y V S V XenVL1 Catfish NuSh
~
~
S P G E R A T L S C [CDRlIW Y Q Q K P G Q A P R S L G Q R A T L S C W Y Q Q K P G Q P P K
S P G E I V L T Q S P G S Q S V S P G D I T M T Q S P P V L S V G L G ~ D I V~ M Tl Q S lP L S L P V T P G
E A Q E
T S T P
V V A A
I T T S
L L I I
T C W H T C [CDRlIW Y T C W Y S C [CDRlIW Y
Q L Q L
Q K Q K Q R Q K
S P E P
G G G G
Q E Q Q
V P K A P K K P S S P Q
HuVKIII
e e e e e e e e F R 3 e e e e L L I Y [CDR21 G I P D R F S G S G S G T D F T L T I S R L E P E D F A V Y I G I P A R F S G S G S R T D F T L T I N P V E A D D V A T Y
Catfish NuSh
L L L L
MuVKIII XenV,I HuVKII
L L L L
I I I I
Y G T Y [CDR2] G V Y G V Y [CDR21 G V
P E R I S F S G S G T D F T L T I S R M E A E D A A D Y
P D R F S G
P D R F
Catfish NuSh
HuVKII
Y Y Y Y Y Y
C [CDR3] F F c c F C [CDR3] F F C C [CDR3] F
G R G T E F S L T I S G V Q S E D A G D Y
S
G S G S G T D G T L K I S R V E A E D V G V Y
FR4
e e
HuVKIII MuVKIII XenVL1
S
S E R F S G S G S G T S F T L T I S N V Q N E D V A D Y
G G G G G G
G G K G K Q
G G G G G G
T T T T T T
K K R K K K
V L V L L V
E I E E I R E
I L V L I
K - K - K - K S R - K -
FIG.18. Comparative alignments of frameworks of VK sequences of diverse vertebrate species. Matches with the human VKIII sequence are shaded. The solid circles indicate invariant positions. Positions of the CDRs are indicated within brackets. Sequences are: Hu VKIII, human VKIII (HuKV325 and FR4 froin CUR): Mu VKIII,mouse VKIII (OC3741);Xen VLI, Xenopus VLI (Zezza et al., 1992); catfish, catfisli \JK; NuSh, nurse shark VK; and HuVKII, human VKII (GM607).The names in brackets for human and mouse are the sequence designations in Kabat et al. (1991). Other sequences are cited in the text.
460
JOHN J. MARCHALONIS et al.
and 3 are considered in the comparison, human and mouse V K are ~ 77% identical to one another. This is consistent with a family originating prior to the species divergence. The result is not so clear-cut, however, because the human V Kand ~ V K show ~ 70% identity, which is borderline for family identification at the protein level. It is striking that the overall degree of identity in this diverse set of V Ksequence is greater than 60% in all cases. Although it is possible that this set of examples is too small to obtain a representative pattern, VK sequences may prove to show more conservation in evolution than do the Vhs. The chicken only has functional genes for h chains so there is no K comparison with birds at this time, and no data are currently available for possible K or A light chains of reptiles. Alignments for the CK domains of these species are given in Fig. 19. The situation here contrasts with that of C h because the conservation between human and mouse sequences is only 57% as opposed to 70% for the CA comparison between the same two species. The degree of identity between the shark and the human CK,however, is 46%, which is comparable to the degree of identity between the sandbar shark and human domains. In addition, to the more clear-cut examples illustrated here, the Xenopus has a class of light chains termed “sigma” which are “K-like” but may be unique to that species (Haire et al., 1996; Schwager et al., 1991; Stewart et al., 1990). Moreover, light chains of other teleost fish also fit the description of being “K-like” but not as clearly as the catfish example illustrated here (Bernstein et al., 1997; Daggfeldt et al., 1993; Ghaffari and Lobb, 1993). IX. Origin and Evolution of Immunoglobulin Heavy Chains
A. EVOLUTION OF VARIABLE DOMAINS Variable regions and joining segments expressing characteristic features of their mammalian homologs are present in all jawed vertebrates. Figure 20 gives comparative alignments of the VH segments of IgM and IgW immunoglobulins of sharks and IgM molecules of bony fishes with a human VH3 sequence given for comparison. There are a number of highly or universally conserved residues and these generally correspond to canonical residues identified by Chothia et al. (1992) as critical for proper inter- and intrachain foldmg and association of the VHand VLdomains. Overall, the VHsof primitive vertebrates tend to show the greatest similarity to the vH3 set of humans (Kirkham and Schroeder, 1994; Schluter et al., 1997; Shen et al., 1996), although critical residues in the FR3 that are involved in binding of the staphylococcal A superantigen (Silverman, 1992) do not occur. Among the highly conserved residues are the cysteines (C) involved in forming the intradomain disulfide bond, the conserved tryptophan (W)
Human Mouse Xenopus Catfish Nurse Sh
Human Mouse Xenopus Catfish Nurse Sh
Human Mouse Xenopus Catfish Nurse Sh
R T V A D N D
. . .
.
.
. K D I N V R D L T V
. .
D R S Q P K L T L L
N I D G S G K D G
R L A L W E K S D W E S
W A G N
A S Q S a
L E D E V
Q R V V I
S Q S K S
K N D E A
L S K E R
Y Y F F F
A T E V S
. ..
P D Q V Q T K
G N S D D
N G S G G
.
S V D V V
Q L V L Q
E N K T T
S S T S S
V Q D A N
T T F E Y
Q D M E L
A Q -
P D E L Q V e..
. -
- - L T Q
S K G S Q C S - R K S E V - S R S A
FIG.19. Comparative alignment of C K domains of diverse vertebrate species. Matches shown are with the human sequence. The solid circles indicate invariant positions. The human and mouse sequences are designated ROY and MOPC41C1, respectively, in Kabat et al. (1991). Other sequences are cited in the text. Nurse Sh is nurse shark K light chain.
FIG.20. Alignment of V, sequence from cartilaginous and bony fishes. A human Vr13sequence is included for comparison and is in boldface. Conserved canonical residues necessary for proper folding of Ig V regions (Chothia et al., 1988, 1992) are shaded. The positions of residues important for intrachain folding are numbered. Residues indicated by a h are involved in interchain folding, and by @ are important for turns.
EMEKGENCX AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
463
located at the beginning of the second framework, a conserved peptide GLEW (hydrophobic) in the second framework, and the conserved sequence YYCA at the end of the third framework (FR3). A three-diinensional model illustrating the position of these coilserved residues was constructed by mapping their positions onto the structure of a VH domain taken from the structure of a human Fab with V,, subgroup 111. Figure 7-21A (see color insert) presents the structure froin the perspective of the outer four-stranded /3-pleated sheet, and Fig. 7-21B (see color insert) shows the five-stranded P-pleated sheet surface, which presents the interdornain contact sites. In the five-stranded P-pleated sheet of VH, the conserved residues include the second half of the intrachain disulfide bond, internal hydrophobic (or aromatic) residues (e.g.,leucine H20, tryptophan H36, and tyrosine H90), and patches of residues important for noncovalent interactions with light chains. Although many residues are involved in the proper docking of VH and VL, leucine H45, tryptophan H47, tyrosine H91, and tryptophan H103 (all conserved) are ainong the most prominent. Note that these residues are all clustered in the lower half of the VHwhere the crossover surfaces with regions of VL are concentrated. The conserved residues of the four-stranded &pleated outer sheet of the VH domain are priniarily hydrophobic. These include one-half of the disulfide bridge, which connects the four- and five-stranded P-pleated sheets. Most of the remaining conserved residues have their side chains directed inward where they help maintain the hydrophobic character of the domain’sinterior. Segments of irregular loops at the ends of the pleated sheets tend to be those favorable for turns, such as the Pro-Gly sequence at positions 14 and 15 and 41 and 42, or the Ser-Gly coinbination at positions 25 and 26. The Glu-Asp-Thr-Ala (EDTA) segment at positions 85-88 forins a distorted helix at the distal ends of these domains. The phylogenetic relationships among the V domains presented in Fig. 22 were analyzed using CLUSTALW with the unrooted tree plotted by PHYLIP DRAWTREE programs. These VH sequences, with the exception of the human VH3, are derived froin either elasinobranchs or bony fishes. Three clusters were disclosed. Two of these clusters, the IgW type and the shark IgM type, are restricted to the elasmobranchs. Within the “IgM” group, the Vli domains of the sandbar shark formed a group separate from that formed by the horned shark type 1 VH sequences and the type 1 Vll of the little skate. Although more than 30 individual sequences were analyzed, there was no evidence for conservation of particular individual Vlls w i t h the elasinobranchs. This contrasts with findings of Anderson and Matasunaga (1995b) who reported that some VI1gene families of teleost fish are stable in evolution for 150-200 million years or longer. A VtI
464
JOHN J. MARCHALONIS et
d.
FIG.22. A radial phylogenetic tree of cartilaginous and bony fish VHsequences. A human vH3 sequence is also included. The tree was calculated using the program CLUSTALW (Thompson et al., 1994) and displayed using the program TreeView (Page, 1996).The VHS sort into three groups, with the cartilaginous VHsclearly forming two distinct classes: the
IgM group and the IgW group.
sequence of the bull shark, a close relative of the sandbar shark, clusters with the sandbar shark VH group. This group can be divided into six families accordmg to the criterion of 80% DNA sequence identity (Shen et al., 1996). The IgW cluster contains the set of Vw sequences from sandbar shark and bull shark IgW molecules (Bernstein et al., 1996b), its ortholog from the nurse shark, IgNARC (Greenberg et al., 1996), and homed shark and little skate VH type I1 sequences (Rast et al., 1994).The third group consists of two major clusters; one is restricted to the teleost fish and shows the clustering of VH genes from widely divergent teleosts such as cod and trout, or catfish and trout, which is consistent with the results of Anderson and Matsunaga (1995b). It also illustrates the relatively large number of VH families in catfish (Ghaffari and Lobb, 1991; Warr et al., 1991) and trout (Anderson and Matsunaga, 1995b). Furthermore, a subgrouping of this set can be considered to be the one leading to the VHfamilies of higher vertebrates and shows particular
EMERGENCE A N D EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
465
relationship to the human vH3 subgroup. VH3-relatedsequences occur in the goldfish (Wilson and Warr, 1992), the catfish (Ghaffari and Lobb, 1991; Warr et al., 1991),the ladyfish (Amemiya and Litman, 1990), the trout (Anderson and Matsunaga, 1995b), and the coelacanth (Amemiya et al., 1993), a Sarcoptyrgian fish possibly related to ancestral tetrapods. Overall, the evolutionary pattern of the emergence, waxing, and waning of VH families is consistent with a “birth-and-death” model of evolution in which new genes are generated by gene duplication, with some duplicated genes retained for long times whereas others are deleted or become nonfunctional because of deleterious mutations (Nei et al., 1997). B. EVOLUTION OF THE CONSTANT DOMAINS OF THE UNIk’ERSAL U , CHAIN Although the p chain is the only one considered to be universally conserved among the gnathostomes (Du Pasquier, 1993; Marchalonis, 1977), the clustered-type (cassette) gene organization in sharks (Hinds and Litman, 1986) presents some possible theoretical difficulties for precise estimates of evolutionary divergence. An immunoglobulin isotype is specified by the heavy chain of the molecule, and in every animal except elasmobranchs, all of the V(D)J and CHgene segments lie within a single gene translocon. By this definition, either the shark has hundreds of isotypes defined by each individual heavy chain gene cluster or it has one “mega isotype” (Marchalonis et al., 1993) in which the heavy chains of all of the clusters are considered to belong to the p isotype. There is variability in comparison among individual C p domains of sharks (Kokubu et al., 1988a), just as there is in comparisons of individual Ch sequences (Hohrnan et al., 1995; Marchalonis et al., 1993; Schluter et al., 1989a,b). However, the degree of this variability is generally less than 10% overall. Thus, in overall phylogenetic comparisons, the error introduced by taking any individual shark C p constant domain for comparison with C p sequences of higher vertebrates where only one copy is available may lead to some variation, but this should not be appreciable. Figure 23 presents an alignment of the Cp1 through Cp4 constant domain sequences of sandbar shark, rainbow trout (Anderson and Matsunaga, 1993), Xenopus (Schwager et al., 1988), chicken (Dahan et al., 1983; Parvari et al., 1988; Reynaud et al., 1989), and human (Friedlander et al., 1990) molecules. Thirty-six universally conserved residues occur. As expected, these correspond to cysteines required to form intrachain disulfide bonds, tryptophans that protect the disulfide bonds, and in C p l , a cysteine required to form the interchain disulfide bond with the light chain.
shark IgM
trout IgM
Xenopus IgM Chicken IgM Human IgM
Illkz;." CDR2
VSWSSSITPVXAPSIFQ AAHYI)IRNI-WSQSVQG
GRIYPGD=~SS~
AAISSXSG
GGII PIFGTANVAQKF~
shark IgM
trout IgM
Xenopus IgM Chicken1 M Human lgb
shark IgM trout IgM Xenopus IgM Chicken1 M Human lgb
Chicken IgM WQQDIAIRVI Iiuman IgM DQGTAIRVFAI
EPLP-- - QSQSVLSAPMAENPENES QPLS---PEmSAF'MPEPQAPGR
FIG.23. Alignment of C and V regions of p chains from shark, trout, Xenopus, chicken, and human. Invariant positions are shaded.
467
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
TABLE IIA RESIDUESCOMPARIUG Ig HEAVYC I I A I N C O N S T A N T REGION DOMAINS OF VARIOUS VEHTEHHATICS -ro DOMAINS OF HUMANIgM
PER(:ENT4GE OF I D E N T I C A L
CH1
Gap penalty Sandhar shark Nurse shark Rainbow trout Xl?n opus Chicken
3.0 30.3 32.0 37.1 32.7 43.6
10.0 25.0 27.9 32.0 28.4 35.0
C H2 3.0 26.5 24.8 26.8 27.0 29.7
10.0 23.8 22.5 22.8 25.5 18.6
CI-I3 3.0 30.6 28.4 22.1 31.0 35.0
10.0 26.0 25.3 17.7 27.6 29.5
CH4 3.0 37.5 37.8 35.8 29.2 42.5
10.0 37.5 36.6 27.0 25.5 42.5
Table I1 presents the percentage of identical residues of Ig heavy chain constant region domains of various vertebrates compared to the domains of the human p chain (Table IIA) and also the percentage of identical residues coinparing the sandbar shark C p domains to those of the nurse shark (Vazquez et al., 1992) and the horned shark (Kokubu et al., 1988a) (Table IIB). In the general phylogenetic comparison (Table IIA), the most highly conserved domain is Cp4, showing approximately 37% identity on the average, with C p I , which shares approximately 35% coverage identity, a close second. The least conserved is Cp2. Within the elasniobranchs, Cp4 clearly shows the strongest identity, with greater than 75% overall. The other three domains show approxiinately 60% identity in the three species compared. Because the ancestral divergence between the nurse shark and the carcharhines was approximately 80 million years and that to the horned shark approxiinately 150 million years (Rast and Litinan, 1994), an evolutionary gradient would be expected. This is suggested in comparisons of C p l , Cp3, and Cp4, but the magnitude of the difference is small. It was previously noted that individual constant domains of Heavy chains have apparently evolved separately of one another and that the key feature in sequence conservation was not the designation of the heavy chain to which they belong, but rather their position within the sequence of heavy chains (Barker et nl., 1978; Liu et al., 1976; Marchalonis and TABIX I I R PERCENTAGE OF IDENTICALRESIIXJES COMPARIT\;(;SANIIHAH SHARKIg HEAVYC r r m C O N S T A N T REGIOND O M A I N S TO NURSE SIIARK AND HOIINED S l f A H K DOMAINS ~~
CHI
Gap penalty Nurse shark Horned shark
3.0 66.0 56.6
10.0 66.0 56.6
CH2
CH3
CH4
3.0 10.0 63.0 63.0 66.0 66.0
3.0 10.0 67.4 67.4 67.6 67.6
3.0 10.0 78.5 78.5 76.2 76.2
468
JOHN J. MARCHALONIS et al.
Schluter, 1989; Schluter et al., 1997). For example, the constant domains most proximal to the V domain formed one group in evolution, whereas the most distal set formed a separate group, whether these were the third Cy domain or the fourth Cp. This “position effect” was analyzed by comparing the phylogenetic relationships among the individual C p domains of mammals, birds, amphibians, bony fishes, and elasmobranchs by two approaches. In the first approach, the unrooted tree showing the phylogenetic relationship of all C p region domains among these vertebrates was constructed with CLUSTALW and plotted by PHYLIP DRAWTREE programs as illustrated in Fig. 24. The individual domains clustered together on the basis of position in the C p sequence. Within the individual clusters, the expected phylogenetic relationships can occur. For example, with respect to C p l , the relationships among the little skate, the sandbar shark, the nurse shark, and the horned shark are those that would be
FIG.24. The phylogenetic relationship of IgM C region domains of various species from sharks to humans. This unrooted tree shows the phylogenetic relationship of C p domains among representative vertebrates. The tree was constructed with CLUSTALW and plotted by PHYLIP DRAWTREE programs.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
469
predicted based on known phylogeny of the elasmobranchs (Compagno, 1988). Interesting differences can occur in that the C p l of the duck and Xenopus cluster together in this analysis. Standard phylogenetic relationships do not always hold because in Cp4, the duck clusters with the mammals whereas the Xenopus domain is associated with those of bony fishes. The overall inter- and intradomain patterns reflect the fact that the individual C p domains arose by tandem duplication extremely early in the emergence of immunoglobulins, and there has been ample time for separate evolution of the domains based on die selective pressures incumbent on their particuIar position. In order to determine whether the evolution of individual C p domains occurred at a constant rate, and whether the individual rates are comparable to those of one another, the authors computed the evolutionary distance of each individual C p domain compared to the corresponding human segment for the representative vertebrate species studied. This distance, 6, is defined as the negative natural log (In) of the fractional identity of the two amino acid sequences (Nei, 1975). For example, if the two sequences are 75% identical, this computes to -ln(0.75), which is equal to 0.288. These values are then plotted against divergence time as estimated from the fossil record, with the time expressed in millions of years. The rate of evolution is obtained from the slope of linear regression analysis. Evolutionary regression analysis for the 4 C p domains is illustrated in Fig. 25. Significant correlation coefficients are obtained for C p l , Cp3, and Cp4, but the correlation shown in Cp2 is weak. The rates of evolution of C p l , Cp3, and Cp4 are fairly constant in the range from 0.9 to 1.2 X lo-’ substitutions per amino acid per year. The rates obtained here are comparable to those estimated by other workers using different sets of divergent vertebrate species; for example, Anderson and Matsunaga (1993) obtained a rate of evolution for the C p domain of 1.4 X lo-’ per amino acid per year. From data presented earlier for constant domains of K and h light chains, the authors estimate a rate of evolution of 1.5 X lo-’ per amino acid per year for Ch domains ranging from sharks to humans. The value obtained from the data set of CK domains in Fig. 19 was approximately 1 X lo-’ per amino acid per year, which suggests that these domains are somewhat more conserved in evolution than are the constant domains of h light chains. The rates of evolution obtained fall in the range of 1-1.5 X lo-’ per amino acid per year for immunoglobulin constant domains. This rate value is relatively conserved by comparison with molecules such as cytochrome c, a slowly evolving protein that has a substitution rate of 0.2 X lo-’ (McLaughlin and Dayhoff, 1972), and fibrinopeptides, the fastest evolving protein known, with a rate of 9.0 X lo-’ (Ohta and Kimura, 1971).
470
JOHN J MARCHALONIS et nl.
m 0.6
o. o.2 00
300
600
00
900
600
900
2 x Divergence time (mya)
2 x Divergence time (mya)
3
300
CP3
1.47
CP4
a
1.o
m 0.8 0.8
B 1.0
I
0.6 0.4 0.2
f(X) = 0.2925 + 1.203 x 10E-3x R = 0.9214
0
300
600
900
2 x Divergence time (mya)
f(x) = 0.203 + 1.104~10E-3x R = 0.9329
0.4
0.2 0
300
600
900
2 x Divergence time (mya)
FIc. 25. Evolutionary rates of IgM C region domains. Plots show the amino acid substration rate of Ig constant region domains of representative vertebrates. The amino acid substitution value [a = In(Ni/N)] of each pair of species compared was plotted against the 2x divergence time of these species.
Using data in the TCR alignments, the authors estimate the rates of evolution of these chains to be as follows: Ca, 3.0 X lo-’ per amino acid per amino acid per year; and Cy, 2.7 X lo-’ per per year; Cp, 2.6 X amino acid per year for the Ig domains. On average, the rates of evolution of the TCR C domains are significantly faster than those of light chains and Cps. Andersson et al. (1995) have built on the types of calculations outlined here to estimate divergence times of various teleost fish in cases where fossil evidence was not available. C. APPEARANCE OF ISOTYPES The secondary effector functions of antibodies, e.g., complement fixation, are generally mediated by the constant regions of heavy chains. Thus, the diversification and apparent increase in numbers of heavy chain isotypes through evolution have generally been thought to be the result of selection to increase the functional diversity of immunoglobulins resulting from the requirement of a more sophisticated immune system in higher vertebrates.
EMERGENCX AND EVOLIJTION OF THE IMMUNOGLOBULIN FAMILY
471
Although this view is probably essentially correct, it is also somewhat of an oversimplification.This is highlighted by data from sharks that indicate that the early iinmune systein was highly advanced in terms of the structure and function of the Igs, the presence of different isotypes (Anderson et nl., 1994; Bernstein et nl., 1996b; Greenberg et nl., 1995, 1996; Harding et nl., 1990), and the expression of a highly diverse binding repertoire (Casson and Manser, 1995; Hinds-Frey et nl., 1993; Hohman et al., 1993; Kokubu e f nl., 1988b). The presence of multi-isotypes in sharks and other lower vertebrates has been unequivocally demonstrated using molecular biology and gene cloning techniques. However, the inicrodiversification in apparently homogeneous isotypes is a phenomenon not generally appreciated, particularly for sharks and fish in which IgM is the major isotype expressed. The shark has hundreds of chain constant genes, each slightly different in sequence. These differences may be functionally significant. One distinction is clear and that is the presence of monoineric and pentaineric forins of IgM. A functional distinction between these two forms is suggested by tlie finding that pentameric molecules are located to the intravascular spaces of the shark body cavity, whereas the monomeric forms are found both in the intra- and extravascular space (Clem and Small, 1967). Similarly, tetraineric IgM is essentially tlie only immunoglobulin secreted in teleost fish, but studies with inonoclonal antibodies have demonstrated considerable heterogeneity within this class (Lobb et al., 1984; Roinbout et nl., 1993; Sanchez et al., 1995; Sanchez and Dominguez, 1991). For example, inucosal and serum immunoglobulin, and the cells that secrete them, can be differentiated using inonoclonal antibodes (Roinbout et nl., 1993).Heterogeneity has been demonstrated at the gene level in salmon, where two isotypic IgM heavy chain gene loci have been identified (Hordvik et nl., 1992). The formation of distinct isotypes is a prominent feature of the phylogeny of tlie iininune system. Delineating ancestral relationships is a difficult task that involves searching for inforination that indicates a coininon phylogenetic history, i.e., the sharing of a common ancestor. This is best done using statistically rigorous methods to deinonstrate hoinology/siinilarity. Various approaches are available, each with different, sometimes significant, underlying assumptions (Doolittle, 1990; Felsenstein, 1981; Huelsenbeck and Rannala, 1997; Saitou and Nei, 1987; Thompson et d.,1994). A major problem in these studies is that we are dealing with extant genes that have undergone separate evolution, soinetirnes for hundreds of millions of years, Thus, the reinnant sequence inforination is at the liinit of our abilities to detect significant relationships. In such cases, we have to rely on additional inforination froin other characteristics, such as conservation of sequence motifs, and analysis of selected highly conserved segments,
472
JOHN J. MARCHALONIS et al.
such as the transmembrane regions. Even when ancestral relationships can be identified with confidence, the exact evolutionary steps may still not be clear. For example, the IgYs are evolutionarily related, but is the chicken Y gene (IgY) a direct lineal descendant of the Xenopus v gene or are these two genes the products of separate descent from a common ancestor that duplicated and diverged before the amphibians appeared (Mussmann et al., 1996)? Despite these caveats, several broad conclusions regarding the appearance and phylogenetic history of heavy chain genes (i.e.,the phylogeny of isotypes) can be stated. The evolution of isotypes is marked by three or four critical duplication events. The first duplication was the formation of IgM and IgW in elasmobranchs. The discovery of IgW (Bernstein et al., 199613; Greenberg et al., 1996) challenges the once undisputed position of IgM as the primordial isotype. The IgW heavy chain shows closer homologies in comparison between its V and C domains than do any other heavy or light chains (Bernstein et al., 1996b).The authors have taken this as evidence that IgW was the primordial molecule based on the assumption that this represents a closer relationship to the prototypical Ig domain existing before the duplication leading to C and V domains. Phylogenetic analysis showing 1: 1 mapping of C p domains with corresponding C w domains (e.g., Cp1 and Cwl) demonstrates that p and w arose from a duplication of a common ancestor (Schluter et al., 1997). The genetic/ phylogenetic distances between corresponding w domains in different species compared to the same distances for p chain domains are approximately the same, indicating that this duplication occurred very early in the evolution of the immune system. These analyses also show that the primordial heavy chain was composed of four constant region domains. We can only speculate about the events leading to the formation of this primordial heavy chain, but it is clear that it was already highly evolved, as these results also show that the individual primordial domains were as differentiated from each other then as is evidenced in constant domains today. This positional effect, as noted earlier, indicates important evolutionary constraints on domains that is overlaid on the selection and diversification of isotypic heavy chains. Heavy chains containing four domains appear to be the most evolutionary fit form of the molecule. Thus, the primordial heavy chain was composed of four domains, all duplication events forming new isotypes involved ancestral genes with four domains, and most immunoglobulins present today are composed of four domains. However, there has been some evolutionary drive for the appearance of heavy chains with fewer domains. For example, the IgX heavy chain in skates contains only two constant domains, but these are clearly orthologous to the first domains of the w
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
473
heavy chain (Schluter et al., 1997). IgX is encoded by separate genes containing only these exons (Boman, 1997) and must have arisen from a duplication and deletion of domains of an ancestral w gene. Similarly, a truncated form of IgY is found in some birds, reptiles, and lungfish that consists of the first two constant domains of the v chain. In this case, however, both intact and truncated forms are derived from the same gene by different RNA processing resulting from the presence of a polyadenylatiodcleavage site situated between the C v 2 and Cv 3 exons (Magor et al., 199413).The functional significance of these isotypes is not known. Indeed, it has been suggested that such immunoglobulins are “maladapted’ and represent a significant disadvantage for host species (Warr et al., 1995). The other type of shortened heavy chain genes consists of three constant domains and is found only in mammals. The major example is IgG. It appears that the y chain was derived from a four-domain molecule by a shortening of the Cn2 domain to form the hinge segment. The obvious advantage of this is the increased flexibility of the Fab segments of IgG (Carayannopoulos and Capra, 1993). Because of the unique cluster “megaisotype” arrangement of immunoglobulin genes in sharks (see next section), the w and the p isotypes are each associated with a separate set of Vn genes. This is in contrast to higher Vertebrates in which the VH genes are common to all isotypes (Max, 1993). Phylogenetic analysis shows that the IgW and IgM V regions form two distinct classes (Fig. 22) (Schluter et al., 1997). This is the case for all elasmobranchs, including carcharhine, horned and nurse sharks, and skates. Thus not only do the two isotypes have unique heavy chains, the V regions associated with them are also very distinct (Schluter et al., 1997). The polarization of the two classes of VH genes suggests that the two isotypes arose concurrently to meet separate functional requirements in the primordial immune system. An interesting question arises concerning the relationship of shark Vo and V p genes to higher vertebrate VH.There are many diverse VHfamilies in higher vertebrates, but these are all members of the same evolutionary set that clusters separately from the elasmobranch VHclasses (Fig. 22). In contrast to light chain variable regions, there is no evidence for the sharing of orthologous VH families between sharks and higher vertebrates. Indeed, this is not even the case for VHfamilies between different species of shark (Schluter et al., 1997; Shen et al., 1996). It appears that the rederivation and formation of individual VH families can occur at each radiation of new species (Kirkham and Schroeder, 1994; Nei, 1975; Shen et al., 1996). Analysis of heavy chain, light chain, and TCR V region sequences from all vertebrate groups (Marchalonis et nl., 1986; Schluter et al., 1997)leads to the conclusion that the elasmobranch V p is the ancestral group from
474
JOHN J , MARCHALONIS ct al
which the higher vertebrate V13set arose. Other investigators have arrived at a similar conclusion, Anderson and Matasunaga (1995a) concluded that the p set of sharks forms an archaic clan that includes some VH sequences of teleost and amphibians. Following this nomenclature, the Vw set has been termed a primordial class (Shen et al., 1996).Kirkham and Schroeder (1994) concluded that some horned shark VHsbelong to the same clan as the human VH 3 family. IgM is found in all phylogenetic groups and there is clearly a lineal evolutionary history of the p constant domains from sharks to humans (Du Pasquier, 1993; Mussmann et al., 1996).In contrast, however, IgW appears to be confined to the elasmobranchs. Searches of sequence databases using the BLAST PROGRAM (National Center for Biotechnology Information, National Library of Medicine, Bethesda, MD) indicate that IgW C regions are weakly related to other higher vertebrate isotypes, particularly IgD and IgA, and raise the possibility that the o gene is the forerunner of certain isotypes in higher vertebrates. However, because p is clearly maintained throughout phylogeny, the presumption is that the formation of other isotypes was initiated by the duplications of ancestral p genes. The second significant duplication event was the formation of the IgD isotype in teleost fish (Wilson et al., 1997). This isotype is functionally similar to IgD in mice and humans, as indicated by (1)a similar genomic arrangement adjacent to the p gene, (2) expression by RNA processing of a single transcript comprising the p and 6 C region exons, and (3) the dual expression of IgD and IgM on some B cells. This molecule is very large, with seven 6 and one p domain in the expressed molecule. It is difficult to discern significant relationships when individual domains are analyzed. However, using maximum-parsimony methods and complete sequences, a phylogenetic relationship between teleost IgD and higher vertebrate IgD is indicated (Wilson et al., 1997). Perhaps one of the most significant events in the phylogeny of immunoglobulin isotypes was the appearance of IgY at the level of the anuran amphibians. This is a widely distributed isotype occurring in birds, reptiles, and amphibian (Warr et al., 1995). It is also very versatile, being the major serum immunoglobulin in most of these species. In the axolotl, IgY also has significant secretory function (Warr et al., 1995). The major physicochemical feature distinguishing IgY from IgM is that it does not form polymers, but is exclusively a monomer. The serum form of IgY is considered the functional equivalent of IgG in lower vertebrates. It also shows some functional similarity to IgE, mediating anaphylactic reactions in some species (Grey, 1967). Furthermore, beyond functional homology, results from molecular genetic studies strongly suggest that IgY is the phylogenetic ancestor of IgG and IgE (Warr et al., 199-5).Actual sequence homologies
EMERGENCE A N D EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
475
are low, reflecting the long phylogenetic history of this isotype, but several lines of' evidence support a phylogenetic relationship. First, several sequence motifs are conseived. All v, E , and some subsets of y chains have an extra intradoinaiii &sulfide bond in the first domain. The secretory piece is a characteristic two residues long for y , v (usually glycine, lysine), and E chains, as opposed to the much longer (approximately 20) residues for p, o,Xenopus IgX, and a chains. Examination of the transmernbrane and cytoplasmic domains of the membrane forms of these isotypes yields similar conclusions. Although the membrane segments show no particular sequence similarity, v, y , arid E chains have a unique pattern of conserved residues (Mussmann et d.,1996).These clones also have long cytoplasmic segments as opposed to much shorter segments in p and IgX chains. Finally, phylogenetic analysis show y , E , and v sequences grouping on a common tree branch (Warr et al., 1995; Wilson et al., 1997). Thus, the appearance of IgY was very important, enabling the development of a functionally versatile isotope that appears to be the homolog and phylogenetic ancestor of IgG, one of the major isotypes in mammals. The next significant developinent was the appearance of IgA; this isotype is primarily secreted at mucosal surfaces, including gut epithelium. As such, it is an important isotype for prevention of infection. Bona$de IgA to date has only been detected to the phylogenetic level of birds (Mansikka, 1992).Interestingly, avian IgA is a four-domain molecule, whereas mammalian a chains consist of only three domains. Interdomain sequence comparisons indlcate that the equivalent of the chicken Ca2 domain is missing in mammalian chains (Mansikka, 1992).This is similar to the y chain in which the second domain condensed to form the hinge region. The importance of a specialized secretory isotype in phylogeny is perlyaps demonstrated by the IgX isotype in Xenopus. The majority of Ig secreted at intestinal mucosa is this isotype (Mussmann et d , 1996) and therefore can be considered an analog of IgA. However, the IgX heavy chain is clearly a duplication of the p chain and there is no evidence to suggest a relationship with mammalian IgA. X. Segmental Gene Organization in Evolution
The basic genomic elements comprising immunoglobulin genes have been essentially conserved during the evolution of jawed vertebrates. This is illustrated by considering the structure of the human K chain locus (Fig. 26). The coding segments comprise a split leader, a V region, J segments, and a constant region. Each of these separate coding segments, or exons, is separated by varying lengths of noncoding intron DNA. The light chain constant region is a single domain encoded by a single exon. Heavy chain
476
JOHN
1. MARCHALONIS
et al.
HUMAN TRANSLOCON
lambda
Cpl CV3 T T M l TM2
heavy chain
’, cp2 cp4 30Ds
JH 1-6
‘*,
Cp *,
,a
’
czl
1100 kb
FIG.26. The translocon type of arrangement for germline immunoglobulin genes. The specific examples illustrated are human heavy and light chain genes. Regions expanded for greater detail are indicated by dashed lines. sec, secretory piece: TM, transmembrane exon; RSS, recombination signal; $, pseudogene.
constant regions are composed of several domains, each ofwhich is encoded by a single exon, and these are separated by intervening introns (Fig. 26). This exon-intron structure of heavy chain constant region genes is also highly conserved, although the intron lengths can vary from approximately 350 bp in humans (Max, 1993) to 1-2 kb in skates (Harding et al., 1990). Two other elements that comprise an immunoglobulin locus are the promoter and enhancer regulatory elements, which control transcriptional activity of the gene (Emst and Smale, 1995a,b). The introns are removed from the functional gene by two processes, DNA rearrangement and RNA splicing, both of which are highly conserved.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
477
A functional V gene is formed by the removal of intervening introns such that V, (D), and J coding elements are physically joined (Max, 1993). For example, in the human heavy chain locus, the D element is first joined to the J segment and then the V segment is rearranged to the D. For light chains, the V is joined directly to the J segment. The rearrangement process is directed by recombination signal sequences (RSS). These are composed of 7-mer (heptamar) and 9-mer (nonamer) sequence elements separated by either 12 ( 21) or 23 (+- 1) bases (Fig. 26). This arrangement of RSSs, as well as the actual heptainer and nonamer sequences, is very highly conserved, being essentially the same for heavy and light chains in sharks, teleosts, amphibians, birds, and mammals (Ramsden et at., 1994). Recombination signal sequences have a precise orientation whereby the hepamer is exactly adjacent to the coding sequence (Fig. 26) (Max, 1993). A RSS is present at the 3’ end of the V regions, the 5‘ and 3’ ends of D regions, and the 5’ end of J regions. Rearrangement follows the 12/23 rule whereby joining occurs only between a RSS with a 12-bp spacer and a RSS with a 23-bp space (Ramsden et al., 1994). Thus, the V,{ segments shown in Fig. 26 cannot join directly to J segments as this would involve rearrangement between two 12-bp RSSs ( M a , 1993). The remaining introns in the split leader, between J and C, and between heavy chain constant domain exons are removed by RNA processing, which is directed by splice signals at the exon-intron boundaries. These splice signals are also relatively conserved in immunoglobulins (Harding et al., 1990; Hohman, et al., 1995; Kokubu et al., 1988a; Max, 1993). Immunoglobulin isotypes can be either secreted or membrane bound. This is determined by the presence of either a secretory or a transmembrane (TM) segment at the 3’ ends of heavy chains. Both forms are expressed from the same gene by alternative RNA processing (Max, 1993). The secretory piece is part of the terminal constant region exon and is followed by a strong polyadenylation signal to end transcription. In certain instances, the regulation of which is not well understood (Harding et al., 1990; Kokubu et al., 1988a; Max, 1993; Warr et al., 1995), transcription continues past this signal to include the two transmembrane exons (Fig. 26). A splice donor site is present in exon 4 at the junction with the secretory segment and is used for the addition of the transmembrane exons and the removal of the secretory segment by RNA processing. This basic arrangement of the TM exons and the splice signals is also conserved from sharks to humans (Kokubu et al., 1988a; Max, 1993). The exception to this rule is in teleosts where the TM exon is spliced to a cryptic splice site in Cp3, thereby eliminating Cp4 (Warr, 1995). The transcription of immunoglobulin and TCR genes is precisely regulated, e.g., only B cells express antibodies. Expression is controlled by
478
JOHN J MAKCHALONIS et ul
promoter and enhancer DNA elements and is achieved by the concerted combinatorial action of multiple regulatory events (Ernst and Smale, 1995b). Promoter regions control the initiation of RNA synthesis from a specific location. Enhancers are cis-acting control regions that can dramatically regulate promoter activity. For example, an enhancer is present downstream from the constant region in the human K gene locus (Fig. 26). Normally, the VK promoter is only weakly active but is upregulated to full activity when brought into the proximity of the enhancer by the rearrangement process (Max, 1993).In higher vertebrates this arrangement of promoters and enhancers is essential for the regulation of transcription. However, enhancers are also important in the control of the rearrangement process itself (Max, 1993) and also in isotype switching (Snapper et al., 1997), and so are critical control regions in immunoglobulin genes and TCR genes. Regulation is achieved by the combined action of trans-acting regulatory factors with cis-acting DNA sequence elements (Ernst and Smale, 1995b). There is a large degree of variation and redundancy in the composition of promoters and enhancers, and some elements are found in both, e.g.,the octamer element is present in immunoglobulin promoters and enhancers (Ernst and Smale, 199513). Thus, significant variations may be found in the promoter/enhancer structures and functions in lower vertebrates. This would seem to be particularly true for the cluster arrangement found in teleost fish and elasmobranchs, as rearrangement distances are very short and unlikely to effect enhancer promoter interaction as in higher vertebrates. Most core promoters consist of the TATA box sequence elements (Ernst and Smale, 1995a). The assembly of the transcription complex is initiated by the binding of a specific protein to the TATA box (Ernst and Smale, 1995a). Essentially all V region genes from all species characterized so far contain a TATA box 5’ to the transcriptional start site (Anderson et al., 1994, 1995; Haire et al., 1991; Hohinan et d., 1993; Klein and O’hUigin, 1993; Max, 1993; Pilstrom and Bengten, 1996; Rast et nl., 1994; Reynaud et al., 1987, 1989; Schwager et al., 1988; Warr, 1995; Wilson and Warr, 1992). Most promoters also contain regulatory elements, which are the binding sites of proteins that interact with the transcription complex. For example, the human VKpromoter has an octamer (Fig. 26) and the human VHpromoter has octamer, hepamer, and E box elements (Ernst and Smale, 1995a). The octamer is a very important and highly conserved regulatory element for immunoglobulin genes and is found, with some variation in position, in all light and heavy chain promoters of elasmobranchs, fish, amphibians, chickens, and mammals (Anderson et al., 1995; Haire et al., 1991; Hohman et al., 1993; Rast et al., 1994; Reynaud et al., 1987, 1989;
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
479
Schwager et al., 1988; Warr, 1995; Wilson and Warr, 1992). The one exception is the elasmobranch VHpromoters (Anderson et al., 1994; Kokubu et al., 1988b).Although a TATA box is present in elasmobranch VHgenes, an octamer segment cannot be identified. Instead a decamer-spacer nonomer motif that has been identified as a regulatory element in mammalian TCR V/3 promoters (Litman et al., 1993) appears to be present in horned shark VH sequences. Curiously, this same element does not appear to be part of the horned shark TCR V/3 promoters (Litman and Rast, 1996). Other than the observed octamer motif, enhancers and promoters are not well characterized in lower vertebrates. Sequence comparisons with known DNA enhancer elements have identified putative elements in the J-Cp intron region gene of skates. Although some are 100% matches (Anderson et al., 1994), the functional significance of this observation is not known. Even if this identification is correct, the mode of action of this putative enhancer must be fundamentally different from that in higher vertebrates because in sharks this region is always in close proximity to the promoter. The only functional testing of promoter and enhancer elements in lower vertebrates has been performed in teleost fish (Ledford et al., 1993; Magor et al., 1994a; Pilstrom and Bengten, 1996). In the catfish, the enhancer has been localized to a 1.7-kb fragment at the 3’ end of the gene. This region includes the TM2 exon and is adjacent to the first 6 exon. In a like manner to the mammalian p intron enhancer, the catfish enhancer directed B-cell-specific expression (Magor et al., 1994a). Crossspecies activity of enhancer/promoter elements in catfish and mouse cells indicates strong evolutionary conservation of the trans-acting regulatory factors that interact with the DNA enhancer and promoter elements. For example, the catfish enhancer mediated B-cell-specific transcription in mouse cells when paired with either a catfish or a mouse VH promoter (Magor et al., 1994a). Murine V K and p enhancers function in B cells from trout (Michard-Vanhee et al., 1994). This activity contrasts with mammalian light chain enhancers, which do not appear to be functional among other mammalian species (Blomberg et al., 1991; Hole et al., 1991). The cross-species activity of the teleost enhancer may reflect the redundancy of enhancer structure or may suggest it as a candidate for a primordial enhancer. E boxes, octamers, and p B mammalian heavy chain enhancer elements are found in the catfish enhancer region, although they are spread over a larger distance (Magor et al., 1997)and some of the putative elements do not contribute to transcriptional activation (Magor et al., 1997). In contrast to the conservation of the basic genetic elements, the arrangement of these elements in the genome has displayed remarkable plasticity throughout phylogeny. Ig gene organization can be classified into two basic types: the translocon arrangement and the cluster arrangement. The human
480
JOHN J. MARCHALONIS et rd.
K locus (Fig. 26) shows the basic properties of a translocon. There are a large number of Vs, several Js, and one constant region. A functional gene is formed by the rearrangement of one V with one of the Js. Antigen specificity is determined by sequences in CDRl and CDR2 in the V region and CDRS formed by the junction of V and J segments (Max, 1993). The joining of the same V with a different J creates a different CDR3 and thus a different antigen specificity. The heavy chain locus (Fig. 26) is very similar, except that CDRS diversity is significantlyincreased by D (diversity) elements. Thus this system has a large inherent capacity for primary diversity generated by the large numbers of genomic elements and the variation in combination of these elements. Joining of segments is not precise and the rearrangement process deletes and/or adds bases at the junctions, thereby amplifylng the degree of diversity. Each locus in most species has approximately 100 Vs, although 50% of the VH genes in humans are nonfunctional (Tomlinson et al., 1992),which may also be the case in other species. Curiously, in humans there are 24 VH regions located on other chromosomes (Tomlinson et al., 1992). These are nonfunctional as there is no associated C region. Another feature of heavy chain translocons is isotype switching. The various isotype heavy chain genes are located downstream of the C p gene (Fig. 26). A switch region located in the J-Cp intron rearranges to one of the switch regions located upstream of each C region, and the intervening DNA is deleted (Snapper et al., 1997). Thus, the same VDH segment is transcribed with a different constant region, resulting in a switch of the isotype secreted. The macroorganization of the conserved immunoglobulin elements is radically different in elasmobranchs. The Ig segments occur in individual cassettes or clusters, each containing a V, two Ds in heavy chains, a J, and one constant region (Fig. 27). Heavy chain clusters are approximately 16 kb in length, and light chain clusters are less than 10 kb. V-D-J introns are 300-500 bases and the J-C introns range from 1to 9 kb. In fundamental contrast to higher vertebrates, many of the clusters contain elements that are already rearranged in the germline. In the sandbar shark, all of the A light chain genes (essentiallythe only type found in serum immunoglobulin) are fused in the germline (Hohman et al., 1993). Similarly, all of the type I light chains in skate are fused (Anderson et al., 1995). However, the genes of the major expressed light chains in homed shark are not fused (Shamblott and Litman, 1989b). About 50% of heavy chain loci contain fused elements (Harding et al., 1990; Kokubu et al., 1988b). The p chain loci generally contain two D elements and three fusion patterns are observed: VD-J, VDJ, and VD-DJ (Fig. 27) (Harding et al., 1990; Kokubu et al., 198813).
EMEKGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
Light chain
481
7
unfused
fused
-1.5kb
1-4kb
Heavy chain 7
unfused
A I\
-2.5kb
fused
+q +%
b
4-10Kb
rn]
n
?
7
3
A
:;
-4-13"kb
-
(I D D D J
C
A
W D &
n
-1 kb
FIG.27. Gerinline cluster type arrangement of immunoglobulin genes in chondrichthytes. The linkage relationships and distance between clusters are unknown. Clusters may even he distributed on different chromosomes.
Typical rearrangement apparently occurs in the unfused loci, but apparently only within clusters (Hinds-Frey et al., 1993). The pattern of RSS arrangement is such that VDlD2J or VDSJ, but not VDIJ, rearrangements can occur. Thus, combinatorial diversity is comparatively low in elasmobranchs. However, junctional diversity generated during the rearrangement process is comparable to that of higher vertebrates (Kokubu et al., 1988b). The D elements can be read in all three reading frames, which also increases CDR3 diversity (Hinds-Frey et al., 1993). Furthermore, CDR3 diversity in the fused germline genes of heavy and light chains is extensive and basically indistinguishable from that generated by normal rearrangement processes (Anderson et al., 1995; Hohman et al., 1993). The number of clusters is high, approximately 100-200 for both light and heavy chains, and examination of VH sequences, both expressed (cDNA) and germline, shows a pattern of diversity in CDRl and CDR2 very reminiscent of somatic mutation in higher vertebrates (Anderson et al., 1995; Hohman et al., 1993; Kokubu et al., 1988b). Although at first glance
482
JOHN J. MARCHALONIS et al.
the cluster type arrangement suggests limited diversity, the actual diversity generated is very high in elasmobranchs. Teleost fish use both the translocon arrangement and the cluster type arrangement (Fig. 28). The single heavy chain locus is essentially the same as the mammalian translocon (Pilstrom and Bengten, 1996; Warr, 1995). Approximately 100VHgenes belonging to four to nine VHfamilies (Amemiya and Litman, 1990;Anderson and Matsunaga, 1995a; Bengten, 1994; Bengten et al., 1994; Ghaffari and Lobb, 1991; Roman and Charlemagne, 1994; Ventura-Holman et al., 1994; Warr et al., 1991) are interdispersed in the locus. The average distance between genes of 3 kb is much shorter than in the human locus. As is found in the mammalian locus, the IgD heavy chain gene is adjacent to the p gene (Fig. 28) (Wilson et al., 1997) and is expressed through RNA processing (Wilson et al., 1997). Light chain genes are organized in clusters in a manner similar to that seen in elasmobranchs (Bengten et al., 1994; Daggfeldt et al., 1993; Ghaffari and Lobb, 1993; Schwager et al., 1996). At least two types of light chain isotypes are present in teleost fish (Fig. 28). These have been well characterized in the catfish whose loci for the two types are organized in slightly different
TELEOSTEI Light chain cluster
3
A - -
G - tvnn -,I--
10-15 kb
rskb
7-1Okb
2kb
Heavy chain translocon
--D1-?
-400 kb
>I0 kb
JHI-9
Cp
2.2 kb 1.8 kb
TMI-TM2 C6
-6 kb
CHONDROSTEI Light chain translocon
_IV:..r 7
-3.5 kb
FIG.28. Cluster and translocon arrangements present in teleost and chondrostian fishes. The catfish gene arrangements are shown for the teleostei and sturgeon for chondrostei,
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
483
manners. The G isotype has one constant segment and one J segment but is associated with at least three V regions (Ghaffari and Lobb, 1993).Two V segments are 5’ of the C segment and one is 3 ’. Interestingly, these V regions are in opposite transcriptional orientation (Ghaffari and Lobb, 1993, 1997), a feature which distinguishes fish and elasmobranch clusters. The only functional significance of this orientation is that rearrangement would occur by inversion rather than deletion (Max, 1993). These loci were characterized using phage A clones and so there is the possibility that the 3’ V segment found on the approximately 20-kb genomic fragments is the proximal 5’ V segment belonging to the next downstream cluster. In which case, the distance between clusters could be as little as 7 kb. The F type light chain clusters are strikingly similar to shark clusters except that the V segments are in opposite transcriptional orientation (Ghaffari and Lobb, 1997). However, these clusters are only approximately 9 kb apart (Fig. 28), raising the possibility that intercluster rearrangement may occur. There are greater than 50 F clusters and 15 G clusters (Ghaffari and Lobb, 1997), numbers consistent with estimates of light chain genes in other species (Pilstrom and Bengten, 1996). Sturgeons constitute a branch of fish between sharks and teleosts and may represent a transitional species. Therefore, it is interesting that the light chain gene in sturgeons (Fig. 28) have a mammalian K-like translocon arrangement with many V segments and at least seven J segments upstream of the constant region (Lundqvist et al., 1996). This result suggests that the translocon organization arose early in evolution after the elasmobranchs and, therefore, the light chain cluster pattern in teleost fish was independently derived. Ig gene organization in amphibians appears to fit the typical mammalian translocon model. This has been best demonstrated in the frog Xenopus (Haire et al., 1991, 1996; Picker and Siegelman, 1993; Schwager et al., 1988, 1991). Three isotypes, IgM, IgY, and IgX, are found in Xenopus. Isotype switching from IgM to IgY is observed during an immune response (Du Pasquier, 1993; Du Pasquier et al., 1989),and the IgX heavy chain has been shown by Southern analysis to be linked to the p gene (Mussmann et al., 1996). Thus, the Y and IgX heavy chain genes appear to be downstream of the p gene in a normal translocon pattern and to be expressed by isotype switching. However, the exact position of these genes has not been fully characterized. Some species have a modification of the basic translocon arrangement, which is characterized by a reduced capacity for combinatorial diversity, and therefore can be classified as limited or restricted translocon (Fig. 29). These translocons have one J, limited numbers of Ds in heavy chains, and low numbers of functional V regions (Butler, 1997; Butler et al., 1996).
484
JOHN J. MARCHALONIS et nl.
PIG
Heavy chain
v
-
%
i
i
3 s
?
CHICKEN Heavy chain -80pseudo VHS
wwvww
wwwwww
-tttrH4
VH’
b
60-80 kb
3.5kb -12kb 200 bp
Light chain -25 pseudo V y
w w w v
w w w w w
I
4-H-HM-l-H: L
4
Ch &-
I
I
b-+-b
A
* -
-20 kb
Jh
Vhl
2.8 kb
1.8kb 1.6 kb
FIG.29. The limited translocon type arrangement is illustrated with examples from the pig and chicken. The arrangement and order of heavy chain isotype genes shown in brackets are unknown. $I indicates pseudogenes.
An extreme example of limited translocon are chicken heavy and light chain genes (Reynaud et al., 1987, 1989). Only one functional V region is found a short distance upstream of J (Fig. 29). However, there are numerous pseudo V regions in the locus, and a large secondary repertoire is generated by templated mutations (gene conversion from these pseudogenes). The IgY and IgA heavy chain genes are probably located downstream of the p gene, although the exact organization is not known. An intermediate type of limited translocon is present in the pig (Fig. 29). There are less than 20 VH genes, which all belong to the same family (Butler et al., 1996). Newborn piglets show a preferential usage of VH and DHgenes (Sun and Butler, 1996), and most of the repertoire is generated by gene conversion and hypermutation. Similar patterns are found in rabbits, sheep, and cattle (Dufour et al., 1996; Dufour and Nau, 1997; Hedrick and Eidelman, 1993). A rabbit VHlocus, for example, has 200 VH
EMERCENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
485
genes, all belonging to the same VHfamily, of which half are pseudogenes (Knight and Crane, 1994; Knight and Winstead, 1997). In about 90% of rearrangements, the proximal VHis utilized and diversity is generated by gene conversion (Berens et al., 1997; Knight and Crane, 1994; Knight and Winstead, 1997). Generation of dwersity by gene conversion and hypermutation is a common feature of animals with the limited type of translocon. The distinguishing feature of the immune system in these animals is that the primary repertoire is generated in gut-associated lymphoid tissue rather than in the bone marrow (Weill and Raynaud, 1996). TCR loci in mice and humans have a translocon arrangement similar to the A light chain loci (Hedrick and Eidelman, 1993). Large numbers of diverse V regions are located upstream of J(D)C cassettes, which may be duplicated in tandem array. For example, in the mouse p chain locus there are two copies of the DJC cassette and four copies of the y JC cluster. Although it is now clear that TCR genes are present in all jawed vertebrates, most work has been done with cDNA and there is very little characterization of genoinic loci structure (Chretien et al., 1997; Fellah et al., 1993a; Hawke et al., 1996; Partula et al., 1995, 1996; Rast et al., 1995, 1997; Rast and Litman, 1994). Initial work in the horned shark suggested that the TCR p locus was in a cluster organization similar to the Ig genes (Rast and Litinan, 1994). However, subsequent work in the skate demonstrates that the TCR constant region exons are present in single copy, although a diverse array of V regions exists (Rast et al., 1997). This result suggests a translocon arrangement, and the results in the horned shark can also be interpreted to reflect a translocon arrangement (Litman and Rast, 1996). XI. Molecular Events Underlying the Explosive Emergence of Immunoglobulins and Their Initial Phases of Evolution
Possibly the most striking feature of the vertebrate combinatorial immune system is its apparent origin as a single burst with the emergence of gnathostomes. Agnathans have complement components and apparent lymphocytes, but all attempts thus far to detect genes for Igs, TCRs, MHC, or RAG proteins have been unsuccessful. Although absence of proof is not proof of absence, these negative results are puzzling because all of these genes that are requisite for the function of the combinatorial immune system are present in the most primitive extant jawed vertebrates, the chondrichthyes. These molecules in chondrichthytes comprise a diverse set but are clearly homologous to their counterparts in mammals. Thus, if the coinbinatorial immune system arose following the ancestral divergence of agnathans and chondrichthytes, it occurred as an explosive event within a relatively short evolutionary span of 20 million years or less.
486
JOHN J. MARCHALONIS et ol.
A speculative model for the emergence of the antigen-specific elements of the vertebrate immune system is shown in Fig. 30. The precursor gene ancestral to both V and C genes duplicated to generate these segments, which remained closely linked but separated by an intron of 1-4 kb consistent with the separation distance within shark clusters (Hohman et al., 1993). It has been speculated that the ancestral Ig domain was related to cell adhesion molecules (Matsunaga and Mori, 1987), although these contain C2 domains rather than the bonajde Ig C1 domains found in MHC and immunoglobulins. The actual ancestor has not been identified and the search must continue in agnathans, protochordates, and deuterostomes. Following the generation of the V-intron-C cassette, the C domain duplicated to generate C1 domains that translocated out of this cluster and evolved into the closely related MHC Ig-C domains and p2microglobulins (see Fig. 3). The next step was the insertion of a J minigene and recombination signal sequences into the intron separating the V and C gene segments. The origin of these is unknown. Once this insertion had taken place, the Vrss-rssJ-C cluster could be acted on by products of RAG genes to rearrange and to be expressed. It has been proposed that the segment of DNA containing RAG-1 and RAG-2 was introduced into the genome of primitive vertebrates by horizontal transfer mediated by retroviruses (Schatz et al., 1992).This hypothesis gained support from the homologies between RAG-1 and site-specific microbial recombinases (Bernstein et al., 1996a; Hughes and Yeager, 1997), by similarities in mechanisms of retroviral recombination and VDJ recombination (Dik et al., 1996), by findings of retrotransposons of the type lacking LTR in sharks, and by potential homology of the shark RAG-1 5'-untranslated region to a rat adenovirus right junction sequence (R. M. Bernstein, S. F. Schluter, and J. J. Marchalonis, unpublished observations). Although the homology to microbial recombinases does not by itself establish that horizontal transfer occurred, it may prove significant that these types of site-specific recombinases (integrases and integration host factors) are present in bacteria, but not in protostomes, lower deuterostomes, or agnathan vertebrates. Their appearance in gnathostomes is either the result of convergent evolution from other systems or the product of horizontal transfer. A number of examples from plants, eukaryotes, and lower deuterostomes provide support for the conjecture that transposable elements introduced a substantial amount of regulatory variation in evolution (Britten, 1996; Kidwell and Lisch, 1997).Whatever the explanation, this is a dear example of the combinatorial immune system coopting elements of other genetic systems. The Vrss-rssJ-C cluster at this stage resembles that of shark type I light chains (Rast et al., 1994). A second major type of light chains (h-like) that
I
VJ fused V
J
V
VJ separated C
Insertion of J minigene and HSS
I
I
C a l C&? C03 C& I
1 Prototype V-intron-C
J
C
J
V
D
-
1
-
*A
1
\ /
V/C Precursor lg domain (unknown)
Agnathadchondrichthyian Ancestor
FIG.30. Hypothetical scheme of the genetic events underlying the generation of the combinatorial immune system of jawed vertebrates. Solid lines represent Ig domains and J segments; transmembrane (TM) and secretory (S) exons are open boxes and solid ellipses, respectively; solid boxes are D segments; and shaded triangles are recombination signal sequences. Introns are indicated by wavy lines.
488
J O H N J. MARCHALONIS et al.
is dominant in carcharhines has the V and J segments fused in register in the germline (Hohman et al., 1993). The orthologous light chains of rays are also fused in the germline (Anderson et al., 1995). The authors suggest that the VJ fusion was a secondary event at this stage. The ancestral heavy chain gene cluster was formed by a duplication of this cassette followed by the tandem duplication of the C segment into four closely linked domains that evolved into the C p l , Cp2, Cp3, Cp4 structure. In addition, a diversity (D) segment utilizing comparable RSS was inserted between the V and the J segments. The primordial light and heavy chain clusters duplicated subsequently to form hundreds of similar cassettes with mutation and selection of V, J, and C segments occurring. The process was a rapid one in that the characteristic distinctions in these segments apparently became stabilized in the 20 million year period followingthe divergence of gnathostomes and ancestral chondrichthytes. It can be proposed that this original burst of duplication and mutation was an explosive event, but selective pressures involved in forming heterodimers and in recognition of antigen resulted rapidly in the stabilization of canonical sequences followed by the domains evolving at moderate constant rates after their appearance in primordial form in the chondrichthytes. Another crucial event was the incorporation of the transmembrane/cytoplasmic exon(s) into the heavy chain cluster that enabled the (lightheavy) 2 structure to serve as a membrane receptor for antigen on B cells. The emergence of the full range of TCR chains likewise preceded the appearance of the ancestors of contemporary elasmobranchs and mammals, approximately 430 million years ago (Rast et al., 1997).These developments for a,0, -y, and 6 gene clusters incorporated the insertion of TM/CYTO exons for all the chains and also D segments into the P and 6 gene array in positions similar to their occurrence in heavy chains. Although it has been proposed that membrane associated TCRs arose prior to secreted immunoglobulins, the alternative should also be considered. Early cladistic analyses of Ig and TCR V-C segments indicated that TCR V domains could have diverged from the light chains following their separation from heavy chains (Beaman et al., 1987). Quantitative comparisons of the rates of evolution of TCR and Ig C domains from sequence data presented earlier indicate that the average rate of TCR C domain evolution, 2.8 0.2 X lo-', was significantly faster than that of Ig light and p chain C domains, 1.1 ? 0.3 X lo-'. Assuming that the rates for each are constant and that 20% identity is the cutoff point for positive identification ofbonufide Ig C domains, TCR C domains appeared between 420 and 540 million years ago, but Ig C domains emerged approximateIy 1000 million years ago. The latter figure (900-1290 million years) is also supported by consideration of the rate of evolution of C p domains, C p l ,
*
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
489
Cp2, Cp3, Cp4, within individual species, including humans, mice, chickens, and sharks. Interestingly, these calculations suggest that TCR arose at approximately the time of the origm of gnathostomes but that the emergence of recognizable Ig C domain precursors preceded this event in evolution. Although the early emergence of TCR as cell surface receptors has been an attractive h,ypothesis, the prior appearance of Ig light and heavy chains is not unreasonable because at least two additional systems-proteosomes and MHC-would have to evolve or be coopted for functional antigen-specific T-cell immunity to arise. The possibilities must be considered either that refinements in the computations will yield Ig values comparable to those for TCR or that, in line with their estimated ages, bona$de Ig constant domains will be found in agnathans, protochordates, or deuterostomes. If we accept the hypothesis that the cluster or cassette organization of Ig segments characteristic of chondrichthtylans is the primordial arrangement, it is not difficult to envision mechanisms for the origin of distinct immunoglobulin isotypes and also to explain why all immunoglobulin domains seem to show 30-40% identity to one another when considered over large evolutionary distances. Possibly in the development of an ancestral osteichthyian species, the chromosomal arrangement was such that the entire array of elasmobranch type clusters was not transmitted to the newly formed species. In the simplest case, one light chain cluster and one heavy chain cluster might be inherited. The heavy chain gene locus would be a p chain, and the light chain could be K-like, h-like, or one of the other types. The transferred clusters probably had the V and J separated with RSS sequences because rearrangement is an essential part of immunoglobulin and TCR activation in species more advanced than chondrichthytes. Light chain clusters occur in teleost, but these contain two VL segments, thereby indicating tandem duplication of these gene segments. Furthermore, in chondrostian fishes such as the sturgeon, an array of distinct VI, genes comparable to that found in mammalian light chain translocons occurs (Lundqvist et al., 1996). The heavy chains of all higher vertebrate species also occur in translocon arrangements. Thus, a second event that followed the inheritance of a p chain cluster was the duplication of variable region segments with these remaining linked to the D, J, and C segments, although these arrangements can cover hundreds of kilobases (Lai et al., 1989). Individual light chain isotypes might have arisen depending on the number of gene clusters inherited from the chondrichthtyian ancestors or, alternatively, they may have arisen by duplication within the phylogenetic development of individual vertebrate classes. All of the vertebrates derived from the ancestral chondrichthtyans would possess IgM heavy chains, and these would be expected to show an evolu-
490
JOHN J. MARCHALONIS et al.
tion in p chains consistent with time and speciation. Heavy chains distinct from the p chain would arise via duplications of the C p cluster and subsequent independent evolution. The o (Bernstein et al., 19968) or NARC (Greenberg et al., 1996) heavy chains of sharks occur in clusters separate and unlinked to those of the p, chain and apparently have not been passed on to higher vertebrates in evolution (Schluter et al., 1997). It is possible that non-p heavy chains forming soluble immunoglobulin dimers of the form (LH)zmay have arisen independently in each vertebrate class (AtweU and Marchalonis, 1975) or, alternatively, there may be a lineal descent of these molecules to some extant in mammals. For example, Warr et al. (1995) argue that the heavy chain of IgY may have been ancestral to mammalian y and E heavy chains. The TCR chains appear to have occurred in stable configuration by the time of the ancestral divergence between chondrichthytes and osteoichtyes and have been maintained in substantially similar form throughout subsequent vertebrate evolution. A major event in the evolution of immunoglobulins was the appearance of IgG immunoglobulins in eutherian mammals, if not all mammals (Marchalonis, 1977), and the capacity for affinity maturation following somatic mutation and selection occurring in B cells in the germinal centers of lymph nodes (Han et al., 1996; Liu et al., 1992). This event required the appearance of histological structures, the emergence of IgG as the dominant immunoglobulin class, and the reappearance of RAG genes to facilitate recombination in adult cells. Despite the unequivocal homologies among, respectively,the V domains, C domains,joining segments, and transmembrane segments in Igs and TCRs of gnathostomes, there has been a tremendous plasticity in the organization of these segments and in the relative dependence on different mechanisms for the generation of diversity in both primary and secondary responses. The chicken, for example, has both a A light chain and a heavy chain gene system that are degenerate forms of the translocon in which there is only one functional Vh or VHand one functional JHor Jh, but a number of pseudogenes in a tandem array with these. Diversity is generated by templated hypermutation (gene conversion) using these pseudogenes (Reynaud et al., 1987, 1989). Primates and rodents have comparable translocon arrangements in both their immunoglobulin and their TCR gene arrangements. In the immunoglobulins,they have relatively large numbers of V regions and these are combinatorially assorted to give diversity. In addition, somatic mutation and selection occur, which are reflected with affinity maturation in the secondary response. The rabbit has a large array of VHgenes comparable in magnitude to those of humans and mice, but utilizes predominantly the VH segment most proximal to the D elements and copies by gene conversion from the other VHsegments (Knight and Crane, 1994).
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
491
Heavy chains of artiodactyls such as pigs (Butler et al., 1996) and sheep (Reynaud et aE., 1991) have only a single J H segment and a small number of VHgenes. The pig depends on gene conversion for generation of dversity, and the sheep relies heavily on nontemplated somatic hypermutation. In broad perspective, the combinatonal immune system of gnathostomes arose in a burst of gene duplication and cooption of genes and mechanism of cell differentiation from other systems and was established prior to the divergence of ancestral chondrichthtians and higher vertebrates. The immunoglobulins underwent evolutionary changes at a relatively conserved rate by comparison with other protein families, and variation in mechanisms for the generation of diversity indicated great plasticity in the organization of the gene segments and a requirement for special processes such as junctional diversification, somatic hypermutation, or gene conversion. XII. Conclusions
The combinatorial immune system defined by the presence of antigenspecific recognition units from the immunoglobulin family (heterodimeric immunoglobulins and T-cell receptors), the genetic machinery necessary for recombination, and cells of the lymphoid series expressing these receptors is fully functional from the earliest extant gnathostomes (sharks and their kin) to mammals. Despite efforts by a number of workers, definitive evidence for genes specifylng immunoglobulins, T-cell receptors, or recombination activating genes has not been documented for more primitive agnathan vertebrates (lampreys and hagfish), for lower deuterostomes such as tunicates and starfish, or for protostome or acoelomate invertebrates. Light chain type molecules form a large and diverse array within cartilaginous fishes, and the complete panoply of T-cell receptor chains (a,p, y , S) has been identified in these species. The combinatorial immune system of jawed vertebrates apparently arose as an evolutionary “big bang” involving the generation and duplication of V and C domains from an unknown precursor and the incorporation of joining segment genes, recombination signal sequences, and transrnembrane/cytoplasmic segments within the brief evolutionary span of 10-20 million years. Mechanisms for cell activation and division were coopted from existing systems widespread in evolution that can be termed inflammatory mechanisms. Orthologs of light chain variable segments may be shared among sharks, mice, and humans, and orthologous relationships among T-cell receptor variable segments of cartilaginous fishes and higher vertebrates have been found. Constant domains of K and y light chains and p heavy chains and Cn, Cp, C y , and CS TCR chains have been identified in diverse vertebrate species. From presently available sequence data, these most probably
492
JOHN J. MARCHALONIS et al.
evolve at constant rates, with the rate of evolution of TCR C domains being more rapid than that of the constant domains of immunoglobulins. Despite the clear homology among members of the immunoglobulin family, particularly among orthologous V domain sequences, there is a marked plasticity in the organization of gene segments; the most extreme differences occur between those of sharks, which are distributed as individual clusters or cassettes, each of which contains a single VL,JL, and CLelement or a single VH,a few Ds, possibly a few JHs, and one C p as opposed to that of mammals where a large number of V segments and a few Js or Ds are associatedwith one CLor a set of heavy chain C segments in a translocon arrangement. Organizational flexibility can differ considerably even within a single vertebrate class such as mammals, where swine have relatively simple heavy chain translocons compared to those of humans or mice and depend on templated hypermutation (gene conversion), whereas primates and rodents generate diversity via rearrangement mechanisms. Chickens also have restricted light chain (A) and heavy chain translocons and utilize gene conversion to V segment pseudogenes in the generation of the primary antibody repertoire. The most recent major step in the evolution of the immune system was the emergence in mammals of germinal centers within the lymph nodes, correlating with the IgM to IgG switch and affinity maturation following from somatic mutation and antigenic selection. The y heavy chain definitive of IgG is present in examples of all mammals, but has not been identified in more primitive vertebrates. Distinct heavy chain isotypes such as IgY of chickens, reptiles, and amphibians and IgW of sharks have also arisen through gene duplication in evolution and these may show distant relationships to mammalian immunoglobulins such as IgD and IgE.
ACKNOWLEDGMENTS This work was supported in part by Grant MCB 9631846 from the National Science Foundation to JJM and Grant CA72803 from the National Cancer Institute, USPHS to ABE. We thank Ms. Diana Humphreys for valuable assistance in the preparation of the manuscript.
REFERENCES Amemiya, C. T., and Litman, G. W. (1990).Complete nucleotide sequence of an immunoglobulin heavy-chain gene and analysis of immunoglobulin gene organization in a primitive teleost species. Proc. Natl. Acad. Sci. USA 87, 811. Amemiya, C. T., Ohta, Y., Litman, R. T., Rast, J. P., Haire, R. N., and Litman, G . W. (1993). VH gene organization in a relict species, the coelacanth Latimeria chalumnae: Evolutionary implications. Proc. Natl. Acad. Sci. USA 90, 6661. Anderson, M., Amemiya, C., Luer, C., Litman, R., Rast, J., Niimura, Y., and Litman, G. (1994). Complete genomic sequence and patterns of transcription of a member of an unusual family of closely related, chromosomally dispersed Ig gene clusters in Raja. Znt. Zmmunol. 6 , 1661.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
493
Anderson, M. K., Shamblott, M. J., Litman, R. T., and Litman, G. W. (1995). Generation of immunoglobulin light chain gene diversity in Raja erinacea is not associated with somatic rearrangment an exception to a central paradigm of B cell immunity. /. Exp. Men. 182, 109. Andersson, E., and Matsunaga, T. (1993). Complete cDNA sequence of a rainbow trout IgM gene and evolution of vertebrate IgM constant domains. Zmniunogenetics 38, 243. Andersson, E., and Matsunaga, T. (19954 Evolution ofimrnunoglobulin heavy chainvariable region genes: A VH family can last for 150-200 million years or longer. Zozmunogenetic.9 41, 18. Anderson, E., and Matsunaga, T. (1995b). Evolutionary stability of iininunoglobulin heavy chain variable gene family: A VH family can last for 150-200 million years or longer. Inmunogenetics 41, 18. Andersson, E., Peixoto, B., Tormanen, V., and Matsunaga, T. (1995). The evolution of the immunoglobulin M constant regon genes of Salmonid fish, rainbow trout (Oncorhynchus mykiss) and Atlantic salmon (Soltielinus alpinus): Implications on the divergence time of species. Iminunogenetics 41, 312. Arden, B., Clark, S. P., Kabelitz, D., and Mak, T. W. (1995a). Human T-cell receptor variable gene segment families. lmtnunogenetics 42, 455. Arden, B., Clark, S. P., Kabelitz, D., and Mak, T. W. (1995b). Mouse T-cell receptor variable gene segment families. Zmmunogenetics 42, 501. Atwell, J. L., and Marchalonis, J. J. (1975). Phylogenetic emergence of immunoglobulin classes dstinct from IgM. J. Zmmunogenetics 1, 367. AtweU, J. L., and Marchalonis,J. J. (1976).In “Comparative Immunology” (J. J. Marchdonis, ed.), p. 276. Blackwell Press, Oxford. Atwell, J. L., Marchdonis, J. J., and Edey, E. H. M. (1973). Major immunoglobulin classes of the echidna (Tuchyglossus aculeatus). Zinmunolugy 25, 835. Ayme-Southgate, A., Vigoreauz, J., Benian, G., and Pardue, M. L. (1991). Drososphila has a twitcllidtitin-related gene that appears to encode projectin. Proc. Natl. Acad. Sci. USA 88, 7973. Baeuerle, P. A,, and Henkel, T. (1994). Function and activation of NK-KB in the immune system. Annu. Reu. Zmmurzol. 12, 141. Barker, W. C., Ketcham, L. K., and Dayhoff, M. 0. (1978). In “Atlas of Protein Sequence and Structure” (M. 0. Dayhoff, ed.), p. 197. National Biomedical Research Foundation, Washington, DC. Bartl, S., Baltimore, D., and Weissman, I. L. (1994). Molecular evolution of the vertebrate immune system. Proc. Natl. Acad. Sci. USA 91, 10769. B a d , S., and Weissman, I. L. (1994).Isolation and characterization of major histocompatibility complex class IIB genes from the nurse shark. Proc. Natl. Acad. Sci. USA 91, 262. Beaman, K. D., Barker, W. C., and Marchalonis, J. J. (1987). In “Antigen Specific T Cell Receptors and Factors” (J. J. Marchalonis, ed.), p. 105. CRC Press, Boca Raton, FL. Beard, J. (1894).The development and probable function of the thymus. Anat. Amy. 9,476. Beck, G., Cooper, E. L., Habicht, G. S., and Marchalonis,J. J. (1994). “Primordial Immunity: Foundations for the Vertebrate Immune System.” New York Academy of Sciences, New York. Beck, G., Vasta, G. R., Marchalonis, J. J., and Habicht, G. S. (1989). Characterization of interleukin-a activity in tunicates. Coinp. Biochem. Physiol. B 92, 93. Bell, J. I., Owen, M. J., and Simpson, E. (1995). “T Cell Receptors.” Oxford University Press, Oxford. Bengten, E. (1994). “The Immunoglobulin Genes in Atlantic Cod (Gadus morhua) and Rainbow Trout (Oncorhynchus ntykiss).” Sweden Uppsda University, Uppsda, Sweden.
494
JOHN J , MARCHALONIS et al.
Bengten, E., Stromberg, S., and Pilstrom, L. (1994). Immunoglobulin VH regions in Atlantic cod (Gadusmorhua L,):Their diversity and relationship to VH families from other species. Dev. Comp. Immunol. 18, 109. Benian, G. M., Kiff, J. E., Neckslmann, N., Moermann, D. F., and Waterston, R. H. (1989). Sequence of an unusually large protein implicated in regulation of myosin activity in C. elegans. Nature (London)342, 45. Bentley, G. A., Boulot, C . , Karjalainen, K., and Mariuzza, R. A. (1995). Crystal structure of the P chain of a T-cell antigen receptor. Science 267, 1984. Bentley, G. A,, and Mariuzza, R. A. (1996). The structure of the T cell antigen receptor. Annu. Rev. Immunol. 14,563. Berens, S. J., Wylie, D. E., and Lopez, 0. J. (1997). Use of a single VH family and long CDR3s in the variable region of cattle Ig heavy chains. Int. Immunol. 9, 189. Bernstein, R. M., Schluter, S. F., Bernstein, H., and Marchalonis, J. J. (1996a).Primordial emergence of the recombination activating gene 1 (RAG1): Sequence of the complete sharkgene indicates homology to microbial integrases. Proc. Natl. Acad. Sci. USA 93,9545. Bernstein, R. M., Schluter, S. F., Lake, D. F., and Marchalonis, J. J. (1994). Evolutionary conservation and molecular cloning of the recombinase activating gene 1. Biochem. Biophys. Res. Commun. 205, 687. Bernstein, R. M., Schluter, S. F., and Marchalonis, J. J. (1997). In “The Physiology of Fishes” (D. H. Evans, ed.), p. 215. CRC Press, Boca Raton, FL. Bernstein, R. M., Schluter, S. F., Shen, S. X., m d Marchalonis, J. J. (1996b). A new high molecular weight immunoglobulin class from the carcharhine shark: Implications for the properties of the primordial immunoglobulin. Proc. Nutl. Acad. Sci. USA 93, 3289. Blomberg, B. B., Rudin, C. M., and Storb, U. (1991). Identification and localization of an enhancer for the human A L chain Ig gene complex. J. Immunol. 147, 2354. Boman, H. G. (1997). Peptide antibiotics: Holy or heretic grails of innate immunity? Scund. J. Immunol. 28, 2. Bork, P., Holm, L., and Sander, C. (1994).The immunoglobulin fold: Structural classification, sequence patterns and common core. J. Mol. Biol. 242, 309. Britten, R. J. (1996).DNA sequence insertion and evolutionaryvariation in gene regulation. Proc. Natl. Acad. Sci. USA 93, 9374. Burnett, R. C., Hanley, W. C., Zhai, S., and Knight, K. L. (1989). The IgA heavy chain gene family in rabbit: Cloning and sequence analysis of 13 C a genes. EMBO J. 8,4047. Butler, J. E. (1997). Immunoglobulin gene organization and the mechanism of repertoire development. Scand. J. Immunol. 45, 455. Butler, J. E., Sun, J., Kacskovics, I., Brown, W. R., and Navarro, P. (1996). The VH and CH immunoglobulin genes of swine: Implications for repertoire development. Vet. Immunol. Immunopathol. 54, 7. Butler, J. E., Sun, J.-S., and Navarro, P. (1996).The swine Ig heavy chain locus has a single JH and no identifiable IgD. Int. Immunol. 8, 1897. Capra, J. D.,and Kehoe, J. M. (1974).Variable region sequences of five human immunoglobulin heavy chains of the VHIII subgroup: Definitive indication of four heavy chain hypervariable regions. Proc. Natl. Acad. Sci. USA 71, 845. Carayannopoulos, L., and Capra, J. D. (1993).In “Fundamental Immunology” (W. E. Paul, ed.), p. 283. Raven Press, New York. Carroll, R. L. (1988). “Vertebrate Paleontology and Evolution.” Freeman, New York. Casson, L. P., and Manser, T. (1995). Evaluation of loss and change of specificity resulting from random mutagenesis of an antibody VH region. J. Immunol. 155, 5647. Chothia, C., Boswell, D. R., and Lesk, A. M. (1988). The outline structure of the T-cell CUPreceptor. EMBOJ. 7, 3745.
EMERGENCE AND EVOLUTION OF T H E IMMUNOGLOBULIN FAMILY
495
Chothia, C., Lesk, A. M., Gherardi, E., Tomlinson, I. M., Walter, G., Marks, J. D., Llewelyn, M. B., and Winter, G. (1992). Structural reportoire of the human H segments. 1. Mol. Biol. 227, 799. Chretien, I., Marcuz, A., Fellah, J., Charlemagne, J., and Du Pasqnier, L. (1997). The T cell receptor P genes of Xenopus. Eur. J. Immi~nol27, 763. Clark, S. P., Arden, B., Kabeliz, D., and Mak, T. W. (1995). Comparison of human and mouse T cell receptor variable gene segment subfamilies. Iinmunogenetics 42, 531. Clem, L. W., and Small, P. A. (1967).Phylogeny of immunoglobulin structure and fnnction. I. Immunoglobulins of the lemon shark. /. Exp. Med. 125, 893. Compagno, L. J. V. (1988).“Sharks of the Order Carcharhiniformes.” Princeton Univ. Press, Princeton, NJ. Cooper, M. D., Chen, C.-H. L., and Lucy, R. P. (1991). Avian T-cell ontogeny. Adu. Inimunol. 50, 87. Daggfeldt, A., Bengten, E., and Pilstrom, L. (1993). A cluster type organization of the loci of the immunoglobulin light chain in Atlantic cod (Gadus morhua L.) and rainbow trout (Oncorhynchus mykisswalbaurn) indicated by nucleotide sequences of cDNAs and hybridization analysis. lwimunogenetics 38, 199. Dahan, A,, Reynaud, C. A., and Weill, J. C. (1983). Nucleotide sequence of the constant region of a chicken p heavy chain immunoglobulin mRNA. Nucleic Acid Res. 11, 5381. Davidson, E. H., Peterson, K. J., and Camerson, R. A. (1995). Origin of bilaterian body plans: Evolution of development regnlatory mechanisms. Science 270, 1319. Davis, M. M., and Bjorkman, P. J. (1988).T-cell antigen receptor genes and T-cell recognition. Nature 334, 393. Dik, C. V. G., Mizuuchi, K., and Gellert, M. (1996). Similaritiesbetween initiation of V(D)J recombination and retroviral integration. Science 271, 1592. Domiati-Saad, R., Attrep, J. F., Brezinschek, H.-P., Cherrie, A. H., Karp, D. R., and Lipsky, P. E. (1996). Staphylococcal enterotoxin D fimctions as a human B cell superantigen by rescuing VH4-expressing B cells from apoptosis. J. Immunol. 156, 3608. Donelson, J. E. (1995). Mechanisms of antigenic variation in Borrelia hennsii and African trypanosomes. 1.Biol. Chem. 270, 7783. Doolittle, R. F. (1990). “Methods in Enzymology.” Academic Press, San Diego. Doolittle, R. F. (1994). Convergent evolution: The need to be explicit. Trends Biochein. Sci. 19, 15. Doolittle, R. F. (1995).The multiplicityof domains in proteins. Annu. Reu. Biochem. 64,287. Doolittle, R. F., Feng, D. F., Anderson, K. L., and Alberro, M. R. (1990). A naturally occurring horizontal gene transfer from a eukaryote to a prokaryote.]. Mol. Euol. 31,383. Dufour, V., Malinge, S., and Nau, F. (1996).The Sheep Igvariable region repertoire consists of a single VH family. 1.Imnmnol. 156, 2163. Dufour, V., and Nau, F. (1997). Genomic organization of the sheep immunoglobulin JH segments and their contribution to heavy chain variable region diversity. Iinmunogenetics 46, 283. Du Pasquier, L. (1993). In “Fundamental Immunology” (W. E. Paul, ed.), p. 199. Raven Press, New York. Du Pasquier, L., Schwager, J., and Flajnik, M. F. (1989). The immune system of xenopus. Annu. Rev. Imniunol. 7, 251. Edelman, G. M. (1987). CAMS and Igs: Cell adhesion and the evolutionary origins of immunity. Immunol. Reu. 100, 11. Edmundson, A. B., Ely, R. R., Abola, E. E., Schiffer, M., and Pagniotopoulos, N. (1975). Rotational allomerision and divergent evolution of domains in immunoglobulin light chains. Biocheinisty 14, 3933.
496
JOHN J. MARCHALONIS et al.
Ernst, P., and Smale, S. T. (1995a). Combinatorial regulation of transcription. I. General aspects of transcriptional control. Itnmunity 2, 311. Emst, P., and Smale, S. T. (1995b). Combinatorial regulation of transcription. 11. The immunoglobulin p heavy chain gene. Immunity 2, 427. Fellah. J. S., Kerfourn, F., Guillet, F., and Charlemagne, J. (1993a). Conserved structure of amphibian T-cell antigen receptor /3 chain. Proc. Nutl. Acud. Sci. USA 90, 6811. Fellah, J. S., Kerfourn, F., Wiles, M. V., Schwager,J., and Charlemagne, J. (199313).Phylogeny of immunoglobulin heavy chain isotypes: Structure of the constant region of Ambystonuz mexicanurn $ chain deduced from cDNA sequence. Iminunogenetics 38, 311. Fellah, J. S., Wiles, M. V., Charlemagne, J., and Schwager,J. (1992).Evolution ofvertebrate IgM: Complete amino acid sequence of the constant region of Ambystornu mericanumn p chain deduced from cDNA sequence. Eur. j . Zmmunol. 22, 2595. Felsenstein, J. (1981). Evolutionary trees from DNA sequences: A maximum likelihood approach. j . Mol. Evol 17, 368. Felsenstein, J. (1993). “PHYLIP (Phylogeny Influence Package).” Department of Genetics, University of Washington, Seattle, WA. Feng, D. F., and Doolittle, R. F. (1990). Progressive alignment and phylogenetic tree construction of protein sequences. Methods Enzymol. 183, 375. Friedlander, R. M., Nnssenzweig, M. C., and Leder, P. (1990). Complete nucleotide sequence of the membrane form of the human IgM heavy chain. Nucleic Acid Res. 18,4278. Gally, J. A,, and Edelman, G. M. (1972).The genetic control of immunoglobulin synthesis. Annu. Rev. Gene. 6, 1. Garcia, K. C., Degano, M., Stanfield, R. L., Brunmark, A,, Jackson, M. J.. Peterson, P. A., Teyton, L., and Wilson, I. A. (1996). An a/3 T cell receptor structure at 2.5 A and its orientation in the TCR-MHC complex. Science 274, 209. Ghaffari, S. H., and Lobb, C. J. (1991). Heavy chain variable region gene families evolved early in phylogeny: Ig complexity in fish. j . Iinmunol. 146, 1037. Ghaffari, S. H., and Lobb, C. J. (1992).Organization of immunoglobulin heavy chain constant and joining region genes in the channel catfish. Mol. Immunol. 29, 151. Ghaffari, S. H., and Lobb, C. J. (1993).Structure and genomic organizationof immunoglobulin light chain in the channel catfish. J. Inimunol. 151, 6900. Ghaffari, S. H., and Lobb, C. J. (1997). Structure and genomic organization of a second class of immunoglobulin light chain genes in the channel catfish.j . Immnunol. 159, 250. Gobel, T. W. F., Chen, C.-L. H., Lahti, J., Kubota, T., Kuo, C.-L., Aebersold, R., Hood, L., and Cooper, M. D. (1994). Identification of T cell receptor a-chain genes in the chicken. Proc. Natl. Acad. Sci. USA 91, 1094. Good, R. A., and Papermaster, B. W. (1964).Ontogeny and phylogeny of adaptive immunity. Ado. Immunol. 4, 1. Greenberg, A. S., Ada, D., Hughes, M., Hughes, A., McKinney, E. C., and Flajnik (1995). A new antigen receptor gene family that undergoes rearrangement and extensive somatic diversification in sharks. Nature 374, 168. Greenberg, A. S., Hughes, A. L., Cuo, J., A d a , D., McKinney, E. C., and Flajnik, M. F. (1996). A novel “chimeric” antibody class in cartilagenous fish: IgM may not be the primordial immunoglobulin. Eur. j . Immunol. 26, 1123. Greenberg, A. S., Steiner, L., Kasahara, M., and Flajnik, M. F. (1993). Isolation of a shark immunoglobulin light chain cDNA clone encoding a protein resembling mammalian K light chains: Implications for the evolution of light chains. Omnunology 90, 10603. Greenhalgh, P., Olsesen, C. E. M., and Steiner, L. A. (1993).Characterization and expression of recombination activating genes (RAG-I and RAG-2) in Xennpus laevis. J. lmmunol. 151,3100.
EMEKGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
497
Greenhalgh, P., and Steiner, L. A. (1995). Recombination activating gene 1 (RAG1) in zebrafish and shark. Immunogenetics 41, 54. Grey, H. M. (1967). Duck immunoglobulins. 11. Biologic and iminunochemical studies, J. Immunol. 98, 820. Grey, H. M. (1969). Phylogeny of immunoglobulins. A& Zmmunol. 10, 51. Haire, R. N., Ohta, Y., Litman, R. T., Amemiya, C. T., and Litman, G. W. (1991). The genomic organization of immunoglobulin VH genes in Xenopus laeuis shows evidence for interspersion of families. Nucleic Acid Res. 19, 3061. Haire, R. N., Ota, T., Rast, J. P., Litman, R. T., Chan, F. Y., Zon, L. I., and Litman, G. W. (1996). A third Ig light chain gene isotype in Xenopus laeuis consists of six distinct VL families and is related to mammalian A genes. J . Immunol. 157, 1544. Han, S., Zheng, B., Schatz, D., Spanopoulou, E., and Kelsoe, G. (1996). Neoteny in lymphocytes: Rag1 and Rag2 expression in germinal center B cells. Science 274, 2094. Hanley, P., Hook, J. W., Raftos, D. A,, Gooley, A. A,, Trent, R., and Raison, R. L. (1992). Hagfish humoral defense protein exhibits structural and functional homologywith mammalian complement components. Proc. Natl. Acad. Sci. USA 89, 7910. Harding, F. A,, Amemiya, C. T., Litman, R. T., Cohen, N., and Litman, G. W. (1990a). Two distinct immunoglobulin heavy chain isotypes in a primitive cartilaginous fish, Raja erinacea. Nucleic Acid Res. 16, 6369. Harding, F. A,, Cohen, N., and Litman, G. W. (1990b). Immunoglobulin heavy chain gene organuation and complexity in the skate, Raja erinacea. Nucbic Acid Res. 18, 1015. Hashimoto, K., Nakanishi, T., and Kurosawa, Y. (1992). Identification of a shark sequence resembling the major histocompatibility complex class I a 3 domain. Proc. Natl. Acad. Sci. USA 89, 2209. Hawke, N. A,, Rast, J. P., and Litman, G. W. (1996). Extensive diversity of transcribed TCR-P in a phylogenetically primitive vertebrate. J. Immunol. 156, 2458. Hayman, J. R., Ghaffari,S. H., and Lobb, C. J. (1993). Heavy chain joining region segments of the channel catfish: Genomic organization and phylogenetic implications.J . Immunol. 151, 3587. Hedrick, S. M., and Eidelman, F. J. (1993). In “Fundamental Immunology” (W. E. Paul, ed.), p. 383. Raven Press, New York. Hedrick, S. M., Neilsen, E.A., Cavalier, J., Cohen, D. I., and Davis, M. M. (1984). Sequence relationships between putative T-cell receptor polypeptides and immunoglobulins. Nature 308, 153. Hein, W. (1994). Structural and functional evolution of the extracellular regions of T cell receptors. Sem. Immunol. 6, 361. Hildemann, W. H. (1974). Some new concepts of immunological phylogeny. Nature 250, 116. Hill, R. L., Delaney, R., Fellows, R. E., and Lebovitz, H. E. (1966).The evohtionaryorigins of the immunoglobulins. Proc. Natl. Acad. Sci. USA 56, 1762. Hinds, K. R., and Litman, G. W. (1986). Major reorganization of immunoglobulin VH segmental elements during vertebrate evolution. Nature (London)320, 546. Hinds-Frey, K. R., Nishikata, H., htman, R. T., and Litman, G. W. (1993). Somatic variation precedes extensive diversification of germline sequences and combinatorial joining in the evolution of immunoglobulin heavy chain diversity.J . Exp. Med. 178, 825. Hoek, R. M., Smit, A. B., Vink, J, M., de Jong-Brink, M., and Geraerts, W. P. M. (1996). A new Ig-superfamily member, molluscan defence molecule (MDM) from Lyinnaea stagnalb is down-regulated during parasitosis. Eur. J. Immunol. 26, 939. Hoffman, J. A,, and Reichhart, J.-M. (1997).Drosophiln immunity. Trends Cell Biol. 7,309.
498
JOHN J. MARCHALONIS et al.
Hohman, V. S., Schluter, S. F., and Marchdonis, J. J. (1992). Complete sequence of a cDNA clone specifying sandbar shark immunoglobulin light chain: Gene organization and implications for the evolution of light chains. Proc. Natl. Acad. Sci. USA 89, 276. Hohman, V. S., Schluter, S. F., and Marchalonis, J. J. (1995). Diversity of Ig light chain clusters in the sandbar shark Carcharhinus plumbeus. J. lmmunol. 155, 3933. Hohman, V. S., Schuchman, D. B., Schluter, S. F., and Marchalonis, J. J. (1993). Genomic clone for sandbar shark A light chain: Generation of diversity in the absence of gene rearrangement. Proc. Natl. Acad. Sci. USA 90, 9882. Hole, N. J. K., Harindranath, N., Young-Cooper, G. O., Garcia, R., and Mage, R. G. (1991). Identification of enhancer sequences 3’ of the rabbit Ig K L chain loci. J. Immunol. 146,4377. Honjo, T., Shimizu, A., and Yaoita, Y. (1989). In “Immunoglobulin Genes” (T. Honjo, F. W. Alt, and T. H. Rabbitts, eds.), p. 123. Academic Press, New York. Hood, L., Rowen, L., and Koop, B. F. (1995). Human and mouse T-cell receptor loci: Genomics, evolution, diversity and serendipity. Ann. N.Y. Acad. Sci. 758, 390. Hordvik, I., Voie, A. M., Glette, J., Male, R., and Endresen, C. (1992). Cloning and sequence analysis of two isotypic IgM heavy chain genes from Atlantic salmon, S a l m salar L. Eur. J. Immunol. 22, 2957. Hsu, E., and Steiner, L. A. (1992).Primary structure of immunoglobulins through evolution. Cum. Opin. Struct. Biol. 2, 422. Huelsenbeck, J. P., and Rannala, B. (1997). Phylogenetic methods come of age: Testing hypotheses in an evolutionary context. Science 276, 227. Hughes, A. L., and Yeager, M. (1997). Molecular evolution of the vertebrate immune system. BioEssays 19, 777. Hultgren, S., Jacob-Dubuisson, F., Jone, C. H., and Branden, C. I. (1992). PapD and superfamily of periplasmic immunoglobulin-like pilus chaperones. Ado. Prot. Chem. 44,99. Ip, Y. T., Reach, M., Engstrom, Y., Kadalyil, L., Cai, H., Gonzalez-Crespo, S., Tatei, K., and Levine, M. (1993). D$, a dorsal-related gene that mediates an immune response in Drosophila. Cell 75, 753. Ishiguro, H., Kobayashi, K., Suzuki, M., Titani, K., Tomonoaga, S., and Kurosawa, Y. (1992). Isolation of a hagfish gene that encodes a complement component. EMBO J. 11, 829. Jin, D., Zhiliang, L., Jin, Q., Yuwen, H., and Hou, Y. (1989). Vaccinia virus hemagglutinin. J. Exp. Med 170, 571. Kabat, E. W., Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. (1991). “Sequences of Proteins of Immunological Interest.” USDHHS Public Health Service, National Institute of Health. Kasahara, M., McKinney, E. C., Flajnik, M. F., and Ishibashi, T. (1993). The evolutionary origin of the major histocompatibility complex: Polymorphism of class I1 a chain genes in the cartilaginous fish. Eur. J. Immunol. 23, 2160. Kasahara, M., Vazquez, M., Sato, K., McKinney, E. C., and Flajnik, M. F. (1992). Evolution of the major histocompatibilitycomplex: Isolation of a class I1 A gene from the cartilaginous fish. Proc. Natl. Acad. Sci. USA. 89, 6688. Kaymaz, H., Dedeoglu, F., Schluter, S. F., Edmundson, A. B., and Marchalonis, J. J. (1993). Reactions of anti-immunoglobulin sera with synthetic T-cell receptor peptides: Implications for the three-dimensional structure and function of the TCR /3 chain. lntl. Immunol. 5, 491. Kidwell, M. G., and Lisch, D. (1997). Transposable elements as sources of variation in animals and plants. Proc. Natl. Acud. Sci. USA 94, 7704.
E M E R G E N C E AND EVOLUTION OF T H E IMMUNOGLOBULIN FAMILY
499
Kirkham, P. M., and Schroeder, H. W., Jr. (1994). Antibody structure and the evolution of iinrnunoglobiilin V gene segments. Sem. hnmunol. 6, 347. Klein, J. (1989). Are invertebrates capable of anticipatory immune responses? Scand. J. Immunol. 29, 499. Klein, J., and O’hUigin, C. (1993). Composite origin of major histocompatibility complex genes. Curr. Opin. Genet Deo. 3, 923. Klein, J., Satta, Y., and O’hUigin, C. (1991).The molecular descent ofthe major histocompatibility complex. Annu. Reo. Immunol. 11, 269. Knight, K. L., and Crane, M. A. (1994). Generating the antibody repertoire in the rabbit. Ado. Immunol. 56, 179. Knight, K. L., and Winstead, C. R. (1997). Generation of antibody diversity in rabbits. Curr. Opin. linmunol 9, 228. Kobayashi, K., and Tomonaga, S. (1988).The second immunoglobulin class is commonly present in cartilaginous fish belonging to the order Rajfomies. Mol. Immunol. 25, 115. Kobayashi, K., Toinonaga, S., and Tanaka, S. (1992).Identification of a second immunoglobulin in the most primitive shark, the frill shark, Chlamydo.selachusarguineus. Dev. Comp. Immunol 16, 295. Kokubu, F., Hinds, K., Litman, R., Shamblott, M. J., and Litman, G. W. (1988a). Complete structure and organization of immunoglobulin heavy chain constant region genes in a phylogenetically primitive vertebrate. EMBO J. 7, 1979. Kokubn, F., Litman, R., Shamblott, M. J., Hinds, K., and Litman, G. W. (198th). Diverse organization of immunoglobulin VH gene loci in a primitive vertebrate. EMBOJ. 7,3413. Koradi, R., Billeter, M., and Wuthrich. K. (1996). MOLMOL: A program for display and analysis of macromoleci~larstructures. J. Mol. Graph. 14, 51. Kronenberg, M., Siu, G., Hood, L. E., and Shastri, N. (1986). The molecular genetics of the T-cell antigen receptor and T-cell antigen recognition. Annu. Rev. Immunol. 4, 529. Lai, E., Wilson, R. K., and Hood, L. E. (1989). Physical maps of the mouse and human immunoglobulin-like loci. Ado. Immunol. 46, 1. Ledford, B. E., Magor, B. G., Middleton, D. L., Miller, R. L., Wilson, M. R., Miller, N . W., Clem, L. W., and Warr, G. W. (1993). Expression of a mouse-channel catfish chimeric IgM molecule in a mouse myeloma cell. Mol. Immunol. 30, 1405. Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J.-M., and Hoffman, J. A. (1996). The dorsoventral regulatory gene cassette spntzle/Toll/cnctus controls the potent antifnngal response in Drosghila adults. Cell 86, 973. Leslie, G. A., and Clem, L. W. (1969). Phylogeny of immunoglobuIin structure and function. 111. ImmunogIobulins of the chicken. J . Exp. Med. 130, 1377. Litman, G. W., Berger, L., Murphy, K., Litman, R., Hinds, K., and Erickson, B. W. (1985a). Immunoglobulin VH gene stnicture and diversity in Heterodontus, a phylogenetically primitive shark. Proc. Natl. Acad. Sci. U S A 82, 2082. Litman, G. W., Finstad, F. J., Howell, J., Pollara, B. W., and Good, R. A. (1970). The evolution of the immune response. :3. Structural studies of the lamprey immunoglobulin. J. Immunol. 105, 1278. Litman, G. W., Murphy, K., Berger, L., Litman, R., Hinds, K., and Erickson, B. W. (198513). Complete nucleotide sequence of three VH genes in Cairnun, a phylogenetically ancient reptile: Evolutionary diversification in coding segments and variation in the structure and organization of recombination elements. Proc. Natl. Acad. Sci. U S A 82, 844. Litman, G. W., and Rast, J. P. (1996).The organization and structure of immunoglobulin and T-cell receptor genes in the most phylogenetically distant jawed vertebrates: Evolutionary implications. Res. Immunol. 147, 226.
500
JOHN J. MARCHALONIS et al.
Litman, C. W., Rast, J. P., Shamblott, M. J., Haire, R. N., Hulst, M., Roer, W., Litman, R. T., Hinds-Frey, K. R., Zilch, A., and Amemiya, C. T. (1993).Phylogenetic diversification of immunoglobulin genes and the antibody repertoire. Mol. B i d . Evol. 10, 60. Liu, T.-Y., Minetti, C. A. S., Fortes-Dias, C., Liu, T., Lin, L., and Lin, Y. (1994). C-reactive proteins, limunectin, lipopolysaccharide-bindingprotein and coagulin. Ann. N.Y. Acad. Scz. 712, 146. Liu, Y.-J., Johnson, G. D., Gordon, J., and MacLennan, I. C. M. (1992). Germinal centres in T-cell dependent antibody responses. Zmmunol. Today 13, 17. Liu, Y. S. V., Low, T. L. K., Infante, A,, and Putnam, F. W. (1976). Complete covalent structure of a human IgAl immunoglobulin. Science 173, 1017. Lobb, C. J., Olson, M. O., and Clem, L. W. (1984). Immunoglobulin light chain classes in a teleost fish. J. Zmmunol. 132, 1917. Luer, C. A,, Walsh, C. J., Bodine, A. B., Wyffels, J. T., and Scott, T. R. (1995). The Elasmobranch thymus: Anatomical, histological and preliminary functional characterization. J. Exp. 2 0 1 . 273, 342. Lundqvist, M., Bengten, E., Stromberg, S., and Pilstrom, L. (1996). Ig Light chain gene in the Siberian sturgeon (Acipenser baed).J. Immunol. 157, 2031. Magor, B. G., Ross, D. A., Middleton, D. L., and Warr, G. W. (1997). Functional motifs in the IgH enhancer of the channel catfish. Zmmunogenetics 46, 192. Magor, €3. G., Wilson, M. R., Miller, N. W., Clem, L. W., Middleton, D. L., and Warr, G. W. (1994a). An Ig heavy chain enhancer of the channel catfish ZctuZurus punctutus: Evolutionary conservation of function but not structure. J. Zmmunol. 153, 5556. Magor, K. E., Higgins, D. A., Middleton, D. L., and Warr, C. W. (1994b). One gene encodes the heavy chains for three different forms of IgY in the duck. J. Zmmunol. 153,5549. Mansikka, A. (1992). Chicken IgA H chains: Implications concerning the evolution of H chain genes. J. Zmmunol. 149, 855. Marchalonis, J. J. (1977). “Immunity in Evolution.” Haward Press, Cambridge. Marchalonis,J. J., and Atwell,J. L. (1972).Phylogenetic emergence of distinct immunoglobulin classes in Letude phylogenetique et ontogenique de la reponse immunitaire et son apport a la theorie immunologique. (P. Liacopoulos and J. Panijel, eds.), p. 153. INSERM, Pans. Marchalonis, J. J., Bernstein, R. M., Shen, S. X., and Schluter, S. F. (1996). Emergence of the immunoglobulin family: Conservation in protein sequence and plasticity in gene organization. Glycobiology 6, 657. Marchalonis, J. J., Ealey, E. H. M., and Diener, E. (1969).Immune response ofthe tuatera, Sphenodon punctutum. Aust. J. Exp. Biol. Med. Sci. 47, 367. Marchalonis, J. J., and Edelman, G. M. (1965). Phylogenetic origins of antibody structure. I. Multichain structure of immunoglobulins in the smooth dogfish (Mustelus canis). J. Exp. Med. 122, 601. Marchalonis, J. J., and Edelman, G. M. (1966a). Phylogenetic origins of antibody structure. 11. Immunoglobulins in the primary immune response of the bullfrog Runa cutesbiuna. 1.Exp. Med. 124, 901. Marchalonis, J. J., and Edelman, G. M. (1966b). Polypeptide chains of immunoglobulins from the smooth dogfish, Mustelus canis. Science 154, 1567. Marchalonis, J. J., and Edelman, G. M. (1968). Phylogenetic origins of antibody structure. 111. Immunoglobulins from the sea lamprey, Petromyzon mudnus. J. Erp. Med. 127,891. Marchalonis, J. J., and Germain, R. N. (1971). Tolerance to a protein antigen in a poikilotherm, the marine toad Bufo marinus. Nature 231, 321.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
501
Marchalonis,J. J., and Germain, R. N. (1980).In “Imn~unologicalTolera~ice” (M. J. Manning, ed.), p. 217. ElseviedNorth Holland Biomedical Press, Amsterdam. Marchalonis, J. J., Hohman, V. S., and Schluter, S. F. (1993). Antibodies of sharks: Novel methods of generation of diversity. Immunologist 114, 115. Marchalonis, J. J., Hohman, V. S., Thomas, C., and Schluter, S. F. (1993). Antibody production in sharks and man: A role for natural antibodies. Dev. Con~p.Immunol. 17, 41. Marchalonis, J. J., Mackel-Vandersteenhoven, A., Vasta, G. R., Schluter, S. F., DeLuca, D., Pandey, J. P., and Wang, A. C. (1987a). In “Antigen Specific T cell Receptors and Factors” (J. J. Marchalonis, ed.), p. 97. CRC Press, Boca Raton, FL. Marchalonis, J. J., and Schluter, S. F. (1989). Evolution of variable and constant domains and joining segments of rearranging immunoglobulins. FASEB 1. 3, 2469. Marchalonis, J. J., and Schluter, S. F. (1990a). On the relevance of invertebrate recognition and defense mechanisms to the emergence of the immune response of vertehrates. Scand. Zmn~nunol.32, 13. Marchdonis, J. J., and Schluter, S. F. (199Ob). Origins of immunoglobulins and immune recognition molecules. BioScience 40, 758. Marchalonis, J. J., Schluter, S. F., and Edmundson, A. B. (1997). The T-cell receptor as immunoglobulin: Paradigm regained. Proc. Soc. Exp. Biol. 216, 303. Marchalonis, J. J., Schluter, S. F., Hubbard, R. A., Diamanduros, A., Barker, W. C., and Pumphrey, R. S. (1988a). Conservation of iinmunoglobulin variable and joining region structure and the design of universal anti-immunoglobulin antibodies reactive with antigen-binding T cell receptors. Znt. Rev. Inmiunol. 3, 241. Marchalonis, J. J., Schluter, S. F., Hubbard, R. A., McCabe, C., and Allen, R. C. (1988b). Immunoglobulin epitopes defined by synthetic peptides corresponding to joining region sequence: Conservation of determinants and dependence upon the presence of an arginyl or lysyl residue for cross-reaction between light chains and T-cell receptor chains. Mol. Zinmunol. 25, 771. Marchalonis, J. J., Schluter, S. F., Rosenshein, I. L., and Wang, A. C. (1988~).Partial characterization of immunoglobulin light chains of carcharhine sharks: Evidence for phylogenetic conservation of variable region and divergence of constant region structure. Deu. Coinp. Zinmunol. 12, 65. Marchalonis, J. J., Schluter, S. F., Wang, E., Dehghanpisheh, K., Lake, D., Yocum, D. E., Edmundson, A. B., and Winfield, J. B. (1994). Synthetic autoantigens of immunoglobulins and T-cell receptors: Their recognition in aging, infection and autoimmunity. Proc. Soc. Exp. Biol. 207, 129. Marchalonis, J. J., Vasta, G. R., Warr, G. W.. and Barker, W. C. (1984). Probing the boundaries of the extended immunoglobulin family of recognition molecules: Jumping domains, convergences and minigenes. Immunol. Today 5, 133. Marchalonis, J. J., Wang, A. C., Galbraith, R. M., and Barker, W. C. (198713). In “The Lymphocyte: Structure and Function” (J. J. Marchalonis, ed.), p. 307. Dekker, New York. Martin, A. P., Naylor, G. J. P., and Palumbi, S. R. (1992). Rates of mitochondria1 DNA evolution in sharks are slow compared with mammds. Nature 357, 153. Matsunaga, T., and Mori, N. (1987). The origin of the immune system: The possibility that immunoglobulin superfamily molecules and cell adhesion molecules of chicken and slime mould are all related. Scnnd. 1. lninwnol. 25, 485. Max, E. E. (1993). I n “Fundamental Immunology” (William E. Paul, ed.), p. 315. Raven Press, Bethesda, MD. McCormack, W. T., Tjoelker, L. W., and Thompson, C. B. (1991). Avian /3 cell deveIopment: Generation of an immunoglobulin repertoire by gene conversion. Ann. Reu. Znzmunol. 9, 219.
I.
502
JOHN J. MARCHALONIS et al.
McKinney, E. C. (1992). Shark lymphocytes: Primitive antigen reactive cells. Annu. Rev. Fish. Dis. 2, 43. McLaughlin, P. J., and Dayhoff, M. 0. (1972).In “Atlas of Protein Sequence and Structure” (M. 0. Dayhoff, ed.), p. 47. National Biomedical Research Foundation, Washington, DC. Metchinkoff, E. (1884). Uber eine Sprospilzkrank-hert der Daphmin Beitrag, zur Lehre uber den Kampf der Phagocyten gegen Krankheitserreger. Virch. Arch. (Zelpathol.) 96, 177. Michard-Vanhee, C., Chourrout, D., Stroinberg, S., Thuvander, A., and Pilstrom, L. (1994). Lymphocyte expression in transgenic trout by mouse immunoglobulin promotedenhance. Immunogenetics 40, 1. Miller, D. A,, and Ratcliffe, N. A. (1989).The evolution of blood cells: Facts and enigmas. Endeavor 13,72. Miller, N. W., Rycyzyn, M. A,, Wilson, M. R., Warr, G. W., Naftel, J. P., and Clem, L. W. (1994). Development and characterization of channel catfish long term B cell lines. 1.Immunol. 152,2180. Mussmann, R., Du Pasquier, L., and Hsu, E. (1996). Is Xenopus IgX an analog of IgA? Eur. I. Immunol. 26, 2823. Mussmann, R., Wilson, M., Marcuz, A,, Courtet, M., and Du Pasquier, L. (1996).Membrane exon sequences of the three Xenopus Ig classes explain the evolutionary origin of mammalian isotypes. Eur. ]. Immunol. 26, 409. Muydermans, S., Atarhouch, T., Saldanha, J., Barbosa, J. A., and Hamers, R. (1994). Sequence and structure of VH domain from naturally occurring camel heavy chain immunoglobulins lacking light chains. Prot. Eng. 7, 1129. Nei, M. (1975). “Molecular Population Genetics and Evolution.” Holland Publishing Co., Amsterdam. Nei, M., Gu, X., and Sitnikova, T. (1997). Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc. Natl. Acad. Sci. USA 94,7799. Noguchi, J. (1903). A study of immunization: Haemolysins, agglutinins, precipitins and coagulins in cold-blooded animals. Centmlbl.f: Bakt. Abt. Orig. 33,353. Nonaka, M., Takhashi, M., and Sasaki, M. (1994).Molecular cloning of alamprey homologue of the mammalian MHC class 111 gene, complement factor B.]. ZmmunoZ. 152,2263. Ohta, T., and Kimura, M. (1971). Functional organization of genetic material as a product of molecular evolution. Nature 233, 118. Page, R. D. M. (1996).An application to display phylogenetic trees on personal computers. Comput. Appl. Biosci. 12, 357. Partula, S.,De Guerra, A., Fellah, J. S., and Charlemagne, J. (1995).Structure and diversity of the T-cell antigen receptor &chain in a teleost fish. 1. Immunol. 155,699. Partula, S., de Guerra, A,, Fellah, J. S., and Charlemagne, J. (1996). Structure and Diversity of the TCR a-chain in a teleost fish. /. Immunol. 157,207. Parvari, R., Avivi, A,, Lentner, F., Ziv, R., Tel-Or, S., Burstein, Y., and Schechter, I. (1988). Chicken immunoglobulin y-heavy chains: Limited VH gene repertoire, combinatorial diversificationby D gene segments and evolution of the heavy chain locus. EMBO]. 7,739. Picker, L. J., and Siegelman, M. H. (1993). In “Fundamental Immunology” (W. E. Paul, ed.),p. 145. Raven Press, New York. Pilstrom, L., and Bengten, E. (1996). Immunoglobulin in fish: Genes, expression and structure. Fish Shellfish Immunol. 6,243. Raftos, D. A., Cooper, E. L., Habicht, G. S., and Beck, G. (1991). Invertebrate cytokines: Tunicate cell proliferation stimulated by an interleukin 1-like molecule. Proc. Natl. Acad. Sci. USA 88, 9518.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
503
Raison, R. L., and Hildemann, W. H. (1984). Immunoglobulin-bearing blood leucocytes in the Pacific hagfish. Dev. Comp. Immunol. 8, 99. Ramsden, D. A., Baetz, K., and Wu, G. E. (1994).Conservation of sequence in recombination signal sequence spacers. Nucleic Acid Res. 22, 1785. Rast, J. P., Anderson, M. K., Ota, T., Litman, R. T., Margittal, M., Shamblott, M. J., and Litman,G. W. (1994). Iminunoglobulin light chain class multiplicity and alternative organizational forms in early vertebrate phylogeny. Immunogenetics 40, 83. Rast, J. P., Anderson, M. K., Strong, S. J., Luer, C., Litman, R. T., and Litman, G. W. (1997). a,p, y , and S T cell antigen receptor genes arose early in vertebrate phylogeny. Immunity 6, 1. Rast, J. P., Haire, R. N., Litman, R. T., Pross, S., and Litman, G. W. (1995). Identification and characterization of T-cell antigen receptor-related genes in phylogenetically diverse vertebrate species. Immunogenetics 42, 204. Rast, J. P., and Litman, G. W. (1994). T-cell receptor gene homologs are present in the most primitive jawed vertebrates. Proc. Natl. Acnd. Sci. USA 91, 9248. Reidl, L. S., Kinoshita, C. M., and Steiner, L. A. (1992). Wild mice express an Ig VA gene that differs from any Vh in Balb/c but resembles a human Vh subgroup. 1. Immunol. 149, 471. Reynaud, C. A,, Anqnez, V., Grimal, H., and Weill, J.-C. (1987). A hyperconversion mechanism generates the chicken light chain preimmune repertoire. Cell 48, 379. Reynaud, C.-A., Dahan, A,, Anquez, V., and Weill, J.-C. (1989). Somatic hyperconversation diversifies the single VH gene of the chicken with a high incidence in the D region. Cell 59, 171. Reynaud, C.-A,, Mackay, C. R., Muller, R. C., and Weil, T.-C. (1991). Somatic generation of lversity in a mammalian primary lymphoid organ: The sheep that Peyer's patches. Cell 64, 995. Roman, T., and Charlemagne, J. (1994). The immunoglobulin repertoire of the rainbow trout (Oncorhynchusmykiss): Definition of nine IgH-V families. Immunogenetics 40,210. Rombout, J. H., Taveme, N., Van de Kamp, M., and Taverne-Thiele, A. J. (1993). Differences in nuxiis and serum immunoglobulin of carp (Cyprinus carpio L). Dev. Comp. Immunol. 17, 309. Rosenshein, I. L., Sehluter, S. F., Vasta, G. R., and Marchalonis, J. J. (1985). Phylogenetic conservation of heavy chain determinants of vertebrates and protochordates. Dev. Comp. Immuriol. 9, 783. Ruben, L. N., Warr, C. W., Decker, J. M., and Marchalonis, J. J. (1977). Phylogenetic origins of immune recognition: Lymphoid heterogeneity and the haptedcamer effect in the goldfish. Carmsius nuratus. Cell. Immunol. 31, 266. Saitou, N., and Nei, M. (1987).The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406. Sanchez, C., Alvarez, A,, Castillo, A,, Zapata, A,, Villena, A., and Dominguez, J. (1995). Two different subpopulations of Ig-bearing cells in lymphoid organs of rainbow trout. Dev. Comp. Immunol. 19, 79. Sanchez, C., and Dominguez, J. (1991). Trout immunoglobulin populations differing in light chains revealed by monoclonal antibodies. Mol. Immunol. 28, 1271. Schacke, H. (1994). Immunoglobulin-like domain is present in the extracellular part of the receptor tyrosine kinase from the marine sponge Geodia cydonium. J. Mol. Recogn. 7,273. Schatz, D. G., Oettinger, M. A,, and Baltimore, D. (1989). The V(D)J recombination activating gene, RAG-1. Cdl 59, 1035. Schatz, D. G., Oettinger, M. A,, and Schlissel, L. (1992). V(D)J recombination: Molecular biology and regulation. Annu. Rev. Immunol. 10, 359.
504
JOHN J. MARCHALONIS et d.
Schiffer, M., Wu, T. T., and Kabat, E. A. (1986). Subgroups of V-genes of &chains of Tcell receptors for antigens. Proc. Natl. Acad. Sci. USA 83, 4461. Schluter, S. F., Beischel, C. J., Martin, S. A., and Marchalonis, J. J. (1989a). Sequence analysis of homogeneous peptides of shark immunoglobulin light chains by tandem mass spectrometry: Correlation with gene sequence and homologies among variable and constant region peptides of sharks and mammals. Mol. lmmunol. 27, 17. Schluter, S. F., Bernstein, R. M., and Marchalonis, J. J. (1997). Molecular origins and evolution of immunoglobulin heavy chain genes of jawed vertebrates. lmmunol. Today 18, 543. Schluter, S. F., Hohman, V. S., Edmundson, A. B., and Marchalonis, J. J. (198913).Evolution of immunoglobulin light chains: cDNA clones specifylng sandbar shark constant regions. Proc. Natl. Acud. Sci. USA 86, 9961. Schluter, S. F., and Marchalonis,J. J. (1986). Antibodiesto synthetic joining segment peptide of the T-cell receptor p chain: Serological cross-reaction between products of T-cell receptor genes, antigen binding T-cell receptors and immunoglobulins. Proc. Nutl. Acud. Sci. USA 83, 1872. Schluter, S. F., Rosenshein, I. L., Hubbard, R. A,, and Marchalonis,J, J. (1987).Conservation among vertebrate immunoglobulin chains detected by antibodies to a synthetic joining segment peptide. Biochem. Biophys. Res. Commun. 145, 699. Schwager, J., Burcket, N., Schwager, M., and Wilson, M. (1991). Evolution ofimmunoglobulin light chain genes: Analysis of Xenopus IgL isotypes and their contribution to antibody diversity. EMBO J. 10, 505. Schwager, J., Grossberg, D., and Du Pasquier, L. (1988a).Organization and rearrangement of immunoglobulin M genes in the amphibian Xenopus. EMBO]. 7, 2409. Schwager, J., Mikoryak, C. A,, and Steiner, L. A. (1988b). Amino acid sequence of heavy chains from Xenopus laeuis IgM deduced from cDNA sequence: Implications for evolution of immunoglobulin domains. Proc. Natl. Acad. Sci. USA 85, 2245. Schwager, P. S., Timmusk, J., Pilstrom, L., and Charlemagne, J. (1996). A second immunoglobulin light chain isotype in the rainbow trout. lminunogenetics 45, 44. Seeger, M. A., Haffley, L., and Kaufman, T. C. (1988). Characterization of amalgam: A member of the immunoglobulin superfamily from Drosophila. Cell 55, 589. Shamblott, M. J., and Litman, G. W. (1989a). Complete nucleotide sequence of primitive vertebrate Ig light chain genes. Proc. Natl. Acad. Sci. USA 86, 5684. Shamblott, M. J., and Litman, G. W. (1989b). Genomic organization and sequences of Ig light chain genes in a primitive vertebrate suggest coevolution of Ig gene organization. EMBO]. 8, 3733. Shen, S. Y., Bernstein, R. M., Schluter, S. F., and Marchalonis, J. J. (1996). Heavy chain variable regions in carcharhine sharks: Development of a comprehensive model for the evolution of VH domains among the gnathanstomes. lmmunol Cell. Biol. 74, 357. Silverman, G. J. (1992).Human antibody responses to bacterial antigens: Studies of a model conventional antigen and a proposed model B cell superantigen. Int. Rev. lmmunol. 9,57. Sims, J. E., March, C. K., Cosman, D., Widmer, M. B., MacDonaId, H. R., McMahan, C. J., Grubin, C. E., Wignall, J. M., Jackson, J. L., Call, S. M., Friend, D., Alpert, A. R., Gillis, S., Urdal, D. L., and Dower, S. K. (1988). cDNA expression cloning of the IL-1 receptor, a member of the immunoglobulin superfamily. Science 241, 585. Sledge, C., Clem, L. W., and Hood, L. E. (1974). Antibody structure: Amino terminal sequence of nurse shark light and heavy chains. 1.lmmunol. 112, 941. Smith, L. C., and Davidson, E. H. (1992).The echinoid immune system and the phylogenetic occurrence of immune mechanisms in deuterostomes. Immunol. Today 13, 356.
EMERGENCE AND EVOLUTION OF THE IMMUNOGLOBULIN FAMILY
505
Snapper, C. M., Marcu, K. B., and Zelazowski, P. (1997). The immunoglobulin class switch: Beyond “accessibility.” 1?n?nunity6, 217. Stewart, S. E., Du Pasquier, L., and Steiner, L. A. (1990). Diversity of expressed V and J regions of immunoglobulin light chains in Xenopus laeuis. Eur. J. Ivnn~unol.23, 1980. Stuge, T. B., Yoshida, S. H., Chinchar, V. G., Miller, N. W., and Clem, L. W. (1997). Cytotoxic activity generated from channel catfish peripheral blood leukocytes in mixed leukocyte cultures. Cell lmmunol. 177, 154. Sun, J., and Butler, J. E. (1996). Molecular characterization of VDJ transcripts from a newborn piglet. Immunology 88, 331. Sun, S. C., Lindstrom, I., Boman, H. G., Faye, I., and Schmidt, 0. (1990). Hemolin: An insect-immune protein belonging to the immunoglobulin superfamily. Science 250,1729. Szenberg, A,, and Warner, N. L. (1962). Dissociation of immunological responsiveness in fowls with a hormonally arrested development of lymphoid tissue. Nature 194, 146. Thompson, C. B. (1995). New Insights into V( D)J recombination and its role in the evolution of the immune system. Immunity 3, 531. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994). CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acid Res. 22, 4673. Tjoekler, L. W., Carlson, L. M., Lee, K., Lahti, J., McCormack, W. T., Leiden, J. M., Chen, C.-L. H., Cooper, M. D., and Thompson, C. B. (1990). Evolutionary conservation of antigen recognition: The chicken T-cell receptor P chain. Proc. Natl. Acad. Sci. USA 87, 7856. Tomlinson, I. M., Walter, G., Marks, J. D., Llewelyn, M. B., and Winter, G. (1992). The repertoire of human germline VH sequences reveals about fifty groups of VH segments with different hypervariable loops. 1. Mu[. B i d . 227, 776. Toyonaga, R., and Mak, T. (1987). Genes of the T cell antigen receptor in normal and inalignant T-cells. Annu. Rev. linmunol. 5, 585. Turchin, A,, and Hsu, E. (1996). The generation of antibody diversity in the turtle. /. lni~iiunol.156, 3797. Uhr, J. W., Finkelstein, M. S., and Franklin, E. C. (1962).Antibodyresponse to bacteriophage ox174 in nonmammalian vertebrates. Proc. Soc. Exp. Bid. Med. 111, 13. Varner, J., Neame, P., and Litman, G. W. (1991). A serum heterodimer from hagfish (Eptatwtus stuutii) exhibits structural similarity and partial sequence homology with immunoglobulin. Proc. Natl. A d . Sci. U S A 88, 1746. Vazquez, M., Mizuki, N., Flajnik, M. F., McKinney, E. C., and Kasahara, M. (1992). Nucleotide sequence of a nurse shark immunoglobulin heavy chain cDNA clone. Mul. lnmiunol. 24, 1157. Ventura-Hohnan, T., Jones, J. C., Ghaffari, S. H., and h b b , C. J. (1994). Structure and genomic organization of VH gene segments in the channel catfish: Members of different VH gene families are interspersed and closely linked. Mol. Immunol. 31, 823. Volanakis, J. E., Xu, Y., and Macon, K. J. (1990). In “Defense Molecules” (J. J. M. A. C. L. Reinisch, ed.), p. 161. Wiley-Liss, New York. Warr, G. W. (1995). The immunoglobulin genes of fish. Dev. Comp. lmmunol. 19, 1. Warr, G. W., DeLuca, D., and Marchalonis, J. J. (1976). Phylogenetic origins of immune recognition: Lymphocyte surface imniunoglobulirrs in the goldfish, Carassius auratzrs. Proc. NatE. Acad. Sci. U S A 73, 2476. Warr, G. W., Magor, K. E., and Higgins, D. A. (1995). IgY: Clues to the origins of modern antibodies. lmmunol. Toclay 16, 392. Warr, G. W., Middleton, D. L., Miller, N. W., Clem, L. W., and Wilson, M. R. (1991). An additional family of VH sequences in the channel catfish. Eur. J. Imniunogenet. 18,393.
506
JOHN
1.
MARCBALONIS et al
Warr, G. W., Miller, N. W., Clem, L. W., and Wilson, M. R. (1992). Alternative splicing pathways of the immunoglobulin heavy chain transcript of a teleost fish, Ictulurus punctutus. lmmunogenetics 35, 253. Weill, J. C., and Raynaud, C. A. (1996). Rearrangementhypermutatiodgene conversion: When, where and why? Immunol. Today 17, 92. Widal, F., and Sicard, E. (1897). Influence de l’organisine sur les proprietes acquises par les humeurs du fait de I’infection (Lagglutination chez quelques animaux a sang-froid). C. R. SOC.Biol. (Paris) 4, 1047. Williams, A. F., and Barclay, A. N. (1988). The immunoglobulin superfamily: Domains for cell surface recognition. Annu. Rev. Immunol. 6, 381. Wilson, M., Bengten, E., Miller, N. W., Clem, L. W., Du Pasquier, L., and Warr, G. W. (1997). A novel chimeric Ig heavy chain from a teleost fish shares similarities to IgD. Proc. Nutl. Acud. Sci. USA 94, 4593. Wilson, M. R., and Warr, G. W. (1992). Fish immunoglobulins and the genes that encode them. Annu. Rev. Fish Dis. 201. Yanagi, T., Yoshikai, Y., Leggett, K., Clark, S., Aleksander, I., and Mak, T. W. (1984). A human T cell specific cDNA clone encodes a protein having extensive homology to immunoglobulin chains. Nature 308, 145. Ying, S.-C., Marchalonis, J. J., Gewurz, A. T., Siege], J. N., Jiang, H., Gewurz, B. E., and Gewurz, H. (1992). Reactivity of anti-human CRP and SAP monoclonal antibodies with Limulin and Pentraxins of other species. Immunology 76, 324. Zapata, A. G., and Cooper, E. L. (1990). “The Immune System: Comparative Histophysiology.” Wiley, Chichester, England. Zezza, D. J., Stewart, S. E., and Steiner, L. A. (1992). Genes encoding Xenopus laevis Ig L chains: Implications for the evolution of K and A chains. J. Immunol. 149, 3968. Zhou, H., Benten, E., Miller, N. W., Warr, G. W., Clem, L. W., and Wilson, M. R. (1997). T cell receptor sequences in the channel catfish. Dev. Comp. Immunol. 21, 238. This article was accepted for publication on December 15, 1997.
ADVANCES IN IMMUNOLOGY,VOL. 70
Current Insights into the ”Antiphospholipid” Syndrome: Clinical, Immunological, and Molecular Aspects DAVID A. KANDIAH,’ ANDREJSAU,~YONGHUA SHENG,’ EDWARDJ. VICTORIA,* DAVID M. MARQUIS,* STEPHEN M. COVrrS,* and STEVEN A. KRIUS’ ‘Depadment of Immunology, Allergy, and lnfeciiaus Disease, Universiiy of New h u h Wales School of Medicine, St. George Haspipol, Kogarah 2217, Austmlia; tRackefeller Universiiy, New York, New Yark 1002 I; and *la Jolh Pharmaceutical Company, k n Diego, California 92 12 1
1. Introduction
In 1983, a distinct syndrome consisting of vascular thrombosis, livedo reticularis, thrombocytopenia, and movement disorders associated with “antiphospholipid (aPL) antibodies was first described (Hughes, 1983). Early studies on aPL antibodies were on patients with systemic lupus erythematosus (SLE) and it was in a subset of patients with SLE that the “antiphospholipid syndrome” (APS) came to be recognized (Hughes, 1985). The association of vascular thrombosis and autoimmune disease was found in the 1960s (Bowie et al., 1963; Alarcon-Segovia and Osmundson, 1965)and laid the foundation for studies to discover the pathogenesis and immunological features of this distinct group of individuals. It was soon noted that a subset of patients had the clinical manifestations of APS without sufficient clinical and immunological criteria to satisfy the 1982 American College of Rheumatology (ACR) diagnostic criteria for SLE (Tan et al., 1982).The definition and criteria for a “primary antiphospholipid syndrome” (PAPS) was first proposed (Asherson, 1988) and the first series was documented the following year (Alarcon-Segovia and Sanchez-Guerrero, 1989). In a 2-year multicenter follow-up study of patients with PAPS and secondary APS (SAPS)in other autoimmune diseases, a lower female/male sex ratio in PAPS was found compared to that in patients with SAPS. Patients with SAPS had more neutropenia, autoimmune hemolpc anemia, endocardial vegetations, and low levels of complements compared to PAPS patients. The incidence of thrombosis was no different in the two groups (Vianna et al., 1994). Although the original concept of the APS was shown to comprise one or more of the clinical manifestations of venous thrombosis, arterial thrombosis, recurrent fetal loss, and thrombocytopenia, more diverse clinical manifestations are now recognized, such as cardiac valvular lesions, adrenal insufficiency, and multiorgan thrombotic complications known as “catastrophic” APS (Asherson, 1992).Despite better understanding of the target antigens of aPL antibodies, original laboratory criteria of moderate to high 507
Copyright 0 1998 by Acditernic Pres.; All nghts of reproduction in any form reserved 0065-2776/98 $25 00
508
DAVID A. KANDIAH et al.
levels of “anticardiolipin” (aCL) antibodies and/or lupus anticoagulant (LA) antibodies are still being used and have now replaced the ACR criteria listing of the lupus erythematosus (LE) cell in the revised list for the diagnosis of SLE (Hochberg, 1997). Current guidelines for the diagnosis of APS are the presence of at least one of the clinical criteria of venous thrombosis, arterial thrombosis, recurrent pregnancy loss, and thrombocytopenia, with one or more of the laboratory criteria of moderate to high levels of IgG and/or IgM aCL antibodies and detection of LA activity in a clotting assay. Current criteria for the detection of LA activity in plasma will be discussed later in Section XI. In a previous review of this subject, the emphasis was on antibody interactions with phospholipids, and a brief introduction was made on the role of P2-glycoprotein I (P2GPI) (McNeil et al., 1991). This review develops on the current insights on antibody interactions with phospholipid-binding plasma proteins, in particular P2GP1, and covers currently recognized clinical associations, immunological aspects, molecular studies, and therapeutic interventions. II. “Antiphospholipid Antibodies
aPL antibodies are a heterogeneous group of autoantibodies that have specificity for a number of phospholipid-binding proteins, phospholipid molecules, and phospholipid-protein complexes. A number of phospholipid-binding proteins have been implicated in APS, including PBGPI, prothrombin, protein C, protein S, kininogens, thrombomodulin, and annexin V. Traditionally, aPL antibodies are detected in LA assays and in solid-phase immunoassays using cardiolipin as the target antigen. The target antigen detected in clotting assays is still unclear, and a number of coagulation proteins have been implicated, particularly PZGPI and prothrombin. It appears, however, that the complexes assembled on the phospholipid surfaces in functional clotting assays are more important than any one protein. The paradoxical phenomenon of prolongation of in vitro phospholipid-dependent clotting tests to detect LA antibodies while being associated in vivo with vascular thrombosis has intrigued scientists and clinicians for years. Studies of APS may throw some light on the pathophysiology of the mode of action of these autoantibodies. Because the original immunoassays for the detection of aPL antibodies used coated cardiolipin on microtiter plates, they were called anticardiolipin antibodies (Loizou et al., 1985). It has been shown that the target antigen in this assay is P2GPI (McNeil et al., 1990; Galli d al., 1990), the original nomenclature is a misnomer, and the antibodies should be known as anti-p2GPI antibodies. For the purpose of this review, antibodies detected on assays employing P2GPI as the coated antigen in the absence
ANTIPIIOSPHOLIPID SYNDROME
509
of phospholipids are referred to as anti-P2GPI antibodies and antibodies detected in a cardiolipin ELISA are referred to as aCL antibodies. These two populations of antibodies in autoimmune patients are identical with few exceptions. As a generic term for anti-P2GPI antibodies and LA antibodies, the term aPL antibodies will be employed. The units of antibody levels in sera expressed in the standard CL-ELISA have been calibrated to known sera from the R a p e Institute, London. One GPIJMPL unit each represents one microgram of affinity-purified antibody per milliliter of serum. Antiphospholipid antibodies are found in “normal” individuals. In a population of 499 blood donors, the prevalence of LA antibodies was 8% and anticardiolipin antibodies was 4.6 (IgG aCL), 4.6 (IgM aCL), and 5.6% (for polyvalent aCL antibodies) (Shi et nl., 1990).When these samples were stratified and the demographics of the blood donors studied, LA antibodies were found in young females. Prospective studies need to be done on these individuals to detect if any clinical problems had developed in subsequent years. As aPL antibodies are not normally distributed with most individuals having undetectable levels, it is more appropriate to use a nonparametric definition of the normal range, such as the 95% central tendency. Antiphospholipid antibodies can be divided into two main groups, classified according to their association with autoimmune or infective conditions. Traditionally these antibodies are termed autoimmune and alloimmune, respectively. Until the discovery that autoimmune antibodies are generally directed to the phospholipid-binding protein p2GPI instead of to the phospholipid molecule itself, the difference between these antibodies was unclear. Although there are exceptions to this rule, autoimmune aPL antibodies detected in solid-phase immunoassays with anionic phospholipids, such as cardiolipin as the coated antigen, are directed to P2GPI captured on a negatively charged surface. Alloimmune aPL antibodies found in chronic infections such as malaria, syphilis, leprosy, tuberculosis, and parvovirus infections do not bind P2GPI but are directed against the anionic phospholipid with p2GPI competing for binding with these antibodies (Hunt et nl., 1992).The binding of this latter group of antibodies has been shown to be charge dependent as high salt-containing buffers abolish binding to cardiolipin (Monestier et al., 1996). A. LUPUS ANTICOAGULANT ANTIBODIES
Plasma that contained proteins that prolonged phospholipid-dependent in vitro clotting assays were first described in SLE in 1952 (Conley and
Hartmann, 1952).Increasing the phospholipid in the assay system appeared to neutralize this LA reaction (Yin and Gaston, 1965). LA activity was
510
DAVID A. KANDIAH d a!.
found in the IgG fraction of serum. It was shown that immunoglobulins with LA activity react with anionic phospholipid but not with zwitterionic phospholipids. The phospholipid configuration appeared important with LA antibodies directed to the hexagonal phase rather than lamellar-phase phospholipids (Rauch et al., 1989). Current evidence would suggest that the antibodies may actually interfere with the assembly of enzymatic procoagulant and anticoagulant complexes on phospholipid surfaces, resulting in clinical vascular complications. Although these complications are predominantly thromboses, reports of hemorrhagic diatheses in patients with LA activity have been described. One of the probable causes of this is the presence of high-affinity antiprothrombin antibodies that complex with prothrombin, resulting in removal of the immune complexes by the reticuloendothelial system. This creates a situation of functional hypoprothrombinemia and bleeding. The dilute Russell’s viper venom time (dRVVT), the dilute activated partial thromboplastin time (dAPTT),and the dilute kaolin clotting time (dKCT) are the most frequently used tests for the detection of LA antibodies in routine practice and for research papers. However, up to 53% of patients with LA antibodies have a prolonged prothrombin time (Horellou et aZ.,1987). Although this may sometimes be due to low factor I1 levels, most of the patients studied have been shown to have normal antigenic levels of prothrombin (Horellou et al., 1987; Fleck et al., 1988). B. ANTI@ GLYCOPROTEIN 1ANTIBODIES Autoantibodies can be produced in response to tissue breakdown as a result of exposure of a target antigen not usually in contact with immunemediated cells. They can also arise after altered expression of cell surface proteins due to external stimuli or to translocation of intracellular antigens to cell surface membranes. The ability of a particular host to handle specific antibody-antigen complexes also predisposes to tissue and organ damage. These factors may all interact to suggest a pathogenic mechanism for the generation of autoantibodies in APS. Following purification of aCL antibodies by ion-exchange chromatography or phospholipid-polyacrylamide affinity chromatography, these antibodies failed to bind to the same phospholipid affinity column unless native or bovine plasma was also present (McNeil et al., 1989). Hence a plasma cofactor had been separated from the antibodies during the purification process that formed part of the antigenic target for these antibodies. The purified antibodies were able to bind in a cardiolipin ELISA where bovine serum is used in diluent and blocking buffers (called a standard CLELISA). This plasma cofactor was purified to homogeneity, sequenced, and identified as P2GPI (McNeil et al., 1990). This phospholipid-binding
FIG. 7-21A. Four-stranded P-pleated sheet (red arrows) of a V, domain, taken from the structure of a human Fab with V, of subgroup I11 (A. B. Edmundson et al., unpublished data). Totally conserved residue positions are colored blue and those 90% conserved are represented as blue chevrons. Residues conserved at lower levels are designated by pale red (75%)or red striped bands (60%).Strategically located residues are numbered to allow correlation of this model with the sequence presented in Fig. 20. This figure was devised by Benjamin Goldsteen and Allen Edmundson, using the program MOLMOL (Koradi et d.,1996).
FIG. 7-21B. Five-stranded P-pleated sheet (white arrows) of the same V, domain. Color coding for the conserved residues and the numbering follow the patterns for A. This figure was devised by Benjamin Goldsteen and Allen Edmundson, using MOLMOL (Koradi et al., 1996).
FIG.8-3. Distribution of charges in the three-dimensional model of human p2GPI-5. Main chain trace of the three-dimensional inodel of P2GPI-5. The positively charged side chains ( h i s , Arg) are shown in blue. The His side chains are not shown, but their main chain is colored blue. The main chains of the negatively charged residues (Asp, Clrr) are shown in red. The phospholipidbinding site is indicated by an arrow. The figure was prepared by program QUANTA (MSI, Sari Diego, CA). Reproduced with permission from Sheng et 01. (1996). 01996. The American Association of Immunologists.
FIG. 8-4. Distribution of charges in the three-dimensional model of human p2GPI-5. Electrostatic potential at the phospholipid-binding region of native and mutant p2GPI-5. (A) Native p2GPI-5 at neutral pH. (B) Lys 42/44/45 + Glu triple mutant at neutral pH. The molecular surfaces of the models are colored by the electrostatic potential, as shown by the color bar on each panel (in units of kT; 1 kT unit = 0.58 kcal/electron mol). The figures were prepared by program GRASP (Nicholls et al., 1991), using the relative dielectric constants of 2 and 78 for protein and solvent, respectively, and the salt concentration of 150 mM. The positively and negatively charged residues are numbered in yellow. Relative to Fig. 3, p2GPI-5 is viewed from the top. Reproduced with permission from Sheng et al. (1996).0 1996.The American Association of Immunologists.
ANTIPHOSPHOLIPID SYNDROME
511
plasma protein is found in relatively high concentrations of 4 p M in plasma or sera and appeared to have a role in the coagulation cascade as a natural anticoagulant based on in vitru experiments. This has led to the proposed theory that anti-P2GPI antibodies found in APS interfere with the natural procoagulant-anticoagulant homeostatic mechanisms, resulting in a procoagulant tendency and clinical thrombosis and atherogenesis. The first clinical study to investigate the role of anti-bZGPI antibodies with thrombosis found that 36% of patients with SLE had these antibodies (Viard et al., 1992). If these antibodies were found with LA antibodies, there was a strong association with thrombosis. Other small retrospective studies were performed to confirm an association of anti-P2GPI antibodies, but were fraught with technical problems in patient selection, retrospective analyses, and the ELISA methods used to detect these antibodies (Martinuzzo et al., 1995; Balestrieri et al., 1995; Cabiedes et al., 1995; Pengo et al., 1996). In the last paper the population of patients selected was done on the basis of their reactivity in a cardiolipin ELISA. It is therefore not surprising that anti-PeGPI antibodies were found in all the patients who had thrombosis, as the target antigen in the standard CL-ELISA is bovine PSGPI, which supports the binding of most human aPL antibodies. There is still some controversy as to whether most patients with autoimmune APS have both anti-pZGPI and antibodies that bind anionic phospholipids directly. Examining the sera of patients with both PAPS and SLE/APS, 68% were found to have true aCL antibodies as demonstrated by reactivity to CL on thin-layer chromatography plates, independent of the presence of PZGPI (Sorice et al., 1996). Using delipidated P2GPI as the antigen, 22.6% of the CL-ELISA positive sera bound P2GPI on immunoblotting. There may also be a population of antibodies that require the complex of PZGPI and phospholipid. Sixteen out of 18 patients with SLE and clinical manifestations of APS with negative IgG and IgM standard CLELISA reactivity had IgG anti-62GPI antibodies (Cabiedes et al., 1995). In a study of 97 patients with IgG and/or IgM anti-02GPI reactivity in a PBGPI-ELISA, 43% of IgM and 8% of IgG antihuman P2GPI antibodies did not bind to purified bovine PBGPI, explaining a negative aCLELISA where bovine PZGPI is the major source of P2GPI (Arvieux et al., 1996). The initial concentration of P2GPI in the patients' sera and the dilution of sera used could determine whether binding occurs in the CL-ELISA. The discovery that PZGPI exhibits genetically determined structural polymorphism with the occurrence of four alleles is another potential source of confusion in the conventional CL-ELISA. PZGPI from certain individuals, hoinozygous for the APOH"3 allele, is unable to bind anionic phospholipid (Kamboh et al., 1995). This group has also found two structural mutations
512
DAVID A. KANDIAI-1 et al.
at codons 306 and 316 in the fifth domain of 02GPI. These inissense mutations affect the structural integrity of the fifth domain of /32GPI affecting phospholipid binding. The authors suggest that there are individuals who are compound heterozygotes for the two mutations, who may be precluded from producing anti-b2GPI autoantibodies (Sanghera et al., 1997).Antibodies with P2GPI reactivity and true CL reactivity exist in the same patient population, with IgG2 subclass restriction of anti-P2GPI antibodies in patients with autoimmune disease (Arvieux et al., 1994). These factors all contribute to the differences in assay results for patients with APS in a conventional CL-ELISA. 111. Clinical Features of the "Antiphospholipid Syndrome
A. CARDIOVASCULAR MANIFESTATIONS APS is associated with a number of clinical manifestations affecting multiple organs. Although the common pathophysiological theme for organ damage appears to be thrombotic microangiopathy, there are other clinical manifestations that cannot be explained by this, e.g., cardiac valvular abnormalities. Patients with APS can have both arterial and venous thrombosis, the only clinical condition that predisposes to this without any structural vascular anomalies. Other inherited conditions tend to predispose to thrombosis in one vascular bed, e.g., homocystinemia and arterial thrombosis, and a number of familial protein deficiencies and genetic mutations that predispose to venous thrombosis. In younger patients and patients with PAPS, the vascular event is often an acute occlusive vascular event instead of being secondary to atherosclerosis. Patients with SLE and secondary APS, however, may have a combination of thrombotic diathesis and atherosclerosis related to other factors, such as long-term steroid administration, hyperlipidemia, and hypertension. Coronary vasculitis is less frequent. The prevalence rates in APS for myocardial infarction have been reported between 0 and 7% (Asherson et al., 1985). A Finnish study showed that the presence of high antibody levels in a standard CL-ELISA was an independent risk factor for myocardial infarction. Subjects with aCL levels in the highest quartile of distribution had a relative risk of myocardial infarction of 2.0 (95% confidence interval, 1.1to 3.5) independent of confounding factors normally associated with coronary vascular disease, such as smoking, age, systolic blood pressure, and hyperlipidemia (Vaaralaet al., 1995).In a series of 83 patients who had undergone coronary artery bypass graft surgery, autoantibodies detected in CL-ELISA were elevated in late bypass graft occlusions (Morton et al., 1986). A placebo group not treated with aspirin with these autoantibodies
ANTIPHOSPHOLIPID SYNDROME
513
had a high rate of coronary artery bypass graft occlusion (Gavaghan et nl., 1987). aPL antibodies may be associated with acute and chronic myocardial dysfunction. The clinical findings of left ventricular isolated and global dysfunction with the presence of insignificant coronary vessel disease as seen on angiography may be associated with aPL antibodies predisposing to coronary microangiopathy (Leung et al., 1990). This can also be found in the context of valvular heart disease in particular mitral regurgitation. The link between aPL antibodies and aseptic vegetations in patients with autoimmune disease was recognized in the 1980s (Anderson et al., 1987; Ford et al., 1988). Studies using echocardiography have shown that autoimmune patients with aPL antibodies have a higher prevalence of valvular vegetations (Khamashta et al., 1990; Cervera et al., 1992, Roldan et al., 1992).Although patients with SLE, especially if they are immunosuppressed with medication to control the disease activity, may have infective endocarditis, it appears that these patients have a higher prevalence of noninfective, thrombotic endocarditis. Linear deposition of IgG aCL antibodies in the subendothelial layer of heart valves in patients with APS has been demonstrated (Ziporen et al., 1996),suggesting a possible pathogenic role of these antibodies in valvular abnormalities. This would need to be studied more extensively in the future. B. NEUROLOGICAL MANIFESTATIONS The cerebral arterial circulation appears to be the most common site for arterial thrombotic episodes in patients with aPL antibodies (Harris et al., 1984). The Antiphospholipid Antibodes in the Stroke Study Group (APASS)found that the presence of autoantibodies above 10 GPL or MPL units in a standard CL-ELISA to be an independent risk factor for a first ischemic stroke in an elderly population without SLE (APASS, 1993). In a prospective study of stroke in patients below the age of 50 years, the risk of stroke recurrence was eight times higher in patients with aPL antibodies than those without (Brey et nl., 1990). It therefore appears that while the presence of aPL antibodies is an important factor to be evaluated for in the context of cerebrovascular events, the positive predictive value is greatest in patients below the age of 50 years. As in other vascular beds, the presence of cigarette smoking and hypercholesterolemia may independently increase the risk of recurrent cerebral ischemia in patients with aPL antibodies (Levine et al., 1990). In another prospective study of patients who presented with focal cerebral ischemia without any prior autoimmune disease, a titer of > 40 GPL units in a standard CL-ELISA conferred a twofold increased risk for a further thromboocclusive event (peripheral or central) or death. This is despite more of these individuals
514
DAVID A. KANDIAH et al.
receiving antiplatelet or anticoagulant therapy or both at the time of followup. These results imply that more specific methods need to be derived to stratify these high-risk patients and to maintain them on suitable therapy after a first vascular occlusive event (Levine et al., 1997). Limited cerebrovascular histopathological data suggest that the vascular abnormality in aPL syndrome is increased fibrin thrombi formation in small- and medium-sized vessels in the absence of vasculitis (Woodard et aZ.,1991).As discussed earlier in the context of cardiac valvular abnormalities, another source of vascular occlusion in patients with aPL antibodies is cardiac emboli, and one-third of the 72 patients studied by the APASS group found cardiac abnormalities in patients with aPL antibodies and cerebral ischemia, predominantly mitral valve abnormalities (APASS, 1990). Although epilepsy is a recognized clinical event in SLE patients, the etiology of this is multifactorial. Hypertension, infection, cerebral ischemia, and vasculitis have all been implicated in patients who develop epilepsy. Epilepsy as a primary neurological event in SLE patients was associated with a high prevalence of aPL antibodies (Herranz et al., 1994). In a study by Verrot et al. (1997), 163 patients with epilepsy were evaluated for autoantibodies in a standard CL-ELISA. The authors found 31 (19%) patients with IgG aCL antibodies of moderate to high titers. None of these patients had any previous clinical events to suggest APS. Brain imagmgs in these patients showed no significant difference in those who were aCL positive and negative (Verrot et al., 1997). Hence there appears to be a group of patients who have epilepsy as a primary clinical manifestation of the presence of aPL antibodies, independent of possible cerebral ischemia. Movement disorders have been associated with SLE initially (Lusins and Szilagyi, 1975) and subsequently with aPL antibodies (Asherson and Hughes, 1988). These movement disorders may be brought out in an estrogen-related hormonal environment, e.g.,in pregnancy, or if the patient was taking the oral contraceptive pill (Asherson et al., 1986a). Movement disorders in patients with aPL antibodies may also follow cerebral infarctions. Another neurological manifestation of aPL antibodies is migraine (Hughes et al., 1986),although the association at the moment is considered tenuous (Hering et al., 1991). Transverse myelopathy has also been described in autoimmune patients with aPL antibodies (Adrianakos et al., 1975).Although there have been a number of case reports of aPL antibodies being found in patients with transverse myelitis (Lavalle et al., 1990), the presence of these antibodies in patients with autoimmune disease may just be part of the spectrum of autoantibodies found in these patients and not be directly responsible for the clinical problem.
ANTIPHOSPHOLIPID SYNDROME
5 15
C. OCULAR ISCHEMIA aPL antibodies may be associated with thromboembolic disease in the visual pathway. A study of patients with cerebrovascular disease and APS found 19% (9/48) had clinical features of amaurosis fugax, ischemic optic neuropathy, and retinal artery occlusions (Levine et al., 1990). Various other studies and case reports have also described patients with aPL antibodies and ocular ischemia, often in the context of more generalized cerebral ischemia (APASS, 1990; Briley et al., 1989). The occurrence of amaurosis fugax in patients under the age of 50 or in patients with frequent episodes ranging from 2 to inore than 100 episodes a week may indicate that APS and aPL antibodies should be screened in these patients (APASS, 1993). A prospective study of 550 patients with SLE revealed that 7.5% of these patients had occlusive ocular vascular disease, and 38% of these patients had LA antibodies investigated by one clotting test only (APT") (Stafford-Brady et al., 1988). Patients presenting with headaches and found to have papilloedema in the presence of a normal cerebral CT scan may have cerebral venous thrombosis. This condition has been associated with aPL antibodies and should be screened for with multiple sensitive tests (Levine et al., 1987). The true incidence of aPL antibodies in optic ischemia and cerebral venous thrombosis has yet to be clearly defined, as more information is available on the detection methods for aPL antibodies. It appears, however, that in patients below the age of 50 years who present with ocular symptoms and have been found to have vasoocclusive disease should be screened for these antibodies by multiple tests.
D. PULMONARY MANIFESTATIONS Recurrent deep venous thromboses are the most common vasoocclusive events that occur in patients with aPL antibodies (Boey et al., 1983). Subsequent pulmonary emboli are not infrequent and often occur in the absence of symptomatic deep venous thromboses (DVTs) (Asherson and Cervera, 1992). Pulmonary hypertension in patients with APS has been documented, but this appears to be mainly associated with SLE and not with thrombotic disease (Asherson, 1990). It is therefore unclear whether SLE patients with pulmonary hypertension and aPL antibodies have the two conditions related or whether the high frequency of aPL antibodies in SLE may mask their true clinical relevance. Intraalveolar pulmonary hemorrhage has been described in patients with SLE and aPL antibodies (Howe et al., 1988). In a retrospective review of inpatients with intraalveolar pulmonary hemorrhage and SLE, six of the eight patients were found to have aCL antibodies (Schwab et al., 1993).
516
DAVID A. KANDIAH et al
This study identifies the problems encountered in retrospective clinical studies of patients with aPL antibodies as the current battery of tests for LA and anti-02GPI antibodies should be performed so as not to miss patients with aPL antibodies. This is particularly true of hemorrhage associated with aPL antibodies where antiprothrombin antibodies should be investigated to detect those high-affinity antibodies that form prothrombin-antiprothrombin complexes that are removed by the reticuloendothelial system. This results in liypoprothrombinemia and hemorrhage in some patients with these antibodies. Hence, pulmonary complications in APS are common but are often related to macrovascular thromboses as part of a systemic hypercoagulable state. Patients may also have adult respiratory distress syndrome as part of multiple organ involvement with extensive microangiopathic thromboses (Ghosh et al., 1993).
MANIFESTATIONS E. RENAL Primary renal disease is increasingly recognized in APS. Antiphospholipid antibody-related intrarenal thromboses may present with systemic hypertension, proteinuria, hematuria, and progressive renal failure, especially in the context of severe thrombotic microangiopathy in catastrophic APS (Asherson, 1993; Piette et al., 1994). Glomerular capillary thrombosis has been found to have a strong association with LA antibodies and predisposes to glomerular sclerosis independent of immune complex disease (Kant et nl., 1981). Renal disease may also arise in the APS with renal artery stenosis. The nature of this vascular occlusion is unclear and may arise as a primary thrombotic phenomenon in the context of aPL antibodies (Ostuni et al., 1990) or may be secondary to previous damage to the renal vessels from atheromatous degeneration in the blood vessel walls or prior renal artery fibromuscular dysplasia (Mandreoli et al., 1992). Renal vein thrombosis may be the cause and result of a thrombotic tendency in the APS. Thrombotic microangiopathy predisposing to the nephrotic syndrome can result in the loss of circulating natural anticoagulants predisposing to vascular thrombosis. However, the thrombophilia associated with aPL antibodies has been shown to occur in the absence of previous renal disease. Comparing two matched groups of SLE patients with renal disease, with and without LA antibodies, no differences were found in their renal biochemical and histological features except in a higher prevalence of intrarenal thromboses in patients with LA antibodies (Farrugia et al., 1992).
F. ADRENAL MANIFESTATIONS The link between adrenal gland hypofunction and aPL antibodies was first recognized in the late 1980s (Grottolo et al., 1988; Asherson and
ANTIPHOSPHOLIPID SYNDROME
517
Hughes, 1989). A number of possible theories exist for this association, including the primary occlusion of the adrenal veins leading to glandular edema and compression of the arterial blood supply and adrenal infarction. Asherson (1994) reviewed 38 cases with adrenal hypofunction and aPL antibodies and showed that 31 of the 38 patients had PAPS and 7 had SLE-associated APS. In 20 of these patients, vascular occlusive events preceded the adrenal hypofunction, whereas in 10 patients, concurrent events occurred in the context of acute adrenal failure. These events were predominantly venous in particular pulmonary emboli (Asherson, 1994). Therefore, patients with aPL antibodies who suddenly develop circulatory collapse need to be investigated for electrolyte disturbances and adrenal structure and function and treatment with adequate and prompt fluid replacement is essential. Marie et al. (1997)identified adrenal failure secondary to bilateral adrenal hemorrhagic infarctions in a 70-yearold patient as a first clinical manifestation of the PAPS who then went on to develop extensive upper limb deep venous thrombosis while on aspirin.
G. HEPATIC MANIFESTATIONS Structural and functional obstruction of venous blood flow in the liver may lead to Budd-Chiari syndrome. In the context of aPL antibodies, this may be the result of thrombosis in the hepatic veins extending to the inferior vena caw. The occurrence of Budd-Chiari syndrome and aPL antibodies was first reported in 1984 (Pomeroy et al., 1984). The majority of patients described with this syndrome in the presence of aPL antibodies have PAPS and have had previous venous occlusive disease and concurrent thrombocytopenia. Hepatic venoocclusive disease resulting in hepatomegaly and ascites secondary to central and sublobular vein occlusions occurs with aPL antibodies. The venous occlusions lead to hepatic sinusoidal congestion, hepatocellular necrosis, and finally fibrosis. The liver has a dual blood supply from the systemic and portal circulation. As such, hepatic infarction is rare unless the patient has generalized thrornbophilia. The first case of hepatic infarction in association with aPL antibodies was described in 1989 (Mor et al., 1989), and primary portal hypertension has been described in APS (Mackworth-Younget al., 1984). Thrombosis of mesenteric vessels has been described in APS resulting in intestinal infarction. Patients may have both arterial and venous occlusions. The presentation is usually with acute abdominal pain or “intestinal angina” (Asherson et al., 1986b). Other abdominal organs that have been described with vascular occlusion and infarction include the spleen (Arnold and Schrieber, 1988) and pancreas (Wang et al., 1992).
518
DAVID A. KANDIAH et a1
H. DERMATOLOGICAL MANIFESTATIONS In line with the common theme of vascular occlusion and insufficiency in end organ disease in APS, skin changes are fairly common in APS. Livedo reticularis, a mottled violaceous discoloration of the skin in a netlike pattern, is found frequently in APS. In patients with livedo reticularis and aPL antibodies, recurrent episodes of cerebral ischemia have been described (Sneddon, 1965).These patients may have a range of neurological manifestations from headache and dizziness, focal neurological deficits, and progressive cognitive deficits (from loss of concentration and memory loss to severe dementia). Livedo reticularis is found sufficiently frequently to be included in the clinical diagnostic criteria proposed (Alarcon-Segovia et al., 1992). Necrotic skin ulcers have been reported since 1963 in association with circulating LA antibodies (Bowieet al., 1963). Superficial cutaneous necrosis has also been described with aPL antibodies. This condition is also found in deficiencies of natural anticoagulants, e.g., protein C and protein S, or in cryoglobulineinia and cryofibrinogenemia,and these abnormalities may be found concurrently with aPL antibodies, with additive risks for the underlying superficial thrombosis. Digital gangrene may also be seen in APS (Alegre et nl., 1989). The histological features of small vessel disease producing the skin and soft tissue manifestations in APS appear to be a noninflammatory thrombosis of small arteries and veins throughout the dermis and subcutaneous fat tissue, occasionally accompanied by endarteritis obliterans. This condition is characterized by narrowing of the vascular lumen with endothelial cell proliferation and fibrohyalinization of the vessel wall (Alegre and Winkelmann, 1988). In livedo reticularis, skin biopsies rarely reveal thrombosis of the small vessels, and vessel wall hyperplasia may be the only histological feature seen. These skin lesions may frequently be the first sign of APS; up to 37% of patients with skin lesions and aPL antibodies develop multisystem thrombotic phenomena in the course of their disease (Alegre et al., 1989). This observation, performed before the antigen specificities of aPL antibodies were better defined, should lead to prospective studies investigating this association with multiple screening tests for aPL antibodies. This could determine the subset of patients who present with skin lesions, who could well go on to develop more severe systemic manifestations that may be prevented or &minished with specific antithrombotic therapy. I. AVASCULAR NECROSISOF BONE aPL antibodies may be associated with clinical avascular necrosis (AVN) of bone. There have been a number of cases in the literature of
ANTIP1IC)SPHOLIPID SYNDROME
519
patients with PAPS who have developed clinical AVN adjacent to various joints (Vela et al., 1991; Seleznick et al., 1991). The prevalence of AVN in patients with aPL antibodies is difficult to ascertain, as patients with SLE are often on corticosteroid therapy, which predisposes to this condition.
J. OBSTETRIC MANIFESTATIONS Lupus anticoagulant antibodies have been associated with pregnancy loss and intrauterine death (Nilsson et al., 1975). Others have shown the same association with aCL antibodies (Lockshin et al., 1985; Harris et al., 1986). It is often suggested that patients with aPL antibodies exist in a prothrombotic state that need some other trigger to precipitate a clinical event. This could well include surgery, oral contraceptive use, and pregnancy. Previous pregnancy failures are an important feature for predicting subsequent pregnancy failure. Autoimmune disease has a variable effect on pregnancy. It has been suggested that the underlying immune abnormality that permitted the development of the autoimmune disease or the autoantibodies that arise may be directly responsible for the fetal loss (Gleicher, 1994). The maternal effects of aPL antibodies in pregnancy are uncommon, but have been reported, including preeclampsia (Scott, 1987), chorea gravidarum (Lubbe and Walker, 1983), and cardopulmonary distress (Branch, 1990). Severe early onset preeclampsia and abruptio placentae may predispose to fetal complications in late pregnancy. However, the fetus appears to be at risk throughout the pregnancy, and detection of aPL antibodies appears to be a useful test in the investigation of autoimmune reproductive failure (Aoki et al., 1995). Abnormal uterine artery flow velocity may predict a poor outcome in cases of aPL antibodies (Caruso et al., 1993). The most common cause of pregnancy loss in the first trimester is chromosomal abnormalities, and this has not been adequately studied in patients with aPL antibodies. Cross-reactivity of aPL antibodies with villous trophoblast cell membrane phospholipids may expose these cells to cytotoxic maternal immune effector cells (Hasegawa et al., 1990; McCrae et al., 1993). One of the essential considerations in pregnancy loss associated with aPL antibodies is whether the aPL-related pregnancy loss(es) may have been triggered by a nonrelated earlier miscarriage in an immunologically susceptible individual. There may be a biphasic pattern of pregnancy loss with embryonal death by 8.5 weeks and fetal complications from week 14 (Goldstein, 1994). Animal models have suggested that aPL antibodes per se appear to predispose to increased fetal resorption (Blank et al., 1991). However, the results of these studies are not conclusive (Blank et al., 1994a and Silver et al., 1997). Late fetal death is the most commonly found obstetric complication in
520
DAVID A. KANDIAH ct nl.
APS. This has been attributed by early (Nilsson et al., 1975) and later studies (de Wolf et al., 1982) to placental infarction. Immune complex deposition on the trophoblast basement membrane has also been implicated in SLE-related fetal loss (Grennan et al., 1978). Elevated aPL levels have been associated with chronic uteroplacental vasculitis in the placental bed (Erlendsson et al., 1993).The occurrence of thrombosis and infarction in non-aPL fetal death, as well as inflammatory changes in aPL-related fetal death, suggests that the end-organ damage (placenta) in aPL disease is multifactorial and may have complex humoral and cellular interactions together with coagulation pathway abnormalities in the maternal circulation, villous trophoblast surface, and within the fetoplacental circulation. Antitrophoblast cytotoxicity initiated by maternally derived aPL antibodies cross-reactive with fetal trophoblast phospholipid epitopes and phosphatidylserine may induce chronic inflammation in the villi (Hasegawa et al., 1990). IV. p2-Glycoprotein I
PSGPI, a plasma protein,was first described in 1961 (Schultze et al., 1961) and has been the subject of extensive research in autoimmune disease. P2GPI is associated with different lipoprotein fractions in plasma and is also designated apolipoprotein H (Lee et at., 1983). PZGPI is a single-chain polypeptide of 326 amino acids with an apparent molecular mass of 50 kDa and is highly gIycosylated (Lozier et al., 1984).The carbohydrate content of P2GPI has been reported as being approximately 18%of its molecular mass (Schultze et al., 1961) and, when tested in phosphate buffer at pH 7.4, exists as 40% P sheet, 30% P turn, and 30% random coil (Walsh et at., 1990). P2GPI is a member of the complement control protein repeat (CCP) or short consensus repeat (SCR) superfamily (Reid and Day, 1989). The SCR is found in proteins involved in the regulation of the complement system (e.g., C4b-binding protein and factor H) and in some noncomplement proteins (selectin family and factor XIII). Although the first four of the five domains are typical examples of this CCP superfamily, the fifth domain is aberrant, containing an additional disulfide bond and a long C-terminal tail. P2GPI is highly conserved among mammalian species, suggesting that it plays an important physiological role (Kandiali and Krilis, 1994). Haptoglobin and factor H, two other members of this superfamily, are not bound by anti-P2GPI antibodies. Haptoglobin is used routinely as a control protein for antibody binding to PBGPI, as nonspecific binding can be detected when compared to p2GPI. p2GPI could not bind complement C3b coated on both activator (zymosan) and nonactivator (sheep
ANTII’HOSPHO1,IPID SYNDROME
521
erythrocytes) of the alternative complement pathway, whereas factor H bound to both surfaces coated with C3b, suggesting that despite structural similarities, these proteins had distinct nonoverlapping functions (Puurunen et nl., 1995). Although P2GPI has been characterized structurally, the tertiary structure of p2GPI and its biological function are not clear. p2GPI has a highly conserved pattern of cysteine residues (Steinkasserer et nl., 1991). Molecular modeling has suggested that a highly positively charged sequence in the fifth domain of P2GPI is surface exposed (Steinkasserer et nl., 1992; Sheng et nl., 1996). This had been predicted previously using the known tertiary structure of factor H, another CCP protein. This surfaceexposed net positive charge could well explain the binding of P2GPI to negatively charged surfaces, e.g., anionic phospholipids (Wurin, 1984), heparin (Polz, 1979), and DNA (Kroll et nl., 1976). Although PZGPI has been shown to be an absolute requirement for autoimmune aPL antibodies to bind in CL-ELISA, a preparation of P2GP1, proteolytically cleaved predominately between Lys317 and Thr318 in the fifth domain, lacked binding to anionic phospholipid (Hunt et al., 1993). This led to further work to map the major phospholipid-binding site on P2GPI initially with peptide inhibition studies (Hunt and Ki-ilis, 1994) and then with sitedirected niutagenesis (Sheng et al., 1996).The lysine-rich segment in the fifth domain ( Lyszxz-Lys2si)has been shown to be the major phospholipidbinding site on PZGPI. Modification of amino acid residues on P2GPI by potassium thiocyanate treatment completely destroys binding capacity, indicating the crucial involvement of lysine residues in the binding of P2GPI to anionic phospholipids (Kertesz et al., 1995). Many of the proposed physiological functions of p2GPI involve its phospholipid-binding properties. The binding of P2GPI to anionic phospholipids had been assessed using multilamellar, predominantly anionic phospholipid vesicles under nonequilibrium conditions (Wurm, 1984). Data suggested a high-affinity interaction of p2GPI with phospholipid in the 10-20 nM range. Zn vivo, however, physiological membranes contain significantly lower concentrations of anionic phospholipids, and normal plasma levels of P2GPI could easily displace Gla-containing proteins from cell membranes disrupting normal homeostatic mechanisms. Moreover, normal plasma concentrations of sodium and divalent cations would inarkedly inhibit this charge-dependent interaction. Physiological concentrations of PZGPI do not have much effect in in vitro coagulation tests unless anti-P2GPI antibodies are also present (Oosting et al., 1992; Roubey et al., 1992; Galli et al., 1992; Matsuda et al., 1993). Extrapolation of the calculation of the apparent dissociation constant for P2GPI binding with physiological anionic phospholipid membranes, to M, suggests that
522
DAVID A. KANDIAH et a1
P2GPI alone may not be able to displace other coagulation proteins from these membranes (Willems et al., 1996).However, the binding of a complex of an IgG anti-P2GPI molecule bivalently to two P2GPI molecules could have a markedly higher affinity for that anionic membrane in vivo and hence displace other coagulation proteins. In vitro, the presence of less procoagulant proteins will result in a delay in the clotting time, a plausible explanation for the lupus anticoagulant phenomenon. Excess phospholipid will allow the capture of other coaguIation proteins again and restore the clotting time back to normal levels. As the target antigen of pathogenic antibodies in APS, much research has gone into studying the interaction of these antibodies with P2GPI. Three hypotheses have been proposed to explain the interactions between P2GPI and anionic phospholipids, allowing subsequent binding of aPL antibodies. They are (1) a shared epitope on the phospholipid-PZGPI complex (McNeil et al., 1990), ( 2 ) a cryptic epitope exposed on P2GPI when it interacts with anionic phospholipids (McNeil et al., 1990), and ( 3 ) increased density of PZGPI captured on the anionic phospholipid (Roubey et al., 1995). There has been increasing evidence to suggest that aPL antibodies bind preferentially to P2GPI immobilized on anionic phospholipids or certain synthetic surfaces (irradiated plates), whereas binding in the fluid phase is weak and often nondetectable. It is unlikely that the same epitope on P2GPI is exposed by different surface interactions, as shown by work done with monoclonal antibodies that are immunoreactive with p2GPI bound to anionic phospholipids and to irradiated plastic wells (Wang et al., 1995). The intrinsic low affinity of anti-P2GPI antibodies in APS is significantly enhanced when there is an increased density of P2GPI bound to a negatively charged surface. It has been found that the binding of purified monoclonal anti-p2GPI antibodies on y-irradiated polystyrene wells was higher than on untreated wells. This binding is most likely due to increased density of BSGPI, as the amount of iodinated p2GPI retained on irradiated wells after the same amount of protein was coated overnight was 200% higher than untreated wells (Kandiah and Krilis, 1996a). Matsuura and colleagues ( 1994) have reported that polyclonal human aCL antibodies and a monoclonal murine aCL antibody bound PZGPI coated on electron or y-irradiated microtiter wells but not on untreated wells. The degree of binding depended on the irradiation dose, and aCL binding to PZGPI adsorbed to these wells correlated well with that of P2GPI complexed to solid-phase CL. Antibodies binding to P2GPI on these irradiated wells were only competitively inhibited by the simultaneous addition of CL-coated latex beads mixed together with P2GP1, but were
ANTIPHOSPHOLJPID SYNDROME
523
unaffected by the addition of excess PBGPI, CL micelles, or CL-coated latex beads. These findings again support the hypothesis that aCL antibodies are low-affinity antibodies and that interaction of antibodies to P2GP1, the target antigen, requires capture of this protein to an appropriately charged surfice. Roubey et al. (1995)have has shown that Fab’ fragments of patient IgG demonstrated little or no binding to PZGPI on y-irradiated polystyrene wells, whereas the whole molecule bound to P2GP1, suggesting a critical role for antibody bivalency. A. EPITOPEMAPPINGOF PHOSPHOLIPIDA N D ANTIBODY-BINDING SITES ON 02-GLYCOPROTEIN 1 Using synthetic peptides spanning the fifth domain of PSGPI, the peptide sequence Cys281-Lys-Asn-Lys-Asp-Lys-Lys-Cys288 inhibited binding of P2GPI to anionic phospholipid in a dose-dependent manner (Hunt and Krilis, 1994). Removal of the flanking cysteines abolished the ability of the peptide to inhibit phospholipid binding of native PSGPI, suggesting that the tertiary structure of P2GPI is important for phospholipid binding. By site-directed mutagenesis of the Lys residues in this amino acid sequence, binding of P2GPI to anionic phospholipid was reduced to about 50% by substituting one Lys residue with an Asp residue and abolished binding with two and three substitutions in this amino acid sequence (Sheng et at., 1996). Using monoclonal antibodies derived from patients with APS and peptide inhibition studies, linear epitopes in the C-terminal end of the fifth domain of P2GPI were recognized by these antibodies (Wang et al., 1995).Constructing two kinds of plasmid expression vectors that express P2GPI and the fifth domain of P2GPI only, polyclonal human anti-PSGPI antibodies were shown to bind the fifth domain of P2GPI directly and could inhibit, in a dose-dependent manner, the binding of these polyclonal antibodies to whole molecule P2GPI coated on irradiated microtiter wells (Yang et al., 1997). Their results suggest that the antigenic epitope for antibody binding is in the fifth domain. B. MOLECULARMODELINGOF P2GPI MODULES 1. Introduction To determine and understand the function of PSGPI, it would be useful to know its three-dimensional (3D) structure. Regrettably, no experimentally determined structure of P2GPI is available. However, it has been shown that the 3D structure of a protein may be calculated with useful accuracy if its amino acid sequence is sufficiently simi!,i to that of a protein with a known 3D structure (Sanchez and Sali, 199713). This comparative or homology modeling technique is particularly
524
DAVID A. KANDIAH ct
01.
useful when only low to medium resolution results are required, such as prediction of exposed regions that may interact with antibodies (de la Paz et al., 1986; Sali et al., 1993) and models of interaction based on electrostatic complementarity (Salemine, 1976; Sali et al., 1993). This review describes a comparative modeling study of the “sushi” domains of P2GPIs from five mammalian species. In particular, the authors review their previously published model of the fifth module in human P2GPI (P2GPI-5) and data on the cardiolipin (CL)-binding site on its surface (Sheng et al., 1996). The relationship between the various P2GPI modules is also discussed.
2. Alignment of SCR Modules in P2GPl and Factor H Amino acid sequences are known for 24 modules in P2GPIs from five mammalian species, including human, bovine, dog, mouse, and rat (Table I). Each P2GPI consists of 5 modules, except for rat PSGPI, which consists of only 4 modules. It has been suggested that the p2GPI modules are related to the SCR modules of factor H (Reid and Day, 1989). Four medium-resolution 3D structures of two different SCR modules from factor H (Table I) have been determined by solution nuclear magnetic resonance ( N M R ) and deposited in the Brookhaven Protein Databank (PDB) (Abola et al., 1987). Structures of the factor H modules and sequences of the 24 P2GPI modules were compared manually to obtain the
TABLE I SOURCESOF STRUCTURAL AND SEQUENCE DATAUSEDIN COMPARATIVE MODELINGOF /32GPI MODULES‘ Name Factor H modules with 3D structures determined by NMR Factor H, module 16 Factor H , module 15 Factor H, modules 15-16 P2GPI sequence Human Bovine Rat Mouse DO%
PDB Code
lHCC lHFI lHFH
Reference
Norinan et al. (1991) Barlow et al. (1993) Barlow et nl. (1993) Kristensen et al. (1991) Bendixen et al. (1992) Aoyama et ril. (1989) Nonaka et nl. (1992) Sellar et al. (1985)
‘ Structures were obtained from the sunliner 1993 release of the Brookhaven Protein Databank (Abola et al., 1987). The deduced amino acid sequences of /32GPIs were obtained from the GenBaik database (Bilofsky and Burks, 1988),except for dog PZGPI, which was obtained from the original paper.
ANTIPHOSPHOLIPID SYNDROME
525
alignment in Fig. 1. Even though sequence identity between factor H modules and P2GPI modules 1-4 is only approximately 20%, their alignment appears to be relatively accurate because there are few gaps and because of the invariability of the two disulfide bonds. Similarly, the alignment of the core regions of factor H and the fifth modules in P2GPIs is strongly determined by the assumption that disulfide bonds 5-50 and 36-61 in H-15 are equivalent to disulfide bonds 3-54 and 39-64 in /32 GPI-5 (Steinkasserer et al., 1993), even though only 9-13 residues out of 62 are identical between factor H modules and p2GP1-5~.The only major ambiguity arises around the 5-residue insertion at residue 21 in human P2GPI-5. There are also three single residue insertions in human P2GPI5 relative to factor H-15 at positions 47 (loop), 51 (loop), and 57 (extended chain). A major difference between factor H modules and 62GPI-Tj modules is that the latter has a 19 residue addition at the C terminus.
3. Overall Sirnilarities anzong Factor H and p2GPl Modules To find the clustering of the modules in factor H and the five /32GPIs, the table of percentage sequence identities for all pairs of the modules was calculated from their alignment (Fig. 2). This matrix was used with the Kitsch computer program (Felsenstein, 1985) to calculate a tree that expresses the relationships among the sequences of the modules, similar to the trees used to deduce the evolution of protein families. In this tree, differences between two groups of sequences are approximated by a vertical distance from the top of the tree to the highest node from which the two groups of sequences branch off. The sequences cluster in six groups. There is one group with the factor H modules H-15 and H-16 and five groups each containing the modules with the same relative position in the five P2GPI sequences. This arrangement suggests that the missing module in rat PZGPI is probably the first module because the group of the first modules does not contain a member from the rat P2GPI. The tree indicates that the first event in evolution of a multidomain PZGPI was gene duplication, which separated P2GPI module 5 from the predecessor of the rest of the modules. This may have been followed by the consecutive appearances of modules 1 and 3, with the final duplication resulting in modules 2 and 4. 4. The Three-Dirnensional Model of the FiBh Domain of Hurnan /32GPI The template structure for comparative modeling of P2GPI-5 was that of the 15th domain of human factor H. H-15 conformation has been determined by solution NMR (Barlow et al., 1993) (PDB code 1HFH). The alignment between P2GPI-5 and H-15 (Fig. 1)was used as input for MODELLER-11 (Saliand Blundell, 1993; Sanchez and Sali, 19974,which
lHCC
"
*o
20
lHFl
60
EKIPCSQPPQIEHGTINSSRSSQ-- ---ESYAHGTKLSYTCEGGFR-ISEENETTCYM-GKWSS-PPQCE EGLPCKSPPEISHGVVAHM--SD------SYQYGEEVTYKCFEGFG-IDGPAIAKCLG-EKWSH-PPSCI
Bovine-1
GRTCPKPDDLPFSTVVPLKT--------FYEPGEEITYSCKPGYVSRGGMRKFICPLTGLWPINTLKCT GRTCPKPDELPFSTVVPLKR--------TYEPGEQIVFSCQPGYVSRGGIRRFTCPLTGLWPINTLKCM
Dog-l
GRTCPKPDDIPFATVVPLKT--------FYDPGEQIAYTCQPGYVFRGLTRRFTCPLTGVWPrNTVRCE
Mouse-1
GRICPKPDDLPFATVVPLKT---------SYDPGEQlVYSCKPGYVSRGGMRRFTCPLTG~lNTLRCV
Human-2
PRVCPFAGILENGAVRYT----------TFEYPNTISFSCNTGFYLNG-ADSAKCTEEGKWSPELPVCA
Bovine2
PRVCPFAGILENGTVRYT----------TFEYPNTISFSCHTGFYLKG-ASSAKCTEEGKWSPDLPVCA PRVCPFAGILENGAVRYT----------TFEYPNTISFACNTGFYLNG-SSSAKCTEEGKWSVDLPVCT PRVCPFAGILENGVVRYT----------TFEYPNTIGFACNPGYYLNG-TSSSKCTEEGKWSE-LPVCA PRVCPFAGILENGIVRYT----------SFEYPKNISFACNPGFFLNG-TSSSKCTEEGKWSPDIPACA PIICPPPSIPTFATLRVYKPSAGN~---NSLYRDTAVFECLPQHAMFG-NDTITCTTHGNWTK-LPECR PITCPPPPIPKFASLSVYKPLAGN----NSFYGSKAVFKCLPHHAMFG-NDTVTCTEHGNWTQ-LPECR RVTCPPPSVPKFATLSVFKPLATN----NSLYGNKAVFECLPHYAMFG-NDTITCTAHGNWTT-LPECR
Human-l
Dog-2 Rat-2 Mouse-2
Human-3 Bovine3
Dog-3 Rat-3
RlTCPPPPlPKFAALKEYKTSVGN---.SSFYQDTVVFKCLPHFAMFG-NDTVTCTAHGNWTQ-LPECR
Mouse-3
RlTCPPPPVPKFALLKDYRPSAGN----NSLYQDTVVFKCLPHFAMIG-NDTVMCTEQGN~R-LPECL
Human-4
EVKCPFPSRPDNGFVNYPAKP-------TLYYKDKATFGCHDGYSLDG-PEEIECTKLGNWSA-MPSCK EVRCPFPSRPDNGFVNHPANP--------VLYYKDTATFGCHETYSLDG-PEEVECSKFGN~A-QPSCK EVKCPFPSRPDNGFVNYPAKQ-------ILYYKDKAMYGCHDTYTLDG-PEVVECNKFGN~A-QPSCK EVKCPFPSRPDNGFVNYPAKP-------VLSYKDKAVFGCHETYKLDG-PEEVECTKTGN~A-LPSCK EVKCPFPPRPENGYVNYPAKP-------VLLYKDKATFGCHETYKLDG-PEEAECTKTGAWSF-LPTCR
B0"l"ed
Dog-4 Rat4 Moue-4
111
Human-5 Bovine5
Dog-5 Rat-5 Mouse-5 PDB-lH(C :
BBB
AF
BBBB
PhD
BBBBBBBBBBB
JMC
BBBB
Homolog GOR
a0
30
ZO
* **
so
70
60
80
ASCKLPVKKATVVYQGERVKIQEKFKNGMLHGDKVSFFCKNKEKKCSYTEDAQCID-GTIE--VPKCFKEHSSLAF~TOASDVKPC ASCKLSIKRATVIYEGERVAIQNKFKNGMLHGQKVSFFCKHKEKKCSYTEDAQCID-GTIE--IPKCFKEHSSLAF~TDASDVKPC ASCKLSVKKATVLYQGERVKLQE~FKDGMLHGQKVSFYCKNKEKKCSYTEDAECID-GTIE--IPKCFKEHSSLAF~TDASDVKPC ASCKLSVKKATVLYQGQRVKIQDQFKNGMMHGDKVHFYCKNKEKKCSYTEEAQCID-GTIE--IPKCFKEHSSLAFWKTDASDVTPC ESCKLPVKKATVLYQGNRVKIQEQFKNGMMHGDKIHFYCKNKEKKCSYTVEAHCRD-GTIE--IPSCFKEHSSLAFVM(TDASELTPC
BBBB
B
HH
H
EBB
BBBBBBBB
B BB
BBBBBB
BBBBB BBBBBBBBB
BBBBB
BBB EBB
HHHHH
BB
HHH
BBB
HHHHHHHHHH H
HH
H
BBBB
BHHHHH HHHHH
HHHHHHHHHH
B
BBB
B
BBBBBB HHHH
BB
BB H H
HHHHHH
527
ANTIPHOSPHOLIPID SYNDROME
100 L
80 k
z w
P
zz w
3
40
2 u)
s
I
15
FIG. 2. Clustering of the SCR modules from human factor H and five mammalian P2GPIs.
produced a model of P2GPI-5 containing all main chain and side chain nonhydrogen atoms. [Modeller is available at URLhttp://guitar.rockefeller.edu:pub/modeller and also as part of Quanta InsightII, and GeneExplorer (MSI, San Diego, CA, USA; e-mail
[email protected])]. The standard automated modeling procedure was used, except that additional distance restraints were imposed on the 19 residue extension at the C terminus. These restraints were obtained as follows. First, four secondary structure prediction methods were applied to human P2GPI-5 (Fig. 1).Three of the four methods resulted in an approximately correct
Frc;. 1. Alignment of the amino acid sequences of the modules from five P2GPIs and modules 15 and 16 from human factor H. The top line refers to the residues in 1HFI. The stars indicate the Lys residues that were mutated to the Glu residues. The line PDB-1HFI contains the secondary structure assignments for lHCC from the corresponding PDB (protein data bank) file. The predictions by the following secondary structure prediction methods are shown: AF, a method based on the physicochemical properties of the residues (Ptitsyn and Finkelstein, 1983); PhD, a neural network method (Rost and Sander, 1993); JMC, a neural network method (Chandonia and Karplus, 1996); Homolog, a method based on residue statistics (Biou et al., 1988); and GOR, a method based on the residue statistics (Biou et ul., 1988). The secondary structure predictions are indicated by H for helix and B for strand.
528
DAVID A. KANDIAH et a[
P
structure content, and all three of these predicted a strand starting at position 71 within the 19 residue C-terminal extension. The method that did not predict a P strand in this region also gave an incorrect secondary structure composition, with many residues predicted as helical. On the basis of these results, the authors predicted that region 71-75 in p2GPI5 was indeed a p strand. The partner of this strand must be one of the known strands in the rest of the molecule. An examination of the preliminary model suggested that the only two possible partners are strands from residue 4-10 and from residue 57-63, with strand 71-75 lying between and/or on top of these two strands. Any other possibility would have involved main chain knots or breaking the disulfide bond to the C-terniinaI Cys residue. Given the assumption that strand 71-75 interacts with strands 4-10 and/or 57-63, there are still four possibilities for the register of the predicted strand with the existing strands. These were explored by generating 3D models that satisfied as well as possible 24 lower distance bounds on the interstrand C ff distances, each restraint set corresponding to one of the four strand registers. The best model was then identified as the model that had minimal restraint violations, best 3D model quality index of Eisenberg and co-workers (Luthy et al., 1992), and the smallest number of residues other than Gly and Asn that had positive angles (Fig. 83, see color insert). The 3D-PROFILES quality index of this representative P2GPI-5 model is 21.5, which is within the allowed range for the protein of the same size as 62GPI (Luthy et al., 1992).This quality index can be compared with the quality indices for the experimental structures of the H-15 and H-16 modules in PDB files lHFI, lHCC, 1HFH-15, and 1HFH16, which are generally, but not always, higher at 30.3, 21.0, 30.1, and 24.4, respectively. The fold of the p2GPI-5 model consists of eight strands, organized in two distorted P sheets with long coiled regions connecting the strands (Fig. 3). There are no helices.
5. Electrostatic Properties of Human P2GPI-5 Electrostatic terms in the potential energy often give rise to specific interactions in complexes (e.g., that between a Lys and a sulfate at contact distance). However, in order to understand or to predict the nature of a complex between two molecules, it is often useful to look at their global electrostatic potential. If the structure of only one ligand is known, it is particularly helpful to examine its electrostatic potential for possible binding sites of the other ligand. This is true in the present case where the interaction between a positive (the protein) and a negative (cardiolipin) ligand is considered and the detailed structure of cardiolipin is not available. Thus, to investigate more closely which particular amino acid residues are critical for phospholipid binding by the intact fifth domain of P2GP1,
ANTIPHOSPHOLIPID SYNDROME
529
electrostatic properties of the 3D model of the fifth domain of human P2GPI (Fig. 8-4, see color insert) were examined. The electrostatic potential on the surface of P2GPI-5 and its mutants was calculated with GRASP (Nicholls et al., 1991), a computer program that uses the finite difference method to solve the linearized PoissonBoltzmann equation. A net charge of -1 was assigned to each Asp and Glu residue and a net charge of +1 to each Lys and Arg residue. Each His was assigned a neutral charge because P2GPI is active in plasma at a pH of about 7.2. Models with all hydrogen atoms and partial charges from the CHARMM-22 force field were used for electrostatic calculations. Although there are considerable uncertainties in the positions of positive charges at the end of long Lys and Arg side chains on the protein surface, these have a small effect on the global features of the electrostatic potential considered below (Sali et nl., 1993). Most of the positively charged side chains (14 out of 16) are located on the surface of two regions. The first of these regions is defined by segments 40-46,63-66, and 81-84 (top face of the module in Fig. 4A). The second region is defined by one long and wide omega loop 3-28 (left face in Fig. 4A). Most of the negatively charged residues (8 out of 11) are located in segments 33, 50-62, and 67-80 (right face in Fig. 4A). The pronounced positive electrostatic potential above the top region in P2GPI-5 is predicted to be significantly reduced if any one of the three Lys residues in the center of the top region is mutated to the glutamic acid residue (Fig. 4B). The sequences of the fifth modules from the five species are highly similar (- 80%),which is reflected in the similarities of the charge distribution and of the electrostatic potentials. For example, the central segment Cys39Cys46 of the top region is identical in all five species, as is Lys66, whereas Lys82 is present in three of the five species.
6. Location uf the Cardiolipin-Binding Region in Human P2GPl-5 Both positively charged faces on P2GPI-5 are likely to attract negatively charged ligands such as cardiolipin (Fig. 4A). However, because the top positively charged face contains peptide 40-46, which is known to bind cardiolipin (Hunt and Krilis, 1994),the three central charges in this particular region are predicted to be part of the binding site for cardiolipin in the intact P2GPI-5 domain. Moreover, the mutation of these three Lys residues to Glu residues is predicted to prevent the interaction between CL and P2GPI-5 (Fig. 4B). These predictions are similar to those made based on homology modeling alone (Steinkasserer et al., 1991, 1992). To test this prediction, the cDNA for human P2GPI was inserted into the baculovirus viral DNA BacPAK 6 for expression in insect cells (Sf21) (Sheng et al., 1996). As discussed previously, site-directed mutagenesis
530
DAVID A. KANDIAH et al.
was then performed to assess the role of the individual amino acids in the Lys40-Lys45 loop in phospholipid binding and anti-P2GPI activity. It was found that residues Lys42, Lys44, and Lys45 were indeed critical for p2GPI binding to anionic phospholipids, but not crucial for direct binding of P2GPI by anti-P2GPI antibodies. As mentioned earlier, it has been shown that cardiolipin binds to an isolated peptide Cys39-Cys46 with the two flanking Cys residues, but not to Lys40-Lys45 or to Ser39-Ser46, where the flanking Cys residues were replaced by Ser (Hunt and Krilis, 1994).Thus, the conformation of segment 40-46 is likely to be critical for phospholipid binding. It appears that the flanking Cys residues form a disulfide bond that favors the peptide conformation in the peptide-phospholipid complex, thus increasing the free energy of binding via reducing its entropy. This is explained by the 3D model ofP2GPI-5 as follows. Even though the two flanking Cys residues are not disulfide bonded to each other in the native molecule, their relative position in the model is consistent with such a bond (Fig. 3). As a consequence, a nonnative disulfide bond between Cys 39 and Cys 47 is expected to favor the native conformation for the intervening peptide segment. The model for interaction is not sufficiently detailed to distinguish between a specific electrostatic interaction that requires a certain peptide sequence and an interaction that relies on charge density without many steric restrictions. Nevertheless, the model did serve as the basis for informed sitedirected mutagenesis experiments that provided more information on the binding of phospholipids to P2GPI. AND CHARACTERIZATION OF THE GENEENCODING C. CLONING MOUSEP2GPI
A mouse ES genomic library in the bacteriophage P1 cloning system was screened using polymerase chain reaction (PCR). A Hind111 fragment was shown to contain the entire mouse P2GPI gene and was ligated into the pBluescript SK vector for further analysis and sequencing. This plasmid clone was digested with different restriction enzymes, and some fragments were further subcloned into pBluescript SK vectors for sequencing. The mouse P2GPI gene was subsequently found to be encoded by eight exons spread over about 18 kb of genomic DNA. Exon 1 contained the 5’untranslated region, the 19 amino acid long signal peptide, and the first 2 amino acids of the mature p2GPI protein. Exons 2-7 contain the rest of the protein-coding sequences. Exon 8 contained the last 19 codons and the entire 3’-untranslated region. The exons correlate well with the structural domains. CCPl is encoded by exon 11, CCP I1 by exons 111 and IV, CCP I11 by exon V, CCP IV by exon VI, and CCP V by exons VII and
531
ANTIPHOSPHOLIPID SYNDROME
VIII (Sheng et al., 1997) (Fig. 5). The mouse PZGPI gene has been localized to distal chromosome 11 (Nonaka et al., 1992), whereas the human P2GPI gene has been assigned to chromosome 17 (Haagerup et al., 1991). However, comparative mapping of human and mouse genomes has shown that the mouse distal chromosome 11 has extensive homology with human chromosome 17 (Buchberg et al., 1989). The amino acid sequence of p2GPI for mammalian species dwovered so far reveals a large degree of homology to the human sequence: mouse (76.1%), bovine (83%), and rat (80%).Alignment of these sequences shows that the fifth domain is the most highly conserved, suggesting that the main functional activities of the protein are present here. ' V. lmmunogenicity and Animal Models
In order to determine if autoantibodies are pathogenic in vivo, suitable animal models need to be studied. A murine model of autoimmune vascular disease (NZW X BXSB/F1) was first described (Hang et al., 1981). Other autoimmune strains predisposed to lupus were studied and showed autoantibodies reactive in a standard CL-ELISA (Gharavi et al., 1989). NZW X BXSB/Fl mice also have thrombocytopenia and were found to have anti-
A
I
ATG
I
AATAAA
STOP
FIG.5. Organization of the mouse PZGPI gene. (A) The structure of the mouse PZGPI gene is shown with restriction enzyme sites. The positions of exons are shown as boxes, and the introns are shown as lines connecting the exons. Restriction sites indicated: X (XhoI), B (BamHI), E (EcoRV), and Xb (XbaI). (B) SCR repeat domain structure of P2GPI. The positions of the translation initiation site (ATG), the polyadenylation site (AATAAA),and the termination codon are indicated. S, signal peptide (Sheng et al., 1997)
532
DAVID A. KANDlAH et 01.
bodies to P2GPI similar to that seen in autoimmune human APS patients (Hashimoto et al., 1992). Two groups have suggested that aPL antibodies have a direct role in causing pregnancy loss in vivo. Passive immunization of normal pregnant mice with human polyclonal (Branch et al., 1990; Blank et al., 1991) antibodies or monoclonal aPL antibodies (Bakimer et al., 1992) have been shown to result in increased fetal loss and fetal resorption and lower mean weights of embryos and placentas compared with mice immunized with normal immunoglobulins. However, one of these groups have since suggested that passive immunization with human IgG polyclonal aPL antibodies had variable effects on murine pregnancy outcome. The rate of fetal death did not increase uniformly with increasing doses of IgG and was unrelated to the individual patient’s medical history (Silver et al., 1997). BALBk mice immunized with CL mixed with PZGPI, CL alone, PZGPI alone, or buffer alone were studied. Mice immunized with CL mixed with PSGPI produced high levels of anti-P2GPI antibodies and antibodies reactive in a standard CL-ELISA. Mice immunized with CL alone did not produce aPL antibodies, and mice immunized with PSGPI alone produced anti-02GPI antibodies (Rauch and Janoff, 1992). Another group suggested that immunization of mice with P2GPI produced a high percentage of fetal resorption in utero when the mice were mated, suggesting an induced model of the APS (Blank et al., 1994b).To study the issue of pathogenesis and thrombosis in an animal model, ~nechanicalstimulus of exposed femoral veins in CD-1 mice was used to promote clot formation. Mice actively immunized with PZGPI and human IgG aPL antibodies from patients with APS developed propagation of the clot and slower dissolution (Pierangeli et al., 1996). Immunization of MLW+ + mice with PZGPI produced aPL antibodies. The development of neurological dysfunction and production of antinuclear and anti-DNA antibodies was controversial, with two opposing conclusions (Cote et al., 1994; Aron et al., 1995). Further research in this area needs to be done as another study has suggested that immunization of BALB/c mice with a monoclonal human aPL antibody (H-3) induces neurological and behavioral defects (Ziporen et al., 1997). Polyclonal antibodies purified from patients with APS, with a PZGPI affinity column, have binding characteristics similar to anti-PZGPI antibodies induced by immunization of a rabbit with human PZGPI. Some of the polyclonal human autoantibodies bound both PZGPI and anionic phospholipids. The binding to anionic phospholipids involves ionic interactions as the binding was reduced significantlyin the presence of high ionic strength buffers (Kouts et al., 1995). Nine monoclonal antibodies derived from NZW X BXSBF1 mice had two populations of antibodies with P2GPI reactivity and anionic phospholipid reactivity in the absence of p2GPI.
ANTIPHOSPHOLIPID SYNDROME
533
These latter antibodies, as with the polyclonal human antibodies, had charge-dependent binding to the anionic phospholipid with abolishment of binding in the presence of high ionic strength buffers (Monestier et al., 1996). Anti-P2GPI antibodies had a clear preference for purified murine P2GPI in a fashion similar to the preference human polyclonal antibodies from patients with APS had for purified human PSGPI. The analysis of the V region sequences of these antibodies suggest that cationic residues in the H chain CDR3 are important for their charge-dependent phospholipid reactivity. Sequence analysis of one of the monoclonal antibodies that recognized P2GPI in a phospholipid-free system, with little change in binding in the presence of high ionic strength buffers, did not reveal any cationic amino acid residues. The structural features of the VH-D-JH junctions of these monoclonal autoantibodies further support the view that an increased frequency of unusual V(D )J rearrangements contribute directly to the development of murine autoimmunity (Monestier et nl., 1996). The presence of these animal models of APS allow for the comparative study of the mechanisms of action of autoantibodies and induced antibodies. This will allow the continuing study of therapeutic interventions that may prevent the clinical manifestations of APS. VI. Prothrombin
Lupus anticoagulant antibodies could potentially inhibit any of four procoagulant phospholipid complexes or two anticoagulant phospholipid reactions. A number of early reports suggested a role for plasma proteins in the activity of lupus anticoagulants. The lupus anticoagulant “cofactor” phenomenon, i.e., the addition of normal plasma to patient plasma, increasing the inhibition of coagulation, was first attributed to the presence of prothrombin (Loeliger, 1959).Autoantibodies to prothrombin were shown in two patients with the lupus anticoagulant-hypoprothrombinemiasyndrome (Bajaj et al., 1983). Circulating prothrombin complexes were found in 74% of patients with LA antibodies and normal prothrombin levels (Fleck et al., 1988). The authors and others have shown that LA can react with human prothrombin directly on phospholipid-free, high-binding (irradiated) ELISA plates (Arvieux et al., 1995, Kandiah and Krilis, 19974, phospholipid-bound prothrombin (Bevers et al., 1991), and phospholipid alone (McNeil et al., 1989; Pierangeli et al., 1993; Kandiah and Krilis, 199713). Antiprothrombin antibodies have been found to have immunological prediction of myocardial infarction in men (Vaarala et al., 1996). In a study of 233 patients with aPL antibodies, 26% had IgG and/or IgM
534
DAVID A. KANDIAH et a1
antiprothrombin antibodies. There was poor correlation between antiprothrombin and anticardiolipin and anti-P2GPI antibodies in this same patient population. Univariate analysis suggested that antiprothrombin IgG correlated well with a history of venous thrombosis, but this effect was lost in the multivariate analysis, whereas anti-PBGPI IgG was the only variable that showed statistical significance ( Forastiero et al., 1997). In another retrospective study of SLE patients with aPL antibodies, the presence of LA antibodies was the only variable that had statistical significance in the multivariate analysis of association with venous thrombosis (Horbach et al., 1996). The varying results obtained by different groups on the prevalence and pathological links of antiprothrombin antibodies cannot be explained by the patient population studied alone. The method of performance of the antiprothrombin-ELISA is different in different studies, varying from the buffers used to dilute the samples and block the prothrombin-coated wells, to the cutoff levels determined. Some investigators use buffer-only wells as controls, which may be important, as it deducts nonspecific binding that can occur with the high binding plates used. If any meta-analysis is to be performed on these studies, to make generalizable deductions on the role of antiprothrombin antibodies, this important variable would need to be considered. It has been shown that LA antibodies in some patients with APS can be separated into antibodies positive in the d R W clotting assay and the dKCT. The immunoreactivity of these separate populations of autoantibodies cannot be explained by their immunoreactivity to P2GPI or prothrombin (Kandiah and Krilis, 1997b). Affinity-purified antiprothrombin antibodies from different patients had different reactivities in these two clotting assays (Kandiah and Krilis, 1997a).This observation was in variance with indirect studies on plasma reactivities, which suggested that anti-P2GPI reactivity corresponded to a prolongation in the d R W assay and antiprothrombin reactivity with prolongation in the dKCT assay (Gdli et al., 1995), but was supported by another study on a large population of patients that did not find a difference in the plasma reactivities in the dRWT and dKCT clotting assays and their anti-p2GPI and antiprothrombin reactivities ( Forastiero et al., 1997). The anticoagulant activity of antiprothrombin antibodies appears to be dependent on their recognition of a phospholipid-human prothrombin complex that inhibits both the conversion of prothrombin into thrombin in the prothrombinase complex (Bevers et al., 1991) and the tenase complex (Permpikul et al., 1994). There appears to be a high species specificity to human prothrombin in the functional assays (Bevers et al., 1991; Rao et al., 1995). LA IgG from two patients inhibited the activation of human but not bovine prothrombin in a purified prothrombin activation system
ANTIPHOSPIIOLIPID SYNDROME
535
(Bevers et al., 1991), whereas LA IgG from a third patient inhibited the activation of both human and bovine prothrombin (Galli et al., 1993). This is in variance to the immunoreactivity of the purified antibodies that recognized both human and bovine prothrombin coated on microtiter wells and on a Western blot, although the binding to human prothrombin was substantially higher in 6 of the 14 preparations studied. Twelve of the preparations showed a significantly increased binding to human prothrombin and 9 to bovine prothrombin in the presence of phosphatidylserine and calcium ions. Further experiments with phosphatidylserine/phosphatidylcholine ( P S R C ) vesicles, soluble prothrombin, and LA IgG failed to explain why LA IgG inhibits human prothrombin activation more effectively than it inhibits bovine prothrombin activation (Rao et al., 1995). In a study of 59 patient plasmas with aPL antibodies, 90% showed reactivity to prothrombin bound to phosphatidylserine in the presence of calcium, whereas only 58% of these plasmas had reactivity to prothrombin coated directly on high binding wells (Galli et al., 1997). These authors suggested that the mode of presentation of prothrombin in solid-phase influenced its recognition by antiprothrombin antibodies. They postulated that these differences were produced either due to clustering and conformational orientation of the prothrombin bound to phosphatidylserine, allowing better capture of the antibodies, or that the capture of prothrombin-antiprothrombin complexes may be better in the presence of calcium ions. This may also be due to the patient population studied, as in patients with antiprothrombin antibodies, the binding to prothrombin coated on irradiated surfaces in the absence of calcium ions was significantly higher than for prothrombin bound to phosphatidylserine in the presence of calcium. This applied to both plasma samples, as in the previous study as well as to affinity-purified antiprothrombin antibodies through a prothrombin column. The dissociation constant calculated for these antibodies was in the region of 200 nM, which showed about 10 times higher affinity than anti-/32GPI antibodies purified from a P2GPI column (Kandiah and Krilis, 1997a). Both studies, however, confirmed the heterogeneity of antiprothrombin antibodies in coagulation assays with no one assay detecting these antibodies consistently. VII. lupus Anticoagulant Antibodies and Protein C Activation
LA antibodies have been shown to have multiple effects on protein C. Results in the literature have been contradictory, with some researchers finding a significant inhibition on the rate of activation of protein C by thrombin on endothelid cells by purified LA IgG (Cariou et al., 1988), whereas others could not confirm this (Oosting et al., 1991; Keeling et al.,
536
DAVID A. KANDIAH et ul
1993). LA antibodies have also been shown to prevent the inactivation of factor Va by protein C. These appeared to be IgGs directed against negatively charged phospholipid-protein complexes of either protein C or protein S (Oosting et al., 1993). APC resistance (i.e., the association of dysfunctional APC with a venous thrombotic tendency) predominates in LA plasma, but is not restricted to the presence of the Arg506-Gln point mutation on factor V (Bokarewa et al., 1995). VIII. lupus Anticoagulant Antibadies and Phosphaiidylethanolamine
The presence of phosphatidylethanolamine (PE) has been shown to augment LA activity and inhibit the anticoagulant effect of activated protein C (APC) in vitro. This effect appeared to arise from interference of LA antibodies with APC activity by binding to PE or the complex of APC and PE (Smirnov et nl., 1995). aPE antibodies have also been shown to require plasma cofactors in their binding to PE, including high and low molecular weight kininogens (HMWK and LMWK) and, less frequently, prekallikrein and factor XI (Sugi and McIntyre, 1995). Kininogens inhibit thrombin-induced platelet aggregation. Kininogendependent IgG aPE markedly increased thrombin-induced platelet aggregation in vitro, whereas kininogen-independent IgG aPE did not (Sugi and McIntyre, 1996). Hence, kininogen-dependent aPE could cause thrombosis in vivo by disrupting the antithroinbotic effects of kininogen. As PE can increase the procoagulant activity of vesicles containing PS and PC, and aPE has been associated with thromboembolic events, the pathogenesis of the thrombosis in these patients is multifactorial. PE undergoes the transition from lamellar to hexagonal I1 phase under certain physiological conditions, and mice immunized with hexagonal PE develop phospholipid-dependent inhibitors of coagulation (LA antibodies) (Rauch and Janoff, 1990). Preincubation of aPL-positive plasma with hexagonal-phase I1 PE has been shown to reduce or even abolish the prolongation of clotting times in phospholipid-dependent coagulation assays, suggesting that this phospholipid is the target antigen for some LA antibodies (Rauch et al., 1989). These authors have suggested that the target antigen for these antibodies may be a complex of PE and human prothrombin (Rauch et al., 1997), although these experiments have been performed in clotting assays and not in a purified system. IX. Antiphospholipid Antibodies and Endothelial Cells
When a more physiological surface, such as endothelial cells, is used for the assembly of the prothrombinase complex, only 18% of IgG fractions
ANTIPHOSPHOLIPID SYNDROME
537
with LA activity were able to inhibit prothrombinase activity (Oosting et al., 1993).Endothelial cells play a central role in the prevention of unwanted activation of the coagulation cascade in intact vessels, This antithrombogenic property of endothelial cell surfaces responds to physiological stimuli and is therefore susceptible to injur>i.aPL antibody-positive SLE sera, but not purified antibody, in the presence of low doses of tumor necrosis factor (TNF)stimulated procoagulant activity by cultured endothelial cells. However, no association was found with clinical thrombotic events (Hasselaar et al., 1989). A high prevalence of antiendothelial cell (AECA)-binding activity is found in sera from patients with APS (Hasselaar et al., 1990; Del Papa et al., 1992). P2GPI is able to bind resting endothelial cells and be recognized by monoclonal and polyclonal anti-P2GPI antibodies (Del Papa et al., 1995; Le Tonqueze et al., 1995). Although platelet binding has been related to the expression of anionic phospholipids on their cell membranes after activation, a comparable phenomenon is unlikely on resting endothelial cells that do not display such phospholipid distribution changes (Del Papa et al., 1992). Both polyclonal and monoclonal antiPSGPI antibodies can upregulate adhesion molecule expression after endothelial cell binding (Del Papa et al., 199s; Del Papa et al., 1997). Del Papa et nl. (1998) showed that P2GPI binds endothelial cell membranes through its fifth domain. The major phospholipid-binding site that mediates the binding of PZGPI to anionic phospholipids is also involved in endothelial binding (Fig. 6). Human umbilical vein endothelial cell (HUVEC) monolayers provide a suitable surface for P2GPI binding comparable to that displayed by anionic phospholipids dried on microtiter wells. The formation of p2GPI and anti-P2GPI complexes induces endothelial activation as supported by E-selectin expression and IL-6 secretion (Del Papa et d., 1998) (Fig. 6). X. Pathogenesis of the Antiphospholipid Syndrome
Patients with APS have a tendency to atherogenesis that is likely related to the multiple immunological abnormalities that occur in this condition. The oxidation of plastic microtiter plates that increases the capture of the target antigens for aPL antibodies may be an in vitro model of the vascular inflammatory processes that result in a high oxidative capacity in vascular walls. Oxidation of plasma proteins and oxygen-mediated endothelial injury decrease the physiological anticoagulant function of endothelium (Vaarala, 1997). Antibodies to oxidized low density lipoproteins (LDL) are associated with carotid atherosclerosis (Salonen et nl., 1992) and myocardial infarction (Puurunen et al., 1994). In patients with SLE, these antibodies have been
538
DAVID A. KANDIAH et a1
FIG.6. PZGPI, anti-PZGPI, and endothelial cells. Anti-PSGPI antibodies bind a synthetic peptide spanning the fifth domain of PZGPI after capture on activated endothelial cells. PZGPI binds activated endothelial cells through the major phospholipid-binding site, KNKEKK (Del Papa et al., 1998). Human polyclonal anti-pSGPI antibodies bind to the same peptide sequence previously shown to support human monoclonal antibody binding after PSGPI bound to a negatively charged surface (Wang et al., 1995). Antibody binding upregulates adhesion molecule (E-selectin) expression and IL-6 secretion (Del Papa et al., 1995, 1998).
shown to cross-react with autoantibodies detected in the standard cardiolipin ELISA to PZGPI (Vaarala et al., 1993). Monoclonal anti-/32GPI antibodies derived from NZW x BXSB/F, mice also cross-react with oxidized LDL (Mizutani et al., 1995). Lipids are transported in blood as lipoproteins, macromolecular complexes of lipids and proteins (apolipoproteins). The properties and functions of apolipoproteins include being structural components of lipoproteins, the regulation of enzyme activity, and binding of lipoproteins to cell surface receptors for internalization and catabolism (Laker and Evans, 1996).Lipoprotein(a), which consists of an LDL particle and apolipoprotein(a), has been shown to be a strong independent risk factor for coronary heart disease (Mbewu and Durrington, 1990; Scott, 1991). Lipoprotein(a) has physiological interactions with coagulation and fibrinolytic systems (Hajjar et al., 1989; Miles et al., 1989). By studying its cDNA sequence, apolipoprotein(a) has been shown to have marked similarities in structure with plas-
ANTIPHOSPI-IOLIPID SYNDROME
539
minogen (McLean et al., 1987). Elevated levels of lipoprotein(a) have been reported in patients with APS, with significantly higher levels in patients with arterial than venous thrombosis (Yamazaki et al., 1994). It has been shown that a protein ligand for apolipoprotein(a) is P2GPI. Using the repetitive apolipoprotein(a) kringle IV type 2 domain as bait to screen a human liver cDNA library by the yeast two-hybrid interaction trap system, 11 clones were identified, of which 8 were P2GPI (Kochl et al., 1997). Coimmunoprecipitation experiments showed specific binding of P2GPI to immobilized apolipoprotein(a), lipoprotein( a), and low density lipoproteins (which had been shown previously).The binding of P2GPI to lipoprotein(a) is via domains 2-4. These observations will lead to further investigations into the role of this P2GPI-apolipoprotein(a) interaction and its role in a prothrombotic tendency. Apolipoprotein(a) may form a multimeric complex with PBGPI, which would be cleared from the circulation by macrophages, a process that could be affected by anti-P2GPI antibodies. In a study of middle-aged men with elevated lipids but no autoimmune disease or history of thrombosis, antiprothrombin antibodies were significantly higher in men who developed myocardial infarctions or cardiac deaths than in controls. When all variables were analyzed, there was an interactive effect of antiprothrombin antibodies with smoking and triglyceride levels independently. Autoantibodes detected in the standard CLELISA and antibodies to oxidized low-density lipoproteins had an additive effect with antiprothrombin antibodies to the risk of cardiac events (Vaarala et al., 1996). The possible link between antibodies to P2GPI and the pathogenesis of thrombosis has been studied extensively. Anionic PLs promote initiation of the contact activation system in blood coagulation, which is inhibited in vitro by physiological concentrations of P2GPI (Schousboe, 1988). The autoactivation of factor XI1 in prekallikrein-deficient plasma in the presence of anionic PLs and cationic zinc is inhibited by PZGPI and the anti-p2GPI antibodieslP2GPI complex, which could behave as a LA (Schousboe and Rasmussen, 1995). One group classified aPLs into two types, depending on their sedimentation characteristics after adsorption with cardiolipin liposomes (Galli et al., 1992). If LA activity cosedimented with the liposomal pellet, antibodies eluted from the liposomes were P2GPI dependent in prolonging dRWT. However, the LA antibody present in the supernatant prolonged dRVVT, irrespective of the presence of PBGPI. This group subsequently suggested that antibodies positive in a d R W T clotting assay were more likely to be associated with clinical thrombosis. This study was done indirectly with patient plasma and retrospective analysis of the clinical histories (Galli et al., 1995). P2GPI at physiological concentrations has been shown to inhibit the generation of factor Xa in the presence of
540
DAVID A. KANDIAH et al.
activated gel-filtered platelets (Shi et al., 1993). aCL antibodies interfered with this inhibition, whereas LA antibodies inhibited this process in a manner similar to PZGPI without any additive effect shown. P2GPI also appears to inhibit the prothrombinase activity of resting nonactivated platelets, lysed platelets, and phosphatidylserine/phosphatidylcholinevesicles (Nimpf et al., 1986), although it is unclear whether it requires small amounts of anti-PZGPI antibodies for this activity (Galliet al., 1993).PZGPI at physiological levels inhibits the factor Va-dependent prothrombinase complex (Mori et aE., 1996). However, in the same system, it potentiates thrombin generation in the presence of activated protein C (APC). This inhibitory effect was diminished by the addition of increasing concentrations of cephalin, suggesting that PZGPI competitively inhibits the binding of APC to the phospholipid surface. This group also showed that the anticoagulant activity of APC was significantly potentiated in P2GPIdepleted plasma, an effect that was reduced with the addition of increasing concentrations of p2GPI. (These in d t r o reactions are illustrated in Fig. 7.) The affinity of P2GPI for phospholipid is increased in the presence of anti-P2GPI antibodies (Willems et al., 1996). Hence in individuals with these antibodies, the P2GPI-anti-PZGPI complex may well displace other coagulation proteins affecting homeostatic mechanisms. Although the affinity constant for polyclonal anti-/32GPI to PZGPI is low at 3.4-7.2 pM (Tincani et al., 1996), this affinity constant is increased with dimerization of P2GPI (Sheng et al., 1998). In a study of 46 patients with SLE and autoantibodies detected in a standard CL-ELISA, comparisons of binding specificities and avidity of binding of the 22 patients with APS and the 24 patients without clinical manifestations of the APS were made. The authors found that while all patients had a positive result in the standard CL-ELISA, the absorbance values were higher in the group with APS. Reactivity in an anti-fi2GPIELISA was significantly more in patients with APS. Urea, a chaotropic agent interfering with electrostatic interactions, was able to reduce binding of the antibodies to p2GPI in both the standard CL-ELISA and the anti-P2GPI-ELISA, mainly in the group without clinical manifestations (Vlachoyiannopoulos et al., 1998). This suggested that patients with APS have a significant increase in autoantibodies detected in an anti-PZGPI ELISA and that these antibodies are generally of high avidity as compared to the autoantibodies found in individuals without clinical manifestations. Auger et al. ( 1995) reported that heparin-induced thrombocytopenia developed in 56.5% of patients with LA antibodies with an increased occurrence of thrombotic events. Although the mechanism is not clear, this could be due to heparin binding to P2GP1, enhancing the antibody binding to anionic phospholipids, e.g., activated platelets. Unfractionated
FIG.7. PZGPI effects in coagulation reactions. In oitro experiments with p2GPI suggest that it inhibits activation of the intrinsic pathway (Schousboe, 1988); inhibits Xa generation (Shi et al., 1993), inhibits prothrombinase activity of human platelets (Nimpf et al., 1986), and modulates the anticoagulant activity of activated protein C on phospholipid (Mori et d.,1995).
542
DAVID A. KANDIAH et a1
heparin has been shown to enhance aPL binding to P2GPI in the presence of phosphatidylserine as well as changing the electrophoretic mobility of P2GP1, which did not occur with low molecular weight heparin (McNally et al., 1994). Hence the perceived anticoagulant properties of PZGPI may be altered by unfractionated heparin and may influence its use in clinical practice for the prophylaxis and treatment of thrombosis associated with these autoantibodies. Another group suggested that heparin reversibly bound aPL antibodies in vitro as shown by depletion of aPL antibodies after passage through a heparin affinity column and by dose-response heparin inhibition of aPL to CL in the presence of PZGPI (standard CLELISA) (Ermel et al., 1995).This effect would be mediated by the binding of PZGPI to heparin. The activation of human umbilical vein endothelial cells by IgG autoantibodies from patients with APS, as measured by increased monocyte adhesion, has been shown to be PZGPI dependent. Interestingly, rabbit polyclonal anti-PSGPI antibodies also activated endothelial cells (Simantov et al., 1995). p2GPI adhesion to endothelium has been described in normal placental vessels. In the placental samples studied, increased P2GPI deposition was found by indirect immunofluorescence on the trophoblast surfaces of placentas from patients with persistently raised titers of aPL antibodies (La Rosa et al., 1994). aPL antibodies have been eluted from placentas of women with elevated serum aPL antibodies. PZGPI was present in placental eluates from both control and aPL affected pregnancies (Chamley et al., 1993). Using reverse transcriptase polymerase chain reaction (RT-PCR),placental cells were shown to synthesize p2GPI transcripts. Immunoblotting experiments suggested that P2GPI is localized in syncytiotrophoblast and extravillous cytotrophoblast. Anti-P2GPI antibodies may therefore bind placental PZGPI, inhibiting their function in vivo (Chamley et al., 1997). The function of placental p2GPI is unclear at the moment, but may have an effect on placental circulatory hemostatic mechanisms. The binding of P2GPI to endothelial cells, via the cluster of Lys residues (Del Papa et al., 1998), suggests the same possibilities as for anionic phospholipid binding. This binding may increase antigen density and/or induce conforinational changes, alIowing for the capture of circulating antiP2GPI antibodies. Anti-peGPI monoclonal antibodies have been shown to exert LA activity in vitro by enhancing the binding of PZGPI to phospholipids (Takeya et al., 1997).This clustering may hinder the lateral mobility of coagulation proteins, affecting the fine balance between their procoagulant and anticoagulant activities. This effect may also occur in PZGPI binding to endothelial cell membranes perturbing endothelial function. It is therefore possible in vivo that autoantibodies directed against P2GPI induce endothelial cell activation, which in the presence of some other
ANTIPHOSPHOLIPID SYNDROME
543
insult may trigger a thrombotic event. Hence aPL antibodies may independently influence atherogenesis by moderating coagulation reactions toward hypercoagulation by an as yet unexplained mechanism. XI. Laboratory Investigations of the Antiphospholipid Syndrome
As aPL antibodies are increasingly shown to be a heterogeneous group of autoantibodies, the need to perform multiple immunoassays and coagulation tests has become imperative. This is especially important in prospective studies of patients with APS as subsets of patients may be identified who are at particular risk of clinical events and may be identified by certain laboratory tests or a combination of tests. Standardization of assays for anticardiolipin antibodies and lupus anticoagulants have been fraught with difficulty, despite numerous attempts to perform this by international standardization workshops and committees. A. ANTICARDIOLIPIN ANTIBODIES At the third international workshop held in 1992,the delegates confirmed that bovine P2 glycoprotein I (PSGPI) supported the binding of purified aPL antibodies to CL (Harris et al., 1994). The authors and others have now found that there are patients who have selective binding to human P2GPI and may therefore have persistent negative aCL titers in conventional immunoassays. The workshop delegates also found binding of purified antibodies to ovalbumin and casein used in the blocking and diluting buffers, which suggest that any protein solutions used in an ELISA system need to be checked for contamination with PSGPI, even in small concentrations (Harris et al., 1994). It has also been found that non-fatty acid-free bovine serum albumin used in a number of laboratory experiments contains sufficient P2GPI to support binding of anti-P2GPI antibodies. The current standard ELISA kits are increasingly using P2GPI as a discriminator of autoimmune aCL antibodies from true aCL antibodies that do not require PZGPI for direct binding to CL. Nine commonly used commercial kits for measuring aCL antibodies were compared with a standardized in-house method. The authors found marked differences in the positivity rate between kits rangng from 31 to 60% for IgG and from 6 to 50% for IgM (Reber et al., 1995). The P2GPI content of the dilution buffers and the wells supplied with the lats were significantly different. Despite extensive efforts over the years to achieve standardization, these results suggest that some technical aspects need to be reevaluated, including cutoff points and the use of controlled amounts of P2GPI or incubation times. P2GPI has been shown to be inhibitory to the binding of antibodies from patients with chronic infections (Hunt et al., 1992). This is due to
544
DAVID A. KANDIAII et al.
the competition of PZGPI with the positively charged antibodies for binding to anionic phospholipids (Monestier et al., 1996). Hence high positivity of aCL titer may be found in some patients with chronic infections, and in the right clinical setting, further tests need to be performed to identify the binding specificities of these antibodies. SLE patients with clinical manifestations of APS but negative for conventional tests in the aCL assay and LA clotting tests were tested for immunoreactivity to various phospholipids, including phosphatidylserine, phosphatidylinositol, phosphatic acid, phosphatidylcholine and phosphatidylethanolamine (Roch et al., 1997). No correlation was found between detected antibodies to phosphatidylethanolamine and clinical manifestations of APS. Overall, in all patient groups, the authors found no additional benefit from testing for immunoreactivity to other phospholipids other than CL (Roch et at., 1997). The use of these new PZGPI immunoassays may supplant the conventional aCL iminunoassays in use because of their improved specificities but not completely replace them because of their relative costs. Clinical studies highlight the importance of detecting aPL antibodies and quantitating their levels and therefore stratifylng the risk for each patient so that optimum treatments could be developed. B. LUPUS ANTICOAGULANT ANTIBODIES The LA/aPL antibody subcommittee has met annually since 1988 to update the nomenclature, methods, and standardization practices of LA testing. Screening tests for LA need to be sensitive, and the amount of phospholipid in the test system is a critical determinant of sensitivity. The reactivity of a particular patient LA antibody is also important. Hence the use of at least two sensitive screening tests is important in detecting the LA antibody. The combination of tests also needs to detect reactivity to different parts of the clotting cascade (Kandiah and Krilis, 1996; Triplett, 1995). Current criteria for the diagnosis of LA antibodies are:
1. Prolongation of at least one PL-dependent clotting test. 2. Evidence of inhibitory activity shown by the effect of patient plasma on pooled normal plasma. 3. Evidence that the inhibitory activity is dependent on PL by the addition or alteration of PL, hexagonal-phase PL, platelets, or platelet vesicles in the test system originally used. 4. LAs must be carefully distinguished from other coagulopathies. Specific factor assays and the clinical history may be helpful in these situations (Brandt et al., 1995).
ANTIPHOSPHOLIPID SYNDROME
545
Predictive tests that identify patients most at risk of the clinical manifestations of APS have not yet been developed. More recent test systems have looked at dot blots to various protein antigens, in vitro thrombin generation, functional assays to detect acquired APC resistance, and inhibition of downregulation of factors Va and VIIIa (Triplett, 1996). Laboratory tests for the detection of aPL antibodies have become more specialized. Hence, clinical and research studies looking at aPL antibodies and clinical features need to be precise in the performance and reporting of the methodology of their tests, which will promote the reproducibility of results across different population groups and allow accurate interpretation of data obtained. XII. Antiphospholipid Syndrome and Future Therapies
Antiphospholipid syndrome belongs to a wide spectrum of clinical disorders that are categorized as autoimmune disorders based on the presence of autoantibodies and/or the finding of lymphocytic infiltrates in the target organs. As in most autoimmune disease, the particular role of the immune system in the initiation and progression of the disease remains uncertain. The finding that plasma proteins are the target antigens for some of the aPL antibodies has gone a long way in investigating the potential pathogenesis in APS, especially in relation to thrombosis and atherogenesis. Some patients respond well to aspirin alone, whereas other patients require high doses of anticoagulation to prevent recurrent vascular events. Autoantibodies in APS have increasingly been shown to be heterogeneous, and the combination of antibodies may be the precipitant for the diverse clinical manifestations of these patients. Not only does the combination of binding specificities of the antibodies need to be identified, but also those patients who respond to existing therapies. In disease in general, and autoimmune disease in particular, the nature of the inciting antigen is central to the definition of the disease and therapeutic interventions. In APS, as in other autoimmune diseases, the three processes that interact to produce disease need to be studied. These are (a) selection in the thymus of a repertoire of T cells that discriminates self from nonself, i.e., tolerance; (b) lymphocyte activation after a potential autoreactive T cell has emerged from the thymus and entered the peripheral circulation; and (c) preprogrammed cell death that eliminates T and B cells with particular autoreactive properties, i.e., apoptosis. In PAPS and APS associated with SLE, there is considerable diversity in the spectrum of autoantibodies and severity of disease among affected indwiduals. One theory proposed for the diversity of autoantibodies is the presence of common structural motifs found in many diverse implicated
546
DAVID A. KANDIAH et a1
molecules, e.g., phosphodiester groups found in single-stranded DNA and phospholipids, and P2GPI- and DR9-binding motif (Fujisao et al., 1996). “Antigenic spreading,” in which the T-cell response to a particular peptide antigen leads to the involvement of other T cells with a progressively wider spectrum of activity (Lehmann et al., 1993),has also been proposed. Thymic tolerance failure may also explain the development of autoimmune disease. This may arise by (a) anatomical sequestration of self-antigens, not exposing developing T cells in the thymus to these self-antigens during their maturation; (b) formation of neoantigens or cryptic antigens when the target antigen is conformationally changed, thus not being recognized by T cells (Fatenejad et al., 1993); and (c) failure to suppress autoreactive T cells in the periphery by some as yet unexplained mechanism (Clark and Ledbetter, 1994). The ability of B cells to internalize antigens and present them as peptide/ class I1 complexes on their cell surface may play a critical role in “antigenic spreading,” promoting a progressively more polyclonal T-cell response against an autoantigen (Mamula and Janeway, 1993). In addition to the crucial processes generating autoreactive T cells, apoptosis and factors leading to the death of cells and their removal in the thymus and the periphery may be important in pathogenesis. It has been shown that M R U p r autoimmune mice with T-cell antigen receptor (Y chain knockout (which lack ap T cells) lack IgG and IgM aCL antibodies compared to their wild-type controls and that CD40 ligand-deficient M R U lpr mice lack IgG aCL antibodies compared to their wild-type controls. These observations suggest that the development of aCL antibodies in these autoimmune mice is dependent on cognate T/B-cell interaction (augmented by the presence of (YPT cells) and is facilitated by the binding of CD40 ligand on activated T cells to CD40 on B cells (Kang et al., 1997). Multiple genes are involved in autoimmune disease pathogenesis. The strongest argument for genes predisposing to disease is increased disease frequency in monozygotic twins. Disease concordance in this twin population, however, is generally less than 30%, and other factors such as random events occurring during the maturation of antigen receptors on T and B cells and environmental factors must also play a significant part. The detection of viral protein epitopes similar to that found in PZGPI may promote further insights into the mechanism of initiation of pathological events and clinical manifestations in APS (Celli et al., 1997). P2GPI is in the CCP family. The complement system and its receptors play an important role in immune defense, linking humoral and cellmediated responses. Complement receptor (CRI), through its binding to activation products of complement C3, can activate cells and induce chemotaxis (Fearon, 1991). B lymphocytes have complement receptor
ANTIPHOSPHOL.IPID SYNDROME
547
CR2, which may form complexes with CD19 that are important in B-cell activation (Matsumoto et al., 1991). Complement receptors also serve to concentrate immune complexes on the surface of B cells, leading to antigen presentation by the B cell (Tuveson et al., 1991). In deriving therapy for an autoimmune disease such as APS, the immune system needs to be modulated so as not to result in total immunosuppression. This modulation may be performed at the level of autoimmune T-cell generation in the thymus or their clonal expansion in the periphery. Current treatment appears to only address the end result of the autoimmune process, i.e., treatment of the thrombosis. Synthetic miniotope peptides, characterized by (a) the inability to activate T cells while (b) retaining the ability to bind immune B cells, may be used to tolerize B cells in an antigen-specific manner. B-cell tolerance entails administering such peptides conjugated to multivalent, stable, nonimmunogenic valency platforms in order to abrogate antibody production via B-cell anergy or clonal deletion after the cross-linking of surface immunoglobulin. The ability to modulate B-cell activity in humans on an antigenspecific basis, using single signal inactivation of target B cells, has been identified as a means of pharmacological intervention in antibody-mediated pathologies (Coutts et al., 1996). Although the exact molecular nature of the target epitopes recognized by aPL antibodies is unknown, the use of peptides derived from epitope libraries may allow for the construction of successful tolerogens. XIII. Summary and Conclusions
Advances in defining the target antigen(s) for the autoantibodies in the APS highlight the inadequacies of the current classification of these autoantibodies into anticardiolipin and LA antibodies. The discovery that p2GPI is the target antigen for the autoantibodies detected in solid-phase immunoassays has opened a number of areas of research linking these autoantibodies to atherogenesis and thrombus formation. Although the role of p2GPI in the regulation of blood coagulation in unclear, current evidence suggests that anti-fi2GPI antibodies interfere with its “normal” role and appear to promote a procoagulant tendency. The expansion of research in this area and the diversity of the clinical manifestations of patients with APS have resulted in the inclusion of molecular biologists and pharmaceutical companies joining immunologists, hematologists, rheumatologists, obstetricians, neurologists, vascular surgeons, and protein and lipid biochemists in attempting to understand the pathophysiology of this condition. Although the published literature may result in conflicting results and introduce new controversies, developing standard-
548
DAVID A. KANDIAH d nl.
TABLE I1 VASCULAR THROMBOSIS Presence of Antiphospholipid antibodies Homocystinemia Protein deficiencies Protein C Protein S Antithrombin 111 APC resistance Genetic point mutations
Arterial Thrombosis
Venous Thrombosis
Yes
Yes
Yes
Yes Yes Yes Yes Yes
ized laboratory methods and extrapolation of in vitro experimental results to the in vivo situation will advance our understanding of the regulation of the immune system and its interaction with normal hemostatic mechanisms. Since the authors’ last review in 1991, the study and understanding of the pathophysiology of APS have evolved from lipid biochemistry to molecular techniques that may eventually provide specific therapies for the clinical manifestations of this condition. Although current treatment has improved the morbidity associated with this condition, especially in improving pregnancy outcomes, future therapies, as outlined in this review, may specifically address the biological abnormalities and have fewer side effects. Better diagnostic tools, such as magnetic resonance imaging with perfusion studies, will allow the study of the true incidence and prevalence of vascular flow changeshssue ischemia and infarction associated with aPL antibodies and help determine treatment and prophylaxis for APS patients. APS is still the only hypercoagulable condition where both arterial and venous beds can be affected independently or in the same individual (Table 11).
ACKNOWLEDGMENTS Work from the authors’ laboratories was supported by grants from the National Health and Medical Research Council (NH&MRC) of Australia and the National Institutes of Health. DAK was supported by a NH&MRC Postgraduate Biomedical Scholarship. AS is a Sinsheimer Scholar and was a Fellow of the Jane Coffin Childs Memorial Fund for Medical Research at the Department of Chemistry, Harvard University.
REFERENCES Abola, E. E., Bernstein, F. C., Biyant, S. H., Koetzle, T., and Weng, J. (1987).In “Crystallographic Databases: Information, Content, Sofhvare Systems, Scientific Applications” (F. H. Allen, G. Bergerhoff, and R. Sievers, eds.), pp. 107-132. Data Commission of the International Union of Crystallography, Cambridge/Chester.
ANTIPHOSPHOLIPID SYNDROME
549
Adrianakos, A. A., Duffy, U. J., Suzuki, M., and Sharp, J. J. (1975). Transverse myelopatliy in systemic lupus erythematosus: A report of three cases and a review of the literature. Ann. Intern. Men. 83, 616-624. Alarcon-Segovia,D., and Osmundson, P. J. (1965). Peripheral vascular syndromes associated with systemic lupus erythematosus. Ann. Intern. Med. 62, 907-919. Alarcon-Segovia, D., Perez-Vasquez, M. E., Villa, A. R., Drenkard, C., and Cabiedes, J. (1992). Preliminary classification criteria for the antiphospholipid syndrome within systemic lupus erythematosus. Semin. Arthritis Rheum. 21, 275-286. Alarcon-Segovia, D., and Sanchez-Guerrero. J. (1989). Primary antiphospholipid syndrome. /. Rheicinntol. 16, 482-486. Alegre, V. A,, Gastineau, D. A., and Winkelmami, R. K. (1989). Skin lesions associated with circulating lupus anticoagulant. Br. /. Dennntol. 120, 419-429. Alegre, V. A,, and Winkelmann, R. K. (1988). Histopathologic and immunofluorescence study of skin lesions associated with circulating anticoagulant. 1. Am. Acnd. D e m t n l . 19, 117-124. Anderson, D., Bell, D., Lodge, R., and Grant, E. (1987). Recurrent cerebral ischaemia and mitral valve vegetation in a patient with lupus anticoagulant.1.Rheumatol. 14, 839-841. The Antiphospholipid Antibodies in Stroke Study Group (APASS) ( 1990). Clinical and laboratory findings in patients with antiphospholipid antibodies and cerebral ischaemia. Stroke 21, 1268-1273. APASS (1993). Anticardiolipin antibodies are a11independent risk factor for first ischaemic stroke. Neurology 43, 2069-2073. Aoki, K., Dudkiewicz, A. B., Matsuura, E., Novotny, M., Kaberlein, G., and Gleicher, N. (1995). Clinical significance of p2-glycoprotein I dependent anticardiolipin antibodies in the reproductive autoimmune failure syndrome: Correlation with conventional antiphospholipid antibody detection systems. Am. 1.Obstet. Gynnecol. 172, 926-931. Aoyarna, Y., Chan, Y. L., and Wool, I. G. (1989).The primary structure of rat p2-glycoprotein I. Nucleic Acids Res. 17, 6401. Arnold, M. H., and Schrieber, L. (1988). Splenic and renal infarction in systemic lupus erythematosus: Association with anticardiolipin antibodles. Chi. Rheumntol. 7,406-410. Aron, A. L., Cuellar, M. L., Brey, R. L., Mckeown, S., Espinoza, L. R., Shoenfeld, Y., and Gharavi, A. E. (1995). Early onset of autoiniinunity in MRL/++ mice following immunization with beta 2-glycoprotein I. Clin. Exp. Immunol. 101, 78-81. Arvieux, J., Roussel, B., Ponard, D., and Colomb, M . G. (1994). IgG2 subclass restriction of anti@ glycoprotein I antibodies in autoiminune patients. Clin. Exp. Immunol. 95, 310-3 15. Arvieux, J.. Darnige, L., Caron, C., Reber, G., Bensa, J. C., and Colomb, M. G. (1995). Developincnt of an ELISA for autoaiitibodies to protlirombin showiilg tlieii- prevalence in patients with lupus anticoagulants. l’hronb. linevaostnsis 74, 1120-1 125. Arvieux. J., Daniige, L., Hadidla, E., R O L I S S B., ~ ~ , Bensa, J. C., and Colomb, M. G. (1996). Species specificity of anti-P2 glycoprotein I autoantibodies and its relevance to anticardiolipiii antibody quantitation. Thronlb. Haeniostnsis 75, 725-730. Asherson, R. A. (1988). A “primary” antiphospholipid syndrome? 1.Rhezrnmtol. 15, 17421746. Asherson, R. A. (1992). The catastrophic antiphospholipid syndrome. 1. Rheumntol. 19, 508-512. Asherson, R. A. (1993).Antiphospholipid antibodies and the kidney.]. Rheumntol. 20,12681272. Asherson, R. A. (1994). Hypoadrenalism and antiphospholipid antibodies: A new cause of “idiopathic”Addison’sdisease. In “Advances in Thomas Addison’s Diseases” (Bhatt, James, Besser. Botazzo, and Keen, eds.), pp, 87-101. Journal of Endocrinology, Bristol, England.
550
DAVID A. KANDIAH el al
Asherson, R. A,, and Cervera, R. (1992). The antiphospholipid syndrome: A syndrome in evolution. Ann. Rheum. Dis. 51, 147-150. Asherson, R. A,, Harris, E. N., Gharavi, A. E., and Hughes, G. R. V. (1986a). Systemic lupus eythematosus, antiphospholipid antibodies, chorea and oral contraceptives. Arthritis Rheum. 29, 1533-1536. Asherson, R. A., Higenbottam, T. W., Dinh Xuan, A. T., Khamashta, M. A., and Hughes, G. R. V. (1990). Pulmonary hypertension in a lupus clinic: Experience with twenty-four patients. 1.Rheumatol. 17, 1292-1296. Asherson, R. A,, and Hughes, G. R. V. (1988). Antiphospholipid antibodies in chorea. 1.Rheumutol. 15, 377-379. Asherson, R. A,, and Hughes, G. R. V. (1989).Recurrent deepvein thrombosis and Addison’s disease in “primary” antiphospholipid syndrome. 1.Rhwmutul. 16, 378-380. Asherson, R. A., Khamashta, M. A., Ordi-Ros, J., Derksen, R. H. W. M., Machin, S. J., Barquinero. J., Outt, H. H., Harris, E. N., Vilardell-Tarres. M., and Hughes, G. R. V. (1989). The “primary” antiphospholipid syndrome: Major clinical and serological features. Medicine 68, 366-374. Asherson, R. A,, Mackworth-Young, C. G., Harris, E., Gharavi, A. E., and Hughes, G. R. V. (1985). Multiple venous and arterial thromboses associated with the lupus anticoagulant and antibodies to cardiolipin in the absence of SLE. Rheumutol. Int. 5, 91-93. Asherson, R. A., Morgan, S. H., Harris, E. N., Gharavi, A. E., Kraus, T., and Hughes, G. R. V. (1986b).Artend occlusion causing large bowel infarction: A reflection of clotting diathesis in SLE. Clin. Rheunwtol. 5, 102-106. Auger, W. R., Permpikul, P., and Moser, K. M. (1995). Lupus anticoagulant, heparin use, and thrombocytopenia in patients with chronic thromboembolic pulmonary hypertension: A preliminary report. Am. 1.Med. 99,392-396. Bajaj, S. P., Rapaport, S. I., Fierer, D. S., Herbst, K. D., and Schwartz, D. B. (1983). A mechanism for the hypoprothrombinemia of the acquired hypoprothrombinemia-lupus anticoaguIant syndrome. BZooc161,684-692. Bakimer, R., Fishman, P., Blank, M., Sredni, B., Djaldetti, M., and Shoenfeld, Y. (1992). Induction of primary antiphospholipid syndrome in mice by immunization with a human monoclonal anticardiolipin antibody (H-3).1. Clin. lnuest. 89, 1558-1563. Balestrieri, G., Tincani, A., Spatola, L., Allegri, F., Prati, E., Cattaneo, R., Valesini, G., Del Papa, N., and Meroni, P. (1995). Anti-~2-glycoprotein I antibodies: A marker of the antiphospholipid syndrome? Lupus 4, 122-130. Barlow, P. N., Steinkasserer, A,, Norman, D. G., Kieffer, B., Wiles, P., Sim, R. B., and Campbell, I. D. (1993). Solution structure of a pair of complement modules by nuclear magnetic resonance. 1. Mol. Biol. 232, 268-284. Bendixen, E., Halkier, T., Magnusson, S., Sottrup-Jensen, L., and Kristensen, T. (1992). Complete primary structuring of bovine beta2-glycoprotein I: Localisation of the disulphide bridges. Biochemistry 31, 3611-3617. Bevers, E. M., Galli, M., Barbui, T., Comfurius, P., and Zwaal, R. F. A. (1991). Lupus anticoagulant IgG’s (LA) are not directed to phospholipids only, but to a complex of lipid-bound human prothrombin. Thromb. Haemstasis 66, 629-632. Bilofsky, H. S., and Burks, C. (1988). The GenBank genetic sequence data bank. Nideic Acids Bas. 16, 1861-1863. Biou, V., Gibrat, J.-F., Levin, J. M., Robson, B., and Gamier, J. (1988). Secondary structure prediction: Combination of three different methods. Prot. Eng. 2, 185-191. Blank, M., Cohen, J., Toder, V., and Shoenfeld, Y. (1991). Induction of anti-phospholipid syndrome in naive mice with mouse lupus monoclonal and human polyclonal anti-cardiolipin antibodies. Proc. Natl. Acad. Sci. USA 88, 3069-3073.
ANTIPHOSPHOLIPID SYNDROME
*551
Blank, M., Faden, D., Tincani, A,, Kopolovic, J., Goldberg, I., Gilhurd, B., Allegri, F., Balestrieri, G., Valesini, G., and Shoenfeld, Y. (199413).Immunization with anticardiolipin cofactor (beta-2-glycoprotein I ) induces experimental antiphospholipid syndrome in naive mice. J. Autoimmun. 7, 441-455. Blank, M., Tincani, A,, and Shoenfeld, Y. (1994a).Induction of experimental antiphospholipid syndrome in naive mice with purified IgG antiphosphatidylserine antibodies.]. Rheumatol. 21, 100-104. Boey, M. L., Colaco, C. B., Gharavi, A. E., Elkon, K. B., Loizou, S., and Hughes, G. R. V. (1983). Thrombosis in systemic lupus erythematosus: Striking association with the presence of circulating lupus anticoagulant. Br. Med. ]. 287, 1021-1023. Bokarewa, M. I., Bremme, K., Falk, G., Sten-Linder, M., Egberg, N., and Blomback, M. (1995). Studies on phospholipid antibodies, APC-resistance and associated mutation in the coagulation factor V gene. Thromb. Res. 78, 193-200. Bowie, E. J. W.,Thompson, J. H., Jr., Pascuzzi, C. A,, and Owen, C. A., Jr. (1963).Thrombosis in systemic lupus erythematosus despite circulating anticoagulants. J. Lab. Clin. Med. 62,416-430. Branch, D. W. (1990). htiphospholipid antibodies and pregnancy: Maternal implications. Semin. Perinatal. 14, 139-146. Branch, D. W., Dudley, D. J., Mitchell, M. D., Creighton, K. A,, Abbott, T. M., Haminond, E. H., and Daynes, R. A. (1990).Immunoglobulin G fractions from patients with antiphospholipid antibodies cause fetal death in BALB/c mice: A model for autoimmune fetal loss. Am. J. Obstet Gynaecol. 163, 210-216 Brandt, J. T., Triplett, D. A,, Alving, B., and Scharrer, I. (1995). Criteria for the diagnosis of lupus anticoagulants: An update. Throw&. Haenlostasis 74, 1185-1190. Brey, R. L., Hart, R. G., Sherman, D. G., and Tegeler, C. T. (1990). Antiphospholipid antibodies and cerebral ischaemia in young people. Neurology 40, 1190-1196. Briley, D. P., Coull, B. M., and Goodnight, S. H. (1989). Neurologic disease associated with antiphospholipid antibodies. Ann. Neurol. 25, 221-227. Buchberg, A. M., Brownell, E., Nagata, S., Jenkins, N. A., and Copeland, N. G. (1989). A comprehensive genetic map of murine chromosome 11 reveals extensive linkage conservation between mouse and human. Genetics 122, 153-161. Cabiedes, J., Cabral, A. R., and Alarcon-Segovia, D. (1995). Clinical manifestations of the antiphospholipid syndrome in patients with systemic lupus erythematosus associate more strongly with anti-@2-glycoprotein-Ithan with antiphospholipid antibodies. J. Rheumatol. 22, 1899-1906. Cariou, R., Tobelem, G., Bellucci, S., Soria, J., Maclouf, J., and Caen, J. (1988). Effect of lupus anticoagulant on antithrombogenic properties of endothelial cells: Inhibition of thronibomodulin-dependent protein C activation. Thromb. Huemostasis 60, 54-58, Caruso, A,, De Carolis, S., Ferrazzani, F., Valesini, G., Caforio, L., and Mancuso, S. (1993). Pregnancy outcome in relationship to uterine artery flow velocity waveforms and clinical characteristics in women with antiphospholipid syndrome. Obstet. Gynnecol. 82,970-977. Celli, C. M., Gharavi, E. E., Cucurull, E., Espinoza, L. R., Chimovich, H., and Gharavi, A. E. (1997). @2-glycoproteinI and phospholipid-binding viral peptides interaction with cardiolipin-containing liposomes. Arthritis Rheum. 40(9), Suppl. 314 [Abstract 17001 Cervera, R., Font, J., Pare, C., Azqueta, M., Perez-Villa, F., Lopez-Soto, A,, and Ingelmo, M. (1992). Cardiac disease in systemic lupus erythematosus: Prospective study of 70 patients. Ann. Rheum. Dis. 51, 156-159. Chamley, L. W., Allen, J. L., and Johnson, P. M. (1997). Synthesis of beta 2-glycoprotein I by the human placenta. Placenta 18, 403-410.
552
DAVID A. KANDIAH el d.
Chamley, L. W., Pattison, N. S., and McKay, E. J. (1993).Elution ofanticardiolipin antibodies and their cofactor beta 2-glycoprotein I from the placentae of patients with a poor obstetric history. J. Reprod lininunol. 25, 209-220. Chandonia, J. M., and Karplus, M. (1996). The importance of larger data sets for protein secondary structure prediction with neural networks. Prot. Sci. 5, 768-774. Clark, E., and Ledbetter, J. (1994). How B and T cells talk to each other. Nature 367, 425-428. Conley, C. L., and Hartmann, R. C. (1952).A haemorrhagic disorder caused by a circulating anticoagulant in patients with disseminated lupus erythematosus. 1.CZin. Invest. 31, 621. Cote, S. A,, Brey, R. L., and Teale, J. M . (1994). Autoimmune neurological disease is accelerated in M R U f mice after apolipoprotein H (Apo H) immunization. Lupus 3,97. Coutts, S . M., Plunkett, M. L., Iverson, G. M., Barstdd, P. A,, and Berner, C. M. (1996). Pharmacological intervention in antibody mediated disease. Lupus 5, 158- 159. de la P a , P., Sutton, B. J., Darsley, M. J., and Rees, A. R. (1986). Modelling of the combining sites of three anti-lysozyme monoclonal antibodies and of the complex between one of the antibodies and its epitope. EMBO J. 5, 415-425. Del Papa, N., Guidali, L., Sala, A,, Buccellati, C., Khamashta, M. A., Ichikawa. K., Koike, T., Balestrieri, G., Tincani, A,, Hughes, G. R. V., and Meroni, P. L. (1997). Endothelial cells as target for antiphospholipid antibodies: Human polyclonal and monoclonal antip2-glycoprotein I antibodies react in vitro with endothelial cells through adherent p2glycoprotein I and induce endothelial activation. Arthritis Rheum. 40, 59 1-561. Del Papa, N., Guidali, L., Spatola, L., Bonra, P., Borghi, M. O., Tincani, A,, Balestrieri, G., and Meroni, P. L. (1995). Relationship between anti-phospholipid and anti-endothelial antibodies 111:p2-glycoprotein I mediates the antibody binding to endothelial membranes and induces the expression of adhesion molecules. CZin. Exp. Rheurnatol. 13, 179-186. Del Papa, N., Meroni, P. L., Tincani, A,, Harris, E. N., Pierangeli, S. S., Barcellini,W., Borghi, M. O., Balestrieri, G., and Zannssi, C. (1992). Relationship between antiphospholipid and antiendothelial antibodies: Further characterization of the reactivity on resting and cytokine-activated endothelial cells. Clin. Exp. Rheurnatol. 10, 37-42. Del Papa, N., Sheng, Y., Raschi, E., Kandiah, D. A., Tincani, A,, Khamashta, M. A., Hughes, G. R. V., Koike, T., Balestrieri, G., Krilis, S. A., and Meroni, P. L. (1998). Human p2glycoprotein I binds to endothelial cells through a cluster of lysine residues that are critical for binding anionic phospholipids and offers epitopes for anti-fi2-glycoprotein 1 antibodies. 1.Iminunol., in press. De Wolf, F., Cameras, L. O., Moerman, P., Vermylen, J., Van Assche, A,, and Renaer, M. (1982). Decidual vasculopathy and extensive placental infarction patient with repeated thromboembolic accidents, recurrent fetal loss, and a lupus anticoagulant. Am. J. Obstet. Gynaecol. 142, 829-834. Erlendsson, K., Steinsson, K., Johansson, J. H., and Geirsson, R. T. (1993). Relation of antiphospholipid antibody and placental bed inflammatoryvascular changes to the outcome of pregnancy in successive pregnancies of 2 women with SLE. J. Rheurnatol. 20, 17791785. Ermel, L. D., Marshburn, P. B., and Kutteh, W. H. (1995). Interaction of heparin with antiphospholipid antibodies (APA)from the sera of women with recurrent pregnancy loss (RPL). Am. J. Reprod. lmrnunol. 33, 14-20. Farrugia, E., Torres, V. E., Gastineau, D., Michet, C. J., and Holley, K. E. (1992). Lupus anticoagulant in systemic lupus erytheniatosus: A clinical and renal pathological study. Am. J. Kidney Dis. 20, 463-471. Fatenejad, S., Mamula, M. J., and Craft, J. (1993). Role of intermolecular/intrastructural B- and T-cell determinants in the diversification of autoantibodies to ribonucleoprotein particles. Proc. Natl. Acad. Sci. USA 90, 12010-12014.
+
ANTIPHOSPHOLIPID SYNDROME
553
Fearon, D. T. (1991). Anti-inflammatory and iinmunosuppressive effects of recombinant soluble complement receptors. Clin. Exp. bnmunol. 86(Suppl. l), 43-46. Felsenstein, J. (1985). Confidence limits on phylogenies: An approach using the bootstrap.
Evolution 39, 783-791. Fleck, R. A., Rapaport, S. I., and Rao, L. V. (1988). Anti-prothrombin antibodies and the lupus anticoagulant. Blood 72, 512-519. Forastiero, R. R., Martinuzzo, M. E., Cerrato, G. S.. Kordich, L. C., and Carreras, L. 0. (1997).Relationship of anti-P2-glycoprotein I and anti-prothrombin antibodies to thrombosis and pregnancy loss in patients with antiphospholipid antibodies. Thromb. Hnemostnsis 78, 1008-1014. Ford, P. M., Ford, S. E., and Lillicrap. D. P. (1988).Association of lupus anticoagulant with severe valvular heart disease in systemic lupus erythematosus. J . Rheumatol. 15,597-600. Fujisao, S., Matsushita, S., Nishi, T., and Nishiinura, Y. (1996). Identification of HLA-DR9 (DRB1 * 0901)-binding peptide motifs iising a phage random peptide library. Hum.
Imniunol. 45, 131-136. Galli, M., Beretta, G., Daldossi, M., Bevers, E. M., and Barbui, T. (1997). Different anticoagulant and immunological properties of anti-prothrombin antibodies in patients with antiphospholipid antibodies. Thrornb. Hnernostnsis 77, 486-491. Galli, M., Bevers, E. M., Comfurius, P., Barbui, T., and Zwaal,R. F. A. (1993). Effect of anti-phospholipid antibodies on procoagulant activity of activated platelets and plateletderived microvesicles. Br. ]. Haemntol. 83, 466-472. Galli, M., Comfurius, P., Barbui, T., Zwaal, R. F., and Bevers, E. M. (1992). Anticoagulant activity of beta 2-glycoprotein I is potentiated by a clistinct subgroup of anticardiolipin antibodies. Thrornb. Haemostnsis 68, 297-300. Galli, M., Comfnrius, P., Massen, C., Heinker, H. C., deBaets, M. H., van Breda-Vriesman, P. J. C., Barhi, T., Zwaal, R. F. A. and Bevers, E. M. (1990). Anticardiolipin antibodies (ACA)directed not to cardiolipin but to a plasma protein cofactor. Lancet 335,1544-1547. Galli, M . , Finazzi, G., Bevers, E. M., and Barbui, T. (1995). Kaolin clotting time and dilute Russell’s viper venom time distinguish between prothroinbin-dependent and P2glycoprotein I-dependent antiphospholipid antibodies. Blood 86, 617-623. Gavaghan, T. P., Krilis, S. A., Daggard, G. E., Baron, D. W., Hickie, J. B., and Chesterman, C. N. (1987). Anticardiohpin antibodies and occlusion of coronary artery bypass grafts.
Lancet 2, 977-978. Gharavi, A. E., Mellors, R. C., and Elkon, K. B. (1989). IgG anticardiolipin antibodies in murine lupus. Clin. Exp. Irnnzrrnol. 78, 233-238. Ghosh, S., Walters, H. D., Joist, J. H., Osborn, T. G., and Moore, T. L. (1993). Adult respiratory distress syndrome associated with antiphospholipid syndrome. J. Rhenmatnl. 20, 1406-1408. Gleicher, N. (1994). Autoantibodies and pregnancy loss. Lancet 343, 747-748. Goldstein, S. R. (1994).Embryonic death in early pregnancy: A new look at the first trimester.
Obstet. Gynaecol. 84, 294-297. Grennan, D. M., McCormick, J. N., Wojtacha, D., Carty, M., and Behan, W. (1978). Iininunological studies of the placenta in systemic lupus erythernatosus. Ann. Rheuni.
Dis. 37, 129-134. Grottolo, A,, Ferari, V., Mariarosa, M., Sainbrnni,. A., Tincani, A,, and Del Bono, R. (1988). Priinary adrenal insufficiency. circulating lupus anticoagulant and anticardiolipin antibodies in a patient with multiple abortions and recurrent thrombotic episodes. Haeninlologin 73, 517-519. Haagemp, A,, Kristensen, T., and Kruse, T. A. (1991).Polylnorphisin and genetic mapping of the gene encoding human p2-glycoprotein I to chromosome 17. Cytogenet. Cell Genet.
58,2004-2010.
554
DAVID A. KANDIAH et al.
Hajjar, K. A,, Gavish, D., Breslow, J. L., and Nachman, R. L. (1989). Lipoprotein(a) modulation of endothelial cell surface fibrinolysis and its potential role in atherosclerosis. Nature 339, 303-305. Hang, L. M., Izui, S., and Dixon, F. J. (1981). (NZW x BXSB)Fl hybrid: A model of acute lupus and coronaryvascular disease with myocardial infarction.]. Exp. Med. 154,216-221. Harris, E. N., Chan, J. K., Asherson, R. A,, Aber, V. R., Gharavi, A. E., and Hughes, G. R. V. (1986). Thrombosis, recurrent fetal loss and thrombocytopenia: Predictive value of the anticardiolipin antibody test. Arch. Intern. Med. 146, 2153-2156. Harris, E. N., Gharavi, A. E., Asherson, R. A,, Boey, M. L., and Hughes, G. R. V. (1984). Cerebral infarction in systemic lupus: Association with anticardiolipin antibodies. Clin. Exp. Rheum. 25, 1271-1277. Harris, E. N., Pierangeli, S., and Birch, D. (1994). Anticardiolipin wet workshop report (fifth international symposium on antiphospholipid antibodies). Am. J. Clin. Pathol. 101, 616-624. Hasegawa, I., Takakuwa, K., Adachi, S., and Kanazawa, K. (1990). Cytotoxic antibody against trophoblast and lymphocytes present in pregnancy with intrauterine fetal growth retardation and its relation to anti-phospholipid antibody. J. Reprod. Immunol. 17, 127-139. Hashimoto, Y., Kawamura, M., Ichikawa, K., Suzuki, T., Sumida, T., Yoshida, S., Matsuura, E., Ikehara, S., and Koike, T. (1992).Anticardiolipin antibodies in NZW x BXSB/Fl mice. ]. Immunol. 149, 1063- 1068. Hasselaar, P., Derksen, R. H. W. M., Blokzijl, L., and de Groot, P. G. (1990). Crossreactivity of antibodies directed against cardiolipin, DNA, endothelial cells and blood platelets. Thromb. Haemostasis 63, 169-173. Hasselaar, P., Derksen, R. H. W. M., Oosting, J. D., Blokzijl, L., and de Groot, P. G. (1989). Synergistic effect of low doses of tumor necrosis factor and sera from patients with systemic lupus erythematosus on the expression of procoagulant activity by cultured endothelial cells. Thromb. Haeinostasis 62, 654-660. Hering, R., Couturier, E. G. M., Asherson, R. A., and Steiner, T. J. (1991).Antiphospholipid antibodes in migraine. Cephalalgia 11, 19-20. Herranz, M. T., Rivier, G., Kharnashta, M. A,, Blaser, K. U., and Hughes, G. R. V. (1994). Association between antiphospholipid antibodies and epilepsy in patients with systemic lupus erythematosus. Arthritis Rheum. 37,568-571. Hochberg, M. C. (for the Diagnostic and Therapeutic Criteria Committee of the American College of Rheumatology) (1997). Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 40, 1725. [letter] Horbach, D. A., van Oort, E., Donders, R. C. J. M., Derksen, R. H. W. M., and de Groot, P. G. (1996). Lupus anticoagulant is the strongest risk factor for both venous and arterial thrombosis in patients with systemic lupus erythematosus. Thromb. Haemostasis 76, 916-924. Horellou, M. H., Aurousseau, M. H., Boffa, M. C., Conrad, J., Wiesel, M. L., and Samama, M. (1987). Biological and clinical heterogeneity of lupus and lupus-like anticoagulant in fifty-seven patients. J Med. 18, 199-217. Howe, H. S., Boey, M. L., Fong, K. Y., and Feng, P. H. (1988). Pulmonary haemorrhage, pulmonary infarction, and the lupus anticoagulant. Ann. Rheum. Dis.47, 869-872. Hughes, G. R. V. (1983).Thrombosis, abortion, cerebral disease and the lupus anticoagulant. Br. Med. J. 287, 1088-1089. Hughes, G. R. V. (1985).The anticardiolipin syndrome. Clin. Exp. Rheumtol. 3,285-286.
ANTIPHOSPHOLIPID SYNDROME
55s
Hughes, G. R. V., Harris, E. N., and Gharavi, A. E. (1986). The anticardiolipin syndrome. /. Rheumtol. 13, 486-489. Hunt, J. E., McNeil, H. P., Morgan G. J., Crameri, R. M., and Krilis S. A. (1992). A phospholipid-P2-gIycoprotein I complex is an antigen for anticardiolipin antibodies occurring in autoimmune disease but not with infection. Lupus 1, 83-90. Hunt, J. E., Simpson, R. J., and Krilis, S. A. (1993). Identification of a region of P2glycoprotein I critical for lipid-binding and anticardiolipin cofactor activity. Pmc. Natl. Acad. Sci. USA 90, 2141-2145. Hunt, J. E., and Krilis, S. A. (1994). The fifth domain of P2-glycoprotein I contains a phospholipid-binding site (Cys281-Cys288) and a region recognized by anticardiolipin antibodies. 1.Zmmunol. 152, 653-659. Kamboh, M. I., Wagenknecht, D. R., and McIntyre, J. A. (1995). Heterogeneity of apolipoprotein H”3allele and its role in affecting the binding of apolipoprotein H (@2-glycoprotein I ) to anionic phospholipids. Hum. Genet. 95, 385-388. Kandiah, D. A,, and Krilis, S. A. (1994). Beta2-glycoprotein I. Lupus 3, 207-212. Kandiah, D. A,, and Krilis, S. A. (1996a). Immunology and methods of detection of antiphospholipid antibodies. In “The Antiphospholipid Syndrome” (R. A. Asherson, R. Cervera, J.-C. Piette, and Y. Shoenfeld, eds.), pp. 29-48. CRC Press, Boca Raton, FL. Kandiah. D. A,, and Krilis, S. A. (1996b).Laboratoly detection of antiphospholipid antibodies. LUPUS5, 160-162. Kandiah, D. A,, and Krilis, S. A. (1997a). Affinity-purified anti-prothrombin antibodies in the antiphospholipid syndrome: Immunological specificity and clotting profiles. Arthritis Rheum. 40(9), Suppl., 104. [Abstract 4431 Kandiah, D. A , , and Krilis, S. A. (199%). The clotting assays dilute Russell’s viper venom time (dRVVT) and dilute kaolin clotting time (dKCT) detect different populations of antibodies in plasmas of individual patients with lupus anticoagulant activity. Arthritis Rheum 40(9), suppl., 298. [Abstract 1604J Kang, I., Craft, J., and Bockenstedt, L. K. (1997). arP T-helper cell and CD40 ligand dependency in the development of anticardiolipin antibody. Arthritis Rheum 40(9), Suppl., 314. [Abstract 17011 Kant, K. S., Pollak, V. E., Weiss, M. A,, Check, H. I., Miller, M. A,, and Hess, E. V. (1981). Glomerular thrombosis in systemic lupns erythematosus: Prevalence and significance. Medicine 60, 71-86. Keeling, D. M., Wilson, A. J., Mackie, I. J., Isenberg, D. A., and Machin, S. J. (1993). Role of beta 2-glycoprotein I and anti-phospholipid antibodies in activation of protein C in vitro. 1. Clin. Pathol. 46, 908-911. Kertesz, Z., Yu, B. B., Steinkasserer, A,, Haupt, H., Benham, A,, and Sim, R. B. (1995). Characterization of binding of human beta 2-glycoprotein I to cardiolipin. Biochem. J . 310,315-321. Khamashta, M. A,, Cervera, R., Asherson, R. A,, Font, J., Gil, A., Coltart, D. J., Vazquez, J. J., Pare, C., Ingelmo, M., Oliver, J., and Hughes, G. R. V. (1990). Association of antibodies against phospholipids with heart valve hsease in systemic lupus erythematosus. Lancet 335, 1,541- 1544. Kochl, S., Fresser, F., Lobentanz, E., Baier, G., and Utermann, G. (1997).Novel interaction of apolipoprotein(a) with /3%glycoprotein I mediated by the kringle IV domain. Blood 90, 1482-1489. Kouts, S., Wang, M. X., Adelstein, S., and Krilis, S. A. (1995). Iininunization of a rabbit with P2-glycoprotein 1 induces charge-dependent crossreactive antibodies that bind anionic phospholipids and have similar reactivity as autoimmune antiphospholipid antibodies. 1.Zmmunol. 155, 958-966.
556
DAVID A. U N D I A H et a1
Kristensen, T., Schousboe, I., Boel, E., Mulvihill, E. M., Hansen, R. R., MoUer, K. B., Moller, N. P. H., and Sottrup-Jensen, L. (1991). Molecular cloning and mammalian expression of human beta 2-glycoprotein I cDNA. FEBS Lett 289, 183-186. Kroll, J., Larsen, J. K., Loft, H., Ezban, M., Wallevik, K., and Faber, M. (1976). DNAbinding proteins in Yoshida ascites tumour fluid. Biochim. Biophys. Actu 434,490-501. Laker, M. F., and Evans, E. (1996). Analysis of apolipoproteins. Ann. C h . Biochem. 33, 5-22.
La Rosa, L., Meroni, P. L., Tincani, A,, Balestrieri, G., Faden, A., Lojacono, A., Morassi, L., Brocchi, E., Del Papa, N., Gharavi, A. E., Sanimaritano, L., and Lockshin, M. A. (1994).p2-glycoprotein I reactivity of monoclonal anticardiolipin antibodies from patients with the antiphospholipid syndrome. J . Rheunratol. 21, 1684-1698. Lavalle, C., Pizarro, S., Drenkard, C., Sanchez-Guerrero, J., and Alarcon-Segovia,D. (1990). Transverse myelitis: Manifestation of systemic lupus etytheniatosus strongly associated with antiphospholipid antibodies. 1.Rheumatof. 17, 34-37. Lee, N. S., Brewer, H. B., and Osborne, J. C. J. (1983). Beta 2-glycoprotein I: Molecular properties of an unusual apolipoprotein, apolipoprotein H.J. Bid. Chem. 258,4765-4770. Lehmann, P. V., Sercarz, E. E., Forsthuber, T., Dayan, C. M., and Gammon, G. (1993). Determinant spreading and the dynamics of the autoimmune T-cell repertoire. Inzrnunol. To&/ 14,203-208. Le Tonqueze, M., Dueymes, M., and Piette, J. C. (1995).Is there a role for p2-glycoprotein I in the anti-phospholipid binding to endothelial cells. Lupus 4, 179-186. Leung, W. H., Wong, K. L., Lau, C. P., Wong, C. K., and Cheng, C. H. (1990).Association between antiphospholipid antibodies and cardiac abnormalities in patients with systemic lupus erythematosus. Am. J. Med. 89, 411-419. Levine, S. R., Deegdn, M. J., Futrell, N., and Welch, K. M. A. (1990). Cerebrovascular and neurologic disease associated with antiphospholipid antibodies: 48 cases. Neurology 40, 1181-1189. Levine, S. R., Kierdn, S., Puzio, K., Feit, H., Patel, S. C., and Welch, K. M. A. (1987). Cerebral venous thrombosis with lupus anticoagulants. Stroke 18, 801-804. Levine, S. R., Salowich-Palm, L., Sawaya, K. L., Perry, M., Spencer, H. J,, Winkler, H. J., Alam, Z., and Carey, J. L. (1997). IgG anticardiolipin antibody titer >40 GPL and the risk of subsequent thrombo-occlusiveevents and death: A prospective cohort study. Stroke 28, 1660-1665. Lockshin, M. D., Druzin, M. L., Goei, S., Qamar, T., Magid, M. S., Jovanovic, L., and Ferenc, M. (1985). Antibody to cardiolipin as a predictor of fetal distress or death in pregnant patients with systemic lupus erythematosus. N . Engl. J . Med. 313, 152-156. Loeliger, A. (1959). Prothrombin as cofactor in the circulating anticoagulant in systemic lupus erythematosus? Thromb. Diath. Haenwwh. 3, 237. Loizou, S., McCrea, J. D., Rudge, A,, Reynolds, A., Boyle, C. C., and Harris, E. N. (1985). Measurement of anticardiolipin antibodies by an enzyme-linked immunosorbent assay (ELISA): Standardisation and quantitation of results. Clin. Exp. biimunol. 62, 738-745. Lozier, J., Takahashi, N., and Putnam, F. W. (1984). Complete amino acid sequence of human plasma beta 2-glycoprotein I. Proc Nutl. Acud. Sci. U S A 81, 3640-3644. Lubbe, W. F., and Walker, E. B. (1983). Chorea gravidarum associated with circulating lupus anticoagulant:Successful outcome of preganncy with prednisone a i d aspirin therapy. Br. J. Obstet. Gynnecol. 90, 487-490. Lusins, j. O., and Szilagyi, P. A. (1975).Clinical features of chorea associated with systemic lupus erythematosus. Am. J . Med. 58, 857-861. Luthy, R., Bowie, J. U., and Eisenberg, D. (1992).Assessment ofprotein models with threedimensional profiles. Nature 356, 83-85.
ANTIPHOSPHOLIPID SYNDROME
5s7
Mackworth-Young, C. G., Gharavi, A. E., Boey, M. L., and Hughes, G. R. V. (1984). Portal and pulnlonaty hypertension in a case of systemic lupus erytheniatosus: Possible relationship with a clotting abnormality. E m J . Hheum. Injam. 7, 71-74. Mamula, M., and Janeway, C. (1993). Do B cells drive the diversification of iminune responses? Zmmutml. Today 14, 151-153. Mandreoli, M., Zuccala, A., and Zucchelli, P. (1992). Fibroinuscular dysplasia of the renal arteries associated with antiphospholipid autoantibodies: Two case reports. Am. J. Kidney Dis.20, 500-503. Marie, I, Levescjue, H., Heron, F.. Cailleux, N., Borg, J. Y., and Courtois, 1-1. (1997). Acute adrenal failure secondaly to bilateral infarction of the adrenal glands as the first manifestation of primary antiphospholipid syndrome. Ann. Rheum. Dis. 56, 567-.568. [Letter] Martinuzzo, M. E., Forastiero, R. R., and Carreras, L. 0. (1995). Anti-@2-glycoproteinI antibodies: Detection and association with thrombosis. B r . J. Haematof. 89, 397-402. Matsuda, J., Gohchi, K., Tsukamoto, M., Saitoh, N., Asanii, K., and Hashimoto, M. (1993). Anticoagulant activity of an anti-P2-glycoprotein I antibody is dependent on the presence of @2glycoprotein I. Am. J. Haematol. 44, 187-191. Matsumoto, A. K., Kopicky-Burd, J., Carter, R. H.,Tuveson, D.A., Tedder, T. F., and Fearon, D. T. (1991). Intersection o f the complement and immune systems: A signal transduction complex of the B lymphocyte-containing complement receptor type 2 and CD19.J. Exp. Med. 173,55-64. Matsuura, E.,Igarashi, Y., Yasuda, T., Triplett, D. A., and Koike, T. (1994). Anticardiolipin antibodies recognise @t-glycoproteinI structure altered by interacting with an oxygen modified solid phase surface. I. Exp. Med. 179, 457-462. Mbewu, A. D., and Durrington, P. N. (1990). Lipoprotein(a): Structure, properties and possible involvement in thrombogenesis and atherogenesis. Atherosclerosis 85, 1-14. McCrae, K.R., DeMichele, A. M., Pandhi, P., Balsai, M. J., Samuels, P., Graham, C., Lala, P. K., and Cines, D. B. (1993). Detection of antitrophoblast antibodies in the sera of patients with anticardiolipin antibodies and fetal loss. Blood 82,2730-2741. McLean, J. W., Tomlinson, J. E., Kuang, W. J., Eaton, D. L., Chen, E. Y., Fless, G. M., Scanu, A. M., and Lawn, R. M. (1987). cDNA sequence of human apolipoprotein(a) is homologous to plasminogen. Nutiire 330, 132-137. McNally, T., Cotterell, S. E., Mac!&, I. J., Isenberg, D. A,, and Machin, S. J. (1994). The interaction o f P2 glycoprotein-I and heparin and its effect on @Z glycoprotein I antiphospholipid antibody cofactor function in plasma. Thronzb. Hnemstnsis 72,578-581. McNeil, H. P., Chesterman, C. N., and Krilis, S. A. (1989). Anticardiolipin antibodies and ~ U ~ L anticoagulants I S comprise separate antibody subgroups with different phospholipid binding characteristics. Br. J. Haenwtol. 73, 506-513. McNeil, H.P., Chesteman C. N., and Kiilis, S. A. (1991). Immunology and clinical importance of antiphospholipid antibodies. Adv. Immunol. 49, 193-280. McNeil, H.P., Simpson, R. J., Chesterman, C. N., and Krilis, S. A. (1990).Antiphospholipid antibodies are Irected to a coniplex antigen that includes a lipid-binding inhibitor o f coagulation: @2-glycoproteinI (apolipoprotein H). Proc. Natl. Acad. Sci. USA 87,41204124. Miles, L. A,, Fless, C. M., Levin, E. G., Scann, A. M., and Plow, E. F. (1989). A potential basis for the thrombotic risks associated with lipoprotein(a). Nature 339, 301-303. Mizutani, H., Knrata, Y., Kosugi, S., Shiraga, M., Kashiwagi, H., Tomiyama, Y., Kanakura, Y., Good, R. A , , and Matsuzawa, Y. (1995). MonocIonaI anticardiolipin autoantibodies established from the ( N Z W x BXSB)Fl niouse model of antiphospholipid syndrome crossreact with oxidised low-density lipoprotein. Arthritis Rheum. 38, 1382- 1388.
558
DAVID A. KANDIAH et nl.
Moll, S., and Ortel, T. L. (1997). Monitoring warfarin therapy in patients with lupus anticoagulants. Ann. Int. Med. 127, 177-185. Monestier, M., Kandiah, D. A,, Kouts, S., Novick, K. E., Ong, G. L., Rddic, M. Z., and Krilis, S. A. (1996). Monoclonal antibodies from NZWx BXSB/Fl mice to p2-glycoprotein I and cardiolipin: Species specificity and charge-dependent binding. J Immunol. 156, 2631-2641. Mor, F., Beigel, Y., Inbal, A,, Goren, M., and Wysenbeek, A. J, (1989). Hepatic infarction in a patient with the lupus anticoagulant. Arthritis Rheum. 32, 491-495. Mori, T., Takeya, H., Nishioka, J., Gabazza, E. C., and Suzuki, K. (1996). p2-glycoprotein I modulates the anticoagulant activity of activated protein C on the phospholipid surface. Thromb. Haemostasis 75, 49-55. Morton, K. E., Gavaghan, T. P., Krilis, S. A,, Daggard, G. E., Baron, D. W., Hickie, J. B., and Chesterman, C. N. (1986). Coronary artery bypass graft failure: An autoimmune phenomenon? Lancet 2, 1353-1356. Nicholls, A,, Sharp, K. A,, and Honig, B. (1991). Protein folding and association: Insights from the interfacial and thermodynamic properties of hydrocarbons. Proteins 11,281-296. Nilsson, I. M., Astedt, B., Hedner, U., and Berezin, D. (1975). Intrauterine death and circulating anticoagulant, “antithromboplastin.” Ada. Med. Scand. 197, 153-159. Nimpf, J., Bevers, E. M., Bomans, P. H. H., Till, U., Wurm, H., Kosher, G. M., and Zwaal, R. F. A. (1986).Prothrombinase activity of human platelets is inhibited by p2-glycoprotein I. Biochim. Biophys. Acta. 884, 142-149. Nonaka, M., Matsuda, Y., Shiroishi, T., Moriwaki, K., Nonaka, M., and Natsuume-Sakai, S. (1992). Molecular cloning of mouse p2-glycoprotein I and mapping of the gene to chromosome 11. Genomics 13, 1082-1087. Norman, D. G., Barlow, P. N., Baron, M., Day, A. J., Sim, R. B., and Campbell, I. D. (1991).Three-dimensional structure of a complement control protein module in solution. 1.Mol. Biol. 219, 717-725. Oosting, J. D., Derksen, R. H. W. M., Bobbink, I. W., Hackeng, T. M., Bouma, B. N., and de Groot, P. G. (1993). Antiphospholipid antibodies directed against a combination of phospholipids with prothrombin, protein C, or protein S: An explanation for their pathogenic mechanism? Blood 81, 2618-2625. Oosting, J. D., Derksen, R. H. W. M., Entjes, H. T. I., Bouma, B. N., and de Groot, P. G. (1992). Lupus anticoagulant activity is frequently dependent on the presence of P2-glycoprotein I. Thromb. Haemostasis 67, 499-502. Oosting, J. D., Derksen, R. H. W. M., Hackeng, T. M., van Vliet, M., Preissner, K. T., Bouma, B. N., and de Groot, P. G. (1991). In-vitro studies of antiphospholipid antibodies and its cofactor, beta 2-glycoprotein I, show negligible effects on endothelial cell mediated protein C activation. Thromb. Haemostasis 66, 666-671. Ostuni, P. A,, Lazzarin, P., Pengo, V., Ruffatti, A., Schiavon, F., and Gambari, P. (1990). Renal artery thrombosis and hypertension in a 13 year old girl with antiphospholipid syndrome. Ann. Rheum. Dis. 49, 184-187. Pengo, V., Biasiolo, A., Brocco, T., Tonetto, S., and Ruffatti, A. (1996). Autoantibodies to phospholipid-binding plasma proteins in patients with thrombosis and phospholipidreactive antibodies. Thromb. Haemostasis 75, 721-724. Permpikul, P., Rao, L. V. M., and Rapaport, S. I. (1994). Functional and binding studies of the roles of prothrombin and p2-glycoprotein I in the expression of lupus anticoagulant activity. Blood 83, 2878-2892. Pierangeli, S. S., Harris, E. N., Gharavi, A. E., Goldsmith, G., Branch, D. W., and Dean, W. L. (1993). Are immunoglobulins with lupus anticoagulant activity specific for phospholipids? Br. J. Haernatol. 85, 124-132.
ANTIPHOSPIiOLIPID SYNDROME
559
Pierangeli, S. S., Liu, S. W., Anderson, G., Barker, J. H., and Harris, E. N . (1996).Thrombogenic properties of murine anti-cardiolipin antibodies induced by beta 2-glycoprotein I and human immunoglobulin G antiphospholipid antibodies. Circulation 94, 1746-1751. Piette, J. C., Cacoub, P., and Wechsler, R. (1994). Renal manifestations of the antiphospholipid syndrome. Semin. Arthritis Rheum. 23, 357-366. Polz, E.(1979). Isolation of a specific lipid-bindlng protein from human serum by affinity chromatography using heparin-sepharose. In “Protides of Biological Fluids” (H. Peeters, ed.), pp. 817-820. Pergamon Press, Oxford, England. Pomeroy, C., Knodell, R. G., Swain, W. R., Ameson, P., and Mahowald, M. L. (1984).BuddChiari syndrome in a patient with a lupus anticoagulant. Gastroenterology 86, 158-161. Ptitsyn, 0.B., and Finkelstein, A. V. (1983). Theory of protein secondary stnicture and algorithm of its prediction. BiopoZytnt.zs 22, 15-25. Puurunen, M., Jokiranta, S., Vaarala, 0.. and Men, S. (1995). Lack of functional similarity between complement factor H and anticardiolipin cofactor, P2-glycoprotein I (apolipoprotein H). Scund. J. Imtrunol. 42,547-550. Puurunen, M., Manttari, M., Manninen, V., Tenkanen, L., Alfthan, G., Ehnholm, C., Vxarala, O., Aho, K., and Palosuo, T. (1994). Antibody against oxidised low-density lipoprotein predicting myocardial infarction. Arch. lntenz. Med. 154, 2605-2609. Rao, L. V. M., Hoang, A. D., and Rapaport, S. I. (1995). Differences in the interactions of lupus anticoagulant IgG with human prothrombin and bovine prothrombin. Thronlb. Haemstasis 73, 668-674. Rauch, J., and Janoff, A. S. (1990). Phospholipid in the hexagoid I1 phase is immunogenic: Evidence for immunorecognition of nonbilayer lipid phases in vivo. Proc. Natl. Acad. Sci. USA 87,4112-4114. Rauch, J., and Janoff, A. S. (1992).The nature of antiphospholipid antib0dies.J. Rheumatol. 19,1782-1785. Rauch, J., Tannenbaum, M., and Fortin, P. R. (1997). Inhibition of lupus anticoagulant activity by hexagonal phase (11) phosphatidylethanolamine requires prothrombin as a cofactor. Arthritis Rheum 40(9), Suppl., 105. [Abstract 4451 Rauch, J., Tannenbaum, M., and Janoff, A. S. (1989).Distinguishing plasma lupus anticoagulants from anti-factor antibodies using hexagonal (11) phase phospholipids. Thromb. Haem.sta.si.s62,892. Reber, G.,Arvieux, J., Comby, E., Degenne, D., de Moerloose, P., Sanmarco, M., and Potron, G. (1995). Multicenter evaluation of nine commercial kits for the quantitation of anticardiolipin antibodies. Thromh. Haeinostasis 73, 444-452. Reid, K. B. M., and Day, A. J. (1989). Structure-function relationships of the complement components. Immunol. Today 10, 177-180. Roch, B., Khamashta, M. A., Atsumi, T., Bertolaccini, M. L., Amengual, O., and Hughes, G. R. V. (1997). Multiple antiphospholipid tests do not increase the diagnostic yield in antiphospholipid syndrome (APS). Arthritis Rheum. 40(9), Suppl., 300. [Abstract 16171 Roldan, C. A,, Shively, B. K., Lau, C. C., Gurule, F. T., Smith E. A,, and Crawford, M. H. (1992).Systemic lupus erythematosus valve disease by transesophageal echocardiography and the role of antiphospholipid antibodies. I . Am. Coll. Carrliol. 20, 1127-1134. Rost, B., and Sander, C. (1993). Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol. 232,584-599. Roubey, R. A. S., Eisenberg, R. A., Harper, M. F., and Winfield, J. B. (1995). Anticardiolipin autoantibodies recognize /32-glycoprotein I in the absence of phospholipid: Importance of Ag density and bivalent binding. J. Immnunol. 154, 954-960. Roubey, R. A. S., Pratt, C. W., Buyon, J. P., and Winfield, J. B. (1992). Lupus anticoagulant activity of autoimmune antiphospholipid antibodies is dependent upon /32-glycoprotein I. J. Clin. Invest. 90, 1100-1104.
560
DAVID A . KANDIAH et al.
Salemme, F. R. (1976). A hypothetical structure for an intermolecular electron transfer complex of cytochromes c and b5. J . Mol. Biol. 102, 563-568. Sali, A., and Blundell, T. L. (1993).comparative protein modelling by satisfaction of spatial restraints. J. Mul. Biol. 234, 779-815. Sali, A,, Matsuinoto, R., McNeil, H. P., Kai-plus, M., and Stevens, R. L. (1993). Threedimensional models of four mouse mast cell cliymases: Identification of proteoglycan binding regions and protease-specific antigenic epitopes. 1.Bid. Chem. 268, 9023-9034. Salonen, J. T., Yla-Herttuala, S., Yamamoto, R., Butler, S., Korpela, H., Salonen, R., Nyyssonen, K., Palinski, W., and Witztum, J. L. (1992). Autoantibody against oxidised LDL and progression of carotid atherosclerosis. Lancet 339,883-887. Sanchez, R., and Sali, A. (1997a). Evaluation of comparative protein structure modeling by (MODELLER-3).Proteins, suppl. 1, 50-58. Sanchez, R., and Sali, A. (1997b). Advances in comparative protein-structure modeling Curt-. Opin. Struct. Biol. 7, 206-214. Sanghera, D. K., Wagenknecht, D. R., McIntyre, J. A,, and Kamboh, M. I. (1997). Identification of structural mutations in the fifth domain of apolipoprotein H (beta 2-glycoprotein I ) which affect phospholipid binding. Hum. Mol. Genet. 6, 311-316. Schousboe, I. (1988). In vitro activation of the contact activation system (Hageman factor system) in plasma by acidic phospholipids and the inhibitory effect of beta 2-glycoprotein I on this activation. Int. J . Biochenz. 20, 309-315. Schousboe, I., and Rasmussen, M. S. (1995). Synchronized inhibition of the phospholipid mediated autoactivation of factor XI1 in plasma by fi2-glycoproteinI and anti-fi2-glycoprotein I. Throinb. Haemostasis 73, 798-804. Schultze, H. E., Heide, K., and Haupt, H. (1961). Uber einbisher unbekanntes niedermole kulares p2 globulin des human serums. Nntumissen Schaften 48, 719-724. Schwab, E. P., Schumacher, R., Jr., Freundlich, B., and Callegari, P. E. (1993). Pulmonary alveolar haemorrhage in systemic lupus erythematosus. Semin. Arthritis Rheum.23,8-15. Scott, J. (1991). Lipoprotein(a). Thrombotic and atherogenic. Br. Med. J. 303, 663-664. Scott, R.A. H. (1987).Anti-cardiolipinantibodies andpre-eclampsia. Br. I. Obstet. Gynaecol. 94,604-605. Seleznick, M. J., Silveira, L. H., and Espinoza, L. R. (1991). Avascular necrosis associated with anticardiolipin antibodies. J . Rheuintol. 18, 1416-1417. Sellar, G. C., Keane, J., Mehdi, H., Peeples, M. E., Browne, N., Whitehead, A. S., Ackers, G. K., and Smith, F. R. (1985). Annu. Rev. Biochem. 54, 597-629. Sheng, Y.,Herzog, H., and Krilis. S. A. (1997). CIoning and characterization of the gene 41, 128-130. encoding the mouse P2-glycoprotein I. Ge~~oinics Sheng, Y., Kandiah, D. A,, and Krilis, S. A. (1998). Anti-Pz-glycoprotein I autoantibodies from patients with the antiphospholipid syndrome bind to &plycoprotein I with low affinity. Diinerization of/3z-glycoproteinI induces a significant increase in anti-&-glycoprotein I antibody affinity. J. Ir~~mnunol., in press. Sheng, Y., Sali, A., Herzog, H., Lahnstein, J., and Krilis, S. A. (1996).Site-directed mutagenesis of recombinant human 02-glycoprotein I identifies a cluster of lysine residues that are critical for phospholipid binding and anti-cardiolipin antibody activity. J. In~mnunol. 157, 3744-3751. Shi, W., Krilis, S. A., Chong, B. H., Gordon, S., and Chesterman, C. N. (1990). Prevalence of lupus anticoagulant and anticardiolipin antibodies in a healthy population. Aust. N.Z . J. Med. 20,231-236. Shi, W., Chong, B. H., Hogg, P. J., and Chesterman, C. N. (1993).Anticardolipin antibodies block the inhibition by &glycoprotein I of the factor Xa generating activity of platelets. Throinb. Haenwstasis 70,342-345.
ANTIl’HOSPHOLIPID SYNDROME
561
Silver, R. M., Smith, L. A,, Edwin, S. S., Oshiro, B. T., Scott, J. R., and Branch, D. W. (1997).Variable effects on murine pregnancy of i~nmunoglobulinG fractions from women with antiphospholipid antibodies. Am. J. Obstet. Cynaecol. 177,229-233. Simantov, R., LaSala, J. M., Lo, S. K., Gharavi, A. E., Saminaritano, L. R., Salmon, J. E., and Silverstein, R. L. (1995). Activation of cultured vascular endothelial cells by antiphospholipid antibodies. 1. Clin. Zrauest. 96, 221 1-2219. Sminiov, M. D., Triplett, D. T., Comp, P. C., Esinoii, N. L., arid Esmon, C. T. (1995). On the role of phosphatidylethanolaniiiie in the inhibition of activated protein C activity by antiphospholipid antibodies. J. Cliii. [riuest. 95, 309-316. Sneddon, I. B. (1965). Cerebral vascular lesion in livedo reticularis. Br. 1. Den~zutol.77, 180- 185. Solice, M., Circella, A,, Grigi, T., Garofalo, T., Nicodenio, G., Pittoni, V., Pontieri, G. M., Lenti, L.,and Valesini, G. (1996). Anticardiolipin and anti-/32-CPI are two distinct populations of autoantibodies. Throinb. Haemostasis 75, 303-308. Stafford-Brady, F. J., Urowitz, M. B., Gladman, D. D., and Easterbrook, M. (1988). Lupus retinopathy: Patterns, associations, prognosis. Arthritis Rheum 31, 1105-1110. Steinkasserer, A,, Estaller, C., Weiss, E. H., Siin, R. B., and Day, A. J. (1991). Coinplete nucleotide a i d deduced amino acid sequence of human P2 glycoprotein I. Biochem. I. 277, 387-391. Steinkasserer, A,, Barlow, P. N., Willis, A. C., Kertesz, Z., Campbell, I. D., Sim, R. B., and Norman, D. G. (1992). Activity, clisulphide mapping and structural modelling of the fifth domain of human /32-glycoprotein I. FEBS Lett. 313, 193-197. Sugi, T., and McIntyre, J . A. (199F;). Autoantibodies to phosphatidylethanolamine (PE) recognize a kininogen-PE complex. Blood 86, 3083-3089. Sugi, T., and McIntyre, J. A. (1996).Autoantibodies to kininogen-phosphatidylethanolaniiiie complexes augment thrombin-induced platelet aggregation. Thromb. Res. 84,97-109. Takeya, H.,Mori, T., Gabazza, E. C., Kuroda, K., Deguchi, H., Matsuura, E., Ichikawa, J. K., Koike, T., and Suzuki, K. (1997).Anti-P2-glycoprotein I monoclonal antibodies with lupus anticoagulant-like activity enhance the P2-glycoprotein I binding to phospholipids. J. Cliri. Invest. 99, 2260-2268. Tan, E. M., Cohen, A. S., Fries, J. F., Masi, A. T., McShane, D. J., Rothfield, N. F., Schaller, J. G., Talal, N., and Winchester, R. J. (1982).The 1982 revised criteria for the classification of systemic lupus eytheinatosus. Arthritis Rheum 25, 1271-1277. Tincani, A,, Spatola, L., Prati, E., Allegri, F.. Ferremi. P., Cattaneo, R., Meroni, P., and Balestrieri, G. (1996). The anti-P2-glycoprotein I activity in human anti-phospholipid syndrome sera is due to rnonoreactive low-affinity autoantibodies directed to epitopes located on native 02-glycoprotein I and preserved during species’ evolution. J. Inunuaol. 157,5732-5738. Triplett, D. A. (1995).Antiphospholipid-proteinantibodies: Laboratory detection and clinical relevance. Tlaromh. Res. 78, 1-31. Triplett, D. A. (1996). Antiphospholipid-protein antibodies: Clinical use of laboratory test results (identification, predictive value, treatment). Hnenwstnsis 26(Suppl. 4), 358-367. Tuveson, D. A,, Ahearn, J. M., and Matsumoto, A. K. (1991). Molecular interactions of complement receptors on B lymphocytes: A CR 1/CR2 complex distinct from the CRY CD19 coliiplex. J. EX^. Med. 173, 1083-1089. Vaarala, 0. (1997). Atherosclerosis in SLE and Hughes syndrome. Lupus 6, 489-490. Vaarala, O., Alfthan, G., Jauhiainen, M., Leirisalo-Repo, M., Aho, K., and Palosuo, T. (1993). Crossreaction between antibodies to oxidised low-density lipoprotein and to cardiolipin in systemic lupus erythe‘matosus. Lancet 341, 923-925.
562
DAVID A. KANDIAH et al.
Vaarala, O., Manttari, M., Manninen, V, Tenkanen, L., Puurunen, M., Aho, K., and Palosuo, T. (1995). Anti-cardiolipin antibodies and risk of myocardial infarction in a prospective cohort of middle-aged men. Circulation 91, 23-27. Vaarala, O., Puurunen, M., Manttari, M., Manninen, V., Aho, K., and Palosuo, T. (1996). Antibodies to prothrombin imply a risk of myocardial infarction in middle-aged men. Thromb. Haemstasis 75, 456-459. Vela, P., Battle, E., Salas, E., and Marco, P. (1991). Primary antiphospholipid syndrome and osteonecrosis. Clin. Exp. Rheumatol. 9, 545-549. Verrot, D., San-Marco, M., Dravet, C., Genton, P., Disdier, P., Bolla, G., Harle, J.-R., Reynaud, L., and Weiller, P.-J. (1997). Prevalence and significance of antinuclear and anticardiolipin antibodies in patients with epilepsy. Am. J. Med. 103, 33-37. Vianna,J. L., Khamashta, M. A., Ordi-Ros,J., Font, J., Cervera, R., Lopez-Soto, A., Tolosa, C., Franz, J., Selva,A,, Ingelmo, M., Vilardell, M., and Hughes, G. R. V. (1994).Comparison of the primary and secondary antiphospholipid syndrome: A european multicenter study of 114 patients. Am. J. Med. 96, 3-9. Viard, J. P., Amoura, Z., and Bach, J. F. (1992). Association of anti-/32-glycoprotein-I antibodies with lupus-type circulating anticoagulants and thrombosis in systemic lupus erythematosus. Am. J. Med. 93, 181-186. Vlachoyiannopoulos, P. G., Tektonidou, M., Petrovas, C., Krilis, S. A,, and Moutsopoulos, H. M. (1998). Antibodies to p2-glycoprotein I: Avidity, binding specificity and association with thrombosis. Submitted for publication. Walsh, M. T., Watzlawick, H, Putnam, F. W., Schmid K., and Brossmer, R. (1990). Effect of the carbohydrate moiety on the secondary structure of /32-glycoprotein I: Implications for the biosynthesis and folding of glycoproteins. Biochemistry 29, 6250-6257. Wang, C . R., Hsieh, H. C., Lee, G. L., Chuang, C. Y., and Chen, C. Y. (1992). Pancreatitis related to antiphospholipid antibody syndrome in apatient with systemiclupus erythematosus. J. Rheunmtol. 19, 1124-1125. Wang, M. X., Kandiah, D. A,, Ichikawa, K., Khamashta, M., Hughes, G. R. V., Koike, T., Roubey, R. A,, and Krilis, S. A. (1995).Epitope specificityof monoclonal anti-/32-glycoprotein I antibodies derived from patients with the antiphospholipid syndrome. J. Immunol. 155, 1629-1636. Willems, G. M., Janssen, M. P., Pelsers, M. M., Comfurius, P., Galli, M., Zwaal, R. F. A,, and Bevers, E. M. (1996). Role of divalency in the high-affinity binding of anticardiolipin antibody-beta 2-glycoprotein I complexes to lipid membranes. Biochemistry 35,1383313842. Woodard, C., Brey, R. L., Hart, R. G., and Kagan-Hallet, K. (1991). Neuropathological findings in stroke associated with antiphospholipid antibodies. Neurology 41, 296. Wurm, H. (1984). Beta 2-glycoprotein I (apolipoprotein H) interactions with phospholipid vesicles. Int. 1.Biochem. 16, 511-515. Yamazaki, M., Asakura, H., Jokaji, H., Saito, M., Uotani, C., Kumabashiri, I., Morishita, E., Aoshima, K., Ikeda, T., and Matsuda, T. (1994). Plasma levels of lipoprotein(a) are elevated in patients with the antiphospholipid sundrome. Thrornb. Huemostasis 71, 424-427. Yang, C. D., Chen, S. L., Shen, N., Pei, J., Xue, F., Lu, Y., Gu, W., and Bao, C. D. (1997). The fifth domain of P2-glycoprotein I contains antigenic epitopes recognised by anticardiolipin antibodies derived from patients with the antiphospholipid syndrome. APLAR J. Rheumatol. 1, 96-100. Yin, E. T., and Gaston, L. W. (1965). Purification and kinetic studies on a circulating anticoagulant in a suspected case of lupus erythematosus. Thromb. Diath. Haemorrh. 14, 88.
ANTIPHOSPHOLIPID SYNDROME
563
Ziporen, L., Goldberg, I., Arad, M., Hojnik, M., Or&-Ros, J., Afek, A,, Blank, M., Sandbank, Y., Vilardell-Tarres. M., de Torres, I., Weinberger, A., Asherson, R. A,, Kopolovic, Y., and Shoenfeld, Y. (1996). Libman-Sachs endocarditis in the antipliospholipid syndrome: Iinmunopathologic findings in deformed heart valves. Lziptts 5, 196-205. Ziporen, L., Shoenfeld, Y., Levy, Y., and Korczyn, A. D. (1997). Neurologcal dysfunction and hyperactive behaviour associated with antiphospholipid antibodies. 1. Clin. Invest. 100, 613-619. This article was accepted for publication on February 4, 1998.
This Page Intentionally Left Blank
INDEX
A Activation-induced cell death cbaracterization, 15-19 T cell deletion, 62 Adaptive immunity, 83-84 ADP-iibosylation Factor. see GTPases Adrenal gland, liypofiinction, 516-517 AICD, see Activation-induced cell death Aiway h~erresponsiveness,154-1.55 Allograft rejection. 155-157 Antibodies anticardiolipin, 543-544 antiprothrombin. 534-535 anti-02-glycoprotein I, 510-512 antiphospholipid, 508-509, 52-537 lupus anticoagiilant efrects. 535-536 laboratory studies, 544-545 location, 509-510 Anticardiolipin antibodies, 543-544 Antigenic spreading, 546 Antigen-presenting cells location, 315 MCH I1 loading, 320-321 targeting soluble antigens, 321-323 Antiphospholipid antibodies characterization, 508-509 endothelial cells, 536-537 A~itjphospholipidlsyndrome adrenal manifestations, 516-5 17 animal models, 531-533 anticardiolipin antihodies, 543-544 antiphospholipid antibodies, 536-537 APS-associated. 515 AVN-associated, 5 1R-5 19 cardiovascular manifestations, 512-513
characterization. 507-508 dermal nranifestations, 518 endothelial cells, 536-537 fuhire therapies, 545-547 02-glycoprotein I antihody binding sites, 523 characterization, 520-523 gene encoding characterization, 530-531 cloning, 530-531 molecular modeling alignment, 524-525 cardiolipin-binding region, 529-530 comparisons, 524-525 electrostatic properties. 528-529 fifth domain, 525, 527-528 structure, 523-524 phospholipid sites. 523 thrombosis link, 539-540 hepatic manifestations, 517 laboratory studies, 543-548 lripiis anticoagulant antibodies, 535-536 laboratmy studies, 544-545 neurological manifestations, 513-514 obstetric manifestations, 519-520 pathogenesis, 537-543 phos~~hatidyletlianolainine, 536 protein C activation. 535-536 piilmonary manifestations, 5 15-5 16 renal rnanifestations. 516 sllnlnlaly, 547-548 APC, we Antigen-presenting cells Apical surface. 399-400 Apoptosis, see dso Programmed cell death AICD, 15-19 IL-2, 14-19 inhibition, 251-253 initiating factor, 258-259
INDEX
mechanism, 269-270 mitochondria1 control, 257-259 promotion, 253-255 protein inhibition, 251-253 APS, see Antiphospholipid syndrome Arterial thrombosis, 513 Arthritis, collagen-induced, 162-164 Autoimmune disease organ-specific, 157-158 pregnancy effects, 519 spontaneous, 165-166 Autoimmune vascular disease, 531-532 Avascular necrosis of bone, 518-519 AVN, see Avascular necrosis of bone
B Bacteria, see also specijic species Bcl-2 homology, 263-264 IL-12 inhibition, 177-181 Basolateral surface, 399 B cells antigenic spreading, 546 CD4' expression, 327-328 CD4+ priming, 325-326 development, 42-47 1L-18 effect, 289 IL-12 effects, 148-151 IL-12 production, 102-104 IL-18R expression, 296-297 tolerizing, 547 BcI-2 apoptosis inhibition, 251-253 promotion, 253-255 cell physiology, 261 cell sunival, 266-269 characterization, 250-251 function, 256-257 ion channels, 265-266 bcalization, 255-256 structure, 263-264 Bcl-x,, CED, 268-269 ion channels, 265-266 structure, 261-263 Bone, avascular necrosis of, 518-519 Bordetelln pertusssis, 180 Borrelia hrgdorferi, 180
Brucelln abortus, 180 Brugin inalnyi, 187
C CnndicZu albicans, 184-185 Cardiolipin, 529-530 Cardiovascular disease, 512-513 Caspase-1, 286-287 CED Bd-x,,, 268-269 characterization, 248-250 mammalian homologs, 260-261 Cell cycle regulation, 11-12 Cell death, see Apoptosis Cell survival, 266-269 Cell viability pathways, 41-42 c-~).Y> 26-27, 41 Chaperones, protein assembly, 385-387 Chinese hamster ovary (CHO) cells, 91 CIA, see Collagen-induced arthritis Ciliary neurotrophic factor, 92-94 c-jttn, 41 c-myc, 26-27, 41 CNTF, see Ciliary neurotrophic factor Coat proteins families, 374-375 functions, 376-377 GTPases regulating, 377-381 Coccidioides inmitis, 186 Colitis, experimental, 164-165 Collagen-induced arthritis, 162-164 Crohn's disease, 164-165 Cyptococms neofonnans, 185 Cutaneous necrosis, superficial, 518 Cytokines, see ulso specijic cytokine CD4+ secretion, 332-334 IL-12-induced, 122-126 production, 146-148 Cytomegalovirus, 168-171
D Delayed-type hypersensitivity, 152-153 Dendritic cells, 113-114 Diabetes mellitus, 303-304 Dilysine motif, 391-392 Diphtheria toxin, 264
567
INDEX
E EAE, see Experiinental-allergic encephalomyelitis EAU, see Experimental uveoretinitis EBV-induced protein 3, 94 Ectrornelia, 344-346 Encephaloinyelitis, experimental-allergic, 160-162 Endocytosis, 39,3-394 Endoplasmic reticulum bidirectional transport, 389-390 degradation studies, 387-389 description, 369-372 MHC 11, 391 retention studies, 387-389 Endosomal system, 372 Endothelid cells, 536-537, 542 Endotoxin-induced liver injury, 301-303 Epidermal growth Factor receptor, 396 Epilepsy, APS-associated, 514 Epitope mapping, 523 ER, see Eiidoplasmic reticulum N-Ethylinaleimide, 381-382 Evolution coinbinatorial system, 418 foundations gene specificity, 421-424 mechanism, 418-420 superfamilies, 420-421 immunoglobulins amphibians. 437 birds, 437-438 bony fishes, 436-438 chondrichthytes, 433-436 emergence, 425-430 gene organization arrangement, 479-480 basic elements, 475-476 intron removal, 476-477 promoters, 478-479 species comparisons, 480-485 transcription, 477-478 heavy chains isotypes, 470-475 universal p, 465, 467-470 variable domains, 460, 463-465 jawed vertebrates, 430-433 light chains, 451-460 mammals, 438-439
molecular events, 485, 488-491 reptiles, 437-438 overview, 417 T-cell receptors, 425-430 jawed vertebrates, 430-433 species comparisons, 441-451 variable domain, 439, 441 Experimental-allergic encephalomyelitis, 160-162 Experimental colitis, 164-165 Experimental uveoretinitis, 162 Experimental viral infections, 168-171
F Factor H, 524-525 Fas receptors, 63
G PZ-Glycoprotein I antibody binding sites, 52-3 characterization. 520-523 gene encoding characterization, 530-531 cloning, 530-531 inoleciilar modeling alignment, 524-525 car&olipin-bin&ng region, 529-530 comparisons, 524-525 electrostatic properties, 528-529 fifth domain, 525, 527-528 structure, 523-524 phospholipid sites, 523 thrombosis link, 539-540 Anti-p2-Glycoprotein I antibodies, 510-512 Golgi network &-, 391 trans-, 371 Graft-uerms-host disease, 153-154 Growth factors activation, model, 19-20 IL-2, 10-11 GTPases subfamilies, 377-381 GTP-binding protein, 393-394
H Hematopoietic stein cells, 119-121 Hemorrhages, intraalveolar pulmonary, 515
568
INDEX
Hepatitis B virus, 340-341 Herpes simplex virus, 348-349 Histoplasma capstdatum, 185-186 HIV, see Human immunodeficiency v i m Human immunodeficiency virus IL-12 production, 172-176 pathological changes, 339-340 Hypercholesteroleinia, APS-associated, 513 Hyperresponsiveness,airway, 154-155 Hypersensitivity, delayed-type, 152-153 Hypoprothrombineriiia, APS-associated, 516
I IDDM, see Insulin-dependent diabetes melIitus Iininature dendritic cells, 316 Iininune system adaptive responses, 83-84 evolution combinatorial system, 418 foundations gene specificity, 421-424 mechanism, 418-420 superfamilies, 420-421 immunoglobulins, 425-430, 463-475 amphibians, 437 birds, 437-438 bony fishes, 436-438 chondrichthytes, 433-436 gene organization, 475-491 heavy chains, 460 jawed vertebrates, 430-439 light chains, 451-460 mammals, 438-439 molecular events, 485, 488-491 reptiles, 437-438 overview, 417,491-492 T-cell receptors, 425-430 jawed vertebrates, 430-439 species comparisons, 441-451 variable domain, 439. 441 function, 313-314 IL-2R role, 58-60 infectious agents and, 85, 351-352 Immunogenicity, 531-533 Immunoglobulin-binding protein, 385-387 Iminunoglobulins evolution amphibians, 437
birds, 437-438 bony fishes, 436-438 chondrichthytes, 433-436 emergence, 425-430 gene organization arrangement, 479-480 basic elements, 475-476 intron removal, 476-477 promoters, 478-479 species comparisons, 480-485 transcription, 477-478 heavy chains isotypes, 470-475 universal p, 465, 467-470 variable domains, 460, 463-465 jawed vertebrates, 430-439 light chains, 451-460 molecular events, 488-491 reptiles, 437-438 IgE, IL-18 effect, 292-294 intracehlar transport, 400-402 Infections, experimental viral infections, 168-171 Inflammatttion, 166-168 Influenza, 341-344 Insulin-dependent diabetes mellitus, 303-304 Insulin receptor substrate-1, 33 Interleukin-12 allograft rejection, 155-157 antitumor effects, 187-193 bacteria, 177-181 B cell response, 148-151 CHO cells, 91 CIA, 162-164 cloning, 86 colitis, 164-165 Crohn’s disease, 164-165 discovery, 85-86 EAE, 160-162 EB13, 94 expression, 114-1 19 fungal pathogens, 184-186 graft rejection, 153-154 helminthic parasites, 186-187 heterodimers, 90-92 homodimers, 90-92 IiyI)erresponsiveness,154-155 hypersensitivity, 152-153 IL-18-regulated, 294-295
INDEX
induction cytoldne, 122-126 dendritic cells, 113-114 infectious pathogens, 106- 107 inflammatory extrucelhlar matrix, 107 modulation, 108-113 T-cell dependent, 107-108 inflainmation. 166-168 mitogenic activity, 127-129 midtiple sclerosis, 160-162 NK cell-induction, 129-131 NK1 T cell induction, 133-134 nonhuinan sources, 89-90 organ-specific autoiininune disease, 157-158 production B cell, 102-104 measurement, 101-102 phagocytic cells. 104-106 protozoan parasites. 181-184 p35 subunit, 87-89 p40 subunit, 86-87, 90-92, 94 purification. 8G receptor. 95- 101 signaling, 95-101 spontaneous autoimmune disease, 165-166 stem cell effects, 119-121 structure, 92-94 T-cell induction, 131-133 Thl cells differentiation. 141-143 generation importance, 138-140 N K cells role, 139-140 process, 143-146 role, 134-136 polarization, 140-141 vaccinations, 148-151 viruses experimental, 168-171 HIV, 172-176 MAIDS, 176-177 ineasles, 171- I72 Interleukin- 18 B cell effect, 289 biologv, 287-294 characterization, 304-305 history, 282-283 host defenses, 298-301 IgE effect, 292-294
IL-12 regidation, 294-295 NK cell effects, 291-292 osteoclast effects, 294 pathological roles, 301-303 processing, 286-287 producing cells, 285-286 receptors, 294-298 role, 281-282 structure, 283-285 T cell effect, 287-289 Interleuldn-2 receptors binding, 3-4 cell cycle regulation, 11-12 cell survival regulation, 14-19 expression, 7-10 fiinction, 2 iinmune fimction, 58-60 lyniphocytes characterization, 42-47 genetic studies, 47-49 stnicture, 53-58 mechanism, 19-21 PHOX region, 24-26 PRHIII, 29-30 signaling cell viability, 41-42 characteristics, 21-22 downstream factors, 37-38 downstream pathways, 24-27 generation, 1-2 intracellular, 21-24 Stat proteins, 27-30 IRS-I, 33 JAK 3 defects, 52-53 lymphocyte effector function, 13-14 M A P kinase pathways, 30-32 mitogenic, 34-37 P13 kinase, 38-39 responses to, 10-19 'SHPS, 33-34 Src fantily kinases, 32-33 STAM role, 40-41 Stat& 38 target genes, 41 TOR role, 39-40 subunit chains cloning, 4-5 sharing, 5-6 striictiise, 4-5
569
570
INDEX
T cells deletion, 61-63 growth, 63-64 IntraceUular transport coat proteins, 374-377 endocytic pathways, 372 description, 392-393 mechanisms, 393-394 MHC 11, 396-398 T-cell activation, 394-396 GTPases regulation description, 377-381 SNARE mediation, 381-384 subfamilies, 377-381 mechanism, 373-374 overview, 369, 402 polarized cells apical surface, 399-400 basolateral surface, 399 description, 398-399 polymeric immunoglobulin receptor, 400-402 secretory pathway bidirectional transport, 389-390 characterization, 385 immunoglobulin-binding protein, 385-387 MHC 11, 390-392 T-cell antigen receptor, 387-389 secretory pathways characterization, 369-372 in vivo cytoskeleton role, 384-385 models, 383-384 organelle structure, 385 Ion channels, 265-266 Ischemia focal cerebral, 513 ocular, 515
J Janus tyrosine kinases mitogenic signaling, 34-35 signaling pathways, 24-27
KDEL proteins, 380, 390 Kidneys disease, 516 PI3 Kinase, 38-39
L LCMV, see Lymphocytic choriomeningitis viiruS Leishmania major, 299 Leishmania spp., 181-183 Lipoproteins, low-density, 537-539 Listeria mononjtogenes, 177-181 Livedo reticularis, 518 Liver diseases, 517 injury, 301-303 Lung diseases, 515-516 Lupus anticoagulant antibodies, 535-536 characterization, 509-510 laboratory studies, 544-545 phosphatidylethanolamine and, 536 LY294002, 39 Lymphocyte effector function, 13-14 Lymphocytes, see also specijic lymphocytes development characterization. 42-47 genetic studies, 47-49 role, 49-52 yL deficiency, 52-53 peripheral, structure, 53-58 Lymphocytic choriomeningitis virus, 168-171, 336,338-339 Lysosomal associated membrane protein-1, 397
x
MAIDS, see Murine AIDS Major histocompatibility complex class I1 antigen presentation, 317-321 endocytic compartments, 396-398 secretory system, 390-392 Malaria, 299-300 MAP kinase pathways, 30-32 MCMV, see Murine cytornegalovims Measles virus CD4+ control, 344 immune deviation, 171-172 Mesenteric vessels, thrombosis, 517 Meth-A tumor, 192 Microangiopathy, 516 Migraines, 514
571
INDEX
Mouse mammary tumor virus. 350-351 Mulitiple organ failure, 303-304 Multiple sclerosis, 160-162 Murine AIDS, 176-177 Murine cytomegalovirus, 168- 171 Mycohacterium auiurn, 179 Mycohacteriurn leprae, 179 Mycobadenurn tuberculosis, 178-179 Myocardial infarction, 533-534
Natural killer cells development, 46-47 IL-18 effect, 291-292 IL-12-induced, 129-131 Th1 generation, 139-140 Natural killer cell stimulatory factor, see Interleukin- 12 NEM-sensitive factor, 381-382 Nyppostrungyliis brasiliensis, 187
0 Organ-specific autoimmune disease, 157-158 Osteoclast. 294
P p35, 87-89 p40, 86-87, 90-92, 94 Permeability transition, 257-259 Phagocytic cells, 298 IL-2 production, 84 Phosphatidylethanolamine, 536 Phospholipids, 523 Phygocytic cells, 104-106 Picornavimses, 349-350 Pla.srorliurn spp., 184 Polarized cells apical surface, 399-400 basolateral surface, 399 characterization. 398-399 polymeric imrnunoglobulin receptor, 400-402 Polio virus, 349-350 Polymeric iniinunoglobulin receptor, 400-402
Poxvinises, 344-346 Pregnancy, failures, 519 Producing cells, 285-286 Programmed cell death, see also Apoptosis genetics, 247-250 pathway, 251 significance, 245-247 Protein assembly, 385-387 Protein C, 535-536 Prothronibin, 533-535 p40 subunit, 90-92, 94
Recombinase activating gene 1, 421-424
Snlnwnellri dublin, 179-180 Scliistosoma IruLnsoni, 186- 187 Serine/threonine kinases, 38-39 Severe combined immunodeficiency IL-2 defective, 59-60 IL-2 factors, 48-49 XSCID, 49-50 Signal recognition particle, 369 Skin lesions, 518 SNARE hypothesis, 373 vesicle fusion, 382-383 Spontaneous autoimmune disease, 165-166 Src family Idnases, 32-33 SRP, see Signal recognition particle STAM protein, 40-41 Stat transcription factors description, 27-30 IL-ZR, 27-30 IL-lZR, 100-101 Stem cells, hematopoietic, 119-121 Struiigyloi&s stercoralis, 187 Synthetic mimotope peptides, 547
T Target of rapamycin, 39-40 T-cell receptors antigen, 387-389
572 evolution emergence, 425-430 jawed vertebrates, 430-433 species comparisons, 441-451 T cells antigenic spreading, 546 CD4' activation administration, 314-319 APC targeting, 321-323 dosage, 314-319 location, 314-319 MHC I1 molecules, 317-321 Th cells, 324-327 B cell priming, 325-326 characterization, 330 effector functions cytokines, 332-334 cytomegalovirus, 347-348 hepatitis B virus, 340-341 herpes simplex virus, 348-349 HIV, 339-340 influenza, 341-344 kinetics, 332 LCMV, 336, 338-339 measles virus, 344 mechanism, 330-331 mouse mammary tumor virus, 350-351 picornaviruses, 349-350 poxviruses, 344-346 vesicular stomatitis virus, 346-347 function, 313-314 overview, 351-352 presentation, 314-319 soluble antigens MHC class 11, 317-319 development, 42-47 growth L 2 R , 63-64 IL-18 effects, 287-289 IL-18 expression, 294-295 IL-12 induced, 107-108 IL-12-induced, 131-133 IL-18R expression, 296-297
INDEX
NK1
IL- l2-induced, 133- 134 proliferation, 10-13 regulation, 394-396 Theiler's virus, 349-350 T helper cells activation, 324-327 autoimmunity, 157-158 differentiation. 141-143, 141-146 effector functions, 330-336 generation, 134-136 IL-12 role, 134-140 NK cells role, 139-140 IL-18 effect, 289-291 polarization, 140-141 stability, 136-138 Thrombosis anitprothrornbin-link, 534 deep venous APS-associated, 515 glomerular capillary, 516 P2-glycoprotein I link, 539-540 mesenteric vessels, 517 TN thymocytes, 46 Toxoplasina gondii, 183 Transporter of antigenic peptides, 390-391 Trichuris muris, 187 Ttypanosom cmzi, 183-184 Tumors, 187-193 Tyrosine kinases, 34-35 Tyrosine phosphatases, 33-34
Uveoretinitis, experimental, 162
v Vaccinations, 148-151 Vaccinia, 344-346 Venoocclusive disease, hepatic, 517 Vesicle fusion, 382-383 Vesicular stomatitis virus, 346-347
CONTENTS OF RECENT VOLUMES
Volume 68
Volume 66
Peptide Bindii'g Specificity"Id HLA 'lass Molecular and Cellular Mechanisms of T Autoimmunity Lymphocyte Apoptosis JUERCXN HAMMER. TIZ~ANA STURNIOI.O, JOSEF M PENNINCER A N D GUIDO A N D FRANCESCO SINIGAGLIA KHOEMEH Prenylation of R a GTPase Superfamily Proteins and Their Function in 1111111unobi 01 ogy ROBERTB. L O ~ E L I .
Role of Cytoldnes in Sepsis C. ERIKHACK, LUCIEN A. AARDEN, ~NI) LAMBERTUS G. TIIIJS Role of Macrophage Migration Inhibitory Factor in the Regulation of the Iininune Response CHRISTINE N. ME= AND R I C H A R J I BUCAL.A
Generation and TAP-Mediated Transport of Peptides for Major Histocompatibility Complex Class I Molecules MOMBURG A N D GUNTHER J. FRANK HAMMERLING
The Intrinsic Coagulation/Kinin-Fori~li~~~ Cascade: Assembly in Plasma and Cell Surfaces in InAainination ALLENP. KAPLAN,KLl5UMAM JOSEPH, YOJI SHIBAYAMA, SESIIA REDDIGARI, BEIIIHANE
Adoptive Tumor Iininunity Mediated by Lymphocytes Bearing Modified AntigenSpecific Receptors BROCKEH AND KLA~JS THOMAS KARJALAINEN
GHEBRESIIWET, A N D M I C H A E L
SILVEHBERC
CDB' Cells in Human Ininiunodeficiency
Membrane Molecules as Differentiation Antigens of Murine Macrophages ANDREWJ. M c K ~ i
Virus Type I Pathogenesis: Cytolytic and
Noncytolytic Inhihition of Viral Repliwtion OlTO 0. YANG A N D BRUCE D. WAL.KER
Major Histocompatibility Complex-Directed Susceptibility to Rheuinatoitl Arthritis GERALD T. NEPOM
INDEX
Volume 67
lnimuno~ogicalTredtnient of Autoimmune Diseases J. R. KALIIEN,F. C. BREEDVELD. €1. B U R K I I A H D T , A N D G R. BURMESTER
CUM~JLATIVE INDEX
INDEX
57-3
574
CONTENTS OF RECENT VOLUMES
Volume 69 Molecular and Cellular Events in Early Tliymocyte Development RODEWALDAND HANS HANS-REIMER JORG FEHLINC Regulation of Immunoglobulin Light Chain Isotype Expression FREDERICK W. ALT A N D JAMES R. GORMAN Role of Iinmiinoreceptor Tyrosine-Based Activation Motif in Signal Transduction from Antigen and Fc Receptors NOAHISAKOV Atypical Serine Proteases of the Complement System GERARDJ. ARLAIJD,JOHN E. VOLANAKIS, NICOLE M. THIELENS, STHANAM V. L. NARAYANA, VERONIQUERossr, AND YUANYUANXu
ISBN 0-Lt-022470-4
Accessibility Control of V(D)J Recombination: Lessons from Gene Targeting ISABELLE LEDUC, WILLIAM M. HEMPEL, NOELLE MATHIEU,RA] KAMAL THIPATHI, A N D PIERRE FERRIER Interactions between the Immune System and Gene Therapy Vectors: Bidirectional Regulation of Response and Expression JONATHAN S. BHOMBERG, LISADEBRUYNE, AND L I ~ I U QIN I Major Histocompatibility Complex Genes Influence Individual Odors and Mating Preferences Pons DUSTIN PENNA N D WAYNE Olfactory Receptor Gene Regulation ANDREW CHESS INDEX
