ADVANCES IN
Immunology VOLUME 71
This Page Intentionally Left Blank
ADVANCES IN
Immu EDITED BY
FRANK J. DIXON The Scripps Research Institute La Jolla, California ASSOCIATE EDITORS
Frederick Alt K. Frank Austen Tadamitsu Kishimoto Fritz Melchers Jonathan W. Uhr
VOLUME 71
ACADEMIC PRESS San Diego London Boston New York Sydney Tokyo Toronto
hhh
This book is printed on acid-free paper.
@
Copyright 0 1999 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 U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-1999 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. 0065-2776/99 $30.00
Academic Press
a division of Harcourr Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.apnet.com Academic Press 24-28 Oval Road. London NWI 7DX, UK http://www.hbuk.co.uWap/ International Standard Book Number: 0-12-02247 1-2 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
apIy8 Lineage Commitment in the Thymus of Normal and Genetically Manipulated Mice
HANSJOHC I7EHI,IN(:,
SUSAN
GILFILIAN, AND RHOIIHICEREDIC:
1 16 Analysis of TCR Gene Rearran ements in c.P and y6 Lineage Cells 19 35 Analysis of TCR Transgenic an! Gene-Targeted Mice 52 Cell Culture Studies 53 Developmental Considerations 54 In Search of a Consensus Model for the &y8 Lineage Split 64 Refihrences
I. Introduction Models of crP/yG Lineage Chmmitnient and Lineage Maintenance
11. 111. I\’. V. VI. VII.
lmmunoregulatory Functions of y8 T Cells w1~1.i Ram, CAHOI. CA
I J~E,5 W A JoNEs-CAHSONAKlhO MICHAELL a m , A N D REBECYA O ’ B R I E ~
MUhASA,
I. Introduction 11. Origin, Lineage and Development, and Distribution III. Specificity IV. Functions V. Concluding Remarks References
77 78 83 93 123 124
STATs as Mediators of Cytokine-Induced Responses
TIMOTHY HOEYA N D MICIIAEI. J. G H U ~ H Y
I. Introduction 11. The STAT Gene Family
111. Structural and Functional Domains in STAT Proteins
145 145
146
vi
CONTEXTS
IV. V. VI. VII.
STAT-Deficient Mice STAT Function in Cellular Proliferation and Disease Regulation of STAT Function Suiniiiary and Perspective References
152 155 156 157 158
CD95(APO-1/Fas)-Mediated Apoptosis: Live and Let Die
PETERH. KRAMMER I. Introduction 11. Death Receptors and Ligands
111. The CD95/CD95L Systein IV. Gene Defects in the CDya5/CD95LSystem V. Role of the CD95/CD95L System ill Deletion of Peripheral T Cells VI. Role of the CD95/CD95L Systein in Liver Homeostasis VII. Signal Transduction of CD95-Mediated Apoptosis VIII. The Death Domain IX. CD95 Associated Si naling Molecules X. Other Signaling Mo ecules Invoked in CD95 signaling XI. Proteins of the Bcl-2 Family XII. The Death-Inducing Signaling Complex (DISC) XIII. Downstream Caspases in CD95 Death Receptor Signaling XIV. Type I and Type I1 Cells XV. FLIPS (FLICE Inhibitory Proteins) XVI. Sensitivity and Resistance of T Lyinphocytes toward CD95-Mediated Apoptosis XVII. The CD95 System and Chemotherapy XVIII. The CD95 Death System in AIDS XIX. Further Considerations on the Role of Apoptosis in the Clinic References
k
163 164 166 167 168 169 169 170 170 172 172 174 176 180 181 182 184 188 190 192
A CXC Chemokine SDF-l/PBSF: A Ligand for a HIV Coreceptor, CXCR4
TAKASHI NACASAWA, KALUNOBU TACHIBANA, A N D KENJIKAWABATA I. Introduction
211
11. Identification, Structure, and Expression of CXC Chemokine
SDF- 1/PBSF 111. Physiological Functions of SDF-l/PBSF
IV. A SDF-l/PBSF Receptor, CXCR4 V. HIV-1 Infection and CXCR4 VI. Perspectives Refirences
212 215 217 219 222 222
vii
CONTEKTS
T Lymphocyte Tolerance: From Thymic Deletion to Peripheral Control Mechanisms BRICITTASTOCKIN(;ER I. Introduction 11. Central Tolerance Induction in the Thymus 111. Peripheral Tolerance References
229 229 240 25 1
Confrontation between lntracellular Bacteria and the Immune System ULHICH
I. II. 111. IV. V. VI. VII. VIII. IX.
E. SCHAIBLE, HELENL. c O l , l , l N S ,
AND
STEFAN H. E. ~
Introduction What Is an Intracellular Pathogen? How to Enter the Host Cell Is How to Survive Induction of Nonspecific Immuni Phagosoine Maturation and Micro ,id Detours Antigen Processin and Presentation Pathways T-cell Subsets an Effector Mechanisms Host Genetics Influencing the Outcome of Infection Immune Intervention Strategies References
T
8
INDEX CONTENTS OF KECEIVTVOINMES
U F M A N N
267 268 269 273 283 292 307 322 326 336 379 385
This Page Intentionally Left Blank
CONTRIBUTORS
Willi Born (77),Departinent of Medicine, National Jewish Medical and Research Center, Denver, Colorado 80206; and Department of Immunology, University of Colorado Health Sciences Center, Denver, Colorado 80962 Carol Cady (Ti),Department o f Iinniunology, University of Colorado Health Sciences Center, Denver, Colorado 80262 Rhodri Ceredig (l),Centre de Recherche d’Iminunologie et Hematologie, F-67091 Strasbourg, France Helen L. Collins (267), M ~ L \Planck Institute for Infection Biology, D-10117 Berlin, Germany Hans Jorg Fehling (11, Basel Institute for Ininiunology, CH-4005 Basel, Switzerland Susan Gilfillan ( l),Basel Institute for Immunology, CH-4005 Basel, Switzerland Michael J. Grusby ( 145), Departinent of Iniinunology and Infectious Diseases, Harvard School of Public Health, and Department of Medicine, Harvard Medical School, Boston, Massachusetts 02115 Timothy Hoey ( 145), Tularik, Inc., South San Francisco, California Jessica Jones-Carson ( T ) Department , of Imniunoloby, University of Colorado Health Sciences Center, Denver, Colorado 80262 Stefan H. E. Kaufmann (267),Max Planck Institute for Infection Biology, D-10117 Berlin, Gerinany Kenji Kawabata (211), Department of Immunology, Research Institute, Osaka Medical Center for Maternal and Child Health, Izumi, Osaka 594-1101, Japan Peter H. Krammer ( 163),Tumor Iuiniunology Prograin, German Cancer Research Center, D-69120 Heidelberg, Germany Michael Lahn (77),Department of Medicine, National Jewish Medical and Research Center, Denver, Colorado 80206 is
X
CONTRIRUTORS
Akiko Mukasa (77),Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts Takashi Nagasawa (21l),Department of Immunology, Research Institute, Osaka Medical Center for Maternal and Child Health, Izumi, Osaka 594-1101, Japan Rebecca O’Brien (77),Department of hiledicine, National Jewish Medical and Research Center, Denver, Colorado 80206; and Department of Immunology, University of Colorado Health Sciences Center, Denver, Colorado 80262 Ulrich E. Schaible (267), Max Planck Institute for Infection Biology, D-10117 Berlin, Germany Brigitta Stockinger (229), Division of Molecular Immunology, National Institute for Medical Research, London NW7 lAA, United Kingdom Kazunobu Tachibana (211), Department of Immunology, Research Institute, Osaka Medical Center for Maternal and Child Health, Izumi, Osaka 594-1101, Japan
\I>\\U( I \ I N l\!\!UhOl~X.\ LO1 i I
ap/ yS lineage Commitment in the Thymus of Normal and Genetically Manipulated Mice HANS JORG FEHUNG,’ SUSAN GILFILLAN,’ AND RHODRI CEREDlGt hsel Instihrte lor Immunology, CH-4005 h e / , Switzedond; ond t Centre de Recherche d’lmmunologie el Hemoblogie, F-6709I Slrosbourg, France
I. Introduction
The purpose of this article is to review the mechanism of @/yS T-cell lineage commitment in tlie tlwmus. As way of introduction, some basic and largely historical aspects ofT-cell development will be reviewed briefly with particular reference to differences that might exist between the generation of a0 versus y6 cells. A more detailed analysis of some of these points can be found elsewhere (Distelhorst and Dubyak, 1998; Kang and Raulet, 1997; Kisielow and von Boelimer, 1995; Robey and Fowlkes, 1998; Rodewald and Feliling, 1998). Although natural killer ( N K ) T cells are discussed with particular reference to their presence in the thymus, detailed analysis of their phenotypic development is beyond the scope of this review.
A. SOMEUNIQUEFLATUHP\ o~ T H E yS T-CELLLINLAGE 1 Identijkation of yS T Cclls
The profound immunodeficiency observed in neonatally thyinectornized mice (Miller, 1961) led to the realization that tlie thymus was a primary lyniphoid organ responsible for tlie generation of thymus-derived, or T cells (Miller, 19S9). Following these seminal observations, it was assumed that there was only one lineage of T cells. With the diwovery that T cells were clearly part of the adaptive inimune system, showing phenomena of specificity and memory, it seemed obvioiis that, like B cells, T cells should also express clonally distributed receptor molecules. However, the fact that T-cell receptor (TCR) molecules were not secreted meant that their identification was protracted and required the advent of inore sophisticated biochemical (Allison and Lanier, 198’i) and molecular biological (Hedrick et nl., 1984) approache5 in the 1980s. It must be emphasized that it was only by such bioclieniical and molecular approaches that yS receptor inolecules were first identified (Raulet, 1989). It would now appear that all jawed vertebrates (gnathostomes) generate two distinct types of T cells, which are characterized by the mutually exclusive expression of (YPor yS T-cell receptor isotypes (Rast et al , 1997). Cells expressing y6 receptors were first identified by flow cytoinetry with monoclonal antibodies to hu1
( op,nplit 0 1YYY h\ 4~id( m~ he,\ 411 ngl+ of rc pndii(tim i n u n l o r n n w r w d
2
I1ANS ]OR(:
FEHLINC: ct 01
man CD3 and ap TCR as CD3+,ap TCR- cells (Brenner et al., 1986). It was then shown that the mouse fetal thymus also contained yS cells that migrated to peripheral lymphoid organs prior to the generation of ap T cells (Fowlkes and Pardoll, 1989; Havran and Allison, 1988; Pardoll et al., 1987). 2. Obvious Functional Difierences between
ap and
yS T Cells
That neonatally thymectomized mice, which presumably contained peripheral y6 T cells, nevertheless showed immunodeficiency (Miller, 1961) raises the important issue of the biological role of 76 T cells. If one considers that the hallmark of cells of the classical adaptive immune system is (1)antigen specificity and (2) memory, then the position of 7 6 cells in such a scenario is perplexing. Recognition of antigen by yS cells is dependent neither on CD4/CD8 co-receptor expression by the T cells themselves nor on expression of classical MHC molecules on the cells being recognized (Haas et al., 1993; Schild et al., 1994). Antigen recognition by yS T cells is quite different from that of ap T cells. Indeed, crystallographic analysis of yS TCR (Li et al., 1998) appears to confirm the idea that recognition of antigen by yS TCR is more like that of antigen by immunoglobulins than that by ap TCR. This becomes particularly relevant when one considers the possible positive (Haas et al., 1993) or negative (Boismenu and Havran, 1997) selection events affecting y6 T-cell development in the thymus and the form, if any, of their selecting ligands. The second issue is that of meniory within the y6 T-cell compartment. Although there appears to be a proliferative response among y6 cells to viral (Cardinget al., 1990),bacterial (Skeen and Ziegler, 1993),and parasitic (Rosat et al., 1993) infections, whether a specific memory component is generated following this proliferative phase is less clear (Mombaerts et al., 1993). This is in striking contrast to that seen in ap T cells where elegant studies have shown selection and expansion of cells with characteristic TCR clonotypes following immunization (Brawand et al., 1998; MacDonald et al., 1993; McHeyzer-Williams and Davis, 1995). In addition, primed ap T cells show modified activation thresholds upon rechallenge (Iezzi et al., 1998), but such information for y6 cells is sparse (Carena et al., 1997). The question of the turnover and life span of yS cells in unimmunized normal and y6 TCR transgenic mice has been addressed (Tough and Sprent, 1998). Thus, most thymic emigrant yS T cells appeared to have a restricted life span as naive cells. However, some yS cells converted to a memory phenotype as judged by acquisition of the CD44'"gh,CD62L'"", HSA'"", CD45RB"ghphenotype. It will be interesting to see if naive and memory phenotype yS cells in normal mice show different activation characteristics.
aplyS IJNE?Z(:E COMMITMENT
3
3. Anatomical Considerations with Regard to Thymic T-Cell Developirient In the intervening period between tlie discovery of T cells and their corresponding receptor molecules, many studies addressed the issue of how the thymus generated T cells. In this regard, differences between ap and yS cells clearly exist. Conibined histological and [ 'Hlthyniidine labeling experiments indicated that the thymus was divided into two main anatomical regions (Metcalf, 1966). First, a predominantly outer cortex comprising 80-90% of cells and where both cell division and death occurred (McPliee et nl., 1979; Shortinan and Jackson, 1974); cortical cells were smaller in diameter and in a compact organization. Second, a predominantly inner medulla where cell division and death were rare (Egerton ct al., 1990); medullary cells were larger in size and inore widely spaced than in the cortex (Metcalf, 1966). From these combined studies, the notion was put forward that T cells were generated in the cortex and that cortical cells were the direct precursors of cells in the medulla (Shortman and Jackson, 1974). It was quickly realized that the vast majority of thyinocytes were destined to die in sitri, a fincling that seemed perplexing before it was realized that the process of apoptosis is a major feature of both B and T lymphocyte development and is linked to t1i.e requirement for receptor selection (Kisielow and von Boehmer, 1995). For the ap lineage, more recent refinements of this cortical to rnedullaiy differentiation model certainly corroborate these earlier findings. However, the situation for yS cells is less clear. By immunohistoclieinical analysis with antibodies to surface y6 TCR, the few y6 cells in the thymus are mostly found in clusters in the cortex (Farr et al., 1990). Little information is available as to their subsequent transit through the thymus, although from labeling experiments it would seem that they probably migrate to the periphery from the medulla (Kellyet nl., 1993).Importantly, as discussed by Tough and Sprent (1998),the kinetics of ap versus yS thymocyte selection may be quite different, with 76 cells being generated and exported from the thymus more rapidly than ap cells, which in tlie adult thymus require a prolonged sojourn in the rnedulla prior to emigration. 4. The Queytion
(fPositive mid Negatiue Selection
The question oftlie developmental site and movement ofy6 cells through the thymus during differentiation is not a trivial one because, for a@cells, the transit from the cortex to the medulla is associated with receptor selection events. Positive selection takes place in the cortex whereas negative selection may take place in both cortex and medulla (Anderson et nl., 1997; Kisielow and von Boehmer, 1995; Merkenschlager et ul., 1997; Punt
et al., 1997). The death of cells during intrathymic development is due either to absence of positive selection or to negative selection. Negative selection of developing y6 cells (Dent et al., 1990) seems to be generally accepted. Whether positive selection also takes place is less certain (Schweighoffer and Fowlkes, 1996; reviewed in Haas et nl., 1993; Robey and Fowlkes, 1998). The issue of positive selection is particularly pertinent with regard to those T-cell subsets expressing invariant 76 receptors. These subsets include so-called dendritic epidermal cells (DECs), which are located in the skin and mostly bear a canonical Vy3’NSl TCR, or y6 cells in the reproductive tract, which express predominantly invariant Vy4N61 TCRs. Experiments by Mallick-Wood et nl. (1998) have revealed that mice lacking the Vy3 chain due to targeted gene disruption are capable of generating almost normal numbers of DECs expressing a y6 TCR with a similar, conserved conformational determinant (idiotype)as found in wildtype mice, despite the use of another nondeleted Vy gene segment. This result provides convincing evidence for the positive selection of y6 cells, at least with regard to this particular y6 subset. At face value, these new findings seem to contradict earlier studies by Asarnow et al. (1993), who used transgenic TCRy minigenes as artificial recombination substates to demonstrate that directed gene rearrangements-even in the absence of the possibility for selection-resulted in efficient formation of the invariant Vy3 junctional sequence. However, both findings can be easily reconciled by assuming that the generation of the highly restricted TCR repertoire of dendritic epidermal cells is the result of two processes: (a) biased gene rearrangements mediated by the recombination machinery and (b) subsequent selection of cells bearing TCRs with the respective invariant determinant. 5. Sensitivity to Glucocorticoids and Cyclosporin A Administration of glucocorticoids to mice results in a dramatic depletion of 100% of cortical and about 50%of medullary thyinocytes 48 hr after drug administration (Blomberg and Andersson, 1971). From this observation, medullary cells, like peripheral “mature” T cells, are “resistant” to glucocorticoids and are therefore called “mature” thymocytes to distinguish them from their glucocortioid-sensitive cortical “immature” partners (Ceredig et al., 1982). The glucocorticoid-mediated death of thymocytes is by apoptosis, and evidence implicates the purinergic receptor P2XI and an inosito1 1,4,5-trisphosphate receptor ( IP3R) in mediating this process (Distelhorst and Dubyak, 1998). For ap cells, the transition from glucocorticoid “sensitive” to “resistant” occurs immediately post-TCR receptor selection
’ Nomenclature throughout this article is according to Garinan et 01.
(1986)
&yG
LINEAGE COM MlTM E NT
i5
(Crompton et al., 1992; Tolosa ct al., 1998). Little information is available on differences in glucocorticoid sensitivity between immature and mature y6 cells. Interestingly, the development of yS but not a0 cells is largely resistant to the administration of cyclosporin A (Robey and Fowlkes, 1998). However, cyclosporin A does have an effect on the phenotypic maturation of intrathymic yS cells (Leclercq et al., 1993).
B. ARRIVINC: AT THE DN + DP + SP MODELOF THYMOPOIESIS 1. CD4 and CD8 as Usefiil, Developnmtal Stage-Specijic Cell Sii $ace Marker,$ The expression of serologically detectable markers was soon found to provide an important parameter for following T-cell development within the thymus. In general, expression of surface markers is used to define different stages within cell lineages, usually a valuable approach (for potential pitfalls, see Section I,B,4). The first such serological marker was the Thy-1 (CD90) antigen (Reif and Allen, 19641, which was considered to define cells of the T lymphoid lineage. By several criteria, CD90 was found not to be uniformly expressed on thymocytes, with small cortical cells expressing more CD90 than their larger medullary descendants (Ceredig et al., 1982). CD90 expression is also low on the very earliest cells in the thyinus and can be practically absent on some cells with T-cell characteristics, notably among intraepithelial lymphocytes ( IEL) (Lefrancois and Goodman, 1989).This variation in CD90 antigen expression may be linked to the presence of multiple promoters within the CD90 gene (Spanopoulou et al., 1991). There are also large species variations in the expression of CD90, with peripheral T cells in rats being mostly CD90- (Hosseinzadeh and Golschneider, 1993). However, at the time, with anti-CD90 reagents no clear dichotomy of peripheral T-cell subsets was observed. The advent of serology identified a series of T lymphocyte (Lyt) alloantigens, which for the first time subdivided mouse peripheral T cells into two phenotypically and functionally distinct populations, namely ( 1) Lyt1 (CDS)+,Lyt-2 (CD8a)-, Lyt-3 (CD8P)- “helper” and (2) Lyt-1-, Lyt2+, Lyt-3’ “cytotoxic cells” (Cantor and Boyse, 1975, 1977).CD5 was later shown to be expressed by a subset of B lyniphocytes (Hardy et al., 1994). Both subsets of peripheral T cells were derived from thymic precursors expressing all three Lyt alloantigens (Kisielow et al., 1975). Application of monoclonal antibody technology to human T cells and thyinocytes resulted in the identification of two subpopulations of mature T cells expressing the antigens CD4 and CD8 in a mutually exclusive fashion (Reinherz et al., 1980; Reinherz and Schlossman, 1980). This dichotomy of T-cell phenotype was particularly attractive given that there
6
H A N S J o H G FEHLJNG rt 01
appeared to be an association between CD4/CD8 phenotype and the specificity of MHC antigen recognition (Swain, 1980). When the human thymus was analyzed, cortical cells were found to express both CD4 and CD8 and were thus called double positive (DP),whereas medullary cells, like peripheral T cells, expressed either CD4 or CD8 and were called single positive (SP) (Janossy et al., 1980). Combining the cortical to medullary anatomical model outlined earlier with the phenotypic results of CD4 and CD8 expression, the DP (cortical) to SP (medullary) transition of human thymocytes was proposed (Reinherz and Schlossman, 1980). With a monoclonal antibody (GK-1.5) to the mouse CD4 antigen (Dialynas et al., 1983), peripheral T cells and medullary thymocytes were found to be either CD4 or CD8 (SP) and cortical cells DP (Ceredig et al., 1983). In addition, these studies identified a small subpopulation of cells in the thymus that expressed neither CD4 nor CD8, so-called double negative (DN) cells. Based on the observation that 100% of cells in the day 15 mouse fetal thymus were DN and that DP cells first appeared at day 16, 2 days before SP cells at days 18 to 19, for ap T cells the DN to DP to SP model of mouse thyrnocyte development was proposed (Ceredig et al., 1983; Fowlkes and Pardoll, 1989) (see Fig. 1A). This scheme appears valid for "classical" ap TCR-expressing cells not bearing the NK-1.1 marker. The DN to DP transition of thymocytes was directly demonstrated by in vitro culture (Ceredig et nl., 1983) and in vivo transfer experiments
FIG.1 . (A) A Simplified schematic representation of adult mouse thyinocyte differentiation. Thyinocyte sul)popiilations are outlined based on their expression of CD4 and CD8 and their relative proportions indicated as a percentage. For conventional (non-NK1. l + ) aP T cells. development progresses from cells expressing neither CD4 nor CD8 (DN) (lower left) to DP cells expressing both antigens (upper right). Efficient transition from DN to DP is contingent on successful TCRP rearrangement and pre-TCR expression (see text for details). Most cells transit directly; howcver, some DN cells proceed to DP via a CD8+/CD4- intermediate; such cells have been called immature single positives (ISP). In some mouse strains, CD4' ISP can also be detected. Following aPTCR receptor selection, DP cells become either CD4 or CD8 single positives (SP). (B) Subpopulations of mouse CD4-/CD8- (DN)thymocytes. A schematic representation of DN thyinocyte subpopulations defined by their expression ofCD25 and CD44. CD44 expression varies from weakly positive (-/low) to bright ( + + ) a n d ,togetherwith CD25, helps define four subsets of DN thymocytes that have been called CD25-/CD44" DN#1, CD2St/CD44' DN#2, CD25'/CD44-""" DN#3, and CD25-/CD44-"""' DN#4. This scheme highlights the heterogeneity of the DN#1 subset. B cells can he distinguished by expression of CD19, NK. and NK T cells by their expression of N K 1 . l in appropriate mouse strains and by weak expression of CD117 (c-kit) and, finally, dendritic precursors by expression of C D l l c and MHC class 11. T precursor cells can be distinguished by bright expression of CD117. See text for further details.
9
I
I
CD8 B DN#3
I I I
+
-/low
CD44-
++
(Fowlkes et al., 1985). Importantly, these in vitro experiments, when combined with the DNA-labeling technique (Ceredig and MacDonald, 1985; Ceredig et al., 1983; Sekaly et al., 1983),indicated that the differentiation to DP cells i n uitro was a process independent of cell division. Later experiments showed that the transition from DN to DP in vim at the population level was accompanied by a burst of rapid cell division (Hoffman et al., 1996; Howe and MacDonald, 1988). However, whether all cells undergoing this transition do so by dividing has not been determined. It should be recalled that in other cell differentiation systems, e.g., gut epithelial cell development, cellular differentiation, as defined by changes in cell phenotype, may occur independently of cell division (Simon and Gordon, 1995). The transition of thymocytes from DP to SP was initially difficult to directly demonstrate in vitro, but has been subsequently confirmed by many groups. Several phenotypic changes are associated with the transition from DP to SP, including changing cell size, downregulation of CD24 and CD90, and upregulation of CD69, and is a topic that has been adequately reviewed elsewhere (Kisielow and von Boehmer, 1995).
2. Heterogeneity cf DN Tliywwcytes Phenotypic analysis of purified DN cells indicated that they were themselves heterogeneous for the expression of several markers, including CD3, TcRyG, TcRaO, CD25, and CD44 (Fowlkes and Pardoll, 1989).DN thymocytes depleted of CD3' ap and yS cells are called triple negative (TN) cells (Godfrey and Zlotnik, 1993). In the adult but not fetal thymus (Antica et d ,1993), the earliest populations of TN cells are weakly CD4 positive, becoming negative at the CD25 stage (Wu et al., 1991). It has been suggested, however, that the CD4 molecules on such cells are passively acquired, presumably froin surrounding DP cells (Michie et al., 1998). Additional refinements to the TN developmental sequence have included c-kit (CD117),the stem cell factor receptor (Godfreyet al., 1992; Matsuzah et al., 1993). Thus, in both fetal and adult thymus, the earliest (TN#1) subset is CD117'/CD2,5-/CD44', which then progresses through a CD 117+/CD25+/CD44'(TN#2) stage to CD 117-/CD25'/CD44-"" (TN#3) and finally to CD117-/CD25-/CD44-"'" (TN#4) cells (Fig. 1B). With sensitive flow cytoinetric techniques, purified TN cells do not show completely biphasic profiles with any of these markers. Indeed, expression of CD117 byTN#l cells is quite heterogeneous in the thymus of recombination activating gene knock-out (RAG KO) mice, ranging from high on a small subset to low on a population of mature N K cells, which are found in the thymus of both RAG KO and normal mice (Carlyle et nl., 1998). In fact, CD117'"" mature NK cells constitute the majority of TN#1 cells in adult RAG KO mice (R. Ceredig, unpublished data).
Several important events take place at the CD25 stage of thymocyte development. For instance, it was demonstrated that the L7Pto (D)JP rearrangement of TCR genes occurs among CD117-/CD25+ TN#3 cells and that subsequent transition to the CD2Fi- TN#4 subset is contingent upon in-frame TcRP rearrangeinents (Mallick cf al., 1993). This process has been called “TCRP selection” and is mediated by the pre-TCR (reviewed by Fehling and von Boehiner, 1997; Kodewald and Fehling, 1998). It should be recalled that D to JP rearrangements are not unique to T cells and that the nioleciilar indicator of T-cell coniinitment is the VP to (D)JP rearrangement. This is equally valid for B cells where the V,, to (D)JHrearrangement inarks B-cell commitment. In the B lyinphocyte lineage, CD25 expression is chiuacteristic of pre-BII cells (Rolink et nl., 1994), a stage following successful Ig,, rearrangements at the CD1 lT/ CD25- pre-B1 cells (Osmond et al., 1998). Based on CD11’7 and CD25 expression, pre-B 1 cells resemble TN#1 thymocytes, cells that contain little, if any, TCR VP+(D)JP rearrangcnients ( Koyis~ict al., 1997). This differing pattern of CD25 expression by developing T and B cells indicates that there is, most likely, no physiologically relevant relationship between receptor gene rearrangement events and CD25 expression. I n contrast, activation of CD25 transcription may be a completely fortuitous event due to the presence of a particular combination of transcription factors at ii given developmental stage ( Ivanov and Ceredig, 1992; Rothenberg and Ward, 1996). Although changes in CD25 expression on niaturing thymocytes itre kipparently of no functional iniportance, CD25 clearly provides a very useful developmental marker, particularly when iised in combination with CD44. Heterogeneous CD25 and CD44 expression has therefore become the most frequently used marker system to subdivide the DN thymocyte population in a developmentally meaningful way. Figure 1 B represents a scheme based on these two markers that illustrates the developmental progression of CD3-CD4-CD8- (TN) cells in the adult thymiis along the four CD25/ CD44-defined stages (TN#l-TN#4). Tlie scheme also reveals the distinct heterogeneity of DN thyniocytes in the adult mouse. The inclusion of a few additional markers leads to further refinement, allowing the attribution of most CD2XD44-defined DN subsets to ii distinct developmental stage or lymphoid cell lineage. Apart from conventional a/3 and y6 T cells, the following cell types can be identified within the DN thymocyte subpopulation. a. B Cells. The thymus contains a distinct population of B cells that can be phenotypically distinguished from peripheral blood B cells transiting the thymus (R. Ceredig, unpublished observations). To exclude thymic B
10
H A N S J o R G FEHLINC, t,t nl
cells from DN or TN preparations, the most frequently used marker has been B220. This is problematic for two reasons. First, B220 expression on thymic B cells is very heterogeneous and can overlap with the negative control. This is most dramatically seen in IL-7 transgenic mice where the number of thymic B cells increases threefold (R. Ceredig, unpublished observations). Second, B220 expression is not unique to B cells and is induced on T cells undergoing apoptosis in vivo (Renno et al., 1996). Whether CD25t DN#3 cells undergoing apoptosis also express B220 is not clear. Some thymic B cells are weakly CD25' and their expression of CD44 is also variable (see Fig. 1B). Thus, B cells, which could potentially contain DP+JP TCR rearrangements, could contaminate all subsets of DN cells. This problem could be overcome either by using CD19 to define B cells or analyzing thymuses from B-less mice, such as membrane p-KO (pmT),animals. Alternatively, fetal mice, whose thymuses contain far fewer B cells, could be used.
b. N K T Cells. The presence of a subpopulation of CD3'"', TcRap"" CD24- CD44' DN cells in the thymus was initially a very puzzling observation (Budd et al., 1987;Cerediget al., 1987; Fowlkes et al., 1987).However, they are now known to belong to a separate lineage of so-called N K T cells (reviewed in Bendelac et al., 1997; MacDonald, 1995; Vicari and Zlotnik, 1996), which transit through the thyinus during their developmental program. N K T cell precursors appear to be present in the fetal mouse thymus (Ceredig, 1988).Importantly, human (Battistini et al., 1997) and mouse (Vicari et al., 1996;Williams et al., 1997) NK T cells expressing y6 TCR have been identified. In the mouse, like their ap partners, they are probably mostly CD24'""', but it will be of interest to determine their TCR receptor repertoire. In the thymus, NK TCRaP cells express CD3 and CD117 weakly but can be identified in appropriate mouse strains using the N K 1 . l marker (Koyasu et al., 1997). NK T cells in the thymus and in the periphery can express CD4 (Arase et al., 1993; Chen et al., 1997; Hoshimoto and Paul, 1994). TCR a/3 N K T cells express surface TCR molecules encoded for by the products of a single Va14Ja281TCRa gene combined with a particular repertoire of three, predominantly V08, but also V/37 and VP2 genes. Apart from being CD3t, NK T cells share many phenotypic properties with NK cells, which are also in the thymus of normal mice. RAG KO mice have been used as a source of DN#1 cells, but as mentioned earlier, these preparations are enriched for NK cells. From gene knockout experiments (Mendiratta et al., 1997),the development of a/3 N K T cells appears to depend on TCR receptor engagement by CD1. The crystal structure of mouse C D l d l has identified a hydrophobicbinding site occupied by glycosylphosphatidylinositol that could constitute
aplyS 1.1h'EhCE COMMITMENT
11
the natural ligand of CDldl-restricted NK T cells (Joyce et d., 1998). In addition, cytokine signaling through the IL-R cominon y (7') chain for commitment and through the IL-7Ra chain for expansion (Boesteanu rt al., 1997) of N K T cells has been demonstrated. These signals appear to 199i),the gene for which is involve interactions with IL-15 (Ohteki et d., regulated by the interferon regulatory factor-1 ( IRF-1) transcription activator. Analysis of IRF-l-deficient mice shows that this factor regulates IL-15 gene expression and thereby N K T-cell development (Ohteki et nl., 1998). Because of their DN and CD3"" phenotype, complete elimination of NK T cells from "TN" preparations may well be difficult to achieve. However, because of their unique TCRa chain and restricted Vp rearrangements, ap NK T cells can be distinguished easily from cells in the major pathway of ap T-cell development (Koyasu et al., 1997).For such analyses, measurement of rearrangements involving Vp8, 7 and 2 genes should be avoided. c. Dendritic Cells. DN#1 cells also contain precursors of dendritic cells. Thymic dendritic cells are CD44'"g'', CD117'"", MHC class 11' and CDllc'. Based on transfer experiments, it was proposed that early thymocytes contain cells capable of forming B/T/NK and dendritic cells (Ardavin et ul., 1993; Wu et al., 1991). None of these experiments were done at the clonal level. In mice with mutations in both CD117 and the IL-R yt chain, the thymus contains an apparently nornial population of antigen-presenting dendritic cells, despite the absence of thyniocyte progenitors. This finding seems more compatible with the concept of separate T/dendritic cell precursors (Rodewald and Fehling, 1998). The remarkable diversity of DN thyinocytes places severe constraints on attempts to study TCR rearrangeinents by single-cell polyinerase chain reactions (PCK) in this population because the heterogeneity of cells within the DN#1 compartment is problematic for the phenotypic identification of T precursor cells. Some contaminants (e.g., N K T cells) are mature T cells with distinct TCR rearrangements, whereas others could contain Dp+Jp rearrangements, yet not be in the T-cell lineage (e.g., B cells). DN#2 cells can be clearly identified as CD117""~1", CD44', CD25' cells, whereas DN#3 have become C D l l 7 and CD44 dull. At the population level, there appears to be a distinct quantitative increase in TCR Vp+(D)JP rearrangements at this DN#2 to DN#3 transition (Koyasu et d ,1997; 1997). As mentioned earlier, qualitative changes in TCR Tourigny et d., Vp+( D)JP rearrangements take place at the DN#3 to DN#4 transition (Mallick c't al., 1993).
12
3. Position of y6 Tliyiiwcytes within the C D 4 K D 8 Decelopnientul S c h i e In the adult mouse, the vast majority of76 cells are clearly part of the DN thyniocyte population, as they Fail to express CD4 and CD8 coreceptors. Initially, it was considered that developing yS cells were exclusively DN (Raulet, 1989). Although this holds true for the vast majority of TCRy6bearing cells in the adult thymus, y6expressing cells that coexpress CD4 and/or CD8 do exist. At day 16 of mouse fetal thymus development, about 50% of y6 cells express CD8 (Fisher and Ceredig, 1991). CD8 can be expressed by y6 cells upon activation (Goodman and Lefrancois, 1988) and is also seen on a subpopulation of intraepithelial y6 cells (Guy-Grand ct d.,1991). Because y6 cells in the fetal thymus are actively cycling (Ceredig, 1990; Fisher and Ceredig, 1991), this may explain why some are CD8’. Expression of CD4 by cloned y6 cell lines (Spits et d., 1991; Wen ct d., 1998), fetal thymocytes (Fisher and Ceredig, 1991), and y6 cells in pre-Ta KO mice (Fehling et al., 1997) has been reported. Additional markers, including CD5, CD45RB, CD62L (Mel-14), and CD24 (HSA), have been used to study the development of fetal intrathymic Vy3+ cells (Leclercq ct al., 1993). These authors concluded that in analoa with a/3 cell development, y6 cell development went froin TCR“””/CD24h’g’’ to TcR‘”g”/CD24‘””and that this transition was affected by cyclosporin A. Another point of view is that HSA”p” y6 cells represent newly formed and HSA“’” activated, or memory, cells (Tough and Sprent, 1998). Unfortunately, at present, the phenotypic analysis of 76 lineage cells is not sufficiently advanced to allow the design of a generally accepted developmental scheme, like the one for a/3 lineage cells. However, a large body of experimental data provides convincing evidence that the differentiation of y6 T cells does not follow a DN to DP to SP transition. For instance, the number of y6 thymocytes is normal or even augmented in many gene knockout mouse strains in which the generation of DP and SP thymocytes is severely hampered, strongly suggesting that CD4 and CD8 expressing thyinocytes are generally not obligatoryintermediates in y6 cell development. Teleologically this makes sense, as the physiologic function of CD4 and CD8 molecules is to interact with MHC, and the analysis of MHC-deficient inice has revealed that this interaction is dispensable for norinal y6 T-cell development. Notwithstanding CD4 and CD8 expression by y6 T cells in certain situations as mentioned earlier, the obligatory steps in y6 maturation therefore seem to occur exclusively within the DN compartment. It thus seems safe to confine a search for possible intermediary stages in the yS developinental pathway to the DN population.
4. A Note of Cnrttion Regnrditig the Us.e of S u f k c c hilurker Exprossioir to ~ c j CCU i Liningcs ~ The variability in cell surface marker expression by TN thymocytes raises the issue, perhaps extreme, of whether it is justified at all in using cell surface phenotyFe to define the lineage relationship of cells. This is particularly relevant for the expression of molecules which apparently play no functional role in thymocyte dcvelopment, like CD2,5, CD44, CD90, and even CD4 aiid CD8, as thr as their expression on the most immature thymocyte precursors is coiicerned. It shoiild also be recalled that detection of antigens by flow microfluoriinetry requires thc expression of several thousand cell surface molecules. For cell surface molecules with receptor function, expression of a few hundred molecules may be sufficient to traiisduce a biological response. Consequently, cells expressing sufficient receptors to respond to the corrcsponding ligand can nevertheless appear negative by flow microfluoriinetry. A particularly cogent example is CD25, the (Y chain of the IL-2 receptor complex. Following the demonstration that CD25 was expressed by DN cells (Ceredig et d . , 1985), much effort was made to demonstrate a role for the corresponding ligand, naniely IL-2, in thymocyte tlevelopment. This was despite the fact that the IL-2 “receptor” on DN cells was of lower affinity arid differed biochemically from I Id-2receptors o i i activated peripliera1 T cells (Lowenthal et d., 1986).It was later sliow~~ that other coniponents of the trimolecular IL-2 receptor coinplcv were not coordinately expressed on DN cells (Falk ct d ,1993). In particular, some IL-2RPexpressing “early” thyinocytes could in fact be y6 T cells (Leclercq et d., 1995). Together with the development of IL-2 KO mice (Schorle et d., 1991),it became apparent that IL-2 wus not an absolute requirement for thyinocyte differentiation. Finally, expression of both CD4 and CD8 needs to be considered, particularly the question of whether expression of both antigens is a unique characteristic of (YPlineage cells. First, several experimental manipulations of RAG KO mice, in which no receptor gene rearrangement can take place [i.e.,activation of thymocytes with anti-CD3 antibody either in t h o ( Jacobs et al., 1994) or itz Gitro (Levelt ct nl., l993), by y-irradiation in uivo (Zuniga-Pfluckeret d . ,1994),or introduction of an activated lck transgene (Mombaerts et nl., 1994)],result in the generation of DP cells. Although these manipulations are generally considered to mimick one or more physiologic functions of the pre-TCH, it has not been formally established that all the resulting DP thymocytes are really genuine ab lineage cells. Second, although the percentage and a l d u t e nunilwr of DP cells are drastically
reduced in TCRP KO mice, a significant and highly variable proportion of DP cells is still present (typically, the thymus of TCRP-I- inice still contains about 20% of DP thymocytes, but because the total thymic cellularity is decreased at least 10-fold in these animals, a proportion of 20% corresponds to less than 2% of the absolute nuinber found in wild-type mice) (Mombaerts et al., 1992).That the proportion of such DP cells was further reduced from -20% of total thymic cellularity in TCRP-I- inice to less than 1%when the S locus w a s also inactivated (Le., in TCRP-I- X TCR&/- mice) could be interpreted to indicate that some DP cells may belong to the yS lineage (Mombaerts et al., 1992; Robey and Fowlkes, 19%) (but also see Section IV,B,2). In conclusion there is no formal proof that the DP phenotype per se is sufficient to indicate ap lineage commitment. Conversely, absence of CD4 and CD8 expression on mature T lineage cells alone cannot be taken iis a reliable phenotype for the y6 lineage, a s a large proportion of mature N K T cells are DN as well. These considerations become particularly iinportant in situations where utilization of the crP or y6 TCR as a reliable lineage marker is compromised, for instance in TCR transgenic or certain TCR knockout mouse strains (see Section IV). C. DEVELOPMENTAL STACEOF apiyS LINEAGE DIVERGEN(:E Three main observations indicate that aP and y6 lineages are derived from a common T-lineage-committed precursor: First, ap and y6 T cells have strilclngly similar phenotypes and patterns of gene rearrangement. In fact, apart from the different TCRs, no single marker has been found to date that unequivocally distinguishes both types of lymphocytes. Second, both ap and y6 T cells differentiate arid mature inside the thymus (with the exception of gut-associated IEL, which will not tie considered here) and, importantly, both lineages develop from phenotypically identical precursor subsets. Finally, and y6 cells can differentiate from "developmentally advanced" T N subpopulations (see later) that have lost the potential to give rise to B cells, N K cells, or thymic DC (dendritic cells), suggesting that the emergence of both cell types defines a final branch point in the development of T-lineage-restricted precursors. Adoptive transfer studies in vivo and repopulation experiments in fetal thymic organ cultures (FTOC) have provided the most direct approach to determine at which developmental stage aP and 76 lineages diverge. Intravenous or intrathyinic injection of Iiiglily purified CD25-CD44+c-kit +CD4'""thyinocytes considered to represent the most immature developmental stage within the thymus (see Section I,B,2) resulted in the generation of mature cells of both lineages (Shortman et nl., 1991; Wu et al., 19911. Later stages (CD25+CD44+or CD2StCD44-""" T N thymocytes)
a/3/ylyS I,I NEAGE C O M M I T M E N T
15
could also generate both ap and yS T cells when incubated in cell culture medium in the presence of IL-7 (Suda and Zlotnik, 1993) or when transferred into deoxyguanosine-treated FTOC in uitro (Godfrey et al., 1993). The generally accepted findlng that all TN subsets at least up to the CD2SfCD44-””“stage, can generate both types of T cells has suggested that the branch point for lineage divergence might reside within the CD25+CD44-””“pre-T-cell population. Intuitively, this assumption seems to be supported by the observation that the CD2rj+CD44-’I”’b subset is the first in which TCRP, y, and 6 rearrangenients can be detected to a significant extent (Dudley et ul., 1995; Godfrey et ul., 1994; Passoni et al., 1997; Petrie et ul., 1995; Tourigny et ul., 1997). Because these molecular events are a prerequisite for the generation of lineage-specific receptors, they might represent, at least in theory, a good starting point for lineage divergence. Productive rearrangement of isotype-specific T-cell receptor genes within a given population would indeed be indicative of a developmental branch point, if such rearrangenients were restricted to the corresponding lineage. However, this i y clearly not the case, as productive y and 6 rearrangements can be found in ap and functional p rearrangements in y6 lineage cells (see later). The onset of 7 , 6, and p rearrangenients predominantly at the CD2Fit TN stage as such is therefore no conclusive evidence for tlie prevnce of a branch point at this stage. A stronger argument in Favor of the CD25’ pre-T-cell stage as the likely point of divergence in c@/yS cell fates is provided by the finding that an important developniental event occurs at this particular stage, namely preTCR-mediated ‘‘0 selection” (see Section I,B,2),which has lineage-specific features because it is required for the normal development of (rp lineage cells, but is completely dispensable for the generation of y6-expressing cells (reviewed in Fehling and von Boehmer, 1997; von Boehrner and Fehling, 1997). The differential dependence of a0 and y 6 cells on /3 selection could therefore represent a developmental division of the CD25+ pre-T-cell population into an cup-committed subset, which depends on pre-TCR signaling for survival, and a y6committed subset that can mature and differentiate in the absence of pre-TCR expression and “ p selection.” However, analyses of TCRP rearrangements in yS lineage cell5 can be interpreted to indicate that many y6-expressing cells in normal mice have actually been subject to p selection (Burtrum et al., 1996; Dudley et id., 1994, 1995) (see later), suggesting that ap and 7 6 cells may diverge at a developmentally later stage. Although CD25-””’ TN cells, which represent the developmental stage immediately following “ p selection,” failed to give rise to significant numbers of 7 6 TCR’ thymocytes in a study involving FTOCs (Godfrey et al., 1993),an earlier report found that C D 2 S CD44-’ h $ T N p ecursors are indeed capable of generating ap and y6 cells, both
16
HANS JORG FEHLINC: rt cd
in vivo after intrathymic injection and in vitro in simple culture medium or medium complemented with cytokines (Petrie et al., 1992). Thus, cup and y6 cells may diverge just prior to the onset of CD4 and CD8 expression. Unfortunately, cell transfer and repopulation assays are vulnerable to at least one serious caveat: they are unable to establish the clonality of precursor-product relationships. Therefore, it cannot be excluded that lineage divergence occurs at a relatively early developmental stage and that precoinmitted cells follow separate pathways, which are phenotypically indistinguishable until after expression of the appropriate lineage marker, i.e., the respective TCR isotype. The reported generation of 76 T cells from CD25-CD44-""" TN precursors is therefore not incompatible with a branch point (stage of commitment) at the CD2S' pre-T-cell stage. It was shown some time ago that single precursor cells obtained from the fetal thymus could reconstitute lymphocyte-depleted FTOC and that the lo5progeny cells that were generated contained multiple TcRP rearrange1986).Importantly, Anderson, Jenkinson, and Owen ments (Williamset d., have shown that upon such in vitro trans'fer, a single sorted CD2St thymocyte can generate both ap and yS cells (personal communication). At present, available data do not allow a precise definition of the developmental stage at which crp and yS lineages separate irreversibly. In fact, instead of being a sudden event that can be ascribed to a certain developmental stage, aP/yG lineage commitment may be a more gradual process involving two or more developmentally successive subpopulations that become increasingly unable to change their developmental fate. The apparent difficulty in demarcating a specific stage in ap/y6 lineage commitment may reflect this situation. In order to solve the issue, some knowledge about the molecular mechanisms that control lineage commitment is clearly necessary. The following section attempts to critically illuminate what is known about these mechanisms at present. II. Models of cup/yij Lineage Commitment and lineage Maintenance
A. SEQUENTIAL REARRANGEMENT MODEL
In the fetal thymus, TCRy, 6, and P gene rearrangements occur significantly before TCRa rearrangements (Fowlkes and Pardoll, 1989).The two T-cell receptor isotypes are expressed on the cell surface in a corresponding fashion: first the y6 TCR at around day 14 and then the crp TCR at day 17/18 (Hedrick and Eidelman, 1993). The discovery of a defined temporal order of TCR gene rearrangements during fetal mouse development led to speculation very early that TCR gene rearrangements may influence or even determine the aPIy6 lineage decision. A common precursor cell may first attempt to generate functional y and 6 rearrangements, and successful
uPlyS I.INEAGE COMMITMENT
17
assembly ofa y6 heterodimer would suppress further TCR rearrangements, forcing the respective cell to differentiate along the y6 lineage. Only when y or 6 gene rearrangements o n both chroinosoines were nonproductive would the cell get the opportunity to commit to the ap lineage and attempt the forination of a fhctional TCRP and eventually TCRa chain. This concept seemed to provide a good rationale for the different timing of TCR rearrangements during fetal tliymopoiesis, which lias become known as the “sequential rearrangement model” or the “model of Allison and Pardoll” (Allison and Lanier, 1987; Pardoll et nl., 1987).
B. COMPETITI\’E REAHRANCEMENT M(>L)EI. Subsequent studies of T-cell development in the adult thymus, however, have shown that y , P. and pro1)ably also 6 rearrangements are initiated and completed essentially at the same developmental stage (Godfrey et nl., 1994; Petrie et nl., 1995), suggestiiig that there is no strict temporal order with respect to these three types of rearrangements in postembryonic thymopoiesis. It is therefore conceivable that y , 6, and P gene segments compete with each other for thc forination of a signaling-competent receptor: if, in uncommitted thymocytes, y and 6 genes were rearranged in a productive fashion first, the cells might commit to the y6 lineage; and if a functional TCRP gene was assembled first, the cell might follow the ap pathway. This concept, known as the “competitive rearrangement niodel,” predicts that lineage commitment occurs during the developmental period at which the vast majority of 2111 y , 6, and P genes rearrange, i.e., at the CD2StCD44-“””pre-T-cell stage. As TCRa gene rearrangements are delayed, occurring during both fetal and postfetal thymopoiesis predominantly when lineage commitment lias already taken place (Burtrurn et nl., 1996; Petrie et nl., 199s; Raulet ct nZ,, 1985; Snodgrass et nl., 1985;Wilson et al., 1996),they are considered irrelevant for the ap/y6 lineage decision, but may participate in lineage maintainance (see later). The “competitive rearrangement model” necessarily implies that ii precursor cell perceives signals from a y6 TCR differently than from a TCRP containing TCR because the differential expression of these TCRs can only be utilized for a lineage decision if the precursor is able to distinguish y6 expression from pre-TCR expression. LINEAGE MODEL C. SEPARATE A very different view of the commitment process is provided by the “separate lineage model,” which, in its purest form, states that the outcome of TCR gene rearrangeinelits is conipletely irrelevant for the lineage decision, and comniitment to the aP or y6 pathway is brought about by some other a s yet unknown mechanism (Winoto and Baltimore, 1989a). The
18
HANS JORG FEMLING et a[.
separate lineage model as such is less satisfying because it does not explain how a lineage decision is achieved, but just points out how it is not achieved. Moreover, it does not make any prediction about the developmental stage at which ap and y6 cells diverge, as “branching” of the a@and y6 lineages can occur before, during, or after the completion of TCR rearrangements (although the separate lineage model is often cited as if it implied that lineage commitment preceded rearrangements). Because it is a very general concept, the separate lineage model is supported experimentally by all results not in line with the more explicit rearrangement model. Direct proof for the concept of separate lineages would require the demonstration that one or more of the TN thyinocyte populations can be sorted into two discrete subsets (separate lineages) with indistinguishable TCR gene rearrangement status but exclusive developmental potential.
D. MAINTENANCE OF THE LINEACE DECISION Two molecular mechanisms have been proposed to explain how a lineage decision, once made, may be implemented and maintained. The first proposal states that cells committing to the ap lineage activate a putative TCRy-specific silencer which then prevents the expression of rearranged TCRy genes, and thus of a y6 TCR (Haas and Tonegawa, 1992). The existence of a cis-acting y silencer element downstream of the TCRy constant region gene has been deduced from experiments involving TCRy transgenic mice (see later). A second proposal states that in cells following the cup pathway, a programmed excision event is triggered that deletes the DS, 16, and C6 regions, thereby preventing permanently the formation of a functional 6 chain (de Villartay and Cohen, 1990; Hockett et al., 1988). This proposal is based on the identification of recombinational elements, termed “6rec,” which are located upstream of the TCRG locus and frequently recombine with a pseudo Ja element ($]a)located immediately 5’ to the J a region (de Villartay et al., 1988; Hockett et al., 1989; Toda et al., 1988) or, in the mouse, also with certain Ja elements (Janowski et al., 199’7; Shutter et al., 1995). Another consequence of such a recombination is the apposition of V a and Jagene clusters, with a potentially advantageous effect on the generation of ap lineage cells, as it might enhance the efficiency of subsequent Va+ J a rearrangements. Although both the “silencer mechanism” and the “deletional mechanism” were originally proposed by proponents of the separate lineage model, these concepts are also fully compatible with the rearrangement models when considered as possible mechanisms for the implementation and maintenance of the lineage decision.
19 111. Analysis of TCR Gene Rearrangements in (YOand 76 lineage Cells
A. G E N E HAL CON 5ID ERATIO N 5 The three models for the a&8 lineage split just described make different predictions with regard to the rearrangement status of the T-cell receptor loci in mature a@and 7 6 lymphocytes. According to the sequential rearrangement model, aP T cells 4iould bear signs of failed y and 6 rearrangements. Consequently, as well as 6 loci must be rearranged extensively in all a6 T cells. Moreover, the frequency of nonproductive rearrangements should be significantly higher than one would expect if the outcome of the rearrangement did not influence the lineage decision, as either y or 6 must be out of frame to allow a cell to undergo subsequent a rearrangement. y6 T cells, however, should be essentially devoid of VP+( D)Jp and Va- J a rearrangements, as they are expected to commit to the y6 lineage before TCRP and TCRa loci have become available for the recombination machinery. The competitive rearrangement model, however, does not exclude the appearance of VP+(D)J@ rearrangements in y6 cells as long as they are out of frame. It is also compatible with a lack of y or 6 rearrangements in some ap T cells because this could reflect a situation in which a functional TCRP rearrangement and subsequent aP lineage commitment has occurred before a y or 6 rearrangement could be initiated. However, the competitive rearrangement model does not allow the complete absence of y or 6 rearrangements in all (or nearly all) CUPT cells. Most important, the key assumption of the competitive rearrangement model that a functional rearrangement at a given locus would promote commitment to the corresponding lineage predicts a bias against in-frame rearrangements of that locu~in the opposite lineage. In other words, if functional y and 6 rearrangements can promote the development of y6 T cells, they should be underrepresented in aP T cells (as predicted by the AllisodPardoll model), and, vice-versa, if productive TCRP rearrangements can direct development into the a@lineage, in-frame P-rearrangements should be underrepresented in y6 cells. In the absence of any kind of selection for or against functional rearrangements, one-third of all genes assembled by V(D)J recombination are expected to be in-frame. Because the outcome of TCR rearrangements is postulated to play no role for the commitment process in the separate lineage model, it does not impose any restrictions on the status of TCR rearrangements in mature aP or 76 cells, at least in its most generalized form, except that there should be no bias against in-frame or 6 rearrangements in cells of the a@lineage or against functional P rearrangements in cells of the y6 lineage.
B. METHODSOF MOLECULAR ANALYSIS The realization that the outcome of TCR gene rearrangements might lineage decision has incited numerous studies to deterinfluence the a p / ~ G mine the status of TCR loci in matiire T cells of both lineages. In the beginning, these analyses were largely confined to individual T cell clones, lines, or hybridomas with the drawback that the statistical significance of the observed frequencies of productive versus nonproductive rearrangements was difficult to assess. Several recent methodological advances have provided the tools for a statistically more solid analysis of rearrangements at the population level. Three sorts of approaches are commonly used. The first and historically earliest approach is based on Southern blotting of genomic DNA, where D N A fragments hybridizing within defined regions of particular T cell receptor loci serve as probes to visualize potential gene rearrangements. This method suffers from two limitations. First, the sensitivity of Southern blotting is not sufficiently high to allow detection of infrequent rearrangements in mixed cell populations. Specific rearrangements in less than 5-10% of cells in a population under study are unlikely to be detected. Low sensitivity is even more of a problem when Southern blotting is used in a “reverse” fashion by measuring decreasing intensities of germline bands as indication of rearrangement. Of course these limitations do not apply when rearrangements in homogeneous cell populations, such as hybridomas or T cell clones, are studied. A second drawback of the Southern blotting technique is its critical dependence on a good knowledge of both the genomic position of the respective probe(s) and the relative order of the rearranging gene fragments in the genome. Insufficient sequence information and the lack of detailed restriction maps have been serious constraints in the past, especially when a comprehensive and detailed analysis of gene rearrangements was required. In the era of genome research and large-scale sequencing, it is obviously just a matter of time before this limitation is completely overcome. The availability of more refined molecular maps and the rapid accumulation of exact sequence information of TCR loci in recent years have already led to a renaissance of Southern blotting for the analysis of TCR gene rearrangements. A dstinct advantage of this method is that it allows the quantitative measurement of multiple gene loci without the possibility of introducing any undue bias, which is always a risk when using amplification-based technologies, such as polymerase chain reaction. All other current approaches to study TCR gene rearrangements rely on PCR. One technique, termed polyrnerase chain reaction-restriction fragment length polymorphism (PCR-RFLP), provides information as to
aIJIyS LINEAGE C:OMMITMENT
21
the ratio of coding to noncoding sequences within a particular subpopulation (Mallick et nl., 1993).Starting with DNA purified from large numbers ( -lo4) of sorted cells, amplification of the rearranged fragment is carried out and the labeled, restriction enzyme-trimnied PCR product is elecrophoresed down a sequencing gel alongside fragments obtained from a control sequencing reaction. Thiis, PCR-amplified products from rearrangements coding for protein will be seen as major hands every three nucleotides. The ratio of coding to noncoding sequences can then be determined by densitometry. Importantly, without additional sequencing, the PCR-RFLP approach does not provide inforination as to whether coding-sized fragments will actually code for a protein nor does it determine what fraction of cells in the starting population contain rearranged fragments. A second, inore quantitative technique determines what fraction of a particular subpopulation of cells contains rearrangements (Mertsching and Ceredig, 1996). This is done by carrying out two parallel PCR reactions on each of two DNA preparations from cells with differing degrees of rearrangements. The control PCR reaction amplifies a DNA fragment within a single exon and i5 used to equilibrate the amount of PCRamplifiable DNA in the two samples. The second PCR reaction amplifies only rearranged fragments. The products of both PCR reactions are transferred to filters and probed with specific internal oligonucleotide probes. By comparing the densitornetiy curves from both reactions, an estimate as to dlfferences in the degree of rearrangement between two subpopulations can be made. Although sensitive enough to detect rearrangements in single cells (Mertsching and Ceredig, 1996),in pooled cells, differences of less than three-fold cannot be measured reliably by this technique. Again, sequencing is required to obtain the qualitative nature of the rearrangements. To gain further insight into the TcR rearrangement status of developing thyinocytes, it would be desirable to apply PCR on the single cell level, as has been done very successfully for B cells in both mice (Loffert et nl., 1996; ten Boekel et nl., 1995, 1997) and humans (Ghia et d.,1996), including subsequent sequence analysis of specific PCR fragments (ten Boekel et d.,1997). However, compared with single cell analysis of lg gene rearrangements in phenotypically defined subsets of developing B cells (Ehlich et aZ., 1994; Osinond et al., 1998),the structure of native TcR loci poses niiinerous additional problems, which make a coinprehensive PCR analysis in single T lineage cells virtually impossible. The mouse 7 locus contains multiple gene clnsters in different orientations (Raulet, 1989).The 6 locus is contained within a (Chien et al., 1987).In particular, VS genes are scattered across the VCYlocus, and although many ap T cells
22
HANS J()HG FEHLING et a!
contain two a rearrangements (Malissen et al., 1992), clearly some have only one and contain detectable 6 rearrangements on the other allele or on extrachromosomal circles (Nakajima et al., 1995). Little is known about allelic exclusion of the mouse 6 locus. The p locus has two DJCp clusters, and sampling of Vp rearrangements to one Jp locus means that no information on the other will be obtained. Some of this p locus complexity can be bypassed by using natural mouse mutants, such as the New Zealand Black (NZB) strain, in which one DJCp cluster is absent (Kotzin et al., 1985). Alternatively, TCR mini loci can be introduced as transgenes in the mouse genome, which then serve as simplified rearrangement substrates (Asarnow et al., 1993; van Meenvijk et al., 1990; Capone et d.,1993a, 1995; Lauzurica and Krangel, 1994; Kang et al., 1995). Targeting such TCR mini loci into their corresponding gene loci, so-called “knock-in” experiments, would refine this experimental approach and may eventually allow a more universal application of the single cell PCR approach, also to T lineage cells. C. RESULTSOF TCR GENEREARRANGEMENT ANALYSES 1. TCRyLocus
Analysis of a large number of T cell clones, T-cell lines, and T hybridomas (reviewed in Raulet et al., 1991) as well as total thymocytes and peripheral T cells (Garman et al., 1986;Kranz et al., 1985)has demonstrated unequivocally that Vy-.Jy gene rearrangements are very abundant in cells of the afl lineage. The fact that TCRy rearrangements are common in ap lineage cells suggests that either common progenitors commit only after the occurrence of Vy+ Jy rearrangements or commitment to the afl lineage does not notably suppress such rearrangements. However, Southern hybridization analysis of DNA from tissues enriched for a/? T cells also revealed that ap T cells can carry TCRy loci in germline configuration (Garman et al., 1986; Kranz et al., 1985). Moreover, T-cell clones with at least one TCRy locus in germline configuration have been found frequently (Heilig and Tonegawa, 1987; Moisan et al., 1989; Reilly et al., 1986). Thus, a/3 T-cell development can occur in the absence of saturating y rearrangements. The recurrent detection of unrearranged TCRy alleles in ap T cells does not support the successive rearrangement model, which advocates exhaustive TCRy rearrangements before commitment to the ap lineage. Whether or not TCRy rearrangements are biased against productive rearrangements in ap lineage cells has been more difficult to address. Sequence analysis of y rearrangements cloned from individual a0 T-cell clones did not allow a firm conclusion. In fact, data suggested that there might be differences with regard to different TCR loci. For instance, the
ap/yG LINEAGE COhlMITMENT
23
frequency of in-frame Vy1.2-Jy2 rearrangements in crp T cells seemed to be close to one-third, as expected for unselected rearrangements (Heilig and Tonegawa, 1987; Kranz et al., 1985; Traunecker et al., 1986),whereas productive Vy2+ Jyl rearrangements were rarely found (Heilig and Tonegawa, 1987; Traunecker et al., 1986). A major point of concern in this type of experimental approach is the statistically limited number of sequences analyzed and the uncertainty whether the particular crp T-cell clones or lines studied are representative for crp lineage cells in general. This problem has been overcome in an elegant experimental approach involving the generation of mice that carried a transgenic TCRy minilocus (Kang et al., 1995). The minilocus encompassed Vy2, Vy3, Vy4, Jyl, and Cyl gene segments in their natural, germline configuration. However, the transgenic Vy genes were engineered to contain a frame-shift mutation that would stop translation prematurely upstream of the V-J junction. These mo&fications were introduced to ascertain that cellular selection could not affect the frequency of rearrangements involving a certain reading frame. The transgenic minilocus therefore provided a reliable internal control for assessing the influence of selection on the frequency of productive TCRy rearrangements in ap T cells. PCR amplification and analysis of almost 100 independent V y 2 4Jrlrearrangements derived from endogeneous TCRy loci and, at the same time, from the transgenic minilocus revealed a significant bias against productive Vy2+ J y l rearrangements in c.0 T cells (Kang et al., 1995):in the absence of selective pressure, approximately 18% of rearrangements (transgenic minilocus-derived sequences) were productive in contrast to only about 4% of the endogeneous Vy2Jyl joints (Kang et al., 1995). Assuming a rearrangement mechanism that does not Favor a particular reading frame and absence of any kmd of selection for or against productive rearrangements, one would expect 33% of VyJr joints to be productive. Efficient selection against cells that can produce a yS TCR (bearing a functionally rearranged TCRy and TCRS allele) should reduce the proportion of productive rearrangements to slightly less than 20% (for calculations, see Dudley et al., 199s). The arithmetically unexpected low frequency of productive rearrangements at both transgenic and endogeneous loci was shown to be due to an in-frame stop codon near the 3’ end of the Vy2 gene segment. When this naturally occurring stop codon was ignored, 32% of transgene rearrangements were in-frame, corresponding to the value (33%) anticipated for random rearrangements. Endogeneous in-frame rearrangements accounted for just 18% of the V y 2 4 Jyl rearrangements, again indicating a significant underrepresentation ( 18%versus 32%) (Kang et al., 1995). Suppression of functional TCRy rearrangements has also been reported in a study involving the PCR-RFLP technique (Dudley et al.,
24
HANS J8RC FEHLINC et al.
1995). Amplification of Vyl. 1+ Jy4 rearrangements from double positive thymocytes and peripheral ap T cells revealed a much lower frequency of in-frame rearrangements (19 and 1896, respectively) than one would have expected from a rearrangement process not subjected to yS counterselection. PCR-RFLP data are convincing, provided the frequency of unselected Vyl.l+Jy4 rearrangements is really 33%.This could not be tested directly because an internal control (e.g., a transgenic Vyl.llCy4 minilocus with an engineered stop codon in the Vyl.1 segment or a mouse with a constant region knockout allele) was not available when the experiments were done. Taken together, current data strongly indicate that the outcome of TCRy rearrangements influences the ap/yS lineage decision in that functional TCRy rearrangements seem to disfavor development along the ap lineage. Combined with the observation that not all TCRy loci or alleles are rearranged in ap T cells, the findings are most compatible with the competitive rearrangement model. 2. TCRSLocus Analysis of TCRS rearrangements in ap T cells has been hampered severely by the fact that TCRS gene segments are located within the TCRa locus and are excised as a circular piece of DNA upon Va+Ja rearrangement (Chien et al., 1987b; Koop et al., 1992) because ap T cells usually rearrange both TCRa aIleles (reviewed by Malissen et al., 1992). TCRS sequences are deleted on both chromosomes and are no longer found in cell lines or hybridomas of the ap lineage. However, the byproducts of a deletional Va-Ja recombination event, circular DNA molecules containing the deleted DNA, are retained to a large degree in thymocytes and newly generated peripheral T cells that have not yet undergone extensive proliferation. These circles have been purified and used to construct “circle DNA libraries,” which were then analyzed for the presence or absence of 6 rearrangements (Takeshita et al., 1989; Winoto and Baltimore, 1989a). Hybridization of a large number of circle DNA-containing h clones with probes flanking the DS2 gene segment and the JSl gene segment, which is by far the most frequently used J element in yS expressing cells, revealed that the overwhelming majority of JSl containing circles hybridized to all probes, implying that DS2 and JSl gene segments were predominantly unrearranged (Winoto and Baltimore, 1989a). This finding seemed to put an end to the “rearrangement models” of lineage commitment. However, a subsequent study using a similar technique reported a different result: the S locus of cloned circle DNA was found to be frequently rearranged involving mainly the JS2 element (Takeshita et al., 1989). This second study was later criticized (Winoto, 1991) for having used a vector
that might have selectively excluded the larger-sized JSl germline fragments and thus introduced a bias for 562 rearrangements. The latter might have been generated by secondary rearrangements occurring after the excision of the TCRG locus, as J62 rearrangements are normally rather rare. The status of TCRG rearrangements in ap lineage cells therefore remained a moot point. The problem was again addressed several years later by Livak and colleagues (1995) who observed that excised TCRG sequences were inaintained in adult mice in both total thymocytes and peripheral T cells to such an extent that they could be studied directly by Southern blot analysis. Potential artifacts associated with the purification and cloning of excised circle DNA could thus be avoided. This observation, together with the fact that the sequence of the TCRdS locus and reliable phosphoimaging methods had become available, allowed a detailed and quantitative Southe m blot analysis of the status of the TCRG locus in DNA extracted from thymocytes or lymph node T cells without further manipulation. The analysis revealed that sequences between D61, D62, and JSl were extensively deleted, whereas sequences between J6l and J62 were largely retained, indicating numerous V( D)J6 rearrangements involving predominantly the JSl gene segment (Livak et nl., 1995), in good agreement with the pattern known from chromosomal TCRG rearrangements in y6 cells (Chien et al., 1987a; Elliott et al., 1988).These results were confirmed with additional probes that directly visualized specific rearrangements. Quantification of band intensities suggested that cup T cells contain at minimum 40% of the retained 6 sequences in a V( D)J rearranged configuration. Importantly, sequence analysis of about 100 independent 6 rearrangements, obtained by PCR amplification of DNA from total thymus and peripheral T cells with VWJ61- and L765/J6l-specificprimers, revealed that only about 20% of these rearrangements were in-frame, significantly less than the 1: 3 ratio expected for random rearrangements, indicating that functional TCRS rearrangements were depleted in cells ofthe ap lineage (Livak et al., 1995). These findings were confirmed and extended in an independent study. Using the PCR-RFLP technique, Dudley et al. (1995) fourid extensive V(D)JS rearrangements in CD25' pre-T cells, CD4'8' thymocytes, and peripheral ap T cells. Densitometric analysis of the PCR-RFLP banding pattern indicated a significant depletion of productive rearrangements in cells of the ap lineage, as only approximately 24 and 19% of the 6 rearrangements detected in DP thymocytes and peripheral T cells, respectively, were in-frame. Interestingly, this depletion was also apparent in the CD25+ pre-T population, suggesting that a set of cells with productive 6 rearrangements had been removed to become 76 T cells already at this early developmental stage. To exclude the possibility that functional TCRS
26
HANS JOHG F E H L N G ct nl
rearrangements are inherently underrepresented in ap lineage cells due to some unknown bias in the recombination mechanism, the analysis was repeated with thymocytes from mutant mice that lack the TCRG constant region gene. Although cells from these mice can undergo (V(D)JG rearrangements, they cannot generate a TCRG chain and thus are not subject to TCRG-mediated selection (Itohara et al., 1993). In CG-’- thymocytes, in-frame TCRG rearrangements were not depleted, constituting about 32%, as expected for random rearrangements (Dudley et al., 1995). A partly different result was reported in a third study. To obtain some information on the status of the TCRdG locus at distinct developmental stages in a normal adult thymus, Wilson et al. (1996) examined the nature and extent of TCRa and TCRG transcripts in developing thymocytes by Northern blotting and sequence analysis of RT-PCR-amplified transcripts. During development, TCRa-specific transcripts were found for the first time at the ISP stage. These transcripts were not full-length, but sterile C a message originating from the so-called TEA (transcription early alpha) region located 3’ of CG and 5’ of the most upstream J a element. TEA transcripts are thought to indicate the opening of the TCRa locus for subsequent rearrangement and thus to reflect commitment to the ap lineage. The detection of TEA transcripts in ISPs is in line with the view that these cells represent a transitional population between DN and DP thymocytes already committed to the ap lineage. Interestingly, ISPs were shown to express full-length TCRG transcripts at high levels, implying that chromosomal V( D)JG rearrangements occur frequently in a@ lineagecommitted cells, well before they are excised by Va- J a rearrangements (Wilson et al., 1996). This result demonstrated that at least a significant fraction of the observed 6 rearrangements in the extrachromosomal circle DNA of ap lineage cells had taken place before excision and thus before lineage commitment, which was not clear from previous studies. However, the study by Wilson et al. (1996) did not confirm a bias against in-frame V(D)JG rearrangements in ap lineage cells: sequencing of 22 PCRamplified V(D)JCG transcripts from ISP-derived hybridomas and of an additional 77 V(D)JCG PCR products amplified from cDNA of freshly sorted ISP thymocytes revealed an overall frequency of productive rearrangements of 29%, which is close to the theoretical value expected for random V(D)JGjoining. A potential problem in this study, however, is the fact that the sequences were obtained by analyzing RNA and not genomic DNA as in other studies. The result could therefore be biased in favor of in-frame rearrangements due to the preferential stability of functional message. In fact, drastically lower steady-state levels of mRNA from Ig and TCR genes harboring premature stop codons are well documented (reviewed by Li and Wilkinson, 1998).
uplyG LINEAGE COMMITMENT
27
Another convincingdemonstration of V( D)J rearrangements at the TCRG locus in thymocytes of the ap lineage was provided by Nakajima and colleagues ( 1995). Southern analysis of genomic DNA from thymocyte subsets of adult and newborn mice, which had been fractionated corresponding to descrete developmental stages, revealed that essentially all C6 genes in these cells were associated with D h J G or V(D)JG rearrangements. Such 6 rearrangements were found irrespective of the developmental stage, including those ap-committed thymocytes in which Va+ Ja rearrangements were not yet completed. Direct evidence for the occurrence of V( D)JG rearrangements in c.p lineage cells before the excision of the TCRG locus was provided by two-dimensional gel electrophoresis of large DNA fragments. Approximately 20% of the rearranged TCRG genes resided on chromosomal DNA in CD3-""" DP thymocytes, which was not the case in more mature subsets. The finding that V(D)JGrearrangements occur in ap-committed thymocytes to a significant extent before 6 loci are excised by Va+ J a rearrangements was confirmed in thymocytes of TCRaP transgenic mice, in which endogeneous a rearrangements are suppressed. In these mice, approximately 50% of rearrangements involving the JGl element were retained on the chromosome, even in mature subsets (Nakajima et al., 1995). The latter finding also demonstrates adventitiously that deletion of the TCRG locus is not a prerequisite for the generation of ap lineage cells. A similar conclusion had been reached in an earlier study, which showed that cytotoxic T-cell lines from ap transgenic mice had committed to the ap lineage without deleting the TCRG locus (Ohashi et al., 1990). Retention of a rearranged TCRG gene on one chromosome in the presence of an a rearrangement on the second has also been found in some T hybridomas generated from thymocytes of nontransgenic mice (Thompson et al., 1990, 1991). Although the presence of chromosomal or extrachromosomal V( D)JG rearrangements in cells of the ap lineage per se is compatible with the rearrangement models as well as the separate lineage model, the findings clearly show that there is no lineage-specific control of rearrangement at the TCRG locus. Two moleculdr mechanisms for lineage determination can therefore be excluded, namely recombinational silencing of the 6 locus in ap cells (Diaz et al., 1994; Winoto and Baltimore, 1989a) and obligatory excision of the 6 locus prior to Va+Ja rearrangement by site-specific recombination involving Grec and +JS recoinbinational elements (de Villartay and Cohen, 1990; Hockett et al., 1988). Experiments indicating a selective depletion of in-frame 6 rearrangements in ap thymocytes (Dudley et al., 1995; Livak et al., 1995) seem to provide strong support for a rearrangement model of lineage commitment.
28
HANS JORG F E H I J N G ct nl.
3. TCRPLOCUS The sequential rearrangement model predicts the absence of VP-( D)JP rearrangements in 76 lineage cells. In contrast, such rearrangements should be frequent, but generally out of frame, according to the competitive rearrangement model. Initial studies involving a limited number of 78 Tcell clones and hybridomas suggested that DO+ JP rearrangements are very common, whereas complete VP+( D)JP rearrangements are rather rare in cells of the y 8 lineage (see reviews by Raulet et al., 1991; Haas and Tonegawa, 1992). More recently, it has become clear that these earlier findings cannot be generalized. With the goal of studying TCRP rearrangements in a representative sample of 78-expressing cells, Dudley and colleagues (1994) analyzed 76 T-cell populations that were sorted from spleen and lymph nodes of TCRadeficient mice. In TCR& mice, y6e.upressing lymphocytes are the predominant population of CD3+ cells, which facilitates their purification and lowers the risk of undue contamination with T cells of the aP lineage. Quantitative Southern blotting with appropriate probes revealed that at least 90% of TCRP loci in sorted y6 cells had DP+JP rearrangements and, most important, at least 50% had VP+(D)JP rearrangements. Interestingly, analysis of the pattern of V( D)JP rearrangements in such y8 T cells with the PCR-RFLP technique and primers specific for VP134P2.2 and VP4iJP2.2 indicated that approximately 68 and 73%, respectively,were in-frame. This result was supported by sequence analysis of 20 Vp13+(D)JP2.2 joints derived from sorted splenic y6 T cells, revealing 70% productive rearrangements (Dudley et al., 1994). Notably, this value is as high as in aP T cells, which are known to be subject to P selection (see Section IV,C,l). The detection of functional VP+( D)JP rearrangements and their apparent overrepresentation in y6 T cells relative to the value expected for random rearrangements has serious implications for all models of aPly8 lineage commitment. It was therefore important to show that this finding was not limited to TCRa-deficient mice. In a subsequent study (Dudley et al., 1995), the analysis was extended to y6 cells sorted from spleen and lymph nodes of normal mice. The PCR-RFLP assay with VP13/JP2.2specific primers again revealed a frequency of approximately 70% in-frame rearrangements. VP-JP rearrangements were also detected in y6 IEL of the gut, which are thought to be partly of thymic origin, using VP4/JP2.5and VfllGIJP2.2-specific primers, and again most of these rearrangements were in-frame. A different pattern was observed in TCRP rearrangements from dendritic epidermal T cells (DECs), which are derived from fetal thymic
afllyS LINEAGE COMMITMENT
29
precursors and regarded as the earliest subset of the y6 lineage (Heyborne et al., 1993; Ikuta et al., 1990). Although rearrangements involving VP4/ JP2.2 and VP5lJP2.6 elements could be clearly detected, the PCR-RFLP patterns were irregular, suggesting fewer or less diverse rearrangements (Dudley et al., 1995). Moreover, in-frame joints were in the minority. The findings regarding DECs were complemented by an independent study focusing on VP6lJP2.5 and VPWJP2.5 rearrangements in day 15 fetal y6 thymocytes (Mertsching and Ceredig, 1996). Semiquantitative PCR analysis revealed a similar degree of TCRP rearrangements in fetal day 15 yS cells as in adult CD8+ ISPs or CD25-CD44-""" DN thymocytes. However, in fetal thymocytes, such rearrangements were mostly (58%)out of frame, as determined by sequencing of 104 PCR-amplified V( D)JP joints. In this context, it is interesting to note that fetal thymocytes provide the precursor cells for dendritic epidermal cells (Allison, 1993; Boismenu and Havran, 1995). The finding of approximately 70% in-frame VP+( D)JP rearrangements in splenic and lymph node-derived 76 T cells (Dudley et al., 1995) suggested that these cells had been selected for functional TCRP chains, possibly in a pre-TCR-mediated process during thymopoiesis. However, because peripheral y6 cells are many steps past primary differentiation events and are likely to have been exposed to a variety of selective events, such as homing or clonal expansion, it is unclear whether these cells reflect events in early differentiation correctly. To avoid this potential pitfall, Burtrum and colleagues (1996) studied TCRP rearrangements in purified 78-expressing cells from the thymus of normal adult mice. Quantitative Southern blot analysis revealed that both DJP clusters were extensively rearranged at levels essentially indistinguishable from those in mature (YP T cells. VP+( D)JP rearrangements were less frequent, but nonetheless substantial, occurring in approximately 15-20% of all alleles, versus approximately 75% of mature (YPT cells. Analysis of the reading frame of complete V( D)JP rearrangements by PCR-RFLP with VPUJP2.6- and VPYJP2.6specific primers revealed that approximately 55 and $5l%, respectively, were productive (Burtrum et al., 1996). Sequencing of 16 independent VP4lJP2.6 joints derived by PCR amplification from genomic DNA of 7 6 thymocytes confirmed the PCR-RFLP data, as 9 (56%) of the rearrangements were in-frame. The results therefore suggest that functional TCRP rearrangements are already overrepresented at an early (thymic) stage in y6 T-cell development, although the predominance of in-frame TCRP rearrangements seems to be less pronounced than in peripheral 76 T cells (50-55% versus 70%). An even lower percentage of productive TCRP rearrangements in 76 thymocytes was found in an independent study. Mertsching and colleagues
30
HANb JORG FEHLING et a/
(1997) analyzed PCR-amplified VP6-(D)JP2.5 rearrangements from CD24’90’ y6 thymocytes, which represent the most numerous subpopulation of thymic y6 cells and supposedly the only subset emigrating from the thymus. Sequencing of 43 distinct VP&(D)JP2.5 joints showed that only 42% were in-frame, which was the same frequency as had been found previously in fetal y6 cells (Mertsching and Ceredig, 1996). Although this value is still 9% higher than the 33% that would be predicted by random joining, it no longer supports the view that ‘‘P selection” is a very common event in y6 T-cell development, as suggested by the previous reports. The reason for the discrepancy in the percentage of productive V( D)JP joints in this study (42%) and the one of Burtrum et al. (50-55%) is unclear at present, but may be related to the analysis of different VP+JP elements or the fact that one analysis (Mertschinget al., 1997) focused on a subpopulation of y6 thymocytes whereas the other (Burtrum et al., 1996) dealt with unfractionated y6 thymocytes. The examination of TCRP rearrangements in 76-expressing cells has provided a number of important findings with relevance for the mechanism of the crPly6 lineage split. The fact that VP+(D)JP rearrangements are common in y6 lineage cells effectively excludes the simple successive rearrangement model, which postulates that TCRP rearrangements are initiated only in those cells that fail to generate a functional yS TCR. A modified version of the successive rearrangement model may still be valid, if it incorporates two assumptions. First, expression of a y6 TCR relegates precursor cells irreversibly into the y6 lineage, but this does not prevent subsequent TCRP rearrangements. Second, P rearrangements that occur after the formation of a functional y6 TCR come too late to influence the lineage decision. The second assumption implies that either the lineagedetermining signal provided by the yS TCR cannot be overridden by signals from a pre-TCR that forms subsequently or, alternatively, a signalingcompetent pre-TCR can no longer form in cells that received a signal by a y6 TCR, e.g., because pTa expression has been switched off concomitantly. Data on TCRP rearrangements in 78-expressing cells also pose a serious threat to the simple version of the competitive rearrangement model. As outlined in detail earlier, this model presumes a competition between y and 6 rearrangements on the one hand and P rearrangements on the other, with the lineage fate being determined by the type of receptor (a y6 TCR or a TCRP-containing pre-TCR) that is generated first. The fact that productive TCRP rearrangements are found frequently in y6 cells excludes the possibility that thymic y6 cells might be derived from a salvage pathway for thymocytes that have failed to produce a TCRP chain. Moreover, it strongly suggests that the formation of a pre-TCR per se does not preclude differentiation along the y6 lineage. This incongruency with the competi-
aO/y6 LINEAGE COMMITMENT
31
tive rearrangement model can be circumnavigated again by evoking tlie same two assumptions made earlier to rescue tlie successive rearrangement model. However, with these two “amendments,” both models become virtually identical. An alternative explanation for the occurrence of productive V( D)JP rearrangements in cells of the y6 lineage is suggested by studies of early B-cell development. The ability of IgH proteins to pair with surrogate light chain plays a critical role in controlling membrane expression of the pre-BCK and hence subsequent differentiation (ten Boekel et nl., 1997, 1998).Keyna and colleagues (1995)identified two p heavy chains expressed in prwursor B cells that were unable to form a pre-BCK, and an analysis of the earliest B-cell precursors in bone marrow revealed that about half of tlie productive V( D)JH rearrangements in these cells encoded p heavy chains unable to pair with surrogate light chains (ten Boekel et al., 1997). In analogy, it remains possible that the in-frame V( D)JP rearrangements observed in y6 cells encode TCRP chains that are unable to associate with pTa, an hypothesis that has not yet been tested experimentally. Failure to form a pre-TCK would provide an explanation why some TCKPexpressing cells are not drawn into the aP differentiation pathway. However, the apparent overrepresentation of productive TCKP rearrangements in y6 cells seems to argue against sucli an interpretation, unless this overrepresentation is brought about by a pre-TCK-independent mechanism (see later). An overrepresentation of in-frame TCKP rearrangements in cells of the y6 lineage has been found repeatedly (Burtnim et al., 1996; Dudley et nl., 1994, 1995), although the extent of this overrepresentation remains a 1997).The high matter of debate (Burtrum et nl., 1996; Mertsching et d., frequency of in-frame V( D)JP rearrangements in y6 cells has come as a surprise because the biologic pressure for this selection is difficult to envisage. From gene knockout experiments it is clear that a functional TCKP chain is not required for normal y6 development (see later). Although the possibility that certain in-frame TCRP gene rearrangements confer a survival advantage on y6 T cells has not yet been excluded, another explanation is currently more popular. It is assumed that tlie accumulation of productive TCKP rearrangements in 76 cells is most likely due to a coincidental expansion of thymic yS precursors by tlie same mechanism that is responsible for the expansion of aP precursors after in-frame TCRPrearrangement. In other words, a varying but substantial number of 76 cells is thought to have been subjected to pre-TCK-mediated P selection (Burtruin et nl., 1996; Mertsching et d., 1997).If correct, this interpretation has two important implications. First, expression of a pre-TCR alone cannot be tlie decisive event for a commitment to the ap lineage, and second,
32
HANS JORC FEIlLlNC et a[
the split between the aP and 76 lineages must occur after the stage at which /3 selection is operating, i.e., at the CD25-44-""" DN stage. Alternatively, y6 cells may be able to undergo P selection well after lineage commitment has taken place. Although pre-TCR-mediated selection certainly provides an attractive and popular explanation for the buildup of in-frame rearrangements in yG cells, alternative scenarios should not yet be neglected. A functional TCRP chain may, for instance, be able to act in a pre-TCR-independent fashion. Experiments have shown that transgenic TCRP chains can signal in the absence of pTa and other T-cell receptor chains, mediating a number of effects, e.g., the generation of small numbers of DP thymocytes in RAG-'- X pTa? mice (Krotkova et al., 1997).Whether the overrepresentation of productive TCRP rearrangements in cells of the yG lineage is indeed due to pre-TCR-mediated selection or some other mechanism can be tested by analyzing rearrangements in y6 cells of pTa-deficient mice (Fehling et al., 1995a), which lack a functional pre-TCR, but possess normal TCRP alleles.
4 . TCRaLocus The presence of Va+ Jarearrangements is widely accepted as a hallmark of cells committed to the aP lineage. This view is based on the following observations. First, Va+ J a rearrangements occur late in thymopoiesis relative to TCRy, 8, and P rearrangements. For instance, significant transcriptional activity of the TCRa locus, which is thought to indicate the opening of the TCRa locus for subsequent rearrangements, can be detected by Northern blotting only at the ISP stage (Wilson et al., 1996). Moreover, the TCRa-specific mRNA in ISP thymocytes consists predominantly of sterile C a transcripts, indicating that the majority of TCRa loci are still in germline eonfiguration at this developmental stage (Wilson et al., 1996). Northern blotting analysis suggests that significant amounts of full-length (VJCa) transcripts are generated only at the DP stage (Wilson et al., 1996). These findings are in good agreement with an earlier study based on quantitative Southern blotting, which demonstrated that substantial TCRa rearrangements do not occur before the CD3-"" DP stage (Petrie et al., 1995).Because ISP and DP thymocytes are generally regarded as cells that belong to the a0 lineage, Va+Ja rearrangements seem to occur predominantly after lineage commitment has taken place and are therefore expected only in aP lineage cells. Second, the particular chromosomal organization of the TCRdG genes results inevitably in the deletion of the TCRG locus upon Va+Ja recombination (Chien et al., 198%; Koop et al., 1992).Continued expression of a yGTCR, and hence the functionality of yG T cells, therefore depends critically on efficient prevention of Va+ J a
aplyG LINEACE COhlMlTMENT
33
rearrangements. Third, transfection studies in ap and y6 T-cell lines have shown that the TCRa enhancer, located 3’ of the C a gene, contains a silencer sequence that prevents enhancer-mediated transcription in y6 but not in ap lineage cells (Winoto and Baltimore, 198913).These findings are supported by experiments with transgenic mice showing that y6 thymocytes selectively fail to express rearranged TCRa transgenes driven by an autologous locus control region (LCR), including the a enhancer and silencers (Diaz et nl., 1994). Other experiments with transgenic mice have demonstrated that the TCRa enhancer also restricts TCR rearrangements to cells of the ap lineage, indicating that the TCRa locus is most likely not only transcriptionally, but also recombinationally silent in y6 lineage cells (Cupone et d . , 1993; Lauz~iricaand Krangel, 1994). Taken together, these considerations suggest that cells of the y6 lineage do not rearrange the TCRa lociis and usually retain both alleles of the 6 locus. This is exactly what has been found in most of the y6 cell lines that have been analyzed (reviewed in Raulet ~t nl., 1991). These findings are supported by an analysis of total y6 thymocytes: quantitative Southern blotting with Ja-specific probes hybridizing to selected intronic regions spaced equilstantly across the Ja gene cluster did not reveal any TCRa rearrangements in the DNA of pools of sorted y6 thymocytes (Burtruni et nl., 1996). Although this type of analysis is not sensitive enough to exclude a low level of Va-Ja rearrangements, it demonstrates that such rearrangements are certainly not common in thymic yS lineage cells. In contrast, the vast majority of ab-expressing cells have rearranged both TCRa loci (Casanova et al., 1991; Malissen et d., 1992). The paucity of Va+ Ja rearrangements in y Sexpressing lymphocytes can in fact provide a useful molecular marker for the identification of y6 lineage cells in situations where a determination of the relevant TCR isotype is either not possible or not informative, e.g., in mice expressing TCR transgenes (Bruno et al., 1996; see later). However, not all y6 cells seem to be completely devoid of Va-Ja rearrangements. Using a semiquantitative PCR assay and primers specific for a single J a element and three distinct V a families, Mertsching et nl. ( 1997) clearly detected Va- J a rearrangements in sorted 76-expressing thyinocytes of adult but not fetal mice. That y6 cells from T C R P P mice contained a similar level of rearrangements suggested that a rearrangement in y6 cells could be generated without expression of a conventional preTCR composed of pTa and TCRP proteins (Mertsching et nl., 1997). Quantitation of the respective PCR bands and comparison with the corresponding bands obtained from SP thymocytes of the c.up lineage suggested that approximately 7% of thymic y6 cells harbored a Va-Ja rearrangement. Further analysis of these y6 thymocytes by RT-PCR revealed a
34
MANS JORC FEIILING el nl
roughly equivalent amount of VJCa transcripts, indicating that all the rearranged TCRa genes were most likely expressed. These data do not necessarily contradict previous findings. Although the percentage of Va+Ja rearrangements reported (7%) is clearly too high to be accounted for by contaminating ap lineage cells, the percentage is sufficiently low to escape detection in Southern blotting analyses or when studying TCRa rearrangements in a limited number of 76 T cell clones or hybridomas. However, why should y6 cells allow TCRa rearrangements when these rearrangements threaten their very existence? The answer might be that such rearrangements occur before lineage commitment is established. In fact, data from the analysis of pTa-’- X TCRa-I- and pTa? X T C R P doubly deficient mice provide clear, albeit indirect, evidence that Va+Ja rearrangements do occur in a few CD25+CD44-“”” pre-T cells (Buer et al., 1997; Mertsching et al., 1997). Some of these cells may eventually enter the y6 pathway. Whatever the reason for low-level TCRa rearrangements in y6 thymocytes may be, the vast majority of y6 lineage cells (>go%) is clearly devoid of any Va+Ja rearrangements and the absence of these rearrangements in mature T cells can be seen as a distinctive feature of the y6 lineage.
5. Concluding Evaluation of T C R Rearrangement Data with Regard to the DifSerent Mode1.s of Lineage Commitment The analysis of TCR gene rearrangements has proven that Vy+Jy, V&( D)J6and VP+( D)JP rearrangements are not lineage-specificevents. Because many aP lineage cells have one or more TCRy loci in germline configuration, it is also clear that a0 T-cell development can proceed in the absence of exhaustive TCRy rearrangements. Moreover, the analysis of the adult thymus did not provide any evidence for a specific temporal order of TCRyIG versus TCRP rearrangements. Taken together, these findings are incompatible with the sequential rearrangement model, at least in its historical, simplest form. This model will therefore not be considered any further. The finding that productive Vy-Jy and V&( D)JS rearrangements are underrepresented in cells of the aP lineage seems to provide strong support for the competitive rearrangement model because the most straightforward interpretation would suggest that progenitors with functional y6 rearrangements are diverted into the y6 lineage. Ilowever, underrepresentation of in-frame y and 6 rearrangements in ap lineage cells does not prove that successful TCRyIS rearrangements guarantee y6 T-cell development. The observed paucity of functional yI6 joins in a0 lineage cells could also be a result of selective cell death or deficits in proliferation, affecting those precursors that express the “wrong” TCR isotype (see later). Moreover,
aply8 LINEAGE COMMITMENT
35
the frequent occurrence of in-frame TCRP rearrangements in 76expressing cells can only be integrated in a competitive rearrangement model by invoking some as yet unproven, ad hoc assumptions. In summary, data obtained from the analysis of TCR gene rearrangements do not allow a distinction between the competitive rearrangement and the separate lineage model; rather, they are compatible with both models, provided some modifications are introduced. That yl8 rearrangements in CUPlineage cells are disproportionately out of frame is certainly a key finding because it demonstrates that these rearrangements are not neutral events but do influence cell fate, a conclusion that must be incorporated in any viable model of aPly8 lineage commitment. IV. Analysis of TCR Transgenic and Gene-Targeted Mice
A. TCR TRANSGENIC MICE In theory, TCR transgenic mice should be a perfect tool for studying the molecular mechanism of the aPIy8 lineage split, as the separate lineage and the rearrangement models make distinct predictions regarding the influence of TCR transgenes on lineage commitment. According to the rearrangement model, introduction of functionally rearranged receptor genes encoding a specific receptor isotype should direct all developing thymocytes into the corresponding lineage and prevent the formation of cells of the opposite lineage. However, if the separate lineage model were correct, the presence of productively rearranged TCR transgenes should not significantly interfere with the formation of either lineage. Although these predictions in theory seem sufficiently straightforward to allow an easy verification, the phenotypes of TCR transgenic mice regarding the development of aP and 76 lineages have turned out to be highly variable and often difficult to interpret. 1 , TCRaP Traiisgcizic Mice
Concerning the effects of TCRaP transgenes on 76 T-cell development, the most comprehensive set of data has been derived from the analysis of HY transgenic mice, which were generated with large genomic fragments encoding a and /3 T-cell receptor chains recognizing a male-specificpeptide (HY) in the context of H2-D” (Kisielow ct nl., 1988). Expression of the TCRP transgene in the presence or absence of the TCRa transgene was shown to suppress endogeneous V y 2 4Jyl rearrangements and to prevent the generation of 78-expressing cells in thymus and lymph nodes (von Boehmer et al., 1988). Together with similar data from another laboratory (Fenton et al., 1988), this result seemed to support the competitive rearrangement model, as it suggested that early expression of a functional
TCRP chain would relegate all T-cell precursors into the ap lineage. However, data from an analysis of highly unusual CD4-8- and CD4-8"" lymphocytes present in HY-TCR transgenic mice advocate an alternative explanation for the absence of y6-expressing cells. The new findings directly support the separate lineage model because these unusual cells were shown to have several features typical for y6 lineage cells (Bruno et d., 1996). First, they were characterized as DN or CD8'"" cells, which is unusual for mature aP lymphocytes. Second, they were not dependent on positive selection, as they were also found in HY-TCR transgenic mice with a nonselecting MHC background. Third, although they expressed the transgenic HY-specific TCR and could be induced to proliferate by antireceptor antibodies, they did not respond to male antigen. Moreover, the fact that these unusual cells accumulated in peripheral lymphoid organs of male mice demonstrated that their precursors were not subject to thymic negative selection, in contrast to HY-TCR-bearing precursors of conventional ap lineage cells. Fourth and most suggestive, they did not rearrange their endogeneous TCRa genes and retained TCRG alleles on both chromosomes, although conventional SP cells, including CD4-8' cells, from HYTCR transgenic mice showed extensive rearangement of endogeneous TCRa genes. Finally, these cells were absent in mice lacking the common cytokine receptor y chain (DiSanto et al., 1996), which has been shown to be specifically required for the development of inature y6 lineage cells (Cao et al., 1995; DiSanto et al., 1995). These particular features strongly suggested that CD4-8- and CD4-8'"" cells expressing the transgenic TCR actually represent cells of the y6 lineage. This interpretation was endorsed by the finding that a significant percentage of these cells coexpressed a y6 TCR when derived from HY-TCR transgenic mice lacking pTa (Bruno et nl., 1996). Therefore, it seems that y6 lineage cells are not really absent in HY and possibly other aPTCR transgenic mice, rather, they are disguised as aP cells that express the transgenic TCR. This interpretation is in line with earlier findings from another TCRaP transgenic mouse model (Capone et al., 1995). Capone and colleagues (1995) generated transgenic mice carrying an unrearranged TCR minilocus that was under the control of either the TCRa or the TCRP enhancer. This minilocus served as an artificial rearrangement substrate. It was found that rearrangement and expression of the /3 enhancer-containing transgenes occurred during thymopoiesis before those containing the a enhancer, with a pattern superimposable on the patterns of endogeneous TCRP or TCRa rearrangements and expression, respectively. These mice were then crossed with TCRaP transgenic mice expressing an alloreactive receptor specific for the MHC class I H2-K" molecule. The thymus of all K"-TCR transgenic animals contained a significant proportion of unusual I P - T C R ~ ~ ~ ~
aD/y/yfi LINEAGE COMMITMENT
37
cells that were CD4-8-. Analogous to the findings in HY-TCR transgenic mice mentioned earlier, these DN, K”-TCR’”gl’cells did not undergo negative selection on a deleting (H2-K”)background. Most interesting, although the transgenic rearrangement substrates under the control of the (Y enhancer were extensively rearranged in all other K”-TCR-bearing thymocytes, rearrangements were barely detectable in the unusual TCR’”zhDN population (Capone et ul., 1995). Although bearing a transgenic (YPTCR, these cells therefore resembled yS tliymocytes, in which V(Y+J(Y rearrangements are known to be very rare (see Section III,C,3). Taken together, data obtained with TCRaP transgenic mice therefore strongly suggest that early expression of an (YP TCR does not prevent tlie formation of y6 lineage cells. Rather, the transgenic ap TCR seems to be able to functionally replace the y6 TCR, allowing yS development, despite tlie absence of the “correct” receptor-a view fully in line with the separate lineage model.
2. TCRyS Trunsgenic Mice y6 transgenic mice should provide a inore sensitive system to distinguish between the different models of aply6 lineage commitment, as a block in the development of ap thymocytes, as predicted by the rearrangement, but not the separate lineage model, would produce a most obvious phenotype. Unfortunately and contrary to expectation, the analysis of TCRyS transgenic mice has revealed very complex and seemingly inconsistent phenotypes. Two distinct sets of TCRyS transgenic mice have been described by Tonegawa’s laboratory. The first set of mice was generated with large cosmid-based fragments encoding rearranged Vy2Jyl and Va5DJSl receptor chains that were derived from the alloreactive yS T-cell hybridoma KN6 specific for an allelic MHC class I molecule encoded in the T l a region of the MHC (T22”).Coinjection of these fragments yielded three transgenic lines with apparently similar phenotypes (Ishida et al., 1990). Thymic cellularity was nornial, as was the number and subset distribution of ap thymocytes. There was a modest increase in thymic 76-expressing cells from 0.,5% in nontransgenic mice to inaxirnally 5% in 76 transgenic mice. Notably, no transgenic y or S transcripts could be detected in cells of the ap lineage, although the respective transgenes were shown to be present in such cells. To investigate the mechanism of this specific block in TCRy and 6 transgene expression, another group of transgenic mice was generated using a genomic DNA fragment that contained essentially the same Vy2Jyl gene as before, but much less flanking sequence. This time, the 6 construct was not injected. Interestingly, the shortened y transgene was strongly expressed in ap lineage cells in all three independent lines of TCRy transgenic mice studied, whereas expression of the
38
H A N S JORC. F E l K I N G ct nl
corresponding endogeneous y genes was still suppressed (Ishida et d., 1990). These results suggested that the C y l gene, and by inference other Cy genes, carried a cis-acting DNA element in their flanking regions that silenced y transcription in ap lineage cells. Based on this interpretation, a new model for the differentiation of ap and y6 T cells was proposed. According to this model, generation of ap and y6 cells would be independent of the outcome of TCRy,S, or /3 rearrangements. Instead, commitment of a given T-cell precursor to the ap lineage would induce a silencing mechanism, which effectively blocks expression of rearranged TCRy genes, thereby preventing the formation of a functional y6 TCR in the inappropriate lineage. This model was tested in a second set of TCRyG transgenic mice constructed with short y transgenes lacking the putative silencer region. Possibly of relevance, this time the transgenes encoded a completely different y STCR, namely the nonpolymorphic “canonical” Vy3Jyl and VSlDJ62 receptor chains known to be expressed exclusively on the surface of embryonic y6 thymocytes that are generated during the first wave of thymopoiesis and give rise to DECs (reviewed in Allison, 1993; Boismenu and Havran, 1995). In both lines of “DEC transgenic” mice that were analyzed, T-cell development was severely perturbed (Bonneville et al., 1989).The absolute number of thymocytes was markedly reduced, and in young mice, ap-expressing thymocytes were virtually absent. In older mice, some ap-expressing thymocytes could be found but their number never reached more than 5% of the level observed in nontransgenic mice. Southe m blot analysis of thymic DNA revealed that D+Jp and Vp+(D)Jp rearrangements were severely blocked, providing a likely explanation for the paucity of ap-expressing cells in DEC transgenic mice. These results were interpreted in favor of the “silencer model” (Bonneville et al., 1989). It was argued that the absence of the silencer in the transgenic construct permitted early expression of both y and 6 transgenes in ap-committed cells, thereby blocking the development of such cells. However, other scenarios are conceivable. In uitru cell culture experiments have suggested that DEC receptor-expressing y6 thymocytes may be selectively depleted in the adult thymus, depending on the particular strain of mice (Iwashima et al., 1991). To explore this directly, Iwashima and colleagues (1991)generated DEC transgenic mice with hstinct genetic backgrounds. The genomic Vy3N61 constructs used for transgenesis were identical to those in Bonneville’s et al. (1989). DEC transgenic mice with a B6 genetic background revealed a similar phenotype as described earlier. However, subsequent experiments established that this phenotype was largely due to the presence of a negatively selecting factor in the B6, but not the C3H strain. For instance, in mice of the latter strain, there was no strong reduction in the total number of thymocytes, and the development of
DP thymocytes was barely inhibited, in sharp contrast to DEC transgenic animals with a BG back$-ound. Fiirtlier expeiiments demonstrated that this depletion was genetically dominant, mapped to chromosome 18, and was mediated by bone-marrow-derived, radiosensitive cells in the thymus. The inhibition of (YOthyrnopoiesis observed in DEC tranvgenic mice therefore seems to be more a function of the genetic background than being related to the presence or absence of a y gene silencer. Wiether a y silencer is important for an unperturbed development of cup lineage cells has been investigated in a third laboratory with yet another set of y6 transgenic mice (Sin1ct al., 1995). The 6 construct in these mice consisted of a VSlJ82CS cDNA, cloned into an expression vector that contained the heterologous H2-K” promoter and Ig heavy chain enhancer in order to support strong expression in lymphoid cells. The y construct consisted of a short geiiornic fragment carrying a productively rearranged Vy4Jyl gene segment with its own regulatory sequences, all Cyl exons, but, most important, not the putative y silencer. Analysis of nine transgenic lines revealed a wide range of phenotypes (Sim et al., 1995).The frequency of 76-expressing cells in the thymus varied between 1.8 and 93% with a spectrum of intermediate values depending on the particular line under investigation. Unfortunately, the influence of the 76 transgenes on the thymic subset composition and on the absolute number of thymocytes was not reported. Analysis of splenocytes, however, revealed that the number of a0 T cells in two lines was essentially normal, whereas in the remaining seven lines the frequency of a/3 T cells was depressed to various extents, ranging from 25 to 75% of normal. The extent of c.p T-cell suppression in the periphery did not correlate wit11 the frequency of 76-expressing cells in the thymus or with transgene copy numbers. Interestingly, in two lines a distinct population of splenic T cells was found to coexpress an a/3 TCK and the transgenic y6 TCR. These results were taken as evidence that expression of a y6 TCR in a0 lineage cells w7as not inconipatible with c.upT-cell development, implying that functioning of the putative y silencer was not essential for a commitment to the ap lineage (Sim et al., 1995). The study also indicated that the expression of functional TCRy and 6 genes was insufficient to direct the differentiation of a precursor T cell into the y6 lineage, clearly contradicting the rearrangement model. Unfortunately, the study did not provide any data on the RNA expression levels of the transgenes during thymopoiesis in the different transgenic lines, raising concerns that at least some of the phenotypes seen might have been due to weak, delayed, or \wiegated transgene expression. Although these are potential problems in TCR transgenic mice in general, they are more common when using highly synthetic, cDNA-based constructs as transgenes.
40
IlANS JOHC, FEIILING ct a/
Another striking example of the vanability in phenotypes of y6 TCR transgenic inice has been provided by Hedrick’s group. These investigators introduced into the mouse gerinline functionally rearranged TCRy and 6 genes derived froin another alloreactive y6 T-cell clone, called G8 (Dent et nl., 1990).The G8 receptor, like the K N 6 receptor in Tonegawa’s mice, was specific for the b (and to a lesser degree the k) allele ofthe nonclassical MHC class I antigen T22. The transgenic constructs used were unaltered genoinic fragments carrying productive Vy2Jyl and V68(V a l 1)JSl rearrangements isolated froin genomic DNA of the G8 hybridoma. The transgenic y construct was relatively short and most likely devoid of the putative y silencer. Surprisingly, mice within a single transgenic line had two rather different phenotypes (Dent et nl., 1990). In some animals, the thymus consisted of approximately 50% of cells expressing the transgenic y6 TCR, whereas cells bearing an a/3 TCR were virtually absent. In these mice, referred to as type I, the absolute iiiirnber of thymocytes was reduced 7- to 10-fold and the residual thyinocytes were predominantly CD4-8-, suggesting that expression of a y6 TCR had blocked the development of a/3 lineage cells. In other mice of the same line, referred to as type 11, inhibition of afl T-cell developinent was not as severe. Although thymic cellularity in these aniinals was also decreased 7- to 10-fold, the percentage of 76-expressing cells was not as strongly elevated as in type I transgenics and a norinal ratio of a/3 to y6 thynocytes was essentially maintained. Moreover, CD4/CD8 profiles and the expression pattern of the a/3 TCR in different thymic subsets were comparable to normal, nontransgenic mice. Why the same line of transgenic mice gave rise to two apparently quite different phenotypes remained unclear. The presence of different genetic background genes could be responsible. Also, a later study by other investigators demonstrated incidentally that the copy number of the transgenes has a critical influence on the degree of inhibition of a/3 development, as a/3 TCR-expressing cells were shown to be, on average, 8- to 9-fold rarer in hornozygous G8 transgenic inice than in heterozygous litterinates (Livhk et al., 1997). G8 transgenic inice with an H-2” background, in which the ligand for the G8 receptor, the MHC class I T22b “autoantigen” was expressed, exhibited a third phenotype (Dent et al., 1990): cells expressing the transgenic y6 TCR were mostly rare, evident only in relatively small numbers with significantly reduced surface expression levels of the G8 TCR. The analysis established that G8-expressing, self-reactive y6 cells were subject to thymic negative selection. Remarkably, the development of a/3 T cells in thymi of such mice was quite efficient. The relative proportions of DP and SP thymocytes and the expression pattern of the aPTCR were similar to nontransgenic mice. In the original description of G8 transgenic/H-2”
aplyS LINEAGE (:OMM lThlENT
41
mice, almost normal numbers of thymocytes were reported (Dent et al., 1990), although later studies revealed that total tliymic cellularity was significantly (4- to -5-fold) reduced in most animals (Livak et al., 1997). Nevertheless, the fact that large numbers of ap T cells were generated in a thymus with a grossly nonnal subset composition demonstrated that self-reactive transgenic y6 T cells could be eliminated by clonal deletion without precluding ap T-cell development. This implied that the ap T cells did not go through a precursor stage at which the transgenic y6 TCR was expressed in a functional form. These findings were in contrast to the phenotype of DEC transgenic mice, in which a deleting background almost completely blocked ap T-cell development ( Iwashiina et nl., 1991). In a separate approach, Hedrick and colleagues created an additional set of y6TCR transgenic niice (Kersh et al., 1995).The receptor, composed of a Vyl.lJy4Cy4 and a VBD62J61C6 chain, was derived from a hybridoina termed “BAS” and appeared to be specific for a ligand expressed by the hybridoina itself. Possibly of importance, the transgenes were constructed in such a way that their expression could not be extinguished in cells of the ap lineage: the genomic y construct was coinpleinented with an extra enhancer element inserted between two Cy4 constant region exons, and the y transgene consisted of the VDJC6 cDNA driven by regulatory elements of the human CD2 gene known to confer T-cellspecific, copy number-dependent, and insertion-independeiit expression. Based on the phenotype, all foiinder lines could be divided into two groups (Kersh et al., 199rj).In one group, there was low expression ofthe transgenic y cliain and little to no expression of transgenic 6. Not surprising, in these mice, ap T-cell development was normal. The other group was characterized by extremely high levels of transgenic y and 6 message in the thymus, a block in Vp+( D)Jp rearrangements, and an essentially complete absence of cells expressing an ap TCR. Surprisingly,even though Vp+( D)Jp rearrangements were almost fully blocked, these mice had virtually iiorinal numbers of DP thyinocytes, which were devoid of CD25 and expressed low levels of the transgenic y6 TCR. Moreover, these y6 expressing thyinocytes had deleted their endogeneous 6 loci and expressed full-length TCRa transcripts, strongly suggesting that they belonged to the a/3 lineage. However, development beyond the DP stage was coinpletely blocked, most likely because the requirement of ap lineage cells for positive selection could not be met by a y6 TCR complex. Collectively, these data indicate that high level and persistent expression of y and 6 (trans)genes in early thymocytes of the a0 lineage can lead to all consequences of pre-TCR expression: downregulation of CD25, proliferative expansion, differentiation into DP thymocytes, and inhibition of complete rearrangements at the entlogeneous TCRP locus, imitating allelic exclusion. To-
42
IlANS J O R G FEII1.INC et n/
gether with the identification of “disguised” y6 cells in HY-TCR transgenic mice described earlier (Bruno et al., 1996), the findings strongly suggest that CD2StCD44-””” pre-T cells do not distinguish among a@, y6-, or pre-TCR, implying that expression of a specific receptor isotype does not determine lineage commitment. Why do TCRy6 transgenic mice exhibit such a medley of distinct phenotypes? At least three parameters are likely to contribute to this rather unexpected variability. First, the outcome of experiments involving TCRtransgenic mice is almost certainly influenced by the strength and timing of TCR transgene expression. For instance, too low or delayed expression of a y or 6 transgene during the period critical for lineage commitment niay fail to reveal any effects, whereas overexpression of the transgenic TCR at a given stage could lead to excessive signaling and unphysiologic responses. The marked phenotypic differences between heterozygous and hornozygous G8 transgenic mice of the same line illustrate how varymg transgene copy numbers, and thus expression levels, can influence the pattern of thymopoiesis (LivAk et al., 1997). Second, the potential presence of undefined thymic autoantigens recognized by the transgenic TCR may contribute to unusual phenotypes. DEC transgenic C57/B16 ( Iwashima et ul., 1991) and G8 transgenic mice with a H2b background (Dent et al., 1990) provide a good example for the profound effect of a deleting autoantigen on thymopoiesis and highlight the potential for a misinterpretation of results regarding the role of the transgenic receptor in the aP/yS lineage decision. In this context, it may be worth pointing out that the phenotypes of G8 transgenic mice on the deleting background and of KN6 transgenic mice described by Tonegawa’s group (Ishida et al., 1990) are not exceedingly different. The possibility that the developmental pattern seen in mice transgenic for the KN6 receptor (also an alloreactive receptor) may be brought about by the presence of a deleting autoantigen rather than the proposed presence of a y silencer should therefore not be completely neglected. Finally, it cannot be excluded that the influence of a transgenic y6 TCR on the crPIy6 lineage decision depends to some extent on which particular kind of y6 rearrangement has been chosen for transgenesis. For instance, y6 T-cell development involves at least two distinct stein cell populations and occurs in, at minimum, three distinct waves, each associated with discrete types of Vy and V6 rearrangements (reviewed in Boismenu and Havran, 1995). The mechanism of lineage Commitment may therefore vary during different stages in ontogeny (see also Section VII). The use of autologously regulated y and 6 transgenes encoding receptors typical for one of the early waves may elicit results differing from those obtained with receptor genes representative for a later wave. The distinct effects of a deleting background on DEC transgenic (Bonneville et nl.,
a/3/yS LINEACE COMMITMEYT
43
1989) versus G8 transgenic mice (Dent et al., 1990) could be a case in point. Because the genes encoding the DEC receptor are representative for the first wave of thymopoiesis, the corresponding transgenes are most likely expressed much earlier in ontogeny than G8-encoding transgenes. Deletion of DEC-expressing thymocytes may therefore include most precursor cells of the ap lineage, whereas a relative delay in the expression of G8 might leave a significant number of ap precursors unperturbed, explaining why on a deleting background ap thyniopoiesis is completely blocked in DEC, but not in G8 transgenic inice. The massive expansion of ap lineage cells expressing a transgenic Vyl. 1JyW&(D)JSl-TCR in one transgenic model (Kersh et al., 1995) may likewise be related to the particular yl6 genes used. It is interesting to note that the V&(D)J61 rearrangement used in that model exhibits random distribution of in-frame joins in normal ap T cells (Livak et al., 1997; Wilson et al., 1996), whereas most other 6 rearrangements are depleted of in-frame joins, suggesting that many V&( D)JSl containing TCRs may behave differently from the most common TCRy6 heterodimers. Moreover, the observation that infranie Vyl.1Jy4 rearrangements are strongly selected in a subset of CD25-44-""\'cells has led to the suggestion that y chains with this particular rearrangement might act, in analoq to pTa, as a pre-Ty (Passoni et al., 1997). In contrast to expectation, the enormous variability in the phenotypes of different TCR transgenic mouse strains demonstrates that this experimental system does not provide the ideal tool to distinguish between different models of lineage determination. A major problem is to recognize which of the observed phenotypes reflect a physiologically relevant situation. However, the experiments clearly show that thymopoiesis is strongly influenced by the onset and level of TCR expression and possibly also by the type of V, D, or J elements of a particular TCR, but it is not clear to what extent if at all, these parameters influence the aply6 lineage decision in the unmanipulated animal. Although TCR transgenic mice at present cannot provide a definitive answer with respect to the different models of lineage commitment, their analysis has given important clues that can be used as a platform for further discussion and experimentation.
B. GENE-TARGETED MICE 1. Mice Unable to Geiierute TCRa, p, or 6 Chains Targeted disruption of the TCRa, p, or 6 constant region genes in embryonic stein cells has been used to generate mice that fail to express specific T-cell receptor chains. The effects of such mutations on T-cell development are unequivocal. TCRa- (Mombaerts et al., 1992b; Philpott
al., 1992) and TCRP-deficient mice (Mombaerts et al., 1992b) lack mature aP lineage cells, but maintain a full complement of 76 cells. Conversely, TCRG knockout mice are deficient in yG-expressing cells, but this deficit does not affect the development and number of aP lineage cells (Itohara et al., 1993).The results demonstrate that cells of one lineage can develop norinally in the complete absence of mature cells of the other lineage. Unfortunately, these findings alone do not provide any further insight into the mechanism of how the ap/yS lineage decision is made, as the fact that both lineages can develop independent of each other is compatible with both models of lineage commitment presented earlier.
r>t
2. Mornentous Clues: Analysis of TCRy and 6 Rearrangements in TCRP Knockout Mice The development of aP thymocytes is severely blocked in TCRP-deficient inice at the transition from the CD25+44-"" to the CD25-44-""" DN stage (Godfrey et al., 1994; Mombaerts et al., 1992b). It is now clear that this transition is controlled by the pre-TCR consisting of a conventional TCRP chain, an invariant molecule called the pre-TCRa (pTa) chain, and certain CD3 subunits (reviewed in Fehling and von Boehmer, 1997; Rodewald and Fehling, 1998; von Boehmer and Fehling, 1997).The cornpetitive rearrangement model suggests that if the generation of a functional TCRP chain and the subsequent formation of a signaling-competent preTCR precede formation of a ySTCR, the respective cell will commit to the ap lineage. Conversely, generation of a productive ySTCR before the generation of a pre-TCR is predicted to result in a decision favoring the yG fate. If this model were correct, one would expect that in the absence of a TCRP chain, precursor cells will fail to enter the ap lineage. This does not seem to be the case, as the thymus of TCRP knockout mice contains a significant number of DP thymocytes (Mombaerts et al., 1992b). An answer to the question of how these residual DP thymocytes arise in the absence of a TCRP chain, and whether they are genuine ap lineage cells, should therefore provide important clues about the molecular mechanism controlling the aP/yG lineage split. In TCRP-'- mice, DP cells constitute typically only about 20% of all thymocytes, which corresponds to approximately 1-5% of the absolute number found in wild-type littermates (although the frequency of these cells varies widely between individual animals). Importantly, in T C R P P X T C R V doubly deficient mice, DP thymocytes are almost absent, comprising fewer than 1% of all thymocytes, which corresponds to less than 0.085%of the absolute number found in normal, unmanipulated animals. The almost complete lack of DP thymocytes in TCRP-'- X TCRGP mice implies that the vast majority of residual DP thymocytes in
T C R P P singly deficient inice is generated by pathways dependent on functional TCRG (and most likelv TCRy) protein. Earlier experiments in severe combined immuno defic‘ient (SCID) mice have shown that the presence of y6expressing cells can induce other DN thymocytes in trans, in some unknown way, to upregulate CD4 and CD8, resulting in the generation of small nurnbers of TCR negative DP thymocytes (Lynch and Shevach, 1993; Shores et al., 1990). In theory, a similar mechanism could operate in the thymus of T C R P P mice, as it contains large numbers of y6 thymocytes. However, data from two independent laboratories clearly demonstrate that expression of a ysTCR can promote the generation of DP thymocytes not only in trans but also in cis. In the first study, Passoni and colleagues (1997) used the PCR-RFLP technique (Section II1,B) to analyze the quality of V&( D)JG rearrangements in the DP thymocyte population of TCRP-I- mice. Similar studies in normal mice (summarized in Sections III,C,l and III,C,2)had previously shown that in-frame TCRyIG rearrangements are generally underrepresented in DP thymocytes, a finding that had been proclaimed as strong evidence for rearrangement models of lineage commitment (Dudley et al., 1995; Livak et al., 1995). Interestingly, the analysis of rearranged 6 loci in residual DP thymocytes of TCRP-’- inice gave the opposite result: all three V&(D)JS rearrangements studied [VM, V65, V&(D)JSl] revealed inore than 75% of in-frame joins, indicating selection of productive TCRG rearrangements (6 selection). The overrepresentation of functional V&( D)J6 rearrangements in DP thymocytes of TCRP-’- mice was shown to be accompanied by y selection, involving primarily Vy2+ Jrljoins. These findings were confirmed and extended by another study ( LivBk et al., 1997). LivBk and colleagues (1997) first reexamined crP versus y6 T-cell development in G8 TCRyG transgenic mice that had been generated and described some years before (Dent et nl., 1990) (see Section IV,A,2). The G8 ySTCR is specific for the nonclassical MHC class I antigen T22”. Although the development of mature c.P T cells was shown to be severely blocked in honiozygous G8 transgenic mice, largely in line with previous data (Dent et nl., 1990),a significant number of DP thymocytes was found. Quantitative Southern blotting revealed that these DP cells were obviously generated by a TCRP-independent pathway, as they were virtually devoid of VP+( D)JP rearrangements. This was confirmed with G8 transgenic, TCRP-I- mice, which also contained a significant number of DP thymocytes. However, the generation of DP cells was shown to be completely blocked when the G8 transgenes, along with the TCRP-I- mutation, were bred on a negatively selecting H-2” (T22”-expressing)background. These data indicated that the development of DP thymocytes in G8 transgenic, T C R P P mice was contingent on expression of the G8 ySTCR, as the
development of DP cells was abrogated in the presence of a negatively selecting ligand. Importantly, although the proportion of TCRy 6’ cells were strongly reduced in mice where negative selection occurred, significant numbers of cells expressing reduced levels of the G8 TCR were still present. Obviously these cells were unable to induce the generation of DP thymocytes in trans. These findings prompted the same researchers to investigate the quality of TCRy and 6 rearrangements in DP thymocytes of nontransgenic TCRP-I- mice, also using the PCR-RFLP technique (LivBk et d., 1997). The results were very similar to those reported by Passoni et al. (1997), in that in-frame V h (D)J6 and in-frame Vy+ Jy rearrangements were found to be strongly overrepresented in sorted DP thymocytes from TCR0-I- mice. Interestingly, the proportion of in-frame TCRS and TCRy rearrangements at all examined loci was nearly as high as in bona fide y6 cells. In line with these findings, additional experiments involving in situ hybridization and Northern blotting demonstrated that the proportion of DP thymocytes expressing TCRG message was much higher in T C R P P than in wild-type mice. Taken together, these data indicated that, in the absence of a TCRP chain, a sizable fraction of DP thymocytes was generated from precursor cells expressing a functional y6 TCR. To confirm that the DP thymocytes under study really represented lineage cells, additional experiments were performed. First, it was shown that DP thymocytes in G8 transgenic, TCRP-’- mice not only expressed CD4 and CD8 at normal levels, but also had downregulated surface expression of CD25 and the y6 TCR. Second, and more significantly, DP cells had initiated TCRa gene rearrangement at multiple J a genes. Initiation of TCRa gene rearrangements was also demonstrated in DP thymocytes from nontransgenic TCRPY mice (Mombaerts et nl., 1992). All these features are typical for ap lineage cells. The two studies just presented make a very important point. The finding that in-frame TCRy and S rearrangements are overrepresented in sorted DP thymocytes of T C R P P mice provides strong evidence that expression of a functional y6 TCR in a precursor cell is not only compatible with a developmental progression to the DP stage, but even necessary for the generation of a significant fraction of DP thymocytes when a TCRP chain is not available. This finding is so important, as it completely contradicts the competitive rearrangement model. However, analysis also shows that this $-driven pathway to the DP stage is very inefficient, as it becomes apparent only when the major, pre-TCR-dependent pathway is blocked. At least three potential reasons for this inefficiency are conceivable, which are not mutually exclusive: First, in contrast to the pre-TCR-driven pathway, which requires only a productive TCRP rearrangement to be opera-
ol/3/yG LINEAGE COMMITMENT
47
tional, the ySinediated generation of DP thymocytes is thought to need both a functional TCRy and a functional TCRG chain. The probability of generating in-frame rearrangements for both genes within one cell may be relatively low. Second, expression of a y6 TCR in DN precursor cells may not be associated with cellular proliferation, the hallmark of preTCR-mediated differentiation to the DP stage. Cell cycle analysis of DN thymocyte subsets in TCRP-’- mice seems to support this hypothesis (Passoni et al., 1997). Finally, the 76-driven pathway may be accessible for a tiny subset of DN T lineage precursors only. One could hypothesize that such cells belong to a special lineage of unknown function. Expression of a productive yS TCR in all other, more conventional precursor cells might then be expected to give rise to y6 lineage cells, fully in line with the predictions of the rearrangement model of lineage commitment. The demonstration that only a relatively insignificant fraction of precursor cells can follow the y6-driven pathway to the DP stage would rescue the competitive rearrangement model.
3. Lineage Commitinent in pTa-De$cient Mice TCRa genes are rearranged and expressed relatively late in T-cell development, and the availability of functional TCRa chains is therefore unlikely to influence the a@yS lineage decision. In accordance with this view, a/? T-cell development proceeds norinally up to the DP stage in TCRadeficient mice, and a change in tlie ratio of ap- to y6-committed thyinocytes cannot be detected (Mombaerts et nl., 1992b; Philpott et al., 1992). A potential role of TCRP rearrangements in the aP/yS lineage decision, as postulated by the rearrangement models, is thus expected to be mediated predominantly by the pre-TCR rather than a mature aP TCR, intimating an important role of pTa in the aP/y6 lineage switch. The phenotype of pTa-deficient mice seeins consistent with this view (Fehling et al., 1995a, 1997):whereas the generation of CUPlineage cells is severely impaired, the development of cells with a y6 TCR proceeds normally. As a result, the proportion of y Sexpressing cells in tlie thymus of pTa? mice is strongly elevated, comprising typically 10 to 20% of the total number of thymocytes. In some mice the percentage of thymic 7 6 cells can be as high as 30%.More important, however, the absolute number of 76-expressing thymocytes is also augmented and is, in general, 3- to 10-fold higher than in pTa’ littermates (Fehling et al., 1997).This can be interpreted as evidence for a direct role of pTa and the pre-TCR in inhibiting the formation of y6 cells in normal mice, as suggested by the rearrangement models of lineage commitment. Interestingly, a pTa-deficient thymus not only harbors larger numbers of 76 cells, but also features a novel population of y6expressing cells that
48
HANS JORC FEIILING ef a[
coexpress CD4-a population that cannot be detected in the adult thymus of normal mice (Fehling et al., 1997). Although these cells have not yet been characterized in detail, the peculiar surface phenotype suggests that the absence of pTa affects y6 T-cell development not only quantitatively, but can lead to a qualitatively different thymic subpopulation-a finding that again might be suggestive of a direct role of pTa in aPly6 lineage decisions. However, pTa-deficient mice are not completely devoid of apexpressing cells. Similar to TCRP-’- mice, they generate significant numbers of DP thymocytes. At least some of these DP cells express an aPTCR and give rise to apparently normal aP T cells (Fehling et al., 1995a), excluding an absolute requirement for pTa and the pre-TCR in directing cells into the ap pathway. The analysis of pTa? X TCRa-’- mice strongly suggests that some cells in pTa-’- mice follow the aP lineage due to functional replacement of the pre-TCR by a prematurely expressed a/3 TCR (Buer et al., 1997). In fact, the majority of mature ap-expressing cells in pTa+ mice are most likely derived from such precursors that, for as yet unknown reasons, happen to rearrange and express TCRa genes unusually early at the CD25+ DN stage. However, ap lineage cells seem to arise even in pTa? X TCRa-’- doubly deficient mice, as indicated by the presence of DP thymocytes in numbers equivalent to those in pTa? single mutant mice (Buer et al., 1997). These cells might be generated in the same way as DP thymocytes in TCRP-’- mice (see Section IV,B,2). If correct, one would predict that PCR-RFLP analysis will reveal an overrepresentation of in-frame y and 6 rearrangements in these pTa+ X TCRa-’- DP thymocytes. The failure to detect 6 chains in pTa-’- X TCRa-/- DP thymocytes by cytoplasmic staining with anti-TCR6 antibodies (Buer et al., 1997) might just signify that these cells expressed a y6 TCR at an earlier stage in their development, but downregulated expression of the y and/or 6 chain upon entering the DP stage. The presence of DP thymocytes in pTa-’- and especially in pTa-’- X TCRa? mice argues against an obligatory role of pTa and the preTCR in lineage commitment. A similar conclusion had been reached after analysis of TCRP-deficient mice (see Section IV,B,2). The function of the pre-TCR could therefore be limited to trigger expansion of a0 lineage cells and not to influence the lineage decision itself. The significant increase in the absolute number of y&expressing thymocytes in pTa-’- mice could be the result of an indirect effect of pTa deficiency, e.g., the provision of additional space in the thymus due to the reduced number of ab lineage cells. Notably, even in pTa-’- and TCRP-’- mice, y6 cells cannot compensate for the loss of aP lineage cells, as total thymic cellularity on the whole remains below 10% of that found in normal mice (Fehling et al., 1995a),
confirming the largely independent regulation of cell number in both lineages, as already described for TCRy6 transgenic mouse strains exhibiting ii block in aP tliymopoiesis. The assumption that pTa is not directly involved in the lineage decision does not exclude a potential negative influence of the pre-TCR on the rearrangement and/or expression of the TCRyIG loci, suggested by two kinds of observations. First, inhibition of endogeneous Vy2+Jyl rearrangements in TCRP transgenic inice (von Boehmer et al., 1988) is much less pronounced in mice carrymg the same P transgene but lacking pTa (Krotkova et al., 1997). A similar phenomenon is seen with regard to V(D)J6rearrangements (A. Krotkova, H. von Boehmer, and H. J. Fehling, unpublished observation). Second, whereas thyniocytes and lymph node cells from HY-TCR transgenic mice with a pTa+ background are essentially devoid of y6 TCRs, a significant fraction of these cells coexpress the transgenic a0 receptor and the endogeneous y6 TCRs when derived from transgenic litterrnates lacking pTa (Bruno et al., 1996). These findings clearly suggest that expression of a functionally rearranged TCRP chain and subsequent forination of a pre-TCR can strongly inhibit not only further rearrangements at the second TCRP allele, but also rearrangements and perhaps expression of y and 6 loci, at least in TCR transgenic mice. It inust be pointed out that this does not automatically imply that the preTCR can iriliibit the forination of y6 lineage cells under physiological conditions, i.e., in normal mice. For instance, it is conceivable that forination of a pre-TCR and inhibition of y and 6 rearrangements occur only after lineage commitment has taken place, just reducing the number of aP lineage cells with y6 rearrangements. In such a scenario, the pre-TCRmediated inhibition of y6 rearrangements would be fully compatible with the separate lineage model because the inhibitory effects of the pre-TCR would be irrelevant for the lineage decision itself. The fact that y6 T-cell development does not depend at all on the presence of pTa may allow cells that coininit to the 76 lineage to inimediately switch off pTa expression. This may, in fact, be necessary to avoid inappropriate expansion of 76-committed cells coincidentally expressing a functional TCRP chain. These considerations lead to an interesting hypothesis: cells that are destined to follow the y6 lineage may no longer express pTa. This hypothesis would predict the existence of two separate precursor populations for aP and y6 lineage cells that can be distinguished based on the presence or absence of pTa. Targeting of a marker gene that can give rise to an easily identifiable cell surface protein, into the pTa locus of embryonic stem cells by hoinologous recombination, should allow the generation of a mouse strain in which such a presumed lineage-specific expression patteiii of pTa can be easily detected, even in the absence
50
HANS JORC FEHL.INC: rt al
of TCRP and other components of the T-cell receptor complex. Such experiments are in progress.
4 . Potential Role uf “Notch” in Lineage Coniinitment Developmental studies in invertebrates, such as Drosophila or Caenorhabditis elegans, have revealed an important role of the transmembrane receptor protein ‘Notch’ and its ligand ‘Delta’ in a number of distinct cell fate decisions (reviewed in Artavanis-Tsakonas et al., 1995; Greenwald and Rubin, 1992; Simpson, 1995). Signaling through ‘Notch’ can divert cells that would otherwise follow a default pathway of differentiation to adopt an alternative, secondary fate. Quantitative differences of ‘Notch’ expression between adjacent cells appear to play a central role in establishing the lineage split, in that initially small differences in ‘Notch’expression between neighboring cells (possibly due to random fluctuations in gene expression) are amplified over time by the ability of ‘Notch’to boost its own expression and inhibit expression of its ligand. This feedback loop results eventually in the formation of two cellular subsets, which express ‘Notch’ and its ligand ‘Delta’in a mutually exclusive fashion, forcing them to follow distinct developmental fates. The finding that at least one of several murine notch homologues is expressed at high levels in developing thymocytes (Hasserjian et al., 1996; Weinmaster et al., 1991) has prompted a group of researches to investigate whether ‘Notch’ might also play a role in lineage decisions in the thymus. A first study revealed that transgenic ‘Notch’can alter the CD4CD8 ratio of developing thymocytes in favor of CD8-bearing cells when expressed in an activated form (Robey et al., 1996). Along with other data, this finding suggested that ‘Notch,’ in concert with the specificity of the TCR, influenced the choice between CD4 and CD8 T-cell lineages. A second study specifically addressed the role of ‘Notch’ in the aPly6 lineage split, and several observations suggested that ‘Notch’signaling might promote preferential development along the a/3 lineage (Washburn et al., 1997). First, reconstitution of R A G - P mice with a 50 : 50 mixture of bone marrow cells from wild-type mice and from mice in which one ‘notch’ allele had been inactivated by targeted mutagenesis revealed a significantly lower ratio of notcht’- versus notcht’+ donor cells among a0 T cells than among y6 T cells, indicating that cells heterozygeous for ‘notch’ contribute less to the a@ than to the y6 population. Second, in reconstituted mice, the proportion of cells heterozygous for ‘notch’was much lower among C D 4 CD8 DP thymocytes, which are considered to belong to the aP lineage, than among their CD25+ DN precursors, which are bipotential. Third, although the absolute number of y6 thymocytes was not decreased in transgenic mice expressing an activated form of ‘Notch,’an almost fourfold
a@yS LINEAGE COMMITMENT
51
increase in the proportion of 76 TCR’ thymocytes expressing the CD8aP heterodimer was found. Introduction of the ‘notch’transgene into ySTCR transgenic mice expressing the G8 receptor resulted in a similar effect, in that the absolute number of 76-expressing cells was not reduced, but the proportion of CD8aP bearing y6 cells was raised from less than 20% to more than 50% of the 76 population. In fact, a significant fraction of these cells was reported to coexpress CD4 along with CD8aP. Because CD4 and CD8aP expression are generally considered markers for the ab lineage, the observed phenotype was taken as a possible hint that many +expressing thymocytes in ‘notch’ transgenic mice might have actually adopted an aP T-cell fate. Fourth, introduction of the ‘notch’ transgene into TCRP-’mice resulted in essentially complete restoration of the DP thymocyte compartment, indicating-according to the authors-a selective promotion of aP T-cell development. The fact that activated ‘Notch’ had no effect on thymic development when introduced in RAG-deficient mice suggested that ‘Notch’could not act alone but required the presence of a y6 receptor. This was strongly supported by the demonstration that productive rearrangements of the TCR6. and :dso y lociis, wcrc ovcrrepresented in DP thymocytcs from TCWfl-’- mice expressing activated ‘Notch,’in contrast to DP thymocytes from normal mice, which carry predominantly nonproductive y and 6 rearrangements. The authors of this study interpreted their results in the framework of rearrangement models of lineage commitment (Washburn et al., 1997). They proposed that the generation of productive TCRP or TCRy6 chains would influence in some unknown way whether a precursor cell could receive a ‘Notch’ signal. The signals from ‘Notch’ and the respective TCR would then be integrated to lead to a specific lineage decision, in that reduced signaling through ‘Notch’would favor the 76 and enhanced signaling the ap pathway. Although the reported results are highly suggestive and consistent with a role of ‘Notch’ in the aP/yS commitment process, alternative explanations are conceivable. For instance, a similar bias in the ratio of y6 versus aP cells, as the one reported for notch+’-/notch+’+ bone marrow reconstituted mice, would be expected if ‘Notch’ provided a tliymocyte-specific survival rather than a differentiation signal and if y6 cells tolerated a reduction in ‘Notch’ signaling more readily then aP cells. The overrepresentation of in-frame y and 6 rearrangements in DP thymocytes of notch-transgenic TCRP KO mice does not provide conclusive evidence for an involvement of ‘Notch’ signaling in lineage commitment decisions either, because newer studies have found such an overrepresentation also in T C R P P mice lacking a ‘notch’ transgene (Passoni et al., 1997; Livlik et al., 1997; see also section IV.B.2). A ‘notch’ transgene is thus not required to allow thymocyte precursors with in-frame y and 6 rearrange-
ments to develop along the ap pathway; it might just boost their proliferative expansion. Clearly, more details about the effects of ‘Notch’ signaling on defined molecular events in thymocyte development are required before a crucial role of ‘Notch’ in ap/yS lineage commitment can be regarded as firmly established. Studying the course of thymopoiesis in the complete absence of ‘Notch’ might actually allow a more conclusive definition of its role in T-cell development in general and the ap/y6 lineage decision in particular. Because conventional ‘notch’ knockout mice die at a stage well before the initiation of thymopoiesis (Conlon et al., 1995; Swiatek et al., 1994), conditional gene disniption should be considered. V. Cell Culture Studies
Cell culture studies have been carried out in order to determine if the commitment to the cup or yS lineage can be influenced in vitro (Schleussner et al., 1991). Briefly, culturing undisrupted mouse fetal thymus lobes at an air-liquid interface, as so-called fetal thymus organ cultures, results in the development of predominantly ap T cells. However, in optimal cultures, the absolute number of recovered ap lineage cells in FTOC is considerably less than that in the corresponding thymus in vivo (Ceredig, 1988). Conversely, the proportion of yS cells in FTOC is higher than in the corresponding thymus in vivo and can be further increased by the addition of IL-7 to FTOC (Plum et al., 1993). The addition of IL-7 also results in the decreased recovery of a/3 cells. In FTOC, the development of cup but not y6 T cells can be inhibited by activation of the CAMPdependent signal transduction pathway and is associated with a dramatic inhibition of TcRa rearrangements (Lalli et al., 1996). When whole fetal thymus lobes are cultured submerged in tissue culture medium gassed with 10% C02,y6 T cells grow preferentially (Ceredig et al., 1989). Again, a significant inhibition of TcRa rearrangement is seen (E. Mertsching and R. Ceredig, unpublished observation). Importantly, the generation of ap T cells in submersion culture can be restored by elevating oxygen concentrations (Ivanov et al., 1993). Although the exact molecular mechanism(s) responsible for the differential growth of (up versus y6 T cells in these cultures has not been defined, these systems demonstrate that a variety of factors usually not considered in lineage commitment models potentially influence the developmental fate. IL-7 is the only cytokine for which a nonredundant function in thymopoiesis has been clearly established. Its potential role regarding the apl y6 lineage decision therefore deserves special comment. In uitro studies have suggested that IL-7 has an important role in promoting the development of y6 cells. These findings are strongly supported by data obtained
aP/y6 LINEAGE COMMITMENT
53
with gene-targeted mice. Notably, TCRy6+ cells are absent in IL-7 (Moore et al., 1996), IL-7Ra (He and Malek, 1996; Maki et al., 1996b), and ILRy ( y o (Cao et nl., 1995; DiSanto et al., 1995, 1996) knockout animals, whereas the development of ap lineage cells in these mouse mutants is impaired h i t not abrugiitcd. The differential sensitivity of up a r i d y6 cells with respect to 1L-7 deficiency could suggest a direct role of IL-7 in tlie process of lineage commitment. This view is supported by studies in IL7Ra-deficient mice indicating that IL-’7 is selectively required for the initiation of Vy+Jy rearrangements (Maki et nl., 1996a). The arrest of y6 but not ap T-cell development could therefore be due to the inability of IL-7Ra chain-deficient mice to rearrange TCRy loci (Maki et al., 1996a). However, an analysis of TCRy and 6 rearrangements in 7,-deficient mice has revealed rearrangements of both types with normal levels of diversity (Rodewald and Haller, 1998). These latter findings clearly showed that y(mediated signaling was not required to obtain diverse TCRy rearrangements. These apparently discrepant results can be reconciled by assuming that signals can be transduced via the IL-7Ra chain in 7,-deficient mice but not by yc in IL-7Ra-deficient mice. It is also possible that ligands other than IL-7, that can bind the IL-7Ra but not yc chain, are required for TCRy rearrangements. Thymic stroma-derived lymphopoietin (TSLP) is an obvious candidate (Peschon et al., 1994). Moreover, it cannot be excluded that the apparent increase in the level of rearrangements in cells cultured in the presence of IL-7 is due to the preferential outgrowth of cells already containing rearrangements. The role of IL-7 in TCR (Candeias et al., 1997; Muegge et al., 1993) and also IgH (Corcoran et al., 1998) rearrangements therefore remains controversial. VI. Developmental Considerations
Does the mechanism of lineage dwergence differ between the fetal and the adult thymus and is this dependent on tlie stem cell population developing at that particular time? It was shown some time ago that the thymus in birds (Jotereau and Le Douarin, 1982) and mice was colonized more than once during the fetal period (Jotereau et al., 1987). There are several differences between tliymocyte development in fetal mice compared with the adult. a. In the fetal thymus, the kinetics of development are more rapid, so that transition from earliest precursors to mature T-cell migration from the thymus takes only 6-7 days. In the embryo and neonatal mouse, mature T cells do not appear to require a prolonged sojourn in the medulla prior to migration to the periphery. Mature ap and y6 T cells are cycling in
54
HANS JOKG FEIILING et nl
the fetal but not the adult thymus (Ceredig, 1990).The cohorts of precursor cells developing in the fetal and adult thymus may have different origins. As reviewed by Rodewald and Fehling (l998), data obtained with Tcf-l (Verbeek et al., 1995) and Zkaros (Wang et al., 1996) KO mice indicate that inactivation of these genes may have different effects on fetal versus adult T-cell progenitors. b. The efficient generation of Vy3+ cells in the fetal mouse thymus, cells that migrate to the skin to form DECs, seems to depend on a combination of heniatopoietic stem cells and fetal thymus stroma (Uchida et al., 1993). Data (Mallick-Woodet al., 1998) from the fetal thymus argue strongly for the presence of a positive selecting ligand for these canonical y6 receptors. Sequence data from day 14 mouse fetal thymus cDNA indicated that TcRy receptors with noncanonical sequences were expressed at this time (Schleussner et al., 1992), supporting a positive selection model for the generation of y6 cells expressing canonical receptors. c. In addition to V gene repertoire (Havran and Allison, 1988) and junctional diversity (reviewed in Gilfillan et al., 1995), differences in the occurrence of Va-Ja rearrangements between fetal and adult y6 cells can be noted. Such rearrangements are detectable at low levels in adult but not in fetal y6 cells (Mertsching et al., 1997). Fetal but not adult y6 cells are generated as cycling cells (Ceredig, 1990), and the cycling status of y6 cells may reduce RAG protein activity, thereby influencing subsequent TcRa and/or p rearrangements. Alternatively, the efficiency of a rearrangements may be developmentally regulated. d. The fetal thymus contains a greatly increased proportion of y6 cells compared with the adult (Havran and Allison, 1988; Pardoll et al., 1987). In the adult, the presence of a@ cells may have a negative influence on the development of y6s, which may involve the Notch signaling pathway (Robey and Fowlkes, 1998) (see Section IV,B,4). It is unclear if there are differences in Notch signaling between the fetal and the adult thymus. VII. In Search of a Consensus Model for the @ / y 6 lineage Split
A. SUMMARY OF EXPERIMENTAL DATA A decade of research into the mechanism of aPIy6 lineage commitment has not yet provided a generally accepted scheme that would explain how a bipotential precursor decides which of the two alternative T-cell pathways it will choose. However, the large collection of diverse experiments described in detail on the previous pages has given nse to a defined set of data that can serve as building blocks for any future model of aPIy6 lineage commitment. These data can be summarized as follows.
afily8 LINEAGE COMhlITMENT
*55
1. The three most immature CD4-8- thymic subsets, defined as CD25-44 'c-kit +, CD25 '44 'c-kit +,and CD2Fj+44-"""c-kit-'""'TN subpopulations, can develop into both ap and 76 lineage cells in thymic organ culture or upon adoptive transfer in RAG-deficient mice. Tlie most advanced TN subset (CD2i3-44-'"''c-kit-''"" cells) seeins to be able to generate y G-expressing thyinocytes as well, when transferred intrathymically (Petrie et al., 1992), although an independent study involving FTOCs failed to confirm this finding (Godfrey et nl., 1993). 2. Vy-Jy and V&(D)JS rearraiigeinents are coininon in a0 lineage cells, and DP-JP as well as complete VP-(D)JP rearrangements are common in yG lineage cells, demonstrating the absence of tight lineagespecific controls with regard to TCRy, 6, and /3 rearrangements. TCRS rearrangements in aP lineage cells occur to a significant extent on chromosomal DNA, before V a 4Ja-mediated excision of the TCRG locus. 3 . The vast majority of mature a0 T lymphocytes have Va-Ja rearrangements on both chromosomes, resulting in the complete deletion of the TCRG locus. However, deletion of the TCRG locus is not necessary to coininit cells to the UP lineage. In 76 lineage cells, Va-Ja rearrangements occur 011 less than 10% of all available alleles and are therefore rather rare. The paucity of Va- J a rearrangements in iriature T cells can be taken as a marker for cells that have developed along the yS pathway. The biological significance of the residual Va- Ja rearrangements in 78 cells is unclear. 4. In-frame Vy-Jy and V&( D)JGrearrangements are generally underrepresented in aP lineage cells. The depletion of productive TCRy and TCRG joints is the result of a selection process and not of a bias introduced by the rearrangement mechanisin. Tlie depletion is detectable in mature aP T lymphocytes,in DP thyniocytes, and, with respect to TCRG rearrangements, apparently also in CD25+44-""" DN thymocytes, suggesting that diversion of yG lineage cells begins at this early developinental stage. 5 . y6 T lymphocytes and yG thymocytes are clearly not devoid of functional TCKP rearrangements. In fact, in-frame VP+( D)JP rearrangements appear to be overrepresented in most peripheral y6 T lymphocytes, except in DECs. An overrepresentation of productive P joints has not been found in day 15 embryonic yS thymocytes, and whether there is a significant overrepresentation of functional TCRP rearrangements in adult 76 thymocytes is still a matter of debate. The apparent dominance of y6 T cells, and possibly yG thymocytes, with in-fi-ame TCRP rearrangements is unexplained, both mechanisi-ically and teleologically. 6. The effect of functionally rearranged TCRyG transgeiles on the development of a0 lineage cells in TCR transgenic inice is extremely variable between different transgenic inoiise lines and sometimes even within a
56
H A N S JCjHG FEIlLlNG et NI
given line. In many TCRy6 transgenic mice, the development o f a p lineage cells is not abrogated. However, development is blocked in DEC transgenic mice, which express a y6 TCR typical for the first wave of tliymic yS cells, and in some mice of the G8 transgenic strain expressing a y6 TCR specific for certain alleles of the nonclassical class I antigen T22. In some situations, where aP development is blocked by y6 transgenes, it is difficult to exclude that negative selection rather than altered lineage decisions are responsible for the observed phenotype. When the development of afi lineage cells is blocked by 76 transgenes, the observed increase in 76-expressing cells never compensates for more than 10% of the total number of cells in a normal thymus, indicating autonomous control of homeostasis for both lineages. TCR can inhibit the formation of y6 7. Expression of a transgenic TCRs, but at the same time rescue the development of y6 lineage cells (Bnino et nl., 1996).Conversely, expression of certain transgenic constructs encoding a y6 TCR can block the rearrangement of endogeneous TCRP loci, and thus the generation of a functional TCRP chain, but at the same time fully rescue the development of aP lineage cells up to the DP stage (Kersh et nl., 1995). 8. The development of y6 lymphocytes proceeds normally in the absence of TCRa, TCRP, and pTa chains. Conversely, the absence of TCRS chains does not affect the generation of lineage cells. These findings imply that the differentiation of cells of one lineage is independent of the presence of mature cells of the opposite lineage. 9. The formation of a functional pre-TCR is not obligatory for thymic precursor cells to enter the aP pathway, as documented in pTa- and TCRP-deficient mice, which contain small but significant numbers of a$ lineage cells. 10. A significant portion of the residual DP thymocytes in TCRP-’mice represent cells selected for in-frame TCRy and 6 rearrangements. A similar situation may exist in pTa-deficient mice. Provided that the respective DP thymocytes in TCRP-I- mice are genuine a0 lineage cells, as many parameters suggest, this seminal finding implies that expression of a ysTCR allows precursor cells to follow an aP fate. This result discounts the rearrangement model of lineage commitment because it demonstrates that the isotype of the TCR is not the determinant of intrathymic lineage commitment, at least in TCRP-deficient mice. The question is now to what extent this observation can be generalized, i.e., whether it applies to most T-lineage-committed precursor cells or only to some atypical subset(s ) that may be dominating in the absence of a functional TCRP gene. 11. The absolute number of y6-expressing thymocytes is 3- to 10-fold increased in both pTa-’- and T C R P P mice. This increase cannot compen-
sate for the dramatic loss of aP lineage cells in these mouse mutants, indicating again an aiitoiioinous control of homeostasis for both T-cell lineages. 12. The expression level of the traiisismembrane receptor ‘Notch’ might influence tlie aply6 lineage decision, as cells with relatively higher levels of ‘Notch’ fivor the a0 pathway and those with relatively lower levels the y6 pathway. 13. Cell culture studies suggest that microeiivironinental factors, such as oxygen or carbon dioxide concentration, presence of cytokines and hormones, and cell-cell contacts, might have the potential to influence the ap versus y6 lineage decision. 14. The fact that fetal and adult thyinopoiesis differ to some degree raises concerns that certain aspects of lineage commitment might change at or around birth. This possibilit\, sliould he kept in mind when interpreting results obtained with young ‘id ti 1t mice. A major problem in finding a unifying sclieme of aP/yS lineage commitmet based on these data is posed by the confusingly diverse phenotypes observed in TCR transgenic inice (suniniaiy:points 6 and 7 ) .If‘one assumes that at least some of these phenotypes represent transgenic-specific peculiarities that do not reflect a physiologically relevant situation, one might decide to disregard the results obtained with TCR transgenic inice altogether and to focus, for the sake of simplicity, first of all on data obtained with norinal and gene knockout mice. If‘this approach led to an inherently concordant model, one coiild then, in a second step, assess to wliat extent the various data derived from transgenic inice support or discredit this conselisus. Following this approach, the following sections present inodified versions of the classical competitive rearrangement and separate lineage models, taking into accoiiiit tlie findings listed earlier. The presented models may help pinpoint “loose ends” and “moot points,” thus providing a platform -for further experimentation and discussion.
B. R E\T, E 11COMPETITIVE RRAHHAN c Based on the competitive rearrangement inodel, the following scheme of a/3/y6 lineage commitment can be devised (Fig. 2 A ) . If a functional TCRP chain is generated in ii bipotential T-cell precursor before the formation of ;I fiinctional ySTCR, the TCKP chain will associated with pTa and form a signaling-conrpetent pre-TCR complex. Signaling through the pre-TCR will then result in coinpletc inhibition of further rearrangemerits at all TCRy, 6. and P loci in tlic rcspective cell, upregulation of ‘Notch’ activity, commitment to the aP lineage, escape from prograiriined cell death, and massive cellular proliferation, followed by lineage-specific
bipotential precursor cells
A
-
tive B rearrangement first
productive y and 6 rearra formation of a y first
formation of a pre-TCR
pTa-
pTa+
* shutdown of further Y
3 shutdown
and S rearrangements of pTa expression I
3 commitment
3 shutdown
to the ve
B
0
of further rearrangements notch activity the ap lineage
3 upregulation of 3 commitment to
rearrangement
+ downregulation of CD25
bipotential precursor cells
ap-committed
*shutdown of pTa expression
no productive p T a + P S ) O rearrangements
rearrangements
tJB
productive p rearrangement reyangement
no productive rearrangement
formation of a yG-TCR
/
productive 1and 6 rearrangements 3 shutdown of further rearrangements
3 shutdown
NO or VERY LITTLE
/
mature yS thymocyte
t
inappropriate TCRF failed positive selection
3
=1formation
of a pre-TCR
of further rearran ements
P
aplyG IJNEAGE COMMITMENT
59
differentiation. If, however, the bipotential T-cell precursor manages to express a functional ySTCR before a pre-TCR, it will stop further rearrangements at the remaining TCRy and 6 loci, rapidly shut off pTa expression, downregulate ‘Notch’ activity, and commit to the y6 lineage. Subsequent TCRP rearrangements may be impaired, but the generation of a functional TCRP chain will have little consequence for the respective cell due to the postulated absence of pTa. Nevertheless, the presence of a functional TCRP protein may give rise to some weak, pre-TCR-independent signal that could result in a veiy modest expansion of the respective cells, similar to that seen in TCRP-transgenic mice lacking pTa (Krotkova et d., 1997),which might explain a potential overrepresentation of productive P rearrangements in y6 thymocytes (summary: point 5). A critical prediction of this modified version of the original competitive rearrangement model remains, namely, that a bipotential precursor must perceive and interpret signals from a pre-TCR very differently from those generated by a y6 TCR.
FIG. 2 . Revised versions of the "competitive) rearrangeinent” (A) arid the “separate lineage” ( B ) models. Tlie classical “separate lineage model” was updated to be compatible with the finding that productive TCRy a r i d 6 rearrangernents are clearly underrepresented in cells of tlie (YPlineage. This obsenution has been incorporated into the revised model (B) by assuming that cells committed to the CUPlirieage but express a y6 TCR will fail to significantly proliferate and thus become a minor subset, in contrast to those precnrsors expressing a pre-TCR. Moreover, y8-expressing ab lineage cells are expected to perish, either on their way toward tlie D P stage because expression of tlie y and/or 6 genes is turned off ( e g , through excision of the S locus aiicl/or activation of a y silencer) or at the D P stage itself because of the presnined inability of a y6 TCR to mediate positive selection of (YPlineage cells. Further modifications of both classical models were necessary to account for the fact that y6-expressing cells are not devoid of functional TCRP rearrangeinents. This is explained in the revised models by assuming that coniniitment to the yS lineage will shut-off pre-TCR signaling, either 1)y extinguishing p Ta expression (as shown) or by sonic other as yet unknown niechanisnt. Alternatively, tlie productive rearrangements seen in y6 lineage cells may give rise exclusively to TCHP chains incapable of forming a preTCR due to pairing problems with pTa and/or the hypothetical VpreT subunit. The demonstration that a significant fraction of‘DP tliyniocytcs in TCRP-’- mice is ( y ) 6selected appears incompatible d i the competitive rearrangeinent inodel depicted in A. To rescue this model, one has to invoke at least wie of’tlie following three ad hoc assumptions: First, the (y)S-selected DP thyinocytes in TCRP-deficient mice are not genuine aP lineage cells. Second, the 76-dependent pathway Ieatling to D P thymocytes in TCRP-deficient mice is a very minor, physiologically irrelevant route, which is not available for tlie vast majority of T-cell precursors. Third, the determining role of tlie TCR isotype for the aPly8 lineage commitment decision is not absolute and can be overruled in some (rare?) situations by other extracellular or cell-autonomous factors.
C. REVISEDSEPARATE LINEAGE MODEL Data obtained in normal and gene knockout mice can also be incorporated into a separate lineage model, as depicted in Fig. 2B. A bipotential precursor commits to the a/3 or yS lineage independent of tlie status of TCR rearrangements, but influenced by patterns of ‘Notch’expression and by other, as yet unidentified, factors and/or cellular interactions. Commitment to the y6 lineage results in an immediate shutdown of pTa expression or, alternatively, in a silencing of the pre-TCR signaling pathway by some other mechanism. Failure to generate a 76 TCR will lead to cell death, whereas expression of a functional y6 TCR will allow the cell to unfold its developmental program and to differentiate along the 76 pathway, albeit without extensive proliferation. Coincidental expression of a functional TCRP chain will not have the same strong effects as in cells committed to the crP lineage, as the formation and/or function of the pre-TCR is coinpromised after commitment to the 7 6 lineage. However, the presence of a functional TCRP chain may give rise to some weak pre-TCR-independent signals, as described earlier, which in turn may allow such cells to proliferate weakly or to survive somewhat longer than their TCRP-negative neighbors, giving them a greater chance to achieve productive y and 6 rearrangements before programmed cell death occurs. Expression of a functional TCRP chain in cells that have committed to the CXPlineage, however, will result in formation of a pre-TCR, inhibition of further rearrangements at TCRy, 6, and @ loci, massive proliferation, and preprogramnied differentiation along the c.P pathway. Inappropriate expression of a y6 TCR in @-committed precursors will also lead to inhibition of further rearrangements and possibly some developmental progression, however, and most importantly, not to extensive cellular proliferation. Expression of a y8 TCR in the a0 lineage is predicted to result eventually in cell death caused by one of several reasons. For instance, cominitinent to tlie ap lineage is most likely associated with a shutdown of tlie transcriptional activity of TCRy and/or TCRG genes and excision of the 6 loci (at a stage when the P locus can no longer be reactivated), depriving developing thyrnocytes of their receptor. Alternatively, or in addition, maturation beyond the DP stage may be blocked, as a y s TCR does not allow positive selection. The distinctive feature of this modified version of the separate lineage model is the assumption that the observed underrepresentation of productive TCRy and 6 rearrangements in CYPlineage cells is not due to a diversion of uncommitted precursors into the y8 lineage, as suggested by the rearrangement model. Rather, the paucity of in-frame yl6 rearrangements in a/3 lineage cells is regarded as the combined result of two effects: (1) elimination of ap-committed cells with in-frame y and 6 gene re-
arrangements, because they are developmentally arrested at the DP stage in the absence an aPTCR; and ( 2 )selective, pre-TCR-mediated expansion of cells with a functional TCRP chain lacking in-frame y/S rearrangements.
D. INTEGRATION OF CONFLICTING: DATA The uncontested finding that productive y and 6 rearrangements are underrepresented in ab lineage cells (suinmary: point 4)strongly supports a model that assumes that the oiitcoine of TCR rearrangements influences the lineage decision. However, the equally uncontested findmg of productive rearrangements at TCRP loci in y6 cells is difficult to incorporate into such a model because it demonstrates that expression of a functional TCRP chain is not sufficient to direct cells into the c.P pathway (summary: point 9). One way to reconcile these apparently contradicting findings is suggested by studies focusing on B-cell development, which have shown that not all functional Ig heavy chains are able to form a pre-B-cell receptor, obviously because of pairing problems with AS (the B-cell analogue of pTa) and/or VpreB chains (ten Boekel et a l , 1998). Whether functionally rearranged TCRP genes in y6 cells encode P polypeptides with similar deficits has not yet been tested and thus remains an interesting possibility. Another way to account for the presence of productive TCRP rearrangements in y6 lineage cells without discarding the competitive rearrangement model would be to amend this model with two ad hoc assumptions: (1)all functional P rearrangements observed in 76 cells have been generated after the formation of a yS TCR, implying that expression of a 76 receptor does not always block further rearrangements at the TCRP lociis; and (2) forination of a functional TCRP chain in cells already expressing a y6 TCR cannot override the lineage decision brought about by earlier signals from the 76 TCR. In fact, one may want to extend this second assumption by postulating that formation of a functional P chain in cells already committed to the y6 lineage also fails to induce massive cellular proliferation or any other effects usually associated with the formation of a pre-TCR. This extension seeins necessary to account for the low frequency of y6expressing cells compared with a p lineage cells in normal mice, and for the fact that the absence of a pre-TCR does not result in a fiirther reduction ofy6expressing cells, but instead in a significant increase (summary: point 11).Unless future studies demonstrate a selective, inherent deficit of y6 cell-specific TCRP chains in pre-TCR formation, the most straightforward explanation for the proposed failure of a TCRP chain to induce pre-TCR-mediated effects seeins to be the absence of pTa in y6committed cells. The most serious threat for the competitive rearrangement model, as depicted in Fig. 2A, arises from the finchng that DP thyinocytes in
62
HANS JOHC FEHLING ct ol
TCRP-I- mice are selected for in-frame TCRy and 6 rearrangements (summary: point 10).The corollary that expression of a yS TCR can stimulate precursor cells to differentiate along the a/3lineage completely contradicts the competitive rearrangement model. To rescue this model one would need to show that the yS-driven pathway to the DP stage is a peciiliarity of TCRP-deficient mice with little physiological relevance. The key question is therefore whether this pathway is available to all T-lineagecommitted precursor cells in a normal thymus or whether it is confined to some rare, atypical cells that may dominate in TCRP-deficient mice. The competitive rearrangement model could also be rescued by arguing that TCRP-’- DP thymocytes with functional ySTCRs are not “genuine” aP lineage cells. Rather, they may represent cells of the yS lineage that express CD4 and CD8 and downregulate CD25 (see also Section I,B,4). It is not easy to rigorously exclude this possibility. The demonstration of Va-+Ja rearrangements in DP thymocytes of TCRP? mice seems to address this issue (Mombaerts et al., 1992; Passoni et al., 1997), but it is hard to prove that these rearrangements occur in the same cells that have undergone (y)6 selection, but not in unselected bystander cells. In an attempt to reconcile experimental data with the competitive rearrangement model, the “codeterminant model” of lineage commitment has been proposed (Passoni et al., 1997). According to this model, fate is determined predominantly, but not exclusively, by the isotype of the TCR. Other factors in the extra- and intracellular environment, for instance, cytokines, cell-cell contacts, or “Notch,” can overrule decisions based on a particular TCR isotype. This adjustment of the competitive rearrangement model would account for the occasional appearance of ap or YC? lineage cells with a “wrong” receptor. However, unless specific “codeterminants” are identified and molecularly defined, this model appears somewhat abstract, especially in light of the fact that all experimental data derived from normal and gene-targeted mice can equally well be accommodated by the “separate lineage” model as presented in Fig. 2B, without postulating a determining role of the TCR isotype in the lineage decision. Which of the models is in best agreement with the findings in TCR transgenic mice? Data summarized under point 7 directly support the separate lineage model and the assumption that expression of a y6 TCR in a@lineage cells will eventually lead to a block of further a/3 lineagespecific development. Conversely, the same data are essentially incompatible with any simple version of the rearrangement model. The separate lineage model is also strongly supported by the finding that a pre-TCR is not obligatory for the formation of aP lineage cells (summary: point 9). Interestingly, both models presented here make similar predictions with regard to the phenotype of TCRy6 transgenic mice, in that early expression
aply8 IJh’EAGE C O M M I T M E N T
63
of a transgenic y6 TCR will obstruct the development along the aP pathway, albeit for different reasons and at different developmental stages (see earlier discussion). The block of (YPT-cell development in some G8 transgenic mice (Dent et nl., 1990)and the partial block observed in seven lines of TCRy6-transgenic mice described by Siin et nl. (1995)are therefore not in disagreement with the separate lineage model. Rather, the question is why does ~$3 T-cell development proceed at all in some TCRyS transgenic lines? Too weak, delayed, or variegated expression of one of the two transgenes in the particular transgenic lines is a possible and commonly presented answer. To avoid such problems and rigorously assess the effect of functionally rearranged TCR transgenes on the aPIy6 lineage decision, it may be necessary to target gene fragments encoding rearranged Tcell receptor chains directly into the respective endogenous TCR loci via homologous recombination in ES cells, which should prevent variation in the phenotype due to transgene-specific effects. While still being technically demanding, such experiments are clearly feasible and promise more reproducible results. Both models presented in Fig. 2 make a number of predictions that are at present not easy to verify experimentally. A major obstacle in studying the mechanism(s) of the a/I/yG lineage split is the absence of lineagespecific markers other than the TCR isotypes themselves. If such markers were available, one could, for instance, assess whether a lineage split exists before and irrespective of TCR rearrangements in rearrangement-deficient mice, which could yield direct proof for the separate lineage model. Conversely, monitoring the effect of targeted TCR transgenes on the expression of such putative lineage-specific markers should providc a conclii+c. answer to the question of whether expression of a specific TCR isotype gives rise to cells of the corresponding lineage and how universal this effect would be. Modern methods of subtractive screening may soon lead to the identification of genes that are expressed in an aP- or 76-specific fashion. Some of these dlfferentially expressed genes might be expected to encode transcription factors directly involved in the in-rplementationof the lineage decision, which would render such a search particularly rewarding. Whether pTa can be used as a reliable molecular marker to distinguish aj3- from y6committed thyinocytes in the absence of TCR expression should soon be known.
ACKNO\VI,EDC~.IENTS Rhodri Ceredig thanks INSERM for support during his leave of absence at the Basel Institute for Immunology. We thank Hans-Reimer Rodewdd ( Basel) for inany helpful coniinents and suggestions. The Basel Institute for Immuno~ogywas founded and is supported by F. Hoffniann La-Roche, Basel.
64
H A N S J O R G FEHLING et 01
REFERENCES Allison, J. P. (1993). y8 T-cell devekpnent. Crin-. Opin. Zrttnwno!. 5, 241-246. Allison, J. P., and Lanier, L. I,. (1987).Structure, function, and serology ofthe T-cell antigen receptor complex. Atmu. R P ~Znintunol. . 5, 503-540. Anderson, G.. Anderson, K. L., Tchilian, E. Z., Owen, J. J. T., and Jenkinson, E. J. (1997). Fil~roblastdependency during early thymocyte development maps to the CD25' CD44t stage and involves interactions with fibroblast matrix molecules. Eur. ]. Znmrztnol. 27, 1200-1206. Antica, M., Wn, L., Shortman, K., and Scollay, R. (1993). Intrathymic lyinphoid precursor cells during fetal thymus development. /. Ztritrtttnol. 151, 5887-5895. Arase, H., Arase. N., Nakagawa, K.. Good, R. A,, and Onoe, K. (1993). NK1.1tCD4tCD8thymocytes with specific lyinpliokine secretion. Eur. J. Ztrtn~iinol.23, 307-310. Ardaivin, C., Wu, L., Li, C.-L., and Shortinan, K. (1993).Thymic dendritic cells and T cells develop siinnltaneously in the thymus from a coininon precursor population. Nature 362, 761-763. Artavanis-Tsak(trias,S., Matsiino, K., and Foitini. M. E. (1995). Notch signaling. Science 268,225-232. Asamow. D. M., Cado, D., and Raulet, D. H. (1993). Selection is not required to produce invnriant T-cell receptor y-gene junctional sequences. Nature 362, 158-160. 13attistini. L., Horsellino, G . . Sawicki, G.. Poccia, F., Salvetti, M., Ristori, G., and Brosnan, C. F. (1997). Phenotypic and cytokine analysis of human peripheral blood y6 T cells expressing N K cell recrptors. /. Zttitntitiol. 159, 3723-3730. Bendelac, A., Rivera, M. N., Park, S. H., and Roark, J. H. (1997). Mouse CDl-specific N K I T cells: development, specificity, and fnnction. A;UUJ.Rev. Zttttnutd 15, 535-562. Blomberg, H., kind Andersson, B. (1971). Characterisation of the immnnoconipetent cells in the niouse thymus: Cell population changes during cortisone-induced atrophy and subsequent regeneration. Cell. Z i t m i i r t o L . 1, 545-560. Boesteaiiu, A., Silva, A., Nakajima, €I., Leonard. W. J., Peschon, J. J.. and Joyce, S. (1997). Distinct roles for signals relayed through the common cytokine receptor y chain and interleukin 7 receptor a chain in natural T cell development.]. Exp. Med. 186, 331-336. Boismenn, R., and Havran, W. L. (1995). T-cell clevelopment: T-cell lineage commitment reLisitrd. Ctrrr. B i d 5, 829-831. Boisinenu. R., and Havrm, W. I,. (1997). An innate view of y8T cells. Cum. Opin. Ittttmrtiol. 9, 57-63. Bonnevilk, M., Ishida, I., Moml)aerts, P., Katsuki, M.,Verbeek, S., Bems, A,, andTonegawa, S. (1989). Blockage of aP T-cell development by TCR y 6 transgeaes. Nature 342, 931-934. Brawand, P.. Biasi, G., Horvath, C., Cerottini, J . C., and MacDonald, H. R. (1998). Flowinicrofluorometricnionitoring of oligoclonal CD8' T cell responses to an immunodominant Moloney leukemia virus-encoded epitope it1 oiuo. /. Ztrimmol. 160, 1659-1665. Brenner, M. B., McLean, J., Dialynas. D. P., Strominger, J. L., Smith, J. A,, Owen, F. L., Seidniiin, J. G . , Ip, S., Rosen, F., and Krangel, M. S. (1986). Identification of a putative second T-cell receptor. Nature 322, 145-149. Bruno, L., Fehling, €1. J., and von Boehmer, H.(1996).The crp T cell receptor can replace the y6 receptor in the tkvehpinent of yS lineage cells. Ziriititinity 5, 343-352. Budtl, R. C., Miescher, G. C., Howe, R. C., Lees, R. K., Bron, C., and MacDonald, H. R. (1987). Developmentally regulated expression of T cell receptor P chain variable domains in immature thymocytes. ]. Exp. Med. 166, 577-582. Buer, J., Aifantis. I., DiSanto, J . P., Fehling, H. J.. and von Boehmer, H. (1997). Role of different T cell receptors in the development of pre-T ce1ls.J. Exp. Med. 185, 1541-1547.
Burtnini. D. B., Kim, S., Dudley, E. C . , Hayday, ‘4.C.. and Petrie, H. T. (1996). TCR gene recombination and a@-ySlineage divrrgence: Productive TCR-@ rearrangenlent is neither exclusive nor preclusive of y6 cell devehpmeut. I. I t i n n i c n o ! . 157, 4293-4296. Candrias. S., Mueggc. K., and Dnruni, S. K. (1997).IL-7 receptor and VDJ recombination: Tropliic versus meclranistic actions. Itrmiutiify 6, 501-508. Cantor, H., antl Boyse, E. A. ( 1975). Functional subclasses of T-lymphocytes 1)eariiigdifferent Ly antigens. I. The generation of finictionally dIstinct T-cell subclasses is a diffcrentiative process independent of antigen. I . Exp. A4rtl. 141, ~ 3 7 6 -1389. Cantor. H., and Boyse. E. A. (1977). Lyniphocytes a s models for the study of ~namnialian cellrilar difkwntiation. Ztnnurno/. Z h . 33, 105- 124. Cao. X., Shores, E. W., Hu Li, J., Anver, M. R., Kelsall, B. L., Russell, S. M.. Drago, J.. Noguclii, M., Crinherg, A., Blooni, E. T., et 01. (1995). Defective lymphoid development in mice lacking expression of the coinnion cytohne receptor y chain. Ztrmzrnity 2,223-238, Capone, M., Cuniow, J.. Bouvier. G., Ferrier, P., and Horvat, B. (1995).T cell development in TCR a@transgenic mice: Andysis iising \’(D)J recombination suhstrates. J . Zrtiniunol. 154, 5165-5172. Capone, M.. Wattin. F.. Fernex, C., Iloivat, B., Krippl. B., ~VII.L., Scollay, R . , and Ferrier, P. (1993). T C R @ and TCR a gene enliaricers confer tissue- antl stage-specificity on V(D)] recombination events. EMBO /. 12, 4335-4346. Carding, S. R.. Allan. W., Kyes, S., IIavday, A . , Bottondy, K., and Doherty, P. C. (1990). Late dominance of tlre inHarnmatory process in niiirine inflnenza by y/S+ T cells. 1.Exp. M d . 172, 1225-1231. Carena, I., Shamshiev, A,. Donda,A,, (:olonna, M., and Libero, <;. D. (1997). Major histocompatibility complex class I molecnles niodnlate activation thresholtl and early signaling of T cell antipm receptor-y/S stimulated by nonpeptidic hgands. /. Exp. Met/. 186, 1769-1774. Carlyle, J. R., Michie, A. M., Clio, S. K.. and Zunig;i-PHiicker, J. K. (1998). Natural laller cell dcvelopinent antl finiction precede a@ T cell differentiation i n nroiise fetal tlrvniic ontogrny. /. Z t i i t u t r r ~ o l . 160, 744-75:3. Casanova, J. L., Romero, P.. Widmann, C., KoiirilsLy, P., and Maryanski, J. L. (1991). T cell receptor genes in a series of class I major Iiistocoinpatibilit). complex-restricted cytotoxicT lymphocyte clones specific for a P/osttmdhtti betphi nonapeptide: Implications for T cell allelic exclrision and aiitigen-specific repertoire. /. Exp. Med. 174, 1371-1383. Ceredig, R. (1988). Differentiation potrntial of 14-day fetal mouse thyinocytes in organ culture: Analysis of CD4/CDH-defined singlepositive antl donble-negative culls. J. Inmuriol. 141, 355-%362. Cc-redig, R. ( 1990).Intrathymic pi-oliferation of perinatal nroiise a@and y8 T cell receptorexpressing inatlire T cells. Znt. I t n r t i i i r i o l . 2, 859-867. Ceredig, R., Dialynas, D. P., Fitcli, F. it‘.. and Macllonald. H. R. (1983). Precursors of T cell growtli factor producing cells in the thymus: C)ntogeny, frequency, and quantitative recovery in a subpopulation of phenotypically mature thyniocytes defined by monoclonal antibody GK-1.5. /. E x p . &fed. 158, 1654-1671. Ceredig, R., Glasebrook, A. L., and M w D o I ~ 13. ~ . H. (1982). Phenotypic and functional properties of niurine thymocytes. I. Prwiirsors of cytolytic T lymphocytes and interleukin 2-producing cells are all contained within a snhpopiilation of “mature” thymocytes as analyzed by monoclonal antibodies and How microHiiorometry.]. Exp.Met!. 155,358-379. Ceredig. H., Loweiithal, J. LV., Nahholz, M.. and MacDonald,H. R. (1985). Expression of interlnekin-2 receptors as a differentiation marker nn intrathymic stem cells. Notzrre 314, 98-100.
66
HANS JORG FEHLING et a1
Ceredig, R., Lynch, F., and Newman, P. (1987).Phenotypic properties, interleukin 2 production, and developmental origin of a “mature” subpopulation of Lyt-2- L3T4- mouse thymocytes. Proc. Nutl. Acud. Sci. USA 84, 85784582. Ceredig, R., and MacDonald, H. R. (1985). Intrathymic differentiation: Some unanswered questions. Sum. lminunol. Res. 4, 87-95. Ceredig, R., Medveczky,J . , and Skulimowski,A. (1989).Mouse fetal thymus lobes cultured i n IL-2 generate CD3+,TCR-y6-expressing CD4-/CD8+ and CD4-/CD8- cells./. Iminnnol. 142, 3353-3360. Ceredig, R., Sekaly, R. P., and MacDonald, H.R. (1983). Differentiation in vitro of Lyt 2+ thymocytes from embryonic Lyt 2- precursors. Nature 303, 248-250. Chen, H., Huang, H., and Paul, W. E. (1997). NK1.1+ CD4+ T cells lose NK1.l expression upon in vitro activation. J. Zmmunol. 158,5112-5119. Chien, Y. H., Iwashima, M., Kaplan, K. B., Eliot, J. F., and Davis, M . M. (1987). A new T cell receptor gene located in the a! locus arid expressed early in T cell differentiation. Nature 327,677-682. Chien, Y. H., Iwashima, M., Wettstein, D. A,, Kaplan, K. B., Elliott, J. F., Born, W., and Davis, M. M. (l987a).T-cell receptor 6 gene rearrangements in early thymocytes. Nature 330, 722-727. Conlon, R. A,, Reaume, A. G., and Rossant, J. (1995). Notch 1 is required for the coordinate segmentation of somites. Development 121, 1533-1545. Corcoran, A. E., Riddell, A,, Krooshoop, D., and Venkitaranian, A. R. (1998). Impaired immunoglobulin gene rearrangement in mice lacking the IL-7 receptor. Nature 391, 904-907. Crompton, T., Ohashi, P., Schneider, S. D., Pircher, H., and MacDonald, H. R. (1992). A cortisone sensitive CD3“” subset of CD4+CD8- thymocytes represents an intermediate stage in intrathymic repertoire selection. lnt. Immunol. 4, 153-161. Dent, A. L., Matis, L. A,, Hooshmand, F., Widacki, S. M., Bluestone, J. A,, and Hedrick, S. M. (1990).Self-reactive yS T cells are eliininated in the thymus. Nature 343,714-719. de Villartay, J. P., and Cohen, D. I. (1990). Gene regulation within the TCR-d6 locus by specific deletion of the TCR-6 cluster. Res linmunol. 141, 618-623. de Villartay, J . P., Hockett, R. D., Coran, D., Korsmeyer, S. J., and Cohen, D. I. (1988). Deletion of the human T-cell receptor 6 gene by a site specific recombination. Nature 335, 170-174. Dialynas, D. P., Wilde, D. B., Marrack, P., Pierres, A,, Wall, K., Havran, W., Otten, G., Loken, M. R., Pierres, M., Kappler, J., and Fitch, F. W. (1983). Characterization of the murine antigenic determinant designate L3T4, recognized by monoclonal antibody GK1.5: Expression of L3T4 by functional T cell clones appears to correlate primarily with class I1 MHC antigen reactivity Zmtminol. Reo. 74, 29-56. Dim, P., Cado, D., and Winoto, A. (1994). A locus control region in the T cell receptor d 6 locus. lmtnunity 1, 207-217. DiSanto, J. P., Guy Grand, D., Fisher, A., and Tarakhovsky, A. (1996). Critical role for the common cytokine receptor y chain in intrathymic and peripheral T cell selection. /. Exp. Med. 183, 1111-1118. DiSanto, J. P., Muller. W., Guy Grand, D., Fischer, A,, and Rajewsky, K. (1995). Lymphoid development in mice with a targeted deletion of the interleukin 2 receptor y chain. Proc. Natl. Acad. Sci. USA 92,377-381. Distelhorst, C. W., and Dnbyak, G. (1998). Role of calcium in glucocorticosteroid-induced apoptosis of thymocytes and lymphoma cells: Resurrection of old theories by new findings. Blood 91,731-734.
aplyS I J J E A G E COMMITMENT
67
D~idley,E. C., Girardi, M., Owen, M. J., and Hayday, A. C. (199*5).Cup and y6 T cells can share a late coininon precursor. Cicrr B i d 5, 659-669. Dudley, E. C., Petrie, H. T., Shah, L. M., Owen, M . I., and Hayday, A. C. (1994). T cell receptor j3 chain gene rearrangement and selection during thyrnocyte development in adult mice. Inininnity 1, 83-93. Egerton, M., Scollay,R., and Shortman, K. (1990).The kinetics of mature T cell development in the thyirms. Proc. Natl. Acarl. Sci. IJSA 90, 10340-10:344. Ehlich, A . , Martin, V., Miiller, W., and Rajrwsky, K. (1994).Analysis of the B-cell progenitor compartment at the level of single cells. Cwr. B i d . 4, 573-583. Elliott, J. F., Rock, E. P., Patten. P. A., Da\ls, M. M.. and Chien, Y. H. (1988). The adult T-cell receptor &chain is diverse and distinct from that of fetal thymocytes. Nature 331, 627-631. Falk. I., Levelt, C. N., and Eichmann, K. (1993). Lineage relationships ofthe fetal thyniocyte subset that expresses the chain of the interleukn-2 receptor. Eur. /. Itnnzrrnol. 23,33733376. Farr, A., Hosier, S., Nelson, A,, Itoliara, S., and Tonegawa, S . (1990). Distribution of thymocytes expressing y6 receptors in the murine thymus during development. /. Zmrnu1101. 144, 492-498. Fehling, H. J., Iritani, B. M., Krotkova, A , , Forhush, K. A,, Laplace, C . , Perlmutter, R. M., and von Boehmer, H. (1997). Restoration of thymopoiesis in pTa-'- mice by anti-CD3e antibody treatment or with transgenes encoding activated Ick or tailless pTa. Immrinity 6, 703-714. Fehling, H. J.. Krotkova, A,, Saint Ruf, C., and von Boehiner, H. (1995). Crucial role of the pre-T-cell receptor 01 gene in development of cup but not y6 T cells. Nrtttire 375, 795-798. Fehling, H. J., and von Boehmer, H. (1997). Early a@T cell development in the thymus of normal and genetically manipulated mice. Curr. Opi88. Zrntiuind. 9, 263-275. Fenton. R. G., Marrack. P., Kappler, J. W., Kanagawa, O., and Seidman, J. G. (1988). Isotypic exclusion of y6 T cell receptors in transgenic mice bearing a rearranged P-chain gene. Science 241, 1089-1092. Fisher, A. G.. and Ceredig, R. (1991). yfi T cells expressing CD8 or CD4'""' appear early i ~ ~1323-1328. ~. in miirine foetal thymus development. Int. Z t r i ~ i i ~ r t3, Fowlkes, 8 . J., Edison, L., Mathieson. B. 1.. and Chused, T. M. (1985). Early T lymphocytes: Differentiation in cioo of adult intrathyniic precursor cells. /. Exp. Med. 162, 802-822. Fowlkes, B. J., Kruisbeek, A. M., Ton-That, I I . , Weston. M. A,, Coligan, J. E., Schwartz, K. H., and Pardoll, D. M. (1987). A novel population of T-cell receptor @-bearing thymocytes which predominantly expresses a single VP gene family. Nature 329,251 -254. Fowlkes, B. J., and Pardoll, D. M. (1989).Molecular and cellular events of T cell development. Ado. Zniniunol. 44, 207-264. Garman, R. D.. Doheih., P. 1.. and Raulet, D. II. (1986). Diversity, rearrangement, and expression of murine T cell y genes. Ccl1 45, 733-742. Ghia. P., ten Boekcl, E., Saw. E., de la Hera, A., Rolink, A., and Melchers, F. (1996). Ordering of Iiuman bone inarrow B Iyinpliocyte precursors by single-cell polymerase chain reaction analyses of the rearrangenient status of the iinmrinoglobuliii H and L chain gene loci. /. E x p . Aled 184, 2217-2229. Gilfillan, S., Benoist, C., and Mathis D. (1995). Mice lacking terminal deoxynucleotidyl transferase: Adult inice with a fetal antigen receptor repertoire. Ini?ntcnol Reo. 148, 201-219. Godfrey, D. I., Kennedy. J., Mombaerts. P., Tonegawa, S., and Zhtnik, A. (1994). Onset of TCR-P gene rearrangement and role of TCR-P expression during CD3-CD4-CD8thymocyte differentiation. /. Imnuiriol. 152, 4783-4792.
68
HANS J 6 R G FEHLING P t (I/
Godfrey, D. I., Keniiedy, J., Suda, T., and Zlotnik, A. (1993). A developmental pathway involving four phenotypically and functionally distinct subsets of CD3-CD4-CD8- triplenegative adult inouse thymocytes defined by CD44 and CD25 expression. J. Irr~n~rrnol. 150, 4244-4252. Godfrey. D. I., and Zlotnik, A. (1993).Control points in early T cell development. Ztttmicnol. Torlny 14,547-553. Gotlfrey, D. I., Zlotnik, A , , antl Suda, T. (1992). Phenotypic and functional characterization of c-kit expression during intrathymic T cell development. J. brmiunol. 149, 2281-2285. Goodman, T., and Lefrancois, L,(1988). Expression of the y8 T-cell receptor on intestinal CD8+ intraepithelial lymphocytes. Nntrrre 333, 855-858. Greenwald, I., and Ruhin, G. M. (1992). Making a difference: The role of cell-cell interactions in establishing separate identities for equivalent cells. Cell 68, 271-281. Guy-Grand, D., Cerf-Bensussan, N . , Malissen, B., Malassis-Seris, M., Briottet, C., and Vassalli, P. (1991). Two gut intraepithelial CD8+ lymphocyte populations with different T cell receptors: A role for the gut in T cell differentiation. J. Exp. Med. 173, 471-481. I-Iaas,W., Pereira, P., arid Tonegawa, S. (1993).ylS cells.Anriri. Reo. Zn~rriunol.11,637-685. Haas, W., and Tonegawa, S. (1992). Developmcint and selection of yS T cells. Cnrr. Opiti. rrrrtrLulloi.4, 147-15s. Hardy, R. R., Carmack, C. E., Li, Y. S., and Haytkawa, K. (1994). Distinctive developmental Reo. 137, 91-118. origins and specificities of murine CDS’ B cells. Zn~n~unol. Hasserjian, R. P., Aster, J. C., Davi, F., Weinberg, D. S., and Sklar, J. (1996). Motlulated expression of Notch 1 during thymocyte development. Blond 88, 970-976. Havran, W. L., and Allison,J. P. (1988). Developmentally ordered appearance of thyniocytes expressing different T cell antigen receptors. Nuttire 335, 443-445. IIe, Y. W.. and Malek, T. R. (1996). Interleukin-7 receptor a is essential for the development of y8’ T cells, hut not natural killer cells. /. Exp. Med. 184, 289-293. Hedrick. S. M., Cohen, D. I.. Nielsen, E. A,. and Davis, M. M. (1984). Isolation of cDNA clonc~sencoding T cehpecific niembrane-associated proteins. Nuttire 308, 149- 153. Hedrick, S. M., and Eidelinan, F. J. (1993).Lyinphocyte antigen receptors. I n “Fundamental Iininunology” (W. E. Paul. ed.), pp. 383-420. Raven Press, New York. Heilig, J. S.. and Tonegawa. S. (1987).T-cell y gene is allelically hut not isotypicallyexcludcd antl is not required in hiown functional T-cell subsets. Proc. N d . Acod. Sci. USA 84, 8070-8074. Heybome. K., Fn,Y.-X., Kalataradi, €1.. Reardon, C., Roark, C.. Eyster, C., Volliiier, M., Born, W., and O’Brian, R. (1993). Evidence that inurine Vy5 and Vy6 yS T cell receptor+ lymphocytes are derived from a coniinon distinct lineage. /. Zmmzinol. 151, 4523-4527. IIockett, R. D., de Villartay, J. P.. Pollock, K., Poplack, D. G., Cohen, D. I.. and Korsineyer, S. J. (1988).Human T-cell antigen receptor (TCR) 8-chain locus and elements responsible for its deletion are within the TCR a-chain locus. Proc. Nntl. Acnd. Sci. USA 85, 96949698. Hockett, R. D.. Nunez, G., and Korsineyer, S. J. (1989). Evolutionary comparison of inurine and human 8 T-cell receptor deleting elements. New Bid. 1, 266-274. Hoffman, E. S., Passoni, L., Cromptori, T., Leu, T. M., Schatz, D. G., Koff, A , , Owen, M. J.. and Hayday, A. C. (1996). Productive T-cell receptor @-chainrearrangement: Coincident regulation of cell cycle and colonality during development iti oioo. Genes Deo. 10,948-962. Hosseinzadeh, 11.. and Golschneider, I. (1993). Recent thymic einigrants in the rat express a unique antigenic phenotype and undergo post-thymic maturation in peripheral lymphoid tissues. /. Zrnrritmnl, 150, 1670-1678. Howe, R. C., and MacDonald, H. R. (1988). Heterogeneity of immature (Lyt-2-/L3T4-! thymocytes: Identification of four major phenotypically &tinct subsets differing in cell cycle status and in uitro activatioii requirements. J . Ztnttiimd. 140, 1047-1055.
Iezzi. G . ,Kai;jalainen, K.. antl Lanzavecchia,A. (1998).The duration of antigenic stimulation chmiiines the fate of naive a r i d effwtor T cells. Ittittiiittif!y 8, 89-95. 1kut;i. K., Kina, T., MacNcil, I., Uchitla. N., Pearilt. B., Cliien. Y.-H.. and \Veissinan. I. L. ( 1890). A developmental switch in thymic lymplioc>+(~ ni;itnnition potrnti;il occiirs at tlie levd of Iieinatopoietic strm cells. C d l 62, 86.3-874. Ishida, I., Verbeek. S.. Bonne\ille, M., I t d i w i , S.,Uerns, A., ;ind Tonrgawi. S. (1990).Tcell receptor y6 and y transgenic mice sriggest a role of a y gene sileiiccr i n tlie geiirration of ap T cells. Proc. Natl. Acrrrl. Sci. l i S A 87, :30(i7-3071. Itohara. S.. Momhaerts, P.. Lafailk, J.. Iacoinini. J., Nelson, A , , Cliirkc. A. K., Hoopcr. M. I>., Farr, A.. and Toiiegawa. S. (1993).T cell rweptor 6grne mntant inice: Independcnt generation of c.0 T ct4s and prograinined rearrangemriits of y6 TCH gcnes. Ccll 72, 337-348. Ivanov,\I.. and Ceredig, R. (1992).Transcription factors in nioiise fetal thymus tltvelopinent. Z n f . r t t l t ~ l l i t l o l .4, 729-737. Ivanov. V.,Merkenschlager, M.. iintl Cercdig. R. (199:3).Aritioxitlant trc~atnientof thymic o r p i cultures decreases NF-kappa H ;tnd TCFl ( a )transcriptioii factor acti\ities antl inhibits a@T cell rlevelopinerit.1.Z i t t t t a r i i o l . 151, 1694-4704. Iwisliima, M., Green. A,, Boiiyhadi. M. Davis, M . M., Allison, J. l',, antl Cliicw. Y.-11. ( 1991). Expression of fetal y6 T-cell rtwptor in iidult niicc triggvrs a noil-hlI IC:-linketl form of selective tlepletion. hit. I t t m i i t t o l . 3, :38%393. Jacobs, H., \'andeputte. D., Tolkainp, L.. tic, Vries, E.. I3nrst. J.. and Hems. A. (1991).CD:3 components at the surface of pro-T cells ciin inrdi:itc pre-T cell clevekipment it1 O i V O . Eur. J , I n m i r t i o l . 24, 934-939. Janossy, G., Tidman. Y., Selhy. \V, S., Tlionias, J. A,, Granger, S.,Kung. P. C . .iind Goltlstc~in, G. (1980).Huinan T 1yiipliocytt.s of indiicer ant1 siippressor pycl occupy tliffcwnt microenvironinents. h'~tirrc288, 81-81. Janowski, K. M.. Ledbetter. S., M a y , M. S.. and Iiockett, K. D. (1997). Identification of a DNA segment rlihibitiiig rearrangemei~tniotlif>ingeffects iipm traiisgcmic 6-tleletillg , 186, 91-100. elenients. J . E . Y ~hlcd. Joterpmi, F., Heuze, F., S;ilonion-Vie, \'., and Giscari, II. ( 1987). Cell kiwtics in the fetal inonse thyniris: Precursor cell input, proliferation, and eniignition. J . Z t t i t t t t r t i o l . 138, 1026- 1030. Jotereau, F. \I.. antl Le Douarin. N . M. (1982). Drnionstration of ii cyclic renewul of t h Iyinplioc>?e precnrsor cells i n the qiiail tllymlis duriiig enil)r).onic antl perinatal life. /. ~ t t t t t t u t i o l . 129, 1869-1877. M'oods, A. S.., Yewtlell, J. \V.. Bcnnink. J. H., Dcs Silva, A. D., Boesteanu. A., , P., Cotter, R. J.. arid Bi-iitkie\vicz,I{. R. (1998). Natural ligand of iiioiise CDldl: r glyccisylpliospliatitlylinositol. Scioicc~279, 1.541- 1541. King. J., Baker. J.. d Raiikt. I). H. ( 1 W S ) . Evitleiicts that protluctivcs rearrangrnicwts of TCR y genes infliieiicc. the conlniitment of progcwitor cc~llsto diffi.rentiilte into ap o r yS T cells. Eirr. J . f t t t t t t i i t i o l . 25, 2706-2709. Kang, 1.. and Karilet, D. 13. ( 1997). E\wits that rcAgrilutc, differentiation of ap TcH' and y 6 TcR' T cells frorii a coniiiioii p r c w n w r . S m i t t . r t w i i t t w ~ 9, . 171-179 Kelly. K , A,, Pearsr, hl., Lc>francois,I,.. and Scollay, 1%. (1993). Einigtxtiol S the addt I I I O I I S ~thynins. I t i f . r t t i t i i i m d . 5, :3O1sthsets of y 6 T C C ~ froin Kersh, G,J., l3oosliin;iild. F. F., and Iletkick, S. M. (1995). Efficieiit niatu l i n e q tliymocytes t o the CD4' CD8' stage in the ahsencr of TCK-@ rc,arrangeiiiriit. 1. r t t ~ t t ~ t t t ~ 154, ~ ) ! . ~706-5711. K e p i . U.. Beck-Engeser. G. B.. Jongstra, J.. Applvcpiist. S. E.. and Jack, H. M . (1YOij). light chain-tlepenclent selectioii o f Ig lieit\,? chain V rcsgions. J . Z t t r t t i t r r r o l .
70
HANS JOHC: FEHLINC: el cil.
Kisielow, P., Blnthniann, II., Staerz, U. D., Steinmetz, M., and von Boehmer, H. (1988). Tolerance in T-cell-receptor transgenic mice involves deletion of nonmature CD4'8' thymocytes. Nnttire 333, 742-746. Kisielow, P., Hirst, J. A,, Shiku, H., Beverley, P., Hoffman, M. K., Boyse, E. A,, and Oettgen, H. F. (1975). Ly antigens as markers for functionally distinct subpopulations of thymus derived lymphocytes of the mouse. Nature 253, 219-221. Kisielow, P., and von Boehmer, H. (1995). Development and selection of T cells: Facts and puzzles. Ado. Inirr~urud.58, 87-209. Koop, B. F., Wilson, R. K., Wang, K., Veniooij, B., Zallwer, D., Kuo. C. L., Seto, D., Toda, M., and Hood, L. (1992).Organization, stnicture, and function of95 kh of DNA spanning the inurine T-cell receptor C d C S region. Geiioinics 13, 1209-1230. Kotzin, B. L., Barr, V. L., and Palmer, E. (1985). A large deletion within the T-cell receptor p-chain gene complex in New Zealand white mice. Science 229, 167-171. Koyasu, S., Clayton, L. K., Heiken, H., Parkes, A., and Reinhertz, E. L. (1997). PreTCR signaling components trigger transcriptional activation of a rearranged TCR a gene locris and silencing of pre-TCR a locus: Implications for intrathymic differentiation. Int. ItnnmTICJ~. 9, 1475-1480. Kranz, D. M., Saito, H., Heller, M., Takagaki, Y., Haas, W.. Eisen, H. N., and Tonegawa, S. (1985). Limited diversity of the rearranged T-cell y gene. Nature 313, 752-755. Krotkova, A,, von Boehmer, H.. and Fehling, H. J. (1997).Allelic exclusion in pTa-deficient mice: No evidence for cell surface expression of two T cell receptor (TCR)-P chains, but less efficient inhibition of endogeneous Vp+( D)Jp rearrangements in the presence of a functional TCR-/3 transgene. /. Exp. Med 186, 767-775. Lalli, E., Sassone-Corsi, P., and Ceredig, R. (19%). Block of T lymphocyte dfferentiation by activation of the CAMP-dependentsignal transduction pathway. EMBO J . 15,528-537. Lauzurica, P.. and Krangel, M. S. (1994). Temporal and lineage-specific control of T cell receptor dS gene rearrangement by T cell receptor a and 6 enhancers. J . Ex?. Med. 179, 1913-1921. Leclercy, G., De Smedt, M., and Plum, J. (l9ga5).Cytokine dependence of Vy3 thymocytes: Matiire but not immature Vy3 cells require eiidogenous IL-2 and IL-7 to survive-evidence for cytokine redundancy. Int. I i n i m m d , 7, 84:3-851. Leclercq, G.. Plum, J., Nandi, D., De Smedt, M., and Allison, J. P. (1993). Intrathymic differentiation of Vy3 T cells. /. E x p . Med. 178, 309-315. Lefrancois, L., and Goodman, T. (1989). I n uioo modulation of cytolytic activity and Thy1 expression in TCR-yS' intraepithelial lymphocytes. Science 243, 1716-1718. Levek C. N., Ehrfeld, A., and Eichmann, K. (1993).Regulation of thytnocyte development through CD3. I. Timepoint of ligation of CD:3&determines clonal deletion or induction of developmental program. J . Exp. Merl. 177, 707-716. Li, H.. Lebedeva. M. I., Llera, A. S., Fields, B. A., Brenner, M . B., and Mariuzza, R. A. (1998). Structiire of the VS domain of a human y6 T-cell antigen receptor. Nature 391,502-506. Li, S., and Wilkinson, M. F. (1998). Nonseiise surveillance in lymphocytes? Immunity 8, 135-141. Livik, F., Petrie, 11. T., Crispe, I . N.. and Schatz, D. G. (1995). In-frame TCR S gene rearrangements play a critical role in the a,f3/y6 T cell lineage decision. Zmmi~nity2, 617-627. Livik, F., Wilsoii, A., MacDonald, H. R., and Schatz, D. G. (1997). Cup lineage-committed thymocytes can be rescued by the y6 T cell receptor (TCR) in the absence of the TCRP chain. Eiir. J. I ? 7 1 i n U i l O ~ .27, 2948-2958.
Liiffert. D., Ehlicli. A.. Miiller, LV..i t n d Kajewsky. K. ( 15196).Surrogate light chain expression is rcqnired to estal)lish iiiiiiiiiiioe;lot)iiliii hea~?.chain allrlic excliision diiring early R cell developnieirt. Zttttttiitiit!/ 4, 1 3 - 1-14. Ixwc*nthal. J. \V.. Howe, K . C . , Cercdig, R.. iintl hl;icDonald. 11. I<. (1986). Fiinctional iniiiiatiilr (Lyt-2 /I;3T4 ) thyntc~cytes. status of intcdnikin 2 reccptors expressed J . I t t i i t t i i t i d . 137, 2579-2584. L!~icIi, F., and Shevach, E. M . (1993). y6 T cells promote CD4 and CDri expression hy SCID thymocytes. I t i t . I i t t t t i i i r i o l . 5, 991-99.5. Macllonald, H. H. (1995). NK1.l’ T cdl receptor-d/3’ cc~lls:Kew cliies to their origin, spc,cificsity, antl function. J . Esp Jfcd 182, 63:3-6:38. MacDonakl, H. K.. Casunova,J. L., Maryanski. J. L.. and Cerottini, J. C . (1993). Oligoclonal cxpansioir o f niiijor liistocompatit)ilih~complex class I-restricted cy-tolytic T lymphocytes during ii priman) ininnine response iti L;ico: Ilirwt monitoring by flow cytonictry and po1ymcrase cllain reaction. J . Esp. j \ f d 177, 1487-1392. Maki, K.. Sunuga. S., and Iknta. K. (1996a). Thc \’-J recom1)ination o f T cell receptor-y genes is hlocked in interlenkin-7 receptor-deficirnt mice. J . Esp. h f i d 184, 2423-2427. Maki. K., Sunuga, S.. Kornagata, Y..Kodaira. Y..Mahuchi. A , , Karasuyiina. H., Yokomuro, K.. Miyazaki. J. 1.. and Ikiita, K. (1996b). Interleiikin 7 rrceptor-tl(.ficieirt inice lack y6 T cells. Pt-or. N d . Actid. Sci. USA 93, 7172-7177, Malissen, M., Trucy. J., Joiivin-Marche. E.. C;mrnave. 1.’ A , , Scollay, H., and Malissen, B. (1992). Kegulation ofa antl /3 gene allelic rxpression dining T-cell tlevelopnient. Z t w t t i m o l . T o d q 13, 315-322. Mallick, C. A., Dndley, E. C., Viney, J, I,.. Owc~n,M. J., and Il;iytlny, A. C. (1993). Hearrangeinent itnd diversity of T crll receptor /3 chain gent's in thynioc>-tes:A critical role for the /3 chain in developnient. C d l 73, 513-519. Mallick-\l’ood, C. A.. Lewis, J. M.. Hichie, L. I . , Owen, M. J., Tigelaar, R. E., antl Hayday, A . C. (1998). Consenfation of T cell rcceptor conforinatitin in epidermal y8 cells with disrupted primary Vy gene usage. Sciriirr 279, 1729- 1733. Matsiizaki, Y., Cyotokn, J., Ogawa. M.. Nishikawa, S..Katsiira. Y.,Gaclielin, G., and Nakanchi. 13. (1993). Chariicteriz&nr of (,-kit positi\zcs iiitrathyinic stem cells that are restricted to lymphoid cliffwentiation. 1. E x p M i d . 178, 1283-1292. McHeyzer-Willianis, M . G.. and DaLis. M. M. (1995). Antigen-specific developinmt of priniary and ~iicmoryT cells iti ciuo. Scii,ttcc, 268, 106-111, McPhee, D., Pyc. S., and Shortman, K. (1979). The diffcrcwtiation of T lyniphocyte. V. Evitle~icefor intrathymi death of most tliymoc~tes.T / i ! / i t i i i , s 1, 151-lfil. Mendiratta, S. K., M:irtin, W.D., Hong. S., Borsteann, A,, Joyce. S.. and Van Kaer. L. (1997). CDltll niut;int mice are deficient i n iratriral T cells that promptly prodlice 11,4.Z r t t t t t ~ i t i i f ! /6, 469-477. Merk(mchlager, M ., Graff, D., Lovatt, M., Boinmhardy, U., Zamoyska, R., and Fisher. A. G.(1997). How inany thynioc)ites ariditiori for se1ection.J. Erp, Mcd. 186, 1149-1 158. Mrrtsching, E., and Ceredig, R. ( 1996).T cc~llrrceptor-y6-c!xprc.ssing fetal mouse thymocy-tcs are generated withoiit T cell receptor \;B scalection. Eirr. J . Z t i i i t t i i ~ i o Z .26, 804-810. Mertsching, E., \Vilson, ‘4.. MacDonaltl, 1-1. R.,md Ceredig, R. (1997). T cell receptor (Y genc rearrangeinent a i d transcription i n adult thymic y6 cells. Eicr. J . Z t t t i t i t i i i o l . 27, 389-396. Mrtcalf. D. (1966).“The TIiyniris: Its Role in Immnne R~~spoiises. Leukacniia Development antl (:arcirrogeiirsis,” Vol. 5 . Springer-\’erlag. I Ieidellierg. Michie. A. M.. Carlyle, J. R.. antl Zuniga-Pfhcker, J. C. (1998). Early intrathyniic precursor phenotype. 1. Zttimutiok 160, 1735- 1741. Miller, J. F. A. P. ( 1961).1inniunologic;tl fiinction of tlie thyinus. Laticd 2, 748.
72
HANS JORC FEHLING at
ni.
Miller, J. F. A. P. (1979). Experimental thymology has come of age. Thymus 1, 3-25. Moisan, J. P., Bonneville, M ., Bonyge, I., Moreau, J. F ., Soulillou, J. P., and Lefranc, M. P. (1989).Characterization of T-cell-receptor y (TRG) gene rearrangements in alloreactive T-cell clones. H u m . Imrtiund. 24, 95-1 10. Mombaerts. P., Anderson, S. J., Perlmntter, R. M., Mak, T. W., and Tonegawa, S. (1994). A n activated Ick transgene promotes thymocje development in RAG-1 mutant mice. Zinrnicnity I, 261-267. Moinbaerts, P., Anioldi, J., Russ. F., Tonegawa, S., and Kaufmann, S. H. (1993). Different roles of ap and y6 T cells in immunity against an intracellular bacterial pathogen. Nature 365, 53-56. Mombaerts. P., Clarke, A. R., Rudnicki, M. A,, lacomini, J., Itohara, S., Lafaille, J. J., Wang, L., Ichikawa, Y., Jaenisch, R., Hooper, M. L., and Tonegawa, S. (1992). Mutations in Tcell antigen receptor genes a and 8 block thymocyte development at different stages. Nature 360, 225-231. Moore, T. A,. von Freeden-Jeffry, U., Murray, R., and Zlotnik, A. (1996). Inhibition of y6 T cell development and early thymocyte maturation in IL-7-’- mice. J. Immunol. 157, 2366-2373. Muegge, K., Vila, M. P., and Durum, S. K. (1993). Interlenkin-7: A cofactor for V(D)J rearrangement of the T cell receptor ,8 gene. Science 261,93-9-5. Nakajima, P. B., Menetski, J. P., Roth, D. B., Gellert, M.. and Bosma, M. J. (1995). V-D-J rearrangements at the T cell receptor 6 locus in mouse thymocytes of the a8 lineage. Iiiimunity 3, 609-621. Ohashi, P. S., Wallace, V. A,, Broughton, H., Ohashi, C. T., Ferrick, D. A,, Jost, V., Mak, T. W., Hengartner, H., and Pircher, H. (1990). Specific deletion of the J-C 6 locus in murine a8 T cell clones and studies using transgenic mice. Eur. 1.In~munol.20,517-522. Ohteki, T., Ho, S., Suzuki, H., Mak, T. W., and Ohashi, P. (1997).Role of IL-15IIL-15p chain in natural killer 1.1‘ T cell receptor-a8 cell development. /. Zmniunol. 159,5931-5935. Osmond, D.G., Rolink, A,, and Melchers, F. (1998). Murine B lymphopoiesis: Towards a unified model. Initnunol. T o d q 19,65-68. Pardoll, D. M., Fowlkes, B. J., Bluestone, J. A,, Kruisbeek, A,, Maloy, W. L., Coligan, J. E., and Schwartz, R. H. (1987). Differential expression of two distinct T-cell receptors during thymocyte development. Nriture 326, 79-81. Passoni, L.,Hoffman, E. S., Kim, S., Cromptori. T., Pao, W., Dong, M.-Q., Owen, M. J., and Hayclay, A. C. (1997).Intrathymic 6 selection events in y6cell development. Zrnrnnnity 7, 83-95. Peschon, J. J . , Morrissey, P. J., Grabstein, K. € I . , Ramsdell, F. J., Maraskovsky, E., Gtiniak, B. C., Park, L. S., Siegler, S. F., Williams, D. E., Ware, C. B., Meyer, J. D., and Davidson, B. L. (1994). Early lymphocyte expansion is severely impaired in interleukin 7 receptordeficient mice. /. Exp. Med. 180, 1955-1960. Petrie, H. T., Livak, F., Burtrnm, D., and M a d . S. (1995).T cell receptor gene recomhination patterns and mechanisms: Cel death, rescue, and T cell production. /. ESP. Med. 182, 121-127. Petrie, H. T., Scollay, R., and Shortman, K. (1992). Commitment to the T cell receptora/3 or -y6 lineages can occur just prior to the onset of CD4 and CD8 expression among immature thymocytes. Eur. J . bnniunol. 22, 2185-2188. Philpott. K. L., Viney, J. L., Kay, G., Rastan, S.. Gardiner, E. M., Chae, S., Hayday, A. C., and Owen. M. J. (1992). Lymphoid development in mice congenitally lacking T cell receptor @-expressing cells. Science 256, 1448-1452. Plum, J.. De Smedt, M., and kclercq, G. (1993). Exogenous IL-7 promotes the growth of CD3- CD4-CD8-CD44+CD2St’-precursor cells and blocks the differentiation pathway of TCH-a8 cells in fetal thymus organ culture. /. Zmmunol. 150, 2706-2716.
aPlyG I J N E A G E COMMITMENT
73
Pmit, J. A.. Hawaii. CV., Abe. K., Sarin, A.. iind Singer, A. (1997).T cell receptor (TCR)iiidiiccd death of immature C:D4’CD8 + thymocytes by two distinct Incchanisnis differing i i i their requireinerit foi- CD28 costiniiilutioii: Iiiiplications for negative selection in the thymus. /. Esp. Alc~d.186, 1911-1922. Rast, J. P.. Anderson, M. K.. Strong, S. J., Liier. C., Litniari, R. T., and Litman, G. W. (1997). a, /3. y , ant1 6 T cell antigen rtsceptor gent’s arose eiirly in vertebrate phylogeny. Z t t i t r i u t t i f ~ /6, 1 1 1. Raiilet. D. II. (1989). The strncture, function and niolecrilar genetics of the $6 T cell receptor. Annrr. Rcc. Z i ~ ~ t ~ i r ~ t 7, a o l 175-207. . Raiilet, D. H., G;iriiian, R. D.. Saito, H., and Tonegawa. S. (1985). Developmental regulation of T-cell receptor gene expression. Notnrc’ 314, 103-107. Raillet. D. H.. Spencer, D. M., Hsiang. Y. H., Goltlinan. 1. l’,, Bix, M., Liao. N. S., Zijstra, M., Jaenisch, R., and Correa, I . (1991). Control of y6 T-crll developnient. Znaiiunol. RCD. 120, 185-204. Reif. A. E.. and Alleri. J. M. V. (1964). The A K K thymic antigen arid its distribution in Iriikemias and nervous tissiie. 1.Erp. M c d . 120, 413-433. Reilly, E. B.. Kranz, D. M., Tonegawa. S.. antl Eiscw, H. N. (1986). A functional y genc formed from known y-gene segments is not riecessay for antigen-specific responses of iniirine cytotoxic T Ijiirphocytes. Nrifirre 321, 878-880. Reinherz, E. L., Kung, P. C., Goldstein. G.. Lev It. H., antl Schlossinan, S. F. (1980). Discrete stages of Iiiiman intrathymic tlifferenti on: Analysis of norinal thyinocytes and liirkeinic lymphoblasts of T-cell lineage. Proc. N d . Arod. S c i USA 77, 1.588- 1592. Reinlierz, E. L.. and Schlossinan. S. F. (1980).The differentiation and fiinctions of human T cells. C d 19, 821-827. Robcy, E., Chang, D., Itano. A.. Cado. I)., Alesantler, I I . , Lans, D., Weininaster, G.,and Salmon, P. (1996). An activated forni of Notch influences the choice between CD4 and CDH T cell lineages. Cell 87, 483-492. Robey, E., and Fowlkes, B. J. (1998). The a@vcrsns y6 T cell lineage choice. Cicrr. @in. z t t l ~ t ~ l r l l o10, ~ . 181- 187. Rodewald, H. H., antl Fehling, 11. J. (1998).Molecnlar mid celliilar events in early thymocytc development. Ark. f t t i t t i u t t r d . 69, ( i n press). Rodewald, H.-K., :ind Haller, C. (1998). Antigen receptor junctional diversity i n growth factor rcccptor deficient mice. DPC. Cortip Z n i t t i i i r i d . ( i n press). Rolink, A,, Crawvnnder. U., \Vinkler, T. H.. Kar2tsriyama, H., and Melchers, F. (1994). IL2 recrlptor a chain (CD2.5.TAC) exprcwioii defines a cnicial stage in pre-B cell development. Znt. Z t t i t t i ~ i t i o l .6, 125-1264. Kosat, J. P., MacDonaltl, H . R., and Louis, J. A. (1993). A role for y6’ T cells during o ~ . 550-555. experimental infection of mice with Lrishiitmio ttu!jrn-./. ~ t n t t i i ~ t i 150, Rothenberg, E. \’., and Ward, S. B. (1996). A dyiiamic assembly of &verse transcription factors integrates activation and cell-tye iiiforirintion for interleukiri 2 gene regulation. Proc. Notl. Acntl. Sci. USA 93, 9358-9365. Davis, M. M., Bluestone, J. A , . Schiltl, H., Mavaddat, N., Litzenberger. C;,, Ehricli. E. W.. Matis, L., Draper, R. K., ; i d Chien. Y. 11. (1994).The nature of major histocompatil)ility complex recognition by y6T cells. Cell 76, 29-37. Schleussner. C., Fisher, A.. and Ceredig, R. (19Y 1 ). Ciiltiire conditions dictate whether iiioiise fetal thymus lobes generate predomiiimtly yl6 or d/3T cells. Crrrr. TOIJ.Microbial. f t r ~ i t ~ r t t ~ 173, o l . 25-28. Schleiissner, C., Kuzio. S., Schultze. N.. Burdet, C., Fisher, A. G.. and Ceredig, R. (1992). T cell receptor repertoire of y6 cells generated from the 14-day embryonic mouse thymiis. Th/trins 20, 195-216. -
Schorle, H., Holtsclike, T., Hunig, T., Schiinpl, A , , and Horak, I . (1991). Development and fiinction of T cells in inice rendered interleiikin-2 deficient by gene targeting. Nature 352, 621-624. Schweighoffer, E., and Fowlkes, B. J. (1996). Positive selection is not required for thymic maturation of transgenic y6 T cells. /. Exp. M d . 183, 2033-2041. Sekaly, R. P., Cerdig, R., arid MacDonald, H. R. I, 1983).Generation of thyrnocyte srlhpopulations in organ culture: Correlated analysis of Lyt-2 phenotype and cell cycle status by flow microfluorometry./. Ittittiunol. 131, 1085-1089. Shores. E. W., Sharrow, S. O., Uppenkamp, 1.. and Singer, A. (1990). T cell receptornegative thymocytes from SCID inice can he induced to enter the CDUCD8 differentiation pathway. E w . /. Itntt~utiol.20, 69-77. Shortman, K., and Jackson, H. (1974). The differentiation of lymphocyte I. Proliferation kinetics and interelationships of subpopillations of mouse thymus cells. Cell. ItrL?nrinol. 12, 230-246. Shortman, K., Wu, L., Kelly, K. A,, and Scollay, R. (1991). The beginning arid the end of the development of TCR y6 cells in the thyinus. Ciwr. Top. Microbial. Itt~tnut~ol. 173, 71-80. Shutter, J., Cain, J. A,, Ldhetter, S., Rogers, M. D., and Hockett, R. D. (199.5). A 6 Tcell receptor deleting element transgenic reporter construct is rearranged in ap but not y6 T-cell lineages. Mu/. Cell. Biol. 15, 7022-7031. Sim, G. K., Olsson, C.. and Augustin, A. (1995).Commitment and maintenance of the ap and y6 T cell lineages. J. ItnnLtino/. 154, 5821-5831. Simon, T. C., and Gordon, J. I. (1995). Intestinal epithelial cell differentiation: New insights from mice, flies and nematodes. Curr. Opin. Genet. Dez;. 5, 577-586. Simpson, P. (1995). Developmental gcnetics: The Notch connection. Nottire 375,736-737. Skeen, M . J., and Zieglrr, H. K. (1993). Induction of niiirine peritoneal yl6 T cells and their role in resistance to bacterial infection. /. Exp. Med. 178, 971-984. Snodgrass, H. R., Kisielow, P., Kiefer, M., Steinmetz, M., and von Boehmer, H. (1985). Ontogeny of the T-cell antigen receptor within the thymus. Nature 313, 592-595. Spanopoulou, E., Gigliere, V., and Grosveld. F. (1991). The functiond domains of the inurine Thy-1 gene promoter. Mol. Cell. B i d 11, 2216-2228. Spits, H., Paliard, X.. and Vandekerckhove, Y., van Vla lar, P., and de Vries, J. E. (1991). Functional differences between CD4' and CD4- T cell receptor-y6clones from peripheral blood. /. Zitiniunol. 147, 1180-1188. Suda. T., and Zlotnik, A . (1993). Origin, differentiation, and repertoire selection of CD3' CD4-CDK thymocytes hearing either a/? or y S T cell receptors. /. Inivminol. 150, 447-455. Swain, S. L. (1980).Association of Lyt phenowes, T cell function and MHC recognition. Fed. Proc. 39, 3110-3113. Swiatek, P. J., Lindsell, C . E., del Amo, F. F., Weinmaster, G., and Gridley, T. (1994). Notch 1 is essential for postimplantation development i n mice. Genes DEV.8, 707-719. Takeshita, S., Toda, M., and Yamagishi, €1. (1989). Excision products of the T cell receptor gene support a progressive rearrangement model of the cd6 locus. E M B O ] . 8,3261 -3270. ten Boekel, E., Melchers, F.. and Rolink, A. (1995). The status of Ig rearrangements in single cells from different stages of B cell development. Int. Z~ntnuno/.7, 1013-1019. ten Boekel, E., Melchers, F., and Rolink, A. G. (1997).Changes in the V(H) gene repertoire of developing precursor B lymphocytes in inouse bone marrow mediated by the pre-B cell receptor: Zmtnrrnity 7, 357-368. ten Boekel, E., Melchers, F., and Rolink. A. G. (1998). Precursor B cell showing H chain allelic inclusion display allelic exclusion at the level of pre-B cell receptor surface eqiression. Ztrrrtiunity 8, 199-207.
Thoiiipson. S. D.. Manzo. A. R., Prlkoneii. J., Larclie, M.. and IIuwitz, J. L. (1991). Dc~vrlopiiientalT cell rcwptor gene rearrangements: relatedness of the d p and ylS precursor. Eirr. J . Z t t ~ t t i i i t i o ~21, . 19:3Y-1950. Thompson, S. D.. Pelkonen. J., and Hunvitz, J. L. (1990). Concoiiiitant T-cell receptor a a i d 8 genv I-rarraiigemciits in indix itha1 T-ct.11 prrciirsors. Proc. Notl. Accirl. Sci. USA 87, 5583-5586. Toda, M.,Fujinioto. S., Iwasato. T., Takeslii w h i , T.. and Y;imagislii, H. (1988). Structure of t.xtracIiroiiiosoii ed from T-cell a i l t i p 1. l?io/, 202, 219-231. J. 1). ( 1998).Tli!~niocy-&t~ glucorticoid resistance alters on and inhibits autoitiiiiiuni~anti IynipIioprbIiferative disease i n MHL-
D. F., and Spreiit, J . (1998).Lift-span of ylS T cells. /. Exp. Mcrl. 187, 357-365, Toiiiigny. M . R., Maicl, S., Burtruiii. 11. €3.. m c l Petrie, H. T. (1997). T cell receptor (TC:H)-Pgeiic, recoiii1,iiiatioii: Dissociatioii from ccll cyclc regulation and develop~nental pr(igression during T cell ontogeiiy. J Exp. i\kd 185, 1,549-1556. Trnnnecker, A., Oliveri, F., Allcw. N., a i d Karjalainen. K. ( 1986).Nornial T cell tlevelopment is possible without “timctional” y chain gcwes. E:iIUO /. 5, 1589-lt5Y3. Ucliida, N., Fleming, W. H.. Alpern. E. J., ;tiid \\’eissman. I . L. (1993). Heterogeneity of heinatopoietic stein cells. Ciirr. Opitt Z i ? t t ) t i / t i o / . 5, 177- 184. van Mecnvijk.J. P.. l3liithinann. H., mid Steinirietz. M. (1990).T-cell specific rearrangement of T-cell receptor p transgenes in illice. E M B 0 J . 9, 10t57-1062, Verlieek. S., Izon, D., Hofliuis, F., Hobaiiiis-Maantlag,E.. t r Riele, H., van de M’rteritig, M., Oostenvegel. M., Wilson. A,, MacDonald, 11. R., and Clevers, H. (1995).An HMGbox-containiiig T-cell factor required for tlrvniocee differentiation. Nrttrirr 374, 70-74. Vicari, A. P., Mocci, S., Openshaw, P., O’Garra, A , , Zlotnik, A. (1996). Mouse y 6 TCR N K 1.1* thvmocytes specifically produce iiiterlwkiii-4, iire inajor liist~~coliipatibilih m p k x class I independent. m t l are tlevelopinentally related to ap TCR+NKl.l+thynioZes. Eicr. 1. Zttiitwnol. 26, 1424- 14%. ri, A. P., a i d Zlotnik, A. (1996). hlouse N K I . I + T cells: A nvw fanlily of T cells. Z i t 1 t i w t w l . Torlny 17, 71-76, von Uoehmer. H., Boniieville, M . , Ishitla. I.. Hyser. S., Lincoln, G., Smith, H. T., Kishi, H., Scott. B., Kisielow, P., ;nid Toiiegawa. S. (1988). Early expression of a T-cell receptor pcliaiii transprne slippresses re;irrangc*nient of the Vy4 p e seginmt. Proc. Nutl. Acnrl. Sci. USA 85, Y729-9732. voii Boclitner, H., a i d Fellling. H. J. ( lY97). Stnictrtre and function of the pre-T cell receptor. Atirir~.Hrt. f t i ~ t i ~ i t t i o f15, . 432-4S2. Wang, J.-H., Niclio~~innopoiiloii, A , . \\’II, I,.. S I I I I ,I,.. Sharpe. A. H., Bigby, M., antl Georgopoulos. K. (19’36). Selectivr clefects in thc tl~~velopment of the fetal a i d adult lyinpliojd systeiri in inice w i t h ;III Ik;ii-os nu11 inlitation. Zir~tuztnity5, 537-549. Waslil)urn, T.. Schweiglioffer. E., Gritllry, T., Chiig. D.. Fowlkes, B. J., Cado, D., antl Hobep, E. (1997).Notch activity influc~iicestlie c@ versiis yS T cell lineage decision. Cell 88, 833-543. \.Yeininaster, G., Rol)c.its, 1’. J., antl 1,eirike. C. i 19911. A homolog of Dro.vophilrz Notch expressed duiiiig niamnialiim tlevchl)nieiit. Dcdopttitwt 113, 199-205. \Veil, L.. Barbrr. D. F.. Pao, LV., Ll’ong, F. S.,O w c ~ iM. . J., and Hayday, A. (1998). Primary y6 crll clones ciin be defined pheiiot$c~dly and fuiictionally as Tlil/Tli2 cells and illustrate 160,1965- 1974. the association of CD4 d i Th2 tlifferenti:ition. J . Zrtitn~it~ol. M’illiains, C;. T., Kingston, R.. Owcn, M. J.. Jcnkinson. E. J., aiid Owen, J. J. T. (1986).A siiigle inicroiiiatiipulatcd stem cell gives rise to niultiplr T-cell receptor gene rearraiigeiiirnts i n thc thymus i t i L;itro. Nuturi>324, 63-64. Toil$.
+
Williarns, I . R., Rawson, E. A,, Manning, L.. Karaoli, T., Rich, B. E., and Kupper. T. S. (1997).IL-7 overexpression in transgenic iiiouse keratinocytes causes a lymphoproliferative skin disease dominated by intermediate TCR cells: Evidence for a hierarchy in IL-7 responsiveness among ciitaneous T cells. 1. Itn.niunol. 159, 3044-3056. Wilson, A,, de V i h t a y , J. P., and MacDondd, H. R. (1996). T cell receptor 6 gene rearrangement and T early a (TEA) expression in immature a/3 lineage thymocytes: Iinplications for aP/yS lineage coinmitment. Iniiiiunity 4, 37-45. Winoto, A. (1991). Regulation of the early stages of T-cell development. Cirrr. Opin, rnLmulloi. 3, 199-203. Winoto, A,, and Baltimore. D. (1989a). Separate lineages o f T cells expressing the a/3 and yS receptors. Nutrrre 338, 430-432. Winoto, A,, and Baltimore, D. (1989h). Cyp lirieage-specific expression of' the a T cell receptor gene by nearby silencers. Cell 59, 649-655. Wu, L., Antica, M., Johnson, G . R., Scollay, R., and Shortman, K. (1991). Devekpnentdl potential of the earliest precursor cells from the adult monse thymus. 1. Exp. Med. 174, 1617-1627. Wu, L., Scollay, R., Egerton, M., Pearse, M., Spangrude, C. J., and Shortman, K. (1991). CD4 expressed on earliest T-lineage precursor cells in the adult murine thymus. Nuturc 349, 71-74. Yoshiinoto, T., and Paul, W. E. (1994).CD4+,NKl.lt T cells promptly produce interleukin 4 in response to in uiuo challenge with anti-CDd. 1. Exp. A4ecl. 179, 1285-1285. Zuniga-Pflucker,J. C., Jiang, D., Schwartzberg, P. L., and Lenardo, M. J. (1994). Sublethal y-radiation induces differentiation of CD4-/CD8- into CD4'/CD8' thymocytes without T cell receptor /3 rearrangements in recombinase activation gene 2+ mice. 1.Exp. h k d . 180, 1517-1521. This article was accepted for pulilication 011 June 1, 1998.
lmmunoregulatory Functions of y6 T Cells Wllll BORN, CAROL CADY, JESSICA JONES-CARSON, AKIKO MUKASA,' MICHAEL LAHN, AND REBECCA OBRIEN Department of f m m u n o l w , UCHSC, Denver, Colorado 80206; and 8righam and Women's Hospital, Harwrd Medical School, Boston, Massachusetts 02146
I. Introduction In the decade after their discovely in the mid-1980s (Bank et al., 1986; Born et a l , 1987; Brenner r t n l , 1986; Chien rt (11, 1987a; Pardoll et a1 , 1987; Saito et d . , 1984; Sowder ct d., 19881, lymphocytes bearing y6 Tcell receptors ( y 6 T cells) have become recognized as regular constituents of higher immune systems (Haas et al., 1993). However, the process of integrating these cells into a coherent picture of the immune response is still in its early stages. A major problem is that we have only the vaguest ideas about what ligands y6 T-cell receptors (TCR) recognize (Born et nl., 199Ob; Chien et al., 1996; DeLibero, 1997; Janeway et al., 1988; Kronenberg, 1994; McMenamin et al., 1994; Raulet, 1994; Wildner et al., 1996). Moreover, experimental modeling did not reveal a clearly defined contribution of y6 cells to that quintessential of the immune functions, the defense against pathogens. With a few exceptions, manipulations of y6 T cells in zjivo have caused merely sinall changes in host resistance, often only noticeable in immune-compromised animals or under conditions of very high pathogenic load. Nonetheless, in some cases experimental y6 T-cell perturbation has led to dramatic changes in the immune response itself, including both the innate and the adaptive arms of the immune system. Such findings gave rise to the concept that y6 T cells are important immunoregulators, controlling both innate and antigen-specific adaptive immune responses (Carding et nl., 1990; D'Souza et d.,1997; Ferrick et nl., 1995; Fu et al., 1994b; Hsieh et al., 1996; McMenainin et a l , 1994; Mengel et nl., 1995), an idea that is consistent with the small numbers of y6 T cells in primates and rodents, their relative inability to recogize and respond to conventional antigens, as compared to a$ T cells and B cells, and their ability to produce large quantities of regulatory mediators rapidly. Therefore, this review focuses on summarizing and interpreting findings which, taken together, support the idea that one role of y6 T cells lies in the regulation of the immune response. However, while immunoregulation may be an important function of y6 T cells, it almost certainly is not their only one. The potential of y6 T cells to exert immune effector functions is evident from studies in vitro. Moreover, as is the case with other cells
-i I
( opnglit 0 1999 h\ A c & i n ~ All n g l h 01 reprrrdiiction i n m v lonn rt. lX)hi 277WJY $31) 00
of the immune system, the biological role ofy6 T cells likely reaches beyond the immune responses. Examples of nonimmune y6 T-cell functions are also included here. y6 T cells have been studied extensively in both mice and humans. This review of y6 T-cell functions, due to the personal bias of the authors, focuses mostly on inurine models and is therefore far from being comprehensive. Moreover, both mice and humans are equipped with coniparatively small populations of y6 T cells. One might predict that in sheep and cattle or some of the other species with larger and more diverse populations, y6 T cells occupy broader and perhaps entirely different functional niches (Cooper et uZ., 1989b; Hein and Mackay, 1991). 11. Origin, Lineage and Development, and Distribution
Current evidence supports the notion that all rearranging genes associated with the adaptive immune response arose early in vertebrate pliylogeny. Whereas attempts to identify such genes in the agnathans, protochordates and invertebrates have failed, both immunoglobulin and TCR genes have been found in the phylogenetically most primitive living jawed vertebrates, the cartilaginous fishes. Four distinct clones of TCR genes homologous to mammalian TCR a!, p, y and 6 have been found in Raja eglanteria, a cartilaginous fish representative of a phyIogenetic lineage that diverged -450 million years ago from a common ancestor with the mammals (Rast et d., 1997).As in mammals, TCR a! and 6 homologs exhibit more sequence identity with each other than with either TCR y and p. This suggests that the two types of genes arose through duplication of a proto-d6 locus, a mechanism also consistent with their current interspersed form of genomic organization in mammalian species (Chien et al., 1987a). It thus appears that the three major known classes of rearranging antigen receptors, including TCR y6, were already present in the common ancestor of presentday vertebrates, more than 450 million years ago. However, contrary to occasional suggestions that y6 T cells are more “primitive” than a@T cells (Herzenberg, 1989), there is no evidence that yp T cells are phylogenetically older than Cup T cells. In present-day mammals, y6 and a!p T-cell lineages have segregated by the time the TCRs appear on the cell surface (Pardoll et al., 1987). Cells siinultaneously bearing both TCRs or mixed TCRs (pa or ya) do not normally arise (Saito et al., 1988), although mixed receptors may occur in some tumor cells (Davodeau et al., l993b; Hochstenbach and Brenner, 1989). Moreover, Cup and y6 T cells differ in their expression of the MHC coreceptors CD4 and CD8. Most thymic and peripheral y6 T cells in mice and humans are CD4-CD8-. A minority of y6 T cells do express CD4,
I h l ~ l U N O I ~ E G l ~ L A T O FRUYN C T I O N S OF yS T CELLS
79
although CD8 is more coiiiinon, especially among certain subsets such as inteTtina1 y6 T cells 111 mice. Finallv, sizes ofy6 and 00 T-cell populations and their relative distributions appear to be independently regulated, at least for the most part (Chrbone clt 0 1 , 1991). When a0 T cells are genetically ablated as is the case in mice lacking fiinctional TCR genes (Monibaei-ts r t nl., 1992),y6 T-cell numbers inav increase but they do not assume the place or diytribution of rr/3 T cells (Mombaerts et (11, 1994). The basis for the lineage segregation between the two T-cell subsets is not yet clear. Current models for this event are essentially divided into those proposing that TCR generation is critical lineage-determining event, and those proposing that lineage cominitmcnt occurs independently of TCR generation. A model known as the “successivegene rearrangement model” is based on the finding that TCH y6 cxpression occurs before TCR expression during fetal thymic ontogeny. According to the model, precursor T cells first attempt to rearrange y and 6 genes to generate a y6 T cell. If they fail, gene rearrangements are initiated and cells with productive 0 gene rearrangements go on to rearrange a to become c.0 T cells (Pardoll et d . , 1987).A similar model suggests that precursor cells rearrange 6, y , and 0 genes roughly at the same time, an assumption supported by studies with fetal thymus hybridomas (Born ct ml., 1985, 1986; Chien ct al., 1987b; Livak et nE, 1995), and that the Fate of the precursor cells depends on their rearrangement status, and perhaps competition between products of functionally rearranged genes. In both models, it is tacitly assumed that fiinctional expression of a TCR heterodiiner results in fate determination of the uncoininitted presursor. The “competitioii models” envisage that successful TCR 0 rearrangement is wfficient for crp T lineage cornmitment whereas both fiinctional y and 6 chain expression is required for y6 lineage commitnrent, a condition favoring crp T lineage differentiation based on probabilities of successful gene rearrangements. However, the possibility that T-cell lineage commitment occurs already prior to TCR gene rearrangements has not been excluded. Several mechanisms have been proposed: that 6 rearrangement is inhibited in 00-committed progenitor cells (Winoto and Baltimore, 198Yb),that the 6 locus is inactivated in af3-conrmitted cells by recombination of the a,, element (DeVillartay et al., 1988), that y genes are silenced transcriptionally in 00-committed thynrocytes (Bonneville rt crl , 1989a; Ishida et al., 1990), and that TCR (Y gene rearrangement and/or expression is inhibited in y6-committed thymocytes (Diaz et al., 1994; Winoto and Baltimore, 1989a). A detailed discussion of these possibilities lias been published (Kang and Raulet,
1997). In both mice and huinans, the first y6 T cells appear before a0 T cells arise in thymus and periphery. The subsets arise in ‘‘wives” at different
80
WILL1 BORN et 01.
times in ontogeny (Havran and Allison, 1988). The “waves” may be controlled independently as targeted deletion of one gene, Vy6, did not perturb expression of the others (Sunaga et al., 1997). The subsets seem to vary in the degree of their thymus dependence (LefranFois, 1991). As they colonize the various tissues, they tend to segregate according to their distinct TCR repertoires. In rodents, these can vary from essentially a single receptor (Asamow et al., 1988) to diverse collections of related receptors (O’Brien et al., 1992).This pattern of tissue-specific TCR distribution early on led to some speculation about the nature of the ligands for y6 TCRs (Janeway, 1988; Janeway et al., 1988). However, there is as yet only scarce evidence for tissue-specific ligands for y6 T cells (Havran et al., 1991; Heybome et al., 1994; O’Brien et al., 1989; Spaner et al., 1995). The development of y6 T-cell subsets has been studied most extensively in mice, but ontogenetic waves and tissue-specific segregation of subsets are also evident in other species, including humans (Bucy et al., 1991; Ishiguro et at., 1993; Machugh et al., 1997; Porcelli et al., 1991).Although it seems probable that y6 T-cell subsets are functionallyspecialized, presently only a few differences in surface markers, profiles of cytokine secretion, and in subset-specific response patterns are known in support of this notion (Bergstresser et a!., 1983; Boismenu and Havran, 1994; Goodman and Lefranqois, 1988, 1989; Groh et al., 1989; Lefraqois and Goodman, 1989; Morita et al.,, 1991; Tschachler et al., 1983). In mice, the first wave of y6 T cells expresses an invariant TCR composed of V75 and V61 (Asamow et al., 1988).These cells appear around day 14 in the fetal thymus and later exclusively colonize the epidermis. Here, they predominate among all lymphocytes, including ap T cells and B cells. Reconstitution experiments revealed that the thymic progenitors of these cells are only present during fetal development and that the fetal thymus environment is required for their maturation (Ikuta et al., 1990). A couple of days later, a second y6 T-cell subset appears in the murine thymus, also expressing an invariant TCR. These cells express a TCR 6 chain identical to that of the first subset but paired with a TCR y chain composed of a different Vy, Vy6. In normal healthy mice, they colonize the mucosal epithelia of the tongue, vagina, uterus, and lung (Hayes et al., 1996; Itohara et al., 1990; Siin et al., 1994). Their appearance is strictly dependent on the presence of IL-7 (Hayes et al., 1996; He and Malek, 1996; Maki et al., 1996; Maki et al., 1996) a cytokine found to be required to a much lesser extent for the developinent of crj3 T cells and B ceIls, but not at all for natural killer (NK) cells (He and Malek, 1996). Interleulan (1L)-7 appears to be required for gene rearrangements at the TCR y locus (Maki et al., 1996), as well as for the subsequent development and expansion of y6 T-cell populations. Unlike Vy5’ cells, Vy6’ cells are not confined to
the tissues they colonize during ontogeny. In pregnancy, they arise, in large numbers and often activated, in the placenta (Heybonie ct nl., 1992). In diseases associated with strong inflaininatory responses, they appear at the sites of inflammation, including the liver. and testis (Mukasa et nl., 1997; Roark et nl., 1996) where they are normally infrequent or even absent. Around the tiine in ontogeny when Vy6’ cells arise, y6 T cells expressing more diverse TCRs begin to appear as well. Among these are cells that express LJy4 together with several different VS genes (Bluestone et nl., 1991; Itohara et al., 1989;Takagaki ef d . ,1989b).Vy4’ cells circulate, appear in the various lymphoid tissues, arid have been found in the lung 1989; Reardon et al., and the lactating niaininary gland (Augustin et d., 1990). Vy4’ cells often express 1’65, and ii yS TCR composed of this V gene combination has been identified that shows specificity for the TL region encoded molecule, T22” (Ito et d ,1990). There is no indication that recognition of T22” requires bound antigen, and the T22”-reactive yp T-cell clone appears to be autoreactive (Bonneville et nl., 198913; Cooper et nl., 1989a; Ito et al., 1990). y6 T cells expressing Vyl arise during late fetal through adult stages of development (O’Brien et nl., 1989, 1992). They express Vyl paired with several V6 genes, usually V6(i or V&, but also including V62, 3, 5 , 7, and 8. Vyl’ cells have been found in the newborn mouse thymus, newborn spleen, intestines, adult spleen and liver, placenta, and, as rare clones, in skin and the lactating maminary gland (Happ et nl., 1989; Nagler-Anderson et nl., 1992; O’Brien et d , 1989, 1992; Reardon et nl., 1990, Roberts et nl., 1991).In the skin, cells expressing Vyl transcripts can become predominant when the thymus-dependent Vy5/V6lt cells are absent, as is the case in mice carrymg the “nude” mutation (Ota ct nl., 1992). The natural ligaiids of Vyl’ cells are not known, but hybridomas expressing these TCRs have been found to respond to a variety of stimuli (potential ligands), including bacteria and bacterial extracts, 60-kDa heat shock protein, small peptides, polyCluTyr, and other polyanionic substances (Born et al., 1990a; Fu et nl., 1993, 1994a; Happ et d., 1989; O’Rrien et d., 1W9, 1992: unpublished data). In addition, they seem to be stimulated by ligands expressed on the hybridonias themselves and on trophoblasts (Ezquerra et nl., 1992; Heyborne et nl., 1994; Nagler-Anderson et nl., 1992; O’Hrien et a/., 1989; Roberts et nl., 1991). The biological significance of these responses is not clear. 7 6 T cells expressing Vy7 tend to express highly diversified TCRs composed of Vy7 paired with VS4, 5 , 6, or 7 . These cells primarily colonize epithelia of the intestines but have also been found in the liver (Asarnow et al., 1989; Itohara et al., 1990; Tsuji et nl., 1996). They differ from most other y6 T cells by their expression of CD8a homodirners and rather variable levels of surface Thy-1, depending on cellular activation (Goodman
and LefranGois, 1988; LefranGois and Goodman, 1989). Because of this and because a large portion of 76 T cells in the liver also express C D 8 a homodimers, it has been suggested that the liver cells belong to the intestinal lineage (Haas and Tonegawa, 1992). Experimental evidence suggests that Vy7' cells can develop independently of the thymus (Bandeira et nl., 1991), but our own data indicate that at least a portion of these cells normally derive from thymic emigrants (K. Kelly, unpublished data). In the adult mouse, most tissues harbor Y S T cells that belong to several of the Vy-defined subsets, although cells of one type often predominate (Augustin et nl., 1989; Bluestone et (11, 1991; Takagak et al., 1989a). Populations expressing invariant TCRs such as the epidermal yS T cells expressing Vy5/V61 and those of the reproductive tract expressing Vy6/ V61 have been found in mice. Moreover, based on Vy gene expression patterns, the two subsets appear to be lineage related (Heyborne et nE., 1993). No equivalent populations have been noted in other species, with the possible exception of the rat (Elbe et al., 1996; Kinebuchi et nl., 1994; Kuhnlein et d.,1994). In humans, a specifically epithelia-associated y6 T-cell subpopulation has been identified in the gut (Deusch et n l , 1991; Soderstroin et nl., 1996). In human blood, two major yS T-cell populations coexist (Porcelli et nl., 1991). The relative frequencies of these subsets change with age. In normal individuals, Vy9/V62 expressing cells increase from approximately 25% of all y6 T cells in umbilical cord blood to >70% in adult peripheral blood (Parker et nl., 1990). In contrast, cells expressing Vy9/ V a l are more frequent very early in development and continually decrease thereafter. More than differential Vy gene expression, the sequential appearance of particular Vy-Jy or VS-DS-JS combinations suggests that, as in the mouqe, human y6 T cells arise in a developmentally ordered fashion (Vietor and Koning, 1990). In addition, both clones and entire subsets of 76 T cells may peripherally expand driven by antigenic stimula1991,1992). tion (De Liberoet al., 1991;Parkeret al., 1990; Uyeinuraet d., Despite similarities in the development of peripheral y6 T-cell populations, enormous differences exist between species regarding the tissue distributions and overall sizes of y6 T-cell subsets (reviewed in Born et n l , 1994). For example, y6 T cells are comparable in numbers to a@T cells in sheep, cattle, pigs, and chickens, despite being rather infrequent in primates and rodents. The biological significance of these differences is not yet understood. If indeed y6 T-cell subsets are functionally specialized (see later), species may vary in their requirements for these functions, e.g., because of their life-styles (Hein and Mackay, 1991) or due to organismic structural properties. It may thus be helpful to consider y6 T-cell subsets within an organism as separate entities and to attempt to identify
and assess their individual functions as an alternative to comparing y6 and ap T cells in a more global way. 111. Specificity
A. IMPI,IC:ATIONS o~ y6 TCR STH~JCTLJHE
The y6 TCR is expressed as a heterodimeric cell surface molecule (Haas et nl., 1993; Raulet, 1989). y and 6 chains each consist of one variable and one constant domain. Cell surface expression of the heterodimer and signaling through the y6 TCR require association with tlie CD3 complex of transmembrane proteins. Thus, tlie overall structure of the y6 TCK complex resembles that of the ap TCR complex far more than that of iiniiiunoglobulins,and the similarity between the two types of TCRs initially suggested that ligands recognized by the two types of T cells may also be similar. A more detailed analysis of y6 TCR structure has revealed further similarities but also substantial differences with a0 TCRs. The genomic organization of the genes encoding TCR y and 6 chains has been described in detail elsewhere (Arden et , 1995; Chien et al., 198%; Clark et nl., 1996; Hayday et al., 1985; Iwasliima et d.,1988; Lefranc et nl., 1989; Lefranc and Rabbits, 1989; Lefranc and Rabbitts, 1985, 1990; Ratlibun et nl., 1988; Raulet, 1989; Zhang ct nl., 1994). Briefly, TCR y genes are clustered at a distinct lociis with considerable variations in genomic organization between the different species. There are only a few Vy genes, each of which typically combines through gene rearrangement only with its most proximal J segment. There are no Dy segments. The V, J, and C gene segments form small clusters. In mice, for example, there are four such clusters and seven Vy-J-C combinations possible within these clusters. Vy-J-C combinations between clusters do occur but are rare. Moreover, the TCR y repertoire is additionally limitcd as junctional variations in rearranged genes are not extensive, and in productive rearrangements involving Vy5 and VyG, are typically absent. These two invariant TCR y chains in mice each form heterodimeric receptors with the same, equally invariant, TCR 6 chain. This lack of diversity in a portion of the y6 TCRs iinplies that ligands recognized by the y6 T-cell populations carrylng these receptors are invariant as well. TCRS and a gene loci are interspersed such that productive TCR a gene rearrangements eliminate most or all gene segments of the TCR 6 cluster, wlierea? TCR 6 rearrangements per se do not prevent subsequent TCR a rearrangements (Cliien ut a/., 1987a,b; Porcelli et al., 1991). This same peculiar genomic organization persists in distant species (e.g., mice and humans), suggestive of evolutionary conservation and functional impor-
84
WI1,L.I BORN rt nl
tance. An obvious consequence of this arrangement is that y6 T cells are far less likely to express TCR a chains, and, inversely, that a@T cells are prevented from expressing TCR S. There are far fewer V6 genes than V a genes (e.g., approximately 10 times fewer in mice.) The V6 genes are interspersed with V a genes, and some are closely related to or even identical with V a genes. Assuming that y6 TCRs make contact with their ligands in a fashion similar to immunoglobulins or a/3 TCRs (Engel and Hedrick, 1988; Jorgensen et al., 1992), i.e., via the three CDR-equivalent loops present in both y and S TCR chains, yS TCRs containing V6's that are also Va's should have similar or identical specificities for two of the six predicted points of ligand contact. The similarities in CDRl and CDR2 equivalents could indicate a structural basis for a similar ligand bias of y6 T cells and a@T cells, e.g., favoring recognition of MHC class I/IIlike molecules. Among the CDR loops, CDRS is of particular interest. For immunoglobulins, X-ray crystallography has shown that CDRS loops of both heavy and light chains are typically involved in antigen contact (Engel and Hedrick, 1988;Jorgensen et al., 1992). In the case of a/3TCRs recognizing peptide/ MHC complexes, the CDRS loops tend to make contact with the bound peptides. Davis and Bjorkman (1988) estimated that the potential number of different y6 TCRs in mice actually exceeds that of a/3TCRs or immunoglobulins. This is primarily due to an enormous potential for junctional diversity within the TCR 6 gene locus. Assuming that CDR3 is also critical for antigen recognition in y6 T cells, Rock et nl. (1994) cornpared CDRS length distributions of the various types of antigen receptors in mice and humans. CDRS lengths were defined as the distance from the J regionencoded Gly-X-Gly motif to the nearest preceding V region-encoded Cys, minus four amino acids. Among transcripts of IgH, -L(Kand A) and TCR a, @, 7 , and 6, CDR3s of IgH and TCR 6 were indeed the longest and most variable in size. In contrast, CDRS size variations of TCR a and @ are much more constrained, with almost indentical CDR3 lengths. In Ig heavy chains, CDR3 loops are often extensive with a wide range of lengths, whereas CDR3 loops of light chains tend to be much shorter with little variations in length. It is unclear whether this difference has any particular significance, but the great variability of heavy chain junctions coincides with the ability of antibodies to recognize structurally diverse antigens. Much in contrast, CDR3 loops of TCR a and p chains are typically similar in lengths. This likely reflects the fact that the size of the ligands for a/3 T cells, which are small peptides nestled within a peptidebinding groove, is almost invariant. Moreover, it suggests that that CDRS loops of TCR a and /3 chains play similar roles, consistent with evidence that both can make contact with the bound peptide fragments (Fremont
Ih.IMlII\’OHEGUI,4TORY FIINCTIOVS OF y6 T CELLS
85
et nl., 1996). Chien and cohhorators (1996) have pointed out that tlie heterogeneity of CDR3 loops in y6 TCRs is more consistent with ligand recognition in a manner similar to that of iiniiiunoglobulins than of cr,P TCRs. This new concept is further supported by their observations of the distinctive ligand requirements of two MHC-reactive y6 T-cell clones (see later).
B. SELECTION OF y6 T CELI.S Most peripheral crp T cells have gone through a screening process that promotes the development of self-MHC-restricted ligand specificities, while preventing the survival of potentially autoaggressive clones. This balancing act is accomplished through positive and negative selection mechanisms that favor cells with relatively low but distinct affinities for self-MHC class 1/11 molecules, but having the potential of much higher affinities for self-MHC-antigen complexes (Robey and Fowlkes, 1994). Because the CDR3 regions of the a@TCR are primarily involved in ligand binding (Engel and Hedrick, 1988; Jorgensen et al., 1992), they are also the primary targets of selection. Nevertheless, the survival of any given cell may also depend on the TCR V genes or even on the accessory molecules it expresses. Because CDR3 sequences of y6 T-cell subsets tend to be nonrandom, it made sense to postulate that the peripheral repertoire of y6 TCRs is shaped by selection as well. However, the generation of particular junctional sequences could also be driven by genetic mechanisms and be entirely independent of TCR-ligand interactions. In this regard, the invariant junctional sequences of the murine VyS/V61 and Vy6NG1-positive subsets have been examined in some detail. Normal development of the Vy5’ subset requires a fetal thymus environment, consistent with the possibility of thymic selection (Ikuta et nl., 1990). Moreover, whereas productive rearrangements of Vy5, Vy6, and V61 genes in fetal thymocytes are essentially invariant,junctional diversity has been seen in nonproductive rearrangements that are not subject to selectional forces ( Itoliara and Tonegawa, 1990; Lafaille et nl., 1989). Finally, when thymic expression of the y6 TCR was prevented by modulation with anti-TCR mAbs, frequencies of productive rearrangements of tlie same genes with noncanonical sequences increased (Lafaille et d . ,1990).These findings are all consistent with thymic selection of the two invariant y6 T-cell subsets in mice. However, other findings indicate that thymic selection is not necessary in the generation of tlie invariant ~6 TCRs. In mice into which nonfunctional TCR y gene substrates were introduced as transgenes, canonical TCR Vy5-Jyl and Vy6-Jyl junctions were generated at high frequencies, even though there was no possibility of selection for certain surface-expressed
86
\VILLI BORN ef d
protein products (Asamow ct al., 1993). Similarly, in mice lacking TCR 6, Vy5-Jyl and Vy6-Jyl canonical junctions are generated in normal frequencies (Itohara et nl., 1993). More extensive experimentation with transgenic mice has emphasized the importance of short homology repeats near these gene junctions and a lack of terminal deoxynucleotidyl transferase (TdT) in the generation of the invariant y6 TCRs (Zhang et al., 1995). Specifically, transgenic recombination substrates revealed that di- and trinucleotide repeats in the coding regions and in P elements have strong effects on the site of recombination. In addition, forced expression of TdT at early developmental stages decreased the frequencies of canonical junctions while increasing the frequency of in-frame noncanonicaljunctions containing N nucleotides. These data seem to indicate that early in development, a directional mechanism of rearrangement, aided by the absence of TdT activity, gves rise to the invariant y6 TCRs in mice (Allison and Havran, 1991; Raulet et al., 1991). Bias in the expressed TCR repertoire has also been noticed among other inurine y6 T-cell subsets. For example, differences between mouse strains in the proportions of V&+ cells among splenic and intestinal yS populations have been reported (LefranGois et al., 1990; Sperling et al., 1992). Among intestinal y6 T-cell subsets, a dominant V&'" phenotype was found to be linked genetically to certain MHC class I1 alleles, expressed in the periphery but not in the thymus (LefranGois t>t al., 1990). In pulmonary y6 T cells of BALB/c mice (but not in BALB/b or C57BLlG mice), characteristic TCR junctional sequences (referred to as BID in rearrangements involving V65, and GxYS in rearrangements involving Vy4) are present in extraordinarily high frequencies (Sim, 1995; Sim and Augustin, 1991a,b).The same junctional sequences were present in mice carrying the nude mutation on a BALB background, suggesting a relatively thymus-independent mechanism. Moreover, although C57BW6 mice lack BID or GxYS sequences in the periphery, they are able to generate them because they could be found in the thymus. The presence of BID arid GxYS cannot be explained by an absence of TdT activity nor have genetic mechanisms been identified that could be responsible for these sequences. However, a correlation between the presence of an endogenous retrovirus and BID sequences was noted (Sim and Augustin, 1993). Therefore, BID and perhaps also GxYS may represent examples of extrathyinic junctional selection of y6 T cells. If y6 T cells are indeed selected, are they subject to similar mechanisms as those operational for a@T cells and, more specifically, are the selecting ligands similar? In mice lachng µglobulin (P,-M), CD8' c.0 T cells fail to develop, presumably because they are not positively selected in the absence of MHC class I expression (Zijlstra et al., 1990). However, the same mice showed normal distributions of y6 T cells in thymus, peripheral
lymphoid tissues, and intraepithelial locations (Koller et d., 1990; Zijlstra ct nl., 1990). Neither abnormalities in Vy or V6 gene usage were noted nor were y6 T cells dysfunctional when stimulated with anti-TCR mAbs. Similarly, normal nuiiibers of splenic y6 T cells were found in double mutant mice lacking both class I and class I1 MHC expression (Grusby et nl., 1993). Nevertheless, in inice expressing transgenic y6 TCRs with specificities for class Ib molecules encoded in the Tla region, PL-Mexpression was required for the normal development of transgene-expressing y6 T cells (Pereira ct n l , 1992; Wells et al., 1991, 1993). In the absence of P,-M, thymic transgene’ y6 T cells exhibited an immature phenotype (HSA’ ), and peripheral y6 T cells were reduced in numbers. Similarly, in mice expressing transgenic class Ib, changes in the peripheral y6 TCR repertoire have been noted. Thus, while the majority of y6 T cells in mice s e e m to develop normally in the absence of MHC class I/II-dependent selection, a minority requires these ligands for their development. This pattern fits well with the observation that in contrast to a0 T cells, specificities for allogeneic MHC among y6 T cells are rare (Bux et nl., 1985; O’Brien et nl., 1989). Nevertheless, the finding that some y6 T cells are selected in a MHC-dependent fashion probably indicates that y6 T cells as a whole are subject to selectional mechanisms, albeit for the most part not involving M HC-type molecules. Alternatively, ligand selection could be exceptional among 76 T cells. only occurring among those cells whose TCRs happen to have aP T cell-like specificities. y6 T cells that fit these ciiteria include a sinall population ofCD8‘ cells in inice and rats mediating tolerance to certain inhaled antigens and apparently endowed with MHCrestricted peptide antigen specificities (McMenainin ct nl , 1994).
C . y6 T CELLSARE STIMULATBL)H1’ M % N Y D I F F E R EMOLECULES ~T Because immunization with soluble antigens has largely failed to elicit antigen-specificy6 T cells in vivo (for possible exceptions, see the examples discussed under tolerance to ingested and inhaled antigens), y6 T cells liavc been screened in uitro with a variety of antigens in the hope of finding specificities by chance, withoiit prior sensitization in vivo This approach is reminiscent of early attempts to find antigen specificities of myeloma proteins, prior to the availability of hybiidomas. The screening of myeloma proteins in binding assays with large collections of various chemical compounds led to tlie discovery of niimerous hapten specificities ( Janeway and Travers, 1997). Similarly, random stimulation of y6 T cells, clones, and hybridomas with various antigens revealed numerous responses to hoth peptidic and nonpeptidic substances. However, the biological significance of these responses remains unclear at present.
Among partially defined antigens, heat-killed bacteria, bacterial extracts, mycobacterial purified protein derivative (PPD, a partially purified culture supernatant of mycobacteria such as M . tuberculosis H37Rv), low molecular weight protease-resistant components of mycobacterial extracts, and polyGT (pGT, a random copolymer of glutamic acid and tyrosine, molecular weight 20-50,000) were all found to stimulate murine and/or human yS T cell responses in vitro (Dembic and Vidovic, 1990; Holoshitz et al., 1989; Kabelitz et al., 1990; O’Brien et al., 1989; Panchamoorthy et al., 1991; Pfeffer et nl., 1990, 1992). With the exception of pGT, all have been reported to elicit polyclonal yet subset-specific reactivity, in this regard reminiscent of the superantigen responses of ap T cells (Herman et al., 1991). The authors’ unpublished data indicate that pGT also elicits polyclonal reactivity of yS T cells in vitro, in the absence of aj3 T cells and without requirement for in uivo priming. Subset specificity and, in some cases, TCR gene transfection indicate that these responses are indeed y6 TCR dependent, but it has remained unclear whether they involve direct binding interactions between the TCR molecules and the antigens. Molecularly defined soluble antigens have also been found to stimulate y6 T-cell responses in vitro. These include tetanus toxoid (Kozbor et al., 1989), mycobacterial 60-kDa heat shock protein (HSP-60) (Haregewoin et al., 1990; O’Brien et al., 1989) HSP-60-derived peptides (Born et al., 1990a; Fu et d . , 1994a), staphyloccocal enterotoxin A, (Rust et al., 1990) listeriolysin 0 (Guo et al., 1995), and lipopolysaccharides (LPS) (Reardon et al., 1995; Tsuji et al., 1996). More recently, nonpeptidic components, mostly derived from mycobacteria, have been isolated that were found to be stimulatory for human but not murine y6 T cells (Biirk et al., 1995; Constant et al., 1994; Schoel et al., 1994; Tanaka et al., 1994, 1995). Chemical characterization of these molecules has revealed that they are all of low molecular inass and contain phosphate groups. Otherwise, their structures are diverse, ranging from nucleotide derivatives to isoprenyls and sugars. As with the less defined antigens listed earlier, these substances all elicit polyclonal yet subset-specific and TCR-dependent responses of yS T cells. Nevertheless, the mechanisms underlying their stimulatory activities remain unresolved. Responses to phosphate-containing antigens have been studied most extensively. Some of these molecules elicit proliferative and cytokine responses at very low molar concentrations (Tanaka et al., 1995). Indirect stimulation of yS T cells can be excluded, as accessory cells are not strictly required to induce these responses. Stimulation of cell lines and clones occurs very rapidly, making it unlikely that the phosphate-antigens function by inducing the de not100expression of cellular ligands (Lang et nl., 1995). A requirement for cell contact has
been interpreted as evidence for antigen presentation (Morita ct al., 1995). However, professional antigen presenters are not required and there is no evidence for antigen processing. Although phosphate ligands could be antigens in the classical sense, it seems e q u J y possible that they function a s niolecular "adjuvants," enhancing preexisting TCR-clependent cellnlar interactions. The unique susceptibility of y6 T cells to phosphate antigens could be of considerable importance for the selective induction of y6 T cells in the course of an immune response to bacterial and perhaps other pathogens (Burk et al., 1997). There is ample evidence that y6 T cells can function as iminunoregiilators (see later), ancl the presence of phosphate antigens might help determine the flavor of their regulatory activities. However, in the absence of rodent responses to phosphate antigens, suitable model systems in which in vim effect\ could be tested remain to be identified.
D. MECHANISM OF LJGAND R I X O ( , N I T I OOF ~ MHC M ~ I . E ( ' U I , E - S P E ~ .y6 I F JT~ ,CEJ.LS In recent years, nnmerous y 6 T-cell clones have been isolated that respond specifically to classical and nonclassical MHC molecules (Bluestone et d., 1988: Bonnedle ct d., 198%; Houlden et d.,1989; Matis and Bluestone, 1991; Matis e t a / , 1987, 1989; Porcelli et al., 1989; Porcelli and Modlin, 1995; Kellahan et a / . , 1991). Most of these cells were selected by allo-antigen stimulation, but others were not, consistent with an inherent or selected bias for the recognition of MHC moleciiles among y6 T cells (Ito et d., 1990). For two MHC-reactive clones, the mechanism of ligand recognition has been analyzed in considerable detail. These were derived from BALB/c nidnu mice (H-2") stimulated with low-density B1O.BR (H2') spleen cells. One clone, LBKS, recognizes the mouse MHC class I1 inolecules I-EL'" (Matis et a l , 1989). LBK5 differs in V( D)J junctional sequences but not V, D, or J segments from another clone, LKDl , derived from a B 10.BR nioiise immunized with B10.D2 (H-2") cells arid specific for 1-A', suggesting that CDR3 determinants dictate antigen specificity for these clones (Rellahan et d., 1991). Schild and collaborators (1994) demonstrated in an elegant set of experiments that LBKS recognizes I-ELindependently of bound pepticles and that no conventional antigen processing pathway is required for the recognition of this molecule by y8 T cells. Even isolated I-EL protein immobilized on a plastic surface stimulated LBKS to an extent similar to that of cells expressing I-EL,thus eliminating the possibility that an additional molecule, such as a superantigen, could be involved in the stimulation. Moreover, epitope mapping using cells expressing mutated I-E molecules indicated that ap T cells ancl LBKS recognize different regions of
90
WILL1 BORU ct nl
I-Ek.Specifically, mutations in the a helices of I - E a and -P that affected a0 T-cell recognition did not alter LBKS stimulation, whereas a mutation at position 79 of I-Ea abolished the response of LBK5 while not interfering with the responses of any of the I-E-specific (YPT cells tested. In addition, the response of LBK5 was influenced by a polymorphic solvent-exposed residue at position 67 of I-EP and by the carbohydrate at a82 (Chien et al., 1996). Another clone termed G8, also derived from alloantigen-immunized BALB/c ndnu mice (Matis et al., 1987), recognizes both the products of the nonclassical class I genes T10 and T22 (Schild et al., 1994;Weintraub et al., 1994).Again, there was no indication of a requirement for conventional antigen-processing pathways in ligand recognition. G8 recognized T10/ T22 molecules expressed in Drosophila cells, which are considered to be incapable of any type of antigen processing or presentation, and it even responded to T10 expressed in Escherichia coli, after the recombinant molecules were immobilized on plastic. That peptide loading in fact is not important for the recognition of the MHC molecule by G8 was also suggested by the failure to elute peptides from the stimulatory T10 molecule and by the finding that T10 does not require peptides for cell surface expression (Chien et al., 1996). T22 is also recognized by another y6 T-cell clone, K N 6 (Ito et al., 1990). G8 and K N 6 express different y6 TCRs. KN6 was derived without alloantigen selection, suggesting that y6 T cells bearing this particular specificity occur more frequently. The T22 gene was also mutagenized, and transfectomas were tested for their ability to stimulate KN6 responses (Moriwaki et al., 1993). In this study, some mutations located at the floor of the predicted peptide-binding groove reduced KN6 reactivity. The response pattern was interpreted to suggest that bound peptides could play a role in ligand recognition, but the alternative possibility that the mutations altered T22 surface expression was not ruled out. In sum, the recognition of TL molecules by y6 T cells remains somewhat less well defined than that of MHC class I1 because possible bound ligands have not been ruled out as strictly and because aP T cells with specificities for TL are not as readily available for comparisons as those with specificities for MHC class I1 (Chien et al., 1996).
E. LICAND RECOGNITIONOF A HERPES SIMPLEX VIHUS-SPECIFIC y6 T-CELLCLONE Another murine y6 T-cell clone, Tg14.4, was found to respond to a herpes simplex virus type I transmexnbrane glycoprotein, gI ( Johnson et nl., 1992). Several observations indicated that gI is also recognized without processing or presentation (Sciammas et al., 1994): (1) anti-gI antibodies
1h.lMUNOKEC:~lLATORY FUNCTIONS OF 7 8 T CELLS
91
blocked the response of Tg14.4, (2) a mutated form of gI not expressed on the cell surface was not stimulatory, and (3) a form of gI expressed as a cell surface protein in the antigen-processing mutant RMA-s remained stirnulatoiy. The TCR of Tg14.4 is coinposed of rearranged V68 and Vy2 variable genes. Junctional sequences have been determined, but a inutational analysis defining requirenients on the TCR structure for the recognition of HSV gI has not yet been reported. Recognition of HSV and its product(s) by y6 T cells could be of considerable importance in host resistance (see later).
F. REQUIREMENTS FOR y6 T-CEI.LAC.TIVATION CARRY IMPI,ICATIONS FOR SPECIFICITY For the activation of a0 T cells through the TCR, multivalent ligands capa1)le of cross-linking the TCRs are required. Foreign peptide antigens are typically rendered polyvalent through their display on the surfice of antigen-presenting cells. Experimentally, a similar situation can be created using iiniiiobilized anti TCR antibodies. TCR cross-linking induces signaling through TCR-associated transmembrane molecules collectively referred to as the CD3 complex. Like a@ TCRs, y6 TCRs are associated with the CD3 complex and require it both for TCR surface expression and for signaling (Haas et nl., 1993). Moreover, like a0 T cells, y6 T cells are activated following TCR cross-linking. First, this has been shown with anti-TCR antibodies and later using antigenspecific 76 T-cell clones. Thus, using y6 T-cell clones recognizing MHC class I1 (I-EL),MHC class I (TlO/T22), and HSV g1 proteins, it was shown that soluble forins of these proteins are only stiinulatory when iininobilized (Chien et nl., 1996; Schild Pt al., 1994; Sciammas et nl., 1994). Interestingly, requirements for TCR cross-linking also seein to exist with substances found to elicit polyclonal y6 T cell responses in citro. Thus, stirnulatory bacterial extracts, randoni amino acid copolymers, and sinall peptides such as those derived from HSP-60 d l are inherently polyvalent, partially insoluble, or need to bc immobilized in order to elicit y6 Tcell responses (Deinbic and Vidovic, 1990; Fu et nl., 1994a, unpublished observations). Similarly, the inore recently discovered low molecular weight phosphate-containing compounds, which are capable of stimulating polyclonal responses of human y6 T cells but are soluble in free form, require cell contact, and thus probably a primitive form of presentation, in order to be stimulatory (Morita et nl., 1995). The requirement for TCR cross-linking in 76 T-cell activation seems to be an indication that y6 T-cell specificities are norinally trained on ligands expressed on the cell surface where they inherently have cross-linking properties, as opposed to soluble ligands. It is still not clear, however,
whether the ligands are primarily heterologous and complexed with autologous cell surfaces or whether they are autologous and perhaps indicators of activation, stress, or inflammation. There are also some indications that signal processing in y6 T cells differs from that in ap T cells. Thus, at least epidermal y6 T cells in mice may use FceRIy for signal transduction instead of CD35. Signaling has not yet been studies in great detail in 76 T cells. Possible implications of the presence of different signal transducers in the two types of T cells have been discussed elsewhere (Leclercq and Plum, 1996).
G. POTENTIAL USE OF sTCR CONSTRUCTS I N DETERMININC; y6 T-CELLLIGAND SPECIFICITIES Candidate ligands for y6 TCRs still need to be confirmed by measuring binding interactions. If yS TCR ligand binding is Ig like (Chien et al., 1996), there is hope that such studies will be facilitated by higher affinities than those of ap TCR ligand interactions, but this remains to be seen. For this type of experiment, soluble and perhaps polyvalent forms of y6 TCRs will be required, similar perhaps to the engineered multivalent ligands for a/3 TCRs (Altman et ul., 1996). Although y6 TCRs can be isolated directly from cells expressing them, conventional methods yield only sinall quantities of mostly denatured protein (Born et al., 1987). As demonstrated with crp TCRs, coinparatively large amounts can be generated using soluble TCR (sTCR) constructs (Fields et al., 1995). The first y6 sTCR reported was derived from a chimeric construct in which the extracellular domains of the mouse Vyl.1-Cy4 and Vy6.2-C6 TCR chains of y6 T-cell hybridoma T195/BW were fused to the hinge region, CH2 and CH3 domains of human IgGl heaLy chain, and transiently expressed in COS cells (Eilat et d., 1992). The chimeric proteins were produced intracellularly at rather high levels, the hvo protein chains formed disulfatelinked, glycosylated heterodimers, and correctly paired receptor chains were found secreted into the culture medium. In addition to confirming the identity of the chimeric secreted TCR yG-IgH heterodimer with Vyland TCR &specific antibodies, reactivity with an anticlonotypic inAb (F10/ 56) suggested that the fusion protein retained a conformation identical or at least similar to that of the native TCR. This chimeric TCR construct has been used in attempts to identifi. a putative autoantigen recognized by hybridoma T195/BW and similar V y l + cell lines (see later), thus far without success. In a CHO expression system, others have demonstrated efficient secretion of nonchimeric disulfidelinked human y6 TCR by introducing translational termination codons upstream from the sequences encoding TCR chain transmembrane regions (Davodeau et al., 1993a). Based on its reactivity with several anti-y and
-6 mAbs, tlie recoinbinant protein appeared to be folded correctly. It also proved to be immunogenic, dlowing the generation of mAbs capable of recognizing both soluble and ineiiibr;ine-l)oiind, native y6 TCRs. A high sensitivity of the interchain disulfide bridge to digestion with papaiii suggested that the sTCR C-terminal portions were in ii more extended configuration than tlie corresponding regions in iiiiinunoglobiilins. However, the soluble y6 heteroclimer remained stable after removal of tlie interchain disulfide link, suggestive of strong noncovalent forces capable of holding the two chains together. More recently, the V6 doniain of another huinan y6 TCR, derived from a clonc specific. for the HLA-A2 molecule (Palliard et nl., 1989), has been ciystallized (Lebedeva ct nl., 1996). Here, the V doniain of TCR 6 was expressed a s a secreted recombinant protein within the periplasinic space ofE. coli (Studieret al., 1990).It was then crystallized in a form suitable for X-ray diffraction analysis (orthorhoiiibic crystals, space group P21212 with unit cell diinensions (1 = 69.9, 11 = 49.0, c = 61.6 diffraction to beyond 2.3 resolution), but the final result of this analysis still awaits publication. Using a baculo\irus expression system, a soluble nonchiineric form of an iiivariaiit inurine y GTCR, initially identified in the tongue and reprocluctive tract but later alw found in several other tissues (Hayes et nl., 1996; Heybome ot nl., 1992; Itohara et d., 1990) and after bacterial infections (Mukasu ct al., 1997: Roark et nl., 199G),has also been generated (Roark, 1995). Reactivity with anti-TCR mAbs suggests that this soluble heterodiiner is also appropriately folded and therefore may be used to generate antibodies against this particular 76 TCR. The structural analysis of y6 TCRs still lags beliind that of ap TCRs, although experiences with ap T-cell-derived molecules likely will accelerate the characterization of y6 TCH heterodiiners. More iinportantly perhaps, as y6 T-cell functions are still poorly defined, crucial insights inay be gained from the structure of their antigen receptors.
A,
A
IV. Functions
A. DISTINCTIVE I N T E H A C T I O N S \VITll
Cl
IMMUNESYSTEM
1 , Acticnted B Celly Stiinidate yS T-Cell Responw,
Early studies showed that y6 T cells are stimulated by R lymphomas in oitro. For example, inurine y6 T cclls responded to the B lymphoma CH12 in syngeneic cell cultures (Sperling and Worti5, 1989). Reactive y6 T cells expressed Thy-1 and R220, but not CD4 or CD8, and proliferated independently of a@T cells. Similarly, human antitumor CTL lines stiiiiulated with autologous Burkitt's lymphoma cells were found to contain B lymphoma-reactive y S T cells (Wright rt nl., 1989). In another study,
94
\VILLI BORN ct nl.
human y6 T-cell clones derived from cultures stimulated with EpsteinBarr virus (EBV)-transformed lymphoblastoid cell lines showed a characteristic pattern of cytotoxicity, mediating cytolysis of Daudi but not Raji cells (Fisch et al., 1990a). Because cytotoxicity against Daudi and another B lymphoblastoid target cell line, Molt4. could be inhibited with anti-TCR mAbs, it was suggested that the y6 CTLs recognized their B-cell targets via the TCR. The response to Daudi was polyclonal, however, involving essentially all y6 T cells expressing Vy9N62 as well as certain Vy9N61positive cells (Fisch et al., 1992),and MVIS not MHC restricted. Most of the Daudi-reactive cells were also stimulated by antigens from mycobacterial extracts, and some experiments indicated the presence of a GroEL (HSP60) homolog on the lymphoma cells. It was therefore suggested that the reactive y6 T cells recognize evolutionarily conserved portions of liomologous heat shock proteins within the HSP-60 family (Fisch et al., 199Ob). In contrast, others had reported Ig isotype specificity of Burkitt’s lymphoma reactive human y6 T cells (Wright et al., 1989). They later found that the same y6 T-cell CTLs lysed heterologous target cells transfected with the tumor Ig A chain gene and provided evidence suggesting that this A chain is recognized as a processed peptide in an idiotype(Id)-specific manner (Kim et nl., 1995). Id-specific ligand recognition did not involve classical MHC molecules, but could be inhibited by antibodies directed against the heat shock protein grp 75 (a member of the HSP-70 family), suggesting a requirement for this molecule in ligand recognition. In a study with mouse y6 T-cell hybridomas derived from intestinal IEL, it was found, however, that cells expressing several different TCR-V combinations all responded to the B-cell lymphoma line A20 (Sano et al., 1993). Neither class I1 MHC molecules nor FcR or the surface Ig expressed on A20 appeared to be required for these responses, although a requirement for TCR expression was evident. Although all of the earlier reports of y6 T-cell responses to B cells involved B lymphoma lines as the stimulus, later studies showed that activated normal B lymphocytes could stimulate y6 T-cell responses as well. Thus, whereas resting human B cells were not stimulatory for human peripheral blood y6 T cells, autologous B cells activated with phytohemagglutinin, pokeweed mitogen, or forinalin-treated Staphylococcus aureus all elicited strong proliferative responses, stimulating more or less the same y6 T-cell populations previously found to respond to the B lymphomas (Hiicker et nl., 1995). Similar findings were reported by another group, comparing activated peripheral blood B cells with EBV-transformed B lymphoblastoid cells lines for the stimulation of human V a l + y6 T cells (Orsini et nl., 1994). This response depended, however, on the expression of B7 and CD39 molecules on the surface of the activated B cells.
It s e e m likely that some of the yST-cell reactivity that has been reported to gram-negative bacteria or 1ipoi)olysaccliaiiclesis based on interactions with activated B cells as well. Thus, in a study with human peripheral blood lymphocytes (PBLs), it was found that responses of y6 T cells to gramnegative bacteria depended on the presence of CDFj-positive B cells (Wilhelm and Tony, 1994).Similarly,a murine epidermal y 6T-cell clone expressing Vy5N61 r&ponded to LPS, but only in the presence of B lymphoid cells (Reardon et ul., 1995), and some y6 T-cell clones, originally‘dcrived from malaria-immunized mice, responded to LPS but required splenocytes as well 1996). B-cell blasts were also found stimulatorv for y6 T cells (Tsuji et d., derived from mice expressing a transgenic, T22”-specificy8 TCR, more so than were resting B cells but less so than activated T cells (Spaner et al., 1995).In sum, there can be little doubt that activated autologous B cells are capable of stimulating y6 T cells in tiitro, consistent with observations of y6 T-cell expansions in diseases associated with polyclonal B-cell activation and proliferation (De Paoli et ul., 1990; Orsini et d.,1993; Rothenberg et al., 1996). A particularly puzzling characteristic is the polyclondity of most of these y6 T-cell responses, which has caused them to be compared to the 1990~;Herman et al., superantigen responses of a@ T cells (Fisch et d., 1991). However, no y6T-cell superantigens expressed on B cells have been identified. The TCR dependence of these responses could reflect TCRmediated ligand recognition: the idiotype-specific responses are especially difficult to reconcile with any other mechanism. However, in most cases, TCR dependency could instead be a consequence of a need for prior TCRmediated activation in the responder populations.
2. yS T Cells Can Alter Antibody Responses Numerous data indicate that y 6 T cells can influence antibody production and Ig switching, both ill tiitro and in tiitio. However, there is currently no convincing evidence for y6 T-cell-mediated antigen-specific help. In syngeneic cocultures with the surface Ig+ mouse lymphoma CH12, a population of Thy-lt, CD4-8-, B220’ y6 T cells enhanced lymphoma Ig secretion, even in the absence of the nominal antigen for the lymphoma cells (Sperling and Wortis, 1989). In another study, two murine y6 T-cell hybridomac capable of responding to a random copolymer of glutarnic acid and tyrosine (poly GT) increased antibody responses to poly GT in a liemolytic plaque assay (Vidovicand Dembic, 1991).In addition, human 7 6 T cells previously reported to a protein of the yeast form of Purucocciclioicles hrusiliensis (Muilk et al., 1984)were found to secrete factors that enhanced B-cell proliferative responses following in tiitro stimulation with P. hrasilimris, as well as fictors enhancing IgM and IgG production in response to this stimulus (Munk et al., 1995).
96
\VILLI BORN rt d.
Similarly, in a study of patients suffering from active lupus nephritis, 1S% of a large collection of patient-derived, IL-2-dependent T-cell clones were found to selectively augment IgG anti-DNA autoantibody production (Rajagopalan et nl., 1990). A substantial fraction of these (7159) were 76 T cells, expressing several different y6 TCR V genes but not Vy9/V62, the predominant pair in normal peripheral blood (Rajagopalanet al., 1992). Because anti-HSP mAbs inhibited proliferative responses of these y6 T cells to autologous EBV-transformed B-cell lines, it was suggested that they recognize and are stimulated by HSPs expressed on the surface of antibody-producing B cells (Fisch et al., 199Oc; Kaur et al., 1993). Intriguingly, more recent studies in two (YPT-cell-deficient mouse models, TCR p and TCR (Y knockout (KO) mice, also suggest that y6 T cells selectively enhance the development of autoantibodies. crp-T-cell-deficient mice do not mount substantial antibody responses against “T-cell-dependent antigens” but develop, in contrast to mice lacking both and y6 T cells, high levels of autoreactive IgG in response to infections (Pa0 et al., 1996; Peng et al., 1996; Wen et al., 1994, 1996). Here, repeated infection of TCR /3 KO mice with the parasite Eimeria vernii,fomnis increased overall antibody production and induced the development of germinal centers, having B cells and y6 T cells in closely adjacent positions. Analysis of y6 TCR gene rearrangements in individual splenic GCs revealed oligoclonal and occasionally monoclonal populations of y6 T cells with frequent VylJy4 or Vy7-Jyl rearrangements. However, in the absence of ap T cells, there was no obvious enrichment for antibodies against the challenging pathogen. Instead, after repeated infections with the parasite, many (although not all) TCR p KO mice developed antinuclear antibodies. Taken together, these studies provide strong evidence that y6 T cells have the ability to augment antibody responses. The underlying mechanism is not obvious, and no relationships between the ligand specificities of the enhanced antibodies and those of the stimulating y6 T cells have been established. Although some data indmte cell contact as a requirement of y6 T-cell help (see later), others indicate that supernatants of stimulated 76 T cells are sufficient. It seems quite likely that y6 T cells can recognize the antibody producing B cells in uivo, as they have been found to respond to activated B cells both in vitro and in viwo. Their help might be triggered by some activation marker expressed on the stimulated B cells, or, alternatively, by the same autoantigens that elicit the antibody responses. However, y6 T cells have also been found to inhibit antibody responses. It has been known for some time that spleen cells from chickens that have been immunized intravenously with sheep red blood cells (SRBC) give very low secondary responses when challenged with SRBC immediately on explantation (Chi et al., 1980), whereas responses are much higher
when the cliallenge is delayed until day 2 of culture. This difference in responsiveness was found to be due to the initial activity of inhibitory T cells, which functionally disappeared within 2 days of culture. After mAbs to chicken T-cell surfice antigens became available, it was found that the inhibitory T cells expressed y6 TCRs and CD8 (Quere et al., 1990). Moreover, they were sensitive to histamine type 2 receptor antagonists, suggesting that their functional activity is, at least in part, controlled by histamine. Similarly,human y6 T cells were found to inhibit IgG production by autologous B cells purified from peripheral blood and stimulated with formalin-treated S. a i m w (Hiicker et d., 199Fj).Because the activation of B cells renders them targets of y6 T cells in uitro, it was suggested that y6 T-cell-mediated IgG inhibition functions as a regulatory circle in vivo. Consistently, in t;ioo depletion ofCD2-4-8- lymphocytes in cattle, presumably a homolog of human y6 T cc4s, led to enhanced antibody responses on in vioo challenge (Howard et 01, 1989). More directly, it was found that murine y6 T cells, isolated from chicken ovalbumin-immunized mice, upon transfer into secondaiy recipients, inhibited IgE production very efficiently (McMenamin et d . , 1994). Essentially the same observations have also been reported in rats (McMenamin et a/., 1991, 1995). Although yS T-cell inhibition of Ig production might sometimes be nonselective, induction of Ig switching can also be experimentally interpreted as Ig suppression. Several studies indicate that y 6 T cells can induce isotype switching in B lympliocfles. Mice deficient in ap T cells, for example, were found to produce Ig of all isotypes, with comparatively high levels of IgG1 and IgE, whereas mice deficient in both ap and 76 T cells were Ig deficient (Wen et al., 1994). This finding suggested a role for y6 T cells in directing isotype switching. Studies with ap T helper cells have shown that interactions between the B-cell surface antigen CD40 and its ligand (CD40L) play an important role in T-cell-dependent isotype switching (Banchereau et al., 1994). Human y6 T cells were also found to express CD4OL, after in uitro stimulation with phorbol ester and ionomycin, albeit in lesser frequencies ant1 at lower levels than unselected ap T cells (Horner rt al., 199s). Furthermore, in the absence of IFN-y, these activated y6 T cells induced IgE synthesis in B cells, in a CD40L-dependent fashion. These data imply that, under appropriate conditions, y6 T cells can induce isotype switchingin B cells, although perhaps less efficientlythan ap T cells. Even anergic y6 T cells appear to be able to regulate immunoglobulin secretion. For example, Vy9-V62+ y6 T cells from a patient with Felty syndrome, a disease characterized by leukopenia and splenomegaly concomitant with seropositive rheumatoid arthritis, did not proliferate after being triggered with anti-CD3 mAbs, but when added to autologous poke-
weed mitogen-stimulated B cells decreased IgM secretion while increasing IgG production (Bank et al., 1995).
3. Activated ap T Cells Stindate y6 T-cell Responses Whereas y6 T cells may primarily inhibit ap T-cell reactivity, activated ap T cells have been found to stimulate y6 T-cell responses, perhaps in this way becoming subject to regulation by y6 T cells. Both in experimental influenza A and in malaria infections in mice, it was noted that proliferative y6 T-cell responses occurred in the wake of a0 T-cell reactivity (Carding et al., 1990; van der Heyde et al., 1993a,b). Similarly, peritoneal y6 T cells from Listerin-immune mice showed an enhanced potential to expand when ap T cells were restiinulated in vitro (Skeen and Ziegler, 1993b). In the lungs of influenza-infected C57BL/6 mice, reactive cup T cells predominate during the first 7 days. After day 8, however, when the infectious virus has been cleared from the respiratory tract, staggered responses of y6 T cells were noted, with maximal numbers of Vy4' cells appearing on day 10 after the infection and of Vy1/2+ cells in day 13 (Carding et al., 1990). Moreover, increases of y6 T cells were found to be, for the most part, dependent on the preceding a$ T cells (Doherty et nl., 1991, 1992).Similarly, in a mouse model of inalaria induced with P. ndnmi, blastogenesis of CD4+ap T cells occurred during ascending parasitemia, whereas increases of y6 T cells were only observed during the subsequent period of descending parasitemia. Moreover, y6 T-cell responses were absent in anti-CD4-treated mice and were much reduced in adoptive cell transfer recipients that had been given CD4-depleted cell fractions (van der Heyde et nl., 1993). In the mouse model of infection with L. monoc!itogenes, y6 T cells expanded in vitro in the presence of specifically stimulated ap T cells, although y6 T cells alone did not respond to such stimuli (Skeen and Ziegler, 1993b). In addition, irradiated ap T cells could enhance proliferation of y6 T cells, whereas irradiated y6 T cells were not stimulatory. This effect appeared to be cytokine mediated and not cell contact dependent. Both interleukin 2 (IL-2) and 7 supported y6 T-cell expansions, whereas IL-7 was only minimally stimulatory for ap T cells. The unique sensitivity of y6 T cells for IL-7 was much increased by accessory cells, which in turn could be replaced by IL-1. All three disease models are consistent with ap ly6 T-cell interactions in vivo. Underlying mechanisms are not directly addressed. The model of L. mnocytogenes infection reveals a distant, cytokine-mediated effect, probably dependent on prior y6 T-cell activation. In contrast, a study with transgenic SCID mice expressing the TlO/T22'-specific y6 TCR of the hybridoma KN6 (Spaner et ul., 1993) indicated that activated ap T cells can provide strong direct stimulation for y6 T cells, presumably recognizing
them via T22” ( S p i e r cf a!., 1995). Nornial KN6-SCID mice contain a uniform population of naive. resting KN6’ y6 T cells. Analysis of mixed lymphocyte cultures generated with KN6’ y6 T cells as responders and niitomycin C-treated stinidator cell preparations revealed a hierarchy of potency among the stimulators: ap T-cell blasts = dendritic cells > Bcell blasts > B cells > resting ap T cells. In contrast, in tiivo, only T cells seemed capable of strongly activating KNGt y6 T cells as measured bv increases in the numbers of splenic KN6’ cells, blastogenesis, and the development of proliferative anergy in the responding KN6’ population, seen in mice following injections of ap T cells. Apparently, the stimulatory activity of these ap T cells depended on their prior activation via KNGSCID alloantigens. Normal T cells derived from mice tolerant to the KN6SCID alloantigens did not readily activate KN6’ responders, but became strong stimulators when previously activated by a different means in zjitro However, a mixture of tliird-partv activated T cells together with T22hpositive non-T cells was not strong& stimulatory. It was therefore concluded that activated T22”-positive ap T cells function as direct stimulators in this model. Whether direct cognate interactions between responding y6 T cells and stimulatory ap T cells are typical or exceptional remains to be determined. 4 . Chaiigcs in y6 T Cells Can Result
iti
Alterocl
ap T-cell Reactitiity
Several observations support the general idea that y6 T cells influence T-cell development and the responses of peripheral ap T effector cells. One of the first involves mice expressing a T-cell receptor Vyl-J4C4 transgene under a H-2K” promoter (Ferrick et al., 1989a,b, 1990,1991 a,b). In these mice, expression of endogenous TCRy genes (Cy4, C y l , and Cy2) was altered, whereas TCR 6, a, and p gene expression remained essentially unchanged. Surprisingly, however, during the first 6 weeks of development, thymi of the transgenic inice were two to three times larger than those of age-matched nontransgenic controls, and a much enlarged medullary compartment, as well as decreased expression of heat-stable antigen by the thyinocytes, indicated that they had accelerated thyrnocyte maturation. Moreover, although peripheral ap T cells in the transgenic inice retained norinal phenotypes and numbers, tliey showed increased responsiveness to the mitogen concanavalin A. The difference in peripheral T-cell reactivity was particularly evident with very young mice (approximately 2 weeks) and disappeared by 12 weeks of age. Splenic T cells from the transgenic inice also showed transiently up to 100-fold increases in cytolytic reactivity with allogeiieic stimulator cells, although it was not resolved whether the stronger reactivity was due to increased responsiveness of individual effector cells or to higher frequencies of CTL precursors.
a0
100
\VILLI BORN ef nl
Even in newborn transgenic mice, allo-reactivity was already detectable. Based on the CD8’ phenotype of the effector cells, this killing was mostly, if not exclusively, ap T cell dependent. There are several explanations for the increased ap T-cell reactivity in Vyl-J4C4 transgenic mice, including &rect effects of the transgene on T-cell maturation and indirect effects caused by the changes in the y6 T-cell compartment. A mere absence of y6 T-cell functions may not be sufficient to increase alp T-cell reactivity, however. In mice lacking all y6 T cells due to a targeted disruption of the gene encoding TCR 6, no evidence for accelerated ap T-cell development or increased ap T-cell effector functions was found (Itohara et al., 1993). Increased alp T-cell reactivity has also been reported after treatment of mice with anti-pan TCR-6 mAb GL3 (Kaufmann et al., 1993). Intraperitoneal injection of this antibody (0.5 mg/mouse) lead to the disappearance of y6 T cells in lymph nodes and spleen for up to 8 days postinjection, whereas no changes were evident in the proportions and phenotypes of other T cells. However, in tissue culture, CD4’ ap T cells of GL3-treated mice proliferated strongly and produced much increased quantities of IL2. The in vivo anti TCR-6 treatment also caused in vitru development of CD8+ alp CTLs capable of recognizing both syngeneic and allogeneic target cells. There was some indication that the treatment with GL3 had merely modulated yS TCR surface expression, and it remained unresolved whether yi3T cells were stimulated or inhibited. However, TCR-dependent y6 T-cell functions were probably disrupted; therefore, changes in ap Tcell functions likely are evidence of impaired “cross talk’ between the two types of T cells. In these experimental models, the association of changes in the y6 Tcell compartment with increased ap T cell reactivity in vitro strongly suggests the capability of yi3 T cells to control, under normal circumstances, ap T-cell responses. Consistently, in vivo y6 T-cell transfer or depletion has been found to alter T-cell-dependent hypersensitivity, tolerance, and host resistance to pathogens (see later). Taken together, these findings suggest that a major function of yi3 T cells is associated with the regulation of alp T-cell responses. However, the means by which y6 T cells exert such regulatory functions are still unclear. In this regard, it is of interest that both human y6 T-cell clones and in vitro-activated y6 T-cell preparations derived from patients with Lyme arthritis were found to express high levels of FAS ligand for extended periods of time and were capable of inducing apoptosis in activated FAS-positive alp T cells from the same source (Vincent et nl., 1996). The kinetics of FAS ligand expression on the y6 T cells differed greatly from the very transient FAS ligand expression on alp T cells, suggestive of a different function in the two types of T cells. FAS ligand expression has also been observed in human CD4-8- T-cell clones
specific for CD1 and capable of lcllling mycobacteria-infected macrophages in vitro (Stenger et u1 , 1997). In addition, it has been found that y6 T cells activated in inurine listeriosis express FAS ligand in uiuo, whereas both y6 and crp T cells express FAS (A. Mukasa et al., in preparation). Concomitant with FAS ligand expression, we also noted high levels of apoptosis in y6 T cells during the Listerin infection. The FAS ligand thus might be an important and distinctive tool of activated y6 T cells, used to restrict the responses of crp T cells or other FAS-positive effector cells, to limit the growth of pathogens dependent on a normal intracellular environment, or to reduce their own activity in the declining phase of an immune response.
5. y6 T Cells nnd Natural Killer Cells As with c.0 T cells, tlie demarcations between y6 T cells and N K cells are not clearly drawn. For instance, yS T cells with NK-like cytolytic activities have been described both in a patient with a malignant lyniphoproliferative disease (Falcao et al., 1992) and in a mouse model of acute graft versus host disease (Ellison et nl., 1995). Also, most y6 T-celldependent cytotoxicity is not MHC restricted, reminiscent to that of NK cells (Mavaddat ct nl., 1993), although in some studies, clear distinctions between target cell specificities of y6 T cells and NK cells were evident (Fisch et al., 199Oa). Numerous examples of y6 T cells expressing N K cell surface markers have also been reported (Koyasu, 1994). Although this may be in part due to abnormal differentiation in uitro, in mice with CD35 deficiency, y6 T cells and cup T cells expressing NK1.l developed in uiuo (Arase et al., 1995). In normal mice, gut IELs expressing both y6 and crp TCRs and exhibiting NK-like cytolytic activities have been described (Guy-Grand et ul., 1996). IIowc.vc.r,unlike classical NK cytotoxicity, their cytolytic activities were not inhibited by cells expressing MHC class I molecules. Also, it has heen shown that cytotoxic y6 and crp T cells expressing the N K cell marker CD56 can be induced to grow out in cultures of human peripheral blood lymphocytes by a combination of the cytolanes IL-12 and IL-2 (Satoh et al., 1996). In addition, and in contrast to the aforementioned findings of Guy-Grand et al. (1996), evidence for the presence of N K cell inhibitor receptors on both inurine and human y6 T cells has been reported (Azuara et al., 1997; Nakajima et d . , 1995, 1997). However, like crp T cells, most freshly isolated y6 T cells do not express NK cell surface markers or have cytolytic activities. In fact, a study in DBAI2 mice indicated that y6 T cells expressing NK1.1 apparently belong to a distinct subset characterized by dull Thy-l expression and the use of a particular pair of TCR V genes, Vyl and the newly discovered V86.4 (Azuara et al., 1997). This situation
102
\VII,IJ HORN
Pf
nl
resembles that of NK1.1-positive a/3T cells, with their distinct TCR repertoire and functional properties (Bendelac et al., 1995; Lantz and Bendelac, 1994; Yoshimoto et al., 1995). Further differences between yS T cells and NK cells were evident in studies with cytokines. Thus, perhaps even more than B cells, murine y6 T cells depend on the presence of IL-7 for their growth and differentiation (Lynch and Shevach, 1992; Skeen and Ziegler, 199313). In contrast, no such dependence was detectedwith N K cells. In mice lacking IL-7 receptor a due to a genetic manipulation ( I L 7 R KO), some defects in T- and B-cell development were found, but, most noticeably, y6 T cells were completely absent (He and Malek, 1996). In contrast, the development and function of N K cells were normal. In a similar study with different IL-7R KO mice, y6 T cells were absent from skin, gut, liver, and spleen, whereas a/3 T cells and B cells were only reduced and NK cells developed normally (Maki et al., 1996). Although some studies have failed to reveal interactions between 76 T cells and NK cells, two papers describe functional interplay between the two types of cells in the course of an infectious disease. In mutant mice rendered deficient in either TCR /3 or TCR 6 genes, NK cell functions were found unimpaired (Mombaerts et al., 1994).This was not unexpected, as normal or even enhanced NK cells activities can be found in mice that are congenitally thymus deficient and in mice carrying the SCID mutation (Bancroft et al., 1987; Unanue, 1997). However, in a mouse model of infection with the bacterium L. monocytogenes, it was found that in oivo depletion of NK1.l-positive cells, in the presence of y6 T cells, led to the accelerated clearance of the pathogen and was accompanied by an enhanced expansion of y6 T cells in the peritoneal cavity, whereas NK cell depletion in the absence of y6 T cells did not accelerate bacterial clearance (Takada et al., 1994). These data suggested a regulatory relationship between the two types of cells in which NK cells might inhibit the protective functions of 76 T cells. In contrast, another group, also using a mouse model of listeriosis, reported that yS T cells control NK-cell-mediated innate resistance (Lade1 et al., 1996). They found that IFN-y production by N K cells in TCR 6 KO mice was impaired in comparison to normal mice, a defect perhaps caused by the reduced production of TNF a! in the absence of y6 T cells. Accordingly, they suggested that y6 T cells provide a link between the innate and the adaptive, antigen-specific immune responses against murine listeriosis, regulating both early NK and late a/3 T-cell reactivities. 6. y6 T Cells and Macrophages Macrophages have been found to serve as potent accessory cells in the responses of y6 T cells to bacteria. Inversely, y6 T cells are capable of activating macrophages via cytokines and probably also chemokines.
Ih.lh.1UhOREClJL.ATOHYFUNC.TIONS OF y8 T CELL5
103
In studies comparing the human T-cell response to live and heat-killed bacteria (Mycohacterium tihxxlosi.7 and Snbnonellci typ/iimiirirtnz), monocytes infected with live bacteria were effective inducers of y6 T-cell expansion (Havlir et al., 1991). In contrast, heat-killed bacteria or purified bacterial proteins induced responses of CD4’ c.p T cells, but very little increase in y6 T-cell numbers. Similarly, live but not heat-killed mycobacteria induced y6T-cell responses in mice (Griffinet al., 1991).The response of human y6 T cells to live M trtbercidosis was found to be dependent on the presence of mononuclear phagocytes, both in the induction phase and for the expansion of previously activated y6 T cells. Nevertheless, unlike CD4’ c.p T cells, y6 T cells were not restricted by class I or class I1 MHC molecules and only responded in the presence of intact bacteria (Boom et d.,1992). Similar MHC nonrestricted responses of human y6 T cells to live pathogens have been observed after stimulation with P. fulcipczmi vi malaria-infected erythrocytes in the presence of autologous and heterologous plastic adherent accessory cells (Goodier et al., 1992). Not only blood monocytes but also alveolar macrophages were found to serve as accessory cells for tlie human y6 T-cell response to M. tuberculosis (Balaji et al., 1995). This is of particular interest because alveolar macrophages form the first line of defense against aerosolized M . tuberciilosis infection in the lung. Two kinds of cellular interactions were observed. Whereas antibacterial responses of activated y6 T cells were always supported by the alveolar macrophages, responses of resting y6 T cells were supported only at low alveolar macrophage to T-cell ratios, but were inhibited at higher ratios (>3 : 1).No such dose-dependent inhibition was seen with blood monocytes, however, which always supported bacteria-induced y6 T-cell responses. This reactivity pattern suggested that alveolar macrophages regulate y6 T-cell reactivity in the lung. In the normal alveolus, the alveolar inacrophage to T cell ratio is equal or greater than 9 : 1, thus providing an environment inhibitory for the induction of y6 T-cell reactivity. However, if y6 T cells hecome activated despite this inhibition, due to strong stimuli, the same environment should enhance these y6 Tcell responses. Moreover, if the ratio of macrophages to T cells is altered, e.g., in the course of an inflammatory response, “diluted” alveolar macrophages could even support de riovo activation of y6 T cells. The molecular nature of these accessory functions is still largely unknown. It is possible that macrophages serve as antigen-presenting cells for y6 T cells (Boom et al., 1992). If so, the absence of MHC restriction in the y6 T-cell responses to mycobacteria and malaria hints at underlyng mechanisms quite different froin classical MHC-restricted antigen presentation and recognition. Alternatively, macrophages may merely produce factors supporting (or suppressing) these 76 T-cell responses. In this regard, it is noteworthy that inurine y6 T cells have been found to respond to
104
WILL1 BORN et a/
IL-15 as a growth factor (Nishimura et al., 1996). IL-15 is a cytokine that uses both p and y chains of the IL-SR for signal transduction and is produced by activated monocytes and macrophages (Grabstein et al., 1994). Murine y6 T cells appearing after primary infection with Salmonella choleraesuis were found to express /3 and y chains of IL-2R and to proliferate in response to rIL-15. Moreover, stimulation with this cytokine induced the y 6 T cells to produce IFN-y and IL-4. In an in vitro model of y6 Tcell/macrophage interactions, 7 6 T cells produced IFN-y in response to the monocyte/macrophage cell line, J774A.1infected with S. choleraesuis, which expressed high levels of IL-15 mRNA. The 7 6 T-cell response was inhibited by cytokine-neutralizing anti-IL-15 mAbs. It thus appeared that IL-15 derived from infected macrophages contributes to the early (ap Tcell independent) activation of y6 T cells (Nishimura et al., 1996) during sal~nonellosisand perhaps also other infectious diseases. y6 T cells appear to be capable of activating macrophages as well. Macrophages derived from mice lacking y6 T cells (TCR-6-’- mice) were found to produce only small amounts of tumor necrosis factor a (TNFa ) in response to LPS as compared to cells derived from normal mice (Nishimura et al., 1995). They also expressed the LPS receptor CD14 at reduced levels. However, this was not an intrinsic defect of macrophages from TCR-8-’- mice, as preincubating them with y 8 T cells from TCRa+’- littermates restored their capacity to produce TNF-a in response to LPS. At least in part, the macrophage priming activity of y6 T cells can be ascribed to their production of IFN-y, as it is inhibited by IFN-y neutralizing mAbs. That y 6 T cells induce macrophage functions was also evident in a mouse model of fungal infections (Jones-Carson et al., 1995). Here, following intraperitoneal inoculation of live (but not heat-killed) Candida albicans, the numbers of y6 T cells and macrophages increased sharply. Based on this observation, in vitro interactions between activated peritoneal y6 T cells and macrophages were examined. Coincubation of macrophages with C. albicans-elicited y6 T cells, but not with y6 T cells activated by other means, resulted in large increases of macrophage-derived nitric oxide (NO). NO production was apparently induced by y8 T-cell-derived IFN-y because it could be abrogated by cytokine-neutralizing anti-IFN-y mAbs, reminiscent of the induction of TNF-a production by macrophages in the LPS system (Nishimura et al., 1995). y6 T cells also appeared to control CandiZida-induced NO production in uiuo, as y6 T-cell depletion resulted in reduced levels of iNOS mRNA. At the same time, systemic y6 T-cell depletion resulted in reduced levels of iNOS mRNA, and increased susceptibility to orogastric candidiasis.
IMMUNORECULATOHY F U M : T I O N S OF yS T CELLS
105
IFN-y may not be the only factor by which y6 T cells influence macrophage functions. In mice infected with L. nwiiocytogeties, it was found that y6 T-cell depletion resulted in prolonged and exacerbated liver inflammation (Fu et al., l994b), apparently because the replacement of neutropliils in the lesions with longer lived mononuclear phagocytes was interrupted. Similar observations have been reported in M . tuberculosisinfected inice (D’Souza et al., 1997). New data indlcate that the delayed recruitment of macrophages seen in the absence of y6 T cells coincides with tlie reduced production of inacrophage-attracting chemokine MCP1 in the absence of y6 T cells (F. DiTirro ct al., unpublished data). This effect could be due to the ability of activated y6 T cells to produce MCP1 themselves (A. M ., unpublished) or, alternatively, by their ability to control MCP-1 production by other cells. In addition, our own recent studies have shown that both murine and human y6 T cells are particularly susceptible to stimulation through macrophage-derived TNF-a, as compared to ap T cells (Lahn et al., 1998). Reciprocal macrophagely6 T-cell interactions therefore likely represent a potent early immune response circuit.
B. H05T
RESI5TANCE TO
PATHOGENS
Responses of y6 T cells are evident in almost every infectious disease examined to date (Born et al., 1991a, 1991b; Haas et al., 1993). However, only in a few animal models is convincing evidence available that y6 T cells contribute to host resistance. In addition, although there are a few exaniples of pathogen-specific y6 T-cell responses, in most other cases, y6 T cells could be only be indirectly protective, e.g., because they help resolve inflammation (see later). A contribution of y6 T cells to host protection could also be obscured by redundant host defenses and only become obvious when the host is immiine compromised or when the pathogenic load is unusually high. Selected examples are described. Both mice and human y6 T cells are stiiiiulated by and lyse herpes simplex virus I (HSV 1)-infected cells in rjitro (Bukowski et al., 1994; Doherty and Zinkemagel, 1974; Maccario et al., 1995; Welsh et al., 1997). However, whereas human T-cell clones did not always show specificity for the virus, they did not lyse niock-infected cells but lysed cells infected with the nonrelated vaccinia virus equally well, at least some mouse y6 T cells specifically recognized HSV glycoprotein I (Johnson et al., 1992; Sciammas et al., 1994). Moreover, infections of mice with HSV-1 elicit protective y6 T cells (Sciammas et al., 1997). To better define tlie role of yS T cells in anti-HSV-I immunity, immunodeficient mice (lacking TCRa,lacking TCR-a and treated with anti-TCR-6 mAbs, or lacking both TCR-P and TCR-6) were infected in the footpad or eye. In the absence
106
WILIJ BORN t>t a1
of y6 T cells, the severity of virus-induced epithelial lesions was much increased. Further, 7 6 T cells reduced mortality by preventing the development of viral encephalitis. Protection by y6 T cells appeared to be based both on the arrest of viral replication and on the reduction of neurovirulence. In a mouse model of infection with vaccinia virus (W),dynamic changes in y6T-cell populations were also noted. These correlated with an increased frequency of cytolytic y6 T cells against W-infected target cells, although the ligand specificitiesof these cells remain unclear, Moreover, mice genetically deficient for y6 T cells had significantly increased viral titers as early as 3-4 days after the infection. The extent of early protection by yS T cells was comparable to that by N K cells. It was therefore suggested that y6 T cells may serve as effectors of natural immunity in this model and perhaps in other viral infections as well (Welsh et al., 1997). Infections with the facultative intracellular bacterium L. rrwnocytogenes elicit yS T-cell responses in mice and humans (Fu et al., 1994b; Hiromatsu et al., 1992a,b;Jouen-Beades et al., 1997; Mukasa et al., 1995, 1997; Ohga et al., 1990; Skeen and Ziegler, 1993b; Usami et al., 1995). Moreover, human yS T cells have been found to react with listeriolysin 0 (Guo et al., 1995),a secreted protein and major virulence factor of this bacterium. y6 T cell responses in murine infections with Listeria have been analyzed in some detail. However, it is not yet clear whether the reactive inurine y6 populations respond to pathogen- or host-derived ligands. Depletion of y6 T cells with anti-TCR-6 rnAb diminished host resistance during the early to intermediate stages of the infection, although bacterial titers were only somewhat reduced. The contribution of y6 T cells to host protection was more pronounced in mice genetically deficient in µglobulin (Roberts et al., 1993) and therefore lacking CD8+ ap T cells, which normally make a decisive contribution toward the sterile elimination of Lis-
teria. Similar findings have been reported with mouse models of rnycobacterial infections (Griffin et al., 1991). Mutant mice lacking a/3 or y6 T cells or all T and B cells (recombinase activation gene deficient, RAG-l-’-) were infected intravenously with M . bovis bacillus Calmette Guerin (BCG),and the relative contributions of ap and y6 T cells to the host immune response were assessed (Ladel et al., 1995b). RAG-1-/- mice as well as TCR p-’mice failed to develop granulomatous lesions and eventually succumbed to the mycobacterial infection, whereas TCR 6-’- mice survived. Antigeninduced IFN-y production by spleen cells in vitro was abrogated in RAG-’- mice and diminished in both TCR p-’- and TCR 8-j- mutants. Moreover, cell reconstitutions suggested that both ap and y6 T cells were required for the antigen-induced production of this cytokine. These results
IhIhlUNORE(;UI.ATORY F1INCTIO”S OF yiS T (:ELL5
107
were interpreted to indicate an essential role for ap T cells in the host response against M . hotis BCG, as well as accessory functions for y6 T cells. A protective role for 76 T cells was also noted in mice injected with M . tuberculosis (Lade1 et al., 1995a). However, in mice infected with low doses of aerosolized 144. tubercrilosis strain Erdman, no significant difference between normal and TCR-6-’- mice in the control and containment of inycobacterial infections was seen (D’Souzaet aZ., 1997).Nevertheless, the mutant mice exhibited a substantial pyogenic form of the granulomatous response as compared to the lymphocytic response seen in normal controls and showed much higher mortality over a period of 120 days. This suggested that y6 T cells do not directly contribute to protection against tuberculosis, except perhaps when bacterial loads are very high. Instead, and reminiscent of the situation in infections with L nlonocytogerm (Fu et al., 1994b), it appeared that y6 T cells modulate the course of a potentially autodestructive inflaininatory response. Although vigorous y6 T-cell responses have been noted after stimulation with grani-negative bacteria and their products (Emoto et al., 1992, 1993; Skeen and Ziegler, 1993a),currently available data about their contribution to host resistance against infections with such bacteria are contradictory. In one study, BALB/c mice were infected orally with Snbrwnella enteriditis (Mixter et al., 1994). After treatment of these inice with T-cell-depleting antibodies, the SO% lethal dose of bacteria decreased in both ap and y6 T-cell-depleted groups, suggesting a contribution by both types of T cells to host resistance against SalnioizeZla. In a different study, mice genetically deficient in crp T cells, y 6 T cells, or both were infected orally or intraperitoneally with Snlriioriella dublin (Weintraub et al., 1997),and the progression of the disease was monitored by determining bacterial numbers in the feces, gut wall, Peyer’s patches, niesenteric lymph nodes, spleen, and liver. No role for intestinal epithelial and niucosal y6 or ap T cells in controlling the invasion of S. chblin into the intestine, or its subsequent replication in Peyer’s patches or the gut wall, was evident. Systemic infections were equally severe for the first 6 days in normal, a0 T-cell-deficient and 7 8 T-cell-deficient mice, arid ap T cells but not y6 T cells were required for clearance of S. clfiblin. Because susceptibility to salmonella infection in mice is affected by the alleles at the Zty locus (Elnoto et al., 1993),T-cellmutant mice carlying the Ity-sensitive or -resistance alleles were tested also, without a significantly different outcome. Nevertheless, at late stages (15-18 days after infection), lower bacterial counts in the livers of inice lacking only ap T cells were noted as compared to mice lacking both ap and y 6 T cells. This was taken to suggest that yS T cells can contribute to acquired immunity against S. cizddin In our own studies with S. typhirnuriiriii ( M . DeGroote et nl., unpublished observations), only a slightly re-
108
WILI,I BORN et nl.
duced antibacterial resistance in mice treated with anti-TCR-6 mAbs was found when bacteria were inoculated intraperitoneally. Surprisingly, however, the same treatment led to increased resistance when bacteria were inoculated parenterally. Moreover, and in contrast to an earlier study by another group, in comparisons of a collection of Salmonella-mutant strains, there were large variations in the induction of 7 6 T-cells responses, but without obvious correlation between the relative virulence and the strength of the y6 T-cell response. Several studies have documented y6 T-cell responses in malaria infection (Elloso et al., 1996; Goodier et al., 1992; Langhorne et al., 1992; van der Heyde et al., 1995). In an attempt to define the role of y6 T cells in the immune response against malaria, one group has monitored the development of liver and blood stages of Plasmodium yoelii, a rodent malaria parasite, in immunized and nonimmunized a0 T-cell-deficient mice (Tsuji et al., 1994). Immunization of ap T-cell-deficient mice with irradiated sporozoites induced an immune response that significantly inhibited the development of the parasite’s liver stages. This inhibitory immune response was abolished by the antibody-mediated transient in vivo depletion of y6 T cells. In addition, y6 T-cell clones were derived from such malariaimmunized a0 T-cell-deficient mice. Adoptive transfer of one of these clones to normal mice inhibited the development of liver stages, following sporozoite inoculation. These results were taken as evidence for y6 Tcell-mediated protective immunity against liver stages of malaria parasites without a need for a0 T cells. In contrast, both normal and y6 T-celldeficient mice cleared the blood stages of P. yoelii, whereas ~$3 T-celldeficient mice failed to control parasiteniia. In a subsequent study, several y6 T-cell clones derived from malaria-infected mice, includmg the protective one, were examined in an attempt to define their TCR structure and antigen specificities (Tsuji et al., 1996). Surprisingly, no evidence for antiparasite specificitieswas found. However, the protective clone differed from the others in a combination of parameters, i.e., TCR structure, expression of additional cell surface molecules, and cytokine production, suggesting that this combination defines the distinct protective phenotype.
C. INFLAMMATION A role for y6 T cells in inflammation was initially suggested based on kinetic studies in a mouse model of influenza infection (Carding, 1990; Carding et al., 1990). In this model, infectious virus is cleared from the mouse lung within 7-10 days of respiratory exposure. During this period, influenza-specific ap T cells and B cells develop and eventually mediate, in conjunction with inflammatory macrophages, the clearance process (Doherty et al., 1992). It was noted that within the first 7 days of the
IMMUNOREGULATORY FUNCTIONS OF y6 T CELLS
109
infection, most of the T cells recovered from the lung were ab T cells, whereas cells expressing y6 TCR mRNA were present at increased levels during the following week. The response of y6 T cells was staged, such that maximal numbers of cells expressing Vy4 appeared on day 10 after respiratory infection, whereas by day 13, Vy1/2+ cells predominated. This late appearance of y6 T cells after viral clearance suggested that their response was not driven by the recognition of viral antigens. Indeed, in secondary infections, there was no demonstrable virus specificity in the y6 T-cell response, although it was proportional to the intensity of the c.b T-cell reactivity (Doherty et al., 1992). Therefore, the y6 T-cell response likely occurred as a reaction to the protective host response, either merely as a bystander effect or perhaps as part of the host’s effort to resolve inflammation. That y6 T cells can indeed contribute to the resolution of inflammation was more directly evident in later studies involving the murine response to the facultative intracellular bacterium L. inoriocytogenes (Fu et al., 199413; Mombaerts et al., 1993). In healthy inimunocompetent mice, Listeria is cleared within 7-10 days after intravenous or intraperitoneal inoculation, following a vigorous inflammatory response and the subsequent development of Listeria-specific cytolytic ap T cells. Given that Listeria tends to multiply within the cells of its target tissues, antilisterial antibodies may develop but do not significantly contribute to host resistance. As already discussed earlier, y6 T cells generally are not required for bacterial clearance and the development of anti-bacterial immunity. However, mice lacking y6 T cells showed dramatic changes in the inflammatory response following bacterial infection ( F u et nl., 1994b; Moinbaerts et al., 1993). The normal host response in the liver, one of the primary target tissues in Listeria infections, is characterized by the development of small granulomatous lesions around infected hepatocytes. Listeria remains confined to these lesions, being initially contained by highly efficient innate host defenses primarily stemming from neutrophils, and eventually cleared by recruited Listeria-specific a0 T cells in conjunction with activated macrophages. 76 T cells seem to appear in these lesions around the same time as (YPT cells. If they are absent, resolution of the inflammatory response is much impaired, leading to severe parenchymal damage. Interestingly, within such enlarged necrotic lesions, Listeria was not confined to individual hepatocytes, but instead appeared to grow extracellularly as well. Observations quite similar to those with Listeria have also been reported in a mouse model of M . tuberculosis infection (D’Souza et al., 1997). A number of ways in which y6 T cells might contribute to the resolution of these inflammatory foci can be envisaged. For example, it has been proposed that y6 T cells regulate responses ofab T cells (Ferrick et d., 1995),
both because they tend to increase simultaneously with or subsequently to ap T-cell responses and because they were found to produce IL-10 in the course of Listerin infection (Hsieh et al., 1996). Downmodulation of the a0 T-cell response during later stages of Listeria infections likely reduces the excess production of proinflammatory cytokines such as IFNy and TNF-a, and thus might help prevent unnecessary tissue damage. However, it is also possible that y6 T cells enhance the clearance process. For instance, it has been found that Listeria-infected mice lacking y6T cells express mRNA encoding macrophage chemoattractant protein I (MCP-1) at much reduced levels ( I . Orme, personal communication), and it has been observed that some y6 T-cell clones contain mRNA for MCP-1 (A. Mukasa et al., unpublished results). y6 T cells might thus control production of a mediator necessary for the efficient recruitment of macrophages to sites of infection. Because macrophages are required for the resolution of inflammation, their (partial) control through y6 T cells could explain the histopathology observed in the absence of y6 T cells. y6 T cells also appear to be involved in early and intermediate stages of the inflammatory response. A murine epidermal y6 T-cell clone was found to produce the chemokines lymphotactin, MIP-la, MIP-10, and RANTES (although not MCP-1) (Boismenu et al., 1996), suggesting that epidermal y6 T cells participate in the early recruitment of iinrnunocytes to sites of epidermal injury. Moreover, various types of y6 T cells were found to produce IFN-7 (Burg et al., 1991; Christmas and Meager, 1990; Ferrick et al., 1995; Tsuji et al., 1996), probably the major macrophageactivating cytokine. Because both NK cells and ap T cells also produce IFN-y (Dunn and North, 1991; Heinzel et al., 1991; Scharton and Scott, 1993; Tripp et nl., 1993), the significance of y6 T cells as a source of this cytolane may be due to an ability to become activated when these other cells are not. AIong this line, it has been suggested that early activated yS T cells producing IFN-7 or IL-4 may help determine whether TH1 or TH2 ap T cells develop preferentially (Ferrick et nl., 1995).In addition, y6 T cells are capable of producing IL-2, IL-3, and granulocyte/macrophagecolony stimulating factors (GM-CSFs),growth factors capable of promoting the expansion of recruited and activated immunocytes (Yokotaet nl., 1996). Given that y6 T cells may play an important role in inflammation, the question of what stimulates their responses again arises. It has been found that y6 T cells are particularly sensitive to stimulation by TNF-a (M. Lahn et al., 1998), one of the earliest inflammatory mediators (Vassalli, 1992). Thus, by the time they encounter antigenic ligands, they may be in a presensitized state. However, engagement of the y6 TCR seems to be a prerequisite for functional responses such as proliferation and cytokine production. Intriguingly, some evidence indicates that foreign antigens are not necessary to elicit y6 T-cell responses in inflammation. First, y6 T
IMMUNOREGULATOHY F U N C l 7 O N S OF yS T CELLS
111
cells expressing similar or even identical TCKs are elicited in numerous different diseases (Born et a l , 1994; Haas et nl., 1993). Second, in the absence of infectious pathogens, y6 T-cell depletion was found to have strong effects on the development of collagen-induced arthritis in mice (Peterman et nl., 1993). Interestingly, early depletion of y6 T cells in this disease model was anti-inflammatory, whereas later depletion was proinflammatory, suggesting that y6 T cells could initially promote tlie autoaggressive response, but later reduce its intensity. Third, in a mouse model of testicular inflammation, in which y6 T cells were also found to prevent excessive tissue damage, the same ~6 T-cell subset responded regardless of whether inflammation was caused by bacterial infection or by autoaggressive T cells (Mukasa et nl., 1997). In this model, orchitis is induced in one testicle by inoculating live L nionocytogenes. Because tlie testis is not a natural target tissue for this bacterium, Listeria grows within the tissue of introduced artificially, but does not spread to the contralateral testicle. Nevertheless, not only the infected but also tlie uiiinfected organ shows inflammation, the latter due to infiltration of autoaggressive T cells elicited with antigens revealed in the infected testicle. y6 T-cell infiltrates are noticeable in both testicles and consist predominantly of cells expressing an invariant TCR composed of VyG and V61. These cells all have the same (canonical) TCR junctions that are also norinally present in tongue, uterus, and vagina (Itohara et n l , 1990) and during pregnancy in the placenta (Heyborne et d., 1992). Because cells expressing this same TCR also arise in the Listerin-infected liver (Roark et nl , 1996) and in the lungs of mice sensitized with aerosolized mycobacterial antigens (Augustin et al., 1989), it seems most likely that \’yG/\’6l+ y6 T cells recognize an autologous ligand associated with the inflammatory host response, and expressed in many different tissues. Intriguingly, iricreases in canonical VyG-Jyl transcripts have also been noted in a mouse model of myelin basic protein-induced experimental allergic encephalomyelitis (EAE), a chronic inflammatory disease of the central nervous system that resembles multiple sclerosis (Olive, 1997). In an effort to modulate EAE by immunization with a peptide representing the canonical VyG-Vyl junction and adjacent amino acids (-YCACWDSSGFHK-), a delay in the onset of EAE and a mild reduction in disease severity in the peptide-treated animals were observed. In contrast to the findings suggesting that VyG/V61’ cells regulate inflammation, this was interpreted to indicate a potentially pathogenic role of these cells in EAE (Olive, 1997). D. C~JTANEOIIS HYPERSEN~ ITY ITIL Studies in mice indicate that y6 T cells affect tlie development of cutaneous contact sensitivity (CS) reactions. In the cutaneous response to picryl
112
WILL1 BORN rt nl
chloride, for example, two different antigen-specific ap T-cell responses, one early and one late, are distinguishable. In addition, y6 T cells were found to be required for the adoptive transfer of CS (Ptak and Askenase, 1992). Here, the y6 T cells were found to be required for the adoptive transfer of CS (Ptak and Askenase, 1992). The y6 T cells assisting CS aP T effector cells were CD8’ and lacked both antigen specificity and MHC restriction. Regulatory yS T cells of this type were present in the spleen of normal but not of “nude” mice, suggesting some degree of thymus dependence (Askenase et al., 1995). Treatment of cell transfer recipients with Bordetalla pertussis antigen (Bp) or with low doses of cyclophosphamide permitted transfer of CS with ap T cells alone. This was interpreted to indicate that Bp elicits recipient-derived y6 T cells, capable of assisting the transferred ap T cells in the CS response, whereas cyclophosphamide prevents a CS-inhibitory response by recipient-derived suppressor cells. y6 T cells capable of assisting CS effector cells were found to express preferentially Vy5 and V64. They did not require prior immunization in order to support aP T-cell responses (Ptak et al., 1996). However, y6 T cells were also found to be involved in the downregulation of CS (Szczepanik et al., 1996). In contrast to the enhancing y6 T cells, CS downregulating y6 T cells showed antigen specificity, although not restriction by classical MHC elements. Because CS could not be induced in the absence of ap T cells, it appears that y6 T cells are primarily involved in the regulation of CS responses, both via antigen-specific and nonspecific mechanisms. E. TOLERANCE 1. Ingested Antigens
The primary mechanisms by which orally introduced antigens induce tolerance rely on the generation of active suppression in the presence of low antigen doses and on clonal anergy with higher doses. Active suppression is mediated via cytokines such as transforming growth factor-p (TGF-P) and IL-4. Regulatory cells induced by oral tolerization are triggered in an antigen-specific fashion, but they suppress subsequent antigen responses in an antigen-nonspecific fashion. Orally induced regulatory cells are capable of suppressing autoimmunity, alloreactivity, and graft rejection (Melamed and Friedman, 1993; Mowat, 1987; Weiner et al., 1994). Studies with animal models have shown that y6 T cells can influence development and tnaintenance of oral tolerance (Fujihashi et al., 1997). In a mouse model of oral tolerance to heterologous red blood cells, it was found that yS T cells can break the antigen-induced tolerant state. Earlier work had shown that abrogation of tolerance in mice to sheep red
IMMIJNOREGULATORY FUNCTIONS O F y6
'r
(:ELLS
113
blood cells is mediated by CD3', CD4-L- (double negative), Vicin villosu lectin-adherent cells. Because athyinic nude (nu/nu) mice contain relatively large niiinbers of Thy-l+,double negative cells that also contain functionally rearranged TCR-y genes, it was tested whether these cells also had the ability to abrogate tolerance (Fujihashi et d , 1989). Examining the phenotype and distribution of CD3' T cells in BALB/c nu/nu mice, they found that double negative cells were the predominant T-cell population in tlie spleen, whereas in inesenteric arid peripheral lymph nodes, approximately one-third of the CD3' cells were double negative. In contrast, CD3' double negative cells represent only a ininor subpopulation in normal mice. The ypleiiic CD3' double negative population in iiu/nu mice contained cells capable of abrogating oral tolerance to SRBC, as indicated by a restoration of antibody responses to SRBC. In order to become regulatory, these cells had to be induced by antigen priming. Antigen-primed cells could restore responsiveness to SRBC in orally tolerant BALB/c mice. Specifically, anti-SRBC PFC responses of the IgM, IgG1, and IgGBb subtypes were increased approximately three-fold over the tolerant background. In marked contrast, responses of the IgA subtype were not increased. Whereas SRBC-primed cells restored responsiveness to SRBC, horse red blood cell (HRBC)-primed cells did not, suggesting a requirement for specific antigen recognition. Among CD3' double negative cells in nude spleen, only V. villo5n-adhereiit cells were capable of abrogating tolerance. Cells capable of breaking oral tolerance to SRBC could also be enriched by selecting for CDS' cells among CD3' double negative cells. To cliaracterize the TCR expressed on CD:3+ tolerogenic cells, a rabbit antibody raised against a synthetic TCR 6 peptide was used. This antibody precipitated a 45-kDa band (presumed to contain TCR 6) from the cell membrane fraction of tlie CD3' double negative cells but not from CD3'/ CD8' cells. Consistently, treatment with an antibody specific for CD3 precipitated both ii 45-and 35-kDa band (presumed to contain TCR-y) froin tlie CD3' double negative cells and 43- and 38-kDa bands (presumably containing TCR p and TCR a,respectively) from CD3'/CD8' cells. Alreadv in this early study, it was concluded that y6 T cells are capable of breaking oral tolerance in an antigen-specific manner. In a second study, the same group extended their original observation to include murine intraepitlielial lymphocytes ( IEL), a population of cells considered to be of preeminent importance in inucosal iniinunity (Fujiliashi et nl., 1990). CD3' IEL purified from the small intestines of mice orally prinied with SRBC could abrogate oral tolerance to this antigen in recipients of adoptively transferred cells. Unlike splenocytes from "nude" mice, IEL restored not only IgM, IgG1, and IgGBb, but also IgA antibody responses to SRBC. As shown with spleen cells, induction of cells capable
114
WILL1 BOHN et a /
of abrogating tolerance was antigen specific, such that SRBC only primed cells capable of restoring responses to SRBC but not HRBC, and vice versa. Among IEL, both CD3+ double negative and CD3t/CD8’ cells could abrogate tolerance. Both subsets restored antibody responses, including all four of the previously mentioned subclasses, but CD8+ cells preferentially restored IgA antibodies. Among CD3+/CD8+cells, only Thy-1 low/negative cells (approximately 80%) abrogated oral tolerance. In this study, TCR expression was reexamined using a mAb with specificity for TCR p framework determinants (mAb H57.597), and it was found that TCR p’ IEL did not abrogate tolerance whereas CD3+,TCR p- cells did. Iminunoprecipitation of membrane fractions from purified CD3+/ CD4-/CD8t/Thy-l- IEL with the previously mentioned anti-TCR 6 peptide polyclonal Ab revealed bands of 45 and 35 kDa, presumed to represent TCR 6 and y chains, respectively. It was concluded that intraepithelial y6 T cells are capable of abrogating oral tolerance whereas intraepithelial cup T cells are not. Moreover, cells capable of breaking oral tolerance were heterogeneous inasmuch as only CD8’ IEL restored IgA responses. As with “nude” spleen cells, the regulatory function of IEL in oral tolerance was limited to the antigen used for piiining. In a more recent study using immunofluorescent cell sorting and a mAb specific for TCR 6, investigators of the same group confirmed that only IEL expressing yS TCRs, not those expressing ap TCRs, are capable of abrogating oral tolerance to heterologous red blood cells (Fujihashi et al., 1992, 1996). Nevertheless, only ap T cells provided help for antibody responses to red blood cells (RBC), presumably via their production of IL-5. The mechanism by which y6 T cells restore antibody responses to RBC remains unresolved. Regulatory y6 IEL were found within the V. uillusa-adherent fraction and these cells contained only low levels of IL5 mRNA, and IL-5 mRNA-positive cells were infrequent. Moreover, exogenous IL-5 was not sufficient to reverse antigen-specific unresponsiveness in cultures containing splenocytes from mice tolerized orally. Interestingly, mice tolerized orally (as opposed to antigen-primed mice) retained y6 IEL capable of abrogating oral tolerance in adoptive transfer recipients. This was interpreted to suggest that y6 IEL maintain appropriate antibody responses, in particular involving the IgA subclass, for the protection of mucosal sites, in the presence of systemic unresponsiveness. The precise mechanism for the abrogation of oral tolerance by y6 IEL, in particular, the basis for the antigen specificity of this activity, was not resolved. Thus, while 78 TCRs could be truly antigen specific, it is also possible that they recognize TCR idiotypes of antigen-specific aj3 T cells or, less likely perhaps, other ligaiids that become available in an antigen-specific manner.
IMMUNOHE(;lILATOHY FUNCTIOSS 01: yS T CELLS
1L5
The role of y6 T cells has also been investigated in a mouse model of oral tolerance to ovalbumin (OVA)(Ke et al., 1997).Here, oral administration of OVA can lead to a state of tolerance associated with the activation of CD8+ T cells capable of transferring unresponsiveness to naive syngeneic hosts. It was further found that injection of a monoclonal anti-TCR 6 antibody (GL3) downmodulated the expression of y6 TCR and inhibited the induction of oral tolerance to OVA, as measured by antibody, CD4', and CD8' T-cell responses. There was no indication that GL3 activated y6 T cells in uivo, suggesting that the antibody treatment inhibited yS T-cell fiinctions. Consistently, oral administration of OVA did riot induce tolerance in TCR 6 KO mice. In a study on the role of y6 T cells in the induction and suppression of experimental autoimmune uveitis in rats, it was also concluded that they can contribute to the establishment of oral tolerance (Wildner et al., 1996). Here, depletion of aj3 T cells but not yS T cells abrogated the disease, establishing a p T cells as the autoaggressive effector cells. However, adoptive transfer of y6 T cells from orally tolerized rats led to suppression of uveitis in an antigen-dependent fashion. Specifically, uveitis induced by a peptide derived from the uveitogenic retinal-soluble antigen was suppressed by y6 T cells derived from rats tolerized with the same peptide. However, HLA peptide B27PD was also tolerogenic. Transfer of ap T cells from tolerized donors, as well as y6 or aj3 T cells from animals fed with control peptides, had no ameliorating effect. Taken together, findings with model? of oral tolerance suggest that yS T cells can have both tolerogenic and antitolerogenic effects, perhaps dependent on the nature of the antigen involved or on the protocol of tolerization.
2. Inhaled Antigens As with ingested antigens, tolerance to inhaled antigens is an active process, involving selective T-cell-mediated immune suppression. Repeated aerosol exposure to antigen at the appropriate doses can elicit apparent low zone immunological tolerance, preceded by a transient phase of antibody production. Tolerance develops most rapidly in the IgE antibody isotype, followed by suppression of delayed-type hypersensitivity and IgG production. Importantly, the tolerance mechanism is operative only under steady-state conditions, when inert antigens are presented via an intact respiratory epithelium. In contrast, inflammatorystimuli tend to override the tolerance induction and promote T-cell sensitization (McMenarnin ct al., 1991).Earlier findings that yS T cells are prevalent at epithelial/mucosal surfaces and tend to accumulate at sites of inicrobially induced inflammation (Augustin et al.,
1989; Janeway et nl., 1988) suggested a potential role of these cells in the maintenance of immune hoemostasis and tolerance. Early evidence in support of this idea came from a study of Brown Norway rats, tolerized to OVA by repeated aerosol exposure (McMenamin et al., 1991). In this model, the state of antigen-specific tolerance was particularly evident in IgE production. The tolerance was transferable using splenic T cells from tolerant rats. Based on sequential depletion, the cells mediating tolerance were CD3’, CD4-, CD5+,and CD8+,TCR ap- but TCR-y mRNA’”, suggesting that they may be part of the y6 T-cell lineage. In this model, as few as 1000-2000 cells were sufficient to adoptively transfer tolerance into adult rats via intraperitoneal injection. In a later study with the rat model (McMenamin et al., 1995), investigators from the same group confirmed that the tolerogenic cells were y6 T cells by staining with antibodies specific for rat 7 6 TCRs (Lawetzky et al., 1990). In addition, these y6 T cells were found to produce high levels of IFN-y in response to OVA stimulation in vitro, suggesting a mechanism for the inhibition of TH2-type antigen responses, and therefore for the selective suppression of JgE antibody production. Based on these findings, it was further proposed that y6 T cells, with their unique potential of selective IgE antibody suppression in response to mucosal antigen exposure, may play an important role in protection against primary allergic sensitization in vivo (McMenamin et al., 1995). Essentially the same experiment was carried out also with C57 BLJ6 mice (McMenamin et a!., 1994). Again, very small numbers of y6 T cells from OVA-tolerized mice selectively suppressed TH2-dependent IgE antibody production, without affecting parallel IgG responses. Challenge of these cells in vitro also resulted in high levels of IFN-y production. Although the mechanism of the tolerogenic effect was not resolved, it seemed to be mediated either by inhibition of the expansion preexisting OVA-specific TH2 cells or of the development of OVA-specific TH2 as opposed to TH1 T cells. The antigen-specific inhibitory effect of y6 T cells suggests that they are themselves capable of recognizing OVA, and perhaps other soluble antigens. Given that systemic immunizations with soluble antigens usually have failed to produce antigen-specific y6 T-cell responses, this could reflect particular, mucosa-associated requirements for y6 T-cell sensitization, a very exciting possibility indeed. However, until clonal specificities for soluble antigens have been identified and characterized, it cannot be excluded that the antigen specificity of inhibition seen in vivo or in bulk cultures in vitro is derived from ap T cells, and selective but antigennonspecific interactions of y6 T cells with them. Observations quite similar to those with OVA have been reported with aerosolized insulin (Harrison et nl., 1996). Insulin acts as an autoantigen
I M b l U N O H E G U L A T O R Y F L J N C T I O N S OF y6 T CELLS
117
in humans and nonobese diabetic (NOD) mice suffering from insulindependent diabetes mellitris (IDDM). When NOD mice were treated with an insulin aerosol after the onset of' subclinical disease, pancreatic islet pathology and the incidence of diabetes were both reduced significantly. Insulin-treated mice showed increased circulating antibodies to insulin, a lack of splenocyte proliferation to the major insulin epitope (insulin B chain amino acids 9-23), and associated increased levels of IL-4 and IL10 secretion. In contrast, splenocyte proliferation in response to glutamic acid decarboxylase, another islet autoantigen, was reduced. Splenocytes from insulin-treated mice were found to be capable of suppressing the adoptive transfer of diabetes to nondiabetic mice using T cells derived from diabetic mice, and the suppression was shown to be dependent on CD8' y6 T cells. However, >lo0 times more cells were needed than in the model of tolerance to OVA, indicating that frequencies of y6 T cells capable of mediating tolerance to insulin are far lower or that they are less effective. Interestingly, in the model of insulin-induced tolerance, not only are insulin-specific responses suppressed, but responses to another unrelated antigen as well, consistent with an antigen-nonspecific regulatory function of y6 T cells. In this regard, the model of insulin tolerance appears to differ from that of OVA tolerance. It thus remains to be seen whether y6 T cells affect tolerance to inhaled antigens primarily by antigen-specific or nonspecific mechanisms and whether y6 T-cell-dependent tolerance to inhaled antigens can be exploited in therapeutic strateges aimed at preventing huinan IDDM and other diseases.
3. Grcfis of Tissue Transplantation of tissues can induce tolerance by the recipient to the alloantigens expressed on the transplant. This tolerance tends to be donor specific and can be transferred by donor specific T cells with suppressor qualities ( Hutchinson, 1986). Investigating the transplant-specific CTL repertoire of a patient with a long-term surviving HLA mismatched hdney graft, the alloantigen specificities of donor-specific CTL clones were determined (Vandekerckhove et al., 1990).Among 14 clones derived from phytohemagglutinin (PHA)-activated peripheral blood mononuclear cells of this patient, two expressed TCR yS. One of these y6 T-cell clones showed specificity for HLA-9 and the other for DQw6. Data indicated that the precursors of CTL specific for donor class I and I1 HLA antigens were not detected from the long-term tolerant recipient, and in fact that part of the donorspecific CTL activity resided within the y6 T-cell population. Late after transplantation ( > 1 year), y6 T cells were also found at increased frequencies in T-cell cultures derived from endoinyocardial biopsies of human heart traiisplant recipients (Vaessen et nl., 1991).In contrast,
118
WILI.1 BORN
ef
nl.
increases of y6 T cells in the peripheral blood of these patients were not seen. Unlike the situation after kidney transplantation, donor-specific cytotoxicity of biopsy-derived y6 T cells was not found. It was suggested that the presence of alloantigen-nonresponsive 7 6 T cells late after heart transplantation, when acute cellular rejection episodes are rare, may reflect their role in downregulating host responses against the allograft. However, in another investigation of patients with heart transplants, inverted Val/ V62 ratios were found among peripheral blood 7 6 T cells (Vaessen et nl., 1996). The relative increases in V a l t y6 T cells were not a result of immunosuppressive medication in the heart transplant patients, since a control group receiving similar doses of cyclosporin A did not show the reverted V6 ratios. Again, no donor HLA-specific y6 T cells were found, suggesting that the response of V a l + and y6 T cells is driven by the recognition of other ligands. In this regard, heat shock proteins have been considered as candidate ligands (Duquesnoy et al., 1991). Unresponsiveness of the host to allogeneic grafts has been studied extensively in animal models. For example, C3H/HeJ mice injected via the portal vein with multiple minor incompatible B1O.BR cells exhibit delayed rejection of subsequent B1O.BR skin grafts. Inhibition of C3H-anti-B1O.BR immunity (in vivo and in vitm) was mediated by and was transferable with CD4-8- y6 T cells (Gorczynski, 1994). This y6 T-cell-mediated delayed skin graft rejection was found to be associated with suppressed IL-2 production, and preferential retention of IL-4 production, by cells stimulated in vitru. In a subsequent study, the same group of investigators found that in vivo y6 T-cell depletion of the recipient with a monoclonal anti TCR6 antibody (mAb UC7.13D5) abolished the prolongation of graft survival normally seen after portal veinous immunization with donor cells (Gorczynski et al., 1996a).In addition, the antibody treatment reversed a bias toward IL-4IL-10 expression normally present in the tolerized animals. This reversal toward TH-2-type cytokines was maintained in vitro. Consistently, antigen-specific delayed rejection of allogeneic skin grafts could not be demonstrated in TCR 6 knockout mice (Gorczynski et al., 1996b). Taken together, data suggested that portal vein immunization leads to oligoclonal expansion of a subset of y6 T cells involved in the inhibition of allograft rejection. Hybridomas generated with y6 T cells derived from Peyer’s patches of tolerized mice responded to alloantigen stimulation. Surprisingly, after adoptive transfer, these hybridomas could delay skin graft rejection. This ability was correlated with the abilities of the hybridomas to produce IL10 in vitro and could be neutrahzed in vivo by injecting the mice with antibodies specific for IL-10 and TGF-P (Gorczynski et nl., 1997). Moreover, on restimulation in vitru, hybridomas producing IL- 10 polarized
cytokine production by freshly isolated inesenteric lymph node cells away from production of IL-2 and IFN-y, toward IL-4, IL-10, and TGF-P. Hybridoinas capable of mediating delayed allograft rejection were also analyzed for the TCR y chain repertoire associated with this function (Gorczynskict ul., 1997; Sun et NI., 1997).Two types of cells were identified: one expressing TCRs with a Vy7-J4 coiiimon junctional sequence and a second with more diverse TCR genes. However, graft prolongation was mediated by liotli and was always associated with IL-10 and/or TGFp production. No obvious correlation between ligand specificities and functional properties was tliscernible. In addition to being capable of delayrig allograft rejection, 76 T cells were also found to protect recipients of cellular allografts against graft versus host disease (GVHD) (Shiohara et a/., 1996). This was found in a mouse model in which a CEi7BWG-derived autoreactive T-cell clone specific for 1-A” was injected iiitradernially into the footpad of syngeneic inice. The grafted cells migrate into the epidermis where they cause histological changes very similar to those seen in human cutaneous GVHD (Shiohara et al., 1988; Shiohara et nl., 1987). In this model, cutaneous GVHD is selflimiting, and the spontaneously recovered epidermis becomes resistant to subsequent induction of GVHD (Shiohara et al., 1990). After grafting of the clone, host-derived y6 and ap T cells were found to be increased locally, leaving it uncertain whicli subset might contribute to host resistance. However, although TCR 6 knockout inice also developed cutaneous GVHD lasting even longer than in normal controls (heterozygous littermates), after recovery they did not become resistant to subsequent cutaneous GVHD, despite the presence of a large population of ap T cells apparently taking the place of epidermal y6 T cells in these mice. Resistance against cutaneous GVHD could be restored by reconstituting epidermal y6 T cells through injections with day 16 fetal thymocytes from normal mice. This indicated that epidermal y6 T cells are responsible for local protection against cutaneous GVHD. Acute GVHD is a major complication of allogeneic bone marrow transplantation. In the early ‘stages, this disease primarily affects the intestinal mucosa, resulting in increased local mitotic activity, crypt hyperplasia and apoptosis, and MHC class I1 expression on the surfaces of villi and crypts. Enteropathy during acute GVHD seems to be mediated by cytokines released by T cells of donor origin responding to the host antigens, but host-derived T cells may also participate in the pathogenesis. In an attempt to determine the relative contributions of these T-cell populations, a kinetic study was performed with (CWBL/C,xDBA/2)Fl mice, after injection of parental (C57BW6) splenocytes (Tsuzuki ct d., 1994). Host-derived intraepithelial lymphocytes were found to increase in the intestines within 3
120
CVILLI BORN et nl
days after injection of the graft, whereas donor-derived cells did not appear within intestinal epithelia until day 12. Around the same time, host-derived IEL, especially y6 T cells, decreased following a sudden rise in apoptotic cells. Whether host-derived y6 T cells initially contribute to or help prevent enteropathy was not resolved. In addition to data indicating that host-derived y6 T cells can downregulate antigraft responses and perhaps diminish GVHD, evidence also suggests that they contribute, under certain conditions, to the graft-versushost response (Ellison et al., 1995). In a mouse model of GVHD essentially identical to the one used for kinetic studies on GVHD-induced enteropathy, it was found that donor-derived y6 T cells exhibit natural killer-like cytotoxicity. Depletion of y6 T cells from donor mice altered the outcome of GVHD with a significant reduction in mortality. It thus appears that whereas host-derived y6 T cells can help increase graft survivaI by modulating alloantigen responses between host and graft, donor-derived y6 T cells sometimes contribute to disease development through their cytolytic effector functions. The importance of the disease-enhancing activity of 76 T cells is difficult to estimate. In a model of skin xenograft rejection in which rat skin was transplanted into mice, no evidence for a rejection enhancing activity of host-derived y6 T cells was found, and ap T cells seemed to be entirely responsible for the rejection reaction (Nishimura et al., 1994). 4. Pregnancy
Because of the fact that specific populations of y6 T cells colonize the nonpregnant female reproductive tract, one might expect that they also play a role during pregnancy. Indeed, monoclonal antibodies specific for human y6 TCRs were found to detect determinants within the human endometrial glandular epithelium, although the nature of these binding sites was initially unclear (Yeh et al., 1990). Characterizations of the lymphocytes present in human and niurine decidual tissues, however, soon revealed the presence of large numbers of y6 T cells. Studies of human decidual leukocytes from early pregnancies showed that 60% of decidual T cells expressed y6 TCRs (Mincheva-Nilssonet al., 1992, 1994).Whereas ap T cells and NK cells were found mostly in large lymphoid cell clusters (LCC) near endometrial glands, y6 T cells were present in LCC but, in addition, also as intraepithelial lymphocytes in glandular epithelium. They expressed a CD4-CD8-CD56+ profile and about half expressed the memoryhctivation markers CD45R0, and Kp43 and/or HML-1, as well as MHC class 11. A similar study in mice also revealed high percentages of y6 T cells at the maternal-fetal interface if compared to maternal spleen (Heyborne et at., 1992). In terms of absolute numbers, we have estimated
that reproductive tract y6 T cells are increased nearly 100-fold in pregnant animals compared with nonpregnant animals. I n allogeneic pregnancies, essentially all y6 T cells at the inaternal-fetal interface were shown to be derived maternally. Many y6 T cells from pregnant mice expressed receptors for IL-2, suggesting their recent activation, whereas y6 T cells from nonpregnant uteri did not. To characterize tlie TCR y6 repertoire in the placentddecidua, 2 1TCR 76-bearing hybridomas generated from lymphocytes in this tissue were analyzed. At least 6 different TCR y6 types were found, but cells expressing a canonical TCR composed of Vy6N61, previously found in nonpregnant uteri, were predominant. Uterine intraepitlielial y6 T cells have also been characterized in sheep (Meeusen et al.,1993). Again, a large fraction w a s found to be activated, indicating that in uiuo activatioii of y6 T cells can occur in tlie absence of any pathological stimuli. It is not yet clear what activates y6 T cells at the maternal-fetal interface. However, it has been found that Iiybridoinas generated with y6 T cells expressing Vyl were stiiniilatetl by freshly prepared trophoblasts and by a trophoblast cell line (Heyborne et al., 1994). This response was TCR dependent, but did not require tlie pr'sence of /32-microglobulinon the stimulating cells. Recognition of troplioblasts by maternal y6 T cells may thus be part of a regulatory ineclianisin in iniirine pregnancy. More recent data suggest that y6 T cells at the inaternal-fetal interfke secrete immunoregulatory cytokines, including TNF-a! and TGF-0 (Suzuki et ol., 199s). In this mouse model of allogeneic pregnancy (C3H/He X AKWJ), numbers of intrauterine y6 T cells increased more than in a syngeiieic control (C3H/He X C3H/He). A y6 T-cell-enriched fraction from the uterine IEL exliibited suppressive activity against alloantigen responses of nonpregnant C3H/€Ielyinph node cells, and the siippression was found to be mediated by TGF-P. It was concliided that y6 T cells among uterine IEL suppress the maternal anti-fetal immune response to prevent rejection of tlie fetus and that this effect is, at least in part, inediated by TGF-0. In a different niouse inodel of dogeneic pregnancy (CBMJ X DBA/2), it was also found that alloiiiiinuiiizatiori against the paternal H-2l antigen shifts cytokine responses of y6 T cells in the uterus away from TH1 and toward TH2 types (Arck ( I t d., 1997; Clark d, 1998), but the importance of y6 T cells in preventing rejection of the fetus was less emphasized. Although essentially all y6 T cells at the niaternd-fetal interface appear to be derived maternally, one early study showed that human fetal liver contains progenitors of y6 T cells capable of cytotoxicity and recognizing 1992). This MHC class I antigens on maternal T cells (Miyagiwa et d., observation was interpreted to suggest that fetal liver-derived y6 T cells 671
122
WILL.1 BORN
(’1
nl
may play a role in protecting the fetus against invading maternal T cells generated in the feto-maternal interaction.
F.
HEALING AND EPITHELIAL HOMEO5TASIS Given that murine Vy5N6lCy6 T cells exclusively colonize the epidermis, it seems most probable that their functions are associated specifically with this tissue. Indeed, it was found that epidermal y6 T cells support the growth of keratinocytes (Boismenu and Havran, 1994). This effect is mediated by y6 T-cell-derived keratinocyte growth factor (KGF),a member of the fibroblast growth factor family (Finch et al., 1989). KGF is a potent mitogen for keratinocytes and appears to play a role in wound healing (Guo et al., 199:3; Werner et a/., 1992, 1994). Interestingly, other types of epitheliallmucosal y6 T cells, but not lymphoid a@ or yS T cells, have also been found to produce KGF, suggesting a specific interdependence between y6 T cells and their surrounding epithelial cells (Boismenu and Havran, 1994; Rakasz et al., 1996). However, because KGF is also produced by dermal fibroblasts (Finch et al., 1989), the question arises as to when y6 T cell-derived KGF production could be important. Although it is possible that KGF production by y6 T cells is merely redundant, it seems far more likely that they produce this factor under conditions when fibroblasts do not, e.g., in response to signals that are only perceived by 76 T cells, and possibly through the 76 TCR. The ligand(s) for murine epidermal y6 T cells is not yet known. Nevertheless, available evidence points to keratinocytes themselves as a source of stirnulatory ligands. Specifically, the coculture of epidermal y6 T cells and keratinocytes led to y6 T-cell activation, whereas other cells were not stimulatory (Havran et al., 1991; Tigelaar and Lewis, 1994). Moreover, only injured (“stressed”) keratinocytes or transformed keratinocyte cell lines had the ability to stimulate. Consistently, keratinocyte damage in situ, caused by contact sensitizing agents, induced localized proliferation of epidermal y6 T cells, whereas the same agents did not stimulate these T cells directly (Huber et al., 1995; Kaminsky et nl., 1993; Rheins et al., 1987). Finally, antibody inhibition with a mAb specific for the TCR, as well as TCR gene transfer experiments, confirmed that the responses of epidermal y6 T cells to keratinocytes are TCR dependent and are therefore probably based on the recognition of a keratinocytederived ligand (Havran et al., 1991). Taken together, these data strongly suggest that keratinocyte damage can stimulate a local yS T-cell response, capable in turn of promoting keratinocyte growth and wound healing (Boismenu et al., 1996). Further evidence that y6 T cells are involved in epithelial homeostasis comes from studies of intestinal epithelia of mice that are congenitally WOUND
Ihlh4UNOREC.IJLATOHY FUNCTIONS OF yS T C:EI.I,S
123
deficient in y6 T cells. Murine intestinal epithelia contain large populations of y6 T cells distinct from those present in the epidermis. It was observed that the absence of y6 T cells correlates with a reduction in epithelial cell turnover, as measured by bromo-deoxy-uridine ( BrdU) staining (Koinano et al., 1995).Also, overall crypt cell numbers were reduced in TCR 6-’- mice. In addition, epithelial differentiation was diminished. Intestinal epithelial differentiation coincides with the migration of the tissue cells from the crypt to thevillus (De Bothetal., 1975;Quaroni and Isselbacher, 1985).Correlated with this migration are changes in the expression of several brush border enzymes and increases in MHC class I1 expression (Kellyet al., 1987; Parr and McKenzie, 1979). In TCR-S-’- niice, surface levels of MHC class I1 molecules were found to be inuch reduced, suggesting that y6 T cells promote the terminal differentiation of intestinal epithelia. The influence of y6 T cells on intestinal epithelial differentiation likely not only affects normal homeostasis but also tissue integrity during infections. Studies with the parasite E. tjemiifonnis, which infects epithelia of inice as well as inany other vertebrates, support this notion. An involvement of y6 T cells in the host response to this pathogen was initially suggested by dynamic changes in these cells during infection (Findly et al., 1993). A later study revealed that whereas inice laclang ap T cells had defects in protective immunity against the parasite, mice lacking 76 T cell suffered from increased intestinal damage, perhaps due to a failure to regenerate injured intestinal tissues and/or to regulate the consequences of the aP T-cell response (Roberts et al., 1996).In this situation, whether the primary contribution of y6 T cells lies in tissue preservation or in iminunoregulatiori or whether the combination of both is a hallmark of y6 T cell function remains to be seen. V. Concluding Remarks
Lymphocytes expressing 76 TCKs appear to have an evolutionary history as extensive as that of the other antigen receptor-bearing subsets. Therefore, their complex response pattern under noriiial and pathological conditions does not seem surprising. To fillly understand y6 T cells, these various responses likely will have to be dissected with as much care as is currently being bestowed upon a0 T cells and B cells. However, a few distinctions of y6 T cells seem to be already evident. Thus, at least in inice and huinans, they show little involvement as effectors of host defenses but seein much involved in their regulation. Specifically,they seein to modulate both innate and adaptive, antigen-specific responses of the immune system, in addition to interactions with cells outside of the immune system. As with ap T cells and B cells, ligand specificities of y6 T cells likely are diverse. Additional
124
\VILLI HORN
('f
rrl
examples of proven ligand recognition are much needed to define the nature of the stimuli likely to trigger TCR-dependent 76 T-cell functions. If y6 T cells indeed have greater bias for autologous ligands than c.P T cells or B cells, this would be consistent with a functional bias for iinmunoregulation in the absence of antigen, e.g., in noninfectious inflammation. However, available data on the role of y6 T cells in antigen-specific tolerance are best reconciled with the existence of at least a small subset of yS T cells capable of recognizing foreign antigens. Finally, polyclonal TCR-dependent and subset-specific responses of y6 T cells to nonpeptidic and peptidic ligands, unlike any of the antigen responses of ~$3 T cells, remain an unexplained phenomenon. Do such polyclonal stimuli initiate immediate defenses without need for prior clonal expansion or do they lower the threshold of more specific, clonotypic reactivity? If y6 T cells are indeed iminunoregulators, they could become a major tool of more physiological immunotherapeutic strategies. Specific targeting of these cells and their functions therefore s e e m to be a worthwhile task for the years ahead. ACKNOWLEDGMENTS
The authors thank their colleagues at National Jewish Medical and Research Center for their support and critical discussions and Sharon Forsberg for her expert assistance in preparing this manuscript.
REFERENCES Allison, J. P., and Iiavran, W. L. (1991). The inimunobiology of T cells with invariant yS antigen receptors. A n w . Rcu. ZttitnutwZ. 9, 679-705. Altnian, J. D., Moss, P. A. H., Goulder, P. J. R., Barouch, D. H., McHeyzer-Williams, M. G . . Bell, J. I., McMichael, A. I., arid Davis. M. M. (1996). Phenotypic analysis of antigen-specific T lymphocytes. Srirnre 274, 94-96. Arasca, H., Ono, S., Arase, N., Park, S. Y., Wakizaka, K., Watanabe, H., Ohno, H., and Saito, T. (1995). Developmental arrest of NKl.lt T cell antigen receptor (TCR) alplidbeta' T cells and expansion of NK1.1' TCR-gamni;ddelta+ T cell development in CD3 zetadeficient mice. /. Exp. Med. 182,891-895. Arck, P. C., Ferrick, D. A., Steele-Noiwood, D., Croitoru, K., and Clark, D. A. (1997). Murine T cell determination of pregnancy outcome. I. Effects of strain, ab T cell receptor, gd T cell receptor, and gd T cell subsets. Am. /. Reprod. h t u m d . 37,492. Arden, B., Clark, S. P., Kabelitz, D., and Mak, T. W. (1995). Mouse T-cell receptor variable gene segment fainihes. Inininnogenetics 42,501-530. Asarnow, D.M., Cado, D., and Raulet, D. H. (1993). Selection is not required to produce invariant T cell receptor ?-gene junctional sequences. Nature 362, 158-160. Asarnow. 11. M., Goodman, T., LeFrancois, L., and Allison, J. P. (1989). Distinct antigen receptor repertoires of two classes of inurine epithelium-associated T cells. Nature 341, 60-62. Asarnow, D. M., Kuziel, W. A., Bonyhadi, M., Tigelaar, R. E., Tucker, P. W., and Allison, J. P. (1988).Limited diversity of y6 antigen receptor genes of Thy-lt dendritic epidermal cells. Ccll 55, 837-847.
Askenase, P. W., Szczepanik, M., I’tak, M., Paliwal, \’.. and Ptak, W. (1995). y6 T cells i n norinal spleen assist inimunizd (YPT cells in the adoptive cell transfer ofcontact sensitikity. /. ~ I t l t I I U l I ~ J154, ~. 3644-36 Augustin, A , . Kubo. R. T., ;in( i n . (;.-K. (1989).Rcsident p u l n i o n a ~lymphocytes espressiiig the y/6 T-cell recepto ntrirr’ 340, 239-241. Azrial-a, V., Levraud, J. P., Lemhezat, M. P., d Pereira. P. (1997). A novel subset of adiilt gamma delta tliymocytes that secretes a distinct pattem ot‘cytoki restricted T cell receptor repertoire. Eirr. /. Z i r i i i w t d . 27, 544 Bahji, K. N., Schwander, S. K., Rich, E. A., antl Booiii, \\’. H. (19Yij).Alveolar inacrophages as accessory cells for hunian yS T cells activated by A4ycobc~rteririii~trthcrcrdosi.u. J r t l t t ~ ~ t l t i o i .154,S Y S ~ - S Y ~ K . Bancliereau, J., Bizan. F., B l a ~ ~ c l D.. ~ d Brii.re. . F., Galizzi, J. P., van Kooten, C.,Liu, Y.J., Roiisset, F., and Saeland, S. (1994). The CD40 antigen antl its ligand. Aiinrr. Rpt;. r ~ ~ ~ t t l f L t12, l o i . 881-922. Bancroft, G. J., Schreiber. R. D., Bosina, G. C . , Bosina. M. J., and Unaiiiie, E. R. (1987). T cell indepentlent ineclianism of niacrophageactivation 11) interferon gainma./. r t r 1 t t r r r t d 139, 1104-1107. Bandeira, A . , Itohara, S.,Boniieville, M., Biirleii-Defrarious, O., Mota-Santos, T., Coutinlio. A., and Tonegawa, S. (1991), Extr;ltliymic origin of intestinal intraepitlielial lymphocytes bearing T-cell antigen receptor y8. Pi-oc. Not!. A c d . Sci. USA 88,43-47. Bank. I., DePinHo, R. A , , Brenner, M. B., Chssiineris. 1.. Alt, F. W., and Chess, L. (1986). A functional T3 inolecule associated with a novel lieterodimer on the surface of immature human thymocytes. Nature 322, 179-181. B m k , I . , Tanay, A , , Hook, M.. antl Lkiieli, A. (1995). VyY-V62‘ gainma delta T cells froni a patient wit11 Felty syidronie that exhibit aberrant response to triggering of the CD3 nioleciile can regulate immunoglot~iilinsecretioii I)y H cells. Cliii. Zinniiinol. Zitmiiiiopathol. 74, 162-169. Brndelac, A , , Lantz, 0..Quimby, M. E., Yewtlell, J. \4’., Ueiinink, J. H.,and Briitkiewiciz, R. R. (1995). CD1 recognition by n i o i w NK1’ T lyitiphocytes.Scicvicc 268,863-865. Bergstresser, P.R., Tigelaar, R. E., Dew, J. 1-1.. a r i d Streilrin. J. W. (1983).Thy-1 antigenbearing dendritic cells popdate niuiine epiderinis. /, ltia&. Dcnticrtol. 81, 286-288. Bluestone, J. A , , Coon, R. Q., Barrett, T. A , . Hodden, B., Sperhng, A. I., Dent. A,, Hetlrick, S.,Rellalian, B., and Matis, I,. A. (19911. Repertoire, drvdopnirnt and ligand specificity of murine TCR y6 cells. Z i t i n i i r i i o / . Hot; 120,5-33. Bluestone, J. A,, Cron, R. Q., Cotterman. M.. Horilden. B. A., and Matis. L. A. (1988). Structure arid specificity of T cell rrwptor y/S on major Iiistocoinpatibili~coinplex antigen-sptlcific CD3’, CD4-, CD8- T Ivmplioc>tes.1.Esp. A4rd. 168, 1899-1916. Boismenu, R., and Havran, if’. I,. (1994). M&latioii ofepithelial cell g r o d i by intraepithelial y6 T cells. Sckv~w266, 1253-1255. Boisinenu, R., Hobbs, M. V., Boiillicr, S.. a d FIavran, W. L. (1996). Molecrilar and cellular biology of dendritic epidermal T cells. S c i n Z t t i i T i i i i i o / , 8, 323-331. Boniieville, M., Isliida, I., Monrbaerts, P., Katsriki, M., Verbeek, S.,Berns, A., and Tonegawa, S. (1989a). Blockage of ap T-cell developinerit by TCK y6 transgenes. Notzrrc 342, 931-934. Bonnecdle, M.. Ito, K., Krecko, E. C;., Itoliara. S.. Kappes, D., Ishida, I., Kanagawa, 0.. Janeway, J. C. A,, Murphy, D. B., and Tonegawa, S. (l989b).Recognition o f a self major histocompatibility coinplex TL region procliict I)?, yS T cell receptors. Pro(,. Ncitl. Acnd. Sci. USA 86,5928-5932. Boom, Vv’. H., Chervenak, K. A , , Mincrk, M. A,, and Elher, J. E. (1992). Role of the inononticlear pliagocyte as an antigeii-presenting cell for hunian y6 T cells activated by live Mycoboctcriim trihc~rcrcltiais.I i l f ~ t .Z i t u i t u i i . 60,:3480-3488.
126
WILL1 BORN el ol
Born, W., Hall, L., Dallas, A,, Boymel, J,, Shinnick, T., Young, D., Brennan, P., and O’Brien, R. (1990a). Recognition of a peptide antigen 11y heat shock reactive y6 T lymphocytes. Scietice 249, 67-69. Born, W., Happ, M. P., Dallas, A,, Reardon, C . , Kubo, R., Shinnick, T., Brennan, P., and O’Brien, R. (1990b). Recognition of heat shock proteins and yS cell function. Zintntmol. Torlny 11, 40-43. Born, W., Kelly. K. A,, and O’Brien, R. L. (1994). In “Handbook of B and T Lyinphocytes” (E. C. Snow, ed.), pp. 179-214. Academic Press, San Diego. Born, W., Miles, C., White, J., O’Brien, R., Freed, J. H., Marrack, P., Kappler, J., and Kubo, R. T. (1987). Peptide sequences of T-cell receptor 6 and y chains are identical to predicted X and y proteins. Nuture 330, 572-574. Born, W., O’Brien, R. L., and Modlin, R. ( 1 9 9 1 ~ Antigen ). specificity of y6 T lymphocytes. F.4SEB 1.5, 2699-2705. Born, W.. Rathbnn, G., Tucker, P., Marrack, P., and Kappler, J. (1986). Synchronized rearrangement of T-cell y and p chaiii genes in fetal thymocyte development. Science 234, 479-482. Born, W., Yague, J., Paliner, E., Kappler, J., and Marrack, P. (1985). Rearrangement of T-cell receptor P-chain genes during T-cell development. Proc. Nntl. Acud. Sci. USA 82,2925-2929. Born, W. K., Harshan, K., Modlin, R. L., and O’Brien, R. L. (1991b). The role of y6 T lymphocytes in infection. C u r . Opiit. Z t t t r r i t u d 3, 455-459. Brenner, M. B., McLean, J., Dialynas, D. P.. Strominger, J. L., Smith, J. A., Owen, F. L., Seidman, J. G., Ip, S., Rosen, F., and Kraugel. M. S. (1986).Identification of a putative second T-cell receptor. Nuttire 322, 145-149. Bucy, R. P., Chen, C.-L. H., and Cooper, M. D. (1991).Analysis o f y 6 T cells in the chicken. Setti. Itnttiunol. 3, 109-1 17. Bukowski. J. F., Morita, C. T., arid Brenner, M. (1994). Recognition and destruction of virus-infected cells hy liuinan y6 CTL. /. Zttiinirtiol. 154, 5133-5140. Burg, G., Duininer, R., W i h e h , M., Nestle, F., Ott, M. M., Feller, A,, Hefner, H., Lanz, U., Schwinn, A., and Wiede, J. (1991). A subcutaneous delta-positive T-cell lymphoma that produces interferon gamma. N. Engl. /. Med. 325, 1078-1081. Burk, M. R.. Donda, A., Mariani, F., Mori, L., and De Libero, G. (1997). Functional inactivation in the whole population of human Vy9N62 T lymphocytes induced by a nonpeptidic antagonist. 1.Exp, M e d 185, 91-97. Burk, M . R., Mori, L., and De Libero, G. (1995).Human Vy9-V62 cells are stimulated in a cross-reactive fashion by a variety of phosphorylated metabolites. Etrr. /. I r r i t r i t i n o l . 25, 2052-2058. Bux, E., Matsunaga, R., Nagatani, T., Walden, P.. Nagy, Z. A,, and Klein, J. (1985).Distribution of alloreactivity among antigen-specific, class 11-restrictedT-cell clones and hybridonias. Zinttiu tiogeitetics 22, 189- 192. Carbone, A., Harbeck, R., Dallas, A.. Nemazee, D., Finkel, T., O’Brien, R., Kuho, R., and Born, W. (1991).Alpha beta T lymphocyte depleted mice, a model for y6 T lymphocyte functional studies. Ztnniwiol. Reo. 120, 35-50. Carding, S. R. (1990). A role for yl6 T cells in the primary immune response to influenza virus. Res. I t t t i t ~ t i r w Z . 141, 603-606. Carding, S. R., Allan, W., Kyes, S.. Hayday. A,, Bottomly, K., and Doherty, P. C. (1990). Late dominance of the inflainmatory process in murine influenza by y6+ T cells. /. Exp. AIM!. 172, 1225-1231. Chi, D. S., Grebenau. M. D., and Thorbecke, G. J. (1980). Antigen-induced helper and suppressor T cells i n normal and agamniaglobulinemic chickens. Ezrr. I, Irnmunol. 140, 20:3.
IMMIIN'OHEGULATOMY FLJNCTIONS OF y6 T CELI.5
127
Chien. Y.-kl., Iwashinia, M.. Kaplan, K.. Elliott, J. F., and Davis, M. M. (1987a). A new T-cell receptor gene located within the alpha locus and expressed early in T-cell differentiation. Nature 327, 677-682. Chien. Y.-H., Iwashiina. M., M'rttstein, D. A , , Kaplan. K. B.. Elliott, J. F., Born, W., arid Da\is, M. M. (198711).T-cell receptor 6 gene reurrangenirnts in early thymocytes. Nature 330, 722-727. Chien. Y.-H.. Jores, H., and Crowley, M. P. (1996).Recognition of $6 T cells. A t m i . Rec. Z t t m i t i nol. 14, 511-532. Christmas, S. E., and Meager. A. (1990). Production of interferon-gainma and turnour necrosis factor-alpha by human T-cell clones expressing different forms of the y6 receptor. Zmrtititiology 71, 486-492. Clark, D. A . , Chaouat. G.. Arck. P. C . Mittruecker. H. M., and Levy, G. A. (1998). Cutting edge: Cytokine-dependent abortion in CBA X DBN2 mice is niediatetl by the procoagulant fg12 prothonibinase. /. Zttiitiiitiol. 160, 545-549. Clark. S. P.. Arden. B., Kabelitz, I)., and Mak, T. W. (1995).Comparison of human and mouse T-cell receptor variable gene segment suhfainilies. Irtittwrnogerietics 42, 531-540. Constant, P.. Davodeau. F., Peyrat, M.-A,, Poqnet, Y.. Piizo, C., Bonneville, M., and Fournii.. I.-]. (1994).Stinidation of huinan y6 T cells by nonpeptitlic inycobacterial ligands. Science 264, 267-270. Cooper, C. L., Mueller, C., Sinchaisri. T.-A,. Pirinez, C., Chan, J., Kaplan, G., Young, S. M. M., Weissnian, I. L., Bloom, B., Kea, T. H., and Modlin, R. L. (1989a). Analysis of naturally occuring delayed-type 1i.yprsensitivity reactions in leprosy by in situ hybiidization. /. EXTJ.Alecl. 169, 1565-1581. R. P., George, 1. F., Lahti. J. M.. Char, D., and Chen, C.-L. H. lopinent in birds. Cold Spritig Narb. Sytnp. Quant. B i d . 54, 69-70. Davis, M. M.. and Bjorkman, P. J. (1988).T cell antigen receptor germ and Tcell recognition. Notic re 334, :395-402. Davodeau, F., Hoiide, I., Boulot, G., Komagni., F.. Necker, A,, Canavo, N., Peyrat, M. A,, Hallet, M. M., \'ii., H., Jaccpes, Y., Mariuzza. R., and Bonne\ille, M. (1993a). Secretion of disulfide-linkedhuman T-cell receptor y6 heterodimers.]. Biol. Chetn. 268, 15455-15460. Davodeau, F., Peyrat, M.-A,, Iloude. I., Wallet, M.-M.. De Libero, G., Vie, H., and Bonneville. M , ( 1993b).Surface expression ofhvo distinct functional antigen receptors on human y6 T cells. Scicrice 260, 1800-1802. De Both, N . J., Van der Kainp, A. W., and Van Dongen. J. M . (1975). The influence of changing crypt cell-kinetics on functional differcwtiation i n the small intestine of the rat. Diffc,rcvrticitLoti 4, 17,5- 182. De Libero. G., Casorati, G.. Giachino, C.. Carbouara, C., Migone, N.. Matzinger, P.. and Lanzavecclria, A. (1991), Selection by two powerfill antigens may account for the presence of the major population of huinan peripheral y/S T cells. 1. Ewp. Med. 173, 1311-1322. DeLihero, G. ( 1997). Sentinel function of broadly reactive human gainina delta T cells. Z r t t t t i u t w / . 7'ochy 18, 22-26. Deinbic, Z., kind L'idovic, D. ( 1990). Specificity and finictiolr of Qa-1 restricted y6 T cells. RES.Z t t w w t i d 141, 600-602. De Paoli. P., Gennari, D., Martelli. P., CaLwzerani, V., Comoretto, and Santini, G. (1990). y6 T crll receptor-hearing lymphocytes during EpsteidBarr virus infvction.1.Ztfect. Di.s. 161, 1013-1016. Deusch, K., Luling, F.. Reich, K., Classcii. M., Wagner, €I..and Pfeffer, K. (1991). A major fraction of huinan intra-rpithelial lymphocytes simultaneously expresses the gamma delta T crll receptor, the CD8 accessoiy molecule and preferentially uses the V61 gene segment. h r . /. Ztntrinrio/. 21, 1053-1059.
128
WILL1 BORX et ol
DeVillartay, J,-P., Hockett, R. D., Coran, D., Korsmeyer, S. J., and Cohen, D. I. (1988). Deletion of the human T-cell receptor &gene by a site-specific recombination. Nature 335, 170-174. Diaz, P., Cado, D., and Winoto, A. (1994). A locus control region in the T cell receptor alphddelta locus. Immunity 1, 207-217. Doherty, P. C., Allan, W., Eichelberger, M., and Carding, S. R. (1991).Heat shock proteins and the y6 T cell response in virus infections: Implications for autoimmunity. Spring. Sent. Zmniunopathol, 13, 11-24. Doherty, P. C., Allan, W., Eichelberger, M., and Carding, S. R. (1992). Roles of cup and y6 T cell subsets in viral immunity. Annu. Reu. Zmmunol. 10, 123-151. Doherty, P. C., and Zinkernagel, R. M. (1974). T-cell mediated iinmunopathology in viral infections. Transplant Rev. 19, 89-120. D'Souza, C. D., Cooper, A. M., Frank, A. A,, Mazzaccaro, R. J., Bloom, B. R., and Orme, I. M. (1997). An anti-inflammatory role for y6 T lymphocytes in acquired immunity to Mycobncteriurn tuberculosis. 1.Zmmunol. 158, 1217-1221. Dunn, P., and North, R. J. (1991). Early gainnia interferon production by natural killer cells is important in defense against inurine listeriosis. Infect. Zmmun. 59, 2892-2900. Duquesnoy, R. J . , Trager, J. D. K., and Zeevi, A. (1991). Propagation and characterization of lymphocytes from transplant biopsies. Grit. Reu. bnmunol. 10, 455-480. Eilat, D., Kikuchi, G. E., Coligan, J. E., and Shevach, E. M. (1992). Secretion of a soluble, chimeric y6 T-cell receptor-immunoglobulin heterodirner. Proc. Natl. Acad. Sci. USA 89, 6871-6875. Elbe, A., Foster, C. A,, and Stingl, G. (1996).T-cell receptor cup and y8 T cells in rat and hnman skin: Are they equivalent? Immunology 8, 341-349. Ellison, C. A,, MacDonald, G. C., Rector, E. S., and Gartner, J. G. (1995). y6 T cells in the pathobiology of murine acute graft-versus-host disease: Evidence that y6 T cells mediate natural killer-like cytotoxicity in the host and that elimination of these cells from donors significantly reduces mortality. /. Zmmunol. 155,4189-4198. Elloso, M. M., van der Heyde, H. C . , Troutt, A,, Manning, D. D., and Weidanz, W. P. (1996).Human y6T cell subset-proliferative response to malarial antigen in uitro depends on CD4+ T cells or cytokines that signal through components of the IL-2R. J. Zmriunol. 157,2096-2102. Emoto, M., Danbara, H., and Yoshikai, Y. (1992). Induction of ylS T cells in murine salmonellosis by an avirulent but not by a virulent strain of Salmonello choleraesuis. /. Exp. Med. 176,363-372. Emoto, M., Naito, T., Nakamura, R., and Yoshikai, Y. (1993). Different appearance o f @ T cells during salmonellosis between Ity' and Ity' mice. 1. Zmmunol. 150, 3411-3420. Engel, I., and Hedrick, S. M. (1988). Site-directed mutations in the VDJ junctional region of a T cell receptor beta chain cause changes in antigenic peptide recognition. Cell 54,473-484. Ezquerra, A., Wilde, D. B., McConnell, T. J., Sturmhnfel, K., Valas, R. B., Shevach, E. M., and Coligan, J. E. (1992). Mouse autoreactive yl6 T cells. 11. Molecular characterization of the T cell receptor. Eur. /. Immunol. 22, 491-498. Falcao, R. P., Voltarelli, J . C., Simoes, B. P., Pestana, D. N., Zago, M. A,, and Figueriredo, hl. S. (1992). Malignant T ylS lymphoproliferative disease with natural killer lyhc activity. Am. 1. Hemutol. 41, 128-131. Ferrick, D. A., Curnano, A., Furlonger, C., Min, X., Iscove, N. N., Paige, C. J., and Mak, T. W. (1991a). Expansion of myelopoietic precursors and inhibition of B-cell precursors in mice that express a T-cell receptor gamma (Vyl.lJy4Cy4)transgene. I. Cell. Physiol. 148, 457-463.
IblMUNOHEGUI.ATOHY FUNCTIONS OF y6 T CELLS
129
Ferrick, D. A,, Min. X., Gajewski, D. A., and Mak, T. W.(1989a). Phenotypic analysis of mice transgenic for the TCH Cy4 (Vyl.1Jy4Cy4)genr. Cold Spring Ifarb. Sytnp. Q m n t . Biol. 54, 57-60. Ferrick, D. A., Ralieintulla, A., Widacki, S. M., Min, X., Allison, J. P., Bluestone, J. A., van Ewijk. W’.. and Mak, T. W. (1990). Exprcssion of a T cell receptor y chain (Vyl.lJy4Cy4) transgene in mice influences T cell receptor ontogeny and thymic architecture during development. J . Znitnirnol. 145, 20-27. Ferrick, D. A., Sambhara, S. R.. Ballhausen. \7., Iwamoto, A,, Pircher, H., \Yalker, C. L., Yokovaina, W. M., Miller, R . G., and Mak, T. W. (l989h).T cell function and expression are tlrainaticdy altered in T cell receptor Vyl. 1Jy4Cy.l transgenic mice. Cell 57,48:3-492. Ferrick, D. A., Schrenzel, M. D., Mnlvania, T., Hsieh, B., Ferlin, W. G., and Lepper, H. (1995).Differential production of interferon-y w d interleukid in response to Thl- and Th2-stiniulating pathogens by y6 T cells i n oico. Nature 373, 255-257. Ferrick, D. A,, Sydora. B., Wallace, V., Geinmell-Hori, L., Kronenberg, M., and Mak, T. W. (1991b).Self-reactive y6 ly-niphoc\;tes:Implications for T-cell ontogeny and reactivity. Itntniin~d.Rec. 120, 51-69. Fields, B. A,, Ober, B., Malchiodi. E. L., Lehedeva, M. I., Braden, B. C., Ysern, X., Kina, J.-K., Shao, X.. Ward, E. S., and Mariiizza. R. A. (1995). Crystal structure of the V, domain of a T cell antigen receptor. Sciencr 270, 1821-1824. Finch. P. W., Hubin. J. S..Miki, T., Ron, D.. and Aaronson, S. A. (1989). Nuinan KGF is FCF-related with properties of a paracrine effector of epithelial cell growth. Science 245, 752-755. Findly, R. C., Roberts, S. J., and Hayday, A. C . (1995). Dynamic response of murine gut intrarpithelial T cells after infection by the coccitlian parasite Einieria. Eur. 1. Itnrnunol. 23, 2557-2564. Fisch, P., Malkovsky, M., Braakinan. E.. Stwin. E., Bolliuis, R. L. H., Piieve, A,, Sosinan, J. A., Lam, V.A , . ant1 Sondel, P. M. (199Oa).ylS T-cell clones and natural killer-cell clones mediate distinct patterns of non-major liistocoinpatiliility complex-restricted cytolysis. J . E . K ~M. d . 171, 1567-1579. Fisch, P., MalkovsLy, M., Klein, B. S., Morrissey, L. Mi., Carper, S. W.. Welch, W. J., and Sondel, P. M. (199Ob). Human Vy9AirS2 T cells recognize a groEL homolog on Daudi Bnrkitt’s lymphoma cells. S c i c t u 250, 1269- 1273. Fisch, P., Malkovsky, M., Kovats, S.. Sturm, E.. Brankinan, E., Klein, B. S., Voss, S. D., DeMars, R., Welch, L!’. J., Bolhuis, R. L. H., and Sondel, P. M. ( 1 9 9 0 ~ ) . Morrissey, L. W.. Recognition by human Vy9A162 T cells o f a GroEI, homolog on Daudi Burkitt’s Iynphonia cells. Scietlcc 250, 1269-1273. Fisch. P., Oettel, K., Fudim, N,,Surfus, J. E., Malkovsky, M., a i d Sondel, P. M. (1992). MHC-unrestricted cytotoxic and proliferative responses of two distinct human ylS T cell subsets to Daudi cells. 1. Zniniuno!. 148, 2:315-2323. Fremont, D. H.. Bees, W. A., and Kozono, H. (1996).Biophysical studies of T-cell receptors and their ligands. Citrr. Opin.Z n i t n i ~ n o / .8, 93-100. Fu. Y.-X., Cranfill. R., \’olliner, M., van der Zee, R.. O’Brien. R. LA., and Born, W. (1993). I n vivo response of murine y6 T cells to ii lieat shock protein-derived peptide. Proc. Natl. Acad. Sci. USA 90 322-326. Fu, Y.-X., Kersh, G., Vollmer, M., Kalataradi, H., Heybonie, K., Reardon, C.,Miles, C., O’Brien. R., and Born. W. (1994a). Structural requirements for peptides that stimulate a subset of y6 T cells. J. Zmrnrrtrol. 152, 1578-1588. Fu, Y.-X., Roark, C. E., Kelly, K., Drevets, D.. Campbell. P., O’Brien, R., and Born, W. (1994b). Immune protection and control of inflanimatoiy tissue necrosis by y6 T cells. J. ~ ) ~ I W I ~ L153, :3ioi-:3iis.
130
WILL1 BORN ef ul
Fujihashi. K., Kiyono, H., Aicher, W. K., Green, D. R., Singh, B., Eldridge, J. H., and McGhee, J. R. (1989). lmmunoregulatory function of CD3', CD4-, and CD8- T cells: y6 T cell receptor-positive T cells from nude mice abrogate oral tolerance. 1. Inm~z~nol. 143, 3415-3422. Fujihashi, K., Kweon, M.-N., Kiyono, H., VanCott, J. L., van Ginkel, F. W., Yamamoto, M., and McGhee, J. R. (1997).AT celVB celVepithelid cell internet for mucosal inflammation and immunity. Spring. Sem Iinrnunoputhol. 18, 477-494. Fujihashi, K., McGhee, J. R., Kweon, M.-N., Cooper, M. D., Tonegawa, S., Takahashi, I., Hiroi, T., Mestecky, I., and Kiyono, H. (1996). ylS T cell-deficient mice have impaired inucosal immunoglobulin A responses. 1. Exp. Med. 183, 1929-1935. Fujihashi, K., Taguchi, T.,Aicher, W. K., McGhee, J. R., Bluestone, J. A,, Eldridge, J. H., and Kiyono, H. (1992).Immunoregulatory functions for nrurine intraepithelid lymphocytes: y l S T cell receptor-positive (TCR') T cells abrogate oral tolerance, while CYIp TCR+T cells provide B cell help. /. Exp. Metl. 175, 695-707. Fujihashi, K., Taguchi, T., McGhee, J. R., Eldridge, J. H.,Bruce, M. G., Green, D. R., Singh, B., and Kiyono, H.(1990). Regulatory function for murine intraepithelial lymphocytes: Two subsets of CD3', T cell receptor-1- intraepithelial lymphocyte T cells abrogate oral tolerance. J. Zmniunol. 145, 2010-2019. Goodier, M., Fey, P., Eichmann, K., and Langhorne, J. (1992). Human peripheral blood y6 T cells respond to antigens of Pl~~modinitafalciparuin. Znt. Zmmunol. 4, 33-41. Goodier, M., Krause-Jauer, M., and Langhome, J. (1992). Quantitative analysis of the response of human T cell receptor Vy9' cells to Pl~srnorliuInfrZci~~aninL. Eur. I. lrntnunol. 22, 2757-2760. Goodman, T., and Lefrancois, L. (1988). Expression of the y-6 T-cell receptor on intestinal CD8+ intraepithelial lymphocytes. Nature 333, 855-858. Goodman, T., and Lefrancois, L. (1989). Intraepithelial lymphocytes: Anatomical site, not T cell receptor form, dictates phenotype and function. /. Exp. Med. 170, 1569-1581. Gorczynski, R. M. (1994). Adoptive transfer or unresponsiveness to allogeneic skin grafts with hepatic y6+ T cells. Ztnmunology 81, 27-35. Gorczynski, R. M., Chen, Z., Hoang, Y., and Rossi-Bergman, B. (1996a). A subset of y6 T-cell receptor-positive cells produce T-helper type-2 cytokines and regulate mouse skin graft rejection following portal venous pretransplant preimmunization. ~mnuctudogy 87, 381-389. Gorczynski, R. M., Chen, Z., Zeng, H., arid Fu, X. M. (1997). Specificity for in vivo graft prolongation in gd T cell receptor+ hybridomas derived from mice given portal vein donar-specific preimmunization and skin allografts./. Zrnrnur~ol 159, 3698-3706. Gorczynski, R. M., Cohen, Z., Leung. Y., and Chen, Z. (199Gb). y6 TCR' hybridomas derived from mice preimmunized via the portal vein adoptively transfer increased skin allograft survivd i n uiuo /. lwnunol. 157, 574-581. Grabstein, K. H., Eisenman, J., Shanebeck, K.. Rauch, C., Srinivasan, S., Fung, V., Beers, C., Richardson, J., Schoenborn, M. A,, Ahdieh, M., Johnson, L., Alderson, M. R., Watson, J. D., Anderson, D. M., and Gin, J. G. (1994). Cloning of a T cell growth Factor that interacts with the p chain of the interleuldn-2 receptor. Science 264, 965-968. Griffin. J. P., Harshan, K. V., Born, W. K., and Orrne, I. M. (1991).Kinetics of accumulation of y6 receptor-bearing T lymphocytes in mice infected with live mycobacteria. Ittfect. Initnun. 59, 4263-4265. Groh, V., Porcelh, S., Fabbi, M., Lanier, L., Picker, L. J.. Anderson, T., Warnke, R. A,. Bhan, A. K., Strominger, J. L., and Brenner, M. B. (1989). Human lymphocytes bearing T cell receptor yl6 are phenotypically diverse and evenly distributed throughout tire lymphoid system. J. Exp. Med. 169, 1277-1294.
Gnishv, M. J.. Auchincloss Jr, H., Lee, R.. Johnson, K. S., Spencer, J. P., Zijlstra, M., Jaeniscli, R., Papiioannou, V.E., and Glirncher, L,. H. (1993).Mice lacking major liistocolnpatibility complex class I and class 11 nioIeculcs. Proc. Nntl. A ~ n t l Sci. . USA 90, 3913-3917. Ciio, I>.,Yu, Q.-C., and Fuclis. E. 11993).Targeting expression of keratinocyte growth factor to keratinocytes elicits striking changes in epietlrelid differentiation in transgenic mice. EMBO J . 12, 97:3-986. Guo, Y.,Ziegkr, H. K.. SaHey. S. A,, Niesel, D. \%'.,\'aid!-a, S., and Klirnpel, G. R. (1995). Hninan T-cell recognition of Listcrin ttioiioc,!/togote.c: Recognition of listeriolysin 0 by TcRap' and TcRy6' T cells. Zrfcct. I n t i n u t i . 63, 2288-2294. Guy-<:rand, D., Cuenod-Jahri. B., Malassis-Seris. M., Srlz, F., and Vassali, P. (3996). Coniplexity of the niouse gut T cell iniinnne system: Identification of two distinct natural killer T cell intrac~pitheliallineages. Eirr. J. Z t n t n i i n o l . 26, 2248-2256. Haas, W., Pereira. P., and Tonegawa, S. ( 199:3),Gamm;t/delta T cc~lls.Aririri. Hec. Ztnrnirrinl. 11, 637-685. Haas, W., and Tonegawa. S . (19923. Devc~lopnientand selection of y6 T cells. Curt-. Opin. Ztiiniunol. 4, 147-155. tliicker, G., Adam. S.. and Wagner, H. (1995). Interaction hetween y6 T cells and B cells regdating IgC production. Z t t i n ~ ~ m d o 84, g / 105- 110. I-Iapp, M. P., Knbo, R. T., Paliner, E., Horn, 1%'.K., and O'Biien, R. L. (1989). Limited receptor repertoire in a inycohacteria-rractive suhset of y6 T lymphocytes. Nature 342,696-698. IIaregcwoin. A , , Singh, B.. Chpta. R. S., a r i d Finl)erg, K. 1%'.(1990). A inycohacterial heat shock protein responsive y6T cell clone also responds to the honiologous huinan lieat shock protein: A possible link between infection and aiitoimmunity.j. Ztfect. Dis. 163, 156-160. Harrison, I,. C., Denipsey-Collier. M., Krainer, 11. R., and Takahashi, K. (1996). Aerosol insulin indnced regulatory CD8 y6 T cells that prevent niiirine insulin-dependent diabetes. J , EXTI.Ml'tl. 184, 2167-2174. Havlir, D. \I., Ellner, J. J., Chenenak. K. .4,, and Booni, I V . H. (1991). Selective expansion of hrlman y6 T cells b y inonocytes infected with live AlrJcobncferiutn tziberciilosis. J. C h . Znmst. 87, 729-733. Iiavrm, \V., and Allison, J. P. (1988). Deve~opnrentallyordered appearance of thymocytes expressing different T cell antigen receptors. A't/frirc 335, 443-445. Havran, \V. L., Chien, Y.-H., and Allison. J. P. (1991 1. Recognition of self antigens by skinderived T cells with invariant y6 antigen receptors. Science 252, 1430-1432. Hayday, A. C., Saito, H., Gillies, S. D., Kranz, D. M.. Tanigawa, G.. Eisen, H., and Tonegawa, S. (1985). Structure, organization. and somatic iwrrangement of T cell gainnia genes. Cell 40, 259-269. Hayes, S. M.. Sirr, A,, Jacob. S., Siin, C.-K.. and Angustin, A. (1996). Role of IL-7 in the shaping of the pulmonary y6 T cell repertoilr. J. Z t w n t i n d . 156, 2723-2729. Ile, Y. W., and Malek. T. K. (1996). Interleukin-; rcwptor alpha is essential for the development of gaminddelta' T cells. hut not natural ldller cells. 1. E x p . Med. 184, 289-293. IIein, W. R., and Mackay, C. R . (1991). Proininence of y6 T cells in the ruminant immune system. Ztizinrittol. Totlny 12, :30-34. Heinzel. F. P.. Sadick, M. D., Mutha, S. S., and Lockslcy, K. M. (1991). Production of interferon y , interleukin 2. interleukin 1. ;tnd interlenkin 10 by CD4' lymphocytes in eico during hraling and progressive iniirine leishmaniasis. Proc. Nntl. Acud. Sci. IJSA 88, TO1 1-701Fj. Herman, A , , Kappler. J. W., Marrack. P., and Piillen, A. M . (1991). Snperantigens: Mechanism ofT-cell stimulation and role in immune responsrs. Arunn. Reo. Inzmrcnol. 9,749-772.
132
WILL1 BORN
ef
ol.
Herzenberg, L. A. (1989). Toward a layered immune system. Cell 59, 953-954. Heyborne, K., Fu, Y.-X., Kalataradi, H., Reardon, C., Roark, C., Eyster. C., Vollmer, M., Born, W., and O'Brien, R. (1993).Evidence that murine Vy5 and Vy6 yGTCR' lymphocytes are derived from a common distinct lineage. J . Ittmwnol. 151, 4523-4527. Heyborne, K., Fu, Y.-X., Nelson, A,, Farr, A,, O'Brien, R., and Born, W. (1994).Recognition of trophoblasts by yS cells. J . Itmnunol. 153, 2918-2926. Heyborne, K. D., Cranfill, R. L., Carding, S. R., Born, W. K., and O'Brien, R. L. (1992). Characterization of yS T lymphocytes at the maternal-fetal interface. J . Itnmzinol. 149, 2872-2878. Hiromatsu, K., Matsuzaki, G., Tauchi, Y . , Yoshikai, Y., and Nomoto, K. (19924. Sequential analysis of T cells in the liver during murine listerial infection. J. Immnunol. 149,568-573. Hiromatsu, K., Yoshikai, Y., Matsuzaki, G., Ohga, S., Muramori, K., Matsumoto, K., Bluestone, J. A,, and Nomoto, K. (1992b). A protective role of yS T cells in primary infection with Listeria monocytogenes in mice. J. Exp. Med. 175, 49-56. Hochstenbach, F., and Brenner, M. B. (1989). T-cell receptor &chain can substitute for (Y to form a /3S lieterodimer. Nature 340, 562-565. Holosliitz, J., Koning, F., Coligan, J. E., De Biuyn, J., and Strober, S. (1989). Isolation of CD4+ CD8- mycobacteria-reactive T lymphocyte clones from rheumatoid arthritis synovial fluid. Nature 339, 226-229. Homer, A. A,, Jabara, H., Ramesh, N., and Geha, R. S. (199,5).ylS T lymphocytes express CD40 ligand and induce isotype switchingin B 1ymphocytes.J.Exp. Med. 181,1239-1244. Houlden, B. A,, Matis, L. A,, Cron, R. Q.. Widacki, S. M., Brown, G. D . , Pampeno, C., Meruelo, D., and Bluestone, J. A. (1989).A TCR yS cell recognizing a novel TL-encoded gene product. Cold Spring Harb. Syinp. ($ant. B i d 54, 45-55. Howard, C. J.. Sopp, P., Parsons, K. R., and Finch, J. (1989). In uiuo depletion of BoT4 (CD4) and of non-T4/T8 lymphocyte subsets in cattle with monoclonal antibodies. Eur. J. Initnunol. 19, 757. Hsieh, B., Schrenzel, M. D., Mulvania, T., Lepper, H. D., DiMolfetto-Landon, L., and Ferrick, D. A. (1996). In vivo cytokine production in murine listeriosis. Evidence for immunoregulation by yS+ T cells. J. Inrmunol. 156, 232-237. Huber, €I., Decossy, P., van Brandwijk, R., and Knop, J. (1995). Activation of murine epidermal TCR-yS T cells by keratinocytes treated with contact sensitizers. J . bntnunol. 155,2888-2894. Hutchinson, I. V. (1986).Suppressor T cells iii allogenic models. Transplantation 41,547. Ikuta, K., Kina, T., MacNeil, I., Uchida, N., Peault, B., Chien, Y.-H., and Weissman, I. L. (1990). A developmental switch in thymic lymphocyte maturation potential occurs at the level of heinatopoietic stem cells. Cell 62, 863-874. Ishida, I., Verbeek, S., Bonneville, M., Itohara, S., Berns, A,, and Tonegawa, S. (1990). Tcell receptor y6 and y transgenic mice suggest a role of a y gene silencer in the generation of ap T cells. Proc. Natl. Acad Sci. USA 87, 3067-3071. Ishignro, N., Ada, Y., Shinagawa, T., and Shinagawa, M. (1993).Molecular structures of cattle T-cell receptor gamma and delta chains predominantly expressed on peripheral blood lymphocytes. lininunogenetics 38, 437-443. Ito, K., Van Kaer, L., Bonneville, M., IIsu, S., Murphy, D. B., and Tonegawa, S. (1990:). Recognition of the product MHC TL region gene (27h)by a mouse yS T cell receptor. Cell 62, 549-561. Itohara, S., Farr, A. G., Lafaille,J. J., Bonneville, M., Takagaki, Y., Haas,.W., and Tonegawa, A. (1990). Homing of a yS thymocyte subset with homogeneous T-cell receptors to mucosal epithelia. Nature 343, 754-757.
IMMUNOREGIJLATORY FUNCTIONS OF y6 T CELLS
133
Itohara, S., Moinbaerts, P.. Lafaille, J.. Iacoinini, J.. Nelson, A., Clarke, A. R., €looper, M. L., Farr. A., and Tonegawa, S. (1993). T cell receptor delta gene mutant mice: Independent generation o falpha beta T cells and programmed rearrangements of gainrna delta TCR genes. Cell 72, 337-348. Itohara, S., Nakanishi, N., Kanagawa. O., Kubo, R., and Tonegawa, S. (1989).Monoclonal antihodies specific to native inurine T-cell receptor 76: Analysis of y6 T cells during thymic ontogeny and in peripheral lymphoid organs. Pmc. Nntl. Arorl. Sci. USA 86, 5094-5098. Itohara, S., and Tonegawa, S. (1990). Srlection of 76 T cells with canonical T-cell antigen receptors i n fetal thymus. Proc. Nutl. Acurl. Sci. U S A 87, 7935-79338. Iwashima, M., Green, A,, Davis, M. M., and Chien, Y.-€I. (1988).Variable region (V6) genc segment most frequently utilized in adult thyniocytes is 3’ of the constant (C6) region. Proc. Not/. Acad. Sci. USA 85, 8161-8165. Janeway, J. C. A,, Jr., and Travers, P. (1997).“Inimunobiology: The Immune System in Health and Disease,” 3rd Ed. (;arland Publishing, New York. Janeway, C. A,, Jr. (1988). Frontiers of the irnmune system. Nntttre 333, 804-806. Janeway, C. A,. Jr., Jones, B., and Hayday, A. (1988). Specificity and function of T cells bearing y6 receptors. Imnictnol. T o d q 9, 73-76. Johnson, R. M., Lancki, D. W.. Sperling, A. I.. Dick. R. F., Spear, P. G., Fitch, F. W.. and Bluestone, J. A. (1992).A inurine CD4 , CD8- T cell receptor-y6 T lymphocyte clone specific for Herpes simplex virus glycoprotein 1. J. Z t t i n w i o l , 148, 983-988. Jones-Carson, J., Vazqnez-Torres. A., van tler Heyde, H., Warner, T., Wagner, R., arid Balish. E. (1995).y 6 T cell-induced nitric oxide production enhances resistance to rnucosal candidiasis. Nufitre Med. 1, 552-557. Jorgensen, J. L., Esser, U., de St. Groth, B. F.. Reay, P. A,, and Davis, M. M. (1992). Mapping T-cell receptor-peptide contacts by variant peptide immunization of singlechain transgenics. Nature 355, 224-230. Jorien-Beades, F., Paris, E., Dieulois, C . , Lemeland, J.-F., Barre-Dezelus, V., Marret, S., Huinbert, G.. Leroy, J., and Tron, F. (3997).In vivo and in Litro activation and expansion infection in humans. Infect. I m r t i t t n . 65,4267of y 6 T cells during Lbterin tt~o~~ic!yfog~~ne~s 4272. Kabelitz, D., Bender. A,, Schondelmaier. S.,Schoel. B., and Kaufmann, S. H. E. (1990). A large fraction of human peripheral 1)lood 76’ T cells is activated by Mycobacteriuni tuberculosis but not by its 65 kD heat shock protein. J. Ewp. Merl. 171, 667-679. Kaminsky, M. J.. Bergstresser, P. R., and Takashima. A. (1993). In vivo activation of mouse dendritic epidermal T cells in sites of contact dermatitis. Etrr. J. Inztnunol. 23, 1715-1718. Kang, J., and Kaulet, D. H. (1997). Events that regulate differentiation of cup TCR’ and y6 TCR’ T cells from a coninion precursor. Setti. Zttittiirnd. 9, 171-179. Kaufinann, S. H. E., Blum, C., and Yamainoto, S. (1993). Crosstalk between (YIPT cells and y / 6 T cells in oioo: Activation of Lyjp T-cell responses after y/S T-cell modulation with the monoclonal antibody CL3. Pmc. Nntl. Acnrl. Sci. USA 90, 9620-9624. Kaur, I., Voss, S. D., Gupta, R. S.. Schell, K . , Fisch, P., and Sondel, P. M. (1993). Human peripheral y6 T cells recognize hsp6O niolc~culeson Daudi Burkitt’s lymphoma cells. J. Znirtmnol. 150, 2046-2055. Ke, Y., Pearce, K., Lake, J. P., Ziegler, H. K., and Kapp, J. A. (1997).y6 T lymphocytes regulate the induction and maintenance of oral tolerance. J . Zttittiunol. 158, 3610-3618. Kelly, J., O’Farrelly. C., O’Mahoney,C.,Weir, D. G.,and Feighery, C . (1987).Immunoperoxidase demonstration of the cellular composition of the normal and coeliac small bowel. Clin. Ewp. Imttznnol. 68, 177- 188. Kim, 11. T., Nelson. E. L., Clayberger, C.. Sanjanwala. M., Sklar, J.. and Krensky. A. M. (1995). y6 T cell recognition of tumor Ig peptide. J . Itnrnrcnol. 154, 1614-1623.
134
WILL1 BORN rf ol
Kinebuchi, M., Kikuchi, K., and Matsuura, A. (1994). Rat T-cell receptor y chain sequences: Identification of'conserved gene segments and a unique chimeric constant region gene, C4L. bnmunogenetics 40, 449-455. Koller, B. H., Marrack, P., Kappler, J. W., and Smithies, 0.(1990). Normal development of mice deficient in B2M, MHC Class I proteins and CD8+ T cells. Science 248, 1227-1230. Koinano, H., Fujiura, Y., Kawaguchi, M., Matsumoto, S., Hashimoto, Y., Ohana, S.. Mombaerts, P., Tonegawa, s.,Yamamoto, H., Itohara, s., Nanno, M., and Ishikawa, H. (1995). Homeostatic regulation of intestinal epithelia by intraepithelial y6 T cells. Proc. Nntl. Acad. Sci. USA 92,6147-6151. Koyasu, S . (1994). CD3tCD16tNK1.1'B220+ large granular lymphocytes arise from both alpha-beta TCR'CD4-CDX- and gamma-delta TCRtCD4-CD8- cells. /. Exp, Med 179, 195-1972, Kozbor, D., Trinchieri, G., Monos, D. S., Isobe, M., Russo, G., Haney, J. A., Zmijewski, C., and Croce, C. M. (1989). Human TCR-y+IS+,CD8' T lymphocytes recognize tetanus toxoid in an MHC-restricted fashion. /. E x p . Med. 169, 1847-1851. Kronenberg, M. (1994). Antigens recognized by y6 cells. C u m Opin. Itnmunol. 6, 64-71, Kuhnlein, P., Park. J.-H., Herrmann, T., Elhe, ,4., and Hiinig, T. (1994). Identification and characterization of rat y / 6 T lymphocytes in peripheral lymphoid organs, small intestine, and skin with a monoclonal antibody to a constant determinant of the y / 6 T cell receptor. 1.liniiiunol. 153, 979-986. Ladel. C. H., Blum, C., Dreher, A., Reifenberg, K., and Kaufmann, S. I I . E. (1995a). Protective role of y / 6 T cells and alp T cells in tuberculosis. Eur. 1.Zrntnunol. 25, 28772881. Ladel, C. H., Blum, C., and Kaufinann, S. H. E. (1996). Control of natural killer cellmediated innate resistance against the intracellular pathogen Listeria monocytogenes by y6 T lymphocytes. Infect. lmnwc. 64, 1741-1749. Ladel, C. H., Hess, J., Daugelat, S., Mombaerts, P., Tonegawa, S., and Kaufmann, S. H. E. (1995b). Contribution of alp and y / 6 T lymphocytes to immunity against Mycobacterium boois bacillus Calmette Gukrin: Studies with T cell receptor-deficient mutant mice. Eur. /. Ztnni~mo~, 25, 838-846. Lafaille, J. J., DeCloux, A., Bonneville, M., Takagaki, Y., and Tonegawa, S. (1989).Junctional sequences of T cell receptor y6 genes: Implications for y6 T cell lineages and for a novel intermediate of V-(D)-Jjoining. Cell 59, 859-870. Lafaille, J. J., Haas, W., Continho, A,. and Tonegawa, S. (1990). Positive selection of y6 T cells. Imrnunol. Toduy 11, 75-78. Lahn, M., Kalataradi, H., Mittelstadt, P., Pflum, E., Vollmer, M., Cady, C., Mukasa, A,. Vella, T. A . , Ikle, D., Harbeck, R., O'Brien, R., and Born, W. (1998). Early preferential . 5221-5230. stimulation of y6 T cells by TNF-a. /. ~ i n i n i i n o ~160, Lang, F., Peyrat, M. A., Constant, P., Davodeau, F., David-Ameline, J., Poquet, Y., Vie. H., Fournik, J. J., and Bonneville. M. (1995). Early activation of human Vy9VS2 T cell broad cytotoxicity and TNF production by nonpeptidic mycobacterial ligands.1.Zmoiiunol. 154, 5986-5994. Langhorne, J., Goodier, M., Cehr, C., and Dubois, P. (1992). Is there a role for y6 T cells in malaria? lmrnunol. Today 13, 298-300. Lantz, O., and Bendelac, A. (1994). An invariant T cell receptor a chain is used by a unique subset of major histocompatibility complex class I-specific CD4 and CD4-8- T cells irt mice and humans.]. Ewp. Med. 180, 1097-1106. Lawetzky, A., Tiefenthaler, G., Kubo, R., and Hunig, T. (1990). Identification and characterization of rat T cell subpopulations expressing T cell receptors alp and y / 6 . Eur. J. lrnniiinol. 20, 343-349. +
Lehedeva. M. I.. Fields. B. A., Spits. H.. C.o\intlaswamy, P., Hrenner. M. B.. nnd Marinzza, R. A. (1996). Cloning. expression, and ciystalliziition of tlw V delta tloiiiain of a huni:un gamma delta T-cell receptor. Protciri Sci. 5, 2638-3642. Leclercq, G., and Pliim, J. ( 1996). Thyinic devrdopnient of \'y3 cells. sfvi. Z r ~ i r i i i i t t c ~ / . 8, 315-321. Lefranc, M.-P., Chuchana. P., Dariavach. P., Nguyrw, C., Hnck. S.. Hrocklj., F., Jordan. B., and Lefranc, G. (1989). Molecular inappingofthe Iiiiman Tcell receptor gamma (TRC;) genes and linkage of the variable and constant regions. Eiir. /. Z ~ r i r r ~ ~ ~19, ~ t o 989-994. l, Lefranc. M.-P., ; t i i d Rabbits. T. H. (1989). The liinniin T-ccll receptor y (TRC:) genes. T r r d s Bioc/wirt. Sci. 2 14-2 18. Lt+dIic. M.-P.. and Rahbitts, T. H. (1985). Two t;indenily organized huinan genes encoding the T-ccll y constant-region seqncnces show multiple reufiingenient in diffbrent T-cell types. Natiirc> 316, 464-466. Lefraiic. M.-P., and Rahbitts, T. 11. (1990). A nonirnclatiire to fit thcl organization of the hunian T-cell receptor y and 6 genes. He.y. Z t r i i t m r i c ~ l . 141, 615-618. I,efranyois, L. (1991) , Extrathvmic differentiation of intriepithelial lymphocytes: Generation of a separate and unequal T-cell repertoire? I i t i t t i i i t i o / . TO^/ 12, 436-438. Lefranpis, L.,and Goodman, T. ( 1989).I n vivo niotlulation of cytolytic activity and Thy1 expression in TCR-y/G' intraepithelid lympliocytes. Sciciicc 243,. 1716-1718. Lefranpis. L., Lecorre, R . , Mayo, J., Blnestone. J. A,. and (hodman. T. (1990).Extrathymic selection of TCR 76' cells by class I1 major Iiistocompatil)ility complex molecules. Cell 63, 333-340. Livak, F., Petric. 1%. T., Crisp?, I. N., and Scliatz, D. C;. (I99S). I n frame TCR 6 gene rearrangements play a critical role in the &y6 T cell lineage decision. Zmtnrinity 2, 617-627. L>iicIi, F.. and Shevacli, E. M. (1992)).Activation reciuirenients of newborn thymic y6 T cells. /. Z t r i r t i i ~ i i o l . 149, 2307-2314. Maccario. R., Conioli. P., Percivdle. E., Montagna, I)., Locatelli, F., and Gerna, G. (1995). Herpes siinplex Liriis-specific human cytotoxic T-cell colonitas expressing either y6 o r a@ T-cell receptor: Role of accessory niolecnlcs on FILA-iinrestrictetl killing of virus-infected targets. Zttirtutnolugy 85, 49-56. Machngh, N. 11.. M b n n i , J. K., Carol, M. J., Wyatt, C;. R., Orden, J. A , , and Davis, M'. C . (1997). Identification of two distinct sril>setsof I)o\ine y6 T cells with unicpie ccll surface phrnotype and tissiie distribntion. Z t t i t i i i i r i d ~ ~ g92, ! ~ :340-:345. Maki. I<., Smiaga, S., ant1 Ikuta, K. (1996). The \'-J recombination of T cell receptor y gent's is blocked in interleukin-7 receptor-deficient mice. /. Exp. Mrd. 184, 2423-2427. Maki, K.. Siniag;i, S., Komagata. Y.. Kotlaira. Y.. Maliuchi, A , , Karasuynna, H., Yokomuro. K.. Miyazaki, J. L.. and Iknta, K. (1996). Interlriikin 7 reccptor-deficient mice lack y6 T cells. Proc. Natl. Acnd. Sci. USA 93, 7172-7177. Matis, L. A,. and Bluestone, J. A. (1991). Specificity of y6 receptor hearing T cells. S P t 7 l . ~ l t L t / l l i T l ~ J 3, ! . 75-80. Matis. L. A., Cron. It., and Bliiestone, J. i\.(1987). Major Iiistoc.ompatihilit>,complex-linked specificity of y6 receptor-bearing T lyniphoe)tes. Notiire 330, 262-264. Matis, L. A,. Fry. A. M., Cron. It. Q., Cotterinan, M. M.. Dick. R. F., and Blnestone, J. A. (1989). Stnictiire and specificity of it c ' I1 alloreactive y6 T cell receptor heterodimer. Scimcr 245, 746-749. Mavadd;tt, N., Rohinson. B. W.. Rosc, A. I I . , Manning. I,. S.. and Garlepp, M. J. (19Ys3). An analysis of the rehtionship between ganinia delta T cell receptor \' gene usage and non-major Iristocornpatability coiiiplex-lcstrictetl cytotoxicity. Z t n t n u r i d . Ccll B i d . 71, 27-37.
136
WII.LI HORN rt rrl.
McMenamin, C., McKersey, M., Kiihnlein, P.. Hunig, T., and Holt, P. G. (1995). y6 T cells down-regulate primary IRE responses in rats to inhaled soluble protein antigens. J. Zmitiunol. 154, 4390-4394. McMenaniin, C., Oliver, J., Girn, B., Holt, B. J., Kees, U. R., Thomas, W. R., and Holt, P. G . (1991). Regulation of T-cell sensitization at epithelial surfaces in the respiratory tract: Suppression of IgE responses to inhaled antigens by CD3+ TcR a-//3-lymphocytes (putative ylS T cells). Zmvnunohg/ 74, 234-239. McMenaniin, C., Pimm, C., McKersey, M., and Holt, P. G. (1994). Regulation of IgE responses to inhaled antigen in mice by antigen-specific y6 T cells. Science 265, 18691871. Meeusen, E., Fox, A,, Brandon, M., and Lee, C.-S. (1993).Activation ofuterine intraepithelial y6 T cell receptor-positive lyinphocytes during pregnancy. Eur. J . Itnrnunol. 23, 11121117. Melanied, D., and Friedman, A. (1993). Direct evidence for anergy in T lymphocytes tolerized by oral administration orf ovalbumin. E m J. Imnarrnol. 23, 935-942. Mengel, J., Cardillo, F., Aroeira, L. S., Williams, O., Russo, M., and Vaz, N. M. (1995).Antiy6 T cell antibody blocks the induction and maintenance of oral tolerance to ovalbumin in mice. Imniunol. Lett 48, 97-102. Mincheva-Nilsson,L., Baranov,V., Yeung, M. M.-W., Hammarstrijm, S., and Hammarstriim, M.-L. (1994). Immunomorphologic studies of human decidua-associated lymphoid cells in normal early pregnancy. J . Immlinol. 152, 2020-2032. Mincheva-Nilsson,L., Hammarstriini, S., and Harnmarstrom, M.-L. (1992).Human decidual leukocytes from early pregnancy contain high numbers of yS+ cells and show selective down-regulation of alloreactivity.J. Inunrrnol. 149, 2203-221 1. Mixter, P. F., Camerini, V., Stone, B. J.. Miller, V. L., and Kronenberg, M. (1994). Mouse lymphocytes that express a y 6 T cell antigen receptor contribute to resistance to salmonella infection in ciuo. Itgect. Inirnzrn. 62, 4618-4621. Miyagawa, Y., Matsuoka, T., Baba, A., Nakaniura, T., Tsuno, T., Tamura, A,, Agematsu, K., Yabuhara, A., Uehara, Y., Kawai, H., and Miyasaka, K. (1992). Fetal liver T cell receptor y/6+ T cells as cytotoxic T lymphocytes specific for maternal alloantigens. J. Exp. Med. 176, 1-7. Mombaerts, P., Arnoldi, J., Russ, F., Tonegawa, S., and Kaufmann, S. H. E. (1993).Different roles of a/3 and y6 T cells in immunity against an intracellular bacterial pathogen. Nature 365, 53-56. Mombaerts, P., Clarke, A. R., Rudnicki, M. A,, Iacoinini, J., Itohara, S., Lafaille, J. J., Wang, L., Ichikawa, Y., Jaenisch, R., Hooper, M. L., and Tonegawa, S. (1992). Mutations in Tcell antigen receptor genes a! and /3 block thymocyte development at different stages. Nature 360, 225-231. Mombaerts, P., Mizoguchi, E., Ljunggren, H. G., Iacomini, J., Ishikawa, H., Wang, L., Grusby, M. J., Climcher, L. H., Winn, H. J., and Bhan, A. K. (1994).Peripheral lymphoid development and function in TCR mutant mice. Znt. Zinrntrno~.6, 1061-1070. Morita, C. T., Beckman, E. M., Bukowski, J. F., Tanaka, Y.. Band, H., Bloom, B. R., Colan, D. E., and Brenner, M. B. (1995).Direct presentation ofnonpeptide prenyl pyrophosphate antigens to human y6 T cells. Immunity 3, 495-507. Morita, C. T., Verma, S., Aparicio, P., Martinez-A, C., Spits, H., and Brenner, M. B. (1991). Functionally distinct subsets of human y / 6 T cells. Eur. J. Inmiunol. 21, 2999-3007. Moriwaki, S., Korn, B. S., Ichikawa, Y., Van Kaer, L., and Tonegawa, S. (1993). Amino acid substitutions in the floor of the putative antigen-binding site of H-2T22 affect recognition of a y6 T-cell receptor. Proc. Natl. Acnd. Sci. USA 90, 11396-11400.
Mowat, A . M. (1987). The regulation o f iiiiniiiiic responses to dietary protein antigcws. z l r t r l ~ l ~ r t ~?br/r!/ l/. 8, 933-98. ;uid Nomoto, K. (1995). Mrlkasa, A , . Hioroniatsri, K., Matsiizaki, (;., O’Brieii. R., B c i r i i ~\i’.. Bacterial infection of the testis leading to ;uitoaggrrssive i l i i i i i r i i i i h triggers apparently opposed responses of a/3 antl y6 T cc~lls.J . Z i r t r r t i ~ r i o / , 155, 2047-2056. Miikasa. A,, Lahn. M.. Pfluiri, E. K.. Borii. it’., antl R. L., 0. B. (1997). E d e n c e that the sanics y6 T cells respond driiiiig iiiti~ction-inducrd ;iid antoinimune inflainmation. J . Z t r i t r i i i r t o / . 159, 5787-,5794. Munk,M . E., Fazioli, R. A., Calich. \’, L. G.. and Kaiifimtin, S. H. E. (199s).P~mcoccit/ir,i[k,s bmsi/iotsis-stiiiriilated hiinian y / 6 T ccsllssupport a n t i l d y production by B cells. Infect. Z r m i i r t i . 63, 160% 1610. Munk. M. E.. Teixeira. H . (;., Kiiner, A . Fazioli, R. A., Gdicli, \’. L. G., and Kaufinann, S. H. E. (1994).Hriinan a/3 a i d yd T cells from iinexposeed individuals respond to protein , ~ i ~ i Ziif. o i ~ Z r /n nf~ ,u t.i d~.6, 171 7-1725 antigens of the yeast form of ~ ~ r ~ ~ ~ ~ ~hrt/silicmi.s. Nagler-Anderson. C., McNair, L. 4.,arid Cradock. A. (1992). Srlf-reactive, T cell receptory6’, Iymphoc\tes from the intestinal e p i t h e h i i ofwcwiling mice. J . Zrritriutid. 149,2315-
2322. Nakajiina, H., Moore, Y. I . , and Takiguchi. M . (1997).Fiinctional expression of the N K B 1 killer cell inhibiton. receptors on a y6 CTL clone. Zrritrirrriogeiictic:v 45, 226-228. Nakajiina, 1-1.. Tomiyaina, H.. and Takigiichi. M . (1995).Inhiliitioii of gamma delta T cell recognition by receptors for MHC class I niolecriles. J . Zttttrtutid 155, 4139-4142. Nishiniiira, 13.. Emoto, M., I-liromatsu, K.. Yani;inioto. S.. Matsuura, K., Gomi, H.. Iketla, T., Itoliara, S . , ant1 Yoshikai, Y. (1995). The role of y6 T cells in priniing macrophages to prodiicc, tiimor necrosis f’actor-a. Eiir-. /. I t r i t r t r r r i o / . 25, 1465-1468. Nishiniiira, H.,Hiroiiiatsii. K.. Kobayaslii. N., C.ral)steiii, K. €1.. P at o n , R., Siigainrira, K., Bhestone, J. A.. and Yoshikai, Y. (1996). IL-15 is a novvl growth fiictor for miirine y6 T cells induced by Snlrrtortc~/k~/infection J . Z r r i t r i r r r t o / . 156, 663-669. Nishiiiiiira, Y., Eto, M.. Maedii, T., Hiroinntsu, K.. K o b a ~ d i i ,N., Noitloto, K., Kong, Y.-Y., antl Nomoto, K. (1994).Inhibitiori of skin scnograft rejection by depleting T-cell reccytor a/3-heariiig cells withorit T-cell rweptor yd-bearing cells or natural killer cells by monoclonal mtil~otly.Ztrtniiiriology 83, 196-204. O’Brieii, R. I,., FII,Y.-X., Crarifill. R.. Dallas. A.. Reardon. C.. Lang, J.. Carding. S. R., Kiiho, R., and Born, M‘. (1992). €Ieat shock protein Hsp-60 reactive y6 cells: A large, divrrsified T Iyiiipliocyte siihset with highly focused specificity. Proc. Nod. Accd. Sci. USA 89, 4348-4:352. O’Brien, R. L.. Happ. M. P.. Dallas, A,. Pa1nic.r. E., Kril)o. R., and Born, tV. K. (1989). receptor yd by an antigen Stimiilatioii of a niajor siibset of lyniplioc derived from Myco1)acteriuiii tuberculo )to, K. (1990). Sequential Oliga, S., Yoshikai, Y., Takeda. Y.. Hiro ty during an i.p. infection qqw;iriiice of $6- and dp-l)earing T c with Listeria monocytogeiies. Eirr. J . Z Olive, C . ( 1997). Modillation of exprrimental all(.rgic c~ncephaloinyelitisin mice by imnirinization with peptide specific for the y6 T cell recvptor. I t t m i i i i o / . Cell Hiol. 75, 102-106. Orsini. D. L., Res. P. S., Van Laar, J. M., Mliller, L. M.. Soprano, A. E., Kooy, Y. M., Tak, P. P.. and Koning, F. (1993). A subset of V61+ T cells proliferates in response to EpsteinBarr \iriis-transfornie(l I3 ce11 hies in \itro. S m d . J . Z r n n i u r i o / . 38, Orsini, D. L., =in Gild, M., Kooy, Y. M.. Struyk, L., Klein. G., vim den Elsen. P., and Koning, F. (1994). Furictional and niolrcrih charactcarization of B-cell responsive V61’ gainina delta T cells. Rut-. J . f r n m r r t t o / . 24, 3199-3204.
138
CZ'I1.LI BORN c/ nl
Ota, Y., Kobata, T., Seki, M., Yagita, H., Shiniada. S., Hrlang, Y.-Y.. Takagaki, Y., and Okiiniiira, K. (1992). Extrathymic origin of VylN66 T cells in the skin. Ettr. J . Iintnnnol. 22,595-598. Palliard, X., Yssel, H,, Blanchard, D., Waitz, J. A,, deVries, J. E., and Spitz, H. (1989). Antigen specific and MHC nonrestricted cytotoxicity of T cell receptor aBi and y6+ human T cell clones isolated in IL-4. I. Ztnmtnol. 143, 452-457. Panchamoorthy, G., McLean, J., Modlin, R. L., Morita, C. T., Ishikawa, S., Brenner, M. B., and Band, H. (1991). A predominance of the T cell receptor of Vy2/V62 subset i n human inycobaceria-responsivc. T cells suggests germline gene encoded recognition. 1. Imnunol. 147, 3360-3369. Pao, W., Wen, L.. Smith, A. L., Culbranson-Judge. A,. Zheng, B., Kelsoe, G., MacLennan, 1. C. M., Owen, M. J., and Hayday, A. C. (1996).y6 T cell help of B cells is induced by repeated parasitic infection, in the absence of other T cells. Cztrr. Biol. 6, 1317-1325. Pardoll, D. M., Fowlkes, B. J., Bluestone, J. A,, Kniisbeek, A., Maloy, W. L., Cohgan, J. E., and Schwartz, R. €I. (1987). Differential expression of two distinct T-cell receptors during thymocyte development. Natnre 326, 79-81. Parker, C. M., Groh, V., Band, H., Porcelli, S. A., Morita, C., Fahhi, M., Class, D.. Strominger, J. L., and Brenner, M. B. (1990). Evidence for extrathymic changes in the T cell receptor y / 6 repertoire. I. E q i . Med. 171, 1597-1612. Parr, E. L., and McKenzie, I. F. C. (1979).Demonstration of Ia antigens on mouse intestinal epithelial cells by immunoferritin labeling. In~nunogenetics8, 499-508. Peng, S. L., Madaio, M. P., Hughes, 1). P. M., Crispe, I. N., Owen, M. J., Wen, L., Hayday, A. C., and Craft. J. (1996). Murine lupus in the absence of (YP T cells. J. I n i ~ i r ~ n d . 156, 4041-4049. Pereira, P., Zijlstra, M., McMaster, J., Loring, J. M., Jaenisch, R., and Tonegawa, S. (1992). Blockade of transgenic y6 T cell development in B2-microglobulindeficient mice. E M B O J. 11, 25-31. Peterinan, G. M., Spencer, C., Sperling. A. I., and Bluestone, J. A. (1993). Kole of y6 T cells in inurine collagen-induced arthritis. J. Immnnol. 151, 6546-6558. Pfeffer, K., Schoel, B., Gulle, H., Kaufinann, S. H. E., and Wagner, H. (1990). Primary responses to human T cells to mycobacteria: '4frequent set of y / 6 T cells are stimulated by protease-resistant ligands. Etrr. 1.Itnmnnol. 20, 1175-1 179. Pfeffer, K., Schoel, B., Plesnila, N., Lipford, G . B., Kromer, S., Deusch, K., and Wagner, H. (1992). A lectin-binding, protease-resistant mycobacterial ligand specifically activates 199' human y6 T cells. J . Zmmcnol. 148, 575-583. Porcelli, S., Brenner, M. B., and Band, H. (1993). Biology oftlie human y6T-cell receptor. I r n m m l . Reo. 120, 137-183. Porcelli, S., Brenner, M. S., Greenstrill, J. K., Balk, S. P., Terhorst, C., and Bleicher, P. (1989). Recognition of cluster of differentional antigens by human CD4-CD8- cytolytic T lymphocytes. Nature 341, 447-450. Porcelli, S. A., and Modlin, R. L. (1995).CD1 arid the expanding universe of T cell antigens. J. Iminunol. 155, 3709-3710. Ptak, W., and Askenase, P. W. (1992). y6 T cells assist ap T cells in adoptive transfer of contact sensitivity. J . Zm?nunol. 149, 3503-3508. Ptak, Mi., Szczepanik, M., Ramabhadran, R., arid Askenase, P. W. (1996). Immune or nonnal y6 T cells that assist a/3 T cells in elicitation of contact sensitivity preferentially me Vy5 and V64 variable regon gene segments. /. Imtnunol. 156, 976-986. Quaroni, A., and Isselhacher, K. J. (1985). Study of intestinal cell differentiation with monoclonal antibodies to intestinal cell surface components. Deo. Rial. 111, 267-279.
Quere, P., Cooper, hl. D., and Tliorhecke, G. J. (1990). Cliaractcrizatiori of' suppressor T cells for aiitibody protluctioii I)? cliickrn spleen cells. I. Antigen-induced suppressor cells are CTH', TcRI' (76) T cells. z t t L t t I f l t i O / ~ J @ /71, 517-522, Rajagopalan. S., Mao, C., and Ilatta, S. K. ( 1992). Patliogenic autoailtibody-inducing y/S T helpw cells from patients with lupiis IirphiitiT express tiiiiisrial T cell receptors. -Glitz. ~ ~ ~ l ~ f i f l f~ I t Ot / l . f t I f / f ~ ~ J62, ~ ~ ~344-350, f ~ l ~ ~ ~ . Hajagopalaii. S., Zordan, T.. Tsokos. G. C., antl DattLi. S . K. (1990). Pathogenic antiDN.4 autoantibocly-iii(1~iciiigT helper cell lines from patients with active Inpus nephritis: Isolation of CD4-8 T helper cell lines tliat cxprvss the yS T-cell antigen receptor. Proc. N ~ t l4cod. Sci. USA 87, 7020-7024. Rakasz, E., Sandor. M.. Hagrn, M.. aiid Lycli, R. G. (1996).Activation features of intraepithelid gminra delta T-cells of the muriiie vagina. Z t t t t t w i i d . Left. 54, 129-1:34. Hast, J . P.. Anderson. hl., Strong, S.J., h e r , (;,, Idtirraii, R. T., ant1 Litinan, G. W. (1997). a, p, y a i d S T cell antigen receptor geiies arose early i n vertebrate phylogeny. Zii~nirrnity 6, 1-11. Ratlibiiil, G.A.. Born. \V., Krizie!, LV. A , . uid Trickt~r,1'. LV. (1988).Diversity of the mouse T cell receptor Cyl gene: Striictiiral aiiiilysis in C57HIJKa. Zi,rtiitriiogr,tLr.fics 27, 121-126. Raulet, 11. H. (1989). The stnicturr, liinctioii. a i i d inolecular genetics o f tlie ylS T cell . Z t t m i m d . 7, 175-207, receptor. L h ~ r i t t Rco. Rarilet, D. 11. (1994). Ilow yS T cells d i e ii hing. Cirrr. B i d . 4, 246-248. Kaulet. D. H., Sprwer, D. M . . Hsiaiig, Y . - H . . C ~ o l c l i i i m ,J. P., Bix, M., Liao. N.-S., Zijlstra, M., Jaenisch, R., antl Correa, I. ( 1991).Control ofy6 T-cell developnient. Z t r i t t w t u d Rrti. 120, 185-204. Rcardon, C., Lefrmcois. L., Farr, A,. Kubo,It.. O'Brien. R.. and Born, M'. (1990). Exprcassion of' ylS T cell receptors on lynipliocytcs from thc Inctatiiig maminary gland. 1.E s p M o d 172, 1263-1266. Reardori, C. L., Heybome, K., Tsuji. M.. Zavala, F., Tigelaar, R. E., O'Brieii K. I>., and Born, W.K. (1995). Miirine epitlrrinal VyFjn'SI-TCR' T cells r q o n d to B cell lines untl liI)ol")lysacctiaridcs. I , Incest. L)wttwtol. 105, 58S-61S. Rella1i:iii. B. L., Bluestone, J. A.. Houlden. B. A . , Cotteriiian, M. M., mid Matis, L. A . (1991). Junctional secpences inHiience the specificity of ylS T cell receptors. /. E x p Med. 173, 503-506. Rliciiis. 1,. A,. Yoiuig, E. M . . Nordlrind, M . I,., Berning. 11. B.. a i d Nordlund, J. J. (19x7). Rapid inchiction of Thy-1 antigenic niarkers on krratinocytes and epiderinal iminuiie cells i n tlie skin of mice following topical tre;itment with coiiiiiioii preservatives used in topical medicatioris ailid in foods. /. Z t w r s f . Z A > t - t f r d o / . 89, 489-494. Itoark, C. E. (1995). A study on iiiuriue liver gaiiui1ddrlta T lymphocytes. Pl1.D. The ' Roark, C . E., Volliner. M . . Cairiphell, P., Borii, W.K.. and O'Biieii, R. L. (1996). Respo~isc~ /. Z t t ~ i t i i m d 156,2214-2220. . o f a yS-TCH irioiioinorphic subset during 1)acteii;iliiift~tioii. Roberts, A. D.. Ordway, D. J., mid Orine. I. M . (1993). Listeria rriot~ocytogc,tir.sinfcction i n 8 2 microglobulin-deficieiit inicc. Ztlji,ct. Z t t u t i u t i , 61, 111:3- 1116. Roberts, K., Y C J ~ [ J ~ I I IW. I ~ IM . . , Kehn, 1'. J.. a i i d Sllcvxh, E. M. (1991). The vitronectin receptor serves as an accessory niolecdr for the activation of a subset of ylS T cells. I. E x p . Met!. 173, 231-240, Koberts, S. J., Smith. A. L., West, '4.B., Wen, L., Fiiidy, R. C., Owen, M. J., and Hayday, A. C . (1996). T-cell L Y and ~ yS' deficient mice display ahioriiial but distinct phenohpes toward a natural, widespread infection o f the intestinal epithelium. Proc. N o t / . Actid. Sci. U S A 93, 1 1774- 11779, Robey, E., and Fowlkrs. B. J. (1994). Selective events in T cell development. h n i l . Hec. Z i t u t i i L t w l , 12, 675-706.
140
\VILLI UOHN ct
NI
Rock, E. P., Sibbald, P. R., Davis, M. M., and Chien, Y.-H. (1994).CDR3 length in antigenspecific immune receptors. J. E x p Med. 179, 323-328. Rothenberg, M. E., Weber, W. E., Longtine, J. A,, and Hafler, D. A. (1996).Cytotoxicgalrllua delta I lymphocytes associated with an Epstein-Barr virus-induced posttransplantation lyinplioproliferative disorder. Clirr. Ittittiunol. Inttiuntpztliol. 80, 1266- 1272. Rust, C. J. J., Verreck, F., Vietor, €I., and Koning, F. (1990). Specific recognition of staphylococcal eliterotoxin A by Iiumaii T cells hearing receptors with the Vy9 region. Nature 346, 572-574. Saito, H., Kranz, D. M., Takagaki, Y., Hayday, A,, Eisen, H., and Tonegawa, S. (1984). Complete primaiy stiuctiire of a heterodimeric T-cell receptor deduced from cDNA sequences. Nature 309, 757-762. Saito. T., Hochstenbach, F., Marusic-Calesic, S., Kruisberk, A. M., Bremier, M., and Germain, R. N . (1988).Surface expression of only yS and/or a/3T cell receptor heterodimers by cells with four (a,P,y,S,)functional recrptor chains. J , Exp. Med 168, 1003- 1020. Sano, Y., Dudley, E., Carding, S., Lin, R.-H., IIayday, A. C., and Janeway, C. A. Jr., (1993).yS T-cell lines isolated from intestinal epithelium respond to a B-cell lymphoma. zfnttlunology 80, 388-394. Satoh, M., Seki. S., Hashimoto, J. W., Ogasawara, K.. Kobayashi, T., Kumagai, K., Matsuno, S., and Takeda, K. (1996). Cytotoxic gammadelta or alphabeta T cells with a natural killer c d l maker, CD56, induced from human peripheral blood lymphocytes by a combination of IL-12 and IL-2. J. Z t t m i t m d . 157, 3886-3892. Scharton, T. M., and Scott, P. (1993). Natural killer cells are a source of interferon y that drives differentiation of CD4+ T cell subsets and induces early resistance to Leishttimiia tnajor in mice. J. Exp. M e d 178, 567-577. Schild, H., Mavaddat, N., Litzenherger, C., Ehrich, E. W.. Davis, M. M., Bluestone, J. A,, Matis, L., Draper, R. K., and Chien, Y.-H. (1994).The nature of major histocompatibility complex recognition by yS T cells. Cell 76, 29-37. Sclioel, B., Sprenger, S., and Kairfinann, S. H. E. (1994).Phosphate is essential for stiniulation afVySVS2 T lymphocytesby inycohacterial low molecular weight ligand. Eur. J. Z t t m u t i d . 24, 1886-1892. Sciammas, R., Johnson, R. M., Sperling, A. I., Brady. W., Liiisley, P. S., Spear, P. G., Fitch, F. W., and Bluestone, 1. A. (1994). Unique antigen recognition by a herpesvirus-specific TCR-yS cell. J. Zn~rncmol.152, 5092-5397. Sciammas, R., Kodukula, P., Tang, Q., Hendricks, H. L., and Bluestone, J. A. (1997).T cell receptor-y/S cells protect mice from herpes simplex virus type 1-induced lethal encephalitis. /. Exp. Med. 185, 1969-1975. Shiohara, R., Moriya, N., Gotoh, C., Sajzawa, K., and Nagashima, M. (1988). 1,ocally administered monoclonal antibodies to lymphocyte function-associated antigen 1 and to L3T4 prevent cutaneous graft-vs-host disease. J. Ittiniund. 141, 2261-2267. Shiohara, T., Moriya, N., Gotoh, C., Hayakawa, J., Nagashima, M., Saizawa, K., and Ishikawa, 13. (1990). Loss of epidermal integrity by T cell-mediated attack induces long-term local resistance to suhsequent attack. I. I ~ i d ~ c t iof~ iresistance i correlates with increased in Thy-lt epidermal cell numbers. J . Exp. Mrd. 171, 1027-1041. Shiohara, T., Moriyi, N., Hayakawa, 1.. Itohara. S., and Ishikawa, H. (1996). Resistance to cutaneous graft-vs.-host disease is not induced in T cell receptor S gene-mutant mice. J . Exp. Met/. 183, 1483-1489. Shiohara, T., Nariinatsu, H., and Nagashima, M. (1987). Induction of cutaneous graft-vshost disease by d o - o r self-Ia-reactivehelper T cells in mice. Transplantation 43,692-698. Sim, G.-K. (1995). Intraepithelial lymphocytes arid the irnrnuiie system. Ado. Iintnunol. 58, 297-343.
1hlhlUYOREC;ULATORY FUNCTIONS OF y6 T C:El,l.b
141
Siin. G.-K., and Augustin. A. (1991a). Dominant expres,sion of the T cell receptor BALB invariant 6 (BID) chain in resident pnhnonaiy l>mphocytes is due to selection. Errt-, /. Z t t i t t i r r d . 21, 859-861. Sini, G.-K., and Augustin, A. (199lb). Extrathvnric positive selection of y6 T cells: \Jy,Jyl rearrangements with “CxYS” junctions. I . Itnttiirtio/. 146, 2349-2445. Sim, G.-K., and Augustin. A. (1993).The presence of a i i endogenons murine leukemia Liirus sequence correlates with the peripheral expansion of y 6 T cells bearing the BALB invariant delta (BID) T cell receptor 6.1. Exp. M L , ~178, . 1819-1824. Sim, G.-K., Rajaserkar, R., Dessing, M.. antl Angustin, A. (1994). Homing antl iri s i t r r diffi2rentiationof resident pulnionay lymphoc>tes. Int. Z t t ~ t ~ ~ r i r i6, o l . 1287-1295. Skeeii. M. J., antl Ziegler, H. K. (1993;i). Intluction of niiirine peritoneal y6 T cells ant1 their role in resistance to biicteiial infection. /. En-11. Med 178, 971-984. Skeen, M. J., and Ziegler, H. K. ( I 99311). Intercellular interactions and cytokine responsivcness of peritoneal d p and y/S T cells from Lis(c,t-ici-iiifected inice: Synerpstic effects of interleukin 1 and 7 on y/6 T cells. I, E.vp hlcrl. 178, 985-996. SGderstroni, K., Bucht, A , , Halapi, E., Gronberg, A,, Magnusson. I., and Kiessling, H. (1W6). Increased frequency of al)nornial ganinia delta T cells in blood of patients mith inflainniatoy bowel diseases. /. Ittitttrrtio!. 156, 2331-2339. Sowder, J. T.. C h i , C.-L. H., Ager, L. L.. C1i;in. M. M.. and Cooper, M. D. (1988). A largr subpopulation of avian T cells express ii Iiomologue oftlie nianinialian Ty/6 receptor. /. Exp. hlerl. 167, 315-322. Spaner, D.. Cohen, B. L., Miller, H. G . , arid Phillips, R. A. ( 1995). Antigen-presenting cells for naive transgenic y6 cells: Potent activation by activated cup T cells. I. Z r t ~ m t r d . 155, :3866-3876. Spaner, D.. Migita, K., Ochi, A,, Shannon, J., Miller. R. G.. PeleiI’d. P.. Tonegawa, S., and Phillips, R. A. (1993). y6 T cells differentiate into a fiinctional h i t nonproliferative state during ii normal irnniune response. Proc. Not/. Acnd. Sci. U S A 90, 8415-8419. Sperling, A. I., Cron, H. Q.. Decker, D. C., Stern, D. A,, and Bluestone, J. A. (1992). Peripheral T cell receptor y6 variable gene repertoire maps to the T cell receptor loci and is influenced by positive selection. /. Z i t i t t i r i t i o ! . 149, 3200-3207. Sperling, A. I., and Wortis, 11. H. (1989).CD4.. CD8- y/6 cells froin normal inice respond to a syigeneic B cell ljinphoma antl can induce its differentiation. Int. Irtttrwnol. 1, 434-442. Stenger, S., Mazzaccaro, R. J., Uyemrira, K..Cho, S., Banies, P. F.. Rosat, J.-P., Sette, A,, Br~nner,M. B.. Porcelli, S. A,, Blooin, B. R., and Modlin, R. L. (1997). Differential effects of‘ cytolytic T cell subsets on intracrllriliir infixction. S ’ C ~ P ~ I C C276, 1684-1687. Studicr, F. \\’.. Rosenberg, A. H., Diinn. J. J., and Dubendorff, J. \V. (1990). Use of T7 RNA plyinerase to direct the expression of cloned genes. hfethody E t m p u d 185, 80-89. Sun, Y,, Chrn, Z., C:hilng, S. W., Zeng, H., and Gorczyiski. R. M. (1997). TCR diversity in y6TCII’ hybridomas derived from niice givcn po1ta1vein donor-specificpreimmunization and skin allografts. Mol. MK/. 3, 89- 102. Sunaga, S., Maki, K., Koniagata. Y.. Miytzaki, J.-l., and I h t a , K. (1997). Developinentally ordered V-J recoinbination in inouse T cell receptor y locus is not perturbed by targeted deletion of the \7y4 gene. 1. Iinttiutio~.158, 422:3-4228. Suzriki. T., Hirornatsu, K., Antlo, Y., Okmioto, T.. Tomotla. Y.. antl Yoshikiu, Y. (199s). Regulatory role of y6 Tcells in uterine intraepithelial lyinphocytes in maternal antifetal iininune response. 1. Zmvmtiol. 154, 4376-4484. Szczepanik, M., Anderson, L. H., Ushio, I I . , Ptak, W., Owcw, M. J., Hayday, A. C.. and Askenase, P. \V. (1996). y6 T cells from tolerized ap T cell receptor (TCR)-deficient mice inliibit contact sensitivity-effectorT cells in vivo. and their interferon-y production in citro. J . Exp. Mcd. 184, 2129-2139.
142
\VILLI BORN
“1
al
Takada, H., Matsuzaki, G., €Iiromatsu, K., and Nornoto, K. (1994). Analysis of the role of natural killer cells in Listerict triotiocytogrne.~infection: Relation between natural killer cc.lls and T-cell receptor y6 T cells in the host defence niechanism at the early stage of infection. Zititt~u~~ology 82, 106-112. Takagaki, Y., DeCloux, A,, Bonneville, M., and Tonegawa, S . (198%). Diversity of y6 Tcell receptors on inurine intestinal intraepitlielial lymphocytes. Notiire 339, 712-714. Takagaki, Y., Nakanishi, N.. Ishida, I., Kanagawa, O., and Tonegawa, S. (198913). T cell receptor-y and -6 genes preferentially utilized by adiilt thymocytes for the surface expression. J . Im?tiutiol. 142, 2112-2121. Tanaka, Y., Morita, C. T., Tanaka, Y., Nieves, E.. Brenner, M. B., and Bloom, B. H. ( 19Yq5). Natural and synthetic non-peptide antigens recognized by human y6 T cells. Notrirc 375, 155-158. Tanaka, Y., Sano, S., Nieves, E., De Libero, G., Rosa, D., Modlin, H. L., Brenner, M. B., Bloom, B. R., and Morita, C. T. (1994). Nonpeptide ligands for human y6 T cells. Yroc. N ( ~ t lAcccd. . sci. USA 91, 8175-8179. Tigelaar, R. E., and Lewis, J. M. (1994).I n “Basic Mechanisms of Physiologic and Abr:rrant Lyniplioproliferation in the Skin” (W. C . Lambert, B. Giannotti, and W. van Vloten, eds.), pp, 39-55. Plenum Press, New York. Tripp, C. S., Wolf, S. F., and Unanue, E. R. (1993). Interleukin 12 and tumor necrosis factor (Y are costirnulators of interferon y production by natural killer cells in severe combined irnir~uiiotleficiei~cy inice with listeriosis, and iiiterleukin 10 is a physiologic antagonist. Proc. Notl. Acrrd. Sci. USA 90, 3725-3729. Tschachler, E., Scliuler, G., Hutterer, J., Leibl. H., Wolff, K., and Stingl, G. (1983). Expression of Thy-I antigen by inurine epiderinal T cells. J . Ztiuest. Denn. 81, 282-285. Tsuji, M., Eyster, C., O’Brien, R. L., Born, M’. K., Bapna, M., Reichel, M., Nussenmeig, R. S., andzavala, F. (1996).Phenotypic and functional properties of murine y8Tcell clones derived from malaria iriimunized (YOT cell-deficient mice. Znt. Zmmunol. 8, 359-366. Tsuji, M., Mombaerts, P., Lefrancois, L., Nussenmeig, R. S., Zavala, F., and Tonegawa, S. (1994). y6 T cells contribute to immunity against the liver stages of malaria in (YOTcell-deficient mice. Proc. Nntl. Acod. Sci. USA 91, 345-349. Tsuzuki, T., Yoshikai, Y., Ito, M., Mori, N., Ohbayashi, M., and Asai, J. (1994). Kinetics of intestinal intraepithelial lymphocytes during acute graft-versus-host disease in mice. Eur. /. Imnnunol. 24, 709-715. Unanue, E. R. (1997).Inter-relationship among macrophages, natural killer cells and neutrophils in early stages of Listerin resistance. Ciirr. Opin. Imtnunol. 9, 35-43. Us:iini, J., Hiromatsu, K., Matsumoto, Y., Maeda, K., Inagaki, H., Suzuki, T., and Yoshikai, Y.(1995).A protective role of y6 T cells in primary infection with L i s t d o t~umxytogenes in autoirnniune non-obese diabetic mice. Ittinircnology 86, 199-205. Uyeinura, K., Deans, R. J., Band, H., Ohintan, J., Pancliainoorthy, G., Morita, C., Rea, T. H., and Modlin, R. L. (1991). Evidence for clonal selection of y/S T-cells in response to a human pathogen. J. E x p . Mcrl. 174, 683-692. Uyemura, K., Klotz, J., Pirinez, C., Ohmen, J., Wang, X.-H., Ho, C., Hoffman, W. L., and Modlin, R. L. (1992). Microanatomic clonality of y6 T cells in human leischinaniasis lesions. J , hi?tumd. 148, 1205-121 1. Vaessen, L. M. H., chiwehand, A. J., Baan, C. C., Jutte, N . H. P. M., Balk, A. 1%. M. M., Claas, F. H. J., and Weimar, W. (1991). Phenotypic and functional analysis of T cell receptor 78-bearing cells isolated from human heart allografts.]. h ~ mt 4 n d .147,846-850. Vaessen, L. M. B., Schipper, F., Knoop, C., Claas, F. H. J., and Weimar, W. (1996). Inverted V61N62 ratio within the T cell receptor (TCR)-yG T cell population in peripheral blood of heart transplant recipients. C . in Exp. Ztrttnunol. 103, 119-124.
IblMUNOHEGU1,ATORY F U N C T I O N S OF yS T CELLS
143
Vandekerckliovt.. R . A. E., Datenia, C.,Ktriniiig, F., Goulniy. E., Persijii, G. G.. \’:in R o o d , J. I., Claas, F. 11. J.. and Dc Vries. J. E. (1990). Analysis o f the donor-specific cytotoxic T Iyiiiphocyte repertoire in a patient with a long term suni\ing allograft: Frequen specificity, and phenot.ype of tlonar-reacti\,e T cell rweptor (TCK)-aP’ ;ind TCR-yG cio1nes..I r t t 1 7 t l l ~ ~ l ~144, )/. iw-12~4. \‘an der Ileytle, H. C . , Elloso, M. M., Chaiig, W.-L.. Kaplan, M., Manning, D. D., and \Veitlanz. CV. 1’. ( 1995). y6 T cells fiiirctioii in cell-inetliated iinmunity to acutcx bloodstage Pkistiioditrrti dtabafidi nr/at?ii m;ilaria. /. Z t t i r i i u i i d . 154, 3985-3990. van tier Heyde, €1. C.. Elloso, M . M., Roopenian. D. C:.. Maiming, D. D.. and Weidariz, \Y. P. (199:3a), Expansion of the C113 ,CIX yS T cell sul)sct in the spleens of niicc tirlrillg rloI~-ietililiiliood-stage I11;tianLI.El4,l 1. r ~ t t ~ r ~ l l 23, r l o ~1846-1850. . \7an tier Heyde, H . C., Manning, 1). D., and Weitlanz, \V.P. (199:3b). Role of CD4’ T cells in the expansion of the CD4-, CDS- yG T cell siibset i i i the spleens of mice during blood-stage inalaria. /. Z r n t i i u t i o / . 151, 6311-6317. Vassalli, P. (1992). The pathophysiolo~of tiii~iornwrosis factors. Antiti. Hrc. Z i t i t i i u i i d . 10, 411-452. Vitlovic, D., and Dembic, Z.(1991). @I-I restricted g:iinma d d t a T cc4ls can help B cells. Curr. Top. Alicrobiol. I r t ~ r t i i r r t o l , 173, 239-244. Vietor, I*., and Koniug, F. (1990). y6 T-cell receptor rcapertoire iii huinaii peripherd blood a i d thyiirris. Z t t i i r t ~ ~ n ~ ~ r 31, i i ~ t340-346. ic~. Vincent, M. S., Roessiler, K . , Lynch. D., \Vilsoii, D.. Cooper. S. M., Tschopp, J., Sigal, L. H., and Budd, R. C. (199G). Apoptosis of Fas-high CD4* spovial T cells by horreliareactive Fas-ligand (high) y6 T cells iii Lyirie d i n t i s . /. bkp. ‘!fed 184, 2109-2117. W e i n t ~ ,11. L., Frietlman, A , . Miller, A , , Kliorrry, S. J., Al-Salhagh. A,, Saritos, L.. Sayegh, %I., Nnssenhlatt. R. B., Trentham, D. E., ; i d Hafler, D . A. (1994). Oral tolerance: Iininiunologic iriechanisms and treatment of:iniirial and human organ-specific autoimmune diseases by oral administration of autoantigens. At lt iu H m ZrnrtitirioZ. 12, 809-837. Weintraul). B., Eckinaiin. L., Okairwto, S.. Hense, M., llediick, S. M., and Fierer, J. (1997). Role of c.P and y6 T cells in tlir host rrsponse to Sdrtioricllri infection as demonstrated iri T-cell-rcceptor-deficient mice or defined Zty genotypes. Infect. Irnitiuii. 65,2306-2312. Weintraub, B. C., Jackson. M. R., and Hetlrick, S. M. (1994). y6 T cells can recognize nonclassical MHC i n the absence of corrveiitional ailtigenic peptides. 1.Zminuno/. 153, 305 1-3058. Wells. F. B., Gahm, S.-J., Hedrick, S. M.. Bluestone. J. A , , Deiit, A,, and Matis, L. A. (1991). Reyuirenwnt for positive selection of yG reccptol--be~iriiigT cells. Science 253, 903-905. W d l s . F. B., Tatsuiini, Y.. Bluestone, J . A , , Hrdrick, S. M., Allison, J. P.. and Matis, L. A. (1993). Plienotypic and fimctional an;ilysis of positive selection in the y/S T cell lineage. 1.E.xp. Mid. 177, 1061-1070. Welsh. R. M., Lin, M.-Y.. Lohtnan, H. I,.. \’arga. S. M.. Zaronainski. C. C., and Selin, L. K. (199i). a11 and y6 T-cell networks a r i d their r o k s in natural resistance to b i r d infc~tions.Itnrrrrttiol. Rcc. 159, 79-93. Wen, I,.. Pao, W., \Vong, F . S., Prng, Q.?Craft, J., Zlieng. B., Kelsoe, G.. Dianda, L., Owen, M. J., and Hayday, A. C. (1996). Cerminal center forination, iminunogkhulin class s\vitchiirg. and autoai1til)ocly protlriction diiveil by “inon a@” T cells. /. Esp. hlrd
183,2271-22&2. Wcn. L., Roberts. S. J., Viriey, J. L.. CYong, F. S., Mallick, C., Findly. R. C . , Peng, Q., Craft, J. E., Owen, M. J.. and Hayday. A . C. (1994). Iinmunoglobulin synthesis arid generalized aritoiiiriiiiiniv in inice congenitally deficient in Cup( +) T cells. Nature 369, 654-658.
144
WILL1 B O R N rf ol.
Werner, S., Peters, K. G., Longaker, M. T., Pace, F. F., Banda, M. J., and Williams, L. T. (1992). Large induction of keratinocyte growth factor expression in the dermis during wound healing. Proc. Natl. Acnd. Sci. USA 89, 6896-6900. Werner, S., Smola, H., Liao, X., Longaker, M. T., Krieg, T., Hofschneider, P. H., and Williams, L. T. (1994).The function of KGF in inorphogenesis of epithelium and reepithelialization of wounds. Science 266, 819-822. Wildner, G., Hiinig, T., and Thurau, S. R. (1996). Orally induced, peptide-specific y/S TCR' cells suppress experimental autoimmune uveitis. Eur. /. Imrriunol. 26,2140-2148. Wilhelm, M., and Tony, H.-P. (1994). An in vitro model for the expansion of Vy962 T lymphocytes during devebpment. S m n d 1.bmnutlo~.40, 521-528. Winoto, A,, and Baltimore, D. (1989a). cup lineage-specific expression of a T cell receptor gene by nearby silencers. Cell 59, 649-655. Winoto, A,, and Baltimore, D. (1989b).Separate lineages of T cells expressing the afi and y6 receptors. Nature 338, 430-432. Wright, A., Lee, J. E., Link, M. P., Smith, S. D., Carroll, W., Levy, R., Clayberger, C., and Krensky, A. M. (1989). Cytotoxic T lymphocytes specific for self tumor immunoglobulin express T cell receptor 6 chain. J. Exp. Med. 169, 1557-1564. Yeh, C.-J. G., Bulmer, J. N., Hsi, B.-L., Tian, W.-T., Rittershaus, C., and Ip, S. H. (1990). Monoclonal antibodies to T cell receptor ylS complex react with human endometrial glandular epithelium. Placenta 11, 253-261. Yokota, K., Ariizumi, K., Kitajima, T., Bergstresser, P. R., Street, N. E., and Takashima. A. (1996). Cytokine-mediated communication between dendritic epidermal T cells and Langerhans cells: In oitro studies using cell lines. /. ZniinuiioL 157, 1529-1537. Yoshimoto, T., Bendelac, A,, Watson, C., Hu-Li, J,, and Paul, W. E. (1995). Role of NK1.I T cells in a TH2response and in immunoglobulin E production. Science 270, 1845-1847. Zhang, X.-M., Tonnelle, C.. Lefranc, M.-P., and Huck, S. (1994). T cell receptor y cDNA in human fetal liver mtl thymus: Variable regions of y chains are restricted to Vyl or V9, due to the absence of splicing of the V10 and V11 leader intron. Eur. /. Iniinunol. 24, 571-578. Zhang, Y., Cado, D., Asaniow, D. M., Komori, T., Alt, F. W., Raulet, D. H., and Allison, J. P. (1995). The role of short homology reports and TdT in generation of the invariant yS antigen receptor repertoire in the fetal thymus. Iintnunity 3, 439-447. Zijlstra, M., Bix, M., Simister, N . E., Loring, J. M., Raulet, D. H., and Jaenisch, R. (1990). fie-microglobulin deficient mice lack CD4-8+ cytolytiic T cells. Nature 344, 742-746. +
This article was accepted for publication on March 24, 1998
STATs as Mediators of Cytokine-Induced Responses TIMOTHY HOEY' AND MICHAEL J. GRUSBYt 'rulorik, Inc., Souh Son Fmncisco, Colifornio 94080 ond beparfment of immunology ond Infectious Diseases, Homrd School of Public Heolh, Deparfment of Medicine, Horvord Medico1 School, Boston, Mossochusetis 02 1 15
I. Introduction
Cytokines are a family of secreted proteins that have important roles in regulating the growth and differentiation of multiple cell types. Although it has long been appreciated that the interaction of cytohnes with their specific cell surfilce receptors results in the induction of new gene transcription, it is only very recently tliat tlie signaling pathways leading from the cytokine receptor to the nucleus have been elucidated. Two novel families of proteins, Jaks ( Janus fhmily tyrosine kinases) and STATs (signal transducers and activators of transcription), have been identified and shown to be important mediators of cytokine-induced signaling. Although originally characterized using interferon (IFN) as a model cytokine, it is now clear that the Jak-STAT signaling pathway is critically important for mediating the biologic effects of a nuniber of different cytokines. Several excellent reviews have lieen published on various aspects of the Jak-STAT signaling pathway (Ihle ct d., 1995; Schindler and Darnell, 199Fj; Ihle, 1996;Darnell, 1997; O'Shea, 1997; Leonard and O'Shea, 1998). The aim of this review is to focus on the biology and biochemistry of the STAT proteins. It reviews how STAT structure is related to function, considers what STAT-deficient mice have told us about the biology of STATs, and discusses new advances in our understanding of how STAT function is regulated. II. The STAT Gene Family
STAT proteins were first identified as components of a DNA-binchng coniplex induced rapidly in response to IFN stimulation. IFN-a induced the activation of a multiprotein coinplex composed of a 11$3-kDaand either a 91- or a 84-kDa protein together with a 48-kDa DNA-binding protein (Fu et a/., 1990). This coinplex was termed interferon-stimulated gene factor-3 ( ISGF-3), and subsequent work demonstrated that tlie 113-kDa protein and tlie 91/84-kDa proteins (which arise from differential splicing of tlie saine gene, see later) became phosphorylated on tyrosine following IFN-a stimulation (Schindler et ol., 1992b). Interestingly, IFN-.)Istimulation led to tlie activation of a different protein complex composed only of the 91-kDa protein and termed ganiina-activated factor (GAF) (Sliuai et 115
( qnnzlat C I W J tn \~.&I!IK Pr?v 211 nglit, 01 r t p r d u c t ~ 111 w nit lnnn re\enc=d OIlBi Zii(dY9 $ 3 0 IJIJ
146
TIMOTIIY IIOEY AND MICHAEL J GRUSBY
al., 1992). The subsequent cloning of the genes encoding these proteins (Fu et al., 1992; Schindler et al., 1992a) led to the eventual identification of the 91-kDa protein (activated by both IFN-a and IFN-7) as Statl and the 113-kDa protein (activated only by IFN-a) as Stat2. Since then, five additional members of this gene family have been cloned, including Stat3 (Akiraet nl., 1994;Zhonget al., 1994), Stat4 (Yamamotoet al., 1994; Zhong et al., 1994), Stats Sa and Sb (Wakao et nl., 1994; Hou et nl., 1995; Liu et nl., 1995; Mui et al., 1995) and Stat6 (Hou et al., 1994; Quelle et al., 1995). It has been pointed out that additional members of the STAT family have been sought through the use of polymerase chain reaction and homologybased screening methods to no avail (Ihle, 1996),suggesting that the seven genes identified to date probably represent the entire gene family. STAT proteins appear to be part of an evolutionarily conserved signal transduction pathway as homologues have been identified in Drosophila (Yan Pt al., 1996) and more recently in Dictyostelium (Kawata et al., 1997). In the mouse, Stat genes are found tightly linked in three clusters (Copeland et al., 1995). Statl and Stat4 are colocalized on chromosome 1, Stat2 and Stat6 are found on chromosome 10, and Stat3, Stada, and StatSb reside on chromosome 11. These observations have led to the suggestion that the family arose through the tandem duplication of an ancestral gene, followed by duplication events of the linked genes and their subsequent dispersion to other chromosomes. As the STAT gene identified in Drosophila is most closely related to StatS, it is possible that this STAT represents the ancestral gene that was then duplicated to yield the Stat3 gene. The high degree of sequence conservation between Stat5a and StatSb suggests that they arose via a more recent gene duplication event. 111. Structural and Functional Domains in STAT Proteins
A. TIIESH2 DOMAIN A N D R~~CEFTOH BINDING The functional domains that have been identified in STAT proteins are shown in Fig. 1. One striking feature of the STAT Family of transcription factors is the presence of a Src homology 2 (SH2) domain. This domain was first identified on receptor tyrosine kinases and functions to bind specific tyrosine-phospliorylated protein sequences (Koch et al., 1991). The STAT SH2 domains, however, represent a subfamily of SH2 domains. For example, this region is approximately 75% identical among STATs 1, 3,4, and 5 , whereas the SH2 domains of Statl and Stat6, the most distantly related members of the STAT family, are 35% identical. In contrast, the conservation between the STAT and the src SH2 domain is much lower. The STAT SH2 domains are only 1Ei-20% identical to src, and the alignment between the STATs and src includes many gaps and nonconservative
STAT\
147
Frc:. 1. Functional doniains of STAT protfsins. Several frinctioiial doniains of STAT molecdrs ha\^ been mapped. These include tliv N-tcrmiiial interaction chnain requircd for tc,tramrrization, a centrd DNA-binding domaiii. ii conserved SH2 tloniain, a conscned tyrosine residue that liecomes phosplio~latetl(111 ;wtivatioii. and ii C-tcrminal transcription activation domain.
substitutions ( Fu, 1992). The highest dcgree of consemition is surrounding the pliospliotyrosine-binding pocket around residue 600 near tlie N-terminal region of the SH2 domain. Several lines of evidence indicate that the SH2 domain-pliospliopeptide interaction provides the specificity for recruitment of the different STAT proteins to various cytokine receI&m. For example, exchange of the Stat1 and Stat2 SH2 domains resulted in an alteration of tlie receptor specificity for the chimeric proteins (Heim et d., 1995). In a reciprocal manner, it was sliown that insertion of the Stat3 docking site sequence YXXQ (see Table I ) derived from a 1 3 0 to the ervthropoietin (EPO) receptor resulted in recruitment of stat3 rather than Stat5 (Stall1 et nl., 1995). In several cases, synthetic peptides d e r i \ ~ from ~ l a cytolune receptor have been found to bind to STAT proteins iti vitro. For example, a phospliorylated peptide derived from the IFN-y receptor intracellular domain was shown to bind STAT1 nionoiners (Creenlund et al., 1994). A useful assay for STAT-peptide interaction is based on the fact that receptorderived peptides can disnipt STAT diiners, leading to an inhibition of DNA-binding activity (Hou et nl., 1994).As mentioned earlier, each STAT SH2 domain appears to have unique peptide-binding specificity. Sequences that can be bound by tlie different STAT proteins, and the receptors from which they are derived, are listed in Table I. Based on studies with src, tlie sequences C-terminal to the phospliotyrosine residue are critical for binding specificity (Songyang et al., 1993). In general, residues at the 1 and + 3 position relative to the phosphotyrosine appear to govern STAT binding to the receptor. Taken together, these data are consistent with the idea that receptor-binding specificity is mediated by the SH2 domain of the STAT protein and the sequence immediately downstream of the phosphotyrosine in the receptor. However, other regions of the STAT protein may also play a role in recruitment to the receptor. The best evidence for this is the observation that deletion of the N-terminal region
+
148
TIMOTHY IlOEY A N D MICHAEL J. G H U S H Y
STAT
Receptor
Sequence
Position
Stat1 state Stat3
IFN-YR IFN-CYH gp130
pYDKPH pYL’FFR pYRHQV pYFKQN pYLPQT pYQPQA pYLPSN pYLSLQ pY LVLD pYKAFS pYKPFQ
440 466 767 814 905 96 1 800 5 10 343 578 606
Stat4 Stat5 Stat6
LIF-R IL-1BR IL-2R Epo-R IL-4n
of Stat2 led to decreased phosphorylation in response to IFN-a, suggesting an inability of the STAT molecule to bind to the receptor (Qureshi et al., 1996). It is not known, however, to what extent this is the case for other STAT Iuroteins as well. For Stat6 at least. the N-terminal domain is not required for tyrosine phosphorylation in response to IL-4 (Mikita et al., 1996). Subsequent to its recruitment to the cytokine receptor, the STAT protein is phosphorylated by Jak kinases on a tyrosine residue located downstream of the SH2 domain around residue 700. Based on in vitro kinase assays and transfection experiments, there does not appear to be much, if any, specificity in the phosphorylation of STAT proteins by the Jaks (reviewed in Ihle, 1996). Furthermore, selective protein interactions between STAT proteins and Jak kinases have not been reported. Thus, the specificity of STAT activation is determined by interaction with the cytokine receptor instead of the Jak kinase.
B. DIMERIZATION Subsequent to their tyrosine phosphorylation, STAT proteins leave the receptor and form dimers through a reciprocal interaction between the SH2 domain of one molecule and the phosphotyrosine of the other. The binding affinities between STAT SH2 domains and receptor-derived peptides are, in general, much greater than those for the interaction of the STAT SH2 domain and the STAT phosphotyrosine peptide and, similar to other SH2 domain-phosphotyrosine peptide interactions, the association and dissociation rates are very fast (Greenlund et al., 1995). An important parameter in STAT dimerization is that the strength of binding in the dimer is the square of the individual SH2-phosphotyrosine peptide affinities.
STATs
149
Therefore, following tyrosine phosphorylation and rapid dissociation from the receptor, STAT proteins have a strong preference for dimerization rather than reassociatioil with the cytokine receptor. It is not known to what extent other regions of the STAT protein contribute to diinerization. Interestingly though, the region from approximately 250 to 310 contains a heptad repeat of hydrophobic residues likely to form an a-helix and similar to the leucine zipper motif involved in the diinerization of other DNA-binding proteins, such as those of the 13-Zip and B-HLH classes (Jones, 1990).
C. N~JCI,EAR LOCALIZATION The transport of proteins from the cytoplasm to the nucleus usually requires a protein sequence termed a nuclear localization signal (NLS). The NLS has not been deterniined for any STAT protein. The activities of Ran, a sniall GTPase that plays a general role in nuclear import, and NPI-1, a component of the nuclear pore-targeting complex, appear to be required for the nuclear transport of Statl after IFN-.)I signaling (Sehrnoto et nl., 1996, 1997). Moreover, experiments with mutant forins of Statl have indicated that dimerization is required for nuclear accuinulation of the protein, a s mutation of Y701 in STAT1 eliminated STAT diinerization as well as nuclear transport (Sekimoto et d ,1996). It has been reported, however, that this same mutant Statl protein is nearly a s effective as wildtype Statl in restoring caspase expression in Statl-deficient cells (Kumar et al., 1997). This observation leaves open the intriguing possibility that low levels of STAT inonomers may be able to niake their way to the nucleus and act as transcriptional regulators through protein-protein rather than protein-DNA complexes. D. DNA BINDING The STAT DNA-binding domain has been identified as a region in the central part of the inolecule between ainino acids 300 and 500. The experimental evidence for this was provided by analyzing the DNA-binding specificity of Stat1 and Stat6 chiineric proteins (Schindler et ut., 1995). Using a similar strategy with Statl and Stat3 hybrids, the DNA-binding specificity region was found to reside between residues 400 and 500 (Horvath ct nl., 199-5).This sequence is not related to any previously characterized DNA-binding stnictural motif. Three classes of DNA-binding sites have been described for the STAT family. The first is the interferon-stimulated regulatory clement ( ISRE) that inediates transcriptional regulation in response to IFN-a. This site binds ISGF-3, consisting of a Statl/Stat2 heterodiiner plus p48 ( F u et cd., 1990). All three proteins in the complex appear to contact DNA directly
150
TIMOTHY HOET AND MI(:HAEL
J.
GHUSRY
(Qureshi et al., 1995). The second class of cis-acting sites was originally defined as IFN-?-activated sequence (GAS) elements. These sites bind homodimers of Statl (Shuai et al., 1993). Subsequently, it was shown that Statl, Stat3, Stat4, and Stat5 can all recognize the same site. This binding site is a palindrome with the sequence TTCNNNGAA ( N 3 site). The highest affinity sites contain CRG ( R = C or G) as the central three nucleotides (Horvath et nl., 1995). IL-4-regulated gene transcription is mediated by a closely related sequence, TTCNNNNGAA (N4 site), that differs in the spacing between the inverted repeats. This binding site corresponds to the optimal recognition site for Stat6 (Schindler et al., 1995). Thus, for STAT hornoQmers, only Stat6 clearly has a distinct specificity and can selectively recognize the IL-4 response element. Given that most STAT proteins bind to a similar DNA sequence, the question arises of how specificity in transcriptional regulation is achieved. Analysis of STAT-binding sites in various promoters has indicated that these sites are often clustered together (Guyer et nl., 1995; John et nl., 1996; Xu et al., 1996). The proximity of the sites allows dimer-dimer interactions that enable STAT proteins to selectively recognize sites that are diverged from the optimal consensus sequence and that are individually lower in affinity (Vinkemeier et al., 1996; Xu et al., 1996). Cooperative binding interactions are mediated by the association of adjacently bound diiners through the N-terminal region of the proteins (Vinkemeier et al., 1996; Xu et d., 1996). Analysis of the crystal structure of the N-terminal domain of Stat4 indicates that this region is composed of eight helices that form a hook-like structure, which appears to mediate dimer-dimer interactions (Vinkemeier et al., 1998). This domain is highly conserved among the STAT family, being between 25 and 50% identical among the six STATs. The N-terminal domain is not essential for dimerization or for binding to single high-affinity site, however (Xu et al., 1996). Interestingly, the sequence of Stat6 in this region is the most diverged among the STAT family. Because Stat6 has the unique ability to recognize the N4 consensus sequence, this protein apparently does not use the mechanism of cooperative binding to adjacent sites to achieve selectivity.
E. TRANSCRIPTIONAL ACTIVATION Subsequent to DNA binding, the STAT proteins function to activate or, in some cases, repress gene transcription. The STAT transcriptional activation domain was first identified as the C-terminal domain in Statl and Stat3 (Wen et al., 1995; Bromberg et al., 1996). The corresponding region also functions in transcription activation in the other STATs (Mikita et al., 1996; Moriggl et al., 1996; Qureshi et al., 1996). Interestingly, this region of the protein is deleted in the STATlP form, which is derived
STAT\
151
from an alternatively spliced form of the Stat 1 transcript. Tlie expression of two naturally occnrring isoforms of STAT proteins that either have (a form) or lack ( p forin) tlie C-terminal domain has also been found for 1996), StatFj (Wang et nl., StatS (Schaefer rt nl., 1995; Caldenlioven clt d., 19961, and Stat6 (Patel et a/., 1998). Generally, the /3 forms of the STAT nioleciiles have been shown to act as dominant negative inhibitors of their a forin counterparts. Statl and Stat3 proteins can also lie serine phosphorylated in their C-terminal domains, and this modification appears to positivily regulate tile potency of the activation domain (Wen ct a/., 1995; Zhang et d., 1995). At present, however, tlie precise niechanisms by which STAT proteins mediate transcriptional activation remain to be elucidated. In addition to the protein interactions between STAT dimers described earlier, interactions of STAT proteins with other transcription factors also occur and, in inany cases, are both critical for transcriptional activation and promoter selectivity. Tlie most well-studied case is the ISGF-3 complex, consisting of StatllStat2 and p48. The domain of Statl that interacts with p48 has been identified as the region between amino acids lS0 and 2SO (Horvath et nl., 1996). Although a StatUStat2 heterodimer alone can hind to a palindrornic N 3 sequence characteristic of the GAS sites and direct IFN-a-induced transcription (Li ct d., 1996), interaction with p48 alters the DNA-binding specificity of the complex such that it now binds to the ISRE site. p48 is a member of the IRF fhmily of transcription factors (Harada et nl., 1989).These proteins have been implicated in many aspects of iinrnune function, including cytokme signaling arid cellular proliferation. Thus. it is tempting to speculate that other members of the IRF gene family may interact with STAT proteins in a manner similar to p48. In iddition to p48, STAT proteins have been shown to interact functionally with several other classes of transcription factors. For example, Statl and Stat2 have been shown to associate with the transcriptional coactivator p300/CBP (Bhattachaiya et d . ,1996; Zliang vt a / ., 1996a).Two interaction sites on Statl, comprising either the N- or the C-terminal domains, were defined for its interaction with p3OO/CBP, although a transcriptional activation fiinction for the N-terminal domain had not previously been reported. In tlie case of State, the C-terniinal domain w a s found to be the site for p300/CBP interaction. Interestingly, the Stat2 activation domain is unique among tlie STAT family in that it is rich in acidic amino acids. I n contrast to its interaction with CREB or jun proteins, the binding of p300/CBP to Statl or Stat2 does not appear' to be dependent on phosphorylation of the STAT protein. Finally, studies with pSOO/CBP have indicated that these proteins have histone acetylase activity (Bannister and Kouzarides, 1996; Ogryzko et ol., 1996). Acetylated histones have been associated with
1.52
TIMOTIIY HOEY AND blI(:I-IAEL J. CRUSBY
active genes and an “open” chromatin configuration (reviewed in Turner, 1993).Thus, it may be that this chromatin remodeling activity allows access to the promoter for other transcription factors and is the result of synergistic interactions between STAT proteins and other transcriptional regulators. An interaction between c-jun and Stilt3 was discovered through a twohybrid screen to look for proteins that can bind the N-terminal domain of c-jun (Shaefer et nl., 1995). Stat3 can stimulate expression of the IL-6 response element in the a2-macroglobulin promoter by worhng together with c-jun. Interestingly, Sliaefer et $. (1995) isolated an alternatively spliced version of Stat3 that lacks the C-terminal transcription activation domain. This /3 form of Stat3 was more effective than the full length when tested in combination with c-jun, and it may be that the short form of Stat3 is specialized for c-jun interaction. Stat5 was found to interact functionally with glucocorticoid and progesterone receptors (Stocklin et nl., 1996).These proteins mediate signaling in response to prolactin and steroid hormones and are involved in activation of the &casein promoter. Similarly, Stat3 binds to, and enhances, the transcriptional activity of the glucocorticoid receptor (Zliang et nl., 1997). Finally, IL-4 response elements are bound by Stat6 and c/EBP proteins (Hou et d , 1994; Delphin and Stavnezer, 1995).Activation of transcription in response to IL-4 from these composite elements requires the transcription activation domain of both proteins, and they appear to interact on DNA to stabilize each other’s binding (Mikita et nl., 1996).A theme that has emerged from these studies is that STAT proteins are often involved in combinatorid interactions with other transcriptional activators. Indeed, in several cases it appears that STAT proteins may not be able to maximally activate gene transcription independently of these interactions. For example, Statl synergistically activates the ICAM-1 promoter in coinbination with Spl; mutation of the Spl site eliminates IFN-?-inducible transcription without affecting Statl binding (Look et al., 1995). IV. STAT-Deficient Mice
With the exception of Stat2, each of the STAT genes has now been functionally inactivated in mice by gene targeting (Table 11). Perhaps the most important conclusion that can be drawn from the collective phenotypes of the STAT-deficient mice is that, in viva, STAT activation seems to be biologically critical for only a select group of cytokines. This is in marked contrast to the in uitro observations demonstrating that almost every known cytokine can activate one or more STAT proteins. For example, a number of ligands other than the IFNs, including IL-6 (Lamer et nl., 1993), IL-10 (Finblooin and Winestock, 1995), and epidermal growth
STAT\
153
factor (Fu and Zhang, 1993), had been sliown, in citro, to lead to the activation of Statl on ligand binding to their specific cell surface receptors. Despite these prior observations, no defects in the in uico responsiveness of Statl-deficient mice to these cytokines and hormones were noted. Siinilarly, leptiri was shown to activate Stat6 in uitro (Chilardi et d., 1996),yet Stat6deficient mice are not obese. Thus, even if these ligands do activate thc Jak-STAT pathway in zjico, then clearly they must be able to activate other biologically relevant signaling pathways as well.
A. STAT~-DICFICIENT MI<E The Statl gene was the first member ofthe STAT family to be disriiptetl d.,1996). Statl-deficient aninials are developmentally nornial and are Iiom at thc expected Mendelian frequencies. Perliaps not unexpectedly, the predominant phenotype of Stat 1deficient mice is the complete lack of responsiveness to both IFN-a and IFN-7. As a resiilt, the transcriptional activation of IFN-inducible genes such 21s IRF-1 a i d MHC class I and class I1 molecules is coiripletely impaired in the mutant mice. The functional consequence of this lack of IFN responsiveness is that Statl-deficient animals are highly susceptible to infections with viruses such as MEW and VSV and with microbial pathogens such as Listeria ~nonoc!itogc.n~~. What was perhaps most surprising from the analysis of this first STAT knockout was the demonstration of the plivsiologic specificity of the JakSTAT signaling pathway. As discussed earlier, the analysis of thcw animals demonstrated that while Statl is critical for inediating biologic responses to IFNs, activation of this transcription factor is clearly not necessary for the responses to a number of other ligaiids. Whetlier the activation of Statl in vitro hy ligands other than IFNs is simply an artifact of nonphysiologic in mice (Diirbin ~t nl., 1996; Meraz ct
154
TIMOTIIY HOEY AN11 MICIIAEL J GRUSBI
stimulation or a minor component of the biologic response to these stimuli remains to be determined.
B. STAT3-DEFIC:IENT MICE Stat3 deficiency leads to embryonic lethality between 6.5 and 7.5 days of gestation (Takeda et al., 1997). Although it is not clear why Stat3deficient embryos fail to complete development, it is interesting to note that a number of different cytokines and growth factors important in growth and differentiation (LIF, EGF, CNTF, G-CSF, etc.) activate Stat3. In this regard, the period of embryonic death observed in Stat3-deficient mice is earlier than that seen in mice deficient in receptors for LIF or CNTF or even the common signal transduction receptor gp130, suggesting that Stat3 deficiency may represent the combined effect of a loss in the receptor signaling of two or more critical growth factors.
c. STAT4-DEFICIENT MICE 1L-12 is the only known ligand capable of activating Stat4, and thus not suprisingly, Stat4-deficient mice show impaired IL- 12-induced responses (Kaplan et al., 1996b; Thierfelder et al., 1996). Thus, IL-12-induced increases in IFN-.), production, cellular proliferation, and natural killer cell cytotoxicity are all abrogated in the absence of Stat4. Moreover, the differentiation of IFN-.), producing T h l cells is markedly impaired in Stat4deficient inice. One group (Kaplan et al., 1996b) noted that this was accompanied by the enhanced development of Th2 cells, consistent with the phenotype of IL-12R-deficient mice (Magram et al., 1996). Interestingly, Stat4 has a very restricted tissue-specific distribution of expression, with high levels expressed in a developmentally restricted manner in the testis (Herrada and Wolgemuth, 1997). Nevertheless, Stat4 deficiency has no discernible effects on spermatogenesis and Stat4-deficient male mice are fertile. D. STAT5-DEFICIENT MICE Perhaps the most surprising finding from the Stat5a and Stat5b knockout mice is their distinct phenotypes, demonstrating that, despite their sequence similarity, these two STAT proteins have distinct functions. Stat5adeficient mice show impaired lobuloaveolar mammopoiesis and females fail to lactate after parturition, consistent with an inability to respond to prolactin (Liu et al., 1997). Although expressed with a similar pattern during mammary gland development, Stat5b clearly cannot compensate for StatFja in maininary gland development. In contrast, Stat5b is required for the sexual dimorphism of body growth rates and growth hormone pulseregulated liver gene expression (Udy et d., 1997).Once again, these results
highlight the degree of biologic specificity in Jak-STAT signaling that has been uncovered in STAT-deficient mice.
E. STATG-DEFKTENT MKT Stat6-deficient mice have been indepentlently generated by three groups, all of which reported similar phenotypes (Kaplan et nl., 1996a; Shimoda et d., 1996; Takeda et nl., 1996b). As mentioned earlier, Stat6 was originally purified and cloned based on its abili? to be activated in response to IL-4 (Hou et a/., 1994), and StatG-deficient mice are niarkedlv impaired in their ability to respond to this cytokine. For example, the expression of genes known to be regulated by IL-4, such a s CD23, IL-4Ra, and MHC class I1 molecules, is not increased following the IL-4 stimulation of Stat6deficient lymphocytes. I n addition, Stat6-deficient mice fiail to generate an IgE response, following either infection with the nematode Nippostrotigylr~sbmsiliensis or immunization with the polyclonal stiinulus anti-IgD. Finally, the differentiation of Th2 cells by Stat6-deficient lymphocytes was f h i d to be niarkedly impaired. Interestingly, however, not all IL4-induced responses are abrogated in Stat6-deficient lymphocytes. For example, IL-4 could rescue Stat6-deficient T cells from undergoing apoptosis (Vella et al., 1997; Kaplan et al., 1998) and probably mediates these effects through an alternate signaling pathway involving the activation of a/., 1996). insulin receptor substrate (Zamorano <>t The IL-4Ra chain is a shared coinpoilent of both IL-4 and IL-13 receptor complexes and. like IL-4, IL-13 is capable of activating Stat6. Not surprisingly, many IL-13-induced responses are also abrogated i'n Stat6-deficient 1996a) mice, including IL-13-driven TI12 differentiation (Kaplan et d., and IL-13-mediated increases in the expression of MHC class I1 antigens and the production of nitric oxide in macropliages (Takeda ut al., 1996a). V. STAT Function in Cellular Proliferation and Disease
The role of STAT proteins in regulating cellular proliferation has been somewhat controversial, although several lines of evidence now suggest that the Jak-STAT signaling pathway is indeed important for this process. Experinients in which cytokine receptors harboring niutations that prevent the docking of STAT proteins often gave conflicting results, in some cases implicating STAT proteins as being important (Danien et al., 1995) and in other cases dispensable (Qiielle a]., 1996) for a proliferative response. In contrast, the expression of a dominant negative Stat5 protein led to impaired IL-3-induced proliferation (Mui et nl., 1996). Perhaps more iinportantly, lympliocytes from inice deficient in Stat4 (Kaplan et nl., 199613; Thierfelder et al., 1996), S t a t 5 (Nakajima et al , l997), and Stat6 (Kaplan c7t
156
TIMOTHY IIOEY A N D MICHAEL J GRUSAY
et nl., 1996a; Sliimoda et al., 1996; Takeda et d., 1996b) show impaired proliferative responses to IL-12, IL-2, and IL-4, respectively. Interestingly, the failure of IL-2 to induce expression of the high-affinity IL-2Ra chain was shown to be responsible for the proliferative defect seen in Stat5adeficient lymphocytes (Nakajiina et nl., 1997). In contrast, failure of IL-4 to upregulate the expression of the IL-4Ra chain was found not to be responsible for the proliferative defect of State-deficient lymphocytes. In thi5 case, Stat6 was shown to regulate tlie expression of ~ 2 7 ~ ' pthe ' , major cdk inhibitor expressed in lymphocytes (Kaplan et al., 1998). Thus, STAT proteins appear to act at multiple levels to regulate cellular proliferation. Perhaps tlie most compelling evidence that STAT proteins are involved in cellular proliferation is tlie observations of constitutive activation of tlie Jak-STAT signaling pathway following the transformation of cells with certain vinises such as HTLV-1 (Migone et nl., 1995) or oncogenes such as v-Src (Yu et nl., 1995), v-Abl (Danial et al., 1995), or v-Eyk (Zong et d , 1996). Moreover, cell lines from a number of patients with lymphoinas and leukemias were sliown to have constitutively activated Jak-STAT activity (Zhang et nl., 1996b; Cliai et al., 1997; Takemoto et al., 1997). Finally, STATs also appear to be important in mediating the antiproliferative effects of IFN-.)I, and do so by regulating the induction of the cdk inhibitor p21""' (Chin et nl., 1996). Interestingly, this activity of Stat1 may be part of the mechanism responsible for thanatophoric dysplasia type I1 dwarfism, where mutations in fibroblast growth factor receptor 3 lead to constitutive tyrosine kniase activity, activation of Statl, and increased expression of p21CLAF' in cartilage from patients with this disease (Su et nl., 1997). VI. Regulation of STAT Function
As outlined earlier, STAT proteins play an important role as regulators of gene transcription in a variety of cellular processes. Thus, it is perhaps not suprising that a number of different mechanisms appear to exist to negatively regulate STAT function. One of these, the generation of alternatively spliced forms that lack a region in the C-terminal domain and thus act as dominant negatives, has already been discussed, although it is at present unclear how the relative levels of the various spliced forms of the STAT molecule are regulated in any given cell. Evidence also shows that the action of an a s yet unidentified pliosphatase results in the dephosphorylation of the critical tyrosine residue required for STAT dimerization and DNA binding (Haspel et nl., 1996; Shuai et nl., 1996). Finally, it has been demonstrated that the degradation of STAT proteins through either the
STAT,
157
1997) or proteosome pathway (Kim and Maniatis, 1996; Rayanade ct d., the action of caspases (King and Goodburn, 1998) can also occur. A novel mechanism for regulating STAT activity has emerged with the identification of two new classes of STAT inhibitors. The first is the CIS/ SOCS/JAB/SSI class of cvtokine-indncible STAT inhibitors that appear to be aide to inhibit STAT activation 1iy binding to either the pliosphorylated Jak kinase or the phosphorylated cytokirie receptor (Yoshimuraet aZ., 1995; Endo ct nl., 1997; Naka ct al., 1997; Starr el nl., 1997). Some 20 proteins containing the so-called SOCS box have now been identified (Hilton et al., 1998), and it will now tie of great interest in determining the exact ~nechanismby which these proteins inhibit specific Jak-STAT signaling events. A second class of proteins, the protein inhibitor of activated STAT (PIAS),has also been identified (Chung ct al., 1997).Two members, PIASZ and PIAS3, have been characterized and shown to interact specifically with Stat1 and Stat3, respectively. Tliur, the PIAS molecules appear to represent specific STAT inhibitors, ;tiid thcre will undoubtedly be great interest in exploring their potential a s tlierapeutic targets. Finally, there are now several exainpks of STAT proteins competing with other transcription factors for overlapping DNA-binding sites. Thus, the Ets fknily protein Elf-1 represses IL-2Ru gene transcription by binding to an Ets-binding site that overlaps with a StatS-binding site (Lecine ct d., 1996). In addition, BCL-6 appears to compete with Stat6 binding to several target genes as BCL-6-deficient mice exhibit augiiiented IL-41997). Finally, Stat6 induced Th2 responses (Dent ct a / . , 1997; Ye et d., can antagonize the binding of other transcription factors, such as NF-KB 1997). suggesting one niedianisin in the E-selectin gene (Bennett et d., by which STAT proteins can act a s transcriptional repressors. VII. Summary and Perspective
In inany respects, it is remarkable the degree to which we understand the biology and biochemistry of the STAT proteins given their recent discovery. Nevertheless, there are still inany important issues related to Jak-STAT signaling that remain to be addrerssed. First, additional structural information of the individual domains in the STAT molecule will be very informative. For example, the STAT SH2 doinuins represent a distinct subfamily relative to the src-related SH2 domains, and therefore it is likely the STAT SH2 domains will have wine unique structural features. Because STAT DNA-binding domains do not show any obvious sequence siinilarity to other DNA-binding motifs, this structure will be the first of its class. Perhaps the most interesting aspect of the structure is that it will allow us
158
TIMOTHY IIOEY AND MICIIAEI,
1
GRUSRY
to gain insight into the interaction of the different domains and how these interactions change during the transition from monoiners to dirners. In addition to structural information, there are inany questions regarding the functions of the STAT family that remain poorly understood. One of the hallmarks of cytokine signaling is the interplay between different signaling pathways. Further investigation of the mechanisms of activation, and inactivation, of STAT proteins will be critical in obtaining a more detailed understanding these processes. In particular, the inechanisrns of transcriptional activation by STAT proteins and their synergistic interactions with other transcription factors will be very important to pursue. Finally, the analysis of STAT-deficient mice has revealed that these proteins are essential for various cytokine signaling pathways during the iininune response. Therefore, we can also look forward to the development of antagonists of STAT function that may prove to be useful in therapeutically modulating immune responses.
REFERENCES Akira, S., Nishio, Y., Inoue, M., Warig, X-J., Wri, S., Matsusaka, T., Yoshida, K., Sudo, T., Nanito, M., iind Kishimoto, T. (1994). Cell 77, 63-71. Baiinister, A. J., and Kouzarides, T. (1996). Il’nttcre 384, 641-643. Bennett, B. L., Cruz, R., Lacson, R. G., a i i c l Manning, A. M. (1997). J , B i d . Chon. 272, 10212-10219. Bhattachalya, S., Eckner, R., Grossinan, S., Oldread, E., Arauy, Z., D’Andrea, A,, and Livingston, D. M. (1996). Naturc, 383, 344-:347. Broinberg, J. F., Morvatli, C. M., Wen, Z.,Srhreiber, R. D., and Damell, J. E. (1996). Proc. N d Accid. Sci. USA 93, 7673-7678. Caldenhoven. E., van Dijk, T. B., Solari, R., Arinstrong, J., Raaijiiiakers, J. A. M., Lainniers, J. J.. Koenderman, L., and de Groot, R. P. (1996). /. Biol. Cherri. 271, 13221-13227. C h i , S. K.. Nichols, G. L., arid Rothmaii, P. ( 1997).1.Imnirrttol. 159, 4720-4728. C h i , Y. E., Kitagawa, M., Su, W.-C. S., You. Z.-H., Iwamoto, Y., and Fu, X.-Y. (1996). Science 272, 71 9-722. Chiuig, C. D., Liao, J., Liu, B., Rao. X.. Jay, l’., Berta. P., and Shuai. K . (1997). Science 278, 1803-1805. Copeland, N. G., Gilbert, D. J.. Schinder, C., Zhorig, Z., Wen, Z., Darnell, J. E., Jr,, Mui, A. L.-F., Miyajima, A., Qiielle, F. W., Ilile, J. N., and Jenkins, N . A. (1995). Genomic:r. 29, 225-228. Damen, J. E., Wakao, H., Miyajinia, A., Krosl, J., Humphries, R. K.. Cutler, R. L., arid Krystal, G.(1995). EMBO 1. 14, 5557-5568. Danial, N. N.. Peruis, A., and Rothinan, P. B. (1995).Science 269, 1875-1877. Damell, J. E., Jr. (1997). Science 277, 1630- 1635. Delphin, S., and Stavnezer, J. (199s)./. E x p Med. 181, 181-192. Dent, A. L., Shaffer, A. L., Yu, X., Allinan, D., arid Staudt, L. M. (1997). Scirrice 276, 589-592. Durbin, J. E., Hackenmiller, H., Sinion, M. C . , and Levy, D. E. (1996). Cell 84, 443-450. Endo, T. A,, Masuliara, M., Yokouchi, M., Suzuki, R., Sakamoto, H., Mitsui, K., Matsumoto, ,4., Taniinura, S., Ohtsubo, M.. Misawa, H., Miyazaki, T., Leonor, N., Taniguchi, T., Fujita, T., Kanakura, Y., Komiya, S., and Yosliimura, A. (1997). Nature 387, 921-924.
STAT.;
Finblooni, D. S.. ant1 Winestock, K. D. (1995).J. I i i i t t w i i o l . 155, 1079-1090, Fu, X.-Y. (1992). Cell 70, 323-335. Fu. X-Y.,Kessler, D. S., \'eals, S. A.. I,e\y, 11. E.. i i ~ d11:irneIl. J. E., Jr. (1990). Proc. Nod. Acntl. Sei. USA 87, 8555-8559. Fu, X-Y., Schinder. C . , lmprota. T., Ael)ersold, K., and Dal.nell, J. E., Jr. (1992). Proc. Not!. Acad. Sci. USA 89, 7840-7843. Fu, X.-Y., and Zliarig, J.-J. (1993).Cell 74, 1135-1145. Cliilardi, N., Zieglrr. S., Wiestiier, A., Stoffel. R.. Hriin, M . H., and Skoda. R. C. (1996). Proc. N d .Actid Sci. U S A 93, 6231-62:35. Greerilund, A. C., Farrar, M . A.. \'iviaiio, B. L., arid Schreiber, R. D. (1994). EMBO /. 13, 1591-1600. Greenluntl, A. C . , Morales, M . O . , Viviaiio, 13. L.. Yan, Y.,Krowleski, J., and Schreiber, R. D. (1995). Z r i i i i i i r i i i t / j 2, 677-687. Guyer. N. B., Sevcms, C . W., Wong. P.. Fengliali, C. A , , antl Wright, T. M . (1995). J . ~ l l i l l i l L l f ~ J155, ~. 3472-3480. Harada, H., Fujita, T.. Miyamoto, M., Kiinuni, Y.,Mariiyima, M . , Furia, A , , Miyata, T.. and Taiiiguchi. T. (1989). Cell 58, 729-739. IIaspel, R. L.. Salditt-Georgieff, M., and Ilarnell, J. E.. Jr. (1996).EhfBOJ. 15, 6262-6268. Heim, M . H., Kerr, I. M., Stark. G. R., antl Darnell, J. E.. Jr. (1995).Scierice 267, 1347-1349, Herrada, G., and Wolgemiith. D. J. (1997).J , Cell. Sci. 110, 1543-1553. Hilton, D. J.. Richardson, R. T.. Alcaxaiitier, \V. S., \'iney, E. M.. Willson, T. A , , Sprogg, N.S., Stair, R., Nicholson. S. E., Metcalf, I)., aiid Nicola, N . A. (1998).Proc. Ntztl. Acad. Sci. USA 95, 114-119. Howath, C. M., Stark, C. R., Kerr, 1. M., and D;iniell, J. E., Jr. (1996). Mol. Cell. B i ~ i l . 13, 3951-3963. Horvath, C. M . , U'en. Z.,and Ilarnell, J. E., Jr. (199.5).Geiics Deo. 9, 984-994. Hou, J.. Schinder, U., Herizel, W.J., Ilo, T. %., Brsseiir, M.,and McKnight, S. L. (1994). Scietice 265, 1701-1706. Hou. J., Schiiidler, U.. Henzel, W. J.. M'ong, S. C., and McKnight, S. L. (1995).Zttimuriily 2, :321-:329. Ihle, J. N. (1996). Cell 84, 3:31-:334. Ihle, J. N.. Witliuln. B. A,, Quelle, F. \V.,Yarnainoto. K., and Silvennoinen, 0. (1995). Aiiiifr, Hec. Zniriiunol. 13, :369-398. John, S.,Robbins, C. M . , and Leonard. M'. J. (1996). EMBO J 15, 5627-5635. Jones, N . (1990). C d l 61, 9-11. Kaplaii, M. €I.,Daniel, C., Schintller, U.. antl Grnshy, M . J. (1998).Mol. Cell. B i d . 18, 19962003. Kaplan. M . H., Schiridlrr, U., Sriiilty, S. T., a r i d Gmshy, M . J. (l996a).Znmrrnity 4,313-:319. Kaplan, bl. H., Sun, Y.-L., Hory. T., antl Grusly, M . J. (1996b). Ntitrrm 382, 174-177. Kawatn. T., Shevclic~iiko.A,, Fukrizawa, M.. Jerinyn. K. A., Totty, N. F., Zliukovskaya, N. \',, Sterling, A. E., Mann, M., and i\'illiuiis, J. G. (1997). Cell 89, 909-916. Kiln, T. K.. and Maniatis, T. (1996). Sciriice 273, 1717-1719. King, P., a i d Goodbum, S. (1998).J. B i d . Clioii. 273, 8699-8704. Koch, C. A , , Anderson, D., Moran, M . F., Ellis, C,, arid Pawson, T. (1991). Sciericc 252, 668-674. Kumar, A , . Coniriiane. M., Flickingrr, T. \V,, H o n d i , C . M., and Stark, G. R. (1997). ScieticP 278, 1630-1632. Lamer, A. C.. Ilavid, M . , Feldman, G. M . , Igarashi, K.-I., Hackett. R. H., Webb, D. S. A . , Sweitzc,r, S. M., Pc~tricoin,E. F., 111. and Finbloom, D. S. (1993). Scieiice 261, 1730-1733.
160
TIMOTHY IIOEY AN11 MICHAEL J. GRUSBY
Lecine, P., Algarte, M., Rameil, P., Beadling, C., Buchler, P., Nabholz, M., and Imbert, J. (1996).Mol. Cell. Biol. 16, 6829-6840. Leonard, W. J., and O’Shea, J. J. (1998). Atinti. Reo. Zrnrnutd 16, 293-322. Li, X., Leung, S., Qureshi, S., Damell, J. E., and Stark, G. R. (1996). I. Biol. Chem. 271, 5790-5794. Lin, X., Robinson, G. W., Gouilleux, F., Groner, B., and Hennighausen, L. (1995). Proc. Natl. Actid. Sci. USA 92, 8831-8835. Liu, X., Robinson, G . W., Wanger, K.-U., Garrett. L., Wynshaw-Boris, A,, and Henighamen, L. (1997). Gcnrs Dev. 11, 179-186. Look, D. C., Pelletier, M. R., Tidwell, R. M., Roswit, W. T., and Holtznian, M. J. (1995). 1. B i d . Chem. 270, 30264-30267. Magrarn, J., Connaughton, S. E., Warner, R. R., Carvajd, D. M., Wu, C.-Y., Ferrante, J., Stewart, C., Sarmiento, U., Faherty, D. A,, and Gately, M. K. (1996).Zmnitcnity 4,471-481. Meraz, M. A,, White, J. M., Sheehan, K. C. F., Bach, E. A,, Rodig, S. J., Dighe, A. S., Kaplan, D. H., Riley, J. K., Greenhind, A. C., Campbell, D., Carver-Moore, K., DuBois, R. N., Clark, R., Agnet, M.. and Schreiber, R . D. (1996). Cell 84, 431-442. Migone, T.-S., Lin, J.-X., Cereseto, A,, Mulloy, J. C., O’Shea, J. J., Franchini, G., and Leonard, W. J. (1995). Science 269, 79-81. Mikita, T., Campbell, D., Wu, P., Williamson, K., and Schintller, U. (1996).Mol. Cell. B i d . 16,581 1-5820. Moriggl, R., Gouilleux-Gruart, V., Jahne, R., Berchtold, S., Gartman, C., Liu, X., Hennighansen, L., Sotiropoulos, A,, Groner, B., and Gouillex, F. (1996).Mol. Cell. Biol. 16,56915700. Mui, A. L.-F., Wakao, I-I., Kinoshita, T., Kitamnra, T., and Miyajima, A. (1996).EMBO J . 15, 242552433. Mni, A. L.-F., Wakao, H., O’Farrell, A,-M., Harada, N., and Miyajirna, A. (1995). E M B O J . 14, 1166-1175. Naka, T., Narazaki, M., Hirata, M., Matsumoto, T., Minamoto, S., Aono, A., Nishirnoto, N., Kajita, T., Taga, T., Yoshizaki, K., A h a , S., and Kishirnoto, T. (1997). Nature 387, 924-929. Nakajiina, H., Liu, X.-W., Wynshaw-Boris, A., Rosenthal, L. A., Imada, K., Finbloom, D. S., Hennighausen, L., and Leonard, W. J. (1997). Znirrwnity 7, 691-701. Ogryzko, V. V., Schlitz, R. L., Russanova, V. Howard, B. H., and Nakatani, Y. (1996). Cell 87,953-959. O’Shea, J. J. (1997). Zntinuriity 7, 1-11. Patel, B. K. R., Pierce, J. H., and LaRochelle. W. J. (1988). Proc. Natl. Acad. Sci. USA 95, 172-177. Quelle, F. W., Shinioda, K., Thierfelder, W., Fischer, C., Kim, A,, Ruben, S. M., Cleveland, J. L., Pierce, J. H., Keegan, A. D., Nelins, K., Paul, W. E., and Ihle, J. N. (1995). Mol. Cell. Biol. 15, 3336-3343. Quelle, F. W., Wang, D., Nosaka, T., Thierfelder, W. E., Stravopodis, D., Weinstein, Y., and Ihle, J. N. (1996). Mol. Cell. B i d . 16, 1622-1631. Qureshi, S. A., Leung, S., Kerr, I. M., Stark, G. R., and Darnell, J. E., Jr. (1996).Mol. Cell. B i d . 16, 288-293. Qureshi, S. A., Salditt-Georgieff, M., and Darnell, J. E., Jr. (1995). Proc. Nntl. Acad. Sci. USA 92, 3829-3833. Rayanade, R. J., Patel, K., Ndubnis, M., Sharnia, S., Omura, S., Etlinger, J. D., Pine, R., and Sehgal, P. B. (1997).J. Biol. Chem. 272, 4659-4662. Schindler, C., and Damell, J. E., Jr. (1995). Armu. Reo. Biochem. 64, 621-651.
STAT\
161
Schindler, C., FII.X.-Y., Improta, T., Aebersold, R., mid Ilarnell, J. E., Jr. (1992a).Proc. N ~ 1 1 Actrd. . Sci. l J S A 89, 7836-7839. Scliiiidler, C., Shuai, K., Prezioso, 1‘.R., and Darnrll, J. E., Jr. (19921)).Scicvicr 257,809-813. Schindlrr. U.,R’II.P.. Rothe, M., Brasseur. M., iciitl McKniglit, S. L. (1YY5). Ztrttnrinif!y 2, 689-697. Sekiiiioto. T., Iiiiaiiioto, N., Nakajiina, K.. Hiraiio, T., iuitl Yoneida, Y. (1997). EMBO J. 16, 7067-7077, Sekiiiioto, T., Nakajinia, K.. Trac1iilxiii:i. T., liiriiiw, T., a d Yoneda, Y. ( 1996). I. B i d . C l u w . 271, 31017-:31020. Shaefer. T. S., Sanders, L. K.. and Nithaiis. D. (19951. Pntc. NafI. Aca4/. Sci. USA 92, 90979101. Shimoda. K., van Deursen, J.. Sarigster. M. I’.,Sariiwar, S. R., Carson, H. T., Tripp, R. A,, Chii, C . , Quelle. F. W., Nosaka, T.. Vigiidi. D. A . A,, I>ohert).,P. C., Grosveld, C . , Paul, \l’ E.,.antl IIilv, J. N . (1996). Notrirc 380, 6:30-63:3. Shuai. K..Liao. J.. ;ant1 Song. M . M . (1Y96). Mo/. CON.Biol. 16, 4932-4941. Shuai, K., Schindler, C., Presiozo. 1’. K., and Ilarncll. J. E., Jr. (1992).Science 258, 18081812. Shiiai. K..Star-k, C . l i . ~Kerr, I. A., ;ind Dtiniell, J. E., Jr. (1993). Science 261, 1744-1746, Sonpmg, Z.,S h o e h i , S. E.. Chaudhiiri. M., Gish. G., Paiwson, T.. liaser. W.G., King, F.. Roberts. T., Ratnofshy. S.. Leclileitlrr. R. J., Ned, H . G., Birge, R. B.. Fajardo, J. E., Chou,M. M.. Hanafusa. H.. Schaffliausen, B., and Cantle), L. C. (1993).Cell 72, 767-778. Stahl. N., Farruggella, T. J , , Uorilton, T. C.. Zhoiig, %., Damell. J. E.. antl Yancopoulos, C:. D. (1995). Scicttrx. 267, 1:349-135:3. Starr, R., Willson, T. A., Vine). E. M., Murray, L. J . I,., Rayiier, 1. R., Jenkins, B. J., Gonda, T. 1.. Alexander, 11;. S., Metcalf. D.. Nicola, N . A,, and Hilton, D. J. (1997). Notrire 387, 917-921. Stocklin, E., Wissler. M., Coilleiis, F.,i n i d Gronei-. B. (1996). Nnfrcr(7 383, 726-728. Su, 11’.-C. S.. Kitagawa. M.. Xiw. N., XY, B., Gm)f’;tlo,S.. Clio. J.. Deng, C., Horton, \V. A , . ant1 FII. X.-Y. (1997). Nature 386, 288-292. Takedii, K., K;iinanaka, M.. Tanaka, T., Kisliimoto. T., and Akira, S. (199Ga).J . Ztwnitnol. 157, 322O-m2. Takeda. K..Nogriclii, K., Slii. U’., Tmaka. T., Matsunioto. M., Yoshitla, N.. Kishiiiioto, T.. and Akira, S. (1995). Proc. Not[. Accid Sci. LISA 94, 3801-3804. Takeda. K., Taiiakn, T.. Shi, \%’.-Matsrinioto, M., Minaini. M., Kasliiwamiira, S., Nakanishi, K., Yoshitla, N.. Kishiinoto. T., a i d Akira, S. (lCJ96h). Notrirc. 380, 625-630. Takemoto, S., Miilloy, J . C., Cereseto, A , . Migone. T.-S.. l’atel, B. K. H.. Matsuoka, M.. Yamaguchi, K., Takatsuki, K.. Kmiiihira. S.. \t’liit(~, J. D., Leonard, \V. 1..M’aldinanii, T., and Franchirii. G. (1997). Proc. .\’of/. .4cvtl. Sci. U S A 94, 1:3897-13902. Thierfdtler, \V. E.. van Deurseii. J. M., l’ainamoto. K,, Tripp, H. A,, Sarawar, S.K.. Carson, R. T.. Sangster, M. Y., \’iguali, D. A. A , . I>olre%. P. C.. Grosveld, G.C.. and Ihle, J. N. (1996). Nntrirc 382, 171-174. Turner, B, M. (199:3). Cell 75, ,543, Udy, (;. B., Towers. K. P., Snell. R. C., \b’ilkins. R. J., Park, S.-II., Rain, P. A,. M’a?cman, D. J.. and Dave?, H . CV. ( 1907). Yroc. NotI. Actill. Sci. U S A 94, 7239-7244. \’ella, A,, Tcague, T. K..Ihle, J., Kapplrr. I., and Marrack. I‘. (1997)./. Eq,. Med. 186, 32S-:3:3O \’inkciiic-ier, U..<;ohen. S. L., Moarefi, I.. Chait, 8 . T., Kuriytn, J., a i d Darnell, J. E., Jr. (1996).ElZlBO /. 15, 5616-5626. \%keiiieier. U., Moarefi. I., Dariirll. 1. E., Jr,, and Kuriyan, J. (1998). Sciettcr 279, 104810,52,
162
T I M O T H Y IIOEY A N D MI(;IIAEI. J GRUSBY
Wakao, H., Gouillei~x,F., and Gronctr, B. (1994).E M B O J . 13, 2182-2191. Wang, D., Stravopodis, D.. T e g l ~ i dS., , Kitazawa, J., a i d Ihle, J. N . (1996).h'fol. Crd. B i d . 16, 6141-6148. Wen, Z., Zhong, Z., and Darnell, J. E., Jr. (1995). Cell 82, 241-250. XU. X., S1in, Y.-L.. d Hoey, T. (1996). S C ~ C T I273, ~ W 794-797. Yamainoto, K., Quelle, F. W., Thierfcltler, W. E., Kreider, B. L., Gilbert, D. J., Jenkins, N. A,, Copeland, N. G . , Silv~nnoinen,O., and lhle, J. N. (1994).Mol. Cell. B i d . 14,43424349. Yan, K., Sniall, S., Desplan, C., Dearolf, C. R., arid Darnell, J . E.,Jr. (1996).Cell 84,421-430. Ye, B. H., Cattoretti, G . , Shen, Q.. Zhang, J., Hawe, N., de Lllaard, R., Leung, C., NouriShirazi, M., Orazi, A., Changanti, R. S. K., Kothnian, P., Stall, A. M., Pandolfi, P.-P., and Dalla-Favera, H. (1997). Nutiire Gmet. 16, 161-170. Yosliirnura, A,, Ohkufto, T., Kiguchi, T., Jenkins. N. A,, Gilbert, D. J., Copelnnd, N. C;., Hara, T.. and Miyajima, A. (1995). E M B O ] . 14, 2816-2826. Yu. C.-L., Meyer, D. J., Cainpbell, G. S., Larner. A. C., Carter-Su, C., Schwartz, J.. and Jove, K. (1995).Science 269, 81-83, Zamorano, J.. Wang, H. Y., Wang, L.-M., Pierce, J. M.,and Keegan, A . D. (1996). 1. I t w w n o l . 157, 4926-4934. Zhang, J. J., Vinkeineier, U., Gu, W., Cliakrawrti, D., Horvath, C. M., and Darnell. J. E., Jr. (1996a).Proc. Nntl. Acnd. Sci. USA 93, 15092-15096. Zhang, O., Nowak, I., Vonderheid, E. C., Hook, A. 14.. Kadin, M. E., Nowell. P. C., Shaw, L. M., and Wasik, M. A. (1996b). Proc. Nntl. Accid. Sci. USA 93, 9148-9153. Zhang, X., Blenis, J., Li, H.-C., Schindler, C., and Chen-Kiang, S. (199s). Science 267, 19901994. Zliang, Z., Jones, S., Hagood, J. S., Fuentrs, N . L,and Fuller. G. M . (1997).J.Biol. Chem. 272, 30607-30610. Zliong, Z., Wen, Z., arid Darnell. J. E., Jr. (1994).Proc. Nntl. Acnd Sci. USA 91,4806-4810. Zong, C., Yan, R., August, A., Darnell, J. E., Jr., and Hanafusa. H. (1996). EMBO J . 15, 4515-4525. This article was accepted for pnhlication on May 21, 1998
\I>\ 4h(
F\ Ih I h l h l l l p \ O l 0C.I \ O I 71
CDSS(AP0-1/Fas)-Mediated Apoptosis: Live and let Die Peter H. Krammer TumorimmunologyPragmm, German Cancer Research Center, 0-69 120 Heide/berg, Germany
I. Introduction
Programmed cell death was discovered by C. Vogt in the middle of the 19th century (Vogt, 1842) by the moi-plmlo\g of dying cells during the metamorphosis of amphibians. In more than hundred years after the initial description, programmed cell death was rediscovered inany times. It was only several years, however, after the landinark paper by Kerr and colleagues (1972),in which the name “apoptosis” for nonnecrotic cell death was coined, that the investiqation of cell death caught the eye of inany in the scientific community. Prrsently, there is a true bull market for apoptosis and apoptosis has almost become a household word. Even nonspecialists know tliat fingers and toes do not simply grow out of the embryonic hand or foot-adage but develop out of the limb buds by also involving the death of cells, apoptosis, in the interdigital spaces. This phenomenon, among inany others, shows that apoptosis plays an essential role in development.. It is now known tliat apoptosis is involved in many physiological processes, and there is hardly any disease whose pathogenesis can be explained without apoptosis, either too much or too little of it. Thus, programmed cell death, apoptosis, is an integral part of life. There are many ways to clie. Apart from cell death by necrosis occurring on tissue injury, several distinct types of apoptosis are observed. Most investigatorswould agree that death by apoptosis is almost always characterized by loss of membrane asymmetry and exposure of pliosphatidylserine, Illebbing (zeiosis),fragmentation of the nucleus, chromatin condensation, and DNA degradation. One way to evoke this program is through death by neglect. Cells die in this nianner when they are deprived of nutrients or growth factors. A proper exaniple is a T lynipliocyte whose growth and activation are dependent 011 cytokines such as interleukin-2 ( IL-2). When T cells are deprived of IL-2, they die by apoptosis in a way analogous to how the Japanese Tamagotchis die, electronic toys deprived of their owner’s care. Mechanistically, death of cells by neglect when tortured by deprivation might even overlap with another inore active type of death triggered through the activation of death receptors expressed on the cell surface. It is this area of death receptor-triggered apoptosis that, momentarily, has ~~
163
Copvnglat 0 1999 In .Ar..alrinic I’nw ,211 nglitr 01 wproductwn 11, .LBI! b r m rewrwd
164
I’ETER H KHAMMER
gained so much attention, worldwide, and that is discussed in the following paragraphs. II. Death Receptors and Ligands
In 1989 a paper was published describing the discovery of an apoptosisinducing death receptor (Trauth et a/., 1989). This receptor was named APO-1, in the anticipation that it might only be the first of a series of apoptosis-inducing receptors to be discovered. Yonehara et a!. (1989) also described a cell surface molecule, which they named Fas, that could be triggered to induce cytotoxicity by an agonistic monoclonal antibody. Sequencing and cloning of the APO-1/Fas proteins and cDNAs, respectively, sometime later showed that APO-1 and Fas were identical (Itoh et al., 1991; Oehm et al., 1992), and the Fifth Workshop on Leukocyte Typing (Schlossman et al., 1993) suggested the name CD95 for the receptor. CD95(APO-l/Fas) belongs to the growing subfamily of death receptors, which is part of a superfamily, the tumor necrosis factor receptor (TNF-R) superfamily,characterized by two to five copies of cysteine-rich extracellular repeats. The prototype receptor of the superfamily is TNF-RI (Table I).The “professional”death receptor subfamilyof this superfamily, likely to expand in the future, is shown in Table I. These receptors have an intracellular death domain (DD; vide infra) essential for transduction of the apoptotic signal and were convincingly shown to be involved in induction of apoptosis. Death receptors such as CD95 and TNF-RI can be activated to induce apoptosis via activating agonistic antibodies. These antibodies proved to be excellent experimental tools. Under physiological conditions, however, death receptors are activated through their natural ligands (L),which have mostly been identified and which have coevolved into a death receptor corresponding death ligand family, the TNF family (Table 11). Except for LTa the death ligands are type I1 transmembrane proteins from which a soluble form can be generated by the activity of proteases. Thus, proteases have been identified for CD95L (Tanaka et al., 1997; Mariani et al., 1996), TRAIL (Mariani et al., 1998), and TNF (TACE; Black et nl., 1997; Moss et nl., 1997). When the crystal structures of TNFa and TNFP alone or TNFP cornplexed to the extracellular domain of TNF-RI were analyzed, a triineric structure in solution was found (Eck et al., 1989; Jones et al., 1992; Eck et al., 1992; Banner et al., 1993). Based on these data, a trimer structure was also modeled for CD95/CD95L (Bajorath and Aruffo, 1997). These data suggest that trimerization of the death receptors by trimerized ligands may be essential for triggering. Normally a ligand binds to its special receptor. For TRAIL, however, four receptors have been published so far (see Table I ) . For TNF receptors
165 TABLE 1 TNF-R SUI’IIIFAMILY
Receptor
ApoptosisInducing Activity
CD9S (APO-I/Fas) DRmRAIL-R 1 DRt5/TRAI L-H2/THICK2
+ + +
TRAI l,-R:3/l)cR l n R i D
-
-
-
OPG DR3/wsl- I/APO-YI’RAMP/LARD
+
CAKl TN F-receptor (TNF-R 1)/CD120a
+ +
TNF-RWCD 12Ob Lynipliotoxin /3 receptor (LTP-Rj CD40 CD30 CD27 GITR 4-1BB HVE M/ATAWR2
ox-40 Low-;iffinit). nerve growth factor receptor ”
+ + + + + n.tl.“
n.d. l1.d. l1.d
+
Reference Itoh et nl. ( 1991): Oehin c’t (11. (1992) Pan Pt (11. (199721) \Valczak et nl. (1997): Pan et d. (199711): Sheritlan et nl. (1997j; Screaton et 01. (1997a); Chaudliary et nl. (1997);Sclineider et 01. (1997) Degli-Esposti et 01. (1997a): Pan et n[. (1997ai: Sclineider ct 01. (1997): Sheridan et 01. (1997) Degli-Esposti et nl. (199711);Marsters ct d.(1997); Mongkolsapaya et nl. (1997) Eniery c’t d . (1998) Chinnaiyin et d.(1996a); Kitson et nl. (1996): Marsters et (11. (1996); Botlmer vt nl. (1997);Screaton et a / . (l99ib) Brojatscli ct nl. (1996) he tsc h e r r f a/. (1990); Schall et nl. (1990); Smith et (21. (1990) Dembic rf 01. (1990) Baens et ul. (1993);( h w e et ol. (1994); Nakamura et nl. (199s) Stainenkobic et 01. (1989) Durkop ef (11. (1992) Caineiini ct (11. (1991) Nocentini et nl. (1997) Kwon arid Weissinan (1989) Montgomery ct nl. (1995); Hsu et al. (1997); Kwon et 01. (1997) Mallrtt (’f nl. (1990) Johnson r t (11. (1986)
Not detrnniiied
the situation is almost equally complex. The soluble forms of‘TNFa and LTa bind to both TNF receptors; the soluble TNFa, however, shows a higher affinity to TNF-RI than the membrane form of TNFa to TNFRII. Furthermore, LTa, in combination with transinernbrane LTP, binds as a heterotrimer to the LTP-R (Grell et al., 1995; Crowe et al., 1994).
166
PETER H. YKAMMER
TABLE I1 TNF FAMILY Ligaiid CD95L TRAIL (APO-2L) TRANCE ( R A N K U OPGL) TNFa
Receptor
Reference
0 9 5 TRAIL-R14 OPG OPG
Snda et al. (1993) Wiley et nl. (1995); Pitti et 01. (1996) Yasuda et 01. (1998); Anderson et al. (1997); Lacey et a[. (1998) Pennica et ol. (1984);Shirai et a[. (1985); Wang et al. (1985) Gray et nl. (1984) Browning et 01. (1993) Graf et (11. (1992) Smith ct al. (1993) Goodw
TNF-RI, TNF-RII
Lyniphotoxin a (LTa) TNF-RI, TNF-RII Lymphotoxiii fi (LTfi) LTP-R CD40L (TRAP/gp39) CD40 CD30 CD30L CD27L (CD70) CD27 4-1BB 4-1BBL OX-40 OX-40L (gp34) NGF NGF-R1
111. The CD95/CD95L System
CD95 is a widely expressed glycosylated cell surface molecule of approximately 45 to Fj2 kDa (335 amino acids), which was named APO-1 (for apoptosis 1) (Trauth et al., 1989). It is a type I transmembrane receptor and can also occur in soluble form (Oehm et al., 1992; Itoh et al., 1991; Cheng et al., 1994).The soluble form of CD95 is generated by differential splicing with the transmembrane part being spliced out. The human gene for CD95, APT, was localized to chromosome 10q23 and the mouse gene to chromosome 19 (Lichter et al., 1992; Watanabe-Fukunaga et al., 1992). Expression of the CD95 gene and cell surface protein is enhanced by interferon (1FN)-y and TNF and by the activation of lymphocytes (Klas et al., 1993; Moller et al., 1994; Leithiiuser et al., 1993). Apoptosis via CD95 can be triggered by agonistic antibodies and by the natural ligand of the receptor, CD95L, expressed in a more restricted way than CD95. CD95L was cloned from the cDNA of a killer cell (PCGO-dlOS) and was shown to be a TNF-related type I1 transmembrane molecule (Suda et nl., 1993).The mouse and human CD95L genes were mapped to chromosome 1 (Takehashi et al., 1994a,b). Killer cells expressing CD95L were shown to kill target cells in a Ca'+-independent fashion via CD95-CD95L interaction (Anel et al., 1995).In addition, human CD95L overexpressed in COS cells was found in the supernatant and induced apoptosis in a soluble form (Tanaka et al., 1995). Soluble CD95L is found as a trimer and is generated from the transmembrane form by the activity of a metalloprotease (Mariani et al., 1995; Yagita et al., 1995; Tanaka et al., 1996). Several papers report
CIlH5( APO- I/F.is i-h.1EDIATED APOPTOSIS
167
activity of the sohlble ligand, whereas other papers show the opposite, depending on the target cells used (Suda ct al., 1997). IV. Gene Defects in the CD95/CD95L System
Several mouse mutations have been identified that cause similar, complex disorders of the immune system, manifested as lymphadenopathy and autoimmunity. One is the lpr (lyinphoproliferation) mutation, a recessive mutation causing lymphadenopath>rand autoimmune disease similar to the human disese systemic lupus erythematosus (SLE). The mutations l p + g (allelic to lpr) and gld (generalized lymphoproliferative disease) cause very similar diseases. In all three cases, aberrant T cells accumulate; these T cells are Thy-1 positive, CD4 and CD8 negative (“double-negative” cells), polyclonal (Y and j3 T-cell receptor positive, and B220 positive. In lpr mice, insertion of a retroviral early transposable element (ETn) into intron 2 of the CD95 gene causes a splicing defect, premature termination, and a greatly reduced expression of CD95 mRNA. In 1pTLX mice a point mutation (isoleucine to asparagine) in the intracellular “death domain” of CD95 abolishes transinision of the apoptotic signal. In gld mice a point mutation in the C terminus of CD95L impairs its ability to interact successfully with CD95 to cause apoptosis. Thus, a failure of apoptosis accounts for the complex immune disorder in lpr and gld mutant mice (for reviews, see Krammer et nl., 1994a,b; Nagata and Golstein, 1995). In humans a similar disease with a dysfunction of the CD95/CD95L system has been reported (Fisher ct al., 1995: Rieux-Laucat et al., 1995). Children with this “autoimrnune lymphoproliferative syndrome” (ALPS) show massive, nonmalignant lymphadenopathy, an altered and enlarged T-cell population, and a massive autoimmune disorder. Many of these children show a crippling mutation in the death domain (DD) but are heterozygous for this defect. Because the parents are not sick, whereas the children are, a secondary as yet unknown defect must exist that is responsible for the appearance of the symptoms. The pathology of Ipr and glrl mutant mice is characterized by the accumulation of aberrant T cells and autoantibody production by B cells. Autoantibody production is the direct result of a defect in the B cells themselves. Transplantation experiments in which T cells could “help” normal or lpr mutant B cells led to the exclusive production of lpr autoantibodles. Data imply that the lpr defect affects B cells directly. Breeding experiments showed that the development of the complete lpr pathology is strongly dependent on the genetic background of the lpr mutation. In some cases, autoimmunity can occur in the absence of lymphadenopathy, which reinforces the view that autoantibody production is not a consequence of excess
168
PETER F I . KRAMMER
aberrant helper T cells. Thus, autoinirnunity and lymphadenopathy are distinct aspects of the lpr pathology, due to B-cell and T-cell defects, respectively (Krammer et nl., 1994a,b). The finding that Ipr and gld mutations are defects in the CD95/CD95L system has helped greatly in determining the physiological role of CD95mediated apoptosis in the immune system. In the normal lymphoid system, apoptosis occurs in primary lymphoid organs, such as the bone marrow, liver, and thymus. There it is used to eliminate useless precursor cells with nonrearranged or aberrantly rearranged nonfunctional antigen receptors. In addition, apoptosis is essential for the deletion of autoreactive T cells in the thymus. This mechanism, therefore, is the basis of central selftolerance. In lymph nodes and the spleen, apoptosis causes deletion in T and B cells. Peripheral deletion by apoptosis is a second line where the immune system establishes self-tolerance and where it downregulates an excessive immune response. Only lymphocytes that survive this process may determine immunological memory. The mechanism of peripheral deletion was previously unknown and has now become elucidated, at least in part. Data show that the CD95/CD95L system contributes substantially to the elimination of peripheral lymphocytes. V. Role of the CD95/CD951 System in Deletion of Peripheral T Cells
To investigate the role of CD95/CD95L in TCR-mediated apoptosis, malignant Jurkat T cells as a model were triggered via the TCR in the presence of blocking anti-CD95 antibodies. These reagents prevented antiCD3-induced apoptosis. Dexamethasone-induced apoptosis, however, was not prevented. The same findings were obtained with activated human Tcell clones and activated peripheral human CD4 positive T cells. Taken together, these results and those by others in vitro and in vivo suggest that CD95/CD95L are involved specificallyin TCR-triggered apoptosis. T-cell apoptosis occurs as “fratricide” by interaction of the membranebound receptor with the membrane-bound ligand on neighboring T cells that kill each other. TCR-triggered CD95-mediated apoptosis is also found in single Jurkat T cells. A single TCR-activated T cell in the absence of costimulation may autonomously decide to die by apoptosis, employing, at least in part, the CD95 pathway. These results suggest a minimal model in which TCR-induced death in activated T cells involves CD95/CD95Lmediated suicide. Collectively, TCR-induced CD95-mediated apoptosis may occur in several forms: fratricide, paracrine death, and autocrine suicide (Dhein et al., 1995). The CD95/CD95L system, however, is not the only death system that plays a role in the deletion of peripheral T
cells. Data from Zheng et nl. (1995) suggest that late after triggering of the TCK in vitro TNF-RII and TNF dominate over the CD95/CD95L system. Elucidation of the TClUCD95RNF-R death mechanism has shed new light on peripheral T-cell tolerance by deletion, on suppression of the immune response, and on the development of memory in the surviving T lymphocyte pool. It can be predicted that as these phenomena are studied in greater depth, other less well-5tudied death system may coinpleinent the ones already invoked in these processes. In addition, it is conceivable, and in fact recent data would point in that direction, that the CD9S/CD95L system has an essential role for B-cell deletion similar to the one described for T cells. Here, however, B-cell deletion may be brought about by CD95L positive T cells that kill CD9S positive B cells (Rothstein et nl., 1995). This review has already described some of the basic features associated with CD95KD95L in t1i-e immune system, but the CD95/CD9FjL system also has an important role in another organ: the liver. VI. Role of the CD95/CD95L System in liver Homeostasis
The liver belongs to the many nonlymphoid tissues that express CD95. CD95 is expressed in the developing (French, 1996) and the mature (Leithhser, 1993) liver. Primary hepatocytes from mice (Ni, 1994) and humans (Galle, 1995) were demonstrated to be sensitive toward CD95mediated apoptosis in uitro. A physiological role of CD9ij in maintaining liver homeostasis has been suggested ;is mice deficient in CD95 develop increased cellularity and substantial liver hyperplasia (Adachi, 1995). CD9SKD95L might have a less prominent role in apoptotic processes in other organs. This role may not be revealed easily in CD95 mutant or knockout mice due to adaptive differentiation and may show better in conditional organ-specific knockout mice. VII. Signal Transduction of CD95-Mediated Apoptosis
Signal transduction of CD9-5has only been clarified recently, at least in part. Thus, oligoinerization of CD95 was shown to be a prerequisite for the transduction of the apoptotic signal. Several subclasses of the agonistic apoptosis-inducing monoclonal antibody. anti-APO-1, have been used to induce the CD9Fj death signal. The best death-inducing isotype, clearly, was the subtype IgG 3. This IgC subtype has the property of self-aggregationvia its Fc part. Other IgG subclasses of the antibody showed a differential cytotoxic activity in the declining order IgG 3 >> IgG 1> IgG 2a > IgG 21). IgG 2b did not show any activity anymore. Because all antibodies carried the same combining site specificity, the reason for this differential
170
PETER M. KKAMMER
apoptotic activity had to be the different activity of the different Fc parts. IgG 2b anti-APO-1 was inactive because it only dimerized CD95 receptors. With additional cross-linking activity the IgG 2b antibody became as active as the self-oligomerizing IgG 3 antibody. Also, apoptosis was not induced with a F(ab’)?anti-APO-1 but with F(ab’)., anti-APO-1 cross-linked with F(ab’)z anti-Ig light chain. Thus, CD95 dimers do not induce apoptosis, whereas oligomerized CD95 receptors show this activity (Dhein et al., 1992). The most probable structure to transmit an apoptotic signal is a CD95 trimer, as this structure corresponds to the predicted trimer structure of members of the TNF-R superfamily by X-ray crystallography (e.g., LTa! complexed with TNF-R1; Banner et al., 1993). VIII. The Death Domain
The intracellular part of CD95 does not possess any consensus sequence that would have predicted the use of a known signaling pathway. A deletion of 15 amino acids of the C terminus of CD95 was shown to increase CD95-mediated apoptosis. Further deletions inhibited the CD95 signal completely. When the sequence of the intracellular part of CD95 was compared with the one of TNF-RI, a homology region of 68 amino acids could be defined. Moreover, using deletion and point mutagenesis, Tartaglia et al. (1993) defined a region of TNF-RI that was essential for the cytotoxicity mediated by the receptor. This stretch was 80 amino acids long, comprises the domain described by Itoh and Nagata (1993), and was later called the death domain. The DD also contains the valine residue (ValzJ8)that is mutated in IpfK mice and which abolishes the signaling of apoptosis. Only several years later were additional DD-containing receptors isolated (DR3, DR4, and DR5). Also, further DD proteins were found, two of which bind to CD95: FADD (MORT 1; Chinnaiyan et al., 1995; Boldin et al., 1995) and RIP (Stanger et al., 1995). These molecules were all cloned in the yeast two-hybrid system with the cytoplasmic part of CD95 used as bait. The yeast two-hybrid system selects for low-affinity interactions. Therefore, some of these molecules still have to prove their importance in an in vivo system. IX. CD95 Associated Signaling Molecules
Almost all published signaling molecules found to bind to the CD95 receptor have been found by the yeast two-hybrid system with the cytoplasmic part of CD95 as bait. These proteins are shown in Table 111. The two DD proteins FADD and RIP bind to the CD95 D D directly. Overexpression of FADD and RIP causes apoptosis. All other molecules
171
Signaling Molecule
Reference
FADD (MORT11
Cliiriiiniym ct nl (1995), Boldlrl cf a1 (199s)
RIP DAXX FAFl UBC-FAP (UBC9) FAP-1 Sentnn
Stmger et (zI (1995) Ynng et nl (1997b) C l l U P t al (1995) Wnght ct nl (1996) L i t o rt (11 (1995) Okurd cf a1 (1996)
(see Table 111) do not possess a DD. Except for FAP-1, binding of all proteins to CD95 is decreased or abrogated by tlie lpeg mutation. Overexpression of DAXX (Yang et al., 1997), FAF 1, and UBC-FAP increases CD95-niediated apoptosis, whereas a blocking effect has been described for FAP-1 and Sentrin. Except for the DD at the C terminus, no other known domain has been identified for FADD. Interestingly, expression of the DD of FADD alone did not cause suicide of tlie cells whereas expression of the N terminus did. This is tlie reason why this part of tlie niolecule is called the “death effector domain” (DED). In the yeast two-hybrid system, RIP showed affinity to CD95 and to a lesser degree to TNF-€31. RIP has a domain homologous to the protein kinases and its DD alone i5 sufficient to kill a cell. DAXX was cloned recently. It increases apoptosis and activates Jim N-terminal kinases ( J N K ) . Its C-terminal DD binds to the DD of CD95 and TNF-RI. When this part of DAXX is expressed alone it inhibits both apoptosis and JNK activation. UBC-FAP is the human hornologue of “ubiquitin-conjugating enzyme 9”(UBC9) in yeast S. cemisiae. UBC9 controls the cell cycle at G2 to M and regulates the degradation of cyclins. A negative regulatory role has been suggested for tlie C terminus of CD95 because the deletion of the last 15 amino acids of CD95 increases the sensitivity toward CD95-mediated apoptosis (Itoh et nl., 1993). This region of CD95 interacts with FAP-1 (for Fas-associated phosphatase l ) ,a protein phosphatase that is capable of inhibiting CD95-mediated apoptosis. Recent reports, however, could not show the interaction of FAP-1 with CD95 in mice (Cuppen et al., 1997).In addition, deletion of the 15 carboxyterminal amino acids of mouse CD95 did not show any effect on CD95mediated apoptosis. Another molecule capable of inhibiting CD95mediated apoptosis is sentrin, which, like RIP and DAXX, also binds to
172
PETER H . KRAMMER
TNF-RI in the yeast two-hybrid system. FAP-1 showed additional homology to ubiquitin, NEDDS, and Sint3, a S. cerevisiae protein. X. Other Signaling Molecules Invoked in CD95 Signaling
Several other partly classical signaling pathways have been described that are thought to play a role in generating the CD95 death signal. Cerainide is cleaved from membrane sphingolipids by sphingomelinases (SMases). Sphingolipids are molecules that form a central part of the cell membrane. Several authors have reported that for CD95-mediated apoptosis the activities of the acid SMase (Kolesnick et al., 1994; Cifone et al., 1994), as well as of the neutral SMase (Tepper et al., 1995), are essential. In addition, tyrosine phosphorylation was described as a regulatory modification on the triggering of many receptors. Thus, the activity of protein tyrosine kinases (PTK) was also found to be essential for CD95mediated apoptosis. Furthermore, protein kinase C (PKC) activity may play a role in modulating the cytotoxic signal. Thus, PKC inhibitors such as H7 and HA 1004 sensitized cells toward CD95-mediated apoptosis, whereas activation with the phorbol ester PMA led to the development of resistance (Copeland et al., 1994). This listing should not create the impression that all molecules mentioned are of equal importance. The direct role of some of them in CD95 signaling needs to be confirmed in the in uivo situation (Table IV). XI. Proteins of the Bcl-2 Family
Bcl-2 was first discovered in the development of human follicular lymphoma (Tsujimoto et al., 1984). A t( 14 : 18) translocation of the bcl-2 gene in this tumor led to a deregulation of its expression because it became controlled by the enhancer of the Ig heavy chain gene (Graninger et al., TABLE IV OTIIER SIC:NAI.INC: MOIXCULESINVOKEDI N CD95 SIGNAI.INC: Signaling Molecule
Bcl-2 family PKC Cerainide
HCP PTK Caspases
Activity Controversial Blocker Activator/stimnlator Activator/stimulator Activator/stirnulator Activator/stimulator
1987). It soon became clear that overexpression of bcl-2 was not sufficient for the transformation of cells. In fact, other genetic alterations were also necessary (Vaux et al., 1988; Reed et al., 1988; Cook et al., 1985; Metcalf et al., 1987). To demonstrate this, elegant experiments using c-myc and bcl-2 transgenic mice were suitable whose offspring showed an increased tumor incidence (Strasser et al., 1990). Although this has been contested lately (HSUet al., 1997), Bcl-2 may heterodiinerize with other proteins in physiological systems. These inolecules interact by domains (BH1 and BH2) that have a high degree of homology (Yin et al., 1994). Together, these molecules form the Bcl-2 family (Sedlack et al., 1995). Interaction with Bax is probably the most important function of Bcl-2 (Oltvai et al., 1993, Yin et nl., 1994; Hanada et al., 1995). Bax occurs in several splice variants; Bax-a, Bax-P, and two forms of Bax-y. Only the function of Bax-a is known. Overexpression of Bax-a! in cells leads to apoptosis, and protection by the simultaneous expression of Bcl-2 is relieved (Oltwai et al., 1993). Thus, Bax-a expression is also found in tissues that show increased apoptosis (Krajewski et al., 1994). Another member of the Bcl-2 family is Bcl-x (Boise et al., 1993). Bcl-x occurs in two splice variants, which act differentially on apoptosis. The long form, Bcl-xL,and Bcl-2 both block apoptosis of cells, e.g., deprived of growth hormones. The short variant of Bcl-x, Bcl-x,, antagonizes the effect of Bcl-2. In different tissues expression of the two variants correlated with the life expectancy of the cells of these tissues (Boise et d., 1993). Bcl-x, and Bcl-2 inay exert their negative effects on different apoptotic pathways. Bcl-x, blocks apoptosis in P-lymphoblastoid cells induced by immunosuppressive agents whereas Bcl2 does not seein to protect at all (Gottschalk et al., 1994; Choi et al., 1995; Cuende et al., 1993). Also, mechanisms of in oivo resistance of T-cell lymphomas do not seein to involve Bcl-2 (Debatin and Krammer, 1995; Boise et al., 1995a). Additional molecules belonging to the Bcl-2 family are Bad, Bik, and Bak-1. Bad is a modulator of the function of Bax (Yang et al., 1995); Bik associates with the adenovirusprotein E1B 19K, Bcl-2, and Bcl-xL(Boyd et al., 1995); and Bak-1 acts like Bad and increases apoptosis (Chittenden et al., 1995; Farrow et al., 1995; Kiefer et al., 1995). In mammals, the A1 (Lin et al., 1993) and Mcl-1 genes (Kozopas et al., 1993) show homology to bcl-2. The protein BHRF was isolated froin the Epstein-Barr virus and tlie protein LMW5-H1 froin tlie African swine fever virus (Neilan et nl., 1993). Both proteins share homology with Bcl-2. BHFR blocks the apoptosis of infected cells whereas no further function is known for LMW5-H1. Another protein, BAG-1, which does not show sequence homology to the Bcl-2 family, interacts with Bcl-2, however, and increases the antiapoptotic effect of Bcl-2 when cotransfected with Bcl-2 (Takayama et al., 1995).
174
PETER 13. KRAMMER
Expression of Bcl-2 in cells protects against different apoptosis-inducing stimuli such as irradiation, DNA damage, glucocorticoids, sodium axide, Ca’+ influx, heat shock, and oxygen radicals (Vaux et al., 1988; McDonell et al., 1989; Strasser et al., 1991b; Sentman et al., 1991; Miyashita and Reed, 1992; Reed, 1994). In contrast to the earlier examples, there are cases, however, in which apoptosis cannot be blocked by Bcl-2. Such a case, for example, is the apoptosis involved in the negative selection of thymocytes (Sentman et al., 1991; Strasser et al., 1991a),cell death induced by TNF (Vanhaesebroeck et al., 1993), or lysis of target cells by cytotoxic T cells (Vauxet al., 1992a).The effect of Bcl-2 on CD95-mediated apoptosis is unclear. The described effects of Bcl-2 vary from complete (Jaattela et al., 199s) or partial inhibition (Itoh et al., 1993; Martin et al., 199Sa; Enari et al., 1995a) to the observation that bcl-2 does not protect cells from CD95-mediated apoptosis at all (Strasser et al., 1995; Memon et al., 1995). These controversial data have been resolved by experiments described by Scaffidi et al. (1998) and will be discussed further after a more complete discussion of the CD95-signaling pathway. XII. The Death-Inducing Signaling Complex (DISC)
The three-dimensional structure of the CD95 DD has been determined by nuclear magnetic resonance spectroscopy. It consists of six antiparallel, amphipathic a helices arranged in a novel fold, which is likely to be important for binding intracellular signaling molecules (Huanget al., 1996). Using a different method than the yeast two-hybrid system, a complex of proteins was identified that associated only with stimulated CD95 (Kischkel et al., 1995).Treatment of CD95-positive cells with the agonistic mAb anti-APO-1 and subsequent immunoprecipitation of CD95(APO-l/Fas) with protein A-Sepharose resulted in the identification of four ytotoxicitydependent APO-1-associated proteins (CAP1-4). The CAP2-4 proteins could be revealed on two-dimensional IEF/SDS gels within seconds after CD95 triggering. Together with CD95 these proteins formed a complex called the death-inducing signaling complex (DISC). Using a specific rabbit antiserum7CAPi and CAP2 couldbe identified as two different serinephosphorylated species of FADD and demonstrated that FADD bound to CD95 in a stimulation-dependent fashion in tjivo. While the binding of FADD to CD95 is stimulation dependent the phosphorylation of FADD is not. Therefore, the role of FADD phosphorylation still remains a mystery. It is not excluded, however, that it is important to give the entire DISC a structural conformation in order to function properly. When FADD-DN (the C-terminal DD-containing part) was stably transfected into cells the DISC also formed. However. FADD-DN was recruited
CD9551APO- l/F.is)-ME UIATEU APOPTOSIS
175
to CD95 instead of the endogenous FADD. Analysis on two-dimensional gels revealed that CAP3 and CAP4 were not part of the DISC anymore ). proteins were therefore prime candidates (Chinnaiyan et al., 1 9 9 6 ~ These for the signaling molecules. Using nanoelectrospray tandem mass spectrometry, sequence information of CAP3 and CAP4 was obtained that led to the retrieval of a full-length clone from a cDNA data base that contained all sequenced peptides (Muzio et al., 1996). This protein contained two DED at its N terminus and showed tlie typical domain structure of an ICE-like protease at its C terminus. It was therefore termed FLICE (for FADD-like ICE). The same molecule was also found by two other groups and was named MACH and MchS (Boldin et nl., 1996; Fernandes-Alnemri et nl., 1996). FLICE belongs to cysteine proteases, which are now called caspases (Alnemri et nl., 1996),and got the name caspase-8. Thus, identification of caspase-8 and its location in the DISC on CD95 triggering connected two levels in the CD95 apoptosis pathway; the CD95 receptor level with tlie intracellular level of the apoptosis executioners, the caspases. The finding that caspase-8 was identified as part of the in civo CD95 DISC suggested that its activation occurred at tlie DISC level. It has been shown that tlie entire cytoplasmic caspase-8 is converted into active caspase-8 subunits at the DISC (Medema et nl., 1997a). After stimulation, FADD and caspase-8 are recruited to CD95 within seconds after receptor engagement. Although this has not yet been shown, direct binding of caspase-8 may cause a structural change in tlie molecule that results in autoproteolytic activation on which the active subunits p10 and p18 are released into the cytoplasm. Part of caspase-8 prodomain stays bound to the DISC. Anti-caspase-8 monoclonal antibodies have been generated that show that from tlie eight published caspase-8 isoforms, only two, caspase8/a and 8/b, were expressed predominantly on the protein level in 13 different cell lines tested (Scaffidi et al., 1997). Both isoforms are recruited to the DISC and are processed with similar kinetics. Recoinbinant caspase8 lacking tlie prodomain has been reported to cleave caspase-8 in uitro, suggesting autocatalytic cleavage of caspase-8 at tlie DISC and an amplification step with caspase-8 at the top of a caspase cascade (Srinivasula et al., 1996a,b; Muzio et al., 1997). However, Medema et al. (1997a) used the DISC and could not confirm this observation. Recombinant caspase8 lacking the prodomain might therefore display a different substrate specificity compared to full-length caspase-8 in ciwo. Most investigators believe that active caspase-8 cleaves various cellular death substrates, including other caspases, such as caspase-3, thus initiating the execution of apoptosis. The caspase-8 gene has been mapped to chromosome 2q33 to 35 (Kisclikel et nl., 1998; Rasper et al., 1998). Furthermore, caspase-8 has been shown to be cleaved and activated by granzyine B in perforin killing
176
PETER 11. KRAMMER
by cytotoxic T lymphocytes. Thus, capsase-8 is the primary, although not the only, initiator of a protease cascade in this type of T killer cell activity (Medema et al., 1997b). XIII. Downstream Caspases in CD95 Death Receptor Signaling
Caspase-8 belongs to a growing family of cysteine proteases (Ahemri et al., 1996). Caspases are crucial for the execution of apoptosis, as is CED-3 during nematode development (Yuan et al., 1993). So far, 11 known caspases can be subdivided into three families based on sequence homology. The ICE-like protease family includes ICE (caspase-1) (Thornberry et al., 1992; Ceretti et al., 1992), WICH-2ACE-relII (caspase-4) (Faucheu et al., 1995; Munday et al., 1995; Kamens et al., 1995), TY/ICEre1111 (caspase-5) (Munday et al., 1995; Faucheu et al., 1996), and ICH3 (caspase-11) (Wanget al., 1996).The CED-3-like family includes CPP3W YAMNapopain (caspase-3) ( Fernandes-Alnemri et al., 1994; Tewari et al., 1995b; Nicholson et al., 1995), Mch2 (caspase-6) ( Fernandes-Alnemri et al., 1995a), Mch3/ICE-LAP3/CMH-l (caspase-7) ( Fernandes-Alnemri et al., 1995b; Lippke et al., 1996; Duan et al., 1996a), caspase-8 (FLICE/ MACH/Mch5) (Muzio et al., 1996; Boldin et al., 1996; Fernandes-Alnemri et al., 1996), MchG/ICE-LAPG (caspase-9) (Duan et al., 1996b; Srinivasula et al., 1996b),and Mch4/FLICE2 (caspase-10) (Fernandes-Alnemri et al., 1996; Vincenz and Dixit, 1997). The third subfamily consists of Nedd2/ ICH-1 (caspase-2) only (Wang et al., 1994; Kumar et al., 1994) (Fig. 1). Caspases are synthesized as proenzymes that are activated by proteolytic cleavage. The active enzyme is a heterotetrameric complex of two large subunits containing the active site and two small subunits, as deduced from the crystal structure of caspase-1 and caspase-3 (Wilson et al., 1994; Walker et al., 1994; Mitt1 et al., 1997). Activation of caspases has been reported for a variety of apoptotic stimuli, including signal transduction through the “death receptors.” Thus, crmA, a viral inhibitor of caspase-1 and other caspases, blocks apoptosis induced by these receptors (Tewari et al., 1995b; Enari et al., 1995b; Los et al., 1995; Miura et al., 1995). As discussed earlier, either caspase-8 or a similar caspase, such as caspase-10, may be involved in a similar fashion as in CD95 signaling in the signal transduction of the other death receptors. Overexpression of caspases in mammalian or insect cells induces apoptosis, but only caspase-3, -6, -7, and -8 were shown to be activated in vivo on triggering of the death receptors (Medema et al., 1997a; Scaffidi et al., 1997; Orth et al., 1996b; Duan et al., 1996a,b; Schlegel et al., 1996; Faleiro et al., 1997). In vitro caspase-8 was shown to cleave caspase-3, -4, -7, -9, and -10 directly. Caspase-2 and -6 were cleaved indirectly by other caspases
177 Caspase-5 (ICEreplIl, TY)
ICE-
subfamily
Caspase-4 (TX, ICH-2, ICE,pII) Caspase-1 (ICE) Caspase-7 (Mch3, ICE-LAPS, CMH-1) Caspase-3 (CPP32, Yama, apopain) Caspase-6 (Mch2) ^^
I,
,”* _,_,__’
Caspase-8 (FLICE, MACH, Mch5) Caspase-10 (Mch4)
I”-.{-
CED-3sub f am I I y
.yLI.-LIII..--x
Caspase-2 (ICH-1) Caspase-9 (ICE-LAPG, Mch6)
Fic:. 1 . The caspasr f’aniilyof ICE/CED-:3-hoiiiologouscystein proteases. Synonyms are given in parentheses. Phylogenetic relationships were generated by PILEUP algorithms of the “Wisconsin COG sequence analysis package.” K n o w i caspase family itieml)ers arc Caeiiorlubrlitis & ~ O J J . S CED-5 and 10 enzynies of hiimaii origin. Similar niolecules have also Ixen identified in other species (e.g., iiiouse, guinea pig. antl Uro.soplda rnclariogcister, not shown). Based on the sequences of ICE and CED-3, the proteins can he divided into hvo subgroups. C. rlqyms CED-3 is closely reliited to huiiian caspise-3. References for caspase-1: Thoriiberr>i et al. (1992); Ceretti ct 01. (1992): caspase-2: Wang et al. (1994); Kuinar et (11. (1994); caspase-3: Fernantles-Alnemri ct (11. (1994); Tewari et al., 199511; Nicholson ~i 01. (1995): caspase-3: Fauclieii ef al. (1995); Munday et 01. (1995); Kamens at rrl. (1995);caspase-5: Munday at a!. (1995); Faucheu c>t a!. (1996);caspase-6: FernandesAlneinri c’t rrl. (lW5a); caspase-7: Feriiaiitles-Alneiiiri et rrl. (1995h); Lippke et nl. (1996); Duan r f 01. (1996a):caspase-8: Muzio ct 01. (l99G):Boldin ct 01. (1996): Fernandes-Aliiemri et (11. (1996); caspise-9: Duan et nl. (199Fb): Srinivasula rt rrl. (1996h); antl caspase-10: Fernandez-Alneniri ei ril. (l99G); ~’iticoiizaiid Dhit (1997).
and activated by caspase-8 in cellular extracts (Muzio et al., 1997). Therefore, caspase-8 is able to start a cascade of caspases. The order of caspases in this cascade is not yet clear. Orth et al. (199611)place caspase-6 upstream of caspase-3 and -7, whereas it has also been demonstrated that activated caspase-3 can cleave and activate caspase-6, -7, and -9 ( Fernandes-Alnemri et al., 1995a,b; Fernandes-Alnemri et al., 1996; Srinivasula et al., 1996b). There is no report so far, however, that demonstrates that a single caspase is crucial for apoptosis signaling by death receptors. Caspase-1 was proposed to be a key molecule in CD95-mediated apoptosis, and thymocytes
178
PETER €1. KRAMMER
from caspase-I-’- mice were reported to be resistant to CD95-induced cell death (Kuida et al., 1995). However, others did not find an impairment of apoptosis in c a s p a s e - P mice (Li et al., 1995b; Smith et al., 1997) or failed to demonstrate activation of caspase-1 on CD95 triggering (Muzio et al., 1997).Therefore, it is unlikely that caspase-1 plays a role in apoptosis signaling through death receptors. Alternatively, another caspase-l-like caspase can substitute for its function in different cellular contexts. Similarly, mice deficient of caspase-3 showed an alteration of brain development and thymocytes were not affected at all (Kuida et al., 1996). However, the activity of caspase-3 might be redundant with other caspases. Therefore, the knockout technology may not always be suitable in investigating the role of a single caspase in apoptosis signaling. Different splice variants of numerous caspases have been reported (Wang et al., 1994; Alnemri et al., 1995; Fernandes-Alnemri et al., 1996; Fernandes-Alneniri et al., 1995; Boldin et al., 1996; Wang et al., 1996c; Vincenz and Dixit, 1997) and shown to function either as activators or as inhibitors of caspase activation. Some ofthem may also represent nonfunctional proteases. Most of the reported splice variants were described on the mRNA level, and the number of isoforms expressed as proteins is clearly more limited (Scaffidi et al., 1997). The activity of caspases characterizes the execution phase of apoptosis. Therefore, the search for substrates cleaved by caspases during apoptosis should provide insight into the more downstream events in apoptosis signaling. Several of these so-called “death substrates” are known so far. First, caspases can cleave and activate other caspases as described earlier. Second, molecules involved in DNA repair, such as poly(ADP-ribose) polymerase (PARP) or the catalytic subunit of the DNA-dependent protein kinase (DNA-PK),were described to be cleaved and thereby inactivated by downstream caspase-3-like caspases (Lazebnik et nl., 1994; Gu et al., 1995a; Casciola-Rosen et al., 1995; Song et al., 1996b).Third, the ribonucleoproteins U1-70kDa, C1 and C2, components responsible for the splicing reaction of precursor mRNA, are inactivated by cleavage through caspases (Casciola-Rosenet al., 1994; Tewari et al., 1995a;Waterhouse et al., 1996). Therefore, caspases may inactivate cellular processes that prevent apoptosis from proceeding. Another set of “death substrates” contains molecules that are involved in the regulation of other signaling processes. One example is the sterol regulatory element-binding proteins SKEBP-1 and SREBP-2 that are cleaved by caspase-3 or -7 (Wang et al., 1995; Pai et al., 1996). The delta isoforin of protein kinase C (PKCG) is also a target for caspases during CD95- or TNF-R-mediated apoptosis. Cleavage activates PKCS, which is associated with chromatin condensation and nuclear fragmentation
(Ghayur et al., 1996; Emoto et al., 1995). Another example is a GDP dissociation inhibitor for the Ras-related Rho family GTPases, D4-GDI cleaved and inactivated during CD95-mediated apoptosis (Na et al., 1996: Danley et al., 1996). This leads to defective Rho GTPase regulation. The PISTLRE kinase, a member of a superfamily of cdc2-like kinases, is also cleaved by caspases during CD95-induced apoptosis. This is accompanied by increased activity of this lanase and may lead to apoptosis (Bunnell et al., 1990; Lahti et al., 1995; Beyaert ct ~ l . 1997). , Cytosolic phospholipase A? (PLA?)was also demonstrated to be a target of caspases during TNFinduced apoptosis, leading to its activation. In addition, the inhibition of PLA? was shown to partially inhibit TNF-induced apoptosis (Wissing et al., 1997). Therefore, the cleavage of these signaling molecules may be involved in the downstream execution of apoptosis. However, the mechanism by which these signaling nioleciiles can transduce the apoptotic signal remains elusive. Another set of death substrates are structural proteins of the cell. Thus, larnin cleavage by caspase-6 may occur during TNF-R- and CD9Fi-induced apoptosis (Orth et al., 1996a; Neamati et al., 1995;Zhivotovsky et al., 1997). Similarly, a-fodrin is cleaved during apoptosis induced by CD95 or by TNF-R1 triggering (Cryns et a/., 1996; Martin et al., 1995a). However, the caspase inhibitor DEVD protects cells from CD95-induced apoptosis but does not prevent fodrin proteolysis. This indicates that cleavage of this protein can be uncoupled from apoptotic cell death (Cryns et al., 1996). Even actin was found to be cleaved proteolytically by caspase-3 on triggering with different apoptotic stimuli. However, contradicting reports exist on the cleavage of actin in oivo during CD95-mediated apoptosis (Mashinia et a!., 1995; Kayalar et al., 1996: Chen et al., 1996: Song et al., 1997; Mashiina et al., 1997). Finally, Gas% a component of the microfilament system, was reported to be cleaved on induction of apoptosis (Brancolini et nl., 1995).Therefore, cleavage of these structural proteins may account for some of the massive morphological changes, such as lnelnbrane blebbing, nuclear fragmentation, and the formation of apoptotic bodies during apoptosis. Some oncoproteins, such as HI) and MDM-2, were found to be cleaved and inactivated by caspases during apoptosis (Bing et al., 1996; Janicke et al., 1996; Ehrhardt et al., 1997).The functional relevance of cleaving these substrates remains unclear. Cleavage and activation of the 112l-activated lilnase PAK2 during CD95and TNF-mediated apoptosis have been reported. PAK2 was also reported to activate the Jun kinase pathway. PAK2 may therefore link the caspases and J N K activation during apoptosis signaling. Interestingly, when the activity of PAK2 is blocked by a dominant negative mutant, the form.
180
PETER 11. KRAMMEH
of apoptotic bodies during CD95-mediated apoptosis is inhibited. Nuclear apoptosis, as well as phosphatidylserine externalization, however, remained unaffected (Rude1and Bokoch, 1997).Thus, this constitutes the first exaniple of how different features of apoptosis might be discriminated at the level of caspase targets. A direct link between caspase-3 and DNA fragmentation was found by cloning a heterodiinerizing factor, DFF, activated to induce DNA fragmentation on isolated nuclei after cleavage of its 45-kDa component by caspase-3 (Liu et al., 1997). Therefore, the theoretical possibility exists of a death receptor signaling pathway involving caspases only. CD95 could activate caspase-8 that cleaves caspase-3, which in turn could activate DFF. Components of this system are probably identical to the murine ICAD/CAD system described by Enari et al. (1998) and Sakahira et al. (1998). This system is believed to function like the IKB/NF-KB system. ICAD, the cytoplasmic inhibitor of CAD, is cleaved by caspase3, and CAD is freed and by means of a nuclear translocation signal translocates to the nucleus where it acts as an endonuclease and cleaves the DNA. Caspase activation does not necessarily lead to apoptosis, as reports find transient activation of caspase-3 during T-cell stimulation with PHA that is not linked to apoptosis (Miossecet al., 1997).Similarly, IL-lP maturation and cleavage of interferon-y inducing factor (IGIF) by caspase-1 are observed during a normal immune response without apoptosis of the secreting cell (Gu et al., 1997; Ghayur et al., 1997; Thornberry et al., 1992; Cerretti et al., 1992). Taken together, the number of caspase substrates should be increasing in the future, and further studies are necessary to unravel the caspase cascade induced by the different death receptors and to identify crucial targets for caspases that establish the link between caspase activation and more downstream events in apoptosis. XIV. Type I and Type II Cells
In the previous discussion regarding the CD95 signaling pathway, a number of controversies are apparent concerning DISC formation, ceramide involvement, Bcl-2 inhibition, DD associating molecules, and so on. Most of these controversies have not yet been resolved, but data by Scaffidi et al. (1998) may provide more clarity to the picture. Two cell types have been identified that each almost exclusively use one of two different CD95 signaling pathways. In both type I and type I1 cells, mitochondria were activated equally on CD95 triggering, and all mitochondria1 apoptogenic activities were blocked by Bcl-2 overexpression. In type I cells, the induction of apoptosis was accompanied by the activation of large amounts of caspase-8 by the DISC, followed by the rapid cleavage
of caspase-3 prior to the loss of mitochondria1 transmembrane potential (A?,,,). In contrast, in type I1 cells, DISC formation was reduced strongly and activation of caspase-8 and caspase-3 occurred following the loss of AT,,,.In type I1 but not in type I cells, Bcl-2 overexpression blocked caspase-8 and caspase-3 activation as well as apoptosis. Overexpression of caspase-3 in the caspase-3 negative cell line MCF7-Fas, which is normally resistant to CD95-mediated apoptosis by overexpression of Bcl-x,,, converted these cells into true type I cells in which apoptosis was not inhibited by Bcl-xl,any longer. Thus, CD95-mediated apoptosis in type I1 cells and in type I cells is dependent and independent of mitochondria1 activity, respectively. It is still unclear how mitochondria are activated in both type I and type I1 cells. Nevertheless, they are only used in type I1 cells to initiate the executionary apoptosis caspase cascade. Thus, it is clear that these data establish that CD95-mediated apoptosis uses two pathways probably dependent on the quantity of caspase-8 initially activated. XV. FLIPS (FLICE Inhibitory Proteins)
An entire family of death effector doinain-containing proteins has been found in the data bases. Some of those proteins are made by class y herpes viruses such as herpes virus Saimiri (HVS), by human herpes virus 8 (HHV 8 ) , a Karposi sarcoma-associated herpes virus, and by molluscum contagiosuin (Peter et nl., 1997b). The proteins have been called V-FLIPS (for viral FLICE inhibitory proteins). V-FLIPSconsist of two death effector dom&ns,;d biochemical analy5is of v-FLIP-transfected cells showed that they bind to the CD95/FADD complex and thus inhibit the recruitment of caspase-8 and a functional DISC formation. In transfected cells, V-FLIP was capable of inhibiting apoptosis induced by several death receptors (CD95, TNF-RI, DR3, and DR4), which suggests that all ofthese receptors use similar signaling pathways (Thome et nl., 1997; Hu et al., 1997a; Bertin et nl., 1997). A human hoinologue of v-FLIP has been found and is called several names: c-FLIP/FLAME/I-FLICE/Casper/CASH/usurpin/MRIT/ CLARP (Irmler et al., 1997; Srinivasula et al., 1997; Hu et al., 1997b; Shu et al., 1997; Goltsev et nl., 1997; Rasper et nl., 1998; Inohara et al., 1997; Han et al., 1997). cFLIPs occur in two forms: a short and a long form. The short form s e e m to act like V-FLIP. The long form has a sequence similar to caspase-8 but with an inactive enzymatic site and interferes with the generation of active FLICE subunits at the receptor level. The gene of c-FLIP/usurpin has been identified. This gene is composed of 13 exons and is clustered within approxiniately 200 kB with the caspase-8 and -10 genes on huiiian chromosome 2q33 to 34 (Rasper et nl., 1998). Taken
182
PETEH H K R A M M E R
together, the function of the c-FLIPS is not yet entirely clear, as data vary and reach from inhibition to induction of apoptosis. XVI. Sensitivity and Resistance of T Lymphocytes toward CD95-Mediated Apoptosis
Peripheral T lymphocytes represent a model system in investigating the modulation of the CD95 signaling pathway. This system has been described by several groups. In vitro, the sensitivity of T lymphocytes toward TCRmediated activation induced T-cell death changes in the course of an immune response (Klas et al., 1993; Zheng et al., 1995; Owen-Schaub et al., 1992). In the early phase of the immune response, peripheral T cells are resistant to the induction of apoptosis. After the immune response, however, activated T cells have to be eliminated. This elimination occurs by the induction of activation-induced T-cell death and is probably regulated by various death systems. Activation-induced T-cell death of CD4 positive T cells seems to be primarily dependent on the CD95/CD95L system (Dhein et al., 1995; Alderson et al., 1995; Brunner et al., 1995; Ju et al., 1995; Singer et al., 1994; Van Parijs et al., 1996), whereas the elimination of CD8-positive T cells might primarily happen by the TNFR/TNF system (Zheng et al., 1995; S y b u et al., 1996; Zimmerman et al., 1996). Nevertheless, mice with a defect in TNF-RI or TNF-RII are still capable of deleting peripheral T lymphocytes (Pfeffer et al., 1993; Rothe et al., 1993; Erickson et al., 1994). As discussed earlier, a defect in CD95 expression or the lack of CD95 results in lymphadenopathy, accumulation of anomalous T cells, and formation of autoantibodies. This phenotype is seen in lpr or the CD95 knockout mouse. In humans, this defect in the CD95 system causes a similar phenotype. Therefore, resistance toward CD95-mediated apoptosis has to be regulated in a precise fashion. It could be shown that CD95 positive resistant T cells show a dysfunction of the DISC based on the lack of recruitment of caspase-8 to the DISC. Only apoptosis-sensitive T cells form a functional DISC. It is still unclear how the recruitment of caspase-8 to the DISC in resistant T cells was prevented. The quantity of caspase-8 in the cytoplasm in resistant and sensitive T cells was comparable. Hypothetically, the association of caspase-8 with a DISC could be prevented by a posttranslational modification of FADD or caspase-8 or by competition of an as yet unknown inhibitor molecule to the death effector domain of these molecules. A new molecule has been cloned that can bind to the DD of the CD95 and the TNF-RI. It was shown that this molecule (sentrin) is capable of inhibiting CD95 as well as TNF-RI-mediated apoptosis.
Sentrin does not contain a DD but binds directly to the DDs of thew receptors but not to FADD. Because the association of FADD to the DISC in resistant and sensitive T cells was unchanged, it is unlikely that the presumed resistance protein is sentrin. It is more likely that the presuined resistance protein contains a death effector domain. As discussed earlier, such molecules have been clonetl bv several groups (c-FLIP/FLAME/ I-FLICE/Casper/CASH/usurpin/MRIT/CLARP). In a publication by Irinler et a1 (1997) it was Fhown that c-FLIP is expressed in resistant T cell5 and is downregulated in T cells as they hecome sensitive to CD95mediated apoptosis. These data suggest that c-FLIP could be one of those regulator proteins. At present, work is i n progress toward finding c-FLIP in tlie DISC. This involves tlie generation of ywcific anti-c-FLIP monoclonal antibodies. Additional experiments showed that caspase-8 is not processed in resistant T cells, i.e., no active caspase-8 subunits could be detected in resistant T cells; they were only found in sensitive T cells. This supports the assuinption that the signaling pathway might be blocked directly at the level of the CD95 receptor. Together with the anomalous DISC formation in resistant T cells, it was shown that a transient upregillation of Bcl-xl correlated with CD95 resistance during T-cell activation. This result is in correlation with many previous reports, which suggest that Bcl-xLmight be a regulator of apoptosis resistance in lymphocytes. In inouse T cells, expression of Bcl-xl but not of Bcl-2 or Bax correlates with apoptosis resistance in activated T cells (Broonie et a1 , 1995). Stimulation of H cells \ia CD40 results in upregulation of Bcl-xLand in resistance toward apoptosis (Choi et d., 1995). Moreover, costimulation of apoptosis-sensitive T cells via CD28 renders these T cells resistant Again, development of resistance in these T cells is correlated with the upregulation of Bcl-xl (Boise et a1 , 1995b). In contrast to these correlative studies there are a number of publications in which no significant inhibition of the CD95 signaling pathway by Bcl-2, which is supposed to act Timilarly to Bcl-xl , was found (Meinon ct al., 1995; Chiu et a l , 1995; Gottschalk et al., 1994; Itoh ct al., 1993). In agreement with these reports, it was not possible to show direct inhibition of Bcl-xlJon DISC forination (Medema et d ,1998). Either Bcl-x, acts on caspase-8 below the DISC or T cells possess, next to the recruitment of caspase-8 to the DISC, another independent niechanism to regulate CD95 apoptosis resistance in which Bcl-xLis involved. In transfectants with BclxI , Chinnaiyan et nl. (1997) showed that Bcl-u, might bind indirectly to caspase-8, which might lead to inhibition of caspase-8 activation. In MCF’iFas cells resistant to CD95-mediated apoptosis by overexpression of BclxL, however, no association of Bcl-x, and caspase-8 could be found (Medema c,t al., 1998). Moreover, activation of caspase-8 in MCF7-Fas cells was not
184
P E T E R 13. K R A M M E R
inhibited. Because resistant T cells do not activate caspase-8 at all, a different mechanism of resistance must apply to T cells. Bcl-xl,may affect the CD95 signaling pathway, but in an as yet unknown fashion or it may regulate an as yet not fully worked out apoptosis signaling pathway initiated by the CD95 receptor in T cells. In any case, however, BcI-xL might be a second safeguard toward induction of apoptosis in T cells apart from the inhibitor activity directly at the receptor level. Several reports support the model that Bcl-x, might influence an apoptosis signaling pathway parallel to the CD95 receptor-mediated caspase cascade (Chinnaiyan et al., 1996b; Strasser et al., 199s). It has been discussed whether PTPases are involved in the regulation of CD95-mediated apoptosis resistance. Thus, the heinatopoietic cell phosphatase HCP was supposed to increase the CD9+5signal, whereas overexpression of PTPase ( FAP-1) inhibited CD95-mediated apoptosis. In the tested T cells, however, no correlation between expression of PTPases and apoptosis sensitivity was found. In addition, it is in agreement with results that tyrosine phosphorylation does not play a decisive role in the CD95 signaling pathway in T cells (Schraven und Peter, 1995). Taken together, data show that resistance and sensitivity of CD95medated apoptosis in activated T cells are regulated on different levels, at least on the level of a functional DISC and on the level of BcI-xL. These mechanisms might be very important for the development of T-cell subpopulations such as memory T cells, which might have an apoptosisresistant phenotype. XVII. The CD95 System and Chemotherapy
The folic acid antagonist aminopterin (methotrexate) was introduced into leukemia therapy by the pediatric oncologist Sidney Farber ( Farber et al., 1948) in the late 1940s. This turned treatment of some leukemias and solid tumors into effective therapy. Especially in the treatment of leukemias and pediatric solid tumors the majority of patients can be cured by therapeutic protocols that use combinations of different cytotoxic drugs given over a certain period of time. In addition to liigh-dose therapy, adjuvant chemotherapy has become an important treatment in cancers of colon and breast. However, the widespread use of cytotoxic drugs, prototypes of which were developed in the late 1960s, revealed that only few tumors are chemosensitive. The most common tumors were especially more or less resistant to chemotherapy (and irradiation). In addition, side effects of chemotherapy limit the development of mega-dose protocols for tumors resistant to conventional doses of cytotoxic drugs. Development of resistance to chemotherapy is also a major problem in chemosensitive tumors such as leukemias or sarcomas. In these tumors the majority of
relapse patients present with tumor cells more resistant to chcniotherapy than the primary cells. Anticancer drugs have primarily been found in assays based on the inhibition of proliferation and clonogenicity. Fiirthcr analysis clemonstrated that most active drugs interfered with celliilar metabolism, mitosis, a n d DNA replication. Consequently, anticancer drugs classify as DNAdamaging agents (cyclospliospliaiiiicle,cisplatin, and doxonibicine),antimetabolites (methotrexate) and 5-flnorouracil), mitotic inhibitors (vincristine), nucleotide analogs (6-mercaptopurine), or inhibitors of topoisomerases (etoposide). It was later found that most drugs used in anticancer therapy today kill target cells by induction of apoptosis (Dive et al., 1992; Gorczyca et al., 1993; Kaufinann ct a1 , 1993; Fisher, 1994). However, the niolecular mechanisms of dnig-induced apoptosis have not been defined until recently. Some drugs indiiced DNA damage sensed by 1353, suggesting that p53 may activate the apoptosis machinen. ( L o w et al., 1993; Miyishita and Heed, 1995). p53 was also shown to Le involved in various forms of apoptosis induced by cellular stress (Levine, 1997).p53 may thus represent a cellular master switch that regulates several distinct cellular responses. Evidence, however, suggests that key downstream elements of the apoptosis machinery are activated dirwtly. Thus, resistance toward chemotherapy in some cases was found to be associated with increased levels of expression of antiapoptotic molecules of tlie Bcl-2 family such as Bcl-2 and BcI-xL(Miyashita and Heed, 1993; Dole ct d., 199s; Suinantran et d , 1995; Minn et al., 1995). The levels of Bcl-2 expression, however, e.g., in lymphoid tumors, chd not correlate directly with the clinical response to chemotherapy. In contrast, the reduced expression of Bax was found to be associated with a poor outcome of therapy in breast cancer (Hermine f>tal., 1996; Coustan-Smith et d . , 1996; Krajewski et d.,1995). p53 was found to lie involved in transcriptional expression of the proapoptotic BX gene following DNA damage. Thus, soine forms of cliernotherapy-induced apoptosis seem to involve upregulation of H~LXexpression mediated by p53 (Miyashita et 01, 1994; Miyashita and Heed, 1995). A major role for the CD95 system in drug-induced apoptosis has been suggested. Cytotoxic drugs commonly used in tlie chemotherapy of leukemias strongly induce CD95L exprejsioii in CD95 positive tumor cells (Hata et ul., 199Fj; Miiller et d., 1997; Friesen ct d., 1996). CD9SL is expressed in a membrane form or secreted by the tumor cells exposed to the drug. Binding of CD95L to CD95 then triggers the apoptosis cascade in chemosensitive cells. This scenaiio is malogous to activatioii-induced death in activated T cc4s following T-cell receptor stimulation (Dhein et al., 1995; Stahnke et nl., 1997).Triggering of the CD95/CD95L system by anticancer drugs was originally discovered in leukemia cells but was also found to be
186
PETEH 11. KKAMMEH
involved in drug-induced apoptosis in other chemosensitive tumors such as hepatoblastoma, neuroblastoma and brain tumors (Hata et al., 1995; Muller et al., 1997; Micheau et al., 1997; Fulda et al., 1997a). In some tumors, treatment with cytotoxic drugs upregulates CD95L and CD95. CD9g upregulation seems to depend on functional, wild-type p53 (Munker et al., 1995; Muller et al., 1997; Fulda et al., 1997a). Regardless of the primary target of the cytotoxic drug in the tumor cell, data would suggest that activation of the key effector molecules for apoptosis, such as the ones in the caspase cascade, is crucial for the antitumor effect, and inhibition of caspase activation in tumor cells is associated with resistance against anticancer drugs (Los et al., 1997; Kaufmann et al., 1993; Bellosillo et al., 1997). It is not clear whether this applies to all drugs, e.g., apoptosis induced by glucocorticoids may use a different effector system (Geley et al., 1997). In cases of drug treatment in which CD95/CD95L is used, it is assumed that the described (see earlier discussion) CD95 signaling pathways are initiated. As described in these situations, caspase-8 plays a crucial role. The role of caspase-8 may even be essential when CD95/CD95L is not used. This is exemplified by treatment of neuroblastomas with Betulinic acid. Here, caspase-8 cleavage is clearly observed (Fulda et al., 1997b). Initially, drugs may affect diverse cellular functions such as cellular metabolism, DNA, or the mitotic apparatus. Cellular damage is sensed by p53 and/or leads to activation of a cellular stress program (Herr et al., 1997). In the following phase a relatively uniform apoptosis program of the cell is activated. Here, CD95LJCD95 interaction represents one of the important trigger mechanisms for the apoptosis cascade leading to protease activation. However, CD95/CD95L represents only a part of an amplifier machinery, as drug-induced production of other apoptosisinducing ligands (TNFa and TRAIL) has also been found (Muller et al., 1997). Alterations in each phase of drug-induced apoptosis may lead to drug resistance. Thus, in addition to established mechanisms of drug resistance, such as increased drug efflux by membrane pumps, the fiilure to activate the apoptosis program represents another mode of drug resistance in tumor cells (Friesen et al., 1997; Knipping et al., 1995). In this respect, a favorable treatment outcome is shown for patients with CD9Fj positive AML undergoing chemotherapy. This situation was compared to CD95 negative AML. Data from this study demonstrate that CD95 expression in tumors in relationship to therapy may be clinically relevant and worth investigating (Min et al., 1996). Thus, the CD95/CD95L system is important for effective chemotherapy. It also participates in reciprocal interactions of tumor cells with cells of the immune system (Debatin, 1997). Thus, CD95 positive tumor cells may become targets for killer T cells,
NK cells, or LAK cells (Micheau et nl., 1997;Yoshihiro et a / . , 1997).Factors that upregulate CD95 expression in tumor cells, such a s cytotoxic drugs or cytohnes, may therefore be involved in the induction of tumor regression by immune cells if these cells resist the attack by CD9SL, which may be produced by the tumor cell itself. These data demonstrate that the treatment of tumor cells with various chemotherapeutic drugs causes an upregulation of the CD95 death system. As more tumor types are checked, it is conceivable that other death systems, such as DR3, TNF-K 1/11, and TRAIL, may be used in a similar fashion as the CD9Fj system. However, the basic features and their impact on tumor treatment will be the same and may cause a change in paradigm in drug treatment and in the evaluation of drug sensitivity and resistance. In a successful antitumor therapy the CD95 signaling cascade, including ICE-like proteases, is switched toward sensitivity. Considerations of drug resistance in the future, therefore, will also have to incorporate testing the functionality of the CD95 and possibly other death systems. It is clear that this has far-reaching therapeutic consequences. The previously described decisive advances in the explanation of the mechanism of chemotherapy may pro\& an explanation of why chemotherapy may have a positive and a negative side. On the one hand, it helps to eliminate the tumor. On the other hand, it may provide the tumor with a lethal weapon, boost immune evasion, and make the tumor an immune privileged site. The lethal weapon is CD95L. CD9FjL can be made by tumors constitutively or its expression might be induced, e.g., a s detailed earlier, by drug treatment of the tumor. CD95L may have a deleterious effect on the immune system, particularly on the activated attacker antitumor T lymphocytes, provided they are in a CD95 apoptosis-sensitive state. The experimental background of this situation is outlined in different in vitro and in oivo model systems in several papers that may change our view of tumor/host interactions (O'Connell et nl., 1996; Hahne et nl., 1996; Strand et nl., 1996). The previous experiments and their interpret,nt'ions have been criticized, and growth of CD9SL positive tumors was observed to be blocked by infiltrating neutrophils and granulocytes (Seino et al., 1997; Arai et nl., 1997).In any case, data point toward a situation in which (1) attacker T lymphocytes, e.g., activated T (killer) cells, express two killing systems: the perforidgranzyme and the CD95/CD95L system; and (2) tumor cells express CD95L and may even have become resistant against its hlling activity (e.g., by downregulating CD95) such as in most hepatomas. Previously, the prevailing view was that the T cells recognize the tumor, do not receive sufficient costiinulatory signals, and are therefore not activated properly. This view now adds a different turn. It says that the tumor is not passive but has the weapon to fight the immune system
188
PETER H. KR-\MMER
(“strike back’) and cause the depletion of attacker lymphocytes and immune suppression. Thus, both T cells and tumor cells are equipped with a set of almost equal weapons. Several parameters may determine who wins the encounter between T cells and tumor cells. The size of the tumor-reactive T-cell pool, the quantity of CD95 and CD95L, and sensitivity and resistance toward CD95 signaling may all be important. In the worst case the tumor has become apoptosis resistant, e.g., by downregulating CD95 or by switching the signaling pathway to insensitive, expresses CD95L, and encounters activated T cells in a CD95 apoptosis-sensitive state. The net result of such an interaction could be T-cell depletion, imiiiunosuppression, and immune escape. The aim of future therapy is to cause tumor rejection by temporarily inducing T-cell resistance toward apoptosis and maintaining sensitivity in the tumor (Krammer et al., 1997, 1998). XVIII. The CD95 Death System in AIDS
In AIDS the number of T cells, e.g., in the peripheral blood, productively infected with HIV is relatively low (in the range of one in several thousand). This implies that T-cell depletion in this disease may also affect noninfected CD4 positive T cells. Depletion of T cells may occur by induction of apoptosis. This assumption raises several questions, mainly: (1) what is the (main) mechanism of apoptosis of noninfected CD4 positive T cells and (2) how is sensitivity toward apoptosis induced in noninfected CD4 positive T cells? The hypothesis that regulatory viral gene products (e.g., HIV-1 Tat) made by HIV-infected cells may penetrate HIV-noninfected cells and render these cells hypersensitive to TCR-induced CD95-mediated apoptosis was tested. HIV Tat might induce a prooxidative state in the affected cells, increase CD95L expression, and facilitate TCR-triggered CD9fj-mediated suicide. Further sensitization of the CD4 positive T cells might result from the binding of HIV a 1 2 0 to the CD4 cell surface receptor and from the cross-linking of bound gpl20 by gp120 antibodies in the patients’ sera. This hypothesis was tested in an in vitro model system using Jurkat T cells or peripheral blood lymphocytes triggered via the TCR and incubated with synthetic Tat or with Tat, gpl20, and anti-gpl20, respectively. TCR-triggered apoptosis was accelerated by the combination of these reagents. Significantly, most but not all of the apoptosis seen under these conditions could be blocked by F(ab’)zanti-CD95 and soluble CD95-Fc receptor decoys (Westendorp et al., 1995; BBuinler et al., 1996; for a review, see Kraminer et al., 1994a,b).These data reinforce the hypothesis that the depletion of noninfected CD4 positive T cells in AIDS might involve TCR-triggered CD95-mediated suicide sensitized by HIV Tat and
a 1 2 0 and anti-gpl20 antibodies in patients' sera. These findings warrant testing this hypothesis in animal model systems and in patients and, if correct, open new exciting possibilities of AIDS therapy by stabilizing the pool of CD4 positive T cells. To test further whether the CD95 system might play a role in AIDS, CD95 expression on isolated peripheral 1)lood lymphocytes was investigated. Although CD95 expression was highly variable, it was increased significantly in CD4 positive and CD8 positive T cells from HIV-1 positive children in comparison to HIV-1 negative controls (BBumler ct d . ,1996). Thus, CD95 expression on T cells ofthe CD4 positive and the CD8 positive subset is increased in HIV infection. The mechanism of this increase niay be directly related to HIV infection or stimulation ljy HIV products or may be caused by a general stiniulation of T cells in this disease. It is interesting that the increase of CD95 expression is observed in both CD4 positive and CD8 positive T cells from HIV- 1 positive children. Assuming a role of CD95 receptor and CD95L in apoptotic depletion of CD4 positive T cells in AIDS requires, thcrefore, that sensitization of the CD95 pathway is mainly directed at the CD4 positive subset of T cells. As suggested earlier, sensitization may involve stimulation of CD4 on these cells by gpl2O and anti-gpl2O antibodies. In addition, other viral gene products such as Tat may have an additional sensitizing effect. The author's data, therefore, provide an interesting potential link between CD95mediated apoptosis and T-cell depletion in AIDS. In view of the discovery of new death receptors, it is conceivable that they are also involved in CD4 positive T-cell depletion in AIDS in a similar fashion as the CD95 system. In addition to CD95-mediated apoptosis, a novel type of apoptosis induced via CXCR4 (fusiidLESTR) and CD4 in CD4 positive T cells as detailed later was found (Bemdt c>td . ,1998).As described earlier, apoptosis can be induced in CD4 positive T lyinphocytes by HIV-1 a 1 2 0 and anti-gpl20 antibodies. Because a 1 2 0 binds to T cells via CD4 and the chernokine receptor CXCR4 ( fiisidLESTR), the role of these molecules was investigated in gpl2O-induced apoptosis. A novel and rapid type of apoptosis induced by both cell surface receptors, 0 4 and CXCR4, in T cell lines, human PBL, and CD4KXCR4 transfectants was found. Significantly, apoptosis was observed exclusively in CD4 positive but not in CD8 positive T cells (Table V). Induction of the new type of cell death did not involve the signaling cascades normally initiated by CD4, CXCR4, and some of the known death receptors (p56''!', G , a , and caspase-8, respectively). The potency of this phenomenon and its specificity for CD4 positive T cells would suggest that it might play a significant role in T helper cell depletion in AIDS. On the basis of these data, the use of antichemokine
190
PETEK I I KHAMMER
TABLE V COMIJAHISON OF NOVELAPOPTOTICCELLDEATH MEDIATEDBY CD4 CD%-MEDIATED APopTosis CD95-Mediated Apoptosis
DNA degradation Kinetics (onset) Changes in FSC/SSC" Loss of niemhrane asylllllletry
Cliroinatin condensation
A !bn! 7-AAD positivity' Inhibition by zVAD-fink" Caspase-3 cleavage Caspase-8 cleavage PARP' cleavage
AND
CXCR4 WITH
CD4- or CXCR4-Mediated Apoptosis
+ + + + + + + + + +
-12 hr
-15 min
+ + + + + -
-
Foward scatter/sideward scatter. Loss of initocliondrial inembranc potcritial. ' Staining of apoptotic cells with the fluorescelit c l y ~7-airiino-actirioiiiycinD " Peptide inhibitor of caspasw ' Poly (ADP ribose) polyimrrasr. " 'I
receptor antibodies meant to prevent HIV-1 infection might be dangerous. The use of tlie natural ligand of CXCR4, SDF-la, or its derivatives, however, could be considered for therapy, as it inhibits infection as well as CXCR4-mediated apoptosis. Studying the apoptotic signaling cascade triggered by CD4 and CXCR4 might prove to be useful in tlie identification of therapeutic strategies aimed at intervening with the progressive loss of CD4 positive T cells in HIV-1-infected individuals. XIX. Further Considerations on the Role of Apoptosis in the Clinic
Apoptosis has come a long way from the description of its morphological features to a partial niolecular understanding of its signaling pathways and its physiological and pathological consequences. Although our understanding of the different routes to deaths is far from complete, the first glimpses at the different signaling pathways make it possible to predict important areas of research in the future. Apoptosis adds a new chapter in the understanding of the pathogenesis of many diseases. Generally, there are diseases with too little apoptosis and diseases with too much apoptosis. As described earlier, cancer could be looked on as a disease with too little apoptosis where the net increase of the tumor burden is the sum of an
increased growth rate and a decreased apoptotic rate. However, AIDS involves too much apoptosis in cells of the lymphoid and nonlymphoid compartment, particularly in CD4 positive T cells. These two diseases are only two examples in which apoptosis and its regulation have gone wrong. The molecular understanding procided by the elucidation of the signaling pathways of apoptosis may make it possible, in the future, to reestablish the normal level of apoptosis. With respect to the CD95 system, which may serve as a paradigm for other death systems, apoptosis could be regulated at several decisive points:
1. the receptor (structure, glycosylation, density, expression as soluble versus membrane-bound receptor), 2. the CD95-mediated pathways (DISC formation, caspase cascade, induction of apoptosis via the mitochondria1 pathway, alteration of the death substrates, stimulation of the apoptotic pathway, inhibition of the apoptotic pathway at the receptor level by, e.g., FLIPS, or further downstream by Bcl-2 or family members or by other inhibitory or stimulating proteins), 3. CD95L (structure, glycosylation, density, expression as soluble ligand versus membrane bound ligand), and 4. regulation of induction versus prevention of apoptosis in different cellular systems. It is important to increase our understanding of the molecular events that determine these particular steps and to develop methods of targeting apoptosis regulatory moleciiles to specific cells of hfferent tissues. Using such methods, the modulation of apoptosis will gain its place in the therapeutic tools used for the treatment of diseases. Alternatively, therapeutic windows for apoptosis modifiers have to be found that only affect diseased cells and leave normal cells intact.
ACKNOWLEDGMENTS I thank all my previous and present collaborators and the previous and present nrcmhers of my group. particularly M. E. Peter, F. C. Kischkel. and C. Scaffidi for their input and discussion, C. Scaffidi for reading of the manuscript, and H . Sauter for expert secretarial assistance. This work was funded by Deutsche Krehshilfe Dr. Mildred Scheel Stiftung: Gerinan-Israeli Cooperation in Cancer Research: AIDS grant German Federal Health Agency; Tumor Centre Heidell)erg/Maiinheiin: BMBF Forderschwerpunkt “Clinicalbiomedical research; AIDS F’erbund Heidelberg; Wilhelm-Sander Stiftung; and SFB Transplanation. I apologize to all my colleagues who have done excellent work in the field and whose papers have not been quoted comprehensively. It was not possible to be encyclopedic in this exponentially growing field.
192
PETER H. KRAMMER
REFERENCES Adachi, M., Suematsu, S., Kondo, T., Ogasawara, J., Tanaka, T., Yoshida, N., and Nagata, S. (1995). Targeted mutation in the Fas gene causes hyperplasia in peripheral lymphoid organs and liver. Nut. Genet. 11, 294-300. 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).Fas ligand mediates activation-induced cell death in human T lymphocytes.]. Exp. Med. 181, 71-77. Alnernri, E. S., Livingston, D. J., Nicholson, D. W., Salvesen, G., Thornberry, N. A,, Wong, W. W., and Yuan, J. (1996). Human ICEICED-3 protease nomenclature. Cell 87, 171. Anderson, D. M., Maraskovsky, E., Billingsley, W. L., Dougall, W. C., Tometsko, M. E., Roux, E. R., Teepe, M. C., DuBose, R. F., Cosnian, D., and Galibert, L. (1997). A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390, 175-179. Anel, A,, Simon, A. K., Auphan, N., Buferne, M., Boyer, C., Gohein, P., and SchmittVerhulst, A. M. (1995). Two signalling pathways can lead to Fas ligand expression in CD8+ cytotoxic T lymphocyte clones. Eur. J. Zmmunol. 25, 3381-3387. Arai, H., Gordon, D., Nabel, E. G., and Nabel, G. J. (1997). Gene transfer of Fas ligand induces tumor regression in oioo. Proc. Natl. Acad. Sci. USA 94, 13862-13867. Baens, M., Chaffanet, M., Cassiman, J. J . , van den Berghe, H., and Marynen, P. (1993). Construction and evaluation of a hncDNA library of human 12p transcribed sequences derived from a somatic cell hybrid. Genomics 16, 214-218. Baumler, C., Herr, I., Bohler, T., Krammer, P. H., and Debatin, K.-M. (1996). Increased ex vivo expression of the CD95 ligand in T cells of HIV infected children. Blood 88,17411748. Bajorath, J., and Aruffo, A. (1997). Prediction of the three-dimensional structure of the human Fas receptor by comparative molecular modelling.]. Comput. AidedMol. Des. l l , 3-8. Banner, D. W., D’Arcy, A,, Janes, W.. Gentz, H., Schoenfeld, H. J., Broger, C., Loetscher, H., and Lesslauer, W. (1993).Crystal structure of the soluble human 55 kd TNF receptorhuman TNF beta complex: Implications for TNF receptor activation. Cell 73, 431-445. Bellosillo, B., Dalmau, M., Colomer, D., and Cil, J. (1997). Involvement of CED3/ICE proteases in the apoptosis of B-chronic lyinphocytic leukemia. Blood 89, 3378-3384. Bemdt, C., Mopps, B., Angerrniilller, S., Gierschik, P., and Krammer, P. H. (1998).CXCR4 and CD4 mediate a novel type of apoptosis in CD4 positive T cells. Submitted for publication. Bertin, J., Armstrong, R. C., Ottilie, S., Martin, D. A., Wang, Y., Banks, S., Wang, G. H., Senkevich, T. G., Alnemri, E. S., Moss, B., Lenardo, M. J., Tomaselli, K. J., and Cohen, J. I. (1997). Death effector domain-containing herpesvirus and poxvirus proteins inhibit both Fas- and TNFR1-induced apoptosis. Proc. Natl. Acad. Sci. USA 94, 1172-1176. Beyaert, R., Kidd, V. J., Cornelis, S., Van de Craen, M., Denecker, G., Lahti, J. M., Gururajan, R., Vandenabeele, P., and Fiers, W. (1997). Cleavage of PITSLRE kinases by ICEICASP-1 and CPP3WCASP-3 during apoptosis induced by tumor necrosis factor. ]. B i d . Chem. 272, 11694-11697. Bing, A., and Dou, Q. P. (1996). Cleavage of retinoblastoina protein during apoptosis: An interleukin 1 beta-converting enzyme-likeprotease as candidate. Cancer Res. 56,438-442. Black, R. A., Rauch, C. T., Kozlosky, C. J., Peschon, J. J., Slack, J. L., Wolfson, M. F., Castner, B. J., Stocking, K. L., Reddy, P., Srinivasan, S., Nelson, N., Boiani, N., Schooley, K. A., Gerhart, M., Davis, R., Fitzner, J. N., Johnson, R. S., Paxton, R. J., March, C. J., and Cerretti, D. P. (1997). A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells. Nature 385, 729-733. Bodnier, J. L., Bums, K., Schneider, P., Hofniann, K., Steiner, V., Thome, M., Bornand, T., Hahne, M., Schroter, M., Becker, K., Wilson, A., French, L. E., Browning, J. L.,
MacDonald, H. R., and Tschopp, J. (1997).TRAMP, a novel apoptosis-mediating receptor with sequence homology to tumor nrcrosis fictor receptor 1 antl Fas(Apo-l/Cl)95). I t t i t n u n i t ! / 6, 79-88. Boise, L. H., Gonzakz Garcia. M.. Postrnia. C. E.. Ding, L., Lindsten, T., Turka, L. A,, Mao, X., Nunez. G.. and Thompson. C;. B. (1993).Ucl-x, a bcl-2-related gene that functions as ii doininant regulator of apoptotic ccll dcath. Cvll 74, 597-608. Boise. L. H., Gottschalk. A. R.. Quintans. J., and Thompson, C. B. (1995a). Bcl-2 and Bcl2-rc-latedproteins in apoptosis regnlation. Curr. T o p Microbial. Zttirtiu?~ol. 200, 107-121. Boisr. L. H., Minn, A. J., Noel. P. J.. Junr, C. f l . . Accavitti, M. A,, Lindsten, T., and Thonipson, C:. B. (199%).CD28 costininlation can promote T cell survival by enhancing the expression of Bcl-X,,.Z t t t t i i t m i f y 3, 87-98. Boldin, M. P.. Goncharov, T. M., Goltscxv, Y. V.. d Wallach, D. (1996). Involvement of MACH, a novel MORTUFADD-interacting protcase. in Fas/APO-1- and TNF receptorindnced cell death. Cell 85, 803-815. Boldin, M . P., Varfolomeev. E. E., Pancrr. Z., Mett, I. L., Canionis, J. H., and Wallach, D. (1995). A novel protein that interacts with the death tloniain of Fas/APO-1 contains a sequence motif related to the death tlomain. J . B i d . Chon. 270, 7795-7798. Boyd. J. M.. Gallo, G . J., Elangovan, B.. Ilonghton, A. B., Malstrom, S., Avery, B. J., Ebb, R. C., Subramanian, T., Chittenden, T., Lritz, R. 1.. rt 01. (199s). Bik, a novel deathinducing protein shares a distinct seqnence motif with Bcl-2 family proteins and interacts with viral arid cehilar suni\~alproiiiotingproteins. 0 t m ) g e n e 11, 1921-1928. Brancolini, C., Benedetti. M., antl Schneich~r,C. (1995). Microfilament reorganization during apoptosis:The role ofGas2. a possible snl)strate for ICE-like proteases. EMBOJ. 14,51795190. Brojatscli, J., Naughton, J., Rolls. M. M . , Zingler, K., ant1 Young, J. A. (1996). CAR1, a TNFR-related protein, is a cellular receptor for cytopatliic avian leukosis-sarcomaviruses and inrdiates apoptosis. Ctdl 87, 845-855. Broonie, 11. E., Dargan, C. M..Krajewski, S., and Reed, J. C. (1995). Expression ofBcl-2, Bcl-x. antl Bax after T cell activation antl IL-2 withtlrawa1.J. I t t i n i t m d 155, 231 1-2317. Urowiing, J. L., Ngani E. K. A,. Lawton, P., DvMarinis, J.. Tizard, R., Chow, E. P., Hession, C.. O'Brine Greco, B., Folry, S. F., and \+';ire, C. F. (1993). Lyniphotoxin beta, a novel member of tlie TNF family that forms il lieteronieric complex with lymphotoxin on the cell surface. C d l 72, 847-8Fj6. Uninner. T., Mogil. R. J., LaFace, D., Yoo, N . J., hlahboubi, A,, Echeverii, F., Martin, S. J., Force, W. I{., Lynch, D. H., Ware, C . F., c't al. (199s). Cell-autonomons Fas (CD9S)/ Fas-ligand interaction inecliates activation-intlucetlapoptosis in T-cell hybridoinas. Nottire 373,44 1-444. B~i~iriell, B. A,, Heath, L. S., Adams, D. E., Lahti, J. M., and Kidd, \', J. (1990). Increased expression of a 58-kDa protein kinase leads to changes in the CHO cell cycle. Proc. N d . A c d . Sci. USA 87, 7467-7471. Camerini, D., Walz. G., Lornen, W.A,, Borst, J., and Seed, B. (1991). The T cell activation antigen C D E is a inember of the nerve gro\vth f;ictor/tumor necrosis fiictor receptor gene f"ami1y./. I t t i t n u n i d , 147, 3165-3169. Casciola Rosen. L. A., Anhalt, G. J., ant1 Rosen, A. (199ij). DNA-dependent protein kinase is one of a subset of autoantigens specifically cleaved early during apoptosis.J . E x p . Med. 182, 1625-1634. Casciola Rosen, L. A,, Miller, D. K., Anhalt, G.J., and Rosen, A. (1994). Specific cleavage of the 70-kDa protein component ofthe U1 sinall nuclear ribonucleoprotein is a characteristic biochernical featnre of apoptotic cell death. J , B i d . Ckem. 269, 30757-30760. Cerretti, D. P., Kozlosky, C. J.. Mosley. B., Nelson, N., Van Ness, K., Greenstreet. T. A,, March. C. J., Kronlieiin, S. R., Dmck, T., Carinizzaro, L. A,, et (11. (1992). Molecular cloning of the interleukin-1 beta converting enzyme. Scimcr 256, 97-100.
194
PETER H . KRAMMEH
ChaudhaI);, P. M., Eby, M., Jasmin, A,, Bookwaker, A., Murray, J., and Hood, L. (1997). Death receptor 5 , a new member of the TNFR family, and DR4 induce FADD-dependent apoptosis and activate the NF-kappaB pathway. Zrnmunity 7, 821-830. Chen, Z., Naito, M., Mashiina, T., and Tsuruo, T. (1996). Activation of actin-cleavable interleukin lbeta-converting enzyme (ICE) Family protease CPP-32 during chemotlierapeutic agent-induced apoptosis in ovarian carcinoma cells. Cancer Res. 56, 5224-5229. Cheng, J., Zhou, T., Liu, C., Shapiro, J. P., Braiier, M. J., Kiefer, M. C., Barr, P. J., and Mountz. J. D. (1994). Protection froin Fas-mediated apoptosis by a soluble form of the Fas molecule. Science 263, 1759-1762. Chinnaiyan, A. M., O’Rourke, K., Tewari, M., m d Dixit, V. M. (1995). FADD, a novel death doinain-containing protein, interacts with the death domain of Fas and initiates apoptosis. Cell 81, 505-512. Chinnaiyan, A. M.. O’Rourke, K.. Yu, G. L., Lyons, R. H., Garg, M., Duan, D. R., Xing, L., Gentz, R., Ni, J., and Dixit, V. M . (1996a). Signal transduction by DR3, a death domain-containing receptor related to TNFR- 1 and CD95. Science 274, 990-992. Chinnaiyan, A. M., Orth, K., O’Rourke, K., D i m , H., Poirier, G. G., and Dixit, V. M. (19961~). Molecular ordering of the cell death pathway: Bcl-2 and Bcl-xL function upstream of the CED-Mike apoptotic proteases. J. B i d . Cheiii. 271, 4573-4576. Chinnaiyan, A. M., Tepper, C. G . , Seldin, M. F., O’Rourke, K., Kischkel, F. C., Hellbardt, S., Krammer. P. H., Peter, M. E., and Dixit, V. M. (1996~). FADD/MORTl is a coininon mediator of CD9S ( FadAPO-1) and tumor necrosis factor receptor-induced apoptosis. J. B i d . Cherri. 271, 4961-4965. Chinnaiyan, A. M., O’Rourke, K., Lane, B. R., and Dixit, V. M. (1997).Interaction of CED4with CED-3 and CED-9: A molecular framework for cell death. Science 275,1122-1126. Chittenden, T., Harrington, E. A,. O’Connor, R.. Fleinington, C., Lutz, R. J., Evan, G. I., and Guild, B. C. (1995). Induction of apoptosis by the BcI-2 hoinologue Bak. Nature 374, 733-736. Chiu, V. K., Walsh, C. M., Liu, C. C., Reed, J. C., and Clark, W. R. (1995). Bcl-2 blocks degranulation but not Fas-based cell-mediated cytotoxicity./. Ztiztnunol. 154,2023-2032. Choi, M. S., Boise, L. H., Gottschalk, A. R., Quintans, J., Thompson, C. B., and Klaus, G. G. (1995).The role of Bcl-Xl,in CD4O-mediated rescue froin anti-mu-induced apoptosis in WEHI-231 B lymphoma cells. Eur. J . Ittiniund 25, 1352-1357. Chu, K., Niu, X., and Williams, L. T. (1995). A Fas-associated protein factor, FAF1, potentiates Fas-mediated apoptosis. Proc. Not/. Acad. Sci. USA 92, 11894-1 1898. Cifone, M. G., De Maria, R., Roncaioli, P., Rippo. M. R., Azuma, M., Lanier, L. L., Santoni, A., and Testi, R. (1994).Apoptotic signaling through CD9g (Fas/Apo-1) activates an acidic sphingoniyelinase.J. E x p Metl. 180, 1547-1552. Cook, W. D., Metcalf, D., Nicola, N. A,, Burgess. A. W., and Walker, F. (1985). Malignant transformation of a growth factor-dependent inyeloid cell line by Abelson virus without evidence of an autocrine mechanism. Cell 41, 677-683. Copeland, K. F., Haaksma, A. G., Goudsmit, J., Krammer, P. H., and Heeney, J. L. (1994). Inhibition of apoptosis in T cells expressing human T cell leukemia virus type I Tax. AIDS Res. Hton. Retrooir. 10, 1259-1268. Coustan-Smith, E., Kitanaka, A,, Oui, C.-H., McNinch, L., Evans, W. E., Rairnondi, S. C., Brehm, F. G., Arico, M., Campano, D. (1996). Clinical relevance of Bcl-2 overexpression in clinical childhood acute lymphoblastic leukemia. B l o o d 87, 1140- 1146. Crowe, P. D., VanArsdale, T. L., Walter, B. N., Ware, C. F., Hession, C., Ehrenfels, B., Browning, J. L., Din, W. S., Coodwin, R. G., and Smith, C. A. (1994). A lymphotoxinbeta-specific receptor. Science 264, 707-710. 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
interleiikin-lheta-coiiverting enzynie/Ced-3 protease distinct from the poly(ADP-iibose) polymerase protease. J . B i d C/trrn. 271, 31277-:31282. Cuende. E., Ales Martinez. J. E.. Ding, L., Gonzalcz Garcia, M., Martinez. C.,and Niinez. G. (1993). Programmed cell death by bcl-2-tlependent and intlependent niechanisins in B Iymphoma cells. EhlBO J . 12, 1555-1560. Wieringa, B.. and Hentlriks, W.(1997). N o evidence for involvement of mouse protein-tyrosine pIiosPhatase-BAS-like Fas-associated phosphatase-1 in Fasmediated apoptosis. 1.B i d . c h i . 48, :3021Fj-:30220. Danley, D. E., Chiiang. T. H.. and Bokoch. G. M. (1996). Defective Rho GTPase regulation by IL-1 Iwta-converting enzvn~e-mediatetlcleavage of D4 GDP dissociation inhibitor. I. ~ I t I t J l f ~ J I O157, ~. 500-50:3. Dbaiho, G. S. ( 1997). Regulation of the stress response by cerainide. Bioclwn. Soc. Trans 25, 557-561. Debatin, K . M.. and Krammer, P. 11. (1'395). Resistance to APO-1 (CD95)induced apoptosis in T-ALL is determined by a BCL-2 independent anti-apoptotic program. Leukcinin 9, 815-8". Debatin, K.-M. (1997). Cytotoxic drugs. progrmiined cell death, and the irninrine system: Defining new roles in an old play. J . Nntl. Caticrr Zti.rt. 89, 750-751. Ilegli-Esposti, M . A.. Sinolak. 1'. J., \Valczak. H., Waiigh, J., Huang, C. P., DuBose, R. F.. Goodwin, R. G.. and Smith, C. .4.(1997a). Cloning and characterization of T R A L R 3 , a novel member of the emerging TRAIL receptor fainily. J . Eq,. Merl. 186, 1165- 1170. Degli-Esposti, M. A,, Dongali, W. C., Sinolak. P., LVaugIi, J. Y.,Smith, C. A,, and Goodwin, R. G. (199711).The novel receptor TRAIL-H4 induces NF-kappaB and protects against TRAIL-metliated apoptosis. yet retains an incomplete death dornain. Zmniutiity 7, 813-820. Demhic, Z., Loetscher. H., Gubler, U.. Pan. Y. C . . Lahm, H. It'., Gentz, R., Brockhaus, M . . and Lesslauer, W.(1990). T\vo huinan TNF receptors have similar extracelliilar, hut distinct intracellular. domain sequences. C!ytokittc, 2, 2 3 - 2 3 ? , Dhein, J.. Daniel, P. T., Trauth, €3. C., Oehm, A,. Moller, P.. and Kraininer, P. H. (1992). Induction of apoptosis by monoclonal antihody anti-APO-1 class switch variants is dependent on cross-linking of APO-1 cell surface antigens. J . ltwttrmd. 149, 3166-3173. Dhein, J., Walczak, H., B#umler, C., Debatin, K. M., and Krammer. P. H. (1995).Aiitocrine T-cell siiicide mediated b!. APO-l( Fas/CD95). Nmtrrre 373, 4.38-441. Dive, C., Evans, C. A,, and Whetton, A. D. (1992). Intliiction of apoptosis-new targets for cancer chcmotherapy. C a w e r B i d . 3, 417-427. Dole. M. G., Jasty, R., Cooper, M. J.. Thonipson, C. R., Nunez, G., and Castle, V. P. (1995). Bcl-xL,is expressed i n neuroblastoma cells and modrilates cheinotlierapy-indiIced apoptosis. Cntuxr H t x 55, 2576-2582. Duan, H.. Cliinnaiyan, A. M., Hudson, P. L., Wing, J. P., He, W.W., and Dixit, V. M . (1996a). ICE-LAP3, a novel niainnialian Iioinologiir~of the Caetior~trrhrlitiselegarts cell death protein Ced-3 is activated during Fas- ant1 tumor necrosis factor-induced apoptosis. J. Biol. C l i ~ i 271, . 1621-1625. Duan, H.. Orth. K . , Chinnaiyin, A. M.. Poirier. G. G.,Froelich, C. J.. Ile, W.W., and Ilixit. V. M. (l996h). ICE-LAP6, a novel nirin1)er of the ICE/Ced-3 gene family, is activated Iiy the cjtotoxic T cell protease granzymc B. J , B i d . C h e m 271, 16720-16724. Durkop, H., Latza, U.. I l i i n i i n c l , M., Eitelbach, F., Seed. B., a r i d Stein, H. (1992). Molecular cloning and expression of a new ineinlwr of thc iiwvc growth factor receptor family that is characteristic for IIodgkin's disease. Cell 68, 421 -427. Eck, M . J., and Sprang, S. R. (1989).The structure of tumor necrosis factor-alpha at 2.6 A resolution: Iinplications for receptor binding. J , B i d . Chetm 264, 17595- 17605.
196
PETER 11 KHkMMER
Eck, M. J., Ultsch, M., Rinderknecht, E., de Vos, A. M., and Sprang, S. R. (1992). The structure of human lymphotoxin (tumor necrosis factor-beta) at 1.9-A resolution. J. Biol. Chem. 267, 2119-2122. Emery, J. G., McDonnell, P., Burke, M. B., Deen, K. C., Lyn, S., Silverman, C . , Dul, E., Appelbaum, E. R., Eichman, C., DiPrinzio, R., et al. (1998). Osteoprotegerin is a receptor for the cytotoxic ligand TRAIL. J. Biol. Chem. 273, 14363-14367. Emoto, Y., Manome, Y., Meinhardt, G., Kisaki, H., Kharbanda, S., Robertson, M., Ghayur, T.. Wong, W. W., Kamen, K., Weichselbaum. R., et al. (1995). Proteolytic activation of protein kinase C delta by an ICE-like protease in apoptotic cells. EMBOJ. 14,6148-6156. Enari, M., Hase, A,, and Nagata, S. (1995a). Apoptosis by a cytosohc extract from Fasactivated cells. E M B O J. 14, 5201-5208. Enari, M., Hug, H., and Nagata, S. (1995b). Involvement of an ICE-like protease in Fasmediated apoptosis. Nature 375, 78-81. Enari, M., Sakahira, H., Yokoyama, H., Okawa, K., Iwamatsu, A,, and Nagata, S. (1998). A caspase-activated Dnase that degrades DNA during apoptosis, and its inhibitor ICAD. Nature 391, 43-50. Erhardt, P., Toniaselli, K. J., and Cooper, G. M. (1997).Identification ofthe MDM2 oncoprotein as a substrate for CPP32-like apoptotic proteases. J , Biol. Chem. 272, 15049-15052. Erickson, S. L., de Sauvage, F. J., Kikly, K., Carver Moore, K., Pitts Meek, S., Gillett, N., Sheehan, K. C., Schreiber, K. D., Goeddel, D. V., and Moore, M. W. (1994). Decreased sensitivity to tumour-necrosis factor but normal T-cell development in TNF receptor-2deficient mice. Nature 372, 560-563. Faleiro, L., Kobayashi, R., Feamhead, 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. Farber, S., Diamond, L. K., Mercer, H. D., et al. (1948). Temporary remissions in acute leukemia in children produced by folic acid antagonist 4-aminopteoylglutamicacid (aminopterin). N. Engl. J , Med. 28, 787. Farrow, S. N., White, J. H., Martinou, I., Raven, T., Pun, K. T., Grinhani, C . J., Martinou. J. C., and Brown, R. (1995). Cloning of a bcl-2 homologue by interaction with adenovirus E1B. Nature 374, 731-733. Faucheu, C., Blanchet, A. M., Collard Dutilleul, V., Lalanne, J. L., and Diu Hercend, A. (1996).Identification of a cysteine protease closely related to interleukin-1 beta-converting enzyme. Eor. 1.Biochein. 236, 207-213. Faucheu, C., Diu, A., Chan, A. W., Blanchet, A. M . , Miossec, C., Herve, F., Collard Dutilleul, V., Gu, Y., Aldape, R. A,, Lippke, J. A,, et al. (1995). A novel human protease similar to the interleukin-1 beta converting enzyme induces apoptosis in transfected cells. EMBO J. 14, 1914-1922. Feniandes-Alnemri, T., Armstrong, H. C . , Krebs, J., Srinivasula, S. M., Wang, L., Bullrich, F., Fritz, L. C., Trapani, J. A,, Tomaselli, K. J., Litwack, G., and Alnemri, E. S. (1996). In vitro activation of CPP32 and Mch3 by Mch4, a novel human apoptotic cysteine protease containing two FADD-like domains. Proc. Natl. Acad. Sci. USA 93,7464-7469. Fernandes-Alnemri, T., Litwack, G., and Alnemri, E. S. (1994). CPP32, a novel human apoptotic protein with homology to Cnenorhahdiitis elegans cell death proteiii Ced-3 and mammalian interleukin-1 beta-converting enzyme. J. Biol. Chein. 269, 30761-30764. Feniandes-Alnemri, T., Litwack, G., and Alnemri, E. S. (1995a). Mch2, a new member of the apoptotic Ced-3/ICE cysteine protease gene family. Cancer Re.s. 55, 2737-2742. Femandes-Alnemri, T., Takahashi, A,, Armstrong, R., Krebs, J., Fritz, L., Tomaselli, K. J., Wang, L., Yu, Z., Croce, C. M., Salveson, G . , et al., (1995b). Mch3, a novel human apoptotic cysteine protease highly related to CPP32. Cancer Res. 55, 6045-6052. Fisher, D. E. (1994).Apoptosis in cancer therapy: Crossing the threshold. Cell 78,539-542.
Fisher, G. N.,Rosenherg, F. J., Straus. S. E., Dale, J. K., Middleton, L. A,, Lin, A. Y.. Strober, W., Lenardo, M. J.. and Prick, J. M. (1995). Dominant interfering Fas gene mutations impair apoptosis in a hriman autoininiunc lyiirplroproliferative syndrome. Cell 81, 935-946. French, L. E., Ilahne, M.. Viard. I., Rdlgnil)er, C;., Zanone, R., Becker, K., Muller, C., and Tschopp, J. (1996).Fas and Fas ligand i n embryos and adult mice: Ligand expression in several irniniine-piivilege(l tissues and coexpression in adult tissues characterized by apoptotic cell turnover. I. Cull. Biol. 133, 335-34:3. Friesen, C.. Herr, I., Kraninirr, P. IJ., and Ilehatin, K.-M. (1996). Involvenrent of the CD95(APO-l/Fas) receptor/ligand system in dnig indnced apoptosis in leiikemia cells. Notiire Mcrl. 2, 574-Fj77. Friesen. C.. Fulda, S., and Debatin, K . - M . (1997). Deficient activation of the CD95 (APO-l/Fas) system i n dnig-resistant cells. Lc.trkettiin 11, 1833-1841. Fnlda, S.. Los, M..Friesen, C., Scaffidi, C;. A,, Mier, W., Benrdict, M., Nufiez, G., Kramrner, P. H . , Peter, M. E., and Debatin, K.-M. (1998). Betulinic acid triggers CD95(APO-l/ Fas) and p53-independent apoptosis via activation of caspases in neuroectoderinal turnors. Concur Re.s. 57, 4956-4964. Fulda, S., Sieverts, H., Friescii, C.. Herr. I.. and Debatin, K.-M. (1997).The CD95 (APO-1/ Fas) system mrdiates dnig indiiced apoptosis in neuroblastoma cells. Coricer Rus. 57, 3823-3829. G a k , P. R., I-Iofinann, W. J., b’alczak. I{., Schdler, H . , Otto, G., Streininel, W., Kraminer, P. I I . , and Riinkel, L. (1995).Involvement ofthe CD95 (APO-1/Fas) receptor and ligand in liver damage. J, Exp. hled. 182, 1223-1230. Geley, S., Hartmann, B. L., Kapelari, K., Egk, A., Villnnger. A,. Heidacher, D. Greil, R., Auer, B., and Kofler, R. (1997). The interleukin 1 beta-converting enzyme inhibitor CrmA prevents APO-l/Fas- hut not glucocorticoid-iiIduced poly(ADP-ribose) polymerase cleavage and apoptosis in 1pphol)lastic leiikemiu cells. FEBS Lett. 402, 36-40. Ghayur, T., Banerjee. S., Hngunin, M., Butler. D., Herzog, L., Carter, A,, Quintal, L., Sekut, L., Talanian, R.. Paskind, M., Wong. W.. Kainen, R., Tracey, D., and Allen, 11. (1997).Caspase-1 processes IFN-gamnra-indiicing factor and regulates LPS-induced IFNgamma production. Nuttrrc 386, 619-623. Ghaynr, T., Hngunin, M., Talanian, R. \’.> Ratnofsky, S.,Quinlan, C., Emoto, Y., Pandey, P., Datta, R.. Huang, Y., Kharhanda, S., Allen, Ii., Kamen, R., Wong, W., and Knfe, D. (1996). Proteolytic activation of protein kinase C delta by an ICE/CED 3-like protease induces clraracteristics of apoptosis. J. EXJJ.Med 184, 2399-2404. Godfrey, W. R., Faponi, F. F.. Harara. M. A,. Bnck, D., and Engleman, E. G. (1994). Identification of a Iiuinan OX-40 ligantl, a costiiniilator of C D 4 f T cells with homology to tumor necrosis factor. J . EX)). Merl. 180, 757-762. Goltsev, Y. V., Kovalenko, A. V.,Arnold, E., Varfolomeev, E. E., Brodianskii, V. M., and Wallach, D. ( 1997). CASH, a novel caspase lroinologue with death effector domains. 1.Biol. Cherti. 272, 19641-19644. Goodwin, R. G.. Alderson, M . R.. Smith, C. A,, Armitage, R. J., VandenBos, T.. JerLy, R., Tough, T. W., Schoenborn. M. A,, Davis Smith, T.. Hennen, K . , ut (11. (1993a). Molecular and biological characterization of a ligand for CD27 defines a new Family of cytokines with IionioloFn, to tninor nwrosis factor. Cell 73, 447-456. Goodwin. R. G., Din, W.S., Davis Smith. T., Anderson, D. M., Girnpel. S. D., Sato, T. A,. Maliszewsla, C. R., Brannan, C. I.. Copeland. N. G., Jenkins, N. A., et ul. (1993b). Molecular cloning of a ligand for the indncible T cell gene 4-1BB: A member of an emerging family of cytokines with homology to tumor necrosis factor. Eur. J . Intmnnol. 23, 2631-2641.
198
PETER H . KRAMMER
Gorczyca, W., Bigman, K., Mittelman, A,, Ahtned, T., Gong, J., Melamed, M. R., and Darzynkiewicz, Z. (1993). Induction of DNA strand breaks associated with apoptosis during treatment of leukemias. Leukemia 7, 659-670. Gottschalk, A. R., Boise, L. H., Thompson, C. B., and Quintans, J. (1994). Identification of iinmunosuppressaiit-induced apoptosis in ii inurine B-cell line and its prevention by bcl-x but not bcl-2. Proc. Natl. Acad. Sci. IJS4 91, 7350-7354. Graf, D., Korthauer, U.. Mages, H. W., Senger, G., and Kroczek, R. A. (1992). Cloning of TRAP, a ligand for CD40 on human T cells. Eur. 1. Imtiuno!. 22, 3391-3194. Graninger, W. B., Seto, M., Boutain, B., Goldman, P., and Korsnieyer, S. J. (1987).Expression of Bcl-2 and Bcl-2-Ig fusion transcripts in normal and neoplastic cells. I. Clin. Inoest. 80, 1512-1515. Gray, P. W., Agganval, B. B., Benton, C. V., Bringman, T. S., Henzel, W. J.. Jarrett, J. A,, Leung, D. W., Moffat, B., Ng, P., Svedersky, L. P., et al. (1984). Cloning and expression of cDNA for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature 312, 721-724. Grell, M., Douni, E., Wajant, H., Lohden, M., Clauss, M., Maxeiner, B., Georgopoulos, S., Lesslauer, W., Kollias, G., Pfizenmaier, K . , et al. (1995). The transmembrane form of tumor necrosis factor is the prime activating ligand of the 80 kDa tumor necrosis factor receptor. Cell 83, 793-802. Gu, Y., Kuida, K., Tsutsui, H., Ku, G., Hsiao, K., Fleming, M. A,, Hayashi, N., Higashino, K., Okaniura, H., Nakanishi, K., Kurimoto, M., Tanimoto, T., Flavell. R. A,, Sato, V., Hardmg, M. W., Livingston, D. J., and Su, M. S. (1997). Activation of interferon-gamma inducing factor mediated by interlerikin-lbeta converting enzyme. Science 275,206-209. Gu, Y., Sarnecki. C., Aldape, R. A,, Livingston, D. J., and Su, M. S. (1995). Cleavage of poly(ADP-ribose)polymerase by interleukin-1 beta converting enzyme and its honiologs TX and Nedd-2. /. Biol. Chem. 270, 18715-18718. Hahne, M., Himoldi, D., Schroter, M., Romero, P., Schreier, M., French, L. E., Schneider, P., Bomand, T., Fontana, A,, Lienard, D., Cerottini, J., and Tschopp, J. (1996). Melanoma cell expression of Fas(Apo-l/CD95) ligand: Irnplications for tumor immune escape. Science 274, 1363-1366. Han, D. K. M., Chaudhary, P. M., Wrigth, M. E., Friedman, C., Trask, B. J., Riedel, R. T., Baskin. D. G., Schwartz, S. M., and Hood, L. (1997). MRIT, a novel death-effector domain-containing protein, interacts with caspases and Bcl-X,, and initiates cell death. Cell Bid. 94, 11333-11338. Hanada, M., Aime Sempe, C., Sato, T., and Reed, J. C. (1995). Structure-function analysis of Bcl-2 protein. Identification of conserved domains important for hornodimerization with Bcl-2 and heterodirnerization with Baw. 1. Biol. Chem. 270, 11962-11969. Hata, EI., Matsuzaki, I%., Takeya, M., Yoshida, M., Sonoki, T., Nagasaki, A,, Kuribayashi, N., Kawano, F., and Takatsuki, K. (1995). Expression of Fas/APO-l(CD95) and apoptosis in tumor cells from patients with plasma cell disorders. Blood 86, 1939-1945. Hermine, O., Haioun, C., Lepage, E., d’Agay, M.-F., Briere, J., Lavignac, C., Fillet, G., Salles, G., Marolleau, J.-P., Diebold, J., Reyes, F., and Gaulard, P. (1996). Prognostic significance of Bcl-2 protein expression in aggressive non-Hodgkin’s lymphoma. Blood 87, 265-272. Herr, I., Bohler, T., Wilhelm, D., Angel, P., and Debatin, K.-M. (1997). Activation of the CD95 (APO-l/Fas) signaling by ceramide mediates cancer therapy induced apoptosis. EMBO J. 16, 6200-6208. Hsu, H., Solovyev, I., Colombero, A,, Elliott, R., Kelley, M., and Boyle, W. J. (1997).ATAR, a novel tumor necrosis factor receptor family member, signals through TRAF2 and TRAF5. /. Biol. Chem. 272, 13471-13474.
Hn, S., Vincenz. C.,Uuller. M., and Disit, \’. M . (1H97a). A novel family of Lira1 death effector domain-containing molecules that inhibit both CI195- and tumor iiecrosis ftactor receptor-1-indiicrtl upoptosis. J . Biol. Clwtti. 272, 9621-9624, I I u , S.,Vincenz. C . , Ni, J.. Centz, R., airtl Ilixit. \’, M . (19971,). I-FLICE, a novel inliilitor of tnmor ncwosis factor rccvptor-1- a i d CD9S-inchiced apoptosis. J . Biol. CIirttL. 272, 172S55-17257. IIiiang. B., Eberstadt, M.. Olejniczak, E. T.. Me~itlows.R. I!. ~ i n t lFesik, S . W.(1996).NMR strnctnre and inntugenesis ofthe Fas (Al’O-I/C1195)cleath tloinain. N(ztrrrc, 384,638-641. Inoliara, N., Koseki, T., Hu, Y..C h i , S..and Niinez, G. (1997). CLARP. a death effector doiliain-containing protein interacts with caspase-8 and regiilates apoptosis. Crll Biol. 94, 10717-10722. lrinler. M., Tlionirs. M., Hahnc>, M., Schnc~itlrr,P., Hofiiiaiin, K., Striner, V.,Bodnicr. J. I,,, Schroter, M., Biinis, K., Mattmarin, C., liiinoltli, D., French, L. E., and Tschopp, J. (1997). Inhihition of death receptor signals hy cellular FLIP. Nottire 388, 190- 19!5. Itoh, N.. Yonehara, S., Ishii. A.. Yoneliara. M., Mizushinin, S.,Sanieshinia, M., Ilase, A , , Seto, Y., and Nagata, S. (1991). The poIy~q)ticleencodetl by the cDNA for hiinian cell surface antigen Fas can nietliate apoptosis. Ccll 66, 233-243. Itoh. N..Tsiijimoto. Y., and Nagata, S. (199:3).Effbct of bcl-2 on Fas antigen-mediated cell death. 1. Z t t i i t ~ r t t i o l .151, 621-627. Jiiiittrlii, M.,Benedict, M., Tewari, M.. Sliayman, J. A,, ant1 Dixit, V. M. (1995). Bcl-x antl Bcl-2 inhibit T N F and Fas-induced apoptosis antl activation of pliospholipase A2 in breast c;ircinoin;i cells. Otirogcwc. 10, 2297-2:305. Janicke, R. U., L%’alker,P. A,, Lin. X Y.,and Porter. A . G. (1996). Specific cleavage of the retinoldastoma protein by an ICE-like protcasc in apoptosis. EA4BO J . 15, 6969-6978. Johnson, D.. Lanahan.A , . Biick, C. R., Schgal. A,, Morgan. C., Mercer, E., Bothwell, M.. and Chao, M. (1986).Expression and striictrire ofthe hnin;in N G F receptor. Cell 47,545-554. Jones. E. Y., Stuart. D. 1.. and Walker, N. P. (1992). (;r).stal stnictnre of TNF. Zirmrcnol. Ser. 56, 93- 127. Ju, S. T.. Panka. D. J., Cni, H., Ettinger, R., el Kliatil), M.. Sherr, D. H., Stanger, B. Z., and Marshak Rothstein. A. ( 1995).Fas(C;DYS)/FasLinteractions required for prograinmed cell deatli after T-cell acti\.ation. Notrrrc 373, 444-448. Kamens, J., Paskintl, M., Hugiinin, M., Talanian, 11. V., Allen, 11.. Banach, D., Buinp, N.. Hackett, M . . Johnston, C. G.,Li. P., ~t al. (199Ti). Itlentification and characterization of I C I 1-2, a novel member of the interleiikin-l I)et;t-coiiverting enzyme family of cysteine proteases. J . Biol. C/iem. 270, 15250- 15256. Kayagaki. N.. Kawasaki. A,. Ehuta, T.. Ohinoto. H.. Ikeda, S.. inoue, S., Yoshino, K.. Okiimnra, K.,and Yagita. H. ( 1995). Metalloprotrinasc-mediated release of human Fas ligand. J . E q i . h f d 182, 1777-1783. Ka)ial;ir. C.. Ord, T., Testa, M . P., Zhong. L. T.. antl Bredesen, D. E . (1996). Cleavage of actin by interlcnkin 1 hrto-converting enqxie to reverse DNase I inhibition. Proc. N d l . Aiwl. Sci. l G A 93, 2234-2238. Kaufniann. S. H., Desnoyers, S., Ottaviiilo. Y.. Davidson, N . E., and Poirier. G. G. (1993). Specific proteolytic cleavage of pol? ( ADP-rit)ose) plyinerase: An early marker of chemotherapy-indiiced apoptosis. CattcrJt-Rex. 53, 3976-33985. Kerr, J. F., Wyllie, A. H., and Cnrrie, A. It. (1972).Apoptosis: A hasic bio1ogic;tl phenomenon with wide-ranging iniplications i n tissnc. kinetics. Br. J . Cottcrr 26, 239-257. Kiefer. M. C., Braner. M. J., Powers. \’. C . , \Vu, J . J., Uniansky, S. R., Toinei. L. D., and Barr. P. J. (1995). Modulation of apoptosis by the widely distril)uted Bcl-2 Iioinologue Bak. Notirrc, 374, 736-739. Kischkel, F. C., IIcllhardt, S.. Belirniann. I., Grrmer, M., Pawlita. M., Krammer. P. 11.. and Peter, M , E. (1995).Cytotoxicity-drpentlent APO-1 ( Fas/CD95)-associated proteins
200
PETER ti. KKAMMEH
form a death-inducing signaling complex (DISC) with the receptor. EMBOJ.14,55795588. Kischkel, F. C., Kioschis, P., Weitz, S., Poustka, A., Lichter, P., and Kraminer, P. H. (1998). Localization of the human and inurine caspase-8 (FLICE/MACH/MchS) genes by FISH. Submitted for publication. Kitson, J., Raven, T., Jiang, Y. P., Goeddel, D. V., Giles, K. M., Pun, K. T., Grinham, C. J., Brown, R., and Farrow, S. N. (1996). A death-domain-containing receptor that mediates apoptosis. Nature 384, 372-375. Klas, C., Debatin, K.-M., Jonker, R. R., and Krammer, P. H. (1993). Activation interferes with the apoptotic pathway in mature human T cells. Znt. Znirnunol. 5(6), 625-630. Knipping, E., Debatin, K.-M., Heilig, B., Eder, A.. and Krammer, P. H. (1995).Identification of soluble APO-1 in supernatants of human B and T cell lines and increased serum levels in B and T cell leukemias. Blood 85, 1562-1569. Kolesnick, R. N., Haiinovitz Friedman, A,, and Fuks, Z. (1994). The sphingoinyelin signal transduction pathway mediates apoptosis for tumor necrosis factor, Fas, and ionizing rahation. Biochem. Cell Biol. 72, 471-474. Kozopas, K. M., Yang, T., Bnchan, H. L., Zhou, P., and Craig, R. W. (1993). MCL1, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to BCLS. Proc. Natl. Acad. Sci. USA 90, 3516-3520. Krajewski, S., Blomqvist, C., Franssila, K., Krajewska, M., Wasenius, V.-M., Niskanen, E., Nordling, S., and Reed, J. C. (1995). Reduced expression of proapoptotic gene Bax is associated with poor response rates to combination chemotherapy and shorter survival in women with metastatic breast adenocarcinoma. Cancer Res. 55, 4471-4478. Krajewski, S., Krajewska, M., Shabaik, A,, Miyashita, T., Wang, H. G., and Reed, J. C. (1994). Immunohistochemical determination of in vivo distribution of Bax, a dominant inhibitor of Bcl-2. Am. J Pathol. 145, 1323-1336. Krammer, P. H. (1997).The tumor strikes back: New data on expression of the CD95(APOW a s ) receptodigand system may cause paradigm changes in our view on drug treatment and tumor immunology. Cell Death Diff 4, 362-364. Krammer, P. H., Behrmann I., Daniel, P., Dhein J., and Debatin K.-M. (19944. Regulation of apoptosis in the immune system. Curr. +in. Zmmnunol. 6, 279-289. Krammer, P. H., Dhein, J., Walczak, H., Behrmann, I., Mariani, S., Matiba, B., Fath, M., Daniel, P. T., Knipping, E., Westendorp, M. 0.. Stricker, K., Baumler, C., Hellbardt, S., Germer, M., Peter, M. E., Debatin, K.-M. (1994b). The role of APO-1 mediated apoptosis in the immune system. Ininwrnol. Rev. 142, 175-191. Krammer, P. H., Galle, P. R., Mdler, P., and Debatin, K.-M. (1998). CDSS(APO-l/Fas)mediated apoptosis in normal and malignant liver-, colon- and hematopoietic cells. I n “Advances in Cancer Research” (G. F. Vande Wonde and G. Klein, eds.), vol. 75, pp. 252-273. Academic Press. 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. Kuida, K., Zheng, T. S., Na, S., Kuan, 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, 368-372. Kumar, S., Kinoshita, M., Noda, M., Copeland, N . G., and Jenkins, N. A. (1994). Induction of apoptosis by the mouse Nedd2 gene, which encodes a protein similar to the product of the Cuenorhubditis elegans cell death gene ced-3 and the mammalian IL-1 betaconverting enzyme. Genes. Dev. 8, 1613-1626.
Kwon, B. S., Tan. K. B., Ni. J., Lee, K. O., Kim, K. K.. Kim, Y.J., \Vang, S., Gentz. R., Y u , C:. L., Harrop, J., Lyn. S. D.. Silverinan, ( ; . ~Porter, T. G.,Triineh, A,, and Yonng, P. 13. (1997).A newly itlentifietl ineml)er of the tiimor necrosis factor receptor superfamily &it11a \vide tissue distril)ntion and involvcwic.nt in Iyinphoc>-te;ictivation. /. Biol. Chcrn. 272, 14272- 14276. Lacry, D. L.. Tininis, E.. Tan, €I. L.,Kelley, M. J., Ihiistan, C . R.. Bnrgess, T., Elliott, R., Colonibero, A.. Elliott, G . , Scrilly, S.. ct cr/. (1998).Osteoprotegerin ligand is a cytokine that regulatc,s osteoclast differentiation and activation. Cull 93, 165-176. Lahti. J. M.. Xi;ing, J.. Heath, I,. S.. C a n p n a , D.. and Kitld, V. J. (1995).PITSLRE protein kinase acti\it\. is associatd with apoptosis. 1 ! 4 0 / . C:c// Biol. 15, 1-1 1. LeithYnser, F.. Uhein, J., Mechtershciiiier. C . , Koretz, K., Bruderlein, S., Ilenne, C., Schniidt. A., Debatin, K. M., Krammrr. P. H,, and Miiller, P. (199.3). Constitutive and intlncetl expression of APO- 1, a ne\v nlrmhrr of the nerve growth factorltumor necrosis Factor receptor superfamily. in normal and neoplastic cells. Lab. Itwest. 69, 415-429. Lenardo, M. J. (1991). Interleukin-2 progrmis inoiisv alpha beta T lymphocytes for apoptosis. Nattirc. 353, 858-861. Levine. A. J. (1997).1353, the cellular gatekeeper for growth antl tli\.ision. Crl/ 88, 323-331. Li, P.. Allen, II., Benrrjee, S.,Franklin, S., Herzog. L.. Johnston. C., McDowt~ll,J., Paskintl. M., Rotlinan. L., Salfeld. J., et i d , (1995). Mice deficient in IL-1 beta-converting enzyme are defective i n production of inaturc. IL-I beta and resistant to ( d o t o x i c shock. C P / / 80,401-411. Liclrtrr, P., \Valczak. II., Ll’eitz, S. Belirnrann. I., and Krammer. P. €1. (1992).The hiinlan APO-I antigen maps to 1Oq23. ii region which is syntenic with iiioiise cliroinosoine 19. Ccnornits 14, 179-180. Lin, E. Y.,Orlofkhy, A,, Berger, M . S.. and Pnstowsky. M. B. (1993). Characterization of Al, ii novel 1iernoI)oietic-spc.cific early-response gene with sequence similarity to 1x1-2. /. r l l l r r l l ~ r t o / .151, 1979-1988. Lippke, J. A,, Gn, Y.. Samecki, C., Caroii, P. R., and SII. M. S . (1996). Identification and characterization of CPP3WMch2 homolog 1, a iiovcl cysteine protease siinilar to CPP32. /. Biol. Chcttt. 271, 1825-1828. Liu, X..Zon, II.. Slaughter. C.. and \\’ang, X. (1997). DFF. a heterodiineric protein that fnnctions downstream of caspse-3 to trigger D N A fi.agnicWation during apoptosis. Cell 89, 17.5-184. Loetscher, H., Pan, Y. C.. Lahin, 14. Lj’., Ckntz, It., Brockliaris. M.. Tahiichi, H.. and Lesslauer, \V, ( 1990).Moleciilar cloning and vspression of the human 55 kd tririior necrosis factor receptor. Crd 61, 351 -359. Los. M.. Herr, I . , Friescn, C.. Fnltla, S.. Scliiilze-Ostlioff, K.. antl Dehatin. K.-M. (1997). Crossresistance of CD9.5- and tlmg-induced apoptosis as a consequence of deficient activation of caspises ( IC:E/Cetl-:3 proteases). Blood 90, 31 18-3129. Los. M., Van de Craen, M., Penning, L. C., Schenk. I I . , Westendorp, M.. Baeurrle, P. A,. Driige, LV.. Kramint~r,P. I.I., Fivrs. \’\:.> antl Schiilze Osthoff, K. (1995). Requirement of iin ICElCED-.3 protease for Fas/APO-I -mcdiatcd apoptosis. Notiire 375, 81-83. Lowe. S.W’.,Kirley, H. E., Jacks, T.. and IIonsnian, D. E. (1993).p53-tlcpendent apoptosis nrodiilates the cytotoxicity of anticancrr agrnts. Ci.11 74, 957-967. Mallett, S.,Fossnin, S.. and Barclay, A . N . i 1990). Characterization of the MRC OX40 antigen of activated CD4 positive T lyinph?-tes: A molecule related to nerve growth factor receptor. EAlBO /. 9, 1063- 1068. Mariani, S. M.. Matilxi, B., Biiiimler. C., and Kraininer, P. H. (1995). Regulation of cell surface APO-I/Fas (CD95) ligand expressioll l)y inrtalloproteases. Eiir. /, I t i i r r i i m o / . 25, 2303-2307.
202
I’ETEH M . KHAMMEH
Mariani, S. M., and Kraminer, P. H. (1998). Ilifferential regulation of TRAlL/APO2L and CD95L in transformed cells of the T and B lymphocyte lineage. Eur. J. I i n t ~ i u d 28, 973-982. Marsters, S. A,, Sheridan, J. P.. Doiiahue, C. J., Pitti, R. M., Gray, C. L., Goddard, A. D., Bailer. K. D., and Ashkenazi, A. (1996). Apo-3, a new member of the tumor necrosis factor receptor family, contains a death domain and activates apoptosis and NF-kappa B. Curr. Bid. 6, 1669-1676. Marsters, S. A,, Sheridan, J. P., Pitti, R. M., Huang, A. D., Skubatch, A. D., Baldwin, A. D., Yuan, A. D., Gurney, A. D., and Goddard, A. D. (1997). A novel receptor for APOZmRAIL contailis a truncated death domain. Cirrr. B i d 7, 1003-1006. Martin, S. J., O’Brien, G. A,, Nishioka, W. K., McGahon, A. J., Mahboubi, A,, Saido, T. C., and Green, D. R. (1995a).Proteolysis of fodiin (non-erythroidspectrin)chringapoptosis. J. B i d . Cheni. 270, 6425-6428. Martin, S. J., Reutelingsperger, C. P., McCahon, A. J., Rader, J. A,, van Schie, R. C., LaFace, D. M., and Green, D. R. (1995b). Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: Inhibition by overexpression of Bcl-2 and Abl. J. Exp. Med. 182, 1545-1SC56. Masliima, T., Naito, M., Fujita, N., Noguchi, K., and Tsuruo, T. (1995). Identification of actin as a substrate of ICE and an ICE-like protease and involvement of an ICE-like protease but not ICE in VP-16-induced U937 apoptosis. Biocherti. Biophys. Re.9. Commun. 217, 1185-1192. Mashima, T., Naito, M., Noguchi, K., Miller, D. K., Nicholson, D. W., and Tsunio, T. (1997).Actin cleavage by CPP-3Yapopain duriiig the development of apoptosis. Oncogene 14, 1007-1012. McDonnell, T. J., Deane, N., Platt, F. M., Nunez, G., Jaeger, U., McKeam, J. P., and Korsrneyer. S. J. (1989). bcl-2-iminiinoglobulin transgenic mice demonstrate extended B cell survival and follicular lymphoproliferation. Cell 57, 79-88. Medema, J. P., Scaffidi, C., Kischkel, F. C., Shevchenko, A,, Mann, M., Krammer, P. H., and Peter, M. E. (19974. FLICE is activated by association with the CD95 death-inducing signaling complex (DISC). E M B O J. 16, 2794-2804. Medenia, J. P., Toes, R. E. M., Scaffidi, C., Zheng, T. S.,Flavell, R. A,, Melief, C. J. M., Peter, M. E., Offringa, R., and Krammer, P. H. (1997b). FLICE activation by granzyme B during CTL mediated apoptosis. Eur. J. Itnmunol. 27, 3492-3498. Medema, J , P., Scaffidi,C., Krammer, P. H., and Peter, M. E. (1998).Bcl-xl,acts downstream of FLICE/caspase-8, J. B i d . Chem. 273, 3388-3393. Memon, S. A,, Moreno, M. B., Petrak, D., and Zacharchuk, C. M. (1995). Bcl-2 blocks glucocorticoid- but not Fas- or activation-induced apoptosis in a T cell hybridonia. J. Ininiunol. 155, 4644-4652. Metcalf, D., Roberts, T. M., Cherington, V., and Dunn, A. R. (1987).The in vitro behavior of hemopoietic cells transformed by polyoma middle T antigen parallels that of primary human inyeloid leukemic cells. EMBO J. 6, 3703-3709. Micheau, 0.. Solary, E., Hammann, A,, Martin, F., and Dimanche-Boitrel, M. T. (1997). Sensitization of cancer cells treated with cytotoxic dnigs to Fas-mediated cytotoxicity. J. Natl. Cancer Inst. 89, 783-789. Min, Y. H., Lee, S.,Lee, J. W., Chong, S. Y., Mahn, J. S., and KO, Y. W. (1996). Expression of Fas antigen in acute inyeloid leukemia is associated with therapeutic response to chemotherapy. Br. J. Haemnutol. 93, 928-930. Miossec, C., Dutilleul, V., Fassy, F., and Diu-Hercend, A. (1997). Evideiice for CPP32 activation in the absence of apoptosis during T lymphocyte stimulation. J. Biol. Chem. 272, 13459-13462.
Mittl, P. R.. Di Marco, S.. Krelis, J. F., Hai. X., Karanewsky, 11. S., Priestle, J. P.. Tomaselli, K. J.. and Chitter, M. C. (1997). Structure of recoinbinant hiiman CPP:32 in complex with the tetrapeptide acctyl-Asp-Val-.41~i-Aspflrioronirthyl ketone. J . Biol. CIwn. 272, (3539-6547. Miura, M., Friedlander, R. M., and Yuan, J. ( 1995).Tunior necrosis factor-indriced apoptosis is niediated by a CrniA-sensitive cell death pathway. Proc. Natl. Acad. Sci. USA 92, 83188322. Miyashita, T., and Reed, J. C. ( 1992). bcl-2 gene transf&r incrcwes relative resistance of S49. 1 and LVEH17.2 lyniphoid cells to cell death antl DNA fragmentatioil indiiced hy glucocorticoids a i i d rnultiple clieiiiotlierapeutic. drugs. Cancer Res. 52, 5407-541 1 , Mipashita, T., and Reed, J , C;. (1993). Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in ii human leukemia cell liiic.. Blootl 81, 151-157. Miyashita, T., Krajewski, S., Krajc,wska. M.. LYang. H. G., Lin, H. K., Lieberinann, D. A,. Hoffman. B., and Heed, J. C. (1994). Tiiirior s iip 1mwo r p53 is a regulator of Bcl-2 and ba); gene expression in vitro antl in vivo. Oticogctie 9, 1799-1805. Miyashita. T., and Reed, J. C. (1995). Tunior suppressor p5C is a tlirect transcriptional activator of the huniaii hi^^ geiir. C d l 80, 293-289. M o k r . P., Koretz, K., Lritliiinsrr, F.. Briiderlein, S., Heiiiie, C., Quentmeier, A , , and Krammer, P. 11. (1994). Expressions of APO-1 (CD95). a member of the NGF/TNF receptor siiperfamily, in norinal and iieoplastic coloii epithelium. Int. /. Cnriccr 57, 1-7. Mongkolsapaya, J., C o y e r , A. E., Xu, X N., Morris, G,, McMichael, A. J., Bell, J. I., antl Screaton. G. R. (1998).Lymphocyte iiihibitor of TRAIL (TNF-related apoptosis-indiicing ligaiitl): A new receptor protecting lyinphocytes from the death ligand TRAIL./. I u i i t i u r w l . 160, :3-6. Montgomery, R. I., Warner, M . S., Luin, 8 . J., iind Spear, P. C. (1996). Herpes simplex virus-1 entiy into cells inetliutetl by ii novel member of the TNF/NGF receptor family. Cell 87, 427-436. Muss, M. L,, Jinq S. !!., M i h > kl. E.. BiirkIi;irt%\V.> ( h r t e r > tl. I , . . Chen. W.J.. <:l;y. u’. C . , 1)idsbury. J . I{., 1 lassler, I ). I l o f f i i i ; i i i , (-;. K . . Knst. T. A,, 1,;imhert. M. H.. Lersnitzer, M. A , . McCauley, P., Mc(;eehan, C . , Mitchell, J., Moyer, M., Pahd, C;.. Rocqiie, W., Overton, L. K.,Schoenrn, F., Seatoii, T., Su, J. L.,Warner, J., Bwherer, J. D., et 01. (1997). Cloning of a disintegrin metalloproteiilase that processes precursor tiinlour-necrosis factor-+ha. Natrrre 385, 733-736. Miillrr, M., Strand, S., Hug. H., Heineniann, E. M., Walczak, H., Hofinann, W.J., et a!. (1997). Drug induced apoptosis in Iiepatonia cells is inediated by the CD9g (AP0-U Fas) receptor/ligand system and involves activation of wild-type p53. J . C h . Inorst. 99, 403-413. Mundav, N. A , , Vaillaiicourt. J. P., Ali, A., Casano, F. J., Miller. D. K.. Molineaux, S. M., Yamin, T. T.,Yu, V. I,,, antl Nicholson. D. W.(1995).Molecular cloning and pro-apoptotic activity of ICErelII and ICErelIII, nienibers o f the ICEICED-8 family of cysteine proteases. /. Biol. Cheiti. 270, lFjH70-15876. Muzio. M.. Chinnaiyan, A. M.. Kischkel, F. C., O’Rourke, K., Shrvchenko, A,, Ni, J.. Scaffidi, C., Bretz. J. D., Zhang, M., Gcntz. I<., Mann, M.. Kraniiner, P. M.,Peter, M. E., and Dixit, V. M. (1996). FLICE, a novel FADD-homologous ICEKED-S-like protease, is recniited to the CD95 ( Fas/APO-1) deatli-inducing signaling compkx. Cell 85, 817-82:, Muzio, M., Salvesen, G. S., and Dixit, \’. M. (1997). FLICE induced apoptosis in a cellfree system: Cleavage of caspase zyniogens. /. Biol. C h e m 272, 2952-2956. Nki, S., Chuang, T. H., Curniingham, A , , Tllri, T. G., Hanke, J. H., Bokoch, G. M., and Danley, U. E. (1996).D4-GD1, a siibstrate of CPP32, is proteolyzed during Fas-induced apoptosis. /. Biol. Cheiti. 271, 11209-1 1213.
204
PETER H. KRAMMER
Nagata, S., and Golstein, P. (1995). The Fas death factor. Science 267, 1449-1456. Nakamura, T., Tashiro. K., Nazarea. M., Nakano, T., Sasayama, S., and Honjo. T. (1995). The murine lyniphotoxin-beta receptor cDNA: Isolation by the signal sequence murine lymphotoxin-betareceptor cDNA: Isolation by the signal sequence trap and chromosomal mapping. Genotnics 30, 312-319. Neamati. N., Fernandez, A,, Wright, S., Kiefer, J., and McConkey. D. J. (1995). Degradation of lamin B1 precedes oligonucleosoinal DNA fragmentation in apoptotic thynocytes and isolated thyinocyte nuclei. J . Ztntniriiol. 154, 3788-3795. Neilan, J. G., Lu, Z., Afonso, C. L., Kutish, C;. F.. Sussman, M. D., and Rock, D. L. ( I 993). An African swine fever vinis gene with similarity to the proto-oncogene bcl-2 and the Epstein-Barr virus gene BHRFI. J . V i r d 67, 4391-4394. Ni, R., Tornita, Y., Matsuda, K., Ichihara, A,, Ishimura, K., Ogasawara, J., and Nagata, S. (1994). Fas-mediated apoptosis in primary cultured mouse hepatocytes. Exp. Cell. RES. 215, 332-337. Nicholson, D. W., Ah, A,, Thornberry, N. A,, \’aillancourt, J. P., Ding, C. K., Gallant, M., Gareau, Y., Griffin, P. R., Labelle, M., Lazebnik, Y. A,, et al. (1995). Identification and inhibition of the ICE/CED-3 protease necessary for inaminalian apoptosis. Natiire 376, 37-43. Nocentini, G.. Giunchi, L., Ronchetti, S., Kraosz, L. T., Bartoli, A,, Moraca, R., Migliorati, G., and Riccardi, C. (1997). A new member of the tumor necrosis factor/nerve growth factor receptor family inhibits T cell receptor-induced apoptosis. Proc. Natl. Acarl. Sci. USA 94, 6216-6221. O’Connell, J., O’Sullivan, G. C., Collins, J. K., arid Shanahan, F. (1996). The Fas counterattack: Fas-mediated T cell killing by colon cancer cells expressing Fas ligand. J , Exp. Met/. 184, 1075- 1082. Ochm, A, Dchrmann, I., Falk. W., Pawlita. M.. Mairr, 6..Klas. C., Li Weher, M . . KidiiircIs. S., Dhein, J., Trauth, B. C., et ol. (1992). Purification and molecular cloning of the APO-I cell surface antigen, a ineinber of the tumor necrosis factodnerve growth fiactor receptor superfamily: Sequence identity with the Fas antigen. J. B i d . Chem. 267,1070910715. Okura, T., Gong, L., Kamitani, T., Wada, T., Okura, I., Wei, C. F., Chang, H. M., and Yeh, E. T. (1996). Protection against Fas/APO-1- and tumor necrosis factor-mediated cell death by a novel protein, sentrin. J. Initnirnol. 157, 4277-4281. Oltvai, Z. N., Millirnan, C. L., and Korsmeyer, S. J. (1993). Bcl-2 heterodiinerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell 74,609-619. Orth, K., Chinnaiyan, A. M., Garg, M., Froelich, C. J., and Dixit, V. M. (1996a). The CED-3/ICE-like protease Mch2 is activated during apoptosis and cleaves the death substrate lamin A. J. B i d . Client. 271, 16443-16446. Orth, K., O’Rourke, K., Salvesen, G. S., and Dixit, V. M. (1996h). Molecular ordering of apoptotic mammalian CED-YICE-like proteases. J. Biol. Cheni. 271, 20977-20980. Owen Schaub. L. B., Yonehara, S., Crump, W. L., and Grimm, E. A. (1992).DNA fragmentation and cell death is selectively triggered in activated human lymphocytes by Fas antigen engagement. Cell. Imrnunol. 140, 197-205. Pai, J. T., Brown, M. S., and Goldstein, J. L. (1996). Purification and cDNA cloning of a second apoptosis-related cysteine protease that cleaves and activates sterol regulatory element binding proteins. Proc. Natl. Acarl. Sci. USA 93, 5437-5442. Pan, G., Ni, J., Wei, Y. F., Yu, G., Gentz, R., and Dixit, V. M. (1997a). An antagonist decoy receptor and a death domain-containing receptor for TRAIL. Science 277, 815-818. Pan, G., O’Honrke, K., Chinnaiyan, A. M., Gentz, H., Ebner, R., Ni, J., and Dixit, V. M. (1997b). The receptor for the cytotoxic ligand TRAIL. Science 276, 111-113.
Pennica, D., Nedwin. G. E.. Hayflick, J. S., Scebnrg. 1’. H.. Derynck, K.,Palladino, M. A,, Kohr. \Y. J., Aggawd, B. B., and Goeddel, D. \’. (1984). Human tumour necrosis factor: Precursor structure, expression and homology to lymphotosin. Nature 312, 724-729. Peter, M. E., Medeina, J. P., and Krainiiier, P. H. (1997).Does the Cmiorlinhditis elegntis protein CED-4 contain a regioii of Iioniolop to the maninialian death effector doinain? Cell Death Diff 4, 523-525. Peter, M. E., Scaffidi, C., Medema, J. P., Kisclikel, F. and Kraniintar, P. 11. (1998). The death receptors. I n “Apoptosis: Biolop, Mechanisnrs and Role in Disease” (S. Knniar, ed.). Springer, Heidelberg. Pfeffc,r, K., Matsnyama, T., Kundig, T. M., Wakc~liani,A,. Kishihara, K., Shahinian. A,, \Vieginaim, K., Ohaslii, P. S., Kronke, M.,and Mak. T. \V. (1993). Mice deficient for the 55 kd tumor necrosis factor receptor arc’ resistant to endotoxic shock, yet succumb to L. monocytogenes infection. Cell 73, 457-467. Pitti, R. M., Marsters, S. A,, Ruppert, S., Donahue, C. J., Moore, A , . and Ashkenazi, A. (1996). Indiiction of apoptosis by Apo-2 ligand, a new member of the tumor necrosis factor cytokine family. /. B i d . Clieni. 271, 12687-12690. Rasprr, D. M., Vaillanconrt, J. P., Hadano, S.. Houtzager, \?. M., Seiden, I., Keen, S . L. C., Tawa, P., Xantliondakis. S., Nasir, J., Martindale, D., Koop, B. F., Peterson, E. P.. Thornberry, N. A,, Huang. J. Q., Black. S. C., Honiung, F., Lenardo. M. 1.. Hayden, M. R.. Roily, S., and Nicholson, D. \f’. (1998). Cell death attenuation by “Usurpin”, a rnaininalian DED-caspase hoinologue that precludes caspase-8 recruitment and activation hy the CD95( Fas, APO-1) receptor complex. In press. Reed, J. C. (1994). Bcl-2 and the regulation ofprogramined cell death.]. Cell Biol. 124, 1-6. Reed, J . C.. Chtldy, M., Slal)iak, T., Crow, C;. M., and Nowell, P. C.(1988). Oncogenic potential of bcl-2 deinonstrated by gcw>transfer. Nriturc, 336, 259-261. Rieux Laucat, F.. Le Deist, F.. Hivroz, C.,Roherts. I. A,, Debatin, K. M., Fischer, A,, and de Villartay. J. P. ( 1995). Mutations in Fas associated with huinan lyinphoproliferative syndrome and autoimmunih. Science 268, 1347-1349. Rothe, J.. Lesslauer, W., Lotsclier. 11.. Lang, Y., Koebel. P., Kontgen, F., Althage, A,, Zinkrmagel, R.. Steinmetz, M.. and Bluethinann. If. (1993). Mice lacking the tumour necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listerio nionocytogenes. Nature, 364, 798-802. Rothstein, T. L., Wang, J. K., Panka, D. J., Foote, L. C . . Wang, Z. U., Stanger, B., Cui, H., JLI, S. T., and Marshak-Rothstein. A. (1995).Protection against Fas-dependent Tlilmediated apoptosis by antigen receptor engagement in B cells. Nature 374, 16C- 16-5. Rudel, T., and Bokocli, G. M . ( 1997). Menibrane and niorphological changes i n apoptotic cells regulated by caspase-mediate.c1activation of PAK2. Science 276, 1571-1574. Sakaliira, H., Enari. M.. and Nagata, S. (1998).Cleavage of CAD inhibitor in CAD activation and DNA degradation during apoptosis. Nntrtre 391, 96-99. Sato, T., lrie, S., Kitada, S., and Heed, J. C . (1995). FAP-1: A protein tyrosine phosphatase that associates with F a . Scienctj 268, 41 1-415. Scaffidi, C., Medenia, J . P., Krammer, P. lI., and Peter, M. E. (1997). FLICE is predominantly expressed as two functionally active isoforins. caspase-8h and caspase-8h. /. B i d . Chcm. 272, 26953-26958. Scaffidi. C . , Fulda. S., Srinivasan, A,, Friesen, C . , Feng, L., Tomaselli, K. J., Debatin. K.-M., Krammer, P. H., and Peter. M. E. (1998). Two CD95(APO-l/Fas) s i p d i n g pathways. E M B O /. 17, 1675-1687. Schall, T. J.. Lewis. M., Koller, K. J., Lee, A,, Rice, G. C.. Wong, G. H., Gatanaga, T.. Granger. G , A,, Lentz, R.. Raab, H., et (11. (1990). Molecular cloning and expression of a receptor for human tumor necrosis fiictor. Crll 61, 361-370.
206
PETEH 11. KKAMMEH
Schlegel, J,, Peters, I., Orrenius, S., Miller, D. K., Thornberry, N. A,, Yamin, T. T., and Nicholson, D. W. (1996). CPP3Wapopain is a key interleukin 1 beta converting enzymelike protease involved in Fas-mediated apoptosis. J. B i d . Cheni. 271, 1841-1844. Sclilossinan, S. F., Bouinsell, L., Gilks, W., €ladan, J. M., Kishinioto, T., Morinioto. C., Ritz, J., Shaw, S., Silverstein, R., Springer, T., Tedder, T. F., and Todd, R. F. (eds.) (1995). “Leucocyte Typing V: White Cell Differentiation Antigens.” Proceedings of the Fifth International Workshop and Conference, Vol. 1. Oxford Univ. Press, London. Schneider, P., Bodiner, J. L., Thoine, M., Hofiiiann, K., Holler, N., and Tschopp, J. (1997). Characterization of two receptors for TRAIL. FERS Lett. 416, 329-334. Schraven, B., and Peter, M. E. (1995). APO-l(CD95)-1riediated apoptosis in Jurkat cells does not involve src kinases or CD45. FEBS Lett. 368, 491-494. Screaton, G. R., Mongkolsapaya, J., Xu, X. N., Cowper, A. E., McMichael, A. J.. and Bell, J. I. ( 1997a). TRICK2, a new alternatively spliced receptor that transduces the cytotoxic signal from TRAIL. C u v . Biol. 7, 693-696. Screaton, G. R., Xu, X. N., Olsen, A. L., Cowper, A. E., Tan, R., McMichael, A. J. and Bell. J. I. (19Y7b). LARD: A new lyinphoid-specific death domain containing receptor regulated by alternative pre-niRNA splicing. Proc. Nod. Acad. Sci. USA 94, 4615-4619. Sedlak. T. W., Olhiai, Z. N., Yang, E., Wang, K., Boise, L. H., Thompson, C. B., and Korsmeyer, S. J. (1995). Multiple Bcl-2 family members denionstrate selective dimerizations with Bax. Proc. Nntl. Acad. Sci. USA 92, 7834-7838. Seino, K., Kayagaki, N., Okuniura, K., and Yagita, €1. (1997). Antitumor effect of locally produced CD95 ligand. Natzirc. hferl. 3, 165-170. Sentnian, C. L., Shutter, J. R., Hockenbery, D., Kanagawa, O., and Korsineyer, S. J. (1991). Bcl-2 inhibits multiple forms of apoptosis but not negative selection in thynocytes. Cell 67,879-888. Sheridan, J. P., Marsters, S. A,, Pitti, R. M., Gurney, A,, Skubatch, M., Baldwin, D., Ramakrishnan, L., Gray, C . L., Baker, K., Wood, W. I., Goddard, A. D., Godowski, P., and Ashkenazi, A. (1997). Control of TRAIL-induced apoptosis by a family of signaling and decoy receptors. Science 277, 818-821. Shirai, T., Yamaguchi, H., Ito, H., Todd, C. W., and Wallace, R. B. (1985). Cloning and expression in Eschericliin coli of the gene for human tiimour necrosis factor. Nature 313, 803-806. Shu, H. B., Halpin, D. R., and Coeddel, D. \’. (1997). Casper is a FADD- and caspaserelated inducer of apoptosis. Ztni)iunity 6, 751-763. Singer, G. G., and Abbas, A. K. (1994). The Fas antigen is involved in peripheral but not thymic deletion ofT lymphocytes in T cell receptor transgenic mice. Inintunity 1,365-371. Smith, C. A., Davis, T., Anderson, D., Solam, L., Beckmann, M. P., Jerzy, R., Dower, S. K., Cosman, D., and Goodwin, R. G. (1990). A receptor for tumor necrosis factor defines an unusual fainily of cellular and viral proteins. Science 248, 1019- 1023. Smith, C. A,, Gruss, €I. J., Davis, T., Anderson, D., Farrah, T., Baker, E., Sutherland, G. R., Brannan, C. I., Copeland, N . G., Jenkins, N . A,, et nl. (1993). CD30 antigen, a marker for Hodgkin’s lymphoma, is a receptor whose ligand defines an emerging family of cytokines with lioinology to TNF. Cell 73, 1349-1360. Smith, D. J., McGuire, M. J., Tocci, M. J., and Thiele, D. L. (1997). IL-1 beta convertase (ICE) does not play a requisite role in apoptosis induced in T lymplioblasts by Fasdependent or Fas-independent CTL effector mechanisms. /. Inimunol. 158, 163- 170. Song, Q., Lees Miller, S. P., Kuniar, S., Zhang, Z., Chan, D. W., Smith, C . C.. Jackson, S. P.. Alnemri, E. s., Litwack, G., Khanna, K. K., and Lavin, M. F. (1996). DNAdependent protein kinase catalytic subunit: A target for an ICE-like protease in apoptosis. E M B O J. 15, 3238-3246.
Song, Q.,Wei, T., Lres M i k r , S., Aheinri, E., \.’l”itters, D.. and Lavin, M. F. (1997). Resistance of actin to cleavage dining apoptosis. Proc. NotI. Accrd. Sci. USA 94, 157-162, Srinivasula, S. M., Ahmad, M., Feinaiitles-Alnenrri. T., Litwick, G., arid Alnemii. E. S. ( 19963). Molecular ordering o f the Fas-apoptotic pathway: The Fas/APO- 1 protcase Mchfj is a CriiiA-inhil)itable protease that activates multiple Ced-YICE-like cysteirie proteases. Proc. Nafl. A c sci. ~ USA 93, 144M-14491. Srinic-as&, S. M.. Ahmid, M., Ottilie, S., Biilliicli, F., Banks. S., Wang, Y.. FernandesAlneniri, T.. Croce, C. M.. Lihvack. G.. Toniaselli. K . J.. Armstrong, R. C., and A h e m r i , E. S. (1997). FLAME-1, a novel FAIIDlike anti-apoptotic inolecnle that regulatcas Fad TNFR1-induced apoptosis. J . Riol. Chcvtr. 272, 18.542-18545. Srinivasrila, S. M., Femandes-Alnemri, T., Zangrilli, J., Roljertson, N., Armstrong, K. C . , Wang, L., Trapmi. J. A., Toniaselli, K.J., Lihvack, G., and Alnemri, E. S. (19961)).The Crd-:3/interleukin lheta converting enzynie-like tionlolog Mch6 and the lamin-cleaving enzyme Mcli2alpha are substrates for the apoptotic n d i a t o r CPP32. /. Riol. Clrrrri. 271, 27099-27106. Stahnkc, C., Hecker. S.. Koline, E.. aiicl Del)atin. K . - M . (1997).Expression of the APO-1 antigen and induction of apoptosis iii hematopoietic cells. Exp Hemutol. (in press). Stamenkovic, I., Clark, E. A., antl Seed, B. ( 1989). A B-lymphocyte activation molecule related to the neive growth fictor rcwptor and indnced hy cytokines in carcinoinas. EMBO /. 8, 1403-1110. Stanger, B. Z.. Leder, P.. Lee. T. [ I., Kim, E., a i d Seed, B. (1995). KIP: A novel protein containing a death domain that interacts mitli Fas/APO-l (CD95) in yeast and causes cell death. C d 81, 513-523. Strand, S., Ilofinann, \I7.J., Hug, H., Miiller, M . , Otto. G., Strand, D., Mariani, S. M., Strcininel, W., Kraniiner, P. H.. and G d k ~ P. , R. (1996). Lymphocyte apoptosis induced by CD95 (APO-UFas) ligaiid-expressing tumor ct,lls: A niechanism of iiriiiinne evasion? Nutiire A 4 ~ t l .2, 1361-1366. Strasser. A , . Harris, A. N’..Bath, M. I,., and Cory. S. (1990). Novel primitive lyniphoid tumours induced in transgenic mice by cooperation behveen niyc and bcl-2. Nature 348, 331 -333. Strasser, A , , Harris. A. W., and Cory, S. (lY9la). Bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship. Cell 67, 889-899. Strasser. A , . EIarris, A. W., Iluang, D. C.. Krammer, P. € I . , and Cory, S. (1995). Bcl-2 antl Fas/APO- 1 regulate distinct pathways to lymp~iocyteapoptosis. EMBO 1.14, 6136-6147, Strasser, A,, \Yhittingham. S., Vaux, D. I,.. Bath, M. LA., Adams, J. M., Cow, S.,and Harris, A. M’.(1991h).Enforced Bcl-2 expression in B-lyiiiphoid cells prolongs antibody responsrs and elicits autoimmune disease. Proc. h’atl. A c ~ d .Sci. USA 88, 8661 -8665. Snda. T., Hashinioto. Il., Tanaka, M.. Oclii, T., m t l Nagatn, S. (1997). Mernbrane Fas ligand kills hriman periphcral blood T Iyniphocytes. and soluble Fas ligand blocks the killing. J . Eq). M i d . 186, 2045-2050. Suda, T.. Takaliaslii, T., Golstein, P., and Nagata, S. (1993).Molecular cloning and expression of the Fas hgand, a novel meinber of the tunior necrosis factor farnily. Cell 75, 1169-1 178. Suda, T., Tanaka, M., Miwa, K., and Nagata, S. (1996). Apoptosis of I I I O U S ~naive T cells iiidiiced by recombinant soluble Fas ligand and activation-induced resistance to Fas ligand. /. Z t i i r t i u t i d . 157, ,3918-3924. Nuiiez. C . , Clarke, M. F., and Wicha, M. S. (1995). Sumantran, V. N., Ealovegy, M. N‘.. Overexpression of Bcl-x, sensitizes MCF-7 cells to cheniotIierap~-indrlcetl apoptosis. Corrccv- Rcs. 55, 2507-2510. Sytwri. H. K., Liblau, R. S., and McDevitt. € I . 0. (1996). The rolcs of Fas/APO-l (CD95) and TNF in antigen-indiiced prograninird cell death in T cell receptor transgenic mice. Ztiirtmtiity 5, 17-30.
208
PETER H. KRAMMEH
Takahashi, T., Tanaka, M., Brannan, C. I., Jenkins, N. A,, Copeland, N. G., Suda, T., and Nagata, S. (1994d. Generalized lyinphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76, 969-976. Takahashi, T., Tanaka, M.. Inazawa, J., Abe, T., Suda, T., and Nagata, S. (1994b).Human Fas ligand: Gene structure, chromosomal location and species specificity. Znt. Zmmfmol. 6, 1567-1574. Takayama, S., Sato, T., Krajewski, S., Kochel, K., Irie, S., Millan, J. A., and Reed, J. C. (1995). Cloning and functional analysis of BAG-1: A novel Bcl-2-binding protein with anti-cell death activity. Cell 80, 279-284. Tanaka, M., Itai, T., Adachi, M., Nagata, S. (1998). Downregulation of Fas ligand by shedding. Nature Med. 4, 31-36. Tanaka, M., Suda, T., Haze, K., Nakamura, N., Sato, K., Kimura, F., Motoyoshi, K., Mizuki, M., Tagawa. S., Ohga, S., Hatake, K., Dminmond, A. H., and Nagata, S. (1996). Fas ligaiid in human serum. Nature Med. 2, 317-322. Tanaka, M., Suda, T., Takahashi, T., and Nagata, S. (1995). Expression of the functional soluble form of human fas ligand in activated lymphocytes. EMBO J . 14, 1129-1135. Tanaka, M., Suda, T., Yatomi, T., Nakamura, N., Nagata, S. (1997).Lethal effect of recombinant human Fas ligand in mice pretreated with Propioriibncteriutn ncnes. J , Zmtnund. 158, 2303-2309. Tartaglia, L. A,, Ayres, T. M., Wong, G. H., and Goeddel, D. V. (1993). A novel domain within the 5S kd TNF receptor signals cell death. Cell 74, 845-853. Tepper, C. C., Jayadev, S., Liu, B., Bielawska, A,, Wolff, R., Yonehara, S., Hannun, Y. A., and Seldin, M. F. (1995). Role for ceramide as an endogenous mediator of Fas-induced cytotoxicity. Proc. Nntl. Acacl. Sci. USA 92, 8443-8447. Tewari, M., Beidler, D. R., and Dixit. V. M. (1995a). CrniA-inhibitable cleavage of the 70-kDa protein component of the U 1 small nuclear ribonucleoprotein during Fas- and tumor necrosis factor-induced apoptosis. J . Biol. Chem. 270, 18738-18741. Tewari, M., and Dixit, V. M. (1995). Fas- and tumor necrosis factor-induced apoptosis is inhibited by the poxvirus crmA gene product. J . Biol. Cheni. 270, 3255-3260. Tewari, M., Quan, L. T., O’Rourke, K., Desnoyers, S., Zeng, Z., Beidler, D. R., Poirier, G. G., Salvesen,G. S.,and Dixit, V. M. (199%).YamdCPPB2 beta, a mammalian homolog of CED-3, is a CrniA-inhibitable protease that cleaves the death substrate poly(ADPribose) polymerase. Cell 81, 801-809. Thome, M., Schneider, P., Hofniann, K., Fickenscher, H., Meinl, E., Neipel, F., Mattmann, C., Bums, K., Bodmer, J. L.. Schroter, M., Scaffidi, C., Krammer, P. H., Peter, M. E., and Tschopp, J. (1997).Viral FLICE-inhibitory proteins (FLIPS)prevent apoptosis induced by death receptors. Nature 386, 517-521. Thornberry, N. A., Bull, H. 6.,Calaycay, J. R., Chapman, K. T., Howard, A. D., Kostura, M. J., Miller, D. K., Molineaux, S. M., Weidiier, J. R., Aunins, J., et al. (1992). A novel heterocheric cysteine protease is required for interleukin-1 beta processing in inonocytes. Nature 356, 768-774. Trauth, B. C., Klas, C., Peters, A. M., Matzku, S., Moller, P., Falk, W., Debatin, K. M., and Krammer, P. H. (1989).Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301-305. Tsujimoto, Y., Finger, L. R . , Yunis, J., Nowell, P. C., and Croce, C. M. (1984).Cloning of the chromosome breakpoint of neoplastic B cells with the t( 14; 18) chromosome translocation. Science 226, 1097- 1099. Uhich, A., Gray, A., Berinan, C., Coussens, L., and Dull, T. J. (1983). Sequence homology of human and mouse beta-NGF subunit genes. Cokl Spring Harb. Symp. Quant. B i d . 48,435-442.
\'an I'arijs, L., Ibraghimov, A.. and Ahbas, A. K. (1Y96).The roles of costiniulation and Fas in T cell apoptosis ant1 peripheral tolrrance. I r t i t t m ~ t i t ! y4, 321-328. Vanhaesebroeck. B., Reed. J. C . , De \'ulck, D., Grooten, J.. Miyasliita. T.. Tanaka, S., Beyieit, R.. k i n Roy, F., and Fiers, I t ' (199:3).Effbct of bcl-2 proto-oncogene expression on c e h h seiisitibity to tumor ntwosis f'~ictor-niediatt,dcytotoxicity. 0rrc:og:unu 8, 10751081. Vans, 11. L., Aguila. H. L.. and \\'eissinm, 1. L. i 1992). Bcl-2 prevents death of factordeprived cells Init fails to prevent apoptosis i i i targets ofcc.11 inetliated killing. h i t . Ir,ir,,unol. 4, 821-824. Vaiis. D. L., COT, S.. and Atlains, J. M . (1988). Bcl-2 gene promotes haemopoietic cell sunival and cooperates \Jith c-niyc to iniinort;tlize pre-€3 cells. Noturr 335, 440-442. Vincenz, C..and Dixit, V. M . (1997). Fas-associated tleath domain protein interleukinlbclta-converting enzyme 2 [ FLICE?), iin ICE/C:etl-:3 Iiomologue, is proxiinally involvetl in C1195- and p5~5-niediateddeath signaling. ]. H i d . Clicwi. 272, 6578-6583. Vogt, C:. (1842). Untersuclinngen ii1x.r (lie EntMicklungsgcsclrichte tler Geblirtslielferkroete (Alytes obstetricans). Walczak, H., Degli-Esposti, M . A., Jo1inso11,I{. S.,Sniolak, 1'. J., LVaiigh, J. Y., Boiani. N., Tiiiiour, M. S.. Gerhart, M . J., Sclioolry, K. A.. Smith. (;. A,, Coodwin, R. G., antl Raucli, C . T. (1997). TRAIL-R2: A novel apoptosis-iiic,dioting receptor for TRAIL. E M B O J. 16, 5386-5397. Walker. N. P., Taliinian, R. \'., Brady. K. D.. Dang, I,. C . , Bunip, N. J.. Ferenz, C. R., Franklin. S., Ghayir. T., Hackctt, M . C.. Ilammill, L. D., ct (11. (1994). Clystal structure of the cysteine protease interleukin- 1 beta-converting enzyme: A (p2O/p10)2 homodinier.
C d l 78, 34:3-352. Wang, A. M., Crease?. A. A . , Ladner, h.1. L3.. Lin, L. S., Strickler. I., Van Arstlell, J. N., Yarnanloto, R., and Mark, D. F. (1Y81j).Moleciikir cloning of the coniplrnientar). DNA for huinan tumor ncwosis fktor. Scicwcc, 228, 149- It54. Wang, L., Miura, M., Bergeron. L., ZIin, H.. antl Y u m i , J. (1994). Ich-1, an Ice/ced-3related gene, encodes both positive m t l negative rc~gnlatorsof programnied cell death. C r d 78, 739-750. Wang, S., Minra, M.. Jnng, Y. K . , Zhu, H., Gagliardini, V., Shi, L., Greenberg, A. I I . , and Yum. J. (1996). Identification and characterization of Ich-3. a nienil)erof the interleukin11)rta converting enzyme ( ICE)/Ced-:3 family and iin upstreani regulator of ICE. ]. Biol. Chcvri. 271, 20580-20587. Wang. X., Pai, J. T.. LViedenfeld, E. A , . M e t h i , J. C.,Slaughter, C. A , , Goldstein, J. L., ant1 Brown, M. S.(1995). Purification o f an interleukin-1 beta converting cnzyine-related cysteine protease that clcaives sterol regulator?. eleinent-l,intling proteins between thc leucine zipper and transmenibrane tioniains. ]. BkJ!. chclri. 270, 18044-18050. Watariabe-Fukuiiaga. R.. Brannan. C . I . , Itoh, N., Yoneliara. S.,Copeland, N . G., Jenkins, N. A,, and Nagata, S. i 1992). The cDNA structnrc., expression, and chromosomal assignment of thr mouse Fas antigen. ]. Z r r t r r i i ~ r i c ~ !148, . 1274- 1279. Waterhousr, N., Knniar, S.. Song, Q., Strike, l'., Sparrow, L., Dreyfuss. G., Alneniri, E. S., Litwack, G., L a i n , M., and Watters. D. (1996). Ileteronnclear ribonucleoproteins C1 and C 2 , coniponeiits ofthe spliceosome, are specific targets of interleukin 1beta-converting . 271, 293:35-29341. enzyine-like proteases in apoptosis. ]. B i ~ l Chrw. Wrstendorp, M. O., Frank, R., Oclisenbaner, C . , Stricker. K., Dhein, J., Walczak. H.. Debatin, K. M., and Kraminer, P. IT. (1995). Sensitization of T cells to CD9t5-~nediated apoptosis by HIV-l Tat ;tiid gp120. Nafirrr, 375, 49-500. Lt'iley, S.R.. Schooley, K., Sniolak, P. J.. Din, \Y. S., Hiiang, C. P., Niclioll. J. K., Sritherlantl, G. R.. Smith, T. D., I kic h, C..Smith, ( z , A,. c't n!. (1995). Identification and characterization of a new nieniber of the TNF family tliat intliicrxsapoptosis. Irnniimit!y 3,673-682.
210
PETER 11. KHAMMER
Wilson, K. P., Black, J. A., Thoinson, J. A,, Kim, E. E., Griffith, J. P., Navia, M. A.. Murcko, M. A,, Chambers, S. P., Aldape, R. A., Raybuck, S. A,, et a!. (1994). Structure and mechanism of interleukin-1 heta couverting enzyme. Ncitnre 370, 270-275. Wissiug, D., Mouritzen, H., Egeblatl, M., Poirier, G. C,,and Jaattela, M. (1997).Involvement of caspase-dependent activation of cytosolic phospholipase. A2 in tumor necrosis factorinduced apoptosis. Proc. Nntl. Acad. Sci. USA 94, 5073-5077. Wright, D. A,. Futcher, B., Ghosh, P., and Ceha, R. S. (1996). Association of human Fas (CD95) with a ~ibiquitin-conjugatingenzyme (UBC-FAP). J , Biol. Cketn. 271, 3103731043. Wu, M., Lee, [I,, Bellas, R. E., Sch;uier, S. L., Arsura, M., Katz, D., FitzCerdld, M . J., Rothsteiu, T. L., Sherr, D. H., and Sonenshein, G. E. (1996). Inhibition of NF-kappaB/ Re1 induces apoptosis of niurine B cells. EMBO J. 15, 4682-4690. Y a p , H., Hanabuchi. S., Asano, Y., Tamura, T., Nariuchi, H., and Okurnura, K. (1995). Fas-mediated cytotoxicity-A new ii111~1~11~oregul~~to~ and pathogenic function of T h l CD+ T cells. In~ttwnol.Heti. 146, 223-239. Yang, E., Zlia, J., Jockel, J., Boise, L. H., Thoiupson, C. B.. and Korsmeyer, S. J. (1995). Bad, a heterodiiueric 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., Ihrado, A. M., Cai, J., Peng, T. I., Jones, D. P., and Wang, X. (1997a). Preventiou of apoptosis by Bcl-2: Release of cytoclirome c from mitochondria blocked. Scie~tce275, 1129- 1132. Yang, X., Khosravi-Far, R., Chang, H. Y., and Baltimore, D. (1997h). Dawx, a novel Fasbinding protein that activates J N K and apoptosis. Cell 89, 1067-1076. Yasuda, H., Shima, N., Nakagawa, N., Yamaguchi, K,,Kinosaki, M., Mochizuki, S.,Tomoyasu, A,, Yano, K., Goto, M., Murakami, A,, el n/. (1998). Osteoclast differentiation factor is a ligand for osteoprotegerin/osteocl~~stoge~~esis-inhibit~~~ factor and is identical to TRANCE/RANKL. Proc. N d . Acnd. Sci. USA 95, 3597-3602. Yin, X. M., Oltvai, Z. N., Veis Novack, D. J., Liuette, G. P., and Korsmeyer, S. J. (1994). Bcl-2 gene family and the regulation of prograniined cell death. Cold Spring Hurl]. Syntp. Qncint. B i d . 59, 387-393. Yonehara, S., Ishii, A,, and Yoirehard, M. (1989). A cell-killing monoclonal antibody (antiFas) to a cell surface receptor antigen co-dowiregulated with the receptor of tumor necrosis factor. J. Exp. M e d 169, 1747-1756. Yosliihiro, K., Zhou, Y. W., Zhang, X. L., Cheu, T. X., Tanaka, S., Azruna, E., and Sakurai, M. (1997). Fas/APO-l(CD95)-mediated cytotoxicity is responsible for the apoptotic cell death of leukaemic cells induced by interleukin-2-activated T-cells. Br. J . Hac.mntol. 96, 147-157. Yuan, J., Shaharn, S., Ledonx, S., Ellis, H. M., and Horvitz, H. R. (1993). The C. elegutts cell death gene ced-3 encodes a protein similar to niammalian interleukin-1 beta-converting enzyme. Cell 75, 641-652. Zheng, L., Fisher, G., Miller, R. E., Peschon, J., Lynch, D. H., and Lenardo, M. J. (1995). Induction of apoptosis in mature T cells by tumour necrosis factor. Nrrture 377,348-351. Zhivotovsky, B., Cedewall, B., Jiang, S., Nicotera, P., and Orrenius, S. (1994). Involvement of Ca'+ in the formation of high molecular weight DNA fragments in thymocyte apoptosis. Biochetn. Biophys. Re,s. Conitnun. 202, 120-127. Ziininerman, C., Rrdusclia Riem, K., Blaser, C . , Zinkemagel, R. M., and Pircher. H. (1996). Visualization, characterization, and turnover of' CD8+ memory T cells in virus-infected hosts. J , Exp. Med. 183, 1367-1375. This article was accepted for publication on March 24, 1998.
A CXC Chemokine SDF- 1 /PBSF: A Ligand for a HIV Coreceptor, CXCR4 TAKASHI NAGASAWA, KAZUNOBU TACHIBANA, AND KENJI KAWABATA Department of Immunology, Research Institute, Osoka Medico1 Center for Maternal and Child Healih, /zumi, Osaka 594- I I0 I, Japan
1. Introduction
The clieniokines are a large fimily of sniall, structurally related 8- to 10-kDa cytokines that have four conserved cysteines forming two disulfide bonds (reviewed in Baggiolini et nl., 1994, 1997; Oppenheiin ct a1 , 1991). There are two subgroups according to tlie arrangement of the first two cysteines, wliicli are either adjacent (CC subfamily) or separated by one amino acid (CXC subfamily). The CC subfamily includes nionocyte clieinoattractant protein-1 ( MCP-l), macropliagc inflainmatory peptidela (MIP-la),RANTES, and eot'ixin. The CXC subfamily includes interleukin ( IL)-8, platelet f'ictor-4 (PF4), and interferon-y-inducible protein (IP-10). As most members of clieniokines liave been reported to be cliemotactic for leukocytes and produced in rcsponse to proinflaminatory agents, they have been designated inducible inflammatory mediators that dictate leukocyte activation and cheinotaxis to the site of inflammation. Administration of an anti-CXC chemokine reagent was sliown to prevent lung reperfiision injury in rabbits, demonstrating that cheinokines play a critical role in pathology (Sekido ef nl., 1993). Gene targeting studies have revealed that MIP-la is required for tlie pathogenesis of coxackievirus-induced myocarditis and influenza viriis-induced pneumonitis (Cook ct d . , 1995). A CXC cliemokine receptor, CXCR2, and CC cliemokine receptors CCRl and CCK2 were shown to be responsible for neutrophil-mediated acute inflammation (Cacalano cf al., 1994) and granuloma formation, a neutrophilmediated host defense against a fungus (Gao et nl., 1997), and clearance of infection by intracellular bacteria (Kuri1i:ira cf d., 1997; Boring et a1 , 1997). However, some cheinokines are thought to have additional functions, including hematopoiesis and angiogenesis. MIP-la was reported to inhibit the growth of heinatopoietic stein cells in vitro (Graham et al., 1990). CXC chemokine, PF4, IL-8, IP-10, and Crop have also been shown to be angiogenic regulators in vitro (reLiewed in Baggiolini et d., 1997). However, their physiological roles have not yet been defined. In 1995, CXCR2 were reported to be negative regulators of niveloid progenitor cells mainly in the preyence of environmental stresses such as bacteria and bacterial products (Cacalano et al., 1994). 21 1
< Il*'\"gl,t
0 I Y Y Y I,\ 4<.Ul<,,,I< Prmr
ill n I Ir,""'l,l
Ionn I., I)lIhj 377h/')Y 630 (XI
111 11,)
212
TAKASHI NA(:ASAWA
t'f
o/
Of interest is the discovery that a member of the CXC chemokine, SDF-l/PBSF, has been shown to be essential for development, including hematopoiesis and cardiogenesis (Nagasawa et al., 1996a). Moreover, a receptor for SDF-l/PBSF, CXCR4, was found to be essential for HIV-1 entry into host cells (Feng et al., 1996). Although infectious agents such as cytomegalovirus (CMV),Herpesvirus sainiiri (HVS), plasrnodiiim vivax, and human herpesvirus 8 (HHV-8) possess open readmg frames encoding the functional chemokine or chemokine receptors (reviewed in Murphy, 1994; Boshoff et al., 1997; Arvanitakis et al., 1997), direct involvement of endogenous chemokines or chemokine receptors in microbial pathogenesis had not been shown. Thus, these findings widened the area of chemokine biology and pathology. This article summarizes advances in SDF- l/E'BSF and CXCR4 research so as to provide a framework for understanding their biological and pathological functions. II. Identification, Structure, and Expression of CXC Chemokine SDF-1/PBSF
A. IDENTIFICATION The murine SDF-1/PBSF was identified independently by two groups of investigators. Tashiro et al. (1993) developed a method to clone cDNAs that carry specific amino-terminal signal sequences, such as those encoding intercellular signal-transducing molecules and receptors, and identified a cDNA of SDF-1/PBSF using a cDNA library from bone marrow-derived stromal cell line ST2, without the use of specific functional assays. Nagasawa et al. (1994)identified the cDNA as growth-stimulating activity for a stromal cell-dependent pre-B-cell clone DW34 by expression cloning strategy using a cDNA library constructed from the bone marrow-derived stromal cell line PA6.
B. GENEAND GENEPIiODUCT The mRNA of SDF-1/PBSF encodes a 89 amino acid protein, comprising a 21 amino acid-cleaved signal peptide (Fig. 1).A form that had been processed at the C-terminal end to generate a 67 residue protein was purified (Bleul et al., 1996b) (Fig. 1).The amino acid sequence of the mature proteins contains four cystein residues conserved among CXC chemokines. Although SDF- l/PBSF belongs to the chemokine family, there are some unique properties of gene and gene products among the members. First, the stnictiire similarity diagram of human CXC chemokines determined by the average linkage cluster analysis revealed that SDF-1/PBSF belongs to a different group than other CXC chemokines (Baggiolini et al., 1997; Bleul et al., 1996b). Second, the amino acid sequence of the SDF-1/PBSF mature protein is highly conserved be-
43
1
hSDF-la/PBSF
KPVSLSYR CPC RFFESH VARANVKHLKILN TPN C ALQIVARLK
hSDF-lP IUSDF-~~
................. I............ ............. ........................................... ................. I............ .............
hSDF-lCt/PBSF
NNNRQV C IDPKLKWIQEYLEKALNK
m8DF-lWPBSF
.........................
mSDF-la/PBSF
44
hSDF-1P
IUSDF-~P
....... ..................RFKM ..........................
68
L..
Fic:. 1. Amino acid sequences of hriinan and murine SDF-1dPBSF and SDF-1P. A form that had been processed at tlie C-terminal end (star) to generate a 67 residue protein was purified (Bled et al., 1996).
tween mouse and human (99%) (Fig. 1)coinpared with other chemohnes, MCP-1 ( 5 5 % ) or MIP-la (75%) (reviewed in Baggiolini e t a ! . , 1994).The murine counterpart of 1L-8 has not been identified, probably as a result of its low interspecies homology. Third, there is an alternatively spliced form of SDF-ldPBSF, designated SDF-lp. The proteins of S D F - l d PBSF and SDF-1P have identical amino acid residues, and the SDF-1p protein has an additional last four residues in the carboxy terminus (Fig. 1). The human and murine SDF-la/PBSF gene consists of three exons, and human SDF-1P consists of four exons (Shirozu et d., 1995; Nagasawa et al., 1996a). The human SDF-10 gene cDNA shares the sequence of the first exon, the second exon, and 87 bases of tlie third exon spliced to the fourth exon (Shirozu et al., 1995). Finally, the gene of SDF-l/PBSF is localized on human chromosome 1Oq (Shirozu et al., 1995) and on inurine chromosome 6 (Noinura et al., 1996), although most of the other CXC cliemokines and CC chemokines are located close together in the human chromosomes 4q and 17q, respectively. C . EXPRESSION Previous studies have revealed that the expression of SDF-l/PBSF mRNA is constitutive whereas that of other chemolunes is inducible. SDF-1/PBSF inRNA is expressed in bone marrow-derived stromal cell lines (PA6, ST2) (Tashiro et al., 1993; Nagasawa et d . ,1994) arid in inany organs in adult mice (brain, thymus, heart, lung, liver, ladney, spleen, stomach, intestine, bone marrow), but not in the murine hematopoietic progenitor cell line (LyD9) (Tashiro et a!., 1993), murine pre-B-cell line
214
TAKASHI ILA<;ASAWA
c’f nl
(702/3), and clone (DW34) (T. Nagasawa, unpublished data). It is also expressed during embryogenesis, e.g., in brain, liver, heart, and bone marrow spindle-shaped stroinal cells (Nagasawa et al., 1996a; T. Nagasawa, unpublished data). D. THREE-DIMENSIONAL STHUCTURE A N D STRUCTURE-ACTIVITY RELATIONS The three-dimensional solution structure of SDF-l/PBSF has been determined by nuclear magnetic resonance ( N M R ) spectroscopy (Crump et al., 1997) (Fig. 2). SDF-1RBSF and SDF-10 are monomers and adapt a chemokine-like fold consisting of an N-terminal region, a loop region that follows the CXC motif, three antiparallel P-strands, and an overlymg C-terminal a helix (Crump et al., 1997). N-terminal eight residues and residues 12-17 (RFFESH) formed a receptor-binding site, and Nterminal two residues comprised the receptor activation motif of SDF-1/ PBSF. In all CXC chemokines studied previously, the sequence Glu-LeuArg (ELR motif), which precedes the first cysteine, is essential for both receptor binding and functional activation (Clark-Lewis et al., 1994).
HOOC
Frc:. 2. Thr striictrire of SDF-1RBSF. SDF-UPBSF consists of an N-terminal region, a loop region that follows the CXC motif, three antiparallel P-strands, and an overlying Cterminal a-helix.
In contrast to those CXC chemokines, SDF-l/PBSF lacks the ELR motif, and the arginine preceding the first cvstein is not absolutely required (Crump et d ,1997). 111. Physiological Functions of SDF-1 /PBSF
A. DE\.EI.OPMENT
The high constitutive expression of SDF-I/PBSF inRNA in embryos raised the possibility that it plays a role in de\~elopment.Mutant mice with targeted disruption for SDF-l/PBSF were sliown to die during embryogenesis (Nagasawa et 01, 1996a). They were present at the expected ratios initil day 15.5 of embryogenesis (E 15.5). However, about half the SDF-1/PBSF -1- embryos were foiind dead at E18.5, and all SDF-1/ PBSF -/- neonates die'd within an hour. Although the cause of their death remains to be elucidated, these results indicate that SDF-1/PBSF is essential for the viability of the embiyo.
B . H E hi ATOH) I E 5 IS The essential functions of SDF-l/PBSF in hematopoiesis have been indicated using fetal liver and bone inarrow of mutant embryos lacking SDF-1FBSF (Fig. 3 ) (Nagasawa ef d.,1996a). Fetal liver is the major liematopoietic organ from around E l 1.5- 12.ijuntil the first postnatal week. In the E18.5 SDF-UPBSF -/- fetal liver, the numbers of pro-B (B220' CD43+)and pre-B (B220' CD43-) cells, which were detected readily in E15.5-18.5 wild-type mice, were reduced severely, although the iiiirnbers of granulocytes ( G r - l + )and monocytes (CDllb' CDl8') in the E18.5
AGM
/\
-....,..........-..... """..L
Yolk sac
1
Bone marrow
/
myelopoiesis
B lymphopoiesis
B lymphopoiesis F I C3. Involvement of SDF-I/PBSF in hciiiatopoiesis during embryogmesis. Hematopoietic sites arc successively active during embryogenesis. SDF-I/PBSF is esseiitial for H lymphopoiesis in liver and bone marrow, iiiid myrlopoiesis in bone inarrow.
216
TAKASIII NAGASAM'A ut ol
mutants were unaltered. The numbers of B-cell progenitors that yield preB-cell colonies (CFU-IL-7) were also reduced in E18.5 mutant mice, although the numbers of primitive, erythroid, and myeloid progenitors that yield erythroid-myeloid mixed colonies (CFU-Mix),erythroid colonies (CFU-E), and inyeloid colonies (CFU-GM), respectively, were not reduced. These results indicate that the involvement of SDF-1/PBSF in fetal liver hematopoiesis is B lymphoid specific. In bone inarrow cavities, where a large number of hematopoietic cells, most of which were developing granulocytic cells, were identified easily in control embryos from E17.5, these cells were virtually absent in E18.5 SDF-l/PBSF -/- embryos. The structure of cortical bones and trabeculae, the formation of bone inarrow cavities, and the number of osteoblastic cells lining the cortical bone and trabeculae were normal in mutants. Consistent with that, the numbers of CFU-Mix, CFU-GM, and CFUIL-7 were reduced greatly in the SDF-1/PBSF -/- bone marrow in comparison with control littermates. Thus, the mutation of SDF-lFBSF eliminated myelopoiesis as well as B lymphopoiesis in the bone marrow. Because the expression of SDF-l/PBSF inRNA was observed in spindleshaped stroinal cells that could be distinguished from osteoblastic cells and hematopoietic cells, SDF-1/PBSF produced by stromal cells may transfer a signal to the hematopoietic progenitor cells in the bone marrow. First, these data indicate that SDF-1/PBSF is an essential cytokine for B lymphopoiesis during embryonic development (Fig. 3). IL-7 is the first example of cytokines that have been shown to be essential for B lymphopoiesis in zjivo (Grabstein et al., 1993; Peschon et al., 1994; Sudo et al., 1993; von Freeden-Jeffry et al., 1995). Although disruption of IL-7 affects the expansion of pre-B cells but not pro-B cells (von Freeden-Jeffry et al., 1995), both pro-B and pre-B cells were reduced in SDF-l/PBSF-deficient mice, representing an earlier developmental defect. Second, SDF-WBSF was shown to be required for myelopoiesis in the bone marrow but not in the fetal liver. Heinatopoietic precursors are thought to migrate from other sites of hematopoiesis, such as liver and the aorta-gonad-mesonephros (AGM) region and colonize the bone inarrow during embryogenesis (Fig. 3).SDF-WPBSF is the first candidate of cytokines that supports colonization of the bone marrow by hematopoietic precursors as a microenvironmental factor. Consistent with that, SDF-l/PBSF was reported to be a chemoattractant for human CD34' hematopoietic progenitors, as described later (Aiuti et al., 1997). C. CARDIOGENESIS Although target cells of most chemokines have been shown to be blood cells, the study using knockout mice indicated that SDF-lFBSF is also
responsible for cardiogenesis (Nagasawa ct ril., 1996a). SDF-l/PBSF -1- embryos had a defect of tlie nieiiibraiioils portion of ventricular septum at E 13.5- 18.5, although tlie aorta antl pulmonary artery outflow tracts and valves were normal. ZII ~itrr hyhridization revealed that tlie expression of SDF-l/PBSF inRNA was detected in the endocardium of the niusciilar ventiicular septuin at E 12.5, wlien the membranous portion of ventricular septum is soon to I)r formed, suggesting a role of SDF-1/ PBSF in ventricular septum formation. A ventiiciilar septum defect was also observed in mice lacking a cytokine, endotlielin-I, and a retinoitl receptor, KXR-a (Kastner ct d., 1994; Kiiriliara ct d . , 1995; Sucov ct ol., 1994). Endotlielin-l-deficient mice also have defects of out flows from the heart, and endothelin-1 is essential for tlie development of neural crestderived tissue, suggesting that ii disturbance in neural crest cell lineage affects carchogenesis (Kurihara rt al., 1995). How SDF-l/PBSF, entlothelin-1, and retinoic acid cooperate in tlie forination of tlie ventriciilar septum is unknown. ACTIVITIESo~ SDF-l/PBSF O N CIt a/., 1997). The result that chemota?.is of CD34’ lieinatopoietic progenitors in response to SDF-1/PBSF is higher in cells from bone marrow than from peripheral blood (Aiuti et d . , 1997) supports the idea that SDF-l/PBSF plays a role in the homing of heinatopoietic cells. On the other hand, SDF-l/PBSF was shown to induce the migration ofmicroglial cells and astrocvtes in the brain (Tanabe ct d . ,1997). IV. A SDF- 1/PBSF Receptor, CXCR4
A. IDENTIFICATION, STRIJCTUHK, .IN]) SIGNAL TIUNSDLI(;TION Cheniokine receptor-like orphans were screened for their ability to induce increases in intracellular Calt in response to human SDF-l/PBSF,
and Chinese hamster ovary cells stably transfected with human HUMSTW HM89/LESTR (Federsppiel et ul., 199:3; Herzog et nl., 1993; Jazin et nl., 1993; Loetscher et al., 1994; Nornura et al., 1993) showed a response to human SDF-l/PBSF (Fig. 4) ( B l e d et nl., 1996a; Oberlin et nl., 1996). Because this receptor has been found to respond to a CXC chemokine, it has been designated CXC chernokine receptor 4 (CXCR4).There are two alternately spliced forms of tnurine CXCR4 that differ by only two amino acids in the N-terminal ectodomain (Heesen et al., 1997; Moepps et nl., 1997). The gene for CXCR4 has been localized to chroinosoine 2q in the vicinity of the gene for other CXC chemokine receptors (Federsppiel et al., 1993). The amino acid sequence of CXCR4 is highly conserved (90%) between mouse and human compared to other CXC cheinokine receptors, IL-8 receptor A (68%) or B (71%) Nagasawa et al., 1996b; Heesen et al., 1996). Little is known about the signal transduction through CXCR4. SDF-1/ PBSF-mediated calcium responses of cells that express CXCR4 are inhibited by pertussis toxin, suggesting that CXCR4 mediates the activation of pertussis toxin-sensitive G-protein (Tanabe et d., 1997). In addition, experiments with insect cells (Sf9) have shown that murine CXCR4 can couple to Gai2ply3 (Moepps et al., 1997).
FIG 4. Amino arid scquencc and trmsmeinbrane model of' 1iunian CXCR4
A
(
x:(:I 1 I7 !VlOKl N F:
s I11:-
l/I'HS F
219
B. EXIWESSION CXCR4 m R N A is expressed constitiitively diiring embryogenesis (E7.5, E11.5, E15.5, and E17.5 mice) (Nagasawa et d . ,1996b) and in many adrilt organs, such as brain (inoiise, rat, bovine, inonkey), lung (rat, monkey), heart (rat, monkey), thymus (nioiise), lyniph node (mouse),spleen (inoiise, monkey), bone inarrow (mouse),stoinach (nioiise),small intestine (mouse), colon (monkey), liver (rat, monkey), and kidney (mouse, rat, bovine) (Federsppiel et al., 199:3; Moepps P t (11.. 1997; Nagasawa ct id., 199613; Rimland et d.,1991). These results are iiiirrored in its ligaiid SDF-1/PBSF and support the idea that inurine CXCR4 is a phvsiological receptor for murine SDF- l/PBSF. Transcripts of CXCK4 were esprcsed in the B-cell progenitor clones and cell lines (DW34, 207, Reh, Nalm-6, 697, 7Ozl3, 18-8, 63-12), the Burkitt lym~~lioma-derived B-cell lines (Ramos, Daudi, Raji), the B lymphoma cell line (A20),peripheral blood lymphocytes, PHA-activated T-cell blasts, thymic T cells (CD4-CD8-, CD4'CD8', CD4'CD8-, CD4-CD8' cells), the T lymphoma cell lines (EL14, Jurkat), peripheral blood neutrophils, peripheral I-)loodiiioiiocytes, the promyelocytic leukemia cell line (HL-60), the monocytic leukemia cell lines (THP-1, U937, WEHI-265.1, IC21), the uterin cervical carcinoina cell line (I-IeLa),the neuroblastoma cell lines (LAN5, SH-SY-SY), the einbryoiiic kidney cell line (293), astrocytes, microglial cells, a i d the microgIiaI cell liile ( ~ 9 and ) were not expressed in the fibroblast cell lines (NIH/3T3 cells, L929), the mesangial tumor cell line (MES-13), or the mast cell clone (MC9) (D'Apiizzo et d , 1997; Federsppiel et nl., 1993; €leesen ct a/., 1996, 1997; Lavi et al., 1997; Loetscher et d , 1994; Moepps et ul., 1997; Nagasawi ct d . ,1996b, Nomura et al., 1993; Tan& ct nl., 1997). Hunian dendritic cells and neurons have been shown to express CXCR4 iriRNA (Ayehunie et d.,1997; GranelliPiperno et al., 1996; Sozzani et d., 1997; Lavi et d . , 1997). Experiments with a mAb to CXCR4 revealed that CXCR4 is expressed on the naive, uiiactivated C1126 low CD45RA' CD45RO- T lymphocyte subset of peripheral blood lympliocytes, and the expression was rapidly uprcgulated on periplierd l h d mononuclear cells during phytoheinagglutinin stimulation and interleiikin 2 piiming ( B l e d ct nl., 1997). Tlie molecular Imis for tlie regulatioii of CXCR4 gene expression has begun to be analyzed. The 5' upstrcam rcgion of the CXCK4 gene contains the TATA box, two potential GC boxes, and a potential iirickar respiratory factor (NRF-1)-bindingsite, which w a s important for the basal and induced activity of tlie CXCR4 promoter (Moriuch et d , 1997). V. HIV-1 Infection and CXCR4
In 1996, CXCR4 was found to fiinction as a coreceptor responsible for the entry of hunian iininuiiodeficiency viiiis (HIV-1) into target cells (Fig.
220
TAEASIII NAGASt\\Z’A ct nl
5 ) (Feng et aZ., 1996). Because CD4, the primary receptor on the host cells, was shown to support viral entiy only when expressed on human cells (Maddon et al., 1986), the existence of a human-specific coreceptor has been suggested since the mid-1970s. This coreceptor is essential not only for the entry of HIV-1 but also for fusion between cells expressing the HIV-1 envelope glycoprotein and cells expressing CD4. Feng et al. (1996) established an assay system in which fusion between these cells leads to activation of a reporter gene. They then adapted the system for functional expression cloning of a fusion coreceptor cDNA and identified a coreceptor for fusion and entry of HIV-1 (Feng et d., 1996). This protein is a G-protein-coupled receptor with seven transrnernbrane helices designated CXCR4, a human SDF-1/PBSF receptor. Shortly aftenvard, physical association of a CD4, HIV-1 envelope glycoprotein (gp120), and CXCR4 was found by irninunoprecipitation (Lapham et al., 1996). The ability of HIV-1 to infect host cells varies from strain to strain and is referred to as cellular tropism (reviewed by Fauci, 1996) (Fig. 5 ) . All strains infect primaiy CD4’ T lymphocytes. Some primary isolates also infect nionocytes but not in transformed T-cell lines and are classified as monocyte- or macrophage-tropic (M-tropic) HIV-1 strains. Other isolates infect transformed T-cell lines but not nionocyte or inacrophage strains and are referred to as T-cell line tropic (T-tropic) HIV-1 strains. As HIV-1 infection progresses, the dominant M-tropic strains are usually replaced by T-tropic strains. A HIV-1 entry coreceptor, CXCR4, was found to act preferentially for T-tropic HIV-1 but not for M-tropic HIV-1 (Feng et al., 1996). However, a cluster of reports showed that the CC chemokine receptor CCR5 functioned as a coreceptor for M-tropic HIV-1 but not Macrophage
M-tropic HIV-1
T cell
T-cell line
T-tropic HIV-I
FIG.5 . Coreceptors for M- and T-tropic strains of HIV-I. The T-tropic HIV-1 strain can infect host cells expressing CXCR4.
for T-tropic HIV-l (Alkhatib et nl., 1996; Choe et al., 1996; Deng 4t nl., 1996; Doranzet d . ,1996; Dragic et a / . , 1996).Consistent with that, primary CD4+ T lyinphocvtes express both CXCR4 and CCR5, and transformed T-cell lines express CXCR4 but not CCRS (Fig. 5). These results indicate a resolution for the long-standing problem of differences in HIV-1 cellular tropism. Although a coreceptor for HIV-1 entry was identified as a hunian-specific molecule, inurine CXCR4 was also found to allow fusion and entry of T-tropic HIV-1 (Bieniasz et nl., 1997; Tachibana et nl., 1997). It has been shown that fiirictional CXCR4 is not expressed in inurine cells; NIH/3T3 that was transfected with human CD4 could not support HIV-l-induced membrane fusion (Bieniaszet al., 1997;Tachibana et nl., 1997). In contrast, inurine CCRS is nonfunctional as a coreceptor for M-tropic HIV-1 (Atchison Pt nl., 1996). Analysis using hnnian-rat CXCR4 chimeras revealed that the first and third extracelliilar tlomains are required for the functional interaction of some T-tropic HIV-l strains with CXCR4 (Brelot et d . , 1997).Thus, the C-terminal cytoplasmic domain ofCXCR4 is not essential for HIV-1 entry (Amara ct d ,1997). The role of intracellular signaling activated by CXCR4 in HIV-1 entry remains to be elucidated. Viral entry does not depend on Gi-coupled vignaling (Cocchi et d.,1996). However, contact between HlV-1 envelope and CXCR4 was shown to initiate a multiple signaling pathway (Davis ct crl., 1997). Although HIV-1 isolates utilize both CXCR4 and CD4 as entry receptors, some HIV-2 isolates were found to use CXCR4 as an alternative receptor in the absence of CD4 (Endres ct nl., 1996). Before the HIV-1 entry coreceptors were identified, the CC cheniokines, RANTES, MIP-la, p, which are ligands for CCR5, were found to block infection with M-tropic HIV-l (Cocchi ct nl., 1995). Then, SDF-l/PBSF was shown to block infection with T-tropic HIV-1 but not with M-tropic HIV-1 (Bleul et a / . , 1996a; Oberliii et al., 1996). SDF-1PBSF induces rapid internalization of CXCR4, md ligand-mediated downregulation may contribute to the HIV-1 suppressive effect of SDF-l/PBSF (Amara et d . , 1997; Signoret et nl., 1997). These results suggest that cheinokines or their derivatives may be ~isefiilagents for the treatment of HIV-1 infection. Three sinall molecule CXCR4 inhibitors that block T-tropic HIV-1 infection have been identified. One is a class of heterocyclic coinpouncls called bicyclains (Donzellact d . ,1998;Schols ct al., 1997),and others are polypeptides of 9 (Doranz et nl., 1997) and 18 (Murakaini et d., 1997) amino acid residues. Their tliree-~mensioiialstructures would assist the rational design of agents capable of inhibiting viral entry mediated by CXCR4, CD4, and the gp120 complex.
Individuals that have a homozygous 32-bp deletion in CCRS are resistant to sexually transmitted HIV-1 (Dean et al., 1996; Liu et al., 1996; Samson et al., 1996). In addition, the highly exposed but uninfected individuals make higher amounts of CC chemokiries (Paxton et al., 1996). Thus, it is important to determine whether mutations in the gene of CXCR4 or SDF-l/PBSF exist among individuals infected with HIV-1, particularly a long-term nonprogressor. Because mice lacking SDF- 1/PBSF are perinatal lethal, homozygous defects in humans for the CXCR4 gene are unlikely to exist (Nagdsawd et al., 1996a). However, it is unclear whether CXCR4 is an essential physiological receptor for SDF-UPBSF. VI. Perspectives
From its initial discovery, the importance of SDF- l/PBSF and CXCR4 has grown to cover a variety of biological and pathological functions. First, it was exciting that a chemokine is crucial for development, including hematopoiesis and cardiogenesis. Of particular interest is the finding that a cytokine plays a critical role in colonization of the bone marrow by hematopoietic precursors, in addition to cell adhesion molecules. Much work remains to be done in defining the functional mechanisms of SDF-l/PBSF and its interactions with other cytokines that bind receptors such as transmeinbrane receptor tyrosine kinases or cytokine receptor family receptors in those developmental processes. In 1996, a chemohne receptor, BLR-1, was shown to be responsible for the localization of B lymphocytes and their migration in response to antigen stimulation. Considering that SDF-l/PBSF and CXCR4 are highly expressed in many adult organs, including lymph node, spleen, and thymus, SDF-l/PBSF may play a critical function in immune response in those organs. Second, because CXCR4 was identified by screening known chemokine receptors for their ability to respond to SDF-l/PBSF, it remains uncertain whether CXCR4 is the only receptor for SDF-1h'BSF. Further studies about the functions of CXCR4 are required. Finally, the discovery about the involvement of CXCR4 in HIV-1 infection raised the possibilities that therapy could be aimed at blocking CXCR4 using peptides or other agents or at regulating the expression of CXCR4 or SDF-l/PBSF. There is no doubt that further understanding of the biological and pathological properties of SDF-1/ PBSF and CXCR4 should be pursued.
ACKNOWLEDGMENT We thank Ms. K. Kawainori for secretarial assistance.
REFEHENCES Aiuti, A., Webb, 1. J., Bled, C. C., Springer, T., and Gutierrez-Kamos, J. C. (1997). The cliemokine SDF-1 is a chemoattractant for human CD34+hematopoietic progenitor cells
and provides B new mechanism to (Jxplain the inobihzation of CD34’ progenitors to peripheral I)lootl. /. Exp. Akd. 185, 11 1-120, Alkhatil), G.,Coinlxidiere. Brotler. C. C.. Fcmg, Y.,Kennedy. P. E., Murphy, P. M., and Berger, E. A. (1996). CKRS: A HANTES, MIP-la. MIP-lP receptor as a fusion cofactor for inacrophage-tropic HI\’-I. Scioict) 272, 1955-1958, Ainara, A., Gall, S. L., Schwartz, O., Salaiiiero, J., Montes, M., Loetscher. P., Baggiolini, M., Virelizier. J.-L.,and Arenzana-Seiscledos, F. (1997).IWcoreceptor do\.nregulation as antivird principle : SDF- la-dependent internalization ofthe cheinokine receptor CXCRl coritrihntes to inhibition of HIV replication. /. Exp. Mcd. 186, 139-146. Arvanitakis, L.. Geras-Raaka, E.. Vamia. A . , G c d i t y p m . M. C . , and Cesarman, E. (1997). Human Iierpes\irus KSHV encodes ii constitutively active C-protc-in-coupled receptor hiked to cell proliferation. h’ntiirc, 385, 347-350. Atchison. R. E.. Gosling. J., Montcclnro. F. S.. Franci, C., Digilio, L., Charo, I. F., and Gokhnith, M . A. ( 1996). Miiltiple c.xtl.acellular eleineiits of CCR5 and HIV-1 entry: Dissociation from r r y o n s r to chemokines. Scicvicc 274, 1924- 1926. Ayehnnie, S . , Garcia-Zepeda. E. A . , Hosic. J. A,. Iloruk, R.. Kupper. T. S . , Luster, A. D., and Riipreclit, R. M. (1997). Hunian iiiiiiiiiiiodeficieii~~ \irus-1 entry into purified blootl deiidritic cells through CC and CXC clieniokinc coreceptors. Blood 90, 1379- 1386. Baggiolini, M . , Ilewald, B., ;uid Moser, H. (1994). Interleukin-8 and related chemotactic cytokines. A r k . I t t i ~ t i i n i o / .55, 97- 179. Raggiolini, M., Dewdd, B., arid Moser, 13. (1997). Hunian clieiiiokines: An update. Anntc.. Her. l t t i t t i i c . t i o / . 15, 675-705. Bienikisz, P. D.. Fridell, R. A . , Anthony, K.,and Cullen, B. R. (1997). Murine CXCR-4 is a functional corecrptor for T-cell-tropic kind diial-tropic strains of human imniiinodrFiciency virns h p c s 1. /. V i r d . 71, 7097-7100. Bleul, C . C:., Farzan, M., Choe. H., Paroliii. C.. Clark-Lewis, I., Sotloroski. J., and Springer, T. A . (1996u).The lymphocyte clieinoattractant SDF-I is ii ligand for LESTWfiisin and blocks Ill\’-I entry. Nntrirc. 382, 829-833. B h I , C . C., Fuhlbrigge, R. C.. Casasnovas, J . M., Ainti, A , , and Springer, T. A. (1996b). A highly efficacious lyinphocytr cliciiioattr~ictant,stromal cell-derived factor 1, J , E x p A4ed 184, 1101-1109. Bled, C. C., Wn, L.. Hoxic, J. A,, Springer, T. A , . and Mackay, C. R. (1997). The HI\’ corweptors CXCK4 and CCKS are tlifferentially expressed and regulated on hiiman T Iymphoc).tes. Pro(,. Nod. Accicl. Sci. USA 94, 1925- 1930. Boring, L., Gosling, J., Chensue. S. W., Kunkel, S. L., Farese, R. V., Jr., Broxmeyer, H. E., and Charo, I. F. (1997). 1mp;tiretl nionoc!te migration and reduced type l(Th1) cytokine responses in C-C chemokine rcwptor 2 knockoiit inice./. C l i x Inuest. 100,25522561. Boshoff, C., Endo Y., Collins, P. D., Taketichi, Y., Keeves, J. D., Schweickart, V. L., Siani, M . A , , Sasaki. T., \Villiams, T. J., Gray, 1’. W.. Moore, P. S., Chang, Y., and Weiss, K . A. (1997). Angiogenic and HIV-irihibitoiy function of KSHV-encoded chemokines. Scicricc 278, 290-294. Brelot, A,, Heveker, N., Pleskotf, O., Sol, N., and Alizon, M . (1997). Role of the first and third extri\cellular domains of CXCR-4 in hunian iiiiiiiuiiodeficieiicy virus coreceptor activity. /. Virol. 71, 4744-4751. Cacaliiino, C., Lee, J., Kikly, K., Ryan, A. M., Pitts-Meek, S.. Hultgren, B., Wood, W. I., and Moore M . W.(1994). Neutrophil and B cell expansion in mice that lack the iininune IL-8 receptor hornolog. Scieticc 265, 682-684. Cliocl, H., FCL~II,M., Sun, Y., Sullivan, N., Rolliris, B., Poriatli, P. D., Lijun, W., Mackay, Gerard, N., Gerard, C.. and Sodoroski, J. (1996).The C. R., Laro5a. G., Newmiin, W..
224
TAKASHI NAGASAWA et (11
j3-chemohne receptors CCR3 and CCRS facilitate infection by primary HIV-1 isolates. Cell 85, 1135-1148. Clark-Lewis, I., Dewald, B., Loetscher, M., Moser, B., and Baggiolini, M. (1994). Structural requirement for interleukin-8 function identified by design of analogs and CXC chemokine hybrids. J. B i d . C h e m 269, 16075-16081. Cocchi, F., DeVico, A. L., Garzino-Demo, A,, Alya, S. K., Gallo, R. C., and LIISSO,P. (1995). Identification of RANTES, MIP-la, and MIP-lP as the niajor €IIV-suppressive factors produced by CD8+ T cells. Science 270, 1811-1815. Cocchi, F., DeVico, A. L., Garzino-Demo, A,, Cara, A,, Gallo, R. C., and Lusso, P. (1996). The V3 domain of the HIV-lgpl20 envelope glycoprotein is critical for chemokinemediated blockade of infection. Nuture M e d . 2, 1244-1247. Cook, D. N., Beck, M. A., Coffinan. T. M., Kirby, S. L., Sheridan, J. F., Pragnell, I. B., and Smithies, 0. (1995). Requirement of MIP-la for an inflammatory response to viral infection. Science 269, 1583- 1585. Cnimp, M. P., Gong, J.-H., Loetscher, P., Rajarathnani, K., Amara, A., Arenzana-Seisdedos, F., Virelizier, J.-L., Baggiolini, M., Sykes, B. D., and Clark-Lewis, I. (1997). Solution structure and h i s for functional activity of stroinal cell-derived factor-1; dissociation of CXCR4 activation frotn binding and inhibition of HIV-1. E M B O J. 16, 6996-7007. D’Apuzzo, M., Rolink, A,, Loetscher, M., Hoxie, J. A,, Clark-Lewis, I., Melcher, F., Baggiolini, M., and Moser, B. (1997). The cheinokine SDF-1 stromal cell-derived factor 1, attracts early stage B cell precursors via the chemokine receptor CXCR4. Eur. J. Zmmunol. 27, 1788-1793. Davis, C., Dikic, I., Unutmaz, D., Hill, C. M., Arthos, J., Siani, M. A,, Thompson, D. A., Schlessinger, J., and Littman, D. R. (1997). Signal transduction due to HIV-1 envelope interactions with chemokine receptors CXCR4 for CCRS. J . Exp. Med. 186, 1793-1798. Dean, M., Carrington, M., Winkler, C., Huttley, G. A,, Smith, M. W., Allikmets, R., Goedert, J. J., Bnchbinder, S. P., Vittinghoff, E., Gomperts, E., Donfield, S., Vlahov, D., Kaslow, R., Saah, A,, Rinaldo, C., Detels, R., Hemophiliagrowth and development study, Multicenter AIDS cohort study, Multicenter hemophilia cohort study, Sail Fransisco city cohort, ALIVE study, and O’Brien, J. S. (1996). Genetic restriction of HIV-1 infection and progression to AIDS by a deletion allele of the CKRS structural gene. Science 273,18561862. Deng, H., Liu, R., Ellineier, W., Choe, S., Unutmu, D., Burkhart, M., Marzio, P. D., Marmon, S., Sutton, R. S., Hill, C . M., Davis, C. B., Peiper, S. C., Schall, T. J., Littinan, D. R., and Landau, N . R. (1996).Identification of a major co-receptor for primary isolates of IIIV-1. Nature 381, 661-666. Donzella, G. A,, Schols, D., Lin, S. W., Este, J. A,, Nagashima, K. A., Maddon, P. J., Allaway, G. P., Sakmar, T. P., Henson, G., Clercq, E. D., and Moore, J. P. (1998). AMD3100, a small molecule inhibitor of HIV-1 entry via the CXCR4 coreceptor. Nature Med. 4, 72-77. Doranz, J. B., Rucker, J . , Yi, Y., Smyth, R. J., Samson, M., Peiper, S. C., Parmentier, M., Collman, R. G., and Doms. R. W. (1996), A dual-tropic primary HIV-1 isolate that uses fusin and the j3-chernokine receptors CKR-5, CKR-3, and CKR-2b as fusion cofactors. Cell 85, 1149-1158. Doranz, B. J., Grovit-Ferbas, K., Sharron, M. P., Mao, S.-H., Goetz, M. B., Daar, E. S., Doms, R. W., and O’Biien, W. A. (1997). A small-molecule inhibitor directed against the cheinokine receptor CXCR4 prevents its use as HIV-1 coreceptor. J. Exp. Med. 186, 1395-1400. Dragic, T., Litwin, V., Allaway, G. P., Martin, S. R., Huang, Y., Nagashima, K. A,, Cayanan, C., Maddon, P. J., Koup, R. A,, Moore, J. P., and Paxton, W. A. (1996). HIV-1 entry into CD4 cells is mediated by the chemohne receptor CC-CKR-5. Nature 381, 667-673.
A C X : (:I4EMOKIiiE SDF-I/PHSI:
225
Endrrs, M. J., Clapliani, P. R., Marsh, M., Ahuja, M., Turner, J. D., McKnigIit, A,, Thomas, J. F.. Stoebenau-Haggart)., B., Choe, S..\’anre. P. J., Wells. T. N. C., Power. C. A,, Sritterwala, S. S.. Doins, K. I[’., Lantlnri, N . I<., and Hoxie, J. A. (1996).CD4-independent infection by HIV-2 is niediatetl by fiisiiJCXCR4. Ccll 87, 745-756. Fauci, A. S. ( 1996). Host factors and the, pithogenesis of f 111’-induced disease. Notiirc 084, 529-534. Federsppiel, B., Melhado, I. G., Duncan, A. M. V., Delancy, A , , Schappert, K., ClarkLewis. I., and J i i i k . F. R. (1993). Molecular cloning of the cDNA and chroinosomal localization of the gene for a putatiw seveii-transiiic,mbrane segment receptor isolated from liuniali spleen. Genonlics 16, 707-712. Feng, Y.. Brotler. C.C., Kennedy, P. E., and Herger, E. A. (1996). HIV-1 entry cofactor: Fiinctional cDNA cloning of a sc~~cn-transm~iiibr;iiie. C; protein-coiipled receptor. Science 272, 872-877. Gao. J.-L.. L V y i i i , T. A.. Chaiig Y., Lee, E. J., Broxnieyer, H. E., Cooper, S., Tiffany. H. L., \I’estphal, H., Kwoii-Chung. J.. a i d Mii~-pIiy.P. M. (1997). Impaired host defense. heinatopoiesis. granulomatous inflainination and t>lw 1-type 2 cytoliine balance in mice lacking CC chernokine receptor 1 . J. Erp. A f d . 185, 1959-1968. Grabstein, K. H.. \Valdschniidt, T. J., Finkelinair. F. I).. [less, B. W., Alpeit, A. R., Boiani, N. E., Namen, A. E., and Morrissey, P. J. (1993).Inhibition ofmurine B a n d T lymphopoiesis in vivo by an anti-interleukin 7 nioiiocloiial antilwcly. J. Exp. hled. 178, 257-264. Graham, G. 1.. Lt’riglit, E. G., Hewick. H., \Volpe, S. D.. Wilkie. N. M., Donaldson, D., Loriinore, S.. and Prugiell, 1. B. ( 1990). Identification and characterization of an inhibitor of hematopoietic stem cell prohferatioii. Scicnct 344, 442-444. Granelli-Piperno, A.. Moser, B., Pope, M., Chen, D., \\lei, Y., Isdell. F., O’Doherty, U., Puton. W., Koup. R., Mojsov, S., Bhardwij, N., Chk-Lewis, I., Baggiolini, M., and Striiiiiian, 13. M. (1996). Efficient interaction of HIV-I with purified dendritic cells via inultiple cliemokine coreceptors. J . Erp. hfrtl. 184, 24433-2438. Ileesen, M.. Bernian, M. A , . Benson, J. D., Ckrard, C . , and Dorf, M . (1996). Cloning of the nioiise fiisin gene, homologue to ii hiiiiian HIV-I co-fktor. J . Irnmutiol. 157, 5455-5480. Heescn, M.. Herinan, M. A,, Benson, J. D., Hopken, U. E., Gerard, N . P., and Dorf, M. ( 1997). ,4lteniattn splicing of inoiisr fiisirKXC chemokine receptor. j . Znininnol. 158, 3562-3561. Herzog, H., IIort, Y. J.. Shine, J.. and Selbie. I,. A . (199:3). Molecular cloning, cliaracterization, and localization o f the Iiumaii homologue to the reported bovine NPY Y3 receptor: Lack of NPY binding and activation. DNA Cell. Biol. 12, 465-471. Jazin, E. E., Yoo, [ I . , Blonicjvist, A. C., Yee, F., Weng, G., Walker, M. W., Salon, J., Larhanimar, D., and Wahlestrdt, C . ( 1993). A proposed h i n e neuropeptide Y receptor cDNA clone. or its hiiman homologue. confer neither NPY binding sites nor NPY responsiveness 0 1 1 t ransfected cells. Regrd. Popt. 47, 247-2Fi8. Kastner, P., Grondona, J. M., Mark. M., Gansuinrillcr, A , , LeMeur, M., Decimo, D., Vonesch, J. L., D o k , P.,and Chambon, P. (1994).Geiietic analysis of RXRa tleveloprnent function: Convergence of RXR and RAK signaling pathways in lieart and eye inorphogenesis. Cell 78, 987-1003. Kurihara, T., !+‘am, G.. Loy, J., and Bravo. H. (1997).Defivts in inacrophage recruitment and host defense in mice lacking the CCR2 chemokine receptor. j . Exp. Med. 186,1757- 1762. Kurihara. Y., Kiiriliara, H., Oda, H.. Mwiirira, K., Nagai, R., Ishikawa, T., and Yazaki, Y. (199rj).Aortic arch nialformations and ventriciilar septa1 defect i n inicc deficient in entlotlielin-1. J. C h i . Znuest. 96, 293-300. Landau, N. K.,Warton, M., a i i d Littmari, D. R . (1988).The envelope glycoprotein of the hiunan iinmunodeficiency birus hinds to the immunoglobulin-like domain of CD4. Nature 334. 159-162.
226
TAKASHI NAGASAWA r f ol
Lapham, C. K., Ouyang, J.. Cliandrasekhar, B., Nguyen, N. Y., Dimitrov, D. S., and Golding, H. (1996).Evidence for cell-surface association between fusin and the CD4-a120 complex in human cell lines. Scierice 274, 602-605. Lavi, E., Strizki, J. M., Ulrich. A. M., Zhang, W., Fn, L., Wang, Q., O’Connor, M., Hoxie, J. A,. and Gonzalez-Scarano, F. (1997). CXCR4 (fnsin), a co-receptor for the type 1 human immunodeficiency virns (HIV-l), is expressed in the human brain in a variety of cell types, including microglia and neurons. Am. /. Padid. 151, 1035-1042. Liu, R., Pawton, W. A., Choe. S., Ceradini, D,, Martin, S. R., Horuk, R., MacDonald, M. E.. Stiihlmann, H., Konp, R. A . , and Landau, N. R. (1996). Homozygous defect in HIV- 1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV1 infection. Cell 86, 367-377. Loetscher. M., Geiser, T., O’Reilly, T.. Zwahlen. R., Baggiolini, M., and M o w , B. (1994). Cloning of a hurnan seven transmembrane domain receptor, LESTR, that is highly expressed in leukocytes. 1.B i d . Chern. 269, 232-237. Maddon, P. J,, Dalgleisli, A. G., McDongal, J. S., Clapham, P. R., Weiss, R. A , , and Axel, R. (1986).The T4 gene encodes the AIDS virus receptor and is expressed in the immune system and the brain. Cell 47, 333-348. Moepps, B., Frodl, R., Rodewald, H. K., Baggiolini, M., and Gierschik, P. (1997). Two murine homologues of the human chemokine receptor CXCR4 mediating stroinal cellderived factor la activation of Ci2 are differentially expressed in vivo. Eur. J. Imnarnt~l. 27,2102-2112. Moriuchi, M., Moriuchi, M.,Turner, W., and Fauci, A. S . (1997).Cloning and analysis of the promoter region of CXCR4, a coreceptor for IIIV-1 entry. /. Irnmnunid. 159,4322-4329. Mnrakami, T., Nakajima, T., Koyanagi, Y., Tachibana, K., Fujii, N., Tamaniura, H., Yoshida, N., Waki, M., Matsumoto, A,, Yoshie, O., Kishimoto, T., Yainamoto, N., and Nagasawa, T. (1997).A sniall molecule CXCR4 inhibitor tliat blocks T cell line-tropic HIV-1 infection. J. Exp. Med. 186, 1389-1393. Murphy, P. M. (1994).The molecular biology of leukocyte chemoattractant receptors. Annu. RUG. ~ I l l I l Z U t Z i J ~12, . 593-633. Nagasawa, T., Hirota, S., Tachibana, K., Takaknra, N., Nishikawa, S. I., Kitamura, Y., Yoshida, N., Kikntani, H., and Kishimoto, T. (19964. Defects of B-cell lymphopoiesis and bone-marrow inyelopoiesis in mice lacking the CXC chemokine PBSF/SDF-1. Nature 382, 635-638. Nagasawa, T., Kikutani, H., and Kishimoto, T. (1994). Molecular cloning and structure of a pre-B-cell growth-stiinulating factor. Proc. Natl. Acad. Sci. USA 91, 2305-2309. Nagasawa, T., Nakajima, T., Tachibana, K., Ilzasa, H., Bled, C. C.. Yoshie, O., Matsushima, K., Yoshida, N., Springer, T. A,, and Kishimoto, T. (1996b). Molecular cloning and characterization of niurine pre-B cell growth stimulating factor/stromal cell-derived factor (PBSF/SDF-I ) receptor, a inurine homolog of HIV-1 entry coreceptor fiisin. Proc. Natl. Acad. Sci. USA 93, 14726-14729. Nomura, H., Nielsen, B. W., and Matsushima. K. (1993). Molecular cloning of cDNAs encoding a LD78 receptor and putative leukocyte cheinotactic peptide receptors. Int. Irrinit~nol.5, 1239-1249. Noinura, M., Matsuda, Y., Itoh, H., Hori, T., and Suzuki, G. (1996). Genetic mapping of the mouse strornal cell-derived factor gene to mouse and rat chromosomes. Cytogenet Cell Genet. 73, 286-289. Oberlin, E., Amara, A., Bachelerie, F., Bessia, C., Virelizier, J.. Areiizana-Seisdedos, F., Schwartz, O., Heard, J., Clark-Lewis, I., Legler, D. F., Loetscher, M., Baggiolini, M., and Moser. B. (1996). The CXC cheinokine SDF-1 is the ligand for LESTIUfusin and prevents infection by T-cell-line-adapted HI\’-l. Nature 382, 833-835.
4 CSC CIIEMOKINE SDF-l/PBSF
227
Oppenlieini. J. J., Zacliariae, C:. O., Mukaida, N., and Matsushima, K. (1991). Properties of the novel proiir H~iiiimotonsupergene intercrine cytokinc family. A t i t l / r . Rro. Z m t t w i i o l , 9, 617-648. Paxton, W.A,. Martin, S. R., Tse, D., O’Brirn, T. R., Skiirnick, J., VanDevavter, N. L., Padian, N., Braun, J. F., Kotlrr, D. P., \vOlinskl\,,S. bl., and Koup, R. A. (1996). Relative resistance to €II\’-l infection of CD4 Iyinpliocytes froin persons who remain uninfectetl despite niultiple high-iisk sesnal exposures. Nature. RIetl. 2, 412-417. Peschon,1. J., Morrissey, P. J.. Grabstein, K. 13.. Ramsdell, F. J., Maraskovsky, E., Gliniak, B. C., Park. L. S., Ziegler, S. F., Willianis. D. E., \V:ire, C;. B., Meyer, J. D., and Davison. B. L,. (1994). Early lymphocyte expression is severely impaired in interleukin 7 receptordeficient inice. J . Exp. Med. 180, 1955- 1960. Rimlantl, J., Xiii, LV., Sweetnani, P., Saijoli, K., Nestler, E. J., and Iluman, R. S. (1991). Seqwnce and expression of a neuroprytide Y receptor cDNA. Md. Phnt71uccol. 40, 869-875. Samson. M., Libcrt, F., Doranz, B. J., Rricker, J., Liesnard, C . . Farber, C.-M., Saragosti, S., I,apoiiineroiilie, C . , Cognaus, J., Forccille. C . , bfiiylderinans, G., Verhofstede. C.. Burtonboy, G., Georges, M.. Imai, T.. Raiia. S., Yi. Y., Sinyth. R. J., Collnian, R. G., Iloms, H. \V., Vassxt, G.. and Parinentier. M. (1996). Resistance to € I N - 1 infection in caucasian intli\;itliials bearing niiitant ;illeles of tlic, CCR-5 chemokine receptor gene. Nattirc 382, 722-725. Schok. D., Stru??: S., Darnnw, J. V., Este. J. A,. Henson, G., and Clereq, E. D. (1997). Inhibition of T-tropic HI\’ strains by selective antagoiiization of the cheinokine receptor . 1383-1388. CXCR4. J , Esp. M P ~ 186, Sekitlo. N., Mukaida, N., Harada. A,, Nukanislii, I.. Watanabe. Y., and Matsushiina, K. (1993).Prevention of lung reperfusion injury in rabbits by a monoclonal antibody against interleukin-8. Natrrrc, 365, 6544657. Shirozii. M . , Nakano. T., Inazawa, J., Tashiro. K., Tatla, H., Sliinohara, T.. and Honjo, T. (1995). Structure and cliromosoiiial localyzation of the humm stronial cell-derived factor 1. Gcwotnic.s 28, 495-500. Sigiioret. N . . Oldridge, J., Pelchen-Matthews, A . , Klasse. P. J.. Tim, T., Brass, L. F., Rosenkildc, M. M.. Schwartz, T. U’.,Holmes, it’., Dallas, W., Luther, M. A,, Wells, T. N. C., IIoxie, J. A , , and Marsh. M . (1997). Phorbol esters and SDF-1 induce rapid endocytosis arid tlown inodiilation ot‘ the cheinokiiie receptor CXCR4. J. Cell Biol. 139, 651-664. Borsatti, A,, Polciitarutti, N., Zhou, D., Pienionti, L., D’Amico, G., Sozzani, S.. Ixini, W., Power, C. A,, Wells, T. N . C., Gobbi, M., Allavena, P., antl Mantovani, A. (1997).Receptor expression and responsiveness of human dendritic cells to a defined set of CC and CXC clieniokines. J. Zmttmnol. 159, 1993-2000. Sucov, H , M.. Dyson, E., Guineringer, C. I>., Price. I., Chien, K. R.. and Evans, R. M. (1994). RXRa mutant inice esta6lisli a genetic hasis for vitamin A signaling in heart niorpliogenesis. Gmes Dec 8, 1007- 1018. Siido. T., Nisliikawa, S., Ohno. N., Akiyania. N.. Euriakoshi, M.. Yosliida, H., and Nishikawwa, S. I. (1993). Expression and fiinctioii of the interlrukin 7 receptor in murine lymphocytes. Proc. N d Acmrl. Sci. USA 90, 9125-9129. Tacliiliana, K., Nakajima, T., Sato, A,, Igarashi, K., Shida, H., lizasa, H., Yoshida, N., Yosliie, O., Kishimoto, T.. antl Nagasawa, T. (1997). CXCR4lfiisin is not a species-specific 1,arricr in miirine cells for HIV-1 entiy. J. E r p M d . 185, 1865-1870. Tan;ibe, S., Heesen, M., Yoshizawa, I., Berman, M. A,, Luo, Y., Bleul, C . C., Springer, T. A,, Okiitla. K,,Gvrard. N., and l h r f , M. E. (1997). Functional expression of the
228
TAKASHI NACASAWA et a1
CXC-cheniokine receptor-4fusin on mouse niicroglial cells and astrocytes. I. Zrntntmol. 159, 905-911. Tashiro, K., Tada, H., Heilker, R., Shirozu, M., Nakano, T., and Honjo, T. (1993). A cloning strategy for secreted protein and type I mernbrane proteins. Science 261, 600-603. voii Freeden-Jeffiy, U., Vieira, P., Lucian, L. A,, McNeil, T., Burdach, S. E. G., arid Murray, R. (1995). Lymphopenia in interleukin-7 gene-deleted mice identifies IL-7 as a nonredundant cytokine. I. Exp. Med. 181, 1519-1526. This article was accepted for publication on March 24, 1998.
h l ) \ h Y ( E \ I Y I M M l I Y O L O C . ~ \ O L 71
T Lymphocyte Tolerance: From Thymic Deletion to Peripheral Control Mechanisms BRlGlTTA STOCKINGER Division of Molecular Immunology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, NW7 IAA United Kingdom
1. Introduction
Our understanding of tolerance induction in lymphocyte subpopulations has been exponentiallyadvanced hy the generation of T- and B-cell receptor transgenic mice. Initial contradictions, which seemed to appear as different transgenic models were published, have given way to overall consensus, at least as far as central tolerance induction in the thymus is concerned. It is now possible to dwriminate between clonal elimination and other ways of rendering lymphocytes unresponsive. The debate of whether tolerance is due to suicide or murder has been decided in fhvor of the former. Despite the assumption that we understand most of the phenomena resulting in central tolerance induction in thymus and bone marrow, a new area of uncertainty lies in our understanding of how tolerance can be induced and maintained in the periphery. This review concentrates on central and peripheral tolerance in the T-cell population only, given that this series has dealt comprehensively with self-tolerance in the B-cell population (Goodnow rt al., 1995). Table 1 lists T-cell receptor (TCR) transgenic mouse strains published to date. II. Central Tolerance Induction in the Thymus
A. CLONAL DELETION AS T I I E MAJORPHIN( IPLE OF TOLERANCE INDUCTION IN THE TIIIMUS
The first direct demonstration for the elimination of potentially autoreactive cells in the thymus came from studies of the T-cell repertoire to endogenous superantigens (Kappler et al., 1987; MacDonald et al., 1988). T cells bearing Vp segments that conferred binding to endogenous mouse inainmary tumor viruses (Mtv)were absent from mice carrying these superantigens in their genome. Analysis of thymic subpopulations established that there was deletion of T cells with Vp segments conferring superantigen reactivity at the immature CD4+8+stage. Shortly thereafter the first T-cell receptor transgenic inoiise strain appeared (Kisielow et al , 1988) bearing T-cell receptors specific for an epitope on the male antigen 229
C opynght 0 l9YY In Academic P r m hll nglit\ 01 rqxmliictwn I I ) d m form n w nrtl “65 27ifd99 $30 00
230 TABLE I T-CELLRECEPTOR TRANSCXNIC MOUSESTRAINS TCR Transgenic Strain 2 c (Sha et al., 1988) H-Y (Kisielow et nl., 1988) C6 (Donek et al., 1996) F5 (Mamalaki et al., 1993) KB5.C20 (Schonrich et al., 1991) BM3 (Spoonas et a/., 1994) OT-1 (Hogquist et al., 1994) 8.3 TCR (Verdaguer et al., 1997) K TCR (Geiger et nl., 1992) F3 (Morahan, et al., 1991) N15 (Ghendler et al., 1997) Clone-4 TCR (Morgan et a/., 1996) P14 (Pircher et al., 1989)
Specificity H2-Ld + QL9 peptide (Sykulev et al., 1994) H-Y peptide from Smcy (Markiewicz et al., 1998) H-Y peptide from Smcy (Scott et al., (1995) InHuenza NP 366-374 H2-K"
MHC Restriction
Clonotypic ab, Rag-/-, scid, or TCRa-1-
H2h class I
1B2 clonotypic mAh
H2-D"
Rag2-1T3.70 chotypic 111.4b
H2-Kk
Clonotypic mAh Va8, Vpll
H2-D" H 2'
Ragl -1Desire clonotypic niAb Ragl -198 clonotypic mAb Ragl-lVa2 mAh
H2' Ovalbuinin 257-264
HZ-K"
Undefined islet ag
H2-K"
SV40 T
h2-k1
H2-Kh VSV nuclear N52-59
H ~ - K ~
Va8 mAb Rag2-I-
InHueiiza hemaglutinin LCMV gp 32-42
H~-K"
DO.ll.10 (Murphy et d., 1990) AD10 (Kaye et al., 1989)
Ovalbntnin 323-339
H2-A'
Pigeon cytochrome c 88-104
H~-E'
2B4 (Berg et al., 1989)
Moth or pigeon cytochronie c Undefined islet ag
H2-E' H2-AI"
scid TCR a-I-
Influenza HA 111-119
H2-E"
Rag2-ITCRa -16.5 clonotypic mAb
IgG2a" 435-451 MBP 1-11 MBP 1-9
H2-A' H2-A" H-2A"
BDC.2.5 (Katz et d, 1993) 14.3d (Kirberg et al., 1994) B5 (Granucci et al., 1996) (Goverman et al., 1993) (Liu et al., 1995)
H2-Db
V a 2 , Vp8.1 mAh Rag2-IKJI-26 clonotypic tnAb
An ti-Vp3 Anti-Val1 Rag2-1-
231
T LYMPHOCYTE TOLERANCE
TABLE I (continued) TCR Trancgenic Strain
Specificity
MHC Restriction
~______
~
Clone 19 (1,afaille et a/ , 1994) Tag TCRl (Forstcxr ct (11, 1995) .3A9 (Ho ct a1 , 1994) 5C C7 (Seder et a l , 1992)
A18 (Zal c’t 0 1 , 1994) A1 (Zelenika et a / , 1998) 4B2A1 (Bogen et a/., 1993) Smarta (Tourne ct a[., 1997) TEa (Grubin P t d.,1997) SEP (Klein e t a ! . , 1998)
DEP (Klein c>t nl , 1998)
TCR-LACK (McSorley et al., 1997) T2.5-5 (Scott pf d., 1994) 4.1 TCR (Schmidt et d . , 1997) ”
Clonotypic ah, Rag-/-, scitl. or TCRa -1-
MRP 1-9
Rag1 -1-
SV40 large T 362-384
H2-A‘
HEI, 46-61 Pigeon cytoclrroine c 88-104 Mouse C5 106-121 HY peptide rniknown A”5 91-101
9H5 clonotypic mAb Rag1-13A9 anti-a antiserum Ragl -1-
H2-Ek 112-Eh 112-E“
LCMV pp 61-80
H2-Ah
Ragl -1Rag1-1scid GBI 13 clonotypic mAb Va2, Vp8.3 mAb
Ea52-68 C-reactive protein 80-94 C-reactive protein 89-101 Lei.dirrurnia mc!jor LACK 158-173 IIA 126-138 Undefined islet ag
112-Ah H2-Ah
Va2, Vp6 inAb
€12-A”
Vall.2 inAb
H2-A”
Va8, Vp4 mAb
Rag2-1-
Reagents identifying T-cell receptor in FACS analysis
H-Y. Again, deletion of T cells specific for H-Y was apparent in immature thymocytes at the double positive CD4’8’ stage in male mice. However, the time point of deletion was not identical for these systems. Generally, superantigen reactive T cells were deleted only at the late CD4+8+stage before transition to single positive mature cells, whereas deletion in CD8 T-cell receptor transgenic mice seemed to affect thymocytes at an earlier stage, leading to profound reduction of the double positive subpopulation and highly reduced thymus cellnlarity. However, CD4 T-cell receptor transgenic strains with physiological ligands showed a phenotype of “late” deletion as seen in superantigen-driven deletion (Douek et al., 1996; Zal et al., 1994: Zelenika et al., 1998).Although the timing of tolerance induction may depend on the localization of antigen (Pircher et al., 1989; Schneider ct al., 1992)or antigen-presenting cells (APC), large numbers of TCR transgenic strains, both of CD8 and CD4 lineage, provided evidence of a
232
BXIGITTA STOCKINGEX
heterogeneity in the timing of deletion, which is best explained by the overall avidity of T-cell receptor/antigen/major histocompatibility complex (MHC) interactions (Oehen et al., 1996). Transgenic strains vary in the time of onset and levels of T-cell receptor expression in different thymic subpopulations and in their coreceptor expression and receptor affinity. In addition there are differences in concentration and localization of cognate antigen and its access to different APC populations. Interactions that exceed a threshold of avidity tend to result in deletion, and an increase in one of the parameters that define the overall avidity can tip the balance. This is illustrated in various transgenic models; e.g., overexpression of CD8 coreceptor was shown to change a positive selection signal to a negative selection signal (Lee et al., 1992; Robey et al., 1991),expression of different levels of transgenic V p l l resulted in different degrees of deletion in the presence of H2-E (Homer et al., 1993), and changes in the density of alloantigen influenced the deletion pattern for thymocytes with a class I alloreactive receptor (Auphan et al., 1991). Deletion of antigen-specific thymocytes apparently requires a lower threshold of antigen stimulation than activation of mature T cells (Pircher et al., 1991; Vasquez et al., 1994). Thymic deletion can be induced in T-cell receptor transgenic mice through the injection of large amounts of cognate peptide. In these cases deletion affects the bulk of immature double positive thymocytes, resulting in a drastic reduction of thymus cellularity. However, a potential complication of this procedure is the concomitant activation of mature single positive thymocytes as well as peripheral T cells, resulting in the release of cytokines such as tumor necrosis factor (TNF) (Mamdaki et al., 1992; Martin and Bevan, 1997; Tarazona et al., 1998). It is currently unclear whether the nonspecific component of deletion is a consequence of direct TNF action or whether TNF acts on stromal cells, which in turn induce an apoptotic process in thymocytes (Lerner et al., 1996). Viral or bacterial infections can lead to transient thymic involution due to the deletion of immature double positive thymocytes, emphasizing that cytokine production in the thymus or periphery following strong T-cell activation can contribute to T-cell deletion in the thymus in a nonspecific manner (Wang et al., 1994; Bonyhadi et nl., 1993; Godfraind et al., 1995).
B. WHICHAPC PRESENT SELF-ANTIGEN FOR NEGATIVE SELECTION? Based on initial studies with bone marrow chimeras, positive and negative selection of thymocytes was ascribed to distinct cell types. Cortical epithelial cells were shown to effect positive selection, whereas bone marrowderived hematopoietic cells were responsible for tolerance induction (Lo and Sprent, 1986; Sprent et al., 1975; Bevan, 1977; Zinkernagel et al.,
1979; Marrack et al., 1988; Bix and Raulet, 1992; Cosgrove et al., 1992; Brocker et al., 1997; Markowitz et al., 1993). However, the involvement of epithelium and especially cortical epithelium remained controversial. Although some reports denied the contribution of cortical epithelium to deletion (Laufer et ul., 1996), others report at least partial effects, such as anergy induction, attributable to epithelium (Hugo et al., 1994; Gao and Sprent, 1990; Roberts et al., 1990; Ramsdell et nl., 1989; Tanaka et al., 1993). The situation was coinplicated further by the choice of antigens and the functional readoiits that were used to assess tolerance. Although thyinocytes with specificity for major liistocoinpatibility antigens appeared to ignore the presence of allo-MHC molecules on thymic epithelium (Ready et d ,1984; von Boehmer and Schubiger, 1984), it became clear subsequently that cortical epithelium and peripheral APC can present distinct sets of self-peptides on allo-MHC (Marrack et al., 1993). Thus, a functional readout of T-cell tolerance on peripheral APC may miss components of epitheliuin-induced tolerance. This was illustrated directly in experiments showing that thymic epithelium could tolerize in a tissuespecific manner (Bonomo and Matzinger. 1993). Furthermore, in contrast with cortical epithelium, medullaiy epithelium may be quite efficient in tolerance induction (Hoffniann et al., 1992; Burkly et al., 1993; Klein et al., 1998) so that results obtained in radiation bone inarrow chimeras do not solely address the tolerance-inducing potential of cortical epithelium. Different experimental systems come to seemingly contradictory conclusions even when addressing tolerance induction to similar antigens. While expression of H2-E in medullary epithelimn was sufficient in one case to induce tolerance in VPFj and V P l l CD4 T cells reactive to superantigen, grafting of H2-E positive branchial clefts, which are free of hematopoietic cells, f d e d to tolerize these T cells, although they induced tolerance to I-E itself (Salaiin c>tal., 1992). Furthermore, the potential to tolerize may depend on antigen presentation pathways available to different thvmic APC. For instance, CD4 T cells specific for pigeon cytochrome c were tolerized whether the antigen was expressed as a membrane protein or targeted to mitochondria after synthesis in the cytoplasm (Oehen et ul., 1996). However, expression of P-gal as a nuclear protein only resulted in tolerance induction when it was expressed by medullary epitheliuin, but not bone marrow-derived cells (Oukka et al., 1996). The tolerance-inducing potential of all thymic M HC class I1 expressing cells for a defined self-antigen in a TCR transgenic system was investigated. This was done in reaggregation cultures with fetal thymocytes from C5 T-cell receptor transgenic mice and purified thymic APC isolated from mice that naturally express the self-antigen CS in their blood circulation
234
B R I C; ITTA STOC KI N C, E R
(Volkmannet al., 1997).Because these experiments relied on endogenously processed self-antigen in APC subpopulations purified from thymus of adult CS+donors, they directly addressed the inherent capacity of different thymic APC to internalize, process, and present for T-cell deletion. In this system, cortical epithelial cells and medullary epithelial cells proved as efficient as dendritic cells in deletion; in fact, the only APC population that was unable to tolerize the T cells were macrophages. However, the levels of endogenous antigen in C5 expressing mice are high (about SO pg/ml in blood) so that the comparative efficiency of cortical epithelial cells for negative selection may not have been put to a stringent test. Taken together, the available collection of data suggests that thymic epithelial cells from cortex and medulla can be involved in negative selection, but they may not be as efficient as dendritic cells in low avidity situations (limiting amounts of antigen, inefficient internalization of exogenous antigen, reduced levels or affinity of T-cell receptors). The ability of thymic macrophages to induce negative selection has been put into question before. They were defective in clonal deletion of H2-E reactive T cells (Miyazaka et al., 1993) and their abject failure to tolerize CS specific T cells was notable. Low levels of MHC class I1 and adhesion molecules such as ICAM-1 (see later) presumably contributed to this phenotype. It remains to be tested whether macrophages are able to induce negative selection for MHC class I-restricted T cells.
C. ROLEOF OTHERCELLSURFACE RECEPTORSI N THYMIC NEGATIVESELECTION Given the importance of costimulatory signals in the activation of mature T cells, it appears likely that negative selection requires more than just a signal through the T-cell receptor. The most obvious candidate costimulatory molecules-B7.1 and B7.2 and their interaction with CD28-do not seem as obligatory for thymic deletion as they are for the induction of responses in peripheral mature T cells (Shahinian et al., 1993; Walunas et al., 1996; Tan et al., 1992). CD28 knockout mice do not show impaired negative selection, and the contribution of B7 molecules to negative selection in various in vitro or in vivo models of negative selection remained ambiguous (Punt et al., 1994; Jones et al., 1993; Amsen and Kruisbeek, 1996). This could partly be due to problems with the assays used for negative selection; only one study addressed physiological negative selection in the presence of self-antigen in viuo. In addition, a likely explanation is that there are more costimulatory molecules operative in negative selection. This is supported by the finding that cortical epithelial cells that do not express B7 molecules are efficient APC for induction of negative selection in C5-specific thymocytes (Volkmann et al., 1997). Additional
T LYMPHOCYTE TOI,EHAN(:E
235
candidates are ICAM-1 (Carlowet al., 1992; Pircher et al., 1993;Kishimoto et nl., 1996), CD40 (Foy et nl., 1995), and CD30 (Amakawa et al., 1996). The role of Fas (CDYS), another member of the TNFR family, and TNF in thymic negative selection is ambiguous. Mice deficient in Fas and Fas ligand have normal T-cell development in the thymus and show apparently unimpaired negative selection of superantigen reactive thymocytes, although one study describes a modulating effect of Fas on the early stages of negative selection (Castro et al,, 1996). In addition, Fas appears to play a role in negative selection of “semimature” T cells in the thymus, a population that has undergone positive selection and downregulated one of the coreceptors, but has not acquired full maturity to exit the medulla yet (Kishimoto and Sprent, 1997). In the majority of studies, negative selection of antigen-specific transgenic thymocytes or of superantigen reactive T cells was reported to occur in the absence of Fas or Fas ligand and blocking of TNF did not interfere with negative selection either (Sidman et al., 1992; Parijs et al., Abbas, 1996; Sytwu et nl., 1996; Adachi et al., 1996). Although most of the negative selection experiments were carried out by injection of peptide, with the potential problems of specificity mentioned earlier, one study used HY TCR transgenic mice arid deletion of transgenic thymocytes by endogenously expressed ligand in male mice was found to proceed normally in Fas ligand or Fas-deficient HY TCR transgenic mice (Schwartz, 1997). Thus, whereas evidence that both Fas and TNF play an important role in the deletion of peripheral T cells following activation (activation induced cell death) is unequivocal, available data do not indicate a major role in thymic negative selection. The regulation of negative selection in the thymus is also influenced by thymus-derived steroid hormones. Nur77, a ligand-dependent transcription factor that belongs to the steroid receptor superfamily, is strongly involved in thymic negative selection. Expression of a dominant negative version of this molecule in transgenic mice interferes with the deletion of self-specific thymocytes (Zhou et al., 1996: Calnan et al., 1995). Interestingly, despite defective clonal deletion in the thymus, peripheral selfspecific T cells that are present in increased numbers in such mice were found to be functionally anergic and prone to increased Fas-mediated deletion. What is the quantitative impact of thymic negative selection on the normal T-cell repertoire? In normal mice, expression of endogenous superantigens has been shown to markedly decrease the total number of reactive T cells reaching maturity (Simpson et al., 1993).In bone marrow chimeras expressing MHC molecules on radioresistant thymic epithelial cells, but not on bone marrow-derived cells, the number of mature thymocytes was increased twofold, indicating that about half of the positively selected
236
BRICITTA STOCKINGER
thymocytes may normally be lost through negative selection (van Meemijk et al., 1997). Although this is a substantial number, it is nevertheless marginal if compared with the extent of thymocyte death due to “neglect” (Surh and Sprent, 1994). D. SIGNALING FOR NEGATIVE SELECTION An important question is what dictates whether contact with antigen will result in deletion or in activation of a T cell. Are tliymocytes inherently more “deletable” than mature T cells? Burnet states in his clonal selection theory that the immune system is geared to distinguish between self and non-self. This, however, should not be interpreted to mean that a T cell “knows” what is a self-antigen to which it is supposed to be tolerant and what is a foreign antigen to which it should develop an immune response. Clearly the developmental stage of a T cell is important for that distinction. While mature peripheral T cells can tie deleted (whether contact is with “self” or “foreign” antigen), this process is preceded by at least partial activation if not full-blown differentiation to effector cell status (see Section 11,2). Developing thymocytes, however, receive signals through engagement of T-cell receptor and coreceptor with antigedMHC complexes before developing the capacity to mount an immune response. Curiously, however, signals through the T-cell receptor can have diametrically opposed outcomes in tlie thymus, leading to either positive or negative selection. The current consensus is that the avidity of such signals determines the fate of developing thyniocytes. Thus, any antigen encountered in the thymus whose interaction with T-cell receptors triggers a signal above a certain threshold would initiate deletion. Although it was not possible to directly establish clonal deletion as the underlying mechanism inducing tolerance in the classical cases of neonatally induced tolerance to alloantigens (Billingham et al., 1953; Owen, 1945; Anderson et al., 1951), several subsequent examples of experimental tolerance induction strongly suggested, or even directly demonstrated, this mechanism (Qin et al., 1989; MacDonald et al., 1988; Ciliak and Lelimann-Grube, 1978). It should be pointed out that this interpretation of the neonatal tolerance experiments does not imply that tlie neonatal period per se is more conducive to tolerance induction rather than induction of immunity. Even within the thymus, susceptibility to deletion extends to a certain window of T-cell differentiation only. Once thymocytes have reached the fully mature single positive state, they respond to anti-TCR stimulation in a manner analogous to peripheral T cells (Stockinger et al., 1996; Volkmann et al., 1997; Ramsdell et al., 1991; Nikolic Zugic and Bevan, 1990; Dyall and Nikolic Zugic, 1995; Barthlott et al., 1997). It is likely, but remains to be formally proven,
'r
LYMPHOWTTE TOLERANCE
237
that the signaling machinery in a mature thymocyte resembles that of a mature peripheral T cell, rather than that of a double positive thymocyte. Insights into TCR signaling pathways have emerged through analysis of nonreceptor protein tyrosine hnases expressed in lymphocytes, such as p59"", p56Ick,and ZAP-70, all of which are required for normal T-cell '~ development and signaling. While the presence of ZAP-70 and ~ 5 6 ' are obligatory for positive selection and negative selection in the thymus and essential for signaling in mature T cells (Negishi et al., 1995; Hashimoto et nl., 1996), differences einerged at distal points in the signaling cascade, indicating that T-cell receptors can deliver different types of signals by variably recruiting distinct, downstream signal transduction pathways (Alberola Ila et nl., 1997). Among the large number of molecules that pass on protein kinase-derived signals are small GTP-bindng proteins, notably p2lrAS, which participates in signaling from a wide variety of protein tyrosine kinases. Expression of a dominant negative form of p21ras was shown to block activation of the IL-2 gene in T-cell lines, indicating its important role in relaJing TCR-mediated signals. Transgenic expression of dominant negative p21" profoundly inhibited thymocyte positive selection, but did not interfere with negative selection, suggesting that TCR-derived signals resulting in negative selection may not use a ras-mediated pathway (Swan et al., 1995). Similarly, dominant negative MAP kinase MEKl affected positive, but not negative, selection (Alberola-Ila et al., 1995). The Rho family GTP exchange factor Vav (Turner et d ,1997), which is critical for proliferation and IL-2 production of mature T cells and essential for positive selection in the thymus, only plays a partial role in negative selection. Vav mutants show draniatically reduced calcium fluxes, suggesting that a defect in this pathway accounts for an absence of positive selection and reduced efficiency of negative selection. Blocking of calcium fluxes with the intracelMar calcium chelator has previously been shown to either abolish negative selection (Vasquez et ul., 1994) or at least to block negative selection to weak, but not strong, ligands (Kane and Hedrick, 1996). Thus, in some situations, interfering with a particular signaling component might affect positive and negative selection differentially because of quantitative differences in the strength of the signals required for these two processes. However, a mere quantitative effect has been excluded, at least for the involvement of the ras signaling pathway, as complete inhibition by simultaneous expression of dominant negative ras and MEKl still only affected positive, but not negative, selection ( Alberola-Ila et al., 1995). E. ESCAPEFROM THYMIC; DELETION While deletion of self-specificthyrnocytes might be considered the major component in the induction of immunologiical tolerance, ample evidence
238
BHIGITTA STOCKINGER
shows that thymocytes can escape from deletion, even if self-antigen is present in the thymus. Escape from thymic deletion is possible for thymocytes that have decreased levels of either coreceptor (Teh et al., 1989) or T-cell receptor (Schonrich et al., 1991; Mamalaki et al., 1995), which may result in decreased avidity and therefore signals that are not sufficient to induce apoptosis. However, it is possible that an apparent phenotype of receptor downregulation is due to the expression of a second endogenously rearranged T-cell receptor in addition to the transgenic T-cell receptor. Such endogenous rearrangements, leading to the expression of additional T-cell receptor a chains, are prominent in most TCR transgenic mouse strains not backcrossed to Rag-/- and can influence the degree of negative selection. Expression of a second T-cell receptor in addition to a selfantigen-specific TCR has been shown to allow escape of T cells from thymic deletion despite the presence of abundant levels of self-antigen (Zal et al., 1996). This potential drawback does not apply to the mice described by Mamalah et al., (1995). In double transgenic Ragl-/- mice expressing an NPspecific CD8 T-cell receptor and endogenous NP peptide, a proportion of CD8 cells appears to be deleted in the thymus, but there are CD8 cells in the periphery. These have reduced levels of receptor and coreceptor and characteristically upregulated levels of CD44, a marker attributed to memory or activated cells (Stout and Suttles, 1992; Budd et al., 1987). Likewise, double transgenic mice that express the antigen influenza hemagglutinin (HA) under control of the K light chain promoter and CD4 T cells specific for HA have antigen nonresponsive CD4 cells in the periphery with increased levels of CD44 (Lanoue et al., 1997). Interestingly, in all reported cases of escape from thymic deletion, despite the presence of self-antigen in the thymus, the escapees in the periphery were functionally inert. In the case of HA and NP-specific T cells their state of unresponsiveness must have been imprinted in the thymus, as already the mature single positive thymocytes abnormally expressed the activation markers found on the peripheral T cells. In other cases (e.g., Zal et al., 1996), dual TCR expressing T cells escaped from thymic deletion in the presence of abundant levels of self-antigen because their levels of the self-antigenspecific (C5) TCR were so low that they remained ignorant of its presence. There is no systematic analysis concerning the life span of peripheral escapees. If there is any correlation with the B-cell system one could assume that functionally unresponsive peripheral T cells probably have a short half-life.
F. CAVEAT. Is OURPERCEPTION OF CLONAL DELETION IN THE THYMUS AN ARTIFACTOF TRANSGENIC MODELS? In the B-cell system an intriguing mechanism preventing wholesale loss of B cells through negative selection has been described, termed receptor
T LYMPHOCYTE TOLEHANCE
239
editing. The original papers based on two different B-cell receptor transgenic mouse models stated that editing of light chains at the pre-B-cell stage allows survival of a proportion of B cells that otherwise would have expressed a self-specific receptor and be deleted (Gay et al., 1993; Tiegs et al., 1993). Pelanda et al. (1997) illustrate that B cells with a self-specific receptor that had been targeted into the endogenous locus by homologous recoinbination are not subject to deletion, but undergo very efficient receptor editing. In conventional, noiitargeted B-cell receptor transgenic mice the extent of clonal deletion might have been overestimated, as the transgenes cannot be deleted by Ig gene rearrangements. If this situation is applied to T-cell selection in the thymus, it is immediately obvious that the occurrence of T cells carrying receptors with other TCR a chains, due to incomplete allelic exclusion at the TCR a locus, are far more prominent in mice that express the self-antigen recognized by the transgenic TCR. This is usually interpreted to be due to the accumulation of such cells in the face of massive deletion of autoreactive cells. No TCR transgenic mouse exists so far that was created by homologous recombination into the correct locus, and the number of T cells with endogenously rearranged receptors varies widely in different TCR transgenic models. This begs the question of whether the assumption of thymic deletion, similar to the B-cell situation, may be an overestimate. Given the fact of massive loss of T cells through Failed positive selection, it might be preferable to avoid deletion following self-antigen recognition by an editing mechanism, such as additional rearrangements of TCR a chains, which would allow a second chance to create an innocuous T-cell receptor. This would be a way to avoid “holes in the repertoire” created by eliminating a wide range of cross-reactive receptor specificities. The existence of autoimmune diseases prompts the question of whether this constitutes failure of thymic negative selection or breakdown of control mechanisms in the periphery. It stands to reason that it must be impossible for every self-antigen in the body to have access to the thymus for tolerance induction. T cells with specificity for autoantigens can be found in the repertoire of healthy humans (Ota et nl., 1990; Pette et al., 1990; Salvetti et nl., 1991; Sommer et al., 1991).Surprisinglythough, a nuinber of presumed tissue-specific proteins have been found to be expressed in the thymus. A number of protein structures considered specific components of the central nervous system are contained in the thymus (Mathisen et al., 1993; Pribyl et al., 1996: Fritz and Zhao, 1996). Transgenic mice constructed with regulatory elements from the insulin gene can express transgenes both in pancreas and in thymus. This was initially thought to reflect the consequence of the promiscuous expression of hybrid insulin-promoter transgenes, but in the meantime thymic expression was demonstrated for a nuinber of pancreas-specific genes, including endogenous, nontransgenic
240
BRIGI'ITA STOCKINGER
elements. (Jolicoeur et nl., 1994; Smith et al., 1997; Antonia et al., 1995). It is impossible to decide whether autoimmunity can arise as a consequence of changes in levels or onset of thymic expression of tissue-specific antigens and whether the target structures recognized reflect thymic expression. In this context the success of experimental tolerance induction affecting dlabetes, by injection of a single protein, such as GAd or insulin (reviewed by Wegmann, 1996), is certainly perplexing, Taken together, it must be assumed that T cells carrying receptors specific for autoantigens are present in the periphery. Whether those T cells ever make contact with self-antigen and what mechanisms exist to prevent their activation and progression to autoimmune disease are the questions facing us under the topic of peripheral tolerance. 111. Peripheral Tolerance
A. T-CELLIGNORANCE, T-CELLRECIRCULATION There must be many self-antigens expressed in peripheral tissues that never access the thymus and therefore cannot be expected to have induced deletion of their specific T cells, yet autoimmune disease is a relatively rare phenomenon. This means that there are effective ways of avoidlng the activation of potentially self-reactive T cells. Conceptually the simplest of possibilities is to avoid contact of T cells with their self-antigen on peripheral tissue. This is termed T-cell ignorance and at its basis is the particular pattern of lymphocyte recirculation, which differs crucially for naive T cells compared with effector and memory T cells (Mackay, 1993). Naive T cells recirculate through secondary lymphoid tissue, but do not access extralymphoid sites such as skin, joints, or pancreas. In contrast, once T cells are activated, they will recirculate through lymphoid and extralymphoid sites (reviewed in Butcher and Picker, 1996). A striking example of T-cell ignorance was described by Ohashi et al. (1991) in a transgenic system. Transgenic mice expressing lyinphochorionieningitis (LCMV) viral glycoprotein in pancreas p islet cells were crossed with mice carrying a transgenic T-cell receptor specific for a peptide within the LCMV glycoprotein. The transgenic T cells remained unresponsive (tolerant) and phenotypically naive unless they were infected with LCMV virus. Infection abolished unresponsiveness and resulted in CD8 T-cell-mediated diabetes. However, T-cell ignorance was not always the outcome when antigen was expressed in the pancreas and the differences may be due to the type of T cell (CD8 or CD4) and the quality of its TCR, as well as its affinity and the amount of antigen expressed (Lo et al., 1992; Morgan et al., 1996: Degermann et al., 1994; Scott et al., 1994; Morahan et a/.),
T LYMPHOCYTE: TOLERANCE
24 1
Some organs, such as the eye, brain, and testis, are considered immuneprivileged sites because antigens expressed in them do not induce immune responses. These organs maintain even stronger barriers to the entry of recirculating lymphocytes than other nonlymphoid sites. Apart from immunological ignorance, mechanisms invoked for maintaining tolerance to such “sequestered” antigens include a switch to liumoral rather than cellmediated immunity (see Section II,D) and the presence of soluble mediators with suppressive characteristics, e.g., TGF-P (Ksander and Streilein, 1994). In cases where ignorance prevails it might be explained with differential migration patterns, as naive T cells are not expected to get into contact with their self-antigen if it is not present in lymphoid organs. However, a number of observations cloud this simplistic view of peripheral tolerance. First, there seems to be large-scale traffic of virgin T cells through extralymphoid sites in early ontogeny (demonstrated in the sheep fetus; Kimpton ct al., 1995), which would indicate that there is contact between naive T cells and tissue-specificantigens early in development. Thymic emigrants were shown to migrate to tlie skin within a few hours after exiting the thymus and it was suggested that these cells might be more susceptible to tolerance induction than circulating peripheral T cells (Washington et al., 1995). Furthermore, it was shown in a transgenic model that neonatal, but not adult, T cells had access to alloantigen expressed exclusively on keratinocytes (Alferink et nl., 1998). The accessibility of peripheral sites appears to be due to a developmental switch in expression of the lymphocyte homing receptor and adhesion molecules, favoring MAdCAM as addressin and a4 p 7 as homing receptors in fetal and early neonatal life (Mebius et nl., 1996). If this is the case one might expect contact of potentially self-reactive T cells and self-antigen on peripheral tissues, at least during a short period in life. Recognition of antigen on nonprofessional APC. such as tissue cells, is unlikely to induce a productive immune response due to the lack of signal 2 (a costiinulatory signal). Such a recognition event may have multiple potential outcomes, such as ignorance, anergy, or partial activation, resulting in differentiation to a regulatory subset. The various mechanisms invoked for induction and maintenance of peripheral tolerance will be discussed in detail. Even more perplexing is the role of dendritic cells in the activation of T cells with specificity for self-antigens expressed on peripheral tissues. To what extent do dendritic cells acquire such antigens, do they present them to T cells, and where? Polly Matzinger, in her “danger” hypothesis, makes tlie case that a productive immune response against any antigen, be it intrimic or foreign to an organism, requires a danger signal that activates APC, notably dendritic cells. The absence of such signals (which
242
BRIGI’ITA STOCKINGER
constitute bacterial or viral products, injury, necrotic death, inflammatory cytokines, etc.) in a quiescent immune system would preclude induction of a T-cell response and thus guarantee self-tolerance in the periphery (Matzinger, 1994). Because the exact nature of the danger signals resulting in activation and migration of dendritic cells are very difficult to define, it is not known if there is a degree of “constitutive” traffic of dendritic cells carrying peripheral self-antigen into draining lymph nodes. A report by Kurts et al. (199713) suggests that dendritic cells (or bone marrowderived APC at least) carry ovalbumin expressed in the plasma membrane of pancreas p cells into the draining lymph nodes. There is apparently no indication of an insult to the pancreas resulting from overexpression of this foreign protein and no inflammation, suggesting constitutive migration of dendritic cells transporting this antigen into lymphoid organs. If this was routinely the case and if dendritic cells are constitutively able to offer costimulatory signals, one might expect a much larger potential for the induction of autoimmune recognition events. More recently, such cross presentation for the activation of CD8 T cells by dendritic cells has been demonstrated to occur as the result of apoptosis, but not necrosis, of cells expressing the antigen in question (Albert et al., 1998). Curiously, no similar clear-cut data exist with respect of antigen pick up by dendritic cells for the activation of CD4 T cells, although one might have thought that it should be easier for exogenous antigens to reach the conventional class I1 presentation pathway rather than the class I presentation pathway. Although the jury is still out to decide whether overexpression of ectopic proteins in various organs will cause disturbance of normal cell physiology, resulting perhaps in increased cell turnover and release of cellular constituents, another point of consideration is the status of maturity of the antigenpresenting dendritic cells. Resting tissue dendritic cells do not constitutively express high levels of MHC class I1 and B7, although these molecules can be upregulated very rapidly. Is it conceivable that constitutively migrating dendritic cells fail to get signals inducing their maturation and that such dendritic cells are tolerogenic rather than immunogenic? It will be very difficult to devise experiments to test this ponit as it is currently impossible to separate dendritic cell migration and maturation or even isolate dendritic cells avoiding their activation.
B. PERIPHERAL DELETION, EXHAUSTION Studies of immune responses to endogenous and exogenous superantigens provided compelling evidence that contact of mature T cells with antigen can lead to their deletion (Webb et al., 1990; Jones et al., 1990; Critchfield et al., 1994; Kawabe and Ochi, 1991).The notion that prolonged exposure of lymphocytes to antigen can lead to “exhaustion” has been
T LYMPIIOCYTE TOLERANCE
243
around much earlier (Byers and Sercarz, 1968; Sterzl and Silverstein, 1967), but in the age of transgenic mice it became possible to study the fate of individual T cells. Deletion of mature T cells as a consequence of excessive stirnulation or antigen persistence is now widely demonstrated not just for superantigens, but also for conventional antigens (MacDonald et al., 1991; Moskophidis et al., 1993; Rocha and von Boehmer, 1991; Kyburz et al., 1993; Bogan, 1996). In contrast to the deletion of iininature thymocytes, mature T cells undergo a period of activation and even expansion before they are deleted, although this may have limited functional consequences (Renno et al., 1995; Forster and Lieheram, 1996; Kurts et al., 1997b; Ehl et al., 1998). Clearly, high antigen dose and chronic stimulation favor elimination of both CD4 and CD8 T cells. The silencing of T cells following persistent engagement of their antigen receptors in the periphery may represent a continuous process, ranging from activation to unresponsiveness to deletion, depending on signal strength and exposure time. Induction of peripheral deletion can be prevented or even reversed in the presence of proinflamniatory substances, such as lipopolysaccharide or poly : IC, or infections with VSV, Listeria, vaccinia virus, or Nipostmngylzis bmsiliensis (Ehl et al., 1998; Chiller and Weigle, 1973; Rocken et al., 1992; Vella et al., 1995), emphasizing the role of additional stimuli other than antigen for the generation of productive immune responses ( Janeway et al., 1996; Matzinger, 1994). One mechanism of deletion is terminal differentiation to short-lived effector cells that may die by apoptosis due to lyinphokine depletion. The latter mechanism could be prominent for CD8 T cells. The absence of CD4 help (Kirberg et al., 1993; Guerder and Matzinger, 1992; K~irtset a?.,1997a)promotes tolerance induction by deletion in the CD8 population, and conversely CD4 T cells can prolong the survival of CD8 T cells, perhaps by supplying cytokines, such as IL-2. Both Fas- and TNF-mediated cell death can contribute to the deletion of effector T cells (Sytwu et aZ., 1996: Crispe, 1994; Parijs et nl., 1996). In addition, a feedback regulatory mechanism that controls the intensity of immune responses is involved in the susceptibility of mature T cells to programmed cell death. The latter operates when T cells are exposed to large concentrations of antigen under fully activating conditions and results from the exposure to IL-2 followed by cell cycle progression (Lenardo et a1 , 1995). The physiological significance of the latter mechanism in self-tolerance is unclear, but it has been proposed as a strategy for high dose antigen therapy of autoimmune encephalomyelitis (Critchfield et a1 , 1994). C. ANERCVA N D PARTIAL SIGNALING: I N T CELLS The activation of T cells requires at least two signals: a signal through the T-cell receptor and a second signal presumed to involve a costimulatory
244
RHI(;ITTA STOCEINGEH
molecule(s) on an antigen-presenting cell (Bretscher and Cohn, 1970; Lafferty et nl., 1983; Mueller et al., 1989). The antigenic signal alone cannot activate T cells and has been shown to result in unresponsiveness, termed anergy (Mueller et al., 1989; Schwartz, 1990; Boussiotis et al., 1994b). Classical T-cell anergy has been defined in vitro using Thl T-cell clones (Jenhns and Schwartz, 1987; Schwartz, 1997); anergic cells were reported to have a block in p21 ras activation (Fields et nl., 1996), decreased activation of several MAP kinases (Li et al., 1996), reduced c-Fos and JunB protein induction (Mondino et al.. 1996), and fail to phosphorylate the AP-1 protein required for IL-2 gene activation (Kang et nl., 1992). Anergy induced in vitro can be reversed by IL-2 and by signaling through the common y chain of the IL-2, 4,and 7 receptors and does not lead to cell death (Boussiotis et al., 19944. While the molecular mechanisms leading to anergy in T cell clones are quite well defined, the situation is far more complex in v i m Anergy is used as a generic term for nonresponsiveness and the underlying reasons and mechanisms for the failure to respond can be manyfold. Anergy in vivo is well documented in the B-cell system; transgenic B cells developing in the presence of soluble self-antigen (HEL) are functionally unresponsive due to a biochemical block in the tyrosine kinase-signaling cascade initiated by IgM and IgD receptors. This inhibits the accumulation of phosphotyrosine on the receptor-associated CD79a and p chains and on the collaborating syk tyrosine kinase (Cooke et al., 1994). Anergic cells were shown to have a very short half-life and failed to enter follicles when competing with normal B cells (Cyster et al., 1994): thus B-cell anergy can be viewed as a slow form of deletion. Induction of tolerance in mature peripheral T cells has a long-standing history. In early studies, tolerance was induced by injection of high or low doses of soluble, deaggregated proteins without any adjuvants (Dresser, 1961;Chiller et al., 1971; Mitchison, 1964; Romball and Weigle, 1993), but the success of this strategy was variable and not extendable to every protein. The underlying mechanisms remained undefined, but could be explained by more recent evidence that antigens presented to T cells in the absence of inflammation (as caused by adjuvants) tend to be tolerogenic (Janeway et al., 1996; Matzinger, 1994; Kearney et d.,1994; Pape et a!., 1997; Kyburz et a/., 1993). Additional mechanisms resulting in high dose tolerance were discussed in Section I1,B. The CTLA-4 molecule, which is induced in T cells on activation, has been invoked in the control ofperipherd tolerance (Chambers et al., 1996). CTLAWB7 interactions regulate T-cell expansion, and CTLA-4 deficient mice have a profound lymphoproliferative disorder (Waterhouse et nl., 1996). It is clear that the uncontained proliferation in CTLA-4 deficient mice is attributable to CD4 T cells (Chambers et al., 1997), suggesting
T LYMPHOCYTE TOI.ERANCE
245
that this molecule is particularly important in the homeostatic control of CD4 T-cell responses. CTLA-4 ligation results in the inhibition of IL-2 production and consequent arrest in cell cycle progression (Kruminel and Allison, 1996). CTLA-4 could also compete for CD28/B7 interactions and thus deliver negative signals, resulting in T-cell mergy (Thompson and Allison, 1997). Induction of anergy following priming with peptide in incomplete Freunds adjuvant was prevented in the presence of antiCTLA-4 antibodies (Perez et d.,1997), suggesting that some form of costimulation is involved in the induction of anergy. The presence of a CTLA-4-specific antibody before the onset of insulitis in mice with a diabetogenic transgenic T-cell receptor precipitated the onset of diabetes dramatically, whereas antibody injection at later stages had no effect S, engagement of CTLA-4 at different (Luhder et al., 1998).T ~ L Idifferential time points in ongoing T-cell responses can have distinct consequences for antigen recognition. One cell type that is frequently associated with the induction of anergy in T cells are B cells. They are shown to be dispensable for the initial priming of naive T cells in viuo (Lassila ct al., 1988; Ronchese and Hausmann, 1993), although some aspects of T-cell priming seem to be impaired in mice lacking B cells (Kurt-Jones et nl., 1988; Ron and Sprent, 1987). Moreover, presentation of antigen by B cells in uivo does not seem a neutral event, but instead induces tolerance (Eynon and Parker, 1992; Fuchs and Matzinger, 1992). It is noteworthy that these experiments all involved antigen nonspecific B cells that either internalized antigen via nonspecific endocytosis or expressed it intracellularly (H-Y antigen in inale B cells). This may make a crucial difference for the functional activity of B cells. First, nonspecific antigen uptake is very inefficient in B cells unless they are activated (Chesnut et a)., 1982), whereas antigen-specific B cells internalize large amounts of antigen via Ig specific endocytosis (Lanzavecchia, 1990). Not only does this guarantee increased occupancy of MHC class I1 molecules with a specific peptide, but cross-linking of Ig receptors (Finkehnan et a1 , 1992) and antigen internalization via Ig receptors also results in the upregulation of B7 inoleciiles (Ho et al., 1994). B cells that do not signal effectively tlirough their antigen receptors fail to activate T cells for CD40 ligand upregulation, IL-2 secretion, and proliferation. However, they may still affect partial activation, resulting in upregulation of activation markers such as CD69 and CD44 (Ho et al., 1994; Croft ct al., 1997). This phenotype, as mentioned earlier, seems to be a hallmark of anergic B cells. Taken together, one may have to view with caution a general assumption that B-cell antigen presentation to naive T cells in uivo will cause anergy, as long as this has not been tested with antigen-specific B cells. B cells that encounter antigen that they can internalize via their
246
BRIGITTA STOCKINGER
Ig receptors may be as capable of inducing productive T-cell responses as dendritic cells. However, high doses of protein antigens or peptides injected in vivo are likely to be targeted to B cells not expressing cognate Ig receptors, given the high frequncy of B cells compared with antigenspecific B cells or with dendritic cells, and would not result in effective B7 upregulation. Under these circumstances it is conceivable that CTLA-4 competes with CD28 for limited B7 molecules so that full activation of T cells is stalled through the curtailment of IL-2 production. The life span of anergized peripheral T cells within the same host (following thymectomy) has not been studied systematically. Instead, the fate of adoptively transferred T cells that were either activated or anergized in the adoptive host has been determined (Rocha et al., 1995). These experiments were performed with donor T cells from Rag-/- transgenic mice, thus avoiding the complication of endogenous, additional T-cell receptors contributing to the survival of such cells (Tanchot and Rocha, 1997). MHC class I-restricted, H-Y-specific T cells are anergized when transferred into host mice that contain the male antigen on every cell in the body. Anergized cells persist for long time periods, provided the host thymus does not generate new T cells. In contrast, they disappear gradually in mice that export T cells into the periphery. This indicates that anergic T cells may have a reduced life span, not necessarily because of intrinsic properties predisposing them to apoptosis, but rather because, like anergic B cells, they lose out in the competition for space in lymphoid organs (Mondino et al., 1996). Is anergy reversible? Anergy induced in T-cell clones in vitro is reversible with IL-2 (Beverley et al., 1992),but this is not usually the case for anergic T cells in vivo (Rocha and von Boehmer, 1991; Lanoue et al., 1997). It is widely believed that a state of T-cell anergy is dependent on the continuous presence of antigen and is reversible upon removal of antigen. This was usually tested by adoptive transfer of previously anergized cells into an antigen-free environment (Rocha et al., 1993; Bachmann et al., 1994; Alferink et al., 1995).This approach can hardly be considered physiological, but an analysis by Tafuri et al. (1995) provided evidence for a transient antigen-specific state of tolerance to paternal alloantigens during pregnancy. Numbers of T cells carrying receptors specific for the paternal alloantigen decreased in thyinectomized pregnant mice and reverted to normal levels after delivery. It was not possible, however, to attribute the changes in specific T cells to a particular mechanism such as anergy, receptor modulation, or deletion because of the large nonspecific fluctuations of spleen cell numbers typical during pregnancy. Ectopic expression of antigen on tissue cells devoid of costimulatory molecules has been used as a means to induce antigen-specific unrespon-
T L Y hl P I 1O( :ITE TO L.EH A NC E
247
siveness. Numerous models of organ-specific TCR transgenic inice were developed whose phenotypes with respect to tolerance induction vary substantially (Adams et al., 1987; Boliine and Pilstrom, 1991; Miller ct al., 1989; Murphy et a/., 1989; Loet c d . , 1991,Cruerderet al., 1994; Hiii-ninerling et nl., 1991).Crosses of mice expressing antigens in a tissue-specific manner with TCR transgenic inice have made it possible to analyze phenotypic and functional effects on specific T cells. Evidence for antigen encounter by peripheral T cells usually is the expression of activation markers such as CD44 or CD69. Graded degrees of peripheral T-cell unresponsiveness, accompanied by T-cell receptor downregulation, were observed in TCR transgenic mic; with tolerance to K” expressed in the liver, glial cells, or slun keratinocytes (Schonrich et (11, 1991, 1992; Ferber et al., 1994). Similarly, the pattern of antigen expression in the thymus and periphery for HEL-specific transgenic T cells resulted in a spectrum of T-cell unresponsiveness, ranging from elimination in the thymus to impaired function of peripheral T cells (Akkaraju Pt al., 1997).In other cases, T cells appeared ignorant of the presence of self-antigen, notably in the case of inyelin basic protein (Goverman et al., 1993; Lafaille ct al., 1994; Liu et d.,199Fj), which might have been due to low avidity of the TCR and/or inefficient presentation of the self-peptide. Apart from the properties of the different T-cell receptors, a likely factor in tlie outcome of antigen recognition by peripheral T cells is the nature and size of the organ expressing self-antigen and the amount of self-antigen present. Th~is,it was suggested that organ vasculature may influence the availability of antigens to the immune system. Antigens expressed on hepatocytes may induce tolerance more easily than those on skin or brain because access-for circulating T cells is easier. In addition, tlie difference in inass of the thyroid coinpared with tlie pancreas was cited as a likely factor for determining the inore pronounced tolerizing effect of the thyroid in this system (Akkaraju et al., 1997). Yet another factor influencing peripheral tolerance induction could be the onset of expression of self-antigens. In a transgenic model of pancreasspecific expression of SV40 T antigen, T cells were tolerant when SV40 T was expressed in the fetal stage. If the antigen was expressed postiiatally only, T cells remained active and caused spontaneoiis autoimmunity (Adams et al., 1987; Geiger ct al., 1992). This might be explained by the different migration patterns of recent thymic emigrants compared with naive, peripheral T cells inentiolied earlier (Section 11,A). It should be stressed, however, that in the case of overexpression of SV40 T or of MHC class I molecules (Miller et al., 1989),the expression per se caused dysplasia and cell death in the pancreas, which would be sufficient to provide a source of self-antigen that could be presented by professional APC such as dendritic cells. Therefore, any potential tolerizing effect of peripheral
248
BHI(;I‘ITA STOCKINGER
tissue cells could have been overwhelmed by the dissemination of selfantigen. Developmentally regulated self-antigens under normal conditions may fail to induce immune responses in the absence of inflammatory stimuli.
D. IMMUNE DEVIATION The term “anergic” suggests functional disability, yet over a number of years anergic T cells have been attributed reactivities that are not compatible with complete inertia. An interesting observation was downregulation of Thl characteristics and a shift to Th2 development in Tho clones treated with “anergizing” stimuli (Gajewski et al., 1994), which suggested that anergized T cells, rather than becoming functionally inert, may instead switch their response potential to a different functional phenotype. A direction of research for which the term “immune deviation” (Asherson and Stone, 1965) was used described a shift from harmful, destructive immune responses to a response mode that did not cause pathology (Rocken and Shevach, 1996).In nonobese diabetic (NOD) mice that spontaneously develop diabetes, the presence of Th1 cells in the islets correlated with disease, whereas the presence of cells producing Th2 cytokines conferred resistance to disease (Liblau et nl., 1995). In a transgenic diabetes model, aniinals with a certain genetic background did not develop autoimmune disease, although their pancreas tissue was infiltrated with activated T cells. This phenotype was attributed to a preferential induction of Th2 rather than Th1 subsets (Scott et al., 1994). Immune deviation is the active mechanism preventing immune responses in the eye (Streilein et al., 1997). In some autoimmune models it was possible to block the induction of disease by the transfer of T cells or thymocytes expressing TI12 cytokines (Powrie and Mason, 1990; Saoudi et al., 1993: Powrie et al., 1994), suggesting immune regulatory properties of the injected cells. In other cases, a regulatory role by Th2 cells was not observed. Thus, it was shown that whereas Th2-like responses correlated with protection from diabetes, these cells could not prevent disease induction by Thl cells (Katz et al., 1995). This illustrates that the view of harmful Th 1 responses being tolerized by a regulatory mechanism involving immune deviation may be an oversimplification.
E. REGULATORY T CELLS After a decade of intensive research into suppressor T cells and suppressor factors, these cells disappeared from center stage following the discovery that the I-J subregion, which supposedly encoded their restriction element (Dorf and Benacerraf, 1985), was not present in the H-2 locus (Kronenberg et nl., 1983). The following interesting statement from a
review on antigen-specific siippressor factors illustrates tlie feelings of a large section of the immunological coininunity: “ F ~ t(Ireas c ofiiizniuriologic. research havc endured srich ctridcnt criticism or oigcnrlered siich faintheartecl ~ y i p o rCIS t the vtiidy of antigeii-Ypw$c \iipprecsion $the i i i i n i i i n c response. . . 1n a very real &iw, tlioce w ~ i pc$)niie(~ o iiiric~ioftlie early work in the j e l d hear responddity Fir the ozitccist statiiy of suppression. With the increciying number of soliiblc; iiier1intor.r c i n d cavcarles qf iiitcr(icting T cells, icliicIi poliidaterl rcciezcc of tlic srihj,jcct in the 198Os, the concept of antigen-s pecijic riipprmsion c i n d sripprewor -factors .siiiiphj hcctimc too complicated and w m disinisscd us artifiict” (O’Hara, 1995). Nevertheless, the notion that immune regulation is the underlying caiise for sonie phenotypes of tolerance sunived. I n contrast to so-called passive mechanisms of tolerance induction, such a s deletion or anergy, immune regulation can be considered an active or doniiiiant form of immunological tolerance. As such, immune regulation offers far wider potential to interfere with tlie autoiminune activation of T cells, which explains the large nuinber of ex-perimental approaches that are uwd to investigate its mechanisms. A number of these approaches ‘ire reviewed in Iiiiinunological Reviews 149 (1996), and will not be discussed in much detail here as the subject area is beyond the scope of this review. One experimental system in wliich the influence of regulatory T cells is invoked is autoimmunity induced by neonatal thymectoiny or generally by the induction of lyinphopenia. This treatment induces a multitude of organspecific autoimmune diseases that can be cured by the transfer of T cells or thyinocytes (reviewed in Gleeson c’t ( I / . , 1996): It is argued that there may be a subset of thymocytes that tire programmed for the control of self-reactive T cells (Saoudi Lit al., 1996: Modigliani et nl , 1996; Le Douarin et al., 199G), but details of their specificity and way of differentiation remain unclear. Irrespective of the problem if and how regulatory T cells are educated in th e t h yin 11s, the re are n u 111e rou s ways for reg11latory in echanisin s invoked in the periphery. Active suppression by regiilatory T cells is a mechanism invoked in oral tolerance (reviewed in M’einer, 1997) and this involves a number of inhibitory cytokines, such as IL-4, IL-10, and TGF-P so that the mode of action is via secreted nonantigen-specific “suppressor factors.” Tolerance to transplantation antigens can be induced if the source of antigen is given at the same time as antibodies to either CD4 or CD8 (Cobbold et d., 1996).This type of tolerancc is “infectioiis,” i.e., it can be transferred by T cells. The primary effector of rejection and the regulatory population both belong to the CD4 T-cell subset, and it appears as if sonie kind of “infectious anergy” induction is underlying the regulation. More recent experimental data may provide a link between cytokine-mediated
250
BRlGlTTA STOCKINGER
suppression and anergy. In addition, they give insight into the problem of the antigen specificity of regulatory T cells, as even in TCR transgenic models where it should be possible to identify T cells with a given antigenic specificity, it often appears that T-cell regulation is found in Rag+ TCR transgenic mice, but not in their Rag-/- counterparts (Forster and Lieberam, 1996: Alferink et al., 1998). This could be taken as evidence that T cells with other specificities than the transgenic receptor generated by endogenous TCR rearrangements are involved in regulation. Alternatively, it could simply indicate that active regulation is difficult to detect in Rag-/- transgenic mice due to the large numbers of monospecific T cells. A Rag-/- TCR transgenic model has been described where T cells specific for hemagglutinin (HA) escape from thymic deletion, but become anergized when crossed with mice expressing HA under control of the immunoglobulin-rc promoter. Anergized HA-specific T cells had impaired ability to cause diabetes following adoptive transfer into mice expressing HA on pancreas islet cells, whereas HA-specific T cells (isolated from mice that did not express endogenous HA) caused diabetes rapidly. Impairment of autoreactive potential was linked with the secretion of large amounts of IL-10 by anergic T cells (Buer et ul., 1998). This cytokine has a longstanding reputation for the downregulation of immune responses and could therefore qualify as a “suppressor factor.” IL-10 was reported to interfere with the costimulatory activity of APC (Ding et al., 1993), the dendritic cell-induced interferon-y production by T cells (Macatonia et al., 1993), or IL-2 production and IL-2R a chain upregulation (de Waal Malefyt et al., 1993; Groux et al., 1996). IL-10 secreting ovalbumin-specific T-cell clones prevented autoimmune inflanirnatory bowel disease when coinjected with the disease-initiating CD4+CD45RBh’subset (Groux et al., 1997). Future experiments, perhaps taking advantage of the fact that antigenspecific T cells can be identified in transgenic systems, should allow clarification of whether particular modes of antigen stimulation, such as stimulation by APC lacking costimulatory molecules or partial triggering of T-cell receptors resulting in altered signaling patterns, can cause a regulatory phenotype within cells of a given specificity. The intriguing possibility is that such regulatory cells may act on other members of the same clone that have not been exposed to the “tolerizing” stimulus themselves and switch their normal response pattern.
ACKNOWLEDGMENTS I thank Liz Simpson, Rose Zamoyska, and Dimitris Kioussis for critical comments on the manuscript. Among their corrections, they all pointed out a few references that were missing. I suspect that the reference list, however long it is, inevitably left out some citations
T LYMHIOCYTE TOLEHANCE
25 1
others might liave considered important. I have cited rebiews wherever applicable so that the reader can obtain original references or further details from those. If there are omissions in the reference list, I can only stress that they have not been intentional. I am grateful to the members of my laboratory for their patience with nie during the final weeks ofwriting.
REFE HENC:ES Adachi, M., Sueniatsu, S., Suda, T., Watanahe, D., Fukuyima, H., Ogasawara, J., Tanaka, T., Yoshida, N., and Nagata, S. (1996). Enhanced and accelerated lyinplioproliferation in Fas-null mice. Proc. Nntl. Acad Sci. USA 93, 2131-2136. Adam, T. E., Alpert, S.. and Hanahan, D. (1987). Non-tolerance and autoantibodies to a transgenic self antigen expressed in pancreatic beta cells. Nottire 325, 223-228. Akkaraju, S., Ho, W. Y., Leong, D., Canaan. K., Davis, M. M., and Goodnow, C . C. (1997). A range of CD4 T cell tolerance: Partial inactivation to organ-specific antigen allows nondestructive thyroiditis or insulitis. Zittrnurtity 7, 255-271. Alberola, Ila, J., Takaki, S., Kerner, J. D., and Perlmutter, R. M. (1997). Differential signaling by lymphocyte antigen receptors. AIIW. Rez;. Z i t i i i t u n d . 15, 125-154. Alberola, Ila, J., Forbush, K. A , , Seger. R., Krebs, E. G.. and Perlmutter, R. M. (1995). Selective requirement for MAP kinase activation in thyinocyte differentiation. Nature 335, 620-623. Albert, M. L., Sauter, B.. and Bhardwaj, N. (1998). Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs N ~ t i r r e392, 86-89. Alferink, J., Schittek, B., Schonrich, G., Hammerling, G. J., and Arnold, B. (1995). Long life spin of tolerant T cells and the role of antigen in maintenance of peripheral tolerance. Int. Zitmu1101 7, 331-336. Alferink, J., Tafuri, A,, Klevenz, A,. Arnold, B., and Hiininierling, G. J. (1998). Tolerance induction in mature T lymphocytes. Nooorti.~Fouitdotiott Syinposiuiti 215, 191- 196. Amakawa, R., Hakem, A,, Kundig, T. M.. Matsnyiina, T., Siniard, J. J., Tiinnis, E., Wakehani, A,, Mittnlecker, H. W., Griesser. II., Takimoto, H., Schniits, R., Shahinian, A,, Ohashi, P., Penninger, J. M., and Mak, T. W. (1996). Impaired negative selection of T cells in Hodgkin’s disease antigen CD30-deficient mice. Cell 84, 551-562. Anisen, D.. and Kruisbeek, A. M. (1996). CD28-I37 interactions function to costiinulate clonal deletion of double-positive thymocytes. Ziit. Z i t t i t ~ u i i d 8, 1927-1936. , E., Lainpkin, G. H., and Meda\yar. P. B. (1951). The use of Anderson, D.. B i l h ~ h a m R. skin grafting to distinguish between nionozygotic and dizygotic twins in cattle. Heredity 5.379-397. Antonia, S. J.. Geiger, T.. Miller, J., and Flavell, K . A . (199.5). Meclianisrns of iiniriune tolerance induction through the thymic exprcssion of a peripheral tissue-specific protein. znt. ZIlt?tiU?tO~.7, -715-725. Asherson. G. L., and Stone, S. I I . (1965). Selective and specific inhibition of 24 hour skin reactions i n the guinea-pig. Immune deviation: Description of the phenomenon and the effect of splenectoiny. Ziti~ttuirology9, 205-217. Auphan, N., Schonricli, G., Malissen, M., Barad, M . , Hlinmerling, G. J., Arnold, B., Malissen, B., and Schmitt-verliulst, A.-M. (1991j . Influence of antigen density on the degree of clonal deletion in T cell receptor transgenic mice. I i t t . Zntitirtnol. 4, 541. Bachmann, M . F.. Rohrer, U . H., Steinhoff, U., Burh. K., Skuntz. S., Arnheiter, H., Hengartner, H.. and Zinkernagel, R. M. (1994). T helper cell unresponsiveness: Kapid induction in ;~ntigen-transgenicand reversion in non-transgenic mice. Eur. /. Zmttunol. 24, 2966-2973. Barthlott, T., Kuliler, H., Pirclier, H., and Eichmann, K. (1997). Differentiation of CD4(high) CD8(low) coreceptor-skewed thymocytes into mature CD8 single-positive cells independent of MHC: class I rpcognition. Eur. /. Iitirttunol. 27, 2024-2032.
252
13HIC:I'ITA STOCKIN<;ER
Berg, L. J., Pullen, A. M., Fazekas de St. Grotli, B., Mathis, D., Benoist, C . , and Davis, M. M. (1989).AntigedMHC specific T cells are preferentially exported from the thymus in the presence of their MHC ligand. Cell 58, 1035-1046. Bewm, M . J. (1977). 111 a radiation chimera host H-2 antigens determine the immune responsiveness of donor cytotoxic cells. Noturf: 269, 417-418. Beverley, B., Kaiig, S. M., Lenartlo, M. J , , and Schwartz, R . H. (1992). Reversal of i t i uitro T cell clonal anergy by IL-2 stimulation. Irit. Immtctml. 4, 661-671. Billingham. R . E., Brent, L., and Medawar, P. €3. (1953). Actively acquired tolerance of foreign cells. Nntnre 172, 603-606. Bix, M., and Raulet, D. (1992). Inefficient positive selection ofT cells directed by hematopoietic cells. Nature 359, 330-333. Bogen, B. (1996). Peripheral T cell tolerance as a tumor escape mechanism: Deletion of CD4' T cells specific for a monoclonal immunoglobulin idiotype secreted by a plasmacytoma. Eitr. /. I t n I n i d , 26, 2671-2679. S. (1993).Clonal deletion of specific thymocytes by an Bogen. B., Dembic, Z., and We in~rnunoglobu~ii~ idiotype. EMBO J. 12,357-363. Biiliine, J., and Pilstrtim, B. (1991).Transgenic mice with ectopic expression of alloantigenic M H C molecdes: Why are they so different and of how much help are they? Zmmunol. Rco. 122,21-32. Bonomo, A,, and Matzinger, P. (1993).Thymus epitlielium induces tissue-specific tolerance. J. Esp. Mcd. 177, 1153-1164. Bonyhadi, M. L.. Rabiri, L., Salinii, S., Brown, D. A,. Kosek, J., McCune, J. M., arid Kaneshima, H. (1993). HIV induces thymus depletion i n oino. Nuture 363, 728-732. Boussiotis, V. A,, Barber, D. L., Nakarai, T., Freeinan, G. J., Gribhen, J. G., Benistein, C;. M., D'Aiidrea, A. D., Ritz, J.. and Nadler, L. M. (1994a). Prevention of T cell anergy by signaling through tlie gamma g chain of the IL-2 receptor. Science 266, 1039-1042. Boussiotis, V. A.. Gribhen, J. G., Freeman, G. J., arid Nadler, L. M. (1994b). Blockade of tlie CD28 co-stimulatory pathway: A means to iriduce tolerance. Ci~rr.Opin. Zmntmol. 6, 797-807. Bretscher, P., and Colin, M. (1970). A theory of self-nonself discrimination: Paralysis and induction involve the recognition of one and hvo determinants oil an antigen, respectively. Science 169, 1042-1049. Brocker, T., Riedinger, M., arid Karjalaine, K. ( I 997).Targeted expression of major histocompatiliility complex (MHC) class I1 iiiolecules demonstrates that dendritic cells can induce negative but not positive selection of thymocytes i n o i w . J , Exp. Med. 185, 541-550. Biitltl, R. C . , Cerottini, J.-C., Horvath. C., Broil, C., Pedrazzini, T., Howe, R. C., and MacDonald, R. H. (1987). Distinction of virgin and memory T lymphocytes: Stable acquisitioii of the Pgp-1 glycoprotein concomitant with antigenic stimulation. J. Irntiuinol, 138, 3120-3129. Buer, J.. Lanoue, A., Franzke, A., Garcia, C., voii Boehmer, H., and Sarukhan, A. (1998). Interleukin 10 secretion and impaired effector function of major histocoiiipatibilitycornplex class 11-restricted T cells aiiergized in vivo. /. ESP. Med. 187, 177-183. Burkly, L. C., Degermann, S., Longley, J., Hagman, J., Brinster, R. L., Lo, D., and Flavell, R. A. (1993). Clonal deletion of V beta 5' T cells by transgenic I-E restricted to thymic medullary epithelium. /. Zrtwiuriul. 151, 3954-3960. Butcher, E. C., and Picker, L. J. (1996). Lymphocyte homing and homeostasis. Science 272, 60-66. Byers. V. S.,and Sercarz, E. E. (1968). The X-Y-Z scheme of immunocyte maturation. IV. TIiv exhaustion of meiiiory cells. 1.Esp. Med. 127, 307-325.
T L Y M PMOCITE TOLE H.AN(:E
253
Cdnan, B. J.. Szycliowski, S.. Chan, F. K., (:ado, D.. and Winoto, A. (1995). A role for the orphan steroid receptor Nur77 in apoptosis accoinpanying ~intigen-iliducetlneptive selection. Zttitriunity 3, 273-282. Carlow, D. A., van, 0. N., Teh. S. J., d Teh, 11. S. (1992). Deletion of antigen-specific iniinature thyniocytes by dendritic cells requires LFA-l/ICAM interactions. J . I t t w I u t d . 148, 1595-1603. Castro, J. E., Listinan, J. A., Jacobson. H. A,. Wang, Y.,Lopez, P. A., Jn, S.. Finn, P. W., and Perkins, D. L. (1996). Fas motlulation of apoptosis during negative selection of thymocytes. Imruritiity 5, 617-627. Chambers, C. A.. Kniinmel, M. F.. Boitel, R., Hurwitz, A., Sullivan, T. J., Fournier, S., Cassell, D., Biunner, M., and Allison, J. P. (1996).The role of CTLA-4 in the regulation and initiation of T-cell responses. I m t t i i i t i o l . Rec. 153, 27-46. Chanibers, C . A.. Sdivan, T. J.. and Allison, J. P. (1997). Lyinl)lioproliferation in CTLA4 deficient mice is mediated by costimul~ition-depc,ndeiitactivation of CD4' T cells. zt?lttL fitlify 7, 885-895. Chesnnt, R. W., Colon, S.. and Grey, H. M. (1982). Antigen presentation by nomial B cells. B cell tumors and inacrophnpes:Functional antl liioclieinical comparison.J. ZtrirnutwZ. 128, 1764-1768. Chiller, J. M., IIabicht, G. S., antl Weiglc, W. 0. (1971).Kinetic clifferences in unresponsiveness of thymus antl bone marrow cells. Scimcc 171, 813-815. Chiller. J. M . . and Weigle. W.0. (1973). Termination of tolerance to human gainma globulin in iiiice by antigen antl bacterial lipopolysocchaiidc (endotoxin).J . Exp. &fed. 137, 740-750. Cihak, J.. antl Lehmaiiii-(:rube, F. ( 1978). Imniunologicul tolerance to lyniphoytic clioriomeningitis virus in neonatally infkctetl \irns carrier inice: Evidence supporting a clonal inactivation n~echanism.I t t i t t ~ i ~ t i ~ i L ~34, i g ~ 265-2i5. y Cobbold, S. P., Adam, E., Marshall, S. E., Davies. J. D.. and Waldmann. 11. (1996). Mechanisnis of peripheral tolerance and siippressioii induced by nionoclonal antibodies to CD4 and CD8. I t t i t n u r i d Reo. 149, 5-33, Cooke. M. P., lleath, A. LV., Shokat, I<. M., Zcwg, Y., Finkelman, F. D.. Linsley, P. S., Howard, M., and Goodnow. C. C. (1994). Iininunoglobulin signal transduction guides the specificity of B cell-T cell interactions and is blocked in tolerant self-reactive B cells. 1. Exp. hfetl. 179, 425-438. Cosgrove, D., Chan. S. H., Lt'altzinger, C.. Benoist, C . , antl Mathis, D. (1992). The thymic coiiipartment responsible for positive selection ofCD4' T cells. I t i t . Z m t t i u t i o / . 4, 707-710. Crispr, I. N. (1994). Fatal intrractions: F;is-inducetl apoptosis of mature T cells. Inirnrmity 1,347-349. Critchfield, J . M., Racke, M. K., Zuniga-Pfliicker,J. C . ,Cannella, B., Raine, C. S., Governian, J.. and Lenardo. M . J. (1994).T cell deletion in high antigen dose therapy of autoimmune enceplialoni)/elitis. Science 263, 1139- 1113. Croft, M., Joseph, S. B., antl Miner, K,T. (1997). Partial activation of naive CD4 T cells and tolerance induction in response to peptide prcwnted by resting B cells. J . Itnnlrtnol. 159, 3257-3265. Cyster. J. G,, Hartley, S. B., and Goodnow. C. C. (1994).Competition for f d i c u h niches exclndes self-reactivr B cells from the recirculating B-cell repertoire. Nature 371, 389-395. Degerinann, S., Reilly, C., Scott, B., Ogata, L., von Boehmer, H., and Lo, D. (19'34).On the various manifestations of spontaneous autoimmune diabetes in rodent inodels. Eur. 1. r l I l t n t i t 1 d . 24, 31s5-:3160.
254
RRIGITTA STOCKINGER
de Waal Malefyt, R., Yssel, € I . , and de Vries, J. E. (1993). Direct effects of IL-10 on subsets of human CD4’ T cell clones and resting T cells: Specific inhibition of IL-2 production and proliferation. J . I t t i t i i u t t o l . 150, 4754-4765. Ding, L., Linsley, P. S., Huang, L. Y., Gerrnain, K. N., and Shevach, E. M. (1993). IL-10 inhibits macrophage costimulatory activity by selectivity inhibiting the up-regulation of B7 expression. J . Ztrirn~tnol. 151, 1224-1234. Diiffinger, R., Klein, T. C., Pepys, M. B., Casanova, J.-L., and Kyewski, B. A. (1997). The MHC class 11-restricted T cell response of C:57/B16 mice to human C-reactive protein: Honioloa to self and the selection ofT cell epitopes and T cell receptors. Mol. Inivittnol. 34, 115- 124. Dorf, M. E., and Benacerraf, 13. (1985). I-J as a restriction element in the suppressor T cell system. Ininiutiol. Reu. 83, 23-40. Douek, D. C., Corley, K. T. T., Zal, T., Mellor, A., Dyson, J., and Altmann, D. (1996). Negative selection by endogenous antigen and superantigen occurs at multiple thymic sites. Znt. l t n t t i r t t i d . 8, 1413-1420. Dresser, D. W. (1961).Effectiveness of lipid and lipidophilic substances as adjuvants. Nature 191, 1169-1171. Dyill, R.. and Nikolic Zugic, J . (1995).The majority of postselection CD4’ single-positive thymocytes requires the thymus to produce long-lived, functional T cells. J . Exp. Med. 181, 235-245. Ehl, S.. Hombach, J., Aichele, P., Rulicke, T., and Pircher, I1.-P. (1998).Viral and bacterial infections interfere with peripheral tolerance induction and activate CD8+ T cells to cause iininunopatlrolo~.J . Exp. Med. 187, 763-774. Eynon, E. E., and Parker, D. C. (1992). Sinall B cells as antigen presenting cells in the induction of tolerance to soluble protein antigen. J. E x p Illed. 175, 131-138. Ferber, I., Schiinrich, G., Schenkel, J., Mellor, A. L., HPmrnerling, G. J., and Arnold, B. (1994). Levels of peripheral T cell tolerance induced by different doses of tolerogen. Scieuce 263, 674-676. Fields, P. E., Chjewski, T. F., and Fitch, F. W. (1996). Blocked Ras activation in allergic CD4‘ T cells. Science 271, 1276-1278. Finkelman, F. D., Lees, A., and Morris, S. C. (1992).Antigen presentation by B lymphocytes to CD4’ T lymphocytesin vivo: Importance for B lymphocyte and T lymphocyte activation. Setti. ltn?niinol, 4, 247-255. Fiirster, I., Hirose, R., Arheit, J. M., Clausen, B. E., and Hanahan, D. (1995). Limited capacity for tolerization of CD4+ T cells specific for a pancreatic beta cell neo-antigen. b n n t r i i i i t y 2, 573-585. Forster, I., and Lieberman, I. (1996). Peripheral tolerance of CD4 T cells following local activation in adolescent mice. Eur. J . I t n r t w t t d 26, 3194-3202, Foy, T. M., Page, D. M., Waldschmitlt. T. J., Schoneveld, A., Laman, J. D., Masters, S. R., Tygrett, L., Ledbetter, J. A,, Aruffo, A,, Claassen, E., et 01. (1995). An essential role for gp39, the ligand for CD40, in thymic selection. J. Exp. Merl. 182, 1377-1388. Fritz, R. B., and Zhao, M. L. (1996).Thymic expression of myelin basic protein (MBP): Activation of MBP-specific T cells by thymic cells in the absence of exogenous MBP. J . Z t n t t i r i r t d . 157, 5249-5253. Fuchs, E. J., and Matzinger, P. (1992). B cells turn off virgin, but not memory T cells. Science 258, 1156-1 159. Gajewski, T. F., Lancki, 11. W., Stack, K., and Fitch, F. W. (1994). “Anergy” of Tho helper T lymphocytes indnces downregulation of Th1 characteristics and a transition to a Th2like phenotype. J. E x p . Merl. 179, 481-491.
T LYMPHOCYTE TOLERANCE
255
Gao, E.-K., arid Spent. J. (1990).StrongT cell tolerance in parent-F1 bone rnarrowclrimeras prepared with siipralethal irradiation. /. Esp bled. 171, 1101-1121. Gay, D., Saunders, T., Camper. S.. and Weigert, M. (1993). Receptor edlting: An approach by antoreactive B cells to escape tolerance. 1,E x p Med. 177, 999-1008, Geiger, T., Gooding, L. R., and Flavell, R. A. (1992). T-crll responsiveness to an oncogenic peripheral protein and spontaneous autoimniiinity i n transgenic mice. Proc. Natl. Acad. Sci. USA 89, 2985-2989. Ghendler, Y., I-Iussey,R. E., Witte, T., Mizoguchi, E., Clayton. L. K., Bhan, A. K., Koyasu, S., Chang, H. C., and Reinherz, E. L. (1997).Doulile-positive T cell receptor (high) tliymocytes are resistant to peptide/rnajor liistoconipatibility coinplrx ligand induced negative selection. Eur. 1. I i i t r i t t m o l . 27, 2279-2289. Gleeson, P. A., Toh. B.-H., and van Driel, I. R. (1996).Organ-specificautoimmunity induced by lyniphopenia. l i i i i i ~ t m o l Reu. . 149, 97- 125. Godfraintl, C., Holmes. K. V., and Coutelier, J. P. (1995).Thyinus involution induced by iiioiise hepatitis virus A59 in BAI,B/c mice. /. Virol. 69, 6541-6547. Goodnow. C;. C., Cyster, J. G., Hartley, S. B., Bell, S. E., Cooke, M. P., Iiealy, J. I., Akkaraju, S.,Rathmell, J. C.. Pogue. S. L., and Shokat, K. P. (1995). Self-tolerance checkpoints in B lyinpliocyte development. Atln. Z r n i i t i i r t o l . 59, 279-368. Goverman, J., Woods, A,, Larson, L., Weiner, L. P., Hood, L., and Zaller, D. M. (1993). Transgenic inice that express a inyelin basic protein-specific T cell receptor develop spontuneous autoiininnnity. Cell 72, 551-560. Granucci, F., Hescigno. M.. Marconi, G., Foti, M., and Ricciardi-Castagnoli,P. (19Y6).Igspecific T cell receptor-transgenic T cells are not deleted in the thyiilus arid are functional in vivo. /. Esp. Mecl. 183, 203-213. Groux. H., Rigler, M., de Vries, J. E., and Roncarolo, M. G. (1996). Interluehn-10 induces a long-term antigen-specificmiergic state in hunian CD4’ Tcel1s.j. Exp. Med. 184, 19-29. Groux, H., O’Garra. A,, Rigler, M., Roiileau, M.. Antonenko, S.,tle Vries, J. E., and Roncarolo, hl. G.(1997). A CD4’ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 389, 737-742. Gnibin, C. E., Kovats, S., deRoos, P., and Rudensky, A. Y. (1997).Deficient positive selection ofCD4 T cells in inice displaying altered repertoires of MHC class 11-boundself-peptides. I r ~ ~ n i n ~ t7, i f y197-208. Guerder, S., and Matzinger, P. ( 1992).A fail-safe inechanisni for maintaining self-tolerance. /. E.rp. Med. 176, 553-564. Guerder, S., Meyerlioff. J., and Flavell. R. (1994). The role of the T cell costiinulator B7-1 in autoiminunity and the induction and niaintenance of tolerance to peripheral antigen. Z i n i i w i i f ! / 1, 155-166. Himinerling, G. J.. Schonrich, C., Momburg, F., Anphan, N., Malissen, M., SchniittVerhulst, A,-M., and Arnold, €3. (1991). Non-deletional mechanisms of peripheral and central tolerance: Studies with transgenic mice with tissue-specific expression of a foreign MHC class I antigen. Z i i t r r t w t o ~ .Ren. 122, 47-68. Hashimoto, K., Solin, S. J., Lrvin, S. D., Tada. T., Perlmutter, R. M., and Nakayinia, T. (1996). Requirement for p561ck tyrosine kinase activation in T cell receptor-mediated thymic selection. /. Esp. hled. 184, Y331L941. Ho, W. Y., Cooke, M. P., Goodnow, C. C., and Davis, M. M. (1994). Resting and anergic B cells are defective in CD28-dependent costiinulatiori of naive CD4’ T cells. 1. Exp. Merl. 179, 1539-1549. Hoffinann, hl. W., Allison, J., and Miller. J. F. A. P. (1992).Tolerance induction by thymic inetliilary epithelium. Proc. Nutl. Acad. Sci. USA 89, 2526-2530.
256
BKIGI'ITA STOCKINCEK
Hogquist, K. A,, Jameson, S. C., Heath, W. R., IIoward, J. L., Bevan, M. J., and Carbone, F. R. (1994).T cell receptor antagonist peptides induce positive selection. Cell 76, 17-27. Homer, R. J., Mamalaki, C., Kioussis, D., and Flavell, R. A. (1993).T cell unresponsiveness correlates with quantitative TCR levels in a transgenic model. l n t . lmmrrnol. 5,1495-1500. Hugo, P., Kappler, J. W., Godfrey, D. I., and Marrack, P. C. (1994).Thymic epithelial cell lines that mediate positive selection can also induce thymocyte clonal deletion. I. Zmnw nol. 152, 1022-1031. Janeway, C. A,, Jr., Goodnow. C. C., and Medzhitov, R. (1996). Danger: Pathogen on the premises! Immunological tolerance. Czrm B i d . 6, 519-522. Jenkins, M. K., and Schwartz, R. H. (1987). Antigen presentation by chemically modified splenocytes induces antigen-specific T cell unresponsiveness in vitro and in vivo. /. Exp. Med. 165, 302-319. Jolicoeur, C., Hanahan, D., and Smith, K. M. (1994). T cell tolerance toward a transgenic /3 cell antigen and transcription of endogenous pancreatic genes in the thymus. Proc. Natl. Acad. Sci. USA 91, 6707-6711. Jones, L. A., Chin, L. T., Longo, D. L., and Kruisbeek, A. M. (1990). Peripheral clonal elimination of functional T cells. Science 250, 1726-1729. Jones, L. A,, Izon, D. J., Nieland, J. D., Linsley, P. S., and Kruisbeek, A. M. (1993).CD28B7 interactions are not required for intrathymic clonal deletion. lnt. Immunol. 5,503-512. Kane, L. P., and Hedrick, S. M. (1996). A role for calcium influx in setting the threshold for CD4+CD8+thymocyte negative selection. /. Immunol. 156, 4594-4601. Kang, S. M., Beverly, B., Tran, A. C., Brorson. K., Schwartz, R. H., and Lenardo, M. J. (1992). Transactivation by AP-1 is a molecular target of T cell clonal anergy. Science 257, 1134-1138. Kappler, J. W., Roehm, N., and Marrack, P. C. (1987).T cell tolerance by clonal elimination in the thymus. Cell 49, 273-280. Katz, J. D., Benoist, C., and Mathis, D. (1995). T helper cell subsets in insulin dependent diabetes. Science 268, 1185-1 189. Katz, J. D., Wang, B., Haskins, K., Benoist, C., and Mathis, D. (1993).Following a diabetogenic T cell from genesis through pathogenesis. Cell 74, 1098-1100. Kawabe, Y., and Ochi, A. (1991). Programmed cell death and extrathymic reduction of Vp8+ CD4+ T cells in mice tolerant to staphylococcus aureus enterotoxin B. Nature 349, 245-248. Kaye, J., Hsu, M.-L., Sauron, M.-E., Jameson, S. C., Gascoigne, N. R. J., and Hedrick, S. M. (1989). Selective development of CD4+ T cells in transgenic mice expressing a class I1 MHC-restricted antigen receptor. Nature 341, 746-749. Kearney, 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. Zmmunity 1,327-339. Kimpton, W. G., Washington, E. A,, and Cahill, R. N. (1995).Virgin alpha beta and gamma delta T cells recirculate extensively through peripheral tissues and skin during normal development of the fetal immune system. lnt. Immunol. 7, 1567-1577. Kirberg, J., Baron, A., Jakob, S., Rolink, A,, Karjalainen, K., and von Boehmer, H. (1994). Thymic selection of CD8+ single positive cells with a class I1 major histocompatibility complex-restricted receptor. /. Exp. Med. 180, 25-34. Kirberg, J., Bruno, L., and von Boehmer, H. (1993). CD4+8- help prevents rapid deletion of CD8' cells after a transient response to antigen. Eur. I. Znmunol. 23, 1963-1967. Kishimoto, H., Cai, Z., Brunmark, A,, Jackson, M. R., Peterson, P. A,, and Sprent, J. (1996). Differing roles for B7 and intercellular adhesion molecule-1 in negative selection of thymocytes. I. Exp. Med. 184,531-537.
T LYXIPIIOCYTE TOLERANCE
257
Kishimoto, II., arid Sprent, J. (1997). Negative selectioii i n the thymus includes semiinatlire T cells. /. Exp. Med 185, 263-271. Kisielow, P., Bliithmann, H., Staerz, LJ., Steininetz. M., and von Boehmer, H, (1988). Tolerance in T cell receptor transgenic mice involves deletion of nonrnature CD4'CD8' thyinocytes. Nnturc 333, 742-i46. Klein, L., Klein, T., Riither, U., and Kyewski. B. (1988).CD4 T cell tolerance to huinan C-reactive protein. a n inducible serum protein, is mediated by thylriic epithelium. /, Exp. Med. 188, 5-16. Kronenberg, M.. Steininetz, M., and Kobori, J. (1983). R N A transcripts for I-J polypeptides are apparently not encotled between the I-A imtl I-E subregions of the muiine major histoconipttibhty complex. Proc. Not/. Acud. Sci. USA 80, 5704-5708. Kiurninel, M. F., and Allison. J. P. (1996).CTLA-4 engagement inhibits 1L-2 accuniulation and cell cycle progression upon activation of resting T cells./. Exp. Med. 183, 2533-2540. Ksantler, B. R.. arid Streilein, J. W. (1994). Regulation of the immune response within privileged sites. Chem. Znimrod 58, 117-145. ~ ~ r t - j o n iE. s , A., Liano, D., H ~ ~ C IK~. A,, S , Benacerraf, B., Syin, M.-s., and ~ b b a sA. , K. (19883. The role of antigen-presentiiig B cells i n T cell priming in vivo. /. Zttitnnnol. 140, 3773-3778, Kurts. C., Carbone, F. R., Bamden, M.. Blanm, E.. Allison, J., Heath, W.R., and Miller, J. F. A. P. (l997a). CD4' T cell help iinpuirs CD8' T cell deletion induced by crosspresentation of self-antigens antl favors autoimmunity. /. Exp. Med. 186, 2057-2062. Kurts, C . , Kosnka. II., Carbone. F. H., Karanialis, F., Miller, J. F. A. P., and Heath, W.R. (1997b). Class I-restricted cross-presentation of exogvnous self antigens leads to deletion of autoreactive CD8' T cells. /. E x p B l c d . 186, 239-245. Kyburz, D., Aichele, P., Speiser, D. E., Ilengartner, 13.. Zinkernagel, R. M., and Pircher, H. (1993). T cell ininiunity after a viral infection versus T cell tolerance induced by soluble viral prptides. Eiir. /. Z t w ~ t ~ n o23, / . 1956-1962. Lafaille, J. J., Nagashiina, K., Katsuki, M.. and Tonegawa, S. (1994). High incidence of spontaneous autoimmune encephalomyelitis in ininiunodeficient anti-myelin basic protein T cell receptor transgenic mice. Cell 78, 399-408. Laffeity, K. J., Prowse, 8.J., and Siineono~ic,C . J. (1983). IniniunobioIogy of tissue transplantation: A retiam to the passenger leukocyte concept. Annu Reu. Z t t u n ~ ~ n1,d 143-173. Lanoue, A,, Bona, C., von Boehmer, H.. and Sanikhan, A. (1997). Conditions that induce tolerance in mature CD4' T cells. /. E s p . Mcd. 185, 405-414. Lanzaveccliia, 4 . (1990).Receptor-mediated antigen uptake and its effect on antigen presentation to class 11-restricted T l y m p h o c p s . Aiatirr. Rcc. Znunzrnol. 8, 773-793. Lassiln, O., Vainio. O., antl Matzinger, P. (1988). Can B cells turn on virgin T cells? Nature 344, 253-255. Laufer, T. M., IkKoning, J., Markowitz, J. S., Lo, D., and Chicher, L. (1996). Unopposed positive selection and autoreactivity in mice expressing class I1 MHC only o n thymic cortex. h7attrre 383, 81-85. Le Douarin, N., Corbel. C., Bandeira, A , , Thomas-\'ash, \'., Motligliani, Y., Coutinho, A., and Salaiin, J. (1Y96). Evidence for a tliyiiius-dep~,iitlentform of tolerance that is not . 35-53. based on elimination or anergy of reactive T cells. Znitnnnol. R ~ c 149, Lee. N. A,, Loh, D. Y.. and Lacy, E. (1992). CD8 surface levels alter the fate ofdphdbeta T cell receptor-expressing thyinocytes in transgenic mice. 1.Exp. Mcd. 175, 1013-1025. Lenardo, M. J., Bohme, S., C h i . L., Coinbadiere, B., Fisher, G., Freedman, M., McFarland, H., Pelfrey, C., and Zheng, L. (1995). Autocrine feedback death and the regulation of mature T lymplioc~yteantigen responses. Ztrtenr. Reo. Z t t u t m r d . 13, 115- 134.
258
BRIGITTA STOCKINGER
Lerner, A,, Clayton, L. K., Mizoguchi, E., Ghendler, Y., Wall Ewijk, W., Koyasu, S . , Bhan, A. K., and Reinherz, E. L. (1996). Cross-linking of T-cell receptors on double-positive thymocytes induces a cytokine-mediated stromal activation process linked to cell death. E M B O 1, 15, 5876-5887. Li, W., Whdey, C. D., Mondino, A,, and Mueller, D. L. (1996). Blocked signal transduction to the ERK and J N K protein kinases in anergic CD4t T cells. Science 271, 1272-1276. Lihlau, R. S . , Singer, S. M., and McDevitt, H. 0. (1995). Thl and Th2 CD4' T cells in the pathogenesis of organ-specific autoimmune diseases. In~nrunol.Today 16, 34-38. Liu, G. Y., Fairchild, P. J., Smith, R. M., Prowle, J. R., Kioussis, D., and Wraith, D. C. (1995). Low avidity recognition of self-antigen by T cells permits escape from central tolerance. Inmiunity 3, 407-415. Lo, D., Freedman, J., Hesse, S., Brinster, R. L., and Sherman, L. (1991). Peripheral tolerance in transgenic mice: Tolerance to cl I1 MHC andnon-MHC transgene antigens. linninnol. Reu. 122, 87-102. Lo, D., Freedman, J., Hesse, S., Palmiter, R. D., Brinster, R. L., and Sherman, L. A. (1992). Peripheral tolerance to an islet cell-specific hernagglutinin transgene affects both CD4+ and CDXt T cells. Eur. /. IininiinoL 22, 1013-1022. Lo, D., and Sprent, J. (1986). Identity of cells that imprint H-2 restricted T-cell specificity in the thymus. Nature 319, 672-675. Luhder, F., Hoglund, P., Allison, J. P., Benoist, C., and Mathis, D. (1998). Cytotoxic T lymphocyte-associatedantigen 4 (CTLA-4) regulates the unfolding of autoimmune diabetes. 1.Exp. Med. 187, 427-432. Macatonia, S. E., Doherty, T. M., Knight, S. C., and O'Garra, a. (1993). Differential effect of IL-10 on dendritic cell-induced T cell proliferation and IFN-gamma production. 1. lnnnrcnol. 150, 3755-3765. MacDonald, H. R., Raschieri, S., and Lees, R. K. (1991).Clonal expansion precedes anergy and death of VP8' peripheral T cells responding to staphylococcal enterotoxin B in vivo. Eur. I. Iininunol. 21, 1963-1966. MacDonald, H. R., Pedrazzini, T., Schneider, R., Louis, J. A,, Zinkernagel, R. M., and Henpartner, H. (1988). Intrathymic elimination of MISa-reactive (V/36+) cells during neonatal tolerance induction to MIS"-encoded antigens. J. Exp. Med. 167, 2005-2010. Mackay, C. R. (1993). Homing of naive, memory and effector lymphocytes. C u r . Opin. Itnmunol. 5, 423-427. Mainalaki, C., Murdjeva, M., Tolaini, M., Norton, T., Chandler, P., Townsend, A,, Simpson, E., and Kioussis, D. (1995). Tolerance in TCWcognate antigen double-transgenic mice mediated by incomplete thymic deletion and peripheral receptor downregulation. Deu. Iininunol. 4, 299-315. Mamalaki, C., Norton, T., Tanaka, Y., Townsend, A. R., Chandler, P . , Simpson, E., and Kioussis, D. (1992).Thymic depletion and peripheral activation of class I major histocompatibility complex-restricted T cells by soluble peptide in T-cell receptor transgenic mice. Proc. Natl. Acad. Sci. USA 89, 11342-11346. Mamalaki, C., Tanaka, Y., Corbella, P., Chandler, P., Simpson, E., and Kioussis, D. (1993). T cell deletion follows chronic antigen specific T cell activation in vivo. Int. Itnrnunol. 5, 1285-1292. Markiewicz, M. A., Girao, C., Opferman, J. T., Sun, J., Hu, Q., Agulnik, A. A,, Bishop, C. E., Thompson, C. B., and Ashton-Rickardt, P. G. (1998). Long-term T cell memory requires the surface expression of self-peptidehajor histocompatibility complex molecules. Proc. Natl. Acad. Sci. USA 95, 3065-3070. Markowitz, J. S., Auchincloss, H. J., Grusby, M. J., and Glimcher, L. H. (1993). Class IIpositive hernatopoietic cells cannot mediate positive selection of CD4' T lymphocytes in class II-deficient mice. Proc. Natl. Acad. Sci. USA 90, 2779-2783.
T 1,YMPHOCI'TE TOLERANCE
259
Marrack, P., Ignatowiu, L., Kappler, J. W., Boyinel, J., and Freed, J. ti. (1993).Comparison of peptides bound to spleen and thyinus class 11. J . Exp. hlctl. 178, 2173-2183. Marrack, P., Lo, D.. Brinster. R., Palmiter, R., Hnrkly, L., Flavell, R . H., and Kappler, J. (1988). The effect of thymus environment on T cell development and tolerance. Cell 53,627-634. Martin, S., and Bevan, M. J. (1997). Antigen-specificand nonspecific deletion of immature cortical thyinocytes caused by antigen iiijection. Eur. /. Zmntrmol. 27, 2726-2736. Mathisen. P. M., Pease, S., Garvey, J., Hood, L., and Readhead, C. (1993). Identification of an embryonic isoform of inyelin basic protein that is expressed widely in the mouse enihryo. Prtic. Natl. Acad. Sci. USA 90, 10125- 10129. Matzinger, P. (1994). Tolerance, danger, and the extended family. Annu. Reo. Zntrnunol. 12, 991-1045, McSorley, S. J., S. Soldera, L. Malherbe. C. Carnautl, Locksley, R. M., R. A. Flavell. and N. Glaichenhaus (1997). Immunological tolerance to a pancreatic antigen as a result of local expression of TNFa by islet fi cells. Z~nmirtity7, 401-409. Mebius, R. E.. Streeter, P. R., Michie, S., Butcher, E. C., and Weissman, I. L. (1996). A developmental switch in lymphocyte homing receptor and endothelial vascular addressin expression regulates lymphocyte homing and permits CD4+ CD3- cells to colonize lymph nodes. Proc. Natl. Acad. Sci. L'SA 93, 11019-11024. Miller, J. F., Morahan, G., Allison, J., Bhathal, P. S.. and Cox, K. 0. (1989).T-cell tolerance in transgenic mice expressing major histocompatihility class I molecules in defined tissues. Ztnttumol. Kec. 107, 109- 123. Mitchison, N. A. (1964).Induction of immunologica~paralysis in two zones of dosage. Proc. Roy. Soc. 161, 275-290. Miyazaka, T., Simiki, G., and Yaniamura, K. (1993). The role of niacrophages in antigen presentation and T cell tolerance. Znt. Zmnuno/. 5, 1023-1033. Modigliani. Y., Bandeira, A,, and Coutinho, A. (1996).A model for developmentally acquired thymiis-dependent tolerance to central and peripheral antigens. Immunol. Rev. 149, 155-220. Mondino, A , . Khonits, A,, and Jenkins, M. K. (1996). The anatomy of T-cell activation and tolerance. Proc. Natl. Acud. Sci. USA 93, 224552252, Mondino, A,, Whaley, C . D., DeSilva, I). R., Li, W., Jenkins, M . K., and Mueller, D. L. (1996). Defective transcription of the 1L-2 gene is associated with impaired expression of c-Fos, FosB, and JiinB in anergic T helper 1 cells. J , Z t t t t m r ~ d157, 2048-2057. Morahan, G., Hoffinann, M. W., and Miller, J. F. (1991).A nondeletional mechanism of peripheral tolerance in T cell receptor transgenic mice. Proc. Natl. Acnd. Sci. USA 88, 11421-11425. Morgan, D. J., Liblau, R., Scott, B.. Fleck, S., McDevitt, H. O., Sarvetnick, N., Lo, D., and Sherman, L. A. (1996).CD8+T cell-rnetliated spontaneous diabetes in neonata! mice. /. Zmmi~rtol.157,978-983. Moskophitlis, D., Lain?, E., and Zinkernagel, R. M. (1993). Peripheral clonal deletion of antiviral memory CD8' T cells. Eur. J . Z)rmr/nd 23, 3306-3311. Mueller, D. L., Jenkins, M. K.. and Shcwartz, R. H. (1989). Clonal expansion versus functional clonal inactivation: A costimulatory signaling pathway determines the outcome of T cell antigen receptor occnpancy. Annu. Reo. Znimuno!. 7, 445-480. Mni-phy, K., Heimberger, A. B., and Loh. D. Y. (1990).Induction by antigen of intrathyinic apoptosis of CD4+CD8+TCR'"thynioc~es in vivo. Scimce 250, 1720- 1722. Murphy, K. M . , Weaver, C. T., Elish, M . , Allen, P. M., and Loti, D. Y. (1989). Peripheral tolerance to dlogeneic 11 histocompatibility antigens expressed in transgenic mice: 1
260
BRICITTA STOCKINCER
Evidence against a clonal deletion mechanisni. Proc. Nod. Acad. Sci. USA 86, 1003410038.
Negishi, I., Motoyama, N., Nakayama, K., Nakayama, K., Senju, S., Hatakeyania, S.. Zhang, Q,, Chan, A. C., and Loh, D. Y. (1995). Essential role for ZAP-70 in both positive and negative selection of thymocytes. Nature 376, 435-438. Nikolic Zugic, J., and Bevan, M. J. (1990). Functional and phenotypic delineation of two subsets of CD4 single positive cells in the thymus. Int. Znimunol. 2, 135-141. Oehen, S., Feng, L., Xia, Y., Siirh, C. D., and Hetlrick, S. M. (1996).Antigen compartnientation and T helper cell tolerance induction. J . E q Med. 183, 2617-2626. O’Hara, R. M. (1995). Antigen-specific snppressor factor: Missing pieces in the pnzzle. Zmmunol. Res. 14, 252-262. Ohashi, P. S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C. T., Odermatt, B., Malissen, B., Zinkernagel, R. M., and Hengartner, H. (1991).Ablation of “tolerance” and induction of diabetes by virus infection in viral antigen transgenic mice. Cell 65, 305-317. Ota, K., Matsui, M., Milford, E. L., Mackin, G. A,, Weiner, H. L., and Hafler, D. A. (1990).T cell recognition of an immunodominant myelin basic protein epitope in multiple sclerosis. Nature 346, 183-187. Oukka, M., Colucci-Guyon, E., Lan Tran, P., Cohen-Tannoudji, M., Babinet, C., Lotteau, V., and Kosmatopoulos, K. (1996). CD4 T cell tolerance to nuclear proteins induced by medullary thymic epithelium. Immunity 4, 545-553. Owen, R. D. (1945).Immunogenetic consequences of vascular anastomoses between bovine twins. Science 102, 400-401. Pape, K. A., Kearney, E. R., Khoruts, A., Mondino, A., Merica, R., Chen, Z. M., Ingulli, E., White, J.. Johnson, J. G., and Jenkins, M. K. (1997). Use of adoptive transfer of T-cell-antigen-receptor-transgenic Tcell for the study of T-cell activation in vivo. Zmnrunol. Reo. 156,67-78. Parijs, L. V., Ibraghimov, A,, and Abbas, A. K. (1996). The role of costimulation and Fas in T cell apoptosis and peripheral tolerance. Immunity 4, 321-328. Pelanda, R., Schwers, S., Sonoda, E., Torres, R. M., Nemazee, D., and Rajewsky, K. (1997). Receptor editing in a transgenic mouse model: Site, efficiency, and role in B cell tolerance and antibody diversification. Zmniunity 7, 765-775. Perez, V. L., Van Parijs, L., Biuckians, A,, Zheng, X. X., Strom, T. B., and Abbas, A. K. (1997). Induction of peripheral T cell tolerance in vivo requires CTLA-4 engagement. Immunity 6, 411-417. Pette, M., Fujita, K., Wilkinson, D., Altmann, D. M., Trowsdale, J., Giegerich, G., Hinkkanen, A,, Epplen, J. T., Kappos, L., and Wekerle, H. (1990). Myelin autoreactivity in multiple sclerosis: Recognition of myelin basic protein in the context of HLA-DR2 products by T lymphocytes of multiple sclerosis patients and healthy donors. Proc. Natl. Acad. Sci. USA 87, 7968-7972. Pircher, H., Brduscha, K., Kasai, M., Mizuochi, T., Zinkernagel, R. M., Hengartner, H., Kyewski, B., and Mliller, K.-P. (1993). Tolerance induction by clonal deletion of CD4+8+ thyrnocytes in vitro does not require dedicated antigen-presenting cells. Eur. J . Imnizrnol. 23, 669-674. Pircher, H., Burki, K., Lang, R., Hengartner, H., and Zinkemagel, R. M. (1989). Tolerance induction in double specific T-cell receptor transgenic mice varies with antigen. Nature 342, 559-561. Pircher, H., Roher, U. H., Moskiphidis, D., Zinkernagel, R. M., and Hengartner, H. (1991). Lower receptor avidity required for thymic clonal deletion than for effector T-cell function. Nature 351, 482-485.
T LYMPklOCYTE TOLERANCE
26 1
Powrie, F., Leach, M. W.,Mauze, S.. Menon, S., Barcomb Caddle, L., and Coffman. R. L. (1994). Inhibition of Thl responses prevents inflammatory bowel disease in scid mice reconstitated with CD45RB'" CD4' T cells. Znwrinity 1, 553-562. Powrie, F., and Mason. D. W. (1990). OX-22'"X'' CD4' T cells induce wasting &ease with multiple organ pathology: Prevention by the OX-22'"" subset. /. Esp. M e d . 172, 1701- 1708. Pribyl, T. M., Campagnoni, C., Kampf, I<.. Handley, V. W., and Campgnoni, A. T. (1996). The major myelin protein genes are cyressed in the hiiman thyinus. /. Neurosci. Re.s, 45, 812-819. Punt, J. A.. Oshorne, B. A,, Takahama, Y., Sharrow. S . O., and Singer, A. (1994). Negative selection of CD4'CD8+ thymocjtes by T cell receptor-induced apoptosis requires a costiinulatory signal that can be pro\itled by CD28. 1,ESP. Merl. 179, 709-713. Qin, Y.. Sun. D., Goto, M., Meyermann, R., nnd Wekcrle, H. (1989). Resistance to experimental autoiinininie enceplialoinyelitis induced by neonatal tolerization to myelin basic protein: Clonal elimination vs. regulation of autoaggressive lymphocytes. Eur. /. Zmmunol. 19, 373-380. Ramsdell. F.. Jenkins, M., Dinh, Q., and Fowlkes. 8. J. (1991). The majority of CD4'8thyniocytes are finictionally immature. J. h 1 ~ 1 1 ~ 1 i o147, 1 . 1779- 1785. Ramsdell, F., Lantz, T.. antl Fowlkes, B. J. (1989).A nondeletional mechanism of thymic self tOlerdIlcr. SCiellcP 246, 1038- 1041. Ready. A. R., Jenkinson, E. J.. Kingston, R . , and Owen, J. J. T. (1984).Successful transplantation across major histocoinpatihility harrier of deoxyguanosine-treateclembryonic thymus expressing class I1 antigens. Nature 310, 231-23:3. Renno, T., Hahne, M., and MacDonaltl, H. R. (1995). Proliferation is a prerequisite for bacterial siiperantigen-induced T cell apoptosis in vivo. J. Exp. bled. 181, 2283-2287, Roberts, J. L., Sharrow, S. O., and Singtar. A. (1990). Clonal deletion an clonal anergy in the thymus induced hy cellular elemcwts with different radiation sensiticities. /. Exp. Merl. 171, 935-940. Robey. E. A,, Ramsdell, F., Kioussis, D., Sha, W.. I,oh, D., Axel, R. A,, and Fowlkes, B. J. (1991).The level of CD8 expression can deterinine the outcome of thymic selection. Cell 69, 108% 1096. Rocha, €3.. Grandien. A,, and Freitas, A. A. (1995). Anergy a i d exhaustion are independent mechanisms of peiiplieral T cell tolerance. 1. Exp. Alcd. 181, 993-1003. Rocha, B., Tanchot, C., and von Boehmer, H. (1993).Clonal anergy blocks in civo growth of mature T cells antl can be reversed in the ahsence of antigen. /. Exp. M e d . 177, 1517-1521. Rocha, B., and voii Boehmer. H. (1991). Peripheral selection of the T cell repertoire. Science 251, 1225-1228. Rocken, M., and Shevach, E. M. (1996). Iminune deviation: The third dimension ofnondeletional T cell tolerance. Z t t t m m d . Reo. 149, 175- 194. Rkken, M., Urban, J. F., and Shevach, E. M. (1992). Infection breaks T-cell tolerance. Nritirrr! 359, 79-82. Roinball. C. G., and Weigle, W.0. (1993). I n \ivo induction of tolerance in inurine CD4' subsets. /. Exp. Mrcl. 178, 1637-1644. Ron, Y., and Sprent. J. (1987).T cell priming in civo: A major role for B cells in presenting antigen to T cells in Ijnph nodes. 1.I n ~ m n n o ~138, . 2848-2856. Ronchese, F., and Hausmann. B. (1993). B lymphocytes in vivo fail to prime naive T cells in the induction of tolerance to soluble protein antigens. /. Exp. M e d . 177, 679-690. Salaiin, J.. Bandeira, A,, Khazaal, I., Burlen-Defranoux, O., Thomas-Vaslin, V..Coltey, M.,
262
HRIGI'ITA STOCKINGER
LeDouarin, N. M., and Coutinho, A. (1992). Transplantation tolerance is unrelated to superantigen dependent deletion and anergy. Proc. Natl. Acad. Sci. USA 89,10420-10424. Salvetti, M., Jung, S., Chang, S.-F., Will, H., Schalke, B. C. G., and Wekerle, H. (1991). Acetylcholine receptor-specific T lymphocyte clones in the normal human immune repertoire: Target epitopes, HLA restriction, and membrane phenotypes. Ann. Neurol. 29, 508-516. Saoudi, A,, Kuhn, J., Huygen, K., de-Kosak, Y,, Velu, T., Goldman, M., Dniet, P., and Bellon, B. (1993). Th2 activated cells prevent experimental autoimmune uveoretinitis, a Thl dependent autoimmune disease. Eur. J, lmmunol. 23, 3096-3103. Saoudi, A,, Seddon, B., Heath, V., Fowell, D., and Mason, D. (1996). The physiological role of regulatory T cells in the prevention of autoimmunity: The function of the thymus in the generation of the regulatory T cell subset. Zmmunol. Reo. 149, 195-216. Schmidt, D., Verdaguer, J., Averill, N., and Santamaria, P. (1997). A mechanism for the major histocompatibility complex-linked resistance to autoimmunity. /. Exp. Med. 186, 1059- 1075. Schneider, R., Pircher, H., Speiser, D. E., Zinkernagel, R., and Hengartner, H. (1992). Kinetics of clonal deletion varies with tolerizing antigen. Thymu,s 20, 5-15. Schonrich, G., Kalinke, U., Momburg, F., Malissen, M., HPmmerling, G. J., and Arnold, B. (1991).Down-regulation of T cell receptors on self-reactiveT cells as a novel mechanism for extrathymic tolerance induction. Cell 65, 293-304. Schonrich, G., Momburg, F., Malissen, M.. Schmitt-Verhulst,A. M., Malissen, B., HPmmerling, G. J., and Arnold, B. (1992). Distinct mechanisms of extrathymic T cell tolerance due to differential expression of self antigen. Znt. Zmmunol. 4, 581-590. Schwartz, R. H. (1990). A cell culture model for T lymphocyte clonal anergy. Science 248, 1349-1355. Schwartz, R. H. (1997). T cell clonal anergy. Curr. Opin. lmmunol. 9, 351-357. Scott, B., Liblau, R., Degermann, S., Marconi, L. A,, Ogata, L., Caton, A. J., McDevitt, H. O., and Lo, D. (1994). A role for non-MHC genetic polymorphism in susceptibility to spontaneous autoimmunity. Immunity 1, 73-82. Scott, D. M., Ehrmann, I. E., Ellis, P. S., Bishop, C. E., Agulnik, A. I., Simpson, E., and Mitchell, M. J. (1995). Identification of a mouse male-specific transplantation antigen, H-Y. Nature 376, 695-698. Seder, R. A., Paul, W. E., Davis, M. M., and Fazekas de St. Groth, B. (1992).The presence of interleukin 4 during in vitro priming determines the lymphokine-producing potential of CD4+ T cells from T cell receptor transgenic mice. 1. Exp. Med. 176, 1091-1098, Sha, W. C., Nelson, C. A,, Newberry, R. D., Kranz, D. M., Russell, J. H., and Loh, D. Y. (1988). Selective expression of an antigen receptor on CD8-hearing T lymphocytes in transgenic mice. Nature 335, 271-274. Shahinian, A., Pfeffer, K., Lee, K. P., Kundig, T. M., Kishihara, K., Wakeham, A,, Kawai, K.. Ohashi, P. S., Thompson, C. B., and Mak, T. W. (1993).Differential Tcell costimulatory requirements in CD28-deficient mice. Science 261, 609-612. Sidman, C. L., Marshall, J. D., and Von Boehmer, H. (1992). Transgenic T cell receptor interactions in the lymphoproliferative and autoimmune syndromes of Ipr and gld mutant mice. Eur. /. Zmmunol. 22, 499-504. Simpson, E., Dyson, P. J., Knight, A. M., Robinson, P. J., Elliott, J. I., and Altmann, D. M. (1993).T cell receptor repertoire selection by mouse mammary tumor viruses and MHC molecules. Zmmunol. Reu. 131, 93-115. Smith, K. M., Olson, D. C., Hirose, R., and Hanahan, D. (1997). Pancreatic gene expression in rare cells of thymic medulla: evidence for functional contribution to T cell tolerance. lilt. lmmunol. 9, 1355-1364.
Sonnner. N., I-larconrt, G. C., Willcox. N., Beeson, D., and h’ewsom-Dii\is, J. (1991 ). Acetylcholine receptor-reactive T Iyinpliocytes froin healthy subjects and myasthenia gravis patients. Nrwrohgy 41, 1270- 1276. Spoonas, A.-M., Toinlinson. P. D.. Antoniou, J., Aiiphan, N., Langlet. C., Malissen, B. Schmitt-Verhulst. A.-M., and Mellor. A . L. (1994). Indnction of tolerance to self MHC class I niolecnles expressed initler the control of inilk protein or P-globin gene promoters. Znf. Zttinwtol. 6, 277-287. Sprent. J., voii Boehmcr, H.. and Nabholz. M. (1975).Association of iminiinity and tolerance to host H-2 determinants i n irradiated F1 Iivhrid mice reconstituted with bone marrow froin one parental strain. /. Exp. M e d 142, 321-331. Sterzl, J.. and Silverstein, A. M. (1967).Developmental aspects of immunity. Ado. Z m t t i u t z d . 6, 337-459. Stockinger, B., Zal, T.. Zal. A., and Cra),, D. (1996). H cells solicit their own help from T cells. J . Exp. M c d 183, 891-899. Stout. R. D., nnd Suttles, J. (1992). T cells hearing the CD44hi “inenio~-y”phenotype display characteristics of activated cells in C1 stage of cell cycle. Cell. l ? t i t t i i m o / . 141, 433-433. Streilein, J. W., Ksander, B. R., and Taylor, A . Iinmune deviation in relation to ocular iinmiinr privilege. /. I i i i i n u n o l . 158, Surli, C. D., and Sprent, J. (1994). T-cell apoptosis detected in sitn during positive and negative selection in the thymus. Nature 372, 100- 103. Swan. K. A., Alberola, I. J., Gross, J. A,, Appleby. M. W., Forbush, K. .4.. Thomas, J. F., and Perlmntter. R. M. (1995). Involveinent of p21ras distiiiguishes positive and negative selection in thyinocytes. EMBO /. 14, 276-385. Sykulev, Y.. Brunniark, A., Tsomitles, T. J., Kageyaiiia, S.. Jackson. M.. Peterson, P. A,. and Eisen. H. N. ( 1994). High-affinity reactions hetween antigen-specific T-cell receptors and peptides associated with allogeneic and sy-ngeneic major histocompatibility complex class I proteins. Proc. Natl. Acmd. Sci. USA 91, 11487-11491. S p w ,H. K.. Liblau, R. S.. and McDe\itt. H. 0. (1996).The roles of Fas/APO-l (CD95) and TNF in antigen-intluced programmed cell death in T cell receptor transgenic inice. Ittrniitnity 5, 17-:30. Tafuri, A,, Alferink, J., Moller. P., IIainnierling, C. J., and Arnold, B. (1995). T cell awareness of paternal alloantigens during pregnancy. Scictice 270, 630-633. Tan, R., Teh, S. J., Ledbetter, J. A,, Linsly. P. S.. and Teh, H. S. (1992). B7 costimnlates proliferation of CD4-8’ T 1yniphoc)ites but is not required for deletion of immature CD4’8’ thyinocytes. J . b t i r t i r t n o t . 149, 3217-3224. Tanaka, Y.,Mamalaki, C., Stockinger, B.. and Kioussis, 1). (1993). In vitro negative selection of ah T cell receptor transgenic thyinocytes by conditionally immortalized thymic cortical epithelial cell lines and dendritic cells. Eitr. J . Zt niim t io l. 23, 2614-2621. Tanchot. C., and Rocha, B. (1997). Peripheral selection of T cell repertoires: The role of continuous thymus outpiit. /. E x p . hlrd. 186, 1099-1 106. Tarazona, R., Williams, O., Moskophidis. D., Snyth, L. A,, Tanaka, Y., Murdjeva. M., Wack, A,, Mamalaki. C., and Kioussis, D. (1998). Susceptibility and resistance to antigen induced apoptosis in the thymus of transgenic mice. /. Iiimttnol. 160, 5397-5403. Teh, H.. Kishi, H., Scott, B.. and von Borliiner, H. (1989). Deletion of antospecific T cells in T cell receptor (TCR) transgenic inice spares cells with normal TCR levels and low levels of CD8 molecules. /. Exp. Med 169, 795-806. Thoinp~oii.C. B.. and Allison, J. P. (1997). The emerging role of CTLA-4 a s an iininune attenuator. Z t i i t ~ i ~ t t i i t !7, y 445-450.
264
RRIG17TA STO(:EINGEH
Tiegs, S. L., Russell, D. M., and Nernazee. D. (1993). Receptor editing in self-reactive hone marrow B cells. J. Exp. Mcd 177, 1009-1020. Tourne, S., Miyazaki, T., Oxeiiius, A,, Klein, L., Fehr. T., Kyewski, B., Benoist, C., and Mathis, D. (1997). Selection of a broad repertoire of CD4+ T cells in H-2Ma‘”’Imice. Itmrtnity 7, 187-195. Turner, M., Mee, J., Walters, A. E., Quinn, M. E., Mellor, A. L., Zamoyska, H., and Tybulewicz, V. L. J. (1997). A requirement for the Rho-family GTP exchange factor Vav in positive and negative selection of thymocytes. Itnrntitiity 7, 451-460. van Merrwijk. J. P., Marguerat, S., Lees, R. K., Germain, R. N., Fowlkes, B. J., and MacDonaltl, H. R. (1897). Quantitative impact of thymic clonal deletion on the T cell repertoire. J. Exp. Med. 185, 377-383. Vasquez, N. J., Kane, L. P., and Hedrick, S. M. (1994). Intracellular signids that mediate thymic negative selection. linnimity 1, 45-56. Vella, A. T., McCormack, J. E., Linsley, P. S., Kappler, J. W., and Marrack, P. (1995). Liiiopolysaccharide interferes with the induction of peripheral T cell death. lrnnitrnity 2, 261-270. Verdagtier, J.. Schmidt, D., Amrani, A,, Anderson, B., Averill, N., and Santamaria, P. (1997). Spontaneous autoimmune diabetes in monoclonal T cell nonobese diabetic mice. J . E x p . Med. 186, 1663-1676. Volkmann, A,, Zal, T., and Stockinger, R. (1997).Antigen-presenting cells in the thymus that can negatively select MHC class 11-restricted T cells recognizing a circulating self antigen. J. b n n l r m l . 158, 693-706. von Boehmer, H., and Sclinliiger, H. K. (1984). Thymocytes appear to ignore class I major histocompatibilitycomplex antigens expressed on thymic epithelial cells. Eur. J. lnwitinol. 14, 1048-1052. Wahinas, T. L., Sperling, A. I., Khattri, R., Thompson, C. R., and Rliiestone, J. A. (1996). CD28 expression is not essential for positive and negative selection of thymocytes or peripheral T cell tolerance. J. Ininiroid. 156, 1006-1013. Wang, S. D., Huang, K. J., Lin, Y. S., and Lei, H. Y. (1994). Sepsis-induced apoptosis of the thymocytes in mice. J. h i r n t m d 152, 5014-5021. Washington, E. A,, Kimpton, W. G., Holder, J. E., and Caliill, R. N. (1995). Role of the tliyiniis in the generation of skin-hoining alpha beta and gamma delta virgin T cells. Eur. J . Ininiunol. 25, 723-727. Waterhouse, P., Marengere, L. E. L., Mittriicker, H.-W., and Mak. T. W. (1996). CTLA4, a negative regulator of T-lymphocyte activation. Itnnznnol. Reu. 153, 183-207. Webb, S., Morris, C., and Sprent, J. (1990). Extrathymic tolerance of mature T cells: Clonal elimination ils a consequence of immunity. C d 63, 1249-1256. Weginann, D. R. (1996). The immune response to islets in experimental diabetes and insulin-dependent diabetes inellitus. Curt-. @in. Inimunol. 8, 860-864. Weiner, H. L. (1997). Oral tolerance: Iinmiine mechanisms and treatment of autoiininuiie diseases. lnitnrinol. Today 18, 335-343. Zal, T., Volkmann, A,, and Stockinger, B. (1994). Mechanisms of tolerance induction in major histocoinpatibilitycomplex class 11-restrictedT cells specific for a blood-home selfantigen. J. Exp. Mcd. 180, 2089-2099. Zal, T., Weiss, S., Mellor, A,, and Stockinger, B. (1996). Expression of a second receptor rescues self-specific T cells from thymic deletion and allows activation of autoreactive effector function. Proc. Nntl. Acnd. Sci. USA 93, 9102-9107. Zelenika, D., Adams, E., Mellor, A., Simpson, E., Chandler, P., Stockinger, B., Waldmann, H., and Cobhold, S. P. (1998). Rejection of H-Y disparate skin grafts by monospecific
CD4’ T helprr 1 (TI111 ant1 T hclprr 2 (Th2) cells: No recjuireinent for CD8’ T cells o r 1% cells. J.
Z t t u f m I i c t / . 161, 1868-1874. Zhoii. T.. C h g . J.. Yang, P.. Ll’iiiig, Z..Liii. C . . SII,X., Hluetlimunn, H., and Mountz, J. D. i1996).Inhil>ition of NiirZ’/Niirrl l e d s to inefficient cloniil tlrletion of self-reactive T cells. J. Esp. Jfcd. 183, 1879-18<)2. Ziiikcmagel, 11. M., Callahan, G. N.. Kleiii, I., and Denneit, G. (1979). Cytotoxic T cells learii specificit?. for self H-2 during tliffrrentiation in the thyinus. Nuturf 271, 2 5 - 2 5 3 .
This Page Intentionally Left Blank
kD\AI\.C F5 IN I M M U N O I O(.I \’OL i l
Confronfation between lntracellular Bacteria and the Immune System ULRICH E. SCHAIBLE, HELEN 1. COLLINS, AND STEFAN H. E. KAUFMANN Max-Planck fnstituk for Infection Biology, D- 10117 Berlin, Germony
1. Introduction
The complex contest between bacteria and tlie mammalian host follows different game plans that are undertaken to ensure the survival of both the pathogen and the host. Conseqiiently, both players have developed tactics to counteract their opponent’s strategies. Some pathogens rely on a “long distance” approach via the production of toxins or noxious virulence factors, which exert their effects systeniically and are counteracted in a similar fashion by the production of antibodies by the host. In contrast, the interplay between intracellular bacteria and their hosts resembles a more intimate game of cat and mouse with frequent changes in tactical advantage. This is mostly because these bacteria use as their habitat one of tlie major defense players for the host, namely macrophages. As a result of this, the foremost line of rapid defense is insufficient for appropriate resistance, and other team players, such as T lymphocytes, must be rallied to ensure victory. Activation of the appropriate T-cell subsets relies on a delicate integration of the whole immune system, which is orchestrated by the early events of infection and finely tuned throughout the duration of the host-pathogen interaction. This review focuses on the early interplay between intracellular bacteria and host cells, which, on the one hand, allows the microbe to gain the initial advantage by establishing itself within an intracellular habitat and, on the other hand, determines the type of T cell rebuttal evoked. The resulting host response is not the exclusive domain of major histocompatibility complex (MHC) class 11-restricted CD4+ T cells as originally thought, but encompasses the whole spectrum of T lymphocytes ranging from MHC class I-restricted CD8+ T cells to unconventional T cells with specificity for nonproteinacious ligands. It is the authors’ opinion that this entire T lyniphocyte repartee is required for optimuni protection, and that at least some of the unconventional T cells have developed as direct counterparts for bacterial pathogens. This review tries to discuss in equal depth the tactics of both host and pathogen engaged in this intricate struggle, with an emphasis 011 the first act and its impact on stimulation of an appropriate T-cell response. The authors are convinced that studies into these interactions not only provide unique opportunities 267
(:o,>)ngl,tA 1999 tl., .4CiUd?llIL(. Pr<% All Iighti 01 r q m d u c t m i i n +y form r e w ~ e d O(KiS-?iiR/YY 9 3 0 (X)
268
UI.1UCI-I E SC:1IAIRLE
et Ill
for expanding our knowledge about tlie extensive potential of the immune system, but also pave the way to the rational design of novel vaccines and improved chemotherapy. After all, conventional vaccines have more or less failed to provide adequate control measures against intracellular bacteria. It is hoped that learning froin nature will provide us with the logic required for constructing a novel generation of more successful vaccines. II. What Is an lntracellular Pathogen?
By definition, pathogenic bacteria that are endowed with strategies to survive in vertebrate cells are termed intracellular bacteria. This capacity protects these pathogens from the defense armainentarium active in the host extracellular milieu such as specific antibodies and complement. The main protective force against infections with intracellular bacteria encompasses T cells expressing a T h l cytokine pattern, in particular interferon y ( IFNy), which activates macrophages. The activated inacrophage with its inicrobicidal machineIy is the main effector cell against such pathogens. In principle, any type of tissue cell, including epithelial cells, endothelial cells, hepatocytes, macrophages and dendritic cells, potentially serves as habitat. However, for most of the intracellular bacteria discussed in this review, the macrophage represents the typical host cell. Professional antigen-presenting cells (APC) such as macrophages and dendritic cells are in charge of uptake and processing of foreign antigens, such as those from invading bacteria, and of induction and maintenance of the specific T-cell response against these antigens. The long-lasting coevolution between pathogen and host leads to the mutual development of unique intracellular survival strategies on the pathogen’s side and reciprocal defense mechanisms on the host’s side. Consequently, the intracellular location of the pathogen is crucial for the type of immune response elicited. After their engulfment, facultative intracellular bacteria survive and replicate in professional phagocytes, but they can also proliferate outside host cells. This group includes the following bacteria and diseases: Salrrwnella spp. (typhoid fever and other salmonelloses);Fruncisella tularensis (tularemia); Brucella abortus (brucellosis); Listeria nmnocytogenes (listeriosis); Mycobacterium tuberculosis (tuberculosis); M . avium-intracellulm-e complex and other atypical inycobacteria causing opportunistic mycobacterial infections in the immunocomproinised host; Legionella pneuunmphila (legionnaire’s disease); and Burkholdem‘n pseudomallei (melioidosis). M . leprae, the agent for leprosy, may also be considered a Facultative intracellular bacterium, although extracellular survival has not yet been proven. Some facultative intracellular bacteria enter nonprofessional phagocytes temporarily, exploiting this capability to invade the host through epithelial cells.
These include Yc ia pestis (plague), k’. entcrocolitica (enteritis), and Sliigella spp. (sliigellosis).Some bactena are preferentially found in endotlielial cells but can also survive in macrop1i;iges such as the cat scratch disease agents Bartonelki hensclae and Afipin fclis A second group of intrucellnlar bacteria depend fully on the- host cell inetabolisin and are hence called obligate intracellular patliogens They cannot survive and replicate outside of their host cells. This group encompasses Rickettsia proiotizecki (louse borne typhus); R rickettsiae (Rocky inountain spotted fever); R tsutsugaititisi (scrub ty~hus); Ehrlichin emir (ehrlichiosis); Chlanzydin tmclionintis, (trachoma, lympliogranuloiiiavenereurn); C. pneuriwniae (pneumonia); C prittaci (psittacosis); ant1 Cosiella biirnettii (Q fever). Some eukarvotic parasites and pathogenic fiingi are also able to survive inside host nkmqhages, such as ineinbers of the Leishiifinin genus (KdlaAzar, etc.);Ti-rjpai~oso~iin c r x i (Chagas’disease); Toxoplasiiia gondii (toxoplasmosis); and Histoplasma crysdritutii (histoplasmosis). Although this review focuses on bacteria, on some occasions it will refer to these organisms whenever it is of general relevance to understand the cell- and immunobiology of intracellnlar parasitism. 111. How to Enter the Host Cell Is How to Survive
The first step toward successfill suivival of an intracellular pathogen in the host is to gain entry into host cells. Several cellular receptors are exploited by intracellular pathogens for binding and induction of engulfineiit. Binding is either directly to the receptor or via host ligands deposited on tlie bacterial surface. Bacteria bound to the cell surface in either way are subsequently taken up as pathogenic cargo. Other pathogens, such as the intracellular parasite T. gondii, force themselves actively into the cell, incuding a vacuole that does not resemble a norinal phagosome ( Joiner, 1997). The uptake mechanism via specific receptors has been terined tlie “zipper mechanisin” to distinguish it from processes where specific adhesion is apparently not a prerequisite for engulfinent (Swanson and Baer, 1995). This latter process, called “trigger ineclianism,” has been proposed for the uptake of salmonellae. At the site of attachment, this organism induces extensive membrane ruffling of the host cell surface. These membrane ruffles subsequently enclose the salmonellae, often together with other inaterial in their vicinity, a ineclianisni termed macropinocytosis (Francis et al., 1993; Alpuche-Aranda et 0 1 , 1994; Swanson and Baer, 1995).AIthougli the cellular receptors involved in engulfment of salmonellae are not yet known, it has been proposed that bacteria exploit the host cell signaling cascade to induce cytoskeletal rearrangements and membrane
270
ULRICII E. SCHAIHLE ef
N/
ruffles. This notion is based on the following two findings: (i) In one mouse cell line (Henle407) S. enterica serovar Typhiinurizm (S. typhimurium) can induce phosphorylation of the epidermal growth factor (EGF)receptor similar to EGF induces endocytosis of its receptor (Galin et d.,1992). However, because S. typhimurium readily invades cells that do not express the EGF receptor, the current view is that this receptor is not involved directly in host cell entry of S. typhimurium ( Jones et al., 1993; Rosenshine et al., 1994; McNeil et al., 1995; Finlay and Fdkow, 1997). (ii) Sabmnellamediated activation of CDC42, a member of the small GTP-binding protein superfamily Ras, has been found to be involved in the induction of cytoskeleta1 rearrangements and membrane ruffling (Chen et al., 1996). Binding to certain receptors induces active phagocytosis. Only professional phagocytes such as neutrophils, macrophages, and immature dendritic cells perform this host-directed internalization process. Several receptors of the so-called “pattern recognition receptor” f-dmily (Medzhitov and Janeway, 1998) can be exploited by intracellular bacteria to induce phagocytosis. These receptors possess lectin-like binding moieties in their extracellular domains that enable them to directly bind a broad array of sugar residues exposed on the surface of various bacteria. Mannose-type receptors such as the macropliage mannose receptor ( M MR) bind mannose, glucose, and fucose residues; galactose and fucose-type receptors bind galactose and fucose, respectively; and both types bind N-acetylglucosamine (Stahl, 1992; Stahl and Ezekowitz, 1998).The MMR is a monomeric transmembrane protein with eight lectin-like domains for carbohydrate recognition (Taylor and Drickamer, 1993).This receptor recognizes terminal inannose residues exposed on mycobacterial surfaces such as mannosecapped lipoarabinoinannan (LAM) or phosphoinositolmannosides (PIM) (Schlesinger, 1993; Emst, 1998; Ehlers and Daffk, 1998). Interestingly, the MMR binds the virulent M . tube?-culosisstrains H37Rv and Erdman but not the avirulent strain H37Ra, although both strains contain the same amount of terminal dimannosyl residues (Schlesinger et al., 1994, 1996). Another pattern recognition receptor, CD 14, mediates phagocytosis through the binding of cell wall components such as the lipopolysaccharides (LPS) of gram-negative bacteria, or mycobacterial LAM with terminal arabinose (Pugm et al., 1994; Wright, 1995; Grunwald et al., 1996). It also facilitates the uptake of nonopsonized M . tuberculosis by human microglial cells (Peterson et al., 1995).In addition, CD14 binds soluble peptidoglycans and muramyldipeptide (Weideinann et al., 1997). A homolog of the Drosophila TOLL protein has been identified as a putative pattern recognition receptor, although its ligand is still unknown (Medzhitov et al., 1997). The macrophage scavenger receptors (MSR-AI, MCR-AII) that usually recognize serum lipids and lipoproteins (Krieger and Herz, 1994) can also
bind cell wall components of gram-negative and gram-positive bacteria and can induce pliagocytosis of various bacterial species, including M . tubcrculosbu (Pearson, 1996; Ziminerli ct nl., 1996). The MSR may play a more prominent role in Imcterial uptake than previously assuined, as mice with interrupted MSR-AIAI genes clear L. niowcytogenes infection less efficiently than wild-type mice (Suzuki et ul., 199S).'This is consistent with the fact that the type I MSR can bind the inajor cell wall component of grani-positive bacteria, lipoteichoic acid (LTA) (Dunne et al., 1994). A third inember of the MSR-A fiiinily, MAKCO, which has been cloned, is apparently involved in binding of bacteria to macrophages (Elomad ef nl., 1995). At present it is unclear whether MSK activate phagocytosis alone or whether additional molecules participate in bacterial uptake. Another mammalian surface protein that 1)inds glycolipids, CD48, has been found to serve as a receptor for the fimbriae of invasive FiinH'-Eschcrichin coli. Members of this species are killed rapidly once within professional phagocytes and hence are generally considered to be extracellular bacteria. However, uptake by CD48 promotes survival of E. coli in macrophages, thus interfering with phagosome maturation (Baorto et ul., 1997). Certain bacteria misuse unique host cell surface molecules to induce uptake as described earlier for the EGF receptor. Thus, L. rrwnocytogenes binds E-cadherin on intestinal epithelial cells via internalin expressed on the bacterial surface (Mengaud ct al., 1996; Finlay and Cossart, 1997; Ireton and Cossart, 1997), whereas invasins of Yersinin spp. and Shigellu spp. mediate adhesion to pl integrins of epithelial cells (Isberg and-Van Nhieu, 1995; Watarai et d . , 1996). The affinity of M . leprne for Schwann cells of peripheral nerves has been explained by specific binding of these niycobacteria to the G domain of the a2 chain of laminin (Rambukkana et al., 1997). Bacteria ;ittach to professional phagocytes indirectly via various serum components such as iininunoglobulins ( Ig), complement components (C), or fibronectin (Fn), via Fc receptors ( FcR), coniplement receptors (CR), or the Fn receptor (FnK), respectively. Specific IgG bind bacteria and mediate uptake via the FcR and, after C fixation, via CR. The role of specific IgG in infections with intracellular bacteria is still unclear, but infected individnals frequently have vast amounts of circulating specific antibodies. The uptake of IgG-opsonized M . tuberculosis both induces a respiratory burst, which produces reactive oxygen intermediates (ROI ), and permits fusion with lysosomes, yet without significant effects on mycobacterial viability (Armstrong and Hart, 1975). All C R induce phagocytosis of bacteria opsonized with C3 breakdown products (Brown, 1991). Like inany pathogens, most intracellular bacteria, including mybacteria, can activate the alternative pathway of C activation,
272
ULKlCll E. SCHAIRLE
at a[
resulting in opsonization by C3b and CSbi. CR1 is a monomeric transmembrane protein that binds C3b and C4b. C3bi is bound by CR3 and CR4, which both belong to the integrin superfamily and contain the common P2 subunit paired with either a M ( C D l l b ) or olx (CD1lc) chains, respectively (Ahearn and Fearon, 1989; Brown, 1991). Internalization via CR does not induce a respiratory burst and therefore represents a way to escape this defense mechanism of the host cell. C. psittaci and one serovar of L. pneumophila can be engulfed by an unusual process called “coiling phagocytosis,”which has been characterized by electron inicroscopy (EM) to be performed by a single pseudopodiuin wrapped around the bacteria in several layers (Wyrick and Brownridge, 1978; Horwitz, 1984). Coiling phagocytosis is promoted by C3 fixation on L. pnezinwplaila and fails to induce ROI, whereas preopsonization with specific IgG facilitates normal pliagocytosis. M . tuberculosis (Schlesinger et al., 1990), M . leprae (Schlesinger and Honvitz, 1991), and L. pnezimphila (Payne and Honvitz, 1987) can induce their own CR-mediated uptake through C3 breakdown and fixation of the resulting products on their surface. This is facilitated by the phenolic glycolipids of M . leprae and the major outer ineinbrane protein of L. pneuinophila. Several inycobacterial species ( M . tuberdosis, M . avium, and M . leprae) can cleave serum C2 to become the C3 convertase, C2a, allowing C3 conversion to C3b and its fixation (Schorey et al., 1997). M . tuberculosis can also bind directly to CR3, independent from C fixation, at a site distinct from the binding site for C3bi (Stokes et al., 1993). This binding can be inhibited by a seaweed-derived 0-glucan, laininarin, by N-acetylglucosamine,and by mycobacterial glucan or mannan that has been described to form a capsule around the bacteria (Cywes et al., 1997; Ehlers and Daff6, 1998). Recognition occurs either via the integrinbinding sequence RGD or, similar to that by pattern recognition receptors, via a lectin-binding site for 0-glucan (Cywes et al., 1996). Binding to the P-glucan site is apparently mediated by the capsular polysaccharide PIM, which can be stripped from the bacterial surface by glass bead treatment (Cywes et al., 1997). The mode of interaction of M . tuberculo.siswith CR3 is apparently of varying complexity dependng on the strains investigated, i.e., some isolates bind CR3 directly whereas others require opsonization with C3 breakdown products (Cywes et al., 1997). In addition to membrane-associated pattern recognition receptors, soluble molecules with lectin-like moieties may play a role in the uptake of intracellular pathogens. One of the most prominent molecules is the pulmonary surfactant protein A (SP-A), which binds M . tuberculosis and mediates its uptake by alveolar macrophages in the lung (Downing et al., 1995; Gaynor et nl., 1995; Pasula et al., 1997).Two potential receptors for SP-A have been described: (i) ClqRp, a 126-kDa protein expressed on
macrophages and endothelial cells, which binds both C Iq and the mannosebinding protein (MBP) (Nepomuceno ef nl., 1997); and (ii) SPRB10, a 210kDa protein on macrophages aiid type I1 pneumocytes that does not bind Clq (Chroneos et al., 1996). Antibodies to SPRBlO block the uptake of SP-A-coated M . bovis Bacillus Cahnette-GuCrin (BCG) by macrophages (Weikert ct nl., 1997). Due to its similarity to Clq, SP-A may also bind to CR1, akhough as yet there is no experiineiital evidence for this. A newly described member of the scavenger receptor family, gp-340, binds lung surfktant protein D and its role in the uptake of inycobacteria remains to be determined (Holinskov ct d., 1997). A similar function has beeii proposed for soluble receptors present in the serum, such as the MBP that opsonizes M . nviicm and A4 tubcwiilosis via LAM and PIM (Hoppe et a/., 1997; Polotsky (it d., 1997). Another serum protein, Fn, can induce phagocytosis of those bacteria expressing Fn-binding proteins, which include a fanlily of secreted 32-kDa proteins of R4. tuberctilosis, the Ag8S coinplex (Abou-Zeid et al., 1988; Borrenians et d . , 1989). The binding of Fn to the FnR is facilitated via the cellular adhesion site represented by the RGD sequence. The importance of Fn for the uptake of niycobacteria is still in debate for the following reasons: (i) Fn is only a weak inducer of phagocytosis and Fn-mediated uptake probably depends on cofactors (Yamada, 1991) arid (ii) the priine role of the 32-kDa protein is to function as mycolyl traiisferase in cell wall synthesis (Belisle et nl., 1997). Moreover, M . triherciilo,siscan also bind heparin, which has been suggested to promote uptake by macrophages (Menozzi ct id., 1996). Finally of interest are observations that M . auiunz organisms isolated from infected macrophage cultures have an increased efficiency to invade other inacrophages via a CH3-independent pathway (Berrnudez et al., 1997). Similarly, L. prieiiinoplzilo invasion of macrophages is increased following culture of the bacteria in their natural host, free-living amoebae (Cirillo ct al., 1997). In both cases it can be speculated that the intracellular milieu induced bacterial genes that are involved in host cell invasion. IV. induction of Nonspecific Immunity
The counteraction of host invasion by pathogens at its roots demands a rapid response, which is generally provided by components of the innate immune system. This nonspecific response develops rapidly and precedes the time-consuming clonal expansion of antigen-specificlymphocytes. Cells critical in these responses are primarily mononuclear phagocytes and natural killer ( N K ) cells, which produce the necessary cytokines and effector inolecules to contain the pathogen in question and to promote the development of an appropriate antigen-specific response.
274
ULHICH E SCHAIRLE et ul.
A. INDUCTION OF CYTOKINES 1. lL-l2/1L-l(YIlFN~ Relatively early after infection with an intracellular pathogen, macrophages, having received the appropriate signals, are activated for the defense mechanisms outlined later. The most important cytokine in this process is IFNy, which is produced in this early stage of infection by NK cells (Bancroft et ul., 1991). However, it is now clear that the production of IFNy by NK cells depends on prior activation of the N K cell by IL12, by interferon y inducing factor (IL-l8), and, to a lesser extent, by tumor necrosis factor a (TNFa). This then results in a proinflammatory feedback loop with IFNy acting on macrophages to sustain IL-12 production, which may be a critical pathway for the control of pathogens that are poor inducers of IL-12 such as M . bovis BCG (Flesch et al., 1995; Matsuinoto et al., 1997) and Toxoplasim (Sousa et nl., 1997b) (see discussion of Thl cell development in Section VII). The induction of IL-12 by pathogens occurs primarily in a T-cellindependent manner as demonstrated by the production of this cytokine in T-cell-deficient mice (Tripp et al., 1994).Thus, components of intracellular bacteria, including LAM and LTA. induce IL-12 (Cleveland et nl., 1996; Yoshida and Koide, 1997). It has been demonstrated that bacterial DNA, specifically unrnethylated CpG motifs, can also induce macrophage IL-12 production, leading to the secretion of IFNy by NK cells (Chace et al., 1997). The importance of this cytokine cascade in host responses to intracellular bacteria has been demonstrated conclusively by the overwhelming susceptibility of IL-12 knockout (KO) mice to infection with M . tuberculosis (Cooper et al., 1997b) in which the increased bacterial load and decreased survival times were linked to the absence of IFNy produced by both innate and acquired immune responses. Indeed, in IFNy KO mice a similar uncontrolled growth of bacilli was observed, reflecting that in the absence of IFNY, macrophages were not activated to an antimycobacterial state (Cooper et al., 1993; Flynn et al., 1993). Patients with a deficiency in either IL-12~40or the p chain of the receptor for this molecule have been described. In both cases, their susceptibility to mycobacteriosis and salmonellosis was increased greatly, resulting in the dissemination of M . bovis BCG and of environmental mycobacteria and salmonellae (Altare et al., 1998; de Jong et al., 1998). A new cytokine, IL-18, has been described that is produced by monocytes and is also a potent inducer of IFNy from T cells and NK cells. This molecule is closely related to the IL-I family of cytokines, although its effects on IFNy production are more potent than either IL-la or IL-lp (Hunter et al., 1997). Although its exact role in antimicrobial infections is
INTRACEl,l,liLAR BA(:TEHIA A N D THE: I M M U N E SYSTEM
27ij
unclear, the demonstration that IL-18 KO inice are deficient in both NK cell fiinction and Tlil cell development (Takeda et al., 1998) suggests that this cytokiiie will also feature as a central mediator in host resistance to intracelliilar bacteria. Recently it has been deinonstrated that murine bone marrow macrophages ( B M M ) can be induced to secrete IFNy by the synergistic action of IL-12 and IL-18, suggesting that macrophages, once triggered by this cytokine combination can promote Thl cell development (Munder ef nl., 1998).
2. 1°F The production ofTNF in bacterial infections is a finely balancedphenomenon, being required for protective processes such as macrophage activation and granulonia formation, and yet TNF is also one of the prime mediators of innnunopathology when present in excess amounts (Beutler and Cerami, 1990).Activation of macrophages in vitro with TNF 1987; Cadranel et d., increases their antibacterial properties against a variety of intracellular pathogens (Langermans and van Furth, 1894;Pasparkis et nl., 1996).Experiments using antibody neutralization in vivo as well as treatment of inice with recombinant cytokine have deinonstrated a key role for TNF in secondary host responses to listeriae and salmonellae (Langermans and van Furth, 1994;Samsom et nl., 1995).These observations have been subseqrientlyconfirmed using KO mice deficient in either the cytokine itself or the p55 TNF receptor (Pfeffer et nl., 1993; Rothe ef nl., 1993; f y n n et al., 1995).Further dissection of responses in these deficient mice suggests that this cytokine is involvedin both the activation and recruitinent ofvarious cell types, including hepatocytes and macrophages in the liver, and the migration of granulocytes and inonocjes from the bone marrow to the circulation (Kindleret al., 1989; van Furth ct al., 1994).Zn oitro, induction ofTNFa in macrophages by mycobacteria has been deinonstrated and appears to be mediated by diverse components, including HspCiO (Peetermans et al., 1995) and LAM (Roach et nl., 1993). Interestingly, this LAM function is restricted to the lipid portion of its mannose-capped form (Roach et al., 1993; Nigou et nl., 1997).The activation of both human and murine macrophages by TNF in vitro enhances their antimycobacterial activity ( Flesch and Kaufinann, 1990b),whereas in tjivo, TNF is found as both mRNA and protein in the lungs ofpatients with pleural tuberculosis who mount a protective immiine response (Barnes et a1 , 1990).
3. IL-6 Interleukin 6 is a proinflammatory cytokine whose primary functions are the regulation of hematopoiesis (Berna ct nl., 19941, the promotion o f differentiation of cytotoxic T lymphocytes (Takai et a l , 1988), and the regulation of acute-phase protein synthesis (Kopfet nl., 1994).This cytokine
276
IJLRI(:H E SCHAIBLE et ol.
is produced in uiuo in mice infected with listeriae and appears to be produced in a TNFa and IFNy-independent manner, whereas in uitro the addition of IL-6 to bone marrow-derived macrophages (BMM) enhances their listericidal (Denis and Gregg, 1991; Have11 and Sehgal, 1991) and antimycobacterial activities ( F l e d and Kaufniann, 1990a). Murarnyl dipeptide and LAM from inycobacteria induce IL-6 production in niacrophages (Sanceau et nl., 1990; Zhang et nl., 1994); however, to date there is still contradictory evidence as to what role IL-6 plays in the infectious process. Mice depleted of IL-6 by mAb treatment in uivo showed increased susceptibility to Lbteria infection (Liu et al., 1994) and a reduced efficacy of BCG vaccination for protection against subsequent challenge with 2tl. aviurn (Appelberg et al., 1994). VanHeyningen and colleagues (1997) found that IL-6 produced from macrophages infected with mycobacteria drastically diminished T-cell responses to both mycobacterially derived and exogenously added, unrelated antigens. However, infection of IL-6 KO mice via aerosol with M. tzibercukJsis did not appear to affect T-cellmediated control of the infection, although KO mice produced less TNFa and IFNy early in infection, suggesting that IL-6 plays a role in innate responses and macrophage activation (Cooper et al., 1997b).After intravenous (iv) inoculation with M. tuberculosis, IL-6 KO mice succumb to infection earlier than wild-type mice, although both are approximately equally resistant to M. bovis BCG (Lade1 et nl., 1997).
4. ZL-1 IL-1 comprises two structurally related molecules, IL-la and IL-lP, with similar biological activities. It is produced by a variety of cells such as monocytes, dendritic cells, and macrophages. Immunoregulatory properties proposed for these molecules include the stimulation of T-cell proliferation and costimulation of antigen presentation. IL-1 can also potentiate the IL-12-mediated induction of IFNy from NK cells and is thus implicated in the T-cell-independent resistance mechanisms to intracellular pathogens (Hunter et nl., 1995). Finally, IL-1 shares proinflamrnatory activities with IL-6 and TNFa. Thus, mice that lack IL-1 receptor antagonist (IL-IRA), a competitive inhibitor of the binding of IL-1 to its receptor, are less susceptible to infection with listeriae, whereas overproducers of this molecule succumb inore rapidly to the infection (Hirsch et al., 1996). In uitro infections of monocytes from HIV-infected individuals with M . nzjiim result in the production of IL-la and IL-lP, although much less than that induced in monocytes from healthy individuals (Johnson et nl., 1997), suggesting that this cytokine plays a role in the control of mycobacterial infections.
INTHr\C:EI.I,ULAR HAC‘TERIA AND TlIE IhfMIJNE: SYSTEM
277
5. Chenzokines Cliemokines are a family of structiirally related secreted proteins that play a fundamental role in leukocyte recruitment and trafficlang. They can he broadly divided into three supergene families based on the expression of a conserved cysteine motif (C).There are C-C cliemokines, including macrophage inflaniniatoryprotein 1P ( MIP-1P) and inacrophage chemotactic proteins (MCP) 1 , 2 , and 3: C-X-C chernokines, including inacrophage inflammatory protein-2 (MIP-2) and IL-8; and C chemokines comprising lymphotactin. The role of these proinflammatory molecules in the course of intracellular bacterial infections has been increasingly appreciated. Thus, chemokines are induced followingthe exposure of macrophages, monocytes, and polymoryhonuclear granulocytes, as well as alveolar epithelial cells, to a variety of bacteria and their coinponmts, such as listeriac (Flesch ct a/., 1997;Arnold and Konig, 1998),L. pieu t w p / i i L ( i , S . tijplziiiznrizm ( Yamainoto et al., 1996b), and mycobacteria (Riedel and Kaufniann, 1997; VouretCraviari et al., 1997; Lin et al., 1998).In the 1)ronchoalveolarlavages of patients during the acute phase of active pulinonary tuberculosis, increased levels of MCP-1 and IL-8 are observed in coinparison with healthy controls (Kurashima et nl., 1997) and correlate with an increase in the numbers of neutrophils and lymphocytes present in the lung. Furthermore, in the murine model of infection with three separate strains ofvinilent M . tuberculosis, a substantial cheinokine response was generated in the lungs of infected mice, although to a varying degree among Ixicterial strains (Rlioades et nl., 1995). A further illustration of the importance of these inolecules in host defense conies from experiments where the early susceptibilityof beige mice to infection with M . nvirini was attributed to the lowered expression of MIP1j3 and MIP-2 in the lungs of these mice, resulting in defective neutrophil recruitment (Floridoet nl , 1997).Knockout mice lacking the CCR2 chemokine receptor, which is the receptor for MCP-1, were shown to be defective in monocyte recruitment and unable to clear infection with listeriae (Kurihara et d ,1997). Furthermore, these mice exhibited both a decrease in granuloma size and a dramatically reduced level of IFNy secretion in response to experimentally induced purified protein derivative (PPD)mediated granulomatous disease (Boring c>t al.,1997). However, separate experiments using mice deficient in MCP-1 itself revealed that although the secondary piillnonary granulomatous response to Schistosomci nuinsoni eggs was diminished, these mice were indistinguishable from wild-type mice in their ability to clear infections with M . tiihercnlosis (Lu et al., 1998). It is tempting to speculate that the discrepancy in these results may relate to the type of infection that the pathogens induce, with the acute, relatively T-cellindependent immune response induced by listeriae being more dependent
278
UL.RICH E. SCHAIULE
i<
rd
on chemokine production than the chronic disease state resulting from mycobacterial infection. The control of the latter infection essentially requires a T-cell response that occurs independently of MCP-1 production.
B. MACROPHAGE ANTIMICRORIAL DEFENSE MECHANISMS 1, Toxic Effector Molecules . . . and How to Avoid Them The highly reactive antimicrobial nitric oxide radical (NO ) is derived from 1.-arginineand molecular oxygen in a reaction catalyzed by nitric oxide synthase (NOS).The NO acts as an oxidizing agent alone or interacts with 0; to form peroxynitrite (ONOO-),which is further transformed to NO via nitrite and nitrate intermediates, which can be detected experimentally as NO. There are three distinct forins of NOS: NOS1 and NOS3, which are calcium dependent and constitutively present in tissues to maintain housekeeping functions, and NOS2 or INOS, a calcium-independent form that can be induced in a variety of cell types in response to immunological stimuli such as IFNy, TNF, and IL-1 (reviewed in Fang, 1997) as well as microbial products such as LPS and LTA (Raetz et al., 1993; English et al., 1996). The role of NO as a primary macrophage antimicrobial effector molecule has been addressed in vitro, where its actions can be specifically blocked using the inhibitor NG-monomethyl-L-arginine (L-NMMA).Using this system it has been shown that macrophages activatedwith IFNy can inhibit the intracellular replication of a variety of intracellular bacteria, including B. pseudomnllei, L. pneunloplzila, M . bovis BCG, and M . tuberculosis in a NOdependent manner (Flesch and Kaufmann, 1991; Chan et al.., 1992; Yamamot0 et d.,1996a; Miyagiet d.,1997; Rhoades and Orme, 1997).In the case of both M . tuberculosis and M . auium, differential susceptibilities of various strains to the effects of activated macrophages have been described (Flesch and Kaufmann, 1988;Appelberg and Orme, 1993).Additionally, such variation in susceptibilityto NO of different strains of M . tuberculosis was apparent in a cell-free system, but again did not correlate with the ability of these bacteria to resist intramacrophage NO-mediated killing (Rhoadesand Orme, 1997).The selection of strains of M . tuberculosis that can survive exposure to NO and ROI has identified a novel antioxidant gene noxrl, which confers enhanced ability to resist killing via NO and ROI both extracellulary and within macrophages (Ehrt et ul., 1997). The cloning and subsequent gene disruption of nos2 (iNOS) have provided further insights into the antimicrobial functions of this molecule during the course of infection. For infections with intracellular pathogens the antimicrobial effects vary with the pathogen investigated. Experimental infection of iNOS KO mice with M. tuberculosis revealed that these mice are extremely susceptible to disease, resulting in a shortened survival time
as well as increased bacterial loads within the lung, spleen, antl liver (MacMicking et al., 1997). Intlcecl, the increased susceptiliility w a s as dramatic in these mice ;is that seen in the IFNy and TNF p55 receptor KO mice, probably reflecting the fact that these two cytolanes are critically involved in iNOS induction (Nathan, 1997). In contrast, iNOS-deficient mice infected with M . m i t i i n siiffer no greater replication of bacteria within the organs examined, but instead reveal tliat lvmpliocytes are suppressed in their responses to mitogen whcw compared to wild-type animals (Doherty and Slier, 1997).This is similar to tlie observed suppression of priiiiai-y immune responses when normal mice are infected with L. Ino,roc2/togo~e,s (Gregory and Wing, 1993). Indeed, iNOS-deficient mice are only slightly more susceptible to listerial infection than wild-type mice ( MacMicking et d . , 19951, a situation that is also apparent in experimental infection of mice with C. trachonintis iii which there is no significant difference in the course of genital infection between iNOS-deficient and \Yild-type mice, iniicrodespite the substantial induction of NO in Chlnl~i2/clic/-infecte(~ phages (Igietseme et nl., 1998). Nitric oxide (NO) production has been iiiiplicated in tlie maintenance and control of microbial latency. Mycobacteria frequently establish asymptomatic chronic infections within the host. Inhibition of NO in experimental models of hf. tri1icrctilosi.r.infection has resulted in prompt reactivation of the growth of the bacilli, suggesting a role for continuous NO production in persistent and latent infections (MacMicking et d., 1997). Despite the experimental models implicating NO as a critical effector molecule in the host antimicrobial defense,there is still controversy about the role of iNOS in hunian discwe a s experimentally it was difficult to detect NO prodiiction from healthy monocytes and macrophages even following i t i z;itro stimulation. However, iNOS expression has been detected in monocytes isolated from patients suffering from a variety of inflammatoryand infectious diseases ( r e ~ i ~ine MacMicking d et d., 1997). For example, inflammatory alveolar macrophages isolated from patients with active tuberculosis produce large amounts of NO (Nicholson et nl., 1996),antl indeed inflammatory macrophages are induced to express iNOS in uitro following M . ttiberculosis infection and appear to use NO to control the intracellular replication of tlie mycobacteria (Nozaki et al., 1997). Additionally, in uitro infection of a human intestinal epithelial cell line with S. entetico serovar Dublin r c d t e d in increased iNOS expression and NO production, provided that the epithelial cells had been primed with IFNy. This upregulation occurred in an LPS-independent manner and was augmented by an enteroinvasive phenotype of the bacteria (Salznian et nl., 1998).These experiments suggest that although the exact fiinction
280
IJLHICII E SCHAIB1.E r f 01
of iNOS in the human system requires further elucidation, this molecule may be important in host defenses, at least at niucosal surfaces. Another process at the macrophages clisposd is the production of ROI, which is initiated by an NADPH oxidase following FcR ligation or activation of the cell by IFNy and LPS. Further metabolic steps catalyzed by superoxide dismutase and appropriate iron molecules result in the production of oxygen ra&cds, which are powerful oxidants and cause DNA damage, as well as alterations in membrane lipids and proteins (reviewed in Fang, 1997). Although it is generally thought that NO is more important in controlling intracellular bacterial replication, there is a certain amount of cooperation between the two systems. Simultaneous production of both NO and ROI can lead to a variety of antimicrobial compounds with distinct properties (Fang, 1997). Indeed, the ability of macrophages to control intracellular replication of €3. abortus depends on both the induction of iNOS and the stimulation of the respiratory burst (Jiang et nl., 1993). Moreover, NO mice deficient in the gp9l protein of reduced NADPH oxidase showed an increased susceptibility to listerid infection (Dinauer et nl., 1997). This was most profound at day 2 postinfection, correlating with the height of neutrophil involvement in the liver. These experiments point to a role for this pathway in early resistance to listeriae, in contrast to the later role of macrophages whose antilisterial activities rely more on NO production. Similarly, although the capacity of macrophages to limit the intracellular growth of mycobacteria is associated with iNOS induction, there is evidence of M . aviuni infections in patients with chronic granulomatous disease (CGD) (Ohga et al., 1997). These patients have a deficiency in the NADPH-oxidase pathway and consequently their neutrophils and monocytes fail to generate oxygen radicals, suggesting that the respiratory burst is required for effective bacterial control. In apparent contrast to this, however, monocytes isolated from CGD patients show no difference in their ability to control M . bovis BCG replication in vitro as compared with those isolated from healthy volunteers (Fazal, 1997). This may reflect either a difference between inycobacterial species as noted earlier for susceptibilityto NO or may again underline the differences between in vitro observations and the more complex interactions of the in vivo situation. Intracellular bacteria have necessarily evolved mechanisms with which to counteract the antimicrobial capacities of macrophages, to enable them to inaintain their intracellular habitat and evade lalling. One of the initial ways in which they do this involves the choice of receptor(s) for entry into the host cell. Thus, entry into the cell via CR fails to induce the generation of ROI, making it a less hazardous route of entry for the pathogen (reviewed in Mosser, 1994). For example, LPS variants of F. tulnrensis are altered in the way that they enter macrophages, and subsequently NO production
and intracehlar growth are also modified (Cowley ct a/., 1996).Add~tionally, soiiie bacteria can proclucc molecules to counteract ROI effects. These include acid phosphates of F. trilnrmrir (Reilly et (11 , 1996) and a periplasinic superoxide disniutasc produced b y salinonellae that can protect this microorganism from the effects of 1x)tIi ROI and NO (De Groote c>t nl., 1997). (>ellwall components, including LAM and phenolic glycolipitl I isolated fromM. leprnu, have becn demonstrated to liave a dowiregulato1y effect on phagocytosis and tlie induction of tlie respiratory burst in murine macrophages (Chan ct d . ,1989; Moiira ct a/., 1997). Manv of these defense mechanisms are not coiistituti] ely expressed in hacteria but are rather induced once tlie microbes are resident inside tlie host cell, via seriso~y regulators, such as soxS and oxyR of salmonellae, which detect concentrations of the toxic molecules. r.hjA, a transcriptional regulator of S. t y p h i i w r i i ~ t n ,has been described, &ich is required for resistance to oxidative stress and is expressed in the intl-acelliilar einiroiiment of the inacropliage (Buclimeier et n l , 1997). Several of the strategies used to counteract tlic macrophage production of ROI also overlap in their effects on NO. The production of low molecular weight thiols, wliicli act as scavenging molecules, are a prime exaniple of this and liave been used to great effect by salinonellae and other enteric pathogens (De Groote et nl., 1996).
2. lritrncelliilnr Iron .
. ~ n Haic d to Get I t
Intracellular bacteria have an obligate requirement for iron to maintain their growtli within cells. In this way a competitive environinent is established between host cell and pathogen for the intracellular iron pool. The mammalian cell has an arra) of specific molecules available to enhance their iron supply, and limiting access of this to the intracellular pathogen is an effective measure for controlling their growth. Extracelliilar iron is bound to transferrin (Tf) and lactoferriii and can be internalized by the host cell via transferrin receptors (TfR), wliicli are trafficked to an early endosomal recycling compartment that inaintains B mild acidic pH fkilitating tlie release of tlie iron from tlie receptor. Within the cell, the majority of available iron is bound to ferritin, whicli acts as ;in iron storage molecule (Harrison and Arosio, 1996). Experiments using L. prieutwphiln revealed that activated macrophages can downregulate tlie expression of TfR and intracellular ferritin, thus limiting iron availability within tlie pliagosome and resulting in death of bacteria that lack efficientiron uptake mechanisms (Byrd and Honvitz, 1989; Gebraii c>t( I / . , 1995). However, because iron is required as a cofactor in other antimicrobial mechanisms. siicli as the generation of KOI, host cells tnust enact a difficult balance to generate sufficient intracelliilar iron to support these effector mechanisms but limit excess production so as not to favor growth of the bacteria.
In light of this, successful intracellular pathogens have been forced to evolve ways to successfully scavenge intracellular iron from the host cell. Perhaps the simplest way is to modify the intracellular compartment in which the bacteria survive, as demonstrated to great effect by M . tuberctclo,sis and M . n2jizm, which arrest maturation of their phagosoines to an early endosoinal stage and thus have access to TfR and bound transferrin (Clemens and Honvitz, 1996; Russell ct nl., 1997). These bacteria have also evolved a number of iron-binding proteins (siderophore;) named exochelins that can remove iron from host proteins and donate them to specialized mycobactin molecules in the bacterial cell wall (Gobin et nl., 1995; Gobin and Honvitz, 1996; Wong et ul., 1996). The production of siderophores has also been described for B. nbortiis, which protect the brucellae from intracellular killing (Leonard et nl., 1997). As in the case of avoidance of toxic effector molecules, many of these iron-scavenging properties are not constitutive but are regulated once in the intracellular en\ironnient. Limitation of intracellular iron diminishes the infectivity of C. trnclzoirlatis in an in vitro system and there is evidence for iron-regulated chlamydia1 proteins, at least one of which is iron inducible (Raulston, 1997). This is similar to the iron-regulatedfrgA gene of Legionella spp., which is speculated to encode a siderophore required for optimal intracellular replication (Hickey and Cianciotto, 1997).A novel ferric reductase has been isolated from M . paratiiberculosis that can scavenge iron from ferritin and transferrin and that has been localized within macrophages in infected bovine tissue (Honiuth ct nl., 1998).
3. Antiinicrobial Peptide.y Much attention has been focused on the role of rnicrobicidal peptides as coinmon medhtors of phagocytic and epithelial cell defenses, following on froin the discovery of the antibacterial effects of lysozyme (Boman, 1996). These compounds include clefensins, bacterial permeabilityinducing protein (BPI), phospholipase A2, and cathelicidins (Ganz and Weiss, 1997). Most of these molecules are found within neutrophils, with the defensins in particular stored within the granules. Following ingestion of salmonellae by neutrophils, these peptides were found in abundance within the phagocytic vacuole (Joiner et al., 1989). They appear to act by pernieabilizing the bacterial membrane and have an inhibitory effect against awidevarietyofpathogeiisinvitro (Porteretal., 1997).The actionsofBPI are niainly manifest against gram-negative bacteria as it displays a high affinityfor LPS and can inhibit endotoxin-mediated signaling. This protein has been detected both in neutrophil precursors and on the surface of inonocytes (Ganz and Weiss, 1997).Although the role of defensins and similar molecules in the control of intracellular bacteria remains unclear, the most likely sce-
nario is that these molecules exert their effects inimediately at tlie port of entry of the invading ptliogens, e.g., at mucosal surfaces such as lung and gut epithelia. In addition, there are considerable differences between liumaris and mice in the repertoire and distiibution of defensins, making it difficult to clraw decisive conclusions for their role in human disease states. V. Phagosome Maturation and Microbial Detours
A. TIIEKEC:ULARM7.4~
Following particle uptake by a professional phagocyte, the newly formed phagosome proceeds through numerous steps of maturation accompanied by continuous reinodeling of its protein composition (Mellman, 1992; Pitt et al., 1992; Kabinowitz et al., 1992; Desjardins ct nl., 1994a,b; Huber and 1995; Gruenbcrg and Mx~field,1995; Griffiths, Peters, 1994; Beron c’t d., 1996; Mellnian, 1996) (Fig. 1).After closure of the phagosome, it quickly loses plasma membrane molecules througli recycling to the cell surfkce (Clemens and Honvitz, 1992). Early phagosomes represent only a brief transient stage during maturation and arc characterized by an alinost neutral pH (pH 6 - 6 5 ) and markers of tlie early/recycling endosome such a s the TfK and its ligand, the iron transpoiter Tf, the M M R , arid a GTPbinding protein, Rab-5. The TfRfR system represents the classical marker for tIii earIy/recycIing endosome: after binding to its receptor, ironsaturated Tf is taken up into an early endosome. Here, iron is released arid both components are recycled to the cell surface. At this early stage of the rccycling/sorting endosome tlip decision is niade whether the engulfed material is directed to the surface or to lysosomes. The next steps of phagosome maturation involve multiple fusion events prior to fusion with lysosomc~s:(i) maiinose-6-pliospli~~te receptor (MGPR) directed delivery of hydrolases such as acid phosplintase, catliepsins B, L, and D from tlie biosynthetic pathway via transport vesicles from the trms-Golgi network; (ii) acquisition of the vacuolar proton ATPase (v-H’ ATPase); and (iii) acidification of the pliagosoint. to p H 4-5.5 (Mellniaii, 1992; SturgillKoszych et d., 1994; Beron et 01.. 1995: Desjardins, 1995; Gruenberg and Maxfield, 1995). Using model phagosomes containing niagnetic Imicls, it has been shown that delivery of hydrolytic enzymes to the phagosome precedes accuinulation of the v-H+ATPase and acidification (SturgillKoszycki et al., 1996). Moreover, the late phagosome is characterized by its hiill fusigenicity with late endosoines transporting endocytosed inaterial such as fluitl-phase markers (dextral) o r ligands of specific receptors, including low-density lipoprotein (LDL), a?-macroglobulin ( a z M ) , and mannosylated/fucosyl~it~(l niolecules (Wileinan ct nl., 1985; Dunn and Maxfield, 1992; Ghosh ct al.. 1994). Lysobiphosphatidic acid (LBPA), a highly
phospholipase-resistant lipid, also localizes to internal membrane stacks in late endosomal compartments where it plays an important role in sorting 1998). Finally, the late phagosome forins a the MGPR (Kobayishi et d., phagolyrosome by fusing with preexisting lysosomes, thus facilitating further digestion and/or recycling of phagocytosed material. At some point during these events, most probably during the late endosornal or lysosomal stage, the internalized material reaches compartments carrying molecules of the M HC class I1 complex and other antigen-presenting molecules, for further processing and subsequent antigen presentation (see later). Various proteins located on the cytoplasniic side of vesicles are directly or indirectly involved in these processes. Small GTP-binding proteins of the Rab family specifically regulate distinct stages of intracellular vesicle trafficking. For example, Rab-Fj and Rab-4 are present on early and recycling endosomes, respectively,whereas Rab-7 is characteristic for late endosoines (Novick and Zerial, 1997). Pliospholipicl- and calcium-binding proteins of the annexin family are located specifically on intracellular vesicles and have been suggested to facilitate differential membrane fusion events (Gruenberg and Emans, 1993). Finally, a number of proteins have been identified that form an essential part of the polypeptide complexes involved in vesicle docking and/or fusion. These include proteins such as the SNARESof both the vesicle and the target membrane, the N-ethylrnaleiiiiide sensitive factor (NSF),and the soluble NSF attachment proteins (SNAP)(Rothman, 1994), which have been detected in macrophage phagosomes ( Hackam et nl., 1996). Moreover, both phagosome formation and intracellular vesicle transport depend on cytoskeletalelements and their rearrangement (Blockeret al., 1996, 1997; reviewed in Haas, 1998).
B. INTRACEI.I,IJLAH COMPAHTMENTS THATSUPPOHT BACTERIAL SUHVIVAL As soon as an intracellular bacterium enters its host cell, the newly formed phagosome usually proceeds through the steps of maturation as described earlier. Intracellular pathogens, however, have developed several strategies to manipulate phagosome maturation to avoid exposure to the antibacterial activities of host macropliages, such as toxic metabolites of the ROI, NO, defensins, and other inicrobicidal peptides, low pH, and lysosornal enzymes (Nathan and Hibbs, 1991; Haas and Goebel, 1992; 1995; O’Brien et nl., 1996). Various microbes exploit distinct Chan et d., intracellular niches for survival and proliferation (Sinai and Joiner, 1997; Russell, 1998), which can influence antigen availability for processing and presentation and consequently the type of immune response elicited. In general, intracellular pathogens follow four main strategies: (i) inhibition of fusion events between the phagosome and endosome, (ii) blocking of
pliagosome niaturation, (iii) adaptation of the pathogen to promote sumval in a late endosoinavlysosomal compartment, and (iv) escape froin the pliagosome into tlie cytoplasm (Fig. 1). 1 Fiision Iuhiliition A5 detailed in Section 11, one wrovar of L. pnezimophiln can be engulfed by coiling pliagocytosis, which subsequently influences tlie fate of tlie phagosome (Clemens and Horwitz, 1992). Although plasma membranederived MHC class I and I1 molecules are excluded rapidly froin the coils, they retain CR3 and Fj'-iiiicleotidase (Clemens and Horwitz, 1992). The mature L pricrimop,hiln pliagosoine is devoid of the lysosome-associated membrane protein-1 (LAMP-l),catliepsin D, and Rab-7, but is found in close proximity initially with mitochondria, later with ribosomes, and is surrounded by rough endoplasinic reticulum (ER) sheets (Swanson and Isberg, 1995, 1996). This has been taken as evidence that L. peunwphilri similar to L e i h i a n i n ) I M ~ X Z C ~ I(see E U later) enters autophagosomes, which are formed from ER sheets and serve as waste and recycling compartments responsible for scavenging, and degradation, of cytoplasmic material arid organelles. However, arguing against this notion, autopliagosomes are usually formed from ribosoine-free ER membranes (Dunn, 1994) and are acidic and highly hydrolytic compartments, whereas the L. pneuirzoplzila phagosome neither acidifies nor contains catliepsin D (Honvitz, 1983; Swanson and Isberg, 1996).Similarly, the intracellular parasite T. gondii actively invades its host cells and forms a vacuole that does not fuse with other intracellular compartnieiits (Sibley et nl , 1985; Sinai and Joiner, 1997). Ijowever, Ig-opsonized T gontlii are passively engulfed via F ~ R and subsequently end up in phagolysosomes (Mordue and Sibley, 1997). The obligate intracellular bacteria, C. pittaci and C. tmclion~zti.s,reside and replicate in a nonacidic compartment, the inclusion body, which is excluded from the classical pliagosome maturation pathway but forins tight associations with mitochondria, most proliably as a source of ATP (reLiewed in Sinai and Joiner, 1997). Independent from the nature of the host cell, the inclusion body neither accumulates extracellularly added tracers (e.g., horseradish peroxidase, dextran) nor does it contain markers for early or late endosomes (such as Tf, LAMP-1, cathepsin D. v-H'ATPase) (Heinzen et al., 1996, Tdraska ct nl , 1996; Sinai arid Joiner, 1997; van Ooij et d . , 1997). However, a study using a human macrophage cell line showed that a significant percentage of C. psittrrci phagosomes acidify and become positive for LAMP-1 and even MHC class I1 molecules (Ojcius et rd., 1997). Nevertheless, chlamydia1 vacuoles are considered sele&ively fusogenic rather tlian nonfusogenic because of their tendency to fuse with each other (hoinotypic fusion). This has been associated with tlie presence
286
ULRlCll E SC:IlAIBI.E
l’t
01
of annexin I1 on the vacuolar ineinbrane (Majeed et d.,1994). In addition, they show differential fusogenicity with Golgi-derived exocytic vesicles carrying sphingolipids but not with those containing glycoproteins as cargo (Hackstadt et al., 1995, 1996; Scidmore et nl., 1996). Although little is known about how chlamydiae influence differential fusogenicity, phagosomes containing heat-lulled chlamydiae, but not those carrying outer membrane ghosts or UV-killed bacteria, fuse with lysosomes. This indicates a role of heat-labile chlamydia1 surface structures in avoiding phagolysosome fusion (reviewed by Sinai and Joiner, 1997).
2. Blocking Plzngosonlc! Maturation In general, mycobacteria reside in tight vacuoles in cultured niurine BMM ( X u et al., 1994) and ex- vivo alveolar macrophages (Moreira et d , 1997), which resist maturation toward the late eiidosomal/lysosoiiial stage (Armstrong and D’Arcy Hart, 1971; Frehel et al., 1986; Clemens and Horwitz, 1995). These vacuoles represent early endosoinal compartments that do not acidify (pH 6.5; Crowle et d., 1991; Sturgill-Koszycki et nl., 1994) due to the paucity of the v-H’ATPase (Sturgill-Koszyckiet nl., 1994; Oh and Straubinger, 1996). Nevertheless, the inycobacterial phagosome is a highly dynamic conipartment that rapidly exchanges molecules with the plasma membrane as evidenced by the following findings: (i) M . tuberculosis phagosomes contain MHC class I and I1 molecules as well as the TfR potentially derived from the plasma membrane (Cleinens and Honvitz, 1995); (ii) M . tuberculosis and M . nviuin phagosomes are readily accessible to plasma membrane-derived GM 1 gangliosides (Russell et nl., 1996) and material bound to cell surface N-acetylglucosamineresidues (de Chastellier et nl., 1995); (iii) M . tuberculosis and M . nuium phagosomes trap exogenously added Tf (Clemens and Honvitz, 1996; Sturgill-Koszyckiet al., 1996; Schaible ct nl., 1998).Additionally, M . bouis BCG phagosomes contain Rab5, the marker for early endosomes, but not the late endosomal marker, Rab-7, and lack the lysosoinal enzyme. p-hexosaminidase (Deretic et al., 1997; Via et nl., 1997; Hasan ct nl., 1997). Pliagosomes containing M . aviuin (10 not have access to markers bound for lysosomes and delivered through either fluid-phase or receptormediated uptake such as dextran, LDL, or a,M (Xu ct al., 1994; Schaible rt nl., 1998). The mycobacterial phagosome membrane is also devoid of organic anion transporters responsible for the sequestration of sinall organic anions (<0.8 kDa) into late endosomes (Schaible et nl., 1998). In contrast, A4 tuberculosis and M . nviuin phagosoines in BMM contain LAMP-1 and the Iysosomal proteases cathepsins B, L, and D, which are most probably derived from the biosynthetic pathway. However, MGPR, which are responsible for the delivery of these enzymes, have not been detected in the
INTH2\(:EI.J,lJLAR
BA(:TEHIA 4NL) THE IMMIJNE SYSTEM
287
mycobacterial phagosome (Xu et d., 1994; Sturgill-Koszycki et a/., 1996). Cathepsins B and L are present in their proteolytically active forins in M . aviuin phagosomes, but processing of cathepsin D to its two mature low molecular weight forms does not proceed under the almost neutral pH conditions (Sturgill-Koszyckiet al., 1996).As detailed later, this finding has important implications for the generation of antigenic peptides from mycobacterial proteins (Geluk et nl., 1997). It has been demonstrated that another mycobacterial species, M . jncirinu 111, also resides in early endosomal coinpartinents of host macrophages (Barker et nl., 1997). Current biew holds that phagosomes of invcobacteria are locked in the stage of the early recycling/sorting endosome that delivers iron-loaded Tf into the phagosome. As iron is limited in the intracellular environment, Tf-derived iron appears to be an important survival factor for mycobacteria inside the host cell. In f k t , when Tf loaded with radioactive iron ("'Fe) was added to infected macrophages, autoradiography indicated that bacilli were able to access the iron (D.G. Russell, personal communication). The question as to how niycobacteria inhibit phagosome maturation cannot be answered as yet. Ammonia, which is produced in vast amounts by mvcobacteria in culture (up to 10 m M ) , has been suggested to buffer the phagosomal content (Gordon ct nl., 1980), which is consistent with the Fact that fusogenicity of vesicles can he influenced by their pH (Clague et al., 1994).Two enzymes secreted by M. ttihc~crilosisinto the phagosomal lumen, urease and glutamine synthetase, most probably contribute to ammonia and polyaniine production in the host cell (Hart11 et nl., 1994; Cleinens et al., 1995). However, preliminary results using an ureasedeficient M . bmis BCG strain (Reyrat et al., 1995) reveal only a minor influence of urease on intracellular survival of BCG (Reyrat et d.,1996). In addition, it has been shown that a glycolipid released by mycobacteria, cord factor (an a trehalose 6,G'-dimycolate),inhibits fusion between phospholipid vesicles ( Fujiwara, 1997). Another possible explanation as to why mycobacteria inhibit phagosome maturation relates to their size and the physical properties of their cell wall, as phagosomes containing beads sinaller than 1 pin and with a hydrophobic surface are also limited in their maturation (de Chastellier and Thilo, 1997). Despite these observations, the relevance of inhibiting phagosoine inaturation to mycobacterial virulence is debatable as it also occurs, at least to some extent, in the attenuated strain M . bocis BCG (Deretic et nl., 1997; Hasan et d . ,1997). However, living in the environment of an early endosoma1 compartment seems important for inycobacterial survival. This is strongly supported by data showing that the M . aoium phagosome matures toward a late endosomal compartment in IFNy-activated macrophages. Under these conditions the phagosomes reach a pH of 5-5.5 due to the
accumulation of the v-H+ATPase. Moreover, they are excluded from Tf/ iron supply and gain access to lysosome-bound tracers such as cu2M (Schaible et al., 1998). In addition, the number of mycobacteria per phagosome is increased from about one in resting to more than five per pttagosorne in activated inacrophages and the phagosornes become more spacious. These alterations are followed by a drop in mycobacterial viability (Scliaible et d . , 1998). It has also been shown that M . bovis BCG containing phagoseines gain access to lysosomal tracers in macrophages activated with IFNy and LPS (Via et nl., 1998). Similar to mycobacteria, phagosomes containing E. chafleensis, a member of the Rickettsiaceae, accuinulate Tf but have limited access to the v-H+ATPaseand acidify only slightly (Barnewall et nl., 1997).Data indicate that phagosomes containing B. abortus also display restricted fusigenicity with lysosomes in inacrophages (Caron et nl., 1994). In nonphagocytic cells they are somehow associated with ER membranes. Although not studied in detail, B. nbortus-derived adenine- and guanine monophosphates apparently block the formation of phagolysosomes (Canning et al., 1986). For A. felis, a filterable factor present in the culture supernatant has been described that inhibits phagolysosomal fusion (Brouqui and Raoult, 1993). 3. Survival in the Phngolysosonle
The bacterial pathogens discussed so far have developed strategies to avoid the acidic and hydrolytic environment in late endosomaVlysosoma1 compartments. However, another group of intracellular pathogens survive and even thrive under these harsh conditions. Even more importantly, the phagolysosomal milieu may provide a vital source of nutrients for these organisms. Unlike its relatives of the Chlamydia genus, Coxiella burnettii resides in a spacious inacrophage phagosome, showing a late endosomaVlysosomal character, i.e., acidification (pH 5.2), sequestration of fluid-phase markers, and access to LAMP-1 and LAMP-2, cathepsin D, acid phosphatase, and v-HtATPase (Heinzen et nl., 1996; Mege et nl., 1997). Interestingly, in addition to homotypic fusion, C. buinettii phagosomes fuse readily with other phagosomes containing yeast particles, latex beads, or the parasite L. nzexiccina (Veras et al., 1994, 1995). The L. niexicana vacuole is also a spacious late endosomalllysosoinal compartment as determined by the following features: sequestration of fluid phase and receptor-delivered tracers, low pH, and presence of LAMP-1, cathepsins B, L, D, MGPR, v-HtATPase, MHC class 11, and LBPA (Prina et al., 1990; Antoine et al., 1991; Russell et nl., 1992; Lang et al., 1994; Sturgll-Koszycki et al., 1994; Schaible et nl., 1998). Hence, fusion of L. rrwxicnna vacuoles with C. burnettii vacuoles indicates that phagosomes trapped in a similar matura-
tion step can fuse with each other. AdditioIially, pliagosoines containing fuse with L. ttrexicatzn vacuoles, an observation live L. nioiioc!jtogene.s.~~cne,s that correlates with tlie lack of annexin I that is present in nonfiisigenic phagosomes carrying (lead listeiiae (Collins et nl., 1997). The vacuole of L. mesicma also gains access to cytoplasmic material via two iiidependent pathways: (i) organic anion transporters in tlie vacuolar membrane and (ii) fusion mitli late autop1iagosoint.s (Schaible et nl., submitted). Whether this is a special feature of the L. ttic..\-ictrnnvacuole or a general characteristic of late enclosoiiiaVlysosoinal compartinents needs further studies. Similar to coxiellae and leishmaniae, F. trilarcvrsis also resists the harsh acidic and hydrolytic conditions of phagolysosomes ( Fortier ef 01.. 1995). Published data on salmonellae-containing phagosomes provide a rather diverse picture, most probably Iiased on the serovars and host cell types studied. It is thus difficult to place sahiionellae in this group. Nevertheless, salmonellae promote phagosome divergence from the degradative pathway but permit phagosome ii&iration toward a later endosomd stage. In ii study using BM M, pliagosoines containing live S. t y p h z r c r i r m developed 1995),which readily matured into spacious vacuoles ( Alpuche-Arxida et d., to late endosoinaVlysosomal vacuoles positive for LAMP-1, cathepsin L, and the fluid-phase tracer, dextran, without significant loss in bacterial 1996). B M M from resistant mice fail to support the viability (Oh et d., generation of spacious vacuoles and consequently survival of bacteria (Alpuche-Aranda et nl., 1995). In contrast, an EM study using B M M from resistant and susceptible mice, a s well as the miiiine macrophage cell line J774, revealed inhibition of phagolysosome fusion (Buchnieier and Heffron, 1991). Similarly, in the murine inacrophage cell line RAW264.7, S. typhimtiriim phagosomes have a mildly acidlc pH, contain LAMP- 1 and acid phosphatase, Init are devoid of cathepsin L arid tlie MGPR. Neither Tf nor dextran as tracers for early and late endosomes/lysosoines, respectively, were delivered to the salmonella vaciiole ill significant amounts (Rathman ct nl., 1996, 1997). I n contrast, salmonellae entering nonphagocytic epithelial cells replicate in lysosoines connected to long filamentous structures containing vast amounts of LAMP-1 and acid phosphatase but not M6PR and cathepsin D (Garcia-del Portillo and Finlay, 1995). Tliese unusual lysosomal structures arc tliought to be induced by the bacteria themselves (reviewed in Garcia-del Portillo, 1996). The .sfA gene responsible for lysosoiiial filament induction by S. tr/plrittrrtririniin epithelial cells has been cloned (Stein et d . ,1996). Of those. bacteria residing in acidic inacrophage phagosomes, S. typliittiuritini is tlie oiily one for which there is at least some knowledge about tlie molecular adaption for intracellular survival (reviewed in Finlay and Falkow, 1997; Guiney, 1997).The two-component signal transducer, PhoPQ, controls the expression of salmonella virulence
290
LJLRIUI E SCHAIBLE r f nl
genes in response to phagocytosis (Groisman et al., 1989: Miller et al., 1989; Alpuche Aranda et al., 1992). Triggered by MgZf,PhoQ phosphorylates PhoP, which regulates the expression of “intracellular survival” genes (Garcia Vescovi et al., 1996). 4 . Escape into Cytoplasm A few intracellular pathogens have chosen a simple but rigorous way to avoid the degradative pathway by escaping from the phagosome into the cytoplasm (Fig. 1). In addition to avoiding the phagolysosomal conditions and associated defense mechanisms, this strategy provides the respective pathogens with a nutrient-rich environment (Moulder, 1985). The cytoplasmic localization also directs the immune response in favor of MHC class I-restricted cytotoxic T lymphocytes (CTL).Listeria monocytogenes is probably the most intensively studied organism of this group. This grampositive rod expresses several virulence factors, including a pore-forming sulfohydril-activatedhemolysin (listeriolysin, LLO) that promotes rupture of the phagosomal membrane and escape into the cytoplasm (reviewed in Cossart, 1997).A lecithinase and two different phospholipase C molecules probably act in concert with LLO for both escape from the phagosome and spreading to other cells. All three virulence factors are positively regulated by the p$A gene product (Cossart, 1997). Shortly after uptake, listeriae traffic to a compartment that shows characteristics of early endosomes (TfR, MMR) and contains key factors for early fusion events (Rab5, NSF, dp-SNAP), but does not contain markers for late endosomes (cathepsin D, LAMP-1, annexin I ) (Alvarez-Dominguezet al., 1996, 1997; Collins et al., 1997).Listeriolysin-dependent perforation of the phagosomal membrane is seen following acidification of the phagosome to a pH of around 5 (Beauregard et al., 1997). After escape into the cytoplasm, the L. mnocytogenes surface protein ActA induces polar actin polymerization and release by recruiting an actin-binding polypeptide of the host cell (Domann et al., 1992; Kocks et al., 1992; Welch et al., 1997). Both ActA and a metalloprotease (Domann et al., 1992), which may also contribute to intracellular survival, are again regulated by the prfA gene product (Cossart, 1997; Ireton and Cossart, 1997). In this way, a polymerized actin tail forms that pushes the bacteria through the cell. According to early EM
FIG.1. A schematic representation of the maturation of phagosomes within rnacrophages and the bacteria that influence this process. Also depicted are routes by which pathogenderived antigens gain access to various antigen-processing pathways for the eventual stimulation of distinct T-cell populations.
Phagolysosorne
Cathepsin D (processed) Other acid hydrolases MGPR flab-7 Fluid-phase markers
Salmonella Coxiella
292
U L R I C I I E SCFIAIHLE
r’f
d
analyses, the host cell plasma menibrune seeins to be forced to protrude into a neighboring cell that subsequently engulfs the listeriae-containing protrusion, resulting in the formation of a double-membraned vesicle. Phospholipases may play a particular role in lysing both the “old” and the “new” host cell membranes (reviewed in Cossart, 1997; Ireton and Cossart, 1997). More recent studies employing video microscopy rather indicate direct fusion between the two plasma membranes and thus introduction of listeriae into the cytoplasm of the new host cell ( J . Wehland, personal communication). Similarly, shigellae and rickettsiae, which mainly persist in epithelial cells, use their own laeinolysins and/or phospholipases to strip themselves from the surrounding phagosoinal ineiaabrane (Maurelli et nl., 198s; Silverinan et nl., 1992; Ojcius et al., 199s; Barzu et al., 1997).Apparently, the harsh lysosoinal conditions are not detrimental for 7’.cruzi, another cytoplasmic pathogen, as escape of this parasite into the host cell cytoplasm is preceded by fusion of its pliagosoine with lysosomes (Andrews et nz., 1990). VI. Antigen Processing and Presentation Pathways
A. CELLS,ANTIGENS,A N D ANTIGEN-PRESENTING MOLECUI,ES The classic antigen-presenting molecules encompass MHC class I and 11 molecules. These polymorphic molecules are encoded by the human leukocyte antigens (HLA) -A, -B, and -C and HLA-DR, -DP, and -DQ loci, respectively, in humans or the H2-K, -D, and - L and H-21-A and -E loci in mice. Classical MHC class I molecules are also termed MHC class Ia molecules to distinguish them from class I-like molecules (MHC class Ib), which are encoded by genes adjacent to the classical MHC class I loci (Fischer Lindahl, 1997). As discussed later, MHC class Ib molecules relevant for bacterial infections comprise H2-M3 and Qa-2 molecules. Moreover, the non-MHC-encoded molecules of the CD1 family can present mycobacterial antigens to certain T-cell populations. MHC class Ia or I1 molecules present antigenic peptides to conventional CD8+ or CD4’ T cells, respectively. Typical MHC class Ia-restricted antigens are endogenous proteins, either host cell or pathogen derived, which are present in the cytosol. Typical MHC class I1 antigens are of exogenous origin taken up by appropriate APC. MHC class I1 molecules are restricted to professional APC such as B cells, dendritic cells, and macrophages, which can specifically take up exogenous antigens and process them for MHC class I1 presentation in endosomal cornparttnents. In contrast, MHC class Ia molecules are expressed on virtually all nucleated cells in the body, which consequently can present antigens in the context of MHC class la. These cells are termed nonprofessional APC. Pathogens
residing mostly in nonprofessional APC such as rickettsiae and shigellae either deliver antigens into the MHC class I a pathway or live unrecognized by the immune svstem. MHC class Ia presentation facilitates recognition of these bacteria as a corollary of their strategy to escape from the phagosoine. Similarly, the imniune response against L. i~iotic,ctitajicncs,which can reside in the cytoplasm of hepatocytes (Szalay cf d . , 199s; Rogers cf nl., 1996), is dominated by MHC class Ia-restricted CD8' T cells (DeLibero and Kaiifinaiin, 1986; H a m and Bevan, 1996). Antigens deiived froin pathogens engulfed by and residing in macrophages or dendritic cells readily enter the MHC class I1 pathway. Antigenspecific B cells are also able to present antigens using MHC class I1 molecules, although these cells eiidocytost~antigens predominantly via specific surface Ig (Drake c f cd , 1997; Watts, 1997). Thus, a notable role of B cells as APC in infections with intracellular bacteria remains questionable, leaving macrophages and dendritic cells a s the central APC for CD4' T-cell stiinulation. Resting macrophages do not express M HC class I1 molecules in appreciable amounts and depend on activation via IFNy or TNFa to become fully competent. Dendritic cells, however, are derived froin phagocytic progenitors that scavenge for foreign antigens in tissues, such as the Langerhans' cells in the s h n . Upon antigen contact, these cells rapidly lose their phagocytic capacity, upregulate surface expression of MHC class I1 and costimulatory iiiolecules, and migrate to the draining lymph nodes where they represent the doininant APC for antigenspecific T-cell priming. Moreover, immature dendritic cells sample vast amounts of antigens in soluble form via niacropinocytosis, i.e., fluid-phase material is captured in large membrane ruffles and endocytosed (Sallusto cf (11 , 1995; Steinman and Swansoii, 1995). Inflaininatory stimuli such a s LPS and TNFa further enhance the potency of clendritic cells to present antigens by increasing the half-life of meinl)rane-associated MHC class I1 molecules-by two orders of magnitiide (Cella ct d., 1997).Although little is known about the ability of dendritic cells to host intracellular pathogens, they can be infected by M ftibwciiloiis, M . hocis BCG, C. trnclionintir, and C. psittnci, as well as I)y IeisIiinania parasites, at least for a sliort tiine (Blank ct nl., 1993; Pancholi ct nl., 199311; Moll ef d., 1995; Ojcius et id., 1998). I n addition to MHC class I1 molecules, human dendritic cells express the unconventional antigc.ii-presentiiig molecules CD1 a, b, and c at high density (Porcelli, 1995; Maher and Kronenberg, 1997). Apart from phdgocytosis ofwhole liacteria, uptake of soluble antigens can occur either without receptor usage by inacropiiiocytosis or by employing Lrarious receptors (see Section 11). These mechanisms represent important pathways for liystander APC to collect antigens secreteellreleased froin bacteria or from infected or lysed cells. Macrophages and dendritic cells
294
ULHICH E SC:IIAIDLE rt n/
use members of the inannose receptor family such as the M M R and DEC205 (expressed on dendritic cells only) for the uptake of inannosylated proteins or glycolipids (Jiang et al., 1995; Sahsto et al., 1995; Engering et al., 1997; Tan et al., 1997).The MMlUligand interaction is labile under mild acidic conditions, facilitating the release of antigens in the early endosome (Diazet al., 1989).This allows several rounds ofantigen uptake, delivery to endosomes, and recycling of the receptors to the cell membrane (Ezekowitz et al., 1990).A similar scenario has been described for other scavenger receptors such as the LDL receptor, which may also contribute to the samp h g of lipid-containing antigens by APC (Goldstein et al., 1985). MSRAI, for example, can bind LTA (Dunne et al., 1994). CD14 and collectins, multimeric proteins that can bind inicrobial sugar residues (e.g.,of LPS and LAM) through peripheral lectin domains, inay also be involved in antigen sampling (Wright, 1995; Epstein et al., 1996; Lu, 1997).Although the relevance of specific antibodies in infections with intracellular bacteria is still in debate and decisive experimental data are not available, the presence of specific antibodies inay facilitate antigen capture via FcR. Like other cellular molecules, receptors involved in antigen uptake are also influenced by the activation stage ofthe cell. Several cytokines influence the up- or downregulation of receptors: (i) M M R and SPR are downregulated by IFNy, (ii) CD14 is downregulated during IL-4GM-CSF-mediated maturation of dendritic cells, and (iii)CR4, MSR, andSPR210areupregulatedandCR3is downregulated during macrophage differentiation (Sastre et al., 1986; Brown, 1991; Chroneos and Shepherd, 1995; reviewed in Ernst, 1998). Protein antigens derived from an intracellular bacterium to be presented by MHC class Ia or I1 molecules must be targeted into the appropriate antigen-processing machinery of the host cell. Secreted proteins or those shed from the bacterial cell wall therefore are superior antigens to those associated with the bacterial cytoplasm and are usually recognized by the immune system at an earlier stage of infection. Proteins localized in the cytoplasm or tightly bound to the cell wall or cell membrane become available only once the bacteria are killed and degraded. The importance of this antigen hierarchy has been evaluated in vaccination experiments using recombinant bacteria such as recombinant S. typhimuriuin and L. mnocytogenes expressing defined antigens in secreted or somatic form (Hess et al., 1996a; Shen et nl., 1998). Proteins expressed exclusively or in vast abundance during the intracellular phase are of special interest for two reasons: (i) these proteins not only represent possible virulence factors that facilitate intracellular survival and phagosome modulation, but (ii) may also be processed and presented preferentially to T cells by the APC. Given that these proteins are secreted, they represent prime antigen candidates for vaccines. Several approaches have led to the identification
IKlR.\CELI.ULAR
BACTERIA AND THE I M M U N E SYSTEM
295
of such proteins. Genes or protein species that are expressed specifically within host cells have been identified by subtractive mRNA analysis, twodimensional gel electrophoresis (2DE), or other methods. The following bacteria have been analyzed: M . tuberculosis (Lee and Honvitz, 1995), M . nviicni (Plum and Clark-Curtiss, 1994; Sturgill-Koszycki et nl., 1997), B. abortus (Lin and Ficht, 1995; Rafie-Kolpin et nl., 1996),L. pneunwplzila (Abu Kwaiket nl., 1993;Abu Kwaik, 1988),and S. tyyhimniArium (Buchmeier and Heffron, 1990; Alpuche Aranda et nl., 1992; Burns-Keliher et al., 1998). Proteins expressed within host cells are partly stress induced and some of them have been identified as heat shock proteins (HSP) such as HSP60. Twenty-four proteins of B. nhortzts, 157 proteins of S. typhinruriimi, and 6 proteins of M . tuberculo.sis have been found to be expressed exclusively by bacteria growing within inacrophages. Studies using 2DE revealed at least 6 proteins that are upregulated or newly expressed in M . bouis BCC, grown within macrophages (U.E.Schaible et nl., in preparation).
B. NOT ENOUGH TROUHI.E: INTRAC:ELLIJI,AR BACTERIAINFLUENCE ANTICENPRESENTATION It is an obvious dilemma for both macrophages arid dendritic cells to serve as primary host cells for intracellular pathogens and, at the same time, to be responsible for the inductiodinaintenance of specific T-cell responses. In fact, in numerous studies the capacity of infected APC to present antigens has been evaluated with varying results. Human and murine macrophages infected with inycobacteria are restricted in their capacity to present unrelated antigens (Gercken ct nl., 1994;VdnHeyningen et al., 1997).This effect is most probably regulated in an autocrine manner through IL-6, which is induced in macrophages upon infection with M . bouis BCG or M . auiiini (VanHeyningen ct nl., 1997). In a different system, IFNy-indiiced surface expression of M HC class I1 molecules in inacrophages was found to be reduced upon infection with M . Zeprne or treatment with its cell wall component, LAM (Sibleyand Krahenbuhl, 1987, 1988; Mshana et al., 1988; Sibley et al., 1988).Similarly, CD1 inolecules on huinan dendritic cells are downregulated upon infection with M. tuberculosis (Stenger et nl., 1998). Although the underlying mechanisms are not clear, following phagocytosis of various bacteria such as E. coli, P.seuclo?nonnsaeruginosci, and salmonellae, human macrophages are hindered in their capacity to present antigens (Pryjma et nl., 1994). Moreover, mycobacteria residing in human macrophages remain undetectable by specific CD4+ T cells (Pancholi et d ,1993a). Howcver, this may be a result of the high infection dose used in this study, as experiments using lower multiplicities of infection (MOI ), i.e., I : 1or 1 : 5 , revealed no defect in specific antigen recognition (Conradt
296
ULHICH E. SCXAIDLE et 01
et al., submitted). A similar downregulation of antigen processing has been described for macrophages infected with Leishmania parasites (Fruth et al., 1993; Prina et al., 1996).The exploitation of the autophagosomal system of the host cell by L. nwxicana as a source for nutrients might direct large quantities of endogenous host antigens into the MHC class I1 processing pathway, therefore diluting out parasite-derived and other exogenous antigens (Schaible et al., submitted). c . MHC CLASSIa-RESTRICTED ANTIGENPROCESSING:T H E DIRECT WAYA N D THE DETOUR MHC class Ia molecules are membrane-bound molecules consisting of three a chains noncovaIently linked to &-microglobulin (Bern).The 20s proteolytic complex in the cytoplasm, the proteasome, consisting of 28 subunits that form a cylinder, is responsible for the generation of MHC class Ia-restricted antigenic peptides (Groettrup et nl., 1996). Proteasomes play a pivotal role in the turnover of cytoplasmic as well as ER-derived ubiquitinated proteins (Groettrup et al., 1996). They generate peptides of 5-15 amino acids (aa) in length, which are suitable as precursors for MHC class Ia-presented peptides, which are generally 8 to 9 amino acids long. Three subunits, LMP2, LMP7, and MECL-1, encoded by genes within the MHC are newly synthesized on IFNy activation and replace certain constitutive subunits within the proteasome. This increases the availability and quality of MHC class Ia-binding peptides (Goldberg and Rock, 1992; Groettrup et al., 1996).Peptides generated by proteasomes are translocated via transporter molecules through the ER membrane (TAP1 and TAPS; transporter associated with antigen processing). In the lumen of the ER, octa- or nonameric peptides are bound by nascent MHC class Ia molecules, which physically associate with TAP with the help of tapasin (Ortmann et al., 1997). Several chaperones, such as calnexin, are involved in this process (reviewed in Hansen and Lee, 1997).After binding of &in in a noncovalent fashion, the MHC class Vpeptide complexes are transported to the cell surface. Due to the cytoplasmic localization of L. monocytogenes, MHC class Ia-restricted CD8+ T-cell responses dominate in the protective effector phase (Kaufmann and Ladel, 1994; Ladel et al., 1994). The lytic activity of the listerial hemolysin, LLO, for the phagosomal membrane is essential and sufficient for cytoplasmic residence of listeriae. This cytolysin has been used to target exogenous proteins or antigens from bacteria such as S. typhimurium or M . bovis BCG into the MHC class Ia presentation pathway in a more efficient manner (Gentchev et al., 1995; Hess et al., 1996a, 1988; Darji et al., 1997).
INI'H.A( :El Lll I A H BACTERIA AND TI1E IhlM IJ N E SYSTEM
297
Four immuriodomiiiant peptides in the context of H-2K' from the three major secreted listerial proteins, LLO (aa9l-99), inurein hydrolase pG0 (aa217-225; aa449-457), and a inetalloproteinase (aa84-92), have been identified (Harty and Bevan, 1992; Villanueva ct at., 1995; Sijts et nl., 1996a; Busch et a/., 1997). In a quantitative assay it was estimated that around 10 times more p60 aa449-4.57 than p60 aa217-225 are generated and presented to CTL (Sijts ct nl., 199%). Since both peptides are derived from the s m e protein and show similar affinities for H-2K" binding and TAP transport, this finding indicates that distinct epitopes are processed with different efficiencies. Moreover, the $0 449-457 peptide dissociates from surfhce MHC class Ia niolecules with a half-life of 1 hr and from intracellular ones with a half-life of only 30 inin, whereas p60 aa217-22.5 has a half-life of G hr on plasma membrane MHC class Ia molecules (Sijts and Pamer, 1997). Despite the differential efficiencies of presentation and intracellular stability of these L nu~nocytogenesepitopes, T-cell populations specific for all four epitopes expand similarly, suggesting that these factors do not determine T-cell responses in vim (Busch et nl., 1998). Blocking of' proteasonie activity abolished generation of these peptides, indicating that secreted proteins from cytoplasmic listeriae have to be processed by proteasomes for MHC class Ia presentation (Sijts et a/., 1996a). Using niutagenizecl pG0, it was sliown that the stability of the antigen in the cytoplasm of the host cell is influenced by its N-terminal region and that antigen degradation and epitope generation are linked (Sijts et nl., 1997). Deletions within the N-terminal part of $0 enhanced cytosolic degradation, whereas C-terminal truncation had a stabilizing effect. A reinarkable stabilization resulted from valine substitutions, whereas aspartic substitutions led to rapid degradation. The rate of p6O degradation correlated with the number of pG0 aa217-225 epitopes generated (Sijts et a1 , 1997). In listerial infection, secreted proteins are superior antigens in protection. This has been demonstrated by expressing a viral CTL epitope in recombinant L. inonocytogenes in either a secreted or a cytosolic form (Shen et a/., 1998). Altlioiigh both forms can prime specific CTL, only those T cells primed with the secreted form can protect aganst infection, which has general implications for vaccine design against intracellular bacteria as detailed later. Previous studies using recombinant S. t y p h z u rium expressing L. nwnocgtogerws proteins pG0, LLO, and superoxide dismutase (SOD) had revealed superior efficacy of secreted over somatic proteins in antilisterial protection (Hess et a / , 199Ga, 1997). The intracellular localization of listeriae and rickettsiae readily explains the potent induction of CD8' T-cell responses, but there are numerous reports describing antigen-specific CD8' T-cell stimulation upon infection
298
ULHICH E. SCHAIBLE et ol.
with intraphagosomal pathogens such as S. typhimurium (Pope and Kotlarski, 1994; Pope et al., 1994; Sztein et al., 1995), M . tuberculosis and M . bovis BCG (DeLibero et d., 1988; Flynn et d., 1992; Lade1 et d., 1995b; Bonato et aE., 1998), B. abortus (Oliveira and Splitter, 1995), C. psittaci (Starnbach et al., 1994), and L. amazonensis (Kima et al., 1997). The important role of CD8’ T cells for protection against experimental M . tuberculosis infection has been demonstrated using mice lacking &m and thus MHC class Ia molecules (Flynn et al., 1992). These observations are not at all surprising in the light of data showing that antigens coupled to the surface of latex beads (Harding and Song, 1994; KovacsovicsBankowski and Rock, 1995), antigens associated with cells or cell fragments (Carbone and Bevan, 1990; Debrick et al., 1991), killed bacteria (Szalay et al., 1995), and antigens expressed by recombinant Salmonella strains 1990; Pfeifer et al., 1993; Turner et (Agganval et al., 1990; Flynn et d., al., 1993;Wick et al., 1994; Hess et al., 1996a) all can enter the MHC class Ia pathway. Similarly, L. monocytogeizcs-specific CD8’ T cells recognize macrophages infected with LLO-deficient listeriae unable to escape into the cytoplasm (Szalay et al., 1994). This “alternative” MHC class Ia pathway, which involves at least some intravesicular steps, is mainly utilized for particulate antigens, such as bacteria, and has only a low efficiency for soluble antigens. Therefore, macrophages are the main APC involved in this antigen-presenting pathway with mast cells, dendritic cells, and B cells being added to the list (Bachmann et al., 1996; Ke and Kapp, 1996; Malaviya et al., 1996; Norbury et al., 1997; Reimann and Kaufmann, 1997; Shen et al., 1997; Svensson et al., 1997). Antigens that are macropinocytosed in vast amounts by dendritic cells are apparently also directed into the MHC class Ia pathway (Norbury et al., 1997). The alternative MHC class Ia processing pathway shows similarities to the one for MHC class I1 presentation, including determination by epitope abundance rather than epitope compartmentalization within the bacterium (Wick et al., 1993, 1994).The underlying cellular mechanisms are not quite clear yet and available data are conflicting: Processing of macropinocytosed antigens in dendritic cells is sensitive to proteasome and Golgi transport inhibitors (brefeldin A) and depends on TAP (Norbury et al., 199i). Macrophages from TAP-’- mice present exogenous antigens less efficiently than those from wild-type mice (Wick and Pfeifer, 1996). Interestingly, the residual antigen-presenting capacity was further abolished by the chloroquine-mediated inhibition of phagosome acidification,indicating involvement of both the classical MHC class Ia and a vesicular pathway (Wick and Pfeifer, 1996). In contrast, MHC class Ia presentation of‘ exogenous ovalbumin was not altered by proteasome inhibitors or by brefeldin A, and decreased presentation by TAP-/- macrophages was explained by a reduced surface expression of
l N T H A C E L L U L A R BACTERIA AND T I I E I M M U N E SYSTEM
299
MHC class Ia molecules (Song and Harding, 1996). Denatured proteins, when administered exogenously are also efficiently presented to CD8’ T cells (Martinez-Kinader et al., 1995; Schirmbeck et al., 1995; Jondal et al., 1996), a phenomenon which again is TAP independent (Schirmbeck and Reimann, 1994).It has been suggested that overloading of phagosomes by indigestible material such as latex beads allows antigens to leak into the cytoplasm and therefore into the classical MHC class Ia pathway (Reis e Sousa and Germain, 1995). Consistent with this notion are infection experiments in Pzm KO mice that suggest participation of CD8+ T cells in protection against M . bovis BCG only after high but not low inocula (Lade1 et al., 1995b). The following scenarios, which are not mutually exclusive, are considered by the authors the most likely explanations for MHC class Ia presentation of antigens from “phagosomal” bacteria. i. Similar to MHC class I1 molecules, a subset of MHC class Ia molecules can associate with the invariant chain, which targets them to endosomal compartments where they are loaded with exogenous peptides (Sugita and Brenner, 1995).Thus, antigenic peptides derived froin the phagosome can be introduced to MHC class Ia molecules in this compartment. ii. A certain number of surface MHC class Ia molecules (H2-L’) are devoid of peptides and recycle froin the cell surface to endosomes where they sample peptides from exogenous antigens (Schirmbeck and Reimann, 1996). As M . tuberadosis phagosomes retain MHC class Ia molecules, empty ones could directly bind peptides in the phagosome (Clemens and Honvitz, 1995). iii. Alternatively,endogenous peptides bound to MHC class Ia molecules could be exchanged with bacterial peptides with higher affinity. This could occur either in the endosomal system or on the plasma membrane with peptides regurgitated by infected cells. Regurgitation would also allow sensitization of bystander APC (Jondal et al., 1996). iv. In a number of bacteria, including salmonellae, a specific secretion apparatus could facilitate the translocation of secreted antigens into the host cell cytoplasm (Jones and Falkow, 1996; Finlay and Falkow, 1997). v. Some intracehlar bacteria express cytolysins that may promote the translocation of secreted proteins into the host cell cytoplasm. Presentation in a TAP-dependent fashion of exogenous ovalbumin, which was engulfed together with M . tuberculosis, was taken as evidence that tubercle bacilli express a membranolytic activity, allowing antigens to leak into the cytoplasm (Mazzaccaro et al., 1996). Similarly, recombinant M . hovis BCG expressing LLO apparently remain in the phagosome but promote MHC class Ia presentation of the bystander antigen ovalbumin (Hess et al., 1998).
300
ULHICH E. SCXA1BI.E el nl
vi. Macrophages can engulf antigen-loaded apoptotic cells or blebs thereof and present these antigens in a TAP-dependent fashion to specific CD8’ T cells (Bellone et al., 1997). Moreover, dendritic cells take up antigen from apoptotic cells and induce an antigen-specific MHC class Irestricted CD8’ T-cell response (Albert et d . ,1998). The pattern recognition receptor, CD14, has been implicated in binding and phagocytosis of apoptotic cells by human macrophages (Devitt et al., 1998). Several “phagosomal” bacteria such as M . tuberculosis, M . aviurn, M . bovis BCG, and salmonellae can induce apoptosis in their host cells on infection (Hayashi et al., 1997; Klingler et al., 1997; Lammas et nl., 1997; Rojas et al., 1997; Zychlinsky and Sansonetti, 1997). This mechanism, therefore, could provide a new pathway for targeting antigens of “phagosomal” bacteria from macrophages into the potent MHC class Ia processing pathway of dendritic cells.
D. MHC CLASSI1 PRESENTATION, THE STRAIGIIT WAY MHC class II-restricted CD4+ T cells represent the central T-cell population in the immune response against intracellular bacteria that reside in phagosomes. M HC class II-presented peptides are generated from exogenous proteins in acidified late endosonial/lysosomal compartments. Both endopeptidases (cathepsins D, E, L, S) and exopeptidases (cathepsins B and H, aminopeptidase) can be involved in this process (Riese et al., 1996; Fineschi and Miller, 1997; Hewitt et al., 1997). Some combinations of proteases can generate certain antigenic peptides but may degrade others (reviewed in Watts, 1997). The IFNy activation of human and murine macrophages selectively induces cathepsins H and S (reviewed in Chapman, 1998) as well as B and L, but not D (Lah et al., 1995). Cathepsins are delivered from the trans-Golgi network by the two M6PR (Kornfeld and Mellrnan, 1989). The compartment where MHC class I1 molecules are loaded with peptides has been termed MIIC or CIIV, depending on the definition of its vesicular stage (Pieters, 1997; Watts, 1997). Although it is still under debate whether these compartments are distinct from other late endosomesAysosomes (Pierre et al., 1996), their multilarnellar morphology is quite distinctive. A similar morphology has been described for autophagosomes (Dunn, 1994), some late endosomal compartments (Kobayashi et al., 1998), and vacuoles harboring L. niexicana (Schaible et al., submitted). Interestingly, the L. mxicana vacuole gains access to MHC class I1 molecules, which may allow peptide loading in the vacuole (Antoine et al., 1991; Russell et al., 1992).Similarly, MHC class I1 molecules engulfed together with mycobacteria may acquire antigenic peptides in the phagosome (Clemens and Horwitz, 1995). In addition, antigens secreted or released from mycobacteria into the phagosomal lumen enter MIIC and
IhTR4(’EI.LL~I,AH BA( TEHIA AND TIlE ILIMLJNE 5Y5TEM
301
other lysosomal compartments, ii situation that can be extrapolated to other pathogens (Xu et al.,1994; Prigozy et al.,1997). Because many intracellular bacteria alter their protein expression pattern after internalization, this mechanism would allow antigens expressed intracellularly to be recognized by the iinniune system. Homodinieric MHC class I1 molecules are targeted by the invariant chain (Ii) as a nonarneric complex to the MIIC/CIIV compartment (Wolf and Ploegh, 1995; Pieters, 1997; Brachet c.t n l , 1997). Subsequently, cathepsins S, L, and probably other proteases digest the invariant chain so that it is trimmed to the peptide aa81- 104 (CLIP), which occupies the peptide-binding groove (Mizuochi et al., 1994; Riese et nl , 1996; Villadangos ct nl., 1997; Chapman, 1998). MHC class I1 molecules are now translocated into a different compartment to associate with HLA-DM, which facilitates the exchange of CLIP with a foreign peptide (Ferrari et al , 1997; Villadangos et nl., 1997). This peptide is still available for further proteolytic trimming by exopeptidases (Nelson ef d.,1997). MHC class II molecules can accommodate peptides with a length of 15 to 22 amino acids. Peptide-loaded M HC class I1 molecules are then transported directly to the plasma membrane (Wubbolts et nl., 1996). Receptors capable of scavenging antigens are also localized in the MIIC/CIIV compartments, such as surfiice Ig in B cells (Dr‘ike et nl., 1997) and the M M R in human 1997). MHC class I1 molecules alone can dendritic cells (Prigozy ef d., also bind wliole protein3 in a partially folded form independent froin HLADM or proteases and transport them to endosomes for further proteolytic trirnming. Although tlie initial proteidMHC class I1 complex is sodium dodecylsulphate (SDS) unstable, it becomes SDS stable after further processing. Both SDS stable and unstable complexes stimulate specific T cells (Lindner and Unanue, 1996). In infections with “phagosomal” pathogens, MHC class II-restricted CD4’ T cells dominate both the induction and effector phases of the immune response. Nevertheless, it is clear from the scheme detailed earlier that the stage of the phagosoine harboring an intracellular pathogen and its connection to compartments where antigen processing and peptide loading of MHC class I1 molecules occur, influences the ensuing iininune response. Proteolytic antigen processing and binding of many peptides to MHC class I1 molecules are optimal at pH 5-6 (Harding et nl., 1991), but processing of both antigens and Ii is less efficient in early endosomes (West et nl., 1994; Swier and Miller, 1995). Under the mild acidic (pH 6.5) and partially hindered proteolytic conditions in the iiiycobacterid phagosorne, peptides with different binding properties are selected. Iinmunodoininant peptides froin tlie inycobacterial antigens, HSP60 and p30/34, are loaded to MHC class I1 DR17 under both acidic and neutral conditions. whereas
302
ULRICH E. SCHAIULE ef rd
a different peptide from p30/34 only binds under neutral conditions (Geluk et al., 1997).The MHC class I1 molecule is probably loadedwith this peptide only at the APC surface or after uptake into the mycobacterial phagosome. Processing of somatic antigens of tubercle bacilli is resistant to lysosomotropic agents such as chloroquine and ammonium chloride, suggesting that processing of these antigens occurs in a nonacidified compartment, such as the mycobacterial pliagosome. Processing of soluble antigens, however, is sensitive to these modulators of the endosomaVlysosoina1pH (Balaji and Boom, 1998). Mycobacterial material, cg., LAM, is trafficked out of the phagosome and targeted to lysosoinalcompartments (Xuetnl., 1994).Effective processing of these antigens could therefore depend on acidification and lysosomal targeting. Similarly, Chlamydia-derived antigens are also trafficked out of the phagosome and may therefore reach compartments competent for antigen processing (Rockey and Rosquist, 1994).
E. UNUSUAL ANTIGENSPRESENTED BY UNUSUAL MOLECULES It is now evident that microbial antigens are presented to unconventional T-cell populations by surface molecules other than MHC class la and I1 or even in the apparent absence of any specialized presenting molecule. The almost exclusive specificity of these modes of presentation for bacterial antigens suggests that they are the outcome of a long-lasting coevolution between the mammalian immune system and intracellular bacteria. 1. Specializedjbr Bacterial Antigen,P, the M H C Class lb Molecules In mice, the nonclassical MHC class Ib molecule H2-M3 presents short peptides containing N-formyl-methionine (N-f-met) to a subset of CD8' T cells (Kurlander et al., 1992; Pamer et al., 1992). This T-cell population confers protection against experimental L. mnocytogenes infection in mice (Kaufmann et nl., 1988; Harty et al., 1996; Pamer et al., 1997). N-f-metcontaining peptides originate from N termini of bacterial or mitochondria1 proteins (Pamer et al., 1992; Lindahl et al., 1997),which are protected from deformylases present in the bacterial cell wall (Lenzet al., 1996).Mitochondria most probably represent descendants of prokaryotes and, aside from around 30 mitochondria1polypeptides, N-f-met is absent from the mammalian proteome (Shawar et al., 1994). Thus, MHC class Ib molecules appear to be specialized to present bacterial antigens, and N-f-met peptide-specific CD8+ T cells are virtually specific for antigenic peptides of listerial origin (Lindahl et al., 1997).This notion is further corroborated by the isolation of MHC class Ia-independent CD8' T cells from mice inoculated with live or dead inycobacteria (DeLibero et al., 1988). Moreover, MHC class Iaindependent CD8+ T cells that are additionally CD l-independent have recently been isolated from tuberculosis patients (Lewinsohn et al., 1998).
INTRAC'ELL,IJLAH BACTERIA AND THE I M M U N E SYSTEM
303
However, N-f-met peptide-specific CD8' T cells have not yet been identified in infections with other intracellular bacteria. Similar to MHC class Ia, H2-M3 is expressed on the cell surface noncovalently linked to P2m (Lenz et 01, 1996). At present, it is still unclear if TAP is needed for the presentation of N-f-met peptides as TAP transport of small peptides is inefficient and formylation impedes TAP transport ( Fischer-Lindahl, 1997). However, processing and peptide loading of H2-M3 can also occur in acidic compartments such as phagosomes or endosornes because APC from TAP-negative mice inoculated with live or killed listeriae can stimulate MHC class Ib-restricted CD8' T cells (Lenz and Bevan, 1996: Lenz et a l , 1996). The peptide-binding groove of H2-M3 can only accommodate small peptides, which is consistent with the Fact that N-f-met peptides from L. inonocytogenes usually represent penta- to heptapeptides (Gulden Pt al , 1996; Lenz et al., 1996). Hydrophobicity is probably the most important determinant for peptide binding to H2-M3 (Wang et al., 199Tj; Fischer-Lindahl, 1997). Although presentation of glycolipids by H2-M3 to L. ni"nor~t"~enes-specificCD8' T cells has been claimed (Huffman et a/., 1996; Nataraj et d., 1996; Kurlander and Nataraj, 1997),it is not clear whether H%M3 can promiscuously bind several structurally unrelated antigenic ligands based exclusively on their hydrophobicity. It should be noted that no direct hoinologs of H2-M3 exist in hurnans and that N-f-met reactive T cells have not been identified in humans Parham, 1994). In mice, the low polymorphic nonclassical MHC class I molecule Qa-1 apparently plays a role in the induction of aiitiinicrobial T-cell responses. Qa-l-specific T cells have been isolated that recognize antigens from L. monocytogenes and confer protection on adoptive transfer into naive mice (Bouwer et nl., 1997).Similar to MHC class Ia molecules, Qa-1 can present nonameric peptides in a TAP-dependent manner (Aldrichet al., 1994).Qa1is probably an antigen-presenting molecule for certain y6T cells (see later). 2 Against the Dognrn. CD1 Mo1ecri1es Present Glijcolipid\ to T Cells CD 1 molecules were initially described as target structures, which are recognized i n the absence of any cognate or externallyadded nominal antigen by autoreactive T cells bearing either ap or y6 T-cell receptors. These T cells express neither CD4 nor the conventional CD8 heterodimer (double negative, DN) but 5ometimes the CD8a &unit as a homodimer (Porcelli, 1995; Kaufinann, 1996).In mice, CD1-dependent DN or CD4' Tcells have been described coexpressingthe N K cell marker, N K 1 . l , in addition to their TCR ap, wlbich consists of a single a chain (Va14-Ja281) combined with a restricted niimber of p chains (Bendelac et a l , 1995). In humans, similar CD1-specific T cells have been described that also express a limited TCR
304
ULHICH E. SCHAIRLE d al
repertoire (Va24Vpll) (Porcelli et al,, 1989; Balk et al., 1991; Exleyet al., 1997).Because the murine Val4 and the human Va24chain are homologous, the two T-cell populations are considered cognates. CDla, -b, and -c are expressed in humans but not in mice whereas the human CDld has its homolog in the murine CD1.1 (Porcelli, 1995). Human CDla, b, c are grouped together as group 1 CD1 molecules, which are mainly expressed on professional APC such as B cells and dendritic cells. Based on sequence homologies, C D l d and CD1.1 belong to group 2 CD1 molecules, which are additionally expressed on macrophages, intestinal epithelial cells, and hepatocytes (Brossay et al., 1997; Kaufmann et al., 1997; Roark et al., 1998; Szalay et al., submitted). The human CD1e and the murine cd1.2 genes probably represent pseudogenes that are not expressed. CD1 molecules share sequence homologies with classical MHC molecules but they are nonpolymorphic and their genes are located outside of the MHC (Porcelli, 1995). Like MHC class I molecules they are linked noncovalently to &m on the cell surface (Brutkiewicz et al., 1995; Bauer et al., 1997; Teitell et al., 1997), but expression is independent of TAP-1, TAP-2, and HLA-DM molecules (Porcelli et al., 1992; de la Salle et al., 1994). Although CD1 molecules are retained in the ER in the absence of P2m, (Sugita et al., 1997),&m-independent expression of CDld has been reported on human intestinal epithelial cells (Balk et al., 1994). Human T-cell lines, most of them DN or CD8+, have been described that recognize mycobacterial glycolipid antigens in the context of CDla, -b, and -c (Porcelli et al., 1992; Beckman et al., 1994, 1996; Sieling et at., 1995; Thonissen et al., 1995; Moody et al., 1997; Rosat et al., 1998). The specific reactivity of these T-cell lines discriminates between the two species, M . tuberculosis and M . leprae (Beckman et al., 1996). To date, the list of antigens presented by C D l b and CDlc includes components derived froin the mycobacterial cell wall: mycolic acids (Beckman et al., 1994), LAM and PIM (Sieling et nl., 1995; Beckman et al., 1996), phospholipids (Rosat et al., 1998), as well as mycolyl glycolipids such as glucose monomycolate (GMM) (Moody et al., 1997). It has been reported that CD1.l can also present glycolipids to murine T cells, although the antigenic ligands identified so far are not of microbial origin, e.g., agalactosylceramide (Kawano et al., 1997) and glycosylphosphatidylinositol (GPI) (Joyce et al., 1998). At present, the biological relevance of these CD1. l-presented glycolipid antigens is not clear. There are also two reports that CD1.l can present peptides to conventional CD8' T cells (Castano et al., 1995; Lee et al., 1998). CDla-reactive T-cell clones with specificity to yet undefined microbial antigens have been isolated from the peripheral blood of healthy donors (Dellabona et al., 1993).
II\’THACELI.IJLAH HACTEHIA AKD THE IhlhlUNE SYSTEM
30s
So far, little is known about the processing requirements of glycolipid antigens presented by CD1. The lysosoinotropic agent cliloroquine abolishes antigen presentation by CDlb, suggesting that binding and probably processing require acidic conditions (Sieling et nl., 1995).This observation is in line with the fact that the cytoplasmic tails of CDlb, c, and d contain a tvrosine-based targeting signal (YQDI) that binds two adapter proteins, AP-1 and AP-2, for trafficking to clatlirin-coated pits and late endosomes (Porcelli, 1995; Sugita et nl., 1996).Truncation of the endosomal targeting signal abolishes trafficking of CD l b to late endosoinal compartments and C D l b mediated antigen presentation (Jackman et nl., 1998). It has been showi that C D l b can sample glycolipid antigens such as LAM delivered by the M M R and possibly other scavenger receptors to late endosomes (Prigozy L’t nl., 1997). Interestingly, the intracellular distribution of CDla is different from CDlb, c, and d and resenihles that of MHC class Ia inolecules in that it is not trafficked to late eiidosomavlysosornal compartments and is mainly found at the cell surface (U. E. Schaible et nl., in preparation). CD1.l is also trafficked to late endosoinal coinpartinents where it colocalizes with H2-DM (Brossay ct nl., 1998). The crystal structure of CD1.1 has heen solved, and a common principle for the binding of glycolipids to all CD1 molecules has been propoqed (Zeng et al., 1997). A h ~ u g hthe overall structure of CD1.l is similar to MHC class I arid I1 molecules. the cleft of the antigen-binding groove is narrower, deeper, and more voluminous due to the elevated a-helix. The electrostatic potential of the groove is neutral and comprises only a few residues capable of forming hydrogen bonds with a ligand. However, the bottom of the groove contains hydrophobic residues that could support the anchoring of hydrophobic entities. Similar to H2-M3, it has been suggested that ligands ;ire bound by CD 1 exclusively through hydrophobic interactions (Zeng et al., 1997), which are facilitated by a conforinational change of CD1 in the acidic environment of the MIIC. Using different GMM constructs, structural featiires of CDlb-binding entities were tentatively defined as two hvdropliobic acyl chains that bind into the groove in a relatively unspecific manner. The hydrophilic sugar moieties are thereby positioned outside of the groove for specific recognition by T cells (Moody et nl., 1997). Moreover, it has been shown that the acyl side chains of ligands, such a5 LAM, PIM, or GMM, are required for high-affinity binding to CDlb. Partial unfolding of the a helices of C D l b at an acidic pH exposes a liydrophobic-binding site to facilitate binding (Emst et al., 1998). However, a-galactosylceraniide does not require trafficking through endosoma1 compartments but binds directly to CD1.l on the surface of fixed APC for NK T-cell stimulation (Burdin ct d., submitted).
306
ULRICH E SCIIAIBLE (4 (11
Employing microbial glycolipids as T-cell antigens would have at least two benefits to the host: (i) They are critical cell wall components, thus limiting the chances that they are altered by mutation (Porcelli et al., 1998),and (ii) because M . tuberculosis-infected macrophages are hindered in their capacity to activate CD4' T cells in an MHC class II-restricted fashion, the CD 1 system may help overcome insufficient CD4' T-cell activation. However, CD 1 molecules are mainly expressed on dendritic cells rather than macrophages and CD1 expression in dendritic cells is downregulated by infection with M . tuberculosis (Stenger et al., 1998).
3. What Do y6 T Cells See? Human y6 T cells represent a small population of all peripheral T lymphocytes that are rapidly expanded upon in uitro stimulation with bacterial components (Modlin et al., 1989; Hara et al., 1992; Garcia et al., 1997).y6 T cells from healthy donors respond strongly to small phosphatecontaining, nonproteinaceous compounds of bacterial origin that include isopentenyl pyrophosphate and alkyl derivatives thereof (Pfeffer et al., 1990; Constant et al., 1994; Schoel et al., 1994; Tanaka et al., 1994, 1995). In most organisms these compounds are natural precursors for various sterol metabolites, suggesting that processing of these compounds for antigen presentation is not required. The list of potential y6 ligands has now been extended toward other phosphate-containing compounds such as nucleotides, phosphosugars, and phosphoesters (Chien et al., 1996).Apparently, y6T-cell ligands are recognized in the absence of any known antigenpresenting molecules, probably by direct interaction between ligand and the yS TCR (Morita et al., 1995). Indeed, evidence has been presented that such ligands are transported from phagosomes containing live mycobacteria to the cell surface where they can then interact with and stimulate 7 6 T cells (Schoel and Kaufmann, 1998).Alveolar macrophages can serve as accessory cells for human y6 T cells activated by M . tuberculosis at low macrophage to T-cell ratios (Balaji et a/., 1995). At higher ratios, alveolar macrophages inhibit 76 T-cell responses, which can be overcome, at least partially, by exogenous IL-2. In mice, y6 T cells have also been described, although these cells do not respond to pliospholigands (Chien et al., 1996; Kaufmann, 1996). Following immunization with heat-killed M . tuberculosis, y6 T cells accumulate at the site of antigen deposition and can undergo antigen-specific proliferation ( Janis ct nl., 1989). Moreover, aerosol administration of mycobacterial PPD, but not ovalbumin, resulted in the accumulation of yS T cells in the lungs of mice (Augustin et al., 1989). However, these cells seem to recognize peptides derived from the ubiquitous HSP6O and some of them probably in a Qa-l-dependent manner (O'Brien et al., 1989; Imani
IVTHACELI,ULAR BACTER14 AND TlIE IhlblUNE SYSTEM
307
and Soloski, 1991). The HSP6O core sequence recognized by murine y6 T cells has been identified as a heptamer (aa181-190) with phenylalanine and leucine at positions 181 and 183, respectively, as essential amino acid residues (Fu et nl., 1994). It is tempting to speculate that the unconventional T-cell populations that mostly recognize bacterial antigens in the context of MHC class Ib, CD 1, or idependent of any antigen-presenting molecule represent the outcome of a long-lasting coevolution between bacteria and their mammalian host. The non- or low-polymorphiccharacter of the antigen-presenting molecules, as well as the skewed TCR repertoire used by the responding T cells, suggests that these APC/T cell systems are at the borderline between the innate and the specific immune response. VII. T-cell Subsets and Effector Mechanisms
A. CON\WVTIONAL T CELLS 1. Helper T Cells
Since the observation by Mosmann and colleagues (1986) that CD4' TCR c.P cells can be divided into two populations according to their cytokine secretion patterns, a wealth of data has been published on Th1 and Th2 cell subsets and the roles these T cells play in infectious dseases (Lucey et nl., 1996). Although the initial identification of these T-cell subsets and their subsequent contribution to polarized immune responses were revealed by experimental infections in mice, there is now increasing evidence that a similar dichotomy of CD4' T cells also exists in humans (reviewed in Romagnani ct nl., 1997). The two populations arise from a single precursor cell termed ThO and are induced to differentiate into subsets primarily depending on the cytokine milieu during antigen activation (reviewed in Seder and Paul, 1994; Abbas et al., 1996; Fearon and Locksley, 1996). Additional factors include cognate interactions between costimrilatory molecules on the surface of T cells and APC, including the CD40/CD40L (CD154) interaction and the B7-1 (CD80), B7-2 (CD86)l CD28, and CTLA4 (CD152) system (reviewed in Croft and Dubey, 1997). TI11 cells produce IFNy and TNFP, whereas Th2 cells secrete IL-4, IL-5, IL-6, and IL-13. It is now established that IL-10, which was originally considered a Th2 cytohne, is also produced by Th1 cells, activated macrophages, and dendritic cells. In addition to. and resulting from differences in their cytokine patterns, these T cell subsets differ in the host immune responses that they induce. IFNy produced by Thl cells promotes cellmediated immunity characterized by macrophage activation, stimulation of CD8' CTL, and the production of opsonizing Ig isotypes IgG2a and
308
ULRICH E. SCHAIBLE et nl
IgG2b. In contrast, Th2 cells control the differentiation of B cells to Igproducing plasma cells mediated by IL-4 and promote Ig class switching to IgE and IgA by IL-4 or IL-5, respectively. Moreover, IL-5 activates eosinophils as important effector cells against helminths. Thus, IL-4 and IL-5 are central to the control of hehninth infections (reviewed in Mosmann et al., 1997). It has been suggested that IL-13 may be an equally, if not more, important Th2 type effector cytokine. Experimental infection of mice with the gastrointestinal nematode Nippostrongylus brasiliensis revealed that whereas IL-4-deficient mice could expel the worms normally, signal transducers and activators of transcription 6 (Stat 6) KO mice, which have a block in IL-4 and IL-13 receptor signaling, as well as mice treated with a specific antagonist for IL-13, failed to expel the parasites (Urban et al., 1998). Such observations have subsequently been confirmed in IL13 KO mice that demonstrate a unique role for this cytokine in Th2 responses and further suggest that IL-13 and IL-4 are not redundant (McKenzie et al., 1998). Although the pattern of cytokines secreted by T helper cells theoretically divides them into two groups, the distinction, especially in disease states, is not absolute, and the search for molecular markers separating the two is still in progress. Candidates under consideration are CD30 as a marker for Th2 cells (Prete et al., 1995) and the p chain of the IL-12 receptor as a marker for Th1 cells (Szabo et al., 1997). More recently, differential expression of chemokine receptors has been proposed, with human Thl clones preferentially expressing CXCR3 and CCR5 and Th2 clones expressing CCR3 and CCR4 (Sallustoet al., 1997; Bonecchi et al., 1998; Loetscher et aZ., 1998). Finally, using differential mRNA display techniques between Thl and Th2 clones, two groups have identified a surface marker, Tl/ST2, that is expressed on Th2 cells but not Thl cells and is critical for Th2 effector cell function (Xu et al., 1998a; Lohning et al., 1998). Infections with intracellular pathogens generally induce a T h l response that is protective (Fig. 2). Probably the best experimental model for defining this phenomenon and dissecting the requirements for the generation of this response has been infection of resistant and susceptible strains of mice with the protozoan pathogen L. major, proving that resistance correlates with Thl responses and the production of IL-12 and IFNy, whereas
FIG.2. A depiction of the regulation of T-cell differentiation and function by cytokines produced as a result of bacterial infection. Only the major regulatory cytokines are included here, and solid lines represent activating functions, whereas downregulatory mechanisms are represented by broken lines.
INNATE “Nonspecific” (Rapid response)
I
4
I
I
I I
I
INTERMEDIATE “Broadly specific” (Delayed response)
“Specific” (Requires more time)
310
ULRICH E. SCHAIBLE el (11
susceptibility is linked to Th2 responses characterized by IL-4 secretion, elevated IgE levels, and failure to resolve the infection (Locksleyand Scott, 1991). The importance of this T-cell dichotomy in human disease states is best exemplified by the host response to infection with M. leprae where the tuberculoid form of the disease is characterized by a Thl type response resulting in activated macrophages, granuloma formation, and few detectable organisms. In contrast, toward the leprornatous pole of the disease, a Th2-like response occurs with increased levels of IgE and IgG1, little, if any, granuloma formation, and an abundance of acid-fast bacilli (reviewed in Lucey et al., 1996). However, it should be noted that in lepromatous skin lesions it is CD8' T cells rather than CD4' T cells that predominate, and these cells produce IL-4, IL-10, and IL-13, which suppress macrophage activation and potentially inhibit the control of bacterial replication (Modlin and Nutman, 1993). For infections with other mycobacteria, however, the question of whether strict ThlK112 cytokine profiles affect pathogenesis of the disease is still under debate. In vitro infections of peripheral blood mononuclear cells with M . bovis BCG resulted in the production of Thl cytokines 4-5 days after infection, and by 10-12 days this response had declined and Th2 cytokines predominated (Sander et al., 1995). The in vivo relevance of this finding, however, is uncertain as BCG vaccination induces strong Thl responses. The development of a Thl response is a regulated event that depends critically on the production of IL- 12 by macrophages and dendritic cells at the initial stages of microbial encounter (Hsieh et al., 1993; Macatonia et al., 1995). The mechanisms of bacterially induced IL-12 secretion have not yet been fully elucidated, although the interaction of bacterial components such as LTA, LPS, and LAM with pattern recognition receptors such as CD14 on mononuclear cells can induce IL-12 secretion and thus provide the initial stimulus (Cleveland et al., 1996; Nigou et al., 1997; Yoshida and Koide, 1997). It has been shown that bacterial DNA with CpG sequences can induce the production of IL-12 in macrophages (Chace et al., 1997). IL-12 signals via Statsl, 3, and 4, with Stat4 being activated exclusively by IL-12 (Jacobson et al., 1995). Consequently, KO mice deficient for either IL-12 or Stat4 show markedly reduced T h l responses (Kaplan et al., 1996; Magram et al., 1996; Thierfelder et al., 1996). One scenario is that IL-12 produced early in infection acts on N K cells and yS T cells in conjunction with TNFa to induce the production of IFNy from NK cells (Unanue, 1997). However, when T cells from transgenic mice are primed in vitro with antigen, APC, and IL-12, but IFNy is neutralized, Thl production is unaffected, suggesting that IL-12 can affect T-cell priming independently of IFNy production. However, in apparent contrast to this, IFNy KO mice display a default Th2 response when challenged with
L.t,iajor (reviewed in Reiner and Setler, l99rj). The early IL- 12 produced by inacropliages in response to infection wit11 M . /xicis BCG tlepentls critically on the rapid production o f IFNy (Flesch et crl., 1995),and findings from Sher and colleagues using 7'. p i d i i revealed that resting inacropliages indeed require IFNy for IL-12 productioii (Sorisa ct al., 1997b).The niain source of this IFNy is the N K cell activated bv IL-12 produced by dendritic cells. Once this is achieved, IFNy acts to &)stirnillate IL-12 production from infected inacropliages. Under the selective pressure of IL-12 and/or IFNy, stable T h l cells are generated and in turn produce IFNy, which can act in two ways to continue the self-pell~etiiationof tlie protective response: (i) this cytokine can forin part of ii feedback loop to fuither enhance tlie production of IL-I 2 from activated macrophages and (ii) it can maintain the expression of IL-12 receptors on CD4' T cells (Guler et d . ,1996).More recent cxperinients liave re~ealedthat IL-18 can potentiate IL-12-driven Th 1 cell development, although it cannot itself induce Th1 cell differentiation (Robinson et nl., 1997). However, IL-18 KO mice show defective Th1 cell development i n response to A.! bouis BCG (Takeda ct nl., 1998). The same conditions that are optinial for priming TliO cells to differentiate into TI11 cells are also inhibitory to the generation of TI12 cells. Th2 responses t~ypicallyocciir in helniintli infections and the critical factor determining Th2 cell priming is IL-4. The cellular soiirce of this early IL-4 remains a matter of debate, although candidate cells include mast cells, basophils, eosinophils, N K l + T cells, and naive 0 4 ' T cells themselves (Paul, 1997). In infections with intracellular bacteria, some production of IL-4 occurs very early in infection but it is short lived, and evidence from tlie authors' laboratory suggests that one fiinction of this cytokine in murine listeriosis is to induce the production of distinct cliemokines such as MCP-1 (Flesch et d.,1997).Indeed, it has now been shown that NK1' T cells, which are potent IL-4 producers, change phenotype to become IFNy prodiicers during the course of infection with M . houis BCG, thus promoting the generation of a protective Th1 response (Elnoto ct d., submitted).This is consistent with data from Clien and Paul ( 1997)showing that repeated stimulation of NK1' T cells i r i uitro converts them from IL-4 to IFNy producers. 2. Cytdytic CDh" T Cells As detailed in Section VI,C, CD8' T cells are MHC class I restricted and generally respond to endogenously generated antigens: citlier selfantigens or those produced by pithogens residing in the cytoplasm of the host-cell, such its listeriae and rickettsiae. However, conclusive evidence now shows that the vacuolar localization of intracellular pathogens such
312
ULRICII E SCHA1BI.E rf cd
as mycobacteria and salmonellae does not prevent bacterial antigens from entering the MHC class I processing pathway to generate a CD8+ T-cell response (see Section V,C). Experimental infection of mice deficient in either CD4+ or CD8" T cells with brucellae or listeriae has proven that protection against these pathogens depends critically on the generation of a CD8+ T-cell response (reviewed in White et al., 1996).Siinilarly, in the murine model of tuberculosis, experiments using Pzm KO mice, which are devoid of functional MHC class I molecules and thus have no MHC class I-restricted ap T cells expressing the CD8 heterodimer, revealed that susceptibility to M . tuberculosis is increased in these mutant mice (Flynn et al., 1992). Furthermore, protection against experimental tuberculosis could be transferred by CD8' T cell clones recognizing inycobacterial HSP6O (Silva et al., 1994). Moreover, CD8+ T cells could be generated from mice infected with M . tuberculosis as well as from human peripheral blood cells following stimulation with live M . bovis BCG and were shown to lyse infected target cells (DeLibero et al., 1988; Turner and Dockrell, 1996; Lewinsohn et al., 1998; Conradt et al., submitted). CD8' T cells that recognize peptides derived from the ESAT 6 antigen of M . tuberculosis have been documented in patients with clinical disease as well as following treatment and in healthy contacts. These cells exhibited peptide-specific IFNy secretion and acquired cytolytic activity following antigenic restimulation in vitro (Lalvaniet al., 1998). Thus there are two possible functions of CTL during the course of infection with intracellular bacteria: (i) target cell killing could directly curtail the growth of the bacteria via cytolysins introduced into the infected host cell by perforin, as demonstrated for M . tuberculosis in both the human and the inurine system (DeLibero et al., 1988; Stenger et al., 1997). One such compound, granulysin, which is specific for human T and N K cells (Pena and Krensky, 1997), has been tentatively described as a candidate molecule in mediating the antimycobacterial effects. (ii) Lysis of incapacitated cells that can no longer control the infection would release the organisms so that other more proficient cells can phagocytose and ultimately kill them (Kaufinann, 1988). Supporting a role for cytolytic mechanisms are experiments using perforin KO mice where a secondary Listeria infection was exacerbated in the absence of CD8' T cells (Kagi et al., 1994). However, in apparent contrast to these findings, studies of experimental infection with M . tuberculosis revealed that mice lacking either perforin- or granzyme-mediated cytolytic mechanisms show no difference in the control of primary inycobacterial infection as compared with wild-type mice (Cooper et al., 1997a; Laochumroonvorapong et al., 1997), although the potential role of these mechanisms in the late phase of infection remains to be addressed.
B. UNCXNVENTIONAI, T CELLS 1. y6 T Ccl1.r.
In both mice and humans, T cells expressing iiii alteniative TCR (yS T cells) coinprise a minor population of 1-5% within tlie peripheral I)lootl and lymphoid organs. In Iiuinaii adults, 50% of the y6 T-cell population express Vy2S2 (DeLibero et d.,1991; Kabelitz c’t d . ,1991);is this population is expanded froin less than Fj% t h i n g ontogeny. This represents an impressive example of the impact of inicrobes on extrathymic T-cell repertoire expansion apparently independent from direct host defense (Parker et al., 1990; Kaufinann. 1996). With regard to intracc3llular infections, these cells appear to participate in the host response, especially toward mycobacteria, but the exact nature of their function reinains to be fully elucid:ited. These T cells accumulate in the lesions of leprosy patients (Modlin et al., 1989)and are expanded in patients with salaionella infkction (Hara et al., 1992). Mycobacterial pliospholigands stimulatc all huiiiiiii y6 T cells expressing the Vy2S2 chain coinbination. Functionally, this stimulation resembles that of ap T cc>llsby superantigens. However, antigen ~)iiidin~recognition occurs at a site that is probably conserved in all Vy2 chains- and close to the highly variiible CDR3 region of the TCR responsible for fine antigen specificity (Rocket nl., 1994; Schild et nl., 1994). Because of this oligoclonal stimulation, the frequency of myco1)acteiiastimulated y6 T cells is at least in the same or even higher order of magnitude ;is tlie frequency of antigen-specific c.p T cells stiiiiulated clonally by bacteria. Evidence shows that growth factors such as IL-2 and IL- 15 produced by activated CD4+ T cells are required for the proliferation of human 76 T cells. Thus, although stimulation of y6 T cells with phosplioligands induces tlie activation markers CD2Fj and CD69, they fail to proliferate (Wesch et nl., 1997). This is further corroborated by findings that Vy2 T cells froin HIV-infected individuals do not respond to mycobacterial antigens due to their deficiency of antigen-specific CD4+ Thl cells ( W t d et nl., 1996). Importantly, because IL-10 production strongly inhibits this response, this mechanism of controlling y6 T-cell proliferation can only occur following the generation of TI11 cytokines (Pechholtl et nl., 1994). Siinilar to human y6 T cells, murine y6 T cells, which do not recognize phospholigands but peptides, expand to an oligoclonal population of y6 T cells expressing the Vyl chain 011 stirnulation hth HSP6O (O’Brien rt (11.. 1989; Fu ~t a / , , 1994; Belles et ( I / . , 1996). Data from many groups have proposcd a role for y6 T cells in inurine listeriosis, a s these cells accuinulate and secrete IFNy in response to Listerici infection (Hiromatsu ct d . ,1992; Skeen and Ziegler, 1993). It has been suggested that this activation of y6
314
U L I W H E. SCHAIRLE cf
(I[
T cells requires the iiiacropliajie-produced cytokines IL-12 and IL-1 (Skeen and Ziegler, 1995) and is also dependent on CD45 (Fujise et al., 1997). Using a comparison between 76 and a@TCR KO mice, the authors’ group found that either a@ or y6 T cells were sufficient for early protection against Listeria infection. However, treatment of (YOTCR KO mice with anti-yS inAb exacerbated the disease, suggesting that compensation inechanisms occur between these two T-cell populations (Moinbaerts et al., 1993). This is also illustrated in mycobacterial infections. Infection of mice iv with low doses of M. tuberciilosis revealed increased bacterial growth in 6 TCR gene disrupted inice between diiys 15 and 30, although by day 120 postinfection both mutant and wild-type mice were able to adequately contain the infection. In contrast, higher doses that were not lethal in control mice rapidly proved fatal for the y6 T-cell deficient KO mice (Ladel et d ,1995a). However, D’Souza and colleagues (1997), using a range of doses of M . tuberculosis of high virulence administered by aerosol, found equivalent survival rates between wild-type and KO mice. A possible scenario to explain the apparent differences in these results is that in the high-dose iv infection of 6 TCR KO mice, the compensatory activation of ap T cells occurs too late to control this overwhelming infection and consequently the mice die. In the case of the lower iv infection and the aerosol infection, the organism do not overwhelm the animals prior to the activation of ap T cells, which then control the infection. During the course of these experiments, it was revealed that the 6 TCR knockout mice exhibited an increased influx of neutrophils into the mycobacterial granuloina in contrast to the priinarily lymphocytic infiltrate in wild-type inice (D’Souza et al., 1997), which may also account for the death of animals when higher doses are administered iv. Based on this and the observation that Listeria-infected S TCR KO mice develop liver abscesses rather than granulomas (Mombaerts et al., 1993), it has been proposed that yST cells primarily play a regulatory role in infections with intracellular bacteria by limiting the inflaiiimatory response that leads to tissue damage. Supporting this notion is the observation that splenic y6 T cells produce IL-10 during the course of a Listeria infection that coincides with maximum IFNy production and a decrease in inflammation and tissue damage, suggesting an important immunoregulatoiy role for these cells in controlling Th1 cell responses (Hsieh et al.. 1996). Consistent with their regulatory function, yS T cells have been implicated in the control of IFNy production by N K cells in response to Listeria infection (Ladel et al., 1996).
2. Cells Controlled by Group I CDl Molecules As discussed in Section V, CD1 molecules are a group of nonpolymorphic MHC-reIated polypeptides with the unique ability to present glycolipid
antigens to unconventional T cells. T cells that respond to CD 1 presented glycolipids express the arj3 TCH and are either CD4- CD8 or CD8' (reviewed in Porcelli et (11, 1996) and are capable of antigen-specific proliferation and IFNy secretion. Furthermore, these cells are c'ipable of lysing antigen-presenting cells infected with live niycobacteria but not uninfected CD1' cells (Stenger et nl., 1997).Although both double negative CD4-CD8 and CD8' CD1-restricted cells lysed infected APCs, they did so via different niechanistns. DN cells affected cell death via the Fas-FasL interaction, whereas CD8' cells einployed perforin- and granzymecontaining cytotoxic granules to mediate tlie lytic effect (Stenger et nl., 1997). Importantly, these elegant experiments also revealed that only lysis of infected macrophages by CD8' CD1-dependent T cells reduced the viability of the intracellular mycobacteria. Thus tliese experiments suggest that T cells reactive with microl)ial glycolipids have ;I dual role in the response to an invading pathogen: an initnunoreplatory role inducing the elimination of infected APC via apoptosis and the granule-dependent killing of bacteria.
3. Group 2 CD1-Coiitrolled T Ccll~-NKl T Cells The murine group I1 CD1 molecules control the development of N K l ' a@ T cells (Bendelac et nl , 1997).Tlie absolute requirement for CD1 in the development of these cells has been confirmed following the generation of CD1.l gene disruption mice that are devoid of NK1 T cells (Chen et nl., 1997; Mendiratta et al., 199'7) These specialized T cells develop mainly in the thymus and display a defined tissue distribution, accounting for the inajority of liver T cells, 20-30% of bone marrow T cells, 10-20% of mature thymocytes, and 0.5-1% of splenocytes, and are also found in the intestine within the lamina propria and the Peyers patches (Vicari and Zlotnik, 1996). Interestingly, the recognition of CD1.l by T cells is highly dependent on tlie cell type in which this molecule is expressed (Park c>t al., 1998) Despite the control by an MHC class I-related molecule, NK1 T cells are not CD8', but rather CD4' or DN. Many experiments have been performed to elucidate a function for these cells in tlie regulation of the immune response in general and specifically whether these cells play a role in tlie host response to infections. What is now undisputed is that these cells rapidly produce IL-4 when stimulated in viuo with anti-CD3 ~ of producing t)otll (Yosliiinoto; i d Paul, 1994)and that they are a l capable IL-4 and IFNy in tjitro when stimulated with anti-CD3 or CD1 (Arase et al., 1993; Chen and Paul, 199'7). It was originally proposed that these T cells produced the IL-4 required to promote the development of coiiventional CD4' TI12 cells (Yosliiinotoet nl., 199,5).However it is now established that nuce devoid of these cells sucli as & i n KO and CD1 KO mice
316
ULRICH E . SCH.41BLE et nl.
are capable of mounting functional Th2 responses, despite the diminished early IL-4 burst (Brown et nl., 1996; Smiley el al., 1997). I n the context of intracellular bacterial infections, experiments using L. iwnocqtogenes and M . bovis BCG have revealed that NK1 T cells in the target organ for Listeria follobing iv infection) rapidly lose this liver T~-il-secretingproperty following infection, and this modulation is mediated via IL-12 (Emoto et al., 1995). Infection of mice with Propionibncter i r i i n acnes confirmed these results and additionally implicated IL-18 in the effect (Matsui et al., 1997). Furthermore, when beige mice were infected with S. entericn serovar Cholerasuis, a suppression of conventional NK cell function was observed and a concomitant expansion of TL-4producing NK1 T cells in the peritoneal cavity of the mutant mice was noted (Enomoto et al., 1997). These mice are unable to mount a successful TI11 response and are consequently more susceptible to the infection, but these effects can be abrogated by the administration of a neutralizing antiIL-4 mAb. Together, these results suggest that although other sources of IL-4 may induce Th2 function, early IL-4 production by NK1 T cells as it default consequence of microbial infection can modulate the outcome of the specific immune response. More recent experinients suggest that during infection NK1 T cells acquire IFNy-secreting properties and lose their ability to produce IL-4. This shift in cytokine production is paralleled by alterations in the density of the cell (Emoto et al., submitted). Thus NK1 T cells may support, rather than counteract, protective immunity to intracellular bacteria.
(a
4. M H C Class lb-Restricted CD8+ T Cells
Experimental infections of mice with listeriae have revealed a subset of CD8' CTL that are restricted by the nonclassical MHC class Ib gene product, H2-M3, and these T cells form a significant component of the total MHC class I-restricted T-cell response (Bouwer et al., 1997; Lenz and Bevan, 1997). As described in Section V, these CD8+T cells recognize N-f-met-containing peptides (Pamer et al., 1992).What role these T cells actually play in protection against intracellular bacteria remains to be seen, although cloned non-M HC-restricted T cells specific for L. rnonocytogenes are protective in a inurine model of infection (Kaufmann et al., 1988). Furthermore, evidence showing that these cells are present and responsive in conventionally housed, nonspecific pathogen-free mice suggests either that the epitopes that are recognized are cross reactive or that priming requirements are not as stringent for these nonclassical MHC molecules (Lenz and Bevan, 1997). Together with the fact that fonnylated proteins are uncommon in the mammalian host and that the CD8+ T-cell responses generated occur across a variety of MHC haplotypes (Gulden et al., 1996),
it has been proposed that stimulation of these cells could form part of a vaccination stratea. It should be noted, however, that no homologous T cells have been identified in humans. C. COSTIMUIATION So far this review has considered the interaction of the MHC-antigen complex arid the appropriate cytokine environment as being critical in activating the specific T-cell response to intracellular pathogens. In addition, cognate receptor interactions such as CD40KD40L and B7/CD28/ CTLA4 are important to ensure that the correct T-cell response is generated. These molecules provide a second signal following engagement of the TCR in the two-step activation process necessary to activate naive T cells, with the failure to receive this second signal resulting in anergy (Jenkins, 1994). CD40L (CD154) is preferentially expressed on T cells, and KO inice with a CDlS4 gene disruption are severely impaired in their primary Tcell responses to protein antigens, as well as in their ability to activate macrophages to produce a variety of products such as TNFa, IL-1, and NO involved in antimicrobial defense (Grewal et nl., 1995). Importantly, the interaction between CD40 and CD154 during antigen presentation to T cells results in tlie secretion of IL-12, which is critical for the generat’1011 of a Th1 response (Grewal and Flavell, 1996). Experimental infection of KO inice with the intracellular protozoan L. mujor revealed that mice deficient either in CD40 or its ligand develop a more severe form of the disease, which correlated with the inability of the macrophages from these mice to produce NO and TNF and to respond to IFNy (Campbell et nl., 1996; Kamanaka ct nl., 1996). I n contrast, preliminary infection studies with L. monocytogenes revealed equivalent growth of this organism in both wild-type arid CD 154-deficient mice witliiii 48 hr and no difference in tlie clearance of a sublethal challenge dose. These findings suggest that both the early T-cell independent control mechanisms and the development of T-cell-mediated immunity to listeriae develop norinally in CD 154-deficient mice (Grewal et al., 1997). One explanation for the discrepancy in these results is that listeriae and other intracellular bacteria are inore potent IL12 inducers, via coinponents such as LTA, and therefore CD40KD154 interactions are less critical. Consistent with this, experiments using CD1-54 KO mice infected with M . ttiberculosis revealed no difference in either survival times or bacterial loads in KO mice as coinpared to CS7BLJ6 wild-type mice (Campos-Net0 cf (11, 1998). As M . triberculosis can exert direct effects on rnacrophages, inducing IL-12, TNF, and NO production (see Section III), this points to an activation of macrophages in a T-
cell-independent manner, thus bypassing the requirement for the CD40/ CD154 interaction. The interaction of CD28/CTLA4 (CDl52) with B7-1 (CD80) or B7-2 (CD86) on APC is a necessary requirement for T cells to produce the autocrine growth factor, IL-2. However, those T cells that utilize IL-4 as a growtli factor appeared not to depend on costirnulation, which has led to the hypothesis that Thl and Th2 cell responses require differential second signals, i.e., the T-cell differentiation toward IL-4-producing Th2 cells is less dependent on B7 than the induction of IFNy-secreting Th1 cells (reviewed in Gause ct al., 1997). Further indication of this differential activation requirement was provided using CD28-deficient mice that show reduced T helper cell activity and decreased Ig class switching (Shahinian et al., 1993). In this case, however, it appeared to be the IL-4-producing cells exhibiting a higher dependency on the presence of CD28. The current hypothesis remains that signaling via members of the B7 family is required to maintain T-cell effector functions whereas the CD40/CD154 interactions appear to primarily play a role in T-cell induction. With the interactions between these costimulatory molecules being critical in initiating and maintaining an effective T-cell response, any modulation of their expression by an intracellular pathogen may provide a means of subverting the host iininune response and enhancing its chances of survival. To date, most data on the modulation of costimulatory molecules by pathogens have been provided by Leishnlania, which induce macrophages to downregulate surface expression of B7-1 following infection both in oitro and in vivo (Kaye et al., 1994; Kaye, 1995). More recent reports have also demonstrated a leishmania1 protein, LeIF, which upregulates B7 expression on inonocytes and macrophages (Probst et al., 1997). A similar situation is seen following infection of macrophages in oitro with live S. t y p l i i n i i i r i t m , where surhce expression of B-7 was downregulated following infection with live bacteria, but was increased if the macrophages were pulsed with dead bacteria (Gupta et at , 1996).Heat-killed B. abortus, which is currently being considered as a vaccine candidate, induces a strong Tlil cell response, partially mediated by its ability to upregulate the expression of B7-1 and B7-2 on liuman inonocytes (Zaitseva ct nl., 1996).Together, these data suggest that high levels of costimulatory molecules are necessary for the initiation of successful Th1 responses, which lead to macrophage activation. D. REMJLATIONOF T H E IMMUNE RESPONSEHY CYTOKlNE5 1 Mncrophnge Activatioii A critical step in the resolution of infection by an intracellular bacterium is the elimination of the organism by the antibacterial properties of the
IKTHACELLIJIAR HACTEHIA A N D THE I M M U N E SYSTEM
319
activated macrophage. The central cytokine required to initiate these events is IFNy, which can synergize with macropliage-produced TNFa to mediate many niicrobicidal mechanisms, such as ROI and NO production discussed in Section I11 (reviewed in Unanue, 1997; MacMicking et al., 1997). The absolute requirement for IFNy in resolving intracellular infections was first deinonstrated by antibody-mediated neutralization of this cytokine in vivo and was subsequently confirmed using KO mice with a deletion of either the lFNy receptor a chain or the IFNy gene itself. All of these inice were highly susceptible to infections with listeriae, mycobacteria, salmonellae, and yersiniae (Buchmeier and Schreiber, 1985; Cooper et ul., 1993; Flynri et a l , 1993; Autenrieth et al., 1994; Hess et al., 199613). Similarly, patients with a inlitation in the receptor for IFNy suffer from an increased susceptibility to mycobacterial infections due to a defect in their ability to activate rnacrophages in response to this cytokme ( Jouanguy et aE., 1996; Newport et ul., 1996). In addition to the requirement for IFNy in the induction of NO, macrophage activation by this cytokme can modulate other cellular functions. Following infection of macrophages, inycobactena reside in intracellular compartments that fail to acidify due to the paucity of the v-H'ATPase (Sturgill-Koszyckiet al., 1994).Activation of infected macrophages with IFNy leads to the accumulation of the vH' ATPase within the inycobacterial phagosome, its subsequent acidification, and segregation froin iron delivery via Tf (Schaible et al., 1998). As a consequence, the antitnycobacterial capacity of IFN 7-activated macrophage5 is enhanced greatly ( F l e d and Kaufniann, 1988; Chan et al., 1992; Schaible ct a l , 1998). In contrast, in hutnans, IFNy fails to consistently induce the antiniycobacterial effects of macrophages in vitro. However, the addition of 1,25-dihydroxyvitaniin D1, the biological active metabolite of vitainin ]I1,maximizes the tiiberculostatic potential of the activated 1987). Moreover, an alternamacrophage (Crowle et al., 1987; Rook et d., tive niicrobicidal mechanism ha5 been described that occurs in the absence of IFNy in which exogenous ATP mediates lysis of M . bovis BCG-infected human macrophages via P2Z receptors, resulting in death of the mycobacteria (Lainmas ct al., 1997). 2 Gr[mulonui Formation A granuloim is an organized lesion formed by infiltrating T cells and inononuclear cells at the site of infection. Forination of granulomas is a characteristic feature of inany infections caused by intracellular bacteria and they are often critical in restricting bacterial replication and confining pathogens to discrete foci. In listeriosis, for example, sterilizing immunity can be accompanied by granuloma formation, although the granulomas formed are often incomplete and are not essential for the eradication of
listeriae (Mielke et nl., 1989). In contrast, a successful imrnune response against mycobacteria is generally accompanied by tuberculoid granulomas. In both cases, similar mechanisms of cell recruitment occur, which are highly orchestrated events regulated largely by cytokines produced by T cells. Infection models with listeriae, brucellae, M . bovis BCG, and M . aviurn revealed that CD4+ T cells are critical for granuloma formation (Mielke, 1991; Lade1 et al., 199Sb; Hgnsch et al., 1996). Furthermore, the use of y6 T-cell KO inice suggests that these cells also exert considerable influence on granuloma formation. A deficiency in this T-cell subset resulted in abscess formation rather than the organized lesions generally observed in listerial infections (Mombaerts et al., 1993) and neutrophildominated inflammation following aerosol infection with M . tuberctilosis (D'Souza et al., 1997). However, in both cases antibacterial resistance was virtually unaffected by the altered tissue response. Early studies by Kindler and colleagues (1989) demonstrated the absolute requirement for TNF in the development of bactericidal granulomas in M . boois BCG infection, which has subsequently been confirmed using p55 TNF receptor KO mice or the soluble TNF receptor (Senaldi et al., 1996). However, KO mice infected with M . tuberculosis revealed that the number of granulomas formed was equivalent to that observed in wildtype mice, yet necrosis was only noted in TNF-competent mice, suggesting that this cytokine plays a role in immunopathology as well as bacterial containment (Flynn et al., 1995). It is now appreciated that other cytokines, particularly IFNy (Mielke et al., 1992) and IL-10, also play a critical role in granuloma formation (Wynn et d . , 1997). Data from a number of laboratories have revealed that the most important step in this process is the activation of nonspecifically immigrating CD4' T cells to produce IFNy, TNF, and IL-2, leading to the activation of macrophages and, together with the appropriate chemokine production (e.g., MCP-l), their attraction to the infection site (Mielke et al., 1997). Extravasation is then mediated via specific adhesion molecules (e.g., LFA-1, ICAM-l), which are upregulated during this process. However, although ICAM-1 KO mice failed to form granulomas following aerosol infection with M. tuberculosis, this inability did not affect the outcome of infection (Johnson et al., 1998). Consistent with this are observations that both ICAM-1 and P-selectin KO mice also control infections with L. monocytogenes as well as M . bovis BCG with equal efficiency to wild-type mice (Steinhoff et al., 1998).Taken together, these results indicate that although granuloma formation may be the desired result of protective immunity, especially against mycobacterial infections, the failure of this process need not adversely influence the outcome of infection in every case.
I N T I ~ A ( ~ ~ I . I . l I L . BACTERIA 41~ .AN11 TI{E IhlMliNE SYSTEM
32 1
response^ As inentioiicd previously, intracellular bacteria do not ally induce a Th2 response followinginfection. However, concomitant with tlie induction of a Thl response to eradicate the liactenn is the risk of inflammatory mediated tissue damage following an uncontrolled iinmune response. Consequently, counterregiilatory cytokines are necessaiy to downregulate the production of inflammatory cytokines such :is IFNy and TNF. One of the most important cytokines in this regulatory process is IL-10, which is produced hy a variety of cell Q p s , including Th2 cells, B cells, and macrophages (Moore ct a / . , 1993). Its major role s e e m to be to inhibit the prodiiction of IL-12 (D'Andrea ct d . , 1993) and to antagonize the effects of IFNy on macrophage activation. For instance, IL-10 induces internalization of MHC class I1 molecules from the surface of APC (Koppelnian et ml., 1997). Thus its regulation plays a critical role in infections where macrophages are the host cells for pathogens. This has been confirmed in experiments where transgenic mice that overproduce IL-10 failed to clear invcobacterial infections, despite the abundmt production of IFNy by T cells, with IL- 10 overriding the antimycobacteiial signals received by the macrophages (Murray r t d., 1997). The iniportance of the regulatory role of IL-10 is emphasized by experiments in which IL-10 KO mice infected with T g m d i i succumb to enhanced liver pathology and necrosis due to incrcwsed production of both IL-12 and IFNy (Gazzinelli et d . , 1996). Interestinglv, it has been shown that IL-12 can induce T cells to produce IL-10, altliough not as efficiently as it mediates IFNy production. Thus the possibility exists that IL-12 can regulate its own production via IL-10 (Meyaard et ol., 1996) and therefore provide a rnechanisrn to downregulate tissue destruction following pathogen-induced iininune responses. 3. Copitrol of Itzapproprioto
4 Efcct
oti
Init,iiine
Lytti),hn),oei~i~
In addition to their biological activities on effector cells of tlie immune system, such as T cells, macrophages, and N K cells, several cytokines including IL-12 and IL-4 can, in combination with other factors, enhance the surviviil and growth of earl?, lieinatopoietic stem cells (reviewed in Chehinii and Trinchieri, 1994; Sonoda, 1994). IL-12 appears to have a dual role in hematopoiesis, in that it enhances the proliferation of inyeloid and B-cell precursors and yet, if NK cells are present, can also inhibit hematopoletic colony forination as a result of the IL-12-induced production of IFNy and TNFa (Cheliimi and Trinchieri, 1994). Sindarly, IL-4 also exhibits diverse effects on hematopoiesis as it can act on committed as well as earl>,progenitors enhancing granulopoiesis but inhibiting mono-
322
ULRICII E. SCHAIBLE et 01.
poiesis from CD34+ bone marrow precursor cells (Snoeck et al., 1996). To date there is little information on what effect, if any, infectious agents may have on these processes. However, the authors have described for the first time the rapid induction of IL-4, but not IL-12, in murine bone marrow cells by M. tuberculosis, M. bovis BCG, and one of their major cell wall components, mannose-capped LAM (Collins et al., 1998). Experiments are currently underway to identify the mechanism of this induction and its possible implications on hematopoiesis and the outcome of infection. Overexpression of IL-4 within the bone marrow in transgenic mouse models has revealed abnormal T-cell development (Tepper et al., 1990) and also an increase in eosinophils with enhanced phagocytic capacity (Sullivan et al., 1992).This suggests that alterations in cytokine concentrations within the bone marrow induced by microbial components may influence the growth and maturation o f cells involved in host defense. VIII. Host Genetics Influencing the Outcome of Infection
The impact of the genetic makeup ofthe host has been well illustrated by a study on an epidemic o f tuberculosis in an isolated population of Yanomami Indians of the Amazon rain forest. This population had no contact with tuberculosis prior to the early 1960s. This study showed an extraordinarily high incidence of active disease, correlating with elevated levels of M. tuberculosis-specificIgG4 antibodies and almost no skin test responses, even among BCG-vaccinated individuals (Sousa et al., 1997a). This emphasizes the importance of host genetic composition on resistance to tuberculosis and, by implication, the evolutionary pressure that microbial pathogens exert on the selection of the human race. It maybe speculated that such high susceptibility reflects the genetic situation at the beginning of a newly introduced pathogen causing epidemic disease as exemplified by tuberculosis in 18th century Europe and more recently by the worldwide spread of HIV.
A. NRAMP For many years it has been appreciated that resistance against intracellular pathogens is, in part, genetically controlled and that inherited factors can modulate the course of disease in infections with a broad clinical spectrum such as leprosy and tuberculosis. In particular these genetic traits can control the progression from infection to disease, which divides individuals into resistant and susceptible groups. Studies in the murine system have revealed a single, dominant autosomal gene on chromosome 1, originally termed Zty/Lsh/Bcg, which controls innate resistance to the unrelated intracellular pathogens, L. major, M . bovis BCG, and S. typhimurium (Blackwell et al., 1994; Zwilling and Hilburger, 1994). Following
IKTHACELLULAH BACTERIA AND TllE IMMIJNE SYSTEM
323
positional cloning and gene targeting, the Nrampl (natural resistanceassociated macrophage protein) gene was identified and shown to be allelic to Bcg (Vidal et al., 1995). This gene belongs to a family of genes that also includes Nramp2 located on mouse chromosome 15. Nrainpl encodes for a hydrophobic, membrane-associated protein expressed exclusively in phagocytic cells. In congenic mouse strains, a single glycine to aspartic acid substitution at position 169 of the fourth transmembrane domain is associated with the Bcg-susceptible (Bcg’) phenotype (Vidal et al., 1996). Conflicting reports exist as to what functional effects this mutation changes, with original data claiming that the amino acid substitution resulted in decreased expression of the molecule (Vidal et al., 1996),whereas more recent data suggest that there is equivalent expression in bone marrow macrophages from BALBlc mice (Bcg’) and CBA mice (BCG‘),indicating that the protein is present but nonfunctional (Atkinson et al., 1997). The exact function of the Nramp1 protein is as yet not fully elucidated, but the use of gene deletion mouse mutants and transfected cell lines have conclusively illustrated the pleiotropic effects on macrophage function that this gene controls. These include the production of ROI, NO, and TNFa from macrophages as well as the regulation of MHC class I1 expression and antigen presentation (Arias et al., 1997; Lang et al., 1997). Using an antibody generated against the C-terminal35 amino acids of murine Nrampl, confocal microscopy has localized this protein to late endosomesAysosornesas well as latex bead phagosomes, which would place the protein in close proximity to an intracellular pathogen (Atkinson et al., 1997; Gruenheid et al., 1997). Moreover, phagosomes containing live M . bovis BCG within macrophages expressing functional Nrampl fuse with V-H+ATPasecarrylng vesicles and subsequently acidify, in contrast to their mutant counterparts (Hackam et al., 1998). Furthermore, a high degree of sequence similarity to the yeast protein SMF-1 has indicated that Nrampl may function as a divalent cation transporter (Blackwell et nl., 1995). Mice with microcytic anemia have been shown to have a mutation in the closely related Nrainp2 gene, which is considered to encode for an iron transporter (Fleming et al., 1997). As many bacteria have an obligate requirement for iron for their intracellular survival (Wooldridge and Williams, 1993), this provides an explanation for an essential role of Nrainpl in resistance and opens up an area for potential therapeutic intervention. The human equivalents NRAA4Pl and NRAMP2 have been identified on chromosomes 2q and 12q, respectively. The gene for human NRAMPl has been cloned and sequenced, and polymorphisms and sequence variants have been described and used in linkage studies for susceptibility to tuberculosis and leprosy (Liu et al., 1995). Early evidence suggested that there was n o evidence for an LshlltylBcg gene hoinologue influencing suscepti-
bility to leprosy (Shaw et cd., 1993). However, a more recent large case control study from The Gambia, West Africa, comparing more than 400 tuberculosis patients with healthy controls, revealed that individuals heterozygous for two N R A M P l alleles were overrepresented among the tuberculosis cases (Belhny et al., 1998). In the murine system, data by Medina and North (1996, 1998; Medina et al., 1996), studying a number of inbred and recombinant mouse strains infected with virulent M. tuberculosis both iv and by aerosol, indicate that Nrampl polymorphisms have no influence on the outcome of infection. In this study, the fact that mouse strains homozygous for the Nrampl susceptibility allele were more resistant than those expressing resistant alleles seems more than coincidental, although as yet this finding cannot be explained. At least it can be concluded that genes other than Nranipl strongly influence susceptibility to experimental tuberculosis infection in mice (Medina and North, 1998). This is in line with data demonstrating that resistance to M . tuberculosis in mice only becomes evident after the onset of specific immunity, i.e., in the lung 30 days postinfection (Nadeau et al., 1995).
B. MHC Segregation analyses in human populations have indicated HLA linkage with various disease states. In terms of intracellular bacteria, most studies have analyzed linkage in leprosy and tuberculosis. Blackwell and colleagues ( 1997) analyzed the irnmunogenetics of both leishmanial and mycobacterial infections. This family study revealed that there was linkage to and allelic association of HLA molecules with leprosy, but not tuberculosis, which is consistent with what was found in a study population in India (Ghosal et nl., 1996). However, a population in Cambodia showed association of an HLA-DQ allele with clinical tuberculosis (Goldfield et al., 1998). Similarly, Rajalingam and colleagues ( 1997) identified an association of HLA-DR2 with mycobacterial disease. It was suggested that these specialized HLA alleles may preferentially present inycobacterially derived peptides. Supporting this notion is the observation that DR17 can only bind its specific peptide at neutral pH rather than at the acidic pH found in most peptide loading compartments (Geluk et al., 1997), which is the pH that the mycobacterial pliagosome maintains (Sturgill-Koszyclaet al., 1994).Studies on experimental tuberculosis in mice provided contradictory results. One study failed to show strong correlation with certain MHC genes, although some minor influence was observed (Medina and North, 1998), whereas in a different set of experiments, H-2k associated with resistance and H2” and H-2“ with susceptibility to tuberculosis (Apt et al., 1993). Although most linkage analyses have been performed for infections with a broad spectrum of clinical disease, there are reports of HLA segregation
in trachoma caused by C tracliowztis. In a study population in Oman, HLA-DR16 was associated with blinding trachoma whereas HLA-DR53 correlated uitli resistance (White ct d . ,1997).With more data now becoining available froin worldwide studies it is clear that HLA associations differ between population groups, making it difficult, or even impossible, to define a universal association for resistance and susceptibility, and underlining the influence of natural selection under the pressure of infection on M HC polymorphism. Within the MHC locus there are genes encoding for TNF and its receptors as well as genes for molecules of the antigen-processing machinery (proteasome subunits and TAP1,2), and these genes have also heen examined for polvmorphisms affecting disease states. In addition to HLA polyinorphisms,-there are indications that the polyinorphism of TAP molecules in rats modifies the spectrum of antigenic peptides presented by MHC class Ia m o l t d e s (Powis, 1997). Although inurine and human TAP molecules show only limited polyinoi-pliisni (Lobigs and Mullbacher, 1993; Yewdell cf a/ , 1993),correlation of certain TAP alleles with disease severity has been claimed in human tuberculosis and leprosy. The frequency of TAPZ-A/F’was increased in patients with pulmonary tuberculosis and that of TAP2-B9 in patients with tuberculoid leprosy (Rajalingam et nl., 1997). Despite the reported associations with M HC, an analysis in The Gambia of the relative contribution of MHC and non-MHC genes in the response to foreign antigens, including those derived from M . tzihcwulosis, indicated that genes outside the MHC lociis were more influential (Jepson et nl , 1997). In a different study, several cases of familial disseminated M . aviuin infections have been explained by a defect in antigen processing that diniinishes IFNy production. APCs from these patients can present influenza hemagglutinin peptides to T cells but are unable to process the native protein for presentation (D’Souza ct 01 , 1996).
R C. O T ~ I EGENES Resistance to sonie intracellular pathogens in the inurine models of infection are not influenced by tither N r m i p or MHC genes. A/J inice are an inbred strain that are niore susceptible to infections with both listeiiae and legionellae. A defect in the phagocyte inflammatory response caused by a deficiency in the C5 complement component was shown to be the major reason for the high susceptibility of these mice to infection with listeriae. Congeiiic inice sufficient for C5 were restored in their ability to control infection (Gervais et d., 1989). The natural resistance or susceptibility of inbred strains of mice to infections with L. pneumoplzitn is controlled by a single, dominant gene on chromosome 13, designated Lgii1. The phenotypic result of expression of this gene is the presence or
326
ULRICH E. SCHAIBLE rt al.
absence of intracellular replication of the bacteria inside host macrophages (Dietrich et al., 1995). Currently, a high resolution linkage map is being prepared of this region to accurately pinpoint and eventually clone the gene (Beckers et al., 1997). In an elegant set of experiments conducted by Murphy and colleagues, another locus has been identified that may influence host resistance to intracellular infections by controlling the T-cell response that is made. The genetic background of mice is an important determinant of IL-12 responsiveness and can influence the onset and severity of disease. CD4' T cells from a DO1l.10 ap TCR transgenic mouse were transferred into either BALB/c or B10.D2 mice, which both made equivalent levels of IL12. Analysis of the T-cell populations revealed that those transferred into the BALB/c mouse made increased IL-4 and decreased IFN-y, whereas the converse was true for the B10.D2 strain (Gorham et al., 1996). As both mice are congenic at the H-2 locus (H-2Dd),non-H-2 genetic loci must be controlling this effect. It has been shown more recently that the downregulation of the IL-12 receptor P chain in the BALB/c mouse but not the B10.D2 mouse strain is involved in this phenomenon (Gorham et al., 1997). A single dominant gene locus termed Tpml (T-cell phenotype modifier) controlling IL-12 responsiveness has been mapped on mouse chromosome 11 and on human chromosome 5 (Gorham et al., 1996). This region of the genome contains a dense cluster of genes of immunological importance, including IL-4, IL-5, and IL-13, as well as T-cell signalling molecules, and currently the molecular identity of Tpml remains to be determined. Consistent with the location of a resistance locus on chromosome 11, serial backcrosses of resistant B10.D2 mice onto the susceptible BALB/c background reveaIed several candidate loci, conferring resistance to L. major infection. However, despite one of these being located on chromosome 11, no single locus was required and the authors postulate that a variety of combinations of these loci may interact, resulting in resistance (Beebe et al., 1997). IX. Immune Intervention Strategies
Detailed knowledge about the biology of infections with intracellular bacteria and the immune response elicited not only allows us to dissect the complex cross talk between pathogen and host, it is also of basic importance for developing rational preventive and therapeutic intervention strategies. Although vaccination is the most successful prophylactic measure against infectious diseases in general, most vaccines currently in use are directed against viruses or extracellular bacteria. The success of these vaccines in principle rests on antibodies that either prevent pathogen
invasion o r neutralize toxins or virulence factors. To date, only hvo vaccines arc in use against intrncellular Iiwcteria-in both cases with only liinited success. The tuberculosis vaccine BCG is an ntteniiated strain derived from tlie virulent M . Iiouis strain, the agent of tuberculosis in cattle and occasionally in human Since its first iisc in 1921. BCC has been administered with lo\v side e ects :3 billion times \vorltl\vide, hut its protective efficacy is highly variable, ranging from 0 to 80% against puliiionaiy tiiberculosis (Ginsberg, 1998). Generally, the protecti6n achieved by BCG is 1iettc.r against niiliary and meningeal tuberculosis in children (46- loo%), but is inefficient against adult tuberculosis, which is usually caused by the reactivation of dormant hi. trihcrm1o.vi.s(Coltlitz rt d . , 1994; Fine, 1995). Increasing incidences of tulwrculosis in developing coiintries and a recent upswing in case numbers in the United States and eastern Europe since the mici-198Os are mainly caused Iiv tIrr HI\’ epidemic, by insufficient compliance wit11 tIie long duration of cIi(.inotlierapy, ancl by globa~uriiariization and pauperization. Moreover, Ad. triherculosis strains liave emerged that are resistant against most of thc cIieiiiotlierapeutic drugs c u r r e d y iwdal)le. This serious 1ie:iltli situation mikes the tlevelopment of an effective vaccine against tuberculosis a high priorit), in adjunct to iinprovetl chemotherapy. The other vaccine ciirrently employed against an iiitracellular bacterium is tlie attenuated S. enfcrictr serovar TfjphiTy21 strain. Becaiise of its low and sIiort-livecI protection, it is only aciministerec1 to travelers to areas where a high risk of coiitl-actingtyplioid exists. There is not only a need for more effective, inexpensive,and easy to apply vaccines against tuberculosis and typhoid fever, h i t also against otherworldwide liealth burdens caused by intricelhilar bacteria such as C. tiaclu,t,zriti.~,which represents the predoininant ciiiise of congenital blindness in developing countries. Rational development of a new generation of vaccines h a s to take into account the unique character of the immiine response that is effective against the respective pathogen. Accordingly, the requirements for such abaccine inc:lutle (i) induction of the “right” T-cell populations arid cytokine patterns that are rapidly mobiliztd and, at the same tinre, mediate protective and long-lasting (nienioiv) immunity; (ii) activation of the “right” effector mechanisins to prevent infection or to eliminate the infectioiis agent; (iii) exclusive specificity for tlie responsiible pathogen to avoid the risk of autoaggression through cross-reactive antigens; (iv) coverage of antigens that are expressed by the hacteria in the host and by various isolates or strains; and (v) iiiiiiiunogenicityiiiogeiricity fbr all MHC haplotypes within 1111111a11 pop111iit’1011s. First of all, a \roccine against an intracchlar bacterium should induce the appropriate combination of specific CD4’ and Cll8’ T cells that produce IFNy to subsequently activate macrophages to eliminate the bac-
328
UI.RICII E SCII4IULF: uf NI
teiia. Vaccines directed against some bacteria have to activate MHC class I-restricted CTL more profoundly than others. This is of particular importance for bacteria that parasitize nonprofessional APC that do not express MHC class I1 molecules or only at low densities. Additional strategies include antigens that stimulate unconventional T cells, such as glycolipids for CD1 reactive T cells and phospholigands for y 8 T cells. Unconventional T cells respond faster than conventional T cells and can therefore promote the generation of Th 1 cells. Glycolipids have the additional advantages (i) that they induce CTL, although they are derived from endo-/phagosomes, and (ii) that CD1 is a nonpolymorphic molecule. However, the overall contribution of unconventional T cells to the efficacy of vaccineinduced protection against intracellular bacteria remains to be determined. To rationally design an effective vaccine, the following aspects are important: i. the choice of antigens and their display and physicochemical nature (intracellular, surface associated or secreted proteins, glycolipids, phospholigands); ii. vaccine types (whole proteins or peptides, whole live attenuated bacteria, recombinant bacteria, naked DNA); and iii. formulations and ways of delivery (higMow doses, adjuvants, immunomodulatory cytokines, immunostimulatory DNA sequences).
Currently, various vaccine strategies against the intracellular bacteria discussed here are being tested with a focus on vaccines against tuberculosis and typhoid fever. In addition, some of these vaccines are also examined as carriers for heterologous antigens. A. IDENTIFICATION OF PROTECTIVE ANTIGENS
During recent years, an array of studies has been published describing proteins from various species of intracellular bacteria and their immunological relevance. Antigens that are released or actively secreted by live bacteria are recognized by the immune system early during infection when the number of degraded bacteria is low, whereas somatic proteins are only recognized after bacterial death and degradation (Kaufmann and Andersen, 1998). Therefore, most studies have concentrated on the identification of secreted proteins as candidate antigens for protection. The superior effectiveness of secreted vs somatic antigens for inducing protective immune responses has been demonstrated most convincingly in studies using either recombinant salmonellae or recombinant listeriae expressing defined antigens (Hess et al., 1996a, 1998; Shen et al., 1998). For M . tuberculosis, more than 200 proteins were found to be released into the culture supernatant, including antigens of potential relevance to protection (Andersen et
nl., 1991; Andersen, 1994; Sonnenberg and Belisle, 1997; Kaufinann and Andersen, 1998). Several studies found that mycobacterial proteins released into the culture supernatants can induce protective T cells (Hubhard ct d . ,1992; Pal and Honvitz, 1992; Andersen, 1994; Roberts et nl., 1995). Vaccination with secreted proteins (including the Fn-binding 30-kDa protein of the Ag85 complex) from M . triliercrilosis protected guinea pigs against aerosol challenge with tuliercle bacilli (Honvitz et nl., 1995). Moreover, Ag8rj and a 29-kDa antigen (CFP29) induce IFNy-secreting T cells in mice (Andersen et al., 1995; Rosenkrands ct nl., 1998). ESAT-6, a secreted 6-ltDa protein that is expressed exclusively in M . tiiherczilosis but not in hl. b i x i y BCG, induced T cells in mice tliat expressed a phenotype potentially associated with protection (Andersen ct d.,1995; Brandt et'crl., 1996; Harboe ef d., 1996). However, it appears that ESAT-6 given in different forins of vaccinations fails to induce satisfactory protection. Other mycobacterial antigens released into tlie culture supernatant, such as the proline-rich coiiiplex (Roinain cf nl., 1993), the chaperones a-crystallin and GroES (Young and Garbe, 1991;Verbon ct nl., 1992), SOD (Andersen et a / . , 1991; Andersen, 1997), MPTS1 (Wiker ct d . , 1992), MPT63, and MPT64 (Haslov et d . ,1995) still await further analysis with regard to their protective potential. Still, these proteins encoinpass the most promising vaccine candidates as dead mycohacteria are insufficient to induce a protective immune response (Orine, 1988). It should be noted that none of the antigens tested proved to be superior over BCG and that the final outcome of such studies will most probahly show that not all secreted proteins are protective. Enormous effects have been put into tlie analysis of the total protein spectniin of various bacteria. A number of groups are working to establish the proteomes of intracellular bacteria, such as M . ttdw-culasis, M . bovis BCG (Sonnenberg and Belisle, 1997; Urquhart et d., 199'7; Welding11 et d . ,1998; Jungblut ct a/., in preparation), S. trplzimurizrni (Qi et d . ,1996; O'Connor ct d.,1997; Burns-Keliher ut d.,1998),Brcicdn ouis, B. melitcrrsis (Teixeira-Gomes et al., 1997a,b),and C. trnclioimtis (Bini et d., 1996). These efforts benefit from, and compleinent greatly, ongoing genome sequence iinalyses of various bacterial pathogens, including M . tnherctiZo.sis, which has been announced to be contiguous (Cole et NZ., 1998). Apart from identification of protein patterns exl&ssed under various conditions, proteoine analyses will also facilitate the determination of posttranslational modifications such as acylation, glycosylation, and phosphorylation of the respective proteins.
B. ATTENUATEDA N D RECOMBINANT VAC:CINE STRAINS Use of live-attenuated bacteria as vaccines has the benefit that the organisms replicate in the host for at least sonie time and therefore gener-
330
U L R I C H E SCIIAIBLE rl
(I/
ally induce long-lasting specific immunity. Moreover, they resemble the fully virulent pathogen to a certain extent and thus may induce a similar iinmune response. Such vaccines currently in use encompass those against tuberculosis, M. bovis BCG, and typhoid fever, S. enterica serovar Typhi Ty21. As mentioned before, M. bovis BCG has been used worldwide as vaccine against tuberculosis and leprosy, although with variable and often low efficacy. Several hypotheses have been brought forward to explain the variable outcome of M. bovis BCG vaccination. These include (i) exposure and sensitization through environmental mycobacteria, which vary greatly between geographical regions (Stanford, 1991; Fine, 1995); (ii) lack of important genedantigens in BCG that are expressed exclusively in virulent mycobacteria (Harboe et d ,1996; Mahairas et al., 1996); and (iii) insufficient induction of CD8+ T cells by A4. bovis BCG, thus excluding an important iinmune mechanism for protection against tuberculosis (Muller et d., 1987; DeLibero et d.,1988; Flynn et al., 1992; Bonato et al., 1998). The latter obstacle may be overcome by the generation of recombinant M. bovis BCG expressing the listerial LLO, which facilitates targeting of BCG antigens into the MHC class I pathway (Hess et al., 1998). This construct will be improved further by including genes encoding M . tuberculosis-specificproteins, with the ultimate goal of targeting protective antigens to both MHC class I and I1 pathways. The abundance of glycolipids with inflammatory properties in the rnycobacterial cell wall and the high proportion of CpG motifs in the mycobacterial DNA most probably further promote a Thl type response. Recombinant vaccine strains such as reconibinant M. bovis BCG expressing heterologous antigens have already been used in vaccine studies against other infections: recombinant M. bovis BCG expressing the surface cysteine protease gp63 of L. nzxicana or the outer surface protein A of Bovelia burgdu$eri can protect against cha1lenge with the respective pathogen (Connell et al., 1993; Stover et al., 1993; Langermann et al., 1994). Another approach aims to identify virulence and survival factors for the generation of KO mutants by genetic deletion. These attenuated strains fail to thrive in the host but retain their immunogenic potential and may be considered safe vaccine strains. Such KO bacteria potentially encompass those lacking metabolically essential genes, i.e., auxotrophic mutants, which die in the host due to a lack of essential nutrients. Using transposon mutagenesis, auxotrophic M. bovis BCC mutants have been generated that are attenuated but still able to induce an immune response comparable to wild-type BCG (McAdam et al., 1995; Guleria et al., 1996). These vaccine strains may provide a safe way to immunize immunodeficient patients, which may develop generalized BCG infection on vaccination. Self-limiting vaccine strains such as the suicide L. nuinucytogenes strains
could proiide a solution to this obstacle because they survive long enough to retain their high iininunogenicity but ultimately die after escape into the cytoplasm (Dietrich ct nl., 1998). Auxotrophic S. trjpliiniuriimi AroAmutants, wliicli were generated by transposon deletion inutagenesis and have a defect in aromatic biosynthesis, inducc protective immunity against salmonella infection in mice (Hosieth and Stocker, 1981; Newton et al., 1989; Horniaeche et 0 1 , 1991; Harrison et a/ , 1997).Attenuated recombinant salmonellae have been genetically engineered to actively secrete various antigens by virtue of the E coli secretion apparatus. These recoinbinant salmonellae have been shown to deliver foreign proteins into the MHC class I and 11 pathway (Hess et nl., 1996a, 1997). Virtually every heterologous antigen could be introduced into this system, probahly also in combination with other antigens. Similarly, recombinant listeriae can also be used as vaccine vehicles (Jensen et (11, 1997; Shen et al., 1998). The recoinbinant salmonellae or recombinant listeriae vaccine carriers have tlie fiirther advantage that they can be administered orally. A different and probably more promising approach is the genetic deletion of certain virulence factor genes, which will diminish the intracellular sui-vival rate of the respective bacteria. By employing transposons tagged with a unique DNA sequence, genes of S. typhiniuriuui were identified that contribute to survival of the salinonellae in mice (Hensel et nl., 1995).This experimental system is currently also applied to other intracelliilar pathogens. However, this approach may be hindered by tlie fact that the capacity to survive intracellularly is based on many factors/genes. Based on the current knowledge of the cytohnes important for controlling intracellular bacteria, i.e., IL-12, GMCSF, and IFNy, recombinant vaccine strai 11s have been constructed that express these cytokines alone or together with defined antigens. Thus, recombinant M . h i s BCG strains expressing IL-2, IFNy, and GMCSF werc found to be superior to wildtype BCG in inducing protection in mice (Murray et d., 1996). Similarly, attenuated recombinant salmonellae exliressnig MIF, TNFa, and IFNy were used siiccessfully to direct the inimune response against L major in susceptible BALB/c mice toward a Th1 type T-cell response (Xu et d., 19981)).Although this approach implies that cvtokines secreted from intracellular bacteria are trafficked to the responsive cells without being degraded, some bacteria may also survive outside of their host cells to release sufficient concentrations of cytokines into the extracellular milieu. C.
S U B U N I T VACCINE5 A N D ADJU\’ANTS
The benefits and drawbacks of‘ purified native or recombinant proteins as vaccine antigens are obvious: These vaccines are highly specific and bear a low risk of side effects; however, when given alone, their immuno-
332
ULRICII E SCHAIBLE r f
n/
stiinulatory and protective efficacy is generally weak and short-lived. For example, after immunization of mice with a mixture of culture supernatant proteins from M . tztbercrrlosis in adjuvant, at best a protection level comparable to BCG vaccinations was achieved. However, BCG-induced protection was long lasting whereas protection induced by the subunit vaccine faded after 5 months (Roberts et al., 1995). Therefore, subunit vaccine formulations depend strongly on inimunostimulatory adjuvants that siniultaneously create the milieu required for the stimulation of protective T cells and generate a reservoir for slow antigen release at the site of application to facilitate long-lasting protection. Alum, the only adjuvant approved so far for use in humans, preferentially induces a Th2 response whereas adjuvants stimulating Thl cells are required for vaccination against intracellular bacteria. Currently, most of these adjuvants are hampered by the fact that their capacity to stimulate a Thl response is accompanied by severe inflammation resulting in tissue destruction. Bacterial glycolipids and oligonucleotides (see later) may be considered “natural” adjuvants due to their proinflammatory properties. As described in Section 111, LPS, LAM, and other bacterial glycolipids are strong inducers of cytokines triggering a Th1 response. Dissecting the adjuvant-associated stimuli that induce inflammation from those that elicit Thl cell responses would strongly improve subunit vaccine strategies. A number of adjuvants have been developed with reduced inflammatory potential but retaining the capacity to induce antigen-specific Thl responses. These include liposoines, microspheres, squalene, and dimethyl deoctadecyl ammonium bromide. Moreover, some adjuvants are endowed with the capacity to introduce antigens into the MHC class I pathway to trigger CD8’ T cells. Of these, iinmuriostiinulatory coinplexes (ISCOM) represent the most advanced adjuvants (Takahashi et nl., 1990; Heeg et nl., 1991; Morein et nl., 1996). Moreover, cytokines, i.e., IL-12, IFNy, or GMCSF, contained in the vaccine formulation can direct the immune reaction toward a Th1 type response (Bermudez and Kaplan, 1995). Peptide vaccines comprising defined protective epitopes while bearing the lowest risk of cross-reactivity because of their unique specificity have the disadvantage of being the least immunogenic. Due to the different peptide motifs defined by the various HLA haplotypes in the human population, only a vaccine encompassing a mixture of antigenic peptides will cover the repertoire of T-cell epitopes present in a heterologous popillation o f vaccinees. Such peptide mixtures are chemically linked during synthesis to form linear or branched multipeptide complexes. To induce an immune response, peptides or complexes thereof are linked to carrier proteins and strong adjuvants have to be incorporated. As a corollary, it is not surprising that up to now only one peptide-based vaccine has been
proven to induce sufficient protection, nainely that against foot and mouth diseuse in cattle (Ada, 1990). The finding that T cells elicited by synthetic peptides can recognize peptide-piilsed APC but not APC loaded with tlie complete protein, which present naturally processed peptides, adds to the arguments against peptide vaccines (Viner et crl , 1996). A different approach is Ixised on immunization with a combination consisting of antigenic peptides and certain HSP inolecules. Vaccination with HSPSO plus peptides has been sliown to mduce protection against challenge with LCMV (Ciupitu ct 01, 1998) and, siinilarly, application of HSP’iO and gp96 plus peptides protected mice against tumor growth (Udono and Srivastava, 1993; Suto, 1995).Although the underlying inolecular mechanisms are not fully understood, it is assumed that HSP preferentially bind peptides for M HC class I processing and subsequently introducc thein into tlie MHC class I processing pathway. Similarly, fusion proteins comprising HSP70 plus HIV1324 or an antigenic ovalb;imin-clerived peptide (SIINFEKL) induced specific iinniune responses including CTL to tlie respective antigens (Suzue and Young, 1996; Suzue et nl., 1997). The feasability of this approach to vaccination against intracellular tiacteria needs further exploration. D. N A ~ E D D N A The latest achievement in vaccine developinent is the direct application of naked plasmid DNA encoding the geneis) of tlie respective antigen(s) (Tang et nl., 1992; Tighe ct 01, 1998). Plasinid DNA is delivered by direct intramuscular injection, by use of a balistic device to “slioot” DNA-coated gold particles into the skin (gene gun), or by means of bacterial carriers (see later) Gene expression is usually under control of a strong viral promoter, siicli as tlie CMV promoter, wliich allows expression of encoded genes by the transfected inamnialian cells, most probably muscle cells. Integration of a secretion signal into tlie plasmid allows export of the antigen by the cell. Subsequently, the iuntigen can stimulate B cells and can also Le taken up by infiltrating nionocytes and dendritic cells, which will then stimulate T helper cells. This vaccine type is safe and easy to produce, store, and administer. So far, DNA vaccination has been einployed successfully in various experiinental infkction models such as influenza virus (Ulmer et al., 1993), Mycoplnsitici p i h u m i s (Barry et a1 , 1995), L. niiljor (Xu and Liew, 1995), an‘d B.hrirgtlorferi (Zliong et d . ,1996; Luke ct nl., 1997). In tlie case of M . pubtionis, expmsion libraries and sublibraries were used for immunization (Barry et nl , 1995). To screen a whole set of antigens at once represents an interesting approacli that is currently applied to other bacteria such as mycobacteria. To date, three studies have demonstrated that DNA vaccination with mycobacterial genes induces a protective iinmune re-
sponse against M ticbercttlosis in mice comparable or only slightly lower as compared with BCG. The antigens used were HSP6O and the 36-kDa proline-rich antigen from M . leprae (Tascon ct d., 1996), the secreted M . tiiberczdosis protein Ag8SA (Huygen ct al., 1996), and the 38-kDa glycolipoprotein (Zhu et al., 1997). In these studies, protection correlated with the induction of high levels of IFNy producing cells and CTL. Moreover, one report showed that a DNA vaccine based on tlie porin gene 0mpC of S. typliimuriuin induces a humoral immune response in mice (Lopez-Macias et d., 199s; reviewed in Sti-Lignell ct al., 1997). Following DNA vaccination, the inflammatory reaction induced by distinct bacterial DNA sequences, such as the CpG-motif, with immunostiniulatory potential promotes tlie infiltration of nionocytes. Similarly, inclusion of genes encoding cytokines and costirnulatory molecules into the vaccine plasmid can promote the appropriate irninune response as shown in the influenza system:The capacity of a plasmid containing the influenza nucleoprotein to stimulate CD8’ T cells wiis improved successfully by the addition of genes encoding IL-12 and GMCSF and the costiinulatory molecules B7-1 and B7-2 (Iwasaki ct al., 1997). Knowledge about the iinmimoinodulatory effect of certain DNA sequences came from early studies trying to identify the molecular entities of M . houis BCG, which are responsible for nonspecific tumor rejection after treatment of tuinor-bearing mice with BCG (Kataoka et al., 1992). The isolated component was identified as BCG-derived DNA, which appeared to be a strong inducer of N K cells and interferons a, b, and y (Sliiniadaet al., 1986; Mashiba et nl., 1988;Yainainoto et nl., 1988; Pisetsky, 1996). The responsible DNA sequences were found to contain at least one palindrornic stretch of the motif 5’-purine-purine-CG-pyrimidinepyriinidine-3’ such as GACGTC, GCXGCC, AGCGCT, and AACGTT (Tokunaga ct nl., 1992; Yamamoto ct d., 1992, 1994). Although the high G/C content in inycobacteria favors abundance of CpG motifs in M . houis BCG, this phenonienon is obviously not restricted to M . bouis BCG-derived DNA, a s DNA from E. coli also induces IFNy via IL-12 and TNFa (Halpern et nl., 1996).Therefore, certain properties of bacterial DNA must be distinct from mainrnalian DNA. The CpG motif occurs more frequently in bacterial (1/16 bp) than in inaininaliaii DNA (1/SO bp), with cytosins being less inethylated in bacteria (<SO%) than in mammals (70-90%) (Pisetsky, 1996;Tighe et d., 1998).The immunogenicity of a DNA vaccine was reduced significantly by methylating its CpG motif (Klinman et al., 1996). CpG containing immunostirnulatory DNA sequences (ISS) have been shown to be strong inducers of IL-6, IFNy, IL-12, and IL-18 and subsequent Th1 and CTL responses i15 well as high antibody titers dominated by the IgG2a isotype (Klinman c>tnl., 1996; Sato et NI., 1996; Lipford
nl., 1997: Koman rt d., 1997).This is most probably due to the infiltration of potent .4PC siich as dendritic cells and niacrophages to the site of intriidernial injection. Moreover, it lias been ~ O L I that I ~ ~ CpC: DNA sustained IL-12 production, thus promoting nonspecific resistance to L. tjwnocgtogcnes infection (Krieg ct nl., su1)mitted). The Th 1-dominated effect seen after intradernial DNA iiijectioii is quite opposite to the outcome of vaccine adininistratioii using a ixilistic device, wliicli ratlier leacis to a TMlike response (Feltquiite ct NI., 1997). It has been proposed that the 100fold lower amount of DNA iised in the gold particle delivery system compared to direct DNA injection l e d s to a dilution of the ininiunostimiilatory DNA sequences, thiis reducing any inflammatory reaction that results in TI11 cell activation (Tighe et a[.,1998). It has been reported that bacteriiil \rectors can be used as transport veliiclcs for recombinant DNA. Tliis vxcinatioii stratea opens the possibility of oral immunixatioii and will therefore additionally activate inucosal immune responses. So far, attenuated recombinant S. jexrieri ( Sizeinore ct d . , 199511, recoinbinant S. t z j p h i t t u i r i r i i j i (Darji rt nl., 1997), and suicide recoinbinant L. I,iotiocglojiciies ( Dietrich r f nl., 1998) liave been used in these qqxwiches. Taken together, a vaccine based on genes of iminunodoininai~tbacterial antigens probably coupled to cytokine ancVor costiniulaton~molecules inducing Th1 responses represent a promising candidate to control infections with intraettllular bacteria. Althongh there were only a few reports that upon DNA vaccination the foreign DNA has been integrated into the host genome (Doerfler et d., 1997; Dietrich et d., 1998), future studies on DNA vaccination have to take these safety aspects into account. Nevertheless, DNA vaccination represents a potent approach to battle infections with intracelliilar bacteria. ef
336
UI.Hl(:H E. SCHAIBLE rt ol.
REFERENCES Ahhas, A. K., Murphy, K. M., arid Shrr, A. (1996). Functional diversity of helper T lynphocytes. Nature 383, 787-793. Abou-Zeid, C., Hatliff, T. L., Wiker, H. G., Harboe, M., Bennedsen, J., and Rook, C . A. W. (1988).Characterization of fibronectin-binding antigens released by Mycobncterium tuberculosis and Mycohcteriurti both BCG. Infect. I m m z i r i . 56,3046-3051. Abu Kwaik, Y. (1998). Induced expression of the Legionellrr pneumoplrila gene encoding a 20-kilodalton protein during intracellular infection. Infect. Inmiun. 66,203-212. Abu Kwaik, Y., Eisenstein, B. I., and Engleberg, N. C. (1993). Phenotypic modulation by Legionella ~~ntrc~nopliila npoii infection of macrophages. Iifect. Ittiniun. 61,1320-1329. Ada, G.L. ( 1990). The iinmiinological principles of vaccination. Lancet 335,523-526. Agganval,A.,Kuinar, S.,Jaffe, H., Hone, D., Gross, M., and Sadoff,J. (1990).Oral Salmonella: Malaria circumsporozoite recombinants induce specific CD8' cytotoxic T cells. J. Exp. Med 172, 1083-1090. Ahearn, J. M., and Fearon, D. T. (1989).Structure and function ofthe complement receptors, CR1 (CD35) and CR2 (CD21). Adu. Irtimurio~.46, 183-219. Albert. M. L., Sauter. B., and Bharclwaj. N . (1998). Dendritic cells acquire mtigen from apoptotic cells and induce class I restricted CTLs. Nature 392,86-89. Altlrich, C. J.. DeCloux, A,, Woods, A. S., Cotter, H. J., Soloski, M. J., and Forinan, J. (1994). Identification of a Tap-dependent leader peptide recognized by alloreactive T cells specific for a class Ib antigen. Cell 79, 649-658. Alpuche-Aranda,C. M., Berthiauine, E. P., Mock. B., Swanson,J. A,, and Miller, S. I. (1995). Spacious phagosonie formation within niouse macrophages correlates with Salmonella serotype pathogenicity and host susceptibility. Itfect. Iminun. 63,4456-4462. Alpuche-Aranda, C. M,, Racoosin, E. L., Swanson. J. A,, and Miller, S. I. (1994). Salmonella stimulate macrophage macropinocytosis and persist within spacious phagosonies. 1.Esp. Med. 179,601-608. Alpuche Aranda, C. M., Swanson, J. A,, Loomis, W. P., and Miller, S. 1. (1992). Salinnnielh typhimuriuni activates virulence gene transcription within acidifed macrophage phagosomes. Proc. Nat. A c Sci. ~ USA 89, 10079-10083. Alhire, F.,Durandy, A., Emile, J.-F., Lamhametli, S., Le deist, F., Laminas, D., Dysdale, P., Jouanguy, E., Doffinger, R., Bernaudin, F., Jeppsson, O., Segal, A. W., Fischer, A,. Kuniararatne. D.. and Casanova. J.-L. ( I 998). IL-12 receptor deficiency selectively impairs rnycobacterial irnmunity Sciciice 280, 143.5-1438. Alvarez-Dominguez, C.,Barbieri, A. M., Beron, W., Wandinger-Ness, A,, and Stahl, P. D. (1996). Phagocytosed live Listeria nionocytogenes influences Rab5-regulated in vitro phagosoine-endosoine fusion. J. B i d . Ch.em. 271, 13834-13843. Alvarez-Dominguez,C.,Roberts, R., and Stahl, P. D. (1997). Internalized Listeria nronocytogenes modulates intracellular trafficking and delays maturation of the phagosome. J. Cell Sci. 110,731-743. Andersen, P. (1994). Effective vaccination of mice against Mycobucterizun tuberculosis infection with a soluble mixture of secreted mycobacterial proteins. Infect. Immrn. 62, 2536-2544. Andersen, P. (1997). Host responses and antigens involved in protective iminunity to Mycobacterium tuhercuZosis. Scand. /. Irnniunol. 45, 115-131. Andersen, P., Andersen, A. B., Sorensen, A. L., and Nagai, S. (1995). Recall of long-lived irnmunity to Mycobaderiuni tzibercldosis infection in mice. 1.Irnnirrnol. 154,3359-3372. Andersen, P., Askgaard. D., Ljungqvist, L, Bennedsen, J., and Heron, I. (1991). Proteins released from Myc(~bncteriumtubercubsis during growth. Infect. bnrnunol. 59, 19051910.
INTHACELLUIAH RA(:TF:HI.A AND THE IMhllJR'E SYSTEhl
337
Aiitoine. J. C.. J ~ ~ i i i i i i i C., e , I,ang, T., Prina, E., de Cliastellier, C., and Frchel, C . (1991). Localizatioii of major 1iistocoinpatil)ilitvcoiiiples class II nrolecules i n phagolvsosomes of niiirine nimup1iage.s infrctetl \tit11 I , ( + d t t r m i o ~ l t t t m l t t f ' J I , Y i . S .It!/&. I t t t r r t i r t i . 59, 764-775 Appellierg, R. Castro, A. G.,Pedrosa, J.. ;urd Miiiopiio. P. (1994). Role of interleiikin-6 iri tht. induction of protecti\v T cells thiiing iirycolxictrrial infections in mice. Ir~triiirtioZogy 82, 361-364. Appelberg. R., and Orme, 1. M. (1993).Effcxtor iiic&misnis involved in cytokine-rrictliated IYl..,. L ri- :iostiisis i ) f nl!/")~(ic.tc.r-iirri~m : i i t t t t iiilrc.~ioiisin iiniiiinc~niiacrrqihiigv~.Irtivizrtt&)gy 80, 352-359. Apt. A. S., Adienko, V. G.. Nikonenko. B. \'.. Kniiiinik, I. B.. Moroz, A. M.,iiiitl Skanienr, E. (1993).Distinct II-2coniplex control of mortality, iirrd iniinuiie responses to tul)crciilosis iiifrction in \irgin arid BCG-v;icciiiatrtf mice. Clitt. Esp. f r r i t r i i r t i ~ ) l .94, 3322-329, Arasr, H., Arue. N., Nakagawa, K.. Good. R. A , . and Oiroe, K. (1993). NKl.l'CD4'CD8thyniocytes with specific^ lynipliokine secretion. Err,.. /, Z r r t t r i i t t i o l . 23, 307-310. Alias, M.. Rojas, M.. Zahaleta. J., Rodiigiiez. J. I., Paris, S. C.. Barrera, L. F., and Giircia, L. F. (1997). 1nliil)ition of virulent 12lyeijbNc,tc.r~rtititic/wrcrtZo.si.rby Bcg(r) and Bcg(s) macrophages correlates with nitric oxide productioii. /. Ir!fecf.Di.v. 176, 1552-1.558. Armstrong. J. A,, and D'Arcy Hart, P. D. (1971). Response of cultured niacrophages to Mycobcrc.trt-iirrtItitbcr~dosis,with ohenations on f i i k i o i i of lysosomes with phagosomes. 1. Esp. Afcrl. 134, 713-740. Arnistrong, J. A,, a n d Hart, P. D. (197.5). Phagosoilre-lvsosoiri~ interactions in cultured macrophages infected with virulent tril)ercle bacilli: Reversal ofthe usuiil nonfiision pattern aiitl olxeiv;itioiis OH hacterial s u n i d /. Exp. hftv1. 142, 1- 16. Arnold, It.. and Konig. W, (199'3). Iiitrrlrukiir-8 release from huiriaii iiriitropliils after phugocytosis of Listc,ricr rnotic)cylogfric,s and Y(Jr.sitiioeritc,rocoliticci. /. M i d . Microbial. 47, 55-62. Athison, P. C.. Blackwell. J. M., ant1 Bartoii, C. f1. (1997. Nrampl locus encvcles a 65 kDa iiiterferon-y-iiitlricihle proteiii i i i niurinc niacrophages. Bioe/wni. 1. 325, 779-786. Augustin, A,. Kubo, R. T., and Sini, G. K. (1989).Resident l~iilmona~lyinphoc~Zes cxpressiiig the yl6 T cell receptor. Natirt-c, 240, T39-241. Autenrieth, I. B., Berr, M., Bohii, E., Kaufinaiin, S. El. E., and Heesemann, J. (1994). Iniinune responses to Yersinirr oitcrocolitictr in siisceptible BALB/c and resistant CSSBW 6 mice: An essential role for ganiina interf(.ron. Itfict. I t t i t t i i t t i , 62, 2590-2599. Bachmann, M. F., Lutz, M. B., Laytoii, C. T., llarris, S. J.. Fc>hr,T., Rescigno, M., and Ricciardi-C'istagiioli. P. ( 1996). Dendritic cc~llsprocess exogenous viral proteins and c i m s like particles for class I preserrtation to CDH' c),totodc T lymphoc\tes. Eirr. J. IrttrriittioZ. 26, 2595-2600. Balaji. K. N..and Boom, \V, H . (1998). Processing of h.l!yco/xrcfen'rrtn tuhrrcirkxis bacilli hy human iiionocytes for CD4' a/3 and y6 T cells: Role of particdate antigen. Irfert. 1 ~ 1 1 ~ t i i ~ t 66, i . 98-106. Balaji, K. N., Schw;irider. S . K., Rich, E. A., and Boonr, W. 11. (1995).Alvcolar iiiacrophages a s accrssory cells for hrinian y6 T cells activated hy A.i!/cobncteritrtri tnbcrcdosis. /. Z t r i t t i i t r t d . 154, 5959-5968. Halk. S. P., Burke, S.. Polischuk, J. E., F r a n k M. E.. Yaiig, L., Porcelli. S.. Colgan, S. P., and Bluiiiberg. R. S. [ 1994). Beta ~-microglohiilin-independentMHC class 111 molecnle expressed by Iinnian intestinal epithrlium. Scic~itc.~ 265, 259-262. Balk, S. P., Ehert, E. C.. Hliiinenthal. R. I.,., McDermott, F. V., Wucherpfennig. K. W., Landau. S. B., arid Hlumberg, R. S. (1991). Oligoclonal expansion and C D l recognition by Iiiiiiiiiii intestinal intrac~pithelialIyiiiphoc)tes. Scirtlce 253, 1411-141 5.
Bancroft, G. J.. Schreiher, H. D., antl Uiiiinue, E. H. (1991). Natural innniiiiity: A T-cellindepeiident pathway of rnacrophage activation, defined in the scitl iiioiisc*. Z r t i i r i i r r i o l . RCL. 124, 5-24. Baorto, D. M., Gao, Z., Malaviya, R.. Dustin, M. L., van der Mcnve, A., Lublin, D. M., and Abraham, S . N . (1997). Survival of FimH-expressing enterobacteria in macrophages relies oii glycolipid traffic. Nature 389, 636-639. Barker, L. P., George, K. M., Falkow, S.. :ind Sinall, P. L. (1997). Differential trafficking of live aiitl d c d M!/c&zctsrirrrrt nwrinurri organisms in inacrophagrs. 1nfc.c.i. Z r r u r i r t > i . 65, 1497-1504. Barnes, P. F., Foiig, S . J., Brennan, P. J., Twomry, P., Mazumtler, A., and Modlin, R. L. ( 1990). 1,ocal production of tumor nccrosis fietor and IFN-y in tuberculous pleuritis. 1.~ I t l ~ t l ~ 145, ~ ~ I ~149J ~ . 154. Barnewall, R. E., Hikihisa, Y., arid Lee, E. [ I. (1997). Elirlicliia chnfJerisis inclusions are early cndosomes which selectively xcuninlate transferrin receptor. Irfect. I r i i r r i i t r i . 65, 1455-1461. Barly, M . A,, Lai, W.C., antl Johnston, S. A. (1995).Protection against mycoplasma intection using expression-library iiiiiiiunization. Nntrrrt, 377, 632-63-5. Barzu, S.. Benjelloiiii-Touinii, Z., Phalipon, A,, Sansonetti, P., and Parsot, C. (1997).Fuiictional analysis of the Shigdln Jurwri IpaC invasin by insertional niutagenesis. Itfict. f r r i t r i u u . 65, 1599- 160lj. Bauer, A,, Huttinger, R., Staffler, G., Hatismann, C., Schmidt. W., Majtlic, 0.. Knapp, DJ., antl Stockingrr, H. ( 1997). Analysis of the rrquirement for beta 2-inicroglobulin for expression and formation of huniaii C13 1 antigens. Brr. J . Z m t m t i o l . 27, 1366- 1373. Beaiiregartl. K . E., Lee, K. D., Collier, H. J.. and Swanson, J. A. (1997).pl-l-dependent perforation of macrophage phagosomes by hsteriolysin 0 from Listerin nionocytugcricx J. Erp. M i d . 186, 1159-1163. Beckers. M . C , Ernst, E., Diez. E., Morisscttca, C., Gewais, F., Hunter, K., Housman, D., Yoshida, S.. Skaniene, E., a i d Gros, P. (19971.High-resolution finkage map of nioiise chroinosome 13 in the vicinity of the host resistance locus Lgtil. Gerioniic,s 39, 254-263. Beckman, E. M., Melian, A. M., Behar, S. M., Sieling, P. A,. -Chatterjee, D., Matsmnoto, R., Rosat, J.-P., Modlin, R. L., and Porcelli, S. A. (1996). CDlc restricts responses of niycobacteria-specific T cells: Evidence for antigen presentation by a second member of the human CD1 family. J. Irritnirnol. 157, 280:3. Beckman, E. M., Porcelli, S. A., Morita, C. T., Behar, S. M., Furlong. S. T., and Breniier, M. B. (1994). Kecogiiition of a lipid antigen by CD1-restricted &3+ T cells. Natrcru 372,691-694. Beebe, A. M., Mauze, S., Schork, N. J., and Coffinan, R. L. (1997). Serial backcross mapping of multiple loci associated with resistance to Leishrrutnici riuqor in mice. frrimmiii!/ 6, 55-557. Belisle, J. T., Vissa, V. D., Sievert, T., Takayama. K., Brennan, P. J.. and Besra, G. S. (1'397). Role of the major antigen of Mycohncteriwri hibPrcdcwis in cell wall biogenesis. Science 276, 1420-1422. Bellainy, R., Ruweude, C., Corrah, T., McAdain, K. P. W. J., Whittle, 14. C., and Hill, A. V. S. (1998).Variations in the NRAhlPl geiic and susceptibility to tul)erculosis in west africans. N . Etigl. J. Med. 338, 640-644. Belles, C., Kiih1, A. K., Donoghue, A. J., Sano, Y., O'Brien, H. L., Born, W.. Bottoinly. K., and Carding, S. R. (1996). Bias in the yS T cell respoiise to Listerin nionoc~togenes: Vy%.3+ cells are a major coinporrent of the 76 T cell response to Listerin rrumoc!yfo~eties. J . I n i n i w i d . 156, 4280-4289.
Bellonr, M., Iezzi. (;., Ro\we, P.. Galati, C., Rochetti. A , , Pia Protti. M . , Davoust. J., Rngiirli, (:., ; i i d hlaiifretli, A. A . (1997). Processing of riigiilfid apoptotic I d e s ~ieltls T cdl cyitopes. J . I t t t t t t ~ ~ t t o /159, . 5591-53399. Bericlelac, A.. Lairtz. O., Qiiiiiih!,. M. E., Yewtlrll. J. \f7.. Bennink, J. R.. aird 13rutkiewvic.z. 1%. R. (1993). CD1 recognition hy nioiise N K I T Iyiiphocytes. Scimcc 268, 86:3-865. Heiidc~lac.A,, Rivera, M. N.. Park. S. 13.. and Koark. J. 13. (1997). Moiise C:Dl-spc~cific NKI T cells: Development. specificih. iind finictioii. Attttu. H(T. Z t t t t t i u t t o l . 15, 535-562. Hermiidez. L. E.. and K q l a n . C.(19953. Reconiliinant c!-tokiiies for controlling iiiycolxicterial infections. Trenrls L\ficrobiol. 3, 22-27. Berniudez, I,. E.. Parker. A , , urid Gootlman. J. 11. ( 1997). GrowtIi \iitlrin macrophagc.s incre:scs tlie (4ficiwcy of nf!t'."bo'.lf,riurti oriri t t i i n invading otlier niacropliages by a coiiipIcwciit recrptor-intlepentleiit patliwa!. It+c.t. Z t t t t t r t i t i . 65, 191~-1925. Bvrna, A., Kopf, M., Kul1)acki. R., Ll'eich. N.. Kdiler, C;.. and (:iittierez-Haiiios. J. C. (1994). Interleiihin-6is reqiiircd in \ivo for tht, rrgul;ition of'stein cells and conimittecl progcwitors of the Ixieniopoietic system. Z t t i t t i t i t t i t ! / 1, 72.5-7:31. Beron. W.,i~lvarrz-Doiiiingii(.z. C.. hluyorga. I,., and Stahl. P. D. (1995). Menilmne trafficking h n g the phagocytic piitlway. Tretrdy Crll Riol. 5, 100-104. Beutler, H.. and C;el.ami, A. (1987). Caclic,ctin: More thiur a tumor necrosis factor. .V. Eirgl. /. M d . 316, 379-3385. Bini, L.. Saiicliez-CainpilltJ, M., Santiicci. I\., Magi. €3.. Marzocclii, €3.. Comaiidiicci. M.,C:hristiansen, (;.> Birkelund, S., Crveiiini. H., \'rc>tori, E., Ratti, G.. and Palliiii. V. ( 1 996). Mapping of C h h q d i c i frclc/rotticitis proteins hy iiiiiiiol,ilinr-pc~l~tc~lanii(!f hvotlirntwsioiial electroplioresis: Spot itlentification hy N-terminal sequencing and iininiinolhtting. E/cc.fn)~,/ion,sis 17, 1&5- 190. Blackwell, J. M..Barton, C . H., \l%ite, J. K., Roucli, T. I., S11;iw. M. A., U'hiteheatl. S. H., Mock, B. A.. Searlc, S.. M'illians, H., ant1 Bakcsr, A. hl. (1994). Genetic regulation of leislimanial and inycolxicteiiiil infections: The Zd.drlZt!/lHcggene stoiv continiirs. Z t t t f r i r m d . Lf~lt.43,9:)- 107. Blackwell, J. M.. Barton, C:. H.. LC'liite. J . K.,S(wle. S.,Baker, A. M., \Villiams, H., and Sliaiv. M. A. (1995). Genoinic organization m t l sequeiice of the hiiinan NRAMP gene: Itlrntificntion antl niapping ol' a proinotrr region I)oIy~~i~)~phism. MoL. M c d 1, 194-2045. Blackwell, J. M..Black. C.F., Pcacock, C.,Miller, E. N., Sil)thorpe, D., Gnananandha, D.. S1i;iw. J. J.. Sihvira, F.. Liiis-Lainson. %.. Kanmr, 17.. Collins. A., and Shaw, M. A. (1997). Iminunogriietics of leisliinaiiial and iiiycohactvrid infections: The Belein Family Study. Plril. Tt-oti,T.Roy, Soc. Lottdori SPn H. Riol. Sci. 352, 1:331-1345. Blaiik, C., Fiichs. 11.. RappersI,erger, I;..Rollinghoff, M.. arid Moll. H . (1993).Parasitisin of epidrrnial 1,angerhans cells i n esperiniental ciit;ineous Icishmaiiiasis with Leislirtimici ttiujor. /. Zrtfi!ct. Dis. 167, 318-425. Blocker, A , . Severin, F. F., Biirkliardt, J . K.-Bingliam, J . B., Yu, H., Olivo, J. C., Scliroer, T. A.. II!wi;in, A. A , , and (:iiffiths. C;. [ 1997). blolecular requireincvits for bi-directional nioveincnt of phagosomes aloiig niicrotul)iiles. /. Ccdl Riol. 137, 11:3-129. Blockrr. A , , Sweriii, F. F.. Ilaberniaiiii. A , , H y i a n , A. A,, C;riffiths. G., antl Burkliardt, J. K. ( 1 996). Microtu~)ule-~issoeiate(1. I,rotein-tlependrnt I)inding of phiigosonies to rnicrotuhiilcs. J . Hiol. C/it,ttt. 271, 3803-381 1. B o n i a i i , H. C;, (1996).I'eptitle xtti1iotic.s: Holy or hcrcstic grails of innntt~iniiiiiiniv. S c m d J . Z t t i r t i u t u ) / . 43,47#5-482. Bonato. \'. I,.. Lima, \'. M., Tuscon, R. E.. Lowrie. I). H., i t r i d Silva, C. L. (1998).Identificatioii antl cli:iructl.riz;ition o f protecti\,c T cells in hspfij DNA-vaccinated and il~!/cohmctc,riritt~ tri/~c~n.rrlo.ri.s-inf~~cted niicc. Zrifict. Z j t i t t t r r t t . 66, 16% 175. +
340
ULHICH E S(:IlAII3LE
Pf
o/
Bonecchi, R., Bianchi, G., Bordignon, P. P., D'Ambrosio, D. A,, Lang, R., Borsatti. A., Sozzani, S., Allavena, P., Gray, P. A,, Mantovani, A., and Sinigaglia, F. (1998). Differential expression of chernokine receptors and chernotactic responsiveness of type 1 T helper cells (Thls) and Th2s. /. Ex/'. Men. 187, 129- 134. Boring, L., Gosling, J., Chensue, S. W., Kunkel, S. L., Farese, R. V., Jr., Broxmeyer, H. E., and Charo, I. F. (1997). Iinpaired inonocyte migration and reduced type 1 (Thl) cytokine responses in C-C cheinokine receptor 2 knockout mice.]. Clin. Znoest. 100,ZSSZ2561. Borremaiis, M., de Wit, L.. Volckaert, G., Ooms, J., De Bruyn, J,, Huygen, K., van Vooren. J,-P,,Stelandre, M., Verhofstadt, R., and Content, J. (1989). Cloning, sequence detennination and expression of a 32-kilodalton protein gene of Mycobncteriuin tuberculosis. Infect. Itntnuri. 57, 3123-3130. Bouwer, H. G., Seainan, M. S., Forrnan. J., and Hinrichs, D. J. (1997). MHC class Ihrestricted cells contribute to antilisterial immunity: Evidence for Qa-lh as a key restricting eleinent for Listeriri-specificCTLs. ]. Ininiuiiol. 159, 2795-2801. Brachet, V., Raposo, G., Aniigorena, S., and Mellman, I. (1997). Ii chain controls the transport of major histocoinpatihility complex class I1 molecules to and from lysosomes. /. Cell Biol. 137, 51-65 Brandt, L., Oettinger, T., Holm, A,, Andersen, A. B., and Andersen, P. (1996). Key epitopes on the ESAT-6 antigen recognized in mice during the recall of protective immunity to Mycobacteriuiii t u b e r a h i s . 1. Immunol. 157, 3527-3533. Brossay, L., Jullien, D., Cardell, S., Sydora, B. C., Burdin, N., Modlin, R. L., and Kronenberg, M. (1997). Mouse CD1 is mainly expressed on heinopoietic-derived cells. I. Ininiunol. 159, 1216-1224. Brossay, L., Tangri, S., Bix, M., Cardell, S., Locksley, R., and Kronenherg, M. (1998). Mouse CD1 autoreactive T cells have diverse patterns of reactivity to CD1' targets. /. Zinmunol. 160, 3681-3688. Broucpi, P., and Raoult, D. (1993). Proteinase K-sensitive and filterable phagosomelysosome fusion inhibiting factor in Afipici felis. Microb. Prithogen. 15, 187-195. Brown, D. R., Fowell, D. J., Corry, D. B., \Vynn, T. A,, Moskowitz, N . H., Cheever, A. W., Locksley, R. M., and Reiner. S. L. (1996). Beta 2-microglobulin-dependent NKI. 1' Tcells are not essential for Thelpercell2 innmine responses.]. Exp. Med. 184,1295-1304. Brown, E. J. (1991). Complement receptors and phagocytosis. Curr. @in. Immzinol. 3, 76-83. Brutkiewicz, R. R., Bennink, J. R., Yewdell, J. W., and Bendelac, A. (1995). TAP-independent, beta 2-microglobulin-dependent surface expression of functional mouse CD1.1. J . Exp. Med 182, 1913-1919. Buchmeier, N.. Bossie, S., Chen, C., Fang, F. C., Guiney, D. G., and Libby, S. J. (1997). SlyA. a transcriptional regulator of Salnumelln t!yp!iimtiriziin, is required for resistance to oxidative stress and is expressed in the intracellular environment of macrophages. Ir$ect. Zintnun. 65, 3725-3730. Buchmeier, N. A,, and Heffron, F. (1990). Induction of Salmonella stress proteins upon infection of inacrophages. Science 248, 730-732. Buchmeier, N. A,, and Heffron, F. (1991). Inhibition of inacrophage phagosoine-lysosome fusion by Srilnu~nellatyphimurium. Infect. lintnun. 59, 2232-2238. Buchmeier, N. A,, and Schreiber, R. D. (1985). Requirement of endogenous interferon-y production for resolution of Listerin numocytogenes infection. Proc. Nut. Acnd. Sci. USA 82, 7404-7408. Brims-Keliher, L., Nickerson, C. A,, Morrow, B., and Curtiss, R., 111 (1998). Cell-specific proteins synthesized by Snlnwnelln typhimurium. Infect. Zttiinun. 66, 856-861.
B ~ ~ s c D. h , H. , Hoiiwer. H. G., Hinrichs, I)..and Panier. E. C . (1997. A nonamer peptide derived from Listwio itiotioclltc)jiotcs irietalloprotease is presented to cytoljic T Iymplioc)tes. Znfict. Ztttitiuti. 65, Biisch. D. I-I.. Pilip, I. M., el l'mner. E. G. (1098).Coortlinate regulation of coinples T cell populations responding to hacterial infection. Ziiirtiuitity 8, Hyrd, T. F.. and Horwitz, M. A. (1989).Iiitrrferoii-y-~icti\~itetl human monocytes downregulate transferrin receptors and inhiliit the- intracc~ll~t1;tr multiplication of Lcgioiiello pti<:rti i i o p h i h by limiting the awilability of iron. /. ( ; / i n . Z i t L ; m t . 83, 1457-1465. Cadranel, J.. Philippr, C., Pcwz, J., Milleron, B.. Akoiin, G., Ardaillon, R., and Band. L. (1990). I n vitro production of tunionr necrosis factor and prostaglantlin E2 by pc~riplieral I~loodmononnclear cells froin tiiberculosis patients. Cliii. E.tp Zrririuenol. 81, 319-324. Campbell, K. A., Oventlalc, P. J.. Kennedy. M. K.. Fanslow, W. C.. Reed. S. G., and Maliszewsli, C . R. (1996). CD40 ligantl is reqiiircd for protectiw cell-inetliated immunit). to I , o i , h r i m i r i itiojor. Zttiitwitit!/ 4, 283-289. Campos-Neto, A., Ovendale, P., Bement. T.. Koppi, T. A., Fanslow, W. C..Rossi, M. A,. and Alderson, M . H. (19%). CD40 ligaiitl is not essential for the developinent of cell niediated ii nmnnity and rwistance to M!/cohoctc~rirc t t t trrbcrcrrlosis.J . I t i t itttciiol. 160,2037204 1. Canning, P. C., Rotli, J. A.. and Deyoe, H. L. (1986). Release of5'-guanosine nionopliospliatr a r ~ adenine l by Bnicellri r i l ~ o r t r r santl their role i n the intracellnlar snMval of the 1)acteria. /. Znfict. Dis. 154, 464-470. Carhone, F. 11.. and Bevaii, M. J. (1990). Class I-restricted processing and presentation of exogenous cell-associated antigen in \ivo. /. Exp. hfcd. 171, 377-38i. Caron, E.. Li;intard, J. P., antl Koliler. S. (1994).The monocytic cell line U-937, physiologically diffrrcmtiatetl Iy retinoic acid and citainin D3, is a niotlcl for intracellular Iwhavior of Bntcrllri spp. A n n N . Y. Acrid Sci. 730, 276-278. Castano. A. I<.. Tangri, S., Miller, J. E. W , Holcombe, H. K., Jackson. M. R., Huse, W. D., Kronenbrrg, M., antl Peterson, P. A. (1995). Peptidr binding and prrsentation by nioiise C D l . Scicmv 269, 22%226. C e h , M., Engering, A.. Pinet, l',, Pietcw, J.. and Lanzaveccliia. A. (1997). Inflammatory stimuli indiice accumiilation of MHC class I 1 coinpleses on dendritic cells. Nntrcrc 388, 782-787. Cliace, J. H., Hooker, N. A,, Mildenstcain, K. L., Krieg, A. M., nnd Cowdrry, J. S. (1997). Bacterial DNA-induced N K cell IFN-y production is dependent on macrophage secretion of IL-12. Clin. I r t f l l t t ~ t ~ oZ ~t .f i ? l f l ~ ~ ~ ~ )84, ~ ) ~185~ ~ l (193. ~ l , Chan, J., Fnjiwara, T.. Brennan. P., McNeil, M.. Turco. S. J., Sibilk, J. C., Snapper, M., Aiscri. P.. a n d Bloom, B. R. (1989). Microl)iol glycolipids: Possible virulence factors that scavenge oxygen radicals. Proc. N d . Accid Sci. USA 86, 2453-2457. Chain. J.. Tan
149-161.
Chen. I I . , and Paul, .I\ E. (1997). Cultnred N K I ,1'CD4' T cells produce large mounts /. 2240-2249. of IL-4 and IFN-y upon activation by anti-C1):3 or CD1. J . ~ l t l i l l ~ i t u J 159, Chen, I,.-M.. f1ohI)ie. S., and C;aldn, J. E. (1996). Requirement of CDC42 for Salinonellainduced cytoskeletal and nuclear responses. Science 274, 21 15-21 18. Clien. Y. H., Chiu. N. M., Mantlal. M., Wang, N., and W a n g . C. R. (1997). Impaired N K l t T cell tlevelopment antl early IL-4 pr(duction i n CDI -tlrficient mice. Ziriitirinity 6,459-467. Cliien. Y.-H., Jores, H.. aiitl Chwley, M. P. (1996). Recognition hy ylS T cells. Artnri. R w .
rtl, V. L. (1995). Differential regulation of the inaiino~eand SPA receptors on niacrophages. Am. J . Ph?/siol. 269, L721-726. Chroneos. Z.C., Abdolrasnlnia. R.. Wliitsett, J. A , , Rice. W. H . , and Shepherd, \7. L. (1996). Pnrification of a cell-surf'ace receptor for surfactant protein A. J . B i d . Chetri. 271, 16375-16383. Cirillo, J. D., Falkow, S., Toinpkins, L. S., and Berinudez, L. E. (1997). Interaction of Mycobncteriiiiit nviiiiii with environniental anioebae enhances \indence. Irlfect. ztiiinif n. 65, 3759-3767. Ciupitii, A. M., Petersson. M.. O'Donnell, C. L., Williains. K., Jindal, S., Kiessling, H . , and Welsh, K . M . ( I 998). Iinmnnization with a lymphocytic chorionieningitis peptide mixed wvitli heat shock protein 70 results in protective ininiunity md specific cytotoxic T lyinpliocytes. J . Esp. bled. 187, 685-691. Clague. M. J., Urlw, S., Aniento. F., and Grnenberg, J. (1994). Vacuolar ATPase activity is required for endosomal carrier vesicle forniation. I. Biol. Chern. 269, 21-24. Cleniens, 11.L.. and IIonvitz, M. A. (1992). Menibrane sortingduringphagocytosis: Selective exclusion of major Iiistocompatibility coiiiplex inolecules but not complement receptor CR3 during converitio~idantl coiling phagoc,*osis. J . E x ~ Jh. l ~ d 175, . 1317-1326. Cleniens, D. L., and Honvitz, M. A. (199.5).(:liai-acterizatioiioftlie M!/cobacfr,riiin1tribercrrkjsis pliagosome antl evidence that pliagosonial maturation is inhibited. /. Esp, McY/. 181, 257-270. Cleniens, 1). L.. and Honvitz, M. A. ( 1996).Thv Mycobncten'riiti tubrrc:trbmisphagosoine interacts with early cnclosomes and is accessil)leto exogeiiously atlininistered transferrin. J . Esp M r d . 184, 13349-1355 .ec, B. Y.,; i t i d IIonviti, M. A. (1995). Purification. characterization and of A4ycobcrczerirl t n tiibercdosis nrease, a potentially ciitical determinant of host-pathogen interaction. J. Bnrtcriol. 177, 5644-5652. Clevelaiid, M. G., Gorhain. J. D., M u i ~ ~ l iT. v , L., 'raonianen, E., and Murphy. K. M. (1996). Lipoteichoic acid preparations of gran-positive bacteria induce iiiterleukin-12 through a CD 14-depentlent pathway. Zilfect. Z i n i i i u i i . 64, 1906- 1912. Colditz, G. A,, Brewer, T., Brrkey, C. S., Wilson. M. E.. Burdick, E., Fineberg, H. \{,, ant1 Mosteller, F. (3994). Efficacy of BCG vaccine in the prevention of tuberculosis: Metaanalysis of the pnhlished literature. J . Am. hid. Assocint. 271, 698-702. Cole, 5'. T.. Brosch, R., Parkliill, J., Gamier, T., Chnrcher, C., Harris, D., Gordon, S. V.. Eiglineier, K., Gas, S., Barry, C. E., Tekaia, F., et d.(1998). Deciphering the ltioloq of Mycobacterirr i t 1 tribrrcri/osis from the complete genome sequence. Nnture 393,537-544. Col!ins, 13. L.. Schaittle. U. E., and Kaufhiann. S. 11. E. (1998). Early lL-4 induction in bone marrow lymphoid precursor crlls by mycobacterial lipoarabinomannan.J . Z r ~ ~ t i i r ~ n o l . In press. Collins. 11. L., Schaihlc, U. E., Ernst, J. D., and Hiissell. D. G. (1997).Transfer ofpliagocytosed particles to the t)ar"sitol)Iioroiis vacuole of Leishtiimiici triexicniin is a transient
ISTRACELL.ULAK BA(:I’ERIA
.ASD THE I M M U N E SYSTEM
343
phenomenon preceding the acquisition of annexin I by the phagosome. J . Cell Sci. 110, 191-200. Connell, N. U., Medina-Acosta, E., McMaster, 117. R., Bloom, B. R, and Russell, D. G. (1993). Effective iinmunization against cutaneous lcislimaniasis with recombinant bacille Calmette-Guerin expressing the Leishmania surface proteinase gp63. Pntc. Not. Acad. Sci. USA 90, 11473-11477. Constant, P., Davodeau, F., Peyrat, M.-A., Poquet, Y., Puzo, G., Bonnevik, M., and Fouriii6, J.-J. (1994). Stimnlation of human y6 T cells by nonpeptidic mycohacterial ligands. Science 264, 267-270. Cooper, A. M., Dalton, D. K., Stewart, T. A,, Griffin, J. P., Russell, D. G., and Orme, I. M. (1993). Disseminated triberculosis in interferon gamma gene-disrupted inice. J . Exp. Med. 178, 2243-2247. Cooper, A. M.,D’Souza, C. D., Frank, A. A,, and Orme, I. M. (1997a). The course of bi‘!/cohnctcriutn trihercii1o.yi.sinfection in the lungs of mice lacking expression of either perforin- or granzyme-mediated cytolytic mechanisms. ZFfect, Z r t i m i m . 65, 1317-1320. Cooper, A. M., Saunders, B. M., D’Soriza, C. D., Frank, A. A,, and Orme, 1. M. (199711). Mycobacteriim tiiherculosis-driven processes in gene-disrupted mice. B i d . Ztxtitut Past. 95,85-96. Cossart, P. (1997). Host/pathogen interactions: Subversion of the inaminalian cell cytoskeleton by invasive bacteria. J . Clin. Znuest. 99, 2307-231 1. Cowky, S. C., Mykseva, S. V., and Nano, F. (1996).Phase variation in Fmnci.sella tulnrensis affecting intracellular growth, lipopolysaccharicle antigenicity and nitric oxide production. h 4 d ilficrobiol. 20, 867-874. Croft, M., and Dubey, C. (1997). Accessoly molecule and costimulation rrcjuirernents for CD4 T cell response. Crit. Rev. Zmniunol. 17, 89-118. Crowle, A. J., Dahl, R., Ross, E., and May, M. H. (1991). Evidence that vesicles containing living, virulent Mycohacteiizm tiihrrculmi.~or Mycobncteriuni mviutn in cultured human macrophages are not acidic. Ztfect. Zrmriun. 59, 1823-1831. Crowle, A. J,, Ross, E. J., and May, M. H. (1987). Inhibition by 1,25(OH)2-vitamin D 3 of the multiplication of virulent tubercle hacilli in ciiltiired human macrophages. Infect. Zmniuti. 55, 2945-2950. C,wes, C., Godenir, N. L., Hoppe, H. C., Scholle, R. R., S t e p , L. M., Kirsch, R. E., and Ehlers, M. R. (1996). Nonopsonic binding of bfycobacterium f i d m m l o s i s to human complement receptor type 3 expressed in Chinese hainstrr ovary cells. Zrfect. Z m t w i i . 64, 5373-5383. Cywes, C., Hoppe, H. C., Daffe, M., and Ehlers, M. R. (1997). Nonopsonic binding of Mycohacteriutii tuberculosis to complement receptor type 3 is mediated by capsular polysaccharides and is strain dependent. Ztf2ct. Zmrnun. 65, 4258-4266. D’Andrea, A,, Aste-Amezaga, M., Valiante, N . M., Ma, X., Kubin, M., and Trinchieri, C:. (199.3). Interleukin 10 (IL-10) inhibits human lyinphocyte interferon-y production hy suppressing natural killer cell stimulatory factor/IL-12 synthesis in accessory cells. J . Esp. A4ed. 178, 1041-1048. Darji, A,, Cuzman, C. A,, Gerstel, B., Wachholz, P., Tirnmis, K. N., Wehland, J., Chakrahorty, T., and Weiss, S. (1997).Oral somatic transgene vaccination using attenuated S. typhiiiiirriuin. Cell 91, 765-775. Dehrick, J. E., Campbell, P. A,, and Staerz, U. D. (1991). Macrophages as accessory cells for class 1 MHC restricted iininune responses. J . Znwiu71ol. 147, 2846-2851. de Cliastellier, C., Lang, T., and Thilo, L. (199.5).Phagocytic processing of the macrophage endoparasite, Mycohacterizrni az;iim, in comparison to phagosomes which contain Racillris siihtilis or latex beads. Eur. 1. Cell B i d . 68, 167-182,
344
UI.RICII E SCHAIBLE et (11
tle Chastellier, C., and Thilo. L. (1997). Phagosonie maturation and fusion with lysosomes in relation to surface property and size of the phagocytic particle. Eur. J. Cell B i d . 74,49-62. De Groote, M. A,, Ochsner, U. A,, Shiloh, M. U., Nathan, C.. McCord, J. M., Dinauer, M. C., Libby, S. J., Vazquez-Torres, A,, Xu, Y., and Fang, F. C. (1997). Periphsmic superoxide dismutase protects Salmonella from products of phagocyte NADPII-oxidase and nitric oxide synthase. Proc. Not. Acnd. Sci. USA 94, 13997-14001. De Groote, M. A,. Testerman, T., Xu, Y., Stauffer, G., and Fang, F. C. (1996). Honiocysteine antagonism of nitric oxide-related cytostasis in Snlttu)nella typhirnzcrinrta. Science 272, 4 14-417. de Jong, R., Haagen, I.-A,, Altare, F., Elferink, D. G., de Boer, T., van Breda Vriesman, P. J. C., Kabel, P. J.. Draaisma, J. M . T., van Dissel, J. T., Casanova, J.-L., Kroon, F. P., and Ottenhoff, T. H. M. (1998). Severe mycobacterial and salmonella infections in interleukin-12-receptor-deficient patients. Science 280, 1432-1435. de la Salk, H., Hanau, D., Fricker, D., Urlacliel-,A,, Kelly, A,, Salainero, J., Powis, S . H., Donato, L., Bausinger, H., Laforrt, M., et ul. (1994). Homozygous human TAP peptide transporter mutation in HLA class I deficiency. Science 265, 237-241. DeLibero, G., Casorati, G., Giachino, C., Carbonara, C., Migone. N., Matzinger, P., arid Lanzavecchia, A. (1991). Selection by two powerful antigens may account for the presence of the major population of human peripheral y/S T cells. J. Exp. Med. 173, 1311-1322. DeLibero, G., Flesch, I. E. A,, and Kaiifmann, S. H. E. (1988). Mycobacteria reactive Lyt2' T cell lines. Eur. J. lrnmuriol. 18, 59-66. DeLiliero, G., and Kaufmonn, S. H. E. (1986). Antigen-specific Lyt-2+ eytolytic T lymphocytes from mice infected with the intracellular bacterium Listeria nlr)noc!/togenes. J. Itrirniinol. 137, 2685-2694. Dellabona, P.. Casorati, G., Friedli, B., Angman, L., Sallnsto, F., Tunnacliffe, A,, Hoosneek, E.. and Lanzavecchia, L. (1993). I n viuo persistance of expanded clones specific for bacterid antigens within the human T cell receptor ~$3CD4-8- subset. J. E x p Mecl. 177, 1763-1771. Denis, M., and Gregg, E. 0. (1991). Recombinant interleiikin-6 increases the intracellular and extracellular growth of Mycobacteriurn noiri tn. Can. J. Microbiol. 37, 479-483. Deretic, V., Via, L. E., Fratti, R. A., and Deretic, D. (1997). Mycobacterial phagosome maturation, rab proteins, and intracellular trafficking. Electrophoresis 18, 2,542-2547. Desjardins, M. (1995). Biogenesis of phagolysosomes: The kiss and run hypothesis. Trends Cell Biol. 5, 183-186. Desjardins, M., Celis, J. E., van Meer, G., Dieplinger, H., Jahraus, A,, Griffiths, G., and Huber, L. A. (1994a). Molecular characterization of phagosomes. J. Biol. Claem. 269, 32194-32200. Desjardins, M., Huber, L. A., Parton, R. G., and Griffiths, G. (1994b). Biogenesis of phagolysosomes proceeds through a sequential series of interactions with the endocytic apparatus. J . Cell Biol. 124, 677-688. Devitt, A., Moffatt, 0..Raykundalia, C., Capra, J. D., Simmons, D. L., and Gregory, C. S. (1998). Human CD14 mediates recognition and phagocytosis of apoptotic cells. NuttLre 392,50S-509. Diaz, R., Wileinan. T. E., Anderson, S. J., and Stahl, P. D. (1989).The use of perineabilized cells to study the ion requirements of receptor-ligand dissociation in endosomes. Biochetn. J . 260, 127-134. Dietrich, G., Buhert, A., Gentschev, I., Sokolovic, Z.,Simm, A., Catic, A,, Kaufmann, S. H. E., Hess. J., Sxalay, A. A., and Goebel, W. (1998).Deliveryof antigen-encohngplasrnid
D l / \ into tllr c!.tosol 01' ~ii;u.ropIl~~gc~s 1)). ;ittc,nr~atcatls~~icitl(a 1,istc~t-it1ttrorroc~t/iogc~ti(~~. S(tl/rt-e,Hiotc,c.lrtiol. 16, 181- 18.5. l)ictricIr, I\.. F.. I):LIII~OII. 11. \ I , . I\l)(arg, K . It.. 1,;111(lc>r, E. S.. :111(lS\V;~IISOII. X I , S, (ISJSJ5). I,gttl, it g ~ i tlr~tt e ~ I v ~ ( ~ I - I I Is~~I ~I (s~c\~ ~ ) t i l )to i l iI,(~giotr(~//(i t! /)rr~~~~ttio/)/ri/(i. 11);11)\to I I I O I I \ C , c l r r o ~ i ~ o s o1:3. l ~ ~Cc,trc~ttric~~ t~ 26, 4-1:3--1.50. l)i~~:uler, 11, C,, ll(~cl<. h1. 13., :t11(1 UI~;IIIIIC,. E. I<.( 1997).hlicv I;tcki~rgr c ~ l ~ ~ c~ricoti~r;t~~ri(Ic~ c(l ;1(1(~11i11c> cli~r~~clc~oti(lv l~lrosi)lr;it(so ~ i ( ! ; t s:~cti\it!. ~ slio\v i~rcrcxscds ~ ~ s c c ~ ~ t i ~to ) i le;trl> it! J . Ittittirttiol. 158, 5.781 -5.58:3. illfi,ctioll \\it11 1,istc~t-irltlrottoc~!/logc~t~c~v. 1)oerflc.r. \\'.. Sclr~~l)l)t~rt, H.. l l ( ~ I l ( 11.. ~ , Kam1lrc.r. (:.. Ililgrr-E\-~rshc.i~~i. K.. KIIoI)I;LII~II. l.1.. ;ultl l l c % ~ r11. ~ ~(1997). ~ s , Illt(-gr:ttio~rol'li)r('ig~~ I I S h ;111(l its c011~(~(~11(~11c(~5 ill I I I ; ~ I I I I I I ; L ~ ~ ; ~ I I s! stc21irs.I'tr~tr(lvAiol(~(~lrtrol. 15, 297L:iOl I)oIrert!. T. XI.. ant1 Sller. A. i 1997). Ilc~k~cts i ~ cc~ll-lrrrtli;ttvtI r i~irlrr~~l~it!al'kxct clrrolric. I ~ l t ~ i o innate. t resista~lcc,ol'~rricc> to J l ! / c ~ o l ) t r c ~ i ~ ~ t - o~.iritti i t ~ t t i iirli~ctio~r.]. Itttttrrttrol. 158, 48"-1% I . L ~ ~ I I I ; IE.. I I I\I\.' r l ~ l i ~ ~J,. i ( lRoli(l(,, . 11.. l'istor, S., 11:trtl. 51.. (;oc-l)(,l.\\... I~(~i~rrt~istt~r-\\';~cl~t(~r, kl.. ;LIICI(:l~;tkral~ort!~. T, ( lS)LI?). :\ I I O \ (,I l);tctc>ri;tl\-irr~Ic~ticc~ gc.~r(~ i l i I,i~t(,t-i(~ rtrotto(~!//og(~~r(~~\ r ( ~ ( ~ ~ ~ti)r i r (11ost ' ( l c~11~~ricro(il~lri('~rt i~~tc.ritctio~~ \\it1111o11101oq to t l ~ l>roli~rc>-ricl~ r r(3gio11 of' \-i~ic~~liir. EJIHO 1. 11, 1981- 1990. Do\\?lil~g,j. I;.. l';is111;1. 11.. \\'rigIrt. J . R.. T\\.i$g, 11. L. 11,. ; I I I ~k1arti11,\\'. 1. 11. (199.5). S~~rlactalrt protc.ilr A pro~rrotc.~ attachmrlit of' Jl!/c~o/)n(~/c,t-i~itti ftr/)c,t-c~rl/o\i.v to ;t1\(~1l:u~riacro~)lragcs tl~lri~rg i~rliactio~l \\-it11lr111ili111i ~ ~ ~ i r i ~ ~ ~ ~ o d\irlls. c . f i e1'1-oc.. i ( ~ ~S(I/. ~ c \ I\(.(id.Sci. rS:\ 92, 4848-48,52. Drxkc~,J. R., \I.cll)stpr. I'., (;a~r~l)ic,r, j. (;.. ;urtl hlrllrr~all.I. i 1997). I)rli\.en 01'13 cc.ll rcxcc.pto~-intel-~rulizrtla11tigc.11to c~lrt1osolnt.s ;urtl cl;~sr11 \c~sicl(.s.J . F:.vl). Jl(,t/. 186, 1.799-1:306. L)'Sot~xa, (;,. (;ool~vr.,,\. XI., F r ; t ~ ~ k 11. , ,I., X1~1~~~1cc;lro. 11. 1.. l3loo111.13. 11.. ;111(l0r111c.. I. XI. (1997). AII a ~ r t i - i ~ r l l a ~ r ~ ~ i rrole ; ~ t ofill~ ? -y6 ?' I!.~r~~>lroc\.te i l l ;~cc~~~irc*tl ilrrlrr~~liit\to J l ! j ( ~ ~ ~ / ) ~ i ( ~ t ( , t - t/~/)(,t-(,ri/o.~i.~. irittr J . Ittitt~t~rr(~/. 158, 1217-1 22 I. D'So~lz;t.S..Lr\ill. XI., Faith. ;\., Ysst~I,11.. 13c.1111c.tt. 13.. I,;tkr, K. A,. R r o \ v ~ ~I . . N.. alicl L:tlril). J . 11. (ID$J(i).Ilef&cti\.cx ;ultigc>~rp~-occ.\silrg;i\soc~i;~tc~cl \\.it11 lb~r~ilial tlissc~trril~atrcl 103, :35-:39. 111)col):~ct(~rio~i5, Cli~r.F:.v/). Zttttt~~i~roI. 1111lrn,K. I\.., aiitl Xl;t\firltl. 1;. 13. (199". I)c.li\c.~? of lig;u~tlsli.o~rrsorting c~1rtloso1i1c.s to of sortiirg erltlosolr~es.1. C d l Biol. 117, :301:310. lato c-nclosonlr~so c c ~ ~ I)!r s ~iiat~~ratiorr 111111n,\\'. A. (1994). ;\~~tol)lr;tp a~itlrel;ttecl ~~~c~clr;ulis~rrs of ~ ~ s o s o r r r c ~ - ~ ~ r c cp~r io; t~ct ~c ti~~ ~ dc.gfittlutio~~. Tt-(,trt/.sCcll Uiol. 4, 139-11:3. ~ I I I I I I I11. ~ ' \Y., . 11cs1rick. 11.. (;r('(xlll)(~rg,J.. Kric'gt'r, XI.. 1111c1 Joili(sr, ti. I\. (lU!J1). 1'11(' t\l>c2I ~ i i a c r o ~ ~ l r sca\c2irgc.r agt~ r c ~ c ~ p t oI~intls r to gfi~~~~-positi\.c. I)actc~ri;ta~itlrctcogni~ca\ lipotc~iclroici~citl.1'1-oc.. .\-of. i\c.nrl. S(,i. 1'SA 91, 1S ( i 3 1867. Elllvrs. XI. I{. \\. .. :t11(1 I);lrf?. 11. (I<)%S). I~rt(~r;tctio~~s I)(>~\\TY,II Jl!~~~(~/)~i~~/(~t-ir~tti i/i/)('t.(,/l/o.\i.\ 6, :328-:3:3,5. :urd Irost cplls: Are I I ~ ! col):tctcxri;~lsltg;lrs thy kc,\ ;' Zt-(,rrt/v Jlic~tr)l~iol, Elrrt, S,. Slrilol~.k1, U,. ~ ~ I I L ~J.,I IClloi. , \ I . , ( ; I I I I L ~ ) L I ~S~., ~S ; I ~ ~ I(;.. ; IXi(,, I I , Q..; I I I ~11iIc,!.. I,. \I7. (1997). :\ no\.rl a~~tio\itla~rt gc5lrc2I'rolrr .\l!/c~o/~c~c~tc.t-i~i~ti trihc'r-(.rrlo~i~ 1. t.xp. .\I(,(/. 186, 1885-1896. Elom;~u,0.. K;tl~g;ts.1.1.. Salrll)c.rg. (;.. T I I I I ~ ~ ; ~J., I I S~~>~I I I. I I I I K I (. ,' ILii~kkil. I. A , , Tlrc~slvl'l'. 1.. Kraal, (:.. ;II~(IT ~ ? . g p a s o ~K.~ .(1995). (;lollil~gof' ;L ~ro\c,lI),~ctc~ria-l)i~rcli~rg reccJptor str~rct~lr;tll\. rc%l;~tc~l to sc.avrl1gc.r rrcc2ptors alrtl c,\;prcas\c~li l l s111)sc~t of' ~rr;~c~.o~)l~agc.\. C,'c,ll 80, 6():3-609. ~ I II<. , I 199.5). 11~t~~rlt~11ki11-4-1~ro~l11ci11g (:l)-i Etrroto. 11.. E111oto.Y,,:11i(1 K ; I I I ~ ' I ~ I :S.~ I11. I K 1 . l ' TCK a/P i~ltcr~r~etliatc. li\ cr !I lrrl)lroe!-tes ;irt3(lo\\ 11-rcg11latec1 I)>- I,i.~tc,~-ici r~ro~roc.~/25, :3:321-:3:325. /ogr.trcJ.u.Errr 1. It~rt~rrttrol.
Engering, A. J.. Cella, M.. Flnitsina. D., Brockhaus, M.. Hoefslnit, E. C., Lanzavecchia. A,. and Pieters, J. (1997). The inannose receptor fiinctions as a high capacity and broad specificity antigen receptor in human dendritic cells. Eur. /. Irnmunol. 27, 2417-2425. English, B. I(.. Patrick, C. C., Orlicek, S. L., McCordic, R., and Shenep. J. L. (1996). Lipoteichoic acid from viridans streptococci induces the production of trnnor necrosis t"ictor and nitric oxide by innrine macropliages. /. Znjkct. Dis. 174, 1348-1351. Enonioto. A,. Nishimiira, II., and Yoshikai, Y. (1997). Predominant appearance of NK1.l' T cells prodircing IL-4 may be involved in thc. increased susceptibility of mice with the beige mutation tlrning Salmonella infection. J . Z t t m w t d 158, 2268-2275. Epstein, J., Eichl,aiuii, Q., Sherrif, S., and Ezekowitz, R. A . (1996).The collectins in innate iniinunity. Crrrr. O p h Zmmund 8, 29-35. Ernst, J. D. (1998). Macrophage receptors for I\lycohnctcririm tuhurcukosi.~.Zt$ect. Ztnrfum 66, 1277-1281. Ernst, W. A,, Maher, J., Cho, S.,Niazi, K. R., Chatterjee, D., Moody. D. H., Besra, G. S., Watanalie, Y., Jensen, P. E., Porcelli, S., Kr~inenberg,M., and Modlin, R. L. (1998). Molecnlar interaction of C D l b with lipoglycan antigens. Irnrrurnity 8, 331-340. Exley, M., Garcia, J., Balk, S. P., and Porcelli, S. (1997).Requirements for CDld recognition by hinnan invariant Va24'CD4-CD8- T cells. J. Exp. Med. 186, 109-120. Ezekowitz, R. A,, Sastry, K., Bailly, P., and W;irner, A. (1990).Molecular characterization of the human macrophage inannose receptor: Demonstration of multiple carbohydrate recognition-likedomains and phagocytosis of yeasts in Cos-1 ccl1s.J. Exp. Med. 172,17851794. Fang, F. C. (1997). Perspectives scries: Host/pathogen interactions. Mechanisms of nitric oxide-related antimicrobial activity. /. Clin. Innest. 99, 2818-2825. Fazal, N. (1997). The role of reactive oxygen species (HOS) in the effector mechanisms of n . B i d . Znt. 43, 399-408. hunian ~intiinycoliacterialiminunity. B i o c h ~ ~Mol. Fearon, D. T., and Locksley, R. M. (1996). The instnictive role of innate ininiunity in the 272, 50-53. acquired immune response. Sci cr~c~ Feltqnate. D.M., Heaney, S.,Webster. R. G., arid Robinson, H. L. (1997).Different Thelper cell types and antibody isotypes generated hy saline and gene gun DNA immunization. /. ~ t f 2 ? f L l t f l ~ J158, l. 2278-2284. Ferniri, C., Knight. A. M., Watts. C., and Picters, J. (1997). Distinct intracellular compartments involved in invariant chain degradation and antigenic peptide loatling of major histocompatiliility complex (MHC) class I1 niolecules. /. Cell Biol. 139, 1433-1446. Fine, P. E. (1995). Variation in protection by BCG: Implications of and for Iieterologous iniinunity. Lnricet 346, 1339-1315. Fineschi, B., ;mtl Miller, J. (1997). Entlosoitial proteases and antigen processing. Trenrls BiucJmi. Sci. 22, 377-382. Finlay. €3. B., and Cossart, P. (1997). Exploitation of mammalian host cell fnnctions by bacterial pathogens. Science 276, 718-725. Finlay, B. B., and Falkow, S. (1997). Coininon themes i n niicrobial pathogenicity revisited. Microbial. &lo/.B i d . Reo. 61, 136-169. Fischer Lindahl, K. (1997). On naming 132 haplotypes: Functional significance of MHC class Ib alleles. I ~ , i , , z u t i ~ ~ ~46, e ~53-62. e~~c.~ Fleming, M. D.. Trenor, C. C., 111, SII, M. A,, Foernzler, D., Beier, D. R., Dietrich, W. F., and Antlrews, N. C. (1997). Microcytic anaemia mice have a mutation in Nramp2, a candidate iron transporter gene. Nnture Genet. 16, 383-386. Flcsch, I. E., Hess, J. I f . . Huang, S.. Aguet, M., Rothe, J.. Bluethniann, I i . , and Kaufirianii, S. H. E. (1995).Early interleukin 12 production by macrophages in response to mycobacte-
rial infection clepeiitls 0 1 1 iritc4i.ron-y and tuiiiorriwrosis f'ktor a , / .Exp. Jlcd 181, 1615-
1621. Flescli, I. E., and Kaiil'inann, S. 1 I . E. (1988). Attempts to characterize tlic mechmisms iiivolveti in iiiycobacterial gro\vth inliibi tion b\-y-iiiterferon-ucti\,ntcd Imir marrow niiicrophages. Ztfict. Z t t i t t i i i t ~ ,56, 1464-1469. Flesch, 1. E., and Kaiifniaiirr, S. H. E. (1990;~). Stiniiilation of antibacterial macrophage acti\,ities l)y B-ccll stiiniilatory f k t o r 2 (interleukin-6). Ztfict. Z i t i t t u t t t . 58, 269-271. Flescli, 1. E., \l'aiidersee, A,, and Karifninnri, S. H. E. (2997). IL-4 secretioii by CD4' NKI ' T cells indiicc~sinonoc!te clieiiioattractairt proteiii-1 in early listvriosis. /. Z i t u i u i t i o / .
159, 7-10. Flesclr, I. E. A,, und Kaiifiiiaiiii, S. I I . E. (l990b).Activation oftulierciilostatic microphage functions liy iiiterfhii-y, interleukill 4 and tninor necrosis fiictor. Zrfect. Z t ~ i t t i i i t i . 58,
267.5-2677. Flescli, 1. F,. A,, aiid Kaufinann, S. 11. E. (1991).'vlcchanisins involved in iiiycobactcsl-ial growth inhibitiori Iiy y-iiitcrli.ron-~icti\-atrtl l ~ o i i c . inarro\v macrophages: I
Fl!nn, J . L., Colclstc~irr,M.M.. C h i , J., Tric~I)oltl,K. J., Pfrffcv, K., I,o\vciistein, C. J., Schreilier, R., Mak, T. \V., a i d U l o o i i i , H. K. (199.5).Tiiiiior nwrosis f:,ictor-ais reqiiirrd iii thcx protective i i n n i i n r c ~rcspoiise against Jl!/cohncfc,r~iitti tidio-culosi.v in iriicc. Z t t i i ~ i i i i i i f y
2, 361-372. FI>ini, J. I,., Coldstein, M . M., Tric~bold,K. J., Kollcr, B., and 13looii1, H. K. (1992). hlajor IiistocoiiipatibiIit>. coinplrx class I-restricted T cells are required for resistance to Jl!/cobnctcriritii tiibcv-crrlosi.y itift+ctioii.Proc. Not. Actrd. Sci. USA 89, 12013- 120 17. Fl>nri, J. L., \\'eiss. \\'. R., Sorris, K. A , , Sc>ifi.rt,11. S., Kuinar, S., and So, M. (1990). Generation of a c)totoxic T-lyiriplioc!-tc, response iisiiig a Saliriotrc~llaantigen-tkliven 111. Jlol.
The
Xlicrohiol. 4,21 1 1 -21 18.
, A. H., Leiby, 1). A , , Kara>.anan, R. H., Asafoadjei, E., Crawford, R. M., Nacy, ., and hfeltzrr, kl. S. (1995).Groivth of Ft-rrricircda trilnrc7nsi.c L\'S in mucropliages: acidic intracellnlar coiiipnrtinerit pro\-idcs essential iron recjiiiretl for gro\vth. Zttfect.
Z i i i t t i i i t i . 63, 147-1483, Francis, C:. L., Ryan, T. A , , Jones, H. D., Sniith, S. J., and Fulkow, S. (1993).Riiffles induced I>>,Saliiioiic,lla arid other stimuli dirpct macropinoc>tosis ofliacteria. Natri rt' 364,639-642. Freliel, C., tle Cliastellier, C,, I,ang, T., and Rastogi, N. (1986). Evidence for inhibition of fusion .' of Iysosoinal and prelysosotiial coiiipartineiits \\it11 phagosoiiies in macrophages infected wit11 pathogenic ,\Z!/cobacterinrti ocziLiti. Zt,fecf. Z t i ~ t t i r ~ n52, . 252-262. Friitli, U., Solioz, I., and Louis, J . A. (1993). L i t i o t i i o ttiujor interfercs \tit11 antigen prcsentatioii by irifectcd niacrophages. /. Z t t i t t i i i t i t d . 150, 1857-1861. FII, Y.-X., Roark, C . E., Kelly, K.. Drcvets, I]., Caiiiphell, P., O'Hrien, R., and Horn, M'. (1994). Immune protection aiid control of inHainmaton tissue necrosis l)y y6 T cc.lls.
/. Z n i ~ t ~ i ~ i r 153, o/. 3101-31 15.
Fujise, S., Kisliiliafii, K., Lee, K. Y.,Matsii7aki, G., and Nonioto, K. (1997). Normal micropliagct functions, but iiripaircd induction of y6 T cc~lls,at tlie site of bacterial infection in CD45 exon 6-deficierit micc,. Erir, 1.Z t t i t t t r L i i o l . 27, 2549-2556.
348
U L H I C I I E SCII.4IULE et
NI
Fujiwara, N. (1997).Distrihiition of antigenic glycohpids among Mycubacterizrm tuberc\lthsis strains and their contribution to virulence. Ktrkknkti 72, 193-205. Gallin, J. E., Pace, J., and Hayman, M. J. (1992). Involvement of the epidermal growth factor receptor in the invasion of cultiired inanimalian cells by S U ~ T I Utyphinmritori. ) ~ ~ N(itrtrc 357, 588-589. Gain, T.. and Weiss, J . ( 1997). Antimicrobial peptides of phagocytes and epithelia. Semiti. Hc,ttuitol. 34, 34:3-354. Garcia-del Portillo, F. (1996). Interaction of Salnionella with lysosomes of eukaryotic cells. Micruhiolugia 12, 259-266. Garcia-del Portillo. F.. and Finlay, B. B. (1995). Targeting of S a h u r t e h t y p h i ~ t i ~ i r i r ~ t t i to vesicles containing lysosoriial iiiembrane glycoproteins b p i sses compartmeiits with maiiiiose 6-phosphate receptors. 1. Cell B i d . 129, 81-97. Garcia, V. E., Sieling, P. A,, Gong, J., Barnes, P. F., Uyeinura, K.. Tanaka. Y., Bloom, B. R., Morita, C. T., ; i d Modliri, R . L. (1997). Single-cell cytokine andysis of$ T cell responses to iionpeptide mycobacterial antigens. 1. I t r i m i n u l , 159, 1328-1335. Gircia Vescovi, E., Soiiciiii, F., arid Groisman, E. A. (1996).Mg2' as an extracellular signal: environiiiental regulation of Salmonella virulence. Cell 84, 165-174. Guuse, W. C . , Ifalvorson, M. J., Lu, P., Greenwald. R., Linsley, P., Urban, J. F., and Finkelman, F. D. (1997). The fiinction of costimulatory inolecules and the development of IL-4-pro~l11cing T cells. Ittimzctcol. Torlay 18, 115-120. Gaynor, C. D., McCorinack. F. X., Voelker, D. R., McGowan, S. E., and Schlesinger, L. S. ( 1995).Pii1inoiiar-ysurfactant protein A mediates enhanced phagocytosis of Mycubacteririrtr tribcr~:rilosisby a direct interaction with human macrophages 1. Irtzmunol. 155, 533433-so5 1. c;uziiie11i, .. ' R. T., Wysocka, M.. Hieny, S., Scharton-Kersteri, T., Cheever, A,, Kuhii, R., Miiller, W., Trinchieri, G., and Slier, A. (1996). 111 the absence of endogenous IL-10, mice acutely infected with Toxojhwtri gotiriii succumb to a lethal immune response dependelit on CD4' T cells and accompanied by overproduction of IL-12, IFN-7 and TNF-a. J. l t m w m ~157, . 798-805. Geliran, S. J., Yamamoto, Y.. Newton, C., Toinioka, M., Widen, R., Klein, T., arid Friedman, H. (1995).LPS inhibits the intracellular growth of Legioiiella pneunwphiln in thioglycollate elicited muriiie peritoneal macrophages by iron-dependent, tryptophan-independent, ouygen-iridependentand arginine-indepc,ndent mechanisms.]. Leukocyte Biol. 57,80-87. Grluk, A.. van Meijgaarden, K. E., de Vries, R. R., Sette, A,, and Ottenhoff, T. 14. (1997). A DRl7-restricted T cell epitope from a secreted Mycobmterizrm tuberculosis antigen only binds to DR17 molecules at neutral 1'1%. Eur. J. Iriiitzunol. 27, 842-847. Gerckm, J., Pryjiiia, J . , Emst, M., :uid Flad, H. D. (1994). Defective antigen presentation by Mycubactet-iurtr ttrberciilosis-infected monocytes. Infect. b n t t i t i t i . 62, 3472-3478. Gervais, F., Desforges, C., and Skamme, E. ( 1989).The C5-sufficient A/]congenic mouse strain. Inflammato~response and resistancc: to Listeria ttwnocytogenes. 1. Ininiunol. 142, 2057-2060. Gllosal, A. C., Maiidal, T. K., Chowdhury, T. K., Mukherjee, K., and Mukherjee, D. (1996). IlLA phenotype frequencies in pulmonary tuberculosis. 1.Indian Med. Associut. 94, 328-330. Ghosh, R. N., Gelman, D. L., anti Mawfield, F. R. (1994). Quantification of low density lipoprotein and trarisferrin endocytic sorting in HEPp2 cells using confocal microscopy. J . CcN Sci. 107, 217772189, Ginsberg, A. M. ( 1998).Tu1)erculosisvaccine development. In The Jordan Report: Accelerated Development of Vaccines, pp. 75-79. Division of Microbiology and Infectious Diseases. N IH .
INTRACE L.1 .UI.AH H A C I E H I A A N D 'I-} 1 E I M M UiiE SYSTEM
349
Gobin, J., and Ilonvitz. M. A. (1996). Exocheliiis of Mycobcrctc,n'rr,,r tubercrtlosis reiiiove iron from Iiuinan iron-bincling proteins arid tlnnatr iron to mycol)actiiisin the ,%f.t~il)e~rcdo.si.s cell wall. I. Exp. A f c d 183, 1527-lS32. Cobin, J.. Moore. C. 11.. R r ~ w e ,J. B., Jr., \Vong, 11. K., Gi1,son. R. W..a d Honvitz, M . A. ( 1995). Iron acquisition 1)y A f ! i c , o ~ / r c ~ c , t ~frhc.rrri/osis: ftrti isolation and cliaracterization of a family of iron-binding exoclieliiis. Proc. Nut. Acmd. Sci. U S A 92, 5389-5193. Goldberg, A. Id., and Rock, K. L. (1992). Proteol proteasonws and antigen presentation. Nnturc, 357, 375-:379. Goklfield. A. E., Drlgado. J. C., Tliiin, S., Bozon, M. V., Ugliaro, A. M . . Turbay, D., Cohen. C., and Yunis, E. J. (1998). Association of im IILA-DQ allele with clinical tulwrculosis. /. h i , hfed. Associnf. 279, 226-2228. Goltlstein, J. L., Brown, M . S., Antkrson. R. G.-Rrissell. D. \i;,, and Schneider, \V. J. (1985). Recc.ptor-niediatet1 rndocytosis: Coiicrpts eiiirrging from tlie LDL receptor system. Aritm H m . Cell Riol. 1, 1-39, Gordon. A. €1 , D'Arcy Hart. I'., and Young, M . R. (19280). Ainiiionia inhibits phagosoirielysosoine fusion in niacropliages. NrrfitrrJ 286, 79-80, G o r h m . J. D . Ciiler, M . I,.. and Muipliy, K. M . (1997). Geiictic control of interleuhinJ. AfOl. h f d . 75, 502-51 I . Irriplications for disease patliogeiie f h h a n i , J. D., Guler, M. I,., Stccw, I{. <;.. Mwkey. A. J., Ddy, M . 1.. Frederick, K.. Dietrich, \V. F., and Murphy, K. M . (1996).Genetic mapping ofa inurine locus controlling tle\&)pment ofT helper liT helpc,r 2 t!pe rcymises. Proc. A7/rt. Acnd. Sci. USA 93, 1246712472, Gregoy, S. H., arid Wing, E. J. (19Y:3). IFN-y inhibits the replication of I,istcn'n rtiortocyto,actit's in liepatocytes. 1,I r t t r r ~ t d .151, 1401- 1409. Grewal. I., Xu J., aiid Flavell, R. L. (1995). Impairment of antigen specific T-cell priming in niice lacking CD40 ligand. A'crfnr-c 378, 617-662. Grewal, 1. S.,Borrow, P.. Panrer, E. C;.. Oldstorir, M . B., and Flavell, H. A. (1997). The CD40-CD154 system in anti-infective host dcfwise. Cum. Opin. I t t t m i m o ~ .9, 491 -497. Grewal, I. S . , md Flavell, H. A. (1996). '4central role of CD4O ligand in the regulation of CD4' T-cdl rrspoiisvs. Ittlttirc r d . Torlwf 17, 4 10-414. C;riffitlis, (;, ( 1996). On vesicles and nirrnbrane coinp;irtmeiits. Protopkrmt~r 195, 37-58, Groettriip, M., Soza, A,, Kuckelkorn, U., a n d Kloetzel, P. M. (1996). Prptide antigen production I)y tlie proteasoine: Coiriplexity pro\itles efficiency. Ittttrtctnol. Todny 17, 429-435. Groisinan. E. I\.. Chiao, E., Lipps, C. J., and Heffroii, F. (1989). Sdvtotudln fyphiriiuriiirit p h o P \irulence gene is a transcriptional regillator. Proc. Not. Accd. Sci. U S A 86, 7077-
7081. Gruenlicq, J., and Enians, N. (1993). Aiinexins in iiieinl)rane traffic. ?'rtw/.s C d l H i d . 3,224-227. Gruenbcrg, J., and Masfield, F. R . (199.5). Menibfiurc transport in the endocytic pathway. Crcrr. Opit,. Crll &o/. 7, 552--563. Gruenlicid, S.. Pinner. E., Desjardiiis, M.. m t l Cros, P. (1997).Natural resistance to infectioii with intracellu1:ir p~itliogens:The Nfiiinpl protein is recruited to the membrane of the pliagoso~ne./ , Exp. hied. 185, 717-730. C:runwald, U., Fan, X.. Jack, R . S..Work;ileniaIiii. G., Kallies, A , , Stelter, F., and Schiitt, C. ( 1996). Moiiocytes ciin pliagocytose grain-wgative Iiacteria by a CD14-drpentlent ~iiechanisirr. 1,Irtt t r w t o / . 157, 4 119-4 125. Guiney, D. G. ( I 997). Regillation of bactcrid v i r u h c e gciie expression by the host environnient. 1. C h i . Ittoest. 99, 565-569.
Grildrn, P. I I . , Fisclicxr. P.. 111, Slirrman, N. E.. W n n g , W.. Engelhard, V. H., Shabanowitz, J , , Hunt, D. F., and Panier. E. C , (1996). A Listerict ttlottocyfogetic's pentapeptide is presented to cjtolytic T lyniphocytes hy the I I%M3 MHC class I b niolecde. Z t t i t t w r i i f ~ / 5 , 73-79. Gulcr, M . L., C;orliani, J. D., Ilsieh, C:. S., Mkickey, A. J., Steen, R. G . , Dietrich, W.F., untl Murphy, K. M. (1996). Genetic susceptibility to Leishniania: IL-12 respoiisivcness in TH1 cell tle\dopment. Seieticc 271, 984-987. Gulrria, I., Teitell)auni, R,,McAdani, R. A,, Kalpina, 6..Jacobs, W.H., Jr., and Bloom. B. R. ( 1996).Auxotrophic vaccines ft~rtnberculosis. Noturf, Mcd 2, 334-3337. Chpt;i, S., L7011r;1, 11.. Saha, H., Nain, C. K.. :tnd Ganguly. N. K. (1996). Macrophuge-T cell interaction in iiiurine salmonellosis: Selrctive tlo~vn-rc.gulationof I C A M - 1 and B7 inolrculcs in infwtetl macrophages and its prol)oble role in cell-nietliated iminunity. Eitr. J . Z t t i t t t u t w / . 26, 563-570. 1la;is, A. (1998). Rrprograniming tlie pliagocytic ~~~ithw;~~-Intracelliilar pitliogcns antl their vaciioles. Mol. Mcttibr. Biol. In Press. Haas, A., antl Chehel, W.(1992). Microbial strategies to prevent ox).geii-deprncleiit killing by phagocytes. F t w Rodic. RM. Corrttrtrot. 16. 137- 157. Hackuii. L). J.. Rotstein, 0. 11.. Bennett, M. K , , Klip, A,, Grinstein, S., and Manolson, M.F. ( 1996). Characterization and subcrlliilar localization of target menit)rane soliihle NSF uttachinent protein receptors (t-SNAKES) in inacropliiiges: Syitaxiiis 2, 3 , and 4 are prcsent on phagosonial niembranes. J . Z t t i t n i i t t o l . 156, 4377-4383, Eiackaim. D. J.. Hotstein, 0. D., %hang,MI.-J., C~ruenlieitl.S., Gros, P., and Grinstein, S. ( 1998).Host reiistance to iiitracellular infection: Mutation of Natiiral resistaiice-associated macropltage protein 1 (Nrmitpl) impairs phagosonial acidification. 1. ibfcrl. 188, 351 -364. Ilackstadt, T . . Rockey. D. I)., Ikinzen, K. A., and Scitlniore, M. A. (1996). C/ilotrtydi(i tmcltottrtitis interrupts an exocytic patlnvny to ;iccpire endogenously synthesized sphingoinyelin i n transit from tlie Golgi apparatus to tlic plasiiia ~iieniliriine.EMBOJ. 15,964-977. Hackstadt, T., Scidmore, M. A., antl Rockey, D. D. (1995). Lipid nietabolisin in Clilnttiydio fmcliontotis-infected cells: Directed trafficking of Golgi-derived sphiiigolipitls to the clilamydial inclasion. Proc. Not. Acod. Sci. USA 92, 4877-3881. Halperii, M.D., Kiirlander, H. J., and Piscts D. S. (1996). Bacterial DNA induces niuiine iitterferon-y prodiiction by stintillation nterlerikiii-l2 antl tumor necrosis fktor-a.
Cell.
Zttitttiitid
167, 72-78.
IIlinsch, [ I . C., Smith, D. A., Mielke, M. E., Hahn, H., Bancroft, G. J., and Ehlers. S. ( 1996). Mechanisms of granuloina forination i n inurine Mycobmfcrirrtti m i u t t i infection: Thra contrihtion of (2134' T cells. I t i t . Ittitiiitnol. 8, 1299-1310. Ilanseii, T. 11.. a r i d Lee. D. R. (1997).Mechanism ofclass I ;isseinblywith beta 2 microglobuhi and loading with peptide. Ado. I t t i t t w t i o / . 64, 105-137. H a ~ i T., . Mizuno, Y., Takaki, K., T a k a d a , El.. Akeda, H., Aoki, T., Naguta, M., Ueda, K.. Matsuzaki, G., Yosliikai. Y., at N / . (1992). Predominant ;ictivation antl expansioit of Vy9bearing y6 T cclls in vivo iis well as in vitro i n Salmonella infection. J . Clitt. Ztice.c.t. 90, 204-210. Harljoe. M., Oettiiiger, T.,Wiker, H. G., Hoseiikrands. I., and A n d m e for occiirrenc~~ of the E S A T - 6 protein in M!/cobacterirrt~tftrbc~rcrh / m T c z r i u t t i />(ttjis ;ind for its ahsence in A~!/cohortPrit~tt~ h i s . BCC;. Infict. Ittwmtt. 64,
16-22. Harding, C. V., Hoof, R. W., Allen, P. M., and Unanue, E. R. (1991). Effects of pH and polysaccliariclc,s on pepticlc binding to class 11 iiiajor Iiistoconipiitibilitycornplex inolecules. Proc. Not. Acnrl. Sci. USA 88, 2740-2744.
Hardiiig. C:. 1' , ;ind Song, K.( 1994). Pliagocytic processing of exogenous particulate antigens by niacrophages for prrsentotion by class I MHC niolrciiles.]. Z t t i t t w t m l . 153,49255493333, Harrison, J. A, \'illarreal-Rainos. B., Xlastroeni, P.. Ileiiiarco de Horniaeche, K., ant1 Morinaeche. C . E. (1997). C;orrehtcls of protection induced by live Aro- Snlttrottelln f!yp/iittiifriwti vaccines in the niurine t)phoitl iiiodel. Z r r t / t t i t i i o h g ! / 90, 618-625. Hnrrison. P. M., and Arosio, P. (1Y96). The ferritiris: Molecular properties, iron storage . 1275, 1631-20:3. fiinction ;in11crlliilar rrgulation. Biocltitrc. f i i o ~ j / r ! y ~.-\ctc/ ILii-tli, G., Clcinens, D. L.. ; i d Honvitz,. M.A. ( 1994).(;lutan>ine synthetase ofhfgc:oboctcrittrtc frrbrtu.tr(rais: Extracellii!ar release. antl c1i;iracterization oi'its enzyinatic activiiy. Proc. Mot. Atatl. Sci. U S A 91, 9:342-9346. Kirt!., J . T.. and Bevan. M. J , (1992). CD8' T cells sprcific for u single nonanier epitope of Iistrv-in ttiottoc.!/togcne,sLire protecti\.c. in vivo. J . E.Y/I..!fee/. 175, lfj-31-1538. Hart\.. J . T., antl Bevan. M. J . (1996).CD8T cell recognition ot'niacrophages and hepatoeytes rcwilts i n ininiunitv to Listetirr tiiotioc'!ytogr,,Ics I t f w t . Z t t i t t i t t t i . 64, 3632-336340, Hart\.. J. T., L ~ n z L. . L.. and Bevan, M. J. ( 19963).l'rinian, and secondary immune responses to Z,isfctic/ ttiotiocgtog[~ti[,.s.Ctrrr. Opiti. Itiitiiutiol. 8, 5%-5:30. Hasan, Z., Scltl~u,C . , K i i l i n , L., L e f h i t s . I., Yoiiiig, I),, Thoks. J., and Pirters, J. (1997). 1sol:ttion antl clianictrrizatiori of tlie niycohacteriul phagosonie: Segregation froin tlit. endosoitia~l~sosoi~i~il pathway. Mol. ,!ficwbiol. 24, 545-55:3. Haslov, K., Antlersen. A,, Nagai, S . , Gottschau, A , . Sormsen, T., and Andersen, P. (1995). Guinea pig celliilar inimnne responses to proteins sccreted by hfycobact~~riiitrt tidwrcihsis Zt!fr.ct. Z t t t t t i i u t . 63, 804-810. Havcll, E. A,, ;tnd Sehgal, P. B. (1991).Tiiiiior necrosis thctor-indrpendent IL-6 production thiring m i i r i n c Iisteriosis. J . I t r u r i r o i o l . 146, 756-761. Hayishi, T., Catanztro, A , , ~ n dI k . S. P. ( 1997). Apoptosis of human inonocytes and inacroplinge< by Mycobnrfcrirr t t i uoircttr sonic;itc,. Ztfict. Zttcttucti, 65, 5262-5271. Heeg. K . . Knon. M'.. and Wagner, II. (1991 ). Vaccinatioii of cliiss I major liist[icoinpatihili~ compleu ( !vI HC-restricted niurine CDX' qtotoxic T lynipliocytes towards soluble antigems: Ininiiiiiostiniulatiiig-ovalbuminconiplrws enter the c l ~ Is MHC-restricted antigcw patliway antl allo\v sensitization against the i i i i i i i i i r i c ) c l o i n i i i ~ i n t peptitle. Eur. J . I t t i t t i i i r i e d , 21, lC52:3-t527. IIeinzen, K. A., Scidmore, tvl. A,. Rocky, I). 11.. and IIackstadt, T. (1996). Differential interactioii \kith rndocytic and exocytic. pttli~vaysdistiiigiiisli parasitophorous vaciioles of Cosicllo but-rietii and Chkrtryrlio trocliotiwtis. Ztfict. Z j t i t t i u t i . 64, 796-809. Hensc.1, M., SlLea. J. E., Gleeson, C., Joiit's, M. D.. Ddton, E., and Holden, D. W. (1995). Siniiiltaneous identification of l)actc,rial \irnleiicr grnes by negative selection. Science 269, 400-403. Hess, J., Dirtrich. G . , C:entsclicv, I.. Miko, D., Goel)el. M'., and Kaufniann, S. H. E. (19%). Protcwtion against niririne listeriosis by air attcwuated recombinant Snbtiotir&r t ! / , d i i t t c t i t i ~ i t t i\.;icciiiestrain that secretes thr, iiatiirally somatic antigen superoxide clisniutasci. Zrficf. Z t r t t t i i c t t . 65, 128&1292. Hess, J.. Ckwtscliev, I., Miko, D., Welzel. M., Latlel, C., Goebel, W., arid Kaufiriaiin, S. H. E. ( lW6a). Siiperior efficacy ot'srcretetl over soiiiatic antigen display in recombinant S c d t t t o t t d l o \wcinr, induced protcvtion against listrriosis. Proc. N o t . Actrd. Sci. USA 93, 1458- 1463:3. Hess, J . , Latlel, C.. Miko. D.. and Kaufinanii, S. 1-1. E. (I996b).Sn/trionc7h f ! / p / i i t t u r r i u r t r aroAinfection in gene targeted immunodeficient mice: Maijor I-oleof CD4 ' TCR-a/3 cells and IFN-y i n 1)acterial clearance indepeiidriit of intracc-lliilar 1ocS;ition.J. Z t n t t m r i o L . 156,3321 :3:326.
Hess, J., Miko. D., Catic, A,, Lehmensiek, V.. Knssell, D. G., and Kanfmann, S. 11. E. (1998). Mycobocteritmt boois bacille Calmette-Guerin strains secreting Iisteriolysin of Listerio ttu)iiocytogetie.s. Proc. N o t . Acod Sci. USA 95, 5299-5304. Hewitt, E. W., Trenmann, A,, Morrice, N., Tatnell. P. J., Kay, J., and Watts, C. (1997).Natural processing sites for human cathepsin E and cxthepsin D in tetanus toxin: Iinp~icationsfor T cell epitope generation. /. I t i i t u u t w l . 159, 4693-4699. EIickey, E. K., and Cianciotto, N. P. (1997). A n iron- and fur-repressed Legionello ptiet’rtiop l i i k r gene that promotes intracellular infection and encodes a protein with similarity to the E.schcrichio ndi aerobiictin synthetases. Zrgect. Znin~un.65, 133- 143. Iliroinatsn, K., Yoshikai, Y., Matsuzaki, G., Ohga, S., Muratnori, K., Matsulrloto, K., Bluestone, J. A,, and Nomoto, K. (1992). A protective role of y6 T cells in primary infection with Listeria itumoc!/togetws in mice. J. Exp. Med. 175, 49-56. Hirsch, E., Itikura, V. M., Paul. S. M., and Hirsh, D. (1996). Functions of interleukin 1 receptor antagonist in gene knockout and overproducing mice. Proc. Not. Amd. Sci. USA 93, 11008-11013. Holniskov, U., Lawson, P.. Teisner, B., Tornoe. I., Willis, A. C., Morgan, C., Koch, C.. and Reid, K. B. (1997). Isolation and cliaracterization of a new niember of the scavenger receptor snperfamily, glycoprotein-340 (gp-340), a s a lung siirfactant protein-D binding niolecnle. J. Biol. Cheni. 272, 13743- 13749. Homutlr, M., Valentin-Weigand. P.. Rolide, M., and Gerlacli, G. F. (1998). Identification and characterization of a novel extracellular ferric reductase from M!ycobncteriuni pnratrcDrrculosis. Zrfect. Initnun. 66, 710-716. Hoppe, H. C., de Wet, B. J., Cywes, C., Daffe, M., and Elders, M. R. (1997). Identification of ~~I~os~~liatidylinositol mannoside BS H inycobxterial adhesin mediating both direct and opsoiiic binding to nonphagocytic mammalim cells. Zrfect. Z i t i t n u n . 65, 3896-3905. Hormaeche, C. E., Joysey, H. S . , Desilva, L., Izliur, M., and Stocker, B. A. (1991). Immunity conferred by Aro- Saltnonella live vaccines. Microb. Puthogen. 10, 149-158. Horuitz, M. A. (1983). The Legionnaires’ disease bacterium legion elk^ p n e u n q h i h ) inhibits pliagosoiiie-lysosoine fusion in Iinman tnonocytes.J. Exp. Alerl. 158,2108-21 26. IIotwitz, M. A. (1984). Phagocytosis of the Legionnaires’ disease bacterium (Legionelln ptieutiiophila) occurs by a novel niechanisin: engulfinent within a pseudopod coil. Cell 36, 27-33, Honvitz, M. A., Lee, B. W.. Dillon, B.J.. and Harth, G. (1995). Protective immunity against tuberculosis induced by vaccination with major extracellular proteins of A.iycob&riutn titbcrcdosis. Proc. Nut. Acad. Sci. USA 92, 1.530-1534. IIosieth, S. K., and Stocker, B. A. (1981). Aromatic-dependent Soltilotiella typhittiurinni are notivirulent and effective as live vaccines. Notiire 291, 238-239. IIsieh, B., Schrenzel. M. D., Mulvania, T.. Lepper, H. D., DiMolfetto-Landon, L., and Ferrick, D. A. (1996). In vivo cytokine production in inurine listeriosis: Evidence for itnmunoregulation by y6‘ T cells. J. Z t t i t w n d . 156, 232-237. 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. Hubbard, R. D.. Floty, C. M., and Collins, F. M. (1992). Immunization of mice with tnycobacterial culture filtrate proteins. Clin. Exp. Zniniuriol. 87, 94-98. Huller, L. A., and Peters, M. E. (1994). Mapping sinall GTP-binding proteins on high resolution two ditnensional gels by a coinbination of GTP binding and liibelling with in situ periodate oxidised GTP. Electrophoresis 15, 283-288. Huffman, G. R., Nataraj, C., and Kiirlander, R. J. (1996). Novel pathway for class I-restricted presentation of an exogenous, non peptide Listeria itiotiocytogenes antigen to CD8’ T cells. FASEB J. 10, A1309.
Ilunter. C. A.. Chizxonite. K..and Remiiigtoii, 1. S. (199.5). IL-1 /3 is required for IL-12 to induce production of IFN-y by NK cells: A role for IL-I /3 in the T cell-irid(~pendelit meclianisni o f resistance against intr;tcellul;ir pathogens. J . ~ l ? l t t l u t t ( J ~155, . 4:347-4:354. Hnnter. f:. A.. Tinims, J.. Pisacane, P.. Menon. S.. G i i , G.,Wdker, W..Astc-Anie.zaga, M . , Chizzonitr. R.. 13azan. J. F.. and Kastelein. R. A . (19971. C o n p i i s o n of the effects of interleukin- la. interleukin- 18 antl iiiterferoii-gaiiiina-intlticing factor on the protllwtion of iriterferon-jianiina by natural killer cells. Errr. J . Z t t i t r t t ~ t t ~ ~27, / . 2787-2792. 1Iuygr11,K.. Content. J.. Denis. 0 . .Montgonieiy, D. I,.. Yawman. A. M.. Deck, H. R., Deil'itt, C. M., Ornic,, I. M., Baldwin. S.. 13'Soiiza, C:.~D r w w t , A , , Lozrs, E., Vantlenl~usscl~e. P., Van Vooreii, J. P., Liu. M. A., and Uliiirr, J. B. (19%). Iiiiinuiiogc,nicitv :ind protcctive c4'ficacy of ii tuberciilosis DNA \wciiic.. Notirrc Mcd. 2 , 893-898. Igietsemr. J. IJ.,Perry. L. L., Ananaba. (;. A , . Uiiri, I . M., Ojior, 0 . 0..Kuinar, S. N., and Caldwrll. H . D. (1998). Clilaiiiydial infi,ctictn i i i inducil)le nitric oxide syithuse knockout inice. It!ff.c.i. Z t u t m t t i . 66, 128% 1286. Iniani, F., and Solosh. M . J. (1991). Heat sliock proteins ciin wgulatc expressioii of thr T1;i region-mcoded c l m Ib niolecule Qa-1, PIYJC. h'rrf.:\cvr/. Sci. C'SA 88, 10475-10479. Ireton, K . , ~ n i t Cossart. l P. (1997). IIost-puthogen iiitcwctions tltiiing ent? iincl actin-hascd nro\zwent of Lisfvrin tttotioc,!/togrric..v.Atttiri Rc~t:.Crttrt. 31, 1 13- 138. Isherg, I<. R., mid \':in Nhieu, G. T. ( 19$J.5).Tlie iiirchanisni of phagocytic uptake proinotrd by invasin-integrin interaction. Trcttrk Cell B~(J/. 5 , 12U-124, Iwasaki, .4,, Stiemliolni. B. J.. Chaii, A . K.,Berinstein, N , L., and Barber, B. 13. (1997). Enhanced CTL rcqonses incdiated !)I plasniid D N A ininiunogens encoding costimulatoq~ niolecules antl c)?okinrs. J. Z t t t t t i t ~ n d 158, 4,591-3601, Jackman, R. M., Stenger. S., Lee, A,. M ~ ~ o d D. y , B.. Rogers. R. A., Niazi, K. R., Sugita. bl.. Modlin, H. L., Peters, P. J., and l'orcelli, S. A. (1998). The tyrosinr-containing cytoplasntic tail od CDlI) is essentid for its ef'ficiciit prcwntation ofbacterial lipid antigem. Z t i ~ t t t i r t i i t 8, ~ ~ :341-351. Jacol)s~n.N. C,, S z h ) , S. J., \\'el)cr-Nordt, I{. M., Zhong, Z., Schreiber, R. 11.. 13arnc41, J. E., Jr,, and Mnrpliy, I<. M. (1995). Interleukin 12 signaling in T helpcr h ~ 1 e( T h l ) cells involves tyrosine pliospliol?ilatioii of signal transducer and activator of transcription (Stat)3 and Stiit4. J . Exp. M d . 181, 17*5,5-1762. Janis, E. M., K;aufinann. S. 11. E.. Schwartz. 11. 11.. and Partloll. D. M. (1989). Actiwtion of' yS T cells i n the primar) immune response to Jf!/cohnctcriritt~ tchercri/osi.9. Scierlcc 244, 713-716. Jenkins. M. K. (1993). The ups m t f d o w i ~ sof' T ct.Il costiinulation. Z t t ~ t ~ t f r t ~1,i f 343-446. ~ Jensen. E.R., Shen. H., Wettstein, F. 0 . .Alrinrd. I<., and Miller. J . F. (1995).Reconil,ii~ciiit Listericr tru,rioc!/f~)getic~,~ as a live vacciiie \xaliicle antl a prol)e for stud~ingcell-nietliatf~d ininirinity. I / t t r t i r m o l . RPC. 158, 147-lS57, Jepson, .4.. Banya, \l',. Sisay-Joof, F., Hassan-King, bl.. Nunes. C., Bennett. S., and Whittle, El. (1997). Quantification of thr relati\.e contrihition of major Iiistocoinpatit)ility complex ( M I I C ) and non-MIIC genes to I I ~ I I Iimmune ~ I ~ rcymisc's to fowigri antigens. Zttfict. Z T ~ I T ~ U ~65, I I . 872-876. Jiang, N',, Swiggard. \V. I., HcriHer, C.. Prng. M., Mirza, A,, Steiiirnan, R. M., antl Nussenzweig, M. C. ( 1995). Thcs receptor DEC-205 expressed by dendritic cells and thymic epitl1rli;tl cells is involved i n antigen pr(iccssing. Nrjtrrw 375, 151-155. Jiang, X.,ltonard, B., Benson, R . , and Baldm~n,C. 1,. (1993). Macropliage control of Bnic.rki r r h r t r r s : Role of reactive nitrogen iiiternietliates and nitric oxide. cd/.~ t t I t f t t i t l ~ J / . 151, :309-3 l9. Johnson. C . M . , C:oopvr, A. M., Frank, A . A,, mtl Orme. I. M. (1998). Atlequate expression of protective iniinilni? i n the alxence ot' granuloma formation on hi!l(,"b~rctcn'rrtrt
354
UL,RICII E. SCIHAIBLE et nl
tziberclilosis-infectedinice with a disruption in the intracellular adhesion niolecule 1 gene. Infect. Znrrnun. 66, 1666- 1670. Johnson, J. L., Shiratsuchi, H., Toossi, Z., and Elher, J. J. (1997). Altered IL-1 expression and compartmentalization in inonocytes from patients with AIDS stimulated with Mycobacterinrn auiurri complex. J, Clin. Zrr~nwnol.17, 387-395 Joiner, K. A. (1997). Membrane-protein traffic in pathogen-infected cells. J. Chn. Inuest. 99, 1814-1817. Joiner, K. A,, <;anz, T., Albert, J.. and Rotrosen, D. (1989). The opsonizing hgand on Snlrnonella t!yphimuriuln inHuences incorporation of specific, but not azurophil, granule constituents into neutrophil phagosomes. J. Cell B i d . 109, 2771-2782. Jondal, M., Schirmbeck, R., and Reimann, J. (1996). MHC class I-restricted CTL responses to exogenous antigens. Znittrutiity 5, 295-302. Jones, B. D., and Falkow, S. (1996). Salmonellosis: Host immune responses and bacterial virulence determinants. Ann. Reo. Zmmmol. 14, 533-561. Jones, B. D., Paterson, H. F., Hall. A,, and Falkow, S. (1993). Sohwtiella typhiitiurittin induces membrane ruffling by a growth factor-receptor-indepeiident mechanism. Proc. Not. Acad. Sci. USA 90, 10390-10394. Jouanguy, E., Altare, F., Lanihamedi, S., Rev, P., Emile, J. F.. Newport, M., Levin, M., Blanche, S., Selioun, E., Fischer, A., and Casanova, J. L. (1996). Interferon-y-receptor deficiency in an infant with fatal bacik Cahnette-Guerin infection. N . Engl. 1. Med. 335, 1956-1961. Joyce, S . , Woods, A. S., Yewdell, J. W., Bennink, J. R., Desilva, A. D., Boesteanu, A., Balk, S. P., Cotter, R. J., and Hrutkiewicz, R. R. (1998). Natural ligand of mouse C D l d l : Cellular glycosyIphosphatidylinositol. Science 279, 1541-1544. Kabelitz, D., Bender, A., Prospero, T., Wesselhorg, S., Janssen, O., and Pechhold, K. (1991). The primary response of human ylS’ T cells to Mycobacterium tuberarlosis is restricted to Vy9-bearing cells. J. E x p . Med. 173, 1331-1338. Kagi, D., Lederrnann, B., Burki, K., Hengartner, H.,and Zinkernagel, R. M. (1994). CD8’ T cell-mediated protection against an intracellular bacterium by perforin-dependent cytotoxicity. Eur. j . Zinrriunol. 24, 3068-3072. Kamiinaka, M., Yu, P., Yasui, T., Yoshida, K., Kawabe, T., Horii, T., Kishimoto, T., and Kikutani, H. (1996).Protective role ofCD40 in Leishtnanici niajor infection at two distinct phases of cell-mediated immunity. Ztnnntnity 4, 275-281. Kaplan, M. H., Sun, Y., Hoey, T., and Crusby, M. J. (1996). Impaired IL-12 responses and enhanced development of Th2 cells in Stat4-deficient mice. Nature 382, 174-177. Kataoka, T., Yarnanloto, S., Yamamoto, T., Kuramoto, E., Kimura, Y., Yano, O., and Tokunaga, T. (1992). Antitumor activity of synthetic oligonucleotides with sequences from cDNA encoding proteins of Mycobactt~riunzbocis BCC. japan. J. Cancer Res. 83, 244-247. Kaufinann, S. H. E. (1988). CD8’ T lyniphocytes in intracellular microbial infections. Zrnniunol. Today 9, 168-174. Kaufmann, S. II. E. (1996). gainmddelta and other unconventional T lymphocytes: What do they see and what do they do? Proc. Nat. Acad. Sci. USA 93, 2272-2279. Kaufniann, S. H. E., and Andersen, P. (1998). Immunity to rnycobacteria with emphasis on tuberculosis: Irnplicatioiis for rational d e s i p of an effective tuberculosis vaccine. In > Chemical Immunology: Immunology of Intracellular Parasitism’’ (F. Y. Liew, ed.), pp, 21-59. Karger, Basel. Kaufrnann, S. 11. E., Emoto, M., Szalay. G., Barsig, J., and Flesch, I. E. (1997). Interleukin 4 and Listeriosis. Zmmutiol. Reu. 158, 95-105. ‘I
Karifinann, S. H. E.. and Ladel, C. H. (1994). Role o f T cell subsets in inimunity against intracehlilr hactcria: Experiniental irifections of knock-out mice \vi t h Li,stct%c triotiocytogt'ties and Al!/coh(icturirtttthoris BCC;. I j t I j r i l [ r ~ i ~ ~ ~91, i ~ ~509-519. [ ) ~ y Kaufiiiann, S. II. E., Kodewak~,H. It., Ilug, E., and DeLibcro, G. (1988). Cloned Lktrrio t~u)noc!/to~enc.s specific non-MHC-restricted Lyt2' T cells with cytolytic and protective activity. 1.Itntrzrcnril. 140, 3173-:3179. Kawano, T., ( h i , J., Koezuka, Y., Toura, I., Kaneko, Y.,Motoki, K., Ueiio, H., Nakagawa, R., Sato, H.. Kondo, E.. Koseki. H., and Taniguchi. $1. (1997).CDld-restricted aiitl TCRmediated activation of \'a14 N K T cells by glycosylcerainicles.Sciutice 278, 1626- 1629. Kaye, P. M . ( 1 995). Costimulation and thc, regulation o f antimicrobial immunity. I m t t t i t n o l . T~d(i!/16, 423-427. Kaye. P. M.. Rogers, N. J.. Curly. A . J.. and Scott. J . C . (1994). Deficient expression of costiniiilatoiy molecules o n Leishmaiiia-iitfcctrtl niacrophages. Eur. /. I t n t n u n o l . 24, 2850-2854. Ke. Y., and Kapp. J. A . i1996).Exogenous antigens gain x c e s s to the major l~istocom1)atiI~ili~ complex class I processing pathway in B cells by receptor mecli;ited uptake. J. E x p Med. 156, 1179-1184. Kiina, P. E., Ruddle. N . H.. and McMahon-Pratt. D. (3997). Presentation via the class I pathway by Leishtnciriia nt,in^otieri.si.s-iiifected macrophages of an endogenoils leishmania1 antigen to CD8+ T cells. /. Ittmutio!, 159, 1828-18:34. Kindler. V.,Sappino. A. P., Grau. G . E.. Pigiwt. P. F.. aiid Vassalli, P. (1989). The indiicing role of tumor necrosis factor in the devc4opment of bactericidal granuloinas during RCG infection. C d l 56, 731-740. Klingler, K., Tchoti-Wong, K.-M., Briiidi, O., Aston. C., Kini, R., Chi, C., and Roin, \Y. N. (1997). Effects of mycohacteria on regulation of apoptosis in inononuclear pliagocytes. Infect l i t i t t i i i t i . 65, 5272-5278. Klinman, D. M., Yi, A . K., Beailcage, S. L., Conover, J., and Krieg, A. M. (1996). CpG motifs present in bacteria D N A rapidly indiice lyinphocytes to secrete interleukin 6 , interleukin 12. and interferon-y. Proc. Not. Actid. Sci. USA 93, 2879-2883. Kobayishi, T.. Stang, E., Fang, K. S., dr Moerloosc, P., Parton, R. G., and C;ruenberg, J . (1998). A lipid associated with the antiphospholipid s y d r o i n e regulates endosoine structure and fiinction. Nature 392, 193-197. Kocks, C., Gouin, E.. Taltouret, M., Berclie, P., Ohayon, H., and Cossart, P. (1992). L. tnotioc!/togeties-indiice~~ actin asseinbly requires the x t A gene product, a surface protein. Cell 68, 521-531. Kopf, M., Barimann, H., Freer. G., Freiidenberg, M., Lariirrs, M., Kisliimoto, T., Zinkenragel, R.. Bluetliniann. H., and Ktihler, G. ( 1994). Impaired iinnrune and acute-phase rcsponses in interleukin-6-deficient niice. Notrrrc 368, 339-342. Koppelman, B , NeeQes, J. J.. de Vries. J. E., and de Wad Malefyt, R. (1997). Interkuhn10 dow-regulates MHC Class I1 a/3 peptide complexes at the plasma membrane of monocytes by affecting arrival and re.c.ycling.I n i m i n i t y 7, 861-871. Konifeltl, S.,and Mellinan. I. ( 1989). The hiogenesis of lysosolnes. Anrrrr. Rel;. Cell Biol. 5 , 180-525. Kovacso\ics-Bankowski, M., and Rock. K. L. (1995).A pliagosoine-to-cytosol pathway for ' I molecules. Science 267, 243-246. exogenous ailtigens presented on MHC Krieger. M.. and Herz, J. (1994). Structures and functions o f multiligand lipoprotein receptors: inacropliage scavenger receptors and LDL receptor-related protein (LRP). Annu Rcw. Biocho!i. 63, 601-637. Kurashima. K. Mukaida, Ii..Fujimiira, M., Yasiii. M., Nakazllmi, Y., Matsuda, T., and Matsushin~a,K. (1997). Elevated cheiiiokine levels in hronchoalveolar lavage fluid of tuhercrilosis patients. Atn. I. R q . Crit. Cnre Mrd. 155, 1474-1477.
356
ULIIICll E. SCHAJHLE rt 01
Kurihara, T., Warr, G., Luy, J., antl Bravo, R. (1997).Defects in macrophage recruitment and host defense in mice lacking the CCR2 cheinokine recept0r.J. Exp. Mcd. 186,1757-1762. Kurlander, R., and Nataraj, C. (1997). Characterization of the niurine H2-M3wt-restricted C1)8 response against a hydrophobic, protease-resistant, pliospliolipid-associat~[lantigen from Listen" ortor~ncytogerics.Imtnurd. Rev. 158, 123-128. Kurlantler, R. J., Shawar, S. M., Brown, M. L., a i d Rich, R. R. (1992). Specialized role for a muriiie class 1-11 MHC molecule in prokayotic host defenses. Science 257, 678-679. Ladel, C. H., Blum, C., Dreher, A,, Reifenberg, K., and Kaufmann, S. H. E. (1995~). Protective role ofyST cells and cub T c e k in tuberculosis. Ettr./. Intmunol. 25,2877-2881. Ladel, C. H., Blum, C., Dreher, A,, Reifenberg, K., Kopf, M., and Kaufinann, S. H. E. (1897). Lethal tuberculosis in interleukin-6-deficient mutant mice. Itlfect. Ztnmrrn. 65, 4843-4849. Ladel. C. €I,, Blum, C., and Kaufinann, S. H . E. (1996). Control of natural killer cellmediated innate resistance against the intracellular pathogen. Listerici trionocytogenes by yS T lymphocytes. Zifect. Zrnnzun. 64, 1744- 1749. Ladel, C. H., Daugelat, S . , and Kaufinann, S.H. E. (1995b).Immune response to MycobnLYen'um bocia bacille Cahnette Guerin infection in major histocompatibihty complex class 1- and 11-deficient knock-out mice: Contribution of CD4 and CD8 T cells to acquired resistance. Errr. I. Zrririirrtzol. 25, 377-384. Ladel, C. H., Flesch, I. E., Amoldi, J., and Kaufmann, S. H. E. (1994). Studies with MHCdeficient knock-out mice reveal impact of both MHC I- and MHC 11-dependent T cell responses on Listeria rrioriorytogenc.~infection. J . Zmtriitnol. 153, 31 16-3122. Lah, T. T., Hawley, M., Rock, K. L., and Goldherg, A. L. (1995). Gamma-interferon causes a selective induction of the lysosoinal proteases, cathepsins B and L, in macrophages. FEBS Lett. 363, 85-89. Lalvani, A,, Brookes, R., Wilkinson, R. J., Mdin, 11.S., Pathan, A. A,, Andersen. P.. Dockrell, H., Pasvol, G., and Hill, A. V. (1998). Human cytolytic and interferon y-secreting CD8' T lymphocytes specific for Mycobncter-iiiiti tubrrarlosis. Pmc. Not. Acad. Sci. USA 95, 270-275. Lammas, D. A,, Stoher, C., Harvey, C. J., Kendrick, N., Panchalingam, S., a i d Kumararatne, D. S. (1997). ATP-induced killing of niycobacteria by human macrophages is mediated hy purinergic P2Z(P2X7)receptors. Itrmunity 7, 433-444. Lang, T., de Chastellier, C., Frehel, C., Hellio, R., Metezeau, P., Leao, S., and Antoine, J. C. (1994). Distributioii of MHC class I and of MHC class I1 molecules in macrophages infected with Leishtrtania cinmzoneiisis. /. Cell Sci. 107, 69-82. Lang, T., Prim, E., Sibtliorpe, D., and Blackwell,J. M. (1997).Nrampl transfection transfers Ity/Lsh/Bcg-relatecl pleiotropic effects on macrophage activation: Influence on antigen processing and presentation. Infect. Z n w i w t . 65, 380-386. Langerniann, S., Palaszynski, S., Satlziene, A,, Stover, C. K., and Koenig, S. (1994). Systemic and mucosal immunity induced by BCG vector expressing outer-surface protein A of Borrdia htirgrlor$et-i. Nuture 372, 552-555. Langermans, I. A., and van Furth, R . (1994).Cytokines and the host defense against Livtcrin rru)nocytogene.uand Soltrwnella typhi7nnrium. Biotherapy 7, 169-178. Laochumroonvorapong, P., Wang, J., Liu, C. C., Ye, W., Moreira, A. L., Elkon, K. B., Freedman, V. H., and Kaplan, G. (1997). Perforin, a cytotoxic molecule which mediates cell necrosis, is not required for the early control of mycobacterial infection in mice. I7lfect. Zmtnrm 65, 127-132. Lee. B. Y., antl HoTuitz. M. A. (1995). Identification of macrophage and stress-induced proteins of Mycobucteriuni tuharcdosis. 1. C h . Znue.st. 96, 245-249.
Lee, D. J.. Abqwtiic,, A.. Carson. D. A.. and Corr. hil. (1998).lntliiction o f m antigen-specific 187, 4:3:3-4:38. CD1-rrstrictcd c!Zotosic T lymphoc?te respoiisc i t t uioo. J . Esp. M~Y/. Lenz. L. L.. and B ( w n , M. J. (1996). H L M 3 restiicted presentiition of 1 k t w i o titottoe!/top i t ' s antigens. Z t ~ i r i i r r t i d Rro. 151, 107-121, Lenz. L. L., i m t l He\un. M . J. (1997).CTL respoiiws to I12-M:3 restricted Listeria epitopes. I t n t ~ t r r r r o l .Rro. 158, 115-121. Lciiz. L. L.. Dere. B.,antl Bcwui. M . 1. ( 19%). Itlentificatioii of ail FI2-M:3-restrictetllisteria c,pitope: Iniplications for antigen preseiitation hy M:3. Zttittirmif!y 5 , 6:3-72. Leonard,B. A., Lopez-Goni. I.. and BaItl\ziii. C. I,. (1997).Bnicclla nbortrts sidrrophorv 2.3tlihydrox?.beiizoicacid protects 1)nicellaefroni killing by inacrophages.Vct.RL's. 28,87-92. Lewiiisohn. D. M., Aldcrson. M . R., Britlen, A . I,.. Ridclell, S. R..Reed. S. C,, antl Gral)steiii, K. 13. (199b). C:lia~tcterization of I l i i i n a n (:l>X' T cells reactive with ,Ilycnhncfrtirrtti ttrbc,rcrr/osi.~-infectetl antigc.ii presentiiig cells. J. Esp. A I d 187, 1633- 1640. Lin, J., antl Ficlit. T. A. (1995). Protein synthesis i n Bt7ccdk/ d ~ ~ r t iiidriced t~s during macrophage infection. 1tdk.t. 1 t t t t w t t . 63, 1409-1414. Liii, Y., Zhang M.. antl Barnes, P. F. ( 1998). Cheinokine production by a Iiiiman ulveolar epitlielial cell line i n respoiiw to ~ ~ I / ~ . ~ i / ~ ~ i ctrrbc~t-cirlo.sis. f ~ ~ t ~ i t t t i ih$ect. I t n t n i r n . 66, 1121 1126. Liiidahl. K. F.. Byers, 11.E., Dabhi. V. M.. Hovik. R.. Jones, E. P., Smith, C:. P., Wang, C . R.. Xiao, H., ant1 Yoshino. M. (1997).H2-M3. ii frill-senicr class I b histocoiiipati1)iIihi ~llltigell.L4rLtf11. &I;. f t t l t t t f l t l ( J / , 15, 851-879. 11 conipleues genenitcd Lindiier, R.. and Unaniie, E. R. (1996).Distinct antigen MllC b?. separiitc processing pathways. EAfBO J . 15, 6810-6920. Lipford, G. R.. Spiwasser. T., Bailer. M . , Zimrnermaiin. S., Koch, E. S., IIeeg, K.. iind L2'agnc.r. 11. (1997).Iinmriiiostiiniilatol-). DNA: srciiieiic(~-depericl[~nt prodiiction of potentially hariiifiil or useful qtokines. Etrr. J . Z t n t u r r t i d . 27, 3420-3426. Liu, J., Fiijiwara, T. M., Briii, N. T., Sanchez, F. O., C c h r , M., Paradis, A. J., Frappier. D., Skamenc., E., Gros, P.. Morgan. K.. et a[. (1995). I(1entification of polymorplrisnIs antl secpencr variiints in tlir Irurii;in Iioiiiologuc of the nioiise natural i-esistaircr,-associiitcd macrophage protein gene. Ant. /. Htrttt. GcttcJt.56, 845-853. Liu. Z., Simpson, R. J., and Cheers, C . ( 1994). Kole of IL-6 in activation of T cells for J . Z t t ~ t t w t o / . 152, 5375-53380. acquired celliilar resistance to Listerio ttiotioc!/togett~,~. Lobigs, M . , and Mullbacher. I\. (1993).The M I IC encotkt1 peptide transporters for class I MI IC: are lixnctioiiiilly not polyniorphic. P n x . Nnt. Auitl. Sei. USA 90, 2676-2680. Imckslry. R. M., antl Scott, P. (3991). Helper T-cell subscts in iiioiisc Ic~ishmaiiiasis:Induction, expansion and effi'ctor function. Z t l i t t u r t i d . Totkq 12, A58-61. I,oetscher, P.. Ugiiccioni, M.. 13ordoli, L., B;iggiolini, M., Moser, B.. Chizzolini, C., and Daym-. J. M. (1998).CCR.5 is cliaractc,ristic of Thl Iymphocytes. Nntrrrc 391, 344-345. Liihning. M.. Stroehmanii. A,. Coylc. A. J.. Grogoa, J. L., Lin, S.. Guttierez-Ramos, J.-C., I,e\inson. D.. Radbnich, A,, and Kiinirxlt. 1'.(1998).TlISTB is prelbntially expressed on iiiurinr TI12 cells independent of interleukin 4. interleiikin 5, and interleukin 10, antl inipoitant h r Th2 effector function. Proc. &'of/. Aced Sci. USA 95, 6930-6935. L,opcz-Macias, C., 1,opez-llrriiandez. M. A.. Gonzalez. C. R.. Isilxisi. A,. and OrtizNiwarrete. 1' (1995). lndiiction of iinti1)odies against Snhioticlln t y p k i OiiipC poiin by naked DNA ininirinization. Attti. N . 1..Arcid. Sci. 772, 285-288. Lri, 13.. Rutledqe. B. J., Gu. L.. Fiorilla. L.. Lukacs. N. b'.,K ~ n k e l .S. Id., North. R., C;er;ird, C.. d Rollins, B. J. ( I Y Y X ) . Abiiorinalities in inoiiocyte recruitment and cytokine exprcwion in nionocyte c1ieiiioattract;uit protein-1 deficient mice. j . Esp. Mcd. 187, 601-608.
358
ULRICH I.: S(:HAIBLE et 01.
Lu, J. (1997). Collectins: Collectors of microorganisms for the innate immune system. Bioessuys 19, 509-518. Lucey. D. R., Clerici, M., and Shearer, G. M. (1996).Type 1 and type 2 cytokine dysregulation in human infectioris, neoplastic, and inflanimatory diseases. Clirr. Microbid Her;. 9, 532-562. Luke, C. J., Carner, K., Liang, X., and Barlmur, A. G. (1997). An OspA-based DNA vaccine protects mice against infection with Borrelin h i t r g d o r j k J. It@. Dis. 175, 91-97. Macatonia, S. E.. Hosken, N . A,, Litton. M., j'ieira, P., Hsieh, C . S.. Culpepper, J. A., Wysocka, M., Trinchieri, G., Murphy, K. M., and O'Garra, A. (1995). Dendritic cells produce 11,-12 and direct the development of Tlil cells from naive CD4' T cells. J. Imtntmol. 154, 5071-5079. MacMicking, J. D., Nathan, C., Hom, G . , Chartrain, N . , Fletcher, D. S., Trumbauer, M., Stevens, K., Xie, Q. W., Sokol, K., Hutchinson, N., et id. (1995). Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide syithase. Cell 81, 641-650. MacMicking, J. D., North, R. J.. Lacourse, R., Mudgett, J. S., Shah. S. K., and Nathan, C. F. (1997).Identification of nitric oxide synthase as a protective locus against tuberculosis. Proc. Nnt. Acnd. Sci. USA 94, 5243-5248. Magram, J., Connaughton, S. E., Warlier, R. R., Carvajal, D. M., Wu, C. Y., Ferrante, J., Stewart, C., Sarmiento, U., Faherty, D. A,, and Gately, M. K. (1996).IL-12-deficient mice are defective in IFNy production and type 1 cytokine responses. Imintiriity 4, 471-481. Mahairas, G. G., Sabo, P. J., Hickey, M. K., Singh, D. C., and Stover, C. K. (199G).Molecular analysis of genetic differences between A4ycobncteriim bouis BCG and virulent A4. boois. J. Bacteriol. 178, 1274-1282. Maher, J. K., and Kronenberg, M. (1997).The role of CD1 molecules in immune responses to infection. Cum. @in. 'Znmunol. 9, 456-461. Majeed, M., Ernst, J. D., Magnusson, K. E., Kihlstrom. E., and Stendahl, 0.(1994). Selective translocation of annexins during intracellular redistribiition of Chlrmiyrh imclionmtis in HeLa and McCoy cells. Infect. I n m u t i . 62, 126-134. Malaviya, R., Twesten, N . J., Ross, E. A,, Abraham, S. N., and Pfeifer, J. D. (1996). Mast cells process bacterial antigens through a phagocytic route for class I MHC presentation to T cells. J. I n i t n i t n o l . 156, 1490-1496. Martinez-Kinader, B.. Lipford, G. B., Wagner, II., and Heeg, K. (1995). Sensitization of M I-IC class I-restricted T cells to exogenous proteins: Evidence for an alternative I-restricted antigen presentation pathway. Iminz~iiology86, 287-295. Mashiba, H., Matsunaga, K., Tomoda, H., Furusawa, M., Jinii, S., and Tokunaga, T. (1988). In vitro augmentation of natiiral killer activity of peripheral blood cells from cancer patients by a DNA fraction from M!/cobucterizrm bouis BCG. Japan.J . Med. Sci. Biol. 41, 197-202. Matsui, K., Yoshimoto, T., Tsutsui, H., Hyodo, Y., Ilayashi, N., Hiroishi, K., Kawada, N., Okamura, H., Nakanishi, K., and Higashino, K. (1997).Pt-[~?ii~iiibncterirrm ncnes treatment diminishes CD4' NK1.1' T cells hut induces type I T cells in the liver by induction of IL-12 and IL-18 prodiiction from Kupffer cells. J. In~munol. 159, 97- 106. Matsumoto, H., Suzuki, K., Tsuyuguchi, K., Tanaka, E., Amitani, R., Maeda, 4.,Yamamoto, K., Sasada, M., and Kuze, F. (1997). Interleukin-2 gene expression in human monocytederived rnacrophages stiniulated with Mycolmcteriuni bouis BCG: Cytokine regillation and effect of NK cells. Infect. Iminun. 65, 4405-4410. Maurelli, A. T., Baudry, B., d'Hauteville, H., Hale, T. L., and Sansonetti, P. J. (1985). Cloning of plasmid DNA sequences involved in invasion of HeLa cells by Shigella.flexneri. Infect. Z t n i w i i . 49, 164-171.
accaro, R. J.. Grtldr, M., Jensen, E. R., \.tin Sainten. I I. M.,Ploegh. Ei. L., Rock, 13looin. 13. R. ( 1996). Major Iiistocoinp~itil)ilityclass I presentation of soluble antigen fiicilitatetl hy dllleo/inrto-iirttt firl~c~tr~irlo.sis infection. h e . A\’rit. Aceid. sci. c’.$i\ 93, 117fi6-11;91. McAtlain. R. A,. \Veisl)rod, T. K.. Martin, J.. Scrideri, J. I).. Brown, A. M.. Ckillo, J. I)., H l o o n i , 13. It., and Jncobs. \I. I<...Jr. ( 19951. I n rii:o gr(i\vth characteristics of lrrlcine a ~ i t l mcdiioniiie urixotrophic nilitants of bf!/ro(,crctet-iir t i t h i ~ BCG s generated by transposon niutagenesi:;. /,!fief. / t n n t i i t t . 63, 1004-1012. McKciizie. (;. J., Bancroft, A.. Greucis. R. I<.. and McKeuzie, A. N . J . (1998). A distinct role for iiiterleiikin-1:3 in TI12 iiicdiatc>.tlinimrint~responsc~s.C:trrt-.Biol. 8, 339-3-12. McNeil. A.. Dunstnn, S . J., C1a-k. S., antl Striigiiell. K. A . (1995). S d t t t c i t i d h t ! / p h i t t t i r r i i r t t i tlisplays normal iiivasion of mice mith tlrfrcti\v epideriiid growth factor rrceptors. It!fict. Z t ~ i i t i ! ( i i .63, 2770-2772, Mediiia, E.. a l l d North. R. J. (199G).The Bcg gem*(Nranipl) does not deteriiiine resistance of mice to \inileiit dlyrobcirfc~r-irrtrttirhc~rcdo.vi.c. &Atiii, A’. Y,Accirl. Sci. 797, 257-”59. Mediiia. E.. arid North. R. J. (1998). Hesistaiice ranking of sonie coninion inl)retl nioiisc strains to ,CI!ycobactr~iirri n tirlxrcitlosis mid rrlatioiisliip to major ~ i i s t o c ( ~ i i i ~ ~ coniplcx tihili~ liaplohpe antl Nrampl grinotye. Z t t t t t t i r t t d o g ! / 93, 270-271. Mediiia, E.. Rogcarson. B. J.. and North, R. J. ( 1996). Tlie Nfiinipl antiniicrobial resistaiice genc segregates independrntly of resistmccs to vinilcwt ~l~yrobcirtrriiitti tihercrdo.vis Ittittuiito/og!/ 88, 479-481. Mrdzhito\J. R., and Janrway. C . A. [ 19981. An anc.i(wt system of host defcnsr. Cirrr. Opitt. itt Z r i i j t t t i t t d i l g ! / 10, 12-15. Medzhito\-. H. l’restoii-Jl~irl1~iri-i.P.. ant! Janeway. C. .4., Jr. (1987).A huiiian Iiomologue of the Drosophila Toll proteiii signals acti\.;ttion of adaptive imnriinity. A’atiirc. 388, 394-397. Mege, J. I,.. Mauriii. M., Ciipo. C., arid Kaoult, D. (1997). Coxic,//n birnictii: The “ q u e ~ ” fever hucteriuiii. i\ niotlel of iniminie siilxwsioii hy H strictly iiitracelliilar niicroorganisni. FEMS Alicn)6. Rc7c. 19, 209-21 7. Mellniaii. I. (1992).The importonce of Iwiiig acid: The role of acidification on intracc~llrilar irieinhrane traffic. /. E.yp B i d 172, :3H-45. Melliiian, I . ( 1996). Endoc?-iosis and iiiolecular sorting. Attit. RLJI;.Ccd. L)c,v. Bici/. 12, 57.5-625. Mendifiitta. S. K., Martin, \t’. D., Hong, S., Boc.stcwiii, 4..Joyce. S., and Van K x r . L. ( 19971. C D l d l niut;irrt inice are deficicwt i l l iiatiiral T cells that proinptl!. produce IL4. Itntttuttit!/ 6, 469-477. Mengaud. J., Ohayon. H.. Goimon. P..blcsge, R.-M.. and Cossart, P. ( 1996). E-cadherin is into the receptor l o r internalin. a siirface protein reqiiircd for entiy of L. tttotioe~togett~~~c cyithelial cells. C d 84, 92:3-932. Menozzi. F. D , Rouse, J. H., Alavi, M . , Liiride-Slinrp, hl., Muller. 1.. Bischoff. K., Brennan, M. J., a i d Lmlit, C:. (1996). 1tlentific;ttion of a Iicl’~iriir-bintlirighcmagghtinin present in mycohacteria. J. Exp. Aid. 184, 993-1001. Meya;trd, L., IIovenkamp, E., Otto, S. A,, and hlicdeiiia, F. (1996). 11,-12-intluced IL-10 production hy huinan T cc~llsits a negative feedback for IL-12-indiicetI i m m ~ ~ responses. nr /, Z i i ~ t i t u i t o / . 156, 2776-2782. Mielkr, M. E. [ 1991). T cell subsets in granri1oiii;ttoiis iiitlaiiimation and inimuiiih to Li.yterici triottoc,!/fogc,r,r.r. and Bnicrlln c h o r f i ~ . B(hriitg ~. Zitstit. Miftcdrrngen. 99-1 11. Mirlkr. M. E.. Niedobitek, G., Stein, El., antl Hahn. H. (1989). Acquired rcsistance to Listc,rici tttot:oc!/fo~:Pties is mediated I)!, I+-i-?+ T crllls independently of the iiiwiix of iiioiiocytes iiito graiiiilomatoiis lesions. /. E . Y ~.\led. 170, 589-594.
K. L., a i i d
360
ULRlCll E. SCH4lRLE el nl
Mielke, M. E., Peters, C., and Hahn, 13. (1997). Cytokines in the induction and expression of T-cell-mediated granuloma forination and protection in tlie inurine model of listeriosis. lniniunol. Rev. 158, 79-93. Mielke, M. E., Rosen, H., Brocke, S., Peters, C.. and Hahn, H. (1992). Protective iininunity and granuloma formation are mediated by two distinct tumor necrosis factor a and y interferon-dependent T cell-phagocyte interactions in murine listeriosis: Dissociation on the basis of phagocyte adhesion mechanisms. lnfect. I n m u n . 60, 1875-1882. Miller, S. I., Kukral, A. M., and Mekalanos, J. J. (1989).A two-component regulatory system (PhoP PIioQ) controls Salnwiiella fyphitnt4ritini vinilence. Pmc. Naf. Acrid. Sci. USA 86, 5054-5058. Miyagi, K., Kawakami, K., and Saito, A. (1997). Role of reactive nitrogen and oxygen intermediates in gamma interferon-stimulated inurine inacrophage bactericidal activity against Burkliolderia pseurkmmllei. Infect. IttinLun. 65, 4108-41 13. Miziiochi, T., Yee, S. T., Kasai. M., Kalauchi, T., Muno, D., and Kominami, E. (1994). Both cathepsin B and cathepsin D are necessary for processing of ovalbumin as well as for degradiition of class I1 MHC invariant chain. Z i n t t i t t ~ i o ~Lett. . 43, 189-193. Mocllin, R. L., and Nntman, T. B. (1993). Type 2 cytokines and negative irninune regulation in human infections. Cum. Opin. lttitnunol. 5, 511-517. Modliri, R. L., Pirmez, C., Hofman, F. M., Torigian, V., Uyemura, K., Rea, T. H., Bloom, B. R., and Brenner, M. B. (1989).Lyniphocytes bearing antigen-specificyST-cell receptors accumulate in human infectious disease lesions. Nutrrre 339, 544-548. Moll, H., Flohe, S., and Blank, C. (1995). Dendritic cells seclude Leishmania parasites that persist in cured mice: A role in tlie maintenance of T-cell memory? Adu. Exp. bled. B i d . 378, 507-509. Moinbaerts, P., Amoldi, J., Russ, F., Tonegawa, S., and Kaufmann, S. H. E. (1993). Different roles of a/3 and yS T cells in immunity against an intracellular bacterial pathogen. Nature 365, 53-56. Moody, D. B., Reinhold, B. B., Guy, M. R., Beckman, E. M., Frederique, D. E., Furlong, S. T., Ye, S., Reinhold, V. N., Sieling, P. A,, Modlin, R. L., Besra, G. S., and Porcelli, S. A. (1997).Structural requirements for glycolipidantigen recognition by CDlb-restricted T cells. Science 278, 283-286. Moore, K. W., O’Garra, A,, de Waal Malefyt, R., Vieira, P., and Mosmann, T. R. (1993). Interleukin-10. Annu. Rev. I n i m ~ m o l .11, 165-190. Mordue, D. G. and Sibley, L. D. (1997).Intracellular fate ofvacuoles containing Toxoplasnui gondii is determined at the time of formation and depends on the mechanism of entry. /. ln~mtinol.159, 4452-4459. Morein, B., Lovgren-Bengtsson,K., and Cox, J. (1996).Modem adjuvants, fnnctional aspects. I n “Concepts in Vaccine Development” (S. H. E. Kaufinann, ed.), pp. 243-264. Walter de Gruyter. BerlirdNew York. Moreira, A. L., Wang, J.. Tsenova-Berkova,L., IIellinann, W., Freedman, V. H., and Kaplan, G . (1997). Sequestration of Mycobncterium tuherculosis in tight vacuoles in vivo in lung macrophages of mice infected by the respiratory route. Infect. Immm. 65, 305-308. Morita, C. T., Beckman, E. M., Bukowski, J. F., Tanaka, Y., Band, H.,Bloom, B. R., Golan, D. E., and Brenner, M. B. (1995). Direct presentation of nonpeptideprenyl pyrophosphate antigens to human yS T cells. Immunity 3, 495-508. Mosmann, T. R., Chewinski, H., Bond, M. W., Giedlin, M. A,, and Coffinan. R. L. (1986). Two types of inurine helper T cell clone. I. Definition according to profiles of 1,pphokine activities and secreted proteins. /. Immonol. 136, 2348-2357. Mosmann, T. R., Li, L., Hengartner. H., Kagi, D., Fu, W., and Sad, S. (1997). Differentiation and functions of T cell subsets. Cihn Found. Symp. 204, 148-154; discussion 154-148.
INTRACEl.l.ULAR BA(:TERIA A N D TllE I h l M U K E SYSTEM
361
Mosser. D. M. ( 1994). Receptors on phagocytic cells involved in iiiicrobial recognition.
Moulder. J. U'.( 1985). Coiiiparative biok+p of intracellular parasitism. Microbial. Rrc. 49, 298-337. Moura, A. C., Modolell, M.. and Mnriano. M. (1997).Do\~~i-rc.gulatory effect ofMycobnctcJrirrrii lcyrtlr, cell wall lipids on pli;igocytosis, osidativc respiratory 1)urst and tuinour cell killing I)? nioiise hone marrow dc,rived inacrophagtl. Scnnd. 1. Z t i i v i u t i d . 46, 5OO-SOFj. Msliana, R. N..Hastings, R . C., and Kralitdiuhl, J. I,. (1988).Infection with live mycobacteria ~ ~ ~ 40-54. ~ ~ [ ~ ~ y inhiliits in vitro detection of Ia antigen on niacropliages. / r t i r i i i ~ i i ~ ) 177, Miiller, I.. Cubbok1, S. P.. \Vak~nianii. 11.. and Kaufinann. S. 11. E. (1987). Iiiipaired resistance to Xfycolmcteriirrii tiihcr-crrk)si.v infwtioii after selective i n civo depletion of L3T-1' and L>t-2' T cells. Itfict. I i i i i ) i t i i i . 55, 20:37-2041. Muntler, M., M d o . M . , Eicliniann, K.. iuid Motlolell. M . (1998). Murine inacrophages secrete interferon gainina upon coml)ined stiindation with interleukin (IL)-12 and IL18: A novel p at h n y . of aiitocrine macrophage activation. 1.E q i . Alctl. 187, 2103-2108. Murray. P. J., Aldo\ini, A.. and Yoring, R. A. ( 1996). Manipulation aiitl potrntiation of aiitiinycohac.teria1 iiiimiiiiity usiiig recoin1,inant b a d e Calniette-Guerin strains that secrcte cytokines. Proc. Nut. Acrid. Sri. I'SA 93, 934-9:39. Murray. P. J. h'uig, L., Oniifryk. C.. Teppvr, R . I., ant1 Young, R. A. (1997). T cellderivrd IL- 10 antagonizes macrophage function in niycobacterial infection. ]. I t n i t i t t r i o / . 158, 3 15-32 1. N d en u . J. J.. 4rl~ickle,L. I)., d Sknmene, E. (1995).Genetic dissection of inflammatory res~'oI1"'s. 1. z t l f l n t i l l i l , 45, "-33. Nataraj, C.. Brown. M . L., Poston. R. M., S h w w , S. M., Rich, R. R., Lindahl, K. F., aiid Kiirlander, R. J . (1996). H2-MSwt-restricted, Zistorio riionocytoecnrs-specific CD 8 T cells recognize ii novel, Iiytlrophobic. proteasc,-resistant, pciiodate-sensitive antigen. Int. Z I 1 1 I t L I i 1 1 0 1 . 8, 367-378. Nathan. C. F., and Hihiis, J. B. Jr. (1991). Role of nitric osidcl s>mthesis in macropliage antiniicrobial activity. Curr. Opiti. I r t i i j i u i i o l . 3, 65-70. Nathan, C. (1997). Intlriciblr nitric osidr syithasc: \$'hat differrnce does it make? . 1.C h i . /r1r;t,st. 100, 2417-2423. Ndson, C . A , . \'idavsly, I., \'iner, N . J.. Gross, M. L.. and Uiianue. E. R. (1997). Aininocomplex class terminal triiiiming of peptides for presentation on major 1iistocoiiip~~til)ility 11 Inolecllles. Pt-oc. Not. Acod. sci. C'SA 94, 628-633. Nelmiiuccwo. R. R.. €Iensclieii-Ediiian. A. H., Br~rgess,W. I I . , and Tciiner, A. J. (1997). cDiia clonins and priin;iry stnictiire analysis of ClqR(P).thr human Clq/Mbl/Spu receptor that inetliates enhanced pliagocytosis iii \itro. Ittiinunity 6, 119-129, Newport, M . J., Hiidcy, C:. M.. Hiiston. S.,Hawylo\vicz, C;. M., Oostra, B. A., Williamson. R., and I,ecin, M. (1996). A iiiutation iii the interft.roii-y-receptor gene and srisceptibilit\ to mycolxicterial infection. N . Engl. ] Mcd. 335, 1941 -1949. Newton. S. M.. Jacob, C. 0..and Stockctr, B. A . (1989). Immunc. response to cholera tosin epitope iiisertetl i n Salnionolla flagellin. Seicricc 244, 70-72. Nicholson. S., Boiiccini-Almeitla, M., Lapa e Silvo, J. R., Nathan, C., Xie, Q. W., Munifortl. R., Weidner. J. R., Calavcy, J.. Geng. J., Borchat. N.. c,t a / . (1996).Indncible nitric oxidr, sylithase in pulnionary alvcdar niacrophages from patients with tiiberciilosis. J . E.yp. &fed. 183, 2293-2302. Nigoii. J., Gilleron. M., Calirizac, €3.. Horlncry, J. I)., Herold. M.. Thnmlier, M., a~id Puzo, G. (1997). The phosphatidyl-inyo-inositol anchor of' the lipoarabinomannaiis froin M!/cobncfcn'roiihoui.~bacillus Calniette G w r i n : Hetcrogeneity. structure, and role in the L . 23094-23103. regii1;ition 6 cytokine secretion. /. Biol. C ~ ~ W272,
362
LJLRICH E. SCHAIBLE et nl.
Norbury, C. C., Chambers, B. J., Prescott. A. R.; Ljunggren, H. G., and Watts, C. (1997). Constitutive macropinocytosis allows TAP-dependent inajor histocompatibility coniplex class I presentation of exogenous soluble antigen by bone marrow-derived dendritic cells. Eur. J. Ziuniimol. 27, 280-288. Novick, P., and Zerial, M. (1997).The diversity of Rab proteins in vesicle transport. Cum. $in. Cell Biol. 9, 496-504. Nozaki, Y., Hasegawa, Y., Ichiyama, S., Nakashima, I., and Shimokata, K. (1997).Mechanism of nitric oxide-dependent killing of Mycobrtcterinm bouis BCG in human alveolar rnacrophages. Infect. Zmniiin. 65, 3644-3647. O'Brien, L., Roberts, B., and Andrew, P. W. (1996). In vitro interaction of Mycobacteriurn tuberculosis and macrophages: Activation of anti-mycobacterial activity of macrophages and mechanisms of anti-mycobacterial iinmunity. Cum. Top. Miorbiol. Znmunol, 215,
97- 101. O'Brien, R. L., Ilapp, M. P., Dallas, A,, Palmer, E., Kubo, R., and Born, W. K. (1989). Stimulation of a major subset of lymphocytes expressing T cell receptor y6 by an antigen derived from M!lcobacterium tuberdosis. Cell 57, 667-674. O'Connor, C. D., Farris, M., Fowler, R., and Qi. S. Y. (1997).The proteome of Salmonella enterica serovar T!yphiniurium: Current progress on its determination and some applications. Elec,-trophore,sis 18, 1483- 1490. Oh, Y. K., Alpuche-Aranda, C., Berthiaunie, E., Jinks, T., Miller, S. I., and Swanson, J. A. ( 1996). Rapid and complete fusion of macrophage lysosomes with phagosomes containing Salnwnellu typhiniurium. Infect. Ininwn. 64, 3877-3883. Oh, Y. K., and Straubinger, R. M. (1996).Intrarellular fate of Mycobacterium aoium Use of dual-label spectrofluorometry to investicate - the influence of bacterial viability and opsonization on phagosomal pH and phagosome-lysosome interaction. h f e c t . Znimun.
64, 319-325. Ohga. S., Ikeuchi, K., Kadoya, R., Okada, K., Miyazaki, C., Suita, S., and Ueda, K. (1997). Intrapulmonary Mycobacterium auinm infection as the first manifestation of chronic granuloinatous disease. J. Znfect. 34, 147-150. Ojcius, D. M., Dedba, Y. B., Kanellopoulos, J. M., Hawkins, R. A,, Kelly, K. A,, Rank, R. G., and Dantlyvarsat, A. (1998). Internalization of Chlamydia by dendritic cells and stimulation of Chlamydia-specificT cells. J . Zmnwtiol. 160, 1297-1303. Ojcius, D. M., Hellio, R., and Dautry-Varsat,A. (1997).Distribution of endosonial, lysosonial, and major histocompatibility complex markers in a monocytic cell line infected with C h h i y d i a psittaci. Infect. Zmnizrn. 65, 2437-2442. Ojcius, D. M., Thibon, M., Mounier, C., and Dautry-Varsat, A. (199s). pH and calcium dependence of heinolysis due to Rickettsia protoazekii: Comparison with phospholipase activity. Infect. Zmmurc. 63, 3069-3072. Oheira, S. C., and Splitter, G. A. (1995). CD8' type 1 CD44'"CD45 RB'" T lymphocytes control intracellular Bnicekz abortus infection as demonstrated in major histocompatibility complex class I- and class 11-deficient mice. Eur. /. Znununol. 25, 2551-2.557. Orme, I. M. (1988).Induction of nonspecific acqtiired resistance and delayed-type hypersensitivity, but not specific acquired resistance in mice inoculated with killed tnycobacterial vaccines. Infect. I n m i i n . 56, 3310-3312. Ortmann, B., Copeman, J., Lehner, P.. Sadasivan, B., Herberg, J. A,, Grandea, A. G., Riddell, S. R., Tanipe, R., Spies, T., Trowsdnle, J., and Cresswell, P. (1997).A critical role for tapasin in the assembly and function of multimeric MHC Class I-TAP complexes. Science 277, 1306-1309. Pal, P. C . ,and Honvitz, M. A. (1992).Immunization with extracellular proteins of Mycobacteriurri tuberculosis induces cell-mediated immune responses and substantial protective iinmunity in a guinea pig model of pulmonary tuberculosis. Infect. Zrnniun. 60,4781-4792.
IKTRt\CEI,LUL.AR BACI'ERIA 4 N l ) THE: I h l M U N F : SYSTEM
363
Pariier, E. C., Wang, C.-R.,Flaherty, L., Fischer Lintlahl, K., arid Bevan, M .J. (1992). I1-2M3 presents ii Listcv-in rru,tioc!ilogenc.I. peptide to c)$otoxic T lympl~ocytes.Cell 70, 215-223. Pamer, E. G.. Sijts, A. J., Villanucva, M. S., Biiscli, D. H., antl Vijh, S. (1997). M I I C class I antigen processing of Listcvia r ~ i [ , r t [ ) ~ , ~ / f ~proteins: ) ~ c ~ ~ t eIinplications .y for doininant and suljdomiiiaiit CTL responses. Z t t t t n t m d . Heti. 158, 129-136. Pancholi, P., Mirza, A., Bhardwaj, N . , and Steininan, R. M. (l993a). Sequestration froin iinmuiie CD4' T cells of mycobactr.ria growing i n Iiuinan macrophages. Science 260, 984-986. Panclroli, P., klirza, A., Schauf. I'., Steininon, R. M., a i i d Hhartlwaj, N. (1993b).Presentation s : of transfer froin infected microof inycobacterial antigens by h u n i m drndritic ~ ~ 4 Lack phages. Ir4ec-t. 1rtmi11ri. 61, 5326-5332. Parham, P. (1!494). The rise and fall of great s I genes. Sc,rtiiti. Z t ~ i t t i i ~ t ~6, o ~373-382. . Park, S. H., Roark, J. H., and Bendelac, A. (1998). Tissiw-specific recognition of iiiouse CD1 moleciiles. /, Zntrriuriol. 160, 3128-3134. Parker, C . M.,Groh, V..Band, ti.,Porcelli, S. A., Morita, C.. Fahbi. M., Glass, D., Strominger, J. L., and Brenner. kl. B. (1990). Evidcmce for extrathymic changes i n the T cell receptor ylS repertoive. /. Exp. Med. 171, 1597-1612. Pasparkis, M., Alexopoulou, L.. Douni, E.. and Kollias. G. (1996). Tumour necrosis factors in iinmune regulation: Everything that is interesting is . . . new! Cytokine Growth Fmtor Rec. 7, 223--229. Pasula, R., Downing, J. F., Wright. J. R., Kacliel, D. L., Davis, T.E.. Jr., and Martin, W.J., I1 (1997). Surfactant protein A (SP-A) mediates attachment of Myc~bactericirri tuhcrcdosi.y to niurine alveolar nwrophages. Atti. /. H e ~ pCdI. A4d B i d . 17, 209-21 7. Paul, \Y. E. (1997).Interleukin 4: Signalling incd~anisinsaiid control ofT cell differentiation. Cibn Fouiid. Syrtip. 204, 208-216: discrission 216-219. Paylie, N . R., ;ind Honvitz, M. A. (1987). Pliago:ocy$osis of L. pnfwrtu)phih is mediated hy human moiioc)te coinplement receptors. /. Exp. Met/. 166, 1077-1389. Pearson, A. M. ( 1996). Scavenger reports in iiinnte immunity. Ciirr. Opirt. Zttititwio/. 8,
20-28. Pechhold, K.. LVesch, D., Schondelmaier, S., and Kabelitz. D. (1994). Primary activation of Vy 9-expressing y6 T cells by M!phcicten'rttti riihr,r-crik)sis:Requirement for Th1 -type CD4 T cell lielp arid inhibition by IL-10. J , IfJittiurio~.152, 4984-4992. Peeteririans. I V . E., Raats, C. J., van Furtli. K., and Langernians, J. A. (1995). Mycobacterial S4-kilodalton heat shock protein induces tninor ncLcrosis fwtor (Y and interleukin 6, reactive nitrogen intermediates, and toxoplasniostatic activity in inurine peritoneal macropliages. Irlfect. Z t r t r r i i i i i . 63, 3454-3458. Pena, S.V.,antl Kreiisb, A . M. (1997).Granulysin, a new huinan cljtolytic aranrile-associated protein witli possible iii~olvementin cell mediated cytotoxicity. Seniin. I i r i r n t i r t d . 9, 1li-125. Peterson, P. K.. Gekkrr, G.. HI], S., Sheiig, U'. S., Anderson. \V. R., Ulevitch, R.J., Tobias, and Cliao, C . C. (1W5).CD14 receptor-mediated P. S., Gustafson, K. V.. Mohtor, T. W., uptake of noiiopsonized Al!i'."bncte,rirrirt ftilwrctiIo,vi,Eby human microglia. Zrlfect. Ztrirtmrt. 63, 1598-1692. Pfeffer, K., Matsiiyiuiia, T., Kundig, T. M., IVakeliani. A . , Kishihara, K., Shahinian, A., Wieginann, K.. Ohashi, P. S., Krolike, M . , arid Mak, T. \V.(1990). Mice deficient for thc 55 kD tumor necrosis factor receptor are resistant to endotoxic shock, yet succumb to L. infection. Cell 73, 457-467. Pfeffer, K., Sckioel, B., Gulle, H., Kaufiiiann, S. H. E., and Wagner, H . (1990). Primary responses of human T cells to inycobactcria: A frequent set of y/S T cells are stimulated by protease-resistant ligands. Eur. /. Z t n r t i u r i d . 20, 1175-1179.
364
ULRICH E SClIAlBLE f't nl
Pfeifer, J. D., Wick, M. J.. Roberts, R. L., Findlay, K., Normark, S. J., and Harding, C. V. (1993). Phagocytic processing of bacterial antigens for class I MHC presentation to T cells. Nature 361, 359-362. Pierre, P., Denzin, L. K., Hamniond, C., Drake, J. R., Amigorena, S., Cresswell, P., and Mellnian, I. (1996). HLA-DM is localized to conventional and unconventional MHC class 11-containing eiidocytic compartments. lrrirnunity 4, 229-239. Pieters, J. (1997). MHC class I1 compartments: specialized organelles of' the endocytic pathway in antigen presenting cells. Biol. Chern. 378, 751-758. Pisetsky, D. S. (1996). Immune activation by bacterial DNA: A new genetic code. Irnmunity 5, 303-310. Pitt, A,, Mayorga, L. S., Stalil, P. D., and Schwaitz, A. L. (1992). Alterations in the protein composition of maturing phagosomes. 1.Clin. Invest. 90, 1978-1983. Plum, G.,andClark-Curtiss, J. E. (1994).Induction ofMycobncteriirinaoiurri gene expression following phagocytosis by human macrophages. Infect. Irnmnun. 62, 476-483. Polotsky, V. Y., Belisle, J. T., Miknsova, K., Ezekowitz, R. A., and Joiner, K. A. (1997). Interaction of human mannose-binding proteiii with Myrobocteriuni auiurn. J . Infect. Dis. 175, 1159-1168. Pope, M., and Kotlarski, I. (1994). Detection of SalrtwrielZf~-specificL3T4' and Lyt-2+ T cells which can proliferate in vitro and mediate delayed-type hypersensiti\ity reactivity. Irnniunolog~8 1, 183- 191. Pope, M., Kotlarski, I., and Doherty, K. (1994). Induction of Lyt-2+ cytotoxic T lymphocytes following primary and secondary Salmonello infection. Ittimunology 81, 177-182. Porcelli, S., Brenner, M. B., Greenstein, J. L., Balk, S. P., Terhorst, C., and Bleicher, P. A. (1989). Recognition of clnster of differentiation 1 antigens by human CD4-CD8cytolytic T lymphocytes. Nature 341, 447-450. Porcelli, S., Morita, C. T.. and Brenner, M. B. (1992).C D l b restricts the response of hnnian CD4-8- T lymphocytes to a microbial antigen. Nature 360, 593-597. Porcelli, S. A. (1995). The CD1 family: A third lineage of aritigen-presenting molecules. Adu. Initnuriol. 59, 1-98, Porcelli, S. A,, Morita, C. T., and Modlin, R. L. (1996). T-cell recognition of non-peptide antigens. curr. Opin. 1mriund. 8, 510-516. Porcelli, S. A,, Segelke, B. W., Sngita, M., Wilson, I. A,, and Brenner, M. B. (1998). The ED1 family of lipid antigen presenting molecules. Irnrrwnol. T o d q 19, 362-368. Porter, E. M., van Dam, E., Valore, E. V., and Ganz, T. (1997).Broad-spectrum antimicrobial activity of human intestinal deferisin 5. Znfict. Zmrrwn. 65, 2396-2401. Powvis, S. J. (1997). Major histocompatibility complex class I molecules interact with both subunits of the transporter associated with antigen processing, TAP1 and TAP2. Eur. J. Inirnuriol. 27, 2744-2747. Prete, G. F. D., Carli, M. D., Alrnerigogna, F., Daniel, C. K., Delioe, M. M., Zancuoghi, G., Pizzolo, G., and Romagnani, S. (1995). Preferential expression of CD30 by human CD4+ T cells producing Th2-type cytokines. FASEB J. 9, 81-86. Prigozy, T. I., Sieling, P. A,, Clemens, D., Stewart, P. L., Behar, S. M., Porcelli, S. A,, Brenner, M. B., Modlin, R. L., and Kronenberg, M . (1997). The mannose receptor delivers lipoglycan antigens to endosoines for presentation to T cells by C D l b molecules. ImftllLYLity 6, 187-197. Prina, E., Antoine, J. C., Wiederanders, B., and Kirschke, H. (1990). Localization arid activity of various lysosonial proteases in Leishnlania a,rla;nnensis-infected macrophages. Infect. lrnniun. 58, 1730-1737. Prina, E., Lang, T., Glaichenhaus, N., and Antoine, J. C. (1996).Presentation ofthe protective parasite antigen LACK by Leishmania-infected macrophages. J. Initrtunol. 156, 43184323.
IN‘THA(XLL1J L A H BA( :TE RIA Ah’ 1) TI 1 Pi: I hl M U N E SYSI’E hf
36.5
Probst, I’., Skeib, Y. A., Steeves. M., Gcwassi. A,, Grabstein, K. H., and Reed, S. G. (1997). A Leishiiiania protein that modiilates iiitcrleukin ( ILL 12, IL- 10 a d tuiiior necrosis factoralpha production and expression of BT- 1 in h u m a i i moiioc)rte-tleriveclantigen-prescnting cells. Etcr J . Z t t m ~ r t t d27, 2634-2642. Pryjma, J., Baran, J.. Ernst, M.. Woloszyn. M., and Flad, H. D. (1994). Altered antigenpresenting cxpacity o f h u m m nionocytcbs after pliagocpsis of bacteria. It+ct. I t t i t t i t i t i . 62, 1961-1967. Pugin, J., Heumann, 1. D., Toniasz. A.. Kravchenko. V. \’,, Akainatsii, Y., Nishijiina, M., Clauser, M. P., Tobias, P. S., antl Ulecitch. K. J. (1994).CD14 is a pattern recognition receptor. I i r i t i i t i n ity 1, 509-5 16. Qi, S. Y., Moir. A., and O’Connor. C. (19%). Proteonir of Sultttonelln typliinitirinm SL1344: Itlentification of novel ahinidant cell envelope proteins and assignment to a twodiinensional reference map. J. Bnctcriol. 178, 5032-50338. Rabinowitz, S., Horstniann, H., Gordon. S., and Criffitlis, G. (1992).Iiniiiunocytoclieniic~il characterization of the endocytic and phagolysosoinal conipartinents in peritoneal ~nacrophages. J . Crll Biol. 116, 95-112. Raetz, C.. Ulevitch, R.. Wright, S., Siblry, C.. Ding, A,, arid Nathan, C. (1991). Cramnegative endotoxin: ,411 estraortlinq lipid with profound effects on eukaryotic signal transduction. FASEB J. 5, 2652-2660. Iiafic.-Kolpin. M., Essrnberg. K. C.. and Wyckoft: 1. H., I11 ( 1996). Identification and comparison of iiiacl.opliage-induce~1proteins antl proteins indnced under various stress conditions iii Bnrctlla ob&ttis. 1ifet.f. I t t t t t t i i i i . 64, 5274-5383. R;ijalingam. R.. Singal, D. P., arid Melira. N.K. (1997).Transporter associated with antigeni n g (TAP)genes and snsceptibilit\ to tuberculoid leprosy antl pullnonary tu1)erculosis. TissirP Arctigetis 49, 168-172. Ramhnkkana. A,, Salzrr, J. L., Yiirchcnco. P. D., antl Triomanen. E. I. (1997). Neural targeting of M~jcobcrtteritiittIelirav nietliated by the G dornain of the laininin-a2 chain. Cell 88, 811-821. Rathnian, M., Barker, L. P., imd Fdko\v, S. (1997). The unique trafficking pattern of SaltttottcUa t ~ ~ ~ ~ l t i ~ ~ i r c r i i ~ t t i - c ~phagosomes i i t ~ i i i i i i i g in niiirine niacrophages is independent of the inec1i;iiiism of bacterial entry. Infict. I t i t n i t t i t . 65, 1475-1485. Rathinan. M., Sjaastad, M. D.. ;md Falko\v, S. (1996).Acidification of phagosomes containing Smltttoi~c/lnt ! / p h i t t c r r r i u t ~ ti n niurine nincrophagt~s.Infict. I t w i t t i n . 64, 2765-2773. Kaulston, J. E. (1997). Response of Ch/nttiydin frcrcliotttntis serovar E to iron restriction i n vitro and evidence for iron-regiilated clilamydial proteins. ltfect. Z r t m ~ i i n .65, 4539-4547. Heilly, T. J., Baron. G. S., Nano, F. E., and Kiihlensclimidt, M. S. (1996).Characterization and sequencing of I I respirator). 1)urst-inhi1,iting acid phosphatase from Frutici.sella fuEareiwi.s. /. B i d Clienc. 271, 10973-109833. Reiinanri, J.. and Kaufiiiann. S. H. E. (1997). Altwnativc, antigen processing pathways for MMC-restricted epitopc presentation i n anti-infective immlmity. Ctrm. Opin h i t t 1 1 L ? l d . 9,462-469. Reiner, S. L., d Sedrr. R. A. (199s). T helper cell diffrrentiatioii i n iinininie response. Cirrr. Opiti. ~ m i c i r i i d7, :360-366. Reis e Sousa, C.. and Gerinain, R. N. (1995). Major liistocompatibili~complex class I presentation of peptitles derived froni soluhle exogenous antigen by a subset of cells engaged in phagocytosis. /. E x p hled. 182, 841-851. R e p t . J. M., Ikrthet. F. X., and Gicquel, B. (1995).The urease locus of M!/cobacterium trrberctikxsis md its utilization for the demonstration of allelic exchange in Mycohucteriz~nt boois bacillus Calmette-Guerin. Proc. N a t . Acnrl. Sci. USA 92, 8768-8772.
Reyrut, J. M., Lopez-Rainirez, G.,Ofrrdo, C., Gicquel, B., and Winter, N. (1996). Urease activity does not contribute dramatically to persistence of Mycolmcteririm boois bacillus Calniette-(~iierin.Irfmtrct.I t n t t i r ~ r i .64, 3934-3936. Rlaoatles. E. R., aud Ormr. 1. M. (1997). Siisceptibility of a panel of viruhit strains of Mycobac-fc~ri:r?n triberrrilosis to reactive nitrogen interinediates. Infect. I r n t t i r m . 65, 11891195. Rhoades, E. R., Coopcr, A. M., and Orine. I. M. (1995). Chemohne response in mice infected with Mycobocteriritti tribrrcrilosis. Infect. Z t t m r m 63, 3871-3877. Kicdel. 13. D., antl Kaufiiiann. S. H. E. (1997).Chenaokine secretion by huniaii polyinorphonuclear granulocytes after stiinulation with Bf!/cobnctcrinn~trrhercrilu niiinnan. Irtfiect. Z m t t i r i t i . 65, 4620-4623. Kiese. R. J., Wolf, P. R., Bromine, D., Natkin, L.. K., Villadangos, J. A,, Ploegh. I-I. L.. and Chapman, H. A. (1996). Essential role for cathepsin S ill M H C class 11-associatedinvariant chain processing and peptide loading. Ztnttirmity 4, 357-366. Roach, T. I., Barton, C. H.,Chatterjee, D., and Blackwell, J. M. (1993). Macrophage activation: li~~o~~rabinoiriaaira~~~a froin avinilent and virulent strains of Myt:obncteritnntuberciilosis differentially induces the rarly genes c-fos, KC, JE. and turnor necrosis factor-a. /. Irmnunol. 150, 1886-1896. h a r k , J. II., Park, S. H., Jayawardena, J., Kavita, U., Shannon, M., and Bendelac, A. (1998). CD1.1 expression by iiiotise iiiitigeii-preseiatingcells and niarginal zone B cells./. I t r i t t m r m l , 160, 3121-3127. Roberts, A. D., Sonnenberg, M. G., Ordway, D. J., F u r y , S . K., Brennan, P. J., Belisle, J. T., and Ornie, I. M. (1995). Characteristics of protective iininirnity engendered by vaccination of mice with purified culture filtrate protein antigens of Mycobmct~v-itcm frrbr~rculosis.I ~ r ~ t n r m ) / o85, g ~ 502-508. Rol)inson, D., Shihuya. K., Mui, A., Zonin, F., Murphy, E., Sam, T., Hartley, S. B., Mrnon, S., Kastelein, K., Bazan, F., and O’C;arra, A. (1997). IGIF does not drive TI11 development but sjnergizes with IL-12 for interferon-y production antl activates IRAK and NFKB. Z 7 ~ l t t ~ f i t i i t !7, / 571-581. Rock, E. P., Sibl)altl, P. H., Davis, M. M., and Cliien, Y.-H. (1994).CDH3 length in antigeiispecific immune receptors. /. Exp. Met/. 179, 323-328. Rockey. D. D., and Rosquist, J. L. (1994).Protein antigens of Clilutnydio psittnci present in infected cells Iiut not detected in the infectious elementary body. Ztfect. [triiririn. 62, 106-112. Rogers, H. W., Caller).. M. P.. Deck, B., and Unantie, E. R . (1996). Listericl t t z ~ r ~ [ ~ c y ~ [ ) ~ ~ , ~ i e . ~ iridiices apoptosis of infrcted liepatocytes. /. l i t i i n i ~ ~ i o 156, l. 679-684. R o j ~ M., , Barrera, L. F., Puzo, G., and Garcia, L. F. (1997). Differential induction of apoptosis liy virulent Mycobactr~riritti ttihercrilosis in resistant and susceptible inurine macrophagw Hole of nitric oxide and mycobacterial products. /. Itnrnzinul. 159, 13521361. Romagnani. S., Parronchi, P., D’Elios, M. M.. Rornagnani, P., Annunziato, F., Piccinni, M. P., Manetti, R., Sanipognaro, S.,Mavilia, C., De Cadi, M., Maggi, E., and Del Prete, C.. F. (1997). An update on human TI11 arid TI12 cells. Znt. Arch. Allergy Z m t ~ i u t i o l . 113, 153-156. Romain, F., Augier. J., Pesclier, P., antl Marchal, G. (1993). Isolation of a proline-rich ~nycol~acterial protein eliciting delayed-tpe hpersensitivity reactions only in guinea pigs iininunized with living inycobacteria. Proc. Not. Acnd. Sci. USA 90, 53224326. Roman, M., Martin-Orozco, E., Goodinan, J. S., Nguyen, M. D., Sato, Y.. Konaghy, A,, Kornbluth, H. S., Richman, D. D., Carson, D. A., and Raz, E. (1997).I ~ n ~ n ~ ~ n o s t i m ~ ~ l a t o r y DNA sequences function as T helper-1-promoting adjuvants. Nature Med. 3, 849-854.
ISTKACELLULAR BACTERIA AND THE I M M U N E SYSTEM
367
Rook, G. A. W., Taverne, J., Leveton, C., and Steele, J. (1987). The role of y-interferon, \.itamin D metabolites and tumor necrosis factor in the ~athogenesisof tuberculosis. I~n~nunology 62, 229-234. Rosat, J.-P., Grant, E. P., Beckman, E . M., Dascher, C. C., Sieling, P. A,, Frederique, D., Modlin, R., Porcilli, S. A,, Furlong, S. T., and Brenner, M. B. (1998). TCR a/3+CD8+ T cells recognize microbial phospholipid antigens in the context of C D 1 presenting molecules. J. Exp. Med. (in press). Rosenkrands, J., Rasmussen, P. R., Carnio, M., Jacobsen, S., Theisen, M., and Andersen, P. (1998). Identification and characterization of a 29-kilodalton protein from Mycohactet-ium tuher~~ilosis culture filtrate recognized by mouse memory effector cells. Infect. Iiizmun. 66, 2728-2735. Rosenshine, I., Ruschkowski, S., Foubister, V., and Finlay, B. B. (1994). Snbnonella typhiiiwTit~ininvasion of epithelial cells: Role of induced host cell tyrosine protein phosphorylatiori. Infect. Iinnlun. 62, 4969-4974. Rothe, J., Lesslauer, W., Lotscher, H., Lang, Y., Koebel, P., Kontgen, F., Althage, A,, Zinkernagel, R., Steinmetz, M., and Bluethmann, 11. (1993). Mice lacking the tumour necrosis Factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria ~nonocytogenes.Nature 364, 798-802. Rothman, J. E. (1994). Mechanisms of intracellular protein transport. Nature 372, 55-63. Russell, D. G. (1998). What does "inhibition of phagosome/lysosome fusion" really mean? Trends Microbiol. 6, 212-214. Russell, D . G., Dant, J., and Sturgill-Koszycki, S. (1996). Mycohacteriz~maoiuin- and Mycobacterium tuberculo.sis-containingvacuoles are dynamic, fusion-competent vesicles that 156, are accessible to glycosphingolipids from the host cell plasmalemma. J. Itr~n~unol. 4764-4773. Russell, D. G., Sturgill-Koszvcki, S., VanIIeyningen, T., Collins, H. L., and Schaible, U. E. (1997).Why intracellular parasitism need not b e a degrading experience for Mycobacterilim. Plzil. Trans. Roy. Soc. Lonrlon Ser. B Biol. Sci. 352, 1303-1310. Russell, D . G., Xu, S., and Chakraborty, P. (1992). Intracellular trafficking and the parasitopl~orousvacuole of Leiskinania imxicana-infected macrophages. J. Cell Sci. 103, 11931210. Sallusto, F., Cella, M., Danieli, C., and Lanzavecchia, A. (1995). Dendritic cells use macropinocytosis and the mannose receptor to concentrate macromolecules in the major histocompatibility complex class I1 compartment: Downregulation by cjtokines and bacterial products. ]. Exp. Med. 182, 389-400. Sallusto, F., Mackay, C. R., and Lanzavecchia, A. (1997). Selective expression of the eotaxin receptor CCR3 by human T helper 2 cells. Science 277, 2005-2007. Salzinan, A. L., Eaves-Pyles, T., Linn, S. C., Denenberg, A. G., and Szabo, C. (1998). Bacterial induction of inducible nitric oxide synthase in cultured human intestinal epithelial cells. Gastroerlterology 114, 93-102. Samsom, J. N., Langermans, J. A,, Savelkoul, H. F., and van Furth, R. (1995). Tumour necrosis factor, but not interferon-y, is essential for acquired resistance to Listeria imnocytogenes during a secondary infection in mice. Irnmunology 86, 256-262. Sanceau, J., Falcoff, R., Beranger, F., Carter, D. B., and Wietzerbin, J. (1990). Secretion of interleulan-6 (IL-6) by human monocytes stimulated by muramyl dipeptide and tumour necrosis factor a. lrrlmunology 69, 52-56. Sander, B., Skansen-Saphir, U., Damm, O., Hakansson, L., Andersson, J., and Andersson, U. (1995). Sequential production of T h l and Th2 cytokines in response to live bacillus Calmette-Guerin. Immunology 86, 512-518.
368
ULRICH E. SCHAIBLE rt n /
Sastre, L., Kishimoto, T. K., Gee, C., Roberts, T., and Springer, T. A. (1986). The mouse leukocyte adhesion proteins Mac-1 and LFA-1: Studies on mRNA translation and protein glycosylation with emphasis on Mac-1. J . Zmiriunol. 137, 1060-1065. Sato, Y., Roman, M., Tighe, H., Lee, D., Corr, M., Nguyen, M. D., Silverman, G. J., Lotz, M., Carson, D. A,, and Raz, E. (1996). Imniunostiniulatory DNA sequences necessary fbr effective intradermal gene immunization. Science 273, 352-354. Schaible, U. E., Sturgill-Koszycki,S., Schlesinger, P. H., and Russell, D. G. (1998).Cytokine activation leads to acidification and increases maturation of Mycobacteriuin aoirrmcontaining phagosomes in murine macrophages. J. Zmmunol. 160, 1290-1296. Schild, H., Mavaddat, N., Litzenberger, C., Ehrich, E. W., Davis, M. M., Bluestone, J. A., Matis, L., Draper, R. K., and Chien, Y.-H. (1994).The nature of major histocompatibility complex recognition by yl6 T cells. Cell 76, 29-37. Schirmbeck, R., Bohm, W., Melber, K., and Reiinann, J. (1995). Processing of exogenous heat-aggregated (denatured) and particulate (native) hepatitis B surface antigen for class I-restricted epitope presentation. J. Zmmirr~ol.155, 4676-4684. Schirinbeck, 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. Ztr~i7tund.24, 1478-1486. Schirmbeck, R., and Reimann, J. (1996). “Empty” Ld molecules capture peptides from endocytosed hepatitis B surFace antigen particles for major histocompatibility complex class I-restricted presentation. Eur. J , Znimunol. 26, 2812-2822. Schlesinger, L. S. (1993). Macrophage phagocposis of virulent but not attenuated strains of Mycobacterii~rntuberculosis is mediated by mannose receptors in addition to complement receptors. J . bnmunul. 150, 2920-2930. Schlesinger, L. S., Bellinger-Kawahara. C. G., Payne, N. R., and Honvitz, M. A. (1990). Phagocytosisof Mycobucterium tuberculosis is mediated by human monocyte complement receptors and complement component C3. J . Immirncd. 144, 2771-2780. Schlesinger, L. S., and Homitz, M. A. (1991). Phenolic glycolipid-1 of Mycobacteriuni l q r a e binds complement component C3 in serum and mediates phagocytosis by human monocytes. J. Exp. Men. 174, 1031-1038. Schlesiriger, L. S., Hull, S. R.. and Kaufinan, T. M. (1994).Binding ofthe tenninal mannosyl units of lipoarabinomannan from a virulent strain of Myc(ibncterizcni tuberculosis to human inacrophages. J. Zrtitnun~d.152, 4070-4079. Schlesinger, L. S., Kaufinann, T. M., Iyer, S., Hull, S. R., and Marchiando, L. K. (1996). Differences in mannose receptor-mediated uptake of lipoarabinomannan from virulent and attenuated strains of Mycobacterium tuhercrilosis by human macrophages.]. Znmiunol. 157,4568-4575. Schoel, B., and Kaufinann, S. H. E. (1998).Influence of mycobacterial virulence and culture condition on y6 T cells. Microb. Patliogen. 24, 197-201. Schoel, B., Sprenger, S., and Kaufmann, S. H. E. (1994).Phosphate is essential for stimulation of Vy9VS2 T lymphocytes by mycobacterial low molecular weight ligand. Eur.J . Immunol. 24, 1886-1892. Schorey, J. S., Carroll, M. C., and Brown, E. J. (1997). A macrophage invasion mechanism of pathogenic mycobacteria. Science 277, 109 1-1093. Scidmore, M. A., Fischer, E. H., and Hackstadt. T. (1996). Sphingolipids and glycoproteins are differentially trafficked to the Chlamydia truchorruitis inclusion. J. Cell Biol. 134, 363-374. Seder, R. A., and Paul, W. E. (1994). Acquisition of lympholane-producing phenotype by CD4+ T cells. Annu. Reu. Z n ~ t ~ n12, ~ l 635-673. .
INTRACELLULAR RA(:TERIA A N D TllE IMhlUNE SYSTEM
369
Senaldi, G.. Yin, S., Sliaklee. C. L., Piguet. P. F., Mak, T. W., and Ulich, T. R. (1996). C o n ~ t t c ~ ~ o ~ f rpclruronr i u t t ~ and M!ycoboctrrium boois bacillus Calmette-G~ierin-i~i~I~~~~ed graiiiilom;i formation is inhibited in TNF receptor I (TNF-RI) knockout Inice and ly treatment with soluble TNF-RI. /. Z t n r ~ ~ i m d157, . 5022-5026, S h a h a n , A., Pfeffer, K., Lee, K. P., Kiindig. T. M.. Kishihara, K., Wakeham, A,. Kawai, K., Ohashi P. S., Tliompson, C . B., and Mak, T. W. (l99:3).Differential Tcell costinilllatory recjuiremeiits in CD28-deficient micc. Scicricc 261, 609-61 2. Shaw, M . A.. Atkinson, S., Dockrell, H., Hnssain. R., Lins-Lainson. Z., Shaw, J.. R u m s , F., Silveira, F.. Melid, S. Q., Kaukal), F., ct 01. (1993). An HFLP map for 2(]33-q37from multicase mycobacterial and leislinianial disease families: No evidence for an Lsl/Ity/ Hcg gene homolope influeiicingsusceptibilityto leprosy. Ann. Htort. Genet. 57,251-271, J. M., Rodgers, J. R.. and Rich, H. R . (1994). Antigen presentation by major histuconipitihility complex class I-B niolecules. Annu. Rec. Irnmunol. 12, 839-880. Shen, H.. Miller. J. F., Fan. X., Kolwyck, D., Ahnied. R., and Ha*, J. T. (1998).Coinpartmentalization of bacterial antigens: Differential effects on priniing of CD8 T cells arid protective irnmunity. Cell 92, 535-545. Shen, Z.. Remikoff. G . , Dranoff, G . , and Rock. K. L. (1997). Cloned dendritic cells can present exogenous antigens on both MHC class I and class 11 molecules. J , I I I I J I ~ I ~ ~ I I ~ , 158, 2723--2730. Shiniada, S . , Yano, 0..and Toknn;iga, T. (19x6). In vivo anginentation of natural killer cell activity with a deoxyribonucleicacid fraction of' BCG. Jopcin. J . Cnticer Res. 77, 808-816. Sibley, L. D., and Krahenbuhl, J. L. ( 1987).Al!/colmferiirtn lcpro~:-burdenedniacropliages are refractory to activation by y-interferon. I?rfccl. Z r t i r r i u t i . 55, 446-450. Sibley, L. D., and Krahenbrilil, J. L. (1988). Defective activation of granulonia macrophages from Alycoboctcritim kproc-infectrd nude inice. 1, Lrukocyte B i d . 43, 60-66. Sibley, L. D. Hunter, S. W.,Brennan, P. J.. and Kralimhiilil, J. L. (1988). Mycobactend 1il'"ar~ibinoiiianiiaii inhibits y-interf(~ron-inecliatedactivation of nixxopliages. Itrfect. Zitittitttl. 56, 1232-1236. Sibky, L. D., Weidner, E., and Kraheiil)iihl, J. I>.(1985). Phagosonie acidification blocked by intracellular Toroplastnci goridii. Notiire 315, 416-4 19. Sieling, P. A.. Chatterjee, D.. Porcelli, S . A,, Prigozy, T. I., Miuzaccaro, R. J., Soriano, T., Bloom, 13. 13., Brenrier, M. B., Kronenl)erg, M., Brennan, P. J., and Modliii, R . L. (1995). CDI -restricted T cell recognition of microbial lipoglycan antigens. Scirwe 269, 227-230. Sijts, A. J., and Pamer. E. G. (1997).Enliancetl intracellular dissociation of major histocompatibility ccrmplex class I-associated peptides: A mechanism for optimizing the spectrum of cell surface-presented cjotoxic T Iyinphocyte epitopes. /. Exp. Med. 185, 1403-141 1. Sijts, A. J.. Vikuiueva. M. S., mid Panier, E. G. (1996a). CTL epitope generation is tightly linked to cellular proteolysis ofa Listc.rin tiiorioc,llfogu,rcsantigen. I. Zmniunol. 156, 14971503. Sijts, A. J. A. M., Neisig, A,, Nref'jes, J. J., and Piinier, E. C;. (l99Gb).Two Listeria nionocytogenes CTL epitopes are processed from the saincb antigen with different efficiencies. /. f r r i t t ~ i i r t d156, 685-692. Sijts, A. J., Pilip, I., and Painer, E. G. (1997). The Listerin tiioriocytogenr,s-secretedp60 prott+i is an N-end rule sulxtrate in the cytosol of infected cells: Implications for major histocompatibility complex class I antigen processing of bacteriiil proteins. /. Biol. Clicni. 272, 19261- 19268. Silva, C. L., Silva, M. F., Pietro. R. C., and Lowrie, D. B. (1994). Protection against tuberculosis by passive transfer with T-cell clones recognizing mycobacterial heat-shock protein 65. Ztrittit4nohg!/ 83, 341-346.
370
ULHICH E. SCIlAIHLE 6't
N/
Silverinan, D. J., Santucci, L. A,, Meyers, N., i d Sekeyova, Z. (1992).Penetration of host cells by Rickettsia rickettsii appears to be mediated by a phospliolipase of rickettsia1 origin. Infect. Ztrittiiiri. 60, 2733-2740. Sinai, A. P., and Joiner, K. A. (1997).Safe haven: The cell hiology of nonfiisogenicpathogen vacuoles. Artnri. Ben. Micmbiol. 51, 415-462. Sizemore, D. R., Branstroin, A. A,, antl Sadoff, J. C. (1995). Attenuated Shigella as a DNA delivery vehicle for DNA-mediated immunization. Science 270, 299-302. Skeen, M. J., and Ziegler, H. K. (1993). Induction of inurine peritoneal y6 T cells and their role in resistance to bacterial infection. /. Exp. Med 178, 971-984. Skeen, M. J., and Ziegler, H. K. (1995). Activation of yS T cells for production of IFN-y is mediated by bacteria via niacropliage-derivecl cytokines IL-1 antl IL-12. J. Z i ~ ~ m i r t t o l . 154,5832-5841. Smiley, S. T., Kciplan, M. H., and Grusby, M. J. (1997). Iinmunoglobulin E production in the alisence of interleukin-4-secreting CD I-dependent cells. Science 275, 977-979. Snoeck, H. W., Lenjou, M., Nys, C., Lardon, F., Peetermans, M. E., Van Bockstaele, D. R., Moulijn. A,, Haenen, L., and Bernentan. Z.N. (199G).Interleukin 4 and interferoiiy costiniulate the expansion of early human myeloid colony-fomiing cells: Proposal of a motlel for the regulation of myelopoiesis by interleukin 4 and interferon-y and its integration with the regulation of the iminune response. Leukcvnin 10, 117- 122. Song, R., and Harding, C. V. (1996). Roles of proteasonies, transporter for antigen presentation (TAP), and beta 2-microglobrilin in the processing of bacterial or particulate antigens via an alternate class I MHC processing pathway. 1. Irtitriirnol. 156, 4182-4190. Sonnenberg, M. G., antl Belisle, J. T. (1997). Definition of Mycobacteritrni tiihercir1osi.s culture filtrate proteins by two-dimensionalpolyaerylaniide gel electrophoresis. N-terminal amino acid sequencing, and electrospray inass spectrometry. Ztlfect. I n i r n u n . 65, 45154524. Sonoda. Y. (1994). Interleuldn-4--a dual regulatory factor in hematopoiesis. Leukemia .!$Ulp'hOtlKl 14, 231 -240. Sousa, A. O., Salem, J. I., Lee, F. K., Vercosa, M . C., Cruaiid, P., Bloom, B. R . , Lagrange, P. H., and David, H. L. (1997a).An epidemic of tuberculosis with a high rate of tuberculin aiiergy among a population pre\ioiisly unexpostd to tuberculosis, the Yanomanii Indians an Amazon. P n x Nat. Acntl. Sci. USA 94, 13227713232, Sousa, C . R., Hieny, S., Scharton-Kerstrn, T.,Jankovic, D.. C h e s t , H., Gerniain, R. N., and Sher, A. (199%). In vivo inicroliial stimulation induces rapid CD40 ligand-iiidependeiit production of interleukin 12 by dendritic cells and their redistrihiition to T cell areas. J. Exp. Med. 186, 1819-1829, Stahl, P. D. (1992). The inannose receptor and other macrophage lectins. Cirrr. Opirt. ltnntrittol. 4, 49-52. Stahl, P. D., and Ezekowitz, R. A. B. (1998).The nianiio~ereceptor is a pattern receptor iiivolved in host defense. Ciirr. Opin. Zrntritinol. 10, 50-55. Stanford, J. L. (1991). Improving on BCG. Apttiis 99, 103-113. Starnbach, M. N.. Bevan, M. J., and Lampe, M. F. (1994).Protective cytotoxic T lymphocytes are indirced during inurine infection with Chkmydia tracimrutis. 1.ftrtrratnol. 153,51835189. Stein, M. A., Leung, K. Y., Zwick, M., Garcia-del Portillo, F.,and Finlay, B. B. (1996). Identification of a Salmonella virulence gene required for formation of filamentous structures containing lysosonial membrane glycoproteinswithin epithelial cells. Mol. Microbiol. 20, 151-164. Steinhoff, U., Kleinm, U., Greiner, M., Bordasch, K., and Kaufinann, S. H. E. (1998). Altered intestinal in~niunesystem but rrormal antibacterial resistance in the abselice of P-selectin and ICAM-1. /. Intttitirtol. 160, 61 12-6120.
Steininan. R 41., and Swanson, J. ( 1995). The rn(loc!!tic actiijih of dendritic cells. /. Exp. ilkd 182, 283-288, Steiiger. S., Mazzacc;iro, K. J.. CJycniiira. K..Clio. S., Barnes, P. F.. Kosat. J. P., Sette, A., Brcnner, M. B.. Porcrlli, S. A.. Bloom, H. R., and Modlin. K. L. (1997). Differential effects of c:ytol>tic T cell siibscBts o i i intracrlliilar infection. S'ciotccj 276, 1684-1687. of C D l on antigen Strnger. S., V i u i . K. H., m t l M o d h i . R. L. (1998). 1~0~11-rrgiilatiori tiihrrcrik).ui.s./. ~ t t t t t i ~ t m (in l . prcss)). presenting cc.lls by infection witti bf!/col~crc.tcririttt Stokes. R. IV., IIaidl, I . D.. Jefferirs. \V. A.. ;mtl Sprert. D. P. (1993). Mycobacteriamacrophage interactions: Macropliagc. phenot)Fe cleterniines tlie nonopsonic binding of ~l!/c~obnctct-iritti Irrbcwdo.ui.uto iriurinc. macrophagrs. /. I t t i t t i l i n o l . 151, 7067-7076. Stover. C. K.,Bansul. C;. P.. Hailson. M . S.. Biirlein. J. E., Palasq-nski. S. R., Youilg, J. F., Koenig. S.. Y o ~ n ~ D. g . B., Satlzienr, A., ant1 Barlioiir. A. G . (1993). Protective immunity eIicitetI by rcconihinaiit IiaciIIe Caliiiettc.-C:ueiiii ( BCG) r~pressingonter siirface protein A (OspA) lipoprotein: A c:iiididate I .yiw disease vuccine. /. E q ) . Mod. 178, 197-209. Striignell, R. A., Drew. D.. Mercitca, J.. DiN;italtl, S.. Firez, N., Dunstan, S. J.. Sirn~nons, C:. P.. and l'adolas, J. (1997).D N A \ x r i n c s for Iiacteiial infections. Z t t i t ) t r ~ t ~ oCcll l. Riol. 75, :364-369. Sturgill-Koszycki. S., Iladtlk P. I,., a i d Russell. D. G. (1997). Tlicl interaction betwreii niycobacteriuni antl the macrophage anal>.zetl l)y t\~~o-dirnensioiialpolyiic~lamitle gel t~IectropIioresis.E~t~ctro~~~ror~.sis IS, 25~8-2,565. Sturgill-Kosz!cki, S.. Scliailde, U. E.. antl Russell, D. C. (1996). Mycobactcrium-containing phagosoincs are accessihle to early c~ndosoirit~s and reflect a transitional statc in normal phagosoiiie biogenrsis. E M B O /. 15, 6960-6968. Sturgill-Koszycki, S.. Schlesinger. P. II . . (liakrahoit\, P.. Hadtlis, P. L., Collins, H. L., Fok, A. K.. Allen. H. I]., Cluck. S. L., lleusrr, J.. and Kiisscll, D. C. (1994). 1,ack of acidification i n Mycolwteriuin phagosonies p r o c h w l hy ~~xc11isioii of the vesicular protoii-ATPase. Scictrcc. 263, 678-681. Sugita, M.. and l3renner, M . B. ( 1 ~ ~ 5Association ). of the invaliant chain with major Iiistocompiitil,ility coiriple.; class Imoleciilrs directs trafficking to endocytic compartments. 1.H i i d C/rc.t)r.270, 1443-1448. Siigita, M., J w k n i a i i , R. M., van Donsc,l;iar, E.. Beliar. S. M., Rogers. H. A.. Peters, P. J., Brenner, b1. B., and PorccIIi, S. A. ( I 996). ~!!top~asiiiictail-depcwient IocaIization of CD 1 b anti~eii-preseiitinginoleciiles to M IICs. ScicJwc 273, 349-:352. Siigita, M . , Pr)rcelli, S. A,. iintl Brtwner, M . B. (1997). Assentbly and reteiition of C D l b Iiea\y cliuins in the endoplasniic reticiiliini. /. Z t t r t t i i ~ t t o / . 159, 2358-2365. S i i h m . L. b"., Bober. I,. A., Grace, M . J., Braun, S., Macosko. 11. D.. Pa)yandi, F., Sivo, C . C., and Nariila, S. K. (1992). Thts in vivo biological activip of inurine IL-4 and the role of entli)gcwous cytokiiitx /. Cdl. Biocltotr. Sirpp/. 16c, 10:3-109. Suzue, K., antl l'onng, H. A . (1996). Acljuviint-free lisp 70 fusion protein system elicits huinoral u i t l celliilar immiinc responses to HI\'-1 1724. /. Zttrtt~irncd 156, 873-879. Suzue, K.. Zhou, X.. Eism. H. N . , a i i d Yoling, R. A . (1997).Heat shock fnsion proteins as veliicles for antigcan tlelive? into the iiiajor Iiistocompatil~iIi~ complex closs I presentation pathway. Proc. Not. Acod. Sci. U S A 94, 13136-13151. Suzuki. H.. Kiirihara. Y.. Takeyi. M.. Kainiida. N.. Kataoka. M.,Jishage, K., Ueda, 0.. Sakaguclii. I I . . Higaslii. T., Suzuki, T., Takashinia, Y..Kawabe. Y., Cyiishi, IIontla, M., Kiirihara, H., Ahuratani, H., Doi, T., Matsunioto. A., Axuma, Toyoda, Y.. Itakura, El.. Yaznki. Y., IIoiinchi, S., Takahashi, K., Kruijt, J. K., vaii Berkel, T. J. C.. Steinbreclwr. U. P., Isliihashi. S., Maeda, N., Gordon, S., and Kotl;tin;i, T. (1997). A role for inicrophage scavengrr receptors in atherosclerosis and siisceptibility to infection. Nrrtiirr, 386. 292-296.
372
ULRICH E. SCMAIBLE el al
Svensson, M., Stockinger, B.. and Wick, M. J. (1997). Bone inarrow derived dendritic cells can process bacteria for MHC-I and MHC-I1 presentation to T cells. J. Immuriol. 158,4229-4236. Swanson, J. A,, and Baer, S. C. (1995).Phagocytosis by zippers and triggers. Trends Cell Biol. 5 , 89-92. Swanson, M. S., and Isberg, R. R. (1995).Association of Legionella pflelinwphlf~with the macrophage endoplasmic reticulum. Infect. Iinnmn. 63, 3609-3620. Swanson, M. S., and Isberg, R. R. (1996). Analysis of the intracellular fate of Legiorieh prieiirrwphila mutants. Atin. N . Y. Acud. Sci. 797, 8-18. Swier, K., and Miller, J. (1995). Efficient internalization of MHC class 11-invariant chain complexes is not sufficient for invariant chain proteolysis and class I1 antigen presentation. J. Imnunol. 155, 630-643. Szabo, S. J., Dighe, A. S., Guliler, U., and Miirphy, K. M. (1997). Regulation of the interleukin (IL)-12R beta 2 subunit expression in developing T helper 1 (Thl) and Th2 cells. I. E x p . Men'. 185, 817-824. Szday, G., Hess, J., and Kaufniann, S. H. E. (1994).Presentation of Listeria monocytogenes antigens by major histocompatibility class 1 inolecules to CD8 cytotoxic T lymphocytes independent of listeriolysin secretion and virulence. Eur. J. Imn~ztnol.24, 1471-1477. Szalay, G . , Ladel, C. H., and Kaufniann. S. H. E. (1995).Stimulation of protective CD8' T lyinphocytes by vaccination with nonliving bacteria. Proc. Nat. Acad Sci. USA 92, 12389- 12392. Sztein, M. B.. Tanner, M. K., Polotsky, Y., Orenstein, J. M., and Levine, M. M. (199Fj). Cytotoxic T lymphocytes after oral imrnrinization with attenuated vaccine strains of Saltnonella t y p h in humans. J. Imrmrnol. 155, 3987-3993. Takahashi, H.. Takesliita, T., Morein, B., Putney, S.,Germain, R. N., and Berzofsky, J. A. ( 1990). Induction of CD8' cytotoxicT cells by immunization with purified HJV-1 envelope protein in ISCOMs. Nature 344, 873-875. Takai, Y., Wong, G. G., Clark, S. C., Burakoff, S. J., and Herrmann, S. H. (1988). B cell stirnulatory factor-2 is involved in the differentiation of cytotoxic T lymphocytes./. Itnrrzrrnol. 140, 508-512. Takeda, K., Tsutsui, H., Yoshimoto, T., Adachi, O., Yoshida, N., Kishimoto. T., Okamura, H., Nakanishi, K., and Akira, S. (1998). Defective N K cell activity and Th1 response in IL-18 deficient mice. Iinrnrrnity 8, -383-390. Tan, M . C., Mommaas, A. C., Drijfhout. J . W., Jordens, R.. Ondenvater, J. J., Venvoerd, D., Mulder, A. A., van der Heiden, A. N., Scheidegger, D., Oomen, L. C., Ottenlioff, T. H., Tulp, A., Neefjes, J. J., and Koning, F. (1997). Mannose receptormediated uptake of antigens strongly enhances HLA class 11-restricted antigen presentation by cultured dendritic cells. Eur. 1. Zrttiniinol. 27, 2426-2435. Tanaka, Y., Morita, C. T., Nieves, E., Brenner, M. B., and Bloom, B. R. (1995). Natural and synthetic non-peptide antigens recognized by human y6T cells. Nature 375,155-158. Tanaka, Y., Sano, S., Nieves, E.. DeLibero, G . , Rosa, D., Modlin, R. L., Brenner, M. B., Bloom, B. R., and Morita, C. T. (1994).Nonpeptide ligands for human yS T cells. Proc. Nat. Acacl. Sci. USA 91, 8175-8179. Tang, D. C., DeVit, M., and Johnston, S. A. (1992). Genetic immunization is a simple method for eliciting an iinrnune response. Nature 356, 152-154. Taraska, T., Ward, D. M., Ajioka, R. S., Wyrick, P. B., and Davis-Kaplan, S. R. (199G).The late chlamydial inclusion membrane is not derived from the endocytic pathway and is relatively deficient in host proteins. Infect. lmmun. 64, 3713-3727. Tascon, R. E., Colston, M. J., Ragno, S., Stavropoulos, E., Gregory, D., and Lowrie, D. B. (1996). Vaccination against tnherculosis by DNA injection. Nature Med. 2, 888-892.
IhTRAC'ELL.IILAH RA(,TEKIA AYII THE I M M U Y E \\STEM
373
Taylor, M. E:.. and Drickainer, K. (199:3).Structural requirements for high affinity binding of complex ligands by the macrophage mannose receptor. 1. Biol. Chent. 268, 399-404. Teitell, M., 1Iolcoinbe. H. I<., Brossay, L., Hagrnbarigh, A,, Jackson, M. J., Pond, L., Balk, S. P., Terliorst, C.. Peterson. P. A., ~ i n dKronenberg, M. (1997). Nonclassical behavior of the iaaoii~eCD1 class I-like molecule. J. Z t t i t n t i t t o l . 158, 2143-2149. Teixeira-C;oiiaes. A. P., Cloeckaeit, A., Uezartl, G., Bowden, R. A,. Dubray, G., and Zygmunt, M. S. (1997a).Itlentification antl ch~iracterizationof Bnicella oais immunogenic proteins using two-dimensional electrophoresis and imiriunohlotting. Electrop,horc..Pi.s 18, 14911497. Teiseiri-Goines, A. P.. Cloeckaert, A.. Bezartl, C;., Diilmy, G., and Zyginunt. M. S. (199%). Mapping and identification of Bnrrch tttelitrnsis proteins by hvo-dimensional electrophoresis and inicrosequencing. El[.ctni))hore.sis 18, 156- 162. Tepper, R. I., Levinson. D. A., Stanger, B. Z., Cainpos-Torres. J., Abbas, A. K., aiid Ledrr, P. ( I 990). IL-4 iiidnces allergic-like inflammatory disease and alters T cell development in transgenic mice. Cell 62, 157-467. Thierfelder, W. E., van Deursen, J. M.. Yamamoto, K., Tripp, R. A,. Sarawar, S. R., Carson. 11. T., Saiagster, M. Y., Vignali, D. A, Doherty, P. C.. Grosveld, G. C., and Ihle, J. N. (1996). Requirement for Stat4 in interleukin-12-1riediatt.d responses of natural killer and T cells. Notiin, 382, 171-174, Thomssen, H., Ivanyi, I.. Espitia, C., Arya. A., antl Londei, M. (1995). Human CD4-CD8Cup+ T-cell receptor T cells recognize different mycolxacteria strains in the context of CDlb. I t t ~ t / ~ t i n d 85, ~ g / :33-40. Tighr, I I . , Con, M., Roman, M., and Raz, E. (1998). Gene vaccination: Plasinid DNA is morr than just a theprint. Irnrnirnol. Torla!/ 19, 89-97. Tokrmaga, T., Y a m , 0.. Kmmaoto. E., Kiimira, Y., Yamamoto, T., Kataoka, T.. and Yamamoto. S. (1992). Synthetic oligonucleotides with particiilar base sequences from the cDNA encoding proteins of M!p)bnctcriiitu boois BCG induce interferons and activate natural killer cells. Microbial. Z t i m i / t i o l . 36, 55-66. Tripp, C. S., Gitely. M. K.. Hakiini, J., Ling, P., antl Unanue, E. R. (1994).Neutralization of IL-12 dtweases resistance to Listrria in SCID and C.R-17 mice. Reversal 11y IFN-y. I , ~ t l l f / l l i t U J / .152, 388-1887. Turner, J., and Dockrell, H. M. (1996). Stimnlatioia of1mmaia peripheral blood inoiioniiclear cells with live M!/cohocturiiinr bocis BCG activates qtolytic CD8' T cells in vitro. I t t m i i n o [ ( J ~ J 87, 339-342. Turner. S. J.. Carhone. F. R., and Strupre11, 11. A. (19%). Sfi/nlone~/n fyphi?nnriu?ndelta aroA delta aroD mutants expressing a foreign rcwmbinant protein induce specific major liistocompatil)ilit>.complex class I-restricted c~ytotoxicT lyinphocytes in mice. Infect. Z m t n i m 61, 537445380, Udono. H . . and Srivastava, P. K. (1993). Heat slrock protein 70-associated peptides elicit slwcific cancer immunity J . Exp. A4id. 178, 1:391- 1396. Ulmer. J. B., Doniielly, J. J., Parker, S. E., Rhodes, G.11.. Felgner, P. L., Dwarki. V. 1.. Gromkowslci. S. H., Deck. R. R.. DeWitt, C. M.. Friedman. A..vt d.(1993).IIeterologous protection against influenza by injection of DNA encoding a viral protein. S c i c w ~ ~ 259, 1745--1749. Unanrie. E. R . (1997). Iiiter-relationship among macrophages, natural hller cells and neutrophils in early stages of Listeria resistance. Cur. @in. Ztnrnmd 9, 35-43. Urban, J. F. J , Nok&-Traiith. N., Donaldson. D. D., Madden, K. B.. Morris, S. C., Collins, M.. and Finkelinan, F. D. (1998). IL-13, IL-4Halpha and stat6 are required for the ex-pulsion oi'the gastrointestinal nematode parasite Nippmtrongybs brrr~iliensis.Zinniiinity 8, 255-264.
374
ULRICII E. SCIIA[RLE
t’t
ol
Urciiiliart. B. L., Atsalos, T. E., Roacli, D., Basseal. D. J., Bjellqvist. B., Britton, W. L., arid Humphery-Smith, I. (1997). “Proteomic contigs” of Mycobncteriurn titbcrcdosis and M!ycohnctcriwn bonk (BCG) iising novel inimobilised p H gradients. EZt,ctrophorc.sis 18, 1384-1392. van Furth. R., van %wet, T. L., Buisinaii, A. M.. ;md van Dissel, J. T. (1994). Anti-tumor necrosis factor antibodies inhibit the inHux o f granulocytes and inonocytcs into an inHaininatory exudate and enhance the growth of Listerin ttuniocytogeties in varions organs. J . Inject. Dis. 170, 234-237. VanHcyningen, T. K., C o l h s , €1. L., and Rnssell, D. G. (1997). IL-6 produced by macrophages infected with Mycobacterium species suppresses T cell responses. J. Ittt?nrino/. 158, 330-337. van Ooij, C., Apodaca, G.,and Engel, J. (1997).Cliaracterization ofthe Chlnttiydin tr[ichottirttis vacuole m d its interaction with the host eridocytic pdiway in HeLa cells. lrgect. ZitLttttltl. 65,758-766. Vrras, P. S., de Chastellier. C., Moreau, M. F.. Villiers, V., Thibon, M., Mattei, D., and Kahinovitcli, M. (1994). Fusion between large phagocytic vesicles: targeting of yeast and other particnlates to phagolysosomes that shelter the bacterium Coxiclln biirnctii or the protozoan Lcishtttnnia nnin;otien.s 11 Chinese hamster ovary cells.J . Crll Sci. 107,30653076. \’eras, P. S., Moulia, C., Dauguet, C.. Tunis. C. T., Thibon, M., and Rabinovitch, M. (1995). Entry and siiMval of Leishntcinin mrtinzonetisis amastigotes mithin ~iliagolysosome-like vacuoles that shelter Coxiello hrmictii iii Chinese hamster ovary cells. Infict. ~ t t t t t i i i t i . 63,3502-3506. Verlion, A,, Kuijper, S., Jansen, H. M., Speelinan. P., and Kolk, A. H. (1992). Antibodies against secreted and non-secreted antigens in mice after infection with live Mycobncteritini tubcrculosi.s. Senrid. J. I t t m r m o l . 36, 371-384. Via, L. E., Deretic, D., Uhner, R. J., Hibler, N . S.,Huber, L. A,, and Deretic, V. (1997). Arrest of mycobacterial phagosoine inatiiration is caused by a block in vesicle fiision between stages controlled by rab5 and rab7.J. B i d . Chettt. 272, 13326-13331, Via, L., Fratti, R., McFalone, M., Pagan-Ramos, E., Deretic, D., and Deretic, V. (1998). Effects of cytokines on inycobacterial phagosome maturation. J . C d l Sci. 111, 897-905, Vicari, A. P.. and Zlotnik. A. (1996). Mouse NK1.I’ T c d s : A new family of T cells. I t i t r t t ~ i i ~T ~ ~ol d. q 17, 71-76. Vidal, S., Tremblay. M. L., Govoni, G., Gauthier, S., Sebastiani, G., Malo, D., Skamene, E., Ohvier, M., Jothy. S., and Gros, P. (1995).The Ity/Lsl/Bcg locus: Natural resistance to infection with intracellular parasites is abrogatrd by disruption of the Nrntitpl gene. J . Exp. Med. 182,655-666. Vidal, S. M., Pinner. E., Lepage, P., Gautliier, S . , and Gros, 1’. (1996). Natiiral resistance to intracellular infections: N m n i p l encodes il iiieinlmne phosplioglycoprotein I,sent . in macrophages from susceptilde (Nrainpl D 169) mouse strains. J . Z r n t t i i m d 157, 3559-
3568. \’illadangos. J. A.. Riese, R. J., Peters, C., Chapinan, H. A,, and Ploegli, H. L. (1997). Degradation of inouse invariant chain: Roles oi’cathepsins S and D and the influence of major histocompatibility complex polyniorphism. J. Exp. Med. 186, 549-560. V i l h i e v a , M.S.,Sijts, A. J., aid Pamcr, E. G. (1995). Listeriolysin is processed efficiently into :in MHC class I-associated epitope in Listeria inonocytogeiie.s-infecte~1cells.J. I t t i t t i u nol. 155,5225-5233. \’her, N.,Nelson, C. A., Deck, B., and U i i a n i i ~ ,E. R. (1996). Complexes generated by the Iiinding of free peptitles to class I1 MHC inolecriles are antigenically diverse coinpard with those generated 11y intracellular processing. J . Ittattiunol. 156, 2365-2368.
INTRACELLULAR BACTERIA AN13 T H E IMbIUNE SYSTEM
375
Vouret-Craviari, \7., Mattencci, C., Peri, G., Poli, G., Introna, M., and Mantovaiii, A. (1997). Expression of a long pentnutin, PTX3, hy monocytes exposed to the mycohacterial cell wall component lipoarabinomannan. Infect. Ininiun. 65, 1345- 1350. Waiig, C.-R., Castano, A. R., Peterson, P. A,, Slanghter, C., Fischer Lindahl, K., and Deisenhofer, J. (19%). Nonclassical binding of forinylated peptide in crystal structure of the MHC class Ib molecnle H2-M3. Cell 82, 655-664. Watarai, M., Fnnato, S., and Sasakawa, C. (1996). Interaction of Ipa proteins of Skigelln Jlesncri with a 5 p l integrin promotes entry of the bacteria into mammalian cells. J . Esp. Med. 183, 991-999. Watts, C. (1997). Capture and processing of' exogenous antigens for presentation on MHC molecules. Annzi. Rru. Irrununol. 15, 821-850. Weidemann, B., Scliletter, J., Dziarski, R., Knsninoto, S., Stelter, F., Rietschel, E. T., Flad, €1. D., arid Ulmer, A. J. (1997). Specific binding of soluble peptitioglycan and muramyldipeptide to CD14 on human monocytes. Infect. Iiimrtn. 65, 8.58-864. Weikert, L. F., Edwards, K., Chroneos, Z.C., IIager, C., Hoffman, L., and Shepherd, i'. L.(1997). SP-A enhances uptake of bacillus Calmette-Guerin by macrophages through a specific SP-A receptor. A m J . Physiol. 272, L989-99.5. Welch, M. D., Iwaniatsu, A., and Mitchison, T. J. (1997). Actin polymerization is induced by Arp2/3 protein complex at the surface ofListeria monocytogcnes. Nntitre 385,265-269. Weldingh, K., Rosenkrands, I., Jacobsen, S., Rasmussen, P. B., Elhay, M. J., and Andersen, P. (1998).Two-dimensional electrophoresis for analysis of Mycohncterium tuberculosis culture filtrate and purification and characterization of six novel proteins. Infect. Iininun. 66, 3492-3500. IVesch, D., Kabelitz, D., Friese, K., and Pechhold, K. (1996). Mycobacteria-reactive y6 T cells in HIV-infected indidnals: Lack of Vy9 cell responsiveness is due to deficiency of antigen-specific CD4 T helpcr g ~ 1ecells. Eur. J . Imiiiunol. 26, 557-562. Wesch, D., Marx, S., and Kahelitz, D. (1997). Comparative analysis of ap and y6 T cell activation by M!lcobncterir/nL tiiberciilosis and isopentenyl pyrophosphate. Eur. J . I i n n i i i nol. 27, 952-9.56. West, M. A,, Lucocq, J. M., and Watts, C. (1994). Antigen processing and class I1 MHC peptide-loading compartments in hriinan B-lymphoblastoid cells. Nature 369, 147-151, White, A. G., Bogh, J., Lehrny,M'., Kuchipiidi, P., Varghese, M., a1 Riyami, H., a1 IIashrni, S., and Daar, A. S. (1997). HLA antigens in Omanis with blinding trachoma: Markers for disease susceptibility anti resistance. Br. J . Ophthalrrud. 81, 431-434. \.17ilson,R. L., and Harty, J. T. (1996). CD8' T cells in intracellular bacterial M7hite, D. W., infections of mice. Re.s. Itninzinol. 147, 519-524. Wick, M. J., Harding, C. V., Norinark, S. J., and Pfeifer, J. D. (1994). Parameters that influence the efficiency of processing antigenic epitopes expressed in Saliiionella typhiiintrium. Infect. Imniun. 62, 4542-4548. Wick, M. J., and Pfeifer, J. D. (1996). Major histocompatibility complex class I presentation of ovalbiimin peptide 257-264 from exogenous sources: Protein context influences the . 2790-2799. degree of TAP-independent presentation. Eur. J . 1 ~ 1 n 1 n i n o l26, Wick, M . J., Pfeifer, J. D., Findlay, K. A , , Harding, C . V., and Normark, S. J. (1993). Compartmentalization of defined epitopcs expressed in Eschericlzici coli has only a minor influence on efficien of phagocytic processing for presentation by class 1 and class 11 major histocornpatihi y complex molecules to T cells. I I f e C t . Inrmun. 61, 4848-4856. Wiker, H. G., Nagai, S., Harhoe, M., and Ljungqvist, L. (1992). A family of cross-reacting proteins secreted by A4ycohacteritin~tubercuhis. Sccinrl. J . Imrniinol. 36, 307-319. Wileman, T., Boshans, R., and Stahl, P. (1985). Uptake and transport of mannosylated ligands by alveolar macrophages: Studies on ATP-dependent receptor-ligand dissociation. J . Biol. Chein. 260, 7387-7393.
376
ULRICH E. SCHAIBLE Pt d.
Wolf, P. R., and Ploegh, €1. L. (1995). How MHC class I1 niolecules acquire peptide cargo: biosyntliesis and trafficking through the endocytic pathway. Annu. Rev. Cell Deo. Biol. 11,267-306. Wong, D. K., Gobin, J., Honvitz, M. A,, a d Gibson, B. W. (1996). Characterization of exochelins of MycoDacteriuniaviicin; Evidence for saturated and unsaturated and for acid and ester forms. J. Bacteriol. 178, 6394-6398. Wooldridge, K. G., and Williams, P. H. (1993). Iron uptake mechanisms of pathogenic bacteria. FEMS Microhiol. Rev. 12, 325-348. Wright, S. D. (1995). CD14 and innate recognition of bacteria. J. Imniunol. 155, 6-8. Wut)bolts, R., Femandez-Borja, M., Ooinen, L., Venvoerd, D., Janssen, H., Calafat. J., Tulp, A,, Dusseljee, S., and Neefjes, J. (1996). Direct vesicular transport of MHC class I1 inolecules from lysosoinal stnictures to the cell surface. J . Cell B i d . 135, 611-622. Wynn, T. A,, Mordwetz, R., Scharton-Kersten, T., Hieny, S., Morse, H. C . , 111, Kuhn, R., Muller, W., Cheever, A. W.. and Slier, A. (1997). Analysis of granuloma formation in double cytokine-deficient inice reveals a central role for IL-10 in polarizing both T helper cell 1- and T helper cell 2-type cytokine responses in civo. J. Immrmol. 159,5014-5023. Wyrick, P. B., and Brownridge, E. A. (1978). Growth of Chlamydia psittaci in macrophages. Infect. Itnnncti. 19, 1054-1060. Xu, D., Chan, W. L., Leung, B. P., Iluang, F.-P., Wheeler, R., Piedrafita, D., Robinson, J. H., and Liew, F. Y. (1998a). Selective expression of a stable cell surface molecule on type 2 but not type 1 helper T cells. J. Exp. Med. 187, 787-794. Xu, D., and Liew, F. Y. (1995). Protection against leishmaniasis by injection of DNA eucoding a major surface glycoprotein, gp63, of L. nmjor. Ininitinology 84, 173-176. Xu. D., McSorley, S. J., Tetley, L., Chatfield, S., Dougan, G., Chan, W. L., Satoskar, A,, David, J. R., and Liew, F. Y. (199%). Protective effect 011 Leishinutiin m j o r infection of migration inhibitory factor. TNF-a and IFN-y administered orally by attenuated Salnuin e h fyphiniicritim.1.Imniund. 160, 1285-1289. Xu, S., Cooper, A., Sturgill-Koszycki,S., VanHeyningen, T., Chatterjee, D., Orme, I., Allen, P., and Russell, D. G. (1994). Intracellular trafficking in Mycobncteriziin tiiberailosis and Mycobacteriuin aviuni-infected macrophages. J . Immttnol. 153, 2568-2576. Yamada, K. M. (1991). Adhesive recoglition sequences. ]. Biol. Chem. 266, 12809-12912. Yainamoto, S., Kuramoto, E., Shimada, S., and Tokunaga, T. (1988). In vitro augmentation of natural killer cell activity and production of interferon-rulpand - y with deoxyribonucleic acid fraction from Mycohucterium hovis BCG. Japan. J. Cancer Res. 79, 866-873. Yainamoto, S., Yamamoto,T., Kataoka, T., Kuramoto, E., Yano, O., and Tokunaga, T. (1992). Unique palindroinic sequences in synthetic oligonucleotides are required to induce IFN and augment IFN-mediated natural killer activity. J. Ininizinol. 148, 4072-4076. Yamamoto,T., Yamamoto, S., Kataoka,T., Komuro, K., Kohase, M., andTokunaga, T. (1994). Synthetic oligonucleotides with certain palindromes stimulate interferon production of human peripheral blood lymphocytes in vitro. Japan.J . Cancer Rex 85, 775-779. Yamamoto, Y., Klein, T. W., and Friedman, A. (1996a). Immunoregulatory role of nitric oxide in Legionella pneuirurl,hila-infected macrophages. Cell. Imniunol. 171, 231-239. Yamamoto, Y., Klein, T. W., and Friedman, H. (199613).Induction of cytokine granulocytemacrophage colony-stimulating factor and chemokine macrophage inflammatory protein 2 inRNAs in macrophages by Legioiielln pnercnwphila or Salnwnella typhimuriuin attachment requires different ligand-receptor systems. Infect. ImnLun. 64, 3062-3068. Yewdell, J. W., Esquivel, F., Arnold, D., Spies, T., Eisenlohr, L. C., and Bennink, J. R. ( 1993).Presentation of numerous viral peptides to mouse major histocompatibilitycomplex (MIIC)class I-restricted T lymphocytes is mediated by the human MHC encoded transporter or by a hybrid mouse-human transporter. ]. Erp. Med 177, 1785-1790.
INTRA(:EI,I,ULAH BACTERIA .4h‘Il THE IhlblUNE SYSTEhf
377
Yoshida, A., aiid Koidr. Y. ( 1997). Aral)iiiofuraiiosyl-teriniriati~tland miiniiosylated 1ipoaral)iiioiniinniins from nlyrol,rrc.fcn‘riitI trrhc,rrii/o.\.isintlucc different levels of interleukin- 12 exprcssion in niiirine iiiacrophages. Iitficf. Z t t t t t t u i t . 65, 1953- 1955. Yoshiinoto, T . Bentlelac. A., N’iitson, C., Hri-Li. 1.. mtl Paul, Lt’, E. (1995).Hole ofNKI.1’ T cells in a ‘TH2 rcsponse antl i n i m ~ ~ ~ ~ ~ ~ i oEg prodtictioil. l o ~ ~ i ~ l i Scicwcc n 270, 1845-1847. Yoshimoto. T and Paul, \t’. E. (1994). CII4’NKl .I’ T cells promptly products interlciikiii 4 iii response to in oico cliallenge uith anti-CD3. J . E . v ~ JAlcd . 179, 128,5-1295. Young, D. B., and Carbe. T. K. (199 1). Heat shock proteins ; u i d antigcns of Ill!/‘.”b//etc,r-iiiiit tnhctrirbni.c.. Itfc,ct. l i i i t i t i r n , 59, 308(i-:3093. Zaitseva. M.. Goltliiig. H., hlmischewvitz. 1.. \t’el)l). D., aiitl Chltliiig, U . ( 19%). Bnrcc.//(r a/~ortri.sa s .I potential vacciiic. canditlatc.: Indrictioii of iiiterleiikin-12 secretion and t w 1i;inced B7.1 antl B7.2 antl intercellrilar ittllirsioii i i i o l t a r l e 1 srirf’ke exprc*ssion in cliitriated huniaii iiionocytes stiiiiiilated by Iieat-inacti\,iitrd B. olrorfris. I i t j k t . l i i t i i t i i i t . 64, 3 1OY-:3 1 17. Zeng, Z.-H., Castano, A. K., Segelke, 13. W., Stiil.a, E. A,, I’c~terson.P. A,. ~ i n d\Vilsoii, I. A. (19971 Crystal striictnre of nioiise CD1: A n MHC-like fold with :I largr hytlrophobic binding groovr. Scimce 277, 3:39-:345. Zhang, Y., Broscr, M.. and Rom. W.N . (1994). Activatioii of the interleukill 6 gems I)? Af~~cobncfcriuii~ tirhf,rcdosis o r Ii~~o~~oI~s~iccIiiiritlt~ is nietlietrtl by niiclrx factors N F-IL6 and NF-KB. Proc.. Naf.Acnd. Sri. USA 91, 222sZliong. W.. Wiesmuller, K. H., Kranirr, M . D.. LVdIicIi, K.,and Siinon, M . M. (1996). Plasmid DNA and protein vaccinatioii of mice to thtt outer siirfiice protein A of Borrdin hirrgr/oij& leads to inductioii of T Iiclpctr cells with specificih for ii major epitopcs x r t l augnientation of protective IgC, aiitiliodies in vivo. h r . J . ~ t t ~ i t t l ~26, m J ~374-2757. . Zliii, X., Veiikataprasad, N.. Thangaraj, €1. S.,[lill. M.. Singll, M.. lvanyi. J., and L’orderinrier, H. M. ( 1997). Functions and specificity of T cells fdhviiig iiiickic acid vuccinatioii of mice against nl!/cobnctc,riiiiit f i i b c n i r l o s i s infection. J . I t t ~ t t i i r i t o / .158, 5921-Fj926. Zinimerli. S.. Edwards, S., and Enist, J. D. (1996). Selective receptor I)lockatlc (hiring pliagocytosis does not alter the siiivi\,al and growtli of nf!/cohnc.tc,il’rrit1 trrbcrctt/osi.r i n human macrophages. Am. J . Rcsp Co// Mo/.Biol. 15, 760-770. Z\Yilling. B. S.. and Hilburpr. M. E. (1994). Macrophag(% resistance. giwes: Bc@ty/Lsh. Z t t t t t t t r r t o / . Scr. 60, 233-245. ZyclilinsLy. A,. m c l Sansonetti, P. J. (1997). Apoptosis a s ii I)roiiifiaiiiniat[)n event: What can we learn from bacteria-indricetl cc-ll de;itli;.’ 7’rmdY Afic.ro6iol. 5, 201 -204. ~
This article was accepted for prhlication
0 1 1 Mil!
21. 1998.
This Page Intentionally Left Blank
INDEX
AIDS, 188-190, see also Human immunodeficiency virus Anergy, 243-248 Antibody responses, 95-98 Antigen-presenting cells bacteiial invasions, 292-295 negative selection, 232-234 Antigens phosphate-containing, 88-89 protective, 328-329 T cell tolerance, 112-115 APC, see Antigen-presenting cells APO-UFas proteins, 164 Apoptosis Bcl-2 proteins, 172-174 capase-induced, 176-180 CD95/CD951Asystem AIDS, 188-190 characterization, 166-167 chemotherapy, 184-188 death domain, 170 death receptor signaling, 175-180 discovery, 165 flips, 182 gene defects, 167-168 hepatic homeostasis, 169 peripheral ‘r cell deletion, 168-169 signaling molecules, 170-172 signal trausduction, 169-170 T cell sensitibrity, 182-184 type I and type I1 cells, 180-182 characterization, 163-164 clinical applicdtions, 190-191 DISC, 174-175 ligands, 164-1 66 mechanism, 163- 164
receptors, 164-166 TNF-induced, 178-179 types, 163 Autoimmune lymphoproliferativesyndrome, 167
B Bax, 172-173 B cells antibody-producing, 96 y6 T cell response, 93-95 negative selection, 238-239 T cell anergy, 245-246 thymus population, 9-10 BCG vaccine, 327-332 Bcl-2 proteins, 172-174
C CAP proteins, 174-176 Cardiogenesis, 216-217 Caspases, 176-180 CDR loops, 84-85 Cell culture studies, 52-53 Cell surface markers application, 13-14 description, 5-8 Cell surface receptors, 234-236 Chemokine induction, 276-278 Chemotherapy, 184-188 Competitive rearrangement model description, 17 revised, 57 379
INDEX
(;oinpleinents CS. 325-326 CD3.91-92 receptors, 27 1-272 Costiniiilation, :317-318 CTLA-4, 244-245 Cirtarieoits hypersensitivity, 111-112 CXC chemokine SD F-llPBSF cardiogenesis, 216-217 CXCR4 receptor cxpression. 219 HIV-1 and, 219-222 identification, 217-218 sipid trmisdrrction, 217-218 stnicturr, 217-218 dr\~elopnlent.215 discoveiy. 2 11-212 expression, 213-214 gene, 212-213 Iwnatopoiesis, 215-216 identification. 2 12 stnictiirc, 214-2'45 Cyclosporin A, 4-5
DAXX proteins, 171 Death-inducing signaling comldrx, 174-176, 182-184 Dendritic cells. 11 DISC, .see Death-inducing signding complex DNA, iiaked, 333-335
E Enceplialoinyelitis. 111 Entlothelin-1. 217 Epitlielial Iiotneostasis, 122-123
G G8 clone, 90 Gene rearrangements BID, 86 general considerations, 19 genr-targeted mice, 44-47 models, 34-35 molecular analysis, 20-22 TCHa locus, 32-34 TCRP locus, 28-32 TCRG lociis, 24-27 TCHy locus, 22-24 Gene-targeted mice lineage commitment, 47-50 rearrangements. 44-47 TCH chain disruptions. 43-44 g/d gene, 167-168 Glucocorticoids receptors, 152 sensitivity, 4-5 Clycolipid presentation, 303-306 Craft ocrsiis host disease, 119-120 Cranilloina formation, 319-32 1 GVHD, see Graft versus host disease
H Heart transplantations, 118 Heat shock proteins, 333 f feniatopoiesis, 215-2 lfi Herpes simplex virus y6 response, 105-106 ligand recognition, 90-91 [Iomeostasis epithelial, 122-123 hepatic, 169 HSV, ,see Herpes simplex virus I Iuinaii immunodeficiency v i r u s , 219-222, see &J AIDS IIypersensitivity, cutaneoos, 111-1 12
F I FADI) proteins C D Y S signaling, 170-172 DISC signaling, 174-176 FAS gene, 101 FLICE inliibitory proteins, 182
lmoiune system cell interactions antibody response, 95-98 B cell/T cell, 93-95
381
INDEX
intr;icellular bacteria conflict. 267-2668 melnory, 2 T cell deviation, 2468 Immunoglohulins. 97-98 InHamin&n. 108-1 11 Influenza hemaggliitinin. 238 Ingested antigens. 112-1 15 Interferon-y granuloma formation, 320 inappropriate responses, 32 I indriction, 274-275 lympliopoeisis, 321-:322 inacropliage .ictivation, 319 macrophage priming, 104- 105 STAT-deficient mice, 152- 155 Interlcukins genetic infection-linkage. 326 granuloina formation, 320 inappropriate responses, 321 intinction. 251-275. 275 lynipliopoeisi<,321 -322 niacrophagr ;ictivation, 104 STAT function, 15.5-156 thyniopoiesis. 52-53 Intracellular bacteria, . s w dr.o s\wrific spc description, 268-269 host invasion antigen picessing, 292-300 uiiconventional, 302-307 c o s t i i n u h m . 317-318 c~toplasmc'scape, 290, 292 tiision inhil)ition. 285-286 granrihna formation, 319-321 host genetics, 322-:326 ininiune failsafe nieiisures, 321 Iynpliopoeisis. 321-322 inacrophagi. activation, 318-319 maturation, 269-273 inrclianism. 269-273 MHC class I1 presentation, 300-,302 phagolysosoine sunival, 288-280 srlnival, 281-292 T cell subsets conventional. 307. 310-312 uncon\,rritiorial, 3 13-31 7 vaccines naked DNA. 333-335 protectivix antigens, 328-329 strains, 329-331
strategies, 326-328 subunits, 331-33:3 1ntracelliil;ir iron, 281-282 Intraepithelial lyinpliocytes, 113-114 Zty grnc, 107-108
J Jak kinascs, 148
K Ker'itinocyte growth factor. 122 K N B clone. 90, 98-99
L LBK5 clone. 89-90 mrsironn, 285, 288 L o i . v h ~ ~ t ~ ~~i i~i cI i I ~ ~ I ~ I 285. I I ~ 288 ~ ~ Lc+shiiwiiicr
~ I ~ ~ ~ I ,
Ligantl recognition HSV-specific, 90-91 MHC-specific, 89-90 sTCR constnicts, 92-93
Li~)ol~olysaccliari~les, 270 Lisfcrio i i ~ ~ ~ t ~ i c ~ t ~ i ~ i ~ i i r ~ r . c)tolytic CDX' T cells, 312 y 6 T cells. 313-314 IL-6 intliiction. 276 IIIouse models, 106-1 11
Liver Iionicostasis, 169
[pr gene, 167- I68 LPS, s w Lil~opolysacchandes Lynipliopoeisis, 321-322
M Macrophages activation. 318-319 antimicrohial defense mechanisms intracelliilar iron, 281-2882, p q t i d e s , 282-283 toxic cffector molecules, 278-2881 cheinotactic proteins, 277 yS T C C ~ and, S 102-105 niannose receptor, 270
INDEX
Major 1ristocoiiipntit)ililycoiiiples class la restricted. 296-300 class Ib, 302-303. 316-317 class 11, 300-302 grnctic inf[,ctioii-linkage, 324-325 yS T-cell rc~trictioris.100 1ig;iricl rccogiiitioii. 8(i-87 T cell esprcssion. 86-87 M E K I . 2-37 Mice ,qelle-targc.ted liiieagc. coiiiiiiitiiirnt. 47-51 I‘(,’ .II.iaiigeiiierits. . 44-47 TCR cliaiii tlisluptions, 43-44 STAT-deficient, 152-15.5 transgenic
IIEC, :38-39,42-43 GX, 40-43 TCHrufi, 35-3i TCRyS, :37-43 M M H , s w Mxrophage iiiaiiiiose receptor ,2f!lcohcictc~riirrricruitoti, 377 ~ \ f ~ / ~ ~ ~ ~ h ~ l /)OL;iS, ~ ~ ~ ~ 330 l 1 ’ l l l l l ,Il!/cobclc.tcrilllll spp., 286-288 ,Il!lcdjnctcriii t I 1 t I 111ijrc71h
P p56, 237 PAK2, 179 I’atliogeii resistance, 105-108 1’li;agocytrsis. 270-272 Pl~agolysosomes,288-390 Phagosouie iiiatufiition blocking, 286-288
cytoplasm escape, 290. 292 ~nechmism.283-286 Pliosi’li~atr-containiiig antigens, 88-89 P h t i o d i i i t n yodii, 108 Pregn;lncy, 120- 122 Programmed cell death, .see Apoptosis Protein tyrosine kiiiases, 172
R RIP proteins, 170-179
i , 9
chviiiokine response. 277 cytol!tic CD8’ T cells, 312 gl.;inuloiiia foriiiaticirr, 320 yS T-crll response, 10:3, 107 y6 T cells, :306-307, 313-314 I L-6 respiis(’, 276 tosic csff‘ector ~noleciilrs,278 vaccines. 327-328 Myelopoiesis, 216
N iiket l 11N A, 33:3-335 Natiiral killcr cells hacterial iii\.usions, 315-31 6 C1195L expression. 166 yS T cells and, 101-102 tlrynius ppiilation. 10- 11 Nitric oxide, 104-105, 278-281 Notch protcin, 50-51 Nramp, :322-324
0 Ontogmetic wavrs, 79-81
S / / / l t l o t i d / f 7 spp.. 107-108 Schisfosotiio spp,, 277 Separate \incage iriodcl, 17- 18, 59-60 Sc~c~uriitial rearrangement riiotlel, 16-17 Signal transcluctioir, 169-170 y Silencer. 3 9 STAT gene tlrficient mice. 152-155 f b i l y , 145-146 fiinction cellular prdiferation. 155- 156 therimtion. 1411-149 disease, 155- 156 DNA I)iirding, 149-150 niiclmr localization, 149 receptor L)intling, 146-148 regulation, 1,56-1,57 SIi2 domain. 146-148 transcl-iptionul activation. 150-152 structrire dinierization, 148-149 DNA hincling, 149-150 nilclear localization, 148 receptor hintling, 146-148
383
INDEX
T T cells 1)acterial inwsiolls conventional. 307. 310-:312 u~~coiivei~tionol. ,313-31 7
C 114/<:D8 ;is cell surface markers. 5-8 y tlt\;liroc:ytes. I2 <:DSJ/C:D95L systenl. 166- 167 deletions, 229, 23 1-2:32 glycolipid presentation. 30:3-:30fi (YPliricagc cell culture studies, 52-53 competitivc rearrangement niodel. 57, 59 coIIsellsIls Inodel, 54-5; data integration. 60-63 LIP thpmocyte rcpresent;itioll, 46-49 y6 interat tion, 98- 104 notcl1 role, 50-51 separate lineage iiiotlrl, 59-60 d y lineage coiiipetiti\~rearrangement niotlel, 17, 19 tlt~vclopnll~nt. 14- 16 fiinction, 2 identification. 1-5 ~iiainteriai,cc~, 18 sc4cxtion. :3-4 separates lineage iiiodel, 17- 18 seqnential rcarrangen~entmotlel, 1lj- 1; TCR gene rrarr;ingenients general coiisitlrwtions. 19 IllOd& :34-:35 ~iiolrciilnranalysis, 20-22 TCRa 1:)ciis. 32-34 TCHP lociis, 2 - 3 2
T(:HS locus. 24-27 TCHy locus. 22-24 y6 lineage activation, 91-92 antibody r'csponses. 95-98 I)wteri:il invasitrris, 306-:307. 313-314 H cell stiniulation. 93-95 cutaneous hfimsensiti\ih, 111-112 de\~eloplllrrlt,78-8:3
tlistril)iition. 78-83 c~volution.123-124 aP interaction, 98-104 infi;uniriation. 108- 1 I 1 liguiiti recognition HS\'-specific. 90-91 MH(;-specific, 89-90 rTCR constn~cts,92-93 mlrcrophages, 102-105 moleculc stinii~lation,87-89 N K cells, 101-102 origin, 78-83 oven~irw,i7-78 pathogen response, 105- 108 sclrction, 85-87 strrictrire specificity, 83-85 tissue spec,ificity, 8O-8:3 tolerance antigens, 112-117 epithelial Iiomeostasis, 122-123
wonrid Iie;iling, 122-12:3 \va\'c~s, 79-81 negntivt. selection, 238-239 pt*ripliei-;il CDS5 apoptosis, 182-184 dek~tion,168-169 tolrrance a n e r p , 243-248 tleletion, 2-42-243 cdiaiistioii, 242-243 i i r i i r i r i n c ~tlc~iation.248 partial signaling, 243-2448 recirculation, 240-242 rrgiilation. 248-2550 sensiti\ity, 4-5 tl~yiriicdevelopnient, 3 TI0 clone. 90 T22 clone, 90 Tgl4.4 clone, 90-91 TI~ynioc!tes (:D4K:I)8 schenle. 12 DN hrtrrogeneit).. 8-1 1 DP gcnr-targrted mice, 4-1-45
(YOlineage rrpreseiitation, 46-49 transgenic mice, 40-41 ISP stage, 3 2 notch role, 50-51
3x4 Thyniopoiesis, 5-8 Thytinis lineage divergence, 53-54 pTa-deficient, 47-48 tolerance clonal deletion, 229-232 negative selection APC, 232-234 cell surface receptors, 234-236 clonal deletion, 237-240 signaling. 236-237 Tissue grafts, 117-120 Tolerance yS T cells antigens, 112-117 epithelial homeostasis, 122-123 pregnancy, 120-122 tissue grafts, 117-120 wound healing, 122-123 peripheral T cell anergy, 243-248 deletion, 242-243 exhaustion, 242-243 immune devkation, 248 partial signaling, 243-248 recircidation, 240-242 regulation, 248-250 tllylnus clonal deletion, 229-232 negative selection APC, 232-234 cell surface receptors, 234-236 signaling, 236-237 Toxic effector molecules, 278-281
INDEX
Transgenic mice DEC, 38-38, 42-43 68. 40-43 TCRaP, 35-37 ‘TCHyS, 37-43 Trigger mechanism, 269 Tumor necrosis factor granuloma formation, 320 inappropriate responses, 321 induced apoptosis, 178-179 induction, 275 lymphopoeisis, 321-322 Tyrosine phosphorylation, 148-149
v Vaccines adjuvants, 331-333 naked DNA, 333-335 protective antigens, 328-329 strains, 329-331 strategies, 326-328 subunit, 331-33:3 Vaccinia virus, 106
Waves. SFP Ontogenetic waves
z ZAP-70, 237 Zipper mechanism, 269
CONTENTS OF RECENT VOLUMES
Major Histocompatibility Complex-llirectcd Siisceptibility to Hheriniatoid Arthritis C;EI~AI,I> T. NEPUM
Volume 68 Posttraiiscriptioiial Regilatioii of niRNAs Important in T Cell Function J A M E S S. ~ ~ A I . T E I < Molecular and Celliilar Mcclranisins of T Lyniphocyte Apoptosis JOSEF M I'ENNINC:ER .ANI) C ~ J I I ~ O KHOEMEH
Generation antl TAP-Mediated Transport of Peptides for Major Histocompatil)ilih Complex Class I Moleciiles
F R ~ NM(!MHI'HC; K HA hl hl E R LI N(:
.4ND c ( : U T H E R
J.
Adoptivr T u i n o r Inrniiinity Mediated by Lymphocytes Bearing Modified .4ntigenSpecific Receptors Ttiobi.As H I ~ ~ ( : K AND E K KI.AL'S K.\HIA~.AINI:N Mriiibrane Molecules as Differentiation Antigcns of Murine Macrophages AVL)HE\VJ . MrKNicii.r .ANI) SLAMON <;OHI)ON
Role of Inrnriiiroreceptor Tyrosine-Based Activation Motif in Signal Transduction from A n t i p i ant1 I:? Receptors NoRtf l s , \ ~ o \
Agl>ical Seiine Protrases of t h * C:oniplcinent Syst e 111 G~HAR J.U AHI.AUI>. J O H N E. VOIAN.AKIS, NICOLEM. T H I E L E N S , STIIAN.AM V. L. NARAYANA, VI~K)NIQU Hossi, E \NU YLIANYUAN Xu 3x5
386
CONTENTS OF RECENT VOLUMES
Accessibility Control of V(D)J Recombination: Interleukin-18: A Novel Cytokine That Augments Both Inirate and Acquired Lessons from Gene Targeting Iinninnity WILLIAM M. HEMPEL,ISABELLE LEDUC, HARUKI OKAMURA, HIHOKCI TSUTSUI, NOELLEMATHIEU, RAJKAMAL TRIPATHI, SIIIN-ICHINO KASHIWAMUHA, TOMOHIRO A N D PIERRE FEHIiIEH YOSHIMOTO,A N D KENJINAKANISHI Interactions between the Iininunc System and Gene Therapy Vectors: Bidirectional CD4’ T-cell Induction and Effector Regulation of Response and Expression Functions: A Comparison of Immunity J O N A T H A N S. RHOMBEHC;,LISADEBHUYNE, against Soluble Antigens and Vird Infections A N D L I H l i l QIN ANNETTEOXENIUS, ROLF M. ZINKERNACEL, A N D HANSHENCARTNEH Major Histocompatibility Complex Genes Influence Individual Odors and Mating Preferences DUSTIN PENNA N D WAYNE Pons Olfactory Receptor Gene Regulation ANDREWCHESS
INDEX
Volume 70 Biology of the Interleukin-2 Receptor BRADH. NELSON A N D DENNIS M WILLERFOIIU Interleukin-12: A Cytokine at the Interface of Inflammation and Iininiinity G i o m o THIN(:IIIERI Recent Progress on the Regulation of Apoptosis by Bcl-2 F m i l y Members ANDYJ. MINK, RACHEL E. SWAIN, AVERIL MA,A N D CRAIG B. THOMP60N
Ciirrent Views in Intracellular Transport: Insights from Studies in Iinmunology VICTORW. HSU AN11 PETERJ. PETERS Phylogenetic Emergence and Molecular Evolution of the Iminunoglohiilin Family J O H N J. MAHCHALONIS, SAMUEL F. SCHLIITER, RALPHM. BERNSTEIN, SHANXIANC SHEN,A N D ALLENR. EDMUNDSON Current Insights into the “Antiphospholipid’ Syndrome: Clinical, Immunological, and Molecular Aspects A. KANDIAH,ANDRE] SALI, DAVID Y O N G H U A SHENG, EDWARD 1. VICTORIA, DAVID M. MARQUIS, STEPHEN M. COUTTS, A N D STEVEN A. KHILIS
INDEX
This Page Intentionally Left Blank
