Advances in Immunology Volume 83 T Cell Subsets: Cellular Selection, Commitment and Identity

ADVANCES IN Immunology T Cell Subsets: Cellular Selection, Commitment and Identity VOLUME 83 This Page Intentionally...

Author: H. Cantor | Laurie Glimcher | Frederick W. Alt

24 downloads 461 Views 4MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

ADVANCES IN

Immunology T Cell Subsets: Cellular Selection, Commitment and Identity VOLUME 83

This Page Intentionally Left Blank

ADVANCES IN

Immunology T Cell Subsets: Cellular Selection, Commitment and Identity EDITED BY HARVEY CANTOR Dana-Farber Cancer Institute Department of Cancer Immunology, Boston, Massachusetts Harvard Medical School Department of Pathology, Boston, Massachusetts

LAURIE GLIMCHER Harvard School of Public Health Department of Immunology and Infectious Disease, Boston, Massachusetts Harvard Medical School Department of Pathology, Boston, Massachusetts

SERIES EDITOR FREDERICK W. ALT Howard Hughes Medical Institute Children’s Hospital, Boston, Massachusetts

VOLUME 83

Elsevier Academic Press 525 B Street, Suite 1900, San Diego, California 92101-4495, USA 84 Theobald’s Road, London WC1X 8RR, UK

This book is printed on acid-free paper. Copyright ß 2004, Elsevier Inc. 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. (www.copyright.com), 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-2004 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. 0065-2776/2004 $35.00 Permissions may be sought directly from Elsevier’s Science & Technology Right Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, E-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://elsevier.com), by selecting ‘‘Customer Support’’ and then ‘‘Obtaining Permissions.’’ For all information on all Academic Press publications visit our Web site at www.books.elsevier.com ISBN: 0-12-022483-6 PRINTED IN THE UNITED STATES OF AMERICA 04 05 06 07 08 9 8 7 6 5 4 3 2

1

CONTENTS

Contributors Introduction

ix xi

Lineage Commitment and Developmental Plasticity in Early Lymphoid Progenitor Subsets David Traver and Koichi Akashi I. Introduction II. Hematopoietic Stem Cells: Clonogenic Precursors of the Hematolymphoid System III. Cell Fate Determination: How do HSCs Commit to Each of the Hematolymphoid Lineages? IV. Overview of T and B Lymphoid Differentiation V. Defining the Earliest Stage in Lymphoid Commitment: Isolation of Common Lymphoid Progenitors (CLPs) VI. Is the CLP Stage Necessary for Adult Lymphoid Differentiation? VII. Common Myeloid Progenitors as the Counterpart of CLPs VIII. Lineage Priming by Promiscuous Gene Expression in Multipotent Stem and Progenitor Cells IX. Lymphoid and Myeloid Promiscuity Demonstrated by Single-Cell Analyses X. Fate Choices Made by Combinations of Instructive Signals at Lineage Promiscuous Stages XI. Lineage Plasticity in Lymphoid Progenitors: Fate Choices are Reversible XII. Comparison of Gene Expression Profiles among Early Hematopoietic Stem and Progenitor Cells XIII. Early Lymphoid Progenitors can Differentiate into Antigen-presenting Dendritic Cells v

1 2 3 6 7 13 17 18 19 22 28 32 33

vi

CONTENTS

XIV. Fetal Hematopoietic Progenitors are Not Fully Committed to the Lymphoid and Myeloid Fates XV. Lymphoid- and Myeloid-restricted Progenitors in Human Bone Marrow XVI. Clinical Relevance XVII. Conclusion References

35 38 39 39 40

The CD4/CD8 Lineage Choice: New Insights into Epigenetic Regulation during T Cell Development Ichiro Taniuchi, Wilfried Ellmeier, and Dan R. Littman I. II. III. IV. V.

Introduction Role of Coreceptors in T Cell Development Regulation of CD4 Gene Expression Regulation of CD8 Gene Expression Conclusion References

55 56 59 77 86 86

CD4/CD8 Coreceptors in Thymocyte Development, Selection, and Lineage Commitment: Analysis of the CD4/CD8 Lineage Decision Alfred Singer and Remy Bosselut I. II. III. IV. V.

Introduction Early Thymocyte Development CD4 and CD8 Coreceptor Molecules on DP Thymocytes Selection and Commitment: Classical Models Kinetic Signaling as an Alternative to Classical Models of Lineage Commitment VI. Conclusions References

91 91 95 99 110 121 121

Development and Function of T Helper 1 Cells Anne O’Garra and Douglas Robinson I. Introduction II. Factors Inducing the Development of Th1 Cells And Their Production of IFN-g III. Role of Th1 Signaling Components in In Vivo Immune Responses

133 135 144

CONTENTS

IV. Transcriptional Regulation for Th1 Cells Producing IFN-g V. New Cytokines: What is Their Role in Th1 Cell Responses? VI. Inherited Disorders of IL-12 and IFN-g-mediated Immunity: Clinical Outcomes of Defects in Th1 Development References

vii 146 149 151 153

Th2 Cells: Orchestrating Barrier Immunity Daniel B. Stetson, David Voehringer, Jane L. Grogan, Min Xu, R. Lee Reinhardt, Stefanie Scheu, Ben L. Kelly, and Richard M. Locksley I. II. III. IV. V. VI. VII. VIII.

Introduction: History and Definitions Activation of IL-4 Expression in Naive CD4 T Cells Stabilization of IL-4 Expression in T Cells Phenotype and Genotype Analysis of Th2 Cells Mutations Impacting IL-4 Expression in T Cells Expression of IL-4 in Non-Th2 Cells Where Does Type 2 Immunity Operate? Concluding Remarks References

163 164 167 170 171 171 177 179 180

Generation, Maintenance, and Function of Memory T Cells Patrick R. Burkett, Rima Koka, Marcia Chien, David L. Boone, and Averil Ma I. II. III. IV. V.

Introduction What Is a Memory Cell? Where Do Memory CD8þ T Cells Come From? How Are Memory Cells Maintained? Conclusions References

191 192 198 212 220 221

CD8þ Effector Cells Pierre A. Henkart and Marta Catalfamo I. II. III. IV.

Effector Cells Defined Secretion as the Mechanism of Effector Function Cytotoxicity In Vivo Cytotoxicity In Vitro

233 235 236 238

viii

CONTENTS

V. VI. VII. VIII. IX.

Perforin-dependent Granule Exocytosis Pathway FasL/Fas Death Pathway Cytokine Secretory Effector Cells CD8þ T Cell Differentiation Conclusions References

238 242 243 245 249 250

An Integrated Model of Immunoregulation Mediated by Regulatory T Cell Subsets Hong Jiang and Leonard Chess I. Introduction and Objectives II. Historical Considerations: Clonal Selection Theory, Immunoregulation, and Regulatory T Cell Subsets III. The T Cell Subsets which Mediate Suppression of the Immune Response IV. An Integrated Model of Immunoregulation by NKT, CD4þCD25þ, and Qa-1 Restricted CD8þ T Cell Subsets References Index Contents of Recent Volumes

253 254 261 277 281 289 301

CONTRIBUTORS

Numbers in parenthesis indicated the pages on which the authors’ contributions begin.

Koichi Akashi (1), Dana-Farber Cancer Institute, Boston, Massachusetts 02115 David L. Boone (191), Department of Medicine and The Ben May Institute for Cancer Research, University of Chicago, Chicago, Illinois 60637 Remy Bosselut (91), Laboratory of Immune Cell Biology, National Cancer Institute, Bethesda, Maryland 20892 Patrick R. Burkett (191), Department of Medicine and The Ben May Institute for Cancer Research, University of Chicago, Chicago, Illinois 60637 Marta Catalfamo (233), National Institutes of Health, Bethesda, Maryland 20892-1360 Leonard Chess (253), Department of Medicine and Pathology, Columbia University College of Physicians and Surgeons, New York, New York 10032 Marcia Chien (191), Department of Medicine and The Ben May Institute for Cancer Research, University of Chicago, Chicago, Illinois 60637 Wilfried Ellmeier (55), University of Vienna, 1235 Vienna, Austria Jane L. Grogan (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 Pierre A. Henkart (233), National Institutes of Health, Bethesda, Maryland 20892-1360 Hong Jiang (253), Department of Medicine and Pathology, Columbia University College of Physicians and Surgeons, New York, New York 10032 Ben L. Kelly (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 Rima Koka (191), Department of Medicine and The Ben May Institute for Cancer Research, University of Chicago, Chicago, Illinois 60637 Dan R. Littman (55), New York University School of Medicine, New York, New York 10016 Richard M. Locksley (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 ix

x

CONTRIBUTORS

Averil Ma (191), Department of Medicine and The Ben May Institute for Cancer Research, University of Chicago, Chicago, Illinois 60637 Anne O’Garra (133), National Institute for Medical Research, London NW7 1AA, United Kingdom R. Lee Reinhardt (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 Douglas Robinson (133), Imperial College London, London SW7 2AZ, United Kingdom Stefanie Scheu (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 Alfred Singer (91), Experimental Immunology Branch, National Cancer Institute, Bethesda, Maryland 20892 Daniel B. Stetson (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 Ichiro Taniuchi (55), Kyushu University, Fukuoka 812-8582, Japan David Traver (1), Dana-Farber Cancer Institute, Boston, Massachusetts 02115 David Voehringer (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143 Min Xu (163), Howard Hughes Medical Institute, University of California San Francisco, San Francisco, California 94143

INTRODUCTION

Until fairly recently, immunologists were most concerned with understanding genetic mechanisms that generated very large numbers of antibodies and T cell receptors from a relatively small set of genes. The discovery of recombination-based genetic mechanisms in T and B cells shifted attention to the obverse problem: how does each cell lineage and sublineage select and use a tiny fraction of the genome to establish its differentiated state? This question has resolved itself into understanding the epigenetic mechanisms that regulate lineage decisions and irreversible lineage commitment. T cell differentiation has, arguably, become the most tractable experimental system for addressing these questions in mammalian cells. The process begins with migration of precursor cells into the thymus, and ends with the formation of two major differentiated subsets. One is equipped to detect and eliminate virally infected cells; a second is programmed to interact with B cells and dendritic cells to induce antibody and/or inﬂammatory reactions. The surface antigens of cells at each differentiative step in this program have been extensively characterized, beginning with hematopoietic stem cells and ending with mature CD8 cells and sublineages of CD4 helper cells. Isolation and characterization of the hematopoietic stem cell (HSC) has been the foundation for analysis of the early events in thymocyte development. Traver and Akashi describe isolation of HSC from bone marrow cells according to reactivity with antibodies to sca-1, c-kit, Thy1, and CD34. To appreciate this remarkable feat, it is only necessary to imagine a baseball game, say between the Red Sox and the NY Yankees at Fenway Park. The 20,000 or so fans represent the differentiated lymphoid, myeloid and erythroid elements within the hematopoietic system. The position players on the ﬁeld represent the committed progenitors for each of these lineages; the pitcher represents the hematopoietic stem cell. Genetic proﬁling of HSC suggests that transcriptional access can predict both their immediate and full differentiative potentials. Traver and Akashi discuss evidence of signiﬁcant plasticity within HSC and progenitor xi

xii

INTRODUCTION

populations, including ﬁndings that some non-hematopoietic genes are transcriptionally accessible and/or active. Although chromatin accessibility and low level transcriptional expression do not normally lead HSC down non-hematopoietic paths, recognition that these precursor cells have alternative genetic potentials is likely to promote a search for inducing agents and receptors that might stimulate HSC to give rise to non-hematopoietic progeny. Mechanisms that underlie CD4/CD8 initial lineage choices and stabilization of these genetic decisions are the subject of reviews by Taniuchi et al. and Singer and Bosselut. Taniuchi et al. delineate the remarkably detailed and complex molecular interactions that regulate CD4 and CD8 expression at discrete stages of T cell development. Expression of the CD8 co-receptor depends mainly on interaction between silencer elements and members of the Runx transcription factor family (and other proteins) to repress CD4 gene expression at early stages of development and to facilitate epigenetic changes that stably quench CD4 gene expression in mature CD8 cells. CD8 gene expression is regulated through a quite different mechanism that depends on tandemly-arranged clusters of cis-acting enhancer elements that sequentially act on CD8 gene expression. Why does regulation of CD4 and CD8 expression depend on many layers of silencing and enhancer elements that are brought to bear at discrete stages in ontogeny? One answer comes from the need for expression of both genes followed by selection-dependent extinction of one of the pair. Additional layers of regulation might be required if additional revisions of co-receptor expression were imposed by positive selection. Singer and Bosselut suggest the possibility that DP thymocytes test TCR/coreceptor compatibility and ﬁx mismatched cells through co-receptor revision. According to this view, TCR signaling of DP cells by MHC (class I or class II) uniformly down regulates CD8 but not CD4 expression. If unabated TCRdependent signaling in this selection intermediate validates the cell’s TCR/ CD8 pairing, the cell continues on to become a full-ﬂedged SP CD4 cell, possibly quenching IL-7R dependent signals along the way. On the other hand, disruption of TCR signaling may direct the cell to reverse its developmental direction, through re-expression of CD8 and repression of CD4 gene expression, to yield a ‘revised’ CD8 cell. Early plasticity followed by irreversible commitment is also a major feature of Th1 and Th2 sublineage development. In the case of Th1 cells, O’Garra and Robinson discuss some of the factors that can impinge upon the STAT1-mediated signaling pathway initiated from the IFN-g receptor that induce expression of the master regulator of Th1 cells, T-bet. Like all good instructors, T-bet-dependent guidance leads to fully independent progeny

INTRODUCTION

xiii

through remodeling of target genes such as IFN-g. This is certainly not the whole story, and new cytokines that enhance the overall expression of IFN-g acting at both early and late stages of Th1 development, mainly through IFN-g, can modify this process. Stetson et al. delineate the central role of GATA-3 as the master regulator of Th2 gene transcription through integration of signals from the T cell synapse and remodeling of chromatin during Th2 development. Additional features of this differentiative path include coordinate regulation of a group of Th2 cytokine genes by inter-genic sequences, designated CNS1 and CNS2 (conserved, non-coding sequences) that facilitate locus-remodeling and epigenetic memory. Terminal cellular commitment that is independent of pioneer transcription factors also involves DNA methylation and heterochromatin patches that stably inhibit non-expressed cytokine loci. This developmental path is placed within the broader biological and clinical context of Th2 immune responses that normally discourage pathogen invasion at epithelial surfaces at the risk of allergic and hypersensitivity reactions. Henkart and Catalfamo and Burkett et al. discuss the generation of memory and cytotoxic effector cells within the CD8 lineage in the context of lineage modeling. Will a single lineage do as CD8 cells pass from effector cells to memory cells? Or do multiple lineages account for memory development, despite the fact that they add complexity to this area. We are also reminded that cytotoxic effector cells are one of the more straightforward elements of the immune response, particularly since much is know about the classic perforin dependent granule exocytosis pathway. Nonetheless, the generation and maintenance of CD8 memory cells has many fascinating biological features. The unexpected interaction in trans between IL-15 and its receptor in the support of CD8 memory, as delineated by Burkett et al., is an excellent case in point. These reviews highlight a series of remarkable new insights into the mechanisms that govern development and function of T cell lineages. As Jiang and Chess point out ‘‘in contrast, the precise biological deﬁnition of immunologic suppressive activity has largely remained an enigma’’. Current approaches to this relatively neglected problem will undoubtedly be sharpened by recent molecular insights into T cell development and potential inhibitory effects of certain cytokines. Jiang and Chess contribute a scholarly review of experimental evidence supporting the view that CD4 and CD8 development may include cells that are genetically programmed to mediate suppressive or effector activity. They cite ﬁndings that the recently-cloned transcription factor Foxp3 is expressed in a subpopulation of regulatory CD4 cells, as a step in this direction. They also summarize the increasingly impressive experimental data that deﬁnes suppressive activity mediated by a sublineage of CD8 cells that

xiv

INTRODUCTION

depend on the MHC class Ib molecule Qa-1. It is ﬁtting that this volume end with an overview of an area of immunology that is just being revived after many years of neglect. Jiang and Chess serve to remind us that the remarkable progress that has been made in understanding epigenetic regulation of T cell development should not be viewed as the last word in this ﬁeld. It may also provide a foundation for future analysis of the development of effector and regulatory sublineages of CD4 and CD8 cells. Harvey Cantor

advances in immunology, vol. 83

Lineage Commitment and Developmental Plasticity in Early Lymphoid Progenitor Subsets DAVID TRAVER AND KOICHI AKASHI Dana-Farber Cancer Institute, Boston Massachusetts 02115

I. Introduction

All blood cell types including lymphocytes are derived from HSCs that both self-renew and maintain multilineage hematopoiesis over the lifetime of the host. Differentiation is deﬁned as the sequence of events through which immature precursors become mature, effector cells. During this stepwise commitment process, it has long been assumed that oligopotent progenitors exist that are daughters of HSCs. Whereas many ﬁndings have retrospectively suggested the existence of such cell types, prospective isolation of clonal progenitors is essential for the precise understanding of both normal and aberrant hematopoiesis. An important issue in the commitment sequence from HSCs to lymphoid cells was whether lymphocytes are directly derived from certain HSC subsets, from bipotent (T-cell/myeloid or B-cell/myeloid) progenitors, or from progenitors that exclusively give rise to all lymphoid cells, including T and B lymphocytes, and natural killer (NK) cells. Since both B and T lymphocytes display similar mechanisms for antigen receptor rearrangement and selection, it was postulated that T and B cells arise from common progenitors that have lost myeloid differentiation potential. Support for the existence of lymphoidrestricted progenitors came from studies using chromosomally marked HSCs (Abramson et al., 1977), from the results of cultured bone marrow cells transplanted into immunodeﬁcient mice (Fulop and Phillips, 1989), and from the loss of both T and B cell subsets in patients with severe combined immunodeﬁciency (SCID) (Fischer, 1992; Hirschhorn, 1990). More recently, a rare population within whole mouse bone marrow was shown to generate all lymphoid subsets but that completely lacked myeloerythroid differentiation potential. These were termed common lymphoid progenitors (CLPs) (Kondo et al., 1997). Subsequently, counterpart myeloerythroid-restricted progenitors (common myeloid progenitors: CMPs) were identiﬁed (Akashi et al., 2000). Continuing studies support the concept that the lymphoid and myeloerythroid pathways diverge downstream of HSCs within the bone marrow of adult mice (Gounari et al., 2002; Igarashi et al., 2002). Using these prospectively isolated populations, lineage commitment can now be investigated directly at each of the major hematopoietic branchpoints (Fig. 1). 1 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

2

DAVID TRAVER AND KOICHI AKASHI

Fig 1 Hematopoietic commitment model based on prospective isolation of lineage-restricted progenitors. HSC, hematopoietic stem cells, CMP, common myeloid progenitors, CLP, common lymphoid progenitors; GMP, granulocyte/monocyte progenitors; MEP, megakaryocyte/erythrocyte progenitors. (See Color Insert.)

II. Hematopoietic Stem Cells: Clonogenic Precursors of the Hematolymphoid System

The search for hematopoietic stem cells began with the observation by Till and McCulloch that bone marrow transplants into lethally irradiated mice generated clonal spleen colonies, and that some of these colonies could generate multilineage hematopoiesis in serially transplanted animals (Becker, 1963; Till and McCulloch, 1961; Wu et al., 1968). Many subsequent retrospective transplantation experiments suggested that rare populations of HSCs existed within total bone marrow (Mulder and Visser, 1987; Visser et al., 1984). The modern era of HSC biology began with the ﬁrst rigorous, prospective isolation of murine HSCs by cell surface phenotype (Spangrude et al., 1988; Uchida and Weissman, 1992; Uchida et al., 1994). These investigators showed that long-term, multilineage reconstitution activity was present only within a population of Lin/loThy-1.1loSca-1þ bone marrow cells. A subset of these cells displayed long-term self-renewing potential (Spangrude et al., 1991) that, at the single-cell level, could give rise to both myeloid and lymphoid outcomes (Smith et al., 1991). Subsequent studies, however, showed that the Lin/loThy-1.1loSca-1þ bone marrow fraction was heterogeneous in terms of self-renewal activity; HSCs with long-term self-renewal activity could be further subfractionated by isolating cells expressing c-Kit, a receptor for Steel

LINEAGE COMMITMENT

3

factor (Slf) (Morrison and Weissman, 1994). Another group showed that within the LinSca-1þc-Kitþ population, only CD34/lo cells are long-term HSCs, whereas the remaining CD34þ cell can display only short-term, multilineage reconstitution (Osawa et al., 1996). Both Thy-1.1loLinSca-1þc-Kitþ and CD34/loLinSca-1þc-Kitþ populations constitute 0.01% of total bone marrow cells, and we have conﬁrmed that >60% of these populations phenotypically overlap. Multilineage, long-term reconstitution from single cells was observed in 20% and 35% of Thy-1.1loLinSca-1þc-Kitþ and CD34/loLinSca-1þc-Kitþ populations, respectively, after competitive reconstitution assays (Osawa et al., 1996; Wagers et al., 2002). Another marker for HSCs, as well as for stem cell subsets in other tissues (Goodell et al., 1997, 2001; Jackson et al., 2001; Storms et al., 2000) is the differential efﬂux of the intracellular dye Hoechst 33342. Hoechstlow cells, termed the side population (SP), almost exclusively contain the long-term HSC subset (Goodell et al., 1997). Puriﬁed SP cells were LinSca-1þc-Kitþ, and contained >30% of CD34/lo cells (Okuno et al., 2002). In a study in 2001, the Thy1.1loLinSca-1þc-Kitþ population was further divided by expression of the fms-like tyrosine kinase-3 (Flt-3 or Flk-2). (Adolfsson et al., 2001; Christensen and Weissman, 2001). Around 60% of Thy-1.1loLinSca-1þc-Kitþ cells express Flk-2, and long-term HSC activity was only found within Flk-2Thy-1.1 lo LinSca-1þc-Kitþ cells, enabling the further enrichment of long-term HSCs. Considering the noted inefﬁciency of intravenously injected cells to seed the bone marrow (Morrison et al., 1996), isolation of long-term HSCs by the above phenotypes has likely reached purity. III. Cell Fate Determination: How do HSCs Commit to Each of the Hematolymphoid Lineages?

A fundamental question is how multipotent cells select one cell fate from a choice of several. Lineage commitment and subsequent differentiation of multipotent cells likely occurs due to the selective activation and silencing of particular gene expression programs. This process likely involves the formation of transcriptional complexes at the regulatory regions of lineage-speciﬁc gene loci (Wadman et al., 1997). Changes in chromatin structure permitting or denying access to transcriptional machinery is probably critical for this process (Berger and Felsenfeld, 2001; Felsenfeld et al., 1996). The development of in vitro clonogenic assays has deﬁned retrospective subsets of myeloerythroid progenitors that appear to have restricted differentiation capacity (Bradley and Metcalf, 1966; Pluznik and Sachs, 1965). Unfractionated bone marrow was found to contain oligopotent colony-forming units (CFU) for all myeloid lineage cells (CFU-GEMMeg or CFU-Mix) (Johnson and Metcalf, 1977; Metcalf et al., 1979), for granulocytes and

4

DAVID TRAVER AND KOICHI AKASHI

macrophages (CFU-GM), and for megakaryocytes and erythrocytes (CFUMegE). Monopotent CFU for granulocytes (CFU-G), macrophages (CFUM), erythrocytes (CFU-E), or megakaryocytes (CFU-Meg) were also found. The combinations of cell types present within these colonies supported the notion of clonal progenitor subsets with progressive loss of lineage potentials, and suggested close relationships between the GM and MegE branches of the hematopoietic tree. If multipotent HSCs indeed give rise to progenitors with progressive lineage restriction in vitro, it is logical to place these progenitors in a hierarchical order in the hematopoietic lineage map (Dexter and Testa, 1980) (Fig. 2A), and the myeloerythroid differentiation pathway might initiate from clonogenic progenitors of all mature cell types, such as CFU-GEMMeg. More precise lineal studies using in vitro blast colony formation, however, did not necessarily support this hierarchical model. Cells harvested from multipotent blast colonies (Nakahata and Ogawa, 1982) were able to efﬁciently form secondary colonies, and paired daughter cells derived from single blast cells frequently gave rise to different combinations of myeloid progeny when split into identical conditions (Suda et al., 1984a,b). Based on this phenomenon, ‘‘stochastic’’ commitment of HSCs to the myeloerythroid fates has been proposed (Ogawa, 1993) (Fig. 2B). This model states that the commitment decision of multipotent cells is essentially random and cell autonomous, and that differentiation is subsequently determined by the availability of survival or growth signals. Another model of hematopoietic commitment was proposed largely based on the analysis of in vitro differentiation potentials of immortalized cell lines. The ‘‘sequential determination model’’ proposed by Brown and colleagues states that there is a predetermined order of developmental choices (Brown et al., 1985, 1987) (Fig. 2C). This model proposes that HSCs undergo an intrinsic decision program to generate cells that can differentiate along one discrete pathway. While all of these commitment models accommodate intermediate progenitor cells, all have the critical caveat that retrospective in vitro lineage outcomes may not reﬂect the full commitment potential of each assayed cell type. First, there exists no single in vitro assay system that is permissive for each blood cell lineage. Whereas retrospective identiﬁcation of a CFU-GEMM colony is often attributed to an early myeloerythroid progenitor, this colony could have equally resulted from a plated HSC that could not produce lymphocytes due to culture limitations. Second, if in vitro colony conditions are fully permissive for all myeloerythroid fates, it may be logically assumed that HSCs should always give rise to CFU-GEMMeg. It is known, however, that single Thy1.1loLinSca-1þc-Kitþ HSCs give rise to many different colony types, including burst-forming unit-erythroid (BFU-E), CFU-E, CFU-GM, CFU-Meg, and multilineage CFU-GEMMeg colonies at high frequencies (Akashi et al., 2000; Heimfeld et al., 1991; Morrison et al., 1996). It is thus still unclear

LINEAGE COMMITMENT

5

Fig 2 Historical hematopoietic commitment models. (A) The hierarchical commitment model based on cell component contained in clonogenic myeloid colonies. (B) The stochastic commitment model based on the analysis of colonies from paired daughter progenitors. (C) The sequential determination model based on the analysis of cell line phenotypes.

whether highly puriﬁed HSCs do indeed commit randomly to the various myeloerythroid fates at least in vitro, or whether these phenomena simply represent the unstable nature of in vitro assay systems. These models have been established mainly on the in vitro behavior of stem and progenitor cells to read-out myeloerythroid, but not lymphoid, differentiation. This is largely because lymphoid assay systems are inefﬁcient compared to myeloerythroid assays, as discussed in the following section. Nonetheless, in each model, there should be a variety of lineage-restricted progenitors in early hematopoiesis that are intermediates between HSCs and mature blood cells. An important approach to understanding hematopoietic commitment was to purify lineage-restricted progenitor subsets and to test their lineal relationships.

6

DAVID TRAVER AND KOICHI AKASHI

IV. Overview of T and B Lymphoid Differentiation

Lymphoid development requires interactions between lymphoid precursor cells and stromal cells; this relationship is responsible for the expansion and selection of immature lymphocytes. The requirement of complex conditions for lymphoid development has thus hindered the precise analysis of lymphoid development and commitment processes. For T lymphocytes, the thymic stromal microenvironment is critical in the selection of developing thymocytes and in the elimination of cells expressing inappropriate T cell receptor (TCR) genes. The thymic microenvironment consists of thymic epithelial cells, macrophages, B cells, and dendritic cells. Lymphoid progenitors from the bone marrow need to seed the thymus for T cell production. It is unclear what cell population homes to thymus in normal T lymphopoiesis, which will be discussed later in this chapter. Within the thymus, the earliest thymic precursor has been identiﬁed within the fraction of CD4loCD8CD3 (triple negative; TN) cells. The earliest thymic progenitors (or proT1 cells) are CD25 þCD44þc-Kitþ, and as a population, they are capable of NK, B, and dendritic cell differentiation, but have lost most myeloid potential (Matsuzaki et al., 1993; Wu et al., 1991a,b). This supports the hypothesis that lymphoid commitment may occur in the bone marrow before cells home to the thymus. In the next proT (proT2 cells) stage, CD25þCD44þc-Kitþ proT cells begin to rearrange TCR genes. The molecules that initiate the recombination of VDJ for TCRb or of VJ for TCRg genes remain largely unclear. This rearrangement requires the two tightly regulated lymphoid-speciﬁc proteins’ recombination activation gene (RAG)-1 and RAG-2, which form a complex resulting in cleavage of DNA (McBlane et al., 1995; van Gent et al., 1996). Mice that lack RAG-1 or RAG-2 genes have a complete block of early T (and early B) cell development (Mombaerts et al., 1992; Shinkai et al., 1992). Ectopic expression of transfected RAG-1 and RAG-2 genes in nonlymphoid tissue does not result in antigen receptor rearrangement (Schatz et al., 1992), suggesting that the RAG genes do not directly instruct lymphoid commitment. Developing preT cells (CD25þCD44c-Kit) are ﬁrst selected based on the appropriateness of rearranged TCRb chains coupled with the invariant preT cell receptor a chain (pTa). Cells that successfully pass through this b-selection stage are further selected by low- and high-afﬁnity major histocompatibility (MHC) molecule interactions that positively select self-restrictive T cells and negatively select autoreactive T cells, respectively. T cell development is thus critically regulated by these two checkpoints that exist speciﬁcally in the thymus. Importantly, to estimate the T cell potential of candidate lymphoid progenitor populations, one should ensure that the assayed cell subset reaches the thymic microenvironment. For example, by intravenous transplantation, high numbers of T cell progenitors are required to obtain T cell reconstitution, whereas injection of

LINEAGE COMMITMENT

7

progenitors directly into the thymus signiﬁcantly increases T cell differentiation efﬁciency (Kondo et al., 1997). This requirement limits the detection of T cell potential by intravenous injection alone, especially when assayed cell types possess limited expansion potential. B cell development also involves selective outgrowth of randomly generated clonal cell populations. Proper rearrangement of B cell antigen receptor (BCR) genes is required for B lymphocyte development. Despite the lack of a requirement for BCR interaction with MHC molecules, B cells need to pass through a number of developmental checkpoints; failure results in apoptotic death. Progressive commitment to the B cell lineage occurs through a phenotypically deﬁned differentiation pathway based on the surface markers AA4.1 and CD43 (Hardy et al., 1991; Li et al., 1996). An early AA4.1þCD43þ preproB cell subset can be further subdivided into A0, A1, and A2 fractions (Allman et al., 1999) based on the additional B220 and CD4 markers. The CD4loB220 A0 fraction expresses heterogeneous levels of Sca-1, c-Kit, and Mac-1, and contains cells that can give rise to myeloid and T cell as well as B cell progeny. This population may contain Thy-1loLinSca-1þCD4lo multipotent progenitors (Morrison and Weissman, 1994). The CD4hiB220þ A1 fraction expresses low levels of Sca-1 and Mac-1, but does not express c-Kit. The A1 fraction lacks myeloid differentiation activity, and possesses minimal T cell differentiation potential as assessed by intrathymic injection. It is unknown whether the A1 fraction contains B and T biopotent precursors. The majority of CD4B220þ A2 cells are B cell-committed. Pro-B cells rearrange DH-JH genes and express B220, CD43, and HSA on their surface (Fraction B), and subsequently express 6C3/BP-1 and undergo V-DJ recombination (Fraction C). Acquisition of CD19 represents an important marker for B cell commitment and corresponds to the ability of late proB cells to proliferate in IL-7 without other stromal cell-derived factors (Hardy et al., 1991; Hayashi et al., 1990). These CD19þ precursors should be categorized in fraction B in Hardy’s classiﬁcation (Tudor et al., 2000). ProB cells in fractions B and C express the surrogate light chain genes, l5 and VpreB, to form pre-BCR. Following rearrangement of the immunoglobulin light chain gene, the heavy and light chains form a complex and are expressed on the surface together with Iga and Igb to become mature B cells. V. Defining the Earliest Stage in Lymphoid Commitment: Isolation of Common Lymphoid Progenitors (CLPs)

To prospectively isolate early lymphoid-commited progenitors, one could either employ the random, biochemical approach that led to HSC isolation or a candidate approach based on lymphoid-speciﬁc cell surface markers. Both Thy-1 and Sca-1, critical markers for HSC fractionation, are also expressed

8

DAVID TRAVER AND KOICHI AKASHI

on the majority of mature T cells. Based on targeted gene disruptions, Thy-1 does not play a critical role in myeloid or lymphoid development (NostenBertrand et al., 1996), and Sca-1 functions in HSC self-renewal and granulocyte development but is not required for lymphoid development (Ito et al., 2003). As has been discussed, much is known regarding the molecular determinants of both T and B lymphocytes. Markers speciﬁc for lymphoid functions could therefore be systematically analyzed to phenotypically delineate early steps in lymphoid differentiation. The ﬁrst such key molecule used to enrich for CLPs was the receptor for interleukin 7 (IL-7), an essential cytokine for both T and B cell development (Peschon et al., 1994; von Freeden-Jeffry et al., 1995) (Fig. 3). As will be discussed, mice carrying reporter genes for other functional lymphoid markers, such as the lymphoid-speciﬁc RAG-1 gene (Mombaerts et al., 1992) and the early T lymphoid preT cell receptor a chain

Fig 3 Identiﬁcation of IL-7Raþ CLPs and myeloid progenitors in mouse bone marrow. (a) Cells negative for lineage (Lin) markers including B220, CD4, CD8, CD3, Gr-1, Mac-1, and TER119 were subdivided into IL-7Ra positive and negative fractions. (b) Sca-1/c-Kit proﬁles of LinIL-7Raþ cells. The LinIL-7RaþSca-1loc-Kitlo cells are CLPs. (c) Sca-1/c-Kit proﬁles of LinIL-7Ra cells. The LinIL-7RaSca-1hic-Kithi subset represents the HSC population, whereas the LinIL-7RaSca-1c-Kithi subset contains all myeloid progenitor populations. (d) The LinIL-7RaSca-1c-Kithi population is subdivided into 3 distinct myeloid progenitor subsets such as CMPs, GMPs, and MEPs, according to their FcgRII/III and CD34 proﬁles.

LINEAGE COMMITMENT

9

Fig 4 Phenotypic differences in primitive lymphoid progenitor subsets possessing T and B lymphoid potential. The early lymphocyte progenitor (ELP) population is deﬁned in the bone marrow as a population that initiates RAG-1 transcription (Igarashi et al., 2002). This population corresponds to a few percent of cells within the conventional short-term HSC subset (LinSca1þc-Kitþ). Although ELPs have ‘‘lymphoid’’ RAG-1 mRNA, they still possess a weak GM potential. ELPs do not express IL-7Ra. Common lymphoid progenitors (CLPs) are cells that begin to express IL-7Ra (Kondo et al., 1997). CLPs are all transcribing the RAG-1 gene, and a small fraction of them also initiate pTa transcription, which is upregulated in thymic proT cells (Miyamoto et al., 2002). In the B lymphoid pathway in the bone marrow, PTa transcription is maintained in a minority of B220þ (c-Kit) cells, which maintain T/B bipotentiality (Gounari et al., 2002). During this sequential lymphoid development in the bone marrow, both Sca-1 and c-Kit are gradually downregulated during the transition of ELPs, CLPs, to B220þ cells. In contrast, the earliest thymic progenitors, called early T lineage progenitors (ETPs) express Sca-1 and c-Kit at a high level, and still possess weak B and GM potentials (Allman et al., 2003). It is unclear whether ETPs originate from ELPs, CLPs, or B220þpTaþ cells. Although the high expression pattern of Sca-1 and c-Kit in ELPs and ETPs may represent their direct precursor/progeny relationship, direct evidence is lacking (see text).

(pTa) (Groettrup et al., 1993), have also been used to search for the earliest lymphoid progenitors that can give rise to all lymphoid components. Lineal relationships and phenotypic overlaps among these lymphoid progenitors deﬁned by each criterion are illustrated in Fig. 4. A. The Role of Flk-2 in Early Stem and Progenitor Cells The Flk-2/Flt-3 receptor, like c-Kit, is a cytokine tyrosine kinase receptor expressed primarily on early hematopoietic precursors (Matthews et al., 1991). Flk-2 was originally isolated based on its expression in HSCs (Mackarehtschian et al., 1995). Mice deﬁcient in Flk-2 or Flt-3 ligand (FL) display loss of early B, NK, and dendritic cell development, and reduced numbers of T cell progenitors (Mackarehtschian et al., 1995; McKenna et al., 2000; Sitnicka et al., 2002). These phenotypes could result due to a requirement for FL signals in common precursors for each lymphoid lineage. As has been discussed, Adolfsson et al. showed that Flk-2 is not expressed in long-term reconstituting HSCs, but is upregulated in non-self-renewing cells within the LinSca-1þc-Kitþ fraction (Adolfsson et al., 2001). Importantly, these authors presented clonal data

10

DAVID TRAVER AND KOICHI AKASHI

indicating that Flk-2þSca-1þc-Kitþ cells contain cells with T, B, and myeloid (GM) potentials, supporting the existence of multipotent progenitors, although MegE potentials were not tested. Christensen et al. further subfractionated the Flk-2þSca-1þc-Kitþ population by using Thy1.1 expression (Christensen and Weissman, 2001). They showed that the Thy1.1loSca-1þc-Kitþ population can be divided into Flk-2 long-term HSC and Flk-2þ short-term HSC populations, and all Thy1.1Sca-1þc-Kitþ multipotent progenitors are Flk-2þ. In an independent study, the Thy1.1Sca-1þc-Kitþ population displayed rapid B cell reconstitution after intravenous transplantation with minimal contributions to the T and myeloid lineages (Searles et al., 2000). Compatible with these data, Flk-2 is expressed in IL-7Raþ CLPs (Kondo et al., 1997), but not in common myeloid progenitors (CMPs) (Akashi et al., 2000) (see following text). Furthermore, FL-deﬁcient mice lack CLPs, but possess normal numbers of CMPs (Adolfsson et al., 2001). These data strongly suggest that Flt-2/FL interactions might play an important role in early lymphoid development. B. CLPs are Defined as the Most Primitive IL-7Ra-Expressing Cells A more deﬁnitive marker for lymphoid commitment might be the expression of the receptor for IL-7, an essential cytokine for both T and B cell development. The IL-7 receptor (IL-7R) is composed of the IL-7Ra chain and the common cytokine receptor g chain (g c) (Kondo et al., 1994; Noguchi et al., 1993). Mice genetically deﬁcient for IL-7 or the IL-7 receptor (IL-7R) lack both T and B cells (Peschon et al., 1994; von Freeden-Jeffry et al., 1995), whereas mice deﬁcient for gc lack NK cells as well as T and B lymphocytes (Cao et al., 1995; Ohbo et al., 1996). NK cell deﬁciency in the latter animals is due to the lack of the IL-15 receptor, which is composed of both the IL-15Ra and gc (Giri et al., 1994). The IL-7R is expressed in early pro-T and pro-B cells (Akashi et al., 1998). The critical role of IL-7 in abT cell differentiation is to maintain survival of developing T cells through the upregulation of a survivalpromoting protein, Bcl-2 (Akashi et al., 1997). Signaling through the IL-7R is necessary for the rearrangement of immunoglobulin heavy chain V segments through the activation of the pax-5 gene (Corcoran et al., 1998), and for V-J recombination of gd TCR genes through the activation of STAT5 (Maki et al., 1996). Administration of neutralizing anti-IL-7 antibodies resulted in severe inhibition of T and B lymphopoiesis, but not myelopoiesis (Bhatia et al., 1995). Taken together, these data suggested that IL-7Ra expression is one of the most reliable functional markers for lymphoid commitment. The isolation of CLPs was performed according to the following steps. First, IL-7Ra-expressing cells were devoid of myeloerythroid potential in vitro and in vivo. Second, within IL-7Raþ bone marrow cells, robust

LINEAGE COMMITMENT

11

differentiation potential to the T, B, and NK cell lineages existed within the LinSca-1loc-Kitlo population in vivo. In mice transplanted with CLPs, donorderived T and B cells began to decline after 4 to 6 weeks, indicating that this population has no signiﬁcant self-renewal activity. Third, by a two-step clonogenic assay, approximately 20% of single IL-7RaþLinSca-1loc-Kitlo cells could give rise to both T and B cells. Thus, IL-7RaþLinSca-1loc-Kitlo cells contain clonogenic T and B cell progenitors and completely lack myeloerythroid differentiation potential. In a subsequent study, more than 40% of single IL-7RaþLinSca-1loc-Kitlo cells could differentiate to NK cells. The vast majority of CLPs express Flk-2 (Sitnicka et al., 2002). The clonogenic T/B bipotentiality of IL-7Ra-expressing CLPs has also been demonstrated by another group (Izon et al., 2001), using a modiﬁed fetal thymic organ culture (FTOC) system (Kawamoto et al., 1997). C. Initiation of Lymphoid Commitment May Occur Prior to the CLP Stage: Early Lymphoid Progenitors Defined by RAG-1 Transcriptional Activation The RAG genes are indispensable for the rearrangement of TCR and Ig genes, and RAG-1 or RAG-2 deﬁcient mice display a complete loss of T and B cells (Mombaerts et al., 1992; Shinkai et al., 1992). Mice carrying GFP knocked into the RAG-1 locus were generated (Igarashi et al., 2001; Kuwata et al., 1999) and similarly assayed for early lymphoid progenitors initiating transcription of the RAG-1 gene (Igarashi et al., 2002). Over 90% of IL-7RaþLinSca-1lo c-Kitlo CLPs expressed GFP in these animals (Igarashi et al., 2002). Interestingly, RAG-1 expression was also observed in 5% of the LinSca1þc-Kitþ HSC fraction. Like CLPs (Sitnicka et al., 2002), GFPþLinSca-1þcKitþ cells also expressed Flk-2, indicating that they are not long-term HSCs (Adolfsson et al., 2001; Christensen and Weissman, 2001) but likely are committed progenitors (Igarashi et al., 2002) and included within the (Flt-2þ)Thy1.1LinSca-1þc-Kitþ population (Christensen and Weissman, 2001; Searles et al., 2000). A vast majority of LinSca-1þc-Kitþ cells and GFPþLinSca-1þcKitþ cells express CD27 (Wiesmann et al., 2000), a member of the TNF receptor family previously shown to play a role in lymphoid proliferation, differentiation, and apoptosis. RAG-1/GFPþLinSca-1þc-KitþCD27þ cells exhibited potent T, B, and NK differentiation potential after transplantation into congenic hosts. The timing of thymic T cell reconstitution of RAG-1/ GFPþLinSca-1þc-KitþCD27þ cells preceded that of CLPs by 7 days, suggesting that these cells were more immature than IL-7RaþLinSca-1locKitlo CLPs. A fraction of RAG-1/GFPþLinSca-1þc-KitþCD27þ cells, however, formed CFU-GM in vitro, indicating that this population is not entirely committed to the lymphoid fates. Furthermore, this study did not demonstrate that T and B cell progeny can originate from single

12

DAVID TRAVER AND KOICHI AKASHI

RAG-1/GFPþLinSca-1þc-KitþCD27þ cells, and therefore the homogeneity of this population is still in question. Nonetheless, these data indicate that the majority of RAG-1/GFPþLinSca-1þc-KitþCD27þ cells are committed to the lymphoid lineages, and that they likely contain cells similar to, or slightly upstream of CLPs. The authors termed this population ‘‘early lymphocyte progenitors (ELPs)’’ (Igarashi et al., 2002). D. Use of pTa Reporter Constructs to Define Early Lymphoid Progenitors The pTa protein pairs with the T cell receptor (TCR) b chain to form the pre-TCR (Groettrup et al., 1993) that plays a critical role in the efﬁcient generation of mature T cells (Fehling et al., 1995). Since pTa is used for preTCR formation after TCRb gene rearrangement, pTa transcription should be a relatively late marker for lymphoid development as compared to RAG-1 transcription. A transgene containing a 9 kb fragment spanning all characterized pTa promoter and enhancer sequences was used to drive expression of a hCD25 minigene (Miyamoto et al., 2002; Reizis and Leder, 2001). In pTa/hCD25 transgenic mice, the expression of hCD25 was highly correlated with that of the endogenous pTa, as quantiﬁed by real-time RT-PCR analyses (Gounari et al., 2002). However, a small fraction of B220þ and CD4þ cells in the bone marrow were hCD25þ, but did not express pTa mRNA, presumably because hCD25 transcripts persist for a while after pTa downregulation. In the thymus, the expression of pTa reported by hCD25 was detected in 30% of mCD25c-KitþCD3CD4lo cells, termed the earliest thymic precursors (Wu et al., 1991b), and in the majority of mCD25þc-Kitþ proT cells, consistent with the normal expression of pTa. In the bone marrow, only 6% of Lin cells were hCD25þ, all of which displayed the IL-7RaþSca-1lo-c-Kitlo CLP phenotype, whereas only 7% of CLPs were positive for hCD25 expression (Miyamoto et al., 2002). This result indicates that CLPs are the earliest population to initiate pTa transcription, but that the majority of CLPs have not initiated pTa transcription. The transcription of pTa does not mark T cell commitment at the CLP stage, since pTa/hCD25þ CLPs differentiated into B cells at the same efﬁciency as pTa/hCD25þ CLPs differentiated into B cells at the same efﬁciency as pTa/hCD25þ CLPs. In two-step clonogenic assays similar to those used in the original CLP report (Kondo et al., 1997), single pTa/hCD25þ CLPs were demonstrated to give rise to both T and B cells (Gounari et al., 2002). This study also showed that CLP activity persists in a small cell fraction of cells phenotypically downstream from CLPs: pTa/ hCD25þCD19B220þc-Kit cells express IL-7Ra, and possess some T and B cell potential, suggesting that CLP potential is maintained in at least a fraction of B220þ cells that have downregulated c-Kit.

LINEAGE COMMITMENT

13

E. Which Markers are Most Useful to Separate the Earliest Lymphoid Progenitors? The phenotype and biological activities of each progenitor subset are summarized in Fig. 4 and Table I. From the already cited studies, it is highly likely that lymphoid commitment occurs in adult mouse bone marrow, and that CLPs are an important intermediate in steady-state hematopoiesis. Although the majority of CLP activity resides in cells within the originally described IL-7RaþLinSca-1loc-Kitlo fraction, similar activity can also be found within a population of LinSca-1þc-Kitþ HSCs (as evidenced by RAG-1/GFP mice) and within B220þCD19c-Kit cells (as evidenced by hCD25-pTa mice). The differences in the distribution of CLP activity described in these studies demonstrates the difﬁculty in ﬁnding markers that precisely correlate with cell functions, although IL-7Ra-expression is still the only available marker for CLPs in normal mice. In separating cells according to positive or negative marker expression, the sensitivity of each marker is critical. The threshold for positive detection will be decided by the sensitivity of the detectors and by the amount and/or longevity of antigens or reporters. For example, although IL7Raþ cells are undetectable in the LinSca-1þc-Kitþ population by FACS, IL-7Raþ cells are detectable by highly sensitive nested RT-PCR analyses. However, such low level of receptor expression may not be sufﬁcient to transmit functional IL-7 signals. In addition, as will be discussed, the expression of lineage-related genes at low levels precedes lineage commitment in multi- or oligopotent cells. Therefore, if the marker is too sensitive, this may result in the isolation of uncommitted cells. In this context, it is likely that the CFU-GM potential detected within the RAG-1/GFPþ LinSca-1þc-Kitþ fraction is a reﬂection of multipotent cells with low level ‘‘priming’’ of lymphoid genes (see following text). Conversely, when a late lymphoid marker is used, such as pTa, a failure to efﬁciently isolate CLPs may result, since <10% of CLPs express pTa transcripts. In this context, the IL-7R is currently the most reliable and efﬁcient marker, since sorting based on IL-7Ra expression yields lymphoid-commited cells with the complete absence of myeloerythroid potential.

VI. Is the CLP Stage Necessary for Adult Lymphoid Differentiation?

As has been discussed, the phenotypic deﬁnition of IL-7RaþLinSca-1locKitlo appears to include the majority of cells having CLP activity in mouse bone marrow. Must all lymphocytes necessarily pass through a CLP stage during their maturation? If this is not the case, then progenitors bipotent for myeloerythroid/T or myeloerythroid/B lineages might exist. Despite an intensive search, such progenitor populations have not yet been isolated.

TABLE I Characteristics of Stem and Primitive Lymphoid Progenitors

Location

Sca-1/c-Kit/Thy1

Fit-3

RAG1 activation

IL-7Ra

pTa activation

Lineage potential as a population

HSC

BM

Sca-1þc-KitþThy1lo

All lineage

ST-HSC

BM

Sca-1þc-KitþThy1

þ

All lineages

ELP CLP

BM BM

Sca-1þc-KitþThy1 Sca-1loc-KitloThy1

þþ þþ

þþ þþ

þþ

þ

T, B, NK GM* T, B, NK

PTNþB220þ

BM

Sca-1c-KitThy1

ND

ND

þþ

þþ

T, B (NK?)

ETP

Thymus

Sca-1þc-Kitþ(Thy1lo)

?

ND

/NN

ND

T, B, NK GM*, Erythroid**

Clonal T/B potential Yes (Osawa et al., 1996) Yes (Adolfsson et al., 2001) ND Yes (Kondo et al., 1997) Yes (Gounari et al., 2002) ND

HSC: hematopoietic stem cells; ST-HSC: short-term HSC; ELP: early lymphocyte progenitors (Igarashi et al., 2002); CLP: common lymphoid progenitors (Kondo et al., 1997); ETP: early T lineage progenitor (Allman et al., 2003). *: the GM potential in ELPs and ETPs was found in 3 and 1%, respectively, by methylcellulose assays. **: In a thymic population corresponding to ETPs, a few percent of cells displayed myelo-erythroid potential (Matsuzaki et al., 1993).

LINEAGE COMMITMENT

15

A population containing B/macrophage bipotent progenitors has recently been reported (Tudor et al., 2000). A B220þCD19þCD43þ(IL-7RaþSca-1) population, consisting of 0.02% of total adult bone marrow cells, contained progenitors capable of differentiation into mature B cells and macrophages. This population could not differentiate into T or NK cells by FTOC, however, nor were myeloerythroid colonies generated in the presence of a wide variety of cytokines. In liquid cultures, these cells gave rise only to macrophages. Under B cell conditions, 1% of these cells could form B cell colonies in vitro. Approximately 3% of single cell-derived colonies contained both B cells and macrophages. These data collectively indicate that a fraction of this population is bipotent for B cells and macrophages (Tudor et al., 2000). The in vivo reconstitution activity of this population was not tested, however. The frequency of combined B cell and macrophage differentiation was very low, and the expansion potential of macrophages from this population is minimal, since differentiated macrophages did not proliferate well to form colonies. The contribution of this population to macrophage development and B lymphopoiesis thus appears to be very limited. It is interesting that this population gave rise to macrophages, but not to monocytes, dendritic cells or other myeloerythroid cells. It is possible that this macrophage readout occurs only under speciﬁc in vitro conditions, representing plasticity of B cell precursors to reactivate macrophage differentiation programs, as will be discussed later in this chapter. Nonetheless, this ﬁnding suggests that an intimate developmental relationship exists between the B cell and macrophage lineages. Similar ﬁndings using prospectively isolated fetal progenitors will be discussed. Another unique population has been found within the CD3CD4loCD8 CD25CD44þ early thymic progenitor subset originally reported by Shortman et al. (Wu et al., 1991a). It has been reported that the majority of CD3CD4loCD8CD25CD44þ early thymic progenitors are c-Kitþ, and these c-Kitþ early thymic progenitors can differentiate into T, B, and NK cells (Matsuzaki et al., 1993). Minor myeloerythroid differentiation activity was also observed in this population (Matsuzaki et al., 1993). Allman et al. divided CD3CD4loCD8CD25CD44þ cells into IL-7Ra/loc-Kitþ and IL-7Raþc-Kit/lo fractions; only the former possessed T cell potential by intrathymic adoptive transfer experiments. The former should thus be identical to the c-Kitþ thymic progenitors (Matsuzaki et al., 1993). The reconstitution potential of T and B cells was compared between bone marrow CLPs and thymic IL-7Ra/loc-Kitþ cells: B cell progeny from thymic IL-7Ra/locKitþ cells was 100-fold less than that from CLPs, and the kinetics of donor B cell appearance was signiﬁcantly delayed. Thymic IL-7Ra/loc-Kitþ cells, termed early T lineage progenitors (ETPs), exhibited sustained production of T cells for 4 weeks in the thymus, whereas T cell production from bone marrow CLPs began to decline 2 weeks after intrathymic transplantation, suggesting

16

DAVID TRAVER AND KOICHI AKASHI

that ETPs are more potent for T cell differentiation than CLPs. This data, however, does not ﬁt the biological potential of CLPs in the original report (Kondo et al., 1997), in which T cell differentiation peaked at 5 weeks after intravenous transplantation of CLPs. Ig-DH-JH rearrangement was found in both CLPs and IL-7Ra/loc-Kitþ cells, whereas TCR-Db1-Jb1 rearrangement was only found from IL-7Ra/loc-Kitþ cells; importantly, B cells derived from this population also rearranged TCR-Db1-Jb1. This result suggests a common origin of T and B cell progeny from thymic IL-7Ra/loc-Kitþ cells, although clonal studies were not performed. Thymic ETPs occasionally formed GM colonies, suggesting that this population is somewhat heterogeneous. How then can IL-7Ra/loc-Kitþ ETPs be integrated into a developmental scheme including CLPs (Kondo et al., 1997) and RAG-1/GFPþSca-1þc-Kitþ ELPs (Igarashi et al., 2002)? Because the IL-7Ra/loc-Kitþ phenotype overlaps with RAG-1/GFPþSca-1þc-Kitþ ELPs that do not express IL-7Ra, and because both populations retain minimal myeloid colony-forming activity, thymic ETPs may be directly derived from RAG-1/GFPþSca-1þc-Kitþ cells and not necessarily via bone marrow CLPs (Allman et al., 2003). However, since the myeloid potential of ETPs (Allman et al., 2003) or ELPs (Igarashi et al., 2002) was observed in only a fraction of each population and was not proven to be associated with their T and B cell potential, they are still heterogeneous and could contain myeloid progenitor contaminants due to their phenotypic overlap. Allman et al. also showed that Ikaros/ mice that display a speciﬁc and early arrest in B cell development (Wang et al., 1996) possess IL-7Ra/loc-Kitþ cells but not CLPs (Allman et al., 2003). These data were interpreted to suggest that CLPs are progenitors mainly for B cells, whereas ETPs may directly originate from HSCs or multipotent progenitors, and may be responsible mainly for T cell development (Allman et al., 2003). However, it is difﬁcult to deﬁne precursor/progeny relationships based on targeted gene disruptions, since each gene could play different roles in different lineages (Weissman, 1994). Ikaros/ mice display impaired longterm HSC activity and decreases in the number of Flt-3þSca-1þc-Kitþ progenitor cells (Nichogiannopoulou et al., 1999) that should contain RAG1/GFPþSca-1þc-Kitþ ELPs (Igarashi et al., 2002). Despite signiﬁcant reductions in primitive progenitor and stem cell numbers, Ikaros/ mice possess relatively normal numbers of myeloerythroid colonies (Nichogiannopoulou et al., 1999), suggesting a compensatory mechanism in vivo. Normal numbers of thymic IL-7Ra/loc-Kitþ cells in Ikaros/ mice could thus be due to similar compensatory mechanisms in the thymus. These data, however, at least provide evidence that bone marrow CLPs and IL-7Ra/loc-Kitþ ETPs are phenotypically and functionally different. The lineal relationships between these populations remain to be determined. T cell precursors that seed the thymus from the bone marrow have been

LINEAGE COMMITMENT

17

intensively characterized (Mori et al., 2001). Two to 4 days after intravenous transplantation of whole bone marrow cells into irradiated hosts, donor-derived cells in the thymus were all CD25c-Kit cells. This surface phenotype is different from either CLPs or IL-7Ra/loc-Kitþ cells, although c-Kit expression in progenitor and stem cells could be downregulated upon rapid expansion (after 5-FU treatment, for example) (Doi et al., 1997; Randall and Weissman, 1997). The CD25c-Kit population predominantly differentiated into T and B cells by secondary intravenous transplantation into congenic hosts (Mori et al., 2001). In contrast, when whole bone marrow cells were injected directly into the thymus, bone marrow cells proliferated and displayed not only lymphoid but also myeloid differentiation in secondary hosts after intravenous transplantation, indicating that the thymic environment does not eliminate cells with myeloid progenitor potential (Mori et al., 2001). Therefore, the thymus appears to normally act as a ﬁlter, allowing lymphoid but not myeloid or multipotent progenitors to seed the thymic microenvironment. It is thus possible that the IL-7Ra/loc-Kitþ population is formed after the selective expansion of thymus-seeding bone marrow progenitors. It is critical to test directly whether intravenously injected CLPs or RAG-1/GFPþSca-1þc-Kitþ ELPs can reconstitute IL-7Ra/loc-Kitþ ETPs in the thymus, and whether IL-7Ra/loc-Kitþ ETPs can clonally differentiate into both T and B cells. VII. Common Myeloid Progenitors as the Counterpart of CLPs

As has been noted, the generation of multilineage CFU-Mix colonies led to the notion of a common myeloerythroid progenitor. Since HSCs or multipotent progenitors can also form CFU-Mix colonies, this is not direct evidence of their existence. With the discovery of CLPs, and the fact that no myeloerythroid potential existed within the IL-7Raþ bone marrow fraction, myeloid-restricted progenitors were searched for by cell surface marker within the IL-7Ra fraction of mouse bone marrow (Akashi et al., 2000). By plating various phenotypic IL-7Ra fractions into methylcellulose cultures, the vast majority of myeloid colony-forming activity was found within a IL-7RaLinSca-1cKitþ population that does not overlap with the phenotype of HSCs or CLPs. This fraction was further divided into three functionally distinct myeloid progenitor populations: FcgRII/IIIloCD34þ common myeloid progenitors (CMPs), FcgRII/IIIloCD34 megakaryocyte/erythrocyte-restricted progenitors (MEPs), and the FcgRII/IIIhiCD34þ granulocyte/monocyte-restricted progenitors (GMPs) (Fig. 3). CMPs gave rise to all myeloerythroid colony types in vitro including CFU-Mix, GMPs generated only GM-afﬁliated colonies (CFU-G, CFU-M, and CFU-GM), and MEPs generated only MegE-afﬁliated colonies (BFU-E, CFU-Meg, and CFU-MegE). Lineal studies demonstrated that CMPs differentiated into phenotypic GMPs and MEPs;

18

DAVID TRAVER AND KOICHI AKASHI

both daughter populations were replated and generated only GM- or MegErestricted progeny, respectively. The majority of single CMPs generated both GM- and MegE-related progeny in two-step culture assays (Akashi et al., 2000). Upon in vivo transfer, these populations displayed short-term production of lineages corresponding to their in vitro activities, indicating that they do not appreciably self-renew (Na Nakorn et al., 2002). These three myeloid progenitor subsets likely represent the major pathways for myeloerythoid cell differentiation since very little colony-forming activity exists outside of the IL-7RaLinSca-1c-Kitþ fraction in steady-state bone marrow. These data also indicate that the commitment of CMPs toward the megakaryocyte/erythrocyte (MegE) or the granulocyte/macrophage (GM) lineages are mutually exclusive events, since MEPs and GMPs never produced colony types outside of their noted potentials (Fig. 1). From the differentiation model illustrated in Fig. 1, it is logical to assume that GMPs and MEPs give rise to monopotent precursors. Some data suggest that this is indeed the case for at least megakaryocyte and mast cell development. Megakaryocyte-restricted progenitors (MKPs) were identiﬁed within the IL-7RaLinSca-1c-Kitþ Thy1.1 marrow fraction using an additional CD9 surface marker (Nakorn et al., 2003). MKPs were found to generate only megakaryocytes in vitro and could give rise to platelets for approximately 3 weeks when transplanted into mice. MKPs could be phenotypically isolated following culture of both CMPs and MEPs, suggesting that they are downstream of both progenitor subsets. Mast cell progenitors (MCPs) (Iwasaki and Akashi, 2001) and eosinophil progenitors (EoPs) (unpublished data) were similarly identiﬁed downstream of GMPs. One potential caveat to both studies is that, by ﬁve-color FACS, only certain combinations of surface markers can be simultaneously analyzed. It will therefore be of interest to expand these studies to nine-color FACS, for example, to simultaneously visualize MKP and MCP cell surface proﬁles within the context of the markers used to isolate the CMP, MEP, and GMP progenitor subsets. Similar approaches will likely yield monopotent progenitors for erythrocytes, monocytes, and granulocyte subsets. VIII. Lineage Priming by Promiscuous Gene Expression in Multipotent Stem and Progenitor Cells

The ability to prospectively isolate lineage-restricted progenitor subsets representing the major hematopoietic branchpoints enables analysis of the molecular mechanisms of lineage commitment. For genetic programs to be activated, chromatin structures that allow RNA polymerase to initiate transcription are required (Berger and Felsenfeld, 2001; Felsenfeld et al., 1996). It has been shown that the activation of chromatin remodeling can occur prior to signiﬁcant expression of genes in the region of interest (Kontaraki et al.,

LINEAGE COMMITMENT

19

2000; Weintraub, 1985). These results led to the hypothesis that open chromatin structure is maintained in early hematopoietic progenitors, enabling multilineage differentiation programs to be readily accessible (Cross and Enver, 1997). This ‘‘priming’’ of genes afﬁliated with multiple lineages would then afford ﬂexibility in cell fate decisions and allow multipotent precursors to rapidly respond to environmental cues. If many genes are, in fact, primed at multi- or oligopotent progenitor stages, these progenitors should then ‘‘promiscuously’’ express genes of multiple lineages at the single-cell level. This hypothesis was ﬁrst tested in myeloerythroid precursors by using a multipotential myeloid cell line, FDCP-mix. In this line, activation of the b-globin locus control region, assayed by erythroid-speciﬁc DNase I hypersensitive sites, occurs prior to erythroid commitment (Jimenez et al., 1992). Furthermore, single FDCP-mix cells coexpress a variety of genes related to the GM and MegE lineages, including MPO and b-globin (Hu et al., 1997). The coexpression of MPO and b-globin was subsequently found in a fraction of human LinCD34þ cells (Hu et al., 1997) and in murine intraembryonic hematopoietic progenitors in the aorta–gonad–mesonephros region (Delassus et al., 1999). These limited data support the lineage priming model; remaining questions are whether promiscuous gene expression is found in normal stem and progenitor cells in vivo, and whether this activity is important in cell fate decisions and the progressive restriction of fate potentials along the hematopoietic hierarchy. These issues have been addressed in puriﬁed progenitor populations at both the single-cell and population levels by RT-PCR and oligonucleotide microarrays, respectively. IX. Lymphoid and Myeloid Promiscuity Demonstrated by Single-Cell Analyses

Since CMPs possess both myeloid and erythroid differentiation potentials and give rise to MEPs and GMPs with restricted, mutually exclusive potentials, highly puriﬁed single progenitors were analyzed for gene-expression proﬁles (Miyamoto et al., 2002). Virtually all single GMPs expressed GM-related genes such as MPO and granulocyte colony-stimulating factor receptor (G-CSFR), but did not express MegE-related genes such as b-globin or EpoR. Conversely, all single MEPs expressed b-globin and/or EpoR but not MPO or G-CSFR. In this way, the genetic ‘‘ﬁngerprints’’ of commited GMPs and MEPs were identiﬁed. In marked contrast, 60% of single CMPs coexpressed all of these GM- and MegE-afﬁliated genes. Analysis of additional myeloerythroid transcription factors at the single-cell level displayed a similar pattern (Miyamoto et al., 2002): More than 50% of single CMPs expressed both PU.1, a master gene for GM and B cell development (McKercher et al., 1996; Scott et al., 1994) and NF-E2, a critical gene for erythroid/megakaryocyte differentiation and/or survival (Romeo et al., 1990; Shivdasani et al., 1995). These data

LINEAGE COMMITMENT

21

directly demonstrate that priming of many myeloerythroid genes occurs in CMPs and that quenching of gene expression inappropriate for GM and MegE-related programs occurs in downstream MEPs and GMPs, respectively (Fig. 5). Lineage promiscuity in lymphoid progenitor populations has similarly been evaluated (Miyamoto et al., 2002). In this study, expression of the T lymphoid genes, GATA-3 and CD3d, and the B lymphoid genes, l5 and Pax-5, were analyzed. GATA-3 is required for the development and/or survival of T cellcommitted precursors (Ho et al., 1991; Ting et al., 1996) and CD3d for proper functioning of the T cell receptor (van den Elsen et al., 1985). Pax-5 is a B-cellrelated transcription factor (Horcher et al., 2001; Nutt et al., 1999) and l5 (Sakaguchi and Melchers, 1986) is indispensable for preB cell receptor formation (Karasuyama et al., 1994). Virtually all single proB cells expressed l5 and Pax-5 but not GATA-3 or CD3d, whereas single proT cells expressed only GATA-3 and CD3d. In contrast, 20% of single CLPs coexpressed all of these lymphoid genes, indicating the existence of promiscuous gene expression at the CLP stage (Miyamoto et al., 2002) (Fig. 5). Cells with lineage promiscuous gene expression were then shown to be capable of differentiation into multiple lineages. Mice harboring an enhanced green ﬂuorescent protein (EGFP) reporter knocked-in to the murine lysozyme M (LysM) locus (Faust et al., 2000) were used to obtain progenitors with active expression of LysM. In the myeloid pathway, 60% of CMPs expressed a signiﬁcant level of LysM/GFP. Both LysM/GFPþ and LysM/GFP CMPs expressed erythroid b-globin demonstating coexpression of myeloerythoid genes, and displayed MegE differentiation in vitro at equal efﬁciencies (Miyamoto et al., 2002). In the lymphoid pathway, the effect of pTa gene expression on B cell differentiation potential was assayed using pTa/hCD25 transgenic mice (Gounari et al., 2002) (also see the previous section). Both pTa/hCD25þ and pTa/hCD25 CLPs displayed similar B cell potentials, and expression proﬁles of CD3d, GATA-1, Pax-5, and l5 by single cell RT-PCRs were identical (Miyamoto et al., 2002). Primed lysozyme M and pTa transcripts, therefore, do not impair cellular viability nor do they necessarily mark Fig 5 Myeloid and lymphoid promiscuity in normal hematopoietic progenitors. Single cell mutiplex-RT-PCR assays for myeloid-related (b-globin and EpoR for erythroid, G-CSFR and MPO for myelomonocytic) and for lymphoid-related genes (l5 and Pax-5 for B lymphoid, GATA-3 and CD3d for T lymphoid). The red rectangles indicate lineage promiscuous expression and white ones indicate no expression. More than 50% of single common myeloid progenitors (CMPs) coexpress both erythroid and myelomonocytic genes, and 20% of single common lymphoid progenitors (CLPs) coexpress both B and T lymphoid-related genes. These data strongly suggest that expression of lineage-related genes precedes commitment, and therefore downregulation of genes of unselected lineage might be important for bipotent progenitors to ultimately commit to certain lineages. (See Color Insert.)

22

DAVID TRAVER AND KOICHI AKASHI

commitment to myelomonocytic or T cell fates, respectively. This evidence indicates that promiscuous gene expression likely plays a key role in providing ﬂexibility in oligopotent precursors and that priming inappropriate to particular lineages is quenched upon differentiation (Miyamoto et al., 2002). X. Fate Choices Made by Combinations of Instructive Signals at Lineage Promiscuous Stages

Single progenitor cells subjected to RT-PCR analyses revealed considerable heterogeneity of gene expression at the oligopotent CMP and CLP stages. Given the snapshotlike nature of these assays, the diversity within each fraction could represent ﬂuctuation of expression in individual cells or distribution of stable expression patterns. Lineage promiscuity might represent a competition of lineage programs at levels below detection or at ‘‘ﬂoating’’ levels, which may be a prelude to exclusive commitment. In priming stages, multiple differentiation programs might collaborate or compete with each other until one becomes dominant. These may include crosstalk among transcription factors or transcriptional complexes. Transcription factor activity can be potentiated (Zhang et al., 1997) or suppressed (Sieweke et al., 1996; Zhang et al., 2000) by interaction with other transcription factors. Changes in levels of key transcription factors (DeKoter and Singh, 2000; Kulessa et al., 1995) could also be critical for lineage decisions. Previous work has suggested that transcription factors are critical for lineage commitment decisions. By using bi- or multipotent cell lines, it has been reported that overexpression of transcription factors, including PU.1, GATA, and C/EBP, results in the preference of one lineage fate from a choice of several (DeKoter and Singh, 2000; Kulessa et al., 1995; Nerlov and Graf, 1998; Nerlov et al., 1998; Scott et al., 1997). Some studies suggest that that the mechanism of ‘‘lineage instruction’’ by transcription factors should also include ‘‘lineage exclusion,’’ where the suppression of speciﬁc differentiation programs may be as important as ‘‘lineage speciﬁcation’’ (Rothenberg and Dionne, 2002). Speciﬁc lineage exclusion by speciﬁc transcription factors may be important to ‘‘stabilize’’ fate potentials upon commitment, presumably by alteration of chromatin. Possible interactions of two competitive transcription factors at primitive progenitor stages are illustrated in Fig. 6. A. PU.1 and GATA-1 in Hematolymphoid Commitment One of the most important molecular examples of fate instruction investigated so far is the interaction between PU.1 and GATA-1. PU.1 is expressed in myeloid cells, B cells, and NK cells, but not in erythrocytes or T cells. PU.1 is expressed in HSCs at a low level, and its expression increases in CMPs and GMPs, but is absent in MEPs (Akashi et al., 2000; Miyamoto et al., 2002).

LINEAGE COMMITMENT

23

Fig 6 Schematic representation of hypothesis of lineage commitment and lineage promiscuity. In this model, commitment is operated if cells express more than threshold levels (broken line) of either of the transcription factors. Transcription factors A and B are simultaneously expressed at low levels at the bipotent stage (lineage promiscuity), and their expression levels may be ﬂuctuated. Once commitment begins, the unused programs will be downregulated, and if either of TFs is upregulated beyond a threshold, their expression is stabilized, presumably by autoregulation. The upregulation of either of TFs might also exclude unrelated developmental programs in order to stabilize their lineage fate choice.

In the lymphoid pathway, PU.1 is expressed in CLPs and proB cells, but not in proT cells (Miyamoto et al., 2002). PU.1 binds to many myeloid gene promoters including the G-CSFR, GM-CSFR, M-CSFR CD11b, and MPO genes to regulate their expression (for review, see Tenen et al., 1997). A 1997 study reported that PU.1 also regulates expression of the IL-7R (Scott et al., 1997), which is necessary for T and B cell development (Akashi et al., 1997; Peschon et al., 1994; von Freeden-Jeffry et al., 1997). PU.1 knockout mice display variable anemia and lack all leukocytes, including monocytes, neutrophils, and B cells (Scott et al., 1994). Mice reconstituted with PU.1/ fetal liver cells exhibit a reduced number of NK cells in addition to loss of T and B cells (Colucci et al., 2001). These ﬁndings suggest that the PU.1 may play a critical role in the differentiation of HSCs into CLPs, and of CMPs into GMPs.

24

DAVID TRAVER AND KOICHI AKASHI

GATA-1 is an essential transcription factor for maturation of both megakaryocytic and erythroid precursors. GATA-1 knockout studies suggest that GATA-1 is necessary for terminal differentiation of erythrocytes and platelets (Fujiwara et al., 1996; Shivdasani et al., 1997). GATA-1 is expressed in HSCs at a low level (Akashi et al., 2000; Miyamoto et al., 2002), and its expression increases in CMPs and MEPs. GMPs or CLPs do not express GATA-1. In chickens, forced expression of GATA-1 reprograms myb-ets transformed progenitors into the erythroid, eosinophilic, and megakaryocytic pathways. During normal hematopoietic development, PU.1 expression increases as stem cells commit to the myeloid lineage, and PU.1 expression declines with erythroid differentiation as GATA-1 expression increases. It is important to understand whether the low-level expression of PU.1 and GATA-1 in HSCs represents an early stage of commitment. When GATA-1 is introduced into HSCs at a high expression level comparable to that seen within MEPs, HSCs immediately lose their self-renewal activity and commit into the MegE lineage (Iwasaki et al., 2002). It is thus highly likely that at the low expression levels normally observed in HSCs, these transcription factors do not exert lineage decisive effects. This, in turn, suggests that changes in expression levels of each transcription factor may be critical for lineage commitment. In cell line studies, PU.1 suppresses GATA-1 activity to block erythroid differentiation (Nerlov and Graf, 1998). Conversely, GATA-1 inhibits binding of PU.1 to cJun, a critical coactivator of PU.1 transactivation of myeloid promoters (Nerlov et al., 2000; Zhang et al., 1999, 2000). This mutually antagonizing effect of PU.1 and GATA-1 may play an important role in the GMP vs MEP commitment at the CMP stage. Expression levels of PU.1 and GATA-1 may be critical for cell fate decisions at hematopoietic branchpoints. For example, in a bipotent cell line, high expression levels of PU.1 instructed macrophage differentiation, whereas B cell differentiation occurred at low levels (DeKoter and Singh, 2000). In addition, once upregulated, GATA-1 and PU.1 can stabilize their own expression by positive autoregulation (Chen et al., 1995; Zhang et al., 1999). The upregulation of GATA-1 or PU.1 itself plays a critical role in exclusion of unselected lineages. A ‘‘lineage exclusive’’ effect has been clearly reported for PU.1 in T cell development (Anderson et al., 2002). In this study, fetal liver cells were retrovirally transduced with PU.1, and were cultured in fetal thymic organ culture (FTOC) systems (Anderson et al., 2002). Enforced PU.1 severely inhibited development of a/bT, g/dT, and B cells, but allowed macrophage differentiation. T cell development was blocked at the CD25þCD4CD8 stage. These immature T cell progenitors obtained in the primary FTOC could differentiate into macrophages when they were transferred into another FTOC culture system containing high oxygen levels. The reason why PU.1 inhibited B cell development may be related to the ﬁnding that high PU.1

LINEAGE COMMITMENT

25

expression directs macrophage commitment at the expense of B cells (DeKoter and Singh, 2000). This lineage exclusive effect has also been reported in the analysis of ectopic GATA-1 expression in GM and lymphoid development (Iwasaki et al., 2002). ProB cells or GM-committed cells retrovirally transduced with GATA-1 did not differentiate into lymphoid cells or GM cells, but immediately underwent apoptotic cell death in vitro (Iwasaki et al., 2002). Enforced expression of Bcl-2 in these cells did not prevent apoptosis. Enforced expression of hGM-CSFR in GATA-1þ GM-committed cells did not restore their GM differentiation (Iwasaki et al., 2002). Thus, GATA-1 appears to directly inhibit lymphoid and GM differentiation rather than inhibit ‘‘permissive’’ survival signals. Mechanisms of differentiation arrest and apoptosis induction remain to be elucidated. Similar competition of lineage programs could be involved at all of branchpoints of hematopoiesis. B. Notch1 in T Cell Commitment Notch genes encode a highly conserved family of transmembrane receptors that regulate cell fate decisions in many developmental pathways (for review, see Fleming, 1998; Robey, 1997). Notch signaling is initiated by ligand binding, including Jagged1, Jagged2, and Delta, upon which Notch is proteolytically cleaved to release its intracellular domain (NICD). NICD, the activated form of the Notch receptor, translocates into the nucleus and pairs with the murine transcription factor CSL (NICD pairs with CBF1 in humans, Su(H) in Drosophila, and Lag-1 in Caenorhabditis elegans). CSL is a DNA-binding protein that normally represses transcription by interacting with several corepressor complexes. The binding of NICD converts CSL into a transcriptional activator and induces the expression of target genes, such as the Hairy Enhancer of Split (HES) family of transcriptional repressors. The HES proteins modulate the activity of tissue-speciﬁc basic helix–loop–helix (bHLH) transcription factors, further propagating the effects of Notch signaling. Recent studies show an essential role of Notch1 signaling in the commitment of lymphoid progenitors to the T cell lineage in the thymus (for review, see Anderson et al., 2001; von Boehmer, 1999). Notch1 is not expressed in B cell progenitors such as proB and preB cells (Pui et al., 1999), whereas it is highly expressed in immature thymic T cells (Hasserjian et al., 1996; Washburn et al., 1997). Targeted conditional disruption of Notch1 (Radtke et al., 1999) or of its downstream effector Hes1 (Tomita et al., 1999) arrests T cell development at the earliest thymic stage. Interestingly, B cell development was considerably facilitated in Notch1deﬁcient mice. In the thymus, Notch1/ cells differentiated mainly into B cells where B lymphopoiesis is normally minimal. Conversely, reconstitution of bone marrow cells retrovirally transduced with activated Notch1 showed that ectopic Notch1 causes a persistent block in early B cell differentiation and

26

DAVID TRAVER AND KOICHI AKASHI

rapid emergence of an immature bone marrow T cell population (Pui et al., 1999). In human CD34þ cord blood cells, Notch1 signaling promotes T and NK development, but inhibits B cell differentiation in vitro (Jaleco et al., 2001). Together, these data suggest that Notch1 may play a decisive role in T versus B lymphocyte commitment, although this effect has not directly been tested at the bipotent CLP stage. Additionally, within commited T cell precursors, activated Notch1 supports abT cell development at the expense of gdT cell differentiation (Washburn et al., 1997). Within abT cells, activated Notch1 skewed differentiation toward CD8þ and away from CD4þ T cells (Robey et al., 1996). Thus, Notch1 appears to play a role at multiple checkpoints in T cell development. How can Notch1 exert such decisive effects along the lymphocyte maturational pathway? One possible mechanistic interpretation is that Notch1 provides permissive signals, such as cell survival or proliferation signals (Jehn et al., 1999; von Boehmer, 1999), since enforced Notch1 signaling stimulated HSC self-renewal in vitro (Varnum-Finney et al., 2000) and blocked TCR-mediated T cell apoptosis (Jehn et al., 1999). Thus, Notch signaling in hematolymphopoiesis may be to maintain cellular viability. In this context, Notch signaling may preferentially support the survival of CD8þ abT cells (von Boehmer, 1999). However, the mechanism by which ectopic Notch1 excludes B cell development remains unknown. It is also important to analyze the interactions with molecules downstream of Notch1 that are essential for T cell development (i.e., HES-1) (Kaneta et al., 2000; Tomita et al., 1999) and B cell development (i.e., E47 (Ordentlich et al., 1998) and CSL reference). The pTa gene was identiﬁed as a transcriptional target of Notch signaling in T cells (Deftos et al., 2000). pTa encodes a transmembrane protein that pairs with the nascent TCRb chain to form the essential pre-TCR signaling complex in thymocytes (Fehling et al., 1995). pTa expression is primed in CLPs, representing ‘‘lymphoid lineage promiscuity’’; the upregulation of pTa coincides with irreversible T cell lineage commitment in both murine and human thymic precursors (Gounari et al., 2002; Miyamoto et al., 2002). pTa is required for ab but not gd T cell development (Fehling et al., 1995). This facilitation of ab lineage commitment is believed to occur by instructive signaling from the pre-TCR (Aifantis et al., 1998). The pTa enhancer can be activated by Notch signaling and contains binding sites for its nuclear effector, CSL (Reizis and Leder, 2001). Mutation of CSL-binding sites abolished enhancer induction by Notch and delayed the upregulation of pTa transgene expression during T cell lineage commitment. These results strongly suggest that Notch1 signals play a key role in T cell speciﬁcation through upregulation of pTa via the CSL signaling pathway (Reizis and Leder, 2001). It is critical to analyze directly the physiological expression level and distribution of Notch1 protein in CLPs in order to test whether Notch signaling dictates the T versus B fate decision.

LINEAGE COMMITMENT

27

C. Pax-5 in B Cell Commitment Pax-5, a key molecule in B cell differentiation (Nutt et al., 1999), also plays a critical role in lymphoid commitment. In steady-state hematopoiesis, Pax-5 expression initiates at the CLP stage, is upregulated in proB cells, but is undetectable in proT or myeloid cells (Miyamoto et al., 2002). Pax-5 appears to maintain late B cell differentiation through several mechanisms including class switching of the IgH gene, regulation of immunoglobulin 30 enhancers, and transcriptional repression of J-chain genes (for review, see Busslinger et al., 2000). Upregulation of Pax-5 in proB cells is thus likely important to restrict differentiation into a B cell fate. Puriﬁed pro-B cells from Pax5-deﬁcient (Pax5/) mice (having recombined D-VJ loci of IgH genes and expressing the l5, E2A, VpreB, Iga, and Igb genes) can differentiate into T cells, NK cells, granulocytes, macrophages, and osteoclasts, as well as B cells in vitro (Nutt et al., 1999). In this report, however, the presence of D-VJ IgH gene rearrangements in the nonlymphoid progeny of assayed pro-B cells was not tested. In contrast, in an adoptive transfer assay, injection of 0.5-1 108 Pax5/ proB cells into sublethally irradiated RAG-2-deﬁcient mice resulted in reconstitution of T cells with D-VJ rearrangement of IgH genes, but not in reconstitution of myeloid cells (Rolink et al., 1999). The discrepancy of myeloid cell production between the in vitro and in vivo assays may be because the myeloid differentiation potential of Pax-5/ proB cells is limited and below detectable levels in vivo. Furthermore, the efﬁciency of T cell differentiation from Pax-5/ proB cells is quite low when compared to CLPs, since 1 103 IL-7Raþ CLPs were minimally required to fully reconstitute T cell development in sublethally irradiated Rag-2/ mice (Kondo et al., 1997). Accordingly, it is possible that only a minor fraction of Pax-5/ proB cells could commit to the T cell lineage in vivo. Nonetheless, these data indicate that, even after the onset of B lineage-associated genes transcription, differentiation programs for other lineages (at least for the T cells) remain accessible. Pax-5/ proB cells can therefore be considered to maintain CLP potential. This is compatible with recent data showing that a fraction of bone marrow B220þ cells maintain the expression of pTa and T cell potential (Gounari et al., 2002). The reduced levels of Pax5 observed in vivo in EBF/E2A double heterozygote mice suggests that these factors are upstream of Pax-5 and may regulate Pax-5 expression in B cells (O’Riordan and Grosschedl, 1999). Accordingly, EBF has been shown to bind the Pax5 promoter and induce its transcription. In mice lacking one copy of the transcription factors EBF and E2A, other genes besides Pax-5 were affected. Speciﬁcally, expression of l5, VpreB, Rag2, and mb-1 genes was substantially reduced (O’Riordan and Grosschedl, 1999). These data suggest that E2A and EBF are expressed prior to B lineage commitment and initiate transcription of many B lineage speciﬁc genes, and

28

DAVID TRAVER AND KOICHI AKASHI

that the induction of Pax5 is an important step in the pathway to B cell commitment. Does Pax-5 have an instructive effect in B cell commitment? A 2001 report formally showed that Pax-5 plays a critical role in T versus B cell development. Pax-5 was conditionally excised by CD19-cre or Mx-cre expression in mice carrying a ﬂoxed Pax-5 allele. These conditional Pax-5-deﬁcient mice displayed preferential loss of mature B cells, inefﬁcient lymphoblast formation and reduced serum IgG levels (Horcher et al., 2001). Furthermore, when Pax-5 was deleted from mature B cells, they rapidly lost B cell markers such as CD19, CD22, CD40, MHC class II, and IgD, but upregulated M-CSFR expression, like Pax-5/ proB cells. Rather than instructing B lymphoid commitment from CLPs, these data suggest that Pax-5 may be critical in ‘‘stabilizing’’ committed B cell precursors, in part, by suppressing myeloid gene expression. It has also been reported that retroviral transduction of Pax-5 into bone marrow progenitors profoundly inhibited myeloid colony formation, presumably by blocking signals from myelomonocytic cytokine receptors such as G-CSFR and GM-CSFR (Chiang and Monroe, 1999). However, in a later study, enforced Pax-5 expression did not suppress myeloid fates (Cotta et al., 2003). The inconsistency between these studies may be due to differences in expression levels of ectopic Pax-5. In a 2002 study, a heterozygous knock-in mouse carrying a Pax-5 gene under the control of the Ikaros locus was reported (Souabni et al., 2002). In this mouse, transgenic Pax-5 is expressed in HSCs as well as in myeloerythroid and lymphoid progenitors. Interestingly, enforced expression of Pax-5 inhibited T lymphopoiesis, whereas it augmented B lymphopoiesis as expected. Myelopoiesis was largely normal, although MegE differentiation was slightly reduced. Interestingly, enforced expression of Pax-5 in Pax-5/ proB cells suppressed Notch1 expression (Souabni et al., 2002), possibly explaining how T lymphopoiesis is inhibited. Thus, enforced Pax-5 does not possess a lineage instructive effect in uncommitted multipotent progenitors but rather functions as a master regulator of B cell development only after cells commit to the lymphoid lineages. In normal hematopoiesis, Pax-5 is expressed at a low level in CLPs together with T lymphoid genes representing their lymphoid lineage priming (Fig. 4) (Miyamoto et al., 2002). It will be interesting to test the effects of overexpression or suppression of Pax-5 and Notch1 at the CLP stage in order to understand their roles in T versus B lymphoid commitment. XI. Lineage Plasticity in Lymphoid Progenitors: Fate Choices are Reversible

The existence of promiscuous gene expression programs at the major hematopoietic branchpoints suggests that the fate potentials of each progenitor may reﬂect differential accessibility to speciﬁc transcriptional programs

LINEAGE COMMITMENT

29

(Akashi et al., 2003). What events are required for the transition from priming to committed stages? It is conceivable that both the upregulation of genes related to the selected lineage and the downregulation of genes irrelevant to the selected lineage are required for commitment. If the expression of critical genes is deregulated by transforming events, for example, do progenitors differentiate to lineages outside their physiological potential? Evidence suggests that this phenomenon can, in fact, occur (Graf, 2002). Lineage conversion was originally reported in murine B cell lines: PreB cell lines transformed by the Abelson virus differentiated into macrophages following treatment with a DNA demethylating agent, 5-azacytidine (Boyd and Schrader, 1982), and B cell lines immortalized with Em-myc transgenes differentiated into macrophages upon overexpression of v-raf (Klinken et al., 1988). Thus, committed B cell precursors harbor latent myelomonocytic potentials that can be activated by speciﬁc genetic alterations. Conversely, granulocytic conversion was reported in a pre-B lymphoma line (Lindeman et al., 1994). These phenomena have postulated a ‘‘close’’ relationship between the B and GM lineages, but it is unclear whether the latent GM potential in B cell lines represents residual oligopotentiality maintained from upstream progenitors or whether B and GM lineages utilize similar genetic programs. In a fraction of human acute myelogenous leukemia (AML), leukemic blasts possess lymphoid markers including monoclonal rearrangement of TCR or Ig genes (Cheng et al., 1986). Conversely, acute lymphoblastic leukemia (ALL) or even myeloma cells occasionally express myeloid markers (Akashi et al., 1991; Grogan et al., 1989). Conversions between lymphoid and myeloid leukemias have also been seen (Murphy et al., 1983; Stass et al., 1984). These ‘‘mixed’’ leukemias have been ascribed to nonphysiological differentiation induced by oncogenic events (lineage inﬁdelity) (McCulloch, 1987; Smith et al., 1983). Recent studies demonstrate that lineage conversion can be induced in committed progenitors simply by enforcing or eliminating single differentiation-related genes such as cytokine receptors and transcription factors. A. Myelomonocytic Conversion of Lymphoid-commited Progenitors by Instructive Cytokine Signals It has been suggested that cytokine signals do not play a role in myeloid lineage determination: Mice with targeted mutations in myeloid cytokines or their receptors do not display a complete loss of speciﬁc lineages (Gillessen et al., 2001; Lieschke et al., 1994; Stanley et al., 1994; Wu et al., 1995), and myeloerythroid progenitors expressing transgenic receptors for myeloid cytokines, such as human granulocyte/macrophage colony-stimulating factor (hGM-CSF) (Nishijima et al., 1995), human granulocyte colony-stimulating factor (hG-CSF) (Yang et al., 1998, 1999), and murine interleukin-5 (IL-5)

30

DAVID TRAVER AND KOICHI AKASHI

(Takagi et al., 1995), did not affect their physiological fates. Constitutive expression of the activated form of EpoR in bone marrow cells also supports GM differentiation (Pharr et al., 1993). Mice with a targeted chimeric mutation in the G-CSF receptor (G-CSFR), where the cytoplasmic domain is replaced by the cytoplasmic portion of the EpoR, do not display any abnormality in their hematopoietic development (Semerad et al., 1999). Signals from these myeloerythroid cytokine receptors may therefore be largely redundant or may simply serve to enhance survival. Thus, the lineage-speciﬁc actions of these myeloid cytokines may rely solely upon the lineage-speciﬁc expression of their cognate receptors. At the level of CLPs, however, myelomonocytic conversion can be induced by ectopic signals from myeloid-related cytokines. GM conversion by ectopic IL-2 or retrovirally induced GM-CSF signals was also observed from proT cells with rearranged T cell receptor b genes but not in downstream preT cells or B cell progenitors (King et al., 2002; Kondo et al., 2000). In another study using hGM-CSFRa/bc double transgenic mice, >50% of hGM-CSFRþ CLPs and >20% hGM-CSFRþ proT cells gave rise to granulocytes, monocytes, and/or dendritic cells, but not MegE lineage cells in the presence of hGM-CSF (Iwasaki-Arai et al., 2003). These data collectively suggest that conversion to GM fates by ectopic GM-CSF signals can occur in CLPs and their downstream lymphoid progeny, but that this potential progressively disappears as cells become more mature. During the conversion of CLPs into the myelomonocytic lineage, a number of GM-related cytokine receptors (e.g., G-CSFR, M-CSFR) and transcription factors (e.g., C/EBPa, PU.1) were reactivated, whereas MegE-related genes such as EpoR and GATA-1 were not (Iwasaki-Arai et al., 2003). Thus, GM-CSF signals can activate GM but not MegE differentiation programs, suggesting their GM-speciﬁc ‘‘instructive’’ role in lineage commitment. Furthermore, this effect is speciﬁcally found with the GM-CSFR, since retrovirally transduced G-CSFR and M-CSFR could not induce GM conversion from CLPs. It should be interesting to test whether the transcription factors downstream of the GM-CSFR can similarly induce GM conversion from CLPs. B. Megakaryocyte/Erythrocyte Conversion from Lymphoid-commited Progenitors The loss of GM and MegE potentials in lymphoid cells might occur simultaneously as multipotent stem cells generate CLPs (Kondo et al., 1997) and common myeloid progenitors (CMPs) (Akashi et al., 2000) since each subset displayed mutually exclusive differentiation potentials. If the inducible GM potential in CLPs reﬂects residual multipotentiality from an upstream progenitor, it is reasonable to expect that MegE conversion could also be induced in CLPs or their lymphoid progeny.

LINEAGE COMMITMENT

31

GATA-1, a critical transcription factor for MegE development, has been reported to instruct MegE and eosinophil commitment. GATA-1 is expressed in erythroblasts, megakaryocytes, eosinophils and, mast cells. The lineage instructive effect of GATA-1 was ﬁrst demonstrated in a multipotential chicken cell line transformed by the Myb-Ets-encoding E26 leukemia virus (Kulessa et al., 1995). This cell line differentiates to the GM lineages following enforced expression of PU.1 whereas enforced expression of GATA-1 caused erythroid or eosinophilic differentiation (Kulessa et al., 1995). Within the myelomonocytic lineages, overexpression of GATA transcription factors can induce MegE phenotypes (for review, see Graf, 2002). Introduction of GATA-1 into myeloid cell lines induces megakaryocyte differentiation (Visvader et al., 1992) with upregulation of MegE-afﬁliated genes such as the erythropoietin receptor (EpoR) and a-globin (Seshasayee et al., 1998; Yamaguchi et al., 1998). Furthermore, both GM-restricted colony-forming cells, which are selectively generated in culture, and puriﬁed GMPs can give rise to erythroblasts, megakaryocytes, and eosinophils by GATA-1 transduction (Heyworth et al., 2002; Iwasaki and Akashi, 2001). These studies suggest that GATA-1 is sufﬁcient to reactivate the MegE and/or eosinophil differentiation programs in immature myelomonocytic cells. In a recent report, GATA-1 was retrovirally introduced into CLPs (Iwasaki et al., 2002). Strikingly, GATA-1 converted CLPs into the MegE lineages, inducing differentiation of hemoglobinized erythroblasts and mature megakaryocytes even in the absence of Tpo or Epo. Furthermore, GATA-1transduced CLPs could not differentiate into T or B cells in vivo, indicating that ectopic GATA-1 inhibited normal lymphoid differentiation from CLPs. GATA-1 altered the expression proﬁles of lineage-afﬁliated genes in CLPs into those observed in MegE-committed MEPs, inducing the upregulation of genes essential for MegE development such as FOG-1, and concomitantly downregulated genes related to the GM and lymphoid lineages including PU.1, Pax5, and IL-7Ra. The reactivation of GATA-1 appears to be sufﬁcient for, and a minimum requirement for, MegE conversion from CLPs (Iwasaki et al., 2002). Together, these data suggest that CLPs are normally lymphoid-restricted because they have downregulated myeloerythroid genes such as GM-CSFR and GATA-1. In turn, latent GM and MegE potentials of CLPs and early T and B lymphoid precursors indicate that the GM and MegE programs are still accessible after physiological lymphoid commitment, presumably by chromatin remodeling. It is of interest to test whether CMPs, the myeloid counterpart of CLPs, possess similar ‘‘plasticity’’ for lymphoid differentiation. Observed plasticity of lineage commitment at the early stages of lymphoid development likely reﬂects residual multipotency from upstream priming stages. This residual multipotency appears to be latent in normal commitment progenitors, presumably by epigenetic programs that control transcriptional accessibility, as

32

DAVID TRAVER AND KOICHI AKASHI

evidenced by gene expression proﬁling in microarray studies (see following text). Stabilization of lineage commitment could be achieved by downregulation of master ‘‘instructive’’ genes that ectopically induce lineage conversion as well as the upregulation of gene sets to directly ‘‘exclude’’ other fate choices. XII. Comparison of Gene Expression Profiles among Early Hematopoietic Stem and Progenitor Cells

That promiscuous gene expression exists at each of the major hematopoietic branchpoints suggests that priming of multiple lineage-afﬁliated programs allows fate choice ﬂexibility at each progenitor stage. The PCR studies that generated these ﬁndings used single cells and assayed only partially representative myeloid and lymphoid genes. Thus, a more global view of gene expression, including hematopoietic and nonhematopoietic genes is essential to unravel the complexity of genetic programs within early hematopoietic subsets. Oligonucleotide microarray analyses were thus performed using highly puriﬁed long-term HSCs, short-term HSCs, CMPs, and CLPs (Akashi et al., 2003). HSCs expressed a variety of myeloid (GM and MegE-afﬁliated) but not lymphoid genes. CMPs coexpressed many GM and MegE-afﬁliated genes, and CLPs coexpressed many T, B, and NK lymphoid-related genes on Affymetrix chips. Thus, genome-wide gene proﬁling revealed that HSCs predominantly exhibit myeloid promiscuity, and that CLPs and CMPs exclusively possess lymphoid and myeloid priming programs, respectively (Akashi et al., 2003). In short-term HSCs (or multipotential progenitors), both lymphoid and myeloid genes were primed. However, as has been mentioned, myeloid and lymphoid gene coexpression was not assayed at the single-cell level in these cell types. Therefore, these ﬁndings in ‘‘short-term HSCs’’ could be due to a combination of heterogeneous expression proﬁles. Nonetheless, this study strongly suggests that lineage promiscuous priming might be a common transcriptional feature in uncommitted stem and progenitor cells, and that primed genes may represent their full and immediate differentiation potentials (Miyamoto et al., 2002). This genome-wide approach discloses another important ‘‘priming’’ event at the level of HSCs. Primitive HSCs expressing CD45, a hematopoietic cellspeciﬁc marker, express approximately 70% of all nonhematopoietic genes, including genes characteristic of neuronal, endothelial, pancreatic, kidney, liver, heart, hair, epithelial, and muscle cell types (Akashi et al., 2003). These nonhematopoietic genes were detectable by nested RT-PCR in CD45þ HCS at the single- to 10-cell levels. This broad transcriptional usage, however, is lost as HSCs generate CMPs and CLPs; these cell types displayed only myeloid and lymphoid expression proﬁles, respectively.

LINEAGE COMMITMENT

33

These data also demonstrate that HSCs possess transcriptional accessibility for nonhematopoietic genes associated with multiple organ systems (Akashi et al., 2003). Recent reports demonstrate that murine bone marrow contains cells capable of differentiation into multiple organs, including endothelial cells, skeletal and cardiac muscle (Orlic et al., 2001), neurons and glia (Priller et al., 2001), parenchymal liver cells (Lagasse et al., 2000), and/or epithelial cells (Krause et al., 2001), as well as hematopoietic cells. While these reports suggest that the bone marrow may be special in harboring precursors of nonhematopoietic fates, the notion of HSC plasticity was challenged by a study in which mice reconstituted with single, GFPþ HSCs were extensively analyzed for GFP expression in nonhematopoietic tissues (Wagers et al., 2002). These clonal experiments demonstrated that HSCs very rarely contributed to nonhematopoietic fates, suggesting that ‘‘transdifferentiation’’ of HSCs is unlikely in this setting (Wagers et al., 2002). It was also reported that cell fusion can occur during co-culture of embryonic stem (ES) cells with HSCs or neural stem cells (NSCs). Although cell fusion was observed at an extremely rare incidence (104 to 105), ‘‘conversion’’ from NSCs (or HSCs) to other cell types could be obtained through spontaneous generation of hybrid cells rather than epigenetic reprogramming of the somatic stem cells (Terada et al., 2002; Ying et al., 2002). Thus, it is likely that ‘‘transdifferentiation’’ is a rare event under physiological conditions (Lemischka, 2002; McKay, 2002). Transdifferentiation in the settings of increased tissue renewal or tissue damage remains to be precisely addressed using prospectively isolated HSCs. Based on the ﬁndings that HSCs normally express many nonhematopoietic genes, it will be similarly interesting to search among these genes for candidate molecules that may instruct nonhematopoietic fate outcomes from bona ﬁde HSCs. XIII. Early Lymphoid Progenitors can Differentiate into Antigen-presenting Dendritic Cells

Dendritic cells (DCs) are bone marrow-derived leukocytes that were initially deﬁned by their high antigen presentation capacities and their ability to prime antigen responsiveness in naive T cells (Banchereau and Steinman, 1998; Hart, 1997). Since monocytes can give rise to DCs in vitro (Inaba et al., 1993), it was initially believed that DCs were of myeloid origin. It is now recognized that DCs develop from both lymphoid and myeloid precursors (Manz et al., 2001). A lymphoid origin for DCs was ﬁrst reported using early thymocyte progenitors (TPs) (CD4loCD44þCD25c-Kitþ) and thymic proT cells (CD44þCD25þcKitþ). Each population is capable of generating CD8aþ DCs in vivo (Ardavin et al., 1993; Wu et al., 1996). Because the majority of thymic DCs

34

DAVID TRAVER AND KOICHI AKASHI

express CD8aþ, it was proposed that CD8a expression is a marker for DCs of lymphoid derivation, whereas conventional CD8a DCs were of ‘‘myeloid’’ origin. This model appeared to be supported by the fact that mice deﬁcient for the myeloid transcription factors, RelB (Wu et al., 1998) or PU.1 (Guerriero et al., 2000) lack CD8a DCs, while ‘‘lymphoid’’ transcription factors, Ikaros (Wu et al., 1997), or Id2 (Hacker et al., 2003) lack only CD8aþ DC subsets. On the other hand, developmental dissociation of CD8aþ DC and T cells was reported in T cell-deﬁcient c-Kit/ gc/ (Rodewald et al., 1999) or Notch1/ (Radtke et al., 2000) mice, suggesting that CD8a is not necessarily a marker of lymphoid derivation. Each myeloid and lymphoid progenitor subset was thus examined for DC differentiation potential. Interestingly, CMPs could generate both CD8aþ and CD8a DCs (Traver et al., 2000). Both CD8aþ and CD8a DCs were produced from CLPs and CMPs with similar efﬁciency on a per-cell basis (Manz et al., 2001; Wu et al., 2001). Thus, expression of CD8a is not indicative of a lymphoid origin, and the presence or absence of CD8a expression among DC subsets is likely to reﬂect maturation status rather than ontogeny. The precursor/progeny relationship between CD8aþ and CD8a DCs, however, is still controversial. Splenic CD8aCD11cþ DCs were reportedly able to give rise to CD8aþCD11cþ DCs (Martinez del Hoyo et al., 2002), but this ﬁnding was not supported by another report (Naik et al., 2003). It has also been suggested that the acquisition of CD8a expression is likely to initiate in splenic DC precursors that are CD11c, since CD8aþCD11c spleen cells gave rise to mature CD8aþCD11cþ DCs (Wang et al., 2002). On the other hand, both CD8aþ and CD8a DCs are reported to originate from CD11cþMHCII DC progenitors which were devoid of other myeloid or lymphoid differentiation potential (Martinez del Hoyo et al., 2002). This population was also shown to give rise to interferon (IFN)-g-producing plasmacytoid DCs (PDCs; see following text) (Martinez del Hoyo et al., 2002). These data suggest that the CD11cþMHCII DC progenitors may represent a DC stage independent of DC pathways from CLPs or CMP. It is more likely, however, that CLPs and/or CMPs generate CD11cþMHCII DC precursors that are upstream of mature CD8aþ and CD8a DCs. Interestingly, DC potential is maintained downstream of CLPs in proT cells and downstream of CMPs in GMPs, whereas DC potential is lost once B cell or megakaryocyte/erythrocyte commitment occurs (Manz et al., 2001). Although murine DCs and their precursors are usually isolated from lymphoid organs or bone marrow, human DCs are usually isolated from peripheral blood (for review, see Shortman and Liu, 2002). Human DCs can be derived from CD34þ progenitors, lymphoid restricted progenitors, and from peripheral blood monocytes, suggesting that human DCs similarly derive from both myeloid and lymphoid pathways (Caux et al., 1996; Hao et al., 2001;

LINEAGE COMMITMENT

35

Randolph et al., 1998; Romani et al., 1994; Sallusto and Lanzavecchia, 1994; Young et al., 1995). In both mice and humans, there appears to be no functional difference between lymphoid- and myeloid-derived DCs as evaluated by mixed leukocyte reactions, immunophenotyping, and cytokine production (Manz et al., 2001; Traver et al., 2000). A precursor population that immediately generates DCs capable of producing interferon a was identiﬁed in humans (Cella et al., 1999; Grouard et al., 1997; Jarrossay et al., 2001; Kadowaki et al., 2000, 2001) and subsequently in mice (Asselin-Paturel et al., 2001; Bjorck, 2001; Nakano et al., 2001). In addition to the conventional DC markers such as CD11c and MHC class II, these cells express CD45RA. This population has been termed plasmacytoid dendritic cells (PDCs) or DC2 (Liu, 2001). In humans, DCs of this phenotype were reportedly generated from at least M-CSFR-expressing myeloid progenitors (Olweus et al., 1996). On the other hand, human CD34þ cells with enforced Id2 or Id3 expression reportedly differentiated into conventional DCs, but not T, B, or PDCs (Heemskerk et al., 1997; Jaleco et al., 1999; Spits et al., 2000), suggesting that PDCs may have a lymphoid origin. Accordingly, PDCs express genes usually found in the lymphoid lineage such as preTa, 14.1, and Spi-B (Bendriss-Vermare et al., 2001; Spits et al., 2000). In our hands, murine PDC equivalents could be induced from both CLPs and CMPs (unpublished data). Thus, it is likely that PDCs as well as conventional DCs originate from both lymphoid and myeloid pathways. CLPs and proT cells, but not preT or proB cells, can differentiate into DCs (Manz et al., 2001; Traver et al., 2000; Wu et al., 1996). It is interesting to note that DC potential is contained within the same lymphoid precursor subsets that can be converted to myeloid fates by ectopic cytokine signals (Iwasaki-Arai et al., 2003) (see previous text). This correlation may suggest that ‘‘lymphoid’’ DC potential within CLPs and proT is a residual function of latent myeloid potential. It is therefore important to characterize the signals that induce DC commitment from lymphoid progenitors. XIV. Fetal Hematopoietic Progenitors are Not Fully Committed to the Lymphoid and Myeloid Fates

Although the hematopoietic hierarchy and effector cell types produced in fetal life are similar to those in adult blood development, some phenotypic and functional differences exist (Holyoake et al., 1999; Jordan et al., 1995; Morrison et al., 1995; Pawliuk et al., 1996). For example, fetal liver HSCs can generate Vg3þ and Vg4þ T cells (Ikuta et al., 1990) and B-1a lymphocytes (Hayakawa and Hardy, 2000) while adult HSCs cannot. Using surface markers similar to those used to isolate adult progenitor populations, fetal liver counterparts have been prospectively puriﬁed. These include the fetal CLPs

36

DAVID TRAVER AND KOICHI AKASHI

(LinIL-7RaþB220/loSca-1loc-Kitlo) (Mebius et al., 2001), CMPs (LinIL7RaSca-1c-KitþAA4.1FcgRII/IIIloCD34þ), GMPs (LinIL-7RaSca1c-KitþAA4.1FcgRII/IIIhiCD34þ), and MEPs (LinIL-7RaSca-1cKitþAA4.1FcgRII/IIIloCD34) (Traver et al., 2001). Each fetal population displays similar differentiation potentials, but, as will be discussed, appear to be somewhat less restricted than their adult counterparts. Clonogenic progenitors for T cells, B cells, and macrophages were demonstrated to exist within the AA4.1þFcgRII/IIIþ fraction in mouse fetal liver (Lacaud et al., 1998). Cumano et al. demonstrated clonogenic B cell and macrophage progenitors within a fetal liver AA4.1þB220Mac-1Sca-1þ fraction, although T cell differentiation potential was not tested (Cumano et al., 1992). B220þc-Kitþ fetal liver cells (also positive for the IL-7Ra chain) differentiated into T cells, B cells, and macrophages at a high frequency (Sagara et al., 1997). Interestingly, IL-7Ra expression is not a deﬁnitive marker to exclude cells with macrophage potential in fetal liver hematopoiesis. A fetal counterpart to adult CLPs, IL-7RaþB220/loSca-1loc-Kitlo cells, comprises 0.5 to 1.2% of E12.5 to E14.5 fetal liver cells. These cells are positive for AA4.1 but negative for FcgRII/III, indicating that there is no overlap with the cells reported by Lacaud et al. (1998). The populations reported by Cumano et al. (1992) and Sagara et al. (1997) should include IL-7RaþB220/loSca-1locKitlo cells. In vitro culture of fetal CLPs on S17 stromal layers showed that 5% of single cells differentiated into macrophages and B cells (Mebius et al., 2001). This population never gave rise to myeloerythroid cells other than macrophages. The burst-size of macrophage differentiation appeared minimal (Mebius et al., 2001). Accordingly, injection of fetal CLPs into the livers (Domen et al., 1998) of sublethally irradiated newborn mice showed, T, B, and NK cell-restricted differentiation without detectable macrophage progeny (Mebius et al., 2001). Taken together, these ﬁndings suggest that macrophage potential is maintained following commitment into the T/B lymphoid lineages during fetal life. In early myelopoiesis, fetal liver CMPs, which form all myeloerythroid colony types, show a relatively high propensity to differentiate into B cells (Traver et al., 2001). In limiting dilution assays, B cell frequency was 0.8% from fetal CMPs. The B cell potential in fetal CMPs has not been tested at the single-cell level and, therefore, this still could be from a minor contaminant of B cell progenitors in the phenotypic CMP fraction. T cell developmental potential, however, was completely absent as no donor-derived progeny were found following intrathymic injection of 10,000 fetal CMPs. Thus, in the model established by prospective isolation of fetal counterparts of adult progenitors, fetal CMPs and CLPs are placed downstream of HSCs as in adult hematopoiesis, allowing them to maintain a minor potential for B cell and macrophage potentials, respectively (Mebius et al., 2001; Traver et al., 2001) (Fig. 7A).

LINEAGE COMMITMENT

37

Fig 7 A developmental model in fetal liver hematopoiesis. (A) A developmental model based on puriﬁcation of phenotypic counterparts of adult CMPs and CLPs in the fetal liver (FL). In this model, B cell potential in fetal CMPs and macrophage potential in fetal CLPs represent their relatively incomplete lineage restriction.

The differentiation potentials of single fetal liver LinSca-1þc-Kitþ HSCs and IL-7RaþLinSca-1þc-Kitþ lymphoid progenitors were carefully evaluated by high-oxygen FTOC in the presence of IL-3, IL-7, and SLF (Kawamoto et al., 1997, 2000). This has been termed the multilineage progenitor (MLP) assay and is powerful since it can support myelomonocytic and lymphoid fates from fetal liver progenitors. Unfortunately, this assay is not suitable to evaluate megakaryocyte/erythrocyte potential. Throughout these experiments, single progenitors giving rise only to T and B cells were never found, although T/GM, B/GM, or T/B/GM outcomes were detectable (Kawamoto et al., 1997). Myeloid progeny in this culture system mainly consisted of macrophages, with minor populations of dendritic cells and granulocytes produced. These data led to the proposal of a unique hematopoietic differentiation model that does not include the CLP stage (Katsura, 2002) (Fig. 7B). In this model, T and B lymphoid development is always associated with myelomonocytic development, and originates from the common myelolymphoid progenitor (CMMP) (Fig. 7B). How can we reconcile this model with the conventional developmental scheme based on prospective puriﬁcation studies? Retrospective studies such as these base upstream developmental sequences on functional endpoints such as colony formation. The ensuing developmental models can only be justiﬁed if the assay system is fully permissive for all fate potentials of plated stem and progenitor cells. In the MLP assay system, only 5% of single fetal liver HSCs exhibited multilineage (T/B/GM) readouts. Nonetheless, the fact that T/B only differentiation was never found from single cells raises the possibility that CLPs are not a requisite step in the generation of fetal lymphocytes. As has been discussed, it is highly likely that macrophage potential is well preserved along the early lymphoid pathway in

38

DAVID TRAVER AND KOICHI AKASHI

the fetal liver (Mebius et al., 2001). This may account for the relatively high incidence of lymphoid/myeloid differentiation from fetal progenitors in the MLP assay. Interestingly, the high oxygen concentration used in the MLP assay dramatically promotes macrophage differentiation (Anderson et al., 2002). Therefore, the macrophage potentials of fetal liver cells could also be overestimated by this assay. Another group has subsequently reported that single, adult CLPs frequently gave rise to both T and B but not myeloid cells in the MLP assay (Izon et al., 2001), highlighting an apparent difference between fetal and adult CLP subsets. It will be interesting to determine the developmental boundaries at which the relatively incomplete lineage restriction observed in fetal progenitors become restricted to the lymphoid and myeloerythroid fates. XV. Lymphoid- and Myeloid-restricted Progenitors in Human Bone Marrow

In human hematopoiesis, similar progenitor populations have been isolated by cell surface phenotypes. A population having signiﬁcant CLP activity has been reported to exist within the LinCD34þCD10þ subset of human bone marrow cells (Galy et al., 1995). Lymphoid potential was determined by reconstitution of human bone and thymus fragments implanted into SCID mice as well as by in vitro culture systems. LinCD34þCD10þThy-1þ multipotent HSC are CD45RA, whereas CD45RAþ LinCD34þCD10þThy-1 CLPs differentiated only into T, B, NK, and dendritic cells (Galy et al., 1995). Similar to mouse CLPs, this population expresses IL-7Ra. Many studies support the existence of human oligopotent myeloid progenitors (de Wynter et al., 2001; Fritsch et al., 1993; Huang et al., 1999; Lansdorp et al., 1990; Olweus et al., 1996, 1997; Rappold et al., 1997). Human counterparts of murine CMPs, GMPs, and MEPs were isolated from human bone marrow and cord blood cells (Manz et al., 2002). All are negative for multiple mature lineage markers, including the early lymphoid markers CD7, CD10, and IL-7Ra, and all are CD34þCD38þ. This CD34þCD38þ fraction was divided by the expression of CD45RA, an isoform of CD45 that can negatively regulate at least some classes of cytokine receptor signaling (Irie-Sasaki et al., 2001), and IL-3Ra, a receptor that, upon activation, supports proliferation and differentiation of primitive progenitors (Kimura et al., 1997). CD45RAIL3Ralo (CMPs), CD45RAþIL-3Ralo (GMPs), and CD45RAIL-3Ra (MEPs) formed distinct myeloid colony types according to their deﬁnitions at a high frequency, but possess little or no LTC-IC activity. The FcgRII/III (CD16/ CD32) that distinguishes mouse CMPs and GMPs was not detectable on human myeloid progenitor populations. CMPs give rise to MEPs and GMPs in vitro and a signiﬁcant proportion of CMPs were demonstrated to possess clonal granulocyte/macrophage and megakaryocyte/erythrocyte potential

LINEAGE COMMITMENT

39

(Manz et al., 2002). Thus, the hierarchical progenitor relationships demonstrated in the mouse also exist in human hematopoiesis and can be prospectively isolated by cell surface phenotype. XVI. Clinical Relevance

Bone marrow transplantation (BMT) into myeloablated individuals often gives rise to graft versus host disease (GVHD) when using total bone marrow or mobilized peripheral blood cells. This is due to the transfer of allogeneic lymphocytes within the transplanted cells. It has been shown that transplantation of puriﬁed CD34þ HSCs alleviates the vast majority of GVHD, presumably because donor-derived lymphocytes can be educated in the host thymus to prevent alloreactivity (Link et al., 1996). Pure HSC transplants, however, are relatively slow to generate sufﬁcient numbers of mature cell types. Using a mouse progenitor transplantation model, it has been shown that high-dose transplantation of either CMPs or MEPs was sufﬁcient for radioprotection over a 1-month interval (Na Nakorn et al., 2002). After this time, rare surviving host HSCs recovered to produce all blood cell subsets (Na Nakorn et al., 2002). Another important complication following BMT is infection. Myeloablative irradiation and/or cytoreductive drugs cause a rapid disappearance of myelomonocytes as well as lymphocytes. This leaves a window following transplantation for opportunistic pathogens such as Aspergillus fumigatus, Pseudomonas aeruginos, and cytomegaloviruses to ﬂourish. Using another mouse model, it was shown that transplantation of CMPs and/or GMPs in conjunction with HSC transplants could protect against otherwise lethal challenges of either pathogen due to increased numbers of myelomonocytic cells (BitMansour et al., 2002). Transplantation of CLPs in conjunction with HSCs also protected mice against murine CMV infection (Arber et al., 2003). Based on these preclinical ﬁndings, it may be of clinical interest to purify human hematopoietic progenitor subsets to augment puriﬁed HSC transplants. XVII. Conclusion

In this chapter, we have provided a current understanding of hematolymphoid development from rare HSCs that give rise to progenitor subsets that progressively lose fate potential as they differentiate down their respective lineages. The ability to prospectively isolate the major branchpoints along the hematopoietic tree allows a molecular proﬁling of each, both at the population and single-cell level. Transcriptional proﬁling at the population level has shown that the ‘‘master regulator’’ genes identiﬁed from mouse knockout studies are expressed in the progenitor subsets upstream of the noted defective lineages. Single-cell proﬁling has supported the hypothesis of ‘‘priming stages,’’ whereby

40

DAVID TRAVER AND KOICHI AKASHI

promiscuous, low-level expression of many genes appears to maintain ﬂexibility in cell fate choices. Interestingly, enforced expression of single, inappropriate genes within progenitors normally exclusively commited to the lymphoid fates can reprogram myeloerythroid outcomes. This demonstrates that the nucleus maintains developmental ﬂexibility and that the progressive loss of fate potential upon differentiation is likely controlled initially by strict regulation of growth factor receptor expression. This ﬁnding may explain the clinical observations of lineage inﬁdelity and mixed-lineage leukemias. Future work with hematopoietic progenitor populations should allow a more precise understanding of the molecular mechanisms of lineage commitment and may help elucidate how stem cells choose between self-renewal and differentiation. References Abramson, S., Miller, R. G., and Phillips, R. A. (1977). The identiﬁcation in adult bone marrow of pluripotent and restricted stem cells of the myeloid and lymphoid systems. J. Exp. Med. 145, 1567–1579. Adolfsson, J., Borge, O. J., Bryder, D., Theilgaard-Monch, K., Astrand-Grundstrom, I., Sitnicka, E., Sasaki, Y., and Jacobsen, S. E. (2001). Upregulation of Flt3 expression within the bone marrow Lin()Sca1(þ)c-kit(þ) stem cell compartment is accompanied by loss of self-renewal capacity. Immunity 15, 659–669. Aifantis, I., Azogui, O., Feinberg, J., Saint-Ruf, C., Buer, J., and von Boehmer, H. (1998). On the role of the pre-T cell receptor in alphabeta versus gammadelta T lineage commitment. Immunity 9, 649–655. Akashi, K., Harada, M., Shibuya, T., Fukagawa, K., Kimura, N., Sagawa, K., Yoshikai, Y., Teshima, T., Kikuchi, M., and Niho, Y. (1991). Simultaneous occurrence of myelomonocytic leukemia and multiple myeloma: Involvement of common leukemic progenitors and their developmental abnormality of ‘‘lineage inﬁdelity.’’ J. Cell Phys. 148, 446–456. Akashi, K., He, X., Chen, J., Iwasaki, H., Niu, C., Steenhard, B., Zhang, J., Haug, J., and Li, L. (2003). Transcriptional accessibility for genes of multiple tissues and hematopoietic lineages is hierarchically controlled during early hematopoiesis. Blood 101, 383–389. Akashi, K., Kondo, M., and Weissman, I. L. (1998). Role of interleukin-7 in T-cell development from hematopoietic stem cells. Immunol. Rev. 165, 13–28. Akashi, K., Kondo, M., von Freeden-Jeffry, U., Murray, R., and Weissman, I. L. (1997). Bcl-2 rescues T lymphopoiesis in interleukin-7 receptor-deﬁcient mice. Cell 89, 1033–1041. Akashi, K., Traver, D., Miyamoto, T., and Weissman, I. L. (2000). A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature 404, 193–197. Allman, D., Li, J., and Hardy, R. R. (1999). Commitment to the B lymphoid lineage occurs before DH–JH recombination. J. Exp. Med. 189, 735–740. Allman, D., Sambandam, A., Kim, S., Miller, J. P., Pagan, A., Well, D., Meraz, A., and Bhandoola, A. (2003). Thymopoiesis independent of common lymphoid progenitors. Nat. Immunol. 4, 168–174. Anderson, A. C., Robey, E. A., and Huang, Y. H. (2001). Notch signaling in lymphocyte development. Cur. Op. Gen. 11, 554–560. Anderson, M. K., Weiss, A. H., Hernandez-Hoyos, G., Dionne, C. J., and Rothenberg, E. V. (2002). Constitutive expression of PU.1 in fetal hematopoietic progenitors blocks T cell development at the pro-T cell stage. Immunity 16, 285–296.

LINEAGE COMMITMENT

41

Arber, C., BitMansour, A., Sparer, T. E., Higgins, J. P., Mocarski, E. S., Weissman, I. L., Shizuru, J. A., and Brown, J. M. Y. (2003). Common lymphoid progenitors rapidly engraft and protect against lethal murine cytomegalovirus infection after hematopoietic stem cell transplantation. Blood 2002, 2012–3834. Ardavin, C., Wu, L., Li, C.-L., and Shortman, K. (1993). Thymic dendritic cells and T cells develop simultaneously in the thymus from a common precursor population. Nature 362, 761–763. Asselin-Paturel, C., Boonstra, A., Dalod, M., Durand, I., Yessaad, N., Dezutter-Dambuyant, C., Vicari, A., O’Garra, A., Biron, C., Briere, F., and Trinchieri, G. (2001). Mouse type I IFNproducing cells are immature APCs with plasmacytoid morphology. Nat. Immunol. 2, 1144–1150. Banchereau, J., and Steinman, R. M. (1998). Dendritic cells and the control of immunity. Nature 392, 245–252. Becker, A., McCulloch, E., and Till, J. (1963). Cytological demonstration of the clonal nature of spleen colonies dervived from transplanted mouse marrow cells. Nature 197, 452–454. Bendriss-Vermare, N., Barthelemy, C., Durand, I., Bruand, C., Dezutter-Dambuyant, C., Moulian, N., Berrih-Aknin, S., Caux, C., Trinchieri, G., and Briere, F. (2001). Human thymus contains IFN-alpha-producing CD11c(), myeloid CD11c(þ), and mature interdigitating dendritic cells. J. Clin. Inv. 107, 835–844. Berger, S. L., and Felsenfeld, G. (2001). Chromatin goes global. Mol. Cell 8, 263–268. Bhatia, S. K., Tygrett, L. T., Grabstein, K. H., and Waldschmidt, T. J. (1995). The effect of in vivo IL-7 deprivation on T cell maturation. J. Exp. Med. 181, 1399–1409. BitMansour, A., Burns, S. M., Traver, D., Akashi, K., Contag, C. H., Weissman, I. L., and Brown, J. M. Y. (2002). Myeloid progenitors protect against invasive aspergillosis and Pseudomonas aeruginosa infection following hematopoietic stem cell transplantation. Blood 100, 4660–4667. Bjorck, P. (2001). Isolation and characterization of plasmacytoid dendritic cells from Flt3 ligand and granulocyte-macrophage colony-stimulating factor-treated mice. Blood 98, 3520–3526. Boyd, A. W., and Schrader, J. W. (1982). Derivation of macrophage-like lines from the pre-B lymphoma ABLS 8.1 using 5-azacytidine. Nature 297, 691–693. Bradley, T. R., and Metcalf, D. (1966). The growth of mouse bone marrow cells in vitro. Aust. J. Exp. Biol. Med. Sci. 44, 287–299. Brown, G., Bunce, C. M., and Guy, G. R. (1985). Sequential determination of lineage potentials during haemopoiesis. Br. J. Cancer 52, 681–686. Brown, G., Bunce, C. M., Howie, A. J., and Lord, J. M. (1987). Stochastic or ordered lineage commitment during hemopoiesis? Leukemia 1, 150–153. Busslinger, M., Nutt, S. L., and Rolink, A. G. (2000). Lineage commitment in lymphopoiesis. Curr. Op. Im. 12, 151–158. Cao, X., Shores, E. W., Hu-Li, J., Anver, M. R., Kelsall, B. L., Russell, S. M., Drago, J., Noguchi, M., Grinberg, A., Bloom, E. T., et al. (1995). Defective lymphoid development in mice lacking expression of the common cytokine receptor gamma chain. Immunity 2, 223–238. Caux, C., Vanbervliet, B., Massacrier, C., Dezutter-Dambuyant, C., de Saint-Vis, B., Jacquet, C., Yoneda, K., Imamura, S., Schmitt, D., and Banchereau, J. (1996). CD34þ hematopoietic progenitors from human cord blood differentiate along two independent dendritic cell pathways in response to GM-CSFþTNF alpha. J. Exp. Med. 184, 695–706. Cella, M., Jarrossay, D., Facchetti, F., Alebardi, O., Nakajima, H., Lanzavecchia, A., and Colonna, M. (1999). Plasmacytoid monocytes migrate to inﬂamed lymph nodes and produce large amounts of type I interferon. Nat. Med. 5, 919–923. Chen, H., Ray-Gallet, D., Zhang, P., Hetherington, C. J., Gonzalez, D. A., Zhang, D. E., MoreauGachelin, F., and Tenen, D. G. (1995). PU.1 (Spi-1) autoregulates its expression in myeloid cells. Oncogene 11, 1549–1560.

42

DAVID TRAVER AND KOICHI AKASHI

Cheng, G. Y., Minden, M. D., Toyonaga, B., Mak, T. W., and McCulloch, E. A. (1986). T cell receptor and immunoglobulin gene rearrangements in acute myeloblastic leukemia. J. Exp. Med. 163, 414–424. Chiang, M. Y., and Monroe, J. G. (1999). BSAP/Pax5A expression blocks survival and expansion of early myeloid cells implicating its involvement in maintaining commitment to the B-lymphocyte lineage. Blood 94, 3621–3632. Christensen, J. L., and Weissman, I. L. (2001). Flk-2 is a marker in hematopoietic stem cell differentiation: A simple method to isolate long-term stem cells. Proc. Natl. Acad. Sci. USA 98, 14541–14546. Colucci, F., Samson, S. I., DeKoter, R. P., Lantz, O., Singh, H., and Di Santo, J. P. (2001). Differential requirement for the transcription factor PU.1 in the generation of natural killer cells versus B and T cells. Blood 97, 2625–2632. Corcoran, A. E., Riddell, A., Krooshoop, D., and Venkitaraman, A. R. (1998). Impaired immunoglobulin gene rearrangement in mice lacking the IL-7 receptor. Nature 391, 904–907. Cotta, C. V., Zhang, Z., Kim, H. G., and Klug, C. A. (2003). Pax5 determines B versus T cell fate and does not block early myeloid-lineage development. Blood 101, 4342–4346. Cross, M. A., and Enver, T. (1997). The lineage commitment of haemopoietic progenitor cells. Cur. Op. Gen. 7, 609–613. Cumano, A., Paige, C. J., Iscove, N. N., and Brady, G. (1992). Bipotential precursors of B cells and macrophages in murine fetal liver. Nature 356, 612–615. de Wynter, E. A., Heyworth, C. M., Mukaida, N., Jaworska, E., Weffort-Santos, A., Matushima, K., and Testa, N. G. (2001). CCR1 chemokine receptor expression isolates erythroid from granulocyte-macrophage progenitors. J. Leukoc. Biol. 70, 455–460. Deftos, M. L., Huang, E., Ojala, E. W., Forbush, K. A., and Bevan, M. J. (2000). Notch1 signaling promotes the maturation of CD4 and CD8 SP thymocytes. Immunity 13, 73–84. DeKoter, R. P., and Singh, H. (2000). Regulation of B lymphocyte and macrophage development by graded expression of PU.1. Science 288, 1439–1441. Delassus, S., Titley, I., and Enver, T. (1999). Functional and molecular analysis of hematopoietic progenitors derived from the aorta–gonad–mesonephros region of the mouse embryo. Blood 94, 1495–1503. Dexter, T. M., and Testa, N. G. (1980). In vitro methods in haemopoiesis and lymphopoiesis. J. Immunol. M. 38, 177–190. Doi, H., Inaba, M., Yamamoto, Y., Taketani, S., Mori, S. I., Sugihara, A., Ogata, H., Toki, J., Hisha, H., Inaba, K., et al. (1997). Pluripotent hemopoietic stem cells are c-kit
LINEAGE COMMITMENT

43

Fritsch, G., Buchinger, P., Printz, D., Fink, F. M., Mann, G., Peters, C., Wagner, T., Adler, A., and Gadner, H. (1993). Rapid discrimination of early CD34þ myeloid progenitors using CD45-RA analysis. Blood 81, 2301–2309. Fujiwara, Y., Browne, C. P., Cunniff, K., Goff, S. C., and Orkin, S. H. (1996). Arrested development of embryonic red cell precursors in mouse embryos lacking transcription factor GATA-1. Proc. Natl. Acad. Sci. USA 93, 12355–12358. Fulop, G. M., and Phillips, R. A. (1989). Use of scid mice to identify and quantitate lymphoidrestricted stem cells in long-term bone marrow cultures. Blood 74, 1537–1544. Galy, A., Travis, M., Cen, D., and Chen, B. (1995). Human T, B, natural killer, and dendritic cells arise from a common bone marrow progenitor cell subset. Immunity 3, 459–473. Gillessen, S., Mach, N., Small, C., Mihm, M., and Dranoff, G. (2001). Overlapping roles for granulocyte-macrophage colony-stimulating factor and interleukin-3 in eosinophil homeostasis and contact hypersensitivity. Blood 97, 922–928. Giri, J. G., Ahdieh, M., Eisenman, J., Shanebeck, K., Grabstein, K., Kumaki, S., Namen, A., Park, L. S., Cosman, D., and Anderson, D. (1994). Utilization of the beta and gamma chains of the IL-2 receptor by the novel cytokine IL-15. EMBO J. 13, 2822–2830. Goodell, M. A., Jackson, K. A., Majka, S. M., Mi, T., Wang, H., Pocius, J., Hartley, C. J., Majesky, M. W., Entman, M. L., Michael, L. H., and Hirschi, K. K. (2001). Stem cell plasticity in muscle and bone marrow. Ann. NY Acad. Sci. 938, 208–218; discussion 218–220. Goodell, M. A., Rosenzweig, M., Kim, H., Marks, D. F., DeMaria, M., Paradis, G., Grupp, S. A., Sieff, C. A., Mulligan, R. C., and Johnson, R. P. (1997). Dye efﬂux studies suggest that hematopoietic stem cells expressing low or undetectable levels of CD34 antigen exist in multiple species. Nat. Med. 3, 1337–1345. Gounari, F., Aifantis, I., Martin, C., Fehling, H. J., Hoeﬂinger, S., Leder, P., von Boehmer, H., and Reizis, B. (2002). Tracing lymphopoiesis with the aid of a pTalpha-controlled reporter gene. Nat. Immunol. 3, 489–496. Graf, T. (2002). Differentiation plasticity of hematopoietic cells. Blood 99, 3089–3101. Groettrup, M., Ungewiss, K., Azogui, O., Palacios, R., Owen, M. J., Hayday, A. C., and von Boehmer, H. (1993). A novel disulﬁde-linked heterodimer on pre-T cells consists of the T cell receptor beta chain and a 33 kd glycoprotein. Cell 75, 283–294. Grogan, T. M., Durie, B. G., Spier, C. M., Richter, L., and Vela, E. (1989). Myelomonocytic antigen positive multiple myeloma. Blood 73, 763–769. Grouard, G., Rissoan, M. C., Filgueira, L., Durand, I., Banchereau, J., and Liu, Y. J. (1997). The enigmatic plasmacytoid T cells develop into dendritic cells with interleukin (IL)-3 and CD40ligand. J. Exp. Med. 185, 1101–1111. Guerriero, A., Langmuir, P. B., Spain, L. M., and Scott, E. W. (2000). PU 1 is required for myeloidderived but not lymphoid-derived dendritic cells. Blood 95, 879–885. Hacker, C., Kirsch, R. D., Ju, X. S., Hieronymus, T., Gust, T. C., Kuhl, C., Jorgas, T., Kurz, S. M., Rose-John, S., Yokota, Y., and Zenke, M. (2003). Transcriptional proﬁling identiﬁes Id2 function in dendritic cell development. Nat. Immunol. 4, 380–386. Hao, Q. L., Zhu, J., Price, M. A., Payne, K. J., Barsky, L. W., and Crooks, G. M. (2001). Identiﬁcation of a novel, human multilymphoid progenitor in cord blood. Blood 97, 3683–3690. Hardy, R. R., Carmack, C. E., Shinton, S. A., Kemp, J. D., and Hayakawa, K. (1991). Resolution and characterization of pro-B and pre-pro-B cell stages in normal mouse bone marrow. J. Exp. Med. 173, 1213–1225. Hart, D. N. (1997). Dendritic cells: Unique leukocyte populations which control the primary immune response. Blood 90, 3245–3287. Hasserjian, R. P., Aster, J. C., Davi, F., Weinberg, D. S., and Sklar, J. (1996). Modulated expression of notch1 during thymocyte development. Blood 88, 970–976.

44

DAVID TRAVER AND KOICHI AKASHI

Hayakawa, K., and Hardy, R. R. (2000). Development and function of B-1 cells. Curr. Op. Im. 12, 346–353. Hayashi, S., Kunisada, T., Ogawa, M., Sudo, T., Kodama, H., Suda, T., and Nishikawa, S. (1990). Stepwise progression of B lineage differentiation supported by interleukin 7 and other stromal cell molecules. J. Exp. Med. 171, 1683–1695. Heemskerk, M. H., Blom, B., Nolan, G., Stegmann, A. P., Bakker, A. Q., Weijer, K., Res, P. C., and Spits, H. (1997). Inhibition of T cell and promotion of natural killer cell development by the dominant negative helix loop helix factor Id3. J. Exp. Med. 186, 1597–1602. Heimfeld, S., Hudak, S., Weissman, I., and Rennick, D. (1991). The in vitro response of phenotypically deﬁned mouse stem cells and myeloerythroid progenitors to single or multiple growth factors. Proc. Natl. Acad. Sci. USA 88, 9902–9906. Heyworth, C., Pearson, S., May, G., and Enver, T. (2002). Transcription factor-mediated lineage switching reveals plasticity in primary committed progenitor cells. EMBO J. 21, 3770–3781. Hirschhorn, R. (1990). Adenosine deaminase deﬁciency. Immunodeﬁc. Rev. 2, 175–198. Ho, I. C., Vorhees, P., Marin, N., Oakley, B. K., Tsai, S. F., Orkin, S. H., and Leiden, J. M. (1991). Human GATA-3: A lineage-restricted transcription factor that regulates the expression of the T cell receptor alpha gene. EMBO J. 10, 1187–1192. Holyoake, T. L., Nicolini, F. E., and Eaves, C. J. (1999). Functional differences between transplantable human hematopoietic stem cells from fetal liver, cord blood, and adult marrow. Exp. Hematol. 27, 1418–1427. Horcher, M., Souabni, A., and Busslinger, M. (2001). Pax5/BSAP maintains the identity of B cells in late B lymphopoiesis. Immunity 14, 779–790. Hu, M., Krause, D., Greaves, M., Sharkis, S., Dexter, M., Heyworth, C., and Enver, T. (1997). Multilineage gene expression precedes commitment in the hemopoietic system. Gene Dev. 11, 774–785. Huang, S., Chen, Z., Yu, J. F., Young, D., Bashey, A., Ho, A. D., and Law, P. (1999). Correlation between IL-3 receptor expression and growth potential of human CD34þ hematopoietic cells from different tissues. Stem Cells 17, 265–272. Igarashi, H., Gregory, S. C., Yokota, T., Sakaguchi, N., and Kincade, P. W. (2002). Transcription from the RAG1 locus marks the earliest lymphocyte progenitors in bone marrow. Immunity 17, 117–130. Igarashi, H., Kuwata, N., Kiyota, K., Sumita, K., Suda, T., Ono, S., Bauer, S. R., and Sakaguchi, N. (2001). Localization of recombination activating gene 1/green ﬂuorescent protein (RAG1/GFP) expression in secondary lymphoid organs after immunization with T-dependent antigens in rag1/gfp knockin mice. Blood 97, 2680–2687. 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 hematopoietic stem cells. Cell 62, 863–874. Inaba, K., Inaba, M., Deguchi, M., Hagi, K., Yasumizu, R., Ikehara, S., Muramatsu, S., and Steinman, R. M. (1993). Granulocytes, macrophages, and dendritic cells arise from a common major histocompatibility complex class II-negative progenitor in mouse bone marrow. Proc. Natl. Acad. Sci. USA 90, 3038–3042. Irie-Sasaki, J., Sasaki, T., Matsumoto, W., Opavsky, A., Cheng, M., Welstead, G., Grifﬁths, E., Krawczyk, C., Richardson, C. D., Aitken, K., et al. (2001). CD45 is a JAK phosphatase and negatively regulates cytokine receptor signalling. Nature 409, 349–354. Ito, C. Y., Li, C. Y., Bernstein, A., Dick, J. E., and Stanford, W. L. (2003). Hematopoietic stem cell and progenitor defects in Sca-1/Ly-6A-null mice. Blood 101, 517–523. Iwasaki, H., and Akashi, K. (2001). Mast cells originate from granulocyte/macrophage-committed progenitors. Blood 98, 275a.

LINEAGE COMMITMENT

45

Iwasaki, H., Mizuno, S., and Akashi, K. (2002). GATA-1 converts lymphoid and myelomonocytic progenitors into the megakaryocyte/erythroid lineages. Blood 100, 258a. Iwasaki-Arai, J., Iwasaki, H., Miyamoto, T., Watanabe, S., and Akashi, K. (2003). Enforced GMCSF signals do not support lymphopoiesis, but instruct lymphoid to myelomonocytic lineage conversion. J. Exp. Med. 197, 1311–1322. Izon, D., Rudd, K., DeMuth, W., Pear, W. S., Clendenin, C., Lindsley, R. C., and Allman, D. (2001). A common pathway for dendritic cell and early B cell development. J. Immunol. 167, 1387–1392. Jackson, K. A., Majka, S. M., Wang, H., Pocius, J., Hartley, C. J., Majesky, M. W., Entman, M. L., Michael, L. H., Hirschi, K. K., and Goodell, M. A. (2001). Regeneration of ischemic cardiac muscle and vascular endothelium by adult stem cells. J. Clin. Inv. 107, 1395–1402. Jaleco, A. C., Neves, H., Hooijberg, E., Gameiro, P., Clode, N., Haury, M., Henrique, D., and Parreira, L. (2001). Differential effects of Notch ligands Delta-1 and Jagged-1 in human lymphoid differentiation. J. Exp. Med. 194, 991–1002. Jaleco, A. C., Stegmann, A. P., Heemskerk, M. H., Couwenberg, F., Bakker, A. Q., Weijer, K., and Spits, H. (1999). Genetic modiﬁcation of human B-cell development: B-cell development is inhibited by the dominant negative helix loop helix factor Id3. Blood 94, 2637–2646. Jarrossay, D., Napolitani, G., Colonna, M., Sallusto, F., and Lanzavecchia, A. (2001). Specialization and complementarity in microbial molecule recognition by human myeloid and plasmacytoid dendritic cells. Eur. J. Immun. 31, 3388–3393. Jehn, B. M., Bielke, W., Pear, W. S., and Osborne, B. A. (1999). Cutting edge: Protective effects of notch-1 on TCR-induced apoptosis. J. Immunol. 162, 635–638. Jimenez, G., Grifﬁths, S. D., Ford, A. M., Greaves, M. F., and Enver, T. (1992). Activation of the beta-globin locus control region precedes commitment to the erythroid lineage. Proc. Natl. Acad. Sci. USA 89, 10618–10622. Johnson, G. R., and Metcalf, D. (1977). Pure and mixed erythroid colony formation in vitro stimulated by spleen-conditioned medium with no detectable erythropoietin. Proc. Natl. Acad. Sci. USA 74, 3879–3882. Jordan, C. T., Astle, C. M., Zawadzki, J., Mackarehtschian, K., Lemischka, I. R., and Harrison, D. E. (1995). Long-term repopulating abilities of enriched fetal liver stem cells measured by competitive repopulation. Exp. Hematol. 23, 1011–1015. Kadowaki, N., Antonenko, S., Lau, J. Y., and Liu, Y. J. (2000). Natural interferon alpha/betaproducing cells link innate and adaptive immunity. J. Exp. Med. 192, 219–226. Kadowaki, N., Ho, S., Antonenko, S., Malefyt, R. W., Kastelein, R. A., Bazan, F., and Liu, Y. J. (2001). Subsets of human dendritic cell precursors express different toll-like receptors and respond to different microbial antigens. J. Exp. Med. 194, 863–869. Kaneta, M., Osawa, M., Sudo, K., Nakauchi, H., Farr, A. G., and Takahama, Y. (2000). A role for pref-1 and HES-1 in thymocyte development. J. Immunol. 164, 256–264. Karasuyama, H., Rolink, A., Shinkai, Y., Young, F., Alt, F. W., and Melchers, F. (1994). The expression of Vpre-B/lambda 5 surrogate light chain in early bone marrow precursor B cells of normal and B cell-deﬁcient mutant mice. Cell 77, 133–143. Katsura, Y. (2002). Redeﬁnition of lymphoid progenitors. Nat. Rev. Immunol. 2, 127–132. Kawamoto, H., Ikawa, T., Ohmura, K., Fujimoto, S., and Katsura, Y. (2000). T cell progenitors emerge earlier than B cell progenitors in the murine fetal liver. Immunity 12, 441–450. Kawamoto, H., Ohmura, K., and Katsura, Y. (1997). Direct evidence for the commitment of hematopoietic stem cells to T, B, and myeloid lineages in murine fetal liver. Int. Immunol. 9, 1011–1019. Kimura, T., Sakabe, H., Tanimukai, S., Abe, T., Urata, Y., Yasukawa, K., Okano, A., Taga, T., Sugiyama, H., Kishimoto, T., and Sonoda, Y. (1997). Simultaneous activation of signals through

46

DAVID TRAVER AND KOICHI AKASHI

gp130, c-kit, and interleukin-3 receptor promotes a trilineage blood cell production in the absence of terminally acting lineage-speciﬁc factors. Blood 90, 4767–4778. King, A. G., Kondo, M., Scherer, D. C., and Weissman, I. L. (2002). Lineage inﬁdelity in myeloid cells with TCR gene rearrangement: A latent developmental potential of proT cells revealed by ectopic cytokine receptor signaling. Proc. Natl. Acad. Sci. USA 99, 4508–4513. Klinken, S. P., Alexander, W. S., and Adams, J. M. (1988). Hemopoietic lineage switch: V-raf oncogene converts Emu-myc transgenic B cells into macrophages. Cell 53, 857–867. Kondo, M., Scherer, D. C., Miyamoto, T., King, A. G., Akashi, K., Sugamura, K., and Weissman, I. L. (2000). Cell-fate conversion of lymphoid-committed progenitors by instructive actions of cytokines. Nature 407, 383–386. Kondo, M., Takeshita, T., Higuchi, M., Nakamura, M., Sudo, T., Nishikawa, S., and Sugamura, K. (1994). Functional participation of the IL-2 receptor gamma chain in IL-7 receptor complexes. Science 263, 1453–1454. Kondo, M., Weissman, I. L., and Akashi, K. (1997). Identiﬁcation of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91, 661–672. Kontaraki, J., Chen, H. H., Riggs, A., and Bonifer, C. (2000). Chromatin ﬁne structure proﬁles for a developmentally regulated gene: Reorganization of the lysozyme locus before trans-activator binding and gene expression. Gene Dev. 14, 2106–2122. Krause, D. S., Theise, N. D., Collector, M. I., Henegariu, O., Hwang, S., Gardner, R., Neutzel, S., and Sharkis, S. J. (2001). Multi-organ, multi-lineage engraftment by a single bone marrowderived stem cell. Cell 105, 369–377. Kulessa, H., Frampton, J., and Graf, T. (1995). GATA-1 reprograms avian myelomonocytic cell lines into eosinophils, thromboblasts, and erythroblasts. Gene Dev. 9, 1250–1262. Kuwata, N., Igarashi, H., Ohmura, T., Aizawa, S., and Sakaguchi, N. (1999). Cutting edge: Absence of expression of RAG1 in peritoneal B-1 cells detected by knocking into RAG1 locus with green ﬂuorescent protein gene. J. Immunol. 163, 6355–6359. Lacaud, G., Carlsson, L., and Keller, G. (1998). Identiﬁcation of a fetal hematopoietic precursor with B cell, T cell, and macrophage potential. Immunity 9, 827–838. Lagasse, E., Connors, H., Al-Dhalimy, M., Reitsma, M., Dohse, M., Osborne, L., Wang, X., Finegold, M., Weissman, I. L., and Grompe, M. (2000). Puriﬁed hematopoietic stem cells can differentiate into hepatocytes in vivo. Nat. Med. 6, 1229–1234. Lansdorp, P. M., Sutherland, H. J., and Eaves, C. J. (1990). Selective expression of CD45 isoforms on functional subpopulations of CD34þ hemopoietic cells from human bone marrow. J. Exp. Med. 172, 363–366. Lemischka, I. (2002). Rethinking somatic stem cell plasticity. Nat. Biotechnol. 20, 425. Li, Y. S., Wasserman, R., Hayakawa, K., and Hardy, R. R. (1996). Identiﬁcation of the earliest B lineage stage in mouse bone marrow. Immunity 5, 527–535. Lieschke, G. J., Grail, D., Hodgson, G., Metcalf, D., Stanley, E., Cheers, C., Fowler, K. J., Basu, S., Zhan, Y. F., and Dunn, A. R. (1994). Mice lacking granulocyte colony-stimulating factor have chronic neutropenia, granulocyte, and macrophage progenitor cell deﬁciency, and impaired neutrophil mobilization. Blood 84, 1737–1746. Lindeman, G. J., Adams, J. M., Cory, S., and Harris, A. W. (1994). B-lymphoid to granulocytic switch during hematopoiesis in a transgenic mouse strain. Immunity 1, 517–527. Link, H., Arseniev, L., Bahre, O., Kadar, J. G., Diedrich, H., and Poliwoda, H. (1996). Transplantation of allogeneic CD34þ blood cells. Blood 87, 4903–4909. Liu, Y. J. (2001). Dendritic cell subsets and lineages, and their functions in innate and adaptive immunity. Cell 106, 259–262. Mackarehtschian, K., Hardin, J. D., Moore, K. A., Boast, S., Goff, S. P., and Lemischka, I. R. (1995). Targeted disruption of the ﬂk2/ﬂt3 gene leads to deﬁciencies in primitive hematopoietic progenitors. Immunity 3, 147–161.

LINEAGE COMMITMENT

47

Maki, K., Sunaga, S., and Ikuta, K. (1996). The V–J recombination of T cell receptor–gamma genes is blocked in interleukin-7 receptor-deﬁcient mice. J. Exp. Med. 184, 2423–2427. Manz, M. G., Miyamoto, T., Akashi, K., and Weissman, I. L. (2002). Prospective isolation of human clonogenic common myeloid progenitors. Proc. Natl. Acad. Sci. USA 99, 11872–11877. Manz, M. G., Traver, D., Miyamoto, T., Weissman, I. L., and Akashi, K. (2001). Dendritic cell potentials of early lymphoid and myeloid progenitors. Blood 97, 3333–3341. Martinez del Hoyo, G., Martin, P., Arias, C. F., Marin, A. R., and Ardavin, C. (2002). CD8alphaþ dendritic cells originate from the CD8alpha dendritic cell subset by a maturation process involving CD8alpha, DEC-205, and CD24 up-regulation. Blood 99, 999–1004. Matsuzaki, Y., Gyotoku, J., Ogawa, M., Nishikawa, S., Katsura, Y., Gachelin, G., and Nakauchi, H. (1993). Characterization of c-kit positive intrathymic stem cells that are restricted to lymphoid differentiation. J. Exp. Med. 178, 1283–1292. Matthews, W., Jordan, C. T., Wiegand, G. W., Pardoll, D., and Lemischka, I. R. (1991). A receptor tyrosine kinase speciﬁc to hematopoietic stem and progenitor cell-enriched populations. Cell 65, 1143–1152. McBlane, J. F., van Gent, D. C., Ramsden, D. A., Romeo, C., Cuomo, C. A., Gellert, M., and Oettinger, M. A. (1995). Cleavage at a V(D)J recombination signal requires only RAG1 and RAG2 proteins and occurs in two steps. Cell 83, 387–395. McCulloch, E. A. (1987). Lineage inﬁdelity or lineage promiscuity? Leukemia 1, 235. McKay, R. (2002). A more astonishing hypothesis. Nat. Biotechnol. 20, 426–427. McKenna, H. J., Stocking, K. L., Miller, R. E., Brasel, K., De Smedt, T., Maraskovsky, E., Maliszewski, C. R., Lynch, D. H., Smith, J., Pulendran, B., et al. (2000). Mice lacking ﬂt3 ligand have deﬁcient hematopoiesis affecting hematopoietic progenitor cells, dendritic cells, and natural killer cells. Blood 95, 3489–3497. McKercher, S. R., Torbett, B. E., Anderson, K. L., Henkel, G. W., Vestal, D. J., Baribault, H., Klemsz, M., Feeney, A. J., Wu, G. E., Paige, C. J., and Maki, R. A. (1996). Targeted disruption of the PU.1 gene results in multiple hematopoietic abnormalities. EMBO J. 15, 5647–5658. Mebius, R. E., Miyamoto, T., Christensen, J., Domen, J., Cupedo, T., Weissman, I. L., and Akashi, K. (2001). The fetal liver counterpart of adult common lymphoid progenitors gives rise to all lymphoid lineages, CD45þCD4þCD3 cells, as well as macrophages. J. Immunol. 166, 6593–6601. Metcalf, D., Johnson, G. R., and Mandel, T. E. (1979). Colony formation in agar by multipotential hemopoietic cells. J. Cell. Phys. 98, 401–420. Miyamoto, T., Iwasaki, H., Reizis, B., Ye, M., Graf, T., Weissman, I. L., and Akashi, K. (2002). Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment. Dev. Cell 3, 137–147. Mombaerts, P., Iacomini, J., Johnson, R. S., Herrup, K., Tonegawa, S., and Papaioannou, V. E. (1992). RAG-1-deﬁcient mice have no mature B and T lymphocytes. Cell 68, 869–877. Mori, S., Shortman, K., and Wu, L. (2001). Characterization of thymus-seeding precursor cells from mouse bone marrow. Blood 98, 696–704. Morrison, S. H., Wandycz, A. M., and Weissman, I. L. (1995). The puriﬁcation and characterization of fetal liver hematopoietic cells. Proc. Natl. Acad. Sci. USA 92, 10302–10306. Morrison, S. J., Wandycz, A. M., Akashi, K., Globerson, A., and Weissman, I. L. (1996). The aging of hematopoietic stem cells. Nat. Med. 2, 1011–1016. Morrison, S. J., and Weissman, I. L. (1994). The long-term repopulating subset of hematopoietic stem cells is deterministic and isolatable by phenotype. Immunity 1, 661–673. Mulder, A. H., and Visser, J. W. M. (1987). Separation and functional analysis of bone marrow cells separated by Rhodamine-123 ﬂuorescence. Exp. Hematol. 15, 99–104.

48

DAVID TRAVER AND KOICHI AKASHI

Murphy, S. B., Stass, S., Kalwinsky, D., and Rivera, G. (1983). Phenotypic conversion of acute leukaemia from T-lymphoblastic to myeloblastic induced by therapy with 20 -deoxycoformycin. Br. J. Haematol. 55, 285–293. Na Nakorn, T., Traver, D., Weissman, I. L., and Akashi, K. (2002). Myeloerythroid-restricted progenitors are sufﬁcient to confer radioprotection and provide the majority of day 8 CFU-S. J. Clin. Inv. 109, 1579–1585. Naik, S., Vremec, D., Wu, L., O’Keeffe, M., and Shortman, K. (2003). CD8{alpha}þ mouse spleen dendritic cells do not originate from the CD8- dendritic cell subset. Blood 102, 601–604. Nakahata, T., and Ogawa, M. (1982). Identiﬁcation in culture of a class of hemopoietic colonyforming units with extensive capability to self-renew and generate multipotential hemopoietic colonies. Proc. Natl. Acad. Sci. USA 79, 3843–3847. Nakano, H., Yanagita, M., and Gunn, M. D. (2001). CD11c(þ)B220(þ)Gr-1(þ) cells in mouse lymph nodes and spleen display characteristics of plasmacytoid dendritic cells. J. Exp. Med. 194, 1171–1178. Nakorn, T. N., Miyamoto, T., and Weissman, I. L. (2003). Characterization of mouse clonogenic megakaryocyte progenitors. Proc. Natl. Acad. Sci. USA 100, 205–210. Nerlov, C., and Graf, T. (1998). PU.1 induces myeloid lineage commitment in multipotent hematopoietic progenitors. Gene Dev. 12, 2403–2412. Nerlov, C., McNagny, K. M., Doderlein, G., Kowenz-Leutz, E., and Graf, T. (1998). Distinct C/EBP functions are required for eosinophil lineage commitment and maturation. Gene Dev. 12, 2413–2423. Nerlov, C., Querfurth, E., Kulessa, H., and Graf, T. (2000). GATA-1 interacts with the myeloid PU.1 transcription factor and represses PU.1-dependent transcription. Blood 95, 2543–2551. Nichogiannopoulou, A., Trevisan, M., Neben, S., Friedrich, C., and Georgopoulos, K. (1999). Defects in hemopoietic stem cell activity in Ikaros mutant mice. J. Exp. Med. 190, 1201–1214. Nishijima, I., Nakahata, T., Hirabayashi, Y., Inoue, T., Kurata, H., Miyajima, A., Hayashi, N., Iwakura, Y., Arai, K., and Yokota, T. (1995). A human GM-CSF receptor expressed in transgenic mice stimulates proliferation and differentiation of hemopoietic progenitors to all lineages in response to human GM-CSF. Mol. Biol. Ce. 6, 497–508. Noguchi, M., Nakamura, Y., Russell, S. M., Ziegler, S. F., Tsang, M., Cao, X., and Leonard, W. J. (1993). Interleukin-2 receptor gamma chain: A functional component of the interleukin-7 receptor. Science 262, 1877–1880. Nosten-Bertrand, M., Errington, M. L., Murphy, K. P., Tokugawa, Y., Barboni, E., Kozlova, E., Michalovich, D., Morris, R. G., Silver, J., Stewart, C. L., et al. (1996). Normal spatial learning despite regional inhibition of LTP in mice lacking Thy-1. Nature 379, 826–829. Nutt, S. L., Heavey, B., Rolink, A. G., and Busslinger, M. (1999). Commitment to the B-lymphoid lineage depends on the transcription factor Pax5. Nature 401, 556–562. O’Riordan, M., and Grosschedl, R. (1999). Coordinate regulation of B cell differentiation by the transcription factors EBF and E2A. Immunity 11, 21–31. Ogawa, M. (1993). Differentiation and proliferation of hematopoietic stem cells. Blood 81, 2844–2853. Ohbo, K., Suda, T., Hashiyama, M., Mantani, A., Ikebe, M., Miyakawa, K., Moriyama, M., Nakamura, M., Katsuki, M., Takahashi, K., et al. (1996). Modulation of hematopoiesis in mice with a truncated mutant of the interleukin-2 receptor gamma chain. Blood 87, 956–967. Okuno, Y., Iwasaki, H., Huettner, C. S., Radomska, H. S., Gonzalez, D. A., Tenen, D. G., and Akashi, K. (2002). Differential regulation of the human and murine CD34 genes in hematopoietic stem cells. Proc. Natl. Acad. Sci. USA 99, 6246–6251. Olweus, J., BitMansour, A., Warnke, R., Thompson, P. A., Carballido, J., Picker, L. J., and Lund-Johansen, F. (1997). Dendritic cell ontogeny: A human dendritic cell lineage of myeloid origin. Proc. Natl. Acad. Sci. USA 94, 12551–12556.

LINEAGE COMMITMENT

49

Olweus, J., Thompson, P. A., and Lund-Johansen, F. (1996). Granulocytic and monocytic differentiation of CD34hi cells is associated with distinct changes in the expression of the PU.1-regulated molecules, CD64, and macrophage colony-stimulating factor receptor. Blood 88, 3741–3754. Ordentlich, P., Lin, A., Shen, C. P., Blaumueller, C., Matsuno, K., Artavanis-Tsakonas, S., and Kadesch, T. (1998). Notch inhibition of E47 supports the existence of a novel signaling pathway. Mol. Cell B. 18, 2230–2239. Orlic, D., Kajstura, J., Chimenti, S., Jakoniuk, I., Anderson, S. M., Li, B., Pickel, J., McKay, R., Nadal-Ginard, B., Bodine, D. M., et al. (2001). Bone marrow cells regenerate infarcted myocardium. Nature 410, 701–705. Osawa, M., Hanada, K., Hamada, H., and Nakauchi, H. (1996). Long-term lymphohematopoietic reconstitution by a single CD34- low/negative hematopoietic stem cell. Science 273, 242–245. Pawliuk, R., Eaves, C., and Humphries, R. K. (1996). Evidence of both ontogeny and transplant dose-regulated expansion of hematopoietic stem cells in vivo. Blood 88, 2852–2858. Peschon, J. J., Morrissey, P. J., Grabstein, K. H., Ramsdell, F. J., Maraskovsky, E., Gliniak, B. C., Park, L. S., Ziegler, S. F., Williams, D. E., Ware, C. B., et al. (1994). Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deﬁcient mice. J. Exp. Med. 180, 1955–1960. Pharr, P. N., Hankins, D., Hofbauer, A., Lodish, H. F., and Longmore, G. D. (1993). Expression of a constitutively active erythropoietin receptor in primary hematopoietic progenitors abrogates erythropoietin dependence and enhances erythroid colony-forming unit, erythroid burstforming unit, and granulocyte/macrophage progenitor growth. Proc. Natl. Acad. Sci. USA 90, 938–942. Pluznik, D. H., and Sachs, L. (1965). The cloning of normal ‘‘mast’’ cells in tissue culture J. Cell Phys. 66, 319–324. Priller, J., Flugel, A., Wehner, T., Boentert, M., Haas, C. A., Prinz, M., Fernandez-Klett, F., Prass, K., Bechmann, I., de Boer, B. A., et al. (2001). Targeting gene-modiﬁed hematopoietic cells to the central nervous system: Use of green ﬂuorescent protein uncovers microglial engraftment. Nat. Med. 7, 1356–1361. Pui, J. C., Allman, D., Xu, L., DeRocco, S., Karnell, F. G., Bakkour, S., Lee, J. Y., Kadesch, T., Hardy, R. R., Aster, J. C., and Pear, W. S. (1999). Notch1 expression in early lymphopoiesis inﬂuences B versus T lineage determination. Immunity 11, 299–308. Radtke, F., Ferrero, I., Wilson, A., Lees, R., Aguet, M., and MacDonald, H. R. (2000). Notch1 deﬁciency dissociates the intrathymic development of dendritic cells and T cells. J. Exp. Med. 191, 1085–1094. Radtke, F., Wilson, A., Stark, G., Bauer, M., van Meerwijk, J., MacDonald, H. R., and Aguet, M. (1999). Deﬁcient T cell fate speciﬁcation in mice with an induced inactivation of Notch1. Immunity 10, 547–558. Randall, T. D., and Weissman, I. L. (1997). Phenotypic and functional changes induced at the clonal level in hematopoietic stem cells after 5-ﬂuorouracil treatment. Blood 89, 3596–3606. Randolph, G. J., Beaulieu, S., Lebecque, S., Steinman, R. M., and Muller, W. A. (1998). Differentiation of monocytes into dendritic cells in a model of transendothelial trafﬁcking. Science 282, 480–483. Rappold, I., Ziegler, B. L., Kohler, I., Marchetto, S., Rosnet, O., Birnbaum, D., Simmons, P. J., Zannettino, A. C., Hill, B., Neu, S., et al. (1997). Functional and phenotypic characterization of cord blood and bone marrow subsets expressing FLT3 (CD135) receptor tyrosine kinase. Blood 90, 111–125. Reizis, B., and Leder, P. (2001). The upstream enhancer is necessary and sufﬁcient for the expression of the pre-T cell receptor alpha gene in immature T lymphocytes. J. Exp. Med. 194, 979–990. Robey, E. (1997). Notch in vertebrates. Cur. Op. Gen. 7, 551–557.

50

DAVID TRAVER AND KOICHI AKASHI

Robey, E., Chang, D., Itano, A., Cado, D., Alexander, H., Lans, D., Weinmaster, G., and Salmon, P. (1996). An activated form of Notch inﬂuences the choice between CD4 and CD8 T cell lineages. Cell 87, 483–492. Rodewald, H. R., Brocker, T., and Haller, C. (1999). Developmental dissociation of thymic dendritic cell and thymocyte lineages revealed in growth factor receptor mutant mice. Proc. Natl. Acad. Sci. USA 96, 15068–15073. Rolink, A. G., Nutt, S. L., Melchers, F., and Busslinger, M. (1999). Long-term in vivo reconstitution of T-cell development by Pax5-deﬁcient B-cell progenitors. Nature 401, 603–606. Romani, N., Gruner, S., Brang, D., Kampgen, E., Lenz, A., Trockenbacher, B., Konwalinka, G., Fritsch, P. O., Steinman, R. M., and Schuler, G. (1994). Proliferating dendritic cell progenitors in human blood. J. Exp. Med. 180, 83–93. Romeo, P. H., Prandini, M. H., Joulin, V., Mignotte, V., Prenant, M., Vainchenker, W., Marguerie, G., and Uzan, G. (1990). Megakaryocytic and erythrocytic lineages share speciﬁc transcription factors. Nature 344, 447–449. Rothenberg, E. V., and Dionne, C. J. (2002). Lineage plasticity and commitment in T-cell development. Immunol. Rev. 187, 96–115. Sagara, S., Sugaya, K., Tokoro, Y., Tanaka, S., Takano, H., Kodama, H., Nakauchi, H., and Takahama, Y. (1997). B220 expression by T lymphoid progenitor cells in mouse fetal liver. J. Immunol. 158, 666–676. Sakaguchi, N., and Melchers, F. (1986). Lambda 5, a new light-chain-related locus selectively expressed in pre-B lymphocytes. Nature 324, 579–582. Sallusto, F., and Lanzavecchia, A. (1994). Efﬁcient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor alpha. J. Exp. Med. 179, 1109–1118. Schatz, D. G., Oettinger, M. A., and Schlissel, M. S. (1992). V(D)J recombination: Molecular biology and regulation. Ann. R. Immun. 10, 359–383. Scott, E. W., Fisher, R. C., Olson, M. C., Kehrli, E. W., Simon, M. C., and Singh, H. (1997). PU.1 functions in a cell-autonomous manner to control the differentiation of multipotential lymphoid-myeloid progenitors. Immunity 6, 437–447. Scott, E. W., Simon, M. C., Anastasi, J., and Singh, H. (1994). Requirement of transcription factor PU.1 in the development of multiple hematopoietic lineages. Science 265, 1573–1577. Searles, A. E., Pohlmann, S. J., Pierce, L. J., Perry, S. S., Slayton, W. B., Mojica, M. P., and Spangrude, G. J. (2000). Rapid, B lymphoid-restricted engraftment mediated by a primitive bone marrow subpopulation. J. Immunol. 165, 67–74. Semerad, C. L., Poursine-Laurent, J., Liu, F., and Link, D. C. (1999). A role for G-CSF receptor signaling in the regulation of hematopoietic cell function but not lineage commitment or differentiation. Immunity 11, 153–161. Seshasayee, D., Gaines, P., and Wojchowski, D. M. (1998). GATA-1 dominantly activates a program of erythroid gene expression in factor-dependent myeloid FDCW2 cells. Mol. Cell B. 18, 3278– 3288. Shinkai, Y., Rathbun, G., Lam, K. P., Oltz, E. M., Stewart, V., Mendelsohn, M., Charron, J., Datta, M., Young, F., Stall, A. M., et al. (1992). RAG-2-deﬁcient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68, 855–867. Shivdasani, R. A., Fujiwara, Y., McDevitt, M. A., and Orkin, S. H. (1997). A lineage-selective knockout establishes the critical role of transcription factor GATA-1 in megakaryocyte growth and platelet development. EMBO J. 16, 3965–3973. Shivdasani, R. A., Rosenblatt, M. F., Zucker-Franklin, D., Jackson, C. W., Hunt, P., Saris, C. J., and Orkin, S. H. (1995). Transcription factor NF-E2 is required for platelet formation independent of the actions of thrombopoietin/MGDF in megakaryocyte development. Cell 81, 695–704.

LINEAGE COMMITMENT

51

Shortman, K., and Liu, Y. J. (2002). Mouse and human dendritic cell subtypes. Nat. Rev. Immunol. 2, 151–161. Sieweke, M. H., Tekotte, H., Frampton, J., and Graf, T. (1996). MafB is an interaction partner and repressor of Ets-1 that inhibits erythroid differentiation. Cell 85, 49–60. Sitnicka, E., Bryder, D., Theilgaard-Monch, K., Buza-Vidas, N., Adolfsson, J., and Jacobsen, S. E. (2002). Key role of ﬂt3 ligand in regulation of the common lymphoid progenitor but not in maintenance of the hematopoietic stem cell pool. Immunity 17, 463–472. Smith, L. G., Weissman, I. L., and Heimfeld, S. (1991). Clonal analysis of hematopoietic stem-cell differentiation in vivo. Proc. Natl. Acad. Sci. USA 88, 2788–2792. Smith, L. J., Curtis, J. E., Messner, H. A., Senn, J. S., Furthmayr, H., and McCulloch, E. A. (1983). Lineage inﬁdelity in acute leukemia. Blood 61, 1138–1145. Souabni, A., Cobaleda, C., Schebesta, M., and Busslinger, M. (2002). Pax5 promotes B lymphopoiesis and blocks T cell development by repressing notch1. Immunity 17, 781–793. Spangrude, G. J., Aihara, Y., Weissman, I. L., and Klein, J. (1988). The stem cell antigens Sca-1 and Sca-2 subdivide thymic and peripheral T lymphocytes into unique subsets. J. Immunol. 141, 3697–3707. Spangrude, G. J., Smith, L., Uchida, N., Ikuta, K., Heimfeld, S., Friedman, J., and Weissman, I. L. (1991). Mouse hematopoietic stem cells. Blood 78, 1395–1402. Spits, H., Couwenberg, F., Bakker, A. Q., Weijer, K., and Uittenbogaart, C. H. (2000). Id2 and Id3 inhibit development of CD34(þ) stem cells into predendritic cell (pre-DC)2 but not into preDC1. Evidence for a lymphoid origin of pre-DC2. J. Exp. Med. 192, 1775–1784. Stanley, E., Lieschke, G. J., Grail, D., Metcalf, D., Hodgson, G., Gall, J. A., Maher, D. W., Cebon, J., Sinickas, V., and Dunn, A. R. (1994). Granulocyte/macrophage colony-stimulating factordeﬁcient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc. Natl. Acad. Sci. USA 91, 5592–5596. Stass, S., Mirro, J., Melvin, S., Pui, C. H., Murphy, S. B., and Williams, D. (1984). Lineage switch in acute leukemia. Blood 64, 701–706. Storms, R. W., Goodell, M. A., Fisher, A., Mulligan, R. C., and Smith, C. (2000). Hoechst dye efﬂux reveals a novel CD7(þ)CD34() lymphoid progenitor in human umbilical cord blood. Blood 96, 2125–2133. Suda, J., Suda, T., and Ogawa, M. (1984a). Analysis of differentiation of mouse hemopoietic stem cells in culture by sequential replating of paired progenitors. Blood 64, 393–399. Suda, T., Suda, J., and Ogawa, M. (1984b). Disparate differentiation in mouse hemopoietic colonies derived from paired progenitors. Proc. Natl. Acad. Sci. USA 81, 2520–2524. Takagi, M., Hara, T., Ichihara, M., Takatsu, K., and Miyajima, A. (1995). Multi-colony stimulating activity of interleukin 5 (IL-5) on hematopoietic progenitors from transgenic mice that express IL-5 receptor alpha subunit constitutively. J. Exp. Med. 181, 889–899. Tenen, D. G., Hromas, R., Licht, J. D., and Zhang, D. E. (1997). Transcription factors, normal myeloid development, and leukemia. Blood 90, 489–519. Terada, N., Hamazaki, T., Oka, M., Hoki, M., Mastalerz, D. M., Nakano, Y., Meyer, E. M., Morel, L., Petersen, B. E., and Scott, E. W. (2002). Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion. Nature 416, 542–545. Till, J. E., and McCulloch, E. A. (1961). A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiat. Res. 14, 1419–1430. Ting, C. N., Olson, M. C., Barton, K. P., and Leiden, J. M. (1996). Transcription factor GATA-3 is required for development of the T-cell lineage. Nature 384, 474–478. Tomita, K., Hattori, M., Nakamura, E., Nakanishi, S., Minato, N., and Kageyama, R. (1999). The bHLH gene Hes1 is essential for expansion of early T cell precursors. Gene Dev. 13, 1203–1210.

52

DAVID TRAVER AND KOICHI AKASHI

Traver, D., Akashi, K., Manz, M., Merad, M., Miyamoto, T., Engleman, E. G., and Weissman, I. L. (2000). Development of CD8alpha-positive dendritic cells from a common myeloid progenitor. Science 290, 2152–2154. Traver, D., Miyamoto, T., Christensen, J., Iwasaki-Arai, J., Akashi, K., and Weissman, I. L. (2001). Fetal liver myelopoiesis occurs through distinct, prospectively isolatable progenitor subsets. Blood 98, 627–635. Tudor, K. S., Payne, K. J., Yamashita, Y., and Kincade, P. W. (2000). Functional assessment of precursors from murine bone marrow suggests a sequence of early B lineage differentiation events. Immunity 12, 335–345. Uchida, N., Aguila, H. L., Fleming, W. H., Jerabek, L., and Weissman, I. L. (1994). Rapid and sustained hematopoietic recovery in lethally irradiated mice transplanted with puriﬁed Thy-1. 1lo Lin-Sca-1þ hematopoietic stem cells. Blood 83, 3758–3779. Uchida, N., and Weissman, I. (1992). Searching for hematopoietic stem cells: Evidence that Thy-1. 1lo Lin-Sca-1þ cells are the only stem cells in C57BL/Ka-Thy-1.1 bone marrow. J. Exp. Med. 175, 175–184. van den Elsen, P., Bruns, G., Gerhard, D. S., Pravtcheva, D., Jones, C., Housman, D., Ruddle, F. A., Orkin, S., and Terhorst, C. (1985). Assignment of the gene coding for the T3-delta subunit of the T3-T-cell receptor complex to the long arm of human chromosome 11 and to mouse chromosome 9. Proc. Natl. Acad. Sci. USA 82, 2920–2924. van Gent, D. C., Ramsden, D. A., and Gellert, M. (1996). The RAG1 and RAG2 proteins establish the 12/23 rule in V(D)J recombination. Cell 85, 107–113. Varnum-Finney, B., Xu, L., Brashem-Stein, C., Nourigat, C., Flowers, D., Bakkour, S., Pear, W. S., and Bernstein, I. D. (2000). Pluripotent, cytokine-dependent, hematopoietic stem cells are immortalized by constitutive Notch1 signaling. Nat. Med. 6, 1278–1281. Visser, J. W. M., Gauman, J. G. J., Mulder, A. H., Eliason, J. F., and de Leeuw, A. W. (1984). Isolation of murine pluripotent hemopoietic stem cells. J. Exp. Med. 59, 1576–1590. Visvader, J. E., Elefanty, A. G., Strasser, A., and Adams, J. M. (1992). GATA-1 but not SCL induces megakaryocytic differentiation in an early myeloid line. EMBO J. 11, 4557–4564. von Boehmer, H. (1999). T-cell development: What does Notch do for T cells? Curr. Biol. 9, R186– 188. von Freeden-Jeffry, U., Solvason, N., Howard, M., and Murray, R. (1997). The earliest T lineagecommitted cells depend on IL-7 for Bcl-2 expression and normal cell cycle progression. Immunity 7, 147–154. von Freeden-Jeffry, U., Vieira, P., Lucian, L. A., McNeil, T., Burdach, S. E., and Murray, R. (1995). Lymphopenia in interleukin (IL)-7 gene-deleted mice identiﬁes IL-7 as a nonredundant cytokine. J. Exp. Med. 181, 1519–1526. Wadman, I. A., Osada, H., Grutz, G. G., Agulnick, A. D., Westphal, H., Forster, A., and Rabbitts, T. H. (1997). The LIM-only protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding complex which includes the TAL1, E47, GATA-1, and Ldb1/NLI proteins. EMBO J. 16, 3145–3157. Wagers, A. J., Sherwood, R. I., Christensen, J. L., and Weissman, I. L. (2002). Little evidence for developmental plasticity of adult hematopoietic stem cells. Science 297, 2256–2259. Wang, J. H., Nichogiannopoulou, A., Wu, L., Sun, L., Sharpe, A. H., Bigby, M., and Georgopoulos, K. (1996). Selective defects in the development of the fetal and adult lymphoid system in mice with an Ikaros null mutation. Immunity 5, 537–549. Wang, Y., Zhang, Y., Yoneyama, H., Onai, N., Sato, T., and Matsushima, K. (2002). Identiﬁcation of CD8alphaþCD11c- lineage phenotype-negative cells in the spleen as committed precursor of CD8alphaþ dendritic cells. Blood 100, 569–577.

LINEAGE COMMITMENT

53

Washburn, T., Schweighoffer, E., Gridley, T., Chang, D., Fowlkes, B. J., Cado, D., and Robey, E. (1997). Notch activity inﬂuences the alphabeta versus gammadelta T cell lineage decision. Cell 88, 833–843. Weintraub, H. (1985). Assembly and propagation of repressed and depressed chromosomal states. Cell 42, 705–711. Weissman, I. L. (1994). Stem cells, clonal progenitors, and commitment to the three lymphocyte lineages: T, B, and NK cells. Immunity 1, 529–531. Wiesmann, A., Phillips, R. L., Mojica, M., Pierce, L. J., Searles, A. E., Spangrude, G. J., and Lemischka, I. (2000). Expression of CD27 on murine hematopoietic stem and progenitor cells. Immunity 12, 193–199. Wu, A., Till, J., Siminovitch, L., and McCulloch, E. (1968). Cytological evidence for a relationship between normal hematopoietic colony-forming cells and cells of the lymphoid system. J. Exp. Med. 127, 455–467. Wu, H., Liu, X., Jaenisch, R., and Lodish, H. F. (1995). Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell 83, 59–67. Wu, L., Antica, M., Johnson, G. R., Scollay, R., and Shortman, K. (1991a). Developmental potential of the earliest precursor cells from the adult mouse thymus. J. Exp. Med. 174, 1617–1627. Wu, L., Scollay, R., Egerton, M., Pearse, M., Spangrude, G. J., and Shortman, K. (1991b). CD4 expressed on earliest T-lineage precursor cells in the adult murine thymus. Nature 349, 71–74. Wu, L., D’Amico, A., Hochrein, H., O’Keeffe, M., Shortman, K., and Lucas, K. (2001). Development of thymic and splenic dendritic cell populations from different hemopoietic precursors. Blood 98, 3376–3382. Wu, L., D’Amico, A., Winkel, K. D., Suter, M., Lo, D., and Shortman, K. (1998). RelB is essential for the development of myeloid-related CD8alpha- dendritic cells but not of lymphoid-related CD8alphaþ dendritic cells. Immunity 9, 839–847. Wu, L., Li, C. L., and Shortman, K. (1996). Thymic dendritic cell precursors: Relationship to the T lymphocyte lineage and phenotype of the dendritic cell progeny. J. Exp. Med. 184, 903–911. Wu, L., Nichogiannopoulou, A., Shortman, K., and Georgopoulos, K. (1997). Cell-autonomous defects in dendritic cell populations of Ikaros mutant mice point to a developmental relationship with the lymphoid lineage. Immunity 7, 483–492. Yamaguchi, Y., Zon, L. I., Ackerman, S. J., Yamamoto, M., and Suda, T. (1998). Forced GATA-1 expression in the murine myeloid cell line M1: Induction of c-Mp1 expression and megakaryocytic/erythroid differentiation. Blood 91, 450–457. Yang, F. C., Tsuji, K., Oda, A., Ebihara, Y., Xu, M. J., Kaneko, A., Hanada, S., Mitsui, T., Kikuchi, A., Manabe, A., et al. (1999). Differential effects of human granulocyte colony-stimulating factor (hG- CSF) and thrombopoietin on megakaryopoiesis and platelet function in hG- CSF receptor-transgenic mice. Blood 94, 950–958. Yang, F. C., Watanabe, S., Tsuji, K., Xu, M. J., Kaneko, A., Ebihara, Y., and Nakahata, T. (1998). Human granulocyte colony-stimulating factor (G-CSF) stimulates the in vitro and in vivo development but not commitment of primitive multipotential progenitors from transgenic mice expressing the human G- CSF receptor. Blood 92, 4632–4640. Ying, Q. L., Nichols, J., Evans, E. P., and Smith, A. G. (2002). Changing potency by spontaneous fusion. Nature 416, 545–548. Young, J. W., Szabolcs, P., and Moore, M. A. (1995). Identiﬁcation of dendritic cell colony-forming units among normal human CD34þ bone marrow progenitors that are expanded by c-kitligand and yield pure dendritic cell colonies in the presence of granulocyte/macrophage colonystimulating factor and tumor necrosis factor alpha. J. Exp. Med. 182, 1111–1119.

54

DAVID TRAVER AND KOICHI AKASHI

Zhang, D. E., Zhang, P., Wang, N. D., Hetherington, C. J., Darlington, G. J., and Tenen, D. G. (1997). Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein alpha-deﬁcient mice. Proc. Natl. Acad. Sci. USA 94, 569–574. Zhang,P.,Behre,G.,Pan,J.,Iwama,A.,Wara-Aswapati,N.,Radomska,H.S.,Auron,P.E.,Tenen,D.G., and Sun, Z. (1999). Negative cross-talk between hematopoietic regulators: GATA proteins repress PU.1. Proc. Natl. Acad. Sci. USA 96, 8705–8710. Zhang, P., Zhang, X., Iwama, A., Yu, C., Smith, K. A., Mueller, B. U., Narravula, S., Torbett, B. E., Orkin, S. H., and Tenen, D. G. (2000). PU.1 inhibits GATA-1 function and erythroid differentiation by blocking GATA-1 DNA binding. Blood 96, 2641–2648.

advances in immunology, vol. 83

The CD4/CD8 Lineage Choice: New Insights into Epigenetic Regulation during T Cell Development ICHIRO TANIUCHI,* WILFRIED ELLMEIER,y AND DAN R. LITTMANz *RIKEN Research Center, Yokohama 230-0045, Japan y University of Vienna,1235 Vienna, Austria z New York University School of Medicine New York, NY 10016

I. Introduction

Differentiation of progenitor cells into various progeny with distinct functional properties is achieved by the regulated activation and silencing of batteries of genes. The ﬁnal differentiated state is thus deﬁned by the expression pattern of these genes, which is heritably maintained as cells undergo subsequent cell divisions. Hematopoiesis offers an ideal developmental system to study how progenitors differentiate into cells of distinct lineage and function and how differentiated states are then stably maintained. A series of transcription factors required for differentiation of erythroid, myeloid, and lymphocyte lineages have been described (Cantor and Orkin, 2001; Singh et al., 1999). Within the lymphoid lineages, activation of the Notch signaling pathway in the thymus favors T cell development, at the expense of B lymphocytes (Izon et al., 2002). This is followed by a series of developmental choices as thymocytes differentiate into T cells with distinct antigen recognition properties and effector functions. Differentiation of thymocytes and their subsequent selection results in export to peripheral secondary lymphoid organs of a large variety of cells: CD4þ T helper cells (Th), CD8þ cytotoxic cells (CTL or Tc), CD4þ regulatory or suppressor cells (Treg), CD8þ intestinal intraepithelial lymphocytes (IEL), NKT cells with restricted T cell antigen receptor (TCR) repertoires, and other T cells with restricted TCRs that localize to various peripheral organs and mucosal regions. The molecular events leading to the differentiation of each of these lineages remain largely unexplored. Expression of the CD4 and CD8 glycoproteins deﬁnes the major lineages of T lymphocyes, and is coupled to the transcriptional programs that characterize the helper and cytotoxic lineages, respectively (Killeen and Littman, 1996). Regulation of expression of the genes encoding these coreceptor molecules has been studied in some detail, with the expectation that such studies could elucidate the mechanisms involved in lineage commitment and establishment of differentiated states. These studies have revealed a complex network of positive regulatory elements in both genes, as well as an extensively characterized negative regulatory region within 55 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

56

ICHIRO TANIUCHI ET AL.

the CD4 gene. In addition, the silencer in the CD4 gene has been shown to be involved in epigenetic regulation, and there are hints that the enhancers in the CD8 gene may also function through epigenetic mechanisms that result in heritable ‘‘on’’ states of expression. The recent ﬁndings on CD4 and CD8 regulation and the general tractability of the T cell system render these genes highly attractive for studies of the general phenomenon of epigenetic regulation during development (Jaenisch and Bird, 2003; Li, 2002). This chapter discusses the role of the CD4 and CD8 co-receptors in T cell development and signal transduction, the regulation of expression of the genes that encode these proteins, and the outstanding questions that need to be addressed. We also emphasize how our understanding of the CD4 silencer can inform future studies on epigenetic regulation during mammalian development. II. Role of Coreceptors in T Cell Development

The thymus is the site where lymphoid progenitors differentiate into T cells bearing a variety of speciﬁcities and functions (von Boehmer et al., 1989). Most cells differentiate along a lineage that gives rise to cytotoxic and helper T cells with TCRab receptors speciﬁc for MHC class I and class II molecules, respectively. Cells of this lineage can be readily classiﬁed according to their expression of the coreceptor molecules, CD4 and CD8. These cell surface glycoproteins bind to membrane-proximal domains of MHC class II and class I molecules, respectively, and their cytoplasmic domains recruit the Src family protein tyrosine kinase p56lck (Lck) (Weiss and Littman, 1994). The most immature cells express neither coreceptor and are classiﬁed as double negative (DN) cells. These cells can be subdivided into four stages, based on surface expression of CD25 and CD44. At the DN3 stage (CD25þCD44lo), there is initiation of VDJ rearrangement at the TCRb locus, and cells that undergo inframe rearrangement and hence express a pre-TCR are selected for further maturation, marked by several rounds of cell division before CD25 is shut off (DN4 stage: CD25CD44lo). Cells that have completed b-selection then turn on expression of both CD4 and CD8ab heterodimers, marking the double positive (DP) stage of differentiation (von Boehmer and Fehling, 1997). These cells, which are the most abundant in the thymus, then undergo VJ rearrangement of the TCRa genes, and cells with productive rearrangements initiate expression of TCRab heterodimers. The DP thymocytes are then subjected to selection based on their aptitude to interact with MHC molecules (Goldrath and Bevan, 1999; Sebzda et al., 1999). Most cells undergo ‘‘death by neglect’’ because they fail to bind with sufﬁcient avidity to self MHC–peptide complexes. The few cells with TCRs that bind with high afﬁnity to MHC–peptide complexes are deleted, through an apoptotic program known as negative selection. Only the rare DP thymocytes with TCRs of intermediate afﬁnity

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

57

for MHC transduce signals that ensure their survival and further differentiation, which is known as ‘‘positive selection.’’ Those cells selected by interacting with MHC class I proceed to become CD4CD8þ T cells (CD8 single positive or SP cells, bearing CD8ab heterodimers) with cytotoxic function, whereas those selected on MHC class II become CD4þCD8 cells (CD4 SP) with helper functions. Other lineages branch off at different stages from this major pathway of differentiation. Cells that express TCRgd differentiate from the DN precursors and migrate to various specialized mucosal regions (Hayday et al., 2001). Many of these give rise to IEL that express CD8aa homodimers (Guy-Grand et al., 2001; Lefrancois and Olson, 1994). A subset of cells with canonical TCRab rearrangements are selected at the DP stage by interacting with the class Ib molecule CD1d, and give rise to NKT cells that migrate to many tissues and make up a large proportion of T cells in the liver (Kronenberg and Gapin, 2002). In 2003, a subset of CD4 SP thymocytes was found to differentiate into regulatory T cells that keep autoreactive cells in check (Hori et al., 2003). The process by which DP thymocytes give rise to CD4 or CD8 SP cells is still poorly understood, but it is known to involve multiple steps. First, the cells undergo positive selection independently of commitment to one of the two lineages. This is known from the phenotype of mutant hd/hd (helper-deﬁcient) mice, which lack CD4 SP cells and mature T helper cells (Keefe et al., 1999). MHC class II-restricted thymocytes undergo positive selection in these mice, but they all differentiate into CD8þ CTL. These animals thus have no defect in positive selection, but they have defective lineage commitment that results in lineage reversal of class II-restricted T cells. How cells commit to the helper versus the cytotoxic lineage remains a matter of some controversy. Early studies suggested that DP thymocytes stochastically shut off either CD4 or CD8, and undergo further selection only if they retain expression of the coreceptor appropriate for the class of MHC recognized by the TCR. Thus, cells speciﬁc for class I would undergo apoptosis if they shut off CD8, which interacts with class I, but would be further selected if they retain expression of CD8 and shut off CD4, which interacts with class II (Davis and Littman, 1994). Later studies have provided strong evidence that the strength of the TCR-initiated signal governs the fate of DP thymocytes. Enhanced activation of the protein tyrosine kinase p56lck, which interacts with both CD4 and CD8 cytoplasmic domains, resulted in biased commitment to the CD4 lineage, such that even class I-restricted thymocytes became CD4 SP cells (Hernandez-Hoyos et al., 2000). Conversely, inhibition of Lck activity resulted in deviation towards the CD8 lineage. Because the proportion of CD4 molecules with Lck bound to them is higher than that of CD8, it has been proposed that the number of Lck molecules

58

ICHIRO TANIUCHI ET AL.

recruited to the TCR complex is the factor that determines the strength of the TCR-initiated signal and, hence, instructs the lineage choice. Most recently, Singer and colleagues have proposed a modiﬁed version of the strength-of-signal instructional model, which they have termed the ‘‘coreceptor reversal’’ or ‘‘kinetic signaling’’ model (Brugnera et al., 2000; Singer, 2002). This model is based on the ﬁnding that, following positive selection signals initiated by TCR interaction with either class I or class II molecules, thymocytes down-regulate expression of CD8, resulting in CD4þCD8lo cells that are precursors of both the CD4 and CD8 SP lineages. These precursor cells have been posited to give rise to cells of either the CD4 or CD8 lineage on the basis of the strength of the TCR/coreceptor signal relative to the strength during earlier selection at the DP stage. Thus, cells selected by interaction with class II would retain a strong signal at the CD4þCD8lo stage, since the CD4 coreceptor continues to be expressed. This would direct the cells towards the CD4 SP lineage. However, if cells were selected by interaction with class I, there would be a reduced signal upon transition to the CD4þCD8lo stage, and this would result in subsequent differentiation to the CD8 SP cytotoxic T cell lineage. The reduced or interrupted signal would therefore initiate the extinction of CD4 expression and the reactivation of CD8 expression. The novelty of this model lies in the proposition that abrogation of a signal dictates a switch in developmental fate. Based on in vitro thymocyte culture experiments, Singer and colleagues have suggested that IL-7R signaling is then required in cells destined for the CD8 SP lineage. This hypothesis is supported by the ﬁnding that conditional inactivation in thymocytes of the SOCS-1 gene, a negative regulator of gc (and hence IL-7R) signaling, results in a considerable increase in CD8 SP cells (Chong et al., 2003). Although the coreceptor reversal model has yet to be experimentally conﬁrmed, its predictions ﬁt well with our current understanding of CD4 and CD8 gene regulation. For example, there is evidence that there are multiple stage-speciﬁc enhancers that regulate CD8 expression, with some active at the DP stage and others at various stages of CD8 SP differentiation. A better understanding of how the coreceptor genes are regulated is, hence, likely not only to contribute to the elucidation of how lineage choice occurs in DP thymocytes, but also may reveal how complex programs of differentiation, to the helper and cytotoxic lineages, are regulated. In addition, other studies have revealed an unexpected novel function of a CD8 gene enhancer that selectively regulates expression of CD8a (Madakamutil et al., 2004). This enhancer has been shown to have a key role in the generation of memory CTL, and its further characterization may provide important insight into the genetic program that governs how a small subset of effector cells is set aside to differentiate into memory cells.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

59

III. Regulation of CD4 Gene Expression

A. Positive Regulatory Elements The CD4 locus is located in syntenic regions of murine chromosome 6 and human chromosome 12. It spans 80 kb, and its exon/intron structure is conserved in mouse and human. The lag-3 gene, which encodes a CD4-related molecule that also binds to MHC class II, is located approximately 20 kb upstream of cd4 in the mouse genome (Baixeras et al., 1992; Triebel et al., 1990). LAG-3 is expressed on all newly activated CD4þ and CD8þ T cells and on some cycling memory cells and, hence, differs signiﬁcantly in its regulation from the closely linked cd4 gene. LAG-3 appears to contribute to the expansion of activated T cells by regulating their survival (Workman and Vignali, 2003). LAG-3 is also expressed on natural killer cells and contributes to the ability of these cells to kill certain target cells (Miyazaki et al., 1996). 1. The CD4 Promoter Candidate cis-regulatory elements were initially idenﬁed using DNase hypersensitivity (HS) analyses of the mouse cd4 locus (Gorman et al., 1987; Sands and Nikolic-Zugic, 1992; Sawada and Littman, 1991; Siu et al., 1994). One of the HS sites is immediately upstream of the transcription initiation site, and a 101-bp fragment containing this sequence had promoter activity in transient transfection assays in T cell lines. Two Myb-binding motifs were required for full promoter activity (Nakayama et al., 1993; Siu et al., 1992). Another mutatagenesis study suggested the involvement of Ets-binding sites for CD4 promoter (P4) activity (Salmon et al., 1993). All studies to date on P4 have been conﬁned to transient expression analyses in vitro, and the promoter has not been analyzed in mice by transgenesis or targeted mutagenesis. This is particularly relevant because, for the activation of reporter genes in vivo, P4 is not sufﬁcient and requires the help of enhancers. It will be important to determine, using chromatin immunoprecipitation (ChIP), whether Myb, Ets, and other transcription factors are associated with P4 in various subsets of thymocytes and in mature CD4 and CD8 lineage cells. Comparison of CD4 and CD8 lineage T cells may also provide valuable information as to whether the promoter is occupied by transcription factors even when the gene is silenced (in CD8þ T cells). It should be noted that lineage speciﬁcity is unlikely to be conferred by P4, since it functions in transgenic mice in conjunction with enhancers from either the cd4 or cd8 genes (Ellmeier et al., 1997). 2. CD4 Enhancer Elements Analysis of gene expression in T cell lines transfected with a reporter gene under the regulation of the CD4 promoter and various sequences corresponding to DNase hypersensitivity (HS) sites resulted in the identiﬁcation of an

60

ICHIRO TANIUCHI ET AL.

enhancer located approximately 13 kb upstream of the transcriptional start site. A 339 bp fragment from this region directed expression of the reporter in an orientation- and position-independent manner, and functioned only in T cells (Sawada and Littman, 1991). The same element and its human homologue, combined with the CD4 promoter and various lengths of the mouse or human CD4 gene, directed expression of CD4 or reporter genes in thymocytes and T cells, but not in other cell lineages (Gillespie et al., 1993; Hanna et al., 1994; Killeen et al., 1993; Sawada et al., 1994; Siu et al., 1994). This T cell-speciﬁc enhancer has been referred to as the CD4 proximal enhancer (E4p). E4p contains three nuclear protein-binding sites that were identiﬁed by DNaseI footprinting with nuclear extracts from Tcell lines (Sawada and Littman, 1991). One of these is a binding motif for the HMG box family proteins, TCF1/LEF1, and the other two contain E-box motifs that bind basic helix–loop–helix proteins. Mutation of tandem E-box motifs (CD4-3) at the 30 end of the enhancer abrogated enhancer activity in transiently transfected T cell lines, whereas mutation of the TCF1/ LEF1 binding motif (CD4-2) and the 50 E-box (CD4-1) had only minor effects (Sawada and Littman, 1991). A heterodimer consisting of E2A and HEB, both basic HLH proteins, was found to bind to one of the E-boxes in CD4–3 and to be required for CD4 enhancer activity in the in vitro assays (Sawada and Littman, 1993). This is discussed in greater detail in a later section. Another HS site, mapped approximately 24 kb upstream of the transcriptional start site in murine cd4, was ﬁrst characterized as a CD4 distal enhancer (E4d) (Wurster et al., 1994). However, combined with the CD4 promoter, E4d does not appear to have enhancer activity in vivo. In addition, transgenes containing both E4d and E4p were expressed not only in T cells, but also in B cells and macrophages (Siu et al., 1994). The location of E4d is adjacent to the 30 end of the neighboring lag-3, which is expressed in activated T cells and in NK cells. It is possible that E4d is involved in regulation of lag-3 expression rather than cd4 expression. The role of E4p in directing expression in all thymocytes and T lymphocytes has been challenged by two reports, which together indicate that positive regulation of CD4 expression at different stages of development is more complex than initially thought. First, von Andrian and colleagues showed that activation of mature T cells by antigen results in down-regulation of expression of a transgenic GFP reporter regulated by E4p and the CD4 promoter (Manjunath et al., 1999). Because endogenous CD4 expression is not affected by activation, this result suggests that regulatory elements other than E4p contribute to the expression of CD4 in antigen-stimulated cells. It is possible that a second enhancer, either a novel or a known one such as E4d, functions alone or in cooperation with E4p to maintain expression of CD4 in effector and memory T cells. Mutation of these enhancers within the endogenous cd4 locus will be required to reveal their physiological functions.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

61

A second report has presented compelling evidence that the proximal CD4 enhancer may not be sufﬁcient to direct expression of transgenic reporter genes in immature DP thymocytes (Adlam and Siu, 2003). These studies employed the 339 bp E4p and approximately 1 kb of promoter sequence 50 to the transcriptional start site. In multiple transgenic mice, this combination of regulatory sequences resulted in expression only in SP thymocytes and mature T cells, but not in DP thymocytes. However, inclusion of a sequence corresponding to a distal DNase HS site, DH17, located approximately 40 kb 30 to the cd4 locus in the ﬁrst intron of the isot gene, restored expression in DP thymocytes (Fig. 1). This sequence was shown to have enhancer activity in transient transfections of a DP thymoma, but it did not function in vivo in the absence of the proximal enhancer. DH17 was shown to contain both an enhancer, in a 1.0 kb fragment that was termed TE (thymocyte enhancer), and an adjacent 0.9 kb locus control region (LCR) that confers copy numberdependent and position-independent expression of transgenes (Adlam and Siu, 2003).

Fig 1 Enhancer and promoter elements in the cd4 locus that contribute to reporter transgene expression in double positive thymocytes versus mature single positive T cells. Constructs that utilize the human CD2 transgene (hCD2) are from Sawada et al., 1994, and unpublished studies of Zou and Littman. Constructs that utilize HLA-B7 are from Adlam and Siu, 2003. E4d, distal enhancer; E4p, proximal CD4 enhancer; P4, CD4 promoter; LCR/TE, locus control region and thymic enhancer located in ﬁrst intron of isot gene. DH denotes DNase hypersensitive sites in the cd4 locus. Sites within the cd4 coding region have been omitted. We propose that DH5 may encode a negative DP thymocyte regulatory element that requires TE or another unidentiﬁed enhancer for expression in immature thymocytes. (See Color Insert.)

62

ICHIRO TANIUCHI ET AL.

The conclusion that the LCR/TE is essential, along with E4p, for expression of CD4 in DP thymocytes is based on results with an impressively large number of transgenic mice. Similarly, many transgenic mice prepared with combinations of E4p, the CD4 promoter (P4), and various parts of the ﬁrst cd4 intron have clearly demonstrated expression in DP thymocytes as well as in more mature T cells. For example, the human CD2 reporter was expressed in DP (and some DN) thymocytes in multiple transgenic lines prepared with E4p and P4 without inclusion of the LCR/TE from DH17 (Sawada et al., 1994). Similarly, CD4-cre mice that have been widely employed to delete genes at the DP stage were prepared with the same combination of regulatory elements (Wolfer et al., 2001). Use of E4p/P4 also directed expression of GFP and of the chemokine receptors CXCR4 and CCR5 in DP thymocytes as well as in mature T cells in both mouse and rat (Ellmeier, Deng, and Littman, unpublished; Keppler et al., 2002; Weninger and von Andrian, unpublished). The consistency observed within each of the studies that examined expression of the transgenes suggests that the discrepancy in results must be due to a subtle, but important, difference in the constructs used by the different groups (see Fig. 1). The discrepancy was discussed in an earlier review, in which it was suggested that the difference may lie in the length of the promoters used by the two laboratories (Ellmeier et al., 1999). Siu and colleagues have used a promoter sequence that includes about 1 kb 50 of the transcriptional start site, while our group has used only 0.5 kb (Adlam and Siu, 2003; Sawada et al., 1994). We propose that a sequence immediately upstream of the CD4 promoter has negative regulatory activity in DP thymocytes and, hence, requires a speciﬁc enhancer activity at this stage of development. This enhancer activity most likely involves E4p functioning in conjunction with another element, possibly the LCR/TE within the isot intron (Fig. 1). Because the region containing ISOT is not required for expression of human CD4 transgenes in DP thymocytes, it is possible that other elements, adjacent to the CD4 coding region, provide this enhancer function in both the mouse and human genes. Alternatively, the mouse and human genes may differ in their requirement for enhancers directing expression in DP thymocytes. The presence of yet another enhancer has been suggested by analysis of mice with targeted deletions of the silencer and adjacent sequences in intron I of cd4. In two separate studies aimed at analyzing silencer function (described in the following text), 6.5 kb or 1.6 kb intron sequences were deleted by homologous recombination in ES cells. In mutant mice generated from these cells, the level of cell surface CD4 expression was reduced in DP thymocytes and peripheral CD4þ T cells (Leung et al., 2001; Zou et al., 2001). Therefore, a putative positive regulatory element may have been deleted along with the CD4 silencer. The low CD4 expression was not due to residual silencer activity, since the level on CD4þ T cells from another mutant strain, in

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

63

which only 429 bp of the CD4 silencer was deleted, was similar to that from wild-type mice (Taniuchi et al., 2002b). Taken together, these results are compatible with the existence of a positive regulatory element, a CD4 intronic enhancer (E4i), adjacent to the CD4 silencer. This enhancer is not necessary for directing expression of reporter transgenes in DP thymocytes, since the combination of E4p and P4 was functional in these cells even in the absence of intron I (Ellmeier et al., 1999; Zou and Littman, unpublished). The functions of the various enhancers previously described (E4p, E4d, LCR/TE, and E4i) have only been inferred from examination of expression patterns in transgenic mice. Because of the complexity of the positive regulation of CD4, an understanding of the roles of the individual elements awaits an analysis following their systematic deletion in the murine germ line. B. Negative Regulatory Elements Involved in CD4 Silencing Early attempts to generate mice in which expression of human CD4 transgenes would faithfully reproduce the pattern of endogenous mouse CD4 expression provided the ﬁrst clue that the gene is negatively regulated in CD8 lineage cells. Human CD4 transgenes that contained E4p were expressed appropriately in DP thymocytes and T helper cells (Donda et al., 1996; Gillespie et al., 1993; Hanna et al., 1994; Killeen et al., 1993; Sawada et al., 1994; Siu et al., 1994). In contrast, when the murine E4p element was combined with the human promoter fused to a cDNA/genomic minigene in which exons 1–6 were derived from the cDNA, expression of CD4 was observed in all thymocytes and in both CD4þ and CD8þ T cells of transgenic mice (Crooks and Littman, unpublished). This suggested that a negative regulatory element, located within the ﬁrst 5 introns of the human CD4 gene, had been omitted in the transgene. Subsequent studies showed that lineage-speciﬁc expression of transgenes requires sequences corresponding to a DNase HS site located 1.6 kb downstream from the ﬁrst exon, within the ﬁrst intron of the murine cd4 locus (Sawada et al., 1994; Siu et al., 1994). A similar negative regulatory sequence in the ﬁrst intron of human CD4 was subsequently described (Donda et al., 1996). Inclusion in transgenic constructs of the murine CD4 silencer (S4), which was narrowed down to a 434 bp fragment from intron 1, resulted in repression of reporter gene expression in DN and CD8 SP thymocytes and in mature CD8 lineage T cells (Sawada et al., 1994; Siu et al., 1994). The activity of the silencer was independent of its orientation or location within the transgene, and it functioned with heterologous enhancers and promoters (Sawada et al., 1994). Importantly, expression of reporter transgenes in which the 434 bp CD4 silencer was combined with E4p and the 0.5 kb murine promoter recapitulated the normal developmental pattern of CD4 expression. Repression was observed not only in CD8 SP thymocytes, but also in a subset of immature CD4CD8DN thymocytes

64

ICHIRO TANIUCHI ET AL.

(Sawada et al., 1994). Based on these observations, a two-stage model was proposed for the temporal activation of the silencer during thymocyte differentiation (Sawada et al., 1994; Siu et al., 1994). In this model, the CD4 silencer is ﬁrst turned on in immature DN thymocytes followed by inactivation at subsequent stages in which both CD4 and CD8 are expressed, then turned on again speciﬁcally in those thymocytes that become cytotoxic-lineage cells. The CD4 silencer was thus proposed to be the key cis-regulatory element regulating lineage speciﬁcity of CD4 expression. Gene-targeting studies in mice have conﬁrmed that the CD4 silencer is essential for shutting off CD4 at both the DN and CD8 SP stages of thymocyte development. Following deletion of intronic segments that included the entire silencer, CD4 was de-repressed in a subset of thymocytes corresponding to immature CD4CD8 DN thymocytes in wild-type mice, in mature (HSAlow) CD8 SP thymocytes, and in peripheral CD8þ Tcells (Leung et al., 2001; Zou et al., 2001). Because the level of CD4 expression in the silencer-deleted mice was lower in immature CD4CD8DN thymocytes than in mature cytotoxic-lineage T cells (Leung et al., 2001; Zou et al., 2001), we cannot rule out the existence of another negative regulatory element operating speciﬁcally in immature DN thymocytes. However, as the 434 bp CD4 silencer is sufﬁcient to completely repress transgene expression in DN thymocytes, it is likely that enhancer activity in these cells is somehow lower than that in the later developmental stages. These results therefore indicate that the CD4 silencer is essential and, most likely, sufﬁcient for full repression of CD4 expression in both DN thymocytes and mature cytotoxic-lineage T cells, and demonstrate it to be the key cis-regulatory element that regulates lineage-speciﬁc activation of the CD4 locus. The search for functional sites within the CD4 silencer, which could potentially explain its selective activity in cytotoxic lineage T cells, has relied on a combination of transfection studies and analysis of transgenic mice, as well as on DNase footprinting analysis with nuclear extracts from T cells. A murine CD4CD8þ thymoma, 1200M, was found to be ideally suited for functional studies of the 434 bp CD4 silencer in transient transfection assays. When the silencer sequence was present in triplicate, reporter gene expression was effectively repressed. This permitted further whittling of the silencer to a 101 bp core fragment (165–265) that retained full activity (Taniuchi et al., 2002b). However, the minimal sequences required for silencer activity differed between transfection assays and transgenic reporter gene assays in mouse. A 147 bp fragment containing the core CD4 silencer was inactive in transgenic mice in the absence of sequences from either the 50 or 30 ﬂanking regions (Sawada and Littman, unpublished). This suggested that the transient repression observed in the T cell line is fundamentally different from the repression or silencing observed in vivo, possibly because changes in chromatin organization depend on speciﬁc sequences in these ﬂanks (see following text).

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

65

DNase I footprinting assays were employed by several groups to identify functionally relevant regions within the murine and human CD4 silencer. Using nuclear extracts from mouse T cell lines, Siu and colleagues identiﬁed three regions (site I, site II, and site III) within the 434 bp murine CD4 silencer (Duncan et al., 1996) (Fig. 2). Analysis of a 50 190 bp human CD4 silencer revealed two sites (site Ih and site IIh), one of which (Ih) partially overlapped with the homologous site I in the murine silencer (Donda et al., 1996). The requirement of these sites for silencer function was examined by mutational analyses in transgenic mice. Although single deletion of any site (site I, site II, or site III) from the 434 bp mouse CD4 silencer did not affect silencer activity, combined deletions of site II with either site I or site III resulted in the loss of activity in transgenic mice (Donda et al., 1996; Duncan et al., 1996). Another study using transient transfection assays identiﬁed two functional sites (site 1 and site 2) in the 101 bp core sequence (Taniuchi et al., 2002b). Site 1 partially overlapped with site II of Siu and colleagues, but site 2 differed from the previously described sites (Fig. 2). Mutation at site 1 or site 2 in the context of the core sequence abrogated silencer function in transfection assays. Similarly, these mutations, placed in the context of the core CD4 silencer with either 50 or 30 ﬂanking regions, resulted in complete loss of silencer activity in transgenic mice. A third site (site 3), located between sites 1 and 2, was ﬁrst identiﬁed by footprinting analysis (Sawada and Littman, unpublished). Although deletion of site 3 in the context of the core silencer resulted in only partial loss of silencer activity in the transfection assay, the same deletion in the context of the silencer lacking the 50 ﬂanking region (131–434) resulted in complete loss of silencer activity in transgenic mice (Mahanta and Littman, unpublished). Thus, the functional requirement for different sites within the silencer differed according to the assay system used and the context in which various mutations were analyzed. Resolution of this problem required development of a reliable system in which the physiological roles of various sequences within the CD4 silencer could be compared. Comparison of the functions of different motifs was accomplished by gene targeting, using homologous recombination in ES cells to introduce numerous mutations within the murine cd4 locus (Taniuchi et al., 2002b). This in vivo mapping study clearly demonstrated a physiological requirement for several sites or subregions within the CD4 silencer. Because the cd4 gene is expressed from both alleles, loss of silencer function in one allele resulted in CD4 derepression that was easily monitored by ﬂow-cytometry, thus allowing analysis in chimeric mice generated with ES cells harboring single mutant alleles. Silencer mutations within the endogenous cd4 locus had full, partial, or no effect on CD4 silencing in DN thymocytes and CD8 lineage T cells (Fig. 2). Importantly, there were differences in the consequence of the mutations at the different stages of T cell development, and this will be discussed in greater

66

ICHIRO TANIUCHI ET AL.

Fig 2 Transcriptional cis-regulatory elements in the murine cd4 locus and functional sites within the murine CD4 silencer. The coding exons are shown as closed bars and the noncoding 50 exon is shown as an open bar. The vertical arrows indicate DNaseI hypersensitive sites. The transcriptional direction of the CD4 gene is shown by the horizontal arrow. Proximal (E4p) and distal (E4d) enhancers (located approximately 13 kb and 24 kb upstream of the transcriptional start site, respectively) and a putative intronic (E4i) enhancer (located immediately 30 to the silencer) are shown as blue square boxes. The promoter (P4) and the silencer (S4) are shown as a purple square box and orange circle, respectively. The expanded region below the map represents the 434 bp murine CD4 silencer and putative factor binding sites. The top bar shows three footprint sites (yellow circles) deﬁned by Siu and colleagues (Duncan et al., 1996) along with putative transacting factors. The second bar shows ﬁve sites deﬁned by transient transfection assays, transgenic reporter assays, and targeted mutagenesis at the cd4 locus (Taniuchi et al., 2002b). The core silencer (sequence 165–265) was sufﬁcient for silencer activity in transfection assays. Site 2 and site 20 are identical to the binding motif for Runt domain transcription factor (Runx) family members. The effects of targeted mutations on CD4 gene silencing in either immature DN thymocytes or peripheral CD8þ mature T cells are shown at the bottom: ‘‘Part/uni’’ denotes partial CD4 derepression in a uniform pattern. ‘‘Part/vari’’ denotes partial CD4 de-repression in a variegated pattern. ‘‘Full’’ represents CD4 de-repression in all of the CD8þ T cells. n.t.: not tested. (See Color Insert.)

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

67

detail in a later section. Deletion of the entire silencer (1–429) or of 90 bp within the core (167–257) resulted in complete CD4 de-repression in all mature CD8-lineage cells and in DN thymocytes. This result clearly showed that the core silencer is indispensable for silencer function. Mutations of single functional sites (site 1, 2, or 3) within the core resulted in CD4 de-repression in only a proportion of mature CD8-lineage cells. However, a combination of mutations at site 1 with mutations at either site 2 or site 3 resulted in CD4 derepression in all of the mature CD8þ T cells, demonstrating a functional redundancy of these sites (Taniuchi et al., 2002b). Analysis of ﬂanking sequences, shown to have variable requirements in transgenic reporter studies, provided additional valuable information. Deletion of site III (260–281), which includes a putative binding motif for SAF (silencer associated factor), a factor reportedly localized in nuclei of CD8þ but not CD4þ T cells, had no effect on CD4 silencing. Deletion of the 30 ﬂanking region (279–429) affected CD4 silencing in only a minor proportion (7%) of mature CD8þ cells. In contrast, deletion of the 50 ﬂanking region (1–130) resulted in CD4 de-repression in a large proportion (80%) of mature CD8þ cells. Deletion of sequence 1–95 resulted in CD4 de-repression in only 10% of CD8þ cells, suggesting the presence of another functional site in 95–130 (referred to as putative site 4) (Taniuchi et al., 2002b). Sequence 75–90 within the 50 ﬂanking region, identiﬁed by footprinting analysis as site I (Duncan et al., 1996), contains a Runx binding motif (referred to as site 20 ) overlapping with an E box motif, a binding site for basic helix–loop–helix proteins. A targeted mutation in which only the Runx binding motif was mutated, leaving an intact E box motif, resulted in CD4 de-repression in a small (6%) subset of mature CD8þ T cells, similar to that observed upon deletion of sequence 1–95. However, a combination of the site 20 mutation with a mutation of the Runx binding motif in the core (site 2) resulted in complete CD4 de-repression in CD8þ T cells (Taniuchi et al., 2002a). This demonstrated a key role for the Runx binding motif in the 50 ﬂanking region of the silencer (see following text). A physiological role for the E box sequence has not been demonstrated by this kind of analysis, although in vitro studies have suggested that it may be a binding site for HES, a target of the Notch signaling pathway (Kim and Siu, 1998). C. Trans-acting Proteins Involved in CD4 Gene Regulation 1. Proteins Regulating CD4 Expression through Enhancers Investigation of trans-acting factors involved in CD4 enhancer function has been conﬁned to the proximal enhancer. A heterodimeric complex of two basic helix–loop–helix proteins, E2A and HEB, was shown to bind to one of the two E-box motifs in CD4-3 (Sawada and Littman, 1993). In the thymus of mice

68

ICHIRO TANIUCHI ET AL.

with a targeted disruption of a functional domain of HEB, there was the appearance of an unusual subset of CD4low/CD8þTCRint cells (Barndt et al., 2000; Zhuang et al., 1996). The kinetics of appearance of this subset in the fetal thymus suggested a delayed up-regulation of the cd4 gene during T cell ontogeny in the absence of a functional HEB protein. A similar population of cells was also observed in mice harboring heterozygous mutations for both heb and e2a (hebþ/ e2aþ/ mice). These results suggest that activation of CD4 is sensitive to the gene-dosage of HEB and E2A, and they are consistent with the involvement of the HEB/E2A heterodimeric complex in CD4 gene activation. This interpretation must be tempered, however, by the observation that compromised HEB and E2A function results in developmental arrest at the b-selection stage in thymopoiesis. The apparent defect in CD4 up-regulation thus occurs in the setting of defective developmental progression, and may not be a direct consequence of reduced HEB/E2A. Some thymocytes can migrate to the periphery in the HEB-deﬁcient mice, and CD4 expression is maintained in splenic T cells in these animals. Therefore, the requirement for HEB in CD4 gene expression may differ between immature thymocytes and mature T cells. This could be explained by the use of different enhancers at the two stages: HEB-dependent elements in E4p may be required in immature (DN and DP) thymocytes, whereas a different enhancer may provide HEB-independent activation in mature T cells. Alternatively, E2A could efﬁciently compensate for HEB function in mature T cells. A third possibility is that an epigenetic mechanism is involved in keeping CD4 active. HEB might therefore be necessary only for opening the CD4 locus in the initiation phase of gene activation, but would not be required subsequently to keep the locus active, due to heritable changes in chromatin. Interestingly, a similar observation was made in mice harboring a compromised CD8 enhancer (discussed in more detail later), although, in that case, there was accumulation of CD4þCD8low/TCRint thymocytes. Thus, compromised enhancer function at the CD4 and CD8 loci may result in inefﬁcient gene activation in a proportion of cells, and lead to subsequent variegated expression. Further analyses, such as a conditional deletion of E4p or the CD8 enhancers, may provide further insight into epigenetic aspects of activation of CD4 or CD8. 2. Trans-acting Proteins that Act on the CD4 Silencer Several transcription factor-binding sites in the CD4 silencer have been shown to be functionally important in transgenic and targeted mutagenesis studies. The basic helix–loop–helix protein HES-1, a target of the Notch signaling pathway, was shown in vitro to bind to one of these motifs, the E box sequence located in site I (Kim and Siu, 1998). In a T cell line, HES-1 repressed reporter gene expression through the E-box sequence in a

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

69

dose-dependent manner (Kim and Siu, 1998). This result was consistent with a proposed function of Notch in regulation of the CD4/CD8 lineage decision. However, inactivation of Notch1 at the DP stage had no effect on lineage choice, suggesting that Notch signaling may not be important in CD4 silencing (Wolfer et al., 2001). In RAG2-defective host mice reconstituted with HES-1-deﬁcient progenitors, T cell development was arrested at the DN stages (Tomita et al., 1999). However, in irradiated host mice, HES-1-deﬁcient progenitors gave rise to mature T cells, including normal mature CD4CD8þ SP cells, suggesting no role for HES-1 in the lineage choice (Kaneta et al., 2000). Further studies are required to determine if the HES-1-binding E box has a role in CD4 gene regulation. Two other factors proposed to have a function in CD4 silencing are c-Myb and SAF, shown in vitro to bind to site II and site III, respectively (Fig. 2) (Allen et al., 2001; Kim and Siu, 1999). Combined mutations of the c-Myb-bindingmotif and the SAF binding site led to the loss of silencer function in transgenic reporter mice (Duncan et al., 1996; Kim and Siu, 1999). Site II overlaps with site 1 identiﬁed by Taniuchi et al. (Fig. 2). However, the Myb-binding motif is not conserved between the mouse and human CD4 silencer, and the consensus sequence is different from the minimum functional sequence of site 1, characterized by ﬁne mapping in transfection assays (Taniuchi et al., 2002b). It is therefore unlikely that Myb binding to this site is required for silencer function. A role for SAF is also difﬁcult to assess at this stage. This homeobox protein was identiﬁed in a screen for factors binding to the site III sequence, and it was reported to localize to the nucleus in CD8þ, but not CD4þ, T cells (Kim and Siu, 1999). However, there has been no functional analysis of the role of SAF in silencing. CD4 silencing in mature CD8þ cells was normal in the absence of the SAF binding motif (Taniuchi et al., 2002b), although it is possible that this site provides a redundant function that can only be revealed by analysis of compound mutations. Whereas evidence for the involvement of many transcription factors in the regulation of CD4 expression remains circumstantial, recent genetic studies have identiﬁed two transcription factors that are clearly required for the negative regulation of CD4 (Taniuchi et al., 2002a). These proteins are members of the Runt domain transcription factor (Runx) family. There are three mammalian members of the Runt domain family, Runx1– 3, that encode a subunits that bind to the Runx binding motif (50 -PuACCACA-30 ) by making heterodimeric complexes with a common b subunit, Cbfb/Pebp2b (Wheeler et al., 2000). Fine mapping analysis in 1200M cell transfection assays revealed that the minimal functional sequence of site 2 (232GACCACA238) in the silencer core corresponded to a consensus Runx binding motif (Taniuchi et al., 2002a). A second Runx binding motif (site20 , 81AACCACA87) was identiﬁed in the

70

ICHIRO TANIUCHI ET AL.

50 ﬂanking region. Both of the Runx binding motifs were required for full CD4 silencer activity. Speciﬁc mutations of either site resulted in variegated CD4 de-repression in mature CD8þ cells, while mutations of both Runx motifs resulted in full CD4 de-repression in all mature CD8-lineage T cells, indicating that Runx sites are indispensable for establishment of epigenetic CD4 silencing. However, in immature DN thymocytes, the level of CD4 expression (or de-repression) when both Runx sites were mutated was lower than that upon deletion of the entire CD4 silencer (Taniuchi et al., 2002a). Thus, the function of Runx sites is likely to be partially compensated by other functional elements in immature DN thymocytes, but not in thymocytes developing into cytotoxic lineage cells, in which epigenetic CD4 silencing is being established. This difference in mechanism of repression or silencing is discussed in greater detail in the next section. Although transcripts for all three runx genes appear to be present in all thymocyte subsets, two of the three genes were shown to be required for regulating CD4 expression at distinct developmental stages (Taniuchi et al., 2002a). Runx1 was shown to be required for CD4 repression in immature DN thymocytes, while Runx3 was shown to be required for establishment of epigenetic CD4 silencing in CD8 lineage T cells. Because null mutations of runx1 result in an early lethal phenotype, the role of this gene in thymocytes was studied by using conditional inactivation of LoxP-ﬂanked runx1 with Cre recombinase expressed at different stages of development. When runx1 was inactivated early in DN thymocytes, using Lck–Cre transgenic mice, CD4 was de-repressed in DN cells to a level similar to that observed when both Runx binding sites were mutated (Taniuchi et al., 2002a). Mice homozygous for runx3 mutations die perinatally, but fetal thymic development can be examined. In Runx3-deﬁcient fetal thymi, CD4 repression in immature DN thymocytes was normal. However, RAG2-null host mice reconstituted with runx3-mutant progenitors displayed either full or variegated de-repression of CD4 in mature CD8-lineage cells, which differed from what was observed with the double Runx site mutations (Taniuchi et al., 2002a). The variegation in CD4 de-repression is likely due to a partial compensatory function of Runx1. Using a different strain of Runx3-deﬁcient mice, in which runx3 inactivation is tolerated beyond the perinatal stage, Groner and colleagues showed full derepression of CD4 in CD8 lineage T cells of mice with the compound runx1þ/ runx3/ mutant genotype (Woolf et al., 2003). The relative contribution of Runx1 to the establishment of epigenetic CD4 silencing remains unclear. There was no apparent loss of CD4 silencing in CD8 lineage cells when runx1 was inactivated at different stages of thymocyte differentiation (Taniuchi et al., 2002a). Runx1 inactivation in DN thymocytes of runx1f/f/ Lck-Cre mice severely blocked the generation of DP and also mature thymocytes and resulted in a reduced number of peripheral mature CD4þ T cells

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

71

(Taniuchi et al., 2002a). However, the efﬁciency of runx1 excision in mature T cells was lower than that in thymocytes, consistent with an essential role for Runx1 during selection events in the thymus. When runx1 was inactivated in DP thymocytes, in runx1f/f/CD4-Cre mice, positive selection was compromised, but there was no CD4 de-repression in those CD8þ T cells that reached the secondary lymphoid organs, despite very efﬁcient inactivation of runx1 in these cells (Egawa, Taniuchi, and Littman, unpublished). It is possible that Runx1 and Runx3 have equivalent abilities to initiate silencing of CD4 in cells destined for the CD8/cytotoxic T cell lineage. Because Runx3 appears to be expressed primarily in CD8 SP medullary thymocytes, while Runx1 is expressed at highest levels in immature cortical thymocytes, the loss of Runx3 would therefore more readily result in CD4 de-repression. A role for a mammalian SWI/SNF chromatin remodeling complex in CD4 silencing has also been proposed. CD4 de-repression was observed in immature DN thymocytes in which the function of the BAF complex, a mammalian counterpart of the yeast SWI/SNF complex, was compromised. Haploinsufﬁciency of Brg1 (ATPase subunit) compounded with expression of a dominant negative form of BAF57 (a DNA binding subunit containing an HMG-box) resulted in partial CD4 de-repression in immature DN thymocytes, and in CD4 de-repression in a minor proportion of mature CD8þ T cells (Chi et al., 2002). CD4 de-repression caused by impaired BAF function was enhanced in mice harboring a mutation in site 1 of the CD4 silencer, but this was observed only in immature DN thymocytes, not in mature CD8þ cells. Association of the BAF complex with the CD4 locus was demonstrated by chromatin immunoprecipitation (ChIP) assays in a T cell line, but it remains unclear whether the BAF complex directly binds to the CD4 silencer. Interestingly, the CD8 gene was poorly up-regulated in a proportion of cells corresponding to the DP stage when BAF function was compromised. Therefore, the BAF chromatin remodeling complex appears to function reciprocally at these two loci in early thymocyte differentiation: it contributes to repressing the CD4 locus and to activating the CD8 locus. Similarly, runx1 inactivation at the immature DN stage resulted in impaired CD8 up-regulation, in addition to the partial defect in CD4 repression. Although further genetic and biochemical studies are required, these results suggest that Runx1 functions at target loci to direct either repression or activation, in part by recruiting BAF complexes. D. Distinct Mechanisms of CD4 Silencing at Two Developmental Stages A common cis-regulatory element, the CD4 silencer, is clearly required for repression of CD4 both in DN thymocytes and in CD8 SP thymocytes that give rise to mature cytotoxic CD8þ T cells. The silencing mechanisms at these

72

ICHIRO TANIUCHI ET AL.

different stages of development are likely to share some general features, but they are also thought to be fundamentally different. The early silencing must be reversible, since thymocytes turn on CD4 following b-selection. In contrast, CD4 remains stably silenced in mature cytotoxic T cells (although there may be an exception in some activated human CD8þ T cells, in which CD4 can be re-expressed (Kitchen et al., 1998) ). Stable or heritable states of gene expression are governed by alterations in the structure of chromatin. Regulation of chromatin structure is achieved largely through post-translational modiﬁcation of histones, which results in recruitment of diverse protein complexes (Jenuwein and Allis, 2001). It has been proposed that CD4 silencing in DN thymocytes is the result of active repression, which is reversible and does not involve extensive remodeling of chromatin, whereas silencing in CD8 lineage thymocytes is irreversible, probably due to an imprint that alters the chromatin structure at the CD4 locus (Taniuchi et al., 2002b). The latter would result in formation of heterochromatin, which is inaccessible to the transcriptional activation machinery. Although there is as yet no direct evidence for differences in histone modiﬁcation or chromatin structure at the repressed CD4 locus in thymocytes at different stages of development, there are several lines of evidence that support this model, and these are discussed here. Targeted mutations within the core sequence of the silencer resulted in compromised CD4 silencing, manifested by partial de-repression of CD4 in both immature DN thymocytes and in mature CD8-lineage T cells. However, the patterns of CD4 de-repression were different at these stages: there was uniform de-repression in immature CD4CD8 DN thymocytes, but variegated de-repression in mature CD8-lineage cells (Taniuchi et al., 2002b). Variegated gene expression, or position effect variegation (PEV), is usually observed in transgenic mice or at sites adjacent to chromosomal translocations, and it is thought to be due to modiﬁcation of chromatin structure (Festenstein et al., 1999). The results suggested that, in DN thymocytes, CD4 is actively repressed by a combination of transcription factors. Loss of binding of any one of these factors would result in partial but uniform de-repression of CD4 expression (Fig. 3). In contrast, in cells committed to the CD8 lineage, loss of binding to any single site in the silencer would reduce the probability of recruiting a complex that modiﬁes chromatin during a critical developmental window. If the complex is recruited, then silencing is established and maintained. However, if the complex is not recruited during this time frame, the locus would not be modiﬁed and CD4 would be de-repressed (Fig. 3). Further analysis showed that CD4 silencing in mature cytotoxic T cells from mice with silencer mutations displayed the characteristic stochastic property attributed to heterochromatin-mediated gene silencing. Comparison of heterozygous and homozygous mice harboring a site 1 silencer mutation showed that the ‘‘on-or-off’’ status was determined stochastically on each allele.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

73

The variegated CD4 de-repression was already observed in mature (TCRhighHSAlow) thymocytes, indicating that the on or off status was likely to be determined in developing thymocytes after positive selection. Once cells succeeded in silencing CD4 in the absence of a binding site within the silencer, the silenced status was maintained stably through subsequent mitoses (Taniuchi et al., 2002b). This is a key feature indicating that the silencer functions epigenetically in CD8 lineage T cells. Proof that the silencer is required for initiation, but not for maintenance, of CD4 silencing in CD8þ T cells was provided by studies of mice in which the CD4 silencer could be deleted at different stages of development. Gene targeting in ES cells was used to ﬂank a 1.6 kb segment of murine cd4 intron 1, which includes the entire silencer, with two loxP sequences (Sil ﬂox allele) (Zou et al., 2001). Expression of Cre recombinase resulted in excision of the loxP-ﬂanked silencer. Similarly to the germline deletion, excision of the CD4 silencer in DP thymocytes, in CD4-Cre transgenic mice, resulted in derepression of CD4 in all mature CD8þ T cells. In contrast, in vitro deletion of the CD4 silencer in puriﬁed CD4CD8þ SP cells from spleen, by using a Cre-encoding retrovirus, did not result in CD4 de-repression, even after several mitoses (Zou et al., 2001). This clearly showed that the silencer is required for establishment, but not for the maintenance, of CD4 silencing. It is likely that a chromatin-modifying complex, recruited by silencer-binding factors during a developmental window as cells differentiate along the CD8 lineage, marks the CD4 locus with an epigenetic tag that is then inherited in the absence of the CD4 silencer. Insight into the nature of the temporal window during which silencing is established comes from mice in which the silencer was excised after cells committed to the CD8 lineage. The Sil ﬂox mice were bred to transgenic mice in which Cre was regulated by an enhancer (E81, see following text) from the murine cd8ab locus that was shown to direct expression speciﬁcally in mature (HSAlow) CD8-lineage thymocytes. Remarkably, CD4 was expressed in CD8þ T cells in which the silencer was deleted, even though CD4 expression had initially been extinguished as the cells progressed to the CD48þHSAhi stage following lineage commitment (Zou et al., 2001). This result suggested that a certain time window was necessary to complete chromatin modiﬁcation for epigenetic CD4 silencing. It will be of considerable interest to determine if the reversible CD4 silencing observed at this early stage following lineage commitment is mechanistically similar to that observed in DN thymocytes. In yeast, a model of stepwise chromatin modiﬁcation was proposed for heterochromatinization at the silent mating type loci. This is initiated by deacetylation of histone by a histone deacetylase (HDAC), followed by methylation of lysine 9 in histone 3 by histone methyl-transferase (HMT), and then recruitment of the yeast homologue of heterochromatin

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

75

protein-1 (HP-1), one of major heterochromatin structural components (Nakayama et al., 2001). In mice, HP-1 was shown to be involved in gene silencing at centromeric regions, because overdosage of HP-1b by use of a transgene enhanced the silencing effect in the peri-centromeric regions, and suppressed PEV (Festenstein et al., 1999). Taniuchi and colleagues showed that the HP-1b transgene suppressed CD4 de-repression in CD8þ T cells from mice harboring a mutation at site 3 in the CD4 silencer. However, the HP-1b transgene did not alter the level of uniform CD4 de-repression in immature CD4CD8 DN thymocytes (Taniuchi et al., 2002b). Therefore, overdosage of HP-1 enhanced CD4 silencing only in mature T cells, which is consistent with the involvement of this heterochromatin component in irreversible epigenetic silencing, at the late stage of thymocyte development, and not in reversible repression of CD4 in DN thymocytes. The distinct roles of Runx1 and Runx3 in silencing at different stages of thymocyte development could also be interpreted as evidence that unique mechanisms are involved at each stage. However, it is also equally likely that the requirement for Runx1 in DN thymocyte active repression and for Runx3 (with some contribution from Runx1) in CD8 SP thymocyte silencing is simply due to the expression patterns of these transcription factors. Runx proteins are known to recruit HDACs by interacting with several corepressor molecules, including Groucho/TLE, which binds to the highly conserved C-terminal sequence found in all Runt domain family members, as well as Sin3A Fig 3 A model for distinct mechanisms of silencer-mediated CD4 repression at two developmental stages. (A) In immature CD4CD8 DN thymocytes, several CD4 silencer binding factors, including the BAF chromatin remodeling complex, together recruit corepressor molecules, resulting in reversible transcriptional repression (active repression). Runx sites (site 2 and site 20 ) would be occupied by Runx1 at this stage. At the transitional stage from CD4þCD8þ DP thymocytes to CD4CD8þ SP cytotoxic-lineage T lymphocytes, lineage speciﬁc modiﬁcations of chromatin structure (small orange circle) are established through the function of distinct machinery (orange square) recruited by the CD4 silencer binding factor complex. We propose that this epigenetic modiﬁcation is preceded by the recruitment of a reversible complex similar to that found in DN thymocytes. Altered chromatin structure would contain HP-1 molecules and serve as a heritable mark for epigenetic maintenance of the silenced status. Runx sites would be occupied mainly by Runx3 at this stage. (B) Compromised silencer function due to a mutation of site 1 results in partial uniform de-repression of CD4 in immature DN thymocytes (left). During the transition from the CD4þCD8þ DP stage to CD8þ SP thymocytes, at which epigenetic silencing is established, the compromised silencer function results in one of two possible outcomes at the CD4 locus. In a fraction of the CD8-lineage T cells, modiﬁcation of chromatin is complete enough for epigenetic inheritance of the silenced status. However, in the rest of these cells, modiﬁcation of chromatin is not complete enough to shut off the CD4 gene. Following subsequent cell divisions, the amount of CD4 transcription would be dependent on the status of chromatin modiﬁcation and would thus vary. The mixture of cells harboring silenced or activated CD4 results in variegated CD4 de-repression. HP-1 contributes to the successful establishment of the epigenetic mark. (See Color Insert.)

76

ICHIRO TANIUCHI ET AL.

(Levanon et al., 1998; Lutterbach et al., 2000; Wheeler et al., 2000). It is therefore possible that the Runx proteins contribute to active repression of CD4 by recruiting HDACs in both DN and CD8 SP thymocytes. In the CD8 SP thymocytes, following transition from the HSAhi to the HSAlo stage, there is likely to be the additional recruitment of the epigenetic machinery, which is dependent on the presence of Runx factors and induces locus heterochromatinization that is stably maintained even in the absence of the CD4 silencer by epigenetic mechanisms. E. Link of CD4 Silencing to Lineage Specification A major incentive for studying signaling pathways that regulate coreceptor expression is the likelihood that these are shared with pathways involved in cell fate determination of DP thymocytes. For this reason, Runx3 was a good candidate for a factor involved not only in CD4 silencing as cells committed to the CD8 lineage, but also in speciﬁcation of the cytotoxic lineage. Indeed, there was a reduction in the number of CD8þ T cells in secondary lymphoid organs of RAG2-deﬁcient mice reconstituted with runx3/ fetal liver cells (Taniuchi et al., 2002a). Similar results were observed in viable runx3/ mice (Woolf et al., 2003). In the thymus, generation of mature (TCRhighHSAlow) CD8þ thymocytes was normal in animals reconstituted with the mutant progenitors but was reduced in the viable mutant mice, although not to the extent observed in the periphery. This phenotype was not due to a defect in lineage commitment, however. Although CD8þ T cells from secondary lymphoid organs of these mice had defective TCR-mediated proliferative responses, after activation with IL-2 they had normal cytotoxic activity and normal expression of perforin, a protein critical for CTL function (Taniuchi et al., 2002a). CD4þ cells in Runx3-deﬁcient mice had normal proliferative responses following TCR crosslinking or alloantigen stimulation. The reduction in number of CD8þ but not CD4þ T cells in secondary lymphoid organs is most likely due to defective homeostatic signaling in the former. Continuous interaction with MHC is required for survival of naive peripheral T cells (Goldrath, 2002), and loss of signaling in the absence of Runx3 would be expected to result in selective loss of the CD8þ T cells. Studies of the half-life of CD8þ versus CD4þ T cells in mice deﬁcient for Runx3 will be important to determine if this is indeed the mechanism accounting for the reduced cellularity. Runx3 is therefore likely to be essential for the TCR-mediated response of CD8þ T cells to antigen stimulation rather than for induction of cytotoxic functions. This may be a common feature of Runx function in T cells, since inactivation of runx1 during thymocyte differentiation resulted in both defective b-selection and positive selection, which require pre-TCR and TCRab signal transduction, respectively. It will be of interest to identify genes, other

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

77

than CD4, that are involved in TCR signaling and are regulated by Runx1 and Runx3 at these different stages of T cell development. It remains unclear how lineage speciﬁcity of CD4 silencing is regulated. Although Runx3 mRNA was detected in CD4 SP thymocytes, albeit at a lower abundance than in CD8 SP thymocytes (Taniuchi et al., 2002a), immunohistochemical analysis suggests that Runx3 protein is present only in CD8 SP and not in CD4 SP thymocytes (Woolf et al., 2003). Runx3 gene expression may therefore be sufﬁcient to induce CD4 silencing. Activation of runx3 transcription may be one component of the CD8 lineage program, and CD4 silencing would therefore be one of the consequences of this program. At this time, it is unclear if expression of Runx3 would be sufﬁcient to impose silencing in developing thymocytes. It is possible that other components that bind to the silencer (e.g., to sites 1, 3, and 4) are constitutively present, and that expression of Runx3 is the key factor that results in silencing. Alternatively, the other factors may likewise be regulated differentially in the CD4 versus CD8 lineage, either at the transcriptional, translational, or post-translational level, and may thus also regulate silencing in the CD8 SP thymocytes. It will be important to identify the other silencer-binding factors and also to characterize the mechanism by which Runx3 expression is regulated, as this may hold the key to the pathway that speciﬁes the cytotoxic T cell lineage. IV. Regulation of CD8 Gene Expression

In contrast to the CD4 coreceptor molecule, which is expressed as a monomer at the cell surface, CD8 is expressed either as a CD8aa homodimer or as a heterodimer formed by the CD8a and CD8b polypeptides. DP thymocytes and conventional TCRab cytotoxic T cells express CD8ab. The CD8a and CD8b genes are closely linked, which suggests that they are coordinately regulated in these cells (Gorman et al., 1988). However, since intraepithelial lymphocytes (IEL) from the gut (Jarry et al., 1990; Lefrancois, 1991) and CD8þ dendritic cells (DC) (Vremec et al., 1992) express only CD8aa homodimers, there must exist independent cis-regulatory elements speciﬁc either for CD8a and/or CD8b. A. Cis-regulatory Elements Involved in CD8 Regulation Early studies on the CD8a and CD8b promoters and on promoter-proximal regions failed to reveal CD8-speciﬁc cis-regulatory elements (for a review on CD8 promoters, see Ellmeier et al., 1999). Subsequent DNase hypersensitivity (DH) analyses over the entire CD8ab locus (referred to hereafter as the cd8 locus in mouse) led to the identiﬁcation of four clusters (I, II, III, and IV) of DH sites within an 80 kb murine genomic fragment (Hostert et al., 1997a)

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

79

and six clusters (I to VI) within a 95 kb human genomic fragment (Kieffer et al., 1997). Transgenic mice generated with these large genomic fragments displayed appropriate developmental stage- and lineage-speciﬁc expression of the transgenic CD8a and b genes, demonstrating that the major cis-regulatory elements are located within these fragments. Interestingly, mosaic expression of the transgene was observed, most likely due to position effect variegation (Festenstein et al., 1996). This suggests that a locus control region (LCR, a cis-acting element that mediates position-independent and copy numberdependent expression of a transgene; for review, see Festenstein and Kioussis, 2000) is either not required to facilitate expression of CD8a and CD8b in their endogenous context or was omitted from the transgenic constructs. Additional transgenic reporter expression assays were used to study and dissect the different DH clusters in more detail, and several genomic fragments involved in the regulation of CD8 expression were identiﬁed (Ellmeier et al., 1997, 1998; Hostert et al., 1998, 1997a,b; Kieffer et al., 1996, 1997; Zhang et al., 1998, 2001). At least four different genomic fragments isolated from the murine cd8 locus were able to direct expression in a developmental stage-, subset-, and lineage-speciﬁc mode (Fig. 4). Since the genomic fragments used in the transgenic assay contained DH sites, it is likely that the enhancer activities co-localize with the hypersensitivity sites. The ﬁrst enhancer identiﬁed, designated E8I or CIII-1,2, had activity only in mature CD8 SP thymocytes and in CD8þ T cells (Ellmeier et al., 1997; Hostert

Fig 4 Summary of characterization of cis-regulatory regions in the murine cd8 locus. Upper part: Schematic map of the cd8a and cd8b loci on mouse chromosome 6. Vertical arrows indicate individual DNase I hypersensitivity (DH) sites that have been grouped to DH clusters I–IV (Hostert et al., 1997a). Horizontal arrows indicate the location and transcriptional orientation of the cd8a and cd8b genes, while the open and closed bars indicate coding and noncoding exons, respectively. All BamHI (B), but only relevant EcoRI (E), sites are shown. The black horizontal bars show the genomic fragments used to deﬁne E8I, E8II, E8III, and E8IV. It is very likely that the enhancer activities within these genomic fragments overlap with some of the DH sites that map within the fragments. Middle part: Graphic representation of the genomic fragments used in transgenic reporter expression assays. The enhancer activity of the various fragments is shown at the right. The references reporting the activities are: T1 and T2 (Hostert et al., 1997a); T3 (Ellmeier et al., 1997; Hostert et al., 1997b); T4 (Hostert et al., 1997b); T5 (Hostert et al., 1998); T6 (Hostert et al., 1998; Zhang et al., 1998); T7–T10 (Ellmeier et al., 1998). þ indicates strong enhancer activity, þ/ weak activity, and – no enhancer activity. Nd: not determined. Lower part: The bars indicate the genomic region deleted in enhancer-deﬁcient mice. The expression of CD8 in the absence of the enhancer is shown at the right. The references reporting the enhancer deletions are: K1 (Ellmeier et al., 1998; Hostert et al., 1998); K2 and K3 (Ellmeier et al., 2002); K4 (Garefalaki et al., 2002). þ indicates normal CD8 expression, þ/ reduced CD8 expression, and – no CD8 expression. ‘‘Var’’ indicates variegated expression of CD8. Nd: not determined. (See Color Insert.)

80

ICHIRO TANIUCHI ET AL.

et al., 1997b). Its up-regulation coincided with the transition of CD8 lineage thymocytes from the TCRintHSAhi to the TCRhiHSAlo stage, which marks the transition of selected cells from the thymic cortex to the medulla. Another enhancer, E8III (CIV-3), directed expression only in immature DP thymocytes. Enhancer E8II (CIV-4,5) directed expression both in DP thymocytes and in CD8þ SP thymocytes and T cells (Ellmeier et al., 1998). Finally, the cisregulatory element E8IV (CIV-1,2) displayed low activity in CD4þ SP and mature T cells in addition to DP and CD8 lineage cells (the signiﬁcance of the low-level activity in helper T cells is currently not understood) (Ellmeier et al., 1998). While all the cis-acting elements were active in thymic-derived T cells (at least at certain developmental stages), only E8I also directed expression in IEL (Ellmeier et al., 1997), which have been proposed to be of extrathymic origin, although this remains controversial (Guy-Grand et al., 2003; Leishman et al., 2002). This suggested that E8I speciﬁcally regulates expression of CD8a (see following text). Furthermore, it was also shown that combinatorial interactions between cis-elements exist (Hostert et al., 1998). The combined activity of DH site cluster II, which by itself has no enhancer activity, and cluster III directs expression of a reporter gene not only in CD8þ T cells (as one would expect since cluster III, containing CIII-1,2(E8I), directs expression in the mature CD8þ T cell lineage), but also in DP thymocytes. Taken together, these studies indicate that lineage-speciﬁc regulation of CD8a and CD8b gene expression during T cell development is achieved through a complex regulatory network that utilizes several closely linked cis-regulatory elements with distinct developmental stage-, subset-, and lineage-speciﬁc functions (Fig. 4). The results from the transgenic reporter expression assays raise the question of why so many different, and probably partially redundant, cis-acting elements are required for the regulation of CD8 expression. One could argue that, in the context of the endogenous CD8 locus, the enhancers that show activity in the same subsets, such as E8I and E8II in CD8 SP and mature CD8þ T cells and E8II and E8III in DP thymocytes, are together required for high-level expression of CD8a and CD8b. When analyzed in transgenic reporter assays isolated from their genomic context, they therefore display similar activities. Another possible reason for the requirement of multiple cis-elements is that some of the enhancers may direct speciﬁcally the expression of either CD8a or CD8b within the thymus-derived T cell lineage. Because it is difﬁcult to experimentally test these possibilities in transgenic reporter assays, new insights have come from studies in mice with speciﬁc targeted mutations in cd8 regulatory sites. To investigate whether different elements in the cd8 locus have unique regulatory functions and to overcome the limitations of the transgenic reporter system, a systematic deletional analysis of the various enhancers and

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

81

cis-elements in the mouse germ line has been initiated. Based on the results from transgenic reporter expression assays, one would have predicted that (1) individual deletions of enhancers should (at least) lead to a reduction of CD8 expression in the subsets in which the enhancers show activity, and that (2) combined deletions of enhancers with similar T cell subset activities should lead to a more dramatic reduction of CD8 expression compared to individual deletions. Dependent on whether some of the enhancers are speciﬁc for CD8a or CD8b, one may also expect a reduction in the expression of only one of these coreceptor genes. These predictions have been recently tested for three cis-elements: the mature CD8þ T cell enhancer E8I, the DP and CD8 SP speciﬁc E8II, and DH site cluster II that contributes to DP-speciﬁc gene expression in transgenic mice. In agreement with the predictions based on the transgenic reporter studies, mice homozygous for deletion of E8I (D1/D1) had a 3- to 5-fold reduction of CD8a expression levels on IEL, particularly on the TCRgdþ cells (Ellmeier et al., 1998; Hostert et al., 1998). This indicates that E8I is the major regulatory element for CD8a expression on IEL. In contrast, even though E8I alone is sufﬁcient to direct expression in CD8 SP thymocytes and in mature CD8þ T cells, expression of CD8ab was largely normal in these thymus-derived cells in the D1/D1 mice. There was a very slight reduction (about 1.3-fold) in the CD8ab level in mature thymocytes, but not in peripheral T cells (Ellmeier et al., 1998). This may be signiﬁcant in the context of sequential functions of CD8 enhancers during thymocyte maturation, and will be discussed in greater detail. Mice with targeted deletions of E8II (D2/D2) had normal expression of CD8 in thymocytes and in both thymus-derived CD8þ T cells and IELs. Since, in transgenic mice, both E8I and E8II direct expression in mature CD8þ T cells, it was possible that these two enhancers could compensate for each other in this subset. In addition, in E8II-deﬁcient mice, E8III or cluster II (in combination with other regulatory regions) could similarly compensate to direct CD8 expression in DP thymocytes. These possibilities were tested by analyzing mice double-deﬁcient for E8I and E8II (D1D2/D1D2 mice). In contrast to individual deletions, deletion of E8I and E8II had a major effect on the expression of CD8 during thymocyte development, but not in mature CD8 lineage cells (Ellmeier et al., 2002). A population of ‘‘CD8-negative’’ DP thymocytes appeared that was indistinguishable from DP thymocytes by expression of other surface markers and by functional phenotype. Remarkably, a very similar phenotype with an even higher proportion of CD8-negative DP thymocytes was observed in mice with a deletion of DH cluster II (Garefalaki et al., 2002). The concurrent appearance of CD8-negative DP thymocytes and DP cells is consistent with variegation of expression of CD8 in the absence of either E8I/E8II or cluster II, which suggests that precursor cells in the mutant mice undergo

82

ICHIRO TANIUCHI ET AL.

stochastic establishment or loss of CD8 gene expression. As a consequence, fewer DP thymocytes were present and fewer mature CD8þT cells developed. These results revealed a novel function of the cis-regulatory elements and enhancers that cannot be predicted by transgenic reporter analyses. They additionally suggested that there is partial redundancy of enhancers involved in initiation of CD8 gene expression in ISP or DP thymocytes. Even in the absence of either both E8I and E8II or cluster II, a majority of thymocytes expressed CD8 in the DP compartment, suggesting that after initiation of expression, an epigenetic mechanism keeps the CD8 locus in an ‘‘on’’ conﬁguration. Surprisingly, expression of CD8 on those cells in which it was up-regulated (i.e., DP thymocytes and peripheral CD8þT cells) was only slightly reduced in the mutant mice. It will be necessary to perform additional targeted mutational analyses in mice to determine if enhancers other than E8I, E8II, and cluster II are required to initiate and sustain expression of CD8 in these cells, or, indeed, whether there is an epigenetic mechanism that renders mature cells independent of further function of these cis-regulatory elements. As has been mentioned, E8I directs reporter gene expression in CD8aaþ IEL in transgenic mice (Ellmeier et al., 1997) and its targeted deletion conﬁrmed that it is the major enhancer that directs expression of CD8a in this lineage (Ellmeier et al., 1998; Hostert et al., 1998). However, a subset of CD8aa-expressing TCRgdþ IEL in E8I-deﬁcient mice continued to express surface CD8aa homodimers, albeit at lower levels (Ellmeier et al., 1998; Hostert et al., 1998). Because none of the other identiﬁed enhancers directed transgene expression in IEL, a possible explanation remained that unknown enhancer elements could compensate for loss of E8I in this particular CD8aalowTCRgdþ IEL subset. Surprisingly, E8II, which did not show any activity in CD8aaþ IEL in transgenic mice (Ellmeier et al., 1998), was found to compensate, at least partially, for the loss of E8I. In D1D2/D1D2 mice, there was almost complete absence of CD8a expression in TCRgdþ IEL and a further reduction of CD8a expression in CD8aaþTCRabþ IEL compared to D1/D1 animals (Ellmeier et al., 2002). E8I and E8II are therefore the cisregulatory elements that direct expression of CD8a in CD8aa homodimerexpressing IEL of the TCRgd lineage. In CD8aaþ IEL of the TCRab lineage, additional elements must be able to direct low-level expression of CD8a. As has been mentioned, CD8aa homodimers are also expressed on splenic DC. However, neither E8I nor E8II was able to direct the expression of a reporter gene in DC (Jung, Ellmeier, and Littman, unpublished). In addition, CD8a expression on DC was unaltered in E8I and E8II single knockout mice or in E8I/E8II double-deﬁcient mice (Ellmeier and Littman, unpublished). Thus, the regulatory elements required for CD8a expression in DC remain to be identiﬁed.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

83

B. Trans-acting Factors and Chromatin Remodeling in CD8 Gene Regulation The results from the CD8 enhancer- and cluster II-deﬁcient mice suggested that the regulation of chromatin is an important step in the activation of CD8a and CD8b during T cell development (Kioussis and Ellmeier, 2002). More direct evidence that chromatin remodeling is indeed required in the regulation of CD8 expression comes from recent studies of the mammalian SWI/SNFlike BAF complex. In addition to its role in CD4 silencing (as has been described), the BAF complex also regulates the activation of CD8 and both BAF57 (an HMG box-containing factor) and Brg1 (an ATPase required for chromatin remodeling) have been implicated in this process (Chi et al., 2002). The ﬁrst indication came from the analysis of CD8a expression levels in BAF57DN transgenic animals. Mice heterozygous for the BAF57DN transgene expressed lower levels of CD8a on DP thymocytes compared to nontransgenic littermates, and this phenotype was even more pronounced in mice homozygous for the transgene. Furthermore, haplo-insufﬁciency of Brg1 (achieved by generating brg1þ/ mice) led to the appearance of a population of CD4þCD8CD3 thymocytes (which was not caused by de-repression of CD4 in DN cells). Although it was not formally demonstrated whether these thymocytes are phenotypically (except with respect to CD8 expression) and physiologically similar to DP cells, this population was reminiscent of the CD8-negative DP cells that develop in mice with either cluster II or combined E8I and E8II deletions. These results suggest a link between some of the cis-regulatory elements and the BAF complex. Another factor that has been implicated in the activation of the CD8a and CD8b loci is Ikaros, a transcription factor with multiple functions in hematopoiesis (Georgopoulos, 2002). A study in which antibodies speciﬁc for Ikaros were used to perform ChIP analysis revealed an association of Ikaros with sequences within the CD8 locus (Harker et al., 2002). Interestingly, the Ikaros-binding regions overlap with DH sites within DH clusters II and III. Furthermore, it was shown that haploinsufﬁciency of Ikaros causes an increase in the variegation of transgenic reporter genes that were driven by genomic fragments containing cluster II and/or cluster III. This effect seemed to be speciﬁc for CD8 regulatory elements, since the degree of variegation caused by control hCD2 regulatory elements was unaffected by Ikaros haploinsufﬁciency (Harker et al., 2002). In addition, the effect of Ikaros (and related family members such as Aiolos) was also seen at the endogenous cd8a and b loci. Compound mutations of the genes encoding Ikaros and Aiolos (Ikarosþ/ Aiolos/) led to the appearance of immature CD4þCD8 cells (Harker et al., 2002), reminiscent of the CD8-negative DP cells observed in enhancer-deﬁcient mice and in mice carrying mutations

84

ICHIRO TANIUCHI ET AL.

in members of the BAF complex. A similar population of cells has also been observed in mice in which runx1 was inactivated with the Lck-Cre transgene (Taniuchi et al., 2002a). Together, these ﬁndings suggest that Ikaros and Runx1 may bind to various enhancers in the CD8 locus, and that they recruit the BAF chromatin remodeling complex, making the locus accessible to the transcriptional machinery. Alternatively, BAFbinding may precede the recruitment of Ikaros, Runx1, and other factors, and may thus remodel chromatin to permit sequence-speciﬁc binding of these factors. Elucidation of the relationship between the transcription factors and the BAF chromatin remodeling complex awaits ChIP analysis at different stages of thymocyte differentiation with strains of mice bearing mutations within the cd8 locus or in genes encoding the various trans-acting factors. C. E8i in Memory Cell Differentiation: Is E8I a Target of TCR Signaling? A surprising new twist to the CD8 regulation story comes from the recent observation by Cheroutre and colleagues that a selected subset of conventional CD8abþ T cells transiently expresses CD8aa homodimers not only upon in vitro TCR stimulation but, more importantly, also upon in vivo activation with antigen (Madakamutil et al., 2004). This induction was revealed by staining with tetramers of the class Ib molecule TL1, which binds selectively to CD8aa homodimers and not CD8ab heterodimers. The excess synthesis of CD8a after induction presumably results in preferential formation of cell surface CD8aa homodimers due to the limiting concentration of CD8b, which is only found in heterodimers. Remarkably, the induction of CD8a was found to be required for survival of the activated effector cells that subsequently differentiate into memory CD8þ T cells. The induction of CD8aa homodimers upon activation of CD8abþ T cells was abrogated in E8I-deﬁcient mice both in vivo and in vitro, indicating that E8I is not only the major enhancer regulating expression of CD8aa on IEL, but is also required for up-regulation of CD8aa on cells destined to become memory T cells (Madakamutil et al., 2004). Although E8I-deﬁcient mice were able to mount a normal primary immune response against LCMV, they failed to develop a signiﬁcant population of LCMV-speciﬁc memory T cells. Thus, activation of E8I by TCR signaling in CD8þ CTL results in up-regulation of CD8aa, which is required for subsequent generation or survival of memory CTL. The upregulation of CD8aa is consistent with the observation of reduced CD8b expression on human memory CTL (Konno et al., 2002) and presumably occurs transiently in only a small subset of effector CD8þ T cells that are destined to differentiate into memory T cells.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

85

This surprising result calls for a reevaluation of the function of E8I, at least beyond the DP stage in the thymus. Expression of CD8aa homodimers in IEL is clearly dependent on E8I, and, as in memory CTL, this may require TCR-mediated signaling. In the thymus, E8I appears to contribute to induction of CD8 expression in DP thymocytes, and it also appears to function late after positive selection to ensure that the level of CD8 is optimal on SP cells. It is possible that at this stage, when cells transit from the TCRintHSAhi to the TCRhiHSAlo phenotype, and presumably migrate from the cortex to the medulla, E8I also functions in response to TCR-mediated signaling. In the context of the ‘‘CD8 reversal model’’ for lineage speciﬁcation (Brugnera et al., 2000), it is expected that CD8 is regulated at different stages of development. Following positive selection signaling, DP expression is extinguished, presumably due to cessation of E8III activity (and of other enhancers that cooperate with E8III). Cells selected by interaction with MHC class I would then undergo CD4 silencing, concomitant with activation of another set of CD8 enhancers, which has yet to be identiﬁed but which could include E8II and E8IV, both of which direct expression in DP and CD8 SP thymocytes. This would restore high levels of CD8ab heterodimer, and the potential for the TCR–coreceptor complex to resume signaling by interacting with self-MHC. We propose that this would result in a TCR-mediated signal that activates E8I (and, possibly, expression of genes required for migration of the cells from the cortex to the medulla). Finally, future studies will have to focus on the analysis of the developmental stage-speciﬁc regulation of enhancer and DH cluster activities and on their interplay to ensure proper expression of CD8. One could imagine that the accessibility of the enhancers for trans-factors is developmentally regulated and therefore determines whether an enhancer is active or not in a particular subset. Thus, the enhancer would contain information for the recruitment of chromatin remodeling activities at the proper developmental stage. Alternatively, a different set of trans-acting factors that are either repressed or induced at the onset of positive selection would regulate CD8 expression in immature versus mature thymocytes, respectively. Whether an enhancer is functional or not would be determined by the presence of the trans-acting factor. The identiﬁcation of the enhancer binding factors will certainly help to reveal which of the models (which are, of course, not mutually exclusive) is correct. D. Enhancer Specificity for CD8a versus CD8b A question that has not been directly addressed in the studies already described is how speciﬁcity of CD8a and b gene expression is achieved in different T cell lineages. Since the clusters of DH sites that contain at least ﬁve

86

ICHIRO TANIUCHI ET AL.

different cis-regulatory elements are in close proximity to each other and since CD8a and CD8b display T cell lineage-dependent differences in their expression pattern, a tight regulatory interaction between the CD8a and CD8b genes and the enhancers must exist. One could speculate that E8I, which is necessary and sufﬁcient to direct expression in CD8aaþ IEL (Ellmeier et al., 1997, 1998; Hostert et al., 1998), is likely to function only in conjunction with the CD8a promoter and may therefore be speciﬁc for regulating CD8a expression. However, additional transgenic expression assays indicate that E8I together with the CD8b promoter is able to direct expression of a reporter gene in IEL (Ellmeier and Littman, unpublished). Thus, incompatibility between E8I and the CD8b promoter does not seem to explain why E8I, which is active in IEL, fails to direct CD8b gene expression in this T cell lineage. It is more likely that an insulator element (Bell et al., 2001) may be localized between E8I and the CD8b promoter, thus preventing enhancer and promoter interactions. V. Conclusion

Close examination of how the CD4 and CD8 genes are regulated during T lymphocyte development, not only in the thymus but also in secondary lymphoid organs, has provided us with valuable insight into mechanisms of lineage speciﬁcation. Further studies in this area promise to uncover how helper versus cytotoxic T cells differentiate, and, possibly, even how effector T cells give rise to subsets of long-lived memory T cells with specialized activation properties. In a more general sense, examination of these loci has provided a unique opportunity to study epigenetic regulation in vertebrate development. The CD4 locus remains the only vertebrate gene currently shown to undergo developmentally regulated epigenetic silencing, and as such, it presents an ideal system for analyzing how heterochromatin is established during developmental processes in higher eukaryotes. Acknowledgments W. Ellmeier is supported by the Austrian Research Fund (FWF) and the START program of the Austrian Ministry of Education, Science, and Culture, and D. R. Littman is an Investigator of the Howard Hughes Medical Institute.

References Adlam, M., and Siu, G. (2003). Immunity 18, 173. Allen, R. D., 3rd, Kim, H. K., Sarafova, S. D., and Siu, G. (2001). Mol. Cell. Biol. 21, 3071. Baixeras, E., Huard, B., Miossec, C., Jitsukawa, S., Martin, M., Hercend, T., Auffray, C., Triebel, F., and Piatier-Tonneau, D. (1992). J. Exp. Med. 176, 327. Barndt, R. J., Dai, M., and Zhuang, Y. (2000). Mol. Cell. Biol. 20, 6677. Bell, A. C., West, A. G., and Felsenfeld, G. (2001). Science 291, 447.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

87

Brugnera, E., Bhandoola, A., Cibotti, R., Yu, Q., Guinter, T. I., Yamashita, Y., Sharrow, S. O., and Singer, A. (2000). Immunity 13, 59. Cantor, A. B., and Orkin, S. H. (2001). Curr. Opin. Genet. Dev. 11, 513. Chi, T. H., Wan, M., Zhao, K., Taniuchi, I., Chen, L., Littman, D. R., and Crabtree, G. R. (2002). Nature 418, 195. Chong, M. M., Cornish, A. L., Darwiche, R., Stanley, E. G., Purton, J. F., Godfrey, D. I., Hilton, D. J., Starr, R., Alexander, W. S., and Kay, T. W. (2003). Immunity 18, 475. Davis, C. B., and Littman, D. R. (1994). Curr. Opin. Immunol. 6, 266. Donda, A., Schulz, M., Burki, K., De Libero, G., and Uematsu, Y. (1996). Eur. J. Immunol. 26, 493. Duncan, D. D., Adlam, M., and Siu, G. (1996). Immunity 4, 301. Ellmeier, W., Sawada, S., and Littman, D. R. (1999). Annu. Rev. Immunol. 17, 523. Ellmeier, W., Sunshine, M. J., Losos, K., Hatam, F., and Littman, D. R. (1997). Immunity 7, 537. Ellmeier, W., Sunshine, M. J., Losos, K., and Littman, D. R. (1998). Immunity 9, 485. Ellmeier, W., Sunshine, M. J., Maschek, R., and Littman, D. R. (2002). Immunity 16, 623. Festenstein, R., and Kioussis, D. (2000). Curr. Opin. Genet. Dev. 10, 199. Festenstein, R., Sharghi-Namini, S., Fox, M., Roderick, K., Tolaini, M., Norton, T., Saveliev, A., Kioussis, D., and Singh, P. (1999). Nat. Genet. 23, 457. Festenstein, R., Tolaini, M., Corbella, P., Mamalaki, C., Parrington, J., Fox, M., Miliou, A., Jones, M., and Kioussis, D. (1996). Science 271, 1123. Garefalaki, A., Coles, M., Hirschberg, S., Mavria, G., Norton, T., Hostert, A., and Kioussis, D. (2002). Immunity 16, 635. Georgopoulos, K. (2002). Nat. Rev. Immunol. 2, 162. Gillespie, F. P., Doros, L., Vitale, J., Blackwell, C., Gosselin, J., Snyder, B. W., and Wadsworth, S. C. (1993). Mol. Cell. Biol. 13, 2952. Goldrath, A. W. (2002). Microbes Infect. 4, 539. Goldrath, A. W., and Bevan, M. J. (1999). Nature 402, 255. Gorman, S. D., Sun, Y. H., Zamoyska, R., and Parnes, J. R. (1988). J. Immunol. 140, 3646. Gorman, S. D., Tourvieille, B., and Parnes, J. R. (1987). Proc. Natl. Acad. Sci. USA 84, 7644. Guy-Grand, D., Azogui, O., Celli, S., Darche, S., Nussenzweig, M. C., Kourilsky, P., and Vassalli, P. (2003). J. Exp. Med. 197, 333. Hanna, Z., Simard, C., Laperriere, A., and Jolicoeur, P. (1994). Mol. Cell. Biol. 14, 1084. Harker, N., Naito, T., Cortes, M., Hostert, A., Hirschberg, S., Tolaini, M., Roderick, K., Georgopoulos, K., and Kioussis, D. (2002). Mol. Cell 10, 1403. Hayday, A., Theodoridis, E., Ramsburg, E., and Shires, J. (2001). Nat. Immunol. 2, 997. Hernandez-Hoyos, G., Sohn, S. J., Rothenberg, E. V., and Alberola-Ila, J. (2000). Immunity 12, 313. Hori, S., Nomura, T., and Sakaguchi, S. (2003). Science 299, 1057. Hostert, A., Garefalaki, A., Mavria, G., Tolaini, M., Roderick, K., Norton, T., Mee, P. J., Tybulewicz, V. L. J., Coles, M., and Kioussis, D. (1998). Immunity 9, 497. Hostert, A., Tolaini, M., Festenstein, R., McNeill, L., Malissen, B., Williams, O., Zamoyska, R., and Kioussis, D. (1997a). J. Immunol. 158, 4270. Hostert, A., Tolaini, M., Roderick, K., Harker, N., Norton, T., and Kioussis, D. (1997b). Immunity 7, 525. Izon, D. J., Punt, J. A., and Pear, W. S. (2002). Curr. Opin. Immunol. 14, 192. Jaenisch, R., and Bird, A. (2003). Nat. Genet. 33 (Suppl.), 245. Jarry, A., Cerf-Bensussan, N., Brousse, N., Selz, F., and Guy-Grand, D. (1990). Eur. J. Immunol. 20, 1097. Jenuwein, T., and Allis, C. D. (2001). Science 293, 1074. Kaneta, M., Osawa, M., Sudo, K., Nakauchi, H., Farr, A. G., and Takahama, Y. (2000). J. Immunol. 164, 256.

88

ICHIRO TANIUCHI ET AL.

Keefe, R., Dave, V., Allman, D., Wiest, D., and Kappes, D. J. (1999). Science 286, 1149. Keppler, O. T., Welte, F. J., Ngo, T. A., Chin, P. S., Patton, K. S., Tsou, C. L., Abbey, N. W., Sharkey, M. E., Grant, R. M., You, Y., Scarborough, J. D., Ellmeier, W., Littman, D. R., Stevenson, M., Charo, I. F., Herndier, B. G., Speck, R. F., and Goldsmith, M. A. (2002). J. Exp. Med. 195, 719. Kieffer, L. J., Bennett, J. A., Cunningham, A. C., Gladue, R. P., McNeish, J., Kavathas, P. B., and Hanke, J. H. (1996). Int. Immunol. 8, 1617. Kieffer, L. J., Yan, L., Hanke, J. H., and Kavathas, P. B. (1997). J. Immunol. 159, 4907. Killeen, N., and Littman, D. R. (1996). Curr. Top. Microbiol. Immunol. 205, 89. Killeen, N., Sawada, S., and Littman, D. R. (1993). EMBO J. 12, 1547. Kim, H. K., and Siu, G. (1998). Mol. Cell. Biol. 18, 7166. Kim, W. W., and Siu, G. (1999). J. Exp. Med. 190, 281. Kioussis, D., and Ellmeier, W. (2002). Nat. Rev. Immunol. 2, 909. Kitchen, S. G., Korin, Y. D., Roth, M. D., Landay, A., and Zack, J. A. (1998). J. Virol. 72, 9054. Konno, A., Okada, K., Mizuno, K., Nishida, M., Nagaoki, S., Toma, T., Uehara, T., Ohta, K., Kasahara, Y., Seki, H., Yachie, A., and Koizumi, S. (2002). Blood 100, 4090. Kronenberg, M., and Gapin, L. (2002). Nat. Rev. Immunol. 2, 557. Lefrancois, L. (1991). Semin. Immunol. 3, 99. Lefrancois, L., and Olson, S. (1994). J. Immunol. 153, 987. Leishman, A. J., Gapin, L., Capone, M., Palmer, E., MacDonald, H. R., Kronenberg, M., and Cheroutre, H. (2002). Immunity 16, 355. Leung, R. K., Thomson, K., Gallimore, A., Jones, E., Van den Broek, M., Sierro, S., Alsheikhly, A. R., McMichael, A., and Rahemtulla, A. (2001). Nat. Immunol. 2, 1167. Levanon, D., Goldstein, R. E., Bernstein, Y., Tang, H., Goldenberg, D., Stifani, S., Paroush, Z., and Groner, Y. (1998). Proc. Natl. Acad. Sci. USA 95, 11590. Li, E. (2002). Nat. Rev. Genet. 3, 662. Lutterbach, B., Westendorf, J. J., Linggi, B., Isaac, S., Seto, E., and Hiebert, S. W. (2000). J. Biol. Chem. 275, 651. Madakamutil, L. T., Christen, U., Lena, C. J., Wang-Zhu, Y., Attinger, A., Sundarrajan, M., Ellmeier, W., von Herrath, M. G., Jensen, P., Littman, D. R., and Cheroutre, H. (2004). Science, in press. Manjunath, N., Shankar, P., Stockton, B., Dubey, P. D., Lieberman, J., and von Andrian, U. H. (1999). Proc. Natl. Acad. Sci. USA 96, 13932. Miyazaki, T., Dierich, A., Benoist, C., and Mathis, D. (1996). Science 272, 405. Nakayama, J., Rice, J. C., Strahl, B. D., Allis, C. D., and Grewal, S. I. (2001). Science 292, 110. Nakayama, K., Yamamoto, R., Ishii, S., and Nakauchi, H. (1993). Int. Immunol. 5, 817. Salmon, P., Giovane, A., Wasylyk, B., and Klatzmann, D. (1993). Proc. Natl. Acad. Sci. USA 90, 7739. Sands, J. F., and Nikolic-Zugic, J. (1992). Int. Immunol. 4, 1183. Sawada, S., and Littman, D. R. (1991). Mol. Cell. Biol. 11, 5506. Sawada, S., and Littman, D. R. (1993). Mol. Cell. Biol. 13, 5620. Sawada, S., Scarborough, J. D., Killeen, N., and Littman, D. R. (1994). Cell 77, 917. Sebzda, E., Mariathasan, S., Ohteki, T., Jones, R., Bachmann, M. F., and Ohashi, P. S. (1999). Annu. Rev. Immunol. 17, 829. Singer, A. (2002). Curr. Opin. Immunol. 14, 207. Singh, H., DeKoter, R. P., and Walsh, J. C. (1999). Cold Spring Harb. Symp. Quant. Biol. 64, 13. Siu, G., Wurster, A. L., Duncan, D. D., Soliman, T. M., and Hedrick, S. M. (1994). EMBO J. 13, 3570. Siu, G., Wurster, A. L., Lipsick, J. S., and Hedrick, S. M. (1992). Mol. Cell. Biol. 12, 1592. Taniuchi, I., Osato, M., Egawa, T., Sunshine, M. J., Bae, S. C., Komori, T., Ito, Y., and Littman, D. R. (2002a). Cell 111, 621.

CD4/CD8 LINEAGE CHOICE: NEW INSIGHTS

89

Taniuchi, I., Sunshine, M. J., Festenstein, R., and Littman, D. R. (2002b). Mol. Cell 10, 1083. Tomita, K., Hattori, M., Nakamura, E., Nakanishi, S., Minato, N., and Kageyama, R. (1999). Genes Dev. 13, 1203. Triebel, F., Jitsukawa, S., Baixeras, E., Roman-Roman, S., Genevee, C., Viegas-Pequignot, E., and Hercend, T. (1990). J. Exp. Med. 171, 1393. von Boehmer, H., and Fehling, H. J. (1997). Annu. Rev. Immunol. 15, 433. von Boehmer, H., Teh, H. S., and Kisielow, P. (1989). Immunol. Today 10, 57. Vremec, D., Zorbas, M., Scollay, R., Saunders, D. J., Ardavin, C. F., Wu, L., and Shortman, K. (1992). J. Exp. Med. 176, 47. Weiss, A., and Littman, D. R. (1994). Cell 76, 263. Wheeler, J. C., Shigesada, K., Gergen, J. P., and Ito, Y. (2000). Semin, Cell Dev. Biol. 11, 369. Wolfer, A., Bakker, T., Wilson, A., Nicolas, M., Ioannidis, V., Littman, D. R., Lee, P. P., Wilson, C. B., Held, W., MacDonald, H. R., and Radtke, F. (2001). Nat. Immunol. 2, 235. Woolf, E., Xiao, C., Fainaru, O., Lotem, J., Rosen, D., Negreanu, V., Bernstein, Y., Goldenberg, D., Brenner, O., Berke, G., Levanon, D., and Groner, Y. (2003). Proc. Natl. Acad. Sci. USA 100, 7731. Workman, C. J., and Vignali, D. A. (2003). Eur. J. Immunol. 33, 970. Wurster, A. L., Siu, G., Leiden, J. M., and Hedrick, S. M. (1994). Mol. Cell. Biol. 14, 6452. Zhang, X. L., Seong, R., Piracha, R., Larijani, M., Heeney, M., Parnes, J. R., and Chamberlain, J. W. (1998). J. Immunol. 161, 2254. Zhang, X. L., Zhao, S., Borenstein, S. H., Liu, Y., Jayabalasingham, B., and Chamberlain, J. W. (2001). J. Exp. Med. 194, 685. Zhuang, Y., Cheng, P., and Weintraub, H. (1996). Mol. Cell. Biol. 16, 2898. Zou, Y. R., Sunshine, M. J., Taniuchi, I., Hatam, F., Killeen, N., and Littman, D. R. (2001). Nat. Genet. 29, 332.

This Page Intentionally Left Blank

advances in immunology, vol. 83

CD4/CD8 Coreceptors in Thymocyte Development, Selection, and Lineage Commitment: Analysis of the CD4/CD8 Lineage Decision ALFRED SINGER* AND REMY BOSSELUT{ {

*Experimental Immunology Branch Laboratory of Immune Cell Biology National Cancer Institute Bethesda, Maryland 20892

I. Introduction

The T cell arm of the adaptive immune system is composed mainly of T lymphocytes expressing ab T cell receptors (TCR) that recognize antigenic peptides presented by Major Histocompatibility Complex (MHC) encoded molecules. ab T cells can be divided into two major subpopulations based on whether their TCR recognize antigenic peptides presented by MHC class I (MHC-I) or MHC class II (MHC-II) molecules. Notably, ab T cells which recognize peptide/MHC-I complexes express CD8 coreceptor molecules and function mainly as cytolytic effector cells, while ab T cells which recognize peptide/MHC-II complexes express CD4 coreceptor molecules and function mainly as helper cells. The concordance among TCR MHC speciﬁcity, coreceptor expression, and function displayed by ab T cells is critical for competent functioning of the adaptive immune system and is established during differentiation in the thymus (Cosgrove et al., 1991; Grusby et al., 1991; Kaye et al., 1989; Koller et al., 1990; Kruisbeek et al., 1983, 1985; Marusic-Galesic et al., 1988; Sha et al., 1988; Teh et al., 1988; Zijlstra et al., 1990). This chapter will focus on the role of CD4/CD8 coreceptor molecules in thymocyte development and on potential mechanisms of the CD4/CD8 lineage decision. II. Early Thymocyte Development

A. T/B Lineage Decision Mature ab T cells are generated from a relatively small number of lymphocyte precursors that initially migrate into the thymus. Even though they have migrated into the thymus, these early lymphocyte precursors still have the potential to differentiate into either T or B cells (Izon et al., 2001, 2002b; Koch et al., 2001; Wilson et al., 2001). Which cell type they actually differentiate into is inﬂuenced by the cell fate determining molecule Notch (Izon et al., 2002a; Koch et al., 2001; Pui et al., 1999; Radtke et al., 1999; Schmitt and Zuniga-Pﬂucker, 2002). Notch is an unusual transmembrane 91 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

92

ALFRED SINGER AND REMY BOSSELUT

protein that is cleaved during biosynthesis so that it consists of two noncovalently associated protein fragments: an extracellular fragment containing the Notch extracellular domain and a membrane-bound fragment containing the Notch transmembrane/intracellular domains (Lai, 2002). Interaction of the Notch extracellular domain with its extracellular ligands triggers further enzymatic cleavage of the Notch intracellular fragment near the inner leaf of the plasma membrane, freeing the intracellular domain from the plasma membrane. The released Notch intracellular domain (Notch-IC) translocates to the nucleus where it activates gene transcription. Activation of Notch-1 in the thymus by interaction with its extracellular ligands Jagged and Delta-like induces lymphocyte precursors to differentiate into T cells (Schmitt and Zuniga-Pﬂucker, 2002); without Notch activation, lymphocyte precursors in the thymus differentiate into B cells (Izon et al., 2002a,b; Koch et al., 2001; Radtke et al., 1999). In fact, Notch-1 can promote differentiation of lymphocyte precursors into T-lineage cells even outside the thymus, in such nontraditional T cell sites as bone marrow and in vitro cultures (Allman et al., 2001; Pui et al., 1999; Schmitt and Zuniga-Pﬂucker, 2002). Reciprocally, interference with Notch activation in the thymus by ectopic expression of Notch ligand modiﬁers, such as Lunatic fringe, interferes with T-lineage commitment and results in the intrathymic accumulation of B cells (Koch et al., 2001). Thus, lymphocyte precursors adopt the B-lineage unless they are directed by activated Notch to differentiate into T cells. B. DN Thymocytes T cell differentiation in the thymus has been subdivided into discrete stages of development based on thymocyte expression of the coreceptor molecules CD4 and CD8 (Fig. 1). The earliest thymocytes are CD48 (DN, double negative), which then differentiate into CD4þ8þ (DP, double positive) cells, which, in turn, ﬁnally differentiate into mature CD4þ8 or CD48þ (SP, single positive) cells. During the DN stage of differentiation, early T lineage cells proliferate, express the Recombination Activating Genes (RAG-1 and RAG-2), initiate TCR gene rearrangements, and undergo selection to identify cells that have generated productive TCRb gene rearrangements. These events occur sequentially in developing DN thymocytes, permitting DN thymocytes to be further subdivided into developmental subsets based on surface expression of CD44 and CD25 (Godfrey and Zlotnik, 1993; Godfrey et al., 1993; Pearse et al., 1989). DN thymocyte subsets are named DN1 through DN4, in order of their appearance during development. The earliest T-committed cells in the thymus are DN1 and are identiﬁed as CD44þ25; DN2 thymocytes are CD44þ25þ; DN3 thymocytes are CD4425þ; and DN4 thymocytes are CD4425 and will differentiate into DP thymocytes (Fig. 1).

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

93

Fig 1 Simpliﬁed scheme of thymocyte development based on surface expression of CD4 and CD8 coreceptors. DN thymocytes can themselves be subdivided into subpopulations based on surface expression of CD44 and CD25.

The ﬁrst of two thymocyte proliferative phases occurs in DN1 and DN2 thymocytes prior to RAG expression and prior to TCR gene rearrangements (Kawamoto et al., 2003). This early proliferative phase is required to expand the number of precursor cells in which TCRb gene rearrangements occur so that as many different productive TCRb gene rearrangements as possible will be expressed in the developing thymocyte pool. Expression of RAG proteins occurs at the DN2 and DN3 stages of differentiation, during which three of the four TCR gene loci (b, g, d) undergo rearrangements. Cells successfully rearranging both TCRg and TCRd gene loci are diverted from the ab lineage to differentiate into gd T cells (Bonneville et al., 1989). Cells successfully rearranging the TCRb gene locus and expressing TCRb protein are identiﬁed at the DN3 stage by a selection event known as ‘‘b-selection’’ that signals continued differentiation. DN3 cells express the invariant signaling components of the TCR complex CD3g, d, e, and zeta which, in mature T cells, require association with both TCRa and TCRb chains to be assembled into a functional signaling complex (Frank et al., 1990). However, in place of TCRa chains, which are generally not expressed in DN thymocytes, DN3 cells express an invariant chain, pre-Ta, (Fehling et al., 1995; Saint-Ruf et al., 1994) that functions as a surrogate for TCRa and can assemble with TCRb to form the ‘‘pre-TCR’’ signaling complex that is transported to the cell surface where it transduces survival signals by, in part, activating NFkb (Aifantis et al., 2001, 2002; Voll et al., 2000). Signaling by pre-TCR complexes appears to require only export from the ER to the cell surface and is thought to occur without ligand engagement (Haks et al., 2003; Irving et al., 1998; O’Shea et al.,

94

ALFRED SINGER AND REMY BOSSELUT

1997). The signals generated by pre-TCR complexes induce differentiation into DN4 stage cells which then undergo a second proliferative burst prior to expression of CD4/CD8 coreceptor proteins and differentiation into DP thymocytes. This second round of thymocyte proliferation assures that each productive TCRb gene rearrangement is replicated in multiple daughter cells prior to initiation of TCRa gene rearrangement in late DN4 and early DP thymocytes. As a result, each TCRb protein has a variety of potential partner TCRa proteins. Thus, the ﬁrst proliferative expansion at the DN1 and DN2 stages increases the diversity of TCRb proteins represented in the repertoire, and the second proliferative expansion at the DN4 stage increases the diversity of potential TCRa proteins that can partner with each TCRb protein. The result is a clonally diverse population of immature DP thymocytes that expresses a hugely diverse ab TCR repertoire. C. DP Thymocytes As a consequence of b-selection, DP thymocytes undergoing TCRa gene rearrangements already contain TCRb proteins. For TCRa, proximal gene rearrangements can be followed by distal gene rearrangements with the result that DP thymocytes have the potential to sequentially rearrange the same TCRa gene locus several times (Guo et al., 2002; Petrie et al., 1993). Productive gene rearrangements result in TCRa proteins that assemble with preexisting TCRb proteins to form functional ab TCR signaling complexes on the surface of DP thymocytes (Kearse et al., 1995). These ab TCR complexes are screened for their ability to engage self peptide/MHC complexes expressed on thymic stromal elements by whether they subsequently transduce a ligandstimulated intracellular signal. Since RAG expression is not terminated in DP thymocytes until TCR signals are transduced (Brandle et al., 1992; Takahama and Singer, 1992; Turka et al., 1991), TCRa gene rearrangements can continue until a TCRa protein is produced that forms an ab TCR complex which can engage intrathymic peptide/MHC complexes with sufﬁcient afﬁnity to stimulate a TCR signal (Buch et al., 2002; Wang et al., 1998). Interestingly, TCR gene rearrangements appear to generate ab TCR that may exhibit an intrinsic bias toward MHC recognition (Zerrahn et al., 1997), which may provide part of the explanation for why MHC complexes are virtually unique among membrane proteins in their ability to stimulate TCR signals in developing DP thymocytes. Initial ab TCR signals in DP thymocytes upregulate CD5 and CD69 surface expression and terminate RAG expression, which precludes further TCRa gene rearrangements and ﬁxes ab TCR speciﬁcity (Bhandoola et al., 1999; Dutz et al., 1995; Swat et al., 1993; Takahama and Singer, 1992; Turka et al., 1991). Unless rescued by ab TCR signals, DP thymocytes undergo apoptosis within 1 to 3 days by a process referred to as ‘‘death by

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

95

neglect’’ (Egerton et al., 1990; Huesmann et al., 1991). So DP thymocytes that either fail to express any ab TCR complexes or express ab TCR with too low afﬁnity for intrathymic peptide/MHC ligands to stimulate TCR signals undergo death by neglect (Kisielow et al., 1988; Mombaerts et al., 1992; Philpott et al., 1992). In contrast, DP thymocytes whose ab TCR have too high afﬁnity for intrathymic peptide/MHC ligands are signaled to undergo apoptosis by ‘‘negative selection’’ (Kappler et al., 1987; Sha et al., 1988; Takahama et al., 1992; Teh et al., 1989). It is only those few DP thymocytes whose ab TCR have intermediate afﬁnity for intrathymic peptide/MHC ligands that are signaled to survive and to further differentiate into SP thymocytes, by a process referred to as ‘‘positive selection’’ (Ashton-Rickardt et al., 1994; Hogquist et al., 1994; Kisielow et al., 1988; Singer et al., 1986, 1987; Teh et al., 1988). Notably, since positive selection of DP thymocytes involves differentiation into either CD4 or CD8 SP T cells, the CD4/CD8 lineage decision is an intrinsic feature of the positive selection process (Janeway, 1988; Kaye et al., 1989; Sha et al., 1988; Teh et al., 1988). III. CD4 and CD8 Coreceptor Molecules on DP Thymocytes

The CD4 and CD8 coreceptor molecules are non-rearranging transmembrane proteins whose extracellular domains bind MHC and whose intracellular domains associate with either the nonreceptor protein tyrosine kinase Lck (Glaichenhaus et al., 1991; Marth et al., 1985, 1989; Shaw et al., 1989, 1990; Turner et al., 1990; Veillette et al., 1988, 1989) or the adapter molecule LAT (Bosselut et al., 1999; Zhang et al., 1998). CD4 and CD8 coreceptor binding to MHC is not peptide speciﬁc, and is distinguished by the fact that CD4 has low afﬁnity for MHC class II complexes while CD8 has low afﬁnity for MHC class I complexes (Doyle and Strominger, 1987; Norment et al., 1988; Wyer et al., 1999). The CD4 and CD8 molecules are considered TCR coreceptors for several reasons (Dembic et al., 1987; Gabert et al., 1987; Janeway, 1988): (1) by virtue of their MHC binding afﬁnity, CD4 and CD8 coreceptors can augment TCR binding to peptide/MHC complexes; (2) because of their intracellular association with Lck, CD4 and CD8 coreceptors can augment the initiation of TCR signal transduction stimulated by peptide/MHC complexes; and (3) as a result of their intracellular association with LAT, CD4 and CD8 coreceptors can augment the propagation of downstream TCR signaling events. Despite their shared ability to function as TCR coreceptors, CD4 and CD8 are structurally distinct molecules. CD4 is a monomer while CD8 is a disulﬁde-linked dimer composed either of two different chains (CD8a and CD8b) which form CD8ab heterodimers or two similar chains which form CD8aa homodimers (Ledbetter and Seaman, 1982; Ledbetter et al., 1981);

96

ALFRED SINGER AND REMY BOSSELUT

CD8bb homodimers are never expressed on the murine cell surface as they are retained within the endoplasmic reticulum (Blanc et al., 1988; Devine et al., 2000; Gorman et al., 1988; Norment and Littman, 1988). The majority of CD8 surface complexes on DP thymocytes consist of CD8ab heterodimers, although CD8aa homodimers are also present on the cell surface (Norment and Littman, 1988). In mouse thymocytes, CD8 heterogeneity is further increased by expression of both full-length and alternatively spliced CD8a RNAs, which encode full-length CD8a and truncated CD8a0 proteins lacking a cytosolic tail (Zamoyska et al., 1985). So CD8 surface complexes on murine DP thymocytes potentially consist of 5 different protein complexes composed of disulﬁde-linked CD8ab, CD8a0 b, CD8aa, CD8a0 a, and CD8a0 a0 chains. Intracellular Lck binding to CD4 and CD8 coreceptor molecules is mediated by a dicysteine motif that is present in the cytosolic tails of both CD4 and CD8a, but absent from CD8b and CD8a0 (Glaichenhaus et al., 1991; Shaw et al., 1989, 1990; Turner et al., 1990; Veillette et al., 1988, 1989). LAT binds to the same or overlapping intracellular sites as Lck on CD4 and CD8a, with the result that Lck and LAT compete with each other for coreceptor binding (Bosselut et al., 1999; Wiest et al., 1996) with Lck binding predominating over that of LAT. Interestingly, more Lck and LAT are bound to CD8ab than to CD8aa surface complexes despite the latter complexes having twice as many potential binding sites (Bosselut et al., 1999; Irie et al., 1995, 1998). Either the CD8b cytosolic tail somehow improves binding of Lck and LAT to the CD8a tail, or two CD8aa chains structurally interfere with one another and diminish Lck’s ability to bind to either chain. In any event, only three of the ﬁve different CD8 complexes expressed on murine DP thymocytes (CD8ab, CD8aa, CD8a0 a) have any ability to bind intracellular Lck, and none of them binds Lck as well as CD4 does. Despite large intracellular pools of Lck in DP thymocytes, the pool of Lck that is available to surface CD4 and CD8 is limiting (Wiest et al., 1993, 1996). It has been estimated that there are four times as many CD8 complexes as CD4 molecules on the surface of DP thymocytes, yet CD4 surface molecules bind quantitatively more Lck than do CD8 surface complexes because Lck occupancy of surface CD4 molecules is 10-fold greater than Lck occupancy of surface CD8 complexes (Wiest et al., 1993). Fully 25 to 50% of surface CD4 molecules on DP thymocytes are associated with intracellular Lck, while only 2.5% of surface CD8 complexes on the same DP thymocytes are associated with intracellular Lck (Wiest et al., 1993). There are several explanations for the signiﬁcantly higher Lck occupancy of surface CD4 than surface CD8 molecules, including: (1) greater afﬁnity of Lck for the tail of CD4 versus CD8a; (2) heterogeneity of surface CD8 complexes, some of which include tailless CD8a0 molecules which cannot bind any Lck at all; and (3) the fact that CD8 surface molecules without associated intracellular Lck can remain on the

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

97

cell surface while CD4 surface molecules without associated intracellular Lck tend to be internalized and removed from the cell surface (Pelchen-Matthews et al., 1991, 1992). The consequence of higher Lck occupancy of CD4 than CD8 coreceptors is that extracellular CD4 engagements induce Lck aggregation and Lck kinase activation far more efﬁciently than do extracellular CD8 engagements (Wiest et al., 1993, 1996). Because experiments examining CD4/CD8 lineage fate may manipulate the amount of CD4 and CD8 on the cell surface, it is important to consider the effects that quantitative changes in coreceptor expression might have on coreceptor function in DP thymocytes. Since the amount of Lck available to surface coreceptors is limiting, surface CD4 and CD8 compete with one another for intracellular Lck binding. As a result, increases or decreases in surface expression of either CD4 or CD8 can alter the Lck signaling competence of either or both coreceptor molecules. Because Lck binds most efﬁciently to CD4, changes in CD4 surface expression have the most pronounced impact on Lck signaling in DP thymocytes. Quantitative decreases in CD4 surface expression increase coreceptor signaling by CD8, as revealed by the observation that absent CD4 expression in CD48 mice results in signiﬁcantly increased Lck association with surface CD8 complexes and signiﬁcantly increased CD8 signaling ability (Wiest et al., 1996). Paradoxically, while quantitative increases in CD4 surface expression decrease CD8 coreceptor signaling (Teh et al., 1991; van Oers et al., 1992; Wiest et al., 1996), marked transgenic CD4 overexpression increases the number of surface CD4 molecules that lack associated intracellular Lck and therefore decreases CD4 coreceptor signaling competence, as extracellular CD4 engagements are now less efﬁcient at aggregating and activating intracellular Lck (Nakayama et al., 1993). While quantitative changes in CD4 expression have the greatest impact on coreceptor signaling function, transgenic overexpression of CD8ab heterodimers increases the relative distribution of available Lck toward CD8, improving CD8 signaling ability (Bosselut et al., 2000, 2001; Chan et al., 1994; Robey et al., 1992). Thus, experimental manipulations that alter the number of either CD4 or CD8 coreceptor molecules on the cell surface can affect the signaling function of both coreceptor molecules. DP thymocytes in the thymus are in close physical contact with cortical thymic epithelial cells that express exceptionally high levels of MHC-II (Jenkinson et al., 1981; Van Ewijk et al., 1980). The high levels of MHC-II engage surface CD4 molecules on DP thymocytes even without concurrent TCR engagement, resulting in ‘‘constitutive’’ low level Lck activation (Nakayama et al., 1989, 1990; van Oers et al., 1994; Wiest et al., 1993). One consequence of such Lck kinase activation by intrathymic CD4/MHC-II interactions is Lck depletion, resulting in decreased amounts of Lck available to surface coreceptors and decreased amounts of Lck available for subsequent

98

ALFRED SINGER AND REMY BOSSELUT

TCR signaling (Wiest et al., 1996). CD4-induced Lck kinase activation also directly affects the structure of surface TCR complexes on DP thymocytes by inducing the phosphorylation of some TCR ITAMs, which result in the p21 form of phospho-zeta rather than the p23 form of phospho-zeta that has been associated with full TCR activation (Madrenas et al., 1995; Nakayama et al., 1989; van Oers et al., 1994). The phosphorylated TCR ITAMs on DP thymocytes recruit ZAP-70 molecules which are not themselves phosphorylated and so remain enzymatically inactive. It has been estimated that over one-third of surface TCR complexes on DP thymocytes contain phosphorylated ITAMs that have been bound by inactive ZAP-70 molecules (Wiest et al., 1996). Such surface TCR complexes on DP thymocytes resemble nonfunctional TCR complexes on mature T cells after engagement of antagonist ligands (Madrenas et al., 1995; Sloan-Lancaster et al., 1994). Indeed, structural changes in surface TCR induced by intrathymic CD4/MHC-II diminish TCR signaling competence in DP thymocytes, at least in response to strong ligands. It has been suggested that these same structural changes may actually increase TCR responsiveness to weak TCR ligands and so may promote positive selection of DP thymocytes in response to intrathymic self-peptide/ MHC complexes (Davey et al., 1998; Lucas et al., 1999; Stefanova et al., 2003). In addition to affecting TCR structure, intrathymic CD4/MHC-II interactions downregulate the number of TCR complexes present on the surface of DP thymocytes (Bonifacino et al., 1990; Cosgrove et al., 1991; Grusby et al., McCarthy et al., 1988; Nakayama et al., 1990; Wiest et al., 1993). Developing DP thymocytes express only 10 to 20% of the number of surface TCR complexes that are on mature SP T cells. Low surface TCR expression on DP thymocytes is primarily a result of low TCR assembly rates, which, in turn, are due to instability of newly synthesized TCRa proteins in the endoplasmic reticulum of immature DP thymocytes (Kearse et al., 1994a,b). Curiously, most of the TCRa proteins that are synthesized in DP thymocytes are degraded before they can be assembled into functional TCR complexes. While it is not clear if Lck activation affects the stability of nascent TCRa proteins in the endoplasmic reticulum of DP thymocytes, activation of Lck by intrathymic CD4/MHC-II interactions clearly contributes to low surface TCR expression, as removal of CD4/MHC-II interactions results in a 2.5-fold increase in surface TCR levels on DP thymocytes (Cosgrove et al., 1991; Grusby et al., 1991; Nakayama et al., 1990; Riberdy et al., 1998; Wiest et al., 1993). Interestingly, CD4/MHC-II interactions are not the only intrathymic interactions that downregulate surface TCR expression on DP thymocytes, as the molecule Src-like Adapter Protein (SLAP) has been recently implicated in downregulation of surface TCR expression on DP thymocytes, perhaps by affecting Lck (Sosinowski et al., 2001).

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

99

To summarize, CD4 and CD8 coreceptor molecules are both expressed on the surface of DP thymocytes and have the potential to activate intracellular Lck molecules in response to MHC binding. Because the amount of Lck available to coreceptors on DP thymocytes is limiting, CD4 and CD8 coreceptor molecules compete with one another for Lck binding. As a result, alterations in surface levels of one coreceptor molecule have the potential to affect the Lck signaling ability of the other coreceptor molecule. CD4 coreceptor molecules on DP thymocytes can engage intrathymic MHC-II molecules without concurrent TCR engagements and these engagements result in low-level Lck kinase activation, which downregulates TCR surface levels and decreases TCR signaling competence, at least to strong TCR ligands. IV. Selection and Commitment : Classical Models

DP thymocytes are the ﬁrst T-lineage cells to express ab TCR complexes, and they are unique among T-lineage cells in expressing both CD4 and CD8 coreceptor molecules. The ab TCR complexes expressed on DP thymocytes are the result of random TCR gene rearrangements that, in aggregate, constitute a hugely diverse ‘‘preselection’’ TCR repertoire. Of the different TCR speciﬁcities expressed in the preselection repertoire, only a relatively small number are thought to have a structure suitable for engaging peptide/MHC complexes, and fewer still have the potential to engage the peptide/ MHC complexes that are actually present in any given thymus (Bhandoola et al., 1999; Merkenschlager et al., 1997). The process of positive selection identiﬁes those relatively few DP thymocytes that express useful TCR speciﬁcities and promotes their differentiation into mature T cells. The efﬁciency of identifying DP thymocytes with potentially useful TCR speciﬁcities is signiﬁcantly improved by CD4 and CD8 coreceptors, as these coreceptor molecules augment the ability of TCR to engage intrathymic peptide/MHC complexes and augment the ability of TCR to transduce intracellular signals (Fung-Leung et al., 1991; Rahemtulla et al., 1991). However, the process of positive selection not only identiﬁes DP thymocytes expressing potentially useful TCR, it must also promote the differentiation of DP thymocytes into mature SP T cells that are of the correct T cell lineage, and so express the appropriate or ‘‘matching’’ coreceptor molecule. How the MHC speciﬁcity of the expressed TCR ultimately dictates CD4/CD8 lineage choice is the central issue of the CD4/CD8 lineage problem. A priori, conversion of DP thymocytes into SP T cells during positive selection need only involve transcriptional termination of one or the other coreceptor gene. In classical models of lineage commitment, transcriptional termination in DP thymocytes of one coreceptor gene is thought to be irreversible

100

ALFRED SINGER AND REMY BOSSELUT

and is equated with CD4/CD8 lineage commitment itself. Classically, DP thymocytes that have terminated CD8 gene transcription are considered to be CD4-committed cells that can only differentiate into CD4SP T cells, and DP thymocytes that have terminated CD4 gene transcription are considered to be CD8-committed cells that can only differentiate into CD8þ T cells (Barthlott et al., 1997a; Benveniste et al., 1996; Chan et al., 1993; Davis et al., 1993; Suzuki et al., 1995; Teh et al., 1988). Lineage-committed cells that are unable to complete their appropriate differentiation program, for whatever reason, are thought to undergo programmed cell death. So, the classical view of CD4/CD8 lineage choice holds that TCR signals in DP thymocytes terminate transcription of one coreceptor gene and irreversibly commit DP thymocytes to differentiating into SP T cells of the opposite coreceptor lineage. Of the two classical models of lineage commitment—instruction and stochastic/selection—instruction is the conceptually simpler one (Robey et al., 1991). In its original version, the instruction model postulated that signals transduced by TCR þ CD4 coengagements were qualitatively distinct from those transduced by TCR þ CD8 coengagements and so instructed DP thymocytes to terminate expression of different coreceptor genes. The most widely accepted current view of instruction is that proposed by Robey and colleagues who suggested that signals transduced by TCR þ CD4 versus TCR þ CD8 coengagements differ quantitatively, not qualitatively, from one another and that it is quantitative differences in ‘‘signal strength’’ that identify which coreceptor gene to terminate (Itano et al., 1996). According to the strength of signal view, TCR þ CD4 coengagements generate relatively strong signals, while TCR þ CD8 coengagements generate relatively weak signals; and strong signals induce CD4-commitment and terminate CD8 gene transcription, while weak signals induce CD8-commitment and terminate CD4 gene transcription. The end result is the differentiation of DP thymocytes into SP T cells whose TCR and coreceptor molecules have ‘‘matching’’ MHC speciﬁcities (i.e., CD4þ T cells with MHC-II speciﬁc TCR and CD8þ T cells with MHC-I speciﬁc TCR). According to the instruction model, every potentially useful TCR speciﬁcity in the preselection repertoire may ﬁnd its way into the mature TCR repertoire (Itano and Robey, 2000). In contrast to instruction, the stochastic/selection model postulates that TCRþcoreceptor signals do not specify lineage direction (Chan et al., 1993, 1994; Corbella et al., 1994; Correia-Neves et al., 2001; Davis et al., 1993; Itano et al., 1994; Robey et al., 1994). Instead, differentiation of DP thymocytes into SP T cells with matching TCR and coreceptor molecules is the result of two distinct signaling steps, the ﬁrst of which occurs in DP thymocytes and the second of which occurs in immature SP (i.e., ‘‘intermediate’’) thymocytes. In the ﬁrst step, TCRþcoreceptor signals induce DP thymocytes to randomly terminate one of the two coreceptor genes and to differentiate into short-lived

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

101

intermediate cells. In the second step, TCRþcoreceptor signals induce shortlived intermediate cells to differentiate into long-lived mature SP T cells; in the absence of TCRþcoreceptor signals, short-lived immature SP thymocytes fail to survive. Thus, according to the stochastic/selection model, TCR-mediated survival signals can only be generated in intermediate thymocytes with matching TCR and coreceptor molecules; intermediate thymocytes with mismatched TCR and coreceptor molecules (i.e., CD4 with MHC-I speciﬁc TCR and CD8 with MHC-II speciﬁc TCR) fail to receive TCR-mediated survival signals and undergo cell death. The end result of TCR-mediated positive selection and rescue is the generation of mature T cells with matching TCR and coreceptor molecules. Because stochastic/selection postulates that lineage choice is a random event for each positively selected thymocyte, approximately 50% of immature SP or intermediate thymocytes would express mismatched TCR and coreceptor molecules, with the result that the TCR speciﬁcities randomly expressed on mismatched intermediate thymocytes are irretrievably lost from the mature TCR repertoire. A. Support for the Strength of Signal Instructional Model TCR signals are generated in DP thymocytes by engagement of either peptide/MHC-I or peptide/MHC-II complexes, but TCR complexes themselves are presumably unable to identify whether they have engaged peptide/ MHC-I or peptide/MHC-II complexes. Consequently, information about MHC class speciﬁcity is most likely provided by coengagement of TCR with either CD4 or CD8 coreceptor molecules, as these coreceptor molecules do distinguish MHC-I from MHC-II complexes and do quantitatively differ from one another in their augmentation of TCR signal transduction. Original support for the strength of signal instructional model came from experiments utilizing coreceptor transgenes that encode chimeric CD8a molecules whose cytosolic tails were replaced with those of CD4, with the result that chimeric CD8/CD4 coreceptor molecules were signiﬁcantly more effective in inducing intracellular Lck kinase activity than normal CD8 molecules (Itano et al., 1996; Seong et al., 1992). Developmentally, such chimeric CD8/ CD4 coreceptor molecules appeared to direct thymocytes expressing an MHC-I restricted TCR transgene to differentiate into CD4þ T cells; whereas, in the presence of transgenic wildtype (rather than chimeric) CD8 coreceptor molecules, the same TCR transgenic thymocytes differentiated into CD8þ T cells. These experiments revealed that lineage choice was not strictly dictated by the MHC class speciﬁcity of the TCR, but could be importantly inﬂuenced by the intracellular tail of the associated coreceptor molecule. However, the CD4 coreceptor tail did not appear to transduce instructional signals because thymocytes expressing low afﬁnity MHC-I restricted TCR speciﬁc for HY failed to differentiate into CD4þ T cells even when their

102

ALFRED SINGER AND REMY BOSSELUT

TCR were coengaged with chimeric CD8/CD4 coreceptor molecules (Itano et al., 1996). Thus, Robey and colleagues proposed that it was the overall strength of signaling by coengaged TCR and coreceptor molecules that determined lineage choice, with quantitative differences in coreceptor/Lck associations being a major determinant of lineage choice. The strength of signaling concept was also supported by several experiments manipulating intracellular kinase activity in DP thymocytes. Csk is a negative regulator of Src family protein kinases such as Lck and Fyn, and experiments in Csk conditional knockout mice revealed that thymocyte differentiation in the absence of Csk expression was uncoupled from TCR signaling as a result of unregulated Lck and Fyn activity (Schmedt et al., 1998). Importantly, Csk-deﬁcient thymocytes differentiated solely into CD4þ T cells, regardless of the TCR they expressed or if they even expressed any TCR at all, indicating that uncontrolled Lck/Fyn kinase activity selectively promotes DP thymocyte differentiation into CD4þ T cells. Other experiments directly examining the inﬂuence of Lck activity on the CD4/CD8 lineage decision utilized transgenes driven by the distal Lck promoter and encoding reengineered Lck molecules (Hashimoto et al., 1996; Hernandez-Hoyos et al., 2000; Sohn et al., 2001). Because the distal Lck promoter is not active until the DP stage of differentiation, Lck activity was not perturbed in such transgenic mice until positive selection. In experiments utilizing the distal Lck promoter to drive expression of transgenic Lck molecules that had been reengineered to be constitutively active, DP thymocytes bearing MHC-I speciﬁc OT-1 transgenic TCR differentiated into CD4þ rather than CD8þ T cells (Hernandez-Hoyos et al., 2000). In some experiments utilizing the distal Lck promoter to drive expression of transgenic Lck molecules that had been reengineered to be enzymatically inactive, DP thymocytes bearing MHC-II speciﬁc AND transgenic TCR differentiated into CD8þ instead of CD4þ T cells (Hernandez-Hoyos et al., 2000). Thus, intracellular Lck kinase activity in DP thymocytes inﬂuenced the CD4/CD8 lineage decision, with greater Lck kinase activity associated with CD4þ T cell differentiation and weaker Lck kinase activity associated with CD8þ T cell differentiation. Experiments manipulating the in vivo activity of Tec family kinases also supported the strength of signal model, in that decreased in vivo Tec kinase activity diminished TCR signal transduction and favored DP thymocyte differentiation into CD8þ T cells (Liao and Littman, 1995; Lucas et al., 2002; Schaeffer et al., 2000). Also supportive of the strength of signal model were experiments manipulating the activity of the downstream extracellular signal-related kinases (Erk), with increased in vivo Erk activity favoring CD4þ T cell differentiation and in vitro inhibition of Erk activity favoring CD8þ T cell differentiation (Bommhardt et al., 1999; Sharp and Hedrick, 1999; Sharp et al., 1997; Wilkinson and Kaye, 2001).

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

103

Intrinsic to the strength of signal concept is the premise that experimental manipulations that weaken intrathymic signaling would divert thymocytes expressing MHC-II restricted TCR into the CD8þ T cell lineage. This prediction was assessed for thymocytes expressing the MHC-II restricted AND transgenic TCR which normally differentiate into CD4þ T cells. In CD4 knockout mice which do not express any CD4 coreceptor molecules, AND thymocytes were found to differentiate into CD8þ T cells, despite their expression of an MHC-II restricted TCR, supporting the concept that strong intrathymic signals mediated by CD4 coreceptors promoted CD4þ T cell differentiation and that weak intrathymic signals resulted in CD8þ T cell differentiation (Matechak et al., 1996). Experiments utilizing anti-TCR antibodies to stimulate positive selection in fetal thymic organ cultures (FTOC) provided further support for the strength of signal model. Addition of anti-TCR antibodies to FTOC induced differentiation of DP thymocytes into SP T cells even in the absence of MHC molecules, demonstrating that TCR engagements could signal positive selection without concomitant coreceptor engagements (Takahama et al., 1994). However, stimulation of positive selection by anti-TCR antibodies generally resulted in differentiation of DP thymocytes into CD4þ T cells, not CD8þ T cells. In fact, addition to FTOC of antibodies that crosslinked surface TCR complexes to other TCR complexes, crosslinked surface TCR complexes to cortical epithelium, or crosslinked surface TCR complexes to CD4/CD8 coreceptor molecules, all induced differentiation of DP thymocytes into CD4þ T cells, even when the antibodies crosslinked TCR and CD8 coreceptors together (Bommhardt et al., 1997, 1999; Muller and Kyewski, 1995; Suzuki et al., 1998). Because anti-TCR antibodies are considered mimics of high afﬁnity TCR interactions that presumably elicit ‘‘strong’’ intracellular signals, the ability of anti-TCR antibodies to drive DP thymocytes to selectively differentiate into CD4þ T cells was consistent with their generation of strong intracellular signals. In an attempt to develop antibodies that might stimulate weak intracellular TCR signals, anti-CD3 antibodies were reengineered to induce only limited bivalent TCR engagements. In contrast to native anti-TCR antibodies, such reengineered anti-CD3 antibodies induced DP thymocytes to selectively differentiate into CD8þ T cells (Bommhardt et al., 1997), supporting the concept that weak TCR signals resulted in CD8þ T cell differentiation. The development of two-stage in vitro suspension culture systems in which DP thymocytes could be stimulated in the absence of other thymic elements made it possible to transiently engage speciﬁc molecules on DP thymocytes in the ﬁrst ‘‘signaling’’ culture and to observe the effects of such engagements on DP thymocyte differentiation in the second ‘‘recovery’’ culture (Cibotti et al., 1997; Ohoka et al., 1996). In one such in vitro system, transient

104

ALFRED SINGER AND REMY BOSSELUT

antibody-induced engagement of surface TCR and CD2 molecules induced DP thymocytes to selectively terminate CD8 gene expression and to appear as CD4þ T cells (Cibotti et al., 1997, 2000). Interestingly, even though antibodyinduced TCR þ CD2 signaling in DP thymocytes selectively terminated CD8 gene expression, it also transiently destabilized both CD4 and CD8 coreceptor mRNAs (Cibotti et al., 2000). In a similar in vitro suspension culture system, stimulation of DP thymocytes with high doses of the pharmacologic agents PMA and ionomycin induced differentiation into CD4þ T cells, while stimulation of DP thymocytes with low doses of PMA and ionomycin induced differentiation into CD8þ T cells (Iwata et al., 1996). In addition, in vitro experiments were performed in a two-stage reaggregate FTOC system in which DP thymocytes expressing transgenic TCR were ﬁrst cocultured with peripheral antigen-pulsed dendritic cells prior to reaggregation and differentiation in the presence of thymic elements (Yasutomo et al., 2000a,b). These experiments revealed that long exposure of DP thymocytes to strong signals induced by antigen-pulsed dendritic cells biased subsequent differentiation toward CD4SP T cells, while short exposure biased subsequent differentiation toward CD8SP T cells. Thus, strong (i.e., long duration) signals induced CD4SP T cell differentiation, whereas weak (short duration) signals induced CD8SP T cell differentiation. Thus, a variety of different experimental approaches both in vivo and in vitro are consistent with the strength of signal concept that strong TCRþcoreceptor signals direct DP thymocytes to differentiate into CD4þ T cells, while weak TCRþcoreceptor signals direct DP thymocytes to differentiate into CD8þ T cells. B. Experiments that Contradict Strength of Signal The core concept of the strength of signal instructional model is that strong TCR signals direct developing thymocytes to differentiate into CD4SP T cells and that weak TCR signals direct developing thymocytes to differentiate into CD8SP T cells. Consequently, reductions in TCR signal strength should result in fewer CD4SP T cells and more CD8SP T cells. But this is not what was observed in two types of experiments that decreased TCR signal strength by either decreasing TCR ITAM number or by decreasing Lck kinase activity. One type of experiment examined T cell development in mice expressing TCR-z chains that had been reengineered to contain fewer than three ITAMs, their normal number (Ardouin et al., 1999; Shores et al., 1994). Each surface TCR complex normally contains 10 ITAMs, which provides signiﬁcant ampliﬁcation of TCR signal transduction. The two z-chains in each surface TCR complex together provide 6 ITAMs, and the other CD3 chains in each surface TCR complex together provide an additional 4 ITAMs. In contrast to normal surface TCR complexes, surface TCR complexes with reengineered TCR-z

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

105

chains contained as few as four ITAMs, providing signiﬁcantly less TCR signal ampliﬁcation than 10 ITAMs (Love et al., 2000; Shores et al., 1997). Thus, T cell development in these mice presented an excellent test of the core concept of the strength of signal model. What was observed in these mice was that decreased TCR signal strength resulted in decreased selection of both CD4SP and CD8SP T cells, but it did not divert CD4þ T cells into the CD8þ T cell lineage (Love et al., 2000; Shores et al., 1997). Rather, reductions in TCR ITAM number resulted in the generation of fewer and fewer SP T cells expressing the MHC-II restricted transgenic AND TCR, but they all remained CD4SP T cells and none were CD8SP T cells. Thus, decreased TCR signal strength resulted in fewer SP T cells, but it did not result in any changes in lineage afﬁliation (Dave et al., 1999). This is an especially compelling result since it directly contradicts the core concept of the strength of signal instructional model. A second type of experiment which tested the core concept of the strength of signal model utilized mice expressing a dominant-negative form of Lck, the expression of which was targeted to DP thymocytes by the distal Lck promoter. TCR signal transduction is initiated in both thymocytes and T cells by activation of the Src-family kinases Lck or Fyn (Iwashima et al., 1994; Marth et al., 1985; Samelson et al., 1990). Activation of Lck and Fyn are both competitively inhibited by expression of catalytically inactive (i.e., kinase-dead) Lck molecules. Consequently, expression of kinase-dead Lck in DP thymocytes diminishes the strength of TCR signal transduction. However, in conﬂict with the strength of signal model, increased expression of kinase-dead Lck in DP thymocytes from normal (i.e., non-TCR transgenic) mice resulted in the generation of fewer numbers of both CD4SP and CD8SP T cells, but it did not result in any change in lineage afﬁliation (Hashimoto et al., 1996; Levin et al., 1993). Expression of kinase-dead Lck in thymocytes bearing the MHC-II restricted DO10 TCR transgene resulted in the generation of fewer CD4SP T cells, but it did not result in the generation of CD8SP T cells expressing the DO10 TCR (Hashimoto et al., 1996). Notably, the failure of kinase-dead Lck to affect the lineage choice of MHC-II restricted DO10 transgenic TCR differs from the ability of kinase-dead Lck to alter the lineage choice of MHC-II restricted AND transgenic TCR (Hernandez-Hoyos et al., 2000). While the discrepant results of kinase-dead Lck on thymocytes expressing DO10 versus AND transgenic TCRs remain to be fully explained, the failure of DO10 transgenic thymocytes to alter lineage direction in response to decreased Lck kinase activity challenges the concept that TCR signal strength is the key determinant of lineage choice. C. Support for Stochastic/Selection In contrast to the strength of signal instructional model, the stochastic/selection model proposes that positive selection signals do not direct lineage choice. Instead, the stochastic/selection model postulates that TCRþcoreceptor

106

ALFRED SINGER AND REMY BOSSELUT

engagements induce DP thymocytes to randomly terminate one of the two coreceptor genes and to differentiate into short-lived intermediate cells that die unless rescued by a second TCR signaling event. In this model, the generation of the TCR rescue signal requires matching TCR and coreceptor molecules, so that intermediate cells with mismatched TCR and coreceptor molecules fail to survive. Thus, stochastic/selection postulates that differentiation of DP thymocytes into SP T cells consists of two TCR-mediated selection events that are separated in time and developmental sequence: the ﬁrst selection event is ‘‘positive selection’’ and results in TCR signaling of DP thymocytes to randomly terminate expression of one coreceptor gene so that they convert into intermediate thymocytes (Chan et al., 1993, 1994; Davis and Littman, 1994; Davis et al., 1993); the second selection event is referred to as ‘‘rescue’’ and results in the selective survival of only those intermediate thymocytes whose TCR and coreceptor molecules having matching MHC speciﬁcities. In the stochastic/selection model, lineage commitment occurs independently of TCR speciﬁcity, so that intermediate thymocytes are committed to one or the other T cell lineage by virtue of having randomly terminated expression of the opposite coreceptor gene; and fully 50% of lineage-committed thymocytes would fail to mature because their TCR and coreceptor speciﬁcities fail to match. To determine if lineage commitment occurs independently of TCR speciﬁcity, experiments have attempted to identify and characterize thymocytes that had recently been signaled to undergo lineage commitment. Initially, such experiments focused on thymocyte subsets whose expression of surface coreceptor molecules was skewed, in that they expressed one coreceptor molecule at high levels and the other coreceptor molecule at low levels (Guidos et al., 1990). The presumption was that CD4þ8lo thymocytes (referred to as CD4transitional cells) were derived from CD4-committed DP thymocytes that had recently terminated CD8 gene expression and were in the process of differentiating into CD4SP T cells, and that CD4lo8þ thymocytes (referred to as CD8transitional cells) were derived from CD8-committed DP thymocytes and were in the process of differentiating into CD8SP T cells (Chan et al., 1993) (Fig. 2, left). In this view, CD4-transitional cells are CD4-committed thymocytes, and CD8-transitional cells are CD8-committed thymocytes (Fig. 2, left). If lineage commitment occurs independently of TCR speciﬁcity, as postulated by stochastic/selection, each transitional cell population should contain two subsets of lineage-committed thymocytes: one subset whose TCR speciﬁcity matched that of the expressed coreceptor and another subset whose TCR speciﬁcity was mismatched with that of the expressed coreceptor. Experimental examination of thymocytes from MHC-I deﬁcient and MHC-II deﬁcient mice appeared to fulﬁll this prediction of the stochastic/selection model. Thymocytes from b2m knockout (MHC-I deﬁcient) mice contained CD4lo8þ, i.e., CD8-transitional, cells that were presumably CD8-committed

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

107

Fig 2 Surface phenotype of thymocytes based on CD4/CD8 staining is displayed as a 2-color ﬂow cytometry plot. Inside each circle is shown the coreceptor mRNA expression of individual thymocytes. The left side of the ﬁgure displays the expected relationship between surface coreceptor protein expression and coreceptor gene expression that led to the interpretation of the original transitional cell experiments. The right side of the ﬁgure displays the relationship between surface coreceptor protein expression and coreceptor gene expression that was subsequently experimentally observed. The arrows represent presumed precursor–progeny relationships.

despite expressing TCR that could only have been signaled by peptide/MHC-II intrathymic complexes (Chan et al., 1993; Crump et al., 1993; van Meerwijk and Germain, 1993), and thymocytes from MHC-II knockout (MHC-II deﬁcient) mice contained CD4þ8lo, i.e., CD4-transitional, cells that were presumably CD4-committed despite expressing TCR that could only have been signaled by peptide/MHC-I intrathymic complexes (Chan et al., 1993). Thus, transitional cell experiments were initially interpreted as fully supporting the stochastic/selection model. To assess the stochastic/selection perspective that lineage-committed thymocytes were initially short-lived intermediate cells whose survival required a TCR-mediated rescue signal, so-called ‘‘coreceptor rescue’’ experiments were undertaken. The concept behind coreceptor rescue was that short-lived intermediate thymocytes with mismatched TCR and coreceptor molecules would avoid cell death if they were forced to express the correct coreceptor molecule. For DP thymocytes expressing MHC-I restricted TCR, 50% would be expected to terminate endogenous CD8 coreceptor gene expression and would therefore be unable to generate the second TCR signal required for continued survival. But introduction of a transgene that forces CD8 expression in all T cells, including intermediate thymocytes, would be expected to rescue such intermediate thymocytes from cell death and promote their differentiation into MHC-I restricted CD4-lineage T cells. For DP thymocytes

108

ALFRED SINGER AND REMY BOSSELUT

expressing MHC-II restricted TCR, the reciprocal would be true, namely, that introduction of a transgene that forces CD4 expression in all T cells, including intermediate thymocytes, would be expected to rescue mismatched intermediate thymocytes from cell death and promote their differentiation into MHC-II restricted CD8-lineage T cells. Thus, stochastic/selection predicted that CD8 and CD4 transgenic mice would contain mature T cells whose TCR were discordant with their coreceptor phenotype. Moreover, stochastic/selection predicted that discordant T cells would represent approximately 50% of T cells in coreceptor transgenic mice. The actual outcome of coreceptor rescue experiments has been somewhat equivocal. As predicted by stochastic/selection, coreceptor rescue experiments did reveal discordant T cells with mismatched TCR and coreceptor molecules (Baron et al., 1994; Chan et al., 1994; Corbella et al., 1994; Davis et al., 1993; Itano et al., 1994). But discordant T cells were only detected at frequencies that were signiﬁcantly lower than the 50% predicted by stochastic/selection (Itano et al., 1994). For example, transgenic expression of CD8ab coreceptor molecules resulted in appearance of discordant CD4þ T cells expressing MHC-I restricted TCR (Chan et al., 1994; Itano et al., 1994), and transgenic expression of CD4 coreceptor molecules resulted in the appearance of discordant CD8SP T cells expressing MHC-II restricted TCR (Baron et al., 1994; Davis et al., 1993). However, the ‘‘rescue’’ of such mismatched T cells required high expression of transgenic coreceptor molecules and resulted in the appearance of relatively low numbers of mismatched T cells. Recently, a coreceptor rescue experiment was accomplished without the use of coreceptor transgenes. Knockout mice were constructed in which the CD4 silencer element was deleted from the endogenous CD4 gene locus, resulting in persistent expression of endogenously encoded CD4 molecules on all T-lineage cells, including CD8þ T cells (Leung et al., 2001). As predicted by stochastic/selection, CD8þ T cells were generated in CD4silencer deﬁcient mice in the absence of MHC-I expression, presumably because persistent CD4 expression rescued intermediate cells with MHC-II restricted TCR that had randomly committed to the CD8þ T cell lineage. While mismatched SP T cells were generated in CD4-silencer knockout mice, the frequency of such mismatched T cells was still signiﬁcantly less than the 50% predicted by stochastic/selection. Thus, there exists experimental support for the stochastic/selection perspective that lineage commitment during positive selection occurs independently of TCR speciﬁcity and results in the generation of short-lived intermediate thymocytes. D. Experiments that Contradict Stochastic/Selection A core requirement of the stochastic/selection model is that the efﬁciency of positive selection must be no higher than 50%, since 50% of potentially selectable DP thymocytes should randomly make the incorrect lineage choice

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

109

and fail to differentiate into mature SP T cells. However, experiments involving BRDU-labeled thymocytes expressing a transgenic TCR have determined that the efﬁciency of positive selection of DP thymocytes into SP T cells could be as high as 90%, a result that contradicts requirements of the stochastic/selection model (Itano and Robey, 2000). The stochastic/selection model also predicts that the few DP thymocytes that can be signaled without concurrent coreceptor engagements would differentiate randomly into CD4SP and CD8SP T cells. Consequently, in coreceptor-deﬁcient mice, T cells expressing either MHC-I or MHC-II restricted TCR should be present in equal frequencies in both CD4 and CD8 lineage cells. However, this is not what has been observed. In CD4 knockout mice, MHC-II restricted T cells reactive against leishmania parasitic antigens were found only among CD4 lineage cells (i.e., CD8 T-helper cells) and not among CD8þ cells (Locksley et al., 1993). Similarly, in CD8 knockout fetal thymic organ cultures, DP thymocytes bearing MHC-I speciﬁc transgenic TCRs only differentiated into CD8 lineage cells (deﬁned as CD4 cells that expressed either CD8b or perforin) and did not differentiate into CD4þ T cells (Goldrath et al., 1997; Sebzda et al., 1997). Thus, lineage choice was not random for thymocytes positively selected by coreceptor-independent TCR signals, in conﬂict with the stochastic/selection model. The support provided to stochastic/selection by the transitional cell experiments referred to earlier hinged on the presumptions that CD4-transitional cells had terminated CD8 gene expression and so were CD4-committed thymocytes, and that CD8-transitional cells had terminated CD4 gene expression and so were CD8-committed thymocytes (Lucas et al., 1995) (Fig. 2, left). That CD8-transitional cells had, in fact, terminated CD4 gene expression was subsequently conﬁrmed (Suzuki et al., 1995). However, the presumption regarding CD4-transitional cells was not correct as demonstrated by two parallel experimental approaches. One approach puriﬁed transitional cell populations and experimentally assessed the coreceptor genes that were expressed by cells within each transitional population. Surprisingly, CD4transitional cells were found to consist of two different subpopulations of thymocytes: one subpopulation which only expressed the CD4 coreceptor gene and a second subpopulation which only expressed the CD8 coreceptor gene (Suzuki et al., 1995). If lineage commitment is equivalent to coreceptor gene expression, then the CD4-transitional thymocyte subpopulation contained both CD4-committed and CD8-committed thymocytes (Fig. 2, right). Why CD8-committed thymocytes would appear as CD4þ8lo (CD4-transitional) cells with reduced surface levels of CD8 was unclear until recently (discussed later in Section V.B), but was conﬁrmed by intrathymic transfer experiments in which puriﬁed CD4-transitional thymocytes gave rise to mature CD8þ T cells (Barthlott et al., 1997b; Lundberg et al., 1995). Thus,

110

ALFRED SINGER AND REMY BOSSELUT

the mere presence of transitional thymocyte populations in MHC-I and MHCII deﬁcient mice could not be taken as support for the stochastic/selection perspective that lineage commitment occurred independently of TCR speciﬁcity. V. Kinetic Signaling as an Alternative to Classical Models of Lineage Commitment

It is fair to say that greater experimental support exists for instruction than for stochastic/selection, but that does not mean that either model is correct. In fact, convincing experiments have been performed that contradict critical requirements of each of these two models. As a result, it is possible that neither instruction nor stochastic/selection may correctly describe how lineage commitment actually occurs in developing thymocytes. While instruction and stochastic/selection represent two alternative perspectives on how lineage commitment might occur during positive selection, the two models disagree only on whether lineage commitment is directed or stochastic. Otherwise, the two models are based on identical presumptions. Speciﬁcally, both models presume that lineage commitment occurs in TCR-signaled DP thymocytes simultaneously with positive selection, and both models presume that lineage commitment results in irreversible silencing of the opposite coreceptor gene. In contrast, a different perspective of lineage commitment, referred to as ‘‘kinetic signaling,’’ is based on a very different set of presumptions (Fig. 3) (Singer, 2002). In the kinetic signaling model, TCR-mediated positive selecting signals do not induce DP thymocytes to undergo lineage commitment. Rather, TCRmediated positive selecting signals induce DP thymocytes to terminate CD8 gene expression and to transcriptionally convert into CD4þ8 intermediate thymocytes, which nevertheless remain lineage-uncommitted cells that retain the potential to differentiate into either CD4SP or CD8SP T cells (Brugnera et al., 2000; Cibotti et al., 2000). However, it is at the CD4þ8 intermediate stage of differentiation that lineage commitment occurs. The lineage decision is based simply on whether TCR-mediated positive selecting signals are still present or have ceased. If positively selecting TCR signals are still present, CD4þ8 intermediate thymocytes differentiate into CD4SP T cells. But, if positively selecting TCR signals are absent for whatever reason, CD4þ8 intermediate thymocytes differentiate into CD8SP T cells—and they do so by reinitiating CD8 transcription and terminating CD4 transcription, molecular events that are collectively referred to as ‘‘coreceptor reversal.’’ Thus, the kinetic signaling model postulates that it is the persistence or cessation of TCR signals in CD4þ8 intermediate thymocytes that determines CD4/CD8 lineage direction. The end result is that sustained (long duration) TCR signals induce DP thymocytes to differentiate into CD4SP T cells, whereas disrupted

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

111

Fig 3 A schematic of the kinetic signaling model of lineage commitment in which cell fate is determined by persistence/cessation of positive selection signaling. This model postulates that DP thymocytes respond to TCR þ coreceptor signals (regardless of their MHC speciﬁcity) by terminating CD8 transcription and transcriptionally converting into CD4þ8 intermediate thymocytes. CD4þ8 intermediate thymocytes are lineage uncommitted cells that retain the potential to differentiate into mature CD4SP or CD8SP T cells. Sustained signaling in CD4þ8 intermediate thymocytes results in differentiation into mature CD4SP T cells, whereas cessation of signaling in CD4þ8 intermediate thymocytes results in ‘‘coreceptor reversal’’ and differentiation into CD4SP T cells. ‘‘Coreceptor reversal’’ refers to the reinitiation of CD8 transcription and termination of CD4 transcription in intermediate thymocytes, transcriptionally converting them from CD4þ8 cells into CD48þ cells. Coreceptor reversal is promoted by IL7 that is constitutively produced in the thymus.

(short duration) TCR signals induce DP thymocytes to differentiate into CD8SP T cells (Fig. 3). Because its underlying presumptions differ substantially from those of classical lineage commitment models, the kinetic signaling model may initially appear to be excessively complex. In actuality, kinetic signaling postulates a very simple binary mechanism by which signaled DP thymocytes can determine their appropriate cell fate (Fig. 3). From an anthropomorphic perspective, the kinetic signaling model can be viewed in the following way: TCR-signaled DP thymocytes are programmed to initially pursue their ‘‘primary’’ cell fate, which is to differentiate into CD4 lineage T cells. In this pursuit, they terminate CD8 gene expression and transcriptionally convert into CD4þ8 cells. If the termination of CD8 gene expression does not interrupt TCR signaling, the cells ‘‘realize’’ that their TCR does not require CD8 gene expression and they complete their differentiation into CD4SP T cells. However, if termination of CD8 gene expression results in a disruption in TCR signaling, the cells realize that their TCR do require CD8 gene expression and so they reverse

112

ALFRED SINGER AND REMY BOSSELUT

developmental direction and differentiate into CD8SP T cells. In this way, lineage commitment becomes a simple binary decision for CD4þ8 intermediate thymocytes in which the presence or absence of TCR signals results in CD4þ or CD8þ T cell differentiation. To summarize, instruction and stochastic/selection models both presume that: (1) lineage commitment occurs in thymocytes at the DP stage of differentiation; (2) lineage commitment occurs simultaneously with positive selection and results in cessation of the opposite coreceptor gene; and (3) cessation of coreceptor gene expression is irreversible and reﬂects commitment to the opposite T cell lineage. In contrast, the kinetic signaling model presumes that: (1) TCR-signaled DP thymocytes initially cease CD8 gene expression, regardless of the speciﬁcity of their TCR; (2) lineage commitment occurs subsequent to the cessation of CD8 coreceptor gene expression in thymocytes that are transcriptionally CD4þ8; and (3) cessation of CD8 gene expression does not necessarily reﬂect lineage commitment. A. Analysis of the Kinetic Signaling Model In the kinetic signaling model, cessation of TCR-mediated positive selection signals in CD4þ8 intermediate thymocytes results in their differentiation into CD8SP T cells. But why should MHC-I restricted TCR signaling cease in CD4þ8 intermediate thymocytes? Two situations based on different TCR afﬁnities can be considered. First, in the top panel of Fig. 4, CD8-dependent TCR/ligand interactions that are of moderate afﬁnity signal DP thymocytes to terminate CD8 transcription and to transcriptionally convert into CD4þ8 intermediate cells; TCR signaling is disrupted when declining CD8 surface levels ﬁnally fall to a level that is too low to maintain TCR/ligand engagements. Cessation of TCR signaling in intermediate thymocytes then results in coreceptor reversal and differentiation into CD8SP T cells. So, for thymocytes with moderate afﬁnity MHC-I speciﬁc TCR, it is the declining expression of CD8 coreceptor molecules on the surface of intermediate thymocytes that causes MHC-I dependent signaling to cease. Second, in the bottom panel of Fig. 4, CD8-dependent TCR/ligand interactions that are of very low afﬁnity also signal DP thymocytes to terminate CD8 transcription and to transcriptionally convert into CD4þ8 intermediate cells; but, because their ligand-binding afﬁnity is so low, TCR engagement cannot be maintained for long and signaling terminates even before surface CD8 protein levels have detectably declined. Cessation of TCR signaling then triggers coreceptor reversal and terminal differentiation into CD8SP T cells. Thus, for thymocytes with very low afﬁnity MHC-I speciﬁc TCR, such as the HY TCR, it is the inability of very low afﬁnity TCR to maintain ligand engagement that causes signal disruption. In this case, signaled DP thymocytes such as those expressing the HY transgenic

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

113

Fig 4 Two examples of signal cessation resulting in CD8SP T cell differentiation. In the top panel, TCR with moderate afﬁnity for MHC-I/peptide complexes transduce positive selection signals in DP thymocytes that terminate CD8 transcription and transcriptionally convert into CD4þ8 intermediate cells. Because CD8 transcription has terminated, CD8 surface protein levels steadily decline. When CD8 surface protein levels fall to the point that they can no longer promote continued TCR/ligand engagement, TCR signaling ceases and results in differentiation into CD8SP T cells. In the bottom panel, TCR with very low afﬁnity for MHC-I/peptide complexes transduce positive selection signals in DP thymocytes that terminate CD8 transcription and transcriptionally convert into CD4þ8 intermediate cells. In this case, TCR afﬁnity for ligand is too low to avoid early disengagement, resulting in signal disruption in CD4þ8 intermediate thymocytes even before CD8 surface protein levels have detectably declined. Because TCR signaling has ceased, CD4þ8 intermediate thymocytes terminally differentiate into CD8SP T cells. It is speculated that a minority of CD8SP T cells, including those expressing the HY transgenic TCR, arise this way.

TCR may differentiate into CD8SP T cells without going through a stage in which they appear as CD4þ8lo cells. In the kinetic signaling model, persistence of positive selection signaling in CD4þ8 intermediate thymocytes drives them to differentiate into CD4SP T cells. But why should MHC-II restricted TCR signaling persist in CD4þ8 intermediate thymocytes? There are two situations based on different TCR afﬁnities that can be considered in this regard as well. First, in the top panel of Fig. 5, CD4-dependent TCR/ligand interactions signal DP thymocytes to

114

ALFRED SINGER AND REMY BOSSELUT

Fig 5 Two examples of signal persistence in CD4þ8 intermediate thymocytes resulting in CD4SP T cell differentiation. In the top panel, TCR with moderate afﬁnity for MHC-II/peptide complexes transduce positive selection signals in DP thymocytes that terminate CD8 transcription and transcriptionally convert into CD4þ8 intermediate cells. Because MHC-II-speciﬁc TCR engagements are not dependent on CD8 coreceptors, signaling persists in CD4þ8 intermediate cells, resulting in differentiation into CD4SP T cells. In the bottom panel, TCR with sufﬁcient afﬁnity for intrathymic MHC/peptide complexes to be independent of both coreceptor molecules transduce positive selection signals in DP thymocytes that terminate CD8 transcription and transcriptionally convert into CD4þ8 intermediate cells. Because coreceptor-independent TCR engagements are not dependent on CD8 coreceptors, signaling persists in CD4þ8 intermediate cells, resulting in differentiation into CD4SP T cells. It is speculated that a minority of CD4SP T cells, including CD4þNK1.1þ T cells, may arise this way.

terminate CD8 transcription and to transcriptionally convert into CD4þ8 intermediate cells; TCR signaling continues because it is unaffected by the loss of surface CD8 proteins on CD4þ8 intermediate cells. Persistent or sustained TCR signaling then results in differentiation of intermediate thymocytes into CD4SP T cells. Second, in the bottom panel of Fig. 5, coreceptorindependent TCR/ligand interactions that are of sufﬁciently high afﬁnity to be independent of both coreceptors signal DP thymocytes to terminate CD8 transcription and to transcriptionally convert into CD4þ8 cells; signaling continues because it is unaffected by the loss of surface CD8 proteins on CD4þ8 intermediate cells. In this situation, which may be characteristic of TCR selected by intrathymic ligands such as CD1, sustained TCR signaling then results in differentiation into CD4SP T cells (Bendelac et al., 1994; Cardell et al., 1995). Thus, thymocytes expressing coreceptor-independent TCR can differentiate into CD4SP T cells even when their selecting ligand is

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

115

an MHC-I/peptide complex, as actually is the case for CD4þNK1.1þ T cells (Bendelac et al., 1997). In addition, it is possible that different expression densities of MHC-I and MHC-II complexes on cortical thymic epthelium may also contribute to TCR signal cessation and persistence. Thymic cortical epithelium expresses a lower density of MHC-I than MHC-II complexes (Jenkinson et al., 1981; Van Ewijk et al., 1980). Consequently, as signaled DP thymocytes migrate through the thymic cortex, it may be more difﬁcult to maintain MHC-I speciﬁc TCR/ligand interactions than to maintain MHC-II speciﬁc TCR/ligand interactions. Because MHC-I speciﬁc interactions may be more readily disrupted in the thymic cortex than MHC-II speciﬁc interactions, the speciﬁcity of the TCR for ligands presented by MHC-I or MHC-II may contribute to whether TCR signals persist or cease and so may itself induce a lineage bias. Thus, kinetic signaling provides a plausible explanation for the experimental observation that thymocytes undergoing positive selection in coreceptor-deﬁcient mice nevertheless differentiated into their appropriate T cell lineage despite the absence of coreceptor engagements (Goldrath et al., 1997; Pena-Rossi et al., 1999; Rahemtulla et al., 1991; Sebzda et al., 1997). In contrast, this experimental observation is incompatible with stochastic/selection but can be explained by the strength of signal model so long as it is presumed that coreceptor-independent TCR interactions with peptide/MHC-I ligands are intrinsically weaker than coreceptor-independent TCR interactions with peptide/MHC-II ligands, a presumption for which there exists no supporting evidence. Notably, the kinetic signaling model is neither instructional nor stochastic, as all signaled DP thymocytes are thought to respond to TCRþcoreceptor engagements identically, regardless of the MHC speciﬁcity of their TCR, by transiently terminating CD8 transcription (Brugnera et al., 2000; Cibotti et al., 2000). As a result, cessation of CD8 coreceptor transcription neither implies nor reveals lineage commitment, and CD4þ8 intermediate thymocytes that are the immediate progeny of signaled DP thymocytes are lineageuncommitted cells that retain the potential to differentiate into either CD4SP or CD8SP T cells. B. Coreceptor Reversal Kinetic signaling evolved from an earlier model (the asymmetric commitment model) formulated before it was recognized that intermediate thymocytes could undergo coreceptor reversal (Benveniste et al., 1996; Suzuki et al., 1995). Coreceptor reversal is a unique and central component of the kinetic signaling model in which intermediate thymocytes that are transcriptionally CD4þ8 respond to cessation of TCR signaling by reinitiating CD8 gene expression and silencing CD4 gene transcription. Coreceptor reversal was

116

ALFRED SINGER AND REMY BOSSELUT

originally discovered by examining TCR-signaled DP thymocytes in a twostage in vitro culture system (Brugnera et al., 2000). In this culture system, transiently signaled DP thymocytes terminate CD8 gene expression and differentiate into intermediate thymocytes that are transcriptionally CD4þ8. Such CD4þ8 intermediate thymocytes were originally thought to be CD4-committed cells that would complete their differentiation into fully mature CD4SP T cells if placed in an inductive environment (Cibotti et al., 1997). However, adoptive transfer of in vitro-generated CD4þ8 thymocytes back into an intact thymus resulted in their becoming CD8SP T cells, not CD4SP T cells (Brugnera et al., 2000). Thus, despite having ceased CD8 gene expression, these CD4þ8 intermediate thymocytes were not CD4 lineage-committed cells as they could undergo coreceptor reversal in vivo and differentiate into CD8SP T cells. Importantly, coreceptor reversal was not limited to thymocytes that had been signaled in vitro, but was a function also possessed by in vivo-signaled CD4þ8 intermediate thymocytes, as they too retained the ability to ‘‘reverse direction’’ and differentiate into CD8SP T cells. While the ability of CD4þ8 intermediate thymocytes to undergo coreceptor reversal is the keystone of the kinetic signaling model, it directly contradicts the classical concept that is fundamental to both stochastic/selection and instructional models, namely, that selective cessation of coreceptor gene expression is irreversible and reﬂects commitment to the opposite T cell lineage. The early transitional cell experiments that focused on thymocytes with skewed surface coreceptor expression observed that CD4þ8lo (i.e., CD4transitional) thymocytes consisted of two subsets: one that was transcriptionally CD4þ8 and one that was transcriptionally CD48þ. It was not surprising that signaled DP thymocytes that had transcriptionally converted into CD4þ8 cells appeared by surface staining as CD4þ8lo thymocytes (Barthlott et al., 1997a; Benveniste et al., 1996; Suzuki et al., 1995). However, at the time these observations were ﬁrst made, it was quite surprising that signaled DP thymocytes that had transcriptionally converted into CD48þ cells appeared by surface staining as CD4þ8lo thymocytes, as their CD4þ8lo appearance was completely discordant with both their coreceptor transcriptional state and their ultimate fate as CD8SP T cells. In 2003, this paradox was resolved by transgenic experiments in which CD8 protein expression was uncoupled from CD8 coreceptor gene expression (Bosselut et al., 2003). These experiments observed that MHC-I signaled DP thymocytes appear as CD4þ8lo cells by surface staining because they initially terminate endogenous CD8 gene expression and transcriptionally convert into CD4þ8 intermediate thymocytes; MHC-I signaled CD4þ8 intermediate thymocytes subsequently undergo coreceptor reversal to transcriptionally convert into CD48þ cells that retain for a time their previous appearance as CD4þ8lo thymocytes even as they are

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

117

Fig 6 The kinetic signaling model proposes that TCR-signaled DP thymoctyes initially terminate CD8 gene expression and appear as CD4þ8lo transitional cells as a result of diminished or absent CD8 gene expression. In the kinetic signaling model, TCR disengagement of MHC-I ligands occurs in CD4þ8lo transitional cells because of diminished surface CD8 coreceptor levels and results in cessation of TCR signaling, reversal of coreceptor gene expression, and terminal differentiation into CD8þ T cells.

differentiating into CD8SP T cells (Fig. 6). Thus, thymocytes within the CD4þ8lo transitional cell population that are transcriptionally CD48þ are the immediate progeny of CD4þ8 thymocytes that have recently undergone coreceptor reversal and whose surface appearance has not yet caught up to their coreceptor transcriptional state (Fig. 6). C. Role of IL-7 in Coreceptor Reversal and CD8þ T Cell Differentiation Most DP thymocytes are short-lived cells that do not express the antiapoptotic protein Bcl-2 and do not express functional cytokine receptors (Linette et al., 1994; Punt et al., 1996). However, TCR signals induce DP thymocytes to upregulate Bcl-2 and IL-7 receptor (IL-7R) expression, both of which remain upregulated for as long as TCR signals persist (Brugnera et al., 2000; Cibotti et al., 1997; Linette et al., 1994; Punt et al., 1996; Sudo et al., 1993). However, cessation of MHC-I speciﬁc TCR signaling in intermediate thymocytes means that such cells require a replacement signal to promote

118

ALFRED SINGER AND REMY BOSSELUT

their survival during coreceptor reversal and differentiation into mature CD8SP T cells. This replacement signal is provided by IL-7, as IL-7R signals maintain Bcl-2 expression and promote survival of developing thymocytes during coreceptor reversal and during differentiation into CD8SP T cells (Brugnera et al., 2000; Yu et al., 2003). In addition to promoting thymocyte survival, IL-7R signals also appear to participate in coreceptor reversal and CD8SP T cell differentiation by enhancing CD4 gene silencing and by promoting the acquistion of functional competence so that CD8SP thymocytes proliferate and secrete cytokines in response to TCR signals (Yu et al., 2003). Experimentally, transiently signaled DP thymocytes differentiate in vitro into CD4þ8 intermediate thymocytes that, upon addition of IL-7, undergo coreceptor reversal and differentiate into fully functional CD8SP T cells that can proliferate and secrete the cytokine IFNg. Reciprocally, blockade of IL-7R signaling upon addition of anti-IL-7R antibodies to late-stage fetal thymic organ cultures selectively abrogates the generation of CD8SP T cells without signiﬁcantly affecting the generation of CD4SP T cells (Brugnera et al., 2000; Yu et al., 2003). Since IL-7R signals promote and enhance coreceptor reversal, how do intermediate thymocytes manage to avoid coreceptor reversal so that they can differentiate into CD4SP T cells? The likely answer is that persistent TCR signaling desensitizes IL-7Rs so that the IL-7Rs cannot signal (Noguchi et al., 1997). Experimentally, it has been observed that TCR signaling in CD4þ8 intermediate thymocytes does block coreceptor reversal, in that signaled cells remain transcriptionally CD4þ8 despite the presence of IL-7 (Brugnera et al., 2000; Yu et al., 2003). This point is strongly supported by experiments examining surface expression on in vivo thymocytes of the glucose transporter molecule glut-1, which is upregulated on thymocytes by IL-7R signals. Even though CD4SP and CD8SP thymocytes both express IL-7Rs, glut-1 expression on in vivo thymocytes was only upregulated on CD8SP thymocytes but not on CD4SP thymocytes, indicating that IL-7R signaling did not occur in developing CD4SP thymocytes, presumably as a result of persistent TCR signaling (Yu et al., 2003). In the kinetic signaling model, since differentiation of CD4þ8 intermediate thymocytes into CD4SP T cells requires persistent TCR signaling, and differentiation into CD8SP T cells requires IL-7R signaling, it is possible to view lineage choice in CD4þ8 intermediate thymocytes as a competition between TCR and IL-7R signaling. That is, for CD4þ8 intermediate thymocytes to continue differentiating into CD4þ T cells, TCR signals must persist and must result in desensitization of the IL-7R; for CD4þ8 intermediate thymocytes to undergo coreceptor reversal and differentiate into CD8þ T cells, TCR signals must either cease or be insufﬁcient to interfere with IL-7R signaling.

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

119

D. Distinction between Signal Intensity and Signal Duration Because strong signals tend to be both more intense and of longer duration than weak signals, it is difﬁcult in many experimental systems to distinguish between the potential effects of signal intensity and signal duration. Consequently, most experiments that have been taken as support for the strength of signal model are equally supportive of signal duration, as per the kinetic signaling model (Singer, 2002). A few examples should sufﬁce to make this point. As described earlier, thymocytes from Csk conditional knockout mice were observed to differentiate exclusively into CD4þ T cells as a result of unregulated Lck and Fyn kinase activity (Schmedt and Tarakhovsky, 2001; Schmedt et al., 1998). But in the absence of Csk, Lck and Fyn kinase activity were increased in both intensity and duration, so that it is not possible to distinguish which parameter was the critical one for the appearance of CD4þ T cells. Essentially the same sort of analysis applies to all other experiments in which in vivo kinase activity was manipulated. So it is not possible to distinguish between the impact of signal intensity versus signal duration in experiments in which Lck, Tec, or Erk kinase activities either were increased and resulted in increased CD4þ T cell differentiation, or were decreased and resulted in decreased CD8þ T cell differentiation. While nearly all experiments that are supportive of strength of signal are equally supportive of kinetic signaling, that is not true for the original experiments that stimulated formulation of the strength of signal model (Itano et al., 1996; Seong et al., 1992). As discussed earlier, these original experiments utilized transgene encoded chimeric CD8/CD4 coreceptor molecules in which the intracellular tail of CD8a was replaced with the intracellular tail of CD4. In the original experiments, mice expressed both transgenically encoded chimeric CD8/CD4 coreceptor molecules and endogenously encoded CD8a molecules, and were found to contain MHC-I speciﬁc T cells that had adopted the incorrect CD4 T cell fate (Itano et al., 1996; Seong et al., 1992). However, in the original experiments, the inﬂuence of chimeric CD8/ CD4 coreceptor proteins on promoting CD4 versus CD8 lineage choice could not be compared, as all transgenic thymocytes also expressed endogenous CD8 coreceptor proteins. Nevertheless, mice expressing CD8/CD4 coreceptor molecules contained many more ‘‘mismatched’’ MHC-I restricted CD4SP T cells than mice expressing only wildtype CD8 coreceptor molecules. This latter comparison suggested that strong intrathymic Lck signals from the CD4 coreceptor tail had induced many MHC-I speciﬁc thymocytes to differentiate into CD4SP T cells that otherwise would have differentiated into CD8SP T cells. What other interpretation of these results might there be? Experiments with chimeric CD8/CD4 coreceptor transgenes were recently redone so that they were essentially identical to the original experiments with

120

ALFRED SINGER AND REMY BOSSELUT

the important exception that chimeric CD8/CD4 coreceptor transgenes were expressed in CD8a knockout mice that did not express endogenously encoded CD8a coreceptor molecules (Bosselut et al., 2001). Since chimeric CD8/CD4 coreceptor molecules were now the only CD8 molecules that thymocytes in these mice expressed, it was possible to obtain a much more complete picture of the effect of the CD4 tail on selection and differentiation of MHC-I restricted thymocytes. In fact, these later experiments yielded a very different picture than was obtained originally. It was found that expression of chimeric CD8/CD4 molecules promoted the differentiation of MHC-I restricted thymocytes into both CD4 and CD8 T cell lineages, with the majority of MHC-I speciﬁc thymocytes still differentiating into CD8SP T cells (Bosselut et al., 2001). In other words, the CD4 coreceptor tail did not preferentially direct MHC-I restricted thymocytes into the CD4 T cell lineage, as most MHC-I restricted thymocytes still differentiated into CD8SP T cells. However, the CD4 coreceptor tail did signiﬁcantly increase the number of MHC-I restricted DP thymocytes that were signaled to undergo positive selection, as evidenced by the greater overall numbers of MHC-I restricted SP T cells (both CD4þ and CD8þ) that were generated. In these experiments, the appearance of MHC-I restricted CD4SP T cells was attributable to sustained MHC-I restricted TCR signaling in intermediate thymocytes as a consequence of persistent expression of transgenically encoded CD8ab molecules, since transcription of the transgene was regulated by heterologous hCD2 transcriptional elements that were unaffected by TCR signaling. The original observation (Itano et al., 1996; Seong et al., 1992) that more MHC-I restricted CD4SP T cells appeared in mice expressing chimeric CD8/CD4 transgenic coreceptors compared to wildtype CD8 transgenic coreceptors was conﬁrmed, but simply paralleled the greater overall numbers of both CD4SP and CD8SP T cells (Bosselut et al., 2001). In other words, in comparison to wildtype CD8 transgenic coreceptor molecules, expression of chimeric CD8/CD4 transgenic coreceptor molecules resulted in greater numbers of both MHC-I restricted CD4SP T cells and MHCI restricted CD8SP T cells. Thus, these latest experiments revealed that differences in Lck signaling by the CD4 and CD8 coreceptor tails do not inﬂuence lineage choice, but do inﬂuence the number of DP thymocytes that can be signaled to undergo positive selection (Bosselut et al., 2001). Importantly, the observation that the CD4 tail promotes positive selection of greater numbers of DP thymocytes than the CD8 tail provides an explanation for why CD4þ T cells outnumber CD8þ T cells in all mammalian species: namely, that CD4-dependent TCR signals stimulate greater numbers of DP thymocytes to undergo positive selection than CD8-dependent TCR signals (Bosselut et al., 2001). Thus, in the kinetic signaling model, TCR signal intensity and TCR signal duration have two very distinct effects: TCR signal intensity determines

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

121

the number of DP thymocytes that are signaled to undergo positive selection, while TCR signal duration determines their lineage direction. VI. Conclusions

How well we understand the CD4/CD8 lineage decision reﬂects how well we understand thymocyte differentiation in general. Certainly, the ﬁnal chapter on the CD4/CD8 lineage decision has not yet been written, but its outlines may be becoming clearer. The initial dichotomy between instructional and stochastic/selection models of lineage commitment fueled a large number of experiments and provided a common intellectual structure for understanding results obtained in widely disparate experimental systems. The more recent formulation of the kinetic signaling model, and the appreciation that signaled thymocytes can reverse course by undergoing coreceptor reversal, adds an entirely new dimension to the lineage commitment debate. Hopefully, it will not be too long before we have both a cellular and a molecular understanding of how bipotential DP thymocytes integrate environmental cues to determine their appropriate cell fate. References Aifantis, I., Borowski, C., Gounari, F., Lacorazza, H. D., Nikolich-Zugich, J., and von Boehmer, H. (2002). A critical role for the cytoplasmic tail of pTalpha in T lymphocyte development. Nat. Immunol. 3, 483–488. Aifantis, I., Gounari, F., Scorrano, L., Borowski, C., and von Boehmer, H. (2001). Constitutive pre-TCR signaling promotes differentiation through Ca2þ mobilization and activation of NF-kappaB and NFAT. Nat. Immunol. 2, 403–409. Allman, D., Karnell, F. G., Punt, J. A., Bakkour, S., Xu, L., Myung, P., Koretzky, G. A., Pui, J. C., Aster, J. C., and Pear, W. S. (2001). Separation of Notch1 promoted lineage commitment and expansion/transformation in developing T cells. J. Exp. Med. 194, 99–106. Ardouin, L., Boyer, C., Gillet, A., Trucy, J., Bernard, A. M., Nunes, J., Delon, J., Trautmann, A., He, H. T., Malissen, B., and Malissen, M. (1999). Crippling of CD3-zeta ITAMs does not impair T cell receptor signaling. Immunity 10, 409–420. Ashton-Rickardt, P. G., Bandeira, A., Delaney, J. R., Van, K. L., Pircher, H. P., Zinkernagel, R. M., and Tonegawa, S. (1994). Evidence for a differential avidity model of T cell selection in the thymus. Cell 76, 651–663. Baron, A., Hafen, K., and von Boehmer, H. (1994). A human CD4 transgene rescues CD4–CD8þ cells in beta 2-microglobulin-deﬁcient mice. Eur. J. Immunol. 24, 1933–1936. Barthlott, T., Kohler, H., and Eichmann, K. (1997a). Asynchronous coreceptor downregulation after positive thymic selection: Prolonged maintenance of the double positive state in CD8 lineage differentiation due to sustained biosynthesis of the CD4 coreceptor. J. Exp. Med. 185, 357–362. Barthlott, T., Kohler, H., Pircher, H., and Eichmann, K. (1997b). Differentiation of CD4(high)CD8(low) coreceptor-skewed thymocytes into mature CD8 single-positive cells independent of MHC class I recognition. Eur. J. Immunol. 27, 2024–2032. Bendelac, A., Killeen, N., Littman, D. R., and Schwartz, R. H. (1994). A subset of CD4þ thymocytes selected by MHC class I molecules. Science 263, 1774–1778.

122

ALFRED SINGER AND REMY BOSSELUT

Bendelac, A., Rivera, M. N., Park, S. H., and Roark, J. H. (1997). Mouse CD1-speciﬁc NK1 T cells: Development, speciﬁcity, and function. Annu. Rev. Immunol. 15, 535–562. Benveniste, P., Knowles, G., and Cohen, A. (1996). CD8/CD4 lineage commitment occurs by an instructional/default process followed by positive selection. Eur. J. Immunol. 26, 461–471. Bhandoola, A., Cibotti, R., Punt, J. A., Granger, L., Adams, A. J., Sharrow, S. O., and Singer, A. (1999). Positive selection as a developmental progression initiated by alpha beta TCR signals that ﬁx TCR speciﬁcity prior to lineage commitment. Immunity 10, 301–311. Blanc, D., Bron, C., Gabert, J., Letourneur, F., MacDonald, H. R., and Malissen, B. (1988). Gene transfer of the Ly-3 chain gene of the mouse CD8 molecular complex: Co-transfer with the Ly-2 polypeptide gene results in detectable cell surface expression of the Ly-3 antigenic determinants. Eur. J. Immunol. 18, 613–619. Bommhardt, U., Basson, M. A., Krummrei, U., and Zamoyska, R. (1999). Activation of the extracellular signal-related kinase/mitogen-activated protein kinase pathway discriminates CD4 versus CD8 lineage commitment in the thymus. J. Immunol. 163, 715–722. Bommhardt, U., Cole, M. S., Tso, J. Y., and Zamoyska, R. (1997). Signals through CD8 or CD4 can induce commitment to the CD4 lineage in the thymus. Eur. J. Immunol. 27, 1152–1163. Bonifacino, J. S., McCarthy, S. A., Maguire, J. E., Nakayama, T., Singer, D. S., Klausner, R. D., and Singer, A. (1990). Novel post-translational regulation of TCR expression in CD4þCD8þ thymocytes inﬂuenced by CD4. Nature 344, 247–251. Bonneville, M., Ishida, I., Mombaerts, P., Katsuki, M., Verbeek, S., Berns, A., and Tonegawa, S. (1989). Blockage of alpha beta T-cell development by TCR gamma delta transgenes. Nature 342, 931–934. Bosselut, R., Feigenbaum, L., Sharrow, S. O., and Singer, A. (2001). Strength of signaling by CD4 and CD8 coreceptor tails determines the number but not the lineage direction of positively selected thymocytes. Immunity 14, 483–494. Bosselut, R., Guinter, T. I., Sharrow, S. O., and Singer, A. (2003). Unravelling a revealing paradox: Why MHC-I signaled thymocytes ‘‘paradoxically’’ appear as CD4þ8lo transitional cells during positive selection of CD8þ T cells. J. Exp. Med. 197, 1709–1719. Bosselut, R., Kubo, S., Guinter, T., Kopacz, J. L., Altman, J. D., Feigenbaum, L., and Singer, A. (2000). Role of CD8beta domains in CD8 coreceptor function: Importance for MHC I binding, signaling, and positive selection of CD8þ T cells in the thymus. Immunity 12, 409–418. Bosselut, R., Zhang, W., Ashe, J. M., Kopacz, J. L., Samelson, L. E., and Singer, A. (1999). Association of the adaptor molecule LAT with CD4 and CD8 coreceptors identiﬁes a new coreceptor function in T cell receptor signal transduction. J. Exp. Med. 190, 1517–1526. Brandle, D., Muller, C., Rulicke, T., Hengartner, H., and Pircher, H. (1992). Engagement of the T-cell receptor during positive selection in the thymus down-regulates RAG-1 expression. Proc. Natl. Acad. Sci. USA 89, 9529–9533. Brugnera, E., Bhandoola, A., Cibotti, R., Yu, Q., Guinter, T. I., Yamashita, Y., Sharrow, S. O., and Singer, A. (2000). Coreceptor reversal in the thymus: Signaled CD4þ8þ thymocytes initially terminate CD8 transcription even when differentiating into CD8þ T cells. Immunity 13, 59–71. Buch, T., Rieux-Laucat, F., Forster, I., and Rajewsky, K. (2002). Failure of HY-speciﬁc thymocytes to escape negative selection by receptor editing. Immunity 16, 707–718. Cardell, S., Tangri, S., Chan, S., Kronenberg, M., Benoist, C., and Mathis, D. (1995). CD1restricted CD4þ T cells in major histocompatibility complex class II-deﬁcient mice. J. Exp. Med. 182, 993–1004. Chan, S. H., Cosgrove, D., Waltzinger, C., Benoist, C., and Mathis, D. (1993). Another view of the selective model of thymocyte selection. Cell 73, 225–236. Chan, S. H., Waltzinger, C., Baron, A., Benoist, C., and Mathis, D. (1994). Role of coreceptors in positive selection and lineage commitment. EMBO J. 13, 4482–4489.

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

123

Cibotti, R., Bhandoola, A., Guinter, T. I., Sharrow, S. O., and Singer, A. (2000). CD8 coreceptor extinction in signaled CD4(þ)CD8(þ) thymocytes: Coordinate roles for both transcriptional and posttranscriptional regulatory mechanisms in developing thymocytes. Mol. Cell. Biol. 20, 3852–3859. Cibotti, R., Punt, J. A., Dash, K. S., Sharrow, S. O., and Singer, A. (1997). Surface molecules that drive T cell development in vitro in the absence of thymic epithelium and in the absence of lineage-speciﬁc signals. Immunity 6, 245–255. Corbella, P., Moskophidis, D., Spanopoulou, E., Mamalaki, C., Tolaini, M., Itano, A., Lans, D., Baltimore, D., Robey, E., and Kioussis, D. (1994). Functional commitment to helper T cell lineage precedes positive selection and is independent of T cell receptor MHC speciﬁcity. Immunity 1, 269–276. Correia-Neves, M., Waltzinger, C., Mathis, D., and Benoist, C. (2001). The shaping of the T cell repertoire. Immunity 14, 21–32. Cosgrove, D., Gray, D., Dierich, A., Kaufman, J., Lemeur, M., Benoist, C., and Mathis, D. (1991). Mice lacking MHC class II molecules. Cell 66, 1051–1066. Crump, A. L., Grusby, M. J., Glimcher, L. H., and Cantor, H. (1993). Thymocyte development in major histocompatibility complex-deﬁcient mice: Evidence for stochastic commitment to the CD4 and CD8 lineages. Proc. Natl. Acad. Sci. USA 90, 10739–10743. Dave, V. P., Allman, D., Wiest, D. L., and Kappes, D. J. (1999). Limiting TCR expression leads to quantitative but not qualitative changes in thymic selection. J. Immunol. 162, 5764–5774. Davey, G. M., Schober, S. L., Endrizzi, B. T., Dutcher, A. K., Jameson, S. C., and Hogquist, K. A. (1998). Preselection thymocytes are more sensitive to T cell receptor stimulation than mature T cells. J. Exp. Med. 188, 1867–1874. Davis, C. B., and Littman, D. R. (1994). Thymocyte lineage commitment: Is it instructed or stochastic? Curr. Opin. Immunol. 6, 266–272. Davis, C. B., Killeen, N., Crooks, M. E., Raulet, D., and Littman, D. R. (1993). Evidence for a stochastic mechanism in the differentiation of mature subsets of T lymphocytes. Cell 73, 237–247. Dembic, Z., Haas, W., Zamoyska, R., Parnes, J., Steinmetz, M., and von Boehmer, H. (1987). Transfection of the CD8 gene enhances T-cell recognition. Nature 326, 510–511. Devine, L., Kieffer, L. J., Aitken, V., and Kavathas, P. B. (2000). Human CD8 beta, but not mouse CD8 beta, can be expressed in the absence of CD8 alpha as a beta beta homodimer. J. Immunol. 164, 833–838. Doyle, C., and Strominger, J. L. (1987). Interaction between CD4 and class II MHC molecules mediates cell adhesion. Nature 330, 256–259. Dutz, J. P., Ong, C. J., Marth, J., and Teh, H. S. (1995). Distinct differentiative stages of CD4þCD8þ thymocyte development deﬁned by the lack of coreceptor binding in positive selection. J. Immunol. 154, 2588–2599. Egerton, M., Scollay, R., and Shortman, K. (1990). Kinetics of mature T-cell development in the thymus. Proc. Natl. Acad. Sci. USA 87, 2579–2582. Fehling, H. J., Krotkova, A., Saint-Ruf, C., and von Boehmer, H. (1995). Crucial role of the pre-T-cell receptor alpha gene in development of alpha beta but not gamma delta T cells. Nature 375, 795–798. Frank, S. J., Samelson, L. E., and Klausner, R. D. (1990). The structure and signaling functions of the invariant T cell receptor components. Semin. Immunol. 2, 89–97. Fung-Leung, W. P., Schilham, M. W., Rahemtulla, A., Kundig, T. M., Vollenweider, M., Potter, J., van Ewijk, W., and Mak, T. W. (1991). CD8 is needed for development of cytotoxic T cells but not helper T cells. Cell 65, 443–449. Gabert, J., Langlet, C., Zamoyska, R., Parnes, J. R., Schmitt-Verhulst, A. M., and Malissen, B. (1987). Reconstitution of MHC class I speciﬁcity by transfer of the T cell receptor and Lyt-2 genes. Cell 50, 545–554.

124

ALFRED SINGER AND REMY BOSSELUT

Glaichenhaus, N., Shastri, N., Littman, D. R., and Turner, J. M. (1991). Requirement for association of p56lck with CD4 in antigen-speciﬁc signal transduction in T cells. Cell 64, 511–520. Godfrey, D. I., and Zlotnik, A. (1993). Control points in early T-cell development. Immunol. Today 14, 547–553. Godfrey, D. I., Kennedy, J., Suda, T., and Zlotnik, A. (1993). A developmental pathway involving four phenotypically and functionally distinct subsets of CD3–CD4–CD8-triple-negative adult mouse thymocytes deﬁned by CD44 and CD25 expression. J. Immunol. 150, 4244–4252. Goldrath, A. W., Hogquist, K. A., and Bevan, M. J. (1997). CD8 lineage commitment in the absence of CD8. Immunity 6, 633–642. Gorman, S. D., Sun, Y. H., Zamoyska, R., and Parnes, J. R. (1988). Molecular linkage of the Ly-3 and Ly-2 genes. Requirement of Ly-2 for Ly-3 surface expression. J. Immunol. 140, 3646–3653. Grusby, M. J., Johnson, R. S., Papaioannou, V. E., and Glimcher, L. H. (1991). Depletion of CD4þ T cells in major histocompatibility complex class II-deﬁcient mice. Science 253, 1417–1420. Guidos, C. J., Danska, J. S., Fathman, C. G., and Weissman, I. L. (1990). T cell receptor-mediated negative selection of autoreactive T lymphocyte precursors occurs after commitment to the CD4 or CD8 lineages. J. Exp. Med. 172, 835–845. Guo, J., Hawwari, A., Li, H., Sun, Z., Mahanta, S. K., Littman, D. R., Krangel, M. S., and He, Y. W. (2002). Regulation of the TCRalpha repertoire by the survival window of CD4(þ)CD8(þ) thymocytes. Nat. Immunol. 3, 469–476. Haks, M. C., Belkowski, S. M., Ciofani, M., Rhodes, M., Lefebvre, J. M., Trop, S., Hugo, P., Zuniga-Pﬂucker, J. C., and Wiest, D. L. (2003). Low activation threshold as a mechanism for ligand-independent signaling in pre-T cells. J. Immunol. 170, 2853–2861. Hashimoto, K., Sohn, S. J., Levin, S. D., Tada, T., Perlmutter, R. M., and Nakayama, T. (1996). Requirement for p56lck tyrosine kinase activation in T cell receptor-mediated thymic selection. J. Exp. Med. 184, 931–943. Hernandez-Hoyos, G., Sohn, S. J., Rothenberg, E. V., and Alberola-Ila, J. (2000). Lck activity controls CD4/CD8 T cell lineage commitment. Immunity 12, 313–322. Hogquist, K. A., Jameson, S. C., Heath, W. R., Howard, J. L., Bevan, M. J., and Carbone, F. R. (1994). T cell receptor antagonist peptides induce positive selection. Cell 76, 17–27. Huesmann, M., Scott, B., Kisielow, P., and von Boehmer, H. (1991). Kinetics and efﬁcacy of positive selection in the thymus of normal and T cell receptor transgenic mice. Cell 66, 533–540. Irie, H. Y., Mong, M. S., Itano, A., Crooks, M. E., Littman, D. R., Burakoff, S. J., and Robey, E. (1998). The cytoplasmic domain of CD8 beta regulates Lck kinase activation and CD8 T cell development. J. Immunol. 161, 183–191. Irie, H. Y., Ravichandran, K. S., and Burakoff, S. J. (1995). CD8 beta chain inﬂuences CD8 alpha chain-associated Lck kinase activity. J. Exp. Med. 181, 1267–1273. Irving, B. A., Alt, F. W., and Killeen, N. (1998). Thymocyte development in the absence of pre-T cell receptor extracellular immunoglobulin domains. Science 280, 905–908. Itano, A., and Robey, E. (2000). Highly efﬁcient selection of CD4 and CD8 lineage thymocytes supports an instructive model of lineage commitment. Immunity 12, 383–389. Itano, A., Kioussis, D., and Robey, E. (1994). Stochastic component to development of class I major histocompatibility complex-speciﬁc T cells. Proc. Natl. Acad. Sci. USA 91, 220–224. Itano, A., Salmon, P., Kioussis, D., Tolaini, M., Corbella, P., and Robey, E. (1996). The cytoplasmic domain of CD4 promotes the development of CD4 lineage T cells. J. Exp. Med. 183, 731–741. Iwashima, M., Irving, B. A., van Oers, N. S., Chan, A. C., and Weiss, A. (1994). Sequential interactions of the TCR with two distinct cytoplasmic tyrosine kinases. Science 263, 1136–1139. Iwata, M., Kuwata, T., Mukai, M., Tozawa, Y., and Yokoyama, M. (1996). Differential induction of helper and killer T cells from isolated CD4þCD8þ thymocytes in suspension culture. Eur. J. Immunol. 26, 2081–2086.

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

125

Izon, D. J., Aster, J. C., He, Y., Weng, A., Karnell, F. G., Patriub, V., Xu, L., Bakkour, S., Rodriguez, C., Allman, D., and Pear, W. S. (2002a). Deltex1 redirects lymphoid progenitors to the B cell lineage by antagonizing Notch1. Immunity 16, 231–243. Izon, D. J., Punt, J. A., and Pear, W. S. (2002b). Deciphering the role of Notch signaling in lymphopoiesis. Curr. Opin. Immunol. 14, 192–199. Izon, D. J., Punt, J. A., Xu, L., Karnell, F. G., Allman, D., Myung, P. S., Boerth, N. J., Pui, J. C., Koretzky, G. A., and Pear, W. S. (2001). Notch1 regulates maturation of CD4þ and CD8þ thymocytes by modulating TCR signal strength. Immunity 14, 253–264. Janeway, C. A., Jr. (1988). T-cell development. Accessories or coreceptors? Nature 335, 208–210. Jenkinson, E. J., Van Ewijk, W., and Owen, J. J. (1981). Major histocompatibility complex antigen expression on the epithelium of the developing thymus in normal and nude mice. J. Exp. Med. 153, 280–292. Kappler, J. W., Roehm, N., and Marrack, P. (1987). T cell tolerance by clonal elimination in the thymus. Cell 49, 273–280. Kawamoto, H., Ohmura, K., Fujimoto, S., Lu, M., Ikawa, T., and Katsura, Y. (2003). Extensive proliferation of T cell lineage-restricted progenitors in the thymus: An essential process for clonal expression of diverse T cell receptor beta chains. Eur. J. Immunol. 33, 606–615. Kaye, J., Hsu, M. L., Sauron, M. E., Jameson, S. C., Gascoigne, N. R., and Hedrick, S. M. (1989). Selective development of CD4þ T cells in transgenic mice expressing a class II MHCrestricted antigen receptor. Nature 341, 746–749. Kearse, K. P., Roberts, J. L., Munitz, T. I., Wiest, D. L., Nakayama, T., and Singer, A. (1994a). Developmental regulation of alpha beta T cell antigen receptor expression results from differential stability of nascent TCR alpha proteins within the endoplasmic reticulum of immature and mature T cells. EMBO J. 13, 4504–4514. Kearse, K. P., Williams, D. B., and Singer, A. (1994b). Persistence of glucose residues on core oligosaccharides prevents association of TCR alpha and TCR beta proteins with calnexin and results speciﬁcally in accelerated degradation of nascent TCR alpha proteins within the endoplasmic reticulum. EMBO J. 13, 3678–3686. Kearse, K. P., Roberts, J. P., Wiest, D. L., and Singer, A. (1995). Developmental regulation of alpha beta T cell antigen receptor assembly in immature CD4þCD8þ thymocytes. Bioessays 17, 1049–1054. Kisielow, P., Teh, H. S., Bluthmann, H., and von, B. H. (1988). Positive selection of antigen-speciﬁc T cells in thymus by restricting MHC molecules. Nature 335, 730–733. Koch, U., Lacombe, T. A., Holland, D., Bowman, J. L., Cohen, B. L., Egan, S. E., and Guidos, C. J. (2001). Subversion of the T/B lineage decision in the thymus by lunatic fringe-mediated inhibition of Notch-1. Immunity 15, 225–236. Koller, B. H., Marrack, P., Kappler, J. W., and Smithies, O. (1990). Normal development of mice deﬁcient in beta 2M, MHC class I proteins, and CD8þ T cells. Science 248, 1227–1230. Kruisbeek, A. M., Fultz, M. J., Sharrow, S. O., Singer, A., and Mond, J. J. (1983). Early development of the T cell repertoire. In vivo treatment of neonatal mice with anti-Ia antibodies interferes with differentiation of I-restricted T cells but not K/D-restricted T cells. J. Exp. Med. 157, 1932–1946. Kruisbeek, A. M., Mond, J. J., Fowlkes, B. J., Carmen, J. A., Bridges, S., and Longo, D. L. (1985). Absence of the Lyt-2-, L3T4þ lineage of T cells in mice treated neonatally with anti-I-A correlates with absence of intrathymic I-A-bearing antigen-presenting cell function. J. Exp. Med. 161, 1029–1047. Lai, E. C. (2002). Notch cleavage: Nicastrin helps Presenilin make the ﬁnal cut. Curr. Biol. 12, R200–202.

126

ALFRED SINGER AND REMY BOSSELUT

Ledbetter, J. A., and Seaman, W. E. (1982). The Lyt-2, Lyt-3 macromolecules: Structural and functional studies. Immunol. Rev. 68, 197–218. Ledbetter, J. A., Seaman, W. E., Tsu, T. T., and Herzenberg, L. A. (1981). Lyt-2 and lyt-3 antigens are on two different polypeptide subunits linked by disulﬁde bonds. Relationship of subunits to T cell cytolytic activity. J. Exp. Med. 153, 1503–1516. Leung, R. K., Thomson, K., Gallimore, A., Jones, E., Van, d. B. M., Sierro, S., Alsheikhly, A. R., McMichael, A., and Rahemtulla, A. (2001). Deletion of the CD4 silencer element supports a stochastic mechanism of thymocyte lineage commitment. Nat. Immunol. 2, 1167–1173. Levin, S. D., Anderson, S. J., Forbush, K. A., and Perlmutter, R. M. (1993). A dominant-negative transgene deﬁnes a role for p561ck in thymopoiesis. EMBO J. 12, 1671–1680. Liao, X. C., and Littman, D. R. (1995). Altered T cell receptor signaling and disrupted T cell development in mice lacking Itk. Immunity 3, 757–769. Linette, G. P., Grusby, M. J., Hedrick, S. M., Hansen, T. H., Glimcher, L. H., and Korsmeyer, S. J. (1994). Bcl-2 is upregulated at the CD4þ CD8þ stage during positive selection and promotes thymocyte differentiation at several control points. Immunity 1, 197–205. Locksley, R. M., Reiner, S. L., Hatam, F., Littman, D. R., and Killeen, N. (1993). Helper T cells without CD4: Control of leishmaniasis in CD4-deﬁcient mice. Science 261, 1448–1451. Love, P. E., Lee, J., and Shores, E. W. (2000). Critical relationship between TCR signaling potential and TCR afﬁnity during thymocyte selection. J. Immunol. 165, 3080–3087. Lucas, J. A., Atherly, L. O., and Berg, L. J. (2002). The absence of Itk inhibits positive selection without changing lineage commitment. J. Immunol. 168, 6142–6151. Lucas, B., Stefanova, I., Yasutomo, K., Dautigny, N., and Germain, R. N. (1999). Divergent changes in the sensitivity of maturing T cells to structurally related ligands underlies formation of a useful T cell repertoire. Immunity 10, 367–376. Lucas, B., Vasseur, F., and Penit, C. (1995). Stochastic coreceptor shut-off is restricted to the CD4 lineage maturation pathway. J. Exp. Med. 181, 1623–1633. Lundberg, K., Heath, W., Kontgen, F., Carbone, F. R., and Shortman, K. (1995). Intermediate steps in positive selection: Differentiation of CD4þ8int TCRint thymocytes into CD4 8þTCRhi thymocytes. J. Exp. Med. 181, 1643–1651. Madrenas, J., Wange, R. L., Wang, J. L., Isakov, N., Samelson, L. E., and Germain, R. N. (1995). Zeta phosphorylation without ZAP-70 activation induced by TCR antagonists or partial agonists. Science 267, 515–518. Marth, J. D., Lewis, D. B., Cooke, M. P., Mellins, E. D., Gearn, M. E., Samelson, L. E., Wilson, C. B., Miller, A. D., and Perlmutter, R. M. (1989). Lymphocyte activation provokes modiﬁcation of a lymphocyte-speciﬁc protein tyrosine kinase (p56lck). J. Immunol. 142, 2430–2437. Marth, J. D., Peet, R., Krebs, E. G., and Perlmutter, R. M. (1985). A lymphocyte-speciﬁc protein– tyrosine kinase gene is rearranged and overexpressed in the murine T cell lymphoma LSTRA. Cell 43, 393–404. Marusic-Galesic, S., Stephany, D. A., Longo, D. L., and Kruisbeek, A. M. (1988). Development of CD4–CD8þ cytotoxic T cells requires interactions with class I MHC determinants. Nature 333, 180–183. Matechak, E. O., Killeen, N., Hedrick, S. M., and Fowlkes, B. J. (1996). MHC class II-speciﬁc T cells can develop in the CD8 lineage when CD4 is absent. Immunity 4, 337–347. McCarthy, S. A., Kruisbeek, A. M., Uppenkamp, I. K., Sharrow, S. O., and Singer, A. (1988). Engagement of the CD4 molecule inﬂuences cell surface expression of the T-cell receptor on thymocytes. Nature 336, 76–79. Merkenschlager, M., Graf, D., Lovatt, M., Bommhardt, U., Zamoyska, R., and Fisher, A. G. (1997). How many thymocytes audition for selection? J. Exp. Med. 186, 1149–1158.

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

127

Mombaerts, P., Clarke, A. R., Rudnicki, M. A., Iacomini, J., Itohara, S., Lafaille, J. J., Wang, L., Ichikawa, Y., Jaenisch, R., Hooper, M. L., et al. (1992). Mutations in T-cell antigen receptor genes alpha and beta block thymocyte development at different stages. Nature 360, 225–231. Muller, K. P., and Kyewski, B. A. (1995). Intrathymic T cell receptor (TcR) targeting in mice lacking CD4 or major histocompatibility complex (MHC) class II: Rescue of CD4 T cell lineage without co-engagement of TcR/CD4 by MHC class II. Eur. J. Immunol. 25, 896–902. Nakayama, T., June, C. H., Munitz, T. I., Sheard, M., McCarthy, S. A., Sharrow, S. O., Samelson, L. E., and Singer, A. (1990). Inhibition of T cell receptor expression and function in immature CD4þCD8þ cells by CD4. Science 249, 1558–1561. Nakayama, T., Singer, A., Hsi, E. D., and Samelson, L. E. (1989). Intrathymic signaling in immature CD4þCD8þ thymocytes results in tyrosine phosphorylation of the T-cell receptor zeta chain. Nature 341, 651–654. Nakayama, T., Wiest, D. L., Abraham, K. M., Munitz, T. I., Perlmutter, R. M., and Singer, A. (1993). Decreased signaling competence as a result of receptor overexpression: Overexpression of CD4 reduces its ability to activate p561ck tyrosine kinase and to regulate T-cell antigen receptor expression in immature CD4þCD8þ thymocytes. Proc. Natl. Acad. Sci. USA 90, 10534–10538. Noguchi, M., Sarin, A., Aman, M. J., Nakajima, H., Shores, E. W., Henkart, P. A., and Leonard, W. J. (1997). Functional cleavage of the common cytokine receptor gamma chain (gammac) by calpain. Proc. Natl. Acad. Sci. USA 94, 11534–11539. Norment, A. M., and Littman, D. R. (1988). A second subunit of CD8 is expressed in human T cells. EMBO J. 7, 3433–3439. Norment, A. M., Salter, R. D., Parham, P., Engelhard, V. H., and Littman, D. R. (1988). Cell–cell adhesion mediated by CD8 and MHC class I molecules. Nature 336, 79–81. Ohoka, Y., Kuwata, T., Tozawa, Y., Zhao, Y., Mukai, M., Motegi, Y., Suzuki, R., Yokoyama, M., and Iwata, M. (1996). In vitro differentiation and commitment of CD4þCD8þ thymocytes to the CD4 lineage, without TCR engagement. Int. Immunol. 8, 297–306. O’Shea, C. C., Thornell, A. P., Rosewell, I. R., Hayes, B., and Owen, M. J. (1997). Exit of the pre-TCR from the ER/cis-Golgi is necessary for signaling differentiation, proliferation, and allelic exclusion in immature thymocytes. Immunity 7, 591–599. Pearse, M., Wu, L., Egerton, M., Wilson, A., Shortman, K., and Scollay, R. (1989). A murine early thymocyte developmental sequence is marked by transient expression of the interleukin 2 receptor. Proc. Natl. Acad. Sci. USA 86, 1614–1618. Pelchen-Matthews, A., Armes, J. E., Grifﬁths, G., and Marsh, M. (1991). Differential endocytosis of CD4 in lymphocytic and nonlymphocytic cells. J. Exp. Med. 173, 575–587. Pelchen-Matthews, A., Boulet, I., Littman, D. R., Fagard, R., and Marsh, M. (1992). The protein tyrosine kinase p561ck inhibits CD4 endocytosis by preventing entry of CD4 into coated pits. J. Cell Biol. 117, 279–290. Pena-Rossi, C., Zuckerman, L. A., Strong, J., Kwan, J., Ferris, W., Chan, S., Tarakhovsky, A., Beyers, A. D., and Killeen, N. (1999). Negative regulation of CD4 lineage development and responses by CD5. J. Immunol. 163, 6494–6501. Petrie, H. T., Livak, F., Schatz, D. G., Strasser, A., Crispe, I. N., and Shortman, K. (1993). Multiple rearrangements in T cell receptor alpha chain genes maximize the production of useful thymocytes. J. Exp. Med. 178, 615–622. 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 alpha beta-expressing cells. Science 256, 1448–1452. Pui, J. C., Allman, D., Xu, L., DeRocco, S., Karnell, F. G., Bakkour, S., Lee, J. Y., Kadesch, T., Hardy, R. R., Aster, J. C., and Pear, W. S. (1999). Notch1 expression in early lymphopoiesis inﬂuences B versus T lineage determination. Immunity 11, 299–308.

128

ALFRED SINGER AND REMY BOSSELUT

Punt, J. A., Suzuki, H., Granger, L. G., Sharrow, S. O., and Singer, A. (1996). Lineage commitment in the thymus: Only the most differentiated (TCRhibcl-2hi) subset of CD4þCD8þ thymocytes has selectively terminated CD4 or CD8 synthesis. J. Exp. Med. 184, 2091–2099. Radtke, F., Wilson, A., Stark, G., Bauer, M., van Meerwijk, J., MacDonald, H. R., and Aguet, M. (1999). Deﬁcient T cell fate speciﬁcation in mice with an induced inactivation of Notch1. Immunity 10, 547–558. Rahemtulla, A., Fung-Leung, W. P., Schilham, M. W., Kundig, T. M., Sambhara, S. R., Narendran, A., Arabian, A., Wakeham, A., Paige, C. J., Zinkernagel, R. M., et al. (1991). Normal development and function of CD8þ cells but markedly decreased helper cell activity in mice lacking CD4. Nature 353, 180–184. Riberdy, J. M., Mostaghel, E., and Doyle, C. (1998). Disruption of the CD4-major histocompatibility complex class II interaction blocks the development of CD4(þ) T cells in vivo. Proc. Natl. Acad. Sci. USA 95, 4493–4498. Robey, E. A., Fowlkes, B. J., Gordon, J. W., Kioussis, D., von Boehmer, H., Ramsdell, F., and Axel, R. (1991). Thymic selection in CD8 transgenic mice supports an instructive model for commitment to a CD4 or CD8 lineage. Cell 64, 99–107. Robey, E., Itano, A., Fanslow, W. C., and Fowlkes, B. J. (1994). Constitutive CD8 expression allows inefﬁcient maturation of CD4þ helper T cells in class II major histocompatibility complex mutant mice. J. Exp. Med. 179, 1997–2004. Robey, E. A., Ramsdell, F., Kioussis, D., Sha, W., Loh, D., Axel, R., and Fowlkes, B. J. (1992). The level of CD8 expression can determine the outcome of thymic selection. Cell 69, 1089–1096. Saint-Ruf, C., Ungewiss, K., Groettrup, M., Bruno, L., Fehling, H. J., and von Boehmer, H. (1994). Analysis and expression of a cloned pre-T cell receptor gene. Science 266, 1208–1212. Samelson, L. E., Phillips, A. F., Luong, E. T., and Klausner, R. D. (1990). Association of the fyn protein–tyrosine kinase with the T-cell antigen receptor. Proc. Natl. Acad. Sci. USA 87, 4358–4362. Schaeffer, E. M., Broussard, C., Debnath, J., Anderson, S., McVicar, D. W., and Schwartzberg, P. L. (2000). Tec family kinases modulate thresholds for thymocyte development and selection. J. Exp. Med. 192, 987–1000. Schmedt, C., and Tarakhovsky, A. (2001). Autonomous maturation of alpha/beta T lineage cells in the absence of COOH-terminal Src kinase (Csk). J. Exp. Med. 193, 815–826. Schmedt, C., Saijo, K., Niidome, T., Kuhn, R., Aizawa, S., and Tarakhovsky, A. (1998). Csk controls antigen receptor-mediated development and selection of T-lineage cells. Nature 394, 901–904. Schmitt, T. M., and Zuniga-Pﬂucker, J. C. (2002). Induction of T cell development from hematopoietic progenitor cells by delta-like-1 in vitro. Immunity 17, 749–756. Sebzda, E., Choi, M., Fung-Leung, W. P., Mak, T. W., and Ohashi, P. S. (1997). Peptide-induced positive selection of TCR transgenic thymocytes in a coreceptor-independent manner. Immunity 6, 643–653. Seong, R. H., Chamberlain, J. W., and Parnes, J. R. (1992). Signal for T-cell differentiation to a CD4 cell lineage is delivered by CD4 transmembrane region and/or cytoplasmic tail. Nature 356, 718–720. Sha, W. C., Nelson, C. A., Newberry, R. D., Kranz, D. M., Russell, J. H., and Loh, D. Y. (1988). Positive and negative selection of an antigen receptor on T cells in transgenic mice. Nature 336, 73–76. Sharp, L. L., and Hedrick, S. M. (1999). Commitment to the CD4 lineage mediated by extracellular signal-related kinase mitogen-activated protein kinase and lck signaling. J. Immunol. 163, 6598–6605. Sharp, L. L., Schwarz, D. A., Bott, C. M., Marshall, C. J., and Hedrick, S. M. (1997). The inﬂuence of the MAPK pathway on T cell lineage commitment. Immunity 7, 609–618.

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

129

Shaw, A. S., Amrein, K. E., Hammond, C., Stern, D. F., Sefton, B. M., and Rose, J. K. (1989). The lck tyrosine protein kinase interacts with the cytoplasmic tail of the CD4 glycoprotein through its unique amino-terminal domain. Cell 59, 627–636. Shaw, A. S., Chalupny, J., Whitney, J. A., Hammond, C., Amrein, K. E., Kavathas, P., Sefton, B. M., and Rose, J. K. (1990). Short related sequences in the cytoplasmic domains of CD4 and CD8 mediate binding to the amino-terminal domain of the p56lck tyrosine protein kinase. Mol. Cell. Biol. 10, 1853–1862. Shores, E. W., Huang, K., Tran, T., Lee, E., Grinberg, A., and Love, P. E. (1994). Role of TCR zeta chain in T cell development and selection. Science 266, 1047–1050. Shores, E. W., Tran, T., Grinberg, A., Sommers, C. L., Shen, H., and Love, P. E. (1997). Role of the multiple T cell receptor (TCR)-zeta chain signaling motifs in selection of the T cell repertoire. J. Exp. Med. 185, 893–900. Singer, A., Mizuochi, T., Munitz, T. I., and Gress, R. E. (1986). Role of self antigens in the selection of the developing T cell repertoire. In ‘‘Progress in Immunology VI’’ (B. Cinader and R. G. Miller, Eds.), pp. 60–66. Academic Press, New York. Singer, A., Munitz, T. I., and Gress, R. E. (1987). Speciﬁcity of thymic selection and the role of self antigens. Transplant Proc. 19, 107–110. Singer, A. (2002). New perspectives on a developmental dilemma: The kinetic signaling model and the importance of signal duration for the CD4/CD8 lineage decision. Curr. Opin. Immunol. 14, 207–215. Sloan-Lancaster, J., Shaw, A. S., Rothbard, J. B., and Allen, P. M. (1994). Partial T cell signaling: Altered phospho-zeta and lack of zap70 recruitment in APL-induced T cell anergy. Cell 79, 913–922. Sohn, S. J., Forbush, K. A., Pan, X. C., and Perlmutter, R. M. (2001). Activated p56lck directs maturation of both CD4 and CD8 single-positive thymocytes. J. Immunol. 166, 2209–2217. Sosinowski, T., Killeen, N., and Weiss, A. (2001). The Src-like adaptor protein downregulates the T cell receptor on CD4þCD8þ thymocytes and regulates positive selection. Immunity 15, 457–466. Stefanova, I., Hemmer, B., Vergelli, M., Martin, R., Biddison, W. E., and Germain, R. N. (2003). TCR ligand discrimination is enforced by competing ERK positive and SHP-1 negative feedback pathways. Nat. Immunol. 4, 248–254. Sudo, T., Nishikawa, S., Ohno, N., Akiyama, N., Tamakoshi, M., and Yoshida, H. (1993). Expression and function of the interleukin 7 receptor in murine lymphocytes. Proc. Natl. Acad. Sci. USA 90, 9125–9129. Suzuki, H., Guinter, T. I., Koyasu, S., and Singer, A. (1998). Positive selection of CD4þ T cells by TCR-speciﬁc antibodies requires low valency TCR cross-linking: Implications for repertoire selection in the thymus. Eur. J. Immunol. 28, 3252–3258. Suzuki, H., Punt, J. A., Granger, L. G., and Singer, A. (1995). Asymmetric signaling requirements for thymocyte commitment to the CD4þ versus CD8þ T cell lineages: A new perspective on thymic commitment and selection. Immunity 2, 413–425. Swat, W., Dessing, M., von Boehmer, H., and Kisielow, P. (1993). CD69 expression during selection and maturation of CD4þ8þ thymocytes. Eur. J. Immunol. 23, 739–746. Takahama, Y., and Singer, A. (1992). Post-transcriptional regulation of early T cell development by T cell receptor signals. Science 258, 1456–1462. Takahama, Y., Shores, E. W., and Singer, A. (1992). Negative selection of precursor thymocytes before their differentiation into CD4þCD8þ cells. Science 258, 653–656. Takahama, Y., Suzuki, H., Katz, K. S., Grusby, M. J., and Singer, A. (1994). Positive selection of CD4þ T cells by TCR ligation without aggregation even in the absence of MHC. Nature 371, 67–70.

130

ALFRED SINGER AND REMY BOSSELUT

Teh, H. S., Garvin, A. M., Forbush, K. A., Carlow, D. A., Davis, C. B., Littman, D. R., and Perlmutter, R. M. (1991). Participation of CD4 coreceptor molecules in T-cell repertoire selection. Nature 349, 241–243. Teh, H. S., Kishi, H., Scott, B., and von Boehmer, H. (1989). Deletion of autospeciﬁc T cells in T cell receptor (TCR) transgenic mice spares cells with normal TCR levels and low levels of CD8 molecules. J. Exp. Med. 169, 795–806. Teh, H. S., Kisielow, P., Scott, B., Kishi, H., Uematsu, Y., Bluthmann, H., and von Boehmer, H. (1988). Thymic major histocompatibility complex antigens and the alpha beta T-cell receptor determine the CD4/CD8 phenotype of T cells. Nature 335, 229–233. Turka, L. A., Schatz, D. G., Oettinger, M. A., Chun, J. J., Gorka, C., Lee, K., McCormack, W. T., and Thompson, C. B. (1991). Thymocyte expression of RAG-1 and RAG-2: Termination by T cell receptor cross-linking. Science 253, 778–781. Turner, J. M., Brodsky, M. H., Irving, B. A., Levin, S. D., Perlmutter, R. M., and Littman, D. R. (1990). Interaction of the unique N-terminal region of tyrosine kinase p56lck with cytoplasmic domains of CD4 and CD8 is mediated by cysteine motifs. Cell 60, 755–765. Van Ewijk, W., Rouse, R. V., and Weissman, I. L. (1980). Distribution of H-2 microenvironments in the mouse thymus. Immunoelectron microscopic identiﬁcation of I-A and H-2K bearing cells. J. Histochem. Cytochem. 28, 1089–1099. van Meerwijk, J. P., and Germain, R. N. (1993). Development of mature CD8þ thymocytes: Selection rather than instruction? Science 261, 911–915. van Oers, N. S., Garvin, A. M., Davis, C. B., Forbush, K. A., Carlow, D. A., Littman, D. R., Perlmutter, R. M., and Teh, H. S. (1992). Disruption of CD8-dependent negative and positive selection of thymocytes is correlated with a decreased association between CD8 and the protein tyrosine kinase, p56lck. Eur. J. Immunol. 22, 735–743. van Oers, N. S., Killeen, N., and Weiss, A. (1994). ZAP-70 is constitutively associated with tyrosine-phosphorylated TCR zeta in murine thymocytes and lymph node T cells. Immunity 1, 675–685. Veillette, A., Bookman, M. A., Horak, E. M., and Bolen, J. B. (1988). The CD4 and CD8 T cell surface antigens are associated with the internal membrane tyrosine–protein kinase p56lck. Cell 55, 301–308. Veillette, A., Bookman, M. A., Horak, E. M., Samelson, L. E., and Bolen, J. B. (1989). Signal transduction through the CD4 receptor involves the activation of the internal membrane tyrosine–protein kinase p56lck. Nature 338, 257–259. Voll, R. E., Jimi, E., Phillips, R. J., Barber, D. F., Rincon, M., Hayday, A. C., Flavell, R. A., and Ghosh, S. (2000). NF-kappa B activation by the pre-T cell receptor serves as a selective survival signal in T lymphocyte development. Immunity 13, 677–689. Wang, F., Huang, C. Y., and Kanagawa, O. (1998). Rapid deletion of rearranged T cell antigen receptor (TCR) Valpha-Jalpha segment by secondary rearrangement in the thymus: Role of continuous rearrangement of TCR alpha chain gene and positive selection in the T cell repertoire formation. Proc. Natl. Acad. Sci. USA 95, 11834–11839. Wiest, D. L., Ashe, J. M., Abe, R., Bolen, J. B., and Singer, A. (1996). TCR activation of ZAP70 is impaired in CD4þCD8þ thymocytes as a consequence of intrathymic interactions that diminish available p56lck. Immunity 4, 495–504. Wiest, D. L., Yuan, L., Jefferson, J., Benveniste, P., Tsokos, M., Klausner, R. D., Glimcher, L. H., Samelson, L. E., and Singer, A. (1993). Regulation of T cell receptor expression in immature CD4þCD8þ thymocytes by p56lck tyrosine kinase: Basis for differential signaling by CD4 and CD8 in immature thymocytes expressing both coreceptor molecules. J. Exp. Med. 178, 1701–1712. Wilkinson, B., and Kaye, J. (2001). Requirement for sustained MAPK signaling in both CD4 and CD8 lineage commitment: A threshold model. Cell Immunol. 211, 86–95.

ANALYSIS OF THE CD4/CD8 LINEAGE DECISION

131

Wilson, A., MacDonald, H. R., and Radtke, F. (2001). Notch 1-deﬁcient common lymphoid precursors adopt a B cell fate in the thymus. J. Exp. Med. 194, 1003–1012. Wyer, J. R., Willcox, B. E., Gao, G. F., Gerth, U. C., Davis, S. J., Bell, J. I., van der Merwe, P. A., and Jakobsen, B. K. (1999). T cell receptor and coreceptor CD8 alphaalpha bind peptide-MHC independently and with distinct kinetics. Immunity 10, 219–225. Yasutomo, K., Doyle, C., Miele, L., Fuchs, C., and Germain, R. N. (2000a). The duration of antigen receptor signaling determines CD4þ versus CD8þ T-cell lineage fate. Nature 404, 506–510. Yasutomo, K., Lucas, B., and Germain, R. N. (2000b). TCR signaling for initiation and completion of thymocyte positive selection has distinct requirements for ligand quality and presenting cell type. J. Immunol. 165, 3015–3022. Yu, Q., Erman, B., Bhandoola, A., Sharrow, S. O., and Singer, A. (2003). In vitro evidence that cytokine receptor signals are required for differentiation of double positive thymocytes into functionally mature CD8þ T cells. J. Exp. Med. 197, 475–487. Zamoyska, R., Vollmer, A. C., Sizer, K. C., Liaw, C. W., and Parnes, J. R. (1985). Two Lyt-2 polypeptides arise from a single gene by alternative splicing patterns of mRNA. Cell 43, 153–163. Zerrahm, J., Held, W., and Raulet, D. H. (1997). The MHC reactivity of the T cell repertoire prior to positive and negative selection. Cell 88, 627–636. Zhang, W., Sloan-Lancaster, J., Kitchen, J., Trible, R. P., and Samelson, L. E. (1998). LAT: The ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell 92, 83–92. Zijlstra, M., Bix, M., Simister, N. E., Loring, J. M., Raulet, D. H., and Jaenisch, R. (1990). Beta 2-microglobulin deﬁcient mice lack CD4–8þ cytolytic T cells. Nature 344, 742–746.

This Page Intentionally Left Blank

advances in immunology, vol. 83

Development and Function of T Helper 1 Cells ANNE O’GARRA* AND DOUGLAS ROBINSON{ *National Institute for Medical Research London NW7 1AA, UK { Imperial College London London SW7 2AZ, UK

I. Introduction

The immune system has developed to be highly specialized and effective in eradicating a wide variety of pathogens with minimum immunopathology. The adaptive arm of the immune response, consisting of antigen-speciﬁc T and B cells, interacts with cells of the innate immune system to mediate an effective response to infectious pathogens. Heterogeneity of T cell responses to pathogens can determine the resistance or susceptibility of a host to such infectious agents. In this regard, at least two subsets of mouse CD4þ T helper cells, termed T helper (Th) 1 and Th2 cells, were identiﬁed in 1986 by Mosmann and Coffman (Cherwinski et al., 1987; Mosmann et al., 1986; Mosmann and Coffman, 1989). These subsets, which were later conﬁrmed to be present in humans (Haanen et al., 1991; Mosmann and Coffman, 1989; Mosmann et al., 1986; Romagnani, 1991; Wierenga et al., 1990), were characterized by the cytokines they produced and were suggested to play distinct roles in ﬁghting infection or, conversely, contributing to immunopathology (Abbas, 1996; Mosmann et al., 1986; Romagnani, 1991; Sher and Coffman, 1992). T helper (Th) 1 cells produce interferon (IFN)-g and lymphotoxin and play a central role in cell-mediated immunity (Mosmann et al., 1986; Sher and Coffman, 1992). The clinical importance of these cells has been conﬁrmed in humans by the demonstration that speciﬁc genetic defects in the Th1 pathway result in extreme susceptibility to microorganisms such as Mycobacteria and Salmonella (Altare et al., 1998; Dofﬁnger et al., 2000a; Newport et al., 1996). Th1 cells have also been implicated in animal models of organ-speciﬁc autoimmune diseases, such as multiple sclerosis and insulin-dependent diabetes, as well as chronic inﬂammatory diseases, including inﬂammatory bowel disease (Liblau et al., 1995; O’Garra and Murphy, 1993; Powrie and Coffman, 1993). However, some studies suggest that not all autoimmune pathologies are attributable to Th1 cell-derived IFN-g (Cua et al., 2003; Watford and O’Shea, 2003), as will be discussed. Th2 cells produce interleukin (IL)-4, IL-5, IL-10, and IL13 and are important in helminth immunity, as well as playing a role in allergic and atopic manifestations (O’Garra, 1998; Romagnani, 1994). 133 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

134

ANNE O’GARRA AND DOUGLAS ROBINSON

In the 1960s, it was demonstrated that humoral and cell-mediated immune responses did not occur in parallel and, in some cases, could be mutually exclusive (Asherson and Stone, 1965; Katsura, 1977; Liew, 2002; Parish, 1972). At the time, this was attributed to observations that CD4þ T cells were heterogeneous (Bottomly and Janeway, 1981; Liew, 2002; Marrack and Kappler, 1975; Tada et al., 1978). Although the array of cytokines produced by the Th1 and Th2 subsets are associated with cell-mediated versus humoral immunity, respectively, such functional correlates are not absolute. Th1 cells, by their production of IFN-g, activate macrophages to produce proinﬂammatory cytokines (Stout and Bottomly, 1989) and to kill intracellular pathogens by mechanisms including NO synthesis (Bogdan et al., 1991). Indeed, it has been shown that Th1 clones can induce DTH responses in vivo (Cher and Mosmann, 1987). However, it is often not noted that Th1 responses are accompanied by an antibody response of mainly the IgG2a isotype induced by IFN-g (Coffman et al., 1991; Snapper and Paul, 1987). IgG2a antibodies are potent inducers of macrophage and killer-cell antibodydependent cellular type cytotoxicity (ADCC) and are also complement ﬁxing (Coffman et al., 1991; Snapper and Paul, 1987). Thus, IFN-g is key in cellmediated immune responses to intracellular pathogens but also in responses in which ADCC, opsonization, and complement-mediated lysis play an important protective role against bacterial and viral pathogens (Coffman et al., 1991; Snapper and Paul, 1987). Th2 cells have always been associated with humoral immune responses, since cytokines which they produce, such as IL-4 and IL-5, are important for B cell growth and differentiation (Coffman et al., 1991; Snapper and Paul, 1987). However, the hallmark of a Th2 response is speciﬁcally the production of IgE and high levels of IgG1 antibodies (Coffman et al., 1991; Snapper and Paul, 1987). These antibody isotypes are inhibited by IFN-g, although low levels of IgG1 antibodies can often be observed within Th1 responses. Both Th1- and Th2-speciﬁc cytokines can promote growth or differentiation of their own respective T cell subset, but additionally can inhibit the development and function of the opposing subset (Abbas et al., 1996). For example, IL-4 can inhibit Th1 development (Hsieh et al., 1992) whereas IFN-g can inhibit the development of Th2 cells (Gajewski and Fitch, 1988). This might explain why Th1 and Th2 responses are often mutually exclusive; hence, the earlier studies that DTH and humoral immune responses do not often coexist (Liew, 2002). The importance of Th1 or Th2 cells in disease outcome was ﬁrst exempliﬁed in the parasitic disease mouse model of Leishmania major infection (Heinzel et al., 1989; Scott et al., 1988). The production of IFN-g by Th1 cells during infection is required for activation of macrophages for the clearance of the parasite via the action of NO and other mediators (Heinzel et al., 1989; Scott et al., 1989). This response leads to resolution of the disease in

FUNCTION OF T HELPER 1 CELLS

135

most mouse strains. BALB/c mice which do not make a Th1 response are susceptible to L. major. Conversely, Th2 cells either injected into resistant mouse strains, or those developing in the BALB/c susceptible strain, by their production of IL-4 inhibit the development of Th1 cells in vivo (Heinzel et al., 1989; Locksley et al., 1991; Scott et al., 1989) and thus prevent parasite clearance. Neutralization of IL-4 during L. major infection of BALB/c mice confers a resistant Th1 phenotype to the mouse and results in resolution of the disease (Sadick et al., 1990). Listeria monocytogenes (Listeria), a gram-positive facultative intracellular bacterium, is associated with severe infections in newborns, the elderly, and immunocompromised individuals (Unanue, 1997). The release of proinﬂammatory cytokines from various cell subsets was found to be pivotal in controlling primary immune responses to Listeria. IFN-g, produced by NK, Th1, and CD8þ T cells, clears Listeria predominantly by its activation of macrophages (reviewed in Unanue, 1997). In addition, TNF plays a role in the clearance of Listeria and this is due only in part to IFN-g since not only does TNF augment IFN-g production by NK cells (Bancroft et al., 1989), but it also synergizes with IFN-g to optimally activate the microbicidal activity of macrophages (Wherry et al., 1991). However, IFN-g does not replace the ability of TNF to resolve a Listeria infection (Tripp et al., 1994). The role of IFN-g in the eradication of Listeria has also been conﬁrmed in IFN-g/ (Harty and Bevan, 1995) and IFN-g receptor/ mice (Huang et al., 1993). In contrast to a primary immune response, the secondary immune response to Listeria is more rapidly induced and is able to overcome a high, normally lethal, dose of bacteria (Mackaness, 1962). This acquired resistance is largely attributable to memory effector CD4þ and CD8þ T cells, in part by their production of IFN-g (Ehlers et al., 1992; Nakane et al., 1989; Tripp et al., 1995a). TNF is also clearly protective in secondary responses to Listeria, and this cytokine may complement the role of IFN-g in bacterial clearance (Nakane et al., 1989; Samson et al., 1995). Unlike infections with L. major, where protective Th1 responses can be inhibited in susceptible strains by Th2 cells producing IL-4, there is no evidence to date that L. monocytogenes ever induces IL-4 production by Th cells, and BALB/c mice clear Listeria infection as readily as other mouse strains. II. Factors Inducing the Development of Th1 Cells and Their Production of IFN-g

A number of studies have indicated that different microbial stimuli, or protein antigens, and the routes or doses whereby they are administered appear to favor the selective induction of either Th1 and Th2 cell subsets (O’Garra, 1998). These factors may favor the involvement of a particular antigen presenting cell as well as the production of particular cytokines by

136

ANNE O’GARRA AND DOUGLAS ROBINSON

either accessory cells/APC or T cells in the microenvironment. These events can then inﬂuence the development of T helper cells producing discrete arrays of cytokines, as discussed in more detail later in this chapter. Two main systems have been used to dissect the molecular mechanisms of Th1 and Th2 development, in terms of both the soluble factors which drive Th1 development and the signaling pathways and transcription factors that are involved (O’Garra and Murphy, 1994). The ﬁrst is the polyclonal stimulation of T cells with, for example, anti-CD3 (plus or minus anti-CD28) and IL-2. The second system makes use of T cells derived from mice expressing transgenic T cell receptors of known speciﬁcities, which can be activated in an antigen-speciﬁc manner. A. IL-12-Driven Th1 Development Using the systems already described, it has been shown that IL-12 is dominant in driving the development of Th1 cells to produce IFN-g upon restimulation (Hsieh et al., 1993a; Manetti et al., 1993; Seder et al., 1993; Trinchieri, 1995). IL-12 is a 75-kDa heterodimer secreted by monocytes (D’Andrea et al., 1992), neutrophils, macrophages, and dendritic cells (reviewed in O’Garra et al., 1995; Trinchieri, 2003). A wide range of DC, when appropriately stimulated, direct the development of Th1 cells largely by their production of IL-12 and/or IFN-a (Cella et al., 2000; de Jong et al., 2002; Ito et al., 2002; Kalinski et al., 1999; Langenkamp et al., 2000; Macatonia et al., 1995; Maldonado-Lopez et al., 2001; Rissoan, 1999; Vieira et al., 2000). IL-12 has multiple activities on NK cells and T cells (D’Andrea et al., 1992; Gazzinelli et al., 1993; Heinzel et al., 1993; Kobayashi et al., 1989; Manetti et al., 1993; Sypek et al., 1993; Tripp et al., 1993) including augmentation of IFN-g production (Chan et al., 1991). IL-12 consists of two subunits, p35 and p40, and upon binding to its cognate receptor, consisting of the IL-12 receptor(R) b1 and the IL-12R b2 chains, activates STAT1, 3, and 4 (Bacon, 1995; Jacobson et al., 1995; Presky et al., 1996; Rogge et al., 1997; Szabo et al., 1995, 1997). Mice deﬁcient in IL-12 and STAT4 had markedly reduced, but not absent, Th1 responses (Kaplan et al., 1996; Magram et al., 1996; Murphy, 1999; Thierfelder et al., 1996). In addition, although IL-12-differentiated CD4þ Th1 cells require STAT4 for subsequent TCR-induced IFN-g production, TCR-triggered CD8þ T cells can produce IFN-g independently of IL-12 and STAT4 (Carter and Murphy, 1999). In contrast to Th1 cells, Th2 cells were shown to lack IL-12 mediated activation of STAT3 and STAT4 (Hilkens, 1996; Szabo et al., 1995, 1997). The molecular basis for this was found to be that Th2 cells down-regulated the IL-12R b chain and so became unresponsive to IL-12 (Rogge et al., 1997; Szabo et al., 1997). Thus, Th1 cells express the IL-12Rb2 and this receptor is upregulated by IFN-g in the mouse (Szabo et al., 1997) and by IFN-a but not IFN-g in the human (Rogge et al., 1997). In contrast, IL-4 down-regulates the expression of

FUNCTION OF T HELPER 1 CELLS

137

IL-12Rb2 on both mouse and human T cells (Rogge et al., 1997; Szabo et al., 1997). Additionally, human allergen-speciﬁc Th2 cells did not signal in response to IL-12 (Hilkens et al., 1996), and furthermore IL-12Rb2 was absent on T cells obtained from the lungs of patients with allergic asthma (a Th2predominant disease) but present on those from patients with sarcoidosis, a Th1-predominant disease (Rogge, 1999). The upregulation of the IL-12Rb2 chain by IFN-g on mouse T cells may, in part, explain earlier observations that IFN-g was required for Th1 development both in vitro and in vivo (Belosevic et al., 1989; Hsieh et al., 1993b; Maggi et al., 1992; Parronchi et al., 1992; Scott and Kaufmann, 1991; Swain et al., 1990), although IFN-g itself could not induce Th1 cell development. The differential expression of the IL-12Rb2 on Th1 and not Th2 cells has been hypothesized to form the basis for Th1 commitment (Rogge et al., 1997; Szabo et al., 1997). B. The Role of Type-I Interferons in Th1 Development As has been discussed, IFN-a and b, but not IFN-g, upregulated the IL12Rb2 on human T cells, enhancing IL-12-driven Th1 development (Rogge et al., 1997). Furthermore, IFN-a and b could themselves induce a greater degree of STAT4 phosphorylation on human but not mouse Th1 cells, as compared to that induced by IL-12 (Rogge et al., 1997, 1998). Indeed, human T cells could be induced to differentiate into Th1 cells producing high levels of IFN-g by Type I IFNs, but not IFN-g, even in the absence of IL-12 (Rogge, 1998). In contrast, IFN-a did not induce the differentiation of mouse Th1 cells nor activate STAT4 phosphorylation (Wenner et al., 1995). Th1 development and activation of STAT4 in human T cells by Type I IFNs has been shown to be due to its recruitment to the IFN-a receptor complex speciﬁcally via the carboxy terminus of STAT2 (Farrar et al., 2000). However, the mouse Stat2 gene harbors a minisatellite insertion that has altered the carboxy terminus and selectively disrupts its capacity to activate STAT4 but not other STATS (Farrar et al., 2000). This defect in murine STAT2 suggests that the signals leading to STAT4 activation and Th1 development in CD4þ T cells are different in in mice and humans. A study has been published showing that IFN-a/b can activate STAT4 in murine T cells although signiﬁcant Type-I IFN-induced IFN-g production was only observed 8 days after LCMV-virus infection in vivo, or after isolation and stimulation in vitro of splenocytes from infected mice at this time point (Nguyen et al., 2002). The authors show that, although a major responding population is the CD8þ T cell, the CD4þ T cells did phosphorylate STAT4 in response to IFN-a/b. They speculate that the apparent contradiction between their ﬁndings and those in previous studies (Farrar et al., 2000; Rogge et al., 1998; Wenner et al., 1995) may result from differences in the concentrations of IFN-a/b and/or strains of mice used (Nguyen et al., 2002). In addition, the same group have previously shown

138

ANNE O’GARRA AND DOUGLAS ROBINSON

that activation of STAT1 by IFN-a/b can inhibit the production of IFN-g and STAT4 phosphorylation (Nguyen et al., 2000) and that the balance between inhibition and activation of IFN-g production may be determined by the relative levels of STAT1 versus STAT4 in T cells during an infection (Nguyen et al., 2002). This intricate interplay may explain differing ﬁndings on the role of IFN-a/b in the activation of STAT4 and Th1 development. C. Restoration of IL-12 Signaling in Th2 Cells Does Not Lead to IFN-g Production As previously discussed, the IL-12Rb2 chain was down-regulated upon differentiation to the Th2 phenotype, preventing IL-12 signaling in these cells (Rogge et al., 1997; Szabo et al., 1997). This down-regulation of the IL-12R was proposed to be a key checkpoint in preventing Th2 cells from responding to IL-12 and producing IFN-g (Szabo et al., 1997). Ectopic expression of IL-12Rb2 in Th2 cells by retroviral expression (Heath, 2000) or as a transgene (Nishikomori et al., 2000) was sufﬁcient to induce STAT4 signaling and proliferation in response to IL-12, which was not observed in control Th2 cells. However, despite the restoration of IL-12-induced STAT4 signaling and proliferation of Th2 cells ectopically expressing the IL-12Rb2, the addition of IL-12 to these cells, in the presence or absence of TCR-signaling, did not lead to IFN-g production (Heath et al., 2000; Nishikomori et al., 2000). This was the case for both polarized Th2 cells and Th2 clones, and furthermore there was no reduction in IL-4 production. These data suggested that IL-12 signaling alone is insufﬁcient to induce the production of IFN-g by Th2 cells. Previous studies had shown that the presence of IFN-g or IL-12 during IL-4-driven Th2 development preserved the ability of these cells to produce IFN-g in addition to preventing down-regulation of the IL-12Rb2 (Szabo et al., 1997). Taken together, the data from all these studies implied that IFN-g provided an extra function, over and above the maintenance of IL-12 signaling, to allow Th2 cells producing IL-4 to also produce IFN-g. More recently, the ability of IFN-g, via STAT1 activation (Afkarian et al., 2002; Lighvani et al., 2001) to induce expression of the Th1-speciﬁc transcription factor T-Bet (Szabo et al., 2000), which can induce remodeling of the IFN-g locus (Mullen et al., 2001), provides a potential explanation for the earlier ﬁndings. This is discussed in more detail later in the text. D. Production of IL-12 and Type-I Interferons by Antigen-presenting Cells A wide range of dendritic cells (DC), when appropriately stimulated, direct the development of Th1 cells largely by their production of IL-12 and/or IFNa (Cella et al., 2000; de Jong et al., 2002; Ito et al., 2002; Kalinski et al., 1999; Langenkamp et al., 2000; Macatonia et al., 1995; Maldonado-Lopez et al., 2001; Rissoan et al., 1999; Vieira et al., 2000). IL-12 production by DC can

FUNCTION OF T HELPER 1 CELLS

139

be stimulated by a large range of microbial products and augmented by CD40ligation or cytokines (Aliberti et al., 2000; Cella et al., 1996, 2000; de Jong et al., 2002; Edwards et al., 2002; Hochrein et al., 2000, 2001; Koch et al., 1996; Langenkamp et al., 2000; Macatonia et al., 1995; Reis e Sousa et al., 1999; Schulz et al., 2000; Vieira et al., 2000). Recognition of some pathogen-derived products involves speciﬁc receptors, such as the Toll-like receptors (TLR), which were initially identiﬁed in Drosophila, but are also expressed in vertebrates (Akira et al., 2001; Medzhitov, 2001; Rock et al., 1998). In humans, distinct DC subsets have been shown to express different TLR and, consequently, to respond to distinct microbial products (Hornung et al., 2002; Jarrossay et al., 2001; Kadowaki et al., 2001). For example, monocyte-derived DC express TLR2, TLR4, and TLR7, whereas plasmacytoid DC express TLR7 and TLR9 (Hornung et al., 2002; Jarrossay et al., 2001; Kadowaki et al., 2001). Interestingly, R848, a ligand for TLR7, induced IFN-a or IL-12 from plasmacytoid-derived DC or monocytederived DC, respectively, both of which resulted in Th1 development (Ito et al., 2002). Similarly to human DC, mouse myeloid DC express TLR2 and TLR4 and respond to peptidoglycan and LPS, respectively, to produce IL-12. These TLR are absent from mouse plasmacytoid-derived DC, in keeping with their lack of responsiveness to peptidoglycan and LPS (Boonstra et al., 2003). Also murine plasmacytoid-derived DC expressed TLR7 and TLR9, in keeping with their production of cytokines, including IL-12 and IFN-a, in response to their respective ligands, R848 and CpG. (Boonstra and O’Garra, unpublished). However, TLR9 expression was not restricted to the mouse plasmacytoidderived DC subset, unlike the human counterpart (Boonstra et al., 2003). In keeping with this differential expression of TLR4 and TLR9, the different DC subsets responded differentially to the respective ligands LPS and CpG to direct Th1 development, at least in part, via production of IL-12 (Boonstra et al., 2003). In addition to TLR ligation, G-protein coupled signaling via CCR5 has also been shown to be critical in the induction of IL-12 production from CD11cþ, CD8-aþ DC (Aliberti et al., 2000). Lack of IL-12 or IFN-a production by certain APC subsets upon stimulation with speciﬁc microbial products may thus reﬂect the lack of expression of the corresponding TLR, and hence their inability to direct Th1 development upon such stimulation (Akira et al., 2001; Hornung et al., 2002; Jarrossay et al., 2001; Kadowaki et al., 2001; Medzhitov, 2001; Wagner, 2001). It is of note that DC may produce IL-12 temporarily and become refractory to further stimuli for the production of IL-12, and this has been referred to as ‘‘exhaustion’’ (Langenkamp et al., 2000) or ‘‘paralysis’’ (Reis e Sousa et al., 1999). Thus, previously stimulated DC (Langenkamp et al., 2000; Reis e Sousa et al., 1999) could help skew the balance away from a protective Th1 response towards a Th0 or Th2 response, perhaps to prevent immunopathology.

140

ANNE O’GARRA AND DOUGLAS ROBINSON

In addition to pathogen-derived products, the strength of signal and/or the antigen dose can affect the development of Th1 or Th2 cells (Constant et al., 1995; Constant and Bottomly, 1997; Hosken et al., 1995; Langenkamp et al., 2000; Ruedl et al., 2000). It is thus possible that distinct DC subsets, by their expression of different levels of MHC class II and/or costimulatory molecules, may change the effective dose of antigen presented by the APC to the T cell. It is therefore important that a range of antigen doses be examined when assessing the ability of DC subsets to direct Th cell development. Indeed, mouse plasmacytoid-derived DC and myeloid DC generated from bone marrow, as well as lymphoid tissue DC, can direct Th1 or Th2 responses, depending on the dose of antigen presented. Although the general ﬁndings were that high dose of antigen favored Th1 and low dose favored Th2 development, the absolute dose of antigen achieving these effects may differ among DC subsets and other APC as well as in response to the prevailing microenvironment (Boonstra et al., 2003). E. IFN-g-Producing Th1 Cells and IL-10: Inhibition of Th1 Development and Function by Effects of IL-10 and Other Inhibitors of APC IL-10 was originally described as a cytokine produced by Th2 cells (Fiorentino et al., 1989); however, it soon became clear that this suppressive cytokine was produced by other cells including Th1 cells, macrophages, and dendritic cells (reviewed in Moore et al., 2001; Trinchieri, 2001). This was perhaps surprising since IL-10 is an inhibitor of Th1 cytokine production (Fiorentino et al., 1989) via its action on the APC to inhibit proinﬂammatory cytokine production and MHC class II and costimulator expression (de Waal Malefyt et al., 1991a,b; Ding and Shevach, 1992; Ding et al., 1993; Fiorentino et al., 1991a,b; Murphy et al., 1994). IL-10 production from Th1 cells was ﬁrst observed in human Th1 clones (Parronchi et al., 1992) and subsequently in mouse polarized Th1 cells from TCR-transgenic mice (Assenmacher et al., 1998), and (O’Garra, unpublished ﬁndings). IL-10 inhibits the ability of macrophages to kill intracellular organisms (Gazzinelli et al., 1992; Moore et al., 2001), the maturation of dendritic cells from monocyte precursors (Allavena et al., 1998; Buelens et al., 1997a,b), and the production of proinﬂammatory cytokines such as IL-12 by dendritic cells and macrophages (D’Andrea et al., 1993; Murphy et al., 1994), a factor important in Th1 development. Surprisingly, macrophages (de Waal Malefyt et al., 1991a; Fiorentino et al., 1991a; Gerber and Mosser, 2001) and dendritic cells (Edwards et al., 2002; McGuirk et al., 2002) may also produce IL-10 under certain conditions, including FcR ligation, and stimulation with microbial products such as zymosan and the ﬁlamentous hemagglutinin from Bordetella pertussis. Whereas certain signals stimulate DC to induce the

FUNCTION OF T HELPER 1 CELLS

141

development of Th1 cells as described earlier, others, such as prostaglandin E2 (Kalinski et al., 1998), cholera toxin, nematode worm, or yeast products, will trigger these DC to direct Th2 development (Bozza et al., 2002; d’Ostiani et al., 2000; Gagliardi et al., 2000; Whelan et al., 2000). This may be through production of IL-10, however, this is not clear in all cases. The relative levels of IL-10 versus IFN-g production by T cells may be an important determinant of the balance between clearance and persistent infection with certain pathogens such as Mycobacterium tuberculosis (Gerosa et al., 1999; Trinchieri, 2001), Mycobacterium avium (Silva et al., 2001), Leishmania major (Kane and Mosser, 2001), Listeria monocytogenes (Silva and Appelberg, 2001; Tripp et al., 1995b), and Toxoplasma gondii (Gazzinelli et al., 1992). Conversely, IL-10 may act as a feedback regulator to suppress immunopathology due to an over-exhuberant Th1 response (Deckert et al., 2001; Li et al., 1999). The source of T cells providing the IL-10 during these responses could be Th1 cells, Th2 cells, or as more recently suggested, regulatory T cells (Belkaid et al., 2001; McGuirk et al., 2002). Such regulatory T cells were ﬁrst described in the setting of regulation of autoimmune and/or inﬂammatory pathologies (Maloy and Powrie, 2001; Sakaguchi, 2000; Shevach, 2000). It is also likely that IL-10 produced by macrophages and dendritic cells early during an immune response will help to limit the Th1 response (as we have discussed). The potential importance of IL-10 versus IFNg responses to clinical response to tuberculosis was demonstrated by the ﬁnding that infected patients with lack of skin responsiveness to M. tuberculosis antigens had reduced peripheral blood T cells proliferation and increased production of IL-10 relative to IFNg (Boussiotis et al., 2000). Furthermore these characteristics persisted after treatment and were antigen speciﬁc: it is suggested that this may contribute to persistent infection in some patients (Delgado et al., 2002). Although it was suggested that the source of IL-10 in these cultures was T cells, antigen presenting cells were also present so may also have contributed. These interactions are summarized in Figure 1. F. Acute Induction of IFN-g Production by IL-12 and IL-18 from Differentiated Th1 Cells IL-18 was ﬁrst described as an IFN-g-inducing factor (IGIF) in a murine model of Propionobacterium acnes-induced shock (Okamura et al., 1995). This cytokine was subsequently characterized as active in promoting proliferation and IFN-g production by Th1 clones and lines and NK cells in both mice and humans and was suggested to have potency similar to that of IL-12 (Kohno et al., 1997; Micallef et al., 1996; Okamura et al., 1995; Ushio et al., 1996). However, the demonstration that Th1 responses were defective in mice with a disrupted IL-12 gene conﬁrmed a central role of IL-12 in Th1 development, showing that IL-18 could not substitute for IL-12 in this response (Magram et al., 1996). Structural analysis and fold recognition suggested that IL-18

142

ANNE O’GARRA AND DOUGLAS ROBINSON

belongs to the IL-1 family as did the observation that IL-18 is synthesized as an inactive precursor that requires cleavage by Caspase-1 for activity (Bazan et al., 1996; Ghayur et al., 1997; Gu et al., 1997). IL-1a was shown to act as a cofactor in IL-12-induced Th1 development in BALB/c but not C57B1/6 mice, whereas IL-1a responsiveness was lost by committed Th1 cells and clones (Lichtman et al., 1988; Shibuya et al., 1998). IL-1a signaling takes place via IL-1-receptor associated kinase (IRAK), which activates a cascade through NIK and CHUCK kinases, leading to activation of nuclear factor-kB (NF-kB) (Cao et al., 1996; Malinin et al., 1997; Regnier et al., 1997). Unlike IL-12, but like IL-1a, IL-18 did not drive Th1 development but potentiated IL-12-induced Th1 development in BALB/c but not C57B1/6 mice (Robinson, 1997). Furthermore, there was marked synergy between IL-18 and IL-12 in inducing IFN-g production from differentiating and committed Th1 cells, suggesting that both IL-12 and IL-18 are required for signiﬁcant expression of the Th1 phenotype. Unlike IL-12, IL-18 did not activate STAT4 activation in Th1 cells, but rather signaled through the IRAK pathway to induce nuclear translocation of the p65/p50 NF-kB complex (Robinson, 1997). In contrast, IL-1a/b showed no effect on differentiated Th1 cells, but activated NF-kB and induced proliferation of Th2 cells, which did not respond to IL-18. Thus, Th1 and Th2 cells were shown to differ in responsiveness to the IL-1 family members IL-18 and IL-1a/b, respectively (Robinson, 1997). Interestingly, the IRAK/NFkB signaling pathway, like the related TLR pathway, is highly conserved throughout evolution and homologous pathways have been found in species as diverse as plants, Drosophila, and mammals (O’Neill and Greene, 1998). The differential activation of a common signaling pathway by the different IL-1 family members, IL-1a/b or IL-18, in either Th1 or Th2 cells, strongly suggested that these cytokines bound to distinct and differentially regulated receptors. The IL-1a receptor consists of the type I IL-1R and the IL-1 accessory protein (Acp) (Greenfeder et al., 1995; Sims et al., 1988). IL-1a/b receptors are lost upon differentiation to the Th1 phenotype (Lichtman et al., 1988). Moreover, antibodies to the type I IL-1R, the type II IL-1R, and the AcP did not block the activity of IL-18 (Hunter et al., 1997). Since then, the IL-18 receptor has been characterized. One chain, designated IL-1Rrp, was ﬁrst identiﬁed as an orphan receptor of the IL-1R family (Parnet et al., 1996) and this was subsequently found to be a receptor for IL-18 (Torigoe et al., 1997; Xu et al., 1998). Consistent with the activity of IL-18 on differentiated Th1 but not Th2 cells, IL-1Rrp expression was shown to be limited to Th1 cells (Torigoe et al., 1997; Xu et al., 1998). Th1 cells derived from IL-1RrPknockout mice failed to activate NFkB in response to IL-18, and their NK cells showed decreased cytolytic activity and decreased IFN-g produced in response to P. acnes (Hoshino et al., 1999). A second chain of the IL-18

FUNCTION OF T HELPER 1 CELLS

143

receptor has been identiﬁed and designated AcPL. Both IL-1Rrp and AcPL are required for IL-18-mediated activation of NFkB and c-jun N-terminal kinase (Born et al., 1998; Debets et al., 2000). IL-1RrP and AcPL (Alpha and Beta) are expressed by Th1 cells and down-regulated in Th2 cells (Debets et al., 2000). Thus, distinct tissue speciﬁcity of IL-1a/b and IL-18 receptors allows these cytokines to utilize a common signaling pathway in distinct Th cell subsets. The molecular mechanism underlying the synergy between IL-18 and IL-12 may be explained, in part, by observations that IL-18 upregulated the IL-12R (Xu, 1998), and IL-12 upregulated the IL-18R (Ahn et al., 1997; Smeltz et al., 2001; Yoshimoto et al., 1998). However, more studies are required to fully understand the molecular nature of the synergy between IL-12 and IL-18 in inducing IFN-g. Strikingly, IL-18 synergized with IL-12 to induce very high levels of IFN-g production from committed Th1 cells in the complete absence of TCR-triggering (Robinson et al., 1997). This IL-12/IL-18-induced IFN-g production was subsequently distinguished from TCR-induced IFN-g production in committed Th1 cells in that the former was insensitive to inhibition by cyclosporin A (Yang et al., 1999). This may have important implications for development of immunopathology during an immune response. Indeed, IL-12- and IL-18driven bystander activation of IFN-g-production by T cells has been reported in Burkholderia pseudomallei infection (Lertmemongkolchai et al., 2001). IL-12 and IL-18-induced IFN-g production correlated with induction of expression of the proteins GADD45b and GADD45g (Yang et al., 2001). Overexpression of GADD45b augmented IFN-g production (Yang et al., 2001) whereas absence of GADD45g decreased IFN-g production, thereby inhibiting Th1 development (Lu et al., 2001). Some evidence indicates that p38 MAP-kinase signaling is involved in Th1 responses (Dong et al., 2002; Rincon et al., 1998) and in the cyclosporin insensitive IL-12- and IL-18-induced IFN-g production (Yang et al., 2001). G. Additional Cell Signaling Pathways in Th1 Development The MAP kinases c-Jun amino-terminal kinase (JNK) and p38 are activated by exposure of cells to cytokines and various environmental stresses. Targets of the p38 and JNK MAP kinase pathways include transcription factors such as c-Jun and ATF2. Activation of c-Jun and ATF2 plays an important role in activating cytokine genes in the innate and adaptive immune response (Rincon and Flavell, 1997; Rincon et al., 1998). Enhanced production of Th2-speciﬁc cytokines was observed in Jnk1-deﬁcient mice, accompanied by nuclear accumulation of NFATc (Dong et al., 2002), under, however, conditions where neither Th1 nor Th2 differentiation was favored. This suggested a modulatory effect rather than a dominant effect on commitment for cytokine gene expression. In Jnk2-deﬁcient T cells, a substantial reduction in IFN-g production was seen,

144

ANNE O’GARRA AND DOUGLAS ROBINSON

which, in part, could be explained by a decreased expression of the IL-12Rb2 and, thus, reduced STAT4 activation required for IFN-g production (Yang et al., 1998). In addition, in one study, a dominant-negative inhibitory mutant of p38 selectively impaired Th1 responses and activation of the IFN-g promoter (Rincon et al., 1998), as did speciﬁc inhibitors of p38 MAP kinase. However, in other studies, p38 MAP kinase was also implicated in Th2 cytokine production (Chen et al., 2000a; Schafer et al., 1999). The small guanosine 50 triphosphate GTP-binding protein Rac2 activates Jnk, p38, and NF-Kb in various cell types and was shown to be Th1-speciﬁc, and to play a central role in Th1 development since dominant negative-Rac inhibited IFN-g production (Li et al., 2000). In addition, T cells from Rac2-deﬁcient mice showed decreased IFN-g production under Th1 conditions in vitro (Li et al., 2000). III. Role of Th1 Signaling Components in In Vivo Immune Responses

Targeted disruption of IL-12 p40 (Magram et al., 1996), IL-12Rb1 (Wu et al., 1997), and STAT4 (Thierfelder et al., 1996) each resulted in mice with a reduced Th1 response and impaired cell-mediated immunity to infectious diseases. IL12 p40 knockout mice or antibodies directed against the IL-12 p40 chain inhibit the protective response to pathogens such as Toxoplasma gondii (Gazzinelli et al., 1993; Yap et al., 2000), which results from a Th1 response where IFN-g stimulates macrophage effector function. The protective Th1 response to Leishmania major was also shown to depend on IL-12 p40 and STAT4 (Mattner et al., 1996; Park et al., 2000; Stamm et al., 1999; Sypek et al., 1993). Furthermore, IL-12 was able to revert an early Th2 response to L. major, leading to Th1-mediated protection (Heinzel et al., 1993). Although important for protection against a number of intracellular pathogens such as L. major, T. gondii, and mycobacteria (Cooper et al., 2002, and reviewed in Sacks and Noben-Trauth, 2002; Trinchieri, 2003), IL-12 appears to be less important in resistance to infections with viruses or chlamydia (Geng et al., 2000; Monteiro et al., 1998; Orange et al., 1995). This key macrophage product also provides protection from Listeria by inducing IFNg production, as shown using antiIL-12 p40 blocking antibodies (Tripp et al., 1994) or IL-12/ mice (Oxenius et al., 1999). In addition, studies in TNF receptor I/ (Pfeffer et al., 1993; Rothe et al., 1993) mice suggest that TNF appears to be more dominant than IL-12 in eliciting protective primary immune responses to Listeria (Tripp et al., 1993). Although IL-12 is also produced in a secondary challenge with this organism, it offers less protection than in primary responses and cannot completely account for the effects of IFN-g (Tripp et al., 1995), indicating that there may be other players involved in IFN-g production at this stage. The importance of IL-18 in Th1 responses has been shown for a variety of pathogens. IL-18 knockout mice were more susceptible to BCG (Takeda et al.,

FUNCTION OF T HELPER 1 CELLS

145

1998), Leishmania major, Shigella (Sansonetti et al., 2000), and to septic arthritis induced by Staphlococcus aureus (Wei et al., 1999). Moreover, antiIL-18 treatment exacerbated Salmonella infection (Dybing et al., 1999; Mastroeni et al., 1999) and Yersinia (Bohn et al., 1998), whereas IL-18 protein treatment ameliorated Yersinia and Vaccinia infections (Bohn et al., 1998; Tanaka-Kataoka et al., 1998) and herpes simplex virus type 1 infection (Fujioka et al., 1999). An IFN-g-independent protective role for IL-18 was seen by showing that IL-18 treatment improved symptoms in IFN-g-depleted mice infected with Vaccinia virus (Tanaka-Kataoka et al., 1998). A role for IL18 in both primary and recall responses to Listeria infections (Neighbors et al., 2000) was observed not only on Th1 cell production of IFN-g but also in the induction of TNF by macrophages, which was observed during Listeria infection. This could account for some of its IFN-g independent effects reported previously in bacterial clearance, and also for its apparently superior effects to IL-12 in both primary and secondary responses to Listeria (Neighbors et al., 2000). When compared with IL-12 p40 knockout mice in their responses to BCG, IL-18-deﬁcient mice showed similarly reduced responses as measured by IFN-g production in a recall assay. Production of IFN-g in response to BCG infection was almost totally abrogated in mice lacking both IL-12 and IL18, suggesting that IL-12 and IL-18 play distinct and nonredundant roles, at least in the response to this pathogen in mice (Takeda et al., 1998). Th1 cells have been implicated in autoimmune and inﬂammatory pathologies including inﬂammatory bowel disease (Liblau et al., 1995; Powrie and Coffman, 1993; Powrie et al., 1994). Furthermore, blocking of IL-12 p40 has been shown to ameliorate a number of autoimmune pathologies such as diabetes and experimental autoimmune encephalomyelitis (Leonard et al., 1995; Segal et al., 1998; Trembleau et al., 1995) and, in addition, administration of IL-12 led to the exacerbation of these diseases (Segal et al., 1998; Trembleau et al., 1995). It has been shown, however, that IL-12 p35-deﬁcient mice remain highly susceptible to the induction of EAE (Becher et al., 2002), while IL-12 p40 mice were protected. Indeed, the same has been shown in hapten-induced colitis (Camoglio et al., 2002). This suggested either that p40/p40 homo-dimers could substitute for IL-12 in IL-12 p35/ mice or that another factor heterodimerizes with IL-12 p40, which is important for mediating EAE. Indeed, this has now been shown to be the case for a relatively novel cytokine named IL-23, which is composed of the IL-12 p40 chain and a novel p19 molecule (Cua et al., 2003; Oppmann et al., 2000). This would indicate either that Th1 responses driven by IL-12 are not required for EAE, or that IL-23 can substitute for IL-12 in directing Th1 responses to induce EAE. Alternatively, it is possible that IL-23 mediates EAE via a different mechanism, as discussed later in this chapter. In addition to IL-23, IL-18 may also play a role in autoimmune andinﬂammatory pathologies, by its action on T cells, as well as by its action

146

ANNE O’GARRA AND DOUGLAS ROBINSON

on macrophages (Xu, 1998). Anti-IL-18R antibody was shown to reduce local joint inﬂammation induced with carrageenan and to reduce mortality associated with LPS-induced shock (Xu, 1998). Thus, since IL-18 has also been implicated in the development of several other autoimmune diseases in the mouse (Gracie et al., 1999; Pizarro et al., 1999; Rothe et al., 1997; Shi et al., 2000; Wildbaum et al., 1998) and plays a dominant role in the initiation and maintenance of cell-mediated immune responses in vivo (Neighbors et al., 2000), antagonists of this molecule may provide a therapeutic strategy not only for the prevention of but also for the treatment of these disorders. A role for Th1 cells producing IFN-g in EAE has already been questioned since IFN-g/ mice surprisingly showed exacerbation of rather than protection against EAE (Chu et al., 2000; Matthys et al., 2001; Willenborg et al., 1996). This is in contrast to the clearly deﬁned role for Th1 cells producing IFN-g for protection against infectious agents such as L. major, Listeria, and myobacteria, as described earlier in this chapter. The role of Th1 responses in mycobacterial infections has been deﬁnitively shown in human genetic disease where mutations in the Th1 signaling pathway render patients susceptible to mycobacterial and salmonella infections, as discussed later in this chapter. IV. Transcriptional Regulation for Th1 Cells Producing IFN-g

A. IFN-g Gene/Locus It has been demonstrated that a region of 8.6 Kb of the IFN-g gene is sufﬁcient to confer its T cell-speciﬁc expression (Hardy, 1985, 1987). However, as yet, the IFN-g gene has not been extensively characterized with respect to its speciﬁc expression in Th1 and not in Th2 cells (reviewed in Agarwal and Rao, 1998a,b; Murphy et al., 2000). Two essential regulatory elements in the IFN-g gene have been described which confer inducibility and cyclosporin sensitivity. Putative binding sites for the transcription factors ATF, Jun, CREB, and Oct-1 have also been described (Penix et al., 1993, 1996). Both proximal and distal elements in transgenic reporter mice have been shown to direct reporter expression in memory effector CD4þ T cells (Aune et al., 1997). The distal sites were shown to contain a consensus GATA motif and a potential regulatory motif found in the promoter regions of other cytokine genes, and factors binding to this site included GATA-3 (Penix et al., 1993). This is of interest since GATA-3 expression has since been shown to be Th2-speciﬁc and critical for differentiation of Th2 cells (Lee et al., 1998, 2000; Zhang, 1997; Zheng and Flavell, 1997, and reviewed later in this chapter). NFAT and NF-kB have been shown to bind similar regions in the ﬁrst intron of the IFN-g gene and to cooperate in these events (Sica et al., 1997). A ubiquitous nuclear factor, Yin-Yang 1 (YY1), has been shown to bind to two regions of the IFN-g promoter and proposed to act as a repressor of basal IFN-g transcription (Ye et al., 1996), although these transcription factors

FUNCTION OF T HELPER 1 CELLS

147

have been proposed to regulate IFN-g expression (Campbell et al., 1996; Sica et al., 1997; Sweetser et al., 1998). A future detailed functional analysis is required to deﬁne their speciﬁc roles and their regulation by cytokines acting to promote or inhibit Th1 development. B. Transcription Factors Involved in Th1 Development The transcription factor T-bet has been shown to have an important role in Th1 development. T-bet was isolated from yeast 2 hybrid analysis of an activated Th1 library to a Th1-speciﬁc portion of the IL-2 promoter, coupled with RDA (Szabo et al., 2000). T-bet expression correlated with IFN-g production by Th1 cells and NK cells, and T-bet was a potent transactivator of the IFN-g gene. Retroviral gene transduction of T-bet into primary T cells or developing Th2 cells could activate IFN-g production (Szabo et al., 2000). T-bet-deﬁcient mice had default Th2 development and developed spontaneous airway changes similar to those of asthma (a Th2-driven disease) (Finotto et al., 2002). It is of note that T-bet did not appear to be required for IFN-g production by CD8þ T cells, although it was required for CD4þ Th1 responses (Szabo et al., 1997). T-bet induction and Th1 development could still occur in Stat4 deﬁcient mice (Mullen et al., 2001) when T cells were stimulated in the presence of anti-IL-4 mAb. Furthermore, ectopic expression of T-bet in a retroviral vector induced IFN-g production from Stat4/ T cells even when cultured in IL-4 and anti-IL-12. T-bet thus appears to act before and independently of IL-12 induced Stat4 activation in Th1 development, and induces chromatin remodeling of the IFN-g locus (Mullen et al., 2001). However, other investigators reported that Stat4 activation could further increase T-bet expression (Grogan et al., 2001). More recently, two groups have shown that T-bet expression is induced upon Stat1 activation, which can occur through the IFN-g receptor (Afkarian et al., 2002; Lighvani et al., 2001). Indeed, Stat1deﬁcient T cells failed to express T-bet despite IFN-g induction under Th1-polarizing conditions (IL-12 and anti-IL-4) (Afkarian et al., 2002). In this study, T-bet-induced IL-12Rb2 expression was independent of Stat1 and there was no evidence for auto-activation of T-bet (Afkarian et al., 2002). Thus, a reason for failure of T-bet deﬁcient T cells to mount Th1 responses could be an inability to induce the IL-12Rb2. Expression of the IL-12Rb2 was seen in Th1 cells derived from Stat1/ mice (Afkarian et al., 2002), which may indicate a T-bet-independent pathway to IL-12Rb2 expression. Mullen et al. also reported that T-bet was essential for maintaining IL-12Rb2 expression in Th1 cells (2001). A genetic interaction has been described between T-bet and a Th1-restricted homeobox transcription factor, Hlx, which was required to initiate optimal induction of IFN-g (Mullen et al., 2002). GATA-3 is expressed by Th2 cells and not Th1 cells (Lee et al., 1998; Zhang et al., 1997; Zheng and Flavell, 1997) and strongly transctivates the IL-5

148

ANNE O’GARRA AND DOUGLAS ROBINSON

Fig 1 Th1 cell development in protection against intracellular pathogens such as mycobacteria: dendritic cells (DC) and macrophages (Mf) produce cytokines (IL-12 and IL-27) that drive Th1 development, while IL-12 and IL-18 are important for maximal IFN-g production from effector Th1 cells to activate macrophage killing and opsonising antibody production. Th1 cells themselves, as well as DC and Mf, can produce IL-10 which acts to control this process and prevent immunopathology. (See Color Insert.)

promoter but only weakly activates the IL-4 promoter (Lee et al., 1998; Ranganath et al., 1998; Zhang et al., 1997) and has been shown to remodel the IL-4 locus (Lee et al., 2000; Ouyang et al., 2000). IFN-g expression was signiﬁcantly reduced by ectopic expression of GATA-3 in developing Th1 cells (Ferber et al., 1999; Ouyang et al., 1998). Thus GATA-3 not only up-regulates production of Th2 cytokines (Ferber et al., 1999; Ouyang et al., 1998; Zheng and Flavell, 1997) but also inhibits IFN-g expression. This has been shown in both IL-4 and STAT6 knockout mice (Ferber et al., 1999; Ouyang, 1998; Ouyang et al., 2000). Ouyang et al. showed that GATA-3 acted during the early stages of Th1 cell development to down-regulate the IL-12Rb2, thus making the cells refractory to IL-12 (Ouyang et al., 1998). This was proposed to be one mechanism by which GATA-3 could down-regulate Th1 development.

FUNCTION OF T HELPER 1 CELLS

149

V. New Cytokines: What is Their Role in Th1 Cell Responses?

IL-27, a new heterodimeric cytokine, is related to IL-12 and acts early together with IL-12 to trigger IFN-g production by naive CD4þ T cells. IL27 is the ligand for TCCR/WSX-1 (Chen et al., 2000b; Yoshida et al., 2001), a novel member of the class I cytokine receptor family, and has been shown to be important for Th1 development. IL-27 is a heterodimer consisting of EBI3, an IL-12 p40-related protein (Devergne et al., 1997) and p28, a newly discovered IL-12 p35-related polypeptide (Pﬂanz et al., 2002). Expression of both components of the IL-27 heterodimer are up-regulated in antigen-presenting cells upon activation with LPS. IL-27 induced signiﬁcant proliferation of naive but not memory T cells (in both mouse and human systems) in the absence of IL-2 (Pﬂanz et al., 2002). These ﬁndings are in keeping with a high level of mRNA expression for TCCR by unidifferentiated CD4þ T cells as compared to Th1- or Th2-polarized cells (Chen et al., 2000b). Since IL-27 synergized with IL-12 for induction of IFN-g production by naive T cells, this suggests that IL27 may prime T cells for subsequent IL-12-induction of IFN-g or, conversely, that both signals are required (Fig. 2). However, Chen et al., have shown that the TCCR/WSX-1 receptor (for IL-27) was not required for the induction of the IL-12R on splenocytes upon activation (Chen et al., 2000b). Thus, IL-27 appears to act at an early stage in Th1 development in a manner distinct from that of IL-12. Two groups have shown that mice made deﬁcient in the TCCR/WSX-1 receptor have signiﬁcant impairment of Th1 responses. Chen et al. showed that TCCR-deﬁcient (TCCR/) mice, upon challenge with protein antigen, had impaired Th1 responses as measured by IFN-g production, as well as reduced IgG2a production, as compared to wild type mice (Chen et al., 2000b). Moreover, there was a profound defect in clearance of Listeria monocytogenes in these TCCR deﬁcient mice. An IL-12-driven Th1 response from enriched CD4þ T cells from TCCR/ mice, stimulated in vitro, was also signiﬁcantly impaired as compared to that in wild-type mice. However, IL-12-induced splenocyte proliferation and upregulation of the IL-12Rb1 and 2 expression upon Con A activation of splenocytes was unimpaired. More recently, Yoshida et al. showed that CD4þ T cells obtained from WSX-1/ mice (an alternative name for TCCR) showed reduced IFN-g production upon primary stimulation under similar Th1-inducing conditions to those used by Chen et al. (IL-12, anti-IL-4, IL-2, Con A, and irradiated spleen cells) (Yoshida et al., 2001). However, in contrast to the ﬁndings of Chen et al., these T cells appeared to recover with respect to IFN-g production upon secondary restimulation in vitro. WSX-1/ mice were markedly more susceptible to Leishmania major infection showing increased footpad swelling and increased numbers of organisms recovered from tissues, as compared to wild-type C57B1/6 mice.

Fig 2 Checkpoints in Th1 development. Presentation of pathogen antigens to naive T cells by antigen-presenting cells (APC) induces IL-12Rb2 expression on T cells. Dendritic cell or macrophage-derived IL-12, together with IL-27, initiates Th1 development. IL-27 signals through the T cell cytokine receptor (TCCR, also termed WSX-1). IL-12 acts via STAT4 to upregulate IFN-g production. Th1 development requires IFN-g itself, which activates STAT1 and induces expression of the transcription factor T-bet. T-bet is a major Th1 commitment factor and transactivates the IFN-g gene, as well as inducing chromatin remodeling of the IFN-g locus. IL-12 induces IL-18 receptor expression, which allows IL-18 to synergize with IL-12 to increase IFN-g production from committed Th1 effectors. IL-18 signals via IL-1 receptor-associated kinase (IRAK) to activate NF-kB, and the combination of IL-12 and IL-18 activates GADD45 and p38 mitogen kinase signaling, all of which are major ampliﬁers of the IFN-g response. Memory CD4þT cells respond to IL-23, an IL-12-related cytokine, which increases proliferation and may increase IFN-g production further, although this may not be its primary function, since IL-23 also acts on macrophages. (See Color Insert.)

FUNCTION OF T HELPER 1 CELLS

151

There was impaired IFN-g production early in the infection (at 2 weeks), although, in accordance with the in vitro data on Th1 development, at 4 weeks after infection, L. major-induced IFN-g production (by draining lymph node cells restimulated in vitro) was equivalent in WSX-1/ and wild-type mice. However, this recovery of IFN-g production was not obviously associated with healing of the disease since footpad swelling and numbers of inﬁltrating organisms remained elevated as compared to wild-type littermates. This early requirement for effective Th1 immunity in L. major clearance was paralleled by increased granuloma formation to BCG. However, in this case, hepatic mycobacterial counts were not affected, suggesting that other mechanisms can control some bacterial infections. Thus, both studies indicate that mice deﬁcient in a functional TCCR/WSX-1 receptor show impaired Th1 responses, accompanied by an impairment in the clearance of pathogens including L. monocytogenes and L. major. However, the latter study suggests that this requirement for signaling through TCCR/WSX-1 may be most important early in an immune response. The deﬁnition of the IL-27 interaction with TCCR/WSX-1 is a further step in understanding the complex checkpoints in Th1 development and underpins the importance of multiple levels of control for IFN-g production for achieving optimal Th1 protective immunity to infection without induction of immunopathology. IL-23, consisting of a novel protein, p19, which binds to the p40 subunit of IL-12, has been reported to act selectively in proliferation and IFNg production from human memory T cells and proliferation of memory T cells in the mouse (Oppmann et al., 2000). Studies of IL-23 receptor expression (Parham et al., 2002) and IL-23 overexpression in transgenic mice (Wiekowski et al., 2001), however, suggest that IL-23 may have a direct effect on macrophage function. This may explain ﬁndings from Cua et al. (2003) showing that IL-23 and not IL-12 has a central role in autoimmune inﬂammation in the mouse EAE model. These authors suggested that unlike IL-12, IL-23 acts more broadly as an end-stage effector cytokine through its direct action on macrophages rather than through induction of IFN-g production from Th1 cells. This is in keeping with observations that while IL-12 p40-deﬁcient mice are protected from EAE induction (because they would lack both IL-12 and IL-23), in contrast, IL-12 p35-deﬁcient mice (which lack only IL-12) remain susceptible (Becher et al., 2002). VI. Inherited Disorders of IL-12 and IFN-g-mediated Immunity: Clinical Outcomes of Defects in Th1 Development

A rare clinical syndrome resulting from defects in various components of the IL-12/IFN-g signaling pathway has been termed Mendelian susceptibility to mycobacterial disease (MSMD) (Dofﬁnger et al., 2000a). Mutations in ﬁve

152

ANNE O’GARRA AND DOUGLAS ROBINSON

different genes (IFN-gR1, IFN-gR2, IL-12p40, IL-12Rb1, and STAT1) have been implicated in this condition, and allelic variation in these mutations has allowed clinical separation of nine distinct inherited disorders. While the individual syndromes are rare, the disease patterns and response to treatment are of considerable interest since they map the potential role of different components of the Th1 pathway in host defense. Furthermore, polymorphisms in the IL-12 promoter have been described (Pravica et al., 2000) and it is possible that this could affect disease susceptibility more broadly. Complete deﬁciency of the chains of the IFN-g receptor (IFN-gR1 or 2) predisposes patients to overwhelming infection with atypical mycobacteria (such as BCG, M. bovis, and M. avium) in early childhood (Dorman and Holland, 1998; Newport et al., 1996). These lesions are associated with poor granuloma formation leading to infections which resemble lepromatous leprosy, with little inﬂammation and massive bacterial load. This variant is generally rapidly fatal without bone marrow transplantation. A partial IFN-gR deﬁciency leads to milder forms of disease with preserved granuloma formation: these children generally respond to antibacterial therapy or IFN-g treatment (Dofﬁnger et al., 2000b; Dupuis et al., 2000). Other genotypes result in either defective STAT1 (Dupuis et al., 2001) or defects in IL-12-induced activation of STAT1 (Gollob et al., 2000b): Both give a similar phenotype to partial IFN-gR deﬁciency. Defects in IL-12 p40 mutations lead to reduced IL-12 production. These patients tended to get recurrent Salmonella enteriditis infections or mycobacterial infection (Altare et al., 1998). A further group of patients had mutations leading to absence of IL-12Rb1 (Altare et al., 1998; de Jong et al., 1998). These patients had either salmonella or mycobacterial infections. Interestingly, in one report, four of these patients were shown to have preserved DTH responses with good skin reactivity to tuberculin and granuloma formation was intact (de Jong et al., 1998). This raises the issue of how these defects lead to susceptibility to infection and suggest that DTH is not sufﬁcient to kill these intracellular pathogens. Whether bactericidal mechanisms, such as macrophage NO production, or antibody responses are affected in these syndromes remains unclear. DTH may be important in persistence or clearance of M. Tuberculosis and this appears to be determined in part by the relative balance of IL-10 and IFNg producion (Boussiotis et al., 2000; Delgado et al., 2002). These syndromes provide graphic conﬁrmation that appropriate Th1 responses are essential to protect against certain intracellular pathogens (Dupuis et al., 2002). In addition, these patients allow further dissection of the role of these components in IFN-g production in humans. For example, mycobacterial antigen-speciﬁc CD4þ T cell clones generated from IL-12Rb1-deﬁcient patients showed a preserved ability to make IFN-g upon stimulation with IL12, which was further augmented by IL-18. Interestingly, this response was not due to STAT4 phosphorylation (which was not seen in these clones in response

FUNCTION OF T HELPER 1 CELLS

153

to IL-12), but was blocked by the inhibitor of p38 MAPK, SB203580 (Verhagen et al., 2000). This suggests that IL-12 Rb2 can signal via p38 MAPK to increase IFN-g in a STAT4-independent pathway. It will be of interest to see whether similar syndromes occur for T-bet, IL-27, or TCCR/WSX-1, and whether speciﬁc targeting of these Th1 checkpoints will be useful for manipulating immunity to infection or ameliorating Th1-induced autoimmune pathology. Acknowledgments We thank Dr. Andre Boonstra for reviewing the manuscript and the graphics Department at NIMR for construction of the ﬁgures. Douglas Robinson is supported by the Wellcome Trust, UK; Anne O’Garra is supported by the Medical Research Council, UK.

References Abbas, A. K., Murphy, K. M., and Sher, A. (1996). Nature 383, 787. Afkarian, M., Sedy, J. R., Yang, J., Jacobson, N. G., Cereb, N., Yang, S. Y., Murphy, T. L., and Murphy, K. M. (2002). Nat. Immunol. 3, 549. Agarwal, S., and Rao, A. (1998a). Current Biol. 10, 345. Agarwal, S., and Rao, A. (1998b). Immunity 9, 765. Ahn, H. J., Maruo, S., Tomura, M., Mu, J., Hanaoka, T., Nakanishi, K., Clark, S., Kurimoto, M., Okamura, H., and Fujiwara, H. (1997). J. Immunol. 159, 2125. Akira, S., Takeda, K., and Kaisho, T. (2001). Nat. Immunol. 2, 675. Aliberti, J., Reis e Sousa, C., Schito, M., Hieny, S., Wells, T., Huffnagle, G. B., and Sher, A. (2000). Nat. Immunol. 1, 83. Allavena, P., Piemonti, L., Longoni, D., Bernasconi, S., Stoppacciaro, A., Ruco, L., and Mantovani, A. (1998). Eur. J. Immunol. 28, 359. Altare, F., Durandy, A., Lammas, D., Emile, J. F., Lamhamedi, S., Le Deist, F., Drysdale, P., Jouanguy, E., Dofﬁnger, R., Bernaudin, F., Jeppsson, O., Gollob, J. A., Meinl, E., Segal, A. W., Fischer, A., Kumararatne, D., and Casanova, J. L. (1998). Science 280, 1432. Asherson, G. L., and Stone, S. H. (1965). Immunology 9, 205. Assenmacher, M., Lohning, M., Scheffold, A., Richter, A., Miltenyi, S., Schmitz, J., and Radbruch, A. (1998). J. Immunol. 161, 2825. Aune, T. M., Penix, L. A., Rincon, M., and Flavell, R. A. (1997). Mol. Cell. Biol. 17, 199. Bacon, C. M., Petricoin, E. F., 3rd., Ortaldo, J. R., Rees, R. C., Larner, A. C., Johnston, J. A., and O’Shea, J. J. (1995). Proc. Natl. Acad. Sci. USA 92, 7307. Bancroft, G. J., Sheehan, K. C. F., Schreiber, R. D., and Unanue, E. R. (1989). J. Immunol. 143, 127. Bazan, J. F., Timans, J. C., and Kastelein, R. A. (1996). Nature 379, 591. Becher, B., Durell, B. G., and Noelle, R. J. (2002). J. Clin. Invest. 110, 493. Belkaid, Y., Piccirillo, C. A., Mendez, S., Shevach, E. M., and Sacks, D. L. (2001). Nature 420, 502. Belosevic, M., Finbloom, D. S., Meide, P. H. V. D., Slayter, M. V., and Nacy, C. A. (1989). J. Immunol. 143, 266. Bogdan, C., Vodovotz, Y., and Nathan, C. (1991). J. Exp. Med. 174, 1549. Bohn, E., Sing, A., Zumbihl, R., Biefeldt, C., Okamura, H., Kurimoto, M., Heeseman, J., and Autenrieth, I. B. (1998). J. Immunol. 160, 299. Boonstra, A., Asselin-Paturel, C., Gilliet, M., Crain, C., Trinchieri, G., Liu, Y.-J., and O’Garra, A. (2003). J. Exp. Med. 197, 101. Born, T. L., Thomassen, E., Bird, T. A., and Sims, J. E. (1998). J. Biol. Chem. 273, 29445. Bottomly, K., and Janeway, C. A., , Jr. (1981). Eur. J. Immunol. 11, 270.

154

ANNE O’GARRA AND DOUGLAS ROBINSON

Boussiotis, V. A., Tsai, E. Y., Yunis, E. J., Thim, S., Delgado, J. C., Dascher, C. C., Berezovskaya, A., Rousset, D., Reynes, J. M., and Goldfeld, A. E. (2000). IL-10-producing T cells suppress immune responses in anergic tuberculosis patients. J. Clin. Invest. 105, 1317–1325. Bozza, S., Gaziano, R., Spreca, A., Bacci, A., Montagnoli, C., di Francesco, P., and Romani, L. (2002). J. Immunol. 168, 1362. Buelens, C., Verhasselt, V., De Groote, D., Thielemans, K., Goldman, M., and Willems, F. (1997a). Eur. J. Immunol. 27, 1848. Buelens, C., Verhasselt, V., De Groote, D., Thielemans, K., Goldman, M., and Willems, F. (1997b). Eur. J. Immunol. 27, 756. Camoglio, L., Juffermans, N. P., Peppelenbosch, M., te Velde, A. A., ten Kate, F. J., van Deventer, S. J., and Kopf, M. (2002). Eur. J. Immunol. 32, 261. Campbell, P. M., Pimm, J., Ramassar, V., and Halloran, P. F. (1996). Transplantation 61, 933. Cao, Z., Henzel, W. J., and Gao, X. (1996). Science 271, 1128. Carter, L. L., and Murphy, K. M. (1999). J. Exp. Med. 189, 1355. Cella, M., Facchetti, F., Lanzavecchia, A., and Colonna, M. (2000). Nat. Immunol. 1, 305. Cella, M., Scheidegger, D., Palmer-Lehmann, K., Lane, P., Lanzavecchia, A., and Alber, G. (1996). J. Exp. Med. 184, 747. Chan, S. H., Perussia, B., Gupta, J. W., Kobayashi, M., Pospisil, M., Young, H. A., Wolf, S. F., Young, D., Clark, S. C., and Trinchieri, G. (1991). J. Exp. Med. 173, 869. Chen, C. H., Zhang, D. H., LaPorte, J. M., and Ray, A. (2000a). J. Immunol. 165, 5597. Chen, Q., Ghilardi, N., Wang, H., Baker, T., Xie, M-H., Gurney, A., Grewal, I. S., and de Sauvage, F. J. (2000b). Nature 407, 916. Cher, D. J., and Mosmann, T. R. (1987). J. Immunol. 138, 3688. Cherwinski, H. M., Schumacher, J. H., Brown, K. D., and Mosmann, T. R. (1987). J. Exp. Med. 166, 1229. Chu, C. Q., Wittmer, S., and Dalton, D. K. (2000). J. Exp. Med. 192, 123. Coffman, R. L., Varkila, K., Scott, P., and Chatelain, R. (1991). Imm. Rev. 123, 1. Constant, S., Pfeiffer, C., Woodard, A., Pasqualini, T., and Bottomly, K. (1995). J. Exp. Med. 182, 1591. Constant, S. L., and Bottomly, K. (1997). Ann. Rev. Immunol. 15, 297. Cooper, A. M., Kipnis, A., Turner, J., Magram, J., Ferrante, J., and Orme, I. M. (2002). J. Immunol. 168, 1322. Cua, D. J., Sherlock, J., Chen, Y., Murphy, C. A., Joyce, B., Seymour, B., Lucian, L., To, W., Kwan, S., Churakova, T., Zurawski, S., Wiekowski, M., Lira, S. A., Gorman, D., Kastelein, R. A., and Sedgwick, J. D. (2003). Nature 421, 744. D’Andrea, A., Rengaraju, M., Valiante, N. M., Chemini, J., Kubin, M., Aste, M., Chan, S. H., Kobayashi, M., Young, D., Nickbarg, E., Chizzonite, R., Wolf, S. F., and Trinchieri, G. (1992). J. Exp. Med. 176, 1387. D’Andrea, A. D., Aste-Amezaga, M., Valainte, N. M., Ma, X., Kubin, M., and Trinchieri, G. (1993). J. Exp. Med. 178, 1041. d’Ostiani, C. F., Del Sero, G., Bacci, A., Montagnoli, C., Spreca, A., Mencacci, A., RicciardiCastagnoli, P., and Romani, L. (2000). J. Exp. Med. 191, 1661. de Jong, E. C., Vieira, P. L., Kalinski, P., Schuitemaker, J. H., Tanaka, Y., Wierenga, E. A., Yazdanbakhsh, M., and Kapsenberg, M. L. (2002). J. Immunol. 168, 1704. de Jong, R., Altare, F., Haagen, I. A., Elferink, D. G., Boer, T., van Breda Vriesman, P. J., Kabel, P. J., Draaisma, J. M., van Dissel, J. T., Kroon, F. P., Casanova, J. L., and Ottenhoff, T. H. (1998). Science 280, 1435. de Waal Malefyt, R., Abrams, J., Bennett, B., Figdor, C., and Vries, J. E. d. (1991a). J. Exp. Med. 174, 1209.

FUNCTION OF T HELPER 1 CELLS

155

de Waal Malefyt, R., Haanen, J., Yssel, H., Roncarolo, M.-G., te Velde, A., Figdor, C., Johnson, K., Kastelein, R., Spits, H., and de Vries, J. E. (1991b). J. Exp. Med. 174, 915. Debets, R., Timans, J. C., Churakowa, T., Zurawski, S., de Waal Malefyt, R., Moore, K. W., Abrams, J. S., O’Garra, A., Bazan, J. F., and Kastelein, R. A. (2000). J. Immunol. 165, 4950. Deckert, M., Soltek, S., Geginat, G., Lutgen, S., Montecinos-Rongen, M., Hof, H., and Schluter, D. (2001). Infect. Immun. 69, 4561. Delgado, J. C., Tsai, E. Y., Thim, S., Baena, A., Boussiotis, V. A., Reynes, J. M., Sath, S., Grosjean, P., Yunis, E. J., and Goldfeld, A. E. (2002). Antigen-speciﬁc and persistent tuberculin anergy in a cohort of pulmonary tuberculosis patients from rural Cambodia. Proc. Natl. Acad. Sci. USA 99, 7576–7581. Devergne, O., Birkenbach, M., and Kieff, E. (1997). Proc. Natl. Acad. Sci. USA 94, 12041. Ding, L., Linsley, P. S., Huang, L. Y., Germain, R. N., and Shevach, E. M. (1993). J. Immunol. 151, 1224. Ding, L., and Shevach, E. M. (1992). J. Immunol. 148, 3133. Dofﬁnger, R., Altare, F., and Casanova, J. L. (2000a). Microbes Infect. 2, 1553. Dofﬁnger, R., Jouanguy, E., Dupuis, S., Fondaneche, M. C., Stephan, J. L., Emile, J. F., Lamhamedi-Cherradi, S., Altare, F., Pallier, A., Barcenas-Morales, G., Meinl, E., Krause, C., Pestka, S., Schreiber, R. D., Novelli, F., and Casanova, J. L. (2000b). J. Infect. Dis. 181, 379. Dong, C., Davis, R. J., and Flavell, R. A. (2002). Annu. Rev. Immunol. 20, 55. Dorman, S. E., and Holland, S. M. (1998). J. Clin. Invest. 101, 2364. Dupuis, S., Dargemont, C., Fieschi, C., Thomassin, N., Rosenzweig, S., Harris, J., Holland, S. M., Schreiber, R. D., and Casanova, J. L. (2001). Science 293, 300. Dupuis, S., Dofﬁnger, R., Picard, C., Fieschi, C., Altare, F., Jouanguy, E., Abel, L., and Casanova, J. L. (2000). Immunol. Rev. 178, 129. Dupuis, S., Dofﬁnger, R., Picard, C., Fieschi, C., Altare, F., Jouanguy, E., Abel, L., and Casanova, J. L. (2002). Mol. Immunol. 38, 903. Dybing, J. K., Walters, N., and Pascual, D. W. (1999). Infect. and Immunity 67, 6242. Edwards, A. D., Manickasingham, S. P., Sporri, R., Diebold, S. S., Schulz, O., Sher, A., Kaisho, T., Akira, S., and Reis e Sousa, C. (2002). J. Immunol. 169, 3652. Ehlers, S., Mielke, M. E. A., Blankenstein, T., and Hahn, H. (1992). J. Immunol. 149, 3016. Farrar, J. D., Smith, J. D., Murphy, T. L., Leung, S., Stark, G. R., and Murphy, K. M. (2000). Nat. Immunol. 1, 65. Ferber, I. A., Lee, H. J., Zonin, F., Heath, V., Mui, A., Arai, N., and O’Garra, A. (1999). Clin. Immunol. 91, 134. Finotto, S., Neurath, M. F., Glickman, J. N., Qin, S., Lehr, H. A., Green, F. H., Ackerman, K., Haley, K., Galle, P. R., Szabo, S. J., Drazen, J. M., De Sanctis, G. T., and Glimcher, L. H. (2002). Science 295, 336. Fiorentino, D. F., Bond, M. W., and Mosmann, T. R. (1989). J. Exp. Med. 170, 2081. Fiorentino, D. F., Zlotnik, A., Mosmann, T. R., Howard, M., and O’Garra, A. (1991a). J. Immunol. 147, 3815. Fiorentino, D. F., Zlotnik, A., Vieira, P., Mosmann, T. R., Howard, M., Moore, K. W., and O’Garra, A. (1991b). J. Immunol. 146, 3444. Fujioka, N., Akazawa, R., Ohashi, K., Fujii, M., Ikeda, M., and Kurimoto, M. (1999). J. Virol. 73, 2401. Gagliardi, M. C., Sallusto, F., Marinaro, M., Langenkamp, A., Lanzavecchia, A., and De Magistris, M. T. (2000). Eur. J. Immunol. 30, 2394. Gajewski, T. F., and Fitch, F. W. (1988). J. Immunol. 140, 4245. Gazzinelli, R. T., Hieny, S., Wynn, T. A., Wolf, S., and Sher, A. (1993). Proc. Natl. Acad. Sci. USA 90, 6115. Gazzinelli, R. T., Oswald, I. P., James, S. L., and Sher, A. (1992). J. Immunol. 148, 1792.

156

ANNE O’GARRA AND DOUGLAS ROBINSON

Geng, Y., Berencsi, K., Gyulai, Z., Valgi-nagy, T., Gonczol, T., and Trinchieri, G. (2000). Infect. Immun. 68, 2245. Gerber, J. S., and Mosser, D. M. (2001). Microbes. & Infect. 2, 131. Gerosa, F., Nisii, C., Righetti, S., Micciolo, R., Marchesini, M., Cazzadori, A., and Trinchieri, G. (1999). Clin. Immunol. 92, 224. Ghayur, T., Banerjee, S., Hugunin, M., Butler, D., Herzog, L., Carter, A., Quintal, L., Sekut, L., Talanian, R., Paskind, M., Wong, W., Kamen, R., Tracy, D., and Allen, H. (1997). Nature 386, 619. Gollob, J. A., Veenstra, K. G., Jyonouchi, H., Kelly, A. M., Ferrieri, P., Panka, D. J., Altare, F., Fieschi, C., Casanova, J. L., Frank, D. A., and Mier, J. W. (2000). J. Immunol. 165, 4120. Gracie, J. A., Forsey, R. J., Chan, A., Gilmour, B. P., Leung, M. R., Greer, K., Kennedy, R., Carter, W. Q., et al. (1999). J. Clin. Invest. 104, 1393. Greenfeder, S. A., Nunes, P., Kwee, L., Labow, M., Chizzonite, R. A., and Ju, G. (1995). J. Biol. Chem. 270, 13757. Grogan, J. L., Mohrs, M., Harmon, B., Lacy, D. A., Sedat, J. W., and Locksley, R. M. (2001). Immunity 14, 205. Gu, Y., Kuida, K., Tsutsui, H., Ku, G., Hsiao, K., Fleming, M. A., Hayashi, N., Higashino, K., Okamura, H., Nakanishi, K., Kurimoto, M., Ikeda, M., Flavell, R. A., Sato, V., Harding, M. W., Livingston, D. J., and Su, M. S. S. (1997). Science 275, 206. Haanen, J. B. G., Maleﬁjt, R. d. W., Res, P. C. M., Kraakman, E. M., Ottenhoff, T. H. M., Vries, R. R. P. d., and Spits, H. (1991). J. Exp. Med. 174, 583. Hardy, K. J., Manger, B., Newton, M., and Stobo, J. (1987). J. Immunol. 138, 1. Hardy, K. J., Matua Peterlin, B., Atchison, R. E., and Stobo, J. D. (1985). Proc. Natl. Acad. Sci. USA 82, 8173. Harty, J. T., and Bevan, M. J. (1995). Immunity 3, 109. Heath, V. L., Showe, L., Crain, C., Barrat, F. J., Trinchieri, G., and O’Garra, A. (2000). ‘‘Cutting edge’’ J. Immunol. 164, 2861. Heinzel, F. P., Sadick, M. D., Holaday, B. J., Coffman, R. L., and Locksley, R. M. (1989). J. Exp. Med. 169, 59. Heinzel, F. P., Schoenhaut, D. S., Rerko, R. M., Rosser, L. E., and Gately, M. K. (1993). J. Exp. Med. 177, 1505. Hilkens, C. M. U., Messer, G., Tesselaar, K., van Rietschoten, A. G. I., Kapsenberg, M. L., and Wieranga, E. A. (1996). J. Immunol. 157, 4316. Hochrein, H., O’Keeffe, M., Luft, T., Vandenabeele, S., Grumont, R. J., Maraskovsky, E., and Shortman, K. (2000). J. Exp. Med. 192, 823. Hochrein, H., Shortman, K., Vremec, D., Scott, B., Hertzog, P., and O’Keeffe, M. (2001). J. Immunol. 166, 5448. Hornung, V., Rothenfusser, S., Britsch, S., Krug, A., Jahrsdorfer, B., Giese, T., Endres, S., and Hartmann, G. (2002). J. Immunol. 168, 4531. Hoshino, K., Tsutsui, H., Kawai, T., Takeda, K., Nakanishi, K., Takeda, Y., and Akira, S. (1999). J. Immunol. 162, 5041. Hosken, N. A., Shibuya, K., Heath, A. W., Murphy, K. M., and O’Garra, A. (1995). J. Exp. Med. 182, 1579. Hsieh, C.-S., Heimberger, A. B., Gold, J. S., O’Garra, A., and Murphy, K. (1992). Proc. Natl. Acad. Sci. USA 89, 6065. Hsieh, C.-S., Macatonia, S. E., Tripp, C. S., Wolf, S. F., O’Garra, A., and Murphy, K. M. (1993a). Science 260, 547. Hsieh, C.-S., Macatonia, S. E., O’Garra, A., and Murphy, K. M. (1993b). Pathogen-induced Th1 phenotype development in CDþ alpha beta-TCR transgenic T cells is macrophage dependent. Int. Immunol. 5, 371–382.

FUNCTION OF T HELPER 1 CELLS

157

Huang, S., Hendriks, W., Althage, A., Hemmi, S., Bluethmann, H., Kanifo, R., Vulcek, J., Zinkernagel, R. M., and Aguet, M. (1993). Science 259, 1742. Hunter, C. A., Timans, J., Pisacane, P., Menon, S., Cai, G., Chizzonitte, R., Bazan, J. F., and Kastelein, R. A. (1997). Eur. J. Immunol. 27, 2787. Ito, T., Amakawa, R., Kaisho, T., Hemmi, H., Tajima, K., Uehira, K., Ozaki, Y., Tomizawa, H., Akira, S., and Fukuhara, S. (2002). J. Exp. Med. 195, 1507. Jacobson, N. G., Szabo, S. J., Weber-Nordt, R. M., Zhong, Z., Schreiber, R. D., Darnell, J. E. J., and Murphy, K. M. (1995). J. Exp. Med. 181, 1755. Jarrossay, D., Napolitani, G., Colonna, M., Sallusto, F., and Lanzavecchia, A. (2001). Eur. J. Immunol. 31, 3388. Kadowaki, N., Ho, S., Antonenko, S., Malefyt, R. W., Kastelein, R. A., Bazan, F., and Liu, Y. J. (2001). J. Exp. Med. 194, 863. Kalinski, P., Schuitemaker, J. H., Hilkens, C. M., and Kapsenberg, M. L. (1998). J. Immunol. 161, 2804. Kalinski, P., Schuitemaker, J. H., Hilkens, C. M., Wierenga, E. A., and Kapsenberg, M. L. (1999). J. Immunol. 162, 3231. Kane, M. M., and Mosser, D. M. (2001). J. Immunol. 166, 1141. Kaplan, M. H., Sun, Y.-L., Hoey, T., and Grusby, M. J. (1996). Nature 382, 174. Katsura, Y. (1977). Immunology 32, 227. Kobayashi, M., Fitz, L., Ryan, M., Hewick, R. M., Clark, S. C., Chan, S., Loudon, R., Sherman, F., Perussia, B., and Trinchieri, G. (1989). J. Exp. Med. 170, 827. Koch, F., Stanzl, U., Jennewein, P., Janke, K., Heuﬂer, C., Kampgen, E., Romani, N., and Schuler, G. (1996). J. Exp. Med. 184, 741. Kohno, K., Kataoka, J., Ohtsuki, T., Suemoto, Y., Okamoto, I., Usui, M., Ikeda, M., and Kurimoto, M. (1997). J. Immunol. 158, 1541. Langenkamp, A., Messi, M., Lanzavecchia, A., and Sallusto, F. (2000). Nat. Immunol. 1, 311. Lee, H. J., O’Garra, A., Arai, K., and Arai, N. (1998). J. Immunol. 160, 2343. Lee, H. J., Takemoto, N., Kurata, H., Kamogawa, Y., Miyatake, S., O’Garra, A., and Arai, N. (2000). J. Exp. Med. 192, 105. Leonard, J. P., Waldburger, K. E., and Goldman, S. J. (1995). J. Exp. Med. 181, 381. Lertmemongkolchai, G., Cai, G., Hunter, C. A., and Bancroft, G. J. (2001). J. Immunol. 166, 1097. Li, B., Yu, H., Zheng, W., Voll, R., Na, S., Roberts, A. W., Williams, D. A., Davis, R. J., Ghosh, S., and Flavell, R. A. (2000). Science 288, 2219. Li, C., Corraliza, I., and Langhorne, J. (1999). Infect. Immun. 67, 4435. Liblau, R., Singer, S., and McDevitt, H. (1995). Immunol. Today 16, 34. Lichtman, A. H., Chin, J., Schmidt, J. A., and Abbas, A. K. (1988). Proc. Natl. Acad. Sci. USA 85, 9699. Liew, F. Y. (2002). Nat. Rev. Immunol. 2, 55. Lighvani, A. A., Furucht, D. M., Jankovic, D., Yamane, H., Aliberti, J., Hissong, B. D., Nguyen, B. V., Gaddina, M., Sher, A., Paul, W. E., and O’Shea, J. J. (2001). Proc. Natl. Acad. Sci. USA 98, 15137. Locksley, R. M., Heinzel, F. P., Holaday, B. J., Muhta, S. S., Reiner, S. L., and Sadick, M. D. (1991). Forum. Immunol. 35, . Lu, B., Yu, H., Chow, C., Li, B., Zheng, W., Davis, R. J., and Flavell, R. A. (2001). Immunity 14, 583. Macatonia, S. E., Hosken, N. A., Litton, M., Vieira, P., Hsieh, C.-S., Culpepper, J., Wysocka, M., Trinchieri, G., Murphy, K. M., and O’Garra, A. (1995). J. Immunol. 154, 5071. Mackaness, G. B. (1962). J. Exp. Med. 116, 381. Maggi, E., Parronchi, P., Manetti, R., Simonelli, C., Piccinni, M.-P., Rugiu, F. S., De Carli, M., Ricci, M., and Romagnani, S. (1992). J. Immunol. 148, 2142.

158

ANNE O’GARRA AND DOUGLAS ROBINSON

Magram, J., Connaughton, S. E., Warrier, R. R., Carvajal, D. M., Wu, C., Ferrante, J., Stewart, C., Sarmiento, U., Faherty, D. A., and Gately, M. K. (1996). Immunity 4, 471. Maldonado-Lopez, R., Maliszewski, C., Urbain, J., and Moser, M. (2001). J. Immunol. 167, 4345. Malinin, N. L., Boldin, N. P., Kovalenko, A. V., and Wallach, D. (1997). Nature 385, 540. Maloy, K. J., and Powrie, F. (2001). Nature Immunol. 2, 816. Manetti, R., Parronchi, P., Guidizi, M. G., Piccinni, M.-P., Maggi, E., Trinchieri, G., and Romagnani, S. (1993). J. Exp. Med. 177, 1199. Marrack, P. C., and Kappler, J. W. (1975). J. Immunol. 114, 1116. Mastroeni, P., Clare, S., Khan, S., Harrison, J. A., Hormache, C. E., Okamura, H., Kurimoto, M., and Dougan, G. (1999). Infect. and Immunity 67, 478. Matthys, P., Vermeire, K., and Billiau, A. (2001). Trends Immunol. 22, 367. Mattner, F., Magram, J., Ferrante, J., Launois, P., Di Padova, K., Behin, R., Gately, M., Louis, J., and Alber, G. (1996). Eur. J. Immunol. 26, 1553. McGuirk, P., McCann, C., and Mills, K. H. (2002). J. Exp. Med. 195, 221. Medzhitov, R. (2001). Nat. Rev. Immunol. 1, 135. Micallef, M. J., Ohtsuki, T., Kohno, K., Tanabe, F., Ushio, S., Namba, M., Tanimoto, T., Torigo, K., Fujii, M., Ikeda, M., Fukuda, S., and Kurimoto, M. (1996). Eur. J. Immunol. 26, 1647. Monteiro, J. M., Harvey, C., and Trinchieri, G. (1998). J. Virol. 72, 4825. Moore, K. W., de Waal Malefyt, R., Coffman, R. L., and O’Garra, A. (2001). Annu. Rev. Immunol. 19, 683. Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L. (1986). J. Immunol. 136, 2348. Mosmann, T. R., and Coffman, R. L. (1989). Ann. Rev. Immunol. 7, 145. Mullen, A. C., High, F. A., Hutchins, A. S., Lee, H. W., Villarino, A. V., Livingstone, D. M., Kung, A. L., Cereb, N., Yao, T. P., Yang, S. Y., and Reiner, S. L. (2001). Science 292, 1907. Mullen, A. C., Hutchins, A. S., High, F. A., Lee, H. W., Sykes, K. J., Chodosh, L. A., and Reiner, S. L. (2002). Nature Immunol. 3, 652. Murphy, E. E., Terres, G., Macatonia, S. E., Hsieh, C.-S., Mattson, J., Lanier, L., Wysocka, M., Trinchieri, G., Murphy, K., and O’Garra, A. (1994). J. Exp. Med. 180, 223. Murphy, K. M., Ouyang, W., Farrar, J. D., Yang, J., Ranganath, S., Asnagli, H., Afkarian, M., and Murphy, T. L. (2000). Annu. Rev. Immunol. 18, 451. Murphy, K. M., Ouyang, S. J., Szabo, S. J., Jacobsen, N. G., Guler, M. L., Gorham, J. D., Gubler, U., and Murphy, T. L. (1999). Curr. Top. Microbiol. Immunol. 238, 13. Nakane, A., Minagawa, T., Kohanawa, M., Chen, Y., Sato, H., Moriwama, M., and Tsuruoka, N. (1989). Infect. & Immunity 57, 3331. Neighbors, M., Barrat, F., Xu, X., Debets, R., Churakova, T., Abrams, J., Bazan, J. F., Kastelein, R., and O’Garra, A. (2000). J. Exp. Med. 194, 343. Newport, M. J., Huxley, C. M., Huston, S., Hawrylowicz, C. M., Oostra, B. A., Williamson, R., and Levin, M. (1996). N. Engl. J. Med. 335, 1941. Nguyen, K. B., Cousens, L. P., Doughty, L. A., Pien, G. C., Durbin, J. E., and Biron, C. A. (2000). Nature Immunol. 1, 70. Nguyen, K. B., Watford, W. T., Salomon, R., Hofmann, S. R., Pien, G. C., Morinobu, A., Gadina, M., O’Shea, J. J., and Biron, C. A. (2002). Science 297, 2063. Nishikomori, R., Ehrhardt, R. O., and Strober, W. (2000). J. Exp. Med. 191, 847. O’Garra, A. (1998). Immunity 8, 275. O’Garra, A., Hosken, N., Macatonia, S., Wenner, C. A., and Murphy, K. (1995). In ‘‘Res. Immunol.’’ (P. Truffa-Bachi, Ed.). Institut. Pasteur Paris, France. O’Garra, A., and Murphy, K. (1993). Curr. Opin. Immunol. 5, 880. O’Garra, A., and Murphy, K. (1994). Curr. Opin. Immunol. 6, 458. O’Neill, L. A., and Greene, C. (1998). J. Leukoc. Biol. 63, 650.

FUNCTION OF T HELPER 1 CELLS

159

Okamura, H., Tsutsul, H., Komatsu, T., Yutsudo, M., Hakura, A., Tanimoto, T., Torigoe, K., Okura, T., Mukuda, Y., Hattori, K., Akita, K., Namba, M., Tanabe, K., Konishi, K., Fukuda, S., and Kurimoto, M. (1995). Nature 378, 88. Oppmann, B., Lesley, R., Blom, B., Timans, J. C., Xu, Y., Hunte, B., Vega, F., Yu, N., Wang, J., Singh, K., Zonin, F., Vaisberg, E., Churakova, T., Liu, M., Gorman, D., Wagner, J., Zurawski, S., Liu, Y., Abrams, J. S., Moore, K. W., Rennick, D., de Waal-Malefyt, R., Hannum, C., Bazan, J. F., and Kastelein, R. A. (2000). Immunity 13, 715. Orange, J. S., Wang, B., Terhorst, C., and Biron, C. A. (1995). J. Exp. Med. 182, 1045. Ouyang, W., Lohning, M., Gao, Z., Assenmacher, M., Ranganath, S., Radbruch, A., and Murphy, K. M. (2000). Immunity 12, 27. Ouyang, W., Ranganath, S. H., Weindel, K., Bhattacharya, D., Murphy, T. L., Sha, W. C., and Murphy, K. M. (1998). Immunity 9, 745. Oxenius, O., Karrer, U., Zinkernagel, R. M., and Hengartner, H. (1999). J. Immunol. 162, 965. Parham, C., Chirica, M., Timans, J., Vaisberg, E., Travis, M., Cheung, J., Pﬂanz, S., Zhang, R., Singh, K. P., Vega, F., To, W., Wagner, J., O’Farrell, A. M., McClanahan, T., Zurawski, S., Hannum, C., Gorman, D., Rennick, D. M., Kastelein, R. A., de Waal Malefyt, R., and Moore, K. W. (2002). J. Immunol. 168, 5699. Parish, C. R. (1972). Transplant. Rev. 13, 35. Park, A. Y., Hondowicz, B. D., and Scott, P. (2000). J. Immunol. 165, 896. Parnet, P., Garka, K. E., Bonnert, T. P., Dower, S. K., and Sims, J. E. (1996). J. Biol. Chem. 271, 3967. Parronchi, P., De Carli, M., Manetti, R., Simonelli, C., Sampognaro, S., Piccinni, M.-P., Macchia, D., Maggi, E., Del Prete, G., and Romagnani, S. (1992). J. Immunol. 149, 2977. Penix, L., Weaver, W. M., Pang, Y., Young, H. A., and Wilson, C. B. (1993). J. Exp. Med. 178, 1483. Penix, L. A., Sweetser, M. T., Weaver, W. M., Hoefﬂer, J. P., Kerppola, T. K., and Wilson, C. B. (1996). J. Biol. Chem. 271, 31964. Pfeffer, K., Matsuyama, T., Kundig, T. M., Wakeham, A., Kishihara, K., Shahinian, A., Wiegmann, K., Ohashi, P. S., Kronke, M., and Mak, T. W. (1993). Cell 73, 457. Pﬂanz, S., Timans, J. C., Cheung, J., Rosales, R., Kanzler, H., Gilbert, J., Hibbert, L., Churakova, T., Travis, M., Vaisberg, E., Blumenschein, W. M., Mattson, J. D., Wagner, J. L., To, W., Zurawski, S., McClanahan, T. K., Gorman, D. M., Bazan, J. F., de Waal Malefyt, R., Rennick, D., and Kastelein, R. (2002). Immunity 16, 779. Pizarro, T. T., Michie, M. H., Bentz, M., Woraratanadharn, J., Smith, M. F., Foley, E. J., Moskaluk, C. A., Bickston, S. J., and Cominelli, F. (1999). J. Immunol. 162, 6829. Powrie, F., and Coffman, R. L. (1993). Immunol. Today 14, 270. Powrie, F., Leach, M. W., Mauze, S., Menon, S., Caddle, L. B., and Coffman, R. L. (1994). Immunity 1, 553. Pravica, V., Brogan, I. J., and Hutchinson, I. V. (2000). Eur. J. Immunogenet. 27, 35. Presky, D. H., Yang, H., Minetti, L. J., Chua, A. O., Nabavi, N., Wu, C. Y., Gately, M. K., and Gubler, U. (1996). Proc. Natl. Acad. Sci. USA 93, 14002. Ranganath, S., Ouyang, W., Bhattarcharya, D., Sha, W. C., Grupe, A., Peltz, G., and Murphy, K. M. (1998). J. Immunol. 161, 3822. Regnier, C. H., Yeong Song, H., Gao, X., Goeddel, D. V., Cao, Z., and Rothe, M. (1997). Cell 90, 373. Reis e Sousa, C., Sher, A., and Kaye, P. (1999). Curr. Opin. Immunol. 11, 392. Reis e Sousa, C., Yap, G., Schulz, O., Rogers, N., Schito, M., Aliberti, J., Hieny, S., and Sher, A. (1999). Immunity 11, 637. Rincon, M., and Flavell, R. A. (1997). Current Biology 7, R729. Rincon, M., Enslen, H., Raingeaud, J., Recht, M., Zapton, T., Su, M. S., Penix, L. A., Davis, R. J., and Flavell, R. A. (1998). EMBO J. 17, 2817.

160

ANNE O’GARRA AND DOUGLAS ROBINSON

Rissoan, M.-C., Soumelis, V., Kadowaki, N., Grouard, G., Briere, F., de Waal Malefyt, R., and Liu, Y.-J. (1999). Science 283, 1183. Robinson, D., Shibuya, K., Mui, A., Zonin, F., Murphy, E., Sana, T., Hartley, S. B., Menon, S., Kastelein, R., Bazan, F., and O’Garra, A. (1997). Immunity 7, 571. Rock, F. L., Hardiman, G., Timans, J. C., Kastelein, R. A., and Bazan, J. F. (1998). Proc. Natl. Acad. Sci. USA 95, 588. Rogge, L., Barberis-Maino, L., Bifﬁ, M., Passini, N., Presky, D. H., Gubler, U., and Sinigaglia, F. (1997). J. Exp. Med. 185, 825. Rogge, L., D’Ambrosio, D., Bifﬁ, M., Penna, G., Minetti, L. J., Presky, D. H., Adorini, L., and Sinigaglia, F. (1998). J. Immunol. 161, 6567. Rogge, L., Papi, A., Presky, D. H., Bifﬁ, M., Minetti, L. J., Miotto, D., Agostini, C., Semenzato, G., Fabbri, L. M., and Sinigaglia, F. (1999). J. Immunol. 162, 3926. Romagnani, S. (1991). Immunol. Today 12, 256. Romagnani, S. (1994). Ann. Rev. Immunol. 12, 227. Rothe, H., Jenkins, N., Copeland, N., and Kolb, H. (1997). J. Clin. Invest. 99, 469. Rothe, J., Lesslauer, W., Lotscher, H., Lang, Y., Koebel, P., Kontgen, F., Althage, A., Zinkernagel, R., Steinmetz, M., and Bluethmann, H. (1993). Nature 364, 798. Ruedl, C., Bachmann, M. F., and Kopf, M. (2000). Eur. J. Immunol. 30, 2056. Sacks, D. L., and Noben-Trauth, N. (2002). Nature Rev. Immunol. 2, 845. Sadick, M. D., Heinzel, F. P., Holaday, B. J., Pu, R. T., Dawkins, R. S., and Locksley, R. M. (1990). J. Exp. Med. 171, 115. Sakaguchi, S. (2000). Cell 101, 455. Samson, J. N., Langermans, J. A. M., Savelkoul, H. F. J., and Furth, S. V. (1995). Immunology 86, 256. Sansonetti, P. J., Phalipon, A., Arondei, J., Thirumalai, K., Banerjee, S., Akira, S., Takeda, K., and Zychlinsky, A. (2000). Immunity 12, 581. Schafer, P. H., Wadsworth, S. A., Wang, L., and Siekierka, J. J. (1999). J. Immunol. 162, 7110. Schulz, O., Edwards, A. D., Schito, M., Aliberti, J., Manickasingham, S., Sher, A., and Reis e Sousa, C. (2000). Immunity 13, 453. Scott, P., and Kaufmann, S. H. E. (1991). Immunol. Today 12, 346. Scott, P., Natovitz, P., Coffman, R. L., Pearce, E., and Sher, A. (1988). J. Exp. Med. 168, 1675. Scott, P., Pearce, E., Cheever, A. W., Coffman, R. L., and Sher, A. (1989). Immunol. Rev. 112, 161. Seder, R. A., Gazzinelli, R., Sher, A., and Paul, W. E. (1993). Proc. Natl. Acad. Sci. USA 90, 10188. Segal, B. M., Dwyer, B. K., and Shevach, E. M. (1998). J. Exp. Med. 187, 537. Sher, A., and Coffman, R. L. (1992). Ann. Rev. Immunol. 10, 385. Shevach, E. M. (2000). Annu. Rev. Immunol. 18, 423. Shi, F., Takeda, K., Akira, S., Sarvetnik, N., and Lundgren, H. (2000). J. Immunol. 165, 3099. Shibuya, K., Robinson, D., Zonin, F., Hartley, S. B., Macatonia, S. E., Somoza, C., Hunter, C. A., Murphy, K. M., and O’Garra, A. (1998). J. Immunol. 160, 1708. Sica, A., Dorman, L., Viggiano, V., Cippitelli, M., Ghosh, P., Rice, and Young, H. A. (1997). J. Biol. Chem. 272, 30412. Silva, R. A., and Appelberg, R. (2001). Antimicrob. Agents Chemother 45, 1312. Silva, R. A., Pais, T. F., and Appelberg, R. (2001). J. Immunol. 167, 1535. Sims, J. E., March, C. J., Cosman, D., Widmer, M. B., MacDonald, H. R., McMahan, C. J., Grubin, C. E., Wignall, J. M., Jackson, J. L., and Call, S. M. (1988). Science 241, 585. Smeltz, R. B., Chen, J., Hu-Li, J., and Shevach, E. M. (2001). J. Exp. Med. 194, 143. Snapper, C. M., and Paul, W. E. (1987). Science 236, 944. Stamm, L. M., Satoskar, A. A., Ghosh, S. K., David, J. R., and Satoskar, A. R. (1999). Eur. J. Immunol. 29, 2524. Stout, R. D., and Bottomly, K. (1989). J. Immunol. 142, 760.

FUNCTION OF T HELPER 1 CELLS

161

Swain, S. L., Weinberg, A. D., and English, M. (1990). J. Immunol. 144, 1788. Sweetser, M. T., Hoey, T., Sun, Y. L., Weaver, W. M., Price, G. A., and Wilson, C. B. (1998). J. Biol. Chem. 273, 34775. Sypek, J. P., Chung, C. L., Mayor, S. E. H., Subramanyam, J. M., Goldman, S. J., Sieburth, D. S., Wolf, S. E., and Schaub, R. G. (1993). J. Exp. Med. 177, 1797. Szabo, S., Dighe, A. S., Gubler, U., and Murphy, K. M. (1997). J. Exp. Med. 185, 817. Szabo, S. J., Jacobson, N. G., Dighe, A. S., Gubler, U., and Murphy, K. M. (1995). Immunity 2, 665. Szabo, S. J., Kim, S. T., Costa, G. L., Zhang, X., Fathman, C. G., and Glimcher, L. H. (2000). Cell 100, 655. Tada, T., Takemori, T., Okumura, K., Nonaka, M., and Tokuhisa, T. (1978). J. Exp. Med. 147, 446. Takeda, K., Tsutsui, H., Yoshimoto, T., Adachi, O., Yoshida, N., Kishimoto, T., Okamura, H., Nakanishi, K., and Akira, S. (1998). Immunity 8, 383. Tanaka-Kataoka, M., Kunikata, T., Takayama, S., Iwaki, K., Ohashi, K., Ikeda, M., and Kurimoto, M. (1998). Cytokine 11, 593. Thierfelder, W. E., van Deursen, J. M., Yamamoto, K., Tripp, R. A., Sarawar, S. R., Carson, R. T., Sangster, M. Y., Vignali, D. A. A., Doherty, P. C., Grosveld, G. C., and Ihle, J. N. (1996). Nature 382, 171. Torigoe, K., Ushio, S., Okura, T., Kobayashi, S., Taniai, M., Kunikata, T., Murakami, T., Sanou, O., Kojima, H., Fujii, M., Ohta, T., Ikeda, M., Ikegami, H., and Kurimoto, M. (1997). J. Biol. Chem. 272, 25737. Trembleau, S., Penna, G., Bosi, E., Mortara, A., Gately, M. K., and Adorini, L. (1995). J. Exp. Med. 181, 817. Trinchieri, G. (1995). Ann. Rev. Immunol. 13, 251. Trinchieri, G. (2001). J. Exp. Med. 194, F53. Trinchieri, G. (2003). Nature Rev. Immunol. 3, 133. Tripp, C. S., Beckerman, K. P., and Unanue, E. R. (1995a). J. Clin. Invest. 95, 1628. Tripp, C. S., Gately, M. K., Hakimi, J., Ling, P., and Unanue, E. R. (1994). J. Immunol. 152, 1883. Tripp, C. S., Kamogawa, O., and Unanue, E. R. (1995b). J. Immunol. 155, 3427. Tripp, C. S., Wolf, S. E., and Unanue, E. R. (1993). Proc. Natl. Acad. USA 90, 3725. Unanue, E. (1997). Immunol. Rev. 158, 11. Ushio, S., Namba, M., Okura, T., Hattori, K., Nukada, Y., Akita, K., Tanabe, F., Konishi, K., Micallef, M., Fujii, M., Torigoe, K., Tanimoto, T., Fukuda, S., Ikeda, M., Okamura, H., and Kurimoto, M. (1996). J. Immunol. 156, 4274. Verhagen, C. E., de Boer, T., Smits, H. H., Verreck, F. A., Wierenga, E. A., Kurimoto, M., Lammas, D. A., Kumararatne, D. S., Sanal, O., Kroon, F. P., van Dissel, J. T., Sinigaglia, F., and Ottenhott, T. H. (2000). J. Exp. Med. 192, 517. Vieira, P. L., de Jong, E. C., Wierenga, E. A., Kapsenberg, M. L., and Kalinski, P. (2000). J. Immunol. 164, 4507. Wagner, H. (2001). Immunity 14, 499. Watford, W. T., and O’Shea, J. J. (2003). Nature 421, 706. Wei, X., Leung, B. P., Niedbala, W., Piedraﬁta, D., Feng, G., Sweet, M., Dobbie, L., Smith, A. J. H., and Liew, F. Y. (1999). J. Immunol. 163, 2821. Wenner, C. A., Macatonia, S. A., O’Garra, A., and Murphy, K. M. (1995). J. Immunol. 156, 1442. Whelan, M., Harnett, M. M., Houston, K. M., Patel, V., Harnett, W., and Rigley, K. P. (2000). J. Immunol. 164, 6453. Wherry, J. C., Schreiber, R. D., and Unanue, E. R. (1991). Infect. Immun. 59, 1709. Wiekowski, M. T., Leach, M. W., Evans, E. W., Sullivan, L., Chen, S. C., Vassileva, G., Bazan, J. F., Gorman, D. M., Kastelein, R. A., Narula, S., and Lira, S. A. (2001). J. Immunol. 166, 7563.

162

ANNE O’GARRA AND DOUGLAS ROBINSON

Wierenga, E. A., Snoek, M., de Groot, C., Chretien, I., Bos, J. D., Jansen, H. K., and Kapsenberg, M. L. (1990). J. Immunol. 144, 4651. Wildbaum, G., Youssef, S., Grabie, N., and Karin, N. (1998). J. Immunol. 161, 6368. Willenborg, D. O., Fordham, S., Bernard, C. C., Cowden, W. B., and Ramshaw, I. A. (1996). J. Immunol. 157, 3223. Wu, C., Ferrante, J., Gately, M. K., and Magram, J. (1997). J. Immunol. 159, 1658. Xu, D., Chan, W. L., Leung, B. P., Hunter, D., Schulz, K., Carter, R. W., McInnes, I. B., Robinson, J. H., and Liew, F. Y. (1998). J. Exp. Med. 188, 1485. Yang, D. D., Conze, D., Whitmarsh, A., Barrett, T., Davis, R. J., Rincon, M., and Flavell, R. A. (1998). Immunity 9, 575. Yang, J., Murphy, T. L., Ouyang, W., and Murphy, K. M. (1999). Eur. J. Immunol. 29, 548. Yang, J., Zhu, H., Murphy, T. L., Ouyang, W., and Murphy, K. M. (2001). Nature Immunol. 2, 157. Yap, G., Pesin, M., and Sher, A. (2000). J. Immunol. 165, 628. Ye, J., Cippitelli, M., Dorman, L., Ortaldo, J. R., and Young, H. A. (1996). Mol. Cell. Biol. 16, 4744. Yoshida, H., Hamano, S., Senaldi, G., Covey, T., Faggioni, R., Mu, S., Xia, M., Wakeham, A. C., Nishina, H., Potter, J., Saris, C. J. M., and Mak, T. W. (2001). Immunity 15, 569. Yoshimoto, T., Takeda, K., Tanaka, T., Ohkusu, K., Kashiwamura, S., Okamura, H., Akira, S., and Nakanishi, K. (1998). J. Immunol. 161, 3400. Zhang, D. H., Cohn, L., Ray, P., Bottomly, K., and Ray, A. (1997). J. Biol. Chem. 272, 21597. Zheng, W.-P., and Flavell, R. A. (1997). Cell 89, 587.

advances in immunology, vol. 83

Th2 Cells: Orchestrating Barrier Immunity DANIEL B. STETSON, DAVID VOEHRINGER, JANE L. GROGAN, MIN XU, R. LEE REINHARDT, STEFANIE SCHEU, BEN L. KELLY, AND RICHARD M. LOCKSLEY Howard Hughes Medical Institute University of California San Francisco San Francisco, California 94143

I. Introduction: History and Definitions

In the mid- to late-1980s, Mosmann and Coffman described the division of antigen-speciﬁc murine CD4 T cell lines and clones into stable subsets based on the patterns of cytokines they produced (Mosmann and Coffman, 1989). Although the cells shared an identical surface phenotype, Th1 cells transcribed and secreted interferon-g (IFNg), IL-2, and lymphotoxin after activation, whereas Th2 cells transcribed IL-4, IL-5, and IL-13 (called p600 at that time). Voluminous contributions over the past 15 years have established that the basic underlying principles extend to both mouse and human CD4 T cells. A fundamental concept, based largely on differentiation of naive, TCR transgenic, T cells in vitro, is the capacity to differentiate a given T cell to either a Th1 or Th2 cell, depending on the stimulation milieu. Although a variety of manipulations, generally characterized as ‘‘strength of signal,’’ inﬂuence the ease with which Th1 or Th2 fates can be generated, key cytokines and cytokine receptor pathways remain the major determinants of cell fate. Th1 and Th2 cells are commonly referred to in association with ‘‘cellular’’ and ‘‘humoral’’ immunity, respectively. The role for IFNg, the cardinal cytokine of Th1 biology, in isotype switching for opsonizing antibodies, and the activation of many cell types by IL-4, the cardinal cytokine of Th2 biology, renders these appellations imprecise. Here, we refer to type 1 immunity as an immune response centrally orchestrated by Th1 cells that stably secrete IFNg. In general, the coordinated activation of phagocytes, production of opsonizing antibodies, and induction of cytolytic T cells by Th1 cells collectively describes the response to systemic invasion. In contrast, we refer to type 2 immunity as an immune response centrally orchestrated by Th2 cells that stably secrete IL-4. As such, the coordinated concentration of activated eosinophils, mast cells, and basophils by Th2 cells collectively describes the response to barrier invasion. This chapter focuses on recent information relative to the generation and activation of IL-4-expressing Th2 cells, and, where appropriate, on comparisons with other IL-4-expressing cells in the coordination of type 2 immunity. 163 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

164

DANIEL B. STETSON ET AL.

II. Activation of IL-4 Expression in Naive CD4 T Cells

A substantial amount of information has accumulated addressing the mechanisms by which naive helper T cells activate IL-4 expression (Murphy and Reiner, 2002). The emerging model is one of lineage differentiation, by which Th2 cells establish a distinct gene expression pattern linked intimately with epigenetically stable alterations in DNA and chromatin (Smale and Fisher, 2002). With the exception of NK T cells, which will be considered later, there is little evidence for activation of IL-4 gene expression during T cell development in the thymus. Activation via the TCR is integrated with co-stimulatory and cytokine signals delivered from dendritic cells and the environmental milieu to establish IL-4 transcription in naive CD4 T cells. Assays correlated with general transcriptional competence suggest that basal modiﬁcations of chromatin impede gene expression in naive T cells, as demonstrated by the absence of DNase I hypersensitivity sites around the IL-4/IL-13 genes (Agarwal and Rao, 1998; Santangelo et al., 2002; Takemoto et al., 1998), hypoacetylation of histones (Avni et al., 2002; Bird et al., 1998; Fields et al., 2002; Valapour et al., 2002), and methylation at CpG sites across the locus (Agarwal and Rao, 1998; Bird et al., 1998; Lee et al., 2002; Santangelo et al., 2002). Genetic evidence suggests that repression is maintained by MeCP1 complexes, including MBD2 and NURD proteins, tightly associated with the locus (Hutchins et al., 2002). Activated T cells from mice deﬁcient in MBD2, a methyl-CpG-binding protein, or in Dnmt1, a maintenance methyltransferase, ectopically express cytokines, including IL-4, consistent with a requirement for basal chromatin-mediated barriers to transcription (Hutchins et al., 2002; Lee et al., 2001b). Differential display was used to identify GATA-3, a zinc-ﬁnger transcription factor, as a highly polarized transcript expressed in Th2, but not Th1, cells (Zhang et al., 1997; Zheng and Flavell, 1997). GATA-3 is essential for normal hematopoiesis and nervous system development, but also for T cell development, and will require a conditionally mutant allele for deﬁnitive genetic analysis of its role in Th2 differentiation. Nevertheless, overexpression, silencing, and dominant negative strategies have together established a key role for GATA-3 in Th2 differentiation. GATA-3 lies genetically upstream of the appearance of chromatin modiﬁcations identiﬁed by DNase I hypersensitivity analysis (Ouyang et al., 2000). A related transcription factor, GATA-4, has the ability to decompact condensed chromatin, a key early step in the induction of albumin expression by fetal liver, using similar zinc ﬁnger motifs that have been implicated in GATA-3 chromatin remodeling (Cirillo et al., 2002; Takemoto et al., 2002). GATA-3 binding sites extend throughout the type 2 cytokine locus encompassing IL-4, IL-13, and IL-5, and, importantly, are present within sequences corresponding to the DNase hypersensitivity regions

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

165

that develop upon Th2 polarization. Assembly of these various sequences with the IL-4 gene as a mini-locus transgene was used to demonstrate the ability of GATA-3 to enhance markedly IL-4 expression within a chromatin context that included these presumptive regulatory regions (Lee et al., 2001a). The emerging model suggests a role for GATA-3 as a ‘‘pioneer’’ factor, likely within the context of as yet unknown additional proteins, to integrate signals emanating from the T cell activation synapse and to engage and decondense chromatin at the type 2 cytokine locus. Possibly, GATA-3 initiates decompaction by displacing repressive MeCP1 complexes, thus initiating a process by which strand methylation could be rapidly diluted during processive cell division, allowing establishment of a permissive histone code. Although deﬁnitive data is lacking, several pieces of evidence favor such a model. First, GATA3 protein is present in naive CD4 T cells, and thus available for targeting immediately after T cell activation. Second, the IL-4 genes are positioned in areas of euchromatin in naive T cells, apart from highly compacted heterochromatin (Grogan et al., 2001). Third, chromatin immunoprecipitations reveal small amounts of GATA-3 bound at IL-4 regulatory elements in resting, naive T cells (Avni et al., 2002). This observation, together with the ﬁnding that IL-4 is transcribed within 30 minutes of T cell activation in a Stat6-independent manner (Grogan et al., 2001), raises the possibility that the IL-4 gene is ‘‘poised’’ for rapid transcription by intermittent stochastic interactions of GATA-3 with the locus. Spontaneous ﬂuidity of nucleosome structure between relatively condensed and permissive states (Narlikar et al., 2001) might reconcile the apparent contradictions inherent between a chromatin-inaccessible state and the ease with which transcription can be initiated at the gene. The frequent co-expression of IL-4, IL-13, and IL-5 from single alleles in Th2 cells (Kelly and Locksley, 2000), together with their genomic co-localization, suggested that conserved cis-acting elements likely contributed to transcription. Indeed, human transgenes containing IL-4, IL-13, and IL-5 were expressed in an appropriate cell- and stimulation-dependent context in mice (Lacy et al., 2000). Comparative sequence analysis across a number of species revealed intergenic sequences, designated CNS-1 and CNS-2 (conserved noncoding sequences), localized to the 50 IL-4/IL-13 intergenic and the 30 IL-4 regions, respectively (Loots et al., 2000). Deletions of either CNS-1 or CNS-2 attenuated IL-4 expression (Mohrs et al., 2001a; Solymar et al., 2002). Intriguingly, these areas map closely with DNase I hypersensitivity sites that appear early after Th2 polarization of naive T cells (Fig. 1). GATA-3 binding sites are located within these regions, supporting the concept of early targeting of highly conserved cis-regulatory elements on either side of the IL-4 gene that serve to recruit chromatin-modifying enzymes, thus conﬁguring the locus in a manner competent for interactions with additional DNA-modifying protein complexes. More extensive analysis deﬁned additional conserved regulatory

166

DANIEL B. STETSON ET AL.

Fig 1 Regulatory elements surrounding the IL4 gene. DNAse I hypersensitivity sites (HSS, arrows) correlate closely with regions known to regulate IL-4 expression. Enhancer elements are indicated in green, and a silencer element is depicted in red. IL-4 exons are numbered and depicted as black boxes. The relative positions of each element are not drawn to scale. (See Color Insert.)

elements within the promoter and the second intronic regions, and yet more distant elements likely contribute to optimal IL-4 expression. Transgenes containing each of these regulatory elements together conferred cell- and condition-speciﬁcity upon IL-4 expression in lymphocytes (Lee et al., 2001a) (Fig. 1). Although GATA-3 ‘‘brackets’’ the IL-4 gene and does not directly transactivate the IL-4 promoter, GATA-3 binds and transactivates the IL-5 and IL-13 proximal promoters directly (Kishikawa et al., 2001; Yamashita et al., 2002; Zhang et al., 1997). Both IL-4 and IL-13 bind to the widely distributed IL4Ra/IL-13Ra1 receptor, but only IL-4 binds to a lymphoid-restricted receptor, IL-4Ra/gc, leading to speculation that evolution has scaffolded lymphocytespeciﬁc regulation onto an IL-4 gene duplicated from related genes more directly activated by GATA-3 in nonlymphoid cells. Ultimate activation of IL-4 expression takes place through contributions from ubiquitous and induced transcription factors that converge upon the GATA-3-modiﬁed locus. Chromatin-modifying complexes required to alter histones and reposition nucleosomes move rapidly to the nucleus after TCR stimulation (Zhao et al., 1998). In addition to being present as protein, GATA-3 transcription accompanies activation, is further augmented by Stat6-mediated signals, and is attenuated by concomitant Stat1- and/or Stat4-mediated signals (Ouyang et al., 2000), which would be generated during induction of type 1 immunity. Both Stat6 and GATA-3, the latter in a Stat6-independent way, promote transcription of c-Maf, a b-ZIP transcription factor expressed in Th2 cells. c-Maf binds and transactivates the IL-4 promoter, and c-Maf-deﬁcient T cells are highly attenuated for IL-4 expression (Kim et al., 1999). In addition to c-Maf, efﬁcient IL-4 expression requires ubiquitous members of the NF-AT and AP-1 families, most notably, NF-ATc1 and JunB, respectively, although complex interplay between activating and repressing members of these transcription families adds regulatory detail (Hartenstein et al., 2002; Ranger et al., 1998; Rengarajan et al., 2002b; Yoshida et al., 1998). Additional

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

167

proteins, including NIP45, IRF4, and TRAF2, have been implicated in modulating IL-4 expression in T cells through various protein–protein interactions (Hodge et al., 1996; Hu et al., 2002; Lieberson et al., 2001; Rengarajan et al., 2002a). In addition to c-Maf, a second critical target of GATA-3 is GATA-3 itself. Auto-activation of the gene by its protein product, together with its transcriptional repression by Stat1/Stat4-mediated signaling, is likely a fundamental process underpinning the polarization of IL-4 expression in differentiating T cell subsets (Ouyang et al., 2000). Additionally, GATA-3 represses IL-12Rb2 expression. Despite its central role in the process of Th2 differentiation, the transcription of GATA-3 itself remains incompletely understood, but the discovery of alternative promoters and silencers suggests careful regulation (Asnagli et al., 2002; Gregoire and Romeo, 1999; Hwang et al., 2002). Activation of the gene involves interplay between NF-AT and NF-kB transcription factors driven to the nucleus in response to TCR stimulation. Transcription is greatly augmented by Stat6 and GATA-3. However, sufﬁciently prolonged signaling is likely to account for observations of IL-4/IL-4R-/Stat6-independent Th2 cell differentiation, since once a threshold has been reached, GATA-3 is alone sufﬁcient to program downstream commitment of the locus independent of these signals (Finkelman et al., 2000; Jankovic et al., 2000; Mohrs et al., 2001b). Deletion of mel-18, a polycomb group gene, resulted in failure to sustain GATA-3 induction with subsequent inability to develop Th2 cells (Kimura et al., 2001). Polycomb genes are typically associated with silencing, raising the possibility that mel-18 restricts expression of inhibitory GATA-3interacting proteins, such as FOG-1 and ROG, which impede GATA-3mediated transcription (Fox et al., 1999), or SOCS5, a Th1-associated negative regulator of IL-4Ra/Stat6 signaling (Seki et al., 2002). Further work is needed. III. Stabilization of IL-4 Expression in T Cells

Activation of IL-4 transcription, initially a Stat6-independent process in naive CD4 T cells, is not sufﬁcient to stabilize the locus for committed expression by daughter cells (Grogan et al., 2001). Persistent signals are required, presumably to achieve threshold levels of GATA-3 required to drive the process in a cell autonomous manner. Although sustained TCR and co-stimulatory signals can be sufﬁcient for Th2 development, this threshold is reached much more readily by the addition of IL-4Ra/Stat6-mediated signals (Ouyang et al., 2000). Stat6 mediates induction of growth factor independent-1 (Gﬁ-1), a transcriptional repressor, which dramatically augments proliferation and survival of cells expressing GATA-3 (Zhu et al., 2002). Requirements for potent, sustained, signaling and/or exogenous IL-4 create thresholds for Th2 development. Using a sensitive knockin GFP reporter,

168

DANIEL B. STETSON ET AL.

analysis after TCR activation suggested that IL-4 transcription is initiated from both alleles in CD4 T cells (Mohrs et al., 2001b). Despite this observation, analysis of differentiated Th2 clones from wild-type or knockin reporter mice have revealed both monoallelic and biallelic expression that is stable over repeated stimulations (Bix and Locksley, 1998; Hu-Li et al., 2001; Riviere et al., 1998). Levels of IL-4 transcribed may also be modulated and reﬂect variability in the extent of chromatin stabilization across the gene (Guo et al., 2002). Taken together, the results are consistent with a gene that is transcriptionally competent at the time of TCR stimulation, but then variably modulated at individual alleles in response to subsequent signals. The resultant cells likely consist of a spectrum from clones with chromatin modiﬁed across the entire IL-4/IL-13/IL-5 cluster to clones modiﬁed across both IL-4 alleles to clones modiﬁed across a single IL-4 allele. A silencer element has been identiﬁed in the 30 untranslated region of the IL-4 gene near areas which may nucleate methylation (Kubo et al., 1997; Lee et al., 2002). Consensus binding sites for the Bcl6 and Runx repressors are within the sequence, but their functional signiﬁcance remains unknown. The pivotal role for Stat6 is borne out by the severe deﬁcit in type 2 immunity in Stat6-deﬁcient mice. These animals, as well as IL-4Ra-deﬁcient mice, are unable to expel intestinal helminths and fail to mount normal allergen-induced mucosal responses (Urban et al., 1998). These models are biologically complex, however, and may represent the conﬂuence of known or likely roles for Stat6 in induction of tissue-recruiting chemokines (Cuvelier and Patel, 2001), appropriate vascular adhesins (Issekutz et al., 2001), and T cell selectin ligands (Wagers et al., 1998). Thus, some of the Stat6-deﬁcient phenotype may reﬂect failure to localize the appropriate effector cells rather than inherent failures in Th2 differentiation itself (Mathew et al., 2001). Rechallenge of Stat6-deﬁcient mice with helminths, however, suggests a primary role for Stat6 in stabilized Th2 effector development. Primary infection of Stat6-deﬁcient mice results in a systemic IL-4 response not different from that in challenged wild-type mice. IL-4 production failed to be maintained, however, and, in contrast to wild-type mice that generated an enhanced memory response, rechallenged Stat6-deﬁcient mice reiterated a primary IL-4 response (Finkelman et al., 2000). These results suggest that IL-4 production by innate immune cells is Stat6-independent. Further, since analysis with knockin reporter mice demonstrated normal IL-4 induction in naive T cells in lymph nodes after helminth challenge, the data suggest that IL-4 production by T cells, while initially Stat6-independent, cannot be sustained in the absence of Stat6 (Mohrs et al., 2001b). Whether these effects are direct or mediated by a requirement for tissue localization of Th2 cells for their terminal differentiation will require further study.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

169

A consistent observation in Th2 cells is the localization of the IL-4 genes to euchromatin domains in the nucleus, in contrast to Th1 cells, which show repositioning of IL-4 alleles in apposition to heterochromatin, consistent with epigenetic silencing (Grogan et al., 2001). Localization to heterochromatin is used to silence other genes in lymphocytes, frequently through interactions with ikaros family proteins that mediate interactions with centromeric domains (Fisher and Merkenschlager, 2002). Ikaros sites are dispersed widely through the type 2 cytokine locus, including 4 sites clustered in the CNS-1 element. CNS-1-deﬁcient Th1 cells fail to reposition the IL-4 genes to heterochromatin, suggesting that the same DNA elements targeted by activating complexes in Th2 cells may be occupied by silencing complexes in Th1 cells that mediate terminal sequestration of the locus (Grogan and Locksley, unpublished studies). The plasticity of IL-4 expression in differentiated Th2 populations remains an area of intense interest. Early studies suggested that repeated stimulation under polarized Th2 conditions drove T cells to irreversible patterns of cytokine expression (Murphy et al., 1996). Whether such conditions occur in vivo is unknown. Studies in both human and mouse, however, have suggested a spectrum of cytokine-producing cells based on their capacity to respond rapidly to antigen recall. In the mouse, where more detail has been developed in studies of CD8 cells, a hierarchical lineage from tissue-occupying effector cells to circulating, effector memory cells to circulating, central memory cells, has been deﬁned (Wherry et al., 2003). However, analysis of human T cell populations has suggested a model of linear differentiation from central memory to effector memory cells (Sallusto and Lanzavecchia, 2001; Sallusto et al., 1999) (although contradictory evidence has been reported in human and mouse systems (Baron et al., 2003; Wu et al., 2002) ). Central memory cells keep their proliferative potential and give rise to effector/memory cells rapidly upon antigen encounter. Emerging data on CD4 T cells, although incomplete, are consistent with a similar lineage scheme (Lohning et al., 2002). By this model, Th2 effector/memory cells are tissue-inﬁltrating cells poised for rapid, stereotyped, type 2 cytokine secretion, which may have limited proliferative potential. In contrast, Th2 central memory cells retain the capacity for clonal expansion and antigen sampling in lymph nodes. Although commitment to transcribe type 2 cytokines is presumably maintained in each of these populations, the capacity to activate other cytokines, including IFNg, may be greater in central memory cells. Studies using human CD4 T cells have suggested substantial cytokine ﬂexibility among polarized effector and memory Th subsets (Messi et al., 2003). It remains unclear how in vitro-generated Th2 cells compare to cells generated in vivo, or how reliable the use of surface markers for the identiﬁcation of various subsets in vivo remains, making

170

DANIEL B. STETSON ET AL.

generalizations difﬁcult. These remain crucially important questions for further study in understanding, for instance, how chronic type 2 afﬂictions such as allergy and asthma are sustained. For instance, whether tissue Th2 effector cells undergo activation-induced cell death, as posited for Th1 tissue effectors (Hayashi et al., 2002), remains unknown. IV. Phenotype and Genotype Analysis of Th2 Cells

Polarization of Th subsets in vitro is associated with modulation of surface receptors that provide information about the cytokine milieu. In vitro, differentiating Th2 cells lose IL-12Rb2 and IL-18 receptor surface expression induced at the time of activation, while IL-4R signaling remains robust, perhaps reﬂecting failure to upregulate SOCS5, as occurs during Th1 differentiation. The IL-1R family member, T1/ST2 (IL-1RL1), is induced on the surface of Th2 cells, and can function as a co-stimulatory signal for proliferation and type 2 cytokine production (Meisel et al., 2001). CRTH2 is a 7-transmembrane, G-protein coupled receptor related to N-formyl peptide receptors that is induced on Th2 cells and facilitates chemotaxis to prostaglandin D2, the major prostaglandin metabolite of activated mast cells (Hirai et al., 2001). The histamine receptor type 1, H1R, which costimulates IFNg production in Th1 cells, is down-regulated on Th2 cells (Jutel et al., 2001). Distinct surface adhesins, including a4b7 integrin, the ligand for mucosal MAdCAM-1, are induced, whereas upregulation of the fucosyltransferase, Fuc-TVII, required for the synthesis of P- and E-selectin ligands, is dependent upon IL-12/Stat4 and, hence, not typically found on Th2 cells. Patterns of Th2 chemokine receptors, while complex and combinatorial, have demonstrated some bias towards recognition by mucosal and allergen-generated chemokines (Luther and Cyster, 2001). While more study is needed, the collective data suggest that Th2 cells acquire the potential to enter mucosal or epithelial sites in response to chemokines commonly associated with allergic inﬂammation. Several groups have used microarrays to compare developing or stable Th1 and Th2 cell lines in efforts to identify subset-speciﬁc transcripts. Transcripts with a bias towards expression in Th2 cells include cytokines (IL-4, IL-13, IL-5, IL-24), transcription factors (GATA-3, Stat6), receptors (T1/ST2), and adhesins (integrin b7). These trends have been seen in some, but not all, studies, perhaps reﬂecting varied differentiation methods and different arrays (Chtanova et al., 2001; Nagai et al., 2001; Rogge et al., 2000). Studies of Th2 cells puriﬁed from distinct tissue compartments in vivo are lacking, and will be informative in understanding the function of Th2 cells as they trafﬁc to sites of infection.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

171

V. Mutations Impacting IL-4 Expression in T Cells

Many mutations affecting IL-4 expression from Th2 cells, including deﬁciencies in IL-4, Stat6, IL-4Ra, GATA-3 (via anti-sense or dominant negative strategies) and c-Maf, or targeted deletions of cis-acting regulatory elements, have already been noted. A number of other gene deletions have effects on Th2 cell development in vitro or in vivo. In general, these can be divided into deletions or additions of exogenous soluble or surface ligands, deletions in signaling pathways, deletions in transcription factors, or imbalances resulting from deletions of key elements required for type 1 immunity. Additional uncharacterized mutations impact pathways likely related to type 2 immunity, such as serum IgE. Examples are listed in Table I, although we focus here on relatively recent observations and do not provide an exhaustive overview. Effects of deﬁciencies in CD4, lck, itk, LAT (linker for activation of T cells), Vav, PKCu and SLP76, however, begin to outline a linear signaling pathway, that, when compromised, impairs activation of IL-4 gene expression (Aguado et al., 2002; Fowell et al., 1997, 1999; Hehner et al., 2000; Yamashita et al., 1998). Small molecule inhibitors targeting aspects of this pathway might be rationally explored for their capacity to impede Th2 differentiation by pharmacologic means.

VI. Expression of IL-4 in Non-Th2 Cells

IL-4 expression has been documented from CD8 T cells, NK T cells, gd T cells, mast cells, basophils, and eosinophils. Other cell types, such as B cells and dendritic cells, have been reported to produce IL-4 under deﬁned circumstances, but such observations remain incompletely examined. Studies of IL-4 expression in non-Th2 cells should be informative in exposing core genetic programs required for hardwiring and expressing genes from the type 2 cytokine locus. Still lacking is deﬁnitive information regarding the requirements for type 2 cytokines generated by innate cells for the terminal differentiation and/or tissue localization of effector Th2 cells. A. Tc2 Cells The name ‘‘Tc2’’ was introduced by Mosmann’s group, who showed that mouse CD8 T cells could be differentiated into stable type 2 cytokineproducing clones in vitro after stimulation with alloantigen and IL-4 (Sad et al., 1995). Evidence for the existence of these cells in vivo came from studies with AIDS patients, who showed relatively high numbers of Tc2 cells in blood and skin (Romagnani et al., 1994). More recent reports show that Tc2 cells are increased in the elderly, and that high numbers of Tc2 cells correlated with a better humoral immune response after inﬂuenza vaccination (Schwaiger et al., 2003; Yen et al., 2000). Tc2 cells can provide B cell help by secretion of

TABLE I Mutations Impacting Type 2 Immunity Mutation

Phenotype

172

Ligand/receptor interactions ICOS/ Reduced IL-2, IL-4, IL-5, IL-13 Increased EAE susceptibility OX40/ Reduced IL-4, IL-5, IgE, decreased eosinophilia OX40L/ Reduced IL-4, IgE, impaired worm expulsion TNF/ Reduced Th2 cytokines DR6/ Increased Th2 cytokines

Model In vitro, asthma, immunization Asthma H. polygyrus Asthma Immunization, in vitro Aspergillus, asthma In vitro, asthma

Reference(s) (Dong et al., 2001)

C3aR/ CD81/

Reduced eosinophilia, reduced IL-4, IL-5, IL-13 Decreased Th2 cytokines, normal serum IgE

Tim-1 polymorphisms TSLP administration

Locus involved in airway hyperreactivity Produced by epithelial cells, favors Th2 differentiation, increased expression in atopic dermatitis Increased mastocytosis, dermatitis, Th2 cytokines in skin

Asthma Human, in vitro

(Jember et al., 2001) (Ekkens et al., 2003) (Matheson et al., 2002) (Liu et al., 2001a; Zhao et al., 2001) (Drouin et al., 2002) (Deng et al., 2000; Deng et al., 2002) (McIntire et al., 2001) (Soumelis et al., 2002)

Spontaneous

(Konishi et al., 2002)

Increased IgE Impaired IgG1 production Increased Th2 cytokines, IgE, eosinophilia

Spontaneous

(Ozaki et al., 2002)

In vivo administration Immunization

(Fort et al., 2001)

IL-18 transgene, skin-speciﬁc IL-21R/ IL-21R/IL-4/ IL-25 administration IL-27/

Decreased IL-4 from NK T cells; conventional T cells normal

(Nieuwenhuis et al., 2002)

Signaling molecules LAT Y136F point mutation Vav dominant negative Itk/ Fyn/ IRS2/ Tyk2/ SAP/

Impaired T cell development, increased IL-4, lymphoproliferation and eosinophilia Decreased IL-4 transcription Impaired IL-4 expression and NF-ATc nuclear translocation Increased Th2 cytokines, eosinophilia

173

P85 PI3K/

Decreased proliferation and Th2 cytokines; normal survival Increased Th2 cytokines, IgE, eosinophilia Decreased Th2 cytokines, IgE, Increased CD8 and Th1 response Defective B cell memory Mutated in XLP Reduced intestinal mast cells, delayed worm expulsion

Gab2/

Decreased mast cell function, impaired FceRI signaling

Transcription factors STAT6/

Mbd2/ Mel-18/

Normal initial IL-4 response Impaired Th2 tissue localization and IL-4 memory response Increased IL-4 expression Decreased Th2 cytokines, impaired worm expulsion

NFATc2/NFATc3/ NFATp/ IRF4/ IRF4 over-expression

IL-4-independent Th2 differentiation Reduced early IL-4 production Decreased Th2 cytokines Increased Th2 cytokines

Spontaneous In vitro In vitro, L. major In vitro, asthma In vitro In vitro, asthma LCMV, L. major

Human Strongyloides venezuelensis in vitro, immunization

(Aguado et al., 2002; Sommers et al., 2002) (Hehner et al., 2000) (Fowell et al., 1999) (Kudlacz et al., 2001; Tamura et al., 2001) (Wurster et al., 2002) (Seto et al., 2003) (Czar et al., 2001; Wu et al., 2001) (Crotty et al., 2003) (Morra et al., 2001) (Fukao et al., 2002) (Gu et al., 2001)

N. brasiliensis

(Finkelman et al., 2000) (Mohrs et al., 2001b)

In vitro In vitro, N. brasiliensis In vitro Immunization In vitro In vitro

(Hutchins et al., 2002) (Kimura et al., 2001) (Rengarajan et al., 2002b) (Hodge et al., 1996) (Rengarajan et al., 2002a) (Hu et al., 2002) (continues)

TABLE I (continued) Mutation Foxp3

174

Miscellaneous Itch/ SPINK5 CC10/

Phenotype

Model

Reference(s)

Mutated in XLAAD (allergy/autoimmunity)

Human

Mutated in Scurfy mouse (lymphoproliferation) Essential for regulatory T cell (Treg) development

Spontaneous In vitro

(Bennett et al., 2001 Chatila et al., 2000) (Brunkow et al., 2001) (Hori et al., 2003)

E3 ubiquitin ligase: enhanced Th2 cytokines, increased serum IgE, IgG1 Serine protease inhibitor: mutated in Netherton syndrome (atopy/allergy) Surfactant cell secreted protein: increased Th2 cytokines, eosinophilia

In vitro

(Fang et al., 2002)

Human

(Chavanas et al., 2000; Walley et al., 2001) (Chen et al., 2001)

Asthma

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

175

IL-4 (Maggi et al., 1994), but, like Tc1 cells, display cytotoxic potential (Sad et al., 1995). Further, it has been shown that Tc2 cells could reduce the growth of established lung metastases in a mouse tumor model (Dobrzanski et al., 1999). Conditions that lead to the differentiation of Tc2 cells in vivo and their true physiological role remain incompletely characterized. B. NK T Cells NK T cells are a relatively small population of T cells that express the NK cell marker, NK1.1, and a restricted repertoire of ab T cell receptors (Kronenberg and Gapin, 2002). The most prevalent NK T cells fortuitously recognize the nonclassical MHC molecule, CD1d, when loaded with agalactosylceramide (a-GalCer), which presumably mimics the unknown selecting ligand. These cells, designated invariant NK T cells (NK Ti) because of their use of the same TCRs in mouse and human, arise from a common T cell precursor after stochastic rearrangements generate a canonical receptor that permits selection on CD1d presented by double-negative thymocytes. In contrast to ab T cells, NK T cells acquire the capacity for rapid IL-4 and IFNg production during thymic development (Benlagha et al., 2002). After migration to liver, bone marrow, and spleen, NK Ti cells undergo terminal differentiation. Essentially all of the early IL-4 that appears in blood after administration of anti-CD3 or a-GalCer is from NK Ti cells. Evidence that these cells are required for type 2 immunity is lacking, however. Their rapid cytokine responses have been exploited to bias immune responses in various infectious diseases, autoimmune, and tumor models. Production of IL-4 by NK Ti cells is unaffected by Stat6-deﬁciency, attenuated in itk-deﬁciency, and, unexpectedly, highly compromised in the absence of EBI3, an IL-12 p40 homolog that dimerizes with IL-12 p35 or its homolog, p28, to create IL-23. Despite data demonstrating induction of IFNg by IL-23 in T cells, NK Ti development and IL-4 production are severely attenuated in the absence of EBI3 (Nieuwenhuis et al., 2002). C. gd T Cells gd T cells constitute 1 to 5% of lymphocytes in the blood of adult animals, but can compose up to 50% of T cells in epithelial tissues such as skin and the gastrointestinal tract. They are not restricted by classical MHC, can recognize soluble protein and nonprotein antigens, and many have cytotoxic function (Carding and Egan, 2002). gd T cells predominate during early fetal development and the expression of IL-4 by fetal thymocytes was ﬁrst described 15 years ago (Tentori et al., 1988). Human pro-T cells isolated from neonatal thymus also produce IL-4 (Barcena et al., 1991). In vitro cultures of these cells in the presence of IL-4 gave rise to gd T cells, suggesting that development of thymic gd T cells might be regulated by autocrine or paracrine stimulatory

176

DANIEL B. STETSON ET AL.

signals provided by IL-4. Some gd T cells isolated in adult life express the NK1.1 marker and secrete IL-4 upon in vitro stimulation (Azuara et al., 1997). This subset of IL-4 producing gd T cells exists predominantly in the liver and spleen (Gerber et al., 1999). In the mouse, gd T cells can be induced to secrete type 1 or type 2 cytokines after polarizing infections (Carding et al., 1993; Ferrick et al., 1995). Further, in a pulmonary allergic asthma model, gd T celldeﬁcient mice had reduced recruitment of effector cells into the lung and attenuated IgE levels. Administration of recombinant IL-4 during immunization of gd T cell-deﬁcient mice restored these responses, suggesting that gd T cells might contribute to early IL-4 production during the development of allergic asthma (Zuany-Amorim et al., 1998). D. Mast Cells and Basophils Mast cells and basophils express the high-afﬁnity IgE receptor and release inﬂammatory mediators and cytokines after IgE cross-linking (Kinet, 1999; Williams and Galli, 2000). Both cell types express IL-4 (Bradding et al., 1992; Brown et al., 1987; MacGlashan et al., 1994; Plaut et al., 1989; Seder et al., 1991). Further, mouse and human basophils dramatically upregulate IL-4 transcription and release IL-4 protein within one hour of allergen challenge (Genovese et al., 2003; Luccioli et al., 2002). This effect was dependent on the high-afﬁnity IgE receptor and was also observed in KitW/KitWv mice, which are genetically mast cell-deﬁcient, suggesting that basophils might be an important source for early IL-4 in vivo. IL-4 expression in mast cells is Stat6- and GATA3-independent, in contrast to Th2 cells, but requires Stat5 and GATA-1/-2 (Sherman, 2001; Weiss and Brown, 2001). E. Eosinophils Distinct from other IL-4-expressing cells, human eosinophils store preformed IL-4 in granules, which can be released rapidly upon stimulation (Bandeira-Melo et al., 2001; Moqbel et al., 1995). In vitro, mouse eosinophils can secrete IL-4 within 18 hours after stimulation, which is blocked by actinomycin D. Thus, mouse eosinophils likely do not store preformed IL-4 but rapidly synthesize it after activation (Justice et al., 2002). In Schistosoma egg granulomas, eosinophils appear to be the major source of IL-4 (Rumbley et al., 1999). Tissue-inﬁltrating eosinophils were also demonstrated to express IL-4 after helminth infection, even in the absence of an adaptive immune response (Shinkai et al., 2002). F. B Cells A 2000 study has shown that B cells can be differentiated into cytokinesecreting effector B cells (Harris et al., 2000). Like T cells, B cells could be polarized to secrete type 1 or type 2 cytokines. Whether they play an important role in vivo remains to be determined.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

177

VII. Where Does Type 2 Immunity Operate?

In vivo, type 2 immunity is most readily apparent after mucosal challenge with helminths and allergens, and, in each, depletion of Th2 cells ablates or substantially attenuates the response. Systemically, type 2 responses can function in a regulatory role by abrogating potential inﬂammatory damage mediated by activated type 1 immune cells (Matsukawa et al., 2001). The propensity for type 2 immune responses to develop at epithelial barriers suggests several hypotheses regarding this apparent geographic predeliction, none of which are mutually exclusive (Fig. 2a). First, antigens acquired across epithelial barriers might possess intrinsic properties that drive type 2 immune responses. Thus, many epithelial antigens, whether derived from exogenous allergens, invading helminths, or biting insects, share the properties of stability to harsh conditions and intrinsic proteolytic activity. Such biophysical attributes may enhance transepithelial delivery to sentinel dendritic cells or contribute to modulating signals via unknown molecular recognition receptors. Crude cocktails of helminthderived antigens frequently contain type 2 polarizing activities (Holland et al., 2000). Despite the explosion of information regarding the role of Toll family proteins as molecular recognition sensors important in mediating type 1 immune responses (Medzhitov, 2001), deﬁnitive identiﬁcation of a pattern recognition receptor critical for initiating type 2 immunity remains elusive (Fig. 2b). A second possibility is that distinct types of dendritic cells populate epithelial and mucosal surfaces. The ability to identify different types of DCs with differing capacity to activate naive T cells to develop into Th1 or Th2 cells has introduced further complexity (Liu et al., 2001b; Patterson, 2000; Reis e Sousa, 2001). Differential recruitment of dendritic cells from the periphery or blood, such as ‘‘pre-DC-2’’ plasmacytoid DCs, may impact T cell polarization through varied display of surface ligands that can affect cytokine production (Table I). IL-4 itself, by regulating expression of sampling receptors, such as DC-SIGN, can affect antigen delivery to the draining lymph node (Relloso et al., 2002). Delivery of antigens to mucosal sites can bias toward type 2 immune responses (Constant et al., 2000). Much more information is needed regarding which populations of DCs become activated during type 2 immunity. A third hypothesis, alluded to earlier, remains the possibility that terminal Th2 differentiation in vivo occurs in tissue, and that signals that mediate this occur predominantly in epithelia. Accumulating evidence supports the speculation that naive CD4 T cells might be ‘‘poised’’ after undergoing clonal expansion in the draining lymph node, but terminally differentiated by cytokine signals, like IL-4, delivered later, perhaps after tissue inﬁltration (Doyle et al., 2001; Mohrs et al., 2003; Wang and Mosmann, 2001). The positioning and/or recruitment of IL-4-expressing innate cells, including mast cells, eosinophils,

178

DANIEL B. STETSON ET AL.

Fig 2 Initiation and expression of type 2 immunity. (a) A schematic diagram depicting three levels at which Type 2 immune responses may be regulated. (b) A simpliﬁed comparison of how type 1 and type 2 responses are initiated. Recognition of pathogen-associated molecular patterns (PAMPS) by pattern recognition receptors (PRRs) results in the production of cytokines which modulate T cell differentiation in the draining lymph nodes. While the basic elements underlying all of these events have been elucidated for Type 1 immunity, very little is known about how recognition of type 2 pathogens is linked to Th2 differentiation. (See Color Insert.)

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

179

and basophils, to invaded barriers would then serve to deliver terminal differentiation signals to poised T cells entering the tissue. In turn, developed Th2 cells reinforce the growth, recruitment, and activation of mast cells, basophils, and eosinophils, resulting in sustained, autoreinforcing, changes at epithelial barriers. The latter, although incompletely characterized, include the development of changes in water and ion exchange, mucus hypersecretion, and, ultimately, subepithelial ﬁbrosis, that together impede access to blood by arthropods or helminths, or tissue invasion by widely prevalent environmental allergens (Madden et al., 2002). Unregulated, however, epithelial type 2 responses may contribute to the clinical manifestations of allergy and asthma. The overlapping phenotype of type 2-associated pathology generated by overexpression of any of the genetically linked type 2 cytokines—IL-4, IL-13, IL-5, and IL-9 (on the same chromosome in humans but the latter split off in the mouse)—suggests integrated functional redundancies (Lee et al., 1997; Rankin et al., 1996; Temann et al., 2002; Zhu et al., 1999). Pairwise comparisons of variable combinations of knockouts established that IL-4 alone, even in the absence of IL-13, IL-5, and IL-9, could mediate all of the attributes of type 2 immunopathology, given a sufﬁciently strong stimulus in vivo (Fallon et al., 2002). Thus, type 2 immunity consists of an orchestrated response underpinned by innate cells but centrally mediated, in its adaptive phase, by Th2 cells. VIII. Concluding Remarks

We have attempted to summarize recent insights regarding the generation of Th2 cells and factors regulating expression of IL-4, the canonical cytokine secreted by these cells. Th2 cells orchestrate a stereotyped cell response characterized by eosinophil and basophil inﬁltration, mast cell hyperplasia, and changes in epithelial mucus and water secretion. Th2 cells have been demonstrated to mediate protection against a variety of intestinal helminths but also ectoparasites that invade at epithelial barriers. When activated systemically, Th2 cells protect against the tissue-damaging proinﬂammatory actions of Th1 cells. When dysregulated, Th2 cells drive chronic atopic diseases of wide prevalence, such as allergy and asthma. Basic insights into the understanding of the generation, activation, tissue localization, and turnover of Th2 cells will provide numerous opportunities for sustained beneﬁts to human health. Acknowledgments The authors regret being unable to cite all of the relevant original literature due to space constraints. Supported in part by the Howard Hughes Medical Institute and grants from the National Institutes of Health (AI26918, AI30663, HL56385). RML is a Larry Ellison Global Infectious Diseases Senior Scholar.

180

DANIEL B. STETSON ET AL.

References Agarwal, S., and Rao, A. (1998). Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 9, 765–775. Aguado, E., Richelme, S., Nunez-Cruz, S., Miazek, A., Mura, A. M., Richelme, M., Guo, X. J., Sainty, D., He, H. T., Malissen, B., and Malissen, M. (2002). Induction of T helper type 2 immunity by a point mutation in the LAT adaptor. Science 296, 2036–2040. Asnagli, H., Afkarian, M., and Murphy, K. M. (2002). Cutting edge: Identiﬁcation of an alternative GATA-3 promoter directing tissue-speciﬁc gene expression in mouse and human. J. Immunol. 168, 4268–4271. Avni, O., Lee, D., Macian, F., Szabo, S. J., Glimcher, L. H., and Rao, A. (2002). T(H) cell differentiation is accompanied by dynamic changes in histone acetylation of cytokine genes. Nat. Immunol. 3, 643–651. Azuara, V., Levraud, J. P., Lembezat, M. P., and Pereira, P. (1997). A novel subset of adult gamma delta thymocytes that secretes a distinct pattern of cytokines and expresses a very restricted T cell receptor repertoire. Eur. J. Immunol. 27, 544–553. Bandeira-Melo, C., Sugiyama, K., Woods, L. J., and Weller, P. F. (2001). Cutting edge: Eotaxin elicits rapid vesicular transport-mediated release of preformed IL-4 from human eosinophils. J. Immunol. 166, 4813–4817. Barcena, A., Sanchez, M. J., de la Pompa, J. L., Toribio, M. L., Kroemer, G., and Martinez, A. C. (1991). Involvement of the interleukin 4 pathway in the generation of functional gamma delta T cells from human pro-T cells. Proc. Natl. Acad. Sci. USA 88, 7689–7693. Baron, V., Bouneaud, C., Cumano, A., Lim, A., Arstila, T. P., Kourilsky, P., Ferradini, L., and Pannetier, C. (2003). The repertoires of circulating human CD8(þ) central and effector memory T cell subsets are largely distinct. Immunity 18, 193–204. Benlagha, K., Kyin, T., Beavis, A., Teyton, L., and Bendelac, A. (2002). A thymic precursor to the NK T cell lineage. Science 296, 553–555. Bennett, C. L., Christie, J., Ramsdell, F., Brunkow, M. E., Ferguson, P. J., Whitesell, L., Kelly, T. E., Saulsbury, F. T., Chance, P. F., and Ochs, H. D. (2001). The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat. Genet. 27, 20–21. Bird, J. J., Brown, D. R., Mullen, A. C., Moskowitz, N. H., Mahowald, M. A., Sider, J. R., Gajewski, T. F., Wang, C. R., and Reiner, S. L. (1998). Helper T cell differentiation is controlled by the cell cycle. Immunity 9, 229–237. Bix, M., and Locksley, R. M. (1998). Independent and epigenetic regulation of the interleukin-4 alleles in CD4þ T cells. Science 281, 1352–1354. Bradding, P., Feather, I. H., Howarth, P. H., Mueller, R., Roberts, J. A., Britten, K., Bews, J. P., Hunt, T. C., Okayama, Y., Heusser, C. H., et al. (1992). Interleukin 4 is localized to and released by human mast cells. J. Exp. Med. 176, 1381–1386. Brown, M. A., Pierce, J. H., Watson, C. J., Falco, J., Ihle, J. N., and Paul, W. E. (1987). B cell stimulatory factor-1/interleukin-4 mRNA is expressed by normal and transformed mast cells. Cell 50, 809–818. Brunkow, M. E., Jeffery, E. W., Hjerrild, K. A., Paeper, B., Clark, L. B., Yasayko, S. A., Wilkinson, J. E., Galas, D., Ziegler, S. F., and Ramsdell, F. (2001). Disruption of a new forkhead/wingedhelix protein, scurﬁn, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat. Genet. 27, 68–73. Carding, S. R., Allan, W., McMickle, A., and Doherty, P. C. (1993). Activation of cytokine genes in T cells during primary and secondary murine inﬂuenza pneumonia. J. Exp. Med. 177, 475–482. Carding, S. R., and Egan, P. J. (2002). Gammadelta T cells: Functional plasticity and heterogeneity. Nat. Rev. Immunol. 2, 336–345.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

181

Chatila, T. A., Blaeser, F., Ho, N., Lederman, H. M., Voulgaropoulos, C., Helms, C., and Bowcock, A. M. (2000). JM2, encoding a fork head-related protein, is mutated in X-linked autoimmunity– allergic disregulation syndrome. J. Clin. Invest. 106, R75–81. Chavanas, S., Bodemer, C., Rochat, A., Hamel-Teillac, D., Ali, M., Irvine, A. D., Bonafe, J. L., Wilkinson, J., Taieb, A., Barrandon, Y., et al. (2000). Mutations in SPINK5, encoding a serine protease inhibitor, cause Netherton syndrome. Nat. Genet. 25, 141–142. Chen, L. C., Zhang, Z., Myers, A. C., and Huang, S. K. (2001). Cutting edge: Altered pulmonary eosinophilic inﬂammation in mice deﬁcient for Clara cell secretory 10-kDa protein. J. Immunol. 167, 3025–3028. Chtanova, T., Kemp, R. A., Sutherland, A. P., Ronchese, F., and Mackay, C. R. (2001). Gene microarrays reveal extensive differential gene expression in both CD4(þ) and CD8(þ) type 1 and type 2 T cells. J. Immunol. 167, 3057–3063. Cirillo, L. A., Lin, F. R., Cuesta, I., Friedman, D., Jarnik, M., and Zaret, K. S. (2002). Opening of compacted chromatin by early developmental transcription factors HNF3 (FoxA) and GATA-4. Mol. Cell 9, 279–289. Constant, S. L., Lee, K. S., and Bottomly, K. (2000). Site of antigen delivery can inﬂuence T cell priming: Pulmonary environment promotes preferential Th2-type differentiation. Eur. J. Immunol. 30, 840–847. Crotty, S., Kersh, E. N., Cannons, J., Schwartzberg, P. L., and Ahmed, R. (2003). SAP is required for generating long-term humoral immunity. Nature 421, 282–287. Cuvelier, S. L., and Patel, K. D. (2001). Shear-dependent eosinophil transmigration on interleukin 4-stimulated endothelial cells: A role for endothelium-associated eotaxin-3. J. Exp. Med. 194, 1699–1709. Czar, M. J., Kersh, E. N., Mijares, L. A., Lanier, G., Lewis, J., Yap, G., Chen, A., Sher, A., Duckett, C. S., Ahmed, R., and Schwartzberg, P. L. (2001). Altered lymphocyte responses and cytokine production in mice deﬁcient in the X-linked lymphoproliferative disease gene SH2D1A/DSHP/ SAP. Proc. Natl. Acad. Sci. USA 98, 7449–7454. Deng, J., Dekruyff, R. H., Freeman, G. J., Umetsu, D. T., and Levy, S. (2002). Critical role of CD81 in cognate T–B cell interactions leading to Th2 responses. Int. Immunol. 14, 513–523. Deng, J., Yeung, V. P., Tsitoura, D., DeKruyff, R. H., Umetsu, D. T., and Levy, S. (2000). Allergeninduced airway hyperreactivity is diminished in CD81-deﬁcient mice. J. Immunol. 165, 5054–5061. Dobrzanski, M. J., Reome, J. B., and Dutton, R. W. (1999). Therapeutic effects of tumor-reactive type 1 and type 2 CD8þ T cell subpopulations in established pulmonary metastases. J. Immunol. 162, 6671–6680. Dong, C., Juedes, A. E., Temann, U. A., Shresta, S., Allison, J. P., Ruddle, N. H., and Flavell, R. A. (2001). ICOS co-stimulatory receptor is essential for T-cell activation and function. Nature 409, 97–101. Doyle, A. M., Mullen, A. C., Villarino, A. V., Hutchins, A. S., High, F. A., Lee, H. W., Thompson, C. B., and Reiner, S. L. (2001). Induction of cytotoxic T lymphocyte antigen 4 (CTLA-4) restricts clonal expansion of helper T cells. J. Exp. Med. 194, 893–902. Drouin, S. M., Corry, D. B., Hollman, T. J., Kildsgaard, J., and Wetsel, R. A. (2002). Absence of the complement anaphylatoxin C3a receptor suppresses Th2 effector functions in a murine model of pulmonary allergy. J. Immunol. 169, 5926–5933. Ekkens, M. J., Liu, Z., Liu, Q., Whitmire, J., Xiao, S., Foster, A., Pesce, J., VanNoy, J., Sharpe, A. H., Urban, J. F., and Gause, W. C. (2003). The role of OX40 ligand interactions in the development of the Th2 response to the gastrointestinal nematode parasite Heligmosomoides polygyrus. J. Immunol. 170, 384–393.

182

DANIEL B. STETSON ET AL.

Fallon, P. G., Jolin, H. E., Smith, P., Emson, C. L., Townsend, M. J., Fallon, R., and McKenzie, A. N. (2002). IL-4 induces characteristic Th2 responses even in the combined absence of IL-5, IL-9, and IL-13. Immunity 17, 7–17. Fang, D., Elly, C., Gao, B., Fang, N., Altman, Y., Joazeiro, C., Hunter, T., Copeland, N., Jenkins, N., and Liu, Y. C. (2002). Dysregulation of T lymphocyte function in itchy mice: A role for Itch in TH2 differentiation. Nat. Immunol. 3, 281–287. Ferrick, D. A., Schrenzel, M. D., Mulvania, T., Hsieh, B., Ferlin, W. G., and Lepper, H. (1995). Differential production of interferon-gamma and interleukin-4 in response to Th1- and Th2stimulating pathogens by gamma delta T cells in vivo. Nature 373, 255–257. Fields, P. E., Kim, S. T., and Flavell, R. A. (2002). Cutting edge: Changes in histone acetylation at the IL-4 and IFN-gamma loci accompany Th1/Th2 differentiation. J. Immunol. 169, 647–650. Finkelman, F. D., Morris, S. C., Orekhova, T., Mori, M., Donaldson, D., Reiner, S. L., Reilly, N. L., Schopf, L., and Urban, J. F., Jr. (2000). Stat6 regulation of in vivo IL-4 responses. J. Immunol. 164, 2303–2310. Fisher, A. G., and Merkenschlager, M. (2002). Gene silencing, cell fate and nuclear organisation. Curr. Opin. Genet. Dev. 12, 193–197. Fort, M. M., Cheung, J., Yen, D., Li, J., Zurawski, S. M., Lo, S., Menon, S., Clifford, T., Hunte, B., Lesley, R., et al. (2001). IL-25 induces IL-4, IL-5, and IL-13 and Th2-associated pathologies in vivo. Immunity 15, 985–995. Fowell, D. J., Magram, J., Turck, C. W., Killeen, N., and Locksley, R. M. (1997). Impaired Th2 subset development in the absence of CD4. Immunity 6, 559–569. Fowell, D. J., Shinkai, K., Liao, X. C., Beebe, A. M., Coffman, R. L., Littman, D. R., and Locksley, R. M. (1999). Impaired NFATc translocation and failure of Th2 development in Itk-deﬁcient CD4þ T cells. Immunity 11, 399–409. Fox, A. H., Liew, C., Holmes, M., Kowalski, K., Mackay, J., and Crossley, M. (1999). Transcriptional cofactors of the FOG family interact with GATA proteins by means of multiple zinc ﬁngers. EMBO J. 18, 2812–2822. Fukao, T., Yamada, T., Tanabe, M., Terauchi, Y., Ota, T., Takayama, T., Asano, T., Takeuchi, T., Kadowaki, T., Hata Ji, J., and Koyasu, S. (2002). Selective loss of gastrointestinal mast cells and impaired immunity in PI3K-deﬁcient mice. Nat. Immunol. 3, 295–304. Genovese, A., Borgia, G., Bjorck, L., Petraroli, A., De Paulis, A., Piazza, M., and Marone, G. (2003). Immunoglobulin superantigen protein L induces IL-4 and IL-13 secretion from human Fc varepsilon RI(þ) cells through interaction with the kappa light chains of IgE. J. Immunol. 170, 1854–1861. Gerber, D. J., Azuara, V., Levraud, J. P., Huang, S. Y., Lembezat, M. P., and Pereira, P. (1999). IL-4producing gamma delta T cells that express a very restricted TCR repertoire are preferentially localized in liver and spleen. J. Immunol. 163, 3076–3082. Gregoire, J. M., and Romeo, P. H. (1999). T-cell expression of the human GATA-3 gene is regulated by a non-lineage-speciﬁc silencer. J. Biol. Chem. 274, 6567–6578. Grogan, J. L., Mohrs, M., Harmon, B., Lacy, D. A., Sedat, J. W., and Locksley, R. M. (2001). Early transcription and silencing of cytokine genes underlie polarization of T helper cell subsets. Immunity 14, 205–215. Gu, H., Saito, K., Klaman, L. D., Shen, J., Fleming, T., Wang, Y., Pratt, J. C., Lin, G., Lim, B., Kinet, J. P., and Neel, B. G. (2001). Essential role for Gab2 in the allergic response. Nature 412, 186–190. Guo, L., Hu-Li, J., Zhu, J., Watson, C. J., Diﬁlippantonio, M. J., Pannetier, C., and Paul, W. E. (2002). In TH2 cells the IL-4 gene has a series of accessibility states associated with distinctive probabilities of IL-4 production. Proc. Natl. Acad. Sci. USA 99, 10623–10628.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

183

Harris, D. P., Haynes, L., Sayles, P. C., Duso, D. K., Eaton, S. M., Lepak, N. M., Johnson, L. L., Swain, S. L., and Lund, F. E. (2000). Reciprocal regulation of polarized cytokine production by effector B and T cells. Nat. Immunol. 1, 475–482. Hartenstein, B., Teurich, S., Hess, J., Schenkel, J., Schorpp-Kistner, M., and Angel, P. (2002). Th2 cell-speciﬁc cytokine expression and allergen-induced airway inﬂammation depend on JunB. EMBO J. 21, 6321–6329. Hayashi, N., Liu, D., Min, B., Ben-Sasson, S. Z., and Paul, W. E. (2002). Antigen challenge leads to in vivo activation and elimination of highly polarized TH1 memory T cells. Proc. Natl. Acad. Sci. USA 99, 6187–6191. Hehner, S. P., Li-Weber, M., Giaisi, M., Droge, W., Krammer, P. H., and Schmitz, M. L. (2000). Vav synergizes with protein kinase C theta to mediate IL-4 gene expression in response to CD28 costimulation in T cells. J. Immunol. 164, 3829–3836. Hirai, H., Tanaka, K., Yoshie, O., Ogawa, K., Kenmotsu, K., Takamori, Y., Ichimasa, M., Sugamura, K., Nakamura, M., Takano, S., and Nagata, K. (2001). Prostaglandin D2 selectively induces chemotaxis in T helper type 2 cells, eosinophils, and basophils via seven-transmembrane receptor CRTH2. J. Exp. Med. 193, 255–261. Hodge, M. R., Chun, H. J., Rengarajan, J., Alt, A., Lieberson, R., and Glimcher, L. H. (1996). NFAT-driven interleukin-4 transcription potentiated by NIP45. Science 274, 1903–1905. Holland, M. J., Harcus, Y. M., Riches, P. L., and Maizels, R. M. (2000). Proteins secreted by the parasitic nematode Nippostrongylus brasiliensis act as adjuvants for Th2 responses. Eur. J. Immunol. 30, 1977–1987. Hori, S., Nomura, T., and Sakaguchi, S. (2003). Control of regulatory T cell development by the transcription factor Foxp3. Science 299, 1057–1061. Hu, C. M., Jang, S. Y., Fanzo, J. C., and Pernis, A. B. (2002). Modulation of T cell cytokine production by interferon regulatory factor-4. J. Biol. Chem. 277, 49238–49246. Hu-Li, J., Pannetier, C., Guo, L., Lohning, M., Gu, H., Watson, C., Assenmacher, M., Radbruch, A., and Paul, W. E. (2001). Regulation of expression of IL-4 alleles: Analysis using a chimeric GFP/IL-4 gene. Immunity 14, 1–11. Hutchins, A. S., Mullen, A. C., Lee, H. W., Sykes, K. J., High, F. A., Hendrich, B. D., Bird, A. P., and Reiner, S. L. (2002). Gene silencing quantitatively controls the function of a developmental trans-activator. Mol. Cell 10, 81–91. Hwang, E. S., Choi, A., and Ho, I. C. (2002). Transcriptional regulation of GATA-3 by an intronic regulatory region and fetal liver zinc ﬁnger protein 1. J. Immunol. 169, 248–253. Issekutz, T. B., Palecanda, A., Kadela-Stolarz, U., and Marshall, J. S. (2001). Blockade of either alpha-4 or beta-7 integrins selectively inhibits intestinal mast cell hyperplasia and worm expulsion in response to Nippostrongylus brasiliensis infection. Eur. J. Immunol. 31, 860–868. Jankovic, D., Kullberg, M. C., Noben-Trauth, N., Caspar, P., Paul, W. E., and Sher, A. (2000). Single cell analysis reveals that IL-4 receptor/Stat6 signaling is not required for the in vivo or in vitro development of CD4þ lymphocytes with a Th2 cytokine proﬁle. J. Immunol. 164, 3047–3055. Jember, A. G., Zuberi, R., Liu, F. T., and Croft, M. (2001). Development of allergic inﬂammation in a murine model of asthma is dependent on the costimulatory receptor OX40. J. Exp. Med. 193, 387–392. Justice, J. P., Borchers, M. T., Lee, J. J., Rowan, W. H., Shibata, Y., and Van Scott, M. R. (2002). Regweed-induced expression of GATA-3, IL-4, and IL-5 by eosinophils in the lungs of allergic C57BL/6J mice. Am. J. Physiol. Lung Cell Mol. Physiol. 282, L302–309. Jutel, M., Klunker, S., Akdis, M., Malolepszy, J., Thomet, O. A., Zak-Nejmark, T., Blaser, K., and Akdis, C. A. (2001). Histamine upregulates Th1 and downregulates Th2 responses due to different patterns of surface histamine 1 and 2 receptor expression. Int. Arch. Allergy Immunol. 124, 190–192.

184

DANIEL B. STETSON ET AL.

Kelly, B. L., and Locksley, R. M. (2000). Coordinate regulation of the IL-4, IL-13, and IL-5 cytokine cluster in Th2 clones revealed by allelic expression patterns. J. Immunol. 165, 2982–2986. Kim, J. I., Ho, I. C., Grusby, M. J., and Glimcher, L. H. (1999). The transcription factor c-Maf controls the production of interleukin-4 but not other Th2 cytokines. Immunity 10, 745–751. Kimura, M., Koseki, Y., Yamashita, M., Watanabe, N., Shimizu, C., Katsumoto, T., Kitamura, T., Taniguchi, M., Koseki, H., and Nakayama, T. (2001). Regulation of Th2 cell differentiation by mel-18, a mammalian polycomb group gene. Immunity 15, 275–287. Kinet, J. P. (1999). Atopic allergy and other hypersensitivities. Curr. Opin. Immunol. 11, 603–605. Kishikawa, H., Sun, J., Choi, A., Miaw, S. C., and Ho, I. C. (2001). The cell type-speciﬁc expression of the murine IL-13 gene is regulated by GATA-3. J. Immunol. 167, 4414–4420. Konishi, H., Tsutsui, H., Murakami, T., Yumikura-Futatsugi, S., Yamanaka, K., Tanaka, M., Iwakura, Y., Suzuki, N., Takeda, K., Akira, S., et al. (2002). IL-18 contributes to the spontaneous development of atopic dermatitis-like inﬂammatory skin lesion independently of IgE/stat6 under speciﬁc pathogen-free conditions. Proc. Natl. Acad. Sci. USA 99, 11340–11345. Kronenberg, M., and Gapin, L. (2002). The unconventional lifestyle of NKT cells. Nat. Rev. Immunol. 2, 557–568. Kubo, M., Ransom, J., Webb, D., Hashimoto, Y., Tada, T., and Nakayama, T. (1997). T-cell subsetspeciﬁc expression of the IL-4 gene is regulated by a silencer element and STAT6. EMBO J. 16, 4007–4020. Kudlacz, E. M., Andresen, C. J., Salaﬁa, M., Whitney, C. A., Naclerio, B., and Changelian, P. S. (2001). Genetic ablation of the src kinase p59fynT exacerbates pulmonary inﬂammation in an allergic mouse model. Am. J. Respir. Cell Mol. Biol. 24, 469–474. Lacy, D. A., Wang, Z. E., Symula, D. J., McArthur, C. J., Rubin, E. M., Frazer, K. A., and Locksley, R. M. (2000). Faithful expression of the human 5q31 cytokine cluster in transgenic mice. J. Immunol. 164, 4569–4574. Lee, D. U., Agarwal, S., and Rao, A. (2002). Th2 lineage commitment and efﬁcient IL-4 production involves extended demethylation of the IL-4 gene. Immunity 16, 649–660. Lee, G. R., Fields, P. E., and Flavell, R. A. (2001a). Regulation of IL-4 gene expression by distal regulatory elements and GATA-3 at the chromatin level. Immunity 14, 447–459. Lee, P. P., Fitzpatrick, D. R., Beard, C., Jessup, H. K., Lehar, S., Makar, K. W., Perez-Melgosa, M., Sweetser, M. T., Schlissel, M. S., Nguyen, S., et al. (2001b). A critical role for Dnmt1 and DNA methylation in T cell development, function, and survival. Immunity 15, 763–774. Lee, J. J., McGarry, M. P., Farmer, S. C., Denzler, K. L., Larson, K. A., Carrigan, P. E., Brenneise, I. E., Horton, M. A., Haczku, A., Gelfand, E. W., et al. (1997). Interleukin-5 expression in the lung epithelium of transgenic mice leads to pulmonary changes pathognomonic of asthma. J. Exp. Med. 185, 2143–2156. Lieberson, R., Mowen, K. A., McBride, K. D., Leautaud, V., Zhang, X., Suh, W. K., Wu, L., and Glimcher, L. H. (2001). Tumor necrosis factor receptor-associated factor (TRAF)2 represses the T helper cell type 2 response through interaction with NFAT-interacting protein (NIP45). J. Exp. Med. 194, 89–98. Liu, J., Na, S., Glasebrook, A., Fox, N., Solenberg, P. J., Zhang, Q., Song, H. Y., and Yang, D. D. (2001a). Enhanced CD4þ T cell proliferation and Th2 cytokine production in DR6-deﬁcient mice. Immunity 15, 23–34. Liu, Y. J., Kanzler, H., Soumelis, V., and Gilliet, M. (2001b). Dendritic cell lineage, plasticity, and cross-regulation. Nat. Immunol. 2, 585–589. Lohning, M., Richter, A., and Radbruch, A. (2002). Cytokine memory of T helper lymphocytes. Adv. Immunol. 80, 115–181. Loots, G. G., Locksley, R. M., Blankespoor, C. M., Wang, Z. E., Miller, W., Rubin, E. M., and Frazer, K. A. (2000). Identiﬁcation of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288, 136–140.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

185

Luccioli, S., Brody, D. T., Hasan, S., Keane-Myers, A., Prussin, C., and Metcalfe, D. D. (2002). IgE(þ), Kit(), I-A/I-E() myeloid cells are the initial source of Il-4 after antigen challenge in a mouse model of allergic pulmonary inﬂammation. J. Allergy Clin. Immunol. 110, 117–124. Luther, S. A., and Cyster, J. G. (2001). Chemokines as regulators of T cell differentiation. Nat. Immunol. 2, 102–107. MacGlashan, D., Jr., White, J. M., Huang, S. K., Ono, S. J., Schroeder, J. T., and Lichtenstein, L. M. (1994). Secretion of IL-4 from human basophils. The relationship between IL-4 mRNA and protein in resting and stimulated basophils. J. Immunol. 152, 3006–3016. Madden, K. B., Whitman, L., Sullivan, C., Gause, W. C., Urban, J. F., Jr., Katona, I. M., Finkelman, F. D., and Shea-Donohue, T. (2002). Role of STAT6 and mast cells in IL-4and IL-13-induced alterations in murine intestinal epithelial cell function. J. Immunol. 169, 4417–4422. Maggi, E., Giudizi, M. G., Biagiotti, R., Annunziato, F., Manetti, R., Piccinni, M. P., Parronchi, P., Sampognaro, S., Giannarini, L., Zuccati, G., et al. (1994). Th2-like CD8þ T cells showing B cell helper function and reduced cytolytic activity in human immunodeﬁciency virus type 1 infection. J. Exp. Med. 180, 489–495. Matheson, J. M., Lemus, R., Lange, R. W., Karol, M. H., and Luster, M. I. (2002). Role of tumor necrosis factor in toluene diisocyanate asthma. Am. J. Respir. Cell Mol. Biol. 27, 396–405. Mathew, A., MacLean, J. A., DeHaan, E., Tager, A. M., Green, F. H., and Luster, A. D. (2001). Signal transducer and activator of transcription 6 controls chemokine production and T helper cell type 2 cell trafﬁcking in allergic pulmonary inﬂammation. J. Exp. Med. 193, 1087–1096. Matsukawa, A., Kaplan, M. H., Hogaboam, C. M., Lukacs, N. W., and Kunkel, S. L. (2001). Pivotal role of signal transducer and activator of transcription (Stat)4 and Stat6 in the innate immune response during sepsis. J. Exp. Med. 193, 679–688. McIntire, J. J., Umetsu, S. E., Akbari, O., Potter, M., Kuchroo, V. K., Barsh, G. S., Freeman, G. J., Umetsu, D. T., and DeKruyff, R. H. (2001). Identiﬁcation of Tapr (an airway hyperreactivity regulatory locus) and the linked Tim gene family. Nat. Immunol. 2, 1109–1116. Medzhitov, R. (2001). Toll-like receptors and innate immunity. Nat. Rev. Immunol. 1, 135–145. Meisel, C., Bonhagen, K., Lohning, M., Coyle, A. J., Gutierrez-Ramos, J. C., Radbruch, A., and Kamradt, T. (2001). Regulation and function of T1/ST2 expression on CD4þ T cells: Induction of type 2 cytokine production by T1/ST2 cross-linking. J. Immunol. 166, 3143–3150. Messi, M., Giacchetto, I., Nagata, K., Lanzavecchia, A., Natoli, G., and Sallusto, F. (2003). Memory and ﬂexibility of cytokine gene expression as separable properties of human T(H)1 and T(H)2 lymphocytes. Nat. Immunol. 4, 78–86. Mohrs, M., Blankespoor, C. M., Wang, Z. E., Loots, G. G., Afzal, V., Hadeiba, H., Shinkai, K., Rubin, E. M., and Locksley, R. M. (2001a). Deletion of a coordinate regulator of type 2 cytokine expression in mice. Nat. Immunol. 2, 842–847. Mohrs, M., Lacy, D. A., and Locksley, R. M. (2003). Stat signals release activated naive th cells from an anergic checkpoint. J. Immunol. 170, 1870–1876. Mohrs, M., Shinkai, K., Mohrs, K., and Locksley, R. M. (2001b). Analysis of type 2 immunity in vivo with a bicistronic IL-4 reporter. Immunity 15, 303–311. Moqbel, R., Ying, S., Barkans, J., Newman, T. M., Kimmitt, P., Wakelin, M., Taborda-Barata, L., Meng, Q., Corrigan, C. J., Durham, S. R., et al. (1995). Identiﬁcation of messenger RNA for IL-4 in human eosinophils with granule localization and release of the translated product. J. Immunol. 155, 4939–4947. Morra, M., Howie, D., Grande, M. S., Sayos, J., Wang, N., Wu, C., Engel, P., and Terhorst, C. (2001). X-linked lymphoproliferative disease: A progressive immunodeﬁciency. Annu. Rev. Immunol. 19, 657–682. Mosmann, T. R., and Coffman, R. L. (1989). TH1 and TH2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7, 145–173.

186

DANIEL B. STETSON ET AL.

Murphy, E., Shibuya, K., Hosken, N., Openshaw, P., Maino, V., Davis, K., Murphy, K., and O’Garra, A. (1996). Reversibility of T helper 1 and 2 populations is lost after long-term stimulation. J. Exp. Med. 183, 901–913. Murphy, K. M., and Reiner, S. L. (2002). The lineage decisions of helper T cells. Nat. Rev. Immunol. 2, 933–944. Nagai, S., Hashimoto, S., Yamashita, T., Toyoda, N., Satoh, T., Suzuki, T., and Matsushima, K. (2001). Comprehensive gene expression proﬁle of human activated T(h)1- and T(h)2-polarized cells. Int. Immunol. 13, 367–376. Narlikar, G. J., Phelan, M. L., and Kingston, R. E. (2001). Generation and interconversion of multiple distinct nucleosomal states as a mechanism for catalyzing chromatin ﬂuidity. Mol. Cell 8, 1219–1230. Nieuwenhuis, E. E., Neurath, M. F., Corazza, N., Iijima, H., Trgovcich, J., Wirtz, S., Glickman, J., Bailey, D., Yoshida, M., Galle, P. R., et al. (2002). Disruption of T helper 2-immune responses in Epstein-Barr virus-induced gene 3-deﬁcient mice. Proc. Natl. Acad. Sci. USA 99, 16951–16956. Ouyang, W., Lohning, M., Gao, Z., Assenmacher, M., Ranganath, S., Radbruch, A., and Murphy, K. M. (2000). Stat6-independent GATA-3 autoactivation directs IL-4-independent Th2 development and commitment. Immunity 12, 27–37. Ozaki, K., Spolski, R., Feng, C. G., Qi, C. F., Cheng, J., Sher, A., Morse, H. C., 3rd, Liu, C., Schwartzberg, P. L., and Leonard, W. J. (2002). A critical role for IL-21 in regulating immunoglobulin production. Science 298, 1630–1634. Patterson, S. (2000). Flexibility and cooperation among dendritic cells. Nat. Immunol. 1, 273–274. Plaut, M., Pierce, J. H., Watson, C. J., Hanley-Hyde, J., Nordan, R. P., and Paul, W. E. (1989). Mast cell lines produce lymphokines in response to cross-linkage of Fc epsilon RI or to calcium ionophores. Nature 339, 64–67. Ranger, A. M., Oukka, M., Rengarajan, J., and Glimcher, L. H. (1998). Inhibitory function of two NFAT family members in lymphoid homeostasis and Th2 development. Immunity 9, 627–635. Rankin, J. A., Picarella, D. E., Geba, G. P., Temann, U. A., Prasad, B., DiCosmo, B., Tarallo, A., Stripp, B., Whitsett, J., and Flavell, R. A. (1996). Phenotypic and physiologic characterization of transgenic mice expressing interleukin 4 in the lung: Lymphocytic and eosinophilic inﬂammation without airway hyperreactivity. Proc. Natl. Acad. Sci. USA 93, 7821–7825. Reis e Sousa, C. (2001). Dendritic cells as sensors of infection. Immunity 14, 495–498. Relloso, M., Puig-Kroger, A., Pello, O. M., Rodriguez-Fernandez, J. L., de la Rosa, G., Longo, N., Navarro, J., Munoz-Fernandez, M. A., Sanchez-Mateos, P., and Corbi, A. L. (2002). DC-SIGN (CD209) expression is IL-4 dependent and is negatively regulated by IFN, TGF-beta, and antiinﬂammatory agents. J. Immunol. 168, 2634–2643. Rengarajan, J., Mowen, K. A., McBride, K. D., Smith, E. D., Singh, H., and Glimcher, L. H. (2002a). Interferon regulatory factor 4 (IRF4) interacts with NFATc2 to modulate interleukin 4 gene expression. J. Exp. Med. 195, 1003–1012. Rengarajan, J., Tang, B., and Glimcher, L. H. (2002b). NFATc2 and NFATc3 regulate T(H)2 differentiation and modulate TCR-responsiveness of naive T(H)cells. Nat. Immunol. 3, 48–54. Riviere, I., Sunshine, M. J., and Littman, D. R. (1998). Regulation of IL-4 expression by activation of individual alleles. Immunity 9, 217–228. Rogge, L., Bianchi, E., Bifﬁ, M., Bono, E., Chang, S. Y., Alexander, H., Santini, C., Ferrari, G., Sinigaglia, L., Seiler, M., et al. (2000). Transcript imaging of the development of human T helper cells using oligonucleotide arrays. Nat. Genet. 25, 96–101. Romagnani, S., Maggi, E., and Del Prete, G. (1994). An alternative view of the Th1/Th2 switch hypothesis in HIV infection. AIDS Res. Hum. Retroviruses 10, iii–ix.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

187

Rumbley, C. A., Sugaya, H., Zekavat, S. A., El Refaei, M., Perrin, P. J., and Phillips, S. M. (1999). Activated eosinophils are the major source of Th2-associated cytokines in the schistosome granuloma. J. Immunol. 162, 1003–1009. Sad, S., Marcotte, R., and Mosmann, T. R. (1995). Cytokine-induced differentiation of precursor mouse CD8þ T cells into cytotoxic CD8þ T cells secreting Th1 or Th2 cytokines. Immunity 2, 271–279. Sallusto, F., and Lanzavecchia, A. (2001). Exploring pathways for memory T cell generation. J. Clin. Invest. 108, 805–806. Sallusto, F., Lenig, D., Fo¨rster, R., Lipp, M., and Lanzavecchia, A. (1999). Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. [see comments] Nature 401, 708–712. Santangelo, S., Cousins, D. J., Winkelmann, N. E., and Staynov, D. Z. (2002). DNA methylation changes at human Th2 cytokine genes coincide with DNase I hypersensitive site formation during CD4(þ) T cell differentiation. J. Immunol. 169, 1893–1903. Schwaiger, S., Wolf, A. M., Robatscher, P., Jenewein, B., and Grubeck-Loebenstein, B. (2003). IL-4-producing CD8þ T cells with a CD62Lþþ(bright) phenotype accumulate in a subgroup of older adults and are associated with the maintenance of intact humoral immunity in old age. J. Immunol. 170, 613–619. Seder, R. A., Paul, W. E., Ben-Sasson, S. Z., LeGros, G. S., Kagey-Sobotka, A., Finkelman, F. D., Pierce, J. H., and Plaut, M. (1991). Production of interleukin-4 and other cytokines following stimulation of mast cell lines and in vivo mast cells/basophils. Int. Arch. Allergy Appl. Immunol. 94, 137–140. Seki, Y., Hayashi, K., Matsumoto, A., Seki, N., Tsukada, J., Ransom, J., Naka, T., Kishimoto, T., Yoshimura, A., and Kubo, M. (2002). Expression of the suppressor of cytokine signaling-5 (SOCS5) negatively regulates IL-4-dependent STAT6 activation and Th2 differentiation. Proc. Natl. Acad. Sci. USA 99, 13003–13008. Seto, Y., Nakajima, H., Suto, A., Shimoda, K., Saito, Y., Nakayama, K. I., and Iwamoto, I. (2003). Enhanced Th2 cell-mediated allergic inﬂammation in Tyk2-deﬁcient mice. J. Immunol. 170, 1077–1083. Sherman, M. A. (2001). The role of STAT6 in mast cell IL-4 production. Immunol. Rev. 179, 48–56. Shinkai, K., Mohrs, M., and Locksley, R. M. (2002). Helper T cells regulate type-2 innate immunity in vivo. Nature 420, 825–829. Smale, S. T., and Fisher, A. G. (2002). Chromatin structure and gene regulation in the immune system. Annu. Rev. Immunol. 20, 427–462. Solymar, D. C., Agarwal, S., Bassing, C. H., Alt, F. W., and Rao, A. (2002). A 30 enhancer in the IL-4 gene regulates cytokine production by Th2 cells and mast cells. Immunity 17, 41–50. Sommers, C. L., Park, C. S., Lee, J., Feng, C., Fuller, C. L., Grinberg, A., Hildebrand, J. A., Lacana, E., Menon, R. K., Shores, E. W., et al. (2002). A LAT mutation that inhibits T cell development yet induces lymphoproliferation. Science 296, 2040–2043. Soumelis, V., Reche, P. A., Kanzler, H., Yuan, W., Edward, G., Homey, B., Gilliet, M., Ho, S., Antonenko, S., Lauerma, A., et al. (2002). Human epithelial cells trigger dendritic cell mediated allergic inﬂammation by producing TSLP. Nat. Immunol. 3, 673–680. Takemoto, N., Arai, K., and Miyatake, S. (2002). Cutting edge: The differential involvement of the N-ﬁnger of GATA-3 in chromatin remodeling and transactivation during Th2 development. J. Immunol. 169, 4103–4107. Takemoto, N., Koyano-Nakagawa, N., Yokota, T., Arai, N., Miyatake, S., and Arai, K. (1998). Th2-speciﬁc DNase I-hypersensitive sites in the murine IL-13 and IL-4 intergenic region. Int. Immunol. 10, 1981–1985.

188

DANIEL B. STETSON ET AL.

Tamura, T., Igarashi, O., Hino, A., Yamane, H., Aizawa, S., Kato, T., and Nariuchi, H. (2001). Impairment in the expression and activity of Fyn during differentiation of naive CD4þ T cells into the Th2 subset. J. Immunol. 167, 1962–1969. Temann, U. A., Ray, P., and Flavell, R. A. (2002). Pulmonary overexpression of IL-9 induces Th2 cytokine expression, leading to immune pathology. J. Clin. Invest. 109, 29–39. Tentori, L., Pardoll, D. M., Zuniga, J. C., Hu-Li, J., Paul, W. E., Bluestone, J. A., and Kruisbeek, A. M. (1988). Proliferation and production of IL-2 and B cell stimulatory factor 1/IL-4 in early fetal thymocytes by activation through Thy-1 and CD3. J. Immunol. 140, 1089–1094. Urban, J. F., Jr., Noben-Trauth, N., Donaldson, D. D., Madden, K. B., Morris, S. C., Collins, M., and Finkelman, F. D. (1998). IL-13, IL-4Ralpha, and Stat6 are required for the expulsion of the gastrointestinal nematode parasite Nippostrongylus brasiliensis. Immunity 8, 255–264. Valapour, M., Guo, J., Schroeder, J. T., Keen, J., Cianferoni, A., Casolaro, V., and Georas, S. N. (2002). Histone deacetylation inhibits IL4 gene expression in T cells. J. Allergy Clin. Immunol. 109, 238–245. Wagers, A. J., Waters, C. M., Stoolman, L. M., and Kansas, G. S. (1998). Interleukin 12 and interleukin 4 control T cell adhesion to endothelial selectins through opposite effects on alpha1, 3-fucosyltransferase VII gene expression. J. Exp. Med. 188, 2225–2231. Walley, A. J., Chavanas, S., Moffatt, M. F., Esnouf, R. M., Ubhi, B., Lawrence, R., Wong, K., Abecasis, G. R., Jones, E. Y., Harper, J. I., et al. (2001). Gene polymorphism in Netherton and common atopic disease. Nat. Genet. 29, 175–178. Wang, X., and Mosmann, T. (2001). In vivo priming of CD4 T cells that produce interleukin (IL)-2 but not IL-4 or interferon (IFN)-gamma, and can subsequently differentiate into IL-4- or IFNgamma-secreting cells. J. Exp. Med. 194, 1069–1080. Weiss, D. L., and Brown, M. A. (2001). Regulation of IL-4 production in mast cells: A paradigm for cell-type-speciﬁc gene expression. Immunol. Rev. 179, 35–47. Wherry, E. J., Teichgraber, V., Becker, T. C., Masopust, D., Kaech, S. M., Antia, R., Von Andrian, U. H., and Ahmed, R. (2003). Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat. Immunol. 4, 225–234. Williams, C. M., and Galli, S. J. (2000). The diverse potential effector and immunoregulatory roles of mast cells in allergic disease. J. Allergy Clin. Immunol. 105, 847–859. Wu, C., Nguyen, K. B., Pien, G. C., Wang, N., Gullo, C., Howie, D., Sosa, M. R., Edwards, M. J., Borrow, P., Satoskar, A. R., et al. (2001). SAP controls T cell responses to virus and terminal differentiation of TH2 cells. Nat. Immunol. 2, 410–414. Wu, C. Y., Kirman, J. R., Rotte, M. J., Davey, D. F., Perfetto, S. P., Rhee, E. G., Freidag, B. L., Hill, B. J., Douek, D. C., and Seder, R. A. (2002). Distinct lineages of T(H)1 cells have differential capacities for memory cell generation in vivo. Nat. Immunol. 3, 852–858. Wurster, A. L., Withers, D. J., Uchida, T., White, M. F., and Grusby, M. J. (2002). Stat6 and IRS-2 cooperate in interleukin 4 (IL-4)-induced proliferation and differentiation but are dispensable for IL-4-dependent rescue from apoptosis. Mol. Cell. Biol. 22, 117–126. Yamashita, M., Hashimoto, K., Kimura, M., Kubo, M., Tada, T., and Nakayama, T. (1998). Requirement for p56(lck) tyrosine kinase activation in Th subset differentiation. Int. Immunol. 10, 577–591. Yamashita, M., Ukai-Tadenuma, M., Kimura, M., Omori, M., Inami, M., Taniguchi, M., and Nakayama, T. (2002). Identiﬁcation of a conserved GATA3 response element upstream proximal from the interleukin-13 gene locus. J. Biol. Chem. 277, 42399–42408. Yen, C. J., Lin, S. L., Huang, K. T., and Lin, R. H. (2000). Age-associated changes in interferongamma and interleukin-4 secretion by puriﬁed human CD4þ and CD8þ T cells. J. Biomed. Sci. 7, 317–321.

TH2 CELLS: ORCHESTRATING BARRIER IMMUNITY

189

Yoshida, H., Nishina, H., Takimoto, H., Marengere, L. E., Wakeham, A. C., Bouchard, D., Kong, Y. Y., Ohteki, T., Shahinian, A., Bachmann, M., et al. (1998). The transcription factor NF-ATc1 regulates lymphocyte proliferation and Th2 cytokine production. Immunity 8, 115–124. Zhang, D. H., Cohn, L., Ray, P., Bottomly, K., and Ray, A. (1997). Transcription factor GATA-3 is differentially expressed in murine Th1 and Th2 cells and controls Th2-speciﬁc expression of the interleukin-5 gene. J. Biol. Chem. 272, 21597–21603. Zhao, H., Yan, M., Wang, H., Erickson, S., Grewal, I. S., and Dixit, V. M. (2001). Impaired c-Jun amino terminal kinase activity and T cell differentiation in death receptor 6-deﬁcient mice. J. Exp. Med. 194, 1441–1448. Zhao, K., Wang, W., Rando, O. J., Xue, Y., Swiderek, K., Kuo, A., and Crabtree, G. R. (1998). Rapid and phosphoinositol-dependent binding of the SWI/SNF-like BAF complex to chromatin after T lymphocyte receptor signaling. Cell 95, 625–636. Zheng, W., and Flavell, R. A. (1997). The transcription factor GATA-3 is necessary and sufﬁcient for Th2 cytokine gene expression in CD4 T cells. Cell 89, 587–596. Zhu, J., Guo, L., Min, B., Watson, C. J., Hu-Li, J., Young, H. A., Tsichlis, P. N., and Paul, W. E. (2002). Growth factor independent-1 induced by IL-4 regulates Th2 cell proliferation. Immunity 16, 733–744. Zhu, Z., Homer, R. J., Wang, Z., Chen, Q., Geba, G. P., Wang, J., Zhang, Y., and Elias, J. A. (1999). Pulmonary expression of interleukin-13 causes inﬂammation, mucus hypersecretion, subepithelial ﬁbrosis, physiologic abnormalities, and eotaxin production. J. Clin. Invest. 103, 779–788. Zuany-Amorim, C., Rufﬁe, C., Haile, S., Vargaftig, B. B., Pereira, P., and Pretolani, M. (1998). Requirement for gammadelta T cells in allergic airway inﬂammation. Science 280, 1265–1267.

This Page Intentionally Left Blank

advances in immunology, vol. 83

Generation, Maintenance, and Function of Memory T Cells PATRICK R. BURKETT, RIMA KOKA, MARCIA CHIEN, DAVID L. BOONE, AND AVERIL MA Department of Medicine and the Ben May Institute for Cancer Research University of Chicago Chicago, IL 60637

I. Introduction

Understanding the cellular and molecular mechanisms that render a vertebrate host refractory to reinfection by a previously encountered pathogen and how this refractory state is both induced and maintained are central immunological questions. Clearly, immunity to pathogens is a complex trait, involving the interactions of many cell types and diverse arrays of effector molecules. The key feature of long-lived immunity is the elicitation of memory B and T cells, which, respectively, produce protective neutralizing antibodies (humoral immunity) or mount rapid recall responses (cellular immunity). The importance of humoral immunity in protection from reinfection was recognized early in the twentieth century and is the basis of protection provided by most current vaccines (Ahmed et al., 2002; Zinkernagel, 2000). By contrast, both the mechanism and relative importance of cellular immunity have only recently been fully appreciated. The development of new techniques that allow precise quantiﬁcation of antigen-speciﬁc T cells has allowed unprecedented analysis of the dynamics of T cell responses to antigen in vivo. In turn, this understanding has greatly facilitated dissection of the speciﬁc cellular and molecular mechanisms by which memory T cells are generated and maintained. Unlike antibodies, T cells can only recognize their cognate antigens in the context of MHC. Therefore, T cell-mediated immunity can never completely prevent reinfection (whether pathological or not) as can high afﬁnity neutralizing antibodies. However, many diseases for which no vaccines (or only poorly effective ones) exist can persist in chronic, semi-controlled states for years. Hypothetically, induction of efﬁcient memory T cells could help the host to maintain control of such infections, preventing disease-induced morbidity and mortality. Thus, the understanding of how memory T cells can be efﬁciently induced and maintained in a functionally protective state may be critical for the future development of vaccines. This chapter examines current theories about how memory T cells are generated and maintained, and how they protect against reinfection. Additionally, while this chapter focuses on memory CD8þ T cells, some important similarities and differences between CD4þ and CD8þ T cell memory will be highlighted. 191 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

192

PATRICK R. BURKETT ET AL.

II. What Is a Memory Cell?

A. Markers and Memory Strictly speaking, memory lymphocytes have recognized and responded to their cognate antigen at some point in the past. This deﬁnition emphasizes that in order to study memory cells, one needs to be able to identify them. The development of MHC Class I tetramers and the ability to observe antigeninduced cytokine expression in single cells has vastly improved researchers’ ability to track antigen-speciﬁc cells in vivo and has greatly facilitated the study of memory T cells of known speciﬁcity (Klenerman et al., 2002). Memory CD8þ T cells also uniquely express a pattern of surface markers, which provides a phenotypic means of separating memory CD8þ T cells from naive and effector CD8þ T cells (Table I). These include molecules involved in trafﬁcking and adhesion (CD44, CD11a, CD62L), cytokine and chemokine receptors (IL-2/15Rb, CCR7), NK cell receptors (Ly49 and NKG2 family molecules), and altered glycosylation patterns on a variety of surface glycoproteins (CD43 and CD45) (Galvan et al., 1998; Harrington et al., 2000; Jamieson et al., 2002; McMahon et al., 2002; Slifka and Whitton, 2000; Slifka et al., 2000). Thus, memory T cells are broadly deﬁned to be CD8þ T cells of a deﬁned speciﬁcity that are known to have encountered their cognate antigen at some prior time and express a deﬁned pattern of surface markers. Intriguingly, naive mice raised in a speciﬁc pathogen-free environment contain a population of CD8þ T cells that exhibit an expression pattern of surface markers, most notably, CD44 and IL-2/15Rb, similar to that of memory CD8þ T cells of known speciﬁcity. These cells, often referred to as

TABLE I Markers of CD8þ T Cell Differentiationa Memory CD8þ T cell Marker

Naive CD8þ T cell

Effector CD8þ T cell

Effector

Central

CD11a CD25 CD44 CD62L CD69 IL-2/15Rb Ly6C CCR7

Lo Lo Lo Hi Lo Lo Lo Hi

Hi Hi Hi Lo Hi Hi Hi Lo

Hi Lo Hi Lo Lo Hi Hi Lo

Hi Lo Hi Hi Lo Hi Lo Hi

a Eight of the most commonly used markers for differentiating naive, effector, and both central and effector subsets of memory CD8þ T cells, and the relative expression of each marker on each stage of CD8þ T cell differentiation are indicated.

MEMORY T CELLS

193

memory-phenotype T cells, are assumed to arise from immune responses against environmental antigens and are commonly used as surrogates for the study of true memory CD8þ T cells. Memory phenotype CD8þ T cells accumulate with age and exhibit many of the same functional and phenotypic characteristics of true, antigen-experienced memory CD8þ T cells, such as the ability to rapidly elaborate effector cytokines upon stimulation (Sprent, 2002; Zhang et al., 2002). However, an important caveat to the use of memory phenotype CD8þ T cells as surrogates for true memory CD8þ T cells is that the exact origins of memory phenotype CD8þ T cells are far from clear. Indeed, naive T cells proliferate upon transfer into lymphopenic hosts, such as RAG-deﬁcient or sublethally irradiated mice (Goldrath and Bevan, 1999). Naive CD8þ T cells undergoing lymphopenia-induced proliferation acquire the phenotype of memory CD8þ T cells, including elevated expression of CD44 and the ability to rapidly produce interferon (IFN)-g (Cho et al., 2000; Goldrath et al., 2000; Murali-Krishna and Ahmed, 2000). Naive CD8þ T cells appear largely dependent on both IL-7 and MHC Class I in order to proliferate in lymphopenic hosts, while naive CD4þ T cells require IL-7 but not MHC Class II (Clarke and Rudensky, 2000; Goldrath and Bevan, 1999; Tan et al., 2001). Since both IL-7 and MHC are constitutively present, it is possible that lymphopenia-induced proliferation may largely be the result of increased availability of MHC and IL-7, because in a lymphopenic host there would be decreased competition among naive T cells for these molecules (Fry and Mackall, 2001; Surh and Sprent, 2000). Interestingly, a 2003 report found that neonatal mice support the lymphopenia-induced proliferation of naive CD4þ T cells, suggesting that physiologically this process may normally occur neonatally (Min et al., 2003). Additionally, other groups have suggested that at least some CD44Hi memory phenotype CD8þ T cells resemble extrathymically derived T cells (Yamada et al., 2001). Thus, memory phenotype T cells may actually be a mixed population, containing true memory cells that have responded to environmental antigens, lymphopenia-induced memory cells, and extrathymically derived cells. Therefore, the conclusions of studies that examine only memory phenotype T cells may not be entirely applicable to true antigen-experienced memory T cells. B. Functions and Properties Exposure of an organism to antigen induces memory T cells that are both quantitatively and qualitatively distinct from their naive predecessors. Quantitatively, the frequency of memory CD8þ T cells speciﬁc for a given peptide/MHC complex is drastically increased in immune versus naive mice (Murali-Krishna et al., 1998). This increased frequency likely plays an important role in the increased speed of recall responses, for mice with decreased numbers of memory CD8þ T cells show decreased recall responses (Borrow

194

PATRICK R. BURKETT ET AL.

et al., 1996; Khan et al., 2002). Additionally, the differential expression pattern of surface markers on memory T cells gives some insight into their distinct functional properties. For instance, while naive T cells are largely restricted to secondary lymphoid tissues, both effector and memory T cells can readily migrate to nonlymphoid tissues, and memory T cells can be maintained in nonlymphoid tissues for extended periods (Marshall et al., 2001; Masopust et al., 2001; Reinhardt et al., 2001; Weninger et al., 2001). This latter point is reﬂected in differences between naive and memory T cell expression of both trafﬁcking and adhesion molecules, such as CD62L and CCR7. Thus, memory T cells are increased in number and present in both lymphoid and nonlymphoid tissues, both of which likely facilitate the rapid recognition and elimination of foreign antigen. The fact that memory CD8þ T cells are present in increased frequency does not solely explain their protective capacity, for they also have qualitatively different proliferative and effector responses to antigen compared to both naive and effector CD8þ T cells of similar speciﬁcity. Upon re-encountering antigen, memory CD8þ T cells can enter cell cycle more rapidly and once they begin cycling, they divide more rapidly than do naive cells (Cho et al., 1999; Veiga-Fernandes et al., 2000). Moreover, in contrast to effector CTLs that show poor proliferation upon restimulation, memory CD8þ T cells can undergo substantial expansion in vivo. The elevated proliferative potential of memory CD8þ T cells correlates with increased TCR-induced Erk1/2 phosphorylation compared to effector CTLs (Kaech et al., 2002a). While both effector and memory CD8þ T cells are capable of rapidly producing effector cytokines, such as IFN-g and TNFa, a given memory CD8þ T cell simultaneously produces both cytokines (Slifka and Whitton, 2000; Slifka et al., 1999). Although the TCR does not undergo afﬁnity maturation in the same way that the BCR does, memory CD8þ T cells can produce similar levels of effector cytokines and promote similar lysis of target cells at lower concentrations of peptide than can T cells earlier in an immune response. The increased sensitivity of memory CD8þ T cells has been observed even in monoclonal T cell populations, further indicating that the functional avidity of memory CD8þ T cells can increase without selecting high-afﬁnity TCRs (Bachmann et al., 1999; Slifka and Whitton, 2001). Memory CD8þ T cells’ increased sensitivity to antigen correlates with increased intracellular levels of lck, suggesting a mechanism for the increased sensitivity of memory CD8þ T cells to TCR-mediated signals. While memory CD8þ T cell production of some effector cytokines requires TCR-induced transcription, memory phenotype T cells can rapidly elaborate the chemokine RANTES independently of transcription (Slifka et al., 1999; Swanson et al., 2002; Walzer et al., 2003). Thus, memory T cells utilize a number of molecular mechanisms for rapid cytokine production, both transcription-dependent and independent. Finally, memory CD8þ T cells

MEMORY T CELLS

195

decline in number far more slowly after mounting a secondary immune response than naive CD8þ T cells do after a primary immune response, possibly due to constitutively higher expression of bcl-2 in memory CD8þ T cells (Grayson et al., 2000, 2002). Thus, memory CD8þ T cells can detect lower concentrations of their cognate antigen, produce a broader array of cytokines on a per-cell basis, expand far more rapidly, and survive for longer periods than can either naive or effector CD8þ T cells. C. Bystander and Heterologous Responses Memory CD8þ T cells not only respond rapidly and potently to their cognate peptide/MHC, they may also provide some degree of protection against heterologous stimuli. Memory T cells continuously proliferate, and this basal proliferation is thought to play a role in their long-term maintenance (Tough and Sprent, 1994). Injection of Toll-like receptor ligands (TLRLs), such as lipopolysaccharide or dsRNA, or innate immune cytokines, such as IFN-a/b, transiently increases the rate of proliferation of memory phenotype T cells, particularly memory phenotype CD8þ T cells (Tough and Sprent, 1994; Tough et al., 1996, 1997). TLRL-induced proliferation, commonly termed bystander proliferation, can occur in the absence of MHC class I, but depends on the induction of one of two cytokine cascades (Eberl et al., 2000; Tough et al., 1996, 1997, 2001; Zhang et al., 1998). The ﬁrst cascade, which is activated by the dsRNA mimic poly inosine: cytosine, is dependent strictly on IFN-a/b (Tough et al., 1996; Zhang et al., 1998). The second bystander proliferation pathway can be activated by the NK T cell ligand a-galactosylceramide and depends on the initial production of IL-12 and IL-18, which together can induce IFN-g (Berg et al., 2002; Eberl et al., 2000; Tough et al., 2001). Both pathways appear to require IL-15 to actually initiate CD8þ T cell proliferation, for it alone can drive CD44Hi CD8þ T cell proliferation both in vivo and in vitro (Zhang et al., 1998). Regardless of the precise pathway activated, antigen-independent TLRL-driven proliferation of memory CD8þ T cells causes these cells to undergo only a few cell divisions and does not signiﬁcantly increase the number of these cells. The lack of expansion is likely due to simultaneous induction of apoptosis in memory phenotype T cells (McNally et al., 2001). Additionally, infection by heterologous viruses can also induce the proliferation of preexisting antigen-speciﬁc memory CD8þ T cells (Kim et al., 2002). Thus, memory CD8þ T cells can be rapidly activated by nonspeciﬁc inﬂammatory stimuli. Memory CD8þ T cells can be recruited to peripheral tissues and mount protective responses after challenge with heterologous pathogens or TLRLs. Recruitment to peripheral tissues appears largely independent of the speciﬁcity of the memory CD8þ T cells, and can occur in the absence of proliferation (Ely et al., 2003; Topham et al., 2001). While the underlying molecular

196

PATRICK R. BURKETT ET AL.

mechanisms of infection-induced memory CD8þ T cell recruitment to peripheral tissues are unclear, these heterologous responses may be protective. For instance, TLRLs stimulate the rapid, antigen-independent secretion of IFN-g by memory phenotype CD8þ T cells. Antigen-independent secretion of IFN-g by memory CD8þ T cells is dependent on both IL-12 and IL-18, and may play an important protective role in early responses to intracellular bacteria such as Burkholderia pseudomallei or Listeria monocytogenes (Berg et al., 2002; Lertmemongkolchai et al., 2001; Yajima et al., 2001). Additionally, LCMVprimed mice can clear a normally lethal vaccinia virus challenge, and this protection correlates with the presence of IFN-g-producing LCMV-speciﬁc CD8þ T cells early in the vaccinia infection (Chen et al., 2001). Intriguingly, heterologous activation can substantially alter the memory CD8þ T cell repertoire. After infection with LCMV, memory CD8þ T cells speciﬁc for distinct LCMV peptides are maintained in a hierarchy of epitope frequencies. While TLRL-induced proliferation does not alter these relative frequencies, heterologous infection with either vaccinia virus or pichinde virus substantially alters the previous hierarchy (Chen et al., 2001; Kim et al., 2002). This suggests that the activation of memory CD8þ T cells by heterologous infections may utilize mechanisms other than the induction of inﬂammatory cytokines, and may involve some degree of TCR cross-reactivity. Thus, memory T cells not only exhibit improved responses to their cognate antigen, but they can also rapidly mount protective responses to heterologous or nonspeciﬁc inﬂammatory stimuli. D. Memory T Cell Subsets Memory T cells of a given antigen speciﬁcity have typically been considered a homogenous population. However, this assumption has recently been called into question. Human and mouse peripheral blood CD4þ and CD8þ memory T cells can be divided into two subsets based on their differential expression of CCR7 and CD62L, both of which regulate T cell homing to secondary lymphoid tissues (Sallusto et al., 1999; Wherry et al., 2003). These two subsets not only differ in their expression of homing molecules, but also in their responses to in vitro stimulation. Both subsets proliferate upon restimulation in vitro but CCR7 CD62L cells (termed effector memory cells) rapidly produce effector cytokines, whereas CCR7þ CD62Lþ cells (called central memory cells) do not (Sallusto et al., 1999). Similarly, antigen-speciﬁc memory T cells isolated from murine nonlymphoid tissues preferentially exhibit ex vivo effector ability (such as IFN-g production or CTL activity), whereas those isolated from secondary lymphoid tissues lack such activity and preferentially produce IL-2 (Masopust et al., 2001; Reinhardt et al., 2001). The respective role of these two memory CD8þ T cell subsets in mediating protection from reinfection and their lineage relationship has only recently begun to be examined. It has been suggested that central memory T cells

MEMORY T CELLS

197

represent incompletely differentiated memory CD8þ T cells, whereas effector memory CD8þ T cells are more differentiated and protective (Champagne et al., 2001; Lanzavecchia and Sallusto, 2002; Sallusto et al., 1999). In contrast to the these ﬁndings, Wherry et al. (2003) found that adoptive transfer of central memory T cells into a naive host resulted in more rapid viral clearance, to both systemic and local infections, than was seen when effector memory CD8þ T cells were transferred. Surprisingly, in several of the models of infection tested, effector memory CD8þ T cells gave no better protection than did naive CD8þ T cells. The increased ability of central memory CD8þ T cells to mediate rapid viral clearance correlated with increased proliferative potential of central memory CD8þ T cells compared to effector memory CD8þ T cells. Additionally, while central memory CD8þ T cells maintain their central memory phenotype upon adoptive transfer into uninfected hosts, effector memory CD8þ T cells transition to a central memory phenotype, and this transition does not require proliferation (Wherry et al., 2003). These data indicate that the increased proliferative potential of central memory CD8þ T cells is crucial for rapid viral clearance in immune mice, and that effector memory CD8þ T cells can slowly differentiate into central memory CD8þ T cells. While these ﬁndings provide signiﬁcant insights into the differentiation of memory CD8þ T cells, multiple questions remain. For instance, if effector memory CD8þ T cells do not signiﬁcantly mediate protection upon reinfection, what is their function? The failure of effector memory CD8þ T cells to provide signiﬁcant protective immunity may be due to constraints imposed by adoptive transfer of puriﬁed populations. Wherry et al. (2003) transferred 75,000 to 300,000 central or effector memory CD8þ T cells, resulting in the recipients having a smaller memory CD8þ T cell pool than would be found in an LCMV-immune mouse. Additionally, Wherry et al. (2003) found that neither central nor effector memory T cells had lytic activity ex vivo, whereas other reports have found that memory CD8þ T cells in nonlymphoid tissues (which are predominantly effector memory T cells) are cytotoxic ex vivo (Masopust et al., 2001). If viral clearance requires a certain frequency of cytolytically active CD8þ T cells, then the efﬁciency of protection of a limited number of adoptively transferred noncytolytic memory CD8þ T cells may depend upon the potential of the transferred population to expand upon challenge. In contrast, a sizable population of cytolytically active effector memory CD8þ T cells might be able to mediate rapid viral clearance without requiring signiﬁcant expansion. A recent paper examining protective immunity following immunization of mice with either live or heat-killed L. monocytogenes (HKLM) supports this latter hypothesis. Vaccination of mice with HKLM induced a memory CD8þ T cell pool that rapidly expanded upon subsequent infection with viable

198

PATRICK R. BURKETT ET AL.

L. monocytogenes (Lauvau et al., 2001). However, despite rapid expansion, HKLM-induced memory CD8þ T cells were unable to rapidly clear the second challenge with live L. monocytogenes, whereas mice that had been initially primed by infection with live L. monocytogenes both expanded and rapidly cleared the bacteria. Interestingly, CD8þ T cells primed with live L. monocytogenes down-regulated CD62L but those primed with HKLM maintained high levels of CD62L, indicating that HKLM priming may preferentially induce central memory CD8þ T cells (Lauvau et al., 2001). Moreover, this data is consistent with the idea that central memory CD8þ T cells provide a reservoir of cells that mediate much of the expansion in a secondary response. Finally, this observation indicates that effector memory CD8þ T cells can rapidly clear certain pathogens upon reinfection (Sprent, 2002). Clearly, further work needs to be done to clarify the respective roles of effector and central memory T cells in immunity to reinfection. The contrasting data of Wherry et al. (2003) and Lauvau et al. (2001) suggest that distinct subsets of memory CD8þ T cells may protect the host from reinfection in quite different ways. Additionally, the fact that challenge with the same pathogen, either live or heat-killed, or in differing doses, can result in marked changes in the subsets present in the memory pool and the protective capacity of that pool has important implications for vaccine design. Ideally, vaccines could bias the subset of memory CD8þ T cells induced so as to best protect the host from infection. Therefore, further work is clearly needed to understand both the factors that regulate the development of memory CD8þ T cells subsets and the optimal structure of a memory pool to protect the host from a given pathogen. III. Where Do Memory CD81 T Cells Come From?

A. The Pattern of T Cell Responses Mature peripheral CD8þ T cells can be placed into three stages of differentiation—naive, effector, and memory—which are functionally distinct and differ in their expression of a variety of surface markers (Table I). While it has long been established that naive CD8þ T cells become effector CTLs upon exposure to antigen, the molecular mechanisms that control differentiation, and the relationship of naive CD8þ T cells and effector CTLs to memory CD8þ T cells have been the subject of much debate. Over the past 10 years, a number of techniques and reagents have been developed that allow antigenspeciﬁc T cells to be readily identiﬁed, such as MHC tetramers or intracellular cytokine staining, thus allowing the dynamics of antigen-driven T cell responses to be examined in great detail. Studies using such techniques and reagents have begun to build a cohesive picture of CD8þ T cell differentiation. When examined graphically, the numbers of antigen-speciﬁc CD8þ T cells form a stereotypical curve (Fig. 1).

MEMORY T CELLS

199

Fig 1 The basic pattern of T cell immune responses. Each line represents the number of either a CD8þ (solid line) or a CD4þ (dashed line) T cells of a given antigen-speciﬁcity over time. The three broad periods of the immune response are indicated. It is currently unclear if memory CD4þ T cells are maintained (Whitmire et al., 1998) or slowly decay in number (Homann, Teyton and Oldstone, 2001).

This curve begins with a given number of naive antigen-speciﬁc CD8þ T cells. In a wildtype mouse, it has been estimated that there are approximately 100 to 200 naive CD8þ T cells speciﬁc for a given peptide/MHC complex (Blattman et al., 2002). Upon exposure to their cognate antigen/MHC complexes, these precursors rapidly expand, until, at the peak of the response, there can be upwards of 1–10 106 CD8þ T cells speciﬁc for a given peptide epitope in the spleen alone. The precise kinetics with which this peak is reached and the size (or amplitude) of this peak will vary depending on the exact experimental system. Typically, the peak of CD8þ T cell expansion occurs slightly after antigen clearance, while the amplitude of the response is proportional to the strength of the initial stimulus (Badovinac and Harty, 2002). The amplitude of this peak appears to be critical for determining the size of the ensuing memory pool (Hou et al., 1994; Murali-Krishna et al., 1998). Once this peak has been reached, the antigen-speciﬁc CD8þ T cells rapidly decline in number, and at least 90% of the responding T cells undergo programmed cell death. However, a signiﬁcant number of antigen-speciﬁc memory CD8þ T cells survive this purge, and this number is maintained at stable levels for the life of the animal (Ahmed et al., 2002; Murali-Krishna et al., 1998). Since T cell-based vaccines would likely require the induction of memory cells from a previously naive host, understanding the regulation of this expansion and contraction and how they relate to the subsequent formation of this memory pool is of great interest.

200

PATRICK R. BURKETT ET AL.

B. The Kinetics of the Response Prior to undergoing rapid antigen-driven expansion and differentiation, naive antigen-speciﬁc T cells must ﬁrst be recruited to the immune response. Several lines of evidence support the idea that the time frame in which CD8þ T cells may be recruited to a response is narrow. In vitro, only a brief initial encounter (2.5 hrs) with antigen appears to be required by CD8þ T cells in order to begin to divide. Once they have begun to divide, they will continue to do so, even in the absence of antigen, provided cytokines such as IL-2 are available (Kaech and Ahmed, 2001; van Stipdonk et al., 2001; Wong and Pamer, 2001). Additionally, if limiting quantities of antigen are added to TCR transgenic cells in vitro, only a fraction of the cells are recruited to undergo cell division, but those cells that are recruited will divide as many times as do cells in nonantigen limiting conditions, and will subsequently differentiate into effector CD8þ T cells (Kaech and Ahmed, 2001). In vivo, adoptive transfer of antigen speciﬁc T cells at various times after Plasmodium yoelii infection showed that the T cells had to be transferred within 24 hours of infection in order to become activated and undergo differentiation (Hafalla et al., 2002). Although the factors that control the period of time in which naive T cells can be recruited to a response are largely undeﬁned, they could include competition of CD8þ T cells for peptide/ MHC complexes on antigen-presenting cells (APCs) (Kedl et al., 2002). The basic pattern of CD8þ T cell expansion, contraction, and memory is similar for a wide range of immunogens, but the exact kinetics of this process may vary. This observation suggests that CD8þ T cells are somehow programmed early during an infection to undergo this pattern with certain kinetics (Badovinac and Harty, 2002). What determines the duration of both the expansion and contraction once CD8þ T cells are recruited to a response? One possibility is the quantity of antigen and length of time for which antigen is present. While experimental determination of these two values is difﬁcult in vivo, two groups have approached this by examining the impact of antibiotic treatment of mice following infection with L. monocytogenes. Postinfection antibiotic treatment eliminates viable bacteria several days faster than normal and also substantially reduces the duration in which L. monocytogenes peptides are presented (Badovinac et al., 2002; Mercado et al., 2000). This correlates with a signiﬁcant reduction in the amplitude of the primary response. However, once a CD8þ T cell response to L. monocytogenes has begun, the kinetics of the response are not altered, even if the infection is rapidly contained by antibiotic treatment of the host mouse. Thus, once antigen has initiated a CD8þ T cell response, the responding cells will continue to expand for a given period of time, after which they will undergo contraction regardless of the continued presence of antigen, suggesting that other factors may regulate the kinetics of the response.

MEMORY T CELLS

201

C. The Size of the Response The kinetics of the CD8þ T cell response may be programmed in an as yet undeﬁned, and likely multifactorial, manner by the conditions of initial priming. However, many of the factors that control the amplitude of the initial expansion are relatively well deﬁned. A great deal of evidence has accumulated over the past several years suggesting that CD8þ T cell responses are initiated by professional APCs, particularly dendritic cells (DCs) (Banchereau and Steinman, 1998). A major reason for this is the increasing recognition that professional APCs, including DCs, are able to process exogenously acquired antigens for presentation in the MHC class I pathway (a process commonly termed cross-presentation) (den Haan and Bevan, 2001). Therefore, the APC does not need to be infected in order to prime CD8þ T cells. Moreover, the ability of DCs to migrate to secondary lymphoid tissues after detection of a TLRL could facilitate rapid initiation of a CD8þ T cell response to infections in such tissues. Finally, DCs preferentially express many accessory molecules that are known to bolster CD8þ T cell responses, including costimulatory molecules and cytokines. Thus, DCs provide not only the antigen, but also the proper environment of cytokines and costimulatory molecules for rapid and controlled expansion of CD8þ T cells. DCs are constantly surveyed by naive T cells, and even in the absence of antigen, DC/T cell interactions can induce biochemical signals that may help tune the relative responsiveness of the T cell (Kondo et al., 2001; Revy et al., 2001; Stefanova et al., 2002). Supporting the idea that DCs are important in priming CD8þ T cells, immunization with puriﬁed antigen-pulsed DCs is sufﬁcient to induce potent primary and protective memory responses (Hamilton and Harty, 2002; Livingstone and Kuhn, 2002; Ludewig et al., 1998, 1999). While it is possible that the inﬂammatory environment induced either by infection or adjuvants may allow nonprofessional APCs to efﬁciently initiate CD8þ T cell responses (Kundig et al., 1995), DC priming of CD8þ T cells may be critical for rapid induction of a CD8þ T cell response (Jung et al., 2002). The CD11c promoter, which is largely DC-speciﬁc, was used to drive transgenic expression of a diphtheria toxin receptor. Treatment of these transgenic mice with diphtheria toxin induced short-term, but selective, deletion of DCs. When CD8þ T cell responses to a variety of antigens or pathogens, including L. monocytogenes, were examined in DC-depleted mice, antigen-speciﬁc CD8þ T cells did not begin to divide as rapidly as did cells in mice with intact DCs (Jung et al., 2002). Although this study did not report whether pathogen clearance or subsequent effector CD8þ T cell responses were similar, it does highlight the importance of DCs for the rapid and efﬁcient initiation of CD8þ T cell responses to antigen. In addition to efﬁciently presenting antigen to T cells, DCs are also important for initiating T cell responses due to their ability to express a number of

202

PATRICK R. BURKETT ET AL.

secreted and surface-bound factors that can modulate the amplitude of CD8þ T cell expansion. These factors may be broadly classiﬁed as either costimulatory molecules or cytokines. Because the impact of costimulatory molecules on T cell responses has been reviewed in detail elsewhere, their roles in regulating the amplitude of CD8þ T cell responses will be discussed only brieﬂy here (Carreno and Collins, 2002; Whitmire and Ahmed, 2000). The majority of costimulatory molecules belong to either the tumor necrosis factor superfamily (TNFSF) or the B7 family. While some costimulatory receptor/ligand pairs modulate both CD4þ and CD8þ T cell responses (B7/CD28), others preferentially modulate the responses of either CD4þ (such as ICOS/B7RP or OX40/ OX40L) or CD8þ (such as 4-1BB/4-1BBL) T cells (Bertram et al., 2002a,b; Halstead et al., 2002; Lee et al., 2002; Rogers et al., 2001; Suresh et al., 2001; Tan et al., 1999). Additionally, costimulatory molecules may help to broaden the spectrum of epitopes recognized by a CD8þ T cell response by preferentially amplifying responses to subdominant epitopes (Halstead et al., 2002). Although blockade or genetic deletion of many costimulatory molecules can result in a reduction in the amplitude of the primary response, those CD8þ T cells that do respond appear to differentiate normally (Whitmire and Ahmed, 2000). This suggests that while these molecules may support the survival and proliferation of responding T cells, and thus regulate the amplitude of the response, they are not required for CD8þ T cells to complete their differentiative program. Cytokines can also regulate the amplitude of the initial CD8þ T cell response. Many signaling pathways involved in the survival and proliferation of T cells are activated by cytokines, including PI-3K and MAPK pathways, as well as STAT family transcription factors. Cytokine receptors are generally multimeric, and one or more receptor chains are commonly shared by related cytokines (Tagaya et al., 1996). Thus, some signaling pathways are activated by many cytokines within a given family. Other signaling pathways can only be activated by a single cytokine via a unique receptor chain. Therefore, cytokines that are structurally and phylogenetically related typically share at least some level of functional homology, due to activation of overlapping signaling pathways. Most cytokines fall into one of three groups, based on structural and functional similarity: common g chain (gc)-dependent cytokines, gp130dependent cytokines, and interferons (IFN). Members of all three of these groups play important roles in regulating T cell responses, although only two, the IFNs and the gc-dependent cytokines, will be discussed here. The IFNs can be divided into two groups, type I IFNs, a multigene family typiﬁed by IFN-a/b, and type II IFNs, of which the only member is the critical effector cytokine IFN-g (Shtrichman and Samuel, 2001). IFN-a/b can promote the survival of super-antigen activated CD8þ T cells in vitro (Marrack et al., 1999; Vella et al., 1998). Unlike gc-dependent cytokines, IFN-a/b

MEMORY T CELLS

203

promote T cell survival without increasing the levels of either bcl-2 or bcl-XL, implying that T cell survival may be promoted by multiple mechanisms, including some independent of bcl-2 or bcl-XL. In vivo, IFN-a/b have been reported to be critical for proper T cell responses after immunization with a mixture of protein and immunostimulatory DNA, although it is unclear to what extent this is due to the direct effect of IFN-a/b on the responding T cells (Cho et al., 2002). Additionally, the role of IFN-a/b in supporting activated T cell survival in vivo may be difﬁcult to determine since IFN-a/b have many immunoregulatory functions and can potently induce the production of other pro-survival cytokines, including IL-15 (Le Bon and Tough, 2002; Mattei et al., 2001). The gc-dependent cytokine group consists of interleukins (IL)-2,-4,-7,-9,-15, and -21 (Vosshenrich and Di Santo, 2001). Of these, IL-2, IL-7, and IL-15 are particularly important in supporting CD8þ T cell responses. Mice or humans with mutations in the gc chain have profound lymphopenia in both primary and secondary lymphoid tissues, with reduced numbers of most lymphoid cell types, including CD4þ and CD8þ T cells, indicating that gc-dependent cytokines are critical for the development and survival of naive T cells (DiSanto et al., 1994, 1995). The reduced numbers of T cells in gc-deﬁcient mice appears to be in large part due to the absence of IL-7 signals, for IL-7 and IL-7Ra deﬁcient mice also have markedly reduced numbers of T cells in primary and secondary lymphoid tissues (Maki et al., 1996; Peschon et al., 1994; von Freeden-Jeffry et al., 1995). Subsequent experiments have found that IL-7 can promote the survival of both naive and activated CD8þ T cell in vitro and in vivo, at least in part by maintaining levels of anti-apoptotic proteins such as bcl-2 or bcl-XL (Rathmell et al., 2001; Schluns et al., 2000; Vella et al., 1998). Although resting naive and memory CD8þ T cells express IL-7Ra, CD8þ T cells down-regulate IL-7Ra upon activation, possibly due to IL-2 induced signals (Schluns et al., 2000; Xue et al., 2002). Additionally, IL-7Ra deﬁcient CD8þ T cells undergo normal antigen-induced proliferation (Schluns et al., 2000). While these data suggest that IL-7 may not play an important role once an antigen-driven CD8þ T cell response has begun, it is interesting to note that when naive T cells are rested in vitro, they become reduced in size, have reduced rates of glycolysis, and take longer to undergo mitosis after stimulation with anti-CD3/CD28 (Rathmell et al., 2001). Incubation of cells with IL-7 rescues the decrease in cell size and glycolytic rate, although it cannot fully overcome the mitotic delay. Thus, IL-7 may not only maintain naive T cell survival by inducing anti-apoptotic bcl-2 family members, but it may also maintain naive T cells in a metabolically active state, which, in turn, may also render them more responsive to antigenic stimulation. In contrast to the dimeric IL-7R, IL-2R and IL-15R are hetero-trimeric receptors composed of unique alpha chains partnered with shared beta and gc chains (Bamford et al., 1994; Grabstein et al., 1994). The two shared receptor

204

PATRICK R. BURKETT ET AL.

chains mediate either all (for IL-2) or the great majority (for IL-15) of signaling events upon ligand engagement (Anderson et al., 1995; Giri et al., 1994). IL-2 is well established to promote the proliferation of activated T cells in vitro and in vivo. However, IL-2 deﬁcient mice have normal CD8þ T cell responses to both LCMV and vaccinia virus, suggesting that IL-2 is not essential for CD8þ T cell expansion and differentiation in vivo (Kundig et al., 1993). Moreover, while young IL-2 and IL-2Ra deﬁcient mice have normal numbers of B and T cells, they rapidly develop a variety of autoimmune disorders marked by severe lympadenopathy (Sadlack et al., 1993; Willerford et al., 1995). Thus IL-2’s nonredundant role in vivo appears to be controlling lymphocyte homeostasis rather than potentiating antigen-induced expansion. Unlike IL-2 and IL-7, which have largely similar effects on both CD4þ and CD8þ T cells, IL-15 appears to preferentially regulate CD8þ T cell function and survival. While IL-15’s in vitro activities closely mimic those of IL-2 (Bamford et al., 1994), IL-15’s in vivo roles are quite distinct from those of IL-2 (Lodolce et al., 2002; Ma et al., 2000). Transcripts for IL-15 are widely expressed in multiple hematopoietic and nonhematopoietic cell types, and are up-regulated by several cell types in response to various innate immune stimuli (Grabstein et al., 1994; Kuniyoshi et al., 1999; Mattei et al., 2001). Mice deﬁcient for IL-15 or IL-15Ra have approximately half of the normal number of naive CD8þ T cells, and the remaining CD8þ T cells have decreased levels of bcl-2, suggesting IL-15 may play a role in naive CD8þ T cell survival (Kennedy et al., 2000; Lodolce et al., 1998; Wu et al., 2002b). After challenge with viral and bacterial pathogens, both IL-15 and IL-15Ra deﬁcient mice have somewhat decreased primary immune responses, while IL-15 transgenic mice have increased primary responses, suggesting that in vivo, IL-15 may augment primary CD8þ T cell responses (Becker et al., 2002; Schluns et al., 2002; Yajima et al., 2001). Supporting this hypothesis, naive undivided T cells can bind IL-15 after TCR stimulation, while IL-2Ra is up-regulated over the course of several cell divisions (Li et al., 2001). These results suggest that functional IL-15R complexes are available to support T cell division early in an immune response while IL-2 signals maybe important later. The early upregulation of IL-15R correlates temporally with the rapid elaboration of IL-15 by professional APCs, which can occur even in the absence of protein synthesis (Neely et al., 2001). Since APC-derived IL-15 is rapidly produced by APCs in response to TLRLs and naive T cells can rapidly respond to it, IL-15 may be an important early signal for amplifying T cell responses in vivo. D. Contraction and the Beginnings of Memory As has been discussed, the dramatic expansion of antigen-speciﬁc CD8þ T cells following exposure to antigen is followed by a rapid decline in cell numbers, prior to reaching a stable plateau of memory CD8þ T cells (Fig. 1).

MEMORY T CELLS

205

This consistent pattern raises an important question: what is the relation of effector CTLs found at the peak of the immune response to the stable memory CD8þ T cell pool found later on? Two broad hypotheses address this question (Fig. 2). The ﬁrst, often termed the linear differentiation model, holds that memory CD8þ T cells are directly descended from effector CTLs and represent a ﬁnal stage of CD8þ T cell differentiation (Fig. 2A). The other hypothesis, or the bi-potential model, suggests that memory CD8þ T cells are a distinct lineage, arising either from a subset of naive CD8þ T cells or under priming conditions distinct from those that lead to CTL differentiation (Fig. 2B) (Bevan, 2002; Kaech et al., 2002b). These two hypotheses have very distinct views on the origin of memory CD8þ T cells, and therefore have very different implications for vaccine design. While some investigators have found evidence that memory CD8þ T cells may arise without becoming effector CD8þ T cells (Manjunath et al., 2001), the majority of current evidence suggests that in vivo most memory CD8þ T cells develop via the linear differentiation model (Kaech et al., 2002a; Opferman et al., 1999).

Fig 2 Models for the generation of memory CD8þ T cells. (A) The linear differentiation model suggests that all responding naı¨ve CD8þ T cells differentiate into effector CTLs, and from this latter population a small percentage survive and give rise to memory CD8þ T cells. (B) The bipotential model suggests that either a subset or under distinct priming conditions naı¨ve CD8 T cells can directly give rise to memory CD8þ T cells. However, the majority of naı¨ve CD8þ T cells still become terminally differentiated effector CTLs, all of which subsequently die. Overall it appears that the majority of memory CD8þ T cells develop via the scheme shown in (A).

206

PATRICK R. BURKETT ET AL.

The linear differentiation model makes several predictions that are not obvious from the decreasing potential model. The ﬁrst of these is that if memory CD8þ T cells are descended from CTLs, then the magnitude of the memory CD8þ T cell pool could be expected to be related to the size of the CTL response. Two experiments clearly illustrate that this is the case. First, when mice are infected with higher numbers of L. monocytogenes, the amplitude of the primary response is increased, although it occurs with similar kinetics. Thus, mice inoculated with high numbers of L. monocytogenes have a quantitatively larger response than do mice inoculated with lower numbers of L. monocytogenes at all stages of the immune response. However, when the number of antigen-speciﬁc CD8þ T cells found in mice inoculated with varying numbers of L. monocytogenes are normalized to the numbers of antigen-speciﬁc CD8þ T cells found at the peak of the response, the curves are virtually identical (Badovinac et al., 2002). This suggests that the size of the subsequent memory pool for a given epitope is highly dependent upon the size of the peak of the immune response for that same epitope. A caveat to these experiments is that the intensity of the innate immune stimulus may also be dependent on the number of L. monocytogenes, and lower numbers of L. monocytogenes may not create the same number of niches for priming memory CD8þ T cells that higher doses do. This has been addressed, to some degree, by examining CD8þ T cell responses to viruses such as LCMV or inﬂuenza that have multiple epitopes with varying degrees of immunodominance. While immunodominance is a poorly understood phenomenon likely due to many factors, more immunodominant epitopes clearly induce larger pools of effector CTLs compared to less immunodominant ones, and both pools of CTLs undergo the same proportional reduction in number as they become memory CD8þ T cells (Marshall et al., 2001; Murali-Krishna et al., 1998). Thus, not only is the size of the memory pool related to the peak of the primary response, but the repertoire of epitopes recognized by the memory pool is also proportionally similar to those recognized by the effector pool. A second prediction of the linear differentiation hypothesis is that a more differentiated CD8þ T cell population will generate memory CD8þ T cells more efﬁciently. Because several cell divisions are required before a naive CD8þ T cell can express the effector molecules that mark a CTL, such as perforin or granzyme B, increased cell division may correlate with increased differentiation, and thus increased potential to become memory CD8þ T cells (Jacob and Baltimore, 1999; Kaech et al., 2002b; Opferman et al., 1999). This hypothesis was examined by taking populations of antigen-stimulated, CFSElabeled CD8þ T cells that had, on average, divided a known number of times and were subsequently adoptively transferred into a new host. Cells that had undergone enough divisions to become CTLs (as assessed by intracellular perforin staining) prior to transfer preferentially gave rise to memory CD8þ

MEMORY T CELLS

207

T cells. Additionally, cells that had undergone more cell divisions prior to transfer possessed a greater potential to give rise to memory CD8þ T cells (Opferman et al., 1999). Further evidence that effector CTLs preferentially give rise to memory CD8þ T cells comes from mice that are transgenic for both human placental alkaline phosphatase (PLAP) under the control of a CD2 promoter and Cre recombinase under the control of the granzyme B promoter. PLAP expression was blocked in naive T cells by a ﬂoxed stop sequence, but upon CD8þ T cell activation, Cre could be induced and this stop sequence could be deleted, permitting expression of PLAP. After infection with LCMV, virtually all memory CD8þ T cells were PLAP positive (but Cre negative), suggesting that memory CD8þ T cells are descended from granzyme B expressing CTLs (Jacob and Baltimore, 1999). Additionally, only brief (24 hrs) contact with cognate peptide is required by naive CD8þ T cells to completely differentiate into effector CTLs and, subsequently, memory CD8þ T cells (Kaech and Ahmed, 2001). These data suggest that an antigen-initiated differentiative program exists that drives naive CD8þ T cells to ﬁrst become CTLs and subsequently memory CD8þ T cells, as would be predicted by the linear differentiation model. Perhaps the most compelling data favoring the linear differentiation model comes from a study that examined the kinetics of acquisition of both the functional and transcriptional proﬁle of memory CD8þ T cells. CTLs and memory CD8þ T cells possess distinct functional properties and these differences are likely due to distinct gene expression proﬁles. If the bi-potential model were correct, and both CTLs and memory CD8þ T cells were present early in an immune response, once the former died off, the remaining cells should immediately possess all the properties of memory CD8þ T cells. However, this was found not to be the case. Memory CD8þ T cells exhibited distinct gene expression patterns compared to either naive or effector CD8þ T cells. The gene expression pattern of antigen-speciﬁc CD8þ T cells isolated at various time points after the peak of an immune response changed quite dynamically before eventually settling into the expression pattern of memory CD8þ T cells. These changes could be observed even after the frequency of antigen-speciﬁc CD8þ T cells had stabilized, suggesting that the unique gene expression pattern of memory CD8þ T cells was only acquired slowly after the resolution of an immune response. While the conversion of effector CD8þ T cells to memory CD8þ T cells occurred largely in the absence of proliferation, substantial changes in the gene expression proﬁle were observed during this transition. These data suggest that effector CD8þ T cells that survive an immune response gradually acquire the gene expression proﬁle of memory CD8þ T cells, rather than a distinct subset of predetermined memory cells expanding over that period (Kaech et al., 2002a). Functional studies of antigen-speciﬁc CD8þ T cell populations also support the idea that effector CTLs differentiate into memory CD8þ T cells. Mice

208

PATRICK R. BURKETT ET AL.

congenitally infected with LCMV are lifelong carriers of the virus and have substantial viral titers in their serum and tissues. Adoptive transfer of CTLs isolated at the peak of the immune response into LCMV-carrier mice only transiently decreases serum virus titers, whereas transfer of memory CD8þ T cells is sufﬁcient for viral clearance. When antigen-speciﬁc CD8þ T cells from time points intermediate between the peak of the response and the stable memory phase were transferred into LCMV-carrier mice, the degree of viral clearance corresponded to the degree of acquisition of a memory gene expression proﬁle (Kaech et al., 2002b). Thus, the development of CD8þ T cell memory appears to largely be the product of an antigen-induced differentiative program whereby naive CD8þ T cells ﬁrst differentiate into CTLs, which then differentiate into memory CD8þ T cells. E. Who Gets to be a Memory Cell? The majority of effector CTLs die in the aftermath of an immune response. Why then do the vast majority (90–95%) of CTLs undergo apoptosis, while a minority (5–10%) escape death and become memory cells? Understanding the molecular mechanisms by which memory CD8þ T cells are generated or selected from CTLs could have important functional implications for vaccine design. In order to answer this question, however, it is ﬁrst necessary to examine why activated T cells die. Activated T cells could either die passively or be instructed to die, and evidence for both mechanisms exists (Hildeman et al., 2002b; Sprent and Surh, 2001). However, the death of activated T cells may be, in large part, a passive physiological consequence of their activation. Immune responses, particularly cellular immune responses, are metabolically intensive affairs, often resulting in damage to self-tissues in order to eliminate the pathogen. CD8þ T cells undergoing antigen-drive expansion divide rapidly and it has been estimated that antigen-speciﬁc cells undergo more than 14 cell divisions in the ﬁrst week of an immune response (Blattman et al., 2002). This rapid expansion of cell numbers is also accompanied by increased cell size, which indicates that activated T cells are more metabolically active than naive cells. Increased metabolic activity leads to the increased production of reactive oxygen species (ROS), which may, in turn, have toxic effects on activated T cells. Indeed, levels of superoxide in activated T cells are substantially higher than those in naive T cells. Moreover, addition of an ROS-scavenging agent correlated with increased mitochondrial membrane potential and decreased apoptosis of activated T cells, suggesting that many activated T cells may die due to oxidative stress (Hildeman et al., 1999). The idea that ROS are an important cause of the death of activated T cells is further supported by examining the impact of two members of the bcl-2 family on the death of activated T cells. The bcl-2 family consists of both pro- and

MEMORY T CELLS

209

anti-apoptotic proteins that appear to mediate their functions largely by regulating mitochondrial function (Chao and Korsmeyer, 1998). Bcl-2 and bcl-XL, both of which are anti-apoptotic, are decreased in activated T cells, while activated bcl-2 transgenic T cells are relatively resistant to death in vivo (Grayson et al., 2000; Hildeman et al., 2002a). Activated T cells deﬁcient for the pro-apoptotic family member bim show substantially reduced cell death in vitro and in vivo, even though they also down-regulate bcl-2, suggesting that bim is downstream of bcl-2. Intriguingly, addition of a ROS-scavenger mitigates the decrease in bcl-2 levels observed after activation, indicating that the increased oxidative stress induced by activation may decrease bcl-2 levels, which then leaves the T cells susceptible to bim-mediated cell death (Hildeman et al., 2002a). While the process of activation may render T cells more susceptible to death, CD8þ T cells face additional hazards upon becoming CTLs. The speciﬁc task of the CTL is to recognize and kill infected cells. In order to perform this task, CTLs use a variety of cytopathic compounds, such as perforin and granzymes, which are stored in specialized cytosolic granules. Even though these granules are only released in a highly localized manner, there is no a priori reason that CTLs should not also be susceptible to their own cytopathic compounds, particularly since a given CTL can kill more than one target cell (McGavern et al., 2002; Rothstein et al., 1978). While CTLs are relatively resilient to being killed by their own effector mechanisms, which may be due to surface-bound cathepsin B cleavage of perforin, mice lacking both perforin and IFN-g have delayed death of antigen-speciﬁc CTLs after challenge with L. monocytogenes (Badovinac et al., 2000; Balaji et al., 2002). Additionally, incubation of CTLs with antigen-loaded target cells increased the number of CTLs exhibiting an apoptotic phenotype in a granule-dependent manner, and the more a CTL population had killed, the less potential it had to give rise to a memory population (Opferman et al., 2001). Finally, other data indicate that a key oxidative repair enzyme is speciﬁcally targeted by granzyme A, and granzyme A-mediated cleavage of this enzyme renders cells more susceptible to CTL-induced death (Fan et al., 2003). Thus, a CTL’s effector molecules may act to render targets more susceptible to ROS-induced death. The idea that activation and effector activity are intrinsically toxic to responding CD8þ T cells would also provide an elegant explanation for the phenomenon of immune exhaustion (Moskophidis et al., 1993). Thus, the combination of oxidative stress that follows activation and the toxicity of their own effector mechanisms may contribute to the pronounced cell death that occurs at the end of a CD8þ T cell response. While CTL death may largely be a cell autonomous process, several lines of evidence suggest that CTLs require extrinsic signals in order to efﬁciently form a memory CD8þ T cell pool. One of the few extrinsic signals with a well-deﬁned role in regulating the conversion of effector CTLs into memory

210

PATRICK R. BURKETT ET AL.

CD8þ T cells is CD40. While many other costimulatory molecules inﬂuence the amplitude of the initial expansion of antigen-speciﬁc CD8þ T cells, and thus impact the size of the subsequent memory pool, CD40 plays a critical role in regulating the survival of activated CD8þ T cells as they begin to decline in number. CD40L is expressed predominantly by activated CD4þ T cells and is an important means by which CD4þ T cells provide help (Bennett et al., 1998; Borrow et al., 1996; Schoenberger et al., 1998; Whitmire et al., 1996; Xu et al., 1994). Interestingly, while CD40L-deﬁcient mice mount normal primary CTL responses to LCMV, they have markedly decreased numbers of memory CD8þ T cells (Borrow et al., 1996). This decrease is due to a more severe death phase in which 99% of responding CD8þ T cells die rather than to a failure to maintain memory CD8þ T cells (Borrow et al., 1998). Addition of an agonistic anti-CD40 monoclonal antibody (mAb) to soluble protein antigen both enhances the primary response and promotes subsequent formation of memory CD8þ T cells (Lefrancois et al., 2000). In vivo anti-CD40 mAb treatment can also induce in vitro-expanded CTLs to survive and mediate protective immunity following adoptive transfer (Tuma et al., 2002). Thus, CD40 ligation appears to be critical to the development of a normal memory CD8þ T cell pool. The exact mechanism by which CD40/CD40L signals increase the survival of activated CD8þ T cells and promote differentiation into memory CD8þ T cells is not entirely clear. One possibility may be CD40L engagement of CD40 on professional APCs. CD40 ligation can activate APCs by inducing expression of costimulatory molecules and cytokines, license APCs to crosspresent exogenous antigens to CD8þ T cells, and also can prolong the survival of DCs in vivo, possibly by upregulating bcl-2 (Bennett et al., 1998; Miga et al., 2001; Nopora and Brocker, 2002; Ridge et al., 1998; Schoenberger et al., 1998). Since DCs express many other costimulatory molecules and pro-survival cytokines, it is possible the effects of CD40 ligation on DC activation and survival might secondarily provide more niches in which activated CD8þ T cells could receive secondary survival signals. In support of this point, treatment of human DCs with CD40L induced the production of IL-15, which, in turn, supported antigen-speciﬁc CTLs in vitro (Kuniyoshi et al., 1999). Prolonged in vivo survival of DCs has been shown to increase both CD4þ and CD8þ T cell responses in vivo and in vitro in a CD40-dependent manner (Medema et al., 2001; Miga et al., 2001). In the case of CD8þ T cells, increased DC survival may be due to the induction of spi-6, an inhibitor of granzyme B, which can protect DCs from cytolysis by CTLs (Medema et al., 2001). Alternatively, CD40 may also be expressed by CD8þ T cells upon activation, and it is possible that direct ligation of CD40 on the CD8þ T cell itself may also be important. In support of this idea, a 2002 report found that while H-Y transgenic CD8þ T cells could efﬁciently eliminate transferred male cells in the absence of CD4þ T cell help, the addition of CD4þ T cells supported the

MEMORY T CELLS

211

subsequent development of memory CD8þ T cells. This effect required both the CD8þ and CD4þ T cells to be able to recognize the same APC, but did not require CD40 expression on the APC. Rather, CD40 expression on the responding CD8þ T cell was required in order for those cells to efﬁciently develop into memory CD8þ T cells (Bourgeois et al., 2002). However, other groups have found that the ability of anti-CD40 mAb to potentiate primary CD8þ T cell responses and support the formation of a memory CD8þ T cell pool can occur in the absence of CD4þ T cell help and requires CD40 expression on a cell type other than the responding CD8þ T cell, suggesting that the precise mechanism of CD40 action may depend on the exact system used (Lefrancois et al., 2000; Ridge et al., 1998). As has been mentioned, memory T cells express increased levels of bcl-2 and transgenic overexpression of bcl-2 results in increased survival of activated CD8þ T cells in vivo (Grayson et al., 2000; Hildeman et al., 2002a). Thus, bcl-2 may be a critical molecule for promoting the subsequent survival of activated CD8þ T cells, and competition for assorted cytokines and costimulatory molecules that can upregulate expression of bcl-2 may represent one important mechanism by which the memory CD8þ T cell pool is selected. It is also possible that other factors play important roles in promoting the survival of activated lymphocytes. A 2001 study examining the molecular basis for improvements in the survival of T cells after activation in the presence of adjuvants, such as LPS and CD40, found that bcl-3, a protein involved in the regulation of the transcription factor NF-kB, could also help promote survival of activated T cells. Bcl-3 is a member of the IkB subfamily, and unlike most other subfamily members, has been suggested to have transactivating activity. Upon activation and retroviral transduction of bcl-3, both activated CD4þ and CD8þ T cells showed improved in vivo and in vitro survival without enhanced proliferation (Mitchell et al., 2001). Intriguingly, the same study found that activation in the presence of LPS and CD40 reduced the relative expression of the NF-kB family member c-Rel, and that retroviral transduction of c-Rel resulted in decreased survival of activated T cells. While it is unclear how bcl-3 and c-Rel, respectively, promote survival and death, these studies suggest that regulation of activated T cell survival can indeed be both positively and negatively modulated by means other than bcl-2 family members. Numerous factors thus control the differentiation of effector CTLs into memory CD8þ T cells. These include signals that promote either the death or survival of effector CTLs. Interestingly, the former are principally cell autonomous molecules, such as ROS or perforin (Badovinac et al., 2000; Hildeman et al., 1999). In contrast, pro-survival signals, which support effector CTL survival and differentiation into memory CD8þ T cells, are largely noncell autonomous signals. For instance, cytokines made by APCs, such as IL-15, can maintain bcl-2 levels in activated T cells, while adjuvants, presumably

212

PATRICK R. BURKETT ET AL.

acting by inducing a variety of pro-survival molecules on APCs, can promote memory Tcell development by inducing the NF-kB transactivator bcl-3 (Mitchell et al., 2001; Vella et al., 1998). This suggests that while effector CTLs are largely destined to die, certain environmental cues from the host can select a subset of this population to be maintained as long-lived memory T cells. F. Whither Memory CD4þ T Cells? While the in vivo dynamics of antigen-speciﬁc CD4þ T cell responses have not been studied as extensively as those of CD8þ T cells, CD4þ T cell responses appear to follow a pattern similar to that of CD8þ T cells (Fig. 1). However, antigen-speciﬁc CD4þ T cells have been shown to divide somewhat more slowly than CD8þ T cells in response to L. monocytogenes (Foulds et al., 2002). Consistent with this observation, the frequency of antigen-speciﬁc CD4þ T cells throughout an immune response to LCMV is substantially less than that of antigen-speciﬁc CD8þ T cells (Homann et al., 2001; Whitmire et al., 1998). Since the reduced frequency of antigen-speciﬁc CD4þ T cells during an immune response makes it difﬁcult to study the dynamics of endogenous antigen-speciﬁc CD4þ T cells in vivo, the responses of adoptively transferred TCR transgenic CD4þ T cells are widely studied (Garside et al., 1998; Kearney et al., 1994; Reinhardt et al., 2001). This approach has been critical to studying the generation of memory CD4þ T cells, although currently the differentiative pathway memory CD4þ T cells is not as clear as it is for memory CD8þ T cells. Some reports have suggested that effector TH2 cells can give rise to memory cells without further proliferation, which would suggest memory CD4þ T cells arise via a linear differentiation process (Hu et al., 2001). However, the ability of in vitro differentiated TH1 cells to give rise to memory cells decreased with prolonged culture period, even though the percentage of cells that rapidly made IFN-g increased over this period, suggesting that memory TH1 cells arise by an alternate mechanism (Wu et al., 2002a). Additionally, a 2003 paper has argued that even though memory phenotype CD4þ T cells have differential DNA acetylation patterns indicative of being either TH1 or TH2 cells, they can alter their cytokine expression patterns upon restimulation (Messi et al., 2003). This observation suggests that while memory CD4þ T cells may be polarized into either TH1 or TH2 cells, they may retain some degree of functional plasticity. Thus, further study of the mechanism by which memory CD4þ T cells are generated is clearly needed. IV. How Are Memory Cells Maintained?

Two hallmarks of immunological memory are that it is exceptionally longlived and that it can mount rapid effector responses upon reencountering antigen. Indeed, individuals who contracted yellow fever were reported to

MEMORY T CELLS

213

have maintained protective antibody levels for over 75 years (Ahmed et al., 2002). Memory T cells are also long-lived, as demonstrated by the fact that once a mouse is immunized, memory CD8þ T cells can be detected for the life of the animal, and in humans vaccinia virus-speciﬁc CD8þ T cells can be detected up to 30 years after smallpox vaccination (Demkowicz et al., 1996; Murali-Krishna et al., 1998). Moreover, memory T cells not only persist for long periods of time, but they also are maintained in a heightened state of activation, as demonstrated by both their cell surface phenotype and functional attributes. This heightened state of activation likely plays an important role in the ability of memory T cells to protect against reinfection. Intriguingly, memory T cells are known to undergo a relatively rapid basal proliferation, and this proliferation has been suggested to play a role in the long-term survival of memory T cells (Tough and Sprent, 1994). Thus, elucidating the mechanisms by which memory T cells both survive for long periods of time and are maintained in a hyperresponsive state are key immunological questions. A. The Role of Persistent Antigen in Memory T Cell Maintenance The heightened state of activation and basal proliferation observed in memory T cells could be consistent with low-level, but persistent, stimulation by residual antigen. Supporting this idea, follicular dendritic cells have been shown to retain antigen for long periods in vivo, and low but detectable levels of LCMV have been shown to persist for many weeks after infection (Ciurea et al., 1999; Gray, 2002). Additionally, naive T cells, and CD8þ T cells in particular, are critically dependent on tonic MHC/TCR signals for their longterm survival (Murali-Krishna et al., 1999; Tanchot et al., 1997). Thus, stimulation of the TCR by small amounts of persistent antigen has long been suggested to be the mechanism by which memory T cells are maintained. However, two elegant studies convincingly demonstrated that memory T cells can survive quite well in the absence of MHC. Memory T cells transferred into MHCdeﬁcient environments also retained the ability to rapidly elaborate effector cytokines, such as IL-4 and IFN-g, and memory CD8þ T cells underwent basal proliferation at rates similar to those observed in MHC Class I competent hosts (Murali-Krishna et al., 1999; Swain et al., 1999). These data indicate that neither contact with MHC nor persistent antigen are necessary for both the basal proliferation and long-term survival of memory T cells. Despite this data, the role of both MHC and persistent antigen in memory T cell homeostasis is not completely resolved. First, H-Y TCR transgenic memory CD8þ T cells appear to require expression of MHC Class I in order to survive, suggesting that there may be some degree of heterogeneity in the dependence of memory T cells on MHC (Tanchot et al., 1997). Second, a recent paper has argued that memory CD4þ T cells require contact with their

214

PATRICK R. BURKETT ET AL.

restricting MHC in order to be maintained in an optimally functional state. Memory CD4þ T cells that were isolated from MHC Class II deﬁcient mice required higher concentrations of peptide in order to proliferate in vitro, and were unable to proliferate normally in response to antigen presented by naive B cells. Intriguingly, even transfer of memory CD4þ T cells from MHC Class II deﬁcient mice into MHC Class II competent hosts was not sufﬁcient to overcome these functional deﬁcits (Kassiotis et al., 2002). While similar experiments have not been conducted for memory CD8þ T cells, these data lend support to suggestions by several investigators that persistent antigen may be critical for the maintenance of memory T cell function, and thus protective immunity (Gray, 2002; Kundig et al., 1996; Zinkernagel, 2002). The use of the lox/cre recombinase technology to inducibly delete genes in vivo has allowed the role of the TCR in the maintenance of memory phenotype T cells to be examined. Upon deletion of the ﬁrst three exons of the TCRb constant region, the numbers of both naive CD4þ and CD8þ T cells begin to decay relatively rapidly. TCRb deﬁcient memory phenotype T cells showed reduced basal proliferation and either were maintained at constant levels (for CD4þ T cells), or declined slowly (CD8þ T cells) (Polic et al., 2001). Thus, while memory phenotype CD8þ T cells do not require MHC Class I in order to proliferate and survive, they may require continued expression of their TCR. Additionally, TCRb deﬁcient T cells had decreased levels of CD5 on their surface, similar to memory CD4þ T cells in the absence of MHC Class II, suggesting that elevated expression of CD5 depends, at least in part, on TCR/MHC interaction (Kassiotis et al., 2002; Polic et al., 2001). Overall, however, memory T cells are far less dependent on TCR/MHC signals for prolonged survival than are naive T cells. B. Cytokines and Memory T Cell Maintenance Cytokines play important roles in the development, survival, and differentiation of naive T cells in vivo. Thus, cytokines are attractive candidates for supporting the survival and proliferation of memory T cells. Two gc-dependent cytokines, IL-7 and IL-15, have been shown to play critical roles in supporting the survival and basal proliferation of memory CD8þ T cells. As discussed earlier, IL-7 is important for the development and survival of naive CD8þ T cells in vivo (Peschon et al., 1994; Schluns et al., 2000; von Freeden-Jeffry et al., 1995). Memory CD8þ T cells express levels of IL-7Ra similar to, or higher than, naive CD8þ T cells, suggesting that IL-7 may also regulate the survival of memory CD8þ T cells (Goldrath et al., 2002; Schluns et al., 2000). In support of this hypothesis, IL-7Ra deﬁcient CD8þ T cells are slower in upregulating bcl-2 levels after an immune response, and give rise to reduced numbers of memory cells (Schluns et al., 2000). Additionally, proliferation of memory CD8þ T cells upon transfer into a lymphopenic environment is

MEMORY T CELLS

215

reduced, but not absent, in either IL-7 deﬁcient mice or in mice treated with neutralizing antibodies to both IL-7 and IL-7Ra (Goldrath et al., 2002; Schluns et al., 2000; Tan et al., 2002). However, while the basal proliferation of memory CD8þ T cells in mice with a full lymphoid compartment was only moderately reduced by administration of blocking antibodies to IL-7Ra, the number of memory CD8þ T cells recovered was substantially reduced (Goldrath et al., 2002). Finally, while IL-7 transgenic mice have an exaggerated number of memory phenotype CD8þ T cells, these T cells do not undergo more rapid proliferation than similar cells in wildtype mice (Kieper et al., 2002). Thus, IL-7 appears to be an important survival factor for memory CD8þ T cells, but largely does not control basal proliferation of memory CD8þ T cells. In contrast to IL-7, which appears to support the survival of both naive and memory CD8þ T cells, the critical in vivo role of IL-15 appears to be the support of both the survival and proliferation of memory CD8þ T cells. While IL-15 and IL-15Ra deﬁcient mice have approximately half the normal number of naive CD8þ T cells, memory phenotype CD8þ T cells, particularly IL-2/ 15RbHi CD8þ T cells, are virtually absent (Kennedy et al., 2000; Lodolce et al., 1998). This appears to be due to a defect in survival, since IL-2/15RbHi memory phenotype CD8þ T cells rapidly decline in number after adoptive transfer into IL-15 deﬁcient mice (Judge et al., 2002). Moreover, mice expressing an IL-15 transgene driven by a MHC Class I promoter have drastically increased numbers of CD8þ T cells, the vast majority of which are of a memory phenotype (Fehniger et al., 2001). In vivo antibody blockade of IL-2/15Rb results in decreased proliferation of memory phenotype CD8þ T cells, which was not observed with antibodies to either IL-2 or IL-2Ra, suggesting that IL-15 is responsible for the basal proliferation of these cells (Ku et al., 2000). As discussed earlier, memory phenotype CD8þ T cells undergo rapid proliferation in vivo upon stimulation with TLRLs, such as LPS or polyI:C (Tough et al., 1996, 1997). TLRLs induce many cytokines in vivo, and memory phenotype CD8þ T cells will proliferate in vivo in response to these cytokines (Tough et al., 1996; Zhang et al., 1998). However, of these cytokines, only IL-15 is capable of driving proliferation of memory phenotype CD8þ T cells in vitro, suggesting it is directly responsible for the observed proliferation in vivo (Zhang et al., 1998). Supporting this point, bystander proliferation to polyI:C requires both IL-15 and IL-15Ra (Judge et al., 2002; Lodolce et al., 2001). Thus, IL-15 is required for the survival of memory phenotype CD8þ T cells, as well as for both basal and TLRL-induced proliferation. In addition to memory phenotype CD8þ T cells, studies have also revealed that antigen-speciﬁc memory CD8þ T cells also require IL-15 for their basal proliferation and long-term survival. IL-15 or IL-15Ra deﬁcient mice infected with either LCMV or VSV are able to mount protective primary responses (albeit of decreased amplitude), and initially possess signiﬁcant numbers of

216

PATRICK R. BURKETT ET AL.

memory CD8þ T cells. IL-15 and IL-15Ra deﬁcient memory cells express a normal pattern of surface markers and produce cytokines rapidly upon restimulation. However, in the absence of either IL-15 or IL-15Ra, memory CD8þ T cells fail to basally proliferate and the memory pool gradually decays in size over time in both lymphoid and nonlymphoid tissues (Becker et al., 2002; Goldrath et al., 2002; Schluns et al., 2002). Conversely, IL-15 transgenic mice generate increased numbers of memory CD8þ T cells after infection with L. monocytogenes (Yajima et al., 2001). Thus, both IL-15 and IL-15Ra are critical to the long-term maintenance of memory CD8þ T cells, and the basal proliferation of memory CD8þ T cells represents a vital homeostatic mechanism for the long-term survival of these cells. How do IL-7 and IL-15 support memory CD8þ T cells (Fig. 3)? Both cytokines have pro-survival effects on CD8þ T cells in vitro, suggesting that they likely mediate their effects by directly acting on the CD8þ T cell. IL-7 appears to be largely produced by radiation-resistant stromal cells and may act directly on the CD8þ T cell, for IL-7Ra deﬁcient memory CD8þ T cells express lower levels of bcl-2 and have impaired survival in vivo (Schluns et al., 2000). In contrast, the mechanism by which IL-15 supports memory CD8þ T cells in vivo may be somewhat more complicated. Unlike many other gc-dependent cytokines, both IL-15 and IL-15Ra are expressed by a diverse array of cell types, including hematopoietically derived

Fig 3 Memory CD8þ T cells are maintained by both IL-15 and IL-7. Memory CD8þ T cells receive survival and proliferation signals from IL-15 and IL-7, both of which are crucial for the long-term maintenance of CD8þ T cell memory. While IL-7 and IL-15 are somewhat redundant in vivo, survival signals are predominantly induced by IL-7, while IL-15 preferentially induces proliferation.

MEMORY T CELLS

217

cells, such as DCs, and nonhematopoietic tissues (Giri et al., 1995; Grabstein et al., 1994; Mattei et al., 2001). Because IL-15 has potent effects on memory CD8þ T cells in vitro, and IL-15 and IL-15Ra mice have very similar phenotypes, it has been widely assumed that IL-15 exerts its actions in vivo by binding to a hetero-trimeric receptor complex composed of the IL-15Ra, IL-2/15Rb, and gc on the surface of the memory CD8þ T cell. However, studies that examined the role of IL-15Ra in supporting bystander proliferation found that IL-15Ra was surprisingly not required on the CD8þ T cell itself, but was required on a bone marrow-derived accessory cell (Lodolce et al., 2001). Similarly, subsequent experiments have found that when naive OT-1 CD8þ T cells are adoptively transferred into IL-15Ra deﬁcient mice, they undergo normal initial antigen-driven expansion, but do not form a stable memory CD8þ T cell pool, in contrast to OT-1 CD8þ T cells in wildtype hosts. Additionally, memory OT-1 CD8þ T cells fail to undergo normal basal proliferation in IL-15Ra deﬁcient hosts (Burkett et al., 2002). Thus, with respect to memory CD8þ T cell homeostasis, it appears that IL-15Ra’s critical role is non-cell autonomous. Moreover, this data calls into question whether IL-15 is directly responsible for supporting memory CD8þ T cells, or if another, as yet undeﬁned, but clearly IL-15/IL-15Ra dependent, molecule or cell type directly supports the proliferation and long-term maintenance of memory CD8þ T cells. While it is possible that both IL-15 and IL-15Ra are required for support of an unknown cell type, or induction of an undetermined molecule, other data suggests that IL-15 may, in fact, be the critical ﬁnal signal for maintaining memory CD8þ T cells. IL-15Ra has long been known to have a high afﬁnity for IL-15 (Kd 1011m), independent of the other two receptor chains (Giri et al., 1995). Additionally, several groups have demonstrated that both monocytes and ﬁbroblasts upregulate surface-bound, biologically active IL-15 upon stimulation with a variety of TLRLs or innate immune cytokines (Briard et al., 2002; Musso et al., 1999; Neely et al., 2001; Rappl et al., 2001). However, it has been shown that this surface-bound IL-15 is not only associated with IL-15Ra, but that IL-15Ra can efﬁciently present IL-15 in trans to an IL-15 dependent cell-line expressing only IL-2/15Rb and gc (Dubois et al., 2002). It is thus possible that in vivo, in order for IL-15 to be either efﬁciently produced or bioavailable, IL-15 must ﬁrst be bound to IL-15Ra and presented in trans. Formal proof of trans presentation of IL-15 by IL-15Ra has yet to be demonstrated in vivo, but several lines of evidence are consistent with this mechanism. First, the majority of biochemical signals induced by IL-15 are thought to be transduced solely by IL-2/15Rb and gc. Although IL-15Ra may be involved in some signaling and improves the afﬁnity of the IL-15R complex for IL-15, IL-15 induced activation of JAK1, JAK3, STAT5, Ras/MAPK, and PI3K signaling pathways are critically dependent on IL-2/15Rb and gc

218

PATRICK R. BURKETT ET AL.

(Bulanova et al., 2001; Bulfone-Paus et al., 1999; Lin et al., 1995). Second, memory phenotype CD8þ T cells do not require IL-15Ra in order to undergo bystander proliferation, but they do require IL-2/15Rb (Judge et al., 2002; Lodolce et al., 2001). Third, DCs coordinately upregulate both IL-15 and IL-15Ra upon stimulation with TLRLs or innate immune cytokines (Mattei et al., 2001). Fourth, recent evidence suggests that NK cells, which also require IL-15 for their peripheral maintenance, require IL-15Ra expression on a secondary cell type in order to be maintained (Koka et al., 2003). Finally, when an Fc-IL-15Ra fusion protein is bound to plastic, it can bind IL-15 and support the survival of both wildtype and IL-15Ra deﬁcient memory CD8þ T cells, suggesting that trans presentation of IL-15 can indeed support memory CD8þ T cells in vitro (Burkett et al., 2003). Thus, trans presentation of IL-15 by IL-15Ra on a supporting cell to IL-2/15Rb and gc on a memory CD8þ T cell may be a critical mechanism by which memory CD8þ T cells are maintained in vivo. The intracellular IL-15-dependent signals that critically support memory CD8þ T cells remains unclear. Memory CD8þ T cells can readily proliferate to low doses of IL-15 in the absence of antigen stimulation, whereas naive CD8þ T cells largely do not (Zhang et al., 1998). This difference has been suggested to be due to differential receptor expression. While naive and memory CD8þ T cells express similar levels of gc, memory CD8þ T cells express high levels of IL-2/15Rb, which may be responsible for the increased responsiveness of memory CD8þ T cells to IL-15. This hypothesis was examined by transgenic expression of a chimeric protein consisting of the extracellular human IL-4Ra fused to the intracellular domains of mouse IL-2/15Rb. Although both naive and memory phenotype CD8þ T cells expressed equal levels of this transgene, only memory phenotype CD8þ T cells readily proliferated in vitro in response to human IL-4 (Gasser et al., 2000). While these data did not evaluate whether elevated IL-2/15Rb expression is required for memory CD8þ T cell survival in vivo, they clearly suggest that factors other than elevated receptor expression can control the increased proliferative response of memory CD8þ T cells to IL-15. While these other factors are unclear, it is interesting that STAT5 transgenic mice have a signiﬁcantly increased proportion of memory phenotype CD8þ T cells (Kelly et al., 2003). Thus, modulation of proteins involved in the regulation of IL-15R signaling, such as STAT5, may also play a role in sensitizing memory CD8þ T cells to respond to IL-15. The importance of IL-15-mediated maintenance of memory CD8þ T cells in protective immunity has been examined for several model pathogens. CD8þ T cell-derived IFN-g is critical for protective immune responses to the intracellular parasite Toxoplasma gondii. If T. gondii-primed mice are treated with soluble Fc-IL-15Ra for several weeks, they are unable to mount a protective response upon challenge with a larger number of parasites, which correlates

MEMORY T CELLS

219

with marked reduction in the number of IFN-g producing CD8þ T cells (Khan et al., 2002). Conversely, protective immunity to T. gondii can be prolonged by administration of exogenous IL-15 (Khan and Casciotti, 1999). Similarly, when equal numbers of bulk CD8þ T cells from L. monocytogenes-primed wildtype or IL-15 transgenic mice are transferred into naive recipients, only those mice receiving CD8þ T cells from IL-15 transgenic mice (which contained a higher proportion of L. monocytogenes-speciﬁc memory CD8þ T cells) were able to control a normally lethal L. monocytogenes infection (Yajima et al., 2001). However, both wildtype and IL-15 deﬁcient LCMV-immune mice could readily control infection by a more virulent strain of LCMV, despite the fact that the IL-15 deﬁcient mice had approximately half the number of memory CD8þ T cells at the time of rechallenge (Becker et al., 2002). Thus, IL-15-mediated maintenance of memory CD8þ T cells is required for protective immunity to some, but not all, intracellular pathogens. Although both IL-15 and IL-7 are involved in the maintenance of memory CD8þ T cells, they do not appear to be critically involved in the maintenance of memory CD4þ T cells. While in vitro assays have shown that IL-15 and IL-7 can drive the proliferation of effector memory CD4þ T cells, and to a lesser extent, central memory CD4þ T cells, a variety of in vivo experiments suggest neither is required (Geginat et al., 2001). Both IL-15 and IL-15Ra deﬁcient mice have largely normal numbers and subsets of CD4þ T cells, as do IL-15 transgenic mice (Fehniger et al., 2001; Kennedy et al., 2000; Lodolce et al., 1998). While IL-7 and IL-15 are critical for lymphopenia-induced proliferation of memory CD8þ T cells, they are not required by memory phenotype CD4þ T cells (Goldrath et al., 2002; Tan et al., 2002). Moreover, although gc is critical for the survival of naive CD4þ T cells, it is dispensable for both antigendriven responses and the survival of memory CD4þ T cells (Lantz et al., 2000). Thus, neither gc-dependent cytokines nor MHC Class II appear to be important for the long-term survival and maintenance of memory CD4þ T cells. Interestingly, a 2001 report suggested that antigen-speciﬁc memory CD4þ T cells normally undergo slow decay in numbers after priming, in contrast to memory CD8þ T cells (Homann et al., 2001). Although other groups have found that CD4þ memory T cells are maintained at steady levels (Whitmire et al., 1998), it remains to be seen what factor or factors do maintain memory CD4þ T cells, or whether they are maintained at all. C. Negative Regulation of Memory T Cells Memory T cells with a given antigen speciﬁcity are present at a higher frequency than their naive precursors. While this increase in cell number likely plays an important role in how memory T cells can mediate protective immunity, it also presents an organism with a theoretical problem. Hypothetically, each time an organism undergoes a T cell response, there

220

PATRICK R. BURKETT ET AL.

will be a net increase in the number of T cells, which could lead to an organism’s amassing a huge number of memory T cells after several infections unless other homeostatic mechanisms control the size of the memory T cell pool. Several possibilities exist for how this is accomplished. First of all, the size of the memory T cell pool could increase over time. While the total number of T cells in mice remains relatively constant throughout life, older mice exhibit a substantial increase in the relative proportion of T cells with a memory phenotype (Zhang et al., 2002). Thus, the relative size of the memory T cell pool increases over time. While that may be one mechanism for accommodating newly generated memory CD8þ T cells, there is also evidence that after infection with a heterologous virus, the frequency of preexisting memory CD8þ T cells can be substantially decreased (Selin et al., 1999). This process, termed attrition, can also be mimicked with polyI:C, and the loss of memory CD8þ T cells occurs rapidly, although their numbers can subsequently recover, likely due to bystander proliferation. The molecular mechanism of attrition is unclear, but it requires IFN-a/b and does not require perforin, IFN-g, or Fas/FasL interactions (McNally et al., 2001). While IFNa/b are known to support the survival of activated T cells in vitro, the kinetics of polyI:C-induced attrition correlate temporally with reduced levels of IL-7 mRNA, raising the possibility that attrition may be due to indirect effects of IFN-a/b on T cell survival (Lodolce et al., 2001; Vella et al., 1998). Intriguingly, attrition may not affect all memory T cells equally. As has been discussed, preexisting memory CD8þ T cells can undergo activation and proliferation after infection with heterologous viruses (Chen et al., 2001; Kim et al., 2002). Unlike in the case of TLRL-induced bystander proliferation, which nonspeciﬁcally drives the proliferation of memory T cells, heterologous activation results in the preferential expansion of memory CD8þ T cells of certain antigen speciﬁcity, and attrition of memory CD8þ T cells of other speciﬁcities. These expanded memory T cells appear to be somewhat crossreactive, and seem selected to be preferentially maintained in the memory pool, whereas CD8þ T cells of more restricted speciﬁcity undergo attrition (Brehm et al., 2002). Thus, subsequent immune responses not only clear space for the generation of new memory T cells, but also actively shape the existing memory T cell repertoire. V. Conclusions

Edward Jenner’s recognition of immune memory represents one of the oldest and most signiﬁcant of immunological observations. However, it has also proved to be one of the most difﬁcult to understand on both a cellular and molecular level. Recently, signiﬁcant advances have been made in understanding where memory T cells, particularly CD8þ T cells, come from, what they can

MEMORY T CELLS

221

do, and what maintains them. Still, a number of signiﬁcant questions remain. Foremost among these are the factors that control memory T cell differentiation and the stable maintenance of the unique phenotype of memory T cells. Memory CD8þ T cells are clearly both functionally and phenotypically distinct from naive and effector CD8þ T cells, possessing properties of both. For instance, memory CD8þ T cells, like naive CD8þ T cells, possess extensive proliferative potential. However, despite being resting, largely quiescent cells, they also can rapidly elaborate effector cytokines in response to cognate antigen, as can CTLs. Moreover, they possess a lengthy lifespan marked by slow basal proliferation. While the factors that control this unique phenotype are unknown, they may be dependent to some extent on IL-7 and/or IL-15 for their expression, given these cytokines’ critical role in the long-term maintenance of memory CD8þ T cells. Intriguingly, LKLF, a transcription factor critical for naive T cells’ survival, can be induced in activated CD8þ T cells by IL-7, suggesting LKLF may also regulate memory CD8þ T cell survival (Schober et al., 1999). Additionally, bcl-6, a transcriptional repressor that blocks terminal B cell development, has been suggested to regulate the survival of memory CD8þ T cells (Ichii et al., 2002). Identiﬁcation of factors that regulate memory CD8þ T cell survival and function and understanding of the regulation of those factors are intriguing areas of future research. References Ahmed, R., Lanier, J. G., and Pamer, E. G. (2002). Immunological memory and infection. In ‘‘Immunology of Infectious Diseases’’ (S. H. E. Kaufmann, A. Sher, and R. Ahmed, Eds.), p. 175. ASM Press, Washington DC. Anderson, D. M., Kumaki, S., Ahdieh, M., Bertles, J., Tometsko, M., Loomis, A., Giri, J., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., et al. (1995). Functional characterization of the human interleukin-15 receptor alpha chain and close linkage of IL-15RA and IL-2RA genes. J. Biol. Chem. 270, 29862. Bachmann, M. F., Gallimore, A., Linkert, S., Cerundolo, V., Lanzavecchia, A., Kopf, M., and Viola, A. (1999). Developmental regulation of Lck targeting to the CD8 coreceptor controls signaling in naive and memory T cells. J. Exp. Med. 189, 1521. Badovinac, V. P., and Harty, J. T. (2002). CD8(þ) T-cell homeostasis after infection: Setting the ‘‘curve.’’ Microbes. Infect. 4, 441. Badovinac, V. P., Porter, B. B., and Harty, J. T. (2002). Programmed contraction of CD8(þ) T cells after infection. Nat. Immunol. 3, 619. Badovinac, V. P., Tvinnereim, A. R., and Harty, J. T. (2000). Regulation of antigen-speciﬁc CD8þ T cell homeostasis by perforin and interferon-gamma. Science 290, 1354. Balaji, K. N., Schaschke, N., Machleidt, W., Catalfamo, M., and Henkart, P. A. (2002). Surface cathepsin B protects cytotoxic lymphocytes from self-destruction after degranulation. J. Exp. Med. 196, 493. Bamford, R. N., Grant, A. J., Burton, J. D., Peters, C., Kurys, G., Goldman, C. K., Brennan, J., Roessler, E., and Waldmann, T. A. (1994). The interleukin (IL) 2 receptor beta chain is shared by IL-2 and a cytokine, provisionally designated IL-T, that stimulates T-cell proliferation and the induction of lymphokine-activated killer cells. Proc. Natl. Acad. Sci. USA 91, 4940.

222

PATRICK R. BURKETT ET AL.

Banchereau, J., and Steinman, R. M. (1998). Dendritic cells and the control of immunity. Nature 392, 245. Becker, T. C., Wherry, E. J., Boone, D., Murali-Krishna, K., Antia, R., Ma, A., and Ahmed, R. (2002). Interleukin 15 is required for proliferative renewal of virus-speciﬁc memory CD8 T cells. J. Exp. Med. 195, 1541. Bennett, S. R., Carbone, F. R., Karamalis, F., Flavell, R. A., Miller, J. F., and Heath, W. R. (1998). Help for cytotoxic-T-cell responses is mediated by CD40 signaling. Nature 393, 478. Berg, R. E., Cordes, C. J., and Forman, J. (2002). Contribution of CD8þ T cells to innate immunity: IFN-gamma secretion induced by IL-12 and IL-18. Eur. J. Immunol. 32, 2807. Bertram, E. M., Lau, P., and Watts, T. H. (2002a). Temporal segregation of 4-1BB versus CD28mediated costimulation: 4-1BB ligand inﬂuences T cell numbers late in the primary response and regulates the size of the T cell memory response following inﬂuenza infection. J. Immunol. 168, 3777. Bertram, E. M., Tafuri, A., Shahinian, A., Chan, V. S., Hunziker, L., Recher, M., Ohashi, P. S., Mak, T. W., and Watts, T. H. (2002b). Role of ICOS versus CD28 in antiviral immunity. Eur. J. Immunol. 32, 3376. Bevan, M. J. (2002). Immunology: Remembrance of things past. Nature 420, 748. Blattman, J. N., Antia, R., Sourdive, D. J., Wang, X., Kaech, S. M., Murali-Krishna, K., Altman, J. D., and Ahmed, R. (2002). Estimating the precursor frequency of naive antigen-speciﬁc CD8 T cells. J. Exp. Med. 195, 657. Borrow, P., Tishon, A., Lee, S., Xu, J., Grewal, I. S., Oldstone, M. B., and Flavell, R. A. (1996). CD40L-deﬁcient mice show deﬁcits in antiviral immunity and have an impaired memory CD8þ CTL response. J. Exp. Med. 183, 2129. Borrow, P., Tough, D. F., Eto, D., Tishon, A., Grewal, I. S., Sprent, J., Flavell, R. A., and Oldstone, M. B. (1998). CD40 ligand-mediated interactions are involved in the generation of memory CD8(þ) cytotoxic T lymphocytes (CTL) but are not required for the maintenance of CTL memory following virus infection. J. Virol. 72, 7440. Bourgeois, C., Rocha, B., and Tanchot, C. (2002). A role for CD40 expression on CD8þ T cells in the generation of CD8þ T cell memory. Science 297, 2060. Brehm, M. A., Pinto, A. K., Daniels, K. A., Schneck, J. P., Welsh, R. M., and Selin, L. K. (2002). T cell immunodominance and maintenance of memory regulated by unexpectedly cross-reactive pathogens. Nat. Immunol. 3, 627. Briard, D., Brouty-Boye, D., Azzarone, B., and Jasmin, C. (2002). Fibroblasts from human spleen regulate NK cell differentiation from blood CD34(þ) progenitors via cell surface IL-15. J. Immunol. 168, 4326. Bulanova, E., Budagian, V., Pohl, T., Krause, H., Durkop, H., Paus, R., and Bulfone-Paus, S. (2001). The IL-15Ralpha chain signals through association with Syk in human B cells. J. Immunol. 167, 6292. Bulfone-Paus, S. S., Bulanova, E., Pohl, T., Budagian, V., Durkop, H., Ruckert, R., Kunzendorf, U., Paus, R., and Krause, H. (1999). Death deﬂected: IL-15 inhibits TNF-alpha-mediated apoptosis in ﬁbroblasts by TRAF2 recruitment to the IL-15Ralpha chain. FASEB J. 13, 1575. Burkett, P. R., Koka, R., Chien, M., Chai, S., Chan, F., Ma, A., and Boone, D. L. (2002). IL-15Ralpha expression on CD8þ T cells is dispensable for T cell memory. Proc. Natl. Acad. Sci. USA 100, 4724. Carreno, B. M., and Collins, M. (2002). The B7 family of ligands and its receptors: New pathways for costimulation and inhibition of immune responses. Annu. Rev. Immunol. 20, 29. Champagne, P., Ogg, G. S., King, A. S., Knabenhans, C., Ellefsen, K., Nobile, M., Appay, V., Rizzardi, G. P., Fleury, S., Lipp, M., Forster, R., Rowland-Jones, S., Sekaly, R. P., McMichael, A. J., and Pantaleo, G. (2001). Skewed maturation of memory HIV-speciﬁc CD8 T lymphocytes. Nature 410, 106.

MEMORY T CELLS

223

Chao, D. T., and Korsmeyer, S. J. (1998). BCL-2 family: Regulators of cell death. Annu. Rev. Immunol. 16, 395. Chen, H. D., Fraire, A. E., Joris, I., Brehm, M. A., Welsh, R. M., and Selin, L. K. (2001). Memory CD8þ T cells in heterologous antiviral immunity and immunopathology in the lung. Nat. Immunol. 2, 1067. Cho, B. K., Rao, V. P., Ge, Q., Eisen, H. N., and Chen, J. (2000). Homeostasis-stimulated proliferation drives naive T cells to differentiate directly into memory T cells. J. Exp. Med. 192, 549. Cho, B. K., Wang, C., Sugawa, S., Eisen, H. N., and Chen, J. (1999). Functional differences between memory and naive CD8 T cells. Proc. Natl. Acad. Sci. USA 96, 2976. Cho, H. J., Hayashi, T., Datta, S. K., Takabayashi, K., Van Uden, J. H., Horner, A., Corr, M., and Raz, E. (2002). IFN-alpha beta promote priming of antigen-speciﬁc CD8þ and CD4þ T lymphocytes by immunostimulatory DNA-based vaccines. J. Immunol. 168, 4907. Ciurea, A., Klenerman, P., Hunziker, L., Horvath, E., Odermatt, B., Ochsenbein, A. F., Hengartner, H., and Zinkernagel, R. M. (1999). Persistence of lymphocytic choriomeningitis virus at very low levels in immune mice. Proc. Natl. Acad. Sci. USA 96, 11964. Clarke, S. R., and Rudensky, A. Y. (2000). Survival and homeostatic proliferation of naive peripheral CD4þ T cells in the absence of self peptide:MHC complexes. J. Immunol. 165, 2458. Demkowicz, W. E., , Jr., Littaua, R. A., Wang, J., and Ennis, F. A. (1996). Human cytotoxic T-cell memory: Long-lived responses to vaccinia virus. J. Virol. 70, 2627. den Haan, J. M., and Bevan, M. J. (2001). Antigen presentation to CD8þ T cells: Cross-priming in infectious diseases. Curr. Opin. Immunol. 13, 437. DiSanto, J. P., Dautry-Varsat, A., Certain, S., Fischer, A., and de Saint Basile, G. (1994). Interleukin-2 (IL-2) receptor gamma chain mutations in X-linked severe combined immunodeﬁciency disease result in the loss of high-afﬁnity IL-2 receptor binding. Eur. J. Immunol. 24, 475. 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 gamma chain. Proc. Natl. Acad. Sci. USA 92, 377. Dubois, S., Mariner, J., Waldmann, T. A., and Tagaya, Y. (2002). IL-15Ralpha recycles and presents IL-15 In trans to neighboring cells. Immunity 17, 537. Eberl, G., Brawand, P., and MacDonald, H. R. (2000). Selective bystander proliferation of memory CD4þ and CD8þ T cells upon NK T or T cell activation. J. Immunol. 165, 4305. Ely, K. H., Cauley, L. S., Roberts, A. D., Brennan, J. W., Cookenham, T., and Woodland, D. L. (2003). Nonspeciﬁc recruitment of memory CD8(þ) T cells to the lung airways during respiratory virus infections. J. Immunol. 170, 1423. Fan, Z., Beresford, P. J., Zhang, D., Xu, Z., Novina, C. D., Yoshida, A., Pommier, Y., and Lieberman, J. (2003). Cleaving the oxidative repair protein Ape1 enhances cell death mediated by granzyme A. Nat. Immunol. 4, 145. Fehniger, T. A., Suzuki, K., Ponnappan, A., VanDeusen, J. B., Cooper, M. A., Florea, S. M., Freud, A. G., Robinson, M. L., Durbin, J., and Caligiuri, M. A. (2001). Fatal leukemia in interleukin 15 transgenic mice follows early expansions in natural killer and memory phenotype CD8þ T cells. J. Exp. Med. 193, 219. Foulds, K. E., Zenewicz, L. A., Shedlock, D. J., Jiang, J., Troy, A. E., and Shen, H. (2002). Cutting edge: CD4 and CD8 T cells are intrinsically different in their proliferative responses. J. Immunol. 168, 1528. Fry, T. J., and Mackall, C. L. (2001). Interleukin-7: Master regulator of peripheral T-cell homeostasis? Trends Immunol. 22, 564. Galvan, M., Murali-Krishna, K., Ming, L. L., Baum, L., and Ahmed, R. (1998). Alterations in cell surface carbohydrates on T cells from virally infected mice can distinguish effector/memory CD8þ T cells from naive cells. J. Immunol. 161, 641.

224

PATRICK R. BURKETT ET AL.

Garside, P., Ingulli, E., Merica, R. R., Johnson, J. G., Noelle, R. J., and Jenkins, M. K. (1998). Visualization of speciﬁc B and T lymphocyte interactions in the lymph node. Science 281, 96. Gasser, S., Corthesy, P., Beerman, F., MacDonald, H. R., and Nabholz, M. (2000). Constitutive expression of a chimeric receptor that delivers IL-2/IL-15 signals allows antigen-independent proliferation of CD8þCD44 high but not other T cells. J. Immunol. 164, 5659. Geginat, J., Sallusto, F., and Lanzavecchia, A. (2001). Cytokine-driven proliferation and differentiation of human naive, central memory, and effector memory CD4(þ) T cells. J. Exp. Med. 194, 1711. Giri, J. G., Ahdieh, M., Eisenman, J., Shanebeck, K., Grabstein, K., Kumaki, S., Namen, A., Park, L. S., Cosman, D., and Anderson, D. (1994). Utilization of the beta and gamma chains of the IL-2 receptor by the novel cytokine IL-15. EMBO J. 13, 2822. Giri, J. G., Kumaki, S., Ahdieh, M., Friend, D. J., Loomis, A., Shanebeck, K., DuBose, R., Cosman, D., Park, L. S., and Anderson, D. M. (1995). Identiﬁcation and cloning of a novel IL-15 binding protein that is structurally related to the alpha chain of the IL-2 receptor. EMBO J. 14, 3654. Goldrath, A. W., and Bevan, M. J. (1999). Low-afﬁnity ligands for the TCR drive proliferation of mature CD8þ T cells in lymphopenic hosts. Immunity 11, 183. Goldrath, A. W., Bogatzki, L. Y., and Bevan, M. J. (2000). Naive T cells transiently acquire a memory-like phenotype during homeostasis-driven proliferation. J. Exp. Med. 192, 557. Goldrath, A. W., Sivakumar, P. V., Glaccum, M., Kennedy, M. K., Bevan, M. J., Benoist, C., Mathis, D., and Butz, E. A. (2002). Cytokine requirements for acute and basal homeostatic proliferation of naive and memory CD8þ T cells. J. Exp. Med. 195, 1515. Grabstein, K. H., Eisenman, J., Shanebeck, K., Rauch, C., Srinivasan, S., Fung, V., Beers, C., Richardson, J., Schoenborn, M. A., Ahdieh, M., et al. (1994). Cloning of a T cell growth factor that interacts with the beta chain of the interleukin-2 receptor. Science 264, 965. Gray, D. (2002). A role for antigen in the maintenance of immunological memory. Nat. Rev. Immunol. 2, 60. Grayson, J. M., Harrington, L. E., Lanier, J. G., Wherry, E. J., and Ahmed, R. (2002). Differential sensitivity of naive and memory CD8þ T cells to apoptosis in vivo. J. Immunol. 169, 3760. Grayson, J. M., Zajac, A. J., Altman, J. D., and Ahmed, R. (2000). Cutting edge: Increased expression of Bcl-2 in antigen-speciﬁc memory CD8þ T cells. J. Immunol. 164, 3950. Hafalla, J. C., Sano, G., Carvalho, L. H., Morrot, A., and Zavala, F. (2002). Short-term antigen presentation and single clonal burst limit the magnitude of the CD8(þ) T cell responses to malaria liver stages. Proc. Natl. Acad. Sci. USA 99, 11819. Halstead, E. S., Mueller, Y. M., Altman, J. D., and Katsikis, P. D. (2002). In vivo stimulation of CD137 broadens primary antiviral CD8þ T cell responses. Nat. Immunol. 3, 536. Hamilton, S. E., and Harty, J. T. (2002). Quantitation of CD8þ T cell expansion, memory, and protective immunity after immunization with peptide-coated dendritic cells. J. Immunol. 169, 4936. Harrington, L. E., Galvan, M., Baum, L. G., Altman, J. D., and Ahmed, R. (2000). Differentiating between memory and effector CD8 T cells by altered expression of cell surface O-glycans. J. Exp. Med. 191, 1241. Hildeman, D. A., Mitchell, T., Teague, T. K., Henson, P., Day, B. J., Kappler, J., and Marrack, P. C. (1999). Reactive oxygen species regulate activation-induced T cell apoptosis. Immunity 10, 735. Hildeman, D. A., Zhu, Y., Mitchell, T. C., Bouillet, P., Strasser, A., Kappler, J., and Marrack, P. (2002a). Activated T cell death in vivo mediated by proapoptotic bcl-2 family member bim. Immunity 16, 759. Hildeman, D. A., Zhu, Y., Mitchell, T. C., Kappler, J., and Marrack, P. (2002b). Molecular mechanisms of activated T cell death in vivo. Curr. Opin. Immunol. 14, 354.

MEMORY T CELLS

225

Homann, D., Teyton, L., and Oldstone, M. B. (2001). Differential regulation of antiviral T-cell immunity results in stable CD8þ but declining CD4þ T-cell memory. Nat. Med. 7, 913. Hou, S., Hyland, L., Ryan, K. W., Portner, A., and Doherty, P. C. (1994). Virus-speciﬁc CD8þ T-cell memory determined by clonal burst size. Nature 369, 652. Hu, H., Huston, G., Duso, D., Lepak, N., Roman, E., and Swain, S. L. (2001). CD4(þ) T cell effectors can become memory cells with high efﬁciency and without further division. Nat. Immunol. 2, 705. Ichii, H., Sakamoto, A., Hatano, M., Okada, S., Toyama, H., Taki, S., Arima, M., Kuroda, Y., and Tokuhisa, T. (2002). Role for Bcl-6 in the generation and maintenance of memory CD8þ T cells. Nat. Immunol. 3, 558. Jacob, J., and Baltimore, D. (1999). Modeling T-cell memory by genetic marking of memory T cells in vivo. Nature 399, 593. Jamieson, A. M., Diefenbach, A., McMahon, C. W., Xiong, N., Carlyle, J. R., and Raulet, D. H. (2002). The role of the NKG2D immunoreceptor in immune cell activation and natural killing. Immunity 17, 19. Judge, A. D., Zhang, X., Fujii, H., Surh, C. D., and Sprent, J. (2002). Interleukin 15 controls both proliferation and survival of a subset of memory-phenotype CD8(þ) T cells. J. Exp. Med. 196, 935. Jung, S., Unutmaz, D., Wong, P., Sano, G., De los Santos, K., Sparwasser, T., Wu, S., Vuthoori, S., Ko, K., Zavala, F., Pamer, E. G., Littman, D. R., and Lang, R. A. (2002). In vivo depletion of CD11c(þ) dendritic cells abrogates priming of CD8(þ) T cells by exogenous cell-associated antigens. Immunity 17, 211. Kaech, S. M., and Ahmed, R. (2001). Memory CD8þ T cell differentiation: Initial antigen encounter triggers a developmental program in naive cells. Nat. Immunol. 2, 415. Kaech, S. M., Hemby, S., Kersh, E., and Ahmed, R. (2002a). Molecular and functional proﬁling of memory CD8 T cell differentiation. Cell 111, 837. Kaech, S. M., Wherry, E. J., and Ahmed, R. (2002b). Effector and memory T-cell differentiation: Implications for vaccine development. Nat. Rev. Immunol. 2, 251. Kassiotis, G., Garcia, S., Simpson, E., and Stockinger, B. (2002). Impairment of immunological memory in the absence of MHC despite survival of memory T cells. Nat. Immunol. 3, 244. Kearney, E. R., Pape, K. A., Loh, D. Y., and Jenkins, M. K. (1994). Visualization of peptide-speciﬁc T cell immunity and peripheral tolerance induction in vivo. Immunity 1, 327. Kedl, R. M., Schaefer, B. C., Kappler, J. W., and Marrack, P. (2002). T cells down-modulate peptide-MHC complexes on APCs in vivo. Nat. Immunol. 3, 27. Kelly, J., Spolski, R., Imada, K., Bollenbacher, J., Lee, S., and Leonard, W. J. (2003). A role for stat5 in CD8(þ) T cell homeostasis. J. Immunol. 170, 210. Kennedy, M. K., Glaccum, M., Brown, S. N., Butz, E. A., Viney, J. L., Embers, M., Matsuki, N., Charrier, K., Sedger, L., Willis, C. R., Brasel, K., Morrissey, P. J., Stocking, K., Schuh, J. C., Joyce, S., and Peschon, J. J. (2000). Reversible defects in natural killer and memory CD8 T cell lineages in interleukin 15-deﬁcient mice. J. Exp. Med. 191, 771. Khan, I. A., and Casciotti, L. (1999). IL-15 prolongs the duration of CD8þ T cell-mediated immunity in mice infected with a vaccine strain of Toxoplasma gondii. J. Immunol. 163, 4503. Khan, I. A., Moretto, M., Wei, X. Q., Williams, M., Schwartzman, J. D., and Liew, F. Y. (2002). Treatment with soluble interleukin-15Ralpha exacerbates intracellular parasitic infection by blocking the development of memory CD8þ T cell response. J. Exp. Med. 195, 1463. Kieper, W. C., Tan, J. T., Bondi-Boyd, B., Gapin, L., Sprent, J., Ceredig, R., and Surh, C. D. (2002). Overexpression of interleukin (IL)-7 leads to IL-15-independent generation of memory phenotype CD8þ T cells. J. Exp. Med. 195, 1533.

226

PATRICK R. BURKETT ET AL.

Kim, S. K., Brehm, M. A., Welsh, R. M., and Selin, L. K. (2002). Dynamics of memory T cell proliferation under conditions of heterologous immunity and bystander stimulation. J. Immunol. 169, 90. Klenerman, P., Cerundolo, V., and Dunbar, P. R. (2002). Tracking T cells with tetramers: New tales from new tools. Nat. Rev. Immunol. 2, 263. Koka, R., Burkett, P. R., Chien, M., Chai, S., Chan, F., Lodolce, J. P., Boone, D. L., and Ma, A. (2003). IL-15Ralpha deﬁcient NK cells survive in normal but not IL-15Ralpha deﬁcient mice. J. Exp. Med. 197, 977. Kondo, T., Cortese, I., Markovic-Plese, S., Wandinger, K. P., Carter, C., Brown, M., Leitman, S., and Martin, R. (2001). Dendritic cells signal T cells in the absence of exogenous antigen. Nat. Immunol. 2, 932. Ku, C. C., Murakami, M., Sakamoto, A., Kappler, J., and Marrack, P. (2000). Control of homeostasis of CD8þ memory T cells by opposing cytokines. Science 288, 675. Kundig, T. M., Bachmann, M. F., DiPaolo, C., Simard, J. J., Battegay, M., Lother, H., Gessner, A., Kuhlcke, K., Ohashi, P. S., Hengartner, H., et al. (1995). Fibroblasts as efﬁcient antigenpresenting cells in lymphoid organs. Science 268, 1343. Kundig, T. M., Bachmann, M. F., Oehen, S., Hoffmann, U. W., Simard, J. J., Kalberer, C. P., Pircher, H., Ohashi, P. S., Hengartner, H., and Zinkernagel, R. M. (1996). On the role of antigen in maintaining cytotoxic T-cell memory. Proc. Natl. Acad. Sci. USA 93, 9716. Kundig, T. M., Schorle, H., Bachmann, M. F., Hengartner, H., Zinkernagel, R. M., and Horak, I. (1993). Immune responses in interleukin-2-deﬁcient mice. Science 262, 1059. Kuniyoshi, J. S., Kuniyoshi, C. J., Lim, A. M., Wang, F. Y., Bade, E. R., Lau, R., Thomas, E. K., and Weber, J. S. (1999). Dendritic cell secretion of IL-15 is induced by recombinant huCD40LT and augments the stimulation of antigen-speciﬁc cytolytic T cells. Cell Immunol. 193, 48. Lantz, O., Grandjean, I., Matzinger, P., and Di Santo, J. P. (2000). Gamma chain required for naive CD4þ T cell survival but not for antigen proliferation. Nat. Immunol. 1, 54. Lanzavecchia, A., and Sallusto, F. (2002). Progressive differentiation and selection of the ﬁttest in the immune response. Nat. Rev. Immunol. 2, 982. Lauvau, G., Vijh, S., Kong, P., Horng, T., Kerksiek, K., Serbina, N., Tuma, R. A., and Pamer, E. G. (2001). Priming of memory but not effector CD8 T cells by a killed bacterial vaccine. Science 294, 1735. Le Bon, A., and Tough, D. F. (2002). Links between innate and adaptive immunity via type I interferon. Curr. Opin. Immunol. 14, 432. Lee, H. W., Park, S. J., Choi, B. K., Kim, H. H., Nam, K. O., and Kwon, B. S. (2002). 4-1BB promotes the survival of CD8þ T lymphocytes by increasing expression of Bcl-xL and Bﬂ-1. J. Immunol. 169, 4882. Lefrancois, L., Altman, J. D., Williams, K., and Olson, S. (2000). Soluble antigen and CD40 triggering are sufﬁcient to induce primary and memory cytotoxic T cells. J. Immunol. 164, 725. Lertmemongkolchai, G., Cai, G., Hunter, C. A., and Bancroft, G. J. (2001). Bystander activation of CD8þ T cells contributes to the rapid production of IFN-gamma in response to bacterial pathogens. J. Immunol. 166, 1097. Li, X. C., Demirci, G., Ferrari-Lacraz, S., Groves, C., Coyle, A., Malek, T. R., and Strom, T. B. (2001). IL-15 and IL-2: A matter of life and death for T cells in vivo. Nat. Med. 7, 114. Lin, J. X., Migone, T. S., Tsang, M., Friedmann, M., Weatherbee, J. A., Zhou, L., Yamauchi, A., Bloom, E. T., Mietz, J., John, S., et al. (1995). The role of shared receptor motifs and common Stat proteins in the generation of cytokine pleiotropy and redundancy by IL-2, IL-4, IL-7, IL-13, and IL-15. Immunity 2, 331. Livingstone, A. M., and Kuhn, M. (2002). Peptide-pulsed splenic dendritic cells prime long-lasting CD8(þ) T cell memory in the absence of cross-priming by host APC. Eur. J. Immunol. 32, 281.

MEMORY T CELLS

227

Lodolce, J. P., Boone, D. L., Chai, S., Swain, R. E., Dassopoulos, T., Trettin, S., and Ma, A. (1998). IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation. Immunity 9, 669. Lodolce, J. P., Burkett, P. R., Boone, D. L., Chien, M., and Ma, A. (2001). T cellindependent interleukin 15Ralpha signals are required for bystander proliferation. J. Exp. Med. 194, 1187. Lodolce, J. P., Burkett, P. R., Koka, R. M., Boone, D. L., and Ma, A. (2002). Regulation of lymphoid homeostasis by interleukin-15. Cytokine Growth Factor Rev. 13, 429. Ludewig, B., Ehl, S., Karrer, U., Odermatt, B., Hengartner, H., and Zinkernagel, R. M. (1998). Dendritic cells efﬁciently induce protective antiviral immunity. J. Virol. 72, 3812. Ludewig, B., Odermatt, B., Ochsenbein, A. F., Zinkernagel, R. M., and Hengartner, H. (1999). Role of dendritic cells in the induction and maintenance of autoimmune diseases. Immunol. Rev. 169, 45. Ma, A., Boone, D. L., and Lodolce, J. P. (2000). The pleiotropic functions of interleukin 15: Not so interleukin 2-like after all. J. Exp. Med. 191, 753. Maki, K., Sunaga, S., Komagata, Y., Kodaira, Y., Mabuchi, A., Karasuyama, H., Yokomuro, K., Miyazaki, J. I., and Ikuta, K. (1996). Interleukin 7 receptor-deﬁcient mice lack gammadelta T cells. Proc. Natl. Acad. Sci. USA 93, 7172. Manjunath, N., Shankar, P., Wan, J., Weninger, W., Crowley, M. A., Hieshima, K., Springer, T. A., Fan, X., Shen, H., Lieberman, J., and von Andrian, U. H. (2001). Effector differentiation is not prerequisite for generation of memory cytotoxic T lymphocytes. J. Clin. Invest. 108, 871. Marrack, P., Kappler, J., and Mitchell, T. (1999). Type I interferons keep activated T cells alive. J. Exp. Med. 189, 521. Marshall, D. R., Turner, S. J., Belz, G. T., Wingo, S., Andreansky, S., Sangster, M. Y., Riberdy, J. M., Liu, T., Tan, M., and Doherty, P. C. (2001). Measuring the diaspora for virus-speciﬁc CD8þ T cells. Proc. Natl. Acad. Sci. USA 98, 6313. Masopust, D., Vezys, V., Marzo, A. L., and Lefrancois, L. (2001). Preferential localization of effector memory cells in nonlymphoid tissue. Science 291, 2413. Mattei, F., Schiavoni, G., Belardelli, F., and Tough, D. F. (2001). IL-15 is expressed by dendritic cells in response to type I IFN, double-stranded RNA, or lipopolysaccharide and promotes dendritic cell activation. J. Immunol. 167, 1179. McGavern, D. B., Christen, U., and Oldstone, M. B. (2002). Molecular anatomy of antigen-speciﬁc CD8(þ) T cell engagement and synapse formation in vivo. Nat. Immunol. 3, 918. McMahon, C. W., Zajac, A. J., Jamieson, A. M., Corral, L., Hammer, G. E., Ahmed, R., and Raulet, D. H. (2002). Viral and bacterial infections induce expression of multiple NK cell receptors in responding CD8(þ) T cells. J. Immunol. 169, 1444. McNally, J. M., Zarozinski, C. C., Lin, M. Y., Brehm, M. A., Chen, H. D., and Welsh, R. M. (2001). Attrition of bystander CD8 T cells during virus-induced T-cell and interferon responses. J. Virol. 75, 5965. Medema, J. P., Schuurhuis, D. H., Rea, D., van Tongeren, J., de Jong, J., Bres, S. A., Laban, S., Toes, R. E., Toebes, M., Schumacher, T. N., Bladergroen, B. A., Ossendorp, F., Kummer, J. A., Melief, C. J., and Offringa, R. (2001). Expression of the serpin serine protease inhibitor 6 protects dendritic cells from cytotoxic T lymphocyte-induced apoptosis: Differential modulation by T helper type 1 and type 2 cells. J. Exp. Med. 194, 657. Mercado, R., Vijh, S., Allen, S. E., Kerksiek, K., Pilip, I. M., and Pamer, E. G. (2000). Early programming of T cell populations responding to bacterial infection. J. Immunol. 165, 6833. Messi, M., Giacchetto, I., Nagata, K., Lanzavecchia, A., Natoli, G., and Sallusto, F. (2003). Memory and ﬂexibility of cytokine gene expression as separable properties of human T(H)1 and T(H)2 lymphocytes. Nat. Immunol. 4, 78.

228

PATRICK R. BURKETT ET AL.

Miga, A. J., Masters, S. R., Durell, B. G., Gonzalez, M., Jenkins, M. K., Maliszewski, C., Kikutani, H., Wade, W. F., and Noelle, R. J. (2001). Dendritic cell longevity and T cell persistence is controlled by CD154-CD40 interactions. Eur. J. Immunol. 31, 959. Min, B., McHugh, R., Sempowski, G. D., Mackall, C., Foucras, G., and Paul, W. E. (2003). Neonates support lymphopenia-induced proliferation. Immunity 18, 131. Mitchell, T. C., Hildeman, D., Kedl, R. M., Teague, T. K., Schaefer, B. C., White, J., Zhu, Y., Kappler, J., and Marrack, P. (2001). Immunological adjuvants promote activated T cell survival via induction of Bcl-3. Nat. Immunol. 2, 397. Moskophidis, D., Lechner, F., Pircher, H., and Zinkernagel, R. M. (1993). Virus persistence in acutely infected immunocompetent mice by exhaustion of antiviral cytotoxic effector T cells. Nature 362, 758. Murali-Krishna, K., and Ahmed, R. (2000). Cutting edge: Naive T cells masquerading as memory cells. J. Immunol. 165, 1733. Murali-Krishna, K., Altman, J. D., Suresh, M., Sourdive, D. J., Zajac, A. J., Miller, J. D., Slansky, J., and Ahmed, R. (1998). Counting antigen-speciﬁc CD8 T cells: A reevaluation of bystander activation during viral infection. Immunity 8, 177. Murali-Krishna, K., Lau, L. L., Sambhara, S., Lemonnier, F., Altman, J., and Ahmed, R. (1999). Persistence of memory CD8 T cells in MHC class I-deﬁcient mice. Science 286, 1377. Musso, T., Calosso, L., Zucca, M., Millesimo, M., Ravarino, D., Giovarelli, M., Malavasi, F., Ponzi, A. N., Paus, R., and Bulfone-Paus, S. (1999). Human monocytes constitutively express membrane-bound, biologically active, and interferon-gamma-upregulated interleukin-15. Blood 93, 3531. Neely, G. G., Robbins, S. M., Amankwah, E. K., Epelman, S., Wong, H., Spurrell, J. C., Jandu, K. K., Zhu, W., Fogg, D. K., Brown, C. B., and Mody, C. H. (2001). Lipopolysaccharide-stimulated or granulocyte-macrophage colony-stimulating factor-stimulated monocytes rapidly express biologically active IL-15 on their cell surface independent of new protein synthesis. J. Immunol. 167, 5011. Nopora, A., and Brocker, T. (2002). Bcl-2 controls dendritic cell longevity in vivo. J. Immunol. 169, 3006. Opferman, J. T., Ober, B. T., and Ashton-Rickardt, P. G. (1999). Linear differentiation of cytotoxic effectors into memory T lymphocytes. Science 283, 1745. Opferman, J. T., Ober, B. T., Narayanan, R., and Ashton-Rickardt, P. G. (2001). Suicide induced by cytolytic activity controls the differentiation of memory CD8(þ) T lymphocytes. Int. Immunol. 13, 411. Peschon, J. J., Morrissey, P. J., Grabstein, K. H., Ramsdell, F. J., Maraskovsky, E., Gliniak, B. C., Park, L. S., Ziegler, S. F., Williams, D. E., Ware, C. B., et al. (1994). Early lymphocyte expansion is severely impaired in interleukin 7 receptor-deﬁcient mice. J. Exp. Med. 180, 1955. Polic, B., Kunkel, D., Scheffold, A., and Rajewsky, K. (2001). How alpha beta T cells deal with induced TCR alpha ablation. Proc. Natl. Acad. Sci. USA 98, 8744. Rappl, G., Kapsokefalou, A., Heuser, C., Rossler, M., Ugurel, S., Tilgen, W., Reinhold, U., and Abken, H. (2001). Dermal ﬁbroblasts sustain proliferation of activated T cells via membranebound interleukin-15 upon long-term stimulation with tumor necrosis factor-alpha. J. Invest. Dermatol. 116, 102. Rathmell, J. C., Farkash, E. A., Gao, W., and Thompson, C. B. (2001). IL-7 enhances the survival and maintains the size of naive T cells. J. Immunol. 167, 6869. Reinhardt, R. L., Khoruts, A., Merica, R., Zell, T., and Jenkins, M. K. (2001). Visualizing the generation of memory CD4 T cells in the whole body. Nature 410, 101. Revy, P., Sospedra, M., Barbour, B., and Trautmann, A. (2001). Functional antigen-independent synapses formed between T cells and dendritic cells. Nat. Immunol. 2, 925.

MEMORY T CELLS

229

Ridge, J. P., Di Rosa, F., and Matzinger, P. (1998). A conditioned dendritic cell can be a temporal bridge between a CD4þ T-helper and a T-killer cell. Nature 393, 474. Rogers, P. R., Song, J., Gramaglia, I., Killeen, N., and Croft, M. (2001). OX40 promotes Bcl-xL and Bcl-2 expression and is essential for long-term survival of CD4 T cells. Immunity 15, 445. Rothstein, T. L., Mage, M., Jones, G., and McHugh, L. L. (1978). Cytotoxic T lymphocyte sequential killing of immobilized allogeneic tumor target cells measured by time-lapse microcinematography. J. Immunol. 121, 1652. Sadlack, B., Merz, H., Schorle, H., Schimpl, A., Feller, A. C., and Horak, I. (1993). Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75, 253. Sallusto, F., Lenig, D., Forster, R., Lipp, M., and Lanzavecchia, A. (1999). Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401, 708. Schluns, K. S., Kieper, W. C., Jameson, S. C., and Lefrancois, L. (2000). Interleukin-7 mediates the homeostasis of naive and memory CD8 T cells in vivo. Nat. Immunol. 1, 426. Schluns, K. S., Williams, K., Ma, A., Zheng, X. X., and Lefrancois, L. (2002). Cutting edge: Requirement for IL-15 in the generation of primary and memory antigen-speciﬁc CD8 T cells. J. Immunol. 168, 4827. Schober, S. L., Kuo, C. T., Schluns, K. S., Lefrancois, L., Leiden, J. M., and Jameson, S. C. (1999). Expression of the transcription factor lung Kruppel-like factor is regulated by cytokines and correlates with survival of memory T cells in vitro and in vivo. J. Immunol. 163, 3662. Schoenberger, S. P., Toes, R. E., van der Voort, E. I., Offringa, R., and Melief, C. J. (1998). T-cell help for cytotoxic T lymphocytes is mediated by CD40–CD40L interactions. Nature 393, 480. Selin, L. K., Lin, M. Y., Kraemer, K. A., Pardoll, D. M., Schneck, J. P., Varga, S. M., Santolucito, P. A., Pinto, A. K., and Welsh, R. M. (1999). Attrition of Tcell memory: Selective loss of LCMV epitopespeciﬁc memory CD8 T cells following infections with heterologous viruses. Immunity 11, 733. Shtrichman, R., and Samuel, C. E. (2001). The role of gamma interferon in antimicrobial immunity. Curr. Opin. Microbiol. 4, 251. Slifka, M. K., Pagarigan, R. R., and Whitton, J. L. (2000). NK markers are expressed on a high percentage of virus-speciﬁc CD8þ and CD4þ T cells. J. Immunol. 164, 2009. Slifka, M. K., Rodriguez, F., and Whitton, J. L. (1999). Rapid on/off cycling of cytokine production by virus-speciﬁc CD8þ T cells. Nature 401, 76. Slifka, M. K., and Whitton, J. L. (2000). Antigen-speciﬁc regulation of T cell-mediated cytokine production. Immunity 12, 451. Slifka, M. K., and Whitton, J. L. (2001). Functional avidity maturation of CD8(þ) T cells without selection of higher afﬁnity TCR. Nat. Immunol. 2, 711. Sprent, J. (2002). T memory cells: Quality not quantity. Curr. Biol. 12, R174. Sprent, J., and Surh, C. D. (2001). Generation and maintenance of memory T cells. Curr. Opin. Immunol. 13, 248. Stefanova, I., Dorfman, J. R., and Germain, R. N. (2002). Self-recognition promotes the foreign antigen sensitivity of naive T lymphocytes. Nature 420, 429. Suresh, M., Whitmire, J. K., Harrington, L. E., Larsen, C. P., Pearson, T. C., Altman, J. D., and Ahmed, R. (2001). Role of CD28-B7 interactions in generation and maintenance of CD8 T cell memory. J. Immunol. 167, 5565. Surh, C. D., and Sprent, J. (2000). Homeostatic T cell proliferation: How far can T cells be activated to self-ligands? J. Exp. Med. 192, F9. Swain, S. L., Hu, H., and Huston, G. (1999). Class II-independent generation of CD4 memory T cells from effectors. Science 286, 1381. Swanson, B. J., Murakami, M., Mitchell, T. C., Kappler, J., and Marrack, P. (2002). RANTES production by memory phenotype T cells is controlled by a posttranscriptional, TCRdependent process. Immunity 17, 605.

230

PATRICK R. BURKETT ET AL.

Tagaya, Y., Bamford, R. N., DeFilippis, A. P., and Waldmann, T. A. (1996). IL-15: A pleiotropic cytokine with diverse receptor/signaling pathways whose expression is controlled at multiple levels. Immunity 4, 329. Tan, J. T., Dudl, E., LeRoy, E., Murray, R., Sprent, J., Weinberg, K. I., and Surh, C. D. (2001). IL-7 is critical for homeostatic proliferation and survival of naive T cells. Proc. Natl. Acad. Sci. USA 98, 8732. Tan, J. T., Ernst, B., Kieper, W. C., LeRoy, E., Sprent, J., and Surh, C. D. (2002). Interleukin (IL)15 and IL-7 jointly regulate homeostatic proliferation of memory phenotype CD8þ cells but are not required for memory phenotype CD4þ cells. J. Exp. Med. 195, 1523. Tan, J. T., Whitmire, J. K., Ahmed, R., Pearson, T. C., and Larsen, C. P. (1999). 4-1BB ligand, a member of the TNF family, is important for the generation of antiviral CD8 T cell responses. J. Immunol. 163, 4859. Tanchot, C., Lemonnier, F. A., Perarnau, B., Freitas, A. A., and Rocha, B. (1997). Differential requirements for survival and proliferation of CD8 naive or memory T cells. Science 276, 2057. Topham, D. J., Castrucci, M. R., Wingo, F. S., Belz, G. T., and Doherty, P. C. (2001). The role of antigen in the localization of naive, acutely activated, and memory CD8(þ) T cells to the lung during inﬂuenza pneumonia. J. Immunol. 167, 6983. Tough, D. F., Borrow, P., and Sprent, J. (1996). Induction of bystander T cell proliferation by viruses and type I interferon in vivo. Science 272, 1947. Tough, D. F., and Sprent, J. (1994). Turnover of naive- and memory-phenotype T cells. J. Exp. Med. 179, 1127. Tough, D. F., Sun, S., and Sprent, J. (1997). T cell stimulation in vivo by lipopolysaccharide (LPS). J. Exp. Med. 185, 2089. Tough, D. F., Zhang, X., and Sprent, J. (2001). An IFN-gamma-dependent pathway controls stimulation of memory phenotype CD8þ T cell turnover in vivo by IL-12, IL-18, and IFNgamma. J. Immunol. 166, 6007. Tuma, R. A., Giannino, R., Guirnalda, P., Leiner, I., and Pamer, E. G. (2002). Rescue of CD8 T cellmediated antimicrobial immunity with a nonspeciﬁc inﬂammatory stimulus. J. Clin. Invest. 110, 1493. van Stipdonk, M. J., Lemmens, E. E., and Schoenberger, S. P. (Stipdonk 2001). Naive CTLs require a single brief period of antigenic stimulation for clonal expansion and differentiation. Nat. Immunol. 2, 423. Veiga-Fernandes, H., Walter, U., Bourgeois, C., McLean, A., and Rocha, B. (2000). Response of naive and memory CD8þ T cells to antigen stimulation in vivo. Nat. Immunol. 1, 47. Vella, A. T., Dow, S., Potter, T. A., Kappler, J., and Marrack, P. (1998). Cytokine-induced survival of activated T cells in vitro and in vivo. Proc. Natl. Acad. Sci. USA 95, 3810. von Freeden-Jeffry, U., Vieira, P., Lucian, L. A., McNeil, T., Burdach, S. E., and Murray, R. (Freeden-Jeffry 1995). Lymphopenia in interleukin (IL)-7 gene-deleted mice identiﬁes IL-7 as a nonredundant cytokine. J. Exp. Med. 181, 1519. Vosshenrich, C. A., and Di Santo, J. P. (2001). Cytokines: IL-21 joins the gamma(c)-dependent network? Curr. Biol. 11, R175. Walzer, T., Marcais, A., Saltel, F., Bella, C., Jurdic, P., and Marvel, J. (2003). Cutting edge: Immediate RANTES secretion by resting memory CD8 T cells following antigenic stimulation. J. Immunol. 170, 1615. Weninger, W., Crowley, M. A., Manjunath, N., and von Andrian, U. H. (2001). Migratory properties of naive, effector, and memory CD8(þ) T cells. J. Exp. Med. 194, 953. Wherry, E. J., Teichgraber, V., Becker, T. C., Masopust, D., Kaech, S. M., Antia, R., Von Andrian, U. H., and Ahmed, R. (2003). Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat. Immunol. 4, 1191.

MEMORY T CELLS

231

Whitmire, J. K., and Ahmed, R. (2000). Costimulation in antiviral immunity: Differential requirements for CD4(þ) and CD8(þ) T cell responses. Curr. Opin. Immunol. 12, 448. Whitmire, J. K., Asano, M. S., Murali-Krishna, K., Suresh, M., and Ahmed, R. (1998). Long-term CD4 Th1 and Th2 memory following acute lymphocytic choriomeningitis virus infection. J. Virol. 72, 8281. Whitmire, J. K., Slifka, M. K., Grewal, I. S., Flavell, R. A., and Ahmed, R. (1996). CD40 liganddeﬁcient mice generate a normal primary cytotoxic T-lymphocyte response but a defective humoral response to a viral infection. J. Virol. 70, 8375. Willerford, D. M., Chen, J., Ferry, J. A., Davidson, L., Ma, A., and Alt, F. W. (1995). Interleukin-2 receptor alpha chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3, 521. Wong, P., and Pamer, E. G. (2001). Cutting edge: Antigen-independent CD8 T cell proliferation. J. Immunol. 166, 5864. Wu, C. Y., Kirman, J. R., Rotte, M. J., Davey, D. F., Perfetto, S. P., Rhee, E. G., Freidag, B. L., Hill, B. J., Douek, D. C., and Seder, R. A. (2002a). Distinct lineages of T(H)1 cells have differential capacities for memory cell generation in vivo. Nat. Immunol. 3, 852. Wu, T. S., Lee, J. M., Lai, Y. G., Hsu, J. C., Tsai, C. Y., Lee, Y. H., and Liao, N. S. (2002b). Reduced expression of Bcl-2 in CD8þ T cells deﬁcient in the IL-15 receptor alpha-chain. J. Immunol. 168, 705. Xu, J., Foy, T. M., Laman, J. D., Elliott, E. A., Dunn, J. J., Waldschmidt, T. J., Elsemore, J., Noelle, R. J., and Flavell, R. A. (1994). Mice deﬁcient for the CD40 ligand. Immunity 1, 423. Xue, H. H., Kovanen, P. E., Pise-Masison, C. A., Berg, M., Radovich, M. F., Brady, J. N., and Leonard, W. J. (2002). IL-2 negatively regulates IL-7 receptor alpha chain expression in activated T lymphocytes. Proc. Natl. Acad. Sci. USA 99, 13759. Yajima, T., Nishimura, H., Ishimitsu, R., Yamamura, K., Watase, T., Busch, D. H., Pamer, E. G., Kuwano, H., and Yoshikai, Y. (2001). Memory phenotype CD8(þ) T cells in IL-15 transgenic mice are involved in early protection against a primary infection with Listeria monocytogenes. Eur. J. Immunol. 31, 757. Yamada, H., Matsuzaki, G., Chen, Q., Iwamoto, Y., and Nomoto, K. (2001). Reevaluation of the origin of CD44(high) ‘‘memory phenotype’’ CD8 T cells: Comparison between memory CD8 T cells and thymus-independent CD8 T cells. Eur. J. Immunol. 31, 1917. Zhang, X., Fujii, H., Kishimoto, H., LeRoy, E., Surh, C. D., and Sprent, J. (2002). Aging leads to disturbed homeostasis of memory phenotype CD8(þ) cells. J. Exp. Med. 195, 283. Zhang, X., Sun, S., Hwang, I., Tough, D. F., and Sprent, J. (1998). Potent and selective stimulation of memory-phenotype CD8þ T cells in vivo by IL-15. Immunity 8, 591. Zinkernagel, R. M. (2000). What is missing in immunology to understand immunity? Nat. Immunol. 1, 181. Zinkernagel, R. M. (2002). On differences between immunity and immunological memory. Curr. Opin. Immunol. 14, 523.

This Page Intentionally Left Blank

advances in immunology, vol. 83

CD81 Effector Cells PIERRE A. HENKART AND MARTA CATALFAMO National Institutes of Health, Bethesda, Maryland 20892-1360

I. Effector Cells Defined

A survey of the literature shows that the term ‘‘effector cells’’ means different things to different immunologists. The word ‘‘effector’’ clearly implies cells with functional activity, but there is no agreement as to what type of activity. For many years, cytotoxicity was the only lymphocyte functional activity that was measurable in short term in vitro assays, and ‘‘effector cells’’ were thus understood to mean cells with such activity. Other in vitro lymphocyte functional activities were clearly more complex, potentially involving multiple cellular activities, and the term ‘‘effector’’ did not seem meaningful if it had to be associated with a particular assay. When it later became possible to quantitatively measure secretion of deﬁned cytokines in short-term in vitro cultures, the term ‘‘effector’’ was logically adopted to describe activity in such assays, and effector cells lost their close association with cytotoxicity. The term effector thus applies to several different types of functions and, for the sake of clarity, the effector function needs to be speciﬁed. Confusion can arise from the multiple types of effector function that now can be measured, not only cytotoxic effector function vs cytokine secretion, but also the two different cytotoxicity pathways and the multiple cytokines T cells make, which are not all regulated similarly during differentiation. Thus, we will refer to cytotoxic effector function and cytokine secretory effector function. The term ‘‘effector’’ has often been used to describe lymphocyte differentiation, particularly in the context of distinguishing memory cells from effector cells. Development of TcR transgenic animals and MHC tetramer staining has allowed a functional deﬁnition of memory cells as those antigen-reactive cells that persist after an immune response and give rise to the enhanced secondary response characteristic of immunological memory. Memory function is thus deﬁned in an in vivo context over a time span of weeks, months, or years, as opposed to effector function that operates within hours. There is no logical reason why memory function should exclude effector function, and indeed most memory cells seem to have cytokine secretory effector functions and, in some cases, also seem to have cytotoxic effector function (see following text). Even more confusion can occur when the term effector is used by some groups to describe cells that gave rise to longer-term biological functions in vivo, such as tumor therapy after adoptive transfer. While this use of the term effector can be defended as logical, it is very different from more 233 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

234

PIERRE A. HENKART AND MARTA CATALFAMO

common usage to describe differentiated cells showing immediate and deﬁned functional activity, because many cellular and molecular interactions occur after injecting cells into an animal and observing their effects some days or weeks later. We would argue that since cytotoxicity and cytokine secretion are the only responses that can presently be measured in short-term assays reﬂecting a deﬁned differentiated T cell phenotype, the term effector should apply only to these functional responses, and we use the term effector accordingly in this chapter. The term effector has also been applied to CD8þ T cell subpopulations deﬁned by phenotypic surface markers, based on correlations of effector function with that subpopulation. This terminology is a result of extensive efforts to identify such correlations, with considerable success for human CD8þ T cells. There is a consensus that naive and cytotoxic effector cells express the CD45RA isoform while memory cells switch to express the smaller CD45RO isoform lacking the CD45RA, B, and C domains. The combination of the CD45 isoform marker with another marker, such as CD27, CD28, CCR7, or CD62L, has been used to delineate four subpopulations, with cytotoxic effector function correlating with low expression of the second marker as well as expression of CD45RA. The same second markers have been proposed to correlate with two functionally distinct subsets of memory cells. The central memory cells bear surface receptors mediating homing to lymphoid organs (such as CD62L) and the CCR7 receptor for chemokines such as CCL19, allowing them to circulate in lymphoid organs. An ‘‘effector memory’’ subpopulation of cells lacking these surface markers is proposed to circulate in nonlymphoid tissue (Sallusto et al., 1999). However, there is presently no consensus as to the precise criteria for deﬁning these memory subpopulations nor, as will be discussed, is there a consensus on whether they also have different effector functions. In any case, to avoid confusion, it should be either speciﬁed or clear from the context when the term effector is used to describe a subpopulation based on surface phenotype vs direct measurements of function. While expression of cytoplasmic perforin or granzymes indicates a potential for cytotoxic effector activity, these granule mediators are necessary but not sufﬁcient for cytotoxicity. An interesting transgenic mouse model to identify effector CD8þ T cells has been constructed, using GFP expressed under the control of the CD4 promoter (T-GFP mice) (Manjunath et al., 1999). As CD8þ T cells differentiate into cytotoxic effector cells, the CD4 promoter becomes inactive. Seven days after T-GFP mice were challenged with vaccinia, all the cytotoxic effector activity was found in the GFP CD8þ T cells, while only the GFPþ cells proliferated in response to TcR ligation. Our previously stated functional deﬁnition of effector cells raises the important caveat that any strictly deﬁned in vitro assay may not be meaningfully

CD8þ EFFECTOR CELLS

235

extrapolated to the in vivo biological phenomenon one is trying to understand. This limitation must be accepted as part of the process of simplifying systems to allow studies under deﬁned conditions. In vitro functional assays should be viewed as a tool to dissect in vivo immune responses in view of the difﬁculties that confront manipulating lymphocytes in vivo. There is no argument against the need for better tools for this, and it is hoped that new technologies will provide more sophisticated means to analyze lymphocyte functions in vivo so that correlations can be developed with the in vitro measurements used to deﬁne effector function. II. Secretion as the Mechanism of Effector Function

The two major molecular pathways responsible for cytotoxic effector function are now understood, and they both rely on the basic cellular process of secretion (Barry and Bleackley, 2002; Russell and Ley, 2002; Trapani and Smyth, 2002). The perforin-dependent granule exocytosis pathway utilizes a regulated secretory mechanism, which is well known in other cell types but in lymphocytes has been exclusively associated with cytotoxicity. This type of secretion involves previously synthesized proteins that are stored intracellularly in secretory granules. In response to receptor crosslinking, the granule membranes undergo fusion with the plasma membrane, which is exocytosis. Lymphocyte secretory granules are similar to those of other hemapoietic cells in that they are modiﬁed lysosomes, containing the usual lysosomal enzymes as well as the secretory proteins and proteoglycans (Stinchcombe and Grifﬁths, 1999). The FasL/Fas pathway involves membrane receptor expression of FasL, which appears to utilize both the regulated secretory pathway in some circumstances and the constitutive secretory pathway in others (Kojima et al., 2002). In the latter case, newly synthesized protein is deposited in the ER, processed in the Golgi, and shuttled to the plasma membrane in small vesicles, where immediate exocytosis results in surface expression. Soluble FasL can sometimes be released later after cleavage of the membrane form by a metalloprotease. In terms of the stated deﬁnition of effector cells, thinking about cytotoxicity in terms of its functional secretory pathways is attractive as a logical connection with functional measurements of immediate cytokine secretion. It should be noted that the major T cell cytokines studied to date all utilize the constitutive secretory pathway. Although their overall secretion is obviously triggered by the TcR, regulation occurs at the level of transcription and protein synthesis, and not the secretory event itself. The classical T cell cytokines are not synthesized previously and stored, as is the case with cytotoxic mediators. Detection of intracellular perforin and granzymes in lymphocytes suggests that they have the potential to be cytotoxic effectors, but it is not clear that

236

PIERRE A. HENKART AND MARTA CATALFAMO

such cells have the ability to exocytose these mediators after TcR engagement. On the other hand, detection of intracellular cytokine indicates that secretion is occurring. In order for the critical effector secretion events to occur, precursor lymphocytes having the ability to recognize antigen generally must have differentiated into cells that also have the ability to secrete rapidly, allowing their detection in short-term assays. There is unfortunately no universal deﬁnition of ‘‘shortterm,’’ although typically 3- to 6-hour incubations are used. Assays of longer duration, e.g., 12 to 24 hours, may be an appropiate readout of effector function in some cases (see following text), but such longer times open the door to measuring antigen-triggered differentiation of precursors into effectors as well as effector function. This issue becomes relevant in studies of memory CD8þ T cells, some of which have no effector function in short-term (4- to 5hour) cytotoxic assays, but acquire it rapidly enough to kill target cells in overnight assays of 16 to 24 hours (Wherry et al., 2003). III. Cytotoxicity In Vivo

Ideally, one would like to be able to measure cytotoxic activity in vivo by measuring target cell death and characterizing the effector cell. One very real problem is that cytotoxic lymphocytes induce an apoptotic response in their target cells, and apoptotic membrane changes trigger phagocytosis by local tissue phagocytes. Thus, any in vivo cytotoxicity assay based on identifying dying target cells faces the difﬁculty that apoptotic cells may rapidly disappear. In favorable cases, CTL-mediated target cell degeneration can be observed in vivo by histological studies or by in situ assays of DNA fragmentation (Ando et al., 1994). However, such approaches may miss many cases of cytotoxicity in vivo, and the nature of the effector cells may be difﬁcult to elucidate. Instead of looking for dead cells, the disappearance of cells of interest can be assumed to be due to their death, and indeed the rejection of grafts or tumors has classically been used as an in vivo measure of cytotoxicity. The importance of CTL was ﬁrst recognized when the disappearance of allogeneic tumors injected into the peritoneal cavity correlated with the appearance of CTL displaying cytotoxic function in vitro (Berke, 1980). A newer approach to measure cytotoxicity in vivo is to inject CFSE-labeled target cells into animals and follow their survival by subsequently sacriﬁcing mice and analyzing the labeled cells using ﬂow cytometry. If different levels of CFSE labeling are used to distinguish between antigen-bearing cells and control cells, this approach can assay for antigen-speciﬁc survival in vivo (Oehen and Brduscha-Riem, 1998). A signiﬁcant issue with measurements of loss of cells from a particular in vivo site is that one may be seeing a redistribution to other organs instead of cytotoxicity, requiring a search of likely compartments to check this possibility.

CD8þ EFFECTOR CELLS

237

In all the above in vivo assays, the nature of the effector cells and the molecular pathways responsible for antigen loss in vivo are difﬁcult to assess and must be approached indirectly, e.g., by selectively manipulating the target cells or animals to block the cytotoxic effect. A major problem with such approaches is that while they can implicate a particular cell or molecule as required for the effect, they may not reﬂect the nature of the direct effector responsible for the depletion of cells in vivo as opposed to one required component in a pathway required for this effect. Effector cells are deﬁned as those directly responsible for a biological phenomenon rather than cells whose activity is required for an upstream step. In some cases, it has been assumed that because effector cells exhibit good cytotoxic activity in vitro, these cells will use the same cytotoxicity pathway to eliminate antigen in vivo. For controlling some viral infections in vivo, the noncytopathic CTL-mediated secretion of IFN-g and TNF-a may be more important than either perforin- or Fas-dependent cytotoxicity (Guidotti and Chisari, 2001). Cell loss in vivo can be due to different mechanisms, and identiﬁcation of potential effector cells and molecular mechanisms can best be done in vitro. However, it is critically important to go back to in vivo systems to test the effector mechanisms proposed from in vitro studies. In some cases, where particular cytotoxicity pathways predominate in vitro, they were found to be not required in vivo (Barchet et al., 2000; Guidotti et al., 1996). Knockout mice lacking perforin and the gld mutant mice defective in FasL have been used to study a variety of in vivo processes for their dependence on their respective cytotoxicity pathways (Kagi et al., 1996; Russell and Ley, 2002). However, this approach requires caution because these mice have abnormal immune systems, compatible with the idea that these cytotoxicity pathways play an important regulatory role in immune responses. Gld mice, in particular, develop lymphoid hypertrophy and, as they age, they accumulate large numbers of nonfunctional lymphocytes (Cohen and Eisenberg, 1991). Perforin knockout mice do not show a normal homeostatic downregulation of expanded T cell numbers after antigenic challenge (Harty and Badovinac, 2002). Perforin knockout mice have provided clear evidence for the operation of this cytotoxicity pathway in vivo, in particular, in resistance to noncytopathic viruses (Kagi et al., 1996). More recently, a role for perforin-dependent cytotoxicity in tumor surveillance has become clear as older perforin knockout mice develop a high rate of spontaneous tumors (Smyth et al., 2000), although this may be due to NK or NKT cells rather than CD8þ CTL (Hayakawa et al., 2002). Perhaps the most dramatic evidence of perforin’s in vivo role came from the identiﬁcation of functional perforin mutations as one cause of the human disease familial hematophagocytic syndrome (FHS) (de Saint Basile and Fischer, 2001; Stepp et al., 1999). After viral infections, FHS children with perforin mutations suffer from massive and potentially fatal tissue inﬁltration

238

PIERRE A. HENKART AND MARTA CATALFAMO

by activated lymphocytes and macrophages, a phenomenon not described in perforin knockout mice. IV. Cytotoxicity In Vitro

Because of the difﬁculties previously outlined, cytotoxic effectors have come to be operationally deﬁned as cells whose cytotoxic function can be detected in short-term in vitro assays. These assays of 3 to 6 hours utilize titrated numbers of effector cells incubated with ﬁxed numbers of target cells, with target lysis classically measured by the 51Cr release assay. Other assays measuring cytolysis are also used, including some measuring release of heavy metal chelates or hydrophilic ﬂuorophores. Target cell apoptosis can also be measured using nuclear DNA fragmentation, or caspase activation using ﬂuorogenic substrates and ﬂow cytometry. All these assays quantitate target cell death, but do not enumerate or directly identify the effector cells responsible. Thus, attribution of effector function to a particular lymphocyte population relies on puriﬁcation of the input effector population in cytotoxicity assays. In the past, cytotoxic effector function was difﬁcult to demonstrate from lymphocytes directly ex vivo, except in patients/animals with acute virus infections or in lymphocytes removed from the peritoneal cavity where allogeneic tumors are undergoing rejection. Cytotoxic function was generally measured from ex vivo lymphocytes (e.g., from human blood) only after culture with antigen and APC for 1 to 2 weeks, followed by a standard short-term assay for cytotoxic function. This approach actually measures an in vitro memory function rather than cytotoxic effector function, as has been deﬁned. However, recent use of TcR transgenic animals has allowed detection of direct ex vivo cytotoxic function since antigen-reactive cells can be identiﬁed and target lysis measured after pulsing with appropriate peptides (see following text). V. Perforin-dependent Granule Exocytosis Pathway

The perforin-dependent granule exocytosis pathway often dominates cytotoxicity measured by in vitro assays. This pathway is triggered by TcR recognition of target antigen, leading to a rapid cytoplasmic polarization of the T cell so that many of its cytoplasmic organelles face the bound target cell, and the granules become positioned for exocytosis. Recent studies of microtubules and MTOC in CTL-target interactions provide insights into this process (Kuhn and Poenie, 2002), but many molecular steps between TcR signaling and MTOC reorganization remain unknown. After polarization, secretion of previously synthesized perforin and granzymes stored in lysosomal secretory granules occurs by exocytosis, which is the fusion of the granule membranes with the plasma membrane.

CD8þ EFFECTOR CELLS

239

Rab 27a has been recently identiﬁed as a molecule required between the CTL polarization and granule fusion steps. This small GTPase family member is defective in humans with one form of Griscelli’s syndrome, a rare genetic disease associated with pigmentation defects (Menasche et al., 2000). Rab27a mutations give rise to a form of Griscelli’s syndrome that is associated with a hemophagocytic syndrome similar to that seen in perforin-defective humans, and CTL from these patients show defective cytotoxic effector activity. The naturally occurring mouse ashen mutation is similarly associated with a pigmentation defect and a defect in Rab27a (Wilson et al., 2000). CTL and NK cells from ashen mice show minimal cytotoxic activity on Fas target cells in vitro, although cytotoxic activity via the Fas pathway is normal (Haddad et al., 2001; Stinchcombe et al., 2001a). TcR-triggered granule exocytosis is defective in ashen CTL, although TcR-induced granule polarization occurs normally. Thus, Rab27a is required for a late step in the regulated secretory pathway in cytotoxic lymphocytes, perhaps in aligning the granule membrane with the plasma membrane to promote the fusion step. TcR-triggered granule exocytosis occurs within the central region of the immunological synapse in CTL (Stinchcombe et al., 2001b). For this secretion to result in cytotoxicity to the target cell, effector cell granules must contain either perforin and granzymes or FasL on the inner surface of the granule membrane. When triggered by immobilized antibodies against the TcR, CTL degranulation rarely releases more than 50% of the total granule contents. Since CTL can kill multiple target cells within a few hours, antigen on target cells also appears to trigger exocytosis of only a fraction of the granules. It is not clear how many of these granule mediators are required for target cell death, although their local concentration in the small volume between effector and bound target cell can be very high, even after an inefﬁcient degranulation. Perforin and granzymes are required for cytotoxicity via the perforindependent granule exocytosis pathway. Perforin is a water-soluble 65kD protein that undergoes a calcium-dependent aggregation leading to membrane insertion and pore formation. Large porelike structures can be seen by microscopy after CTL or NK killing (Dourmashkin et al., 1980; Podack and Dennert, 1983), although it is not certain that these structures are required for target death. The sieving behavior of macromolecular markers resealed inside red cell ghosts attacked by NK cells indicated that proteins up to 500kD but not larger could escape, compatible with the observed structural pores (Simone and Henkart, 1980), and electrical measurements on planar lipid bilayers demonstrate perforin’s ability to form functional membrane pores (Young and Cohn, 1986). However, no satisfactory molecular mechanism has been elucidated to explain perforin’s oligomerization, membrane insertion, and pore formation. The recognition that a lipid-binding C2 domain is activated by a

240

PIERRE A. HENKART AND MARTA CATALFAMO

granule proteolytic processing step is an important step in assigning functional domains to this molecule (Uellner et al., 1997). Because perforin knockout CTL show a complete lack of cytotoxic effector function on Fas targets, perforin is clearly necessary for this cytotoxicity pathway. However, pore formation by perforin in target membranes is not the lethal injury resulting in the death of nucleated target cells. When mast cells were transfected to express perforin in their granules, they became highly cytotoxic to red blood cell targets but had minimal ability to kill nucleated target cells (Shiver and Henkart, 1991). Only when granzymes A and B were expressed in addition to perforin did mast cells attain cytotoxic activity comparable to CTL (Nakajima et al., 1995). Thus, perforin’s required role in this cytotoxicity pathway appears to be the facilitation of granzyme entry into target cells. The mechanism of this permeabilization is still not clear. Perforin pores inserted into target membranes may be sealed by normal repair processes to avoid lysis, but while they are open, they may allow granzyme entry that actually leads to target death. However, granzymes are secreted as a complex with large proteoglycans; it is not clear that perforin pores are large enough to allow their passage through the target membrane (Metkar et al., 2002; Raja et al., 2002). An alternative model for perforin’s function is that it permeabilizes endosomal vesicles in target cells after membrane-bound granzymes and perforin are taken up from the surface by endocytosis (Froelich et al., 1996), but whether this occurs in lethally injured CTL target cells is still unclear. Granzymes are a subfamily of serine proteases that are uniquely expressed in CTL and NK cells (Kam et al., 2000), with granzymes A and B the predominant enzymes in CTL in vivo. These proteases have very different substrate speciﬁcity, with the former cleaving after lysine and the latter after aspartic acid. Because granzyme B can directly process and activate procaspases that are widely expressed in the cytoplasm of all cells, introduction of granzyme B into the target cytoplasm provides a clear route to target cell apoptosis and death (Darmon et al., 1995). However, it appears that granzyme B-induced apoptosis is more complex, requiring mitochondrial damage to suppress the naturally occurring caspase inhibitors (Goping et al., 2003; Sutton et al., 2003). When caspase activation is blocked by caspase inhibitors, target cell lysis by CTL still occurs normally, although apoptotic nuclear changes are blocked along with some apoptotic cytoplasmic changes (Sarin et al., 1997, 1998). Thus, caspase-independent death pathways can also lead to rapid target cell death. Granzyme B itself cleaves a number of target cell proteins, and it has been suggested that these cleavages might be important in the generation of some forms of autoimmunity (Andrade et al., 1998; Casciola-Rosen et al., 1999). Unlike perforin knockout mice, CTL from granzyme knockout mice retain cytotoxic effector activity in this pathway, although granzyme B knockout CTL

CD8þ EFFECTOR CELLS

241

cause substantially less target apoptotic activity and somewhat less lysis than control CTL (Heusel et al., 1994; Shresta et al., 1995). The lack of a strong cytotoxicity defect in granzyme-deﬁcient CTL can be explained by the potential redundancy of granzyme proteases, which include several that are not well characterized. Granzyme A is the other well-expressed granzyme in primary CTL, and it does not directly process procaspases like granzyme B. When introduced into the cytoplasm of target cells, granzyme A triggers a novel form of apoptosis, associated with DNA single-strand breaks (Beresford et al., 1999, 2001). This target damage pathway is initiated by granzyme A cleavage of SET, part of an ER complex containing an inactive DNAse which becomes active and translocates to the nucleus (Fan et al., 2003a, 2002). Another substrate of granzyme is the oxidative repair protein Ape-1, whose degradation may contribute to the initiation of apoptosis in target cells (Fan et al., 2003b). CTL from granzyme A knockout mice have no detectable defect in cytotoxic effector function (Ebnet et al., 1995), although they are defective in their ability to resist some virus infections (Mullbacher et al., 1996; Pereira et al., 2000; Riera et al., 2000). A model system for studying target apoptosis induced by secreted perforin and granzymes has been extensively investigated by several groups. This involves addition of puriﬁed granzymes and sublytic amounts of perforin or other membrane permeabilizing agents to cells in medium. This model allows the use of deﬁned cytotoxic mediators, but their concentration at the target membrane is many orders of magnitude less than occurs after CTL exocytosis. As reviewed elsewhere, granzyme B-induced apoptosis has been extensively studied with this system (Barry and Bleackley, 2002; Smyth et al., 2001). Studies in this model system have led to the proposal that granzyme B binds to target cell surfaces by a mannose-6-phosphate receptor (Motyka et al., 2000), but critical ﬁndings supporting this model are controversial (Trapani et al., 2003). The issue of how cytotoxic effector cells themselves are not killed after degranulation has long been a subject of discussion. A recent model to explain this proposes that the lysosomal protease cathepsin B is exposed on the effector surface after exocytosis, allowing proteolytic degradation of cytotoxic mediators coming back to the effector cell (Balaji et al., 2002). In support of this model, it was shown that active surface cathepsin B appears on the CTL surface after degranulation, and its inhibition sensitizes CTL to rapid perforin-dependent suicide upon degranulation. From a practical standpoint, comparing in vitro cytotoxic function of CTL from perforin knockout mice relative to normal mice has often been used to implicate the involvement of this cytotoxicity pathway. Another approach to test the operation of this pathway in vitro is via the use of concanamycin A, a

242

PIERRE A. HENKART AND MARTA CATALFAMO

drug that blocks the granule proton pump and leads to proteolytic degradation of perforin within the granules (Kataoka et al., 1997). VI. FasL/Fas Death Pathway

The second pathway for cytotoxic effector function is the FasL/Fas pathway. In this case, TcR ligation leads to surface FasL expression, either by de novo transcription and translation followed by FasL surface expression, or by exocytosis of preformed FasL resident on the inside of the lysosomal secretory granule membranes containing perforin and granzymes (Bossi and Grifﬁths, 1999; Haddad et al., 2001; Kojima et al., 2002). After its biosynthesis, FasL is localized to granules by virtue of a proline-rich motif in its carboxy terminus (Blott et al., 2001). Exocytosis results in surface expression of this FasL as a result of the fusion of the granule membrane with the plasma membrane. Cytotoxicity by the FasL/Fas pathway requires that the target cell express a functional Fas death pathway, which includes not only surface Fas expression measurable by ﬂow cytometry, but also a functionally intact internal death pathway that is not blocked by internal death pathway inhibitors, such as FLIPs. Blocking the surface FasL –Fas interactions with noncrosslinking IgG anti-Fas antibodies or soluble Fas–IgG constructs provides an alternative approach to implicating the Fas pathway in functional cytotoxicity assays. Whether effector FasL expression results from exocytosis of preformed FasL in granules or by de novo synthesis can be addressed by use of inhibitors of protein/RNA synthesis. The relative use of the two cytotoxic effector pathways (perforin-dependent vs Fas-dependent) in vivo will depend on properties of both effector and target cells (Kafrouni et al., 2001). In vivo cytotoxicity by CD8þ T cells may utilize other pathways that do not seem to dominate in vitro. In particular, TNF-a was originally identiﬁed as a mediator of tumor cytotoxicity in vitro, but although CTL secrete this cytokine, it does not seem to contribute to short term CTL-mediated cytotoxicity in vitro (Ratner and Clark, 1993). However, when the two major cytotoxicity pathways are inoperative, longer-term (16- to 24-hour) assays may detect cytotoxicity due to membrane-bound or secreted forms of TNF-a expressed by CTL. Although such cytotoxicity could be due to antigen-induced differentiation followed by mediator secretion, some target cells die slowly in the presence of high concentrations of puriﬁed TNF-a, so it is possible that longer assays are required to measure this cytotoxicity pathway. In vivo CD8þ T cell-mediated tumor cell killing by this route could be important, particularly for immunotherapy. With target cells that are susceptible to TNF-a-induced apoptosis, the potential for cytotoxicity by secreted TNF-a would appear to be best assayed by measuring TNF-a secretion as will be described.

CD8þ EFFECTOR CELLS

243

VII. Cytokine Secretory Effector Cells

Detection of cytokine mRNA by in situ hybridization or RT-PCR can be used to show activation of cytokine genes in tissue sections, principally to aid in clinical diagnosis. However, since cytokine secretion may not be tightly linked to transcription or mRNA levels, this would not appear to be a satisfactory readout of effector function. Immunohistochemical detection of cytokine protein in tissues has also been reported, but this appears technically demanding and is not quantitative. More satisfying results have come from analysis of cytokine secretion in short cultures of the effector cells in the presence of antigenic stimuli, with measurement of the resulting cytokine secretion by several approaches. Various protein or peptide antigenic stimuli have been used, along with various APC anti costimulating antibodies. Other stimuli used include a combination of phorbol myristyl acetate (PMA) and ionomycin, anti-CD3 with or without costimuling antibodies such as antiCD28, or superantigens. These assays are parallel to cytotoxicity assays except that secretion is measured, using several different approaches as will be described. 1. A straightforward assay of cytokine secreted into the medium by ELISA or proliferation assays, which is homologous to cytotoxicity assays in that a quantitative measure of secretion by the input cell population is obtained. Supernatant-induced proliferation of selected cell lines was formerly widely used to measure cytokines, but the identity of the cytokine responsible may be in doubt because some responding cell lines have multiple cytokine receptors, each of which triggers proliferation. Neutralization by speciﬁc antibodies can help satisfy such concerns, but proliferation-based cytokine detection assays have largely been replaced by molecularly based detection systems like ELISA. 2. The ﬁrst assay developed which read out the number of cells secreting a speciﬁc cytokine of interest was the ELISPOT assay, in which cytokinesecreting cells are cultured with antigen or other stimulus (typically, for 24 hours) on a surface containing immobilized anticytokine antibody so that the secreted cytokine is captured and detected with a second anticytokine antibody (Czerkinsky et al., 1991). The readout is a visual count of positive cells based on spots of captured cytokine. 3. Detection of cytokines secreted by individual living activated T cells using a capture technique employing heteroconjugated antibodies between the CD45 surface antigen and an anticytokine antibody. The captured cytokine is then detected with a second, ﬂuorochrome-conjugated anticytokine antibody (Brosterhus et al., 1999). This method can permit simultaneous characterization of surface phenotype, but depends on the availability of the appropriate heteroconjugated antibodies.

244

PIERRE A. HENKART AND MARTA CATALFAMO

4. A very powerful approach for assaying cytokine-secreting effectors is based on ﬂow cytometric detection of cytoplasmic cytokine after treatment with drugs that block release of the newly synthesized mediators from cells (Jung et al., 1993). Typical protocols call for stimulation of effector cells with antigen, anti-CD3, or PMA and ionomycin for 6 hours, with Brefeldin A added for the last 4 hours to block release from the Golgi. (Some protocols use monensin in place of Brefeldin A.) The cells are then ﬁxed and permeabilized, stained with ﬂuorophore-labeled anticytokine antibody, and analyzed by ﬂow cytometry. This approach allows the enumeration of effector cells and analysis of relative levels of secretion on a per-cell basis. Like the previously cited approach, it permits simultaneous characterization of the effector cell surface phenotype by staining ﬁrst for surface antigens detectable in different channels. Kits are available (e.g., from BD Biosciences) for such assays of interferon-g, TNF-a, IL-2, IL-4, IL-5, IL-10, and others, along with antibodies for codetection of surface markers. The possible secretion of cytokines such as TNF-a and IL-10 by APC can be eliminated from consideration if cytokine-positive cells are identiﬁed by appropriate T cell surface markers. One deﬁnitional issue for cytokine secretory effector cells is which cytokines are being measured, and in particular whether IL-2 should be considered as an ‘‘effector cytokine.’’ IL-2 is secreted by naive and memory human CD4þ and CD8þ T cells after TcR ligation, but not by differentiated CD8þ cells that kill and secrete IFN-g (Hamann et al., 1997; Sallusto et al., 1999). This different pattern has led to the exclusion of IL-2 as an effector cytokine, and we will not consider it further. Parallel to the Th1 vs Th2 cytokine secretion patterns of CD4þ T cells, CD8þ T cells follow a tendency to become differentiated along similar patterns (Mosmann et al., 1997). After in vitro activation in the appropriate cytokines, both Tc1 and Tc2 cells have potent in vitro cytotoxic activity via the perforin-dependent granule exocytosis pathway and the FasL/Fas pathway. However, TcR engagement triggers Tc1 cells to secrete predominantly IFN-g and IL-2, whereas Tc2 cells secrete predominantly IL-4, IL-5, and IL-10. Although no phenotypic surface markers have been described that completely correlate with the Tc1 vs Tc2 cytokine secretion patterns, a 2003 report indicates that within the CD8þ central memory subset of human blood lymphocytes (see following text), CCR4 is expressed on cells that secrete IL-4 within 24 hours after activation, while the CCR4 cells make low amounts of IL-4 (but do make IFN-g) under the same conditions (Geginat et al., 2003). When tested in various animal models, Tc1 and Tc2 cells have been shown to have different activities, presumably due to the differences in their cytokine

CD8þ EFFECTOR CELLS

245

secretion. For example, in a mouse tumor therapy model, Tc1 cells were considerably more potent than Tc2 cells, an effect that depended on their ability to secrete IFN-g (Helmich and Dutton, 2001). For immunotherapy of leukemias, Tc2 cells were found to be advantageous over Tc1 cells because of their greater tumor protective graft-vs-leukemia effect vs a deleterious graftvs-host effect (Fowler and Gress, 2000). Tc1 vs Tc2 differentiation has not been shown to play a crucial role in in vivo immune responses, but this is an area of considerable clinical interest.

VIII. CD81 T Cell Differentiation

The role of cytotoxic effector cells in the antigen-triggered differentiation pathway from naive to memory CD8þ T cells has long been a subject of debate. The possibility that functional memory cells may also have cytotoxic or cytokine secretory effector function makes for possible difﬁculties with interpreting simpliﬁed differentiation schemes such as those depicted in Fig. 1. The simplest differentiation model (Fig. 1A) is that there is a linear antigen-induced differentiation from naive to cytotoxic effector cells, followed by survival of some of the latter as memory cells. Opposed to this is the branched model (Fig. 1B) which proposes that antigen induces the formation of separate lineages of effector cells and pre-memory cells, and none of the memory cells derive from cytotoxic effector cells with time. There is evidence in favor of both the linear and branched models (Kaech et al., 2002), which we will not attempt to summarize here, except to point out that there is no strong evidence against the simple linear model. The dissection of memory cells into two subpopulations based on homing and chemokine receptor expression (central and effector memory cells) demands for more complex models, depicted in Fig. 1C–F. Some experiments strongly support a progression from effector memory to central memory cells in the mouse LCMV system (Wherry et al., 2003), compatible with models C or D. A. Effector Functions of CD8þ T Cell Phenotypic Subsets 1. Naive Phenotype Dozens of published studies are in general agreement that when antigen is ﬁrst encountered, naive CD8þ T cells from humans and mice respond by secreting IL-2 but not other cytokines. One exception to this is a report that naive mouse CD8þ T cells do make TNF-a, although not IFN-g (Walzer et al., 2000). There is a consensus of reports that naive CD8þ T cells do not kill antigen-bearing target cells in short-term assays, nor do they express detectable cytoplasmic perforin or granzymes.

246

PIERRE A. HENKART AND MARTA CATALFAMO

Fig 1 Possible CD8þ T cell differentiation pathways. N, naive; E, effector; EM, effector memory; CM, central memory.

CD8þ EFFECTOR CELLS

247

2. Central Memory Phenotype As deﬁned by high expression of CCR7 but low expression of CD45 RA, human central memory CD8þ T cells from blood were reported to express undetectable levels of cytoplasmic perforin and to secrete negligible amounts of IFN-g in response to anti-CD3 and PMA (Sallusto et al., 1999). When EBVspeciﬁc CD8þ T cells from chronically EBV-infected human donors were examined for their ability to secrete IFN-g, the CCR7þ and CD62Lþ subsets did not respond while the CCR7 and CD62L subsets did, conﬁrming the previously stated depiction of central memory cells as lacking cytokine secretory effector function (Hislop et al., 2001; Tussey et al., 2000). A similar but perhaps not identical subpopulation of human CD8þ cells as deﬁned by high expression of CD27 but low expression of CD45RA was found to express no cytoplasmic perforin and low levels of cytoplasmic granzyme B, and to lack cytotoxic effector activity (Hamann et al., 1997). In contrast to the report using CCR7 expression, the CD27lo subpopulation was reported to be capable of rapidly secreting IFN-g, TNF-a, IL-4 after stimulation by PMA and ionomycin. Further evidence that human CD8þ central memory cells have cytokine secretory effector function was provided by a report that both CCR7þ and CCR7 memory cells against EBV and ﬂu make IFN-g in response to peptide antigens (Ravkov et al., 2003). In the mouse, earlier reports examined memory cells from spleen or a combination of spleen and lymph nodes, without deﬁning their surface phenotype. They found that such memory cells expressed perforin and were cytotoxic to antigen-pulsed target cells in short-term assays (Cho et al., 1999; Opferman et al., 1999), and gave rapid IFN-g secretion after activation (Cho et al., 1999). Similar cytotoxic effector function was reported in human memory cells (Hislop et al., 2001). Another study showed that mouse memory CD8þ T cells from spleen were not cytotoxic in short-term assays, as opposed to CD8þ memory cells from liver, lung, and lamina propria (Masopust et al., 2001). However, these memory cells from spleen as well as other organs were capable of rapid IFN-g secretion after activation. As deﬁned by CCR7 expression, murine central memory CD8þ T cells were reported to be cytotoxic effectors and capable of secreting IFN-g and TNF-a (Unsoeld et al., 2002). It appeared that the cytotoxic activity of these cells may have been declining over a period of weeks, however. Another study deﬁned murine central memory cells based on high CD62L expression, and found that such cells expressed low levels of cytoplasmic granzyme B and were cytotoxic in 18-hour but not 5-hour assays (Wherry et al., 2003). When isolated from either spleen or lymph node, these cells secreted IFN-g and TNF-a rapidly after activation.

248

PIERRE A. HENKART AND MARTA CATALFAMO

3. Effector Memory Phenotype Among the different laboratories studying CD8þ memory cells, there is a consensus that effector memory cells have cytokine secretory effector activity and, indeed, this ability gave rise to the name ‘‘effector memory.’’ The consensus includes both the mouse (Masopust et al., 2001; Unsoeld et al., 2002; Wherry et al., 2003) and human systems (Hamann et al., 1997; Sallusto et al., 1999), deﬁned by the different approaches already discussed, and applies to IFN-g, TNF-a, IL-4, and IL-5, although not all laboratories examined all these cytokines. With respect to the cytotoxic effector activity of effector memory cells, the limited published data does not show a consensus. In patients infected with hepatitis B and hepatitis C, tetramer staining of blood cells showed a predominance of CD45RA, CCR7 antigen-reactive cells for both viruses during and after the acute viral infection (Urbani et al., 2002). However, the cytoplasmic perforin content was clearly lower in the CD45RA, CCR7 subpopulation after hepatitis C infection than after hepatitis B infection, suggesting that the surface phenotype may not correlate with function. In mice, CD8þ CCR7 memory cells were reported to have cytotoxic effector activity in short-term assays (Unsoeld et al., 2002), and CD8þ memory cells from nonlymphoid organs such as the liver and lung were also described as having such activity (Masopust et al., 2001). On the other hand, CD62Llow memory cells from mice infected with LCMV more than a month previously were found to be active in 18-hour but not 5-hour cytotoxic assays (Wherry et al., 2003). No satisfying generalizations can be made at present regarding either cytotoxic or cytokine secretory effector activities of memory phenotype cells. Part of the problem is a lack of consensus on the deﬁnition of memory cells and subsets in both mouse and human, but it also seems likely that different biological systems give different patterns of differentiation. While many reports indicate that memory cells have cytokine secretory effector activity, their cytotoxic effector activity seems to depend on the system being studied. 4. Effector Phenotype Cytotoxic effector function in a subset of CD8þ T cells from normal human blood was ﬁrst demonstrated in CD11b cells using aCD3-redirected cytotoxicity (Azuma et al., 1993). Subsequently, the CD8þ effector phenotype population, deﬁned by expression of CD45RA but low expression of CD27, also showed potent cytotoxic effector function by redirected cytotoxicity, as well as the ability to rapidly secrete IFN-g and INF-a, but not IL-2 (Hamann et al., 1997). When stained with an anti-granzyme A antibody, this effector subpopulation shows a bright granular stain similar to that of NK cells (Baars et al., 2000). The same study showed that these effectors secrete granzyme

CD8þ EFFECTOR CELLS

249

A and B after treatment with PMA and ionomycin, and that they express intracellular FasL. Interestingly, the frequency of CD45RAþCD27 CD8þ T cells in the blood of healthy children was recently found to correlate with prior infection with CMV but not EBV, varicella-zoster, or with MMR vaccination (Kuijpers et al., 2003). The frequency of these effector phenotype cells rose to 20% of the CD8þ T cells during the acute CMV infection and remained stable over a period of 3 years. In parallel to the these results, a CD45RAþ CCR7 human blood CD8þ effector phenotype subpopulation was reported to express high levels of intracellular perforin and to rapidly make IFN-g but not IL-2 (Sallusto et al., 1999). In the mouse, no useful CD45RA antibodies are available, and no equivalent surface markers have been developed to deﬁne this population of CD8þ T cells. However, as has been discussed, cytotoxic effector function is measurable from CD8þ T cells isolated directly from mice under a variety of circumstances. IX. Conclusions

Although the term ‘‘effector cell’’ has been used by immunologists in different ways, the term best describes differentiated T cells with immediate TcR-triggered cytotoxicity or cytokine secretion, as detected in short-term assays, which are currently only feasible in vitro. All investigators agree that antigen recognition by naive CD8þ T cells does not result in either type of effector function (the IL-2 produced by naive T cells is not considered an effector cytokine). After an initial encounter with antigen, a differentiation process occurs enabling effector function in subsequent antigen encounters. However, different effector functions have different differentiation pathways, and some subsets of post-naive CD8þ T cells have potent TcR-triggered secretory function for some cytokines but not for cytotoxic effector function. Major cellular and molecular questions about these differentiation pathways remain unanswered. Cytotoxic effector cells function by two different molecular pathways, both of which utilize TcR-triggered exocytosis of secretory granules. Cytotoxic effector cells are differentiated to express the preformed mediators perforin, granzymes, and Fas Ligand, intracellularly in granules. Differentiation further provides for a functional TcR-induced cytoplasmic polarization and granule exocytosis, with Rab27 playing a role in the latter. In the case of cytokine secretory effector cells, differentiation alters transcriptional regulatory elements to allow rapid cytokine gene expression, protein synthesis, and secretion by the constitutive pathway in response to TcR signaling. Even in fully differentiated cytotoxic effector cells, the different secretory processes for cytotoxicity and cytokine secretion are controlled by different branches of

250

PIERRE A. HENKART AND MARTA CATALFAMO

the TcR-initiated signaling pathway. Signaling for exocytosis of cytotoxic granules involves cytoplasmic events while cytokine secretion is typically limited by nuclear events controlling transcription, reﬂecting a yet-to-be-deﬁned branch point in the signaling pathway. Attempts to ﬁnd human lymphocyte subsets deﬁned by surface phenotype that correlate with effector functional activities have provided a useful cytotoxic effector phenotype based on expression of the CD45RA isotype and low expression of a second marker such as CD27 or CCR7. In contrast, no clear phenotypic markers have been found to correlate with cytokine secretory effector activity. Immunological memory is based on an expansion of antigen-reactive T cell clones as well as the ability of those cells to respond rapidly to antigen. The basic cellular mechanism of these T cell responses is secretion, both of cytotoxic mediators and cytokines. The differentiation process that confers secretory effector function on T cells thus lies at the heart of mammalian immunity and will be studied for years to come. References Ando, K., Guidotti, L. G., Wirth, S., Ishikawa, T., Missale, G., Moriyama, T., Schreiber, R. D., Schlicht, H.-J., Huang, S., Chisari, F. V., Schlicht, H. J., and Huang, S. N. (1994). J. Immunol. 152, 3245. Andrade, F., Roy, S., Nicholson, D., Thornberry, N., Rosen, A., and Casciola-Rosen, L. (1998). Immunity 8, 451. Azuma, M., Phillips, J. H., and Lanier, L. L. (1993). J. Immunol. 150, 1147. Baars, P. A., Ribeiro Do Couto, L. M., Leusen, J. H., Hooibrink, B., Kuijpers, T. W., Lens, S. M., and van Lier, R. A. (2000). J. Immunol. 165, 1910. Balaji, K. N., Schaschke, N., Machleidt, W., Catalfamo, M., and Henkart, P. A. (2002). J. Exp. Med. 196, 493. Barchet, W., Oehen, S., Klenerman, P., Wodarz, D., Bocharov, G., Lloyd, A. L., Nowak, M. A., Hengartner, H., Zinkernagel, R. M., and Ehl, S. (2000). Eur. J. Immunol. 30, 1356. Barry, M., and Bleackley, R. C. (2002). Nat. Rev. Immunol. 2, 401. Beresford, P. J., Xia, Z., Greenberg, A. H., and Lieberman, J. (1999). Immunity 10, 585. Beresford, P. J., Zhang, D., Oh, D. Y., Fan, Z., Greer, E. L., Russo, M. L., Jaju, M., and Lieberman, J. (2001). J. Biol. Chem. 276, 43285. Berke, G. (1980). Prog. Allergy 27, 69. Blott, E. J., Bossi, G., Clark, R., Zvelebil, M., and Grifﬁths, G. M. (2001). J. Cell Sci. 114, 2405. Bossi, G., and Grifﬁths, G. M. (1999). Nature Med. 5, 90. Brosterhus, H., Brings, S., Leyendeckers, H., Manz, R. A., Miltenyi, S., Radbruch, A., Assenmacher, M., and Schmitz, J. (1999). Eur. J. Immunol. 29, 4053. Casciola-Rosen, L., Andrade, F., Ulanet, D., Wong, W. B., and Rosen, A. (1999). J. Exp. Med. 190, 815. Cho, B. K., Wang, C., Sugawa, S., Eisen, H. N., and Chen, J. (1999). Proc. Natl. Acad. Sci. USA 96, 2976. Cohen, P. L., and Eisenberg, R. A. (1991). Ann. Rev. Immunol. 9, 243. Czerkinsky, C., Andersson, G., Ferrua, B., Nordstrom, I., Quiding, M., Eriksson, K., Larsson, L., Hellstrand, K., and Ekre, H. P. (1991). Immunol. Rev. 119, 5. Darmon, A. J., Nicholson, D. W., and Bleackley, R. C. (1995). Nature 377, 446. de Saint Basile, G., and Fischer, A. (2001). Curr. Opin. Immunol. 13, 549.

CD8þ EFFECTOR CELLS

251

Dourmashkin, R. R., Deteix, P., Simone, C. B., and Henkart, P. A. (1980). Clin. Exp. Immunol. 43, 554. Ebnet, K., Hausmann, M., Lehmann-Grube, F., Mu¨llbacher, A., Kopf, M., Lamers, M., and Simon, M. M. (1995). EMBO J. 14, 4230. Fan, Z., Beresford, P. J., Oh, D. Y., Zhang, D., and Lieberman, J. (2003a). Cell 112, 659. Fan, Z., Beresford, P. J., Zhang, D., and Lieberman, J. (2002). Mol. Cell. Biol. 22, 2810. Fan, Z., Beresford, P. J., Zhang, D., Xu, Z., Novina, C. D., Yoshida, A., Pommier, Y., and Lieberman, J. (2003b). Nat. Immunol. 4, 145. Fowler, D. H., and Gress, R. E. (2000). Leuk. Lymphoma 38, 221. Froelich, C. J., Orth, K., Turbov, J., Seth, P., Gottlieb, R., Babior, B., Shah, G. M., Bleackley, R. C., Dixit, V. M., and Hanna, W. (1996). J. Biol. Chem. 271, 29073. Geginat, J., Lanzavecchia, A., and Sallusto, F. (2003). Blood 101, 4260. Goping, I. S., Barry, M., Liston, P., Sawchuk, T., Constantinescu, G., Michalak, K. M., Shostak, I., Roberts, D. L., Hunter, A. M., Korneluk, R., and Bleackley, R. C. (2003). Immunity 18, 355. Guidotti, L. G., and Chisari, F. V. (2001). Annu. Rev. Immunol. 19, 65. Guidotti, L. G., Ishikawa, T., Hobbs, M. V., Matzke, B., Schreiber, R., and Chisari, F. V. (1996). Immunity 4, 25. Haddad, E. K., Wu, X., Hammer, J. A., and Henkart, P. A. (2001). J. Cell Biol. 152, 835. Hamann, D., Baars, P. A., Rep, M. H.G., Hooibrink, B., Kerkof-Garde, S. R., Klein, M. R., and Van Lier, R. A.W. (1997). J. Exp. Med. 186, 1407. Harty, J. T., and Badovinac, V. P. (2002). Curr. Opin. Immunol. 14, 360. Hayakawa, Y., Kelly, J. M., Westwood, J. A., Darcy, P. K., Diefenbach, A., Raulet, D., and Smyth, M. J. (2002). J. Immunol. 169, 5377. Helmich, B. K., and Dutton, R. W. (2001). J. Immunol. 166, 6500. Heusel, J. W., Wesselschmidt, R. L., Shresta, S., Russell, J. H., and Ley, T. J. (1994). Cell 76, 977. Hislop, A. D., Gudgeon, N. H., Callan, M. F., Fazou, C., Hasegawa, H., Salmon, M., and Rickinson, A. B. (2001). J. Immunol. 167, 2019. Jung, T., Schauer, U., Heusser, C., Neumann, C., and Rieger, C. (1993). J. Immunol. Methods 159, 197. Kaech, S. M., Wherry, E. J., and Ahmed, R. (2002). Nat. Rev. Immunol. 2, 251. Kafrouni, M. I., Brown, G. R., and Thiele, D. L. (2001). J. Immunol. 167, 1566. Kagi, D., Ledermann, B., Burki, K., Zinkernagel, R. M., and Hengartner, H. (1996). Annu. Rev. Immunol. 14, 207. Kam, C. M., Hudig, D., and Powers, J. C. (2000). Biochim. Biophys. Acta 1477, 307. Kataoka, T., Togashi, K., Takayama, H., Takaku, K., and Nagai, K. (1997). Immunol. 91, 493. Kojima, Y., Kawasaki-Koyanagi, A., Sueyoshi, N., Kanai, A., Yagita, H., and Okumura, K. (2002). Biochem. Biophys. Res. Commun. 296, 328. Kuhn, J. R., and Poenie, M. (2002). Immunity 16, 111. Kuijpers, T. W., Vossen, M. T., Gent, M. R., Davin, J. C., Roos, M. T., Wertheim-van Dillen, P. M., Weel, J. F., Baars, P. A., and van Lier, R. A. (2003). J. Immunol. 170, 4342. Manjunath, N., Shankar, P., Stockton, B., Dubey, P. D., Lieberman, J., and von Andrian, U. H. (1999). Proc. Natl. Acad. Sci. USA 96, 13932. Masopust, D., Vezys, V., Marzo, A. L., and Lefrancois, L. (2001). Science 291, 2413. Menasche, G., Pastural, E., Feldmann, J., Certain, S., Ersoy, F., Dupuis, S., Wulffraat, N., Bianchi, D., Fischer, A., Le Deist, F., and de Saint Basile, G. (2000). Nat. Genet. 25, 173. Metkar, S. S., Wang, B., Aguilar-Santelises, M., Raja, S. M., Uhlin-Hansen, L., Podack, E., Trapani, J. A., and Froelich, C. J. (2002). Immunity 16, 417. Mosmann, T. R., Li, L., and Sad, S. (1997). Semin. Immunol. 9, 87. Motyka, B., Korbutt, G., Pinkoski, M. J., Heibein, J. A., Caputo, A., Hobman, M., Barry, M., Shostak, I., Sawchuk, T., Holmes, C. F., Gauldie, J., and Bleackley, R. C. (2000). Cell 103, 491. Mullbacher, A., Ebnet, K., Blanden, R. V., Hla, R. T., Stehle, T., Museteanu, C., and Simon, M. M. (1996). Proc. Natl. Acad. Sci. USA 93, 5783.

252

PIERRE A. HENKART AND MARTA CATALFAMO

Nakajima, H., Park, H. L., and Henkart, P. A. (1995). J. Exp. Med. 181, 1037. Oehen, S., and Brduscha-Riem, K. (1998). J. Immunol. 161, 5338. Opferman, J. T., Ober, B. T., and Ashton-Rickardt, P. G. (1999). Science 283, 1745. Pereira, R. A., Simon, M. M., and Simmons, A. (2000). J. Virol. 74, 1029. Podack, E. R., and Dennert, G. (1983). Nature 302, 442. Raja, S. M., Wang, B., Dantuluri, M., Desai, U. R., Demeler, B., Spiegel, K., Metkar, S. S., and Froelich, C. J. (2002). J. Biol. Chem. 277, 49523. Ratner, A., and Clark, W. R. (1993). J. Immunol. 150, 4303. Ravkov, E. V., Myrick, C. M., and Altman, J. D. (2003). J. Immunol. 170, 2461. Riera, L., Gariglio, M., Valente, G., Mullbacher, A., Museteanu, C., Landolfo, S., and Simon, M. M. (2000). Eur. J. Immunol. 30, 1350. Russell, J. H., and Ley, T. J. (2002). Annu. Rev. Immunol. 20, 323. Sallusto, F., Lenig, D., Forster, R., Lipp, M., and Lanzavecchia, A. (1999). Nature 401, 708. Sarin, A., Haddad, E. K., and Henkart, P. A. (1998). J. Immunol. 161, 2810. Sarin, A., Williams, M. S., Alexander-Miller, M. A., Berzofsky, J. A., Zacharchuk, C. M., and Henkart, P. A. (1997). Immunity 6, 209. Shiver, J. W., and Henkart, P. A. (1991). In ‘‘NK Cell-Mediated Cytotoxicity: Receptors, Signaling, and Mechanisms’’ (R. B. Herberman and E. Lotzova, Eds.), p. 341. CRC Press, Miami, FL. Shresta, S., Maclvor, D. M., Heusel, J. W., Russell, J. H., and Ley, T. J. (1995). Proc. Natl. Acad. Sci. USA 92, 5679. Simone, C. B., and Henkart, P. (1980). J. Immunol. 124, 954. Smyth, M. J., Kelly, J. M., Sutton, V. R., Davis, J. E., Browne, K. A., Sayers, T. J., and Trapani, J. A. (2001). J. Leukoc. Biol. 70, 18. Smyth, M. J., Thia, K. Y., Street, S. E., MacGregor, D., Godfrey, D. I., and Trapani, J. A. (2000). J. Exp. Med. 192, 755. Stepp, S. E., Dufourcq-Lagelouse, R., Le Deist, F., Bhawan, S., Certain, S., Mathew, P. A., Henter, J. I., Bennett, M., Fischer, A., de Saint Basile, G., and Kumar, V. (1999). Science 286, 1957. Stinchcombe, J. C., Barral, D. C., Mules, E. H., Booth, S., Hume, A. N., Machesky, L. M., Seabra, M. C., and Grifﬁths, G. M. (2001a). J. Cell Biol. 152, 825. Stinchcombe, J. C., Bossi, G., Booth, S., and Grifﬁths, G. M. (2001b). Immunity 15, 751. Stinchcombe, J. C., and Grifﬁths, G. M. (1999). J. Cell Biol. 147, 1. Sutton, V. R., Wowk, M. E., Cancilla, M., and Trapani, J. A. (2003). Immunity 18, 319. Trapani, J. A., and Smyth, M. J. (2002). Nat. Rev. Immunol. 2, 735. Trapani, J. A., Sutton, V. R., Thia, K. Y., Li, Y. Q., Froelich, C. J., Jans, D. A., Sandrin, M. S., and Browne, K. A. (2003). J. Cell Biol. 160, 223. Tussey, L., Speller, S., Gallimore, A., and Vessey, R. (2000). Eur. J. Immunol. 30, 1823. Uellner, R., Zvelebil, M. J., Hopkins, J., Jones, J., MacDougall, L. K., Morgan, B. P., Podack, E., Waterﬁeld, M. D., and Grifﬁths, G. M. (1997). EMBO J. 16, 7287. Unsoeld, H., Krautwald, S., Voehringer, D., Kunzendorf, U., and Pircher, H. (2002). J. Immunol. 169, 638. Urbani, S., Boni, C., Missale, G., Elia, G., Cavallo, C., Massari, M., Raimondo, G., and Ferrari, C. (2002). J. Virol. 76, 12423. Walzer, T., Joubert, G., Dubois, P. M., Tomkowiak, M., Arpin, C., Pihlgren, M., and Marvel, J. (2000). Cell Immunol. 206, 16. Wherry, E. J., Teichgraber, V., Becker, T. C., Masopust, D., Kaech, S. M., Antia, R., von Andrian, U. H., and Ahmed, R. (2003). Nat. Immunol. 4, 225. Wilson, S. M., Yip, R., Swing, D. A., O’Sullivan, T. N., Zhang, Y., Novak, E. K., Swank, R. T., Russell, L. B., Copeland, N. G., and Jenkins, N. A. (2000). Proc. Natl. Acad. Sci. USA 97, 7933. Young, J. D., and Cohn, Z. A. (1986). Cell 46, 641.

advances in immunology, vol. 83

An Integrated Model of Immunoregulation Mediated by Regulatory T Cell Subsets HONG JIANG AND LEONARD CHESS Department of Medicine and Pathology Columbia University College of Physicians and Surgeons New York, New York 10032

I. Introduction and Objectives

The immune system is capable of mounting effective immune responses to a virtually inﬁnite variety of foreign pathogens or tumor cells while avoiding harmful immune responses to self. This is achieved by central thymic selection and peripheral regulation. It is clear that thymocytes expressing T cell receptors (TCR) with high afﬁnity for self-peptide/MHC complexes undergo apoptosis and are deleted centrally in the thymus. However, recent experiments have highlighted the fact that central thymic negative selection eliminates the majority potentially pathogenic self-reactive T cells, however, many selfreactive T cells with low to intermediate afﬁnity for self antigen escape thymic negative selection and are released into the periphery. Although these selfreactive T cells display relatively low-avidity to self peptide/MHC complexes, they are capable of self peptide-driven proliferation and some may differentiate into potentially pathogenic effector cells (Bouneaud et al., 2000; Goldrath and Bevan, 1999; Jiang et al., 2003; Kuchroo et al., 2002). Thus, mechanisms that normally regulate the outgrowth or function of these self-reactive T cells may ultimately control the initiation and progression of autoimmune disease. Unfortunately, these carefully designed physiologic systems that normally prevent the emergence of functional autoreactive cells sometimes fail, with the subsequent development of clinical autoimmunity. In this chapter, we summarize our current views of the cellular basis of physiologic immune regulation. Moreover, we present a model of immunoregulation that emphasizes the unique roles played by three distinct subsets of regulatory T lymphocytes in the control or suppression of the peripheral immune response. These subsets include the NKT cells and the CD4þ regulatory T cells which exist, in part, as naturally occurring regulatory cells and the CD8þ regulatory cells which are speciﬁcally induced to differentiate into effector regulatory cells following speciﬁc interaction with antigen-activated T cells. We propose that these three subsets of regulatory cells work in concert to regulate peripheral immunity at different phases of the immune response (Fig. 1). First, we present some of the key ideas that we feel are essential to understanding current paradigms of immunoregulation and especially the role 253 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00

254

HONG JIANG AND LEONARD CHESS

Fig 1 Homeostatic control of the outgrowth of antigen-activated CD4þ T cells, The control of the peripheral immunity is accomplished by mechanisms intrinsic to antigen activation of CD4þ T cells, including apoptosis, induction of anergy, and differentiation into Th subsets which are independent of other types of regulatory cells. In addition, superimposed on these intrinsic mechanisms are control mechanisms mediated by distinct subsets of NKT, CD4þ, and CD8þ regulatory (suppressor) T cells.

of the regulatory cells in this process. We then summarize the evidence supporting an integrative model of immunoregulation which proposes that the NKT, CD4þ, and CD8þ regulatory T cell subsets display unique immunoregulatory functions that are predominant during the primary and secondary phases of the immune response. This immunoregulatory model proposes that regulatory T cells work in concert with other immunoregulatory mechanisms, including those mediated by cytokines, to prevent the outgrowth of pathogenic autoreactive clones. Moreover, we propose that some of these same immunoregulatory mechanisms also ﬁne tune the peripheral T cell repertoire in response to foreign antigens and thus play a functional role in the afﬁnity maturation of the T cell repertoire to foreign antigens. II. Historical Considerations: Clonal Selection Theory, Immunoregulation, and Regulatory T Cell Subsets

The idea that immune responses must be tightly regulated was implicit in the early ideas of clonal selection as proposed by Ehrlich in 1900. Ehrlich envisioned that during the ontogeny and outgrowth of the immunocompetent

IMMUNOREGULATION AND T CELL SUBSETS

255

clones responsive to foreign antigens there had to be mechanisms to control the outgrowth of clones reactive with self (Ehrlich, 1900). In Ehrlich’s conception, the failure to control the outgrowth of autoreactive cells would lead to a state of ‘‘horror autoxicus’’ or autoimmunity. Ehrlich’s ideas were ampliﬁed and reﬁned in the 1950s by MacFarlane Burnet and Neils Jerne, with the elaboration of the Clonal Selection Hypothesis (Burnet, 1957; Jerne, 1955, 1975). This hypothesis states that: (1) multiple clones of immunocompetent cells displaying unique antigen-speciﬁc receptors exist prior to the introduction of foreign antigens; (2) following exposure to a particular antigen, the antigen interacts with its unique receptor on the cell surface of an immunocompetent cell and this speciﬁc receptor–antigen interaction triggers cells to divide and express their immunological functions; and (3) the majority of cells bearing receptors for self-antigens must be eliminated during differentiation or carefully regulated. These tenets of clonal selection have proven, over the years, to be essentially correct. A corollary of these notions, as has been noted, is that autoimmune disease may arise from the failure to eliminate or inactivate immunocompetent cells during their ontogeny and/or the failure of the immune system to control the outgrowth or function of self-reactive clones in the periphery. Thus, immunocompetent self-reactive cells circulating in the periphery, from a teleological point of view, had to be either incapable of responding to self antigens (anergic or tolerant to self) or, alternatively, had to be speciﬁcally suppressed. Indeed, during the past three decades evidence has continued to emerge that has substantiated the idea that the immune system, not unlike all other complex physiologic and biochemical systems, is homeostatically regulated and often involves interacting feedback suppressor mechanisms. In the immune system, these suppressor mechanisms are aimed at retarding the outgrowth of potentially pathogenic autoreactive clones while permitting the robust expansion of clones reactive to foreign antigens. In a large number of autoimmune diseases in man, including rheumatoid arthritis, type 1 diabetes mellitus, and multiple sclerosis, the pathogenic autoreactive clones are predominately CD4þ cells. As depicted in Fig. 1, four distinct types of mechanisms can be shown under various experimental conditions to control the outgrowth of autoreactive CD4þ T cells. One set of mechanisms involves activation-induced cell death, in which the encounter between self antigen with autoreactive clones expressing high afﬁnity T cell receptors (TCR) delivers apoptotic signals leading to DNA fragmentation and cell death. This activation-induced cell death usually occurs intrathymically as a mechanism of central thymic selection. In addition, activated self-reactive clones can also express deathinducing cell surface molecules like Fas ligand or TNF like molecules (Lenardo et al., 1999). These molecules interacting with speciﬁc receptors expressed on activated clones may also induce apoptotic signals leading to the

256

HONG JIANG AND LEONARD CHESS

fratricide or suicide of self-reactive clones. The basic importance of these intrinsic apoptotic mechanisms to the regulation of self-reactive cells and the control of autoimmunity is illustrated by the observations that mice and people genetically deﬁcient in Fas or its receptor develop signiﬁcant lymphadenopathy, splenomegaly, and autoantibodies as well as other autoimmune characteristics such as systemic lupus erythematosus and rheumatoid arthritis (Lenardo et al., 1999; Wu et al., 1996). A second general set of regulatory mechanisms is initiated during the initial TCR/MHC/peptide interaction, when other receptor ligand interactions become pivotal in ultimately dictating the functional fate of T cells. For example, one of the earliest antigen activation-induced cell surface molecules expressed by T cells is CD40L (Lederman et al., 1992). A critical consequence of the interaction of CD40L with CD40 expressed on APCs is the upregulation of other key co-stimulatory molecules, including CD80 and CD86 (Caux et al., 1994; Klaus et al., 1994). These molecules interact with CD28 or CTLA-4 molecules expressed on T cells to determine whether the outcome of antigen triggering will either be functional T cell activation or, alternatively, induction of anergy and tolerance (Durie et al., 1994; Koulova et al., 1991; Lenschow et al., 1996). Thus, antigen triggering in the absence of CD80 or CD86 triggering is known to induce anergic T cells (Boussiotis et al., 1996; Jenkins, 1994; Lenschow et al., 1996). Similarly, blockade of the CD40L/CD40 pathway can lead to tolerance induction (Durie et al., 1994). Clearly, T cell anergy in the periphery has been considered to be one of the major mechanisms for the peripheral self-tolerance. This mechanism could be either triggered by activation of T cell clones without co-stimulatory signals or as one of the effector mechanisms the regulatory T cells use to downregulate peripheral immune responses. A third general set of regulatory mechanisms, also a consequence of the initial MHC/peptide triggering of CD4þ T cells, is the further differentiation of the CD4þ T cells into the functional distinct TH1 and TH2 subsets phenotypically distinguished, in part, by the elaboration of distinct sets of cytokines (Coffman and Mosmann, 1991; Mosmann and Coffman, 1989; Mosmann et al., 1986). In this regard, IFN-g secreted by TH1 cells is known to downregulate the differentiation and function of TH2 cells, and conversely, IL-4 and IL-10 inhibits TH1 cell differentiation (Fitch et al., 1993; Mosmann and Coffman, 1989; Seder and Paul, 1994). In addition, a third type of Th cell capable of secreting the immunosuppressive cytokines TGFb and IL-10, termed Th3, has been observed (Roncarolo and Levings, 2000). Together, the intrinsic immunoregulatory properties of these cytokines have been shown to signiﬁcantly modify the induction and natural history of a variety of autoimmune diseases. For example, the downregulatory inﬂuences of IL-4, IL-10, and TGFb on Th1 subset differentiation and function have been shown to be anti-inﬂammatory

IMMUNOREGULATION AND T CELL SUBSETS

257

in models of the Th1 cell-mediated diseases noted previously. Indeed, a widely prevalent view is that the balance between the emergence of Th1 and Th2 CD4þ T cells following antigen activation plays a major role in the outgrowth and functions of self-reactive clones (Charlton and Lafferty, 1995; Del Prete, 1998). Superimposed on these ﬁrst three mechanisms of homeostatic regulation is a fourth general mechanism mediated by suppressor T cells which control the induction and/or outgrowth of antigen-activated T cells. The idea that suppressor T cells may be critical to the control of immune responses arose in the laboratories of Byron Waksman and Richard Gershon at Yale. Indeed, the ﬁrst evidence for the existence of suppressor T cells arose in the Gershon laboratory in the late 1960s and early 1970s in studies of immunetolerance to foreign antigens. Gershon showed initially that adoptive transfer of T cells from animals made tolerant to foreign antigen X could speciﬁcally suppress the production of anti-X antibodies in recipient animals (Gershon and Kondo, 1970, 1971). Importantly, it was subsequently found that the suppressor activity of T cells could also downregulate autoreactive cells as well as delayed type hypersensitivity responses (DTH) and tumor immunity (Askenase et al., 1975; Gershon et al., 1974). These experiments were conﬁrmed in a variety of distinct experimental models during the 1970s (Gershon, 1974, 1975, 1976). It is of interest to point out that these initial suppressor cell experiments strongly suggested that suppressor cells were not only important in the peripheral regulation of self-reactive cells but were also important in the regulation of cells responding to foreign antigens. The ﬁrst insights into potential models of how suppressor T cells may function in the speciﬁc regulation of immunity arose from the seminal experiments of Cantor and Boyse, who showed that the genetically well-deﬁned Lyt allo-antisera could be used to deﬁne phenotypically stable and functionally distinct subsets of T cells expressing either Lyt1 or Lyt2,3 alloantigens (Cantor and Boyse, 1975a,b). These experiments demonstrated that it was possible using antibodies to Lyt1 and Lyt2,3 to isolate the Lyt1 and Lyt2,3 subsets from nonimmune animals and show that these phenotypically stable populations of cells were genetically preprogrammed prior to interaction with antigen to mediate distinct immunologic programs. Thus, the Lyt1 subset were programmed to mediate inducer functions involved in the T cell help required for B cell differentiation and IgG antibody synthesis as well as the activation of macrophages and antigen presenting dendritic cells to mediate DTH responses. On the other hand, the Lyt2 T cells were programmed to differentiate into cytotoxic T lymphocytes (CTL) capable of lysing tumor cells and allogeneic targets. In addition, the Lyt2 subset was found to contain suppressor cells capable of speciﬁcally downregulating the responses to foreign antigens. For example, it was found that Lyt1 T cells generate helper activity to foreign

258

HONG JIANG AND LEONARD CHESS

antigens like sheep erythrocytes (SRBC). In contrast, after priming with SRBC, Lyt2,3 cells have suppressive activity. Thus, both cytotoxic and speciﬁc suppressor functions are mediated by T cells of the Lyt2,3 subset, which can be distinguished from helper T cells, by their Lyt phenotypes. It was unknown whether killing and suppression are functionally interrelated properties of a single Lyt2,3 subset or whether the Lyt2,3 population comprises two subsets whose surface phenotypes are not yet distinguishable. Thus, helper and suppressor immunoregulatory functions as expressed in Lyt1 and Lyt2,3 subsets were found to be manifestations of separate T cell differentiative pathways programmed prior to interaction with antigen (Cantor et al., 1976). These ﬁndings were later extended to other species, including humans (Chess and Schlossman, 1977; Evans et al., 1977) and with the advent of monoclonal antibodies, the CD4 and CD8 surface molecules were identiﬁed and used to study T cell differentiation and immunoregulation in more precise detail (Kung et al., 1979; Reinherz and Schlossman, 1980). The inducer Lyt1 population was exclusively found in the CD4þ population and the cytotoxic and suppressor Lyt2,3 populations were largely contained within the CD8þ population (Dialynas et al., 1983; Swain et al., 1984). In addition to these functional distinctions, CD4þ and CD8þ T cell subsets also differed in a cognitive sense. Thus, CD4þ T cells were found to be MHC class II restricted whereas CD8þ T cells were found to be MHC class I restricted in both human and mouse (Rao et al., 1983; Reinherz and Schlossman, 1980; Thomas et al., 1983). It is of interest that several years following this initial delineation of the CD4 and CD8 differentiative pathways that the CD4 and CD8 molecules were cloned and their structure and biochemical functions identiﬁed (Littman et al., 1985; Maddon et al., 1985, 1987). Thus, CD4 molecules were found to bind to conserved regions of MHC class II molecules and the CD8 molecule was found to bind to conserved regions of MHC class I molecules. Moreover, the cytoplasmic tails of both CD4 and CD8 molecules were found to bind to the lck tyrosine kinase involved in the antigen-induced T cell receptor triggering. As a consequence of these ﬁndings, the CD4 and CD8 molecules were clearly not only stable cell surface markers which deﬁned the major stable lineages of functional T cell subsets prior to interaction with antigen but were also functionally involved, as co-receptors that act together with the T cell receptor in the recognition of MHC/peptide complexes and the subsequent triggering of the TCR (Gay et al., 1987; Littman, 1987). The association of CD4 expression with the inducer or helper functions of T cells correlates with fact that the CD4 molecule binds MHC class II molecules which are preferentially expressed on B cells and other antigenpresenting cells. These are precisely the cells which are activated and induced to differentiate by activated helper CD4þ T cells. Similarly, it was of interest that CD8 molecules which bind MHC class I mark the cytotoxic lymphocyte

IMMUNOREGULATION AND T CELL SUBSETS

259

(CTL) functional subset. Teleologically, this makes sense because MHC class I molecules are expressed on virtually all nucleated cells and one important physiologically relevant function of CTL is to eradicate virally infected cells and tumor cells which could involve any nucleated cell. On the other hand, the functional signiﬁcance of the relationship of CD8 expression—MHC class I recognition with suppressor cell function—was not initially understood, but, in retrospect, suggested that the speciﬁcity of CD8þ regulatory T cell function may involve recognition of MHC class I/peptide complexes on the cellular targets of suppression (Jiang and Chess, 2000). In both mice and humans, suppression was shown in vitro to require interactions between CD8þ cells and CD4þ cells (Cantor et al., 1978a,b; Eardley et al., 1978; Thomas et al., 1983). These experiments led to the ﬁndings that a subset of activated CD4þ T cells were required to induce CD8þ suppressor cells. Moreover, the CD8þ effector cells then could mediate suppression by suppressing the activity of the CD4þ T cell inducing population. From these experiments, models emerged in which extraordinarily complex cellular circuits were envisioned (Asherson et al., 1986; Benacerraf et al., 1982; Dorf and Benacerraf, 1984; Germain and Benacerraf, 1980; Green et al., 1983). For example, a CD4þ suppressor inducer T cell subset were characterized by the expression of cell surface molecules identiﬁed by alloantisera, one termed Qa-1 (Eardley et al., 1978; Stanton et al., 1978). The Qa-1 alloantisera mapped to the 30 end of MHC class I region on chromosome 17 in a region known to encode other genes, such as TL (Kastner et al., 1979; Stanton and Boyse, 1979). It is of interest that eventually Qa-1 was cloned and found to be a MHC class Ib molecule capable of presenting endogenous as well as exogenous peptides (Connolly et al., 1993; Lalanne et al., 1985; Shawar et al., 1994) to CD8þ T cells. However, the potential signiﬁcance of Qa-1 expression on suppressor inducer cells was not delineated until the mid-1990s (Jiang et al., 1998, 1995) (see following text). Thus, in experiments which we will discuss in more detail, it has been shown that Qa-1 is preferentially expressed on activated T cells and antibodies to Qa-1 block CD8þ T cell mediated suppression in vitro (Jiang and Chess, 2000). These data suggest that a subset of CD8þ T cells may express TCRs restricted by Qa-1 or other MHC class Ib molecules and that the speciﬁcity of suppression may be dictated by the expression of certain Qa-1/peptide complexes selectively expressed on some antigen activated CD4þ T cells. In this regard, it is important to emphasize that the original suppressor cell circuits were initially conceived and/or deduced at a time when molecular immunology was in its infancy. For example, the nature of the TCR receptor was unknown, as was the precise structure and function of MHC molecules in restricting T cell activity. With respect to suppressor models, a key surface molecule which seemed to identify suppressor inducer cells was

260

HONG JIANG AND LEONARD CHESS

Qa-1 (Cantor et al., 1978a), which was not known at the time to be an MHC class Ib molecule. In addition, the great majority of the cytokines which are now known to regulate immune functions were largely unknown. In addition, it was also unknown that antigen activation of the TCRs induced the differentiation of CD4þ T cells into Th1 and Th2 subsets which elaborated distinct sets of regulatory cytokines as has been described. Clearly, understanding the precise role of these lymphokines in suppression would have signiﬁcantly inﬂuenced the interpretation of data, suggesting that an array of antigenspeciﬁc and-nonspeciﬁc suppressor factors were uniquely secreted by suppressor inducer and effector cells. In fact, the identiﬁcation of antigen-speciﬁc suppressor molecules could not be reproduced and veriﬁed and this, in turn, called into question the very existence of suppressor cells. These concerns led to general skepticism for the role of T–T interactions between CD4þ and CD8þ T cells in the downregulation of immune responses. As a consequence, interest in the models of T cell suppression mediated by CD8þ T cells waned by the mid-1980s (Bloom et al., 1992a,b; Janeway, 1988; Moller, 1988). In fact, for more than a decade, the very term suppressor cell was barely heard in immunologic circles and even today, when there has been somewhat of a resurgence in interest in immunoregulation largely due to the discovery of CD4þ suppressor cells (Hori et al., 2003; Sakaguchi, 2000; Shevach, 2000, 2002), the idea of CD8þ T cell suppression is not mentioned in otherwise scholarly reviews of immunoregulation. Clearly, although much of the skepticism concerning T cell suppression mediated by CD8þ T cells in the mid-1980s was justiﬁed, it is our view that undoubtedly the baby went out with the bath water. We believe that some of the important ideas, data, and models of immunoregulation discredited were essentially correct and have far-reaching biological and clinical signiﬁcance. For example, some studies suggested a central role of the Qa-1, a MHC class Ib molecule, in the speciﬁcity of suppression mediated by CD8þ suppressor T cells. Moreover, the models envisioned 25 years ago suggested that vaccination procedure employing pathogenic autoreactive clones could be used to prevent autoimmune disease (Ben-Nun et al., 1981; Cohen and Weiner, 1988). This, in fact, has been veriﬁed in murine models and, importantly, the protection mediated by T cell vaccination is abrogated by depletion of CD8þ T cells in the vaccinated animal (Jiang et al., 1998, 2001). It is important to point out that part of the waning interest in suppression in the mid-1980s was also a consequence of the spectacular advances that were being made in other avenues of immunologic investigation and, in particular, the introduction of molecular biological approaches to the solution of immunologic problems. These advances included the cloning and structural characterization of the ab TCRs, the MHC molecules, as well as the CD4 and CD8 molecules. In addition, other advances included the identiﬁcation and

IMMUNOREGULATION AND T CELL SUBSETS

261

structural characterization of a vast number of lymphokines as well as the elucidation of the Th1 and Th2 paradigms already noted. In contrast, to this day, the precise biochemical deﬁnition of how suppressor cells speciﬁcally regulate immune responses remains an enigma. It seem to us that it is highly unlikely that speciﬁc suppression of the immune response will be carried out solely by antigen nonspeciﬁc suppressor T cells elaborating cytokines that also are not antigen speciﬁc. It seems more likely that speciﬁc immune suppression will be mediated, at least in part, by speciﬁc cognate interactions between suppressor T cells and antigen activated CD4þ T cells. In this regard, we will present evidence that the interaction between the regulatory CD8þ T cells and the CD4þ T cell targets is based on the speciﬁc recognition of Qa-1/peptide complex preferentially expressed on certain activated CD4þ T cells by the ab TCR of the regulatory CD8þ T cells (Jiang et al., 1998, 2001, 2003). As has been noted, during the past decade there has been a resurgence of interest in suppressor T cells, largely due to the discoveries that in addition to CD8þ T cells, other types of cells including CD4þ T cells and NKT cells can also function to suppress the outgrowth of self-reactive T cells. Next, we brieﬂy review the nature of the suppression mediated by the various types of CD4þ and NKT regulatory cells and contrast these mechanisms to the suppression mediated by the CD8þ regulatory T cells. We conclude by proposing a provisional general model of immunoregulation which suggests that these various suppressor populations functioning in concert and superimposed on suppressor independent mechanisms of immunoregulation (apoptotic, anergic, and cytokine mediated mechanisms) ultimately regulate the peripheral immune responses.

III. The T Cell Subsets which Mediate Suppression of the Immune Response

A. The CD4þ T Suppressor Cells The ﬁrst evidence suggesting that suppressor function is not exclusively a property of CD8þ T cells but could also be detected within highly puriﬁed populations of CD4þ T cells, at least in vitro, arose in the early 1980s in experiments showing that co-culture of graded numbers of polyclonally activated human CD4þ T cells to autologous resting CD4þ T cells inhibited the capacity of the resting CD4þ T cells to induce B cell differentiation and Ig synthesis. The CD4þ inducer function of B cell differentiation was found to be radioresistant whereas suppressor capacity detected in the activated CD4þ T cells was radiosensitive (Thomas et al., 1981, 1982). It was difﬁcult to further characterize the activated CD4þ T cells at the time of these experiments because the vast array of other cell surface molecules expressed on activated

262

HONG JIANG AND LEONARD CHESS

CD4þ T cells, as well as the cytokines that were to distinguish resting from activated or memory human T cells, were not yet discovered. However, in vivo experiments in the mouse in the mid-1980 placed the idea of CD4þ suppressor cells on a ﬁrm biological foundation. In particular, studies of the organ-speciﬁc autoimmune disease induced in mice following neonatal thymectomy were quite revealing. These thymectomized mice were shown to have reduced numbers of Lyt1þ, Lyt 2,3- cells. Furthermore, reconstitution of thymectomized mice by highly puriﬁed population of Lyt1 but not by Lyt2,3þ cells from syngeneic normal mice completely inhibited disease development (Sakaguchi et al., 1985). A decade later in 1995, Sakaguchi showed that this suppression was mediated predominantly by CD4þCD25þ cells (Sakaguchi et al., 1995). The general experimental protocols employed to characterize these suppressor cells in vivo took advantage of the observation that mice deﬁcient in T cells (neonatally thymectomized, nu/nu mice, or Rag-/-mice) do not develop autoimmune disease following adoptive transfer of normal syngeneic spleen cells unless the spleen cells are depleted of CD4þCD25þ T cells. Thus, when CD4þ normal splenic T cells prepared from normal mice are depleted of CD25þ cells and the remaining CD4þ T cells were transferred to syngeneic T cell-deﬁcient mice, the recipients spontaneously developed various organ-speciﬁc autoimmune diseases (including diabetic-like insulinitis, thyroiditis, and gastritis) and systemic wasting disease. Reconstitution of the CD4þCD25þ population inhibited the autoimmune development (Sakaguchi, 2000; Shevach, 2000). The CD4þ suppressor cells have been further phenotypically distinguished, in large part, by cell surface molecules and/or cytokine secretion proﬁles. The cell surface molecules in addition to CD25 include CD45RB low, CTLA-4, and glucocorticoid-induced tumor necrosis factor receptor family-related gene (GITR) (Sakaguchi, 2000). All of molecules are actually well-deﬁned activation molecules expressed for a period of time on the majority of CD4þ T cells following antigen triggering of the TCR. Thus, CD4þ T cell clones with helper/ inducer functions also express these molecules following T cell activation and, as such, they are not speciﬁc markers of suppressor cells within the normal CD4þ T cell population. The search is still on to deﬁne this subset of cells with speciﬁc markers (Shevach, 2002). However, it has been noted that some CD4þ suppressor cells constitutively express either CD25 and/or CTLA-4. Although this point is controversial, it is possible that cells co-expressing CD25 and/or CTLA-4 may be enriched in the ‘‘true suppressor cells,’’ which may express a phenotypically stable marker of suppression that remains to be identiﬁed. In addition, the functions of these molecules are well known but do not together necessarily suggest a paradigm that enlightens us about the functional signiﬁcance of these markers for suppression. Thus, CD25 is a functional component of the IL-2 receptor important in the growth and

IMMUNOREGULATION AND T CELL SUBSETS

263

proliferation of all T cells and NK cells and, unless the suppression mediated by CD25þ T cells is due to depletion of circulating IL-2, it is not clear how the quintessential marker of the suppressor phenotype is itself functionally involved in suppression. The CD45 molecule is a tyrosine phosphatase known to regulate TCR signaling on all T cells. However, the CTLA-4 molecule is a receptor for CD80 (B71) and CD86 (B72) and CTLA-4/B7 interaction is known to inhibit TCR triggering (Janeway and Bottomly, 1994) and may be of special interest with respect to the function of CD4þCD25þ regulatory cells because like CD25, CTLA-4 is thought to be constitutively expressed on the CD4þ suppressor populations (Takahashi et al., 2000). Moreover, blockade of the CTLA-4/B7 interactions can abrogate suppression and T cells triggered via CTLA-4 predominantly secrete TGFb, a cytokine with suppressive functions (Salomon et al., 2000). Taken together, these data point to CTLA-4’s being a key receptor regulating CD4þ T cell suppression. Indeed, CTLA-4 deﬁcient mice exhibit profoundly enhanced inﬂammatory responses as do T cell deﬁcient mice adoptively transfused with CD4þ cells depleted of CD25þ cells, which may also render the mice deﬁcient in CTLA-4þ suppressor cells. On the other hand, in vitro experiments have not been able to document a blocking effect of anti-CTLA-4 on suppressor cell function (Shevach et al., 2001). With respect to the general role of cytokines in the suppression mediated by CD4þ regulatory T cells, the cytokine proﬁles found in the various populations of CD4þ suppressor cells include a variety of combinations of the already known immunoregulatory cytokines (i.e., IL-10, TGFb, IL-4, IFN-g). For example, there are CD4þ CD45Rblow activated suppressor cells that secrete large quantities of either IL-10 and IL-4 (termed Tr1 cells) and other CD4þ CD45Rblow suppressor T cells that secrete large quantities of TGFb (termed Th3 cells) (Roncarolo and Levings, 2000). The functional signiﬁcance of these cytokine-secreting CD4þ T cells is supported by the ﬁndings that TGFbdeﬁcient mice develop autoimmune disease (Gorelik and Flavell, 2000) and that administration of neutralizing antibodies to IL-4 or TGFb abrogates the in vivo prevention of autoimmunity or tolerance-inducing activity of CD4þ T cells in some models (Seddon and Mason, 1999; Zhai and Kupiec-Weglinski, 1999). The relationship of these cells to the CD4þCD25þCTLA-4þ cells is unclear. For example, the in vitro capacity of CD4þCD25þ T cells to suppress immune responses is thought to be contact dependent and not due to the IL-10, TGFb, and IL-4 cytokines. It is possible the Th3 and Tr1 populations arise from ‘‘conventional’’ resting CD4þCD25 T cells, which following antigen activation express CD25. In this regard, these Th3- or Tr1-like populations of CD4þCD25 cells may account for the in vivo experiments showing that transgenic mice expressing a T cell receptor for the EAE-inducing autoantigen MBP do not develop EAE unless they are bred with Rag-/- mice. The

264

HONG JIANG AND LEONARD CHESS

transgenic/rag-/- mice develop severe EAE. Importantly, the EAE is abrogated by adoptive transfer of either CD4þCD25þ or CD4þCD25- Tcells to the Rag-/mice. These suggest that the CD4þCD25- ‘‘suppressor’’ cells are functional in vivo but are unlikely to be a lineage-speciﬁc suppressor population. It is of great interest that a recently cloned transcription factor, termed Foxp3, a member of the forkhead family of DNA-binding transcription factors, is not expressed in naive CD4þCD25- cells but is highly expressed in antigenactivated CD4þCD25þ cells. In CD4þCD25þ T cells, expression of Foxp3 functions as a repressor of transcription factors that regulate T cell activation (Schubert et al., 2001). For example, Foxp3 binds to NFAT and overexpression of Foxp3 inhibits IL-2 production as well as IFN-g, IL-4, and IL-10 (Hori et al., 2003; Schubert et al., 2001). Importantly, mutational defects in the Foxp3 gene result in the fatal autoimmune and inﬂammatory disorder of the ‘‘scurfy mouse’’ and in the clinical and molecular features of the immuno-dysregulation, polyendocrinopathy, enteropathy, X linked (IPEX) syndrome in humans. Both scurfy mice and IPEX patients have defects in T cell activation and reduced numbers and reduced suppressor functions mediated by the CD4þCD25þ T cells (Bennett et al., 2001; Brunkow et al., 2001; Wildin et al., 2001). In Foxp3-overexpressing mice, both CD4þCD25- and CD4CD8þ T cells show suppressive activity, suggesting that expression of Foxp3 is linked to suppressor functions. The forced expression of Foxp3 also delays disease in CTLA-4-/- mice, indicating that the Scurﬁn and CTLA-4 pathways may intersect (Khattri et al., 2003). Taken together, these data strongly support the idea that Foxp3 may uniquely deﬁne the subset of naturally occurring CD4þ suppressor T cells. The in vivo data demonstrate that CD4þCD25þ suppressor T cells function to preferentially suppress self-reactive T cells and yet preserve normal immune functions. This apparent cognitive capacity to distinguish self from nonself is, on the surface, difﬁcult to reconcile with a number of studies that have provided convincing evidence that the CD4þCD25þ T regulatory cells are not antigen speciﬁc. So how do they distinguish self from nonself? The lack of antigen speciﬁcity has been reviewed (Shevach et al., 2001). Thus, in elegant experiments, the Shevach lab showed that when T cells from TCR transgenic mice are activated with their peptide/MHC ligand and expanded in vitro in IL-2, the activated suppressors are subsequently capable of suppressing the responses of T cells from mice that express a different transgenic TCR (Shevach et al., 2001; Thornton and Shevach, 2000). Moreover, no MHC restriction is observed in the interaction of the activated suppressors and the responding targets (Piccirillo and Shevach, 2001). Thus, the T cell receptors employed by these regulatory CD4þ T cells are likely to be quite diverse and their capacity to distinguish self from nonself may be independent of the T cell receptor they express. It has been suggested that insight into the mechanism

IMMUNOREGULATION AND T CELL SUBSETS

265

by which these antigen nonspeciﬁc T cells can still distinguish self from nonself may have come from recent studies showing that the CD4þ T regulatory cells when nonspeciﬁcally activated by LPS express tolllike receptors (Sakaguchi, 2003). Thus, these naturally occurring regulatory T cells may represent the regulatory component of the innate immune system responsive to ‘‘dangerlike signals.’’ Perhaps self-reactive T cells preferentially express ligands which are recognized by tolllike receptors expressed by the regulatory cells. Clearly, the elucidation of the precise target structures recognized by CD4þCD25þ regulatory cells may help deﬁne the receptors employed by the regulatory cells to distinguish self from nonself. We will present the evidence that CD8þ suppressor T cells are not naturally occurring cells but instead are speciﬁcally induced during the primary adaptive immune response and are triggered to differentiate into effector suppressor cells distinguishing among the clones of autoreactive cells responding to a singe peptide. Unlike the naturally occurring CD4þCD25þ suppressor T cells, the NKT cells and CD8þ T suppressor cells employ their conventional T cell receptor to recognize and distinguish the targets of suppression. B. The NKT Cells NKT cells are a unique population of cells that coexpress receptors of the NK lineage as well as the ab TCR. Most murine NKT cells express an invariant TCRa chain encoded by the Va14-Ja281 gene segment, paired preferentially to Vb8, Vb7, and Vb2 chains (Bendelac, 1995). Va14þ NKT cells recognize lipid antigens presented by the MHC class Ib-like molecule CD1d (Bendelac et al., 1997). Similar subsets of CD1d-restricted NKT cells are also present in humans (Exley et al., 1997; Lee et al., 2002a) and they express an invariant Va24-JaQ TCR. CD1d-restricted NKT cells are mainly of CD4þ or CD4CD8- phenotype (Eberl et al., 1999). Studies have indicated that NKT cells are involved in various immune responses, including tumor rejection (Cui et al., 1997), early protection from infectious agents (Bendelac et al., 1997), and regulation of autoimmune diseases (Godfrey et al., 2000; Gombert et al., 1996). These in vivo roles in immune responses are likely to be directly linked to the observations that following TCR stimulation, NKT cells develop augmented cytotoxicity directed at tumor targets as well as certain immature cells (Cui et al., 1997) and promptly produce large amounts of cytokines, including IL-4 and IFN-g, as well as TGFb and IL-10 (Bendelac et al., 1995, 1997; D’Orazio and Niederkorn, 1998; Sharif et al., 2002). Although the physiological ligand for CD1-restricted NKT cells is unknown, a-galactosylceramide (a-GalCer), a glycolipid isolated from marine sponges that speciﬁcally binds CD1d, has been shown to be a potent activator for Va14þ NKT cells (Kawano et al., 1997). After injection of a-GalCer into mice, the activation of NKT cells results in rapid production of cytokines,

266

HONG JIANG AND LEONARD CHESS

which are known to be involved in the activation of cell types involved in both innate immunity (NK cells and dendritic cells) and adaptive immunity (T cells and B cells). In addition, there is evidence that NKT cells via cytokine secretion can either inﬂuence or work in concert with other regulatory cells to control autoimmunity. For example, the secretion of cytokines by NKT cells favors Th2 differentiation and also can regulate the function of CD8þ regulatory T cells (Bendelac et al., 1997; Wilbanks and Streilein, 1990). On the other hand, regulatory CD4þ T cells can suppress NKT function (Azuma et al., 2003). Thus, there is a rather complex set of interactions between innate immune cells like the NKT cell and the control of adaptive immune responses mediated by either CD4þ or CD8þ regulatory T cells. As a consequence, NKT cells have been shown to inﬂuence the course of a variety of autoimmune disease models in mice and also appear to involved in a number of human autoimmune diseases including type 1 diabetes and multiple sclerosis. It is of interest that prominent among the diseases effected by NKT cells are those primarily induced by Th1 cells, including diabetes and EAE models (Furlan et al., 2003; Godfrey et al., 2000; Singh et al., 2001). In these diseases, the evidence strongly suggests that the Th2-favoring cytokines IL-4 and IL-10 play an important role. For example, it was found that the number and function of NKT cells are diminished in the thymus and peripheral lymphoid organs of young NOD mice (Baxter et al., 1997; Hammond et al., 1998; Sharif et al., 2002). In particular, anti-CD3-induced IL-4 secretion is barely detectable in young NOD mice. Adoptive transfer of cell populations enriched for NKT cells prevents IDDM in NOD recipients (Baxter et al., 1997; Falcone et al., 1999; Sharif et al., 2001). Moreover, depletion of NKT cells early in the evolution of diabetes in the NOD mice accelerates the onset of diabetes (Frey and Rao, 1999). Recently, the question of whether functional deﬁciency in NKT cells has a causative effect on the onset of diabetes was addressed by comparing the cumulative incidence of diabetes and progression of insulitis in CD1-deﬁcient NOD mice with CD1-heterozygous controls. In addition, in these studies, the effect of a-GalCer treatment on the development of diabetes in NOD and CD1-deﬁcient NOD mice was assessed. It was found that lack of CD1-restricted NKT cells promotes the development of diabetes, whereas activation of Va14þ NKT cells by a-GalCer suppresses disease in the NOD model. Similarly, in a model of colitis that is induced by dextran sodium sulfate, depletion of NKT cells accelerates the onset of disease (Saubermann et al., 2000). In the latter study, treatment with a glycolipid ligand that is recognized by and triggers NKT cells resulted in a signiﬁcant improvement of colitis; this effect was abrogated in CD1d-deﬁcient mice. Similarly, administration of a-GalCer prevents C57BL/6 mice from myelin antigen (MOG)-induced induced EAE.

IMMUNOREGULATION AND T CELL SUBSETS

267

A reduction in number or altered function of NKT cells has also been correlated with autoimmune disease in humans. In patients with multiple sclerosis who are in relapse, the frequency of Va24–JaQ NKT cells was greatly reduced in comparison with normal donors or patients with other autoimmune/ inﬂammatory neurological diseases. In addition, diabetic individuals had lower frequencies of Va24–JaQ NKT cells in comparison with their nondiabetic monozygotic twins (Wilson et al., 1998). The few Va24–JaQ NKT cell clones that could be isolated from these diabetic patients displayed a speciﬁc defect in IL-4 production. Taken together, these converging studies suggested a prominent role of NKT cells in natural protection against destructive Th1-mediated autoimmunity in type 1 diabetes. To directly test this hypothesis, a direct and highly speciﬁc CD1d/a-GalCer tetramer-based methodology was devised by the Bendelac laboratory to test whether humans with type 1 diabetes (T1D) have associated NKT cell defects. Although marked and stable differences in NKT cells among individuals was observed, the study of diabetic patients and healthy controls, including discordant twin pairs, demonstrates that NKT cell frequency and IL-4 production are conserved during the course of type 1 diabetes. These results do not necessarily refute the hypothesis that NKT cell defects underlie T1D but instead, taken together with the previous studies, may indicate that immunoregulation of autoimmune disease is mediated by several subsets of immunoregulatory cells working in concert (Lee et al., 2002b). Thus, in any particular disease, autoimmunity may not reﬂect a single deﬁciency in one subset, as one can sometimes create artiﬁcially in a genetically altered mouse, but instead a defect in an integrated system of immunoregulation mediated at different levels by multiple T cell subsets. Thus, as a member of the family of immunocompetent cells participating in the innate immune response, NKT cells are positioned to inﬂuence and interact with other regulatory T cells during the early phases of the autoimmune response (Bendelac and Fearon, 1997). Their interaction with CD4þ T cells is implicit in their capacity to secrete cytokines which, like IL-4 and IL-10, may shift the Th1–Th2 balance. In addition, there is evidence that NKT cells may also interact with CD8þ T cells in immunoregulation. For example, NKT cells have been shown to have a unique role in tolerance induction in immune-privileged sites (Sonoda et al., 1999). For example, mice deﬁcient in CD1, which speciﬁcally lack NKT cells, could not be tolerized to OVA in the ACAID (anterior chamber-associated immune deviation) model unless they were reconstituted with NKT cells together with CD1expressing APCs (Sonoda et al., 2001). However, NKT cells could not transfer tolerance, suggesting that they have a role in the differentiation of another Tr cell subset which may be CD8þ regulatory T cells that were previously implicated in ACAID (Wilbanks and Streilein, 1990). The mechanisms through

268

HONG JIANG AND LEONARD CHESS

which NKT cells may regulate tolerance induction are complex and involve interactions with CD4þ and CD8þ regulatory cells. C. The CD8þ T Suppressor Cells As noted in Section II, although the role of CD8þ T cells in regulating immune responses to both self and foreign antigens dominated the immunologic literature for more then a decade, interest had waned in the mid-1980s. This interest in CD8þ suppression was rekindled in the early 1990s by experiments in the EAE model induced by the 1-9Nac MBP in B10PL mice. This model was particularly useful because it permitted a precise dissection of the control of the antigen-speciﬁc immune response to a pathogenic self-antigen during the induction, recovery, and relapse from disease. Moreover, because mice become resistant to EAE following the initial induction of diseases, it permitted unequivocal experiments showing that CD8þ T cells control this resistance (Jiang et al., 1992; Koh et al., 1992). Thus, in this model, it is known that unlike chronic relapsing EAE models, which have spontaneous alterations of remissions and relapses without further antigen stimulation, 1-9Nac MBP induced EAE in B10PL mice is an acute monophasic model. The animals develop clinical symptoms of EAE about two weeks after EAE induction and the symptoms persist for about 2 weeks and are followed by spontaneous lifelong recovery. Importantly, most EAErecovered mice are resistant to rechallenge with MBP. The essential observation made in the early 1990s was that this resistance is abolished by depleting the animals of CD8þ T cells using anti-CD8 monoclonal antibodies prior to the second induction of EAE. Interestingly, depletion of CD8þ T cells before initial induction of EAE did not alter the onset, incidence, or severity of disease, demonstrating that CD8þ T cells are not involved in regulating the ﬁrst episode of disease. When these mice recover from the ﬁrst episode of disease and are allowed to recover normal levels of CD8þ T cells, they, unlike control mice, are no longer resistant to reinduction of disease with 1-9Nac MBP and develop EAE upon reimmunization (Jiang et al., 1992, 2001). Thus, CD8þ T cells require priming during the ﬁrst episode of EAE to regulate CD4þ T cells triggered by secondary MBP stimulation in vivo. This requirement for priming of the CD8þ T cells is, of course, a characteristic of immunocompetent T cells during the evolution of antigen-speciﬁc adaptive immune responses. The importance of CD8þ regulatory T cells was also observed in studies of CD8-/- ‘‘knockout’’ mice. Thus, when CD8-/- knockout mice are bred with the EAE-susceptible PL/J strain, the CD8-/- mice develop more chronic EAE than do the wild-type PL/J mice, reﬂected by a higher frequency of relapses in the CD8-/- mice (Koh et al., 1992). These experiments provide evidence that CD8þ T cells play a key role both in inducing resistance to autoimmune EAE

IMMUNOREGULATION AND T CELL SUBSETS

269

and in abrogating recurrent relapsing episodes of autoimmunity in vivo. The mechanisms responsible for the spontaneous recovery from the ﬁrst episode of EAE in CD8þ T cell intact mice are not completely understood, but regulatory CD8þ T cells are not required. Presumably, the mechanisms of spontaneous recovery during the ﬁrst episode in normal mice involve other homeostatic mechanisms, including the innate immune response as well as the regulatory CD4þ T cells that have been discussed. In addition to the involvement of CD8þ T cells in the regulation of autoimmune responses, the involvement of CD8þ T cells in the control of the immune response to foreign antigen as well as superantigens was also rekindled in the 1990s. For example, it was shown that CD8þ T cells control the immune response of H-2d mice to the conventional foreign antigen Hen Egg Lysozyme (HEL) (Nanda and Sercarz, 1996). Thus, Vba haplotype mice which lack 10 TCR Vb gene segments respond efﬁciently to the HEL peptide 74–96. In contrast, normal Vbb H-2d mice do not respond to the HEL peptide 74–96. Importantly, the normal Vbb H-2d mice depleted of CD8þ T cells in vivo (by anti-CD8 Mab treatment) respond vigorously to this antigen. These studies demonstrate that CD8þ T cells are involved in the control of the immune response to conventional foreign antigens like HEL. Together with the experiments already reviewed, these studies suggest that CD8þ T cells govern immune responses to both self and foreign antigens. In addition, studies of the control of the immune response to superantigen were of interest. It was known (Kawabe and Ochi, 1991) that there is a deletion of 30 to 40% of CD4þVb8þ cells 7 to 14 days after a single injection of SEB in mice. It was found that the downregulation of CD4þVb8þ T cells below baseline is not observed in mice depleted of CD8þ cells by treatment either with monoclonal anti-CD8 antibody or in CD8þ T cell deﬁcient b2 M-/-mice. Moreover, following SEB administration, splenic and lymph node CD8þ T cells preferentially recognizing CD4þVb8þ T cells are generated and can be speciﬁcally expanded in vitro. These CD8þ T cells are cytotoxic to autologous CD4þVb8þ T cells but not to autologous CD4þVb8- T cells. Furthermore, this cytotoxicity is dependent on recognition of b2-microglobulin-associated molecules and is inhibited by antiserum speciﬁc for Qa-1 molecules but not by antibody to classical MHC class I-a molecules. Together, these data support the idea that the regulation of the immune response to SEB may be mediated in part by Qa-1 restricted CD8þ T cells (Jiang et al., 1995, 1998). These data are consistent with a model of speciﬁc immunoregulation in which following superantigen activation, Qa-1/self-peptide motifs expressed on activated CD4þ T cells are recognized by the ab TCR expressed by precursor regulatory CD8þ T cells. These CD8þ T cells are induced to differentiate and downregulate CD4þ T cells expressing the same particular Qa-1/self-peptide

270

HONG JIANG AND LEONARD CHESS

motifs. In principle, the Qa-1 binding peptide/s could be TCR Vb peptide/s or alternatively other self-peptide/s bound to Qa-1 after T cell activation. This model of immunoregulation predicts that vaccination of mice with SEB-activated CD4þ T cells will induce the Qa-1 restricted regulatory CD8þ T cells. This prediction was shown to be correct by experiments in which mice vaccinated with SEB activated CD4þVb8þ T cells were shown to induce splenic CD8þ T cells, which preferentially lysed Vb8þ targets in a Qa-1 restricted manner (Jiang et al., 1998). In addition, in the EAE model, T cell vaccination with MBP reactive CD4þVb8þ T cells induced both CD8þ T cell dependent protection from EAE and the emergence of CD8þ T cells capable of recognizing MBP-reactive CD4þ T cell clones in a Qa-1 restricted manner. Moreover, T cell hybrids were derived from mice vaccinated with an MBP reactive CD4þVb8þ T cell clone, which speciﬁcally recognized some but not all MBP-reactive Vb8þ T cell clones. Importantly, the speciﬁc recognition could be blocked by antibodies to Qa-1, CD8, and the ab TCR (Jiang et al., 1998, 2001). These data conﬁrmed that the interaction between the regulatory CD8þ T cells and target CD4þ T cells is through the recognition of Qa-1/selfpeptide complex expressed on CD4þ T cells by ab TCR on regulatory CD8þ T cells. The tri-molecular interaction between the Qa-1 restricted regulatory CD8þ T cells and target CD4þ T cells is illustrated by Fig. 2. Furthermore, these CD8þ T cell hybrids were generated from mice vaccinated with antigen triggered CD4þ TH1 cells during T cell vaccination (TCV) and, in vitro, distinguish mature MBP-reactive TH1 from TH2 cells in a Qa-1 restricted manner (Jiang et al., 2001). In vivo, protection from EAE induced by TCV is dependent on CD8þ T cells and MBP-reactive TH1 Vb8þ clones, but not TH2 Vb8þ clones, employed as vaccine T cells, protect animals from subsequent induction of EAE. Moreover, in vivo depletion of CD8þ T cells during the ﬁrst episode of EAE results in skewing of the TH phenotype toward TH1 upon secondary MBP stimulation. These data provide evidence that CD8þ T cells control autoimmune responses, in part, by regulating the TH phenotype of self-reactive CD4þ T cells (Jiang et al., 2001). The biological function of regulatory CD8þ T cells was further studied in the murine EAE model of autoimmunity. Interestingly, CD4þ T cells isolated from EAE-recovered animals do respond to 1–9 Nac MBP, in vitro (Jiang et al., 2001). Thus, there is an abundance of self-reactive T cell clones in the periphery of both EAE-recovered and CD8þ T cell-depleted EAE-recovered (CD8-/EAE) mice, despite the fact that EAE-recovered mice are resistant but CD8-/EAE mice are susceptible to EAE reinduction. This implies that although both EAE-recovered and CD8-/EAE mice have T cell clones proliferating in response to MBP, they were likely to be qualitatively very different because the biological consequences of secondary immunization with MBP had profoundly different clinical outcomes. This reasoning led us to propose

IMMUNOREGULATION AND T CELL SUBSETS

271

Fig 2 Tri-molecular interaction between regulatory CD8þ T cells and activated autologous CD4þ T cells. This ﬁgure illustrates that the speciﬁc tri-molecular interaction between the regulatory CD8þ T cells and target CD4þ T cells is through the recognition of Qa-1/self-peptide complex expressed on certain activated CD4þ T cells by ab TCR on regulatory CD8þ T cells. (See Color Insert.)

the hypothesis that the clonal composition of the peripheral MBP-reactive CD4þ TCR repertoire is regulated by CD8þ T cells and thus will be different in EAE-recovered mice and EAE-recovered mice depleted of CD8þ T cells. This hypothesis was tested by comparing the composition of the MBP-reactive TCR repertoire, following MBP immunization, in naive mice (control), mice recovered from EAE, and CD8þ T cell depleted, EAE-recovered mice (CD8-/EAE) (Jiang et al., 2003). The outline of the experimental design and a schematic illustration of representative results are shown in Fig. 3. The studies demonstrate that one mechanism for limiting the outgrowth of potentially pathogenic self-reactive clones in the periphery is the selective downregulation of these clones by CD8þ T cells. Evidence that the control of CD4þ Vb8.2þ MBP-reactive T cells was selective was revealed by PCRbased CDR3 length distribution analysis of the TCR repertoire and by direct sequence analysis of the CDR3 regions. In CD8-/EAE mice, 60% of the MBP-reactive Vb8.2 repertoire is dominated by the signiﬁcant outgrowth of only a few clones with unique sequences. The signiﬁcant outgrowth of these dominant clones, each representing >2.5% of the repertoire, was not observed in EAE-recovered mice with intact CD8þ T cells. In contrast, the TCR

272

HONG JIANG AND LEONARD CHESS

Fig 3 Regulatory CD8þ T cells control self-reactive TCR Vb repertoire in EAE mice. The ﬁgure shows the outline of the experimental design and a schematic illustration of representative results demonstrating that regulatory CD8þ T cells control self-reactive TCR Vb repertoire in EAE mice. The MBP-reactive TCR repertoire, following MBP immunization, in naive mice (control), mice recovered from EAE and CD8þ T cell-depleted, EAE-recovered mice (CD8-/EAE) (Jiang et al., 2003) were compared by TCR Vb surface staining, PCR-based CDR3 length distribution analysis of the TCR repertoire, and direct sequence analysis of the CDR3 regions. The studies demonstrated that regulatory CD8þ T cells selectively downregulate certain but not all MBP-reactive T cells within the TCR Vb8.2 family in EAE-recovered mice.

repertoire of MBP-reactive CD4þ T cells in the EAE-recovered mice was composed predominantly of a highly diverse set of self-reactive clones with limited outgrowth. Because the CD8þ T cell depleted EAE-recovered mice are highly susceptible to clinical EAE, whereas CD8þ T cell intact EAErecovered mice are not, it is likely that the dominant clones which emerge following secondary MBP immunization in CD8þ T cell depleted EAErecovered mice contain the potentially encephalitogenic CD4þ T cells. This idea was further supported by our observation that MBP-reactive CD4þ Vb8.2þ T cell clones derived from CD8þ T cell depleted EAErecovered mice were more likely to induce EAE following adoptive transfer into naive mice than clones derived from EAE-recovered mice with intact CD8þ T cells. Furthermore, adoptive transfer of CD8þ T cells isolated from EAE-recovered mice but not naive mice modiﬁed MBP-reactive TCR Vb8.2 but not Vb6 repertoire in recipient mice (Jiang et al., 2003). Taken together, these in vivo and in vitro studies provide evidence that in addition to their TCR Vb speciﬁcity, CD8þ T cells only selectively downregulate certain

IMMUNOREGULATION AND T CELL SUBSETS

273

but not all self-reactive T cells within the TCR Vb8.2 family. These regulatory CD8þ T cells play a key role in controlling self-reactive TCR repertoire by selectively downregulating the clones enriched in the potentially pathogenic self-reactive T cells in the periphery. The residual highly diverse nonpathogenic self-reactive TCR repertoire is preserved by this selective downregulation. Taken together, it thus appears that the regulatory CD8þ T cells described are Qa-1 restricted and that they regulate immune response in a TCR Vb and TH type speciﬁc manner. Since regulatory CD8þ T cells control peripheral immune responses to both self and foreign antigens by selectively downregulating certain but not all antigen-activated T cells thus regulating peripheral TCR repertoire, it is of interest to consider why only some but not all antigenactivated clones are downregulated. As has been noted, it is known that Qa-1 is only minimally expressed on resting lymphoid cells and that, unlike classical MHC class Ia molecules, its expression is dependent on activation. Although the precise Qa-1 binding peptide/s in this system have not been identiﬁed yet, it is likely that the major determinant of whether or not a CD4þ T cell will process and present Qa-1/self-peptide will be the consequence of initial cognitive encounter between the particular ab TCR expressed by antigen-reactive CD4þ T cells and the MHC class II/antigen peptide complex presented on antigen-presenting cells. One possible explanation for the speciﬁc recognition of certain clones that could be envisioned is an idiotype model in which the TCR expressed by regulatory CD8þ T cells will preferentially recognize Qa-1 complexed with only certain TCR Vb peptides. Alternatively, one can envision an afﬁnity model in which the precise threshold for antigen-activated CD4þ T cells to process and present Qa-1/self-peptides may be a function of the afﬁnity/avidity of the interaction between antigen-speciﬁc CD4þ T cells and MHC class II/antigen peptides expressed on antigen-presenting cells. We have previously proposed an ‘‘afﬁnity model’’ that hypothesizes that CD4þ T cells will process and present their own Qa-1/self-peptide as a function of the afﬁnity/avidity of their TCRs for the MHC/antigen peptide on APC. Thus, TCRs with either low or high afﬁnity interaction with MHC/ antigen peptide will not express Qa-1/peptide complexes whereas TCRs with intermediate afﬁnity will express Qa-1/peptide. As a consequence, only T cells with ‘‘intermediate’’ afﬁnity, above and under certain thresholds, will be regulated by the Qa-1 restricted CD8þ regulatory cells (Jiang and Chess, 2000). Evidence in support of this hypothesis came from our observation in EAE that MBP-reactive CD4þ clones which are under the control of CD8þ T cells are the clones with higher growth potential responding to MBP (Jiang et al., 2003). Because high afﬁnity self-reactive clones are deleted intrathymically and the low afﬁnity clones are not properly triggered by MBP, these higher growth potential clones presumably possess intermediate afﬁnity to MBP. The precise

274

HONG JIANG AND LEONARD CHESS

threshold for these CD4þ T cells to process and present self-peptide coupled to Qa-1 molecules will not only be a function of the quality, intensity, and duration of the initial tri-molecular complex interaction between antigenspeciﬁc CD4þ T cells and antigen-presenting cells but may also be inﬂuenced by signaling via co-stimulatory molecules including possible CD28/B7, CTLA-4/B7, and CD40/CD40L interactions. Moreover, these same factors may determine whether the surface Qa-1 molecules on CD4þ T cells are predominately composed of Qa-1/Qdm or Qa-1/self-peptide or both. The expression of Qa-1/self-peptide will induce speciﬁc regulatory CD8þ T cells, which may be further regulated by the Qa-1/Qdm expression through their CD94/NKG2 receptor. Thus, either the particular Qa-1 binding peptide/s (idiotype model) or the afﬁnity/avidity of antigen-reactive T cells (afﬁnity model), or both, determines the susceptibility of antigen-activated CD4þ T cells to the selective downregulation by the CD8þ T cells. We initially proposed that the source of Qa-1 binding peptide/s would be derived from the TCR Vb proteins. This could clearly explain the Vb speciﬁcity of the regulation by the CD8þ T cells. Based on the newly developed experimental evidence in analyzing the MBP-reactive TCR Vb repertoire that CD8þ T cells selectively downregulate certain but not all MBP-reactive Vb8.2 and Vb13 T cells which have higher growth potential responding to MBP in vivo in EAE mice (Jiang et al., 2003), we have modiﬁed our theory. The modiﬁed theory suggests that whether or not a particular CD4þ T cell is subject to CD8þ T cell downregulation depends on whether the CD4þ T cell is capable of presenting self-peptide/s coupled to Qa-1 rather than which Qa-1 binding self-peptide the CD4þ T cell presents. The selective expression of Qa-1/self-peptide/s on certain activated CD4þ T cells is determined by the afﬁnity/avidity of T cells as a consequence of activation. This modiﬁed theory suggests that a non-TCR Vb peptide, which could be common in all activated T cells, could also, like Vb peptide/s, bind to Qa-1 and be recognized by the Qa-1 restricted regulatory CD8þ T cells. For example, heat shock proteins are induced by T cell activation and heat shock protein peptide/s are known to bind to Qa-1 (Imani and Soloski, 1991). In either case, the expression of Qa-1/self-peptide complex by activated CD4þ T cells could be governed by the afﬁnity/avidity of the TCR on CD4þ T cells and the co-stimulatory molecules involved during the initial encounter of TCR on CD4þ T cells and MHC class II/antigen peptide on antigen-presenting cells. In either case, certain TCR Vb speciﬁc regulation may be observed due to the preferential usage of certain TCR Vb chains by the responding T cell population to a particular antigen. For example, we have shown that CD8þ T cells selectively downregulate certain but not all MBP-reactive T cell clones in EAE-recovered mice only in TCR Vb8.2 and Vb13 but not in other Vb families (Jiang et al., 2003). It is of interest that in B10PL mice, the Vb8.2

IMMUNOREGULATION AND T CELL SUBSETS

275

and Vb13 families of CD4þ T cells represent the two major T cell populations that are activated by 1-9Nac MBP and are largely responsible for clinical EAE (Acha-Orbea et al., 1988; Zamvil et al., 1985). In this case, either ‘‘Vb peptide’’ or ‘‘non-Vb peptide’’ or both could account for the ‘‘Vb speciﬁc’’ phenomenon. It is important to emphasize that CD8þ T cells require priming during the ﬁrst episode of EAE to regulate the outgrowth of potentially pathogenic CD4þ T cells triggered by secondary MBP challenge in vivo. The evidence that regulatory CD8þ T cells require priming during the ﬁrst episode of EAE is simply that B10PL mice depleted of CD8þ T cells during the initial induction of EAE recover from EAE normally but are not resistant to rechallenge with 1–9Nac MBP. In contrast, EAE-recovered mice with CD8þ T cells primed during the ﬁrst episode are resistant to reinduction of EAE unless they are depleted of CD8þ cells prior to reinduction of EAE (Jiang et al., 1992, 2001, 2003). Based on these observations, we envision that the cellular events during the evolution of natural EAE are as follows: MBP-reactive CD4þ T cells are activated by the ﬁrst encounter with MBP to induce both clinical EAE and regulatory CD8þ T cells. The regulatory CD8þ T cells are not induced in time to downregulate MBP-reactive CD4þ T cells during the ﬁrst episode of EAE but are primed to downregulate MBP-reactive CD4þ T cells activated by secondary MBP immunization. In this study, CD8þ T cells were depleted during the induction of EAE (ﬁrst episode). These mice developed EAE and clinically recovered. Newly generated CD8þ T cells reappear during the recovery and are present at the time of secondary MBP immunization in vivo. When the MBP-reactive TCR Vb repertoire of these mice was compared to that of EAE-recovered mice with primed CD8þ T cells, the profound effect of the primed CD8þ T cells on the CD4þ MBP-reactive TCR Vb repertoire was observed. This sequence of events is reminiscent of the general biology of CD8þ T cells involved in the response to viruses. During the initial infection with most viruses, CD8þ T cells are not involved in the recovery that may occur within the ﬁrst week or so (this initial recovery is mediated, in part, by the innate immune response, including macrophages, NK and NKT cells, and gd T cells) but they are primed during the initial infection period and are clearly involved in resistance to reinfection or persistent virus infection. In summary, a general model of peripheral regulation by Qa-1 restricted regulatory CD8þ T cells is illustrated in Fig. 4. There are two unique features of Qa-1 restricted regulatory CD8þ T cells, which distinguish them from other regulatory NKT or CD4þ T cells. First, the Qa-1 restricted regulatory CD8þ T cells selectively downregulate certain but not all antigen-activated T cells. The molecular interaction between the regulatory CD8þ T cells and target CD4þ T cells is through the recognition of Qa-1/self-peptide complex, which may be only preferentially expressed by T cells with intermediate afﬁnity/avidity to the antigens, by ab TCR on regulatory CD8þ T cells

276

HONG JIANG AND LEONARD CHESS

Fig 4 Model of cognate interactions in the induction and function of Qa-1 restricted regulatory CD8þ T cells. The Qa-1 restricted regulatory CD8þ T cell selectively downregulate certain but not all antigen-activated CD4þ T cells based on the speciﬁc recognition of Qa-1/self-peptide/s expressed on certain CD4þ T cells by the ab TCR on the CD8þ T cells. In this regard, we have demonstrated in the EAE model that self-reactive CD4þ T cells which are selectively downregulated by the CD8þ T cells are enriched in potentially pathogenic self-reactive T cell clones (Jiang et al., 2003). (See Color Insert.)

(Jiang et al., 1998, 2001). Second, the Qa-1 restricted regulatory CD8þ T cells require priming in vivo. The regulatory CD8þ T cells are induced by antigenactivated T cells expressing TCRs of intermediate afﬁnity/avidity during the primary immune response, perhaps with help from professional antigenpresenting cells, and then, in turn, selectively downregulate those T cells in the later stage of immunity. Ultimately, the regulatory CD8þ T cells, in concert with other regulatory mechanisms, control the peripheral TCR repertoire during the course of immune responses to both self and foreign antigens. With respect to the self-reactive clones, the peripheral repertoire is composed of mainly low and intermediate afﬁnity/avidity CD4þ T cells. The intermediate afﬁnity T cells are enriched in the potentially pathogenic self-reactive T cell clones and are controlled by CD8þ regulatory cells. The very high afﬁnity selfreactive clones have already been eliminated in the thymus so that the biological consequence of selective regulation of these intermediate clones will be the control of autoimmunity. However, with respect to foreign antigens,

IMMUNOREGULATION AND T CELL SUBSETS

277

the repertoire is primarily composed of intermediate and high afﬁnity clones. The biological consequence of the selective downregulation of the intermediate afﬁnity but not the high afﬁnity foreign-reactive T cell population will be to promote afﬁnity maturation. Thus, the biological consequences of the selective downregulation of T cells with intermediate afﬁnity/avidity TCRs to antigens by the Qa-1 restricted CD8þ T cells is to shape the peripheral T cell repertoire in order to ensure peripheral self-tolerance and facilitate afﬁnity maturation to foreign antigens. IV. An Integrated Model of Immunoregulation by NKT, CD41CD251, and Qa-1 Restricted CD81 T Cell Subsets

In this chapter, we have emphasized the fact that the immune system has evolved a variety of regulatory mechanisms mediated by distinct T cell subsets to control peripheral immunity. In particular, we have discussed various pathways of immunoregulation mediated by suppressor subsets of CD4þ and CD8þ T cells as well as by NKT cells. The regulation of immune responses mediated by these regulatory T cell subsets is superimposed on the regulatory mechanisms induced by antigen activation of the immunocompetent cells which are independent of the regulatory T cells and include antigen-induced apoptosis, anergy as well as antigen-induced differentiation of precursor T cells into Th subsets secreting regulatory cytokines (see Fig. 1). Each of the regulatory T cell subsets expresses distinct receptors, employs different effector mechanisms, and functions predominately at different stages during the evolution of the peripheral immunity (see Table I). Thus, the NKT cells and CD4þCD25þ regulatory cells exist in the early neonatal period as ‘‘natural suppressor cells’’ prior to antigen activation and function primarily during the early ‘‘innate’’ and/or primary immune responses (see Table I and Fig. 5). The NKT cells are endowed with ab TCRs which speciﬁcally recognize glycolipid molecules often expressed by various pathogens and are presumably expressed also by tumor cells, activated blasts, and injured ‘‘apoptotic’’ cells. These NKT cells are poised to secrete IL-4 and IL-10, which are known to inﬂuence the balance of Th1 or Th2 cells that emerge during the primary immune response. It is of interest that in addition to NKT cells, the CD4þCD25þ regulatory subset also exists in the periphery in the early neonatal period and persist in the peripheral lymphoid system as cells capable of potently suppressing the outgrowth of immunocompetent cells. These cells, like the NKT cells, can function during the primary immune response and do not require speciﬁc induction. In vitro, the suppressor function of these cells can be shown to dependent of cell-cell contact but these cells can also secrete immunoregulatory cytokines including TGFb which may be involved in the their suppressor function in vivo. The precise speciﬁcity of these cells for their targets remain

TABLE I Characteristics of NKT, CD4þCD25þ and Qa-1 restricted CD8þ Regulatory T Cell Subsets Subsets of regulatory T cells NKT cells

CD4þCD25þ regulatory T cells CD8þ regulatory cells

Target cells for suppression

Restriction element/antigen

Stage of immunity affected Innate

Regulatory mechanisms

Tumor cells, pathogenactivated T cells, and/or APCs expressing CD1d/glycolipid T cells, probably APCs

CD1d/glycolipid

IL-4, IL-10 TGFb, IFN-g

Unknown; not MHC restricteda

Primary, earlyb

Requires cell–cell contact, cytokines

Certain activated T cells expressing Qa-1/self-peptide/s

Qa-1/hydrophobic self-peptides

Secondary, lateb

Cytotoxicity; requires cell–cell contact; cytokines

In vivo function Destruction of tumors and infectious pathogens; regulation of Th1-mediated autoimmune diseases Prevention of a variety of autoimmune diseases, regulation of allo-graft rejection Fine-tune peripheral TCR repertoire; maintain self-tolerance and control autoimmune disease

The speciﬁc interaction between CD4þCD25þ regulatory T cells and the target T cells is currently unknown. CD4þCD25þ regulatory T cells isolated from naive unprimed mice protect recipient animals from autoimmune diseases when adoptively transferred. In contrast, Qa-1 restricted CD8þ regulatory T cells require priming during primary immune response in order to regulate the secondary immune response in vivo. a

b

IMMUNOREGULATION AND T CELL SUBSETS

279

Fig 5 Regulatory T cells control the peripheral induction and clonal outgrowth of self-reactive T cells. This ﬁgure illustrates various pathways of immunoregulation mediated by suppressor subsets of NKT, CD4þ, and CD8þ T subsets. Each of the regulatory T cell subsets expresses distinct receptors, employs different effector mechanisms, and functions predominately at different stages during the course of the peripheral immune response. The NKT and CD4þCD25þ regulatory cells are ‘‘natural suppressor cells’’ and are present prior to antigen activation and function primarily during the early ‘‘innate’’ and/or primary adaptive immune responses. In contrast, the CD8þ regulatory cells are induced to differentiate into suppressor effector cells during the primary immune response and they function as effector suppressor cells predominately during the secondary and memory phases of immunity. (See Color Insert.)

unknown and it is not clear whether antigen presenting cells and/or T cells are the targets of CD4þCD25þ mediated suppression. Moreover, although the CD4þCD25þ suppressor cells express conventional ab TCRs, the evidence suggests that these TCRs are not involved in the recognition of the targets of suppression. In contrast to both the NKT cells and the CD4þCD25þ regulatory cells, the CD8þ regulatory T cells are not observed in naive animals prior to antigen encounter. As a consequence, adoptive transfer of CD8þ T cells from naive animals has no effect on the outcome of autoimmune responses and depletion of CD8þ T cells prior to the ﬁrst induction of EAE has no effect on the ﬁrst episode of disease. However, the CD8þ regulatory cells function like classical

280

HONG JIANG AND LEONARD CHESS

immunocompetent cells activated during adaptive immune responses. Thus, the CD8þ regulatory cells are induced to differentiate into suppressor effector cells during the primary immune response and they function as effector suppressor cells predominately during the secondary and memory phases of immunity. Thus, adoptive transfer of CD8þ T cells from self-antigen primed mice will retard the outgrowth of the potentially pathogenic self-reactive T cells. In this regard, it is of interest that CD8þ regulatory T cells are known to mediate resistance to autoimmunity following initial recovery from disease and to decrease the incidence and severity of relapse of the disease. In contrast to the CD4þCD25þ T cells, the CD8þ regulatory T cells utilize their ab TCRs to recognize target cells in a MHC-restricted fashion. Thus, the CD8þ regulatory cells are Qa-1 restricted and selectively downregulate Qa-1/self-peptide complexes preferentially expressed on certain but not all activated T cells. We propose a general synthesized model of peripheral regulation of immunity incorporating pathways mediated by the three different T cell subsets we have discussed. Here, we use regulation in control of autoimmunity as an example. As shown in Fig. 5, we envision that during the initiation of autoimmune disease, cell injury or death leads to the release of self-peptide/s which can be processed and presented by antigen-presenting cells to selfreactive CD4þ T cells, which subsequently induce and effect immune responses to self. During these early stages of the primary autoimmunity, the NKT and CD4þCD25þ regulatory cells will directly inﬂuence the emergence of self-reactive clones by controlling the balance of Th1 to Th2 differentiation and cytokine release. In addition, CD4þCD25þ T cells by cell–cell contact will directly suppress certain self-reactive cells by mechanisms that are currently unknown. We envision that the majority of the activated self-reactive T cells are controlled by these regulatory T cells working in concert with the conventional Th1 and Th2 cells as well as activated CD4þCD25- T cells which uniquely express immunosuppressive cytokines (i.e., the Tr1 and Th3 cells). This level of control prevents the majority of potentially pathogenic self-reactive T cells from functioning and therefore generally maintains self-tolerance in the periphery. However, by analogy to thymic negative selection which deletes the majority but not all potentially pathogenic self-reactive T cells, the very existence of autoimmune disease in mammals implies that this ﬁrst level of peripheral regulatory defense mechanisms may not control all potentially pathogenic self-reactive T cell clones. This leak in the ﬁrst level of control may be due to subtle defects in the mechanisms that normally control CD4þCD25þ T cell differentiation, such as defects in Foxp3 or subtle defects in the differentiation of the CD1-restricted NKT cells. Thus, we believe that the immune system developed a second level of control mechanisms, including the pathway

IMMUNOREGULATION AND T CELL SUBSETS

281

mediated by Qa-1 restricted CD8þ T cells, to ﬁne-tune the TCR repertoire of self-reactive T cells that have escaped both thymic selection and peripheral regulation of the primary response to self-reactive T cells. Presumably, this was necessary because self-reactive T cells that have escaped both thymic deletion and primary peripheral regulation may contain the potentially pathogenic autoreactive T cells. The Qa-1 restricted regulatory CD8þ T cellmediated pathway represents the only regulatory pathway capable of ﬁnetuning the TCR repertoire. Thus, we envision that self-reactive T cells, which escape the ﬁrst level of regulatory defense mechanisms, are activated by further self-antigen triggering, in vivo, and undergo clonal growth. Some of these outgrowth clones are potentially pathogenic and probably possess intermediate afﬁnity/avidity to self. We have proposed that the intermediate afﬁnity/avidity T cells, which are enriched in potentially pathogenic clones, express Qa-1/self-peptide complexes on their surface. The activated CD4þ T cells expressing Qa-1/self-peptide/s are capable of inducing Qa-1 restricted regulatory CD8þ T cell and also are subject to the speciﬁc downregulation by the CD8þ regulatory T cells. On the other hand, the low afﬁnity self-reactive clones are not pathogenic and they persist in the periphery by escaping downregulation by the CD8þ T cells. In summary, the peripheral immunity is controlled by a well-designed program consisting of a variety of regulatory mechanisms. Thus, despite the abundance of self-reactive clones in the periphery, clinical autoimmunity is usually well controlled. Acknowledgments The research was supported by NIH grants AI39630, AI39675, and National Multiple Sclerosis Society grant RG2938A to HJ and NIH grant U19 AI/46132 to LC.

References Acha-Orbea, H., Mitchell, D. J., Timmermann, L., Wraith, D. C., Tausch, G. S., Waldor, M. K., Zamvil, S. S., McDevitt, H. O., and Steinman, L. (1988). Limited heterogeneity of T cell receptors from lymphocytes mediating autoimmune encephalomyelitis allows speciﬁc immune intervention. Cell 54, 263–273. Asherson, G. L., Colizzi, V., and Zembala, M. (1986). An overview of T-suppressor cell circuits. Annu. Rev. Immunol. 4, 37–68. Askenase, P. W., Hayden, B. J., and Gershon, R. K. (1975). Augmentation of delayed-type hypersensitivity by doses of cyclophosphamide which do not affect antibody responses. Journal of Experimental Medicine 141, 697–702. Azuma, T., Takahashi, T., Kunisato, A., Kitamura, T., and Hirai, H. (2003). Human CD4þCD25þ regulatory T cells suppress NKT cell functions. Cancer Res. 63, 4516–4520. Baxter, A. G., Kinder, S. J., Hammond, K. J., Scollay, R., and Godfrey, D. I. (1997). Association between alphabetaTCRþCD4-CD8- T-cell deﬁciency and IDDM in NOD/Lt mice. Diabetes 46, 572–582. Ben-Nun, A., Wekerle, H., and Cohen, I. R. (1981). Vaccination against autoimmune encephalomyelitis with T-lymphocyte line cells reactive against myelin basic protein. Nature 292, 60–61.

282

HONG JIANG AND LEONARD CHESS

Benacerraf, B., Greene, M. I., Sy, M. S., and Dorf, M. E. (1982). Suppressor T cell circuits. Annals of the New York Academy of Sciences 392, 300–308. Bendelac, A. (1995). Mouse NK1þ T cells. Curr. Opin. Immunol. 7, 367–374. Bendelac, A., and Fearon, D. T. (1997). Innate pathways that control acquired immunity. Curr. Opin. Immunol. 9, 1–3. Bendelac, A., Lantz, O., Quimby, M. E., Yewdell, J. W., Bennink, J. R., and Brutkiewicz, R. R. (1995). CD1 recognition by mouse NK1þ T lymphocytes. Science 268, 863–865. Bendelac, A., Rivera, M. N., Park, S. H., and Roark, J. H. (1997). Mouse CD1-speciﬁc NK1 T cells: Development, speciﬁcity, and function. Annu. Rev. Immunol. 15, 535–562. Bennett, C. L., Christie, J., Ramsdell, F., Brunkow, M. E., Ferguson, P. J., Whitesell, L., Kelly, T. E., Saulsbury, F. T., Chance, P. F., and Ochs, H. D. (2001). The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat. Genet. 27, 20–21. Bloom, B. R., Modlin, R. L., and Salgame, P. (1992a). Stigma variations: Observations on suppressor T cells and leprosy. Annu. Rev. Immunol. 10, 453–488. Bloom, B. R., Salgame, P., and Diamond, B. (1992b). Revisiting and revising suppressor T cells. Immunol. Today 13, 131–136. Bouneaud, C., Kourilsky, P., and Bousso, P. (2000). Impact of negative selection on the T cell repertoire reactive to a self-peptide: A large fraction of T cell clones escapes clonal deletion. Immunity 13, 829–840. Boussiotis, V. A., Freeman, G. J., Gribben, J. G., and Nadler, L. M. (1996). The role of B7-1/B72:CD28/CLTA-4 pathways in the prevention of anergy, induction of, productive immunity, and down-regulation of the immune response. Immunol. Rev. 153, 5–26. Brunkow, M. E., Jeffery, E. W., Hjerrild, K. A., Paeper, B., Clark, L. B., Yasayko, S. A., Wilkinson, J. E., Galas, D., Ziegler, S. F., and Ramsdell, F. (2001). Disruption of a new forkhead/wingedhelix protein, scurﬁn, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat. Genet. 27, 68–73. Burnet, F. M. (1957). A modiﬁcation of Jerne’s theory of antibody production using the concept of clonal selection. Aust. J. Sci. 20. Cantor, H., and Boyse, E. A. (1975a). Functional subclasses of T lymphocytes bearing different Ly antigens. II. Cooperation between subclasses of Lyþ cells in the generation of killer activity. J. Exp. Med. 141, 1390–1399. Cantor, H., and Boyse, E. A. (1975b). Functional subclasses of T-lymphocytes bearing different Ly antigens. I. The generation of functionally distinct T-cell subclasses is a differentiative process independent of antigen. J. Exp. Med. 141, 1376–1389. Cantor, H., Hugenberger, J., McVay-Boudreau, L., Eardley, D. D., Kemp, J., Shen, F. W., and Gershon, R. K. (1978a). Immunoregulatory circuits among T-cell sets. Identiﬁcation of a subpopulation of T-helper cells that induces feedback inhibition. J. Exp. Med. 148, 871–877. Cantor, H., McVay-Boudreau, L., Hugenberger, J., Naidorf, K., Shen, F. W., and Gershon, R. K. (1978b). Immunoregulatory circuits among T-cell sets. II. Physiologic role of feedback inhibition in vivo: Absence in NZB mice. J. Exp. Med. 147, 1116–1125. Cantor, H., Shen, F. W., and Boyse, E. A. (1976). Separation of helper T cells from suppressor T cells expressing different Ly components. II. Activation by antigen: After immunization, antigen-speciﬁc suppressor and helper activities are mediated by distinct T-cell subclasses. J. Exp. Med. 143, 1391–1401. Caux, C., Burdin, N., Galibert, L., Hermann, P., Renard, N., Servet-Delprat, C., and Banchereau, J. (1994). Functional CD40 on B lymphocytes and dendritic cells. [Review] [21 refs]. Res. Immunol. 145, 235–239.

IMMUNOREGULATION AND T CELL SUBSETS

283

Charlton, B., and Lafferty, K. J. (1995). The Th1/Th2 balance in autoimmunity. Curr. Opin. Immunol. 7, 793–798. Chess, L., and Schlossman, S. F. (1977). Human lymphocyte subpopulations. Adv. Immunol. 25, 213–241. Coffman, R. L., and Mosmann, T. R. (1991). CD4þ T-cell subsets: Regulation of differentiation and function. [Review] [14 refs]. Res. Immunol. 142, 7–9. Cohen, I. R., and Weiner, H. L. (1988). T-cell vaccination. Immunol. Today 9, 332–335. Connolly, D. J., Cotterill, L. A., Hederer, R. A., Thorpe, C. J., Travers, P. J., McVey, J. H., Dyson, J., and Robinson, P. J. (1993). A cDNA clone encoding the mouse Qa-1a histocompatibility antigen and proposed structure of the putative peptide binding site. J. Immunol. 151, 6089–6098. Cui, J., Shin, T., Kawano, T., Sato, H., Kondo, E., Toura, I., Kaneko, Y., Koseki, H., Kanno, M., and Taniguchi, M. (1997). Requirement for Valpha14 NKT cells in IL-12-mediated rejection of tumors. Science 278, 1623–1626. D’Orazio, T. J., and Niederkorn, J. Y. (1998). A novel role for TGF-beta and IL-10 in the induction of immune privilege. J. Immunol. 160, 2089–2098. Del Prete, G. (1998). The concept of type-1 and type-2 helper T cells and their cytokines in humans. Int. Rev. Immunol. 16, 427–455. Dialynas, D. P., Quan, Z. S., Wall, K. A., Pierres, A., Quintans, J., Loken, M. R., Pierres, M., and Fitch, F. W. (1983). Characterization of the murine T cell surface molecule, designated L3T4, identiﬁed by monoclonal antibody GK1.5: Similarity of L3T4 to the human Leu-3/T4 molecule. J. Immunol. 131, 2445–2451. Dorf, M. E., and Benacerraf, B. (1984). Suppressor cells and immunoregulation. Ann. Rev. Immunol. 2, 127–158. Durie, F. H., Foy, T. M., Masters, S. R., Laman, J. D., and Noelle, R. J. (1994). The role of CD40 in the regulation of humoral and cell-mediated immunity. [Review]. Immunology Today 15, 406–411. Eardley, D. D., Hugenberger, J., McVay-Boudreau, L., Shen, F. W., Gershon, R. K., and Cantor, H. (1978). Immunoregulatory circuits among T-cells sets. I. T-helper cells induce other T-cell sets to exert feedback inhibition. J. Exp. Med. 147, 1106–1115. Eberl, G., Lees, R., Smiley, S. T., Taniguchi, M., Grusby, M. J., and MacDonald, H. R. (1999). Tissue-speciﬁc segregation of CD1d-dependent and CD1d-independent NKT cells. J. Immunol. 162, 6410–6419. Ehrlich, P. (1900). The Croonian lecture: On immunity. Proc. Royal Soc., London 66, 424. Evans, R. L., Breard, J. M., Lazarus, H., Schlossman, S. F., and Chess, L. (1977). Detection, isolation, and functional characterization of two human T-cell subclasses bearing unique differentiation antigens. J. Exp. Med. 145, 221–233. Exley, M., Garcia, J., Balk, S. P., and Porcelli, S. (1997). Requirements for CD1d recognition by human invariant Valpha24þ CD4-CD8- T cells. J. Exp. Med. 186, 109–120. Falcone, M., Yeung, B., Tucker, L., Rodriguez, E., and Sarvetnick, N. (1999). A defect in interleukin 12-induced activation and interferon gamma secretion of peripheral natural killer T cells in nonobese diabetic mice suggests new pathogenic mechanisms for insulin-dependent diabetes mellitus. J. Exp. Med. 190, 963–972. Fitch, F. W., McKisic, M. D., Lancki, D. W., and Gajewski, T. F. (1993). Differential regulation of murine T lymphocyte subsets. Annu. Rev. Immunol. 11, 29–48. Frey, A. B., and Rao, T. D. (1999). NKT cell cytokine imbalance in murine diabetes mellitus. Autoimmunity 29, 201–214. Furlan, R., Bergami, A., Cantarella, D., Brambilla, E., Taniguchi, M., Dellabona, P., Casorati, G., and Martino, G. (2003). Activation of invariant NKT cells by alphaGalCer administration

284

HONG JIANG AND LEONARD CHESS

protects mice from MOG35-55-induced EAE: Critical roles for administration route and IFNgamma. Eur. J. Immunol. 33, 1830–1838. Gay, D., Maddon, P., Sekaly, R., Talle, M. A., Godfrey, M., Long, E., Goldstein, G., Chess, L., Axel, R., Kappler, J., et al. (1987). Functional interaction between human T-cell protein CD4 and the major histocompatibility complex HLA-DR antigen. Nature 328, 626–629. Germain, R. N., and Benacerraf, B. (1980). Helper and suppressor T cell factors. [Review]. Springer Seminars in Immunopathology 3, 93–127. Gershon, R. K. (1974). T cell control of antibody production. Contemp. Top. Immunobiol. 3, 1–40. Gershon, R. K. (1975). A disquisition on suppressor T cells. Transplant Rev. 26, 170–185. Gershon, R. K. (1976). The role of suppression in immunoregulation. Adv. Exp. Med. Biol. 66, 557–563. Gershon, R. K., and Kondo, K. (1970). Cell interactions in the induction of tolerance: The role of thymic lymphocytes. Immunology 18, 723–737. Gershon, R. K., and Kondo, K. (1971). Infectious immunological tolerance. Immunology 21, 903–914. Gershon, R. K., Mokyr, M. B., and Mitchell, M. S. (1974). Activation of suppressor T cells by tumour cells and speciﬁc antibody. Nature 250, 594–596. Godfrey, D. I., Hammond, K. J., Poulton, L. D., Smyth, M. J., and Baxter, A. G. (2000). NKT cells: Facts, functions and fallacies. Immunol. Today 21, 573–583. Goldrath, A. W., and Bevan, M. J. (1999). Selecting and maintaining a diverse T-cell repertoire. Nature 402, 255–262. Gombert, J. M., Herbelin, A., Tancrede-Bohin, E., Dy, M., Carnaud, C., and Bach, J. F. (1996). Early quantitative and functional deﬁciency of NK1þ-like thymocytes in the NOD mouse. Eur. J. Immunol. 26, 2989–2998. Gorelik, L., and Flavell, R. A. (2000). Abrogation of TGFbeta signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 12, 171–181. Green, D. R., Flood, P. M., and Gershon, R. K. (1983). Immunoregulatory T-cell pathways. Annu. Rev. Immunol. 1, 439–463. Hammond, K. J., Poulton, L. D., Palmisano, L. J., Silveira, P. A., Godfrey, D. I., and Baxter, A. G. (1998). Alpha/beta-T cell receptor (TCR)þCD4-CD8-(NKT) thymocytes prevent insulindependent diabetes mellitus in nonobese diabetic (NOD)/Lt mice by the inﬂuence of interleukin (IL)-4 and/or IL-10. J. Exp. Med. 187, 1047–1056. Hori, S., Nomura, T., and Sakaguchi, S. (2003). Control of regulatory T cell development by the transcription factor Foxp3. Science 299, 1057–1061. Imani, F., and Soloski, M. J. (1991). Heat shock proteins can regulate expression of the Tla regionencoded class Ib molecule Qa-1. Proc. Natl. Acad. Sci. USA 88, 10475–10479. Janeway, C., Jr., and Bottomly, K. (1994). Signals and signs for lymphocyte responses. [Review]. Cell 76, 275–285. Janeway, C. A., Jr. (1988). Do suppressor T cells exist? A reply [editorial]. Scand. J. Immunol. 27, 621–623. Jenkins, M. K. (1994). The ups and downs of T cell costimulation. Immunity 1, 443–446. Jerne, N. K. (1955). The natural selection theory of antibody formation. PNAS 41, 849. Jerne, N. K. (1975). The immune system: A web of V domains. Harvey Lect 70, 93–110. Jiang, H., Braunstein, N. S., Yu, B., Winchester, R., and Chess, L. (2001). CD8þ Tcells control the TH phenotype of MBP-reactive CD4þ Tcells in EAE mice. Proc. Natl. Acad. Sci. USA 98, 6301–6306. Jiang, H., and Chess, L. (2000). The speciﬁc regulation of immune responses by CD8þ T cells restricted by the MHC class IB molecule QA-1. Annu. Rev. Immunol. 18, 185–216. Jiang, H., Curran, S., Ruiz-Vazquez, E., Liang, B., Winchester, R., and Chess, L. (2003). Regulatory CD8þ T cells ﬁne tune the MBP-reactive T cell receptor Vb repertoire during EAE. PNAS 100, 8378–8383.

IMMUNOREGULATION AND T CELL SUBSETS

285

Jiang, H., Kashleva, H., Xu, L. X., Forman, J., Flaherty, L., Pernis, B., Braunstein, N. S., and Chess, L. (1998). T cell vaccination induces T cell receptor Vbeta-speciﬁc Qa-1-restricted regulatory CD8(þ) T cells. Proc. Natl. Acad. Sci. USA 95, 4533–4537. Jiang, H., Ware, R., Stall, A., Flaherty, L., Chess, L., and Pernis, B. (1995). Murine CD8þ T cells that speciﬁcally delete autologous CD4þ T cells expressing V beta 8 TCR: A role of the Qa-1 molecule. Immunity 2, 185–194. Jiang, H., Zhang, S. I., and Pernis, B. (1992). Role of CD8þ T cells in murine experimental allergic encephalomyelitis. Science 256, 1213–1215. Kastner, D. L., Rich, R. R., and Chu, L. (1979). Qa-1-associated antigens. II. Evidence for functional differentiation from H-2K and H-2D antigens. J. Immunol. 123, 1239–1244. Kawabe, Y., and Ochi, A. (1991). Programmed cell death and extrathymic reduction of Vbeta8þ CD4þ T cells in mice tolerant to Staphylococcus aureus entertoxin B. Nature 349, 245–248. Kawano, T., Cui, J., Koezuka, Y., Toura, I., Kaneko, Y., Motoki, K., Ueno, H., Nakagawa, R., Sato, H., Kondo, E., et al. (1997). CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides. Science 278, 1626–1629. Khattri, R., Cox, T., Yasayko, S. A., and Ramsdell, F. (2003). An essential role for Scurﬁn in CD4þCD25þ T regulatory cells. Nat. Immunol. 4, 337–342. Klaus, S. J., Pinchuk, L. M., Ochs, H. D., Law, C. L., Fanslow, W. C., Armitage, R. J., and Clark, E. A. (1994). Costimulation through CD28 enhances T cell-dependent B cell activation via CD40–CD40L interaction. J. Immunol. 152, 5643–5652. Koh, D.-R., Fung-Leung, W.-P., Ho, A., Gray, D., Acha-Orbea, H., and Mak, T.-W. (1992). Less mortality but more relapses in experimental allergic encephalomyelitis in CD8-/-mice. Science 256, 1210–1213. Koulova, L., Clark, E. A., Shu, G., and Dupont, B. (1991). The CD28 ligand B7/BB1 provides costimulatory signal for alloactivation of CD4þ T cells. J. Exp. Med. 173, 759–762. Kuchroo, V. K., Anderson, A. C., Waldner, H., Munder, M., Bettelli, E., and Nicholson, L. B. (2002). T cell response in experimental autoimmune encephalomyelitis (EAE): Role of self and cross-reactive antigens in shaping, tuning, and regulating the autopathogenic T cell repertoire. Annu. Rev. Immunol. 20, 101–123. Kung, P., Goldstein, G., Reinherz, E. L., and Schlossman, S. F. (1979). Monoclonal antibodies deﬁning distinctive human T cell surface antigens. Science 206, 347–349. Lalanne, J. L., Transy, C., Guerin, S., Darche, S., Meulien, P., and Kourilsky, P. (1985). Expression of class I genes in the major histocompatibility complex: Identiﬁcation of eight distinct mRNAs in DBA/2 mouse liver. Cell 41, 469–478. Lederman, S., Yellin, M. J., Krichevsky, A., Belko, J., Lee, J. J., and Chess, L. (1992). Identiﬁcation of a novel surface protein on activated CD4þ T cells that induces contact-dependent B cell differentiation (help). J. Exp. Med. 175, 1091–1101. Lee, P. T., Benlagha, K., Teyton, L., and Bendelac, A. (2002a). Distinct functional lineages of human V(alpha)24 natural killer T cells. J. Exp. Med. 195, 637–641. Lee, P. T., Putnam, A., Benlagha, K., Teyton, L., Gottlieb, P. A., and Bendelac, A. (2002b). Testing the NKT cell hypothesis of human IDDM pathogenesis. J. Clin. Invest. 110, 793–800. Lenardo, M., Chan, K. M., Hornung, F., McFarland, H., Siegel, R., Wang, J., and Zheng, L. (1999). Mature T lymphocyte apoptosis—immune regulation in a dynamic and unpredictable antigenic environment [in process citation]. Annu. Rev. Immunol. 17, 221–253. Lenschow, D. J., Walunas, T. L., and Bluestone, J. A. (1996). CD28/B7 system of T cell costimulation. Annu. Rev. Immunol. 14, 233–258. Littman, D. R. (1987). The structure of the CD4 and CD8 genes. Annu. Rev. Immunol. 5, 561–584.

286

HONG JIANG AND LEONARD CHESS

Littman, D. R., Thomas, Y., Maddon, P. J., Chess, L., and Axel, R. (1985). The isolation and sequence of the gene encoding T8: A molecule deﬁning functional classes of T lymphocytes. Cell 40, 237–246. Maddon, P. J., Littman, D. R., Godfrey, M., Maddon, D. E., Chess, L., and Axel, R. (1985). The isolation and nucleotide sequence of a cDNA encoding the T cell surface protein T4: A new member of the immunoglobulin gene family. Cell 42, 93–104. Maddon, P. J., Molineaux, S. M., Maddon, D. E., Zimmerman, K. A., Godfrey, M., Alt, F. W., Chess, L., and Axel, R. (1987). Structure and expression of the human and mouse T4 genes. Proc. Natl. Acad. Sci. USA 84, 9155–9159. Moller, G. (1988). Do suppressor T cells exist? Scand. J. Immunol. 27, 247–250. Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L. (1986). Two types of murine helper T cell clone. I. Deﬁnition according to proﬁles of lymphokine activities and secreted proteins. J. Immunol. 136, 2348–2357. Mosmann, T. R., and Coffman, R. L. (1989). TH1 and TH2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7, 145–173. Nanda, N. K., and Sercarz, E. (1996). A truncated T cell receptor repertoire reveals underlying immunogenicity of an antigenic determinant. J. Exp. Med. 184, 1037–1043. Piccirillo, C. A., and Shevach, E. M. (2001). Cutting edge: Control of CD8þ T cell activation by CD4þCD25þ immunoregulatory cells. J. Immunol. 167, 1137–1140. Rao, A., Allard, W. J., Hogan, P. G., Rosenson, R. S., and Cantor, H. (1983). Alloreactive T-cell clones. Ly phenotypes predict both function and speciﬁcity for major histocompatibility complex products. Immunogenetics 17, 147–165. Reinherz, E. L., and Schlossman, S. F. (1980). The differentiation and function of human T lymphocytes. Cell 19, 821–827. Roncarolo, M. G., and Levings, M. K. (2000). The role of different subsets of T regulatory cells in controlling autoimmunity. Curr. Opin. Immunol. 12, 676–683. Sakaguchi, S. (2000). Regulatory T cells: Key controllers of immunologic self-tolerance. Cell 101, 455–458. Sakaguchi, S. (2003). Control of immune responses by naturally arising CD4þ regulatory T cells that express toll-like receptors. J. Exp. Med. 197, 397–401. Sakaguchi, S., Fukuma, K., Kuribayashi, K., and Masuda, T. (1985). Organ-speciﬁc autoimmune diseases induced in mice by elimination of T cell subset. I. Evidence for the active participation of T cells in natural self-tolerance; deﬁcit of a T cell subset as a possible cause of autoimmune disease. J. Exp. Med. 161, 72–87. Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., and Toda, M. (1995). Immunologic selftolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J. Immunol. 155, 1151–1164. Salomon, B., Lenschow, D. J., Rhee, L., Ashourian, N., Singh, B., Sharpe, A., and Bluestone, J. A. (2000). B7/CD28 costimulation is essential for the homeostasis of the CD4þCD25þ immunoregulatory T cells that control autoimmune diabetes. Immunity 12, 431–440. Saubermann, L. J., Beck, P., De Jong, Y. P., Pitman, R. S., Ryan, M. S., Kim, H. S., Exley, M., Snapper, S., Balk, S. P., Hagen, S. J., et al. (2000). Activation of natural killer T cells by alphagalactosylceramide in the presence of CD1d provides protection against colitis in mice. Gastroenterology 119, 119–128. Schubert, L. A., Jeffery, E., Zhang, Y., Ramsdell, F., and Ziegler, S. F. (2001). Scurﬁn (FOXP3) acts as a repressor of transcription and regulates T cell activation. J. Biol. Chem. 276, 37672–37679. Seddon, B., and Mason, D. (1999). Regulatory T cells in the control of autoimmunity: The essential role of transforming growth factor beta and interleukin 4 in the prevention of autoimmune

IMMUNOREGULATION AND T CELL SUBSETS

287

thyroiditis in rats by peripheral CD4(þ)CD45RC-cells and CD4(þ)CD8() thymocytes. J. Exp. Med. 189, 279–288. Seder, R. A., and Paul, W. E. (1994). Acquisition of lymphokine-producing phenotype by CD4þ T cells. Annu. Rev. Immunol. 12, 635–673. Sharif, S., Arreaza, G. A., Zucker, P., and Delovitch, T. L. (2002). Regulatory natural killer T cells protect against spontaneous and recurrent type 1 diabetes. Ann. N.Y. Acad. Sci. 958, 77–88. Sharif, S., Arreaza, G. A., Zucker, P., Mi, Q. S., Sondhi, J., Naidenko, O. V., Kronenberg, M., Koezuka, Y., Delovitch, T. L., Gombert, J. M., et al. (2001). Activation of natural killer T cells by alpha-galactosylceramide treatment prevents the onset and recurrence of autoimmune Type 1 diabetes. Nat. Med. 7, 1057–1062. Shawar, S. M., Vyas, J. M., Rodgers, J. R., and Rich, R. R. (1994). Antigen presentation by major histocompatibility complex class I-B molecules. Annu. Rev. Immunol. 12, 839–880. Shevach, E. M. (2000). Regulatory T cells in autoimmmunity. Annu. Rev. Immunol. 18, 423–449. Shevach, E. M. (2002). CD4þ CD25þ suppressor T cells: More questions than answers. Nat. Rev. Immunol. 2, 389–400. Shevach, E. M., McHugh, R. S., Piccirillo, C. A., and Thornton, A. M. (2001). Control of T-cell activation by CD4þ CD25þ suppressor T cells. Immunol. Rev. 182, 58–67. Singh, A. K., Wilson, M. T., Hong, S., Olivares-Villagomez, D., Du, C., Stanic, A. K., Joyce, S., Sriram, S., Koezuka, Y., and Van Kaer, L. (2001). Natural killer T cell activation protects mice against experimental autoimmune encephalomyelitis. J. Exp. Med. 194, 1801–1811. Sonoda, K. H., Exley, M., Snapper, S., Balk, S. P., and Stein-Streilein, J. (1999). CD1-reactive natural killer T cells are required for development of systemic tolerance through an immuneprivileged site. J. Exp. Med. 190, 1215–1226. Sonoda, K. H., Faunce, D. E., Taniguchi, M., Exley, M., Balk, S., and Stein-Streilein, J. (2001). NK T cell-derived IL-10 is essential for the differentiation of antigen-speciﬁc T regulatory cells in systemic tolerance. J. Immunol. 166, 42–50. Stanton, T. H., and Boyse, E. A. (1979). A new serologically deﬁned locus, Qa-1, in the T1a-region of the mouse. Immunogenetics 3, 525–531. Stanton, T. H., Calkins, C. E., Jandinski, J., Schendel, D. J., Stutman, O., Cantor, H., and Boyse, E. A. (1978). The Qa-1 antigenic system. Relation of Qa-1 phenotypes to lymphocyte sets, mitogen responses, and immune functions. J. Exp. Med. 148, 963–973. Swain, S. L., Dialynas, D. P., Fitch, F. W., and English, M. (1984). Monoclonal antibody to L3T4 blocks the function of T cells speciﬁc for class 2 major histocompatibility complex antigens. J. Immunol. 132, 1118–1123. Takahashi, T., Tagami, T., Yamazaki, S., Uede, T., Shimizu, J., Sakaguchi, N., Mak, T. W., and Sakaguchi, S. (2000). Immunologic self-tolerance maintained by CD25(þ)CD4(þ) regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4. J. Exp. Med. 192, 303–310. Thomas, Y., Rogozinski, L., and Chess, L. (1983). Relationship between human T cell functional heterogeneity and human T cell surface molecules. Immunol. Rev. 74, 113–128. Thomas, Y., Rogozinski, L., Irigoyen, O. H., Friedman, S. M., Kung, P. C., Goldstein, G., and Chess, L. (1981). Functional analysis of human T cell subsets deﬁned by monoclonal antibodies. IV. Induction of suppressor cells within the OKT4þ population. J. Exp. Med. 154, 459–467. Thomas, Y., Rogozinski, L., Irigoyen, O. H., Shen, H. H., Talle, M. A., Goldstein, G., and Chess, L. (1982). Functional analysis of human T cell subsets deﬁned by monoclonal antibodies. V. Suppressor cells within the activated OKT4þ population belong to a distinct subset. J. Immunol. 128, 1386–1390.

288

HONG JIANG AND LEONARD CHESS

Thornton, A. M., and Shevach, E. M. (2000). Suppressor effector function of CD4þCD25þ immunoregulatory T cells is antigen nonspeciﬁc. J. Immunol. 164, 183–190. Wilbanks, G. A., and Streilein, J. W. (1990). Distinctive humoral immune responses following anterior chamber and intravenous administration of soluble antigen. Evidence for active suppression of IgG2-secreting B lymphocytes. Immunology 71, 566–572. Wildin, R. S., Ramsdell, F., Peake, J., Faravelli, F., Casanova, J. L., Buist, N., Levy-Lahad, E., Mazzella, M., Goulet, O., Perroni, L., et al. (2001). X-linked neonatal diabetes mellitus, enteropathy, and endocrinopathy syndrome is the human equivalent of mouse scurfy. Nat. Genet. 27, 18–20. Wilson, S. B., Kent, S. C., Patton, K. T., Orban, T., Jackson, R. A., Exley, M., Porcelli, S., Schatz, D. A., Atkinson, M. A., Balk, S. P., et al. (1998). Extreme Th1 bias of invariant Valpha24JalphaQ T cells in type 1 diabetes. Nature 391, 177–181. Wu, J., Wilson, J., He, J., Xiang, L., Schur, P. H., and Mountz, J. D. (1996). Fas ligand mutation in a patient with systemic lupus erythematosus and lymphoproliferative disease. J. Clin. Invest. 98, 1107–1113. Zamvil, S., Nelson, P., Trotter, J., Mitchell, D., Knobler, R., Fritz, R., and Steinman, L. (1985). T cell clones speciﬁc for myelin basic protein induce chronic relapsing paralysis and demyelination. Nature 317, 355–358. Zhai, Y., and Kupiec-Weglinski, J. W. (1999). What is the role of regulatory T cells in transplantation tolerance? Curr. Opin. Immunol. 11, 497–503.

INDEX

A

Attrition, affect of memory T cells, 220 Autoantibodies, 256 Autoimmune diseases cause of, 255 cytokines and, 256 IL-18 and, 146 initiation of, 279f, 280 NK T cell function on, 265–267 types of, 255 Autoimmune pathologies, 133 Autoimmune responses, CD8þ T cell controlled, 270 Autoimmunity, 255 CD8þ T cells and, 268–269 characteristics of, 256 control of, 256 NK T cells and, 266 Autoreactive cells, 253 controlling outgrowth of, 255

Activated T cells Bcl-3 and survival of, 211 death of, 208 ROS and, 208–209 Alloantigens, 257–258 Alloantisera, Qa-1, 259–260 Anergy, 256 Antibodies, 145–146 anti-CD3, 103–104 anti-TCR, 103 memory B cells producing, 191 monoclonal, 258 Antibody-dependent cellular type cytotoxicity (ADCC), 134 Antigens activation of immunocompetent cells, 277 doses of, 269 effector function and, 249 epithelial, 177 foreign, 194, 257, 276–277 HEL, 269 self, 255 type 2 immune responses and, 177, 178f in vivo loss of, 237 Antigen-specific CD8þ T cells expansion/decline of, 199f, 204 gene expression pattern of, 207 Apoptosis antigen induced, 277 of DP thymocytes, 94–95 granzyme A triggered, 241 granzyme B induced, 240 target cell, 238 TNF-a induced, 242 Apoptotic mechanisms, 255 regulation of self-reactive cells and, 256

B B cells CLPs as progenitors for, 16 differentiation, 15, 257, 261 IL-4 expression by, 176 lymophocyte precursors differentiation into, 92 Pax-5 and commitment of, 27–28 Basophiles, IL-4 expression in, 176 Bcl-3, 211 Bi-potential model, origin of memory CD8þ T cells and, 205, 205f

C

CD4þ cells, B cell differentiation inducer function of, 261 CD4 genes c-Myb/SAF binding in silencing of, 66f, 69 289

290

INDEX

CD4 genes (continued) enhancer regulated expression of, 67–68 HEB/E2A heterodimeric complex in activation of, 68 trans-acting proteins and regulation of, 67–71 transcription factors for negative regulation of, 69–71 CD4 intronic enhancer, 63 CD4 promoter (P4), 59–60 expression in DP thymocytes and, 62 CD4 proximal enhancer nuclear protein-binding sites of, 60 in transgenic reporter genes expression, 61, 61f CD4þ regulatory cells, 55 CD4 silencer, 62–63 in CD4 silencing, 73 functional sites in, 64–65, 66f mammalian SWI/SNF chromatin remodeling complex in, 71 mechanisms of, 71–73, 75–76, a98f mutations in, 65, 66f, 67 temporal activation of, 64 trans-acting proteins acting on, 68–71 CD4 SP cells, 56 process of, 57–58 CD4þ suppressor cells biological foundation of, 262 expressing CD25/CTLA-4, 262 Foxp3 and, 264 CD4þ T cells, 193 antigen-reactive, 273 as autoreactive clones, 255 CD8þ T cells downregulation of, 274 CD40 direct ligation on, 210–211 CD40/CD40L signals increase survival of, 210 cytokine secretion of, 244 encephalitogenic, 272 inducing Qa-1 restricted regulatory CD8þ T cells, 281 LAG-3 expressed on, 59 Lck-deficient thymocytes differentiation into, 102 MHC class II restricted, 258 outgrowth of, 255 Qa-1 restricted regulatory CD8þ T cells interaction with target, 270, 271f Qa-1/self-peptide and, 273–274

regulatory, 253–254 response patterns of, 199f, 212 response patterns of antigen-specific, 212 SEB activated, 270 self-reactive, 280 suppressed by CTLA-4, 263 suppression by, 261 surface Qa-1 molecules on, 274 TCR employed by, 264 Th1 development in, 137 TH1/TH2 subsets of, 256 CD4þ Th cells, 55 subsets of, 133 CD4þ8 intermediate thymocytes, 113–115, 114f coreceptor reversal as function of, 116 CD4/CD8 coreceptor molecules, 99 CD4þCD8 T cells. See CD4 SP cells CD4CD8þ T cells. See CD8 SP cells CD4þCD25þ cells Foxp3 expressed in, 264 self/nonself distinguished by, 264–265 TCRs expressed by, 279 CD4þCD25þ regulatory cells characteristics of, 278t immunocompetent cells suppressed by, 277 self-reactive clones influenced by, 280 CD4þCD25þ T cells, immunosuppresive cytokines expressed by, 280 CD8þ cytotoxic cells, 55 CD8þ effector cells, 259 CD8þ effector T cells acquiring gene expression profiles, 207 cell division into, 200 as defined, 234 identifying, 234 memory CD8þ T cells v., 194–195 CD8 genes cis-regulatory element and regulation of, 77, 78f, 79–84 DP thymocytes terminate, 117f expression regulation of, 77, 78f, 79–86 genomic fractions in regulation of, 78f, 79 regulation of, 83–84 trans-acting factors in, 83–84 CD8þ regulatory cells, 253–254 recognition of Qa-1/peptide complex, 261 T cells regulated by, 273

INDEX

CD8þ regulatory T cells acting as effector suppressor cells, 279–280 autoimmunity resistance mediated by, 280 biological function of, 270 characteristics of Qa-1, 278t features of Qa-1 restricted, 275–277, 276f importance of, 268–269 peripheral regulation by Qa-1 restricted, 275, 276f prior to antigen encounter, 279 CD8 SP cells, 56 CD4 silencing in, 71–72 enhancer directed expression in, 80 process of, 57–58 CD8þ suppressor T cells autoimmunity and, 268–269 EAE model and, 268 as induced during immune response, 265 role in regulating immune responses, 268 CD8þ T cell phenotypic subsets central memory, 247 effector, 248–249 effector memory, 248 naive, 245 CD8þ T cells APC initiated response of, 201 autoimmune responses controlled by, 270 bcl-2 and survival of, 211 CD4 silencing in, 73 CD4þ T cells downregulation by, 274 cell death during response of, 209 costimulatory regulation of responses of, 202 cytokine secretion of, 244 DC and expansion of, 201–202 differential receptor expression, 218 differentiation of, 198–99, 199f, 245, 246f, 247–249 EAE mice and, 271–272, 272f enhancer expression in, 81–82 expansion of, 199, 199f expansion/contraction/memory pattern of, 200 generating memory CD8þ T cells, 206–207 IL-2 and expansion differentiation of, 204 IL-7 role in differentiation of, 117–118 IL-15 and function survival of, 204 ILs supporting responses of, 203 immune exhaustion and, 209 immune response to foreign/self antigens governed by, 269

291

LAG-3 expressed on, 59 memory development of, 208 MHC class I restricted, 258 naive, 217–218 peripheral MBP-reactive CD4þ TCR repertoire regulated by, 270–272, 272f priming during EAE of, 275 response kinetics of, 200–201 Runx 3 and TCR-mediated response of, 76 selective downregulation of self-reactive T cells by, 272–273 suppression by, 260–261 surface markers of, 192t, 198, 199f viral clearance and, 197 CD8a expression of, 58 induction of, 84 CD28, 256 CD40 CD40L reaction with, 256 direct ligation of CD48þ T cell, 210–211 survival of activated CD8þ T cells regulated by, 210 CD45 molecule, TCR signaling on T cells regulated by, 263 CD48, 256 CD80, 256 CDL. See Cytotoxic lymphocytes Cells. See also specific cells activation-induced death of, 255 central memory, 196 death/injury of, 279f, 280 distinguishing self from nonself, 264–265 division of, 200 experiments of transitional, 106–108, 107f multiple lineage differentiation of, 21 Qa-1 restricted CD8þ T cells, 277 regulatory, 261 self-reactive, 256 suppressor, 260 suppressor T, 257 Cellular immunity, 191 Central thymic selection, 253 CFU. See Colony-forming units Chromatin GATA-4 in, 164 remodeling of, 18–19, 83 Clinical autoimmunity, 253

292

INDEX

Clonal selection, 254–255 Clones, 254–255 autoreactive, 255 downregulation of antigen-activated, 273 possessing intermediate affinity/avidity to self, 281 recognition of, 273 self-reactive, 255–257, 271, 272f, 276, 280 T cell, 270 CLPs. See Common lymphoid progenitors CMPs. See Common myeloid progenitors Colony-forming units (CFU), 3–4 Common lymphoid progenitors (CLPs), 1, 2f, 9f activity of, 13 adult lymphoid differentiation and, 13, 15–17 CMPs as counterpart of, 17–18 fetal, 35–37, 37f IL-7R and, 8, 8f as IL-7Ra-expressing cells, 10–11 isolation of, 7, 8f, 13 lymphoid lineage priming in, 20f, 28 Notch 1 distribution in, 26 Pax-5 in, 28 pTa transcription initiated by, 12 Common myeloid progenitors (CMPs), 10 as CLPs counterpart, 17–18 myeloerythroid gene priming in, 19, 20f, 21 Coreceptor genes, 58 Coreceptor molecules, 91. See also CD4/CD8 coreceptor molecules auto-immune diseases and, 146 CD4þ T cell/CD8þ T cell differentiation and, 103 on DP thymocytes, 95–99 IL-12 synergy between, 143 intracellular Lck binding to, 96–97 IRAK pathway signaling of, 142 lineage choice of, 100–101 protective role of, 145 regulation/expression of genes encoding, 55–56 structural difference between, 95–96 T cell development role of, 56–58 T cell differentiation development stages and, 92, 93f as TCR coreceptors, 95 in Th1 responses, 144–145

tissue specificity of, 142–143 Coreceptor reversal model. See Kinetic signaling model; Strength of signal instructional model Costimulatory molecules, regulating amplitude of CD8þ T cell response by, 202 CPS, 264–265 CTL. See Cytotoxic T lymphocytes CTLA-4 molecule, 256 CD4þ T cell suppression regulated by, 263 Cytokine secretory effector cells, 243–245 Cytokine secretory effector function memory cells with, 233 Cytokines, 256, 260 autoimmune diseases and, 256–257 autoimmunity and, 266 immunosuppresive, 280 memory CD8þ T cells production of, 194 memory T cell maintenance and, 214–215 new, 149, 151 NK T cells producing, 265–266 production of proinflammatory, 140–141 regulation of CD8þ T response, 202 role in suppression by, 263 secretion of, 235–236, 243–244 signals, 29–30 Th cell produced, 134 Th2-specific, 143 Cytotoxic effector cells degranulation of, 241–242 differentiation of, 245, 246f Cytotoxic effector function CD45RA expression and, 234 FasL/Fas death pathway and, 235–236, 242 measuring, 238 memory cells with, 233 perforin-dependent granule exocytes pathway and, 235–236, 238–239 Cytotoxic lymphocytes (CDL) CD8 molecules and, 258–259 function of, 259 Rab 27a required by, 239 Cytotoxic T cells CD4 silencing in, 72–73 TCRab, 77 Cytotoxic T lymphocytes (CTL), 236, 257. See also Effector CTLs extrinsic signals and, 209–210 memory, 58, 84

293

INDEX

memory CD8þ T cells and, 206 recognize/kill infected cells function of, 209 TcR-triggered granule exocytosis in, 239 Cytotoxicity Antibody-dependent cellular type, 134 measuring, 236 NK T cells and, 265 perforin-dependent, 237 in vitro, 238 in vivo, 236–238

D DCs. See Dendritic cells Delayed type hypersensitivity (DTH), 257 Dendritic cells (DCs) CD8þ T cells expansion and, 201 CD8a expression on, 82 CD40/CD40L and, 210 cytokine production by, 148f early lymphoid progenitors differentiate into, 33–35 IL-12 production by, 139–140 initiating T cell responses, 201–202 markers of, 35 maturation of, 140 T cell polarization and, 172t–174t, 177 Th cells development and, 140 Th1 cells development and, 136, 138 DH. See DNase hypersensitivity DN cells, 56 DN thymocytes. See Double negative thymocytes DNase hypersensitivity (DH) cluster II deletion of, 81 clusters of, 79, 85 sites of, 77, 78f, 79–80 Double negative (DN) thymocytes CD4 repressed in, 72, 74f CD4 silencing in, 71–72 development of, 92–93, 93f negative regulatory element in, 64 proliferative phases of, 93–94 selection of, 56–57 subsets of, 92, 93f Double positive (DP) thymocytes, 92 Bcl-2/IL-7R upregulated by, 117 CD4 expression in, 62 CD4/CD8 coreceptor molecules expressed by, 99

CD4-dependent TCR/ligand interactions signal to, 113–114, 114f CD4SP/CD8SP T cells differentiation from, 104, 111–112, 111f, 115 CD8-dependent TCR/ligand interactions signal to, 112, 113f CD8-negative, 81 cell fate determination of, 76 coreceptor function in, 97 development of, 94–95 differentiation into CD4þ T cells of, 102–103 enhancer directed expression in, 80–82 intracellular kinase activity in, 102 intracellular pools of Lck in, 96 Lck depletion and, 97–98 Lck signaling and signaled by, 120 as lineage-committed cells, 100 reporter transgene expression in, 61, 61f SP T cell conversion of, 99–100 stochastic/selection model and, 105–108 TCR complexes on, 98 TCR signaling of, 106 TCR signals generated in, 101 DP thymocytes. See Double positive thymocytes DTH. See Delayed type hypersensitivity

E EAE, 263–264, 268 cellular events during, 275 IFN-y production in, 146 induction of, 145 NK T cells and, 266 TCV induced protection from, 270 Early lymphoid progenitors differentiation into DCs of, 33–35 pTa proteins defining, 12 Early T lineage progenitors (ETPs) myeloid potential of, 16 T cell production by, 15–16 Early thymic progenitors, c-Kitþ, 15–16 Effector cells, 253. See also Effector function as defined, 233–235, 236, 249 Effector CTLs memory CD8þ T cells by, 207–208, 211–212 molecular mechanism change of CD8þ T cells to, 208 Effector function. See also Cytokine secretory effector function; Cytotoxic effector function

294

INDEX

Effector function (continued) secretion as mechanism of, 235–236 as subpopulation, 234 Effector memory CD4þ T cells, Listeria and, 135 Effector memory CD8þ T cells clear pathogens upon reinfection, 197–198 differentiate into central memory CD8þ T cells, 197 Listeria and, 135 protective immunity and, 197 Effector molecules, 191 Eosinophils, IL-4 storage/expression by, 176 ETPs. See Early T lineage progenitors Extrinsic signals, 209 CD40, 210

F Fas ligand, 255 FasL/Fas death pathway, 235, 242 Fetal hematopoietic progenitors, as not fully committed to lymphoid/myeloidfates, 35–38, 37f Flk-2, in early stem and progenitor cells, 9–10 Foreign antigens, 257 clones and, 276–277 immune tolerance to, 257 memory T cell recognition/elimination of, 194 Foxp3 CD4þCD25þ regulation and, 264 defects in, 280 mutational defects in, 264

G GATA-1 in hematolyphoid commitment, 22, 24–25 lineage instructive effect of, 31 GATA-3, 146 binding sites of, 164–165 T cell activation synapse signals and, 165 Th1 development down-regulated by, 148 Th2 cells expression of, 147–148 Th2 differentiation role of, 164 GATA-4, in decompacting condensed chromatin, 164 Genes activation/silencing of batteries of, 55 priming of, 19 GMPs, GM-related genes expressed by, 19

Granzymes, 239 A, 241 B, 240–241

H Helminth immunity, Th2 cells and, 133 Hematopoietic stem cells (HSCs), 2–3 commitment models of, 4–5, 5f commitment sequence to lymphoid cells from, 1, 2f fetal, 35 gene expression profiles and, 32–33 hematolymphoid lineages commitment of, 3–5, 5f Hetero-trimeric receptors, 203–204 Homeostatic regulation, 256–257 HSCs. See Hematopoietic stem cells Human bone marrow, lymphoid- and myeloid-restricted progenitors in, 38–39 Humoral immunity reinfection protection and, 191

I IEL. See Intestinal intraepithelial lymophocytes IFN. See Interferons IFN-g IL-12 signaling and production of, 138 IL-18/IL-12 induced production of, 142 macrophage effector function stimulated by, 144 regulating elements in, 146 responses of, 134 role of, 163 secretion of, 237 T cell produced, 141 TCR induced production of, 143 Th1 cell produced, 134, 256 IFN-g receptor (IFN-gR), 152 IFN-g-inducing factor (IGIF), 141 IFN-gR. See IFN-g receptor IgG, antibody synthesis, 257 IGIF. See IFN-g-inducing factor IL. See Interleukins IL-2 controlling lymphocyte homeostasis, 204 as effector cytokine, 244 IL-4, 256 in coreceptor reversal, 117–118 expression activation of, 166

INDEX

mutations affecting expression of, 171 non-Th2 cell expression of, 171, 175–176 T cell expression of, 167–170 T cell production of, 168 IL-7 ab T cell differentiation role of, 10 memory CD8þ T cells support by, 216, 216f naive T cells and, 203 as survival factor for memory CD8þ T cells, 215 IL-7 receptor (IL-7R) DP thymocyte upregulation of, 117 PU.1 regulated, 23 role of signals of, 118 T/B cell development and, 8, 8f T/B cell expression of, 10 IL-7R. See IL-7 receptor IL-10, 256 cell production of, 140–141 IL-12 DC stimulation by, 139–140 G protein coupled signaling via CCR5 and, 139 IL-18 synergy between, 143 p40, 144–145 as secreted, 136 signaling, 138 Th1 cells development and, 136, 141–142 IL-15 maintaining memory of memory CD8þ T cells by, 217 memory CD8þ T cells support by, 216–217, 216f protective immunity to intracellular pathogens and, 219 regulating CD8þ T cells function/ survival, 204 STATS and memory CD8þ T cells response to, 218 as survival/proliferation factor for antigen-specific memory CD8þ T cells, 215–216 as survival/proliferation factor for memory CD8þ T cells, 215 IL-18 Autoimmune diseases and, 146 IL-12 synergy between, 143 in Th1 cells responses of, 144–145 IL-23, 145, 151

295

Immune memory, 220–221 Immune responses, 208 cytotixicity pathways role in, 237 to foreign pathogens, 253 NK T cells and tumor rejection, 265 NK T cells primary, 277 regulation of, 254 to SEB, 269 suppressor cells regulating, 261 suppressor T cells control of, 257 of T cells, 199f triggers of, 256 Immune system, 133, 255 suppressor mechanisms of, 255 Immunity cellular, 191 helminth, 133 humoral, 191 peripheral regulation of, 280 protective, 197–198 suppressor T cells regulation of, 257 Immunocompetent cells, 268 antigen activation of, 277 autoimmune disease and, 255 NK T cells as, 267 Immunodominance, 206 Immuno-dysregulation, polyendocrinopathy, enteropathy, X linked syndrome (IPEX), 264 Immunological memory, 212 Immunopathology development of, 143 prevention of, 148f Th subsets contribution to, 133 Immunoregulation, 269–270 helper/suppressor functions of, 258 NK T cells and, 267 paradigms of, 253–254 by regulatory T cell subsets, 277 Immunosuppresive cytokines, Th3 secreting, 256 Interferons (IFN), 202–203 type I, 137–138 Interleukins (IL), 203. See also IL-2; IL-4; IL-7; IL-10; IL-12; IL-15; IL-18; IL-23 Intestinal intraepithelial lymophocytes (IEL), 82 CD8þ, 55 Intracellular pathogens, 144 Th1 caused death of, 134

296

INDEX

IPEX. See Immuno-dysregulation, polyendocrinopathy, enteropathy, X linked syndrome

Lymphokines, structural characterization of, 261 Lyt2, 257 SRBC and, 258

K Kinetic signaling model, 111f. See also Strength of signal instructional model analysis of, 112–115, 113f, 114f coreceptor reversal in, 115–119 lineage commitment of DP thymocytes in, 110–111 MHC-I-signaled thymocytes differentiation into CD8SPT cells in, 117f signal intensity/duration distinctions of, 119–121

L Ligand, engagement, 204 Lineage commitment models classical, 100–101 Linear differentiation model origin of memory CD8þ T cells and, 205, 205f predictions of, 206–207 Listeria monocytogenes (Lysteria), controlling primary immune responses to, 135 LKLF, 221 Lymphadenopathy, 256 Lymphocytes development of, 6 differentiation of T/B, 6–8 IL-4 expression of, 166, 166f manipulating, 235 markers of, 13 memory, 192 precursors, 91 promiscuity of, 19, 20f, 21–22 regulatory T, 253 secretory granules of, 235 Lymphoid progenitors as differentiate into T cells, 56 lineage elasticity in, 28–32 lineage promiscuity in, 20f, 21 Lymphoid-committed progenitors instructive cytokine signaling myelomonocytic conversion of, 29–30 megakaryocyte/erythrocyte conversion from, 30–32

M Macrophages cytokine production by, 148f differentiation of, 15 effector functions of, 144 fetal CLPs and, 36–38, 38f TNF induction by, 145 Major Histocompatibility Complex (MHC), 99, 258–259, 264 CD4/CD8 coreceptor binding to, 95 class I, 91 class II, 91 class specificity of, 101 positive/negative selection by, 6 role in memory T cell homeostasis, 213–214 TCR signals in DP thymocytes stimulated by, 94 Mast cells, IL-4 expression in, 176 MBP, immunization with, 270 MeCP1 complexes, 164 Megakaryocyte-restricted progenitors (MKPs), 18 Memory B cells, antibodies produced by, 191 Memory CD4þ T cells functional plasticity and, 212 maintenance of, 219 Memory CD8þ T cells, 191 in absence of IL-15, 216 acquisition kinetics of, 207 attrition of, 220 CD8þ effector T cells v., 194–195 CD8þ T cells generating, 206–207 cytokine production of, 194 differentiation of, 197 effector CTLs create, 207–208, 211–212 effector CTLs relation to, 205, 205f expansion of, 194 frequency of, 193 heterologous stimuli and, 195–196 HKLM-induced, 197–198 homeostasis of, 217 IL-7 as survival factor for, 215 IL-15 and maintaining memory of, 217 IL-15 survival/proliferation factor for, 215 immunodominance and, 206

297

INDEX

LCMV infection of, 196 LKLF and survival of, 221 magnitude of pool/size of CTL response of, 206 maintenance of, 218 memory pool of, 198 naive CD8þ T cells/effector CTLs relationship to, 198 protective immunity to intracellular pathogens of, 219 recruited to peripheral tissues, 195–196 response to IL-15 and STAT5, 218 role of subsets of, 196–197 secondary response mediated by, 198 support by IL-7/IL-15, 216–217, 216f Memory T cells antigen specific, 196 cellular immunity of, 191 foreign antigen recognition/elimination by, 194 induced by organism/antigen, 193 induced/maintained, 191 negative regulation of, 219–220 persistent antigen and function of, 214 protective immunity mediated by, 219 surface markers on, 192, 192t, 194 tonic MHC/TCR signals and survival of, 213 Memory-phenotype CD8þ T cells, 193 Mendelian susceptibility to mycobacterial disease (MSMD) gene mutations and, 151–152 MHC. See Major Histocompatibility Complex MHC I CD8 SP cells with cytotoxic function and, 57 TCR/ligand interactions specific to, 115 MHC II CD4 SP cells with helper functions and, 57 TCR/ligand interactions specific to, 115 MKPs. See Megakaryocyte-restricted progenitors Molecular immunology, 259–260 Molecules, 262 antigen activation-induced cell surface, 256 antigen-specific suppressor, 260 CD45, 263 costimulatory, 202 CTLA-4, 263 cytoplasmic tails of, 258 effector, 191 glycolipid, 277

MHC, 258–259, 264 types of, 255 Monoclonal antibodies, 258 MSMD. See Mendelian susceptibility to mycobacterial disease

N Naive CD4 T cells IL-4 expression activation in, 164–167 IL-4 genes in, 165 Natural killer (NK) cells, 1 NK cells. See Natural killer cells NK T cells activators for, 265–266 autoimmunity controlled by, 266 characteristics of, 278t diabetes and, 266–267 function on autoimmune disease, 265–267 IL-4 expression in, 175 immunoregulation and, 267 as primary immune response, 277 with restricted TCRs, 55, 253–254 self-reactive clones influence by, 280 Notch 1, in T cell commitment, 25–26 Notch intracellular domain (Notch-IC), gene transcription activated by, 92 Notch molecule cell fate determined by, 91 lymphocyte precursors differentiation into T cells by, 92 Notch-IC. See Notch intracellular domain

P P4. See CD4 promoter Pathogens, 141 foreign, 253 immunity to, 191 intracellular, 134, 152 Pax-5, in B cell commitment, 27–28 Perforin, 239–240 Perforin-dependent granule exocytosis pathway cytotoxic activity via, 244 perforin/granzymes and, 239 process of, 238 secretion of, 235 Peripheral immunity regulation of, 253, 254f regulatory mechanisms controlling, 281

298

INDEX

Peripheral regulation, 253 of immune responses, 261 Physiologic immune regulation, 253 Progenitor cells, differentiation of, 55 PU.1, in hematolyphoid commitment, 22–25

Q Qa-1, 273 Qa-1 restricted CD8þ T cells biological consequences of down regulation of, 277 CD4þ T cells inducing, 281 Qa-1/peptide complex, 273–274 CD8þ regulatory cells recognition of, 261 TCRs expressing, 273

R Rab 27a molecule, mutations in, 239 RAG. See Recombination Activating Genes RAG proteins DNA cleavage by, 6 thymocyte proliferative phase expression of, 93 RAG-1, ELP defined by transcriptional activation of, 11–12 Reactive oxygen species (ROS), 208 Receptor ligand interactions, 256 Recombination Activating Genes (RAG), 92. See also RAG-1 DNA cleavage by, 6 function of, 11 Regulatory mechanisms MHC/peptide triggering CD4þ cells cause of, 256 sets of, 256–257 TCR/MHC/peptide interaction initiating, 256 Regulatory T cells as component of immune system, 265 subsets of, 277 ROS. See Reactive oxygen species Runx genes CD4 silencing and, 70–71 Runx 3, 76 silencing roles of, 75–76 TCR signaling in, 76–77

S SEB, immune response to, 269 Self peptide/MHC complexes, T cell reaction to, 253

Self-reactive cells regulation of, 256 suppression of, 273 Self-reactive clones, 255 CD4þ T cells and, 276 death of, 256 limiting outgrowth of, 271–272, 272f outgrowth/functions of, 256–257 Self-reactive T cells CD8þ T cells selective downregulation of, 272–273 clonal growth of, 281 express ligands, 265 Signal instructional model DP thymocytes and, 102 strength of, 101–104 Single positive (SP) cells. See CD4 SP cells; CD8 SP cells SP cells. See Single positive cells Splenomegaly, 256 Stat6, Th2 effector development role of, 168 Stochastic/selection model contradictions to, 107f, 108–110 support for, 105–108 Strength of signal instructional model, 58. See also Kinetic signaling model contradictions to, 104–105 signal intensity/duration distinction of, 119–121 Superantigens, 269 Suppressor cells, 260 immune responses regulated by, 261 Suppressor T cells, control of immune responses and, 257

T T cell receptors (TCR), 253, 264 activated transgenic, 136 antigen activity of, 260 CD4þ8 signaling of, 113, 113f CD8 gene expression and, 111–112 cell death and, 255 downregulation and, 272–273 DP thymocytes and, 94, 97–98 DP thymocytes expressing, 99 expressed by CD4þCD25þ, 279 expressing Qa-1/peptide, 273 heterogeneity to pathogens of, 133 IL-7R desensitized by signaling of, 118 of intermediate affinity, 56–57

INDEX

intermediate thymocytes signaling of, 112 memory CD8þ T cells and, 194 role in memory phenotype T cell maintenance, 214 signal intensity/duration of, 120–121 stochastic/selection model and, 106–108 in Th2 development, 167 transgenic cells and, 200 T cell responses DCs initiating, 201–202 increase of T cells due to, 219–220 T cell vaccination (TCV), CD8þ T cell hybrids and, 270 T cells. See Tumor cells T helper cells (Th cells) DC subsets on development of, 140–141 disease outcome and, 134–135 IL-12 and, 141 protein antigens and induction of, 135–136 specific cytokines, 134 subsets of, 133 T-bet, 147 Tc1 cells, 244–245 Tc2 cells, 244–245 expression of IL-4 in, 171, 175 TCR. See T cell Receptors TCV. See T cell vaccination TGFb, 256 Th cells. See T helper cells Th1 cells, 256. See also T helper cells autoimmune pathologies and, 133 cell-mediated immunity role of, 133 commitment, 137 DC and development of, 136, 138 development of, 148f, 150f downregulatory influences on, 256–257 IFN-y secreted by, 134, 256 IL-4 location in, 169 IL-10 production by, 140 IL-12 driven development, 136–137 IL-18 in responses of, 144–145 impaired responses of, 149, 151 intracellular pathogen protection by, 152–153 proinflammatory cytokines produced by, 134 as subset of CD4þ Th cells, 133 systematic invasion response of, 163 T-bet and, 147 Th1 clones, DTH responses induced by, 134 Th2 cells, 256. See also T helper cells

299

development molecular mechanisms of, 136 development of, 167 differentiation and function of, 256 GATA-3 in differentiation of, 164 helminth immunity and, 133–134 IL-4 expression plasticity in differentiated, 169 IL-4-expressing, 163 IL-4/IL-4R/Stat6-independent differentiation of, 167 IL-12 signaling and, 138 interleukins produced by, 133 mutations affecting IL-4 expression from, 171 naive T cell polarization by, 165, 166f phenotype/genotype analysis of, 170 production of Ige/IgG1 antibodies, 134 responses of, 144 as subset of CD4þ Th cells, 133 terminal differentiation of, 177–178 type 2 immunity orchestrated by, 163, 179 as unresponsive to IL-12, 136 Thymic negative selection, T cell elimination through, 253, 255 Thymocytes apoptosis of, 253 CD4-CD8-DN, 63–64, 72 CD8 lineage, 80 CD8 single positive, 63–64 into CD8þ T cell lineage, 103 CD8SP T cells differentiation from, 117f differentiation of, 55, 92 kinetic signaling model and, 110–115, 111f, 113f, 114f lineage commitment development in, 101–110 Runx sites and, 70 TCR transgenic, 101–102 TLR. See Toll like receptors TLR9 expression of, 139 TLRLs. See Toll like receptor ligands TNF, Listeria clearance by, 135 TNF-a apoptosis induced by, 242 secretion of, 237 Toll like receptor ligands (TLRLs), 195, 201 cytokines induced by, 215

300 Toll like receptors (TLR), pathogen-derived products recognition by, 139 Transcription factors Notch1 as, 25–26 Pax-5, 27–28 PU.1/GATA/1 as, 22–25 Transgenes CD4, 63 lineage-specific expression of, 63 Tumor cells (T cells) activation regulation of, 264 autoimmune disease and, 253 autoreactive, 281 CD4þ/CD8þ interaction, 260 CD8abþ, 84 DC interaction with, 201–202 DC40L expressed by, 256 death of, 199, 208 differentiation, 15–17, 258 division of, 204 enhancer directed expression in, 80 functional fate of, 256 growth and proliferation of, 262–263 Hybrids, 270 IL-4 production by, 168 IL-7 and, 203 immune response of, 199f Lyt2, 257–258 mediated immunity, 191

INDEX

memory, 191 memory-phenotype, 192–193 mutations impacting IL-4 expression in, 171 polyclonal stimulation of, 136 pTa gene as Notch target in, 26 recruitment of antigen-specific, 200 as regulated by CD8þ regulatory cells, 273 response to foreign antigens of, 254 Runx function in, 76–77 scurfy mice/IPEX patients defects in activation of, 264 self-reactive, 261, 265, 272–273, 281 stabilization of IL-4 expression in, 167–170 suppressor, 257 survival of, 202–203 thymic negative selection elimination of, 253 Type 2 immunity, 163 initiation/expression of, 177, 178f, 179 mutations impacting, 172t–174t Type I IFNs STAT4 and, 137 in Th1 development, 137–138

V Vaccines, 260 humoral immunity and, 191 subsets of memory CD8þ T cells and, 198 T cell based, 199

CONTENTS OF RECENT VOLUMES

Volume 74

Ju¨rgen Hess, Ulrich Schaible, Ba¨rbel Raupach, and Stefan H. E. Kaufmann

Biochemical Basis of Antigen-Specific Suppressor T Cell Factors: Controversies and Possible Answers Kimishice Isihzaka, Yasuyuki Ishii, Tatsumi Nakano, and Katsuji Sugik

The Cytoskeleton in Lymphocyte Signaling A. Bauch, F. W. Alt, G. R. Crabtree, and S. B. Snapper

The Role of Complement in B Cell Activation and Tolerance Michael C. Carroll

TGF- Signaling by Smad Proteins Kohei Miyazono, Peter ten Dijke, and Carl-Henrik Heldin

Receptor Editing in B Cells David Nemazee

MHC Class II-Restricted Antigen Processing and Presentation Jean Pieters

Chemokines and Their Receptors in Lymphocyte Traffic and HIV Infection Pius Loetscher, Bernhard Moser, and Marco Bacciolini

T-Cell Receptor Crossreactivity and Autoimmune Disease Harvey Cantor

Escape of Human Solid Tumors from T-Cell Recognition: Molecular Mechanisms and Functional Significance Francesco M. Marincola, Elizabeth M. Jaffee, Daniel J. Hicklin, and Soldano Ferrone

Strategies for Immunotherapy of Cancer Cornelis J. M. Meliey, Rene E. M. Toes, Jan Paul Medema, Sjoerd H. van der Burg, Ferry Ossendorp, and Rienk Offringa

The Host Response to Leishmania Infection Werner Solbacii and Tamas Laskay

Tyrosine Kinase Activation in the Decision between Growth, Differentiation, and Death Responses Initiated from the B Cell Antigen Receptor Robert C. Hsueh and Richard H. Scheuermann

Index

Volume 75

The 30 IgH Regulatory Region: A Complex Structure in a Search for a Function

Exploiting the Immune System: Toward New Vaccines against Intracellular Bacteria 301

302

CONTENTS OF RECENT VOLUMES

Ahmed Amine Khamlichi, Eric Pinaud, Catherine Decourt, Christine Chauveau, and Michel Cogne´

Human Basophils: Mediator Release and Cytokine Production John T. Schroeder, Donald W. MacGlashan, Jr., and Lawrence M. Lichtenstein

Index

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

Volume 77 The Actin Cytoskeleton, Membrane Lipid Microdomains, and T Cell Signal Transduction S. Celeste Posey Morley and Barbara E. Bierer Raft Membrane Domains and Immunoreceptor Functions Thomas Harder

Btk and BLNK in B Cell Development Satoshi Tsukada, Yoshihiro Baba, and Dai Watanabe Diversity and Regulatory Functions of Mammalian Secretory Phospholipase A2s Makoto Murakami and Ichiro Kudo The Antiviral Activity of Antibodies in Vitro and in Vivo Paul W. H. I. Parren and Dennis R. Burton Mouse Models of Allergic Airway Disease Clare M. Lloyd, Jose-Angel Gonzalo, Anthony J. Coyle, and Jose-Carlos Gutierrez-Ramos Selected Comparison of Immune and Nervous System Development Jerold Chun Index

Volume 78 Toll-like Receptors and Innate Immunity Shizuo Akira Chemokines in Immunity Osamu Yoshie, Toshio Imai, and Hisayuki Nomiyama Attractions and Migrations of Lymphoid Cells in the Organization of Humoral Immune Responses Christoph Schaniel, Antonius G. Rolink, and Fritz Melchers Factors and Forces Controlling V(D)J Recombination

CONTENTS OF RECENT VOLUMES

David G. T. Hesslein and David G. Schatz T Cell Effector Subsets: Extending the Th1/Th2 Paradigm Tatyana Chtanova and Charles R. Mackay MHC-Restricted T Cell Responses against Posttranslationally Modified Peptide Antigens Ingelise Bjerring Kastrup, Mads Hald Andersen, Tim Elliot, and John S. Haurum Gastrointestinal Eosinophils in Health and Disease Marc E. Rothenberg, Anil Mishra, Eric B. Brandt, and Simon P. Hogan Index

Volume 79 Neutralizing Antiviral Antibody Responses Rolf M. Zinkernagel, Alain Lamarre, Adrian Ciurea, Lukas Hunziker, Adrian F. Ochsenbein, Kathy D. McCoy, Thomas Fehr, Martin F. Bachmann, Ulrich Kalinke, and Hans Hengartner Regulation of Interleukin-12 Production in Antigen-Presenting Cells Xiaojing Ma and Giorgio Trinchieri Mechanisms of Signaling by the Hematopoietic-Specific Adaptor Proteins, SLP-76 and LAT and Their B Cell Counterpart, BLNK/SLP-65 Deborah Yablonski and Arthur Weiss Xenotransplantation David H. Sachs, Megan Sykes, Simon C. Robson, and David K. C. Cooper Regulation of Antibacterial and Antifungal Innate Immunity in Fruitflies and Humans Michael J. Williams

303

Functional Heavy-Chain Antibodies in Camelidae Viet Khong Nguyen, Aline Desmyter, and Serge Muyldermans Uterine Natural Killer Cells in the Pregnant Uterus Chau-Ching Liu and John Ding-E Young Index

Volume 80 Protein Degradation and the Generation of MHC Class I-Presented Peptides Kenneth L. Rock, Ian A. York, Tomo Saric, and Alfred L. Goldberg Proteoanalysis and Antigen Presentation by MHC Class II Molecules Paula Wolf Bryant, Ana-Maria Lennon-Dume´nil, Edda Fiebiger, Ce´cile Lagaudrie´re-Gesbert, and Hidde L. Ploegh Cytokine Memore of T Helper Lymphocytes Max Lo¨hning, Anne Richter, and Andreas Radbruch Ig Gene Hypermutation: A Mechanism is Due Jean-Claude Weill, Barbara Bertocci, Ahmad Faili, Said Aoufouchi, Ste´phane Frey, Annie De Smet, Se´bastian Storck, Auriel Dahan, Fre´de´ric Delbos, Sandra Weller, Eric Flatter, and Claude-Agne´s Reynaud Generalization of Single Immunological Experiences by Idiotypically Mediated Clonal Connections Hilmar Lemke and Hans Lange The Aging of the Immune System B. Grubeck-Loebenstein and G. Wick Index

304

CONTENTS OF RECENT VOLUMES

Volume 81 Regulation of the Immune Response by the Interaction of Chemokines and Proteases Jo Van Damme and Sofie Struyf Molecular Mechanisms of Host-Pathogen Interaction: Entry and Survival of Mycobacteria in Macrophages Jean Pieters and John Gatfield

Tumor Vaccines Freda K. Stevenson, Jason Rice, and Delin Zhu Immunotherapy of Allergic Disease R. Valenta, T. Ball, M. Focke, B. Linhart, N. Mothes, V. Niederberger, S. Spitzauer, I. Swoboda, S.Vrtala, K. Westritschnic, and D. Kraft

B Lymphoid Neoplasms of Mice: Characteristics of Naturally Occurring and Engineered Diseasse and Relationships to Human disorders Herbert Morse et al.

Interactions of Immunoglobulins Outside the Antigen-Combining Site Roald Nezlin and Victor Ghetie

Prions and the Immune System: A Journey Through Gut Spleen, and Nerves Adriano Aguzzi

The Roles of Antibodies in Mouse Models of Rheumatoid Arthritis, and Relevance to Human Disease Paul A. Monach, Christophe Benoist, and Diane Mathis

Roles of the Semaphorin Family in Immune Regulation H. Kikutani and A. Kumanogoh HLA-G Molecules: from Maternal-Fetal Tolerance to Tissue Acceptance Edgardo Carosella et al. The Zebrafish as a Model Organism to Study Development of the Immune System Nick Trede et al. Control of Autoimmunity by Naturally Arising Regulatory CD4þ T Cells S. Sakaguchi Index

Volume 82 Transcriptional Regulation in Neutrophils: Teaching Old Cells New Tricks Patrick P. McDonald

MUC1 Immunology: From Discovery to Clinical Applications Anda M. Vlad, Jessica C. Kettel, Nehad M. Alajez, Casey A. Carlos, and Olivera J. Finn Human Models of Inherited Immunoglobulin Class Switch Recombination and Somatic Hypermutation Defects (Hyper-IgM Syndromes) Anne Durandy, Patrick Revy, and Alain Fischer The Biological Role of the C1 Inhibitor in Regulation of Vascular Permeability and Modulation of Inflammation Alvin E. Davis, III, Shenghe Cai, and Dongxu Liu Index

Traver and Akashi, Fig 1 Hematopoietic commitment model based on prospective isolation of lineage-restricted progenitors. HSC, hematopoietic stem cells, CMP, common myeloid progenitors, CLP, common lymphoid progenitors; GMP, granulocyte/monocyte progenitors; MEP, megakaryocyte/erythrocyte progenitors.

Traver and Akashi, Fig 5 Myeloid and lymphoid promiscuity in normal hematopoietic progenitors. Single cell mutiplex-RT-PCR assays for myeloid-related (b-globin and EpoR for erythroid, G-CSFR and MPO for myelomonocytic) and for lymphoid-related genes (l5 and Pax-5 for B lymphoid, GATA-3 and CD3d for T lymphoid). The red rectangles indicate lineage promiscuous expression and white ones indicate no expression. More than 50% of single common myeloid progenitors (CMPs) coexpress both erythroid and myelomonocytic genes, and 20% of single common lymphoid progenitors (CLPs) coexpress both B and T lymphoid-related genes. These data strongly suggest that expression of lineage-related genes precedes commitment, and therefore downregulation of genes of unselected lineage might be important for bipotent progenitors to ultimately commit to certain lineages.

Taniuchi ET AL., Fig 1 Enhancer and promoter elements in the cd4 locus that contribute to reporter transgene expression in double positive thymocytes versus mature single positive T cells. Constructs that utilize the human CD2 transgene (hCD2) are from Sawada et al., 1994, and unpublished studies of Zou and Littman. Constructs that utilize HLA-B7 are from Adlam and Siu, 2003. E4d, distal enhancer; E4p, proximal CD4 enhancer; P4, CD4 promoter; LCR/TE, locus control region and thymic enhancer located in ﬁrst intron of isot gene. DH denotes DNase hypersensitive sites in the cd4 locus. Sites within the cd4 coding region have been omitted. We propose that DH5 may encode a negative DP thymocyte regulatory element that requires TE or another unidentiﬁed enhancer for expression in immature thymocytes.

Taniuchi ET AL., Fig 2 Transcriptional cis-regulatory elements in the murine cd4 locus and functional sites within the murine CD4 silencer. The coding exons are shown as closed bars and the noncoding 50 exon is shown as an open bar. The vertical arrows indicate DNaseI hypersensitive sites. The transcriptional direction of the CD4 gene is shown by the horizontal arrow. Proximal (E4p) and distal (E4d) enhancers (located approximately 13 kb and 24 kb upstream of the transcriptional start site, respectively) and a putative intronic (E4i) enhancer (located immediately 30 to the silencer) are shown as blue square boxes. The promoter (P4) and the silencer (S4) are shown as a purple square box and orange circle, respectively. The expanded region below the map represents the 434 bp murine CD4 silencer and putative factor binding sites. The top bar shows three footprint sites (yellow circles) deﬁned by Siu and colleagues (Duncan et al., 1996) along with putative trans-acting factors. The second bar shows ﬁve sites deﬁned by transient transfection assays, transgenic reporter assays, and targeted mutagenesis at the cd4 locus (Taniuchi et al., 2002b). The core silencer (sequence 165–265) was sufﬁcient for silencer activity in transfection assays. Site 2 and site 20 are identical to the binding motif for Runt domain transcription factor (Runx) family members. The effects of targeted mutations on CD4 gene silencing in either immature DN thymocytes or peripheral CD8þ mature T cells are shown at the bottom: ‘‘Part/uni’’ denotes partial CD4 de-repression in a uniform pattern. ‘‘Part/vari’’ denotes partial CD4 de-repression in a variegated pattern. ‘‘Full’’ represents CD4 de-repression in all of the CD8þ T cells. n.t.: not tested.

Taniuchi

ET AL.,

Fig 3

(continues)

Taniuchi ET AL., Fig 3 A model for distinct mechanisms of silencer-mediated CD4 repression at two developmental stages. (A) In immature CD4CD8 DN thymocytes, several CD4 silencer binding factors, including the BAF chromatin remodeling complex, together recruit corepressor molecules, resulting in reversible transcriptional repression (active repression). Runx sites (site 2 and site 20 ) would be occupied by Runx1 at this stage. At the transitional stage from CD4þCD8þ DP thymocytes to CD4CD8þ SP cytotoxic-lineage T lymphocytes, lineage speciﬁc modiﬁcations of chromatin structure (small orange circle) are established through the function of distinct machinery (orange square) recruited by the CD4 silencer binding factor complex. We propose that this epigenetic modiﬁcation is preceded by the recruitment of a reversible complex similar to that found in DN thymocytes. Altered chromatin structure would contain HP-1 molecules and serve as a heritable mark for epigenetic maintenance of the silenced status. Runx sites would be occupied mainly by Runx3 at this stage. (B) Compromised silencer function due to a mutation of site 1 results in partial uniform de-repression of CD4 in immature DN thymocytes (left). During the transition from the CD4þCD8þ DP stage to CD8þ SP thymocytes, at which epigenetic silencing is established, the compromised silencer function results in one of two possible outcomes at the CD4 locus. In a fraction of the CD8-lineage T cells, modiﬁcation of chromatin is complete enough for epigenetic inheritance of the silenced status. However, in the rest of these cells, modiﬁcation of chromatin is not complete enough to shut off the CD4 gene. Following subsequent cell divisions, the amount of CD4 transcription would be dependent on the status of chromatin modiﬁcation and would thus vary. The mixture of cells harboring silenced or activated CD4 results in variegated CD4 de-repression. HP-1 contributes to the successful establishment of the epigenetic mark.

Taniuchi ET AL., Fig 4 Summary of characterization of cis-regulatory regions in the murine cd8 locus. Upper part: Schematic map of the cd8a and cd8b loci on mouse chromosome 6. Vertical arrows indicate individual DNase I hypersensitivity (DH) sites that have been grouped to DH clusters I–IV (Hostert et al., 1997a). Horizontal arrows indicate the location and transcriptional orientation of the cd8a and cd8b genes, while the open and closed bars indicate coding and noncoding exons, respectively. All BamHI (B), but only relevant EcoRI (E), sites are shown. The black horizontal bars show the genomic fragments used to deﬁne E8I, E8II, E8III, and E8IV. It is very likely that the enhancer activities within these genomic fragments overlap with some of the DH sites that map within the fragments. Middle part: Graphic representation of the genomic fragments used in transgenic reporter expression assays. The enhancer activity of the various fragments is shown at the right. The references reporting the activities are: T1 and T2 (Hostert et al., 1997a); T3 (Ellmeier et al., 1997; Hostert et al., 1997b); T4 (Hostert et al., 1997b); T5 (Hostert et al., 1998); T6 (Hostert et al., 1998; Zhang et al., 1998); T7–T10 (Ellmeier et al., 1998). þ indicates strong enhancer activity, þ/ weak activity, and no enhancer activity. Nd: not determined. Lower part: The bars indicate the genomic region deleted in enhancer-deﬁcient mice. The expression of CD8 in the absence of the enhancer is shown at the right. The references reporting the enhancer deletions are: K1 (Ellmeier et al., 1998; Hostert et al., 1998); K2 and K3 (Ellmeier et al., 2002); K4 (Garefalaki et al., 2002). þ indicates normal CD8 expression, þ/ reduced CD8 expression, and no CD8 expression. ‘‘Var’’ indicates variegated expression of CD8. Nd: not determined.

O’Garra and Robinson, Fig 1 Th1 cell development in protection against intracellular pathogens such as mycobacteria: dendritic cells (DC) and macrophages (Mf) produce cytokines (IL-12 and IL-27) that drive Th1 development, while IL-12 and IL-18 are important for maximal IFN-g production from effector Th1 cells to activate macrophage killing and opsonising antibody production. Th1 cells themselves, as well as DC and Mf, can produce IL-10 which acts to control this process and prevent immunopathology.

O’Garra and Robinson, Fig 2 Checkpoints in Th1 development. Presentation of pathogen antigens to naive T cells by antigen-presenting cells (APC) induces IL-12Rb2 expression on T cells. Dendritic cell or macrophage-derived IL-12, together with IL-27, initiates Th1 development. IL-27 signals through the T cell cytokine receptor (TCCR, also termed WSX-1). IL-12 acts via STAT4 to upregulate IFN-g production. Th1 development requires IFN-g itself, which activates STAT1 and induces expression of the transcription factor T-bet. T-bet is a major Th1 commitment factor and transactivates the IFN-g gene, as well as inducing chromatin remodeling of the IFN-g locus. IL-12 induces IL-18 receptor expression, which allows IL-18 to synergize with IL-12 to increase IFN-g production from committed Th1 effectors. IL-18 signals via IL-1 receptor-associated kinase (IRAK) to activate NF-kB, and the combination of IL-12 and IL-18 activates GADD45 and p38 mitogen kinase signaling, all of which are major ampliﬁers of the IFN-g response. Memory CD4þT cells respond to IL-23, an IL-12-related cytokine, which increases proliferation and may increase IFN-g production further, although this may not be its primary function, since IL-23 also acts on macrophages.

Stetson ET AL., Fig 1 Regulatory elements surrounding the IL4 gene. DNAse I hypersensitivity sites (HSS, arrows) correlate closely with regions known to regulate IL-4 expression. Enhancer elements are indicated in green, and a silencer element is depicted in red. IL-4 exons are numbered and depicted as black boxes. The relative positions of each element are not drawn to scale.

Stetson ET AL., Fig 2 Initiation and expression of type 2 immunity. (a) A schematic diagram depicting three levels at which Type 2 immune responses may be regulated. (b) A simpliﬁed comparison of how type 1 and type 2 responses are initiated. Recognition of pathogen-associated molecular patterns (PAMPS) by pattern recognition receptors (PRRs) results in the production of cytokines which modulate T cell differentiation in the draining lymph nodes. While the basic elements underlying all of these events have been elucidated for Type 1 immunity, very little is known about how recognition of type 2 pathogens is linked to Th2 differentiation.

Jiang and Chess, Fig 2 Tri-molecular interaction between regulatory CD8þ T cells and activated autologous CD4þ T cells. This ﬁgure illustrates that the speciﬁc tri-molecular interaction between the regulatory CD8þ T cells and target CD4þ T cells is through the recognition of Qa-1/ self-peptide complex expressed on certain activated CD4þ T cells by ab TCR on regulatory CD8þ T cells.

Jiang and Chess, Fig 4 Model of cognate interactions in the induction and function of Qa-1 restricted regulatory CD8þ T cells. The Qa-1 restricted regulatory CD8þ T cell selectively downregulate certain but not all antigen-activated CD4þ T cells based on the speciﬁc recognition of Qa-1/self-peptide/s expressed on certain CD4þ T cells by the ab TCR on the CD8þ T cells. In this regard, we have demonstrated in the EAE model that self-reactive CD4þ T cells which are selectively downregulated by the CD8þ T cells are enriched in potentially pathogenic self-reactive T cell clones (Jiang et al., 2003).

Jiang and Chess, Fig 5 Regulatory T cells control the peripheral induction and clonal outgrowth of self-reactive T cells. This ﬁgure illustrates various pathways of immunoregulation mediated by suppressor subsets of NKT, CD4þ, and CD8þ T subsets. Each of the regulatory T cell subsets expresses distinct receptors, employs different effector mechanisms, and functions predominately at different stages during the course of the peripheral immune response. The NKT and CD4þCD25þ regulatory cells are ‘‘natural suppressor cells’’ and are present prior to antigen activation and function primarily during the early ‘‘innate’’ and/or primary adaptive immune responses. In contrast, the CD8þ regulatory cells are induced to differentiate into suppressor effector cells during the primary immune response and they function as effector suppressor cells predominately during the secondary and memory phases of immunity.