Advances in Immunology Volume 89

Contents Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i...

Author: Frederick W. Alt

42 downloads 947 Views 5MB 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

Contents

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Posttranscriptional Mechanisms Regulating the Inflammatory Response Georg Stoecklin and Paul Anderson 1. 2. 3. 4. 5. 6.

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coordinate Regulation of Pro-Inflammatory Proteins . . . . . . . . . . . Control of mRNA Stability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Translational Inhibition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Animal Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 5 16 19 23 25 26

Negative Signaling in Fc Receptor Complexes Marc Dae¨ron and Renaud Lesourne 1. 2. 3. 4. 5.

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fc Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Positive Signaling by Activating FcRs . . . . . . . . . . . . . . . . . . . . . . . Negative Signaling by Activating FcRs . . . . . . . . . . . . . . . . . . . . . . Negative Signaling by Inhibitory FcRs . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v

39 40 44 50 60 70 74

c on t e n ts

vi

The Surprising Diversity of Lipid Antigens for CD1-Restricted T Cells D. Branch Moody 1. 2. 3. 4. 5. 6. 7. 8.

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction: From Molecules to Functions . . . . . . . . . . . . . . . . . . CD1 Protein Expression on Antigen-Presenting Cells . . . . . . . . . . Subcellular Lipid Antigen Processing Pathways. . . . . . . . . . . . . . . . 3-Dimensional Structures of CD1-b2-Microglobulin-Lipid Complexes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microbial Antigens and Infectious Disease . . . . . . . . . . . . . . . . . . . Self Antigens, Autoreactivity, and Autoimmune Disease . . . . . . . . . Synthetic Lipid Antigens and Prospects for Immunotherapy . . . . . Conclusion: Prospects for Immunotherapy . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87 87 90 94 102 111 119 125 126 128

Lysophospholipids as Mediators of Immunity Debby A. Lin and Joshua A. Boyce Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. LPL Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Cell Surface Receptors for LPLs and Their Signaling Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Expression of LPA and S1P Receptors by Immune Cells and Their Functions in In Vitro Studies . . . . . . . . . . . . . . . . . . . . . 5. In Vivo Functions of LPLS in Immune Responses and Inflammation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Clinical Applications of S1P Receptor Agonists. . . . . . . . . . . . . . . . 7. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

141 141 143 145 148 156 159 160 160

Systemic Mastocytosis Jamie Robyn and Dean D. Metcalfe 1. 2. 3. 4.

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mast Cell Biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biology of Kit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mastocytosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

169 169 170 184 193

c o nt e n t s 5. 6. 7. 8. 9.

Prognosis and Predictive Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Supportive Care and Long-Term Management . . . . . . . . . . . . . . . . Future Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii 209 209 216 218 220 220

Regulation of Fibrosis by the Immune System Mark L. Lupher, Jr. and W. Michael Gallatin 1. 2. 3. 4. 5. 6. 7.

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fibrotic Disease Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cellular Mediators of Fibrosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inflammatory Chemokines that Regulate Fibrosis. . . . . . . . . . . . . . The Role of Integrins in Regulating the Fibrotic Response . . . . . . Other Potential Targets for Anti-Fibrotic Therapy . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

245 245 246 249 261 264 268 270 273

Immunity and Acquired Alterations in Cognition and Emotion: Lessons from SLE Betty Diamond, Czeslawa Kowal, Patricio T. Huerta, Cynthia Aranow, Meggan Mackay, Lorraine A. DeGiorgio, Ji Lee, Antigone Triantafyllopoulou, Joel Cohen-Solal, and Bruce T. Volpe 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lupus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neuropsychiatric Lupus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanisms of NPSLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mouse Models of NPSLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peptide Reactivity of DNA-Reactive Antibodies . . . . . . . . . . . . . . . Presence of DWEYS-Reactive Antibodies in Murine and Human SLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proteins Harboring the D/E W D/E Y S/G Concensus Sequence: Glutamate Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . Antibody-Mediated Neurotoxicity . . . . . . . . . . . . . . . . . . . . . . . . . . A Murine Model for Antibody-Mediated Neuronal Death . . . . . . . Antibody-Mediated Neurotoxicity in the Amygdala. . . . . . . . . . . . .

289 290 293 295 296 298 298 300 300 302 304 307

viii 12. 13. 14. 15. 16.

c on t e n ts Evidence that Antibodies Are Involved in NPSLE in Patients . . . . Anti-Peptide Antibody Activates Prolactin Secretion . . . . . . . . . . . . Anti-Peptide Antibodies and Manifestations of NP-SLE. . . . . . . . . Implications for SLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Implications for Human Pathobiology . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

308 309 310 310 311 311

Immunodeficiencies with Autoimmune Consequences Luigi D. Notarangelo, Eleonora Gambineri, and Raffaele Badolato Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Autoimmune Manifestations in Primary Immune Deficiencies: Relevance and Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. AIRE, Central Tolerance, and the Pathophysiology of Autoimmune Polyendocrinopathy. . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Defective AIRE Expression Explains the Pathophysiology of Autoimmunity in Omenn Syndrome . . . . . . . . . . . . . . . . . . . . . . . . 4. CD4þ CD25þ Regulatory T Cells and the Pathophysiology of IPEX (Immunodysregulation – Polyendocrinopathy – Enteropathy – X-Linked) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Concluding Remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

321 322 324 337

346 357 359

Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 Contents of Recent Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385

Contributors

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

Paul Anderson (1), Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts Cynthia Aranow (287), Department of Medicine, Columbia University Medical Center, New York, New York Raffaele Badolato (319), ‘‘Angelo Nocivelli’’ Institute for Molecular Medicine, Department of Pediatrics, University of Brescia, Brescia, Italy Joshua A. Boyce (141), Departments of Medicine and Pediatrics, Harvard Medical School, Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Boston, Massachusetts Joel Cohen-Solal (287), Department of Medicine, Columbia University Medical Center, New York, New York Marc Dae¨ron (39), Unite´ d’Allergologie Mole´ culaire et Cellulaire, De´partement d’Immunologie, Institut Pasteur, Paris, France Lorraine A. DeGiorgio (287), Department of Neurology and Neuroscience, The Burke Medical Research Institute, Weill Medical College of Cornell University, White Plains, New York Betty Diamond (287), Department of Medicine, Columbia University Medical Center, New York, New York W. Michael Gallatin* (245), ICOS Corporation, Bothell, Washington Eleonora Gambineri (319), Department of Pediatrics, University of Florence, Florence, Italy Patricio T. Huerta (287), Department of Neurology and Neuroscience, The Burke Medical Research Institute, Weill Medical College of Cornell University, White Plains, New York Czeslawa Kowal (287), Department of Medicine, Columbia University Medical Center, New York, New York *Current address: Frazier Healthcare Ventures, Seattle, Washington.

ix

x

c o n tr i b u t o rs

Ji Lee (287), Department of Microbiology and Immunology, Albert Einstein College of Medicine, Bronx, New York Renaud Lesourne (39), Unite´ d’Allergologie Molećulaire et Cellulaire, De´partement d’Immunologie, Institut Pasteur, Paris, France Debby A. Lin (141), Department of Medicine, Harvard Medical School, and Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Boston, Massachusetts Mark L. Lupher, Jr. (245), ICOS Corporation, Bothell, Washington Meggan Mackay (287), Department of Medicine, Columbia University Medical Center, New York, New York Dean D. Metcalfe (169), Laboratory of Allergic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland D. Branch Moody (87), Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts Luigi D. Notarangelo (319), ‘‘Angelo Nocivelli’’ Institute for Molecular Medicine, Department of Pediatrics, University of Brescia, Brescia, Italy Jamie Robyn (169), Laboratory of Allergic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland Georg Stoecklin (1), Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts Antigone Triantafyllopoulou (287), Department of Medicine, Montefiore Medical Center, Bronx, New York Bruce T. Volpe (287), Department of Neurology and Neuroscience, The Burke Medical Research Institute, Weill Medical College of Cornell University, White Plains, New York

Posttranscriptional Mechanisms Regulating the Inflammatory Response Georg Stoecklin and Paul Anderson Division of Rheumatology, Immunology, and Allergy; Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts

1. 2. 3. 4. 5. 6.

Abstract............................................................................................................. Coordinate Regulation of Pro‐Inflammatory Proteins ................................................. Control of mRNA Stability .................................................................................... Translational Inhibition......................................................................................... Animal Models ................................................................................................... Clinical Significance............................................................................................. Conclusions........................................................................................................ References .........................................................................................................

1 1 5 16 19 23 25 26

Abstract The inflammatory response is a complex physiologic process that requires the coordinate induction of cytokines, chemokines, angiogenic factors, effector‐ enzymes, and proteases. Although transcriptional activation is required to turn on the inflammatory response, recent studies have revealed that posttranscriptional mechanisms play an important role by determining the rate at which mRNAs encoding inflammatory effector proteins are translated and degraded. Most posttranscriptional control mechanisms function to dampen the expression of pro‐inflammatory proteins to ensure that potentially injurious proteins are not overexpressed during an inflammatory response. Here we discuss the factors that regulate the stability and translation of mRNAs encoding pro‐inflammatory proteins. 1. Coordinate Regulation of Pro‐Inflammatory Proteins In prokaryotes, the coordinate expression of genes encoding components of a metabolic pathway is often accomplished by expressing an mRNA transcript that encodes more than one protein. For example, the Lactose operon in Escherichia coli transcribes a single mRNA that encodes b‐galactosidase, permease, and transacetylase—proteins required for the utilization of lactose as a carbon source (Beckwith, 1967). Unlike their prokaryotic ancestors, processed eukaryotic transcripts generally encode a single protein. In eukaryotes, the coordinate expression of protein components of a complex functional program is accomplished by the use of regulatory nucleic acid sequences

1 advances in immunology, vol. 89 # 2006 Elsevier Inc. All rights reserved.

0065-2776/06 $35.00 DOI: 10.1016/S0065-2776(05)89001-7

2

G E O R G S T O E C K L I N A N D PA U L A N D E R S O N

that determine rates of transcription, translation, and mRNA decay. Coordinate transcriptional control is accomplished by the selective use of promoter elements that are recognized by specific transcription factors. Activation of a transcription factor that turns on the expression of several genes involved in a common biological pathway can be sufficient to confer coordinate protein expression. One example is the heat shock response in which stress‐induced activation of heat shock factor 1 confers de novo transcription of mRNAs encoding several heat shock proteins (Voellmy, 2004). Mammalian cells have evolved posttranscriptional mechanisms that further coordinate the expression of proteins involved in a common cellular function. In some cases, regulation of mRNA stability and translation may be sufficient to coordinately regulate the expression of distinct functional classes of proteins. This appears to be the case in Saccharomyces cerevisiae in which gene array profiling has shown that mRNAs encoding subunits of complex structures such as proteasomes or ribosomes tend to have similar half‐lives (Wang et al., 2002). In many cases, posttranscriptional regulation is conferred by cis‐acting elements located in the 30 ‐untranslated region (30 UTR) of individual mRNAs. These regulatory elements recruit specific RNA‐binding proteins that, directly or indirectly, regulate mRNA translation and/or stability. A striking example of this type of regulation is provided by a family of Pumilio‐ Fem‐3‐binding factor (Puf) proteins. The Puf proteins bind to variants of a UGUR tetranucleotide motif found in the 30 UTR of selected mRNAs to regulate mRNA stability and/or translation (Wickens et al., 2002). In Saccharomyces cerevisiae, gene array profiling has revealed that Puf proteins can coordinately regulate the expression of mRNAs encoding proteins with a common function (Gerber et al., 2004). Thus, Puf1p and Puf2p were found to bind to mRNAs encoding membrane‐associated proteins, Puf3p binds to mRNAs encoding mitochondrial proteins, and Puf4p and Puf5p bind to mRNAs encoding nuclear proteins. Each Puf protein binds to a specific variant of the UGUR motif to confer specificity. These results strongly support the contention that RNA‐binding proteins coordinately regulate groups of mRNAs encoding proteins of a common function. This general mechanism has been referred to as a ‘‘post‐transcriptional operon’’ (Keene and Tenenbaum, 2002). The inflammatory response is an example of a functional program that requires the coordinate induction of proteins involved in a common function. During the inflammatory response, coordinate transcriptional control is accomplished by the selective use of promoter elements that are recognized by specific transcription factors. For example, the transcription factor Nuclear factor (NF)‐kB moves from the cytoplasm to the nucleus in T‐cells and macrophages that are exposed to inflammatory stimuli (Muller, 2001). Nuclear

P O S T T R A N S C R I P T I O N A L C O N T R O L O F I N F L A M M AT I O N

3

NF‐kB binds to a nucleotide sequence element common to the promoters of several genes that encode inflammatory effector proteins (Muller, 2001). Although transcriptional activation is required for the synthesis of mRNA encoding inflammatory effector proteins, it does not determine the level of protein expression. Rather, it is the posttranscriptional regulation of mRNA stability and translation that determines levels of protein expression. Remarkably, mRNAs encoding proteins that regulate every facet of the inflammatory response are subject to these posttranscriptional control mechanisms. At the initiation phase of inflammation, Granulocyte‐macrophage colony‐stimulating factor (GM‐CSF) primes neutrophils, macrophages, and dendritic cells for subsequent pro‐inflammatory action (Hamilton, 2002). Moreover, the expression of chemokines that recruit neutrophils, eosinophils, monocytes, and dendritic cells to sites of inflammation is strongly regulated at the posttranscriptional level. Effects on cellular recruitment are potentiated by the posttranscriptional regulation of the endothelial adhesion factor Vascular cell adhesion molecule (VCAM)‐1. During the effector phase of inflammation, the expression of pro‐inflammatory cytokines such as Tumor necrosis factor (TNF)a, Interleukin (IL)‐1b, IL‐6, and Interferon (IFN)g is also subject to posttranscriptional control. Excessive production of these cytokines orchestrates tissue damage in patients with inflammatory arthritis and inflammatory bowel disease. Proteases such as Matrix metalloprotease (MMP)‐9 and MMP‐ 13 participate in the tissue damage by breaking down components of the extracellular matrix. The effector phase is further exacerbated by enzymes that synthesize pro‐inflammatory mediators such as Cyclooxygenase (COX)‐2, 15‐Lipoxygenase (LO), and inducible Nitric oxide synthase (iNOS), which are also regulated at the level of mRNA stability and translation. Taken together, it is clear that posttranscriptional control mechanisms coordinately regulate multiple aspects of the inflammatory response. In general, these mechanisms conspire to dampen inflammation. This additional level of control may be required to prevent the pathological overexpression of proteins that are potentially injurious to the host. The synchronized anti‐inflammatory effects of these regulatory programs make them ideal targets for the development of anti‐ inflammatory drugs. In this review, we will focus on two mechanisms that regulate the fate of mRNAs in the cytoplasm: control of translation efficiency and control of mRNA stability. Although earlier steps (pre‐mRNA splicing, export of the mRNA from the nucleus into the cytoplasm) and later events (protein secretion, activation by proteolytic cleavage) are extensively regulated as well, they will not be discussed here. Table 1 lists pro‐inflammatory proteins whose expression is regulated at the level of translation or mRNA stability. Most of these proteins fall into one of three categories: cytokines, chemokines, and enzymes required for the synthesis of pro‐inflammatory mediators.

4


Table 1 Pro‐Inflammatory Proteins Regulated at the Posttranscriptional Level mRNA

Mechanism

Element

References

Cytokines IL‐1b

mRNA stability

AREa

IL‐1F7b IL‐2

mRNA stability mRNA stability

CRDb ARE, JREc

IL‐3

mRNA stability

ARE

IL‐6

mRNA stability

ARE

IL‐11

mRNA stability

ARE

IL‐18 TNFa

mRNA stability mRNA stability, translation

n.d. ARE, CDEd

GM‐CSF

mRNA stability, translation

ARE

G‐CSF

mRNA stability

ARE, SLDEe

IFNb

mRNA stability

ARE, CRD

IFNg

mRNA stability

ARE

VEGF

mRNA stability

ARE, 50 UTR, CRD

MCP‐1 (CCL2) Eotaxin (CCL11) GROa (KC, CXCL1) GROb (MIP‐2, CXCL2) GROg (CXCL3) IL‐8 (CXCL8)


n.d. ARE

(Pastore et al., 2005) (Atasoy et al., 2003)

mRNA stability

ARE

mRNA stability

ARE


ARE ARE

IP‐10 (CXCL10)

mRNA stability

n.d.

(Biswas et al., 2003; Sirenko et al., 1997; Stoeckle, 1991) (Rousseau et al., 2002; Stoeckle, 1991) (Stoeckle, 1991) (Holtmann et al., 1999; Stoeckle, 1991; Winzen et al., 1999) (Vockerodt et al., 2005)

(Fenton et al., 1988; Sirenko et al., 1997) (Bufler et al., 2004) (Chen et al., 1998; Lindsten et al., 1989; Ogilvie et al., 2005) (Nair et al., 1994; Wodnar‐Filipowicz and Moroni, 1990) (Akashi et al., 1990; Neininger et al., 2002; Stoecklin et al., 2001; Winzen et al., 1999) (Bamba et al., 2003; Yang and Yang, 1994) (Bufler et al., 2004) (Brook et al., 2000; Carballo et al., 1998; Han et al., 1990; Stoecklin et al., 2003) (Carballo et al., 2000; Grosset et al., 2004; Koeffler et al., 1988; Shaw and Kamen, 1986) (Brown et al., 1996a; Koeffler et al., 1988; Putland et al., 2002) (Raj and Pitha, 1983; Whittemore and Maniatis, 1990) (Hodge et al., 2002; Lindsten et al., 1989; Mavropoulos et al., 2005) (Coles et al., 2004; Dibbens et al., 1999; Ikeda et al., 1995; Levy et al., 1998)

Chemokines

5


Table 1 (Continued) mRNA

Mechanism

Element

References

Enzymes and other pro‐inflammatory proteins VCAM‐1

mRNA stability

n.d.

MMP‐9 MMP‐13 15‐LO iNOS

mRNA stability translation translation mRNA stability

ARE ARE DICE f ARE

COX‐2

mRNA stability, translation

ARE

(Croft et al., 1999; Iademarco et al., 1995) (Huwiler et al., 2003) (Yu et al., 2003) (Ostareck et al., 1997) (Di Macro et al., 2005; Rodriguez‐Pascual et al., 2000) (Dixon et al., 2003; Lasa et al., 2000; Mukhopadhyay et al., 2003; Ridley et al., 1998; Ristimaki et al., 1994)

a

ARE: AU‐rich element. CRD: coding region determinant of instability. c JRE: JNK response element. d CDE: constitutive decay element. e SLDE: stem‐loop destabilizing element. f DICE: differentiation control element. b

2. Control of mRNA Stability 2.1. mRNA Decay Mediated by the AU‐Rich Element Many of the pro‐inflammatory transcripts that are regulated at the posttranscriptional level have an adenine/uridine‐rich element (ARE) in their 30 UTR, which targets individual transcripts for rapid cytoplasmic degradation. AREs were initially discovered by their characteristic nucleotide pattern, and the high degree of sequence conservation between different mammalian species was an early indication of their important regulatory role (Caput et al., 1986). ARE‐mediated mRNA decay (AMD) was first demonstrated by Shaw and Kamen (1986), who inserted the ARE of GM‐CSF into the 30 UTR of a normally stable reporter mRNA (in this case a b‐globin transcript), and observed that it strongly reduced reporter gene expression by destabilizing the mRNA (Shaw and Kamen, 1986). Conversely, deletion of the ARE from the 30 UTR of IL‐3 or TNFa enhances gene expression by stabilizing the mRNA (Kontoyiannis et al., 1999; Stoecklin et al., 1994). Besides inducing AMD, quantitative analysis of TNFa reporter gene expression showed that the ARE also inhibits translation of the mRNA (Han et al., 1990). AREs are U‐rich sequences that contain several copies of a canonical AUUUA pentamer.

6


According to a classification that takes into account deadenylation kinetics, the typical class II AREs of cytokine mRNAs (e.g., GM‐CSF, IL‐3, and TNFa) contain a cluster of 4–7 partially overlapping AUUUA pentamers within a U‐rich context (Chen and Shyu, 1995). Class I AREs (e.g., c‐myc, c‐fos) have fewer and more scattered AUUUA pentamers, while class III AREs (e.g., c‐jun) lack this motif. Class II AREs trigger a very rapid, asynchronous shortening of the poly‐A tail (processive deadenylation activity) prior to decay of the mRNA body, whereas class I and III AREs are associated with a less rapid, synchronous shortening of the poly‐A tail (distributive deadenylation activity) (Chen et al., 1995). For class II AREs, the AUUUA pentamers are of crucial importance, since point mutations within the pentamers efficiently abrogate AMD (Stoecklin et al., 1994). The extended UUAUUUAUU nonamer is the minimal sequence that can induce mRNA degradation (Lagnado et al., 1994; Zubiaga et al., 1995). This sequence is also the minimal binding site for tristetraprolin (TTP), an ARE‐binding protein required for AMD (see below). Compared to this minimal sequence, typical AREs are much longer (50–100 nucleotides), indicating that AREs of different mRNAs may vary with respect to their destabilizing activity, their ability to inhibit translation, and the regulatory mechanisms that control these activities. Since AREs were first discovered in cytokine transcripts, the biological importance of controlling mRNA stability has primarily been established in cells of the immune system. Nevertheless, AMD is a highly conserved regulatory mechanism that functions in yeast (Vasudevan and Peltz, 2001), trypanosomes (Quijada et al., 2002), Drosophila (Jing et al., 2005), and virtually all mammalian cell lines. Based on the frequency of AU‐rich sequences in the human genome, it has been estimated that 5–10% of all mRNAs may contain an ARE (Bakheet et al., 2003). Although experimental data is still needed to confirm this high estimate of functional AREs, it is clear that AMD is a general mechanism that regulates gene expression in most, if not all, eukaryotic cells. 2.2. The Different Roles of ARE‐Binding Proteins More than 20 different proteins that bind to AREs have so far been identified, but only a subset of them has been shown to influence the stability or translation efficiency of their target mRNAs. Table 2 gives an overview of ARE‐binding proteins (ARE‐BPs) that are known to have a regulatory function: TIA‐1, TIAR, FXR1P, and CUGBP2 inhibit the translation of ARE‐ mRNAs; HuR and YB1 stabilize ARE‐mRNAs; whereas AUF1, KSRP, RHAU, and the TTP/BRF family of proteins are involved in destabilizing ARE‐mRNAs.

7


Table 2 ARE‐Binding Proteins ARE‐BP TIA‐1 TIAR FXRIP CUGBP2 HuR

Domains

Function

References

RRM (3)*

Inhibits translation

(Piecyk et al., 2000)

RRM (3) KH (2), RGG (1) RRM (3) RRM (3)

Inhibits translation Inhibits translation Inhibits translation Stabilizes mRNA

(Yu et al., 2003) (Garnon et al., 2005) (Mukhopadhyay et al., 2003) (Fan and Steitz, 1998; Lal et al., 2004; Peng et al., 1998; Wang et al., 2000) (Lal et al., 2004; Mazan‐Mamczarz et al., 2003) (Kullmann et al., 2002) (Capowski et al., 2001; Chen et al., 2000; Coles et al., 2004) (Lal et al., 2004; Loflin et al., 1999; Sarkar et al., 2003) (Xu et al., 2001) (Dean et al., 2002) (Chen et al., 2001; Gherzi et al., 2004) (Tran et al., 2004) (Carballo et al., 1998; Chen et al., 2001; Lai et al., 1999; Ming et al., 2001) (Lai et al., 2000; Stoecklin et al., 2000, 2002) (Lai et al., 2000)

Activates translation

YB1

CSD (1)

Inhibits translation Stabilizes mRNA

AUF1 (hnRNP D0)

RRM (2)

Destabilizes mRNA

AUF2 (CBF‐A) KSRP

RRM (2) KH (4)

Stabilizes mRNA Stabilizes mRNA Destabilizes mRNA

RHAU TTP (TIS11)

Helicase (1) C3H (2)

Destabilizes mRNA Destabilizes mRNA

BRF1 (TIS11b)

C3H (2)

Destabilizes mRNA

BRF2 (TIS11d)

C3H (2)

Destabilizes mRNA

*Designates the number of domains.

Genetic studies have provided evidence that the zinc finger proteins TTP, BRF1, and BRF2 play a central role in the degradation of ARE‐mRNAs. The TTP/BRF proteins bind to AREs with high specificity through their characteristic tandem C3H zinc finger domains (Blackshear, 2002; Varnum et al., 1991). The function of TTP was discovered by the study of TTP‐deficient mice, which develop generalized inflammatory symptoms that arise from increased levels of TNFa and GM‐CSF (Taylor et al., 1996). Cytokine overproduction in TTP/ mice is due to an increased stability of TNFa, GM‐CSF, and IL‐2 mRNAs (Carballo et al., 1998, 2000; Ogilvie et al., 2005). TTP normally binds to the ARE of these mRNAs and promotes their rapid degradation by targeting them to the cellular RNA degradation machinery (see below). Since TTP is mainly expressed in activated macrophages and T‐cells (Cao et al., 2004; Raghavan

8


et al., 2001), its principal function appears to be the control of cytokine expression in the immune system. Yet TTP is also expressed in a variety of other tissues such as lung, liver, and intestine (Cao et al., 2004; Lu and Schneider, 2004), and may thus be involved in regulating other processes. BRF1 and BRF2 are homologues of TTP which share about 70% sequence identity in the highly conserved zinc finger domain, but differ substantially in the N‐ and C‐terminal domains (Varnum et al., 1991). BRF1‐deficient mice are not viable due to embryonic lethality at day 11 post‐fertilization, probably as a result of severe placental dysfunction (Stumpo et al., 2004). A mutant cell line generated by chemical mutagenesis and selected for a defect in AMD has provided evidence for the important role of BRF1 (Stoecklin et al., 2000, 2002). This mutant cell line lacks expression of BRF1 due to point mutations in both alleles, and is unable to rapidly degrade reporter transcripts containing the AREs of TNFa, GM‐CSF, IL‐2, IL‐3, and IL‐6 (Stoecklin et al., 2001). The results obtained with TTP/ mice and BRF1/ cells indicate that cytokine transcripts with class II AREs are subject to a common degradation pathway. Overexpression studies have shown that BRF2, the second homologue of TTP, can also induce AMD (Lai et al., 2000). Mice expressing an N‐terminally deleted form of the BRF2 protein are viable, but exhibit complete female infertility due to block of early embryonic development at the two‐cell stage (Ramos et al., 2004). Taken together, the data from the knock‐out mice reveal that TTP plays a critical and non‐redundant role in the immune system, whereas both BRF1 and BRF2 appear to be important regulators of embryonic development. It is noteworthy that prolonged overexpression of TTP and BRF1 causes apoptotic cell death in a variety of cell lines (Johnson and Blackwell, 2002; Johnson et al., 2000), indicating that the TTP/BRF family of proteins may target yet unidentified mRNAs. Characterizing the entire spectrum of mRNAs that are regulated by TTP/BRF will help to uncover biological processes for which posttranscriptional control of mRNA stability is relevant. Although enforced tethering of TTP to an mRNA that does not contain an ARE (through a heterologous RNA‐protein interaction) is sufficient to induce rapid degradation of the mRNA (Lykke‐Andersen and Wagner, 2005), it is clear that other ARE‐BPs are also required for destabilizing ARE‐mRNAs. KSRP was initially identified as an RNA‐binding protein that enhances splicing (Min et al., 1997), and later found to interact with the 30 UTR of IL‐2 transcripts (Chen et al., 2001). KSRP binds to different AREs (e.g., IL‐2, TNFa, c‐fos) through its KH domains, and simultaneously interacts with the exosome, a large complex of about ten 30 ‐50 exonucleases and associated helicases (Mitchell and Tollervey, 2000; Raijmakers et al., 2004). Importantly, KSRP was shown to be required for AMD both in vitro and in live cells (Chen et al., 2001; Gherzi et al., 2004). The helicase RHAU was identified as a protein that


9

binds to the ARE of urokinase plasminogen activator (uPA) and also interacts with the exosome (Tran et al., 2004). RHAU accelerates degradation of uPA mRNA, but it is not clear whether RHAU facilitates the degradation of other ARE‐mRNAs as well. The ARE‐BP AUF1 was identified as an activity that destabilizes an ARE‐ containing RNA in vitro (Brewer, 1991; Zhang et al., 1993). The destabilizing role of AUF1 has been confirmed by studies using siRNA to suppress AUF1 expression (Lal et al., 2004; Raineri et al., 2004), as well as by overexpression studies (Loflin et al., 1999; Sarkar et al., 2003). Its ability to interact with the exosome may provide the mechanistic basis for its activity (Chen et al., 2001). At least one study, however, has indicated that overexpression of AUF1 can also cause stabilization of ARE‐containing reporter mRNAs (Xu et al., 2001). This apparent contradiction may arise from the fact that AUF1 is expressed as four different isoforms (Wagner et al., 1998), which can differ in their activity (Raineri et al., 2004). Interestingly, overexpression of AUF2, a close homologue of AUF1, also stabilizes an ARE‐mRNA (Dean et al., 2002). HuR and its neural‐specific homologues HuB, HuC, and HuD form the embryonic lethal abnormal visual (ELAV) family of RNA‐binding proteins, all of which have high affinity for AU‐ and U‐rich sequences (Antic and Keene, 1997; Ma et al., 1996). Various studies have shown that HuR, the ubiquitously expressed member of the ELAV family, stabilizes a large number of ARE‐ mRNAs, including transcripts containing the ARE of GM‐CSF, IL‐3, and VEGF (Fan and Steitz, 1998; Ford et al., 1999; Levy et al., 1998; Ming et al., 2001; Peng et al., 1998; Raineri et al., 2004). HuR is required for the induced expression of iNOS (Di Macro et al., 2005; Rodriguez‐Pascual et al., 2000), and has also been implicated in the stabilization of IL‐8, eotaxin, COX‐2, and MMP‐9 mRNA (Atasoy et al., 2003; Cok et al., 2003; Huwiler et al., 2003; Winzen et al., 2004). Class II AREs typically contain 4 to 7 AUUUA pentamers within a U‐rich sequence of 50 or more nucleotides. This contrasts with the fact that TTP requires only the nonameric sequence UUAUUUAUU for efficient binding (Blackshear et al., 2003a; Worthington et al., 2002), and that a single C3H zinc finger of BRF2 makes direct contact with just four nucleotides: UAUU (Hudson et al., 2004). This would indicate that physiological AREs are large enough to accommodate more than one, and perhaps several ARE‐BPs. A simple model would postulate that stabilizing proteins such as HuR compete with destabilizing proteins such as TTP or AUF1 for binding to the ARE, and thereby prevent decay of the mRNA. It appears, however, that HuR does not bind within the core region of class II AREs (which consists of clustered AUUUA pentamers), but to an auxiliary, mainly U‐rich region upstream of the core region (Chen et al., 2002; Winzen et al., 2004). Paradoxically, this

10


auxiliary region both enhances the destabilizing activity of the ARE core region, and allows for stabilization of the mRNA by HuR. Such an upstream auxiliary region has been described for the AREs of c‐fos, TNFa, GM‐CSF, and IL‐8 (Chen et al., 2002; Stoecklin et al., 2001; Winzen et al., 2004). Very few studies have so far directly addressed the question whether ARE‐BPs bind individually to the ARE and compete for binding, or whether they bind cooperatively to form multi‐protein ARE‐mRNPs. One study found that KSRP and TTP together compete with HuR for binding to the Pitx2 mRNA (Briata et al., 2003). Using a genome‐wide approach, another study was able to distinguish between mRNAs that are simultaneously bound by both AUF1 and HuR, and mRNAs bound by the two proteins individually (Lal et al., 2004). Given the number of proteins which can bind to AREs (at least 20), and the fact that individual AREs differ quite considerably in their sequence, it is likely that each ARE recruits a different subset of ARE‐BPs, which together determine the stability and translation efficiency of that particular transcript. At present, we lack a broader understanding of the composition of specific ARE‐ mRNP complexes, and the changes these complexes may undergo when the degradation of the mRNA is activated or inhibited. 2.3. The Pathway of ARE‐mRNA Degradation In general, degradation of mRNAs in the cytoplasm is initiated by removal of the poly‐A tail, followed by exonucleolytic decay in the 30 ‐50 or the 50 ‐30 direction. At the 30 end, the mRNA is degraded through the exosome. At the 50 end, the 7‐methyl guanosine cap is removed by the decapping complex Dcp1/Dcp2, and the mRNA body is subsequently degraded by the 50 ‐30 exonuclease Xrn1 (Cougot et al., 2004b; Fillman and Lykke‐Andersen, 2005; Parker and Song, 2004). Dcp1/Dcp2 and Xrn1 form a larger complex with the Lsm1‐7 proteins, all of which are concentrated in small cytoplasmic foci termed processing bodies (Cougot et al., 2004a; Ingelfinger et al., 2002; van Dijk et al., 2002). Processing bodies are considered to be sites of mRNA degradation in the cytoplasm. In vitro decay studies have indicated that ARE‐mRNAs are degraded in the 30 ‐50 direction by the exosome (Chen et al., 2001), and one of the exosome components, Pm‐Scl‐75, was shown to directly bind to the ARE (Mukherjee et al., 2002). On the other hand, the ARE also stimulates decapping of the RNA in vitro (Gao et al., 2001). As illustrated in Fig. 1, the TTP/BRF family of proteins plays a central role in the degradation of ARE‐mRNAs. Both in vitro and when overexpressed in live cells, TTP enhances deadenylation of ARE‐ mRNA (Fig. 1A) (Lai et al., 1999, 2003). TTP and BRF1 further interact with both the exosome (Fig. 1B) and with components of the decapping/Xrn1 complex (Fig. 1C) (Chen et al., 2001; Gherzi et al., 2004; Lykke‐Andersen


11

Figure 1 The TTP/BRF family of proteins plays a central role in the degradation of ARE‐mRNAs. (A) Cytoplasmic mRNA degradation is initiated by deadenylation. (B) 30 ‐50 decay is mediated by the exosome (green), a complex of exonucleases and helicases. (C) At the 50 end, the mRNA is decapped by Dcp1/2 (yellow), followed by Xrn1‐mediated 50 ‐30 degradation. These proteins form a larger complex with Lsm1‐7, and are concentrated in cytoplasmic processing bodies. (D) The RNA‐induced silencing complex (RISC, in blue) and microRNA 16 are also required for ARE‐ mRNA degradation. The zinc finger proteins TTP/BRF1 (red) bind with high affinity to the ARE and induce degradation of the mRNA by interacting with the exosome, the decapping/Xrn1 complex, and the RISC complex.

and Wagner, 2005). This suggests that TTP/BRF1 are capable of targeting ARE‐mRNAs to both decay pathways. We have recently shown that the 50 ‐30 pathway is indeed important for degrading ARE‐mRNAs (Stoecklin et al., 2005). Using siRNA to target individual components of both pathways, Xrn1 and Lsm1 were found to be essential for AMD, whereas exosome components are less important. TTP and BRF1 also colocalize with processing bodies, a further indication that AMD uses the 50 ‐30 pathway (Kedersha et al., 2005). A recent study by Jing and coworkers (2005) revealed that AMD is dependent on the RNA‐induced silencing complex (RISC) and the microRNA miR16 (Fig. 1D). TTP interacts with two proteins of the RISC complex, Argonaute 2 and 4, which in turn allows the microRNA miR16 to base‐pair with the ARE. Interestingly, Argonaute proteins also colocalize with processing bodies and mRNAs targeted by miRNAs become concentrated in these structures (Liu et al., 2005). These novel findings indicate that the ARE actually consist of a dual code: one part of the ARE serves to recruit RNA‐binding proteins which interact with the cellular degradation machinery, while another part of the ARE (or adjacent sequences) base‐pair with regulatory microRNAs, which may contribute to the sequence specificity of the destabilizing element. A challenge of future studies will be to break this code in a way that will allow

12


us to predict for individual AREs which proteins bind to it, which microRNAs are associated, and what the functional consequences are for mRNA stability and translation efficiency. 2.4. Regulation of ARE‐mRNA Decay The ARE not only targets the mRNA for rapid degradation, but also allows the tight regulation of mRNA stability in response to extracellular cues. In a variety of cell types, many cytokine mRNAs have a very short half‐life (in the range of 10–30 minutes) in resting cells, effectively preventing cytokine protein production. Upon cell stimulation, the rapid induction of cytokine expression requires both transcriptional activation and stabilization of the mRNAs. In activated cells, cytokine mRNA half‐lives are in the range of several hours, which can easily account for a 10‐fold increase in protein production. In the post‐induction phase, ARE‐mRNAs are no longer stabilized and their degradation ensures rapid return to the low, usually undetectable levels of basal cytokine production. Thus, the dynamic regulation of mRNA stability is an important means by which cytokine levels are controlled in a time‐dependent manner. The following examples (listed in Table 1) document the general importance of this mode of posttranscriptional regulation. In T‐cells stimulated with anti‐CD3 and anti‐CD28 antibodies, or phorbol ester, induction of IL‐2, IFNg, TNFa, and GM‐CSF coincides with a marked increase in the stability of the corresponding mRNAs (Bickel et al., 1990; Lindsten et al., 1989). Stabilization of IL‐3 and GM‐CSF mRNA occurs in mast cells after treatment with calcium ionophores (Wodnar‐Filipowicz and Moroni, 1990), and GM‐CSF mRNA is likewise stabilized in stimulated eosinophils (Esnault and Malter, 1999). The production of IFNg in activated NK cells involves stabilization of the mRNA (Hodge et al., 2002; Mavropoulos et al., 2005). Macrophages produce a variety of cytokines and chemokines in response to activation by lipopolysaccharide (LPS), and mRNA stabilization contributes to the induction of IL‐1b, IL‐18, TNFa, GM‐CSF, Growth regulated oncogene (GRO)a and GROb expression (Biswas et al., 2003; Brook et al., 2000; Bufler et al., 2004; Carballo et al., 2000; Rousseau et al., 2002). In monocytes activated through adhesion, mRNAs encoding IL‐1b and GROa are stabilized (Sirenko et al., 1997). Regulation of ARE‐mRNA stability is not restricted to the cells of the immune system. Virally infected fibroblasts produce IFNb in the early phase of infection. During this phase, IFNb mRNA is stable, whereas rapid degradation of IFNb mRNA during the late phase accounts for the post‐induction turnoff of IFNb expression (Raj and Pitha, 1983; Whittemore and Maniatis, 1990). Stimulation of fibroblasts and other cell types with pro‐inflammatory


13

cytokines such as TNFa or IL‐1 induces the production of various cytokines and chemokines. In these cells, mRNA stabilization contributes to the induced expression of IL‐6, IL‐11, G‐CSF, GM‐CSF, Macrophage chemoattractant protein (MCP)‐1, eotaxin, GROa, GROb, GROg, and IL‐8 (Akashi et al., 1990; Atasoy et al., 2003; Bamba et al., 2003; Koeffler et al., 1988; Pastore et al., 2005; Stoeckle, 1991; Yang and Yang, 1994). Expression of the adhesion molecule VCAM‐1 on endothelial cells and fibroblast‐like synoviocytes is induced by TNFa and IL‐4, and involves stabilization of the mRNA (Croft et al., 1999; Iademarco et al., 1995). Since VCAM‐1 is overexpressed in rheumatoid arthritis synovium (Morales‐Ducret et al., 1992), it may promote the inflammation by recruiting monocytes and lymphocytes to the affected joint. The inducible nitric oxide synthase (iNOS) enzyme is upregulated in many cell types in response to bacterial products or pro‐inflammatory cytokines, leading to the sustained production of nitric oxide (NO). NO is a diffusible molecule that has beneficial antimicrobial effects in the defense against pathogens, and acts as a potent messenger in regulating tissue perfusion, epithelial permeability, and the immune response. On the other hand, overproduction of NO can be detrimental as it is associated with tissue damage and chronic inflammation (Kolios et al., 2004; MacMicking et al., 1997). The induction of iNOS occurs both at the transcriptional level, and at the posttranscriptional level through stabilization of the mRNA (Carpenter et al., 2001; Di Macro et al., 2005; Lahti et al., 2003; Perez‐Sala et al., 2001; Rodriguez‐Pascual et al., 2000). Another important effector‐enzyme of the inflammatory response, COX‐2, is also regulated at the posttranscriptional level. In a variety of cell types, IL‐1 induces the expression of COX‐2 by a mechanism that involves stabilization of the mRNA (Ridley et al., 1998; Ristimaki et al., 1994), and the same is observed in LPS‐treated monocytes (Dean et al., 1999). Taken together, these examples demonstrate that mRNA stabilization is a general mechanism by which different cell types achieve proper expression of a large group of pro‐inflammatory proteins. In most cases, an ARE in the 30 UTR plays a pivotal role in regulating the stability of the mRNA. For this reason, a major effort in the field is to identify the molecular mechanisms by which ARE‐mRNAs are stabilized. Stabilization of ARE‐mRNAs is dependent upon the activation of different signal transduction pathways. In T‐cells and mast cells, activation of the c‐jun N‐terminal kinase (JNK) pathway is required for the stabilization of IL‐2 and IL‐3 mRNA (Chen et al., 1998; Ming et al., 1998). In NIH3T3 cells, the phosphatidylinositol 3‐kinase (PI3K) pathway contributes to ARE‐mRNA stabilization (Ming et al., 2001). Induction of iNOS was linked to protein kinase Cd activation and may also require the JNK pathway (Carpenter et al., 2001;

14


Lahti et al., 2003). In most cases, however, the mitogen‐activated protein kinase p38 (p38‐MAPK) pathway plays a central role in stabilizing ARE‐ mRNAs. As shown in a variety of cell types, p38‐MAPK and the downstream MAPK‐activated protein kinase 2 (MK2) are required for the posttranscriptional induction of IL‐1b, IL‐3, IL‐6, IL‐11, TNFa, IFNg, GROa, GROb, IL‐8, IP‐10, and COX‐2 (Bamba et al., 2003; Brook et al., 2000; Dean et al., 1999; Holtmann et al., 1999; Kotlyarov et al., 1999; Lasa et al., 2000; Mavropoulos et al., 2005; Ming et al., 2001; Neininger et al., 2002; Ridley et al., 1998; Rousseau et al., 2002; Sirenko et al., 1997; Vockerodt et al., 2005; Winzen et al., 1999). Given that such a large number of pro‐inflammatory proteins are induced posttranscriptionally through the p38‐MAPK–MK2 axis, developing inhibitors of this pathway has been a major focus in the search for novel anti‐inflammatory drugs. Recent findings provide first insights into the molecular mechanism that links the p38‐MAPK–MK2 pathway to regulators of AMD. Again, the TTP/ BRF proteins appear to play a central role in this process. Both p38‐MAPK and MK2 phosphorylate TTP (Carballo et al., 2001; Mahtani et al., 2001; Zhu et al., 2001), and the direct phosphorylation of TTP by MK2 at serine 52 and serine 178 leads to complex formation between TTP and the adaptor protein 14‐3‐3 (Chrestensen et al., 2003; Johnson et al., 2002; Stoecklin et al., 2004). As a consequence of 14‐3‐3 binding, TTP activity is reduced and ARE‐mRNAs thereby stabilized (Stoecklin et al., 2004). Interestingly, the activity of BRF1 is also inhibited by phosphorylation and consecutive binding of 14‐3‐3 (Schmidlin et al., 2004). Although one study questions this model (Rigby et al., 2005), phosphorylation‐induced complex formation of the TTP/BRF proteins with 14‐3‐3 may be a general mechanism by which ARE‐mRNAs are stabilized. Another target of the p38‐MAPK pathway is the ARE‐BP hnRNP‐A0 (Rousseau et al., 2002). In LPS‐stimulated macrophages, MK2 phosphorylates hnRNP‐A0 at serine 84, but it is not clear whether this contributes to stabilization of ARE‐mRNAs. The anti‐inflammatory cytokine IL‐10 is a potent inhibitor of macrophage activation that reduces the expression of many cytokines (e.g., TNFa, IL‐1, IL‐6, GM‐CSF), several chemokines, and COX‐2 (Moore et al., 2001). Although most of the inhibitory effect of IL‐10 on the expression of cytokines appears to occur at the transcriptional level in a STAT3‐dependent manner (Takeda et al., 1999; Williams et al., 2004a), there is evidence that posttranscriptional mechanisms are also involved. In LPS‐treated macrophages, IL‐10 was found to reduce TNFa, IL‐1a, and IL‐1b mRNA levels without affecting their rates of transcription, indicating that IL‐10 accelerates the decay of these mRNAs (Bogdan et al., 1992). The suppressive effect of IL‐10 on TNFa expression requires the TNFa 30 UTR (Denys et al., 2002), and another study


15

indicates that IL‐10 reduces the translation of TNFa through inhibition of the p38‐MAPK–MK2 pathway (Kontoyiannis et al., 2001). Moreover, IL‐10 was found to destabilize GROa (murine KC), GM‐CSF, and G‐CSF, as well as its own mRNA (Biswas et al., 2003; Brown et al., 1996b; Kishore et al., 1999). Given that some of these results are contradictory, the contribution of posttranscriptional mechanisms to the suppressive effect of IL‐10 remains a controversial issue (Williams et al., 2004b). 2.5. Non‐ARE Decay Elements The ARE is probably the most common element that regulates mRNA stability, yet other regulatory elements have been described in a number of mRNAs encoding pro‐inflammatory proteins. For example, TNFa contains a second, constitutive decay element (CDE) in the 30 UTR, located downstream of the ARE. As opposed to AMD, CDE‐mediated mRNA decay is not inhibited by stimulation of macrophages with LPS, nor by activation of the PI3K or p38‐MAPK pathways (Stoecklin et al., 2003). The CDE may thus serve as a safeguard element that maintains stringent control over TNFa production, thereby reducing the risk of accumulating potentially injurious TNFa levels. G‐CSF is another cytokine that contains in its 30 UTR two independent destabilizing elements: an ARE and the stem‐loop destabilizing element (SLDE). The SLDE, similar to the CDE of TNFa, prevents stabilization of G‐CSF mRNA under conditions where AMD is inhibited (Brown et al., 1996a; Putland et al., 2002). The induction of IL‐2 expression in stimulated T‐cells results from both transcriptional activation and posttranscriptional mRNA stabilization (Lindsten et al., 1989; Musgrave et al., 2004). In unstimulated cells, the ARE in the 30 UTR mediates rapid decay of IL‐2 mRNA. Stabilization of IL‐2 mRNA occurs in response to JNK activation and requires, besides the ARE, a JNK‐response element (JRE) in the 50 UTR of the transcript (Chen et al., 1998). Two proteins bind to the JRE, nucleolin and the cold shock domain (CSD) protein YB1, and both are required for stabilizing IL‐2 mRNA (Chen et al., 2000). Hypoxia‐induced expression of VEGF, a major angiogenic factor, is regulated at both transcriptional and posttranscriptional levels (Ikeda et al., 1995; Levy et al., 1996; Stein et al., 1995). Under normoxic conditions, VEGF mRNA is labile. Three distinct destabilizing elements mediate rapid decay of the VEGF transcript: an ARE in the 30 UTR, a coding region determinant of instability (CDR), and an element in the 50 UTR (Dibbens et al., 1999). Interestingly, the three elements can independently mediate rapid decay of a reporter transcript, yet all three elements together are required to allow the mRNA to be stabilized in response to hypoxia (Dibbens et al., 1999). The

16


ARE‐BP HuR binds to the 30 UTR of VEGF mRNA and contributes to mRNA stabilization (Levy et al., 1998). A complex that contains YB1 and polypyrimidine tract binding protein (PTB) binds to two sites in the 50 UTR and one site in the 30 UTR, and participates in the stabilization of VEGF transcripts (Coles et al., 2004). These binding sites are very similar to the binding site of YB1 in the 50 UTR of IL‐2, indicating that YB1‐containing complexes may play a general role in stabilizing different mRNAs (Coles et al., 2004). YB1 was also found to enhance the stability of GM‐CSF mRNA, although in this case, YB1 appears to directly interact with the ARE of GM‐CSF (Capowski et al., 2001). The examples of TNFa, IL‐2, and VEGF illustrate that the ARE does not operate as an isolated element, but cooperates with other elements to achieve complex control over mRNA stability and translation. By recruitment of ARE‐ BPs together with other RNA‐binding proteins, multiple transcripts can be regulated specifically and expressed differentially according to cellular requirements and extracellular cues. We are still far away from understanding the complex interplay between the factors that contribute to the stability and translation efficiency of individual mRNAs. Future studies may reveal that combining non‐ARE regulatory elements with an ARE is a common mechanism of mRNAs that is regulated at the posttranscriptional level. 3. Translational Inhibition 3.1. Translational Silencing of ARE‐Transcripts Several RNA‐binding proteins repress the translation of mRNAs encoding proteins that regulate the inflammatory response. TIA‐1 and TIAR are closely related members of the RNA‐recognition motif (RRM) family of RNA‐binding proteins that inhibit the translation of TNFa transcripts in macrophages (Anderson and Kedersha, 2002a,b; Piecyk et al., 2000), but not in T lymphocytes (Saito et al., 2001). Although LPS‐activated macrophages derived from wild‐type and TIA‐1/ mice express similar amounts of TNFa transcripts, macrophages lacking TIA‐1 produce significantly more TNFa protein than wild‐type controls (Piecyk et al., 2000; Saito et al., 2001). In macrophages lacking TIA‐1, the percentage of TNFa transcripts found in polysomes is increased, suggesting that TIA‐1 functions as a translational silencer (Piecyk et al., 2000). The overexpression of TNFa protein in macrophages lacking TIA‐1 is strain dependent. TIA‐1/ macrophages derived from BALB/c mice produce 3–5 times more TNFa than wild‐type controls, whereas TIA‐1/ macrophages derived from C57BL/6 mice produce nearly 10 times more TNFa than wild‐type controls (Saito et al., 2001). Moreover, C57BL/6 mice


17

spontaneously develop mild arthritis (Phillips et al., 2004). Thus, unidentified genetic modifiers determine whether TIA‐1/ mice develop arthritis. TIA‐1 similarly represses the translation of COX‐2 (Dixon et al., 2003), an enzyme that converts arachidonic acid into pro‐inflammatory prostaglandins (Sundy, 2001; Schnitzer and Hochberg, 2002). Pharmacologic inhibitors of COX‐2 are potent anti‐inflammatory agents that significantly reduce the severity of inflammatory arthritis (Schnitzer and Hochberg, 2002). The 30 UTR of COX‐2 transcripts contains an ARE that recruits a multimeric protein complex that includes TIA‐1, TIAR, hnRNP U, and HuR (Cok et al., 2003). TIA‐1 null fibroblasts express 2–3 times more COX‐2 and 2 times more prostaglandin E2 than wild‐type controls (Dixon et al., 2003). Moreover, the expression of COX‐ 2 in colon cancer‐derived cell lines is inversely correlated with the expression of TIA‐1 (Dixon et al., 2003). In colon cancer cell lines expressing abundant TIA‐1, sucrose gradient analysis shows that COX‐2 mRNA is concentrated in low‐density fractions that contain untranslated mRNPs (Dixon et al., 2003). These results suggest that TIA‐1 can coordinately repress the expression of TNFa and COX‐2 to dampen the inflammatory response. TIAR represses the translation of MMP‐13 (Yu et al., 2003), a TNFa‐ induced collagenase that has been implicated in pro‐inflammatory, angiogenic, and destructive processes within the joints of patients with rheumatoid arthritis (Konttinen et al., 1999; Moore et al., 2000; Vincenti and Brinckerhoff, 2002; Wernicke et al., 2002). Although the mechanism of TIAR‐induced translational silencing has not been investigated, the fact that TIAR/ macrophages, like TIA‐1/ macrophages, overexpress TNFa (Piecyk et al., 2000) suggests that TIA‐1 and TIAR use similar mechanisms to inhibit protein translation. Interestingly, a 17 amino acid peptide derived from an alternatively spliced TIAR exon enhances the expression of MMP‐13 (Yu et al., 2003), suggesting that this exon plays an important role in TIAR function. The ability of TIA‐1 and TIAR to coordinately regulate the expression of several pro‐inflammatory proteins supports the concept that posttranscriptional mechanisms play a major role in regulating the inflammatory response. CUGBP2, an RNA‐binding protein that is structurally related to HuR, has also been implicated in the translational silencing of ARE‐containing transcripts (Mukhopadhyay et al., 2003). CUGBP2 binds to two distinct AREs found in the 30 UTR of COX‐2 mRNA. Paradoxically, expression of recombinant CUGBP2 stabilizes luciferase reporter transcripts bearing the COX‐2 ARE, but reduces expression of the luciferase protein. Differential centrifugation analysis revealed that luciferase transcripts are excluded from polysomes in cells transfected with CUGBP2, but not CUGBP1 or HuR. Thus CUGBP2, like TIA‐1 and TIAR, appears to function by preventing its associated transcripts from

18


moving to polysomes. It is not known whether CUGBP2, TIA‐1, and TIAR cooperate to repress the translation of ARE‐containing transcripts. The Fragile X‐related protein FXR1P has also been identified as a translational silencer that targets ARE‐containing transcripts (Garnon et al., 2005). The FXR family of RNA‐binding proteins (FMRP, FXR1P, and FXR2P) possesses two hnRNP K‐homology (KH) domains and an RGG box (Kaufmann et al., 2002). These proteins bind to both U‐rich and G‐rich sequences, and the RGG box allows FMRP to bind to the G‐quartet RNA motif (Darnell et al., 2001, 2004). Thus, the RNA specificity of the FXR proteins could be quite broad. FXR1P binds specifically to the ARE in the 30 UTR of TNFa mRNA, and FXR1P/ macrophages express approximately 2 times more TNFa than wild‐type macrophages (Garnon et al., 2005). Polysome profiles suggest that FXR1P, like TIA‐1 and TIAR, can exclude TNFa mRNA from polysomes. These results suggest that FXR1P, TIA‐1, and TIAR may work together to bring about translational silencing of TNFa transcripts. It remains to be determined whether FXR1P also represses the translation of additional pro‐inflammatory proteins. 3.2. Mechanisms of Translational Silencing Much of what we know about the mechanism of TIA‐1‐induced translational repression comes from studies of the general translational arrest triggered by environmental stress (e.g., heat, oxidative conditions, and energy deprivation) (Anderson and Kedersha, 2002a,b; Kedersha and Anderson, 2001). Stress‐ induced translational arrest is characterized by the activation of one or more members of a family of serine/threonine kinases (e.g., double‐stranded RNA‐ dependent protein kinase R, PKR‐like endoplasmic reticulum kinase, GCN2, and heme‐regulated inhibitor kinase) (Harding et al., 2000a,b; Sood et al., 2000; Williams, 1999). These kinases phosphorylate eIF2a, a component of the ternary complex that loads tRNAiMet onto the small ribosomal subunit to initiate protein synthesis (Dever, 2002). Phosphorylation of eIF2a inhibits protein translation by reducing the availability of active ternary complex (Krishnamoorthy et al., 2001). Under these conditions, TIA‐1 promotes the assembly of a non‐canonical translation‐initiation complex that is directed to discrete cytoplasmic foci known as stress granules (Anderson and Kedersha, 2002b; Kedersha et al., 1999, 2000, 2002). Like TIA‐1, the RNA‐binding proteins TIAR, FMRP, and FXR1P are concentrated at SGs, suggesting that these proteins might work together to bring about stress‐induced translational silencing. A non‐canonical translation initiation complex assembled during stress has reduced amounts of eIF2 and eIF5 (Kedersha et al., 2002). It is possible that the assembly of these non‐canonical translation initiation complexes contribute to


19

the reduced translation of ARE‐containing transcripts mediated by TIA‐1, TIAR, and FXR1P. These complexes may be unable to recruit the large ribosomal subunit, preventing the assembly of polysomes. Although stress granules are visible only when large numbers of transcripts are simultaneously subjected to translational arrest, the underlying translational control mechanism (i.e., assembly of stalled initiation complexes) regulates protein expression in both stressed and unstressed cells (Anderson and Kedersha, 2002a,b). 3.3. Translational Silencing of Non‐ARE Transcripts The translational silencing of ARE‐containing transcripts has the potential to coordinately regulate the expression of multiple pro‐inflammatory proteins. Translational silencing also regulates the expression of non‐ARE‐containing transcripts encoding proteins that control the inflammatory response. A wellcharacterized example is the translational regulation of 15‐LO, an enzyme that participates in the conversion of arachidonic acid into either leukotriene A4 (LTA4) or lipoxin A4 (LXA4) (Kuhn et al., 2002). Whereas LTA4 is converted into LTB4, a potent pro‐inflammatory lipid, LXA4 is an anti‐inflammatory lipid that promotes the resolution of inflammation (Serhan, 2001). Because of this functional dichotomy, 15‐LO has both pro‐ and anti‐inflammatory effects in a variety of experimental systems (Kuhn et al., 2002). During erythrocyte maturation, the mRNA encoding 15‐LO is expressed in the early stages of erythropoiesis, but 15‐LO protein is only expressed at a late stage of maturation (van Leyen et al., 1998). The developmental stage‐specific expression of 15‐LO is achieved by translational silencing in early erythroid progenitors. Translational silencing of 15‐LO mRNA is dependent upon a differentiation control element (DICE) in the 30 UTR (Ostareck et al., 2001). mRNP K/E1 binds hnRNP K and hnRNP E1, which prevent the recruitment of 60S ribosomal subunits to the translation initiation complex (Ostareck et al., 2001). Thus, translational silencing by TIA‐1, TIAR, and hnRNP K/E1 may all involve the exclusion of the 60S ribosomal subunit from the initiation complex. The effect of translational regulation of 15‐LO on the production of pro‐ and anti‐inflammatory lipids remains to be determined. 4. Animal Models 4.1. TNFDARE Knock‐In Mice The importance of posttranscriptional pathways in the regulation of inflammation has been dramatically demonstrated in mutant mice lacking these regulatory controls. In general, posttranscriptional control mechanisms dampen

20


the expression of pro‐inflammatory proteins, and thereby prevent the pathological overexpression of these potentially dangerous products. This is certainly true in the case of TNFa, in which transgene‐derived overexpression is sufficient to induce inflammatory arthritis (Keffer et al., 1991). The contribution of posttranscriptional mechanisms to the normal regulation of TNFa production was elegantly revealed by the analysis of knock‐in mice that lack the ARE in the 3’UTR of one of the TNFa alleles (TNFDARE). Macrophages and synovial fibroblasts from TNFDARE mice overexpress TNFa mRNA and protein, and these animals spontaneously develop inflammatory arthritis and inflammatory bowel disease (Kontoyiannis et al., 1999). Because the ARE has not been implicated in the transcriptional regulation of TNFa production, these results imply that posttranscriptional controls are essential for the prevention of spontaneous inflammatory disease. Importantly, the production of TNFa from TNFDARE macrophages is not affected by inhibitors of p38‐MAPK, confirming that the ARE is required for the ability of p38‐MAPK to stabilize and promote the translation of TNFa transcripts (Kontoyiannis et al., 1999). The development of inflammatory bowel disease in TNFDARE mice requires the function of CD8þ, but not CD4þ T lymphocytes (Kontoyiannis et al., 2002). The full spectrum of intestinal inflammation requires the Th1 cytokines IL‐12 and IFN‐g, but not the Th2 cytokine IL‐4 (Kontoyiannis et al., 2002). Restricted expression of TNFa by either macrophages or T lymphocytes is sufficient to induce intestinal inflammation (Kontoyiannis et al., 2002). Moreover, studies of bone marrow chimeras revealed that the presence of the TNFDARE mutation in either bone marrow‐derived or tissue stroma‐derived cells is sufficient to induce intestinal inflammation. It is therefore likely that redundant cellular pathways act downstream of TNFa to bring about intestinal inflammation. 4.2. TTP Knock‐Out Mice Studies of TNFDARE macrophages reveal that the ARE is a destabilizing element that promotes the degradation of TNFa transcripts. The decay of TNFa mRNA is dependent upon TTP, a zinc‐finger protein that targets ARE‐ containing transcripts to the mRNA degradation machinery (see Fig. 1). Mutant mice lacking TTP develop a syndrome of cachexia, arthritis, dermatitis, and autoimmunity that results from the pathological overexpression of TNFa (Taylor et al., 1996). The administration of neutralizing antibodies reactive with TNFa prevents all aspects of this syndrome, indicating that TNFa plays a central role in disease pathogenesis. The importance of TNFa was confirmed by breeding TTP‐deficient mice with mice lacking TNF‐Receptor (R)1 or TNFR2. Whereas TTP/ TNFR1/ mice did not develop cachexia or


21

arthritis, TTP/ TNFR2/ developed more severe arthritis than mice lacking TTP alone (Carballo and Blackshear, 2001). Thus, TNFa binds to TNFR1 to bring about the pathological features of this syndrome. In contrast, interactions with TNFR2 may prevent pathology. TTP/ mice also develop bone marrow and peripheral blood granulopoiesis, suggesting that TTP may regulate the production of granulocyte growth and/or survival factors. Interestingly, this phenotype was preserved in TTP/ mice lacking TNFR1, TNFR2, or both TNF receptors, suggesting that granulopoiesis is not mediated by overexpressed TNFa. The mechanism of granulopoiesis remains to be determined. Unlike TNFDARE mice, TTP/ mice do not develop inflammatory bowel disease (Taylor et al., 1996). It is not surprising that the phenotypes of TNFDARE and TTP/ mice are different. TNFDARE macrophages lack the ability to bind to several ARE‐binding proteins, including the TTP homologues BRF1 and BRF2, or translational silencers such as TIA‐1 and FXR1P. Thus, the phenotype of TNFDARE mice is likely to be more severe than that of TTP/ mice. Because TNFa transgenic mice develop arthritis, but not inflammatory bowel disease (Keffer et al., 1991; Kollias, 2004), the level of TNFa expression may not be sufficient to explain why bowel inflammation is observed in the TNFDARE strain. It is possible that tissue‐specific regulation of TNFa expression determines whether there is inflammation in the joints, the intestines, or both. 4.3. TIA‐1 and TIAR Knock‐Out Mice On the Balb/c background, mice lacking TIA‐1 do not develop spontaneous inflammatory disease. When bred onto the C57Bl/6 background, however, TIA‐1/ mice develop mild non‐erosive arthritis (Phillips et al., 2004). Thus, genetic modifiers are important in determining the disease phenotype. Mutant mice lacking both TTP and TIA‐1 develop spontaneous inflammatory arthritis that is significantly more severe than the arthritis observed in mice lacking either TIA‐1 or TTP alone (Phillips et al., 2004). Although macrophages derived from TIA‐1/ TTP/ mice overexpress TNFa mRNA, they express less TNFa protein than TIA‐1/ or TTP/ macrophages (Phillips et al., 2004). Whether this results from defective nuclear export of TNFa transcripts, defective translation, or defective processing of TNFa protein remains to be determined. The source of arthritigenic cytokine in mice lacking both TIA‐1 and TTP appears to be an expanded population of neutrophils found in the bone marrow and peripheral blood. The marked increase in neutrophils in TTP/ bone marrow and peripheral blood is potentiated in mice that also lack TIA‐1. The mechanism whereby TIA‐1 and TTP cooperate to increase the maturation and/or survival of neutrophils is not known. Whereas wild‐type neutrophils produce little or no

22


TNFa in response to LPS stimulation, neutrophils lacking TTP secrete significant amounts of TNFa in response to LPS (Phillips et al., 2004). It is therefore possible that neutrophils are an important source of this pro‐arthritic cytokine in mice lacking both TIA‐1 and TTP. Although neutrophils clearly contribute to the pathogenesis of arthritis (Edwards and Hallett, 1997; Pillinger and Abramson, 1995), our understanding of their precise contribution to the inflammatory process is incomplete. Neutrophils are the predominant cell type found in inflammatory synovial fluid (but not synovial pannus) derived from patients with rheumatoid arthritis. They are an important source of arachadonic acid‐derived inflammatory mediators (e.g., prostaglandins, leukotrienes, and lipoxins). In animal models of inflammatory arthritis, neutrophils are essential components of the inflammatory process (Jonsson et al., 2005; Lawlor et al., 2004; Wipke and Allen, 2001). The cooperative regulation of both neutrophil maturation and function by TIA‐1 and TTP suggests that these cells are a major source of inflammatory cytokine production in a subset of patients with rheumatoid arthritis. 4.4. G‐CSF and GM‐CSF Transgenic Mice The importance of neutrophils in inflammatory arthritis was dramatically demonstrated in mice lacking G‐CSF, an important regulator of neutrophil production. These mice exhibit impaired granulopoiesis and are protected from collagen‐induced arthritis (Lawlor et al., 2004). Moreover, administration of anti‐G‐CSF antibodies protects wild‐type mice from collagen‐induced arthritis (Lawlor et al., 2004). These results reveal that G‐CSF is required to produce the mature neutrophils which are essential effectors of synovitis. The 30 UTR of the G‐CSF mRNA encodes two regulatory elements that dampen the production of G‐CSF at the posttranscriptional level. Thus, G‐CSF may be one of the key cytokines whose posttranscriptional regulation can determine susceptibility to arthritis. Like G‐CSF, GM‐CSF contributes to the growth, differentiation, and function of granulocytes and macrophages. In both in vitro and in vivo studies, GM‐CSF has been shown to ‘‘prime’’ neutrophils and macrophages for their effector functions (Hamilton, 2002). In this capacity, GM‐CSF may play an important role in the inflammatory response. Consistent with this prediction, GM‐CSF/ mice are resistant to methylated BSA/IL‐1‐induced arthritis (Yang and Hamilton, 2001). Moreover, anti‐GM‐CSF antibodies reduce the severity of both collagen‐ and BSA/IL‐1‐induced arthritis (Cook et al., 2001; Yang and Hamilton, 2001). These results place GM‐CSF in the category of a pro‐inflammatory protein whose expression is regulated at the posttranscriptional level. The importance of posttranscriptional regulation of GM‐CSF


23

production was demonstrated in transgenic mice expressing recombinant GM‐CSF with or without the 30 ARE under control of a CMV promoter. At embryonic day 14, the ARE‐deleted transcript, but not the ARE‐containing transcript, was expressed in all tissues in which the CMV promoter was active (Houzet et al., 2001). Mice with the ARE‐deleted construct exhibit increased proliferation of granulocytes and macrophages. Taken together, these results implicate GM‐CSF as an important pro‐inflammatory cytokine subject to posttranscriptional control. 5. Clinical Significance The coordinate regulation of pro‐inflammatory proteins by posttranscriptional control pathways suggests that components of these pathways may be attractive targets for drug development. Indeed, chemical inhibitors of p38‐MAPK potently reduce the expression of pro‐inflammatory proteins in activated macrophages, and strongly support a role for posttranscriptional control in the regulation of inflammation (Pargellis and Regan, 2003). p38‐MAPK inhibitors reduce the expression of pro‐inflammatory cytokines and chemokines such as TNFa, IL‐1b, IL‐6, IL‐8, and MCP‐1. These drugs also reduce the expression of pro‐inflammatory enzymes such as COX‐2, iNOS, MMP‐1, MMP‐3, MMP‐9, and MMP‐13, as well as endothelial adhesion molecules such as E‐selectin, intracellular adhesion molecule (ICAM)‐1, and VCAM‐1 that recruit inflammatory cells to the rheumatoid synovium (Westra et al., 2004a,b, 2005). These drugs also inhibit the development of inflammatory arthritis in different animal models of arthritis (Pargellis and Regan, 2003). Several pharmaceutical companies are testing p38‐MAPK inhibitors in clinical trials involving patients with rheumatoid arthritis. In phase I clinical trials, single dose administration of these agents has been well tolerated (Parasrampuria et al., 2003). If the safety profile of these drugs is acceptable, they could become important treatments for a variety of immune‐mediated inflammatory diseases. MK2, a specific substrate for p38‐MAPK, has also been implicated in the posttranscriptional regulation of pro‐inflammatory protein expression. Macrophages derived from mutant mice lacking MK2 overexpress TNFa, IFNg, IL‐1, IL‐6, and nitric oxide following LPS activation (Kotlyarov et al., 1999). Thus, MK2 inhibitors, similar to p38‐MAPK inhibitors, have the potential to coordinately reduce the expression of several classes of proteins with pro‐inflammatory activity. The effects of MK2 on mRNA stability and translation are complex and poorly understood. In macrophages lacking MK2, the half‐life of IL‐6 mRNA is reduced more than 10 fold, whereas the half‐ life of TNFa mRNA is essentially unchanged (Neininger et al., 2002).

24


This could be due to the presence of a second decay element, the CDE, in the TNFa 30 UTR (Stoecklin et al., 2003). It remains to be determined whether differences in the spectrum of pro‐inflammatory proteins regulated by p38‐MAPK and MK2 will affect the clinical efficacy of these drugs. Comparative pre‐clinical and clinical evaluation of p38‐MAPK and MK2 inhibitors will be required to determine the potential for these drugs in the treatment of inflammation. A number of other compounds have also been shown to interfere with cytokine production at the posttranscriptional level. In macrophages, the radicicol analogue A inhibits IL‐1b, IL‐6, and TNFa production by accelerating mRNA decay in an ARE‐dependent manner (Kastelic et al., 1996). In monocytes, the cannabinoid ajulemic acid reduces LPS‐induced IL‐1b production, which may involve accelerated IL‐1b mRNA decay (Bidinger et al., 2003). In osteoblasts, a tetracycline derivative was shown to inhibit IL‐1b‐ induced IL‐6 production by accelerating decay of IL‐6 mRNA (Kirkwood et al., 2003). Thalidomide, once in use as a sedative but withdrawn from the market after its severe teratogenic effects were discovered, reduces TNFa and COX‐2 expression by accelerating the degradation of the corresponding mRNAs. Due to its anti‐inflammatory activity, thalidomide has been reintroduced for the treatment of erythema nodosum leprosum, aphtous stomatitis, Behcet syndrome, chronic cutaneous systemic lupus erythematosus, and graft‐versus‐host disease (Calabrese and Fleischer, 2000). Studies have also shown efficacy against aphthous ulcers and other HIV‐associated conditions including Kaposi sarcoma, as well as in the treatment of multiple myeloma. By interfering with posttranscriptional control mechanisms, these drugs have the capacity to dampen the expression of several pro‐inflammatory proteins simultaneously, and may therefore prove useful for the treatment of a variety of inflammatory conditions. The importance of posttranscriptional control mechanisms in the regulation of inflammation suggests that individual components of these regulatory pathways may influence disease susceptibility. Patients expressing single nucleotide polymorphisms that reduce the expression and/or function of TTP, TIA‐1, TIAR, and FXR1P could have an increased risk of developing inflammatory arthritis or inflammatory bowel disease. Sequence analysis has identified 13 polymorphisms in the coding regions of TTP and its two homologues BRF1 and BRF2 (Blackshear et al., 2003b). Six of these mutations would result in amino acid changes that could alter protein function. Another mutation was a dinucleotide substitution that would prevent splicing of the single intron in ZFP36L1, the gene encoding for BRF1. Analysis of lymphoblasts from this individual confirmed that the expression of ZFP36L1 mRNA was reduced by 50%. Case control association studies will be required to determine whether


25

any of these gene polymorphisms are associated with increased susceptibility to immune‐mediated inflammatory disease. 6. Conclusions Posttranscriptional regulation of mRNA stability and translation are major determinants of pro‐inflammatory protein expression. An ARE in the 30 UTR of a large number of mRNAs encoding pro‐inflammatory proteins plays a central role by recruiting a specific set of ARE‐BPs, which exert their functions by (i) repressing translation of the bound mRNA, (ii) targeting the mRNA for rapid degradation, or (iii) stabilizing the mRNA in response to extracellular cues. In addition, non‐ARE elements participate in regulating stability and translation of certain mRNAs, and provide a further level of complexity. A molecular understanding of the underlying mechanisms will help to characterize new targets for the development of a next generation of anti‐inflammatory drugs. Genetic polymorphisms within the genes encoding posttranscriptional regulatory proteins are likely to identify individuals with enhanced susceptibility to immune‐mediated inflammatory diseases. As our understanding of posttranscriptional regulatory pathways is still rudimentary, this growing area of research will continue to provide important insights into the fundamental principles that determine gene expression. Much remains to be learned. It is not clear how posttranscriptional control pathways discriminate between the many transcripts that contain regulatory elements in their 30 UTRs. Gene array analysis of mRNAs that co‐precipitate with TIA‐1 or TIAR has revealed that these RNA‐binding proteins can associate with thousands of transcripts (Lopez de Silanes et al., 2005; M. Gorospe, personal communication). Yet functional studies suggest that the mRNAs regulated by these proteins may be much more restricted. It is likely that multiple RNA‐binding proteins expressed in individual cells work together to select mRNAs that are subject to post‐ transcriptional regulation. We know very little about how different ARE‐PBs cooperatively regulate mRNA stability and translation. The potential participation of microRNAs provides further opportunities for fine‐tuning the mRNA selection process. Defining the complex interplay of all factors associated with ARE and non‐ARE elements will help us to fully understand the posttranscriptional operon that governs the expression of pro‐inflammatory proteins. Acknowledgments We would like to thank Nancy Kedersha (Brigham and Women’s Hospital, Boston) for helpful comments on the manuscript. PA was supported by National Institute of Health grants AI‐33600 and AI‐50167.

26


References Akashi, M., Loussararian, A. H., Adelman, D. C., Saito, M., and Koeffler, H. P. (1990). Role of lymphotoxin in expression of interleukin 6 in human fibroblasts. Stimulation and regulation. J. Clin. Invest. 85, 121–129. Anderson, P., and Kedersha, N. (2002a). Stressful initiations. J. Cell Sci. 115, 3227–3234. Anderson, P., and Kedersha, N. (2002b). Visibly stressed: The role of eIF2, TIA‐1, and stress granules in protein translation. Cell Stress Chaperones 7, 213–221. Antic, D., and Keene, J. D. (1997). Embryonic lethal abnormal visual RNA‐binding proteins involved in growth, differentiation, and posttranscriptional gene expression. Am. J. Hum. Genet. 61, 273–278. Atasoy, U., Curry, S. L., Lopez de Silanes, I., Shyu, A. B., Casolaro, V., Gorospe, M., and Stellato, C. (2003). Regulation of eotaxin gene expression by TNF‐alpha and IL‐4 through mRNA stabilization: Involvement of the RNA‐binding protein HuR. J. Immunol. 171, 4369–4378. Bakheet, T., Williams, B. R., and Khabar, K. S. (2003). ARED 2.0: An update of AU‐rich element mRNA database. Nucleic Acids Res. 31, 421–423. Bamba, S., Andoh, A., Yasui, H., Makino, J., Kim, S., and Fujiyama, Y. (2003). Regulation of IL‐11 expression in intestinal myofibroblasts: Role of c‐Jun AP‐1‐ and MAPK‐dependent pathways. Am. J. Physiol. Gastrointest. Liver Physiol. 285, G529–G538. Beckwith, J. R. (1967). Regulation of the lac operon. Recent studies on the regulation of lactose metabolism in Escherichia coli support the operon model. Science 156, 597–604. Bickel, M., Cohen, R. B., and Pluznik, D. H. (1990). Post‐transcriptional regulation of granulocyte‐ macrophage colony‐stimulating factor synthesis in murine T cells. J. Immunol. 145, 840–845. Bidinger, B., Torres, R., Rossetti, R. G., Brown, L., Beltre, R., Burstein, S., Lian, J. B., Stein, G. S., and Zurier, R. B. (2003). Ajulemic acid, a nonpsychoactive cannabinoid acid, induces apoptosis in human T lymphocytes. Clin. Immunol. 108, 95–102. Biswas, R., Datta, S., Gupta, J. D., Novotny, M., Tebo, J., and Hamilton, T. A. (2003). Regulation of chemokine mRNA stability by lipopolysaccharide and IL‐10. J. Immunol. 170, 6202–6208. Blackshear, P. J. (2002). Tristetraprolin and other CCCH tandem zinc‐finger proteins in the regulation of mRNA turnover. Biochem. Soc. Trans. 30, 945–952. Blackshear, P. J., Lai, W. S., Kennington, E. A., Brewer, G., Wilson, G. M., Guan, X., and Zhou, P. (2003a). Characteristics of the interaction of a synthetic human tristetraprolin tandem zinc finger peptide with AU‐rich element‐containing RNA substrates. J. Biol. Chem. 278, 19947–19955. Blackshear, P. J., Phillips, R. S., Vazquez‐Matias, J., and Mohrenweiser, H. (2003b). Polymorphisms in the genes encoding members of the tristetraprolin family of human tandem CCCH zinc finger proteins. Prog. Nucleic Acid Res. Mol. Biol. 75, 43–68. Bogdan, C., Paik, J., Vodovotz, Y., and Nathan, C. (1992). Contrasting mechanisms for suppression of macrophage cytokine release by transforming growth factor‐beta and interleukin‐10. J. Biol. Chem. 267, 23301–23308. Brewer, G. (1991). An AþU‐rich element RNA‐binding factor regulates c‐myc stability in vitro. Mol. Cell. Biol. 11, 2460–2466. Briata, P., Ilengo, C., Corte, G., Moroni, C., Rosenfeld, M. G., Chen, C. Y., and Gherzi, R. (2003). The Wnt/beta‐catenin–>Pitx2 pathway controls the turnover of Pitx2 and other unstable mRNAs. Mol. Cell 12, 1201–1211. Brook, M., Sully, G., Clark, A. R., and Saklatvala, J. (2000). Regulation of tumour necrosis factor alpha mRNA stability by the mitogen‐activated protein kinase p38 signalling cascade. FEBS Lett. 483, 57–61.


27

Brown, C. Y., Lagnado, C. A., and Goodall, G. J. (1996a). A cytokine mRNA‐destabilizing element that is structurally and functionally distinct from A þ U‐rich elements. Proc. Natl. Acad. Sci. USA 93, 13721–13725. Brown, C. Y., Lagnado, C. A., Vadas, M. A., and Goodall, G. J. (1996b). Differential regulation of the stability of cytokine mRNAs in lipopolysaccharide‐activated blood monocytes in response to interleukin‐10. J. Biol. Chem. 271, 20108–20112. Bufler, P., Gamboni‐Robertson, F., Azam, T., Kim, S. H., and Dinarello, C. A. (2004). Interleukin‐1 homologues IL‐1F7b and IL‐18 contain functional mRNA instability elements within the coding region responsive to lipopolysaccharide. Biochem. J. 381, 503–510. Calabrese, L., and Fleischer, A. B. (2000). Thalidomide: Current and potential clinical applications. Am. J. Med. 108, 487–495. Cao, H., Tuttle, J. S., and Blackshear, P. J. (2004). Immunological characterization of tristetraprolin as a low abundance, inducible, stable cytosolic protein. J. Biol. Chem. 279, 21489–21499. Capowski, E. E., Esnault, S., Bhattacharya, S., and Malter, J. S. (2001). Y box‐binding factor promotes eosinophil survival by stabilizing granulocyte‐macrophage colony‐stimulating factor mRNA. J. Immunol. 167, 5970–5976. Caput, D., Beutler, B., Hartog, K., Thayer, R., Brown‐Shimer, S., and Cerami, A. (1986). Identification of a common nucleotide sequence in the 30 ‐untranslated region of mRNA molecules specifying inflammatory mediators. Proc. Natl. Acad. Sci. USA 83, 1670–1674. Carballo, E., and Blackshear, P. J. (2001). Roles of tumor necrosis factor‐alpha receptor subtypes in the pathogenesis of the tristetraprolin‐deficiency syndrome. Blood 98, 2389–2395. Carballo, E., Cao, H., Lai, W. S., Kennington, E. A., Campbell, D., and Blackshear, P. J. (2001). Decreased sensitivity of tristetraprolin‐deficient cells to p38 inhibitors suggests the involvement of tristetraprolin in the p38 signaling pathway. J. Biol. Chem. 276, 42580–42587. Carballo, E., Lai, W. S., and Blackshear, P. J. (1998). Feedback inhibition of macrophage tumor necrosis factor‐alpha production by tristetraprolin. Science 281, 1001–1005. Carballo, E., Lai, W. S., and Blackshear, P. J. (2000). Evidence that tristetraprolin is a physiological regulator of granulocyte‐macrophage colony‐stimulating factor messenger RNA deadenylation and stability. Blood 95, 1891–1899. Carpenter, L., Cordery, D., and Biden, T. J. (2001). Protein kinase Cdelta activation by interleukin‐ 1beta stabilizes inducible nitric‐oxide synthase mRNA in pancreatic beta‐cells. J. Biol. Chem. 276, 5368–5374. Chen, C.‐Y. A., and Shyu, A.‐B. (1995). AU‐rich elements: Characterization and importance in mRNA degradation. Trends Biochem. Sci. 20, 465–470. Chen, C. Y., Del Gatto‐Konczak, F., Wu, Z., and Karin, M. (1998). Stabilization of interleukin‐ 2 mRNA by the c‐Jun NH2‐terminal kinase pathway. Science 280, 1945–1949. Chen, C. Y., Gherzi, R., Andersen, J. S., Gaietta, G., Jurchott, K., Royer, H. D., Mann, M., and Karin, M. (2000). Nucleolin and YB‐1 are required for JNK‐mediated interleukin‐2 mRNA stabilization during T‐cell activation. Genes Dev. 14, 1236–1248. Chen, C. Y., Gherzi, R., Ong, S. E., Chan, E. L., Raijmakers, R., Pruijn, G. J., Stoecklin, G., Moroni, C., Mann, M., and Karin, M. (2001). AU Binding Proteins Recruit the Exosome to Degrade ARE‐Containing mRNAs. Cell 107, 451–464. Chen, C. Y., Xu, N., and Shyu, A. B. (1995). mRNA decay mediated by two distinct AU‐rich elements from c‐fos and granulocyte‐macrophage colony‐stimulating factor transcripts: Different deadenylation kinetics and uncoupling from translation. Mol. Cell. Biol. 15, 5777–5788. Chen, C. Y., Xu, N., and Shyu, A. B. (2002). Highly selective actions of HuR in antagonizing AU‐rich element‐mediated mRNA destabilization. Mol. Cell. Biol. 22, 7268–7278.

28


Chrestensen, C. A., Schroeder, M. J., Shabanowitz, J., Hunt, D. F., Pelo, J. W., Worthington, M. T., and Sturgill, T. W. (2004). MAPKAP kinase 2 phosphorylates tristetraprolin on in vivo sites including Ser 178, a site required for 14‐3‐3 binding. J. Biol. Chem. 279, 10174–1084. Cok, S. J., Acton, S. J., and Morrison, A. R. (2003). The proximal region of the 30 ‐untranslated region of cyclooxygenase‐2 is recognized by a multimeric protein complex containing HuR, TIA‐ 1, TIAR, and the heterogeneous nuclear ribonucleoprotein U. J. Biol. Chem. 278, 36157–36162. Coles, L. S., Bartley, M. A., Bert, A., Hunter, J., Polyak, S., Diamond, P., Vadas, M. A., and Goodall, G. J. (2004). A multi‐protein complex containing cold shock domain (Y‐box) and polypyrimidine tract binding proteins forms on the vascular endothelial growth factor mRNA. Potential role in mRNA stabilization. Eur. J. Biochem. 271, 648–660. Cook, A. D., Braine, E. L., Campbell, I. K., Rich, M. J., and Hamilton, J. A. (2001). Blockade of collagen‐induced arthritis post‐onset by antibody to granulocyte‐macrophage colony‐stimulating factor (GM‐CSF): Requirement for GM‐CSF in the effector phase of disease. Arthritis Res. 3, 293–298. Cougot, N., Babajko, S., and Seraphin, B. (2004a). Cytoplasmic foci are sites of mRNA decay in human cells. J. Cell Biol. 165, 31–40. Cougot, N., van Dijk, E., Babajko, S., and Seraphin, B. (2004b). ‘Cap‐tabolism’. Trends Biochem. Sci. 29, 436–444. Croft, D., McIntyre, P., Wibulswas, A., and Kramer, I. (1999). Sustained elevated levels of VCAM‐1 in cultured fibroblast‐like synoviocytes can be achieved by TNF‐alpha in combination with either IL‐4 or IL‐13 through increased mRNA stability. Am. J. Pathol. 154, 1149–1158. Darnell, J. C., Jensen, K. B., Jin, P., Brown, V., Warren, S. T., and Darnell, R. B. (2001). Fragile X mental retardation protein targets G quartet mRNAs important for neuronal function. Cell 107, 489–499. Darnell, J. C., Warren, S. T., and Darnell, R. B. (2004). The fragile X mental retardation protein, FMRP, recognizes G‐quartets. Ment. Retard. Dev. Disabil. Res. Rev. 10, 49–52. Dean, J. L., Brook, M., Clark, A. R., and Saklatvala, J. (1999). p38 mitogen‐activated protein kinase regulates cyclooxygenase‐2 mRNA stability and transcription in lipopolysaccharide‐treated human monocytes. J. Biol. Chem. 274, 264–269. Dean, J. L., Sully, G., Wait, R., Rawlinson, L., Clark, A. R., and Saklatvala, J. (2002). Identification of a novel AU‐rich‐element‐binding protein which is related to AUF1. Biochem. J. 366, 709–719. Denys, A., Udalova, I. A., Smith, C., Williams, L. M., Ciesielski, C. J., Campbell, J., Andrews, C., Kwaitkowski, D., and Foxwell, B. M. (2002). Evidence for a dual mechanism for IL‐10 suppression of TNF‐alpha production that does not involve inhibition of p38 mitogen‐activated protein kinase or NF‐kappa B in primary human macrophages. J. Immunol. 168, 4837–4845. Dever, T. E. (2002). Gene‐specific regulation by general translation factors. Cell 108, 545–556. Dibbens, J. A., Miller, D. L., Damert, A., Risau, W., Vadas, M. A., and Goodall, G. J. (1999). Hypoxic regulation of vascular endothelial growth factor mRNA stability requires the cooperation of multiple RNA elements. Mol. Biol. Cell 10, 907–919. Di Marco, S., Mazroui, R., Dallaire, P., Chittur, S., Tenenbaum, S. A., Radzioch, D., Marette, A., and Gallouzi, I. E. (2005). NF‐k B‐mediated MyoD decay during muscle wasting requires nitric oxide synthase mRNA stabilization, HuR protein, and nitric oxide release. Mol. Cell. Biol. 25, 6533–6545. Dixon, D. A., Balch, G. C., Kedersha, N., Anderson, P., Zimmerman, G. A., Beauchamp, R. D., and Prescott, S. M. (2003). Regulation of cyclooxygenase‐2 expression by the translational silencer TIA‐1. J. Exp. Med. 198, 475–481. Edwards, S., and Hallett, M. (1997). Seeing the wood for the trees: The forgotten role of neutrophils in rheumatoid arthritis. Immunol. Today 18, 320–324.


29

Esnault, S., and Malter, J. S. (1999). Primary peripheral blood eosinophils rapidly degrade transfected granulocyte‐macrophage colony‐stimulating factor mRNA. J. Immunol. 163, 5228–5234. Fan, X. C., and Steitz, J. A. (1998). Overexpression of HuR, a nuclear‐cytoplasmic shuttling protein, increases the in vivo stability of ARE‐containing mRNAs. EMBO J. 17, 3448–3460. Fenton, M. J., Vermeulen, M. W., Clark, B. D., Webb, A. C., and Auron, P. E. (1988). Human pro‐ IL‐1 beta gene expression in monocytic cells is regulated by two distinct pathways. J. Immunol. 140, 2267–2273. Fillman, C., and Lykke‐Andersen, J. (2005). RNA decapping inside and outside of processing bodies. Curr. Opin. Cell Biol. 17, 326–331. Ford, L. P., Watson, J., Keene, J. D., and Wilusz, J. (1999). ELAV proteins stabilize deadenylated intermediates in a novel in vitro mRNA deadenylation/degradation system. Genes Dev. 13, 188–201. Gao, M., Wilusz, C. J., Peltz, S. W., and Wilusz, J. (2001). A novel mRNA‐decapping activity in HeLa cytoplasmic extracts is regulated by AU‐rich elements. EMBO J. 20, 1134–1143. Garnon, J., Lachance, C., Di Marco, S., Hel, Z., Marion, D., Ruiz, M. C., Newkirk, M. M., Khandjian, E. W., and Radzioch, D. (2005). Fragile X‐related protein FXR1P regulates proinflammatory cytokine tumor necrosis factor expression at the post‐transcriptional level. J. Biol. Chem. 280, 5750–5763. Gerber, A. P., Herschlag, D., and Brown, P. O. (2004). Extensive association of functionally and cytotopically related mRNAs with Puf family RNA‐binding proteins in yeast. PLoS Biol. 2, E79. Gherzi, R., Lee, K. Y., Briata, P., Wegmuller, D., Moroni, C., Karin, M., and Chen, C. Y. (2004). A KH domain RNA binding protein, KSRP, promotes ARE‐directed mRNA turnover by recruiting the degradation machinery. Mol. Cell 14, 571–583. Grosset, C., Boniface, R., Duchez, P., Solanilla, A., Cosson, B., and Ripoche, J. (2004). In vivo studies of translational repression mediated by the granulocyte‐macrophage colony‐stimulating factor AU‐rich element. J. Biol. Chem. 279, 13354–13362. Hamilton, J. A. (2002). GM‐CSF in inflammation and autoimmunity. Trends Immunol. 23, 403–408. Han, J., Brown, T., and Beutler, B. (1990). Endotoxin‐responsive sequences control cachectin/ tumor necrosis factor biosynthesis at the translational level. J. Exp. Med. 171, 465–475. Harding, H., Novoa, Y., Zhang, H., Zeng, R., Wek, M., Schapira, M., and Ron, D. (2000a). Regulated translation initiation controls stress‐induced gene expression in mammalian cells. Mol. Cell 6, 1099–1108. Harding, H. P., Zhang, Y., Bertolotti, A., Zeng, H., and Ron, D. (2000b). Perk is essential for translational regulation and cell survival during the unfolded protein response. Mol. Cell 5, 897–904. Hodge, D. L., Martinez, A., Julias, J. G., Taylor, L. S., and Young, H. A. (2002). Regulation of nuclear gamma interferon gene expression by interleukin 12 (IL‐12) and IL‐2 represents a novel form of posttranscriptional control. Mol. Cell. Biol. 22, 1742–1753. Holtmann, H., Winzen, R., Holland, P., Eickemeier, S., Hoffmann, E., Wallach, D., Malinin, N. L., Cooper, J. A., Resch, K., and Kracht, M. (1999). Induction of interleukin‐8 synthesis integrates effects on transcription and mRNA degradation from at least three different cytokine‐ or stress‐ activated signal transduction pathways. Mol. Cell. Biol. 19, 6742–6753. Houzet, L., Morello, D., Defrance, P., Mercier, P., Huez, G., and Kruys, V. (2001). Regulated control by granulocyte‐macrophage colony‐stimulating factor AU‐rich element during mouse embryogenesis. Blood 98, 1281–1288. Hudson, B. P., Martinez‐Yamout, M. A., Dyson, H. J., and Wright, P. E. (2004). Recognition of the mRNA AU‐rich element by the zinc finger domain of TIS11d. Nat. Struct. Mol. Biol. 11, 257–264.

30


Huwiler, A., Akool el, S., Aschrafi, A., Hamada, F. M., Pfeilschifter, J., and Eberhardt, W. (2003). ATP potentiates interleukin‐1 beta‐induced MMP‐9 expression in mesangial cells via recruitment of the ELAV protein HuR. J. Biol. Chem. 278, 51758–51769. Iademarco, M. F., Barks, J. L., and Dean, D. C. (1995). Regulation of vascular cell adhesion molecule‐1 expression by IL‐4 and TNF‐alpha in cultured endothelial cells. J. Clin. Invest. 95, 264–271. Ikeda, E., Achen, M. G., Breier, G., and Risau, W. (1995). Hypoxia‐induced transcriptional activation and increased mRNA stability of vascular endothelial growth factor in C6 glioma cells. J. Biol. Chem. 270, 19761–19766. Ingelfinger, D., Arndt‐Jovin, D. J., Luhrmann, R., and Achsel, T. (2002). The human LSm1‐7 proteins colocalize with the mRNA‐degrading enzymes Dcp1/2 and Xrnl in distinct cytoplasmic foci. RNA 8, 1489–1501. Jing, Q., Huang, S., Guth, S., Zarubin, T., Motoyama, A., Chen, J., Di Padova, F., Lin, S. C., Gram, H., and Han, J. (2005). Involvement of microRNA in AU‐rich element‐mediated mRNA instability. Cell 120, 623–634. Johnson, B. A., and Blackwell, T. K. (2002). Multiple tristetraprolin sequence domains required to induce apoptosis and modulate responses to TNFalpha through distinct pathways. Oncogene 21, 4237–4246. Johnson, B. A., Geha, M., and Blackwell, T. K. (2000). Similar but distinct effects of the tristetraprolin/TIS11 immediate‐early proteins on cell survival. Oncogene 19, 1657–1664. Johnson, B. A., Stehn, J. R., Yaffe, M. B., and Blackwell, T. K. (2002). Cytoplasmic localization of Tristetraprolin involves 14–3‐3‐dependent and ‐independent mechanisms. J. Biol. Chem. 277, 18029–18036. Jonsson, H., Allen, P., and Peng, S. L. (2005). Inflammatory arthritis requires Foxo3a to prevent Fas ligand‐induced neutrophil apoptosis. Nat. Med. 11, 666–671. Kastelic, T., Schnyder, J., Leutwiler, A., Traber, R., Streit, B., Niggli, H., Mac Kenzie, A., and Cheneval, D. (1996). Induction of rapid IL‐1 beta mRNA degradation in THP‐1 cells mediated through the AU‐rich region in the 30 UTR by a radicicol analogue. Cytokine 8, 751–761. Kaufmann, W. E., Cohen, S., Sun, H. T., and Ho, G. (2002). Molecular phenotype of Fragile X syndrome: FMRP, FXRPs, and protein targets. Microsc. Res. Tech. 57, 135–144. Kedersha, N., and Anderson, P. (2001). Stress granules: Sites of mRNA triage that regulate mRNA stability and translatability. Biochem. Soc. Trans 30, 963–969. Kedersha, N., Chen, S., Gilks, N., Li, W., Miller, I. J., Stahl, J., and Anderson, P. (2002). Evidence that ternary complex (eIF2‐GTP‐tRNA(i)(Met))‐deficient preinitiation complexes are core constituents of mammalian stress granules. Mol. Biol. Cell 13, 195–210. Kedersha, N., Cho, M. R., Li, W., Yacono, P. W., Chen, S., Gilks, N., Golan, D. E., and Anderson, P. (2000). Dynamic shuttling of TIA‐1 accompanies the recruitment of mRNA to mammalian stress granules. J. Cell Biol. 151, 1257–1268. Kedersha, N. L., Gupta, M., Li, W., Miller, I., and Anderson, P. (1999). RNA‐binding proteins TIA‐ 1 and TIAR link the phosphorylation of eIF‐2 alpha to the assembly of mammalian stress granules. J. Cell Biol. 147, 1431–1442. Kedersha, N., Stoecklin, G., Ayodele, M., Yacono, P., Lykke‐Andersen, J., Fitzler, M. J., Scheuner, D., Kaufman, R. J., Golan, D. E., and Andersen, P. (2005). Stress granules and processing bodies are dynamically linked sites of mRNP remodeling. J. Cell Biol. 169, 871–884. Keene, J. D., and Tenenbaum, S. A. (2002). Eukaryotic mRNPs may represent posttranscriptional operons. Mol. Cell 9, 1161–1167. Keffer, J., Probert, L., Cazlaris, H., Georgopoulos, S., Kaslaris, E., Kioussis, D., and Kollias, G. (1991). Transgenic mice expressing human tumour necrosis factor: A predictive genetic model of arthritis. EMBO J. 10, 4025–4031.


31

Kirkwood, K., Martin, T., Andreadis, S. T., and Kim, Y. J. (2003). Chemically modified tetracyclines selectively inhibit IL‐6 expression in osteoblasts by decreasing mRNA stability. Biochem. Pharmacol. 66, 1809–1819. Kishore, R., Tebo, J. M., Kolosov, M., and Hamilton, T. A. (1999). Cutting edge: Clustered AU‐rich elements are the target of IL‐10‐mediated mRNA destabilization in mouse macrophages. J. Immunol. 162, 2457–2461. Koeffler, H. P., Gasson, J., and Tobler, A. (1988). Transcriptional and posttranscriptional modulation of myeloid colony‐stimulating factor expression by tumor necrosis factor and other agents. Mol. Cell. Biol. 8, 3432–3438. Kolios, G., Valatas, V., and Ward, S. G. (2004). Nitric oxide in inflammatory bowel disease: A universal messenger in an unsolved puzzle. Immunology 113, 427–437. Kollias, G. (2004). Modeling the function of tumor necrosis factor in immune pathophysiology. Autoimmun. Rev. 3(Suppl. 1), S24–S25. Kontoyiannis, D., Boulougouris, G., Manoloukos, M., Armaka, M., Apostolaki, M., Pizarro, T., Kotlyarov, A., Forster, I., Flavell, R., Gaestel, M., Tsichlis, P., Cominelli, F., and Kollias, G. (2002). Genetic dissection of the cellular pathways and signaling mechanisms in modeled tumor necrosis factor‐induced Crohn’s‐like inflammatory bowel disease. J. Exp. Med. 196, 1563–1574. Kontoyiannis, D., Kotlyarov, A., Carballo, E., Alexopoulou, L., Blackshear, P. J., Gaestel, M., Davis, R., Flavell, R., and Kollias, G. (2001). Interleukin‐10 targets p38 MAPK to modulate ARE‐ dependent TNF mRNA translation and limit intestinal pathology. EMBO J. 20, 3760–3770. Kontoyiannis, D., Pasparakis, M., Pizarro, T. T., Cominelli, F., and Kollias, G. (1999). Impaired on/off regulation of TNF biosynthesis in mice lacking TNF AU‐rich elements: Implications for joint and gut‐associated immunopathologies. Immunity 10, 387–398. Konttinen, Y. T., Salo, T., Hanemaaijer, R., Valleala, H., Sorsa, T., Sutinen, M., Ceponis, A., Xu, J. W., Santavirta, S., Teronen, O., and Lopez‐Otin, C. (1999). Collagenase‐3 (MMP‐13) and its activators in rheumatoid arthritis: Localization in the pannus‐hard tissue junction and inhibition by alendronate. Matrix Biol. 18, 401–412. Kotlyarov, A., Neininger, A., Schubert, C., Eckert, R., Birchmeier, C., Volk, H. D., and Gaestel, M. (1999). MAPKAP kinase 2 is essential for LPS‐induced TNF‐alpha biosynthesis. Nat. Cell Biol. 1, 94–97. Krishnamoorthy, T., Pavitt, G. D., Zhang, F., Dever, T. E., and Hinnebusch, A. G. (2001). Tight binding of the phosphorylated alpha subunit of initiation factor 2 (eIF2alpha) to the regulatory subunits of guanine nucleotide exchange factor eIF2B is required for inhibition of translation initiation. Mol. Cell. Biol. 21, 5018–5030. Kuhn, H., Walther, M., and Kuban, R. J. (2002). Mammalian arachidonate 15‐lipoxygenases structure, function, and biological implications. Prostaglandins Other Lipid Mediat. 68–69, 263–290. Kullmann, M., Gopfert, U., Siewe, B., and Hengst, L. (2002). ELAV/Hu proteins inhibit p27 translation via an IRES element in the p27 50 UTR. Genes Dev. 16, 3087–3099. Lagnado, C. A., Brown, C. Y., and Goodall, G. J. (1994). AUUUA is not sufficient to promote poly (A) shortening and degradation of an mRNA: The functional sequence within AU‐rich elements may be UUAUUUA(U/A)(U/A). Mol. Cell. Biol. 14, 7984–7995. Lahti, A., Jalonen, U., Kankaanranta, H., and Moilanen, E. (2003). c‐Jun NH2‐terminal kinase inhibitor anthra(1,9‐cd)pyrazol‐6(2H)‐one reduces inducible nitric‐oxide synthase expression by destabilizing mRNA in activated macrophages. Mol. Pharmacol. 64, 308–315. Lai, W. S., Carballo, E., Strum, J. R., Kennington, E. A., Phillips, R. S., and Blackshear, P. J. (1999). Evidence that tristetraprolin binds to AU‐rich elements and promotes the deadenylation and destabilization of tumor necrosis factor alpha mRNA. Mol. Cell. Biol. 19, 4311–4323.

32


Lai, W. S., Carballo, E., Thorn, J. M., Kennington, E. A., and Blackshear, P. J. (2000). Interactions of CCCH zinc finger proteins with mRNA. Binding of tristetraprolin‐related zinc finger proteins to Au‐rich elements and destabilization of mRNA. J. Biol. Chem. 275, 17827–17837. Lai, W. S., Kennington, E. A., and Blackshear, P. J. (2003). Tristetraprolin and its family members can promote the cell‐free deadenylation of AU‐rich element‐containing mRNAs by poly(A) ribonuclease. Mol. Cell. Biol. 23, 3798–3812. Lal, A., Mazan‐Mamczarz, K., Kawai, T., Yang, X., Martindale, J. L., and Gorospe, M. (2004). Concurrent versus individual binding of HuR and AUF1 to common labile target mRNAs. EMBO J. 23, 3092–3102. Lasa, M., Mahtani, K. R., Finch, A., Brewer, G., Saklatvala, J., and Clark, A. R. (2000). Regulation of cyclooxygenase 2 mRNA stability by the mitogen‐activated protein kinase p38 signaling cascade. Mol. Cell. Biol. 20, 4265–4274. Lawlor, K. E., Campbell, I. K., Metcalf, D., O’Donnell, K., van Nieuwenhuijze, A., Roberts, A. W., and Wicks, I. P. (2004). Critical role for granulocyte colony‐stimulating factor in inflammatory arthritis. Proc. Natl. Acad. Sci. USA 101, 11398–11403. Levy, A. P., Levy, N. S., and Goldberg, M. A. (1996). Post‐transcriptional regulation of vascular endothelial growth factor by hypoxia. J. Biol. Chem. 271, 2746–2753. Levy, N. S., Chung, S., Furneaux, H., and Levy, A. P. (1998). Hypoxic stabilization of vascular endothelial growth factor mRNA by the RNA‐binding protein HuR. J. Biol. Chem. 273, 6417–6423. Lindsten, T., June, C. H., Ledbetter, J. A., Stella, G., and Thompson, G. B. (1989). Regulation of lymphokine messenger RNA stability by a surface‐mediated T cell activation pathway. Science 244, 339–343. Liu, J., Valencia‐Sanchez, M. A., Hannon, G. J., and Parker, R. (2005). MicroRNA‐dependent localization of targeted mRNAs to mammalian P‐bodies. Nat. Cell Biol. 7, 719–723. Loflin, P., Chen, C. Y., and Shyu, A. B. (1999). Unraveling a cytoplasmic role for hnRNP D in the in vivo mRNA destabilization directed by the AU‐rich element. Genes Dev. 13, 1884–1897. Lopez de Silanes, I., Galban, S., Martindale, J. L., Yang, X., Mazan‐Mamczarz, K., Indig, F. E., Falco, G., Zhan, M., and Gorospe, M. (2005). Identification and functional outcome of mRNAs associated with RNA‐binding protein TIA‐1. Mol. Cell. Biol. 25, 9520–9531. Lu, J. Y., and Schneider, R. J. (2004). Tissue distribution of AU‐rich mRNA‐binding proteins involved in regulation of mRNA decay. J. Biol. Chem. 279, 12974–12979. Lykke‐Andersen, J., and Wagner, E. (2005). Recruitment and activation of mRNA decay enzymes by two ARE‐mediated decay activation domains in the proteins TTP and BRF‐1. Genes Dev. 19, 351–361. Ma, W.‐J., Cheng, S., Campbell, C., Wright, A., and Furneaux, H. (1996). Cloning and characterization of HuR, a ubiquitously expressed Elav‐like protein. J. Biol. Chem. 271, 8144–8151. MacMicking, J., Xie, Q. W., and Nathan, C. (1997). Nitric oxide and macrophage function. Annu. Rev. Immunol. 15, 323–350. Mahtani, K. R., Brook, M., Dean, J. L., Sully, G., Saklatvala, J., and Clark, A. R. (2001). Mitogen‐ activated protein kinase p38 controls the expression and posttranslational modification of tristetraprolin, a regulator of tumor necrosis factor alpha mRNA stability. Mol. Cell. Biol. 21, 6461–6469. Mavropoulos, A., Sully, G., Cope, A. P., and Clark, A. R. (2005). Stabilization of IFN‐gamma mRNA by MAPK p38 in IL‐12‐ and IL‐18‐stimulated human NK cells. Blood 105, 282–288. Mazan‐Mamczarz, K., Galban, S., Lopez de Silanes, I., Martindale, J. L., Atasoy, U., Keene, J. D., and Gorospe, M. (2003). RNA‐binding protein HuR enhances p53 translation in response to ultraviolet light irradiation. Proc. Natl. Acad. Sci. USA 100, 8354–8359.


33

Min, H., Turck, C. W., Nikolic, J. M., and Black, D. L. (1997). A new regulatory protein, KSRP, mediates exon inclusion through an intronic splicing enhancer. Genes Dev. 11, 1023–1036. Ming, X. F., Kaiser, M., and Moroni, C. (1998). c‐jun N‐terminal kinase is involved in AUUUA‐ mediated interleukin‐3 mRNA turnover in mast cells. EMBO J. 17, 6039–6048. Ming, X. F., Stoecklin, G., Lu, M., Looser, R., and Moroni, C. (2001). Parallel and Independent Regulation of Interleukin‐3 mRNA Turnover by Phosphatidylinositol 3‐Kinase and p38 Mitogen‐ Activated Protein Kinase. Mol. Cell. Biol. 21, 5778–5789. Mitchell, P., and Tollervey, D. (2000). Musing on the structural organization of the exosome complex. Nat. Struct. Biol. 7, 843–846. Moore, B. A., Aznavoorian, S., Engler, J. A., and Windsor, L. J. (2000). Induction of collagenase‐3 (MMP‐13) in rheumatoid arthritis synovial fibroblasts. Biochim. Biophys. Acta 1502, 307–318. Moore, K. W., de Waal Malefyt, R., Coffman, R. L., and O’Garra, A. (2001). Interleukin‐10 and the interleukin‐10 receptor. Annu. Rev. Immunol. 19, 683–765. Morales‐Ducret, J., Wayner, E., Elices, M. J., Alvaro‐Gracia, J. M., Zvaifler, N. J., and Firestein, G. S. (1992). Alpha 4/beta 1 integrin (VLA‐4) ligands in arthritis. Vascular cell adhesion molecule‐1 expression in synovium and on fibroblast‐like synoviocytes. J. Immunol. 149, 1424–1431. Mukherjee, D., Gao, M., O’Connor, J. P., Raijmakers, R., Pruijn, G., Lutz, C. S., and Wilusz, J. (2002). The mammalian exosome mediates the efficient degradation of mRNAs that contain AU‐rich elements. EMBO J. 21, 165–174. Mukhopadhyay, D., Houchen, C. W., Kennedy, S., Dieckgraefe, B. K., and Anant, S. (2003). Coupled mRNA stabilization and translational silencing of cyclooxygenase‐2 by a novel RNA binding protein, CUGBP2. Mol. Cell 11, 113–126. Muller, C. W. (2001). DNA recognition by NF kappa B and STAT transcription factors. Ernst Schering Res. Found. Workshop 14, 3–166. Musgrave, B. L., Watson, C. L., Haeryfar, S. M., Barnes, C. A., and Hoskin, D. W. (2004). CD2‐ CD48 interactions promote interleukin‐2 and interferon‐gamma synthesis by stabilizing cytokine mRNA. Cell. Immunol. 229, 1–12. Nair, A. P., Hahn, S., Banholzer, R., Hirsch, H. H., and Moroni, C. (1994). Cyclosporin A inhibits growth of autocrine tumour cell lines by destabilizing interleukin‐3 mRNA. Nature 369, 239–242. Neininger, A., Kontoyiannis, D., Kotlyarov, A., Winzen, R., Eckert, R., Volk, H. D., Holtmann, H., Kollias, G., and Gaestel, M. (2002). MK2 targets AU‐rich elements and regulates biosynthesis of tumor necrosis factor and interleukin‐6 independently at different post‐transcriptional levels. J. Biol. Chem. 277, 3065–3068. Ogilvie, R. L., Abelson, M., Hau, H. H., Vlasova, I., Blackshear, P. J., and Bohjanen, P. R. (2005). Tristetraprolin down‐regulates IL‐2 gene expression through AU‐rich element‐mediated mRNA decay. J. Immunol. 174, 953–961. Ostareck, D., Ostareck‐Lederer, A., Shatsky, I., and Hentze, M. (2001). Lipoxygense mRNA silencing in erythroid differentiation: The 30 UTR regulatory complex controls 60S ribosomal subunit joining. Cell 104, 281–290. Ostareck, D. H., Ostareck‐Lederer, A., Wilm, M., Thiele, B. J., Mann, M., and Hentze, M. W. (1997). mRNA silencing in erythroid differentiation: hnRNP K and hnRNP E1 regulate 15‐lipoxygenase translation from the 30 end. Cell 89, 597–606. Parasrampuria, D. A., de Boer, P., Desai‐Krieger, D., Chow, A. T., and Jones, C. R. (2003). Single‐ dose pharmacokinetics and pharmacodynamics of RWJ 67657, a specific p38 mitogen‐activated protein kinase inhibitor: A first‐in‐human study. J. Clin. Pharmacol. 43, 406–413. Pargellis, C., and Regan, J. (2003). Inhibitors of p38 mitogen‐activated protein kinase for the treatment of rheumatoid arthritis. Curr. Opin. Investig. Drugs 4, 566–571. Parker, R., and Song, H. (2004). The enzymes and control of eukaryotic mRNA turnover. Nat. Struct. Mol. Biol. 11, 121–127.

34


Pastore, S., Mascia, F., Mariotti, F., Dattilo, C., Mariani, V., and Girolomoni, G. (2005). ERK1/2 regulates epidermal chemokine expression and skin inflammation. J. Immunol. 174, 5047–5056. Peng, S. S., Chen, C. Y., Xu, N., and Shyu, A. B. (1998). RNA stabilization by the AU‐rich element binding protein, HuR, an ELAV protein. EMBO J. 17, 3461–3470. Perez‐Sala, D., Cernuda‐Morollon, E., Diaz‐Cazorla, M., Rodriguez‐Pascual, F., and Lamas, S. (2001). Posttranscriptional regulation of human iNOS by the NO/cGMP pathway. Am. J. Physiol. Renal Physiol. 280, F466–F473. Phillips, K., Kedersha, N., Shen, L., Blackshear, P. J., and Anderson, P. (2004). Arthritis suppressor genes TIA‐1 and TTP dampen the expression of tumor necrosis factor alpha, cyclooxygenase 2, and inflammatory arthritis. Proc. Natl. Acad. Sci. USA 101, 2011–2016. Piecyk, M., Wax, S., Beck, A. R., Kedersha, N., Gupta, M., Maritim, B., Chen, S., Gueydan, C., Kruys, V., Streuli, M., and Anderson, P. (2000). TIA‐1 is a translational silencer that selectively regulates the expression of TNF‐alpha. EMBO J. 19, 4154–4163. Pillinger, M., and Abramson, S. (1995). The neutrophil in rheumatoid arthritis. Rheum. Dis. Clin. North Am. 21, 691–714. Putland, R. A., Sassinis, T. A., Harvey, J. S., Diamond, P., Coles, L. S., Brown, C. Y., and Goodall, G. J. (2002). RNA destabilization by the granulocyte colony‐stimulating factor stem‐loop destabilizing element involves a single stem‐loop that promotes deadenylation. Mol. Cell. Biol. 22, 1664–1673. Quijada, L., Guerra‐Giraldez, C., Drozdz, M., Hartmann, C., Irmer, H., Ben‐Dov, C., Cristodero, M., Ding, M., and Clayton, C. (2002). Expression of the human RNA‐binding protein HuR in Trypanosoma brucei increases the abundance of mRNAs containing AU‐rich regulatory elements. Nucleic Acids Res. 30, 4414–4424. Raghavan, A., Robison, R. L., McNabb, J., Miller, C. R., Williams, D. A., and Bohjanen, P. R. (2001). HuA and tristetraprolin are induced following T cell activation and display distinct but overlapping RNA binding specificities. J. Biol. Chem. 276, 47958–47965. Raijmakers, R., Schilders, G., and Pruijn, G. J. (2004). The exosome, a molecular machine for controlled RNA degradation in both nucleus and cytoplasm. Eur. J. Cell Biol. 83, 175–183. Raineri, I., Wegmueller, D., Gross, B., Certa, U., and Moroni, C. (2004). Roles of AUF1 isoforms, HuR and BRF1 in ARE‐dependent mRNA turnover studied by RNA interference. Nucleic Acids Res. 32, 1279–1288. Raj, N. B., and Pitha, P. M. (1983). Two levels of regulation of beta‐interferon gene expression in human cells. Proc. Natl. Acad. Sci. USA 80, 3923–3927. Ramos, S. B., Stumpo, D. J., Kennington, E. A., Phillips, R. S., Bock, C. B., Ribeiro‐Neto, F., and Blackshear, P. J. (2004). The CCCH tandem zinc‐finger protein Zfp36l2 is crucial for female fertility and early embryonic development. Development 131, 4883–4893. Ridley, S. H., Dean, J. L., Sarsfield, S. J., Brook, M., Clark, A. R., and Saklatvala, J. (1998). A p38 MAP kinase inhibitor regulates stability of interleukin‐1‐induced cyclooxygenase‐2 mRNA. FEBS Lett. 439, 75–80. Rigby, W. F., Roy, K., Collins, J., Rigby, S., Connolly, J. E., Bloch, D. B., and Brooks, S. A. (2005). Structure/Function Analysis of Tristetraprolin (TTP): p38 Stress‐Activated Protein Kinase and Lipopolysaccharide Stimulation Do Not Alter TTP Function. J. Immunol. 174, 7883–7893. Ristimaki, A., Garfinkel, S., Wessendorf, J., Maciag, T., and Hla, T. (1994). Induction of cyclooxygenase‐2 by interleukin‐1 alpha. Evidence for post‐transcriptional regulation. J. Biol. Chem. 269, 11769–11775. Rodriguez‐Pascual, F., Hausding, M., Ihrig‐Biedert, I., Furneaux, H., Levy, A. P., Forstermann, U., and Kleinert, H. (2000). Complex contribution of the 30 ‐untranslated region to the expressional


35

regulation of the human inducible nitric‐oxide synthase gene. Involvement of the RNA‐binding protein HuR. J. Biol. Chem. 275, 26040–26049. Rousseau, S., Morrice, N., Peggie, M., Campbell, D. G., Gaestel, M., and Cohen, P. (2002). Inhibition of SAPK2a/p38 prevents hnRNP A0 phosphorylation by MAPKAP‐K2 and its interaction with cytokine mRNAs. EMBO J. 21, 6505–6514. Saito, K., Chen, S., Piecyk, M., and Anderson, P. (2001). TIA‐1 regulates the production of tumor necrosis factor alpha in macrophages, but not in lymphocytes. Arthritis Rheum. 44, 2879–2887. Sarkar, B., Xi, Q., He, C., and Schneider, R. J. (2003). Selective degradation of AU‐rich mRNAs promoted by the p37 AUF1 protein isoform. Mol. Cell. Biol. 23, 6685–6693. Schmidlin, M., Lu, M., Leuenberger, S. A., Stoecklin, G., Mallaun, M., Gross, B., Gherzi, R., Hess, D., Hemmings, B. A., and Moroni, C. (2004). The ARE‐dependent mRNA‐destabilizing activity of BRF1 is regulated by protein kinase B. EMBO J. 23, 4760–4769. Schnitzer, T. J., and Hochberg, M. C. (2002). COX‐2‐selective inhibitors in the treatment of arthritis. Cleve. Clin. J. Med. 69(Suppl. 1), SI20–SI30. Serhan, C. N. (2001). Lipoxins and aspirin‐triggered 15‐epi‐lipoxins are endogenous components of antiinflammation: Emergence of the counterregulatory side. Arch. Immunol. Ther. Exp. (Warsz) 49, 177–188. Shaw, G., and Kamen, R. (1986). A conserved AU sequence from the 30 untranslated region of GM‐ CSF mRNA mediates selective mRNA degradation. Cell 46, 659–667. Sirenko, O. I., Lofquist, A. K., De Maria, C. T., Morris, J. S., Brewer, G., and Haskill, J. S. (1997). Adhesion‐dependent regulation of an A þ U‐rich element‐binding activity associated with AUF1. Mol. Cell. Biol. 17, 3898–3906. Sood, R., Porter, A. C., Olsen, D. A., Cavener, D. R., and Wek, R. C. (2000). A mammalian homologue of GCN2 protein kinase important for translational control by phosphorylation of eukaryotic initiation factor‐2alpha. Genetics 154, 787–801. Stein, I., Neeman, M., Shweiki, D., Itin, A., and Keshet, E. (1995). Stabilization of vascular endothelial growth factor mRNA by hypoxia and hypoglycemia and coregulation with other ischemia‐induced genes. Mol. Cell. Biol. 15, 5363–5368. Stoeckle, M. Y. (1991). Post‐transcriptional regulation of gro alpha, beta, gamma, and IL‐8 mRNAs by IL‐1 beta. Nucleic Acids Res. 19, 917–920. Stoecklin, G., Colombi, M., Raineri, I., Leuenberger, S., Mallaun, M., Schmidlin, M., Gross, B., Lu, M., Kitamura, T., and Moroni, C. (2002). Functional cloning of BRF1, a regulator of ARE‐ dependent mRNA turnover. EMBO J. 21, 4709–4718. Stoecklin, G., Hahn, S., and Moroni, C. (1994). Functional hierarchy of AUUUA motifs in mediating rapid interleukin‐3 mRNA decay. J. Biol. Chem. 269, 28591–28597. Stoecklin, G., Lu, M., Rattenbacher, B., and Moroni, C. (2003). A constitutive decay element promotes tumor necrosis factor alpha mRNA degradation via an AU‐rich element‐independent pathway. Mol. Cell. Biol. 23, 3506–3515. Stoecklin, G., Mayo, T., and Anderson, P. (2005). ARE‐mRNA degradation requires the 50 ‐30 decay pathway. EMBO Rep. In Press. Stoecklin, G., Ming, X. F., Looser, R., and Moroni, C. (2000). Somatic mRNA turnover mutants implicate tristetraprolin in the interleukin‐3 mRNA degradation pathway. Mol. Cell. Biol. 20, 3753–3763. Stoecklin, G., Stoeckle, P., Lu, M., Muehlemann, O., and Moroni, C. (2001). Cellular mutants define a common mRNA degradation pathway targeting cytokine AU‐rich elements. RNA 7, 1578–1588. Stoecklin, G., Stubbs, T., Kedersha, N., Wax, S., Rigby, W. F., Blackwell, T. K., and Anderson, P. (2004). MK2‐induced tristetraprolin:14‐3‐3 complexes prevent stress granule association and ARE‐mRNA decay. EMBO J. 23, 1313–1324.

36


Stumpo, D. J., Byrd, N. A., Phillips, R. S., Ghosh, S., Maronpot, R. R., Castranio, T., Meyers, E. N., Mishina, Y., and Blackshear, P. J. (2004). Chorioallantoic fusion defects and embryonic lethality resulting from disruption of Zfp36L1, a gene encoding a CCCH tandem zinc finger protein of the Tristetraprolin family. Mol. Cell. Biol. 24, 6445–6455. Sundy, J. S. (2001). COX‐2 inhibitors in rheumatoid arthritis. Curr. Rheumatol. Rep. 3, 86–91. Takeda, K., Clausen, B. E., Kaisho, T., Tsujimura, T., Terada, N., Forster, I., and Akira, S. (1999). Enhanced Th1 activity and development of chronic enterocolitis in mice devoid of Stat3 in macrophages and neutrophils. Immunity 10, 39–49. Taylor, G. A., Carballo, E., Lee, D. M., Lai, W. S., Thompson, M. J., Patel, D. D., Schenkman, D. I., Gilkeson, G. S., Broxmeyer, H. E., Haynes, B. F., and Blackshear, P. J. (1996). A pathogenetic role for TNF alpha in the syndrome of cachexia, arthritis, and autoimmunity resulting from tristetraprolin (TTP) deficiency. Immunity 4, 445–454. Tran, H., Schilling, M., Wirbelauer, C., Hess, D., and Nagamine, Y. (2004). Facilitation of mRNA deadenylation and decay by the exosome‐bound, DExH protein RHAU. Mol. Cell 13, 101–111. van Dijk, E., Cougot, N., Meyer, S., Babajko, S., Wahle, E., and Seraphin, B. (2002). Human Dcp2: A catalytically active mRNA decapping enzyme located in specific cytoplasmic structures. EMBO J. 21, 6915–6924. van Leyen, K., Duvoisin, R. M., Engelhardt, H., and Wiedmann, M. (1998). A function for lipoxygenase in programmed organelle degradation. Nature 395, 392–395. Varnum, B. C., Ma, Q. F., Chi, T. H., Fletcher, B., and Herschman, H. R. (1991). The TIS11 primary response gene is a member of a gene family that encodes proteins with a highly conserved sequence containing an unusual Cys‐His repeat. Mol. Cell. Biol. 11, 1754–1758. Vasudevan, S., and Peltz, S. W. (2001). Regulated ARE‐mediated mRNA decay in Saccharomyces cerevisiae. Mol. Cell 7, 1191–1200. Vincenti, M. P., and Brinckerhoff, C. E. (2002). Transcriptional regulation of collagenase (MMP‐1, MMP‐13) genes in arthritis: Integration of complex signaling pathways for the recruitment of gene‐specific transcription factors. Arthritis Res. 4, 157–164. Vockerodt, M., Pinkert, D., Smola‐Hess, S., Michels, A., Ransohoff, R. M., Tesch, H., and Kube, D. (2005). The Epstein‐Barr virus oncoprotein latent membrane protein 1 induces expression of the chemokine IP‐10: Importance of mRNA half‐life regulation. Int. J. Cancer 114, 598–605. Voellmy, R. (2004). Transcriptional regulation of the metazoan stress protein response. Prog. Nucleic Acid Res. Mol. Biol. 78, 143–185. Wagner, B. J., De Maria, C. T., Sun, Y., Wilson, G. M., and Brewer, G. (1998). Structure and genomic organization of the human AUF1 gene: Alternative pre‐mRNA splicing generates four protein isoforms. Genomics 48, 195–202. Wang, W., Caldwell, M. C., Lin, S., Furneaux, H., and Gorospe, M. (2000). HuR regulates cyclin A and cyclin B1 mRNA stability during cell proliferation. EMBO J. 19, 2340–2350. Wang, Y., Liu, C., Storey, J., Tibshirani, R., Herschlag, D., and Brown, P. (2002). Precision and functional specificity in mRNA decay. Proc. Natl. Acad. Sci. USA 99, 5860–5865. Wernicke, D., Schulze‐Westhoff, C., Brauer, R., Petrow, P., Zacher, J., Gay, S., and Gromnica‐Ihle, E. (2002). Stimulation of collagenase 3 expression in synovial fibroblasts of patients with rheumatoid arthritis by contact with a three‐dimensional collagen matrix or with normal cartilage when coimplanted in NOD/SCID mice. Arthritis Rheum. 46, 64–74. Westra, J., Doornbos‐van der Meer, B., de Boer, P., van Leeuwen, M. A., van Rijswijk, M. H., and Limburg, P. C. (2004a). Strong inhibition of TNF‐alpha production and inhibition of IL‐8 and COX‐2 mRNA expression in monocyte‐derived macrophages by RWJ 67657, a p38 mitogen‐ activated protein kinase (MAPK) inhibitor. Arthritis Res. Ther. 6, R384–R392.


37

Westra, J., Kuldo, J. M., van Rijswijk, M. H., Molema, G., and Limburg, P. C. (2005). Chemokine production and E‐selectin expression in activated endothelial cells are inhibited by p38 MAPK (mitogen activated protein kinase) inhibitor RWJ 67657. Int. Immunopharmacol. 5, 1259–1269. Westra, J., Limburg, P. C., de Boer, P., and van Rijswijk, M. H. (2004b). Effects of RWJ 67657, a p38 mitogen activated protein kinase (MAPK) inhibitor, on the production of inflammatory mediators by rheumatoid synovial fibroblasts. Ann. Rheum. Dis. 63, 1453–1459. Whittemore, L.‐A., and Maniatis, T. (1990). Postinduction turnoff of beta‐interferon gene expression. Mol. Cell. Biol. 64, 1329–1337. Wickens, M., Bernstein, D. S., Kimble, J., and Parker, R. (2002). A PUF family portrait: 30 UTR regulation as a way of life. Trends Genet. 18, 150–157. Williams, B. R. (1999). PKR; a sentinel kinase for cellular stress. Oncogene 18, 6112–6120. Williams, L., Bradley, L., Smith, A., and Foxwell, B. (2004a). Signal transducer and activator of transcription 3 is the dominant mediator of the anti‐inflammatory effects of IL‐10 in human macrophages. J. Immunol. 172, 567–576. Williams, L. M., Ricchetti, G., Sarma, U., Smallie, T., and Foxwell, B. M. (2004b). Interleukin‐10 suppression of myeloid cell activation—a continuing puzzle. Immunology 113, 281–292. Winzen, R., Gowrishankar, G., Bollig, F., Redich, N., Resch, K., and Holtmann, H. (2004). Distinct domains of AU‐rich elements exert different functions in mRNA destabilization and stabilization by p38 mitogen‐activated protein kinase or HuR. Mol. Cell. Biol. 24, 4835–4847. Winzen, R., Kracht, M., Ritter, B., Wilhelm, A., Chen, C. Y., Shyu, A. B., Muller, M., Gaestel, M., Resch, K., and Holtmann, H. (1999). The p38 MAP kinase pathway signals for cytokine‐induced mRNA stabilization via MAP kinase‐activated protein kinase 2 and an AU‐rich region‐targeted mechanism. EMBO J. 18, 4969–4980. Wipke, B. T., and Allen, P. M. (2001). Essential role of neutrophils in the initiation and progression of a murine model of rheumatoid arthritis. J. Immunol. 167, 1601–1608. Wodnar‐Filipowicz, A., and Moroni, C. (1990). Regulation of interleukin 3 mRNA expression in mast cells occurs at the posttranscriptional level and is mediated by calcium ions. Proc. Natl. Acad. Sci. USA 87, 777–781. Worthington, M. T., Pelo, J. W., Sachedina, M. A., Applegate, J. L., Arseneau, K. O., and Pizarro, T. T. (2002). RNA Binding Properties of the AU‐rich Element‐binding Recombinant Nup475/ TIS11/Tristetraprolin Protein. J. Biol. Chem. 277, 48558–48564. Xu, N., Chen, C. Y., and Shyu, A. B. (2001). Versatile role for hnRNP D isoforms in the differential regulation of cytoplasmic mRNA turnover. Mol. Cell. Biol. 21, 6960–6971. Yang, L., and Yang, Y. C. (1994). Regulation of interleukin (IL)‐11 gene expression in IL‐1 induced primate bone marrow stromal cells. J. Biol. Chem. 269, 32732–32739. Yang, Y. H., and Hamilton, J. A. (2001). Dependence of interleukin‐1‐induced arthritis on granulocyte‐macrophage colony‐stimulating factor. Arthritis Rheum. 44, 111–119. Yu, Q., Cok, S. J., Zeng, C., and Morrison, A. R. (2003). Translational repression of human matrix metalloproteinases‐13 by an alternatively spliced form of T‐cell‐restricted intracellular antigen‐ related protein (TIAR). J. Biol. Chem. 278, 1579–1584. Zhang, W., Wagner, B. J., Ehrenman, K., Schaefer, A. W., De Maria, C. T., Crater, D., De Haven, K., Long, L., and Brewer, G. (1993). Purification, characterization, and cDNA cloning of an AU‐rich element RNA‐binding protein, AUF1. Mol. Cell. Biol. 13, 7652–7665. Zhu, W., Brauchle, M. A., Di Padova, F., Gram, H., New, L., Ono, K., Downey, J. S., and Han, J. (2001). Gene suppression by tristetraprolin and release by the p38 pathway. Am. J. Lung Cell Mol. Physiol. 281, L499–L508. Zubiaga, A. M., Belasco, J. G., and Greenberg, M. E. (1995). The nonamer UUAUUUAUU is the key AU‐rich sequence motif that mediates mRNA degradation. Mol. Cell. Biol. 15, 2219–2230.

Negative Signaling in Fc Receptor Complexes Marc Dae¨ron and Renaud Lesourne Unite´ d’Allergologie Molećulaire et Cellulaire, De´partement d’Immunologie, Institut Pasteur, Paris, France

1. 2. 3. 4. 5.

Abstract............................................................................................................. Fc Receptors ...................................................................................................... Positive Signaling by Activating FcRs ...................................................................... Negative Signaling by Activating FcRs..................................................................... Negative Signaling by Inhibitory FcRs..................................................................... Conclusion ......................................................................................................... References .........................................................................................................

39 40 44 50 60 70 74

Abstract Cell activation results from the transient displacement of an active balance between positive and negative signaling. This displacement depends in part on the engagement of cell surface receptors by extracellular ligands. Among these are receptors for the Fc portion of immunoglobulins (FcRs). FcRs are widely expressed by cells of hematopoietic origin. When binding antibodies, FcRs provide these cells with immunoreceptors capable of triggering numerous biological responses in response to a specific antigen. FcR‐dependent cell activation is regulated by negative signals which are generated together with positive signals within signalosomes that form upon FcR engagement. Many molecules involved in positive signaling, including the FcRb subunit, the src kinase lyn, the cytosolic adapter Grb2, and the transmembrane adapters LAT and NTAL, are indeed also involved in negative signaling. A major player in negative regulation of FcR signaling is the inositol 5‐phosphatase SHIP1. Several layers of negative regulation operate sequentially as FcRs are engaged by extracellular ligands with an increasing valency. A background protein tyrosine phosphatase‐dependent negative regulation maintains cells in a «resting» state. SHIP1‐dependent negative regulation can be detected as soon as high‐affinity FcRs are occupied by antibodies in the absence of antigen. It increases when activating FcRs are engaged by multivalent ligands and, further, when FcR aggregation increases, accounting for the bell‐shaped dose‐ response curve observed in excess of ligand. Finally, F‐actin skeleton‐associated high‐molecular weight SHIP1, recruited to phosphorylated ITIMs, concentrates in signaling complexes when activating FcRs are coengaged with inhibitory FcRs by immune complexes. Based on these data, activating and inhibitory FcRs could be used for new therapeutic approaches to immune disorders.


0065-2776/06 $35.00 DOI: 10.1016/S0065-2776(05)89002-9

40

M A R C D A E¨ R O N A N D R E N A U D L E S O U R N E

When sensitized with IgE antibodies, mouse mast cells and human basophils release granular mediators and secrete pro‐inflammatory cytokines and chemokines in response to stimulation by specific antigen. These biological responses depend on high‐affinity receptors for the Fc portion of IgE antibodies (FceRI) that are expressed by the two cell types (Ishizaka et al., 1970; Metzger et al., 1986; Prouvost‐Danon and Binaghi, 1970). For a given concentration of IgE used for sensitization, mediator release increases with the concentration of antigen used for challenge up to a maximum. Release then decreases as the concentration of antigen further increases (Dembo et al., 1978). Peritoneal mouse mast cells also degranulate when challenged by preformed IgG immune complexes (Prouvost‐Danon et al., 1966). IgG‐ induced responses depend on low‐affinity receptors for the Fc portion of IgG (FcgRIIIA) (Dae¨ron et al., 1992; Hazenbos et al., 1996) that bind immune complexes with a high avidity. Bone Marrow‐derived Mast Cells (BMMC) do not respond or respond very poorly to IgG immune complexes, although they express FcgRIIIA (Benhamou et al., 1990). Likewise, human blood basophils release no or little histamine in response to immune complexes (Van Toorenenbergen and Aalberse, 1981), although they express another type of low‐affinity IgG receptor (FcgRIIA) which can activate mast cells (Dae¨ron et al., 1995a). These observations have long been interpreted as resulting from an inefficient engagement of activating receptors by high concentrations of antigen or by IgG immune complexes. Actually, these experiments show that negative regulation occurs in Fc Receptor (FcR) complexes. One is an example of autonomous negative regulation of activating FcRs; others are examples of negative regulation by inhibitory FcRs. These examples were selected from studies of FcRs in mast cells and basophils. FcR‐ dependent negative signaling is not peculiar to these cells. Mast cells are, however, convenient models to study FcR signaling, and they will often be used as examples throughout this review. 1. Fc Receptors 1.1. FcRs, the Third Type of Immunoreceptors Receptors for the Fc portion of immunoglobulins are immunoreceptors of the third type. They ‘‘recognize’’ neither native antigens as B Cell Receptors (BCRs) do, nor the association of antigen‐derived peptides with Major Histocompatibility Complex molecules, as T Cell Receptors (TCRs) do, but antigen‐ antibody complexes. Even though they do not themselves bind to antigen, they enable cells to respond specifically to antigen. Antibodies indeed function as extracellular adapter molecules when their Fab and Fc portions bind

N E G AT I V E S I G N A L I N G I N F C R E C E P T O R C O M P L E X E S

41

simultaneously to specific epitopes on antigen and to FcRs on cell membrane, respectively. BCRs, TCRs, and FcRs are receptors for the three forms under which any given antigen can interact with and deliver signals to cells of the immune system. BCRs and TCRs are assigned specificity at an early stage during B and T cell differentiation through somatic DNA rearrangements. Combinations of variable gene segments determine the clonally restricted specificity of lymphocytes. Specificity persists over cell divisions, as it is transmitted to the progeny within a given clone. These unique features of lymphocytes have several consequences. Altogether, the lymphocytes of an individual can recognize virtually all antigens this individual can be exposed to. Their number being finite, only a small number of naı¨ve lymphocytes can respond to a given antigen. Lymphocytes therefore need first to undergo clonal expansion for significant numbers of cells expressing antigen receptors with any given specificity to be generated and to mount an adaptive immune response. In addition, B and T lymphocytes are not ready‐to‐work effector cells. They need to differentiate into antibody‐producing plasma cells and into helper, regulatory, or cytoxic T cells, respectively, before they can act on antigen. Unlike lymphocytes, large numbers of differentiated cells of hematopoietic origin are capable of exerting a variety of biological activities without needing to proliferate and/or to differentiate. These mostly myeloid cells are the primary effectors of innate immunity. They are equipped with pattern‐ recognition receptors which enable them to interact with structures borne or secreted by microorganisms, but they lack antigen receptors. Most myeloid cells, however, express FcRs. FcRs provide these cells with immunoreceptors and a bona fide immunological specificity. Antigen specificity is provided by antibodies that happen to be present in the environment and bind to FcRs. As these antibodies, polyclonal in nature, have different specificities, one FcR‐ expressing cell can respond specifically to a wide repertoire of different antigens. This repertoire can, theoretically, be as wide as that of the whole population of B cells. In the presence of specific antibodies, FcRs enroll in adaptive immunity, the many cells involved in innate immunity. Besides endowing them with specificity, FcRs can indeed generate intracellular signals which modulate their biological activities. Some FcRs activate, whereas others inhibit cellular responses. 1.2. Activating FcRs Most FcRs are activating receptors (Dae¨ron, 1997; Ravetch and Bolland, 2001; Ravetch and Kinet, 1991). Activating FcRs comprise receptors for IgA (FcaRI), IgE (FceRI), and IgG (FcgRI, FcgRIIA/C, FcgRIIIA, and FcgRIV).

42


They include high‐affinity receptors (FcaRI, FceRI, FcgR, and FcgIV) which can bind monomeric immunoglobulins, and low‐affinity receptors (FcgRIIA/C and FcgRIIIA) which cannot, but which can bind multivalent antigen‐antibody complexes and immunoglobulin aggregates with a high avidity (Hulett and Hogarth, 1994). As a consequence, a proportion of high‐affinity FcRs are occupied in vivo, whereas low‐affinity FcRs remain free in spite of the high concentrations of immunoglobulins present in the extracellular milieu. With one exception in humans (FcgRIIA/C), activating FcRs are multi‐chain receptors composed by one immunoglobulin‐binding FcRa subunit and one (FcRg) or two (FcRg and FcRb) common transduction subunits. As for other immunoreceptors, the cell‐activating properties of FcRs depend on the presence of Immunoreceptor Tyrosine‐based Activation Motif(s) (ITAMs) in the intracytoplasmic domains of their transduction subunits (Reth, 1989). Activating FcRs are expressed by myeloid cells and by lymphoid cells with no classical antigen receptor (i.e., NK cells [Perussia et al., 1989] and intraepithelial g/d T cells of the intestine [Deusch et al., 1991; Sandor et al., 1992; Woodward and Jenkinson, 2001]). They are not expressed by mature T and B lymphocytes. Lymphocytes therefore do not express more than one type of antigen receptor, and activating FcRs do not interfere with lymphocyte activation triggered by clonally expressed antigen receptors. Interestingly, however, activating FcRs are transiently expressed by pre‐B and pre‐T cells, before they express a functional BCR or TCR, respectively (Sandor and Lynch, 1992). Low levels of FcgRIIIA were recently reported to be expressed on a subset of self‐specific murine CD8 T cells and to efficiently trigger antibody‐dependent cell‐ mediated cytotoxicity (Dhanji et al., 2005). Differing from other immunoreceptors, which induce both cell activation and proliferation, FcRs induce cell activation only. Activating FcRs do not induce unique biological responses, but biological activities that can be induced by other receptors in the same cell. 1.3. Inhibitory FcRs Inhibitory FcRs consist of one family of low‐affinity receptors for IgG, referred to as FcgRIIB (Dae¨ron, 1997; Ravetch and Bolland, 2001). FcgRIIB are single‐chain receptors, encoded by one gene named fcgr2b, which generates two (FcgRIIB1 and FcgRIIB2 in humans) or three (FcgRIIB1, FcgRIIB1’, and FcgRIIB2 in mice) isoforms of membrane receptors, by alternative splicing of sequences encoded by the first intracytoplasmic exon (Hibbs et al., 1986; Latour et al., 1996; Lewis et al., 1986; Ravetch et al., 1986). One distinctive feature of the fcgr2b gene is indeed that one exon encodes the transmembrane domain and three others the intracytoplasmic domain of FcgRIIB (in other FcR genes, a single exon encodes both the transmembrane


43

and the intracytoplasmic domains) (Brooks et al., 1989; Hibbs et al., 1988). The inhibitory properties of FcgRIIB depend on an Immunoreceptor Tyrosine‐ based Inhibition Motif (ITIM) (Dae¨ron et al., 1995a), encoded by the third intracytoplasmic exon of the fcgr2b gene, and located in the intracytoplasmic domain of all murine and human FcgRIIB isoforms. FcgRIIB are expressed by myeloid and, with two exceptions, by lymphoid cells. The two exceptions are NK cells and resting T cells which express a variety of other inhibitory receptors involved in cell–cell interactions (Long, 1999). FcgRIIB can negatively regulate cell activation triggered by all ITAM‐containing receptors (Dae¨ron et al., 1995a) as well as cell proliferation triggered by growth factor receptors with an intrinsic kinase activity (Malbec et al., 1999). In order to exert their inhibitory properties, FcgRIIB must be co‐engaged with activating receptors by a common extracellular ligand at the surface of the same cell (Dae¨ron et al., 1995b). The specificity of negative regulation is therefore under the control of two antigen‐specific recognition processes: that of IgG antibodies which engage FcgRIIB and that of immunoreceptors with which FcgRIIB are coaggregated. 1.4. Activating and Inhibitory FcRs in Physiology and Pathology The aggregation of identical FcRs only (homo‐aggregation) is a rare situation in physiology. Even when cells express one type of FcR only (e.g., FcgRIIB in murine B cells, or FcgRIIIA in murine NK cells), immune complexes can co‐ engage FcRs with other immunoreceptors (BCRs in B cells, or NK Receptors on NK cells). Several FcRs are coaggregated when IgG immune complexes interact with cells that co‐express several FcgRs (FcgRI, FcgRIIB, and FcgRIIIA on macrophages or dendritic cells, for instance, or FcgRIIA and FcgRIIB on human B cells) or with cells that co‐express FcRs for several classes of antibodies (mouse mast cells, for instance, where IgG immune complexes can coaggregate FcgRs and FceRI‐bound IgE). Hetero‐aggregation, that is, the coaggregation of different types of FcRs or the coaggregation of FcRs with other immunoreceptors, is actually a rule, rather than an exception, under physiological conditions. Because there are FcRs for all antibody classes, because immune complexes contain more than one class of antibody, and because most cells express more than one type of FcRs, various combinations of FcRs can be engaged at the cell surface to form hetero‐aggregates with a non‐predetermined composition. FcRs can thus generate a variety of signaling complexes, depending on the relative proportion of receptors of the various types that are co‐engaged by immune complexes. The in vivo biological significance of FcgRIIB‐dependent negative regulation of activating FcR‐dependent physio‐pathological processes has been

44


established using mice rendered deficient for FcgRIIB by homologous recombination (Ravetch and Bolland, 2001). Compared to wt mice, FcgRIIB‐ deficient mice were found to produce more antibodies (Takai et al., 1996), to exhibit exaggerated anaphylactic reactions and Arthus reactions of a higher intensity (Takai et al., 1996; Ujike et al., 1999), to be more susceptible to collagen‐induced arthritis (Kleinau et al., 2000; Yuasa et al., 1999), and, in the C57BL/6 background, to spontaneously develop Lupus‐like syndromes (Bolland and Ravetch, 2000). FcgRIIB were also shown to critically determine the protective effects of anti‐tumor therapeutic antibodies in a murine model of melanoma (Clynes et al., 1998), and of IVIG in a model of idiopathic thrombopenic purpura (Samuelsson et al., 2001). 2. Positive Signaling by Activating FcRs 2.1. Positive Signaling in Resting Cells Positive signals are generated even before immunoreceptors are engaged by extracellular ligands. This can be readily unraveled by treating cells with the tyrosine phosphatase inhibitor pervanadate. Pervanadate‐treated cells display an array of tyrosyl‐phosphorylated molecules, including immunoreceptors, indicating that protein tyrosine kinases are active in resting cells but that their substrates are constantly dephosphorylated by tyrosine phosphatases. It follows that cell activation results from a transient displacement of a physiological balance between positive and negative signals that controls cellular responses. Interestingly, the expression of multi‐subunit immunoreceptors such as BCRs was found to be required (and sufficient?) for intracellular signaling molecules to be phosphorylated in pervanadate‐treated cells (Wossning and Reth, 2004), suggesting that signaling complexes can be organized by immunoreceptors even in the absence of known extracellular ligands, but that positive signals emanating from such complexes are either insufficient to lead to cell activation or are dampened by an autonomous type of negative regulation of immunoreceptor signaling. The displacement of the constitutive balance between positive and negative signals that leads to biological responses primarily depends on extracellular ligands which engage surface receptors. 2.2. FcR Engagement and the Constitution of Signalosomes Activating FcRs trigger signals when aggregated by antibody and multivalent antigen. Dimerization was, long ago, shown to be the minimal degree of FceRI aggregation capable of generating activation signals sufficient for triggering


45

mediator release by mast cells (Segal et al., 1977; Siraganian et al., 1975). Intracellular signals are generated within juxta‐membrane signaling complexes that assemble under FcR aggregates and form signalosomes. Signalosomes are transient structures which contain the signaling complexes generated at a given time and at a given location, in which signaling molecules can meet and interact with each other. These comprise receptors that are co‐engaged by common extracellular ligands, molecules that are recruited underneath, and molecules that are contained in subcellular compartments into which receptor aggregates translocate. Signalosomes are dynamic structures which evolve with time and with their intracellular location. Molecules are sequentially recruited first, as complexes build up and get organized around transmembrane adapters. Recruitment depends in part on inducible molecular changes, such as phosphorylation, on the generation of specific molecules, and on location or relocation of molecules into subcellular compartments. It is stabilized by cooperative interactions between molecules with several binding sites and by cytosolic adapters. The composition of signalosomes then rapidly changes as recruited enzymes meet substrates and act on them. Finally signalosomes are dismantled as signaling molecules are ubiquitinated and degraded by the proteasome. 2.3. Generation of Positive Signals by Activating FcRs An initial event in signal transduction by activating FcRs is the activation of src‐family protein tyrosine kinases. In resting cells, these kinases are maintained in an inactive state as a result of the phosphorylation of a regulatory C‐terminal tyrosine by the C‐terminal tyrosine Src kinase Csk (Okada et al., 1991). This confers the molecule a closed conformation that prevents substrates to have access to the catalytic site of the kinase (Cole et al., 2003). The regulatory tyrosine is dephosphorylated by the transmembrane protein tyrosine phosphatase CD45 (Burns et al., 1994; Thomas and Brown, 1999). Supporting a role of CD45 in FceRI signaling, CD45‐deficient mast cells displayed reduced IgE‐induced mediator release and CD45‐deficient mice were refractory to IgE‐induced systemic anaphylaxis (Berger et al., 1994). How CD45 becomes involved upon FcR receptor engagement is unclear. Whatever the mechanism, src kinases are activated and they can phosphorylate tyrosine residues in the ITAMs of FcR transduction subunits. In most cases, the responsible kinase is Lyn. Whether src kinases are constitutively associated with FcR subunits and transphosphorylate ITAMs upon FcR aggregation (Pribluda et al., 1994), or whether ITAMs are phosphorylated in lipid rafts, where src kinases are concentrated (Brown and London, 2000), upon translocation of FcR aggregates into these microdomains (Field et al., 1997) still

46


needs to be clarified. In any case, phosphorylated ITAMs provide docking sites that mediate the recruitment of SH2 domain‐containing molecules, among which is the two‐SH2 domain‐containing protein tyrosine kinase Syk (Benhamou et al., 1993). Once recruited, Syk is tyrosyl‐phosphorylated by src kinases and it further auto‐phosphorylates (Kimura et al., 1996). This activates its catalytic activity. Syk then phosphorylates tyrosines in multiple molecules (Costello et al., 1996). Among these are the cytosolic adapter molecule SH2 domain‐containing Leukocyte Protein of 76 kDa (SLP‐76) (Hendricks‐Taylor et al., 1997; Kettner et al., 2003) and the raft‐associated transmembrane adapter Linker for Activation of T cells (LAT) (Wonerow and Watson, 2001). A parallel series of src kinase‐initiated events was described, following FceRI aggregation in mouse mast cells. Fyn was indeed found to tyrosyl‐ phosphorylate the cytosolic adapter Gab2, thus enabling its association with the p85 subunit of Phosphatidylinositol 3‐kinase (PI3K) via its SH2 domain, and the subsequent activation of the p110 catalytic subunit of this enzyme (Parravincini et al., 2002). PI3K generates phosphatidyl (3,4,5)tris‐phosphate [PI(3,4,5)P3] by adding a phosphate group at position 3 in phosphatidyl (4,5) bis‐phosphate. Several molecules that contain a Pleckstrin Homology (PH) domain are recruited to the membrane by PI(3,4,5)P3. 2.4. Organization of FcR Signaling Complexes by Adapter Proteins The many molecular interactions that occur in signalosomes generate signals that are organized by tyrosine‐rich adapter molecules which, when phosphorylated, function as scaffold proteins. These include cytosolic and transmembrane adapters. SLP‐76 is one such cytosolic adapter. Besides its N‐terminal SH2 domain, SLP‐76 contains a central proline‐rich region and multiple C‐terminal tyrosines (Jackman et al., 1995). Once phosphorylated, it binds to a variety of molecules including the exchange factor Vav (Tuosto et al., 1996) and other adapters such as Gads, Nck, and SLAP‐130 (Boerth et al., 2000). Based on studies of cells from SLP‐76‐deficient mice, SLP‐76 was shown to contribute to the activation of phospholipase C‐g (PLC‐g) and to the activation of Mitogen‐Activated protein (MAP) kinases (Pivniouk et al., 1999). Transmembrane adapters consist of a short extracellular domain, unlikely to bind extracellular ligands; a single transmembrane domain; and a long intracellular domain devoid of molecular interaction domains, but rich in tyrosine residues. When phosphorylated upon FcR engagement, these tyrosines function as inducible docking sites for cytosolic molecules having SH2 domains.


47

Transmembrane adapters are of two types, depending on the presence, in their intracytoplasmic domain, of a juxtamembrane CxxC motif which targets them to lipid rafts. The Protein Associated with GEMs/Csk‐binding protein (PAG/Cbp), LAT, Non‐T cell Activation Linker/Linker of Activation for B cells (NTAL/LAB), and Lck‐Interacting Membrane protein (LIME) have such a palmitoylation site. The T cell Receptor‐Interacting Molecule (TRIM), SHP‐2‐Interacting Transmembrane adapter (SIT), and Linker of Activation for X cells (LAX) do not, and they are excluded from lipid rafts (Kliche et al., 2004; Togni et al., 2004). LAT was shown to support positive signaling triggered not only by TCR, but also by FceRI, and the mechanisms by which it concurs to mast cell activation and to T cell activation are thought to be similar. FceRI aggregation in BMMC from LAT/ mice triggered a reduced phosphorylation of SLP‐76 and of PLC‐g, resulting in a decreased Ca2þ mobilization and MAP Kinase activation and, ultimately, in a decreased release of preformed mediators and secretion of cytokines (Saitoh et al., 2000). FcRb/FcRg ITAMs and Syk were phosphorylated as in wt cells. These observations suggested that LAT primarily serves as a coupling molecule between immunoreceptors and intracellular signaling pathways leading to cellular responses (Sommers et al., 2004). LAT contains many tyrosines (9 in mice, 10 in humans) in its intracytoplasmic domain (Weber et al., 1998; Zhang et al., 1998a). It is tyrosyl‐phosphorylated by Syk following FceRI engagement, and serves as a scaffold molecule by providing multiple docking sites for additional SH2 domain‐containing cytosolic enzymes and adapters to be recruited. These include PLC‐g, protein tyrosine kinases of the Tec family, the p85 subunit of PI3K, exchange factors of the Vav family, and the adapters Gads, Grap, and Grb2 (Weber et al., 1998; Zhang et al., 1998a, 2000). Works based on mutational analysis of LAT identified critical tyrosine residues involved in the recruitment of these molecules in T cells (Zhang et al., 2000; Zhu et al., 2003). These were the four distal tyrosines (Y132, Y171, Y191, and Y226 in humans, and their homologues in mice Y136, Y175, Y195, and Y235). Specifically, Y132/136 was demonstrated as being the major binding site for PLC‐g, and the three distal tyrosines (Y171/175, Y191/195, and Y226/235) binding sites for Gads, Grap, and Grb2 (Zhang et al., 2000). The two sets of binding sites also contribute to the recruitment of other molecules such as SLP‐76 via Gads and they cooperate to stabilize the binding of molecules recruited by each other. A mutational analysis of the four distal tyrosines of LAT, in LAT/ BMMC reconstituted in vitro with wt or mutant LAT (Saitoh et al., 2003), confirmed that, once phosphorylated upon FceRI engagement, these residues play critical roles for FceRI signaling by recruiting the same set of signaling molecules in mast cells as in T cells.

48


2.5. Intracellular Propagation of FcR Signals Molecules recruited and activated in signalosomes concur to the activation of metabolic pathways which propagate signals intracellularly up to the nucleus and back to the plasma membrane. Several pathways are used by activating FcRs. They are, with variations, the same as pathways used by other immunoreceptors. Some lead to the calcium response, while others lead to the activation of transcription factors. These pathways are not linear, but tightly interconnected. We will briefly underline only critical steps that either contribute to or are targets of negative regulation. The calcium response results from the recruitment and activation of PLC‐g1 and/or 2, depending on the cell type (Wang et al., 2000; Wen et al., 2002). The recruitment of PLC‐g involves the interaction of one of its SH2 domain with phosphorylated Y136 on LAT, and the interaction with the adapter Gads which binds to phosphorylated LAT terminal tyrosines. PLC‐g is also recruited to the membrane through the interaction of its PH domain with newly formed PI (3,4,5)P3. PLC‐g is subsequently activated as a result of the phosphorylation of specific tyrosine residues by Syk and by the Tec kinase Btk (Humphries et al., 2004), respectively. PLC‐g generates inositol (1,4,5)tris‐phosphate [IP(1,4,5) P3 or IP3] and Diacyl Glycerol (DAG). IP3 triggers an efflux of intracellular Ca2þ from the endoplasmic reticulum and, secondarily, an influx of extracellular Ca2þ. The result is a markedly increased intracellular Ca2þ concentration. Intracellular Ca2þ is critical for exocytosis in mast cells. It also activates calcineurin. This phosphatase dephosphorylates the Nuclear Factor of Activation for T cells (NF‐AT), which enables its translocation from the cytosol to the nucleus (Stankunas et al., 1999). DAG upregulates the catalytic activity of several among the many serine‐ threonine kinases of the Protein Kinase C (PKC) family. Following further activation as a result of the phosphorylation of several serine/threonine and tyrosine residues, PKCs phosphorylate a variety of substrates involved in the activation of MAP kinases (Kawakami et al., 1998) and of transcription factors (Turner and Cantrell, 1997), and in mast cell degranulation (Buccione et al., 1994). PKCs can also threonyl‐phosphorylate FcRg (Pribluda et al., 1997), which contributes to the activation of Syk (Swann et al., 1999). Another substrate of DAG‐activated PKCs is the serine Protein Kinase D (PKD) (Valverde et al., 1994). PKD is abundant in mast cells and, when activated upon FceRI engagement, it contributes to the regulation of transcriptional activity of NF‐kB (Johannes et al., 1998). NFkB activation was observed upon FceRI aggregation in mast cells (Hundley, Blood, 2004) and dendritic cells (Kraft et al., 2002), preceded by the seryl‐phosphorylation and degradation of IkB, and it was reported to be


49

involved in the generation of several cytokines (Marquardt and Walker, 2000). NF‐kB was also activated in human monocytes (Drechsler et al., 2002) and mesangial cells (Duque et al., 1997), as a consequence of FcgRs and FcaRI aggregation, respectively. Three sets of MAP kinases are activated upon FcR engagement: Erk1/2, JNK, and p38 (Dong et al., 2002). Erk1/2 are the terminal effector kinases of the Ras pathway, JNK and p38, effector kinases of the rac pathway. Ras and Rac are small G proteins which are in an inactive form when associated with GDP, and in an active form when associated with GTP. The replacement of GDP by GTP on Ras and Rac depends on the exchange factors Sos and Vav, respectively (Cantrell, 1998; Downward, 1996). It initiates a cascade of serine/ threonine phosphorylations, the ultimate substrates of which are MAP kinases. Phosphorylated MAP kinases are translocated into the nucleus where they can phosphorylate transcription factors. These associate with NF‐AT to form a complex which can bind to specific sites in the promoter of cytokine genes and initiate their transcription. 2.6. Ligand Valency Influences FceRI‐Dependent Mast Cell Secretory Responses Several observations recently challenged the widely accepted concept that the binding of monomeric IgE to FceRI generates no detectable signal and no detectable response. An exposure of mast cells to IgE in the absence of antigen was indeed reported (1) to up‐regulate the expression of membrane FceRI (Hsu and MacGlashan, 1996; MacGlashan et al., 1997), (2) to increase the survival of mast cells in the absence of growth factors (Asai et al., 2001; Kawakami and Galli, 2002), and (3) to induce cytokine secretion (Kalesnikoff et al., 2001; Kohno et al., 2005; Pandey et al., 2004). The effect of monomeric IgE on mast cell survival and cytokine secretion was found to depend on the FcRg ITAM (Kohno et al., 2005), but not the upregulation of FceRI expression. The effect on receptor expression was shown to result from slowing down the removal of FceRI from the membrane and its subsequent degradation without affecting the rate of FceRI synthesis (Borkowski et al., 2001). As a consequence, FceRI accumulate on the mast cell membrane without requiring detectable intracellular signals. By contrast, the effects on mast cell survival and cytokine secretion were found to be restricted to anti‐DNP/TNP IgE, to vary markedly from one mAb IgE to another (Kitaura et al., 2005), and most importantly, to be inhibited by a monovalent hapten such as DNP‐lysine (Tanaka et al., 2005). These effects, therefore, must be understood as resulting from FceRI aggregation, whatever the mechanism, that is, to obey the general rule. Interestingly, however, quantitative variations of receptor aggregation were found to result in qualitative

50


variation in cellular responses. A low level of FceRI aggregation, induced by incubating mast cells with IgE in the absence of (known) antigen, triggered intracellular signals leading to the secretion of IL‐3 or MCP‐1, but not to degranulation, whereas a high level of receptor aggregation, induced by incubating with multivalent antigen mast cells sensitized with the same IgE, triggered both degranulation and cytokine secretion (Gonzalez‐Espinosa et al., 2003; Kohno et al., 2005; Yamasaki et al., 2004). Molecular mechanisms that enable quantitative differences in receptor aggregation to produce qualitatively different responses remain to be elucidated. 3. Negative Signaling by Activating FcRs FcR‐dependent cell activation is negatively regulated by several inhibitory mechanisms generated by activating FcRs themselves. Some are triggered together with activation mechanisms by ITAM‐containing FcRs and contribute to their own, autonomous control. Others can be triggered by activating FcRs in the absence of detectable positive signals, although they depend on ITAMs, and can negatively regulate signaling triggered by other activating receptors expressed by the same cell. 3.1. Autonomous Negative Regulation of Activating FcRs When engaged by antibody and antigen, activating FcRs not only generate positive signals, but also negative signals. This autonomous negative regulation controls the intensity and duration of positive signals. Negative signaling depends on several mechanisms involving a variety of molecules. Interestingly, many among the proteins which contribute to negative regulation are the same as those which contribute to positive regulation. These include receptor subunits, protein tyrosine kinases, adapter molecules, and phosphatases. 3.1.1. FcRb The mast cell‐specific FcRb subunit was first understood to function as an amplifier of signals generated by FcRg upon FceRI aggregation (Adamczewski et al., 1995; Dombrowicz et al., 1998; Lin et al., 1996). Differing from mouse or rat FceRIa, which need to associate with both FcRg and FcRb in order to be expressed at the mast cell membrane, human FceRIa need to associate with FcRg only. As a consequence, FceRI can be expressed in human mast cells with or without FcRb. They can also be expressed by human monocytes, macrophages, and eosinophils, which do not express FcRb (Maurer et al., 1994), but not in corresponding murine cells. Signals triggered by FcRb‐

51


associated FceRI were found to be of a higher intensity than signals triggered by FceRI associated with FcRg only (Dombrowicz et al., 1998). FcRb also enhances IgE‐induced allergic responses by up‐regulating the surface expression of FceRI (Donnadieu et al., 2003). Recently, however, FcRb was found to generate ITAM‐dependent negative signals. The FcRb ITAM has a unique feature. Compared to other ITAMs, the FcRb ITAM contains an additional tyrosine residue, in the 6‐residue sequence that separates the two canonical YxxL motifs:

Human Mouse

FcRg

FcRb

YTGLSTRNQETYETL YTGLNTRSQETYETL

YEELNIYSATYSEL YEELNVYSPIYSEL

Based on a mutational analysis, this additional tyrosine was shown to be involved in the negative regulation of IgE‐induced signals. The activation of the MAP kinases Erk and p38, the activation of NF‐kB, and, ultimately, the secretion of IL‐6, IL‐13, and TNF‐a were indeed enhanced when this residue was mutated into phenylalanine (Furumoto et al., 2004). No marked effect was observed on the activation of PLC‐g, the Ca2þ response, the generation of leukotrienes, and the release of b‐hexosaminidase, suggesting that this tyrosine is not critical in signal amplification. These altered responses were reminiscent of the phenotype of mast cells derived from Lyn‐deficient mice (Odom et al., 2004). Indeed, pull‐down experiments using beads coated with phosphopeptides corresponding to a wt or an altered FcRb ITAM showed that the additional tyrosine could mediate the binding of Lyn, and also of the SH2 domain‐containing inositol phosphatase SHIP1. Supporting an in vivo significance of this in vitro analysis, slightly less FcRb coprecipitated with Lyn, and SHIP1 was less phosphorylated following FceRI engagement, when the additional tyrosine was mutated in FcRb. FcRb may therefore contribute to the involvement of SHIP1 and to the recruitment of Lyn in FceRI signaling complexes. Increasing evidence supports the idea that this src family protein tyrosine kinase contributes to negative regulation of immunoreceptor signaling and, possibly, more than to positive regulation as originally thought. 3.1.2. Lyn The src‐family protein tyrosine kinase Lyn was shown to play a critical role in the initiation of IgE‐induced signal transduction in mast cells. Lyn was indeed demonstrated to be responsible for the phosphorylation of both FcRb and FcRg ITAMs upon FceRI aggregation and for the initial phosphorylation of

52


Syk, when the latter has been recruited to the phosphorylated FcRg ITAM (Jouvin et al., 1994; Kihara and Siraganian, 1994; Scharenberg et al., 1995). Lyn was therefore considered first as a major player in positive signaling. When Lyn‐deficient mice became available, it became apparent that Lyn is involved in a variety of negative regulatory processes. B cells from Lyn/ mice were found to be hyper‐responsive to BCR engagement (Chan et al., 1997; Hibbs et al., 2002), and to IL‐4 stimulation (Janas et al., 1999). Lyn/ mast cells were also more responsive to proliferative signals delivered by IL‐3 or Stem Cell Factor (Hernandez‐Hansen et al., 2004). Importantly, Lyn/ mast cells were more responsive to FceRI‐dependent activation signals (Kawakami et al., 2000; Nishizumi and Yamamoto, 1997). As expected, IgE‐induced phosphorylation of FceRI ITAMs was reduced in Lyn/ BMMC (Kawakami et al., 2000; Kovarova et al., 2001). The phosphorylation of Csk‐binding protein (Cbp) was abrogated and, as a consequence, the coprecipitation of Csk with this scaffold adapter protein observed in wt mast cells was lost in Lyn/ mast cells. Noticeably, the catalytic activity of Fyn was increased in these cells, and hyperactive Fyn was phosphorylated on tyrosine 417, in the activation loop of the kinase. The hyper‐responsiveness of Lyn/ mast cells to IgE could be ascribed to this kinase as this phenotype was abrogated in BMMC derived from doubly deficient Lyn//Fyn/ mice (Odom et al., 2004). Altogether these data provided the following explanation to the negative role of Lyn in mast cell activation. In wt cells, Lyn phosphorylates Cbp which recruits Csk. Csk phosphorylates the regulatory tyrosines 508 and 528 of Fyn and thereby inhibits its catalytic activity (Odom et al., 2004). Interestingly, the phenotype of Lyn/ mice was reminiscent of an ‘‘allergic’’ phenotype which could not be accounted for by the hyper‐reactivity of mast cells only. As these mice grew older, they displayed an increased serum IgE concentration, an upregulation of FceRI expression on mast cells, increased numbers of peritoneal mast cells and eosinophils, and elevated levels of plasma histamine (Odom et al., 2004). Most of these allergy‐associated traits could be ascribed to a screwed isotypic switch toward IgE during B cell differentiation due to the hyper‐responsiveness of Lyn/ B cells to IL‐4, and to the consequences of an increased IgE serum concentration. Finally, besides its first recognized role in positive signaling by immunoreceptors, a critical role of Lyn kinase in negative signaling that dampens cell activation by these receptors must be considered. Whether Lyn primarily contributes to positive or to negative signaling may depend on the cell type and on engaged receptors. 3.1.3. LAT LAT has been first understood to organize signalosomes generated by activating receptors and to couple them with downstream signaling pathways leading


53

to cellular responses (Sommers et al., 2004). LAT is critical for TCR signals involved in early T cell differentiation and, indeed, LAT‐deficient mice display an arrest in thymocyte development with a block in both TCRab and gd T cell differentiation (Zhang et al., 1999). Unexpectedly, knock‐in mice, expressing LAT with a single point mutation of the PLC‐g‐binding site (Y136F), displayed an aberrant T cell development characterized by a partial block in early T cell differentiation and polyclonal lymphoproliferative disorder, resulting in abnormally high numbers of CD4þ TCRab T cells that secreted abnormally high levels of TH2 cytokines in the periphery. As a consequence of this exaggerated TH2 polarization, serum IgG1 concentrations were 5000‐fold higher than in wild‐type mice, serum IgE concentrations were in the range of several mg/ml, instead of a few mg/ml, and peripheral tissues were massively infiltrated with eosinophils. The differentiation of TCRgd T cells was unaffected (Aguado et al., 2002; Sommers et al., 2002). Likewise, knock‐in mice bearing point mutations of the adapter‐binding three distal tyrosines of LAT (Y175F, Y195F, and Y235F) displayed a complete block in the differentiation of TCRab T cells and an abnormal differentiation of TCRgd T cells, also resulting in an exaggerated TH2 polarization and massive proliferation. As a result, IL‐4 secretion was increased, and the serum concentrations of IgG1 and IgE were 500‐ and 1000‐fold higher than in normal mice, respectively (Nun˜ez‐Cruz et al., 2003). Although they affect two distinct T cell lineages, respectively, the two types of LAT tyrosine mutations therefore seemed to inhibit a negative regulation that normally controls terminal T cell differentiation. This suggested that, besides positive signals, LAT might support negative signals that normally regulate terminal T cell differentiation and proliferation, and that this regulation, which differentially affects TCRab and TCRgd signaling, depends on distinct tyrosine residues. Our analysis of IgE‐induced biological responses of cultured mast cells derived from the same knock‐in mice led to the same conclusion for FceRI signaling. A systematic comparison of biologic responses observed in pairs of mutants enabled us to dissect the respective roles played by LAT tyrosines in mast cells (Malbec et al., 2004). As expected, Y136 and the three distal tyrosines differentially contributed to exocytosis and the secretion of cytokines, on the one hand, and to the generation or the activation of major cytosolic effectors such as intracellular Ca2þ and the terminal MAP kinases of the ras pathway, Erk1/2, on the other hand. Interestingly, mutations unraveled the existence of negative signals, generated by distinct LAT tyrosines. Thus Y136 had a negative effect on mediator release when Y175, 195, and 235 were mutated and, conversely, Y175, 195, and 235 had a negative effect when Y136 was mutated. Positive and negative signals generated by different segments of the LAT molecule are apparently additive. Thus, sequences containing the five proximal tyrosines

54


could abrogate the negative effects of Y136 in the absence of the three distal tyrosines or the negative effects of the three distal tyrosines in the absence of Y136, observed on b‐hexosaminidase release in BMMC. Importantly, LAT can integrate positive and negative signals even when in a wt configuration. Thus, the four distal tyrosines together had a positive effect on b‐hexosaminidase release in BMMC, but of a lower magnitude than the intense positive effects of either Y136 alone or of the three distal tyrosines alone. These observations would be best explained if LAT could promote the assembly of a signaling complex composed of a mixture of intracellular molecules with antagonistic properties. 3.1.4. NTAL Another transmembrane adapter was recently cloned as a result of a search for the B cell homologue of LAT and was named Linker for activation of B cells (LAB) (Janssen et al., 2003). Because it is expressed not only by B cells, but also by monocytes, NK cells, mast cells, and platelets, this molecule was also named Non‐T cell Activation Linker (NTAL) (Brdicka et al., 2002). NTAL is encoded in humans by the WBSCR5 gene, on chromosome 7 (Brdicka et al., 2002; Martindale et al., 2000). It consists of a single polypeptide resembling LAT, with a short extracellular domain, a transmembrane domain with a potential palmitoylation CxxC motif, and a long intracytoplasmic domain containing nine tyrosine residues that are phosphorylated upon immunoreceptor engagement and provide multiple binding sites for SH2 domain‐containing molecules. Grb2, Sos, Gab1, and c‐Cbl indeed coprecipated with phosphorylated NTAL in monocytes and B cells (Brdicka et al., 2002). Differing from LAT, NTAL contains no PLC‐g binding site. When expressed in LAT‐deficient T cells, NTAL could partially restore TCR signaling (Koonpaew et al., 2004), and LAT/ mice expressing an NTAL transgene under the control of the CD2 promoter had a phenotype resembling that of LAT Y136F knock‐in mice (Janssen et al., 2004). Based on these observations, NTAL was proposed to play, in B cells, a similar role as the one LAT plays in T cells (Brdicka et al., 2002). Because mast cells co‐express LAT and NTAL and because FceRI signaling was reduced, but not abrogated in BMMC derived from LAT‐ deficient mice, the two adapters were thought to play complementary roles in mast cell activation. Surprisingly, the genetic deletion of NTAL resulted in increased, rather than diminished, IgE‐induced release of granular mediators and secretion of cytokines by mast cells (Volna et al., 2004; Zhu et al., 2004). The tyrosyl‐phosphorylation of Syk, LAT, and PLC‐g1 and 2 were increased, as well as the phosphorylation of the Erk, p38, and JNK MAP kinases in BMMC from NTAL‐deficient mice. The activity of PI3K, the concentration of PI


55

(3,4,5)P3, the amount of IP3, and the Ca2þ response were also increased. NTAL therefore appears to negatively regulate FceRI signaling. The mechanism of inhibition still needs to be elucidated. Whether NTAL recruits an inhibitory molecule, such as a phosphatase, is one possibility that has not been convincingly demonstrated. Whether a competition between LAT and NTAL, which recruit a common set of adapter molecules, exists is supported by the observation that the phosphorylation of LAT is increased in the absence of NTAL (Volna et al., 2004) and that, reciprocally, the phosphorylation of NTAL is increased in the absence of LAT. The augmented phosphorylation of LAT in NTAL/ mast cells is likely to explain the increased phosphorylation of PLC‐g and its consequences on IP3 production and Ca2þ mobilization. Also, NTAL lacks a PLC‐g binding site, and the recruitment of PLC‐g by LAT requires the cooperative binding of several among the adapters that are recruited by both LAT and NTAL and that NTAL might sequester. Against the competition hypothesis, LAT and NTAL were found to reside in distinct lipid rafts on the plasma membrane (Volna et al., 2004). Whether these different microdomains could possibly merge during FceRI signaling, as it was suggested (Rivera, 2005), needs to be demonstrated. Interestingly, NTAL may not only generate negative signals, but also contribute to positive signals in mast cells. These could be observed in the absence of LAT. Inhibition of mediator release was indeed reported to be more pronounced in BMMC from LAT‐ and NTAL‐doubly deficient mice than in BMMC from LAT‐deficient mice (Volna et al., 2004; Zhu et al., 2004). A positive role of NTAL could be seen on Ca2þ responses in T and B lymphocytes (Brdicka et al., 2002; Janssen et al., 2004) and on Stem Cell factor‐ induced activation of human mast cells (Tkaczyk et al., 2004). A recent analysis performed in DT40 B cells proposed that, when recruiting Grb2, phosphorylated NTAL removes a Grb2‐dependent inhibitory effect on the BCR‐induced influx of extracellular Ca2þ (Stork et al., 2004). This inhibitory effect could be due to protein tyrosine phosphatases and inositol phosphatases which associate with Grb2 in different conditions. 3.1.5. Protein Tyrosine Phosp hatases Protein tyrosine phosphatases are thought to negatively regulate FcR signaling. Supporting evidence is, however, scarce. The SH2 domain‐containing Protein Tyrosine Phosphatase SHP‐1 has been implicated in FceRI signaling by using trapping mutants (Xie et al., 2000). SHP‐1 was reported to associate with the phosphorylated ITAM of FcgRIIA, Syk, the p85 subunit of PI3K, and Dok‐1, and to decrease the tysrosyl‐phosphorylation of intracellular proteins, upon FcgRIIA aggregation in the macrophage‐like THP‐1 cells (Ganesan

56


et al., 2003). SHP‐1 also contains several consensus binding motifs for the SH2 domain of Grb2, and the inhibitory effect of Grb2‐SHP‐1 complexes was observed on cytokine receptor signaling (Minoo et al., 2004). The possible role of the second SH2 domain‐containing Protein Tyrosine Phosphatase SHP‐ 2 in the negative regulation of FcR‐dependent cell activation remains to be demonstrated. Several protein phosphatase devoid of SH2 domain were also found to be activated upon FceRI aggregation and to dephosphorylate ITAMs (Swieter et al., 1995). 3.1.6. Inositol Phosp hatases Inositol phosphatases, by contrast, play a prominent role in controlling FcR‐ dependent cell activation. The inositol 3‐phosphatase PTEN was involved in FcgR signaling as Akt and MAP kinase phosphorylation induced upon FcgRIIIA aggregation was enhanced in macrophages from PTEN/ mice, resulting in enhanced cytokine secretion (Cao et al., 2004). SHIP2 was tyrosyl‐ phosphorylated upon FcgRI engagement in THP‐1 cells or upon FcgRIIA engagement in human peripheral blood monocytes following upregulation by LPS, and it associated via its SH2 domain to the phosphorylated ITAM of this receptor (Pengal et al., 2003). Finally, SHIP1 was described to inhibit FcgRIIA‐dependent phagocytosis in THP‐1 cells (Nakamura et al., 2002), to coprecipitate with phosphorylated FcgRIIA, and to negatively regulate NFk‐ B‐mediated gene transcription during phagocytosis in human myeloid cells (Tridandapani et al., 2002). SHIP1 activity was reported to associate with the phosphorylated z subunit and to negatively regulate FcgRIIIA‐dependent ADCC in human NK cells (Galandrini et al., 2002). SHIP1 was found to bind in vitro to phosphopeptides corresponding to the FcRb ITAM (Kimura et al., 1997) and to interact with FcRb when examined by yeast triple hybrid assay (Osborne et al., 1996). The possible in vivo recruitment of this phosphatase in FceRI signaling complexes remains elusive as, so far, it was not reported to coprecipitate with FceRI, including the FcRb subunit, following receptor engagement in mast cells. SHIP1 was, however, understood to play a central regulatory role in the autonomous negative regulation of FceRI signaling. This conclusion was based on studies of SHIP1‐deficient mice. As they get older, SHIP1/ mice spontaneously develop a splenomegaly and a progressive lung infiltration by myeloid cells that leads to a waste syndrome and, ultimately, to a shortened life span. Their myeloid progenitor cells are hyper‐responsive to cytokines, such as IL‐3, and growth factors, such as Granulocyte/Macrophage Colony‐Stimulating Factor and Stem Cell Factor (Helgason et al., 1998). Interestingly, BMMC derived from SHIP1/ mice are hyper‐responsive not only to Stem Cell Factor‐, but also to IgE‐dependent


57

stimulation. Such cells indeed release more b‐hexosaminidase than do BMMC derived from wt mice in response to FceRI aggregation by IgE and antigen. Supporting the conclusion that BMMC from SHIP1/ mice could respond to a lower degree of receptor aggregation, IgE anti‐DNP alone could trigger these cells, but not wt‐type cells, to release b‐hexosaminidase, as well as an array of cytokines. These antigen‐independent responses were inhibited by a monovalent hapten such as DNP‐lysine (Huber et al., 1998; Kalesnikoff et al., 2001). IgE‐induced increased degranulation was correlated with augmented and sustained Ca2þ mobilization and Erk1/2 activation. The phosphorylation of Shc, which associates constitutively to SHIP1, was reduced in the absence of SHIP1, but, surprisingly, FcRb phosphorylation was increased. Based on these data, SHIP1 was proposed to raise the threshold of FceRI aggregation needed to generate activation signals and to function as a ‘‘gatekeeper’’ of mast cell degranulation (Huber et al., 1998). SHIP1 is constitutively active. By contrast with SHPs, the phosphatase activity of SHIP1 is not up‐regulated when its SH2 domain binds to a tyrosyl‐phosphorylated motif, but when it is translocated close to the membrane (Bolland et al., 1998). The expression of a membrane‐targeted CD8‐SHIP1 chimera in COS cells constitutively induced a three‐fold higher enzymatic activity than the expression of a cytosolic form of SHIP1 (Phee et al., 2000). A simple explanation is that, under these conditions, SHIP1 is located close to its membrane substrate. SHIP1 removes 5‐phosphate groups in the inositol ring of 3‐phosphorylated inositides and phosphatidylinositides. Its substrates are inositol (1,3,4,5)tetrakis‐phosphate [I(1,3,4,5)P4] and PI(3,4,5)P3, which are hydrolyzed into inositol (1,3,4)tris‐phosphate and into phosphatidylinositol (3,4)bis‐phosphate, respectively (Damen et al., 1996). SHIP1 can therefore prevent PI(3,4,5)P3‐dependent critical upstream events leading to the Ca2þ response and, as a consequence, inhibit cell responses (Scharenberg and Kinet, 1998; Scharenberg et al., 1998). Another role of SHIP1 in autonomous negative regulation was recently unraveled. This regulation accounts for the bell‐shaped curve of mast cell activation as a function of antigen concentration. Inhibition of biological responses in excess of antigen is unique neither to FceRI nor to mast cells. It was long interpreted as resulting from a progressive decrease in receptor aggregation, due to a competition of high concentrations of antigen for efficiently crosslinking FceRI‐bound IgE (Dembo et al., 1978; Wofsy et al., 1978), although negative regulation had previously been hypothesized as an explanation, resulting from an excess of receptor aggregation (Magro and Alexander, 1974). Supporting experimental predictions deduced from a mathematical analysis (Delisi and Siraganian, 1979), recent works provided evidence that intracellular signals do not decrease, but increase, as the concentration of

58


antigen increases. Thus, the tyrosyl‐phosphorylation of intracellular proteins in whole cell lysates and, more specifically, of FcRb and PLC‐g were of a higher magnitude in BMMC stimulated with supra‐optimal concentrations of antigen than in BMMC stimulated with an optimal antigen concentration. The secretory response decreases, however, because negative signals increase and become dominant over positive signals. Supporting this interpretation, the inducible tyrosyl‐phosphorylation of SHIP1 dose‐dependently increased with the concentration of antigen, even after supra‐optimal concentrations were reached. Most importantly, inhibition of secretion induced by an excess of antigen in mast cells derived from wt mice was abrogated in mast cells derived from SHIP1‐deficient mice (Gimborn et al., 2005). These data altogether indicate that SHIP1, possibly recruited by FcRb when heavily phosphorylated as a result of supra‐optimal receptor aggregation, is the effector of autonomous negative regulation of FceRI signaling that dampens mast cell activation in excess of ligand. 3.1.7. Cbl Finally, ubiquitination of receptors and signaling molecules, followed by proteasomal degradation, were shown to terminate cell activation. Thus, following FceRI engagement, FcRb and FcRg, as well as Syk, undergo rapid c‐Cbl‐ dependent E3 ligase‐mediated ubiquitination (Gimborn et al., 2005). Lyn also associates with c‐Cbl and is ubiquitinated and degraded in IgE‐activated mast cells (Kyo et al., 2003). Likewise, Syk and ZAP‐70 are ubiquitinated following FcgRIIIA engagement in human NK cells (Paolini et al., 2001). 3.2. Promiscuous Negative Regulation of Activating FcRs by FcaRI ITAM‐containing FcRs were recently demonstrated to have the ability of generating not only positive and negative signals which regulate each other, but also negative signals which can affect positive signals delivered by other activating FcRs in the same cell. FcaRI are such receptors. They bind monomeric IgA with a moderate affinity and dimeric IgA with a high avidity (Wines et al., 2001). FcaRI are encoded by genes of the Leukocyte Receptor Complex, on chromosome 19. They share with receptors encoded by this gene family a KIR‐type orientation, instead of an FcR‐type orientation of their extracellular domains (Herr et al., 2003). Although FcaRI can be expressed without, FcaRI associate with FcRg and, upon aggregation by IgA immune complexes, they trigger cell activation like other ITAM‐containing immunoreceptors. They are expressed by a variety of myeloid cells which contribute to inflammation (Monteiro and Van De Winkel, 2003).


59

Surprisingly, the engagement of FcRg‐associated FcaRI by monomeric ligands—Fab fragments of mAbs against the extracellular domains of human FcaRI or human serum IgA—was found to negatively regulate the in vitro phagocytosis of IgG‐opsonized bacteria by human monocytes or IgE‐ dependent exocytosis in the rat mast cell line RBL‐2H3 transfected with cDNA encoding FcaRI. When administered intraperitoneally into human FcaRI transgenic mice, anti‐FcaRI Fab fragments also inhibited bronchial constriction and airway infiltration by inflammatory cells induced by IgE and antigen in a murine model of allergic asthma. Using chimeric molecules made of the a subunit of FcaRI the transmembrane domain of which had a point mutation preventing the association with FcRg and the intracytoplasmic domain of which was replaced by that of FcRg (FcaRI/FcRg chimeras) expressed in RBL transfectants, the authors demonstrated that inhibition depended on the FcRg ITAM, and that both tyrosines were required for inhibition. These tyrosines were phosphorylated following monovalent engagement of FcaRI/ FcRg chimeras, but to a much lower extent than following plurivalent engagement. Inhibition was a slow process, taking 6 hrs to be complete. Interestingly, inhibition induced by monovalent ligands was correlated with the coprecipitation of SHP‐1 with weakly phosphorylated FcaRI/FcRg chimeras. Indeed, SHP‐1 did not detectably coprecipitate with chimeras that were heavily phosphorylated following cell activation induced by multivalent ligands. Finally, the coprecipitation of SHP‐1 with FcaRI/FcRg chimeras was dose‐dependently inhibited by a MEK inhibitor, suggesting a positive role of Erk in SHP‐1 recruitment. Intriguingly, when engaged by monovalent Fab fragments of a mAb against the extracellular domain of FcgRIIB, FcgRIIB/FcRg chimeras failed to inhibit IgE‐induced mediator release in the same cells, suggesting that, beside the intracytoplasmic ITAM, the ligand and/or the extracellular domain of the chimera were critical for inhibition (Pasquier et al., 2005). Also, inhibition is unlikely to depend on the mere membrane recruitment of SHP‐1. IgE‐induced mediator release and intracellular signaling were indeed not impaired in RBL transfectants expressing FcgRIIB whose intracytoplasmic domain had been replaced by the catalytic domain of SHP‐1 (Hardre´‐Lieńard and Dae¨ron, unpublished data). Whatever the mechanism of inhibition, these results have several important implications. First, they support the evidence that, although not able to fully activate cells, interactions of ITAM‐containing immunoreceptors with monovalent ligands can generate intracellular signals. Second, they indicate that FcaRI can generate either positive or negative signaling depending on extracellular ligands available (i.e., depending on whether IgA are in complexes with specific antigen or not). Whether other ITAM‐containing receptors may exert similar dual functions or whether it is a unique feature of FcaRI is not

60


known. FceRI do not seem to inhibit cell activation by other ITAM‐containing receptors when occupied by monomeric IgE as they are under physiologic conditions. Third, they suggest that FcaRI may negatively regulate activation signals triggered by many other receptors. Negative regulation by ITAM‐ containing receptors apparently did not require that inhibitory and activating receptors be coaggregated at the cell surface. If this conclusion proves to be correct, one can expect that many biological responses be affected by monovalent ligand‐induced negative regulation by FcaRI. SHP‐1 can indeed inhibit most if not all activation processes triggered by receptors whose signaling depends on tyrosyl‐phosphorylation of proteins. Finally, these findings provide a possible explanation and molecular basis to the paradox that, although IgA receptors can activate inflammatory cells (Patry et al., 1995), IgA have long been known to have general anti‐inflammatory effects (Russell et al., 1997) and to the observation that selective IgA deficiencies are correlated with increased susceptibility to autoimmune and allergic diseases (Schaffer et al., 1991). 4. Negative Signaling by Inhibitory FcRs By contrast with FcaRI‐dependent negative regulation, FcgRIIB‐dependent negative regulation requires that the inhibitory receptors be coaggregated with activating receptors by a common extracellular ligand and affects cell signaling triggered by these receptors. 4.1. Inhibitory FcRs and ITIMs The inhibitory properties of FcgRIIB lie on the presence of an ITIM in their intracytoplasmic domain. First identified in FcgRIIB (Dae¨ron et al., 1995a), ITIMs were subsequently found in a large number of inhibitory receptors that control the biologic activities of hematopoietic cells (Long, 1999). Sequence alignments of these ITIMs made it possible to define ITIMs structurally. ITIMs consist of a sequence containing a single tyrosine (Y) followed by a hydrophobic residue (I, V, or L) at position Y þ 3 and preceeded by a less conserved hydrophobic residue at position Y 2 (Vivier and Dae¨ron, 1997). One consequence of the coaggregation of FcgRIIB with activating receptors is the phosphorylation of their ITIM. FcgRIIB are not tyrosyl‐ phosphorylated when aggregated at the cell surface. They become phosphorylated when they are coaggregated with activating immunoreceptors (D’Ambrosio et al., 1995) because these provide the src kinase which phosphorylates both ITAMs and ITIMs in receptor coaggregates (Malbec et al., 1998). Due to this peculiarity, FcgRIIB are not inhibitory in resting cells. They do not establish a threshold that must be overcome by activating receptors.


61

They become functional ‘‘upon request’’ only, when cell activation has been launched. The phosphorylation of the FcgRIIB ITIM is indeed critical to initiate negative regulation. 4.2. The Recruitment of SHIP1 by FcgRIIB Inhibitory receptors carrying phosphorylated ITIMs (pITIMs) were shown to recruit SH2 domain‐containing cytosolic phosphatases that interfere with signals transduced by ITAM‐bearing receptors (Bolland and Ravetch, 1999). Four such phosphatases have been identified in mice and in humans: the two‐ SH2 domain‐containing Protein Tyrosine Phosphatases SHP‐1 and SHP‐2 and the single‐SH2 domain‐containing inositol 5‐phosphatases SHIP1 and SHIP2. Phosphorylated ITIMs differ from phosphorylated ITAMs by their specificity for SH2‐containing molecules. ITIMs recruit phosphatases only, whereas ITAMs recruit protein tyrosine kinases, adapter molecules, and phosphatases. FcgRIIB were found to differ from other ITIM‐containing receptors by being capable of recruiting SHIP1 and SHIP2. The FcgRIIB ITIM indeed has an affinity for the SH2 domain of SHIPs that other ITIMs lack. Our investigation of the bases of this unique specificity identified several parameters as being critical for SHIP1 to be recruited by FcgRIIB. 4.2.1 The Y þ 2 Leucine Determines the Affinity of the FcgRIIB ITIM for SHIP1/2 First of all, the affinity of FcgRIIB for SHIPs depends on a specific amino acid at position Y þ 2 in the ITIM. As expected from studies that established the molecular bases of the affinity of SH2 domains of other molecules for tyrosyl‐ phosphorylated peptides, the affinity of pITIMs for the SH2 domains of these phosphatases required the conservation of both the Y and the Y þ 3 residues. Synthetic peptides corresponding to pITIMs of all ITIM‐bearing molecules were found to bind SHP‐1 and SHP‐2 in vitro (Burshtyn et al., 1996; D’Ambrosio et al., 1995). The in vitro binding of SHP‐1 and SHP‐2 to the pITIMs of KIR2DL3 and FcgRIIB depends on the Y 2 residue (Ve´ly et al., 1997). Phosphorylated peptides corresponding to the FcgRIIB ITIM, but not phosphorylated peptides corresponding to the KIR2DL3 ITIMs, bound also SHIP1 and SHIP2 (Muraille et al., 2000; Ono et al., 1996). To identify the SHIP‐binding site in FcgRIIB, we exchanged residues between the FcgRIIB ITIM and the N‐terminal ITIM of KIR2DL3. Loss‐of‐function and gain‐of‐ function substitutions identified the Y þ 2 leucine, in the FcgRIIB ITIM, as determining the binding of both SHIP1 and SHIP2, but not the binding of SHP‐1 or SHP‐2. Conversely, the Y 2 isoleucine that determines the in vitro

62


binding of SHP‐1 and SHP‐2 affected neither the in vitro binding nor the in vivo recruitment of SHIP1 or SHIP2 (Bruhns et al., 2000). One hydrophobic residue in the ITIM of FcgRIIB therefore determines the affinity for SHIPs. This residue is symmetrical to another hydrophobic residue that determines the affinity of all ITIMs for SHPs. It defines a SHIP‐binding site, distinct from a SHP‐binding site, which confers FcgRIIB their ability to recruit SHIP1 and SHIP2. 4.2.2. The Density of pITIM Determines the Selective Recruitment of SHIP1/2 by FcgRIIB Intriguingly, these two binding sites are not used in vivo. Although agarose beads coated with phosphorylated peptides corresponding to the FcgRIIB ITIM bind in vitro both SHIP1/2 and SHP‐1/2, phosphorylated FcgRIIB, recruit selectively SHIP1/2 in vivo (Fong et al., 1996; Muraille et al., 2000; Ono et al., 1996). When investigating the reasons for this discordance, we found that beads coated with low amounts of pITIM bound SHIP1, but not SHP‐1, i.e., they behaved in vitro like phosphorylated FcgRIIB in vivo. The same was found when examining the binding of pITIM‐coated beads to GST fusion proteins containing the SH2 domain of SHIP1 or the two SH2 domains of SHP‐1. The reason is that the affinity of the SH2 domain of SHIP1 is high enough for binding to pITIM‐coated beads, but not that of either the N‐ or the C‐terminal SH2 domain of SHP‐1 (Lesourne et al., 2001). SHP‐1 indeed requires its two SH2 domains to bind to two pITIMs that are close enough to enable a cooperative interaction. This condition is fulfilled in vitro when beads are coated with sufficient amounts of pITIMs or in vivo when two tandem pITIMs are present in the intracytoplasmic domain of inhibitory receptors such as KIR2DL3. The deletion (Bruhns et al., 1999) or the mutation (Burshtyn et al., 1996) of either ITIM indeed abrogated the ability of KIR2DL3 to recruit SHP‐1. This is not fulfilled by FcgRIIB when coaggregated with activating receptors. When trying to increase FcgRIIB phosphorylation in B cells and mast cells, we found that concentrations of extracellular ligands optimal for FcgRIIB phosphorylation failed to induce the recruitment of SHP‐1. SHP‐1 was, however, recruited by FcgRIIB when the receptors were hyperphosphorylated following cell treatment with pervanadate (Lesourne et al., 2001). These data suggest that, although it can be reached under non‐physiological conditions, a high enough level of FcgRIIB phosphorylation may not be reached, under physiological conditions, to enable the in vivo recruitment of SHP‐1. Whether a regulatory mechanism limits the phosphorylation of FcgRIIB and whether (pathological?) conditions that would lead to the hyperphosphorylation of FcgRIIB might enable the


63

recruitment of SHP‐1 that would dephosphorylate signaling molecules are interesting possibilities that remain to be demonstrated. 4.2.3. The Recruitment of SHIP1 by FcgRIIB Requires the Cooperative Recruitment of Cytosolic Adapters Surprisingly, we found that, although sufficient for binding SHIP1 or SHIP2 in vitro, the FcgRIIB pITIM is not sufficient for the receptors to recruit these phosphatases in vivo. It is a general consensus that the FcgRIIB ITIM is both necessary and sufficient for inhibition of cell activation. The conclusion that it is necessary was based on the pioneer work by Amigorena and coworkers who showed that a 13‐amino acid deletion, which was later understood to encompass the ITIM, abrogated inhibition in B cells (Amigorena et al., 1992). A point mutation of the ITIM tyrosine also abrogated FcgRIIB‐dependent inhibition of mast cell and T cell activation (Dae¨ron et al., 1995a), and abolished (Muta et al., 1994) or reduced (Fong et al., 2000) the calcium response in B cells. The conclusion that the ITIM is sufficient was based on works by Muta and coworkers who showed that a chimeric molecule whose intracytoplasmic domain contained the murine FcgRIIB ITIM retained inhibitory properties in B cells (Muta et al., 1994). A C‐terminal deletion of the intracytoplasmic domain of murine FcgRIIB, which left the ITIM intact, however, prevented SHIP1 from being detectably coprecipitated, and reduced the inhibitory effect of FcgRIIB on BCR signaling (Fong et al., 2000). Our recent study showed that this C‐terminal sequence contains a second tyrosine‐based motif that mediates the recruitment of the cytosolic adapter proteins Grb2 and Grap via their SH2 domain and that contributes to the recruitment of SHIP1. The recruitment of the phosphatase indeed required an intact adapter‐binding motif and, conversely, the recruitment of adapters required an intact phosphatase‐binding motif. The reason is that Grb2 and Grap are constitutively associated with SHIP1 via their C‐terminal SH3 domain, and this association increases upon coaggregation of BCR with FcgRIIB. Grb2/Grap thus form a tri‐molecular complex with SHIP1 and FcgRIIB. This stabilizes the binding of the phosphatase to the ITIM and enables its recruitment by murine FcgRIIB. Supporting this conclusion, SHIP1 failed to coprecipitate with FcgRIIB, when tyrosyl‐phosphorylated upon coligation with BCR in mutant DT40 cells lacking both Grb2 and Grap (Isnardi et al., 2004). This requirement may not be peculiar to the interactions between FcgRIIB1, SHIP1, and Grb2. As discussed above, molecules that contain two SH2 domains require the cooperative binding of these two domains to two sequences containing phosphorylated tyrosines in order to be recruited in vivo. The recruitment of ZAP‐70 and Syk (Bu et al., 1995; Kurosaki et al., 1995), or SHP‐1 (Lesourne et al., 2001),

64


required the conservation of their two SH2 domains and the conservation of the two tyrosines of ITAMs in immunoreceptors (Kimura et al., 1996) or of the two ITIMs in KIRs (Bruhns et al., 1999; Burshtyn et al., 1999), respectively. Moreover, molecules that contain a single SH2 domain were found to require the cooperation of other SH2 domain‐containing molecules in order to be recruited (Yamasaki et al., 2003). One can therefore propose that one SH2 domain alone may not be sufficient for enabling stable interactions between signaling molecules. 4.3. SHIP1 Accounts for FcgRIIB‐Dependent Negative Regulation 4.3.1. SHIP1 is Necessary and Sufficient for FcgRIIB‐Dependent Negative Regulation Once it has been stably recruited, SHIP1 is the effector of FcgRIIB‐ dependent negative regulation. Evidence supporting this conclusion is as follows. FcgRIIB‐dependent negative regulation was abolished in cultured mast cells derived from the bone marrow of SHIP1‐deficient mice (Malbec et al., 2001), but not in mast cells derived from the bone marrow of moth‐eaten mice which are deficient in SHP‐1 (Fong et al., 1996). FcgRIIB‐dependent inhibition of Ca2þ mobilization was abolished in SHIP1‐deficient chicken DT40 B cells, but not in SHP‐1‐deficient (Ono et al., 1997) or in SHP‐2‐ deficient (Isnardi et al., unpublished observation) DT40 cells. Noticeably, FcgRIIB‐dependent inhibition was only reduced in B cells from SHIP1‐ deficient mice (Brauweiler et al., 2000), possibly because SHIP‐2 could partially replace SHIP1. Although also present in mast cells, SHIP2 could, however, not mediate FcgRIIB‐inhibition in SHIP1‐deficient mast cells. Inhibition was also partially reduced in B cells from moth‐eaten mice (D’Ambrosio et al., 1995). One possible reason is that SHP‐1‐deficient B cells are constitutively hyper‐activated (Pani et al., 1995), which might make BCR‐dependent signaling more difficult to inhibit. These data indicate that SHIP1 is necessary for FcgRIIB‐dependent inhibition of mast cell activation and, most probably, of B cell activation. Evidence that SHIP1 is also sufficient is as follows. B cell (Ono et al., 1997) and mast cell (Malbec et al., 2001) activation were comparably inhibited when BCRs or FceRI were coaggregated with wt FcgRIIB or with FcgRIIB whose intracytoplasmic domain had been replaced by the catalytic domain of SHIP1. In an analysis of a series of FcgRIIB‐SHIP chimeras, we found that, when coaggregated with BCR in the FcgR‐deficient cell line IIA1.6, SHIP1 chimeras abolished IL‐2 secretion, Ca2þ mobilization, Akt phosphorylation, and Erk1/2 phosphorylation. Under the same conditions, SHIP2 chimeras inhibited Akt phosphorylation, but did not affect Erk1/2


65

phosphorylation, Ca2þ mobilization, and IL‐2 secretion (Hardre´‐Lieńard and Dae¨ron, unpublished data). 4.3.2. Two Effector Mechanisms Are Used By SHIP1 in FcgRIIB‐Dependent Negative Regulation SHIP1 mediates FcgRIIB‐dependent inhibition by at least two distinct mechanisms. One depends on its catalytic activity, the other does not. By dephosphorylating PI(3,4,5)P3, SHIP1 prevents the recruitment of PH domain‐containing molecules such as PKB/Akt. The serine/threonine phosphorylation of PKB/Akt observed following BCR or FceRI aggregation was indeed abrogated upon coaggregation of these immunoreceptors with FcgRIIB (Jacob et al., 1999; Malbec et al., 2001). PKB/Akt phosphorylation depends on the membrane translocation of PKB/Akt and of PDK1, the responsible kinase. Both contain one PH domain which targets both the substrate and the enzyme to PI(3,4,5)P3‐rich membrane regions. PKB/Akt phosphorylation is therefore an indirect means to estimate the amount of membrane PI(3,4,5)P3 (Carver et al., 2000). Supporting this approximation, when transfected into B cells, a GFP construct containing the PH domain of Akt that is diffusely distributed in the cytosol of resting cells, translocates to the membrane following BCR aggregation. This translocation was prevented when BCR were coaggregated with FcgRIIB (Astoul et al., 1999). PKB/Akt phosphorylation is critical for mechanisms that prevent apoptosis. Although useful to assess PI(3,4,5)P3 degradation, and although it was recently reported to promote IgG immune complex‐induced phagocytosis in murine macrophages (Ganesan et al., 2004), PKB/Akt is not known to be a major player in signaling pathways leading to cell activation. PLC‐g and Tec kinases are. Like PKB/Akt, PLC‐g and Tec kinases contain a PH domain which mediates or contributes to their membrane recruitment via PI(3,4,5)P3. When translocated to the membrane, Tec kinases are thought to be tyrosyl‐phosphorylated/activated by Lyn and, together with Syk, to phosphorylate PLC‐g. The mechanism by which SHIP1 can negatively regulate the activity of Tec kinases was recently documented. SHIP1, as well as SHIP2, were reported to bind preferentially to the Tec kinase itself, and to inhibit its activity. Binding occurs through the SH3 domain of Tec, and mutations of this domain generated a hyperactive form of Tec. Constitutively active Tec could also be generated by introducing mutations that targeted this kinase to the membrane. Since Tec activity is positively regulated by its membrane localization, mostly via its recruitment to PI(3,4,5)P3, it was proposed that, by hydrolyzing PI(3,4,5)P3, SHIP1/2 could prevent the membrane recruitment and, hence, the activation of Tec (Tomlinson et al., 2004). This explanation of the inhibition of Ca2þ responses observed upon coaggregation of FcgRIIB

66


with immunoreceptors and the Fyn/Gab2/PI3K pathway that was described in mast cells (Parravincini et al., 2002) are not readily compatible. This Fyn‐ initiated pathway leads to the generation of PI(3,4,5)P3 by PI3K, whereas the Lyn/Syk/LAT/PLC‐g leads to Ca2þ mobilization. The mechanism of SHIP1‐ mediated FcgRIIB‐dependent inhibition of the Ca2þ response is more difficult to understand if the substrate of SHIP1 does not belong to the same pathway as that which leads to PLC‐g activation. These apparently conflicting data may be reconciled if one considers that bridges exist between the two pathways as suggested by the decreased phosphorylation of PLC‐g observed in Gab2/ mice (Gu et al., 2001). PLC‐g is indeed recruited both by PI(3,4,5)P3 and by LAT, as well as Btk, via Gads and SLP76. PI3K is recruited both by Gab2 and, via Gads, by LAT (Schraven et al., 1999). The coaggregation of FcgRIIB with immunoreceptors markedly inhibits the phosphorylation/activation of MAP kinases. SHIP1‐dependent PI(3,4,5)P3 degradation may affect the recruitment of the exchange factor Vav, which is translocated to the membrane via its PH domain, and the subsequent generation of Rac‐GTP that leads to the activation of JNK and p38. Inhibition of Erk1/2 activation also depends on SHIP1. It, however, does not depend on the phosphatase activity of the enzyme. SHIP has a tyrosine‐rich C‐terminal segment which contains NPXY motifs. It is constitutively tyrosyl‐phosphorylated. It is further phosphorylated following immunoreceptor‐dependent cell activation, and even further when recruited by FcgRIIB. The responsible kinase is thought to be Lyn. The phosphorylation of SHIP1 does not affect its enzymatic activity, but it confers this phosphatase the properties of an adapter molecule which can affect positive signals, independently of its catalytic activity. This conclusion stemmed from the observation that the adapter molecule Dok‐1 becomes heavily phosphorylated following the coaggregation of BCR with FcgRIIB in murine B cells (Tamir et al., 2000). Dok‐1 is a member of a family of adapter proteins that are tyrosyl‐phosphorylated upon engagement of a variety of cytokine receptors, growth factor receptors, and immunoreceptors. Dok phosphorylation depends on its membrane recruitment, and membrane‐targeted Dok‐1 was constitutively phosphorylated. Dok‐1 can be phosphorylated by Lyn or by Tec. Stem Cell factor‐induced Dok‐1 phosphorylation was, however, prevented in mast cells derived from Lyn/ mice, indicating that Lyn is primarily responsible for Dok‐1 phosphorylation in these cells (Liang et al., 2002). When tyrosyl‐phosphorylated, Dok‐1 recruits a variety of SH2 domain‐containing molecules including rasGAP which negatively regulates Ras activation. Dok‐1 contains an N‐terminal PH domain, a PTB domain, and a proline/tyrosine‐rich C‐terminal sequence. The role of Dok‐1 in FcgRIIB‐dependent negative regulation was analyzed using chimeric molecules made by replacing the intracytoplasmic domain of


67

FcgRIIB by the PH and PTB domain‐containing N‐terminal half of Dok‐1 or the proline/tyrosine‐rich C‐terminal half of Dok‐1. SHIP1 coprecipitated with the N‐terminal Dok chimera, whereas rasGAP coprecipitated with the C‐terminal Dok chimera when chimeras were coaggregated with BCR (Tamir et al., 2000). Ras‐GAP contains an SH2, an SH3, another SH2, and a PH domain, followed by a catalytic domain which can enhance the auto‐catalytic activity of ras‐GTP. As a consequence, Ras‐GTP is converted into RasGDP, and the Ras pathway is extinguished. Indeed, Erk1/2 activation seen upon BCR aggregation was inhibited upon coaggregation of BCR with the C‐terminal Dok chimera, but not with the N‐terminal Dok chimera (Tamir et al., 2000). Based on these data, it was proposed that, when recruited by FcgRIIB and tyrosyl‐phosphorylated, SHIP1 recruits Dok‐1 via the PTB domain of the latter. Dok‐1 becomes tyrosyl‐phosphorylated and recruits rasGAP via the SH2 domain of the latter. rasGAP turns Ras off and prevents the activation of Erk1/2. Similar results were observed when FcgRIIB were coaggregated with FceRI in mast cells (Ott et al., 2002). Supporting this scenario, MAP kinase activation was enhanced in response to BCR aggregation, and inhibition of cell proliferation in response to the coaggregation of BCR with FcgRIIB was abolished in B cells from Dok‐1‐deficient mice (Yamanashi et al., 2000). 4.4. FcgRIIB Amplify the Autonomous Negative Regulation of Activating FcRs 4.4.1. FcgRIIB‐Dependent Negative Regulation of FceRI Signaling Does not Occur in Lipid Rafts Lipid rafts are cholesterol/glycosphygolipid‐rich membrane micro‐domains (Brown and London, 2000; Horejsi, 2003) that diffuse laterally within the plasma membrane (Pralle et al., 2000). They play a critical role in positive signaling by FceRI. Disruption of rafts, using cholesterol‐depleting drugs, dramatically decreases early phosphorylation events induced upon FceRI aggregation (Sheets et al., 1999). According to a current model, FceRI are excluded from rafts in resting mast cells, whereas signaling proteins that are covalently associated with saturated fatty acids, such as Lyn (Young et al., 2003) and LAT (Zhang et al., 1998b), are concentrated in these domains. Upon aggregation, a fraction of FceRI transiently translocate into rafts (Field et al., 1997), bringing FceRI and raft‐associated signaling proteins close to each other. Kono and coworkers reported that FcgRIIB can translocate into lipid rafts upon aggregation in RBL‐2H3 cells (Kono et al., 2002) and Aman and coworkers reported that, when coaggregated with BCRs in A20 lymphoma

68


B cells, FcgRIIB recruited SHIP1 preferentially in low‐density detergent‐ resistant membrane compartments (Aman et al., 2000). We failed to observe a detectable translocation of FcgRIIB into lipid rafts, when coaggregated with FceRI. Actually the coaggregation of FcgRIIB with FceRI partially inhibited the translocation of FceRI into lipid rafts. The recruitment of SHIP1 by FcgRIIB is therefore not likely to take place in lipid rafts in mast cells. Because FcgRIIB are phosphorylated by the raft‐associated protein tyrosine kinase Lyn upon coaggregation with FceRI (Malbec et al., 1998), FcgRIIB may, however, transiently translocate into rafts where they are possibly phosphorylated. 4.4.2. FcgRIIB Associate with the Submembranous F‐Actin Skeleton When analyzing the contents of subcellular fractions prepared from RBL‐2H3 cells, we observed that FcgRIIB and SHIP1 were located in different subcellular compartments in resting cells. Following cell disruption in hypotonic buffer, differential centrifugation and solubilization of resulting fractions, most, if not all, FcgRIIB were indeed recovered in membrane fraction, whereas SHIP1 was recovered in the cytosolic and in the F‐actin skeleton fractions. The submembranous F‐actin skeleton, which connects F‐actin‐associated proteins with membrane proteins and phospholipids (Luna and Hitt, 1992), is another subcellular compartment. Unlike rafts, the submembranous F‐actin skeleton is not critical for FceRI‐dependent positive signaling. Rather, it seems to be involved in constitutive negative regulation of FceRI signaling. Indeed, drugs such as latrunculin, which prevent actin polymerization, enhance mast cell degranulation (Frigeri and Apgar, 1999). Interestingly, inhibition of degranulation observed in excess of antigen was markedly reduced in cells treated with latrunculin B, and actin could coprecipitate with SHIP1 in BMMC (Gimborn et al., 2005). Since FcgRIIB inhibit mast cell activation by recruiting SHIP1, the two molecules must meet somewhere. We found that, when coaggregated with FceRI, FcgRIIB heavily translocated into the F‐actin skeleton compartment. This translocation did not require that FcgRIIB be coaggregated with FceRI as FcgRIIB were similarly translocated upon aggregation by specific ligands. Surprisingly, it did not require either the intracytoplasmic domain of FcgRIIB as tail‐less FcgRIIB behaved similarly as intact receptors. Like FcgRIIB, FceRI were found in the membrane fraction in resting cells and, albeit in lower proportions, they dose‐dependently translocated into the F‐actin skeleton fraction when aggregated by IgE and antigen. The coaggregation with FcgRIIB did not increase but facilitated FceRI translocation which reached comparable levels at lower concentrations of antigen. Since tail‐less FcgRIIB


69

could enhance the translocation of FceRI into the F‐actin skeleton fraction but failed to inhibit mast cell activation, when coaggregated with FceRI (Lesourne et al., 2005), this effect of FcgRIIB on FceRI cannot alone account for negative regulation. 4.4.3. FcgRIIB Concentrate SHIP1 Close to FceRI Signaling Complexes in the F‐Actin Skeleton Filamin 1 is an actin‐binding protein that was previously reported to associate with SHIP2 in platelets (Dyson et al., 2001, 2003). We found that SHIP1 and Filamin 1 were recovered in the same sub‐cellular fractions as SHIP1 and that SHIP1 coprecipitated with filamin 1 in unstimulated RBL‐2H3 cells. Noticeably, the high‐molecular weight isoform of SHIP1 was predominant in the F‐actin skeleton fraction and it preferentially coprecipitated with filamin 1, whereas the two main SHIP1 isoforms were equally distributed in the cytosolic fraction. SHIP2 was proposed to associate with filamin via its proline‐rich C‐terminal region that is conserved in high‐molecular weight isoforms of SHIP1, but is spliced out in low‐molecular weight isoforms. Interestingly, the high‐molecular weight isoform of SHIP1 also preferentially coprecipitated with phosphorylated FcgRIIB, following their coaggregation with FceRI. These data altogether suggested that FcgRIIB could recruit filamin‐bound SHIP1 in the submembranous F‐actin skeleton compartment. This possibility was examined in intact cells by confocal microscopy. Upon coaggregation, FcgRIIB and FceRI rapidly formed small FcR patches on the plasma membrane. Both SHIP1 and filamin 1, but not F‐actin, co‐patched with FcRs. As the size of patches enlarged with time, higher amounts of SHIP1 colocalized with FcR patches. Surprisingly, filamin 1, as well as F‐actin, were excluded from large FcR patches (Lesourne et al., 2005). Based on these data, we propose a dynamic model according to which the translocation of FcgRIIB into the cytoskeleton enables these receptors to meet filamin‐bound SHIP1. The high‐avidity cooperative interactions between SHIP1, Grb2, and FcgRIIB are likely to displace SHIP1 from filamin and to concentrate the phosphatase in FcR signaling complexes. Supporting this critical role of the cytoskeleton, FcgRIIB‐dependent negative regulation of IgE‐induced mediator release was markedly reduced in latrunculin B‐treated cells. As for the exclusion of filamin and F‐actin from large FcR patches, one may hypothesize that the increased local degradation of PI(3,4,5)P3 by SHIP1 might decrease the rate of actin polymerization. Actin is indeed constantly polymerized and depolymerized and actin polymerization depends on PI3K (Bhargavi et al., 1998). Finally, we propose that FcgRIIB negatively regulate FceRI signaling by two mechanisms. First, they facilitate the translocation of FceRI into the F‐actin skeleton

70


compartment, thus enhancing SHIP1‐dependent constitutive negative regulation of FceRI at low antigen concentrations. Second, FcgRIIB concentrate SHIP1 in the vicinity of FceRI. Supporting this interpretation, SHIP1 readily coprecipitates with phosphorylated FcgRIIB but not with with FceRI. It follows that FcgRIIB act as amplifiers of SHIP1‐dependent constitutive negative regulation of FceRI signaling. 5. Conclusion FcRs are critical molecules of the immune system as they mediate most biological activities of the main effectors of the so‐called humoral immunity, that is, antibodies. Because they are ubiquitously expressed (mostly, but not only) by cells of hematopoietic origin, and because antibodies circulate in the blood stream, FcRs are involved in a wide array of biological activities in physiology. They also contribute to a variety of pathological processes. FcRs can trigger the release of potentially harmful—in some cases, life‐ threatening—inflammatory mediators, and induce destructive cytotoxic mechanisms, but (or therefore?) their activating properties are tightly controlled by regulatory mechanisms. As a consequence, immune responses are normally nonpathogenic. These regulatory mechanisms are primarily based on negative signaling that counterbalances positive signaling. Several levels of negative regulation can act on a given activating FcR. Negative regulation depends on different molecular mechanisms that may be used sequentially, depending on the conditions. A critical condition is the aggregation state of FcRs. Protein tyrosine phosphatase‐dependent negative regulation operates in resting cells when multi‐subunit FcRs are expressed on the plasma membrane and not yet engaged by any ligand (Fig. 1A). SHIP1‐ dependent negative regulation operates in mast cells whose FceRI are occupied by ‘‘monomeric’’ IgE (Fig. 1B). Unknown regulatory mechanisms account for the selective expression of some cytokine genes in mast cells exposed to IgE in the absence of antigen. Promiscuous SHP‐1‐dependent negative regulation is also triggered in cells whose FcaRI are occupied by monomeric IgA. Negative regulation involving multiple molecules that generate negative signals of different types operates as soon as positive signals are generated by activating FcRs. These include receptor subunits, kinases and phosphatases, and cytosolic and transmembrane adapter molecules. SHIP1 is a major player in the negative regulation that controls antigen‐induced IgE‐dependent mast cell activation (Fig. 1C). When further aggregated by supra‐optimal concentrations of ligand, FceRI associate with the F‐actin skeleton where the filamin 1‐bound high‐molecular weight isoform of SHIP1 resides. SHIP1 extinguishes positive signals and prevents mediator release. One, however, does not know


71

which molecular interaction(s) enable its recruitment in FceRI signaling complexes (Fig. 1D). When they are co‐engaged by IgG immune complexes, FcgRIIB facilitate the association of FceRI with the F‐actin skeleton, and tyrosyl‐phosphorylated FcgRIIB recruit and concentrate high‐molecular weight SHIP1 in the signaling complex, where it dephosphorylates PI(3,4,5) P3, becomes C‐terminally tyrosyl‐phosphorylated, and recruits Dok‐1 (Fig. 1E). As consequences, both the Ca2þ response and the activation of MAP kinases are inhibited. Noticeably, negative signaling often uses molecules that are also involved in positive signaling. The ITAM‐containing FcR subunit FcRb generates positive signals that complement FcRg‐dependent signaling. It contributes to bring Lyn in the signalosome and, possibly SHIP1. Lyn phosphorylates not only FcR ITAMs and Syk, but also SHIP1, enabling this phosphatase to inhibit the Ras pathway via the sequential recruitment of Dok‐1 and rasGAP. Lyn phosphorylates also Cbp, enabling Csk to be recruited and to prevent Fyn from being activated and to lead to the activation of PI3K. Grb2 can be recruited via its SH2 domain by phosphorylated adapters such as LAT, NTAL, or Shc, in activating FcR signaling complexes and contribute to positive regulation, but also by FcgRIIB and contribute to negative regulation. It is constitutively associated, via its N‐terminal SH3 domain, with the exchange factor Sos which activates Ras, but also, via its C‐terminal SH3 domain, with SHIP1 which inhibits Ras. Grb2 can also interact, via its SH2 domain, with phosphorylated SHP‐1 which dephosphorylates signaling molecules. LAT is critical for positive TCR‐ and FcR‐dependent signaling but, as revealed by knock‐in mice expressing LAT with selective tyrosine mutations, it also contributes to generate negative signals. NTAL may function both as a LAT equivalent in B cells and as a LAT antagonist in mast cells and, in these cells, its overall dominant negative effect results from an integration of negative and positive signals. Noticeably, molecules involved in negative regulation such as SHP‐1 (Xie et al., 2000) and SHIP1 (Giallourakis et al., 2000) can also have positive effects when overexpressed. Finally, depending on the ligand valency—IgA alone or in complex with multivalent antigen—FcaRI, can either prevent or induce inflammatory responses. Altogether, data listed above lead to the conclusion that molecules have no biological functions, but biological properties only. What ultimately determines a ‘‘function’’ is the context in which a set of molecules interact in sequence with each other. This context depends on the organization of signaling complexes that transiently form and function in different subcellular compartments where different molecules reside or are translocated. As a consequence, and as learnt from the study of knock‐out mice, therapeutic approaches aiming at targeting any specific molecule can be expected to have

72


Figure 1 Five levels of negative regulation in FcR complexes. Molecules in black are primarily involved in the generation of positive signals, molecules in red are primarily involved in the generation of negative signals, molecules in blue are involved in the generation of both positive and negative signals. (A) Positive and negative regulation in resting cells. Protein tyrosine kinases


73

‘‘paradoxical’’ unwanted effects. An alternative is to act on the balance between positive and negative signaling in appropriate cells. Since most cells are constitutively equipped with both activating and inhibitory FcRs, these can be used as therapeutic tools. One way is to increase the expression of FcRs of one type or of the other. This is apparently what happens when intravenous immunoglobulins (IVIG) are administered and upregulate the expression of FcgRIIB (Bruhns et al., 2003). Another way is to bring more FcRs of one type into complexes of FcRs of the other type. In vitro and in vivo proofs of concepts were recently provided that one can favor negative regulation using bispecific synthetic molecules capable of co‐engaging FceRI and FcgRIIB on human mast cells and basophils, and reduce IgE‐dependent human mast cell activation (Tam et al., 2004), allergen‐induced systemic anaphylaxis, and airway hyper‐responsiveness in transgenic mice expressing human FceRI (Zhu et al., 2005). Similar approaches can be envisioned in other diseases requiring immune responses to be dampened. Conversely, other molecules can be tailored to favor positive regulation in pathological situations requiring immune responses to be boostered. For these approaches to develop and be mastered, further investigations are needed in order to understand what determines the ratio of activating and inhibitory FcRs expressed at the cell and protein tyrosine phosphatases constitutively phosphorylate and dephosphorylate, respectively, intracellular proteins. Possibly resulting activation signals do not lead to a detectable cellular reponse. (B) SHIP1 as a gatekeeper of mast cell activation. Positive signals triggered by IgE in the absence of antigen are constitutively negatively regulated by SHIP1. As a result, wt mast cells usually do not degranulate when sensitized with IgE and are not challenged with antigen, but SHIP1‐deficient mast cells do. (C) Negative signals generated together with positive signals by activating FcRs. Upon aggregation of activating FcRs by antibodies and multivalent antigens, both the Lyn/Syk/PLC‐g and the Fyn/Gab2/PI3K pathways are activated, leading to cell activation. These positive signals are counterbalanced by negative signals. By phosphorylating Cpb, Lyn enables Csk to be recruited and to inhibit Fyn. By phosphorylating SHIP1, Lyn enables Dok1 to be recruited and to inhibit Ras via rasGAP. SHIP1 is possibly recruited by phosphorylated FcRb. NTAL also negatively regulates FceRI signaling by not yet clear mechanisms. Biological responses of the cell results from the integration of these antagonistic signals. (D) Negative signals generated by activating FcRs in excess of ligand. When supra‐optimally engaged by an excess of ligand, FceRI aggregates associate with the F‐actin skeleton, where the high molecular isoform of SHIP1 is constitutively associated with Filamin 1. As a consequence, more SHIP1 is involved in negative regulation as indicated by its increased phosphorylation. The result is a dose‐dependent inhibition of degranulation. (E) Negative regulation by FcgRIIB. When coaggregated with FceRI, FcgRIIB are phosphorylated by Lyn, associate with the F‐actin skeleton, and recruit F‐actin‐associated SHIP1. The recruitment of SHIP1 involves the interactions of its SH2 domain with specific residues in the FcgRIIB phosphorylated ITIM and of its C‐terminal prolin‐rich region with the C‐terminal SH3 domain of Grb2 which, itself, binds to the phosphorylated C‐terminal tyrosine of FcgRIIB via its SH2 domain. FcgRIIB thus concentrate SHIP1 in FceRI signaling complex and, by inhibiting both the Ca2þ response and the activation of MAP Kinases, extinguish all cellular responses.

74


surface, whether activating or inhibitory FcRs can be preferentially engaged by antibodies, how FcRs generate positive and negative signals, and how these signals are integrated within cells. Acknowledgments Our works discussed in this review were supported in part by the Institut National de la Sante´ et de la Recherche Me´dicale (INSERM), the Fondation pour la Recherche Me´dicale (FRM), the Association pour la Recherche sur le Cancer (ARC), and the Institut Pasteur. RL was the recipient of fellowships from the ARC and from the Socie´te´ Française d’Allergologie et d’Immunologie Clinique (SFAIC).

References Adamczewski, M., Numerof, R. P., Koretzky, G. A., and Kinet, J. P. (1995). Regulation by CD45 of the tyrosine phosphorylation of high affinity IgE receptor b‐ and g‐chains. J. Immunol. 154, 3047–3055. 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. Aman, M. J., Walk, S. F., March, M. E., Su, H. P., Carver, D. J., and Ravichandran, K. S. (2000). Essential role for the C‐terminal noncatalytic region of SHIP in FcgRIIB1‐mediated inhibitory signaling. Mol. Cell. Biol. 20, 3576–3589. Amigorena, S., Bonnerot, C., Drake, J., Choquet, D., Hunziker, W., Guillet, J. G., Webster, P., Saute`s, C., Mellman, I., and Fridman, W. H. (1992). Cytoplasmic domain heterogeneity and functions of IgG Fc receptors in B‐lymphocytes. Science 256, 1808–1812. Asai, K., Kitaura, J., Kawakami, Y., Yamagata, N., Tsai, M., Carbone, D. P., Liu, F. T., Galli, S. J., and Kawakami, T. (2001). Regulation of mast cell survival by IgE. Immunity 14, 791–800. Astoul, E., Watton, S., and Cantrell, D. (1999). The dynamics of protein kinase B regulation during B cell antigen receptor engagement. J. Cell Biol. 145, 1511–1520. Benhamou, M., Bonnerot, C., Fridman, W. H., and Dae¨ron, M. (1990). Molecular heterogeneity of murine mast cell Fcg receptors. J. Immunol. 144, 3071–3077. Benhamou, M., Ryba, N. J., Kihara, H., Nishikata, H., and P., S. R. (1993). Protein tyrosine kinase p72syk in high‐affinity IgE receptor signaling. Identification as a component of pp72 and association with the receptor g chain after receptor aggregation. J. Biol. Chem. 268, 23318–23324. Berger, S. A., Mak, T. W., and Paige, C. J. (1994). Leukocyte common antigen (CD45) is required for immunoglobulin E‐mediated degranulation of mast cells. J. Exp. Med. 180, 471–476. Bhargavi, V., Chari, V. B., and Singh, S. S. (1998). Phosphatidylinositol 3‐kinase binds to profilin through the p85 a subunit and regulates cytoskeletal assembly. Biochem. Mol. Biol. Int. 46, 241–248. Boerth, N. J., Judd, B. A., and Koretzky, G. A. (2000). Functional association between SLAP‐130 and SLP‐76 in Jurkat T cells. J. Biol. Chem. 275, 5143–5152. Bolland, S., Pearse, R. N., Kurosaki, T., and Ravetch, J. V. (1998). SHIP modulates immune receptor responses by regulating membrane association of Btk. Immunity 8, 509–516. Bolland, S., and Ravetch, J. V. (1999). Inhibitory pathways triggered by ITIM‐containing receptors. Adv. Immunol. 72, 149–177.


75

Bolland, S., and Ravetch, J. V. (2000). Spontaneous autoimmune disease in FcgRIIB‐deficient mice results from strain‐specific epistasis. Immunity 13, 277–285. Borkowski, T. A., Jouvin, M. H., Lin, S. Y., and Kinet, J. P. (2001). Minimal requirements for IgE‐ mediated regulation of surface FceRI. J. Immunol. 167, 1290–1296. Brauweiler, A., Tamir, I., Dal Porto, J., Benschop, R. J., Helgason, C. D., Humphries, R. K., Freed, J. H., and Cambier, J. C. (2000). Differential regulation of B cell development, activation, and death by the src homology 2 domain‐containing 50 inositol phosphatase (SHIP). J. Exp. Med. 191, 1545–1554. Brdicka, T., Imrich, M., Angelisova, P., Brdickova, N., Horvath, O., Spicka, J., Hilgert, I., Luskova, P., Draber, P., Novak, P., Engels, N., Wienands, J., Simeoni, L., Osterreicher, J., Aguado, E., Malissen, M., Schraven, B., and Horejsi, V. (2002). Non‐T cell activation linker (NTAL): A transmembrane adaptor protein involved in immunoreceptor signaling. J. Exp. Med. 196, 1617–1626. Brooks, D. G., Qiu, W. Q., Luster, A. D., and Ravetch, J. V. (1989). Structure and expression of human IgG FcRII (CD32). Functional heterogeneity is encoded by the alternatively spliced products of multiple genes. J. Exp. Med. 170, 1369–1386. Brown, D. A., and London, E. (2000). Structure and Function of Sphingolipid‐ and Cholesterol‐ rich Membrane Rafts. J. Biol. Chem. 275, 17221–17224. Bruhns, P., Marchetti, P., Fridman, W. H., Vivier, E., and Dae¨ron, M. (1999). Differential roles of N‐ and C‐terminal ITIMs during inhibition of cell activation by killer cell inhibitory receptors. J. Immunol. 162, 3168–3175. Bruhns, P., Samuelsson, A., Pollard, J. W., and Ravetch, J. V. (2003). Colony‐stimulating factor‐1‐ dependent macrophages are responsible for IVIG protection in antibody‐induced autoimmune disease. Immunity 18, 573–581. Bruhns, P., Ve´ly, F., Malbec, O., Fridman, W. H., Vivier, E., and Dae¨ron, M. (2000). Molecular basis of the recruitment of the SH2 domain‐containing inositol 5‐phosphatases SHIP1 and SHIP2 by FcgRIIB. J. Biol. Chem. 275, 37357–37364. Bu, J. Y., Shaw, A. S., and Chan, A. C. (1995). Analysis of the interaction of ZAP‐70 and Syk protein‐tyrosine kinases with the T‐cell antigen receptor by plasmon resonance. P.NA.S. 92, 5106–5110. Buccione, R., Di Tullio, G., Caretta, M., Marinetti, M. R., Bizzarri, C., Francavilla, S., Luini, A., and De Matteis, M. A. (1994). Analysis of protein kinase C requirement for exocytosis in permeabilized rat basophilic leukaemia RBL‐2H3 cells: A GTP‐binding protein(s) as a potential target for protein kinase C. Biochem. J. 298(Pt 1), 149–156. Burns, C. M., Sakaguchi, K., Appella, E., and Ashwell, J. D. (1994). CD45 regulation of tyrosine phosphorylation and enzyme activity of src family kinases. J. Biol. Chem. 269, 13594–13600. Burshtyn, D. N., Lam, A. S., Weston, M., Gupta, N., Warmerdam, P. A., and Long, E. O. (1999). Conserved residues amino‐terminal of cytoplasmic tyrosines contribute to the SHP‐1‐mediated inhibitory function of killer cell Ig‐like receptors. J. Immunol. 162, 897–902. Burshtyn, D. N., Scharenberg, A. M., Wagtmann, N., Rajogopalan, S., Berrada, K., Yi, T., Kinet, J.‐P., and Long, E. O. (1996). Recruitment of tyrosine phosphatase HCP by the killer cell inhibitory receptor. Immunity 4, 77–85. Cantrell, D. (1998). Lymphocyte signalling: A coordinating role for Vav? Curr. Biol. 8, R535–R538. Cao, X., Wei, G., Fang, H., Guo, J., Weinstein, M., Marsh, C. B., Ostrowski, M. C., and Tridandapani, S. (2004). The inositol 3‐phosphatase PTEN negatively regulates Fc g receptor signaling, but supports Toll‐like receptor 4 signaling in murine peritoneal macrophages. J. Immunol. 172, 4851–4857. Carver, D. J., Aman, M. J., and Ravichandran, K. S. (2000). SHIP inhibits Akt activation in B cells through regulation of Akt membrane localization. Blood 96, 1449–1456.

76


Chan, V. W., Meng, F., Soriano, P., De Franco, A. L., and Lowell, C. A. (1997). Characterization of the B lymphocyte populations in Lyn‐deficient mice and the role of Lyn in signal initiation and down‐regulation. Immunity 7, 69–81. Clynes, R., Takechi, Y., Moroi, Y., Houghton, A., and Ravetch, J. V. (1998). Fc receptors are required in passive and active immunity to melanoma. Proc. Natl. Acad. Sci. USA 95, 652–656. Cole, P. A., Shen, K., Qiao, Y., and Wang, D. (2003). Protein tyrosine kinases Src and Csk: A tail’s tale. Curr. Opin. Chem. Biol. 7, 580–585. Costello, P. S., Turner, M., Walters, A. E., Cunningham, C. N., Bauer, P., Downward, J., and Tybulewicz, V. L. J. (1996). Critical role for the tyrosine kinase Syk in signalling through the high affinity IgE receptor of mast cells. Oncogene 13, 2595–2605. D’Ambrosio, D., Hippen, K. H., Minskoff, S. A., Mellman, I., Pani, G., Siminovitch, K. A., and Cambier, J. C. (1995). Recruitment and activation of PTP1C in negative regulation of antigen receptor signaling by FcgRIIB1. Science 268, 293–296. Dae¨ron, M. (1997). Fc Receptor Biology. Annu. Rev. Immunol. 15, 203–234. Dae¨ron, M., Bonnerot, C., Latour, S., and Fridman, W. H. (1992). Murine recombinant FcgRIII, but not FcgRII, trigger serotonin release in rat basophilic leukemia cells. J. Immunol. 149, 1365–1373. Dae¨ron, M., Latour, S., Malbec, O., Espinosa, E., Pina, P., Pasmans, S., and Fridman, W. H. (1995a). The same tyrosine‐based inhibition motif, in the intracytoplasmic domain of FcgRIIB, regulates negatively BCR‐, TCR‐, and FcR‐dependent cell activation. Immunity 3, 635–646. Dae¨ron, M., Malbec, O., Latour, S., Arock, M., and Fridman, W. H. (1995b). Regulation of high‐ affinity IgE receptor‐mediated mast cell activation by murine low‐affinity IgG receptors. J. Clin. Invest. 95, 577–585. Damen, J. E., Liu, L., Rosten, P., Humphries, R. K., Jefferson, A. B., Majerus, P. W., and Krystal, G. (1996). The 145‐kDa protein induced to associate with Shc by multiple cytokines is an inositol tetraphosphate and phosphatidylinositol 3,4,5‐trisphosphate 5‐phosphatase. Proc. Natl. Acad. Sci. USA 93, 1689–1693. Delisi, C., and Siraganian, R. P. (1979). Receptor cross‐linking and histamine release. II. Interpretation and analysis of anomalous dose‐response patterns. J. Immunol. 122, 2293–2299. Dembo, M., Goldstein, B., Sobotka, A., and Lichtenstein, L. (1978). Histamine release due to bivalent penicilloyl haptens: Control by the number of crosslinked IgE antibodies on the basophil plasma membrane. J. Immunol. 121, 354–358. Deusch, K., Pfeffer, K., Reich, K., Gstettenbauer, M., Daum, S., Luling, F., and Classen, M. (1991). Phenotypic and functional characterization of human TCR g dþ intestinal intraepithelial lymphocytes. Curr. Top. Microbiol. Immunol. 173, 279–283. Dhanji, S., Tse, K., and Teh, H. S. (2005). The low affinity Fc receptor for IgG functions as an effective cytolytic receptor for self‐specific CD8 T cells. J. Immunol. 174, 1253–1258. Dombrowicz, D., Lin, S., Flamand, V., Brini, A. T., Koller, B. H., and Kinet, J. P. (1998). Allergy‐ associated FcRb is a molecular amplifier of IgE‐ and IgG‐mediated in vivo responses. Immunity 8, 517–529. Dong, C., Davis, R. J., and Flavell, R. A. (2002). MAP kinases in the immune response. Annu. Rev. Immunol. 20, 55–72. Donnadieu, E., Jouvin, M. H., Rana, S., Moffatt, M. F., Mockford, E. H., Cookson, W. O., and Kinet, J. P. (2003). Competing functions encoded in the allergy‐associated FceRIb gene. Immunity 18, 665–674. Downward, J. (1996). Control of ras activation. Cancer Surv. 27, 87–100. Drechsler, Y., Chavan, S., Catalano, D., Mandrekar, P., and Szabo, G. (2002). FcgR cross‐linking mediates NF‐kappaB activation, reduced antigen presentation capacity, and decreased IL‐12 production in monocytes without modulation of myeloid dendritic cell development. J. Leukoc. Biol. 72, 657–667.


77

Duque, N., Gomez‐Guerrero, C., and Egido, J. (1997). Interaction of IgA with Fc a receptors of human mesangial cells activates transcription factor nuclear factor‐kappa B and induces expression and synthesis of monocyte chemoattractant protein‐1, IL‐8, and IFN‐inducible protein 10. J. Immunol. 159, 3474–3482. Dyson, J. M., Munday, A. D., Kong, A. M., Huysmans, R. D., Matzaris, M., Layton, M. J., Nandurkar, H. H., Berndt, M. C., and Mitchell, C. A. (2003). SHIP‐2 forms tetrameric complex with filamin, actin, and GPIb‐IX‐V: Localization of SHIP‐2 to the activated platelet actin cytoskeleton. Blood 102, 940–948. Dyson, J. M., O’ Malley, C. J., Becanovic, J., Munday, A. D., Berndt, M. C., Coghill, I. D., Nandurkar, H. H., Ooms, L. M., and Mitchell, C. A. (2001). The SH2‐containing inositol polyphosphate 5‐phosphatase, SHIP‐2, binds filamin and regulates submembraneous actin. J. Cell Biol. 155, 1065–1079. Field, K. A., Holowka, D., and Baird, B. (1997). Compartmentalized Activation of the High Affinity Immunoglobulin E Receptor within Membrane Domains. J. Biol. Chem. 272, 4276–4280. Fong, D. C., Brauweiler, A., Minskoff, S. A., Bruhns, P., Tamir, I., Mellman, I., Dae¨ron, M., and Cambier, J. C. (2000). Mutational analysis reveals multiple distinct sites within Fcg receptor IIb that function in inhibitory signaling. J. Immunol. 165, 4453–4462. Fong, D. C., Malbec, O., Arock, M., Cambier, J. C., Fridman, W. H., and Dae¨ron, M. (1996). Selective in vivo recruitment of the phosphatidylinositol phosphatase SHIP by phosphorylated FcgRIIB during negative regulation of IgE‐dependent mouse mast cell activation. Immunol. Lett. 54, 83–91. Frigeri, L., and Apgar, J. R. (1999). The role of actin microfilaments in the down‐regulation of the degranulation response in RBL‐2H3 mast cells. J. Immunol. 162, 2243–2250. Furumoto, Y., Nunomura, S., Terada, T., Rivera, J., and Ra, C. (2004). The FceRIb immunoreceptor tyrosine‐based activation motif exerts inhibitory control on MAPK and IkappaB kinase phosphorylation and mast cell cytokine production. J. Biol. Chem. 279, 49177–49187. Galandrini, R., Tassi, I., Mattia, G., Lenti, L., Piccoli, M., Frati, L., and Santoni, A. (2002). SH2‐ containing inositol phosphatase (SHIP‐1) transiently translocates to raft domains and modulates CD16‐mediated cytotoxicity in human NK cells. Blood 100, 4581–4589. Ganesan, L. P., Fang, H., Marsh, C. B., and Tridandapani, S. (2003). The protein‐tyrosine phosphatase SHP‐1 associates with the phosphorylated immunoreceptor tyrosine‐based activation motif of FcgRIIa to modulate signaling events in myeloid cells. J. Biol. Chem. 278, 35710–35717. Ganesan, L. P., Wei, G., Pengal, R. A., Moldovan, L., Moldovan, N., Ostrowski, M. C., and Tridandapani, S. (2004). The serine/threonine kinase Akt Promotes Fc g receptor‐mediated phagocytosis in murine macrophages through the activation of p70S6 kinase. J. Biol. Chem. 279, 54416–54425. Giallourakis, C., Kashiwada, M., Pan, P. Y., Danial, N., Jiang, H., Cambier, J., Coggeshall, K. M., and Rothman, P. (2000). Positive regulation of interleukin‐4‐mediated proliferation by the SH2‐ containing inositol‐50 ‐phosphatase. J. Biol. Chem. 275, 29275–29282. Gimborn, K., Lessmann, E., Kuppig, S., Krystal, G., and Huber, M. (2005). SHIP down‐regulates FceRI‐induced degranulation at supraoptimal IgE or antigen levels. J. Immunol. 174, 507–516. Gonzalez‐Espinosa, C., Odom, S., Olivera, A., Hobson, J. P., Martinez, M. E., Oliveira‐Dos‐Santos, A., Barra, L., Spiegel, S., Penninger, J. M., and Rivera, J. (2003). Preferential signaling and induction of allergy‐promoting lymphokines upon weak stimulation of the high affinity IgE receptor on mast cells. J. Exp. Med. 197, 1453–1465. 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.

78


Hazenbos, L. W., Gessner, J. E., Hofhuis, F. M. A., Kuipers, H., Meyer, D., Heijnen, I. A. F. M., Schmidt, R. E., Sandor, M., Capel, P. J. A., Dae¨ron, M., van de Winkel, J. G. J., and Verbeek, J. S. (1996). Impaired IgG‐dependent anaphylaxis and Arthus reaction in FcgRIII (CD16) deficient mice. Immunity 5, 181–188. Helgason, C. D., Damen, J. E., Rosten, P., Grewal, R., Sorensen, P., Chappel, S. M., Borowski, A., Jirik, F., Krystal, G., and Humphries, R. K. (1998). Targeted disruption of SHIP leads to hemopoietic perturbations, lung pathology, and shortened life span. Genes Dev. 12, 1610–1620. Hendricks‐Taylor, L. R., Motto, D. G., Zhang, J., Siraganian, R. P., and Koretzky, G. A. (1997). SLP‐76 is a substrate of the high affinity IgE receptor‐stimulated protein tyrosine kinases in rat basophilic leukemia cells. J. Biol. Chem. 272, 1363–1367. Hernandez‐Hansen, V., Mackay, G. A., Lowell, C. A., Wilson, B. S., and Oliver, J. M. (2004). The Src kinase Lyn is a negative regulator of mast cell proliferation. J. Leukoc. Biol. 75, 143–151. Herr, A. B., Ballister, E. R., and Bjorkman, P. J. (2003). Insights into IgA‐mediated immune responses from the crystal structures of human FcaRI and its complex with IgA1‐Fc. Nature 423, 614–620. Hibbs, M. L., Bonadonna, L., Scott, B. M., McKenzie, I. F. C., and Hogarth, P. M. (1988). Molecular cloning of a human immunoglobulin G Fc receptor. Proc. Natl. Acad. Sci. USA 85, 2240–2244. Hibbs, M. L., Harder, K. W., Armes, J., Kountouri, N., Quilici, C., Casagranda, F., Dunn, A. R., and Tarlinton, D. M. (2002). Sustained activation of Lyn tyrosine kinase in vivo leads to autoimmunity. J. Exp. Med. 196, 1593–1604. Hibbs, M. L., Walker, I. D., Kirszbaum, L., Pietersz, G. A., Deacon, N. J., Chambers, G. W., McKenzie, I. F. C., and Hogarth, P. M. (1986). The murine Fc receptor for immunoglobulin: Purification, partial amino acid sequence, and isolation of cDNA clones. Proc. Natl. Acad. Sci. USA 83, 6980–6984. Horejsi, V. (2003). The roles of membrane microdomains (rafts) in T cell activation. Immunol. Rev. 191, 148–164. Hsu, C., and Mac Glashan, D. J. (1996). IgE antibody up‐regulates high affinity IgE binding on murine bone marrow‐derived mast cells. Immunol. Letters 52, 129–134. Huber, M., Helgason, C. D., Damen, J. E., Liu, L., Humphries, R. K., and Krystal, G. (1998). The src homology 2‐containing inositol phosphatase (SHIP) is the gatekeeper of mast cell degranulation. Proc. Natl. Acad. Sci. USA 95, 11330–11335. Hulett, M. D., and Hogarth, P. M. (1994). Molecular basis of Fc Receptor function. Adv. Immunol. 57, 1–127. Humphries, L. A., Dangelmaier, C., Sommer, K., Kipp, K., Kato, R. M., Griffith, N., Bakman, I., Turk, C. W., Daniel, J. L., and Rawlings, D. J. (2004). Tec kinases mediate sustained calcium influx via site‐specific tyrosine phosphorylation of the phospholipase Cg Src homology 2‐Src homology 3 linker. J. Biol. Chem. 279, 37651–37661. Ishizaka, K., Tomioka, H., and Ishizaka, T. (1970). Mechanism of passive sensitization. I. Presence of IgE and IgG molecules on human leukocytes. J. Immunol. 105, 1459–1467. Isnardi, I., Lesourne, R., Bruhns, P., Fridman, W. H., Cambier, J. C., and Dae¨ron, M. (2004). Two distinct tyrosine‐based motifs enable the inhibitory receptor FcgRIIB to cooperatively recruit the inositol phosphatases SHIP1/2 and the adapters Grb2/Grap. J. Biol. Chem. 279, 51931–51938. Jackman, J. K., Motto, D. G., Sun, Q., Tanemoto, M., Turck, C. W., Peltz, G. A., Koretzky, G. A., and Findell, P. R. (1995). Molecular cloning of SLP‐76, a 76‐kDa tyrosine phosphoprotein associated with Grb2 in T cells. J. Biol. Chem. 270, 7029–7032. Jacob, A., Cooney, D., Tridandapani, S., Kelley, T., and Coggeshall, K. M. (1999). FcgRIIb modulation of surface immunoglobulin‐induced Akt activation in murine B cells. J. Biol. Chem. 274, 13704–13710.


79

Janas, M. L., Hodgkin, P., Hibbs, M., and Tarlinton, D. (1999). Genetic evidence for Lyn as a negative regulator of IL‐4 signaling. J. Immunol. 163, 4192–4198. Janssen, E., Zhu, M., Craven, B., and Zhang, W. (2004). Linker for activation of B cells: A functional equivalent of a mutant linker for activation of T cells deficient in phospholipase C‐g 1 binding. J. Immunol. 172, 6810–6819. Janssen, E., Zhu, M., Zhang, W., Koopnaw, S., and Zhang, W. (2003). LAB: A new membrane‐ associated adaptor molecule in B cell activation. Nature Immunol. 4, 117–123. Johannes, F. J., Horn, J., Link, G., Haas, E., Siemienski, K., Wajant, H., and Pfizenmaier, K. (1998). Protein kinase Cmu downregulation of tumor‐necrosis‐factor‐induced apoptosis correlates with enhanced expression of nuclear‐factor‐kappaB‐dependent protective genes. Eur. J. Biochem. 257, 47–54. Jouvin, M. H., Adamczewski, M., Numerof, R., Letourneur, O., Valle´, A., and Kinet, J. P. (1994). Differential control of the tyrosine kinases Lyn and Syk by the two signaling chains of the high affinity immunoglobulin E receptor. J. Biol. Chem. 269, 5918–5925. Kalesnikoff, J., Huber, M., Lam, V., Damen, J. E., Zhang, J., Siraganian, R. P., and Krystal, G. (2001). Monomeric IgE stimulates signaling pathways in mast cells that lead to cytokine production and cell survival. Immunity 14, 801–811. Kawakami, T., and Galli, S. J. (2002). Regulation of mast‐cell and basophil function and survival by IgE. Nat. Rev. Immunol. 2, 773–786. Kawakami, Y., Hartman, S. E., Holland, P. M., Cooper, J. A., and Kawakami, T. (1998). Multiple signaling pathways for the activation of JNK in mast cells: Involvement of Bruton’s tyrosine kinase, protein kinase C, and JNK kinases, SEK1 and MKK7. J. Immunol. 161, 1795–1802. Kawakami, Y., Kitaura, J., Satterthwaite, A. B., Kato, R. M., Asai, K., Hartman, S. E., Maeda‐ Yamamoto, M., Lowell, C. A., Rawlings, D. J., Witte, O. N., and Kawakami, T. (2000). Redundant and opposing functions of two tyrosine kinases, Btk and Lyn, in mast cell activation. J. Immunol. 165, 1210–1219. Kettner, A., Pivniouk, V., Kumar, L., Falet, H., Lee, J. S., Mulligan, R., and Geha, R. S. (2003). Structural requirements of SLP‐76 in signaling via the high‐affinity immunoglobulin E receptor (FceRI) in mast cells. Mol. Cell. Biol. 23, 2395–2406. Kihara, H., and Siraganian, R. P. (1994). Src homology 2 domains of Syk and Lyn bind to tyrosine‐ phosphorylated subunits of the high affinity IgE receptor. J. Biol. Chem. 269, 22427–22432. Kimura, T., Sakamoto, H., Apellea, E., and Siraganian, R. P. (1996). Conformational changes induced in the protein tyrosine kinase p72Syk by tyrosine phosphorylation or by binding of phosphorylatedimmunoreceptor tyrosine‐based activation motif peptides. Mol. Cell. Biol. 4, 1471–1478. Kimura, T., Sakamoto, H., Appella, E., and Siraganian, R. P. (1997). The negative signaling molecule SH2 domain‐containing inositol‐polyphosphate 5‐phosphatase (SHIP) binds to the tyrosine‐phosphorylated b subunit of the high affinity IgE receptor. J. Biol. Chem. 272, 13991–13996. Kitaura, J., Eto, K., Kinoshita, T., Kawakami, Y., Leitges, M., Lowell, C. A., and Kawakami, T. (2005). Regulation of Highly Cytokinergic IgE‐Induced Mast Cell Adhesion by Src, Syk, Tec, and Protein Kinase C Family Kinases. J. Immunol. 174, 4495–4504. Kleinau, S., Martinsson, P., and Heyman, B. (2000). Induction and suppression of collagen‐ induced arthritis is dependent on distinct fcg receptors. J. Exp. Med. 191, 1611–1616. Kliche, S., Lindquist, J. A., and Schraven, B. (2004). Transmembrane adapters: Structure, biochemistry and biology. Semin. Immunol. 16, 367–377. Kohno, M., Yamasaki, S., Tybulewicz, V. L., and Saito, T. (2005). Rapid and large amount of autocrine IL‐3 production is responsible for mast cell survival by IgE in the absence of antigen. Blood 105, 2059–2065.

80


Kono, H., Suzuki, T., Yamamoto, K., Okada, M., Yamamoto, T., and Honda, Z. (2002). Spatial raft coalescence represents an initial step in FcgR signaling. J. Immunol. 169, 193–203. Koonpaew, S., Janssen, E., Zhu, M., and Zhang, W. (2004). The importance of three membrane‐ distal tyrosines in the adaptor protein NTAL/LAB. J. Biol. Chem. 279, 11229–11235. Kovarova, M., Tolar, P., Arudchandran, R., Draberova, L., Rivera, J., and Draber, P. (2001). Structure‐function analysis of Lyn kinase association with lipid rafts and initiation of early signaling events after Fce receptor I aggregation. Mol. Cell. Biol. 21, 8318–8328. Kraft, S., Novak, N., Katoh, N., Bieber, T., and Rupec, R. A. (2002). Aggregation of the high‐affinity IgE receptor FceRI on human monocytes and dendritic cells induces NF‐kB activation. J. Invest. Dermatol. 118, 830–837. Kurosaki, T., Johnson, S. A., Pao, L., Sada, K., Yamamura, H., and Cambier, J. C. (1995). Role of the Syk autophosphorylation site and SH2 domains in B cell antigen receptor signaling. J. Exp. Med. 182, 1815–1823. Kyo, S., Sada, K., Qu, X., Maeno, K., Miah, S. M., Kawauchi‐Kamata, K., and Yamamura, H. (2003). Negative regulation of Lyn protein‐tyrosine kinase by c‐Cbl ubiquitin‐protein ligase in FceRI‐mediated mast cell activation. Genes Cells 8, 825–836. Latour, S., Fridman, W. H., and Dae¨ron, M. (1996). Identification, molecular cloning, biological properties and tissue distribution of a novel isoform of murine low‐affinity IgG receptor homologous to human FcgRIIB1. J. Immunol. 157, 189–197. Lesourne, R., Bruhns, P., Fridman, W. H., and Dae¨ron, M. (2001). Insufficient Phosphorylation prevents FcgRIIB from recruiting the SH2 Domain‐Containing Protein Tyrosine Phosphatase SHP‐1. J. Biol. Chem. 6327–6336. Lesourne, R., Fridman, W. H., and Dae¨ron, M. (2005). Dynamic interactions of FcgRIIB with filamn‐bound SHIP1 amplify filamentous actin‐dependent negative regulation of FceRI signaling. J. Immunol. 174, 1365–1373. Lewis, V. A., Koch, T., Plutner, H., and Mellman, I. (1986). A complementary DNA clone for a macrophage‐lymphocyte Fc receptor. Nature 324, 372. Liang, X., Wisniewski, D., Strife, A., Shivakrupa, Clarkson, B., and Resh, M. D. (2002). Phosphatidylinositol 3‐kinase and Src family kinases are required for phosphorylation and membrane recruitment of Dok‐1 in c‐Kit signaling. J. Biol. Chem. 277, 13732–13738. Lin, S., Cicala, C., Scharenberg, A. M., and Kinet, J.‐P. (1996). The FceRIb subunit functions as an amplifier of FceRIg‐mediated cell activation signals. Cell 85, 985–995. Long, E. O. (1999). Regulation of immune responses through inhibitory receptors. Annu. Rev. Immunol. 17, 875–904. Luna, E. J., and Hitt, A. L. (1992). Cytoskeleton–plasma membrane interactions. Science 258, 955–964. Mac Glashan, D. W., Bochner, B. S., Adelman, D. C., Jardieu, P., Togias, A., and Lichtenstein, L. M. (1997). Serum IgE levels drives basophil and mast cell IgE receptor display. Int. Arch. Allergy Immunol. 113, 45–47. Magro, A. M., and Alexander, A. (1974). Histamine release: In vitro studies of the inhibitory region of the dose‐response curve. J. Immunol. 112, 1762–1765. Malbec, O., Fong, D., Turner, M., Tybulewicz, V. L. J., Cambier, J., C., Fridman, W. H., and Dae¨ron, M. (1998). FceRI‐associated lyn‐dependent phosphorylation of FcgRIIB during negative regulation of mast cell activation. J. Immunol. 160, 1647–1658. Malbec, O., Fridman, W. H., and Dae¨ron, M. (1999). Negative regulation of c‐kit‐mediated cell proliferation by FcgRIIB. J. Immunol. 162, 4424–4429. Malbec, O., Malissen, M., Isnardi, I., Lesourne, R., Mura, A.‐M., Fridman, W. H., Malissen, B., and Dae¨ron, M. (2004). Linker for activation of T cells integrates positive and negative signaling in mast cells. J. Immunol. 173, 5086–5094.


81

Malbec, O., Schmitt, C., Bruhns, P., Krystal, G., Fridman, W. H., and Dae¨ron, M. (2001). The SH2 domain‐containing Inositol 5‐Phosphatase SHIP1 mediates cell cycle arrest by FcgRIIB. J. Biol. Chem. 276, 30381–30391. Marquardt, D. L., and Walker, L. L. (2000). Dependence of mast cell IgE‐mediated cytokine production on nuclear factor‐kappaB activity. J. Allergy. Clin. Immunol. 105, 500–505. Martindale, D. W., Wilson, M. D., Wang, D., Burke, R. D., Chen, X., Duronio, V., and Koop, B. F. (2000). Comparative genomic sequence analysis of the Williams syndrome region (LIMK1‐ RFC2) of human chromosome 7q11.23. Mamm. Genome. 11, 890–898. Maurer, D., Fiegiger, E., Reininger, B., Wolffwiniski, B., Jouvin, M.‐H., Kilgus, O., Kinet, J.‐P., and Stingl, G. (1994). Expression of a functional high affinity immunoglobulin E receptor (FceRI) on monocytes of atopic individuals. J. Exp. Med. 179, 745–750. Metzger, H., Alcaraz, G., Hohman, R., Kinet, J. P., Pribluda, V., and Quarto, R. (1986). The receptor with high affinity for immunoglobulin E. Annu. Rev. Immunol. 4, 419–470. Minoo, P., Zadeh, M. M., Rottapel, R., Lebrun, J. J., and Ali, S. (2004). A novel SHP‐1/Grb2‐ dependent mechanism of negative regulation of cytokine‐receptor signaling: Contribution of SHP‐1 C‐terminal tyrosines in cytokine signaling. Blood 103, 1398–1407. Monteiro, R. C., and Van De Winkel, J. G. (2003). IgA Fc receptors. Annu. Rev. Immunol. 21, 177–204. Muraille, E., Bruhns, P., Pesesse, X., Dae¨ron, M., and Erneux, C. (2000). The SH2 domain containing inositol 5‐phosphatase SHIP2 associates to the immunoreceptor tyrosine‐based inhibition motif of FcgRIIB in B cells under negative signalling. Immunol. Letters 72, 7–15. Muta, T., Kurosaki, T., Misulovin, Z., Sanchez, M., Nussenzweig, M. C., and Ravetch, J. V. (1994). A 13‐amino‐acid motif in the cytoplasmic domain of FcgRIIB modulates B‐cell receptor signalling. Nature 368, 70–73. Nakamura, K., Malykhin, A., and Coggeshall, K. M. (2002). The Src homology 2 domain‐ containing inositol 5‐phosphatase negatively regulates Fcg receptor‐mediated phagocytosis through immunoreceptor tyrosine‐based activation motif‐bearing phagocytic receptors. Blood 100, 3374–3382. Nishizumi, H., and Yamamoto, T. (1997). Impaired tyrosine phosphorylation and Ca2þ mobilization, but not degranulation, in lyn‐deficient bone marrow‐derived mast cells. J. Immunol. 158, 2350–2355. Nun˜ez‐Cruz, S., Aguado, E., Richelme, S., Chetaille, B., Mura, A.‐M., Richelme, M., Pouyet, L., Jouvin‐Marche, E., Xerri, L., Malissen, B., and Malissen, M. (2003). LAT regulates gd T cell homeostasis and differentiation. Nature Immunol. 4, 999–1008. Odom, S., Gomez, G., Kovarova, M., Furumoto, Y., Ryan, J. J., Wright, H. V., Gonzalez‐Espinosa, C., Hibbs, M. L., Harder, K. W., and Rivera, J. (2004). Negative regulation of immunoglobulin E‐dependent allergic responses by Lyn kinase. J. Exp. Med. 199, 1491–1502. Okada, M., Nada, S., Yamanashi, Y., Yamamoto, T., and Nakagawa, H. (1991). CSK: A protein‐ tyrosine kinase involved in regulation of src family kinases. J. Biol. Chem. 266, 24249–24252. Ono, M., Bolland, S., Tempst, P., and Ravetch, J. V. (1996). Role of the inositol phosphatase SHIP in negative regulation of the immune system by the receptor FcgRIIB. Nature 383, 263–266. Ono, M., Okada, H., Bolland, S., Yanagi, S., Kurosaki, T., and Ravetch, J. V. (1997). Deletion of SHIP or SHP‐1 reveals two distinct pathways for inhibitory signaling. Cell 90, 293–301. Osborne, M. A., Zenner, G., Lubinus, M., Zhang, X., Songyang, Z., Cantley, L. C., Majerus, P., Burn, P., and Kochan, J. P. (1996). The inositol 50 ‐phosphatase SHIP binds to immunoreceptor signaling motifs and responds to high affinity IgE receptor aggregation. J. Biol. Chem. 271, 29271–29278. Ott, V. L., Tamir, I., Niki, M., Pandolfi, P. P., and Cambier, J. C. (2002). Downstream of kinase, p62 (dok), is a mediator of FcgIIB inhibition of FceRI signaling. J. Immunol. 168, 4430–4439.

82


Pandey, V., Mihara, S., Fensome‐Green, A., Bolsover, S., and Cockcroft, S. (2004). Monomeric IgE stimulates NFAT translocation into the nucleus, a rise in cytosol Ca2þ, degranulation, and membrane ruffling in the cultured rat basophilic leukemia‐2H3 mast cell line. J. Immunol. 172, 4048–4058. Pani, G., Kozlowski, M., Cambier, J. C., Mills, G. B., and Siminovitch, K. A. (1995). Identification of the tyrosine phosphatase PTP1C as a B cell antigen receptor‐associated protein involved in the regulation of B cell signaling. J. Exp. Med. 181, 2077–2084. Paolini, R., Molfetta, R., Piccoli, M., Frati, L., and Santoni, A. (2001). Ubiquitination and degradation of Syk and ZAP‐70 protein tyrosine kinases in human NK cells upon CD16 engagement. Proc. Natl. Acad. Sci. USA 98, 9611–6916. Parravincini, V., Gadina, M., Kovarova, M., Odom, S., Gonzalez‐Espinosa, C., Furumoto, Y., Saitoh, S., Samelson, L. E., O’ Shea, J. J., and Rivera, J. (2002). Fyn kinase initiates complementary signals required for IgE‐dependent mast cell degranulation. Nature Immunol. 3, 741–748. Pasquier, B., Launay, P., Kanamaru, Y., Moura, I. C., Pfirsch, S., Ruffie, C., Henin, D., Benhamou, M., Pretolani, M., Blank, U., and Monteiro, R. C. (2005). Identification of FcaRI as an inhibitory receptor that controls inflammation: Dual role of FcRg ITAM. Immunity 22, 31–42. Patry, C., Herblin, A., Lehven, A., Bach, J.‐F., and Monteiro, R. C. (1995). Fca receptors mediate release of tumour necrosis factor‐a and interleukine 6 by human monocytes following receptor aggregation. Immunology 86, 1–5. Pengal, R. A., Ganesan, L. P., Fang, H., Marsh, C. B., Anderson, C. L., and Tridandapani, S. (2003). SHIP‐2 inositol phosphatase is inducibly expressed in human monocytes and serves to regulate Fcg receptor‐mediated signaling. J. Biol. Chem. 278, 22657–22663. Perussia, B., Tutt, M. M., Qui, W. Q., Kuziel, W. A., Tucker, P. W., Trinchieri, G., Bennett, M., Ravetch, J. V., and Kumar, V. (1989). Murine natural killer cells express functional Fcg receptor II encoded by the FcgRa gene. J. Exp. Med. 170, 73–86. Phee, H., Jacob, A., and Coggeshall, K. M. (2000). Enzymatic activity of the Src homology 2 domain‐containing inositol phosphatase is regulated by a plasma membrane location. J. Biol. Chem. 275, 19090–19097. Pivniouk, V. I., Martin, T. R., Lu‐Kuo, J. M., Katz, H. R., Oettgen, H. C., and Geha, R. S. (1999). SLP‐76 deficiency impairs signaling via the high‐affinity IgE receptor in mast cells. J Clin Invest 103, 1737–1743. Pralle, A., Keller, P., Florin, E. L., Simons, K., and Horber, J. K. (2000). Sphingolipid‐ cholesterol rafts diffuse as small entities in the plasma membrane of mammalian cells. J. Cell Biol. 148, 997–1008. Pribluda, V. S., Pribluda, C., and Metzger, H. (1994). Transphosphorylation as the mechanism by which the high affinity receptor for IgE is phosphorylated upon aggregation. Proc. Nat. Acad. Sci. USA 91, 11246–11250. Pribluda, V. S., Pribluda, C., and Metzger, H. (1997). Biochemical evidence that the phosphorylated tyrosines, serines, and threonines on the aggregated high affinity receptor for IgE are in the immunoreceptor tyrosine‐based activation motifs. J. Biol. Chem. 272, 11185–11192. Prouvost‐Danon, A., and Binaghi, R. (1970). In vitro sensitization of mouse peritoneal mast cells with reaginic antibody. Nature 228, 66–68. Prouvost‐Danon, A., Queiroz‐Javierre, M., and Silva‐Lima, M. (1966). Passive anaphylactic reaction in mouse peritoneal mast cells in vitro. Life Sci. 5, 1751. Ravetch, J. V., and Bolland, S. (2001). IgG Fc receptors. Annu. Rev. Immunol. 19, 275–290. Ravetch, J. V., and Kinet, J. P. (1991). Fc Receptors. Annu. Rev. Immunol. 9, 457–492. Ravetch, J. V., Luster, A. D., Weinshank, R., Kochan, J., Pavlovec, A., Portnoy, D. A., Hulmes, J., Pan, Y. C. E., and Unkeless, J. C. (1986). Structural heterogeneity and functional domains of murine Immunoglobulin G Fc receptors. Science 234, 718–725.


83

Reth, M. G. (1989). Antigen receptor tail clue. Nature 338, 383–384. Rivera, J. (2005). NTAL/LAB and LAT: A balancing act in mast‐cell activation and function. Trends Immunol 26, 119–122. Russell, M. W., Sibley, D. A., Nikolova, E. B., Tomana, M., and Mestecky, J. (1997). IgA antibody as a non‐inflammatory regulator of immunity. Biochem. Soc. Trans. 25, 466–470. Saitoh, S., Arudchandran, R., Manetz, T. S., Zhang, W., Sommers, C. L., Paul, E. Love, Rivera, J., and E., S. L. (2000). LAT is essential for FceRI‐mediated mast cell activation. Immunity 12, 525–535. Saitoh, S.‐I., Odom, S., Gomez, G., Sommers, C. L., Young, H. A., Rivera, J., and Samelson, L. E. (2003). The Four Distal Tyrosines Are Required for LAT‐dependent Signaling in FceRI‐ mediated Mast Cell Activation. J. Exp. Med. 198, 831–843. Samuelsson, A., Towers, T. L., and Ravetch, J. V. (2001). Anti‐inflammatory activity of IVIG mediated through the inhibitory Fc receptor. Science 291, 484–486. Sandor, M., Houlden, B., Bluestone, J., Hedrick, S. M., Weinstock, J., and Lynch, R. G. (1992). In vitro and in vivo activation of murine g/d T cells induces the expression of IgA, IgM, and IgG Fc receptors. J. Immunol. 148, 2363–2369. Sandor, M., and Lynch, R. G. (1992). Lymphocyte Fc receptors: The special case of T cells. Immunol. Today 14, 227–231. Schaffer, F. M., Monteiro, R. C., Volanakis, J. E., and Cooper, M. D. (1991). IgA deficiency. Immunodefic. Rev. 3, 15–44. Scharenberg, A. M., El‐Hillal, O., Fruman, D. A., Beitz, L. O., Li, Z., Lin, S., Gout, I., Cantley, L. C., Rawlings, D. J., and Kinet, J.‐P. (1998). Phosphatidylinositol‐3,4,5‐triphosphate (PtdIns‐ 3,4,5‐P3)/Tec kinase‐dependent calcium signaling pathway: A target for SHIP‐mediated inhibitory signals. EMBO J. 17, 1961–1972. Scharenberg, A. M., and Kinet, J. P. (1998). PtdIns‐3,4,5‐P3: A regulatory nexus between tyrosine kinases and sustained calcium signals. Cell 94, 5–8. Scharenberg, A. M., Lin, S., Cueńod, B., Yamamura, H., and Kinet, J.‐P. (1995). Reconstitution of interactions between tyrosine kinases and the high affinity IgE receptor which are controlled by receptor clustering. EMBO J. 14, 3385–3394. Schraven, B., Marie‐Cardine, A., Hubener, C., Bruyns, E., and Ding, I. (1999). Integration of receptor‐mediated signals in T cells by transmembrane adaptor proteins. Immunol. Today 20, 431–434. Segal, D. M., Taurog, J. D., and Metzger, H. (1977). Dimeric immunoglobulin E serves as a unit signal for mast cell degranulation. Proc. Natl. Acad. Sci. USA 74, 2993–2997. Sheets, E. D., Holowka, D., and Baird, B. (1999). Critical role for cholesterol in Lyn‐mediated tyrosine phosphorylation of FceRI and their association with detergent‐resistant membranes. J. Cell Biol. 145, 877–887. Siraganian, R. P., Hook, W. A., and Levine, B. B. (1975). Specific in vitro histamine release from basophils by bivalent haptens: Evidence for activation by simple bridging of membrane‐bound antibodies. Immunochemistry 12, 149–155. Sommers, C. L., Park, C. S., Lee, J., Feng, C., Fuller, C. L., Grimberg, A., Hildebrand, J. A., Lacana, E., Menon, R. K., Shores, E. W., Samelson, L. E., and Love, P. E. (2002). A LAT mutation that inhibits T cell development yet induces lymphoproliferation. Science 296, 2040–2043. Sommers, C. L., Samelson, L. E., and Love, P. E. (2004). LAT: A T lymphocyte adapter protein that couples the antigen receptor to downstream signaling pathways. Bioessays 26, 61–67. Stankunas, K., Graef, I. A., Neilson, J. R., Park, S. H., and Crabtree, G. R. (1999). Signaling through calcium, calcineurin, and NF‐AT in lymphocyte activation and development. Cold Spring Harb. Symp. Quant. Biol. 64, 505–516.

84


Stork, B., Engelke, M., Frey, J., Horejsi, V., Hamm‐Baarke, A., Schraven, B., Kurosaki, T., and Wienands, J. (2004). Grb2 and the non‐T cell activation linker NTAL constitute a Ca2þ‐ regulating signal circuit in B lymphocytes. Immunity 21, 681–691. Swann, P. G., Odom, S., Zhou, Y. J., Szallasi, Z., Blumberg, P. M., Draber, P., and Rivera, J. (1999). Requirement for a negative charge at threonine 60 of the FcRg for complete activation of Syk. J. Biol. Chem. 274, 23068–23077. Swieter, M., Berenstein, E. H., and Siraganian, R. P. (1995). Protein tyrosine phosphatase activity associates with the high affinity IgE receptor and dephosphorylates the receptor subunits, but not Lyn or Syk. J. Immunol. 155, 5330–5336. Takai, T., Ono, M., Hikida, M., Ohmori, H., and Ravetch, J. V. (1996). Augmented humoral and anaphylactic responses in FcgRII‐deficient mice. Nature 379, 346–349. Tam, S. W., Demissie, S., Thomas, D., and Dae¨ron, M. (2004). A bispecific antibody against human IgE and human FcgRII that inhibits antigen‐induced histamine release by human mast cells and basophils. Allergy 59, 772–780. Tamir, I., Stolpa, J. C., Helgason, C. D., Nakamura, K., Bruhns, P., Dae¨ron, M., and Cambier, J. C. (2000). The RasGAP‐binding protein p62dok is a Mediator of Inhibitory FcgRIIB Signals in B cells. Immunity 12, 347–358. Tanaka, S., Mikura, S., Hashimoto, E., Sugimoto, Y., and Ichikawa, A. (2005). Ca2þ influx‐ mediated histamine synthesis and IL‐6 release in mast cells activated by monomeric IgE. Eur. J. Immunol. 35, 460–468. Thomas, M. L., and Brown, E. J. (1999). Positive and negative regulation of Src‐family membrane kinases by CD45. Immunol. Today 20, 406–411. Tkaczyk, C., Horejsi, V., Iwaki, S., Draber, P., Samelson, L. E., Satterthwaite, A. B., Nahm, D. H., Metcalfe, D. D., and Gilfillan, A. M. (2004). NTAL phosphorylation is a pivotal link between the signaling cascades leading to human mast cell degranulation following Kit activation and FceRI aggregation. Blood 104, 207–214. Togni, M., Lindquist, J., Gerber, A., Kolsch, U., Hamm‐Baarke, A., Kliche, S., and Schraven, B. (2004). The role of adaptor proteins in lymphocyte activation. Mol. Immunol. 41, 615–630. Tomlinson, M. G., Heath, V. L., Turck, C. W., Watson, S. P., and Weiss, A. (2004). SHIP family inositol phosphatases interact with and negatively regulate the Tec tyrosine kinase. J. Biol. Chem. 279, 55089–55096. Tridandapani, S., Wang, Y., Marsh, C. B., and Anderson, C. L. (2002). Src homology 2 domain‐ containing inositol polyphosphate phosphatase regulates NF‐k B‐mediated gene transcription by phagocytic FcgRs in human myeloid cells. J. Immunol. 169, 4370–4378. Tuosto, L., Michel, F., and Acuto, O. (1996). p95vav associates with tyrosine‐phosphorylated SLP‐ 76 in antigen‐stimulated T cells. J. Exp. Med. 184, 1161–1166. Turner, H., and Cantrell, D. A. (1997). Distinct Ras effector pathways are involved in FceR1 regulation of the transcriptional activity of Elk‐1 and NFAT in mast cells. J. Exp. Med. 185, 43–53. Ujike, A., Ishikawa, Y., Ono, M., Yuasa, T., Yoshino, T., Fukumoto, M., Ravetch, J., and Takai, T. (1999). Modulation of immunoglobulin (Ig)E‐mediated systemic anaphylaxis by low‐affinity Fc receptors for IgG. J. Exp. Med. 189, 1573–1579. Valverde, A. M., Sinnett‐Smith, J., Van Lint, J., and Rozengurt, E. (1994). Molecular cloning and characterization of protein kinase D: A target for diacylglycerol and phorbol esters with a distinctive catalytic domain. Proc. Natl. Acad. Sci. USA 91, 8572–8576. Van Toorenenbergen, A. W., and Aalberse, R. C. (1981). IgG4 and passive sensitization of basophil leukocytes. Int. Arch. Allergy Appl. Immunol. 65, 432–440.


85

Ve´ly, F., Olivero, S., Olcese, L., Moretta, A., Damen, J. E., Liu, L., Krystal, G., Cambier, J. C., Dae¨ron, M., and Vivier, E. (1997). Differential association of phosphatases with hematopoietic coreceptors bearing Immunoreceptor Tyrosine‐based Inhibition Motifs. Eur. J. Immunol. 27, 1994–2000. Vivier, E., and Dae¨ron, M. (1997). Immunoreceptor tyrosine‐based inhibition motifs. Immunol. Today 18, 286–291. Volna, P., Lebduska, P., Draberova, L., Simova, S., Heneberg, P., Boubelik, M., Bugajev, V., Malissen, B., Wilson, B. S., Horejsi, V., Malissen, M., and Draber, P. (2004). Negative regulation of mast cell signaling and function by the adptor LAB/NTAL. J. Exp. Med. 200, 1001–1013. Wang, D., Feng, J., Wen, R., Marine, J. C., Sangster, M. Y., Parganas, E., Hoffmeyer, A., Jackson, C. W., Cleveland, J. L., Murray, P. J., and Ihle, J. N. (2000). Phospholipase Cg2 is essential in the functions of B cell and several Fc receptors. Immunity 13, 25–35. Weber, J. R., Orstavik, S., Torgersen, K. M., Danbolt, N. C., Berg, S. F., Ryan, J. C., Tasken, K., Imboden, J. B., and Vaage, J. T. (1998). Molecular cloning of the cDNA encoding pp36, a tyrosine‐phosphorylated adaptor protein selectively expressed by T cells and natural killer cells. J. Exp. Med. 187, 1157–1161. Wen, R., Jou, S. T., Chen, Y., Hoffmeyer, A., and Wang, D. (2002). Phospholipase C g 2 is essential for specific functions of FceR and FcgR. J. Immunol. 169, 6743–6752. Wines, B. D., Sardjono, C. T., Trist, H. H., Lay, C. S., and Hogarth, P. M. (2001). The interaction of FcaRI with IgA and its implications for ligand binding by immunoreceptors of the leukocyte receptor cluster. J. Immunol. 166, 1781–1789. Wofsy, C., Goldstein, B., and Dembo, M. (1978). Theory of equilibrium binding of asymmetric bivalent haptens to cell surface antibody: Application to histamine release from basophils. J. Immunol. 593–601. Wonerow, P., and Watson, S. P. (2001). The transmembrane adapter LAT plays a central role in immune receptor signalling. Oncogene 20, 6273–6283. Woodward, J., and Jenkinson, E. (2001). Identification and characterization of lymphoid precursors in the murine intestinal epithelium. Eur. J. Immunol. 31, 3329–3338. Wossning, T., and Reth, M. (2004). B cell antigen receptor assembly and Syk activation in the S2 cell reconstitution system. Immunol. Lett. 92, 67–73. Xie, Z. H., Zhang, J., and Siraganian, R. P. (2000). Positive regulation of c‐Jun N‐terminal kinase and TNF‐a production but not histamine release by SHP‐1 in RBL‐2H3 mast cells. J. Immunol. 164, 1521–1528. Yamanashi, Y., Tamura, T., Kanamori, T., Yamane, H., Nariuchi, H., Yamamoto, T., and Baltimore, D. (2000). Role of the rasGAP‐associated docking protein p62Dok in negative regulation of B cell receptor‐mediated signaling. Gene Develop. 14, 11–16. Yamasaki, S., Ishikawa, E., Kohno, M., and Saito, T. (2004). The quantity and duration of FcRg signals determine mast cell degranulation and survival. Blood 103, 3093–3101. Yamasaki, S., Nishida, K., Sakuma, M., Berry, D., McGlade, C. J., Hirano, T., and Saito, T. (2003). Gads/Grb2‐Mediated Association with LAT Is Critical for the Inhibitory Function of Gab2 in T Cells. Mol. Cell. Biol. 23, 2515–2529. Young, R. M., Holowka, D., and Baird, B. (2003). A lipid raft environment enhances Lyn kinase activity by protecting the active site tyrosine from dephosphorylation. J. Biol. Chem. 278, 20746–20752. Yuasa, T., Kubo, S., Yoshino, T., Ujike, A., Matsumura, K., Ono, M., Ravetch, J. V., and Takai, T. (1999). Deletion of Fcg Receptor IIB Renders H‐2b Mice susceptible to Collagen‐induced Arthritis. J. Exp. Med. 189, 187–194.

86


Zhang, W., Irvin, B. J., Trible, R. P., Abraham, R. T., and Samelson, L. E. (1999). Functional analysis of LAT in TCR‐mediated signaling pathways using a LAT‐deficient cell line. Int. Immunol. 11, 943. 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. Zhang, W., Trible, R. P., and Samelson, L. E. (1998). LAT palmitoylation: Its essential role in membrane microdomain targeting and tyrosine phosphorylation during T cell activation. Immunity 9, 239–246. Zhang, W., Trible, R. P., Zhu, M., Liu, S. K., McGlade, C. J., and Samelson, L. E. (2000). Association of Grb2, Gads and phospholipase C‐g1 with phosphorylated LAT tyrosine residues: Effect of LAT tyrosine mutations on T cell antigen receptor signaling. J. Biol. Chem. 275, 23335. Zhu, D., Kepley, C. L., Zhang, K., Terada, T., Yamada, T., and Saxon, A. (2005). A chimeric human‐ cat fusion protein blocks cat‐induced allergy. Nat. Med. 11, 446–449. Zhu, M., Janssen, E., and Zhang, W. (2003). Minimal requirements of tyrosine residues of linker for activation of T cells in TCR signaling and thymocyte development. J. Immunol. 170, 325–333. Zhu, M., Liu, Y., Koonpaew, S., Granillo, O., and Zhang, W. (2004). Positive and negative regulation of FceRI‐mediated signaling by the adaptor protein LAB/NTAL. J. Exp. Med. 200, 991–1000.

The Surprising Diversity of Lipid Antigens for CD1‐Restricted T Cells D. Branch Moody Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts

1. 2. 3. 4. 5. 6. 7. 8.

Abstract............................................................................................................. Introduction: From Molecules to Functions ............................................................. CD1 Protein Expression on Antigen‐Presenting Cells ................................................ Subcellular Lipid Antigen Processing Pathways......................................................... 3‐Dimensional Structures of CD1‐b2‐Microglobulin‐Lipid Complexes .......................... Microbial Antigens and Infectious Disease ............................................................... Self Antigens, Autoreactivity, and Autoimmune Disease.............................................. Synthetic Lipid Antigens and Prospects for Immunotherapy........................................ Conclusion: Prospects for Immunotherapy ............................................................... References .........................................................................................................

87 87 90 94 102 111 119 125 126 128

Abstract CD1 proteins have been conserved throughout mammalian evolution and function to present lipid antigens to T cells. Crystal structures of CD1‐lipid complexes show that CD1 antigen‐binding grooves are composed of four pockets and two antigen entry portals. This structural information now provides a detailed understanding of how CD1‐binding grooves capture a surprisingly diverse array of lipid ligands. CD1‐expressing APCs are able to acquire lipid antigens from their own pool of lipids and from exogenous sources, including microbial pathogens, bystander cells, or even the systemic circulation. CD1 proteins bind to certain antigens using high stringency loading reactions within endosomes that involve low pH, glycosidases, and lipid transfer proteins. Other antigens can directly load onto CD1 proteins using low stringency mechanisms that are independent of cellular factors. New evidence from in vivo systems shows that CD1‐restricted T cells influence outcomes in infectious, autoimmune, and allergic diseases. These studies lead to a broader view of the natural function of ab T cells, which involves recognition of both cellular proteins and lipids. 1. Introduction: From Molecules to Functions The discovery of the CD1 antigen presentation system represents an advance in immunology because it shows that T cells scan and respond to changes in the lipid content of target cells. The central questions in the study of


0065-2776/06 $35.00 DOI: 10.1016/S0065-2776(05)89003-0

88

D. BRANCH MOODY

CD1‐restricted T cells have unfolded in a very different way from those that led to the discovery of the major histocompatibility complex (MHC). The study of the MHC started by establishing the role of T cells in controlling viral infection and rejecting transplanted tissues (McDevitt and Benacerraf, 1969; Zinkernagel and Doherty, 1974) and represented a three‐decade‐long search for the molecules that mediate T cell activation (Garboczi et al., 1996; Garcia et al., 1996; Kappler et al., 1983; McIntyre and Allison, 1983; Oldstone et al., 1988). In contrast, detailed molecular information about CD1 proteins and CD1 genes was available for many years prior to the first evidence that CD1 proteins control the activation of T cells. McMichael and Milstein discovered CD1 proteins using the monoclonal antibody, NA1/34, which bound a protein on human thymocytes with an apparent molecular weight of 45 kilodaltons. The ligand bound by this antibody showed some biochemical similarities to the MHC class I protein, but could be distinguished by its lack of recognition by an MHC class I‐specific antibody, W6/32 (McMichael et al., 1979). Initially known as human thymocyte antigen‐1 (HTA‐1), this protein was renamed as the first cluster of differentiation antigen (CD1) because its discovery represented the first use of monoclonal antibodies to define a specific cell surface marker on human lymphocytes. Cellular studies showed that the CD1 heavy chain associates with b‐2 microglobulin to form heterodimers (Cotner et al., 1981; Terhorst et al., 1981). Subsequently, the five CD1 genes (CD1A, CD1B, CD1C, CD1D, CD1E) were cloned from human thymocytes and mapped to chromosome 1 (Calabi and Milstein, 1986). The first studies of CD1 function showed that human ab and gd T cell clones were directly reactive with CD1 in the sense that their activation could be blocked with antibodies against CD1 proteins, but the activation did not require an exogenous antigen (Porcelli et al., 1989). Shortly thereafter, it was found that antigens from microbial pathogens could be taken up into endosomes for processing reactions within human dendritic cells and that these microbial antigens were lipidic in nature (Beckman et al., 1994; Porcelli et al., 1992). Twenty‐five years after the discovery of CD1 proteins, current work in the field seeks to understand the precise functions of CD1‐ restricted T cells in influencing infectious, autoimmune, allergic, vascular, and neoplastic diseases. In considering the natural functions of CD1‐restricted T cells in immune response, two populations within the CD1‐restricted T cell repertoire have been recognized and categorized, based on whether or not they express conserved or diverse TCRs (Fig. 1). In general, CD1‐restricted T cells are capable of expressing diverse TCR a and b chains and recognize structurally diverse classes of lipid antigens. Within the larger CD1‐restricted T cell repertoire, NK T cells are an abundant and functionally important subset

LIPID ANTIGENS FOR CD1‐RESTRICTED T CELLS

89

Figure 1 The larger repertoire of diverse CD1‐restricted T cells is composed of clones expressing TCRs with diverse structural elements, including differing TCR a‐ and b‐variable gene segments with varied N‐region additions. NK T cells are a large subpopulation of CD1‐restricted T cells that recognize CD1d and express conserved TCRs with an invariant TCR a chain and limited diversity of TCR Vb genes.

that expresses nearly invariant TCRs composed of a canonical a chain (Va14Ja18 in mice and Va24Ja18 in humans) paired with a limited number of b‐chains. The discovery of NK T cells involved the identification of T cells that lack CD4 and CD8 coreceptors, express Vb8‐containing TCRs and could be identified by staining C‐type lectins that are also found on NK cells such as NK1.1 (CD161) (Budd et al., 1987; Fowlkes et al., 1987). Separately, knockout of MHC class II proteins revealed a set of CD4þ T cells that express TCR a chains with limited diversity (Bendelac et al., 1994; Cosgrove et al., 1991). The molecular targets of recognition of NK T cells were solved in two key studies, which showed that CD1d expression on APCs was necessary for their activation (Bendelac, 1995) and that the CD1d‐mediated activation was greatly enhanced by treating APCs with synthetic glycolipids related in structure to a‐galactosyl ceramides (Kawano et al., 1997). The use of the term NK T cell to refer to CD1d‐restricted T cells with invariant TCRs persists for these historical reasons and because NK cells and NK T cells have some limited functional similarities. For example, both NK cells and NK T cells use T‐bet transcription factors (Townsend et al., 2004), expand in the presence of IL‐15 (Matsuda et al., 2002), and have at least some reliance on NK1.1 for activation (Exley et al., 1998). However, the primary activation signals for NK T cells involve TCRs and CD1d, not NK complex‐ encoded receptors. Further, many conventional T cells express NK1.1, and many CD1d‐restricted T cells with invariant TCR a chains do not express NK1.1 (Gumperz et al., 2002). The relative lack of sensitivity and specificity of NK1.1 and other NK complex‐encoded proteins for distinguishing NK T

90

D. BRANCH MOODY

cells from MHC‐restricted T cells has led to the waning use of these markers as the primary defining features of NK T cells. Instead, NK T cells are identified by the presence of invariant TCR a chains and activation by CD1d in combination with glycolipid antigens (Fig. 1). This review of CD1‐restricted T cells follows the path of evolution of CD1 research—from molecular structures to in vivo studies—with emphasis on the newest advances in each area.

2. CD1 Protein Expression on Antigen‐Presenting Cells 2.1. Three Groups of CD1 Proteins The human CD1 locus encodes five genes: CD1A, CD1B, CD1C, CD1D, and CD1E (Calabi and Milstein, 1986). Prior to accumulation of any knowledge of the functions of CD1 proteins, a grouping system was proposed based on levels of amino acid sequence homology among the five members of the human CD1 family (Calabi et al., 1989). CD1a, CD1b, and CD1c were most homologous to one another and were designated as the group 1 isoforms. CD1d demonstrated the lowest overall homology to other CD1 proteins and was designated as the group 2 isoform. CD1e was similarly homologous to both groups, so it was not originally classified, but is now known as the group 3 isoform. Although this classification system was originally devised based on amino acid sequence, it remains useful because more detailed studies of gene regulation and CD1 function have generally, but not universally, found differing functions for group 1, group 2, and group 3 CD1 proteins. For example, each of the group 1 isoforms (CD1a, CD1b, CD1c) bind and present microbial lipid antigens to diverse CD1‐restricted T cells, whereas the group 2 isoform (CD1d) is more clearly associated with a function in activating invariant NK T cells. Because NK T cells generally function to regulate other cells, whereas diverse T cells are thought to have direct effector functions against cells with altered lipids, these differences in antigen profiles may also translate into differing functions in vivo. There are, however, exceptions to these apparently differing immunoregulatory and effector functions of group 1 and 2 CD1 proteins, as NK T cells have been recently found to directly recognize microbial lipids (Kinjo et al., 2005; Mattner et al., 2005), and T cells that are directly reactive to group 1 CD1 proteins regulate DC maturation (Vincent et al., 2003). The function of the group 3 isoform, CD1e, is distinct because it is not detected at the surface of APCs and does not function in antigen display (Angenieux et al., 2000, 2005). Instead, it functions in the processing and transfer of lipids to other CD proteins (de la Salle et al., 2005).


91

2.2. Evolutionary Conservation of CD1 CD1 proteins are expressed in every mammalian species that has been examined to date, including monkeys, horses, sheep, cows, rabbits, guinea pigs, rats, and mice. Because the human system was the first to be described, CD1 nomenclature designates non‐human sequences according to the human ortholog with which they share the highest level of overall sequence homology. When a species expresses more than one ortholog of a given human gene, each protein is given an arbitrary numerical designation. For example, guinea pigs express three CD1c orthologs (CD1c1, CD1c2, CD1c3), and rabbits express two orthologs of CD1a (CD1a1 and CD1a2) (Dascher et al., 1999; Hayes and Knight, 2001). This nomenclature reflects the fact that most mammalian species have larger numbers of CD1 genes, with mice and rats being the notable exceptions. Recent advances point to new insights in the evolutionary origins of the CD1 system. MHC class I and II molecules are conserved in bony and cartilaginous fish but are lacking in jawless fish. Based on fossil evidence that dates this split, MHC antigen‐presenting molecules are thought to have arisen approximately 500 million years ago, contemporaneously with recombinases that allowed production of rearranged TCRs, forming the basis of the acquired immune system (Laird et al., 2000). Sequence analyses of MHC I, MHC II, and CD1 genes show similar levels of relatedness (despite differences in domain organization), and it has been argued that these three families of antigen‐presenting molecules are derived from a single ancestral gene (Hughes, 1991; Porcelli, 1995). However, there has been controversy as to whether the CD1 system is an ancient branch of the adaptive immune system or evolved from MHC class I proteins under selective pressure over a relatively short span. For example, computer‐based homology analyses had predicted that CD1 may have diverged early in vertebrate evolution, whereas others suggested that divergence might have occurred later, perhaps even after the bird–mammal (synapsid–diapsid) split, approximately 300 million years ago (Hughes, 1991; Porcelli, 1995). Direct evidence that ancestral CD1 genes predate the bird–mammal divergence in the form of CD1 gene sequences in early vertebrates was lacking until recently. Two studies have independently isolated avian CD1 orthologs in chickens (Gallus gallus), which have sequence or structural homologies with mammalian CD1 genes (Miller et al., 2005; Salomonsen et al., 2005). Although the ability of these avian CD1 gene products to present antigens has not been analyzed, these chicken CD1 genes are predicted to encode proteins with requisitely high amino acid homology with mammalian CD1 proteins. Also, chicken CD1 proteins have hydrophobic amino acids in the predicted

92

D. BRANCH MOODY

antigen‐binding domains. Importantly, both studies found that chicken CD1 genes are tightly linked with the MHC complex, providing the first example of direct linkage between CD1 and MHC genes. These studies indicate that the CD1 system predates evolution of mammalian species and more directly support prior arguments that CD1 and MHC genes diverged from a common ancestral gene. 2.3. Constitutive and Inducible Expression of CD1 on APCs With the exception of epithelial cells that line the gastrointestinal tract (Somnay‐Wadgaonkar et al., 1999; van de Wal et al., 2003), the expression of CD1 proteins is generally limited to hemopoietically derived cells with specialized functions in antigen presentation and immune response. All five human CD1 proteins are expressed on thymocytes at high levels (Calabi and Milstein, 1986). Most evidence suggests that thymocytes downregulate CD1 proteins prior to their exit to the periphery. However, there is some evidence for expression of CD1a and CD1c proteins on a small subset of T cells in the peripheral blood (Salamone et al., 2001a,b). B cells represent another effector population that constitutively expresses CD1 proteins at high levels. Although CD1c and CD1d are found on a minority of peripheral blood B cells, many studies of secondary lymphoid tissues show that these two isoforms are expressed on large numbers of B cells that localize to the interface of B cell‐ and T cell‐rich areas of secondary lymphoid tissues and co‐express markers of marginal zone B cells. In fact, formal analysis of patterns of CD1c and CD1d expression have found high correlation with localization to the marginal zone, such that CD1c and CD1d can be considered relatively specific markers of marginal zone B cells (Weller et al., 2004). Marginal zone B cells are positioned so that they can rapidly sample blood‐borne antigens, and new evidence indicates that the expression of CD1d at this site regulates Ig production in a way that can alter infection. For example, marginal zone B cells from CD1d‐ deficient mice produce lower levels of borrelia‐specific IgM than wild‐type mice, and passive transfer of IgM partially rescues the worsened infection seen in CD1d‐deficient mice infected by Borrelia burgdorferii and Borellia hermsii (Belperron et al., 2005; Kumar et al., 2000). In contrast to the apparently constitutive expression of CD1 on thymocytes, B cells, and epithelia, the patterns of expression of individual CD1 isoforms on myeloid cells are complex and change in response to cellular activation. CD1d proteins are expressed at low but consistently detectable levels on monocytes. In contrast, fresh human monocytes from peripheral blood lack expression of CD1a, CD1b, and CD1c, but can be induced to express these group 1 isoforms at high levels after treatment with GM‐CSF, IL‐4, and other factors that


93

promote DC maturation (Porcelli et al., 1992; Sallusto and Lanzavecchia, 1994). Indeed, group 1 CD1 protein expression is increasingly used as a lineage marker that allows myeloid DCs to be distinguished from activated macrophages. However, modulation of CD1 expression on DCs has only recently come under study as a possible physiological means of regulating the functions of CD1‐restricted T cells, leading to the identification of molecular signals that turn off or on CD1 antigen presentation in DCs. One study has shown that infection of CD1‐expressing dendritic cells by Mycobacterium tuberculosis, a pathogen that makes several classes of CD1‐presented lipids, resulted in the complete loss of detectable CD1b proteins over a period of several days. Such downregulation of CD1 expression in vitro led to the speculation that this could represent a mechanism of immune evasion in vivo (Stenger et al., 1998b). However, similar studies of in vitro‐derived dendritic cells have found only partial or no downregulation of CD1 proteins in response to mycobacterial infection (Giuliani et al., 2001; Henderson et al., 1997). Also, studies testing the effect of Mycobacterium tuberculosis infection of fresh monocytes have found upregulation of CD1a, CD1b, and CD1c after infection. The mechanism of CD1 upregulation in response to mycobacteria involves lipid agonists of Toll‐like receptor 2 (TLR‐2), such as lipoarabinomannan, which promotes transcription and translation of group 1 CD1 proteins (Roura‐Mir et al., 2005b). A similar process likely occurs in vivo, as Mycobacterium leprae infection increases group 1 CD1 protein expression on DCs within the skin of patients with tuberculoid leprosy (Krutzik et al., 2005; Sieling et al., 1999). Also, group 1 CD1 proteins are expressed in peribronchial tissues and brochioalveolar lavage samples of tuberculosis patients (Buettner et al., 2005; Uehira et al., 2002). Thus, the dominant effect of mycobacteria is to generate a local inflammatory reaction that leads to increased expression of group 1 CD1 proteins on myeloid cells at sites of infection. New insights into the differential timing of expression of CD1a, CD1b, CD1c, CD1d, and CD1e on maturing myeloid DCs point to potentially distinct biological functions of each of the three groups of CD1 proteins. Kinetic measurements of group 1 CD1 induction at the cell surface in response to cytokines, TLR agonists, or mycobacterial lipids show that cell surface expression is first detected 2 days after activation, peaking at day 3. This delayed expression of group 1 CD1 proteins relates to the cellular mechanism of group 1 CD1 induction, which requires synthesis of new proteins, rather than redirected trafficking of pre‐formed proteins, as is the case with MHC class II (Roura‐Mir et al., 2005b). Certain stimuli that promote group 1 CD1 expression lead to the concurrent downregulation of CD1d and vice versa. For example, the upregulation of CD1d expression by agonists of

94

D. BRANCH MOODY

the peroxisome proliferator activated receptor g (PPAR‐g) pathway is accompanied by decreased expression of CD1a (Szatmari et al., 2004), and mycobacterial cell wall products that increase group 1 CD1 expression decrease CD1d expression. The differing timing and opposite responses of group 1 and group 2 CD1 proteins imply distinct functions. Rapid activation of NK T cells over minutes to hours by constitutively expressed CD1d proteins is one criterion supporting the description of NK T cells as ‘‘lymphocytes of innate immunity’’ (Benlagha and Bendelac, 2000). In contrast, the delayed expression of group 1 CD1 proteins on DCs, as well as their regulation by antecedent activation of pattern recognition receptors of the innate immune system, are more in keeping with a slightly delayed pathogen recognition system that becomes effective during the transition from innate to adaptive immune responses. These new insights derived from in vitro studies of the agonists, receptors, and kinetics of expression of CD1 now form a rationale for further investigation of the dynamic patterns of expression of CD1 proteins in vivo. Given the clear evidence for cell‐type and activation‐dependent patterns of expression of CD1 proteins, there currently exists remarkably little information regarding signaling molecules, promoter elements, or transcriptional mechanisms that control CD1 expressions, at the surface of myeloid APCs. Agonists of TLRs and PPAR‐g both signal through nuclear factor kappa B (NFkB), implicating this signaling pathway in the control of CD1 expression. There has been some preliminary mapping of CD1 promoters, and new evidence shows that E26 transformation‐specific (ets) transcription factors control CD1d expression (Calabi et al., 1989; Geng et al., 2005). 3. Subcellular Lipid Antigen Processing Pathways New studies indicate that the loading of lipid antigens onto cellular CD1 proteins is regulated by vesicular ATPases, cell surface lectins, lipid transfer proteins, and other cellular cofactors with distinct patterns of expression within subcellular compartments of APCs. These studies are beginning to coalesce into models of coordinated subcellular pathways of lipid antigen processing. Whereas protein antigen processing centrally involves cleaving antigens to generate smaller peptides, lipid antigen processing is coming to be understood as a multi‐stage transfer process for moving relatively insoluble antigens through aqueous solutions for loading into the hydrophobic grooves of CD1 proteins. Viewed from this perspective, the trafficking of CD1 proteins through the secretory, cell surface, and endosomal pathways is not merely a means to disperse CD1 proteins throughout diverse lumenal compartments of APCs. Instead, the complex subcellular trafficking patterns offer a means to


95

regulate the loading of differing subclasses of antigens into CD1 proteins and thereby influence the repertoire of lipids that are ultimately displayed to T cells. In particular, the routing of CD1 trafficking through late endosomes and lysosomes renders CD1 proteins able to capture antigens that cannot be captured within other cellular subcompartments. 3.1. CD1 Translation and Egress to the Cell Surface Leader sequences in the CD1 genes result in cotranslational insertion into the endoplasmic reticulum (ER), so that the a1, a2, and a3 portions of the heavy chain are inserted in the lumen, leaving 6 or more amino acids extending into the cytoplasm (cytoplasmic tails). The folding of CD1 heavy chains is regulated by calnexin and calreticulin and appears to be followed by rapid association with b‐2 microglobulin in the ER (Bauer et al., 1997; Huttinger et al., 1999; Kang and Cresswell, 2002a; Sugita and Brenner, 1994). Whether or not ER‐resident lipids associate with CD1 proteins during the initial folding process has not been directly assessed in cells, but there is indirect evidence that lipids promote CD1 folding. For example, addition of lipids to unfolded recombinant CD1 heavy chains increases the rate of folding and assembly with b‐2 microglobulin in vitro, likely due to the ability of the aliphatic hydrocarbon chains to stabilize the formation of the inner hydrophobic surface of the a1–a2 superdomain (Karadimitris et al., 2001). Further supporting this hypothesis, it has been possible to elute phosphatidylinositol‐containing lipids from cellular CD1d proteins, and some evidence indicates that lipid association occurs prior to transit of CD1 through the golgi apparatus (De Silva et al., 2002; Joyce et al., 1998). Microsomal triglyceride transfer protein (MTP), an ER‐resident protein with known function in assembling lipoprotein particles, has been recently shown to associate with CD1d, and MTP deletion results in reduced levels of NK T cells activation in vitro and in vivo (Brozovic et al., 2004; Dougan et al., 2005). This result implies that altered lipid transfer or loading onto CD1 within the secretory pathway affects the subsequent ability of cell surface CD1d proteins to regulate NK T cell activation. The ability of phosphatidylinositol and other endogenous lipids to load onto CD1 proteins early in their trafficking pathways has led to the speculation that these lipids might function as chaperones that are exchanged for exogenously acquired lipids in endosomes or subsequently encountered compartments, analogous to the mechanism by which MHC class II invariant chain peptide (CLIP) regulates loading of antigenic peptides. During golgi transit, CD1 proteins undergo N‐linked glycosylation at multiple sites so that mature CD1 proteins are decorated with three or more

96

D. BRANCH MOODY

glycans. The trans‐golgi can be considered the point at which the trafficking patterns of each human CD1 proteins most clearly begin to diverge into separate pathways that lead to CD1 accumulation in endosomes via direct transport from the trans‐golgi or recycling from the cell surface. At least two mechanisms have been identified whereby sorting events at the trans‐golgi lead to direct transport of CD1 proteins to endosomal compartments. First, the cytoplasmic tail sequences of some CD1 proteins contain modified dileucine motifs, which generally promote sorting into vesicles that traffic to endosomes (Fig. 2). This role was first shown in human CD1d proteins, when mutagenesis of the modified dileucine motif in the cytoplasmic tail was shown to decrease steady state localization of CD1d proteins in endosomes (Rodionov et al., 1999, 2000). A second means of direct transport of CD1 proteins from the golgi to endosomes involves MHC class II‐invariant chain complexes, which have been detected in substoichiometric amounts in association with mouse CD1d (Jayawardena‐Wolf et al., 2001; Kang and Cresswell, 2002b). Although the precise mechanism of redirected trafficking is not yet known, the cytoplasmic tails of the MHC class II b‐chain and the invariant chain contain modified dileucine motifs (Fig. 2). Therefore, mouse CD1d proteins, which lack dileucine motifs, may nevertheless undergo dileucine‐based sorting using the tails of other proteins in heteromultimeric complexes. The recycling pathway involves rapid transport to the cell surface via the default secretory pathway, followed by regulated entry into the endosomal network via reinternalization. The relative proportions of CD1 proteins that reach endosomes by the direct and recycling pathways likely differs for each CD1 isoform. CD1e accumulates in the golgi apparatus at steady state and can be detected in smaller amounts in CD63‐expressing lysosomes with a multi‐lammellar appearance in electron micrographs. Because CD1e proteins have not been detected at the cell surface and do not mediate internalization of anti‐CD1e antibodies, CD1e appears to solely use direct transport from the trans‐golgi network to reach the endosomes (Angenieux et al., 2000, 2005). Human CD1e has a particularly long cytoplasmic tail, which contains a putative dileucine motif (Fig. 2), but the role of signaling sequences in the cytoplasmic tail has not yet been investigated. For other CD1 isoforms, the recycling pathway appears to be quantitatively dominant. This conclusion is supported by pulse‐chase studies showing that CD1b and CD1d proteins rapidly appear at the cell surface in a time frame typical of proteins in the secretory pathway (Briken et al., 2002; Jayawardena‐Wolf et al., 2001). Once CD1a, CD1b, CD1c, and CD1d proteins reach the surface, they recycle to varying extents to distinct subcompartments of the endosomal network.


97

Figure 2 The cytoplasmic tails of CD1 proteins regulate intracellular trafficking. The cytoplasmically oriented sequences of transmembrane proteins are annotated for sequences that conform to YXXZ (bold) or modified dileucine motifs (underlined). The mechanism of action of YXXZ motifs is known to involve direct binding to the m‐subunit of adaptor protein (AP) complexes, which mediates packing of transmembrane cargo proteins into transport vesicles. Interspecies differences in trafficking of orthologous proteins are illustrated by human CD1d, which uses dileucine‐based sorting and YXXZ‐mediated interactions with AP‐2. Murine CD1d lacks an identifiable dileucine motif but has a YXXZ motif, which interacts with AP‐2 and AP‐3.

3.2. CD1 Recycling to Endosomes The internalization of CD1 proteins from the cell surface to the endosomal network is regulated by cytosolic adaptor protein complexes (AP), which bind to tyrosine‐containing amino acid sequence motifs in the cytoplasmic tails of CD1b, CD1c, and CD1d proteins. Adaptor protein complexes (AP‐1, AP‐2, AP‐3, AP‐4) are heterotetramers composed of two large (c, a, d, or paired

98

D. BRANCH MOODY

with b1, b2, b3, or b4), one medium (m1, m2, m3, or m4), and one small (s1, s2, s3, s4) protein. The binding of AP complexes to transmembrane cargo proteins is mediated by the insertion of two amino acid residues from the cytoplasmic tail of cargo proteins, one tyrosine and one hydrophobic, into two clefts in the m‐subunit of AP complexes (Fig. 2). The spacing of these two interactions is favored when the tyrosine and hydrophobic residues are separated by two intervening amino acids (Ohno et al., 1995; Owen and Evans, 1998). Thus, AP‐interacting proteins can be recognized by the presence of a tyrosine‐containing sequence in their cytoplasmic tails, YXXZ, in which Y is a tyrosine, Z is a hydrophobic amino acid, and XX are any two amino acids that function as spacers. Although the intracellular distribution of AP complexes is somewhat overlapping, AP‐1 complexes localize to the trans‐golgi; AP‐2 complexes are abundant at the cell surface, and AP‐3 complexes are found in endosomal compartments (Bonifacino and Traub, 2003). These complexes localize to the cytoplasmic face of cellular membranes, where they regulate packaging of transmembrane proteins into transport vesicles and can be thought of as a network that influences the steady state intracellular localization of many cargo proteins. The functions of AP complexes in regulating the trafficking and antigen‐ presenting functions have been extensively studied for human and murine CD1 proteins. A clear function for the cytoplasmic tail in regulating CD1 localization was first shown for human CD1b using mutants that lacked the entire cytoplasmic tail (Sugita et al., 1996). Mutant CD1b proteins were found to have significantly reduced localization in MHC class II compartments (MIIC) at steady state, and this resulted in impairment of CD1b presentation of exogenously derived mycobacterial lipids to T cells (Jackman et al., 1998). Subsequently, it was found that the particular amino acids that comprise the YXXZ motif in human CD1b and mouse CD1d are capable of interacting with the m‐subunit of AP‐2 and AP‐3 complexes, whereas human CD1d and human CD1c only interact with the m‐subunit of AP‐2 (Briken et al., 2002; Elewaut et al., 2003; Sugita et al., 2002). Thus, human CD1b and mouse CD1d reinternalization appears to involve a two‐stage process in which AP‐2 complexes promote transport from the surface to endosomes, and then AP‐3 works within early endosomes to promote transport to late endosomes and lysosomes. In contrast, human CD1c and human CD1d undergo the first step involving reinternalization from the surface to endosomes, but lack AP‐3‐ mediated redistribution to late endosomes and lysosomes. Human CD1a has a short cytoplasmic tail lacking any discernable sorting motif, a feature which distinguishes human CD1a from other human CD1 isoforms as well as non‐human CD1a proteins (Fig. 2). This molecular analysis of AP complex interactions with tail motifs in various CD1 proteins predicts that human


99

CD1b and mouse CD1d recycle to late endosomes most efficiently, followed by human CD1d, human CD1c, and human CD1a. Indeed, immunoelectron microscopic analysis of CD1 in multi‐lammellar compartments and multi‐color immunofluorescence analysis of CD1 colocalization bear out this prediction (Sugita et al., 1999). Thus, the apparent hierarchy of the degree of late endosomal localization among CD1 proteins can be largely accounted for by their known differences in interactions with AP‐2 and AP‐3 complexes. 3.3. CD1 Trafficking Controls Lipid Presentation Recycling from the cell surface to endosomes has important functional consequences for the ability of CD1 proteins to subsequently present lipid antigens at the cell surface. This conclusion is supported by many cellular studies in which lipid‐mediated T cell activation is abrogated by deleting endosomal targeting sequences located in CD1 tails, blocking trafficking by membrane fixation, or pharmacologically inhibiting acidification of endosomes. For example, reversal or blockade of endosomal acidification by treating APCs with concanamycin or chloroquine leads to a severe or complete loss of lipid‐ mediated T cell activation (Gilleron et al., 2004; Moody et al., 1999; Porcelli et al., 1992; Roberts et al., 2002; Sieling et al., 1995). In addition, CD1b proteins that have intact antigen‐binding domains, but lack cytoplasmic tail sequences for targeting to endosomes, fail to efficiently mediate T cell activation in response to mycolate antigens (Jackman et al., 1998). Similarly, deletion of the cytoplasmic tail of mouse CD1d, including the YXXZ motif, leads to the loss of activation and positive selection of NK T cells with invariant (Va14) TCRs (Chiu et al., 1999, 2002). The altered T cell activation seen in both types of studies likely occurs due to reduced levels of endosomal recycling and not other effects, because total cellular pools and cell surface density of CD1 proteins are generally maintained or increased after cytoplasmic tail deletion. In addition, alteration of endosomal trafficking by cytoplasmic tail mutation in mouse CD1d does not lead to the loss of ability to activate Va14 T cells, and human tail‐deleted CD1b retains its ability to present certain kinds of lipid antigens (Chiu et al., 1999; Moody et al., 2002). These studies have been interpreted to mean that altered CD1b or CD1d trafficking does not merely reduce the efficiency of presentation, but can meaningfully alter the subset of lipids that are displayed on the cell surface. Endosomal recycling also promotes T cell activation by CD1c‐presented mannosyl mycoketides and CD1a‐presented dideoxymycobactins (Sugita et al., 1999, 2000). However, recognition of these antigens is not absolutely dependent on intact recycling pathways, and fewer numbers of antigens presented

100

D. BRANCH MOODY

by CD1a and CD1c have been studied, so it is not yet possible to draw generalizations about the role of recycling for these two isoforms. Although it is clear that passage through endosomes is important for presenting certain lipid antigens, it is not yet known whether the two means of access to endosomes, the direct and recycling pathways, have distinct or overlapping functions. Quantitatively, tyrosine motif‐mediated recycling from the surface appears dominant because the effects of deletion of dileucine motifs in CD1d and invariant chain are not clearly seen, unless the recycling pathway is also interrupted by mutation of tyrosine motifs (Jayawardena‐Wolf et al., 2001; Kang and Cresswell, 2002b). Also, the nearly complete loss of NK T cells after deletion of the CD1d tail implies that MHC class II‐invariant chain‐related mechanisms mediating endosomal delivery are not sufficient to rescue the functional defect in AP‐mediated recycling (Chiu et al., 2002). However, the precise localization and function of CD1 in the subcompartments of the late endosome–lysosome continuum are not yet fully understood. Therefore, it remains possible that these two cellular mechanisms may target CD1 proteins for subcompartments with distinct immunological functions. 3.4. Molecular Events in Endosomal Processing The targeting of CD1 proteins and antigens to endosomes leads to more efficient T cell activation, supporting the concept that endosomes represent a coordinated subcellular pathway that regulates antigen display. Whereas processing of protein antigens centrally involves covalent cleavage to generate peptides so they can fit within the groove of MHC‐encoded proteins, most known natural lipid antigens contain lipid tails that approximate the volume of their respective CD1 grooves. Thus, the cellular events involved in converting lipids to a recognizable form do not universally require modification of antigen structure through covalent cleavage. Instead, endosomal lipid antigen processing appears to be centrally concerned with transfer of relatively insoluble lipids through aqueous biological solutions to and from membranes, lipid‐ protein aggregates, and the hydrophobic groove found in CD1 proteins. Existing data support four separate, but not mutually exclusive events, which contribute to the generation of antigenic CD1 complexes within endosomes: concentration of lipids in proximity to CD1 proteins, acid‐mediated loading of lipids into the CD1 groove, glycosidase‐mediated trimming of glycans, and lipid‐binding protein‐mediated antigen transfer to CD1. CD1 proteins are expressed on myeloid DCs, Langerhans cells, B cells, and other professional APCs, which use receptor‐mediated and other mechanisms to concentrate certain classes of lipids within endosomes. Cell surface receptors that have been shown to mediate uptake and presentation of glycolipid


101

antigens include two lectins, the mannose receptor and langerin, whose expression on DCs contributes to glycolipid‐mediated T cell activation (Hunger et al., 2004; Prigozy et al., 1997). In addition, new evidence shows that biotin‐specific B cell antigen receptors (BCR) can promote T cell recognition of biotinylated glycolipids, leading to the speculation that BCRs on glycolipid‐specific B cells might promote uptake and processing of natural glycolipid antigens (Lang et al., 2005). Also, a new study shows that deletion of expression of apolipoprotein E (apoE) affects presentation of bacterial antigens to diverse CD1‐restricted T cells and NK T cells (van den Elzen, et al., 2005). ApoE is expressed at high levels in endosomes and also functions as a major constituent of very low density lipoprotein (VLDL) particles in the bloodstream, where it mediates binding to the low density lipoprotein (LDL) receptors. Identification of lipoprotein pathways for antigen delivery provides a link between systemic circulation of lipids and their transport to APCs, illustrating how cellular pathways that had been previously known to control lipid metabolism also control immune recognition of lipids. In addition to these receptor‐mediated mechanisms, maturing DCs regulate macropinocytosis of antigen‐containing aggregates in ways that affect uptake of CD1‐presented lipid antigens (Roura‐Mir et al., 2005b). Last, lipids with particularly long alkyl chains preferentially accumulate in CD1b‐containing lysosomes via as yet undefined mechanisms (Moody et al., 2002). Collectively, these studies support the concept that the lumen of endosomes or perhaps the inner leaflet of endosomal membranes represents an antigen depot, where certain types of lipids accumulate to high local concentrations. These studies show how endosomes, in contrast to other subcellular compartments, are enriched for exogenously acquired lipids, including those that are selectively imported by pattern recognition receptors or phagocytosis triggered by direct encounter with pathogens. The ability of CD1 proteins to encounter distinct subsets of lipids during passage through endosomes may lead to their preferential display based on mass‐action, independent of any processes that increase the efficiency of antigen loading at this site. A second general mechanism by which endosomal cofactors promote lipid antigen recognition involves the acidic environment of late endosomes and lysosomes, which is maintained by vesicular ATPase proton pumps. The well documented ability of endosomal acidification inhibitors to block antigen processing may relate to the ability of low pH to partially denature CD1 proteins, as suggested by studies of recombinant CD1b binding with lipoarabinomannan and glucose monomycolate antigens (Ernst et al., 1998). These in vitro studies point to a model in which CD1 proteins remain in a closed conformation at neutral pH when moving through the secretory pathway, but open up to accept ligands, especially ligands with longer alkyl chains, as they

102

D. BRANCH MOODY

traffic through late endosomes. Low pH may also indirectly promote antigen complex formation by activating pH‐dependent protein cofactors other than CD1, such as lipid transport proteins or hydrolases that act to alter the structure of endosomally localized antigens. Although lipases that might modify the lipid anchors of antigens in ways that affect binding to CD1 have not been identified, endosomally localized, pH‐dependent glycosidases can alter glycan structures to reveal TCR epitopes. For example, deletion of a‐galactosidase A prevents the conversion of synthetic digalactosyl‐a‐ceramides to antigenic monogalactosyl ceramides (Prigozy et al., 2001). Similarly, deletion of b‐hexosaminidase B reduces activation of NK T cells in vivo, by altering the self glycolipids present in endosomes (Zhou et al., 2004b). Last, certain lipid transfer proteins, including the saposin and apoE, localize to endosomes at relatively high concentrations, so that this environment may be particularly rich in lipid transfer proteins. Saposin lipid transfer proteins, also known as sphingolipid activator proteins, are a family of proteins that are derived from cleavage of the inactive precursor protein, prosaposin. At low pH, saposins bind to the inner leaflet of endosomal membranes, where they partially insert into the membrane causing local disruptions in lipid packing, which is thought to allow extraction of lipids into the lumen. Saposin family proteins play a key role in regulating the glycolipid transport and content, such that altered expression of saposins cause human lipid storage diseases (Sandhoff and Kolter, 2003). Deletion of prosaposin in mice or in human cells leads to decreased activation and positive selection of CD1‐restricted T cells both in vitro and in vivo (Kang and Cresswell, 2004; Winau et al., 2004; Zhou et al., 2004a). The observed loss of CD1‐restricted T cell function might result indirectly from the global alterations in endosomal glycolipid content that follow prosaposin deletion. However, in vitro studies show that saposins transport certain lipids between vesicles, providing evidence for an intermembrane antigen transport function, suggesting a direct role in antigen loading onto CD1 (Zhou et al., 2004b). ApoE is also expressed at high levels in endosomes, so that its effects in antigen recognition may relate to a dual role in mediating the capture of lipoprotein particles by LDL‐R and in solubilization of endosomally localized antigens (van den Elzen et al., 2005). 4. 3‐Dimensional Structures of CD1‐b2‐Microglobulin‐Lipid Complexes 4.1. Conserved Features of CD1 Structure Both CD1 and MHC class I are comprised of heavy chains of similar length, which are organized into three extracellular domains (a1, a2, and a3) and bind b‐2 microglobulin to form heterodimers in cells. The crystal structure of


103

murine CD1d showed that the a1–a2 superdomain forms a hollow groove, which is delimited by two anti‐parallel a‐helices that sit atop six b‐strands, forming a b‐sheet. The three‐dimensional folding of mouse CD1 parallels that of MHC class I and II, such that it is possible to overlay the a‐carbon backbone traces and compare the structures of their antigen‐binding grooves (Zeng et al., 1997). However, several key features distinguish the overall architecture of the antigen‐binding grooves of CD1 from those found in MHC‐encoded antigen‐presenting molecules. First, CD1 grooves have a larger internal volume, which results in large part from a series of bulky amino acids at positions 18, 40, and 49 in the consensus sequence (Fig. 3). These residues are known as the a1 scaffold because their large side chains are positioned between the b‐sheet floor and the a1 helix, where they function to displace the a1 helix vertically. The scaffold creates greater depth for CD1 grooves, which are much larger (1350 to 2200 A˚3) than those found in MHC‐encoded antigen‐presenting molecules. The a1 scaffold has been observed in all CD1 proteins crystallized to date—murine CD1d, human CD1b, human CD1a, and human CD1d (Batuwangala et al., 2003; Gadola et al., 2002; Giabbai et al., 2005; Koch et al., 2005; Zajonc et al., 2003, 2005a,b; Zeng et al., 1997). Furthermore, bulky amino acids are present at consensus positions 18, 40, and 49 in most or all CD1 sequences in mammalian and avian species studied to date (Fig. 3). Thus, deep grooves represent a highly conserved feature of mammalian CD1 protein structure, which allow lipids to bind such that a substantial portion of the ligands lies within, rather than on top of, the a1–a2 superdomain. This mode of binding allows the aliphatic hydrocarbon chains of antigens to be extensively sequestered from the aqueous biological solutions that surround the CD1 protein. A second contrasting feature of CD1‐ and MHC‐binding grooves is the presence of non‐polar amino acids that are positioned at the inner surface of the CD1 groove. These amino acids form a nearly continuous hydrophobic surface that interacts with the aliphatic hydrocarbon chains of lipid ligands (Zeng et al., 1997). This situation contrasts with the polar and charged amino acids that allow for an extensive hydrogen bonding network and ionic interactions with peptide side chains with MHC groove surfaces. These differences in electrostatic topography of the grooves point to a basic difference in the chemical basis for ligand capture by CD1 and MHC antigen‐presenting molecules. Hydrophobic interactions, which predominate in CD1, generate binding force. Unlike salt bridges and hydrogen bonds, they do not require precise positioning of the individual chemical elements that constitute the binding pairs. Thus, any particular methylene unit within the alkyl chain can interact with nearly any point on the hydrophobic surface of CD1, and binding is optimized when the total number of hydrophobic interactions is maximized, as

104

D. BRANCH MOODY

Figure 3 Functionally dominant amino acids near the antigen‐binding groove. This analysis aligns the a1 and a2 domains non‐human CD1 sequences based on a previously defined consensus


105

would be the case when the lipid ligand fully occupies the groove. This contrasts with ionic and hydrogen bonding interactions which require that binding pairs be positioned at a discrete distance from its binding partner. Whereas MHC grooves are open for direct access to solvent throughout their length, interdomain interactions between the a1 and a2 helices close the superior margin of the groove at one end. In murine CD1d, this closure is formed by interdomain interactions between Leu66‐Thr167 and Phe70‐ Leu163 and can be thought of as a roof that encloses part of the superior aspect of the groove. This roof structure has been seen in crystal structures of murine CD1d, human CD1d, human CD1b, and human CD1a (Fig. 4), and the roof‐forming amino acids at consensus positions 66, 70, 162, and 166 are generally conserved in non‐human CD1 sequences (Fig. 3). Thus, whereas the MHC class I and II grooves are directly exposed to solvent over most of their 25 A˚ length, CD1 grooves are accessed by a smaller, 10 to 17 A˚ long gap. Because this gap is relatively small and connects the outer surface of CD1 to a large interior cavity, the main opening to the groove can be thought of as a portal. 4.2. CD1 Pockets and Portals A series of seven CD1‐lipid crystal structures were recently solved in a short time period, providing a wealth of new structural data that forms the basis for a standardized nomenclature of groove‐associated structures and sheds light on

sequence for human and mouse proteins (Porcelli, 1995), and shows ribbon and schematic diagrams of the CD1b‐phoshatidylinositol structure (Gadola et al., 2002). Amino acids are shown for those positions in the consensus sequence that form specialized structural features near the antigen‐binding groove as identified in crystal structures of human CD1a, human CD1b, human CD1d, or mouse CD1d proteins: a1 scaffold (bulky residues that raise the a1 helix), A0 roof (interdomain contacts involving leucine, isoleucine, threonine, or size‐conserved amino acids that close the superior margin of the A0 pocket), A0 pole (hydrophobic and bulky residues that form a vertical axis in the A0 pocket), A0 terminus (residues with side chains that are equal to or larger than valine and located near the inferior margin of the A0 pocket), C0 origin (small non‐aromatic residues that create a cavity forming the opening to the C0 pocket), C0 portal (two cysteine residues that form a disulfide bond to stabilize the C0 portal), and T0 tunnel (two glycine residues that create a cavity that forms the T0 tunnel). Grey shading identifies residues that are conserved in size or chemical characteristics as compared with residues that define the structural element in crystallized CD1 proteins. This analysis suggests that the a1 scaffold, the A0 roof, and the A0 pole are likely conserved in mammalian CD1 proteins, whereas other features are likely to be found in only certain CD1 isoforms

106

D. BRANCH MOODY

Figure 4 CD1 antigen‐binding groove surfaces. Viewed from the superior aspect of the a‐helical surfaces of CD1, the antigen‐binding grooves of human CD1a, human CD1b, and murine CD1d proteins are rendered as transparent surfaces and shown with bound ligands. Ligands traverse a portal above the F0 pocket, whereas the top of the A0 pocket is covered by a roof‐like structure. Only CD1b has a second portal, the C0 portal, which connects the bottom of the groove with the outer, lateral surface of CD1b. All three grooves have an ovoid A0 pocket that encircles the A0 pole formed by phenylalanine at position 70 (F 70) and valine or cysteine at position 12 (V12, C12). Whereas the A0 pockets of CD1b and CD1d join directly with the F0 pocket (upper left corner), the A0 pocket of CD1a is terminated by a valine at position 28 (V28). This difference in structure allows the longer glucose monomycolate ligand to encircle the A0 pole of CD1b and exit to other pockets, whereas the terminus of the A0 pocket of CD1a places a limit on the length of alkyl chains that can bind in this groove. Groove surfaces are based on previous depictions or crystal structures of CD1‐ lipid complexes (Batuwangala et al., 2003; Giabbai et al., 2005; Moody et al., 2005; Zajonc et al., 2005b).

mechanisms of lipid antigen capture. These structures are CD1b bound to phosphatidylinositol (CD1b‐PI), CD1b‐GM2 ganglioside, CD1b‐glucose monomycolate, CD1a‐sulfatide, CD1a‐lipopeptide, CD1d‐phosphatidylcholine, and CD1d‐a‐galactosyl ceramide (Batuwangala et al., 2003; Gadola et al., 2002; Giabbai et al., 2005; Koch et al., 2005; Zajonc et al., 2003,


107

2005a,b). One clear conclusion derived from these new studies is that the antigen‐binding grooves of CD1a (1300 A˚3), CD1d (1650 A˚3), and CD1b (2200 A˚3) differ significantly in volume, suggesting that each isoform is specialized for binding antigens with lipid anchors of differing size. The overall structures of CD1‐binding grooves are composed of up to six named component structures. In human CD1b, there are four pockets (A0 , C0 , F0, T0 ) and two portals (F0, C0 ), whereas human CD1a and mouse and human CD1d proteins have only two pockets (A0 , F0 ) and one portal (F0 ). Three of the pockets, A0 , C0 , and F0, take their names based on their positions relative to the A, C, and F pockets in MHC class I (Zeng et al., 1997). The fourth pocket, known as the T0 tunnel, is named for its location at the bottom of the CD1b groove, where it connects A0 and F0 pockets to form the A0T0 F0 superchannel (Fig. 3). Because all CD1 proteins studied to date have a roof structure above the A0 pocket, the main portal for ligand entry is positioned somewhat ectopically on the distal surface of CD1, above the F0 pocket, so it is known as the F0 portal (Figs. 3 and 4). In human CD1b, there is a second, smaller portal located beneath the a2 helix at the end of the C0 pocket and is known as the C0 portal (Figs. 3 and 4). The differing structures of CD1 antigen‐binding grooves can be understood by visualizing how these pockets and portals connect together in different ways. 4.2.1. F0 and A0 Pockets The main entrance to the grooves of CD1 proteins is the F0 portal, which is located at the top of the F0 pocket. The F0 pocket is a somewhat vertically oriented cavity, which serves to connect the outer, TCR‐binding surface of CD1 to the interior of the groove. The alkyl chains of antigens, which serve to anchor antigens into CD1 proteins, descend through the F0 pocket and then wind around the A0 pocket, an ovoid concavity positioned beneath the A0 roof. The A0 pocket is partially or completed bisected by the A0 pole, which is formed by larger, hydrophobic residues at positions 12 and 70 (Fig. 3). These amino acids form an axis around which the alkyl chains of antigens are wrapped. An A0 pole is present in all CD1 structures crystallized to date, and larger, non‐polar amino acids are generally conserved at these positions in most CD1 sequences, providing evidence for conservation of this structure among mammalian CD1 proteins (Fig. 3). A key difference among CD1 isoforms is the directionality and the degree to which alkyl chains can encircle the A0 pole. For example, the A0 pocket of CD1a is abruptly terminated, so that the A0 pocket is like a narrow bent tube with a terminus that does not allow long alkyl chains to protrude into other

108

D. BRANCH MOODY

pockets (Fig. 4). This terminus creates a situation in which lipids of only a discrete length can fit within the pocket, leading to the description of the A0 pocket of CD1a as a ‘‘ruler’’ for alkyl chains (Zajonc et al., 2003). Supporting the hypothesis that CD1a selectively presents lipid antigens with a discrete chain length, lipopeptide analogs with alkyl chains that match the volume of the A0 pocket (C20) show higher potency for activating T cells than analogs with shorter (C18) alkyl chains (Moody et al., 2004; Zajonc et al., 2005b). The abrupt terminus of the A0 pocket in CD1a contrasts with the open‐ended A0 pockets of CD1b and CD1d, which connect directly to pockets within the groove. Also, glucose monomycolate and phosphatidylcholine traverse the donut‐shaped A0 pocket in opposite directions (Fig. 4), illustrating the more flexible ways in which CD1b and CD1d can bind lipids within the A0 pocket. The more interconnected nature of the A0 and F0 pockets in CD1 isoforms other than CD1a appears to allow binding of antigens whose lipid anchors vary more significantly in length. The particularly complex and interconnected nature of the pockets within CD1b led to the description of this groove as a ‘‘maze’’ for alkyl chains (Gadola et al., 2002). 4.2.2. The C0 Pocket and C0 Portal in CD1b CD1b has the largest and most complex groove structure of CD1 proteins studied to date. Whereas the grooves of other isoforms are completely enclosed so that the antigen‐binding cavity can be accessed only via the F0 portal, CD1b contains a second portal, the C0 portal, which takes its name from its location at the bottom of the C0 pocket (Gadola et al., 2002). The opening that forms the C0 portal is stabilized by a disulfide bond involving positions Cys131 and Cys145 in the consensus sequence. This pair of cysteines is not found in any other human or mouse CD1 isoform, suggesting that this is an isoform‐ specific adaptation of CD1b among human CD1 proteins (Fig. 3). The origin of the C0 pocket is created by a gap that results from the presence of a relatively small valine residue, compared to larger aromatic residues found at this position in other human CD1 isoforms (Fig. 3). The C0 pocket is open at both ends, so that it likely binds shorter lipids entirely within the pocket and allows longer lipids to protrude from the groove. Although crystal structures of CD1‐lipid complexes have not unambiguously depicted ligand density protruding through the C0 portal, CD1b is known to present antigens that match or exceed the volume of the CD1b groove (Gilleron et al., 2004; Moody et al., 2002). This leads to speculation that the C0 portal of CD1b functions as an escape hatch allowing presentation of antigen with particularly long alkyl chains.


109

4.3. CD1 Isoform Specificity The apparent promiscuity with which a given CD1 protein can bind lipid anchors of differing structures has led to interesting controversies regarding the possible isoform‐specific antigen‐capturing functions of CD1 proteins. On the one hand, CD1a, CD1b, CD1c, and CD1d have distinct patterns of expression on and within APCs, implying distinct functions. Further, the obvious differences in size and shape of CD1a, CD1b, and CD1d antigen‐binding grooves strongly suggest that CD1 isoforms are specialized to bind distinct classes of lipid antigen (Fig. 4). On the other hand, chemical motifs that describe the structures of antigens that bind to one CD1 isoform but not to others have not yet been well established. For example, it is known that despite the large differences in size and shape seen in the CD1b and CD1d antigen‐binding grooves, both proteins can present sphingolipids and diacylglycerols with similar or identical lipid anchors, which have an overall length of C32–44 (Fischer et al., 2004; Gumperz et al., 2000; Joyce et al., 1998; Kawano et al., 1997; Shamshiev et al., 1999, 2000; Sieling et al., 1995). Similarly, sulfatide antigens are presented by human CD1a, CD1b, CD1c, and CD1d proteins (Shamshiev et al., 2002). These studies raise the possibility that each CD1 isoform might present essentially the same spectrum of antigens. A complete answer to this question of isoform‐specific binding motifs will require elution of lipids from cellular CD1 proteins and better methods to measure specific binding to CD1 proteins in vitro. Nevertheless, certain insights from molecular and cellular studies indicate that isoform‐specific patterns of antigen capture likely do exist. For example, CD1c‐presented dolichols and mycoketides have a single alkyl chain (Matsunaga et al., 2004; Moody et al., 2000), whereas most antigens presented by other CD1 isoforms have two. The hypothesis that CD1c may be specialized to present lipids with a single lipid anchor has not yet been proven. However, it already seems clear that the large size of the CD1b groove and the existence of the C0 portal as a functional escape hatch represent isoform‐specific adaptations that allow presentation of long chain (C80) antigens that cannot fit within the confines of CD1a or CD1d grooves. Secondly, even when two CD1 isoforms bind the same antigen, they may do so with differing molecular mechanisms, so that the antigens may differ in the overall hierarchy of antigens presented by each CD1 isoform. For example, CD1b and CD1d both present sphingolipids. However, a‐galactosyl ceramides occupy the full capacity of the CD1d groove (Zajonc et al., 2005a), whereas GM2 ganglioside occupies only two of the four pockets present in CD1b (Gadola et al., 2002). Consistent with the markedly differing mechanisms by which these CD1 proteins bind the same type of ceramide anchor, CD1d presents a‐galactosyl ceramides at much

110

D. BRANCH MOODY

greater efficiency (nanomolar) than CD1b presents GM1 (micromolar) (Kawano et al., 1997; Shamshiev et al., 1999). 4.4. Functionally Dominant Aspects of CD1 Structure CD1 nomenclature and conventional analyses of relatedness among CD1 orthologs are based on overall levels of relatedness of proteins at the amino acid level. However, new information from CD1 crystal structures and functional studies have identified the particular amino acids that play dominant roles in CD1 function based on their abilities to mediate interactions with AP complexes or form the C0 portal, the T0 tunnel, and the terminus of the A0 pocket. Natural or experimentally induced point mutations in these functionally dominant positions lead to small changes in the overall sequence, but potentially large changes in the function of the protein. Figure 3 summarizes these functionally important amino acid positions, as determined from crystal studies of human and murine CD1 proteins, and then compares conservation of these residues among non‐human CD1 sequences. This analysis leads to questions about whether non‐human CD1 proteins that are denoted as orthologs of a given human CD1 isoform do indeed have the same or similar functions as their human counterparts. For example, the ability of human CD1b proteins to present particularly large lipids likely comes about in part due to the T0 tunnel, which results from two particularly small residues, glycines at positions 98 and 116 in the CD1 consensus sequence. These small amino acids leave a gap that forms the T0 tunnel and connects the A0 and F0 pockets to form a long A0T0 F0 superchannel. Whereas glycines are generally conserved at these positions among non‐human proteins that are designated as CD1b orthologs, guinea pig CD1b4 has larger valines at these positions and may therefore lack the T0 tunnel (Fig. 3). Likewise, the C0 portal contributes to the ability of human CD1b proteins to present large lipids. However, guinea pig CD1b4 and rabbit CD1b1 lack the cysteines at positions 131 and 145, suggesting that these proteins may lack an escape mechanism for larger alkyl chains. Thus, guinea pig CD1b4 provides an example of a protein that is designated as an ortholog of human CD1b based on its overall sequence, yet appears to lack key amino acids that form the structures that allow the groove in human CD1b to present lipids with long alkyl chains. Related to this, guinea pig CD1b3 might fail to capture lipids in late endosomes, as it is unique among CD1b orthologs in its lack of an identifiable tyrosine or dileucine endosomal targeting sequence (Fig. 2). These observations point to possible differences in function among guinea pig CD1b orthologs and might explain the basis of how this species retained a particularly large family of group 1 CD1 proteins (Dascher et al., 1999).


111

Other questions arise from inspection of non‐conserved amino acids at functionally important positions in the consensus sequence. For example, non‐human CD1a sequences differ from human CD1a in key residues that are known to control antigen binding and endosomal trafficking. The valine at position 28 in human CD1a forms the terminus of the A0 pocket, which allows this pocket to bind lipids with a discrete chain length. Yet, the rabbit CD1a sequence at position 28 is more like that of human CD1b and mouse CD1d proteins, which use their open‐ended A0 pockets to interconnect with other pockets (Fig. 3). Similarly, human CD1a has been proposed to have a specialized function in surveying the content of early endosomes, based on the lack of endosomal targeting sequences in its cytoplasmic tail. However, non‐human human CD1a proteins have longer tails with putative endosomal localization sequences (Fig. 2). Although these questions related to the conservation of function in mammalian CD1a and CD1b proteins remain unanswered, the differential use of dileucine and tyrosine (YXXZ) motifs in the tails of CD1d orthologs provides a well documented example of how two species use distinct cellular mechanisms to promote recycling to late endosomes. Human, but not mouse CD1d has a dileucine motif that promotes steady state localization to endosomes (Rodionov et al., 1999, 2000). Conversely, the YXXZ motif in murine CD1d (YQDI) can interact with AP‐2 and AP‐3, whereas the human CD1d tail sequence (YQGV) binds only to the m‐subunit of AP‐2 (Elewaut et al., 2003; Sugita et al., 2002). One hypothesis is that this represents an unusual example of convergent evolution in which two closely related species evolved separate mechanisms to navigate into late endosomal compartments. Similarly, it has been proposed that the mouse CD1 system, which lacks an ortholog of CD1b, may have acquired point mutations that enable its YXXZ motif to allow CD1d binding to AP‐3, so that the mouse CD1d protein can sample lysosomal antigens in a way that is accomplished by CD1b in human systems (Dascher and Brenner, 2003). Both of these hypotheses emphasize the view that certain kinds of antigen processing reactions do not merely require antigen delivery to endosomes, but also require the particularly acidic environment ( pH 4.5) of late endosomes and lysosomes. 5. Microbial Antigens and Infectious Disease The range of known antigens presented by the CD1 system has grown rapidly in recent years. Whereas the early studies of CD1‐restricted T cells emphasized T cell activation by a‐galactosyl ceramide and several specialized lipids found only in mycobacteria, CD1 proteins have now been shown to mediate T cell activation in response to at least one example of most major classes

112

D. BRANCH MOODY

of lipids found in mammalian cells, including phosphatidylinositols, glycosyl phosphatidylinositols, phosphatidylcholines, phosphatidylethanolamines, gangliosides, sulfatides, isoglobosides, and polyisoprenols. In addition, CD1 proteins present chemically diverse bacterial lipids such as mycolates, polyketides, acylated carbohydrates, and lipopeptides (Fig. 5). The apparent structural diversity of lipid antigens for the CD1 system is increasingly leading to the view that CD1 proteins do not solely present lipids with highly specialized structures or functions, but instead broadly sample the lipid content of cells for display. This in turn raises important practical questions about how CD1 proteins can bind to such structurally diverse ligands and how cells regulate the process by which different classes of self and foreign lipids compete for loading onto CD1. 5.1. Microbial Antigens The antigen presentation functions of CD1 were discovered through the study of human T cell responses to Mycobacterium tuberculosis. After isolation of T cell clones that respond to antigens presented by CD1a, CD1b, or CD1c, the stimulatory compounds were isolated and their structures identified as long chain, a‐branched, b‐hydroxy fatty acids known as mycolic acids (Beckman et al., 1994), lipoarabinomannans (Sieling et al., 1995), glucose monomycolates (Moody et al., 1997), mannosyl phosphomycoketides (Matsunaga et al., 2004; Moody et al., 2000), acylated sulfotrehaloses (Gilleron et al., 2004), and dideoxymycobactin lipopeptides (Moody et al., 2004) (Fig. 5). There is evidence that all of these lipids are made by mycobacterial species that infect mammals, and many of these antigens have been shown to correlate with or cause virulence. For example, mannosyl phosphomycoketides and acyl sulfotrehaloses are generally lacking in saprophytic mycobacteria, but are present in species capable of infecting cells (Domenech et al., 2004; Matsunaga et al., 2004). Deletion of pathways leading to the synthesis of mycobactin siderophores and dideoxymycobactin antigens reduce growth of M. tuberculosis in human cells (De Voss et al., 2000). Whereas most mycobacterial antigens are presented by group 1 CD1 proteins, there is new evidence that mycobacterial phosphatidylinositol mannosides (PIM) can bind to CD1d and activate NK T cells (Fischer et al., 2004). Two recent studies have provided convincing molecular evidence that NK T cells recognize structurally related, a‐linked sphingolipids from Sphingomonas capsulata, Sphingomonas paucimobilis, Sphingomonas yanoikuyae, or Ehrlichia muris (Kinjo et al., 2005; Mattner et al., 2005). Although is it not clear whether sphingomonas species, which have only rarely been shown to infect immunocompromised humans, are of clinical significance for human disease,


113

Figure 5 Glycolipid stimulants of CD1‐restricted T cells. Human and mouse CD1 proteins are known to mediate T cell responses to structurally diverse classes of mammalian, microbial, and synthetic glycolipids.

this new insight is notable for several reasons. First, these naturally produced bacterial antigens have glucuronic or galacturonic acids that are in a‐anomeric linkage with the sphingosine base (Fig. 5). This a‐linkage represents a chemical feature that distinguishes these foreign lipids from most mammalian

114

D. BRANCH MOODY

glycosyl ceramides, which typically are in b‐linkage to the sphingosine base. The difference in linkage serves to pivot the carbohydrate moiety into distinct positions. Therefore, the unusual a‐anomeric linkage plausibly serves as a chemical feature that allows these lipids to be recognized as foreign in comparison to mammalian sphingolipids that are otherwise similar in structure. In addition, this a‐linkage parallels a key feature of previously identified marine sponge‐derived a‐galactosyl ceramides, which are potent activators of NK T cells, but are not known to be produced by any organism that normally comes into contact with the mammalian immune system (Kawano et al., 1997). Therefore, identification of a‐linked sphingolipids in bacteria represents a step forward in identifying natural foreign antigens for NK T cells. Last, it is notable that the sphingomonas cell walls, unlike most gram negative bacteria, lack lipopolysaccharide (LPS) agonists of TLR‐4. This leads to the speculation that the invariant TCRs on NK T cells serve as a substitute for the well known means of innate activation through LPS and TLR‐4. In addition to these bacterial antigens for NK T cells, glycolipid stimulants of NK T cells have been identified from Leishmania, Trypanosoma, and Plasmodium species. CD1d knockout increases susceptibility of mice to infection by Leishmania donovani, and CD1‐tetramers loaded with lipophosphoglycan components of the outer cell wall of leishmania stain NK T cells, suggesting that activation by this antigen occurs through the TCR (Amprey et al., 2004). Glycosyl phosphatidylinositol anchors that resemble those found in Plasmodium and Trypanosoma species activate NK T cells in a CD1d‐dependent manner in vitro, and these glycolipids likewise bind to CD1d tetramers. In addition, under certain conditions, CD1d knockout can lead to altered B cell responses or outcomes of experimental models of malaria infection (Hansen et al., 2003a,b; Schofield et al., 1999). However, other studies have suggested that CD1d does not contribute to control of experimental infection with Plasmodium berghei (Molano et al., 2000) or activation of NK T cells (Procopio et al., 2002). In addition to the microbial compounds of known structure summarized in Fig. 5, CD1‐restricted T cells have been reported to be activated by a variety of bacteria, fungi, pathogenic protozoa, and viruses, although the precise structures of any antigenic compounds produced by these pathogens remain to be determined. 5.2. Diverse CD1‐Restricted T Cells in Infectious Disease In follow up to studies showing that human T cell clones could limit the in vitro growth of M. tuberculosis via cytolysis, granulysin, and g‐interferon (Stenger et al., 1997, 1998a), more recent studies have sought to detect lipid‐reactive T cells during in vivo infection. Because most or all antigens were presented by


115

group 1 CD1 isoforms that lack orthologs in mice, in vivo studies were carried out with human patients with resolving subclinical M. tuberculosis infections (Agea et al., 2005; Moody et al., 2000; Ulrichs et al., 2003) or pulmonary tuberculosis (Gilleron et al., 2004). Each of these clinical studies found similar results. Fresh peripheral blood lymphocytes reactive against mannosyl phosphodolichols, mycolic acids, glucose monomycolates, or acyl sulfotrehaloses were not detectable in healthy controls, but were found in the majority of infected patients, with some evidence that they occur with high precursor frequency. These studies provide evidence that CD1‐ and lipid‐reactive T cells are activated during the acute phases of a human infectious disease and have prompted ongoing studies as to whether such responses persist such that they can be considered memory responses. Whether or not such lipid‐reactive T cells are protective against infection awaits further studies in animal models that express group 1 CD1 proteins. One study of guinea pigs, which express orthologs of human group 1 CD1 proteins, found that immunization with a mixture of lipid and protein antigens substantially reduces the size of pulmonary granulomata and provides some reduction in mycobacterial burden after challenge (Dascher et al., 2003). 5.3. NK T Cells and Infectious Disease CD1d and NK T cells have been implicated in affecting outcomes in response to a wide variety of infections by bacteria, viruses, fungi, and protozoa. Interpretation of the biological significance of these many findings hinges in part on certain aspects of study design relating to the antigens used and the means of deleting NK T cells. One general issue relates to experimental means to separately study the effects of invariant NK T cells with Va14Ja18 TCRs (previously known as Va14Ja281 TCRs) versus the larger repertoire of T cells that recognize CD1d, which includes T cells that do not express this particular TCR (varied NK T cells). The latter population was originally identified based on in vitro analysis of CD1d‐restricted T cells taken from mice and humans, and is receiving more scrutiny based on in vivo studies of infection showing that this population can mediate effects that are independent of invariant NK T cells (Behar and Cardell, 2000). CD1d deletion results in loss of both variable and invariant NK T cells. Invariant NK T cells can be selectively removed by germline deletion of the sequences containing the Ja18 segment (Kawano et al., 1997) or be selectively augmented by sorting for T cells that stain a‐galoctosyl ceramide‐loaded tetramers or antibodies against TCR a chains with the Va14 gene segment. Taking advantage of these methods, recent studies have suggested a role for variable NK T cells in mediating tissue damage in response to Coxsackie virus

116

D. BRANCH MOODY

or hepatitis B virus (HBV) genes. In the former study, it was found that post‐ viral myocarditis was attenuated by CD1d deletion, but not Ja18 deletion, implicating variable NK T cells in causing the post‐infectious myocardial damage (Huber et al., 2003). Likewise, HBV gene‐mediated liver damage could be transferred by CD1d‐restricted cells that did not stain for Va14 or bind to a‐galactosyl ceramide‐loaded CD1d tetramers, implicating variable NK T cells in this model of hepatitis (Baron et al., 2002). A second general issue relates to outcomes of infection after treatment of animals with a particularly potent synthetic agonist of NK T cells, a‐galactosyl ceramide. Administration of this compound leads to systemically detectable levels of g‐interferon, a broadly acting anti‐microbial Th1 cytokine. It has long been speculated that a‐galactosyl ceramides are superagonists in the sense that they activate NK T cells more strongly than naturally occurring antigens for NK T cells. Recent studies have identified naturally occurring stimulants of NK T cells, including isoglobosides, sphingomonas‐derived ceramides, lipophosphoglycans, and phosphatidylinositol mannosides. Comparison of these lipids with synthetic a‐galacotsyl ceramides has generally found that naturally occurring antigens less potently activate NK T cells or mediate tetramer staining of NK T cells, as compared with a‐galactosyl ceramide (Amprey et al., 2004; Fischer et al., 2004; Kinjo et al., 2005; Mattner et al., 2005; Zhou et al., 2004b). Because a‐galactosyl ceramide appears to act much more strongly than natural antigens, evidence that a‐galactosyl ceramide mediates protection provides useful evidence supporting the use of this glycolipid as an immunotherapeutic agent, but does not necessarily imply a natural function for NK T cells in infection. For example, one study found that deletion of Va14Ja18 NK T cells does not affect the outcome of acute cytomegaloviral (CMV) infections, but administration of a‐galactosyl ceramide provides some protection (van Dommelen et al., 2003). The apparent differences in outcomes after these two experimental manipulations might be explained if NK T cells do indeed have an antiviral function, but the pathogen fails to provide a stimulus of comparable potency to that of a‐galactosyl ceramide. Several recent studies show that deletion of CD1d results in worsened outcomes of experimental viral infections. For example, deletion of CD1d or the combined deletion of NK and NK T cells leads to increased susceptibility to infections with intravaginal herpes simplex virus type 2 infections in mice (Ashkar and Rosenthal, 2003). Mouse infections with herpes simplex virus type 1 (HSV 1) are worsened in mice lacking either CD1d or the Ja14Va18 TCR, implicating a role for invariant NK T cells in the response (Grubor‐Bauk et al., 2003). Last, there is one case report of disseminated infection of a human with an attenuated strain of varicella zoster virus in a human with low levels of NK T


117

cells (Levy et al., 2003). Collectively, these reports suggest that NK T cells can play a determinative role in infections with herpes viruses, a class of viruses that persist long term in mammalian hosts. Conversely, there is some evidence that CD1d and NK T cells do not affect outcomes in more acute infections with the SARS carona virus and lymphocytic choriomeningitis virus (Glass et al., 2004; Spence et al., 2001). 5.4. Virally Mediated Alterations in CD1 Expression Coupled to this evidence that CD1d protects against viral infections, there is new evidence that viruses modulate CD1 expression in ways that might lead to immune evasion. Kaposi sarcoma‐associated herpes virus (KSHV) downregulates cell surface CD1d expression via modulator of immune recognition (MIR)‐mediated reinternalization and ubiquitination (Sanchez et al., 2005). Similarly, the human immunodeficiency virus (HIV‐1) or other viral vectors expressing the HIV nuclear envelope factor (nef) gene partially downregulate human expression of CD1a and CD1d proteins on infected cells (Cho et al., 2005; Shinya et al., 2004). In both cases, nef‐mediated downregulation is lessened when the cytoplasmic tail of CD1 proteins is truncated, implicating altered trafficking as the mechanism of loss of cell surface CD1. This redirected trafficking of CD1 may be related to the previously known mechanism by which HIV nef induces downregulation of MHC class I by revealing a cryptic cytoplasmic tail sequence that interacts with adaptor protein complexes (Le Gall et al., 1998). However, the mechanisms involving CD1 and MHC class I are at least somewhat distinct, as MHC class I accumulates in the golgi, and CD1a loss appears to involve redirected transport to lysosomes. Given the emerging evidence that CD1d mediates protective responses during viral infections, especially herpes viruses, immune evasion through downregulation of CD1 expression is a plausible and interesting new concept. Understanding how partial downregulation of CD1 expression on the surface of infected cells translates into biologically meaningful immune evasion requires further work on several basic questions. Some evidence indicates that CD1 downregulation is a cell‐autonomous effect (Quaranta et al., 2002), raising questions about how viruses that persist in cells that do not normally express CD1 can transfer inhibitory effects to CD1‐expressing cells. Also, research into immune evasion by pathogens has in some cases outpaced work to determine whether CD1‐restricted T cells play a determinative role in infection. CD1 proteins and NK T cells are not uniformly active against all viruses, so strong support for active subversion of immune recognition must involve the search for evidence that CD1 proteins are determinative for the natural infectious process itself.

118

D. BRANCH MOODY

5.5. Indirect Means of T Cell Activation by Glycolipids The primary means of activation of CD1‐restricted T cells involves direct contact of CD1‐lipid complexes with TCRs, as shown by experiments in which antigen‐dependent T cells can be stained with fluorescent, lipid‐loaded tetramers (Benlagha et al., 2000; Karadimitris et al., 2001; Matsuda et al., 2000). Also supporting this conclusion, truncated TCRs bind with high affinity to CD1‐lipid complexes, and TCR‐lipid‐CD1 complexes can be modeled based on the crystal structures of TCRs and CD1‐lipid complexes (Grant et al., 2002; Sidobre et al., 2002; Zajonc et al., 2005b). The strongest evidence that any given glycolipid functions as a cognate antigen that directly contacts the TCR is seen when (1) it activates T cells, (2) activation requires CD1 proteins, (3) purified CD1‐lipid complexes activate T cells, and (4) the lipid lacks adjuvant or mitogenic properties. Alpha‐galactosyl ceramide likely meets all of these criteria and can be considered to act as a cognate antigen, based on the ability of CD1d‐a‐galactosyl ceramide complexes to bind to and activate TCRs. Similarly, gangliosides, sulfatides, mycolates, and glucose monomycolates can stimulate T cells when added in complex with recombinant CD1 proteins, strongly pointing to the likelihood that their mechanism of action involves direct contact with variable regions of the TCR. Thus, many lipid stimulants of T cells act as classical cognate antigens in the sense that they provide stimulatory signals through the variable regions of the TCR. Several recent studies have shown that bacteria, bacterial lipids, or agonists of toll‐like receptors in ways that do not primarily involve TCR contact. Instead effects are mediated through increased CD1 expression, increases lipid antigen uptake, altered the production of endogenous lipids or secretion of activating cytokines by APCs (Brigl et al., 2003; De Libero et al., 2005; Krutzik et al., 2005; Mattner et al., 2005; Roura‐Mir et al., 2005b). These studies suggest bacterial lipids, especially those with the ability to agonize TLR‐2 or TLR‐4, may activate CD1‐restricted T cells by an indirect mechanism that requires two separate signals, APC activation combined with TCR‐CD1 contacts. For example, in vitro studies show that gram negative bacterial lipopolysaccharide leads to NK T cell activation through an indirect mechanism that involves TLR‐4‐mediated secretion of IL‐12 by myeloid APCs, rather than direct binding to CD1d and presentation to TCRs (Brigl et al., 2003). In fact, the failure of b‐hexosaminidase‐deficient mice to show NK T cell responses to LPS suggests that the role of LPS in vivo may involve stimulation of production of endogenous isogloboside antigens within the APC, rather than LPS‐ mediated contact with T cell receptors (Mattner et al., 2005). These new studies raise questions about whether some of the lipid stimulants depicted


119

in Fig. 5 act mainly as cognate antigens for TCRs or function via indirect mechanisms by activating APCs. Further studies of the mechanisms of glycolipid‐mediated T cell activation have obvious importance for designing immunotherapies (Section 7) and may solve controversies relating to discordant results with the same glycolipid by different types of APCs. 6. Self Antigens, Autoreactivity, and Autoimmune Disease 6.1. Structures of Self Lipid Antigens Early studies of CD1‐mediated T cell activation showed that CD1‐autoreactive T cells could be selectively activated by CD1‐expressing APCs, but unlike T cells described in the prior section, did not require any exogenous or foreign lipid antigen (Porcelli et al., 1989). Such CD1 autoreactivity might have occurred as a result of TCR recognition of unliganded CD1 proteins or by CD1 proteins that were bound by endogenous self lipid antigens. Studies showing that GM1 ganglioside‐CD1b complexes can activate T cells provided more direct evidence for presentation of antigenic self lipids to T cells (Shamshiev et al., 1999, 2000). The ability of CD1 proteins to bind and present self lipids is now further supported by studies that have directly eluted phosphatidylinositols from CD1 and shown that they can activate NK T cells under certain circumstances (Gumperz et al., 2000; Joyce et al., 1998). In addition, sulfatides (Jahng et al., 2004; Shamshiev et al., 2002), ganglioside GD3 (Wu et al., 2003), phosphatidylethanolamine (Rauch et al., 2003), and other diacylglycerols (Agea et al., 2005) can lead to activation of human CD1‐ restricted T cells (Fig. 5). 6.2. Candidate Mechanisms for Regulating Responses to Self Antigens The identification of self cellular lipids as stimulants for CD1‐restricted T cells in vitro now raises basic questions about how T cell responses to self antigens are regulated in vivo. Although there is evidence for positive selection of NK T cells in the thymus (Benlagha et al., 2002), it is not yet known whether diverse CD1‐restricted T cells undergo negative selection in a process that is equivalent to mechanisms that are well established to control central tolerance in MHC‐restricted T cells (Wei et al., 2005). In fact, until recently there was little insight into the mechanisms of regulating autoreactivity to cellular lipids. However, new studies suggest several candidate mechanisms that could act in the periphery to regulate T cell stimulation by self lipids.

120

D. BRANCH MOODY

6.2.1. Alterations of Self Lipid Antigen Production in Cells One possibility is that self lipid antigens with an ability to stimulate T cells are not constitutively synthesized, but instead are produced in response to activating stimuli. For example, a recent study has shown that knockout of a lysosome enzyme, b‐hexosaminidase b (Hex b), leads to loss of NK T cells in the periphery. Separately, it was found that isoglobotrihexosylceramide (iGb3), a mammalian glycolipid produced by this enzyme, can activate invariant NK T cells (Zhou et al., 2004b). Isoglobosides are known to exist in mammalian cells, but are not so abundant that their functions and cellular distributions are well understood. These findings now suggest that as yet unidentified upstream stimuli, which alter the production, intracellular localization, or loading of isogloboside lipids onto CD1d proteins, regulate the activation of NK T cells in vivo. 6.2.2. Adjuvants for CD1‐Restricted T Cells A second general possibility is that APCs require a second stimulus, in addition to self antigens bound in the groove, to optimally present lipid antigens to T cells. As discussed earlier, in vitro studies of antigen presentation and CD1 expression show that microbial cell wall products can alter levels of CD1 expression, rates of lipid antigen delivery to endosomes, and other factors that convert monocytes into competent antigen‐presenting cells (De Libero et al., 2005; Krutzik et al., 2005; Roura‐Mir et al., 2005b). Myeloid DCs and their precursors migrate widely among tissues and initiate T cell responses. In theory, a system that limits group 1 CD1 expression to those myeloid DC precursors that have directly encountered pathogens, and thereby received activating stimuli, may limit unwanted presentation of self lipid antigens in other situations. A recent study has shown that infection of human DCs with bacteria or treatment with bacterial products increases the ability of myeloid cells to present self gangliosides and sulfatides to CD1a‐ and CD1b‐restricted T cell clones, thereby highlighting the possibility that bacteria turn on pathways that lead to presentation of both self and foreign lipids (De Libero et al., 2005). 6.2.3. Regulation by Subcellular Antigen Processing Pathways A third possibility is that regulated mechanisms of antigen loading within endosomes can alter the spectrum of self and foreign lipids presented to T cells. This hypothesis can be considered by comparing emerging concepts in subcellular pathways of lipid antigen presentation with well‐established subcellular pathways of peptide presentation. To generate peptide antigens, nearly


121

all proteins require cellular processing, which occurs in one of two distinct pathways, depending on whether peptides are generated in the endosomes or in the cytosol (Cresswell, 1994). The general cellular requirements for lipid antigen presentation differ from those of proteins in two important ways. First, individual lipid antigens presented by the CD1 system vary in the extent to which their recognition requires cellular processing, such that certain antigens absolutely require cellular processing and others can be readily recognized after binding to recombinant CD1 proteins at neutral pH. Secondly, most or all cellular cofactors for lipid antigen recognition identified to date act on antigens within the lumen of intracellular compartments, especially late endosomes. This leads to a model in which the separate pathways of lipid antigen presentation are defined in part based on the identity of the antigen and the stringency of its loading requirements, with most high‐stringency loading interactions occurring in endosomes (Table 1). The antigen‐binding domains of CD1 proteins remain topologically confined to lumenal compartments and cell surface of APCs, so most or all antigens likely contact the CD1 groove after delivery from the lumenal leaflet of membranes, intralumenal lipid‐protein complexes, intralumenal microbes, or, in the case of cell surface CD1 proteins, the outside of the cell. Thus, the total pool of antigens available for binding onto CD1 includes both endogenous self lipids that comprise the membranes or lumenal compartments of CD1‐expressing APCs and exogenous antigens taken up into endosomal compartments. Among these, certain types of lipid antigens appear to involve high stringency loading pathways in that they absolutely require cellular factors for loading onto CD1 proteins. These antigens include mycolic acids, lipoarabinomannan, glucose monomycolates with long alkyl chains (C80 GMM), and acylated sulfotrehaloses. The functional studies of the loading or recognition of these lipid antigens indicate that their loading takes place in acidic endosomes using mechanisms that involve low pH, receptor‐mediated delivery to endosomes, or lipid transfer systems involving saposins or apoE. Other lipid antigens readily bind to recombinant CD1 proteins at neutral pH in experimental systems using low stringency mechanisms. These antigens tend to be self lipids and have relatively short alkyl chains, such as phosphatidylinositol, phosphatidylcholine, phosphatidylethanolamine, ganglioside, sulfatide, and glucose monomycolate with short alkyl chains (C32 GMM) (Table 1). Many studies of the differing loading requirements of these antigens point to a general model in which high and low stringency loading interactions occur in distinct subcellular compartments. Under normal conditions, dynamically trafficking CD1 proteins are predicted to bind low stringency antigens in almost any lumenal compartment, including the ER, secretory pathway, cell

122

D. BRANCH MOODY

Table 1 High and Low Stringency Pathways for Presentation of Lipids to T Cells High stringency Location

Endosomes

Representative antigens

mycolic acid, C80 glucose monomycolate, lipoarabinomannan acyl sulfotrehaloses, endogenous antigens for NK T cells No

Presentation by recombinant CD1 proteins at neutral pH Inhibition by fixation of APC membranes Inhibition by neutralization of APC endosomes Inhibition by endosomal targeting motif deletion in CD1 Potency of presentation Half‐life after antigen pulse References

Low stringency ER, Golgi, cell surface and endosomes Sulfatide, ganglioside, phosphatidylinositol, phosphatidylcholine, phosphatidylethanolamine C32 glucose monomycolate Yes

Yes

No

Yes

No

Yes

No

High (nanomolar)

Low (micromolar)

Long

Short

(Chiu et al., 1999; Gilleron et al., 2004; Jackman et al., 1998; Moody et al., 2002; Porcelli et al., 1992; Sieling et al., 1995)

(Agea et al., 2005; Joyce et al., 1998; Shamshiev et al., 1999, 2000, 2002)

Antigens can bring CD1b and CD1d proteins via stringent mechanisms that require endosomal cofactors or low pH or bind via non‐stringent mechanisms that likely occur in many cellular compartments.

surface, and endosomes. Such non‐selective interactions may lead to the capture of antigens based on their abundance, which might account for the elution of phosphatidylinositol‐containing compounds from cellular CD1d proteins (Joyce et al., 1998). When CD1 proteins enter the specialized endosomal microenvironment, they encounter antigens concentrated in this location by selective capture pathways and are loaded under more stringent conditions involving lipid transfer proteins, pH‐mediated CD1 denaturation, and other factors that selectively promote lipid exchange. Despite delay and greater complexity of cellular interactions that lead to loading of high


123

stringency antigens, the studies summarized in Table 1 generally show that high stringency antigens have a greater potency for activating T cells (nanomolar) as compared with exogenously added low stringency antigens (micromolar). Further, some evidence suggests that high stringency antigens have longer half‐lives of binding to CD1 or activation of CD1‐restricted T cells (Moody et al., 2002). The general prediction of this model is that CD1 proteins bind and present many classes of self antigens. CD1 transit through endosomal compartments leads to acquisition of high affinity antigens, which skews the overall repertoire of subsequently displayed antigens toward those subclasses that enter the cells from the outside or have unusual chemical features (such as long chain length), which allow them to be retained within the groove. This may promote presentation of foreign antigens, if they are concentrated in endosomes by virtue of their entry into cells by phagocytosis or pattern recognition receptor‐mediated internalization and other mechanisms (Blander and Medzhitov, 2004; Hunger et al., 2004; Prigozy et al., 1997). Although this model is consistent with most published studies, certain missing information is the subject of ongoing study. Does CD1 randomly capture low stringency lipids, or are there dominant self antigens that serve a groove blocking function? Do lipid antigens generally undergo structural alterations in endosomes? Which endosomal cofactors are involved in the dynamic process of lipid insertion into the CD1 grooves? Also, it should be emphasized that the existence of high and low stringency loading pathways has been most clearly demonstrated for antigens presented by CD1b and CD1d, whereas much less information is available for CD1a and CD1c. 6.3. Autoimmune Recognition of Self Antigens The identification of mouse and human T cell clones that directly recognize CD1 proteins or ubiquitous self antigens raises the possibility that autoreactive T cells could mediate autoimmune injury in vivo. One model posits that activation of immunoregulatory NK T cells generally protects against autoimmunity and that loss or weakening of NK T cell immunoregulation leads to autoimmunity. This hypothesis is supported by several studies showing that patients with autoimmune disease have reduced total numbers of NK T cells or have NK T cells that are biased toward Th1 cytokine profiles (Sumida et al., 1995; van der Vliet et al., 2002; Wilson et al., 1998), although one study of human autoimmune diabetes patients has questioned these findings (Lee et al., 2002). Separately, deletion of CD1 or NK T cells generally worsens progression of autoimmune diabetes in animals (Duarte et al., 2004; Falcone et al., 2004; Naumov et al., 2001).

124

D. BRANCH MOODY

A second type of model involves the hypothesis that diverse CD1‐restricted T cells might directly recognize lipids expressed within the target cells and that recognition leads to local tissue destruction. For example, sulfatides, which are expressed at particularly high levels in myelin, can be the targets of variable CD1d‐restricted T cells in vivo, and treatment of mice with sulfatides ameliorates experimental allergic encephalomyelitis (Jahng et al., 2004). Similarly, diverse CD1‐restricted T cells recognizing gangliosides, which are likewise expressed at high levels in myelin, have been detected in human patients with multiple sclerosis (Shamshiev et al., 1999). Also, studies of human patients with Grave’s disease and Hashimoto’s thyroiditis have identified autoreactive T cells that recognize CD1a and CD1c proteins and home to the thyroid gland (Roura‐Mir et al., 2005a). Because diverse CD1‐restricted T cells lack surface markers or structurally invariant TCRs that would allow direct tracking in vivo or ex vivo, it is not yet known whether such cells are more numerous in autoimmune disease patients versus normal controls. However, these studies do provide proof of concept that CD1 and self lipid recognition can occur at the site of autoimmune tissue destruction. 6.4. CD1 and Allergic Disease There is new evidence that activation of NK T cells contributes to the development of allergic diseases. One of the striking features of the function of NK T cells is that they are able to secrete large amounts of IL‐4 in a primary stimulation, in contrast to MHC‐restricted T cells. This led to the hypothesis that NK T cells function to provide a source of IL‐4 at early time points in immune response and could consequently polarize MHC‐restricted T cells towards Th2 differentiation pathways or otherwise regulate Th2 immune responses. The first studies of CD1 knockout in mice found that IL‐4 production, IgE responses, and other surrogates of Th2 immunity were generally intact in CD1d‐deficient mice, suggesting that NK T cells were not absolutely required for these particular outcomes (Chen et al., 1997; Smiley et al., 1997). However, the ability of NK T cells to secrete IL‐4 in large amounts has never been questioned, and more detailed studies of CD1d knockout mice now suggest a role for NK T cells in mouse models of asthma and some forms of allergic hypersensitivity reactions that are mediated by Th2 cytokines. A role of CD1d and NK T cells in a mouse model of OVA‐induced airways hyperreactivity was clearly shown in a study in which deletion of either CD1d or Va14Ja18þ T cells resulted in marked reduction of inflammation and airway responsiveness. Further, it was shown that reconstitution with NK T cells reversed the effects, as long as the transferred NK T cells were capable of secretion of IL‐4 or IL‐13 (Akbari et al., 2003). A role for CD1d knockout in


125

worsening pulmonary airways disease has been confirmed in several other studies and has extended to models of ragweed‐induced pulmonary inflammation (Araujo et al., 2004; Bilenki et al., 2004; Lisbonne et al., 2003). Further, treatment of animals with a‐galactosyl ceramide results in increased levels of lung eosinophils, IgE, and IL‐4 (Kim et al., 2004). It is notable that effects of CD1d‐restricted NK T cells in animal models of airway hyperreactivity induced by pulmonary challenge with protein antigens are stronger than those seen with experiments carried out to assess IgE or other systemic measures of allergic inflammation, raising the possibility that NK T cells have particularly strong effects in the lung. In addition to these effects on murine NK T cells, pollen phospholipids that differ in structure from those found in mammalian cells can activate CD1‐restricted T cells in humans, and some evidence indicates that this effect correlates with seasonal allergy (Agea et al., 2005). 7. Synthetic Lipid Antigens and Prospects for Immunotherapy 7.1. T Cell Fine Specifity for Lipid Structure Based on emerging information about how the precise chemical features of antigens influence their recognition by T cells, it is now possible to design altered lipid ligands that activate T cells in ways that influence their cytokine profiles and other effector functions. Most structure‐function studies of lipid recognition have found that alteration of the carbohydrate or peptide moieties of antigens abrogate their ability to stimulate T cells. However, small changes in the overall chain length or saturation state of the alkyl units of CD1‐ presented antigens can preserve antigen recognition, while bypassing certain APC processing requirements or altering the potency or Th1–Th2 polarization properties of antigens. For example, mycobacterial glucose monomycolate antigens, normally produced with long (C80) alkyl chains, are presented by DCs after transport to endosomes. A version of glucose monomycolate that has the same carbohydrate structure, but a shorter C32 alkyl chain, is able to bypass the need for processing in endosomes so that it can be more efficiently presented by B cells that lack efficient mechanisms for uptake into endosomes (Moody et al., 2002). Similarly, several systems have shown that mono‐ or polyunsaturated fatty acids generally increase the potency of phosphatidylethanolamines, lipopeptides, and phosphatidylcholines for activating T cells (Agea et al., 2005; Moody et al., 2004; Rauch et al., 2003). It has been argued that the diunsaturated fatty acids that occur naturally in pollens are responsible for the increased responses compared to mammalian phosphatidylcholines that contain predominantly monounsaturated or saturated fatty acids.

126

D. BRANCH MOODY

7.2. Synthesis of Altered Lipid Ligands The ability of the fine structure of the lipid moieties of antigens to influence T cell responses in vitro has facilitated efforts to design altered a‐galactosyl ceramides that produce desired immunological outcomes in vivo. The immunological properties of a‐linked ceramide lipids were initially identified by screening large chemical libraries for the ability of compounds to promote tumor regression. This screen identified marine sponge antigens, which served as the basis for synthesis of an a‐galactosyl ceramide with a C26:0 fatty acyl unit and a C18:0 sphingosine base. This synthetic compound likewise activates NK T cells and is known by the proprietary name, KRN 7000 (Kawano et al., 1997). Production of analogs with shorter fatty acyl and sphingosine bases, C24 and C9, respectively, resulted in a compound called OCH, a form of a‐galactosyl ceramide that causes NK T cells to secrete higher levels of IL‐4 and protects against experimental allergic encephalomyelitis in mice (Miyamoto et al., 2001). This effect, which has also been seen in mouse models of autoimmune diabetes, arthritis, and colitis, is related to the shorter alkyl chains, which appear to less effectively anchor the lipid to CD1d, resulting in faster off rates and altered TCR signaling (Chiba et al., 2004; Mizuno et al., 2004; Oki et al., 2004; Ueno et al., 2005). A separate approach validates the concept of synthesizing a‐galactosyl ceramides that preferentially stimulate Th2 cytokine secretion, as synthesis of an a‐galactosyl ceramide analog containing a C20:2 fatty acyl chain likewise allowed potent induction of IL‐4 by NK T cells in vitro and in vivo, as compared to KRN 7000 (Yu et al., 2005). 8. Conclusion: Prospects for Immunotherapy The isolation of natural antigens and synthesis of altered lipid antigens now offers a new means for immunotherapy, whereby CD1‐presented lipids are given as immunomodulatory agents that activate, deactivate, or polarize the responses of CD1‐restricted T cells. As evidence accumulates that lipid ligands of CD1 proteins alter outcomes in animal models of autoimmunity, infections, and cancer, CD1‐presented lipids are now being used as immunomodulatory agents in clinical trials of human diseases (Nieda et al., 2004). Many glycolipid antigens, including a‐galactosyl ceramides, are orally bioavailable, so that they could be readily administered in conventional oral formulations (Silk et al., 2004). Certain aspects of the basic biology of lipid synthesis and presentation suggest that lipid‐based immunotherapy could improve or complement existing approaches to immunization. The success of antigen‐specific T cell modulation by MHC‐presented peptides has been limited due to the diversity


127

of human MHC protein structures, which requires design of peptides, which promiscuously bind to many MHC proteins. The lack of common or functionally important CD1 polymorphisms raises the prospect that lipid antigens would be presented similarly by individuals within genetically diverse human populations. Related to this, it has been argued that the low levels of allelic polymorphism in CD1 proteins results from the lack of rapid change in lipid structure over evolutionary time. Peptide antigens can escape detection by T cells with a single point mutation, which often disrupts TCR contact but does not have an appreciable affect on the overall function of the larger protein. In contrast, lipids and glycolipids are made by biosynthetic pathways involving multiple enzymes. Although changes in glycolipid structure can occur through genetic deletion of enzyme function, such changes usually alter the glycolipid in a more general way so that its overall function is lost. There is increasing evidence that CD1‐restricted T cells target mycobacterial lipids that have non‐redundant functions in growth or function as pathogenicity factors (Gilleron et al., 2004; Van Rhijn et al., 2005; Matsunaga et al., 2004). These observations are consistent with the hypothesis that pathogens cannot as readily form functionally intact escape mutants because the lipid antigens contribute directly to the ability to infect and persist in the host. Two types of lipid‐based immunotherapeutic strategies, which parallel the separate immunoregulatory and effector functions of diverse and invariant CD1‐restricted T cells (Fig. 1), are currently being developed. One strategy involves injection of a‐galactosyl ceramide or other agonists of NK T cells to promote cytokine release, leading to dendritic cell maturation and altered MHC‐restricted T cell activation. In this approach, glycolipids are considered adjuvants because the main effect of NK T cells involves activation of antigen‐ presenting function of DCs in mice (Fujii et al., 2003) and humans (Nieda et al., 2004). Concomitant treatment with conventional peptide antigens can focus the subsequent MHC‐restricted T cell response on a desired target tissue. For example, co‐administration of a‐galactosyl ceramide with HLA‐A2‐presented peptide antigens markedly increases the priming of CD8þ peptide‐specific T cells and promotes regression of melanoma tumors (Silk et al., 2004). A variant of this strategy has been used in animal models of Th1‐mediated autoimmune diseases. Altered lipid ligands increase the ratio of Th2/Th1 cytokines produced by NK T cells, which can downmodulate MHC‐restricted T cell function (Miyamoto et al., 2001; Van Kaer, 2005; Yu et al., 2005). A second strategy, which is just beginning to be developed, involves stimulating diverse CD1‐restricted T cells, which may directly carry out effector functions against those cells that express lipids that are selectively made by pathogens, tumors, or within tissues with specialized functions. This approach does not involve MHC‐restricted T cells and relies on the premise that human

128

D. BRANCH MOODY

T cells can be activated to recognize lipid antigens with highly selective patterns of expression on a limited number of target cells. Some candidates already exist, as mycolates, sulfotrehaloses, and phosphomycoketides are found only in pathogens, and certain sphingolipids are somewhat selectively expressed in myelin (Fig. 5). Despite decades of work on tumor‐associated glycolipid targets of antibody responses, well‐defined tumor‐associated glycolipids presented by the CD1 system have not yet been extensively investigated. Because many of these antigens for diverse CD1‐restricted T cells are presented by inducibly expressed group 1 CD1 proteins, immunization strategies will likely involve glycolipid adjuvants to increase group 1 CD1 expression and function on myeloid DCs. These new therapeutic strategies derive from a simple observation, made about a decade ago, that human T cells are activated by both lipid and protein components of target cells. Acknowledgments The author thanks C. Anthony DeBono, Carme Roura‐Mir, Christopher Dascher, and Dirk Zajonc for analysis of CD1 sequence, design of CD1 schematics, and helpful discussions. This work is supported by the Pew Foundation Scholars in the Biomedical Sciences, the Cancer Research Institute, the Mizutani Foundation for Glycoscience, and the NIH (AI049313 and AR48632).

References Agea, E., Russano, A., Bistoni, O., Mannucci, R., Nicoletti, I., Corazzi, L., Postle, A. D., De Libero, G., Porcelli, S. A., and Spinozzi, F. (2005). Human CD1‐restricted T cell recognition of lipids from pollens. J. Exp. Med. 202, 295–308. Akbari, O., Stock, P., Meyer, E., Kronenberg, M., Sidobre, S., Nakayama, T., Taniguchi, M., Grusby, M. J., DeKruyff, R. H., and Umetsu, D. T. (2003). Essential role of NKT cells producing IL‐4 and IL‐13 in the development of allergen‐induced airway hyperreactivity. Nat. Med. 9, 582–588. Amprey, J. L., Im, J. S., Turco, S. J., Murray, H. W., Illarionov, P. A., Besra, G. S., Porcelli, S. A., and Spath, G. F. (2004). A subset of liver NK T cells is activated during Leishmania donovani infection by CD1d‐bound lipophosphoglycan. J. Exp. Med. 200, 895–904. Angenieux, C., Fraisier, V., Maitre, B., Racine, V., Van Der, W. N., Fricker, D., Proamer, F., Sachse, M., Cazenave, J. P., Peters, P., Goud, B., Hanau, D., Sibarita, J. B., Salamero, J., and de la Salle, H. (2005). The cellular pathway of CD1e in immature and maturing dendritic cells. Traffic 6, 286–302. Angenieux, C., Salamero, J., Fricker, D., Cazenave, J. P., Goud, B., Hanau, D., and de la Salle, H. (2000). Characterization of CD1e, a third type of CD1 molecule expressed in dendritic cells. J. Biol. Chem. 275, 37757–37764. Araujo, L. M., Lefort, J., Nahori, M. A., Diem, S., Zhu, R., Dy, M., Leite‐de‐Moraes, M. C., Bach, J. F., Vargaftig, B. B., and Herbelin, A. (2004). Exacerbated Th2‐mediated airway inflammation and hyperresponsiveness in autoimmune diabetes‐prone NOD mice: A critical role for CD1d‐ dependent NKT cells. Eur. J. Immunol. 34, 327–335.


129

Ashkar, A. A., and Rosenthal, K. L. (2003). Interleukin‐15 and natural killer and NKT cells play a critical role in innate protection against genital herpes simplex virus type 2 infection. J. Virol. 77, 10168–10171. Baron, J. L., Gardiner, L., Nishimura, S., Shinkai, K., Locksley, R., and Ganem, D. (2002). Activation of a nonclassical NKT cell subset in a transgenic mouse model of hepatitis B virus infection. Immunity 16, 583–594. Batuwangala, T., Shepard, D., Gadola, S. D., Gibson, K. J. C., Zaccai, N. R., Besra, G. S., Cerundolo, V., and Jones, E. Y. (2003). The crystal structure of human CD1b with a bound bacterial glycolipid. J. Immunol. 238, 2–2388. Bauer, A., Huttinger, R., Staffler, G., Hansmann, C., Schmidt, W., Majdic, O., Knapp, W., and Stockinger, H. (1997). Analysis of the requirement for beta 2‐microglobulin for expression and formation of human CD1 antigens. Eur. J. Immunol. 27, 1366–1373. Beckman, E. M., Porcelli, S. A., Morita, C. T., Behar, S. M., Furlong, S. T., and Brenner, M. B. (1994). Recognition of a lipid antigen by CD1‐restricted ab T cells. Nature 372, 691–694. Behar, S. M., and Cardell, S. (2000). Diverse CD1d‐restricted T cells: Diverse phenotypes, and diverse functions. Semin. Immunol. 12, 551–560. Belperron, A. A., Dailey, C. M., and Bockenstedt, L. K. (2005). Infection‐induced marginal zone B cell production of Borrelia hermsii‐specific antibody is impaired in the absence of CD1d. J. Immunol. 174, 5681–5686. Bendelac, A. (1995). Positive selection of mouse NK1þ T cells by CD1‐expressing cortical thymocytes. J. Exp. Med. 182, 2091–2096. 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. Benlagha, K., and Bendelac, A. (2000). CD1d‐restricted mouse V alpha 14 and human V alpha 24 T cells: Lymphocytes of innate immunity. Semin. Immunol. 12, 537–542. 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. Benlagha, K., Weiss, A., Beavis, A., Teyton, L., and Bendelac, A. (2000). In vivo identification of glycolipid antigen‐specific T cells using fluorescent CD1d tetramers. J. Exp. Med. 191, 1895–1903. Bilenki, L., Yang, J., Fan, Y., Wang, S., and Yang, X. (2004). Natural killer T cells contribute to airway eosinophilic inflammation induced by ragweed through enhanced IL‐4 and eotaxin production. Eur. J. Immunol. 34, 345–354. Blander, J. M., and Medzhitov, R. (2004). Regulation of phagosome maturation by signals from toll‐ like receptors. Science 304, 1014–1018. Bonifacino, J. S., and Traub, L. M. (2003). Signals for sorting of transmembrane proteins to endosomes and lysosomes. Annu. Rev. Biochem. 72, 395–447. Brigl, M., Bry, L., Kent, S. C., Gumperz, J. E., and Brenner, M. B. (2003). Mechanism of CD1d‐ restricted natural killer T cell activation during microbial infection. Nat. Immunol. 4, 1230–1237. Briken, V., Jackman, R. M., Dasgupta, S., Hoening, S., and Porcelli, S. A. (2002). Intracellular trafficking pathway of newly synthesized CD1b molecules. EMBO J. 21, 825–834. Brozovic, S., Nagaishi, T., Yoshida, M., Betz, S., Salas, A., Chen, D., Kaser, A., Glickman, J., Kuo, T., Little, A., Morrison, J., Corazza, N., Kim, J. Y., Colgan, S. P., Young, S. G., Exley, M., and Blumberg, R. S. (2004). CD1d function is regulated by microsomal triglyceride transfer protein. Nat. Med. 10, 535–539. Budd, R. C., Miescher, G. C., Howe, R. C., Lees, R. K., Bron, C., and MacDonald, H. R. (1987). Developmentally regulated expression of T cell receptor beta chain variable domains in immature thymocytes. J. Exp. Med. 166, 577–582.

130

D. BRANCH MOODY

Buettner, M., Meinken, C., Bastian, M., Bhat, R., Stossel, E., Faller, G., Cianciolo, G., Ficker, J., Wagner, M., Rollinghoff, M., and Stenger, S. (2005). Inverse correlation of maturity and antibacterial activity in human dendritic cells. J. Immunol. 174, 4203–4209. Calabi, F., Jarvis, J. M., Martin, L., and Milstein, C. (1989). Two classes of CD1 genes. Eur. J. Immunol. 19, 285–292. Calabi, F., and Milstein, C. (1986). A novel family of human major histocompatibility complex‐ related genes not mapping to chromosome 6. Nature 323, 540–543. Chen, Y. H., Chiu, N. M., Mandal, M., Wang, N., and Wang, C. R. (1997). Impaired NK1þ T cell development and early IL‐4 production in CD1‐deficient mice. Immunity 6, 459–467. Chiba, A., Oki, S., Miyamoto, K., Hashimoto, H., Yamamura, T., and Miyake, S. (2004). Suppression of collagen‐induced arthritis by natural killer T cell activation with OCH, a sphingosine‐ truncated analog of alpha‐galactosylceramide. Arthritis Rheum. 50, 305–313. Chiu, Y. H., Jayawardena, J., Weiss, A., Lee, D., Park, S. H., Dautry‐Varsat, A., and Bendelac, A. (1999). Distinct subsets of CD1d‐restricted T cells recognize self‐antigens loaded in different cellular compartments. J. Exp. Med. 189, 103–110. Chiu, Y. H., Park, S. H., Benlagha, K., Forestier, C., Jayawardena‐Wolf, J., Savage, P. B., Teyton, L., and Bendelac, A. (2002). Multiple defects in antigen presentation and T cell development by mice expressing cytoplasmic tail‐truncated CD1d. Nature Immunol. 3, 55–60. Cho, S., Knox, K. S., Kohli, L. M., He, J. J., Exley, M. A., Wilson, S. B., and Brutkiewicz, R. R. (2005). Impaired cell surface expression of human CD1d by the formation of an HIV‐1 Nef/ CD1d complex. Virology 337, 242–252. 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. Cotner, T., Mashimo, H., Kung, P. C., Goldstein, G., and Strominger, J. L. (1981). Human T cell surface antigens bearing a structural relationship to HLA antigens. Proc. Natl. Acad. Sci. USA 78, 3858–3862. Cresswell, P. (1994). Assembly, transport, and function of MHC class II molecules. [Review] [151 refs]. Ann. Rev. Immunol. 12, 259–293. Dascher, C. C., and Brenner, M. B. (2003). Evolutionary constraints on CD1 structure: Insights from comparative genomic analysis. Trends Immunol. 24, 412–418. Dascher, C. C., Hiromatsu, K., Naylor, J. W., Brauer, P. P., Brown, K. A., Storey, J. R., Behar, S. M., Kawasaki, E. S., Porcelli, S. A., Brenner, M. B., and LeClair, K. P. (1999). Conservation of a CD1 multigene family in the guinea pig. J. Immunol. 163, 5478–5488. Dascher, C. C., Hiromatsu, K., Xiong, X., Morehouse, C., Watts, G., Liu, G., McMurray, D. N., LeClair, K. P., Porcelli, S. A., and Brenner, M. B. (2003). Immunization with a mycobacterial lipid vaccine improves pulmonary pathology in the guinea pig model of tuberculosis. Int. Immunol. 15, 915–925. de la Salle, H., Mariotti, S., Angenieux, C., Gilleron, M., Garcia‐Alles, L. F., Malm, D., Berg, T., Paoletti, S., Maitre, B., Mourey, L., Salamero, J., Cazenave, J. P., Hanau, D., Mori, L., Puzo, G., and De Libero, G. (2005). Assistance of microbial glycolipid antigen processing by CD1e. Science 310, 1321–1324. De Libero, G., Moran, A. P., Gober, H. J., Rossy, E., Shamshiev, A., Chelnokova, O., Mazorra, Z., Vendetti, S., Sacchi, A., Prendergast, M. M., Sansano, S., Tonevitsky, A., Landmann, R., and Mori, L. (2005). Bacterial infections promote T cell recognition of self‐glycolipids. Immunity 22, 763–772. De Silva, A. D., Park, J. J., Matsuki, N., Stanic, A. K., Brutkiewicz, R. R., Medof, M. E., and Joyce, S. (2002). Lipid protein interactions: The assembly of CD1d1 with cellular phospholipids occurs in the endoplasmic reticulum. J. Immunol. 168, 723–733.


131

De Voss, J., Rutter, K., Schroeder, B. G., Su, H., Zhu, Y., and Barry, C. E. (2000). The salicylate‐ derived mycobactin siderophores of Mycobacterium tuberculosis are essential for growth in macrophages. Proc. Natl. Acad. Sci. USA 97, 1252–1257. Domenech, P., Reed, M. B., Dowd, C. S., Manca, C., Kaplan, G., and Barry, C. E., III (2004). The role of MmpL8 in sulfatide biogenesis and virulence of Mycobacterium tuberculosis. J. Biol. Chem. 279, 21257–21265. Dougan, S. K., Salas, A., Rava, P., Agyemang, A., Kaser, A., Morrison, J., Khurana, A., Kronenberg, M., Johnson, C., Exley, M., Hussain, M. M., and Blumberg, R. S. (2005). Microsomal triglyceride transfer protein lipidation and control of CD1d on antigen‐presenting cells. J. Exp. Med. 202, 529–539. Duarte, N., Stenstrom, M., Campino, S., Bergman, M. L., Lundholm, M., Holmberg, D., and Cardell, S. L. (2004). Prevention of diabetes in nonobese diabetic mice mediated by CD1d‐ restricted nonclassical NKT cells. J. Immunol. 173, 3112–3118. Elewaut, D., Lawton, A. P., Nagarajan, N. A., Maverakis, E., Khurana, A., Honing, S., Benedict, C. A., Sercarz, E., Bakke, O., Kronenberg, M., and Prigozy, T. I. (2003). The adaptor protein AP‐3 is required for CD1d‐mediated antigen presentation of glycosphingolipids and development of Valpha14i NKT cells. J. Exp. Med. 1133–1146. Ernst, W. A., Maher, J., Cho, S., Niazi, K. R., Chatterjee, D., Moody, DB, Besra, G. S., Watanabe, Y., Jensen, P. E., Porcelli, S. A., Kronenberg, M., and Modlin, R. L. (1998). Molecular interaction of CD1b with lipoglycan antigens. Immunity 8, 331–340. Exley, M., Porcelli, S., Furman, M., Garcia, J., and Balk, S. (1998). CD161 (NKR‐P1A) costimulation of CD1d‐dependent activation of human T cells expressing invariant V alpha 24 J alpha Q T cell receptor alpha chains. J. Exp. Med. 188, 867–876. Falcone, M., Facciotti, F., Ghidoli, N., Monti, P., Olivieri, S., Zaccagnino, L., Bonifacio, E., Casorati, G., Sanvito, F., and Sarvetnick, N. (2004). Up‐regulation of CD1d expression restores the immunoregulatory function of NKT cells and prevents autoimmune diabetes in nonobese diabetic mice. J. Immunol. 172, 5908–5916. Fischer, K., Scotet, E., Niemeyer, M., Koebernick, H., Zerrahn, J., Maillet, S., Hurwitz, R., Kursar, M., Bonneville, M., Kaufmann, S. H., and Schaible, U. E. (2004). Mycobacterial phosphatidylinositol mannoside is a natural antigen for CD1d‐restricted T cells. Proc. Natl. Acad. Sci. USA 101, 10685–10690. Fowlkes, B. J., Kruisbeek, A. M., Ton‐That, H., Weston, M. A., Coligan, J. E., Schwartz, R. H., and Pardoll, D. M. (1987). A novel population of T‐cell receptor alpha beta‐bearing thymocytes which predominantly expresses a single V beta gene family. Nature 329, 251–254. Fujii, S., Shimizu, K., Smith, C., Bonifaz, L., and Steinman, R. M. (2003). Activation of natural killer T cells by alpha‐galactosylceramide rapidly induces the full maturation of dendritic cells in vivo and thereby acts as an adjuvant for combined CD4 and CD8 T cell immunity to a coadministered protein. J. Exp. Med. 198, 267–279. Gadola, S. D., Zaccai, N. R., Harlos, K., Shepherd, D., Castro‐Palomino, J. C., Ritter, G., Schmidt, R. R., Jones, E. Y., and Cerundolo, V. (2002). Structure of human CD1b with bound ligands at 2.3 A, a maze for alkyl chains. Nature Immunol. 3, 721–726. Garboczi, D. N., Ghosh, P., Utz, U., Fan, Q. R., Biddison, W. E., and Wiley, D. C. (1996). Structure of the complex between human T‐cell receptor, viral peptide and HLA‐A2. Nature 384, 134–141. Garcia, K. C., Degano, M., Stanfield, R. L., Brunmark, A., Jackson, M. R., Peterson, P. A., Teyton, L., and Wilson, I. A. (1996). An alphabeta T cell receptor structure at 2.5 A and its orientation in the TCR‐MHC complex. Science 274, 209–219. Geng, Y., Laslo, P., Barton, K., and Wang, C. R. (2005). Transcriptional Regulation of CD1D1 by Ets Family Transcription Factors. J. Immunol. 175, 1022–1029.

132

D. BRANCH MOODY

Giabbai, B., Sidobre, S., Crispin, M. D., Sanchez‐Ruiz, Y., Bachi, A., Kronenberg, M., Wilson, I. A., and Degano, M. (2005). Crystal structure of mouse CD1d bound to the self ligand phosphatidylcholine: A molecular basis for NKT cell activation. J. Immunol. 175, 977–984. Gilleron, M., Stenger, S., Mazorra, Z., Wittke, F., Mariotti, S., Bohmer, G., Prandi, J., Mori, L., Puzo, G., and De Libero, G. (2004). Diacylated Sulfoglycolipids Are Novel Mycobacterial Antigens Stimulating CD1‐restricted T Cells during Infection with Mycobacterium tuberculosis. J. Exp. Med. 199, 649–659. Giuliani, A., Prete, S. P., Graziani, G., Aquino, A., Balduzzi, A., Sugita, M., Brenner, M. B., Iona, E., Fattorini, L., Orefici, G., Porcelli, S. A., and Bonmassar, E. (2001). Influence of Mycobacterium bovis bacillus Calmette Guerin on in vitro induction of CD1 molecules in human adherent mononuclear cells. Infect. Immun. 69, 7461–7470. Glass, W. G., Subbarao, K., Murphy, B., and Murphy, P. M. (2004). Mechanisms of host defense following severe acute respiratory syndrome‐coronavirus (SARS‐CoV) pulmonary infection of mice. J. Immunol. 173, 4030–4039. Grant, E. P., Beckman, E. M., Behar, S. M., Degano, M., Frederique, D., Besra, G. S., Wilson, I. A., Porcelli, S. A., Furlong, S. T., and Brenner, M. B. (2002). Fine specificity of TCR complementarity‐determining region residues and lipid antigen hydrophilic moieties in the recognition of a CD1‐lipid complex. J. Immunol. 168, 3933–3940. Grubor‐Bauk, B., Simmons, A., Mayrhofer, G., and Speck, P. G. (2003). Impaired clearance of herpes simplex virus type 1 from mice lacking CD1d or NKT cells expressing the semivariant V alpha 14‐J alpha 281 TCR. J. Immunol. 170, 1430–1434. Gumperz, J., Roy, C., Makowska, A., Lum, D., Sugita, M., Podrebarac, T., Koezuka, Y., Porcelli, S. A., Cardell, S., Brenner, M. B., and Behar, S. M. (2000). Murine CD1d‐Restricted T Cell Recognition of Cellular Lipids. Immunity 12, 211–221. Gumperz, J. E., Miyake, S., Yamamura, T., and Brenner, M. B. (2002). Functionally distinct subsets of CD1d‐restricted natural killer T cells revealed by CD1d tetramer staining. J. Exp. Med. 195, 625–636. Hansen, D. S., Siomos, M. A., Buckingham, L., Scalzo, A. A., and Schofield, L. (2003a). Regulation of murine cerebral malaria pathogenesis by CD1d‐restricted NKT cells and the natural killer complex. Immunity 18, 391–402. Hansen, D. S., Siomos, M. A., Koning‐Ward, T., Buckingham, L., Crabb, B. S., and Schofield, L. (2003b). CD1d‐restricted NKT cells contribute to malarial splenomegaly and enhance parasite‐ specific antibody responses. Eur. J. Immunol. 33, 2588–2598. Hayes, S. M., and Knight, K. L. (2001). Group 1 CD1 genes in rabbit. J. Immunol. 166, 403–410. Henderson, R. A., Watkins, S. C., and Flynn, J. L. (1997). Activation of human dendritic cells following infection with Mycobacterium tuberculosis. J. Immunol. 159, 635–643. Huber, S., Sartini, D., and Exley, M. (2003). Role of CD1d in coxsackievirus B3‐induced myocarditis. J. Immunol. 170, 3147–3153. Hughes, A. L. (1991). Evolutionary origin and diversification of the mammalian CD1 antigen genes. Mol. Biol. Evol. 8, 185–201. Hunger, R. E., Sieling, P. A., Ochoa, M. T., Sugaya, M., Burdick, A. E., Rea, T. H., Brennan, P. J., Belisle, J. T., Blauvelt, A., Porcelli, S. A., and Modlin, R. L. (2004). Langerhans cells utilize CD1a and langerin to efficiently present nonpeptide antigens to T cells. J. Clin. Invest 113, 701–708. Huttinger, R., Staffler, G., Majdic, O., and Stockinger, H. (1999). Analysis of the early biogenesis of CD1b: Involvement of the chaperones calnexin and calreticulin, the proteasome and b2‐microglobulin. Int. Immunol. 11, 1615–1623. Jackman, R. M., Stenger, S., Lee, A., Moody, D. B., Rogers, R. A., Niazi, K. R., Sugita, M., Modlin, R. L., Peters, P. J., and Porcelli, S. A. (1998). The tyrosine‐containing cytoplasmic tail of CD1b is essential for its efficient presentation of bacterial lipid antigens. Immunity 8, 341–351.


133

Jahng, A., Maricic, I., Aguilera, C., Cardell, S., Halder, R. C., and Kumar, V. (2004). Prevention of autoimmunity by targeting a distinct, noninvariant CD1d‐reactive T cell population reactive to sulfatide. J. Exp. Med. 199, 947–957. Jayawardena‐Wolf, J., Benlagha, K., Chiu, Y. H., Mehr, R., and Bendelac, A. (2001). CD1d endosomal trafficking is independently regulated by an intrinsic CD1d‐encoded tyrosine motif and by the invariant chain. Immunity 15, 897–908. Joyce, S., Woods, A. S., Yewdell, J. W., Bennink, J. R., De, S. A., Boesteanu, A., Balk, S. P., Cotter, R. J., and Brutkiewicz, R. R. (1998). Natural ligand of mouse CD1d1: Cellular glycosylphosphatidylinositol. Science 279, 1541–1544. Kang, S. J., and Cresswell, P. (2002a). Calnexin, calreticulin, and ERp57 cooperate in disulfide bond formation in human CD1d heavy chain. J. Biol. Chem. 277, 44838–44844. Kang, S. J., and Cresswell, P. (2002b). Regulation of intracellular trafficking of human CD1d by association with MHC class II molecules. EMBO J. 21, 1650–1660. Kang, S. J., and Cresswell, P. (2004). Saposins facilitate CD1d‐restricted presentation of an exogenous lipid antigen to T cells. Nat. Immunol. 5, 175–181. Kappler, J., Kubo, R., Haskins, K., White, J., and Marrack, P. (1983). The mouse T cell receptor: Comparison of MHC‐restricted receptors on two T cell hybridomas. Cell 34, 727–737. Karadimitris, A., Gadola, S., Altamirano, M., Brown, D., Woolfson, A., Klenerman, P., Chen, J. L., Koezuka, Y., Roberts, I. A., Price, D. A., Dusheiko, G., Milstein, C., Fersht, A., Luzzatto, L., and Cerundolo, V. (2001). Human CD1d‐glycolipid tetramers generated by in vitro oxidative refolding chromatography. Proc. Natl. Acad. Sci. USA 98, 3294–3298. Kawano, T., Cui, J., Koezuka, Y., Toura, I., Kaneko, Y., Motoki, K., Ueno, H., Nakagawa, R., Sato, H., Kondo, E., Koseki, H., and Taniguchi, M. (1997). CD1d‐restricted and TCR‐mediated activation of Va14 NKT cells by glycosylceramides. Science 278, 1626–1629. Kim, J. O., Kim, D. H., Chang, W. S., Hong, C., Park, S. H., Kim, S., and Kang, C. Y. (2004). Asthma is induced by intranasal coadministration of allergen and natural killer T‐cell ligand in a mouse model. J. Allergy Clin. Immunol. 114, 1332–1338. Kinjo, Y., Wu, D., Kim, G., Xing, G. W., Poles, M. A., Ho, D. D., Tsuji, M., Kawahara, K., Wong, C. H., and Kronenberg, M. (2005). Recognition of bacterial glycosphingolipids by natural killer T cells. Nature 434, 520–525. Koch, M., Stronge, V. S., Shepherd, D., Gadola, S. D., Mathew, B., Ritter, G., Fersht, A. R., Besra, G. S., Schmidt, R. R., Jones, E. Y., and Cerundolo, V. (2005). The crystal structure of human CD1d with and without alpha‐galactosylceramide. Nat. Immunol. 6, 819–826. Krutzik, S. R., Tan, B., Li, H., Ochoa, M. T., Liu, P. T., Sharfstein, S. E., Graeber, T. G., Sieling, P. A., Liu, Y. J., Rea, T. H., Bloom, B. R., and Modlin, R. L. (2005). TLR activation triggers the rapid differentiation of monocytes into macrophages and dendritic cells. Nat. Med. 11, 653–660. Kumar, H., Belperron, A., Barthold, S. W., and Bockenstedt, L. K. (2000). Cutting edge: CD1d deficiency impairs murine host defense against the spirochete, Borrelia burgdorferi. J. Immunol. 165, 4797–4801. Laird, D. J., De Tomaso, A. W., Cooper, M. D., and Weissman, I. L. (2000). 50 million years of chordate evolution: Seeking the origins of adaptive immunity. Proc. Natl. Acad. Sci. USA 97, 6924–6926. Lang, G. A., Illarionov, P. A., Glatman‐Freedman, A., Besra, G. S., and Lang, M. L. (2005). BCR targeting of biotin‐{alpha}‐galactosylceramide leads to enhanced presentation on CD1d and requires transport of BCR to CD1d‐containing endocytic compartments. Int. Immunol. 17, 899–908. Le Gall, S., Erdtmann, L., Benichou, S., Berlioz‐Torrent, C., Liu, L., Benarous, R., Heard, J. M., and Schwartz, O. (1998). Nef interacts with the mu subunit of clathrin adaptor complexes and reveals a cryptic sorting signal in MHC I molecules. Immunity 8, 483–495.

134

D. BRANCH MOODY

Lee, P. T., Putnam, A., Benlagha, K., Teyton, L., Gottlieb, P. A., and Bendelac, A. (2002). Testing the NKT cell hypothesis of human IDDM pathogenesis. J. Clin. Invest. 110, 793–800. Levy, O., Orange, J. S., Hibberd, P., Steinberg, S., LaRussa, P., Weinberg, A., Wilson, S. B., Shaulov, A., Fleisher, G., Geha, R. S., Bonilla, F. A., and Exley, M. (2003). Disseminated varicella infection due to the vaccine strain of varicella‐zoster virus, in a patient with a novel deficiency in natural killer T cells. J. Infect. Dis. 188, 948–953. Lisbonne, M., Diem, S., de Castro, K. A., Lefort, J., Araujo, L. M., Hachem, P., Fourneau, J. M., Sidobre, S., Kronenberg, M., Taniguchi, M., Van Endert, P., Dy, M., Askenase, P., Russo, M., Vargaftig, B. B., Herbelin, A., and Leite‐de‐Moraes, M. C. (2003). Cutting edge: Invariant V alpha 14 NKT cells are required for allergen‐induced airway inflammation and hyperreactivity in an experimental asthma model. J. Immunol. 171, 1637–1641. Matsuda, J. L., Gapin, L., Sidobre, S., Kieper, W. C., Tan, J. T., Ceredig, R., Surh, C. D., and Kronenberg, M. (2002). Homeostasis of V alpha 14i NKT cells. Nat. Immunol. 3, 966–974. Matsuda, J. L., Naidenko, O. V., Gapin, L., Nakayama, T., Taniguchi, M., Wang, C. R., Koezuka, Y., and Kronenberg, M. (2000). Tracking the response of natural killer T cells to a glycolipid antigen using CD1d tetramers. J. Exp. Med. 192, 741–754. Matsunaga, I., Bhatt, A., Young, D. C., Cheng, T. Y., Eyles, S. J., Besra, G. S., Briken, V., Porcelli, S. A., Costello, C. E., Jacobs, W. R., Jr., and Moody, D. B. (2004). Mycobacterium tuberculosis pks12 Produces a Novel Polyketide Presented by CD1c to T Cells. J. Exp. Med. 200, 1559–1569. Mattner, J., Debord, K. L., Ismail, N., Goff, R. D., Cantu, C., III, Zhou, D., Saint‐Mezard, P., Wang, V., Gao, Y., Yin, N., Hoebe, K., Schneewind, O., Walker, D., Beutler, B., Teyton, L., Savage, P. B., and Bendelac, A. (2005). Exogenous and endogenous glycolipid antigens activate NKT cells during microbial infections. Nature 434, 525–529. McDevitt, H. O., and Benacerraf, B. (1969). Genetic control of specific immune responses. Adv. Immunol. 11, 31–74. McIntyre, B. W., and Allison, J. P. (1983). The mouse T cell receptor: Structural heterogeneity of molecules of normal T cells defined by xenoantiserum. Cell 34, 739–746. McMichael, A. J., Pilch, J. R., Galfre, G., Mason, D. Y., Fabre, J. W., and Milstein, C. (1979). A human thymocyte antigen defined by a hybrid myeloma monoclonal antibody. Eur. J. Immunol. 9, 205–210. Miller, M. M., Wang, C., Parisini, E., Coletta, R. D., Goto, R. M., Lee, S. Y., Barral, D. C., Townes, M., Roura‐Mir, C., Ford, H. L., Brenner, M. B., and Dascher, C. C. (2005). Characterization of two avian MHC‐like genes reveals an ancient origin of the CD1 family. Proc. Natl. Acad. Sci. USA 102, 8674–8679. Miyamoto, K., Miyake, S., and Yamamura, T. (2001). A synthetic glycolipid prevents autoimmune encephalomyelitis by inducing TH2 bias of natural killer T cells. Nature 413, 531–534. Mizuno, M., Masumura, M., Tomi, C., Chiba, A., Oki, S., Yamamura, T., and Miyake, S. (2004). Synthetic glycolipid OCH prevents insulitis and diabetes in NOD mice. J. Autoimmun. 23, 293–300. Molano, A., Park, S. H., Chiu, Y. H., Nosseir, S., Bendelac, A., and Tsuji, M. (2000). Cutting edge: The IgG response to the circumsporozoite protein is MHC class II‐dependent and CD1d‐ independent: Exploring the role of GPIs in NK T cell activation and antimalarial responses. J. Immunol. 164, 5005–5009. Moody, D. B., Briken, V., Cheng, T. Y., Roura‐Mir, C., Guy, M. R., Geho, D. H., Tykocinski, M. L., Besra, G. S., and Porcelli, S. A. (2002). Lipid length controls antigen entry into endosomal and nonendosomal pathways for CD1b presentation. Nature Immunol. 3, 435–442. Moody, D. B., Reinhold, B. B., Guy, M. R., Beckman, E. M., Frederique, D. E., Furlong, S. T., Ye, S., Reinhold, V. N., Sieling, P. A., Modlin, R. L., Besra, G. S., and Porcelli, S. A. (1997). Structural requirements for glycolipid antigen recognition by CD1b‐restricted T cells. Science 278, 283–286.


135

Moody, D. B., Reinhold, B. B., Reinhold, V. N., Besra, G. S., and Porcelli, S. A. (1999). Uptake and processing of glycosylated mycolates for presentation to CD1b‐restricted T cells. Immunol. Lett. 65, 85–91. Moody, D. B., Ulrichs, T., Muhlecker, W., Young, D. C., Gurcha, S. S., Grant, E., Rosat, J. P., Brenner, M. B., Costello, C. E., Besra, G. S., and Porcelli, S. A. (2000). CD1c‐mediated T‐cell recognition of isoprenoid glycolipids in Mycobacterium tuberculosis infection. Nature 404, 884–888. Moody, D. B., Young, D. C., Cheng, T. Y., Rosat, J. P., Roura‐Mir, C., O’Connor, P. B., Zajonc, D. M., Walz, A., Miller, M. J., Levery, S. B., Wilson, I. A., Costello, C. E., and Brenner, M. B. (2004). T cell activation by lipopeptide antigens. Science 303, 527–531. Moody, D. B., Zajonc, D. M., and Wilson, I. A. (2005). Anatomy of CD1‐lipid antigen complexes. Nat. Rev. Immunol. 5, 387–399. Naumov, Y. N., Bahjat, K. S., Gausling, R., Abraham, R., Exley, M. A., Koezuka, Y., Balk, S. B., Strominger, J. L., Clare‐Salzer, M., and Wilson, S. B. (2001). Activation of CD1d‐restricted T cells protects NOD mice from developing diabetes by regulating dendritic cell subsets. Proc. Natl. Acad. Sci. USA 98, 13838–13843. Nieda, M., Okai, M., Tazbirkova, A., Lin, H., Yamaura, A., Ide, K., Abraham, R., Juji, T., Macfarlane, D. J., and Nicol, A. J. (2004). Therapeutic activation of Valpha24 þ Vbeta11 þ NKT cells in human subjects results in highly coordinated secondary activation of acquired and innate immunity. Blood 103, 383–389. Ohno, H., Stewart, J., Fournier, M. C., Bosshart, H., Rhee, I., Miyatake, S., Saito, T., Gallusser, A., Kirchhausen, T., and Bonifacino, J. S. (1995). Interaction of tyrosine‐based sorting signals with clathrin‐associated proteins. Science 269, 1872–1875. Oki, S., Chiba, A., Yamamura, T., and Miyake, S. (2004). The clinical implication and molecular mechanism of preferential IL‐4 production by modified glycolipid‐stimulated NKT cells. J. Clin. Invest 113, 1631–1640. Oldstone, M. B., Whitton, J. L., Lewicki, H., and Tishon, A. (1988). Fine dissection of a nine amino acid glycoprotein epitope, a major determinant recognized by lymphocytic choriomeningitis virus‐specific class I‐restricted H‐2Db cytotoxic T lymphocytes. J. Exp. Med. 168, 559–570. Owen, D. J., and Evans, P. R. (1998). A structural explanation for the recognition of tyrosine‐based endocytotic signals. Science 282, 1327–1332. Porcelli, S., Brenner, M. B., Greenstein, J. L., Balk, S. P., Terhorst, C., and Bleicher, P. A. (1989). Recognition of cluster of differentiation 1 antigens by human CD4‐CD8‐ cytolytic T lymphocytes. Nature 341, 447–450. Porcelli, S., Morita, C. T., and Brenner, M. B. (1992). CD1b restricts the response of human CD48 T lymphoyctes to a microbial antigen. Nature 360, 593–597. Porcelli, S. A. (1995). The CD1 family: A third lineage of antigen‐presenting molecules. Adv. Immunol. 59, 1–98. Prigozy, T. I., Naidenko, O., Qasba, P., Elewaut, D., Brossay, L., Khurana, A., Natori, T., Koezuka, Y., Kulkarni, A., and Kronenberg, M. (2001). Glycolipid antigen processing for presentation by CD1d molecules. Science 291, 664–667. Prigozy, T. I., Sieling, P. A., Clemens, D., Stewart, P. L., Behar, S. M., Porcelli, S. A., Brenner, M. B., Modlin, R. L., and Kronenberg, M. (1997). The mannose receptor delivers lipoglycan antigens to endosomes for presentation to T cells by CD1b molecules. Immunity 6, 187–197. Procopio, D. O., Almeida, I. C., Torrecilhas, A. C., Cardoso, J. E., Teyton, L., Travassos, L. R., Bendelac, A., and Gazzinelli, R. T. (2002). Glycosylphosphatidylinositol‐anchored mucin‐like glycoproteins from Trypanosoma cruzi bind to CD1d but do not elicit dominant innate or adaptive immune responses via the CD1d/NKT cell pathway. J. Immunol. 169, 3926–3933.

136

D. BRANCH MOODY

Quaranta, M. G., Tritarelli, E., Giordani, L., and Viora, M. (2002). HIV‐1 Nef induces dendritic cell differentiation: A possible mechanism of uninfected CD4(þ) T cell activation. Exp. Cell Res. 275, 243–254. Rauch, J., Gumperz, J., Robinson, C., Skold, M., Roy, C., Young, D. C., Lafleur, M., Moody, D. B., Brenner, M. B., Costello, C. E., and Behar, S. M. (2003). Structural features of the acyl chain determine self‐phospholipid antigen recognition by a CD1d‐restricted invariant NKT (iNKT) cell. J. Biol. Chem. 278, 47508–47515. Roberts, T. J., Sriram, V., Spence, P. M., Gui, M., Hayakawa, K., Bacik, I., Bennink, J. R., Yewdell, J. W., and Brutkiewicz, R. R. (2002). Recycling CD1d1 molecules present endogenous antigens processed in an endocytic compartment to NKT cells. J. Immunol. 168, 5409–5414. Rodionov, D. G., Nordeng, T. W., Kongsvik, T. L., and Bakke, O. (2000). The cytoplasmic tail of CD1d contains two overlapping basolateral sorting signals. J. Biol. Chem. 275, 8279–8282. Rodionov, D. G., Nordeng, T. W., Pedersen, K., Balk, S. P., and Bakke, O. (1999). A critical tyrosine residue in the cytoplasmic tail is important for CD1d internalization but not for its basolateral sorting in MDCK cells. J. Immunol. 162, 1488–1495. Roura‐Mir, C., Catalfamo, M., Cheng, T. Y., Marqusee, E., Besra, G. S., Jaraquemada, D., and Moody, D. B. (2005a). CD1a and CD1c activate intrathyroidal T cells during Graves’ disease and Hashimoto’s thyroiditis. J. Immunol. 174, 3773–3780. Roura‐Mir, C., Wang, L., Cheng, T. Y., Matsunaga, I., Dascher, C. C., Peng, S. L., Fenton, M. J., Kirschning, C., and Moody, D. B. (2005b). Mycobacterium tuberculosis Regulates CD1 Antigen Presentation Pathways through TLR‐2. J. Immunol. 175, 1758–1766. Salamone, M. C., Mendiguren, A. K., Salamone, G. V., and Fainboim, L. (2001a). Membrane trafficking of CD1c on activated T cells. J. Leukoc. Biol. 70, 567–577. Salamone, M. C., Rabinovich, G. A., Mendiguren, A. K., Salamone, G. V., and Fainboim, L. (2001b). Activation‐induced expression of CD1d antigen on mature T cells. J. Leukoc. Biol. 69, 207–214. 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. Salomonsen, J., Sorensen, M. R., Marston, D. A., Rogers, S. L., Collen, T., van Hateren, A., Smith, A. L., Beal, R. K., Skjodt, K., and Kaufman, J. (2005). Two CD1 genes map to the chicken MHC, indicating that CD1 genes are ancient and likely to have been present in the primordial MHC. Proc. Natl. Acad. Sci. USA 102, 8668–8673. Sanchez, D. J., Gumperz, J. E., and Ganem, D. (2005). Regulation of CD1d expression and function by a herpesvirus infection. J. Clin. Invest. 115, 1369–1378. Sandhoff, K., and Kolter, T. (2003). Biosynthesis and degradation of mammalian glycosphingolipids. Philos. Trans. R. Soc. Lond. B Biol. Sci. 358, 847–861. Schofield, L., McConville, M. J., Hansen, D., Campbell, A. S., Fraser‐Reid, B., Grusby, M. J., and Tachado, S. D. (1999). CD1d‐restricted immunoglobulin G formation to GPI‐anchored antigens mediated by NKT cells. Science 283, 225–229. Shamshiev, A., Donda, A., Carena, I., Mori, L., Kappos, L., and De Libero, G. (1999). Self glycolipids as T‐cell autoantigens. Eur. J. Immunol. 29, 1667–1675. Shamshiev, A., Donda, A., Prigozy, T. I., Mori, L., Chigorno, V., Benedict, C. A., Kappos, L., Sonnino, S., Kronenberg, M., and De Libero, G. (2000). The ab T cell response to self‐ glycolipids shows a novel mechanism of CD1b loading and a requirement for complex oligosaccharides. Immunity 13, 255–264. Shamshiev, A., Gober, H. J., Donda, A., Mazorra, Z., Mori, L., and De Libero, G. (2002). Presentation of the same glycolipid by different CD1 molecules. J. Exp. Med. 195, 1013–1021. Shinya, E., Owaki, A., Shimizu, M., Takeuchi, J., Kawashima, T., Hidaka, C., Satomi, M., Watari, E., Sugita, M., and Takahashi, H. (2004). Endogenously expressed HIV‐1 nef down‐regulates


137

antigen‐presenting molecules, not only class I MHC but also CD1a, in immature dendritic cells. Virology 326, 79–89. Sidobre, S., Naidenko, O. V., Sim, B. C., Gascoigne, N. R., Garcia, K. C., and Kronenberg, M. (2002). The V alpha 14 NKT cell TCR exhibits high‐affinity binding to a glycolipid/CD1d complex. J. Immunol. 169, 1340–1348. Sieling, P. A., Chatterjee, D., Porcelli, S. A., Prigozy, T. I., Mazzaccaro, R. J., Soriano, T., Bloom, B. R., Brenner, M. B., Kronenberg, M., and Brennan, P. J. (1995). CD1‐restricted T cell recognition of microbial lipoglycan antigens. Science 269, 227–230. Sieling, P. A., Jullien, D., Dahlem, M., Tedder, T. F., Rea, T. H., Modlin, R. L., and Porcelli, S. A. (1999). CD1 expression by dendritic cells in human leprosy lesions: Correlation with effective host immunity. J. Immunol. 162, 1851–1858. Silk, J. D., Hermans, I. F., Gileadi, U., Chong, T. W., Shepherd, D., Salio, M., Mathew, B., Schmidt, R. R., Lunt, S. J., Williams, K. J., Stratford, I. J., Harris, A. L., and Cerundolo, V. (2004). Utilizing the adjuvant properties of CD1d‐dependent NK T cells in T cell‐mediated immunotherapy. J. Clin. Invest. 114, 1800–1811. Smiley, S. T., Kaplan, M. H., and Grusby, M. J. (1997). Immunoglobulin E production in the absence of interleukin‐4‐secreting CD1‐dependent cells. Science 275, 977–979. Somnay‐Wadgaonkar, K., Nusrat, A., Kim, H. S., Canchis, W. P., Balk, S. P., Colgan, S. P., and Blumberg, R. S. (1999). Immunolocalization of CD1d in human intestinal epithelial cells and identification of a beta2‐microglobulin‐associated form. Int. Immunol. 11, 383–392. Spence, P. M., Sriram, V., Van Kaer, L., Hobbs, J. A., and Brutkiewicz, R. R. (2001). Generation of cellular immunity to lymphocytic choriomeningitis virus is independent of CD1d1 expression. Immunology 104, 168–174. Stenger, S., Hanson, D. A., Teitelbaum, R., Dewan, P., Niazi, K. R., Froelich, C. J., Ganz, T., Thoma‐Uszynski, S., Melian, A., Bogdan, C., Porcelli, S. A., Bloom, B. R., Krensky, A. M., and Modlin, R. L. (1998a). An antimicrobial activity of cytolytic T cells mediated by granulysin. Science 282, 121–125. Stenger, S., Mazzaccaro, R. J., Uyemura, K., Cho, S., Barnes, P. F., Rosat, J. P., Sette, A., Brenner, M. B., Porcelli, S. A., Bloom, B. R., and Modlin, R. L. (1997). Differential effects of cytolytic T cell subsets on intracellular infection. Science 276, 1684–1687. Stenger, S., Niazi, K. R., and Modlin, R. L. (1998b). Down‐regulation of CD1 on antigen‐ presenting cells by infection with Mycobacterium tuberculosis. J. Immunol. 161, 3582–3588. Sugita, M., and Brenner, M. B. (1994). An unstable beta 2‐microglobulin: Major histocompatibility complex class I heavy chain intermediate dissociates from calnexin and then is stabilized by binding peptide. J. Exp. Med. 180, 2163–2171. Sugita, M., Cao, X., Watts, G. F., Rogers, R. A., Bonifacino, J. S., and Brenner, M. B. (2002). Failure of trafficking and antigen presentation by CD1 in AP‐3‐deficient cells. Immunity 16, 697–706. Sugita, M., Grant, E. P., van Donselaar, E., Hsu, V. W., Rogers, R. A., Peters, P. J., and Brenner, M. B. (1999). Separate pathways for antigen presentation by CD1 molecules. Immunity 11, 743–752. Sugita, M., Jackman, R. M., van Donselaar, E., Behar, S. M., Rogers, R. A., Peters, P. J., Brenner, M. B., and Porcelli, S. A. (1996). Cytoplasmic tail‐dependent localization of CD1b antigen‐ presenting molecules to MIICs. Science 273, 349–352. Sugita, M., van Der, W., Rogers, R. A., Peters, P. J., and Brenner, M. B. (2000). CD1c molecules broadly survey the endocytic system. Proc. Natl. Acad. Sci. 97, 8445–8450. Sumida, T., Sakamoto, A., Murata, H., Makino, Y., Takahashi, H., Yoshida, S., Nishioka, K., Iwamoto, I., and Taniguchi, M. (1995). Selective reduction of T cells bearing invariant Va24JaQ antigen receptor in patients with systemic sclerosis. J. Exp. Med. 182, 1163–1168.

138

D. BRANCH MOODY

Szatmari, I., Gogolak, P., Im, J. S., Dezso, B., Rajnavolgyi, E., and Nagy, L. (2004). Activation of PPARgamma specifies a dendritic cell subtype capable of enhanced induction of iNKT cell expansion. Immunity 21, 95–106. Terhorst, C., van Agthoven, A., Le Clair, K., Snow, P., Reinherz, E. L., and Schlossman, S. F. (1981). Biochemical studies of the human thymocyte cell surface antigens T6, T9 and T10. Cell 23, 771–780. Townsend, M. J., Weinmann, A. S., Matsuda, J. L., Salomon, R., Farnham, P. J., Biron, C. A., Gapin, L., and Glimcher, L. H. (2004). T‐bet regulates the terminal maturation and homeostasis of NK and Valpha14i NKT cells. Immunity 20, 477–494. Uehira, K., Amakawa, R., Ito, T., Tajima, K., Naitoh, S., Ozaki, Y., Shimizu, T., Yamaguchi, K., Uemura, Y., Kitajima, H., Yonezu, S., and Fukuhara, S. (2002). Dendritic cells are decreased in blood and accumulated in granuloma in tuberculosis. Clin. Immunol. 105, 296–303. Ueno, Y., Tanaka, S., Sumii, M., Miyake, S., Tazuma, S., Taniguchi, M., Yamamura, T., and Chayama, K. (2005). Single dose of OCH improves mucosal T helper type 1/T helper type 2 cytokine balance and prevents experimental colitis in the presence of valpha14 natural killer T cells in mice. Inflamm. Bowel. Dis. 11, 35–41. Ulrichs, T., Moody, D. B., Grant, E., Kaufmann, S. H., and Porcelli, S. A. (2003). T‐cell responses to CD1‐presented lipid antigens in humans with Mycobacterium tuberculosis infection. Infect. Immun. 71, 3076–3087. van de Wal, Y., Corazza, N., Allez, M., Mayer, L. F., Iijima, H., Ryan, M., Cornwall, S., Kaiserlian, D., Hershberg, R., Koezuka, Y., Colgan, S. P., and Blumberg, R. S. (2003). Delineation of a CD1d‐restricted antigen presentation pathway associated with human and mouse intestinal epithelial cells. Gastroenterology 124, 1420–1431. van den Elzen, P., Garg, S., Leon, L., Brigl, M., Leadbetter, E., Gumperz, J., Dascher, C., Cheng, T., Iaiaronov, P., Besra, G., Kent, S. C., Moody, D. B., Michael, B., and Brenner, M. B. (2005). An apolipoprotein mediated pathway of exogenous lipid antigen presentation. Nature 437, 906–910. van der Vliet, H. J., von Blomberg, B. M., Hazenberg, M. D., Nishi, N., Otto, S. A., van Benthem, B. H., Prins, M., Claessen, F. A., van den Eertwegh, A. J., Giaccone, G., Miedema, F., Scheper, R. J., and Pinedo, H. M. (2002). Selective decrease in circulating V alpha 24 þ V beta 11 þ NKT cells during HIV type 1 infection. J. Immunol. 168, 1490–1495. van Dommelen, S. L., Tabarias, H. A., Smyth, M. J., and Degli‐Esposti, M. A. (2003). Activation of natural killer (NK) T cells during murine cytomegalovirus infection enhances the antiviral response mediated by NK cells. J. Virol. 77, 1877–1884. Van Kaer, L. (2005). alpha‐Galactosylceramide therapy for autoimmune diseases: Prospects and obstacles. Nat. Rev. Immunol. 5, 31–42. Van Rhijn, I., Zajonc, D. M., Wilson, I. A., and Moody, D. B. (2005). Presentation of lipopeptide antigens to T cells. Curr. Opin. Immunol., in press. Vincent, M. S., Gumperz, J. E., and Brenner, M. B. (2003). Understanding the function of CD1‐ restricted T cells. Nat. Immunol. 4, 517–523. Wei, D. G., Lee, H., Park, S. H., Beaudoin, L., Teyton, L., Lehuen, A., and Bendelac, A. (2005). Expansion and long‐range differentiation of the NKT cell lineage in mice expressing CD1d exclusively on cortical thymocytes. J. Exp. Med. 202, 239–248. Weller, S., Braun, M. C., Tan, B. K., Rosenwald, A., Cordier, C., Conley, M. E., Plebani, A., Kumararatne, D. S., Bonnet, D., Tournilhac, O., Tchernia, G., Steiniger, B., Staudt, L. M., Casanova, J. L., Reynaud, C. A., and Weill, J. C. (2004). Human blood IgM ‘‘memory’’ B cells are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire. Blood 104, 3647–3654.


139

Wilson, S. B., Kent, S. C., Patton, K. T., Orban, T., Jackson, R. A., Exley, M., Porcelli Schatz, D. A., Atkinson, M. A., Balk, S. P., Strominger, J. L., and Hafler, D. A. (1998). Extreme Th1 bias of invariant Valpha24JalphaQ T cells in type 1 diabetes. Nature 391, 177–181. Winau, F., Schwierzeck, V., Hurwitz, R., Remmel, N., Sieling, P. A., Modlin, R. L., Porcelli, S. A., Brinkmann, V., Sugita, M., Sandhoff, K., Kaufmann, S. H., and Schaible, U. E. (2004). Saposin C is required for lipid presentation by human CD1b. Nat. Immunol. 5, 169–174. Wu, D. Y., Segal, N. H., Sidobre, S., Kronenberg, M., and Chapman, P. B. (2003). Cross‐ presentation of Disialoganglioside GD3 to Natural Killer T Cells. J. Exp. Med. 198, 173–181. Yu, K. O., Im, J. S., Molano, A., Dutronc, Y., Illarionov, P. A., Forestier, C., Fujiwara, N., Arias, I., Miyake, S., Yamamura, T., Chang, Y. T., Besra, G. S., and Porcelli, S. A. (2005). Modulation of CD1d‐restricted NKT cell responses by using N‐acyl variants of alpha‐galactosylceramides. Proc. Natl. Acad. Sci. USA 102, 3383–3388. Zajonc, D. M., Cantu, C., Mattner, J., Zhou, D., Savage, P. B., Bendelac, A., Wilson, I. A., and Teyton, L. (2005a). Structure and function of a potent agonist for the semi‐invariant natural killer T cell receptor. Nat. Immunol. 6, 810–818. Zajonc, D. M., Crispin, M. D., Bowden, T. A., Young, D. C., Cheng, T. Y., Hu, J., Costello, C. E., Rudd, P. M., Dwek, R. A., Miller, M. J., Brenner, M. B., Moody, D. B., and Wilson, I. A. (2005b). Molecular mechanism of lipopeptide presentation by CD1a. Immunity 22, 209–219. Zajonc, D. M., Elsliger, M. A., Teyton, L., and Wilson, I. A. (2003). Crystal structure of CD1a in complex with a sulfatide self antigen at a resolution of 2.15 A. Nat. Immunol. 4, 808–815. Zeng, Z., Castanõ, A. R., Segelke, B. W., Stura, E. A., Peterson, P. A., and Wilson, I. A. (1997). Crystal structure of mouse CD1: An MHC‐like fold with a large hydrophobic binding groove. Science 277, 339–345. Zhou, D., Cantu, C., III, Sagiv, Y., Schrantz, N., Kulkarni, A. B., Qi, X., Mahuran, D. J., Morales, C. R., Grabowski, G. A., Benlagha, K., Savage, P., Bendelac, A., and Teyton, L. (2004). Editing of CD1d‐bound lipid antigens by endosomal lipid transfer proteins. Science 303, 523–527. Zhou, D., Mattner, J., Cantu, C., III, Schrantz, N., Yin, N., Gao, Y., Sagiv, Y., Hudspeth, K., Wu, Y. P., Yamashita, T., Teneberg, S., Wang, D., Proia, R. L., Levery, S. B., Savage, P. B., Teyton, L., and Bendelac, A. (2004). Lysosomal glycosphingolipid recognition by NKT cells. Science 306, 1786–1789. Zinkernagel, R. M., and Doherty, P. C. (1974). Restriction of in vitro T cell‐mediated cytotoxicity in lymphocytic choriomeningitis within a syngeneic or semiallogeneic system. Nature 248, 701–702.

Lysophospholipids as Mediators of Immunity Debby A. Lin* and Joshua A. Boyce{ *Department of Medicine, Harvard Medical School, and Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Boston, Massachusetts { Departments of Medicine and Pediatrics, Harvard Medical School, Division of Rheumatology, Immunology, and Allergy, Brigham and Women’s Hospital, Boston, Massachusetts Abstract............................................................................................................. Introduction ....................................................................................................... LPL Synthesis .................................................................................................... Cell Surface Receptors for LPLs and Their Signaling Pathways ................................... Expression of LPA and S1P Receptors by Immune Cells and Their Functions in In Vitro Studies .................................................................................................. 5. In Vivo Functions of LPLS in Immune Responses and Inflammation............................ 6. Clinical Applications of S1P Receptor Agonists ......................................................... 7. Summary ........................................................................................................... References .........................................................................................................

1. 2. 3. 4.

141 141 143 145 148 156 159 160 160

Abstract Lysophospholipids (LPLs) are lipid‐derived signaling molecules exemplified by lysophosphatidic acid (LPA) and sphingosine 1‐phosphate (S1P). Originally identified as serum‐associated growth factors, these mediators now are known to signal through a family of diverse G protein‐coupled receptors (GPCRs). Virtually all cells that participate in the immune response express multiple receptors for LPLs. The development of antibody reagents that recognize the receptors for each LPL and the derivation of receptor‐selective agonists and receptor‐null mouse strains have provided insights into the widely diverse functions of LPLs in immune responses, particularly the role of S1P in lymphocyte trafficking. This review focuses on the biology of the LPLs as these molecules relate to functional regulation of immune cells in vitro and to the regulation of integrated immune responses in vivo. 1. Introduction Lysophospholipids (LPLs) are lipid signaling molecules that derive from cell membrane‐associated precursors. They can be generated rapidly by cells in response to a variety of perturbations, are abundant in normal biologic fluids, and have broad and potent actions. The two major classes of LPLs, lysoglycerolphospholipids and lysosphingophospholipids, are exemplified by


0065-2776/06 $35.00 DOI: 10.1016/S0065-2776(05)89004-2

142

DEBBY A. LIN AND JOSHUA A. BOYCE

Figure 1 Chemical structures of LPA (A) and S1P (B). The structure of LPA 18:2, the most abundant molecular species of LPA in biologic fluids, is depicted. Other molecular species include both unsaturated (16:1, 18:1) and saturated (12:0, 14:0, 16:0, and 18:0) members.

lysophosphatidic acid (LPA) and sphingosine 1‐phosphate (S1P), respectively. These structurally related lipids are defined by the absence of one fatty acid from one of two potential acylation positions on a glycerol backbone (Fig. 1). Originally recognized as serum‐associated growth factors for endothelial cells, both LPA and S1P serve diverse functions in the growth, motility, and differentiation of a wide array of cells. High affinity receptors exist for each ligand, and genetic strategies for targeted deletion of some of these receptors have revealed critical functions for LPA in neurodevelopment (Contos et al., 2000) and for S1P in the development of the vascular system (Liu et al., 2000). The recognition that these receptors are widely expressed by the cells responsible for both innate and adaptive immune responses and the development of receptor‐selective agonists and receptor‐null mouse strains have permitted studies of the functions of these endogenous mediators in immunity and inflammation in vivo. A comprehensive understanding of these functions for each LPL in immunity is emerging, especially the recognition of a prominent role for S1P in the homeostatic control of lymphocyte migration. Moreover, the fact that LPA stimulates a range of responses from lymphocytes, eosinophils, macrophages, and mast cells in vitro suggests that it has additional functions in vivo in the induction or amplification of immune or inflammatory responses. This review summarizes the functional characteristics of the LPLs in immunity and inflammation, with particular emphasis on LPA and S1P, their receptors, the putative functions of each mediator, the validated role of S1P in lymphocyte homing, and prospects for therapeutic development.

LY S O P H O S P H O L I P I D S A S M E D I AT O R S O F I M M U N I T Y

143

2. LPL Synthesis LPLs arise as a consequence of mobilization of cell membrane‐associated precursors, both by homeostatic mechanisms and during activation responses. The synthetic pathways for each ensure the production of LPA and S1P, respectively, at quantities sufficient to maintain concentrations of each mediator close to the micromolar range in the extracellular fluid (Pages et al., 2001; Yatomi et al., 1997). These pathways are detailed below. 2.1. LPA LPA circulates at high (low micromolar) concentrations in normal serum. It is bound by serum albumin (Tigyi and Miledi, 1992), which is a high‐capacity but low‐affinity carrier for LPA, and gelsolin (Goetzl et al., 2000a), which has less capacity but higher affinity and specificity. It is likely that these carriers regulate the presentation, strength, and duration of receptor activation by LPA. Several molecular species of LPA, including both unsaturated (16:1, 18:1, and 18:2) and saturated (12:0, 14:0, 16:0, and 18:0) forms, are present in serum, with LPA 18:2 being the most abundant form (Fig. 1). The existence of multiple pathways for LPA production (Pages et al., 2001) (Fig. 2) likely reflects the importance of LPA in cell homeostasis. Although some LPA may be generated at inner membrane leaflets by phospholipase D, most LPA is likely produced at the outer membrane leaflet or extracellularly from LPA precursor, lysophosphatidylcholine (LPC). LPC is generated from phosphatidyl choline (PC) on the leaflets of cell membranes through the actions of extracellular (secretory) phospholipase A2 (sPLA2) enzymes, such as group IIA

Figure 2 Pathways for LPA synthesis. Platelets are the dominant source of LPA in vivo. PC is converted to lysophosphatidyl choline (LPC) by secretory phospholipase A2 (sPLA2), phospholipase A1 (PLA1), or lecithin‐cholesterol acyltransferase (LCAT). LPA can then be produced directly from LPC via lysoPLD (also known as autotaxin) or can be converted directly from PA. LPA is hydrolyzed to monoacylglycerol MAG by lipid phosphate phosphatases.

144


PLA2. Platelets comprise the major source of LPC (and thus, of LPA) in normal serum (Sano et al, 2002); experimental depletion of platelets in rats decreases serum LPC concentrations by roughly half (Aoki et al., 2002). Other cells that can provide LPC (or that generate and secrete small quantities of LPA directly) include fibroblasts, adipocytes (which secrete LPA in response to stimulation with adrenergic agonists) (Pages et al., 1999), and potentially mononuclear phagocytes (which can also express sPLA2). The sources of LPC suggest mechanisms that may enhance LPA generation in diverse tissue injury, possibly to induce inflammation and/or initiate tissue healing through growth factor‐like activities. Furthermore, group IIA PLA2 is inducible in some cell types by interleukin (IL)‐1b, which suggests a mechanism by which LPA levels can increase during inflammation (Degousee et al., 2001). LPC can also be converted from membrane‐associated PC by phosphatidic acid‐specific PLA1, and it can be liberated by lecithin‐cholesterol acyltransferase from PC bound to serum lipoproteins (Aoki et al., 2002), thereby comprising a potential extracellular reservoir. Although both adipocytes and platelets can directly generate and release small amounts of LPA from endogenous enzymatic metabolism of LPC when they are activated by physiologic agonists, most extracellular LPA is likely converted from extracellular LPC by a serum‐ associated lysophospholipase D (lysoPLD), known also as autotaxin, which was originally identified as a serum factor that facilitated the chemotaxis of ovarian tumor cells. Lipoprotein‐bound LPC is constitutively in serum at concentrations approaching 100 mM (Aoki et al., 2002). Ovarian cancer cells generate especially large quantities of LPA, which is a potent autocrine growth factor for these cells (Xu et al., 1995b). An alternate pathway for LPA production involves conversion of phosphatidic acid (PA) directly to LPA by low molecular weight sPLA2 (Pages et al., 2001). Recently, sphingomyelinase D, which is an ecto‐enzyme of ticks and Corynebacterium species, has been identified as a virulence factor that converts host‐derived LPC to LPA (van Meeteren et al., 2004). In this instance, the local overproduction of LPA is pathogenetic and initiates regional inflammation and thrombosis. This finding suggests that LPA may initiate the innate immune response in vivo. The strength and duration of LPA‐mediated signaling is likely controlled by membrane‐associated lipid phosphate phosphatases, which hydrolyze LPA into monoacyl glyceride (MAG) (Smyth et al., 2003). 2.2. S1P Whereas LPA is generated predominantly in the extracellular compartment, S1P (which is present in physiologic fluids in a concentration range similar to that of LPA) is generated intracellularly, and its presence in the serum reflects


145

its release from the cells of origin. The cellular sources of S1P include platelets (Yatomi et al., 1997) and resident tissue mast cells (Jolly et al., 2004). S1P synthesis begins with the conversion of endogenous membrane‐derived sphingomyelin to ceramide (Cer) by sphingomyelinase, then to sphingosine (Sph) by ceremidase, which is converted to S1P via phosphorylation by one of two Sph kinases (SphK1 or SphK2) (Olivera and Spiegel, 2001) (Fig. 3). SphK1 translocates to the cell membrane of activated mast cells by a mechanism requiring the actions of PLD. Once synthesized, S1P serves both intracellular and extracellular signaling functions (Spiegel and Milstein, 2000). Intracellular S1P binds putative endoplasmic reticulum‐associated receptors to facilitate the release of intracellular stores of calcium into the cytosol during cell activation, accounting for a portion of intracellular calcium flux in mast cells (Melendez and Khaw, 2002) and macrophages (Melendez et al., 1998). S1P is also released by mast cells and platelets into the extracellular space where, like LPA, it is carried by albumin and is typically found in high nanomolar to low micromolar concentrations. Signaling initiated by S1P is limited by its dephosphorylation by two S1P phosphatases (Mandala, 2001) or an S1P lyase (Zhou and Saba, 1998), and it is counterbalanced by antagonistic signaling induced by Cer and Sph (see below). 3. Cell Surface Receptors for LPLs and Their Signaling Pathways 3.1. Receptors LPLs were initially recognized as serum constituents with growth factor‐like activity that were susceptible to inhibition by pertussis toxin (PTX), suggesting that these actions were mediated by specific G protein‐coupled receptors (GPCRs) coupled to Gi/o family proteins. In the late 1990s, a group of eight homologous GPCRs known as endothelial differentiation and growth (Edg) receptors was identified. Of these, three (initially designated the Edg‐2, Edg‐ 4, and Edg‐7 receptors) selectively bind LPA with high affinity (low nM concentration range); these receptors are now designated the LPA1, LPA2, and LPA3 receptors, respectively (An et al., 1998a; Bandoh et al., 1999; Hecht et al., 1996). An orphan GPCR (p2y9/GPR23) recently was shown to bind LPA selectively and with high affinity when expressed by transfection in a cell line (Noguchi et al., 2003). This putative fourth LPA receptor (now termed the LPA4 receptor) shares little sequence homology with the Edg family members, being more closely related to the purinergic (P2Y) receptor family members that recognize extracellular nucleotides. In some cell types, LPA‐mediated effects on cell growth were found to be independent of GPCRs. This observation was explained by the demonstration that LPA can also function as a transcellular agonist for the nuclear peroxisome proliferator‐activated receptor gamma

146


Figure 3 Pathway for S1P synthesis. Sphingomyelin is converted to Cer by sphingomyelinase. Cer is converted to Sph by ceramidase. Sph is phosphorylated by SphK1/2 to produce S1P, which is in turn dephosphorylated by S1P phosphatase.


147

(PPAR‐g) (McIntyre et al., 2003). In a model of atherosclerosis in rats, PPAR‐g was central to LPA‐induced neointima formation (Zhang et al., 2004). A total of five Edg receptors (Edg‐1, Edg‐5, Edg‐3, Edg‐6, and Edg‐8) selectively bind S1P with high affinity and specificity; these receptors are also designated the S1P1, S1P2, S1P3, S1P4, and S1P5 receptors, respectively (An et al., 1997; Graeler et al., 1998; Hla and Maciag, 1990; Im et al., 2000; Okazaki et al., 1993; Yamaguchi et al., 1996). An additional orphan receptor, GPR63, binds S1P with low affinity (Niedernberg et al., 2003). The existence of multiple receptors for S1P (as well as LPA) implies that the functions for each mediator may be considerably diverse. Such diversity is further suggested by the observation that S1P receptors can also form homo‐ and heterodimers (van Brocklyn et al., 2002), a process that expands the repertoire of signaling processes initiated by other GPCR systems. 3.2. Signaling and G Proteins Each receptor for LPA and S1P induces biologic responses by initiating signaling events through G protein subunits. When expressed by transfection into heterologous cell lines, LPA1 receptors primarily couple to PTX‐sensitive Gi family proteins (An et al., 1998b) and potently stimulate cell growth through Ras‐dependent activation of mitogen‐activated protein kinase (MAPK) cascades. LPA2 associates with G12/13, prominently stimulating the Rho/Ras/Rac GTPases (Moolenaar, 1999; Radeff‐Huang et al., 2004); this event is critical to stimulating cell motility through effects on the actin cytoskeleton. In some cell types, LPA2 receptors stimulate calcium flux in part by inducing intracellular generation of S1P by SphK1 (Young et al., 2000), suggesting a potential mechanism for cross‐talk among these respective LPLs. LPA3 receptors induce calcium flux using Gq proteins that stimulate phospholipase C (PLC) (Radeff‐Huang et al., 2004). The LPA4 receptor differs sharply from the Edg‐type LPA receptors in its ability to stimulate the accumulation of cAMP in transfected cells (Noguchi et al., 2003), likely reflecting coupling to Gs family proteins. GPCRs exhibiting this property induce ligand‐initiated signaling events that inhibit secretory functions of activated leukocytes and that counteract contractile responses in smooth muscle cells. Studies with transfected cells show that S1P1 receptors signal exclusively through Gi proteins (Windh et al., 1999), whereas S1P2 and S1P3 receptors signal using Gi, Gq, and G12/13 (Pyne and Pyne, 2000). Xenopus oocytes exhibited S1P‐induced calcium flux when transfected with either the S1P2 or the S1P3 receptor, but not with the S1P1 receptor (Ancellin and Hla, 1999). Although studies in transfectants indicate that each receptor has preferences for specific G proteins, there is strong evidence that each LPA receptor can couple to two or three different G proteins when

148


naturally expressed on primary cells. These findings further validate the diversity of signaling events and biologic responses to LPA and S1P. 4. Expression of LPA and S1P Receptors by Immune Cells and Their Functions in In Vitro Studies The receptors for LPA and S1P are broadly expressed on cell types of the immune system, including monocytes (Lee et al., 2002), B‐ and T‐lymphocytes (Girkontaite et al., 2004; Goetzl et al., 2000b), dendritic cells (DCs) (Idzko et al., 2002), mast cells (Bagga et al., 2004), natural killer (NK) cells (Maghazachi, 2003), neutrophils (Itagaki et al., 2005), and eosinophils (Roviezzo et al., 2004). The sections below outline the receptor‐mediated functions that have been identified for each cell type in vitro. 4.1. Lymphocytes Although transformed human T cells were reported to express several Edg receptor family members, the profile of these receptors expressed by primary human T cells is more limited. Jurkat leukemic T cells responded to LPA in vitro by migrating through an experimental connective tissue‐like barrier, accompanied by enhanced generation of matrix metalloproteases (Zheng et al., 2001). Freshly isolated human CD4þ T cells from peripheral blood predominantly express mRNA encoding the LPA2 receptor (Goetzl et al., 2000b), along with the corresponding protein. The generation of IL‐2 that is induced by stimulation of these primary cells with a combination of antibodies against CD3 and CD28 is decreased in a dose‐dependent manner when these cells are treated with LPA. The suppression of mitogen‐induced IL‐2 generation and decreased migration in response to LPA were both reproduced by activating LPA2 receptor‐selective monoclonal antibodies. Interestingly, no LPA receptors were identified on the CD8þ cells isolated from the same individuals in this study, and these cells did not respond functionally to LPA. Human CD4þ lymphocytes stimulated ex vivo with either a mitogenic lectin or with anti‐ CD3/anti‐CD28 exhibited inducible expression of the LPA1 receptor, accompanied by suppressed expression of the LPA2 receptor (Zheng et al., 2000). These changes in the receptor profile of CD4þ T cells were accompanied by striking corresponding changes in their functional responses to LPA. Specifically, LPA no longer induced chemotaxis itself and inhibited chemotaxis to exogenous chemokines; however, it stimulated (rather than suppressed) IL‐2 production. These in vitro experiments indicate that LPA receptor expression in CD4þ T cells is regulated dynamically by activation and suggest that naı¨ve T cells may respond initially to LPA as a chemoattractant and an inhibitor of mitogen‐induced IL‐2 production. It is tempting to speculate that


149

the transition to effector T cell state upon arrival at the site of antigen presentation may be accompanied by a loss of LPA‐induced chemotaxis and migrational arrest, but LPA may amplify the effector function of T cells (i.e., IL‐2 production) at this stage. In contrast to the suppressive effects of LPA on the production of IL‐2 by freshly isolated human CD4þ T cells, LPA stimulated Jurkat T cells for calcium flux, proliferation, and IL‐2 production (Xu et al., 1995a). Similarly, immortalized human B lymphoblasts responded to LPA with calcium flux, MAPK activation, and immunoglobulin production (Rosskopf et al., 1998), whereas LPA induced calcium flux and chemotaxis of cultured human Th1 and Th2 cells (Wang, L. et al., 2004). In a separate study, LPA (1 mM) reduced the proliferation of mouse CD4þ and CD8þ T cells by 16% and 11%, respectively, in response to the combination of stimulatory antibodies to CD3 and CD28 (Dorsam et al., 2003). LPA has also been reported to inhibit apoptosis of a human T lymphoblast cell line (Goetzl et al., 1999). Since recently derived mice lacking LPA1 and LPA2 receptors have yet to be subjected to a rigorous analysis of immune function, the true physiologic role of LPA in T cell trafficking and function in vivo has yet to be determined. Like LPA, S1P mediates multiple biologic functions of both mouse and human lymphocytes, although, unlike LPA, these effects are not limited to the CD4þ subset. Mouse C57BL/6 T cells express S1P1 and S1P4 receptors (Graeler and Goetzel, 2002), while human peripheral blood T cells express S1P1, S1P3, S1P4, and S1P5 receptors (Jin et al., 2003a). At low nanomolar concentrations well within the physiologic range in normal serum, S1P is chemotactic for cultured mouse splenic T cells across a Matrigel membrane (Graeler and Goetzl, 2002). S1P at similarly low concentrations also augmented chemotaxis of mouse T cells to the chemokines CCL‐21 and CCL‐5, but it inhibited chemotaxis to these ligands when it was present at high nanomolar‐ to‐micromolar levels. These observations suggest the potential for differential modulatory effects of S1P on lymphocyte trafficking that depend on regional S1P concentrations. The direct effects of S1P on chemotaxis were reproduced in rat HTC4 cells that were transfected with S1P1 receptors, but not those transfected with S1P4 receptors, suggesting that the former receptor was the most crucial for chemotactic function. Both the enhancing and inhibitory effects of S1P were blocked by T cell activation through the T cell receptor (Graeler et al., 2002). This maneuver downregulates the expression of both the S1P1 and S1P4 receptors, while simultaneously upregulating the expression of S1P3 and S1P5 in cultured human peripheral blood T cells (Jin et al., 2003a). The activation‐dependent downregulation of the S1P1 receptor is transient and recovers in mouse T cells by a mechanism that is dependent on protein kinase (PK) C e and AP‐1 transcriptional activity (Graeler et al., 2003). Thus, as is the

150


case for LPA, the response of T cells to S1P is dynamic and regulated in part by inducible changes in the levels and repertoire of expressed receptors, particularly that of the S1P1 receptor. These findings likely reflect events that are reproduced in vivo as demonstrated by models of lymphocyte trafficking (see below). Lymphocytes isolated from humans and mice exhibit responses to S1P in vitro in addition to chemotaxis. At relatively high (0.1–10 mM) concentrations, S1P inhibited the proliferation of human peripheral blood T cells induced by a combination of stimulating antibodies directed against CD3 and CD28, by stimulation with PMA/ionomycin, or by co‐culture with antigen‐loaded DCs. The inhibitory effects of S1P on PMA/ionomycin stimulation could be blocked by pre‐treatment of the cells with PTX. These findings imply the involvement of a Gi‐coupled S1P receptor family member, whereas PTX did not interfere with the effect of S1P on proliferation induced by anti‐CD3/anti‐CD28 or stimulation with DCs (Jin et al., 2003a). At slightly lower doses (0.01–1 mM), S1P inhibited the proliferation of splenic CD4þ and CD8þ T lymphocytes from C57BL/6 mice that occurred in response to stimulation with anti‐CD3/ anti‐CD28 or anti‐CD3 plus IL‐7 (Dorsam et al., 2003). Thus S1P has the potential to interfere with the proliferation of both human and mouse T cells and can do so by both Gi/o‐dependent and ‐independent signaling pathways. S1P (1–10 mM) also amplifies the production of IL‐2 and interferon (IFN)‐g by human T cells stimulated by anti‐CD3/anti‐CD28. This costimulatory effect was inhibited by either exogenous Sph or Cer (Jin et al., 2003a). S1P also enhances IL‐10 production by mouse CD4þ CD25þ regulatory T cells, facilitating the capacity of these cells to suppress IL‐2 production and proliferation of CD4þ CD25 T cells (Wang, W. et al., 2004). In another study, S1P decreased the production of IFN‐g and IL‐4, but not that of IL‐2, by antigen receptor‐stimulated CD4þ T cells from C57BL/6 mice. S1P inhibition of IFN‐g production was dependent on S1P1 receptor expression (Dorsam et al., 2003). Thus S1P either can directly augment the secretion of certain cytokines by lymphocytes or can also suppress lymphocyte functions, depending on context and (likely) on the profile of S1P receptors expressed by a given lymphocyte cell population. Since this repertoire is dynamically regulated, it is not surprising that the nature of the responses to S1P by lymphocytes in vitro vary from study to study. 4.2. Natural Killer Cells Human NK cells that are stimulated with IL‐2 express the LPA1, LPA2, and LPA3 receptors. Stimulation of these primed NK cells with LPA induced a chemotactic response in vitro that was inhibited by pre‐treatment of the cells


151

with PTX. LPA and PA induced calcium flux and modestly augmented IFN‐g production by NK cells (Jin et al., 2003b). Whereas S1P receptors were not profiled in this study and S1P did not induce chemotaxis of these activated NK cells, in another study human NK cells expressed S1P1, S1P4, and S1P5 receptors and responded chemotactically to S1P in a PTX‐sensitive manner (Kveberg et al., 2002). Thus, like lymphocytes, NK cells express a diverse profile of LPA and S1P receptors and exhibit responses to these ligands that can be modulated by activation. The role of NK cells in anti‐tumor immunity raises the possibility that LPLs may provide an innate stimulus for the recruitment of NK cells to sites of tumor invasion. These studies also suggest the existence of additional functions for LPLs in anti‐viral immunity. 4.3. Dendritic Cells DCs, the most potent antigen‐presenting cells, can be differentiated from human monocytes in vitro. The addition of LPC, the precursor of LPA, to cultured human peripheral blood monocytes augmented their differentiation into mature DCs and was associated with extracellular signal‐regulated kinase (ERK) MAPK phosphorylation. In a mixed lymphocyte reaction, the LPC‐ treated monocytes enhanced the production of IL‐2 and IFN‐g by T cells. The effect of LPC on CD86 expression was partially inhibited by PTX and BN52021, an antagonist of the platelet‐activating factor receptor (Coutant et al., 2002). During antigen processing, maturing DCs undergo tightly coordinated changes in cell surface phenotype and responses to endogenous chemoattractants so as to facilitate their migration to regional lymph nodes. Although both immature and mature human DCs express mRNA transcripts for LPA1, LPA2, and LPA3 receptors (Panther et al., 2002) as well as for S1P1, S1P2, S1P3, and S1P4 receptors (Idzko et al., 2002), the responses of DCs to LPA and S1P change substantially as the cells mature. Both LPA and S1P elicited PTX‐sensitive calcium flux, actin polymerization, and chemotaxis in immature DCs. These chemotactic responses were lost after the cells were stimulated with lipopolysaccharide (LPS) to induce their maturation. However, both LPA and S1P (at high nanomolar/low micromolar range concentrations) inhibited the LPS‐mediated generation of IL‐12 and TNF‐a by DCs, and augmented their production of IL‐10 in a PTX‐insensitive manner (Idzko et al., 2002; Panther et al., 2002; Renkl et al., 2004). The changes in cytokine profile were associated with a reduced ability of these LPS‐stimulated DCs to polarize T cells toward the production of IFN‐g. Both the selective S1P receptor agonist FTY720 and its phosphorylated analogue, FTY720‐P, inhibited the chemotactic response of immature and mature DCs to S1P. Furthermore, mature DCs treated with FTY720 and FTY720‐P exhibited reduced IL‐12

152


production and increased IL‐10 production. T cells co‐cultured with these DCs shifted from a Th1 to a Th2 cytokine profile (Muller et al., 2005). Collectively, these observations suggest that either LPA or S1P could favor polarization of T cells toward a pro‐allergic phenotype during the development of adaptive immunity through effects mediated by DCs. 4.4. Monocytes/Macrophages Peripheral blood monocytes and/or tissue macrophages in mice, humans, and rats all express receptors for LPA and S1P (Duong et al., 2004; Hornuss et al., 2001). Mouse peritoneal macrophages express the LPA1, S1P1, and S1P3 receptors. Stimulation of these mouse macrophages with LPA or S1P upregulated their expression of IL‐1b and TNF‐a transcripts and protein, and downregulated IL‐2 transcription (Lee et al., 2002). Human monocytes, macrophages, and the Mono Mac 6 (MM6) monocytic cell line express mRNA transcripts for LPA1, LPA2, S1P1, S1P2, and S1P4. Both LPA and S1P in nanomolar range concentrations stimulated calcium flux in LPS‐activated MM6 cells via a PLC and Gi pathway (Fueller et al., 2003). An important role for the S1P3 receptor in the localization of MOMA1þ metallophilic macrophages and endothelial cells within mouse splenic marginal sinus zones has been recognized in studies of mice lacking this receptor (Girkontaite et al., 2004). 4.5. Granulocytes Human peripheral blood eosinophils express mRNA encoding both LPA1 and LPA3 receptors. At low micromolar range concentrations, LPA induced chemotaxis, actin cytoskeletal rearrangement, CD11b upregulation, and oxidative burst by human eosinophils. These responses were sensitive to inhibition by PTX and were blocked by a dual‐selective antagonist of the LPA1 and LPA3 receptors, diacylglycerol pyrophosphate (Idzko et al., 2004). Human eosinophils also express S1P1 receptors and, to a lesser extent, S1P2 and S1P3 receptors. Importantly, ex vivo stimulation of human eosinophils with S1P sharply upregulated their expression of the mRNA encoding the critical chemokine receptor CCR3, as well as that of the CCR3 ligand RANTES (Roviezzo et al., 2004). Thus both S1P and LPA may induce or modify patterns of eosinophil recruitment in certain contexts. This possibility is supported by the observation that the injection of S1P induces recruitment of eosinophils into the footpads of rats (Roviezzo et al., 2004). Like eosinophils, human peripheral blood neutrophils respond to both LPA and S1P. LPA induced calcium flux and oxidative burst from neutrophils, neither of which was inhibited by PTX, thus suggesting a potential alternative


153

pathway for neutrophil activation (Itagaki et al., 2005). However, local anesthetics have been shown to inhibit LPA‐induced chemotaxis and priming of neutrophils (Fischer et al., 2001), which may be important for their anti‐ inflammatory effects. LPA also stimulates neutrophil degranulation, in part through enhanced PLD activity (Tou and Gill, 2005). In an in vivo model, inhalation of LPA by guinea pigs increased eosinophil and neutrophil numbers in bronchoalveolar lavage fluid as well as superoxide production (Hashimoto et al., 2003). In contrast, other studies have shown that LPA inhibits the neutrophil metabolic burst after stimulation of the cells with fMLP and phorbol 12‐myristate 13‐acetate (PMA) (Chettibi et al., 1994). S1P stimulates calcium flux in human neutrophils independently of GPCRs and instead acts through store‐operated calcium entry channels (Itagaki and Hauser, 2003). Inhibition of SphK led to diminished neutrophil chemotaxis in vitro. In a rat model of hemorrhagic shock, SphK inhibition also decreased both CD11b expression, a marker of neutrophil activation, and lung injury as measured by permeability to dye (Lee et al., 2004). 4.6. Mast Cells Mast cells reside in perivascular spaces in all organs and are ideally situated to receive LPL signals from the vasculature. Human mast cells derived in vitro from umbilical cord blood express transcripts encoding all four LPA receptors (Bagga et al., 2004), and exhibit cytofluorographic expression of LPA1, LPA2, and LPA3 receptors (Lin and Boyce, 2005). In an in vitro model culture system for mast cell development from human cord blood‐derived progenitor cells, LPA potently stimulated the proliferation of mast cells, providing a synergistic signal with the obligate mast cell growth factor, stem cell factor (SCF). At peak concentrations (5 mM), LPA increased mast cell numbers by 10‐fold and strongly enhanced the formation of secretory granules and the expression of alpha and beta tryptases. PTX blocked LPA‐induced proliferation virtually completely (Bagga et al., 2004) and partly interfered with LPA‐induced calcium fluxes (Lin and Boyce, 2005) (Fig. 4). Whereas unprimed human mast cells did not generate cytokines or chemokines in response to LPA, mast cells primed with the Th2 cytokine IL‐4 produced abundant quantities of the chemokines macrophage inflammatory protein (MIP)‐1b, monocyte chemotactic protein (MCP)‐1, and IL‐8. The IL‐4‐dependent priming upregulated expression of mitogen‐activated protein kinase‐kinase, permitting LPA to induce strong phosphorylation of ERK. Interestingly, whereas LPA‐dependent proliferation was completely blocked by a dual LPA1/LPA3 receptor‐selective antagonist VPC32179, this compound did not affect LPA‐mediated chemokine generation, which instead was mimicked by fatty alcohol phosphate‐12, a

154


Figure 4 Effect of LPA on mast cell growth and activation. LPA induces mast cell proliferation synergistically with SCF and stimulates tryptase expression through a pathway involving PPAR‐g, LPA1, and/or LPA3. In contrast, LPA induces chemokine (MIP‐1b, MCP‐1, and IL‐8) production through LPA2 receptors, and this production depends on IL‐4 priming and MAPK phosphorylation. KIT, c‐kit tyrosine kinase.

selective agonist of the LPA2 receptor. Thus LPA may both support reactive mastocytosis (a feature observed in several disease states) and may also serve as an amplifier of mucosal inflammation where mast cell hyperplasia is mediated by a Th2 cytokine‐based mechanism. Cer, Sph, and S1P exert complex and generally opposing effects on mast cell activation, a phenomenon referred to as the ‘‘sphingolipid rheostat’’ (Olivera and Rivera, 2005). Whereas the addition of exogenous Cer or Sph to mouse bone marrow‐derived mast cells (BMMCs) in vitro induces apoptosis (Itakura et al., 2002), S1P promotes proliferation and effector functions (Jolly et al., 2004; Prieschl et al., 1999) (Fig. 5). Exogenous Sph inhibited MAPK activation and the generation of leukotrienes and cytokines by BMMCs classically activated by cross‐linkage of the high‐affinity Fc receptor for IgE (FceRI) but had minimal effect on cell degranulation. The addition of S1P reversed Sph‐ induced inhibition of activation. A rat mast cell line, RBL‐2H3, produced MIP‐1b and MCP‐1 in response to stimulation with exogenous S1P. The mouse mast cell line CPII responded to stimulation with exogenous S1P with exocytosis and leukotriene secretion, as well as with TNF‐a production when S1P was provided in conjunction with ionomycin (Prieschl et al., 1999).


155

Figure 5 Effect of S1P on mast cell activation. Antigen/IgE crosslinking of FcRI on mast cells activates SphK1 via a PLD‐dependent pathway to produce S1P. S1P acts intracellularly to induce Ca2þ release from the endoplasmic reticulum, which is necessary for exocytosis. S1P also acts extracellularly on S1P2 receptors to mediate calcium release via PLC/IP3, and on S1P1 to mediate antigen‐directed chemotaxis. The ability of mast cells to secrete S1P in response to activation suggests a potential mechanism for lymphocyte recruitment in allergic disease.

Although the receptor(s) responsible were not identified in these studies, both BMMCs and RBL‐2H3 cells express both S1P1 and S1P2 receptors, but do not express S1P3, S1P4, and S1P5 receptors. In addition to the paracrine effects already discussed, S1P also regulates mast cell activation in an autocrine manner. FcRI‐induced activation of CPII cells, BMMCs, and cultured human mast cells results in the generation and/or secretion of S1P. In human and mouse BMMCs, FcRI cross‐linkage induced the translocation of SphK1 to the cell membrane, a necessary prerequisite to S1P generation (Choi et al., 1996; Jolly et al., 2004; Melendez and Khaw, 2002). SphK1 required activation of PLD in human mast cells, and S1P production was in turn required for the calcium flux and exocytosis that occurred in response to FcRI cross‐linkage, presumably through the effects of S1P actions at the endoplasmic reticulum to liberate intracellular stores of calcium. Endogenous S1P generation was also required for optimal exocytosis of granule contents by BMMCs; however, this event reflected transactivation of the S1P2 receptor by secreted S1P, rather than intracellular actions (Jolly et al., 2004). Additionally, transactivation of the S1P1 receptor was required for migration of BMMCs toward antigen in this study, based on RNA interference

156


with the pertinent receptors. Collectively, these studies support the likely importance of the Cer/Sph/S1P rheostat in the autocrine regulation of activation‐induced mast cell signaling and establish clearly distinct functions for the S1P1 and S1P2 receptors. Moreover, the ability of activated mast cells to secrete S1P in response to activation may have implications for chemotaxis of lymphocytes and other S1P‐responsive cells in allergic responses. 5. In Vivo Functions of LPLS in Immune Responses and Inflammation 5.1. LPA Recent studies have addressed LPA‐mediated functions in inflammation using direct challenges. The intranasal administration of LPA to immunologically naı¨ve C57BL/6 mice resulted in increased MIP‐2 and neutrophil recruitment in BAL fluid (Cummings et al., 2004). These effects were attributed to stimulatory effects on bronchial epithelial cells, since this same study demonstrated that LPA induced IL‐8 production in human bronchial epithelial cells that was dependent on NF‐kB transcription and protein kinase C‐d. The injection of LPA into the perivascular tissues of rat carotid arteries induced the formation of neointima, accompanied by local proliferation of macrophages (Zhang et al., 2004). These events were attributed to activation of PPAR‐g. Although no studies have specifically addressed the role of any specific LPA receptor in immunity or inflammation, the availability of mouse strains that lack LPA1 receptors (which have defective forebrain development), LPA2 receptors (which are phenotypically normal), or both receptors should permit such investigations (Contos et al., 2002). 5.2. S1P In contrast to the relatively limited information concerning the role of LPA in immune function in vivo, an abundant amount of literature supports the role of S1P. Intravenously administered S1P induced a rapid lymphopenia in experimental animals, an effect attributed to the sequestration of both T cells and B cells in secondary lymphoid organs. Thymocytes acquire S1P1 receptor expression during their maturation into CD4þ and CD8þ lymphocytes (Matloubian et al., 2004). Constitutive levels of S1P in the blood are sufficient to induce the egress of these naı¨ve T cells into the circulation. S1P1 receptor expression is then cyclically modulated, being down‐regulated in the blood, subsequently upregulated in the secondary lymphoid tissues, and then down‐modulated once again with transit to the lymph (Lo et al., 2005) (Fig. 6).


157

Figure 6 Regulation of lymphocyte trafficking by S1P1 receptor signaling. S1P1 receptor expression is acquired during thymic development by single positive CD4þ or CD8þ T cells. Lymphocytes migrate into the blood and then to secondary lymphoid organs where the relatively low levels of S1P allow upregulation of S1P1 expression. S1P1 signaling facilitates lymphocyte egress into lymph. The administration of FTY720 downregulates S1P1 receptor expression and inhibits lymphocyte egress from thymus and lymphoid organs.

Both antigen‐induced and S1P‐induced stimulation of lymphocytes likely account for the down‐modulation of S1P1 receptor expression in the peripheral blood. The down‐modulation of S1P1 receptor expression likely permits chemokine‐dependent migration of T cells to regional lymph nodes, a process that requires the function of endogenously generated cysteinyl leukotrienes and the leukotriene‐specific multidrug transporter (Honig et al., 2003; Yopp et al., 2004). The acquisition of the S1P1 receptor during T cell development in the thymus and its re‐acquisition by mature lymphocytes in the lymph nodes permits S1P to promote egress of nascent T cells from the thymus and induce the re‐circulation of mature T cells from the secondary lymphoid organs, respectively. Whereas S1P1 receptor deficiency caused embryonic lethality due to a defect in the development of the microvasculature (Liu et al., 2000), mice with conditional deletion of the S1P1 receptor exhibited thymic hyperplasia due to retention of CD4þ and CD8þ T cells in the thymus (Allende et al., 2004). Moreover, the adoptive transfer of fetal liver‐derived hematopoietic progenitor cells from S1P1 receptor‐deficient mice into lethally irradiated recipient mice resulted in sequestration of T lymphocytes in the thymus and lymph nodes (Matloubian et al., 2004). Conversely, the transfer of T cells from mice transgenically over‐expressing S1P1 receptors resulted in increased numbers of T cells in the circulation and decreased numbers in the lymph nodes (Graeler et al., 2005). In transgenic cells, S1P1 surface expression was reduced in the blood and lymph and increased in spleen and lymph nodes (Lo et al., 2005). In studies of S1P1 receptor‐null mice, T lymphocytes and precursors could enter the thymus and peripheral lymphoid tissues but could not leave (Allende et al., 2004; Matloubian et al., 2004). Collectively, these

158


studies support a major role for the S1P1 receptor in the regulation of T cell trafficking and indicate that the expression of this receptor is dynamically regulated in vivo. The ability of systemically administered S1P to elicit lymphopenia is similar to that of 2‐amino‐2‐(2‐[4‐octylphenyl]ethyl)‐1,3‐propanediol (FTY720) (Adachi et al., 1995), an immunosuppressant isolated from the fungus Isaria sinclairii, which induced profound lymphopenia in rats. The systemic administration of S1P, FTY720, or FTY720‐P induced sequestration of lymphocytes in peripheral lymphoid tissues (Brinkmann et al., 2002; Mandala et al., 2002), but did not affect the numbers of myelomonocytic cells. These similarities were explained by the subsequent recognition that FTY720 and FTY720‐P are potent agonists at all S1P receptors except for the S1P2 receptor (Brinkmann et al., 2002; Mandala et al., 2002). This discovery greatly accelerated the acquisition of knowledge about the physiologic functions of S1P. In in vitro studies, FTY720 crosses cell membranes and is phosphorylated by SphK2 to a greater extent than by SphK1, although much less efficiently than is Sph. FTY720 also inhibits the phosphorylation of Sph by SphK2 to a greater extent than by SphK1 (Paugh et al., 2003). In studies of rat HTC4 cells engineered to express S1P receptor subtypes, the administration of FTY720 caused pronounced internalization and degradation of the S1P1, S1P2, and S1P5 receptors but did not affect the expression of S1P3 or S1P4 receptors (Graeler and Goetzl, 2004). Systemically administered FTY720 depleted both T and B lymphocytes in thoracic duct, peripheral blood, and spleen (Brinkmann et al., 2002; Chiba et al., 1998; Mandala et al., 2002). FTY720‐induced lymphocyte homing to secondary lymphoid organs depends on CD49d, CD62L, and CD11a, as indicated by experiments with specific antibody blockade (Chiba et al., 1998). FTY720 also strikingly inhibits thymocyte egress from thymus (Matloubian et al., 2004). Indeed, systemically administered FTY720 induces equal depletion of naı¨ve and effector/memory T cells from the peripheral circulation. Many of the bioactive effects of FTY720 and its phosphorylated metabolite reflect its capacity to induce striking internalization and subsequent down‐modulation of S1P receptors, particularly that for S1P1. The administration of S1P, FTY720, or FTY720‐P to wild‐type mice sharply down‐modulates S1P1 receptor expression in vivo (Matloubian et al., 2004), thus arresting emigration of nascent CD4þ and CD8þ lymphocytes from the thymus while preventing the recirculation of mature T cells from the regional nodes. Another S1P receptor agonist, 2‐amino‐4-4‐heptyloxyphenyl‐2‐methylbutanol (Kiuchi et al., 2000), also inhibits thymic egress, which is accompanied by loss of CD69 on CD4þ and CD8þ thymocytes in C57BL/6 mice (Rosen et al., 2003). Thus, T lymphopenia induced by S1P receptor agonists relates directly to the requirement for S1P1 receptor signaling at two distinct developmental stages in T cells.


159

S1P receptor agonism induced the migration of splenic B cells from the marginal zone into lymphoid follicles of mice. Based on studies of S1P1‐ deficient fetal liver chimeric mice, B cell localization to the marginal zone of the spleen depends on S1P1 (Cinamon et al., 2004). Activated B cells from C57BL/6 mice modified to overexpress S1P1 demonstrated decreased localization to the white pulp of the spleen compared to control cells. Treatment of the mice with FTY720 restored trafficking of the B cells to the white pulp. In studies of S1P3 null and reconstituted mice, S1P3 expression was required for normal migration and organization of splenic marginal sinus B lymphocytes (Girkontaite et al., 2004). Thus B cell lymphopenia induced by FTY720 reflects the arrest of normal S1P‐induced migration patterns in the spleen, which in turn reflect contributions from both the S1P1 and S1P3 receptors. 6. Clinical Applications of S1P Receptor Agonists Studies in animal models demonstrate that FTY720 is a potently immunosuppressant in vivo, with beneficial effects on allograft survival and in autoimmune disease models. FTY720 promoted skin graft survival in MHC incompatible rats. This effect was attributed to the redistribution of lymphocytes to the peripheral lymphoid organs rather than to transplanted tissues (Chiba et al., 1998, 1999). Such altered lymphocyte trafficking could have a potentially beneficial effect. In a mouse model, FTY720 diminished graft‐ versus‐host disease without affecting graft‐versus‐leukemia, due to lymphocyte migration into lymph nodes rather than tissues (Kim et al., 2003). In a mouse model of experimental autoimmune encephalitis, the administration of FTY720 or FTY720‐P resulted in clinical improvement (Webb et al., 2004). Furthermore, the administration of FTY720 prevented pulmonary inflammation induced by allergen challenge in mice that received adoptive transfer of antigen‐specific Th1 or Th2 cell clones, and abrogated antigen‐induced airway hyperresponsiveness (Sawicka et al., 2003). These observations in mice collectively support the therapeutic potential of S1P receptor agonism in organ transplantation, autoimmunity, and allergic diseases. Importantly, and in contrast to other immunosuppressive drugs, FTY720 has been reported not to impair immunity to specific pathogens. FTY720 (0.3 mg/kg) did not inhibit cell‐mediated or antibody responses in mice infected with lymphocytic choriomeningitis virus or vesicular stomatitis virus (Pinschewer et al., 2000). FTY720 did not alter IL‐4 and IFN‐g production by T cells, IgG production by B cells, or T cell proliferation responses in mixed‐lymphocyte reactions or after TCR stimulation (Brinkmann et al., 2001). However, in other studies, FTY720 diminished the immune response against specific antigens. FTY720 reduced the number of antigen‐specific CD4þ T cells in the draining lymph

160


nodes and peripheral blood of mice after antigen challenge. This effect was attributed to reduced recirculation of naı¨ve and antigen‐specific T cells from lymph nodes, rather than inhibition of T cell proliferation (Xie et al., 2003). FTY720 also reduced the production of high‐affinity IgG1 antibodies against the T‐cell‐dependent antigen, alum‐precipitated nitrophenyl acetyl conjugated to chicken g‐globulin, in immunized mice. Antibody responses to the T cell‐ independent antigen, nitrophenyl‐Ficoll, were not affected. FTY720 (at doses 20 ng/ml, while normal subjects had levels 80%

CM, SM, SM‐AHNMD CM SM‐AHNMD ASM SM SM CM SM‐AML SM‐familial CM, ISM SM‐AHNMD SM‐AHNMD SM‐AHNMD

Advances in Immunology Volume 89

Advances in Agronomy, Volume 89

Advances in Agronomy, Volume 89

Advances in Immunology Volume 45

Advances in Immunology Volume 34

Advances in Immunology Volume 1

Advances in Immunology Volume 10

Advances in Immunology Volume 30

Advances in Immunology Volume 82

Advances in Immunology Volume 31

Advances in Immunology Volume 52

Advances in Immunology Volume 48

Advances in Immunology Volume 56

Advances in Immunology Volume 12

Advances in Immunology Volume 21

Advances in Immunology Volume 41

Advances in Immunology Volume 71

Advances in Immunology Volume 84

Advances in Immunology Volume 43

Advances in Immunology Volume 73

Advances in Immunology, Volume 61

Advances in Immunology Volume 14

Advances in Immunology Volume 46

Advances in Immunology, Volume 106

Advances in Immunology Volume 19

Advances in Immunology Volume 99

Advances in Immunology Volume 81

Advances in Immunology Volume 4

Advances in Immunology Volume 80

Advances in Immunology Volume 64

Advances in Immunology Volume 89

Advances in Agronomy, Volume 89

Advances in Agronomy, Volume 89

Advances in Immunology Volume 45

Advances in Immunology Volume 34

Advances in Immunology Volume 1

Advances in Immunology Volume 10

Advances in Immunology Volume 30

Advances in Immunology Volume 82

Advances in Immunology Volume 31

Advances in Immunology Volume 52

Advances in Immunology Volume 48

Advances in Immunology Volume 56

Advances in Immunology Volume 12

Advances in Immunology Volume 21

Advances in Immunology Volume 41

Advances in Immunology Volume 71

Advances in Immunology Volume 84

Advances in Immunology Volume 43

Advances in Immunology Volume 73

Advances in Immunology, Volume 61

Advances in Immunology Volume 14

Advances in Immunology Volume 46

Advances in Immunology, Volume 106

Advances in Immunology Volume 19

Advances in Immunology Volume 99

Advances in Immunology Volume 81

Advances in Immunology Volume 4

Advances in Immunology Volume 80

Advances in Immunology Volume 64

Recommend Documents