ADVANCES /N
Immunology EDITED BY FRANK J. DIXON Scripps Clinic and Research Foundation La Jolla, California ASSOCIATE EDITORS
K. FRANK AUSTEN JONATHAN
W. UHR
TADAMITSU KISHIMOTO FRITZMELCHERS FREDERICK ALT
VOLUME 54
ACADEMIC PRESS, INC. A Division of Harcourl Brace & Company
San Diego New York Boston London Sydney Tokyo Toronto
This book is printed on acid-free paper.@
Copyright 0 1993 by ACADEMIC PRESS, INC. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Academic Press, Inc.
1250 Sixth Avenue, San Diego, California 92101-431 1
United Kingdom Edition published by
Academic Press Limited
24-28 Oval Road, London NW 1 7DX International Standard Serial Number: 0065-2776
International Standard Book Number: 0- 12-022454-2
PRINTED IN THE UNITED STATES OF AMERICA 9 3 9 4 9 5 9 6 9 1
98
BB
9
8
7
6
5
4
3
2
1
ADVANCES IN IMMUNOLOGY, VOL. 54
Interleukin-6 in Biology and Medicine SHIZUO AKIRA,. TETSUYA TAGA,' AND TADAMITSU K l S H l M O T W
* Institute for Molecular and Cellular Biology, Osaka University, Osaka 565, Japan,
t
and Department of Medicine Ill, Osaka University Medical School, Osaka 553, Japan
1. Introduction
Interleukin-6 (IL-6) is a cytokine with pleiotropic activities that plays a central role in host defense (Kishimoto, 1989; Shegal, 1990; Akira et al., 1990a; Hirano et al., 1990; Van Snick, 1990).IL-6 can exert growth-inducing, growth-inhibitory, and differentiation-inducing activities, depending on the target cells. These activities include (1)terminal differentiation (secretion of immunoglobulins) in B cells, (2) growth promotion on various B cells (myeloma/plasmacytoma/ hybridoma cells), (3) support of multipotential colony formation by hematopoietic stem cells, (4) elicitation of hepatic acute-phase response, ( 5 ) differentiation and/or activation of T cells and macrophages, and (6) neural differentiation. In previous studies this molecule was described with various designations such as B cell stimulatory factor 2 (BSF-2),interferon-& (IFN-&), hybridoma growth factor (HGF), and hepatocyte-stimulating factor (HSF).The name IL-6 was proposed because the nucleotide sequences of all these proteins were found to be identical (Table I). IL-6 has now been implicated in the pathology of many diseases including multiple myeloma, mesangial proliferative glomerulonephritis, rheumatoid arthritis, and acquired immunodeficiency syndrome (AIDS). Selective inhibition of the synthesis or of the action of IL-6 may have therapeutic benefit against the IL-6-associated diseases. On the other hand, IL-6 has potent antitumor activity against certain types of tumors. Application of IL-6 is promising in cancer treatment as well as in treatment of radiation- or chemotherapy-induced myelosuppression. II. Historical Overview
A. B CELLSTIMULATORYFACTOR 2 On antigenic stimulation, B cells proliferate and differentiate into antibody-producing cells under the control of T cells and macrophages. This process was found to be mediated by soluble factors 1 Copyright 8 1993 hy Academic Press, Inc.
All rights of reproduction in any form reserved.
2
SHIZUO AKIRA ET AL.
TABLE I FACTORS THATHAVETURNED OUT TO BE IDENTICAL TO INTERLEUKIN-6
B cell stimulatory factor-2 (BSF-2) B cell differentiation factor (BCDF) Interferon-B2(IFN-&) 26-kDa protein Hybridoma/plasmacytomagrowth factor (HPGF) Interleukin hybridoma plasmacytoma 1 (ILHP-1) Plasmacytoma growth factor (PCT-GF) Hepatocyte-stimulating factor (HSF) Macrophage granulocyte-inducing factor 2 ( M G1-2) Cytotoxic T cell differentiation factor (CDF) Thrombopoietin
(Dutton et al., 1971; Schimple and Wecker, 1972; Kishimoto and Ishizaka, 1973). In the early 1980s it was shown that at least two different factors, B cell growth factor (BCGF) and B cell differentiation factor (BCDF), were involved in the regulation of B cell differentiation (Yoshizaki et al., 1982). Since then a variety of factors have been reported to be involved in the regulation of proliferation and differentiation of B cells into antibody-producing cells (Kishimoto, 1985). B cell stimulatory factor 2 (BSF-2) was identified as a factor in the culture supernatants of phytohemagglutinin (PHA-stimulated) (Muraguchi et al., 1981) or antigen-stimulated (Teranishi et al., 1982) peripheral mononuclear cells that induced immunoglobulin (Ig) synthesis in EpsteinBarr virus (EBV)-transformed B cell lines and was originally called BCDF. This molecule, BCDF/BSF-2, was separable from other factors such as IL-2 and BCGF (Yoshizaki et aZ., 1982; Hirano et al., 1984). It was also demonstrated that BSF-2 functions in the late phase of Staphylococcus aureus Cowan I (SAC)-stimulated normal B cells (Hirano et al., 1984)or EBV-transformed cells (Yoshizakiet al., 1982)to induce Ig production, provided other factors such as IL-2 and BCGF are available. Furthermore, BSF-2 was found to act on B cell lines and augment the levels of mRNA and protein of secretory-type Ig (Kikutani et al., 1985). BSF-2 was purified to homogeneity (Hirano et aZ., 1985) from the culture supernatant of a human T cell leukemia virus type I (HTLV1)-transformed T cell line and its partial N-terminal amino acid sequence was determined (Hirano et aZ., 1987).On the basis of the amino
INTERLEUKIN-6 I N BIOLOGY AND MEDICINE
3
acid sequence, the corresponding cDNA was cloned from a T cell line (Hirano et al., 1986). B. HYBRIDOMA/PLA.SMACYTOMA GROWTHFACTOR In 1972, Namba and Hanaoka demonstrated that a murine adherent phagocytic cell line produces a growth factor(s) that promotes the growth of the MOPC 104E plasmacytoma cell line (Namba and Hanaoka, 1972). Growth factors for plasmacytoma were also reported (Metcalf, 1974; Corbel and Melchers, 1984; Nordan and Potter, 1986). A growth factor(s) for murine hybridoma was found in supernatants of human endothelial cells (Astaldi et al., 1980) and human monocytes (Aarden et al., 1985). A hybridoma growth factor designated interleukin hybridoma plasrnacytoma 1 (IL-HPl) (Van Snick et al., 1986)and a molecule termed plamacytoma growth factor (PCT-GF) (Nordan et al., 1987) were purified from a murine helper T cell clone and a murine macrophage cell line P388D1, respectively, and their partial Nterminal amino acid sequences were determined, demonstrating that both growth factors were identical. Human hybridoma/plasmacytoma growth factor (HPGF) was also purified from an osteosarcoma cell line MG-63 (Van Damme et al., 1987b) and peripheral blood monocytes (Brakenhoff et al., 1987). Although the N-terminal amino acid sequence of murine HPGF was found to have no homology with that of human HPGF, molecular cloning of murine IL-HPl demonstrated that murine HPGF has a sequence homology with the human equivalent (Van Snick et al., 1988).
c. INTERFERON-&/26-kDa PROTEIN In 1980, Weissenbach et al. reported that human fibroblasts contain a novel interferon (IFN) mRNA that is inducible by poly(rI).poly(rC) and cycloheximide. This mRNA has a different size (1.3kb) and translation product (26 kDa) from IFN-/3 (Wiessenbach et al., 1980).The corresponding 26-kDa protein was given the name interferon-& (IFN-&) because of an antiviral activity that could be inhibited by antisera against IFN-P. Content et al. (1982) cloned the same mRNA species but concluded that the 26-kDa protein had no antiviral activity and was not serologically related to IFN-p. The 26-kDa protein was shown to be induced in fibroblasts on stimulation with IL-1 (Content et al., 1985; Van Damme et al., 1985).The nucleotide sequences of the cDNAs encoding human IFN-& and the 26-kDa protein were determined and showed the identity of these molecules (Zilberstein et al., 1986; Haegemann et al., 1986).
4
SHIZUO AKIRA ET AL.
D. HEPATOCYTE-STIMULATING FACTOR Acute inflammation is accompanied by changes in many plasma protein concentrations. As the acute-phase proteins are synthesized in the liver and injury to another part of the body results in increased synthesis of acute-phase proteins, the existence of hormone-like mediators was proposed. The isolation and characterization of a regulatory molecule(s) controlling plasma protein biosynthesis have been a major interest in several laboratories for the past two decades. Substances in the leukocyte extracts were found to drastically change the concentration of certain plasma proteins produced by hepatocytes. The leukocyte product was named hepatocyte-stimulating factor (HSF) by several groups. Gauldie et al. (1987) as well as Andus et al. (1987) demonstrated that HSF is identical to the cytokine known as IFN-&, BSF-2, or HPGF. 111. Structure and Expression of Interleukin-6
A. STRUCTURE Interleukin-6 is a glycoprotein with a molecular mass in the range 20 to 30 kDa, depending on the cellular source and preparation. The molecular weight heterogeneity of IL-6 results from post-translational modifications such as N- and 0-linked glycosylation and phosphorylation (May et al., 1988a,b); however, these differences in molecular size do not seem to play a major role in the biological activities of IL-6. cDNAs encoding human, murine, and rat IL-6 have been cloned. Human IL-6 (Hirano et al., 1986)consists of 212 amino acids including a 28-amino-acid signal peptide, whereas mouse (Van Snick et al., 1988) and rat (Northemann et al., 1989) IL-6 consist of 211 amino acids with a 24-amino-acid signal sequence. Comparison of the cDNA sequence of mouse IL-6 with that of human IL-6 shows a homology of 65% at the DNA level and 42% at the amino acid level, although the murine and rat protein sequences are 93% identical. Despite the low amino acid homology between human IL-6 and its murine counterpart, human IL-6 works on murine cells with the same biological activity as mouse IL-6, but not vice versa. There is very little homology in the N-terminal region, but the central portion is more conserved (57% for the region spanning residues 42-102). In particular, the four cysteine residues of the protein, which are located in this region, can be perfectly aligned. The same motif of four cysteines is also found in human and mouse granulocyte colony-stimulating factor (G-CSF) and in chicken myelomonocytic growth factor (MGF), suggesting an evolutionary relation
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
5
between these molecules (Leutz et al., 1989). The two disulfide bridges in mouse IL-6 have been located between Cys 46-Cys 52 and Cys 75-Cys 85 (Simpson et al., 1988). With the exception of the short segment (amino acids 1-28) at the N terminus, the entire primary sequence of human IL-6 is shown to be required for biological activity (Brakenhoff et al., 1989; Kriittgen et al., 1990; Snouwaert et d., 1991). IL-6 is predicted to have a common tertiary fold similar to the four-ahelix bundle structure found in growth hormone, despite little similarity in amino acid sequence (Bazan, 1990a, b) (Fig. 1).The evolutionary relationship between G-CSF and 1L-6 is cemented by a common pattern of exons and introns in their respective genes. The predicted helices (labeled A to D) and loops (two long A-B and C-D loops, one short B-C loop) map to distinctive exon-encoded parts of the aligned chains. Other cytokines including G-CSF, MGF, prolactin (PRL), and erythropoietin (EPO) adopt the similar helical fold-and-loop topology. The region of greatest similarity between these molecules lies at the C-terminal end (helix D) of the alignment of protein chains. Mutagenetic, deletion, and neutralizing antibody studies for PRL, EPO, IL-6, and G-CSF suggest that the surface of predicted helix D is the primary receptor-binding structure.
FIG.1. Schematic ternary structure of interleukin-6. IL-6 protein is predicted to be composed of four (I helices (A through D) and loops connecting them.
6
SHIZUO AKIRA ET AL.
B. INDUCERS AND PRODUCERS Interleukin-6 is produced by many different cell types including monocytes/macrophages, fibroblasts, keratinocytes, endothelial cells, mesangial cells, glial cells, chondrocytes, osteoblasts, smooth muscle cells, T cells, B cells, granulocytes, mast cells, and certain tumor cells (Table 11). Constitutive IL-6 production is reported in a number of tumor cell lines such as cardiac myxoma, cervical carcinoma, renal carcinoma, and bladder carcinoma cells. Except for tumor cells that produce IL-6 constitutively, normal cells do not produce IL-6 unless appropriately stimulated. The production of IL-6 is positively or negatively regulated by a variety of stimuli. Such positive and negative TABLE I1 CELL SOURCES AND INDUCERS OF INTERLEUKIN-6 Cell Type Fibroblasts
Monocyte/macrophage Endothelial cells Keratinocytes Endometrial stromal cells Bone marrow stromal cells B cells T cells Mast cells Neutrophils Osteoblasts Astrocytes Kupffer cells Intestinal epithelial cells Vascular smooth muscle cells Anterior pituitary cells Adrenal gland cells Myocytes
Inducer IL-1, TNF, PDGF, IFN-p, LPS, viruses, serum, poly(I).poly(C), adenyl cyclase activators (forskolin, cholera toxin, isobutylmethylxanthine, dibutylyl CAMP),calcium ionophore A-23187, cycloheximide, PMA, diacylglycerol, prostaglandin E l , sodium fluoride, ouabain LPS, IL-6, IFN-7, PMA, GM-CSF, IL-I, CSF-1, viruses (HIV), adherence, muramyl dipeptide(MDP), C5a LPS, IFNy, IL-I, TNF, IL-4 IL-1, PMA, LPS, Con A, UV irradiation, IL-4 IL-1, TNF, IFN-7 IL-1, IL-6 Staphylococcus aureus Cowan I, IL-4, IL-1, TFN-a PHA + TPA, PHA + monocyte, anti-CD3 antibody, HTLV infection Antigen + IgE, PMA, Con A, calcium ionophore A-23187 GM-CSF, TNF IL-I, TNF, LPS, bradykinin, parathyroid hormone IL-1, TNF, LPS, calcium ionophore A-23187 LPS, IL-1, TNF TGF-/3 IL-1
PMA, LPS, vasoactive intestinal peptide(VIP), IL-1 IL-1, protein kinase C activators, LPS, calcium ionophore A-23187, prostaglandin Ez, forskolin, dibutylyl CAMP, angiotensin 11, ACTH Serum, calcium ionophore, A-23187
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
7
regulation of IL-6 production varies, depending on the cell type. Lipopolysaccharide (LPS) enhances IL-6 production in monocytes and fibroblasts (Helfgott et al., 1987). Various viruses induce IL-6 production in fibroblasts (Sehgal et al., 1988; Van Damme et al., 1989) or in the central nervous system (Frei et al., 1988). A variety of peptide factors, such as IL-1, tumor necrosis factor (TNF), IFN-P, and plateletderived growth factor (PDGF) enhance IL-6 production in fibroblasts and certain tumor cell lines (Kohase et al., 1986; Walther et al., 1988, Van Damme et al., 1987a). Interferon-y (IFN-y) induces IL-6 production by macrophages and endothelial cells (Leeuwenberg et al., 1990; Sanceau et al., 1989). IL-6 does not induce IL-1 or TNF. Rather, IL-6 suppresses endotoxin-induced IL-1 and TNF production (Aderka et al., 1989; Schindler et al., 1990). Moreover, in certain conditions IL-6 is capable of inducing its own production (Shabo et al., 1989; Miyaura et al., 1989). IL-4 is a potent inducer of IL-6 production in normal B cells, keratinocytes, and endothelial cells (Smeland et al., 1989; Howell et al., 1991, Colottia et al., 1991), whereas IL-4 inhibits IL-6 production in monocytes, fibroblasts, and synoviocytes (te Velde et al., 1990; Gibbons et al., 1990; Cheung et al., 1990; Lee et al., 1990; Miossec et al., 1992). Tumor growth factor (TGFP) downregulates IL-6 production by human monocytes (Musso et al., 1990), but it enhances IL-6 production by intestinal epithelial cells (McGee et al., 1992). IL-10 inhibits the production of IL-6 by macrophages (Fiorentino et al., 1991; de Waal Malefyt et al., 1991). Thus, the multiple interactions between cytokines (cytokine network) exist for the regulation of IL-6 production. Second-messenger agonists such as diacylgycerol, phorbol ester, forskolin, isobutylmethy/xanthine, and calcium ionophore A-23187 also stimulate IL-6 gene expression (Sehgal et al., 1987a; Zhang et al., 1988). At least two signal pathways are involved in IL-6 induction by IL-1 or TNF, one requiring protein kinase C (PKC) activation (Sehgal et al., 1987a) and the other involving adenylate cyclase (Zhang et al., 1988); however, these two mechanisms, singly or in combination, cannot completely account for IL-6 induction by TNF or IL-1, suggesting that other, as yet unidentified signal transduction mechanisms also play a role in IL-6 induction by these cytokines. Production of IL-6 can be further superinduced by cycloheximide treatment, suggesting that regulation of production of these cytokines is under the control of labile repressor proteins. Dexamethasone and other glucocorticoids can markedly suppress the production of IL-6 (Helfgott et al., 1987). This suppression is shown to occur at both the transcriptional and the post-transcriptional levels. IL-6 contains re-
8
SHIZUO AKIRA ET AL.
peating AU-rich sequences in its 3' untranslated region, which are commonly observed in the 3'untranslated regions of mRNAs for lymphokines, cytokines, and proto-oncogenes, and thought to be involved in mRNA stability (Shaw and Kamen, 1986).These AU-rich sequences may act to confer cycloheximide superinduction or dexamethasone suppression. 17P-Estradiol also inhibits IL-6 gene expression in endometrial stromal cells (Tabibzadeh et al., 1989).
c. REGULATIONOF INTERLEUKIN-6 GENEEXPRESSION The chromosomal genes for human and murine IL-6 were isolated (Yasukawa et al., 1987; Tanabe et al., 1988).The complete human and mouse IL-6 genes are approximately 5 and 7 kb long, respectively and both consist of five exons and four introns. The gene organization of IL-6 is very similar to that of G-CSF, suggesting that these genes might have evolved from a common ancestor. The genes for human and mouse IL-6 are located on the short arm of chromosome 7 ( 7 ~ 2 1and ) the proximal region of chromosome 5, respectively (Sehgal et al., 1987b; Ferguson-Smith et al., 1988; Mock et al., 1989). There is considerable polymorphism in the human IL-6 gene; three MspI, two BglII, and at least four BstNI alleles that segregate independently have been identified by restriction fragment length polymorphism (Bowcock et al., 1988). Thus, theoretically, there exist up to 24 different IL-6 haplotypes in the human population. The MspI and BglII alleles represent point mutations in introns within the IL-6 gene, whereas the BstNI alleles represent high-frequency insertion/deletion events occurring on the 3' side of the IL-6 gene; however, none of the polymorphisms appears to affect the structure of IL-6 mRNA or its translation product. The region extending -350 bp upstream of the transcriptional start is highly homologous between human and mouse IL-6 genomic genes (Tanabe et al., 1988). Potential transcriptional control elements are identified within the conserved region of the IL-6 promoter, as indicated in Fig. 2. Isshiki et al. (1990) showed that the -180 to -122 region in the IL-6 promoter is involved in IL-1 induction. They also identified a nuclear factor, termed NF-ZL6, binding to a 14-bp nucleotide (ACATTGCACAATCT)within the IL-1-responsive element. Ray et al. (1989) showed that a 23-bp IL-6 multiresponse element (MRE) (-173 to -151)is responsible for induction by IL-1, TNF, and serum as well as by the activators of protein kinase A (forskolin) and protein kinase C (phorbol ester). They also identified several sequence-specific complexes that were increased in intensity in HeLa cell nuclear extracts after stimulation. In this region, a CRE motif as well as an upper half of
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
MRE
-173
NF-IL6
NF-KB
(Bindingrite) - 1 6 -73 (Binding site) -64
GCTAAAGGACGT~CA~GCACAATCT~
GRE GRE
9
GCA"CC4
AP-1 CRENF-IL6 C-fOERCE N F - a TATA c-fos SRE homology homology
FIG.2. Transcriptional regulatory elements identified in the human interleukin-6 promoter. c-fos SRE homology represents a 70%nucleotide sequence identity across a 50-nucleotide stretch of the c-Jos enhancer, including the SRE (serum response element). GRE, glucocorticoid response element; CRE, cyclic AMP response element; AP-1, activation protein 1; RCE, retinoblastoma control element; MRE, multiresponse element.
the 14-bp palindrome are present. An NF-KB binding motif is also present at position -73/-64. Several groups demonstrated that NF-KBis responsible for IL-6 induction (Shimizu et al., 1990; Liberman and Baltimore, 1990; Zhang et al., 1990). Thus, three regulatory regions have been shown to be involved in regulation ofthe IL-6 gene (Fig. 2).
1. NF-IL6 NF-IL6 was initially identified as a nuclear factor binding to a 14-bp palindromic sequence (ACATTGCACAATCT) within an IL-1responsive element in the human IL-6 gene (Isshiki et al., 1990). The gene encoding NF-IL6 was cloned from a Xgtll cDNA expression library of LPS-stimulated human peripheral monocytes by a southwestern method (Akira et aZ., 1990b). The cloned NF-IL6 contained a region highly homologous to the C-terminal portion of C/EBP, the first nuclear factor proposed to contain a leucine zipper structure (Landschulz et aZ., 1988a,b). The highly homologous region includes a basic domain and a leucine zipper structure essential for DNA binding and dimerization, respectively. NF-IL6 recognizes the same nucleotide sequences as C/EBP. Both proteins recognize a variety of divergent nucleotide sequences with different affinities, and the consensus sequence is T(T/G)NNGNNAA(T/G).Expression of these two proteins is, however, quite different. C/EBP is expressed in liver and adipose tissues and is supposed to regulate several hepatocyte- and adipocytespecific genes. By contrast, NF-IL6 is expressed at an undetectable or minor level in all normal tissues, but is drastically induced by stimulation of LPS, IL-1, TNF, or IL-6. NF-IL6 can bind to the regulatory region of various genes including IL-8, G-CSF, IL-1, immunoglobulin,
10
SHIZUO AKIRA ET AL.
and acute-phase protein genes (Natsuka et al., 1992).NF-IL6 has been shown to be identical to IL-GDBP, the DNA-binding protein responsible for IL-6-mediated induction of several acute-phase proteins (Poli et al., 1990b). These results indicate that NF-IL6 may be a pleiotropic mediator of many inducible genes involved in acute-phase, immune and inflammatory responses (Akira and Kishimoto, 1992). 2 . NF-KB
NF-KBwas originally characterized as a kappa immunoglobulin enhancer DNA-binding protein. Binding sites for NF-KBare present in the regulatory regions of certain cytokine genes (including the TNF, lymphotoxin, IL-6, IL-8, and P-IFN genes), the IL-2 receptor gene, class I and I1 histocompatibility antigen genes, several acute-phase response genes, and several viral enhancers including human immunodeficiency virus type 1 (HIV-1) (Lenardo and Baltimore, 1989). NF-KB is a complex of two proteins of 50 and 65 kDa. NF-KB preexists in the cytoplasm of most cells in an inactive form, complexed to IKB. Stimulation by a number of agents such as phorbol ester, LPS, and TNF results in the dissociation of the IKB-NF-KB complex, probably by phosphorylation of IKB. Subsequently the NF-KB heterodimer migrates to the nucleus, where it binds to its cognate DNA binding sites and activates transcription. The genes encoding the p50 and the p65 subunits of NF-KB were cloned and found to be highly homologous to the proto-oncogene c-re1 and the Drosophila maternal effect gene dorsal within a large domain required for DNA binding and dimerization (Kieran et al., 1990; Gohsh et al., 1990; Nolan et al., 1991). D. REPRESSIONOF INTERLEUKIN-6 EXPRESSION It is well known that glucocorticoids have an inhibitory effect on the production of many cytokines including IL-6, TNF, and IL-1. Ray et al. (1991) have investigated the molecular basis for the repression of the IL-6 promoter by the glucocorticoid dexamethasone. The results showed that the activated GR binds to the inducible enhancers (MRE and NF-IL6 binding sites) as well as to the basal transcription regulatory regions (TATA box and RNA start sites) in the IL-6 promoter, and its binding interferes with the binding of positive-acting inducible and basal transcription factors, resulting in the highly efficient repression of this gene by dexamethasone. It has recently been shown that wild-type (wt) p53 or wt RB can repress the IL-6 promoters in serum-induced HeLa cells by CAT assay, suggesting that p53 and RB may be involved as transcriptional repressors in IL-6 gene expression (Santhanam et al., 1991b). p53 and
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
11
RB are considered to be tumor suppressor proteins that are frequently mutated in a variety of neoplasms. Increased IL-6 production is demonstrated in neoplastic cells that exhibit mutations in p53 and RB. The molecular mechanism underlying p53- or RB-mediated repression of the IL-6 promoter awaits further investigation. c-fos repression by RB was shown to be mediated through a cis-acting element in the c-fos promoter called the retinoblastoma control element (RCE)(Robbins et al., 1990).In this respect it is of interest that a region highly homologous to the RCE is also present between positions -126 and -101 in the IL-6 promoter (see Fig. 2). Furthermore, a variety of viral proteins including SV40 large T antigen, adenovirus E1B protein, and human papillomavirus proteins interact avidly with p53 and RB. In cells transformed or infected with adenovirus or SV40, the transforming proteins of these viruses form physical complexes with p53 or RB protein, which may inactivate the normal regulatory function of p53 or RB protein. At the same time, this inactivation of p53 or RB protein may lead to enhanced or dysregulated production of IL-6. IV. Interleukin-6 Receptor
A. INTERLEUKIN-6 RECEPTOR COMPLEX
Human interleukin-6 receptor (IL-6R) cDNA was cloned by use of the COS7 cell expression system (Yamasaki et al., 1988).On the basis of the deduced amino acid sequence, human IL-6R consists of an extracellular region of 339 amino acids, a membrane-spanning region of 28 amino acids, and a cytoplasmic region of 82 amino acids. Human IL-GR as well as its mouse and rat homologs cloned by crosshybridization has a domain of about 90 amino acids, in the amino terminus of the extracellular region, which fulfills the criteria for the constant 2 (C2) set of the immunoglobulin supergene family (Yamasaki et al., 1988; Sugita et al., 1990; Baumann et al., 1990).The remaining extracellular region of IL-6R was found to share structural similarity with subsequently cloned several receptors for, especially, hematopoietic cytokines (Bazan, 1990a,b). This finding has defined a hematopoietic cytokine receptor family that includes such receptors as IL-2R (p and y chains), IL-3R, IL4R, IL-5R, IL-GR, IL-7R7 IL-9R, EPO-R, G-CSF-R, GM-CSF-R, and LIF-R (Bazan, 1990a,b; Taga and Kishimoto, 1992; and references therein). They share an approximately 200-amino-acid-homologous module, which is characterized by four conserved cysteine residues distributed in the amino-terminal half and (WSXWS in one-letter symbols) motif located a Tip-Ser-X-Trp-Ser
12
SHIZUO AKIRA ET AL.
at the carboxyl-terminal end. This homologous module comprises two tandemly placed fibronectin type I11 domains which often appear in proteins functioning in cell-to-cell adhesion, implying that the members of the hematopoietic cytokine receptor family might have emerged from a cell surface molecule that could have been involved in communication between cells via direct contact (Bazan, 1990b; Patthy, 1990). Each of these two fibronectin type I11 domains is composed of seven folds ofp strands positioned antiparallely so as to form a “barrellike” shape (Fig. 3). A trough formed between two “barrels” is believed to function as a ligand-binding pocket. In fact, in uitro-mutated IL-6R protein lacking the immunoglobulin-like domain but not the
0
immunoglobulinlike domain
0
type 111 domain tibronectin
U
FIG.3. Structure ofthe interleukin-6 receptor and gp130. The extracellular region of IL-6R comprises the immunoglobulin-like domain and the cytokine receptor family module which is composed oftwo fibronectin type 111 domains. The extracellular region of gp130 comprises six fibronectin type 111 domains, the second and the third of which constitute the cytokine receptor family module. The schematic ternary structure of the cytokine receptor family module is depicted on the left. Amino acid residues are indicated in one-letter codes.
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
13
cytokine receptor family module in its extracellular region retains IL-6 binding capability, and according to the predicted ternary structure of the cytokine receptor family module, many of the amino acid residues critical for IL-6 binding are distributed to the hinge region between the two barrel-like fibronectin type I11 domains (Yawata et al., 1993). The WSXWS motif characteristic of the family is predicted to be in this hinge region. All the IL-6R mutants possessing amino acid substitutions in either of the four cysteine residues or the WSXWS motif conserved in the family lack IL-6 binding capability. The cytoplasmic region of IL-6R is relatively short and has no obvious catalytic domain, and deletion of this region did not affect the cellular responsiveness to IL-6. This suggested a receptor component that is associated with IL-6R and is responsible for signal transduction. Such a molecule, now called gp130 because it is a 130-kDa glycoprotein, was discovered by immunoprecipitation of IL-6R protein; gp130 was co-immunoprecipitated with 1L-6R from digitonin lysates of IL-6-stimulated cells (Taga et al., 1989). The IL-6 triggered association of IL-6R and gp130 takes place at 37°C within 5 minutes but not at WC, suggesting a requirement for membrane fluidity in this step. The association of the two membrane proteins is revealed to occur between their extracellular regions, because the extracellular soluble form of IL-6R associates with gp130 in the presence of IL-6. Based on the amino acid sequence deduced from the cDNA cloned by immunoscreening of an Escherichia coli expression library, human gp130 is predicted to consist of an extracellular region of 597 amino acids, a membrane-spanning region of 22 amino acids, and a cytoplasmic region of 277 amino acids (Hibi et al., 1990).The entire extracellular region of human as well as subsequently cloned mouse gp130 comprises six repeats of the fibronectin type I11 domain. Among these repeats, the second and third type I11 domains compose the hematopoietic cytokine receptor family module, possessing the characteristic four cysteine residues in the former and the WSXWS motif in the latter (see Fig. 3). It has been demonstrated that despite its lack of IL-6 binding property, gp130 is involved in the formation of high-affinity binding sites and is critical for IL-6 signal transduction (Hibi et al., 1990; Taga et al., 1992): (1) cDNA-expressed IL-6R shows only lowaffinity IL-6 binding property, but coexpression of IL6-R with gp130 confers both high- and low-affinity binding sites. (2)Anti-gpl30 monoclonal antibody blocks the formation of high- but not low-affinity IL-6 binding sites and also inhibits IL-6-induced biological responses. (3)The IL-WsIL-GR complex induces DNA synthesis in a transfectant expressing gp130, but not in its parental pro-B cell line. In conclusion,
14
SHIZUO AKIFU ET AL.
the IL-6R complex comprises two functionally different membrane proteins: a ligand-binding chain (80-kDa IL-6R) and a nonbinding but signal-transducing chain (gp130). When IL-6R is occupied by IL-6, these two chains associate extracellularly to form the high-affinity functional receDtor. The signal-transducing chain of the IL-6R complex, that is, gp130, has been shown to be expressed in nearly all human and mouse cell lines examined (Hibi et al., 1990; Saito et al., 1992). In the case of the mouse, such tissues as brain, thymus, heart, lung, spleen, liver, and kidney are shown to express gp130. As for expression of IL-6R, although it is also widely distributed in accordance with the pleiotropic nature of IL-6, it does not seem to be as ubiquitous as gp130. In mouse and human peripheral blood mononuclear cells, monocytes and T cells (both CD4’ and CD8+ subpopulations) express IL-GR as detected by flow cytometric analysis and binding assay (Taga et al., 1987; Coulie et al., 1989; Bauer et al., 1989; Hirata et al., 1989). IL-6R is barely detectable on freshly isolated mouse thymocytes, but appears after thymocytes are cultured in vitro for 2 days (Coulie et al., 1989; Kobayashi et al., 1992). B cells do not usually express IL-6R, but they come to express IL-6R when stimulated in vitro with mitogen (Taga et al., 1987; Hirata et al., 1989). Peyer’s patch B cells, many of which are class-switched (to IgA) and considered to be at a more differentiated stage, express IL-6R before in vitro activation (Fujihashi et al., 1991; Kobayashi et al., 1992). These findings may reflect the functional difference of IL-6 in these cells: IL-6 induces (1)differentiation of macrophages (Shabo et al., 1988; Miyaura et al., 1988), (2) production of immunoglobulins in activated but not resting B cells (Muraguchi et al., 1988; Fujihashi et al., 1991), and (3)proliferation of peripheral T cells or mitogen-cocultured thymocytes (Garman et al., 1987; Lotz et al., 1988). In inflammatory contexts, hepatocytes produce acute-phase proteins, the process of which involves IL-6. IL-GR transcripts expressed in hepatocytes are increased when the mice are inoculated with complete Freund’s adjuvant or turpentine (Baumann et al., 1990; Nesbitt and Fuller, 1992). Under inflammatory conditions, in addition to IL-6, glucocorticoid is upregulated as a consequence of activation of the pituitary-adrenal axis. Recombinant IL-6 as well as synthetic glucocorticoid, dexamethasone, enhances IL-6R expression in hepatocytes and hepatoma cells (Bauer et al., 1989; Nesbitt and Fuller, 1992; Rose-John et al., 1990; Saito et al., 1992). As for the level of gp130 transcripts in hepatocytes, although it is indeed upregulated by the addition of turpentine, IL-6, or dexamethasone, the extent of upregulation is relatively small, or gp130 expression remains rather stable,
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
15
compared with the IL-6R levels (Nesbitt and Fuller, 1992; Schooltinck et al., 1992; Snyers and Content, 1992). These results suggest that under inflammatory conditions, hepatocytes become more sensitive to IL-6 to produce acute-phase proteins by upregulating the levels of receptor components, especially that of IL-6R rather than gp130. The preceding observations of regulated expression of IL-6R and ubiquitous expression of gp130 suggest that the functions of the pleiotropic cytokine IL-6 may be controlled by expression of the ligand-binding chain but not the signal-transducing chain of the receptor complex. In this context, a case of dysregulated expression found in a plasmacytoma cell line has to be noted. In the P3U1 plasmacytoma cell line the cDNA encoding rearranged IL-6R was isolated; a DNA segment corresponding to the cytoplasmic region was replaced with part of the long terminal repeat (LTR)of the intracysternal A-particle gene (Sugita et al., 1990).The rearranged mouse IL-6R retains the ability to mediate IL-6 signals, and this rearranged, but functionally normal, IL-6R is overexpressed in P3U1 cells, probably because of the internal enhancer in the LTR sequence. As IL-6 is a potent growth factor for plasmacytomas, this suggests that overexpression of rearranged IL-6R might be responsible for the development of this particular plasmacytoma cell line. In MRL/lpr autoimmune mice, splenic B cells express abnormally high levels of IL-6R without any in uitro stimulation, suggesting its contribution to B cell hyperreactivity in this strain (Kobayashi et al., 1992). B. SIGNALS THROUGH gp130, WHICHIs S H A R E D BY SEVERAL CYTOKINES Since the finding of the IL-6R-associated signal transducer gp130, it has been hypothesized that functional redundancy, a characteristic feature of the actions of many cytokines, could be explained if several different cytokine receptors were to interact with a common signaltransducing component, such as gp130. Leukemia-inhibitory factor (LIF)and oncostatin M (OM)were initially identified as growth inhibitors for a mouse myeloid leukemia cell line and human melanoma cell line, respectively. LIF and OM are multifunctional cytokines whose biological functions overlap each other and those of IL6, for example, induction of acute-phase protein synthesis in hepatocytes and macrophage differentiation of M 1 cells (Hilton and Gough, 1991; Rose and Bruce, 1991; Kishimoto et al., 1992). Although LIF-responding cells express both high- and low-affinity LIF binding sites, cDNAexpressed LIF-R protein on COS cells shows only low-affinity LIF binding property, suggesting an additional high-affinity converting
16
SHIZUO AKIRA ET AL.
subunit of the LIF-R complex (Gearing et al., 1991). This converter was later revealed to be identical to the IL-6 signal transducer, gp130; coexpression of the cDNAs for LIF-R and gp130 was shown to generate high- and low-affinity LIF binding sites (Gearing et al., 1992). In addition, although gp130 or LIF-R binds OM with low or null, respectively, intrinsic affinity, coexpression of the two proteins confers intermediate-affinity OM binding sites, suggesting gp130 and LIF-R also associate to form an OM-R complex; however, because some OM-responsive cell lines that express high- and low-affinity OM binding sites do not express LIF-R, an unidentified receptor component for the OM-R complex is also suggested (Gearing and Bruce, 1992). Antigp130 monoclonal antibodies completely block the acute-phase protein synthesis in hepatoma cells induced by either IL-6, LIF, or OM, and the growth inhibition of melanoma cells induced by either IL-6 or OM (Liu et al., 1992; Taga et al., 1992). Furthermore, any of these cytokines rapidly induces tyrosine phosphorylation of gp130 (Ip et al., 1992; Taga et al., 1992). The results confirmed that gp130 is critical for the signaling process triggered by these three cytokines. Among the three cytokines, LIF also shows pleiotropic functions within the neural system which are overlapped by the functions of ciliary neurotrophic factor (CNTF), for example, promotion of the survival of sensory and motor neurons and induction of switch from adrenergic to cholinergic phenotype in cultured neurons (Yamamori et al., 1989; Stockli et al., 1989; Kishimoto et al., 1992). Based on the information from the cloned cDNA, CNTF-R shows the highest sequence homology to IL-6R. It lacks transmembrane and cytoplasmic regions, but is instead anchored to the membrane via a glycosylphosphatidylinositol (GPI)linkage (Davis et al., 1991).These observations suggest that gp130 may also be associated with CNTF-R. CNTF actions on neuronal cells were completely blocked by anti-gpl30 antibodies, indicating that CNTF signaling processes involve gp130 (Ip et al., 1992; Taga et al., 1992). CNTF stimulation rapidly induces tyrosine phosphorylation of both gp130 and gpl30-associated 190-kDa protein. This 190-kDa protein is most likely LIF-R, suggesting that the functional CNTF-R complex may include LIF-R in addition to gp130 and CNTF-R (Fig. 4). Interleukin-11 exerts multiple biological functions similar to those of IL-6, implying that similar signaling processes may be operating in the IL-6 and IL-11 systems (Baumann and Schendel, 1992).Although a specific receptor for IL-11 has not yet been molecularly cloned, anti-gp 130 monoclonal antibodies (mAbs) inhibited IL-ll-induced TF-1 cell proliferation, suggesting that gp130 is essential for 1L-11 signal trans-
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
17
FIG.4. Receptor complexes for interleukin-6 (A), CNTF (B), LIF (C),and OM (D).A signal transducing component, gp130, is shared by these receptor complexes and is essential for initiating their respective cytoplasmic signaling processes. Cytokine stimulation induces oligomerization of the receptor components which is postulated to stabilize interaction with a downstream molecule. See text for details.
duction and may be a component of the IL-11R complex (Y.-C. Yang et al., unpublished). It should be noted that although the ubiquitously expressed gp130 is involved in mediating signals elicited by all the previously mentioned cytokines, the ability of a cell to respond to each of these factors appears to be regulated by the specific expression of distinct receptor chains. This may explain why these cytokines with overlapping biological functions do show their own specific activities as well.
18
SHIZUO AKIRA ET AL.
No enzymatic activities or sequence motifs known for signal transduction have been found in gp130 protein; however, a series of studies have provided clues to understanding how gp130 initiates a cascade of cytoplasmic signals. Stimulation of cells by either IL-6, LIF, OM, CNTF, or IL-11 has been shown to induce tyrosine-specific phosphorylation of cellular proteins including gp130 (Nakajima and Wall, 1991; Murakami et al., 1991; Yin et al., 1992; Schieven et al., 1992; Taga et al., 1992).The addition of tyrosine kinase inhibitors blocks cellular responses induced by IL-6 (Nakajimaand Wall, 1991).Furthermore, stimulated gp130 has been shown to be associated with tyrosine kinase activity, suggesting that tyrosine kinase may be associated with a130 on stimulation. Stimulation of the cells with IL-6 induces the formation of gp130 homodimers (M. Murakami et al., unpublished). Thus, dimerization of gp130 may form divalent contact surfaces important for stable interaction with a cytoplasmic molecule such as a tyrosine kinase (see Fig. 4). At present, it remains unclear exactly what sort of tyrosine kinase interacts with gp130. As gp130 has been shown to be essential for transduction of the IL-6, LIF, OM, CNTF, and IL-11 signals, it is necessary to examine whether the gp130 molecule in each of the receptor complexes for these cytokines interacts with the same or a different kinase. In the 277-amino-acid cytoplasmic region of gp130, a -60-amino-acid portion, proximal to the transmembrane domain, was shown to be essential for at least IL-6-induced DNA synthesis in a mouse pro-B cell line (Murakami et al., 1991). In this critical cytoplasmic region, two short stretches of amino acids exist that are highly conserved among many receptors or signal transducers belonging to the hematopoietic cytokine receptor family, suggesting that a similar mechanism might be involved in the signaling processes of various cytokines. A candidate signaling process that operates downstream of the tyrosine kinase step is protein serinehhreonine phosphorylation. Serine/threonine kinase inhibitors block IL-6-mediated cellular responses without affecting IL-6-induced tyrosine phosphorylation (Nakajima and Wall, 1991). Another candidate molecule functioning downstream of the putative tyrosine kinase is the p21 Ras protein, In PC12 cells, formation of GTP-bound Ras is considered to be essential for their neural differentiation (Szeberenyi et al., 1990). The GTPbound form of p21 Ras (activated Ras) is upregulated on stimulation of PC12 cells by IL-6. A tyrosine kinase inhibitor was observed to block the IL-6-induced formation of GTP-bound Ras (Nakafuku et al., 1992). With respect to the involvement of serinelthreonine kinase and Ras in IL-6 signal transduction, a recent finding on a nuclear factor, NF-IL6, has provided a clue to the better understanding of the pathway from
INTERLEUKIN-6 I N BIOLOGY AND MEDICINE
19
receptor to nuclear factor, an ultimate target of the IL-6 signal. NF-IL6 was originally identified as a sequence-specific nuclear factor involved in IL-l-induced IL-6 gene expression (Akira et at., 1990b).NF-IL6 was subsequently found to interact with an IL-6-responsive element in the promoters of acute-phase protein genes whose expression in liver is regulated by IL-6. Thus, IL-6-induced activation of acute-phase genes in hepatocytes may be controlled by modification of NF-IL6. Serineand threonine-specific phosphorylation of NF-IL6 has been demonstrated to be critical for its activation. Involvement of a serine/ threonine kinase such as MAP kinase in this phosphorylation process has also been suggested. Transfection of the oncogenic rus gene enhances the phosphorylation and transactivating function of NF-IL6 (Nakajima et al., 1993). These results imply that the IL-6-induced activation of gpl30-associated tyrosine kinase may lead to the activation of Ras and, subsequently, MAP kinase, resulting in the functional activation of NF-IL6.
c. SOLUBLE FORMS OF INTERLEUKIN-6 RECEPTORCOMPONENTS Naturally produced soluble IL-6R protein, which retains its ability to bind IL-6, was observed first in human urine. ILS-binding protein, whose partial amino acid sequence was identical to that of previously cloned IL-6R, was purified by IL-6-coupled column chromatography (Novick et al., 1989). In human and mouse serum, the presence of natural soluble IL-6R, which is capable of binding IL-6 and mediating signals via membrane-anchored gp130, has been demonstrated (Novick et d . , 1991; Honda et d . , 1992; Suzuki et d . ,submitted). This implies a possible physiological role for serum-soluble IL-6R in modulating IL-6 functions. Although the mechanism of production of soluble IL-6R has not fully been elucidated, identification of mRNA encoding the transmembrane domainless form, possibly created by alternatively splicing, was reported (Lust et aZ., 1992). Autoimmuneprone MRL/lpr mice appear to produce serum-soluble IL-6R at a significantly high concentration, compared with normal mice. In addition, an age-associated increase in serum-soluble IL-6R levels is observed in mice of this strain (Suzuki et al., submitted). In MRL/lpr mice, the serum level of IL-6 also increases according to age (Tang et al., 1991).Taken together, these elevations might be involved in the pathogenesis of autoimmune symptoms in MRL/lpr mice. A soluble forms of the IL-6 signal-transducing receptor component, gp130 is also found in human serum. Serum-soluble gp130 is shown to negatively regulate the IL-6 signal, suggesting its physiological role (M. Narazaki et al., submitted).
20
SHIZUO AKIRA ET AL.
V. Biological Function of Interleukin-6
A. GROWTH REGULATORY FUNCTIONS IL-6 enhances, inhibits, or has no effect on cell proliferation depending on the cell type.
1 . Growth-Stirnulatory Effect Interleukin-6 promotes growth and may be an autocrine growth factor in a number of plasmacytomas and myelomas (Vink et al., 1990; Kawano et al., 1988), EBV-transformed B lymphocytes (Tosato et al., 1990; Yokoi et al., 1990), several T and B lymphomas (Shimizu et al., 1988; Yee et al., 1989), mesangial cells (Horii et al., 1989), vascular smooth muscle cells (Nabata et al., 1990), Kaposi sarcomaderived cells (Miles et al., 1990), renal carcinoma (Miki et al., 1989), Pagetic osteoclasts (Roodman et al., 1992) and psoriatic keratinocytes (Grossman et al., 1989). 2. Growth-lnhibitory Effect Dose-dependent growth inhibition of IL-6 is observed in a number of human breast carcinoma cell lines including ductal carcinomas and adenocarcinomas (Chen et al., 1988). IL-6 also affects the cell morphology and behavior of breast cancer cells: colonies lose epithelial morphology and contain dispersed fusoid cells, with few contacts, increased motility, loss of cytoskeletal organization, and loss of desmosomes (Tam et al., 1989). Although IL-6 stimulates growth of plasmacytomas, myelomas, and several B lymphomas as described, it can also inhibit growth of certain chronic B leukemic cells and B lymphomas (Chen et al., 1988). Thus, IL-6 acts as a positive and negative regulator of B lymphocyte growth. M12 murine B lymphoma cells, which lack the IL-6 receptor but have the gp130 transducer, became growth inhibited by IL-6 on transfection by the IL-6 receptor gene (Taga et al., 1989). IL-6 inhibits the growth of certain myeloid leukemic cells and induces the differentiation of these cells into mature macrophage-like cells (Shabo et al., 1988; Miyaura et al., 1988; Lotem et al., 1989). Inhibition of acute myelogeneous leukemia (AML) development by IL-6 treatment was observed (Givon et al., 1992), IL-6 inhibits the proliferation of bone marrow and tissue macrophages (Riedy and Stewart, 1992). IL-6 also inhibits melanocyte proliferation (Morinage et al., 1989) and the growth of early-stage melanoma cells (Lu et al., 1992). IL-6 has been reported to inhibit the proliferation of endothelial cells and diploid fibroblasts under particular experimental conditions (Kohase et al., 1986).
INTERLEUKIN-6 I N BIOLOGY AND MEDICINE
21
B. IMMUNE REGULATION 1. Effects on B Cells Interleukin-6 was first recognized as a T cell-derived factor acting on B cells to induce immunoglobulin (Ig) secretion. IL-6 acts mainly on the late phase of the B cell differentiation pathway, consistent with the finding that IL-6R is expressed on activated B cells but not resting B cells (Taga et al., 1987). IL-6 acts on mitogen-activated B cells to induce IgM, IgG, and IgA production without stimulating B cell proliferation (Muraguchi et al., 1988; Beagley et al., 1989). In this case, IL-6 shows no differential effects among IgM-, IgG-, and IgA-committed B cells, in contrast to other interleukins, such as IL-4 and IL-5, which display preferential effects on IgE and IgA secretion. Anti-IL-6 antibody completely inhibits Ig production. The essential role of IL-6 is also demonstrated in polysaccharide-specific Ig production (Ambrosino et al., 1990), IL-4-dependent IgE response (Vercelli et al., 1989), tetanus toxoid-specific Ig production (Brieva et al., 1990), and influenza A virus-specific primary response (Hilbert et al., 1989). In uiuo antigen-stimulated lymphoblastoid B cells responded well to IL-6 and differentiated into antibody-producing cells in uitro (Lue et al., 1991). IL-6 enhanced the in uiuo secondary anti-SRBC antibody production in mice (Takatsuki et al., 1988). These experimental results demonstrate that IL-6 functions as a B cell differentiation factor in uitro as well as in uiuo. Interleukin-6 is also a potent growth factor for hybridoma/ plasmacytoma/myeloma cells and only 2 pg/ml rIL-6 could induce 50% of the maximal proliferation in myeloma cell lines (Van Damme et al., 1987b; Aarden et al., 1987; Nordan et al., 1987). This concentration of IL-6 is 100-fold less than that required for Ig induction in B cells. Therefore, IL-6-dependent B cell hybridoma lines, such as B9 and 7TD1, provide us with an extremely specific and sensitive bioassay for IL-6. IL-6 is found to increase the frequency of development of hybridomas producing monoclonal antibodies and to augment cloning efficiency; therefore, IL-6 is now used to establish hybridoma cell lines (Matsuda et al.,1988; Harris et al., 1992). IL-6 also promotes the proliferation of EBV-infected B cells and permits their growth at low cell densities (Tosato et al., 1988).
2. Effect on T Cells Interleukin-6 is involved in T cell activation, growth, and differentiation. IL-6 stimulates the proliferation of peripheral T cells and mature thymocytes activated with lectins or anti-T cell receptor mono-
22
SHIZUO AKIRA ET AL.
clonal antibodies (Garman et al., 1987; Lotz et al., 1988; Uyttenhove et al., 1988).IL-6 induces not only proliferation but also differentiation of cytotoxic T lymphocytes (CTLs) (Okada et al., 1988; Takai et al., 1988, Renauld et al., 1989). Anti-IL-6 antibodies completely block cytolytic and proliferative T cell responses, supporting the importance of IL-6 in the growth and differentiation of T cells. In most cases, the effect of IL-6 is evident in the presence of other stimuli, including IL-1, TNF, and PHA. IL-6 and IL-1 synergize for T cell proliferation (Houssiau et al., 1988a). IL-1 induces IL-6 production and increases the sensitivity to IL-6 (Helle et al., 1988). The involvement of IL-2 in IL-6 (or IL-1 + IL-6)-mediated T cell proliferation and differentiation was demonstrated by the observation that T cells stimulated with IL-6 produced IL-2 and that anti-IL-2 receptor mAb completely inhibited IL-6 (or IL-1 + IL6)-induced murine T cell proliferation and differentiation. These results suggests that IL-1 and IL-6 could exert their T cell growth activity by upregulating the production of and the response to IL-2. Part of the synergistic interaction between IL-6 and IL-1 seems to result from the mechanism by which IL-6 converts T cells to an IL-2-responsive state b y the transition from Go to an early stage in GI and the induction of IL-2 receptor (Tac antigen) expression (Noma et al., 1987; Le et al., 1988), whereas IL-1 and IL-6 both act on IL-2 production (Garman et al., 1987; Houssiau et al., 1988a). In primary CD4+ or CD8+ T cells, T cell receptor (TCR) crosslinking by anti-CD3 mAb induces both IL-2 receptor expression and responsiveness to exogenous IL-2, but is not sufficient to induce either IL-2 secretion or T cell proliferation. IL-2 secretion and proliferation require both TCR crosslinking and antigen presenting cell (APC)-derived costimulatory signals. It is demonstrated that either IL-1 or IL-6 can replace the requirement for APC-derived costimulatory signals for IL-2 secretion and proliferation (Kasahara et al., 1990; Lorre et aZ., 1990). IL-6 also augments the activity of human natural killer cells (Luger et al., 1989). C. HEMATOPOIESIS 1 . Effects on Hematopoietic Progenitor Cells Hematopoiesis is regulated by a variety of growth- and differentiation-inducing factors. In the steady state, the majority of hematopoietic stem cells are dormant and reside in the Go phase of the cell cycle. Ikebuchi et al. (1987) showed that IL-6 acted synergistically with IL-3 in uitro to hasten the appearance of multilineage blast cell colonies grown from murine spleen cells. A similar synergy
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
23
between IL-6 and IL-3 was shown using purified human bone marrow progenitors. In this case, IL-6 induced a Go-to-G1 progression of hematopoietic stem cells, whereas IL-3 did not trigger their emergence from the Go phase, but was necessary for the proliferation of these cells. Continuous perfusion of IL-6 into normal mice increased splenic CFU-s numbers (Suzuki et al., 1989). Bone marrow transplanted mice that were subsequently treated with 1L-6 exhibited both enhanced hematopoietic repopulation and enhanced survival (Okano et d., 1989). IL-6 or a combination of IL-3 and IL-6 given to mice with radiation-induced hematopoietic suppression was shown to facilitate multilineage recovery. Rennick et al. (1989) demonstrated the ability of IL-6 to interact with IL-4, G-CSF, M-CSF, and GM-CSF to selectively enhance the clonal growth of progenitor cells at specific stages of lineage commitment and maturation. IL-1, IL-6, and KL had limited ability to stimulate the proliferation of murine hematopoietic progenitor cells. IL-6 and IL-3 were able to stimulate an immature population of progenitor cells, and IL-6 was shown to increase the number of high-proliferative-potential colonyforming cells (HPP-CFC) stimulated by IL-3, IL-4, G-CSF, M-CSF, and GM-CSF from d4 5-FU BM cells. When used alone, IL-6 was shown to directly support the in uitro proliferation of murine GM progenitors, as well as to directly promote megakaryocyte maturation in uitro. Much smaller amounts of IL-6 induced a neutrophilia, a slight lymphopenia, and a reticulocytosis (Ulich et al., 1989). Interleukin-6 may be useful in the application of retroviral gene transfer methods to human cells. Retroviral-directed gene integration requires active cell cycle and reconstitution requires maintenance of self-renewal capacity in the donor cell population. The combination of IL-3 and IL-6 or Steel factor and IL-6 has been shown to improve the efficiency of retroviral-mediated gene transfer into reconstituting hematopoietic stem cells (Bodine et al., 1989; Dick et al., 1991; Apperley et al., 1991; Luskey et al., 1992).
2. Effects on Megakaryocytes Human megakaryocytopoiesis is a complex phenomenon that includes proliferation of committed megakaryocytic progenitor cells, and cellular maturation comprising nuclear polyploidization, growth in size, and generation of cytoplasmic lineage markers. IL-3, GMCSF, erythropoietin, and Steel factor are able to induce proliferation and differentiation of the committed progenitors, but none of these factors has been shown to be specific for the megakaryocyte lineage. In contrast, late stages, including polyploidization and maturation,
24
SHIZUO AKIRA ET AL.
were under the control of a specific factor(s) called thrombopoietin. IL-6 has been found to function as thrombopoietin (Ishibashi et al., 1989a,b; Asano et al., 1990; Hill et aZ., 1990). IL-6 induces not only the in uitro maturation of megakaryocytes (increase in ploidization, size, and acetylcholine esterase activity) but also in in uiuo increase in platelet counts in mice and monkeys. Transgenic mice carrying human IL-6 are shown to have an increased number of megakaryocytes in their marrow (Suematsu et al., 1989). IL-6 and IL-6R are constitutively expressed by human megakaryocytes, suggesting that normal human megakaryocytopoiesis might be regulated in part by an autocrine loop (Navarro et al., 1991). Whether or not IL-6 is the physiological regulator of thrombopoiesis is, however, controversial (Hill et al., 1992b; Straneva et al., 1992). Reactive or secondary thrombocytosis is observed in various conditions such as inflammation, following surgery, or in trauma and malignancy, accompanied by increased levels of serum IL-6. On the other hand, no elevation of serum IL-6 levels has been observed in patients with myeloproliferative disorders and idiopathic thrombocytopenic purpura. Administration of anti-IL-6 antibody does not cause thrombocytopenia in uiuo. Furthermore, decreased numbers of circulating platelets are not associated with increased levels of serum IL-6, although increased IL-6 levels are more closely related with the existence of ongoing inflammatory processes. Therefore, IL-6 may not be required for steady-state thrombopoiesis, but may play a role in situations of hematopoietic stress. In any case, the effects of IL-6 on thrombocytes are of potential clinical importance for the treatment of thrombocytopenia. 3. Effects on Macrophage Differentiation Human and mouse myeloid leukemic cell lines can be induced to differentiate into macrophages in uitro by several factors including G-CSF, macrophage-granulocyte-inducing factor 2 (MGI-2), and leukemia-inhibitory factor (LIF). IL-6 was shown to be identical to MGI-2, which could induce the differentiation of a murine myeloid leukemia cell line, M 1 (Shabo et d.,1988; Miyaura et al., 1988; Chiu et al., 1989; Lotem et al., 1989). IL-6 inhibits the growth of human U937 and murine M 1 myeloid leukemic cell lines, and induces the differentiation of these cells into mature macrophage-like cells. At the same time, IL-6 enhances phagocytosis and expression of a number of macrophage differentiation antigens including Mac-1 and Mac-3 and yFc receptors, major histocompatibility complex class I, nonspecific esterases, lysozyme, 2',5'-A-synthetase, c-fms (M-CSF re-
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
25
ceptor), indicating the functional differentiation into mature macrophages. In human myeloid cell line U937, IL-6 has some effects alone, which can be enhanced by association with other cytokines such as IFN-y, IL-1, LIF, G-CSF, and GM-CSF. HL-60 cell growth is reduced by IL-6 with GM-CSF and differentiation is seen with TNF and IFN-y. Shabo et al. (1989) proposed the hypothesis that IL-6 and GM-CSF are autocrine differentiating factors which are synthesized in myeloid cells as a second messenger of other growth factors, based on the fact that all CSFs including IL-3 can induce the production of IL-6 and GM-CSF in normal myeloid precursors.
D. ACUTE-PHASE PROTEIN SYNTHESIS IN HEPATOCYTES Inflammation is accompanied by the acute-phase response which is characterized by significant alterations in the serum levels of several plasma proteins, known as acute-phase proteins (APPs) (Kushner, 1982; Koj, 1985). APP production is reflected in the increase in erythrocyte sedimentation rate, which is used as a cursory indicator of inflammation in humans. The acute-phase response is well preserved throughout phylogeny and is considered to serve host defense systems by protecting the generalized tissue destruction associated with inflammation. In fact, many APPs are antiproteinases, opsonins, or blood-clotting and wound-healing factors. Acute-phase proteins are synthesized mainly by the liver. Both an increase and a decrease in synthesis of APPs are seen in the acutephase response. Concentrations of several plasma proteins increase dramatically. For example, more than 1000-fold increased levels of both C-reactive protein (CRP) and serum amyloid A (SAA) are observed in sera of severely infected individuals. Other plasma proteins increase moderately (e.g., fibrinogen, a1AT, complement proteins factor B and C3). In contrast, there is a decrease in several plasma proteins such as albumin, transferrin, cyz-globulin, and transthyretin. The biosynthesis of acute-phase proteins by hepatocytes is regulated by the hepatocyte-stimulating factor (HSF). It has been shown that HSF is indentical to IL-6 (Gauldie et al., 1987; Andus et al., 1987). rIL-6 induced synthesis and secretion of a wide spectrum of acute-phase proteins from primary hepatocytes and hepatocyte cell lines, including SAA, CRP, haptoglobin, al-antitrypsin, al-acid glycoprotein, and al-antichymotrypsin, whereas albumin and transferrin were decreased. Induction of CRP, SAA, and fibrinogen by conditioned medium was completely inhibited by antibodies to rIL-6. In vivo administration of IL-6 in rats induced a typical acutephase reaction similar to that induced by turpentine (Geiger et al.,
26
SHIZUO AKIRA ET AL.
1988). An IL-6-mediated increase in serum acute-phase proteins was demonstrated in the monkey (Asano et al., 1990). It was also reported that serum levels of IL-6 correlated well with those of CRP and fever in patients with severe burns (Nijsten et al., 1987), and an increase in serum IL-6 was observed before an increase in serum CRP in patients undergoing surgery (Nishimoto et d., 1989; Shenkin et d., 1989). The results confirmed the in uiuo effect of IL-6 in the acutephase reaction. Besides IL-6, several cytokines have been found capable of directly inducing acute-phase proteins from the liver: IL-1, TNF-a, IL-11, LIF, TGF-P, and oncostatin M. IL-1/TNF and IL-6 act on some genes (e.g., al-acid glycoprotein) in a synergistic manner and on other genes (e.g., fibrinogen) in an additive or negative manner. LIF was discovered because of its ability to induce terminal differentiation of M 1 myeloid leukemia cells (Gearing et al., 1987) and inhibit differentiation of embryonic stem cells (Williams et al., 1988), and is now found to be a multifunctional cytokine like IL-6. Oncostatin M is a cytokine expressed in activated human T lymphocytes and monocytes, orginally identified as a growth regulator for certain tumor cell lines (Zarling et al., 1986). LIF and oncostatin M regulate the same set of genes as does IL-6 (Baumann and Wong, 1989; Richard et al., 1992). In this regard, it is noteworthy that the highaffinity receptors for LIF and oncostatin M share gp130, a signal transducer of IL-6 (Gearing et al., 1992). IL-11, the recently discovered factor (Paul et al., 1990), is shown to exert IL-6-like effects on liver cells (Baumann and Schendel, 1992). TGF-P can affect hepatic synthesis and secretion of a subset of acute-phase proteins, both directly and by modulating the effect of IL-6 (Mackiewicz et al., 1990; Morrone et al., 1989). The affected group of plasma proteins is distinct from those affected by IL-6. It has been shown that glucocorticoids potentiate the effect of cytokines on induction of some, but not all, human acute-phase proteins, although they have no stimulatory action on their own. Aside from its role as an inducer of acute-phase proteins, IL-6 acts on the central nervous system to induce fever and to elicit the release of adrenocorticotropic hormone (ACTH) (LeMay et al., 1990; Naitoh et al., 1988).ACTH in turn increases the synthesis of glucocorticoids in adrenal glands. Elevated levels of circulating glucocorticoids synergize in IL-6 in inducing the increased hepatic synthesis and secretion of acute-phase proteins. Glucocorticoids also upregulate high-affinity IL-6 receptors and gp130 on hepatocytes and thereby augment the synergy between glucocorticoids and IL-6 in the
INTERLEUKIN-6 IN BIOLOGY A N D MEDICINE
27
synthesis of acute-phase proteins by hepatocytes (Snyers et al., 1990; Schooltinck et al., 1992). On the other hand, glucocorticoids decrease the monocytes/macrophages to produce IL-6, IL-1, and TNF-a feedback mechanism preventing excessive production of inflammatory cytokines. These findings show an important regulatory interaction between the immune and neuroendocrine systems.
E. A PROINFLAMMATORY AS WELLAS AN
ANTI-INFLAMMATORY CYTOKINE Interleukin-1 and T N F are known as potent inducers of proteinase production by fibroblasts, synoviocytes, and chondrocytes, and are recognized as the principal mediators of inflammatory connective tissue destruction. IL-1 and T N F activate the endothelial cells to stimulate the synthesis of intercellular adhesion molecule 1 (ICAM-1) and to induce the expression of endothelial-leukocyte adhesion molecule 1 (ELAM-l), causing neutrophils, monocytes, and lymphocytes to adhere. In contrast to IL-1 and TNF, IL-6 does not induce adhesion molecules by endothelial cells. IL-6 does not stimulate the production of collagenase, matrix metalloproteinase, or stromelysin. Rather, IL-6 is identified as a potent inducer of a tissue inhibitor of metalloproteinase l/erythroid potentiating activity (TIMP/EPA) (Sato et al., 1990; Lotz and Guerne, 1991). The acute-phase proteins regulated primarily by IL-6 are antiproteinases, oxygen scavengers, and clotting factors, although the true physiologic role of the acute-phase proteins remains unclear. Although IL-1 and T NF have extremely high toxicity in uiuo, IL-6 is tolerable at a high concentration in sera. IL-6 does not cause shock in mice (Neta et al., 1988), dog (Preiser et al., 1991) or primates (Asano et al., 1990) regardless of the amount given either alone or with TNF, although there are reports that antibodies to IL-6 reduced LPS-caused mortality in mice (Starnes et al., 1990; Heremans et al., 1992). Furthermore, IL-6 inhibited significantly the acute neutrophilic exodus and T N F production caused by an intratracheal injection of LPS, providing evidence that IL-6 may represent an endogenous negative feedback mechanism to inhibit endotoxin-initiated cytokine-mediated acute inflammation (Ulich et al., 1991). It has, however, been shown that IL-6 is the major inducer of phospholipase A2 (PLA2) gene expression in human hepatoma cells (Crow1 et al., 1991). PLA2 is an enzyme that plays an important role in inflammation by producing potent lipid mediators, such as leukotrienes, prostaglandins, and plateletactivating factor. Serum levels of PLA2 activity are elevated in septic shock and rheumatoid arthritis. Furthermore, IL-6 potentiated IL-1-
28
SHIZUO AKIRA ET AL.
and TNF-stimulated collagenase and PGE2 production by chondrocytes, although IL-6 by itself had no capacity to induce either collagenase or PGEz production by nasal chondrocytes (Smith et al., 1992). It is feasible to assume that inflamed synovium-derived IL-6, IL-1, and TNF-a might interact to augment proteinases and prostanoid production by chondrocytes in certain pathological conditions. Thus, IL-6 has two aspects as a proinflammatory as well as an antiinflammatory factor.
F. EFFECTON BONEMETABOLISM Interleukin-6 is an important regulator of bone remodeling. In normal young adults, there is a balance between the processes of bone formation by osteoblasts and bone resorption by osteoclasts. Such bone remodeling is regulated by local factors referred to as osteotropic cytokines that are generated in the microenvironment of the remodeling unit. The osteotropic cytokines include IL- 1, TNF, CSFs, and IL-6. IL-6 is produced locally in bone by osteoblasts under the direction of parathyroid hormone or other cytokines such as IL-1 and TNF (Lowik et al., 1989; Ishimi et al., 1990; Feyen et al., 1989). Osteoclasts also produce IL-6 as do a variety of other cells in the marrow microenvironment. IL-6 stimulates early osteoclast precursor formation from cells present in GFU-GM colonies (Kurihara et al., 1990). Moreover, IL-6 stimulates the recuitment as well as the formation of osteoclasts and the release of 45Ca from prelabeled fetal mouse bone, and induces bone resorption cooperatively with IL-1 in vitro. Also, nude mice inoculated with Chinese hamster ovary cells transfected with the human IL-6 gene exhibited a significant increase in blood calcium which was associated with increased levels of serum IL-6, demonstrating that IL-6 stimulates bone resorption (Black et al., 1991). Evidence for the direct role of IL-6 in osteoclastogenesis in vivo has been demonstrated. IL-6 production by bone and marrow stromal cells is suppressed by 17P-estradiol in oitro (Girasole et al., 1992). In mice, ovariectomy (estrogen loss) enhanced osteoclast development. The enhanced osteoclast formation was prevented by administration of anti-IL-6 antibody in d u o , suggesting that estrogen loss upregulates osteoclastogenesis through an increase in the production of IL-6 in the microenvironment of the marrow (Jilka et al., 1992). G. EFFECTON SKIN Skin is one of the major sites of IL-6 production. Keratinocyte cell lines can be induced to express high levels of IL-6 mRNA and protein by a number of agents including LPS, phorbol esters, various
INTERLEUKINS IN BIOLOGY A N D MEDICINE
29
toxins, and cytokines such as IL-1 and TNF. IL-6 stimulates proliferation of normal human keratinocytes (Grossman et al., 1989; Yoshizaki et al., 1990). In contrast to other cell types including fibroblasts and macrophages, IL-4 induces IL-6 in the keratinocytes. Physicochemical agents such as thermal injury and ultraviolet irradiation lead to increased production of IL-6 by the skin (Kirnbauer et al., 1991). Increased systemic levels of IL-6 can be detected in human volunteers and in experimental animals following ultraviolet irradiation (Urbanski et al., 1990). Acute exposure to ultraviolet light causes cutaneous inflammation, malaise, somnolence, chills, and fever. Plasma IL-6 levels correlate remarkably with the course of fever followed by an increase in acute-phase proteins such as CRP. IL-6, which is released by keratinocytes following ultraviolet exposure, may gain access to the circulation and may function as an important mediator of systemic sunburn reaction.
H. EFFECTON BLOODVESSELS The vascular endothelial cells that form the inner lining of blood vessels play an active role in mediating an inflammatory response. IL-1, LPS, and oncostatin M induce IL-6 production in endothelial cells (Sironi et al., 1989; Jirik et al., 1989; Brown et al., 1990). IL-6 caused a significant increase in the mRNA level of platelet-derived growth factor (PDGF) in cultured human endothelial cells (Calderon et ul., 1992). PDGF stimulates the proliferation and migration of vascular smooth muscle cells (VSMCs) and fibroblasts. PDGF is also chemotactic for monocytes and neutrophils and induces them to release inflammatory mediators such as superoxide anion and lysozyma1 enzymes. IL-6 is also shown to increase the permeability of endothelial cells (Maruo et d., 1992). These series of events are considered to contribute to vasculitis and atherosclerogenesis. VSMCs also express IL-6 in response to IL-1 (Loppnow and Libby, 1990). IL-6 stimulates the proliferation of VSMCs in a PDGFdependent manner (Nabata et al., 1990; Ikeda et al., 1991). Therefore, IL-6 is released by VSMCs and promotes the growth of VSMCs in an autocrine manner via induction of endogenous PDGF production. Endothelial cells continuously form nitric oxide from L-arginine. This basal release of nitric oxide by the endothelium accounts for the biological properties of the so-called endothelium-derived relaxing factor (EDRF) and is involved in the regulation of regional vascular resting tone of different vascular beds, including the proximal and resistance coronary vessels. It has been demonstrated that IL-6 inhibits heart contractility in a concentration-dependent, reversible
30
SHIZUO AKlRA ET AL.
manner (Finkel et al., 1992). The nitric oxide synthase inhibitor N G monomethy-L-arginine (L-NMMA) blocked the negative inotropic effect, whereas L-arginine reversed the inhibition by L-NMMA, suggesting that the direct negative inotropic effect of IL-6 is mediated through a myocardial nitric acid synthase. Serum IL-6 levels are reported to become elevated in patients of myocardial infarction (Entman et al., 1991; Ikeda et al., 1992b). The effects of L-NMMA on coronary blood flow and resistance are shown to be attenuated substantially in postinfarction reactive cardiac hypertrophy (Drexler et al., 1992). Septic shock is characterized by cardiocirculatory insufficiency in which vascular hyporesponsiveness is a major determinant of mortality. Correlations between nitric oxide production and hypotension in shock have been observed (Fleming et al., 1990; Westenberger et al., 1990). Therefore, in both myocardial postinfarction and septic shock, monokines including IL-6 may depress the contractile function by activating the L-argininehitric oxide pathway of the vascular smooth muscles. I. EFFECTON NEURONALCELLS Interleukin-6 is produced in neuronal cells by specific stimuli, and it also exerts some effects on them. LPS, IL-1, and y-interferon (IFN-y) induce expression of mRNA for IL-6 in cultured astroglial cells and microglia. Either IL-lP or TNF exerts a strong inducing signal for IL-6 in primary rat astrocytes (Lieberman et al., 1989). Virus-infected microglial cells and astrocytes produce IL-6 (Frei et al., 1988). IL-6 stimulates astrocytes and some other neural cells to proliferate. IL-6 was found to induce the differentiation of PC12 cells into neural cells (Satoh et al., 1988). Human IL-6 can support the survival of the cultured cholinergic neurons in addition to regulating dopamine synthesis (Hama et al., 1989). Noteworthy is the presence of bidirectional communication between the immune and neuroendocrine systems. The neuroendocrine system, particularly the hypothalamic-pituitary-adrenal (HPA) axis, can modulate immune responses, whereas inflammatory cytokines can modulate neuroendocrine activities. Centrally administered IL-6 rapidly exerts a stimulatory effect on ACTH release in conscious male rats (Naitoh et al., 1988). IL-6 also stimulates the release of a variety of anterior pituitary hormones, such as prolactin, growth hormone, and luteinizing hormone (Spangelo et al., 1989). IL-1 injected into the lateral brain ventricle of rats increases circulating IL-6 levels in hypophysectomized and adrenalectomized rats (De Simoni et al., 1990). Corticotropin-releasing factor (CRF) is
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
31
shown to work as an autocrine or paracrine inflammatory cytokine. IL-1 and IL-6 can induce the synthesis and secretion of CRF by hypothalamic cells (Lyson et al., 1991; Navarra et al., 1990);conversely, this neuropeptide can induce the synthesis and secretion of IL-1 and IL-6 (Leu and Singh, 1992). Anterior pituitary cells by themselves are found to produce IL-6 (Spangelo and MacLeod, 1990). IL-6 is produced by anterior pituitary cells in response to LPS and phorbol ester and agents that elevate intracellular CAMP concentrations in uitro. Also, IL-6 production by anterior pituitary cells is stimulated by vasoactive intestinal peptide (VIP), which stimulates adenylate cyclase activity, causing a concentration-dependent enhancement of IL-6 production (Spangelo and MacLeod, 1990). In addition to increasing ACTH and CRF release from the hypothalamus and pituitary, endotoxin and IL-1 also cause the adrenal to release IL-6 directly (Judd et al., 1990). Induction of IL-6 production by IL-1 may be potentiated markedly by the stress-induced elevation of ACTH levels, because the effects of ACTH and IL-1 together on IL-6 production are greater than the sum of their effects separately. IL-6 in turn may stimulate the release of glucocorticoid from the adrenal cortex. IL-6 not only stimulates basal corticosterone release, but potentiates ACTH-stimulated corticosterone release (Salas et al., 1990). Therefore, under chronic stress, IL-6 produced in the adrenal gland may amplify the response of the adrenal cortex to ACTH. J. ROLE DURING EMBRYONIC DEVELOPMENT
The potential role of cytokines during embryonic development and particularly their possible function in the development of pluripotent hemopoietic stem cells have been studied by several investigators. Murray et al. (1990) detected mRNA transcripts for IL-6 and LIF but not for GM-CSF or IL-3 in mouse blastocysts at 3.5 days of gestation, suggesting that IL-6 and LIF may regulate the growth and development of trophoblasts or embryonic stem cells. Schmitt et al. (1991) and Burkert et al. (1991) independently detected transcriptional activation of several cytokines and the corresponding receptor genes (Epo, CSF-1, IL-4, and IL-6) during embryonic stem cell development; however, IL-3 and GM-CSF were not expressed during the first 24 days of embryonic stem cell differentiation, strongly suggesting that IL-3 and GM-CSF are not critical to early hematopoiesis. Rothstein et al. (1992) investigated expression of cytokines in cDNA libraries from unfertilized eggs and two-cell, eight-cell, and blastocyst-stage mouse embryos, and identified IL-6 transcripts as early as the eight-cell stage, persisting into the blastocyst stage;
32
SHIZUO AKIRA ET AL.
however, the role of IL-6 in embryogenesis awaits further experimentations.
K. ROLE DURING PLACENTAL/FETAL DEVELOPMENT Interleukin-6 has been speculated to influence placental/fetal development. The ovarian steroids estrogen and progesterone regulate cellular and molecular changes that occur in the uterus during the estrous cycle. During the estrous cycle, uterine cells undergo cycles of proliferation, differentiation, and death. Freshly explanted human endometrial cells secrete IL-6 (Semer et aZ., 1991). Angiogenesis accompanies the cyclic destruction and reconstitution of the endometrium. IL-6 mRNA is transiently expressed during the angiogenesis that accompanies folliculogenesis and formation of the maternal decidua during early postimplantation development (Motro et al., 1990). Uterine stromal and endothelial cells secrete IL-6, and its production by these cells is inhibited by estrogen and/or progesterone. Activated leukocyte products including IL-6 arrest embryonic development at the two-cell to morula stage (Hill et al., 1987). Furthermore, IL-6 added to blastocysts on laminin-coated tissue culture wells results in a transient inhibition of the rate of blastocyst attachment and, to a greater extent, an inhibition of the rate of embryo outgrowth (Jacobs et aZ., 1992). These data indicate that IL-6 can have inhibitory effects on preimplantation embryos. Interleukin-6 is produced by extraembryonic tissues later in gestation and may act on maternal tissue to control interactions between the fetus and the mother, such as angiogenesis, the formation of new blood vessels. This important process accompanying the development of the placenta and the uterus following inplantation of the embryo may be under fetal as well as maternal control. Studies with human placental trophoblasts or whole placental tissue, which is extraembryonic but primarily fetal in origin, have demonstrated the presence of both IL-6 mRNA and biologically active protein (Kameda et aZ., 1990; Duc-Goiran et aZ., 1989). IL-6 produced by human trophoblasts could act on trophoblasts to stimulate the release of human chorionic gonadotropin through a pathway distinct from the one involving gonadotropin-releasing hormone (Nishino et al., 1989). As hCG in turn stimulates the placenta to produce progesterone, which then acts on uterine tissue to maintain an abundant supply of blood vessels, placental-derived IL-6 may be important in the initiation of this angiogenic process. Biological functions of IL-6 are summarized in Table 111.
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
33
TABLE 111 BIOLOGICAL FUNCTIONS O F INTERLEUKIN-6 Effect on €3 cells
Effect on T cells
Effect on hematopoietic progenitor cells Effect on megakaryocytes Effect on macrophages Effect on hepatocytes Effect on bone metabolism Effect on blood vessels Effect on neuronal cells Effect on placenta
Immunoglobulin production Proliferation of hybridoma/plasmacytoma/myeloma cells Proliferation of EBV-infected B cells Proliferation and differentiation of T cells Differentiation of cytotoxic T lymphocytes Induction of IL-2 receptor (Tac antigen) expression and IL-2 production Augmentation of natural killer activities Enhancement of multipotential hematopoietic colony formation Megakaryocyte maturation Growth inhibition of myeloid leukemic cell lines Macrophage differentiation of myeloid leukemic cell lines Acute-phase protein synthesis Stimulation of osteoclast formation Induction of bone resorption Induction of platelet-derived growth factor Proliferation of vascular smooth muscle cells Negative inotropic effect on heart Neural differentiation of Pc12 cells Support of survival of cholinergic neurons Induction of adrenocorticotropic hormone synthesis Secretion of human chorionic gonadotropin by trophoblasts
VI. Interleukin-6 and Disease
A. B CELLNEOPLASIA Multiple myeloma is a human B cell neoplasm characterized by accumulation, in the bone marrow, of plasma cells that secrete monoclonal immunoglobulins and by multiple osteolytic lesions. IL-6 is important for in vivo growth of murine plasmacytomas and human myelomas (Van Damme et a[., 1987b; Kawano et al., 1988), suggesting a possible involvement of IL-6 in the generation of plasmacytomas/myelomas. There is a significant association between the occurrence of plasma cell neoplasias and chronic inflammation (Isobe and Osserman, 1971; Isomaki et al., 1978). Plasmacytomas can be induced in BALB/c mice by mineral oils such as pristane that are potent inducers of chronic inflammation and IL-6 production (Potter
34
SHIZUO AKIRA ET AL.
and Boyce, 1962). Indomethacin inhibits plasmacytogenesis (Potter et al., 1985) and inhibits the elevation of IL-6 in pristane-treated mice (Shacter et al., 1992). Kawano et al. (1988) reported that IL-6 is a possible autocrine growth factor for human myeloma cells, including human myeloma cell line U266, on the basis of three pieces of evidence: (1) IL-6 induced in uitro growth of myleoma cells freshly isolated from patients with multiple myeloma. (2) Myeloma cells spontaneously produced IL-6 and expressed the IL-6 receptor. (3) In uitro growth of myeloma cells was specifically inhibited by anti-IL-6 antibody. In support of the existence of IL-6-mediated autocrine growth, Schwab et al. (1991) have shown that the addition of a neutralizing anti-IL-6 monoclonal antibody or IL-6 antisense oligonucleotides can inhibit proliferation of the human myeloma cell line U266, and these effects are reversed by adding IL-6. A similar finding is presented by Levy et al. (1991). It have also been demonstrated that IL-1 or IFN-a stimulated the growth of human myeloma cells by inducing autocrine production of IL-6 in myeloma cells (Kawano et al., 1989; Jourdan et al., 1991); however, Klein et al. (1989) were unable to confirm the autocrine hypothesis in human myelomas and proposed a paracrine hypothesis instead. They demonstrated significant IL-6 mRNA expression in bone marrow stromal cells, not purified myelomas cells, from most of the myeloma patients in uiuo (13119 patients). Bone marrow stromal cells from multiple myelomas actively produce IL-6, although normal bone marrow stromal cells do not (Nemunaitis et al., 1989). Activated multiple myeloma stromal cells may play a role in supporting the growth and final differentiation of malignant B cells of peripheral origin and further promoting the recruitment of circulating osteoclast precursors and the growth of osterclasts, resulting in bone destruction (Caligaris-Cappio et al., 1991). Indeed, a significant increase in bone resorption is observed from the early stage of multiple myeloma. Together, these pieces of evidence indicate that IL-6 functions as an autocrine as well as paracrine factor in myeloma cells. Serum levels of IL-6 reflect disease severity in plasma cell dyscrasias. In the sera of patients with advanced multiple myeloma increased levels of IL-6 have been described, whereas in early stages IL-6 is usually not elevated (Bataille et al., 1989). It is possible that, in some tumors, high serum levels of IL-6 might reflect a progression from paracrine to autocrine growth in myeloma cells (JernbergWiklund et al., 1992). Activation of the IL-6 and the IL-6R genes through viral insertion has been reported. A murine plasmacytoma cell line, MPC11, consti-
INTERLEUKINS IN BIOLOGY AND MEDICINE
35
tutively produced IL-6 owing to the insertion of an intracisternal A particle (IAP) retrotransposon 18 bp 5’ of the transcriptional start site of the IL-6 gene (Blankenstein et al., 1990). The IL-6 gene but not the IL-6 receptor gene is occasionally rearranged in patients with mutiple myelomas (Fiedler et al., 1990). Enhanced expression of the IL-6 receptor was observed in a murine plasmacytoma cell line, P3U1, in which the intracytoplasmic region of the IL-6R was replaced with part of the long terminal repeat (LTR) of the IAP gene (Sugita et al., 1990). Transfection with an IL-6 cDNA was recently shown to increase the tumorigenicity of an IL-6-dependent B cell hybridoma and a mouse plasmacytoma (Tohyama et al., 1990; Vink et al., 1990). The increase in tumorigenicity was inhibited in the animals treated with monoclonal antibodies capable of blocking the binding of IL-6 to its receptor. A similar autocrine growth mechanism is also suggested in EBVtransformed B cells. EBV is a herpesvirus that preferentially infects B lymphocytes and induces proliferation, Ig secretion, and immortalization. EBV infection appears to play a pathogenetic role in the development of endemic Burkitt lymphoma and in the recently identified AIDS-associated Burkitt lymphoma. The continuous proliferation of EBV-infected B cells appears to depend on autocrine secretion of cytokines, including IL-1 and IL-6 (Scala et al., 1985; Tosato et al., 1990; Yokoi et al., 1990). The constitutive expression of the exogenously introduced IL-6 gene in EBV-infected B cells led to an altered pattern of growth and to a malignant phenotype, as shown by clonogenicity in soft agar cultures and turmorigenicity in nude mice (Scala et al., 1990). These data suggest that the combined action of EBV, which exerts an immortalizing function, and IL-6, which has growth-promoting activity, can give rise to fully transformed B cell tumors in immunodeficient subjects. A possible autocrine role for IL-6 has also been reported in non-Hodgkin’s lymphomas, chronic lymphocytic leukemias, and acute myeloid leukemias (Freeman et al., 1989; Yee et al., 1989; Biondi et al., 1989; Oster et al., 1989). An increased incidence of lymphoproliferative disorders [posttransplant lymphoproliferative disorders (PTLD)] is reported in cardiac transplant recipients receiving OKT3 monoclonal antibody (Swinnen et al., 1990). An extremely high incidence of EBVassociated PTLD (38%) is observed in those patients having received multiple courses of OKT3. Before exerting its potent immune suppressive properties, the OKT3 monoclonal antibody induces the activation of T cells and monocytes (Abramowicz et al., 1989; Chatenoud et al., 1989). The observation that OKT4 monoclonal antibody and
36
SHIZUO AKIRA ET AL.
cyclosporin A enhanced transcription of the IL-6 gene suggests that IL-6 may play a role in the development of B cell lymphomas in transplant recipients (Bloemena et al., 1990; Walz et al., 1990; Goldman et al., 1992).
B. BACTERIALAND PARASITE INFECTION A marked increase of IL-6 levels is observed in the cerebrospinal fluid of patients with acute bacterial infection of the central nervous system and in the serum of patients with severe burns or sepsis (Hack et al., 1989; Waage et al., 1989; Helfgott et al., 1989). Healthy individuals have undetectable or very low concentrations of IL-6 in their sera (always <50 pg/ml; in most cases,
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
37
onts of P. berghei (Vreden et al., 1992). IL-6 is also shown to directly induce the cytotoxicity of platelets toward Schistosoma mansoni (Pancre et al., 1990). C. VIRALINFECTION Various virus infections have been shown to effectively induce the production of IL-6 (Van Damme et al., 1989; Sehgal et al., 1988; Becker et nl., 1991; Devergne et al., 1991). Numerous observations indicate that the levels of IL-6 in body fluids increase rapidly after viral infection. IL-6 in cerebrospinal fluid of HTLV-1-associated myelopathy was increased (Ohbo et al., 1991; Nishimoto et al., 1992). By infection with experimental lymphocytic choriomeningitis virus in mice, IL-6 levels were also approximately 60 times higher in CSF than in serum (Frei et al., 1988). The patients positive for IL-6 generally had more severe clinical symptoms and signs than those negative for IL-6. Regarding the interaction between viral infection and IL-6, one of the most extensively studied viruses is human immunodeficiency virus (HIV), which is discussed in the next section. It has been demonstrated that hepatitis B virus (HBV) uses IL-6 as a target for attachment to cells. HBV, a member of the family of hepadnaviridae, is a major human pathogen implicated in primary hepatocarcinoma. HBV receptors were detected on human liver and hepatoma cells, on B lymphocytes, on monocytes activated by LPS, and on T cell lines activated by Con A. It was also shown that IL-6 carries a binding site for HBV and mediates HBV-cell interactions. Therefore, HBV belongs to an apparently expanding family of viral pathogens using interactions between virally encoded proteins and cytokines or cytokine receptors as steps in replication (Neurath et al., 1992).
D. ACQUIREDIMMUNODEFICIENCY SYNDROME 1. Human Immunodeficiency Virus Infection Accumulating data have indicated that IL-6 has a role in the clinical manifestations associated with HIV infection. Initial infection with HIV is usually followed by a long asymptomatic period characterized by viral latency or low-level virus replication. This latency persists until the appropriate activation signals stimulate viral transcription. IL-6 participates in an autocrine loop which amplifies HIV-1 replication and expression. IL-6 has been shown to be involved in the activation of HIV expression in infected monocytic cells alone and in synergy with TNF (Poli et al., 1990a). On the other hand, in vitro infection of normal monocytes/macrophages with HIV has been found to induce gene expression and secretion of IL-6 (Nakajima et al., 1989).
38
SHIZUO AKIRA ET AL.
Increased levels of IL-6 have been reported in the serum of HIVinfected patients (Breen et al., 1990; Birx et al., 1990; Honda et al., 1990). In addition to progressive T cell function failure, functional abnormalities of B cells are recognized in patients with AIDS; hypergammaglobulinemia, an increase in Ig-secreting cells, in uiuo expression of activation marker on circulating B cells, and presence of autoantibodies in serum (Lane et al., 1983; Martinez-Maza et al., 1987). Spontaneous in uitro Ig production by PBMCs from HIV-infected donors was inhibited by the addition of anti-IL-6 serum, suggesting that IL-6 overproduction in HIV infection could contribute to B cell hyperstimulation and to hypergammaglobulinemia (Amadori et al., 1989). Patients with AIDS show an increased frequency of lymphoid malignancies, most of which are high-grade, non-Hodgkin’s I3 cell lymphomas. Although these lymphomas are mostly monoclonal, and often show positivity for the EBV genome, in no case has the existence of the HIV-1 genome been reported. The increased levels of serum IL-6 and polyclonal B cell activation may be associated with the increased frequencies of B cell malignancies seen in AIDS patients. Infection of the brain with HIV often accompanies neurological manifestations, clinically (memory loss, dementia, motor deficits, fever, drowsiness, weakness, and pain) and histopathologically (vasculitis, vascular necrosis, astrogliosis, demyelination, neuron loss, and central nervous system lymphoma growth). The presence of high levels of IL-6, but not TNF, in the cerebrospinal fluid of patients with AIDS suggests that IL-6 may actually be involved in these neurological manifestations of AIDS (Gallo et al., 1989).
2. Kaposi’s Sarcoma Kaposi’s sarcoma (KS) is the most frequent tumor seen in individuals with AIDS. KS cells are of vascular origin and have features in common with endothelial and smooth muscle cells. The tumor often rapidly accelerates at the time of opportunistic infection or progression of the underlying immunodeficiency. AIDS-KS-derived cell lines produce several cytokines and growth factors, including IL-1, GM-CSF, and basic fibroblast growth factor (bFGF). Addition of anti-IL-1 or antibFGF sera to AIDS-KS cell lines results in decreased cellular proliferation, indicating that these factors may be operating as autocrine/ paracrine growth factors for AIDS-KS cells. Also, various AIDS-KSderived cell lines have been shown to secrete substantial amounts of IL-6, to contain intracellular IL-6 by immunoperoxidase staining, and to display high levels of IL-6 mRNA. AIDS-KS cells express the recep-
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
39
tor for IL-6. IL-6 has been shown to stimulate proliferation of AIDSKS-derived cell lines (Miles et al., 1990). IL-6 antisense oligodeoxynucleotides decreased IL-6 production by these cells and dramatically inhibited their growth (Miles et al., 1990). Together, these results suggest that IL-6 is an autocrine growth factor for AIDS-KS cells. Oncostatin M is found to be a potent mitogen for AIDS-KS cells (Miles et al., 1992; Nair et al., 1992).After exposure to oncostatin M, AIDS-KS cells assumed a spindle morphology, had an increased ability to proliferate in soft agar, and secreted increased amounts of IL-6. Furthermore, antisense IL-6 oligonucleotides inhibited by approximately 50% the mitogenic effects of oncostatin M, suggesting that both IL-6- and non-1L-6-dependent mechanisms may be involved in the effects of oncostatin M on AIDS-KS cells. HIV tat is a virally encoded, transcription-enhancing molecule, which can interact with a sequence within the HIV-LTR region to upregulate viral gene expression. Exogenously added Tat protein entered cells and transactivated the HIV-LTR. It has been reported that transgenic mice generated with the tat gene under the control of HIV-LTR developed a KS-like disease similar to the human counterpart (Vogel et al., 1988).Tat protein is shown to induce IL-6 production in KS cells and stimulate growth of KS cells (Ensoli et al., 1990). The Tat-induced increase in IL-6 expression and KS cell proliferation was specifically inhibited by antisense IL-6 oligonucleotides, suggesting that at least part of the mechanism by which Tat increases KS cell proliferation is through an IL-6-dependent mechanism.
E. INFLAMMATORY OR AUTOIMMUNE DISEASES 1. Cardiac Myxoma The first example of a disease associated with IL-6 overproduction was cardiac myxoma, a type of benign intraatrial heart tumor (Hirano et al., 1987). Patients with cardiac myxoma frequently display hypergammaglobulinemia and various kinds of autoantibodies and an increase in acute-phase proteins. As these symptoms disappear after resection of the tumor, the myxoma cells or products derived from them were suspected to induce the symptoms. Cultured cardiac myxoma cells were shown to produce a large amount of IL-6 protein constitutively. This finding was confirmed by Jourdan et al. (1990), who reported a case of cardiac myxoma in which after surgical removal of the tumor, the significantly elevated levels of serum IL-6 returned to undetectable levels, with regression of the immunological features. A dosedependent relationship is also reported between plasma levels of IL-6,
40
SHIZUO AKIRA ET AL.
but not other cytokines (IL-1, TNF-a, IFN-y), and the immunological features of the patient. This evidence demonstrates that IL-6 is responsible for the clinical manifestations seen in cardiac myxoma.
2. Rheumatoid Arthritis Rheumatoid arthritis (RA) patients characteristically show polyclonal plasmacytosis, presence of autoantibodies, and increases in acutephase proteins and platelets. These symptoms may be explained by overproduction of IL-6. In fact, elevated levels of IL-6 can be detected in the synovial fluid from affected joints and sera of patients with active RA (Hirano et al., 1988; Houssiau et al., 1988b; Bhardwaj et al., 1989; Guerne et al., 1989). Significant correlations have also been shown between the concentrations of synovial IL-6 and IgG, as well as between serum IL-6 activity and levels of a variety of acute-phase proteins such as CRP (Houssiau et al., 198813; Hermann et al., 1989).T cells, B cells, synoviocytes, and chondrocytes have been identified as sources of IL-6. Several other cytokines, such as IL-1 and TNF, which are also present in synovial fluid, are potent inducers of IL-6. IL-6 can contribute to the proliferation of synovial lining cells, which results in the marked buildup of inflammatory tissue (pannus) in the joints of patients with RA (Firestein et al., 1990). Inflammatory cytokines and the network of their interactions may therefore be involved in RA. Interleukin-6 production was observed in type I1 collagen-induced arthritis in mice (Takai et al., 1989). An age-associated increase in serum IL-6 was also observed in MRL/lpr mice, which spontaneously develop autoimmune diseases with lymphoid hyperplasia associated with an infiltration of plasma cells, arthritis, hypergammaglobulinemia, and a high incidence of monoclonal or oligoclonal IgGs (Tang et al., 1991). Pristane can cause not only plasma cell neoplasias, but also arthritis in certain mouse strains (Potter and Wax, 1981).Together, these facts suggest that IL-6 may be involved in the development of autoimmune arthritis. 3. Castleman’s Disease Abnormal IL-6 production and polyclonal plasmacytosis have also been observed in patients with Castleman’s disease (Yoshizaki et al., 1989). This chronic disease with benign hyperplastic lymphadenopathy is characterized by large lymph follicles with intervening sheets of plasma cells. The patients show hypergammaglobulinemia and increases in acute-phase proteins and platelets. B blastoid cells in the germinal centers of such hyperplastic lymph nodes were found to produce IL-6. Clinical improvement and decrease in serum IL-6 levels
INTERLEUKIN-6 IN BIOLOGY AND. MEDICINE
41
were observed following resection of the hyperplastic lymph node. Major systemic effects resembling Castleman's disease have been reported in congenitally anemic W / W mice reconstituted with hematopoietic cells transfected with the coding sequences of murine IL-6 (Brandt et al., 1990). 4 . Mesangial Proli$erative Glomerulonephritis Mesangial proliferative glomerulonephritis (MPG) is histologically characterized by proliferation of mesangial cells. IL-6 is demonstrated to he a possible autocrine growth factor for rat mesangial cells and to be constitutively produced by renal mesangial cells in patients with MPG (Horii et al., 1989; Ruef et al., 1990). Furthermore, IL-6 can be detected in the urine samples of patients with MPG, but not in those of patients with other types of glomerulonephritis. There is also a correlation between the level of urine IL-6 and the progressive stage of MPG. 5. Psoriasis
Psoriasis is characterized by epidermal hyperplasia, altered epidermal maturation, and local accumulation of acute-phase and chronic inflammatory cells. Various abnormalities have been observed in the serum and plasma of patients with psoriasis. Increases in CRP and an-macroglobulin appear related to severity. Significantly elevated amounts of circulating IL-6 are observed in sera of psoriatics and are likely to account for the increases in CRP and an-macroglobulin (Grossman et al., 1989; Neuner et al., 1991).In contrast to normal or uninvolved skin, keratinocytes in psoriatic lesions are remarkably positive for IL-6 as detected by immunohistochemistry and in situ hybridization (Grossman et al., 1989). IL-6 also stimulates the proliferation of keratinocytes and may account for the hyperkeratosis observed in psoriatic plaques. The diminution of hyperkeratosis following therapy was accompanied by a decrease in IL-6 staining in the psoriatic plaque (Sehgal, 1990);however, overproduction of IL-6 in skin of mice expressing the keratin promoter-driven IL-6 transgene did not result in hyperkeratosis (Truksen et al., 1992). 6. Systemic Lupus Erythematosus Systemic lupus erythematosus (SLE) is a chronic autoimmune disease involving almost any organ system. One of its major immunological abnormalities is B cell hyperactivity. IL-6 levels were elevated in some of the sera (Swaak et al., 1989) and in the cerebrospinal fluids (Hirohata and Miyamoto, 1990) of patients with SLE. Patients with SLE showed increased IL-6 gene transcription by T cells, B cells, and
42
SHIZUO AKIRA ET AL.
monocytes (Tanaka et al., 1988; Linker-Israeli and Deans, 1989). IL-6 preferentially stimulates resting SLE B cells (Kitani et al., 1989). SLE B cells show spontaneous activation by in vitro culture without any stimulation. The elevated spontaneous in vitro production of IgG by SLE mononuclear cells is enhanced by exogenous IL-6, and is partially inhibited by neutralizing anti-IL-6 antibodies (Linker-Israeli et al., 1991).Cultured mononuclear cells from patients with SLE, but not patients with RA or normal donors, secreted IL-6 after exposure to ultraviolet light, which may be involved in the exacerbations of SLE provoked by photosensitivity (Pelton et al., 1992). The mechanism leading to the constitutive expression of IL-6 in SLE has not yet been elucidated, It has been reported that SLE patients lack a cell subset that specifically suppresses IL-6 production (Warrington and Rutherford, 1990). In the plasma of SLE patients, nucleosomes are found as the major component of circulating DNA (Rumore and Steinman, 1990). Both DNA and nucleosomes have been demonstrated to bind specifically to the surface of human peripheral blood mononuclear cells and to stimulate the release of IL-6 (Hefeneider et nl., 1992). IL-6 secretion, as a consequence of DNA and nucleosome interaction with cell surface binding molecules, may have relevance to the polyclonal activation of immunoglobulin production seen in SLE patients.
7. Alzheimer’s Disease Alzheimer’s disease is the most common cause of dementia in the elderly. The characteristic neuropathological findings are PA4containing plaques, neurofibrillary tangles, and amyloid infiltration of cerebrovascular walls. There is a significant correlation between PA4 deposition and the clinical severity of dementia. IL-6 has been implicated in Alzheimer’s disease (AD), where along with IL-1, it may modulate amyloid protein precursor synthesis (Griffin et al., 1989; Vandenabeele and Fiers, 1991). A major proteinaceous constituent of senile plaques in AD, called PA4, is derived from PA4 precursor protein [amyloid precursor protein (APP)]. Normal secretion of the major amino-terminal part of the membrane-inserted APP molecules is achieved by cleavage of APP by a proteolytic enzyme inside the PA4 sequence, thus excluding the formation of the pathogenic PA4 under normal physiological conditions. Protease inhibitors such as az-macroglobulin (a2-M) and alantichymotrypsin prevent normal APP processing in AD, allowing formation of the pathogenic PA4. a2-M is synthesized by human neuronal cells, but only after stimulation of the cells with IL-6 (Ganter et al., 1991). Strong immunohistochemical staining for a2-M and IL-6 was
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
43
found in senile plaques of AD cortices (Bauer et al., 1991). ( ~ 1 Antichymotrypsin that is induced by IL-6 was also a component of amyloid deposits in neuritic plaques and blood vessel walls from brains of AD patients.
8 . Cachexia Animals and humans with chronic infections or cancer may develop the syndrome of cachexia, which is characterized by severe wasting of both protein and fat stores. The syndrome is often associated with elevated serum triglycerides and reduced serum LPL activity. LPL plays an important role in adipocyte metabolism by hydrolyzing circulating triglycerides to fatty acids for subsequent storage as fat in the adipocytes. Therefore, a reduction in the activity of LPL contributes to a decrease in total body fat stores. TNF has been considered as a mediator of cachexia (Tracey e t al., 1988).Anti-TNF antibodies block tumor-induced cachexia, but the blocking effect is only partial. Elevated TNF concentrations have generally not been found in cancer patients with cachexia. On the other hand, introduction of a retroviral expression vector containing the IL-6 coding sequence into the bone marrow of mice or introduction of IL-6-transfected Chinese hamster ovary cells resulted in mice with marked weight loss (Black et al., 1991).Strassmann et al. (1992) have presented direct evidence of the contribution of endogenous IL-6 to the development of cachexia. Circulation levels of IL-6 in C-26.IVX-bearing mice correlated directly with tumor size and the extent of cachexia. In such models IL-6, but not TNF or IL-1, was detected in the peripheral circulation. If the primary tumors were resected, mice gained weight and the levels of IL-6 in the serum were reduced significantly. Moreover, monoclonal antibody to murine IL-6 (but not anti-TNF antibody) was able to significantly suppress the development of cachexia in tumor-bearing mice. Furthermore, it is reported that IL-6 inhibits lipoprotein lipase activity in adipose tissue and this effect may contribute to the observed weight loss seen in the tumor-bearing host (Greenberg et al., 1992). The hypertriglycemia observed in cachexia may be the result of the increased hepatic production of lipoprotein stimulated by IL-6 (Feingold and Grunfeld, 1992); however, these data are in striking contrast to the observation that administration of exogenous IL-6 to mice or other experimental animals does not elicit a weight loss syndrome. Perhaps in conjunction with TNF, IL-1, IFN-y, and other as yet unknown cytokines, IL-6 may play a role in the development of the cachectic syndrome.
44
SHIZUO AKIRA ET AL.
F. OTHERDISEASES Constitutive production of IL-6 has been demonstrated in renal cell carcinoma (Miki et al., 1989), ovarian carcinoma (Watson et al., 1990), pleural mesothelioma (Schmitter et al., 1992), parotid gland adenoma (Gallo et al., 1992), pheochromocytoma (Fukumoto et al., 1991), and glioblastoma cells (Van Meir et al., 1990). Patients with these malignancies often exhibit paraneoplastic symptoms, which include fever, leukocytosis, erythrocytosis, thrombocytosis, hypercalcemia, and elevation of acute phase reactants. Many of the paraneoplastic symptoms parallel the actions of IL-6. Elevated IL-6 concentrations are detected in sera of patients with multiple sclerosis (Maimone et al., 1991; Frei et al., 1991), alcoholic hepatitis (Hill et al., 1992a; Sheron et al., 1991),acute hepatitis (Sun et al., 1992), polyarteritis nodosa (Nakahama et al., 1992), keloid (McCauley et al., 1992), and HTLV-1-associated myelopathy (Ohbo et al., 1991; Nishimoto et al., 1990). Patients with these diseases sometimes manifest many aspects of the acute-phase response and systemic B cell activation, which may be explained by actions of IL-6. Interleukin-6 may be an important mediator in the pathogenesis of autoimmune insulin-dependent diabetes mellitus (Bendtzen et al., 1989; Cambell et al., 1989, 1991), thyroiditis (Bendtzen et al., 1989), and Paget’s disease (Roodman et al., 1992). High-level expression of IL-6 mRNA was detected in active inflammatory bowel disease specimens from ulcerative colitis and Crohn’s disease patients (Stevens et al., 1992) and in atherosclerotic lesions of WHHL rabbit aortae, which are strikingly similar to those observed in human familial hypercholesterolemia (FH)(Ikeda et al., 1992a). Levels of IL-6 were significantly elevated in the bronchoalveolar lavage fluid of patients with symptomatic compared with asymptomatic asthma (Mattoli et al., 1991; Broide et al., 1992). In contrast, a recent report shows that IL-6 may act as an agent that downregulates an inflammatory response. Denis (1992) examined the role of IL-6 in the development of chronic lung inflammatory conditions, using a mouse model of hypersensitivity pneumonitis established by intranasal instillation of the thermophilic actinomycete. Neutralization of endogenous IL-6 by anti-IL-6 antibody brought about a significant increase in fibrosis. Conversely, direct IL-6 intratracheal infusion was associated with a diminished fibrotic response. This decrease in fibrosis was shown to be the result of the decreased neutrophi1 influx and the inhibitory action of IL-6 on the production of TNF, which plays an important role in mouse hypersensitivity pneumonitis.
INTERLEUKINS I N BIOLOGY AND MEDlCINE
45
Diseases associated with abnormal IL-6 production are summarized in Table IV.
G. INTERLEUKIN-6TRANSGENIC MICE Transgenic animals or animals engrafted with a tumor genetically engineered to produce a given cytokine provide models for studying the in vivo effects of cytokines. To analyze the biological effects of very TABLE IV INTERLEUKIN-6 AND CLINICAL DISORDERS Acute inflammatory disorders Meningitis Intrauterine infection Acute transplant rejection Burns Surgical trauma Chronic diseases Rheumatoid arthritis Mesangial proliferative glomerulonephritis Psoriasis Castleman’s disease Alzheimer’s disease Systemic lupus erythematosrls Autoimmune insulin-dependent diabetes Thyroiditis Multiple sclerosis Alcoholic hepatitis Acute hepatitis Bronchial asthma Paget’s disease Polyarteritis nodosa Crohn’s disease Scleroderma Cachexia Keloid HTLV-associated myelopathy Arteriosclerosis Neoplasms Multiple myeloma Leukemia Post-transplant lymphoproliferative disorders Kaposi’s sarcoma Renal cell carcinoma Pheochromocytoma Pleural mesothelioma Cardiac myxoma Parotid eland adenoma
46
SHIZUO AKlRA ET AL.
high and continuous levels of IL-6, IL-6 transgenic mice were generated. The IL-6 transgenic mice (C57BL/6) were produced using the human IL-6 genomic gene fused with the human immunoglobulin heavy chain enhancer to ensure a high level expression of IL-6 in B lineage cells (Suematsu et al., 1989). In these IL-6 transgenic mice a fatal plasmacytosis, histologically indistinguishable from plasmacytoma, was generated. High concentrations of human IL-6 and a polyclonal increase in IgGl were detected in the sera of all transgenic mice; however, the plasma cells were not transplantable to syngeneic mice and did not contain obvious c-myc gene rearrangements, which are observed in almost all pristane-induced plasmacytoma cells. Susceptibility to plasmacytoma development is genetically determined and most inbred strains other than BALB/c are resistant. Backcrossing the C57BL/6 IL-6 transgenic mice to BALB/c induced transplantable plasmacytomas with myc translocation (Suematsu et al., 1992). The other interesting findings in the IL-6 transgenic mice are the generation of MPG and the increase in mature megakaryocytes in the bone marrow. The former evidence taken together with the clinical and experimental data described in the previous section indicate that abnormal IL-6 expression plays a critical role in the generation of MPG. The latter evidence is in complete agreement with the recent findings that IL-6 induces the maturation of megakaryocytes as a thombopoietic factor in uitro and induces an increase in platelet number in both mice and monkeys. The effects of IL-6 in uiuo were assessed b y inoculating into nude mice Chinese hamster ovarian (CHO) cells that were transfected with the murine IL-6 gene (Black et al., 1991). Nude mice bearing CHO cells expressing IL-6 developed hypercalcemia, leukocytosis, thrombocytosis, and cachexia, suggesting that increased circulating concentrations of IL-6 in patients with malignant disease may contribute to a number of paraneoplastic syndromes including hypercalcemia, cachexia, leukocytosis, and thrombocytosis. It has been demonstrated that mice transplanted with bone marrow cells infected with an IL-6-producing retrovirus develop a fatal myeloproliferative disease within 4 weeks of engraftment (Hawley et a1., 1992).The mice manifest elevated peripheral leukocyte counts with a predominance of neutrophilic granulocytes, increased mesangial cell proliferation in the kidney, frequent liver abnormalities, and marked alterations in plasma protein levels. To explore the role of IL-6 in skin, transgenic mice were constructed by using a human keratin 14 promoter to express IL-6 in the basal cells of stratified squamous epithelia (Truksen et al., 1992). Mice expressing
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
47
the K14-IL-6 transgene were smaller than normal and exhibited retarded hair growth. IL-6 expression did not lead to enhanced epidermal proliferation but it did result in a thicker stratum corneum without leukocytic infiltration, making it unlikely that it has direct proinflammatory activity in skin. The promotion of increased stratum corneum formation may help to counteract otherwise deleterious effects, such as accelerated growth of basal epidermal cells and subsequent sloughing of the differentiating cells from the skin surface, observed in psoriatic skins. Alternatively, it may be that only in combination with other factors such as IFN-y, GM-CSF, and IL-1, is this cytokine capable of mediating proinflammatory effects in the skin. VII. interieukin-6 as a Diagnostic Marker Interleukin-6 may be a diagnostic cytokine for intraamniotic infection (Romero et al., 1992). Second- or third-trimester amniotic fluid contained detectable but low levels of IL-6 (range, 50-1500 pg/ml). In preterm labor associated with intraamniotic infection there was a dramatic increase in the concentration of IL-6, up to 5 pg/ml (Romero et al., 1990; Santhanam et al., 1991a). Patients with raised IL-6 concentrations but without demonstrable infection showed histological evidence of chorioamnionitis and/or failed to respond to tocolysis (pharmacological inhibition of labor). These findings indicate that the measurement of amniotic fluid concentrations of IL-6 could be ofvalue in the diagnosis of microbial invasion of the amniotic cavity. Patients with an increased amniotic fluid IL-6 level, despite negative amniotic fluid cultures, are at risk for histological chorioamnionitis and preterm delivery. Interleukin-6 is an unspecific marker for the inflammatory state and does not allow discrimination of various inflammatory diseases; however, Heney et al. (1992) have measured plasma IL-6 to evaluate its diagnostic value in the assessment of febrile neutropenia and to study its relationship with CRP. There was a statistically significant difference between median admission IL-6 levels of patients with gram-negative and -positive infections and unexplained fevers. These groups could not be differentiated by CRP levels alone. The measurements of plasma IL-6 on admission provides a more sensitive predictive marker than does CRP of subsequent disease severity and may help to differentiate diagnostic groups of children with febrile neutropenia. Furthermore, IL-6 may be a diagnostic aid during the first hours of an inflammatory state, particularly if CRP levels had not yet increased.
48
SHIZUO AKlRA ET AL.
Interleukin-6 is also a prognostic factor in multiple myeloma and renal cell carcinoma. Myeloma patients whose IL-6 serum concentration at the time of diagnosis was less than 7 pg/ml showed significantly longer survival than patients whose IL-6 level was 7 pg/ml or higher. The respective 50%survival rates were 53.7 months as compared with only 2.7 months (Ludwig et al., 1991). Also, renal cell carcinoma patients with high serum IL-6 levels had a shorter time between the diagnosis of their primary tumor and the development of metastases, as well as a shorter survival (Blay et al., 1992). It has been shown that the measurement of urinary IL-6 is a helpful tool for monitoring the progression of IgA nephropathy (Dohi et d., 1991). Interleukin-6 measurements may be useful in monitoring early rejection in transplant patients. Van Oers et al. (1988) demonstrated that IL-6 was increased in both blood and urine at the time of rejection following renal transplantation. In liver transplantation, it has been reported that IL-6 in bile is useful as a marker of acute rejection (Umeshita et al., 1992). VIII. Clinical Application of Interleukin-6
A. TREATMENT OF MYELOSUPPRESSION AND THROMBOCYTOPENIA Neutropenia and thrombocytopenia are major factors contributing to morbidity and mortality after radiotherapy or chemotherapy of cancer. IL-6 may be therapeutically useful in the treatment of radiation- or chemotherapy-induced myelosuppression and thrombocytopenia (Burstein et al., 1992; Herodin et al., 1992). It has been shown that rIL-6 stimulates multilineage hematopoiesis and accelerates recovery from radiation- or chemotherapy-induced hematopoietic depression (Patchen et al., 1991; Takatsuki et al., 1990).IL-6 has been also shown to stimulate megakaryocyte maturation in uitro and to produce increased peripheral platelet levels in uiuo. Considering the fact that rhIL-3 markedly expands the pool of megakaryocyte progenitors but lacks significant stimulatory activity on the terminal phase of platelet production, whereas rhIL-6 causes a dose-dependent increase in blood platelets, the sequential administration of rhIL-3 and rhIL-6 represents a novel and powerful strategy to stimulate thrombopoiesis in uiuo, and may have therapeutic potential in conditions with decreased platelet production (Geissler et al., 1992).
B. CANCER TREATMENT Application of cytokines has shown promise in cancer treatment. Many cytokines are now in clinical trials. IL-6 has been demonstrated to have potent antitumor activity in certain types of tumors. Adminis-
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
49
tration of rIL-6 resulted in a reduction in the number of micrometastases from sarcomas and adenocarcinomas in mice (Mul’e et al., 1990).This antitumor activity may be associated with the development of specific cytotoxic T cells, because IL-6 induces tumor-specific cytotoxic T lymphocyte generation via the activation of helper T cells in uiuo (Kitahara et al., 1990; Mul’e et al., 1992). rIL-6 also augments the antibody-dependent cellular cytotoxicity (ADCC) activity of human peripheral blood mononuclear cells using antibodies to human tumor antigens and human tumor cells as targets, suggesting a potential role for IL-6 in combination with anticancer antibodies for cancer therapy (Tsang et al., 1991). Because of the very short in uiuo half-life of rIL-6, continuous infusion or regular injections are necessary for obtaining the significant effects, and this may cause systemic toxicity. To avoid this drawback “tumor cell-targeted cytokine gene therapy” has been developed, in which the tumor cells are genetically engineered to produce a given cytokine in uitro and their injection into mice provides a locally enhanced cytokine concentration (Russell, 1990). Tumor cell-targeted cytokine gene therapy in animal trials has been successfully used with IL-2, IL-4, IFN-y, TNF-a, and IL-6, providing therapeutic potential against human cancer. IL-6 transfection into the high-metastatic, lowimmunogenic D122 clone of Lewis lung carcinoma resulted in a significantly lower metastatic competence of the tumor cells in syngeneic C57B1/6 mice. Reduction of their tumorigenic properties and suppression of their metastatic competence were correlated with the extent of IL-6 production (Porgador et al., 1992). Thus, it seems that the prospects for antimetastatic immunotherapy by cellular vaccination of tumor cells genetically manipulated to express IL-6, either alone or in combination with genes of the MHC or other cytokines, deserve investigation. In the therapy of malignant disease, suppression of malignancy by inducing differentiation is an alternative approach to the use of anticancer compounds that kill normal cells as well as tumor cells. The fact that IL-6 promotes differentiation of myeloid leukemias suggests the potential therapeutic usefulness of IL-6 in certain myeloid leukemia cells (Givon et al., 1992). In both transplantable and radiation-induced acute myelocytic leukemia (AML) in SJL/J mice, administration of IL-6 reduced the incidence of leukemia development and increased survival. In addition, in uitro liquid cultures of leukemic blood cells obtained from AML patients showed that IL-6 inhibited growth and decreased the proportion of blasts with an increase in more mature myeloid elements; however, IL-6 was suspected to act as a growth factor for AML blasts of certain AML patients. To determine the possi-
50
SHIZUO AKIRA ET AL.
bility of using IL-6 in the treatment of human AML, the in uitro action of IL-6 on growth and differentiation of AML should be examined in both liquid and semisolid cultures. IX. Clinical Application of Interleukin-6 Inhibitors
A. ANTI-INTERLEUKIN-6ANTIBODY A clinical trial in which anti-IL-6 antibodies were used for the therapy of multiple myeloma was performed by Klein and his colleagues (1991). Injection of anti-IL-6 antibodies into multiple myeloma patients with terminal disease and extramedullary proliferation completely blocked myeloma cell proliferation in uiuo and completely inhibited the serum IL-6 bioactivity and serum CRP levels. The clinical status of patients improved throughout treatment and no major side effects were observed. Even in patients with advanced or terminal disease, in uitro myeloma cell proliferation was found to be dependent on IL-6. The reduction in serum calcium levels during treatment is inconsistent with data showing a role for IL-6 in bone resorption in vitro (Bataille et al., 1992). B. CYTOKINES AND CHEMICALS INHIBITING INTERLEUKIN-6 SYNTHESIS Corticosteroids and estrogens are potent inhibitors of IL-6 production as a result of their inhibition of IL-6 gene transcription. The inhibitory effect of corticosteroids on myeloma growth seems to be mediated by a steroid-induced decrease in IL-6 expression. Adriamycin, vincristine, and cyclophophamide, used in the treatment of multiple myeloma, are reported to suppress the release of IL-6 from LPS-stimulated human peripheral blood mononuclear leukocytes (Hasan et al., 1992). IL-4 has also been shown to interfere with growth of plasmacytoma cells by blocking endogenous IL-6 synthesis (Taylor et al., 1990; Herrmann et al., 1991). Retinoic acid downregulates the number of IL-6Rs on the myeloma cell line AF10, a variant that was cloned from U266, and, correspondingly, inhibits cell proliferation (Side11 et al., 1991). The antiproliferative action of retinoic acid on AFlO cells may be caused by reduction of IL-6R expression and subsequent inhibition of IL-6-mediated autocrine growth. B chronic lymphocytic leukemia (B-CLL)and hairy leukemia (HCL) produce IL-6, which may induce positive feedback growth loops. IFN-a abrogates IL-6-induced proliferation of HCL and B-CLL cells in uiuo (Heslep et al., 1990). Taken together, these data suggest that
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
51
IL-4, retinoic acid, and IFN-a may exert therapeutic effects in these malignancies by blocking the IL-6.IL-6R autocrine loop. Interleukin-4 has recently been shown to inhibit the secretion of proinflammatory cytokines such as IL-1, TNF, and IL-6 by activated PBMCs, monocytes/macrophages, or synovium (Briolay et al., 1992). IL-4 strongly inhibits the production of proinflammatory cytokines by rheumatoid synovium, at both the protein and mRNA levels (Miossec et d.,1992).IL-4 also has the ability to inhibit cytokine-induced bone resorption in uiuo (Watanabe et aZ., 1990). The inhibition of proinflammatory cytokine production by IL-4 might provide the rationale for the clinical use of IL-4 either systemically or locally at the site of inflammation in chronic inflammatory diseases such as rheumatoid arthritis. C. CHIMERIC TOXINS Ira Pastan and his colleagues developed a chimeric toxin in which IL-6 was fused to a mutant form of psuedomonal exotoxin (Siegall et al., 1988).The chimeric toxin has been previously shown to specifically kill malignant hepatic, prostatic, epidermoid, and myeloma cell lines in uitro (Siegall et al., 1990, 1991).The chimeric toxin is being considered for a clinical trial in treating multiple myeloma (Kreitman et al., 1992), in either phase I in uiuo treatment or ex uiuo purging of myeloma cells in autologous transplant protocols. Current problems in the use of recombinant fusion toxins in humans are the killing of numerous normal cells, particularly hepatocytes, that express IL-6 receptors and the appearance of antitoxin antibodies in some patients. X. Conclusions
Interleukin-6 may be described as a double-edged sword. There can be little doubt that IL-6 plays a pivotal role in hematopoiesis, immune reactions, and acute-phase responses; however, it has been demonstrated that the abnormal expression and dysregulation of IL-6 are involved in the pathogenesis of certain diseases. Therefore, the inhibition or modulation of IL-6 could have profound therapeutic benefits, although in most diseases it is apparent that more than one cytokine is involved and that the full clinical manifestations of the diseases result from the dysregulated cytokine network. It is now clear that blockade of IL-6 function can inhibit myeloma cell proliferation and improve the clinical status of patients with multiple myeloma. The cell biology of the intracellular events that link transduction to gene regulation is an important area, and work on these topics may
52
SHIZUO AKIRA ET AL.
help us understand such phenomena as multifunction of IL-6 (acutephase protein synthesis from hepatocytes, neural outgrowth of PC 12 cells, differentiation to macrophages, and immunoglobulin production in B cells) and bidirectional effects of cell growth (growth-inhibitory and -stirnulatory effects) depending on the cell type. Furthermore, selective inhibition of transmembrane signaling events may have therapeutic potential, as demonstrated in cyclosporin and FK506, which inhibit the synthesis of NFAT-1, a transcription factor involved in IL-2 gene expression in T cells. Much work remains to be done in defining the roles for IL-6 in embryogenesis and development, because IL-6 is expressed very early in embrogenesis. Through the development of knockout mice lacking IL-6 or the 1L-6 receptor complex, many of the unsolved issues about the roles of IL-6 in normal and different pathological conditions should be resolved in the near future.
ACKNOWLEDGMENTS We thank Ms. K. Kubota and Ms. K. Ono for their excellent secretarial assistance.
REFERENCES Aarden, L. A., Lansdorp, P., and DeCroot, E. R. (1985). A growth factor for B cell hybridomas produced by human monocytes. Lymphokines 10,175-185. Aarden, L. A., DeCroot, E. R., Schaap, 0. L., and Lansdorp, P. M. (1987). Production of hybridoma growth factor by human monocytes. Eur.]. Immunol. 17, 1411-1416. Abramowicz, D., Schandene, L., Goldman, M., Grusiaux, A,, Vereerstraeten, P., De Pauw, L., Wybran, L., Kinnaert, P., Dupont, E., and Toussaint, C. (1989). Release of tumor necrosis factor, interleukin-2, and gamma-interferon in serum after injection of OKT3 monoclonal antibody in kidney transplant recipients. Trunspluntotion 47,606608. Aderka, D., Le, J., and Vilcek, J. (1989). 1L-6 inhibits liposaccharide-induced tumor necrosis factor production in cultured human monocytes, U937 cells, and in mice. 1.lmmunol. 143,3517-3523. Akira, S., Hirano, T., Taga, T., and Kishimoto, T. (1990a). Biology of multifunctional cytokines: IL 6 and related molecules (IL 1 and TNF). FASEBJ. 4,2860-2867. Akira, S., Isshiki, H., Sugita, T., Tanabe, O., Kinoshita, S., Nishio, Y., Nakajima, T., Hirano, T., and Kishimoto, T. (1990b). A nuclear factor for IL-6 expression (NF-IL6) is a member of a ClEBP family. E M B O J . 9,1897-1906. Akira, S., and Kishimoto, T.(1992). IL-6 and NF-IL6 in acute-phase response and viral infection. Immunol. Rev. 127,25-50. Amadori, A., Zamarchi, R., Ciminale, V., Del Mistro, A., Siervo, S., Alberti, A., Colombatti, M., and Chieco-Bianchi. (1989). HIV-specific B cell activation: A major constituent of spontaneous B cell activation during HIV-infection. J . Immunol. 143, 21462152. Ambrosino, D. M.. Delaney, N. R., and Shamberger, R. C. (1990). Human polysaccharide-specific B cells are responsive to pokeweed mitogen and IL-6. J . Immunol. 144,1221-1226. Andus, T., Geiger, T., Hirano, T., Northoff, H., Canter, U.,Bauer, J., Kishimoto, T., and
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
53
Heinrich, P. C. (1987).Recombinant human B cell stimulatory factor 2 (BSF-2IIFN-p 2) regulates P-fibrinogen and albumin mRNA levels in Fao-9 cells. FEBS Lett. 221, 18-22. Apperley, J. F., Luskey, B. D., and Williams, D. A. (1991). Retroviral gene transfer of human adenosine deaminase in murine hematopoietic cells: Effect of selectable marker sequence on long-term expression. Blood 78,310-317. Asano, S., Okano, A., Ozawa, K., Nakahata, T., Ishibashi, T., Koike, K., Kimura, H., Tanioka, Y.,Shibuya, A,, Hirano, T., Kishimoto,T., Takaku, F., and Akiyama, Y.(1990). In oioo effects of recombinant human interleukin-6 in primates: Stimulated production of platelets. Blood 75,1602-1605. Astaldi, G. C. B., Janssen, M. C., Landsdorp, P. M., Willems, C., Zeijlemaker, W. P., and Oosterhof, F. (1980). Human endothelial culture supernatant (HECS): A growth factor for hybridomas. J . Immunol. 125, 1411-1414. Bataille, R., Jourdan, M., Zhang, X.G., and Klein, B. (1989). IL-6 is a potent growth factor for plasma cells and is elevated in overt myeloma and plasma cell leukemia. J . Clin. Inoest. 84,2008-2011. Bataille, R., Chappard, D., and Klein, B. (1992). The critical role of interleukin-6, interleukin-lp and macrophage colony-stimulating factor in the pathogenesis of bone lesions in multiple myeloma. Znt.]. Clin.Lab. Res. 21,283-287. Bauer, J., Bauer, T. M., Kalb, T.,Taga, T., Lengyel, G., Hirano, T., Kishimoto,T., Acs, G., Mayer, L., and Gerok, W. (1989). Regulation of interleukin 6 receptor expression in human monocytes and monocyte-derived rnacrophages.]. Exp. Med. 170,1537-1549. Bauer, J., and Herrmann, F. (1991). Interleukin-6 in clinical medicine. Ann Hemutol. 62, 203-2 10. Bauer, J., Strauss, S., Sohreiter-Gasser, U., Canter, U., Schlegel, P., Witt, I., Yolk, B., and Berger, M. (1991). Interleukin-6 and alpha-2-macroglobulin indicate an acute-phase state in Alzheimer’s disease cortices. FEBS Lett. 285, 111-114. Baumann, H., and Schendel, P. (1992). Interleukin-11 regulates the hepatic expression of the same plasma protein genes as interleukin-6.1. Biol. Chem. 266,20424-20427. Baumann, H., and Wong, G. G. (1989). Hepatocyte-stimulating factor I11 shares structural and functional indentity with leukemia-inhibitory fact0r.J. Zmmunol. 143,11631167. Baumann, M., Baumann, H., and Fey, G. H. (1990). Molecular cloning, characterization and functional expression of the rat liver interleukin-6 receptor. J . Biol. Chem. 265, 19853-19862. Bazan, J. F. (1990a). Haemopoietic receptors and helical cytokines. Zmmunol. Today 11, 350-354. Bazan, J. F. (1990b). Structural design and molecular evolution of a cytokine receptor superfamily. Proc. Nutl. Acud. Sci. U.S.A.87,6934-6938. Beagley, K. W., Eldridge, J. H., Lee, F., Kiyono, H., Everson, M. P., Koopman, W. J., Hirano, T., Kishimoto, T., and McGhee, J. R. (1989). Interleukins and IgA synthesis: Human and murine interleukin 6 induce high rate IgA secretion in IgA-committed B cel1s.J. E x p . Med. 169,2133-2148. Becker, S., Quay, J., and Soukup, J. (1991). Cytokine (tumor necrosis factor, IL-6, and IL-8) production by respiratory syncytial virus-infected human alveolar macrophages. J. Immunol. 147,4307-4312. Bendtzen, K., Buschard, K., Diamant, M., Horn, T., and Svenson, M. (1989). Possible role of IL-1, TNF-a, and IL-6 in insulin-dependent diabetes mellitus and autoimmune thyroid disease. Lymphokine Res. 8,335-340. Bhardwaj, N., Santhanam, U., Lau, L. L., Tatter, S. B., Ghrayeb, J., Rivelis, M., Steinman,
54
SHIZUO AKIRA ET AL.
R. M., Sehgal, P. B., and May, L. T. (1989). IL-GIIFN-pB in synovial effusions of patients with rheumatoid arthritis and other arthritides. J. Zmmunol. 143,2153-2159. Biondi, A., Rossi, V., Bassan, R., Barbui, T., Bettoni, S., Sironi, M., Mantovani, A., and Rambaldi, A. (1989). Constitutive expression of the interleukin-6 gene in chronic lymphocytic leukemia. Blood 73, 1279-1284. Birx, D., Redfield, R. R., Tencer, K., Fowler, A., Burke, D. S., and Tosato, G . (1990). Induction of interleukin-6 during human immunodeficiency virus infection. Blood 76, 2303-2310. Black, K., Garrett, I. R., and Mundy, G. R. (1991). Chinese hamster ovarian cells transfected with the murine interleukin-6 gene cause hypercalcemia as well as cachexia, leukocytosis and thrombocytosis in tumor-bearing nude mice. Endocrinology 128, 2657-2659. Blankenstein, T., Qin, Z., Li, W., and Diamantstein, T. (1990). DNA rearrangement and constitutive expression of the interleukin 6 gene in a mouse plasmacytoma. J . E x p . Med. 171,965-970. Negrier, S., Combaret, V., Attali, S., Goillot, E., Merrouche, Y.,Mercatello, A., Blay, J.-Y., Ravault, A., Tourani, J. M., Moskovtchenko, J. F., et al. (1992). Serum level ofinterleukin 6 as a prognosis factor in metastatic renal cell carcinoma. Cancer Res. 52,33173322. Bloemena, E., Ten Berge, I. J. M., Surachno, J.. and Wilmink, J. M. (1990). Kinetics of interleukin 6 during OKT3 treatment in renal allograft recipients. Transplantation 50, 330-331. Bodine, D. M., Karlsson, S., and Nienhuis, A. W. (1989). Combination of interleukin 3 and 6 preserves stem cell function in culture and enhances retrovirus-mediated gene transfer into hematopoietic stem cells. Proc. Natl. Acad. Sci. U.S.A.86,8897-8901. Bowcock, A. M., Kidd, J. R., Lathrop, M., Danshvar, L., May, L. T., Ray, A., Sehgal, P. B., Kidd, K. K., and Cavallisforza, L. L. (1988).The human “beta-2 interferon/hepatocyte stimulating factor/interleukin-6” gene: DNA polymorphism studies and localization to chromosome 7p21. Genomics 3,8-16. Brakenhoff, J. P., de Groot, E. R., Evers, R. F., Pannekoek, H., and Aarden, L. A. (1987). Molecular cloning and expression of hybridoma growth factor in Escherichia coli. J . Zmmunol. 139,4116-4121. Brakenhoff, J. P., Hart, M., and Aarden, L. A. (1989). Analysis of human IL-6 mutants expressed in Escherichia coli: Biologic activities are not affected by deletion of amino acids 1-28.J. Zmmunol. 143,1175-1182. Brandt, S. J., Bodine, D. M., Dunbar, C. E., and Nienhuis, A. W. (1990). Dysregulated interleukin 6 expression produces a syndrome resembling Castleman’s disease in mice.]. Clin. Invest. 86,592-599. Breen, E. C., Rezai, A. R., Nakajima, K., Beall, G. N., Mitsuyasu, R. T., Hirano, T., Kishimoto, T., and Martinez-Maza, 0. (1990). Infection with HIV is associated with elevated IL-6 levels and production. J . Zmmunol. 144,480-484. Brieva, J. A., Martin, R. A., Martinez-Maza, O., Kagan, J., Merrill, J., Saxon, A., Van Damme, J., and Stevens, R. H. (1990). Interleukin-6 is essential for antibody secretion by human in oioo antigen-induced lymphoblastoid B cells. Cell. Zmmunol. 130,303310. Briolay, J., Dechanet, J., Blanchard, D., Banchereau, J., and Miossec, P. (1992). Interleukin 4 inhibits polyclonal immunoglobulin secretion and cytokine production by peripheral blood mononuclear cells from rheumatoid arthritis patients.]. Clin. Immunol. 12,36-44. Broide, D. H., Lotz, M., Cuomo, A. J., Coburn, D. A., Federman, E.C., and Wasserman,
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
55
S. I. (1992). Cytokines in symptomatic asthma airways. J . Allergy Clin. Immunol. 89, 958-967. Brown, T. J., Rowe, J. M., Lie, J., and Shoyab, M. (1990). Regulation of IL-6 expression by oncostatin M.J. Immunol. 147,2175-2180. Burkert, U., von Ruden, T., and Wagner, E. F. (1991). Early fetal hematopoietic development from in oitro differentiated embryonic stem cells. New Biol. 3,698-708. Burstein, S. A., Downs, T., Friese, P., Lynam, S., Anderson, S., Henthorn, J., Epsteine, R. R., and Savage, K. (1992). Thrombocytosis in normal and sublethally irradiated dogs: Response to human interleukin 6. Blood 80,420-428. Calderon, T. M., Sherman, J., Wilkerson, H., Hatcher, V. B., and Berman, J. W. (1992). Interleukin 6 modulates c-sis gene expression in cultured human endothelial cells. Cell. Immunol. 143,118-126. Caligaris-Cappio, F., Bergui, L., Gregoretti, M. G., Gaidano, G., Gaboli, M., Schena, M., Zallone, A. Z.. and Marchisio, P. C. (1991). Role of bone marrow stromal cells in the growth of human multiple myeloma. Blood 77,2688-2693. Cambell, I. L., Cutr, A., Wilson, A., and Harrison, L. C. (1989). Evidence for IL-6 production by and effects on the pancreatic beta-cell. J . Immunol. 143, 1188-1191. Cambell, 1. L., Kay, T. W., Oxbrow, L., and Harrison, L. C. (1991). Essential role for interferon-y and interleukin-6 in autoimmune insulin-dependent disbetes in NOD/ Wehi mice.]. Clin. Inoest. 87,739-742. Chatenoud, L., Ferran, C., Reuter, A., Legendre, C., Gevaert, Y., Kreis, H., Franchimont, P., and Bach, J. F. (1989). Systemic reaction to the anti-T-cell monoclonal antibody OKT3 in relation to serum levels of tumor necrosis factor and interferon-gamma. N . Engl.J. Med. 320,1420-1421, Chen, L., Mory, Y., Zilberstein, A., and Revel, M. (1988). Growth inhibition of human breast carcinoma and leukemia/lymphoma cell lines by recombinant interferon /32. Proc. Natl. Acad. Sci. U.S.A. 85,8037-8041. Cheung, D. L., Hart, P. H., Vitti, G. F., Whitty, C. A., and Hamilton, J. A. (1990). Contrasting effects of interferon-y and interleukin-4 on the interleukin-6 activity of stimulated human monocytes. Immunology 71,70-75. Chiu, C. P., Lee, F., Fero, T. J., Johnson, A., Everitt, J., and Malik, A. B. (1989). IL-6 is a differentiation factor for M 1 and WEHI3B myeloid leukemic cells. J. Immunol. 142, 1909- 1915. Colottia, F., Sironi, M., Borre, A., Luini, W., Maddalena, F., and Mantovani, A. (1992). Interleukin 4 amplifies monocyte chemotactic protein and interleukin 6 production by endothelial cells. Cytokine 4,24-28. Content, J., De Wit, L., Pierard, D., Derynck, R., De Clercq, E., and Fiers, W. (1982). Secretory proteins induced in human fibroblasts under conditions used for the production of interferon fi. Proc. Natl. Acad. Sci. U.S.A. 79,2768-2772. Content, J., De Wit, L., Poupart, P., Opdenakker, G., Van Damme, J., and Billiau, A. (1985). Induction of a 26-kDa-protein mRNA in human cells treated with an interleukin-1-related, leukocyte-derived factor. Eur. J . Biochem. 152, 253257. Corbel, C., and Melchers, F. (1984). The synergism of accessory cells and of soluble factors derived from them in the activation of B cells to proliferation. Immunol. Rev. 78,51-74. Coulie, P. G., Stevens, M., and Van Snick, J. (1989). High- and low-affinity receptors for murine interleukin 6. Distinct distribution on B an T cells. Eur. J . Immunol. 19, 2107-21 14. Crowl, R. M., Stoller, T. J., Conroy, R. R., and Stoner, C. R. (1991). Induction of phospho-
56
SHIZUO AKIRA ET AL.
lipase A2 gene expression in human hepatoma cells by mediators of the acute phase response.]. Biol. Chem. 266,2647-2651. Davis, S., Aldrich, T. H., Valenzuela, D. M., Wong, V., Furth, M. E., Squinto, S. P., and Yancopoulos, G. D. (1991). The receptor for ciliary neurotrophic factor. Science 253, 59-63. Denis, M. (1992). Interleukin-6 in mouse hypersensitivity pneumonitis: Changes in lung free cells following depletion of endogenous IL-6 or direct administration of C . 52,197-201. IL-6.1. L ~ U C OBi01. Denis, M., and Gregg, E. 0. (1990). Recombinant tumor necrosis factor alpha decreases whereas recombinant interleukin 6 increases growth of a virulent strain of Mycobacterium avium in human macrophages. Zmmunology 71,139-141. De Simoni, M. G., Sironi, M., DeLuigi, A., Manfridi, A., Mantovani, A., and Ghezzi, P. (1990). Intracerebroventricular injection of interleukin 1 induces high circulating levels of interleukin 6.1. Exp. Med. 171,1773-1778. Devergne, O., Peuchmaur, M., Hunbert, M., Navratil, E., Leger-Ravet, M. B., Crevon, M. C., Petit, M. A., Galanaud, P., and Emilie, D. (1991). I n oioo expression of ILlbeta and IL-6 genes during viral infections in human. Eur. Cytokine Netw. 2, 183194. De Waal Malefyt, R., Abrams, J., Bennett, B., Figdor, C. G., and d e Vries, J. E. (1991). Interleukin 10 (IL-10) inhibits cytokine synthesis by human monocytes: An autoregulatory role of IL-10 produced by monocytres.]. E x p . Med. 174,1209-1220. Dick, J. E., Kamel-reid, S., Murdoch, B., and Doedens. M. (1991). Gene transfer into normal human hematopoietic cells using in uitro and in oioo assays. Blood 78,624634. Dohi, K., Iwano, M., Muraguchi, A., Horii, Y., Hirayama, T., Ogawa, S., Shiiki, H., Hirano, T., Kishimoto, T., and Ishikawa, H. (1991). The prognostic significance of urinary interleukin 6 in IgA nephropathy. Clin Nephrol. 35,l-5. Drexler, H., Hablawetz, E., Lu, W., Riede, U.,and Christes, A. (1992). Effects of inhibition of nitric oxide formation on regional blood flow in experimental myocardial infarction. Circulation 86,255-256. Duc-Goiran, G . P., Chany, C., and Doly, J. (1989). Unusually large interferon-a-like mRNAs and high expression of interleukin-6 in human fetal annexes.]. Biol. Chem. 26, 16507- 16511. Dutton, R. W., Falkoff, R., Hirst, J. A., Hoffman, M., Kappler, J. W., Kattman, J. R., Lesley, J. F., and Vann, D. (1971).Is there evidence for a nonantigen specific diffusable chemical mediator from the thymus cell in the initiation of the immune response? Prog. fmmunol. 1,355-368. Ensoli, B., Barillari, G., Salahuddin, S. Z., Gallo, R. C., and Wong-Staal, F. (1990). Tat protein of HIV-1 stimulates growth of cells derived from Kaposi’s-sarcoma lesions of AIDS patients. Nature (London)345,M-86. Entman, M. L., Michael, L., Rossen, R. D., Dreyer, W. F., Anderson, D. C.,Taylor,A. A., and Smita, C. W. (1991). Inflammation in the course of early myocardial ischemia. FASEB]. 5,2529-2537. Feingold, K. R., and Grunfeld, C. (1992). Role of cytokines in inducing hyperlipidemia. Diabetes 41,97-101. Ferguson-Smith, A. C., Chen, Y. F., Newman, M. S., May. L. T., Sehgal, P. D., and Ruddle, F. H. (1988). Regional localization of the interferon-P2R-cell stimdatory factor 2/hepatocyte stimulating factor gene to human chromosome 7p15-21. Genomics 2,203-208. Feyen, J. H. M., Elford, P., DiPadova, F. E., and Trechsel, U. (1989). Interleukin-6 is
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
57
produced by bone and modulated by parathyroid hormone. J. Bone Miner. Res. 4, 633-638. Fiecller, W., Weh, H. J., Suciu, E., Wittlief, C., Stocking, C., and Hossfeld, D. K. (1990). The IL-6 gene but not the IL-6 receptor gene is occasionally rearranged in patients with multiple myeloma. Leukemia 4,462-465. Finkel, M. S., Oddis, C. V., Jacob, T. D., Watkins, S. C., Hattler, B. G., and Simmons, R. L. (1992). Negative inotropic effects of cytokines on the heart mediated by nitric oxide. Science 257,387-389. Fiorentino, D. F., Zlotnik, A., Mosmann, T. R., Howard, M., and O’Garra, A. (1991). IL-10 inhibits cytokine production by activated macr0phages.J. lmmunol. 147,38153822. Firestein, G . S., Alvaro Gracia, J. M., Maki, R., Alvaro, and Gracia, J. M. (1990). Quantitative analysis of cytokine gene expression in rheumatoid arthritis. J . Zmmunol. 144, 3347-3353. Fleming, I., Gray, G. A., Julou-Schaeffer, G., Parratt, J. R., and Stoclet, J. C. (1990). Incubation with endotoxin activates the L-arginine pathway in vascular tissue. Biochem. Biophys. Res. Commun. 171,562-568. Flesch, I. E., and Kaufmann, S. H. (1990). Stimulation of antibacterial macrophage activities by B-cell stimulatory factor 2(interleukin 6). Infect. lmmun. 58,269-271. Freeman, G . J., Fredman, A. S., Rabinowe, S. N., Segil, J . M., Horowitz, J., Rosen, K., Whihnan, J. F., and Nadler, L. M. (1989). Interleukin 6 gene expression in normal and neoplastic B cells. J . Clin. Znoest. 83, 1512-1518. Frei, K., Leist, T. P., Meager, A., Gallo, P., Leppert, D., Zindernagel, R. M., and Fontana, A. (1988). Production of B cell stimulatory factor-2 and interferon y in the central nervous system during viral meningitis and encephalitis. J. Erp. Med. 168,449-453. Frei, K., Fredrikson, S., Fontana, A., and Link, H. (1991). Interleukin-6 is elevated in plasma in multiple sclerosis. J . Neuroimmunol. 31, 147-153. Fujihashi, K., McGhee, J. R., Lue, C., Beagley, K. W., Taga, T., Hirano, T., Kishimoto, T., Mestecky, J., and Kiyono, H. (1991). Human appendix B cells naturally express receptors for and respond to interleukin 6 with selective IgAl an IgA2 synthesis. J . Clin. lnuest. 88,248-252. Fukumoto, S., Matsumoto, T., Harada, S., Fujisaki, J., Kawano, M., and Ogata, E. (1991). Pheochroniocytoma with pyrexia and marked inflammatory signs: A paraneoplastic syndrome with possible relation to interleukin-6 production. J. Clin. Endocrinol. Metab. 73,877-881. Gallo, P., Frei, K., Rordorf, C., Lazdins, J., Tavolato, B., and Fontana, A. (1989). Human immunodeficiency virus type 1(HIV-1) infection of the central nervous system: An evaluation of cytokines in the cerebrospinal fluid. J . Neuroimmunol. 23, 109-1 16. Gallo, O., Bani, D., Toccafondi, G., Almerigogna, F., and Storchi, 0. F. (1992). Characterization of a novel cell line from pleomorphic adenoma of the parotid gland with niyoepithelial phenotype and producing interleukin-6 as an autocrine growth factor. Cancer 70,559-568. Canter, U., Strauss, S., Jonas, U., Weidemann, A., Beyreuther, K., Volk, B., Berger, M., and Bauer, J. (1991). a2-Macroglobulin synthesis in interleukin-6-stimulated human neuronal(SH-SY5Y neuroblastoma) cells. Potential significance for the processing of Alzheimer p-amyloid precursor protein. FEBS Lett. 282,127-131. Garman, R. D., Jacobs, K. A., Clark, S. C., and Raultet, D. H. (1987). B-cell-stimulatory factor 2(p2 interferon) functions as a second signal for interleukin 2 production by mature murine T cells. Proc. Notl. Acad. Sci. U.S.A.84,7629-7633. Gauldie, J . , Richards, C., Haarnish, D., Lansdorp, P., and Baumann, H. (1987). Interferon
58
SHIZUO AKIRA ET AL.
P2/B cell stimulating factor type 2 shares identity with monocyte-derived hepatocyte stimulating factor and regulates the major acute phase protein response in liver cells. Proc. Natl. Acad. Sci. U.S.A. 84,7251-7255. Gearing, D. P., and Bruce, A. G. (1992). Oncostatin M binds the high-affinity leukemia inhibitory factor receptor. Gearing, D. P., Cough, N. M., King, J. A., Hilton, D. J., Nicola, N. A., Simpson, R.J., Nice, E. C., Kelso, A., and Metcalf, D. (1987). Molecular cloning and expression of cDNA encoding a murine myeloid leukaemia inhibitory factor (LIF).EMBOJ. 6,3995-4002. Gearing, D. P., Thut, C . J., Vandenbos, T., Gimpel, S. D., Delaney, P. B., King, J., Price, V.,Cosman, D., and Beckmann, M. P. (1991). Leukemia inhibitory factor receptor is structurally related to the IL-6 signal transducer, gp130. EMBOJ. 10,2839-2848. Gearing, D. P., Comeau, M. R., Friend, D. J., Gimpel, S. D., Thut, C. J., McGourty, J., Brasher, K. K., King, J. A., Gillis, S., Mosley, B., Ziegler, S. F., and Cosman, D. (1992). The IL-6 signal transducer, gp130: an oncostatin M receptor and affinity converter for the LIF receptor. Science 255,1434-1437. Geiger, T., Andus, T., Klapproth, J., Hirano, T., Kishimoto, T., and Heinrich, P. C. (1988). Induction of rat acute-phase proteins by interleukin-6 in uiuo. Eur. J . Zmmunol. 18, 717-72 1. Geissler, K., Valent, P., Bettelheim, P., Sillaber, C., Wagner, B., Kyrle, P., Hinterberger, W., Lechner, K., Liehl, E., and Mayer, P. (1992). I n uiuo synergism of recombinant human interleukin-3 and recombinant human interleukin-6 on thrombopoiesis in primates. Blood 79,1155-1160. Gibbons, R.,Martinez, O., Matli, M., Heinzel, F., Bernstein, M., and Warren, R. (1990). Recombinant IL-4 inhibits IL-6 synthesis by adherent peripheral blood cells in oitro. Lymphokine Res. 9,283-293. Girasole, G., Jilka, R. L., Passeri, G., Boswell, S., Boder, G., Williams, D. C., and Monolagas, S. C. (1992). 17 Beta-estradiol inhibits interleukin-6 production by bone marrow-derived stromal cells and osteoclasts in oitro: A potential mechanism for the antiosteoporotic effect of estrogen. J . Clin. Invest. 89,883-891. Givon, T., Slavin, S., Haran-Ghera, N., Michalcviz, R.,and Revel, M. (1992). Antitumor effects of human recombinant interleukin-6 on acute myeloid leukemia in mice and in cell cultures. Blood 79,2392-2398. Gohsh, S., Gifford, A. M., Riviere, L. R., Tempest, P., Nolan, G. P., and Baltimore, D. (1990). Cloning of the p50 DNA subunit of the transcription factor NF-KB.p50 reveals remarkable homology at the amino terminal end to the oncogene v-rel. Cell 62, 1019-1029. Goldman, M., Gerard, C., Abramowicz, D., Schandene, L., Durez, P., De Pauw, L., Kinnaert, P., Vereerstreten, P., and Velu, T. (1992). Induction of interleukin-6 and interleukin-10 by the OKT3 monoclonal antibody: Possible relevance to posttransplant lymphoproliferative disorders. Clin. Transplant. 6,265-268. Greenberg, A. S., Nordan, R. P., McIntosh, J., Calvo, J. C., Scow, R. O., and Jablons, D. (1992). Interleukin 6 reduced lipoprotein lipase activity in adipose tissue of mice in uioo and in 3T3-Ll adipocytes: A possible role for interleukin 6 in cancer cachexia. Cancer Res. 52,4113-4116. Griffin, W. S. T., Stanley, L. C., Ling, C., White, L., MacLeod, V., Perrot, L. J., White, C. L., 111, and Araoz, C. (1989). Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc. Natl. Acad. Sci. U.S.A 86, 7611-7615. Grossman, R. M., Krueger, J., Yourish, D., Granelli-Piperno, A., Murphy, D. P., May,
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
59
L. T., Kupper, T. S., Sehgal, P. B., and Gottlieb, A. B. (1989). Interleukin 6 is expressed in high levels in psoriasis skin and stimulates proliferation of cultured human keratinocytes. Proc. Natl. Acad. Sci. U.S.A.86,6367-6371. Guerne, P.-A., Zuraw, B. L., Vaughan, J. €I., Carson, D. A., and Lotz, M. (1989). Synovium as a source of interleukin 6 i n oitro. Contribution to local and systemic manifestations of arthritis. J. Clin. lnoest. 83,585-592. Hack, C . E., de Groot, E. R., Felt-Bersma, R. J. F., Nuijens, J. H., Van Schijndel, R. J. M., Eerenberg-Belmer, A. J. M., Thijs, L. G., and Aarden, L. A. (1989). Increased plasma levels of interleukin-6 in sepsis. Blood 74, 1704-1710. Haegeman, G., Content, J., Volckaert, G., Derynck, R., Tavernier, J., and Fiers, W. (1986). Structural analysis of the sequence coding for an inducible 26-kDa protein in human fibroblasts. Eur. J . Biochem. 159,625-632. Hama, T., Miyamoto, M., Tsukui, H., Nishio, C., and Hatanaka, M. (1989). Interleukin-6 as a neurotrophic factor for promoting the survival ofcultured basal forebrain cholinergic neurons from postnatal rats. Neurosci. Lett. 104,340-344. Harris, J. F., Hawley, R. G., Hawley, T. S., and Crawford-Sharpe, G. C. (1992). Increased frequency of both total and specific monoclonal antibody producing hybridomas using a fusion partner that constitutively expresses recombinant IL-6. J . Immunol. Methods 148,199-207. Hasan, S. I., Blaney, B. A. D., and Yurk, J. L. (1992). Effect ofanticancer drugs on the release of interleukin-6 in vitro. Cancer lmmunol. lmmunother. 34,228-232. Hawley, R. G., Fong, A. Z., Burns, B. F., and Hawley, T. S. (1992). Transplantable myeloproliferative disease induced in mice by an interleukin 6 retr0virus.J. E x p . Med. 176,1149-1163. Hefeneider, S. H., Cornell, K. A., Brown, L. E., Bakke, A. C., McCoy, S. L., and Bennett, R. M. (1992). Nucleosomes and DNA bind to specific cell-surface molecules on murine cells and induce cytokine production. Clin. Zmniunol. Zmmunopathol. 63,245-251. Helfgott, D. C., May, L. T., Sthoeger, Z., Tamm, I., and Sehgal, P. B. (1987). Bacterial lipopolysaccharide (endotoxin) enhances expression and secretion ofP2 interferon by human fibroblasts. J . E x p . Med. 166, 1300-1309. Helfgott, D. C., Tatter, S. B., Santhanam, U., Clarick, R. H., Bhardwaj, N., May, L. T., and Sehgal, P. B. (1989). Multiple forms of IFN-PIIL-6 in serum and body fluids during an acute phase bacterial infection. J . lmmunol. 142,948-953. Helle, M., Brakenhoff, J. P. J., DeGroot, E. R., and Aarden, L. A. (1988). Interleukin-6 is involved in interleukin-1-induced activities. Eur. J . lmmunol. 18,957-959. Heney, D., Lewis, I. J., Evans, S. W., Banks, R., Bailey, C. C., and Whicher, J. T. (1992). Interleukin-6 and its relationship to c-reactive protein and fever in children with febrik neutropenia. J. Infect. Dis. 165,886-890. Heremans, H., Dillen, C., Put, W., Van Damme, J., and Billiau, A. (1992). Protective effect of anti-interleukin(IL)-6antibody against endotoxin, associated with paradoxically increased IL-6 levels. Eur.J. lmmunol. 22,2395-2401. Hermann, E., Fleischer, B., Mayet, W.-J., Poralla, T., and Meyer zuin Buschenfelde, K.-H. (1989). Correlation of synovial fluid interleukin 6 (IL-6) activities with IgG concentrations in patients with inflammatory joint disease and osteoarthritis. Clin. E x p . Rheumatol. 7,411-414. Herodin, F., Mestries, J.-C., Janodet, D., Martin, S., Mathieu, J., Gascon, M.-P., Pernin, M.-0.. and Ythier, A. (1992). Recombinant glycosylated human interleukin-6 accelerates peripheral blood platelet count recovery in radiation-induced bone marrow depression in baboons. Blood 80,688-695.
60
SHIZUO AKIRA ET AL.
Herrmann, F., Ardreeff, M., Gruss, H. J., Brach, M. A., Lubbert, M., and Mertelsmann, R. (1991). Interleukin-4 inhibits growth for multiple myelomas by suppressing interleukin-6 expression. Blood 78,2070-2074. Heslep, H. E., Bianchi, A. C., Cordingley, F. T.,Turner, M., Chandima, W., DeMel, C. P., Hoffbrand, A. V., and Brenner, M. K. (1990). Effects of interferon alpha on autocrine growth factor loops in B lymphoproliferative disorders. J . Exp. Med. 172, 17291734. Hibi, M., Murakami, M., Saito, M., Hirano, T., Taga, T., and Kishimoto, T. (1990). Molecular cloning and expression of an IL-6 signal transducer, gp130. Cell (Cambridge, Mass.) 63,1149-1157. Hilbert, D. M., Cancro, M. P., Scherle, P. A., Nordan, R. P., Van Snick, J., Gerhard, W., and Rudikoff, S. (1989). T cell derived IL-6 is differentially required for antigenspecific antibody secretion by primary and secondary B cel1s.J. lmmunol. 143,40194024. Hill, D. B., Marsano, L., Cohen, D., Allen, J., Shedlofsky, S., and McClain, C. J. (1992a). Increased plasma interleukin-6 concentrations in alcoholic hepatitis. J . Lab. Clin. Med. 119,547-552. Hill, J. A., Haimovici, F., and Anderson, D. J. (1987). Production of activated lymphocytes and macrophages inhibit mouse embryo development in uitro.]. Immunol. 139, 2250-2254. Hill, R.J., Warren, M. K., and Levin, J. (1990).Stimulation of thrombopoiesis in mice by human recombinant interleukin-6. J . Clin. Invest. 85, 1242-1247. Hill, R.J., Warren, M. K., Levin, J., and Gauldie, J. (1992b). Evidence that interleukin 6 does not play a role in the stimulation of platelet production after induction of acute thrombocytopenia. Blood 80,346-351. Hilton, D. J., and Cough, N. M. (1991). Leukemia inhibitory factor: A biological perspective. J. Cell Biochem. 46,21-26. Hirano, T., Teranishi, T., Lin, B., and Onoue, K. (1984). Human helper T cell factor(s) IV. Demonstration of a human late-acting B cell differentiation factor acting on Staphylococcus aureus Cowan I-stimulated B cells. J . Immunol. 133,798-802. Hirano, T., Taga, T., Nakano, N., Yasukawa, K., Kashiwamura, S., Shimizu, K., Nakajima, K., Pyun, K. H., and Kishimoto, T. (1985). Purification to homogeneity and characterization of human B-cell differentiation factor (BCDF or BSFp-2). Proc. Natl. Acad. Sci. U.S.A.85,5490-5494. Hirano, T., Yasukawa, K., Harada, H., Taga, T., Watanabe, Y., Matsuda, T., Kashiwamura, S.-I., Nakajima, K., Koyama, K., Iwamatsu, A., Tsunasawa, S., Sakiyama, F., Matsui, H., Takahara, Y., Taniguchi, T., and Kishimoto, T. (1986). Complementary DNA for a novel human interleukin (BSF-2) that induces B lymphocytes to produce immunoglobulin. Nature (London) 324,73-76. Hirano, T., Taga, T.,Yasukawa, K.,Nakajima, K., Nakano, N., Takatsuki, F., Shimizu, M., Murashima, A., Tsunasawa, S., Sakiyama, F., and Kishimoto, T. (1987). Human B cell differentiation factor defined by a anti-peptide antibody and its possible role in autoantibody production. Proc. Natl. Acad. Sci. U.S.A. 84,228-231. Hirano, T., Matsuda, T., Turner, M., Miyasaka, N., Buchan, G., Tang, B., Sato, K., Shimizu, M., Maini, R., Feldman, M., and Kishimoto, T. (1988). Excessive production of interleukin 6/B cell stimulatory factor-2 in rheumatoid arthritis. Eur.].lmmunol. 18, 1797- 1801. Hirano, T., Akira, S., Taga, T., and Kishimoto, T. (1990). Biological and clinical aspects of interleukin 6. lmmunol. Today 11,443-449. Hirata, Y., Taga, T., Hibi, M., Nakano, N., Hirano, T., and Kishimoto, T. (1989). Charac-
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
61
terization of IL-6 receptor expression by monoclonal and polyclonal antibodies. J . lmmunol. 143,2900-2906. Hirohata, S., and Miyamoto, T. (1990). Elevated levels of interleukin-6 in cerebrospinal fluid from patients with systemic lupus erythematosus and central nervous system involvement. Arthritis Rheum. 33,644-649. Honda, M., Kitamura, K., Mizutani, Y., Oishi, M., Arai, M., Okura, T., Igarhi, K., Yasukawa, K., Hirano, T., Kishimoto, T., Mitsuyasu, R., Chermann, J.-C., and Tokunaga, T. (1990). Quantitative analysis of serum IL-6 and its correlation with increased levels of serum IL-2R in HJV-induced diseases. J . Zmmunol. 145,4059-4064. Honda, M., Yamamoto, S., Cheng, M., Yasukawa, K., Suzuki, H., Saito, T., Osugi, Y., Tokunaga, T., and Kishimoto, T. (1992). Human soluble IL-6 receptor: Its detection and enhanced release by HIV infection.]. lmmunol. 148,2175-2180. Horii, Y., Muraguchi, A., Iwano, M., Matsuda, T., Hirayama, T., Yamada, H., Fujii, Y., Dohi, K., Ishikawa, H., Ohmoto, Y., Yoshizaki, K., Hirano, T., andKishinioto, T. (1989). Involvement of interleukin-6 in mesangial proliferative glomeru1onephritis.J. lmmunol. 143,3949-3955. Houssiau, F. A., Coulie, P. G., Olive, D., and Van Snick, J. (1988a). Synergistic activation of human T cells by interleukin 1 and interleukin 6. Eur.J.lmmunol. 18,653-656. Houssiau, F. A., Devogelaer, J. P., Van Damme, J., De Deuxchaisnes, C. N., and Van Snick, J. (1988b). Interleukin-6 in synovial fluid and serum of patients with rheumatoid arthritis and other inflammatory arthritides. Arthritis Rheum. 31, 784-788. Howell, G., Pham, P., Taylor, D., Foxwell, and Feldmann, M. (1991). Interleukin 4 induces interleukin 6 production by endothelial cells: Synergy with interferon-y. Eur. J . lmmunol. 21,97-101. Ikebuchi, K., Wong, G. G., Clark, S. C., Ihle, J. N., Hirai, Y., Ogawa, M. (1987). Interleukin-6 enhancement of interleukin-3-dependent proliferation of multipotential hemopoietic progenitors. Proc. Natl. Acad. Sci. U.S.A.84,9035-9039. Ikeda, U . , Ikeda, M., Oohara, T., Oguchi, A., Kamitani, T., Tsuruya, Y., and Kano, S. (1991). Interleukin-6 stimulates the growth of vascular smooth muscle cells in a PDGF-dependent manner. A m . J .Physiol. 260, H1713-1717. Ikeda, U.,Ikeda, M., Seino, Y., Takahasi, M., Kano, S., and Shimada, K. (1992a). Interleukin 6 gene transcripts are expressed in atherosclerotic lesions of genetically hyperlipidemic rabbits. Atherosclerosis 92,213-218. Ikeda, U.,Ohkawa, F., Seino, Y., Yamamoto, K., Hidaka, Y., Kasahara, T., Kawai, T., and Shimada, K. (1992b). Serum interleukin 6 levels become elevated in acute myocardial infarction. J . Mol. Cell. Cardiol. 24,579-584. Ip, N. Y., Nye, S. H., Boulton, T. G., Davis, S., Taga, T., Li, Y., Birren, S. J., Yasukawa, K., Kishimoto, T., Anderson, D. J., Stahl, N., and Yancopoulos, G . D. (1992). CNTF and LIF act on neuronal cells via shared signaling pathways that involve the IL-6 signal transducing receptor component gp130. Cell (Cambridge, Mass.) 69,1121-1 132. Ishibashi, T., Kimura, H., Shikawa, Y., Uchida, T., Kariyone, S., Hirano, T., Kishimoto, T., Takatsuki, F., and Akiyama, Y. (1989a). Interleukin 6 is a potent thrombopoietic factor in uioo in mice. Blood 74, 1241-1244. Ishibashi, T., Kiniura, H., Uchida, T., Kariyone, S., Friese, P., and Burstein, S. A. (198913). Human interleukin 6 is a direct promoter of maturation of megakaryocytes in uitro. Proc. Nut/. Acad. Sci. U.S.A.86,5953-5957. Ishimi, Y., Miyaura, C., HeJin, C., Akatsu, T., Abe, E., Nakmura, Y., Yamguchi, A., Yoshiki, S., Matsuda, T. A,, Hirano, T., Kishimoto, T., and Suda, T. (1990). IL-6 is produced by osteoblasts and induces bone resorption. J . lmmunol. 145, 32973303.
62
SHIZUO AKIRA ET AL.
Isobe, T., and Osserman, E. F. (1971). Pathologic conditions associated with plasma cell dyscrasias: A study of 806 cases. Ann. N.Y.Acad. Sci. 190,507-517. Isomaki, H. A., Hakulinen, T., and Joutsenlahti, U. (1978). Excess risk of lymphomas, leukemia and myeloma in patients with rheumatoid arthritis. J . Chron. Dis. 31,691696. Nakajima, T., Shimamoto, T., Hirano, T. and KishiIsshiki, H., Akira, S., Tanabe, 0.. moto, T. (1990).Constitutive and IL-1 inducible factors interact with the IL-1 responsive element in the IL-6 gene. Mol. Cell. Biol. 10,2757-2764. Jacobs, A. L., Sehgal, P. B., Julian, J., and Carson, D. P. (1992). Secretion and hormonal regulation of interleukin-6 production by mouse uterine stromal and polarized epithelial cells cultured in oitro. Endocrinology 131, 1037-1046. Jernberg-Wiklund, H., Pettersson, M., Carlsson, M., and Nilsson, K. (1992). Increase in interleukin 6 (IL-6) and IL-6 receptor expression in a human multiple myeloma cell line, U-266, during long-term in oitro culture and the development of a possible autocrine IL-6 loop. Leukemia 6,310-318. Jilka, R. L., Hangoc, G., Girasole, G., Passeri, G., Williams, D. C., Abrams, J. S., Boyce, B., Broxmeyer, H., and Manolagas, S. C. (1992). Increased osteoclast development after estrogen loss: Mediation by interleukin-6. Science 257,8841. Jirik, F. R., Podor, T. J., Hirano, T., Kishimoto, T., Loskutoff, J., Carson, D. A., and Lotz, M. (1989). Bacterial lipopolysaccharide and inflammatory mediators augment IL-6 secretion by human endothelial cells. J . Immunol. 142, 144-147. Jourdan, M., Battaille, R., Seguin, J., Zhang, X.G., Chaptal, R. A., and Klein, B. (1990). Constitutive production of interleukin-6 and immunologic features in cardiac myxomas. Arthritis Rheum. 33,398-402. Jourdan, M., Zhang, X.G., Portier, M., Boiron, J.-M., Bataille, R., and Klein, B. (1991). IFN-o! induces autocrine production of IL-6 in myeloma cell lines. J . Immunol. 147, 4402-4407. Judd, A. M., Spangelo, B. L., and MacLeod, R. M. (1990). Rat adrenal zonal glomerulosa cells produce interleukin 6. Prog. Neuroendocrinol. Immunol. 3,282-292. Kameda, T., Matsuzaki, N., Sawai, K., Okada. T., Saji, F., Matsuda, T., Hirano, T., Kishimoto, T., and Tanizawa, 0.(1990). Production of interleukin-6 by normal human trophoblast. Placenta 11,205-213. Kasahara, Y.,Miyawaki, T., Kato, K., Kanegane, H., Yachie, A., Yokoi, T., and Taniguchi, N. (1990). Role of interleukin 6 for differential responsiveness of naive and momory CD4'T cells in CD2-mediated activati0n.J. Exp. Med. 172, 1419-1424. Kaufmann, S. H., and Flesch, I. E. (1990). Cytokines in antibacterial resistance:possible applications for irnmunomodulation. Lung 168,1025-1032. Kawano, M., Hirano, T., Matsuda, T., Taga, T., Horii, Y.,Iwato, K., Asaoku, H., Tang, B., Tanabe, O., Tanaka, H., Kuramoto, A., and Kishimoto, T. (1988).Autocrine generation and requirement of BSF-B/ILS for human multiple myelomas. Nature (London)332, 83-85. Kawano, M., Tanaka, H., Ishikawa, H., et 01. (1989). Interleukin-1 accelerates autocrine growth of myeloma cells through interleukin-6 in human myeloma. Blood 73,21452148. Kern, P., Hemmer, C. J., Van Damme, J., Gruss, H. J., and Dietrich, M. (1989). Elevated tumor necrosis factor alpha and interleukin-6 serum levels as markers for complicated plasmodium falciparum malaria. Am. J . Med. 87,139-143. Kieran, M., Blank, V., Logeat, F., Vandekerckhove, J., Lottspeich, F., LeBail, O., Urban, M. B., Kourilsky, P., Baeuerle, P. A., and Israel, A. (1990).The DNA binding subunit of NF-kB is identical to factor KBF-1 and homologous to the re1 oncogene product. Cell (Cambridge,Mass.) 62,1007-1018.
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
63
Kikutani, H., Taga, T., Akira, S., Kishi, H., Miki, Y., Saiki, O., Yamamura, Y., and Kishimoto, T. (1985).Effect of B cell differentiation factor (BCDF) on biosynthesis and secretion of immunoglobulin molecules in human B cell lines. J. Immunol. 134, 990-995. Kirnbauer, R.,Kock, A., Neuner, P., Forster, E., Krutmann, J., Urbanski, A., Schauer, E., Ansel, J . C., Schwarz, T., and Luger, T. A. (1991). Regulation of epidermal cell interleukin-6 production by UV light and corticosteroids. /. Inuest. D e m a t o l . 96, 484-489. Kishimoto, T. (1985). Factors affecting B cell growth and differentiation. Annu. Rev. Zntmunol. 3, 133-157. Kishimoto, T. (1989).The biology of interleukin-6. Blood 74, 1-10. Kishimoto, T., and Ishizaka, K. (1973). Regulation of antibody response in oitro. VII. Enhancing soluble factors for IgC and IgE antibody response. /. Immunol. 111, 1194-1205. Kishimoto, T., Akira, S., and Taga, T. (1992).Interleukin-6 and its receptor: A paradigm for cytokines. Science 258,593-597. Kitahara, M., Kishimoto, S., Hirano, T., Kishimoto, T., and Okada, M. (1990).The in oioo anti-tumor effect of human recombinant interleukin-6. Jpn. J . Cancer Res. 81, 10321038. Kitani, A., Hara, M., Hirose, T., Norioka, K., Harigai, M., Hirose, W., Suzuki, K., Kawakami, M., Kawagoe, M., and Nakamura, H. (1989).Heterogeneity of B cell responsiveness to interleukin 4, interleukin 6 and low molecular weight B cell growth factor in discrete stages of B cell activation in patients with systemic lupus erythematosus. Clin. Exp. Immunol. 77,31-36. Kleiri, B., Zhang, X. G., Jourdan, M., Content, J., Houssiau, F., Aarden, L., Piechaczyk, M., and Bataille, R. (1989).Paracrine rather than autocrine regulation of myeloma-cell growth and differentiation by interleukin-6. Blood 73,517-526. Klein, B., Wijdenes, J., Zhang,X. G., Jourdan, M., Boiron, J. M., Brochier, J., Liautard, J., Merlin, M., Clement, C., and Mortel-Fourniel, B. (1991). Murine anti-interleukin-6 monoclonal antibody therapy for a patient with plasma cell leukemia. Blood 78, 1 198- 1204. Kobayashi, I., Matsuda, T., Saito, T., Yasukawa, K., Kikutani, H., Hirano, T., Taga, T., and Kishimoto, T. (1992).Abnormal distribution of IL-6 receptor in aged MRLllpr mice: Elevated expression on B cells and absence on CD4+ cells. Int. Immunol. 4, 14071412. Kohase, M., Henriksen-DeStefano, D., May, L. T., Vilcek, J., and Sehgal, P. B. (1986). Induction of @-interferon by tumor necrosis factor: A homeostatic mechanism in the control of cell proliferation. Cell (Cambridge, Mass.) 45,659-666. Koike, K., Nakahata, T., Takagi, M., Kobayashi, T., Ishiguro, A., Tsuji, K., Nagnuam, K., Okano, A., Akiyama, Y., and Akabane, T. (1988).Synergism of BSF-elinterleukin 6 and interleukin 3 on development ofmultipotential hemopoietic progenitors in serum-free culture. J. E x p . Med. 168,879-890. Koj, A. (1985). In “The Acute-Phase Response to Injury and Infection” (A. H. Cordon and A. Koji, eds.), Vol. 10, p. 139. Elsevier, AmsterdamINew York. Kreitman, R. J., Siegall, C. B., FitzCerald, D. J., Epstein, J., Barlogie, B., and Pastan, I. (1992).Interleukin-6 fused to a mutant form of Pseudomonas exotoxin kills malignant cells from patients with multiple myeloma. Blood 79, 1775-1780. Kriittgen, A., Rose-John, S., Moller, C., Wroblowshi, B., Woller,A., Miillberg, J., Hirano, T., Kishimoto, T., and Heinrich, P. C. (1990). Structure-function analysis of human interleukin-6: Evidence for the involvement of the carboxy-terminus in function. FEBS Lett. 273,95-98.
64
SHIZUO AKlRA ET AL.
Kurihara, N., Bertolini, D., Suda, T., Akiyam, Y., and Roodman, G . D. (1990). IL-6 stimulates osteoclast-like multinucleated cell formation in long term human marrow cultures by inducing IL-1 release. J. Immunol. 144,4226-4230. Kushner, I. (1982). The phenomenon of the acute phase response. Ann. NY Acad. Sci. 389,39-48. Landschulz, W. H., Johnson, P. F. and McKnight, S. L. (1988a). The leucine zipper: A hypothetical structure common to a new class of DNA binding proteins. Science 240, 1759- 1764. Landschulz, W. H., Johnson, P. F., Adashi, E. Y., Graves, B. J., and McKnight, S. L. (1988b). Isolation of a recombinant copy of the gene encoding C/EBP. Genes Deu. 2, 786-800. Lane, H. C., Masur, H., Edgar, L. C., Whalen, G., Rook, A. H., and Fauci, A. (1983). Abnormalities of B e l l activation and immunoregulation in patients with the acquired immunodeficiency syndrome. N. Engl. J. Med. 309,453-458. Le, J.. Fredrickson, G., Reis, L., Diamanstein, T., Hirano, T., Kishimoto, T., and Vilcek, J. (1988). Interleukin-2-dependent and interleukin-2-independent pathways of regulation of thymocyte function by interleukin 6. Proc. Natl. Acad. Sci. U.S.A. 85,509514. Lee, J. D., Swischer, S. G., Minehart, E. H., McBride, W. H., and Economou, J. S. (1990). Interleukin-4 downregulates interleukin-6 production in human peripheral blood mononuclear cells. J . Leukocyte Biol. 47,475-479. Leeuwenberg, J. F. M., Von Asmuth, E. J. V., Jeuhomme, T. M. A., and Buurman, W. A. (1990). IFN-y regulates the expression of the adhesion molecule ELAM-1 and IL-6 production by human endothelial cells in oitr0.J. Immunol. 145,2110-2114. LeMay, L. G . ,Vander, A. J,, and Kluger, M. J. (1990). Role ofinterleukin 6 in fever in rats. Am. J . Physiol. 258, R798-R803. Lenardo, M. J., and Baltimore, D. (1989). NF-KB:A pleiotropic mediator ofinducible and tissue-specific gene control. Cell (Cambridge,Mass.)58,227-229. Leu, S.-J. C., and Singh, U. K. (1992). Stimulation of interleukin-6 production by corticotropin-releasing factor. Cell. Immunol. 143,220-227. Leutz, A., Damm, K., Sterneck, E., Kowenz, E., Ness, S., Frank, R.,Gausepohl, H., Pan, Y. C. E., Samrt, J., Hayman, M., and Graf, T. (1989). Molecular cloning of the chicken niyelomonocytic growth factor (cMGF)reveals relationship to interleukin-6 and granulocyte colony stimulating factor. EMBOJ. 8, 175-181. Levy, Y.,Tsapis, A., and Brouet, J. C. (1991). Interleukin-6 antisense oligonucleotides inhibit the growth of human myeloma cell 1ines.J. Clin. Invest. 88,696-699. Liberman, T. A., and Baltimore, D. (1990). Activation of interleukin-6 gene expression through the NF-KBtranscription factor. Mol. Cell. Biol. 10,2327-2334. Lieberman, A. P., Pitha, P. M., Shin, H. S., and Shin, M. L. (1989). Production oftrimor neurosis factor and other cytokines by astrocyte stimulated with lipopolysaccharide or a neurotropic virus. Proc. Natl. Acad. Sci. U.S.A.86,6348-6352. Linker-Israeli, M., and Deans, R. (1989). Dysregulated lymphkine production in systemic lupus erythematosus. Ann. NY Acad. Sci. 557,567-571. Linker-Israeli, M., Deans, R. J., Wallace, D. J., Prehn, J., Ozeri-Chen, T., and Klinenberg, J. R. (1991). Elevated levels of endogenous IL-6 in systemic lupus erythematosus: A putative role in pathogenesis.]. Inmunol. 147, 117-123. Liu, J., Modrell, B., Aruffo, A., Marken, J. S., Taga, T., Yasukawa, K., Murakanii, M., Kishimoto, T., and Shoyab, M. (1992). Interleukin-6 signal transducer gp130 mediates oncostatin M signaling. J . Biol. Chem. 267,16763-16766. Loppnow, H., and Libby, P. (1990). Proliferating or interleukin 1-activated human
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
65
vascular smooth muscle cells secrete copious interleukin 6. J . Cliti. Itioest. 85, 731738. Lorre, K., Van Danime, J., Venvilghem, J., Baroja, M. L., and Ceuppens, J. L. (1990). IL-6 is an accessory signal in the alternative CD2-mediated pathway o f T cell activation. J. Znimunol. 144,4681-4687. Lotem, J., Shabo, Y., and Sachs, L. (1989). Clonal variation in susceptibility to differentiation by different protein inducers i n the myeloid leukemia cell line M1. Leukemia 3, 804-807. Lotz, M., and Guerne, P.-A. (1991). Interleukin-6 induces the synthesis of tissue inhibitor of nietalloproteinases-l/erythroidpotentiating activity (TIMP-l/EPA).J . B i d . Chein. 266,2647-2651. Lotz, M., Jirik, F., Kabouridis, R., Tsoukas, C., Hirano, T., and Kishimoto, T. (1988). BSF-2/IL-6 is a co-stimulant for human thymocytes and T lymphocytes. J . E x p . Med. 167,1253-1258. Lowik, C. W. G. M.,van der Pluijm, G., Blyoys, H., Hoekman, K., Bijvoet, 0. L. M., Aarden, L. A., and Papapoulos, S. E. (1989). Parathyroid hormone (PTH) and PTH-like protein (PLP) stimulate interleukin-6 production by osteogenic cells: A possible role of interleukin-6 in osteoclastogenesis. Biochem. Biophys. Res. Coinniun. 162, 15461552. Lu, C., Vickers, M. F., and Kerbel, R. S. (1992). Interleukin 6: A fibroblast-derived growth inhibitor of human melanoma cells from early but not advanced stages oftumor progression. Proc. Nutl. Acud. Sci. U.S.A.89,9215-9219. Ludwig, H.. Nachbaur, D. M.,Fritz, E., Kramer, M., and Huber, H. (1991). Interleukin-6 is a prognostic factor in multiple myeloma. Blood 77,2794-2795. Lue, C., Kiyono, H., McGhee, J. R., Fjihashi, K., Kishimoto, T., Hirano, T., and Mestecky, J. (1991). Recombinant human interleukin-6 (rhIL-6) promotes the terminal differentiation of in oioo-activated human B cells into antibody-secreting cells. Cell Inrmuiiol. 132,423-432. Luger, T. A, Krutman, J., Kirnbauer, R., Urbanski, A., Schwarz, T., Klappacher, G., Kock, A,, Micksche, M.,Malejczy, K. J., Schauer, E., May, L. T., and Sehgal, P. €3. (1989). IFNPIIL-6 augments the activity of human natural killer cells. J . Imtnutiol. 143, 1206- 1209. Luskey, B. D., Rosenblatt, M.,Zsebo, K., and Williams, D. A. (1992). Stem cell factor, interleukin-3, and interleukin-6 promote retroviral-mediated gene transfer into murine hematopoietic stem cells. Blood 80,396-402. Lust, A. J., Donovan, K. A., Kline, M. P., Greipp, P. R., Kyle, R. A., and Maihle, N. J. (1992). Isolation of an mRNA encoding a soluble form of the human interleukin-6 receptor. Cytokine 4,96-100. Lyson, K., Milenkovic, and McCann, S. M. (1991).The stimulatory effect ofinterleukin-6 on corticotropin-releasing factor and thyrotropin-releasing hornione secretion i n oitro. Prog. Neuroendocrinol. Imniunol. 4, 161-165. Mackiewicz, A,, Ganapathi, M. K., Schultz, D., Brabenec, A., Weinstein, J., Kelley, M.F., and Kushner, I. (1990). Transforming growth factor /3 1 regulates production ofacutephase proteins. Proc. Natl. Acad. Sci. U.S.A. 87, 1491-1495. Maimone, D., Gregory, S., Arnason, B. G., and Reder, A. T. (1991).Cytokine levels in the cerebrospinal fluid and serum of patients with multiple sclerosis. J . Neuroiminunol. 32,67-74. Martinez-Maza, O., Crabb, E., Mitsuyasu, R. T., Fahey, J. L., and Giorgi, J. V. (1987). Infection with the human immunodeficiency virus (HIV) is associated with an in oioo increase in B lymphocyte activation and immaturity. J . Immunol. 108,3720-3725.
66
SHIZUO AKIRA ET AL.
Maruo, N., Morita, I., Shirao, M., and Murota, S. (1992). IL-6 increases endothelial permeability in oitro. Endocrinology 131,710-714. Matsuda, T., Hirano, T., and Kishimoto, T. (1988).Establishment of a IL-6lBSF-2dependent cell line and preparation ofanti-IL-6lBSF-2 monoclonal antibodies. Eur.J. lmmunol. 18,951-956. Mattoli, S., Mattoso, V. L., Soloperto, M., Allegra, L., and Fasoli, A. (1991). Cellular and biochemical characteristics of bronchoalveolar lavage fluid in symptomatic nonallergic asthma. J . Allergy Clin. lmmunol. 87,794-802. May, L. T., Grayeb, J., Santhanam, U., Tatter, S . B., Sthoeger, Z., Helfgott, D. C., Chiorazzi, N., Grieninger, G., and Sehgal, P. B. (1988,). Synthesis and secretion of multiple forms of p2-interferonlB-cell differentiation factor 2lhepatocyte-stimulating factor by human fibroblasts and monocytes. J . Biol. Chem. 263,7760-7766. May, L. T., Santhanam, U., Tatter, S. B., Helfgott, D. C., Ray, A., Ghrayeb, J., and Sehgal, P. B. (1988b). Phosphorylation of secreted forms of human P2-interferonlhepatocytestimulating factorlinterleukin-6. Biochem. Biophys. Res. Commun. 152, 1144-1 150. McCauley, R. L., Chopra, V., Li, Y.-Y., Herndon, D. N., and Robson, M. C. (1992). Altered cytokine production in black patients with keloids. J . Clin. Zmmunol. 12, 300-308. McCee, D. W., Beagley, K. W., Aicher, W. K., and McCee, J. R. (1992). Transforming growth factor-p enhances interleukin-6 secretion by intestinal epithelial cells. lmmunology 77,7-12. McIntosh, J. K., Jablons, D. M., Mule, J. J., Nordan, R. P., Rudikoff, S., Lotz, M. T., and Rosenberg, S. A. (1989). In oioo induction of IL-6 by administration of exogenous cytokines and detection of de nooo serum levels of IL-6 in tumor-bearing mice. J . lmmunol. 143, 162-167. Metcalf, D. (1974). The serum factor stimulating colony formation in oitro by murine plasniacytoma cells: Response to antigens and mineral oil. J . lmmunol. 113, 235243. Miki, S., Iwano, M., Miki, Y., Yamamoto, M., Tang, B., Yokokawa, K., Sonoda, T., Hirano, T., and Kishimoto, T. (1989). Interleukin-6 (IL-6) functions as an in oitro autocrine growth factor in renal cell carcinomas. FEBS Lett. 250,607-610. Miles, S. A., Rezai, A. R., Salazar-Gonzales, J. F., Meyden, M. V., Stevens, R. H., Logan, D. M., Mitsuyasu, R. T., Taga, T., Hirano, T., Kishimoto, T., and Martinez-Maza, 0. (1990). AIDS Kaposi sarcoma-derived cells produce and respond to interleukin 6. Proc. Natl. Acad. Sci. U.S.A. 87,4068-4072. Rezai, A., Magpantay, L., Salahuddin, S. Z., Nakamura, Miles, S . A., Martinez-Maza, 0.. S., Radka, S. F., and Linsley, P. (1992). Oncostatin-M as a potent mitogen for AIDSKaposi sarcoma-derived cells. Science 255, 1432-1434. Miossec, P., Briolay, J., Dechanet, J., Wijdenes, J., Martines-Valdez, H., and Banchereau, J. (1992). Inhibition of the production of proinflammatory cytokines and immunoglobulines by interleukin-4 in an ex oiuo model of rheumatoid synovitis. Arthritis Rheum. 35,874-883. Miyaura, C., Onozaki, K., Akiyama, Y., Taniyama, T., Hirano, T., Kishimoto, T., and Suda, T. (1988). Recombinant human interleukin 6 (B-cell stirnulatory factor 2) is a potent inducer of differentiation of mouse myeloid leukemia cells ( M l ) . FEBS Lett. 234,17-21. Miyaura, C., Jin, C. H., Yamaguchi, Y., Tomida, M., Hozumi, M., Matsuda, T., Hirano, T., Kishimoto, T., and Suda, T. (1989). Production of interleukin 6 and its relation to the macrophage differentiation of mouse myeloid leukemia cells ( M l ) treated with differentiation-inducingfactor and la-dihydroxyvitamin D3. Biochem. Biophys. Res. Commun. 158,660-666.
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
67
Mock, B. A., Nordan, R. P., Justice, M. J., Kozak, C., Jenkins, N. A., Copeland, N. G., Clark, S. C., Wong, G . G., and Rudikoff, S. (1989). The murine IL-6 gene maps to the proximal region of chromosome 5. J . Zmmunol. 142,1372-1376. Molyneux, M. E., Taylor, T. E., Wirima, J. J.. and Grau, G . E. (1991). Tumor necrosis factor, interleukin-6, and malaria. Lancet 337, 1098. Morinaga, Y., Suzuki, H., Takatsuki, F., Akiyama, Y.,Taniyama,T., Matsushima, K., and Onozaki, K. (1989). Contribution of IL-6 to the anti-proliferative effect of IL-1 and tumor necrosis factor on tumor cell lines. J . Immunol. 143,4538-3542. Morrone, G., Cortese, R., and Sorrentino, V. (1989). Post-transcriptional control of negative acute phase genes by transforming growth factor beta. EMBOJ. 8,3767-3771. Motro, B., Itin, A., Sachs, L., and Keshet, E. (1990). Pattern of interleukin 6 gene expression in oioo suggests a role for this cytokine in angiogenesis. Proc. Natl. Acad. Sci. U.S.A. 87,3092-3096. Muench, M. O., Schneider, J. G . , and Moore, M. A. S. (1992). Interaction among colony stimulating factors, IL-lb, IL-6, and Kit-ligand in the regulation of primitive niurine hematopoietic cells. Exp. Hematol. 20,339-349. Mul’e, J. J., McIntosh, J. K., Jablons, D. M., and Rosenberg, S. A. (1990). Antitumor activity of recombinant interleukin 6 in mice. J . E x p . Med. 171,629-636. Mul’e, J. J., Custer, M. C., Travis, W. D., and Rosenberg, S. A. (1992). Cellular mechanisms of the antitumor activity of recombinant IL-6 in mice.J. Zmmunol. 148,26222629. Muraguchi, A,, Kishinioto, T., Miki, Y., Kuritani, T., Kaieda, T., Yoshizaki, K., and Yamamura, Y. (1981). T cell-replacing factor (TRF)-inducedIgG secretion in human B blastoid cell line and demonstration of acceptors for TRF.]. Zmmunol. 127,412-416. Muraguchi, A., Hirdno, T., Tang, B., Matsuda, T., Horii, Y., Nakajima, K., and Kishinioto, T. (1988).The essential role ofB cell stimulatory factor 2 (BSF-2IIL-6) for the terminal differentiation of B cel1s.J. E x p . Med. 167,332-344. Murakami, M., Narazaki, M., Hibi, M., Yawata, H., Yasukawa, K., Hamaguchi, M., Taga, T., and Kishimoto, T. (1991). Critical cytoplasmic region of the IL-6 signal transducer, gp130, is conserved in the cytokine receptor family. Proc. Nutl. Acnd. Sci. U.S.A.88, 11i349-11353. Murray, R., Lee, F., and Chiu, C.-P. (1990). The gene for leukemia inhibitory factor and interleukin-6 are expressed in mouse blastocysts prior to the onset of hemopoiesis. Mol. Cel. Biol. 10,4953-4956. Musso, T., Espinoza-Delgado,J., Pulkki, K., Gussela, G. L., Longo, D. L., and Varesio, L. (1990). Transforming growth factor beta downregulates interleukin-1-induced IL-6 production by human monocytes. Blood 76,2466-2469. Nabatd, T., Morimoto, S., Koh, E., Shiraishi, T., and Ogihara, T. (1990). Interleukin-6 stimulates c-myc expression and proliferation of cultured vascular smooth muscle cells. Biocheni. Int. 20,445-453. Nair, B. C., DeVico,A. L., Nakamura, S., Copeland, T. D., Chen,Y., Patel, A., O’Neil, T., Oroszlan, S., Gallo, R. C., and Sarngadharan, M. G. (1992). Identification of a major growth factor for AIDS-Kaposi’s sarcoma cells as oncostatin M. Science 255, 14301432. Naitoh, Y., Fukata, J., Tominaga, T., Tamai, S., Mori, K., and Imura, H. (1988). Interleukin-6 stimulates the secretion ofadrenocorticotropic hormone in conscious, freelymoving rats. Biochem. Biophys. Res. Commun. 155, 1459-1463. Nakafuku, M., Satoh, T., and Kaziro, Y. (1992). Differentiation factors, including nerve growth factor, fibroblast growth factor, and interleukin-6, induce accumulation of an active ras-GTP complex in rat pheochromocytoma PC12 cells. J . B i d . Cheni. 267, 19448-19454.
68
SHIZUO AKIRA ET AL.
Nakahama, H., Okada, M., Miyazaki, M., Imai, N., Yokokawa, T., and Kubori, S. (1992). Distinct responses of interleukin-6 and other laboratory parameters to treatment in a patient with polyarteritis nodosa-A case report. Angiology 43,512-516. Nakajima, K., and Wall, R. (1991). Interleukin-6 signals activating junB and tisll gene transcription in a B-cell hybridoma. Mol. Cell. Biol. 11, 1409-1418. Nakajima, K., Martinez-Maza. O., Hirano, T., Breen, E. C., Nishanian P. G . , SalazarGonzalez, J. F., Fahey, J. L., and Kishimoto, T. (1989). Induction of IL-6 (B cell stimulatory factor 2/IFNIP2) production by HIV. J. lmmunol. 142, 144-147. Nakajima, T., Kinoshita, S., Sasagawa, T., Sasaki, K., Naruto, M., Kishimoto, T., and Akira, S. (1993). Phosphorylation at threonine 235 by a ras-dependent MAP kinase canscade is essential for transcriptional activation of NF-IL6. Proc. N a t . Acad. Sci. U.S.A.,90,2207-2211. Namba, Y., and Hanaoka, M. (1972). Immunocytology of cultured IgM-forming cells of mouse. I. Requirement of phagocytic cell factor and its role in antibody formation. J . lmmunol. 109,1193-1200. Natsuka, S . , Akira, S., Sugita, T.,Hashimoto, S., Nishio, Y., Isshiki, H., and Kishimoto, T. (1992). Macrophage differentiation specific expression of NF-IL6, a transcription factor for IL-6. Blood 79,460-466. Navarra, P., Tsagarakis, S., Faria, M., Rees, L. H., Besser, M., and Grossman, A. B. (1990). Interleukin-1 and -6 stimulate the release of corticotropin-releasing hormone-41 from rat hypothalamus in oitro via eicosanoid cyclooxygenase pathway. Endocrinology 128, 37-44. Navarro, S., Debili, N., LeCouedic, J. P., Klein, B., Breton-Gorius, J., Doly, J., and Vainchenker, W. (1991). Interleukin-6 and its receptor are expressed by human niegakaryocytes: In oitro effects on proliferation and endoreplication. Blood 77,461-471. Nemunaitis, J.,Andrew, F. D. F., Mochizuki, D. Y., Lilly, M. B., and Singer, J . W. (1989). Human marrow stromal cells: response to interleukin-6 and control of IL-6 expression. Blood 74,1929-1935. Nesbitt, J. E., and Fuller, G . M. (1992). Differential regulation of interleukin-6 receptor and gp130 gene expression in rat hepatocytes. Mol. Biol. Cell 3, 103-112. Neta, R., Vogel, S. N., Sipe, J. D., Wong, G. G., and Nordan, R. P. (1988). Comparison of in oioo effects of human recombinant IL1 and human recombinant IL6 i n mice. Lymphokine Res. 7,403-412. Neuner, P., Nurbani, A., Trautinger, F., Moler, A., Kirnbauer, R., Kapp, A., Schopf, E., Schwarz, T., and Lugar, T. A. (1991). Increased IL-6 production by monocytes and keratinocytes in patients with psoriasis. J . lnoest. Dermatol. 97,27-33. Neurath, A. R., Strick, N., and Sproul, P. (1992). Search for hepatitis B virus ccll receptors reveals binding site for interleukin 6 on the virus envelope protein.J. E x p . Med. 175, 46 1-469. Nijsten, M. W. N., DeCroot, E. R., TenDuis, H. J.. Klasen, H. J., Hack, C. E., and Aarden, L. A. (1987). Serum levels of interleukin-6 and acute phase responses. Lancet 2,921. Nishimoto, N., Yoshizaki, K., Tagoh, M., Monden, S., Kishimoto, T., and Hirano, T. (1989). Elevation of serum interleukin-6 prior to acute phase proteins on the inflammation by surgical operation. Clin. lmmunol. lmmunopathol. 50,399-401. Nishimoto, N., Yoshizaki, K., Eiraku, N., Hachigashira, K.,Tagoh, H., Ogata, A., Kuritani, T., Osame, M., and Kishimoto, T. (1990). Elevated levels ofinterleukin-6 in serum and cerebrospinal fluid of HTLV-I-associated myelopathy/tropical spastic parapasesis. J . Neurol. Sci. 97, 183-193. Nishino, E., Matsuzaki, N., Masuhiro, K., Kameda, T., Taniguchi, T., Takagi, T., Sagi, F., and Tanizawa, 0. (1989). Trophoblast-derived interleukin-6 (IL-6) regulates hunian
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
69
chorionic gonadotropin release through IL-6 receptor on human trophoblasts. J . Clin. E ndocrinol. Metabol. 71,536-541. Nolan, G. P., Ghosh, S., Liou, H.-C., Tempst, P., and Baltimore, D. (1991). DNA binding and IKBinhibition of the cloned p65 subunit of NF-KB,a related polypeptide. Cell 64, 961-969. Nonia, T., Mizuta, T., Rosen, A., Hirano, T., Kishimoto, T., and Honjo, T. (1987). Enhancement of the interleukin-2 receptor expression on T cells by multiple Blyinphotropic lymphokines. lmmunol. Lett. 15,249-253. Nordan, R., and Potter, M. (1986). A macrophage-derived factor required by plasniacytomas for survival and proliferation in oitro. Science 233,566-569. Nordan, R., Pumphrey, J., and Rudikoff, S . (1987). Purification and NHz-terminal amino acid sequence of a plasmacytoma growth factor derived from the murine macrophage cell line, P388D1. J . lmmunol. 139,813-817. Northemann, W., Braciak, T. A., Hattori, M., Lee, F., and Fey, G. H. (1989). Structure of the rat interleukin 6 gene and its expression in macrophage-derived cells. J. B i d . Chem. 264,16072-16082. Novick, D., Engelmann, H., Wallach, D., and Rubinstein, M. (1989). Soluble cytokiiie receptors are present in normal human urine../. E x p . Med. 170, 1409-1414. Novick, D., Engelmann. H., Revel, M., Leitner, 0..and Rubinstein, M. (1991). Monoclonal antibodies to the soluble human IL-6 receptor: affinity purification, ELISA, and inhibition of ligand binding. Hybridoma 10, 137-146. Nussler, A., Pied, S . , Goma, J., Renia, L., Miltgen, F., Grau, G. E., and Mazier, D. (1991). TNF inhibits malaria hepatic stages in oitro via synthesis of IL-6. lnt. lmmunol. 3, 317-321. Ohbo, K., Suganiura, K., Sekizawa, T., and Kogure, K. (1991). Interleukin-6 in cerebrospinal fluid of HTLV-1-associated myelopathy. Neurology 41,594-595. Okatfa, M., Kitahara, M., Kishimoto, S., Matsuda, T., Hirano, T., and Kishimoto, T. (1988). IL-6/BSF-2 functions as a killer helper factor in the in oitro induction of cytotoxic T cells. J . lmmunol. 141, 1543-1549. Okaiio, A., Suzuki, C., Takatsuki, F., Akiyama, Y.,Koike, K., Nakahata, T., Hirano, T., Kishimoto, T., Ozawa, K., and Asano, S. (1989). In oitro expansion of inurine pluripotent hemopoietic stem cell population in response to interleukin 3 and interleukin 6: Application to bone marrow transplantation. Transplantation 48, 495-498. Oster, W., Cicco, N. A., Klein, H., Hirano, T., Kishimoto, T., Lindenmann, A., Mertelsniann, H. H., and Herrmann, F. (1989). Participation ofthe cytokines interleukin-6, tumor necrosis factor-alpha, and interferon-beta secreted by acute myelogenous leukeniia blasts in autocrine and paracrine leukemia growth control. /. Clin. Invest. 84, 451-457. Pancre, V., Monte, D., Delanoye, A., Capron, A., and Auriault, C. (1990). Interleukin-6 is the main mediator of the interaction between monocytes and platelets in the killing of Schistosoma mansoni. Eur. Cytokine Network 1, 15-19. Patchen, M. L., MacVittie, T. J., Williams, J . L., Schwartz, G. N., and Souza, L. M. (1991). Administration of interleukin-6 stimulates multilineage hematopoiesis and accelerates recovery from radiation-induced hematopoietic depression. Blood 77,472-480. Patthy, L. (1990) Honiology ofa domain ofthe growth hormone/prolactin receptor family with type 111 modules of fibronectin. Cell (Cambridge,Mass) 61, 13-14. Paul, S.R., Bennett, F., Calvetti, J. A., Kelleher, K.,Wood, C. R., O'Hara, R. M. Jr, Leary, A. C., Sibley, B., Clark, S. C., Williams, D. A., and Yang, Y.-C. (1990). Molecular cloning of a cDNA encoding interleukin 11, a stromal cell-derived lymphopoietic and heniatopoietic cytokine. Proc. Natl. Acad. Sci. U.S.A. 87,7512-7516.
70
SHIZUO AKIRA ET AL.
Pelton, B. K., Hylton, W., and Denman, A. M. (1992). Activation of IL-6 production by UV irradiation of blood mononuclear cells from patients with systemic lupus erythematosus. Clin. E x p . Zmmunol. 89,251-254. Pied, S., Renia, L., Nussler, A., Miltgen, F., and Mazier, D. (1991). Inhibitory activity of IL-6 on malaria hepatic stages. Parasite Zmmunol. 13,211-217. Poli, G., Bressler, P., Kinter, A., Duh, E., Timmer, W. C., Rabson, A., Justement, J. S., Stanky, S., and Fauci, A. S. (1990a). Interleukin-6 induces human immunodeficiency virus expression in infected monocytic cells alone and in synergy with TNF-a by transcriptional and post-transcriptional mechanisms. J . E x p . Med. 172, 151-158. Poli, V., Mancini, F. P., and Cortese, R. (1990b). IL-GDBP, a nuclear protein involved i n interleukin-6 signal transduction, defines a new family of leucine zipper proteins related to C/EBP. Cell (Cambridge,Mass) 63,643-653. Porgador, A., Tzehoval, E., Katz, A., Vadai, E., Revel, M., Feldman, M., and Eisenbach, L. (1992). Interleukin 6 gene transfection into Lewis lung carcinoma tumor cells suppresses the malignant phenotype and confer immunotherapeutic competence against parental metastatic cells. Cancer Res. 52,3679-3686. Potter, M., and Boyce, C. (1962). Induction of plasma cell neoplasms in strain Balb/c mice with mineral oil and mineral oil adjuvants. Nature (London)193, 1086-1087. Potter, M., and Wax, J. S. (1981). Genetics of susceptibility to pristane-induced plasmacytomas in BALB/cAn: reduced susceptibility in BALB/cJ with a brief description of pristane-induced arthritis. /. Zmmunol. 127, 1591-1595. Potter, M., Wax, J. S., Anderson, A. D., and Nordan, R. P. (1985). Inhibition of plasmacytoma development in BALB/c mice by ind0methacin.J. E x p . Med. 161,996-1012. Preiser, J. C., Schmartz, D., Van der Linden, P., Content, J., Vanden-Russche, P., Buurman, W., Sebald, W., Dupont, E., Pinsky, M. R.,andVincent, J. L. (1991). Interleukin-6 administration has no acute hemodynamic or hematologic effects in the dog. Cytokine 3, 1-4. Ray, A., Sassone-Corsi, P., and Sehgal, P. B. (1989). A multiple cytokine- and second messenger-responsive element in the enhancer of the human interleukin-6 gene: Similarities with c-fos gene regulation. Mol. Cell. Biol. 9,5537-5547. Ray, A., LaForge, K. S., and Sehgal, P. B. (1991). On the mechanism for efficient repression of the interleukin-6 promoter by glucocorticoids: Enhancer, TATA box, and RNA start site (Inr motif) occlusion. Mol. Cel. Biol. 10,5736-5746. Renauld, J. C., Vink, A., and Van Snick, J. (1989). I L l and IL6 in CTL inductionaccessory signals in murine cytolytic T cell responses: Dual requirement for IL1 arid IL6. J . Zmmunol. 143,1894-1898. Rennick, D., Jackson, J.. Yang, G., Wideman, J.. Lee, F., and Hudak, S. (1989). Interleukin-6 interacts with interleukin-4 and other heniatopoietic growth factors to selectively enhance the growth of megakaryocytic, erythroid, niyeloid, and multipotential progenitor cells. Blood 73, 1828. Richard, C. D., Brown, T. J., Shoyab, M., Baumann, H., and Gauldie, J. (1992). Recombinant oncostatin M stimulates the production of acute phase proteins in HepC2 cells and rat primary hepatocytes in oitro.]. Zmmunol. 148, 1731-1736. Riedy, M. C., and Stewart, C. C. (1992). Inhibitory role of interleukin-6 in macrophage proliferation. J . Leukocyte Biol. 52, 125-127. Robbins, P. D., Horowitz, J. M., and Mulligan, R. C. (1990). Negative regulation ot' human c-fos expression by the retinoblastomal gene product. Nature (London)346, 668-67 1. Romero, R., Avila, C., Santhanam, U., and Sehgal, P. B. (1990). Amniotic fluid IL-6 i n preterm labor: Association with infection. J . Clin. Znoest. 85, 1392- 1400.
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
71
Romero, R., Sepulveda, W., Kenney, J. S., Archer, L. E., Allison, A. C., and Sehgal, P. B. (1992). Interleukin 6 determination in the detection of microbial invasion of the amniotic cavity. I n “Polyfunctional Cytokines: IL-6 and LIF,” Ciba Found. Symp. 167, pp. 205-223. Wiley, Chicherster. Roodman, G. D., Kurihara, N., Ohsaki, Y., Kuritd, A., Hosking, D., Demulder, A., Smith, J. F., and Singer, F. R. (1992). Interleukin 6. A potential autocrine/paracrine factor in Paget’s disease of bone. J. Clin. lnoest. 89,46-52. Rose, T. M., and Bruce, A. G. (1991). Oncostatin M is a member of a cytokine family that includes leukemia-inhibitory factor, granulocyte colony-stimulating factor, and inter88,8641-8645. leukin 6. Proc. Nutl. Acud. Sci. U.S.A. Rose-John, S., Schooltink, H., Lenz, D., Hipp, E., Dufhues, G., Schmitz, H., Schiel, X., Hirano, T., Kishimoto, T., and Heinrich, P. C. (1990). Studies on the structure and regulation of the human hepatic interleukin-6 receptor. E u r . J . Biochem. 190,79-83. Rothstein, J. L., Johnson, D., DeLoia, J. A., Skowronski, J., Solter, D., and Knowles, B. (1992). Gene expression during preiniplantation mouse development. Genes Deo. 6, 1 190- 1201. Ruef, C., Budde, K., Lacy, J., Northemann, W., Baumann, M., Sterzel, R. B., and Coleman, D. L. (1990). Interleukin 6 is an autocrine growth factor for niesangial cell. Kidney 1nt. 38,249-257. Rumore, R. M., and Steinman, C. R. (1990). Endogenous circulating DNA in systemic lupus erythematosus. 1.Clin.lnoest. 86,69-74. Russell, S. L. (1990).Lymphokine gene therapy for cancer. Irnniunol. Today 11, 196200. Saito, M., Yoshida, K., Hibi, M., Taga, T., and Kishimoto, T. (1992). Molecular cloning of a murine IL-6 receptor-associated signal transducer, gp130, and its regulated expression i n oioo. J . fnimunol. 148,4066-4071. Salas, M. A., Evans, S. W., Levell, M. J., and Whicher, J. T. (1990). Interleukin-6 and ACTH act synergistically to stimulate the release of croticosterone from adrenal gland cells. Clin. E x p . lmmutiol. 79,470-473. Sanceau, J., Beranger, F., Gaudelet, C., Wieetzerbin, J . (1989). IFN-y is an essential cosignal for triggering IFN-@2/BSF-2gene expression in human nionocytic cell lines. Ann. N . Y . Acud. Sci. 557, 130-143. Santhanam, U., Avila, C., Romero, R., Vignet, H., Ida, N., Sakurai, S., and Sehgal, P. B. (l99la). Cytokines in normal and abnormal partition: Elevated amniotic fluid interleukin-6 levels in women with premature rupture of membranes associated with intrauterine infection. Cytokines 3, 155-163. Santhanam, O., Ray, A., and Sehgal, P. B. (1991b). Repression ofthe interleukin 6 gene promoter by p53 and the retinoblastoma susceptibility gene product. Proc. Nutl. Acud. Sci. U.S.A. 88, 7605-7609. Sato, T., Ito, A., and Mori, Y. (1990). Interleukin 6 enhances the production of tissue inhibitor of metalloproteinases (TIMP) but not that of matrix metalloproteinases by human fibroblasts. Biochem. Biophys. Res. Comtnun. 170,824-829. Satoh, T., Nakaniura, S., Taga, T., Matsuda, T., Hirano, T., Kishimoto, T., and Kaziro, Y. (1988). Induction of neural differentiation in PC12 cells by B cell stimulatory factor 2/interleukin 6. Mol. Cell. Biol. 8,3546-3549. Scala. G . , Kuang, Y. D., Hall, R. E., Muchmore, A. V., and Gyenheim, J. J. (1985). Accessory cell function of human B cells. I. Production of both interleukin-1 like activity and an interleukin-1 inhibitory factor by an EBV-transformed human B cell line. J . E x p . Med. 159, 1637-1652. Scala, G., Quinto, T., Ruocco, M. R., Arcucci, A., Mallardo, M., Caretto, P., Forni, G., and
72
SHIZUO AKIRA ET AL.
Venuta, S. (1990). Expression of an exogenous interleukin 6 gene in human Epstein Barr virus B cells growth advantage and i n uiuo tumorigenicity. ./. E x p . Med. 172, 61-68. Schieven, G. L., Kallestad, J. C., Brown, T. J., Ledbetter, J. A., and Linsley, P. S. (1992). Oncostatin M induces tyrosine phosphorylation in endothelial cells and activation of p62YeStyrosine kinase. ./. Zmmunol. 149, 1676-1682. Schimple, A., and Wecker, E. (1972). Replacement ofT cell function by a T cell product. Nature (London)237,15-17. Schindler, R., Mancilla, J., Ghorbani, R., Clark, S. C., and Dinarello, C. A. (1990). Correlations and interactions in the production of interleukin-6 (IL-6). IL-1. and tumor necrosis factor (TNF) in human blood mononuclear cells: IL-6 suppresses IL-1 and TNF. Blood 75,40-47. Schmitt, R. M., Bruyn, E., and Snodgrass, H. R. (1991). Hemapopoietic development of embyonic stem cells: Cytokine and receptor gene expression. Genes Deu. 5,728-740. Schmitter, D., Lauber, B., Fagg, B., and Stahel, R. A. (1992). Heniatopoietic growth factors secreted by seven human pleural mesothelioma cell lines: interleukin-6 production as a common feature. Znt. ./. Cancer 51, 196-301. Schooltinck, H., Schmitz-Van derleur, H., Heinrich, P. C., and Rose-John, S. (1992). Up-regulation of the interleukin-6-signal transducing protein (gp130) by interleukin-6 and dexamethasone in HepG2 cells. F E B S J . 3,263-265. Schwab, G . , Siegall, C. B., Aarden, L. A., Neckers, L. M., and Nordan, R. P. (1991). Characterization of an interleukin-6-mediated autocrine growth loop in the human niultiple myeloma cell line, U266. Blood 77,587-593. Sehgal, P. B. (1990). Interleukin-6: a regulator of plasma protein gene expression in hepatic and non-hepatic tissues. Mol. Biol. Med. 7, 117-130. Sehgal, P. B., Walther, Z., and Tamm, I. (1987a). Rapid enhancement ofP2-IFNIBSF-2 gene expression in human fibroblasts by diacylglycerols and the calcium ionophore A23187. Proc. Natl. Acad. Sci. U.S.A.84,3633-3667. Sehgal, P. B., Zilberstein, A., Ruggieri, R. M., May, L. T., Ferguson-Smith, A,, Late, D. L., Revel, M., and Ruddle, F. (1987b). Human chromosome 7 carries the p2 interferon gene. Proc. Natl. Acud. Sci. U.S.A.84,3633-3637. Sehgd, P. B., Helfgott, D. C., Santhanam, U., Tatter, S. B., Calarick, R. H., Ghrayeb, J., and May, L. T. (1988). Regulation of the acute phase and immune responses in viral disease: Enhanced expression of the /32-interferonlhepatocyte-stimulating factor/ interleukin 6 gene in virus-infected human fibroblasts. ./. E x p . Med. 167, 19511956. Semer, D., Reisler, K., MacDonald, P. C., and Casey, M. L. (1991). Responsiveness of human endometrial stromal cells to cytokines. Ann. N.Y. Acad. Sci. 622,99-110. Shabo, Y.,Lotem, J., Rubinstein, M., Revel, M., Clark, S. C., Wolf, S. F., Kamen, R., and Sachs, L. (1988). The myeloid blood cell differentiation-inducing protein MGI-2A is interleukin 6. Blood 72,2070-2073. Shabo, Y., Lotem, J., and Sachs, L. (1989). Autoregulation of interleukin 6 and granulocyte-macrophage colony-stimulating factor in the differentiation of myeloid leukemic cells. Mol. Cell. B i d . 9,4109-4112. Shacter, E., Arzadon, G. K., and Williams, J. (1992). Elevation of interleukin-6 in response to a chronic inflammatory stimulus in mice: Inhibition by indomethacin. Blood 80,194-202. Shaw, G., and Kamen, R. (1986). A conserved A U sequence from the 3’ untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell (Cantbridge, M U S S )46,659-667. Shenkin, A., Fraser, W. D., Series, J., Winstanley, F. P., McCartney, A. C., Burns,
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
73
H. J. G., and Van Danime, J. (1989). The serum interleukin-6 response to elective surgery. Lymphokine Res. 8, 123-127. Sheron, N., Bird, G., Goka, J., Alexander, G., and Williams, R. (1991). Elevated plasma interleukin-6 and increased severity and mortality in alcoholic hepatitis. Clin. E x p . lmmunol. 84,449-453. Shimizu, S., Hirano, T., Yoshioka, R., Sugai, S., Matsuda, T., Taga, T., Kishimoto, T., and Konda, S. (1988). Interleukin-6 (B-cell stimulatory factor-2)-dependent growth of a Lennert’s lymphoma-derived T-cell line(KT-3). Blood 72, 1826-1828. Shimizu, H., Mitomo, K., Watanabe, T., Okamoto, S., and Yamamoto, K-I. (1990). Involvement of a NF-KB like transcription factor in the activation of the interleukin-6 gene by inflammatory lymphokines. Mol. Cell. Bio2. 10,561-568. Shiratsuchi, H., Johnson, J. L., and Ellner, J. J. (1991). Bidirectional effects ofcytokines on the growth of Mycobacterium avium within human monocytes. 1.Immunol. 146, 3165-3 170. Sidell, N., Taga, T., Hiano, T., Kishimoto, T., and Saxon, A. (1991). Retinoic acid-induced growth inhibition ofa human myeloma cell line via down-regulation ofIL-6 receptors. J . Immunol. 146,3809-3814. Siegall, C. B., Chaudhary, V. K., FitzGerald, D. J., and Pastan, I. (1988). Cytotoxic activity of an interleukin 6-Pseudomonas exotoxin fusion protein on human myeloma cells. Proc. Natl. Acad. Sci. U.S.A.85,9738-9742. Siegall, C. B., Schwab, G., Nordan, R. P., FitzGerald, D. J., and Pastan, I. (1990). Expression of the interleukin 6 receptor and interleukin 6 in prostate carcinoma cells. Cancer Res. 50,7786-7788. Siegall, C. B., Kreitman, R. J., FitzGerald, D. J., and Pastan, I. (1991). Anti-tumor effects of IL-6-Pseudomonas exotoxin chimeric molecules against the human hepatocellular carcinoma PLC/PRF/5 in mice. Cancer Res. 51,2831-2836. Simpson, R. J., Moritz, R. L., Van Roost, E., and Van Snick, J. (1988).Characterization of a recombinant murine interleukin-6: Assignment ofdisulfide bonds. Biochem. Biophys. Res. Commun. 157,364-372. Sironi, M., Breviariao, F., Proserpio, P., Biondi, A., Vecchi, A., Van Danime, J., Dejana, E., and Mantovani, A. (1989). IL-I stimulates IL-6 production in endothelial cells. 1.lmmunol. 142,549-553. Smeland, E. B., Blonihoff, H. K., Funderud, S., Shalaby, M. R., and Espevik, T. (1989). Interleukin 4 induces selective production of interleukin 6 from normal human B lymphocytes.]. E x p . Med. 170, 1463-1468. Smith, R. J., Justen, J. M., Ulrich, R. G., Lund, J. E., and Sam, L. M. (1992). Induction of neutral proteins and prostanoid production in bovine nasal chondrocytes by interleukin-1 and tumor necrosis factor a:Modulation of these cellular responses by interleukin-6 and platelet-derived growth factor. Clin. Immunol. Ininiunopathol. 64, 135-144. Snouwaert, J. N., Kariya, K., and Fowlkes, D. M. (1991). Effects ofsite-specific mutations on biologic activities of recombinant human IL-6. 146,585-591. Snyers, L., and Content, J. (1992). Enhancement of IL-6 receptor /3 chain (gp130)expression by IL-6, IL-1 and TNF in human epithelial cells. Biochem. Biophys. Res. Commun. 185,902-908. Snyers, L., De Wit, L., and Content, J. (1990). Glucocorticoid up-regulation of highaffinity interleukin 6 receptors on human epithelial cells. Proc. Natl. Acad. Sci. U.S.A. 87, 2838-2842. Spangelo, B. L., and MacLeod, R. M. (1990). Regulation ofthe acute phase response and neuroendocrine function by interleukin 6. Prog. Neuroendocrinol. Immunol. 3, 167175.
74
SHIZUO AKIRA ET AL
Spangelo, B. L., Judd, A. M., Isakson, P., and MacLeod, R. M. (1989). Interleukin-6 stimulates anterior pituitary hormone release in uitro. Endocrinology 125,575-577. Starnes, H. F., Pearce, M. K., Twari, A., Yim, J. M. ,ZOU,J.-C., and Abrams, J. S. (1990). Anti-monoclonal antibodies protect against lethal Escherichia coli infection and lethal tumor necrosis factor (I challenge in mice.J. Zmmunol. 145,4185-4191. Stevens, C., Walz, G., Singaram, C., Lipman, M. L., Zanker, B., Muggia, A., Antonioli, D., Peppercorn, M. A., and Strom, T. B. (1992).Tumor necrosis factor-alpha, interleukin-1 beta, and interleukin-6 expression in inflammatory bowel disease. Dig. Dis. Sci. 37, 818-826. Stokli, K. A,, Lottspeich, F., Sendtner, M., Masaiakowski, P., Carroll, P., Gotz, R., Lindholm, D., and Thoenen, H. (1989). Molecular cloning, expression and regional distribution of rat ciliary neurotrophic factor. Nature (London)342,920-923. Straneva, J. E., vann Besien, K. W., Derigs, G., and Hoffman, R.(1992). Is interleukin 6 the physiological regulator of thrombopoiesis? E x p . Hematol. 20,47-50. Strassmann, G., Fong, M., Kenney, J. S., and Jacob, C. 0. (1992). Evidence for the involvement of interleukin 6 in experimental cancer cachexia. 1. Clin. Invest. 89, 1681- 1684. Suematsu, S., Matsuda, T., Aozasa, K., Akira, S., Nakano, N., Ohno, S., Miyazaki, J., Yamamura, K., Hirano, T., and Kishimoto, T. (1989). IgGl plasmacytosis in interleukin-6 transgenic mice. Proc. Natl. Acad. Sci. U.S.A. 86,7547-7551. Suematsu, S., Matsusaka, T.,Matsuda, T., Ohno, S., Miyazaki, J., Yamamura, K., Hirano, T., and Kishimoto, T. (1992). Generation of plasmacytomas with the chromosomal translocations t(12;15) in IL-6 transgenic mice. Proc. Natl. Acad. Sci. U.S.A. 89,232235. Sugita, T., Totsuka, M., Saito, M., Yarnasaki, K., Taga, T., Hirano, T., and Kishimoto, T. (1990). Functional murine interleukin 6 receptor with the intracisternal A particle gene product at its cytoplasmic domain: Its possible role in plasmacytomagenesis. J . Exp. Med. 171,2001-2009. Sun, Y., Tokushige, K., Isono, E., Yamguchi, K., and Obata, H. (1992).Elevated seruni interleukin-6 levels in patients with acute hepatitis. J . Clin. Zmmunol. 12, 197-200. Suzuki, C., Okano, A., Takatsuki, F., Miyasaka, Y., Hirano, T., Kishimoto, T., Ejima, D., and Akiyama, Y. (1989). Continuous perfusion with interleukin 6(IL-6) enhances production of hematopoietic stem cells(CFU-s). Biochem. Biophys. Res. Commtcn. 159,933-938. Swaak, A. J., van Rooyen, A., and Aarden, L. A. (1989). Interleukin-6 (IL-6) and acute phase proteins in the disease course of patients with systemic lupus erythematosus. Rheumotol. Znt. 8,263-268. Swinnen, L. J., Costanzo-Nordin, M. R.,Fisher, S. G., O’Sullivan, E. J., Johnson, M. R., Heroux, A. L., Dizikes, G. J., Pifarre, R.,and Fisher, R. I. (1990). Increased incidence of lymphoporliferative disorder after immunosuppression with the monoclonal antibody OKT3 in cardiac-transplant recipients. N. Eng1.J. Med. 323, 1723-1728. Szeberenyi, J.. Cai, H., and Cooper, G. Y. (1990). Effect of a dominant inhibitory Ha-ras mutation on neuronal differentiation of PC12 cells. Mol. Cell. Biol. 10,5324-5332. Tabibzadeh, S. S., Santhanam, U., Sehgal, P. B., and May, L. T. (1989).Cytokine-induced production of IFN-/32/156 by freshly expanted human endometrial stromal cells. J . Zmmunol. 142,3134-3139. Taga, T., and Kishimoto, T. (1992).Cytokine receptors and signal transduction. FASEBJ. 6,3387-3396. Taga, T., Kawanishi, T., Hardy, R. R.,Hirono, T., and Kishimoto, T. (1987). Receptors for B cell stimulatory factor 2 (BSF-2): Quantitation, specificity, distribution, and regulation of their expression. J . Exp. Med. 166,967-981.
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
75
Taga, T., Hibi, M., Hirata, Y., Yamasaki, K., Yasukawa, K., Matsuda, T., Hirano, T., and Kishimoto, T. (1989). Interleukin-6 triggers the association of its receptor with a possible signal transducer, gp130. Cell (Cambridge,Mass) 58,573-581. Taga, T., Narazaki, M., Yasukawa, K., Saito, T., Miki, D., Hamaguchi, M., Davis, S., Shoyab, M., Yancopoulos, G. D., and Kishimoto, T. (1992). Functional inhibition of hematopoietic- and neurotrophic-cytokines (LIF, OM, and CNTF) by blocking IL-6 signal transducer, gp130. Proc. Natl. Acad. Sci. U.S.A. 89, 10998-11001. Takai, Y., Wong, G. G., Clark, S. C., Burakoff, S . J., and Herrmann, S. H. (1988). B cell stimulatory factor-2 is involved in the differentiation of cytotoxic T lymphocytes. 1.lmmunol. 140,508-512. Takai, Y., Seki, N., Senoh, H., Yokota, T., Lee, F., Hamaoka, T., and Fiijiwara, H. (1989). Enhanced production of interleukin-6 in mice with type I1 collagen-induced arthritis. Arthritis Rheum. 32,594-600. Takatsuki, F., Okano, A., Suzuki, C., Chieda, R., Takahara, Y., Hirano, T., Kishinioto, T., Hamuro, J., and Akiyama, Y. (1988). Human recombinant IL-G/R cell-stimulatory factor 2 augments inurine antigen-specific antibody responses in oitro and in oioo. 1.Itntnunol. 141,3072-3077. Takatsuki, F., Okano, A., Suzuki, C., Miyasaka, Y., Hirano, T., Kishinioto, T., Ejinia, D., and Akiyama, Y. (1990). Interleukin 6 perfusion stimulates reconstitution of the immune and hematopoietic systems after 5-fluorouracil treatment. Cancer Res. 50,28852890. Tam, I., Cardinale, I., Krueger, J., Murphy, J. S., May, L. T., and Sehgal, P. B. (1989). Interleukin 6 decreases cell-cell association and increases motility of ductal breast carcinoma cells.]. E x p . Med. 170, 1649-1669. Tanabe, O., Akira, S . , Kamiya, T., Wong, G. G., Hirano, T., and Kishimoto, T. (1988). Genomic structure ofthe murine IL-6 gene. High degree of conservation of potential regulatory sequences between mouse and human.]. lmmunol. 141,3875-3881. Tanaka, Y., Saito, K., Shirakawa, F., Ota, T., Suzuki, H., Eto, S., andyamashita, U. (1988). Production of B cell stimulating factors by B cells in patients with SLE.]. lmmunol. 141,3043-3049. Tang, B., Matsuda, T., Akira, S., Nagata, N., Ikehara, S., Hirano, T., and Kishimoto, T. (1991). Age-associated increase in interleukin 6 in MRL/lpr mice. I n t . lmmuiiol. 3, 273-278. Taylor, C. W., Grogan, T. M., and Salmon, S . E. (1990).Effects of interleukin-4 on the in uitro growth of human lymphoid and plasma cell neoplasms. Blood 75, 1114-1 118. Te Velde AA, Huijbens R. J. F., Hieje K, De Vries, J. E., and Figdor, C. F. (1990). Interleukin-4 (IL-4) inhibits secretion of IL-IP, tumor necrosis factor-cr and IL-6 by human monocytes. Blood 76,1392- 1397. Teranishi, T., Hirano, T., Naoniichi, A., and Onoue, K. (1982). Human helper T cell factor(s) (ThF). 11. Induction of IgG production in B lymphoblastoid cell lines and identification of T cell-replacing factor-(THF) like factor(s). J . Inln~unol.128, 193198.
Tohyama, N., Karasuyama, H., and Tada, T. (1990). Growth autonomy and tuniorigenicity of' interleukin-6 dependent B cells transfected with interleukin-6 cDNA. 1. E x p . Med. 171,389-400. Tosato, G., Seamon, K. B., Goldman, N. D., Sehgal, P. B., May, L. T., Washington, G. C., Jones, K. D., and Pike, S. E. (1988). Monocyte-derived human B-cell growth factor identified as interferon-P2(BSF-2, IL-6). Science 239,502-504. Tosato, G., Tanner, J., Jones, K. D., Revel, M., and Pike, S . E. (1990). Identification of interleukin-6 as an aiitocrine growth factor for Epstein-Barr virus-imrnortalized B cells.]. Virol. 64,3033-3041.
76
SHIZUO AKIRA ET AL.
Tracey, K. J., Wei, H., Manogue, K. R., Fong, Y., Hesse, D. G., Nguyen, H. T., Kuo, G. C., Beutler, B., Cotran, R. S., Cerami, A., and Lowry, S. F. (1988). Cachectinkumor necrosis factor induces cachexia, anemia, and inflammation. J . E x p . Med. 167, 12111227. Truksen, K., Kupper, T. A., Degenstein, L., Williams, I., and Fuchs, E. (1992). Interleukin 6: insights to its function in skin by overexpression in transgenic mice. Proc. Natl. Acad. Sci. U.S.A.89,5068-5072. Tsang, K. Y., Finch, M. D., Primus, F. J., and Schlom, J. (1991). Human reconibinant interleukin-6 enhances antibody-dependent cellular cytotoxicity of human tumor cells mediated by human peripheral blood niononuclear cells. Cancer Zmmunol. Zmmunother. 34,9-16. Ulich, T. R., del Castillo, J., and Guo, K. (1989). I n oioo hematologic effect of IL-6 on hematopoiesis and circulating numbers of RBCs with WBCs. Blood 73, 108-1 10. Ulich, T. R., Guo, K., Remick, D., Del Cstillo, and Yin, S. (1991). Endotoxin-induced cytokine gene expression in oioo. I11 IL-6 mRNA and serum protein expression, and the in oioo hematologic effects of IL-6. J . Zmmunol. 146,2316-2323. Umeshita, K., Monden, M., Tono, T., Hasuike, Y., Kanai, T., Gotoh, M., Mori, T., Seu, P., Shaked, A., and Busuttil, R. W. (1992). Interleukin-6 in bile as a marker of allograft rejection following liver transplantation in man. Transplant. Proc. 24, 1462-1464. Urbanski, A., Schwarz, T., Neuner, P., Krutmann, J., Kirnbauer, R., Kock, A., and Luger, T. A. (1990).Ultraviolet light induces increased circulating interleiikin-6 in humans. J . Znoest. Dermatol. 94,808-811. Uyttenhove, C., Coulie, P. G., and Van Snick, J. (1988). T cell growth and differentiation induced by interleukin-HPl/IL-6, the murine hybridoma/plasmacytoma growth fwtor. J . E x p . Med. 167,1417-1427. Van Damme, J.,De Ley, M., Opdenakker, G., Billiau, A,, De Somer, P. (1985). Homogenous interferon-inducing 22K factor is related to endogenous pyrogen and interleiikin1. Nature (London)314,266-268. Van Damme, J., Cayphas, S., Opdenakker, G., Billiau, A., and Van Snick, J. (1987a). Interleukin-1 and poly(rI)poly(rC) induce production of a hybridonia growth factor by human fibroblasts. E u r . J . Zmmunol. 17, 1-7. Van Damme, J., Opdenakker, G., Simpson, R. J., Rubira, M. R., Cayphas, S., Vink, A., Billiau, A., and Van, Snick, J. (1987b). Identification of the human 26-kD protein, interferon p2, as a cell hybridoma/plasmacytoma growth factor induced by interleukin-1 and tumor necrosis factor. J . E x p . Med. 165,914-919. Van Damme, J., Schaafsma, M. R., Fibbe, W. E., Falkenburg, J. H., Opdenakker, G . , and Billiau, A. (1989). Simultaneous production of interleukin 6, interferon-beta and colony-stimulating activity by fibroblasts after viral and bacterial infection. Eur. J . Immunol. 19,163-168. Vandenabeele, P., and Fiers, W. (1991). Is amyloidogenesis during Alzheimer’s disease due to an IL-l-/IL-6- mediated ‘acute phase response’ in the brain? Zmmunol. Toduy 12,217-219. Van Meir, E., Sawamura, Y.,Diserens, A. C., Hamou, M. F., and d e Tribolet, N. (1990). Human glioblastoma cells release interleukin 6 in oioo and in uitro. Cancer Res. 50, 6683-6688. Van Oers, M. H. J., van der Heyden, A. A. P. A. M., and Aarden, L. A. (1988). Interleukin6 (IL-6) in serum and urine of renal transplant recipients. Clin. E x p . lrnmunol. 71, 314-319. Van Snick, J. (1990). Interleukin-6: An overview. Annu. Rev. Zmmunol. 8,253-278. Van Snick, J., Cayphas, S., Vink, A., Uyttenhove, C., Coulie, P. G., Rubira, M. R., and Simpson, R. J. (1986). Purification and NHz-terminal amino acid sequence ofa T-cell-
INTERLEUKIN-6 IN BIOLOGY AND MEDICINE
77
derived lymphokine with growth factor activity for B-cell hybridomas. Proc. Natl. Acad. Sci. U.S.A.83,9679-9683. Van Snick, J., Cayphas, S., Szikora, J.-P., Renauld, J.-C., Van Root, E., Boon, T., Simpson, R. J., Van Roost, E., Boon, T., and Simpson, R. J. (1988). cDNA cloning of murine interleukin-HP1: Homology with human interleukin-6. Eur.J. lmmunol. 18,193-198. Vercelli, D., Jabara, H. H., Arai, K., Yokota, T., and Geha, R. S. (1989). Endogenous interleukin-6 plays an obligatory role in interleukin-4 dependent human IgE synthesis. Eur. J . lmmunol. 19, 1419-1424. Vink, A., Coulie, P. G., Warnier, G., Renauld, J.-C., Stevens, M., Donckers, P., and Van Snick, J. (1990). Mouse plasmacytoma growth in oioo: Enhancement by interleukin-6 and inhibition by antibodies directed against IL-6 or its receptor. ./. Exp. Med. 172, 997-1000. Vogel, J.,Hinrichs, S. H., Reynolds, R. K., Luciw, P. A., and Jay, G. (1988).The HIV tat gene induces dermal lesions resembling Kaposi's sarcoma in transgenic mice. Nature (London)335,606-611. Vreden, S. G., van den Broek, M. F., Oettinger, M. C., Verhave, J. P., Meuwissen, J. H., and Sauerwein, R. W. (1992). Cytokines inhibit the development of liver schizonts of the malaria parasite Plasmodium Oerghei in oioo. Eur. J . lmmunol. 22,2271-2275. Waage, A., Brandtzaeg, P., Halstensen, A., Kiernlf, P., and Espevik, T. (1989). The complex pattern of cytokines in serum from patients with meningococcal septic shock. Association between interleukin 6, interleukin 1 and fatal outcome../. Exp. Med. 169, 333-338. Walther, Z., May, L. T., and Sehgal, P. B. (1988). Transcriptional regulation of the interferon-P2/B cell differentiation factor BSF-2/hepatocyte-stimulating factor gene in human fibroblasts by other cytokines. J . lmmunol. 140,974-977. Walz, G . , Zanker, B., Melton, L. B., Suthanthiran, M., and Strom, T. B. (1990).Possible association of the immunosuppressive and B cell lymphoma-promoting properties of cyclosporine. Transplantation 49, 191-194. Warrington, R. J., and Rutherford, W. J . (1990). Normal mitogen-induced suppression of the interleukin-6 response and its deficiency in sytemic lupus erytheniatosus. ./. Clin. lmmunol. 10, 52-60. Watanabe, K., Tanaka, Y., Morimoto, I., Yahata, K., Zeki, K., Fujihira, T., Yamashita, U., and Eto, S. (1990). Interleukin-4 as a potent inhibitor of bone resorption. Biocliem. Biophys. Res. Commun. 172, 1035-1041. Watson, J. M., Sensintaffar,J. L., Berek, J. S., and Martinez-Maza, 0. (1990).Constitutive prodnction of interleukin 6 by ovarian cancer cell lines and by primary ovarian tumor cultures. Cancer Res. 50,6959-6965. Weissenbach, J.. Chernajovsky, Y., Zeevi, M., Shulman, L., Sorecq, H., Nir, U., Wallach, D., Perricaude, M., Tiollais, P., and Revel, M. (1980). Two interferon mRNA in human fibroblasts: I n oitro translation and Escherichia coli cloning studies. Proc. Nutl. Acud. Sci. U.S.A.77,7152-7156. Westenberger, U., Thanner, S., Ruf, H. H., Gersonde, K., Sutter, G., and Trentz, 0. (1990). Formation of free radicals and nitric oxide derivative of hemoglobin in rats during shock syndrome. Free Radicals Res. Comttiun. 11, 167-178. Williams, R. L., Hilton, D. J., Pease, S., Willson, T. A., Stewart, C. L., Gearing, D. P., Wagner, E. F., Metcalf, D., Nicola, N. A., andcough, N. M. (1988). Myeloidleukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nuture (London)336,684-687. Yamamori, T., Fukada, K., Aebersold, R., Korsching, S., Fann, M.-J., and Patterson, P. H. (1989). The cholinergic neuronal differentiation factor from heart cells is identical to leukemia inhibitory factor. Science 246, 1412-1416.
78
SHIZUO AKlRA ET AL
Yamasaki, K., Taga, T., Hirata, Y., Yawata, H., Kawanishi, Y., Seed, B., Taniguchi, T., Hirano, T., and Kishimoto, T. (1988). Cloning and expression of the human interleukin-6 (BSF-2IIFNP2)receptor. Science 241,825-828. Yasukawa, K., Hirano, T., Watanabe, Y., Muratani, K., Matsuda, T., and Kishimoto, T. (1987). Structure and expression of human B cell stimulatory factor-2(BSF-2/IL-6) gene. EMBO J. 6,2939-2945. Yawata, H., Yasukawa, K., Natsuka, S., Murakami, M., Yamasaki, K., Hibi, M., Taga, T., and Kishimoto, T. (1993). Structure-function analysis of human IL-6 receptor: dissociation of amino acid residues required for IL-6-binding and for IL-6 signal transduction through gp130. EMBO]., in press. Yee, C., Biondi, A., Wang, X. H., Iscove, N. N., d e Sousa, J., Aarden, L. A., Wong, G. G., Clark, S. C., Messner, H. A,, and Minden, M. D. (1989). A possible autocrine role for interleukin-6 in two lymphoma cell lines. Blood 74,798-804. Yin, T., Miyazawa, K., and Yang, Y. C. (1992).Characterization of interleukin-11receptor and protein tyrosine phosphorylation induced by interleukin-1 1 in mouse 3T3-Ll cells. J. Biol. Chem. 267,8347-8351. Yokoi, T., Miyawaki, T., Yachie, A., Kato, K., Kasahara, Y., and Taniguchi, N. (1990). Epstein-Barr virus-immortalized B cells produce IL-6 as an autocrine growth factor. Immunology 70,100-105. Yoshizaki, K., Nakagawa, T., Kaieda, T., Muraguchi, A., Yamamura, Y., and Kishimoto, T. (1982). Induction of proliferation and Ig production in human B leukemic cells by anti-immunoglobulins and T cell factors. J . Zmmunol. 128, 1296-1301. Yoshizaki, K., Matsuda, T.,Nishimoto, N., Kuritani, T., Taeho, L., Aozasa, K., Nakahata, T., Kawai, H., Togo, H., Komori, T., Kishimoto, S., Hirano, T., and Kishimoto, T. (1989). Pathological significance of interleukin 6(IL-6/BSF-2)in Castleman’s disease. Blood 74,1360-1367. Yoshizaki. K., Nishimoto, N., Matsumoto, K., Tagoh, H., Taga, T., Deguchi, Y.,Kuritani, T., Hirano, T., Hashimoto, K., Okada, N., and Kishimoto, T. (1990). Interleukin-6 and expression of its receptor on epidermal keratinocytes. Cytokines 2,381-387. Zarling, J. M., Shoyab, M., Marquardt, H., Hanson, M. B., Lioubin, M. N., and Todaro, G. J. (1986). Oncostatin M:A growth regulator produced by differentiated histiocytic lymphoma cells. Proc. Nutl. Acud. Sci. U.S.A.83,9739-9743. Zhang, Y., Lin, J. X., and Vilcek, J. (1988). Synthesis of interleukin 6 (interferoi~p2lBcell stimulatory factor 2) in human fibroblasts is triggered by an increase in intracellular cyclic AMP. J . Biol. Chem. 263,6177-6182. Zhang, X. G., Klein, B., and Bataille, R. (1989). Interleukin-6 is a potent myeloma-cell growth factor in patients with aggressive multiple myeloma. Blood 74, 11-13. Zhang, Y., Lin, J. X., and Vilcek, J. (1990). Interleukin-6 induction by tumor necrosis actor and interleukin-1 in human fibroblasts involves activation of a nuclear factor binding to a KB-like sequence. Mol. Cell. Biol. 10,3818-3823. Zilberstein, A., Ruggieri, R., Kom, J. H., and Revel, M. (1986). Structure and expression of cDNA and genes for human interferon-p2, a distinct species inducible by growthstimulatory cytokines. EMBO J. 5,2529-2537. This article was accepted for publication on 10 February 1993.
ADVANCES I N IMMUNOLOGY, VOL. 54
Interleukin-9 I.-C RENAULD, F. HOUSSIAU, 1. LOUAHED, A. VINK, J. VAN SNICK, AND C. UYTTENHOVE Ludwig Institute for Cancer Research, Brussels Branch, ond Experiment01Medicine Unit, Catholic Uniwrsify of louvain, B- 1200 Brussels, Belgium
1. Introduction
It has long been known that activated helper T cells contribute to immune responses by producing various soluble factors that regulate a panel of effector cells. Progress in molecular biology has led to the characterization of an ever-growing number of these proteins. Pleiotropy and redundancy have turned out to be the hallmarks of the cytokine network: a single factor usually has various activities on unrelated targets, and unrelated factors frequently display the similar activity on a single target cell. Thus, as were most cytokines, interleukin-9 (IL-9) was functionally identified in various systems using distinct approaches. Originally, IL-9 was described as a murine T cell growth factor produced b y activated T cells and characterized by a narrow specificity for certain helper T clones (1).Stable T cell lines could be derived in the absence of feeder cells and antigen, provided supernatants from activated helper T cells were added to the cultures. The growth factor present in such supernatants was further purified and designated P40, and its corresponding cDNA was cloned (2). Independently, Hultner and collaborators reported that a factor produced by activated splenocytes was able to enhance the proliferation of mast cell lines induced by 1L-3 or IL-4 (3,4). This activity, designated mast cell growth-enhancing activity, (MEA), was also found in the supernatant from a murine Mls-reactive Th cell line derived by Schniitt and collaborators, who had observed that these cells produced a T cell growth factor TCGF-III(5). The molecular cloning of a murine P40 cDNA and the availability of recombinant protein led to the demonstration that the same factor, namely P40/IL-9, was responsible for all these biological activities (6). In the human, IL-9 was initially identified and cloned by Yang and colleagues as a mitogenic factor for a human megakaryoblastic leukemia (7), whereas the same human cDNA was isolated by crosshybridization with the mouse probe (8). More recently, IL-9 targets were found to encompass erythroid progenitors (9), human T cells (lo), B cells (1l), fetal thymocytes (12), and thymic lymphomas (13). 79 Copyright 0 1993 hy Academic Press, Inc. All rights of reproduction in any form reserved.
80
J.-C. RENAULD ET AL.
Although the biology of IL-9 might not be completely unraveled, we can already envision its physiological significance. In this review, we summarize the information available on the biology of IL-9 and its receptor, with particular emphasis on its activity on T cells and mast cells. II. Characterization and Cloning of Mouse and Human Interleukin-9
The first description of IL-9 was based on the observation that supernatants from activated helper T cells are capable of supporting the long-term growth of some mouse T-cell clones in the absence of antigen and antigen presenting cells. By the use of such supernatants, stable cell lines were derived and used to follow the purification of the factor responsible for this activity. The purified protein, originally designated P40 on the basis of its apparent size in gel filtration, was characterized by an elevated PI (- 10)and a high level of glycosylation (1).Partial amino acid sequences obtained after cyanogen bromide treatment allowed the cloning of a full-length cDNA encoding the murine IL-9 protein (2). In parallel with the cDNA cloning, a complete sequencing of the purified protein has been achieved (14), confirming the deduced amino acid sequence of the mature protein. The IL-9 protein sequence consists of 144 residues with a typical signal peptide of 18 amino acids. The presence of four potential N-linked glycosylation sites could explain the discrepancy observed between the predicted relative molecular mass (14,150) and the relative molecular mass measured for native IL-9. The sequence is also characterized by the presence of 10cysteines in the mature protein and a strong predominance of cationic residues, which explains the elevated PI (-10). The human homolog of mouse IL-9 was cloned independently by expression cloning of a factor stimulating the growth of a human megakaryoblastic leukemia (7) and by cross-hybridization with the mouse gene (8).Like the mouse protein, human IL-9 consists of 144 amino acids, with four potential N-linked glycosylation sites. The general structure of the two molecules is well conserved, with 10 cysteines perfectly matched and overall homologies of 69% at the nucleotide level and 55% at the protein level (Fig. 1).Cross-reactivity between species is, however, limited as the murine protein is biologically active on human cells whereas human IL-9 is not active on murine cells. The mouse IL-9 gene has been localized on chromosome 13(15) and the human IL-9 gene was mapped on chromosome 5, in the 5q31-q35 region (16). Interestingly, this region which has been shown to be
81
INTERLEUKIN-9
-10 1 10 20 30 h MLLAMVLTSALLLCSVAGQGCPTLAGILDINFLINKMQEDPASKCHCS II I I II II II I I I I I1 II I I * * I m MLVTYILASVLLFSSVLGQRCSTTWGIRDTNYLIENLKDDPPSKCSCS 40
50
60
70
h A&V&VSCLCLGIPSDNCTRPCFSERLSQMTmMQTRYPLIFSRVKKSV IIII*I-l I II. I I I I n I I Ill I m GbJV?JSCLCLSVPTDDCTTPCYREGLLQLTEQKSRLLPVFHRVKRIV 80
90
100
110
120
h EVLKNNKCPYFSCEQPCWTAGNALTFLKSLLEIFQKEKRGKI
I I I I I * I 11.1 I*III Ill I IIIIII Ill I m EVLKNNCPSFSCEKPCNJgaGNTLSFLKSLLGTFQKTEMQRQKSRP
FIG. 1. Alignment of murine and human interleukin-9 protein sequences. Vertical lines indicate the conserved residues, black dots show the 10 cysteines of the mature polypeptide, and underlining indicates the potential N-linked glycosylation sites. Amino acid numbering is based on the N terminus of the mature protein.
deleted in a series of hematological disorders, also contains various growth factor and growth factor receptor genes such as IL-3, IL-4, IL-5, colony-stimulating factor (CSF)-1, and CSF-1 receptor (CSF-1R). Radiation hybrid mapping analysis has recently located the IL-9 gene between the IL-3 and the early growth response (EGR)-1 genes (17). The human and murine IL-9 genes share a similar genomic organization, with 5 exons and 4 introns stretching over about 4 kb. The 5 exons are identical in size for both species and show homology levels of 56, 67,64,73, and 74% respectively. In contrast, although the size of the 4 introns is only slightly different between the human and mouse genes, no significant sequence homology was found in the introns; however, 3' and 5' untranslated regions show a high level of identity supporting a possible involvement of these sequences in the transcriptional or post-transcriptional regulation of IL-9 expression (18). A classical TATA box was identified in both genes 22 to 24 nucleotides upstream of the transcription start mapped by S1 nuclease protection. Potential recognition sites for several tissue plasminogen activator (TPA)-inducible transcription factors such as AP-1 and AP-2 were identified in the 5' flanking region of the genes, providing a structural basis for the induction of IL-9 expression by phorbol esters. A consensus sequence for interferon regulatory factor (1RF)-1was also identified in both promoters but its physiological relevance remains elusive (18).Recently, Kelleher and co-workers identified other consensus sequences in the 5' untranslated region of the human gene (SP1, NF-KB,octamer, AP-3, AP-5, glucocorticoid responsive element,
82
J.-C. RENAULD ET AL.
cAMP response element, etc.) and suggested that the NF-KBsite and cAMP response element could be involved in the constitutive expression of IL-9 by human T lymphotropic virus type 1 (HTLV-1)transformed T cells (19). Functional analysis using deletion or substitution mutants of the IL-9 promotor should clarify the regulatory mechanisms underlying IL-9 expression. 111. The Interleukin-9 Receptor
The existence of specific high-affinity receptors for IL-9 was demonstrated on a variety of mouse hematopoietic cells including T cells, mast cells, and macrophages. Scatchard analysis performed on a T cell clone revealed a single class of binding sites with a KD of 100 pM. Crosslinking studies suggested that IL-9 binds to a 64-kDa glycoprotein, the molecular mass of which is reduced to 54 kDa on treatment with N-glycosidase F (20). A murine IL-9 receptor cDNA has been identified by expression cloning in COS cells (21). The deduced protein contains 468 amino acids and two hydrophobic domains spanning residues 15-37 and 271-291, probably corresponding to the signal peptide and transmembrane domain, respectively. The extracellular domain, composed of 233 amino acids, contains two potential N-linked glycosylation sites and 6 cysteines. The position of these cysteines, as well as the presence of a WSEWS motif located 26 residues upstream from the transmembrane region, indicates that the IL-9 receptor is a member of the hematopoietin receptor superfamily (22). The human IL-9 receptor cDNA, isolated by cross-hybridization with the mouse probe, encodes a 522-amino-acid protein with a 53% homology to the mouse IL-9 receptor. The extracellular region is particularly conserved with 67% identity, whereas the cytoplasmic domain is significantly larger in the human receptor (231 versus 177 residues). The demonstration that this cDNA encoded a functional IL-9 receptor was provided by the observation that its expression in a mouse T cell clone conferred responsiveness to human IL-9 (21). As already observed for many members of the hematopoietic receptor superfamily, IL-9 receptor messenger RNAs have been identified that lack the sequences encoding the transmembrane and cytoplasmic domains, as a result of alternative splicing. This phenomenon, as well as the use of alternative polyadenylation signals, is responsible for the multiple bands observed in Northern blot. For instance, in the human, an intriguing heterogeneity of the IL-9R messages is generated by alternative splicing of an intron located 28 nucleotides downstream the
INTERLEUKIN-9
83
putative initiation condon. In most clones identified in different cell types, this intron was only partially spliced by using alternative 5' and/or 3' splicing sites. Little is known on the mechanisms of signal transduction through the IL-9 receptor. IL-9 stimulation of human megakaryoblastic leukemia MOTE induced at least four unidentified tyrosine phosphorylated bands in Western blot analysis. In contrast, IL-9 neither induced nor enhanced the phosphorylation of the serine-threonine kinases Raf-1 and MAP, an event frequently observed on stimulation with cytokines (23).The cytoplasmic domain of the IL-9 receptor does not contain any motif characteristic of a potential tyrosine kinase or serine-threonine kinase activity; however, a large percentage of serine and proline residues were found in this region, as already observed for most cytokine receptors. Moreover, an intriguing stretch of nine successive seriiies was observed in the human receptor. Finally, some sequence homology with other receptors was noted proximally to the transmembrane domain, in particular a Pro-X-Pro sequence preceded by a cluster of hydrophobic residues that partially fits a recently described consensus sequence shared by many cytokine receptors (IL-4R7IL-'IR, IL-3R, EPOR, ILSRP, G-CSFR) (24). Downstream from this motif, a striking homology was observed with the /3 chain of the IL-2 receptor and with the erythropoietin receptor. As a result, for the first 33 amino acids of the cytoplasmic domain, 40% identity was noted between the human IL-9R and the IL2RP and 27% between IL-9R and the erythropoietin receptor. Interestingly, these two receptors interact with the cytokines that have been shown to synergyze with IL-9 for the proliferation of fetal thymocytes and erythroid progenitors, respectively. A schematic representation of the human and mouse IL-9 receptors is shown in Fig. 2. IV. Interleukin-9 in the Hematopoietic System
The first evidence of involvement of IL-9 in the hematopoietic system was provided by the identification and cloning of the human protein as a growth factor for the megakaryoblastic leukemia line Mo7E (7,25). This finding prompted several groups to evaluate the activity of IL-Y on normal hematopoietic precursors. Although IL-9 did not seem to be active on megakaryoblastic precursors, it was found to support the clonogenic maturation of erythroid progenitors in the presence of erythropoietin (9). This activity was not blocked by a neutralizing antiserum against IL-3 or GM-CSF and was observed with highly purified progenitors after sorting for CD34+ cells and T
84
J.-C. RENAULD ET AL.
Extracellular
Cytoplasmic domain
0 mIL9R
1
\ a
homology with ILZRB
, Serine-rich region
hIL9R
FIG.2. Schematic representation of murine and human interleukin-9 receptors.
cell depletion (26,27). By contrast, granulocyte or macrophage colony formation (CFU-GM, CFU-G, or CFU-M) was not influenced by IL-9. Taken together, these data indicate that, within the adult hematopoietic system, IL-9 is a rather specific regulator of erythropoiesis. A similar erythroid burst-promoting activity has been described in the mouse but, surprisingly, seems to be dependent on the presence of T cells (28). . Experiments comparing the effects of IL-9 on fetal and adult progenitors have shown that IL-9 is more effective on fetal cells, suggesting that it supports the maturation of a primitive subset of fetal BFU-E. Moreover, addition of IL-9 to cultures of fetal progenitors induced maturation of CFU-Mix and CFU-GM, activities not detected in adult cultures (29).The observation that murine day 15 fetal thymocytes (12), but not adult thymocytes (30),responded synergistically to IL-9 and IL-2 lends further credence to the hypothesis that the spectrum of activity of IL-9 is larger on fetal progenitors.
INTERLEUKIN-9
85
V. Interfeukin-9 and Human T Cells
Studies with human T cells have clarified two aspects of the biology
of IL-9, namely, the mechanisms underlying its expression in T cells
and its stimulating activity on peripheral T cells.
A. EXPRESSION OF INTERLEUKIN-9 BY HUMAN T CELLS We initially found that T cell mitogens such as phytohemagglutinin (PHA) and anti-CD3 monoclonal antibody (mAb) induced substantial IL-9 expression in freshly isolated peripheral blood mononuclear cells (PBMCs), which was further enhanced by addition of phorbol myristate acetate (PMA). Sorting experiments confirmed that IL-9 was preferentially produced by T cell-enriched lymphocyte populations and, more specifically, by CD4' T cells (18).The delayed IL-9 expression in PBMCs, with a peak at 28 hours, together with the inhibitory effect of cycloheximide, suggested the involvement of secondary signals in this process. Further experiments identified IL-2 as a required mediator for IL-9 induction in T cells because (1) anti-IL-2R mAb blocked IL-9 expression after stimulation with PMA and anti-CD3 mAb, and (2) IL-2 was the only cytokine that synergized with PMA for the induction of IL-9 (31). The central role played by IL-2 in IL-9 expression in freshly isolated human T cells was recently confirmed in the mouse, inasmuch as IL-2 was capable of inducing IL-9 in fresh murine splenocytes costimulated with PMA (P. Monteyne, personal communication). It should, however, be stressed that the regulation of the IL-9 gene in fresh T cells might be different from that in T cell lines, as suggested recently by the observations that IL-1, not IL-2, serves as a secondary signal for IL-9 expression in murine T cells lines (32)and HTLV-l-transformedT cells produce IL-9 constitutively (7,19).
B. T CELLGROWTH-STIMULATING ACTIVITYOF HUMAN INTERLEUKIN-9 Data obtained with human T cells have shed new light on the T cell growth factor activity of IL-9 (10).We observed that all human T cell lines derived by weekly stimulation of T cells with PHA, IL-2, and irradiated allogeneic PBMCs feeders expressed a strong IL-9R message and proliferated in response to IL-9, after only a few weekly passages (Table I). By contrast, unlike other T cell growth factors such as IL-2, IL-4, and IL-7, IL-9 did not induce the proliferation of freshly isolated T cells, neither alone nor in synergy with various cytokines or T cell costimuli. Interestingly, significant proliferations were induced
86
J.-C. RENAULD ET AL.
TABLE I RESFQNSEOF HUMAN T CELLLINES TO INTERLEUKIN-W Thymidine incorporation (cpm) in the presence of Line
No Factor
IL-9
IL-2
HC 15 HC 17 HC 18 HC 22
1350 1710
4,760 11,350 26,510 40,030
197,400 163,500 92,660 208,400
900
4540
a T cell lines derived from PBMCs by weekly stimulation with PHA (1/1OOO),IL-2 (20 U/ml), and irradiated allogeneic PBMCs (l@/ml) as feeders were seeded in microtiter plates (10s cells/well) in the presence or in the absence of IL-9 (35 Ulml) or 11-2 (50 U/ml). Thymidine incorporation was measured in triplicate cultures (SEM < 10%)on day 2.
by IL-9 when PBMCs were activated for only 10 days with PHA, IL-2, and irradiated allogeneic feeder cells, thereby indicating that responses to IL-9 might require previous activation. The importance of a state of activation was further reinforced by our observations that responses to IL-9, unlike those elicited by IL-2, varied depending on the time after restimulation of the cultures: responses were optimal when the cells had reached a fully blastic stage and decreased when the cells were tested later in the culture, namely, when they had undergone a size reduction (Fig. 3). These observations suggest that the response to IL-9 varies according to the stage of cell activation. These results, obtained with human T cells, differ in several ways from those initially reported in the murine model (1,33). Murine T cells respond to IL-9 only after months of culturing. Moreover, when murine T cells become responsive to IL-9, factor-dependent lines are readily derived that grow independently of their antigen and feeder cells. These discrepancies could be related to the distinct experimental systems used and, in particular, to the differences in the stimulation procedures. Human IL-9-responsive T cells are strongly activated with PHA, IL-2, and feeder cells on a weekly basis, whereas a much less potent, antigen-specific stimulation is used in the murine model. This difference could explain why freshly raised murine T cell clones do not respond to IL-9. On the other hand, IL-9 responsiveness of murine
INTERLEUKIN-9
87
h
0
J
days after restimulation
FIG.3. Response of human T cell lines to interleukin-9 varies according to the stage of activation. A human IL-9-responsive T cell line was seeded in microtiter plates (lo5cells/well) on days 4, 7, 10, and 14 after restimulation. Proliferation in response to IL-9 30 Ulml (left) or to IL-2 50 U/ml (right) was measured in triplicate cultures after 2 days. Shaded areas indicate background proliferation; hatched areas indicate specific cytokine-induced proliferation.
cells is acquired at a later stage when T cells have undergone a transformation process, as described in more detail later. The fact that no permanent IL-9-dependent human T cell lines or clones could be derived so far suggests that such an in uitro transformation process does not take place in the human model. The activity of IL-9 on murine T cell clones is apparently restricted to the helper T phenotype. To address this issue in the human, we derived human T cell clones either from established PHA-stimulated T cell lines or by direct cloning from freshly isolated purified T cells. We found that these clones proliferated in response to IL-9, irrespectively of their CD4 or CD8 phenotype. We also observed that tumorspecific cytolytic T cell clones expressed IL-9R messages (Fig. 4) and responded to IL-9 by proliferation or increased survival. Taken together, these experiments indicate that the T cell-stimulating activity
88
J.-C. RENAULD
ET AL.
FIG.4. Northern blot analysis of interleukin-9 receptor expression in human CTL clones. Poly(A)+RNA was extracted by cesium chloride purification and oligo(dT)beads from CRL 1572 teratocarcinoma cells (lane I), Mo7E cells (lane 2), and human antimelanoma CTL clones 159/3 and 159/5 (lanes 3 and 4). Poly(A)+ RNA was electrophoresed in an agarose gel and hybridized with an IL-9R or a p-actin probe.
of IL-9 is not restricted to a given phenotype, but rather depends on a particular stage of activation. VI. Interleukin-9 and Mouse T Cellr: A Role in Tumorigenetis?
The response of murine helper T clones to IL-9 is gradually acquired by long-term in uitro culture and mimicks, to some extent, tumoral transformation. Before describing this multistep process, it must be remembered how these clones are derived and cultured. Helper T clones are derived from lymph nodes ofmice immunized with proteins and maintained in uitro by repeated stimulation with antigen and syngeneic antigen presenting cells. After stimulation, two distinct phases can be observed based on T cell morphological changes. During the first days after antigenic stimulation, T cells become blastic, proliferate intensively, and secrete large amounts of lymphokines. This blastic phase is followed by a rest period during which the cells undergo a considerable size reduction and stay quiescent for 1 or 2 more weeks without any reduction in numbers. At the end of such a culture, cells are diluted and restimulated with antigen and feeder
89
INTERLEUKIN-9
cells. Most importantly, a rest period is required before restimulation; otherwise no stable T cell clones can be maintained in culture. Responses to cytokines are evaluated in the absence of antigen and feeder cells at the end of the rest period. Using this experimental system, we observed that responses to IL-9 varied not only between independent clones but also for a single clone depending on the total number of in uitro stimulations. When early T cell clones were exposed to IL-9, no response was detected at all, although significant proliferations were induced by IL-2. In a second stage, helper T clones responded to IL-9 by blast formation and IL-6 secretion; no significant proliferation was observed but cell survival was greatly enhanced (Fig. 5). After some more weeks in culture, T cells started proliferating in response to IL-9 but only in the presence of another factor, a strong synergy being observed with IL-4 and IL-3. In the final stage, IL-9 alone induced significant proliferation. The responsiveness to IL-9 at these different stages is illustrated in Fig. 6. When the cells become fully responsive to IL-9, as assessed in a microtiter proliferation assay, major changes appear in the correspond-
100
z
80
M
." .-
40
v1
20
0
0
20
10
Days
FIG.5. Survival of helper T clones in the presence of interleukin-9. Fourteen days after antigenic stimulation, cells of the TUC2.15 helper T clone were seeded in microtiter plates (5 x 104/well)without antigen and presenting cells. Culture medium was supplemented or not with recombinant murine IL-9 (500 U/ml). Cell survival was measured by trypan blue dye exclusion.
90
J.-C. RENAULD ET AL. 50
Stage 1
25
0
50
Stage 2
25
0 100
Stage 3
I
!!
v 3
Y
50
0
IL9
IL3
ILQ+lL3
FIG.6. Response of helper T clones to interleukin-9 at different stages of transformation. At the end of an antigenic stimulation, TUC2.15 helper T cells were seeded (5 x 104/well) in microtiter plates. Culture medium was supplemented with IL-3 (3 ng/ml) and/or IL-9 (500 U/ml). Thymidine incorporation was measured in triplicate cultures on day 3.
ing cultures. Cells do not go through a rest period anymore and remain blastic. Moreover, they lose their capacity to produce cytokines on lectin stimulation. Finally, they die during the second week of the culture period. At this stage, these cultures can be rescued by addition of IL-9, and permanent IL-9-dependent cell lines can be derived. An increase in cell size is noted, as are an accelerated growth rate and the progressive disappearance of all T cell markers. Thus, in a few weeks time, Thy.1, CD4, CD3, and T cell receptor (TCR) expression is totally lost, TCR rearrangements at the DNA level being the only T cell marker left. Most of our IL-9-dependent lines also lose the (Y chain of the IL-2 receptor as well as the capacity to proliferate in the presence of IL-2. In contrast, they remain IL-4 responsive, as all IL-9-dependent lines, derived from THI or TH2 clones, can be grown in IL-4, with one exception, the ST2.K9 line, which is responsive to IL-2 and IL-9 but
INTERLEUKIN-9
91
not to IL-4 (33).Leukemia inhibiting factor (LIF/HILDA)and insulin have also been shown to support the long-term proliferation of some of these IL-9-dependent lines but the significance of this observation remains elusive (34; F. Lehmann et al., unpublished data). The progressive deregulation undergone by T cells during acquisition of IL-9 responsiveness is reminiscent of the multistep process leading to cell transformation (35,36). T o examine whether such a tumorigenic transformation occurred in our cell lines, w e transfected an IL-9-dependent T cell line with IL-9 cDNA. The transfectant cells secreted IL-9 and grew autonomously in uitro. After injection of these cells into syngeneic C57BL/6 mice, a very high tumor incidence was observed. Animals died in 3 to 4 months as a result of widespread lymphomas (37).This finding suggests that dysregulated production of IL-9 by T cells could be part of T cell transformation processes. T o test this hypothesis, we generated transgenic mice expressing high levels of IL-9 constitutively (Renauld et al., manuscript in preparation). Although no major morphological changes were noted in the lymphoid organs of most mice, the animals showed a high incidence of thymic lymphomas. About 5% of the mice spontaneously developed lymphomas and all transgenic mice developed such tumors after injection of doses of a mutagen (N-methyl nitrosourea) that was totally innocuous in control mice. These tumors involved primarily the thymus and invaded other lymphoid organs such as the spleen and the lymph nodes. Most tumors had rearranged the TCRP locus and expressed both CD4 and CD8 antigens, whereas CD3 expression varied between tumors. Noteworthy, the instability of the transgene in one of the five independent founders led to the development of chimeric mice, in which some ofthe cells had lost the transgene. This phenomenon allowed the occurrence in these mice of tumors that had lost the transgene. One of these tumors initially failed to grow in vivo after transplantation into syngeneic normal mice; however, injection of these cells into transgenic mice rapidly resulted in tumor formation, thereby demonstrating that the influence of IL-9 was not restricted to the initial steps of oncogenesis. Injection of large numbers of cells into normal mice also resulted, in some experiments, in tumor formation after a prolonged period. The observation that the latter tumors are subsequently able to grow indistinctly in normal or transgenic mice suggested that they had undergone additional genetic alterations leading to IL-9-independent growth. Incidentally, this last step in T cell transformation seems to occur spontaneously in uitro as all IL-9dependent cell lines generate autonomous cells at frequencies ranging from <1/107 to l/105. Injected in uiuo, such autonomous cells induce
92
J.-C. RENAULD ET AL.
tumors histologically comparable to those obtained with IL-9transfected cells. The mechanisms underlying this process are presently unclear. The hypothesis that autonomous cells produce their own IL-9 is, however, unlikely as conditioned medium failed to sustain the growth of IL-9-dependent lines and as no IL-9 message was detected by polymerase chain reaction (PCR) analysis. The occurrence of thymic lymphomas in IL-9-transgenic mice prompted us to investigate whether IL-9 played a similar role in other models of thymic lymphomas. In this respect, we have found that IL-9 significantly stimulates the in vitro proliferation of primary lymphomas induced either by chemical mutagenesis in DBA/2 mice or b y X-ray radiation in B6 mice (13).Moreover, these studies have demonstrated a strong synergy between IL-9 and IL-2 for several lymphomas (Fig. 7), a synergy reminiscent of the activity of IL-9 on murine fetal thymocytes (12). A role for IL-9 in human lymphomas was suggested by the observation that lymph nodes from patients with Hodgkin’s and large cell
FIG.7. Interleukin-9 and interleukin-2 synergistically stimulate the proliferation of thymic lymphomas. Fresh tumor cells isolated from the thymic lymphomas N M l l and NM 19 (See Ref. 13) were seeded in microtiter wells (5 x lo4cellslwell). Proliferation in response to IL-2 (100 U/ml) and/or IL-9 (500 U/ml) was measured in triplicate cultures after 2 days.
INTERLEUKIN-9
93
anaplastic lymphomas constitutively produce IL-9 as demonstrated by Northern blot and in situ hybridization analyses (38).Constitutive IL-9 expression was also detected in HTLV-l-transformed T cells (19) and in Hodgkin’s cell lines (38,39). The recent demonstration of an autocrine loop for the in vitro growth of one of these Hodgkin’s cell lines (39) suggests a potential involvement of IL-9 in this disease. VII. Interleukin-9 and Mast Cells
hliurine mast cell lines are readily derived by cultivating bone marrow progenitors with lectin-activated T cell supernatants. Such cell lines, which are phenotypically and functionally related to mucosal mast cells, can be maintained in uitro in a state of factor-dependent growth for several weeks (40). Originally, IL-3 was identified as the growth factor required by these cells (41), but subsequently it appeared that IL-4 provided a comitogenic signal (42,43). Finally, a third factor, present in spleen cell conditioned medium was characterized on the basis of its synergistic effect with IL-3 for the proliferation of permanent bone marrow-derived mast cell lines (BMMCs) such as L138.8A (3).This factor, provisionally designated mast cell growthenhancing activity (MEA), was identified as IL-9 as (1)recombinant IL-9 displayed a similar mast cell growth factor activity and (2) the activity of semipurified MEA was inhibited by an anti-IL-9 rabbit antiserum (6). The proliferative activity of IL-9, alone or in synergy with IL-3, was confirmed on other permanent mast cell lines such as MC-6, H7, and MC-9 (28; unpublished data). As observed with certain TH clones, the response of BMMCs to IL-9 varies according to the time in culture. When primary BMMCs were derived from hematopoietic progenitors. IL-9 alone was not sufficient to sustain mast cell growth, but synergistically enhanced the proliferation induced by IL-3 or the combination of IL-3 and IL-4 (Fig. 8). Moreover, IL-9 significantly increased the survival of these primary BMMCs in the absence of other factors (6; unpublished data). When stable mast cell lines were obtained, IL-9 alone became capable of inducing proliferation, without the need for additional factors. Finally, autonomous cells, which induced tumors in syngeneic animals, could be derived from factor-dependent cell lines (44). In addition to its proliferative activity, IL-9 also induces functional changes such as IL-6 secretion by mast cell lines (6, 44). Moreover, investigations in our laboratory of the activity of IL-9 at the molecular level indicate that IL-9 is a potent regulator of the expression of protease genes like those belonging to the granzyme family. We also
94
J.-C. RENAULD ET A L
FIG.8. Activity of IL-9 on the proliferation ofprimary bone marrow-derived mast cell cultures. BALB/C bone marrow cells (10' cells) were cultured for 6 weeks with or without IL-3 (1 nglml), IL-4 (50 U/ml), and IL-9 (500 U/mI). Cultures were divided every week.
noted that expression of the alpha chain of the high-affinity IgE receptor is upregulated by IL-9, indicating that this factor could be a key mediator of mast cell responses (Louahed et al., manuscript in preparation). Preliminary data indicating that IL-9 synergizes with IL-4 for IgE production by murine and human B cells further reinforce this view (11;B. Dugas, personal communication). In uiuo, mastocytosis occurs in mice infected with various helminthic worms. After infection with Trichinella spiralis, IL-9 production is induced in mesenteric lymph nodes, as is production of other cytokines such as IL-3, IL-4, and IL-5 (45). Moreover, resistance to Trichuris muris was found to correlate with the production of IL-5 and IL-9 by mesenteric lymph node cells (46). The kinetics of cytokine production induced by Heligmosomoides pol ygyrus or Nippostrongylus brasiliensis, known to induce a TH2 response (47), was extensively studied by Gause and collaborators. Within 3 to 6 hours of oral inoculation of H . polygyrus larvae, IL-9 mRNA levels increased considerably in Peyer's patches and remained elevated several days. Similarily, when N . brasiliensis larvae were inoculated subcutaneously, IL-9
INTERLEUKIN-9
95
mRNA induction was found in the lungs by 24 hours and in mesenteric lymph nodes after 4 days. As the larvae reached the lungs within 1day and the gut after 3 to 4 days, IL-9 expression appears to be an early event in the immune response and actually precedes the production of other cytokines such as IL-4. Interestingly, pretreatment of mice with anti-CD4 and anti-CD8 monoclonal antibodies failed to suppress this early IL-9 production, suggesting that T cells would not be the exclusive source of this factor (F. Finkelman, personal communication). The role of IL-3 and IL-4 in Nippostrongylus-induced mastocytosis was demonstrated by the observation that either anti-IL-3 or anti-IL-4 antibodies partially (40-50%) inhibit mucosal mast cell hyperplasia; however, when both anti-IL-3 and anti-IL-4 antibodies were combined, no more than 85% suppression was observed, suggesting the involvement of a third factor (48). The observation that IL-9 was induced in these models, together with its activity on mast cells in uitro, raised the possibility that the molecule contributed to the mastocytosis observed in these animals. Although an anti-IL-9 antiserum by itself did not influence mast cell hyperplasia, combination of this antibody with anti-IL-3 and anti-IL-4 antibodies increased reproducibly the suppression of the mast cell response from 85 to 95%. Moreover, when mice received suboptimal doses of anti-IL-3 plus anti-IL-4 antibodies, suppression of mastocytosis increased from 60 to 94% by injection of an anti-IL-9 antiserum (F. Finkelman, personal communication). Taken together, these results suggest that IL-9, though not absolutely required for the development of mucosal mast cell hyperplasia induced by worm infections, amplifies this response, particularly under conditions in which IL-3 and IL-4 are limiting.
VIII. Conclusions
An increasing number of observations indicate that IL-9 is a multifunctional cytokine with diverse effects on a variety ofcell types. Since its discovery as a growth factor for helper T cell clones, IL-9 has turned out to be involved in other important biological systems, such as erythropoiesis and mast cell activation. In addition to their physiological significance, these properties raise obvious therapeutic possibilities. On the one hand, the ability of IL-9 to stimulate erythroid lineage recovery requires further investigation. On the other hand, the possibility that IL-9 antagonists could be used to downregulate excessive mast cell activation is not too farfetched. Finally, should the involvement of IL-9 be confirmed in hematological malignancies, IL-9
96
J.-C. RENAULD ET AL.
blockade could provide us with a complementary approach to the management of these malignancies.
REFERENCES 1 . Uyttenhove, C., Simpson, R. J., andVan Snick, J. (1988).Proc. Natl. Acad. Sci. U.S.A. 85,6934-6938. 2. Van Snick, J., Goethals, A,, Renauld, J-C., Van Roost, E., Uyttenhove, C., Rubira, M. R., Moritz, R. L., and Simpson, R. J. (1989).]. E x p . Med. 169,363-368. 3. Hultner, L., Moeller, J., Schmitt, E., Jager, G., Reisbach, G., Ring, I., and Dormer, P. (1989).J Immunol. 142,3440-3446. 4. Moeller, J., Hultner, L., Schmitt, E., and Dormer, P. (1989).]. Immunol. 142,3447345 1. 5. Moeller, J., Hultner, L., Schmitt, E., Breuer, M., and Dormer, P. (1990).J . lmmunol. 144,4231-4234. 6. Hultner, L., Druez, C., Moeller, J., Uyttenhove, C., Schmitt, E., Rude, E., Dormer, P., and Van Snick, J. (1990).Eur.]. Immunol. 20,1413-1416. 7 . Yang, Y., Ricciardia, S., Ciarletta, A., Calvetti, J., Kelleher, K., and Clark, S. C. (1989). Blood 74,1880-1884. 8. Renauld, J-C., Goethals, A., Houssiau, F., Van Roost, E., and Van Snick, J. (1990) Cytokine 2,9-12. 9. Donahue, R. E., Yang, Y. C., and Clark, S. C. (1990). Blood 75,2271-2275. 10. Houssiau, F., Renauld, J. C., Stevens, M., Lehmann, F., Coulie, P. G., and Van Snick, J. (1993).]. Immunol., in press. 1 1 . Petitfrkre, C., Dugas, B., Braquet, P., and Mencia-Huerta, J. M., (1993).Immunology, in press. 12. Suda, T., Murray, R., Fischer, M., Tokota, T., and Zlotnik, a. (1990).J . Immunol. 144, 1783- 1787. 13. Vink, A., Renauld, J-C., Warnier, G., and Van Snick, J. (1993).Eur.]. Immunol., in press. 14. Simpson, R. J., Moritz, R. L., Gorman, J. J., and Van Snick, J. (1989).Eur. J . Biochem. 183,715-722. 15. Mock, B. A., Krall, M., Kozak, C. A., Nesbitt, M. N., McBride, 0.W., Renauld, J-C., and Van Snick, J. (1990). lmmunogenetics 31,265-270. 16. Modi, W. S., Pollock, D. D., Mock, B. A., Banner, C., Renauld, J-C., and Van Snick, J. (1991).Cytogenet. Cell. Genet. 57,114-116. 17. Warrington, J., Bailey, S., Armstrong, E., Aprelikova, O., Alitalo, K., Dolganov, G., Wilcox, A., Sikela, J., Wolfe, S., Lovett, M., and Vasmuth, J. (1992). Genomics 13, 803-808. 18. Renauld, J-C., Goethals, A., Houssiau, F., Merz, H., Van Roost, E., and Van Snick, J. (1990).]. Immunol. 144,4235-4241. 19. Kelleher, K., Bean, K., Clark, S. C., Leung, W. Y., Yang-Feng, T. L., Chen, J. W., Lin, P. F., Luo, W., and Yang, Y. C. (1991). Blood 77, 1436-1441. 20. Druez, C., Coulie, P., Uyttenhove, C., and Van Snick, J. (1990).J . Immunol. 145, 2494-2499. 21. Renauld, J-C., Druez, C., Kermouni, A., Houssiau, F., Uyttenhove, C., Van Roost, E., and Van Snick, J. (1992). Proc. Natl. Acad. Sci. U.S.A.89,5690-5694. 22. Idzerda, R. J., March, C. J., Mosley, B., Lyman, S. D., Vanden Bos, T., Gimpel, S. D., Din, W. S., Grabstein, K. H., Widmar, M.B., Park, L., Cosman, D., and Beckmann, M. P. (1990).]. E x p . Med. 171,861-873.
INTERLEUKIN-9
97
23. Miydzdwd, K., Hendrie, P., Kim, Y-J., Mantel, C., Yang, Y-C., Se Kwon, B., and Broxmeyer, H. (1992). Blood 80, 1685-1692. 24. Murakami, M., Narazaki, M., Hibi, M., Yawata, H., Yasukawa, K., Hamaguchi, M., Taga, T., and Kishimoto, T. (1991). Proc. Natl. Acad. Sci. U.S.A.88, 11349-11353. 25. Hendrie, P., Miyazawa, K., Yang, Y-C., Langefeld, C., and Broxmeyer, H. (1991). E x p . Hemutol. 19, 1031-1037. 26. Lu, L., Leemhuis, T., Srour, E., and Yang, Y-C. (1992). E x p . Hematol. 20,418-424. 27. Birner, A., Hultner, L., Mergenthaler, H-G., Van Snick, J., and Dormer, P. (1992). E x p . Hemutol. 20,541-545. 28. Williams, D. E., Morrissey, P. J., Mochizuki, D. Y., de Vries, P., Anderson, D., Cosman, D., Boswell, H. S., Cooper, S., Grabstein, K. H., and Broxmeyer, H. E. (1990). Blood 76,906-911. 29. Holbrook, S . T., Ohls, R. K., Schribler, K. R., Yang, Y. C., and Christensen, R. D. (1991). Blood 77,2129-2134. 30. Suda, T., Murray, R., Guidos, C., and Zlotnik, A. (1990).J . Zmmunol. 144,3039-3045. 31. Houssiau, F., Renauld, J-C., Fibbe, W., and Van Snick, J. (1992).J . Zmmunol. 148, 3147-3 151. 32. Schmitt, E., Beuscher, HU., Huels, C., Monteyne, P., van Brandwijk, R., Van Snick, J., and Rude, E. (1991).J . Zmmunol. 147,3848-3854. 33. Schmitt, E., van Brandwijk, R., Van Snick, J., Siebold, B., and Rude, E. (1989).Eur.J. Zmmunol. 19,2167-2170. 34. Van Damme, J., Uyttenhove, C., Houssiau, F., Put, W., Proost, P., and Van Snick, J. (1992). Eur. J . Zmmunol. 22,2801-2808. 35. Ball, P., Conroy, M., Heusser, C., Davis, J., andconscience, J-F. (1983).Differentiation 24,74-78. 36. Nowell, P. (1976). Science 194,23-28. 37. Uyttenhove, C., Druez, C., Renauld, J-C., Herin, M., Noel, H., and Van Snick, J. (1991).J . E x p . Med. 173,519-522. 38. Merz, H., Houssiau, F., Orscheschek, K., Renauld, J-C., Fliedner, A., Herin, M., Noel, H., Kadin, M., Mueller-Hermelink, K. H., Van Snick, J., and Feller A. C. (1991).Blood 78,1311-1317. 39. Gruss, H. J., Brach, M. A., Drexler, H. G., Bross, K. J., and Herrman, F. (1992). Cancer Res. 52,1026-1031. 40. Jarrett, E., and Haig, D. M. (1984). lmmunol. Today5, 115-119. 41. Ihle, J., Keller, J., Oroszlan, S., Henderson, L., Copeland, T., Fitch, F., Prystowsky, M., Goldwasser, E., Schrader, J., Palaszynski, E., Dy, M., and Lebel, B. (1983). J. Zminunol. 131,282-287. 42. Mossman, T., Bond, M., Coffman, R., Ohara, J., and Paul, W. (1986).Proc. Natl. Acad. Sci. U.S.A. 83,5654-5658. 43. Schmitt, E., Fassbender, B., Beyreuther, K., Spaeth, E., Schwarzkopf, R., and Rude, E. (1987). lmmunobiology 174,406-419. 44. Hultner, L., and Moeller, J. (1990). E x p . Hematol. 18,873-877. 45. Grencis, R. K., Hultner, L., and Else, K. J. (1991). Immunology 74,329-332. 46. Else, K., Hultner, L., and Grencis, R. (1992). Zmmunology 75,232-237. 47. Urban, J. F., and Madden, K. B., Svetic, A., Cheever, A., Trotta, P. P., Cause, W. C., Katona, I. M., and Finkelman, F. D. (1992). Immunol. Reo. 127,205-220. 48. Madden, K., Urban, K., Ziltener, H., Schrader, J., Finkelman, F., and Katona, I. (1991).J . Zmmunol. 147,1387-1391.
This article was accepted for publication on 10 February 1993.
ADVANCES IN IMMUNOLOGY, VOL. 54
Superantigens and Their Potential Role in Human Disease BRIAN 1. KOTZlN*,S,Il, DONALD Y. M. LEUNG*,§,JOHN KAPPLERt,S,II, AND PHILIPPA MARRACKf,S,# Deporiments of Pediatrics and Medicine, and the j Howard Hughes Mdical Insthte ot Denwr, the Notionol Jewish Center for Immunology and Respimtory Medicine, Denwr, Colorado 802w Deporiments of $ Modicine, Pediatrics, 11 Microbiology and Immunology, and # Biochemiw, Biophysics, and Genetics, Uniwrsify of Cdomdo Hwlth Sciences Center, Denwr, Colotudo a262
1. Introduction
The term superantigen was introduced to describe a group of microbial antigens that differ in several respects from conventional protein or peptide antigens (White et al., 1989). Most importantly, the recognition of superantigens by T cell receptors (TCR) appears to depend almost entirely on the variable domain of the TCR /3 chain (Vp) with little regard for the other diversity components (i.e., Dp, Jp,Va, Ja).A schematic representation of this fundamental difference in recognition is shown in Fig. 1.Because the relative number of Vp genes is limited, a given superantigen is capable of interacting with a large fraction (-5-30%) of the T cell repertoire, whereas the responding frequency to a conventional antigen is usually much less than 1 in 1000. Like peptide antigens, superantigens are presented by class I1 major histocompatibility complex (MHC) molecules, but they do not engage the MHC peptide-binding groove (Fig. 1). Instead, the intact (unprocessed) superantigen seems to interact with conserved amino acid residues that are on the outside of the peptide-binding cleft. Polymorphic differences in MHC that affect peptide binding do not usually affect superantigen binding or presentation to TCR Vp, and recognition of superantigens is not normally MHC restricted. Two general classes of superantigens with markedly different structures have been studied in detail. The first type was noticed nearly 20 years ago when Festenstein observed marked T cell responses in primary mixed lymphocyte reactions with cells from certain MHCidentical strains (Festenstein, 1973). The stimulating antigens were called minor lymphocyte-stimulating (Mls) antigens to differentiate them from MHC antigens. Since this pioneering work, immunologists have realized that mice contain many genes that have the properties of Mls. As described below, these superantigens are encoded by endogenous retroviral genes. The second class of superantigens is represented by a growing list of bacterial, mycoplasmal, and viral proteins, 99 Copyright 0 1993 hy Academic Press, Inc. All rights ofreproduction in any form reserved.
100
BRIAN L. KOTZIN ET AL.
FIG.1. Left:Diagram ofthe interactions between a conventional (peptide) antigen, T cell receptor (TCR), and major histocompatibility complex (MHC) molecule. Conventional peptide antigens lie within a groove formed by the a and p chains of the class I1 MHC molecule. T cell recognition of the peptide-MHC complex usually involves multiple elements of both the TCR a and p chains. The numbers shown correspond to the TCR Va and Vfl complementarity-determining regions (CDRs) or hypervariable regions (HVRs) involved in the peptide-MHC recognition. CDR3 is formed by residues within the junctional regions (Vp-Dp-j/3 and Va-Ja), and is frequently critically important in peptide-MHC recognition. Right: Diagram of the interactions between superantigen (SAG), TCR, and MHC class I1 molecules. Superantigens do not bind within the peptide-binding groove of the MHC molecule and interact primarily with the Vp region of the T cell receptor. Evidence indicates that the superantigen primarily interacts with the fourth hypervariable region (HVR4) of the TCR /3 chain.
capable of producing a variety of pathological effects after infection. Proteins of this group include the prototypic staphylococcal enterotoxins and streptococcal exotoxins. In a short period of time, the study of superantigens has provided new insight into many basic biological questions. Furthermore, evidence has indicated a potential role for superantigens in the pathogenesis of several human diseases. This article examines the rapidly expanding literature on this important group of viral and bacterial molecules. II. Endogenous Murine Superantigens: Murine Mammary Tumor Virus
A. IDENTIFICATION OF Mls ANTIGENS AS PRODUCTS OF MURINE TUMOR VIRUS MAMMARY The murine Mls antigens were originally discovered in the early 1970s because of their enormous T cell stimulatory capacity, unexplained by MHC differences, in primary mixed lymphocyte cultures (Festenstein, 1973).The superantigen nature of these antigens and the in vivo consequences of expression were not, however, appreciated
SUPERANTIGENS
101
until recently (reviewed in Acha-Orbea and Palmer, 1991; Herman et al., 1991a; Janeway, 1991; Huber, 1992; Marrack et al., 1993). This breakthrough in understanding was dependent on the development of monoclonal antibodies that are specific for T cells expressing a particular VP domain. With such antibodies, several groups noted the absence of VP-bearing subsets in murine strains expressing Mls antigens. For example, in mice that carry the Mlsa (Mls-1")gene, initial experiments noted that T cells bearing VP8. 1and VP6 were absent from the peripheral lymphoid tissues (Kappler et al., 1988; MacDonald et al., 1988b). Related to the recognition of the Mls-1" antigen as a self-antigen, these T cell subsets were shown to be deleted during thymic maturation as part of the self-tolerance process. Thus, Vp8.l' T cells were present among immature (CD4+, CD8') thymocytes but were absent from mature (CD4+, CD8- or CDS', CD4-) thymocyte populations and from mature peripheral T cells (Kappler et al., 1988). The in uiuo deletion of T cells caused by Mls-la expression is now known to encompass T cells expressing VP6, 7,8.1, and 9 (Table I) (Kappler et al., 1988; MacDonald et al., 1988b; Okada and Weissman, 1989; Happ et al., 1989; Okada et al., 1990; Waanders and MacDonald, 1992). Other Mls antigens were shown to be associated with deletion of different VP-expressing T cell subsets (Table I). For a time, it seemed as if the murine class I1 MHC molecule, I-E, was another superantigen involved in the deletion of T cells expressing VP5, 11, 12, 16, or 17a (Kappler et al., 1987a,b; Woodland et al., 1990; Bill et al., 1989; Vacchi0 and Hodes, 1989; Singer et al., 1990; Vacchio et al., 1990). However, studies showed that I-E was necessary but not sufficient for the deletion of these VP subsets (Bill et al., 1989; Woodland et al., 1990). An analysis of different inbred and recombinant inbred strains revealed that another endogenous antigen (termed "cotolerogen" ) was also necessary. When presented by an I-E molecule, these cotolerogens thus functioned in a manner similar to the Mls antigens; that is, both mediated deletion ofT cells and, in both cases, recognition depended mostly on the VP part of the TCR. The retroviral origin of the endogenous superantigens was first suggested by genetic studies attempting to map the positions of the M l s and Mls-like genes. In one set of studies, Woodland et al. (1990,1991a) showed that the Etc-l cotolerogen (mediating deletion of VP5+ and VP11+T cells) was genetically linked to the position of an integrated mouse mammary tumor provirus (MMTV) on chromosome 12 (Mtu-9; Table I). Subsequent experiments were unable to segregate genetically the presence of Mtu-9 and the phenotype of VP-specific T cell deletion. At about the same time, several other groups of investigators
102
BRIAN L. KOTZIN ET AL.
TABLE I MAMMARY TUMOR VIRUSSUPERANTICENS IN MICE" Superantigen
Vp specificity
Mammary tumor virus
6,7,8.1,9 3, 17 3,5, 17 3 3 5.1.5.2, 11, 12, 17 11, 12, 17 11,12 16 14,15 14 6, 7, 8.1,9
7 1 6 13 3 9 8 11 ? Exogenous MMTV Exogenous MMTV Exogenous MMTV
Chromosome
~
Mls-la Mls-2" Mls-3" Ml~-2~-like vSAG-3 Etc-1, DVp11-2 DVpll-1 DVpl1-3 Unnamed MMTV (C3H) MMTV (GR) MMTV (SW)
1 7 16 4 11 12 6 14 ?
None None None
a Data for this table were obtained from the following:Kappler et al. (1987a,b,1988);MacDonald et al. (1988);Okada and Weissman (1989); Happ et al. (1989);Okada et al. (1989);Pullen et al. (1988, 1989,1991,1992);Bill etal. (1989);Woodlandetal. (l990,1991a,b);Vacchioefal. (1990);Singeretal. (1990);Fairchild et 01. (1991); Frankel et al. (1991);Dyson ef al. (1991);R.Abe ef al. (1991);Marrack et al. (1991);Acha-Orbea et al. (1991, 1992); Held et al. (1992);Gollob and Palmer (1992);JouvinMarche et 01. (1992,1993).Only a partial list of MMTVs is provided. Several MMTV proviruses with similar Vp specificities similar to those listed or with undefined specificities are not included in this table. Examples include Mfo-43 (Rudy et al., 1992) from strain MAlMyJ (VSS, 7, 8.1, and 9) and Mto-Mai from strain MA1 (VP3, 17). Some additional Vfl specificities may also be described for the MMTVs listed. For example, considering similarities in vSAG carboxy-terminalsequence (see Fig. 3). it seems likely that vSAC 13 and vSAC 3 are also recognized by V/317' T cells.
noted that separate Mls antigens also mapped to different MMTV loci dispersed throughout the murine genome (Table I) (Dyson et al., 1991; Frankel et d.,1991). All described Mls loci are now known to be genetically inseparable from endogenous MMTV loci. Marrack et al. (1991) provided additional evidence for the role of MMTV-encoded products in VP-specific T cell deletions. It was noted that Vp14+ T cells were deleted in C3H/HeJ mice, but only in F1 offspring if the mother was of C3HIHeJ origin. This led to the discovery that T cell deletion was dependent on the transfer of a milk-borne infectious mammary tumor virus [infectious MMTV (C3H)l during nursing. Specific deletion of different VP subsets has been observed with infectious MMTV transmitted in other strains (Table I) (Acha-Orbea et al., 1991; Held et aZ., 1992).
103
SUPERANTIGENS
B. STRUCTURE-FUNCTION STUDIES OF MURINE MAMMARY TUMOR VIRUSSUPERANTICENS The MMTV superantigens have been shown to be encoded within the 3’ long terminal repeat (LTR)sequences of the virus (Fig. 2). Thus, transfection of the DNA within this region conferred VP-specific stimulatory capabilities to recipient B cells (Choi et al., 1991, 1992; Woodland et ul., 1991b; Beutner et al., 1992; Pullen et al., 1992). Furthermore, expression of this MMTV genetic region as a transgene resulted in VP-specific T cell deletion in vivo (Acha-Orbea et al., 1991). Originally referred to as the ORF protein for the open reading frame found within the MMTV 3’ LTR, it has been proposed that this protein be renamed vSAG (viral superantigen) for its functional properties (Choi et al., 1991; Coffin, 1992; Marrack et al., 1993). The MMTV open reading frame that encodes vSAG activity produces a protein of about 320 amino acids (Fig. 2), and such a product can be produced in cell-free systems (Choi et al., 1992; Korman et al., 1992). These studies indicate that the vSAGs are type I1 membranebound proteins. Most membrane proteins are type I, that is, their
MMTV
I
LTR
I
gag
IP‘O
I
LTR
I
I
/ORFI
4
0
OR1:
0
0
0
0
0
J
0
0
100
“.I
r rr
0
0
0
0
0
0
0
0
0
0
0
0
0
1 1 1 1 1 1 1 1 1 1
I
VSXG FIG.2. Diagram of the open reading frame (ORF) or vSAG protein encoded in the 3’ long terminal report (LTR) of MMTV.Amino acids are numbered. The hatched area denotes the transmembrane region and the shaded COOH-terminal area of the vSAG represents a highly variable region that is predominantly involved in V/3 specificity (see Fig. 3). The arrow denotes a highly basic proteolytic cleavage site.
104
BRIAN L. KOTZIN ET AL.
NH2-terminal amino acids are found outside and their COOH-terminal amino acids lie inside the cell. Type I1 membrane proteins, on the other hand, have a single internal hydrophobic region, which acts both as a signal peptide and as a transmembrane region, and the COOH terminus is found extracellularly, whereas the NH2 terminus is found in the cytoplasm. Inspection of the amino acid sequences of different vSAGs reveals considerable homology except for the 20-30 COOHterminal residues (Fig. 3) (Fasel et aZ., 1982; Donehower et aZ., 1983; Crouse and Pauley, 1989; King et al., 1990; Acha-Orbea and Palmer, 1991; Pullen et al., 1992; Rudy et aZ., 1992; Brandt-Carlson et al., 1993).In fact, these superantigens can be assigned to one of more than five different families based on their COOH-terminal sequence, which also appears to be important in determining Vp specificity (Fig. 3).For example, members of the group of MMTV superantigens that react with T cells bearing Vp5 and/or Vpll are closely related throughout
MMTV
8
9 11
13
---------I H---V-YN SR-E-KRH I I - - - K - L P L A F ---------I H---V-YN S R - E - K R H I I - - - K - L P L A F --------- I H---V-YN S R - E - K R H I I - - - K - L P L A F --------- I H---V -Y N R-E-KRH I 1 ---K- L P L A F
7
---------
1 6
3
43
sw
S
NF-----D--E---I
-K I -Y NMKY T H G - R V G F D P F
--------- NF-----D--E---V-KI-YNMKY
---------NF-----
D--E---I
THN-RI C F D P F
-K I I Y N - K Y T H C - R I C F D P F
C3H GR
FIG.3. Amino acid sequences of the variable carboxy termini of different MMTV vSAGs. The MMTV are grouped by V/3 specificity. A remarkable correlation with carboxy-terminal sequence is shown. [Adapted from Choi et al. (1991);Acha-Orbea et al. (1991); Pullen et al. (1992); Acha-Orbea and Palmer (1991); Held et al. (1992); Huber (1992); Rudy et (11. (1992); Jouvin-Marche et al. 1992, 1993).]
SUPERANTICENS
105
their sequences, including the COOH-terminal end. Members of the group interacting with Vp3 are almost identical to each other and differ from the Vp5/11 group most markedly in the last 20-30 residues (Pullen et al., 1992). Mutational analyses also confirm a direct role for this COOH-terminal region in Vp specificity and superantigen function (Y. Choi, P. Marrack, and J. Kappler, unpublished observations, 1992). The exposed and variable nature of these COOH-terminal residues and correlation with VP specificity strongly support the hypothesis that the MMTV vSAG directly interacts with the TCR Vp, rather than indirectly inducing superantigen activity. Several groups of investigators have made monoclonal antibodies against the vSAG that can block superantigen function (Acha-Orbea et al., 1992; Winslow et al., 1992; Mohan et al., 1993). These studies clearly demonstrate that the vSAG protein directly interacts with the Vp chain of the TCR. One set of antibodies generated to a C-terminal fragment of vSAG-7 [the vSAG from Mtu-7 (Mls-la)] was shown to block anti-Mls T cell responses in uitro and to rescue V/36+T cells from clonal deletion in uiuo in strains carrying Mtv-7 (Acha-Orbea et al., 1992). These antibodies could also inhibit a VP6-specific proliferative response in uiuo induced by infectious MMTV (SW). Interestingly, cell surface staining experiments could easily demonstrate vSAG expression on the surface of cells transfected with a baculovirus carrying the gene, but expression could not be demonstrated on activated B cells that were stimulatory in culture. With a different antibody directed to the same C-terminal region of vSAG-7, Winslow et al. (1992) were able to detect cell surface staining on B cell blasts derived from an Mtv-7-positive strain, but not on resting B cells, resting T cells, or activated T cells from the same strain. The expression level on activated B cells was, however, low, suggesting that only a small number of molecules is expressed on the cell surface. Immunoblotting studies of total cellular proteins from vSAG7-expressing cells revealed two bands of 18.5and 45 kDa (Winslow et al., 1992). The 45-kDa band was found in internal labeling experiments, but only the 18.5-kDa band could be found on the surfaces of the cells. The results suggested that the MMTV vSAG protein is synthesized from the full-length open reading frame, transported into the endoplasmic reticulum where it is glycosylated, and then clipped prior to expression on the cell surface. A highly basic cleavage site that explains the truncated cell surface form can be identified (Fig. 2). As discussed below, it is thought that both MMTV and bacterial superantigens must bind class I1 MHC proteins before they can be recognized by TCR Vps. Interestingly, immunoprecipitation of surface-labeled cells with the anti-vSAG7 an-
106
BRIAN L. KOTZIN ET AL
tibody did not reveal an expected protein fragment of the vSAG, nor was there any sign on such gels of coprecipitation of the COOHterminal half of vSAG7 with, for example, MHC class 11. C. CELLULAR EXPRESSION OF MURINEMAMMARY TUMOR VIRUSSUPERANTIGENS Murine mammary tumor viruses are type B retroviruses that can be transmitted exogenously as milk-borne infectious retroviral particles. Infection of mammary epithelial cells and gene expression are essential for continuation of the viral life cycle. In mammary tissue, transcription of MMTV genes is regulated through the hormone response element located in the 5' LTR (reviewed in Giinzburg and Salmons, 1992; Corley et al., 1992). Hormones such as progesterone and prolactin are believed to be particularly important in the transcriptional activation of MMTV during pregnancy and lactation. As discussed above, MMTV can also be inherited as an integrated provirus. Although it was once thought that most MMTV proviruses are transcriptionally silent, the deletion of V/3 subsets associated with distinct MMTV loci indicate that expression of at least the gene encoding the superantigen must be taking place. Within the hematopoietic lineage, B cells account for most of the MMTV expression and most B cell lines and normal B cells express MMTV transcripts (reviewed in Corley et al., 1992). In contrast to mammary tissue, expression of MMTV in B cells is hormone independent, and can be upregulated by B cell stimulation with lipopolysaccharide (LPS) or various cytokines such as interleukin 4 (IL-4) (Corley et al., 1992; Gollob and Palmer, 1991). Studies have demonstrated a markedly restricted distribution of MMTV superantigen expression, and B cells are clearly the best stimulators of Mls responses (Webb et al., 1989; Molina et al., 1989). Other class II-bearing cells such as macrophages or dendritic cells do not appear to express the MMTV vSAG on their surface in a functional manner. Transfection of T cell lines derived from Mtu-7 (Mls-la)positive strains with class I1 molecules also has not resulted in stimulatory capabilities characteristic of vSAG-7-expressing B cells. Stimulation of B cells with LPS and IL-4 enhances their ability to stimulate T cells via expressed superantigens (Gollob and Palmer, 1991).Although these studies indicate that immune responses to the superantigen are controlled by B cells, other cells may express MMTV vSAG proteins. For example, CD8+, but not CD4+, T cells derived from an Mtu-7 (Mls-1")-positive strain can be induced to express vSAG mRNA (Mohan et d.,1993).The injection of these CD8+ cells into Mtu-7-negative (MZs-lh) mice results in the induction of specific tolerance and to
SUPERANTIGENS
107
deletion of Mls-1"-responsive VPS' T cells (Webb and Sprent, 1990b). The mechanism by which Mls antigens are functionally presented by class 11-expressing cells in these animals remains unclear.
D. ROLEOF MURINEMAMMARY TUMOR VIRUSSUPERANTICEN IN LIFE CYCLE OF VIRUS AND HOST The life cycle of infectious MMTV involves transmission of the virus to babies through infected milk, uptake of the virus through the gut, a harboring of infectious virus during early life, and infection of mammary epithelial cells in the adult (reviewed in Acha-Orbea and Palmer, 1991). During pregnancy, the production of high titers of infectious virus allows the continued transmission of virus. Mammary tumors actually appear to be a nonessential part of the life cycle and are dependent on where the virus integrates into the murine genome. Evidence indicates that persistent infection in early life is critical for establishment of active infection, and is dependent on a functional immune system. For example, nude mice that lack most functional T cells are not susceptible to MMTV-induced tumor development. These nude mice, however, develop tumors when T cells from infected mice are injected (Tsubura et al, 1988). Experiments have further shown that MMTV superantigens are critically involved in this early life phase of the life cycle of the virus. I n one set of studies, BALB/c mice congenic for the Mtu-3 integrant were studied for their susceptibility to experimental MMTV infection (Hainaut et al., 1990). The presence of this provirus conferred resistance to exogenous MMTV infection, as shown by measuring viral antigens released in the milk of infected mice and by recording the incidence of early mammary tumors. In a separate experimental system, mice transgenic for the gene encoding the exogenous MMTV (C3H) superatigen were produced (Golovkina et al., 1992). Unlike their nontransgenic littermates, transgenic mice expressing high levels of this vSAG protein specifically deleted their Vp14+ T cells. Mice that had early and marked deletion were resistant to infection with the exogenous C3H MMTV when it was present in their mother's milk. Thus, in the life history of the virus, stimulation of target T cells appears to be necessary for infection. In the suckling recipient of infected milk, a pool of activated lymphocytes appears to be created in which the virus may survive. Because presentation of the vSAG probably involves B cells, both activated B and T cells may result, and both subsets may harbor infectious virus in this early period (Held et al., 1993).Later in life, the developed mammary glands can be infected and the virus can again be transmitted to progeny via infected milk.
108
BRIAN
L. KOTZIN ET AL.
Endogenous MMTV proviruses are present in the germ line of all inbred mice, and most strains contain approximately three to eight proviruses at various genomic locations (Kozak et al., 1987; Abe et aZ., 1991).Over 40 different endogenous MMTV loci have been described, and most have been associated with VP-specific T cell deletions. Most of the proviruses are defective and do not produce infectious virus but have retained a functional superantigen gene and express a vSAG protein that results in T cell subset deletion. Experiments described immediately above suggest that the endogenous MMTV proviruses and especially their superantigen genes may be retained in the genome as an antiviral defense mechanism. Murine mammary tumor proviruses and their respective T cell deletions have also been observed in wild-type mice, and it has been difficult to find mice that do not have VP deletions (Pullen et al., 1990a; Jouvin-Marche et al., 1989, 1992, 1993). In some mice, nearly half of the potential T cell repertoire can be deleted; yet no deleterious effect has been apparent. To date, endogenous superantigens such as those encoded by MMTV proviruses have not been found in humans, and some investigators have presented evidence against their existence (Baccala et al., 1991). However, with the newer tools available to study the human T cell repertoire, studies have identified VP deletions that resemble those induced by MMTV (J. Donahue, P. Marrack, J. Kappler, and B. L. Kotzin, unpublished observations, 1993). Thus these deletions occur in both CD4+ and CD8+ populations, can be shown to segregate in only certain families, and cannot be ascribed to either MHC or TCR p-chain gene complexes. There also has been great interest in the possibility that human infectious retroviruses, especially those associated with immunodeficiency diseases such as AIDS, may encode superantigens that are involved in their pathogenicity. These data are reviewed below. 111. Exogenous Superantigens
A. STAPHYLOCOCCAL AND STREPTOCOCCAL TOXINS A number of different bacterial toxins appear to stimulate T cells through particular Vp chains in a manner similar to the MMTV superantigens (reviewed in Marrack and Kappler, 1990). Staphylococcus aut-eus, a common human pathogen, produces several enterotoxins, designated SEA (staphylococcal enterotoxin A) through SEE, which can be responsible for food poisoning and occasionally shock in humans. Some S. aureus isolates also produce toxic shock syndrome toxin 1(TSST-l), which has been implicated in the majority of cases of human toxic shock syndrome, as well as the exfoliative toxins (ExFT),
109
SUPERANTIGENS
which are associated with the scalded skin syndrome. Streptococcus pyogenes, or group A streptococcus, another human pathogen of skin and pharynx, also produces toxins with superantigenic properties. These have been designated streptococcal pyrogenic exotoxins A-C (SPE-A-C), also referred to as streptococcal erythrogenic toxins or scarlet fever toxins. These toxins (especially SPE-A) have been implicated in the induction of scarlet fever as well as a severe toxic shocklike illness associated with group A streptococcal infection. Table I1 summarizes many of the known S. aureus and Str. pyogenes superantigens and the Vfl regions with which they interact. Their potential role in disease is discussed in later sections. The structures of the staphylococcal and streptococcal enterotoxins relevant to superantigen function have been reviewed (Marrack and Kappler, 1990).Most are globular proteins of about 24 to 30 kDa. Not surprisingly, because many of them act enterically, they are relatively resistant to digestion. Some of the toxins are closely related; for exam-
TABLE I1 Vp SPECIFICITIES OF SOME EXOGENOUS BACTERIALSUPERANTICENS" V/i specificity Toxin
Mouse
Humanb
SEA SEB SECl SEC2 SEC3 SED SEE TSST-1 ExFT SPE-A SPE-B SPE-C MAM
1,3, 10,11, 17 7, 8.1-8.3 3, 8.2, 8.3, 11 3,8.2, 10, 17 7,8.2 3, 11, 17 11, 15, 17 15,16 10, 11, 15 ND" ND ND 5.1,6,8.1-8.3
1.1,5,6's, 7.3-7.4,g.l 3, 12, 14, 15, 17,20 3,6.4,6.9, 12, 15 12, 13.2, 14, 15, 17,20 3,5, 12, 13.2 5, 12 5.1, 6's, 8, 18 2 2 8, 12, 14, 15 2,8 1,2,5.1, 10 3, 17
Data for this table were taken from Marrack and Kappler (1990);Callahan et 01. (1990); Kappler et al. (1989a); Choi et al. (1989); Cole et al. (1990b);J. Abe et al. ( 1 9 9 1 ) ; Friedman et al. (1991a);Baccala et 01. (1992); Hudson et al. (1993);Tomai et al. (1992); Mollick et al. (1993). The complete responding human Vp repertoire has not been defined for many of the exogenous superantigens, and additional Vp populations are likely to be added to this list. ND, Not determined or undefined.
110
BRIAN L. KOTZIN ET AL.
ple, SEA and SEE are more than 90% alike in amino acid sequence, and SEB is closely related to the SEC toxins. In contrast, it is difficult to align sequences of TSST-1 and the exfoliative toxins with the others, and one would not have predicted that these proteins have similar immunological properties. As discussed below, all of these toxins bind to multiple MHC class I1 molecules and all are recognized predominantly by the Vp region of the T cell receptor. Overall, comparison of the sequences indicates that the staphylococcal toxins fall into at least three groups, one with SEA, SEE, and SED, another with SEB and the SEC toxins, and a third that includes TSST-1 and exfoliative toxins (Marrack and Kappler, 1990). Unlike members of the first two groups, TSST-1 does not have an internal disulfide loop. Streptococcal pyrogenic toxin A (SPE-A) has significant structural similarity to the SEB/ SEC group, and SPE-C appears to have some structural similarities with SPE-A (Bohach et al., 1989, 1990; Weeks and Ferretti, 1986; Bohach and Schlievert, 1987; Goshorn and Schlievert, 1988). Streptococcal pyrogenic toxin B, on the other hand, appears to be a variant of a streptococcal proteinase precursor and is unrelated to any of the other toxins (Hauser and Schlievert, 1990). Table I1 shows that the pattern of Vp specificity for the different S. aureus toxins corresponds loosely with their groupings by sequence similarity. However, there are clear exceptions. For example, although _ _ T cells bearing Vpl2, SECS is SEB and the SECs all stimulate human the only member of this group that stimulates T cells bearing Vp5.2, and SEC2 and SECS (but not SECl or SEB) stimulate T cells bearing Vp13.2. Furthermore, differences in murine and human Vp subsets responding to SEA vs SEE have allowed an analysis of critical enterotoxin residues involved in Vp binding (see below). Regions of SEB involved in its ability to function as a superantigen have been elucidated. In a mutational analysis, both the binding of toxin to MHC and the ability to interact with TCR Vp mapped to the NHz-terminal portion of the toxin (Kappler et aZ., 1992).Three regions were identified (Fig. 4). Mutations within region 2 (residues 41-53) were shown to be important for MHC class I1 binding, whereas mutations within the two-amino acid region 3 (residues 60 and 61) appeared to influence Vp binding. Other mutations such as in region 1(residues 9-23) appeared to influence both functions, although mutations at position 23 (region l b ) primarily affected Vp interactions. Interestingly, when the sequences ofthe s.aureus enterotoxins are aligned for maximal homology (Marrack and Kappler, 1990), this asparagine residue at position 23 is conserved among all of the enterotoxins and TSST-1 as well. Similarly, critical residues in the MHC-binding region
111
SUPERANTICENS
T
30 UI M a 0 p--LQ I SEB d s w D P r P d % u o t s s r P m ~ ~ ~ I ~ I O Q P L Y P ~ ~ ~ I I ( D T ~ " ~ ~ ~ ~ - ~ l l ~ ~
.... .... .... .... .... .... ....
.... .... .... .... .... .... ....
. ....
.... .... .... .... .... .... .... ....
. . . . . . .
. ..". . .... . .... . .... . .... . .... . ....
. .... - - I . .... .... . .... .... . .... .... . .... .... . .... ..-.
FIG.4. A mutational analysis showing regions of SEB involved in TCR and MHC binding. The partial amino acid sequence of SEB and mutants used in the study are shown. A dash indicates identity to native SEB. Region 1 corresponds to a region involved in both MHC and TCR binding. Region l a appears to be involved mostly in MHC binding whereas mutations in residue 23 (region lb) primarily affect Vp interactions. Mutations within region 2 primarily affect MHC binding, and residues in region 3 (residues 60 and 61) appear to be in the Vp recognition site. [Adapted from Kappler et al. (1992),J. Exp. Med. 175,390, by copyright permission of Rockefeller Univ. Press.]
also appear to be conserved among the different enterotoxins. The three-dimensional crystal structure of SEB has been determined (Swaminathan et al., 1992) and can be interpreted based on the above mutational analysis. Figure 5 shows areas of the two-domain structure that are involved in MHC binding. Important residues implicated in MHC binding are separated on the molecule, and do not coalesce into a single site although they appear to face in the same direction. The TCR-binding site encompasses a shallow cavity formed by both domains, and thus residues in both the NH2- and the COOH-terminal domains may be involved in the TCR interaction. Residues of SEA and SEE important in Vp-binding specificity have also been evaluated (Irwin et al., 1992; Hudson et al., 1993; Mollick et al., 1993). In these experiments, advantage was taken of the fact that although the amino acid sequences of SEA and SEE are more than 90% similar (82% identical), they activate murine and human T cells bearing different Vp elements (see Table 11).A series of hybrid molecules were prepared that exchanged residues between the two toxins, and these molecules were tested for their ability to stimulate different Vp
112
BRIAN L. KOTZIN ET AL.
FIG.5. A ribbon drawing showing the two-domain structure of SEB and residues determined to be important in the binding of SEB to MHC and TCR. Two regions facing in the same direction but separated on the molecule appear to be important in MHC binding. A shallow cavity in the molecular surface forms the TCR-binding site. After alignment with SEB residues, amino acids of SEA involved in Vp recognition can also be placed on a rim ofthis cavity. [Adapted from Kappler et al. (1992)and Swaminathan et al. (1992), with permission.]
subsets. Two amino acid residues (residues 206 and 207) near the carboxy terminus of the enterotoxins were found to be primarily responsible for the differences in Vp specificity. Residues in the NH2 terminus had a more minor, but significant, influence on Vp specificity (Mollick et al., 1993). After aligning residues in SEA and SEB, and based on the SEB crystal structure described above, the important SEA residues could be placed on the outer rim of the pocket binding to the TCR and in close proximity to SEB NHz-terminal region residues determined to be important in TCR binding (see Fig. 5). The binding of different enterotoxins to MHC class I1 appears to involve complex interactions that vary among the toxins (discussed below). As noted above, analysis of SEB residues involved in MHC binding disclosed two sites, primarily determined by NH2-terminus residues, that were geographically separated on the toxin molecule. Residues in the NHz-terminal third of the SEA molecule also appear to be responsible for the stronger binding of SEA to MHC class I1 compared with SEE (Pontzer et al., 1989, 1991; Mollick et al., 1993). However, unlike SEB, histidine residues in the carboxy-terminal re-
SUPERANTIGENS
113
gions of SEA and SEE also appear to be crucial for MHC binding, and binding also requires a zinc ion (Fraser et al., 1992). A histidine residue in the HLA-DR fl chain is also required for SEABEE binding, perhaps stabilizing the zinc binding and the toxin-MHC interaction (Herman et al., 1991b; Karp and Long, 1992). THATFUNCTION AS SUPERANTICENS B. OTHERBACTERIALPRODUCTS 1 . Mycoplasma arthritidis Mitogen Mycoplasmas are a diverse group of cell wall-free microorganisms and the smallest self-replicating prokaryotes. Although frequently part of the normal flora of humans and other mammals, many species cause a variety of pathological consequences, including experimental arthritis (Cole et al., 1985; Cole and Atkin, 1991).Mycoplasma arthritidis is known primarily for its ability to induce acute and chronic relapsing arthritis in rats and mice. This naturally occurring arthritogen of rodents can also cause conjunctivitis, uveitis, urethritis, and paralysis. It has been known for some time that a soluble factor produced by this organism can act as a potent T cell mitogen for both murine and human T cell(s) [termed M . arthritidis mitogen (MAM)] (Cole et al., 1981, 1982; Atkin et al., 1986).Although not fully defined, it is a highly basic protein of 15-30 kDa that is acid and heat labile. Partial NHz-terminal sequencing apparently has indicated that it is distinct from other bacterial proteins, including other bacterial toxins (Cole and Atkin, 1991). Over the last several years, it has become clear that MAM is a superantigen. Thus, stimulation of murine and human T cells is TCR Vfl specific (Cole et d.,1989, 1990b; Friedman et d.,1991a; Baccala et d., 1992) (see Table II), Furthermore, stimulation is dependent on antigen-presenting cells bearing class I1 molecules, primarily I-E in murine systems and HLA-DR in human systems, but is not MHC restricted and MAM does not require processing to be stimulatory (Cole et al., 1981,1986,1990a; Bekoff et al., 1987; Matthes et al., 1988; Baccala et al., 1992). In culture, MAM is also a potent stimulator of T cell-dependent polyclonal B cell activation and immunoglobulin production (Tumang et d.,1990).These investigators have suggested that this property distinguishes MAM from, for example, the staphylococcal enterotoxins. It is, however, not clear that MAM results in a qualitatively different form of stimulation to account for the marked B cell activation (Mourad et aZ., 1989). Although MAM appears to be involved in a toxic shock-like illness induced in animals after intravenous injection of viable organisms, there is evidence that MAM may not be involved in the induced arthritis (Cole and Atkin, 1991).
114
BRIAN L. KOTZIN ET AL.
2 . Clostridial and Pseudomonas Toxins A 34-kDa enterotoxin produced by Clostridium perfringens (CPET) has also been shown to stimulate human peripheral blood T cells in a manner that simulates other exogenous superantigens (Bowness et al., 1992).Thus, marked T cell stimulation was apparent with small quantities of toxin. Furthermore, human T cells bearing Vp6.9 and VpZ2 were preferentially stimulated, dependent on class 11-expressingcells in culture, and processing of CPET was not required to induce proliferation. Interestingly, like the S. aureus enterotoxins, CPET is a major cause of food poisoning. Pseudomonas exotoxin A (PE) is an ADP-ribosyl transferase with a molecular mass of 66 kDa (Misfeldt, 1990). Its sequence and crystal structure have been determined (Gray et al., 1984; Allured et al., 1986),and there appears to be no homology with the bacterial superantigens discussed above. Although stimulation has been determined to be Vp selective [VpS in mouse (Legaard et al., 1991); not determined in humans], other characteristics of stimulation, such as a requirement for processing (Legaard et al., 1991), are unusual when compared to the other known superantigens. As discussed below, Vp can be the prominent TCR element in the recognition of certain peptide antigens, and further investigation will be necessary to define the superantigenic properties of PE.
3. Streptococcal M Proteins Streptococcus pyogenes, in certain susceptible individuals, can trigger the onset of autoimmune diseases such as acute rheumatic fever, rheumatic heart disease, and acute glomerulonephritis (reviewed in Stollerman, 1990).The major virulence factor of these bacteria is the M protein molecule that is expressed on the bacterial cell surface. The family of M proteins includes structurally related proteins that are highly variable at the amino terminus. M proteins have been shown to elicit strong T cell stimulation, and a pepsin-purified 22-kDa fragment of M protein from type 5 organisms (pepM5) was shown to be a potent stimulant of human T cells in culture (Dale et al., 1981; Kotb et al., 1989). T cell stimulation was Vp specific, required class 11-expressing presenting cells and was not MHC restricted, suggesting the superantigen nature of the stimulating molecule (Tomai et al., 1990). Human T cells expressing VP2,4, and 8 were shown to be preferentially stimulated (Tomai et al., 1991). These investigators have proposed that the large number of T cells stimulated may be important in the role of M protein in the development of poststreptococcal diseases (Kotb, 1992). The streptococcal M proteins are structurally dissimilar from the
SUPERANTIGENS
115
other well-studied streptococcal and staphylococcal toxin superantigens. Although the purified M protein products used in the above studies were homogeneous and without contamination by gel analysis, concern has been expressed regarding the possibility that minute quantities of toxin contamination could be responsible for the superantigenic properties of the M proteins (Fleischer et al., 1992; Braun et al., 1993). This seems unlikely, however, based on several pieces of evidence, including differences in responding VP populations (Tomai et al., 1992), differences in responses among species (Dale et al., 1981; Robinson and Kehoe, 1992; Leonard et al., 1991), and the ability to block M protein responses without affecting toxin stimulation (Dale et al., 1981; Wang and Kotb, 1992). Furthermore, studies with recombinant molecules expressed in cell-free systems (and therefore free of contaminating streptococcal proteins) have apparently confirmed the VP-specific stimulatory capacity of M 5 (Robinson and Kehoe, 1992; M. Kotb, unpublished observations, 1993).
C. EXOGENOUS SUPERANTICENS PRODUCED BY INFECTIOUSVIRUSES Superantigen production has been implicated in several viral systems. Some viruses that are associated with human disease and that may involve a superantigen in pathogenicity (e.g. HIV) are discussed in more detail in Sections V,G and V,H. 1 . Mouse Acquired Zmmunodeficiency Disease Syndrome Studies have suggested that certain infectious retroviruses may encode or induce molecules with superantigenic properties. Infection of certain strains of mice with a replication-defective murine leukemia virus (MuLV) can lead to a severe immunodeficiency syndrome, and paradoxically, to marked proliferation of CD4+ T cells in a VP-specific manner, suggestive of an Mls-like superantigen. Initial studies suggested that the viral Gag protein was responsible for the VP5-selective T cell stimulation (Hugin et al., 1991).At this time, however, it remains unclear whether a molecule actually encoded by this virus is functioning as a superantigen. 2 . Herpesuirus Saimiri and Epstein-Barr Virus Herpesvirus saimiri is an infectious virus of monkeys. It establishes a persistent asymptomatic infection in its natural host (squirrel monkey, Saimiri sciureus), but infection of other New World primates can lead to rapidly progressive T cell lymphomas (Fleckenstein and Desrosiers, 1982). Studies have shown that an abundant transcript expressed by herpesvirus saimiri at “immediate-early” times of lytic infection potentially encodes a product with homology to the MMTV
116
BRIAN L. KOTZIN ET AL.
superantigens described above (Nicholas et al., 1990; Thomson and Nicholas, 1991). This molecule also has the structure of a type I1 transmembrane glycoprotein and has a small amino-terminal intracellular domain. The similarities between this herpesvirus saimiri protein and MMTV superantigens suggests that they may have similar functions, and experiments are in progress to determine whether this protein acts as a superantigen. Interestingly, herpesvirus saimiri and Epstein-Barr virus (EBV) encode closely related proteins and have a similar organization of coding sequences in their genomes (Gompels et al., 1988). Epstein-Barr virus is the cause of a number of benign and malignant lymphoproliferative syndromes, including infectious mononucleosis and progressive B cell lymphomas in immunosuppressed individuals (see below). Studies have shown that an EBV-transformed B cell line (Raji)expresses a superantigen that stimulates human Vp3+ cells (Donahue et al., 1993). Although other EBV-transformed cell lines did not stimulate Vp3+ T cells, it is possible that different EBV isolates encode superantigens with other Vp specificities, similar to MMTV.
3. Rabies Nucleocapsid The nucleocapsid of the rabies virus has been shown to be a potent activator of human peripheral blood lymphocytes in vitro (Herzog et al., 1992), and this activation has been shown to resemble that induced by a superantigen (Lafon et al., 1992). The nucleocapsid is composed of a negative-strand RNA coated with three internal proteins: N, NS, and L proteins. Stimulation of T cells in culture with nucleocapsid was shown to require class 11-bearing antigenpresenting cells. Direct binding to cells, dependent on MHC class I1 expression, also could be demonstrated. Fixation of the antigenpresenting cells did not prevent stimulation, indicating that processing was not necessary. Analysis of Vp expression in stimulated peripheral blood cells from either unprimed or primed donors revealed selective stimulation of Vp8' T cells, and depletion of Vp8' T cells from these cell populations prevented the T cell-proliferative response to nucleocapsid. The superantigen-like properties of the nucleocapsid were recapitulated by using purified or recombinant N protein. Interestingly, stimulation induced b y the intact rabies virus did not simulate that of the nucleocapsid or N protein. The above studies raise a number of provocative questions. For example, is there a role for the superantigen (and therefore activated T cells) in the infectivity of rabies virus or in the pathogenesis of central nervous system infection? The potential role of a nucleocapsid superan-
SUPERANTIGENS
117
tigen in vaccine-induced neutralizing antibodies also needs to be addressed. IV. Immune Response to Superantigens
The large number of T cells stimulated by various superantigens initially suggested that these molecules may function as T cell mitogens, that is, as nonspecific T cell stimulators similar in activity to plant lectins such as concanavalin A (ConA). However, there are at least two important differences between superantigens and T cell mitogens. The most important is that stimulation by a mitogen is not dependent on a particular variable part of the T cell receptor, whereas superantigens stimulate T cells that bear only certain VP regions. In addition, despite the need for antigen-presenting cells, there is usually no requirement for presentation by MHC class I1 molecules in mitogen-induced stimulation. Similar to conventional protein or peptide antigens, the T cell response to superantigens almost always requires both (1)presentation of MHC molecules on antigen-presenting cells, and (2) recognition by variable elements of the T cell receptor. However, the characteristics of each of these interactions clearly separate the two classes of antigens. A. BINDINGTO MAJORHISTOCOMPATIBILITY COMPLEX Convincing evidence has been provided that the staphylococcal enterotoxins and closely related molecules bind to class I1 MHC molecules, and that this interaction is frequently required for T cell stimulation (Fleischer and Schrezenmeier, 1988; Carlsson et al., 1988; Chatila et al., 1988; Fraser, 1989; Scholl et al., 1989a,b, 1990a,b; Mollick et a/., 1989; White et al., 1989; Hermann et al., 1989, 1990). Dissociation constants for the binding of SEA and TSST-1 to human class I1 molecules have been estimated to be about to 10-"M (Fischer et al., 1989;Scholl et al., 1989a). The process ofbinding toxins to class I1 MHC molecules is much more permissive than that seen with the binding of conventional peptide antigens (White et al., 1989; Janeway et al., 1989; Scholl et al., 1990a). Whereas peptide antigens are dependent on allelic MHC residues for binding, toxin superantigeris bind to a wide variety of allelic and isotypic forms of class I1 MHC molecules in mouse and humans. In spite of this promiscuity in presentation, different MHC class I1 isotypes do differ in their ability to present these antigens. Several groups have studied the presentation of S. aureus enterotoxins by different human class I1 isotypes
118
BRIAN L. KOTZIN ET AL.
and determined that there is frequently a hierarchy of presentation with HLA-DR > HLA-DQ > HLA-DP (Scholl et al., 1989a,b, 1990a,b; Hermann et al., 1989; Herman et al., 1.990; Karp et al., 1990; Pontzer et al., 1991; Mollick et al., 1991). Studies of the murine H-2 molecules have demonstrated stronger binding and presentation by I-E (analog of HLA-DR) compared with I-A (analog of HLA-DQ) for certain toxins and differential binding among various alleles (White et al., 1989; Janeway et al., 1989; Scholl et al., 1990b; Mollick et al., 1991; Braunstein et al., 1992). For the presentation of certain exogenous superantigens, such as MAM, I-E expression in mice or HLA-DR expression in humans appears to be necessary (Cole et al., 1981,1986,1990a; Bekoff et al., 1987; Baccala et al., 1992).As discussed below, however, certain individual HLA-DR molecules can be extremely poor in binding and presenting. The interaction with class I1 MHC molecules appears to involve more than one site and varies among the toxins. For example, analysis of SEB residues important for MHC binding revealed two binding sites separated on the molecule (Kappler et al., 1992; Swaminathan et al., 1992), indicating that there may be two distinct MHC regions involved in the interaction. It has also been shown that SEB and TSST-1 bind to distinct sites on HLA-DR, in that these two toxins will not compete with each other for binding (Scholl et al., 1989b; Chintagumpala et al., 1991). Staphylococcal enterotoxin A (as well as SEE and SED) will, however, compete with both SEB and TSST-1, also suggesting that SEA binds to at least two separate regions (See et al., 1990; Chintagumpala et al., 1991). Consistent with this model, SEB or TSST-1 compete poorly with SEA. Staphylococcal enterotoxin A and SEE are usually presented well by HLA-DR. However, certain individual human DR moleciiles (e.g., DRw53) are extremely poor at binding and presenting these toxins (Herman et al., 1990). These data suggest that at least one of the critical interactions involves the DR /3 chain because the DR a chain is invariant. The sequences responsible for this binding were localized on the DR /3 chain by measuring toxin binding to a panel of chimeric class I1 molecules expressed on transfected cells as well as introducing point mutations in the DRw53 /3 chain (Herman et al., 1991b; Karp and Long, 1992). These studies showed that a histidine residue at position 81 on the outside of the a-helical wall of the peptide-binding groove was critically involved in toxin binding. It has been hypothesized that this residue promotes toxin binding by stabilizing a zincbinding site on SEAISEE (Fraser et al., 1992). This DR @chain residue is not important for binding either SEB or TSST-1 (Herman et
SUPERANTICENS
119
al., 1991b; Pontzer et al., 1991; Fraser et al., 1992). Major histocompatibility residues important for TSST-1 binding have also been identified by similar experimental approaches (Braunstein et al., 1992; Panina-Bordignon et al., 1992). Residues in both the a and /3 chains were shown to be involved, and all of the important residues could be placed on the a-helical walls of the peptide-binding groove. Although mutations in the a chain could eliminate high-affinity binding, low levels of binding persisted, again suggesting multiple attachment sites for these toxins. Overall, the above data indicate distinct sites for different toxins and more than one binding site for an individual toxin. Interaction with one of the binding sites can usually be shown to be stronger than the other. These sites appear to involve both the a and /3 chains of the MHC, and it is possible that the toxinMHC complex involves crosslinking of different MHC molecules. This crosslinking may be essential in the presentation of bacterial toxins to T cells. Evidence indicates that toxin superantigens bind to MHC at sites outside the allelically hypervariable groove that binds conventional peptide antigens (Dellabona et al., ~~1990; Karp ~ _ et_ al., 1990; Herman et al., 1991b; Karp and Long, 1992; Braunstein et al., 1992; PaninaBordignon et al., 1991). Thus, mutations in residues that point outward affect toxin but not peptide binding. In addition, mutations in the amino acids that make up the insides of the antigen-binding groove can have dramatic effects on the presentation of conventional peptides but litte influence on the presentation of toxins. These and other studies have also shown that the association of toxins with murine class I1 molecules does not inhibit presentation of an authentic antigenic peptide derived from a foreign protein. As emphasized above, exogenous superantigens also differ from conventional antigens in their lack of requirement for processing by antigen-presenting cells (APCs) (Janeway et aZ., 1989; Marrack and Kappler, 1990; Herman et al., 1990, 1991a; Fleischer and Schrezenmeier, 1988; Fleischer et al., 1989; Mollick et al., 1989; Scholl et al., 1989; Fraser, 1989; Kappler et al., 1992). Antigen-presenting cells that have been metabolically inactivated are still capable of presenting s. aureus enterotoxins but cannot present intact conventional protein antigens to T cells, and denaturation of SEA or SEB, or disruption of the intramolecular disulfide bond, results in a loss of superantigenic activity. Furthermore, as discussed above, point mutations dispersed throughout the amino-terminal region of SEB can inhibit MHC binding and/or TCR recognition. There has, however, been some controversy as to the processing requirements for certain ~
120
BRIAN L. KOTZIN ET AL.
exogenous superantigens, such as the Pseudomonas aeruginosa exotoxin A (Legaard et al., 1991) and the M . arthritidis mitogen (MAM) (Bauer et al., 1988).Other studies have reported that binding of SEA to MHC and SEA-induced T cell stimulation can be blocked with synthetic peptides matching SEA amino-terminal sequences, and that amino-terminal or carboxy-terminal fragments of certain S. aureus toxins can retain T cell-stimulating activity (Pontzer et al., 1989, 1991). At this time, these latter studies cannot be easily reconciled with the large body of data indicating that the globular twodomain conformation of the bacterial toxins is required for MHC binding and T cell stimulation. It is thought that MMTV superantigens (vSAGs) also must bind to class I1 MHC molecules before they can be recognized by T cell receptors. Evidence for this includes experiments showing that recognition of vSAGs by T cells can easily be inhibited by anti-class I1 antibodies (Peck et al., 1977). Additional studies have shown that presentation of vSAGs depends on the isotypes and alleles of class I1 molecules expressed. For example, animals that are H-2q (I-Aq,I-E-) cannot present the vSAG protein of Mtu-7 (vSAG-7),and other vSAGs (i.e., MTV-8, 9, 11) require an I-E molecule for presentation. Analogous to exogenous superantigens, for almost all MMTV vSAGs, I-E molecules appear to present in a more effective manner. Studies with an antibody to vSAG-7 also have shown that cells bearing I-Aq cannot express vSAG-7 on their surfaces (Winslow et al., 1992). Studies with murine T cell hybridomas also have shown that T cell recognition of MMTV superantigens, such as vSAG-7 and vSAG-9, can be strongly influenced by MHC molecules, resulting in distinct patterns of fine specificity (Blackman et al., 1992; Woodland et al., 1993). In further support of the role for class 11, many studies have also shown that expression of endogenous vSAGs in mice almost always deletes CD4+ cells better than CD8+ T cells bearing the target VP. However, despite all of the above evidence, almost nothing is known about the binding of the MMTV vSAGs to MHC class 11. These type I1 proteins and their degradation products [i.e., the 18.5-kDa product found on the cell surface (Winslow et al., 1992)l have structures that do not resemble the bacterial toxins and, thus far, it has not been possible to obtain a product that can be added exogenously to result in functional stimulation. It is also disappointing that immunoprecipitation studies with an anti-vSAG antibody could not demonstrate an association of the vSAG with MHC class I1 (Winslow et al., 1992).
SUPERANTIGENS
121
B. RECOGNITION BY T CELLRECEPTOR Residues of the toxin superantigens that affect binding to TCR Vps are discussed above. In this section, we focus on the regions of TCR VP that are involved in the complex. Several groups have proposed that the conservation of critical amino acids among immunoglobulin and TCR V domains predicts that the structure of the TCR V regions will be similar to those of immunoglobulin (Chothia et al., 1988; see White et al., 1993). The TCR V domains have, therefore, been envisioned as forming a series of antiparallel P strands with the loops corresponding to the three immunoglobulin complementarity-determining regions (CDRs) brought to one face to form the binding site for peptide antigen-MHC (Fig. 6).
FIG.6. Structure ofthe TCR Vp domain modeled from immunoglobulin. The view is of the solvent-exposed face composed of /3 strands B, D, and E in combination with a fourth nonantigen-binding loop, designated hypervariable region 4 (HVR4). Residues critically involved in superantigen binding are predicted to lie in HVR4. CDRl-3 are spatially separated from HVR4 and are involved in peptide-MHC recognition. Buried strands (C, C', F,and G) are predicted to be involved in the interaction between the aand P-chain V domains, and these strands (not labeled in diagram) would be sequestered from the solvent. [Adapted from White et al. (1993),J. Exp. Med. 177,122, by copyright permission of Rockefeller Univ. Press.]
122
BRIAN L. KOTZIN ET AL.
Some of the p strands (C, C’, F, G) are predicted to be involved in the interaction between the a-chain and p-chain V domains, and these strands would be sequestered from the solvent. Other p strands (A, B, D, E) in combination with a fourth non-antigen-binding loop [hypervariable region 4 (HV4)] between the I) and E strands are predicted to form a lateral solvent-exposed face (Fig. 6). In support of this model, amino acids important for peptide--MHC recognition have been shown to predominate in those regions that correspond to the CDR loops, especially CDR3, which is formed from the junctional Va-Ja and Vp-Dp-Jp gene segments. In contrast, residues involved in superantigen recognition appear to be in a different region of the TCR, away from the CDR loops. Choi et al., (199Oa) examined the SECB- and SEC3-binding site on human Vp13.2 by using chimeric human Vp/ mouse Va T cell receptor molecules. A 10-amino acid stretch in HV4 (residues 67-77) of the p chain was found to determine superantigen reactivity. In analogous studies, residues were exchanged between mouse Vp3, which binds SEC3, and mouse Vp17, which does not. The residues important for binding turned out to be HV4 residues 66,68, 72, and 74 (White et al., 1993). These residues were not involved in recognition of a peptide-MHC complex (Cyto C/I-Ek), whereas an allelic residue in CDRl was shown to be required for recognition of this peptide-MHC combination but was uninvolved in SEC3 recognition. Residues critically involved in vSAG recognition appear to be on the same solvent-exposed face but tend to include a slightly different part. In studies examining engagement of vSAG-7, Pullen et al. (1990b, 1991) showed that crucial residues of Vp8.Z were within the HV4 loop and a p strand leading up to CDRl (strand B) adjacent to this loop (see Fig. 6). Again, a residue in CDRl (position 24) of Vp8.2 was found to be critical for recognition of a peptide-I-A” complex without effect on vSAG recognition. In support of these observations, another group has shown that in wild mice mutations in HV4 (position 71) of Vp17 affected self-reactivity and deletion of Vp17+ T cells (Cazenave et al., 1990). Together, these data indicate that both endogenous and exogenous superantigens engage Vp on a solventexposed side face of the T cell receptor in a region distinct from that used to recognize peptide-MHC. Models of superantigen recognition by T cell receptors must also incorporate the required presentation by class 11 MHC. Thus far, there have been only rare examples of superantigen-mediated T cell stimulation in the absence of class I1 (Hewitt et al., 1992), and it has not been possible to detect direct binding of superantigens to T cells. One hypothesis at this time is that contact of the T cell receptor with
SUPERANTIGENS
123
both superantigen and class I1 is needed to generate adequate avidity for stimulation. Some studies suggest that orientation of the TCR is similar in both peptide-MHC and superantigen-MHC complex recognition, and data showing occasional contributions of Va to stabilization of the TCR-superantigen-MHC complex may support this model (see below). Alternatively, however, superantigens may undergo a conformational change after interaction with class I1 or undergo some type of crosslinking (see Section IV,A) such that direct binding to VP can take place. Furthermore, a given TCR may recognize a particular superantigen in association with different allelic and isotypic forms of MHC class 11, and upwardly pointing mutations in the a-helical barrels of MHC, which interfere with TCR recognition of peptide-MHC, usually have no effect on recognition of bound superantigen (see White et al., 1993). The overriding importance of HV4 also is consistent with the concept that few essential interactions between TCR and MHC take place during superantigen recognition. If a conformational change of the superantigen is critical, examination of the crystal structure of SEB does not reveal an obvious manner by which it occurs (Swaminathan et al., 1992). The anxiously awaited crystal structure of a class I1 MHC-toxin complex should more definitively address this question.
Modijcations of "V/3 rule" As discussed above, the evidence overwhelmingly suggests that there is a direct interaction of the TCR V/3 with the presented superantigen, whether bacterial toxin or MMTV superantigen (Choi et al., 1990a; Pullen et al., 1990b, 1991; Gascoigne and Ames, 1991; White et al., 1993). However, there is also evidence that other components of the TCR can influence this interaction. For example, in primary cultures in which T cells from Mls-1"-negative mice are being stimulated with Mls-1"-positive cells, skewing of Va use can be demonstrated in the responding VP6' T cell population (Vacchio et al., 1992). Other studies examining in vitro responses of T cell hybridomas to MMTV superantigen-expressing stimulator cells and to different toxins have shown that Va can influence the VP-superantigen interaction (Smith et al., 1992; Woodland et al., 1993; Blackman et al., 1992).A correlation with in uiuo deletion could be demonstrated only for those interactions that were relatively low affinity (Smith et al., 1992; Vacchio et al., 1992).These and other studies also have shown that other elements of the TCR (i.e., DPJP or Ja)may contribute to superantigen recognition (Smith et al., 1992; Candeias et al., 1991; Woodland et al., 1993).T h e participation of other TCR elements may be most important when the
124
BRIAN L. KOTZIN ET AL.
VP-superantigen interaction is of relatively low affinity. The mechanism by which V a sometimes influences superantigen recognition is unclear. Although V a residues may interact directly with the superantigen, it seems more likely that certain Vas may interfere with the VP-superantigen interaction or that certain Va-Vp combinations may be favored because of a stronger affinity for class I1 MHC. In support of this latter hypothesis, studies have clearly shown that MHC allelic residues can influence whether non-VP elements participate in superantigen recognition (Blackman et al., 1992; Woodland et al., 1993). Superantigens do not appear to be the only molecules recognized predominantly by TCR VP. Occasional peptide antigens have been described for which the analysis of TCR sequences expressed by responding T cell clones has disclosed a predominant VP and extensive @chain CDR3 and a-chain diversity. Thus far, the best example of this phenomenon involves the recognition of a tetanus toxin-derived peptide (p880-888) by human VP2+ T cells (Boitel et al., 1992). Interestingly, binding of this peptide to MHC is also promiscuous in that it can be presented by different HLA-DR molecules (PaninaBordignon et al., 1989; O’Sullivan et al., 1991). Residues of the VP critically involved in recognition of these peptides have not been further elucidated. OF ACCESSORY CD4 OR CD8 MOLECULES C. INVOLVEMENT In general, stimulation of T cells by most exogenous bacterial superantigens appears to be relatively CD4 independent. Both in culture and in patients with toxic shock syndrome (see Section V,B below), CD4 and CD8 subsets expressing the appropriate VPs appear to be stimulated nearly equally (Kappler et al., 1989a; Choi et al., 1990b; Marrack and Kappler, 1990). For occasional Vp-toxin combinations in uitro proliferation appears to take place mostly in the CD4 population (J. Abe et al., 1991). These may represent situations in which the TCR affinity for the superantigen-MHC complex is relatively low and stabilization by a CI34-class I1 MHC interaction is required. The injection of bacterial toxins into animals usually results in a greater degree of deletion or tolerance in the responding CD4 population, although both CD4 and CD8 subsets are usually affected (Kawabe and Ochi, 1990; see Section IV,D below). The in uitro response to Mls antigens (MMTV superantigens) is usually dominated by CD4+ T cells and can be inhibited by anti-CD4 antibodies (reviewed in Janeway et al., 1989; Janeway, 1991). However, some response within the CD8 population can usually be seen
SUPERANTIGENS
125
(MacDonald et al., 1990; Larsson-Sciard et al., 1990; Webb and Sprent, 1990a; Chvatchko and MacDonald, 1991; Hermann and MacDonald, 1993). In the in uiuo recognition of MMTV superantigens, intrathymic deletion occurs at the double-positive (CD4+, CD8+) stage of development and deletion is correspondingly apparent in both the mature CD4+ and CD8' subsets. The role of CD4 in this deletion process is somewhat controversial. In experiments blocking with antiCD4 antibody, VP-specific deletion in the CD8 population was prevented, suggesting a critical role for CD4 in the deletion process (Fowlkes et al., 1988; MacDonald et al., l988a). In contrast, experiments i n mice made deficient for CD4 by gene targeting demonstrated no need for CD4 except in situations in which the TCR affinity was lowest (Wallace et al., 1992). The injection of MMTV superantigenbearing cells into adult mice or infection with exogenous MMTV leads to VP-specific deletion and anergy preferentially in the responding CD4 population (Rammensee et al., 1989; Webb et al., 1990; Marrack et ul., 1991; Ignatowicz et al., 1992; see Section IV,D below). AND ANERCY IN ADULTANIMALSINJECTED D. T CELLDELETION WITH SUPERANTICENS From the information presented above, it might be inferred that the binding of an appropriate T cell to the complex of superantigen and class I1 MHC results in the intrathymic deletion of immature T cells as in the case of endogenous superantigens, or in the activation of mature T cells. For example, initial experiments showed that the in uiuo administration of SEB to neonatal animals causes specific deletion of VP8' T cells, whereas administration of a large dose to adult mice causes expansion of T cells bearing VP8 (White et al., 1989). However, subsequent experiments by other groups indicated that the initial expansion in adult animals is usually a transient response to superantigen exposure and that stimulated cells are eventually deleted and/or anergized (Kawabe and Ochi, 1990, 1991; Rellahan et al., 1990). This peripheral (nonthymic) tolerance phenomenon involving both anergy and deletion has been demonstrated after exposure to either MMTV or after injection of exogenous bacterial superantigens (Rammensee et al., 1989; Webb et al., 1990; Blackman et al., 1990; Jones et al., 1990; Dannecker et al., 1991; Ignatowicz et al., 1992; Kawabe and Ochi, 1990, 1991; Rellahan et aZ., 1990; MacDonald et al., 1991; McCormick et al., 1993; Gaur et al., 1993; Ochi et al., 1993). The in uiuo response to SEA injection has been studied in great detail (McCormick et al., 1993). Mice were acutely or chronically
126
BRIAN L. KOTZlN ET A t .
exposed to varying doses of SEA, and the percentage of T cells expressing different Vps was followed over time in different tissues, including lymph nodes, spleen, and blood. With most doses, an initial sharp rise in the proportion of reactive T cells (Vp3+and Vpll' T cells) was followed by a dramatic decline in all compartments studied. The disappearance occurred far too quickly to be explained by thymic deletion and appeared to be primarily secondary to death of target peripheral T cells. Responses were observed in both the CD4+ and CD8+ T cell subsets, although there were subtle differences in the sensitivity to different doses as well as kinetics of activation and deletion. In contrast to single doses, which caused the disappearance of SO-70% of reactive T cells, repeated exposure to even the smallest doses given (-10 ng given intraperitoneally every 2 days) was followed by almost complete deletion of target T cells. Interestingly, at these doses deletion was demonstrated in the apparent absence of a prior expansion, arguing against the hypothesis that the stronger the proliferative response, the more marked the deletion (Webb et al., 1992). Depletion of T cells was not permanent, recovery being mostly related to newly generated cells in the thymus in the absence of toxin. The mechanism responsible for the T cell deletion and/or anergy after experimental administration of superantigens is currently not known. It is possible that the ability of the exogenous superantigens to bind directly to all class I1 molecules (including those already carrying a peptide in their binding groove) may allow cells that are not appropriate antigen-presenting cells to interact with responding T cells. For example, resting B cells express class I1 MHC molecules and could bind superantigens but would not be able to present in an effective manner. Such incomplete presentation has been shown to lead to eventual nonresponsiveness rather than stimulation and expansion (Jenkins et al., 1987; Eynon and Parker, 1992). It is also possible that T cell tolerance is related to the size of the responding T cell population and perhaps to inappropriate cytokine release. In this regard, experiments have shown that IL-2 receptor (IL-2R) blockade can prevent SEB-induced deletion of Vp8' cells (Lenardo, 1991). In contrast, high doses of cyclosporin A can enhance peripheral T cell deletion caused b y SEB injection (Vanier and Prud'homme, 1992). Certain T cell populations appear to be differentially susceptible to deletiodanergy after engagement of superantigen in viuo. For example, although SEB-tolerant CD4+ T cells demonstrate a profound downregulation of their ability to produce IL-2 in
SUPERANTIGENS
127
uitro, they have the capacity to expand and produce IL-4 after infection with Nippostrongylus brasiliensis (Rocken et al., 1992). Others have also suggested that Thl-type cells [producing IL-2 and interferon y (IFN-y)3 may be more affected compared to IL-4 producing Th2-type cells after injection of superantigens.
E. T CELLSIGNALING IN RESPONSETO SUPERANTIGENS VERSUS CONVENTIONAL ANTIGENS Some studies have suggested that T cells engaging a superantigen may undergo signaling processes that are different when compared to those that occur after responding to conventional antigen-MHC. For example, Ca2+ fluxes and phosphoinositide metabolism have been reported to be absent in T cells proliferating in response to a superantigen (O’Rourke et al., 1990; Liu et al., 1991; Oyaizu et al., 1992). Distinct patterns of cytokine production have also been observed (Liu et al., 1991; Patarca et al., 1991). Other studies have not supported such claims (Hodes and Abe, 1993), and the reasons for the discrepancy in results are currently unclear. Whether superantigens deliver qualitatively rather than quantitatively different signals remains a controversial issue. F. DIRECTNON-TCELLEFFECTS As discussed above, superantigens directly bind to non-T cells via class I1 MHC molecules and have the ability to mediate signaling through this interaction. Signaling also may be enhanced by possible crosslinking of class I1 molecules. Direct non-T cell effects have been observed in a variety of systems where exogenous superantigens have been added to class 11-expressing non-T cells (Mourad et al., 1990, 1992; Fuleihan et al., 1991; Trede et al., 1991). The in uitro release of IL-1 and tumor necrosis factor a (TNF-a) from monocytes after addition of toxins has been observed by several groups of investigators (Ikejima et al., 1984, 1988; Parsonnet et al., 1985; Parsonnet and Gillis, 1988; Jupin et al., 1988; Fuleihan et aZ., 1991; reviewed in Parsonnet, 1989). It has been proposed that this release from non-T cells may be involved in the pathogenesis of toxic shock syndrome. However, even in in uitro systems, there is evidence that the presence of superantigen-reactive T cells greatly potentiates the release of such cytokines (See et al., 1992). Furthermore, the in uiuo release of TNF and subsequent consequences after exposure to staphylococcal enterotoxins appear to require the presence of responsive T cells (Marrack et al., Miethke et d.,1992).
128
BRIAN L. KOTZIN ET AL.
V. Superantigens and Human Disease
A. FOOD POISONING Staphylococcal enterotoxins are among the most common causes of food poisoning in humans (Bergdoll, 1979, 1985). All of these molecules are able to produce vomiting and diarrhea in humans and other primates following oral administration or duodenal instillation. These toxins, however, appear to have little, if any, enterotoxic effect after oral administration in other laboratory animals. The illness in humans after accidental ingestion of contaminated food (frequently improperly refrigerated baked ham, poultry, meat or potato salads, or cream-filled bakery goods) is almost always self-limited, and most cases are not brought to the attention of a physician or other medical personnel, To date, no studies addressing T cell activation have been performed in these patients. It has been suggested that the enterotoxic effects of the staphylococcal enterotoxins are secondary to their superantigen activity, that is, dependent on T cell stimulation (Marrack and Kappler, 1990). However, some evidence suggests that these two activities of the toxins may be distinct, including (1)the absence of enterotoxic effects in nonprimate animals despite their ability to respond to these molecules as superantigens, (2) the apparent mediation of the enteric reaction by neural cells with subsequent liberation of histamine from gut mast cells (Scheuber et al., 1985), and (3)the ability to separate enterotoxic from T cell stimulatory properties by using modified molecules (e.g., carboxymethylated SEB) or blocking molecules (Reck et al., 1988; Alber et al., 1990). Still, it is curious why (and how) these molecules with considerably different primary structures have continued to maintain both enterotoxic and superantigenic functions. Vomiting and diarrhea can also occur in patients with toxic shock syndrome, where the toxin may not interact directly with gastrointestinal cells and the evidence for required T cell stimulation is much stronger (see the next section).
B. TOXICSHOCK SYNDROME Toxic shock syndrome remains a serious disease characterized by rapid onset of fever, rash, shock, and multiorgan involvement (Todd and Fishaut, 1978; Davis et al., 1980; Shands et al., 1980; reviewed in Todd, 1988; Chesney, 1989; Bohach et ul., 1990). The clinical criteria for making the diagnosis are given in Table 111. Although the majority of cases are associated with menstruation and tampon use, similar manifestations have been increasingly recognized in other settings involving focal S. aureus infections (Todd, 1978). Stuphylococcus
SUPERANTIGENS
129
TABLE 111 CLINICAL CRITERIA USEDIN CLASSIFICATION OF TOXIC SHOCKSYNDROME^ Major criteria Fever [s102°F (38.8"C)I Rash (early erythroderma with desquamation 1 to 2 weeks after onset) Hypotension (shock) Minor criteria Diarrhea and/or vomiting Muscular involvement Mucous membrane hyperemia Renal (decreased renal function or pyuria) Hepatic (elevated liver enzymes) Hematological (thrombocytopenia) Central nervous system (disorientatiodmental status changes) a All major criteria and three or more minor criteria are required. Exclusions include positive throat cultures for group A streptococcus, positive bacterial cultures from blood or cerebrospinal fluid, or positive serological tests for Rocky Mountain spotted fever, leptospirosis, or measles. [Taken from Shands et al. (1980);Todd (1978); Bohach et d.(1990).]
aureus toxins, particularly toxic shock syndrome toxin 1 (TSST-l), have been implicated in the pathogenesis of the disease (Bergdoll et al., 1981; Schlievert et al., 1981; Todd, 1988; Chesney, 1989: Bohach et al., 1990). Greater than 90% of S. aureus strains isolated from menstruation-related cases produce TSST-1, and the absence of specific antibody to TSST-1 appears to be a major risk factor for acquiring the disease. In nonmenstrual cases, TSST-1 is the causal toxin in less than 50% of cases, and SEB and SEC production assume a greater role in pathogenesis (Bohach et al., 1990). Animal studies with purified TSST-1 have also provided strong support for its important role in the cause of toxic shock syndrome (de Azavedo, 1989). With the elucidation of its superantigenic nature, the mechanism by which TSST-1 results in disease has been clarified (see below). As reviewed above, a number of studies have demonstrated that S. uureus toxins, including TSST-1, have the properties of superantigens and are powerful stimulators of human T cells in culture. Using a modified polymerase chain reation (PCR) to quantitate TCR Vp expression in a population of T cell blasts, Choi et u2. (1989) demonstrated that the majority of T cells stimulated by TSST-1 in culture specifically expressed VP2. Based on these in uitro findings, studies were initiated in patients with toxic shock syndrome to determine if expansion of Vp2' T cells is a marker of the in uiuo disease process (Choi et al., 1990b). Figure 7 shows PCR results from a group of patients demonstrating marked and selective expansions of Vp2' T
130
BRIAN L. KOTZlN ET AL.
0 0.6
3
e
d
8 0.4 -
e
\
s
0.2
-
0 @ @
-*a
0 -
TSS Patients FIG.7. Vp2 expression in peripheral blood 'Tcells from 10 patients with toxic shock syndrome. The data are presented as a Vp/Caratio, as determined by a quantitative PCR technique. The broken line indicates two standard deviations above the mean value for a group of 15 control individuals. T cells expressing other Vps were not increased in patients compared with controls. [Taken from Choi et al. (1990b) and B. L. Kotzin, N. Ricalton, J. Kappler, and P. Marrack, unpublished observations, 1992.1
cells (Choi et al., 1990b and B. L. Kotzin, N. Ricalton, J. Kappler, and P. Marrack, unpublished observations, 1992). The increase in VP2 expression in peripheral blood T cells has been verified with specific monoclonal anti-VP2 antibodies (Kotzin et al., 1993). In some patients early after disease, over 60% of the peripheral blood T cells expressed VP2 with a reciprocal decrease in other VP-expressing populations. These studies have also indicated that, consistent with previous an uitro experiments, stimulation and expansion of VP2+ T cells takes place in both the CD4+ and CD8+ T cell populations. Serial evaluations of patients showed a gradual return of VP2 levels toward normal over a 1-to 4-month period (Choi et a!., 1990b).The fact that levels do normalize indicates that the increase in VP2' T cells occurs after the onset of toxic shock syndrome and is not a factor influencing susceptibility. It is important to note that the VPZ' T cell stimulation and expansion in toxic shock syndrome appear to be different from the peripheral T cell deletion and anergy observed in mice after systemic injec-
SUPERANTICENS
131
tion of toxins (see previous section). The gradual return toward normal VP2 levels in patients greatly contrasts with the marked decrease in VP8' or VP3+ T cell percentages to below pretreatment levels seen in mice approximately 1-2 weeks after injection of SEB or SEA, respectively (Kawabe and Ochi, 1990,1991; McCormick et al., 1993). Studies have also failed to document evidence of anergy in the VP2 population at any time after toxic shock syndrome (Kotzin et al., 1993). The reasons for these apparent differences in response to superantigens are currently unclear, but factors such as toxin characteristics, dose of toxin, rate of exposure, route (mucosal vs systemic), cytokines liberated during the toxic shock syndrome, effects of the marked inflammatory response, or other host differences may be involved. There is little information from human studies defining how TSST1 interacts with the host to produce the diverse manifestations of toxic shock syndrome (see Table 111).The hypotension appears to be associated with decreased vasomotor tone and increased capillary leakage, and the signs of shock in various organs are not accompanied by significant inflammatory lesions (Todd, 1988). Humoral mediators such as IL-1 and especially TNF-a have been implicated as a cause of shock, and elevated levels of TNF-a have been documented during acute disease (Tracey et al., 1986; Beutler and Cerami, 1987; Parsonnet, 1989).As discussed above, these cytokines can be shown to be released in cultures of monocytes or macrophages after the addition of TSST-1 (Ikejima et al., 1984, 1988; Parsonnet et al., 1985; Jupin et al., 1988; Parsonnet and Gillis, 1988; Fast et al., 1989; Parsonnet, 1989). However, in support of a critical role for T cells in the human disease, the above T cell repertoire studies in patients indicate that during toxic shock syndrome, T cell stimulation occurs on a massive scale. The responding T cell frequencies are orders of magnitude greater than that observed in response to conventional antigens. It is likely that these activated T cells (in both CD4+ and CD8+ populations) release large quantities of lymphokines such as IL-2 IFN-y, and lymphotoxin (TNF-P), all of which could be contributing to the induction of shock and other manifestations. Shocklike syndromes have been observed in patients with cancer being given large amounts of IL-2 and in allograft recipients being treated with anti-CD3 monoclonal antibodies (Belldegrun et al., 1987; Gaynor et al., 1988; Chatenoud et al., 1989). Perhaps equally important, the T cell activation process could also greatly enhance or be required for the release of IL-1 and T N F by non-T cells such as macrophages (see below). Animal studies have provided additional evidence suggesting that
132
BRIAN L. KOTZIN ET AL.
toxin-induced T cell stimulation is required for fully expressed toxic shock syndrome (Marrack et al., 1990; Miethke et al., 1992). Unmanipulated mice generally do not manifest shock after injection of S. aureus toxins such as SEB. Marrack et al. (1990) showed that intraperitoneal administration of large doses of SEB resulted in progressive weight loss. Nude mice (which almost completely lack T cells), mice treated with cyclosporin A (which blocks the T cell release of cytokines), or mice made deficient in those T cells able to respond to SEB, suffered little or no weight loss after SEB was given to them. Pretreatment of mice with D-galactosamine (D-Gal) leads to the development of lethal shock when mice are challenged parenterally with low doses of either LPS or SEB (Galanos et al., 1979; Lehmann et al., 1987; Miethke et al., 1992). Shock in both cases can be shown to be mediated by TNF. Interestingly, the lethal shock triggered by SEB was shown to be mediated by T cell activation, a conclusion based on the observation that T cell repopulation of SCID mice reconstituted sensitivity to SEB and cyclosporin A conferred protection (Miethke et al., 1992). Endotoxin-induced shock appeared to be independent of T cells. This study demonstrates that although TNF represents a key mediator in inducing shock after both endotoxin or superantigen administration, only disease induced by the latter is dependent on T cells. A severe and often fatal toxic shock-like syndrome, associated with group A streptococcus, has been reported with increasing frequency (Cone et al., 1987; Stevens et al., 1989; Begovac et al., 1990; Belani et al., 1991; Musser et al., 1991).The increased number of cases has been associated with the reemergence of bacterial isolates that produce streptococcal erythrogenic toxin A (SYE-A),and it has been postulated that the toxic shocklike syndrome is caused by SPE-A production in vivo (Cone et al., 1987; Stevens et al., 1989; Bohach et al., 1990; Hauser et al., 1991; Cleary et al., 1992). Many of the clinical characteristics are similar to those described above for s. aureus-associated disease (see Table 111). However, in contrast to S. aureus-related toxic shock syndrome, in which infection is almost always focal and frequently insignificant, the group A streptococci-induced cases have been more often associated with disseminated infection. In certain instances, it has been difficult to separate clinical features of sepsis from those directly related to exotoxin production. There are no published reports describing selective T cell subset stimulation in these patients. Studies analyzing peripheral blood T cells from a few patients with this syndrome have not been able to demonstrate selective stimulation of those VP-expressing subsets predicted to be affected
SUPERANTICENS
133
from in vitro studies, that is, Vp8, 12, and 14 (B. L. Kotzin, J. Lafferty, and N. Ricalton, unpublished observations, 1992). C. KAWASAKISYNDROME Kawasaki syndrome (KS) is an acute multisystem disease of unknown etiology that primarily affects infants and young children (Kawasaki, 1967; Morens et al., 1980; Rauch and Hurwitz, 1985; reviewed in Melish and Hicks, 1990; Melish, 1991). Although KS occurs worldwide in children of all racial groups, it is clearly most prevalent in Japan and in children of Japanese ancestry. Over 100,000 cases have been reported in Japan alone. Its primary clinical features are listed in Table IV. Of interest, a number of the features of acute KS are reminiscent of toxic shock syndrome. Indeed, occasional adult patients initially suspected to have a diagnosis of KS have not infrequently turned out to have toxic shock syndrome. Although the acute illness is generally self-limited, coronary artery abnormalities secondary to blood vessel inflammation (vasculitis) develop in 15 to 25% of untreated patients (Fujiwara and Hamashima, 1978; Kato et aZ., 1982). The consequences of this possible autoimmune complication can be devastating. In Japan and the United States, KS has become the most common cause of acquired heart disease in children. Studies have demonstrated that intravenous administration of high doses of gamma globulin during the acute phase of KS significantly decreases the occurrence of subsequent development of coronary artery abnormalities (Newberger et al., 1986, 1991; Nagashima et d., 1987; Furusho et al., 1984; Rowley et al., 1988). TABLE IV DIAGNOSTIC CRITERIA FOR KAWASAKISYNDROME
1. Fever for at least 5 days 2. Presence of at least four of the five following signs“:
a. Bilateral nonexudative conjunctival injection b. One of the following changes in the oropharynx: injected or fissured lips, “strawberry tongue,” or injected pharynx c. One of the following extremity changes: erythema of the palms or soles, edema of the hands or feet, or periungal desquamation d. Rash, primarily polymorphous e. Acute nonsuppurative cervical lymphadenopathy 3. Illness not explained by any other known disease a Patients with fewer than four of these five principal signs can be diagnosed as atypical KS if coronary aneurysms are present.
134
BRIAN L. KOTZIN ET AL.
Although its etiology is unknown, compelling evidence indicates that KS is secondary to infection (reviewed in Melish and Hicks, 1990; Melish, 1991). Geographical clustering has been noted for years. Epidemics and even pandemics in Japan have been described. The disease also has a seasonal variation, for example, outbreaks in the United States usually occur in the winter and spring. It should also be reemphasized that KS is a disease of young children. The observation that children less than 6 months or greater than 8 years of age are rarely affected suggests that the lack of a protective antibody against the putative KS agent may be an important risk factor for development of KS. Finally, in support of its infectious etiology, the acute onset of disease has clinical features that are strongly suggestive of infection. Studies have demonstrated that KS is associated with marked activation of T cells and monocytes/macrophages (reviewed in Leung, 1990, 1993). Based on these immunological features as well as the clinical features resembling a toxic shock syndrome, studies were initiated to determine if KS is associated with exposure to a superantigen such as a bacterial toxin (Abe et al., 1992, 1993). Using a quantitative PCR technique to analyze Vp expression, patients with acute KS demonstrated significantly elevated levels of circulating Vp2' T cells and, to a lesser extent, Vp8.l' 'Icells ' compared to other control groups (Fig. 8; Abe et al., 1992). None of the other 20 Vp populations analyzed were found to be significantly altered. Furthermore, during the convalescent phase of KS, there was a normalization of the levels of Vp2' and Vp8.l' T cells. Using monoclonal antibodies directed to Vp2 and Vp8.1, immunofluorescence analysis confirmed the selective increase in these subsets, primarily in the CD4' population. Increased percentages were noted only in the acute phase of disease. The percentages of Vp2' T cells as determined by monoclonal antibody reactivity and flow cytometry correlated linearly with Vp expression as quantitated by PCR. Interestingly, however, T cells from acute KS patients appeared to express proportionately higher levels of Vp2 transcripts per cell as compared to healthy controls or convalescent KS patients (Abe et al., 1993). As reviewed above, superantigens stimulate T cells almost solely through the Vp portion of the TCR and therefore induce an expansion of T cells mostly independent of the TCR p-chain junctional or CDR3 region. To determine whether Vp2' and Vp8' T cells were expanded in a superantigen-like (polyclonal) as opposed to an oligoclonal manner in acute KS, sequences of random cDNA clones containing Vp2 and Vp8 segments were analyzed (Table V) (Abe et al.,
135
SUPERANTICENS VDWCa ratio 0.5
-
P.0.002+ P=O.Ool-~ 8 I
0.4
-
D
0 Mlxed
0.3
0.2
-
0 I I 0 8 0 I
D I
m a 0
I
-
,
8
I
0.1
0.2
0.1
-
I D
D D
100
.O I
0.0 T
Ache KD
'
Disk
Heitthy . ' Heilhy Children Contmlr Adulr
0.04
I
Acute KD
- Healthy - Disease . Healthy 1
I
I
Children Controls
Adub
FIG.8. Vp2 and Vb8.l expression in T cells from patients with acute KS compared with various control groups. Data are presented as VpICa ratios as determined by a quantitative PCR technique. None of the other 20 Vp gene segments analyzed were abnormally increased in the acute KS group. [Taken from Abe et al. (1992) with permission. ] TABLE V ANALYSISOF T CELLRECEPTOR@-CHAIN JUNCTIONAL SEQUENCES CONTAINING Vp2 OR Vp8.l GENESEGMENTS FROM ACUTEKAWASAKI SYNDROME PATIENTS ~~
Analysis 1. Vp2
2. VP8
VpICa
Number of clones with same sequencesh
CDR3 regionsb
KD3 KD4 KD5 Normal mean
0.26 0.29 0.46 0.10
0119 0118 0118
Diverse Diverse Diverse
KD3 Normal mean
0.14 0.07
0117
Diverse
Donor
ratio"
-
-
-
Vp2 or Vp8 expression (determined by quantitative PCR) in peripheral blood T cells of individual KS patients compared to the mean value for a group of normal controls. cDNA clones encoding PCR-amplified Vp2 and Vp8 fragments were randomly sequenced. None of the cDNA clones from these KS patients had the same junctional sequence, and the junctional regions varied greatly in terms of residue number, sequence, and Jpusage (data not shown). [Adapted from Abe et al. (1993)l.
136
BRIAN L. KOTZIN ET AL.
1993). In patients with a two- to threefold increase in Vp2 gene expression, one might expect 60% of the sequences to reflect the abnormally expanded T cells. None of the Vp2 sequences was found to be identical to another, and considerable diversity among the sequences within each individual was apparent. The CDR3 regions varied greatly in terms of residue number and sequence, and a conserved motif was not apparent (Abe et al., 1993). These data therefore indicate that in the expansion of Vp2' cells in acute KS, Vp itself plays the dominant role in recognition. Together with the high frequency of responding cells in the circulation (>1 in 50), which is disproportionate to that expected for a response to a conventional antigen, the results appear to be most consistent with the hypothesis that T cell activation during the acute phase of KS is mediated by a superantigen. The hypothesis that a Vp2 (and, to a lesser extent, Vp8)-stimulating superantigen may be involved has provided a new approach to uncovering the infectious etiology of KS. In this regard, it was noted -that the pattern of VP stimulation was remarkably similar to that observed for certain streptococcal toxins (J. Abe et aZ., 1991,1992). Interestingly, some epidemiological studies in Japan analyzing antibodies to streptococcal proteins supported this hypothesis (Y. Abe et al., 1990). Results from one study, however, in which bacteria were cultured from KS patients and controls, suggest a stronger association of KS with TSST-1-producing S. aureus (Leung et al., 1993). In this study, S . aureus and group A streptococci were cultured from the throat, rectum, axilla, and/or groin of patients and tested for production of toxins, particularly those with Vp2-stimulatory activity. Toxins with superantigenic activity were isolated from 13 of 16 KS patients but from only 1 of 16 controls. Toxic shock syndrome toxin 1-secreting S. aureus was isolated in 11 cases whereas streptococcal toxins SPE-B and SPE-C were found in the other two. Of note, the toxin-producing S . aureus isolated from these patients had a highly unusual appearance and could easily have been mistaken for coagulase-negative staphylococci. The mechanism by which a superantigen can lead to the clinical manifestations of KS is unclear. As discussed above, bacterial toxins can result in marked production of IL-1 and TNF, and the production of increased levels of these as well as other cytokines has been reported to occur during the acute phase of KS (reviewed in Leung, 1990, 1993). As in toxic shock syndrome, T cell stimulation may be critical for cytokine release in acute KS. In KS, the subsequent induction of adhesion molecules that localize inflammatory cells to the
SUPERANTIGENS
137
vascular wall may also play an important role in the pathogenesis of vasculitis (Pober, 1988; Leung et al., 1989). It is also conceivable that superantigens could induce autoreactive T cells that are involved in the later and more persistent manifestations of KS (see Section V,D). D. RHEUMATOIDARTHRITIS Rheumatoid arthritis (RA) is an autoimmune disease characterized by inflammation of multiple joints. Mononuclear cell infiltration of the
synovial membrane is associated with destruction of articular cartilage and surrounding structures, and RA is a major cause of chronic disability in adults. Although the pathogenesis of this disease remains unresolved, both genetic and environmental factors have been implicated in the disease process. One major element of the genetic predisposition is related to the association of RA with certain class I1 MHC haplotypes, in particular, subtypes of HLA-DR4 and HLA-DR1 (reviewed in Gregerson et al., 1987). This association suggests that the disease may involve CD4+ T cells bearing ap T cell receptors. In support of this idea, CD4+ T cells represent a major component of the mononuclear cells in the synovial fluid and tissue of patients with RA, and the majority of these T cells bear surface markers that reflect in uiuo activation. Furthermore, partial elimination of T cells or T cell functions by a variety of techniques has led to an amelioration of disease activity in certain patients. Using a quantitative polymerase chain reaction technique, Paliard et al. (1991) analyzed the TCR p chain repertoires of synovial T cells in RA and compared these levels to the percentages of T cells bearing particular Vps in the peripheral blood of the same individual. Results were obtained from nine patients with rheumatoid arthritis and three patients with other types of inflammatory arthritis. All nine RA patients had a lower percentage of Vplcbearing T cells in their peripheral blood than in synovial fluid. This was not true in the three control patients. In fact, no Vpl4-bearing T cells could be detected in the majority of the RA patients (Fig. 9). A similar technique was used to examine Vp14 expression in a number of normal individuals, including several expressing HLA-DR4. In no cases were the percentages of Vp14+ T cells in peripheral blood undetectable. To determine the nature of the Vp14+ T cells in the joint, for example whether they had been expanded by a local response, the junctional sequences of the Vp14 gene segments expressed in synovial fluid T cells were determined. The results indicated that one or a few clones dominated
138
BRIAN L. KOTZIN ET AL.
0
0
0 0
2
<1
8 0
0
.f' @8
.AI
I
RA Patients
Controls
FIG.9. Evidence for the effects of a superantigen in rheumatoid arthritis (RA). A subset of RA patients studied had undetectable levels of Vp14+ T cells in their peripheral blood. Such low levels were not found in any normal individuals studied or in patients with other forms of arthritis. [Adapted from Paliard et al. (1991).]
the Vp14+ population in the synovial fluid of individual RA patients, whereas oligoclonality was less marked for other Vps (Paliard et al., 1991). The above results strongly implicate Vpl4-bearing T cells in the pathology of RA. Furthermore, they suggest that the etiology of RA may involve initial activation of VP14' T cells by a Vpl4-specific superantigen. Figure 10 depicts one model that could account for the low percentages of Vp14+ T cells in the peripheral blood but the presence of a few dominant clones of Vp14+ T cells in the synovial fluid of RA patients. The model is partially based on experiments in adult mice (described above) examining the fate of T cells confronted with viral or bacterial superantigelis after in d u o administration. These studies have revealed that, although initially inducing Vpspecific T cell activation, exposure to superantigens often leads to the eventual disappearance or inactivation of responding T cells. Therefore, Paliard et a2. (1991)postulated that RA patients have encountered a microbial superantigen specific for Vp14. This leads to the activation of most cells bearing Vp14, a few of which, because of their
SUPERANTICENS
139
Patient Encounters
c
T Cells Bearing Vp14 Activated
Small Subset of Activated Vp14 + T Cells CrossReact with a
T Cells Are
Migrate to the Joints and Produce Local Inflammation
c c
I T Cells (With Various Vps
Destructive Joint Inflammation Characteristic of Rheumatoid Arthritis
FIG. 10. A hypothetical mechanism proposed by Paliard et al. (1991) by which a microbial superantigen might lead to the development of rheumatoid arthritis. [Taken from Drake and Kotzin (1992) with permission.]
cross-reaction with particular self-antigens, home to the joints, where they are involved in the joint disease and are maintained by long-term stimulation with synovial self-antigens. The resultant inflammatory process also leads to the recruitment of additional T cells bearing various VP elements. The superantigen subsequently causes the elimination or anergy of most of the remaining VP14’ T cells in the periphery of patients. Based on the manner in which superantigens are presented, it seems unlikely that the HLA-DR4 subtype association with disease can be explained at the level of superantigen presentation. Instead, it is predicted that this association would be at the level of presentation of self-antigen in the synovium. Several studies in animal models of autoimmunity support the scheme proposed above, especially the hypothesis that superantigens can trigger disease exacerbations in susceptible or “primed” individuals. In experimental autoimmune encephalomyelitis (EAE),susceptible animals can be induced to develop central nervous system damage and
140
BRIAN L. KOTZIN ET AL.
neurological disease after immunization with brain antigens such as myelin basic protein (MBP)(reviewed in Zamvil and Steinman, 1990). The disease is dependent on the activation of CD4+ T cells reactive with MBP. In mice the disease is usually acute and, in those mice that survive the initial disease, complete recovery from neurological deficits usually occurs. In a subset of immunized mice, disease may not develop despite evidence for MBP-reactive CD4+ T cells in the peripheral lymphoid tissues. Studies have demonstrated that immunized mice that have either recovered from acute disease or not developed disease can be triggered to relapse by injection of staphylococcal enterotoxin (Schiffenbauer et d., 1993; Brocke et d . , 1993). The Vpspecificity of this effect is currently unclear. For example, in several murine strains, disease is dependent on MBP-reactive T cells that express Vp8.2. Bacterial toxins that target this Vp subset (e.g., SEB) triggered neurological disease. Surprisingly, SEA, which is not recognized by Vp8' T cells, also caused relapses. In analogous studies, superantigens were addressed as a triggering agent in an arthritis induced in rats by peptidoglycan-polysaccharide polymers isolated from the cell walls of group A streptococci (Schwab et al., 1993) and collagen-induced arthritis in mice (Cole et al., 1993).In these cases, Vp specificity of the injected toxin appeared to be important. For example, TSST-1, but not SPE-A, was able to trigger severe relapses and progressive joint destruction in the rat model, which was otherwise a nondestructive monophasic disease. Similarly, MAM and SEB, but not SEA, were able to exacerbate collagen-induced arthritis. Another group analyzing the T cell receptor repertoire of activated T cells in the synovium of rheumatoid arthritis patients has proposed that superantigens may be involved in the disease process (Howell et al., 1991b). Three gene families (Vp3, V1114, and Vp17) were found to be overrepresented in the majority of the five synovial samples analyzed. In individual patients, the repertoires expressing these Vps were dominated by a single rearrangement, indicative of clonal expansion in the synovium. These investigators noted the high sequence similarity among these Vps, especially in the HV4 region. Based on this homology and the known function of this region in binding superantigens (see above), it was hypothesized that a superantigen may be involved in the selection of T cells expressing such related T cell receptors.
ASSOCIATEDWITH E. AUTOIMMUNEDISEASES STREPTOCOCCAL INFECTION As reviewed above, infections with group A streptococci have been associated with disease syndromes that suggest a direct role for associated superantigen production. For example, scarlet fever and a toxic
SUPERANTICENS
141
shock syndrome-like illness have been linked to the production of the streptococcal erythrogenic toxins, especially SPE-A (Cone et al., 1987; Stevens et al., 1987; Bohach et al., 1990; Hauser et al., 1991; Cleary et al., 1992). In addition, it is well known that infection with group A streptococci may evoke autoimmune disease syndromes in the susceptible host, such as rheumatic fevedrheumatic heart disease and poststreptococcal glomerulonephritis. A major virulence factor of these organisms appears to be the surface antigen M proteins. Furthermore, especially in the case of rheumatic heart disease, evidence suggests that autoimmune-mediated damage may be related to molecular mimicry between M proteins and host cardiac tissue proteins (Dale and Beachey, 1985; reviewed in Stollerman, 1990).As reviewed above, it is interesting that studies suggest that streptococcal M proteins have the properties of a superantigen (Tomai et al., 1990,1991,1992). Studies currently underway are attempting to determine whether superantigen-mediated stimulation by M proteins is involved in their ability to induce autoimmunity.
F. SYSTEMIC ( LUPUS-LIKE) AUTOIMMUNE DISEASES Previous studies have shown that a chronic graft-versus-host disease in mice can lead to IgG autoantibody production, including IgG antiDNA antibody production, and a lupus-like renal disease (reviewed in Portanova and Kotzin, 1988). This induced autoimmune disease appears to involve a direct interaction between the T cells of the donor and the B cells of the host. Friedman et al. (1991b) have proposed that a similar situation could result after superantigen exposure. Thus an exogenous superantigen could “bridge” a generalized CD4+ T cell-B cell interaction, resulting in B cell activation and potential autoantibody production. These authors have demonstrated that one particular bacterial superantigen, M. arthriditis mitogen (MAM),can result in marked B cell activation and immunoglobulin production in vitro (Tumang et al., 1990). Although provocative, there are currently no data to support a potential role for microbial superantigens in lupuslike disease and, thus far, T cell repertoire studies in both human systemic lupus and murine models of lupus have not found VP-specific T cell expansions characteristic of such an exposure (Kotzin et al., 1988, 1989a,b; K. Hobbs and B. L. Kotzin, unpublished observations, 1993). Furthermore, as discussed above, in contrast to the prediction of generalized T cell activation leading to B cell stimulation, a number of studies have noted eventual tolerance of the responding T cell populations after systemic exposure to microbial superantigens (see above). Indeed, toxin (SEB) administration to mice with lupus-like disease (MRL-’l”llpr) has been shown to selectively delete target VP
142
BRIAN L. KOTZIN ET AL.
populations and to prevent autoantibody production and autoimmune disease (Kim et al., 1991).
G. IMMUNODEFICIENCY DISEASE A number of different groups have proposed that HIV-1, the cause of human acquired immunodeficiency syndrome (AIDS), encodes a superantigen that may be involved in infectivity or host CD4+ T cell depletion. Support for the idea that a retrovirus-encoded superantigen can enhance infectivity and replication is reviewed above in Section II,D, which describes relevant studies of MMTV. It also has been proposed that induced apoptosis of T cells is a component in the CD4+ T cell depletion observed in AIDS patients and other HIV-infected individuals. One model suggests that engagement of CD4 by viral gp120 predisposes the cell to undergo cell death when its TCR is engaged (Newel1 et al., 1990; Terai et al., 1991; Groux et al., 1992; Banda et al., 1992). A virally encoded superantigen could therefore serve as a "death blow" resulting in Vp-selective deletion. Laurence et al. (1992) have presented evidence that a superantigen encoded by HIV affects infectivity arid viral replication. CD4+ T cell lines expressing different TCR Vps were infected in uitro with HIV-1. Markedly different titers of HIV-1 virion production resulted, dependent on the Vp expressed. For example, Vpl2' T cell lines from several different donors reproducibly yielded up to 100-fold more gag gene product ( ~ 2 4antigen) ~ " ~ compared to lines expressing other Vps. Consistent with the notion that this involved a presented superantigen, Vp selectivity of virion production depended on antigenpresenting cells expressing class I1 MHC but was not MHC restricted. Enhanced virion production in the V1312+ cells appeared to depend on T cell activation because polyclonal stimulation of lines expressing other Vps enhanced production to levels of the Vpl2' lines. These investigators also demonstrated evidence for in vivo superantigen expression. Thus Vpl2' cells freshly isolated from AIDS patients were enriched for both gp120 expression and HIV-1 gag DNA. Furthermore, antigen-presenting cells from HIV-1-seropositive, but not HIV-1negative, donors preferentially stimulated Vpl2' T cells. Of note, these investigators could not demonstrate that Vp12+ T cells were preferentially deleted in infected patients. Overall, the above data are consistent with the hypothesis that an HIV-encoded superantigen is presented by non-T cells and results in target T cell proliferation that facilitates viral replication. Other investigators have proposed that CD4+ T cell deletion during HIV infection is Vp selective and involves a superantigen. In one study, a quantitative PCR technique was used to examine TCR Vp and
SUPERANTIGENS
143
Vct expression in patients with AIDS (Imberti et al., 1991).A number of Vp genes were noted to be expressed at low levels in patients compared to controls. However, because the deleted Vp families were mostly ones that are poorly represented in the normal circulating T cell pool, the selective decreases observed may have represented a more general (and nonspecific) decrease in TCR gene expression. The Vp17 deletion predicted from this study has not been verified with monoclonal anti-VD antibodies (Laurence et al., 1992; Posnett et al., 1993). H. LYMPHOPROLIFERATIVE DISEASES 1 . Epstein-Burr Virus Epstein-Barr virus (EBV) is a ubiquitous human pathogen that has evolved an effective strategy for infection, presistence, and dissemination (reviewed in Strauss et al., 1993). It occurs in more than 90% of the population, most often without clinical consequences. Many of the genes encoded by this virus appear to ensure its persistence in B lymphocytes, but the clinical outcome of infection appears to depend on host immune responses. Infectious mononucleosis develops in a subset of individuals, usually infected adolescents. With rare exceptions, infectious mononucleosis is a self-limited disease. In immunosuppressed individuals, lymphoproliferation of virus-infected B cells can lead to serious sequelae, including B cell tumors, and a fatal outcome. In certain populations, Burkitt’s lymphoma, nasopharyngeal carcinoma, non-Hodgkin B cell lymphomas, and central nervous system (CNS) lymphomas, rare T cell lymphomas, and some thymomas have been associated with EBV infection. The immunological manifestations of EBV infection suggest a possible role for a superantigen. For example, in infectious mononucleosis, marked polyclonal T cell (mostly CD8+ T cell) activation is almost always observed (reviewed in Strauss et ul., 1993). In one study using a quantitative PCR to analyze V@ expression, selectivity for certain Vp populations was suggested (Smith et al., 1993).In patients with immunodeficiency leading to progressive and sometimes malignant B cell proliferation, T cell stimulation can also be documented. The TCR repertoire of this latter population is undefined. In patients with Burkitt’s lymphomas, there is little information about possible T cell influences. A murine model to study EBV-induced human B cell tumor development has evolved. SCID mice can be successfully reconstituted with human adult lymphocytes (Mosier et al., 1988).If the latter come from donors with serological evidence of previous EBV infection, the mice
144
BRIAN L. KOTZIN ET AL.
frequently develop a fatal lymphoproliferative disease associated with the presence of oligoclonal EBV+ tumors of human B cell origin (Mosier et al., 1988; Cannon et al., 1990; Rowe et al., 1991; Purtilo et al., 1991).The outgrowth of these tumors in SCID mice simulates EBV+ tumors arising in immunosuppressed humans. Interestingly, the development of EBV+ tumors in SCIJI mice appears to require coinjected T cells, and administration of agents that prevent T cell activation, such as cyclosporin A, also prevents B cell tumor development (Veronese et al., 1992). The nature of the B cell-T cell interaction or cell influence on tumor development is currently unclear. There is currently no direct evidence that EBV encodes a superantigen or that a superantigen is involved in any of the clinical consequences of disease. As discussed in Section 111,C,2, herpesvirus saimiri, which has a genomic organization similar to that of EBV (Gompels et al., 1988),has an open reading frame within its genome that encodes a product with homology to the MMTV superantigens (Nicholas et al., 1990; Thomson and Nicholas, 1991). The genetic region of interest in EBV is, however, distinct from herpesvirus saimiri, and the products from this region are undefined (Gompels et al., 1988; Parker et al., 1990). Studies have shown that an EBV-transformed B cell line (Raji) stimulates T cell hybridomas expressing Vp3 irrespective of TCR (Y chain and with other characteristics indicative of stimulation by a superantigen (Donahue et al., 1993). Although other EBVtransformed B cell lines did not stimulate VP3+ cells, it is unclear whether superantigens with other VP specificities are encoded by different EBV isolates, similar to MMTV.
2. B Cell Lymphomas There is evidence that some human B cell lymphoma cells can elicit strong responses from T cells, and that their growth can be enhanced by T cell stimulation and T cell-derived lymphokines (Veronese et al., 1992).The early events that promote B cell dysregulation and oligoclonal expansion are mostly unclear. The hypothesis that in some cases the in viuo progressive growth of a B cell lymphoma may involve expression of a superantigen and T cell stimulation is supported by murine studies. In SJL/Jmice, lymphomas of the B cell lineage spontaneously develop with age in nearly 90% ofmice (Nakauchi et al., 1987). These tumors have been referred to as reticulum cell sarcomas (RCSs) and resemble histopathologically certain types of human B cell lymphomas. Studies have shown that the SJL/J tumor cells stimulate syngeneic CD4+ T cells, which is dependent on MHC class I1 expression (Ponzio et al., 1977). Furthermore, T cell responsiveness to the tumor in uiuo is required for progressive growth or to support the
SUPERANTICENS
145
growth of transplanted tumors (Lerman et al., 1976; Katz et aZ., 1980; Sakano and Bonavida, 1986; Ohnishi and Bonavida, 1987; Lasky et al., 1988; Alisauskas and Ponzio, 1989). Studies have indicated that stimulation of T cells in uitro is Vp selective (Katz et al., 1988,1989; Tsiagbe et al., 1991, 1993a,b). Furthermore, elimination of responding T cells by anti-Vp antibodies or by genetic depletion can prevent tumor growth (Katz et al., 1980,1988,1989). Because two groups of investigators have noted responding subsets expressing different Vps (Katz et al., 1988, 1989; Tsiagbe et al., 1993a,b), it is possible that the putative tumor-expressed superantigen may vary in different SJL strains, similar to that described above for MMTV. In one set of studies, stimulation of Vpl6+ T cells by the spontaneous SJL B cell lymphomas was correlated with expression of a particular MMTV vSAG (Tsiagbe et al., 1993b). VI. Summary
In the past few years, there has been a virtual explosion of information on the viral and bacterial molecules now known as superantigens. Some structures have been defined and the mechanism by which they interact with MHC class I1 and the Vp region of the T cell receptor is being clarified. Data are accumulating regarding the importance of virally encoded superantigens in infectivity, viral replication, and the life cycle of the virus. In the case of MMTV, evidence also suggests that superantigens encoded by a provirus may be maintained by the host to protect against future exogenous MMTV infection. Experiments in animals have also begun to elucidate the dramatic and variable effects of superantigens on responding T cells and other immune processes. Finally, the role of superantigens in certain human diseases such as toxic shock syndrome, some autoimmune diseases like Kawasaki syndrome, and perhaps some immunodeficiency disease such as that secondary to HIV infection is being addressed and mechanisms are being defined. Still, numerous important questions remain. For example, it is not clear how superantigens with such different structures, for example, SEB, TSST-1, and MMTV vSAG, can interact with MHC and a similar region of the TCR in such basically similar ways. It remains to be determined whether there are human equivalents of the endogenous murine MMTV superantigens. The functional role of bacterial superantigens also remains to be explained. Serious infection and serious consequences from toxin-producing bacteria are relatively rare events, and it is questionable whether such events are involved in the selection pressure to maintain production of a functional superantigen.
146
BRIAN L. KOTZIN ET AL.
Hypotheses to explain these molecules, which can differ greatly in structure, include T cell stimulation-mediated suppression of host responses or enhancement of environments for bacterial growth and replication, but substantiating data for these ideas are mostly absent. It also seems likely that only the tip of the iceberg has been uncovered in terms ofthe role of superantigens in human disease. Unlike toxic shock syndrome, other associations, especially with viral superantigens, may be quite subtle and defined only after considerable effort. The definition of these molecules and mechanisms of disease may result in new therapeutic strategies. Finally, it is a.pparent that superantigens have dramatic effects on the immune system. One wonders whether these molecules or modifications of them can be used as specific modulators of the immune system to treat disease.
REFERENCES Abe, J., Forrester, J., Nakahara, T., Lafferty, J. A., Kotzin, B. L., and Leung, D. Y. (1991). Selective stimulation ofhuman T cells with streptococcal erythrogenic toxins A and B. J . Zmmunol. 146,3747-3750. Abe, J., Kotzin, B. L., Jujo, K., Melish, M. E., Glode, M. P., Kohsaka, T., and Leung, D. Y. (1992). Selective expansion of T cells expressing T-cell receptor variable regions Vp2 and Vp8 in Kawasaki disease. Proc. Natl. Acad. Sci. U.S.A. 89,4066-4070. Abe, J., Kotzin, B. L., Meissner, C., Melish, M .E., Takahashi, M., Fulton, D., Romagne, F., Malissen, B., and Leung, D. Y. M. (1993). Characterization of T cell repertoire changes in acute Kawasaki disease.J. Exp. Med. 177,791-796. Abe, R., Vacchio, M. S., Fox, B., and Hodes, €1. J. (1988). Preferential expression of the T-cell receptor Vp3 gene by Mlsc reactive T cells. Nature (London) 335,827-830. Abe, R., Foo-Phillips, M., and Hodes, R. J. (1991). Genetic analysis of the MIS system. Formal MIS typing of the commonly used_ inbred _ strains. Zrnmunogenetics 33,62-73. Abe, Y.,Nakano, S., Nakahara, T , Gmezawa, Y.,Kato, I., Ushijima, H., Yoshino, K., Ito, S., Noma, S., and Okitsu, S. (1990). Detection of serum antibody by the antimitogen assay against streptococcal erythrogenic toxins. Age distribution in children and the relation to Kawasaki disease. Pediatr. Res. 27, 11-15. Acha-Orbea, H., and Palmer, E. (1991). Mls--a retrovirus exploits the immune system [see comments]. Zmmunol. Today 12,356-361. Acha-Orbea, H., Shakhov, A. N., Scarpellino, L., Kolb, E., Muller, V., Vessaz-Shaw, A., Fuchs, R., Blochlinger, K., Rollini, P., and Billotte, J. (1991). Clonal deletion of Vpl4-bearing T cells in mice transgenic for mammary tumor virus. Nuture (London) 350,207-21 1. Acha-Orbea, H., Scarpellino, L., Shakov, A. N., Held, W., and MacDonald, H. R. (1992). Inhibition of mouse mammary tumor virus-induced T cell responses in oioo by antibodies to an open reading frame pr0tein.J. Exp. Med. 176, 1769-1772. Alber, G., Hammer, D. K., and Fleischer, B. 1990). Relationship between enterotoxicand T lymphocyte-stimulating activity of staphylococcal enterotoxin Br. J. Zmmunol. 144,4501-4506. Alisauskas, R. M., and Ponzio, N. M. (1989). T-helper-cell-specific monoclonal antibody inhibits growth of B-cell lymphomas in syngeneic SJL/J mice. Cell Zrnrnunol. 119, 286-303. 1:
SUPERANTIGENS
147
Allured, V. S., Collier, R. J., Carroll, S. F., and McKay, D. B. (1986).Structure ofexotoxin A of Pseudomonas aeruginosa at 3.0-b; resolution. Proc. Natl. Acad. Sci. U.S.A. 83, 1320- 1324. Atkin, C. L., Cole, B. C., Sullivan, G . J., Washburn, L. R., and Wiley, B. B. (1986). Stimulation of mouse lymphocytes by a mitogen derived from Mycoplasma arthritidis. V. A small basic protein from culture supernatants is a potent T cell mitogen. J. Irnmunol. 137, 1581-1589. Baccala, R., Kono, D. H., Walker, S., Balderas, R. S., and Theofilopoiilos, A. N. (1991). Genoniically imposed and somatically modified human thymocyte Vp gene repertoires. Proc. Natl. Acud. Sci. U.S.A. 88,2908-2912. Baccala, R., Smith, L. R., Vestberg, M., Peterson, P. A., Cole, B. C., and Theofilopoulos, A. N. (1992).Mycop/asma arthritidis mitogen. VP engaged in mice, rats, and humans, and requirement of HLA-DRct for presentation. Arthritis Rheum. 35,434-442. Banda, N. K., Bernier, J., Kurahara, D. K., Kurrle, R., Haigwood, N., Sekaly, R. P., and Finkel, T. H. (1992). Crosslinking CD4 by human immunodeficiency virus gp120 primes T cells for activation-induced apoptosis. J . Exp. Med. 176, 1099-1 106. Bauer, A., Rutenfranz, I., and Kirchner, H. (1988). Processing requirements for T cell activation by Mycoplasma arthritidis-derived mitogen. Eur. J . Immunol. 18, 21092112. Beachey, E. H., Stollerman, G . H., Johnson, R. H., Ofek, I., and Bisno, A. L. (1979). Human immune response to immunization with a structurally defined polypeptide fragment of streptococcal M protein. J . Exp. Med. 150,862-877. Begovac, J., Marton, E., Lisic, M., Beus, I., Bozinovic, D., and Kuzmanovic, N. (1990). Group A P-hemolytic streptococcal toxic shock-like syndrome. Pediatr. Infect. Dis. J . 9,369-370. Bekoff, M. C., Cole, B. C., and Grey, H. M. (1987). Studies on the mechanism of stimulation of T cells by the Mycoplusma arthritidis-derived mitogen. Role of class I1 IE molecules. J . Irnmunol. 139,3189-3194. Belani, K., Schlievert, P. M., Kaplan, E. L., and Ferrieri, P. (1991). Association of exotoxin-producing group A streptococci and severe disease in children. Pediatr. Infect. Dis. J. 10,351-354. Belldegrun, A., Webb, D. E., Austin, H. A., Steinberg, S. M., White, D. E., Linehan, W. M., and Rosenberg, S. A. (1987). Effects of interleukin-2 on renal function in patients receiving immunotherapy for advanced cancer. Ann. Intern. Med. 106,817822. Ben-Nun, A., Liblau, R. S., Cohen, L., Lehmann, D., Tournier-Lasserve, E., Rosenzweig, A., Zhang, J. W., Raus, J. C., and Bach, M. A. (1991). Restricted T-cell receptor VP gene usage by myelin basic protein-specific T-cell clones in multiple sclerosis: Predominant genes vary in individuals. Proc. Natl. Acad. Sci. U.S.A. 88, 2466-2470. Bergdoll, M. S. (1979). Staphylococcal intoxications. In “Foodborne Infections and Intoxications” (H. Riemann and F. L. Bryan, eds.), pp. 443-493. Academic Press, New York. Bergdoll, M. S. (1985). The staphylococcal enterotoxins-an update. In “The Staphylococci” (J. Jeljaszewics, ed.), pp. 247-266. Gustav Fisher Verlag, New York. Bergdoll, M. S., Crass, B. A., Reiser, R. F., Robbins, R. N., and Davis, J. P. (1981).A new staphylococcal enterotoxin, enterotoxin F, associated with toxic-shock-syndrome Staphylococcus aureus isolates. Lancet 1, 1017-1021. Betley, M. J. and Mekalanos, J. J. (1985). Staphylococcal enterotoxin A is encoded by phage. Science 229, 185-187.
148
BRIAN L. KOTZIN ET AL.
Beutler, B., and Cerami, A. (1987). Cachectin: hdore than atumor necrosis factor. N . E n g l . J. Med. 316,379-385. Beutner, U.,Frankel, W. N., Cote, M. S., Coffin, J. M., and Huber, B. T. (1992). Mls-1 is encoded by the long terminal repeat open reading frame of the mouse mammary tumor provirus Mtv-7. Proc. Natl. Acad. Sci. U.S.A 89,5432-5436. Bill, J., Kanagawa, O., Woodland, D. L., and Pa!mer, E. (1989).The MHC molecule I-E is necessary but not sufficient for the clonal deletion of Vp1l-bearing T cells. 1. E x p . Med. 169,1405-1419. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C. (1987). The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature (London)329,512-518. Blackman, M. A., Burgert, H. G., Woodland, D. L., Palmer, E., Kappler, J. W., and Marrack, P. (1990). A role for clonal inactivation in T cell tolerance to Mls-la. Noture (London)345,540-542, Blackman, M. A,, Lund, F. E., Surman, S., Corley, R. B., and Woodland, D. L. (1992). Major histocompatibility complex-restricted recognition of retroviral superantigens by V/317+T cells. J. E x p . Med. 176,275-280. Blanco, L., Choi, E. M., Connolly, K., Thompson, M. R.,and Bonventre, P. F. (1990). Mutants of staphylococcal toxic shock syndrome toxin 1: Mitogenicity and recognition by a neutralizing monoclonal antibody. Infect. Immun. 58,3020-3028. Bohach, G. A., and Schlievert, P. M. (1987). Nucleotide sequence of the staphylococcal enterotoxin C1 gene and relatedness to other pyrogenic toxins. Mol. Gen. Genet. 209, 15-20. Bohach, G. A., Handley, J. P., and Schlievert, P. M. (1989). Biological and immunological properties of the carboxyl terminus of staphylococcal enterotoxin C1. Infect. Immun. 57,23-28. Bohach, G. A,, Fast, D. J., Nelson, R. D., and Schlievert, P. M. (1990). Staphylococcal and streptococcal pyrogenic toxins involved in toxic shock syndrome and related illnesses. Crit. Rev. Microbiol. 17,251-272. Boitel, B., Ermonval, M., Panina-Bordignon, P., Mariuzza, R. A., Lanzavecchia, A., and Acuto, 0. (1992). Preferential V/3 gene usage and lack ofjunctional sequence conservation among human T cell receptors specific for a tetanus toxin-derived peptide. J. E x p . Med. 175,765-777. Bowness, P., Moss, P. A., Tranter, H., Bell, J. I., and McMichae1,A. J. (1992). Clostridium perfringens enterotoxin is a superantigen reactive with human T cell receptors Vb6.9 and Vp22.J. E x p . Med. 176,893-896. Brandt-Carlson, C . , Butel, J. S., and Wheeler, D. (1993). Phylogenetic and structural analyses of MMTV LTR ORF sequences ofexogenous and endogenous origins. Virology (in press). Braiin, M. A., Gerlach, D., H a w i g , U. F., Ozegowski, J.-H., RomagnB, F., Carrel, S., Kohler, W., and Fleischer, B. (1993). Stirnillation of human T cells by streptococcal “superantigen” erythrogenic toxins (scarlet fever toxins).]. Immunol. 150,2457-2466. and Karp, D. (1992). SeBraunstein, N. S., Weber, D. A., Wang, X. C., Long, E. 0.. quences in both class I1 major histocompatibility complex a and p chains contribute to the binding of the superantigen toxic shock syndrome toxin 1.1. E x p . Med. 175, 1301- 1305. Brocke, S., Gaur, A,, Piercy, C., Gijbels, K., Fathman, C. G., and Steinman, L. (1993). Exacerbation or activation of subclinical EAE mediated by exposure to SEB. (in press). Brown, J . H., Jardetzky, T., Saper, M. A., Samraoui, B., Bjorkman, P. J., and Wiley, D. C. (1988). A hypothetical model of the foreign antigen binding site of class I1
SUPERANTIGENS
149
histocompatibility molecules [published erratum appears in Nature (London) 23;333(6175):786]. Nuture (London) 332,845-850. Burns, F. R., Li, X. B., Shen, N., Offner, H., Chou, Y. K., Vandenbark, A. A., and Heber-Katz, E. (1989). Both rat and mouse T cell receptors specific for the encephalitogenic determinant of myelin basic protein use similar Vcx and Vp chain genes even though the major histocompatibility complex and encephalitogenic determinants being recognized are different.]. Exp. Med. 169,27-39. Buxser, S . , Bonventre, P. F., and Archer, D. L. (1981). Specific receptor binding of staphylococcal enterotoxins by murine splenic lymphocytes. Infect. Immun. 33,827833. Callahan, J. E., Herman, A., Kappler, J. W., and Marrack, P. (1990). Stimulation of B1O.BR T cells with superantigenic staphylococcal toxins. ]. lmmunol. 144, 24732479. Candeias, S., Waltzinger, C., Benoist, C., and Mathis, D. (1991). The Vp17+ T cell repertoire: Skewed JP usage after thymic selection; dissimilar CDR3s in CD4+ versus CD8+ cells.]. Exp. Med. 174,989-1000. Cannon, M. J., Pisa, P., Fox, R. I., and Cooper, N. R. (1990). Epstein-Barr virus induces aggressive lymphoproliferative disorders of human B cell origin in SCID/hu chimeric mice.]. C h i . lnoest. 85, 1333-1337. Carlsson, R., Fischer, H., and Sjogren, H. 0. (1988). Binding of staphylococcal enterotoxin A to accessory cells is a requirement for its ability to activate human T cells. J . lmniunol. 140,2484-2488. Chatenoud, L., Ferran, C., Reuter,A., Legendre,C., Gevaert,Y.,Kreis, H., Franchimont, P., and Bach, J. F. (1989). Systemic reaction to the anti-T-cell monoclonal antibody OKT3 in relation to serum levels of tumor necrosis factor and interferon-gamma [corrected] [letter] [published erratum appears in N Engl] Med 6;321(1):63]. N . Engl. ]. Med. 320, 1420-1421. Chatila, T., Wood, N., Parsonnet, J., and Geha, R. S. (1988). Toxic shock syndrome toxin-1 induces inositol phospholipid turnover, protein kinase C translocation, and calcium mobilization in human T cells. ]. Zmmunol. 140, 1250-1255. Chesney, P. J. (1989). Clinical aspects and spectrum of illness of toxic shock syndrome: Overview. Reo. Infect. Dis. ll(Supp1. I), S1-S7. Chintagumpala, M. M., Mollick, J. A., and Rich, R. R. (1991). Staphylococcal toxins bind to different sites on HLA-DR.]. lmmunol. 147,3876-3881. Choi, Y., Kotzin, B., Herron, L., Callahan, J., Marrack, P., and Kappler, J. (1989). Interaction of Staphylococcus aureus toxin “superantigens” with human T cells. Proc. Natl. Acad. Sci. U.S.A.86,8941-8945. Choi, Y., Herman, A., DiGiusto, D., Wade, T., Marrack, P., and Kappler, J. (1990a). Residues of the variable region of the T-cell-receptor P-chain that interact with S. aureus toxin superantigens. Nuture (London) 346,471-473. Choi, Y., Lafferty, J. A., Clements, J. R., Todd, J. K., Gelfand, E. W., Kappler, J., Marrack, P., and Kotzin, B. L. (1990b). Selective expansion of T cells expressing Vp2 in toxic shock syndrome.]. Exp. Med. 172,981-984. Choi, Y., Kappler, J. W., and Marrack, P. (1991). A superantigen encoded in the open reading frame of the 3’ long terminal repeat of mouse mammary tumor virus. Nature (London)350,203-207. Choi, Y.,Marrack, P., and Kappler, J. W. (1992). Structural analysis of a mouse mammary tumor virus superantigen. ]. E x p . Med. 175,847-852. Chothia, C., Boswell, D. R., and Lesk, A. M. (1988). The outline structure ofthe T-cell cxp receptor. EMBO]. 7,3745-3755. Chvatchko, Y., and MacDonald, H. R. (1991). CD8+ T cell response to Mls-la determi-
150
BRIAN L. KOTZIN ET AL.
nants involves major histocompatibility complex class I1 molecules. J . E x p . Med. 173, 779-782. Cleary, P. P., Kaplan, E. L., Handley, J. P., Wlazlo, A., Kim, M. H., Hauser, A. R., and Schlievert, P. M. (1992). Clonal basis for resurgence of serious Streptococcus pyogenes disease in the 1980s. Lancet 339,518-521. Coffin, J. M. (1992). Superantigens and endogenous retroviruses: A confluence of puzzles. Science 255,411-413. Cole, B. C., and Atkin, C. L. (1991). The Mycoplasma arthritidis T-cell mitogen, MAM: A model superantigen. lmmunol. Today 12,271-276. Cole, B. C., Daynes, R. A., and Ward, J. R.(198 1). Stimulation of mouse lymphocytes by a mitogen derived from Mycoplasma arthritidis. I. Transformation is associated with an H-2-linked gene that maps to the I-E/I-C subregion. J . lmmunol. 127, 1931-1936. Cole, B. C., Washburn, L. R., Sullivan, G. J., and Ward, J. R. (1982). Specificity of a mycoplasma mitogen for lymphocytes from human and various animal hosts. Infect. lmmun. 36,662-666. Cole, B. C., Washburn, L. R., and Taylor-Robinson, D. (1985). Mycoplasma-induced arthritis. In “The Mycoplasmas” ( S . Razin and M. F. Barile, eds.), Vol. IV, pp. 107160. Academic Press, New York. Cole, B. C., Araneo, B. A,, and Sullivan, G. J. (1986). Stimulation of mouse lymphocytes by a mitogen derived from Mycoplasma arthritidis. IV. Murine T hybridoma cells exhibit differential accessory cell requirements for activation by M . arthritidis T cell mitogen, concanavalin A, or hen egg-white lysozyme. J. lmmunol. 136,3572-3578. Cole, B. C., Kartchner, D. R., and Wells, D. J. r(1989).Stimulation of mouse lymphocytes by a mitogen derived from Mycoplasma arthritidis. VII. Responsiveness is associated with expression of a product(s) of the VP8 gene family present on the T cell receptor alp for antigen. J. lmmunol. 142,4131-4137. Cole, B. C., David, C. S., Lynch, D. H., and Kartchner, D. R. (1990a). The use of transfected fibroblasts and transgenic mice establishes that stimulation of T cells by the Mycoplasma arthritidis mitogen is mediated by Ea. J. Immunol. 144,420-424. Cole, B. C., Kartchner, D. R., and Wells, D. J. (1990b). Stimulation ofmouse lymphocytes by a mitogen derived from Mycoplasma arthritidis (MAM). VIII. Selective activation of T cells expressing distinct V@T cell receptors from various strains of mice by the “superantigen” MAM. J . Immunol. 144,425-431. Cole, B. C., Atkin, C. L., and Griffiths, M. M . (1993). Mycoplasma superantigens and autoimmune arthritis. J . Cell Biochem. (Suppl.) 17(B), 132 (Abstr.). Concannon, P., Pickering, L. A., Kung, P., and Hood, L. (1986). Diversity and structure of human T-cell receptor P-chain variable region genes. Proc. Natl. Acad. Sci. U.S.A.83, 6598-6602. Cone, L. A., Woodard, D. R., Schlievert, P. h4., and Tomory, G. S. (1987). Clinical and bacteriologic observations of a toxic shock-like syndrome due to Streptococcus pyogenes. N . Engl. J . Med. 317,146-149. Corley, R. B., Lund, F. E., Randall, T. D., King, L. B., Doerre, S., and Woodland, D. L. (1992). Mouse mammary tumor proviral gene expression in cells of the B lineage. Semin. lmmunol. 4,287-296. Crouse, C. A., and Pauley, R. J. (1989). Molecular cloning and sequencing of the MTV-1 LTR: Evidence for a LTR sequence alteration. Virus Res. 12, 123-137. Dale, J. B., and Beachey, E. H. (1985). Multiple, heart-cross-reactive epitopes of streptococcal M pr0teins.J. E x p . Med. 161, 113--122. Dale, J. B., Simpson, W. A., Ofek, I., and Beachey, E. H. (1981). Blastogenic responses of human lymphocytes to structurally defined polypeptide fragments of streptococcal M protein. J. Immunol. 126, 1499-1505.
SUPERANTIGENS
151
Dannecker, G., Mecheri, S., Staiano-Coico, L., and Hoffman, M. K. (1991). A characteristic Mls-la response precedes Mls-la anergy in oioo. J . Immunol. 146,2083-2090. Davis, J. P., Chesney, P. J., Wand, P. J., and Laventure, M. (1980). Toxic-shock syndrome: Epidemiologic features, recurrence, risk factors, and prevention. N . Engl. J . Med. 303,1429-1435. Davis, M. M., and Bjorkman, P. J. (1988). T-cell antigen receptor genes and T-cell recognition [published erratum appears in Nature (London) 20;335(6192):744]. Nature (London)334,395-402. de Azavedo, J . C. (1989). Animal models for toxic shock syndrome: Overview. Reo. Infect. Dis. 11 (Suppl. l), S205-S209. Dellabona, P., Peccoud, J., Benoist, C., and Mathis, D. (1989). T-cell recognition of superantigens: Inside or outside the groove? Cold Spring Harbor S y m p . Quant. Biol. 54 (Pt. l), 375-381. Dellabona, P., Peccoud, J., Kappler, J., Marrack, P., Benoist, C., and Mathis, D. (1990). Superantigens interact with MHC class I1 molecules outside of the antigen groove. Cell 62,1115-1121. Donahue, J. P., Marrack, P. C., and Kappler, J. W. (1993).A hVP3.1-specific superantigen is expressed by the human cell line Raji. Submitted. Donehower, L. A., Fleurdelys, B., and Hager, G. L. (1983). Further evidence for the protein coding potential of the mouse mammary tumor virus long terminal repeat: Nucleotide sequence of an endogenous proviral long terminal repeat. J . Virol. 45, 94 1-949. Drake, C. G., and Kotzin, B. L. (1992). Superantigens: Biology, immunology, and potential role in disease. J . Clin. Immunol. 12, 149-162. Dyson, P. J., Knight, A. M., Fairchild, S., Simpson, E., and Tomonari, K. (1991). Genes encoding ligands for deletion of V P l l T cells cosegregate with mammary tumor virus genomes [see comments]. Nature (London)349,531-532. Eynon, E. E., and Parker, D. C. (1992). Small B cells as antigen-presenting cells in the induction of tolerance to soluble protein antigens.]. Exp. Med. 175, 131-138. Fairchild, S., Knight, A. M., Dyson, P. J., and Tomonari, K. (1991). Cosegregation of a gene encoding a deletion ligand for Tcrb-V3+ T cells with Mtv-3. Immunogenetics 34, 227-230. Fasel, N., Pearson, K., Buetti, E., and Diggelmann, H. (1982). The region of mouse mammary tumor virus DNA containing the long terminal repeat includes a long coding sequence and signals for hormonally regulated transcription. EMBOJ. 1,3-7. Fast, D. J., Schlievert, P. M., and Nelson, R. D. (1989).Toxic shock syndrome-associated staphylococcal and streptococcal pyrogenic toxins are potent inducers of tumor necrosis factor production. Infect. lmmun. 57,291-294. Festenstein, H. (1973). Immunogenetic and biological aspects of in oitro lymphocyte allotransformation (MLR) in the mouse. Transplant. Reo. 15,62-88. Fischer, H., Dohlsten, M., Lindvall, M., Sjogren, H. O., and Carlsson, R. (1989). Binding of staphylococcal enterotoxin A to HLA-DR on B cell lines. J . Immunol. 142, 31513157. Fleckenstein, B., and Desrosiers, R. C. (1982). Herpesoirus saimiri and herpesviruses ateles. In “The Herpesviruses” (B. Roizman, ed.), Vol. 1, pp. 253-332. Plenum, New York. Fleischer, B. (1989). Bacterial toxins as probes for the T-cell antigen receptor. Immunol. Today 10,262-264. Fleischer, B., and Schrezenmeier, H. (1988).T cell stimulation by staphylococcal enterotoxins. Clonally variable response and requirement for major histocompatibility complex class I1 molecules on accessory or target cells. J . Exp. Med. 167, 1697-1707.
152
BRIAN L. KOTZIN ET AL.
Fleischer, B., Schrezenmeier, H., and Conradt, P. (1989).T lymphocyte activation by staphylococcal enterotoxins: Role of class I1 molecules and T cell surface structures. Cell Zmmunol. 120,92-101. Fleischer, B., Gerardy-Schahn, R.,Metzroth, B., Carrel, S., Gerlach, D., and Kohler, W. (1991).An evolutionary conserved mechanism ofT cell activation by microbial toxins. Evidence for different affinities of T cell receptor-toxin interaction. J . Zmmunol. 146, 11-17. Fleischer, B., Schmidt, K. H., Gerlach, D., and Kohler, W. (1992). Separation of Tcell-stimulating activity from streptococcal M protein. Inject. Zmmun. 60, 17671770. 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 +bearing thymocytes which predominantly expresses a single V/3 gene family. Nature (London) 329,251-254. Fowlkes, B. J., Schwartz, R. H., and Pardoll, D. M. (1988). Deletion of self-reactive thymocytes occurs at a CD4+8+precursor stage. Nature (London) 334,620-623. Frankel, W. N., Rudy, C., Coffin, J. M.,and Huber, B. T. (1991).Linkage of Mls genes to endogenous mammary tumour viruses of inbred mice [see comments]. Nature (London) 349,526-528. Fraser, J. D. (1989). High-&nity binding of staphylococcal enterotoxins A and B to HLA-DR. Nature (London)339,221-223. Fraser, J. D., Urban, R. G., Strominger, J. L., and Robinson, H. (1992).Zinc regulates the function of two superantigens. Proc. Natl. Acad. Sci. U S A . 89,5507-5511. Friedman, S. M., Crow, M. K., Tumang, J. R., Tumang, M., Xu,Y. Q., Hodtsev, A. S., Cole, B. C., and Posnett, D. N. (1991a).Characterization ofhuman T cells reactive with the Mycoplasma orthritidis-derived superantigen (MAM): Generation of a monoclonal antibody against V/317, the T cell receptor gene product expressed by a large fraction of MAM-reactive human T cells.]. Exp. Med. 174,891-900. Friedman, S. M., Posnett, D. N., Tumang, J. H.,Cole, B. C., and Crow, M. K. (1991b).A potential role for microbial superantigens in the pathogenesis of systemic autoimmune disease. Arthritis Rheum. 34,468-480. Fujiwara, H., and Hamashima, Y. (1978).Pathology of the heart in Kawasaki disease. Pediatrics 61, 100-107. Fuleihan, R.,Mourad, W.,Geha, R. S., and Chatila, T. (1991).Engagement of MHC-class I1 molecules by staphylococcal exotoxins delivers a comitogenic signal to human B cells. J . Zmmunol. 146,1661-1666. Furusho, K., Kamiya, T., Nakano, H., Kiyosawa, N., Shinomiya, K., Hayashidera, T., Tamura, T., Hirose, O., Manabe, Y.,and Yokoyama, T. (1984).High-dose intravenous gammaglobulin for Kawasaki disease. Lancet 2, 1055-1058. Galanos, C., Freudenberg, M. A., and Reutter, W. (1979).Galactosamine-induced sensitization to the lethal effects of endotoxin. Proc. Notl. Acad. Sci. U.S.A. 76,5939-5943. Gascoigne, N. R.. and Ames, K. T. (1991).Direct binding of secreted T-cell receptor B chain to superantigen associated with class I1 major histocompatibility complex protein. Proc. Notl. Acad. Sci. U S A . 88,613-616. Gaur, A,, Fathman, C. G., Steinman, L., and Brocke, S. (1993). SEB induced anergy: Modulation of immune response to T cell determinants of myoglobin and myelin basic protein. J . Zmmunol. (in press). Gaynor, E. R.,Vitek, L., Sticklin, L., Creekmore, S. P., Ferraro, M. E., Thomas, J. X., Jr., Fisher, S. G., and Fisher, R. I. (1988).The hernodynamic effects of treatment with interleukin-2 and lymphokine-activated killer cells. Ann. Intern. Med. 109, 953958.
SUPERANTIGENS
153
Gollob, K. J., and Palmer, E. (1991). Physiologic expression of two superantigens in the BDFl mouse. J. Immunol. 147,2447-2454. Gollob, K. J., and Palmer, E. (1992). Divergent viral superantigens delete Vp5' T lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 89,5138-5141. Golovkina, T. V., Chervonsky, A., Dudley, J. P., and Ross, S. R. (1992). Transgenic mouse mammary tumor virus superantigen expression prevents viral infection. Cell 69,637645. Compels, U. A., Craxton, M. A., and Honess, R. W. (1988). Conservation of gene organization in the lymphotropic herpesviruses herpesvirus Saimiri and Epstein-Barr virus. J , Virol. 62,757-767. Goshorn, S. C., and Schlievert, P. M. (1988). Nucleotide sequence of streptococcal pyrogenic exotoxin type C. Infect. Immun. 56,2518-2520. Gray, G . L., Smith, D. H., Baldridge, J. S., Harkins, R. N., Vasil, M. L., Chen, E. Y.,and Heyneker, H. L. (1984). Cloning, nucleotide sequence, and expression in Escherichiu coli of the exotoxin A structural gene of Pseudomonas aeruginosa. Proc. Natl. Acad. Sci. U.S.A.81,2645-2649. Gregersen, P. K., Silver, J.,and Winchester, R. J. (1987). The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum. 30, 1205-1213. Groux, H., Torpier, G., Monte, D., Mouton, Y., Capron, A., and Ameisen, J. C. (1992). Activation-induced death by apoptosis in CD4+ T cells from human immunodeficiency virus-infected asymptomatic individuals. /. Exp. Med. 175,331-340. Gunzburg, W. H., and Salmons, B. (1992). Factors controlling the expression of mouse mammary tumor virus. Biochem. J. 283,625-632. Hainaut, P., Castellazzi, M., Gonzales, D., Clausse, N., Hilgers, J., and Crepin, M. (1990). A congenic line of the BALB/c mouse strain with the endogenous mouse mammary tumor virus proviral gene Mtv-3: Tissue-specific expression and correlation with resistance to mouse mammary tumor virus infection and tumorigenesis. Cancer Res. 50,3754-3760. Happ, M. P., Woodland, D. L., and Palmer, E. (1989). A third T-cell receptor p-chain variable region gene encodes reactivity to Mls-la gene products. Proc. Natl. Acad. Sci. U.S.A. 86,6293-6296. Hauser, A. R., and Schlievert, P. M. (1990). Nucleotide sequence of the streptococcal pyrogenic exotoxin type B gene and relationship between the toxin and the streptococcal proteinase precursor. J . Bacteriol. 172,4536-4542. Hauser, A. R., Stevens, D. L., Kaplan, E. L., and Schlievert, P. M. (1991). Molecular analysis of pyrogenic exotoxins from Streptococcus pyogenes isolates associated with toxic shock-like syndrome. 1.Clin. Microbial. 29, 1562-1567. Held, W., Shakhov, A. N., Waanders, G., Scarpellino, L., Luethy, R., Kraehenbuhl, J. P., MacDonald, H. R., and Acha-Orbea, H.(1992).An exogenous mouse mammary tumor virus with properties of Mls-la (Mtv-7),J. E x p . Med. 175,1623-1633. Held, W., Shakhov, A. N., Izui, S., Waanders, G. A., Scarpellino, L., MacDonald, H. R., and Acha-Orbea, H. (1993). Superantigen-reactive CD4+ T cells are required to stimulate B cells after infection with mouse mammary tumor virus. J. Exp. Med. 177, 359-366. Herman, A., Croteau, G., Sekaly, R. P., Kappler, J., and Marrack, P. (1990). HLA-DR alleles differ in their ability to present staphylococcal enterotoxins to T cells. J . Exp. Med. 172,709-717. Herman, A., Kappler, J. W., Marrack, P., and Pullen, A. M. (199la). Superantigens: Mechanism of T-cell stimulation and role in immune responses. Annu. Rev. Immunol. 9,745-772.
154
BRIAN L. KOTZIN ET AL.
Herman, A., Labrecque, N., Thibodeau, J., Marrack, P., Kappler, J. W., and Sekaly, R. P. (1991b). Identification of the staphylococcal enterotoxin A superantigen binding site in the p-1 domain ofthe human histocompatibility antigen HLA-DR. Proc. Natl. Acad. Sci. U.S.A. 88,9954-9958. Hermann, T., and MacDonald, H. R. (1993). The CD8 T cell response to staphylococcal enterotoxins. Semin Immunol. 5,33-39. Herrmann, T., Accolla, R. S., and MacDonald, H. R. (1989). Different staphylococcal enterotoxins bind preferentially to distinct major histocompatibility complex class I1 isotypes. Eur.]. Immunol. 19,2171-2174. Herzog, M., Lafage, M., Montano-Hirose, J. A.. Fritzell, C., Scott-Algara, D., and Lafon, M. (1992). Nucleocapsid specific T and B cell responses in humans after rabies vaccination. Virus Res. 24,77-89. Hewitt, C . R.,Lamb, J. R., Hayball, J., Hill, M., Owen, M. J., and O’Hehir, R.E. (1992). Major histocompatibility complex independent clonal T cell anergy by direct interaction of Staphylococcus aureus enterotoxin B with the T cell antigen receptor. J . Exp. Med. 175,1493-1499. Hodes, R.J., and Abe, R. (1993). T cell recognition of Mls-like superantigens: Analysis of TCR requirements, superantigenic ligands, and signal transduction. Semin. Immunol. 4,319-327. Howell, M. D., Diveley, J. P., Esty, A., Lundeen, K. A., Winters, S. T., Carol, D. J., and Brostoff, S.W. (1991a). Clonal infiltrates of activated Vp17+ T cells in synovial tissues of rheumatoid arthritis patients. ]. Cell. Biochem. Suppl. 15A, 295. Howell, M. D., Diveley, J. P., Lundeen, K. A., Esty, A., Winters, S. T., Carlo, D. J., and Brostoff, S. W. (1991b). Limited T-cell receptor p-chain heterogeneity among interleukin 2 receptor-positive synovial T cells suggests a role for superantigen in rheumatoid arthritis. Proc. Natl. Acad. Sci. U.S.A.88, 10921-10925. Huber, B. T. (1992). Mls superantigens: How retroviruses influence the expressed T cell receptor repertoire. Semin. Immunol. 4,313-318. Hudson, K. R.,Robinson, H., and Fraser, J. D. (1993).Two adjacent residues in staphylococcal enterotoxins A and E determine T cell receptor V p specificity.]. E x p . Med. 177, 175- 184. Hugin, A. W., Vacchio, M. S., and Morse, H. C. (1991).A virus-encoded “superantigen” in a retrovirus-induced immunodeficiency syndrome of mice. Science 252, 424427. Ignatowicz, L., Kappler, J., and Marrack, P. (1992). The effects ofchronic infection with a superantigen-producing virus. J . E x p . Med. 175,917-923. Ikejima, T., Dinarello, C. A., Gill, D. M., and Wolff, S. M. (1984). Induction of human interleukin-1 by a product of Staphylococcus aureus associated with toxic shock syndrome.]. Clin. Inoest. 73,1312-1320. Ikejima, T., Okusawa, S., van der Meer, J. W., and Dinarello, C. A. (1988). Induction by toxic-shock-syndrome toxin-1 of a circulating tumor necrosis factor-like substance in rabbits and of immunoreactive tumor necrosis factor and interleukin-1 from human mononuclear cells. ]. Infect. Dis. 158,1017--1025. Imberti, L., Sottini, A., Bettinardi, A,, Puoti, M., and Primi, D. (1991). Selective depletion in HIV infection of T cells that bear specific T cell receptor Vp sequences. Science 254,860-862. Irwin, M. J., Hudson, K. R., Fraser, J. D., and Gascoigne, N. R. (1992). Enterotoxin residues determining T-cell receptor V p binding specificity. Nature (London) 359, 84 1-843. Janeway, C. A., Jr. (1990). Self superantigens? Cell 63,659-661.
SUPERANTICENS
155
Janeway, C. A., Jr. (1991). Selective elements for the Vp region ofthe T cell receptor: MIS and the bacterial toxic mitogens. Ado. lmmunol. 50, 1-53. Janeway, C. A., Jr. and Katz, M. E. (1985).The immunobiology ofthe T cell response to Mls-locus-disparate stimulator cells. I. Unidirectionality, new strain combinations, and the role of Ia antigens. J. Zntmunol. 134,2057-2063. Janeway, C. A., Jr., Yagi, J., Conrad, P. J., Katz, M. E., Jones, B., Vroegop, S., and Buxser, S . (1989). T-cell responses to MIS and to bacterial proteins that mimic its behavior. Immunol. Reo. 107,61-88. Jenkins, M. K., Pardoll, D. M., Miznguchi, J.,Quill, H., and Schwartz, R. H. (1987).T-cell unresponsiveness in oioo and in oitro: Fine specificity of induction and molecular characterization of the unresponsive state. lmmunol. Reo. 95, 113- 135. Jones, L. A., Chin, L. T., Longo, D. L., and Kruisbeek, A. M. (1990). Peripheral clonal elimination of functional T cells. Science 250, 1726-1729. Jouvin-Marche, E., Trede, N. S., Bandeira, A,, Tomas, A., Loh, D. Y.,and Cazenave, P. A. (1989). Different large deletions of T cell receptor Vp genes in natural populations of mice. Eur. J. lmmunol. 19, 1921-1926. Jouvin-Marche, E., Cazenave, P. A., Voegtle, D., and Marche, P. N. (1992). Vp17 T-cell deletion by endogenous mammary tumor virus in wild-type-derived mouse strain. Proc. Nutl. Acud. Sci. U.S.A.89,3232-3235. Jouvin-Marche, E., Marche, P. N., and Cazenave, P. A. (1993). Clonal deletion ofVpl7 T cells in mice from natural populations. Semin. lminunol. 4,305-312. Jupin, C., Anderson, S., Damais, C., Alouf, J. E., and Parant, M. (1988). Toxic shock syndrome toxin 1as an inducer of human tumor necrosis factors and gamma interferon. J. E x p . Med. 167,752-761. Kain, S . H.,Platt, E. J., Brown, K. S., Black, N., and Firestone, G. L. (1992). Disruptions in intracellular membrane trafficking and structure preclude the glucocorticoiddependent maturation of mouse mammary tumor virus proteins in rat hepatoma cells. J. B i d . Chein. 267, 8128-8135. Kappler, J. W., Roehm, N., and Marrack, P. (1987a).Tcell tolerance by clonal elimination in the thymus. Cell 49,273-280. Kappler, J. W., Wade, T., White, J., Kushnir, E., Blackman, M., Bill, J., Roehm, N., and Marrack, P. (1987b). A T cell receptor Vp segment that imparts reactivity to a class I1 major histocompatibility complex product. Cell 49,263-271. Kappler, J. W., Staerz, U., White, J., and Marrack, P. C. (1988). Self-tolerance eliminates T cells specific for Mls-modified products of the major histocompatibility complex. Nature (London)332,35-40. Kappler, J. W., Kotzin, B., Herron, L., Gelfand, E. W., Bigler, R. D., Boylston, A., Carrel, S., Posnett, D. N., Choi, Y.,and Marrack, P. (1989a). Vp-specific stimulation ofhunian T cells by staphylococcal toxins. Science 244,811-813. Kappler, J. W., Kushnir, E., and Marrack, P. (1989b). Analysis ofVpl7a expression in new mouse strains bearing the Vpa haplotype.]. E x p . Med. 169, 1533-1541. Kappler, J. W., Herman, A., Clements, J.. and Marrack, P. (1992). Mutations defining functional regions of the superantigen staphylococcal enterotoxin B.J. E x p . Med. 175, 387-396. Karp, D. R., and Long, E. 0. (1992). Identification of HLA-DRl p chain residues critical for binding staphylococcal enterotoxins A and E.J. E x p . Med. 175,415-424. Karp, D. R., Teletski, C. L., Scholl, P., Geha, R., and Long, E. 0.(1990).The a1 domain of the HLA-DR molecule is essential for high-affinity binding of the toxic shock syndrome toxin-1. Nature (London)346,474-476. Kato, H., Ichinose, E., Yoshioka, F., Takechi, T., Matsunaga, S., Suzuki, K., and Rikitake,
156
BRIAN L. KOTZIN ET AL.
N. (1982). Fate of coronary aneurysms in Kswasaki disease: Serial coronary angiography and long-term follow-up study. Am. J . Cardid. 49, 1758-1766. Katz, I. R.,Lerman, S. P., Ponzio, N. M., ShrefAer, D. C., and Thorbecke, G . J. (1980). Growth of SJLIJ-derived transplantable reticulum cell sarcoma as related to its ability to induce T-cell proliferation in the host, I. Dominant negative genetic influences of other parent haplotype in F1 hybrids of SJL/J mice. J . E x p . Med. 151,347-361. Katz, J. D., Ohnishi, K., Lebow, L. T., and Bonavida, B. (1988).The SJLlJ Tcell response to both spontaneous and transplantable syngeneic reticulum cell sarcoma is mediated predominantly by the Vp17a+ T cell clonotype. J . E x p . Med. 168,1553-1562. Katz, J. D., Lebow, L. T., and Bonavida, B. (1989). The in uioo depletion of Vp17af T cells results in the inhibition of reticulum cell sarcoma growth in SJL/J mice. Evidence for the use of anticlonotypic antibody therapy in the control of malignancy. J . Immunol. 143, 1387-1395. Kawabe, Y., and Ochi, A. (1990). Selective allergy of Vp8' T cells in staphylococcus enterotoxin B-primed mice. J . E x p . Med. 172, 1065-1070. Kawabe, Y., and Ochi, A. (1991). Programmed cell death and extrathymic reduction of Vp8' CD4+ T cells in mice tolerant to Staphylococcus aureus enterotoxin B [see comments]. Nature (London)349,245-248. Kawasaki, T. (1967). [Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children]. Arerugi 16, 178-222. Kim, C., Siminovitch, K. A., and Ochi, A. (1991). Reduction of liipus nephritis in MRL/lpr mice by a bacterial superantigen treatment. J . E x p . Med. 174,1431-1437. Kimura, N., Toyonaga, B., Yoshikai, Y., Du, R.P., and Mak, T. W. (1987). Sequences and repertoire of the human T cell receptor (I and @ chain variable region genes in thymocytes. Eur. J . Immunol. 17,375-383. King, L. B., Lund, F. E., White, D. A., Sharma, S., and Corley, R. B. (1990). Molecular events in B lymphocyte differentiation. Inducible expression of the endogenous mouse mammary tumor proviral gene, Mtv-9. J. Immunol. 144,3218-3227. Knight, A. M., Harrison, G. B., Pease, R. J., Robinson, P. J., and Dyson, P. J. (1992). Biochemical analysis of the mouse mammary tumor virus long terminal repeat product. Evidence for the molecular structure of an endogenous superantigen. Eur. J . Irnmunol. 22,879-882. Korman, A. J., Bourgarel, P., Meo, T., and Rieckhof, G . E. (1992). The mouse mammary tumour virus long terminal repeat encodes a type I1 transmembrane glycoprotein. EMBO J . 11,1901-1905. Kotb, M. (1992). Role of superantigens in infectious diseases and their sequela. Curr. Opin. Infect. Dis. 5,364-374. Kotb, M., Courtney, H. S., Dale, J. B., and Beachey, E. H. (1989).Cellular and biochemical responses of human T lymphocytes stimulated with streptococcal M proteins. J . Immunol. 142,966-970. Kotzin, B. L., Babcock, S. K., and Herron, I,. R. (1988). Deletion of potentially selfreactive T cell receptor specificities in L3'1'4-, Lyt-2- T cells of Ipr mice [published erratum appears J E x p Med 169(4):1515-15161. J . E x p . Med. 168,2221-2229. Kotzin, B. L., Herron, L. R.,Babcock, S. K., Portanova, J. P., and Palmer, E. (1989a). Self-reactive T cells in murine lupus: Analysis of genetic contributions and development of self-tolerance. Clin. Immunol. Immunopathol. 53, S35-S46. Kotzin, B. L., Kappler, J. W., Marrack, P. C., arid Herron, L. R. (1989b). T cell tolerance to self antigens in New Zealand hybrid mice with lupus-like disease. 1. Immunol. 143, 89-94.
SUPERANTIGENS
157
Kotzin, B. L., Ricalton, N., Leung, D. Y., Kappler, J. W., and Marrack, P. (1993). Characterization of T cell alterations in toxic shock syndrome. Manuscript in preparation. Kozak, C., Peters, G . , Pauley, R., Morris, V., Michalides, R., Dudley, J., Green, M., Davisson, M., Prakash, 0.. and Vaidya, A. (1987). A standardized nomenclature for endogenous mouse mammary tumor viruses. J . Virol. 61,1651-1654. Lafon, M., Lafage, M., Martinez-Arends, A., Ramirez, R., Vuillier, F., Charron, D., Lotteau, V., and Scott-Algara, D. (1992). Evidence for a viral superantigen in humans. Nature (London)358,507-510. Langford, M. P., Stanton, G. J., and Johnson, H. M. (1978). Biological effects ofstaphylococcal enterotoxin A on human peripheral lymphocytes. Infect. Immun. 22,62-68. Larsson-Sciard, E. L., Spetz-Hagberg, A. L., Casrouge, A., and Kourilsky, P. (1990). Analysis of T cell receptor Vp gene usage in primary mixed lymphocyte reactions: Evidence for directive usage by different antigen-presenting cells and Mls-like determinants on T cell blasts. Eur.J. Immunol. 20, 1223-1229. Lasky, J. L., Ponzio, N. M., and Thorbecke, G . J. (1988). Characterization and growth factor requirements of SJL lymphomas. I. Development of a B cell growth factordependent in oitro cell line, cRCS-X [published erratum appears in J . Immunol. 140(7):2478].J . Immunol. 140,679-687. Laurence, J., Hodtsev, A. S., and Posnett, D. N. (1992). Superantigen implicated in dependence of' HIV-1 replication in T cells on TCR V beta expression. Nuture (London) 358,255-259. Lee, B. K., and Eicher, E. M. (1990). Segregation patterns of endogenous mouse mammary tumor viruses in five recombinant inbred strain sets [published erratum appears in J Virol65(3):16661.J.Virol. 64,4568-4572. Legaard, P. K., LeGrand, R. D., and Misfeldt, M. L. (1991). The superantigen Pseudonionas exotoxin A requires additional functions from accessory cells for T lymphocyte proliferation. Cell Immunol. 135,372-382. Lehniann, V., Freudenberg, M. A., and Galanos, C. (1987).Lethal toxicity of lipopolysaccharide and tumor necrosis factor in normal and D-galactosamine-treated mice. J . Exp. Med. 165,657-663. Lenardo, M. J. (1991). Interleukin-2 programs mouse ap T lymphocytes for apoptosis. Nuture (London)353,858-861. Leonard, B. A., Lee, P. K., Jenkins, M. K., and Schlievert, P. M. (1991).Cell and receptor requirements for streptococcal pyrogenic exotoxin T-cell mitogenicity. Infect. Im? I I U ~59, ~ . 1210-1214. Lerman, S. P., Carswell, E. A., Chapman, J., and Thorbecke, G. J. (1976). Properties of reticulum cell sarcomas in SJL/J mice. 111. Promotion of tumor growth in irradiated mice by normal lymphoid cells. Cell Immunol. 23,53-67. Leung, D. Y., Cotran, R. S., Kurt-Jones, E., Burns, J. C., Newburger, J. W., and Pober, J. S. (1989). Endothelial cell activation and high interleukin-1 secretion in the pathogenesis of acute Kawasaki disease [see comments]. Lancet 2, 1298-1302. Leung, D. Y. (1990). In~n~unologic aspects of Kawasaki syndrome.]. Rheumutol. S u p p l . 24, 15-18. Leung, D. Y. M. (1993). Kawasaki disease. CUM.Opinion Rheumatol. 5,41-50. Leung, D. Y. M., Meissner, H. C., Fulton, D. R., Kotzin, B. L., and Schlievert, D. M. ( 1993). Identification of a novel strain of toxic shock syndrome toxin (TSST)-producing Staphylococcus uureus in Kawasaki syndrome. Clin. Res. (in press). Li, Y., Szabo, P., Robinson, M. A., Dong, B., and Posnett, D. N. (1990).Allelic variations in the human T cell receptor Vp6.7 gene products [published erratum appears in J. Exp. Med. 171(3):973].J.E x p . Med. 171,221-230.
158
BRIAN L. KOTZIN ET AL.
Liu, H., Lampe, M. A., Iregui, M. V., and Cantor, H. (1991). Conventional antigen and superantigen may be coupled to distinct and cooperative T-cell activation pathways. Proc. Natl. Acad. Sci. U.S.A. 88,8705-8709. Lund, F. E., and Corley, R. B. (1991). Regulated expression of mouse mammary tumor proviral genes in cells ofthe B 1ineage.J. Exp. Med. 174,1439-1450. MacDonald, H. R.,Hengartner, H., and Pedrazzini, T. (1988a). Intrathymic deletion of self-reactive cells prevented by neonatal anti-CD4 antibody treatment. Nature (London) 335,174-176. MacDonald, H. R.,Schneider, R.,Lees, R. K., Howe, R.C., Acha-Orbea, H., Festenstein, H., Zinkernagel, R. M., and Hengartner, €3. (1988b). T-cell receptor VP use predicts reactivity and tolerance to Mlsa-encoded antigens. Nature (London) 332, 4045. MacDonald, H. R.,Lees, R.K., and Chvatchko, Y.(1990). CD8+ T cells respond clonally to Mls-la-encoded determinants.J. Exp. Med. 171,1381-1386. MacDonald, H. R.,Baschieri, S., and Lees, R. K. (1991). Clonal expansion precedes anergy and death of Vp8' peripheral T cells responding to staphylococcal enterotoxin B in oioo. Eur. J . Immunol. 21, 1963-1966. Marrack, P., and Kappler, J. ( 1990). The staphylococcal enterotoxins and their relatives. Science 248,1066. Marrack, P., Blackman, M., Kushnir, E., and Kappler, J. (1990). The toxicity of staphylococcal enterotoxin B in mice is mediated by T cel1s.J. E x p . Med. 171,455-464. Marrack, P., Kushnir, E., and Kappler, J. (1991). A maternally inherited superantigen encoded by a mammary tumour virus [see cornments]. Nature (London)349,524-526. Marrack, P., Winslow, G., Choi, Y.,Scherer, hi., Pullen, A., White, P., and Kappler, J. (1993). The bacterial and mouse mammary tumor virus superantigens; two different families of proteins with the same functions. Zmmunol. Reo. 131,79-103. Matthes, M., Schrezenmeier, H., Homfeld, J., Fleischer, S., Malissen, B., Kirchner, H., and Fleischer, B. (1988). Clonal analysis of human T cell activation by the Mycoplasma arthritidis mitogen (MAS). Eur. J. Inimunol. 18, 1733-1737. McCormick, J. E., Callahan, J. E., Kappler, J., and Marrack, P. (1993). Profound deletion of mature T cells in oioo by chronic exposure to exogenous superantigen.J. Zmmunol. (in press). Melish, M. E. (1991). Clinical and epidemiological aspects of Kawasaki disease. Clin. Cardiol. 14(suppl. 11),3-10. Melish, M. E., and Hicks, R. V. (1990). Kawa:iaki syndrome: Clinical features. Pathophysiology, etiology and therapy. J . Rheumatol. Suppl. 24,2-10. Miethke, T., Wahl, C., Heeg, K., Echtenacher, B., Krammer, P. H., and Wagner, H. (1992). T cell-mediated lethal shock triggered in mice by the superantigen staphylococcal enterotoxin B: Critical role of tumor necrosis factor. J . E x p . Med. 175, 9198. Misfeldt, M. L. (1990). Microbial superantigens. Infect. Immun. 58,2409-2413. Mohan, N., Mottershead, D., Subramanyam, M.,Beutner, U., and Huber, B. T. (1993). Production and characterization of an MIS-1-specific monoclonal antibody. J . E x p . Med. 177,351-358. Molina, I. J., Cannon, N. A., Hyman, R., and Huber, B. T. (1989). Macrophages and T cells do not express Mlsa determinants.J. Zmmunol. 143,39-44. Mollick, J. A., Cook, R.G., and Rich, R. R.(1989). Class I1 MHC molecules are specific receptors for staphylococcus enterotoxin A. Science 244,817-820. Mollick, J. A., Chintagumpala, M., Cook, R. G., and Rich, R. R. (1991). Staphylococcal exotoxin activation ofT cells. Role of exotoxin-MHC class I1 binding affinity and class I1 isotype. J . Immunol. 146,463-468.
SUPERANTIGENS
159
Mollick, J. A., McMasters, R. L., Grossman, D., and Rich, R. R. (1993).Localization of a site on bacterial superantigens that determines T cell receptor B chain specificity. J . E x p . Med. 177,283-293. Morens, D. M., Anderson, L. J., and Hunvitz, E. S . (1980). National surveillance of Kawasaki disease. Pediatrics 65,21-25. Mosier, D. E., Gulizia, R. J., Baird, S. M., and Wilson, D. B. (1988). Transfer of a functional human immune system to mice with severe combined immunodeficiency. Nature (London)335,256-259. Mourad, W., Scholl, P., Diaz, A,, Geha, R., and Chatila, T. (1989). The staphylococcal toxic shock syndrome toxin 1 triggers B cell proliferation and differentiation via major histocompatibility complex-unrestricted cognate TIB cell interaction. J . E x p . Med. 170,2011-2022. Mourad, W., Geha, R. S., and Chatila, T. (1990).Engagement of major histocompatibility complex class I1 molecules induces sustained, lymphocyte function-associated molecule 1-dependent cell adhesion. J . E x p . Med. 172, 1513-1516. Mourad, W., Mehindate, K., Schall, T. J., and McColl, S. R. (1992). Engagement ofmajor histocompatibility complex class I1 molecules by superantigen induces inflammatory cytokine gene expression in human rheumatoid fibroblast-like synoviocytes. J. Exp. Med. 175,613-616. Musser, J . M., Hauser, A. R., Kim, M. H., Schlievert, P. M., Nelson, K., and Selander, R. K. (1991). Streptococcus p y o g e n e s causing toxic-shock-like syndrome and other invasive diseases: Clonal diversity and pyrogenic exotoxin expression. Proc. Natl. Acud. Sci. U.S.A.88,2668-2672. Nagashima, M., Matsushima, M., Matsuoka, H., Ogawa, A., and Okumura, N. (1987). High-dose gammaglobulin therapy for Kawasaki disease. J . Pediatr. 110,710-712. Nakauchi, H., Osada, H., Yagita, H., and Okumura, K. (1987). Molecular evidence that SJL reticulum cell sarcomas are derived from pre-B cell. Clonal rearrangement of heavy chain but not of light chain immunoglobulin genes. J . Immunol. 139,28032809. Newburger, J . W., Takahashi, M., Bums, J . C., Beiser, A. S., Chung, K. J., Duffy, C. E., Glode, M. P., Mason, W. H., Reddy, V., and Sanders, S. P. (1986). The treatment of Kawasaki syndrome with intravenous gamma globulin. N . Engl. J . Med. 315, 341347. Newburger, J. W., Takahashi, M., Beiser, A. S., Burns, J. C., Bastian, J., Chung, K. J., Colan, S. D., Duffy, C. E., Fulton, D. R., and Glode, M. P. (1991).A single intravenous infusion of gamma globulin as compared with four infusions in the treatment of acute Kawasaki syndrome [see comments]. N . Eng1.J. Med. 324,1633-1639. Newell, M. K., Haughn, L. J., Maroun, C. R., and Julius, M. H. (1990). Death of mature T cells by separate ligation ofCD4 and the T-cell receptor for antigen. Nature (London) 347,286-289. Nicholas, J., Smith, E. P., Coles, L., and Honess, R. (1990). Gene expression in cells infected with gammaherpesvirus saimiri: Properties of transcripts from two immediate-early genes. Virology 179, 189-200. Ochi, A., Yuh, K., and Migita, K. (1993). Not every superantigen induces tolerance in vivo. Semin Immunol. 5,57-63. Ohnishi, K., and Bonavida, B. (1987). Regulation of Ia+ reticulum cell sarcoma (RCS) growth in syngeneic SJL/Jmice. I. Inhibition of tumor growth by passive administration of L3T4 monoclonal antibody before or after tumor inocu1ation.J. Immunol. 138, 4524-4529. Okada, C. Y., and Weissman, I. L. (1989). Relative V beta transcript levels in thymus and peripheral lymphoid tissues from various mouse strains. Inverse correlation of I-E and
160
BRIAN L. KOTZIN ET AL.
Mls expression with relative abundance of several Vp transcripts in peripheral lymphoid tissues.]. E x p . Med. 169, 1703-1719. Okada, C. Y., Holzmann, B., Guidos, C., Palmer, E., and Weissman, 1. L. (1990). Characterization of a rat monoclonal antibody specific for a determinant encoded by the Vp gene segment. Depletion ofVp7+ T cells on mice with Mls-la haplotype.]. lmmunol. 144,3473-3483. O’Rourke, A. M.,Mescher, M., and Webb, S. R. (1990). Activation of polyphosphoinositide hydrolysis in T cells by H-2 alloantigen but not Mls determinants. Science 249, 171- 174. O’Sullivan, D., Arrhenius, T., Sidney, J., Del (hercio, M. F., Albertson, M., Wall, M., Oseroff, C., Southwood, S., Colon, S. M., and Gaeta, F. C. (1991). On the interaction of promiscuous antigenic peptides with different DR alleles. Identification of common structural motifs. J . Immunol. 147,2663-2663. Otten, A. D., Sanders, M. M., and McKnight, G. S. (1988). The MMTV LTR promoter is induced by progesterone and dihydrotestosttxone but not by estrogen. Mol. Endocrinol. 2,143-147. Oyaizu, N., Chirmule, N., Yagura, H., Pahwa, R., Good, R. A., and Pahwa, S. (1992). Superantigen staphylococcal enterotoxin-B induced T-helper cell activation is independent of CD4 molecules and phosphatidylinositol hydrolysis. Proc. Natl. Acad. Sci. U.S.A. 89,8035-8039. Paliard, X., West, S. G., Lafferty, J. A., Clements, J. R., Kappler, J. W., Marrack, P., and Kotzin, B. L. (1992). Evidence for the effects of a superantigen in rheumatoid arthritis. Science 253,325-329. Palmer, E. (1991). Infectious origins of superantigens. Curr. Biol. 1,74-76. Panina-Bordignon, P., Tan, A., Termijtelen, A., Demotz, S., Corradin, G., and Lanzavecchia, A. (1989). Universally immunogenic 1’ cell epitopes: Promiscuous binding to human MHC class I1 and promiscuous recognition by T cells. Eus. J. lmmunol. 19, 2237-2242. Panina-Bordignon, P., Fu, X. T., Lanzavecchia, A,, and Karr, R. W. (1992). Identification of HLA-DRa chain residues critical for binding of the toxic shock syndrome toxin superantigen.]. Exp. Med. 176,1779-1784. Parker, B. D., Bankier, A., Satchwell, S., Barrell, B., and Farrell, P. J. (1990). Sequence and transcription of Raji Epstein-Barr virus DNA spanning the B95-8 deletion region. Virology 179,339-346. Parsonnet, J. (1989). Mediators in the pathogenesis of toxic shock syndrome: Overview. Reo. Infect. Dis.11 (Suppl. l), S263-S269. Parsonnet, J., and Gillis, Z. A. (1988). Production of tumor necrosis factor by human monocytes in response to toxic-shock syndrome toxin-1. J. Infect. Dis. 158, 10261033. Parsonnet, J., Hickman, R. K., Eardley, D. D., and Pier, G. B. (1985). Induction of human interleukin-1 by toxic-shock-syndrome toxin-1. ]. Infect. Dis.151,514-522. Patarca, R., Wei, F.-Y., Iregui, M. V., and Cantor, H. (1991). Differential induction of interferon y gene expression after activation of CD4+ T cells by conventional antigen and Mls superantigen. Proc. Natl. Acud. Sci. U.S.A.88,2736-2739. Peavy, D. L., Adler, W. H., and Smith, R. T. (1970). The mitogenic effects of endotoxin and staphylococcal enterotoxin B on mouse spleen cells and human peripheral lymphocytes.]. lmmunol. 105,1453-1458. Peck, A. B., Janeway, C. A., Jr., and Wigzell, H. (1977). T lymphocyte responses to Mls locus antigens involve recognition of H-2 I region gene products. Nuture (London) 266,840-842.
SUPERANTIGENS
16 1
Peters, G., Placzek, M., Brookes, S., Kozak, C., Smith, R., and Dickson, C. (1986). Characterization, chromosome assignment, and segregation analysis of endogenous proviral units of' mouse mammary tumor virus. J . Virol. 59,535-544. Pober, J. S. (1988). Warner-Lambert/Parke-Davis award lecture. Cytokine-mediated activation of vascular endothelium. Physiology and pathology. Am. J . Pathol. 133, 426-433. Pontzer, C. H., Russell, J. K., and Johnson, H. M. (1989). Localization of an immune functional site on staphylococcal enterotoxin A using the synthetic peptide approach. J . Immunol. 143,280-284. Pontzer, C. H., Russell, J. K., and Johnson, H. M. (1991). Structural basis for differential binding of staphylococcal enterotoxin A and toxic shock syndrome toxin 1 to class I1 major histocompatibility molecules. Proc. Natl. Acud. Sci. U.S.A.88, 125-128. Ponzio, N. M., David, C. S., Shreffler, D. C., and Thorbecke, G. J. (1977). Properties of reticulum cell sarcomas in SJL/J mice. V. Nature of reticulum cell sarcoma surface antigen which induces proliferation of normal SJL/J T cells.]. Exp. Med. 146, 13214.5. Portanova, J. P., and Kotzin, B. L. (1988). Lupus-like autoimmunity in murine graftversus-host disease. Concepts. Immunopathology 6,119-140. Posnett, D. N., Wang, C. Y., and Friedman, S. M. (1986). Inherited polymorphism of the human T-cell antigen receptor detected by a monoclonal antibody. Proc. Natl. Acad. Sci. U.S.A83,7888-7892. Posnett, D. N., Kabak, S., Hodtsev, A. S., Goldberg, E. A., and Asch, A. (1993). TCR-VP subsets are not preferentially deleted in AIDS. AlDS (in press). Pullen, A. M., Marrack, P., and Kappler, J. W. (1988). The T-cell repertoire is heavily influenced by tolerance to polymorphic self-antigens. Nature (London) 335, 796802. Pullen, A. M., Marrack, P., and Kappler, J. W. (1989). Evidence that MIS-2 antigens which delete Vp3+ T cells are controlled by multiple genes. J. lmmunol. 142,30333037. Pullen, A. M., Potts, W., Wakeland, E. K., Kappler, J., and Marrack, P. (1990a). Surprisingly uneven distribution of the T cell receptor Vp repertoire in wild mice. J. Exp. Med. 171,49-62. Pullen, A. M., Wade, T., Marrack, P., and Kappler, J. W. (1990b). Identification of the region of T cell receptor p chain that interacts with the self-superantigen Mls-la. Cell 61, 1365-1374. Pullen, A. M., Bill, J., Kubo, R. T., Marrack, P., and Kappler, J. W. (1991). Analysis of the interaction site for the self superantigen Mls-la on T cell receptor VP.1. Exp. Med. 173, 1183-1 192. Pullen, A. M., Choi, Y., Kushnir, E., Kappler, J., and Marrack, P. (1992). The open reading frames in the 3' long terminal repeats of several mouse mammary tumor virus integrants encode VP3-specific superantigens. J. Exp. Med. 175,41-47. Purtilo, D. T., Falk, K., Pirruccello, S. J., Nakamine, H., Kleveland, K., Davis, J. R., Okano, M., Taguchi, Y.,Sanger, W. G., and Beisel, K. W. (1991). SCID mouse model of Epstein-Barr virus-induced lymphomagenesis of immunodeficient humans. Int. J . Cuncer 47,510-517. Ranimensee, H. G., Kroschewski, R., and Frangoulis, 8. (1989). Clonal anergy induced in mature VP6' T lymphocytes on immunizing Mls-lb mice with Mls-la expressing cells. Nature (London)339,541-544. Rauch, A. M. and Hurwitz, E. S. (1985). Centers for Disease Control (CDC) case definition for Kawasaki syndrome [letter]. Pediatr. Infect. Dis. 4,702-703.
162
BRIAN L. KOTZIN ET AL.
Reck, B., Scheuber, P. H., Londong, W., Sailer-Kramer, B., Bartsch, K., and Hammer, D. K. (1988). Protection against the staphylococcal enterotoxin-induced intestinal disorder in the monkey by anti-idiotypic antibodies. Proc. Natl. Acad. Sci. U.S.A.85, 3170-3 174. Rellahan, B. L., Jones, L. A., Kruisbeek, A. M., Fry, A. M., and Matis, L. A. (1990). In vivo induction of anergy in peripheral Vp8' T cells by staphylococcal enterotoxin B.J. E x p . Med. 172,1091-1100. Robinson, J. H., and Kehoe, M. A. (1992). Group A streptococcal M proteins: Virulence factors and protective antigens. Zmmunol. Today 13,362-367. Robinson, M. A. (1991). The human T cell receptor p-chain gene complex contains at least 57 variable gene segments. Identification of six Vp genes in four new gene families.]. Zmmunol. 146,4392-4397. Rocken, M., Urban, J. F., and Shevach, E. M. (1992). Infection breaks T-cell tolerance. Nature (London)359,79-82. Rowe, M., Young, L. S., Crocker, J., Stokes, H., Henderson, S., and Rickinson, A. B. (1991). Epstein-Barr virus (EBV)-associated lymphoproliferative disease in the SCID mouse model: Implications for the pathogenesis of EBV-positive lymphomas in man. J. Exp. Med. 173, 147-158. Rowley, A. H., Duffy, C. E., and Shulman, S. T. (1988). Prevention of giant coronary artery aneurysms in Kawasaki disease by intravenous gamma globulin therapy [see comments]. J . Pediatr. 113,290-294. Rudy, C. K., Kraus, E., Palmer, E., and Huber, B. T. (1992). Mls-I-like superantigen in the MA/MyJ mouse is encoded by a new mammary tumor provirus that is distinct from Mtv-7.1. E x p . Med. 175,1613-1621. Russell, J. K., Pontzer, C. H., and Johnson, H. M. (1990). The I-A p b region (65-85) is a binding site for the superantigen, staphylococcal enterotoxin A. Biochem. Biophys. Res. Commun. 168,696-701. Russell, J. K., Pontzer, C. H., and Johnson, H. h4. (1991). Both a-helices along the major histocompatibility complex binding cleft are required for staphylococcal enterotoxin A function. Proc. Nutl. Acad. Sci. U.S.A.88,7228-7232. Sakano, T., and Bonavida, B. (1986). Inhibition of IA positive reticulum cell sarcoma tumor cell growth in syngeneic SJL/J mice hy passive administration of monoclonal anti-IA antibody. Cancer Res. 46,3245-3248. Scheuber, P. H., Golecki, J. R., Kickhofen, B., Scheel, D., Beck, G., and Hammer, D. K. (1985). Skin reactivity of unsensitized monkey upon challenge with staphylococcal enterotoxin B. Infect. Immun. 50,869-876. Schiffenbauer,J., Soos, J., Johnson, H., Butfiloski, E. J., and Wegrzyn, L. (1993). Staphylococcal enterotoxin can re-activate experimental allergic encephalomyelitis. Proc. Natl. Acad. Sci. U S A . (in press). Schlievert, P. M., Shands, K. N., Dan, B. B., Schmid, G. P., and Nishimura, R. D. (1981). Identification and characterization of an exotoxin from Staphylococcus aureus associated with toxic-shock syndrome. J . Infect. Dis. 143,509-516. Scholl, P., Diez, A., Mourad, W., Parsonnet, J., Geha, R. S., and Chatila, T. (1989a). Toxic shock syndrome toxin 1binds to major histocompatibility complex class 11 molecules [published erratum appears in Proc. Natl. Acad. Sci. U.S.A.86(18):7138]. Proc. Natl. Acud. Sci. U 3 . A 86,4210-4214. Scholl, P. R., Diez, A., and Geha, R. S. (1989b). Staphylococcal enterotoxin B and toxic shock syndrome toxin-1 bind to distinct sites on HLA-DR and HLA-DQ molecules. J. Immunol. 143,2583-2588.
SUPERANTIGENS
163
Scholl, P. R., Diez, A., Karr, R., Sekaly, R. P., Trowsdale, J., and Geha, R. S. (1990a). Efrect of isotypes and allelic polymorphism on the binding of staphylococcal exotoxins to MHC class I1 molecules.]. lmmunol. 144,226-230. Scholl, P. R., Sekaly, R. P., Diez, A,, Glimcher, L. H., and Geha, R. S . (1990b). Binding of toxic shock syndrome toxin-1 to murine major histocompatibility complex class 11 molecules. Eur. ]. lmmunol. 20, 1911-1916. Schwab, J. H., Brown, R. R., Anderle, S . K., and Schlievert, P. M. (1993). Superantigen can reactivate bacterial cell wall-induced arthritis. J. lmmunol. (in press). See, R. H., Krystal, G., and Chow, A. W. (1990). Binding competition of toxic shock syndrome toxin L and other staphylococcal exoproteins for receptors on human peripheral blood mononuclear cells. Infect. Immun. 58,2392-2399. See, R. H., Kum, W. W., Chang, A. H., Goh, S. H., and Chow, A. W. (1992). Induction of tumor necrosis factor and interleukin-1 by purified staphylococcal toxic shock syndrome toxin 1 requires the presence of both monocytes and T lymphocytes. Infect. lmmun. 60,2612-2618. Shands, K. N., Schmid, G. P., Dan, B. B., Blum, D., Guidotti, R. J., Hargrett, N. T., Anderson, R. L., Hill, D. L., Broome, C. V., Band, J. D., and Fraser, D. W. (1980). Toxic-shock syndrome in menstruating women: Association with tampon use and Stciphylococcus aureus and clinical features in 52 cases. N . Engl. J. Med. 303, 14361442. Shortman, K., Wilson, A,, Pearse, M.,Gallagher, P., and Scollay, R. (1988). Mouse strain differences in subset distribution and T cell antigen receptor expression among CD4-CD8- thymocytes. lmmunol. Cell Biol. 66,423-433. Singer, P. A., Balderas, R. S., and Theofilopoulos, A. N. (1990). Thymic selection defines multiple T cell receptor VP “repertoire phenotypes” at the CD4/CD8 subset level. E M B O J . 9,3641-3648. Smith, H. P., Le, P., Woodland, D. L., and Blackman, M. A. (1992). T cell receptor alpha-chain influences reactivity to Mls-1 in VP8.1 transgenic mice. J . lmmunol. 149, 887-896. Smith, T. J., Terada, N. W., Robinson, C. C., and Gelfand, E. W. (1993). Acute infectious niononucleosis stimulates the expansion ofVP6.1-6.3 and VP7 T cells. Blood (in press). Stevens, D. L., Tanner, M. H, Winship, J., Swarts, R., Ries, K. M., Schlievert, P. M.,and Kaplan, E. (1989). Severe group A streptococcal infections associated wtih a toxic shock-like syndrome and scarlet fever toxin A [see comments]. N . Eng1.J. Med. 321, 1-7. Stollernian, G. H. (1990). Rheumatogenic group A streptococci and the return of rheumatic fever. Adu. Intern. Med. 35, 1-25. Straws, S . E., Cohen, J. I., Tosato, G., and Meier, J. (1993). Epstein-Barr virus infections: Biology, pathogenesis, and management. Ann. Intern. Med. 118,45-58. Swaminathan, S . , Furey, W., Pletcher, J., and Sax, M. (1992). Crystal structure of staphylococcal enterotoxin B, a superantigen. Nature (London)359,801-806. Takahama, Y., Kosugi, A., and Singer, A. (1991). Phenotype, ontogeny, and repertoire of CD4-CD8- T cell receptor aP + thymocytes. Variable influence of self-antigens on T cell receptor VP usage.]. lmmunol. 146,1134-1141. Terai, C., Kornbluth, R. S., Pauza, C. D., Richman, D. D., and Carson, D. A. (1991). Apoptosis as a mechanism of cell death in cultured T lymphoblasts acutely infected with HIV-1.1. C l h . lnuest. 87, 1710-1715. Thonison, B. J., and Nicholas, J. (1991). Superantigen function [letter]. Nature (London) 351,530.
164
BRIAN L. KOTZIN ET AL.
Todd, J. K. (1988). Toxic shock syndrome. Clin Microbiol. Reo. 1,432-446. Todd, J. and Fishaut, M. (1978).Toxic-shock syndrome associated with phage-group-I staphylococci. Lancet 2, 1116-1118. Tomai, M., Kotb, M., Majumdar, G., and Beachey, E. H. (1990). Superantigenicity of streptococcal M protein. J. E x p . Med. 172,359-362. Tomai, M. A,, Aelion, J . A,, Dockter, M. E., Majumdar, G., Spinella, D. G., and Kotb, M. (1991). T cell receptor V gene usage by human T cells stimulated with the superantigen streptococcal M pr0tein.J. E x p . Med. 174,285-288. Tomai, M. A., Schlievert, P. M., and Kotb, M. 1992). Distinct T-cell receptor Vp gene usage by human T lymphocytes stimulated with the streptococcal pyrogenic exotoxins and pep M 5 protein. Infect. Zmmun. 60,701-705. Tracey, K. J., Beutler, B., Lowry, S. F., Merryweather, J., Wolpe, S., Milsark, I. W., Hariri, R. J., Fahey, T. J., Zentella, A., and Albed, J. D. (1986). Shock and tissue injury induced by recombinant human cachectin. Science 234,470-474. Trede, N. S., Geha, R. S., and Chatilla, T. (1991).Transcriptional activation of IL-1p and tumor necrosis factor-a genes by MHC class I1 1igands.J. Zmmunol. 146,2310-2317. Tsiagbe, V. K., Rabinowitz, J. L., and Thorbecke, G. J. (1991). I-E expression does not by itself inHuence growth of or T cell unresponsiveness to SJL lymphomas. Cell Zrnmu1101.136,329-339. Tsiagbe, V. K., Asakawa, J., Miranda, A., Sutherland, R.M., Paterson, Y., and Thorbecke, G. J. (1993a).The syngeneic response to SJL follicular B cell lymphoma (RCS)cells is primarily in Vpl6+, CD4+ T ce1ls.J. Zmmunol. (in press). Tsiagbe, V. K., Yoshimoto, T., Asakawa, J., Cho. S. Y., Meruelo, D., and Thorbecke, G. J. (1993b). Linkage of a superantigen-like stimulation of syngeneic T cells in a mouse model of follicular center B cell lymphoma to transcription of endogenous mammary tumor virus. E M B O J . (in press). Tsubura, A., Inaba, M., Imai, S., Murakami, A , Oyaizu, N., Yasumizii, R.,Ohnishi, Y., Tanaka, H., Morii, S., and Ikehara, S. (1988). Intervention of T-cells i n transportation of mouse mammary tumor virus (milk factor) to mammary gland cells in uioo. Cancer Res. 48,6555-6559. Tumang, J. R.,Posnett, D. N., Cole, B. C., Crow, M. K., and Friedman, S. M. (1990). Helper T cell-dependent human B cell differentiation mediated by a mycoplasrnal superantigen bridge. J . E x p . Med. 171,2153-2158. Uchiyama, T., Tadakuma, T.,Imanishi, K., Araake, M., Saito, S., Yan, X.J., Fujikawa, H., Igarashi, H., and Yamaura, N. (1989). Activation of murine T cells by toxic shock syndrome toxin-1. The toxin-binding structures expressed on murine accessory cells are MHC class I1 molecu1es.J. Immunol. 143,3175-3182. Vacchio, M. S., and Hodes, R.J. (1989). Selective decreases in T cell receptor Vp expression. Decreased expression of specific Vp families is associated with expression of multiple MHC and non-MHC gene pr0ducts.J. E x p . Med. 170,1335-1346. Vacchio, M. S., Ryan, J. J., and Hodes, R. J. :1990). Characterization of the ligand(s) responsible for negative selection of Vpll- and Vpl2-expressing T cells: Effects of a new MIS determinant.J. E x p . Med. 172,807413. Vacchio, M. S., Kanagawa, 0..Tomonari, K., and Hodes, R.J. (1992). InHuence of T cell receptor V a expression on Mls" superantigen-specific T cell responses. J . E x p . Med. 175,1405-1408. Vanier, L. E., and Prud'homme, G. J. (1992). Cyclosporin A markedly enhances superantigen-induced peripheral T cell deletion and inhibits anergy induction.]. E x p . Med. 176,37-46.
SUPERANTIGENS
165
Veronese, M. L., Veronesi, A., D'Andrea, E., Del Mistro, A., Indraccolo, S., Mazza, M. R., Mion, M., Zamarchi, R., Menin, C., and Panozzo, M. (1992). Lymphoproliferative disease in human peripheral blood mononuclear cell-injected SCID mice. I. T lymphocyte requirement for B cell tumor generation.]. Exp. Med. 176, 1763-1767. Waanders, G. A. and MacDonald, H. R. (1992). Hierarchy of responsiveness i n cjioo and in oitro among T cells expressing distinct Mls-la-reactive V/3 domains. E u r . ] . lmmunol. 22,291-293. Wallace, V. A., Rahemtulla, A., Timnis, E., Penninger, J., and Mak, T. W. (1992). CD4 expression is differentially required for deletion of MLS-la- reactive T cells. ]. E x p . Med. 176,1459-1463. Wang, B., and Kotb, B. (1992). Mapping of immunologically functional epitopes of the streptococcal superantigen, pep M5.32nd Intersci. Conf. Antimicro. Agent. (abstract). Wang, C. Y., Bushkin, Y., Pica, R., Lane, C., McGrath, H., and Posnett, D. N. (1986). Stimulation and expansion of a human T-cell subpopulation by a monoclonal antibody to T-cell receptor molecule. Hybridoma 5, 179-190. Webb, S. R., and Sprent, J. (1987). Downregulation of T cell responses by antibodies to the T cell receptor. J . E x p . Med. 165,584-589. Webb, S. R., and Sprent, J. (199Oa).Response of mature unprimed CD8' T cells to MIS" determinants.]. E x p . Med. 171,953-958. Webb, S. R., and Sprent, J. (199Ob). lnduction of neonatal tolerance to MIS"antigens by CD8+ T cells. Science 248,1643-1646. Webb, S. R., Okamoto, A., Ron, Y., and Sprent, J. (1989). Restricted tissue distribution of MIS"determinants. Stimulation of MIS"-reactive T cells by B cells but not by dendritic cells or niacrophages. J . E x p . Med. 169, 1-12. Webb, S. R., Morris, C., and Sprent, J. (1990). Extrathymic tolerance of mature T cells: Clonal elimination as a consequence of immunity. Cell 63, 1249-1256. Webb, S. R., O'Rourke, A. M., and Sprent, J. (1992). Factors influencing the fate ofT cells responding to MIS antigens. Semin. lmmunol. 4,329-336. Weeks, C. R. and Ferretti, J. J. (1986). Nucleotide sequence of the type A streptococcal exotoxin (erythrogenic toxin) gene from Streptococcus pyogenes bacteriophage T12. Infect. lmmun. 52, 144-150. White, J., Herman, A., Pullen, A. M., Kubo, R., Kappler, J. W., and Marrack, P. (1989). The Vp-specific superantigen staphylococcal enterotoxin B: Stimulation of mature T cells and clonal deletion in neonatal mice. Cell 56,27-35. White, J., Pullen, A., Choi, K., Marrack, P., and Kappler, J. W. (1993).Antigen recognition properties of mutant VB3+ T cell receptors are consistent with an immunoglobulinlike structure for the receptor.]. Exp. Med. 177, 119-125. Winslow, G. M., Scherer, M. T., Kappler, J. W., and Marrack, P. (1992). Detection and biochemical characterization of the mouse mammary tumor virus 7 superantigen (Mls-la). Cell 71,719-730. Woodland, D., Happ, M. P., Bill, J., and Palmer, E. (1990). Requirement for cotolerogenic gene products in the clonal deletion of I-E reactive T cells. Science 247, 964-967. Woodland, D. L., Happ, M. P., Gollob, K. J., and Palmer, E. (199la). An endogenous retrovirus mediating deletion o f a/3 T cells? [see comments]. Nature (London) 349, 529-530. Woodland, D. L., Lund, F. E., Happ, M. P., Blackman, M. A., Palmer, E., and Corley, R. B. (199lb). Endogenous superantigen expression is controlled by mouse mammary tumor proviral loci. 1.E x p . Med. 174,1255-1258.
166
BRIAN L. KOTZIN ET AL.
Woodland, D. L., Smith, H. P., Surman, S., Le, P., Wen, R.,and Blackman, M. A. (1993). Major histocompatibility complex-specific recognition of Mls-1 is mediated by multiple elements to the T cell receptor. J . Exp. Med. 177,433-442. Wucherpfennig, K. W., Ota, K., Endo, N.,Seidman, J. G., Rosenzweig, A., Weiner, H. L., and Hafler, D. A. (1990). Shared human T cell receptor Vp beta usage to immunodominant regions of myelin basic protein. Science 248, 1016-1019. Yagi, J., Baron, J., Buxser, S., and Janeway, C. A., Jr. (1990). Bacterial proteins that mediate the association of a defined subset of T cell receptor:CD4 complexes with class I1 MHC. J . Immunol. 144,892-901. Yui, K., Katsumata, M., Komori, S., Gill-Morse, L., and Greene, M. I. (1992).Response of Vp8.l' T cell clones to self Mls-la: Implications for the origin of autoreactive T cells. Int. Immunol. 4,125-133. Zamvil, S. S., and Steinman, L. (1990). The 'Tlymphocyte in experimental allergic encephalomyelitis. Annu. Reo. Immunol. 8,579-621. Zinkernagel, R. M., and Doherty, P. C. (1975).H-2 compatability requirement for T-cellmediated lysis of target cells infected with lymphocytic choriomeningitis virus. Different cytotoxic T-cell specificities are associated with structures coded for in H-2K or H-2D.J . Exp. Med. 141,1427-1436. This article was accepted for publication on 3 March 1993.
ADVANCES IN IMMUNOLOGY. VOL. 54
Interleukin-1 Receptor Antagonist WILLIAM P. AREND Division of Rheumotology, Department of Medicine, Univemivof ColomdoHealth Sciences Center, Denwr, Colorado 80262
1. Introduction
A network of cytokines mediates communication between cells of the immune and inflammatory systems. These cytokines are primarily active locally in the microenvironment of cells in tissues, but also may carry out effects distally in an intact organism in a hormone-like fashion. Cytokines are released by particular cells in response to a variety of stimuli, and activate target cells after binding to specific cell surface receptors. Regulation of the cytokine network is achieved by many different mechanisms including inhibition of cytokine production, binding in the soluble phase, competitive blocking of receptor binding, and decreased postreceptor induction of intracellular signals. Interleukin-1 (IL-1) was first identified as a macrophage-derived lymphocyte-activating factor (LAF) in the early 1970s (Gery et al., 1971, 1972; Gery and Waksman, 1972). In retrospect, multiple different biological activities in cell supernatants and body fluids, including that of endogenous pyrogen, probably were due, at least in part, to the molecule labeled in the mid-1980s as IL-1 (reviewed in Dinarello, 1990). IL-1 is a 17-kDa glycoprotein that is produced primarily by monocytes and macrophages, but may be synthesized by a variety of cells in multiple organs. Both IL-la and IL-lP bind to two different receptors that are present constitutively on many different target cells, or their production may be stimulated in these cells. Because of the seemingly ubiquitous production of IL-1, the apparent pleiotropic nature of its biological effects, and the presence of other cytokines with redundant properties, the precise role of IL-1 in normal physiology has not been clearly established. Most importantly, however, IL-1 has been implicated as a mediator of tissue destruction in many human diseases (Dinarello, 1991), including rheumatoid arthritis (Arend and Dayer, 1990). Given the widespread presence of IL-1 in different organs, its production by a variety of cells, and its hypothesized proinflammatory role in human diseases, many investigators had long sought the existence of naturally occurring inhibitors of IL-1. This work culminated in the 167 Copyright 8 1993 by Academic Press, Inc.
All rights of reproduction in any form reserved.
168
WILLIAM P. AXEND
description of a specific receptor antagonist of IL-1 (reviewed in Arend, 1990, 1991; Dinarello and Thompson, 1991). The molecule now known as IL-lra was first discovered as a 22- to 25-kDa IL-1 inhibitory bioactivity in the supernatants of human monocytes cultured on adherent immunoglobulin G (IgG) (Arend et al., 1985)and in the urine of patients with fever or myelomonocytic leukemia (Balavoine et al., 1985, 1986). This unique molecule represents the first known naturally occurring protein that functions as a specific receptor antagonist of any cytokine or hormone-like molecule. In an 8-year span (1985-1993) the story of IL-lra has rapidly evolved from an uncharacterized biological activity in monocyte supernatants to a recombinant molecule that is being evaluated for the treatment of at least seven different human diseases. This article summarizes the historical background of IL-1 inhibitors and the experimental work that led to the initial identification of the molecule now known as IL-lra. The purification, cloning, and expression of IL-lra, the characterization of IL-lra protein and gene structures, and the regulation of production of IL-lra also are discussed. The multiple reports on the in uitro and in viuo effects of IL-lra are reviewed, including early results from studies in human diseases. Lastly, some speculations are offered on the possible biological relevance of different structural variants of IL-lra.
A. INTERLEUKIN-1 INHIBITORS Many early studies described the presence of IL-1 inhibitory bioactivities in human body fluids or in the supernatants of a variety of human or animal cell cultures. These studies have previously been reviewed (Larrick, 1989; Arend et al., 1989; Dayer and Seckinger, 1989; Shaw, 1991; Higgins and Postlethwaite, 1991; Arend and Dayer, 1993). These IL-1 inhibitory activities were heterogeneous in size and biological properties and largely have remained unpurified and uncharacterized. In retrospect, however, many of these IL-1 inhibitors may have been IL-lra or soluble forms of IL-1 receptors. Molecules resembling IL-lra in size or biological properties were found under a variety of conditions: in the sera of normal donors injected with endotoxin (Dinarello et aZ., 1981); in the supernatants and lysates of stimulated human neutrophils (Tiku et al., 1986); secreted by Epstein-Barr virus (EBV)-infected human monocytes (Lotz et al., 1986a); in the supernatants of cultured synovial fluid mononuclear cells from rheumatoid arthritis patients (Lotz et al., 1986b); in normal human urine (Svenson and Bendtzen, 1988); in the supernatants of lipopolysaccharide (LPS)-stimulated Kupffer cells (Shira-
INTERLEUKIN-1 RECEPTOR ANTAGONIST
169
hama et al., 1988);secreted by cultured human alveolar macrophages (Gosset et al., 1988);and produced by the human monocytic leukemia cell line THP-1 (Gaffney et al., 1989). In addition to IL-lra, many of these studies may have described the presence of soluble IL-1 receptors (reviewed in Arend and Dayer, 1993). A variety of 95- to 100-kDa IL-1 inhibitory bioactivities have been described in the supernatants of virally infected macrophages or lymphocytes (Scala et al., 1984; Rodgers et al., 1985; Roberts et al., 1986; McCarthy et al., 1989; Salkind et al., 1991). In recent studies, soluble IL-1 receptors of 33 to 44 kDa were described in the supernatants of vaccinia or cowpox virus-infected CV-1 cells (monkey kidney fibroblast-like; Spriggs et al., 1992), encoded by the vaccinia virus gene B15R (Alcami and Smith, 1992). Whether the larger-sized IL-1 inhibitors produced by virally infected cells represent aggregates of soluble IL-1 receptors or more novel molecules remains to be determined. Lastly, a 52-kDa IL-1 inhibitor in the supernatants of human myelomonocytic M20 cells is not IL-lra, as its activity is not affected by an antiserum specific for IL-lra (Barak et al., 1986,1991). B. DISCOVERY OF INTERLEUKIN-1 RECEPTOR ANTAGONIST The IL-1 receptor antagonist was discovered by two laboratories independently searching for IL-1 and IL-1 inhibitory bioactivities in human monocyte supernatants or urine. My laboratory had explored the consequences on human monocytes and macrophages of binding different forms of immune complexes. Adherent immune complexes are found in the tissues of patients with various autoimmune diseases, including on the articular cartilage and in periarticular tissues of patients with rheumatoid arthritis and along the glomeruli of patients with specific types of nephritis. Monocytes crawl across this adherent IgG as they migrate into these diseased tissues and the cells differentiate into macrophages. As little was known about the consequences of this interaction on monocyte function, in the late 1970s my laboratory began to examine this issue in an in vitro experimental system. We showed that human monocytes cultured on adherent IgG secreted the neutral proteases plasminogen activator and elastase (Ragsdale and Arend, 1979; 1981). In addition, these cells exhibited a modulation of detectable Fc receptors on the exposed portion of the cell (Ragsdale and Arend, 1980), and of complement receptors as well when complement was included within the immune complex (Arend and Massoni, 1981). At the same time other laboratories were describing various bioactivities produced by monocytes and macrophages under specific culture conditions. In the mid-1980s these activities were decided to
170
WILLIAM P. AREND
be due to one molecule, termed interleukin-1; cytoplasmic DNAs (cDNAs) for IL-la and IL-l/? were cloned in 1985and 1986 (reviewed in Dinarello, 1990). Simultaneously, other investigators hypothesized that IL-l-like molecules may be present in rheumatoid synovial tissue and may mediate joint destruction by inducing the release of prostaglandins and neutral proteases by fibroblast-like cells in the synovium and chondrocytes in the underlying articular cartilage. Because macrophages in the synovial pannus were adherent to IgG in the form of immune complexes in the underlying articular cartilage, my laboratory proposed that this might be the mechanism of stimulation and production of IL-1 in the rheumatoid joint. In extensive studies we failed to find any IL-1 bioactivity released by human monocytes cultured on adherent IgG in uitro. We had previously explored for the presence of IL-1 inhibitory activity in the supernatants of monocytes stimulated with various agents. Therefore, we assayed for the presence of IL-1 inhibitory bioactivity in supernatants of monocytes cultured with various forms of immune complexes. In monocytes cultured on adherent immune complexes or IgG alone, a potent IL-1 inhibitory activity was observed, as assayed by inhibition of TL-1 enhancement of phytohemagglutinin (PHA)-induced proliferation of murine thymocytes or of IL-l-induced secretion of collagenase from cultured rabbit chondrocytes (Table I) (Arend et al., 1985). This IL-1 inhibitor had a molecular weight of -22 kDa and appeared to be specific for IL-1, having no effect on IL-2-induced thymocyte proliferation. On the basis of the earlier observations by many laboratories of the presence of IL-1 inhibitors in human urine, Dayer and colleagues pursued this direction. An IL-1 inhibitor of -25 to 35 kDa was found in the urine of three patients with myelomonocytic leukemia (Balavoine et aZ., 1985, 1986). This material inhibited the effects of IL-1 in the murine thymocyte assay, did not bind to a concanavalin A-Sepharose column, and was susceptible to denaturation by boiling or urea. This laboratory went on to show that the semipurified IL-1 inhibitor from urine blocked the stimulatory effects of both IL-la and IL-lP, but not of tumor necrosis factor a (TNF-a), on both murine thymocytes and human fibroblasts (Seckinger et al., 1987a). Most importantly, they demonstrated that the urine-derived material was a competitive inhibitor of IL-1 binding to murine EL4-6.1 thymoma cells (Seckinger et al., 1987b). My laboratory showed that the semipurified IL-1 inhibitor from monocytes cultured on adherent IgG exhibited the same properties (Arend et al., 1989). In addition to inhibiting IL-1 effects on murine thymocytes and rabbit articular chondrocytes, the monocyte-
171
INTERLEUKIN-1 RECEPTOR ANTAGONIST
TABLE I STIMULATION OF PRODUCTION OF INTERLEUKIN-1 INHIBITOR Material incubated with the monocytes (concentration)
ChondrocYte stimulation (% of control IL-1 activity)"
Thymocyte stimulationb % of control IL-1 activity
% of control IL-2 activity
LPS (20 ng/ml) HSA (100 pg/ml) Soluble anti-HSA (100 pg/ml) Soluble ICs (100 pg/ml) Precipitated ICs (100 pg/ml) Adherent HSA (2 ng/mm2) Adherent anti-HSA (4 ng/mm2) Adherent ICs (18 ng/mm2) IgC-coated sheep red blood cells
120 92.5 87.7 105.6 118.8 95.6 4.1 8.3 117.8
97.2 101.0 83.7 73.8 80.3 102.8 0.1 5.8 86.9
108.8 159.8 74.4 56.1 100.3 85.1 116.7 81.9 74.7
Ability of monocyte supernatants to inhibit IL-l-induced collagenase production from cultured rabbit articular chondrocytes. Supernatants were harvested after incubation of monocytes for 24 hours with human serum albumin (HSA, the antigen), rabbit anti-HSA antibodies, or different forms of immune complexes (ICs).The data are expressed as percentages ofcontrol collagenase production observed with IL-I alone. 'Ability of monocyte supernatants to inhibit either IL-1 or IL-2 augmentation of PHA-induced proliferation of murine thymocytes. For further details, see the original paper (Arend et al., 1985).
derived IL-1 inhibitor also blocked IL-1 induction of prostaglandin Ez (PGE2) production by human foreskin fibroblasts and synovial cells. The monocyte-derived IL-1 inhibitor was not transforming growth factor /3 (TGF-/3), was not immunologically cross-reactive with IL-1, and also blocked IL-1 binding to specific receptors on EL4-6.1 cells. Thus, two different laboratories demonstrated the existence of a receptor antagonist of IL-1, present either in urine or in supernatants of monocytes cultured on adherent IgG. II. Purification, cDNA Cloning, and Expression of Interleukin-1ra
During the mid- to late 1980s, many laboratories were engaged in purifying and characterizing this apparent receptor antagonist of IL-1. In early 1990 collaborative studies between scientists at Synergen and my laboratory described the purification of three apparent species of IL-lra from the supernatants of monocytes cultured on adherent IgG (Hannum et al., 1990).The cells were obtained by leukopheresis from normal donors and were plated on IgG-coated or fetal calf serum (FCS)-coated tissue culture wells in the presence of a low concentration of [35S]methionine and in the absence of added serum. The mono-
172
WILLIAM P. M E N D
cyte supernatant was harvested after 24 hours of culture and was fractionated by anion-exchange chromatography on a Mono-Q column. Three peaks of radioactivity, coinciding with peaks of IL-1 inhibitory bioactivity, were present in the supernatants from monocytes cultured on IgG and were absent in the supernatants of cells cultured on FCS. Thus, IL-lra was a major IgG-induced protein in these cells. Subsequent purification steps were greatly aided by being able to follow both radioactivity and bioactivity. The three peaks of IL-1 inhibitory activity from Mono-Q were termed x (18 kDa), a (22 kDa), and p (22 kDa). Each of these materials were purified to homogeneity by repeat anion-exchange chromatography followed by gel filtration chromatography on Superose 12 and C4 reversed-phase HPLC. The final preparations represented a yield of ~ 3 of %the IL-lra estimated to be present in the starting material and demonstrated identical single bands by both silver-stained polyacrylamide gel electrophoresis (PAGE) and autoradiography after PAGE. This IL-lra purification scheme is summarized in Table 11. TABLE I1 PURIFICATION OF INTERLEUKIN-1rP FROM Igc-INDUCED MONOCYTESUP ERN AT ANTS".^
Total protein Sample Crude QM salt supernatant Mono Q X
a
P
Total Superose 12 X
a
P
Total CCRPHPLC X
a
P
Total
(mg)
1448 111 0.80 1.73 4.74 7.27
Purification
(-fold)
-
Yield (96) -
10.3
78.5
101.8
51.0
0.047 0.075 0.165 0.287
0.00111 0.00156 0.00137 0.00404
1,162
23.0
10,894
3.0
a Purification of ILlra from the Supernatants of human monocytes cultured for 24 hours on adherent IgG. Details of the purification procedure can be found in Hannum et al. (1990). [Reprinted with permission from Nature (Hannum et al., 1990). Copyright (1990) Macmillan Magazines, Ltd.]
INTERLEUKIN-1 RECEPTOR ANTAGONIST
173
The three IL-1 inhibitors purified from IgG-induced monocytes were subjected to N-terminal sequence analysis. The x and a proteins exhibited the same N-terminal sequence through 20 residues determined, but the /3 protein could not be sequenced (Hannum et al., 1990).Peptides of the a and /3 proteins were obtained after endopeptidase digestion and were sequenced. Both the a and /3 proteins exhibited overlapping peptides with the same sequence. In addition, one /3 peptide demonstrated the same N-terminal sequence as did the intact x and a proteins. A composite sequence consisted of -60% each protein and indicated identical amino acid structures. This molecule was unique as no similar sequences could be located in the Protein Information Resource Databank. Subsequent digestion of the a and p proteins with N-glycanase reduced their size to 18 kDa, identical with the x protein. Thus, this monocyte-derived IL-1 inhibitor appeared to be one protein consisting of a nonglycosylated 18-kDa form and two N-linked glycosylated 22-kDa forms. These purified native IL-1 inhibitors each exhibited equivalent competitive inhibition of 12'I-IL-la binding to EL4-6.1 cells, proving that this molecule was a receptor antagonist of IL-1 and did not require the carbohydrate moiety for this activity. Furthermore, this molecule failed to exhibit agonist activity against dermal fibroblasts (i.e., no PGEz production). The cDNA for this IL-1 receptor antagonist, now termed ZL-Ira, was cloned from a AgtlO library prepared from IgG-induced monocytes after 17 hours of culture (Eisenberg et al., 1990). The library was screened with 15- and 20-residue oligonucleotide probes that were based on the amino acid sequences KMQAF and KFYFQED. One clone was identified with both probes; DNA was isolated from this clone and digested with EcoRI, yielding a 1.8-kb fragment. This DNA fragment hybridized in a Southern blot to five different oligonucleotide probes based on peptide sequences in the purified protein. This 1.8-kb fragment was cloned into the EcoRI site of the phage M13mp19 and its nucleotide sequence was determined. The cDNA contained one open reading frame encoding for a protein of 177 amino acids. A short 5' UTR of 14 nucleotides and a long 3' UTR of 1133 nucleotides also were present in this 1.8-kb cDNA. The 3' UTR of IL-lra mRNA lacked the AUUUA sequence implicated in regulating the mRNA halflife of many cytokines and growth factors (Shaw and Kamen, 1986). Further analysis indicated that the N-terminal 25 amino acids possessed an internal hydrophobic stretch resembling a signal sequence. A protein of 152 amino acids would result from removal of this leader peptide and, most importantly, would possess an N-terminal sequence identical to that of the monocyte-derived protein. Thus, this predicted
174
WILLIAM P. 4REND
IL-lra molecule would have a molecular weight of 17,115 and a pl of 5.2. The recombinant IL-lra molecule was obtained by cloning the cDNA into the T7 expression vector pT5T (Eisenberg et d.,1990). Expression was carried out in Escherichia coli using IPTG to enhance production. The bacterial lysates contained a dominant band identical in size to nonglycosylated native IL-lra. Furthermore, these lysates inhibited IL-1 induction of PGEz production in foreskin fibroblasts, as did lysates ofCOS cells transfected with the IL-lra cDNA in a mammalian expression vector. The recombinant IL-lra molecule was purified by serial ion-exchange and gel filtration chromatographies and then sequenced, It showed 26% amino acid sequence homology to human IL-lP and 19% homology to IL-la. In addition, IL-lP and IL-lra exhibited similar hydrophilicity plots. Thus, IL-lra shows some structural similarity to the two described forms of IL-1. The degree of homology between IL-la and IL-lP is similar to that between IL-lra and these two other forms of IL-1. Studies have shown that the four cysteines present in the IL-lra molecule are not disulfide-linked (Steinkasserer et al., 1992b). Another laboratory subsequently reported the purification, sequencing, cDNA cloning, and expression of an identical human IL- 1receptor antagonist protein, termed ZRAP (Carter et al., 1990). This molecule was obtained from supernatants of the human myelomonocytic leukemia cell line U937 after differentiation with phorbol myristate acetate (PMA) and stimulation with granulocyte-macrophage colonystimulating factor (GM-CSF). The purification scheme included sizeexclusion FPLC on a Superose 12 column and C4 reverse-phase HPLC. IRAP sequencing and cloning were carried out as for the monocyte-derived IL-lra molecule and revealed an identical structure. Recombinant IRAP was obtained by expression in E. coli and was purified by anion-exchange chromatography on QAE Sepharose and TSK DEAE 5PW columns, followed by sizing on a TSK Bio-Sil-125 HPLC column. The recombinant molecule exhibited an apparent molecular weight of 22 kDa and a pl of 6.0 by two-dimensional PAGE. An identical family of glycoproteins was also found in the supernatants of PMA-differentiated THP-1 cells, another human myelomonocytic leukemia cell line (Bienkowski et al., 1990). Four different IRAP molecules were isolated from the THP-1 supernatants, all being glycosylation derivatives of the same polypeptide. Interestingly, one form of IRAP exhibited a N-terminal sequence that started seven residues in from the usual N terminus. This was probably a product of proteolysis as protease treatment of recombinant IL-lra yielded an identical form
INTERLEUKIN-1 RECEPTOR ANTAGONIST
175
(Arend, unpublished observations). All of the IL-1 receptor antagonist forms isolated in these two studies possessed equivalent abilities to inhibit IL-1 binding to receptors on murine thymocytes or T cells. A third laboratory reported the partial purification and characterization of human IL-lra. This IL-1 inhibitor bioactivity was found in the supernatants of three additional human myelomonocytic leukemia cell lines, H-161, AML-193, and HL-60, after PMA differentiation and GM-CSF stimulation (Mazzei et al., 1990a). The molecule was partially purified from these cell supernatants by NH4S04 precipitation, ion-exchange chromatography on phenyl-Sepharose and Mono-Q columns, and gel filtration on a Superose 12 column. The IL-1 inhibitor also was partially purified from human urine by sequential ionexchange chromatography, hydrophobic chromatography, gel filtration chromatography, and negative immunosorption (Mazzei et al., 1990b). Although in neither study was the purified IL-1 inhibitor sequenced, both molecules had size and charge characteristics similar to those of IL-lra. In addition, the urine-derived IL-1 inhibitor was identified as a 23-kDa band in a Western blot by antibodies specific for IL-lra (Seckinger et al., 1990b). Thus, initial studies described IL-lra in human urine and in the supernatants of a variety of human myelomonocytic cell lines as well as in the supernatants of human monocytes. The apparently varying values for reported size of nonglycosylated IL-lra are due to the fact that this molecule migrates anomalously in PAGE. Although the actual molecular weight is 17,115, values in the literature range up to 21,000 to 22,000. Three laboratories subsequently have reported the cloning and expression of murine IL-lra (Shuck et al., 1991; Zahedi et al., 1991; Matsushime et al., 1991).In addition, similar studies have been published on rat (Eisenberg et al., 1991) and rabbit (Goto et al., 1992) IL-lra. The mouse and rat IL-lra proteins were each synthesized as 178-amino-acid precursors and were processed to 152-amino-acid mature proteins, as was the human mature IL-lra. Mouse and rat IL-lra displayed 77 and 75% amino acid sequence homology, respectively, to human IL-lra (Eisenberg et al., 1991). Although rabbit IL-lra was synthesized as a 177-amino-acid precursor, it was processed to a smaller molecule of 143 residues (Coto et al., 1992). Thus, the mature form of rabbit IL-lra was shorter by nine amino acids at the N terminus in comparison to the human, mouse, or rat protein. Another difference was that the native rabbit IL-lra molecule appeared not to be glycosylated, although it possessed the same potential N-glycosylation site at Asn 109 as did IL-lra from the other species. Rabbit IL-lra exhibited 77% amino acid sequence homology to the human molecule. Thus,
176
WILLIAM P. 4REND
IL-lra appears to be a highly conserved molecule, similar to I L - l a and IL-lP. It will be seen later that structural variation at the N terminus is characteristic of IL-lra. 111. Structural Variants of Interleukin-1ra Protein
A structural variant of IL-lra was described in 1991, with the mRNA found in keratinocytes and other epithelial cells (Haskill et al., 1991). These investigators had previously identified 14 different mRNAs rapidly induced in human monocytes by adherence to plastic or to substrates containing connective tissue components or immune complexes (Sporn et al., 1990).These cloned cDNAs were sequenced and many appeared to code for unique proteins not previously described. One sequence that was rapidly induced in monocytes by adherence, MAD 15, had a 1.8-kb transcript. A single full-length cDNA corresponding to this sequence was subsequently isolated from a human blood cell library. This cDNA was identical to the IL-lra sequence earlier reported by Eisenberg et ul. (1990)except at the 5’ end, where 85 bp was replaced by a different sequence of 130 bp. This structural variant appeared to be created when a different first exon was spliced into an internal acceptor site within the first exon for monocytederived IL-lra in the region encoding for the leader sequence (between bp 87 and 88 of the cDNA). Thus, the keratinocyte variant of IL-lra contained 159 amino acids, the C-terminal 152 residues being identical to those of monocyte-derived IL-lra. The additional seven amino acids in keratinocyte IL-lra exhibited three different residues at the N terminus (MAL), with the remaining four extra residues being identical to the leader sequence of monocyte-derived IL-lra (ETIC) (Fig. 1).Because the keratinocyte variant lacked a full secretory peptide, it remained intracellular in keratinocytes. Thus, this form of the molecule was termed icIL-2ru (for intracellular IL-lra), as opposed to sZL-2 ru (secretory IL-lra) from monocytes. Throughout this article, the term IL-1ra refers to the monocyte-derived molecule unless noted otherwise. It is likely that other structural variants remain to be characterized as extracts of various cells or tissues exhibit either smaller (Haskill et al., 1991) or larger (Haskill; Gahring and Arend, unpublished observations) bands on Western blot analysis that are recognized by antibodies specific for IL-lra. These additional forms of IL-lra may be generated by proteolysis or by various post-transcriptional mechanisms, including alternative RNA splicing, or may represent primary products of different genes.
INTERLEUKIN-1 RECEPTOR ANTAGONIST
152 aa
I
152 aa
177
Precursor sll-1 ra
I
Mature sll-lra icll-1ra
FIG.1. Structures of three forms of human IL-lra, all possessing the same 152-aminoacid C-terminal sequence. Precursor sIL-Ira represents the synthesized form with a 25-amino-acid leader sequence. The icIL-lra molecule is 152 amino acids in length; the N-terminal M, A, and L residues are unique to this molecule and the penultimate E, T, I, and C residues are identical to those of the sIL-lra leader sequence.
IV. Chromosomal localization and Structures of Interleukin-1ra Genes
The gene for the human IL-lra has been localized to the long arm of chromosome 2, mapping to bands 2q13-14.1 (Lennard et al., 1992; Steinkasserer et al., 1992c; Patterson et al., 1993). Most interestingly, this same region of human chromosome 2 also contains the genes for IL-la, IL-1P, type I IL-1 receptor (IL-lRI), and type I1 IL-1 receptor (IL-1RII);however, the gene for IL-lra is not immediately adjacent to those for IL-la and IL-1P because none of these genes colocalized to the same yeast artificial chromosome (YAC) (Patterson et al., 1993). The gene for murine IL-lra also localizes to chromosome 2, in the proximal region between the centromere and Spna2 (Zahedi et al., 1991).The genes for murine IL-lra and IL-lP also map to this same region; however, the genes for murine IL-1RI and IL-1RII map to chromosome 1 in a region known to be syntenic to a portion of human chromosome 2 (reviewed in Patterson et al., 1993). T h e significance remains unclear of the common chromosomal localization in the human of the genes for three different forms of IL-1 as well as for two different IL-1 receptors. The gene structure for human IL-lra is similar to those for IL -l a and IL-lP (Fig. 2) (Eisenberg et al., 1991); however, the IL-lra cDNA contains four exons and three introns, whereas the portion of the cDNA encoding for the mature 17-kDa forms of IL-la and IL-lP contains three exons and two introns. The first exon of the IL-lra gene appears to be unique, as the N-terminal portion ofthe IL-lra protein has little homology with the same regions of IL-la and IL-1P proteins. On the basis of these observations, Eisenberg et al. (1991) have hypothesized that the divergence of IL-lra from IL-1 is due to a partial duplication of an
precursor a.a. # c3
mature a-a. #
mature =.a. # precursor a.a. #
IL-Ira mature a.a. #
*
1-16 16-32 33
-
107 107
-
-
164 164 205
206
-
m
-.
* * +
FIG.2. Gene structures for IL-la, IL-lp, and IL-lra. The boxes correspond to exons and the broken horizontal bars are introns. The left end of the gene is the start of transcription. Empty boxes are untransiated regions, lightly stippled areas are translated sequences, and heavily stippled areas represent sequences encoding mature proteins. Amino acid (a.a.) numbering for the precursor proteins and the mature proteins is based on the human sequences. Initial amino acids for mature IL-la, IL-1@,and IL-lra correspond to precursor amino acid numbers 113, 117, and 26, respectively. (Adapted from Eisenberg et al., 1991, with permission.)
INTERLEUKIN-I RECEPTOR ANTAGONIST
179
ancestral IL-1 gene. The first exon and a small part of the second exon of the IL-lra gene appear to b e derived from a different segment of the genome than gave origin to the IL-la and IL-lP genes. From the calculated mutational rates for these genes, it was concluded that the divergence of IL-lra from IL-1 was a very early event. Thus, similar to IL-1, it is likely that IL-1 receptor antagonist molecules will be identified in lower phylogenetic forms of life. Lastly, a triallelic variablelength polymorphism has been described in intron 2 of the human IL-lra gene (Steinkasserer et al., 1991). This polymorphism appears to be due to a variable copy number of an 86-bp sequence (Tarlow et al., 1993).Whether this allelic variation in intron 2 influences any function or property of IL-lra, such as relative gene expression, remains to be determined. The full genomic structure of 1L-lra, including the exon for the intracellular variant, has recently been described (Bienkowski et al., 1993).The sIL-lra gene is 6.4 kb long and the first exon for icIL-lra lies another 9.6 kb upstream (Fig. 3).Thus, the promoter for sIL-lra actually lies within intron 1for icIL-lra. The 5' flanking region ofexon 1for icIL-Ira has recently been sequenced and mapped (Bienkowski et d.,1993). This promoter appears to lack a traditional TATAA or CAAT motif so it must use an alternative mechanism of transcriptional initiation. The sequence and detailed characterization of the sIL-lra promoter have been reported (Smith et al., 1992): 1680 bp of 5' flanking
-1
I II 111
6.4kb
IV
1-
v icll-1ra sll-1 ra
FIG.3 . Interleukin-lra genomic DNA and mRNA structures. The gene for each form of IL-lra possesses four exons and three introns. The first exon is different between icll-lra and slL-lra, and the two lie 9.6 kb apart in genomic DNA. The relative sizes of different regions are not proportionately represented in these figures.
180
WILLIAM P. AREND
region for the sIL-lra gene was sequenced, revealing a TATAA box at -26 with consensus sequences for possible NFkB-, NFIL-lpA-, AP-1and CRE-binding sites further upstream. Transfection studies showed that the IL-lra promoter was active only in those human or murine cell lines (macrophage) that could be induced to produce IL-lra endogenously. A series of 5’-deletional mutants was generated to determine the cis-acting transcriptional elements responsible for both baseline and LPS-induced IL-lra promoter activities. Both the -294 to - 148 and - 148 to -85 regions exhibited potent regulatory function in both deletional and heterologous promoter studies. Further experiments are characterizing the precise DNA sequences in these proximal regions responsible for transcriptional regulation after stimulation with LPS or adherent IgG, as well as the interacting nuclear protein factors. Similar studies also are in progress on the icIL-lra promoter, as the two IL-ha promoters appear to have different patterns of specificity regarding both cell type and degree of cell differentiation (Smith and Arend, unpublished observations). V. Regulation of Interleukin-1ra Production
IL-lra was originally described as a net IL-1 inhibitory bioactivity in the supernatants of human monocytes cultured on adherent IgG. Under these conditions the cells also produce some IL-1p; however, the amount of IL-lra is much greater and obscures detection of any IL-1 bioactivity. Both IL-1 and IL-lra also can be produced by other cells including in uitro-derived or tissue macrophages, neutrophils, fibroblasts, chondrocytes, and hepatic cells, as reviewed later. Macrophage cell lines have been observed to produce both IL-1p and IL-lra simultaneously (Smith and Arend, unpublished observations). Published studies also indicate that both proteins can be present at the same time within a single human monocyte (Anderson et al., 1992). The effects of LPS on production of IL-la, IL-Lp, and IL-lra were examined in human monocytes at the single-cell level using specific antibodies and indirect immunofluorescence techniques with double staining. All three cytokines showed peak production, as percentage of monocytes positive, at 4 to 6 hours after culture with 100 ng/ml LPS; however, a maximum of only 48% of monocytes produced IL-lra as compared with 75 and 80% for IL-la and IL-lp, respectively. In addition, the early patterns of intracellular staining were different for these three cytokines: IL-la and IL-1p exhibited both filamentous cytoplasmic and weak membrane staining, whereas at 4 to 6 hours IL-lra was localized primarily to the Golgi complex. This pattern of IL-la and IL-1p locali-
INTERLEUKIN-1 RECEPTOR ANTAGONIST
181
zation within the monocyte persisted through 72 hours. In contrast, the Golgi localization of IL-lra in LPS-stimulated monocytes decreased markedly by 12 hours to be replaced by a second wave of IL-lra appearance diffusely in the cytoplasm at 24 to 48 hours. This second pattern of IL-lra in the cells, however, was seen in a maximum of only 20 to 30% of the cells (Andersson et al., 1992). Although not proven by these studies, this late cytoplasmic localization of IL-lra in monocytes may represent synthesis of the intracellular form. Both alveolar macrophages (Quay et al., 1993) and fibroblasts (Chan et al., 1992) are capable of producing both forms of IL-lra mRNA, as reviewed later. A. DIFFERENTIAL REGULATION OF INTERLEUKIN-1P AND INTERLEUKIN-1ra PRODUCTION IN MONOCYTES
Recent studies have shown that production of the agonist IL-lP and production of the receptor antagonist IL-lra by human monocytes are differentially regulated (Arend et al., 1991; Poutsiaka et al., 1991). In these studies IL-1P and IL-lra production was studied at both the mRNA level, using in vitro hybridization with specific cDNA probes, and at the protein level, using specific radioimmunoassays (RIAs) or enzyme-linked immunoadsorbent assays (ELISAs) (Malyak et al., 1991). Adherent human monocytes stimulated with LPS produced near-equivalent amounts of IL-lP and IL-lra proteins over 24 hours in culture; relative steady-state mRNA levels paralleled the protein levels (Arend et al., 1991). Enhanced rates of transcription for IL-lP and IL-lra, as determined by nuclear run-on assays, were observed in LPS-stimulated monocytes. In addition, the mRNA half-lives were similar for IL-1P and IL-lra ( t %of -2 to 4 hours) in LPS-induced cells. In contrast, monocytes cultured on adherent IgG exhibited a low level of IL-1P transcription with little to no protein production; however, IgG-induced monocytes displayed a high and sustained rate of IL-lra transcription as well as a marked prolongation in mRNA stability ( t % > 15 hours). Paradoxically, the addition of high concentrations of LPS (10 ng/ml to 1 pg/ml) to monocytes cultured on IgG led to a reduction in levels of IL-lra produced to the levels seen with LPS alone, without altering levels of IL-1P production. This LPS effect was mediated at both the transcriptional and post-transcriptional levels. Thus, LPS induces production of both IL-1P and IL-lra in human monocytes but adherent IgG selectively stimulates IL-lra protein production. Additional studies by another laboratory further substantiated that IL-lp and IL-Ira production by monocytes was regulated in a different and possible differential fashion (Poutsiaka et al., 1991). In this study,
182
WILLIAM P. 4REND
mixed peripheral blood mononuclear cells (PBMCs) were incubated in suspension culture at the bottom of polypropylene tubes. In comparison to adherent monocyte culture, the monocytes in this suspension culture are more concentrated and in greater contact with lymphocytes. The addition of 1% human AB serum to this culture system enhanced IL-lra production seven fold over 24 hours without altering the level of IL-1p production. Soluble IgG (2.5 to 100 pg/ml) or GMCSF (1-100 ng/ml) similarly enhanced IL-lra production without altering IL-1p. Conversely, rocking the cultures to prevent settling out of the cells and closer interactions did not change the level of IL-1P production but decreased by 75 to 80% the amount of IL-lra produced. Thus, these results indicate that monocyte production of IL-lra is enhanced by serum, IgG, GM-CSF, and cell-cell contact (Poutsiaka et al., 1991). Although not examined, it is likely that the IgG effect was due to the presence of aggregates, as monomeric IgG in solution is a weak inducer of IL-lra. Furthermore, the effect of serum may not have been due to the presence of IgG, as IL-lra production by monocytes may be induced by growth factors or other serum response factors, particularly after a short period of serum starvation (Gahring and Arend, unpublished observations). The results of earlier studies indicated that the effect of adherent IgG on induction of IL-lra production in monocytes was mediated through receptors for the Fc portion of IgG (Schur et al., 1990). Adherent Fc fragments but not F(ab’)z fragments were effective, whereas neither fibronectin nor three of its fragments exhibited any stimulatory activity. Adherent human IgG of all four subclasses enhanced IL-lra production by the monocytes in the order IgGl > > > IgG3 > IgG2 > IgG4. Lastly, adherent immune complexes with complement stimulated levels of IL-lra production by monocytes equivalent to those stimulated without complement, but detectable levels of IL-1 production were now observed.
B. EFFECTSOF OTHER CYTOKINES ON INTERLEUKIN-Ira PRODUCTION BY MONOCYTES The results of studies discussed earlier indicated that IL-1P production and IL-lra production could occur in the same monocyte, but that each protein was regulated by different mechanisms. Furthermore, IL-lra mRNA and protein production by these cells seemed to lag behind IL-1, suggesting that IL-lra production may be influenced by materials released at earlier times by these cells or by lymphocytes in the culture system. As was reported for macrophage cell lines, GMCSF enhanced IL-lra production by monocytes cultured over 1 to 4
INTERLEUKIN-I RECEPTOR ANTAGONIST
183
days (Shields et al., 1990; Jenkins and Arend, 1993); however, this cytokine seems to be more potent as a differentiating factor than as a direct inducer of IL-lra production (see later text). Although IL-lP alone is a weak inducer of IL-lra production by PBMCs in suspension culture, it may enhance the effects of IgG (Poutsiaka et d.,1991). IL-la also is a weak inducer of IL-lra production by adherent PBMCs, but IL-3 alone seems to have more potent stimulatory effects (Jenkins and Arend, 1993). Thus, production of IL-1 and IL-3 in mixed cell suspension cultures may enhance the stimulatory effects of other agents on monocyte production of IL-lra (Table 111). An even more potent effect is observed with IL-4. IL-4 alone does not induce IL-lP production in monocytes. One laboratory reported that IL-4 alone enhanced IL-lra production by PBMCs cultured in suspension (Vannier et al., 1992); however, other laboratories failed to find any stimulatory effects of IL-4 alone on IL-lra production by elutriated monocytes in suspension culture (Fenton et al., 1992; Orino et al., 1992). In addition, IL-4 alone is only a weak inducer of IL-lra production by PBMCs in adherent culture (Jenkins and Arend, 1993). IL-4 potently upregulates LPS-induced IL-lra production by monocytes in suspension or adherent culture, while decreasing LPSinduced IL-1P production (Table 111) (Vannier et al., 1992: Fenton et al., 1992; Orino et al., 1992; Jenkins and Arend, 1993). IL-4 appears to have the same differential effect on IL-la induction of these two TABLE 111 EFFECTS OF OTHER CYTOKINES ON INTERLEUKIN-1Ta PRODUCTIONa Cell
Inducing cytokine (strength)
Monocyte
IL-1 (weak inducer) IL-3 (moderate inducer) IL-4 (weak alone, strong enhancer of LPS) IL-10 (weak alone, moderate enhancer of LPS) GM-CSF (strong differentiating agent) GM-CSF (moderate inducer) TNF-a (moderate inducer) GM-CSF (primarily enhancer of LPS) IL-4 (primarily enhancer of LPS) TNF-a (moderate inducer)
Macrophage Neutrophil Keratinocyte
Effects of other cytokines on induction of IL-Ira protein production in human cells. Further details can be found in the followingpublications: monocytes (Jenkins and Arend, 1993; Jenkins et al., 1993); macrophages (Jansonet d ,1991,1993); neutrophils(Malyak et al., 1993b);keratinocytes (Kutschet al., 1993).
184
WILLIAM P. AREND
cytokines, enhancing IL-lra while depressing IL-lP production (Vannier et al., 1992).The inhibitory effects of IL-4 on LPS-induced IL-lP production are accompanied by an increase in the half-life of IL-lP mRNA (Donnelly et aZ., 1991), whereas the half-life for IL-lra mRNA remains unchanged (Fenton et al., 1992). Thus, IL-4 may enhance LPS- or IL-l-induced IL-lra production by monocytes primarily at the transcriptional level. IL-10 also differentially regulates IL-lP and ILIra production by monocytes, potently inhibiting production of IL-1P and other cytokines by monocytes stimulated with LPS or interferon-y (Fiorentino et al., 1991; de Waal Malefyt et al., 1991; Ralph et al., 1992); however, IL-10 alone is a weak stimulator of IL-lra production by monocytes and only a moderate enhancer of LPS-induced IL-lra (Jenkins et al., 1993). Neither IL-3 nor GM-CSF has an enhancing effect similar to that of IL-4 or IL-10 on LPS-induced IL-lra production by adherent PBMCs (Jenkins and Arend, 1993). In addition, my laboratory has found no effect of IL-3, IL-4, or GM-CSF on IgG-induced IL-lra production by adherent PBMCs (Jenkins and Arend, 1993), whereas high concentrations of LPS exhibited an inhibitory effect as previously published (Arend et al., 1991). Another laboratory reported that GM-CSF enhanced IgG-induced IL-lra production by PBMCs in suspension culture (Poutsiaka et al., 1991). Lastly, a variety of other cytokines including IL-2, IL-6, M-CSF, G-CSF, interferon-y, TNF-a, platelet-derived growth factor (PDGF-BB), epidermal growth factor (EGF), and acidic and basic fibroblast growth factor (FGF), all have no direct effects on IL-lra production by adherent PBhlCs, either alone or added to adherent IgG (Jenkins and Arend, 1993). One laboratory has reported that TGF-P is a direct stimulant of IL-lra production by adherent monocytes (Turner et al., 1991),but my laboratory cannot confirm this observation (Jenkins and Arend, 1993). The expression of TGF-P receptors is, however, sensitively regulated on these cells and variability in TGF-P receptor expression may explain these different results. In addition, some of the variability in effects of other cytokines on IL-lra production discussed earlier may result from differences in the techniques of cell preparation and culture. The important point is that IL-lra production by monocytes in uitro is regulated differently than IL-1P production and may be influenced by many factors including the presence of other cytokines. As monocytes in uiuo, particularly locally in tissues, are exposed to other cytokines, the net resultant level of IL-lra production may not be as predictable as might be believed from the results of in uitro studies.
INTERLEUKIN-I RECEPTOR ANTAGONIST
185
c. INTERLEUKIN-1ra PRODUCTION BY OTHER CELLS 1. In Vitro-Deriued Macrophages In addition to freshly isolated human monocytes, numerous other cells have demonstrated IL-lra production. Differentiation of monocytes into macrophages during in uitro culture led to greatly enhanced levels of IL-lra production (Roux-Lombard et al., 1989; Janson et al., 1991). In contrast to monocytes, the in uitro-derived macrophages exhibited IL-lra production without further stimulation; however, differentiation in GM-CSF led to enhanced levels of IL-lra synthesis without appreciable secretion. Levels of 1L-lra protein production by these cells were paralleled by relative steady-state IL-lra mRNA levels, suggesting that transcription may be the major mechanism involved. Again, in contrast to monocytes, levels of IL-lra production by in uitro-derived macrophages were not induced by LPS or culture on adherent IgG (Janson et al., 1991). The IL-lra protein that was secreted by these cells was the 22- to 25-kDa variably glycosylated species, whereas lower-molecular weight forms were found inside the cells. The IL-lra protein in the supernatants of in uitro-derived macrophages appeared to be equipotent to recombinant 17-kDa nonglycosylated protein in inhibiting IL-1 effects in the murine thymocyte assay.
2 . Alveolar Macrophages Numerous investigators have demonstrated that alveolar macrophages also produce large amounts of IL-lra (Galve-de Rochemonteix et al., 1990; Nagai et al., 1991;Takeuchi et al., 1992; Moore et al., 1992; Kline et al., 1992; Quay et al., 1993; Janson et al., 1993). Some of these studies showed that alveolar macrophages obtained from the bronchoalveolar lavage fluid of normal donors or of patients with interstitial lung diseases produced an IL-1 inhibitory bioactivity during in uitro culture. This IL-1 inhibitory activity was found to block the lZ5I-IL-la binding to EL4-6.1 murine thymoma cells (Galve-de Rochemonteix et ul., 1990) or to murine thymocytes (Takeuchi et al., 1992), and thus probably represented IL-lra protein. Later studies showed that unstimulated human alveolar macrophages produced IL-lra mRNA and protein during in uitro culture and, similar to in uitro-derived macrophages, failed to enhance production in response to LPS or adherent IgG (Moore et al., 1992; Kline et al., 1992; Janson et al., 1993). In contrast, alveolar macrophage production of IL-lra was increased by culture in IL-4 (Moore et al., 1992), GM-CSF (Janson et al., 1993) or serum (Kline et al., 1992). Freshly isolated cells failed to contain any
186
WILLIAM P.
AREND
detectable IL-lra mRNA or protein, indicating that these cells were primed in uiuo to exhibit enhanced levels of IL-lra production during in uitro culture (Janson et al., 1993). The production of IL-lra by alveolar macrophages during in uitro culture was not dependent on adherence or the presence of serum. Equivalent levels of IL-lra production were exhibited by alveolar macrophages from normal donors who were either nonsmokers or smokers; however, cells from nonsmoking patients with interstitial lung diseases (ILD) spontaneously produced over twofold more IL-lra than did cells from healthy nonsmokers (Janson et al., 1993). Interestingly, this enhanced IL-lra production was not demonstrated by cells from the lungs of smoking ILD patients. Thus, smoking may dampen IL-lra production in alveolar macrophages from patients with fibrotic lung diseases. Again, the IL-lra protein production was paralleled by mRNA levels, the IL-lra secreted by alveolar macrophages was the larger glycosylated species, and this IL-lra was able to block receptor binding of IL-1 (Janson et al., 1993). Some aspects of IL-lra production by freshly isolated monocytes, in uitro-derived macrophages, and alveolar macrophages are summarized in Table IV. The reasons for the failure of the latter two types of cells to respond to LPS or adherent IgG are not clear but may relate to changes in Fc receptor expression or in signal transduction mechanisms. Lastly, alveolar macrophages are capable of producing the mRNA for both structural variants of 1L-lra. Culture of these cells with respiratory syncytial virus enhanced levels of both mRNA, although mRNA for sIL-lra remained in great excess ofthat for icIL-lra (Quay et al., 1993). Production of the mRNA for both variants of IL-lra has also been observed in fibroblasts (see later text). TABLE IV INTERLEUKIN-1rBPRODUCTION BY MONOCYTES AND MACROPHAGES' Monocytes Constitutive Stimulated by LPS Induced by adherent IgG Enhanced by GM-CSF
-
+ +++ f
In oitro-derived Alveolar macrophages macrophages
+
+
+++
++
-
-
a Summary of studies on induction of IL-lra production by monocytes and macrophages. For further details see the following publications: monocytes (Arend et al., 1989); in oitro-derived macrophages (Janson et al., 1991); alveolar macrophages (Jansonet al., 1993).
INTERLEUKIN-1 RECEPTOR ANTAGONIST
187
3. Synovial Macrophages IL-lra production also has been described with synovial macrophages from either the fluid or tissue. Synovial fluid macrophages obtained from patients with inflammatory arthritis spontaneously released IL-1 inhibitory bioactivity during in vitro culture (Rouxbinding to IL-1 Lombard et al., 1992). This material blocked 1251-IL-la receptors on EL4-6.1 cells and contained a 21.5-kDa band in Western blot analysis using a polyclonal antiserum specific for IL-lra. It was not ascertained in these studies whether freshly isolated synovial fluid macrophages contained IL-lra mRNA or protein; however, it does appear that synovial fluid macrophages, like alveolar macrophages, are capable of constitutively producing IL-lra during in vitro culture in the absence of further stimuli. This IL-lra appears to be the larger glycosylated form. IL-lra mRNA and protein also are present in synovial tissue macrophages in histological sections from the synovium of patients with rheumatoid arthritis (RA) or osteoarthritis (OA) (Firestein et al., 1992a; Koch et al., 1992). The predominant staining for IL-lra protein in RA tissues was in perivascular macrophages in the sublining region, with staining also in the lining cells. In OA, however, the IL-lra protein was localized primarily to the lining cells. No IL-lra was detected in the synovia of patients without arthritis, suggesting that in both inflammatory and noninflammatory forms of arthritis, IL-lra production in the synovium is induced (Firestein et al., 1992a). This conclusion is further supported by the findings that IL-lra mRNA and protein are present in macrophages isolated from RA synovial tissue (Koch et al., 1992) and the cultured cells from RA or OA synovial tissue produce biologically active IL-lra protein (Firestein et al., 1992a). In addition to macrophages, it appears that synovial fibroblast-like cells also produce IL-lra (Firestein et d., 1992b). In fact, the mRNAs for both secretory and intracellular forms of IL-lra are found in extracts of both RA and OA synovium. Thus, macrophages and/or fibroblasts from diseased synovium may produce both forms of IL-lra, similar to alveolar macrophages and dermal fibroblasts. Synovial fluids also contain an IL-I inhibitory bioactivity that is due, at least in part, to IL-lra (Roux-Lombard et al., 1992). Elevated IL-lra protein levels, up to 45 ng/ml, are present in the synovial fluids from -80% of patients with rheumatoid arthritis as well as in =30%of fluids from patients with infectious or inflammatory, nonrheumatoid arthropathies (Maylak et al., 1993a).The IL-lra levels in synovial fluids correlate with concentrations of neutrophils. Furthermore, isolated synovial fluid neutrophils produce IL-lra, as do peripheral blood neutrophils
188
WILLIAM P. AREND
(see later). Lastly, cultured chondrocytes from articular cartilage also appear to constitutively produce IL-lra, in addition to IL-la and IL-lP (Tiku et al., 1992).Thus, IL-lra in synovial fluids may be derived from neutrophils and macrophages present in the fluid, macrophages and fibroblasts in the synovial tissue, and chondrocytes present in the underlying articular cartilage. 4 . Peritoneal Macrophages, Hepatoc ytes, and Uterine Stromal Cells Macrophages were purified from the peritoneal fluids of women who underwent laparoscopy as part of an evaluation for infertility. Low y blot analysis in the levels of IL-lra mRNA were found l ~ Northern cells of 30% or less of women with no disease, postinflammatory adhesions, and stage I or I1 endometriosis (Mori et al., 1992);however, the peritoneal macrophages from four of five women with stage I11 or IV endometriosis exhibited high levels of IL-lra mRNA. The presence of IL-lra protein in these cells and the ability of peritoneal macrophages to produce IL-lra during in uitro culture were not examined. By use of specific polymerase chain reaction (PCR) primers, a secretory IL-lra transcript was isolated from the hepatoma cell line HepG2 and from total liver RNA (Steinkasserer et al., 1992a). Thus, IL-lra mRNA is present in hepatic cells but it has not been determined whether these cells can produce IL-lra protein or whether hepatic macrophages, i.e., Kupffer cells, also contain IL-lra mRNA and can synthesize protein. Lastly, IL-lra mRNA was present in whole ovarian material as well as in macrophage-free preovulatory follicular aspirates (Hurwitz et al., 1992). This later finding suggests that IL-lra may be derived from granulosa cells in the ovary; however, no detectable IL-lra mRNA was found in cultured ovarian granulosa or theca cells, in either the absence or the presence of the stimulatory agent forskolin. Thus, it appears that IL-lra mRNA may be found constitutively in granulosa cells, suggesting that this protein may be produced by these cells in uiuo and may play some self-regulatory role in the ovary (Hurwitz et al., 1992).
5. Fibroblasts Recent studies have shown that fibroblasts cultured from skin biopsies of normal human adults contain IL-lra mRNA and are capable of synthesizing the protein (Chan et al., 1992). PCR amplification of reverse-transcribed mRNA indicated that icIL-lra mRNA was present in unstimulated fibroblasts with levels increased after culture in PMA. In contrast, low levels of sIL-lra mRNA were detected only in PMAstimulated fibroblasts. Immunofluorescence studies with a murine monoclonal antibody specific for IL-lra protein of both structural vari-
INTERLEUKIN-1 RECEPTOR ANTAGONIST
189
ants revealed the presence of protein diffusely throughout the cytoplasm, as well as occasionally in the nucleus, of both resting and stimulated fibroblasts. In addition, IL-lra protein was detected by ELISA in the lysates of both unstimulated and PMA-induced cells, but appeared not to be secreted by the fibroblasts. The ELISA detects both forms of IL-lra. These results strongly suggest, however, that icIL-lra is the predominant form of the protein synthesized by dermal fibroblasts. Thus, dermal as well as synovial fibroblasts are capable of transcribing the mRNA for both forms of IL-lra but may produce primarily icIL-lra protein. 6 . Neutrophils The original description of cloning of icIL-lra cDNA indicated that only sIL-lra mRNA was present in human peripheral blood neutrophils (Haskill et ul., 1991). This observation has been pursued further in recent publications, Studies on induction of protein synthesis in neutrophils noted that GM-CSF and TNF-a stimulated the synthesis and secretion of a 23-kDa protein, later noted to be IL-lra (McColl et uZ., 1990, 1992; Beaulieu et aZ., 1992). In addition, LPS also induced IL-lra production in neutrophils and both IL-4 and GM-CSF enhanced the stimulatory effects of LPS (Malyak et d . ,1993b). Further studies indicated that small amounts of IL-lra protein in the absence of mRNA were detected in the lysates of freshly isolated neutrophils from both peripheral blood and synovial fluid (Malyak et ul., 1993a,b). The absence of IL-1p mRNA further substantiated that these cells were not activated during the isolation procedure. Thus, it appears that the neutrophils may have been stimulated to transcribe IL-lra mRNA and synthesize protein earlier in their life cycle, with some residual protein remaining in or on the cells. The IL-lra protein synthesized by GM-CSF- or TNF-a-stimulated polymorphonuclear neutrophils (PMNs) largely remains intracellular and greatly exceeds the amounts of synthesized IL-1p (McColl et al., 1992; Malyak et al., 1993b).Furthermore, the IL-lra protein in supernatants of stimulated neutrophils is largely in the 22- to 25-kDa glycosylated form, whereas both glycosylated and nonglycosylated IL-lra is present in neutrophil lysates (Malyak et al., 1993b). Neutrophils synthesize l%or less the amount of IL-lra per cell in comparison to monocytes or macrophages (Table V). Large numbers of neutrophils present in inflammatory lesions may be a major local source of this protein. This conclusion is supported by the results of recent studies examining IL-lra levels in various organs after intravenous (IV) or intratracheal (IT) administration of LPS in rats (Ulich et al., 1992). After IV injection, IL-lra mRNA was observed to
190
WILLIAM P. AREND TABLE V INTERLEUKIN-Ira PRODUCTION BY DIFFERENT CELLS' IL-lra
Cell Monocyte Macrophage (alveolar) Neutrophil Keratinocyte
Stimulus
None
Adherent IgC
GM-CSF
None Adherent IgG GM-CSF None GM-CSF
TNF-a None
TNF-a
ng/106 cells
1.6 48.5 8.2 24.0 24.0 34.0 0.08 0.35 0.23 38.4 70.4
% i n lysates
61 14
49 50 50 46
64 69 64 99 99
Total IL-lra production (supernatants and lysates) by different cells, as determined by a specific ELISA, and percentages of the total found in the lysates. Further details can be found in the following publications: monocytes (Arend et al., 1989); alveolar macrophages (Janson et al., 1993);neutrophils (Malyak et 01.. 1993b);keratinocytes (Bigler et al., 1992).
peak at 2 to 4 hours in the lung, liver, and spleen, following peak levels of IL-la and IL-lP mRNAs at 1hour. IL-lra production in uiuo after IV LPS injection in baboons or humans also lags after IL-1, suggesting a possible role for IL-lra as an endogenous negative regulator of IL-1 effects (see later text). After IT administration of IL-lra in rats, however, IL-lra mRNA levels in whole lung peaked at 6 hours coincident with a massive influx of neutrophils (Ulich et al., 1992). Furthermore, IL-lra mRNA was present in large amounts in neutrophils obtained by bronchoalveolar lavage from rats injected intratracheally with LPS. These results indicate that neutrophils may be a major source of IL-lra in lungs after local stimulation with LPS. 7. Kera tinocytes The structural variant of IL-lra produced by human keratinocytes and other epithelial cells is described in Section 111. In addition, icIL-lra also may be produced by alveolar macrophages and by both dermal and synovial fibroblasts. The characteristics of keratinocyte production of icIL-lra and its presence in the skin of psoriasis patients have been described. Cultured human keratinocytes constitutively transcribed icIL-lra mRNA and produced large amounts of icIL-lra
INTERLEUKIN-1 RECEPTOR ANTAGONlST
191
protein that remained almost entirely intracellular (Gniaz-Chatellard et al., 1991; Bigler et al., 1992). The intracellular variant of IL-lra in keratinocytes was not glycosylated, and the small amounts of icIL-lra variably found in the supernatants of cultured keratinocytes also were not glycosylated. Thus, keratinocytes did not appear to actively secrete icIL-lra. A variety of cytokines were examined for their ability to enhance keratinocyte production of icIL-lra; only TNF-(Uhad this ability and it increased IL-la production as well (Kutsch et al., 1993). In addition, culture of keratinocytes on substrates of various connective tissue proteins did not affect the baseline production of 1L-lra and IL-la or alter the enhanced production by TNF-a stimulated cells. Differentiation of keratinocytes by culture in high concentrations of Ca2+ (1.0 mM or higher) led to enhanced levels of spontaneous IL-lra synthesis (Gruaz-Chatellard et al., 1991; Bigler et al., 1992).Intracellular IL-lra purified from keratinocytes by affinity chromatography exhibited a level of biological activity equivalent to that of recombinant sIL-lra in inhibition of IL-1 augmentation of PHA-induced murine thymocyte proliferation (Bigler et al., 1992). In addition, recombinant icIL-lra inhibited various effects of IL-1 on human endothelial cells including production of IL-6, IL-8, monocyte chemotactic protein, and adhesion molecules (Bertini et al., 1992). Furthermore, stimulants of endothelial cell functions such as LPS, IL-1, TNF-a, and GM-CSF failed to induce any detectable IL-lra mRNA. Thus, the endothelial cell is one of the few cells capable of making IL-1 that does not appear to produce IL-lra. The presence of icIL-lra mRNA and protein in intact skin in uivo has been evaluated extensively in published studies (Hammerberg et al., 1992). Extracts of keratome sections of normal skin and of intact or lesioned skin from psoriasis patients were chromatographically fractionated. IL-la and IL-Ira proteins were detected by ELISA in overlapping peaks of extracts from all three types of skin. The ratios of IL-Ira to IL-la proteins were 123 in normal skin, 60.5 in uninvolved psoriatic skin, and 1076 in involved psoriatic skin. The absolute levels of IL-lra were 30% reduced in psoriatic skin lesions compared with uninvolved skin (Hammerberg et al., 1992), a finding confirmed b y immunofluorescent studies (Kristensen et al., 1992). IL-la levels were, however, decreased 10- to 20-fold in psoriatic skin lesions, confirming the observations of many laboratories. PCR amplification of reverse-transcribed mRNA from skin cytosols showed the presence of low levels of both sIL-lra and icIL-lra mRNAs in normal skin. Psoriatic skin lesions, however, exhibited 3- to 4-fold increases in each mRNA, the ratio of icIL-lra mRNA to sIL-lra mRNA remaining =lo
192
WILLIAM P. AREND
(Hammerberg et al., 1992). The origin of the sIL-lra mRNA in normal and psoriatic skin was not established by these studies but may have been infiltrating macrophages. IL-lra protein (both structural variants) was localized in skin biopsy specimens using immunofluorescence techniques. The IL-lra protein was concentrated in the stratum granulosum of normal skin, consistent with the observation from in vitro studies of enhanced production by differentiated keratinocytes. In contrast, staining for IL-lra protein in psoriatic skin lesions was most intense in the basal-misbasal layers: reflecting the abnormal pattern of differentiation exhibited by psoriatic keratinocytes in uiuo. In summary, icIL-lra is a major intracellular protein of keratinocytes both in uitro and in uiuo, with differentiation leading to enhanced levels of production. Both icIL-lra and IL-la are produced constitutively by keratinocytes, with T N F-a! stimulating further transcription and translation of both proteins. The meaning of these observations for epidermal cell function in uiuo remains unclear. The possibility exists that icIL-lra plays some unique biological role inside cells. The presence of excessive amounts of IL-lra over IL-la in keratinocytes indicates that a net anti-inflammatory effect would result when these proteins are released from dead cells in the stratum corneum. In any case, the skin is a major reservoir of both IL-la and icIL-lra.
VI. Receptor Binding of Interleukin-1ra
Interleukin-lra was initially characterized as a receptor antagonist by its ability to inhibit IL-1 binding to cell surface receptors without inducing any detectable intracellular agonist effects. The absence of agonist effects of IL-lra in uitro has been observed with multiple cells: murine thymocytes (proliferation, receptor internalization, and phosphorylation of EGF receptor); human fibroblasts ( PGEz production); rabbit chondrocytes (PGE2 and collagenase production); human monocytes, macrophages, and neutrophils (cytokine production); human fibroblasts (expression of the immediate early response genes forfosljun and of the late genes for procollagenase and prostromelysin); and human endothelial cells (cytokine production and expression of adhesion molecules). The existence of a naturally occurring receptor antagonist of a cytokine or hormone-like molecule is, however, singularly unique in biology. The possibility should remain open that an agonist effect of this molecule may yet be described. In various high concentrations in certain assays IL-lra preparations may exhibit slight apparent agonist effects. It remains unclear whether
INTERLEUKIN-1 RECEPTOR ANTAGONIST
193
these effects are not induced by trace contaminants or whether they represent an inhibition of autocrine IL-1 effects. A. INTERLEUKIN-1ra BINDINGTO TYPE1 INTERLEUKIN-1 RECEPTORS Extensive studies have been carried out on IL-lra binding to the 80-kDa type I IL-1 receptor (IL-lRI), the type found on T cells, endothelial cells, fibroblasts, and chondrocytes. Human IL-lra binds to IL-1RIs on murine EL4-6.1 thymoma cells with an affinity equal to that of IL-la and IL-1P, K D of = 150 pM from equilibrium studies (Dripps et al., 1991a). A Scatchard plot of direct IL-lra binding to these cells exhibited a straight line, indicating the presence of only one class of binding sites of = 3300/cell. Interestingly, the affinity of IL-lra for EL4 cells increased threefold with a decrease in pH from 8.5 to 6.0, whereas IL-la and IL-lP binding did not show similar changes. The association rate constant for IL-lra binding to EL4 cells was 4.3 x lo7 M-' min-', a value 5-fold lower than that of IL-la. The dissociation rate constant for IL-lra binding was = 1.25 x min-', a value 20-fold lower than that of IL-la. The K D calculated from the ratio of the dissociation and association rate constants was similar to the K D obtained from the equilibrium binding studies. The meaning of the slower receptor association and dissociation of IL-lra in comparison to IL-la and IL-1P is not clear, although these differences could be due to technical considerations. The results of these binding studies indicate that an IL-lra molecule dissociating from an IL-1 receptor has a higher probability of leaving the cell surface than rebinding to a nearby vacant receptor (Dripps et al., 1991a). Human IL-lra binds to IL-1RIs on cultured rheumatoid synovial fibroblast-like cells with kinetics similar to those observed with T cells (Arend and Coll, 1991). IL-lra bound to these cells in adherent culture with a K D of 213 pM and a Ki of 134 pM, values similar to those of IL-la binding. IL-Ira and IL-la appeared to bind to the same receptors on the synovial cells and not to overlapping subsets of receptors, as the binding of each ligand could be 100% inhibited by high concentrations of the other. The results of additional crossinhibition studies indicated, however, that although IL-la and IL-lra bound to IL-1RIs on synovial cells with the same avidity, they did not bind identically. The hypothesis was raised that these two ligands may bind to different regions of the extracellular portion of the IL-1R1, yet sterically inhibit binding of each other. The possibilities of how IL-lra may bind to IL-1RIs and not activate cells are discussed later.
194
WILLIAM P. AREND
B. INTERLEUKIN-1ra BINDING TO TYPE11 INTERLEUKIN-1 RECEPTORS The binding of IL-lra to the 68-kDa type I1 IL-1 receptors (IL1RIIs) on B cells, neutrophils, or monocytes has been more controversial. Initial studies indicated that human IL-lra did not inhibit at all or blocked very weakly '251-IL,-la binding to IL-1RIIs on the murine pre-B cell line 70Z/3 (Hannum et al., 1990; McMahan et al., 1991; Stylianou et al., 1992); however, IL-lra inhibited IL-lainduced NFKB transcription in these cells, indicating the presence of a small number of functionally active IL-1RIs (Stylianou et al., 1992). In addition, human IL-lra bound to recombinant murine IL-1RIIs with an avidity equal to that of IL-la, indicating that these receptors in the intact cell were sterically not available to IL-lra but they were to IL-la (McMahan et al., 1991). This finding suggests that IL-lra binding to IL-lRIIs, similar to IL-lRIs, may be to a different region of the extracellular portion of the receptor than interacts with IL-la. Furthermore, IL-lra bound to human IL-1RIIs with an affinity equal to that of IL-la, including cell surface IL-1RIIs on the human B cell line CB23 and recombinant IL-1RIIs in solution. In contrast to ILlRIs, human IL-1RIIs bound IL-1P with an avidity 10-fold greater than that of IL-la and IL-lra, which exhibited equivalent binding to each other (McMahan et al., 1991). The binding of human IL-lra to human IL-1RIIs has been s.ubstantiated from studies on neutrophils, monocytes, and Raji B lymphoma cells. IL-lra competitively inhibited the binding of '251-IL-lP or 1251IL-la to IL-1RIIs on human neutrophils or Raji cells (Granowitz et al., 1991a). In these studies, however, IL-lra binding to IL-1RIIs appeared to be less avid than that of either IL-la or IL-1P. In addition, chemical crosslinking studies confirmed that IL-lra bound to IL-1RIIs on these cells. Similar results were obtained with the same cells by other investigators. IL-lra bound to IL-1RIIs on human B cells with a KO of 15 nM and to neutrophils with a K D of 8 mM, both equal to the binding of IL-la but = 18fold lower than that of IL-lP (Dripps et al., 1991b). These studies also showed that human, mouse, and rat IL-lra all bound extremely weakly to IL-1RIIs on the murine pre-B cell line 70Z/3, suggesting that murine IL-1RIIs on the intact cells were sterically inaccessible to IL-lra of different species. It should be noted that the avidity of IL-lra binding to human IL-1RIIs is z100-fold lower than binding to IL-1RIs. Thus, when both IL-1 receptors are present on the same cell, as seems to be the case with B cells, neutrophils, and monocytes where small numbers of IL-1RIs
INTERLEUKIN-1 RECEPTOR ANTAGONIST
195
are present, IL-lra will preferentially associate with IL-1RIs over IL-IRIIs. This difference may be biologically relevant as only ILlRIs appear to be capable of transducing an intracellular signal. ILlRIIs possess only a short cytoplasmic domain and may be functionally inert. Thus, the biological role of cell surface IL-1RIIs remains unclear. IL-lra also blocks IL-1 receptor-mediated effects on human monocytes, inhibiting IL-1-induced production of IL-1, IL-6, or TNF-a! in these cells (Granowitz et al., 1992a). Even though ‘251-IL-lfi chemically crosslinked to the 68-kDa IL-1RII on the monocytes, it is likely that the biological response to IL-1 was mediated by small numbers of higher-affinity IL-1RIs. These investigators also observed that addition of IL-lra to the monocytes up to 8 hours after stimulation with IL-lP still resulted in some inhibition of responses. This surprising result suggested either that IL-1 receptors were recycling on these cells or that a subpopulation of cells expressed IL-1 receptors newly acquired during culture.
c. INTERLEUKIN-1ra BINDINGTO SOLUBLE
RECEPTORS INTERLEUKIN-1 The cDNAs have been cloned for murine and human 80-kDa ILlRIs and 68-kDa IL-1RIIs with some interesting structural differences in the proteins determined. Mature human IL-1RI is a 552amino-acid polypeptide with extracellular, transmembrane, and cytoplasmic domains of 319,20, and 213 residues, respectively (Sims et ul., 1989). In contrast, the mature human IL-1RII is a 386-amino-acid protein with extracellular, transmembrane, and cytoplasmic domains of 332,26, and 29 residues, respectively (McMahan et al., 1991). Both types of IL-1 receptors are members of the Ig superfamily possessing three Ig-like domains in the extracellular portions. Despite all three IL-1 ligands binding to both types of IL-1 receptors, the extracellular portions of the two receptors exhibited only 28% amino acid sequence homology (McMahan et al., 1991). The areas of greatest conservation in the extracellular portions of these two IL-1 receptors are near the C terminus of domain 2 and the N terminus of domain 3. These areas may represent contact points for ligand binding, as discussed later. The most important structural difference between the two IL-1 receptors, however, is in the intracytoplasmic portions where IL-1RII possesses only a short 29-amino-acid structure. Thus, it appears that IL-1RII is not internalized after ligand binding and does not induce intracellular signals.
196
WILLIAM P. AREND
A soluble murine IL-1RI containing only 316 amino acids of the extracellular portion was prepared by expressing a constructed cDNA in HeLa cells (Dower et al., 1989). The most N-terminal residues of the extracellular portion of the intact IL-1RI are missing from this truncated protein. The soluble IL-1RI is 60-kDa but contains considerable carbohydrate, as N-glycanase digestion reduced the molecular weight to 34 kDa (Dower et al., 1989). The mouse soluble IL-1RI bound IL-la with avidity equal to that of the intact cell surface IL-1RI. Although the soluble IL-1RI has not heretofore been described to occur naturally, small amounts may be present in normal human serum. The interaction between IL-lra and a 60-kDa serum binding factor was inhibited by rabbit antibodies specific for the truncated soluble form of human IL-1RI (Svenson et al., 1993). Surprisingly, this purified human serum binding factor interacted with high avidity with IL-lra (KD of 70 pM), but exhibited no binding of IL-la or IL-1p. In addition, a human version of the recombinant soluble IL1RI bound IL-lra 200-fold more avidly than to IL-la or IL-1p (Svenson et al., 1993), differing from the cell surface IL-1RI that binds all three forms of IL-1 with near-equal avidities. The possibility exists that variations in the C-terminal structure of soluble IL-1RI lead to differences in relative binding of IL-la, IL-lp, and IL-lra. In any case, these results suggest that samll amounts of soluble IL-1RI may be present in normal serum and selectively bind IL-lra. Whether this protein-bound IL-lra is still capable of interacting with cell surface receptors has not been determined. Similar to soluble receptors for other cytokines, however, soluble IL-1RI may serve to maintain ILIra in circulation and possibly deliver it to tissues. Soluble IL-1RIIs have been described in normal human plasma (Eastgate et al., 1990), synovial fluids (Symons et al., 1990), supernatants of PHA-stimulated human mononuclear cells (Symons et al., 1990), and supernatants of the Raji human B cell line (Symons and Duff, 1990; Giri et al., 1990). A serine protease inhibitor prevented release of the soluble IL-1RII from Raji cells, suggesting that it may be proteolytically cleaved from the cell surface (Symons and Duff, 1990). Purification of this binding factor showed it to be a 47-kDa protein that bound IL-1p with the same avidity as cell surface ILlRII ( K D = 2 nM) (Symons et al., 1991); however, the soluble ILlRII failed to bind IL-la or IL-lra at all, differing from the cell surface IL-1RII that binds IL-la! and IL-lra weakly. The soluble ILlRII inhibited binding of IL-lP to the EL-4 NOB.l T cell line and to the Raji B cell line, but not binding of IL-la to the T cell line (Sy-
197
INTERLEUKIN-1 RECEPTOR ANTAGONIST
mons et al., 1991). Thus, the soluble form of IL-1RII found in human plasma and body fluids may function as a selective inhibitor of IL-1p in uivo. Binding of IL-la, IL-1P, and IL-lra to various forms of human 1L-1 receptors is summarized in Table VI. These findings may be relevant to regulation of the bioavailability and effects of these cytokines produced in uivo. Soluble IL-1RI is presently being evaluated as a therapeutic agent in human autoimmune and inflammatory diseases.
D. HOW DOESINTERLEUKIN-1ra FUNCTION AS A RECEPTORANTAGONIST? Extensive studies have been carried out by a number of laboratories to decipher the structural aspects of receptor binding and cell activation of IL-la and IL-1p. Although these questions have not been answered definitively, some general principles have emerged. There may exist three binding sites on the surface of IL-1P that interact separately with the three lg domains on the extracellular portion of the IL-1 receptor (Clore et al., 1991). Furthermore, it is likely that the contact points for receptor binding may differ somewhat between IL-la, IL-lP, and IL-lra (Labriola-Tompkins et al., 1991). In addition, different sites may be involved in each ligand for binding to IL-1RIs and IL-1RIIs. Two major categories of IL-1 function in vivo are the effects on immune and inflammatory cells. Studies on IL-lp peptides have been able to separate the regions of the molecule responsible for mediating these functions. A synthetic nonapeptide corresponding to fragment 163-171 of human pro-IL-lp demonstrated potent immunomodulatory properties in uivo in the absence of inducing metabolic changes associated with inflammation (Boraschi et al., 1988). Furthermore, a monoclonal antibody to this region of IL-lp TABLE VI BINDING OF INTERLEUKIN-1 PROTEINS TO DIFFERENT INTERLEUKIN-1 RECEPTORS' Receptor
IL-la
IL-lp
IL-lra
Intact IL-1RI Soluble IL-1RI Intact IL-1RII Soluble IL-1RII
+++
+++
+++ +++ +
k
+
-
r
+++ +++
-
' Relative binding of IL-la, IL-lP, and IL-lra to intact cell surface and soluble (extracellular domain) forms of IL-1RI and IL-IRII. See text for details and references.
198
WILLIAM P. AREND
selectively inhibited the immunostinlulatory effects of IL-1p in viuo in mice without altering pyrogenic activity (Boraschi et al., 1989). A peptide corresponding to residues 208 to 240 of human pro-IL-lp appeared to mediate the pyrogenic and inflammatory effects of the molecule (Opp et al., 1992). Thus, the possibility seems likely that different regions of IL-1P are responsible for binding to IL-lRIs, binding to IL-lRIIs, inducing inflammatory events, and mediating effects on cells of the immune system. The structural aspects of receptor binding of IL-lra have been examined in an effort to understand the reason for an apparent absence of agonist effects. A point mutation of Arg to Gly at position 11 in mature IL-1p (127 in the pro-IL-lp) reduced the bioactivity of this molecule 100-fold toward T cells while only modestly reducing receptor binding by 25% (Gehrke et al., 1990). This residue in IL-lra is a tryptophan. In addition, mutating Asp to Lys at position 145 in mature IL-1p led to a 90% reduction in biological activity toward T cells without any alteration in the affinity of receptor binding ( J u et al., 1991). Rather than indicating the presence of active sites, these point mutations may be affecting IL-2 activities by altering conformation. IL-la and IL-1p both consist of antiparallel /%pleated sheets with the additional presence of large loops (Wilder et al., 1992). The secondary and tertiary structures of IL-lra are similar to those of IL-lp, but different regions of the primary sequence appear to constitute the p strands (Stockman et al., 1992). In addition, IL-lra and IL-1p differ in the locations of numerous positively charged side chains, regions possibly important in receptor binding and activation. Additional studies on point mutants of IL-1 have generated a working hypothesis regarding the structiiral aspects of receptor binding and cell activation. The IL-1p Arg-to-Gly mutein at position 11 stimulates immediate early gene expression in fibroblasts (i.e., fos and j i n ) without inducing late gene expression (procollagenase and prostromelysin) (Conca et aZ., 1991). IL-lra binds to IL-1RI equally well but fails to trigger either level of cell response. Thus, induction of late transcription events appears to involve ligand-receptor interactions that are present in intact IL-1p but absent in the IL-lp 11 mutein or in IL-lra. Although the Arg-to-Gly mutation at position 11 of IL-1P leads to a loss of biological activities, substitution of other amino acids at the same site does not produce a similar change in function (Auron et al., 1992). The results of molecular modeling analysis indicate that the Arg-to-Gly mutation resulted in a disruption of the structural integrity of the antiparallel p strand 1/12 pair. Collapse of P strand 1 into a hydrated space between strands 1, 2, and 4 ap-
INTERLEUKIN-1 RECEPTOR ANTAGONIST
199
peared to structurally alter a cleft in IL-1p that contained a cluster of conserved amino acids. One of these amino acids now made inaccessible was Asp 145; thus, this residue was thought to represent a late trigger for IL-1 biological activity. The early trigger for other IL-1 biological activities was hypothesized to reside in a p-bulge domain located between strands 4 and 5 of intact IL-1p and of IL-1@ 11 mutein (Auron et al., 1992). Most importantly, neither of these structural features thought to be involved in mediating IL-lP activities are present in IL-lra. In fact, mutation of Lys to Asp at position 145 in IL-lra led to the acquisition of weak agonist activities toward both T cells and fibroblasts ( J u et al., 1991). The structure-function relationships of IL-1@and IL-lra are summarized in Fig. 4. A common receptor binding epitope for IL-1p and IL-Ira may be formed by the charged surface of the rim of a @-pleated
IL-1 Receptor
IL-1 p
Closed End p - Sheet "Cap"
Open End p - Barrel "Face"
Three lg-Superfamily Domains
FIG.4. General conceptual model for an IL-l-receptor interaction, suggesting possible combined features of type I and type I1 receptors. The open end of the IL-1 barrel, which exhibits a threefold symmetry, is shown interacting with a putative threefold IL-1 extracellular receptor domain (soluble IL-1 receptor). The Rl20 and D261 residues of precursor IL-1p (4 and 145 in mature IL-lp) that are postulated to play a role respectively in ligand binding and activity triggering are indicated, as are the hydrophobic FIG2 and F2M (46 and 150 in mature IL-lp) (indicated by the hexagonal plates). These hydrophobic residues are located at the IL-1 surface in a location that would position them at the ligand-receptor interface. The putative charged residues shown on the IL-1 receptor may interact with the Rlzo and residues of precursor IL-lP (4 and 145 in mature IL-lp), leading to stimulation of late gene expression in target cells. The P-bulge region, possibly responsible for early gene induction, is shown as a shaded area on the face ofthe IL-1p molecule. IL-lra is capable of binding to the IL-1 receptor, but lacks these latter two structural features necessary for induction of a biological response in target cells after receptor binding. [Reprinted from Auron et al. (1992) with permission. Copyright ( 1992) American Chemical Society.]
200
WILLIAM P. AREND
band structure plus a hydrophobic core represented by two phenylalanine side chains at positions 46 and 150 of mature IL-lj3 (LabriolaTompkins et al., 1991); however, IL-lra lacks two additional structural features thought to be important in mediating the biological activities of IL-lP: a @-bulgebetween strands 4 and 5 (site for early response gene induction) and a region around aspartic acid at residue 145 (site for stimulation of late gene expression). As IL-la does not possess these same structures as IL-lP, receptor binding and activation by IL-la may involve different structural determinants. VII. In Witro and in Wivo Effects of Interleukin-lra
Since the cloning and expression of IL-lra were described in early 1990, numerous laboratories have examined the effects of this recombinant protein in a variety of biological systems. These studies have further substantiated the initial impression that IL-lra lacks any direct positive agonist effects. It has been somewhat more difficult, however, to prove conclusively that IL-lra does not induce a negative, or inhibitory, intracellular signal in some target cells. What may be interpreted as a negative signal generated by IL-lra may actually be the receptor blocking effects against IL-1 of paracrine or autocrine origin in mixed cell populations (see below). The remainder of this article summarizes the in uitro and i n uiuo effects of IL-lra by organ system or disease. A few general principles of IL-lra effects should be emphasized before a discussion of observations from studies with specific tissues or organs.
A. RATIO OF INTERLEUKIN-1raTO INTERLEUKIN-1 A great molar excess of IL-lra over IL-1 is necessary to inhibit biological responses to IL-1 in uitro or in uiuo. This general principle was first recognized in early 1990 and has been confirmed by multiple investigators over the past 3 years working in different experimental systems. The initial observations were that amounts of IL-lra 5 or 40 times greater than those of IL-l@or IL-la, respectively, were required to yield 50% inhibition of the augmenting effects of IL-1 in the murine thymocyte assay (Arend et al., 1990). In addition, 50% inhibition of the stimulatory effects of IL-la or IL-lP on PGEz production by cultured rheumatoid synovial cells or rabbit chondrocytes, and on collagenase production by the synovial cells, required up to 100-fold excess amounts of IL-lra (Table VII). These observations are explained by the “spare receptor effect.” Target cells may express up to 2000 or more IL-1 receptors per cell; however, only a
201
INTERLEUKIN-I RECEPTOR ANTAGONIST
TABLE VII INHIBITION OF INTERLEUKIN-~-~NDUCED PROSTAGLANDIN PRODUCTION BY INTERLEUKIN-1Ta
Cell Human synovial cells" Rabbit articular chondrocytesa
Inducing protein IL-la IL-lP IL-la IL-lP
50% inhibition (IL-ha, pg/ml)b 1,600
2,800
5,700 15,500
Excess over IL-1 (-fold)
29 14 102 75
" Cultured human rheumatoid synovial cells or rabbit articular chondrocytes were incubated for 16 hours with 3 U/ml IL-l(56 pglml IL-la or 207 pg/ml IL-16) and serially increasing concentrations of recombinant IL-Ira. 'The concentrations of IL-lra giving 50% inhibition of IL-1-induced PGEp production, as measured in cell supernatants using a specific ELISA, were determined. For further details, see the original paper (Arend et al., 1990).
few receptors per cell need to be occupied by IL-1 to induce a full biological response. Thus, target cells appear to be uniformly sensitive to very small amounts of IL-1. Even though IL-lra binds to ILlRIs with avidity equal to that of IL-1, the presence of excess receptors requires that IL-lra flood the system to block occupancy of only a few receptors by IL-1. Hypotheses on the possible in vivo relevance of IL-lra must always take into consideration this requirement for excess amounts of IL-lra over IL-1.
B. AUTOCRINEOR PARACRINE EFFECTSOF INTERLEUKIN-1 A variety of exogenous cell stimuli may result in the local production of IL-la or IL-1P that, in turn, act on the producing cells to induce further responses. Thus, IL-lra may appear to be inhibiting the responses to the initial stimulus but actually may only be blocking the receptor binding of the locally produced IL-1. An example of the importance of this mechanism is seen in the effects of LPS on monocytes. Pretreatment of these cells with IL-lra reduced LPSinduced cytokine synthesis: IL-la by 33%,IL-1P by 43%,and TNF-a by 20% (Granowitz et al., 1992~).Furthermore, this effect of IL-lra was not limited to LPS stimulation of monocytes. IL-lra also reduced IL-1P synthesized in response to IL-2 by 44%,toxic shock syndrome toxin-1 by 26%, and PMA by 76%. These results indicate that IL-1P produced in monocytes in response to a variety of stimulatory agents may induce further IL-1 in a delayed fashion. The effects of this autocrine or paracrine IL-1 are inhibited by exogenous or endogenous
202
WILLIAM P. AREND
IL-lra. It should be remembered that LPS induces both IL-1 and IL-lra, so that endogenously-produced IL-lra may limit the autocrine and paracrine effects of IL-1. In a similar fashion, IL-lra was found to inhibit IL-8 production stimulated by LPS in purified human neutrophils or mononuclear cells (DeForge et al., 1992) or in PBMCs induced by IL-lp, LPS, or live Borrelia burgdorferi organisms (Porat et aZ., 1992). Other investigators also have recently described the apparent inhibitory effects of IL-lra on LPS-induced IL-1 (Conti et al., 1992a,c),including inhibition of LPS-induced PGE2 and leukotriene B4 production by human monocytes (Conti et al., 1992b). Thus, the autocrine or paracrine effects of IL-1 may be important in amplifying many cell responses, and IL-lra may dampen or limit these effects. VIII. Interleukin-lra Effects on the Immune System
IL-1, either soluble or membrane-bound, is thought by many investigators to represent a second signal for inducing T cell proliferation during antigen presentation; however, other cytokines may also augment T cell growth and the prime importance of IL-1 has been questioned. The possible effects of 1L-lra on the normal response of the immune system pose an important question facing the potential therapeutic administration of this protein in human diseases. A. IMMUNE CELLFUNCTION in Vitro IL-lra clearly inhibited IL-l-induced T cell proliferation, using either soluble IL-1 (Arend et al., 1990) or membrane-associated IL-1 (Seckinger et al., 1990b). Furthermore, IL-lra inhibited the proliferation of murine TH2 clones in response to antigen presented by splenocytes or macrophages, whereas the proliferation of TH1 clones was not affected (Chang et aZ., 1990). These results indicated that endogenous IL-1 production by the antigen-presenting cell was a necessary factor in TH2 cell proliferation. In contrast, IL-lra failed to inhibit human T cell proliferation induced by mitogens, soluble antigens, or allogeneic determinants, although IL-lra did inhibit augmented responses induced in these in vitro systems by exogenously added IL-1 (Nicod et al., 1992). These results suggested that the endogenous production of IL-1 was not a necessary cofactor for human T cell proliferation, different from the results with the murine TH2 clones; however, the possibility cannot be eliminated that IL-lra could not effectively penetrate the interface between the antigen-presenting
INTERLEUKIN-1 RECEPTOR ANTAGONIST
203
cell and the T cell in the human cell experiments to inhibit effectively the local effects of soluble or membrane-associated IL-1. Additional studies showed that IL-lra reduced IL-2-induced TNF mRNA expression in human lymphokine-activated killer (LAK) cells and decreased the lytic activity of these cells (Fujiwara and Grimm, 1992). These results suggested the role of autocrine production of IL-1 in LAK cell induction by IL-2, although somewhat different results were obtained by another investigator (Conti et al., 1991a). In other studies, 1L-lra reduced concanavalin A-induced blastogenesis of human lymphocytes (Conti et al., 1991c), and PHA-induced DNA synthesis in human lymphocytes (Conti et al., 1991b), again suggesting a role for autocrine IL-1. Lastly, IL-lra inhibited the costimulation of lg synthesis by B cells induced by T cells with added IL-1 (Tucci et al., 1992). Futhermore, IL-lra also inhibited human B cell responsiveness induced by PMApreactivated murine EL-4 T cells in the absence of added IL-1. As the B cells did not acquire IL-1-dependent TNF-a responsiveness, these investigators argued that IL-1 was not produced by the preactivated T cells and that IL-lra may provide a direct or indirect inhibitory signal interfering with IL-1-independent B cell activation. The nature of such an IL-lra-induced signal was not, however, explored and this possibility must remain hypothetical.
B. IMMUNE RESPONSES in Viuo A variety of antigen-specific immune responses were not inhibited by the in uivo administration of human IL-lra to mice (Faherty et al., 1992). These T cell responses included the cytolytic T cell response to allogeneic splenocytes and the delayed type hypersensitivity response to oxazolone. The B cell responses that also were refractory to IL-ha included the splenic plaque response to sheep red blood cells and the IgG or IgM response to trinitrophenyl-keyhole limpet hemocyanin. Twice-daily administration of IL-lra or the monoclonal antibody 35F5 inhibitory to IL-1RI function or continuous infusion of IL-lra in uiuo did not reduce the T or B cell responses. It was concluded that IL-1 may not be required for these antigen-induced B and T cell responses in uiuo. Alternatively, as both IL-lra and the monoclonal antibody 35F5 block only IL-1RIs and not IL-lRIIs, the possibility remains that murine THI cells could have provided sufficient help for the examined immune responses through a small number of IL-lRIIs, if these are functional. IL-lra also did not inhibit steady-state hematopoiesis in mice during a 14-day continuous infu-
204
WILLIAM P. .4REND
sion (Fibbe et al., 1992).Additional studies are necessary to examine the possible short- and long-term effects of human IL-lra on human T and B cell responses in uiuo. IX. Interleukin-1ra in Infection and Sepsis in Vivo
The effects of endotoxin (LPS) on the in uitro production of IL-1 and IL-lra by human monocytes were reviewed in Sections V,A. and V,C. LPS induces the transcription and translation of both cytokines, with the kinetics of appearance of mRNA and protein for IL-1P usually preceding those of IL-lra by a short time interval. In addition, IL-1 itself is a weak inducer of I L - h a production by monocytes. Thus, it might be hypothesized that IL-lra production occurs as part of a physiological regulatory response to limit the proinflammatory effects of IL-1. Multiple other cytokines that may be present in diseased tissues in uiuo may influence the relative production of IL-1 and IL- Ira. Whether an overall proinflammatory or anti-inflammatory environment is present in a diseased tissue will depend on many factors, including the local ratio of IL-1 to IL-lra; the production of other cytokines such as TGF-P, GM-CSF, IL-4, and IL-10; the presence of other inflammatory molecules such as prostaglandins, oxygen radicals, and nitric oxide; and the local release of other modulating molecules by infiltrating inflammatory cells. A two-phase approach is necessary to understand the biological role of IL-lra in uiuo in these complex situations. The first phase is to determine the amounts and type of IL-lra (and of IL-1) endogenously produced in diseased tissues. The second phase is to examine the effects of administered IL-lra on various parameters of disease activity. The remainder of this article summarizes published information on these two phases of the involvement of IL-lra in experimental animal or human diseases.
A. INTERLEUKIN-Ira IN ANIMALMODELSO F SEPSIS Considerable data exist to indicate that both IL-1 and TNF-a are produced in large amounts in experimental animal models of infection and sepsis; however, the relative importance of either molecule in subsequent pathophysiological events has remained unclear. In studies on a rabbit model of meningitis, the intracisternal administration of rabbit TNF-a or IL-1P led to significant levels of cerebrospinal fluid pleocytosis and inflammation (Ramilo et al., 1990). Inhibition of meningeal inflammation as a result of intracisternal injection of LPS was observed when monoclonal murine antibodies to TNF-a
INTERLEUKIN-1 RECEPTOR ANTAGONIST
205
or a polyclonal rabbit antiserum to IL-1j3 was administered together with the LPS. The most marked inhibition of inflammation occurred when antibodies to both TNF-a and IL-lp were given with the LPS. Thus, the endogenous production of both TNF-a and IL-lj3 is implicated in this model of LPS-induced meningitis. Furthermore, the coadministration of a great excess of recombinant human IL-lra over rabbit IL-lj3 also led to a marked reduction in cerebrospinal fluid pleocytosis (Ramilo et al., 1990).In other studies, IL-lra blocked local accumulation of PMNs after intraperitoneal injection of IL-lj3 or LPS in mice (McIntyre et al., 1991). Additional in uiuo effects of IL-1 that were inhibited by subcutaneous injection of IL-lra in these studies included egress of PMNs from the bone marrow and increased serum levels of hepatic acute phase proteins, IL-6, or corticosterone. IL-lra also inhibited the induction of colony-stimulating factor (CSF) and of early endotoxin tolerance in LPS-treated mice (Henricson et al., 1991).A moderate single dose of IL-lra, 0.5 mg/kg given intraperitoneally, only partially suppressed LPS-induced fever in rats and did not reduce serum levels of IL-6 (Smith and Kluger, 1992). Thus, ILIra is able to inhibit numerous markers and processes of inflammation induced in mice or rats by the in uiuo administration of LPS or IL-1, but a constant infusion or repeated doses of IL-lra are more effective. IL-lra also is effective in both blocking physiological abnormalities and reducing mortality in animal models of sepsis. IL-lra blocked LPS-induced shock in rabbits in a dose-responsive fashion; a 90% reduction in mortality was seen with the highest dose of ILIra, a total of 100 mg/kg body weight given intravenously in injections every 2 hours over 24 hours (Ohlsson et al., 1990). IL-lra was effective when administered at the same time as the lethal dose of LPS or 2 hours after. IL-Ira also reduced physiological abnormalities and death in rabbits infused with E. coli organisms (Wakabayashi et al., 1991). In these studies, IL-lra largely reversed the hypotension, decrease in cardiac output, and increase in systemic vascular resistance observed in this animal model of septic shock. In recent studies, IL-lra was shown to inhibit three potentially toxic cardiovascular ef'fects of IL-lj3 in rabbits: a rapid prostaglandin synthesis, a slower nitric oxide synthesis, and an enhanced responsiveness to kinins that are agonists for j31 receptors (Petitclerc et al., 1992). Mice receiving a lethal injection of E. coli LPS also exhibited a marked improvement in survival when treated with IL-lra every 4 hours for 24 hours (Alexander et ul., 1991). Interestingly, in both rabbits (Wakabayashi et al., 1991) and mice (Alexander et al., 1991), serum levels of TNF-a and
206
WILLIAM P. AREND
IL-1p proteins were not changed by IL-lra administration. Lastly, IL-lra markedly reduced hemodynamic abnormalities and improved survival from 20 to 71% in a rat model of septic shock induced b y ligation and puncture of the cecum (Alexander et al., 1992). Studies in primates have revealed a similar beneficial effect of ILIra on the consequences of sepsis. Baboons receiving a sublethal injection of LPS or IL-la exhibited tachycardia, hypotension, decreased serum levels of amino acids, and increased levels of lactic acid and IL-6 (Fischer et al., 1991). Simultaneous administration of IL-la and a 1000-fold excess amount of IL-lra completely reversed these physiological abnormalities. In addition, a baboon injected with IL-lra alone (10 mg/kg body weight) exhibited no changes in any of the hemodynamic, metabolic, or hormonal parameters examined. Subsequent studies showed that the response to IL-lra was influenced by the relative amount of endogenous IL-lp produced (Fischer et al., 1992a). Sublethal endotoxemia in baboons was accompanied by no detectable serum levels of IL-1p; IL-lra exhibited only minimal effects on the modest hemodynamic and metabolic abnormalities. In contrast, lethal E. coli-induced shock was marked by high circulating levels of IL-lp, and a significant ameliorative effect of IL-lra was observed on hemodynamic and metabolic derangements as well as on death. Thus, an exaggerated systemic IL-lp response is characteristic of overwhelming E. coli septic shock in baboons and IL-lra can effectively prevent the injurious consequences. In another animal model of local inflammation or infection, the intratracheal administration of LPS or IL-1p led to an acute inflammatory lung disease in rats (Ulich et al., 1991). The coadministration of IL-lra inhibited the effects of either LPS or IL-1p on inducing PMN influx into the lung. Furthermore, intratracheal injection of LPS induced a progressive increase in IL-lra mRNA expression in the whole lung. This local IL-lra production most likely occurred in infiltrating PMNs (Ulich et al., 1992). In addition, following the intravenous injection of LPS in rats IL-la and IL-1p mRNA expression peaked at 1 hour in the lung, liver, spleen, and bowel; IL-lra mRNA expression in the same organs peaked later at 2 to 4 hours. Thus, LPS induces a widespread cytokine response in these animals and the local administration of IL-lra into the lungs can block the induction of acute inflammatory events in this organ. In addition, muscle proteolysis in rats injected with endotoxin can be partially blunted by repeated treatments with IL-lra (Zamir et al., 1992). Somewhat unusual or paradoxical aspects of IL-lra production or
INTERLEUKIN-1 RECEPTOR ANTAGONIST
207
effects have been observed in some particular types of infection. First, live Borrelia burgdorferi organisms, the pathological agent in human Lyme disease, preferentially induced IL-1P gene expression and protein production over IL-lra in human PBMCs (Miller et al., 1992). This effect of the Borrelia organisms may explain the rather severe inflammatory arthritis that sometimes occurs in patients with Lyme disease. Second, IL-1P may enhance the growth of virulent strains of E. coli; this unusual effect of IL-1 is inhibited by IL-lra, indicating that the organisms express IL-1 receptors (Porat et al., 1991). The in v i m relevance of this observation to the course of infectious diseases remains unclear, although the growth-promoting effect of IL-1P was observed only during the log phase of bacterial growth. Third, IL-1 may play an important role in antibacterial resistance to infection with Listeria monocytogenes and other intracellular organisms. Mice treated with exogenous IL-la exhibited enhanced resistance to Listeria infection, and treatment with monoclonal antibody 35F5, which blocks IL-lRIs, resulted in greatly enhanced bacterial growth in the livers and spleens (Have11 et al., 1992). The effects of IL-Ira in murine listeriosis have not been described. Theoretically, IL-lra administration could lead to possible unwanted consequences in infections with intracellular organisms where IL-1 plays a role in host resistance.
B. INTERLEUKIN-Ira I N HUMANINFECTIONS AND SEPSIS On the basis of extensive experience with IL-lra in experimental animal models of septic shock, studies have commenced on the therapeutic use of IL-lra in patients with sepsis syndrome. These clinical trials have been preceded by studies on IL-lra production in LPSinjected normal volunteers, the presence of circulating IL-lra in patients with severe infections, and the pharmacokinetics of IL-lra administered to normal volunteers. 1 . Circulating Interleukin-lra in Humans IL-1 inhibitory bioactivity was initially described in the plasma of normal volunteers who received a single IV injection of E. coli endotoxin (Spinas et al., 1990). This bioactivity was detected as inhibition of IL-l-induced PGEz production by fibroblasts. The kinetics of IL-1 inhibitory activity in plasma coincided with peak levels of fever and disappeared with decreasing body temperature. Later studies indicated that this IL-1 inhibitory bioactivity was due, at least in part, to the presence of IL-lra. IL-lP and IL-lra levels were measured in plasma by radioimmunoassays over 24 hours following injection of E.
208
WILLIAM P. AREND
coli endotoxin into nine healthy volunteers (Granowitz et ul., 199lb). These individuals experienced transient chills and generalized myalgias after 1 to 2 hours and a mean temperature increase of 1.6"C occurred 4 to 5 hours after endotoxin injection. Plasma IL-1P was first detected at 1hour, reached a maximum of 79 pg/ml at 2 hours, and then decreased over the next 22 hours. In contrast, IL-lra was first detected in plasma at 2 hours after endotoxin injection, reached a maximum of 6400 pg/ml at 3 to 6 hours, and slowly declined thereafter. Thus, after endotoxin injection of normal volunteers, IL-lra levels in plasma appeared after IL-1P but reached a peak almost 100-fold higher. These excess amounts of IL-Ira over IL-lP are certainly sufficient to block the biological activities of IL-1 on diverse target cells (Arend et al., 1990). In addition, elevated levels of IL-lra to 54.3ng/ml were found in the plasma of 12 acutely ill surgical patients with or without accompanying sepsis (Fischer et al., 1992b). A detailed study of IL-lra plasma levels over time in patients with sepsis syndrome has not yet been published.
2. Interleukin-1 ra Administration in Humans A phase I study examined the pharmacokinetics and possible consequences of IV infusion of IL-lra into 25 normal volunteers (Granowitz et d.,1992b). Total doses of between 1 and 10 mg/kg were administered over 3 hours in a continuous infusion, with the plasma levels of IL-lra being 3.1 and 29 pg/ml, respectively, at 3 hours. The plasma levels of IL-lra declined rapidly over the 9 hours after the infusion, with an initial half-life of 2 1minutes and a terminal half-life of 108 minutes. The plasma clearance of IL-lra was 2.0 k 0.3 mg/ min/kg and, most interestingly, less than 3.2% of the dose of IL-lra infused was found in the urine. These pharmacokinetic characteristics suggest that IL-lra is widely distributed in the body after IV administration. Whether this IL-lra is bound to plasma proteins or to IL-1 receptors on tissue cells was not investigated in this study. ILIra administration in humans appeared to be quite safe as no changes were noted in symptoms, physical examination, complete blood counts, mononuclear cell phenotypes, blood chemistry profiles, and serum iron and serum cortisol levels (Granowitz et al., 1992b). Furthermore, these results establish that IL-lra in uiuo appears to have no obvious agonist properties, confirming the results of numerous in uitro studies. The effects of IL-lra in a phase 1/11 study of patients with sepsis syndrome have been described in a preliminary publication (Gordon
INTERLEUKIN-1 RECEPTOR ANTAGONIST
209
et al., 1992). IL-lra was infused IV over 3 days at doses of 17,67, and 133 mg/hr into 25 patients in each group; 25 control patients received a control infusion of saline. The treatment of the sepsis syndrome was otherwise the same in all four groups. The mortality at Day 28 from all causes was 44% in the control group and 32, 25, and 16%, respectively, in the three IL-lra treatment groups. On the basis of these promising preliminary observations, IL-lra is currently being examined in a phase I11 clinical trial in 900 patients with sepsis syndrome. X. Interleukin-1ra in Inflammatory Arthritis
Studies describing IL-lra production by synovial fluid cells and synovial tissue macrophages are reviewed in Section V,C,3. In this section studies on the in uitro effects of IL-lra on cells found in the joint, the effects of IL-lra in animal models of arthritis, and the effects of IL-lra in human joint diseases are reviewed. A. In VitrO STUDIES ON INTERLEUKIN-ha The results of early studies indicated that IL-lra blocked the stimulatory effects of soluble IL-1 (Arend et al., 1990) or of membraneassociated IL-1 (Seckinger et al., 1990a) on PGE2 and collagenase production by human synovial fibroblasts or rabbit articular chondrocytes. Furthermore, IL-lra inhibited the enhanced production of collagenase, gelatinase, caseinase, and PGE2 in IL-l-stimulated, but not TNF-a-treated, bovine nasal chondrocytes (Smith et al., 1992). In additional studies, IL-lra blocked IL-l-induced overproduction of hyaluronic acid by cultured human synovial cells (Seckinger et al., 199Oc), underproduction of sulfated glycosaminoglycans by human chondrocytes (Seckinger et al., 199Oc), and both matrix degradation and decreased glycosaminoglycan production in bovine nasal cartilage explants (Smith et al., 1991). The ability of IL-1 to induce neutral protease release and to alter synthetic function of these cells in uiuo is thought to be an important mechanism of tissue damage in rheumatoid arthritis and other forms of chronic inflammatory arthritis (Arend and Dayer, 1990). In addition, IL-l-induced enzyme release by articular chondrocytes may play a role in cartilage damage in osteoarthritis; anti-inflammatory drugs may enhance production of ILIra by chondrocytes in this disease (Herman et al., 1991). Lastly, IL-lra blocked IL-l-induced bone resorption and PGEz production in cultured mouse calvariae (Seckinger et al., 1990b). Thus, from
2 10
WILLIAM P. AREND
these in uitro studies IL-lra theoretically could be of benefit in both inflammatory and noninflammatory joint disease, as well as possibly in osteopenic conditions.
B. INTERLEUKIN-Ira I N ANIMALMODELSOF ARTHRITIS Initial studies on the effects of IL-lra in different animal models of arthritis have yielded somewhat mixed results. IL-lra was described to be beneficial in collagen-induced arthritis in mice (Wooley et al., 1990) but not in antigen-induced arthritis in rabbits (Wooley et al., 1990; Lewthwaite et al., 1992). In another animal model, reactivation of bacterial cell wall-induced arthritis in rats, IL-lra modestly reduced evidence of joint inflammation (Schwab et al., 1991). In this model, monoarticular arthritis was induced by the intraarticular injection of peptidoglycan-polysaccharide polymers isolated from the cell walls of group A streptococci. Reactivation of the arthritis resulted after an IV injection of the same material 20 days after the initial induction. IL-lra administered subcutaneously or intraperitoneally at the time of reactivation and at 6 hour intervals thereafter led to a 60% reduction in joint swelling and a decrease in the histological severity of synovitis and cartilage erosion (Schwab et al., 1991). Lastly, IL-lra injected intravenously into rabbits largely inhibited PMN emigration into the joint and loss of proteoglycan from articular cartilage following intraarticular injection of IL-1 (Henderson et al., 1991). The variable effects of IL-lra in different animal models of arthritis may reflect the fact that multiple cytokines are involved in the pathophysiological mechanisms or that adequate tissue levels of IL-lra were not attained.
c. INTERLEUKIN-Ira I N PATIENTS WITH ARTHRITIS Elevated levels of an IL-1 inhibitory bioactivity were present in the serum and urine of patients with juvenile rheumatoid arthritis during a febrile episode (Prieur et al., 1987). This bioactivity may represent IL-lra produced in a feedback regulatory mechanism in response to the earlier release of IL-1 in large amounts. Enhanced IL-lra production also most likely occurs in active rheumatoid arthritis, as evidenced by the very high levels of this protein found in the synovial fluids of over 80%ofpatients with this disease (mean IL-lra level 17.1 ng/ml) in the absence of detectable IL-lP (Malyak et al., 1993a). Why do these patients exhibit active arthritis in the presence of high synovial fluid levels of IL-lra? The answer to this question is not known, but many possibilities exist. First, IL-lra in synovial fluid may be bound to other proteins, such as az-macroglobulin or soluble IL-lRIs, and may not be
INTERLEUKIN-1 RECEPTOR ANTAGONIST
211
available to inhibit IL-1 stimulation of target cells. Second, IL-lra in the fluid phase may simply not penetrate well into the synovial tissue. Third, IL-lra production in rheumatoid synovial tissue may not be in sufficient excess to inhibit the effects of locally produced IL-1. Preliminary evidence suggests that this may be the case, as cultured rheumatoid synovial cells produce IL-lra only at an excess of twofold over IL-1 (Firestein et al., 1992b). Lastly, the apparent paradox of ongoing active synovitis in the presence of massive levels of synovial fluid IL-lra may reflect the relative importance of other cytokines, such as TNF-a, in the initiation and perpetuation of the disease process. Clinical studies suggest that IL-lra may be ofbenefit in patients with active arthritis. In patients with acute synovitis secondary to Lyme disease, high levels of synovial fluid IL-lra with respect to IL-1 correlated with a more rapid recovery from the episode of arthritis in comparison to patients with the opposite cytokine pattern (Miller et al., 1993). IL-lra in the fluid phase either is able to penetrate into the synovial tissue or mirrors events occurring in the tissue. The preliminary results of a phase I clinical trial of IL-lra in rheumatoid arthritis suggest some early efficacy (Lebsack et al., 1991).After a single subcutaneous injection of IL-lra or after 7 days of once-daily treatment, decreases were observed in the mean tender joint count, erythrocyte sedimentation rate, and level of C-reactive protein. A phase 11clinical trial of IL-lra in rheumatoid arthritis is currently underway and more information soon should be available on the safety and possible efficacy of this agent over a longer period. XI. Interleukin-lra in the Nervous System
One of the benefits of the availability of recombinant IL-lra for in uitro and in uiuo studies is that the role of IL-1 in normal physiology might be clarified. Extensive evidence exists to suggest that IL-1 might be involved in normal, physiologic regulation of sleep (reviewed in Opp and Krueger, 1991). Intracerebroventricular administration of ILIra led to a transient reduction in non-rapid eye movement sleep (NREMS) in normal rabbits. Furthermore, IL-lra completely prevented the fever and NREMS induced by injection of IL-1 (Opp and Krueger, 1991). In more recent studies, a peptide representing the amino acid sequence 208-240 of pro-IL-lp induced both fever and NREMS in rabbits or rats (Opp et al,, 1992). Pretreatment with human IL-lra markedly reduced both the pyrogenic and somnogenic effects of this IL-lj3 peptide from the same species. The IL-lj3 208-240 fragment
212
WILLIAM P. AREND
did not affect T cell function, whereas the 163-171 peptide fragment was active in T cell assays but did not induce fever or PGEz production (Boraschi et al., 1988). Whether IL-lra blocks the in uitro or in uiuo effects of the proinflammatory IL-1p 208-240 fragment has not been reported; however, the possibility exists that IL-lra not only inhibits the receptor binding of IL-1, but also blocks secondary interactions of IL-1 with cell surface receptors that trigger intracellular biological responses. Studies on IL-lra peptides along with the pyrogenic or immunomodulatory peptides of IL-lP should clarify this possibility. Other effects of IL-lra on central or peripheral nervous system function or cells have been reported. IL-lra blocked the IL-lp-induced decreases in firing rates of both wann- and cold-sensitive neurons in the preoptic areas of guinea pig brain slices (Xin and Blatteis, 1992). These temperature-sensitive neurons are thought to be the central nervous system (CNS) targets for endogenous pyrogens. The pyrogenic and somnogenic effects of IL-1 in the CNS may be secondary, at least in part, to its ability to augment y-aminobutyric acid A receptor expression on neuronal cells and to induce a subsequent increase in chloride transport into the cells. IL-ha specifically inhibited this effect of IL-1 on cultured primary neuronal cells (Miller et al., 1991). In addition, IL-Ira markedly reduced neuronal death in rats secondary to focal cerebral ischemia and IL-lra decreased excitotoxic injury caused by infusion of a glutamate derivative (Relton and Rothwell, 1992). These interesting results suggest that IL-1 may be a mediator of ischemic and excitotoxic brain damage possibly mediated by glutamate receptors on neuronal cells, In other studies, intraperitoneal (IP) injection of IL-lra reduced the depressive effects of LPS (IP)on social exploration and body weight in rats (BluthC et al., 1992). Intracerebroventricular injection of IL-lra had no effects, suggesting that IL-lra was acting at a peripheral and not a central level. In other rat studies, however, intracerebroventricular injection of IL-lra prevented the IL-lp-induced suppression of food and water intake (Plata-Salaman and Ffrench-Mullen, 1992). Lastly, IL-lra impeded peripheral nerve regeneration mediated by IL-1induced nerve growth factor in Schwann cells (GuCnard et al., 1991). A monograph summarizes the evidence for the presence of IL-1 and IL-1 receptors in the CNS and the possible diverse effects of 1L-1 on CNS function (Rothwell and Dantzer, 1992). IL-lra mRNA appears to be produced in the same areas in the brain where IL-1 mRNA is found, in particular in the paraventricular nucleus of the hypothalamus, the hippocampus, and the cerebellum (Licinio et al., 1991). Thus, the relative production of IL-lra by microglial or neuronal cells in the CNS
INTERLEUKIN-1 RECEPTOR ANTAGONIST
213
may influence the possible local stimulatory effects of IL-1. IL-lra mRNA is also found in the choroid plexus, suggesting that this protein may be secreted into the cerebrospinal fluid. XII. Interleukin-1ra in Malignancies
The in uitro growth of numerous different malignant cells is stimulated or inhibited by the autocrine production of growth factors or interleukins. An example is a human ovarian carcinoma cell line that expresses IL-1 receptors; IL-1 inhibited proliferation of this cell line in uitro (Kilian et al., 1991). IL-lra inhibited IL-1 binding to this cell line in an equimolar fashion and 100-fold excess amounts of IL-lra blocked the antiproliferative effects of IL-1. IL-lra alone, however, exhibited no effect on spontaneous cell proliferation. A different situation exists with malignant myeloid cells from patients with acute or chronic myelogenous leukemia (AML or CML). These cells also express specific IL-1 receptors but IL-1 enhances in uitro growth. Unstimulated peripheral blood or bone marrow cells from AML patients contained both IL-1P and GM-CSF (Rambaldi et d.,1991). IL-lra inhibited both the spontaneous proliferation of these cells in uitro and the production of GM-CSF and IL-1P. Thus, it appears that IL-1 is an autocrine growth factor for these cells in uitro, and IL-1 produced by the cells induces further production of IL-1 as well as of GM-CSF. IL-lra mRNA was absent or barely detectable in these AML cells, suggesting that an imbalance in IL-1 and IL-lra synthesis may have predisposed to the unrestricted growth of these cells. A somewhat different growth pattern was observed with CML cells, where IL-1 augmented interferon-y-sensitive proliferation of cells from a subset of patients (Estrov et al., 1991); however, cells from other patients were resistant to interferon-y-induced enhancement in growth in uitro. These cells appeared to represent a more advanced stage of the disease; IL-1P did not enhance their growth in uitro, suggesting that they were endogenously producing maximal amounts of IL-1. This conclusion is supported by the observation that IL-lra inhibited the spontaneous in uitro growth of cells from interferon-yresistant CML cells by up to 56%, whereas lower levels of growth inhibition were observed with interferon-y-sensitive cells (Estrov et al., 1991). These observations on the growth inhibitory effects of ILIra on AML or CML cells in uitro have led to a clinical trial of IL-lra in CML. It is possible that this material may find a role as an adjunct to other forms of therapy in these diseases. Serum levels of IL-lra are increased in patients with Hodgkin’s
214
WILLIAM P. AREND
disease, particularly in those lacking systemic symptoms (Gruss et al., 1992). This observation suggests that these symptoms may be seen in Hodgkin’s disease when the level of IL-lra production is not adequate to block the effects of IL-1 fully. XIII. Interleukin-1ra in Diabetes Mellitus
IL-1 has been implicated in pancreatic /3 cell damage in patients with type I diabetes mellitus (reviewed in Sandler et al., 1991). IL-lp inhibits insulin production by rat pancreatic p cells in uitro as well as induces cytotoxicity. These two effects may be mediated by different mechanisms possibly involving the production of neutral proteases and nitric oxide. IL-lra was found to prevent the IL-1-induced decrease in insulin content and release in cultured rat islet cells in uitro (Dayer-M6troz et al., 1989).IL-lra also blocked the decrease in growth of a rat insulinoma cell line treated with IL-1 (Eizirik et al., 1991). In more recent studies IL-lra was shown to block both an early phase of IL-1 enhancement in glucose-induced insulin release in cultured rat pancreatic cells in uitro and a late phase of decreased insulin synthesis and release (Eizirik et al., 1992). Studies from another laboratory indicated that IL-lra was more inhibitory to effects of IL-1 on pancreatic a than p cells and that inhibition of the latter cells may require 1000-fold or greater excess amounts of IL-lra (J. Nerup, unpublished observations). Studies on the in uiuo effects of IL-lra in spontaneous or induced animal models of diabetes have not yet been reported. XIV. Interleukin-1ra in Inflammatory Bowel Disease
IL-1 may be involved in the colonic lesions in inflammatory bowel disease. IL-lra was shown to have anti-inflammatory effects in animal models of ulcerative colitis. In one study, colitis was induced in rabbits by the instillation of formalin into the colon followed by immune complexes (Cominelli et al., 1990). IL-1 gene expression and protein synthesis were demonstrated in colonic tissue at early times, followed by the appearance of PGE2 and leukotriene Bq. Furthermore, the tissue levels of IL-1 correlated with the degree of tissue inflammation in this rabbit model of colitis. The IV administration of IL-lra before and during the development of this disease reduced all of the indices of inflammation (Cominelli et al., 1990). Additional studies revealed that IL-la! levels in colon tissue were not changed by IL-lra treatment; however, PGE2 and leukotriene B4 levels were decreased (Cominelli et al., 1992).IL-lra also was effective in reducing colonic inflammation
INTERLEUKIN-1 RECEPTOR ANTAGONIST
215
if administered 3 hours, but not 12 hours, after induction of the colitis. IL-lra also appeared to be effective in a rat model of colitis induced by the local instillation of acetic acid (Thomas et al., 1991). Again, based on these observations in experimental animal models of colitis, a clinical trial of IL-lra in patients with inflammatory bowel disease has been initiated. W . Other Effects of Interleukin-1ra in Witro and in Wivo
A. GRAFT-VERSUS-HOST DISEASE Graft-versus-host disease (GVHD) is a major complication of allogeneic bone marrow transplantation. Both TNF-a and IL-1 have been implicated in the pathophysiology of GVHD in animal models and humans; however, the relative importance of each cytokine remains unclear. The administration of IL-lra reduced both the immunosuppression and the mortality of murine GVHD after bone marrow transplantation without impairing the engraftment of hematopoietic stem cells (McCarthy et al., 1991). Furthermore, IL-lra did not alter the restoration in cell-mediated immunity. In more recent studies, IL-lra has been found to rescue mice from established GVHD; IL-lra infusion from Days 10 to 20 after bone marrow transplantation increased survival at Day 50 from 27 to 75% (Ferrara et al., 1992a). Preliminary results of a phase 1/11 clinical trial of IL-lra in steroidresistant human GVHD indicated efficacy in some patients (Ferrara et al., 1992b).Ten patients receiving allogeneic bone marrow transplants developed GVHD of grade I1 or greater. These patients were resistant to 6 days of high-dose (3-10 mg/kg/d) methylprednisolone therapy, and they also received cyclosporine A and methotrexate. All patients received a 7-day IV infusion of IL-lra, but at three different doses of 400, BOO, or 1600 mg/d. Four patients exhibited a partial or full response to IL-lra, four patients showed a minimal to no response, and two patients died before completing the 7-day infusion. The beneficial response was not related to IL-lra dose and this material appeared to be safe. More extensive studies on IL-lra in patients with GVHD after bone marrow transplantation are in progress. B. PULMONARY DISEASES
A variety of pulmonary diseases may be associated with the production and effects of IL-1, and IL-lra may ameliorate the inflammatory consequences. Immune complex-induced lung injury follows the IV injection of antigen [bovine serum albumin (BSA)] and the intratracheal instillation of anti-BSA antibodies in rats (Mulligan and Ward,
216
WILLIAM P. AREND
1992).IV infusion of IL-lra prevented the infiltration of PMNs into the lungs and the subsequent lung damage. The effects of IL-lra on lung inflammation following inhalation of LPS is discussed in Section V,C,6 (Ulich et al., 1992). In another study, IL-lra inhibited bronchial airway reactivity following inhalation of aerosolized antigen and priming with an IV injection of substance P in guinea pigs (Selig and Tocker, 1992). Lastly, IL-lra protein has been demonstrated by immunohistochemical techniques to be present in granulomas associated with Mycohacterium tubercuzosis infection, sarcoidosis, and foreign bodies (Chensue et al., 1992). Thus, IL-lra inhibits acute inflammation in animal models of pulmonary disease and is present in the lung lesions of patients with chronic granulomatous disease. The relevance of endogenous production of IL-lra to the course of acute and chronic human lung diseases and the effects of exogenous administration of IL-lra remain to be determined. C. SYNTHESIS OF ACUTEPHASEPROTEINS IL-6 is a potent inducer of acute phase protein synthesis by the liver in acute and chronic inflammatory diseases. IL-lra mRNA was present in a total liver cell RNA extract as well as in a hepatoma cell line (Steinkasserer et aZ., 1992a). IL-lra also blocked the IL-1-induced increase in IL-6 in uivo in mice (Mengozzi et al., 1991). Lastly, IL-lra inhibited the IL-1-induced increase in serum amyloid A protein production and decreases in synthesis of both albumin and haptoglobin by the human hepatoma cell line HUH-7 in oitro (Bevan and Raynes, 1991).The effects of IL-lra on synthesis of acute phase proteins in uivo have not yet been reported.
D. PREGNANCY AND PRETERM DELJVERY Recent studies have demonstrated that IL-lra levels are very high in amniotic fluid (Romero et aZ., 1992).Amniotic fluid was obtained from women in the midtrimester of pregnancy, at term pregnancy, and in preterm labor. IL-lra levels, as determined by a specific radioimmunoassay, were elevated in all amniotic fluid samples in a wide range up to 70 ng/ml. The mean IL-lra levels were the same in all patient groups, either without or with infection. Most interestingly, IL-la and IL-lP were generally undetectable in amniotic fluids, except in samples from infected patients at delivery. In addition, IL-lra reduced IL-1pinduced PGEz production by primary cultures of amnion and chorion cells; IL-lra alone exhibited no agonist effects on these cells (Romero et al., 1992). IL-1 has been strongly implicated in premature labor secondary to intrauterine infection. IL-lra administration prevented
INTERLEUKIN-1 RECEPTOR ANTAGONIST
217
IL-1-induced preterm parturition in pregnant mice without demonstrating any adverse effects on the course or outcome of the pregnancy (Romero and Tartakovsky, 1992). The origin of the very high IL-lra levels in normal amniotic fluid is unknown but suggests a possible role in protecting the pregnancy against adverse effects of IL-1. Lastly, IL-lra may be of therapeutic benefit in the prevention of premature delivery in women with intrauterine infections. E. OTHER EFFECTS OF INTERLEUKIN-1ra As might be imagined, IL-lra inhibits virtually every IL-1-induced effect in uitro or in uiuo mediated through IL-1 receptors. Other examples include ocular inflammation induced by the intravitreous injection of IL-1 but not LPS (Rosenbaum and Boney, 1992); the autocrine constitutive production of IL-la and IL-lP in aging fibroblasts (Kumar et al., 1992); the IL-1-induced production of IL-8 and monocyte chemotactic peptide by human mesangial cells (Brown et al., 1992); and the IL-1-induced inhibition of both human chorionic gonadotropinstimulated testosterone production and cytochrome P450 mRNA expression in cultured rat Leydig cells (Lin et al., 1991a,b). These examples again illustrate the point that the balance between IL-lra and IL-1 in various tissues may influence the degree of IL-1 effects that may result. Whether these effects of IL-1 are important in normal physiological functions and whether IL-lra inhibits those functions in viuo remain largely unknown. Theoretically, agents that enhance IL-lra synthesis or secretion or inhibit IL-1 production i n uiuo could have multiple effects, either beneficial or possibly harmful. XVI. Biological Relevance of interieukin-1ra
Studies to date on IL-lra have raised a number of interesting questions about the biological relevance of this molecule. Why has this highly conserved molecule survived throughout evolution? What role does IL-lra play in normal physiology or in the host response to injurious processes? Does IL-lra exist solely to balance or regulate the potentially proinflammatory effects of IL-1 or does it serve another more physiological function? The answers to these questions are not known at this time but ongoing studies in many laboratories are addressing these issues. IL-1 was initially described as a lymphocyte-activating factor but probably is more important as an endogenous pyrogen and for its effects on cells of the inflammatory system. The precise role of IL-1 in normal physiology has not been established but exciting possibilities
218
WILLIAM P. AREND
exist in the CNS, the neuroendocrine system, the ovaries, and other organ systems. It would appear that IL-lra of either structural variant is produced by most cells in the body that make IL-1. The biological consequences of extracellular secretory IL-lra in various organs may depend on the relative ratio of IL-lra to IL-1. This ratio would be determined by the rates of production and degradation locally in a particular tissue. This article has discussed some of the cytokines and other factors that influence IL- Ira production. Particular emphasis has been placed on those factors that appear to regulate differentially or reciprocally IL-lra and IL-lP production, like adherent IgG, GM-CSF, or IL-4. A greater understanding of the biological relevance of extracellular IL-lra may evolve from studies in which this molecule is either overexpressed or underexpressed in uiuo, in transgenic or knockout mice. High IL-lra levels exist both in the circulation and in human body fluids, such as synovial fluids, in particular diseases. In addition, very high IL-lra levels are present in normal amniotic fluids. This finding indicates that the intact organism is capable of locally making large amounts of IL-lra. Is IL-lra in circulation or body fluids biologically active and available to bind to IL-1 receptors on cells in solution or tissues? What is the relevance of the observation that soluble forms of IL-1RIs may preferentially bind IL-lra and soluble forms of IL-1RIIs may selectively bind IL-lp? The levels ofthese soluble IL-1 receptors in the circulation, body fluids, or tissues of patients with various diseases are not known. It would appear that soluble IL-1RII binding of IL-1p prevents interaction with cell surface receptors. Whether IL-lra is prevented by soluble IL-1RIs from binding to IL-1 receptors on cells is not known. Theoretically, the relative levels of soluble IL-1RIs and IL-1RIIs could locally influence the ratio of active IL-lra to IL-1p. Lastly, the existence of an intracellular structural variant of IL-lra raises additional questions. This molecule seems to be a constitutive product of keratinocytes and other epithelial cells, with a level of production greatly exceeding that of IL-la. Low levels of icIL-lra mRNA, and possibly also protein, are observed in synovial and dermal fibroblasts. In addition, icIL-lra mRNA may be a delayed product of macrophages under particular conditions. These observations suggest that intracellular IL-lra may play additional roles inside cells, such as in regulation of IL-1 receptor expression or of IL-1-induced gene transcription. The possibility also exists that intracellular effects of IL-lra may involve mechanisms other than binding to IL-1 receptors, as detectable IL-1 receptors have not been described inside cells. Studies using techniques of contemporary cell and molecular biology
INTERLEUKIN-1 RECEPTOR ANTAGONIST
219
should provide answers to these questions. Until further information is available, it seems reasonable to believe that IL-lra exists primarily to regulate the extracellular effects of IL-1 in a delayed fashion. Intracellular IL-lra may serve a function only when released from dying or dead cells, along with IL-lP in macrophages and IL-la in keratinocytes. WII. Summary and Conclusions
IL-lra is the first described naturally occurring receptor antagonist of any cytokine or hormone-like molecule. IL-lra is a member of the IL-1 family by three criteria: amino acid sequence homology of 26 to 30% to IL-lP and 19% to IL-la; similarities in gene structure; and common gene localization to human chromosome 2q14. Two structural variants of IL-lra exist: sIL-lra, a secretory molecule produced by monocytes, macrophages, neutrophils, fibroblasts, and other cells; and icIL-lra, an intracellular molecule produced by keratinocytes and other epithelial cells, macrophages, and fibroblasts. IL-lra production by monocytes, macrophages, and neutrophils may be regulated in a differential fashion with IL-1P. Human IL-lra binds to both human IL-1RIs and IL-1RIIs on cell surfaces, although with 100-fold greater avidity to IL-1RIs. IL-lra may bind preferentially to soluble IL-1RIs and not at all to soluble IL-1RIIs. IL-lra competitively inhibits binding of both IL-la and IL-1P to cell surface receptors without inducing any discernible intracellular responses. All three forms of IL-1 may bind to IL-1 receptors in a similar fashion but IL-lra may lack the secondary interactions necessary to trigger cell responses. A 100-fold or greater excess of IL-lra over IL-1 may be necessary to inhibit biological responses to IL-1 both in uitro and in uiuo. The roles of sIL-lra and icIL-Ira in normal physiology or in host defense mechanisms remain unclear. The administration of IL-lra blocks the effects of IL-1 in some animal models of septic shock, inflammatory arthritis, graft-versus-host disease, and inflammatory bowel disease. The preliminary results of clinical trials in humans indicate possible efficacy of IL-lra in sepsis syndrome, rheumatoid arthritis, and GVHD.
REFERENCES Alcami, A., and Smith, G . L. (1992). Cell (Cambridge, Mass.) 71,153-167. Alexander, H. R., Doherty, G . M., Buresh, C. M., Venzon, D. J., and Norton, J. A. (1991). 1.E x p . Med. 173,1029-1032. Alexander, H. R., Doherty, G . M., Venzon, D. J., Merino, M. J., Fraker, D. L., and Norton, J. A. (1992). Surgery 112,188-194.
220
WILLIAM P. AREND
Andersson, J., Bjork, L., Dinarello, C. A., Towbin, H., and Andersson, U. (1992).Eur.]. Zmmunol. 22,2617-2623. Arend, W. P. (1990).Prog. Growth Factor Res. 2,193-205. Arend, W. P. (1991).]. Clin. Inoest. 88,1445-1451. Arend, W. P., and Coll, B. P. (1991).Cytokine $407-413. Arend, W. P., and Dayer, J.-M. (1990).Arthritis Rheum. 33,305-315. Arend, W. P., and Dayer, J.-M. (1993).In “Immunopharmacology of Joints and Connective Tissue’’ (M. E. Davies, and J. T. Dingle, eds.). Academic Press, London, (in press). Arend, W. P., and Massoni, R.J. (1981).Immunology 44,717-725. Arend, W. P., Joslin, F. G., and Massoni, R. J. (1985).J.Zmmunol. 134,3868-3875. Arend, W. P., Joslin, F. G., Thompson, R. C., and Hannum, C. H. (1989).]. Zmmunol. 143, 1851-1858. Arend, W. P., Welgus, H. G., Thompson, R. C., and Eisenberg, S. P. (1990).J. Clin. Znoest. 85, 1694-1697. Arend, W. P., Smith, M. F., Jr., Janson, R. W., and Joslin. F. G. (1991).]. Zmmunol. 147, 1530-1536. Auron, P. E., Quigley, G. J,, Rosenwasser, L. J., and Gehrke, L. (1992).Biochemistry 31, 6632-6638. Balavoine, J.-F., d e Rochemonteix, B., Cruchaud, A,, and Dayer, J.-M. (1985). In “The Physiologic, Metabolic, Immunologic Actions of Interleukin-1” (M. J. Kluger, J. J. Oppenheim, and M. C. Powanda, eds.), pp. 429-436. Alan R. Liss, New York. Balavoine, J.-F., d e Rochemonteix, B., Seckinger, P., Cruchaud, A., and Dayer, J.-M. (1986).]. Clin. Znoest. 78, 1120-1124. Barak, V., Treves, A. J,, Yanai, P., Halperin, M., Wasserman, D., Biran, S., and Braun, S. (1986).Eur.]. Zmmunol. 16, 1449-1452. Barak, V.. Peritt, D., Flechner, I., Yanai, P., Halperin, T., Treves, A. J., and Dinarello, C. A. (1991).Lymphokine Cytokine Res. 10,437-442. Beaulieu, A. D., Paquin, R., Rathanaswami, P., and McColl, S. R. (1992).]. Biol. Chem. 267,426-432. Bertini, R., Sironi, M., Martin-Padura, I., Colotta, F., Rambaldi, S., Bernasconi, S., Ghezzi, P., Haskill, S. J., Liu, D., and Mantovani, A. (1992).Cytokine 4,44-47. Bevan, S., and Raynes, J. G. (1991).]. Zmmunol. 147,2574-2578. Bienkowski, M. J., Eessalu, T. E., Berger, A. E., Truesdell, S. E., Shelly, J. A., Laborde, A. L., Zurcher-Neely, H. A., Reardon, I. M., Heinrikson, R. L., Chosay, J. C., and Tracey, D. E. (1990).]. B i d . Chem. 265,14505-14511. Bienkowski, M. J., Shuck, M. E., Nagpala, P., Drong, R. F., Sieu, L. C., Carter, D. B., and Slightom, J. L. (1993).Submitted for publication. Bigler, C. F., Norris, D. A., Weston, W. L., and Arend, W. P. (1992).J.Inoest. Dermatol. 98,38-44. Bluth6, R.-M., Dantzer, R., and Kelley, K. W. (1992).Brain Res. 573,318-320. Boraschi, D., Nencioni, L., Villa, L. K., Censitii, S . , Bossh, P., Ghiara, P., Presentini, R., Perin, F., Frasca, D., Doria, G., Fomi, G., Musso, T., Giovarelli, M., Ghezzi, P., Bertini, R., Besedovsky, H. O., Del Rey, A., Sipe, J. D., Antoni, G., Silvestri, S., and Tagliabue, A. (1988).J. E r p . Med. 168,675-686. Boraschi D., Volpini, G., Villa, L., Nencioni, L., Scapigliati, G., Nucci, D., Antoni, G., Matteucci, G., Cioli, F., and Tagliabue, A. (1989).]. Zmmunol. 143, 131-134. Brown, Z., Strieter, R. M., Neild, G. H., Thompson, R. C., Kunkel, S. L., and Westwock, J. (1992).Kidney Znt. 42,95-101. Carter, D. B., Deibel, M. R.. Jr., Dunn, C. J., Tomich, C . 4 C., Laborde, A. L., Slightom, J. L., Berger, A. E., Bienkowski, M. J., Sun, F. F., McEwan, R. N., Harris, P. K. W., Yem, A. W., Waszak, G. A., Chosay, J. G., Sieu, L. C., Hardee, M. M., Zurcher-Neely,
INTERLEUKIN-1 RECEPTOR ANTAGONIST
22 1
H. A., Reardon, I. M., Heinrikson, R. L., Truesdall, S. E., Shelly, J. A., Eessalu, T. E., Taylor, B. M., and Tracey, D. E. (1990).Nature (London)344,633-638. Chan, L. S., Hammerberg, C., Kang, K., Sabb, P., Tavakkol, A., and Cooper, K. D. (1992). J . Invest. Dennatol. 99,315-322. Chang, T-L., Shea, C. M., Urioste, S., Thompson, R. C., Boom, W. H.,and Abbas, A. K. (1990).J . Immunol. 145,2803-2808. Chensue, S. W., Warrnington, K. S., Berger, A. E., and Tracey, D. E. (1992).Am.J. Pathol. 140,269-275. Clore, G. M., Wingfield, P. T., and Gronenborn, A. M. (1991). Biochemistry 30,23152323. Cominelli, F., Nast, C. C., Clark, B. D., Schindler, R., Llerena, R., Eysselein, V. E., Thompson, R. C., and Dinarello, C. A. (1990).J . Clin. Invest. 86,972-980. Cominelli, F., Nast, C. C., Duchini, A., and Lee, M. (1992). Gastroenterology 103, 65-7 1. Conca, W., Auron, P. E., Aoun-Wathne, M., Bennett, N.,Seckinger, P., Welgus, H. G., Goldring, S. R., Eisenberg, S. P., Dayer, J.-M., Krane, S. M., and Gehrke, L. (1991). J . Biol. Chem. 266, 16265-16268. Conti, P., Dempsey, R. A., Reale, M., Barbacane, R. C., Panara, M. R., Bongrazio, M., and Mier, J. W. (1991a).Immunology 73,450-456. Conti, P., Reale, M., Barbacane, R. C., Panara, M. R., Bongrazio, M., Dempsey, R. A., and Dinarello, C. A. (1991b). Immunology Lett. 28, 19-26. Conti, P., Reale, M., Panara, M. R., Barbacane, R. C., Bongrazio, M., and Dempsey, R. A. FEBS Lett. 286,137-141. (1991~). Conti, P., Feliciani, C., Barbacane, R. C., Panara, M. R., Reale, M., Placido, F. C., Sauder, D. N., Dempsey, R. A., and Amerio, P. (1992a).Immunology 77,245-250. Conti, P., Panara, M. R., Barbacane, R. C., Placido, F. C., Bongrazio, M., Reale, M., Dempsey, R. A., and Fiore, S. (199213).Cell. Immunol. 145, 199-209. Conti, P., Panara, M. R., Porrini, A. M., Gambi, D., Barbacane, R. C., Reale, M., Bongrazio, M., and Dempsey, R. A. (1992~).Scand. J. Immunol. 36,27-33. Dayer, J.-M., and Seckinger, P. (1989).In “Interleukin-1, Inflammation and Disease” (R. Bomford and B. Henderson, eds.), pp. 283-302. Elsevier Science (Biomedical Division), Amsterdam. Dayer-Metroz, M.-D., Wollheim, C. B., Seckinger, P., and Dayer, J.-M. (1989).J. Autoimmunity 2, 163-171. DeForge, L. E., Tracey, D. E., Kenney, J. S., and Remick, D. G. (1992).Am. J . Pathol. 140,1045-1054. de Waal Malefyt, R., Abrams, J., Bennett, B., Figdor, C. G., and de Vries, J. E. (1991). J . Exp. Med. 174,1209-1220. Dinarello, C. A. (1990). In “Advances in Immunology” (F. J. Dixon, ed.), Vol. 44, pp. 153-205. Academic Press, Orlando. Dinarello, C. A. (1991). Blood 77,1627-1652. Dinarello, C. A,, and Thompson, R. C. (1991).Zmmunol. Today 12,404-410. Dinarello, C. A., Rosenwasser. L. J., and Wolff, S. M. (1981).J . Immunol. 127,2517-2519. Donnelly, R. P., Fenton, M. J., Kaufman, J. D., and Gerrard, T. L. (1991).J . Zmmunol. 146,3431-3436. Dower, S. K., Wignall, J. M., Schooley, K., McMahan, C. J., Jackson, J. L., Prickett, K. S., Lupton, S., Cosman, D., and Sims, J. E. (1989).J . Immunol. 142,4314-4320. Dripps, D. J., Brandhuber, B. J., Thompson, R. C., and Eisenberg, S. P. (1991a).J. Biol. Chem. 266,10331-10336. Dripps, D. J., Verderber, E., Ng, R. K., Thompson, R. C., and Eisenberg, S. P. (1991b). J . Biol. Chem. 266,20311-20315.
222
WILLIAM P. AREND
Eastgate, J. A., Symons, J. A., and Duff, G. W. (1990).FEBS Lett. 260,213-216. Eisenberg, S. P., Evans, R. J., Arend, W. P., Verderber, E., Brewer, M. T., Hannum, C. H., and Thompson, R. C. (1990).Nature (London) 343,341-346. Eisenberg, S. P., Brewer, M. T., Verderber, E.,Heimdal, P., Brandhuber, B. J., and Thompson, R.C. (1991). Proc. Natl. Acud. Sri. U.S.A.88,5232-5236. Eizirik, D. L., Tracey, D. E., Bendtzen, K., and Sandler, S. (1991). Diubetologia 34, 445-448, Eizirik, D. L., Tracey, D. E., Bendtzen, K., and Sandler, S. (1992). Autoimmunity 12, 127-133. Estrov, Z . , Kurzrock, R.,Wetzler, M., Kantajian, H., Blake, M., Harris, D., Gutterman, J. U., and Talpaz, M. (1991).Blood 78,1476-1484. Faherty, D. A., Claudy, V., Plocinski, J. M., Kamta, K., Kilian, P., Thompson, R. C., and Benjamin, W. R.(1992).J . Immunol. 148,766-771. Fenton, M. J., Buras, J. A,, and Donnelly, R. P. (1992).J. Immunol. 149,1283-1288. Ferrara, J. L. M., Abhyankar, S., and Gilliland, D. G. (1992a).Blood 80(Sl), 270a. Ferrara, J. L. M., Weinstein, H. J., Guinan, E. C., Parsons, S., Sieff, C. A., Rimm, I. J., Gilliland, D. G., Bierer, B. E., Burakoff, S. J., Bloedow, D. C., Abate, L. M., Lebsack, M. E., Catalano, M. A., and Antin, J. H. (199313).Blood 80(S1),270a. Fibbe, W. E., Hamilton, M. S., Laterveer, L., Duinkerken, N., Thompson, R. C., van Damme, J., and Willemze, R. (1992).Blood 80(S1),91a. Fiorentino, D. F., Zlotnik, A., Mosmann, T. R.,Howard, M., and O’Garra, A. (1991). J . Immunol. 147,3815-3822. Firestein, G. S., Berger, A. E., Tracey, D. E., Chosay, J. G., Chapman, D. L., Paine, M. M., Yu, C., and Zvaifler, N. J. (1992a).J. Zmmunol. 149,1054-1062. Firestein, G. S., Paine, M. M., Boyle, D. L., Yu, C., Zvaifler, N. J., and Arend, W. P. (1992b).Arthritis Rheum. 35, S143. Fischer, E., Marano, M. A., Barber,A. E., Hudson, A., Lee, K., Rock,C. S., Hawes, A. S., Thompson, R. C., Hayes, T. J., Anderson, T. D., Benjamin, W. R., Lowry, S. F., and Moldawer, L. L. (1991).Am J. Physiol. 261, R4424452. Fischer, E., Marano, M. A., Van Zee, K. J., Rock, C . S., Hawes, A. S., Thompson, W. A., DeForge, L., Kenney, J. S., Remick, D. G., Bloedow, D. C., Thompson, R. C., Lowry, S. F., and Moldawer, L. L. (1992a).J. Clin. Znuest. 89, 1551-1557. Fischer, E., Van Zee, K. J., Marano, M. A., Rock, C. S., Kenney, J. S., Poutsiaka, D. D., Dinarello, C. A,, Lowry, S. F., and Moldawer, L. L. (1992b). Blood 79,2196-2200. Fujiwara, T., and Grimm, E. A. (1992).J. Immunol. 148,2941-2946. Gaffney, E. V., Stoner, C. R., Lingenfelter, S. Ei., and Wagner, L. A. (1989).J. Immunol. Methods 122,211-218. Calve-de Rochemonteix, B., Nicod, L. P., Junod, A. F., and Dayer, J.-M. (1990). Am. J . Respir. Cell Mol. Biol. 3,355-361. Gehrke, L., Jobling, S. A., Paik, L. S. K., McDonald, B., Rosenwasser, L. J., and Auron, P. E. (1990).J . Biol. Chem. 265,5922-5925. Gery, I., and Waksman, B. H. (1972).J. E r p . Msd. 136,143-155. Gery, I., Gershon, R. K., and Waksman, B. H. (1971).J . Immunol. 107,1778-1780. Gery, I., Gershon, R. K., and Waksman, B. H. (1972).J . E r p . Med. 136,128-142. Giri, J. G., Newton, R. C., and Horuk, R. (1990).J. Biol. Chem. 265, 17416-17419. Cordon, G. S., Fisher, C. J., Slotman, G . J., Opal, S. M., Pribble, J. P., Stiles, D. M., and Catalano, M. A. (1992).Clin. Res. 40,254A. Cosset, P., Lassalle, P., Tonnel, A. B., Dessaint, J. P., Wallaert, B., Prin, L., Pestel, J., and Capron, A. (1988).Am. Reo. Respir. Dis. 138,40-46. Goto, F., Goto, K., Miyata, T., Ohkawara, S., Takao, T., Mori, S., Furukawa, S., Maeda, T., Iwanaga, S., Shimonishi, Y., and Yoshinaga, M.(1992). Immunology 77,235-244.
INTERLEUKIN-1 RECEPTOR ANTAGONIST
223
Granowitz, E. V., Clark, B. D., Mancilla, J., and Dinarello, C. A. (1991a).J.Biol. Chem. 266,14147-14150. Granowitz, E. V., Santos, A. A., Poutsiaka, D. D., Cannon, J. G., Wilmore, D. W., Wolff, S. M., and Dinarello, C. A. (1991b). Lancet 338,1423-1424. Granowitz, E. V., Clark, B. D., Vannier, E., Callahan, M. V., and Dinarello, C. A. (1992a). Blood 79,2356-2363. Granowitz, E. V., Porat, R., Mier, J. W., Pribble, J. P., Stiles, D. M., Bloedow, D. C., Catalano, M. A., Wolff, S. M., and Dinarello, C. A. (1992b). Cytokine 4, 353360. Granowitz, E. V., Vannier, E., Poutsiaka, D. D., and Dinarello, C. A. (1992~). Blood 79, 2364-2369. Gruaz-Chatellard, D., Baumberger, C., Saurat, J.-H., and Dayer, J.-M. (1991). FEBS Lett. 294,137-140. Gruss, H.-J., Dolken, G., Brach, M. A., Mertelsmann, R., and Henmann, F. (1992). Lancet 340,968. G u h a r d , V., Dinarello, C. A., Weston, P. J., and Aebischer, P. (1991).J . Neurosci. Res. 29,396-400. Hammerberg, C., Arend, W. P., Fisher, G. J., Chan, L. S., Berger, A. E., Haskill, J. S., Voorhees, J. J., and Cooper, K. D. (1992).J.Clin. Znoest. 90,571-583. Hannum, C. H., Wilcox, C. J., Arend, W. P., Joslin, F. G., Dripps, D. J., Heimdal, P. L., Armes, L. G., Sommer, A., Eisenberg, S. P., and Thompson, R. C. (1990). Nature (London)343,336-340. Haskill, S., Martin, G., Van Le, L., Morris, J., Peace A., Bigler, C. F., Jaffe, G. J., Hammerberg, C., Sporn, S. A., Fong, S., Arend, W. P., and Ralph, P. (1991). Proc. Natl. Acud. Sci. U.S.A. 88,3681-3685. Havell, E. A., Moldawer, L. L., Helfgott, D., Kilian, P. L., and Sehgal, P. B. (1992). J . Zmmunol. 148,1486-1492. Henderson, B., Thompson, R. C., Hardingham, T., and Lewthwaite, J. (1991). Cytokine 3,246-249. Henricson, B. E., Neta, R.,and Vogel, S. N. (1991). Znfec. Zmmunity 59, 1188-1191. Herman, J. H., Sowder, W. G., and Hess, E. V. (1991). J . Rheumatol. 18(S27), 124126. Higgins, G. C., and Postlethwaite, A. E. (1991). In “Monoclonal Antibodies, Cytokines and Arthritis” (T. F. Kresina, ed.), pp. 133-172. Marcel Dekker, New York. Hurwitz, A., Loukides, J., Ricciarelli, E., Botero, L., Katz, E., McAllister, J. M., Garcia, J. E., Rohan, R., Adashi, E. Y., and Hernandez, E. R. (1992). J . Clin. Znoest. 89, 1746- 1754. Janson, R.W., Hance, K. R.,and Arend, W. P. (1991).J.Zmmunol. 147,4218-4223. Janson, R.W., King, T. E., Jr., Hance, K. R.,and Arend, W. P. (1993).Am. Reo. Resp. Dis., in press. Jenkins, J. K., and Arend, W. P. (1993). Cytokine, in press. Jenkins, J. K., Malyak, M., and Arend, W. P. (1993). Submitted for publication. Ju, G., Labriola-Tompkins, E., Campen, C. A., Benjamin, W. R.,Karas, J., Plocinski, J., Biondi, D., Kaf€ka, K. L., Kilian, P. L., Eisenberg, S. P., and Evans, R.J. (1991). Proc. Natl. Acad. Sci. U.S.A. 88,2658-2662. Kilian, P. L., Kafllca, K. L., Biondi, D. A., Lipman, J. M., Benjamin, W. R.,Feldman, D., and Campen, C. A. (1991). Cancer Res. 51,1823-1828. Kline, J. N., Monick, M. M.,and Hunninghake, G. W. (1992). J . Appl. Physiol. 73, 1686-1692. Koch, A. E., Kunkel, S. L., Chensue, S. W., Haines, G. K., and Strieter, R. M. (1992).Clin. Zmmunol. Immunopathol. 65,23-29.
224
WILLIAM P. AREND
Kristensen, M., Deleuran, B., Eedy, D. J., Feldmann, M., Breathnach, S. M., and Brennan, F. M. (1992). Br.J. Dermatol. 127,305-311. Kumar, S., Millis, A. J. T., and Baglioni, C. (1992). Proc. Natl. Acad. Sci. U.S.A. 69, 4683-4687. Kutsch, C. L., Norris, D. A., and Arend, W. P. (1993).J . Inoest. Dennatol., in press. Labriola-Tompkins, E., Chandran, C., Koffka, K. L., Biondi, D., Graves, B. J., Hatada, M., Madison, V. S., Karas, J., Kilian, P. L., and Ju. G. (1991). Proc. Natl. Acad. Sci. U.S.A. 88,11182-1 1186. Larrick, J. W. (1989). Imrnunol. Today 10,61-66. Lebsack, M. E., Paul, C. C., Bloedow, D. C., Burch, F. X.,Sack, M. A,, Chase, W., and Catalano, M. A. (1991).Arthritis Rheum. 34, S45. Lennard, A., Gorman, P., Carrier, M., Griffiths, S., Scotney, H., Sheer, D., and Solari, R. (1992).Cytokine 4,83-89. Lewthwaite, J. C., Hardingham, T. E., and Henderson, B. (1992). Br. J . Rheumatol. 31(S2), 114. Licinio, J., Wong, M.-L., and Gold, P. W. (1991).Endocrinology 129,562-564. Lin, T., Guo, H., Calkins, J. H., Wang, D., and Chi, R. (1991a).Mol. Cell.Endocrinol. 78, 205-209. Lin, T., Wang, D., Nagpal, M. L., Calkins, J. H., Chang, W., and Chi, R. (1991b). Endocrinology 129,1305-1311. Lotz, M., Tsoukas, C. D., Fong, S., Dinarello, C. A., Carson, D. A., and Vaughan, J. H. (1986a).J . Immunol. 136,3643-3648. Lotz, M., Tsoukas, C. D., Robinson, C. A., Dinarello, C. A., Carson, D. A., and Vaughan, J. H. (1986b).J. Clin. Inoest. 78,713-721. Malyak, M., Joslin, F. G., Verderber, E. L., Elsenberg, S. P., and Arend, W. P. (1991). J . Immunol. Methods 140,281-288. Malyak, M., Swaney, R. E., and Arend, W. P. (1993a).Arthritis Rheum., in press. Malyak, M., Smith, M. F., Jr., and Arend, W. P. (1993b). Submitted for publication. Matsushime, H., Roussel, M. F., Matsu'shima,K., Hishinuma, A., and Sherr, C. J. (1991). Blood 78,616-623. Mazzei, G. J., Bernasconi, L. M., Lewis, C., Mermod, J.-J., Kindler, V., and Shaw, A. R. (1990a).J. Immunol. 145,585-591. Mazzei, G. J., Seckinger, P. L.,Dayer, J.-M., and Shaw, A. R. (1990b).Eur.1. Immunol. 20,683-689. McCarthy, D. O., Domurat, F. M., Nichols, J. E., and Roberts, N. J., Jr. (1989).J. Leuk. Biol. 46, 189-198. McCarthy, P. L., Jr., Abhyankar, S., Neben, S., Newman, G., Sieff, C., Thompson, R.C., Burakoff, S. J., and Ferrara, J. L. M. (1991).Blood 78, 1915-1918. McColl, S. R.,Paquin, R.,and Beaulieu, A. D. (1990).Biochem. Biophys. Res. Commun. 172,1209-1216. McColl, S. R., Paquin, R., MBnard, C., and Beaulieu, A. D. (1992).J . E x p . Med. 176, 593-598. McIntyre, K. W., Stepan, G. J., Kolinsky, K. D., Benjamin, W. R.,Plocinski, J. M., Kaftka, K. L., Campen, C. A., Chizzonite, R. A., and Kilian, P. L. (1991).J . E x p . Med. 173, 931-939. McMahan, C. J., Slack, J. L., Mosley, B., Cosman, D., Lupton, S. D., Brunton, L. L., Grubin, C. E., Wignall, J. M., Jenkins, N. A., Brannan, C. I., Copeland, N. G., Huebner, K., Croce, C. M., Cannizzarro, L. A., Benjamin, D., Dower, S. K., Springgs, M. K., and Sims, J. E. (1991).EMBOJ. 10,2821-2832. Mengozzi, M., Bertini, R.,Sironi, M., and Ghezzi, P. (1991).Lymphokine Cytokine Res. 10,405-407.
INTERLEUKIN-I RECEPTOR ANTAGONIST
225
Miller, L., G., Galpern, W. R., Dunlap, K., Dinarello, C. A., and Turner, T. J. (1991).Mol. Pharmucol. 39,105-108. Miller, L. C . , Isa, S., Vannier, E., Georgilis, K., Steere, A. C., and Dinarello, C. A. (1992). J . Clin. Invest. 90,906-912. Miller, L. C., Lynch, E. A., Isa, S., Logan, J. W., Dinarello, C. A., and Steere, A. C. (1993). Lancet, 341,146-148. Moore, S. A., Strieter, R. M., Rolfe, M. W., Standiford, T. J., Burdick, M. D., and Kunkel, S. L. (1992).Am. J . Respir. Cell Mol. Biol. 6,569-575. Mori, H., Itoh, N., Sawairi, M., Wada, K., Nakagawa, M., and Tamaya, T. (1992). Fertility Sterility 57,535-542. Mulligan, M. S., and Ward, P. A. (1992).J.lmmunol. 149,331-339. Nagai, S., Aung, H., Takeuchi, M., Kusume, K., and Izumi, T. (1991). Chest 99,674-680. Nicod, L. P., Habre, F. E., and Dayer, J.-M. (1992). Cytokine 4,29-35. Ohlsson, K., Bork, P., Bergenfeldt, M., Hageman, R., and Thompson, R. C. (1990).Nature (London)348,550-552. Opp, M. R., and Krueger, J. M. (1991).Am.]. Physiol. 260, R4534457. Opp, M. R., Postlethwaite, A. E., Seyer, J. M., and Krueger, J. M. (1992).Proc. Natl. Acad. Sci. U.S.A.89,3726-3730. Orino, E., Sone, S., Nii, A., and Ogura, T. (1992).J.lmmunol. 149,925-931. Patterson, D., Jones, C., Hart, I., Bleskan, J., Berger, R., Geyer, D., Eisenberg, S. P., Smith, M. F., Jr., and Arend, W. P. (1993). Genomics. 15, 173-176. Petitclerc, E., Abel, S., deBlois, D., Poubelle, P. E., and Marceau, F. (1992).J . Cardiovasc. Pharmacol. 19,821-829. Plata-Salaman, C . R., and Ffrench-Mullen, J. M. H. (1992). Psychol. Behau. 51, 12771279. Porat, R., Clark, B. D., Wolff, S. M., and Dinarello, C. A. (1991). Science 254,430-432. Porat, R., Poutsiaka, D. D., Miller, L. C., Granowitz, E. V., and Dinarello, C. A. (1992). FASEB J . 6,2482-2486. Poutsiaka, D. D., Clark, B. D., Vannier, E., and Dinarello, C. A. (1991). Blood 78, 1275-128 1. Prieur, A.-M., Kaufmann, M.-T., Griscelli, C., and Dayer, J.-M. (1987). Lancet 12401242. Quay, J., Arend, W. P., and Becker, S. (1993).J.1nJ Dis., in press. Ragsdale, C. G., and Arend, W. P. (1979).J.E x p . Med. 149,954-968. Ragsdale, C . G., and Arend, W. P. (1980).J . E x p . Med. 151,32-44. Ragsdale, C . G., and Arend, W. P. (1981).Clin. Exp. lmmunol. 46,214-224. Ralph, P., Nakoinz, I., Sampson-Johannes, A., Fong, S., Lowe, D., Min, H.-Y., and Lin, L. (1992).J . lmmunol. 148,808-814. Rambaldi, A., Torcia, M., Bettoni, S., Vannier, E., Barbui, T., Shaw, A. R., Dinarello, C. A., and Cozzolino, F. (1991). Blood 78,3248-3253. Ramilo, O., SBez-Llorens, X., Metersola, J., Jafari, H., Olsen, K. D., Hansen, E. J., Yoshinaga, M., Ohkawara, S., Nariluchi, H., and McCracken, G . H., Jr. (199O).J.E x p . Med. 172,497-507. Relton, J. K., and Rothwell, N. J. (1992). Brain Res. Bull. 29,243-246. Roberts, N. J., Jr., Prill, A. H., and Mann, T. N. (1986).J.E x p . Med. 163,511-519. Rodgers, B. C., Scott, D. M., Mundin, J., and Sissons, J. G. P. (1985).J . Virol. 55,527532. Romero, R., and Tartakovsky, B. (1992).Am. J . Obstet. Gynecol. 167,1041-1045. Romero, R., Sepulveda, W., Mazor, M., Brandt, F., Cotton, D. B., Dinarello, C. A., and Mitchell, M. D. (1992).Am. J . Obstet. Gynecol. 167,863-872. Rosenbaum, J. T., and Boney, R. S. (1992).Arch. Ophthalmol. 110,547-549.
226
WILLIAM P. AREND
Rothwell, N., and Dantzer, R. (1992). “Interleukin-1 in the Brain.” Pergamon Press, Oxford. Roux-Lombard, P., Modoux, C., and Dayer, J.-M. (1989). Cytokine 1,45-51. Roux-Lombard, P., Modoux, C., Vischer, T., Grassi, J., and Dayer, J.-M. (1992).J. Rheumatol. 19,517-523. Salkind, A. R., McCarthy, D. O., Nichols, J. E., Domurat, F. M., Walsh, E. E., and Roberts, N. J., Jr. (1991).J. Infect. Dis. 163,71-77. Sandler, S., Eizirik, D. L., Svensson, C., Strandell, E., Welsh, M., and Welsh, N. (1991). Autoimmunity 10,241-253. Scala, G., Kuang, Y. D., Hall, R. E., Muchmore, .4. V., and Oppenheim, J. J. (1984).J . E x p . Med. 159,1637-1652. Schur, P. H.,Chang, D.-M., Baptiste, P., Uhteg, L. C., and Hanson, D. C. (1990). Clin. Immunol. Immunopathol. 57,45-63. Schwab, J. H., Anderle, S. K., Brown, R. R., Dalldorf, F. G., andThompson, R. C. (1991). Infect. Immunity 59,4436-4442. Seckinger, P., Lowenthal, J. W., Williamson, K., Dayer, J.-M., and MacDonald, H. R. (1987a).J . Immunol. 139, 1546-1549. Seckinger, P., Williamson, K., Balavoine, J.-F., Mach, B., Mazzei, G., Shaw, A., and Dayer, J.-M. (1987b).J. Immunol. 139, 1541-1545. Seckinger, P., Kaufmann, M.-T., and Dayer, J.-M. (1990a). lmmunobiology 180,316327.
Seckinger, P., Klein-Nulend, J., Alander, C., Thompson, R. C., Dayer, J.-M., and Raisz, L. G . (1990b).J. Immunol. 145,4181-4184. Seckinger, P., Yaron, I., Meyer, F. A., Yaron, M., and Dayer, J.-M. (1990~).Arthritis Rheum. 33,1807-1814. Selig, W., and Tocker, J. (1992). Eur. J . Pharmucol. 213,331-336. Shaw, A. (1991). In “Cytokines and Inflammation” (E. S. Kimball, ed.), pp. 35-57. CRC Press, Boca Raton. Shaw, G., and Kamen, R. (1986). Cell (Cambridge, Mass) 46,659-667. Shields, J., Bernasconi, L. M., Benotto, W., Shaw, A. R., and Mazzei, G. J. (1990). Cytokine 2,122- 128. Shirahama, M., Ishibashi, H., Tsuchiya, Y., Kurokawa, S., Hayashida, K., Okumura, Y., and Niho, Y. (1988). Scand. J. Immunol. 28, ‘719-725. Shuck, M. E., Eessalu, T. E., Tracey, D. E:., and Bienkowski, M. J. (1991). Eur. J. Immunol. 21,2775-2780. Sims, J. E., Acres, B., Grubin, C. E., McMahan, C. J., Wignall, J. M., March, C. J., and Dower, S. K. (1989).Proc. Natl. Acad. Sci. U.S.A. 86,8946-8950. Smith, B. K., and Kluger, M. J. (1992).Am. J . Physiol. 263, R653-R655. Smith, M. F., Jr., Eidlen, D., Brewer, M. T., Eisenberg, S. P., Arend, W. P., and Gutierrez-Hartmann. A. (1992).J. Immunol. 149,2000-2007. Smith, R. J., Chin, J. E., Sam, L. M., and Justen, J. M. (1991). Arthritis Rheum. 34,7883.
Smith, R. J., Justen, J. M., Ulrich, R. G., Lund, J. E., and Sam, L. M. (1992). Clin. Immunol. Immunopathol. 64,135-144. Spinas, G. A., Bloesch, D., Kauhann, M.-T., Keller, U., and Dayer, J.-M. (1990). Am. J . Physiol. 259, R993-R997. Sporn, S. A,, Eierman, D. F., Johnson, C. E., Morris, J., Martin, G., Ladner, M., and Haskill. S. (1990).J. Immunol. 144,4434-4441. Spriggs, M. K., Hruby, D. E., Maliszewski, C. R., and Pickup, D. J., Sims, J. E., Buller, R. M. L., and VanSlyke, J. (1992). Cell (Cambridge, Mass) 71,145-152. Steinkasserer, A., Koelble, K., and Sim, R. B. (1991). Nucleic Acids Res. 19,5095.
INTERLEUKIN-1 RECEPTOR ANTAGONIST
227
Steinkasserer, A., Estaller, C., Weiss, E. H., and Sim, R. B. (1992a). FEBS Lett. 310, 60-62. Steinkasserer, A., Solari, R.,Mott, H. R.,Aplin, R.T., Robinson, C. C., Willis, A. C., and Sim, R. B. (1992b). FEBS Lett. 310,63-65. . 13, Steinkasserer, A., Spurr, N. K., Cox, S., Jeggo, P., and Sim, R. B. ( 1 9 9 2 ~ )Genomics 654-657. Stockman, B. J.. Scahill, T. A., Roy, M.,Ulrich, E. L., Strakalaitis, N. A., Brunner, D. P., Yem, A. W., and Deibel, M. R.,Jr. (1992). Biochemistry 31,5237-5245. Stylianou, E., O’Neill, A. J. L., Rawlinson, L., Edbrooke, M. R.,Woo, P., and Saklatvala, J. (1992).J . Biol. Chem. 267, 15836-15841. Svenson, M., and Bendtzen, K. (1988). Scand. J . Immunol. 27,593-599. Svenson, M., Hansen, M. B., Heegaard, P., Abell, K., and Bendtzen, K. (1993). Submitted for publication. Symons, J. A., and Duff, G. W. (1990). FEBS Lett. 272,133-136. Symons, J. A., Eastgate, J. A., and Duff, G. W. (1990). Cytokine 2, 190-198. Symons, J. A., Eastgate, J. A,, and Duff, G. W. (1991).J . Exp. Med. 174, 1251-1254. Takeuchi, M., Nagai, S., Nakada, H., Aung, H., Satake, N., Kaneshima, H., and Izumi, T. (1992). Clin. E x p . Immunol. 88, 181-187. Tarlow, J. K., Blakemore, A. I. F., Lennard, A., Solari, R.,Hughes, H. N., Steinkasserer, A., and Duff, G. W. (1993). Hum.Gene., in press. Thomas, T. K., Will, P. C., Srivastava, A., Wilson, C. L., Harbison, M., Little, J., Chesonis, R. S., Pignatello, M., Schmolze, D., Symington, J., Kilian, P. L., and Thompson, R.C. (1991).Agents Actions 34,187-190. Tiku, K., Tiku, M. L., Liu, S., and Skosey, J. L. (1986).J . Immunol. 136,3686-3692. Tiku, K., Thakker-Varia, S., and Tiku, M. L. (1992).Arthritis Rheum. 35, S123. Tompkins-Labriola, E., Chandran, C., KafRa, K. L., Biondi, D., Graves, B. J., Hatada, M., Madison, V. S., Karas, J., Kilian, P. L., and Ju, G. (1991). Proc. Natl. Acad. Sci. U.S.A. 88,11182-11186. Tucci, A., James, H., Chicheportiche, R.,Bonnefoy, J.-Y., Dayer, J.-M., andzubler, R. H. (1992).J . Immunol. 148,2778-2784. Turner, M., Chantry, D., Katsikis, P., Berger, A., Brennan, F. M., and Feldman, M. (1991). Eur. J . Immunol. 21,1635-1639. Ulich, T. R.,Yin, S., Guo, K., del Castillo, J., Eisenberg, S. P., and Thompson, R. C. (1991).Am. J . Pathol. 138,521-524. Ulich, T. R.,Guo, K., Yin, S., del Castilo, J., Yi, E. S., Thompson, R.C., and Eisenberg, S. P. (1992).Am. J . Pathol. 141,61-68. Vannier, E., Miller, L. C., and Dinarello, C. A. (1992). Proc. Natl. Acad. Sci. U.S.A.89, 4076-4080. Wakabayashi, G., Gelfand, J. A., Burke, J. F., Thompson, R. C., and Dinarello, C. A. (1991). FASEB J . 5,338-343. Wilder, C. L., Friedrich, A. D., Potts, R.O., Daumy, G. O., and Francoeur, M. L. (1992). Biochemistry 31,27-31. Wooley, P. H., Whalen, J. D., Chapman, D. L., Berger, A. E., Aspar, D. G., and Richard, K. A. (1990).Arthritis Rheum. 33(9), S20. Xin, L., and Blatteis, C. M. (1992). Brain Res. 569,348-352. Zahedi, K., Seldin, M. F., Rits, M., Ezekowitz, A. B., and Whitehead, A. S. (1991). J . Immunol. 146,4228-4233. Zamir, O., Hasselgren, P. O., O’Brien, W., Thompson, R. C., and Fischer, J. E. (1992). Ann. Surg. 216,381-387. This article was accepted for publication on 19 January 1993.
ADVANCES IN IMMUNOLOGY, VOL. 54
Mechanism and Regulation of Immunoglobulin lsotype Switching ROBERT 1. COFFMAN,' DEBORAH A. LEBMAN,t A N D PAUL ROTHMANS
* Department of Immunology,DNAX Reswrch Instht., Palo A h , California 94304, ond
t Department of Microbiology and Immunology, Medical College of Viginio, Richmond, Viginia
23298, ond # Dopartnwnt of Medicine, Co1l.g. of Htysicians ond Sugwns of Columbia University, New Yo& New York 10032
1. Introduction
The basic subunit of all immunoglobulin molecules consists of four polypeptides, two identical heavy (H) chains and two identical light (L) chains. Each of these chains consists of an amino-terminal variable (V) region followed by a constant (C) region. The variable regions of each HL pair form the antigen-binding site of the molecule, whereas the constant region of the heavy chain confers specific effector functions on the molecule. There are nine heavy chain constant region (CH ) genes in the haploid genome of humans, eight in mouse and rat, and several fewer in many other species. It is the CH region of an immunoglobulin molecule that defines its isotype, that is, its class or subclass. Antibodies of different isotypes differ significantly in their abilities to fix complement, bind to CH-specific receptors (Fc receptors) present on a wide variety of cell types, cross mucosal and placental barriers, and form polymers of the basic four-chain immunoglobulin molecule. The phenomenon known as class or isotype switching was initially characterized as the appearance in the serum first of IgM and later of IgG antibodies in response to primary and secondary challenges with foreign antigens (Bauer et al., 1963; Uhr and Finkelstein, 1963). The relative rate at which this switch occurred, the maximum titers of each isotype that were reached, and whether IgG appeared during the primary response varied in responses to different antigens and adjuvants. The sequential order of appearance of IgM and IgG, however, seemed independent of these variables. In general, IgM responses diminished or disappeared entirely following secondary and subsequent challenges, whereas secondary IgG responses were greater and developed as rapidly as the IgM responses. The elegant and difficult analyses of Nossal and colleagues (Nossal et al., 1964) suggested that single antibody-forming cells (AFCs) could secrete both IgM and IgG antibodies at a certain point during a primary immune response. 229 Copyright 0 1993 hy Academic Press, Inc. All rights of reproduction in any form reserved.
230
ROBERT L. COFFMAN ET AL.
For many years, the key question was whether the switch in the serological response reflected a switch in the isotype produced by a single AFC (or clone of AFCs), or represented the progressive dominance of the response by a distinct lineage of AFC committed to IgG production? In other words, was the isotype switching an antigen-driven differentiation step occurring within clones of mature B cells, or did B cells expressing different isotypes arise as separate lineages at an earlier stage of B cell development? Even before the genetic basis of isotype switching was elucidated, several lines of evidence demonstrated convincingly that isotype switching was primarily a clonal, antigen-induced differentiation of mature B cells-the last in a series of remarkable genetic rearrangements that occur in the development of the B cell lineage. In both embryonic chickens (Cooper et al., 1972) and neonatal mice (Lawton and Cooper, 1993; Manning, 1974; Murgita et al., 1973)treatment with anti-p chain antibodies caused a marked reduction in serum IgG and IgA and in IgG' B cells, as well as the expected deficit in IgM. Similarly, the addition of anti-p antibodies to in uitro cultures blocked the development of antigen-specific (Pierce et al., 1972) and polyclonal (Andersson et al., 1978; Kearney et al., 1976) IgG responses. Press and Klinman (1973) first demonstrated in splenic fragment cultures that multiple isotypes could be produced by a single clone of antigenstimulated B cells. This was a direct demonstration that switching occurred during antigen-dependent clonal expansion of B cells. Similarly, it was shown that the primary IgM and IgG1 responses to a hapten developed entirely from B cells that bore membrane IgM (mIgM) and mIgD, but not mIgG1, receptors for that hapten (Coffman and Cohn, 1977).These latter two experiments also demonstrated that the B cells could switch their expression of heavy chain constant (CH ) regions, yet retain their original antigen specificity, and, presumably, their heavy chain variable (VH) region. The retention of V H regions during isotype switching was confirmed with idiotypes as clonal markers in serological responses of individual rabbits (Oudin and Michel, 1969), in human myeloma (Wang et nl., 1970), and in mouse splenic fragment cultures (Gearhart et al., 1975). Although isotype switching appears to occur largely in response to antigen challenge, there remains evidence that it can also occur during early stages in B cell development (Cooper et al., 1977; Abney et al., 1978). Isotype switching could be induced by the direct actions of mitogens or antigens on B cells, as shown by the ability of a variety of Tindependent antigens and mitogens to induce IgG, especially IgG3, in mice (Rosenberg, 1982; Kearney et al., 1976; Mongini et al., 1981;
IMMUNOGLOBULIN ISOTYPE SWITCHING
23 1
Andersson et al., 1978; Press and Klinman, 1973; Severinson et al., 1982). However, IgG, IgA, and IgE responses to most antigens are far more dependent on the presence of T cells (specifically helper, or, as they are now called, CD4+ T cells) than are IgM responses (Rosenberg, 1982; Ishizaka, 1976; Mitchell et al., 1972; Grumet, 1972; Pierce and Klinman, 1975). This dependence of IgG responses on T cells could also be shown for nominally T-independent antigens (Schrader, 1975).Although a number of early studies suggested that switching to each isotype might be mediated by a specific subset of T cells, clonal progeny of single T cells were shown to stimulate the production of multiple isotypes in splenic fragment cultures (Pierce et al., 1978). Studies over the past decade with cloned Th lines have also failed to confirm the existence of Th cells specific for single isotypes, although Th cells that express different cytokines can differ in their abilities to induce switching to some isotypes (Coffman et al., 1988). It now well established, in most cases, that isotype switching involves an intrachromosomal recombination that brings the fully assembled and expressed V H gene into close proximity to a new CH gene, with the resulting deletion of C,, Cs, and any other intervening CH genes. The consequence of this essentially irreversible cellular differentiation is the production of antibody molecules that retain the same antigen specificity (because they retain the same L chain and VH ), but connect it with different effector functions. In this article we focus first on the molecular mechanisms of switch recombination and other possible modes of isotype switching. The remainder of the article will be devoted to the regulation of isotype switching, especially by T cells and the cytokines they produce. Virtually all of the analyses of the mechanisms and regulation of switching have been done in either mouse or human, with a disproportionate amount done in the mouse. For most of this article, the two species will be discussed separately, although the basic lessons learned from the mouse are, for the most part, valid for humans as well. II. Molecular Mechanisms of lsotype Switching
A. ORGANIZATION OF HEAVY CHAINLocus IN MOUSEAND HUMAN The immunoglobulin heavy chain constant region genes are found in a locus 3’ of the heavy chain VH genes. The murine CHlocus is on chromosome 12 and is composed of eight different genes that encode the different isotypes of murine immunoglobulin (Fig. 1).The entire locus spans approximately 200 kb and the CH genes are ordered
232 VDJ
ROBERT L. COFFMAN ET AL.
Cp CS
Cyl
Cy2b
C@a
CE
Ca
Mouse VDJ
Cp C6
Cy3 Cyl
Cal
w
c$2 cys
CE ca2
Human FIG.1. The organization of the immunoglobulin heavy chain constant region genes in mouse and human. The individual exons of each gene are not represented in this diagram.
5' ~-6-y3-yl-y2b-y2a-&-a 3' (Shimizu et al., 1982). The CH genes are
thought to have evolved from a common ancestral gene, most similar to C, , which contains four separate exons encoding separate protein domains (Kawakami et al., 1980; Rogers et al., 1980).The C, gene also contains four exons (Ishida et al., 1982; Liu et al., 1982), whereas the C , and C , genes consist of three exons and a short hinge exon (Tucker et al., 1979, 1981; Sakano et al., 1979). The hinge region, which is found between the first and second CHexons, encodes a short stretch of amino acids rich in proline and cysteine residues. The hinge region is thought to impart flexibility between the two binding regions of IgG and IgA antibodies, thereby allowing them to more easily bind two antigen determinants. The Ca gene only has two exons plus a hinge region exon (Tucker et al., 1980). The human CH locus spans over 300 kb on chromosome 14 and is composed of nine functional genes and two pseudogenes (Fig. l ) , 3' (Hofker et al., 1989; ordered 5' p-S-y3-yl-+&-al-J-y2-y4-~-a2 Kirsch et al., 1982). The structure of this locus is believed to have arisen from a tandem duplication of the y-y-&-a segment of a primordial locus (Flanagan and Rabbitts, 1982). In addition, there is another C, gene (+C,,) located on chromosome 9. This pseudogene does not contain introns or a 5' switch region, and is thought to represent the integration of a processed RNA intermediate (Battey et al., 1982). Otherwise the structure of the other human C H genes are similar to those described above for the murine genes. Human and murine immunoglobulin can be found in either membrane or secreted forms. During early stages of B cell development the membrane form predominates, but after B cells are stimulated by
IMMUNOGLOBULIN ISOTYPE SWITCHING
233
antigen they can terminally differentiate into plasma cells that produce large amounts of secreted immunoglobulin. The membrane form of immunoglobulin differs from secreted immunoglobulin by the addition of one (6) or two ( p ) exons to the 3’ end of the CH gene. These exons encode transmembrane and intracytoplasmic domains (Alt et al., 1980). The mRNA encoding the membrane or secreted forms of immunoglobulin results from differential RNA splicing and involves the use of alternative polyadenylation sites (reviewed in Kemp et al., 1983). The mechanisms that determine which form of immunoglobulin is produced are not well understood.
B. SWITCHREGIONS The switch from the expression of C, and Cs to the expression of downstream isotypes normally involves a recombination that deletes the intervening CH genes from the chromosome and positions the new CH gene directly 3’ of the assembled VH gene. The sites of these recombinations are in regions of tandem repeats, termed switch (S) regions, which are present 5‘ of each CH gene, except Cs (Davis et al., 1980; Maki et al., 1981; Kataoka et al., 1981). Switch regions vary in size from 1kb (S, ) to 10 kb & I ) , and are composed of tandem repeats that vary both in length and sequence (reviewed in Gritzmacher, 1989).The murine S, region is about 3 kb and can be divided into a 3’ region with sequences of [( GAGCT), GGGGT], ,where n = 1-7 and m = 150 (Nikaido et al., 1981), and a 5’ region in which these two pentamers are interspersed with the heptamer sequence (C/ T)AGGTTG (Marcu et al., 1982).The human S, locus is slightly different in that the heptamer sequence is distributed throughout the region (Takahashi et al., 1982; Mills et al., 1990). Although other switch regions contain more complex patterns of repeated sequences, all switch sequences contain multiple copies of the pentameric sequences GAGCT and GGGGT (Nikaido et al., 1982; Stanton and Marcu, 1982). The pentamers ACCAG, GCAGC, and TGAGC are also commonly found in switch regions (Gritzmacher, 1989). In addition, the heptameric repeat (C/T)AGGTTGis abundantly present in switch region sequences and is found near many, but not all, switch recombination sites that has been characterized in plasmacytomas and hybridomas (Marcu et al., 1982). At the murine S, and S, loci there are sequences 40 (Nikaido et al., 1982) and 80 (Davis et al., 1980; Obata et al., 1981) bp in length, respectively, that are tandemly repeated. These sequences are somewhat homologous to S, , especially in areas of the repeats containing the GAGCT pentamer. Both human and murine S, regions are much
234
ROBERT L. COFFMAN ET AL.
less homologous to S, than are the S, and S, regions. The homology of murine S, regions to S, decreases with the position along the chromosome ( S y 3 > Syl > S72b > Syza). The Inurine S, regions are composed of tandem repeats of 49 or 52 bp (SYza) sequences. Within the 49- or 52-bp repeats, the pentameric sequences TGGGG, GCAGC, and ACCAG are commonly found (Kataoka et al., 1981; Mowatt and Dunnick, 1986; Nikaido et al., 1981; 1982; Stanton and Marcu, 1982; Szurek et al., 1985; Wu et al., 1984).
c. MECHANISMSOF SWITCH RECOM5INATION The production of isotypes other then IgM and IgD requires a recombination event that juxtaposes a new C H gene directly 3' of the productive VHDJH rearranged gene. 'This recombination can occur in several ways.
1 . Looping Out cind Deletion The primary mechanism for switching involves intrachromosomal recombination, with looping out and deletion of the CH genes between the sites of recombination. Although the analysis of cell lines and hybridomas has been supportive of this mechanism, the identification and characterization of the deleted products of recombination have added conclusive evidence that intrachromosomal deletion is the major mechanism of switch recombination. Several groups have found that the intervening regions are deleted as a circle that can be isolated (von Schwedler et al., 1990a; Iwasato et al., 1990; Matsuoka et al., 1990). Analysis of these deleted circles isolated from murine B cells has revealed that these circles are the products of intrachromosomal recombination between two S regions. The looped region, which was usually lost from the cell, could also reintegrate either back into the heavy chain locus, resulting in an inversion (Jack et al., 1988), or (rarely) into other regions of the genome (De Pinho et al., 1984).
2. Sister Chromatid Exchange Another mechanism of switching involves unequal crossing over between sister chromatids (Obata et al., 1981; Coleclough et al., 1980; Tilley and Birshtein, 1985). This recombination results in the asymmetric segregation of CHgenes in daughter cells and the accumulation of duplicated C Hgenes on the same chromosome. Therefore a cell that switches from C , to Cy2b will contain one C , gene, whereas the sister cell will contain three C , genes. However,. there are several lines of evidence that indicate sister chromatid exchange is an infrequent mechanism of switching. In one analysis of pre-B cell lines that switch
IMMUNOGLOBULIN ISOTYPE SWITCHING
235
in culture, cells with three C, genes could not be detected (Wabl et al., 1985).In addition, the identification of deleted switch circles indicates that normal B cells utilize intrachromosomal deletion, and not sister chromatid exchange, as the major mechanism of switch recombination. Another mechanism of switching involves the mitotic interchromosoma1 recombination between homologous chromosomes (Kipps and Herzenberg, 1986; Davis et al., 1980).This mechanism could explain the “back switches” that occur at low frequency in myeloma cell lines (Radbruch et al., 1980), but it does not appear to be the common mechanism of switch recombination (reviewed in Gritzmacher, 1989; von Schwedler et al., 1990b). 3. Comparisons between VDJ and Switch Recombination To understand the molecular mechanism involved in switch recombination, it is helpful to contrast it to VDJ recombination, the other form of somatic recombination involved in B cell development that also involves looping out and deletion of DNA. All germline variable region gene segments (VH , D, and JH) are flanked by recognition sequences that consist of conserved palindromic heptamer and nonamer sequences separated by either a 12- or 23-bp spacer region (Sakano et aZ., 1980).In general, recombination occurs only between gene segments flanked by recognition sequences that have different-length spacers (referred to as the 12/23rule). The sites of VDJjoins are within coding exons. Therefore, two-thirds of all joins occur out of frame and generate a nonproductive allele that cannot encode a protein product (reviewed in Blackwell and Alt, 1989a). In contrast, the site of switch recombination lies within the intron between the rearranged VDJ gene and the CH gene and, thus, switching would not be expected to generate a high frequency of nonproductive genes. Another distinction between these two somatic recombination events is their timing in B cell development. VDJ recombination, which requires the products of the RAG I and RAG 2 genes, occurs in the pre-B cell stage of B cell development within the fetal liver and bone marrow (reviewed in Alt et al., 1992). In contrast, switch recombination occurs, for the most part, later in B cell development, when mature peripheral B cells (mIgM+/mIgD+)are stimulated by antigen. The RAG I and RAG 2 genes do not appear to be actively transcribed during this later stage of B cell development. The differences between VDJ and switch recombination suggest there are at least partly distinct “recombinases” for these two types of rearrangements. Studies of the sequences involved in switch recombination have proved to be more difficult to interpret. A large number of the switch
236
ROBERT L. COFFMAN ET AL.
recombination joins that have been mapped in hybridomas and tumors have occurred 5’ of the S, pentameric repeats (Gritzmacher, 1989), raising the question of whether switch recombination actually requires switch sequences. More recent studies, however, indicate that switch recombination most commonly does involve switch regions. The sites of recombination characterized in deleted switch circles were shown to be in the actual Sp region (von Schwedler et al., 1990a; Iwasato et al., 1990; Matsuoka et al., 1990). In addition, deletional recombination within an artificial recombination substrate (Section II,C,4) occurred in switch regions and requires the presence of switch sequences (Ott et al., 1987; Ott and Marcu, 1989). Within the S region, recombination sites are dispersed throughout the region, without apparent predilection for a particular area within the switch region. In an attempt to determine whether there is any site specificity in switch recombination, Dunnick et al. (1993) have compiled the sequences of over 150 recombination sites. This analysis included sequences of the recombination joins left in the chromosome and the reciprocal products that are deleted as circular forms. This analysis confirmed that in contrast to VDJ recognition sequences, there does not seem to be specific sequences present at, or juxtaposed to, all switch joins, implying that switch recombination is not site specific. However, the repeated pentameric and heptameric sequences are found near many joins. It is still unclear whether these sequences play a direct role in recombination or if their presence at joins reflects their frequency within switch regions. In addition, the analysis revealed that, despite the homology between different switch regions, the recombination sites frequently lack donor-acceptor homology. One of the difficulties in studying switch recombination is the possibility that artifacts are generated during the amplification and cloning of switch joins. Certain bacterial extracts have been shown to mediate recombination between S regions in vitro (Kataoka et al., 1983). One possible explanation for this is that chi sequences, which promote generalized homologous recombination in Escherichia coli (e.g., Smith et al., 1981), are homologous to switch sequences and indeed the chi sequence itself, GCTGGTTGG, is found within the S,Z, and S,I regions (Kenter and Birshtein, 1981). Therefore, recombination sites that are cloned and amplified in bacteria can undergo recombination in vitro and the sites identified would not be the sites of the initial B cell recombination event. In addition, the repetitive nature of S regions may make artifacts more likely during amplification by polymerase chain reaction (PCR). These technical difficulties have led to the possibility that some of the sites of switch recombination that are
IMMUNOGLOBULIN ISOTYPE SWITCHING
237
isolated and amplified are not indicative of the sites utilized during the recombination event in B cells. 4. Experimental Systems for Studying Switch Recombination
To learn more about the switch recombinase machinery, several investigators have developed switch substrate constructs similar to those used in analyzing the mechanisms that control VDJ recombination (reviewed in Alt et al., 1988; Lieber, 1991).In one set of experiments, a retroviral construct containing S, and SyZb sequences flanking a herpes simplex virus thymidine kinase (HSV-tk) gene was introduced into lymphoid and nonlymphoid cell lines that did not express their endogenous tk genes (Ott et al., 1987; Ott and Marcu, 1989). When bromodeoxyuridine (BUdR) was added to the cells to select for cells that had deleted the HSV-tk gene, there was a much higher frequency of survival among B lymphoid cell lines vs nonlymphoid cells. Analysis of surviving clones showed that deletion of tk in the B lymphoid cell lines involved switch sequences. In addition, if the construct contained only the S, region 5' of the tk gene and no S region 3' of tk, recombination was not seen. These findings indicate that at least some component(s) of S region recombination are mediated by a B lymphoid-specific mechanism and that switch sequences are important in the recombination process. More recently, an in vitro switch assay has been developed in which the recombination construct is transiently transfected into B cells stimulated with lipopolysaccharide (LPS) and subsequently analyzed for recombination, utilizing a bacterial selection system (Leung and Maizels, 1992). The authors observed increased levels of recombination if a transcriptional element (enhancer)was present in the plasmid construct. However, most of the recombination observed in the plasmid did not involve the S, sequences, and it remains unclear whether the mechanism of this recombination is the same as normal switch recombination. In addition, the tissue specificity of this system has not been examined. Hopefully, further development of this transient transfection system will allow further definition of the switch recombinase system. 5. Znvolvement of DNA Replication
There are several lines of evidence that DNA replication is associated with switch recombination. This link between DNA replication and switching is supported by studies in which inhibition of DNA replication in B cells with aphidicolin inhibits switching (Chu et al., 1992a). Another link between switching and DNA replication is that
238
ROBERT L. COFFMAN ET AL.
agents that are mitogenic to B cells, such as LPS and anti-IgM antibodies, stimulate switching (reviewed in Severinson et al., 1982). Furthermore, s, deletion seems to occur during the first S phase in B cells stimulated with LPS (Kenter and Watson, 1987). One of the more intriguing insights into the mechanism of recombination came from a study of the switch products in the 1.29 cell line. When daughter cells of a single switching event were examined, the five separate clones isolated could be divided into two groups that differed in a number of small mutations (either substitutions or deletions) at the recombination join (Dunnick et al., 1989). These mutations likely arose during the DNA replication that occurred early after recombination. Interestingly, the mutations appeared to arise on only one strand of DNA (Dunnick and Stavnezer, 1990). In other studies, mutations have also been found to occur at the sites of switch recombination (Ott et al., 1987; Winter et al., 1987; Shapira et al., 1991). This high rate of mutation suggests that the DNA replication associated with switch recombination may be error prone. 6. Sequential Switching In addition to switching between S, and downstream S regions, the study of deleted switch circles (Yoshida et al., 1990) provided clear evidence that switching can also oc:cur between two downstream S regions, as had been inferred from a number of previous studies (Davis et al., 1980; Kataoka et al., 1981; Nikaido et al., 1982; Schultz et al., 1990). This indicates that, during differentiation, B cells can switch sequentially to two or more different isotypes. Although sequential switching could potentially occur during switching to any region, it seems to be prominent in cells that switch to IgE. The analysis of circles containing deleted S, isolated from mice infected with helminth parasites revealed that most IgE-producing cells arose by sequential switching, first to S y l , then to S, (Yoshida et al., 1990). In addition, others have found that the switch regions found in the expressed IgE heavy chain locus after switching often contain SyI sequences, remnants of a sequential switch (Siebenkotten et al., 1992; Mills et al., 1992). These molecular data are consistent with several experiments that have characterized the relationship between the cells that produce IgCl and IgE. When mIgM+ B cells are cultured with LPS and interleukin 4 (IL-4), mIgGl+/mIgE+ double-positive staining cells appear after 3 days of culture and these cells secrete predominantly IgE on reculture (Snapper et al., 1988b; Mandler et al., 1993). The high frequency of sequential switching from y l to E has led to the
IMMUNOGLOBULIN ISOTYPE SWITCHING
239
suggestion that this pathway is required for IgE production. However, preliminary studies of mice in which the Z y l region has been removed by gene targeting reveal that, despite the inability of these mice to switch to y l (Jung et al., 1993), they produce normal levels of IgE when infected by parasites (S. Jung and A. Radbruch, personal communication, 1992). These results suggest that, although switching to yl may commonly precede switching to E , it is not an essential prerequisite.
7 . Proteins Znuolued in Switch Recombination Studies have sought to identify and characterize DNA-binding proteins specific for switch region sequences, with the hope that these will include components of the switch recombination mechanism. Some of these experiments have demonstrated that transcription factors bind to switch sequences in uitro. For instance, the octamerbinding protein (Oct) binds to a sequence within Syl (Schultz et al., 1991)and NFKBbinds to Sy3 (Wuerffel et al., 1992).Other investigators have identified B cell-specific nucleoprotein complexes by using probes from various switch regions. Waters et al. (1989) described a B cell-specific complex, termed SaBP, that binds to two sequences within the a locus. Binding of this factor is competed with probes from S, and S,I. Xu et al. (1992)showed that this same factor, termed Sp-B1, bound to another sequence within S,. It appears that the recently identified BSAP transcription factor (Barberis et al., 1990) may be responsible for all of these binding activities (Liao et al., 1992; Xu et al., 1992; Adams et al., 1992).In fact it appears that this factor can bind to multiple regions within the heavy chain locus (Liao et al., 1992).The BSAP factor, which is encoded for by the Pax-5 gene (Adams et al., 1992), can act as a transcription factor at the CD19 gene. The same factor appears to be essential for the production of the germline E transcript (see below) and also binds near the promoter of the germline y2a transcript (Liao et al., 1992). Although this factor has an important role in the transcription of several B cell-specific promoters, it is not yet clear whether the binding of BSAP to the switch sequences has any physiological role or is simply a by-product of its promiscuous binding properties. Two groups have identified factors that bind to different sequences within S, and are present in extracts of B cells stimulated with LPS (Wuerffel et al., 1990; Williams and Maizels, 1991). Although the sequences to which these factors bind have been defined, there is no direct evidence yet that these binding activities are related to switch
240
ROBERT L. COFFMAN ET AL.
recombination. Further understanding should come from the cloning of the proteins responsible for these binding activities and the development of better in uitro switching assays. AND GERMLINE CHTRANSCRIPTION D. DIRECTEDSWITCHING
Isotype-specific immune responses could potentially be generated either by stochastic class switching to random isotypes, and subsequent selection of B cells that are able to express a particular type of immunoglobulin, or by directed switching in which particular recombination events are directed by the immune response. It appears that the major mechanism by which mice generate specific responses is to direct switch recombination to specific regions. The best evidence for directed switching in mice comes from studies of the status of the nonproductive heavy chain allele in B lymphoid cells (reviewed in Radbruch et al., 1986a; Schultz et a%.,1990). Because these genes do not encode a protein product, selection should not play a role in these rearrangements. Several studies have shown that, in both spleen cells and hybridomas that produce IgGl ,u p to 60% of the cells had switched to the S,1 region on the nonproductive allele (Radbruch et al., 1986b; Winter et al., 1987).The idea of directed switching is also supported by studies of c-myc protooncogene translocations in B cell tumors. The immunoglobulin isotype produced hy these tumors frequently correlated with the S region to which the c-myc translocation has taken place at the nonproductive CH locus (reviewed in Cory, 1986).Additional evidence for directed switching has come from studies of several murine cell lines (discussed below) that spontaneously switch to a restricted set of isotypes in culture. [n similar experiments examining human B cell tumors, however, there has been no correlation between the S region recombinations on the productive and nonproductive allele (Borzillo et al., 1987; Webb et al., 1985). Although the total number of examples so far analyzed is small, these studies suggest differences in the mechanisms of control of switch recombination between mouse and human.
1, Germline C,, Transcripts At least two distinct, but not mutually exclusive, mechanisms have been proposed for the regulation of switch recombination to specific isotypes. The sequence differences found within the switch regions suggest that there could be switch region-specific recombinases that mediate recombinations specifically to certain switch regions (Davis et al., 1980).A second mechanism, termed the accessibility model, proposes that there is a nonspecific S region recombination complex (S
IMMUNOGLOBULIN ISOTY PE SWITCHING
24 1
recombinase) that derives its specificity not from the S region sequence, but from the “accessibility” of different S regions (Yancopoulos et al., 1986; Stavnezer-Nordgren and Sirlin, 1986). A critical distinction between these two models is that factors that regulate switching to certain isotypes (Thcells, cytokines, mitogens) would, in the first case, regulate the expression of isotype-specific recombinases, whereas in the second case they would regulate the accessibility of specific S regions for recombination. Support for the accessibility model came first from the analyses of CH transcription in tumor cell lines that exhibit class-switch recombination in culture. Abelson murine leukemia virus (AMuLV)transformed pre-B cell lines that spontaneously switch in culture from p to y2b by S region recombination produce a truncated C,,zb-hybridizingtranscript (termed y2b germline transcript) that does not hybridize to a VH probe, but does hybridize to a probe from the germline genomic region 5‘ to Sy2b (a region deleted after switching to y2b) (Yancopoulos et al., 1986; Lutzker and Alt, 1988a). The y2b germline transcript consisted of a 5’ exon, termed Zy2b, spliced to a complete Cy2bgene (Fig. 2). Some AMuLV-transformed cell lines have also been shown to switch to y3 and these lines can produce a germline y3 transcript that is analogous in structure to the y2b germline transcript (Rothman et al., 1990b). In these lines, LPS stimulation increased the level of the y2b and y3 germline transcripts prior to the induction of switching to these genes (Lutzker et al., 1988; Rothman et al., 1990b). I S
CH1
CH2
CH3s
m
DNA
I
alternate mRNA splicing, . polyadenylation
germline transcript
FIG. 2. General structural features of germline CH transcripts. The final mRNA consists of I region sequences spliced to the 5’ end of the CH gene in a manner analogous to the splicing of V H to CH in productive immunoglobulin mRNA transcripts. [Reprinted with permission from Schultz and Coffman (1992). Copyright CRC Press, Boca Raton, FL.]
242
ROBERT L. COFFh4AN ET AL.
Similar observations were made with the 1.29 lymphoma cell line that can spontaneously switch in culture from p to a.When 1.29 cells are induced with either LPS or anti-idiotype antibody, switching to a is markedly increased and some switching to y2a and E is seen (Stavnezer-Nordgren and Sirlin, 1986). 1.29 cells produce a variety of germline C, transcripts and the level of these germline C, transcripts in subclones of the 1.29 line correlates with the ability of the subclone to switch to a. When induced, this line also produces Cyeaand C , germline transcripts, but not transcripts from the CH regions to which it does not switch. In addition, the CH genes to which 1.29 switches are hypomethylated, as compared to other CH regions, a phenomenon likely related to the active transcription of these regions (Stavnezer et al., 1988a). A large amount of work has centered on identifying and characterizing the expression pattern of germline CH transcripts (see Section II,D,3). These studies have demonstrated a strict correlation between the expression of specific CH germline transcripts and subsequent switching to the corresponding CH gene. A few experiments using gene targeting to alter the heavy chain locus in cells have directly addressed the role of germline transcripts or transcription in the regulation of class switching. In one instance, the Z,,1 region was removed from the heavy chain locus by homologous recombination. Heterozygous mice with one wild-type and one mutant allele did not switch to y l on the mutant allele but were able to switch on the normal allele (Jung et al., 1993). This suggests that sequences within the ZY1 region are necessary for switching to yl, with the implication that the promoter activity associated with this region is a necessary component for targeting the region for switching. Preliminary studies of murine B cells generated from ES cells lacking the I,& region, also show that the Z region DNA is important in class switching (F. Alt, personal communication, 1992). Homologous recombination has also been used to mutate the immunoglobulin heavy chain locus of the 18.81A20 murine pre-B cell line (Xuet al., 1993). The heavy chain locus of the mutated cell line contains the immunoglobulin heavy chain enhancer and immunoglobulin variable region gene promoter in place of the LPSIIL-4-responsive germline E promoter. When the parent cell line is cultured in the presence of both LPS and IL-4, germline E transcription is induced prior to class switching to E . The mutant cell line constitutively transcribes the E locus in the absence of IL-4. Strikingly, this constitutive germline E production allows the mutant cell line to switch to E in the absence of IL-4 whereas the parental cell line requires IL-4 to switch
IMMUNOGLOBULIN ISOTYPE SWITCHING
243
to E . This result strongly argues that germline E transcription plays a direct role in switching to the E locus and accounts for the specific actions of IL-4 in IgE regulation.
2. Role of Gemline Transcription The studies summarized above have demonstrated a clear relationship between germline CHtranscription and isotype switching. There are several lines of evidence, from both prokaryotes and eukaryotes, that indicate that transcription may be an essential part of the recombination process. Studies of VDJ recombination have correlated transcription of specific variable region gene segments with developmentally controlled recombination of those genes (reviewed in Blackwell and Alt, 1989b) and have indicated that the heavy chain enhancer is essential for VDJ recombination at the heavy chain locus (Ferrier et al., 1990). The possibility that transcription itself may contribute to the process of recombination is also supported by studies in lower organisms. Double-strand breaks, which are the initiating lesion for switching of yeast mating type, occurred in only those mating type genes that were actively transcribed (Strathern et al., 1982). In yeast, homologous recombination between two adjacent genes increased when they were being actively transcribed (e.g., Keil and Roeder, 1984; Thomas and Rothstein, 1989). In addition, some site-specific recombinations in bacteria have been shown to require an enhancer (reviewed in Craig, 1985). In the above systems, the mechanisms by which transcription stimulates recombination are not yet known. One model proposes that the DNA breaks mediated by RNA polymeraseassociated topoisomerases stimulate recombination. Another suggests that the supercoiling associated with transcriptionally active DNA promotes recombination (reviewed in Thomas and Rothstein, 1991). 3. Possible Roles for Germline RNA Transcripts Although transcription itself may be sufficient to target recombination, another, not mutually exclusive, possibility is that the germline CH transcripts themselves may have a functional role in class switch recombination. The murine p (Lennon and Perry, 1985), y3 (Rothman et al., 1990b), y l (Xu and Stavnezer, 1990), y2b (Lutzker and Alt, 1988a), y2a (W. Dunnick, pcsonal communication, 1992), E (Rothman et al., 1990a; Gerondakis, 1991), and a (Lebman et al., 1990b; Radcliffe et al., 1990; Gaff and Gerondakis, 1990) germline transcripts along with the human p (Neale and Kitchingman, 1991), y l (Sideras et al., 1992), y3 (Sideras et al., 1992), a (Nilsson et al., 1991), and E (Gauchat et al.,
244
ROBERT L. COFFMAN ET AL.
1990) germline transcripts, have been cloned and share many structural features (Fig, 2). In general, germline transcripts initiate in regions (termed I regions) that are found 5‘ of the corresponding switch region. The transcripts initiate from multiple sites over this region, probably due to the lack of TATA box and CCAAT motifs at the promoters. The initial transcription proceeds through the S region and CH gene. The transcript is processed to a stable form, which consists of a 5’ exon derived from the corresponding I region spliced to the CH gene in much the same way that the V gene is spliced to CH in transcripts encoding a complete heavy chain. This conserved structure raises the possibility that the transcripts themselves perform a functional role. Interestingly, the I, exon of the gennline p transcript is still present after most switches, and transcripts that contain the I, exon spliced to a C y z b gene have been identified in cells that have switched to y2b (S. Li and F. Alt, personal communication, 1992). These transcripts could play a role in sequential switching from y2b to downstream regions. The 5’ exons of different germline CHtranscripts do not contain significant homology nor are their sequences conserved between human and murine I regions. Nearly all of these transcripts contain stop codons in the open reading frame of the CH gene and therefore they cannot encode large proteins. The exception may be the murine germline y l transcript, which is reported to contain an open reading frame for 27 amino acids in frame with the CH region (Goodman et al., 1993). In addition, there are short open reading frames within many I exons and it has been suggested that part of the germline I, transcript may be translated into a short peptide (Radcliffe et al., 1990). Germline transcripts could also play a direct role in class switching by directly interacting with either the recombination enzyme complex or with the S region chromatin. Although the stable germline transcripts have spliced out the S regions, they represent the mature product of primary transcripts that include S region sequences. Work has revealed that, in uitro, sequences from within the S, region can be found in a unique intramolecular triple-strand complex that was dependent on the presence of an RNA molecule (Reaban and Griffin, 1990). Potentially, germline CH transcripts might play a role in stabilizing switch region DNA in alternative chromatin structure, thereby rendering it accessible for recombination. However, in the experiments presented in this article the RNA present was synthesized in the opposite orientation from the documented germline a transcript (Stavnezer, 1991).This study suggests that the interaction of the germline transcript with S region chromosomal DNA could be important in a g e t i n g switch recombination. It is also possible that germline RNA
IMMUNOGLOBULIN ISOTYPE SWITCHING
245
could interact with other regions of DNA or with protein molecules involved in switch recombination. Another potential role for germline CHtranscripts is as a substrate for RNA trans-splicing. Trans-splicing of a germline CH transcript to a VDJ-C, RNA could account for several previous results, including switching without DNA recombination and double isotype-producing B cells (Section 11,E). If trans-splicing normally occurs between germline transcripts and RNA encoding the heavy chain protein, then these double isotype-expressing cells would be precursors to switched cells and therefore could potentially be targets for regulating switching.
E. EXPRESSION OF DOWNSTREAM ISOTYPES WITHOUT SWITCH RECOMBINATION During normal ontogeny B cells coexpress on their surface IgM
and IgD molecules that contain the same binding specificity (Coffman and Cohn, 1977; Pernis et al., 1977). This is accomplished without DNA recombination by differential RNA splicing of long transcripts that contain both C, and CS (Maki et al., 1981; Knapp et al., 1982). Expression of p along with other, non-8 CH isotypes, has been reported in both normal B cells and B cell tumors in which the CH locus appears in the germline configuration. However, the existence of double isotype-producing cells that have not undergone switch recombination remains controversial. In some cases, the expression of the non-p isotype may be due to its binding to Fc receptors on the p-producing cells (Katona et al., 1985). Other dual-isotype cells, especially ones that appear transiently during immune responses, may represent cells that have undergone S region rearrangement but retain VDJ-C, mRNA, which they continue to translate to IgM protein (Nossal et al., 1964). One mechanism proposed to explain nondeletional isotype switching is the differential splicing of long RNA transcripts, analogous to the mechanism of p and 6 coexpression. Evidence for long transcripts was demonstrated with a sandwich hybridization technique that revealed the presence of large nuclear RNA that hybridize to two different CH genes in cells that were dual-isotype producers (Perlmutter and Gilbert, 1984). A more recent model for isotype switching without deletion suggested that the trans-splicing of VDJ-C, RNA to germline CH transcripts was responsible for this phenomenon (Shimizu et al., 1989; Lutzker and Alt, 1988b). Unlike cissplicing, which involves the processing of a single RNA molecule, trans-splicing involves the splicing of two separate RNA species (re-
246
ROBERT L. COFFMAN ET AL.
viewed in Agabian, 1990). This process has been well characterized in trypanosomes and nematodes. Mammalian cells transfected with the namatode substrates for trans-splicing have been shown to transsplice RNA (Bruzik and Maniatis, 1992). Two different B cell systems in which trans-splicing may occur have been characterized. The BCLI. B cell line was one of the first examples of a clonal line that expressed two immunoglobulin isotypes without apparent deletion of the intervening C H genes (Chen et al., 1986). It was shown that clones of BCLl that express both IgM and IgGl contained both VDJ-C, R.NA and germline yl transcripts (Nolan-Willard et al., 1992). Nuclear run-off experiments revealed that transcription in these cells was not contiguous from the C , to the C y l locus but was detectable only at high levels around the C , gene, at the Cy3 gene, and around the C y l loci, suggesting that long transcripts were not present but that the germline y l transcripts were being trans-spliced to the VDJ-C, RNA. The other extensively characterized example of dual-isotype expression without recombination is the finding that the VDJ region of a human VDJ-C, transgene can be found in the IgG fraction of transgenic mice (Shimizu et al., 1989, 1991). One group has shown that transgenic murine heavy chain genes can undergo interchromosomal recombination to the endogenous heavy chain locus (Durdik et al., 1989). However, in transgenic mice with the human heavy chain gene there did not seem to be detectable interchromosomal recombination. In these mice, detection of transgenic VDJ with different murine C H regions correlates with the induction, by cytokines, of the corresponding germline transcripts (Han et al., 1991), suggesting that germline transcripts were being trans-spliced to a VDJ-C, RNA. Although trans-splicing could explain the above findings it should be noted that biochemical proof of trans-splicing in B cells has remained elusive. In summary, there is compelling evidence that the transcriptional activation or increased accessibility of a CH locus is a prerequisite to switch recombination itself. Switch recombination to the various C H loci does not occur with a fixed probability, but is regulated and can be directed to one or more specific CH. Studies with B cell tumors that can be induced to switch to specific isotypes suggest that the control of transcriptional activation is the means by which this regulation occurs. It is the regulation of this transcriptional activation by T cells and the cytokines they produce that is the subject of the remainder of the article.
IMMUNOGLOBULIN ISOTYPE SWITCHING
247
111. Regulation of Switching by Interleukin-4
A. INTERLEUKIN-4-STIMULATEDIgE AND IgG, PRODUCTION Interleukin-4 was the first cytokine shown to differentially regulate the expression of specific isotypes, due largely to the pioneering efforts of the laboratories of E. Severinson and E. Vitetta. Both groups initially described the ability of certain T cell supernatants to enhance IgGl and inhibit IgGZb and IgG3 production when added to cultures of LPS-stimulated murine B cells (Isakson et al., 1982; Bergstedt-Lindqvist et al., 1984). Isakson et al. (1982) presented evidence that this activity stimulated significant IgGl production from mIgG1- B cells, suggesting that the activity promoted isotype switching. Purification of the active component of these supernatants revealed a single factor that had all three activities and copurified with the B cell growth factor. IL-4 (at that time called BSF-1) (Vitetta et al., 1985; Sideras et al., 1985), and a monoclonal antibody to IL-4 neutralized the IgG1-enhancing activity (Vitetta et al., 1985). Further proof of the identity of the two activities came with the cloning of the IL-4 molecule (Noma et al., 1986; Lee et al., 1986), in one case using the IgG1-enhancing activity to detect cDNA clones encoding the active molecule (Noma et al., 1986). In a separate study, Coffinan and Carty (1986) found that supernatants from some, but not all, T h clones could induce a 100- to 1000fold increase in IgE production in cultures of LPS-stimulated B cells. All supernatants that contained the IgE-enhancing activity also enhanced IgG1, and the two activities copurified in initial biochemical characterizations. Proof that the IgE-enhancing activity was also mediated by IL-4 was again obtained by comparison with highly purified IL-4 (BSF-l), and by neutralization with a monoclonal anti-IL-4 antibody (Coffman et al., 1986). The IgE-enhancing activity of IL-4 was also confirmed with recombinant IL-4 (Table I) (Lee et al., 1986; Coffman et al., 1987; Snapper and Paul, 1987). T cell supernatants that contained the IgG1- and IgE-stimulating activities frequently inhibited IgG3 production 5- to 10-fold, and IgGzb production to a lesser degree (Bergstedt-Lindqvist et d.,1984; Layton et al., 1984; Coffman and Carty, 1986). These activities were also shown to be mediated by purified natural and recombinant IL-4 (Coffman et al., 1986, 1987; Snapper and Paul, 1987). All of these activities of IL-4 were the result of direct actions on the B cell, rather than indirect action on a T cell or accessory cell (Coffman and Carty, 1986; Snapper and Paul, 1987). The critical period for IL-4 to act was
248
ROBERT L. COFFMAN ET AL.
TABLE I EFFECTOF INTERLEUKIN 4 ON L I ~ ~ L Y S A C C ~ R I D E - S SPLENIC TIM~A B CELLS" ~D Supernatant levels at day 7 (ng/ml) Addition
IgE
IgA
IgC,
IgC%
IgGPb
IgG,
IgM
None rIL-4'
<1 920
89 120
1,080 22,300
84 66
761 248
3,460 1,040
132,000 95,000
'Reprinted with permission from Schultz and Coffman (1992). Copyright CRC Press, Boca Raton, FL. rlL-4 (500U/ml) was added to cultures 1 day after stimulation with LPS.
1to 3 days after LPS stimulation, which is at least 1 to 2 days prior to the onset of significant immunoglobulin production in this culture system ( C o h a n and Carty, 1986; Snapper et al., 1988a). Thus, in order to stimulate IgGl and IgE production, IL-4 had to be present before, but not during, the time these isotypes were being actively produced and secreted, a timing consistent with stimulation of an isotype switch to these two isotypes. Interleukin-4 was both necessary and sufficient for the induction of IgE and the enhancement of IgGl in LPS-stimulated B cell cultures and IL-4 accounted for essentially all of the IgE- and IgG1-enhancing activity found in Th supernatants (Coffman et al., 1986). To date, no cytokine has been described that further enhances the activity of optimal concentration of IL-4, although IL-5 will significantly enhance the effect of suboptimum IL-4 concentrations, both in mouse (R. L. C., unpublished observations, 1988) and in humans (Pene et al., 1988~).
B. INTERLEUKIN-4 ENHANCES ISOTYPE SWITCHING TO IgE AND IgG, Studies on the mechanism of IgE and IgGl enhancement centered originally on the question of whether IL-4 increased the frequency with which B cells switched to these isotypes or whether it was a selective growth or differentiation factor for B cells that had switched to those isotypes independent of IL-4. The first indications that IL-4 might enhance isotype switching were from experiments showing that IL-4-induced IgGl production came from mIgM+ B cells with undetectable levels of mIgG1 at the initiation of the culture (Isakson et aZ., 1982; Layton et al., 1984; Yuan et al., 1985). Although this result indicated that the IgG1 production was the consequence of a switch that occurred during the culture period, it did not distinguish between the two mechanisms listed above. To address that issue, several groups have performed limiting dilution cultures of LPS-
IMMUNOGLOBULIN ISOTYPE SWITCHING
249
stimulated B cells and demonstrated that IL-4 significantly increased the frequency of precursors that gave rise to IgG1- and/or IgEsecreting cells (Layton et al., 1984; Bergstedt-Lindqvist et al., 1988; Savelkoul et al., 1988).A major limitation of this technique, however, was the low cloning efficiency in these cultures. It was formally possible that different subsets of B cells responded to LPS and LPS plus IL-4. More definitive data were obtained in a microculture system in which well over 50% of individual B cells gave rise to clones of imniunoglobulin-secreting cells (Lebman and Coffman, 1988). In this system, single B cells were stimulated polyclonally by a rabbit Igspecific Th2 clone, with rabbit anti-mouse IgM antibodies acting as the antigen. In this system the majority of B cell clones (78%) produced IgE in the presence of optimum concentrations of IL-4. The addition of anti-IL-4 antibodies substantially reduced both the frequency of IgE-producing clones and the amount of IgE produced per clone. Neither addition had a measurable effect on the cloning frequency. These results demonstrated clearly that IL-4 increased switching to IgE rather than preferential growth or differentiation of cells that had already switched to IgE. Further evidence that IL-4 induces isotype switching has come from studies of surface immunoglobulin isotypes of B cells in LPSstimulated cultures. In general, many B cells express mIgG1, mIgE, or both after 4 days of culture with LPS and IL-4, whereas few B cells cultured without IL-4 expressed these isotypes (Isakson et al., 1982; Layton et al., 1984; Savelkoul et al., 1988; Yuan et al., 1985; Snapper et al., 1988b; Radbruch et al., 1986b). Most of these mIgG1-positive cells had gene rearrangements at one or both of the C,1 loci (Winter et aE., 1987; Radbruch and Sablitzky, 1983). The mIgGl-positive cells were derived from B cells that were positive for mIgM/mIgD, but negative for mIgG, (Isakson et al., 1982; Layton et al., 1984; Yuan et al., 1985; Radbruch et al., 1986b).At certain times in these cultures, a significant number of cells coexpress mIgGl and mIgE and these cells may represent cells undergoing a sequential switch from IgGl to IgE (Snapper et al., 1988b; Mandler et al., 1993).
c. ROLE OF INTERLEUKIN-4 I N T CELL-DEPENDENT
ISOTYPESWITCHING Most isotype switching occurs in response to T-dependent, rather than T-independent, antigens and there is a considerable evidence, summarized in the introduction, that Th cells stimulate and regulate most isotype switching. Yet, essentially all of the basic in vitro studies on the role of 1L-4 and other cytokines on the process of isotype
250
ROBERT L. COFFMAN ET AL.
switching were done using B cells stimulated with the mitogen LPS. Lipopolysaccharide is a convenient and reproducible for murine B cells, but it is not easy to directly extrapolate results ot this system to B cells stimulated by antigen and helper T (Th) cells, for two reasons. First, LPS alone stimulates both proliferation and differentiation, whereas these are separable processes requiring distinct signals from T h cells. Second, LPS itself induces isotype switching, preferentially to IgG, (van der Loo et al., 1979; Radbruch et al., 1986a). For these reasons, it is important to investigate the control of switching in T h stimulated B cells. The requirement of IL-4 for switching to IgE has been demonstrated in a variety of Th-dependent systems, both in vitro and in vivo. Only clones of the IL-4-producing Th2 subset were able to stimulate primary IgE responses, the Thl clones, which produce interferon y (IFN-y), but not IL-4, inhibited IgE production (Coffman et al., 1988). IgE production induced by Th2 clones could be almost completely inhibited by anti-IL-4 antibodies, whereas antibodies to other cytokines caused no inhibition. Large IgE responses could be induced by Thl clones, but only in the presence of anti-IFN-y antibodies and exogenous IL-4. The role of IL-4 in IgE switching has been confirmed in viuo as well. Thus, anti-IL-4 inhibited 95-99% of the primary IgE response to a number of T-dependent stimuli, including helminth infections (Finkelman et al., 1986), protein antigens (Finkelman et al., 1988b), and anti-IgD antibodies (Finkelman et al., 1986, 1988a). The IgE responses to secondary challenge with some of these antigens, however, is much less inhibited by anti-IL-4 treatment (Finkelman et al., 1988b), as would be expected if part of the secondary response came from memory B cells that had already switched to IgE. The finding that IL-4 “knockout” mice have no detectable IgE production, even after helminth infection, suggests that no other cytokine made during this response can induce significant switching to IgE (Kuhn et al., 1991). That these IL-4-dependent IgE responses involve switch recombination has been shown by the isolation from responding mice of circular extrachromosomal DNA pieces containing the deleted intervening CH loci (Yoshida et al., 1990). In contrast to IgE, Th-stimulated switching to IgGl is relatively independent of IL-4. IgGl responses induced by Th2 clones are partly, but not completely, inhibited by anti-IL-4 (Coffman et al., 1988; Stevens et al., 1988) and many T h l clones can induce significant IgGl responses (Coffman et al., 1988). In viuo anti-IL-4 has little effect on primary IgGl responses to a variety of antigens, even
IMMUNOGLOBULIN ISOTYPE SWITCHING
251
though the antibody treatments were sufficient to inhibit virtually all of the IgE responses (Finkelman et al., 1990). Furthermore, IgGl responses are still prominent in IL-4 “knockout” mice, although their magnitude is reduced severalfold from control mice (Kuhn et al., 1991). The implication of these observations is that there may be other factors or signals that can induce switching to IgG1, albeit not as efficiently as IL-4. Although IL-4 is required for switching to IgE and is a potent stimulus for switching to IgG1, there is little evidence that IL-4 can cause either switch unless the B cells are appropriately activated. In mitogen-stimulated cultures, the mitogen itself (e.g., LPS) provides the B cell activation signal. It is now clear, however, that B cell activation induced b y Th cells is mediated by direct contact, not by soluble factors (reviewed by Parker, 1993).This contact-mediated activation signal could be delivered by activated, but not resting, T cells and induced significant proliferation, but virtually no differentiation to immunoglobulin production (Parker, 1993).The addition of IL-4 to Th-activated B cell cultures resulted in significant immunoglobulin production and significant switching to IgE and IgGl (Coffman et al., 1988).This contact-mediated activation signal did not require living T cells and could be effectively delivered by plasma membranes isolated from activated T h cells (Brian, 1988; Hodgkin et al., 1990; Noelle et al., 1991). Addition of IL-4 to B cells stimulated with Th membranes also induced immunoglobulin production, the majority of which consisted of IgE and IgG, (Hodgkin et al., 1990). Both Thl and T h 2 clones, or membranes from either, appear to deliver a comparable activation signal that can lead to efficient switching to IgGl and IgE in the presence of IL-4 (Coffman et al., 1988; Hodgkin et al., 1991). Although IL-4 is able to induce switching to IgE in B cells activated by T h contact or by a variety of mitogens, Snapper et al. (1991) have shown that IL-4 induces IgG1, but not IgE, switching in B cells activated with anti-6 antibodies conjugated to dextran polymers. This contact-mediated activation signal can itself influence isotype switching. Stimulation of resting B cells with activated Thl membranes induces significant germline transcription from the y l locus, but not from the 73, y2b, E, or a CH loci (Schultz et al., 1992). This germline y l induction is not inhibited by anti-IL-4 and occurs in B cells from IL-4 “knockout” mice, suggesting that this T h contact signal is a second switch factor for IgGI. In the absence of any cytokines, however, this germline transcription does not appear to lead to switch recombination, suggesting the T cell-mediated signal is in-
252
ROBERT L. COFFMAN ET AL.
sufficient to induce some part of the recombinase mechanism (Schultz and Coffman, 1993). Under normal conditions, however, the T h contact signal would be linked directly to cytokine production, and would lead to preferential recombination to C,1. This switch-inducing activity associated with the Th contact-mediated activation signal may explain the many observations of IL-4-independent IgGl switching. It should be noted that a second group, using a similar strategy, was unable to detect germline y l transcripts in the absence of IL-4 (Noelle et al., 1992). The nature of this contact-mediated signal has been clarified with the cloning of the T cell ligand (CD40L) for the B cell surface molecule, CD40 (Armitage et al., 1992). CD40L is rapidly induced on T cell activation and is expressed by both T h l and Th2 clones. It appears to deliver an activation and proliferation signal that is essentially equivalent to that delivered by activated T h cells or membranes. The addition of IL-4 to B cell cultures stimulated via the CD40-CD40L interaction leads to substantial IgE production (Armitage et al., 1992; Spriggs et al., 1992). These results are consistent with the many reports (discussed in Section VI) that antibodies to CD40 activate human B cells to switch to IgE and IgG4 in the presence of IL-4 (Jabara et al., 1990; Rousset et al., 1991). D. MOLECULAR MECHANISMS OF INTERLEUKIN-4-INDUCED SWITCHING The evidence, presented in Section 11, suggests that germline transcription of a C H gene is a prerequisite for switch recombination to that isotype. This implies that factors that induce or inhibit germline transcription can regulate the frequency of switching to that isotype. Studies on the mechanisms by which IL-4 regulates switching to several isotypes, provide us with the clearest examples of this. Interleukin-4 was first shown to inhibit germline transcription of the y2b gene in LPS-stimulated B cells (Lutzker et al., 1988), and this correlated well with the previous reports of an inhibition of IgGzb production in such cultures (Bergstedt-Lindqvist et al., 1984; Coffman and Carty, 1986). Subsequent studies have shown that IL-4 induces germline E (Rothman et al., 1988; Stavnezer et al., 1988a) and y l (Berton et al., 1989; Esser and Radbruch, 1989) and inhibits germline y3 (Rothman et al., 1990b; Severinson et al., 1990) in similar LPSstimulated cultures. Induced germline transcripts can be detected less than 24 hours after stimulation with LPS and IL-4, and precede by 1-2 days both the appearance of VDJ-containing transcripts of the same isotype (Rothman et al., 1988; Berton et al., 1989) and actual
253
IMMUNOGLOBULIN ISOTYPE SWITCHING
switch recombination (Chu et al., 1992b). Because switched CH loci can no longer make germline transcripts, the levels of germline transcripts decrease as VDJ-containing transcripts with the same CH increase. Thus, in LPS-stimulated B cell cultures, all changes in the frequency of switching induced by IL-4 are foreshadowed by parallel changes in germline transcription. A diagram of this sequence of events, using IL-4-induced IgE switching as an example, is shown in Fig. 3. Although other modes of B cell stimulation have not been as thoroughly studied as LPS, IL-4 has been shown to have the expected effects on germline y l and E transcription in B cells stimulated with Th clones (Berton and Vitetta, 1992) or Th membranes (Schultz et d., 1992). Along with germline transcription, IL-4 has also been shown to induce hypomethylation (Burger and Radbruch, 1992) and to inSy3 Cy3
VDJ Sp Q DNA mRNA
nRNA
Sy2bCy2b
m
I
VDJ Q
Sy3 Cy3
VDJ Sp Q
DNA
Syl Cyl
Syl Cyl
SyZaCy2a
k& 0
Sa Ca
~ S Q E
Sa Ca
LPS + IL-4
Sy2bCy2b
Sy2aCy2a
% m\ 0 0 0 0 0
VDJQ
\
\ \ 0
0
0
V D J $ S E ~ H ~Sa Ca
mRNA
kck
b
\
m VDJ CE
0
0
0
0
0
0
+
FIG.3. Sequence of events in IL-4-induced switching to IgE in the mouse. Interleukin 4 first induces germline E transcripts in activated B cells. This is followed by switch recombination to the active C . locus, which involves looping out and deletion of all intervening CH genes. [Reprinted with permission from Schultz and Coffman (1992). Copyright CRC Press, Boca Raton, FL.]
254
ROBERT L. COFFhlAN, ET AL.
duce a DNase I-hypersensitive site near the site of initiation of germline y l transcription (Berton and Vitetta, 1990; Schmitz and Radbruch, 1989). Further understanding of the molecular mechanisms by which exogenous factors, such as IL-4, regulate germline CH transcription requires characterization of the cis-responsive promoter elements that control germline CH transcription. Analysis of the sequences upstream of the initiation sites of germline transcripts has revealed the presence of regions highly conserved in mammalian evolution (Mills et al., 1990). For instance, there is a region of homology present upstream of the murine y2b and y3 I regions that was also found upstream of the human y l and y3 regions (Sideras et al., 1989; Rothman et al., 1990b). In addition, there are conserved sequences around the initiation sites of the murine and human germline E transcripts (Rothman et al., 1990a; Gerondakis, 1991; Gauchat et al., 1990) and human and murine a transcripts (Nilsson et al., 1991; Lebman et al., 1990b; Radcliffe et al., 1990). The conserved regions upstream of the initiation sites of the germline transcripts are likely candidates for cytokine-responsive transcription elements. The promoters that control transcription at the murine e, yl., and a loci have been characterized and contain some of these conserved elements. In addition, an IL-4-responsive enhancer has been described at the human y3 loci (Kuze et al., 1991). A cis-controlling region responsible for the LPS/IL-4 induction of germline E transcripts has been identified and characterized (Fig. 4).
L
conserved
region
Germline E Promoter FIG.4. Control regions of rnurine germline C. transcription.
IMMUNOGLOBULIN ISOTYPE SWITCHING
255
These studies show that a 179-base pair (bp) minimal element at the E locus can impart LPS/IL-4-inducible transcription to a heterologous reporter gene. This region includes sequences that are highly conserved between human and mouse. The induction imparted by this 179bp region is sensitive to the level of IL-4 used, and can be downregulated by IFN-y (P. Rothman, unpublished observations, 1992), in a pattern similar to the regulation, by IL-4 and IFN-y, of endogenous germline E transcripts (Rothman et al., 1988; Gerondakis, 1991; Stavnezer et al., 1988b; Severinson et al., 1990). Electrophoretic mobility shift assays have identified several nuclear proteins that can bind to this region (Rothman et al., 1991). Recent work has shown that the transcription factor BSAP (Barberis et ul., 1990) (also termed NFHP), which is encoded by the PAX-5 gene (Adams et al., 1992), binds to an evolutionarily conserved sequence of DNA present directly upstream of the initiation site of the germline E transcripts (Laio et al., 1993)that is essential for germline transcription (Rothman et al., 1991). BSAP seems to bind to several regions within the heavy chain locus, including the region around the initiation site of the germline y2a transcript and the 3' C, enhancer (Liao et al., 1992). Although BSAP functions as an inducer oftranscription at the germline E promoter, it appears that at the 3' C, enhancer region BSAP may act as a repressor (Singh and Birshtein, 1993). Two IL-4-inducible nucleoprotein complexes, which bind to sequences at the initiation site of germline E transcripts (just downstream of the BSAP-binding site) (Rothman et al., 1991),are indistinguishable from those that bind to the BRE sequence within an IL-4-inducible enhancer at the MHC A, locus (Boothby et al., 1988). The proteins responsible for these complexes have not been cloned. It has been shown that the nonhistone chromosomal protein HMG-I(Y) (Johnson et al., 1988) can bind to this same sequence (M. Boothby, personal communication, 1992). The study of the murine y l promoter has provided further understanding of the mechanism by which IL-4 regulates switching (Xu and Stavnezer, 1992). A minimal 352-bp fragment was shown to provide promoter activity to a heterologous promoter. The transcription from this promoter was shown to be induced by phorbol myristate acetate (PMA) and synergistically increased by IL-4 in a pattern similar to that described for the germline y l transcript. Mutational analysis revealed that several elements within this region were important for the promoter activity. One of these elements, which was important for both the PMA activation and IL-4 synergistic activity, was homologous to the BRE element (Boothby et al., 1988) described
256
ROBERT L. COFFMAN ET AL.
above. In addition, mutations of several CACCC boxes (Dierks et al., 1983) present within the promoter markedly reduced both basal activity and IL-4-inducible transcription. Another group has shown that an IL-4-inducible nucleoprotein complex is formed b y using the region upstream of the initiation site of germline y l transcripts as a probe in electrophoretic mobility shift studies (Illges and Radbruch, 1992). Although the exact sequences that bind this IL-4-inducible factor were not identified, a BRE element was contained in this probe. The cloning of the IL-4-inducible factors that bind to these elements will assist in understanding the mechanism by which IL-4 controls germline CH transcription. IV. Interferon y Regulation of isotype Switching
The first indication that IFN-y could play a role in regulation of isotype switching was the observation that it inhibited IL-4 induction of both IgG, and IgE secretion (Coffman and Carty, 1986). The reciprocal roles of IL-4 and IFN-y in IgE expression coupled with the finding that IL-4 and IFN-y were produced b y different CD4' T cell subsets, Th2 and Thl, respectively, led to the model that regulation of IgE in uivo was dependent on the number and/or state of activation of the two subsets of T helper cell (Coffman et al., 1988). Interferon y can also stimulate isotype switching to certain isotypes. For example, IFN-y increases tgG2, in cultures of B cells stimulated with either LPS or dextran-conjugated anti-IgD (Snapper and Cellular studies demonstrated that Paul, 1987; Snapper et al., 1988~). IFN-y caused both a three- to five-fold increase in the number of IgGz,-expressing clones and an increase in the number of IgG2,expressing cells in those clones, suggesting that it acted to stimulate isotype switching (Snapper and Paul, 1987; Snapper et al., 1988~). Furthermore, pretreatment of mlgG- cells with IFN-y led to enhanced IgG2, secretion following the addition of LPS (Snapper et al., 1988~).It has been shown that IFN-y increases germline y2a transcripts ( J. Collins and W. Dunnick, personal communication, 1992). Taken together these findings support the conclusion that IFN-y promotes isotype switching to IgGz, in oitro in LPS-stimulated B cells. Two lines of evidence indicate that IFN-y regulates isotype expression similarly in uiuo and in vitro (Finkelman et al., 1988a). Mice injected with goat anti-mouse IgD have substantially increased levels of serum IgGl and IgE, both of which can be suppressed by the injection of IFN-y. In contrast, IFN-y stimulates the IgGz, response in these mice. In addition, immunization of mice with killed, fixed
IMMUNOGLOBULIN ISOTYPE SWITCHING
257
Brucella abortus, a potent inducer of IFN-y, enhanced IgG,, production. Analyses of the role of IFN-y in isotype switching have shown that the efyects of cytokines on isotype secretion may depend OR the mode of activation of the B cells. Interferon y has been shown to inhibit IgG3 production by LPS-activated B cells (Snapper and Paul, 1987). However, when B cells are activated with anti-IgD conjugated to dextran, IFN-y stimulates IgG3 secretion (Snapper et al., 1992). The secreted IgGa in these cultures is derived from cells that are mIgG-. In addition, IFN-y specifically increases the steady state level of germline y3 mRNA. Thus, both cellular and molecular evidence suggest that IFN-y stimulates anti-IgD dextran-stimulated B cells to switch to IgG,. V. Regulation of lsotype Switching by Transforming Growth Factor p
The initial report on the effect of transforming growth factor (TGF-P) on the pattern of isotypes secreted by LPS-stimulated B cells demonstrated that TGF-P induced a 5- to 10-fold increase in the level of IgA secretion concomitantly with a decrease of similar magnitude in the other isotypes (Fig. 5 ) (Coffman et al., 1989). Although IL-2 alone caused at best a 2- to %fold increase in IgA secretion by LPS-stimulated B cells, the addition of IL-2 to cultures containing TGF-P caused a 30- to 50-fold enhancement in IgA secretion, suggesting that the two cytokines act synergistically (Coffman et al., 1989). A similar interaction of TGF-P and IL-5 was also observed (Coffman et al., 1989; Sonoda et al., 1989). At the cellular level, there was considerable evidence that TGF-P alone stimulated isotype switching to IgA and that either IL-2 or IL-5 stimulated maturation of mIgA-expressing cells to IgA-secreting cells, but did not stimulate the actual switch to IgA (Coffman et al., 1989; Sonoda et al., 1989; Kim and Kagnoff, 1990a,b; Lebman et al., 1990a). Transforming growth factor /3 was shown to enhance IgA secretion by both mIgAPeyer’s patch cells and mIgA- splenic B cells (Coffman et aZ., 1989; Sonoda et aZ., 1989), although it inhibited IgA secretion by mIgAf Peyer’s patch cells. Limiting dilution analysis demonstrated that TGF-P increased the frequency of IgA-secreting cells (Kim and Kagnoff, 1990b). In addition, TGF-/3 alone caused mIgA- cells to express mIgA whereas IL-2 had no effect on either the number or proportion of mIgA-expressing cells (Lebman et al., 1990a). On a molecular level, TGF-/3 alone was able to induce both productive and germline a mRNAs in LPS-stimulated normal B cells,
258
ROBERT L. COFFMAN ET AL.
1000 '
/IgA
1
TGF-P (ng/mi) FIG.5. Titration of TGF-P into cultures of LPS-stimulated murine B cells. Data for each isotype are expressed as the percentage of the control response without TGF-P addition. [Reprinted with permission from Coffman et al. (1989)J.E r p . Med. 170,1039, by copyright permission of Rockefeller Univ. Press.]
and IL-2 neither induced a mRNA, nor increased the steady state level of TGF-@-induced a! mRNA (Lebman et al., 1990a,b). The germline a! mRNA transcripts in these cells were shown to consist of a 126-bp exon (I,) located approximately 1.5 kb 5' to S , spliced to the first exon of C, (Lebman et al., 1990b). Subsequent studies demonstrated that TGF-@ induced deletional switch recombination to C, (Matsuoka et al., 1990; Iwasato et al., 1992). Thus a considerable body of cellular and molecular data supports the concept that TGF-@ stimulates isotype switching to IgA. The 1.29 B cell lymphoma has been used as model for isotype switching to IgA. Prior to stimulation, mIgM+ 1.29 cells expressed germline a mRNA (Radcliffe et al., 1990). The predominant type of germline a nRNA in the 1.29 cell line differs from the one induced by TGF-@in LPS-stimulated normal B cells. In 1.29 cells, I, is either 460 or 397 nucleotides, depending on the initiation site (Radcliffe et al., 1990). I , cloned from normal B cells initiates at the same site as the
IMMUNOGLOBULIN ISOTYPE SWITCHING
259
397-bp exon, but is only 126 bp long, suggesting that the different germline transcripts result from a difference in splicing between normal cells and the cell line (Radcliffe et al., 1990; Lebman et al., 1990b). Transforming growth factor P enhanced the LPS-induced increase in mIgA+ 1.29 cells and caused a substantial increase in the steady level of germline a mRNA transcripts prior to the appearance of mIgA+ cells (Shockett and Stavnezer, 1991). Furthermore, the increase in steady state level of germline a transcripts was associated with a comparable increase in transcription rate. Analysis of the ability of the region 5’ to I , to control expression of the luciferase reporter gene transfected into 1.29 cells demonstrated that at least two distinct elements are involved in constitutive and TGF-P-induced transcription (Lin and Stavnezer, 1992). An ATF/CRE site in the region from -1 to -106 appears to be involved in basal level expression. Transforming growth factor P inducibility is due in part to a separate element, that is, a tandemly repeated sequence in the region between - 127 and - 105. Thus, that data from studies on IgA expression in 1.29 are consistent with a model in which TGF-P mediates isotype switching to IgA by inducing transcription of germline a mRNA. Transforming growth factor P clearly stimulates isotype switching to IgA in both normal cells and cell lines, but it is not clear if TGF-/3 is a primary or a secondary signal in this process. Studies with the CHl2.LX cell line suggest that TGF-P is a secondary signal that causes isotype switching in a population that is already committed to high-frequency switching to IgA (Whitmore et al., 1991). Although TGF-/3 in combination with IL-2 causes a 30- to 50-fold increase in IgA secretion in LPS-stimulated B cell cultures, TGF-P causes only 3% of the cells to express mIgA (Coffman et al., 1989; Lebman et al., 199Oa). Significantly, the secreted IgA in these cultures is derived from cells that are first induced to express mIgA (Lebman et al., 1990a). Transforming growth factor /3 also induced only small increases in mIgA expression in B cells activated with either anti-IgD conjugated to dextran or Th2 helper cells (Ehrhardt et al., 1992). These studies imply that TGF-P is either only a partial switch signal (Ehrhardt et al., 1992) or it acts on a population that has already received a signal that drives it toward IgA. Studies demonstrating that TGF-j3 acts posttranscriptionally to increase germline a mRNA accumulation by LPS-stimulated splenic B cells (Lebman et al., 1993) are consistent with the latter hypothesis. Although it is clear that TGF-/3 enhances isotype switching in uitro, it has not yet been determined whether TGF-/3 is an important inducer of isotype switching in uiuo.
260
ROBERT L. COFFMAN ET AL.
VI. lrotype Switching in Humans
Virtually all of the pioneering studies on the regulation of isotype switching were done in the murine system. There is now considerable data available from studies in the human system, and it has become clear that there are numerous analogies between the two species. Not only are the effects of homologous cytokines similar, but the similarity continues to be evident in situations in which data on molecular mechanisms in both species exist. Consequently, it is likely that the similarities observed between mouse and human in the mechanism of IgE regulation will extend to other isotypes and permit generalizations that apply to other mammals. Studies with human B lineage tumors have suggested that the germline transcription of some or all C, and C, loci may be less specifically regulated than in mouse (Sideras et al., 1992; Kerr and Burrows, 1991), but further study will be needed to determine if this is true for normal B cells. A. INTERLEUKIN-4 REGULATION OF ISOTYPE SWITCHING Several cellular studies indicated that the roles of IL-4 and IFN-y in the regulation of human IgE synthesis were analogous to their roles in the murine system (Pene et al., 1988a,b; Thyphronitis et al., 1989; Vercelli et al., 1989; Jabara et al., 1988).The major difference has been that most of the initial murine studies were done with cultures containing only mitogen-stimulated purified B cells and cytokines, whereas most human systems required more complex culture conditions. For example, in early studies antibody secretion was measured in cultures of enriched human B cells cocultured with activated alloreactive T cell clones or their supernatants (Pene et al., 1988b). In these cultures, the relative levels of IgE secretion correlated with the relative levels of IL-4 and IFN-y produced by the different T cell clones such that clones producing high levels of IFN-.)Icould not induce IgE synthesis unless anti-IFN-y was present (Pene et al., 1988b). Subsequently it was demonstrated that IL-4 induced IgE secretion by enriched human B cells and that IFN-.)Iand IFN-a blocked IgE induction (Pene et al., 1988a). However, IgE secretion required the presence of accessory cells because preparations of purified. B cells could not be induced to secrete IgE, suggesting that IL-4 did not act directly on B cells to induce IgE secretion. The work of several laboratories ultimately demonstrated that IL-4 could act directly on B cells, but the B cells required an additional activation signal (Thyphronitis et al., 1989; Gascan et al., 1991a,b; Jabara et al., 1990). This signal could be provided
IMMUNOGLOBULIN ISOTYPE SWITCHING
261
by infection with Epstein-Barr virus (EBV)(Thyphronitis et al., 1989), monoclonal antibodies directed against CD40 (Jabara et al., 1990; Rousset et al., 1991), and activated CD4+ T cells (Gascan et al., 1991a,b). Molecular studies on the role of IL-4, IFN-y, and the activating signal on IgE expression suggest a model for the generation of B cells expressing secondary isotypes. Interleukin 4 alone was shown to be sufficient for the expression of germline E transcripts and IFN-y decreased the steady state level of germline E mRNA (Gauchat et al., 1990).The human germline E transcripts, which are similar in structure to those described in the murine system, consist of an exon ( I , ) located 5’ to the switch region spliced to the first exon of C, (Gauchat et al., 1990).Treatment of purified mIgE- B cells with IL-4 alone caused the induction of germline, but not productive, E mRNA (Jabara et al., 1991; Shapira et al., 1992). When purified mIgE- B cells were stimulated with IL-4 together with either EBV or anti-CD40 both germline and productive E mRNA transcripts were induced although neither EBV nor anti-CD40 alone were capable of inducing any C,-containing transcripts (Jabara et al., 1991; Shapira et aZ., 1992). The combination of EBV and IL-4 resulted in deletional switch recombination that involved a direct joining of S, to S, (Shapira et al., 1991). A similar deletional switch recombination occurred in the presence of IL-4 and anti-CD40 but not IL-4 alone (Shapira et al., 1992). Taken together these findings suggest that two separate signals are necessary and sufficient for isotype switching to IgE. The first signal is B cell activation, which may also induce a switch recombinase able to cause switching to any active CH gene. The second signal is provided by IL-4, which induces germline transcription of the C, gene, making it a substrate for the recombinase. The effect of IL-4 on isotype switching in humans is not limited to IgE expression. The addition of IL-4 to B cells cultured with irradiated thymoma cells and phorbolI2-myristate 13-acetate increased the number of cells with cytoplasmic IgG4 (Lundgren et al., 1989). Subsequent clonal analysis of the effect of IL-4 on the pattern of isotype production by B cells activated with CD4+ T cell clones demonstrated that IL-4 stimulated isotype switching to IgG4 (Gascan et al., 1991b). A second cytokine, IL-13, has been found to induce isotype switching to IgE and IgG4 in humans (Punnonen et al., 1993). Interleukin-13 is the human homolog of the murine Th2-SpeCifiC cytokine p600 (Brown et al., 1989). Interleukin-13 acts much like IL-4 and it has been suggested that the two cytokines may share some
262
ROBERT L. COFFMAN ET AL.
components of their receptors, because they cross-compete for the same high-affinity receptors on cells (G. Zurawski, personal communication, 1992).
B. TRANSFORMING GROWTHFACTOR /3 AND IgA EXPRESSION Cellular and molecular studies indicate that TGF-/3 stimulates isotype switching to IgA in humans. Two different approaches were used to address this issue on a cellular level (Defrance et al., 1992; Van Vlasselaer et al., 1992). Utilizing the relative level of mIgD expression to separate tonsillar cells into populations that are enriched for either naive or germinal center B cells, DeFrance and colleagues observed that TGF-/3in combination with IL-10 induced IgA secretion by naive B cells activated with anti-CD40 (Defrance et al., 1992). Limiting dilution analyses demonstrated that the increased IgA secretion was associated with an increase in the frequency of IgA-producing cells. However, TGF-P inhibited IL-10-induced secretion of IgA in the germinal center B cell-enriched population, suggesting that, similar to the murine system, TGF-P could exert dichotomous effects on the development of IgA-secreting cells in the human system such that it stimulates IgA expression by mIgA- cells while inhibiting IgA secretion by mIgA+ cells (Defrance et al., 1992; Coffman et al., 1989). Additional cellular evidence that TGF-P mediates isotype switching is derived from studies demonstrating that TGF-P stimulated IgA production b y mIgA-, but not mIgA+, human splenic B cells stimulated with pokeweed mitogen and activated cloned CD4+ T cells. On a molecular level, TGF-/3 induces a1 and a 2 germline transcripts, which are structurally similar to the murine germline a transcripts (Islam et al., 1991; Nilsson et al., 1991).Thus there are both cellular and molecular data to support a similar role for TGF-P in IgA regulation in both human and murine systems. ACKNOWLEDGMENTS The authors wish to thank the following colleagues for sharing their thoughts and unpublished results: F. Alt, B. Birshtein, M. Boothby, J. deVries, W. Dunnick, D. Goodman, S. Jung, A. Radbruch, C. Snapper, and G . Zurawski. R.L.C. also wishes to thank J. Yorke for help in organizing the bibliography. The DNAX Research Institute is supported by the Schering-Plough Corporation. D. A. Lebman is an RJR Nabisco Research Scholar in Immunology and is supported by the Jeffress Memorial Trust and USPHS Grant DK 42892. P. Rothman is supported by a Pfizer Scholars Award, a Pew Scholar Award, and by a National Institutes of Health Grant R 0 1 AI33450-01.
IMMUNOGLOBULIN ISOTYPE SWITCHING
263
REFERENCES Abney, E. R., Cooper, M. D., Kearney, J. F., Lawton, A. R., and Parkhouse, R. M. (1978). J . lmmunol. 120,2041-2049. Adams, B., Dorfler, P., Aguzzi, A., Kozmik, Z., Urbanek, P., Maurer-Fogy, I., and Busslinger, M. (1992). Genes Dev. 6, 1589-1607. Agabian, N. (1990). Cell (Cambridge,Mass.) 61, 1157-1160. Alt, F. W., Bothwell, A. L., Knapp, M., Siden, E., Mather, E., Koshland, M., and Baltimore, D. (1980). Cell (Cambridge,Mass.) 20,293-301. Alt, F. W., Ferrier, P., Malynn, B., Lutzker, S., Rothman, P., Berman, J., Blackwell, K., Mellis, S., Pollock, R., Furley, A., Rathbun, G., Yancopoulos, G., Logtenberg, T., Morrow, M., Cook, W., Heinke, B., and Suh, H. (1988).Ann. N.Y.Acud. Sci. 546,9-24. Alt, F. W., Rathbun, G., Oltz, E., Taccioli, G., and Shinkai, Y. (1992).Ann.N.Y.Acad. Sci. 651,277-294. Anderson, J . , Coutinho, A., and Melchers, F. (1978). Eur. J . Immunol. 8,336-343. Armitage, R., Fanslow, W., Strockbine, L., Sato, T., Clifford, K., Macduff, B., Anderson, D., Gimpel, S., Davissmith, T., Maliszewski, C., Clark, E., Smith, C., Grabstein, K., Cosnian, D., and Spriggs, M. (1992). Nature (London)357,80-82. Barberis, A., Widenhorn, K., Vitelli, L., and Busslinger, M. (1990). Genes Dev. 4,849859. Battey, J., Max, E. E., McBride, W. O., Swan, D., and Leder, P. (1982). Proc. Natl. Acad. Sci. U.S.A. 79,5956-5960. Bauer, D. C., Mathies, M. J., and Stavitsky, A. B. (1963).J. Exp. Med. 117,889. Bergstedt-Lindqvist, S . , Sideras, P., MacDonald, H. R., and Severinson, E. (1984). Immunol. Rev. 78,25-50. Bergstedt-Lindqvist, S., Moon, H. B., Persson, U., Moiler, G., Heusser, C., and Severinson, E. (1988). E u r . J . lmmunol. 18,1073-1077. Berton, M. T., and Vitetta, E. S. (1990).J . Exp. Med. 172,375-378. Berton, M. T., and Vitetta, E. S. (1992).Int. Immunol. 4,387-396. Berton, M. T., Uhr, J. W., and Vitetta, E. S. (1989). Proc. Natl. Acad. Sci. U.S.A. 86, 2829-2833. Blackwell, T. K., and Alt, F. W. (1989,). J . Biol. Chem. 264, 10327-10330. Blackwell, T. K., and Alt, F. W. (1989b).Annu. Rev. Genet. 23,605-636. Boothby, M., Gravallese, E., Liou, H.-C., and Glimcher, L. H. (1988). Science 242, 1559-1562. Borzillo, G. V., Cooper, M. D., Kubagawa, H., Landay, A., and Burrows, P. D. (1987). J . lmmunol. 139,1326-1335. Brian, A. A. (1988). P r G C . Natl. Acad. Sci. U.S.A. 85,564-568. Brown, K. D., Zurawski, S. M., Mosmann, T. R., and Zurawski, G. (1989).J . lmmunol. 142,679-687. Bruzik, J. P., and Maniatis, T. (1992). Nature (London)360,692-695. Burger, C., and Radbruch, A. (1992). Mol. lmmunol. 29, 1105-1112. Chen, Y. W., Word, C., Dev, V., Uhr, J. W., Vitetta, E. S., and Tucker, P. W. (1986).J. E x p . Med. 164,562-579. Chu, C. C., Paul, W. E., and Max, E. E. (1992a).Abstr. 8th Int. Congr. Immunol., p. 34. Chu, C. C., Paul, W. E., and Max, E. E. (1992b). Proc. Natl. Acad. Sci. U.S.A. 89, 6978-6982. Coffman, R. L., and Carty, J. (1986).J. Immunol. 136,949-954. Coffman, R. L., and Cohn, M. (1977).J. Immunol. 118,1806-1815. Coffman, R. L., Ohara, J., Bond, M. W., Carty, J.. Zlotnik, A., and Paul, W. E. (1986). J . Immunol. 136,4538-4541.
264
ROBERT L. COFFMAN ET AL.
Coffman, R. L., Lee, F., Yokota, T., Arai, K., and Mosmann, T. R. (1987). “UCLA Symposium: Immune Regulation by Characterized Polypeptides” (G. Goldstein, J. Bach, and H. Wigzell, eds.), pp. 523-532. Alan R. Liss, New York. Coffman, R. L., Seymour, B. W., Lebman, D. A., Hiraki, D. D., Christiansen, J. A., Shrader, B., Cherwinski, H. M., Savelkoul, H. F., Finkelman, F. D., Bond, M. W., and Mosmann, T. R. (1988). Immunol. Rev. 102,5-28. Coffman, R. L., Lebman, D. A., and Shrader, B. (1989).J.E x p . Med. 170,1039-1044. Coleclough, C., Cooper, D., and Perry, R. P. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 1422-1426. Cooper, M. D., Lawton, A. R., and Kincade, P. W. (1972).Contemp. Top. Immunobiol. 1, 33-47. Cooper, M. D., Kearney, J. F., Lydyard, P. M., Grossi, C. E., and Lawton, A. R. (1977). Cold Spring Harbor Symp. Quant. Biol. 41 (I?. l), 139-145. Cory, S. (1986).Ado. Cancer Res. 47,189-234. Craig, N. L. (1985). Cell (Cambridge,Mass.) 41,649-650. Davis, M. M., Kim, S. K., and Hood, L. E. (1980). Science 209,1360-1365. De Pinho, R., Kruger, K., Andrews, N., Lutzker. S., Baltimore, D., and Alt, F. W. (1984). Mol. Cell. Biol. 4,2905-2910. Defrance, T., Vanbervliet, B., Briere, F., Durand, I., Rousset, F., and Banchereau, J. (1992).J . E x p . Med. 175,671-682. Dierks, P., van Ooyen, A., Cochran, M. D., Dobkin, C., Reiser, J., and Weissmann, C. (1983).Cell (Cambridge, Mass.) 32,695-706. Dunnick, W., and Stavnezer, J. (1990). Mol. Cell. Biol. 10,397-400. Dunnick, W., Wilson, M., and Stavnezer, J. (1959). Mol. Cell. Biol. 9, 1850-1856. Dunnick, W. A,, Hertz, G. Z., Scappino, L., and Gritzmacher, C. (1993). Nucleic Acids Res. 21,365-372. Durdik, J., Gerstein, R.M., Rath, S., Robbins, F’. F., Nisonoff, A., and Selsing, E. (1989). Proc. Natl. Acad. Sci. U.S.A. 86,2346-2350. Ehrhardt, R. O., Strober, W., and Harriman, G. R. (1992).J. Immunol. 148,3830-3836. Esser, C., and Radbruch, A. (1989). EMBOJ. 8t2), 483-488. Ferrier, P., Krippl, B., Blackwell, T. K., Furley, A. J., Suh, H., Winoto, A., Cook, W. D., Hood, L., Costantini, F., and Alt, F. W. (1990). EMBOJ. 9,117-125. Finkelman, F. D., Katona, I. M., Urban, J. F., Jr., Snapper, C. M., Ohara, J., and Paul, W. E. (1986). Proc. Natl. Acad. Sci. U.S.A.83,9675-9678. Finkelman, F. D., Katona, I. M., Mosmann, T. R.,and Coffman, R.L. (1988a).J.Immunol. 140,1022-1027. Finkelman, F. D., Katona, I. M., Urban, J. F., Jr., Holmes, J., Ohara, J., Tung, A. S., Sample, J. vG., and Paul, W. E. (1988b).J.Immunol. 141,2335-2341. Finkelman, F. D., Holmes, J., Katona, I. M., IJrban, J. F., Jr., Beckmann, M. P., Park, L. S., Schooley, K. A., Coffman, R. L., Mosmann, T. R.,and Paul, W. E. (1990).Annu. Rev. Immunol. 8,303-333. Flanagan, J. G., and Rabbitts, T. H. (1982). Nature (London)300,709-713. Gaff, C., and Gerondakis, S. (1990). Znt. Immunol. 2,1143-1148. Gascan, H., Gauchat, J. F., Aversa, G., Van Vlasselaer, P., and d e Vries, J. E. (1991a). J . Immunol. 147,8-13. Gascan, H., Gauchat, J. F., Roncarolo, M. G., Yssel, H., Spits, H., and de Vries, J. E. (1991b).J. E x p . Med. 173,747-750. Gauchat, J. F., Lebman, D. A., Coffman, R. L., Gascan, H., and d e Vries, J. E. (1990). J. Exp. Med. 172,463-473.
IMMUNOGLOBULIN ISOTYPE SWITCHING
265
Gearhart, P. J., Sigal, N. H., and Klinman, N. R. (1975). Proc. Natl. Acad. Sci. U.S.A.72, 1707-1711. Gerondakis, S. (1991).Proc. Natl. Acad. Sci. U.S.A. 87, 1581-1585. Goodman, D. J., Gaff, C., and Gerondakis, S. (1993). Znt. Zmmunol. 5, 199-208. Gritzmacher, C. A. (1989). Crit. Reo. Zmmunol. 9,173-200. Grumet, F. C. (1972).J.E x p . Med. 135,110-125. Han, H., Okamoto, M., Honjo, T., and Shimizu, A. (1991).Znt. Zmmunol. 3, 1197-1206. Hodgkin, P. D., Yarnashita, L. C., Coffman, R. L., and Kehry, M. R. (199O).J.Immunol. 145,2025-2034. Hodgkin, P. D., Yamashita, L. C., Seymour, B., Coffman, R. L., and Kehry, M. R. (1991). J . Zmmunol. 147,3696-3702. Hofker, M. H., Walter, M. A., and Cox, D. W. (1989).Proc. Natl. Acad. Sci. U.S.A. 86, 5567-557 1. Illges, H., and Radbruch, A. (1992). Mol. Zmmunol. 29, 1265-1272. Isakson, P. C., Pure, E., Vitetta, E. S., and Krdmmer, P. H. (1982).J . E x p . Med. 155, 734-748. Ishida, N., Ueda, S., Hayashida, H., Miyata, T., and Honjo, T. (1982). EMBO J . 1, 1117-1123. Ishizaka, K. (1976).Adv. Zmmunol. 23,l-75. Islam, K. B., Nilsson, L., Sideras, P., Hammarstrom, L., and Smith, C. I. (1991). Znt. Zmmunol. 3,1099-1106. Iwasato, T., Shimizu, A., Honjo, T., and Yamagishi, H. (1990). Cell (Cambridge, Mass.) 62, 143-149. Iwasato, T., Arakawa, H., Shimizu, A., Honjo, T., and Yamagishi, H. (1992).J.E x p . Med. 175,1539-1546. Jabara, H. H., Ackerman, S. J., Vercelli, D., Yokota, T., Arai, K.-I., Abrams, J,, Dvorak, A. M., Lavigne, M. C., Banchereau, J., De Vries, J., Leung, D. Y. M., and Geha, R. S. (1988).I. Clin. Zmmunol. 8,437-446. Jabara, H. H., Fu, S. M., Geha, R. S., and Vercelli, D. (l99O).J.E r p . Med. 172,1861-1864. Jabara, H. H., Ahern, D. J., Vercelli, D., and Geha, R. S. (1991). J . Zmmunol. 147, 1557-1560. Jack, H. M., McDowell, M., Steinberg, C. M., and Wabl, M. (1988).Proc. Natl. Acad. Sci. U.S.A.85, 1581-1585. Johnson, K. R., Lehn, D. A., Elton, T. S., Barr, P. J., and Reeves, R. (1988).J.B i d . Chem. 263,18338-18342. Jung, S., Rajewsky, K., and Radbruch, A. (1993). Science 259,984-988. Kataoka, T., Miyata, T., and Honjo, T. (1981). Cell 23,357-368. Kataoka, T., Takeda, S., and Honjo, T. (1983). Proc. Natl. Acad. Sci. U.S.A. 80, 26662670. Katona, I. M., Urban, J. F., Jr., and Finkelman, F. D. (1985).Proc. Natl. Acad. Sci. U.S.A. 82,511-515. Kawakami, T., Takahashi, N., and Honjo, T. (1980). Nucleic Acids Res. 8,3933-3945. Kearney, J. F., Cooper, M. D., and Lawton, A. R. (1976).J.Zmmunol. 116,1664-1668. Keil, R. L., and Roeder, G. S. (1984). Cell 39,377-386. Kenip, D. J., Morahan, G., Cowman, A. F., and Harris, A. W. (1983).Nature (London)301, 84-86. Kenter, A. L., and Birshtein, B. K. (1981). Nature (London) 293,402-404. Kenter, A. L., and Watson, J. V. (1987).J.Zmmunol. Meth. 97, 111-117. Kerr, W. G., and Burrows, P. D. (1991). Int. Immunol. 3,1059-1065.
266
ROBERT L. COFFMAN ET AL.
Kim, P. H., and Kagnoff, M. F. (1990a).J.Zmmunol. 144,3411-3416. Kim, P. H., and Kagnoff, M. F. (1990b).J.Zmmunol. 145,3773-3778. Kipps, T. J., and Herzenberg, L. A. (1986). EMBOJ. 5,263-268. Kirsch, I. R., Morton, C. C., Nakahara, K., and I,eder, P. (1982). Science 216,301-303. Knapp, M. R., Liu, C. P., Newell, N., Ward, R. B., Tucker, P. W., Strober, S., and Blattner, F. (1982). Proc. Natl. Acad. Sci. U.S.A. 79,2996-3000. Kuhn, R., Rajewsky, K., and Muller, W. (1991). Science 254,707-710. Kuze, K., Shimizu, A., and Honjo, T. (1991). Znt. Zmmunol. 3,647-655. Laio, F., Birshtein, B., and Rothman, P. (1993). Submitted. Lawton, A. R., and Cooper, M. D. (1993). Contemp. Top. lmmunobiol. 3, 193-225. Layton, J. E., Vitetta, E. S., Uhr, J. W., and Gammer, P. H. (1984).J . Exp. Med. 160, 1850- 1863. Lebman, D. A., and Coffman, R. L. (1988).J.E x p . Med. 168,853-862. Lebman, D. A,, Lee, F. D., and Coffman, R. L. (1990a).J . Zmmunol. 144,952-959. Lebman, D. A., Nomura, D. Y.,Coffman, R. L., and Lee, F. D. (1990b).Proc. Natl. Acad. Sci. U.S.A. 87,3962-3966. Lebman, D. A,, Park, M., Hansen-Bundy, S., and Pandya, A. (1993).In preparation. Lee, F., Yokota, T., Otsuka, T., Meyerson, P., Villaret, D., Coffman, R., Mosmann,T. R., Rennick, D., Roehm, N., Smith, C., Zlotnik, A., and Arai, K. (1986). Proc. Natl. Acad. Sci. U.S.A. 83,2061-2065. Lennon, G. G., and Perry, R. P. (1985). Nature (London) 318,475-478. Leung, H., and Maizels, N. (1992). Proc. Natl. .%cad.Sci. U.S.A.89,4154-4158. Liao, F., Giannini, S. L., and Birshtein, B. K. (1.992).J . Zmmunol. 148,2909-2917. Lieber, M. R. (1991). FASEB J . 5,2934-2944. Lin, Y. C., and Stavnezer, J. (1992).J.Zmmunol. 149,2914-2925. Liu, F. T., Albrandt, K., Sutcliffe, J. G., and Katz, D. H. (1982). Proc. Natl. Acad. Sci. U.S.A.79,7852-7856. Lundgren, M., Persson, U., Larsson, P., Magnusson, C., Smith, C. I., Hammarstrom, L., and Severinson, E. (1989). Eur.J. lmmunol. 19,1311-1315. Lutzker, S., and Alt, F. W. (1988a). Mol. Cell. Biol. 8,1849-1852. Lutzker, S. G., and Alt, F. W. (198813).“Mobile I)NA”(D. E. Bergand M. M. Howe, eds.), 1988, pp. 691-714. American Society for Microbiology, Washington, D.C. Lutzker, S., Rothman, P., Pollock, R., Coffman, R. L., and Alt, F. W. (1988). Cell (Cambridge, Mass.) 53,177-184. Maki, R., Roeder, W., Traunecker, A., Sidman, C., Wabl, M., Raschke, W., and Tonegawa, S. (1981). Cell (Cambridge, Mass.) 24,353-365. Mandler, R., Finkelman, F. D., Levine, A. D., and Snapper, C. M. (1993).J.Zmmunol. 150,407-418. Manning, D. D. (1974).J . Zmmunol. 113,455-163. Marcu, K. B., Lang, R. B., Stanton, L. W., and Harris, L. J. (1982).Nature (London)298, 87-89. Matsuoka, M., Yoshida, K., Maeda, T., Usuda, S., and Sakano, H. (1990). Cell (Cambridge, Mass.)62, 135-142. Mills, F. C., Brooker, J. S., and Camerini-Otrro, R. D. (1990). Nucleic Acids Res. 18, 7305-7316. Mills, F. C., Thyphronitis, G., Finkelman, F. D., and Max, E. E. (1992).J.Zmmunol. 149, 1075- 1085. Mitchell, G. F., Grumet, F. C., and McDevitt, H. 0. (1972).J.Exp. Med. 135, 126-135. Mongini, P. K., Stein, K. E., and Paul, W. E. (1981).J.Exp. Med. 153, 1-12. Mowatt, M. R., and Dunnick, W. A. (1986).J.Zmmunol. 136,2674-2683.
IMMUNOGLOBULIN ISOTYPE SWITCHING
267
Murgita, R. A., Mattioli, C. A., and Tomasi, T. B., Jr. (1973).]. E x p . Med. 138,209-228. Neale, G. A., and Kitchingman, G. R. (1991).Nucleic Acids Res. 19,2427-2433. Nikaido, T., Nakai, S., and Honjo, T. (1981). Nature (London)292,845-848. Nikaido, T., Yamawaki-Kataoka,Y., and Honjo, T. (1982).]. Biol. Chem. 257,7322-7329. Nilsson, L., Islam, K. B., Olafsson, O., Zalcberg, I., Samakovlis, C., Hammarstrom, L., Smith, C. I., and Sideras, P. (1991). l n t . Immunol. 3, 1107-1115. Noelle, R. J., Daum, J., Bartlett, W. C., McCann, J., and Shepherd, D. M. (1991). ]. Immunol. 146,1118-1124. Noelle, R. J., Shepherd, D. M., and Fell, H. P. (1992).]. lmmunol. 149, 1164-1169. Nolan-Willard, M., Berton, M. T., and Tucker, P. (1992).Proc. Natl. Acad. Sci. U.S.A.89, 1234-1238. Noma, Y., Sideras, P., Naito, T., Bergstedt-Lindquist, S., Azuma, C., Severinson, E., Tanabe, T., Kinashi, T., Matsnda, F., Yaoita, Y., and Honjo, T. (1986).Nature (London) 319,640-646. Nossal, G. J. V., Szenberg, A., Ada, G. L., and Austin, C. M. (1964).J.E x p . Med. 119,485. Obata, M., Kataoka, T., Nakai, S., Yamagishi, H., Takahashi, N., Yamawaki-Kataoka, Y., Nikaido, T., Shimizu, A., and Honjo, T. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 2437-2441. Ott, D. E., and Marcu, K. B. (1989). lnt. Immunol. 1,582-591. Ott, D. E., Alt, F. W., and Marcu, K. B. (1987). EMBOJ. 6,577-584. Oudin, J., and Michel, M. (1969).]. E x p . Med. 130,619-642. Parker, D. C. (1993).Annu. Reo. lmmunol. 11 (in press). Pene, J., Rousset, F., Briere, F., Chretien, I., Bonnefoy, J. Y., Spits, H., Yokota, T., Arai, N., Arai, K., Banchereau, J., and de Vries, J. (1988a). Proc. Natl. Acad. Sci. U.S.A. 85, 6880-6884. Pene, J., Rousset, F., Briere, F., Chretien, I., Paliard, X.,Banchereau, J., Spits, H., and d e Vries, J. E. (1988b).]. Immunol. 141,1218-1224. Pene, J., Rousset, F., Briere, F., Chretien, I., Wideman, J., Bonnefoy, J. Y., and DeVries, J. E. (1988~). Eur.]. lmmunol. 18,929-935. Perlmutter, A. P., and Gilbert, W. (1984). Proc. Natl. Acad. Sci. U.S.A.81,7189-7193. Pernis, B., Forni, L., and Luzzati, A. L. (1977). Cold Spring Harbor Symp. Quant. Biol. 41 (Pt. l), 175-183. Pierce, C. W., Solliday, S. M., and Asofsky, R. (1972).]. E x p . Med. 135,675-697. Pierce, S. K., and Klinman, N. R. (1975).]. E x p . Med. 142, 1165-1179. Pierce, S. K., Cancro, M.P., and Klinman, N. R. (1978).]. E x p . Med. 148,759-765. Press, J. L., and Klinman, N. R. (1973).]. E x p . Med. 138,300-305. Punnonen, J., Aversa, G., Cocks, B. G., McKenzie, A. N. J., Menon, S., Zurawski, G., d e Waal Malefyt, R., and de Vries, J. E. (1993). Proc. Natl. Acad. Sci. U.S.A. 90.(in press). Radbruch, A., and Sablitzky, F. (1983). EMBO]. 2,1929-1935. Radbruch, A., Liesegang, B., and Rajewsky, K. (1980). Proc. Natl. Acrid. Sci. U.S.A. 77, 2909-29 13. Radbruch, A,, Burger, C., Klein, S., and Muller, W. (1986a).Immunol. Rev. 8 9 , 6 9 4 3 . Radbruch, A., Muller, W., and Rajewsky, K. (1986b). Proc. Natl. Acad. Sci. U.S.A.83, 3954-3957. Radcliffe, G., Lin, Y. C., Julius, M., Marcu, K. B., and Stavnezer, J. (1990).Mol. Cell. Biol. 10,382-386. Reahan, M. E., and Griffin, J. A. (1990). Nature (London)348,342-344. Rogers, J,, Early, P., Carter, C., Calame, K., Bond, M., Hood, L., and Wall, R. (1980).Cell 20,303-312. Rosenberg, Y. J. (1982).lmmunol. Reo. 67,33-58.
268
ROBERT L. COFFMAN ET AL.
Rothman, P., Lutzker, S., Cook, W., Coffman, R., and Alt, F. W. (1988).J. Exp. Med. 168, 2385-2389. Rothman, P.,Chen, Y.-Y., Lutzker, S., Li, S. C., Stewart, V., Coffman, R., and Alt, F. W. (1990a).Mol. Cell. Biol. 10, 1672-1669. Rothman, P., Lutzker, S., Gorham, B., Stewart, V., Coffman, R.,and Alt, F. W. (1990b). Int. Immunol. 2,621-627. Rothman, P., Li, S. C., Gorham, B., Glimcher, L., Alt, F., and Boothby, M. (1991).Mol. Cell. Biol. 11,5551-5561. Rousset, F., Garcia, E., and Banchereau, J. (1951).J. E r p . Med. 173,705-710. Sakano, H., Rogers, J. H., Huppi, K., Brack, C.. Traunecker, A., Maki, R., Wall, R., and Tonegawa, S. (1979).Nature (London)277,627-633. Sakano, H., Maki, R., Kurosawa, Y., Roeder, W., and Tonegawa, S. (1980).Noture (London) 286,676-683. Immunol. 141, Savelkoul, H. F., Lebman, D. A., Benner, R.,and Coffman, R. L. (1988).J. 749-755. Schmitz, J., and Radbruch, A. (1989).Int. Immirnol. 1,570-575. Schrader, J. W. (1975).J. Immunol. 114,1665-1669. Schultz, C. L.,and Coffman, R. L. “IL-4: Structure and Function,” (H. Spits, ed.), 1992, pp. 76-88.CRC Press, Boca Raton, Florida. Schultz, C. L., and Coffman, R. L. (1993).Submitted. Schultz, C.,Petrini, J., Collins, J., Claflin, J. L., Denis, K. A., Gearhart, P., Gritzmacher, Immunol. 144,363-370. C., Manser, T.,Shulman, M., and Dunnick, W. (199O).J. Schultz, C. L., Elenich, L. A., and Dunnick, W. A. (1991).Int. Immunol. 3, 109-116. Schultz, C. L.,Rothman, P., Kiihn, R., Kehry, M., Miiller, W., Rajewsky, K., Alt, F., and Coffman, R. L. (1992).J. Immunol. 149,6044. Severinson, E., Bergstedt-Lindqvist, S., van der Loo, W., and Fernandez, C. (1982). Immunol. Rev. 67,7345. Severinson, E., Fernandez, C., and Stavnezer, J. (1990). Eur.J.Immunol. 20,1079-1084. Shapira, S. K., Jabara, H. H., Thienes, C. P., Ahern, D. J., Vercelli, D., Could, H. J., and Geha, R. S. (1991).Proc. Natl. Acad. Sci. U.S.A. 88,7528-7532. Shapira, S. K., Vercelli, D., Jabara, H. H., Fu, S. M., and Geha, R. S. (1992).].E x p . Med. 175,289-292. Shimizu, A., Takahashi, N., Yaoita, Y., and Honjo, T. (1982).Cell (Cambridge,Mass.)28, 499-506. Shimizu, A., Nussenzweig, M. C., Mizuta, T-H., Leder, P., and Honjo, T. (1989).Proc. Natl. Acad. Sci. U.S.A.86,8020-8023. Shimizu, A., Nussenzweig, M. C., Han, H., Sanchez, M., and Honjo, T. (1991).J. E x p . Med. 173,1385-1393. Shockett, P., and Stavnezer, J. (1991). J . Immunol. 147,4374-4383. Sideras, P., Bergstedt-Lindqvist, S., and Severinson, E. (1985).Eur. J. Immunol. 15, 593-598. Sideras, P., Mizuta, T. R., Kanamori, H., Suzuki, N., Okamoto, M., Kuze, K., Ohno, H., Doi, S., Fukuhara, S., Hassan, M. S., Hammarstrom, L., Smith, E., and Honjo, T. (1989).Int. Immunol. 1,631-642. Sideras, P., Nilsson, L., Islam, K. B., Quintana, I. Z., Freihof, L., Rosen, A., Juliusson, G., Hammarstrom, L., and Smith, C. I. (1992).J. Immunol. 149,244-252. Siebenkotten, G., Esser, C., Wabl, M., and Radbruch, A. (1992). Eur. J . Immunol. 22, 1827-1834. Singh, M., and Birshtein, B. K. (1993).Mol. C d l . Biol. 13 (in press).
IMMUNOGLOBULIN ISOTYPE SWITCHING
269
Smith, G. R., Kunes, S. M., Schultz, D. W., Taylor, A., and Triman, K. L. (1981).Cell 24, 429-436. Snapper, C. M., and Paul, W. E. (1987). Science 236,944-947. Snapper, C . M., Finkelman, F. D., and Paul, W. E. (1988a).J.E x p . Med. 167,183-196. Snapper, C . M., Finkelman, F. D., Stefany, D., Conrad, D. H., and Paul, W. E. (1988b). J . Immunol. 141,489-498. Snapper, C. M., Peschel, C., and Paul, W. E. (1988c).J.Immunol. 140,2121-2127. Snapper, C. M., Pecanha, L. M., Levine, A. D., and Mond, J. J. (1991).J.Zmmunol. 147, 1163-1170. Snapper, C . M., Mclntyre, T. M., Mandler, R., Pecanha, L. M., Finkelman, F. D., Lees, A., and Mond, J. J. (1992).J.E x p . Med. 175, 1367-1371. Sonoda, E., Matsumoto, R., Hitoshi, Y., Ishii, T., Sugimoto, M., Araki, S., Torninaga, A., Yamaguchi, N., and Takatsu, K. (1989).J.E x p . Med. 170,1415-1420. Spriggs, M. K., Armitage, R. J., Strockbine, L., Clifford, K. N., Macduff, B. M., Sato, T. A., Maliszewski, C. R., and Fanslow, W. C. (1992).J.E x p . Med. 176, 1543-1550. Stanton, L. W., and Marcu, K. B. (1982).Nucleic Acids Res. 10,5993-6006. Stavnezer, J. (1991). Nature (London)351,447-448. Stavnezer, J., Radcliffe, G., Lin, Y-C., Nietupski, J., Berggren, L., Sitia, R., and Severinson, E. (1988a).Proc. Natl. Acad. Sci. U.S.A.85,7704-7708. Stavnezer, J., Radcliffe, G., Lin, Y. C., Nietupski, J., Berggren, L., Sitia, R., and Severinson, E. (1988b). Proc. Natl. Acad. Sci. U.S.A.85,7704-7708. Stavnezer-Nordgren, J., and Sirlin, S. (1986). EMBOJ. 5,95-102. Stevens, T. L., Bossie, A., Sanders, V. M., Fernandez-Botran, R., Coffman, R. L., Mosmann, T. R., and Vitetta, E. S. (1988). Nature (London)334,255-258. Strathern, J. N., Klar, A. J., Hicks, J. B., Abraham, J. A., Ivy, J. M., Nasmyth, K. A,, and McGill, C. (1982). Cell (Cambridge,Mass.) 31, 183-192. Szurek, P., Petrini, J., and Dunnick, W. (1985).J.Immunol. 135,620-626. Takahashi, N., Ueda, S., Obata, M., Nikaido, T., Nakai, S., and Honjo, T. (1982). Cell 29, 671-679. Thomas, B. J., and Rothstein, R. (1989). Cell (Cambridge, Mass.) 56,619-630. Thomas, B. J., and Rothstein, R. (1991). Cell (Cambridge, Mass.) 64,l-3. Thyphronitis, G., Tsokos, G. C., June, C. H., Levine, A. D., and Finkelman, F. D. (1989). Proc. Natl. Acad. Sci. U.S.A.86,5580-5584. Tilley, S. A., and Birshtein, B. K. (1985).J.E x p . Med. 162,675-694. Tucker, P. W., Marcu, K. B., Slightom, J. L., and Blattner, F. R. (1979). Science 206, 1299- 1303. Tucker, P. W., Liu, C. P., Mushinski, J. F., and Blattner, F. R. (1980). Science 209, 1353-1360. Tucker, P. W., Slightom, J. L., and Blattner, F. R. (1981).Proc. NatLAcad. Sci. U.S.A.78, 7684-7688. Uhr, J. W., and Finkelstein, M. S. (1963).J.E x p . Med. 117,457. van der Loo, W., Gronowicz, E. S., Strober, S., and Herzenberg, L. A. (1979).J.Zmmunol. 122,1203-1208. Van Vlasselaer, P., Punnonen, J., and d e Vries, J. E. (1992).J.Immunol. 148,2062-2067. Vercelli, D., Jabara, H. H., Arai, K., and Geha, R. S. (1989)J. E x p . Med. 169,1295-1307. Vitetta, E. S., Ohara, J., Myers, C. D., Layton, J. E., Krammer, P., and Paul, W. E. (1985). J . E x p . Med. 162, 1726-1731. von Schwedler, U., Jack, H.-M., and Wabl, M. (1990a).Nature (London)345,452-456. von Schwedler, U., Jack, H. M., and Wabl, M. (1990b).New Biol. 2,657-662.
270
ROBERT L. COFFMAN ET AL.
Wabl, M., Meyer, J., Beck-Engeser, G.. Tenkhoff, M., and Burrows, P. D. (1985).Nature (London)313,687-689. Wang, A. C., Wilson, K. S., Hopper, J. E., Fudenberg, H. H., and Nisonoff, A. (1970). Proc. Natl. Acad. Sci. U.S.A.66,337-343. Waters, S. H., Saikh, K. U., and Stavnezer, J. (1989).Mol. Cell. Biol. 9,5594-5601. Webb, C. F., Cooper, M. D., Burrows, P. D., and Griffin, J. A. (1985).Proc. Natl. Acad. Sci. U.S.A. 82,5495-5499. Whitmore, A. C., Prowse, D. M., Haughton, G., and Arnold, L. W. (1991).Int. lmmunol. 3,95-103. Williams, M., and Maizels, N. (1991).Genes Dao. 5,2353-2361. Winter, E., Krawinkel, U., and Radbruch, A. (1987).EMBO]. 6,1663-1671. Wu, T. T., Reid-Miller, M., Perry, H. M., and Kabat, E. A. (1984).E M B O ] . 3,2033-2040. Wuerffel, R. A., Nathan, A. T., and Kenter, A. L. (1990).Mol. Cell. Biol. 10, 1714-1718. Wuerffel, R., Jamieson, C. E., Morgan, L., Merkulov, G. V., Sen, R., and Kenter, A. L. (1992).J . E r p . Med. 176,339-349. Xu, L., Kim, M. G., and Marcu, K. B. (1992).I n t . Zmmunol. 4,875-887. Xu, L., Gorham, B., Li, S . C., Bottaro, A., Alt, F. W., and Rothman, P. (1993).Proc. Natl. Acad. Sci. U.S.A. 90,3705-3709. Xu, M., and Stavnezer, J. (1990).Deo. lmmunol. 1,ll-17. Xu, M. Z., and Stavnezer, J. (1992).E M B O J . 11,145-155. Yancopoulos, G. D., DePinho, R. A., Zimmennan, K. A., Lutzker, S. G., and Rosenberg, N. (1986).EMBO]. 5,3259-3266. Yoshida, K., Matsuoka, M., Usuda, S., Mori, A., Ishizaka, K., and Sakano, H. (1990).Proc. Natl. Acad. Sci. U.S.A. 87,7829-7833. Yuan, D., Weiss, E. A., Layton, J. E., Krarnmer, P. H., and Vitetta, E. S. (1985). 1.lmmunol. 135, 1465-1469. This article was accepted for publication on 3 March 1993.
ADVANCES IN IMMUNOLOGY, V OL. 54
CD44 and Its Interaction with Extracellular Matrix JAYNE LESLEY,, ROBERT H Y M N , * A N D PAUL W. K l N C A D e 'Department of Cancer Biology, The Salk Institute, Son Diego, California 92186, and the tlmmunobiology and Cancer Program, OklahomaMedical Research Foundation, Oklahoma City, Oklohoma 73104
1. Introduction
CD44 is a broadly distributed family of cell surface glycoproteins that has been studied independently by many investigators in a variety of systems and under a variety of names. It is now generally believed to be a cell adhesion molecule with proposed functions in extracellular matrix (ECM)binding, cell migration, lymphopoiesis, and lymphocyte homing. Two reviews have sought to clarify the diverse historicaI nomenclature and present the evidence that has brought this assortment of molecules and their proposed functions together under the designation CD44 (Gallatin et al., 1991; Haynes et al., 1989). Most recently, study of the structure of the (probably) single gene that encodes CD44 has begun to reveal how the great variety of molecular forms of CD44 may be generated (see Section 11). Our interest in CD44 was prompted by its involvement in lymphocyte development (see Section IV), during which it participates both in the earliest stages of T and B cell differentiation and in later stages of T and B cell activation in response to immunological stimuIi. In these and other contexts, CD44 seems to function by mediating cell-cell or cell-substrate interactions through recognition of elements of the extracellular matrix, intercellular milieu, and/or pericellular layer (Section 111). Perhaps all of the suggested functions of CD44 on motile cells, including metastatic cells (Section VI), can be explained as consequences of this recognition function. However, downstream events subsequent to ligand recognition by CD44 may vary greatly depending on cell type and on other stimuli in the cellular environment, perhaps accounting for the diversity of cellular responses observed to result after binding CD44 by ligand or antibody (Section IV). Of particular interest to us is the finding that ligand recognition by CD44 is not constitutive in many CD44-expressing cells. Rather, like a number of other lymphocyte adhesion receptors (Dustin and Springer, 1991), CD44 receptor function is strictly regulated (Section V). Regulated, transient receptor activation may provide specificity for what would otherwise be an uncontrolled interaction between a broadly 271 Copyright 0 1993 hy Academic Press, Inc. All rights of reproduction in any form reserved.
272
JAYNE LESLEY ET AL.
distributed cell surface molecule and common components of the extracellular environment. The association of elevated or altered CD44 expression with malignancy (Section VI) may be an indication of the importance of regulation of CD44 function to normal cell behavior. II. Molecular lsoforms and Posttranslational Modifications
A. DISTRIBUTION AND HETEROGENEITY OF CD44 CD44 has been characterized in particular detail on lymphocytes, macrophages, fibroblasts, epithelial cells and keratinocytes (Table I). However, it has also been detected on many other cell types, and several tissues previously considered to lack CD44 were subsequently shown to be positive under at least some circumstances. For example, a number of papers cite endothelial cells and erythrocytes as lacking this marker (Trowbridge et al.,1982; Kansas et al.,1989; Alho and Underhill, 1989; Picker et al., 1989a). However, other reports have demonstrated expression on these cell types (Spring et al.,1988; Lucas et al., TABLE I SOMEWELL-CHARACTERIZED FOHMS OF CD44 Tissue"
Apparent M , ( X
Protein isoformsb ~~
Lymphoid" M yeloidd Erythroide Fibroblastsf Epitheliala Nervous systemh Endothelial'
85-95 and 180-200 85-95 82-92 80-85and180 177-250 74-86
116
A, B, D, E, G A, G A A A, E, F, G, H, 1, J, K, L, M, N A, C ?
Selected references (below) given for lymphoid and myeloid cells include leukemias; those for fibroblasts include fibrosarcomas; epithelial cells include carcinomas, keratinocytes, and melanoma; and the nervous system includes neuroblastomas, glioblastomas, and astrocytes. Also noteworthy is a thorough biochemic:d characterization of CD44 in placenta (St. Jacques et al., 1993). " Protein isoforms correspond to Fig. 2. 'Jalkanen et al. (less), Jalkanen and Jalkanen (1992), Picker et al. (1989), Omary et al. (1988). Trowbridge et al. (1982). Kalomiris and Bourguignon (1988), Flanagan et al. (1989), Gallatin et al. (1989). Hughes et al. (1983), Camp et al. (1991). Lucas et al. (1989), Spring et al. (1988). 'Hughes and August (198l), Camp et al. (19!Jl), Kansas et al. (1989), Carter and Wayner (1988), Brown et al. (1991),Tarone et al. (1984), Culty et al. (1990). Brown et al. (1991). Kansas et al. (1989), Haggerty et al. (1992). Asher and Bignami (1992). Vogel et al. (1998).Girgrah et 01. (1991b). Bourguignon et al. (1992).
'
CD44 AND ITS INTERACTION WITH ECM
273
1989; Bourguignon et al., 1992; Heider et al., 1993). These discrepan-
cies may result from different experimental techniques, iise of cultured
or transformed versus normal tissues, heterogeneity among endothelial cells, or genetic polymorphisms that dramatically influence levels of CD44 expression (see below). The few tissues that have consistently been described as CD44 negative include liver hepatocytes, kidney tubular epithelium, cardiac muscle, portions of the skin, and testis (Picker et al., 1989a; Flanagan et al., 1989; Wang et al., 1992; Heider et al., 1993).Expression in the nervous system of young normal individuals is restricted to the white matter, including astrocytes and glial cells, whereas its appearance in gray matter accompanies age and disease (Flanagan et al., 1989; Girgrah et al., 1991a,b; Vogel et al., 1992; Asher and Bignami, 1992). CD44 has been characterized in a number of species, which in addition to human and mouse, also includes hamster, baboon, rat, sheep, dog, and cow (Mackay et al., 1988; Idzerda et al., 1989; Aruffo et al., 1990; Sandmaier et aZ., 1990; Gunthert et al., 1991; Gallatin et al., 1991; Bosworth et al., 1991). The most abundant form of CD44 on lymphocytes has an apparent molecular mass on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of 85-95 kDa (Table 1). This is commonly termed the “standard” or “hematopoietic” form. The presence of intramolecular disulfide bonds accounts for a slightly larger apparent molecular mass when run under reducing conditions. A similar, but not necessarily identical, type of CD44 has been characterized on macrophages, fibroblasts, fibrosarcomas, and astrocytes. The protein is acidic, with estimates of isoelectric point ranging from 4.2 to 5.8 (Jalkanen et al., 1988; Kalomiris and Bourguignon, 1988; Picker et al., 1989,; Culty et al., 1990).Most of this charge is accounted for by the presence of sialic acid. CD44 can incorporate radiolabeled sulfate, which may be attached directly to the protein core (Jalkanen et al., 1988; Brown et al., 1991). The hematopoietic form is synthesized as a polypeptide of apparent molecular mass 42 kDa that undergoes subsequent N- and 0-linked glycosylation (Brown et al., 1991; Lokeshwar and Bourguignon, 1991).Also, depending on the tissue, CD44 is usually phosphorylated (Isacke et al., 1986; Kalomiris and Bourguignon, 1988; Carter and Wayner, 1988; Camp et al., 1991; Neame and Isacke, 1992). It was apparent from biochemical studies that multiple forms of CD44 must exist (Omary et al., 1988; Kansas et al., 1989). A 180- to 200-kDa form of CD44 on lymphocytes was found to be sensitive to chondroitinase ABC, thus demonstrating that the molecule can exist as a proteoglycan (Jalkanen et al., 1988).This modification is not trivial
274
JAYNE LESLEY
ET AL.
because it confers new ligand-binding ability on the molecule and unique functions to the cells that express it (Jalkanen and Jalkanen, 1992; Fassen et al., 1992)(see Section 111,D).Other size differences in CD44 have in some cases been accounted for by glycosylation (Quackenbush et al., 1990; Camp et al., 1991). Indeed, half of the apparent molecular mass of the molecule is contributed by N-linked and 0linked carbohydrate and, in addition to chondroitin sulfate, some species of CD44 bear heparan sulfate (Brown et al., 1991). Some CD44 molecules with a larger apparent molecular mass first noted in immunochemical studies (Omary et al., 1988; Kansas et al., 1989) were later shown to result from alternative exon utilization (see below). It is not clear whether glycosylation differences and/or alternative exon splicing confer tissue specific properties on CD44.
B. SOLUBLECD44 Receptors of many kinds are frequently detectable in the circulation in soluble form and their levels may reflect disease activity (Fernandez-Botran, 1991). Substantial quantities (5 pg/ml) of CD44 may be present in human serum and plasma (Dalchau et al., 1980; Telen et al., 1983; Lucas et al., 1989; Picker et al., 1989b; Baiil and HoiejSi, 1992). An inverse relationship was found between soluble CD44 levels in synovial fluid and numbers of immigrating blood cells in arthritic joints (Haynes et al., 1991). The cellular origin of soluble CD44 is not known and, depending on the individual, either one or two prominent species of CD44 were demonstrable (Baiil and HofejSi, 1992). Neutrophil granulocytes, but not lymphocytes, readily shed CD44 when stimulated in culture (Campanero et al., 1991; Baiil and HoiejSi, 1992). The size of the recovered material was somewhat smaller than that associated with the cells and there is evidence that this material is generated by an endogenous proteolytic mechanism. Crosslinkage of CD44 on lymphocytes and neutrophils, but not fibroblasts, resulted in loss of some surface antigen (Baiil and HoiejSi, 1992; Jacobson et al., 1984). Soluble CD44 has also been recovered from the culture medium of keratinocytes and carcinoma cells (Brown et al., 1991; Haggerty et al., 1992). It is not known if soluble CD44 has biological importance. However, at least one other cellular adhesion molecule (F3) is functionally competent when presented in soluble form (Durbec et al., 1992). Cell surface CD44 is sensitive to a number of proteolytic enzymes. Trypsin releases a 65-kDa fragment from murine lymphocytes or macrophages (Trowbridge et al., 1982; Hughes et al., 1983), CD44 is cleaved in two steps by a bacterial glycoproteinase (Sutherland et al.,
CD44 AND ITS INTERACTION WITH ECM
275
1992), and bromelain releases all but 22 kDa of CD44 from T cells (Hale and Haynes, 1992). It has been noted that alternative splicing mechanisms can introduce arginine dipeptides into human CD44 isofornis (Dougherty et al., 1991). These sites in other receptors and proteoglycans correspond to sites involved in proteolytic release from the membrane. In mice, there is a similar site near the lipid bilayer in all known splice variants and an additional one in high molecular weight variants (He et al., 1992). C. PROTEIN ISOFORMS The molecular cloning of CD44 from hematopoietic cells was immediately informative with respect to structure and possible functions. The mature core protein is predicted to be 37-38 kDa and can be subdivided into several domains (Zhou et al., 1989; and see Fig. 1). There is one potential membrane-spanning region that shows 80-90% sequence homology among different species. A stretch of some 90 relatively hydrophobic residues comprises the amino terminus of the extracellular domain and also shows 80-90% sequence similarity among species. This domain has homology to cartilage link protein and other proteins known to interact with hyaluronan (HA) (Stamenkovic et al., 1989; Goldstein et al., 1989; Zhou et al., 1989; Nottenburg et al., 1989; Wolffe et al., 1990; Bosworth et al., 1991; Gunthert et al., 1991) (see Section III,B,l). However, CD44 lacks the conserved basic residues found in cartilage link and proteoglycan core proteins and thought to be involved in HA recognition (Goldstein et al., 1989; Wolffe et al., 1990). This region contains six cysteines that might be utilized to form a single globular domain (Goldstein et al., 1989).In the human, there are six potential sites for N-linked carbohydrate addition in this domain. The membrane-proximal domain is less well conserved, showing only approximately 50% sequence similarity among species. This domain contains three or four consensus sites for chrondroitin sulfate attachment and many potential sites of 0-glycosylation. The cytoplasmic domain is highly conserved, showing 80-90% sequence similarity among species. Of the six serine residues in human CD44 cytoplasmic domain, five are conserved among the human, mouse, baboon, cow, and hamster (Stamenkovic et al., 1989; Zhou et al., 1989; Idzerda et al., 1989; Bosworth et al., 1991; Aruffo et al., 1990). Four of these residues are conserved in the rat; however, serine residue 317 is replaced by threonine (Gunthert et al., 1991). Of the conserved serine residues, Ser-296 (the first amino acid of the mature human protein is residue 1; for other species, residue numbers have been changed to correspond to the human sequence) i s not phosphorylated in intact epithelial cells (Neame and Isacke, 1992), although this
276
JAYNE LESLEY E T AL.
FIG.1. Schematic representation of the human CD44 protein, “hematopoietic” or “standard” form. The solid filled area indicates the conserved region, amino acids 12 to 101, which shows sequence similarity to HA-binding domains of cartilage proteoglycan core protein and link protein. [Amino acid numbers correspond to the sequence given in Stamenkovic et al. (1989), with the first amino acid of the mature protein designated as residue 1.)The heavily stippled area indicates the membrane-proximal region, which shows more variability among species than other parts of the molecule, and which has been deleted in a mutant construct expressed in transfected cells (ANC in Table IV; He et al., 1992). The lightly stippled area indicate:; the transmembrane domain. 0-, Potential N-linked glycosylation sites; 0--,possible 0-linked glycosylation sites; -, potential sites for addition of chondroitin sulfate; -, site of insertion of alternately spliced exons (between amino acids 202/203); @-, four serines in the cytoplasmic domain that are conseved among all species and that are potential sites for phosphorylation; @-, serines 303 and 305,which have been shown to be required for phosphorylation in vivo (Neame and Isacke, 1992; Camp et al., 199.3a); S-S, probable, and S-+, possible, disulfide bonds between cysteine residues.
+
CD44 AND ITS INTERACTION WITH ECM
277
residue is a potential substrate for CAMP- and cGMP-dependent protein kinases (Wolffe et al., 1990). Both Ser-303 and -305 may be phosphorylated. Mutation of either abolishes the ability of the hematopoietic form of CD44 to be phosphoryIated in epithelial cells (Neame and Isacke, 1992). In other studies in which mutant constructs were transiently expressed in COS cells, mutation of residue 305 abolished phosphorylation, whereas mutation of residue 303 reduced it (Camp e t al., 1993a). As discussed by Neame and Isacke (1992), neither of these two residues appears to be a substrate for protein kinase C of CAMPdependent protein kinases. Ser-271, which is located close to the transmembrane region, may be a substrate for protein kinase C (Wolffe et al., 1990), but it is not clear whether this residue is phosphorylated in intact cells. Mutation of this residue did not affect phosphorylation of CD44, nor did deletion of serine 317, when mutant constructs were expressed in COS cells (Camp et al., 1993a). Almost all studies have concluded that CD44 represents a single gene, located on the short arm of chromosome 11 in humans (Goodfellow et al., 1982; Franke et al., 1983) and on chromosome 2 in mice (Lesley and Trowbridge, 1982; Colombatti et al., 1982). There are at least 19 exons (Screaton et al., 1992),as indicated in Fig. 2, and unpublished studies with rodents suggest there is an additional exon between exons 5 and 6 (Screaton et d., 1992; U. Gunthert, personal communication, 1992). Variation in the cytoplasmic tails of CD44 results from differential utilization of one of two exons (exon 18 or 19) (Screaton et al., 1992). The first three amino acids common to both tails are coded by exon 17. Exon 18 encodes an A+T-rich untranslated region, which might confer instability on the mRNA. The additional amino acids present in the long cytoplasmic tail are encoded in exon 19. It is not certain whether mature CD44 molecules with the short tail are expressed; however, CD44 mRNA for a short-tailed form has been detected by the polymerase chain reaction (Goldstein and Butcher, 1990). In the most simple pattern of expression, at least three major CD44 mRNA species are observed in Northern blots. These are approximately 1.6,2.2, and 4.8 kb in humans (Stamenkovic et al., 1989; Goldstein et al., 1989; Quackenbush et al., 1990; Brown et al., 1991; Shtivelman and Bishop, 1991)and 1.6,3.3, and 4.6 kb in mice (Wolffe et al., 1990; Haegel and Ceredig, 1991).This size variation apparently results from the use of multiple polyadenylation signals (Shtivelman and Bishop, 1991; Ham et al., 1991). At least 18 CD44 transcripts have been described to date (Fig. 2). This heterogeneity results from the facts that 12 of the 19 exons can undergo alternative splicing and that consensus splice donor/acceptor
278
JAYNE LESLEY ET AL.
MON I
2
3
4
5
1 6
7
8
9 10 I1
12 13 14 15 I6 I7 I8 19
lM
1.....I .....a....9.....n.................................................................... n..p.....Jl.......... 1 0...1..., ,, .n.., In......................................................... ...........u.,.P.J ...!.El U.1.J ................ 0 I..................................._......... ............ B..? .....El..........fl C ...I...RJ ...0......................*............*I .......................... ... a....a...B..........1 0....@..,.I ...Q..0.. ...................................................... .....a,.a,p.,. 4..........0 E U...,.~....~.,..Q....n........................................................ N.*Jt.R...J ....I.....*...J P
[email protected].,..4....n................................................ n..OI.OP*O.~...J .,. 1.......... c n..I.,J..a..1................. ..........4.J..Jl., RO.U..R..,.R........J u....i.,.l.%.n.,.n........... ...l..,lJ..ul.I.O.n..J ................ ......4........... ...1 I ' n...i.,.i...n ..n........... ~,..E.~~4..~..,~."~...U...R.,.~..........U 0..R..J....n..n.......).......R.~~.~.~,r..~n..~.. ~RUn...R..4..........0 K n....l.... 1.. .. .......-........1.J....R.u~..~n..~..~~."~.O.U...R ....1..........D .I 9,.1.*,.kQ .. ....... .......1............... 4..P.R.4.J ...a.......... 1 ...IJ...QJ. ......................a..p.LB..J.. .............. U..B....._.... 0 0..l..J ..a..a.......".......B.J..,a.,B..,s ..........._*........ " ........... ....._....1 O 0..1.J ..a..0.................. I..)................8,.n*II, ...... I....-....D P 1..4.J..n.. .n.......).......1............................. m..*R....-... ..... ...I..... a..9...n................................................. I..! ....."..J...? ....1.....-....0 A
*.
nm......... .*,2
Q
*-*
FIG.2. Multiple transcription products of the CD44 gene. The genomic structure of human CD44 as described by Screaton et ol. (1992) is shown at the top with leader peptide (LP) and transmembrane domains (TM)indicated. Exons (2, 3) encode the putative HA-binding domain. An additional exon may exist between exons 5 and 6 [Screaton et al. (1992) and U. Gunthert, personal communication]. Also illustrated are splice variants in published cDNA sequences that may encode unique protein isoforms. The corresponding reference citations are as follows: A (Stamenkovic et al., 1989; Wolffe et ol., 1990; Zhou et al., 1989; Aruffo et al., 1990; Gallatin et ol., 1991; Idzerda et al., 1989; Bosworth e t al., 1991; Nottenburg et al., 1989; He et al., 1992, Harn e t ol., 1991);B (Goldstein and Butcher, 1990; Goldstein et al., 1989; Screaton e t al., 1992); C (Shtivelman and Bishop, 1991); D (Arch et al., 1992); E (Jackson et al., 1992; Cooper et al., 1992);
CD44 AND ITS INTERACTION WITH ECM
279
sites occur within two exons (exons 5 and 7). Splicing within exons sometimes accounts for deletion of part of exon 5 in neuroblastoma cells (Fig. 2C) (Shtivelman and Bishop, 1991) and part of exon 7 in keratinocytes (Fig. 2L) (Hofmann et al., 1991). Furthermore, one example has been found in which a “constant” exon (exon 15)is deleted (Fig. 2N) (Gunthert et al., 1991). This is an impressive amount of diversity. However, in the case of another cell adhesion molecule, NCAM, which can alternatively splice 12 of its 27 exons, as many as 192 proteins are possible (Barthels et al., 1992; Screaton et al., 1992).A satisfactory nomenclature for CD44 transcripts and corresponding core proteins has not yet been introduced. Although the “hematopoietic” or “standard” isoform (Fig. 2A) is the most prominant on blood cells, it is also expressed by fibroblasts and melanoma cells (Stamenkovic et al., 1989; Idzerda et al., 1989; Zhou et al., 1989; Nottenburg et al., 1989; Wolffe et al., 1990;Aruffo et al., 1990; Gallatin et al., 1991; Bosworth et al., 1991; He et al., 1992). Similarly, transcripts corresponding to the “epithelial” form of CD44 (Fig. 2G) have been described in leukemias (Dougherty et al., 1991; Gunthertet al., 1991; Stamenkovic et al., 1991; Jackson et al., 1992). Although it is not yet certain that every possible splice variant is translated into a mature protein, expression of a number of splice variants has been demonstrated. For example, long CD44 transcripts have been found in keratinocytes and carcinoma cells (Hofmann et al., 1991; Kugelman et al., 1992; Haggerty et al., 1992)and deglycosylated core proteins of a large size have been characterized in such cells (Brown et al., 1991; Haggerty et al., 1992). When compared with the “standard” CD44 transcript that predominates in hematopoietic cells (Fig. 2A), this long core protein contains an additional 338 amino acids in the extracellular domain. Besides doubling the length, three potential N-glycosylation sites, numerous 0-glycosylation sites, and two chondroitin sulfate attachment sequences are added to the molecule. Consensus attachment sequences for heparan sulfate side chains have not been described, but it is clear that they are present in this form of CD44 (Brown et al., 1991; Haggerty et al., 1992). As noted above, the F (He et al., 1992; Jackson et al., 1992);G (Dougherty e t al., 1991; Gunthert et al., 1991; Stamenkovic et al., 1991; Hofmann et al., 1991; Jackson e t al.,1992; Brown et al., 1991; He et ul., 1992); H (Jackson et al., 1992); I (He et ul., 1992); J (He et al., 1992); K (Kugelman et id.,1992; Haggerty et al., 1992; Hofmann et al., 1991);L (Hofmann et al., 1991); M (Jackson et ul., 1992; Hofmann et al., 1991);N (Gunthert et al., 1991); 0, P, Q, and R (Hofmann et al., 1991).Additional splice variants have been reported by Herrlich et a1. (1993),but their sequences are not yet published.
280
JAYNE LESLEY ET AL.
inserted sequences also contain potential protease cleavage sites. Individual cells can simultaneously synthesize multiple isoforms of CD44 and our experiments with murine carcinoma cells revealed that the apparent molecular weights of the proteins corresponded to mRNA length (He et al., 1992). It is uncertain whether aberrent splicing occurs in malignant cells, or whether the diversity of transcripts frequently observed in them is a consequence of elevated levels of CD44 transcripts that are normally only minor components and/or only present in activated cells. Activation of normal lymphocytes correlated with the presence of a splice variant (Arch et al., 1992; Koopman et al., 1993). Also, a change in CD44 size, which may have been caused by posttranslational modification, occurred on stimulation of normal macrophages (Camp et ul., 1991). Transformed cells might therefore be representative of some normal differentiation stage or activation state. If splice variants are selectively expressed in tumor cells, this could provide a useful diagnostic tool, because 1tumor cell expressing a variant transcript may be detectable among lo6blood cells, using the polymerase chain reaction (Matsumura and Tarin, 1992). In addition, particular CD44 isoforms correlate with metastasis in certain cases (Gunthert et al., 1991) (see Section VI). Expression and alternative splicing of CD44 are clearly regulated in a way that is tissue and differentiation stage specific. In lymphocytes, these are also subject to change on activation (see Section V). Analysis of upstream regulatory sequences in the CD44 gene has been carried out in human neuroblastoma cells (Shtivelman and Bishop, 1991)and the human and mouse promoters have been analyzed by Herrlich and colleagues (1993). The neuroblastoma gene showed no TATA or CCAAT boxes, but contained a GC-rich region; however, the human and mouse genes cloned by Herrlich and colleagues (1993) show a TATA box at -31 to -35. In the neuroblastoma gene, RNA initiation sites were localized by primer extension experiments to 128 and 136 nucleotides upstream of the translation initiation codon, but the possibility of additional start sites in different tissues was not excluded. Functional promoter activity and the existence of multiple negative regulatory elements were demonstrated by transfection assays (Shtivelman and Bishop, 1991). Both the human and mouse genes show potential binding sites for the SP-1transcription factor, although it is not certain whether these sites function in intact cells (Shtivelman and Bishop, 1991; Herrlich et al., 1993).An AP-1 binding site at - 108 to -114 in the human sequence and -109 to -115 in the mouse sequence appears to function, because mutation of this site reduces
CD44 AND ITS INTERACTION WITH ECM
28 1
transcription of an indicator gene in transient transfection assays, reduces ability to respond to induction by phorbol ester, and reduces the increased transcription seen after cotransfection of j u n or ras constructs (Herrlich et al., 1993). Several observations suggest that one or more genes act to influence the level of CD44 expression on the cell surface. A polymorphism in mice correlates with CD44 mRNA content (Haegel and Ceredig, 1991). The In" blood group antigen of humans is carried on the CD44 molecule and its expression on erythrocytes and monocytes, but not lymphocytes, is influenced by the dominant inhibitor gene Zn(Lu) (Telen et al., 1983; Spring et al., 1988).An inverse relationship may also exist between expression of CD44 and the multidrug resistance gene (Cianfriglia et al., 1991).
D. CYTOPLASMIC DOMAIN Functional importance has been demonstrated for the cytoplasmic domains of many cell adhesion molecules and this is also the case for CD44 (Lesley et al., 1992; Thomas et al., 1992; see Sections II1,C and V,C,l). It has long been known that CD44 interacts at least indirectly with components of the cytoskeleton, including actin and ankyrin (Jacobson et al., 1984);Tarone et al., 1984; Lacy and Underhill, 1987; Carter and Wayner, 1988; Kalomiris and Bourguignon, 1988; Geppert and Lipsky, 1991; Bourguignon et al., 1992). Binding to erythrocyte ankyrin by purified CD44 has been demonstrated by in uitro assays (Kalomiris and Bourguignon, 1988); however, there is no sequence similarity between murine CD44 and the region of erythrocytic band 3 that is thought to be involved in ankyrin binding (Wolffe et al., 1990). The ability of CD44 to bind ankyrin in uitro is influenced by acylation (Bourguignon et al., 1991).Two conserved cysteines, located in the transmembrane domain, and at the interface of transmembrane and cytoplasmic domains, are potential sites for attachment of palmitic acid. Some similarity has been noted between sequences in the CD44 tail and those of members of the G-protein superfamily (Lokeshwar and Bourguignon, 1992). Purified CD44 bound GTP and displayed some GTPase activity in in uitro assays (Lokeshwar and Bourguignon,
1992).
Phosphorylation of the cytoplasmic tail of CD44 has also been thought to affect cytoskeletal association. In uitro phosphorylation of the molecule may enhance its affinity for ankyrin as measured in uitro assays (Lokeshwar and Bourguignon, 1992). As discussed in more detail below (Section V,C,l), however, it is not certain how these in uitro observations relate to observations in intact cells. In macro-
282
JAYNE LESLEY ET AL.
phages, only nonphosphorylated CD44 was present in the detergentinsoluble pool (Camp et al., 1991). Activation of T lymphocytes by phorbol ester and ionomycin reduced the apparent interaction of CD44 with the cytoskeleton (Geppert and Lipsky, 1991), as did phorbol ester treatment of macrophages (Camp et al., 1993a). However, it is not clear that phorbol ester treatment affects phosphorylation of CD44 in intact cells. No enhanced phosphorylation was seen in epithelial cells after phorbol ester treatment (Neame and Isacke, 1992), although transient increases in phosphorylation were seen in macrophages within minutes after phorbol ester treatment (Camp et al., 1993a). Two closely spaced and well-conserved serines in the cytoplasmic tail (residues 303 and 305 in human CD44) may be phosphorylated (Isacke et al., 1986;Carter and Wayner, 1988; Neame and Isacke, 1992; Camp et al., 1993a). Mutation of either serine residue prevented (Neame and Isacke, 1992)or greatly reduced (Camp et al., 1993a)phosphorylation. Although deletion of the cytoplasmic tail of CD44 prevented localization of the molecule to the basolateral membrane of transfected polarized epithelial cells, mutation of either serine residue 303 or 305 such that phosphorylation was prevented did not influence localization and did not affect the association of CD44 with the cytoskeleton (Neame and Isacke, 1992). Enhanced phosphorylation induced by phorbol ester was seen in COS cells transfected with a mutant CD44 construct in which serine residues 271, 303, and 305 were changed to glycine or alanine. This observation implies that the transient phorbol ester-induced phosphorylation was not occurring on these residues (Camp et al., 1993a).
111. CD44 and Extracellular Matrix
A. GENERAL FEATURES OF EXTRACELLULAR MATRIX The extracellular matrix (ECM) fills the spaces between cells. Although it has long been recognized that the components of the ECM perform an important structural role, it has been realized more recently that the ECM communicates with the cell interior and thus modulates cell adhesion, proliferation, and differentiation (Toole, 1991; Schubert, 1992). Major constituents of the ECM include collagenous proteins (Linsenmayer, 1991) and proteoglycans (Wight et al., 1991). The latter consist of one or more glycosaminoglycans-linear polymers of repeating disaccharides-covalently bound to a protein core.
CD44 AND ITS INTERACTION WITH ECM
283
Hyaluronan (HA) is a glycosaminoglycan consisting of a linear polymer of [u-glucuronic acid (1-p-3) N-acetyl-D-glucosamine (l-p-4)],* (Laurent, 1989; Wight et al., 1991; Laurent and Fraser, 1992), which differs from other glycosaminoglycans in not being covalently linked to a core protein. The HA polymer can have a molecular weight of up to several million. In solution, the polymer behaves as a random coil (Laurent, 1989). A large quantity of solvent is trapped within the coil and the molecule can be considered to be a highly hydrated sphere (Laurent and Fraser, 1992).There is evidence for secondary structure via hydrophobic bonding (Scott, 1992), and the possibility of aggregation of HA molecules leading to infinite meshworks has been raised (Scott, 1989, 1992). Within the ECM, large aggregates are formed by noncovalent interactions between proteoglycan molecules and HA (Morgelin et al., 1988). The basic unit is a ternary complex consisting of proteoglycan monomer, HA, and link protein, a molecule of approximately 40 kDa that interacts with both the proteoglycan monomer and HA to stabilize the proteoglycan-HA complex (Neame et al., 1986;Wight et al., 1991). Aggrecan, the aggregating chondroitin sulfate proteoglycan of cartilage, contains a protein core of 220 kDa to which numerous glycosaminoglycan side chains are covalently attached. Two globular domains, termed G1 and G2, are present at the amino terminus (Doege et al., 1987, 1991). The G1, but not the G2, domain interacts noncovalently with HA. The function of the G2 domain is uncertain and some proteoglycans, such as those found in aorta, lack a G2 domain (Morgelin et al., 1989). Proteoglycan and HA form complexes that appear in the electron microscope as “necklace-like” structures, with proteoglycan monomers randomly attached along the HA chain (Morgelin et ul., 1988). Addition of link protein converts this structure to a much more densely packed structure with a continuous coat of proteoglycan extending along the HA strand (Morgelin et al., 1988, 1989), reflecting stabilization of the proteoglycan-HA interaction by the link protein. The G1 domain of aggrecan is composed of subdomains containing disulfide-bonded loop structures of about 100 amino acids (Doege et al., 1987, 1991). These domains show sequence similarity to tandem repeat regions in the C-terminal region of link protein (Neame et al., 1987; Doege et al., 1987, 1991) that have been demonstrated to be involved in the binding of link protein to HA (Perin et al., 1987; Goetinck et al., 1987). Other HA-binding proteoglycans also show regions of sequence homology to the tandem repeat regions found in aggrecan and link protein: versican, isolated from fibroblasts (Le Baron et al., 1992), and TSG-6, a protein inducible by tumor necrosis factor a
284
JAYNE LESLEY ET AL
and interleukin 1 (IL-1) in fibroblasts and inducible by mitogens in peripheral blood mononuclear cells (Lee et al., 1992).
B. CD44 AS A RECEPTORFOR HYALURONAN 1. Sequence of CD44 Implies a Role in Hyaluronan Recognition A portion of the N-terminal domain of CD44 in humans and mouseresidues 12-101 in humans (the first amino acid of the mature protein is residue 1; see Fig. 1)-shows approximately 30% sequence similarity to the second (B) subdomain of cartilage proteoglycan core and link proteins (Goldstein et al., 1989; Stamenkovic et al., 1989; Idzerda et al., 1989; Wolffe et al., 1990; Doege et al., 1991). This sequence similarity increases to about 50% if conservative amino acid substitutions are considered (Goldstein et al., 1989). There is also a lower, but significant, sequence similarity to the B1 subdomain of cartilage proteoglycan core and link proteins (Goldstein et al., 1989). Th'is sequence similarity does not necessarily imply that this region of CD44 is actually functionally involved in the binding of CD44 to HA (Wolffe et al., 1990; Doege et al., 1991). Both link protein and the two HAbinding proteoglycan core proteins aggregan and versican contain at least two disulfide-bonded subdomains that form tandem-repeated loop structures and it may be that binding of more than one subdomain to HA is required to provide sufficient affinity for a stable ternary complex between link protein, HA, and proteoglycan core protein (Goetinck et al., 1987).Also, the interaction of link protein with HA is thought to be largely of an ionic nature (Goetinck et nl., 1987). If so, then it should be noted that the arginine and lysine residues of link protein thought to be important in this interaction are not conserved in CD44 (Goldstein et al., 1989; Wolffe et al., 1990). There is an arginine residue at position 70 of the human sequence and position 72 of the mouse sequence (Goldstein et al., 1989; Zhou et al., 1989; Wolffe et al., 1990), but it is not known whether this residue serves the same function as the basic residues found in the homologous region of the link protein. Nevertheless, although there is only limited sequence similarity between the HA-binding domains of proteoglycan core and link proteins and CD44, CD44 does function as a receptor for HA. 2. Evidence that CD44 Is a Receptor f o r Hyaluronan a. Inhibition of Hyaluronan Binding by CD44-Specific Antibodies It has been known for many years that HA can bind to the surface of nonchondrogenic cells in culture (Underhill and Toole, 1979). Many cell types are surrounded by a chondrocyte-like pericellular matrix
CD44 AND ITS INTERACTION WITH ECM
285
(Goldberg et al., 1984; Knudson and Knudson, 1991). Formation of this pericellular matrix occurs on addition of HA and cartilage proteoglycan monomer to cells, but matrix formation occurs only when the cells are able to bind HA through an HA-specific cell surface receptor (Knudson and Knudson, 1991). This receptor is present on a variety of cell lines. In t)it)o, it is expressed on epithelial cells, especially proliferating epithelial cells such as in the basal layers of stratified epithelium and on the basolateral surfaces of cells at the base of the crypts of Lieberkuhn of the intestine, although it is not found on cells lining the mouth of the crypts or on the villi (Alho and Underhill, 1989). Because this receptor has been best characterized on fibroblast cell lines, it will be termed the “fibroblast HA receptor.” It is distinct from other HA receptors found on liver endothelial cells (Yannariello-Brown et al., 1992a,b) and ras-transformed tumor cells (Hardwick et al., 1992). The fibroblast HA receptor is a glycoprotein of M,85,000 that binds to six sugar residues (three repeating disaccharide units) on the HA polymer (Underhill et al., 1983). It is phosphorylated and a proportion is associated with cytoskeletal actin filaments (Underhill et al., 1985, 1987; Lacy and Underhill, 1987; Culty et al., 1990). These properties resemble those of CD44 (Lesley et al., 1990a; Culty et al., 1990). Binding of [3H]HA to mouse cells was blocked by the CD44-specific monoclonal antibody (mAb) KM201 and this antibody depleted detergent cell lysates of material able to bind [3H]HA. These latter experiments indicate the identity of the fibroblast HA receptor and CD44 (Culty et al., 1990). A number of cell types, including fibroblasts and CD44-positive hematopoietic cells and cell lines, aggregate in the presence of exogenous HA (Green et al., 1988; Lesley et al., 1990a).This HA-dependent aggregation can be blocked if the cells are preincubated with antibody specific to the fibroblast HA receptor (Green et al., 1988) or with CD44-specific mAb (Lesley et al., 1990a). Binding of fluoresceinconjugated HA to lymphoid cell lines or adhesion of these cells to HA-coated plates is also specifically inhibited by CD44-specific antibodies (Lesley et al., 1990a; Miyake et al., 1990b). Adhesion of a B-lineage hybridoma to a bone marrow-derived stromal cell line can be blocked by CD44-specific antibody or by pretreatment of the stroma1 cell line with hyaluronidase (Miyake et al., 1990b), demonstrating that cell adhesion can be dependent on CD44 and HA. Only a subset of CD44-specific mAbs inhibits HA-dependent binding, indicating that only particular epitopes on the CD44 molecule contribute to (or border) the HA-specific binding site. However, not all CD44-positive cells bind HA (Lesley et al., 1990a; Miyake and
286
JAYNE LESLEY ET AL.
Kincade, 1990). This result indicates that although CD44 is necessary for HA binding, there is not a one-to-one correspondence between the expression of CD44 on the cell surface and the ability to bind HA. This point is extremely important and its implications will be discussed in detail below (Section V).
b. Transfection and Expression of CD44 cDNA Constructs
Several groups have used transfection and expression of CD44 constructs to demonstrate that HA is a ligand for CD44. Aruffo and colleagues (1990) expressed a soluble human CD44-immunoglobulin construct in COS cells. The gene product of this construct was a dimeric molecule consisting of the extracellular domain of CD44 and the hinge, cH2, and c H 3 domains of human IgG,. The soluble CD44 immunoglobulin fusion protein bound to lymph node high endothelial cells in primary culture. Binding was abolished by pretreatment of the target cells with hyaluronidase but not by enzymes that did not have activity against HA. Binding of the CD44-immunoglobulin fusion protein was inhibited in the presence of HA but not other glycosaminoglycans. In complementary experiments, St. Jacques and colleagues (1993) showed that radiolabeled CD44 purified from placenta by affinity chromatography could bind to immobilized HA. Binding was inhibited by preincubation of the CD44 with soluble HA and the bound material was sensitive to hyaluronidase. Much less binding was seen to collagen I, collagen VI, fibronectin, or heparin. These results strongly imply that HA is a ligand for CD44. Aruffo and colleagues (1990) isolated a cDNA clone for the hamster CD44 molecule and transfected this clone into COS cells. The transfected cells reacted with antibody specific for the hamster fibroblast HA receptor, which, as discussed above (Section 11I7B,2,a),is CD44. In other experiments, the same group transfected the CD44-negative human Burkitt B cell lymphoma line Namalwa with the hematopoietic form of human CD44 and examined binding of the transfectants to primary cultures of lymph node high endothelial cells (Stamenkovic et al., 1991). B cells transfected with the hematopoietic form bound to high endothelial cells and binding was blocked if the assay was done in the presence of HA or a polyclonal CD44-specific antibody or if the high endothelial cells were pretreated with hyaluronidase. Lesley and colleagues (1992) examined the HA-binding phenotype of the CD44.2-negative murine T cell lymphoma AKRl transfected with a murine CD44.1 cDNA. Transfection of CD44 conferred the ability to bind fluorescein-conjugated HA from solution and to bind to immobilized HA. Binding could be blocked by preincubation with
CD44 AND ITS INTERACTION WITH ECM
287
CD44-specific mAb and/or by competition with unconjugated HA. Transfection also conferred a CD44-dependent and hyaluronidasesensitive increase in adhesion to a lymph node endothelial cell line. These experiments provide conclusive evidence that CD44 functions as a receptor for HA. However, as noted above (Section III,B,2,a), not all cells that express the hematopoietic form of CD44 bind ligand. This observation implies that the cell is able to regulate the ability of the CD44 molecule to bind ligand. Thus it is reasonable to assume that some cells will not bind HA even if transfected with CD44 cDNA constructs whose expression leads to acquisition of HA-binding function when they are transfected into other cells. This, in fact, is the case (see Section V,A,4). The interesting questions are why CD44 is able to bind HA only in certain cellular environments and what factors are required to confer an HA-binding phenotype. These points will be discussed further in Section V. C. FEATURES OF CD44 MOLECULEMEDIATING INTERACTION WITH HYALURONAN The coding region of CD44 can be divided into four domains: an amino-terminal highly conserved domain, a less conserved membrane-proximal domain containing the alternatively spliced exons, the transmembrane domain, and the highly conserved cytoplasmic domain (Zhou et al., 1989; and see Fig. 1).To examine the role of these regions in mediating the ability of CD44 to bind HA, mouse CD44 constructs that code for a product lacking either the membraneproximal domain or the cytoplasmic domain have been transfected into the mouse T cell lymphoma AKR1. Transfectants that do not express the membrane-proximal region (between Val-161 and Arg-244) bind HA (He et al., 1992). This result indicates that the amino-terminal two-thirds of the CD44 molecule is sufficient for HA recognition. That this amino-terminal region is involved in binding of HA has been inferred from the fact that the region of sequence similarity to proteoglycan core and link proteins is located in the amino-terminal domain between residues 12 and 101 (the first residue of the mature human protein is residue 1) (Goldstein et al., 1989; see Section III,B,l). St. Jacques et. al. (1993) suggested that within this region, only residues 18-30 and 88-112 are likely to be exposed at the surface. An antibody to a peptide comprising residues 18-30 did not inhibit binding of CD44 to HA, suggesting that these residues do not contribute to the HA-binding site. AKRl transfectants expressing a mutant CD44 construct with a stop codon at Gly-276 and coding for only the first six amino acids of the
288
JAYNE LESLEY ET AL.
cytoplasmic domain do not bind HA, from solution, although AKRl transfectants expressing the wild-type molecule bind soluble HA well (Lesley et al., 1992).The amount of CD44 expressed on the surface of cells transfected with this mutant CD44 construct was roughly equal to the amount expressed on control transfectants expressing wild-type CD44 (Lesley et al., 1992). This is an important point, because the ability to bind HA from solution appears proportional to the amount of CD44 expressed on the cell surface (He et al., 1992). Thus, in experiments in which the HA-binding activity of two cell lines is to be compared, it may be necessary to use fluorescence-activated cell sorting or other methods to isolate cell lines expressing equal amounts of CD44 in order for valid comparisons to be made. Both sets of transfectants bound to HA immobilized on plates and showed hyaluronidasesensitive binding to a lymph node endothelial cell line. However, the HA-specific binding of transfectants expressing the mutant construct lacking most of the cytoplasmic domain was lower than that of transfectants expressing a wild-type CD44 construct. Similar results were observed by Thomas and colleagues, who transfected melanoma cells with human CD44 constructs expressing only the first 6 or 16 amino acids of the cytoplasmic domain (Thomas et al., 1992). Transfectants not expressing most of the cytoplasmic domain bound to HA-coated plates less well than transfectants expressing the wild-type CD44 molecule. Also, melanoma transfectants lacking most of the cytoplasmic domain of CD44 did not migrate on HA-coated surfaces, even though the cells did attach. Therefore interaction of the cytoplasmic domain of CD44 with the cytoskeleton may be important in mediating postattachment events, such as cell motility. Although the AKRl transfectants expressing a molecule lacking most of the cytoplasmic domain did not bind soluble HA, they could be induced to do so if the transfected cells were “activated” by pretreatment with the CD44-specific mAb IRAWB 14 (Lesley et al., 1992). This observation suggests that one function of the cytoplasmic domain is to allow preexisting CD44 molecules to enter a state whereby they are able to bind HA. Although these transfectants do not bind soluble HA unless “activated” by antibody, they do bind to immobilized HA, although less well than transfectants expressing wild-type CD44. These points will be discussed in more detail in Section V. Although the membrane proximal domain is not required for HA recognition, insertion of exons in this region may modulate the ability to bind HA. Stamenkovic and colleagues found that the Burkitt lymphoma cell line Namalwa transfected with the human CD44 epithelial isoform [equivalent to the M2 isoforni of the mouse (He et al., 1992);
CD44 AND ITS INTERACTION WITH ECM
289
Fig. 2G] did not show HA-dependent binding to primary cultures of lymph node high endothelial cells (Stamenkovic et al., 1991) or to immobilized HA in microwells (Sy et al., 1991), whereas the same Burkitt lymphoma cell line transfected with the hematopoietic form did show HA-dependent binding. Flow cytometric analysis indicated that both sets of transfectants expressed approximately equal levels of CD44 (Stamenkovic et al., 1991; Sy et al., 1991). Melanoma cells transfected with a human CD44 epithelial isoform construct also bound poorly to immobilized HA (Thomas et al., 1992).These observations have led to speculation that the epithelial isoform and possibly other isoforms of CD44 containing exons inserted in the membrane proximal region may be unable to bind HA. The inserted sequences contain potential sites of glycosylation that might affect ability to bind HA. If so, then the particular cells into which the constructs have been transfected might determine the extent to which CD44-dependent binding to HA is observed. Higher molecular weight CD44 species purified from placenta (which is likely to comprise a mixture of isoforms) did not bind to immobilized HA (St. Jacques et al., 1993). In contrast, transfectants of the mouse T cell lymphoma AKRl expressing several higher molecular weight mouse CD44 isoforms (F, G, I, and J in Fig. 2), including the mouse equivalent of the human epithelial form, did bind HA (He et al., 1992). Binding of soluble HA was proportional to the level of CD44 expression, and for all isoforms, the level of HA binding was markedly increased if the transfectants were first “activated” by pretreatment with IRAWB 14 mAb. All of the isoforms mediated CD44-dependent adhesion to immobilized HA (He et al., 1992). This result indicates that CD44 isoforms containing inserts in the membrane-proximal region are not intrinsically unable to bind HA, although it is not excluded that the presence of the insert may have quantitative effects on the ability of CD44 to bind HA when expressed in particular cellular environments (e.g., by affecting the affinity of the binding site or the overall avidity of a particular cell for HA). It is interesting to note that Namalwa cells transfected with the epithelial form of CD44 did bind to BHK cells (Stamenkovic et al., 1991). Although no evidence was presented that this binding was HA dependent, BHK cells do express high levels of cell surface HA (Underhill and Toole, 1982). There is another possible explanation for these seemingly differing results. The last two amino acids derived from exon 5 (Fig. 2) are alanine and threonine. These are near the boundary of nearly all splice variants of CD44 and have been independently found in both genomic and cDNA sequences in many laboratories for human, rat, hamster,
290
JAYNE LESLEY ET AL.
and mouse CD44 (see, e.g., Aruffo et al., 1990; He et al., 1992; Screaton et al., 1992). A single exception is in the “epithelial” CD44 cDNA described by Stamenkovic et al. (1991), in which the nonpolar alanine is replaced by a basic arginine residue. It seems possible that this nonconservative change influences the HA recognition capability of this particular clone of the human epithelial form of CD44. OF CD44 WITH OTHER EXTRACELLULAR D. INTERACTION MATRIXPROTEINS There is evidence that CD44 can interact with other extracellular matrix proteins in addition to HA. The details, however, are somewhat contradictory, and it is not clear to what extent CD44 functions in intact cells as an adhesion receptor for ECM components other than HA. The class I11 extracellular matrix receptor (ECMR 111)was characterized as a phosphorylated glycoprotein of M, 90,000 that bound to type I and type VI collagen in affinity chromatography experiments or when ECMR I11 was incorporated into liposomes (Wayner and Carter, 1987; Carter and Wayner, 1988). One-dimensional peptide mapping and sequential immunoprecipitation experiments indicated that ECMR I11 and CD44 were closely related or identical entities (Gallatin et al., 1989). Although binding of CD44 to collagen was demonstrated in these experiments when CD44 was incorporated into liposomes, only low levels of binding of radiolabeled soluble CD44 isolated from placenta to collagen I or VI could be demonstrated (St. Jacques et al., 1993). It seems possible that the affinity of CD44 for collagen may be quite low, and it is not clear under what circumstances (or whether) CD44 mediates the adhesion of intact cells to collagen. Fassen and colleagues (1992) have characterized a chondroitin sulfate-containing proteoglycan of mouse melanoma cells that binds to type I collagen. This molecule appears related to CD44, because it is recognized by CD44-specific antibodies on Western blots, although only after treatment with chondroitinase. The melanoma molecule does not seem to function as an adhesion receptor for collagen I, because treatment with P-D-xyloside, which inhibits synthesis of chondroitin sulfate proteoglycans, or chondroitinase does not inhibit attachment of melanoma cells to collagen I. Migration of the cells within collagen gels is, however, inhibited after treatment with P-D-xyloside, suggesting that this molecule functions in mediating cell movement after the cells have attached via other receptors and that the chondroitin sulfate modification is necessary for it to do so. CD44 purified from lymphocytes that has been covalently modified by addition of chondroitin sulfate has been shown to bind fibronectin
CD44 AND ITS INTERACTION WITH ECM
29 1
and laminin as well as collagen I in in vitro binding assays (Jalkanen and Jalkanen, 1992). The presence of the chondroitin sulfate moiety is essential for binding activity, because only the M , 180,000-200,000 chondroitin sulfate-substituted form of CD44 has binding activity; the 90,000 form is inactive. Heparin, heparan sulfate, and chondroitin sulfate inhibit the binding of purified CD44 to fibronectin. These observations suggest that CD44 is interacting with fibronectin via its covalently attached chondroitin sulfate side chains. CD44 purified from placenta, however, binds fibronectin or collagen I only poorly (St. Jacques et al., 1993). Whether the interaction of chondroitin sulfate moieties on lymphocytes with fibronectin occurs in vivo and whether this interaction is important in mediating lymphocyte homing to specific sites is uncertain. IV. CD44 and Lymphocyte Homing
A. EXPRESSION OF CD44 ON HEMATOPOIETIC CELLS 1. Distribution of CD44 during Hematopoietic Development in Mouse and Human Most hematopoietic cells of mouse and human express CD44 at some level, although the degree of expression can be quite heterogeneous, with some cells expressing close to background amounts whereas other cells express large amounts (Trowbridge et al., 1982; Kansas et al., 1989,1990; Horst et al., 1990a). In the mouse, it has been shown that pluripotent stem cells are CD44+ (Trowbridge et al., 1982; Spangrude et al., 1989), and CD44' cells are found in every hematopoietic lineage. In hematopoietic lineages that have been studied in detail, there are differentiation-related changes in the level of CD44 expression (see below). It has been proposed that these changes in CD44 expression during differentiation are related to differences in the adhesion requirements of sessile vs migratory lymphocytes (Horst et al., 1990a) and to bone marrow stromal cell-progenitor cell interactions in the early stages of hematopoiesis (Kansas et al., 1990; Miyake et al., 1990a). In the mouse, the prothymocyte, a progenitor cell in bone marrow capable of homing to and populating the thymus, expresses CD44 (Trowbridge et al., 1982; Spangrude et al., 1989). It is likely, but not certain, that this progenitor cell is a multipotent stem cell, which becomes committed to the T cell lineage at some time after colonization of the thymus (Spangrude and Scollay, 1990; Wu et al., 1991a,b). Once in the thymus, this progenitor retains the CD44+ phenotype and
292
JAYNE LESLEY ET AL.
the ability to home to and repopulate the thymus for an unknown period of time as a “thymus-homing thymocyte progenitor” (Lesley et al., 1985a; Wu et aZ., 1991b). As differentiation of the bone marrowderived progenitor proceeds in the thymus, the a chain of the interleukin 2 receptor (IL-2R) is expressed and CD44 is lost. The kinetics of appearance of populations of “double negative” (CD4-, CD8- ) thymocytes in irradiated thymuses during repopulation by bone marrowderived progenitors indicates the following progression: CD44+/ IL-2R- + CD44+/IL-2R++ CD44-/IL-2R+ + CD44-/IL-2R-, followed by CD4, CD8, and T cell receptor expression (Lesley et al., 1990b; Petrie et al., 1990; Scollay, 1991). CD44 reappears on some more mature thymocytes; however, the number of mature thymocytes expressing CD44 varies among mouse strains (Lesley et al., 1988; Lynch and Ceredig, 1989). Mouse strains expressing the CD44.1 allele, especially BALB/c mice, have relatively large numbers of CD44+ thymocytes, which include relatively mature “single-positive” cells, whereas CD44.2 strains, such as C57BL/6 and AKR/J mice, have few CD44+ thymocytes. The appearance of CD44+ cells in the mouse fetal thymus supports the sequence of CD44 expression described above (Lesley et al., 1985b; Husmann et al., 1988; Penit and Vasseur, 1989). The first cells that can be phenotyped in the mouse fetal thymus are CD44+ (at 12-13 days). As these cells expand and differentiate, IL-2R is expressed and CD44 is lost. This occurs rapidly, over about 3 days, preceding expression of CD4, CD8, and T cell receptor. Reexpression of CD44 on more mature CD8+ and CD4+ single-positive thymocytes of fetal and newborn mice shows the strain variation described above (Lynch and Ceredig, 1989). In peripheral T cells of the mouse, consistent with the differences in CD44 expression in the thymus, all BALB/c T cells are detectably CD44+ at some level, whereas AKR/J and C57BL/6 T cells contain a CD44-negative subpopulation (Lynch and Ceredig, 1989; Lee and Vitteta, 1991; Lesley and Hyman, 1992). Recent thymus emigrants in the periphery of AKR/J mice are heterogeneous in CD44 expression (our unpublished results, 1992)and the thymus is the source of CD44negative peripheral T cells in C57BLi6 mice (Budd et al., 1987b);thus CD44 expression does not appear to be a factor in determining emigration from the thymus. In the human fetal thymus, a pattern of CD44 expression similar to that described in the mouse is observed (Horst et al., 1990b). The earliest lymphoid immigrants to the thymus express CD44 at relatively high levels. CD44 is then downregulated or lost during early differen-
CD44 AND ITS INTERACTION WITH ECM
293
tiation and reexpressed later. Thus most cortical thymocytes are CD44-, whereas medullary thymocytes are CD44+ (Isacke et al., 1986; de 10s Toyos et al., 1989; Kansas et al., 1989; Horst et al., 1990b). The first lymphocytes to appear in fetal lymph nodes are CD44+ T lymphocytes. As lymph node (LN) organization is completed with delineation of B and T cell areas (around 19 weeks of gestation), all lymphocytes are CD44+ (Horst et al., 1990b). Later in LN development and in the adult, LN lymphocytes are strongly CD44+, except for germinal center B cells, which are CD44'" to negative (Kansas et al., 1989; Horst et al., 1990b). In human peripheral blood also, most lymphocytes are CD44+ (Isacke et al., 1986; de 10s Toyos et al., 1989; Kansas et al., 1989; Horst et al., 1990b). B lineage expression of CD44 in human bone marrow cells was studied by using the expression of CDlO to identify immature pre-B cells and CD20 to identify mature B cells (Kansas and Dailey, 1989). CD44 expression was low on immature B cells and upregulated on mature B cells. B cell progenitors were not examined in this study, because they are a minor subpopulation of bone marrow cells, (but see Section IV,B,4). A large panel of non-Hodgkin's lymphomas and lymphoid leukemias thought to represent a complete spectrum of T and B cell development shows a pattern of expression similar to that seen in normal development, with the most immature stages expressing high CD44 levels, immature cells showing low or negative CD44 expression, and mature cells again expressing relatively high CD44 levels (Horst et al., 1990a). From these studies in both mouse and human the general conclusion emerges that CD44 is expressed at relatively high levels during early stages of lymphoid development. At immature stages, the CD44 antigen is temporarily lost from the cell surface to be reacquired later in maturation. CD44 expression in human bone marrow during myeloid and erythroid differentiation was studied by using a panel of cell surface markers to identify different stages of maturation in these lineages (Kansas et al., 1990). In agreement with studies in the mouse, lineage negative undifferentiated cells, presumed to be very early progenitors, were high in CD44 expression. Subsequent changes in CD44 expression showed different patterns in monocyte, granulocyte, and erythroid lineages. In another study, human bone marrow cells were separated by fluorescence-activated cell sorting into CD44hi, CD44m"d, and CD44l" subpopulations and assayed for myeloid and erythroid progenitor activity. Granulocyte-macrophage colonyforming unit and erythroid burst-forming unit precursors were pre-
294
JAYNE LESLEY ET AL.
dominantly in the CD44hifraction (Lewinsohn et al., 1990).It has been suggested that CD44 is a general hematopoietic adhesion molecule, mediating interactions with bone marrow stromal cells (Kansas et al., 1990; Lewinsohn et al., 1990; Miyake et al., 1990a),which are thought to be important in the survival and differentiation of hematopoietic progenitors (Dexter et d.,1977; Whitlock and Witte, 1982).
2. CD44 Expression on Memory I’ Cells and Activated T Cells In C57BL/6 mice, which express low numbers of CD44+ T cells in the thymus and periphery (Lynch and Ceredig, 1989), memory cytotoxic T cell precursors, elicited in response to several different antigens, have been shown to be almost exclusively in the CD44+, CD8+ population of peripheral T cells (Budd et al., 1987a,b). Before immunization, alloresponsive CTLp (cytotoxicT lymphocyte precursors) were found equally in both CD44+ and CD44- populations of CD8+ cells. Eight to 12 weeks after immunization, the frequency of allospecific CTLp increased only in the CD44’ population and, unlike the responding cells of unimmunized mice and responding CD44- cells of immunized mice, the cytotoxicity of these “memory” CD44’ cells was not inhibited by CD8-specific antibody, indicating a higher avidity for the target antigen (Budd et al., 1987a). For nonallogeneic antigens, to which no responding cells were detectable in unimmunized mice, 8to 30-fold higher frequencies of antigen-specific CTLp were found in the CD44’ population than in the CD44- population of CD8+ cells of immunized mice (Budd et al., 1987a,b). These results indicate that elevated CD44 expression is stably acquired as a result of antigenic stimulation in viuo. It has also been shown that murine T cells expressing low levels of CD44 upregulate CD44 expression on stimulation in vitro (Budd et al., 1987b; Lynch et al., 1987).As pointed out by Lynch and Ceredig (1989), the usefulness of CD44 as marker for “memory” or previously activated T cells is limited to mouse strains that express low numbers of CD44+ cells in mature thymus and peripheral T cell populations. Elevated CD44 expression also occurs in helper (CD4+) memory T cells of C57BL/6 mice, defined by functional responses to antigen and by coordinate expression of high CD44 levels with other markers for memory T cells such as low CD45RB and low Mel-14 expression (Butterfield et al., 1989; Swain et al., 1990; Lee and Vitteta, 1991). Furthermore, this “memory” population secretes a different array of lymphokines, characteristic O f Th2 helper cells, on in uitro stimulation as compared to naive helper cells (Swain et al., 1990,1991; Vella et al., 1992). When resting CD44hiand CD.44’” helper T cells (CD4’ ) were
CD44 AND ITS INTERACTION WITH ECM
295
compared for DNA, RNA, and protein content, Stout and Suttles (1992) concluded that CD44himemory T cells (also defined by high levels of asialo-G, and low levels of CD45RB) are maintained in G1 (but not necessarily cycling) rather than resting out of cycle in Go, as were CD44’” CD4+ cells. This conclusion was based on the higher mean RNA content, higher mean total protein content, and stronger responses to in uitro stimulation for CD44hicells as compared to CD44’” cells. Both CD8+ and CD4+ peripheral T cells of the CD44’” or -negative phenotype are gradually depleted following adult thymectomy of C57BL/6 mice (Budd et al., 1987b; Swain et al., 1990), whereas CD44”’ T cells are depleted after a course of treatment with anti-mouse thymocyte serum given intraperitoneally, which preferentially affects circulating T cells (Swain et al., 1990). Furthermore, T lymphocytes expressing high levels of CD44 accumulate with age in mice (Lerner et al., 1989) and in the abnormal T cell populations of mouse strains that develop lymphoproliferative disease (Davidson et al., 1986; Budd et al., 1991). All these data support the proposal that the thymus is the source of CD44’” or CD44- T cells (depending on the mouse strain), and that stimulation in the periphery results in stably elevated CD44 expression. Differences in adhesion receptor expression between previously activated and naive T cells are thought to contribute to the different recirculation patterns (MacKay et al., 1990) and to the different endothelium-binding specificities (Pober and Cotran, 1991) of these two cell types. Two studies directly demonstrate the participation of stimulated cells expressing high levels of CD44 in immunological responses in uiuo. Mobley and Dailey (1992) detected the appearance of a minor Mel-14-negative, CD44hipopulation of CD8’ cells in LN draining an allograft. This population contained all detectable cytolytic activity. A similar Mel-14-negative, CD44hipopulation represented the majority of CD8+ cells in a grafted sponge matrix containing allogeneic cells, and the adhesion molecules LFA-1 and ICAM-1, were also elevated on this population. Rodrigues et al. (1992) employed adoptive transfer of antigen-specific cloned cytolytic cell lines to study the requirements for a protective response against malaria infection. Two sets of clones, which did not differ in epitope fine specificity or ability to lyse target cells in vitro, were either “protective” or “nonprotective” in naive mice injected with malaria sporozoites. High expression of CD44 and VLA-4 correlated with in uiuo protective function, and the importance of CD44 expression was further substantiated by sorting CD44hi and CD44’” subpopulations from a partially protective clone. Sorted
296
JAYNE LESLEY ET AL.
CD44hicells were protective, whereas CD44'"cells were not. Histo-
logical studies of infected liver revealed that injected radiolabeled protective cells were found closely associated with parasitized hepatocytes, whereas nonprotective cells, although present in equal numbers in the liver, were not in close contact with infected hepatocytes. This is the most direct demonstration that CD44 participates in the effector stage of an immunological response. Studies of human lymphocytes also suggest that elevation of CD44 expression is a consequence of activation. "Memory" T cells, recognized by their high expression of LFA-3, LFA-1, and CD2 and by their absence in fetal umbilical cord blood, showed a two-fold elevation in CD44 expression compared to naive cells, and this population gave enhanced responses to soluble antigen in uitro relative to the adhesion receptor-low population (Sanders et nl., 1988). In vitro stimulation of human T cells induced a two-fold increase in CD44 expression and induced increased adhesion to and migration on lymphokine-activated HUVEC (human umbilical vein endothelial cells), which was partially inhibited by mAb (Hermes-3) against CD44 (de 10s Toyos et al., 1989; Oppenheimer-Marks et al., 1990). Haynes et al. (1991) found that CD44 expression was high on lymphocytes and macrophages in synovial tissue of rheumatoid arthritis patients. Two studies have presented evidence that higher molecular weight isoforms of CD44 are transiently expressed in response to antigenic stimulation in uiuo in rats (Arch et al., 1992)and in human T cells in response to phorbol ester or anti-CD3 plus IL-2 stimulation in vitro (Koopman et al., 1993). In studies of macaque lymphocytes, Willerford et al. (1989) concluded that activated lymphocytes were CD44hi,expressing 5- to 10fold more CD44 than CD44l" cells, and that CD44'" lymphocytes upregulated CD44 during in vitro activation. They suggest that CD44hi and CD44l" lymphocytes follow different migratory pathways in the circulation, because splenic and peripheral blood cells were mostly CD44hi, whereas recent emigrants from lymph node and Peyer's patches, collected from the thoracic duct, were predominantly CD44'".
B. PHYSIOLOGICAL EXPERIMENTS IMPLYING A ROLE FOR CD44 IN LYMPHOCYTE HOMING AND HEM.4TOPOIESIS 1. Relationship to Hermes Antigen Mediating Peripheral Lymph Node Adhesion Lymphocyte migration into secondary lymphoid tissues occurs via adhesion to a specialized endotheliurn identified histologically as high endothelial venules (HEV) in lymph nodes and Peyer's patches (Woodruff et aZ., 1987). Stamper and Woodruff (1976)developed an in
CD44 AND ITS INTERACTION WITH ECM
297
uitro assay to measure the capability of circulating lymphocytes to specifically recognize and bind to HEV. In this assay, lymphocyte populations are incubated with frozen sections of lymphoid tissue. After washing off unbound cells, cells adhering to the high endothelium of blood vessels are quantitated microscopically. Using this assay, Mel-14 mAb, which recognizes murine L-selectin (Lasky et al., 1989), was identified on the basis of its ability to inhibit murine lymphocyte and lymphoid cell line adhesion to HEV (Gallatin et al., 1983). The Hermes antigen of human lymphocytes was identified by crossreactivity with Mel-14 mAb and given the designation “lymphocyte homing receptor” (Jalkanen et al., 1986), but antibodies raised against Hermes antigen were later found to recognize human CD44 (Gallatin et al., 1989; Picker et al., 1989b; Stefanovd et al., 1989; St. John et al., 1990). One of the Hermes-specific antibodies, Hermes-3, and a polyclonal antiserum produced against the isolated antigen blocked adhesion of human lymphocytes to frozen sections of HEV (Jalkanen et al., 1987), as Mel-14 does in the murine system. Pals et al. (1989a) found that two other mAbs specific for human CD44 caused partial inhibition of lymphocyte HEV binding. It is surprising that different cell adhesion molecules appear to mediate this activity in different species, yet this seems to be the case: Culty et al. (1990) showed that binding of murine cells to HEV is not sensitive to hyaluronidase or to mAb known to block HA binding by CD44. In the mouse, carbohydrate ligands other than HA have been shown to mediate adhesion via Mel-14 (Imai et al., 1990). Thus there is no evidence that CD44-HA interactions participate in lymphocyte adhesion to HEV in the mouse, although they can mediate adhesion to cultured endothelial cell lines (Lesley et al., 1992; Uhlig et al., 1993). Hermes-3 mAb, which blocks human lymphocyte adhesion to frozen sections of HEV, does not block HA-dependent adhesion of a B cell line transfected with human CD44 to cultured rat endothelium (Stamenkovic et al., 1991) or [ 3H]HA binding by detergent extracts of human CD44’ cells (Culty et al., 1990). This is consistent with the report that the determinant recognized by the Hermes-3 mAb maps to the membrane-proximal portion of the external domain of CD44 (Goldstein et al., 1989), whereas the predicted HA-binding sequences are near the amino-terminal, membrane-distal portion of the molecule (see Section 111). These results imply that Hermes-3-sensitive binding of human lymphocytes to HEV comprises a CD44-ligand interaction distinct from the CD44-HA interaction described in Section 111. However, because the role of HA in CD44-mediated binding of human lymphocytes to HEV in the frozen section assay has not been tested,
298
JAYNE LESLEY ET AL.
other explanations are possible. Hermes-3 mAb could indirectly affect CD44-mediated HA binding in normal lymphocytes, for example, by interfering with CD44 interaction with other molecules on the cell surface involved in regulating its HA-binding function (see Section V). Such regulatory interactions may not occur in the transfected B cell system studied by Aruffo et at. (1990) or in the solubilized receptor[ 3H]HA-binding assay of Culty et al. (1990). In conclusion, it has not been shown, in either murine or human systems, that CD44-mediated binding to HA is involved in the proposed “homing receptor” activity of CD44, defined experimentally by CD44-dependent lymphocyte binding to HEV in the frozen section assay (Jalkanen et al., 1987). However, CD44-mediated adhesion of lymphocytes to cultured endothelial cells and cell lines, and to stromal cell lines, has been shown to involve HA binding (Aruffo et at., 1990; Stamenkovic et al., 1991; Miyake et al., 1990b; Lesley et al., 1992), indicating that CD44 binding of HA may be involved in lymphocyteendothelial interactions under certain circumstances.
2. In Vivo Role of CD44 in Lymphocyte Migration The adhesion of lymphocytes to HEV in frozen sections of lymphoid tissue in uitro is thought to reflect in uiuo interactions of circulating blood cells with specific endothelial structures that determine selective entry of cells into distinct organ:;. In mice, extravasation of lymphocytes has been studied in uiuo by intravenous injection of radiolabeled lymphocytes and determination of the distribution of injected cells by counting radioactivity in various tissues. Uhlig and colleagues (1993) found no effect of Fab fragments of CD44-specific mAb on gross localization of labeled mature lymphocytes one hour after injection into mice. In similar experiments, LFA-l-specific mAb and Mel-14 mAb specific for L-selectin, which inhibits murine lymphocyte adhesion to HEV in frozen section assays, did inhibit localization in LN (Gallatin et al., 1983; Hamann et al., 1988, 1991).Injection of CD44specific antibody in uiuo causes loss ofCD44 from the lymphocyte cell surface (Camp et al., 1993b; see Section V,C,4). Camp et al. (1993b) found that peripheral L N lymphocytes, stripped of cell surface CD44 by in uiuo exposure to CD44-specific mAb, entered lymphoid organs normally. These studies argue against a role for CD44 in the extravasation of lymphocytes into lymphoid organs during normal trafficking in the mouse. Removal of cell surface CD44 by exposure to CD44-specific mAb, however, did result in inhibition of edema and leukocyte infiltration at a site of cutaneous delayed-type hypersensitivity 24 hours after chal-
CD44 AND ITS INTERACTION WITH ECM
299
lenge (Camp et al., 1993b). Subsequently, a normal inflammatory response developed with the infiltrating cells reexpressing CD44. Even in these mice, which were responding to immunological challenge, there was no effect of CD44 mAb treatment on the number of lymphocytes in draining LN, although the majority of these lymphocytes remained CD44 negative at 72 hours postchallenge. The authors suggest that CD44 is required for extravasation into an inflammatory lesion involving nonlymphoid tissue, but not for normal leukocyte recirculation through lymphoid organs.
3. Inhibition of Migration of Thymocyte Progenitors CD44 was shown to be present on pluripotent bone marrow stem cells that give rise to multilineage 10-day spleen colonies and on the bone marrow prothymocytes that populate the thymus (Trowbridge et al., 1982). Cytotoxic depletion of bone marrow, using CD44-specific mAb and complement, dramatically reduced these activities in irradiated, bone marrow-repopulated mice. The possibility that CD44 might be involved in homing of these two progenitors was suggested by the fact that incubation with mAb alone was as effective at inhibiting repopulation as incubation with mAb and complement. Similarly, repopulation by a thymus homing progenitor resident in the thymus whose progeny can be detected in the thymus at 12-15 days postirradiation was prevented by pretreatment of progenitor cell enriched populations with CD44-specific mAb alone (Lesley et al., 1985a). It is possible, however, that antibody acted to remove these progenitors from the circulation via opsonization and phagocytosis in the liver rather than by blockade of a homing receptor, because the experiments used intact antibody (IgG2, and IgGZb) rather than Fab antibody fragments. 4 . Inhibition of Hematopoiesis in Vitro by CD44-Spec$c Antibodies Monoclonal antibodies to CD44 were found to inhibit B cell lymphohematopoiesis completely in long-term bone marrow cultures (Miyake et al., 1990a). The antibodies used in these experiments were initially prepared and selected on the basis of their ability to inhibit adhesion of a B lineage hybridoma to a cloned stromal cell line. Further definition of that model revealed that CD44 on the lymphoid cells was recognizing HA on the surface of stromal cells (Miyake et al., 1990b). However, the actual mechanism of inhibition of long-term cultures remains uncertain. Addition of CD44-specific antibodies to cultures that were already well established had no effect, whereas antibodies to either VCAM-1 or VLA-4 dislodged B lymphocyte pre-
300
JAYNE LESLEY ET AL.
cursors from the stroma at that time (Miyake et al., 1991).This indicates that CD44 and its ligands are important for an early step in culture initiation, but not for later interactions of relatively abundant lymphohematopoietic progenitor cells with the microenvironment. Because of the ubiquitous nature of CD44 and the complexity of in vivo experiments, the role of this molecule and its ligands in normal lymphohematopoiesis remains an open issue.
5. Signaling through CD44-Ligand Znteractions A number of studies have shown that treatment of human T cells with antibodies specific for CD44 in conjunction with other activation signals can result in either enhancement or inhibition of lymphocyte responses (see Table 11). These experiments are believed to mimic possible effects of engagement of CD44 with its ligand on stimulation through other receptor-ligand interactions, such as T cell receptor binding to antigen. On the basis of observed stimulatory effects, it is concluded that interaction of CD44 with an unknown ligand on endothelial cells or antigen-presenting cells may enhance the ability of circulating T cells to respond to antigen-specific signals. This conclusion is indirect, and a few points should be noted about these experiments: CD44-specific mAbs alone usually do not elicit any activity; not all CD44-specific mAbs have the described activities; the primary stimuli being influenced by CD44-specific mAbs are often suboptimal; some of the effects seen are dependent on the presence of monocytes in peripheral blood lymphocyte (PBL) cultures, so different results are obtained with purified T cells. Denning et al. (1990) suggested that CD44 may enhance T cell activation by increasing T cell-monocyte interactions through stimulation of LFA-l/ICAM-mediated or CD2/LFA-S-mediated adhesion. It has been shown that CD44-specific mAb treatment of human T lymphocytes can induce LFA-l-mediated cell aggregation (Pals et al., 1989a; Koopman et al., 1990).Also, overexpression of baboon CD44 in murine fibroblasts resulted in increased spontaneous cell aggregation that probably did not involve LFA-1 (St. John et al., 1990). Only a few studies, involving myeloid lineage cells, have examined responses of hematopoietic cells to HA. Hiro et al. (1986) found that soluble HA stimulated IL-1 production by cultured human monocytes and rabbit macrophages. Although CD44 was not directly implicated in this study, the specificity of the reaction for HA, but not other glycosaminoglycans, suggests that CD44 may be the receptor. More recently, Noble et al. (1993)showed that murine bone marrow-derived macrophages are stimulated by HA to synthesize mRNA for IL-lp,
TABLE I1 In Vitro EFFECTS OF C D ~ ~ - S P E C IMONOCLONAL FIC ANTIBODIESON HUMAN PERIPHERAL BLOOD LYMPHOCYTE RESPONSES CD44-specific mAb H90 and Fab of H90 NIH-44.1 A3D8 A3D8 Fab of A3D8 212.3
Suboptimal, CD2-mAb pairs Immobilized CD&mAb Suboptimal, CD2-mAb pair Suboptimal, immobilized CD3-mAb None None
Increased T cell proliferation Increased T cell proliferation, induced ILPR expression Increased T cell proliferation Increased T cell proliferation Partial inhibition of E rosettes Inhibition of T cell E rosettes
Suboptimal, CD3-mAb Suboptimal, CD2-mAb pair High levels, CD2-mAb pair Suboptimal, CD2-mAb pair Optimal, immobilized CD3-mAb
Increased T cell proliferation Increased T cell proliferation (monocyte dependent) Increased T cell proliferation (monocyte independent) Increased T cell proliferation (monocyte dependent) Inhibited T cell proliferation, IL-2 production, IL-2R expression, CaZ+ flux Increased T cell proliferation No effect Inhibited T cell proliferation Redirected CTL-mediated lysis to antigen-negative, Fc-receptor+ target Inhibited CTL-mediated lysis of antigen+ target Stimulated lymphokine production by IL-2-cultured T cells (LAK-T cells) Inhibited lymphokine production by LAK-T cells Enhanced NK lysis of NK targets Enhanced NK lysis of NK targets Increased T cell proliferation Increased T cell proliferation Increased T cell proliferation Induced monocytedependent T cell proliferation Induced monocyte-independent T cell proliferation
Suboptimal, CD2-mAb pair PHA, PMA" Fab of 212.3 Optimal, immobilized CD3-mAb 9F3 None F(ab')e of 9F3 None NIH-44 Immobilized, CD3-mAb CD4-mAb + tumor cells s5 None FabofS5 None 8B2.5 Submitogenic PMA Immobilized CD3-mAb CDZ-mAb Immobilized None 8B2.5 None a
Effect (result)
Costimulation
PHA, Phytohemagglutinin; PMA, phorbol myristate acetate.
Ref. Huet et al. (1989) Shimizu et al. (1989) Hale and Haynes (1989) Denning et al. (1990)
Rothman et al. (1991)
Seth et al(l991) Chong et al. (1992) Tan et al. (1993) Pierres et 01. (1992)
302
JAYNE LESLEY ET AL.
tumor necrosis factor a (TNF-a) and insulin-like growth factor 1 (IGF-1) and to secrete IGF-1. This activity was inhibited by CD44specific mAb, suggesting that a CD44-HA interaction on the macrophage cell surface might initiate a cytokine cascade. Stimulation of macrophages with CD44-specific mAb can result in release of cytokines: TNF, IL-1, and macrophage colony-stimulating factor (M-CSF) (Webb et al., 1990; Gruber et al., 1992). Hyaluronan has also been shown to stimulate neutrophil migration in Boyden chamber assays, in conjunction with fibronectin (Hakansson and Venge, 1985), and this could be mediated by CD44, which is abundant on neutrophils. V. Regulation of Interaction of CD44 with Extracellular Matrix
A. EVIDENCE THATHYALURONAN BINDINGIs REGULATED 1 . Cell Lines Evidence that the ECM receptor function of CD44 is regulated first came from studies of HA binding by murine hematopoietic cell lines expressing CD44. Although some CD44-positive cell lines bind HA in a CD44-dependent manner, as demonstrated by the inhibition of HA binding with certain CD44-specific mAb, many CD44-expressing cell lines do not bind HA (Table 111; and Lesley et al., 1990a; Miyake and Kincade, 1990). All of these cell lines express the hematopoietic isoform of CD44. Hyaluronan receptor activity is not masked by endogenous HA in these cell lines, because treatment with hyaluronidase or chondroitinase ABC does not reveal HA-binding function (Lesley et al., 1990a).Some T cell lines could be induced to bind HA after culture in phorbol ester. Maximal induction required 16 hours of incubation and was accompanied by increased CD44 expression but no change in the isoform expressed (Lesley et al., 1990a; Hyman et al., 1991).It does not appear, however, that differences in the HA-binding capacity of CD44-positive cell lines can be accounted for solely by quantitative differences in the level of CD44 expressed on the cell surface: a CD44hivariant selected by fluorescence-activated cell sorting for high levels of CD44 expression did not bind HA, whereas an HA-binding variant selected from the same parental line expressed less CD44 than the CD44hivariant (Hyman et al., 1991). Several B cell lines could not be induced to bind HA by culture in phorbol ester or lipopolysaccharide. Among CD44-positive B cell lines, HA-binding activity was observed only in the most mature cell types, those secreting immunoglobulin (Lesley et al., 1990a; Miyake and Kincade, 1990, and our unpublished results, 1992).
303
CD44 AND ITS INTERACTION WITH ECM
TABLE I11 CD44 EXPRESSION AND HYALURONAN BINDING BY CELLLINES^
Cell type
T cell
B cell
Fibroblast
Cell line AKR 1 s49 SAKRTLS 12 CD44hivariant of SAKRTLS Ha+ variant of SAKRTLS EL4 BW5147 S 194 (myeloma) C1.18 (myeloma) BM2 (hybridoma) 70Z/3 (pre-B) RAW 253 (pre-B) Sp2/0 (myeloma) XS63 (myeloma) L cell
CD44 expressionb
HA binding‘
Induced HA bindine
ndr
-
+ + +
+
-
+ + + + + + +
-
+ + + + +
nd nd
+
Adhesion to immobilized HAe
+ + + +
nd nd -
nd nd -
+ + +
nd nd nd nd
+ -
+ +
nd
” References: Lesley et al. (1990a. 1992). Miyake and Kincade (1990),Hyman et al. (1991); and our unpublished results (1992). CD44 expression was determined by flow cytometery with fluorescein-conjugated mAb, IM7; + indicates fluorescein staining greater than threefold over background of unstained cells. HA binding was determined by flow cytometry with fluorescein-conjugated HA; + indicates fluorescein staining greater than threefold over background of unstained cells. Induced HA binding was determined by flow cytornetry with fluorescein-conjugated HA and the inducing rnAb, IRAWB14; + indicates an increase in binding at least threefold over that seen with fluoresceinconjugated HA alone. Adhesion was determined by counting radiolabeled cells bound to HA immobilized on plastic culture wells. ’nd, Not determined.
Certain CD44-specific mAbs, such as IRAWB 14 (Lesley et al., 1992), “activated” HA binding by CD44 in some T cell lines that expressed CD44 but did not bind HA constitutively (see Table 111). The term “activate” is used to signify the rapid conversion of preexisting CD44 molecules on the cell surface from a state that does not bind soluble HA to a state that does bind HA. The rapid activation of CD44 HA-binding function by mAb was clearly distinguishable from induction of HA binding observed after exposure of cell lines to phorbol ester. Phorbol ester induction occurred over several hours in culture at 37°C and was accompanied by large increases in CD44 expression, indicating cell differentiation (Lesley et al., 1990a; Hyman et al.,
304
JAYNE LESLEY ET AL
1991). Activation by CD44-specific mAb occurred at 0°C and in the absence of divalent cations (Lesley and Hyman, 1992). It did not involve intracellular signaling, because it occurred immediately on binding antibody at 0°C and in cells that had been prefixed. These results suggest a direct effect of the antibody on CD44 molecules on the cell surface. Monovalent Fab fragments were inactive, indicating that crosslinking of CD44 molecules is required and suggesting that HA-binding function may involve clustering of CD44 into a multivalent structure (Lesley et al., 1993). Not all cell lines that express CD44 and fail to bind HA were activated by mAb. Pre-B lines, such as RAW 253, could not be induced to bind HA by IRAWB 14, nor could L cells (Table 111). Some CD44-positive cell lines that did not bind soluble HA did, however, bind HA immobilized on plastic surfaces. This was true of the T cell lines that could be induced by CD44-specific mAb to bind soluble HA (see Table 111),and of a transfected T cell line expressing a mutant CD44 molecule with a truncated cytoplasmic domain (Lesley et aZ., 1992). This suggests that either (1) the immobilized ligand represents a different structure that can be recognized by CD44 on these cells whereas soluble ligand cannot, or (2) the immobilized ligand itself activates the HA-binding function of CD44, perhaps b y stabilizing low-affinity interactions through clustering of CD44 molecules into multivalent receptors. The CD44-positive pre-B cell line, RAW 253, which is not inducible by IRAWB 14 mAb, does not bind immobilized HA in the presence or absence of the mAb. Data on HA binding by human hematopoietic cell lines or cells of other species are lacking. However, several brain-derived human cell lines were able to bind HA via CD44 only if they were pretreated with hyaluronidase (Asher and Bignami, 1992). In these studies also, not all CD44' lines bound HA, although all expressed the same hematopoietic CD44 isoform. A human melanoma line sorted for high and low CD44 expression bound immobilized HA, and the extent of binding correlated with the level of CD44 expression (Birch et al., 1991).
2. Normal Hematopoietic Cells In view of the broad tissue distribution of both CD44 and its proposed ligands, it is not surprising that receptor-ligand interactions are restricted, especially among circulating cells. Indeed, in the mouse, normal, resting hematopoietic cells that express CD44 could not be shown to bind either soluble or immobilized HA (Lesley et al., 1990a, 1992; Murakami et al., 1990, 1991; Lesley and Hyman, 1992). Hyal-
CD44 AND ITS INTERACTION WITH ECM
305
uronidase treatment of bone marrow cells, among which myeloid precursors express especially high CD44 levels, did not expose HA binding function. In uitro activation of T cells with phorbol ester, ionomycin, concanavalin A (ConA), CD3-specific antibody, and several lymphokines after both short term (20 min at 37°C) and longer term (overnight culture) treatment, and of B cells with lipopolysaccharide (overnight culture), did not induce detectable HA binding (Lesley et al., 1990a; Lesley and Hyman, 1992). A population of HA-binding B cells could, however, be induced by culture for several days in IL-5 or by a chronic graft-versus-host reaction in uiuo (Murakami et al., 1990, 1991). The HA-binding cells induced under these conditions were secreting IgM antibody. The conditions and the time required to induce these HA-binding B cells were indicative of selection and differentiation processes. Although in uitro stimulation of normal T cells did not induce HA binding, it has been possible to demonstrate that mouse T cells expressing CD44 can bind both soluble and immobilized HA after exposure to certain CD44-specific mAbs. The extent of HA binding induced by the CD44-specific mAb IRAWB 14 in splenic T cells was proportional to the level of CD44 expressed (Lesley and Hyman, 1992), which varies among mouse strains (Lynch and Ceredig, 1989; Lesley et al., 1988) and with cell activation (see Section IV,A,2). The same antibody, however, did not induce bone marrow cells or splenic B cells to bind HA. As with T cell lines, the mAb IRAWB 14 activated HA binding on normal T cells immediately on binding to preexisting cell surface CD44 molecules in a temperature- and cation-independent manner. These demonstrations of CD44-mediated HA binding by normal T and B cells under specific conditions suggest that CD44-dependent HA binding may occur naturally in uiuo. Yet, if this is true, the absence of binding activity among hematopoietic cells taken from normal mice indicates that CD44-mediated binding of HA must be tightly regulated. The circumstances under which the HA receptor function of CD44 may be called on in uiuo, and the mechanisms by which receptor function is activated at the appropriate time and place, are unknown. Indeed, the role of CD44-mediated ECMbinding activity in hematopoietic cell function is unknown. But the association of elevated levels of adhesion receptors, including CD44, with effector cells engaged in immunological reactions and with memory cell phenotype and the effects of CD44-specific mAb on T cell function (see Section IV) suggest that CD44 does participate in lymphocyte function.
306
JAYNE LESLEY ET AL.
3. Three Activation States of Hematopoietic CD44 The evidence from both normal murine lymphocytes and lymphoid cell lines suggests that there are three possible states of HA receptor function for CD44 (Lesley and Hyman, 1992, He et al., 1992): (1)nonactivatable, corresponding to resting B cells, some pre-B lymphomas, and bone marrow cells, (2) activatable (able to rapidly convert preexisting CD44 to function to bind HA, e.g., by IRAWB 14 monoclonal antibody), represented by resting CD44+ T cells and some T lymphomas (Lesley et al., 1992), and ( 3 )constitutively active, or able to bind HA without activation, represented by some B and T cell lines (Table 111) and IgM-secreting B cells induced by IL-5 in vitro (Murakami et aZ., 1990) or graft-versus-host reaction in uiuo (Murakami et al., 1991). The conversion of CD44-expressing cells from one HA receptor state to another may be a relatively slow process involving maturation/differentiation, or it may be rapid, as in the conversion of activatable cells to HA-binding function by CD44specific mAb. A survey of B lineage cell lines “frozen” in different stages of differentiation (Coffman, 1982) indicates a progression from state 1, to state 2, to state 3 during differentiation from pre-B to immunoglobulin-secreting cells (our unpublished results, 1992). Among normal B cells, only nonactivatable (state 1; Lesley and Hyman, 1992) and constitutively active (state 3; Murakami et al., 1990, 1991) states have been observed. Resting T cells that express CD44, on the other hand, were activatable (state 2), although their ability to bind HA once activated was also a function of the level of CD44 expression (Lesley et al., 1992; Lesley and Hyman, 1992). The activatable state could allow rapid and reversible engagement of CD44-HA interactions under specific circumstances. As noted above (see Section V,A,2), HA binding in normal T cells was not activated by a number of reagents that induce or mimic stimulation through the T cell receptor and that have been shown to increase ligand binding by other adhesion molecules (see below, Section V,B). Thus the physiological inducers of CD44 activation in T cells are unknown. Possible mechanisms are discussed below (Section V,C). 4 . Control of Regulation of CD44 Function by Cellular Environment Differences among cells in the HA-binding function of CD44 imply that this function is closely regulated by the cell. The best evidence that the cellular environment influences the ECM-binding function of CD44 comes from the transfection of identical CD44 constructs into different cell types (see Table IV). When the CD44-
307
CD44 AND ITS INTERACTION WITH ECM
TABLE IV POSITIVE AND NEGATIVEREGULATION OF HYALURONAN BINDING IN TRANSFECTED CELLLINESO CD44 expression Parental cell line
Transfection
None Wild-type CD44.1e ACYf ANCg MI isoform (variant F)” M2 isoform (variant G) M3 isoform (variant I) M4 isoform (variant J) L cells (CD44.2+) None Wild-type CD44.1 None S49 (CD44.1-) Wild-type CD44.2 RAW 253 (CD44.1+) None Wild-type CD44.2 AKRl (CD44.2-)
CD44.1
HA CD44.2b binding
InducedHA bindingd
-
+ + + +
+ + + -
+
-
+ +
References: Lesley et al. (1992); He et al. (1992); and our unpublished results (1992).
’Expression of CD44 alleles was determined by flow cytometry with fluorescein-conjugated
allele-specific mAb, RAMB44 for CD44.1 and C71/26 for CD44.2; + indicates fluorescein staining greater than threefold over background of unstained cells. Hyaluronin binding was determined by flow cytometry with fluorescein-conjugatedHA; + indicates fluorescein staining greater than threefold over background of unstained cells. Induced HA binding was determined by flow cytometry with fluorescein-conjugated HAand the inducing mAb, IRAWB14; + indicates an increase in binding ofat least threefold over that seen with fluorescein-conjugated HA alone. wt, wild-type CD44, “hematopoietic” or “standard’ form. JACY is a transfectant expressing a mutant CD44.1 molecule lacking all but the first six amino acids of the cytoplasmic domain (Lesley et ol., 1992). 8 ANC is a transfectant expressing a mutant CD44.1 molecule lacking amino acids 162 to 244 of the membrane-proximal region of the external domain (He et al., 1992). M1-M4 isoforms are transfectants expressing CD44.1 isoforms with inserted sequences in the membrane-proximal region of the external domain, corresponding to splice variants F, G, I, and J illustrated in Fig. 2 (He et al., 1992).
308
JAYNE LESLEY ET AL.
negative murine T cell lymphoma AKRl (genotype CD44.2) was transfected with the hematopoietic form of CD44.1 cDNA, the resulting transfectants were able to bind HA as measured in several different assay systems (Lesley et al., 1992). This cell line was also permissive for HA binding when it expressed higher molecular weight CD44 isoforms (He et al., 1992). We have also transfected a construct coding for the hematopoietic form of CD44.1 into L cells. Although L cells express endogenous CD44.2, they do not bind HA and cannot be induced to bind by mAb (Table 111). The transfected CD44.1 construct was expressed at levels equivalent to the endogenous CD44.2, but it did not confer HA-binding function on L cells (Table IV). Similarly, a construct coding for the CD44.2 hematopoietic form was transfected into a CD44-negative (CD44.1 genotype) T lymphoma, S49, and into the CD44.l-positive pre-B lymphoma RAW 253, which does not bind HA in either the absence or presence of the inducing mAb IRAWB 14. The S49 transfectants acquired CD44-dependent HA-binding function, but the RAW 253 transfectants remained unable to bind HA (Table IV). It is noteworthy that the transfected CD44 molecules expressed in L cells and RAW 253 cells, which were identified and immunoprecipitated with allele-specific mAb (Lesley and Trowbridge, 1982), were of the same molecular weight as the endogenous CD44 molecules, indicating that they were subject to the same posttranslational modifications. The molecular weight of the CD44.2 molecule expressed in RAW 253 was different than that of the same construct expressed in S49 cells, indicating differences in the posttranslational modification of CD44 in the two cell types (our unpublished results, 1992).
B. PARALLELS WITH REGULATORY MECHANISMS OF OTHER ADHESIONMOLECULES The failure of many CD44-expressing cells to bind HA, and the ability to rapidly activate HA binding under specific conditions, suggests that CD44 is a member of a growing list of cell adhesion molecules that require activation to bind ligand most efficiently. Indeed, the ability to rapidly and transiently increase ligand-binding affinity may be a general property of leukocyte adhesion molecules (Dustin and Springer, 1991; Spertini et al., 1991). This list includes many of the integrins (see Dustin and Springer, 1991; Hynes, 1991) and Lselectin (Spertini et al., 1991). These adhesion systems exhibit regulatory features that are similar to the CD44-HA interactions we have described: they exhibit transitions in ligand-binding function on cell
CD44 AND ITS INTERACTION WITH ECM
309
differentiation and activation, and some show rapid activation by certain receptor-specific mAb. 1. Platelet gpIIb-IIIa (all&) lntegrin Best studied is the platelet integrin gpIIb-IIIa, which on resting circulating platelets does not bind soluble ligands, although it can bind immobilized fibrinogen. Soluble fibrinogen could be bound after platelet cell activation by phorbol ester or physiological agonists, and this activation involved intracellular signaling (Shattil and Brass, 1987). Thus the activated platelet is able to modulate the affinity of the integrin receptor for its ligand from within. Receptor activity could also be induced by certain mAbs that bound the integrin (O’Toole et al., 1990) and by binding of ligands (Du et al., 1991). Solubilized integrin could be activated to bind soluble ligand by mAb as a monovalent Fab fragment (O’Toole et al., 1990), and the activated integrin could be specifically recognized by certain mAbs that did not bind integrin on resting cells (Shattil et al., 1985). On the basis of these and other observations, platelet integrin activation is thought to involve a conformational change in the molecule. The cytoplasmic domain of gpIIb-IIIa has been implicated in controlling transitions between activation states (O’Toole et al., 1991).
2. LFA-1 The a ~ & integrin (LFA-1) on T lymphocytes participates in the interaction between T lymphocytes and antigen-presenting cells and
between T lymphoctyes and target cells, only if activated through stimulation of the T cell receptor or CD2 (Dustin and Springer, 1989; van Kooyk et al., 1989). As in the case of CD44 activation, an LFA-1specific mAb, NK1-L16, could activate LFA-1-mediated adhesion in some T cells (Keizer et al., 1988); however, unlike the CD44 case, in which phorbol esters failed to activate HA binding in normal T cells, phorbol ester treatment did induce rapid LFA-1 activation. Figdor et al. (1990) suggested three activation states of LFA-1 on human PBLs: (1)inactive LFA-1, on resting PBLs that express little or none of the LFA-1 epitope recognized by NK1-L16, (2) intermediate or activatable LFA-1, expressing the NKl-L16 epitope but not LFA-1 adhesive activity, and (3)active LFA-1, which mediates adhesion. The authors proposed that progression from the resting to the activatable state involves maturation and perhaps Ca2+-dependent clustering of receptors (Figdor et al., 1990). As with the platelet receptor gpIIbIIIa, it is thought that cells regulate the activity of LFA-1 through
310
JAYNE LESLEY ET AL.
intracellular interactions with the cytoplasmic domain of the integrin (Hibbs et al., 1991).
3. Mac-1 /CR3 The a& integrin on monocytes and neutrophils, also referred to as CR3 (complement receptor 3) or Mac-1, was the first integrin found to be activatable. On monocytes and neutrophils, the complement receptor function could be rapidly activated by phorbol esters (Wright and Silverstein, 1982) and this was thought to involve receptor clustering (Detmers et al., 1987). A change in receptor conformation upon activation to bind fibrinogen could be detected by a mAb specific for the activated state (Altieri and Edgington, 1988) and was dependent on extracellular calcium (Graham and Brown, 1991). Although cations are thought to play a role in activation of most of the integrins, there is no evidence for any such requirement for CD44 function. 4 . L-Selectin L-Selectin, a member of the family of adhesion receptors containing a lectin-like domain, expressed on lymphocytes and neutrophils, was activated by lineage-specific stimuli to increase its affinity for a carbohydrate ligand (Spertini et al., 1991). Lymphocyte L-Selectin was activated by crosslinking of T cell receptor or CD2, whereas L-Selectin activity on neutrophils was stimulated by G-CSF, GMCSF, and TNF-a. These transitions occurred in minutes and without changes in the level of receptor expression. L-Selectin-mediated binding of murine lymphocytes to HEV was also increased by these treatments. C. POSSIBLE MECHANISMSFOR REGULATING CD44 RECEPTORFUNCTION So far, we have presented evidence that CD44 on hematopoietic cells can exhibit three states of HA-receptor funtion. The functional state of CD44 in a particular cell type is related to the cell type and its state of differentiation (Tables I11 and IV, and Section V,A). The conversion between functional states may take place through differentiation and thus involve the synthesis of an array of new molecules that could influence CD44 function, or it might take place by rapid activation of preexisting CD44 molecules, as is seen on exposure to CD44-specific mAb. Below, we discuss several possible nonexclusive mechanisms by which cells might regulate the ECM receptor function of CD44: (1)interaction of the cytoplasmic domain with in-
CD44 AND ITS INTERACTION WITH ECM
311
tracellular proteins, such as elements of the cytoskeleton, (2) modification of the cytoplasmic domain by phosphorylation, (3)posttranslational modification of the extracellular domain by carbohydrate addition, (4) expression of alternate CD44 isoforms, (5) interaction with other cell surface molecules, (6) interaction with other extracellular ligands, and (7) masking or shedding of cell surface CD44.
1 . Regulation through the Cytoplasmic Domain and the Possible Role of Phosphorylation Evidence for a role for the cytoplasmic domain in regulating HA binding comes from transfection experiments with CD44 constructs coding for CD44 molecules with truncated cytoplasmic domains (Lesley et al., 1992; Thomas et al., 1992). Although the wild-type hematopoietic form of CD44 binds HA when transfected into the T lymphoma AKR1, a “tailless” CD44 mutant molecule lacking all but the first six amino acids of the cytoplasmic domain does not bind HA in solution and cells transfected with this construct adhere poorly to immobilized HA (Lesley et al., 1992). Thomas et al. (1992) also found that melanoma cells transfected with mutant CD44 constructs coding for deletions in the cytoplasmic domain bound poorly to immobilized HA and did not migrate on an HA substrate, whereas transfectants expressing a wild-type CD44 construct did (see Section 111). These studies suggest two possible roles for the cytoplasmic domain of CD44: (1) an inside-out control of CD44-ligand-binding function, possibly by influencing the cell surface distribution of CD44 and/or conformation of the external domain {as has also been suggested for some integrins [see Dustin and Springer (1991) and Hynes (1992)l and for L-selectin (Spertini et uZ. 1991)]}, and (2) an outside-in signaling function, whereby cell motility or other cell functions could be activated in response to CD44 binding to ligand, possibly by engagement of the cytoskeleton or by interaction with intracellular signaling mechanisms. The tailless CD44 construct expressed in AKRl cells does not bind soluble HA but can be activated to bind b y exposure to the CD44specific mAb IRAWB 14 (Lesley et al., 1992). As described above (Section V,A, l), this antibody-induced activation does not involve intracellular signaling and cannot be mediated by monovalent Fab fragments (Lesley et al., 1993).This suggests that clustering of CD44 molecules into a multimeric configuration may be required for HA binding, and that the cytoplasmic domain of CD44 may influence HA binding by mediating the formation of a multimeric receptor. CD44 aggregation on the cell surface could be influenced by interaction of
312
JAYNE LESLEY ET AL.
the cytoplasmic domain with cytoskeletal elements or with other intracellular molecules, and this in turn might involve changes in phosphorylation of the cytoplasmic domain. It is also possible that alterations in the cytoplasmic domain could mediate a conformational change in the external domain of CD44. Interactions of CD44 with cytoskeletal elements, through its cytoplasmic domain, have been reported in several cell systems. In fibroblasts, Jacobson et al. (1984) related the distribution of actin stress fibers to that of CD44 crosslinked into patches by primary plus secondary antibodies. About 50% of cell surface gp85 (CD44) in BHK cells described by Tarone et al. (1984) was found to be Triton X-100 insoluble and distributed with actin filaments. Lacey and Underhill (1987) showed that a portion of the HA receptor of fibroblasts (CD44) was associated with cytoskeletal actin filaments in the Triton X-100insoluble fraction of Swiss 3T3 cells. Bourguignon et nl. (1986) found an association between lymphoma gp85 and an ankyrin-like protein that also binds to actin and fodrin in lymphoma cells whose surface molecules were crosslinked with lectin or antibody. Kalomiris and Bourguignon (1988) later showed that purified gp85 (CD44) bound to purified erythrocyte ankyrin, and that phosphorylation of CD44 by exogenously added brain protein kinase C increased this interaction (Kalomiris and Bourguignon, 1989). Camp et al. (1991), however, found a negative correlation between phosphorylation of CD44 and its association with the cytoskeleton in macrophages isolated from the peritoneal cavity of mice. In resident macrophages, the fraction (about half) of CD44 that was associated with the cytoskeleton [1% Nonidet P-40 (NP-40) insoluble] was not phosphorylated, whereas detergent-soluble CD44 in these cells was phosphorylated. Cell activation induced a change in CD44 association with the cytoskeleton: elicited macrophages showed no association of CD44 with the cytoskeleton and, again, detergent-soluble CD44 was phosphorylated. Geppert and Lipsky (1991) also saw a reduction in cytoskeletal association of CD44 after phorbol myristate acetate (PMA) activation of human PBLs. Neame and Isacke (1992) examined the localization of CD44 in the basolateral membrane of polarized epithelial (MDCK) cells. This distribution was dependent on the cytoplasmic domain of CD44 because cells transfected with human CD44 constructs coding for molecules lacking most of cytoplasmic domain showed a diffuse distribution of human CD44, whereas cells transfected with wildtype constructs showed normal localization of the human CD44 to the basolateral membrane. This targeting presumably involves the interaction of the cytoplasmic domain of CD44 with other, intracelluI
CD44 AND ITS INTERACTION WITH ECM
313
lar molecules. Phosphorylation does not play a role in localization, however, because cells transfected with a mutant construct, in which the two serines responsible for in vivo phosphorylation in these cells were mutated to alanine or glycine, showed normal basolateral localization of the human CD44 molecules. The absence ofthe phosphorylation sites also did not affect cytoskeletal association of the mutant CD44 molecules expressed in 3T3 fibroblasts (Neame and Isacke, 1992). Among the studies of CD44 interactions with the cytoskeleton and correlations with phosphorylation of the cytoplasmic domain, none has looked at binding of ECM components by CD44 in relation to these activities. Thus, although the association of CD44 with cytoskeletal elements seems a possible mechanism of regulating its receptor function, the data relating cytoskeletal interactions, phosphorylation, and ECM-receptor function are incomplete and, in some cases, contradictory. The apparent contradictions, could, however, be the result of the use of different experimental systems (i.e., different cell types and cell-free versus intact cell systems). Two reports have defined intracellular molecules that influence the function of integrins, revealing additional mechanisms by which adhesion receptor-ligand interactions might be regulated from inside the cell. Hermanowski-Vosatka et al. (1992) isolated a small lipid from stimulated polymorphonuclear leukocytes that can activate ligand binding by CR3 integrin in unstimulated polymorphonyclear leukocytes and by the purified integrin. Levels of expression of this lipid in polymorphonuclear leukocytes paralleled the state of activation of CR3 function. In another study, Pullman and Bodmer (1992) used adhesion to collagen to select transfectants expressing a cDNA clone coding for a molecule that increased integrin-mediated adhesion to components of the ECM. The encoded protein of 142 amino acids had an N-terminal myristylation motif and a consensus tyrosinephosphorylation site at the C terminus, which was required for the enhancement of adhesion. The authors suggested that this is a signal transduction molecule that contributes to integrin function. These studies indicate potential means by which the function of cell adhesion molecules might be regulated by the intracellular environment. These, or other unknown mechanisms, could modulate the HAbinding function of CD44.
2 . Modijication of Extracellular Domain Among human cell lines, a wide variety of glycosylation patterns of CD44 has been described (see Section 11; e.g., Omary et al., 1988; Brown et al., 1991; Jalkanen et al., 1988), with each cell type seem-
314
JAYNE LESLEY ET AL.
ing to have “decorated” CD44 with a unique array of carbohydrate structures. There is no doubt that these modifications may influence CD44 function and its interactions with the ECM. In considering a possible role for cell type-specific glycosylation patterns in regulating CD44 interactions with the ECM, one must take into account potential effects of the use of alternate CD44 isoforms. In most of the studies of differences in glycosylation patterns, it is not certain which core proteins underlie the diversity of molecular weight patterns observed. The inserted amino acid sequences in the nonstandard (nonhematopoietic) isoforms of CD44 all contain numerous potential glycosylation sites, as do the predicted HA-binding amino-terminal region and the membrane-proximal region of the extracellular domain of “standard” CD44 (see Section 11). To address the functional role of the variant CD44 isoforms, it will be important to determine their glycosylation status in uiuo, which requires knowing the in uiuo cellular distribution. The development of mAbs specific to the inserted domains of the different isoforms (Matzku et d . , 1989; Koopman et aZ., 1993) should soon make this problem amenable to analysis. There is evidence for the transient expression of alternate (nonstandard) CD44 isoforms in hematopoietic cells in response to in uiuo immunological challenge (Arch et al., 1992) and in uitro activation (Koopman et aZ., 1993). Thus, switching isoforms provides another mechanism for regulating CD44-mediated interactions with ECM. Below, we discuss several studies bearing directly on the question of alternate CD44 isoforms and the role of posttranslational carbohydrate additions in CD44 interactions with the ECM. Fassen et uZ. (1992) found that migration of a melanoma cell line into a collagen matrix was dependent on the presence of a chondroitin sulfate-modified cell surface glycoprotein. The approximately 110-kDa core protein, remaining after chondroitinase digestion, reacted with CD44-specific mAbs, suggesting that it is probably a higher molecular mass isoform of CD44. Removal of chondroitin sulfate prevented cell migration into collagen and prevented binding of collagen by the isolated glycoprotein. Jalkanen and Jalkanen (1992) found that only a relatively minor form of chondroitin sulfatemodified standard form CD44 purified from human PBL was able to bind fibronectin. Chondroitinase treatment did not affect PBL binding to HEV in a frozen section assay, but did reduce, slightly, the binding of a B cell line to fibronectin. Thus, although interaction of chondroitin sulfate-modified CD44 with fibronectin may contribute to cell adhesion, it does not appear to be involved in CD44dependent adhesion to HEV.
CD44 AND ITS INTERACTION WITH ECM
315
The ability of different CD44 isoforms to bind HA has been studied by transfection of constructs coding for different isoforms into cell lines (see Section 111,C). Human hematopoietic and epithelial forms of CD44 expressed in a CD44-negative Burkitt lymphoma line differed in their ability to bind HA, with the epithelial form of CD44 failing to mediate adhesion to immobilized HA (Sy et al., 1991) or HA-dependent adhesion to cultured endothelial cells (Stamenkovic et al., 1991). The epithelial form of CD44 expressed in melanoma cells also bound poorly to HA, whereas the hematopoietic form conferred HA-binding function (Thomas et al., 1992). But several high molecular weight isoforms of murine CD44, including that which is homologous to human epithelial CD44, bound HA when expressed in the CD44-negative lymphoma AKRl (He et al., 1992). These clearly conflicting results could result from differences in posttranslational modifications to the inserted isoform sequences mediated by different cellular environments. Another possibility, discussed further below, is that other cell surface molecules, or even soluble cell products, may interact with the external domain of CD44 to modulate its affinity for HA. Such external factors may have specificity for different CD44 isoforms and may differ for different cell types. 3. Regulation by Interactions of the Extracellular Domain of CD44 with Other Molecules The discovery of multiple CD44 isoforms with inserted sequences in the membrane-proximal portion of the extracellular domain (Section I1,C) gives new inpetus to the suggestion that the HA-receptor function of CD44 might be regulated by interaction with other molecules at the cell surface (Lesley et al., 1990a). Alternate sequences in different isoforms could allow interaction of CD44 with different regulatory molecules in different cell types and tissue environments, adding to the possibilities of specific control of an otherwise simple interaction between a single amino-terminal, HA-binding region and a ubiquitous ligand HA. There is, to date, no direct evidence for such heteromolecular interactions involving CD44 and other cell surface molecules. A number of studies have examined the effects of crosslinking cell surface molecules with mAbs. Antibody cross-linking is presumed to mimic the effect of binding of a receptor to its ligand. Thus these types of studies have been used to imply possible influences of the engagement of one receptor-ligand pair on the function of other receptorligand interactions. Rosenman et al. (1993) have observed cocapping of CD2 and of L-selectin with CD44 capped by antibody crosslinking
316
JAYNE LESLEY ET AL.
in human T cells from peripheral blood. Although the observed cocapping is thought to be mediated by the engagement of the cytoskeleton, the proximity of different adhesion receptors induced by cocapping could result in interactions of their external domains. Costimulation of CD44 with cell surface molecules such as CD2 and T cell receptor of human T cells, using combinations of antibodies to crosslink these cell surface molecules, can result in enhancement or inhibition of activation signals (see Table 11). It is not known whether interactions take place between the external domains of these molecules. Nevertheless, the observations indicate that CD44 interactions with other cell surface molecules may be dynamic and may change depending on the array of ligands available. This is consistent with the activatable state of normal T cells in the mouse. In other cell types, however, such as resting B cells, CD44 might be maintained in an inactive state by more stable interactions with a different set of cell surface molecules, which would be altered only by differentiation events. Yet another possibility is that cell surface CD44 function might be influenced by direct interaction of its external domain with soluble extracellular factors. Factors that are secreted in restricted tissue environments and/or in response to specific stimuli, such as inflammation, could be specific ligands for the alternative sequences of different CD44 isoforms. Activation of CD44 HA-binding function by extracellular soluble factors would allow rapid and flexible engagement of CD44-HA-mediated adhesion and mobility, and expression of additional isoforms by stimulated cells could increase the repertoire of factors that might influence CD44 function. CD44 has been shown to contribute to cell mobility on ECM substrates (Fassen et al., 1992; Thomas et al., 1992). A number of motility factors have been shown to influence cell migration of specific leukocyte populations. For example, members of the IL-8 chemotactic factor family induce migration in different subpopulations of neutrophils, monocytes, and lymphocytes, but the receptors for these factors are unknown (Larsen et al., 1989; Schall, 1991). One family member, RANTES, is specific for monocytes and for memory T cells (Schall et al., 1990). Another family member, MIP-1P, was shown to be present on LN endothelium and to influence adhesiveness of specific T cell subpopulations (Tanaka et d., 1993). Although there is no evidence that these factors influence CD44 function, factors of this type might be capable of directly and specifically activating CD44 by binding to sequences in its external domain (Kincade, 1992).
CD44 AND ITS INTERACTION WITH ECM
317
4 . Regulation by Shedding, Masking, and Soluble CD44
Downregulation of CD44 by shedding of the external domain is a potentially rapid mechanism of regulating CD44 expression and, hence, ECM-binding function. Loss of cell surface CD44 by shedding of a protease-cleaved fragment of the cell surface molecule has been reported for human neutrophils and lymphocytes. Campanero et al. (1991) observed downregulation of cell surface CD44 and CD43 in human neutrophils after a 30-minute exposure to TNF, PMA, calcium ionophore, and fMLP (formyl-Met-Leu-Phe). This downregulation was believed to result from proteolytic cleavage, as it was inhibited by protease inhibitors. Baiil and HoiejSi (1992) also observed downregulation of cell surface CD44 (after 12 hours) in granulocytes in response to stimulation with PMA and, to a lesser degree, ionomycin, and in both granulocytes and lymphocytes in response to immobilized or soluble antibody specific for CD44. Again, shedding was believed responsible for the loss of cell surface antigen, in this case because an '251-labeled CD44-specific antibodyreactive species of reduced molecular mass (compared to cell surface CD44) could be isolated from supernatants of '251-surface labeled stimulated cells. A number of other hematopoietic cell surface molecules have been found to be downregulated by shedding in response to external stimuli: TNF receptor (Porteau and Nathan, 1990, CD6 (Huizinga et al., 1988), CD14 (Baiil and Strominger, 1991), Lselectin (Griffin et al., 1990; Kishimoto et al., 1989), ICAM-1 (Becker et al., 1991), CD23 (Guy and Gordon, 1987), and CD32 (Sarmay et al., 1991). A soluble form of CD44 has been found in human serum. Lucas et al. (1989) reported an In-Lu antigen-related species in serum, and Baiil and HoiejSi (1992) found two species, isolated by immunoaffinity chromatography, of approximately 60 to 80 kDa and 100 to 150 kDa. These two species contained different peptides when analyzed after digestion with V8 protease, suggesting that they may be products of different CD44 isoforms. The presence of soluble CD44 in serum suggests that shedding of CD44 may take place in uiuo Baiil and HoiejSi, 1992). Experiments of Camp and colleagues (1993b) demonstrated that shedding of CD44 from the surface of murine lymphocytes was induced by CD44-specific mAb administered in uiuo and resulted in a 1.5- to twofold increase in soluble CD44 in serum. Soluble CD44 may also represent another mechanism of modulating the function of cell surface CD44, through competition for
318
JAYNE LESLEY ET AL.
ligand between soluble and cell-associated forms (Haynes et al., 1989, 1991). The ability of CD44 to bind HA could be modified by blocking of its HA-binding sites with either cell-associated molecules or soluble HA. Asher and Bignami (1992) were able to detect CD44-mediated HA-binding in brain-derived cell lines and primary cells only after treatment with hyaluronidase, indicating that the HA-binding domain of CD44 was occupied by endogenously produced HA. Knudson and Knudson (1991) showed that accumulation of a pericellular coat around some cell types in culture was dependent on the expression of a cell surface receptor for HA. Thus, in some cell types, CD44 may function in the assembly of a protective coat and perhaps to anchor a cell in a particular location, rather than to promote migration. In hematopoietic cells, however, HA-receptor function was not exposed by treatment of cell lines or bone marrow cells with hyaluronidase or chondroitinase ABC (Lesley et al., 1990a). VI. CD44 and Tumor Cell Migration: Possible Role of CD44 in Metastasis
A. METASTASIS AND EXTRACELLULAR MATRIX Tumor metastasis is the end result of a multistep process (Schirrmacher, 1985; Fidler, 1990). As discussed in Fidler (1990) and Blood and Zetter (1990), the primary malignant neoplasm must become vascularized if it is to grow beyond a minimum size (several cubic millimeters). Following vascularization, the tumor cell mass penetrates the basement membrane and invades lymphatics, venules, or capillaries. Cells from this invasive tumor cell mass must then detach and enter the circulation [often forming emboli of small tumor cell aggregates, lymphocytes and platelets (Fidler, 1990; Blood and Zetter, 1990)], be transported through the circulation, and arrest in and adhere to the vessels of the target organ. The adhering tumor cells must then exit from the vasculature, penetrate the subendothelial basement membrane, and enter the extracellular space (Blood and Zetter, 1990). For successful metastasis to occur, the tumor cells must be able to proliferate in the new location-a process that is highly dependent on both the local microenvironment and the ability of the invading tumor cells to respond to this microenvironment (Schirrmacher, 1985; Fidler, 1990). Given the complexity of the metastatic process, there has been considerable debate as to whether successful metastasis is the end result of the selection of a “metastatic phenotype” or whether it is a consequence of the chance survival of a tumor cell through a series of
CD44 AND ITS INTERACTION WITH ECM
319
random traumatic events (Schirrmacher, 1985; Kerbel, 1990; Aslakson and Miller, 1992). There is good evidence that tumor cell populations show heterogeneity in metastatic potential and that tumor cells that show an enhanced ability to metastasize to specific sites preexist in the tumor cell population and can be selected, at least experimentally (Schirrmacher, 1985; Fidler, 1990; Kerbel, 1990). It is not reasonable to conclude from this fact, however, that there is a single “metastatic molecule” whose presence (or absence) can account for metastatic behavior. Independent subpopulations derived from a single tumor may fail to disseminate at different points of the metastatic process, suggesting a marked heterogeneity even within a single tumor (Aslakson and Miller, 1992). Multiple changes in tumor cell phenotype that each lead to an only moderate increase in tumor cell survival at one or another step of the metastatic process may be the difference between success or failure in terms of the ability of a given tumor to metastasize. Successful tumor metastasis might best be thought of as the end result of a sequential process containing both random and selective elements (Schirrmacher, 1985). This complexity means that it is difficult to make overall global generalizations from experimental models. As discussed by Aslakson and Miller (1992), properties measured in uitro that appear to correlate with metastasis may not be the same properties responsible for metastatic behavior in uiuo (for which in uitro assays may not exist). Conversely, properties that are expressed in both metastatic and nonmetastatic cells may be crucially important for metastasis, but the particular nonmetastatic cells examined may all have downstream defects that do not allow them to complete later steps in the metastatic process and thus the importance of the property might not be appreciated. Also, the metastatic behavior of a particular tumor is a reflection both of the properties of the cells of the tumor itself and the organ environment in which these tumor cells are placed (Fidler, 1990).Thus, the behavior of a tumor injected subcutaneously or intraavenously by the experimenter may be different than the behavior of the same tumor in its true in uiuo site. The necessity for metastasizing tumor cells to move through the extracellular spaces and penetrate basement membranes implies that recognition of extracellular matrix components is an essential factor mediating tumor metastasis. Many factors contibute to tumor cell adhesion and motility in the extracellular matrix. Attachment to components of the matrix via a variety of cell surface receptors (Van Roy and Marcel, 1992) may be followed by localized degradation of the extracellular matrix (Liotta, 1986; Blood and Zetter, 1990; Liotta et
320
JAYNE LESLEY ET AL.
al., 1991). The tumor cells then can migrate into the region of local-
ized degradation (Liotta, 1986), using the same or a different set of adhesion receptors as those used for the initial attachment. Hyaluronan has been implicated in tumor cell migration, and some of these migratory events may be relevant to metastasis. Tumor cells may directly secrete HA or may stimulate neighboring fibroblasts to secrete HA (Knudson et al., 1989). This secreted HA is incorporated into the extracellular matrix (Yoneda et al., 1988; Knudson and Knudson, 1991). It may function to separate tissues, affect collagen fibril formation and packing, or affect adhesive forces between the cells and their substrate (Docherty et al., 1989). In model systems in uitro, HA has been reported to enhance fibroblast mobility in monolayers (Turley et al., 1985) and in three-dimensional collagen matrices (Docherty et al., 1989). Hyaluronan is thus likely to be involved in tumor cell migration within extracellular spaces. Presumably, increased HA synthesis could enhance migration and in this way contribute to metastasis. Hyaluronan is also present around most, but not all, basement membranes (Underhill, 1989) and thus may contribute to the movement of tumor cells into and out of the vasculature. There are some data consistent with these ideas. Hyaluronan accumulation measured in collagen matrices has been correlated with tumor progression of preneoplastic mouse mammary cells (Hitzeman et d., 1992), and the peripheral invasive areas of human breast carcinomas showed a higher content of HA than the central noninvasive areas or adjacent normal tissue (Bertrand et al., 1992). Factors produced by human mesothelioma cells stimulate HA production by normal mesothelial cells and fibroblasts (Asplund et al., 1993). It is conceivable that enhanced HA production could act in a positive feedback loop to facilitate tumor cell growth and migration. Similarly, increased expression of CD44 could contribute to adhesion to and mobility on an HA substrate. Thomas et al. (1992) showed in an in uitro assay that transfectants expressing high levels of the hematopoietic form, but not the epithelial form, of CD44 showed mobility on HA-coated surfaces. Degradation of extracellular matrix components (Blood and Zetter, 1990) and of HA in particular (West and Kumar, 1989) has been implicated in angiogenesis. Thus, HA might be important in the establishment of the initial tumor and of metastases at distant sites by inducing the vascularization of tumor cell foci. Degradation products of the HA polymer containing approximately 4-25 disaccharides induce the formation of blood vessels on the chick chorioallantoic membrane (West et al., 1985) and induce the synthesis of new
CD44 AND ITS INTERACTION WITH ECM
321
proteins in cultures of bovine aortic endothelial cells (Kumar et al., 1987). Degradation products of extracellular matrix proteins have also been thought to be chemotactic for tumor cells (Blood and Zetter, 1990). Although there seems to be no evidence directly implicating HA as a causal agent of tumor cell chemotaxis, HA has been shown to enhance the chemotactic response of granulocytes for fibronectin (Hakansson and Venge, 1985). CD44 has been reported to participate in the internalization and degradation of HA (Culty et al., 1992; Sampson et ul., 1992),although the rate of internalization noted in these studies is slow. CD44 has a long half-time on the cell surface (Jacobson et al., 1984), and it is not certain whether CD44 is on the endocytic pathway. CD44 does have a sequence similar to the internalization recognition motifs of constitutively recycling receptors (Trowbridge, 1991). A calciumindependent HA receptor that is thought to be distinct from CD44 and is on the endocytic pathway has been reported in rat liver sinusoidal endothelial cells (Yannariello-Brown et al., 1992a,b). The relative contribution of this receptor, of the HA receptor studied by Turley and colleagues (Hardwick et d., 1992), and of CD44 to HA degradation in tissues that are major sites of HA uptake in uitio (e.g., lymph node; Fraser et al., 1988) or in the vicinity of proliferating tumor cells remains to be determined. As noted above (Section III,D), chondroitin sulfate-substituted forms of CD44 have been reported to bind fibronectin, collagen I, and laminin (Jalkanen and Jalkanen, 1992). Tumor cells expressing this form of CD44 might have the ability to bind to these ECM components, and it is possible that acquisition of this ability may act to enhance tumor cell migration at several steps of the metastatic process. Perhaps most interesting in this regard is the demonstration that melanoma cells bearing a CD44-like chondroitin sulfate proteoglycan show enhanced mobility in collagen gels (Fassen et al., 1992). It is not unreasonable to think that this property could contribute to the success of tumor cell invasion. In certain in tiitro experimental situations, CD44 has been shown to confer an adhesive phenotype on cells (St. John et al., 1990; Belitsos et al., 1990) distinct from the aggregation that occurs when CD44positive cells are exposed to HA. It is not clear to what extent such CD44-mediated adhesion occurs in uiuo or whether it occurs at all. Nevertheless, there is at least the possibility that CD44-CD44 interactions could contribute to tumor cell survival or movement. In this connection, CD44 has been reported on vascular endothelium (Pals et al., 1989a) and umbilical vein endothelial cells (Oppenheimer-
322
JAYNE LESLEY ET AL.
Marks et al., 1990) and a “CD44-like” molecule has been reported on a chemically transformed cell line derived from bovine aorta (Bourguignon et al., 1992). Thus the possibility exists that tumor cells bearing HA or CD44 itself could bind to-CD44 on the walls of blood vessels and that, therefore, CD44 could be of importance in adherence of tumor cells to blood vessel walls and in intravasation or extravasation of tumor cells from the circulation. This is all, however, entirely speculative and there is no experimental evidence.
B. CD44 EXPRESSION AND METASTASIS There are many reports of CD44 expression in human tumor cell lines (Nemec et al., 1987; Picker et al., 1989a; Stamenkovic et al., 1989; Quackenbush et al., 1990; Hoffinann et al., 1991; Dougherty et al., 1991; Kuppner et al., 1992; Jackson et al., 1992; Heider et al., 1993; Koopman et al., 1993). The pattern of expression in cell lines is variable, with some cell lines expressing no CD44, some expressing only the hematopoietic form, and others expressing various higher molecular weight isoforms. It is difficult to discern a clear pattern in these studies. Some of these differences may reflect differences in CD44 expression on the tissues from which the cell lines were derived (Picker et al., 1989a; Heider et al., 1993); however, this does not seem to be the case in every instance (Heider et al., 1993). In long-term cultured cell lines, it is not always certain that the CD44 expression pattern is the same as that of the tumor from which the cell line was derived. CD44 expression on some, but not all, primary tumors may be enhanced over the level generally seen on the corresponding normal cell type. Also, differences in the expression of isoforms may be seen when normal and tumor tissue is compared. Although no direct cause-and-effect relationship can be demonstrated, there are reports that increased levels of CD44 expression in non-Hodgkin’s lymphoma are correlated with increased tumor “aggressiveness.” Pals and colleagues (Pals et al., 198913; Horst et al., 199Oc) and Jalkanen and colleagues (1991) examined CD44 expression in non-Hodgkin’s lymphoma and found a statistically significant relationship between high levels of CD44 expression and clinical stage, tumor spread, and a poor response to treatment. A large proportion, although by no means all, aggressive non-Hodgkin’s lymphomas express an antigenic determinant encoded by exon 10, indicating that these cells express a variant isoform (Koopman et al., 1993). This variant isoform is not expressed at significant levels on resting lymphoctyes, although transient expression has been seen when lymphocytes are
CD44 AND ITS INTERACTION WITH ECM
323
activated (Arch et al., 1992; Koopman et al., 1993). CD44 has also been reported to be present on 95% of tumor cells from childhood acute lymphoblastic and acute myeloblastic leukemia (Kreindler et al., 1990). A series of primary gastrointestinal malignant B cell lymphomas showed a heterogeneous pattern of CD44 expression that did not indicate a role for CD44 in the localization of these tumors (Moller et al., 1991). In tumors of the breast and colon, enhanced expression of CD44 variant isoforms is seen as compared to the corresponding normal tissue (Matsumura and Tarin, 1992; Heider et al., 1993). The significance of these Observations is uncertain and clearly isoform expression does not correlate with metastasis. Variant isoform expression is seen in breast and colon carcinoma with and without metastases (Matsumura and Tarin, 1992; Heider et al., 1993) and most cases of adenomatous polyps examined also show expression of variant CD44 isoforms (Heider et al., 1993). CD44 expression on a series of brain tumors was variable, with high-grade gliomas showing strong expression (Kuppner et al., 1992). As pointed out above (Section 11), it is possible that enhanced CD44 expression and/or isoform expression in tumor cells may reflect the activation state of the cell in which the neoplastic event occurred. Clearly, relating observations on quantitative levels of CD44 expression and the expression of CD44 isoforms to the metastatic behavior of tumors in vivo requires the development of meaningful experimental models (Fidler, 1990; Rettig, 1992). Birch and colleagues (1991) used fluorescence-activated cell sorting to isolate mutants expressing high levels of CD44 from the human melanoma cell line LT1. Stable clones could be isolated that expressed about five- to sevenfold higher levels of the hematopoietic form of CD44 relative to CD44’” clones. Compared to the clones expressing lower levels of CD44, these CD44hi clones showed an enhanced ability to adhere to immobilized HA, enhanced homotypic aggregation, and an increased ability to migrate in a “wounded” monolayer. Although both CD44hi and CD44l” clones showed a tumor incidence approaching 100%when injected subcutaneously, the CD44hi clones showed an increased capacity for lung colonization when injected intravenously (only a single cell dose was examined in these experiments). The mechanism responsible for this enhanced lung colonization is uncertain; however, HA has been shown to b e present in perivascular areas of the lung (Underhill, 1989), whereas CD44 is present in alveolar macrophages (Green et al., 1988; Underhill, 1989) and on alveolar lining epithelium (Picker et al., 1989a). Sy et u1. (1991) examined the rate of tumor growth and metastatic
324
JAYNE LESLEY ET AL.
ability of the human Burkitt lymphoma Namalwa transfected with either the hematopoietic or epithelial forms of human CD44 and injected into nude mice. Transfectants for the hematopoietic form bound HA much more efficiently than transfectants for the epithelial form (see Section 111,C). The transfectants for the hematopoietic form showed an increased rate of tumor development when injected subcutaneously or intravenously and a somewhat greater tendency to metastasize, especially to the bone marrow, than untransfected Namalwa cells or transfectants for the epithelial form of CD44. Although these results do correlate with ability to bind HA, the mechanism responsible for the increased frequency of tumor takes and metastasis is unclear. Treatment of the mice with a soluble CD44-immunoglobulin fusion protein at the time of tumor injection and at intervals thereafter suppressed the incidence of tumors (Sy et al., 1992), suggesting that binding to HA may be involved. The mechanism of this latter inhibition is unclear, however, because the fusion protein disappears from the circulation of these mice within minutes after injection and it seems unlikely that at the doses used all possible binding sites in the animal are saturated. Gunthert and colleagues have used as a model the rat pancreatic adenocarcinoma cell line BSp73 (Gunthert et al., 1991).A metastatic cell line and a nonmetastatic cell line have been established that either do or do not form metastases in the lymph node and the lungs when cells are injected into the footpad and the local tumors are removed 10 days later. The metastatic cell line expresses a unique antigenic determinant, which expression cloning showed was related to the presence of a CD44 isoform containing an alternatively spliced exon 10 (see Fig. 2; Gunthert et al., 1991; Arch et al., 1992; Seiter et al., 1993; Herrlich et al., 1993). Transfection of a CD44 construct expressing this isoform into the nonmetastatic cell line conferred metastatic behavior on it (Gunthert et al., 1991; Seiter et al., 1993). Antibodies specific for the variant isoform inhibited metastatic spread when injected with the tumor, but not when injected later, suggesting that the antibodies interfered with migration from the site of injection in the footpad to the local lymph node (Reber et al., 1990; Seiter et al., 1993). Because the antibody is not directly cytotoxic or cytostatic, at least in uitro, it may act by interfering with interactions between the tumor cell and other cells and/or the ECM, rather than acting to enhance effector immune mechanisms (Seiter et al., 1993). At least in this model system, expression of a particular CD44 isoform does seem to be essential for metastasis. How general this result
CD44 AND ITS INTERACTION WITH ECM
325
will turn out to be and how it relates to the other studies in which higher levels of the hematopoietic form of CD44 correlated with tumorigenicity and/or metastatic ability are uncertain. At least one other series of metastatic cell lines in the rat also expresses this isoform (Gunthert et al., 1991) and transfection of the isoform into a rat fibrosarcoma cell line confers metastatic capability (Herrlich et al., 1993). As discussed above, however, there is not yet enough information on the pattern of isoform expression in human tumors and tumor cell lines to allow generalizations to be made regarding the importance of CD44 isoform expression in metastasis. The expression of these isoforms in nonmetastatic, and even noncancerous, cells indicates that cellular factors other than CD44 isoform expression must contribute to metastasis. How the presence of the extra exon affects ligand recognition by CD44 is unclear as well (Gunthert et al., 1991; Herrlich et al., 1993). Although the alternatively spliced exons found in the nonconserved domain have potential sites of O-glycosylation and chondroitin sulfate addition, it is not clear that cells expressing only exon 10 have any additional posttranslational modifications not found in the standard CD44 molecule (Herrlich et al., 1993). Whether the presence of this exon confers new ligand-binding specificities is unknown. Interestingly, the cytoplasmic domain does not appear to contribute to metastatic behavior, because transfectants of deletion constructs lacking this domain but including exon 10 exhibit metastatic behavior (Herrlich et al., 1993). Although there are clearly many unanswered questions, these studies and the others discussed above have stimulated much interest in the details of CD44-ligand recognition. For a more coherent picture to emerge, it will be necessary to develop methodology that will relate advances in the molecular biology of CD44 to biologically meaningful models of tumor metastasis. VII. Summary
It is now generally accepted that CD44 is a cell adhesion receptor and that hyaluronan is one of its ligands. Like many cell adhesion receptors, CD44 is broadly distributed, and its ligand, hyaluronan, is a common component of extracellular matrices and extracellular fluids. Yet a great variety of responses has been reported to result from CD44 ligation. These include cell adhesion, cell migration, induction (or at least support) of hematopoietic differentiation, effects on other cell adhesion mechanisms, and interaction with cell activation signals.
326
JAYNE LESLEY ET AL.
This diversity of responses indicates that downstream events following ligand binding by CD44 may vary depending on the cell type expressing CD44 and on the environment of that cell. CD44 is expressed on cells in the early stages of hematopoiesis and has been shown to participate in at least some aspects of the hematopoietic process. In mature lymphocytes, CD44 is upregulated in response to antigenic stimuli and may participate in the effector stage of immunological responses. Along with other adhesion receptors that show alterations in expression after activation, CD44 probably contributes to differences in the recirculation patterns of different lymphocyte subpopulations. CD44 ligand-binding function on lymphocytes is strictly regulated, such that most CD44-expressing cells do not constitutively bind ligand. Ligand-binding function may be activated as a result of differentiation, inside-out signaling, and/or extracellular stimuli. This regulation, which in some situations can be rapid and transient, potentially provides exquisite specificity to what would otherwise be a common interaction. CD44 is not a single molecule, but a diverse family of molecules generated by alternate splicing of multiple exons of a single gene and by different posttranslational modifications in different cell types. It is not yet clear how these modifications influence ligand-binding function. The significance of the multiple isoforms of CD44 is not understood, but association of some isoforms with malignancies has been observed. And in at least some experimental systems, a contribution of CD44 isoforms to metastatic behavior has been demonstrated.
ACKNOWLEDGEMENTS We would like to thank all ofour collegues who shared their unpublished manuscripts with us. This work was supported by NIH Grants CA-13287 and AI-31613 and NSF Grant DCB-8900579 to R. Hyman, NIH Grants AI-19884 and AI-20069 to P. W. Kincade, National Cancer Institute Core Grant CA-14195 to the Salk Institute, and by the Hansen Foundation.
REFERENCES Alho, A. M., and Underhill, C. B. (1989).J. Cell B i d . 108,1557-1565. Altieri, D. C., and Edgington, T. S.(1988).J. Zmmunol. 141,2656-2660. Arch, R., Wirth, K., Hofmann, M., Ponta, H., Matzku, S., Herrlich, P., and Zoller, M. (1992). Science 257,682-685. Aruffo, A,, Stamenkovic, I., Melnick, M., Underhill, C. B., and Seed, B. (1990). Cell 61, 1303- 1313. Asher, R., and Bignami, A. (1992). Exp. Cell Res. 203,80-90. Aslakson, C., and Miller, F. (1992). Cancer Res. 52,1399-1405. Asplund, T., Versnel, M., Laurent, T., and Heldin, P. (1993). Cancer Res. 53,388-392.
CD44 AND ITS INTERACTION WITH ECM
327
Barthels, D., Vopper, G., Boned, A., Crerner, H., and Wille, W. (1992).Eur.J.Neurosci. 4, 327-337. BaZil, V., and HoiejSi, V. (1992).]. Zmmunol. 149,747-753. BaZil, V., and Strorninger, J. L. (l99l).J.Zmmunol. 147, 1567-1574. Becker, J. C., Durnrner, R., Hartmann, A. A., Burg, G., and Schmidt, R. E. (1991). J . Zmmunol. 147,4398-4401. Belitsos, P., Hildreth, J., and August, J. (199O).J.Zmmunol. 144, 1661-1670. Bertrand, P., Girard, N., Delpech, B., Duval, C., d’Anjou, J., and Dauce, J. (1992).Znt. J. Cancer 52,l-6. Birch, M., Mitchell, S., and Hart, I. R. (1991).Cancer Res. 51,6660-6667. Blood, C., and Zetter, B. (1990).Biochim. Biophys. Acta 1032,89-118. Bosworth, B. T., St. John, T., Gallatin, W. M., and Harp, J. A. (1991).Mol. Zmmunol. 28, 1131-1 135. Bourguignon, L. Y. W., Walker, G., Suchard, S. J., and Balazovich, K. (1986).J.Cell B i d . 102,2115-2124. Bourguignon, L. Y.W., Kalorniris, E. L., and Lokeshwar, V. B. (1991).J.Biol. Chem.266, 11761- 1 1765. Bourguignon, L. Y. W., Lokeshwar, V. B., He, J., Chen, X., and Bourguignon, G. J. (1992). M o l . Cell. Biol. 12,4464-4471. Brown, T. A., Bouchard, T., St. John, T., Wayner, E., and Carter, W. G. (1991).J.Cell Biol. 113,207-221. Budd, R. C., Cerottini, J.-C., and MacDonald, H. R. (1987a)J Zmmunol. 138,1009-1013. Budd, R. C., Cerottini, J.-C., Horvath, C., Bron, C., Pedrazzini, T., Howe, R. C., and MacDonald, H. R. (1987b).J.Zmmunol. 138,3120-3129. Budd, R. C., Schurnacher, J. H., Winslow, G., and Mosrnann, T. R. (1991). Eur. J . Immunol. 21,1081-1084. Butterfield, K., Fathrnan, C. G., and Budd, R. C. (1989).J.E x p . Med. 169, 1461-1466. Camp, R. L., Kraus, T. A., and Pure, E. (1991).J.Cell Biol. 115,1283-1292. Camp, R., Haspel, R., Kraus, T., Scheynius, A., and Pure, E. (1993a).(Submitted for publication.) Camp, R. L., Scheynius, A,, Johansson, C., and Pure, E. (1993b).(Submitted for publication.) Campanero, M. R., Pulido, R.,Alonso, J. L., Pivel, J. P., Pirnentel-Muinos, F. X., Fresno, M., and Sanchez-Madrid, F. (1991).Eur.J.Zmmunol. 21,3045-3048. Carter, W. G., and Wayner, E. A. (1988).J.Biol. Chem. 263,4193-4201. Chong, A. S.-F., Boussy, I. A., Graf, L. H., and Scuder, P. (1992).Cell. Zmmunol. 144, 69-79. Cianfriglia, M., Yassen,A., Tornbesi, M., Samoggia, P., Barca, S., andCaserta, M. (1991). lnt. J . Cuncer 49,394-397. Coffman, R. (1982).Zmmunologicul Rev. 69,5-23. Colornbatti, A., Hughes, E. N., Taylor, B. A., and August, J. T. (1982).Proc. Natl. Acad. Sci. U.S.A.79, 1926-1929. Cooper, D. L., Dougherty, G., Harn, H.-J., Jackson, S., Baptist, E. W., Byers, J., Datta,A., Phillips, G., and Isola, N. R. (1992).Biochem. Biophys. Res. Commun. 182,569-578. Culty, M., Miyake, K., Kincade, P. W., Sikorski, E., Butcher, E. C., and Underhill, C. B. (199O).J.Cell Biol. 111,2765-2774. Culty, M., Nguyen, H., and Underhill, C. (1992).J.Cell Biol. 116, 1055-1062. Dalchau, R., Kirkley, J., and Fabre, J. W. (1980).Eur.]. Zmmunol. 10,745-749. Davidson, W. F., Durnont, F. J,, Bedigian, H. G., Fowlkes, B. J., and Morse 111, H. C. (1986).J . Zmmunol. 136,4075-4084.
328
JAYNE LESLEY ET AL.
d e 10s Toyos, J,, Jalkanen, S., and Butcher, E. C. (1989). Blood 74,751-760. Denning, S . M., Le, P. T., Singer, K. H., and Haynes, B. F. (199O).J.Zmmunol. 144,7-15. Detmers, P. A., Wright, S. D., Olsen, E., Kimball, B., and Cohn, Z. A. (1987)J. Cell Biol. 105,1137-1145. Dexter, T. M., Allen, T. D., and Lajtha, L. G. (1977).J.Cell Physiol. 91,335-344. Docherty, G., Forrester, J., Lackie, J., and Gregory, P. (1989).J. Cell Sci. 92,263-270. Doege, K., Sasaki, M., Horigan, E., Hassell, J., and Yamada, Y. (1987)J Biol. Chem. 262, 17757- 17767. Doege, K., Sasaki, M., Kimura, T., and Yamada, Y. (1991).J.Biol. Chem. 266,894-902. Dougherty, G. J., Lansdorp, P. M., Cooper, D. L., and Humphries, R. K. (1991).J . Exp. Med. 174, 1-5. Du, X., Plow, E. F., Frelinger 111, A. L., O’Toole, T. E., Loftus, J. C., and Ginsberg, M. H. (1991). Cell 65,409-416. Durbec, P., Gennarini, G., Goridis, C., and Rougon, G. (1992)J. Cell Biol. 117,877-887. Dustin, M. L., and Springer, T. A. (1989).Nature (London)341,619-624. Dustin, M. L., and Springer, T. A. (1991).Annu. Reo. Immunol. 9,27-66. Fassen, A. E., Schrager, J. A., Klein, D. J., Oegema, T. R., Couchman, J. R., and McCarthy, J. B. (1992).J.Cell Biol. 116,521-531. Fernandez-Botran, R. (1991). FASEBJ. 5,2567-2574. Fidler, I. (1990). Cancer Res. 50,6130-6138. Figdor, C. G., van Kyook, Y.,and Keizerg, G. D. (1990). Immunol. Today 11,277-280. Flanagan, B. F., Dalchau, R., Allen, A. K.,Daar, A. S., and Fabre, J. W. (1989). ZmmunolOgy 67,167-175. Franke, U., Foellmer, B., and Haynes, B. F. (1983). Somatic Cell Genet. 9,333-344. Fraser, J., Kimpton, W., Laurent, T., Cahill, R., and Vakadis, N. (1988). Biochem. J . 256, 153- 158. Gallatin, W. M., Weissman, I. L., and Butcher, E. C. (1983).Nature (London)304,30-34. Gallatin, W. M., Wayner, E. A., Hoffman, P. A., St. John, T.,Butcher, E. C., and Carter, W. G. (1989). Proc. Natl. Acad. Sci. U.S.A. 86,4654-4658. Gallatin, W. M., Rosenman, S. J., Ganji, A., and St. John, T. P. (1991). “Cellular and Molecular Mechanisms of Inflammation: Vascular Adhesion Molecules” (C. G. Cochrane and M. A. Gimbrone, eds.), pp. 131-150. Academic Press, San Diego, California. Geppert, T.D., and Lipsky, P. E. (1991).J.Zmmunol. 146,3298-3305. Girgrah, N., Ackerley, C. A., and Moscarello, M. A. (1991a). Neuroreport 2,441-444. Girgrah, N., Letarte, M., Becker, L. E., CNZ,T. F., Theriault, E., and Moscarello, M. A. (1991b).J . Neuropathol. E x p . Neurol. 50,779-792. Goetinck, P., Stirpe, N., Tsonis, P., and Carlone, D. (1987).J.Cell Biol. 105,2403-2408. Goldberg, R., Seidman, J., Chi-Russo, G., and Toole, B. (1984). J . Biol. Chem. 259, 9440-9446. Goldstein, L. A,, and Butcher, E. C. (1990). Zmmunogenetics 32,389-397. Goldstein, L. A., Zhou, D. F. H., Picker, L. J., Minty, C. N., Bargatze, R. F., Ding, J. F., and Butcher, E. C. (1989). Cell 56,1063-1072. Goodfellow, P. N., Banting, G., Wiles, M. V., Tunnacliffe, A., Parker, M., Solomon, E., Dalchau, R., and Fabre, J. W. (1982). Eur.J. Zmmunol. 12,659-663. Graham, I. L., and Brown, E. J. (1991).]. Zmmunol. 146,685-691. Green, S., Tarone, G., and Underhill, C. (1988). E x p . Cell Res. 178,224-232. Griffin, J. D., Spertini, O., Ernt, T. J., Belvin, M. P., Levine, H. B., Kanakura, Y., and Tedder, T. (199O).J.Zmmunol. 145,576-584. Gruber, M. F., Webb, D. S., and Gerrard, T. L. (1992).J.Immunol. 148, 1113-1118.
CD44 AND ITS INTERACTION WITH ECM
329
Gunthert, U., Hofmann, M., Rudy, W., Reber, S., Zoller, M., Haubmann, I., Matzku, S., Wenzel, A., Ponta, H., and Herrlich, P. (1991). Cell (Cambridge,Mass.) 65, 13-24. Guy, G. R.,and Gordon, J. (1987). Proc. Natl. Acad. Sci. U.S.A.84,6239-6243. Haegel, H., and Ceredig, R. (1991).Eur.1. Zmmunol. 21, 1549-1553. Haggerty, J. G., Bretton, R. H., and Milstone, L. M. (1992).J . Znoest. Dermatol. 99, 372-380. Hakansson, L., and Venge, P. (1985).J . Zmmunnol. 135,2735-2739. Hale, L. P., and Haynes, B. F. (1989). F A S E B J . 3, A482. Hale, L. P., and Haynes, B. F. (1992).J . Zmmunol. 149,3809-3816. Hamann, A., Jablonski-Westrich, D., Duijvestijn, D., Butcher, E. C., Baisch, H., Harder, R.,and Thiele, H. G. (1988).J.Zmmunol. 140,693-699. Hamann, A., Jablonski-Westrich, D., Jonas, P., and Thiele, H.-G. (1991). Eur.J.Zmmunol. 21,2925-2929. Hardwick, C., Hoare, K., Owens, R.,Kohn, H., Hook, M., Moore, P., Cripps, V., Austen, L., Nance, D., and Turley, E. (1992).J.Cell Biol. 117,1343-1350. Harn, H.-J., Isola, N., and Cooper, D. L. (1991). Biochem. Biophys. Res. Commun. 178, 1127-1 134. Haynes, B. F., Telen, M. J., Hale, L. P., and Denning, S . M. (1989). Zmmunol. Today 10, 423-428. Haynes, B. F., Hale, L. P., Patton, K. L., Martin, M. E., and McCallum, R. M. (1991). Arthritis Rheum. 34,1434-1443. He, Q., Lesley, J., Hyman, R., Ishihara, K., and Kincade, P. W. (1992).J.Cell Biol. 119, 1711-1719. Heider, K.-H., Hofmann, M., Hors, E., van den Berg, F., Ponta, H., Herrlich, P., and Pals, S. (1993).J . Cell Biol. 120,227-233. Hermanowski-Vosatka, A., Van Strijp, J. A. G., Swiggard, W. J., and Wright, S. P. (1992). Cell (Cambridge,Mass.) 68,341-352. Herrlich, P., Rudy, W., Hofmann, M., Arch, R., Zoller, M., Zawadski, V., Tolg, C., Hekele, A., Pals, S., Heider, K.-H., Sleeman, J., and Ponta, H. (1993). In “Cell Adhesion Molecules.” Edigref, Rome (in press). Hibbs, M. L., Xu, H., Stacker, S. A., and Springer, T. A. (1991).Science 251,1611-1613. Hiro, D., Ito, A., Matsuta, K., and Mori, Y. (1986).Biochem. Biophys. Res. Commun. 140, 715-722. Hitzeman, J., Woost, P., and Hosich, H. (1992). In Vitro Cell Dew. Biol. 29A, 284-292. Hofmann, M., Rudy, W., Zoller, M., Tolg, C., Ponta, H., Herrlich, P., and Gunthert, U. (1991).Cancer Res. 51,5292-5297. Horst, E., Meijer, C. J., Radaskiewicz, T., van Dongen, J. J., Pieters, R., Figdor, C., Hooftman, A., and Pals, S . T. (1990a). Leukemia 4,383-389. Horst, E., Moyer, C. J,, Duijvestijn, A. M., Hartwig, N., Van der Harten, H., and Pals, S. T. (1990b). Eur.1. Zmmunol. 20,1483-1489. Horst, E., Meijer, C. J., Radaskiewicz, T., Ossekoppele, G., Van Krieken, J., and Pals, S . Leukemia 4,595-599. (1990~). Huet, S., Groux, H., Caillou, B., Valentin, H., Prieur, A.-M., and Bernard, A. (1989). J . Zmmunol. 143,798-801. Hughes, E. N., and August, J. T. (1981).J.Biol. Chem. 256,664-671. Hughes, E. N., Colombatti, A., and August, J. (1983).J.Biol. Chem. 258,1014-1021. Huizinga, T. W. J., van der Schoot, C. E., Jost, C., Klassen, R.,Kleijer, M., von den Borne, A. E., Roos, D., and Tetteroo, P. A. T. (1988).Nature (London)333,667-669. Husmann, L., Shimonkevitz, R.,Crispe, N., and Bevan, M. (1988). J . Zmmunol. 141, 736-740.
330
JAYNE LESLEY ET AL
Hyman, R.,Lesley, J., and Schulte, R. (1991). Zmmunogenetics 33,392-395. Hynes, R.0.(1992). Cell 69, 11-25. Idzerda, R. L., Carter, W. G., Nottenburg, C., Wayner, E. A., Gallatin, W. M., and St. John, T. (1989).Proc. Natl. Acad. Sci. U.S.A.86,4659-4663. Imai, Y.,Trug, D. D., Singer, M., and Rosen, S. D. (199O).J.Cell Biol. 111, 1225-1232. Isacke, C. M., Sauvage, C. A., Hyman, R., Lesley, J., Schulte, R.,and Trowbridge, I. S. (1986). Zmmunogenetics 23,326-332. Jackson, D. G., Buckley, J., and Bell, J. I. (1992).J . B i d . Chem. 267,4732-4739. Jacobson, K., O’Dell, D., Holifield, B., Murphy, T. L., and August, T. (1984).J.Cell B i d . 99,1613-1623. Jalkanen, S., and Jalkanen, M. (1992).J.Cell. B i d . 116,817-825. Jalkanen, S. T., Bargatze, R. F., Herron, L. R.,and Butcher, E. C. (1986).Eur.J.Zmmunol. 16,1195-1202. Jalkanen, S . T., Bargatze, R. F., de 10s Toyos, J., and Butcher, E. C. (1987).J.Cell B i d . 105,983-990. Jalkanen, S., Jalkanen, M., Bargatze, R.,Tammi, M., and Butcher, E. C. (1988).J.Zmmunol. 141,1615-1623. Jalkanen, S., Joensuu, H., Soderstrom, K.-O., and Klemi, P. (1991).J . Clin. Znoest. 87, 1835-1840. Kalomiris, E. L., and Bourguignon, L. Y. W. (1988).J. Cell B i d . 106,319-327. Kalomiris, E. L., and Bourguignon, L. Y. W. (1989).J.Biol. Chem. 264,8113-8119. Kansas, G . S., and Dailey, M. 0. (1989).J.Zmmunol. 142,3058-3062. Kansas, G . S., Wood, G. S., and Dailey, M. 0. (1989).J . Zmmunol. 142,3050-3057. Kansas, G . S., Muirhead, M. J., and Dailey, M. 0. (1990). Blood 76,2483-2492. Keizer, G . D., Visser, W., Vliem, M., and Figdor, C. (1988).J.Zmmunol. 140,1393-1400. Kerbel, R. (1990).Ado. Cancer Res. 55,87-132. Kincade, P. W. (1992).Nature (London)361,15-16. Kishimoto, T., Jutila, M., Berg, E., and Butcher, E. (1989).J.Zmmunol. 143,3318-3324. Knudson, W., and Knudson, C. (1991).J . Cell Sci. 99,227-235. Knudson, W., Biswar, C., Li, X.-Q., Nemec, R.,and Toole, B. (1989). I n “The Biology of Hyaluronan, CIBA Foundation Symposium 143,” pp. 150-169, Wiley, Chichester, England. Koopman, G., van Kooyk, Y., de G r d , M., Meyer, C. J.. Figdor, C. G., and Pals, S. T. (1990).J . Zmmunol. 145,3589-3593. Koopman, G., Heider, K.-H., Horst, E., Adolf, G. R., van den Berg, F., Ponta, H., Herrlich, P., and Pals, S. T. (1993).J.Exp. Med. 177,897-904. Kriendler, D., Petsche, D., Hrincu, A., Gouges, A,, Quackenbush, E., Freedman, F., Gelfand, E., and Letarte, M. (1990). Leukemia Lymphoma 3,7-18. Kugelman, L. C., Ganguly, S., Haggerty, J. G., Weissman, S. M., and Milstone, L. M. (1992).J . Znoest. Dennatol. 99,381-385. Kumar, S., West, D., and Ager, A. (1987). Dif-et-entiation 36,57-70. Kuppner, M., Van Meir, E., Gauthier, T., Hamon, M.-F., and de Tribolet, N. (1992).Znt.1. Cancer 50,572-577. Lacy, B. E., and Underhill, C. B. (1987).J . Cell B i d . 105,1395-1404. Larsen, C. G., Anderson, A. 0..Appella, E., Oppenheim, J. J., and Matsushima, K. (1989). Science 243,1464-1466. Lasky, L. A., Singer, M. S., Yednock, T. A., Dowbenko, D., Fennie, C., Rodriguez, H., Nguyen, T., Stachel, S., and Rosen, S. D. (1989). Cell 58, 1045-1055. Laurent, T. (1989). I n “The Biology of Hyaluronan, CIBA Foundation Symposium 143,” pp. 1-5. Wiley, Chichester, England.
CD44 AND ITS INTERACTION WITH ECM
33 1
Laurent, T., and Fraser, J. (1992).FASEBJ. 6,2397-2404. Le Baron, R., Zimmerman, D., and Ruoslahti, E. (1992). J . Biol. Chem. 267, 1000310010. Lee, W. T., and Vitetta, E. S. (1991).Cell. Immunol. 132,215-222. Lee, T., Wisniewski, H.-G., and Vilcek, J. (1992).J. Cell Biol. 116,545-557. Lerner, A., Yamada, T., and Miller, R. A. (1989).Eur. J . Immunol. 19,977-982. Lesley, J., and Hyman, R. (1992).Eur. J . Immunol. 22,2719-2723. Lesley, J., and Trowbridge, I. S. (1982).Immunogenetics 15,313-320. Lesley, J., Hyman, R., and Schulte, R. (1985a).Cell. lmmunol. 91,397-403. Lesley, J.. Trotter, J., and Hyman, R. (1985b). lmmunogenetics 22, 149-157. Lesley, J., Schulte, R., Trotter, J., and Hyman, R. (1988). Cell. Immunol. 112,40-54. Lesley, J., Schulte, R., and Hyman, R. (1990a).Exp. Cell Res. 187,224-233. Lesley, J., Trotter, J., Schulte, R., and Hyman, R. (1990b). Cell. lmmunol. 128, 6378. Lesley, J., He, Q., Miyake, K., Hamann, A., Hyman, R., and Kincade, P. W. ( 1 9 9 2 )Exp. ~ Med. 175,257-266. Lesley, J., Kincade, P. W., and Hyman, R. (1993). (Submitted for publication.) Lewinsohn, D. M., Nagler, A., Ginzton, N., Greenberg, P., and Butcher, E. C. (1990). Blood 75,589-595. Linsenmayer, T. (1991).I n “Cell Biology ofthe Extracellular Matrix” (E. Hay, ed.), 2nd Ed., pp. 7-44. Plenum, New York. Liotta, L. (1986).Cancer Res. 46,l-7. Liotta, L., Steeg, P., and Statler-Stevenson, W. (1991). Cell (Cambridge, Mass.) 64, 327-336. Lokeshwar, V. B., and Bourguignon, L. Y.(1991).J. Biol. Chem. 266, 17983-17989. Lokeshwar, V. B., and Bourguignon, L. Y.(1992).J. Biol. Chem. 267,22073-22078. Lucas, M. G., Green, A. M., and Telen, M. J. (1989).Blood 73,596-600. Lynch, F., and Ceredig, R. (1989).Eur.1. Immunol. 19,223-229. Lynch, F., Chaudri, G., Allan, J. E., Doherty, P. C., and Ceredig, R. (1987). Eur. J . lmmunol. 17,137-140. Mackay, C. R., Maddox, J. F., Wijffels, G. L., MacKay, I. R., and Walker, I. D. (1988). lmmunology 65,93-99. MacKay, C. R., Marston, W. L., and Dudler, L. (1990).]. Exp. Med. 171,801-817. Matsumura, Y.,and Tarin, D. (1992).Lancet 340, 1053-1058. Matzku, S., Wenzel, A., Liu, S., and Zoller, M. (1989).Cancer Res. 49, 1294-1299. Miyake, K., and Kincade, P. (1990).Curr. Topics Microbiol. lmmunol. 166,87-90. Miyake, K., Medina, K. L., Hayashi, S.-I., Ono, S., Hamaoka, T., and Kincade, P. (1990a). J . Exp. Med. 171,477-488. Miyake, K., Underhill, C., Lesley, J., and Kincade, P. (1990b).J. Exp. Med. 172,69-75. Miyake, K., Medina, K., Ishihara, K., Kimoto, M., Auerbach, R., and Kincade, P. W. (1991).J. Cell Biol. 114,557-565. Mobley, J. L., and Dailey, M. 0. (1992).J. lmmunol. 148,2348-2356. Moller, P., Eichelmann, A., Mechtersheimer, G., and Koretz, K. (1991).1nt.J. Cancer44, 846-855. Morgelin, M., Paulsson, M., Hardingham, T., Heinegard, D., and Engel, J. (1988). Biochem.J . 253,175-185. Morgelin, M., Paulsson, M., Malmstrom, A., and Heinegard, D. (1989).J . Biol. Chem. 264,12080-12090. Murakami, S., Miyake, K., June, C. H., Kincade, P. W., and Hodes, R. J. (1990). J . lmmunol. 145,3618-3627.
332
JAYNE LESLEY ET AL.
Murakami, S., Miyake, K., Abe, R.,Kincade, P. W., and Hodes, R. J. (1991).J.Zmmunol. 146,1422-1427. Neame, P., Christner, J., and Baker, J. (1986).J.Biol. Chem. 261,3519-3535. Neame, P., Christner, J., and Baker, J. (1987).J.B i d . Chem. 262, 17768-17778. Neame, S. J,, and Isacke, C. M. (1992). EMBOJ. 11,4733-4738. Nemec, R., Toole, B., and Knudson, W. (1987). Biochem. Biophys. Res. Commun. 149, 249-257. Noble, P. W., Lake, F. R., Henson, P. M., and Riches, D. W. (1993). (Submitted for publication.) Nottenburg, C., Rees, G., and St. John, T. (1989). Proc. Natl. Acad. Sci. U.S.A. 86, 8521-8525. Omary, M. B., Trowbridge, I. S., Letarte, M., Kagnoff, M. F., and Isacke, C. M. (1988). Zmmunogenetics 27,460-464. Oppenheimer-Marks, N., Davis, L. S., and Lipsky, P. E. (1990).J . Zmmunol.145, 140148. O’Toole,T. E., Loftus, J. C., Du, X.,Class, A. A., Ruggeri, Z. M., Shattil, S. J., Plow, E. F., and Ginsberg, M. H. (1990). Cell Reg. 1,883-893. O’Toole, T. E., Mandelman, D., Forsyth, J., Shattil, S. J., Plow, E. F., Ginsberg, M. H. (1991). Science 254,845-847. Pals, S. T., Hogemorst, F., Keizer, G. D., Thepen, T., Horst, E., and Figdor, C. C. (1989a). 1.Zmmunol. 143,851-857. Pals, S., Horst, E., Ossekoppele, G., Figdor, C., Scheper, R., and Meijer, C. (1989b). Blood 73,885-888. Penit, C., and Vasseur, F. (1989).J. Zmmunol. 142,3369-3377. Perin, J.-P., Bonnet, F., Thurieau, C., and Jolles, P. (1987).J . Biol. Chem. 262, 1326913272. Petrie, H., Hugo, P., Scollay, R.,and Shortman, K. (1990).J.Erp. Med. 172, 1583-1588. Pierres, A., Lipcey, C., Mawas, C., and Olive, D. (1992). Eur.J.Zmmunol. 22,413-417. Picker, L. J., Nakache, M., and Butcher, E. C. (1989a).J. Cell Biol. 109,927-937. Picker, L. J., de Los Toyos, J., Telen, M. J., Haynes, B. F., and Butcher, E. C. (1989b). J . Zmmunol. 142,2046-2051. Pober, J. S.,and Cotran, R. S. (1991).Ado. Zmmunol. 51,261-302. Porteau, F., and Nathan, C. (1990).J.Erp. Med. 172,599-607. Pullman, W. E., and Bodmer, W. F. (1992). Nature (London)356,529-532. Quackenbush, E. J., Vera, S., Greaves, A,, and Letarte, M. (1990). Mol. Zmmunol. 27, 947-955. Reber, S., Matzku, S., Gunthert, U., Ponta, H., Herrlich, P., and Zoller, M. (1990).Znt.J. Cancer 46,919-927. Rettig. W. (1992). Curr. Opinion Zmmunol. 4,630-640. Rodrigues, M., Nussenzweig, R. S., Romero, P., and Zavala, F. (1992).J.E x p . Med. 175, 895-905. Rosenman, S. J., Dubb, L., and Callatin, M. (1993).J.Leuk. B i d . 53, 1-10. Rothman, B. L., Blue, M.-L., Kelley, K. A., Wunderlich, D., Mien, D. V.,a nd Aune, T. M. (1991).J . Zmmunol. 147,2493-2499. Sampson, P., Rochester, C., Frendlich, B., and Elias, J. (1992). J . Clin. Invest. 90, 1492- 1503. Sanders, M. E., Makgoba, M. W., Sharrow, S. O., Stephany, D., Springer, T. A., Young, H. A., and Shaw, S. (1988).J.Zmmunol. 40,1401-1407.
CD44 AND ITS INTERACTION WITH ECM
333
Sandmaier, B. M., Storb, R., Appelbaum, F. R., and Gallatin, W. M . (1990). Blood 76, 630-635. Sarmay, G., Rozsnyay, Z., Szabo, I., Biro, A., and Gergely, J. (1991).Eur.J.lmmunol. 21, 541-549. Schall, T. J. (1991). Cytokine 3, 165-183. Schall, T. J., Bacon, K., Toy, K. J., and Goeddel, D. U. (1990). Nature (London) 347, 669-671. Schirrmacher, V. (1985).Ado. Cancer Res. 43, 1-73. Schubert, D., (1992). Trends Cell Biol. 2,63-66. Scollay, R. (1991). Curr. Opinion lmmunol. 3,204-209. Scott, J. (1989).In “The Biology of Hyaluronan, CIBA Foundation Symposium 143,” pp. 6-20. Wiley, Chichester, England. Scott, J. (1992). FASEBJ. 6,2639-2645. Screaton, G. R., Bell, M. V., Jackson, D. G., Cornelis, F. B., Gerth, U., and Bell, J. I. (1992). Proc. Natl. Acad. Sci. U.S.A.89, 12160-12164. Seiter, S., Arch, R., Reber, S., Komitowski, D., Hofmann, M., Ponta, H., Herrlich, P., Matzku, S., and Zoller, M. (1993).J.Exp. Med. 177,443-455. Seth, A., Cote, L., Nagarkatti, M.,and Nagarkatti, P. S. (1991). Proc. Natl. Acad. Sci. U.S.A.88,7877-7881. Shattil, S. J., and Brass, L. F. (1987).J.Biol. Chem. 262,992-1000. Shattil, S. J.. Hoxie, J. A., Cunningham, M.C., and Brass, L. F. (1985).J.Biol. Chem.260, 11107-11114. Shimizu, Y., Van Seventer, G. A., Siraganian, R., Wahl, L., and Shaw, S. (1989). J . Zmmunol.143,2457-2463. Shtivelman, E., and Bishop, J. M. (1991). Mol. Cell. Biol. 11,5446-5453. Spangrude, G., and Scollay, R. (1990).J . lmmunol. 145,3661-3668. Spangrude, G., Klein, J., Heimfeld, S., Aihara, Y., and Weissman, 1. (1989).J.Zmmunol. 142,425-430. Spertini, O., Kansas, G. S., Munro, J. M., Griffin, J. D., and Tedder, T. F. (1991). Nature (London)349,691-694. Spring, F. A., Dalchau, R., Daniels, G. L., Mallinson, G., Judson, P. A., Parsons, S. F., Fabre, J. W., and Anstee, D. J. (1988). Immunology 64,37-43. St. Jacques, S., Dadi, H. K., and Letarte, M. (1993). Placenta 14 (in press). St. John, T., Meyer, J., Idzerda, R., and Goldstein, W. M. (1990).Cell (Cambridge,Mass.) 60,45-52. Stamenkovic, I., Amiot, M., Pesando, J. M., and Seed, B. (1989). Cell (Cambridge,Mass.) 56, 1057-1062. Stamenkovic, I., Aruffo, A., Amiot, M., and Seed, B. (1991). EMBOJ. 10,343-348. Stamper, H. B., and Woodruff, J. J. (1976).J.Exp. Med. 144,828-833. Stefanovi, I., Hilgert, l., Batil, V., KriStovii, H., and HotejSi, V. (1989). Zmmunogenetics 29,402-404. Stout, R. D., and Suttles, J. (1992). Cell. lmmunol. 141,433-443. Sutherland, D. R., Abdullah, K. M., Cyopick, P., and Mellors, A. (1992).1.Zmmunol. 148, 1458-1464. Swain, S. L., Weinberg, A. D., and English, M. (1990).J.lmmunol. 144,1788-1799. Swain, S. L., Bradley, L. M., Croft, M., Tonkonogy, S., Atkins, G., Weinberg, A. D., Duncan, D. D., Hedrick, S. M., Dutton, R. W.,and Houston, G. (1991). Zmmunol. Reo. 123.115-144.
334
JAYNE LESLEY ET AL.
Sy, MA., Guo, Y.-J., and Stamenkovic, I. (1991).J. E x p . Med. 174,859-866. Sy, M.-S., Guo, Y.-J., and Stamenkovic, I. (1992).J. E x p . Med. 176,623-627. Tan, P. H., Santos, E. B., Rossbach, H.-C., and Sandmaier, B. M. (1993)J Immunol. 150, 1-9. Tanaka, Y., Adams, D. H., Hubscher, S., Hirano, H., Siebenlist, U., and Shaw, S. (1993). Nature (London) 361,79-82. Tarone, G., Ferracini, R., Galetto, G., and Comoglio, P. (1984)J.Cell B i d . 99,512-519. Telen, M. J., Eisenbarth, G. S., and Haynes, B. F. (1983).J. Clin. Invest. 71,1878-1886. Thomas, L., Byers, H. R., Vink, J., and Stamenkovic, I. (1992).].Cell B i d . 118,971-977. Toole, B. (1991).In“Cel1 Biology ofthe Extracellular Matrix” (E. Hay, ed.), pp. 305-341. Plenum, New York. Trowbridge, I. (1991).Curr. Opinion Cell B i d . 3,634-641. Trowbridge, I. S.,Lesley, J., Schulte, R., Hyman, R., and Trotter, J. (1982).lmmunogenetics 15,299-312. Turley, E., Bowman, P., and Kytryk, M. (1985).J. Cell Sci. 78,133-145. Uhlig, H., Rebstock, S., Lesley, J., Jablonski-Westrich, D., and Hamann, A. (1993). (Submitted for publication.) Underhill, C. (1989).In “The Biology of Hyaluronan, CIBA Foundation Symposium 143,”pp. 87-106.Wiley, Chichester, England. Cell Biol. 82,475-484. Underhill, C., and Toole, B. (1979).J. Underhill, C., and Toole, B. (1982).].Cell. Physiol. 110,123-128. Underhill, C., Chi-Rosso, G., and Toole, B. (1983).J. Biol. Chem. 258,8086-8091. Underhill, C., Thurn, A,, and Lacy, B. (1985).J. B i d . Chem. 260,8128-8133. Underhill, C., Green, S., Comoglio, P., and Tarone, G. (1987).J . Biol. Chem. 262, 13142-13146. van Kooyk, Y., van de Wiel-van Kemenade, P., Weder, P., Kuijpers, T. W., and Figdor, C. G. (1989).Nature (London) 342,811-813. Van Roy, F., and Marcel, M. (1992).Trends Cell Biol. 2,163-169. Vella, A. T., Hulsebosch, M. D., and Pearce, E. J. (1992).J . lmmunol. 149,1714-1722. Vogel, H.. Butcher, E. C., and Picker, L. J. (1992).].Neurocytol. 21,363-373. Wang, C., Tammi, M., and Tammi, R. (1992).Histochemistry 98,105-112. Wayner, E.,and Carter, W. (1987).J. Cell Biol. 105,1873-1884. Webb, D. S., Shimizu, Y., Van Seventer, G. A., Shaw, S., and Gerrard, T. I. (1990). Science 249,1295-1297. West, D., and Kumar, S. (1989).In “The Biology of Hyaluronan, CIBA Foundation Symposium 143,”pp. 187-207.Wiley, Chichester, England. West, D., Hampson, I., Arnold, F., and Kumar, S. (1985).Science 228, 1324-1326. Whitlock, C. A., and Witte, 0. N. (1982).Proc. Natl. Acad. Sci. U.S.A.79,3608-3612. Wight, T., Heinegard, D., and Hascall, V. (1991).In “Cell Biology of the Extracellular Matrix” (E. Hay, ed.), 2nd Ed., pp. 45-78.Plenum, New York. Willerford, D. M., Hoffman, P. A., and Gallatin, W. M. (1989).J . Immunol. 142,34163422. Wolffe, E. J., Cause, W. C., Pelfrey, C. M., Holland, S. M., Steinberg, A. D., and August, J. T. (1990).].Biol. Chem. 265,341-347. Woodruff, J. J., Clark, L. M., and Chin, Y. H. (1987).Annu. Reo. Immunol. 5,201-222. Wright, S. D., and Silverstein, S. C. (1982).].E x p . Med. 158,2016-2023. Wu, L.,Scollay, R., Egerton, M., Pearse, M., Spangrude, G., and Shortman, G. (1991a). Nature (London)349,71-74. Wu, L., Antica, M., Johnson, G., Scollay, R., and Shortman, K. (1991b).].E x p . Med. 174, 1617-1627.
CD44 AND ITS INTERACTION WITH ECM
335
Yannariello-Brown, J., Frost, S., and Weigel, P. (1992a).J . Biol. Chem. 267, 2045120456. Yannariello-Brown,J . , McCary, C., and Weigel, P. (1992b).J.Cell. Biochem. 4 8 , 7 3 4 0 . Yoneda, M., Shimizu, S., Nishi, Y.,Yamagata, M., Suzuki, S., and Kimata, K. (1988).J.Cell Sci. 90,275-286. Zhou, D. F. H., Ding, J. F., Picker, L. J., Bargatze, R. F., Butcher, E. C., and Goeddel, D. V. (1989).J . Zmmunol. 143,3390-3395. This article was accepted for publication on 3 March 1993.
ADVANCES IN IMMUNOLOGY. VOL. 54
Immunoglobulin Receptor-Associated Molecules NOBUO SAKAGUCHI, TATSUYA MATSUO, JUN NOMURA, KAZUHIKO KUWAHARA, HIDEYA IGARASHI, AND SElJl INUl Depariment of Immunology,School of Lib Science, Faculty of Medicine, Totton. University, Yonago 683, Japan
1. introduction
The immune system requires antigen-specific responses to protect our body from invading microorganisms or pathogenic antigens in a manner that does not elicit nonspecific responses against the body’s own organs and tissues. At the end of the 19th century, the German scientist, Paul Ehrlich, proposed an exciting model called the “side chain theory” (Ehrlich, 1900), in which an antibody itself plays the acceptor function specific to the antigen on the surface of B lymphocytes. Kitazato and Boehring had just discovered the antibody as a specific serum component that recognizes and neutralizes the antigen molecule. It had already been suggested that antigenic stimulation induces the activation of antigen-specific B lymphocytes and this concept was further extended in “the clonal selection theory” by Burnet (1959). Extensive studies dealing with the antibody molecule, one of the most important functional molecules in the immune system, have clarified the structure and expression of immunoglobulin molecules composed of tetramer polypeptide components, heavy and light chains. Studies of the organization and rearrangement of immunoglobulin genes in the development and maturation of B lymphocytes demonstrated the molecular mechanism of immunoglobulin variable (V) region diversity, which is the source of antibody specificity (Tonegawa, 1983).The antibody repertoire created by gene rearrangement and somatic mutation covers enormous antibody specificity ( lo9 to 10’’). It is expected that the same immunoglobulin receptor (IgR) repertoire is expressed on the surface of B lymphocytes. During development and throughout life, to maintain a healthy condition in the conventional environment with massive exposure to pathogenic antigens, the immune system has to expand a specific portion of the antibody repertoire and has to restrain the response so that the threshold necessary to maintain the repertoire is not exceeded. Until the 1970s, the molecular basis of antigen-specific activation in B lymphocytes remained unclear. Normal B lymphocytes are composed of many kinds of individual antigen-specific B cells, and this 337 Copyright 0 1993 by Academic Press, Inc. All rights of reproduction in any form reserved.
338
NOBUO SAKAGUCHI ET AL.
heterogeneous population hampered study of the B cell activation mechanism. Early studies, however, demonstrated that antiimmunoglobulin (anti-Ig) stimulation induces the polyclonal activation of most surface immunoglobulin receptor-positive (sIgR') B cells with IgR capping and patching on B cells (Taylor et al., 1971).This IgR crosslinking leads to internalization of the antigen/IgR complex in cells, which suggests that the cytoskeletal/IgR interaction is part of the process of B cell activation. Unfortunately, this suggestive evidence has been accepted for decades, although the intracellular portion of IgR has a short taillike structure with only 3 to 28 amino acids. It has been difficult to understand or to study the further downstream events of the intracellular signal transduction mechanism in different types of cells. The T cell receptor (TCR), whose structure appeared only in 1982 (Allison et al., 1982; Meuer et al., 1983; Haskins et al., 1983), became a target of extensive studies of transport, expression, and antigen-specific signal transdiiction. TCR associates with such T cellspecific surface molecules as CD3 y, 6, E , (, and r) chains to express and mediate the antigen-specific signals given as antigen on class I or class I1 molecules (Borst et al., 1983; Patty et al., 1987; Samelson et al., 1985; Weiss et al., 1986). In B cells, however, it took a long time to demonstrate the IgR-associated molecules using monoclonal antibodies (mAbs) or molecular cloning. Here, we review the current knowledge concerning the molecules associated with IgR and advances in the study of the B cell activation mechanism. At present, it is necessary to present our current understanding of the molecular mechanism of IgR-mediated signal transduction in B cells and also it is important to identify what is still unclear. II. B Cell Differentiation and the Expression of Immunoglobulin Receptor
B cells are considered to originate from the multipotent hematopoietic precursor cells in the fetal liver or in the bone marrow of adult mice (Osmond and Everett, 1964; Osmond and Nossal, 1974; Melchers et al., 1975; Raff et al., 1975a; Owen et al., 1975; Osmond, 1985; Kincade, 1987; Muller-Sieburg et al., 1986; Hardy et al., 1991). The developmental stages of B lineage cells are characterized with respect not only to expression of IgR and the immunoglobulin gene configurations but also to expression of B cell-specific surface antigens (Kincade et al., 1981; Bruce et al., 1981;Coffman and Weissman, 1981; Coffman, 1982; McKearn et al., 1984; McKeam and Rosenberg, 1985; Strasser, 1988). The earliest identifiable B lineage precursor cells are the so-called pro-B cells, which are characterized as surface IgR negative but DJ
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
339
rearranged in either one of the immunoglobulin heavy-chain genes. The next developmental stage of precursor B cells is the pre-B cell, characterized classically by the expression of cytoplasmic p chain without sIgM expression (Raff et al., 1975a; Rosenberg et al., 1975; Burrows et al., 1978; Alt et al., 1981; Paige et al., 1978; Sakaguchi et al., 1980). The immunoglobulin gene configuration is VDJ rearranged as in the IgH gene, but either germline or no transcription of IgL mRNA occurs even with the rearranged IgL gene. In the early development of B lineage cells, there is no evidence of antigen-specific stimulatory factors to the IgR; therefore, their development is considered antigen independent (Cooper, 1978). These precursor B cells are considered IgR-negative cells whose growth or activation is dependent on soluble lymphokines or direct cell contact with stromal cells in the fetal liver or bone marrow. The growth conditions and the requirement for growth factors of these early precursor cells are well characterized by Whitlock/Witte-type bone marrow culture of pre-B cells (MullerSieburg et al., 1986; Witte et al., 1987). B lineage cells require direct cell contact with the stromal cell layer and also require soluble growth or differentiation factors, either IL-3, IL5, or IL-7 (Kincade et al., 1989; Dorshkind, 1990; Rolink and Melchers, 1991; Namen et al., 1988; McNiece et al., 1991; Nishikawa et al., 1988; Hayashi et al., 1990; Rolink et al., 1991; Inui and Sakaguchi, 1992). A light chain-like gene named A5 was found as a pre-B cell-specific cDNA clone by the subtractive hybridization method (Sakaguchi et al., 1986; Sakaguchi and Melchers, 1986). A5 mRNA is expressed in the restricted stage of early pro-B and pre-B cells in mice but is not expressed in mature sIgR-positive B cells with the rearrangement of IgL chain genes (Sakaguchi et aZ., 1986; Kudo et al., 1992). Putative A5 gene product has a light chain-like structure with highly conserved JA and CA segments, but does not contain the V region-like portion (Sakaguchi et al., 1986;Kudo et al., 1987a,b).The gene is not rearranged in B cell development but is expressed specifically at the restricted stage of pro-B and pre-B cells in both mouse and human (Sakaguchi et al., 1986; Bauer et al., 1988; Melchers et al., 1993; Kudo et al., 1992; Takai et al., 1992; Schiff et al., 1989, 1990). The 5' part of the putative A 5 gene product is quite different from the VA sequence, and in a search 5' upstream of the A5 genomic segment, Kudo et aZ. found two DNA segments (VpreB1 and VpreB2) that are homologous to the V region segment (Melchers et al., 1993). Both VpreB and A 5 transcripts are expressed in early B lineage cells (Sakaguchi et al., 1986; Kudo et al., 1987a; Kudo and Melchers, 1987; Rolink and Melchers, 1991; Melchers et d., 1993). The proteins encoded by VpreB (16 kDa) and A 5
340
NOBUO SAKAGUCHI ET AL.
(22 kDa) are expressed on the surface of DNA transfectants (Karasuyama et al., 1990; Tsubata and Reth, 1990). These two proteins probably correspond to two IgL-like proteins named iota (L) and omega (w) that coprecipitate with p chain or DHJHCH protein (64kDa) (Pillai and Baltimore, 1987, 1988; Herayil and Pillai, 1991; Misener et al., 1990; Tsubata et al., 1991).A pseudo-light-chain and p-chain complex ( p chain/VpreB/A5 complex) structure is shown to be expressed on the surface of B lineage precursor cells as late pre-B cells (Nishimoto et al., 1991; Ishihara et al., 1991). Deletion of the membrane domain of IgR induced the arrest of pre-B cell development, which suggested the presence of the regulatory function of the complete p chain protein in early B cell development (Kitamura et al., 1991; Kitamura and Rajewsky, 1991). A5 knockout mice obtained by the homologous recombination method lacked the normal B lineage cell differentiation but have an abnormal rearrangement of B cell clones (Kitamura et al., 1992), which suggests the important functional role of A5/VpreB expression in early B cell differentiation (Gu et al., 1991; Tsubata et al., 1992; Misener et al., 1991). The molecules associated with this surrogate IgR complex on the surface of early B cells are now under extensive study (Takemori et al., 1990; Chen et al., 1990; Iglesias et al., 1991, Ishihara et al., 1992a,b; Matsuo et al., 1993; Karasuyama et al., personal communication; Nunez et al., personal communication), which is probably important to demonstrate the function of this pre-B cell IgR. Matsuo et al. demonstrated the contribution of a molecule (MB-1) associated with mature B cell IgR. Functional molecules composed possibly of scr-type tyrosine kinases are also considered to be involved in the signal transduction of pro-B and pre-B cells expressing surrogate light c h a i d p chain complex (Matsuo et al., 1993). A model of B cell development is shown in Fig. 1.During maturation of B lineage cells, the expression, structure, and function of IgR differ with the differentiation stage. In the early differentiation of the bone marrow environment, pro-B cells and pre-B cells have to undergo rearrangement of both IgH and IgL chain genes to store the functional repertoire of the antibody. At the immature B cell stage, IgR mediates the inhibitory antigenic signals for the proliferation of B cells (Raff et al., 1975b; Metcalf and Klinman, 1976; Isakson and Vitteta, 1980; Cooper et al., 1980;Nossal, 1983; Teale and Klinman, 1980; DeFranco et al., 1982; Scott et al., 1987). Mature B cells expressing surface IgM+/IgD+,which are present in peripheral lymphoid organs of adult mice, mediate the antigenic signals as the initiator for proliferation and differentiation into antibody-secreting plasma cells after several steps (Parker, 1975; Kishimoto and Ishizaka, 1975; Siekman et al., 1978;
34 1
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
6 cell dMferentiatimpathway fetal liver or bone marrow
,
peripheral lymphoid organ antibody secreting cell
stem cell
Ig gene rearrangement
1118
unctionsoflgR
=
b
B29 expression
1.11.
/ / negative stimulation for the proliferation
positive stirnulation for the proliferation and differentiation
FIG.1. Model of B cell differentiation, B lineage cells originated from stem cells in the bone marrow differentiate into pro-B cells (DHJH-rearranged in the IgH gene) and then into pre-B cells with complete rearrangement of the IgH gene. These early B cells express p/VpreB/X5 surrogate IgR complex on the cell surface, which is possibly used as the receptor for stimulation of early B cell differentiation. The IgR-positive B cells, which migrate to the peripheral organ, will receive the antigen-dependent stimulation to the IgR on B cells. The mb-1 and B29 transcripts are expressed in the wide spectrum ofB cell differentiation pathways; however, these two genes and the gene products (MB-1 and B29) are detected with slight differences in the various developmental stages of B lineage cells. Possible functions of the signals through each IgR-related structure are different depending on the maturational stage of B cells.
Julius, 1987). Antigen stimulation triggers several early biochemical changes in B lineage cells, as discussed later; however, the final effect of the IgR-mediated signals requires a longer interval in uiuo or longer culture in uitro. B lineage cells have different responses to IgRmediated signals, especially depending on the B cell stage. Therefore, it is necessary to study the structure and functions of IgR using monoclonal or homogeneous B lineage cells and to compare the differences among the various stages of B lineage cells.
\)p
342
NOBUO SAKACUCHI
ET
AL.
111. Membrane and Secreted Immunoglobulin Molecules
B cells undergo a transition from expression of membrane-bound IgM to IgM secretion during maturation and activation (Melchers and Cone, 1973). The secreted immunoglobulin heavy chain and membrane-bound receptor immunoglobulin heavy chain are structurally different (Melchers and Uhr, 1973). Membrane and secreted p chains are distinct polypeptide species, with separate biosynthetic intermediates from the translation stage. The p polypeptides of membrane and secretory IgM are initially translated with a cleavable leader peptide, 73 kDa in size, containing the glycosylated side chains. These two distinct chains obtain additional glycosylated side chains: 11 kDa for membrane p and 13 kDa for secretory p during maturation of the p chain. These sizes correlate with numbers of C region asparaginelinked carbohydrate units in membrane p (4) and secretory p (5), respectively (Kehry et al., 1980). Finally, the 73-kDa membrane p is converted to the 78-kDa membrane form with further carbohydrate addition. Secretory p chain has further intracellular intermediates of 70 and 76 kDa which are detected as the mixed heterogeneous forms of 76 kDa on average. Resting B cells express surface p (78 kDa) and an intracellular precursor p (73 kDa), but do not produce detectable levels of secretory p or the later intermediate forms of secretory p chain (Sidman, 1981). Major bands of p chain-specific RNA on total RNA Northern blot analysis are 2.4 kb, presumably encoding p, polypeptides, and 2.7 and 3 kb, encoding pm polypeptides, both of which are synthesized from a single p chain gene by selecting alternate RNA processing pathways (Early et al., 1979,1980; Nishikura and Vuocolo, 1984; Rogers and Wall, 1984; Mather et al., 1984; Alberini et al., 1990). By in vitro translation, 2.7- and 2.4-kb RNA species in the B cell lines correspond to 67- and 64-kDa polypeptide proteins (Alt et al., 1980). In plasmacytoma, selective expression of pspolypeptide by the differential splicing and expression of J chain (Koshland, 1975) is not the only reason that plasmacytoma does not express membrane IgR. DNA transfection of membrane form IgM did not induce the membrane IgR in plasmacytoma. Failure of the expression of membrane IgR by the DNA transfection of membrane form, rearranged IgM genes clearly suggested that some unknown mechanism, which is necessary to transport IgM molecules to the Golgi apparatus and to express them on the surface, is missing in the fully differentiated plasmacytoma cells, (Hombach et al., 1988b). Membrane domain plays important functions for IgR in capping,
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECIJLES
343
endocytosis, expression, and IgR-mediated signal transduction (Sitia et al., 1990; Shaw et al., 1990; Parikh et al., 1992; Williams et al., 1990). The hybrid IgR molecule whose membrane domain is replaced by that of MHC class I (Dubois et al., 1992) or class I1 (Webb et al., 1989)was introduced by DNA transfection but failed to respond in the capping, endocytosis, and Ca2+ mobilization response. Shaw et al. (1990) mutated the membrane domain of p chain as tyrosine-serine residues to valine-valine dipeptide, which resulted in the loss of Ca2+ mobilization response. Antigen uptake and presentation activity of B cells are preserved by the amino acid residue (587) with hydroxyl group, and loss of this group causes marked reduction of antigen presentation to T cell clones. The Cys-575 in the tailpiece of secretory-type IgM molecules is involved in the retention of IgR in the plasma membrane (Sitia et al., 1990).The mutation of Cys-575 to Ala causes the hypersecretion of monomeric IgM molecules in the plasmacytoma cells. The carboxyl two-thirds of the transmembrane domain is necessary for the induction of Ca2+ mobilization response, whereas the complete transmembrane domain is required for antigen presentation (Parikh et al., 1992). IV. lsotypes of Immunoglobulin Receptor
Most peripheral B cells express two distinct Ig isotypes (IgM and IgD) simultaneously, and both IgR isotypes are capable of binding the same antigen or anti-idiotypic ligands (Fu et al., 1975; Knapp et al., 1981). These two receptors are translated from the differentially spliced RNA molecules coding for membrane IgM and IgD from the identical rearranged immunoglobulin heavy-chain gene segments. Despite extensive studies on their functions or roles in activation and differentiation of mature B cells, transformed B cell lines, and malignant tumor cells, they have remained somewhat of an enigma. In cells coexpressing IgM and IgD, the p-6 locus downstream ofVDJ is identical to that in germline cells (or nonlymphoid tissues). The C8 encoding elements are located 2.5 to 12 kb downstream of the C, elements that are apparently part of a complex transcriptional unit that can simultaneously generate p mRNA and 6 mRNA components by alternative RNA processing (Liu et al., 1991). Qualitative and quantitative changes in p-8 expression occur in the development of B cells. Earlystage I3 cell lymphomas express the transcription of the entire 25 kb of the p-6 locus and the relative contents of p and 6 mRNA are determined at the level of mRNA processing. In contrast, IgM-secreting cells express the transcription that is terminated between the p and 6
344
NOBUO SAKAGUCHI ET AL.
genes, precluding the production of 6 mRNA (Perry, 1981).The accumulated evidence indicates that sIgD has functions similar to those of sIgM with respect to the activation and proliferation of mature B cells (Siekmann, 1978; Cambier and Ransom, 1987; Harnett et al., 1989). There are, however, reports that suggest that functional differences exist between sIgM and sIgD (Tisch et al., 1988; Ales-Martinez et al., 1990). sIgD appears on the surface in mature peripheral B cells, and sIgM+/sIgD+ B cells are more resistant to tolerance (Waldschmidt and Vitetta, 1985; Scott et al., 1987; Cambier et al., 1978; Mongini et al., 1989).Self-reactive IgM receptor was lost on the surface of doubletransgenic mice with antilysozyme and lysozyme genes. These B cells expressed only IgD molecules with the same antigen binding activity but they did not respond to the antigenic stimulation as observed in the IgM class receptor. Changes in the IgM/IgD receptor ratio induced in the double-transgenic mice also support the notion of the differential role of IgD and IgM in B cell tolerance (Goodnow et al., 1988; Adams et al., 1990). Surface IgD is expressed at only a low level on sIgM+ B cells in the spleen marginal zone (Gray et al., 1982)and on Ly-l/CD5+ B cells (Herzenberg et al., 1986). Brink et al. (1992) compared the activation, deletion, or anergy of antilysozyme B cells through the signals from sIgM or sIgD receptors in the transgenic mice, but they could not detect the obvious difference in these classes of receptors. Monoclonal B cell lines or lymphoma cells expressing both classes of receptors are used simultaneously to study these differential roles of sIgM and sIgD. A murine B cell lymphoma, ECH408-1, expressing idiotypically and allotypically distinguishable transfected IgM and endogenous IgD receptor, showed a clear difference in response to isotype specific stimulation. Anti-IgM induces growth inhibition but anti-IgD does not, although both kinds of stimulation induce calcium mobilization (Ales-Martinez et al., 1990). Iwabuchi et al. (1992) constructed the IgR expression vector of 6 / p mand p/&, consisting of exons encoding extracellular 6 and p domains and membrane regions of different isotypes and introduced them into WEHI-231 cells. They analyzed the IgR-mediated growth inhibition of WEHI-231 cells and observed the inability through 6 / p m construct, which suggests the possible functional importance of an additional extracellular region of p chain in this kind of response (Iwabuchi et al., 1992). Similarly, a human B lymphoma cell line, B104, expressed both receptors on the surface. A panel of anti-IgM antibodies inhibited the growth of B104 cells, but anti-IgD antibodies could not do so at all (Kim et al., 1991).
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
345
V. Functions of Immunoglobulin Receptor and Associated Molecules
A. ACTIVATIONOF B CELLSFOR PROLIFERATION AND DIFFERENTIATION Antigenic stimulation or IgR crosslinking induces B cells to proliferate and differentiate into antibody-secreting cells with the help of T cells or soluble mediators (Kishimoto, 1985). Crosslinking of IgR with anti-IgM antibody is the efficient stimulatory signal as the polyclonal activator of individual B cells. Antigenic stimulation of B cells by thymus-dependent and -independent antigens induces specific plasma membrane depolarization on a large proportion of antigenbinding cells within 2 hours of stimulation (Monroe and Cambier, 1983a,b). But only thymus-independent antigens induces the subsequent Go to GI transition (blastogenesis), which suggests that the membrane IgR crosslinking signal is sufficient to induce membrane depolarization and increased class I1 antigen expression, but is insufficient to drive B cells into the cell cycle. Maino et al. (1975) first discovered that signals through IgR stimulate the phosphoinositide signaling pathway by measuring phosphatidylinositol. Subsequent results also demonstrated the activation of phosphatidylinositol metabolism in IgR-mediated signal transduction (Kenny et al., 1979; Chien and Aschman, 1983; Coggeshall and Cambier, 1984; Grupp and Harmony, 1985; Kritz et al., 1986). Antigenic stimulation with various forms of TNP-carrier protein induced similar changes in phospholipidinositol metabolism in purified TNP-binding B cells (Myers et al., 1987; Grupp et al., 1987). In earlier studies of IgR-associated protein components, Choi et al. (1983)used a selective radioiodination method of IgR on the surface of chicken B cells with antigen-lactoperoxidase (LPO) (Choi et aE., 1983). Human gamma globulin, antigen bound to surface IgR of immunized chicken B cells and the conjugated LPO, readily catalyzed the radioiodination of IgR complex. They detected the coprecipitated protein as 55 kDa, with most of the radiolabeled protein IgM molecules (Choi et al., 1983).In 1976, Marchalonis proposed a model for the IgR-associated complex to explain the specific triggering of B cells by an analogy with certain peptide hormones whose receptor consisted of a recognition element, a regulator element, and an effector element. Haustein and Tshammer tried to identify the associated molecules with IgM and IgD receptors expressed on nonstimulated and stimulated B cells (Haustein and Tsammer, 1987; Haustein and van der Ahe, 1986). Two-dimensional sodium dodecyl sulfate-polyacrylamide gel
346
NOBUO S A U G U C H I ET AL.
electrophoresis (unreduced in the first dimension, reduced in the second dimension) of surface iodinated B cell lysates showed that various polypeptides (56, 50, 46, 42, 35, and 30 kDa) could be identified as being covalently linked to IgM and IgD half-molecules. Chemical crosslinking treatment of intact B cells further demonstrated additional 56- and 46-kDa polypeptides as noncovalently linked to IgM halfmolecules. Sanders et al. (1987) identified a 60-kDa IgM-binding protein that is expressed on the surface of activated B cells. The 60-kDa protein induced by stimulation of PMA or by crosslinkage with antiIgR antibody could bind secreted and membrane-bound IgM molecules. But the function and the physiological role of all these components are still undetermined.
B. CYTOSKELETAL INTERACTIONS The cytoskeletal apparatus forms an internal lattice that pervades the cytoplasmic networks and connects the plasma membrane with the nucleus. The cytoplasmic matrix composed of microfilaments, microtubules, and intermediate filaments is the possible structural support for the regulation of the mobility, localization, and expression of membrane IgR on B cells (Petrini et al., 1983; Albrecht and Noelle, 1988; Albrecht et al., 1990). IgR-mediated stimulation induces actin polymerization through signaling pathways coupled to GTP-binding proteins (Melamed et al., 1991). One of the candidate molecules is a lymphoma membrane-associated 41-kDa protein that shares a number of structural and functional similarities with the a1 subunit of the GTP-binding protein (Bourguignon et al., 1990). This 41-kDa protein is closely associated with cytoskeletal proteins of actin, myosin, and fodrin which colocalize under an IgR capped site. This association requires the activation of phospholipase C (PLC). Stimulation of B cells with anti-IgM antibody also induces the phosphorylation of a lymphocyte myosin light chain of 22 kDa (Fechheimer and Cebra, 1982). Myosin light chain-specific kinase and phosphatase are identified, and both may contribute to the regulation of movements performed by lymphocytes that are necessary for immunological reactions. Melamed et al. (1991)demonstrated that crosslinking the IgR on human B cells induced the conversion of G-actin to F-actin, with increased tyrosine phosphorylation resulting in actin polymerization. Actin-tyrosine phosphorylation by the stimulation of IgR also occurs in chicken bursa1 B cells (Rosenspire and Choi, 1988). In murine B cells, anti-IgR antibody does not induce actin polymerization, which is induced by cytochalasin stimulation (Wilder and Ashman, 1991).
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
347
OF ASSEMBLED IgR STRUCTURE C. TRANSPORTATION RETAININGFUNCTIONS
There is no certain evidence of the requirement for the molecules involved in the transportation of IgR molecules from a site of protein synthesis, such as the endoplasmic reticulum (ER) or Golgi apparatus, to the plasma membrane. The simple transportation of IgR to the cell surface does not indicate the complete integrity of IgR functions as the signal transducing unit of the receptor-mediated signals. In contrast to membrane IgM, secretory IgM is rapidly degraded in B cells under a post-ER/pre-trans-Golgi proteolytic pathway (Amitay et al., 1992). On treatment with brefeldin A, degradation of IgM in a 38C B lymphoma cell was strongly inhibited, as was seen in IgM-secreting hybridoma cells. Amitay et al. proposed that B cells avoid secretion by active and selective targeting of secretory IgM to a developmentally regulated post-ER degradation pathway in which degradation is mediatdd by a cysteine protease. Bip is a 78-kDa ER protein which was identified in association with the free nonsecreted Ig heavy chain synthesized in B lineage cells, including Abelson virus-transformed pre-B cell lines and plasmacytoma cells (Haas and Wabl, 1983). Bip binds to the CH1 domain of heavy chain protein in the cytoplasma and was considered a necessary molecule to transport proteins (Hendershot et al., 1987; Kerr et al., 1989). Bip also binds to Ig light chain and probably stabilizes the Bip-Ig light chain complex (Kuittler and Haas, 1992). Later studies showed that Bip is identical to a 78-kDa glucose-regulated protein whose physiological functions are not yet determined (Hendershot et al., 1988, Munro and Pelham, 1986). Both Bip and the glucoseregulating protein have the identical N-terminal amino acid sequence: N -(X)-(X)-(X)-D -K- K- (X)-D -V- (X)-V-V- (X)-I -D -L -(X)-TT-Y-(X)-(X)-V-. Bip is expressed in both fibroblasts and lymphoid cells as several forms are detected by phosphorylation on serine and threonine residues or by labeling with [3H]adenosine (Hendershot et al., 1988). The phosphorylated Bip on serine and threonine residues binds to the ER, whereas heavy chain-binding Bip is not phosphorylated or adenosine labeled (Hendershot et al., 1988; Freiden et al., 1992). Although Bip is one of the candidate components in transport of the IgR complex structure onto the surface of B cells, it will be necessary to identify a new component that promotes the transport or maintains the assembly of the functional receptor complex.
348
NOBUO SAKAGUCHI ET AL.
D. ENDOCYTOSIS AND PROCESSING OF MEMBRANE IgR/ANTIGEN COMPLEX Another important function of membrane (IgR) (mIgR) is the internalization of the mIgR/antigen complex after antigen binding to the receptor on the surface of antigen-specific B cells. This process is probably important in the effective condensation of specific antigen, transportation of processed peptide antigenic determinant onto class I1 molecules, and presentation of antigen on class I1 to specific T cells (Abbas et al., 1985).Several reports suggest that mIgR is endocytosed in the absence of specific antigen (Antoine and Avrameas, 1974; Davidson et d . ,1990) and that about 60% of the recycling pool of membrane Ig is in the cells (Davidson et al., 1990). Pure and Tardelli (1991) demonstrated the inhibition of ligand-induced internalization of IgR on B cells with the protein tyrosine kinase inhibitors. Tyrosine residues in cytoplasmic domains of endocytosed receptors are potentially key residues for receptor internalization and the fate of endocytosed receptors in a model system where receptor recycling and ligand processing take place. It is still not determined whether IgR/antigen complex either simply dissociates and is processed without further utilization as the IgR, or internalized antigen and IgR are rapidly separated and non-antigen-bound IgR is recycled onto cell surface. Moller et al. argued about the function of IgR on B cells in receptor signal transmission. They considered that IgR does not transmit any signals at all, but rather its role is to bind antigen and focus activation signals to T cells by presentation of the processed antigen through IgR-mediated internalization (Moller et al., 1991). Contribution of IgR for specific antigen presentation on B cells is observed under conditions where surface IgR binds and internizes antigen for subsequent processing (Casten and Pierce, 1988). B cells effectively stimulate specific T cells when provided with 1/1000th the concentration of cytochrome c covalently coupled to anti-IgR antibodies of both p and 6 isotypes (Pierce et al., 1988). VI. B Cell-Specific Proteins Involved in the Immunoglobulin Receptor Complex Identified by Molecular cDNA Cloning
A. B CELL-SPECIFIC GENESINVOLVED IN THE IgR COMPLEX Evidence is accumulating that suggests that functional molecules are required to act as the signal transducer for receptor-mediated signal transduction and transporter molecules are required to express IgR on the B cell plasma membrane (Choi e t al., 1983; Haustein and
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECIJLES
349
Tsammer, 1987; Haustein and van der Ahe, 1986; Sanders et al., 1987; Deans et al., 1984; Nishikura and Vuocolo, 1984; Sitia et al., 1990; Hombach et al., 1988a; Parkhouse, 1990).The protein chemistry could not prove with certainty the existence of functional molecules, because in specific immunoprecipitations there are often high background proteins under usual conditions. To overcome these difficulties, molecular cloning of the genes expressed as B cell-specific protein is a useful approach. Several B cell lineage-specific cDNA libraries were constructed by the subtractive hybridization method (Sakaguchi et aZ., 1986; Sakaguchi and Melchers, 1986, 1987; Yancopolous et al., 1990; Hermanson et al., 1988). A pre-B-minus T cell (702/3-K62 T cell hybridoma) cDNA library was successful in cloning B cell-specific genes with stage-restricted expression. A pool of 200 individual B cellexpressed clones were further screened by the cloning strategy to focus on the one encoding a membrane molecule expressed as B cell specific and also specific in the restricted stage of B cells from pre-B to mature surface IgM + B cells. One of the clones, named mb-1, was first identified as a B cell-specific cDNA clone than can encode the 25 to 35-kDa membrane glycoprotein on the surface of B lineage cells from sIgM- pro-B cell to mature sIg+ B cells. The deduced mb-1 gene product (MB-1) has an immunoglobulin-like domain in the extracellular portion with 116 amino acids and contains 4 cysteines and 2 N-linked glycosylation sites. In the membrane domain there are an attachment site for 0-linked sugar and a consensus structure with CD3like function (Sakaguchi et al., 1988). In the cytoplasmic portion, 61 amino acids, there are two phosphorylatable tyrosine residues, one of which is quite similar to the muscle phosphorylase b tyrosine phosphorylation sites. Hombach et al. (1990b) and Campbell and Cambier (1990) demonstrated an IgR-associated heterodimer component of 34 kDa and 39 kDa by diagonal sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis in the presence of digitonin buffer. Campbell and Cambier (1990)demonstrated that these proteins are induced to phosphorylate by stimulation with aluminum fluoride. The 34-kDa protein, named IgM-a, is expected as the mb-1 gene product, but the 39-kDa component, named Ig-p, is not known at that time. Reth and colleagues compared the cytoplasmic sequences of the B cell-specific mb-1 and B29 genes with the cytoplasmic portions of CD3 molecules y, 6, and E . They found an interesting motif (D/ ExxxxxxxD/ExxYxxLxxxxxxxYxxL/I)which consists of two negatively charged amino acids and two tyrosines followed, by a leucine or an isoleucine (Reth, 1989, 1992; Reth et al., 1991). The motif is considered an important structure for the transport or expression of anti-
350
NOBUO SAKAGUCHI ET AL.
gen receptor molecules on the cell surface that are not being expressed without these functional molecules. Amino-terminal peptide sequencing of the IgR-associated heterodimer components (Ig-a/Ig-P) showed complete identity in nine amino acids (Hombach et al., 1990a; Campbell et al., 1991b) between Ig-a and the deduced amino acid sequence of the mb-1 gene product, and also in 12 amino acids (Hombach et al., 1990a) and 7 amino acids (Campbell et al., 1991b) between Ig-p and the B29 gene product. The IgD-associated a component was initially considered to be different from IgM-a for several reasons. Because of its slightly larger mass, IgD-a migrates differently in NEPHEGE/SDS-PAGE analysis with a slight difference in pl. J5586m2.6 transfectant expressing only mIgD did not express mb-1 transcript, which strongly suggests that IgD-a is very similar in function to but is not encoded b y the mb-1 gene (Hombach et al., 1988a). Wienand and co-workers demonstrated that IgDassociated components consisting of two proteins of 35 kDa (IgD-a) and 39 kDa (Ig-p) partly differ from IgM-associated components, which is quite interesting to account for the different functions of different classes of IgR in B cell activation (Wienand et al., 1990; Reth et al., 1991). Thus, many investigators searched for the genes coding the IgD-associated proteins. And some studies also dealt with IgG receptor-associated components in B cells of autoimmune disease mice. But in a later study, they appeared to be identical in proteolytic cleavage, in Western blot analysis against anti-MB-1 antibody, and in N-glycanase treatment (Campbell et al., 1991a). Surface IgD expression on MB-1-negative cells was through glycosyl-phosphatidylinosito1 linker (Wienand and Reth, 1991, 1992). The p component of IgR-associated molecules has been characterized. The B29 cDNA clone had been reported in 1988, almost at the same time as the mb-1 gene (Hermanson et al., 1988).But it was just a B cell-specific gene that may encode a protein expressed on the surface of B cells. Hombach et al. (1990a) and Campbell et al. (1991b) performed the microsequence analysis of the Ig-p component and found that the N-terminal sequence is identical to the deduced amino acid sequence of B29 cDNA. Ishihara et al. (1992a) finally prepared an antibody specific for the B29 gene product using fusion protein with TrpE and detected the 36- to 47-kDa protein covalently associated with 32- to 34-kDa protein considered as the MB-1 protein. Surface expression of B29 protein correlates exactly with the expression of membrane IgR on the surface of B cells by flow-cytometric analysis, which contrasts markedly with MB-1 protein expressed before the appearance of IgR at the pre-B cell stage (Nomura et al., 1991).
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
351
Lin and Justment (1992) demonstrated the association of' multiple src-type protein kinases with MB-1 and B29 proteins. A model IgR complex including kinase molecules is shown in Fig. 2. Tables I and I1 summarize IgR-associated components that may be involved in IgRmediated signal transduction. B. SIGNALTRANSDUCTION THROUGH THE MB-1 MOLECULE To characterize the functions of the MB-1 molecule, Matsuo et al. (1991) prepared anti-MB-1 antibody by immunizing synthetic oligopeptides corresponding to the extracellular portion of the deduced MB-1 protein. Using the anti-MB-1 antibody, the predominant 34-kDa protein is the monomer MB-1, but it is also present as the heterodimer
FIG.2. Putative structure ofthe IgR complex on mature B cells that is connected with several functional molecules. An Ig-a (34 kDa)lIg-p (39 kDa) heterodimer component, encoded by mb-l and B29 genes, respectively, is associated with IgR. The physical association is shown according to the results in the text, which were mostly obtained with digitonin buffer as detergent.
TABLE I IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES Name
Protein
Synonym
A. Membrane IgR-Associated Molecules mb-1 1gM-a 34kDa
B29
8
Reference Sakaguchi et at., 1988; Hombach et at., 19l38a. 1990b Wienand et al., 1990; Campbell and Cambier. 1990; Campbell et at., 1991a; Veokitaraman et at., 1991 Hermanson et al., 1988;Hombach et al.. 199oa; Campbell et al., 199lb
+ +
Probable Probable
+(tyrosine) +(tyrosine)
b-6
39 kDa
+
+
Probable
+(mine)
78kDa
+
+
Undetermined
+
Haas and Wabl. 1983;Munrn and Pelham, 1986; Hendershot et 01.. 1988
22 kDa
+ (pre-B cell specific)
+
Possible
-
+ (pre-B cell specific)
+
Possible
-
Sakaguchi and Melchers, 1986; Pillai and Baltimore, 1987; W Karasuyama et al., 1990; Tsubata and Reth, 1 Kudo and Melchers, 1987; Pillai and Baltimore, 1W; IlaraFuyama et at., 1990; Tsubata and Retb, I990
+ + + + +
Probable Probable Probable Probable Probable Probable Probable Probable
+
Probable
Glucoser e m u pmtein
Iota
D. Tymsine kinases
P53lYn p56b
p56blk p561ck p56hck p59fyn ~ 7 2
E.Other Functional Molecules Phosphatidylinositol% b e
Phosphorylation
+ +
C. Light Chain-Like Structure Omega
pso
Signal transduction
35 kDa
AS
VpreB
Ig association
IgDQ
B. InIracellularAssociation with IgM Bip
surface expression
16 kDa
53 kDa 56 kDa 56 kDa 56 kDa 56 kDa 59 kDa 72 kDa 80 kDa 85 kDa/
110 kDa
+ +
+
Yamanashi et at., 1989,1991a Yamanashi et at., 1989, 1991a Dymecki et at., 1990; Burkhardt et at.. 1991 Cambell and Sefton, 1992 Li et at., 1992 Burkbardt et at.. 1991 Hutchcmft et at., 1991 Campbell and Sefton, 1990
+
Gold et d.,1992; Clark et at.,1992; Yamanashi et at., 1982
TABLE I1 FUNCTIONAL MOLECULESINVOLVED IN IMMUNOGLOBULIN RECEPTOR-MEDIATED Classification Ca2+-dependentK+ channels Guanine nucleotide phosphatebinding protein (G-protein) Phospholipase C (PLC) Phosphatidylinositol3-kinase Protein kinase C Tyrosine phosphorylated subshates Mouse
Human
Other phosphoproteins
SIGNAL TRANSDUCTION
Molecule
References
PKC-0 (82-80 kDa) PKC-/31//311(82-80 kDa)
MacDougall et al., 1988 Bourguignon et al.. 1990 Graziadei et aZ.. 1990 Carter et al., 1991; Hempel and DeFranco, 1991; Roifman and Wang, 1992 Gold et al., 1992; Yamanashi et al., 1992; Clark et al., 1992 Marquez et al., 1991 Cambier et al.,1987b;Sarthou et aZ.,1989
41 kDa
p21ma PLC7l PLCyz
150-200 kDa 123-129 kDa 101 kDa 81 kDa 69-73 kDa 62-65 kDa 52-56 kDa 47.48 kDa 42 kDa 40 kDa 37 kDa 35 kDa 80 kDa 76 kDa 63 kDa 45 kDa 32 kDa 160 kDa 7.5 kDa 48 kDa 39 kDa 22 kDa Gactin F-actin 52 kDa (CAMP-induced) 23 kDa (CAMP-induced)
Gold et al.,1990
Campbell and Sefton, 1990
Leprince et 01.,1992
Whialer et al..1992
354
NOBUO SAKAGUCHI ET AL.
component linked with B29, as Ig-dig-p. The monomer MB-1 was also detected by Gold et al. (1991). In contrast to MB-1, B29 protein is detected only as a heterodimer component. Using rat mAbs against the same epitope of MB-1 protein, Nomura et al. observed the Ca2+ mobilization response in pre-B cells but not in sIgM-positive WEHI-231 cells. This anti-MB-1 mAb-induced Ca2+ mobilization response is detected in the pre-B cell line obtained by Whitlock/Witte in a longterm culture method or fresh C57BL/6 bone marrow B220+ cells (Fig. 3).The mAbs inhibited the proliferation of pre-B cell lines (Nomura et IgM IgD
pro-B
pre-B
0
u
immature B
mature B
WEHI-231
1000
--%-M?,
BALl7
100
1c
bone marrow cells
lo ol 10
:
0
1
W/W pre-B cells
loool ;
0
1
FIG. 3. Induction of the calcium mobilization response in early B lineage cells through the MB-1 molecule. Murine B cell lymphomas and partially purified B lineage cells are loaded with Fura-2 and tested to see if there is an increase in intracellular calcium after stimulation with rat monoclonal anti-MB-1 antibodies.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
355
al., 1991). Furthermore, the mAbs induced intracellular tyrosine phosphorylation in early B lineage bone marrow cells (Matsuo et al., 1993). These results on induction of Ca2+ mobilization response, effect on proliferation, and induction of tyrosine kinase activity demonstrated that MB-1 itself can mediate the extracellular signals into B lineage cells. It still remains to be determined why sIgM-positive B cells cannot respond effectively to anti-MB-1 antibody stimulation. C. GENOMIC DNA STRUCTURE OF mb-1 GENE CD3 molecules are organized into a rather compact gene segment that contains three CD3 y, 6, and E genes. These three genes have high homology and are conserved between mouse and human (Weiss et al., 1986). The initial analysis of the mb-1 gene demonstrated additional bands on EcoRI digestion in most murine strains, and we thought that these additional bands might encode other genes related to mb-1 and this DNA sequence might encode mb-1-like IgR-associated molecules with IgD or another class of IgR (Kashiwamura et al., 1990). We have isolated genomic DNA clones from the BALB/c cosmid library and characterized over 120 kb by restriction mapping and Southern blot analysis as the chromosomal walking method. We found only a single copy of mb-1 hybridizing sequence within this 120-kb stretch. There is only a single band of 1kb on Northern blot analysis of mRNAs from 18 murine B lineage cell lines (Sakaguchi et al., 1988; unpublished data). We searched for the mb-1-related gene by screening spleen and bone marrow (BALB/c) cDNA libraries; 18 various cDNA clones were chosen at random but the sequencing study did not find a new related sequence coding for association with the IgD receptor (Sakaguchi et al., unpublished data). A BamHI 5.7-kb DNA fragment containing five exons was sequenced. One interesting finding obtained by the homology search on the EMBL database is the segment homologous to the bcl-2 intron sequence which might be related to B lineage-related expression.
D. REGULATORYELEMENTS OF mb-1 GENEEXPRESSION IN THE RESTRICTEDSTAGE OF €3 CELLDIFFERENTIATION In the 5‘ region of the mb-1 gene we found an octamer-like sequence that contains one nucleotide insertion, in contrast to the B29 gene which contains a complete octamer sequence at -125 (5’-ATTTGCAT3’)(Kashiwamura et al., 1990). Expression of B29 is very similar to that of the octamer-binding protein (Oct-2) (Hermanson et al., 1988)which can be considered to regulate Ig gene expression in B lineage cells until the plasma cell stage. The mb-1 gene is expressed from early
356
NOBUO SAKAGUCHI ET AL.
pro-B cells to mature B cells, but is not expressed in plasma cells on which sIgR is not expressed. This octamer-like sequence found at 5' in the mb-1 gene might be responsible for causing the slight difference in gene expression between mb-l and B29 genes in the B lymphoid lineage cells. Hagman, and Feldhaus, and their co-workers studied DNA binding protein [early B cell factor (EBF), Hagman et al., 19911 and BLyF (Feldhaus et al., 1992) around this 5' promoter region sequence 5'-CAAGGGAAT-3' and characterized a novel lineage-specific gene transcription in early B lineage cells. EBF/BLyF appears to consist of at least two polypeptides of around 70-75 and 80-85 kDa. The EBF/BLyF binding site is important for maximal mb-1 gene expression in early B cell stages. But this EBF/BLyF activity is strong in ELA T cell lines, which probably causes the small amount expression of mb-1 mRNA in EL4 cell line (Feldhaus e t al., 1992).This expression of mb-1 mRNA in T cell lines is also observed in human T cell lines but we did not detect the MB-1 protein in T cell lines with a small amount of mb-1 mRNA (Ichigi et al., submitted for publication). It is necessary to study the physiological roles of the trace amounts of mb-1 mRNA detected in thymoma cells or T lineage cells by Northern blot analysis or reverse transcriptase/polymerase chain reaction analysis. The TRElike sequence is also found in the 5' region of the mb-1 gene. Another interesting DNaseI hypersensitive site 1 kb downstream of the mb-1 gene was found in a B cell-specific manner. As human and murine 3' DNA translated and untranslated sequences have highly homologous segments, it is likely that this 3' region controls the stage-specific expression of mb-l on early precursor cells to the sIgM- and sIgDpositive B cells. We recently demonstrated that even before IgR gene rearrangement, MB-1 protein can be expressed on the surface of early B lineage cells that obviously do not express surface IgR (Ichigi et aZ., 1993).It is interesting to note that MB-1 protein functions without sIgR and gene rearrangement. E. HUMAN COUNTERPART OF I@-ASSOCIATED PROTEINS Initial screening of human mb-1 cDNA clones using crosshybridization with murine mb-1 cDNA isolated the human counterpart of the Daudi cDNA clones that had high homologies in the transmembrane and intracellular portions (Sakaguchi et aZ., 1988). This high homology was confirmed in the cDNA library of Raji cells (Reth et aZ., personal communication); Ha et al., 1992) and Ramos cells (Yu and Chang, 1992).This IgR complex structure is quite similar in human B cells: they are associated with IgM and IgD receptors, the size difference is due to the different types of gycosylation, and MB-1 expression was also detected in slg-negative pre-B cells (van Noesel et aZ., 1990,
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECtJLES
357
1991,1992).The size of human MB-1 is considered to be 47 kDa, larger than that of murine MB-l(34 kDa) (van Noesel et al., 1990; Sakaguchi et al., unpublished data). This size difference is probably due to the six N-glycosylated side chains in the extracellular portion of the deduced human MB-1 protein compared with the four sites in mouse MB-1 (Sakaguchi et al., 1988).The size of the human Ig-p chain is 37 kDa which is probably encoded by human B29 gene (van Noesel et al., 1990).The alphabetical nomination based on the size cannot account for the molecule in human IgR-associated components. Comparison of both mouse and human B29, MB-1, and TCR-associated CD3 proteins showed close similarities in all extracellular, transmembrane, and cytoplasmic domains. The extracellular domains contain Ig-like domains (Kashiwamura et al., 1990), from seven p strands which fold into two sheets of three p strands each (A, B, E, and G, F, C). B29 and MB-1 transmembrane regions are clearly defined by hydrophobic amino acids. B29 and MB-1 transmembrane domains showed high conservation; seven amino acids were identical, and the charged glutamic acids as well as the glutamines were located close to the extracellular ends. All the cytoplasmic regions of MB-1, B29, and CDSy, 8, and 6 displayed the antigen receptor tail motif D-(X),-E/D-(X),-Y-(X)2-L(X)7-Y-(X)z-L/I. MB-1 and B29 possessed the extended conserved sequence motif D-(X)S-E-X-E/D-(X)~-Y-E-G-L-N-X-D-(X)*-YE-D-I which contained the YEGLN and YED sequences suggested to act as specific substrates for tyrosine kinases (Reth, 1992; Clark et al., 1992). The mAbs recognizing synthetic peptides of human MB-1 cytoplasmic domain or mouse B29 cytoplasmic domain could detect the 47-kDa and 39-kDa proteins, which are very similar in size to the proteins coprecipitated with membrane IgR molecules in the presence of digitonin (Mason et al., 1991, 1992). These mAbs stained the cytoplasm of B lineage human neoplasms or normal B cells. Using these mAbs, Mason et al. (1992)observed the differential expression of MB-1 and B29 proteins during human B cell differentiation as reported in murine B cells (Sakaguchi et al., 1988; Nomura et al., 1991; Hermanson et al., 1988; Ishihara et al., 1992a). Synthesis of B29 proteins begin laters in precursor B cells than synthesis of MB-1, and ceases before the terminal plasmacyte phase. VII. Attempts to Induce the Membrane Immunoglobulin M Receptor by Immunoglobulin Gene Transfection
Expression of the heavy-chain genes was studied in lymphoid cell lines at several stages of lymphoid differentiation (Deans et al., 1984; Nishikura and Vuocolo, 1984; Sitia et al., 1990) in transgenic mice
358
NOBUO SAKAGUCHI ET AL.
(Grosscheld et al., 1984; Rusconi and Kohler, 1985). Nishikura et al. and Sitia et al. studied the expression of membrane IgM on the surface of myeloma cells by introducing an intact p gene or one capable of directing p, RNA synthesis. The transfectant cells produced'abundant pmRNA but they did not express membrane IgM on the surface. The p, and pmdo not differ in the assembly of light chains or in their stability in the cells. Hombach et al. (1988a) tried to express membrane IgR by introducing the NP-binding p chain gene into myeloma cell J558L containing A light chain only. Most of the transfectants produce secretory IgMlA, but a small proportion of the cell expressed sIgM using fluorescent staining. They enriched the sIgM-positive transfectants and demonstrated the coprecipitation of 34-kDa protein with membrane IgM purified by NP column chromatography. Surface IgM expression is correlated with the expression of mb-1 transcription, which strongly suggests that the 34-kDa protein is encoded by mb-1. Using J558L cells, Hombach et al. (1988a) introduced mb-1, p chain genes and obtained the sIgM-positive transfectants. Introduction of mb-1, p chain genes caused the expression of sIgM receptor on the surface of J558L cells that already contained A light chains and B29 proteins; however, the sIgM-positive transfectant cells did not possess the ability to transport the IgR-mediated signal as in the Ca2+ mobilization response (Justment et al., 1990), which suggested that simple IgR expression is not sufficient for signal transduction. Justment et al. demonstrated the requirement for CD45 molecules in IgR-mediated signal transduction using J558Lpm3 cells that express sIgM by the DNA transfection of p heavy-chain and mb-1 genes but lack the calcium mobilization response to anti-IgR stimulation. Introduction of CD45 gene encoding membrane protein with the tyrosine phosphatase activity could enable the transfectant, J558Lpm3-17.~-46.s,to transduce a Ca2+ mobilization response through the IgR complex, suggesting that CD45 regulates signal transduction by the action of tyrosine phosphatase activity in close association with the IgR complex (Justment et al., 1991; Lin et al., 1992). Stimulation of B cells with anti-CD45 mAb and subsequent crosslinking induced the cytoskeletal association of CD45 and the increased tyrosine phosphorylation of MB-1 and B29 proteins (Lin et al., 1992). These results would suggest that regulation of tyrosine phosphorylation and dephosphorylation is required to maintain the active state of IgR complex on B cells. Venkitaraman et al. (1991)transfected NIH3T3 cells with mb-1,B29, A light-chain gene, and heavy-chain genes of different isotypes. All the isotypes of p, 6, y , a, and E heavy-chain genes could successfully induce IgR on fibroblast cells, which suggests that mb-1 and B29 gene products are capable of IgR expression.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECLJLES
359
Matsuuchi et al. (1992) introduced these four genes (mb-1, B29, p chain, and A chain) into endocrine AtT20 cells. The transfectant expressed IgR by which anti-IgR crosslinking induces the tyrosine kinase and phosphatidyl inositol3-kinase activity. This experiment suggested that some intracellular signaling occurs through these four molecules expressed on the surface of endocrine cells. Costa et al. (19923 prepared the IgMR-positive T cell transfectant by the introduction of IgM, mb-1,and B29 genes into a human T cell line, Jurkat cells. The IgM receptor on T cells recognized PC antigen as a free antigen and mediated the signal into T cells with Ca2+mobilization response to secrete IL-2. These results clearly demonstrate that IgM, MB-1, and B29 are acting as the signal transducer unit with the functional molecules such as src-type tyrosine kinase also expressed in T cells. Although the components involved in TCR complex may also act with B cell antigen receptor, it is confirmed that MB-1 or B29 is necessary to mediate the antigenic signal into cells. Costa et al. studied the associated molecules with transfected functional IgM receptor and showed 52-, 44-,39-, and 34-kDa proteins with surface iodinated samples. The four gene transfectants are summarized in Fig. 4. VIII. Signal Transduction and Its Pathway through the Immunoglobulin Receptor in B Cells
A. INVOLVEMENTOF PROTEIN TYROSINE KINASEIN IgR-ASSOCIATED TRANSMEMBRANE SIGNALING
In B cells, anti-IgR antibody induces the activation of phosphoinosito1 metabolism, accompanied by an increase in subsequent secondmessenger molecules such as inositol triphosphate and diacylglycerol, causing signal transduction through a Ca2+-dependentpathway(s) and activation of phospholipid-dependent serine/threonine kinase and protein kinase C (PKC) (Coggeshall and Cambier, 1984; Bijsterbosch et al., 1985; Cambier et al., 1987a).Whether these are the only molecular changes in IgR-mediated activation remains to be determined. Some receptors for growth factors or hormones possess tyrosine kinase activity, for example, insulin, epidermal growth factor (EGF), platelet-derived growth factor (PDGF), insulin-like growth factor 1 (IGF-1), and colony-stimulating factor 1 (CSF-1). Receptor tyrosine kinases catalyze the phosphorylation of intracellular substrates as well as the ligand-induced autophosphorylation of tyrosine residues within the cytoplasmic domain, leading to increased catalytic activity of the tyrosine kinase itself and interaction between the receptor and the cellular protein for the subsequent signal transduction. Receptormediated multiple signaling pathways are considered to exist because
360
NOBUO SAKACUCHI ET AL.
slgM
1
...)
m b 1 gene P gene
JWL slgM
2
IgHgenes A gene
*.-
NIH3T3
slgE
3 Induction of tyrosine kinase and PI-3 kinase activity
829 gene A gene
endocrine AtT20
4 829 gene P gene A gene
1-
TCR+ CD3+
..)
Induction of tyrosine kinase activity and IL-2 secretion
FIG.4. Expression of IgR on the surface of several kinds of cells by DNA transfection. (1) A myelomacell, JSSSL,expresses A light-chain protein in the absence of heavy chain. DNA transfection of p chain and mb-l genes induces the expression of IgR on the cell surface (Hombach et aZ.,1990). (2)A fibroblast cell line, NIH3T3, expresses surface IgR after DNA transfection of one of five IgH genes (p, 6, y , a,and E chains) when added
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECIJLES
36 1
of the direct phosphorylation of substrates such as phospholipase C, ( PLC-.)I), phosphatidylinositol 3-kinase (PtdIns-SK), ras GTPaseactivating protein (GAP), and rag The src-type tyrosine kinase family includes eight genes (fyn, fgr, hck, lck, lyn, src, yes, and blk), all of which encode membrane-associated protein tyrosine kinase without a membrane-spanning domain (Semba et al., 1985, 1986; Nishizawa et al., 1986; Ziegler et al., 1987; Marth et al., 1985; Yamanashi et al., 1987; Takeya and Hanafusa, 1983; Dymecki et al., 1990). The aminoterminal domain of 60 to 90 amino acids is not homologous in each member, suggesting that this portion is the recognition sequence that interacts with other molecules. Only a few studies have dealt with tyrosine kinase activity (Earp et al., 1984; Harrison et al., 1984) in B cells, for example, 56- and 60-kDa protein components that are inducibly phosphorylated by stimulation with anti-IgR Ab (Nel et al., 1984, 1985). Several src-type tyrosine kinase molecules such as ~ 5 6 ' (Marth "~ et al., 1985; Voronova and Sefton, 1986),blk (Dymecki et al., 1990),lyn (Yamanashi et al., 1989), hck (Quintrell et al., 1987; Ziegler et al., 1987), andfyn (Semba et al., 1986) gene products are expressed in B cells and B lymphoma cells (Gold et al., 1990; Campbell and Sefton, 1990; Ziegler et al., 1987; Yamanashi et al., 1989; Dymecki et al., 1990). Anti-IgR (both IgM and IgD) antibody stimulation induced an early increase (within 30 seconds) in the tyrosine phosphorylation of several cellular substrate proteins (101, 80, 76,63, and 32-35 kDa in addition to the constitutive tyrosine phosphorylation of 101,60, and 29 kDa) in normal B cells, WEHI-231 B lymphoma cells (Gold et al., 1990; Campbell and Sefton, 1990; Yamanashi et al., 1991a),and BAL17 cells (Mizuguchi et al., 1992). Yamanashi et al. (1991a,b) showed that p53"J" and p56'Y" coprecipitate with IgR in both WEHI-231 and BAL 17 cells. Gold et al. (1991) demonstrated that anti-IgM antibody induces tyrosine phosphorylation of MB-1IIg-/3 protein and a 54-kDa unidentified phosphoprotein associated with IgR by using anti-P-Tyr and anti-MB-1 antibodies. Campbell and Sefton (1990) reported that the 80-kDa phosphoprotein coprecipitates with IgR as the major tyrosine phosphorylated protein, A tyrosine phosphoprotein around this together with mb-1, B29, and A light chain genes (Venkitaraman et 01.,1991). (3) An endocrine cell, AtT20, expresses IgR after DNA transfection of four genes (mb-1,B29, p chain, and h light-chain genes). The IgR on endocrine cells could mediate the signals to induce tyrosine kinase and phosphatidylinositol 3-kinase activities (Matsuuchi et ul.,1992).(4)A human T cell line, Jurkat,expresses IgR after DNA transfection o f p chain, h light-chain, and B29 genes and this IgR recognizes PC antigen which results in the induction of tyrosine kinase activity and secretion of IL-2 (Costa et a1.,1992).
362
NOBUO SAKAGUCHI ET AL.
size has been reported in the response of B cells to anti-IgM antibody (Campbell and Sefton, 1990),anti-IgD antibody (Cold et al., 1990),and anti-class I1 antibody (Lane et al., 1990). The equivalent molecule is also characterized as a 72-kDa protein tyrosine kinase, and a 72-kDa protein is detected in both B cells and T cells (Hutchcroft et al., 1991) (Table I). Developmental expression of some tyrosine kinases is also demonstrated in B lineage cells. blk is identified as the tyrosine kinase gene with B cell-specific expression (Dymecki et al., 1990) coding for the 56-kDa tyrosine kinase, which shows the strongest activation in response to anti-IgM antibody in resting B cells (Burkhardt et al., 1991). The ltk receptor-type tyrosine kinase (56- and 64-kDa phosphoproteins) is also expressed in pre-B cells and cerebral neurons (Bernards and de la Monte, 1990);however, an association with IgR-related components has not yet been determined in pre-B cells. lyn encodes protein tyrosine kinases p56'un and p53'un by an alternative splicing that is expressed in B cells but is not detected in T cells (Yamanashi et al., 1989; Yamanashi 1991a; Mizuguchi et al., 1992).PMA also induces tyrosine phosphorylation from 20 minutes to 48 hours in the Triton X-100-soluble substrates of 75, 66, 43,and 28 kDa and in the 56- to 61-kDa Triton X-100-insoluble materials. Stimulation of resting B cells with anti-IgM or anti-IgD antibodies induces protein tyrosine phosphorylation. Of the known members of the src family of protein tyrosine kinases, blk, lyn, lck, hck, and fyn protein tyrosine kinases are shown to be abundantly expressed in normal resting murine splenic B cells and WEHI-231 B lymphoma cells (Burkhardt et al., 1991; Li et al., 1992; Campbell and Sefton, 1992). Lin et al. (1992) demonstrated the coprecipitation of src-type tyrosine kinase molecules with IgR and MB-l/B29 heterodimer components in the presence of digitonin lysis buffer. The detergent containing 1% NP-40 dissociates these tyrosine kinase molecules from IgR but they still coprecipitated with the MB-l/B29 heterodimer complex. It is estimated that in resting B cells, only a small fraction (below 1-3%) of MB-UB29 heterodimers appear to be coprecipitated with protein tyrosine kinases and only a single species of tyrosine kinase might be complexed to each MB-l/B29 complex (Lin et al., 1992). Li et al. (1992) also demonstrated the potential physical interaction of lyn, lck, and hck with IgR, but not blk association with IgR, in an immune complex kinase assay with anti-IgM antibody under digitonin and Brij 96 lysis conditions. This result-that multiple tyrosine kinases are involved in IgR complex-is also supported by the unchanged B cell development in mutant mice that do not express either of these ty-
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
363
rosine kinase molecules as p59fYnby homologous recombination (Stein et al., 1992; Appleby et al., 1992). Takagi et al. (1991) showed the redistribution of tyrosine phosphoprotein beneath the IgR cluster induced by anti-IgM antibody stimulation as cappinglpatching in B cells under microscopic observation. But the movement of IgR on B cells did not correlate with tyrosine kinase activity, as Genistein inhibited the tyrosine kinase activity but did not change the capping/patching of IgR. Downregulation of the one tyrosine kinase activity in normal cells by the other src-type tyrosine kinase was demonstrated in T cells. The catalytic activity of ~ 5 6 ' "is~ normally suppressed by phosphorylation of a carboxyl-terminal tyrosine, Tyr-505, by 50-kDa protein tyrosine kinase, p5OCsk,which specifically phosphorylates Tyr-505 of ~ 5 6 ' " The ~ . data support the concept that the catalytic activity of ~ 5 6 ' "is~regulated by the regulatory "~ action of two enzymes displaying opposite activities: ~ 5 0 ' phosphorylates the suppressive Tyr-505 site of ~ 5 6 ' and " ~ the membrane-bound phosphotyrosine phosphatase CD45 dephosphorylates at this site, resulting in activation (Mustelin et al., 1989).Although the Ser-Ser-Thr motif contained in p5OCskis a candidate site for the regulatory contribution of a serinelthreonine-specific kinase, there is no evidence of phosphorylation at serine or threonine, even after treatment of cells with PMA. It is necessary to study and accumulate the results of this kind of regulatory interaction in kinase molecules. Anti-Ig-induced proliferation of normal human B cells results in increased tyrosine phosphorylation of several cellular proteins (Nel et al., 1984,1985; Lane et al., 1991).Anti-IgR antibody stimulation, PMA, and anti-CD45 mAb crosslinking induced the appearance of tyrosine phosphorylated proteins, although the pattern is distinct in each stimulation of human B cells (Lane et al., 1991). Anti-CD45 mAb inhibited the anti-IgM-induced initial rise in calcium mobilization and the Go-G1 transition of human B cells (Mitter et al., 1987; Gruber et al., 198Y), which suggests the involvement of tyrosine phosphorylation and tyrosine phosphatase interchange reaction in IgR-mediated signal transduction. In human B cells, an IgR complex composed of IgM and 48-kDa (MB-1) and 39-kDa (B29) molecules physically associates with 160- and 75-kDa tyrosine phosphorylated proteins; 160-, 75-, 48-, and 39-kDa proteins with serinelthreonine phosphorylation; and p56lYn, ~56'"~ and , p59fY" tyrosine kinases (Leprince et al., 1992) (Tables I and 11). Of seven identified src-type tyrosine kinases-src, yes, lyn,fyn, lck, blk, and hck-B lineage cells at the pro-B cell phenotype maintained with IL-3 (the F-7 clone) expressed substantial amounts of active fyn
364
NOBUO SAKAGUCHI ET AL
and lyn but blk was barely detectable using a very sensitive in uitro kinase reaction method in combination with the specific immunoprecipitated proteins. This F-7 clone was transfected with IL-2 receptor @chain DNA and expressed the functional IL-2 receptor (IL-2R-a and -p chains), but lacked ~ 5 6 ' "as~the related src-type tyrosine kinase. This artificial transfectant of early B220+ B lineage cells contains the functional coupling of IL-2 receptor with lyn kinase instead of ~ 5 6 ' as "~ the signal transducing tyrosine kinase molecule. Only 2% of the Zyn kinase is physically associated with the membrane IL-2 receptor p chain transfected, indicating that the receptors are capable of mediating external signals into the cells by using one of the kinase molecules, if one of the appropriate kinases is missing (Torigoe et al., 1992). This redundancy of the tyrosine kinase molecules for the membrane receptor is probably important in maintaining the signal-transducing machinery of lymphocytes.
B. INVOLVEMENT OF G-BINDING PROTEINS It is clear that a family of similar guanine nucleotide-binding proteins (G-proteins) exists in most eukaryotic cells and is involved in transmembrane signal transduction. One of the substrate proteins phosphorylated by the tyrosine kinase in receptor-mediated signal transduction has been identified as the phosphoinositide-specific PLC (Carter et al., 1991). Involvement of a G-protein in IgRmediated signaling has been demonstrated with permeabilized cells in which G-protein can be directly activated by a nonhydrolyzable GTP analog, GTP-yS, causing PI-specific PLC activation and inositol phosphate production (Hamett and Klaus, 1989; Gold et aZ., 1987). Melamed et al. (1992) showed that phosphoinositol turnover is regulated by a pertussis-sensitive G-protein possibly through the activity of PLC-yl. Treatment of B cells with pertussis blocked tyrosine phosphorylation of PLC-yl and generation of inositol phosphates, inhibition of c-fos mRNA expression, and DNA synthesis in anti-IgM antibody-stimulated B cells, which suggests that IgR-associated tyrosine kinase activity is regulated by G-protein. G-proteins are heterotrimers consisting of a,p, and T subunits. The a1 subunit contains the binding site for GTP and the activity of GTPase. GI is known to be involved in the activation of PLC, but it is not still clear what kind of G-binding proteins are involved in the IgR complex. A family of membrane-associated proteins ( ~ 2 1 ' 9that bind GTP and contain GTPase activity are also candidate components detected by immunofluorescence microscopy beneath the pole of IgR capping (Graziadei
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
365
et al., 1990). A 41-kDa protein is expressed on the lymphoma plasma membrane that shares structural and functional similarities with GTP-binding protein. The 41-kDa protein stimulates the capping of functional molecules when added to permeabilized lymphoma cells. This protein is closely associated with several cytoskeletal proteins such as actin, myosin, and fodrin and is preferentially accumulated beneath receptor-capped structures (Bourguignon et al., 1990) (Table I). C. ACTIVATIONOF PHOSPHOLIPASE C In many kinds of receptors, the stimulatory ligands enhance turnover of membrane phosphoinositides (Pls). Bijsterbosch et at. (1985) and Ransom et at. (1986) demonstrated that anti-IgR antibody induces an early increase (within 2-3 minutes) in the level of [3H]inositol triphosphate in resting murine B cells, probably by activating phospholipase C. A similar response was measured as generated inositol 174,5-triphosphateand diacylglycerol in immature WEHI-231 B lymphoma cells (Fahey and DeFranco, 1987); however, other activation signals such as lipopolysaccharide and PMA failed to induce this response. In WEHI-231 cells anti-IgM antibody stimulation induces the activation of phospholipase accompanied by the Ca2+ mobilization response and activation of protein kinase C, but ionomycin and PMA could not replace action of anti-IgM antibody (Page and DeFranco, 1988). Anti-IgD antibody also induces the activation of inosito1 polyphosphate production in B cells (Harnett et al., 1989; Brunswick et al., 1989a,b), and this effect is marked when antibody is conjugated to dextran (Brunswick et al., 1989a,b). Hivroz et al. (1990) reported two kinds of human B chronic leukemia cells in which the signal transduction cascade is as simple as sIgR/G-protein/Pl-PLC or the cells in this cascade reaction, are defective. In human B cell chronic leukemia (BCLL), crosslinking of membrane IgM does not always induce both the phosphatidylinositol/Ca2+ mobilization response and cell proliferation (Hivroz et al., 1990). The fibroblast growth factor receptor is a typical model system and can be referred as a signal transduction pathway through activation of receptor-type tyrosine kinase. FGF stimulates tyrosine kinase activity, including autophosphorylation of tyrosine at 766 and PLC-yl, inducing dimerization and association with the FGF receptor. This event subsequently induces phosphatidylinositol metabolism to produce diacylglycerol and phosphatidyl triphosphate. A DNA transfection experiment using point mutation of the FGF receptor in tyrosine residue 766 demonstrated that FGF-induced receptor association to
366
NOBUO SAKAGUCHI ET AL.
PLC-yl mediating [Ca2'] increase and PKC activation are not necessary for the proliferation of fibroblast cells (Peters et al., 1992; Mohammadi et d., 1992). These results suggest that the tyrosine kinase pathway of other molecules is involved in the growth signals, but the activation of PLC-yl is involved in the different responses of the cells. Possible mechanisms coupling the IgR complex to PLC have suggested a role for a G-protein on phosphatidylinositol diphosphate (PIPZ) hydrolysis induced by stimulation of IgR. Ligation of IgR increased immunoprecipitable PLC activity two-fold by 90 seconds and four-fold by 30 minutes. Hempel and DeFranco (1991) studied the mRNA expression of five isozymes of PLC in various B lineage cell lines with the cDNA probe obtained using polymerase chain reaction. All of the B cell lines tested express high levels of PLC-a (1.9 kb) and PLC-y2 (4 kb) mRNA, but do not express PLC-p (7.5 kb) or PLC-6 (2.4 kb) by Northern blot analysis. PLC-71 expression varies in many B lineage cell lines and is not correlated with the developmental stage of the cell line. Tyrosine phosphorylation occurs in both PLC-yl (Carter et al., 1991; Roifman and Wang, 1992) and PLC-y2 (Roifman and Wang, 1992) after 60 seconds of stimulation in B cells, but the selective phosphorylation occurs in PLC-yZ rather than in PLC-yl in response to anti-IgM or anti-IgD antibodies (Hempel et al., 1992). There is no clear understanding of the functional difference between PLC-72 and PLC-yl, both of which share high sequence homology and contain the src-like SH2 and SH3 sequences in the noncatalytic domains involved in the recognition of and binding to phosphorylated tyrosine. As tyrosine phosphorylation induces PLC activity and the binding of this enzyme to the plasma membrane (Todderud et al., 1990), activation of PLC-y2 is considered to be involved in the IgR-mediated signal transduction to the downstream cascade reactions in B cells; however the mode of activation of this PLC activity in B cells is not identical to that of the receptor protein tyrosine kinases of fibroblasts, which induces predominantly the single response of cellular proliferation. Gilliland et al. (1992) studied the PLC-yl-associated tyrosine phosphorylated proteins of T and B cells using a fusion protein containing the two SH2 domains from human PLC-yl and the immunoglobulin constant region of human IgG1. The fusion protein detected lineage-restricted association with a 74-kDa phosphoprotein in T cells and a 93-kDa phosphoprotein in B cells. In contrast to activated EGF receptor in fibroblasts, which coprecipitates with protein tyrosine kinase activity, PLC-yl-associated protein tyrosine kinase ac-
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
367
tivity was not detected in activated lymphocytes, suggesting that lymphocyte PLC-yI SH2-binding proteins are cell lineage specific and may be transiently associated with activated PLC-yI (Table I). The noncovalent coupling of IgR complex to PTK mediating PLC activation may contribute to the various B cell responses of the different stages in the maturation pathway. The defect in IgR-mediated signaling occurred in several mutant cell lines from WEHI-231 cells at the certain point in the signal transduction pathways for the growth arrest after stimulation with anti-IgR treatment accompanied with inducible tyrosine phosphorylation, PLC activation, and Ca2+ mobilization response (Monroe et al., 1989; Page et al., 1991). Several mutant cell lines named VS2.12c1.2 (Monroe et al., 1989) and W52.1, W53.1, W62.1, W306.1 (Page et al., 1991) did not show the inhibitory effect of proliferation by antiIgR stimulation possibly at the proximal point of PLC activation. The other mutant named W305.1 showed the alteration in response to diacylglycerol and calcium (Page et al., 1991). These results would suggest that PLC activation is partly related to the growth inhibitory signals in WEHI-231 cells. Another series of experiments using DNA transfection for IgD receptor on WEHI-231 cells demonstrated that anti-IgD stimulation could not induce growth arrest, although both anti-IgD and anti-IgM Ab stimulation caused inositol phospholipid hydrolysis (Tisch et al., 1988). This difference in transfectant suggested that anti-IgR induced growth arrest may be mediated only through IgM receptor linked to an unknown pathway in addition to the well-characterized PLC activation pathway, although not linked to the IgD-mediated pathway. TO SIGNALS THROUGH IgR D. Ca2+ MOBILIZATIONRESPONSE One of the early biochemical changes induced in activated B cells by anti-IgR antibody is an increase in intracytoplasmic free Ca2+ measured as 45Ca2+(Broun et al., 1979) and by direct determination using fluorescent dye (Pozzan et al., 1982). Monroe and Kass (1985) showed that pharmacological inhibition of either protein kinase C activity or channel-mediated Ca2+ influx completely abrogates the increase in RNA synthesis associated with anti-IgR antibodystimulated B cell activation. The anti-IgR antibody induced Ca2+ mobilization response was detected by signals through receptors of IgM, IgD, and IgG classes in B cell tumors and B cell hybridomas bearing p, 6, or y chains on their surface (Mizuguchi et al., 1986a,b; Harnett et al., 1989).The human B lymphoma cell line Daudi also responded to anti-IgR antibody with Ca2+ mobilization (Kalunta et al., 1988).
368
NOBUO SAKAGUCHI ET AL.
Receptor capping is not dependent on the Ca2+ mobilization response induced by IgR crosslinking (Nachshen et al., 1986). This increase in [Ca2+]i results in the probable activation of Ca2+dependent K+ channels in human B cells (MacDougall et al., 1988). Ca2+ channel blockers (Verapamil) inhibit the proliferation of human B cells (Brent et al., 1990). The patch clamp technique showed two types of whole-cell currents in resting B cells based on activation kinetics. It is considered that an ion channel capable of conducting K+ in normal human B cells is inactivated by increases in [Ca2+Iiand this Ca2+ influx through plasmalemmal Ca2+ channels may be important in the early phase of B cell activation during the transition from Go to G1 (Brent et aZ., 1990). The anti-IgR-stimulated Ca2+ response is partly attributed to the release of intracellular Ca" stores (Ransom et al., 1988), but is also dependent on the influx of extracellular free Ca2+ produced by the increased permeability of the plasma membrane for Ca2+ (Bijsterbosch et aZ., 1986; Wilson et al., 1987). The change in calcium is caused by IgR signaling via activation of hydrolysis of phosphoinositdes into inositol 1,4,5-triphosphate (IP3), not through a voltage-dependent calcium channel that has been described in cardiac muscle and in various neuronal cell types (LaBaer et al., 1986). Lazarus et al. tested antigen-induced Ca2+ signaling in B cells. Antigen stimulation of a mouse TA3 hybridoma B cell transfectant that expresses SP6 anti-TNP-specific IgM with a low molar ratio TNP1-OVA antigen (at any dose up to 500 ng/ml) did not induce an increase in [Ca2'li, but high molar TNP-OVA induced changes in [Ca2+lias observed with anti-IgM crosslinking of IgR (Lazarus et al., 1990). Brunswick et al. observed the induction of B cell activation with a low dose of anti-IgR conjugated to dextran in the absence of obvious elevation in [Ca2+Ii(Brunswick et al., 1989a). These results suggest the possible other signal transduction pathway in B cells that does not correlate with the [Ca2'Ii elevation.
E. PHOSPHATIDYLINOSITOL 3-KINASE Initial tyrosine kinase activation triggered by antigen binding leads to tyrosine phosphorylation of several membrane and cytoplasmic tyrosine residues and probably induces the association of IgR components with other functional molecules. These signaling molecules include serine kinase, PLC-y, GTPase-activating protein (GAP), and PI 3-kinase. PI-3 kinase phosphorylates PI at the 3-position of the inositol ring, which generates phosphatidylinositol3-monophosphate (PI-3-P), and is further phosphorylated by phosphatidylinositol monophosphate (PIP) kinase to form phosphatidylinositol diphosphate. PI-3 kinase, the metabolism and physiological functions of
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
369
which are unknown, associates with middle T antigen in polyoma virus-transformed cells (Whitman et al., 1985) and with v-src protein in v-src-transformed cells (Courtneidge and Heber, 1987; Fukui and Hanafusa, 1989). PI-3 kinase is now considered to be a complex structure consisting of p85a, p85& and pllO associated with PDGF-P receptor. The cDNA structure of p85a and p85p demonstrates that two p85 proteins contain an SH3 domain and two SH2 domains homologous to domains found in several receptor-associated enzymes (Escobedo et al., 1991; Otsu et al., 1991; Skolnik et al., 1991). The SH2 domain acts to facilitate efficient tyrosine phosphorylation. The SH3 domain is also a noncatalytic domain of about 50 amino acids that is shared among many SH2-containing proteins. SH3 domains are found in cytoskeletal proteins such as spectrin and fodrin, which suggests an interaction between the SH3 domain and the membrane or submembrane cytoskelton. It also appears in molecules related to cell growth (Broek et al., 1987; Hughes et al., 1990), cell fusion, and putative transcription factors (Kitamura et al., 1989; Katzav et al., 1989). The pllO component encoded by the gene as 1068 amino acids associates with p85a into an active p85a-pl10 complex that binds the activated receptor (Hiles et al., 1992). Gold et al. (1992) demonstrated increased tyrosine phosphorylation of PI-3 kinase precipitated with anti-p-Tyr antibody in B lymphoma cells after crosslinking (2-3 minutes) by anti-IgM or anti-IgD antibodies. Yamanashi et al. (1992) demonstrated the PI-3 kinase and its activity in antip53'Y"and p56"" immunoprecipitates after anti-IgR crosslinking of WEHI-231 cells and Daudi B lymphoma cells. Clark et u1. (1992) used the MB-1 and B29 cytoplasmic sequence motif NH,-(Asp or Glu)-XXXXXXX-(Asp or G1u)-Tyr-XXX-Leu-XXXXXXX-Tyr=-(Leu or 1le)-COOH to search for the molecules involved in the IgR-mediated signal transduction in B cells by chimeric protein synthesized in uitro. They found that MB-1 associated with Lyn, Fyn, PI-3 kinase, and an unidentified 38-kDa phosphoprotein. B29 motif detected PI-3 kinase and unidentified 40-kDa and 42-kDa phosphoproteins. These results clearly demonstrate that MB-1 and B29 proteins are physically associated with several functional molecules such as the downstream components in signal transduction through IgR. F. PROTEIN KINASE C PATHWAY The ubiquitously expressed calcium/phospholipid-dependent serine/threonine kinase, protein kinase C (PKC), is involved in the regulation of a variety of cellular signal transductions for proliferation, differentiation, and metabolism. Cellular activation of PKC ac-
370
NOBUO SAKAGUCHI ET AL.
tivity appears to be mediated by diacylglycerol, which is produced in the course of inositol phospholipid breakdown. The effect of anti-IgR antibody can be mimicked in B cells by the use of calcium ionophores and PMA, an activator of PKC (Monroe and Kass, 1985; Klaus et al., 1986). Both reagents induce membrane depolarization (Monroe and Cambier, 1983a; Monroe et al., 1983, 1984) and an increase in the expression of surface major histocompatibility complex class I1 molecules (Mond et al., 1981). Monroe et al. tested the effect of phorbol diesters on mouse B lymphocyte kinase C activity. Phorbol diesters, PMA, and 4P-PPD, which are potent tumor promoters, activate partially purified B cell PKC and stimulate B cell membrane depolarization and increased membrane class I1 expression. These events are quite similar to the responses induced by the stimulation of membrane IgM receptor crosslinking with anti-IgM antibody; which suggests that PMA stimulation mimics anti-IgM antibody on B cells. Crosslinking of IgR leads to the phosphorylation of membrane proteins at serine and threonine residues which are very similar as detected by PKC activation (Dasch and Stavitsky, 1985; Hornbeck and Paul, 1986). As an initial step in elucidating the role of plasma membrane kinases in the signal transmission of lymphocytes, the endogenous substrate protein components were analyzed by one-dimensional and two-dimensional PAGE analysis (Chaplin et al., 1979). Hornbeck and Paul extended such a study to clearly determine the inducible phosphoprotein components in activated B cells. They used twodimensional gel analysis of phosphoproteins after stimulation of resting B cells and B lymphoma cells compared with those in such other cell lineages as T lymphoma cells (EL4), glioma cells (C6), neuroblastoma cells (NB2a), and fibroblasts (DAP). PMA also induced a phosphorylation pattern on proteins quite similar that induced by anti-IgM antibody stimulation. All of the phosphoproteins (Nos. 1- 12) compared were phosphorylated at serine or threonine residues, but none were detected on tyrosine residues in this study. Interestingly, most of the phosphorylation induced occurred in membrane-associated and cytoskeletal-bound protein components in B cells, for example, vimentin and lamin (personal communication with Dr. W. E. Paul). One of the interesting molecules that is inducibly phosphorylated is a MARKS protein recently identified as a specific PKC substrate that binds to calmodulin and actin, with inducible phosphorylation during phagocytic activation in macrophage cells (Graff et al., 1989). These results strongly suggest that the anti-IgMstimulated signal transduction pathway is closely related to the PKC pathway in B cells.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
371
The effect of PMA on the stimulation of B cells is not consistent or not a single response. PMA did not induce the Go-to-GI transition of B cells. Anti-IgM antibody induces resting B cells to enter the GI phase of the cell cycle, whereas PMA requires costimulation with a calcium ionophore to induce a similar change (Cambier and Ransom, 1987). Mizuguchi and colleagues demonstrated that PMA inhibits the anti-IgM or anti-IgD response of resting B cells in DNA synthesis, phosphoinositol metabolism, the calcium mobilization and response (Mizuguchi et al., 1986a; Bijsterbosch and Klaus, 1987). In contrast to resting B cells, B cells activated for 29 hours show a positive response to anti-IgR antibody in the presence of PMA. A feedback inhibitory effect of PMA on the IgR-mediated response was observed in an immature B cell model, WEHI-231 B lymphoma cells. PMA treatment of WEHI-231 cells blocked the anti-IgR antibody-induced production of inositol phosphates and accumulation of phosphatidic acid as the phosphorylated product of diacylglycerol and an increase in [Ca2'Ii (Gold and DeFranco et al., 1987). PKC isozymes are a family of cytosolic kinases that translocate from the soluble fraction, as often observed in the plasma membrane, on stimulation. PKC activity is known to be dependent on phospholipids which is provided from the associated form with various cytoskeletal proteins, including spectrin. It is reported that spectrin and PKC-011 are colocalized in untreated lymphocytes and that these two proteins are coincidentally translocated to the focal aggregate within the cytoplasm following stimulation through the antigen-specific receptor or direct activation of PKC by phorbol esters (Gregorio et al., 1992). Sarthou et al. (1989) studied the location of PKC in response to PMA or anti-IgR antibody stimulation. PMA induced a rapid and almost complete redistribution of cytosolic PKC to the plasma membrane fraction in WEHI-231 cells, but anti-IgR antibody did not modify the compartmentalization of PKC, which suggests that IgR-mediated signaling involves an additional pathway independent of PKC activation (Sarthou et al., 1989). cDNA cloning of seven different protein kinase C isozymes (a,p, y, 6, E , 6, and q) and structural analysis demonstrated that each isozyme has different calcium and phospholipid requirements for catalysis. Analysis of myeloid clones derived from bipotential B lineage progenitor cell lines suggests that the B cell phenotype is associated with the expression of PKC-a, although it is also expressed in T cells. Interestingly, differentiation of B cells into antibody-secreting plasma cells is accompanied by an increase in PKC-a and loss of PKC-/3 expression. In normal B cells, two predominant PKC species of 82 and 67 kDa localized in the cytosol of nonstimulated cells (Cambier et al., 1987b; Sarthou et al., 1989). Stimula-
372
NOBUO SAKAGUCHI ET AL.
tion signals in PKC activation lead to disappearance of the 82-kDa polypeptide and a marked increase in association with the membrane fraction. Cambier et al. (1987a) demonstrated the translocation of both components into the nuclear fraction after stimulation with anticlass I1 antibodies. Anti-IgR antibody or PMA stimulation could not induce compartmentalization of the 65-kDa polypeptide in WEHI231 cells in which anti-IgR antibody induced the inhibition of the proliferation (Sarthou et al., 1989). This difference indicates the presence of an alternative signal transduction mechanism through the IgR complex. Depletion of PMA-sensitive protein kinase activity by prolonged PMA treatment could not cause the nonresponsiveness to anti-IgR antibody stimulation in normal murine and human B cells (Mond et al., 1987; Franqois et al., 1988).Marquez et al. (1991) demonstrated pIIpII and a isoenzymes in spleen B cells but did not detect y and E species. They partially purified PKC species using DE-52 chromatography, a hydroxyapatite column, and subsequent resolution by immunoblot analysis with a mAb specific for the p1lpII and (Y species (Marquez et al., 1991). Borner et al. (1992) studied the functional association of individual isoforms of PKC in R6 embryo fibroblast cells transfected with one of the myc, neulerbB2, mos, ras, src, and fos oncogenes. Transformation of R6 cells by the oncogenes ras, src, andfos differentially alter the expression of three isoforms of PKC; it increased PKC-a (8.1-and 3.5-kb mRNA, 81-kDa protein) and PKCd (3.2-kb mRNA, 76- and 74-kDa proteins) and decreased PKC-E (89-kDa protein) in cells, whereas myc, neulerb-B2, or mos did not cause such changes (Borner et al., 1992), which suggests the distinct roles of specific oncogenes in fibroblast transformation. These results also suggest that individual isoforms may play distinct roles in activation of and signal transduction in B cells. Stimulation of IgR leads to a series of metabolic events including the rapid induction of a protein tyrosine kinase activity that can phosphorylate a series of substrates, as yet mostly unidentified (Gold et al., 1990; Lane et al., 1991). Leprince et al. (1992) demonstrated, in human B cells, the presence of protein tyrosine kinase and protein serine kinase activities in slgM immunoprecipitates, and four major substrates with molecular weights of 160, 75, 48, and 39 kDa were found phosphorylated on both tyrosine and serinehhreonine residues. The 48- and 39-kDa components were possibly MB-1 and B29 proteins. The 160- and 75-kDa proteins were unable to be iodolabeled, suggesting that these phosphoproteins are intracellular or inner membrane proteins. The 160-kDa component is unlikely to be an isoform of PLC-7 as the authors could not detect the PLC activity
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
373
from the slgM complex nor could they immunoprecipitate this 160kDa portein with specific antibodies against PLC-yl or -79. Although inducible phosphorylation in the IgM complex in B cells is quite similar to that in T cells, where p5gfun and ~ 5 6 ' are " ~ major elements of signal transduction for the TCR (Rudd, 1988; Samelson et al., 1986, 1990; Cooke et al., 1991), protein serine kinase activity is also found in the IgR complex in B cells. Their observation of both protein tyrosine kinase and protein serine kinase activities in the IgR complex would suggest that a protein tyrosine kinase activity is switched on after slgM crosslinking. Elevation of intracellular cAMP levels inhibits or stimulates the proliferation of human and mouse B cells in response to anti-IgR antibody stimulation (Muraguchi et al., 1984; Simkin et al., 1987) and ionomycin plus PMA (Holte et al., 1988). Treatment of B cells with forskolin or dibutyryl-CAMP results in an increase in protein phosphorylation of the 23- and 52-kDa polypeptides, which were not markedly phosphorylated by stimulation with PMA (Whisler et al., 1992). These two CAMP-dependent kinase-reactive substrates correspond closely to Rap protein (23-kDA)and to the 51-kDa regulatory subunits of PKA I1 which has an autophosphorylation site at Ser-95. A specific 46-kDa/50-kDa protein, named vasodilator-stimulated phosphoprotein (VASP), is the substrate for both CAMP-dependent protein kinase and cGMP-dependent protein kinase and is expressed in platelets, T cells, B cells, and other cells. The increase in the intracellular cAMP level modulates inositol phospholipid hydrolysis in a murine helper T cell clone (Alava et al., 1992). Treatment of permeabilized T cells with purified CAMP-dependent protein kinase resulted in inhibition of the TCR/CD3-mediated proliferative response which is associated with phosphorylation of PLC-71 in the absence of phosphorylation components of the TCR/CD3 complex. Their results strongly suggest that PKA-mediated phosphorylation of PLC may regulate TCR/CDS-induced inositol phospholipid hydrolysis. The molecular linkage of intracellular cAMP level and PKA activity to the signal transduction pathways such as PLC activity and the second messengers for PKC activation in B cells remains for future analysis. Lipopolysaccharide induces both B cells proliferation and differentiation to lg secretion. Costimulation with the F(ab')z fragment of anti-IgM antibody leads to inhibition of Ig secretion but not of proliferation, with decreased expression of ps,K and J chains but not of the IL-2R 55-kDa chain and mb-l mRNA (Berberich and Schimpl, 1992). This effect can be replaced by activating the PKC pathway with pho-
374
NOBUO SAKAGUCHI ET AL.
rbol esters alone or with a combination of phorbol esters and the calcium ionophore ionomycin. An inhibitor of PKC, staurosporine, partially inhibits the differentiation of B cells but allows continued proliferation of LPS-stimulated B cells (Yuan et al., 1992).This inhibitory effect on B cell differentiation is at the transcriptional level of the IgH chain gene, which suggests that LPS allows mediation of the signal through PKC in the distal differentiative events of B cell activation. The synergistic effect of IL-4 on submitogenic doses of anti-IgR antibody in B cell proliferation suggests that IL-4 may play some role in the T cell-dependent helper effect for IgR-mediated signal transduction. Harnett et al. (1991) showed that IL-4 synergizes with nonmitogenic concentrations of anti-IgR antibody to induce translocation of PKC from cytosol to membranes. These results suggest that both types of stimulation might share some part of the signal transduction pathway in B cells. G. DOWNREGULATION Anti-IgM also induces the inhibitory or negative stimulation of B cells in the downregulation of several genes. Anti-IgM antibody induces a marked reduction in CD20 and class I phosphorylation which is probably mediated through PKC activation (Valentine et al., 1989).One interesting downregulation in mature B cells was recently demonstrated by Ma et a2. (1992). They constructed c-myc or N-myc oncogene transgenic mice and obtained the pre-B and B cell lines expressing E p m y c oncogenes. By measuring the recombinase activity after stimulation with anti-IgM antibody, they observed the rapid, specific, and reversible downregulation of RAG-1 and RAG-2 gene expression, which might suggest another important function of IgMmediated signal transmission in early B cell differentiation. Crosslinking of IgR induced expression of the Egr-l gene, a murine early growth factor-inducible gene coding a zinc finger protein, in the majority of B cells excited into cell cycle. But in an immature B cell model of sIgM-positive-WEHI-231 cells, which do not respond to cell growth and proliferation, anti-IgM antibody did not induce an increase in Egr-l or c-fos gene expression (Seyfert et al., 1989). Both anti-IgM and PMA failed to induce the upregulation of Egr-l mRNA in WEHI231 cells, but did induce it in normal B cells and sIgM/IgD-positive BALL7 cells. Egr-l gene product is a key molecule in studying the differences in IgR-mediated signals that result in cellular proliferation in mature peripheral B cells but in induced unresponsiveness in immature B cells.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
375
H . REGULATION OF IMMUNOGLOBULIN GENEREARRANGEMENT It has been postulated that a part of immunoglobulin heavy chain or heavy chain gene expression regulates the further rearrangement to form a productive VH-DH-JH rearrangement. Reth et al. (1987) suggested in a series of transfection experiments that high-level expression of D p protein prevented further rearrangements in heavychain loci to form VH-DH-JH in Abelson virus-transformed pre-B cell lines. Introduction of complete membrane p chain gene into an Abelson virus-transformed pre-B cell line containing only DJ H3 rearranged, 300-19, induced activation of the Ig-K locus for rearrangement. But expression of the secreted form (ps) did not result in activation of the Ig-K locus (Reth et al., 1987). Expression of endogenous immunoglobulin gene rearrangement or expression of immunoglobulin molecules on B cells was tested in transgenic mice containing the rearranged immunoglobulin p chain gene. In this experiment, endogenous immunoglobulin gene was also used as the functional IgR in the presence of IgM transgene, which suggests that an additional genetic mechanism regulates the expression of immunoglobulin receptor, considered as the allelic exclusion (Stall et al., 1988).Rath et al. (1991)tested the expression of endogenous lg genes in rearranged Ig-p transgenic mice and observed a 3- to 10-fold increase of Ig-A light chain-expressing B cells, suggesting quantitative regulation of Ig light-chain isotype expression with the rearranged p chain gene or p chain protein. Gu et al. (1991)also observed that membrane-bound D, protein arrested B cell differentiation by preventing VH-DH-JH joining in knockout mice of the membrane exon of the p chain. T h e D p polypeptide can be expressed on the surface of Abelson virustransformed pre-B cell lines and coprecipitates with the surrogate p/A5/VpreB complex (Tsubata et al., 1992). I. DEVELOPMENTAL REGULATION OF ASSOCIATIONOF FUNCTIONAL COMPONENTS OF IgR Several differences in IgR-mediated signal transduction were reported in IgM-positive B cells. IgR-mediated signals induce unresponsiveness in immature B cells (sIgM+sIgD-) (Nossal, 1983; Cooper et al., 1980; Metcalf and Klinman, 1976). The IgR-mediated signals also induce an apoptotic change in certain B lymphomas with only slgM-positive WEHI-231 cells. This cell line is considered representative of immature B cells and can be converted into the tolerant state for antigen stimulation (Hasbold and Klaus, 1990).Crosslinking of IgR on WEHI-231 cells results in inositol phospholipid
376
NOBUO SAKAGUCHI ET AL.
hydrolysis, producing diacylglycerol and activating PKC and inositol 1,4,5-triphosphate, causing release of calcium from intracellular stores and cell cycle arrest at GI phase (Page and DeFranco, 1990). Meucchi and Ashman (1986) suggested that the intrinsic unresponsiveness is part of the mechanism of B cell tolerance in BDFl mice injected with 2,4,6-trinitrobenzenesulfonicacid (TNBS). YellenShaw et al. (1991) compared IgR-mediated signaling between neonatal and adult IgR-positive B cells. In neonatal B cells, anti-IgR antibody crosslinking does not cause phosphatidylinositol turnover but induces the calcium mobilization response, which may be involved in the receptor-mediated signals for B cell anergy. It may result from a different association of the components with IgM receptors on immature B cells and mature B cells. Yellen-Shaw and Monroe (1992) pointed out the difference between the coprecipitated proteins especially in 56-kDa protein. Multiple src family tyrosine kinases such as lyn, fyn, and lck can potentially interact with IgR of any B cell type, but the patterns of expression of these kinases vary in B cells of various subsets and differentiation stages (Campbell and SeAon, 1992). B cell unresponsiveness also occurs in B cells in certain circumstances such as in transgenic mouse using hen egg lysozyme and is considered immunological anergy. To explain this unresponsiveness of B cells, Cambier et al. (1990) suggested that the antigen induces receptor desensitization as documented in a variety of receptor systems. Ligand-induced desensitization in B cells appears to be unrelated to PKC activation and is possibly mediated by an uncoupling of membrane IgR from G-proteins that regulate PLC, but the molecular basis of the ligand-induced desensitization remains unclear. Mond et al. (1990) examined the effect of several protein kinase activators on IgR-mediated B cell stimulation. Four kinds of PKC activatorsPMA, indolactam, bryostatin, and mezerein-inhibited (PIPZ) hydrolysis and elevations in intracellular calcium, but only PMA and bryostatin inhibited cell proliferation induced with anti-IgM antibody. These results suggested that a high level of PIP2 hydrolysis and an increased level of intracellular calcium are not essential for anti-IgR-mediated B cell proliferation. B cells of xid (X-linked immunodeficiency) model mice express a high ratio of surface IgM to IgD, express no Lybd differentiation antigen, and cannot respond to the T-independent antigen type 2 (Mosier et al., 1977; Scher, 1982), caused by one or several X-linked xid gene(s) possibly in a family of XLR genes as previously characterized (Cohen et al., 1985a,b; Siege1 et al., 1987). The xid B cells probably provide useful information on normal B cell development and activation. Grupp and Harmony
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
377
(1985) indicated that anti-IgR antibody does not induce inositol phospholipid turnover in xid B cells. To stimulate xid B cells or neonatal B cells, anti-IgR antibody has to be conjugated with Sepharose or dextran as a solid matrix, in contrast to normal mature B cells which respond to the soluble form of anti-IgR antibody (Mond et al., 1983; Klaus et al., 1986). Interestingly, soluble anti-IgM antibody could induce early tyrosine phosphorylation, as evidenced by the active signal transduction through IgR in xid B cells but with a different pattern of phosphorylated proteins compared with normal adult B cells (Lindsberg et al., 1991). Anti-IgM antibody stimulation also induces inositol phospholipid hydrolysis and subsequent calcium mobilization in xid B cells (Rigley, 1989). These signals in xid B cells do not result in the subsequent proliferation, but a combination of phorbol ester and calcium ionophore effectively stimulates proliferation of xid and normal B cells (Klaus et al., 1986; Lindsberg et al., 1991), which may suggest that the defect in xid B cells is distal to the IgRtyrosine kinase events but proximal to the PKC active point. Murine peritoneal B cells are composed of a unique subset of Ly-1 + B cells characterized by a number of functions such as predisposition to autoantibody secretion, association with many autoimmune diseases, and association with many B lineage malignant tumors. B cells from transgenic mice of antierythrocye antibody are eliminated from peripheral lymphoid organs but remain in the peritoneal cavity as Ly-1 + B cells (B-1 cells) (Murakami et al., 1992). A few Ly-1 + B cells that escape apoptotic death remain in the peritoneum possibly because of a failure to encounter erythrocyte antigen. This is one mechanism considered for the autoimmunue state in uiuo. Peritoneal B cells contain higher PKC activity (60% higher) than conventional peripheral B cells, as measured on a per weight of protein basis. PMA treatment induced a fivefold increase in PKC activity at 4 hours after stimulation in both membrane and cytoplasmic fractions of peritoneal B cells in comparison to normal €3 cells (Cohen and Rothstein, 1991). Conventional B cells and peritoneal B cells both contain similar amounts of PKC-P (80-kDa) but peritoneal B cells contain more PKC-a (80-kDa), as detected by Western blot analysis using isozyme-specific mAbs. PKC-a is considered the most resistant to the proteolytic pathway of downregulation (Ase et al., 1988). This isoenzyme difference in peritoneal B cells probably causes a long-lasting effect of PKC activity or may contribute to the isoenzyme-specific downstream substrate(s), which can alter the response of peritoneal B cells. The long-term stimulation of normal B cells with anti-IgR antibody induces constitutive expression of the
378
NOBUO SAKAGUCHI ET AL.
Ly-1 + phenotype on the responding B cells which also behave as peritoneal Ly-1 + B cells (Ying-zi et al., 1991; Rothstein et al., 1991). This transition from Ly-1 - to Ly-1 + expression in activated B cells might be associated with the acquisition of responsiveness to PMA (Rothstein et al., 1991). IX. What Occurs after Stimulation of the immunoglobulin Receptor-Mediated Signal Transduction Cascade?
Crosslinking of sIgR by anti-IgM antibody leads to a number of well-characterized early metabolic changes including tyrosine phosphorylation of several substrate molecules in the cytoplasm and activation of PLC, which results in phosphoinositide hydrolysis with the production of two second-messenger molecules, inositol triphosphate, and diacylglycerol (Maino et al., 1975). Diacylglycerol in conjunction with Ca2' activates PKC, resulting in translocation of the enzyme from the cytosol to the plasma membrane. These events lead to subsequent downstream events mediated by protein phosphorylation (Coggeshall and Cambier, 1984; Bijsterbosch et al., 1985), followed by the expression of growth-related genes that include proto-oncogenes (Snow et al., 1986; McCormack et al., 1984; Fisher et al., 1991). In contrast to resting cells which express a low level of c-mycmRNA, anti-IgR stimulation induces an early and transient rise in c-myc gene expression at the level of gene transcription and post-transcription (Phillips and Parker, 1987; Buckler et al., 1988; Klemsz et al., 1989). A more prolonged increase in c-myc transcription occurs in B cells stimulated with anti-IgR in the presence of cytochalasin which perturbs the cytoskeletal structure (Buckler et al., 1990). It might be interesting to study the molecular mechanism of cytochalasin on the level of c-myc mRNA for the search for further downstream events in receptormediated signal transduction pathways. As a candidate for the next messenger to induce nuclear events resulting from cytoplasmic signaling cascade rections, sequence-specific DNA binding proteins (NF-KB) can be considered to act in the transcriptional activation responsible for the anti-Ig-mediated signal transduction process. NF-KB binding activity was induced after surface IgR crosslinking (Liu et al., 1991; Rooney et al., 1988) with a Ca2+-mediated-mechanism but through a signal transduction pathway that was either PKC dependent (Liu et al., 1991)or independent (Rooney et al., 1988).The p50 subunit of NF-KB synthesized as a precursor molecule of 105 kDa subsequently releases the amino-terminal p50 polypeptide with re1 homology, DNA binding activity, and activity for gene transcription. One of
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
379
two mRNAs, 4.0 and 2.6 kb, derived from the p105 locus (Ghosh et al., 1990) is differentially expressed in the B lineage cells. The 2.6-kb mRNA is found predominantly in pro-B (LyD9) and pre-B cells (38B9, 70213, and 230-238) and is barely detected in more mature B cell lines (WEHI-231, S194, and MPC11) (Liou et al., 1992),which suggests that the 2.6-kb ( IKBY)mRNA is regulated developmentally in B cell differentiation (Inoue et al., 1992). Another candidate for distal coupling of gene expression of IgR is AP-1, which is produced by the formation of heterodimers between c-Jun- and c-Fos-related proteins, leading activation of gene transcription through binding to a specific cis-acting element, TRE[5’-TGA(C/G)TCA-3’1. Phorbol ester and anti-p antibody treatment have similar but not identical effects on B cells. Stimulation of B cells to enter S phase of the cell cycle by treatment with a combination of phorbol ester and calcium ionophore also stimulated nuclear TRE binding activity (Chiles et al., 1991), which suggested that in primary B cells nuclear TRE binding activity represents a downstream signaling event resulting from changes in PKC activity and intracellular calcium. Crosslinking of IgR stimulates the de nouo synthesis of Jun-B and c-Jun proteins and induces the nuclear AP-1 (Chiles and Rothstein, 1992). X. Future Perspectives
Intensive studies on IgR-mediated signal transduction provide much information concerning the molecular mechanism of B cell activation. Two IgR-associated components (MB-1 and B29) have universal roles for the expression and signal transduction of IgR; however, many questions remain to be clarified on the developmental difference in IgR-mediated signals. It is necessary to demonstrate the functional molecules involved in the IgR-related complex of B lineage cells at various differentiational stages. In B cells, it is still difficult to apply the known signal transduction pathways used in receptors for many kinds of growth factors. It is necessary to confirm each of the pathways or the downstream molecular events in B cells. It is also important to find the target genes that are upregulated or downregulated by signals through IgR in B cells. IgR-mediated signals seem to stimulate proliferation, surface expression of several other functional receptors of soluble lymphokines, differentiation of B cells toward class switching, affinity maturation of V regions of IgR, and transformation into antibody-secreting plasma cells. For these cellular events, B cells probably have to run into cell cycle and divide several times. Our knowledge on the molecular events that occur 30 seconds to
380
NOBUO SAKAGUCHI ET AL.
1 hour after stimulation has only begun to be acquired. To approach these questions, we need improvements in the technology used to chase the moelcular events for a longer period. It is also necessary to find the unknown component involved in the signal transduction machinery in B lineage cells. Fortunately, many investigators are now studying these molecules with stepwise advancement. It will not be long before we know the activation mechanism of B cells and can apply the results to the study of autoimmune disease or immunodeficient states.
ACKNOWLEDGMENT We thank Ms. Yasuoka for her excellent assistance in preparing this manuscript.
REFERENCES Abbas, A. K., Haber, S., and Rock, K. L. (1985).J . Zmmnol. 135,1661. Adams, E., Basten, A,, and Goodnow, C. C. (1990).Proc. Natl. Acad. Sci. U.S.A. 87, 5687-5691. Alava, M. A.,DeBell, K. E., Conti, A., Hoffman, T., and Bonvini, E. (1992).Biochem. J . 284,189-199. Alberini, C. M., Bet, P., Milstein, C., and Sitia, R. (1990).Nature (London)347,485-487. Albrecht, D. L., and Noelle, R. J. (1988).J.Immunol. 141,3915-3922. Albrecht, D., Millis, J. W., and Noelle, R. J. (199O).J. Zmmunol. 144,3251-3256. Ales-Martinez, J. E.,Warner, G. L., and Scott, D. W. (1990). Cell. Immunol. 127,527534. Allison, J.. McIntyre, B., and Bloch, D. (1982).J. Immunol. 129,2293. Alt, F. W., Bothwell, A. L. M., Knapp, M., Koshland, M., and Baltimore, D. (1980).Cell (Cambridge, Mass.)24,293-301. Alt, F.,Rosenberg, N.,Lewis, S., Thomas, E., and Baltimore, D. (1981). Cell (Cambridge, Mass.)27,38. Amitay, R., Schchar, I., Rabinovich, E., Haimovich, J., and Bar-Nun, S. (1992).J . Biol. Chem. 267,20694-20700. Antoine, J-C., and Avrameas, S. (1974).Eur. J. Immunol. 4,468. Appleby, M. W., Gross, J. A., Cooke, M. P., Levin, S. D., Qian, X.,and Perlmutter, R. M. (1992).Cell (Cambridge, Mass.) 70,751-763. Ase, K., Berry, N., Kikkawa, U., Kishimoto, A., and Nishizuka, Y. (1988).FEBS Lett. 236, 396. Bauer, S . R., Kudo, A., and Melchers, F.(1988).EMBOJ. 7,111-116. Berberich, I., and Schimpl, A. (1992).Eur.]. Zmmunol. 22,525-529. Bernards, A.,anmd de la Monte, S. M. (1990).EMBOJ. 9,2279-2287. Bijsterbosch, M. K., and Klaus, G. G. B. (1987).Eur.J. Zmmunol. 17,113-118. Bijsterbosch, M. K., Meade, C. J.,Turner, G. A,, and Klaus, G. G . (1985).Cell (Cambridge, Mass.)41,999-1006. Bijsterbosch, M. K., Rigley, K. P., and Klaus, G. G. B. (1986).Biochem. Biophys. Res. Commun. 137,500-506. Borner, C., Guadagno, S. N., Hsiao, W. W.-L., Fabbro, D., Barr, M., and Weinstein, I. B. (1992).J.B i d . Chem. 267, 12900-12910. J . Biol. Chem. 258,5135-5141. Borst, J., Alexander, S., Elder, I., and Terhorst, C. (1983).
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECIJLES
38 1
Bourguignon, L. Y. W., Walker, G., and Huang, H. S. (1990).J . Immunol. 144,2242-2252. Brent, L. H., Butler, J. L., Woods, W. T., Jr., and Bubien, J. K. (1990).J . Immunol. 145, 2381-2389. Brink, R., Goodnow, C. C., Crosbie, J., Adams, E., Eris, J., Mason, D. Y.,Hartley, S., and Basten, A. (1992).J.Exp. Med. 176,991-1005. Broek, D., Toda, T., Michaeli, T., Levin, L., Birchmeier, C., Zoller, M., Powers, S., and Wigler, M. (1987). Cell (Cambridge, Mass.) 48,789-799. Broun, J., Sha’afi, R. I., and Unanue, E. R. (1979).J.Cell Biol.82,755-766. Bruce, J., Symington, F. W., McKern, T. J., and Sprent, J. W. (1981).J . Immunol. 127, 2496. Brunswick, M., June, C. H., Finkelman, F. D., Dintizis, H. M., Inman, J. K., and Mond, J. J. (1989a).Proc. Notl. Acad. Sci. U.S.A.86,6724-6728. Brunswick, M., June, C. H., Finkelman, F. D., and Mond, J. J. (1989b).J.Immunol. 143, 1414- 1421. Buckler, A. J., Rothstein, T., and Sonenshein, G. E. (1988).Mol. Cell. Biol. 8,1371-1375. Buckler, A. J., Rothstein, T., and Sonenshein, G. E. (199O).J.Immunol. 145,732-736. Burkhardt, A. L., Brunswick, M., Bolen, J. B., and Mond, J. J. (1991). Proc. Natl. Acad. Sci. U.S.A.88,7410-7414. Burnet, F. M. (1959). “The Clonal Selection Theory of Acquired Immunity.” Cambridge Univ. Press, Cambridge. Burrows, P. D., Kearney, J. F., Lawton, A. R., and Cooper, M. D. (1978).J . Immunol. 120, 1526. Cambier, J. C., and Ranson, J. T. (1987).Annu. Rev. Immunol. 5,175-199. Cambier, J. C., Ligler, F. S., Uhr, J. W., Kettman, J. R., and Vitetta, E. S. (1978). Proc. Natl. Acad. Sci. U.S.A.75,432-435. Cambier, J. C., Justment, L. B., Newell, K., Chen, Z. Z., Harris, L. K., Sandoval, V. M., Klemsz, M. J., and Ransom, J. T. (1987a).Zmmunol. Rev. 95,37-57. Cambier, J. C., Newel], M. K., Justment, L. B., McGuire, J . C., Leach, K. L., and Chen, Z. Z. (1987b).Nature (London)327,629-632. Cambier, J. C., Fisher, C. L., Pickles, H., and Morrison, D. C. (199O).J.Immunol. 145, 13-19. Campbell, K. S., and Cambier, J. C. (1990).EMBOJ. 9,441-448. Campbell, M.-A., and Sefton, M. B. (1990). EMBOJ. 9,2125-2131. Campbell, M.-A., and Sefton, M. B. (1992). Mol. Cell. Biol. 12,2315-2321. Campbell, K. S., Hager, E. J., and Cambier, J. C. (1991a).J. Immunol. 147, 1575-1580. Campbell, K. S., Hager, E. J., Friedrich, R. J., and Cambier, J. C. (1991b). Proc. Natl. Acad. Sci. U.S.A.88,3982-3986. Carter, R. H., Park, D. J., Rhee, S. G., and Fearon, D. T. (1991). Proc. Natl. Acad. Sci. U.S.A.88,2745-2749. Casten, L. A., and Pierce, S. K. (1988).J.Immunol. 140,40. Cattaneo, A. and Neuberger, M. S. (1987). EMBOJ. 6,2753-2758. Chaplin, D. D,, Wedner, H. J,, and Parker, C. W. (1979). Biochem.]. 182,527-546. Chen, J.. Stall, A. M., Herzernberg, L. A., and Herzenberg, L. A. (1990). EMBOJ. 9, 2117-2124. Chien, M. M., and Aschman, R. F. (1983).J.Immunol. 130,2568-2573. Chiles, T. C., and Rothstein, T. L. (1992).J . Immunol. 149,825-831. Chiles, T. C., Jialing, L., and Rothstein, T. L. (1991).J.Immunol. 146, 1730-1735. Choi, Y. S., Lee, M. S., and Rosenpire, A. J. (1983). Mol. Immunol. 20, 1249-1257. Clark, M. R., Campbell, K. S., Kazlauskas, A,, Johnson, S. A., Hertz, M., Potter, T. A., Pleiman, C., and Cambier, J. C. (1992). Science 258, 123-126.
382
NOBUO SAKAGUCHI ET AL
Coffman, R. L. (1982). Zmmunol. Rev. 69,5. Coffman, R. L., and Weissman, I. L. (1981). Nature (London)289,681. Coggeshall, K. M., and Cambier, J. C. (1984).J.Zmniunol. 133,3382. Cohen, D. I., Hedrick, S. M., Nielsen, E. A,, D’Eustachio, P., Ruddle, F., Steinberg, A. D., Paul, W. E., and Davis, M. M. (1985a).Nature (London)317,369-371. Cohen, D. I., Steinberg, A. D., Paul, W. E., and Davis, M. M. (1985b).Nature (London) 317,372. Cohen, D. P., and Rothstein, T. (1991).J.Zmmunol. 146,2921-2927. Cooke, M. P., Abraham, K. M., Forbush, K. A,, and Permutter, R. M. (1991). Cell (Cumbridge, Mass.) 65,281-291. Cooper, D. D. (1978).In “B Lymphocytes in the Immune Respone” (M. D. Cooper, D. E. Mosier, I. Sceh, and E. S. Vitteta, eds.), p. 63. Elsevier, Amsterdam. Cooper, M. D., Kearney, J. F., Gathings, W. E., and Lawton, A. R. (1980). Zmniunol. Rer;. 52,29. Cooper, M. D., Mulvaney, D., Countiho, A., and Cazenave, P. A. (1986).Nature (London) 321,616. Costa, T. E., Franke, R. R., Sandchez, M., Misulovin, Z., and Nussenzweig, M. C. (1992). J . E x p . Med. 175, 1669-1676. Courtneidge, S. A., and Heber, A. (1987). Cell (Cambridge,Mass.) 50, 1031-1037. Dasch, J. R., and Stavitsky, A. B. (1985). Mol. Zmmunol. 22,379-389. Davidson, H. W., West, M. A., and Watts, C. (199O).J.Zmmunol. 144,4104-4109. Deans, R. J., Denis, K. A., Taylor, A,, and Wall, R. (1984).Proc. Natl. Acud. Sci. U.S.A.81, 1292-1296. DeFranco, A. L., Davis, M. M., and Paul, W. E. (1982). “UCLA Symposia on Molecular and Cellular Biology” (E. S. Vitetta, ed.), Vol. 24, p. 445. Academic Press, New York. DeFranco, A. L., Gold, M. R., and Jakway, J. P. (1987). Zmmnol. Reo. 95, 162-175. Dorshkind, K. (1990).Annu. Reo. Zmmunol. 8,111-137. Dubois, P. M., Stepinski, J., Urbain, J., and Sibley, C. H. (1992). Eur. J . Zmmunol. 22, 851-857. Dymecki, S. M., Niederhuber, J. E., and Desiderio, S. V. (1990). Science 247,332-336. Early, P., Davis, M. M., Kaback, D. B., Davidson, N., and Hood, L. (1979). Proc. Nutl. Acad. Sci. U.S.A. 76,857-861. Early, P., Rogers, J., Davis, M., Calame, K., Bond, M., Wall, R., and Hood, L. (1980).Cell (Cambridge, Moss.) 20,313-319. Earp, H. S., Austin, K. S., Buessow, S. C., Dy, R., and Gillespire, G. Y. (1984).Proc. Natl. Acad. Sci. U.S.A. 81,2347-2351. Ehrlich, P. (1900). The Croonian lecture: On immunity. Proc. R . SOC. Lond. (Biol.)66, 424. Escobedo, J. A., Navankasattusas, S., Kavanaugh, W. M., Milfay, D., Fried, V. A., and Williams, L. T. (1991). Cell (Cambridge, Mass.) 65,75-82. Fahey, K., and DeFranco, A. L. (1987).J.Zmrnunol. 138,3935-3942. Fechheimer, M., and Cebra, J. J. (1982).J.Cell Biol. 93,261-268. Feldhaus, A. L., Mbangkollo, D., Arvin, K. L., Klug, C. A., and Singh, H. (1992). Mol. Cell. Biol. 12, 1126-1133. Fisher, C. L., Ghysdael, J., and Cambier, J. C. (1991).]. Zmmunol. 146,1743-1749. Francois, D. T., Katona, I. M., June, C. H., Wahl, L. M., Feuerstein, N., Huang, K. P., and Mond, J. J. (1988).J.Zmmunol.140,3338-3343. Freiden, P. J., Gaut, J. R., and Hendershot, L. M. (1992). EMBOJ. 11,63-70. Fu, S. M., Winchester, R. J., and Kunkel, H. G . (1975).J.Zmmunol. 114,250-252. Fukui, Y., and Hanafusa, H. (1989).Mol. Cell. Biol. 9, 1651-1658.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
383
Ghosh, S., Gifford, A. M., Rivere, L. R., Tempest, P., Nolan, G. P., and Baltimore, D. (1990).Cell (Cambridge,Mass.) 62, 1019-1029. , and Ledbetter, Gilliland, L. K., Schieven, G. L., Norris, N. A., Kanner, S. B., A ~ f f oA., J. A. (1992).J.Biol. Chem. 267,13610-13616. Gold, M. R., and DeFranco, A. L. (1987).J.Immunol. 138,868-876. Gold, M. R., Jakway, J. P., and DeFranco, A. L. (1987).J.Immunol. 139,3604-3613. Gold, M. R., Law, D. A,, and DeFranco, A. L. (1990).Nature (London)345,810-813. Gold, M. R., Matsuuchi, L., Kelly, R. B., and DeFranco, A. L. (1991).Proc. Natl. Acad. Sci. V.S.A.88,3436-3440. Gold, M. R., Chan, V. W.-F., Turck, C. W., and DeFranco, A. L.(1992).J.Immunol. 148, 2012-2022. Goodnow, C. C., Crosbie, J., Adelstein, S., Lavoie, T. B., Smith-Gill, S. J., Brink, R. A., Pritchard-Briscoe, H., Wotherspoon, J. S., Loblay, R. H., Raphael, K., Trent, R. J., and Basten, A. (1988).Nature (London)334,676-682. Graff, J. M., Young, T. M., Johnson, J. D., and Blackshear, P. J. (1989).J.Biol. Chem. 264, 2 1818-2 1823. Gray, D., MacLennan, I. C. M., Bazin, H., and Kahn, M. (1982).Eur.J.Immunol. 12,564. Graziadei, L., Riabowol, K., and Bar-Sagi, D. (1990).Nature (London)347,396-400. Grosscheld, R., Weaver, D., Baltimore, D., and Costantini, F. (1984).Cell (Cambridge, MUS S .)38,647-658. Gruber, M. F., Bjorndahl, J. M., Nakamuro, S., and Fu, S. M. (1989).J.Immunol. 142, 4144-4 152. Grupp, S. A., and Harmony, J. A. K. (1985).J.Immunol. 134,4087-4094. Grupp, S. A., Snow, E. C., and Harmony, J. A. K. (1987).Cell. Immunol. 109,181-191. Gu, H., Kitamura, D., and Rajewsky, K. (1991).Cell (Cambridge, Mass.)65,47-54. Gupta, S. K., and Woda, B. A. (1988).J.Immunol. 140, 176-182. Ha, H., Kubagawa, H., and Burrows, P. D. (1992).J.Immunol. 148,1526-1531. Haas, I., and Wabl, M. (1983).Nature (London)306,387-389. Hagman, J., Travis, A., and Grosschedl, R. (1991).EMBOJ. 10,3409-3417. Halbrugge, M., Eigenthaler, M., Polkr, C., and Walter, U. (1992). Cell. Signaling 4, 189-199. Hardy, R. R., Carmack, C. E., Shinton, S. A., Kemp, J. D., and Hayakawa, K. (1991). 1.E x p . Med. 173,1213-1225. Harnett, M. M., and Klaus, G. G. B. (1988).J.Immunol. 140,3135-3139. Harnett, M. M., Holman, M. J., and Klaus, G. G. B. (1989).Eur.J. Immunol. 19, 19331939. Harnett, M. M., Holman, M. J., and Klaus, G. G. B. (1991).J.Immunol. 147,3831-3836. Harrison, M. L., Low, P. S., and Geahlen, R. L. (1984).J.Biol. Chem. 259,9348-9350. Hasbold, J., and Klaus, G. G . B. (1990).Eur.J. Immunol. 20,1685-1690. Haskins, K., Kubo, R., White, J., Pigeon, M., Kappler, J., and Marrack, P. (1983).J.Exp. Med. 157,1149. Haustein, D., and Tsammer, TH. (1987).Bering Znst. Mitt. 81,100-109. Haustein, D., and van der Ahe, D. (1986).Eur.J.Immunol. 16, 113-115. Hayashi, S. I., Kunisada, T., Ogawa, M., Sudo, T., Kodama, H., Suda,T., Nishikawa, S., and Nishikawa, S. I. (1990).J . E x p . Med. 171, 1683-1695. Hempel, W. M., and DeFranco, A. L. (1991).J.Immunol. 146,3713-3720. Hempel, W. M., Schatzman, R. C., and DeFranco, A. L. (1992).J.Immunol. 148,30213027. Hendershot, L., Bole, D., Kohler, G., and Kearney, J. (1987).J.Cell Biol. 104,761-767. Hendershot, L. M., T;-g, J., and Lee, A. S. (1988).MoZ. Cell. Biol. 8,4250-4256.
384
NOBUO SAKAGUCHI ET AL
Herayil, B. J., and Pillai, S. (1991).J . Exp. Med. 173,111-116. Hermanson, G. G., Eisenberg, D., Kincade, P. W., and Wall, R.(1988).Proc. Natl. Acad. Sci. USA. 85,6890-6894. Herzenberg, L. A., Stall, A. M., Lalor, P. A,, Sidman, C., Moore, W. A., Parks, D. R.,and Henenberg, L. A. (1986).Immunol. Reo. 12,564. Hiles, I. D., Otsu, M., Volinia, S., Fry, M. J., Gout, I., Dhand, R., Panayotou, G., Ruiz-Larrea, F., Thompson, A., Totty, N. F., Hsuan, J. J., Courtneidge, S. A., Parker, P. J., and Waterfield, M. D. (1992).Cell (Cambridge,Mass.) 70,419-429. Hivroz, C., Geny, B., Brouet, J.-C., and Grillot-Couvalin, C. (1990).J. Immunol. 144, 2351-2358. Hollis, G. F., Evans, R. J., Stafford-Hollis, J. M., Korsmeyer, S. J., and McKearn, J. P. (1989).Proc. Natl. Acad. Sci. U.S.A.86,5552-5556. Hoke, H., Torjesen, P., Blomhoff, H. K., Ruud, E., Funderud, S., and Smeland, E. B. (1988).Eur. J . Immunol. 18,1359-1366. Hombach, J,, Leclerq, L., Radbruch, A., Rajewsky, K., and Reth, M. (1988a).EMBOJ. 7, 3451-3456. Hombach, J., Sablitzky, F., Rajewsky, K., and Reth, M. (1988b).J . E x p . Med. 167, 652-657. Hombach, J., Lottspeich, F., and Reth, M. (1990a).Eur. J . Immunol. 20,2795-2799. Hombach, J,, Tsubata, T., Leclercq, L., Stappert, H., and Reth, M. (1990b).Nature (London)343,760-762. Hornbeck, P., and Paul, W. E. (1986).J . Biol. Chem. 261, 14817-14824. Hughes, D. A,, Fukui, Y.,and Yamamoto, M. (1990).Nature (London)344,355-357. Hutchcroft, J. E., Harrison, M.L., and Geahlen, R.L. (1991).J.Biol. Chem. 266,1484614849. Ichigi, Y., Muraguchi, A., Tokushima, M., Naitoh, K., Nomura, J., Fujiyama, C., Kuwahara, K., Igarashi, H., Inui, S., Kimoto, M.,and Sakaguchi, N. (1993) submitted for publication. Iglesias, A,, Kopf, M., Williams, G. S., Bu her, B., and Kohler, G. (1991).EMBO J . 10, 2 147-2 156. Inoue, J.-I., Kerr, L. D., Kakizuka, A., and Verma, I. M. (1992).Cell (Cambridge,Mass.) 68,1109-1 120. Inui, S., and Sakaguchi, N. (1992).Immunol. Lett., 34,279-288. Isakson, P. C., and Vitetta, E. S. (1980).J.E x p . Med. 148, 1628. Ishihara, K., Medina, K., Hayashi, S-I., Pietrangeli, C. E., Namen,A. E., Miyake, K., and Kincade, P. W. (1991).Deo. Immunol. 1,149-161. Ishihara, K.,Wood, W. J., Jr., Damore, M., Hermanson, G. G., Wall, R.,and Kincade, P. W. (1992a).Proc. Natl. Acad. Sci. U.S.A. 89,633-637. Ishihara, K., Wood, W. J., Jr., Wall, R., Sakaguchi, N., Michnoff, C., Tucker, P. W., and Kincade, P. W. (1992b).J. Immunol., in press. Iwabuchi, N., Williams, D. B., Ngugen, H. P., Wu, Y., Tisch, R., Azuma, T., Phillips, M. J., and Hozumi, N. (1992).Eur. J . Immunol. 22,2507-2511. Julius, M. H. (1987).Immunol. Reo. 95,177-194. Justment, L. B., Wienands, J., Hombach, J., Reth, M., and Cambier, J. C. (1990). J . Immunol. 144,3272-3280. Justment, L. B., Campbell, K. S., Chien, N. C., and Cambier, J. C. (1991).Science 252, 1839- 1842. Kalunta, C. I., Kaptein, J. S., Niedzin, H., Scott, S. J., Lee, G. H. D., and Lad, P. M. (1988). Proc. Natl. Acad. Sci. U.S.A.85,9204-9208. Karasuyama, H., Kudo, A., and Melchers, F. (1990).J . E x p . Med. 172,969-972.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
385
Kashiwamura, S., Koyama, T., Matsuo, T.,Steinmetz, M., Kimoto, M., and Sakaguchi, N. (1990).J . Zmmunol. 145,337-343. Katzav, S., Martin-Zanca, D., and Barbacid, M. (1989). EMBOJ. 8,2283-2290. Kehry, M., Edward, S., Douglas, R.,Sibley, C., Rashke, W., Fambrough, D., and Hood, L. E. (1980). Cell (Cambridge,Mass.) 21,393-406. Kenny, J. J., Martinez-Maza, O., Fehninger, T., and Aschman, R. F. (1979).J.Zmmunol. 122,1287-1284. Kerr, W. G.,Cooper, M. D., Feng, L, Burrows, P. D., and Hendershot, L. M. (1989). Znt. Zmmunol. 1,355-361. Kim, K.-M., Yoshimura, T., Watanabe, H., Ishigami, T., Nambu, M., Hata, D., Higaki, Y., Sasaki, M., Tsutsui, T., Mayumi, M., and Mikawa, H. (1991).J . Zmmunol. 146, 819825. Kincade, P. W. (1987).Ado. Zmmunol. 41, 181-267. Kincade, P. W., Lee, G., Watanabe, T., Sun, L.,and Scheid, M. P. (198l).J.Zmmunol. 127, 2262. Kincade, P. W., Lee, G., Pietrangeli, C. E., Hayashi, S. I., and Gimble, J. M. (1989).Annu. Reo. Zmmunol. 7,111-143. Kishimoto, T. (1985). Annu. Reo. Zmmunol. 3, 133. Kishimoto, T., and Ishizaka, K. (1975).J . Zmmunol. 114,585. Kitamura, D., and Rajewsky, K. (1992). Nature (London) 356,154-156. Kitamura, D., Kaneko, H., Miyagoe, Y., Ariyasu, T., and Watanabe, T. (1989).Nucleic Acids Res. 17,9367-9379. Kitamura, D., Roes, J., Kuhn, R.,and Rajewsky, K. (1991).Nature (London)350,423-426. Kitamura, D., Kudo, A., Schaal, S., Muller, W., Melchers, F., and Rajewsky, K. (1992). Cell (Cambridge,Mass.) 69,823-831. Klaus, G. G . B., O’Garra, A., Bijsterbosch, M. K., and Holman, M. (1986). Eur.1. Zmmunol. 16,92-97. Klemsz, M. J., Justment, L. B., Palmer, E., and Cambier, J. C. (1989).J . Zmmunol. 143, 1032. Knapp, M. R.,Strober, S., Lin, C. P., Tucker, P. W., Newell, N., and Blattner, F. (1981).In “B Lymphocytes in the Immune Response” (Functional, Developmental, and Interactive Properties)” (N. Klinman, D. Mosier, I. Scher, and E. V. Vitetta, eds.), p. 19. ElseviedNorth-Holland, Amsterdam. Koch, N., and Haustein, D. (1983). Mol. Zmmunol. 20,33-37. Koshland, M. E. (1975).Ado. Zmmunol. 20,41-69. Kritz, M. K., Vitetta, E. S., and Sullivan, T. J. (1986).J.Zmmunol. 137,478-483. Kudo, A., and Melchers, F. (1987). EMBOJ. 6,2267. Kudo, A., Pravtcheva, D., Sakaguchi, N., Ruddle, F. H., and Melchers, F. (1987a). Genomics 1, 1. Kudo, A., Sakaguchi, N., and Melchers, F. (1987b). EMBOJ. 6, 103. Kudo, A., Thalmann, P., Sakaguchi, N., Davidson, W. F., Pierce, J. H., Kearney, J. F., Reth, M., Rolink, A., and Melchers, F. (1992). Znt. Zmmunol. 4,831-840. Kuittler, M. R.,and Haas, I. G . (1992). EMBOJ. 11,1573-1581. LaBaer, J., Tsien, R. Y., Fahey, K. A., and DeFranco, A. L. (1986). J. Zmmunol. 137, 1836-1844. Lane, P. J. L., McConnel, F. M., Schieven, G. L., Clark, E. A., and Ledbetter, J. A. (1990). J . Immunol. 144,3684-3692. Lane, P. J . L., Ledbetter, J. A., McConnel, F. M., Draves, K. E., Deans, J., Schieven, G. L., and Clark, E. A. (199l).J.Zmmunol. 146,715-722. Lazarus, A. H., Mills, G. B., and Delovitch, T. L. (1990).J . Zmmunol. 144,4147-4155.
386
NOBUO SAKAGUCHI ET AL.
Leprince, C., Draves, K. E., Ledbetter, J. A,, Torres, R.M., and Clark, E. A. (1992).Eur.1. Immunol. 22,2093-2099. Li, Z-X., Mahajan, S., Prendergast, M. M., Fargnoli, J., Zhu, X., Klages, S., Adam, D., Scheven, G. L., Blake, J., Bole, J. B., and Burkhard, A. L. (1992). Biochem. Biophys. Res. Commun. 30,1536-1544. Liao, F., Giannini, S. L., and Birshtein. B. K. (1992).J.Zmmunol. 148,2909-2917. Lin, J,, and Justment, L. B. (1992).]. Zmmunol. 149, 1548-1555. Lin, J,, Brown, V. K., and Justment, L. B. (1992).J.Zmmunol. 149,3182-3190. Lindsberg, M-L., Brunswick, M., Yamada, H., Lees, A., Inman, J., June, C. H., and Mond, J. J. (1991).J.Zmmunol. 147,3774-3779. Liou, H-C., Nolan, G. P., Ghosh, S., Fujita, T., and Baltimore, D. (1992). EMBO J . 8, 3003-3009. Liu, J,, Chiles, T. C., Sen, R.,and Rothstein, T. L. (1991).J . Zmmunol. 146, 1685-1691. Ma, A,, Fischer, P., Dildrop, R., Oltz, E., Rathbun, G., Achacoso, P., Stall, A., and Alt, F. (1992).EMBOJ. 11,2727-2734. MacDougall, S. L., Grinstein, S.,and Gelfand, E. W. (1988).J.C h . Znoest. 81,449-454. Maino, V. C., Hayman, M. J., and Crumpton, M. J. (1975). Bi0chem.J. 146,247-252. Marchalonis, J. J. (1976).In “Contemporary Topics in Immunobiology” (H. N. Eisen and R.A. Reisfeld, eds.), Vol. 5, pp. 125-160. Plenum Press, New York/London. Marquez, C., Martinez-A., C., and Bosca, L. (1991).J.Immunol. 147,627-632. Marth, J. D., Peet, R.,Krebs, E. G., and Perlmutter, R. M. (1985). Cell 43,393-404. Mason, D. Y., Cordell, J. L., Tse, A. G. D., Van Dongen, J. J. M., van Noesel, C. J. M., Micklem, K., Borst, J., and Gatter, K. C. (1991).J. Zmmunol. 147,2474-2482. Mason, D. Y., van Noesel, C. J. M., Cordell, J. L., Comans-Bitler, W. M., Micklem, K., Tse, A. G. D., van Lier, R. A. W., and van Dongen, J. J. M. (1992). Eur.J.Zmmunol. 22, 2753-2756. Mather, E. L., Nelson, K. J., Haimovich, J., and Perry, R. P. (1984). Cell (Cambridge, MUSS.)36,329-338. Matsuo, T., Kimoto, M., and Sakaguchi, N. (1991).J.Immunol. 146,1584-1590. Matsuo, T., Nomura, J., Kuwahara, K., Igarashi, H., Inui, S., Hamaguchi, M., Kirnoto, M., and Sakaguchi, N. (1993).J.Zmmunol. 150,l-11. Matsuuchi, L., Gold, M. R.,Travis, A., Grosschedl, R., DeFranco, A. L., and Kelly, R.B. (1992).Proc. Natl. Acad. Sci. U.S.A.89,3404-3408. McCormack, J. E., Pepe, V. H., Kent, R. B., Dean, M., Marshak-Rothstein, A., and Sonenshein, G. E. (1984).Proc. Natl. Acad. Sci. U.S.A.81,5546-5550. McKearn, J., Baum, C., and Davie, J. (1984).]. Zmmunol. 132,332-339. McKearn, J. P., and Rosenberg, N. (1985).J.Zmmunol. 15,295-298. McNiece, I. K., Langley, K. E., and Zsebo, K. M. (1991).]. Zmmunol. 156,3785-3790. Melamed, I., Downey, G. P., and Roifman, C. M. (1991). Biochem. Biophys. Res. Commun. 176,1424-1429. Melamed, I., Wang, G., and Roifman, C. M. (1992).J.Zmmunol. 149, 169-174. Melchers, F., and Cone, R. E. (1973). Transplant. Reo. 25,26-53. Melchers, F., von Boehmer, H., and Phillips, R. A. (1975). Transplant Reo. 25,26. Melchers, F., Karasuyarna, H., Haasner, D., Bauer, S., Kudo, A., Sakaguchi, N., and Rolink, A. (1993). Zmmunol. Today 14,60-68. Melchers, U., and Uhr, J. W. (1973).J.Exp. Med. 138, 1282-1287. Metcalf, E. S., and Klinrnan, N. R. (1976).J . Exp. Med. 143,1327. Meucchi, V., and Ashman, R. F. (1986).J . Zmmunol. 136,2339-2341. Meuer, S., Fitzgerald, K., Hussey, R., Hodgdon, J., Schlossman, S., and Reinherz, E. (1983).1.Exp. Med. 157,705.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLEClJLES
387
Misener, V., Jongstra-Bilen, J., Young, A. J., Atkinson, M. J., Wu, G. E., and Jongstra, J. (1990).J . Immunol. 145,905-909. Misener, V., Downey, G. P., and Jongstra, J. (1991). Znt. Immunol. 3, 1129-1136. Mittler, R. S., Greenfield, R. S., Schacter, B. Z., Noel, R. F., and Hoffman, M. K. (1987). J . Zmmunol. 138,3159-3166. Mizuguchi, J., Bevan, M. A., Li, J. H., and Paul, W. E. (1986a). Proc. Natl. Acad. Sci. U.S.A.83,4474-4478. Mizuguchi, J., Tsang, W., Morrison, S. L., Beave, M. A., and Paul, W. E. (1986b). J . Immunol. 137,2162-2167. Mizuguchi, J., Yamanashi, Y., Ehara, K., Tamura, T., Nariuchi, H., Gyotoku, Y., Fukazawa, H., Uehara, Y., and Yamamoto, T. (1992).J.Zmmunol. 148,689-694. Mohammadi, M., Dionne, C. A., Li, W., Spivak, L. T., Honegger, A. M., Jaye, M., and Schlessinger, J. (1992). Nature (London)358,681-684. Moller, G., Alarcon-Riquelme, M., Clinchy, B., Gontijo, C. M., and Hoiden, I. (1991). Scand.1. Zmmunol. 33,111-116. Mond, J. J., Feuerstein, N., Finkelman, F. D., Huang, F., Huang, K. P., and Dennis, G. (1987). Proc. Natl. Acad. Sci. U.S.A.84,8588-8592. Mond, J. J., Balapure, A,, Feruerstein, N., June, C. H., Brunswick, M., Lindsberg, M-L., and Witherspoon, K. (199O).J.Zmmunol. 144,451-455. Mond, J. J., Seghal, E., Kung, J., and Finkelman, F. D. (1981).J.Immunol. 116, 1656. Mond, J. J., Schaefer, M., Smith, J., and Finkelman, F. D. (1983). J . Immunol. 131, 2 107-2 109. Mongini, P. K. A., Blessinger, C., Posnett, D. N., and Rudich, S. M. (1989).J.Zmmunol. 143,1565-1574. Monroe, J. C., and Cambier, J. C. (1983a).J.Exp. Med. 157,2073. Monroe, J. C., and Cambier, J. C. (1983b).J.Zmmunol. 133,2641. Monroe, J. G., and Kass, M. J. (1985).]. Zmmunol. 135,1674-1682. Monroe, J. CG., Seyfert, V. L., Owen, C. S., and Sykes, N. (1989).J . E x p . Med. 169, 1059- 1070. Monroe, J. G., Niedel, J. E., and Cambier, J. C. (1983).J.Zmmunol. 131,2641-2644. Monroe, J. G., Niedel, J. E., and Cambier, J. C. (1984).J.Immunol. 132,1472-1478. Mosier, D. E., Mond, J. J,, and Goldings, E. A. (1977).J.Immunol. 119, 1874-1878. Muller, B., Cooper, L., and Terholst, C. (1992).Eu7.J. Zmmunol. 22, 1621-1625. Muller-Sieburg, C. E., Whitlock, C. A., and Weissman, I. L. (1986). Cell (Cambridge, Mass.) 44,653. Munro, S., and Pelham, H. R. B. (1986). Cell (Cambridge,Mass.) 48,899-907. Muraguchi, A., Miyazaki, K., Kehrl, J. H., and Fauci, A. S. (1984). J. Immunol. 133, 1283-1287. Murakami, M., Tsubata, T., Okamoto, M., Shimizu, A,, Kumagai, S., Imura, H., and Honjo, T. (1992). Nature (London) 357,77-80. Mustelin, T., Coggeshall, M.K., and Altman, A. (1989).Proc. Natl. Acad. Sci. U.S.A.86, 6302-6306. Myers, C. D., b i z , M. K., Sullivan, T. J., and Vitetta, E. S. (1987). J . Immunol. 138, 1705- 1711. Nachshen, D. A., Pfeffer, L. M., and Tamm, I. (1986).J.Biol. Chem. 261,15134-15139. Namen, A. E., Lupton, S., Hjerrid, K., Wangnall, J., Mochizuki, D. Y.,Schmierer, A., Mosley, B., March, C., Urdal, D., Gillis, S., Cosman, D., and Goodwin, R. G. (1988). Nature (London) 333,571-573. Nel, A. E., Landreth, G. E., Tung, H. E., and Galbraith, R. M. (1984).Biochem. Biophys. Res. Commun. 125,859-866.
388
NOBUO SAKACUCHI ET AL.
Nel, A. E., Nawvailies, M., Rosberger, D. F., Landreth, G. E., Goldschmidt-Clermont, P. J., Baldwin, C. J., and Galbraith, R. M. (1985).J.Zmmunol. 135,3448-3453. Nishikawa, S. I., Ogawa, M., Nishikawa, S., Kunisada, T., and Kodama, H. (1988).Eur.1. Immunol. 18,1767-1777. Niskikura, K., and Vuocolo, C. A. (1984).E M B O J . 3,689-699. Nishimoto, N., Kubagawa, H., Ohno, T., Cartland, G. L., Stamkovic, A. K., and Cooper, M. D. (1991).Proc. Natl. Acad. Sci. U.S.A. 88,6284-6288. Nishizawa, M.,Semba, K., Yoshida, M., Yamamoto, T., Sasaki, M., and Toyoshima, K. (1986).Mol. Cell. Biol. 6,511-517. Nomura, J., Matsuo, T., Kubota, E., Kimoto, M., and Sakaguchi, N. (1991).Znt. Zmmunol. 3,117-126. Nossal, G. J. V. (1983).Annu. Reo. Zmmunol. 1,33. Osmond, D. G. (1985).J. Znoest. Dennatol. 85,2-9. Osmond, D.G., and Everett, N. B. (1964). Blood 23,l-17. Osmond, D. G., and Nossal, G. J. V. (1974).Cell. Immunol. 13,132. Otsu, M., Hilles, I., Gout, I., Fry,M. J., Ruiz-Larrea, F., Papayotou, G., Thompson, A., Dhand, R., Hsuan, J., Totty, N., Smith, A. D., Morgan, S. J., Coutneidge, S. A., Parker, P. J., and Waterfield, M. D. (1991).Cell (Cambridge,Mass.) 65,91-104. Owen, J. J. T., Raff, M. C., and Cooper, M. D. (1975).Eur. J . Zmmunol. 5,468. Page, D. M.,and DeFranco, A. L. (1988).J. Zmmunol. 140,3717-3726. Page, D. M., and DeFranco, A. L. (1990).Mol. Cell. Biol. 10,3003-3012. Page, D. M.,Cold, M. R., Fahey, K. A,, Matsuuchi, L.,and DeFranco, A. L. (1991).J. Biol. Chem. 266,5563-5574. Paige, C. J., Kincade, P. W., and Ralph, P. (1978).J. Zmmunol. 121,641. Parikh, V. S., Bisop, G. A., Liu, K.-J., Do, B. T., Ghosh, M. R., Kim, B. S., and Tucker, P. W. (1992).J. Exp. Med. 176,1025-1031. Parker, D. C. (1975).Nature (London)258,361-366. Parkhouse, R. M. E. (1990).Immunology 69,298-302. Patty, C. L., Alarcon, B., Malin, R., Weinberg, K., and Terholst, C. (1987).].Biol.Chem. 262,4854-4859. Perry, R. P. (1981).]. Cell. Biol. 91,28-38. Peters, K.G., Marie, J., Wilson, E., Ives, H. E., Escobedo, J., Rosario, M. D., Mirda, D., and Williams, L. T. (1992).Nature (London)358,678-681. Petrini, M., Emerson, D. L., and Galbraith, R. M. (1983). Nature (London)306,73-74. Phillips, N. E., and Parker, D. C. (1987).Mol. Zmmunol. 24,1199. Pierce, S. K., Morris, J. F., Grusby, M. J., Kaumaya, P., Buskirk, A. V., Srinivasan, M., Crump, B., and Smolenski, L. A. (1988).Zmmunol. Rev. 106,149-178. Pillai, S., and Baltimore, D. (1987).Nature (London)329, 172-174. Pillai, S.,and Baltimore, D. (1988).Curr. Top. Microbiol. Zmmunol. 137,136-139. Pozzan, T., Arslan, P., Tsien, R. Y.,and Rink, T. J. (1982).J. Cell Biol. 94,335-340. Pure, E., and Tardelli, L. (1991). Proc. Natl. Acad. Sci. U S A . 89,114-117. Quintrell, N.,Lebo, R., Varmus, H., Bishop, J. M., Pettenati, M. J., Le Beau, M. M., Diaz, M. O., and Rowley, J. D. (1987).Mol. Cell. Biol. 7,2267-2275. Raff, M. C., Messon, M., Owen, J. J. T., and Cooper, M. D. (1975a).Nature (London)259, 224. Raff, M. C., Owen, J. J. T., Cooper, M. D., Lawton, A. D., Megson, M., and Gathings, W. E. (1975b).J. Exp. Med. 142,1052. Zmmunol. 137,708-714. Ransom, J., Hams, L. K., and Cambier, J. C. (1986).J. Ransom, J., Chen, M., Sandoval, V. M., Pastemak, J. A., Digiusto, D., and Cambier, J. C. Zmmunol. 140,3150-3155. (19W.J.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
389
Rath, S., Nisonoff, A., Selsing, E., and Durdik, J. (1991).J.Zmmunol. 146,2841-2847. Reth, M. (1989). Nature (London)338,383-384. Reth, M. (1992).Antiu. Reo. Zmmunol. 10,97-121. Reth, M., Petrac, E., Wiese, P., Lobel, L., and Alt. F. (1987). EMBO I. 6, 32993305. Reth, M., Hombach, J., Wienands, J., Campbell, K. S., Chien, N., Justment, L. B., and Cambier, J. C. (1991).Zmmunol. Today 12,196-201. Rigley, K., Harnett, M. M., Phillips, R. J., and Klaus, G. G. B. (1989).Eur.1. Zmmunol. 19, 2081-2086. Rijkers, G. T., Griffioen, A. W., Zegers, B. J. M., and Cambier, J. C. (1990). Proc. Natl. Acad. Sci. U.S.A.87,8766-8770. Rogers, J., and Wall, R. (1984).Adu. Zmmunol. 35,39-59. Roifman, C. M., and Wang, G. (1992).Biochem. Biophys. Res. Commrrn. 183,411-416. Rolink, A., and Melchers, F. (1991). Cell (Cambridge,Mass.) 66, 1081-1094. Rolink, A., Streb, M., Nishikawa, S-I., and Melchers, F. (1991). Eur. J. Immunol. 21, 2609-2612. Rooney, J. W., Dubois, P. M., and Sibley, C. H. (1988).Life. Sci. 42,497-504. Rosenberg, N., Baltimore, D., and Sher, C. D. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 1932. Rosenspire, A. J., and Choi, Y. S. (1988). Life Sci. 42,497-504. Rothstein, T. L., Kolber, D. L., Murphy, T. P., and Cohen, D. P. (1991).]. Zmmunol. 147, 3728-3735. Rudd, C. E., Trevillyan, J. M., Dasgupta, F. E., Wong, L. L., and Schlossman, S. F. (1988).Proc. Natl. Acad. Sci. U.S.A.85,5190-5194. Rusconi, S., and Kohler, G. (1985). Nature (London)314,330-334. Sakaguchi, N., and Melchers, F. (1986). Nature (London)324,579-582. Sakaguchi, N., Kishimoto, T., Kikutani, H., Watanabe, T., Yoshida, N., Shimizu, A., Yamawaki-Kataoka, Y.,Honjo, T., and Yamamura, Y. (1980).J. Zmmunol. 125,26542659. Sakaguchi, N., Berger, C. N., and Melchers, F. (1986). EMBOJ. 5,2139-2147. Sakaguchi, N., Berger, C. N., and Melchers, F. (1987).Curr. Top. Microbiol. Zmmunol. 135, 139-148. Sakaguchi, N., Kashiwamura, S., Kimoto, M., Thalmann, P., and Melchers, F. (1988). EMBO J. 7,3457-3464. Samelson, L. E., Harford, J. B., and Klausner, R. D. (1985).Cell (Cambridge, Mass.)43, 223-231. Samelson, L. E., Patel, M. D., Weissman, A. M., Harford, J. B., and Klausner, R. D. (1986).Cell (Cambridge,Mass.) 46,1083-1090. Samelson, L. E., Philips, A. F., Loung, E. T., and Klausner, R. D. (1990). Proc. Natl. Acad. Sci. U.S.A.87,4358-4362. Sanders, S. K., Kubagawa, H., Suzuki, T., Butler, J. L., and Cooper, M. D. (1987). J. Immunol. 139,188-193. Sarthou, P., Hernry-Tourlme, N., and Cazenave, P-A. (1989). Eur. J. Zmmunol. 19, 1247-1252. Scher, I. (1982). lmmunol. Reo. 64,117. Schiff, C., Milili, M., and Fougereau, M. (1989). Eur.]. Zmmunol. 19,1873-1878. Schiff, C., Bensmana, M., Guglielmi, P., Milili, M., Lefranc, M. P., and Fougereau, M. (1990). Znt. Zmmunol. 2,201-207. Scott, D. W., Chase, J. H., Warner, G. L., O’Garra, A., Klaus, G. G. B., and Quill, H. (1987).Zmmunol. Reo. 99,153.
390
NOBUO SAKAGUCHI ET AL.
Semba, K., Yamanashi, Y., Nishizuka, M., Sukegawa, J., Yoshida, M., Sasaki, M., Yamamoto, T., and Toyoshima, K. (1985).Science 227, 1038-1040. Semba, K., Nishizawa, M., Miyajima, N., Yoshida, M. C., Sukegawa, J., Yamanashi, Y., Sasaki, M., Yamamoto, T., and Toyoshima, K. (1986).Proc. Natl. Acad. Sci. U.S.A.83, 5459-5463. Seyfert, V. L., Sukhatme, V. P., and Monroe, J. G. (1989).Mol. Cell. Biol. 9,2083-2088. Shaw, A. C., Mitchell, R. N., Weaver, Y. K., Campos-Torres, J., Abbas, A. K., and Leder, P. (1990).Cell (Cambridge,Mass.) 63,381-392. Sidman, C. (1981).Cell (Cambridge,Mass.) 23,379-389. Sidman, C. L., Vercovici, T., and Gitler, C. (1980).Mol. Zmmunol. 17, 1575-1583. Siege], J. N., Turner, C. A., Klinman, D. M., Wilkinson, M., Steinberg, A. D., Mcleod, C. L., Paul, W. E., Davis, M. M.,andCihen, D. I. (1987).J. E x p . Med. 166,1702-1715. Siekman, D. G.,Asofsky, R., Mosier, D. E., Sitron, I. M., and Paul, W. E. (1978).J. Exp. Med. 147,815. Sirnkin, N. J., Jelinek, D. F., and Lipsky, P. E. (1987).J. Zmmunol. 138,1074-1081. Sitia, R., Neuberger, M. S., and Milstein, C. (1987).EMBOJ. 6,3969-3977. Sitia, R., Neuberger, M., Alberini, C., Bet, P., Fra, A,, Valetti, A., Williams, G., and Milstein, C. (1990).Cell (Cambridge,Mass.)60,781-790. Skolnik, E. Y., Margolis, B., Mohammadi, M., Lowenstein, E., Fischer, R.,Drepps, A., Ullrich, A., and Sclessinger, J. (1991).Cell (Combridge,Mass.) 65,83-90. Snow, E. C., Fetherston, J. D., and Zimmer, S. (1986).J. E x p . Med. 164,944-949. Stall, A. M., Kroese, G . M., Gadus, F. T., Sieckmann, D. G., and Herzenberg, L. A. (1988). Proc. Natl. Acad. Sci. U.S.A.85,3546-3550. Stein, P. L.,Lee, H-M., Rich, S., and Soriano, P. (1992).Cell (Comhridge, Mass.) 70, 741-750. Storb, U., Denis, K. A,, Brinster, R. L., and Witte, 0. N. (1985). Noture (London)316, 356-358. Strasser, A. (1988).Eur.J.Zmmunol. 8, 1803-1810. Takagi, S., Daibata, M., Last, T. J., Humphreys, R. E., Parker, D. C., and Sairenji, T. (1991).J. E x p . Med. 174,381-388. Takai, Y., Sakata, T., Iwagami, S., Tai, X.,Kita, Y., Harnaoka, T., Sakaguchi, N., Yamagishi, H., Tsuruta, Y., Teraoka, H., and Fujiwara, H. (1992).J . Zmmunol. 148,13291337. Takemori, T., Mizuguchi, J., Miyazoe, I., Nakanishi, M., Shigemoto, K., Kimoto, H., Shirasawa, T.,Maruyama, N., and Taniguchi, M. (1990).EMBOJ. 9,2493-2500. Takeya, T., and Hanafusa, H. (1983).Cell (Combridge,Mass.) 32,881-890. Taylor, R. B., Duffus, W. P. H., Raff, M. C., and de Petris, S. (1971).Nature (London)223, 225. Teale, J. M., and Klinman, N. R.(1980).Nature (London)288,385. Thores, B., Schults, M.-F., and Vassalli, P. (1985).EMBOJ.4,361-368. Tisch, R., Roifman, C. M., and Hozumi, N. (1988).Proc. Natl. Acad. Sci. U.S.A. 85, 6914-6918. Todderud, G., Wahl, M. I., Rhee, S. G . ,and Carpenter, G. (1990).Science 249,296-298. Tonegawa, S. (1983).Nature 302,575-581. Torigoe, T., Saragovi, H. U.,and Reed, J. C. (1992).Proc. Natl. Acad. Sci. U.S.A. 89, 2674-2678. Tsokos, G. C., Kinoshita, T.,Thyphronitis, G., Patel, A. D., Dickler, H. B., and Finkelman, F. D. (199O).J. Zmmunol. 144,239-243. J . E x p . Med. 172,973-976. Tsubata, T., and Reth, M. (1990). Tsubata, T., Tsubata, R.,and Reth, M. (1991).Eur. J . Zmmunol. 21,1359-1363.
IMMUNOGLOBULIN RECEPTOR-ASSOCIATED MOLECULES
39 1
Tsubata, T., Tsubata, R.,and Reth, M. (1992). Int. Immunol. 4,637-641. Ullrich, A., and Schlessinger, J. (1990). Cell (Cambridge,Mass.) 61,203-212. Valentine, M. A., Meier, K. E., Rossie, S., and Clark, E. A. (1989).J. Biol. Chem. 264, 11282-1 1287. Van Noesel, C. J. M., Borst, J., De Vries, E. F. R.,and van Lier, R. A. W. (1990). Eur. J . Immunol. 20,2789-2793. Van Noesel, C. J. M., Lier, R.A. W., Cordell, J. L., Tse, A. G. D., Van Schijndel, G. M. W., De Vries, E. F. R., Mason, D. Y., and Borst, J. (1991).J.Immunol. 146,3881-3888. Van Noesel, C. J. M., Brouns, G. S., van Schijndel, G. M. W., Bende, R.J., Mason, D. Y., Borst, J., and van Lier, R. A. W. (1992).J.E x p . Med. 175, 1511-1519. Venkitaraman, A. R., Williams, G. T., Dariavach, P., and Neuberger, M. S. (1991).Nature (London) 352,777-781. Voronova, A. F., and Sefton, B. M. (1986). Nature (London)319,682-685. Waldshmidt, T. J., and Vitetta, E. S. (1985).J.Immunol. 134, 1436. Webb, C. F., Nakai, C., and Tucker, P. W. (1989). Proc. Natl. Acad. Sci. U.S.A.86, 1977- 1981. Weiss, A., Imboden, J., Hardy, K., Manger, B., Terholst, C., and Stobo, J. (1986).Annu. Rev. Immunol. 4,593-615. Wetzel, G . D. (1991). Cell. Immunol. 137,358-366. Whisler, R. L., Beiqing, L., Grants, 1. S., and Newhouse, Y. G . (1992). Cell. Immunol. 142,398-415. Whitman, M., Kaplan, D. R.,Schaffhausen, B. S., Cantley, L., and Roberts, T. M. (1985). Nature (London) 315,239-242. Wienand, J., and Reth, M. (1991). Eur.1. Immunol. 21,2373-2378. Wienand, J., and Reth, M. (1992).Nature (London) 356,246-248. Wienand, J., Hombach, J., Radbruch, A., Riesterer, C., and Reth, M. (1990). EMBOJ. 9, 449-455. Wilder, J . A., and Ashman, R. F. (1991). Cell. Immunol. 137,514-528. Williams, G. T., Venkitaraman, A. R.,Gilmore, D. J., and Neuberger, M. S. (199O).J.Exp. Med. 171,947-952. Wilson, H. A., Greenblatt, A. D., Taylor, C. W., Putney, J. W., Tsien, R. Y., Finkelman, F. D., and Chused, T. M. (1987).J.Immunol. 138,1712-1718. Witte, P. L., Robinson, M., Henley, A,, Low, M. G., Stiers, D. L., Perkins, S., Fleischman, R. A., and Kincade, P. W. (1987).Eur. J . Immunol. 17, 1473. Yamanashi, Y., Fukushige, S., Semba, K., Miyajima, N., Matsubara, K., Yamamoto, T., and Toyoshima, K. (1987).Mol. Cell. Biol. 7,237-243. Yamanashi, Y., Mori, S., Yoshida, M., Kishimoto, T., Inoue, K., Yamamoto, T., and Toyoshima, K. (1989).Proc. Natl. Acad. Sci. U.S.A.86,6538-6542. Yamanashi, Y., Kakiuchi, T., Mizuguchi, J., Yamarnoto, T., and Toyoshima, K. (1991a). Science 251, 192-194. Yamanashi, Y., Miyasaka, M., Teakeuchi, M., Ilic, D., Mizuguchi, J., and Yamamoto, T. (1991b). Cell Re&. 2,979-987. Yamanashi, Y., Fukui, Y., Wongsasant, B., Kinoshita, Y., lchimori,Y., Toyoshima, K., and Yamamoto, T. (1992).Proc. Natl. Acad. Sci. U.S.A. 89, 1118-1122. Yancopoulos, G. D., Oltz, E. M., Rathbun, G., Berman, J. E., Smith, R. K., Lansford, R. D., Rothman, P., Okada, A., Lee, G., Morrow, M., Kaplan, K., Prockop, S., and Alt, F. W. (1990). Proc. Natl. Acad. Sci. U.S.A. 87,5759-5763. Yellen-Shaw, A. J., and Monroe, J. G. (1992).]. E x p . Med. 176, 129-137. Yellen-Shaw, A. J., Glenn, W., Sukhatme, V. P., Cao, X., and Monroe, J. G. (1991). J . Immunol. 146,1446-1454.
392
NOBUO SAKAGUCHI ET AL.
Ying-zi, C., Rabin, E., and Wortis, H. H. (1991). Int. Zmmunol. 3,467-476. Yu, L. M., and Chang, T. W. (1992).I. Immunol. 148,633-637. Yuan, D., Wilder, J. A., Moore, B. B., and Dang, T. (1992).1.Zmmunol. 148,3652-3657. Ziegler, S. F., Marth, J. D., Lewis, D. B., and Perlmutter, R. M. (1987). Mol. Cell. Biol. 7, 2276-2285. This article was accepted for publication on 10 February 1993.
ADVANCES I N IMMUNOLOGY, VOL. 54
Analysis of B Cell Tolerance in Vitro DAVID W. SCOlT Division of Immundogy, Uniwrsity of Rochostor Cancer Contor, and Dopahnont of Microbiology and Immunology, Uniwrsify of Rochester School of Medicino and Dentistry, Rochostor, Now rork 14642
1. Introduction and Rationale
Distinguishing self from nonself is a hallmark of the immune system. As T cell recognition and provision of help to B cells are critical for responsiveness to virtually all protein antigens, unresponsiveness at the T cell level could be expected to account for self tolerance. Why then would tolerance at the B cell level be necessary? After all, self antigen-binding B cells have been described in normal individuals (Casali and Notkins, 1989; Ternyck and Avrameas, 1986), although there is no evidence that their autoreactive products have pathological consequences. Indeed, for some self antigen-binding cells, it is not clear whether these cells can be activated or are merely anergic. For several reasons, however, tolerance induction in high-affinity clones must be induced to preserve the integrity of the host. Viral infection can induce the upregulation of major histocompatibility antigen expression, which could lead to novel peptides being presented to activate nontolerant T cells. These helper cells would then provide help for dormant self-reactive B cells. Moreover, the development of somatic mutation during the immune response might allow high-affinity antiself B cells potentially to arise. Therefore, B cell tolerance is necessary to prevent the occurrence of autoreactivity, as well as to shape immunologic diversity. Despite the need for purging the B cell repertoire, proof that tolerance in B cells occurs required years of analysis in uitro and awaited definitive proof in transgenic mice. The purpose of this review is not to cover B cell tolerance in general, as this subject has been broadly examined in several articles. (Nossal, 1992; Scott et al., 1993), but rather to chronicle the evolution of our understanding of B cell tolerance through in uitro systems. We then evaluate these studies in terms of our evolving knowledge from transgenic models defining B cell tolerance in uiuo and signal transduction in uitro. 393 Copyright 0 1993 by Academic Press,Inc. All rights of reproduction in any form reserved.
394
DAVID W. SCOTT
B CELLTOLERANCE DEFINED AND THE VALUEOF in Vitro STUDIES By definition, B cell tolerance is a specific hyporesponsive or unresponsive state induced by prior exposure to a given antigen (epitope). Proof of tolerance requires that such an epitope potentially be immunogenic. Thus, an inert hapten or nonimmunogenic peptide is neither tolerogenic nor immunogenic per se, but can be either depending on its carrier association and mode of presentation (use of adjuvants, route of administration, etc.). In fact, it is not clear whether B cell tolerance exists for many self epitopes because significant immunogenic challenge to those epitopes may never be engendered in uiuo. Interestingly, Paul Ehrlich attempted to do just that at the turn of the century in experiments that led him to coin the phrase horror autotoxicus (Ehrlich and Morgenroth, 1901). For decades, the mechanism(s) of unresponsiveness at the B cell level could not be studied. Indeed, it was not until the early 1970s that evidence for B cell tolerance was obtained by Weigle’s group (Chiller et al., 1971). Simultaneously, the ability to culture rodent lymphocytes to elicit antibody responses in uitro was first reported in 1967 by Mishell and Dutton (1967) and simultaneously by Marbrook (1967). These systems allowed the first detailed analysis of B cell tolerance mechanisms. That is, manipulation of culture conditions or the effects of specific metabolic inhibitors on tolerance now could be investigated under more defined in uitro conditions. Although certain caveats must be considered (such as the disruption of the organization of lymphoid tissues and the homeostatic role of lymphocyte recirculation), one could now use in uitro cultures to understand the mechanism(s) of tolerance induction. It is worth noting in introducing this subject that Carrel and Ingebrigtsen demonstrated an immune response in uitro from rabbit lymph node fragments in 1912. They reported then that responsiveness was consistently preceded by phagocytosis of particulate antigens. Ironically, these workers never challenged nonresponsive cultures with phagocytized (immunogenic) antigen as a rigorous test for tolerance. Nearly 60 years later, as a graduate student with Byron Waksman, I began my journey in tolerance by testing this notion (Scott, 1968). II. Model Systems for the Analysis of B Cell Tolerance
A. HISTORY As noted earlier, Carrel and Ingebrigtsen (1912) described the first antibody responses in uitro. Their observation of the occurrence of phagocytosis in cultures that eventually produced “antibody” led
ANALYSIS OF B CELL TOLERANCE in Vitro
395
Fishman (1959) to feed bacteriophage antigen to peritoneal macrophages in uitro to elicit a neutralizing antibody response. Again, challenge of nonresponding cultures directly exposed to phage antigen was never attempted. Using another approach, Diener and Armstrong (1969) reported that mouse splenocytes cultured with high (microgram) doses of SaZmoneZZa adelaide polymerized flagellin were subsequently unable to form antibody to a lower (nanogram), immunogenic challenge. This fulfilled the criteria for the analysis of B cell tolerance in uitro; moreover, because the challenge was with a thymusindependent antigen, there could be no doubt that tolerance was occurring at the B cell level. These classic studies established the kinetics and temperature dependence of tolerance in uitro. That is, unresponsiveness required incubation at 37°C and at least 3 to 6 hours to occur, the first results suggesting that tolerance was an active process involving cellular metabolism (Diener and Armstrong, 1969). Subsequently, Feldmann and Diener (1970) established that immune complexes containing less than a few picograms of antiflagellin IgG were extremely efficient at inducing unresponsiveness in uitro. These studies paved the way toward our recognition of the role of IgG carriers and immune complexes, especially the importance of Fc receptor crosslinking, in tolerance (for review, see Fedyk et al., 1993). We confirmed that tolerance could be induced in rat spleen fragments exposed in uitro to a high concentration of ultracentrifuged intact bovine gamma globulin (Scott and Waksman, 1968); however, the latter studies did not discern whether the reduced IgM and IgG responses observed were the consequence of B cell tolerance or T cell unresponsiveness, as decreased delayed hypersensitivity reactions were also found in recipients of “tolerant” spleen cells (Scott and Waksman, 1968). Subsequent studies, using haptenated IgG tolerogens and challenge with the same hapten on a different carrier, confirmed that B cell tolerance was inducible under these conditions in uitro (Metcalf and Klinman, 1976; Scott, 1976).These studies are now reviewed.
B. HAFTEN-SPECIFIC B CELLTOLERANCE 1 . Splenic Fragment Cultures and Limiting Dilution a . Role of B Cell MaturationaZ Stage Burnet’s (1959)clonal selection theory suggested that lymphocytes
undergo a period of sensitivity to tolerance during development. Thus, one would predict that both B cells and T cells might be rendered tolerant at an early, immature stage. The most likely phase of tolerance susceptibility is that of an IgM+ IgD- B cell that has just rearranged its
396
DAVID W. SCOTT
light chain genes and has emerged from a pre-B cell to an immature B cell stage. Whether this cell has only an obligate tolerogenic pathway is debatable (see later), but it clearly is sensitive to tolerance induction. For example, Nossal and Pike (1978), Cambier et al. (1977), and we (Venkataraman and Scott, 1977, 1979) showed that neonatal rodent B cells had a significantly reduced precursor frequency when exposed to a tolerogen in uitro or in uiuo. The specific acquisition of maturity and “tolerance resistance” was mapped out in elegant studies by Metcalf and Klinman (1976), using a splenic focus technique. This combined in uiuo and in uitro approach proved invaluable for studies on the mechanism of tolerance and is worth describing briefly herein. B cells, like T cells, recirculate and will migrate from blood to lymph after an intravenous injection. Klinman and colleagues injected titered doses of B cells intravenously into lethally irradiated [hemocyanin (Hy) ] carrier-primed recipients (summarized in Klinman, 1976). Injected B cells migrated to and lodged (at least temporarily) in the splenic white pulp within hours. The spleens could then be removed and chopped into 1-mm fragments that maintained macroscopic tissue integrity. As “helper cell” activity was relatively radioresistant and in excess, those hapten-specific B cells that migrated into a given fragment would be in a milieu containing carrier-specific helper T cells. Thus, culture of individual fragments with hapten-Hy conjugates would lead to antibody production by the fragment provided a hapten-specific B cell localized in a given fragment. Poisson analysis of the frequency of positive fragments could then be used to determine the hapten-specific B cell precursor frequency of the expressed repertoire. What Klinman and colleagues did was to culture fragments with hapten on an unrelated carrier for 24 hours prior to challenge with hapten on Hy. This “tolerance induction” step allowed one to measure a number of parameters of the tolerogenesis process, as well as to point out that immature B cells were not obligately tolerance sensitive, at least at this stage of differentiation. That is, if the hapten were provided on a carrier to which the donor spleen was primed (Hy), then an immune response occurred. If not, tolerance was observed. These studies definitively identified the late fetal to early neonatal period as one of high sensitivity to tolerance (Metcalf and Klinman, 1976).Adult B cells were completely resistant to tolerance induction under these conditions; interestingly, the development of resistance was apparent by the end of the first week of murine life (Metcalf and Klinman, 1976), a point at which no other markers of differentiation were apparent, including the expression of IgD (see later).
ANALYSIS OF B CELL TOLERANCE in Vitro
397
Cambier et al. (1977)also found that B cells from neonatal mice were far more tolerance sensitive (>looox) in a completely in vitro approach with both T cell-dependent (TD) and T cell-independent (TI) forms of challenge. In fact, with one TI challenge (trinitrophenylcoupled to Brucella abortus), both adult and neonatal B cells appeared tolerance sensitive. We confirmed these results, but showed that sensitivity to tolerance did not occur with all forms of TI challenge (Scott et al., 1979a). Therefore, it was not generalizable that B cells that responded to TI forms of antigen were more immature and, therefore, tolerance sensitive. It is, however, worth noting that TD responses require cytokines, which, we will show later, can convert tolerogenic to immunogenic signals.
b. lmportance of Affinity and Valence The affinity of the immunoglobulin receptors on a B cell is pre-
dictive of that cell’s ability to capture antigen and be activated or rendered tolerant. It is logical that high-affinity B cells would be more efficient at antigen capture and, hence, susceptible to tolerogenic signals. Therefore, the net effect of tolerance induction would be a lower overall response, as the higher-affinity clones would have been functionally inactivated; however, the appearance of greater susceptibility of higher-affinity B cells to tolerance induction in vivo could be an indirect effect of tolerance at the T cell level because a lack of helper activity might fail to drive the repertoire normally to a higher-affinity response. Hence, it was important to directly measure affinity in B cell tolerance in the absence of T cell effects. Klinman and colleagues (Metcalf et al., 1977) found that tolerance induction at a given molar concentration of hapten led to responsiveness of only lower-affinity clones, thus confirming the in vivo results of Werblin and Siskind (1974). It was also noted that tolerance to dinitrophenyl also led to unresponsiveness to the cross-reactive trinitrophenol hapten, although cross-triggering could not be observed with these two haptens (Metcalf et aZ., 1977). Using haptenated (fluorescein) gelatin to purify B cells, Venkataraman and Scott (1979) further demonstrated that high affinity fluorescein-specific B cells were rendered tolerant prior to exposure to an antigenic challenge (that could drive affinity maturation). That is, hapten-purified B cells of relatively high affinity were directly susceptible to tolerance in vitro. Therefore, the greater sensitivity of highaffinity B cells to tolerance was proven by these in vitrn approaches. Most of the studies in vitro employed multivalent haptenated proteins as tolerogens. This was not only convenient, but also reflected the difficulty of preparing a univalent haptenated tolerogen. Studies by
398
DAVID W. SCOTT
Nossal and Pike (1978), as well as Waldschmidt et al. (1983), provided evidence for the optimal ratio of hapten to carrier to induce tolerance. For immunoglobulin carriers, 6 to 12 hapten molecules per 150,000 Da worked well. A molar ratio of less than 2 to 4 hapten molecules per carrier led to significantly lower tolerogenicity; greater than 12 hapten molecules usually reduced the efficacy of the tolerogen, presumably as a result of denaturation of the carrier. These results were extended by Waldschmidt et al. (1983)for different Ig isotype carriers. The implication of these studies was that a critical level of crosslinking of receptors with a mimimal threshold binding affinity (Teale and Klinman, 1980) was required for optimal tolerance induction. These studies begged the question of whether crosslinking of B cell Ig receptors was required for tolerance induction. Metcalf and Klinman ( 1976) directly tested this hypothesis by preparing dinitrophenol S-papain, a univalent hapten-carrier conjugate, which was totally without activity in their in uitro splenic focus tolerance induction protocol. How, then, do univalently expressed epitopes on protein antigens induce tolerance naturally in uiuo? Or do they? Evidence from transgenic mice, cited later, suggests that univalent epitopes can induce B cell tolerance, but the possibility that such epitopes could be presented on a cellular surface in a multivalent array can not be eliminated. Thus, this point is still moot.
2 . Use of Affinity-Purtjied B cells to Analyze Tolerance The rare frequency of B cells specific for a single epitope made the analysis of the cellular and molecular mechanisms of tolerance a daunting task. That is, it is virtually impossible to study the signaling pathways of rare (<1in 1000)antigen-specific B cells with background noise from over 99.9% of other cells in the absence of some amplifying mechanism (e.g., see Scott, 1973). Before the advent of immunoglobulin transgenic mice in which virtually every B cell expressed the same (transgenic) Ig receptor, this was one of the major challenges in unraveling tolerance mechanisms. Efforts to resolve this dilemma included attempts at the cloning of B cells in uitro (Howard et al., 1981; Aldo-Benson and Scheiderer, 1983),with or without the purification of hapten-specific B cells by affinity techniques (Haas and Layton, 1975). Unfortunately, continued expansion of B cells in uitro proved to be much more difficult than T cell cloning, perhaps because B cell growth factors (such as IL-4 and IL-5) also drive significant terminal differentiation. In addition, other B cell cloning techniques are short-term and may alter the state of activation of the specific B cells or block tolerance induction by the use of mitogenic agents (Pillai and Scott, 1981).
ANALYSIS OF B CELL TOLERANCE in Vitro
399
Nonetheless, purification of hapten-specific B cells still allows one to examine small numbers of B cells from normal or tolerant donors. Probably the most extensively used approach was affinity purification on haptenated gelatin-coated petri dishes (Haas and Layton, 1975), extensively employed by Nossal and colleagues (Nossal and Pike, 1978; Pike and Nossal, 1979; Pike et al., 1982; cf. Nossal, 1992). This group demonstrated by limiting dilution techniques that the precursor frequency for simple haptens, like nitroiodophenyl (NIP), dinitropheno1 (DNP), and fluorescein (FL), could be increased more than 150 to 500-fold by binding to and isolation from specific hapten-coated gelatin. Ironically, these purified hapten-specific B cells required “filler cells” such as thymocytes or irradiated spleen cells to be stimulated for an antibody response! Fortunately, exposure of these hapten-specific cells to a given tolerogen would lead to unresponsiveness to subsequent challenge with filler cells. This permitted limited analysis of the initial signaling events in tolerance, the target maturational stage, and the requirement for (at least) bivalence to induce tolerance. For example, Nossal and Pike (1978) found that 50% of enriched FL-specific neonatal B cells were rendered tolerant by <0.1 pg/ml F L S . ~human gamma globulin (FL-HGG), whereas adult B cells showed no reduction in precursor frequency at 25 pg/ml tolerogen. Subsequently, they observed that 280 pg/ml FL3.6-HGG was required for a 50% reduction in precursor frequency of adult haptenspecific B cells, again revealing a 1000-folddifference between neonatal and adult B cells. Moreover, the presence of lipopolysaccharide during in vitro culture prevented tolerance induction, a result confirming the notion that a “second signal” received with tolerogenic crosslinking can sometimes overcome tolerogenesis. When the tolerogen contained 12 moles of FL per HGG carrier, unresponsiveness in adult B cells was found at 1 pg/ml. Interestingly, Pike et al. (1981) confirmed later that higher hapten: carrier ratios favored tolerance in adult B cells and that intact IgG conjugates were more efficacious than FL-Fab or FL-conjugated to bovine serum albumin. These studies emphasized the unique qualities of IgGs as tolerogenic carriers (Bore1 and Kilham, 1974; Waters and Diener, 1983) and implied that the FC portion of immunoglobulin played a functional role in tolerogenesis (see Section II,C,3). This was carefully analyzed in vitro by Waldschmidt et al. (1983), who not only established that Fc receptor binding IgGs were the most efficient tolerogens, but also that higher epitope densities improved tolerogenicity. Studies with clonable B cells, using the CFU-B technique of Kincade (1978), were undertaken by Pillai and Scott (1981). In these
400
DAVID W. SCOTT
studies, we first found that hapten-specific B cells from tolerant mice formed CFU-B at a normal frequency, a result that would be expected from the strong mitogenic conditions of the cloning procedure; however, when a high concentration of tolerogen (FL-sheep gamma globulin) was included in the agar, a subset of normal adult B cells failed to clone! The presence of macrophages during the cloning procedure allowed this subset to grow. This suggests that certain cytokines, produced by macrophages, may provide a second signal overriding tolerogenic signals in uitro. In this cloning system, those cells that did form colonies in agar in the presence of tolerogen could not be subsequently triggered to form antibody in uitro (Pillai and Scott, 1981). In contrast, Schad and Phipps (1988) found that exposure of purified FL-binding B cells to tolerogen on macrophages facilitated unresponsiveness. This effect was later shown to result from the production of prostaglandin E2 (PGE2) by macrophages, as free tolerogen or immune complexes plus PGEz also led to increased unresponsiveness (Schad and Phipps, 1988; Stein and Phipps, 1989). It is tempting to speculate that under CFU-B cloning conditions, macrophage-produced PGEz is absent or is modulated by other cytokines. Sensitivity to tolerance induction with hapten-purified B cells correlated with the relative affinity of the responding clones, as would be predicted from the in uiuo studies of Werblin and Siskind (1974).Thus, B cells that were both FL-gelatin-purified and FACS-sorted for highaffinity epitope binding were even more sensitive to unresponsiveness then single-cycle FL-gelatin-purified cells (Nossal and Pike, 1978). Confirming the splenic focus data of Metcalf and Klinman (1976), Nossal and Pike (1978), in a suspension culture system, also found that susceptibility to tolerance decreased during the first month of life, although not as rapidly as in the splenic focus system. This result may reflect the effect of thymus-dependent challenge in the latter model and the fact that higher-affinity B cells were “read out” in the former. As IgD expression on maturing B cells increases gradually during the first month of life, a role of IgD in resistance to tolerance may be considered; this is discussed in Section II,B,3. The properties of tolerant hapten-specific B cells have been thoroughly examined and summarized by Chace and Scott (1988a,b),who were able to stimulate FL-binding B cells with a mixture of cytokines in the absence of filler cells. This allowed direct analysis of cell cycle staging and other parameters of B cell responsiveness. Basically, “tolerant” B cells existed in adult mice, as had been reported earlier (Venkataraman and Scott, 1977,1979; Pillai and Scott, 1983),but were
ANALYSIS OF B CELL TOLERANCE in Vitro
40 1
“anergic” in terms of antibody formation and entry into cycle by stimulation by specific antigen (Venkataraman and Scott, 1980; Chace and Scott, 1988a). FL-binding cells from tolerant donors expressed normal amounts of IgM, IgD, and class I1 major histocompatibility complex (MHC) antigens and could be stimulated to express increased class I1 levels and enter GI by antigen or anti-Ig exposure (Chace and Scott, 1988a,b). Early events, such as calcium mobilization and membrane depolarization, as expected, were not defective in tolerant long-term B cell lines (Aldo-Benson, 1986; Aldo-Benson and Tsao, 1987). This firmly established the existence ofthe anergic B cell in adult tolerance, a result elegantly confirmed in transgenic mice (Goodnow et al. 1988). 3. Differential Signaling via Zmmunoglobulins M and D
Antigen specificity in B cells is provided by the expression of surface immunoglobulin receptors. The earliest B cells express only surface IgM, but, as they mature, these cells express both IgM and IgD of the same epitope specificity, that is, with the same V regions (Goding et al., 1977). Based on the exquisite sensitivity of neonatal, immature B cells to tolerance induction in uitro, it was inferred that IgM could deliver only a tolerogenic signal, whereas IgD provided a counteracting immunogenic trigger (Uhr and Vitetta, 1975). The gradual appearance of IgD during the first few weeks of murine life generally correlated with the acquisition of resistance of tolerance in the studies of Nossal and Pike (1978); however, Metcalf and Klinman (1976)found that sensitivity to tolerance induction in the splenic focus system was lost abruptly at 7 days of age, a time when only a few B cells expressed both isotypes. This implies that some IgM-only B cells must be resistant to tolerance. Thus, IgM receptors are not restricted to deliver only tolerogenic signals. This conclusion could also be reached because fetal or neonatal IgM+IgD- B cells exposed to antigen and provided with T cell help produced an antibody response (Metcalf et al., 1977). What, then, is the role of surface IgD? Vitetta and colleagues (1977) and Scott et al. (1977) independently found that the presence of IgD could modulate potentially tolerogenic signals in developing B cells. In the former case, adult B cells were treated with papain, which differentially removed IgD from the surface. These cells showed a dose-response curve for sensitivity to tolerance that mirrored that of neonatal B cells (Vitetta et al. 1977). We directly modulated IgD by treatment of young, adult B cells with an anti-IgD allotype and found that their resistance to an in uitro tolerance induction protocol was shifted to greater sensitivity. Although hindsight might suggest alternative interpretations, these studies are consistent with a
402
DAVID W. SCOTT
differential role for IgD in tolerance induction, a conclusion supported by studies with a polyclonal anti-Ig tolerance model (Gaur et al., 1993) and in uiuo tolerance studies in transgenic mice (Carsetti et al., 1993). This result is even more surprising because the initial biochemical pathways triggered by crosslinking of either IgM or IgD appear to be very similar. Thus, events downstream of these signals must differ for these isotypes; however, these studies do not consider the developmentally regulated expression of specific kinases or their substrates that may modulate B cell responsiveness (Yellen-Shaw et al., 1992; Yao and Scott, 1993; see Sections II1,B and C). The exact functional role of IgD in tolerance requires further study.
C. ANTI-IMMUNOGLOBULIN M SURROGATE SYSTEMS Using Escherichia coli lipopolysaccharide (LPS) as a polyclonal B cell activator, Lawton and Cooper (1974) and Raff et al. (1975) found that anti-immunoglobulin treatment of adult B cells prevented LPSdriven differentiation with no adverse effects on proliferation. In contrast, neonatal (immature) B cells were blocked in terms of both proliferation and differentiation. In fact, in uiuo treatment of developing mice with anti-IgM prevented B cell development and led to depletion of this pool (Lawton and Cooper, 1974). This system, in which anti-Ig serves as a surrogate for specific antigen, has served as a useful model for tolerance in lieu of hapten-specific B cell purification. Thus, crosslinking of surface IgM and/or IgD with anti-Ig would be expected to mimic the effects of epitope binding by immature and mature B cells. 1 . Capping and Resynthesis of Receptors When surface immunoglobulin receptors are crosslinked with antiIg at 37"C, they aggregate and move to one pole of the cell in an energy-dependent pattern known as capping. In adult B cells, capping leads to subsequent shedding or internalization of the receptor complex, presumably to endosomes for processing. This process is followed by reemergence of surface IgM in 4 to 6 hours after de n o w protein synthesis (Greeley et al., 1974). In contrast, neonatal B cells cap sluggishly and fail to reexpress their IgM receptors normally (Sidman and Unanue, 1975). Hence, the greater sensitivity of neonatal B cells to tolerance could be the result of failure of cells at this stage of development, after initial antigen contact, to reexpress the surface IgM needed to be triggered by antigen. This phenomenon fails to explain the loss of LPS responsiveness, however, as mitogenicity is independent of surface Ig expression on B cells. Indeed, adult B cells, treated
ANALYSIS OF B CELL TOLERANCE in Vitro
403
with anti-IgM overnight and then washed are also unresponsive to LPS challenge in terms ofantibody synthesis (Warner and Scott, 1991), as will be reviewed in Section II,C,3. In the latter studies, surface IgM levels of anergic B cells are slightly reduced, whereas IgD expression is normal. This phenotype resembles that of circulating anergic B cells in 1ysozyme:antilysozyme double transgenic mice (Goodnow et al., 1988). In the in vitro system of Diener and colleagues (Diener and Armstrong, 1969), tolerance induction with high concentrations of polymerized flagellin led to the appearance of antigen-binding B cells whose membranes appeared “frozen” and unable to cap with immunogenic flagellin concentrations (Diener and Paetkau, 1972).This pattern may be an artifact of the highly polymerized nature of the tolerogen and may be typical of paralysis with polysaccharide antigens, as lack of capping has not been observed with other protein tolerogens such as IgGs (Venkataraman and Scott, 1977). 2 . Role of Apoptosis in Neonatal B Cell Tolerance As noted earlier, treatment of either neonatal or adult splenic B cells with anti-IgM leads to unresponsiveness to challenge with either LPS or a specific antigen (e.g., FL-Brucella abortus). In the former case, little B cell proliferation to LPS is observed (Raff et al., 1975; M. Borrello and D. Scott, unpublished), a result that suggests that immature B cells are either eliminated or functionally inert to stimulation. In support of deletion, it was shown that mice treated in vivo from birth show a depletion of B cells in their peripheral lymphoid organs (Lawton and Cooper, 1974; Raff et al., 1975).We reinvestigated this process and found that in vivo or in vitro treatment of neonatal spleen cells with anti-IgM leads to a loss of IgM+, B220+ B cells (Scott et al., 1992; Brown et al., 1992). By flow cytometric analysis of the forward versus 90-degree light scattering properties of treated cells, it could be discerned that a significant subset of neonatal B cells displayed apoptotic bodies typical of programmed cell death (Brown et al., 1992). Presumably, both neonatal IgM+ B cells and emerging pre-B cells, as they nascently express surface IgM, receive a signal that ultimately results in the initiation of cell death via apoptosis. The dose dependence of this process was followed carefully in vitro by Pike et al. (1982). Using FACS-sorted IgM- pre-B cells, they first observed that up to 40% of these would express IgM within 48 hours of culture. In the presence of 1to 10 pg/ml monoclonal anti-p, the emergence of surface IgM+ B cells was almost completely blocked; lower concentrations had little effect on surface IgM expression. Function-
404
DAVID W. SCO’lT
ally, however, cells treated with <1 Fglml anti-IgM were anergic to LPS in terms of both proliferation and differentiation. Therefore, at low concentrations of antigen, emerging neonatal B cells might receive a negative signal that renders them anergic to even mitogenic triggering or they are deleted in the presence of sufficient antigen (or its surrogate), presumably via apoptosis.
3. Anti-immunoglobulin M-induced Unresponsiveness in Adult B Cells Consistent with the failure of adult B cells to be activated by LPS in the presence of anti-IgM (Raff et al., 1975),we previously reported that preculture with anti-IgM has the same effect (Warner and Scott, 1991). This system has allowed the separation of a putative tolerance induction step from the subsequent challenge, as in classical tolerance protocols. Anergy in terms of B cell differentiation was anti-Ig dose dependent, required bivalent antibody, and needed at least a 4-hour incubation at 37”C, analogous to the clonal “abortion” studies of Nossal and colleagues (Nossal and Pike, 1978; Pike and Nossal, 1979; Pike et al., 1982)and the original flagellin tolerance model (Diener and Armstrong, 1969). Interestingly, adult B cells were rendered unresponsive to LPS challenge by either F(ab’)g or intact IgG anti-mouse Fab or anti-mouse IgM; tolerance to antigen-specific (FL-Brucella or FL-LPS) challenge required not only an interaction with IgM, but also Fc receptor crosslinking (Warner and Scott, 1991). The role of Fc receptor downregulation of B cell signaling will be considered later; at this point, it is sufficient to note that intact anti-Ig fails to stimulate a sustained increase in intracellular calcium and leads to abortive phosphatidylinositol (PI) hydrolysis (Bijsterbosch and Klaus, 1985). Adult B cells express both IgM and IgD, with a higher surface density of the latter isotype. Gaur et al. (1993) recently examined the roles of IgM and IgD in our surrogate model for tolerance. We found that both polyclonal and monoclonal anti-IgM were consistently tolerogenic in the same in uitro protocol used by Warner and Scott (1991). In contrast, anti-IgD failed to induce unresponsiveness in adult B cells under identical culture conditions (Gaur et al., 1993). This is an interesting result considering the previous studies suggesting that acquisition of IgD may signify a “tolerance-resistant” state (Vitetta et al., 1977; Scott et al., 1977), typical of adult B cells with a TD challenge (Metcalf and Klinman, 1976). Note, however, that a multivalent antigen should theoretically be able to crosslink neighboring IgM and IgD molecules on the same adult B cell surface. In fact, anti-Fab should have the same surface Ig crosslinking potential, although this would
ANALYSIS OF B CELL TOLERANCE in Vitro
405
depend on the actual density and array of each isotype on the surface. A hindsight interpretation of the studies in which IgD was removed is that the modulating effect of IgD on tolerance signaling was eliminated by enzyme treatment (Cambier et al., 1977)or capping (Scott et al., 1977), so that the IgM signal not only dominated, but was exaggerated. When both anti-IgM (at limiting doses) and anti-IgD were added in this model, the result was surprising. Gaur et al. (1993) found that unresponsiveness was enhanced by crosslinking both isotypes independently or even sequentially. That is, crosslinking both receptors could lead to synergy in tolerance. One interpretation of these results is that crosslinking both isotypes leads to an accumulation of second messengers not achieved with a limiting amount of anti-IgM only. Anti-Fab or antigen might preferentially crosslink IgM to IgM or IgD to IgD depending on the surface density and array of each isotype on a given B cell. It should be noted that these experiments examined only the effects of crosslinking IgM to IgM and IgD to IgD, but not IgM to IgD! It is hoped that evaluation of the biochemical consequences of signaling with these reagents independently or together will unravel this mystery.
4 . Tolerance in Other B Cell Subsets Unlike T cells, there are few definitive markers for different stages of B cell differentiation other than surface 1gM:IgD density. The major exception is the CD5 surface antigen, normally found on T cells, but expressed at low levels on a subset of B cells (Hayakawa et al., 1984). Murine B cells that express CD5 (Le., the so-called €31 subset) are found predominantly in the peritoneal cavity, are rare in the spleen (<5% in normal strains), and are virtually absent from lymph nodes. Interestingly, the neonatal splenic B cell population, although small in absolute numbers, is approximately 50% CD5'. Moreover, it has been suggested (Kearney and Vakil, 1986) that these CD5+ B cells play a role in immune regulation by producing antibodies that are polyreactive and/or express anti-idiotype specificities. It was therefore important to determine the sensitivity of these cells to tolerance induction. Using the anti-Ig preculture approach of Warner and Scott (1991), Liou et al., (1992) tested the sensitivity of peritoneal B cells (50% CDS+) and found that this population was resistant to this tolerance induction protocol relative to adult splenic B cells, which were sensitive under the same conditions. In over 20 experiments, anti-Ig treatment reduced splenic B cell responsiveness to LPS by 77 to 83%, whereas peritoneal B cells were inhibited only 3 to 16%.On the basis
406
DAVID W. SCOTT
of cell mixing experiments, this effect could not be explained by contaminating macrophages or other cells (Liou et al., 1992),but appeared to reflect an intrinsic defect in signal transduction. For example, peritoneal B cells were unable to respond to challenge with LPS or specific antigen after treatment with phorbol esters, which directly activate protein kinase C (PKC) and by pass surface Ig receptors. Moreover, these B1 cells displayed a reduced calcium mobilization response to anti-IgM in uitro. Although further studies are necessary, these results suggest that peritoneal B cells may be deficient in a pathway regulated by PKC activation. Evidence for altered PKC activation in B1 cells recently has been obtained by Cohen and Rothstein (1991). The relative resistance of this CD5-enriched population to tolerance contrasts with the exquisite sensitivity of neonatal B cells (half of which are CD5+)to unresponsiveness and apoptosis; however, it is not known whether it is the CD5+ or the CD5- B cells in the neonatal spleen that are being rendered tolerant by this treatment, as there have been no attempts at separation in these experiments. Nonetheless, further data are consistent with the notion that B1 subset (at least in the adult) is relatively tolerance resistant. First, spleen cells from motheaten mice contain large numbers of CD5+ B cells and are prone to autoimmune diseases; these B cells are also.tolerance resistant (L-B. Liou and D. W. Scott, in preparation). Second, transgenic mice expressing the rearranged IgM receptor from a CD5+ B cell lymphoma not only possess an increased percentage of CD5+ B cells in the spleen, but also are relatively resistant to this tolerance induction protocol (Liou et al., 1993).This suggests that the expressed receptor is restricted to a particular B cell lineage and, therefore, functional behavior pattern. Examination of CD5 expression and VH gene utilization in Ig transgenic mice should prove informative (cf. Goodnow, 1992). It should be noted that Cong and co-workers (1991)found that normal splenic (CD5-) B cells could be induced to express the CD5 marker by culture for several days in uitro with high concentrations of anti-IgM. This was not due to preferential expansion of CD5+ B cells. On the basis of our previous experience (Warner and Scott, 1991),one might expect these anti-Ig cultured B cells to be anergic to further stimulation (for antibody production) via surface Ig or LPS, although this has not been formally tested. Alternatively, these newly derived CD5+ B cells may be altered in their responsivity to anti-Ig and could resemble peritoneal B1 cells. We examined this hypothesis by testing the tolerance sensitivity of B cells primed either in uivo or in vitro with haptenated Ficoll, a multivalent T cell-independent antigen that might be expected to crosslink surface Ig and stimulate the expression
ANALYSIS OF B CELL TOLERANCE in Vitro
407
of CD5. Yao and Scott (1992) found that these primed B cells were completely resistant to anti-Ig-induced tolerance in an epitopespecific manner. That is, FL-Ficoll-primed splenic B cells were resistant to tolerance induction by IgM crosslinking in terms of anti-FL, but were sensitive when anti-trinitrophenol (a specificity for which they were unprimed) was measured. It has not been determined yet whether the FL-Ficoll-primed B cells now express the C D 5 antigen or other markers of the B1 subset. Nevertheless, their signaling properties have been altered by this exposure to hapten in a multivalent array. Identification of these processes is under investigation. I n contrast to these results with TI antigen-primed B cells, Linton et al. (1991) observed that secondary B cells may possess a window of exquisite sensitivity to tolerance normally missing from the adult B cell repertoire. Using the relatively low expression of the J l l d (heatstable antigen) as a marker for secondary B cells, they found that shortly after T D antigen priming in uitro, Jlld'"" precursors for secondary IgG responses were rendered unresponsive by hapten on an unrelated carrier. This window of sensitivity to tolerance is predictable from the need to purge the evolving secondary repertoire of potential antiself specificities generated from somatic mutation after initial antigen contact. Interestingly, Yao and Scott (1992) did not find that T D antigen could prime for resistance to tolerance at any time. Therefore, the development of resistance may be a property of the method of B cell priming and possible (TI) activation of the CD5+ subset of B cells, whereas TD priming opens a window of tolerance sensitivity to purge somatically mutated antiself reactivities (Linton et az., 1991).
5 . Abortive Signaling and Second Signals in Adult B Cells As shown earlier, treatment of splenic B cells with anti-Ig has been used as model for unresponsiveness, which is dose dependent, leads to anergy with adult cells, and can cause deletion with neonatal splenocytes. The initial consequences of Ig receptor crosslinking have been extensively studied. Ironically, this system has been used as a model for B cell activation, as all B cells, by definition, bear surface immunoglobulin, and signaling events in epitope-specific B cells via antigen can be extrapolated from treatment with anti-Ig. In mature B cells, crosslinking IgM or IgD rapidly initiates tyrosine phosphorylation of a number of substrates, activation of phospholipase C-y, PI hydrolysis, PKC activation, mobilization of intracellular calcium, as well as the influx of extracellular Ca2+ (cf. DeFranco, 1987; Cambier and Campbell, 1992; Yao and Scott, 1993, for references). Although immature B cells in the neonatal spleen and adult splen-
408
DAVID W. SCOTT
ocytes treated with anti-IgM produced similar changes in C a y , the former cells fail to hydrolyze inositol phospholipids (Yellen-Shaw and Monroe, 1992).Coupled with the lack of an unidentified 56-kDa kinase (Yellen-Shaw and Monroe, 1992), it appears that immature cells are defective in an anti-Ig-driven second-messenger signaling process and, therefore, fail to be activated productively into cycle in the absence of T cell-derived cytokines. Thus, the most likely pathway for immature B cells encountering antigen is an abortive signaling process, which may be synonomous with tolerance. Does abortive/incomplete signaling correlate with tolerance induction in mature B cells? First, both F(ab’)zfragments ofanti-IgM and anti-IgD can drive adult sglenic B cells productively into the GI phase of the cell cycle, yet only the former produces unresponsiveness (Gaur et al., 1993). Moteover, intact anti-Ig against both of these isotypes abortively stimulates PI hydrolysis (Bijsterbosch and Klaus, 1985)with differential tolerogenic properties. Therefore, it is necessary to look downstream of calcium mobilization 2 phosphatidyl inositol hydrolysis to understand the pathways of adult B cell tolerance. An obvious distal barrier to cell cycle progression beyond Go to G1 entry is GI into S. To approach this question, we (Warner and Scott, 1989; Warner et al., 1991a)employed inhibitors such as cholera toxin and cyclosporin A (CSA), which block anti-Ig-stimulated B cell entry into S at different points. Cholera toxin effectively inhibits B cell stimulation into cycle by anti-Ig, but has little effect on class I1 increases or tolerance induction (Warner et al., 1989). In contrast, CSA blocks both entry into S and tolerance induction in a dose-dependent fashion. Cyclosporin A inhibits B cell as well as T cell activation, although its targets in B cell stimulation are unclear. It is known that CSA prevents the transcription of class I1 MHC antigens driven by anti-Ig, whereas cholera toxin does not. This suggests that the transcriptional events in tolerance induction and in upregulation of class I1 expression may have common elements. An overview of these and related results is provided in Table I (Warner et al., 1991a). Interestingly, several groups (Yuan, 1987; Flahart and Lawton, 1987a,b; Martensson et al., 1989; Chen et al., 1991) have found that anti-Ig treatment of adult B cells decreases the steady-state levels of p heavy-chain RNA via a trans-acting element (Flahart and Lawton, 1987a).Identification of the 5’ regulatory regions affected by anti-Ig in the p heavy-chain regulatory region and upstream of class I1 will, therefore, be of importance. The involvement of common transcription factors, such as NFKB,AP-1, and OTF-2, may provide clues (Lenardo and Baltimore, 1990; Chen et al., 1991; Chiles and Rothstein, 1992).
TABLE I LAT TI ON SHIP BETWEEN PRODUCTIVE OR ABORTIVEB CELL ENTRY INTO CYCLE, CLASS 11HYPEREXPRESSION, AND THE INDUCTION OF HYPORESPONSIVENESS IN ADULT B CELLS"
Stimulus F(ab'), anti-Ig Phorbol myristate acetate + ionomycin IgG anti-Ig Concanavalin A Phorbol myristate acetate Ionomycin F(ab')Z anti-Ig + CT F(ab')z or IgG anti-Ig + cyclosporin A
Entry into cycle Yes Yes Abortivec Abortive Abortive Abortive Abortive Abortive
DNA Synthesis
MHC Class I1 Hyperexpression
Yes Yes No No No No No No
Yes Yes Yes Yes Yes Yes Yes No
B cell hyporesponsivenessb FL-BA
LPS
No Yes Yes
Yes Yes Yes
Yes No
Yes Yes Yes No
Yes -
No
Yes
Modified from Warner et al. (1991a) with permission. Hyporesponsiveness on challenge with either tluorescen-Brucello abortus (FL-BA) or lipodysaccharide (LPS). Data are from Warner and Scott (1989, 1991)and Warner et al. (1991a). 'Abortive cell cycle entry includes exit from Go but failure to proceed through GI and into S.
410
DAVID W. SCOTT
The data from this in oitro system with anti-Ig as a surrogate for antigen are consistent with the hypothesis that abortive cell cycle progression (into GI) is a critical step in B cell activation that leads mature B cells toward a tolerogenic pathway, provided that T cellderived cytokines are not provided. This was evident from the studies of Warner and Scott (1991), who could prevent tolerance by pretreatment with IL-4; however, the provision of T-cell help in the form of IL-4 does not convert a tolerogenic signal per se if given concomitantly with anti-Ig (Warner et al., 1989). This result rather may indicate the need for direct T cell contact to prevent tolerance. Clearly, signals that lead to increased class I1 MHC expression allow for T cell-B cell interactions that promote cell cycle progression and differentiation toward antibody synthesis. A unifying hypothesis for positive signaling and tolerance in adult B cells would suggest that signal 1 (oligovalent antigen or anti-Ig crosslinking) alone can be tolerogenic unless signal 2 (T cell interactions) is provided. D. B CELLLYMPHOMA MODELS B cell lymphomas are considered to be the neoplastic counterparts of normal B cells. Although they are transformed, many of these lymphomas functionally respond to anti-Ig or anti-idiotype treatment. Indeed, these lymphomas have proven valuable for the study of signal transduction pathways and for the transfection of known genes that might alter these pathways (reviewed in Yao and Scott, 1993; Scott et al., 1993a,b). Since 1982, several groups have used a series of B cell lymphomas as models for tolerance because their growth can be inhibited by anti-Ig treatment (Ralph, 1979; Boyd and Schrader, 1981; DeFranco et al., 1982; Pennell and Scott, 1986; Scott et al., 1985) and growth arrest overcome by T cell help (Scott et al., 1989; Valentine and Licciardi, 1992). These studies have shown that growth inhibition reflects the arrest of these cells at the G1:S interface secondary to a crosslinking event in early G1 (Scott et al., 1986);thereafter, these cells die via programmed cell death or apoptosis (Benhamou et al., 1990; Hasbold and Klaus, 1990; Warner et al., 1991b).Therefore, these lymphomas allow us to examine in uitro the events that can be assumed to accompany clonal deletion via apoptosis in uiuo.
1 . Growth Arrest and Apoptosis lnduced by Anti-immunoglobulin M Typical of the B cell lymphomas that are growth arrested at G1:S by anti-IgM are WEHI-231 and CH31; these cells express IgM and generally have neither surface IgD nor 6 chain message, as is the case with
ANALYSIS OF B CELL TOLERANCE i n Vitro
41 1
neonatal splenocytes. When exponentially growing WEHI-231 and CH31 cells are treated with anti-IgM, they decrease in cell size and gradually accumulate in late GI over the next 24 to 48 hours. Laddering of their DNA and an increase in the appearance of nuclear apoptotic bodies also are increasingly apparent during this time interval (Benhamou et al., 1990; Hasbold and Klaus, 1990; Warner et al., 1991b; Fischer et al., 1993a). These studies suggest that apoptosis correlates chronologically with cell cycle arrest, but they do not prove that one is required for the other. Apoptosis in lymphomas is considered to be a model for deletional B cell tolerance in the neonatal period of development. As stated earlier, the deletion of B cells during development can be achieved by antiIgM treatment initiated at the neonatal period (Lawton and Cooper, 1974). The appearance of a subset of apoptotic cells in neonatal splenocytes treated with anti-IgM has been reported by Brown et al. (1992), although this is difficult to detect by standard techniques in vitro because of the high level of background apoptosis seen in control cultures (Borrello and Scott, unpublished). The appearance of apoptosis in transgenic models for deletional tolerance should directly confirm these studies with anti-IgM in neonatal splenocytes and in lymphomas; these are discussed later. Our working hypothesis has been that engagement of B cell surface Ig receptors at a critical point in G1 leads to cell cycle arrest (Scott and Klinman, 1987); such arrested ceIIs commit to an apoptotic pathway if an appropriate second signal is not delivered shortly thereafter.
2. Mechanism of Cell Cycle Controls in Tolerance a . myc, Transforming Growth Factor p, and p R B Evidence that the cell cycle is strictly controlled by the cyclins and a variety of oncogenes and anti-oncogenes has accumulated during the last few years (reviewed in Draetta, 1990; Hunter and Pines, 1991). Using B cell lymphomas as models for tolerance, several laboratories have focused on the roles of the cellular oncogene c-myc, the growth competence gene egr, as well as anti-oncogenes such as the retinoblastoma gene product pRB, as critical elements in the tolerance process leading to apoptosis (McCormack et al., 1984; Seyfert et al., 1990; Warner et al., 1991b; Mashesaran et al., 1991). For example, Sonenshein and colleagues (McCormack et ul., 1984; Mashesaran et ul., 1992) demonstrated that c-myc transcription increased within 30 minutes in the growth-inhibitable WEHI-231 B cell lymphoma treated with anti-IgM and that myc protein appears to be phosphorylated within 1 hour under these conditions. Message levels for c-myc de-
4 12
DAVID W. SCOTT
crease thereafter to undetectable levels within 8 to 24 hours of treatment. As myc protein is believed to function as a cell cycle-controlling transcriptional element (Spencer and Groudine, 1991; Blackwood and Eisenman, 1991; Hekkila et al., 1987), the changes in c-myc observed in WEHI-231 are assumed to be causative for growth arrest. In contrast, Tisch et al. (1988), using IgD-transfected WEHI-231 cells, found that anti-IgD also induced the same initial rise, but not the subsequent loss, of c-myc message. Consistent with the observation that IgD does not transmit inhibitory signals in normal (Gaur et al., 1993) and transformed B cells (Tisch et al., 1988; Ales-Martinez et at., 1988),these data support the idea that myc protein levels are required for cell cycle progression in these cells. How myc protein regulates cell growth is still unclear, but recent evidence suggests that this factor plays a role in apoptosis (Evan et al., 1992; Shi et al., 1992). This is reviewed in Section II,D,2. Cognizant of the fact that anti-IgM fails to drive neonatal B cells productively into cycle, Monroe and co-workers (Seyfert et al., 1989, 1990) examined the role of the growth competence gene egr in both immature normal B cells and different lymphomas. They found that egr transcription was induced in adult but not neonatal or bone marrow IgM- B cells (Seyfert et al., 1989) or two immature (growthinhibitable) lymphomas. Although this could be due the result ofa lack of p56-dependent kinase activity, it was also found that the egr gene was hypermethylated in WEHI-231 cells, thus blocking its transcription. Treatment of these lymphomas with 6-methyl cytosine to prevent hypermethylation led to egr transcription (Seyfert et at., 1990). Whether this reversed anti-IgM inhibition of these lymphomas (and prevented myc downregulation) or promoted neonatal B cell cycle entry was not reported. The 110-kDa retinoblastoma gene product pRB is a nuclear phosphoprotein that is a potent anti-oncogene regulating cell cycle progression (DeCaprio et al., 1989; Buchkovich et al., 1989; Mihara et at., 1990; Goodrich et al., 1991).pRB is phosphorylated during mid- to late G1 and remains in that state until G2:M, when a PP1-type phosphatase dephosphorylates it (Ludlow et al., 1990; J. Ludlow, personal communication). In fact, phosphorylated pRB is required to cross the G1:S border. The state of phosphorylation of pRB is also modulated by transforming growth factor (TGF)-/3,a cytokine that is a potent growth regulator in many cell types (Laiho et al., 1990; Massagub, 1990). It is significant that TGF-/3-treated cells are arrested at the G1:S border and possess underphosphorylated pRB. What are the roles of TGF-P and pRB in tolerance? Using the WEHI-231 B cell lymphoma, we (Warner
ANALYSIS OF B CELL TOLERANCE in Vitro
413
et al., 1991b) found that active TGF-P was produced by B cell lymphomas treated with anti-IgM and that pRB was underphosphorylated in these arrest cells, in agreement with the work of Sonenshein’s group (Mashesaran et al., 1991). These studies have been reproduced with human Burkitt lymphoma cells that are inhibitable by anti-IgM (S. Kent, L. Joseph, and D. W. Scott, unpublished). Moreover, further studies with synchronized lymphoma cells (Fischer et al., 1993) demonstrated that the underphosphorylated pRB was the cause, not the result, of cell cycle arrest. This suggests that an investigation of the production of TGF-P, or a cytokine like it, and the state of pRB phosphorylation in developmental tolerance models would be fruitful. Interestingly, Shull et al. (1992) found that TGF-P knockout mice develop a wasting syndrome resembling systemic autoimmune diseases. Analysis of B cell (and T cell) tolerance in these knockout mice is in progress (C. Sidman, personal communication).
b. Blocking Negative Signaling with Antisense myc Finally, how do these findings relate to apoptosis? The role ofmyc in apoptosis was tested using antisense oligonucleotides by Fischer et al. (1993). Based on the observation of Evan et al. (1992) that accumulation of myc protein at cell cycle borders signals apoptosis in rat fibroblasts, we treated WEHI-231 or CH31 cells with antisense oligos for the first coding sequence of c-myc and found that both growth arrest and apoptosis induced by anti-IgM (or TGF-P) were prevented. Moreover, this effect was c-myc specific, as neither nonsense myc nor antisense fos had little effect on thymidine incorporation and none on apoptosis. Antisense myc-protected cells were also able to phosphorylate pRB normally (Fischer et al., 1993). Because Northern analysis showed that myc message was stabilized by antisense myc, it is believed myc protein could play a pivotal role in this model of tolerance and cell cycle control. Cohen and Rothstein (1991)used subtractive hybridization to isolate several candidate clones that were specific for the apoptosis process. We used antisense ologonucleotides for one of these, RP-8 (Owens et aZ., 1991), to determine if it was critical in apoptosis induced by antiIgM in these lymphomas. Unfortunately, we found that antisense RP-8 oligos neither prevented apoptosis, nor blocked cell cycle arrest in anti-IgM-treated B-cell lymphomas (G. Fischer and D. Scott, unpublished data, 1993). These data are consistent with the hypothesis that myc protein plays an important role in the growth arrest process that leads to apoptosis, but that putative apoptosis genes may not be transcribed in this process. Similar results have been obtained by Honjo’s
4 14
DAVID W. SCOTT
group (Ishida et al., 1992).It will be important to test the roles ofthese products in normal B cell tolerance models, especially in comparison to the anti-apoptotic properties of the bcl-2 gene product (Strasser et al., 1991).
E. PRELIMINARY STUDIES WITH TRANSGENIC SPLEEN CELLS in Vitro The ability to follow the fate of antigen-specific B cells during tolerogenesis had been a major impediment to molecular studies of this process. The alternative models (hapten-specific B cell isolation, antiIgM as a surrogate for antigen, and B cell lymphomas) all have inherent advantages and disadvantages. During the last 5 years, transgenic mice with rearranged immunoglobulin receptors, as well as the creative use of both transgenic and normal mice expressing a potential tolerogen, have enormously aided progress in the field of tolerance. As these studies have been extensively reviewed (Nossal, 1992; Goodnow, 1992, Scott et al., 1993),my goal herein is only to relate these studies to the in uitro models cited earlier and to summarize new evidence obtained with transgenic B cells exposed to a tolerogenic stimulus in uitro. The major accomplishment of transgenic models for B cell tolerance was to silence the naysayers (who denied the existence of B cell tolerance), as well as to provide definitive proof of both anergy and deletion. These studies involved the exposure of B cells expressing the rearranged transgenic Ig receptors in uivo to a given antigen during development, acutely in radiation chimeras or following upregulation in adults (Nemazee and Burki, 1989; Goodnow et al., 1988,1989).The phenotypic properties and in uitro responsiveness of anergic B cells have been thoroughly analyzed by Goodnow and colleagues (1988), whose observations were confirmed by Warner et al. (1991a) in the anti-IgM model called “the poor man’s transgenic.” What are the results of exposing transgenic B cells to a putative tolerogen in uitro? Predictably, the initial signals following exposure to antigen resemble those recorded for anti-Ig crosslinking (C. Goodnow, personal communication; D. Nemazee, personal communication). Although the detailed downstream events have not been reported, definitive evidence of apoptosis has been observed in two of three deletional models (Carsetti et nl., 1992; Murakami et al., 1992). On the other hand, Tiegs et al. (1993) found a high incidence of rearrangements in “tolerant” transgenic mice, a novel pathway suggesting an escape mechanism in deletional self tolerance. It will be important to use these systems in a thorough analysis of the signaling pathways from receptor ligation to apoptosis (or anergy) in uitro. Whether rearrangements following tol-
ANALYSIS OF B CELL TOLERANCE in Vitro
415
erogenic stimuli occur in general needs confirmation, but evidence cited in Section II1,A is consistent with this as a tolerogenesis escape pathway. That is, cells that cannot rearrange are deleted; those that rearrange, survive. Finally, a caveat about the use of transgenics to study B cell tolerance: By their very nature, these mice have limited endogenous immunoglobulin rearrangements. It is possible that normal repertoire purging steps may be silenced in these mice, and moreover, development of B cell subsets may b e limited. Thus, it is not surprising that many Ig transgenics have small spleens and fewer B cells (summarized in Goodnow, 1992). Indeed, in some cases, a skewing toward the B1 (CD5) subset has been observed (S. Clarke, personal communication). As tolerance susceptibility may vary with a given subset (see Section I17C,2), studies in Ig transgenic mice must be kept in perspective. Nonetheless, it is fascinating that virtually all of the systems for analyzing B cell tolerance in uitro have revealed a multiplicity of pathways for B cell tolerance (Scott et al., 1979b). 111. BiochemicalAnalyses of B Cell Tolerance Signal Transduction Pathways
A. USEOF INHIBITORS TO BLOCK in Vitro TOLERANCE The major advantage of in uitro models for tolerance is the ability to manipulate the conditions leading to unresponsiveness. Inhibitors of DNA, RNA, and protein synthesis, as well as agents that block membrane receptor interactions or mimic second-messenger pathways have all been used. The first definitive approach was taken by Desaymard (1981), who studied tolerance induced in adult splenocytes cultured with DNP- or TNP-coupled polysaccharide tolerogens. Desaymard (1981) found that cytochalasin B (CB), Vinblastine (VB), and colchicine (CC), as well as lanthanum or zinc chloride (LZC) and sodium azide, blocked tolerance induction. All of these reagents interfere with receptor capping or membrane fluidity (CB, VB, CC, LZC) and energy-dependent internalization (azide). It should be noted that this system requires multivalent crosslinking of receptors and may be highly dependent on these membrane events. Moreover, reagents that affect microtubules or microfilaments may also interfere with mitosis; indeed, energy is needed for cell cycle progression, not just receptor redistribution. In contrast, Teale and Klinman (1984) used many of the aforementioned reagents and found that these inhibitors did not prevent neonatal B cell tolerance in the splenic focus technique,
4 16
DAVID W. SCOTT
which measures a T cell-dependent response. Rather, they found that 3-deaza-adenosine, a methylation inhibitor affecting microfilament activity, was the only receptor-targeted reagent effective at blocking tolerance. Importantly, inhibitors of RNA and DNA synthesis, but not cell division, also prevented unresponsiveness. These data are consistent with tolerance involving cell cycle progression up to and including DNA synthesis. It is also possible that Ig gene rearrangements may be prevented by these drugs, a result consistent with the data of Tiegs et al. (1993). As noted earlier, Warner and colleagues (Warner and Scott, 1989; Warner et al., 1991a) found that cyclosporin A, but not cholera toxin, inhibited tolerance induction by anti-IgM in adult B cells. These data imply that the (abortive) cell cycle progression may lead down a tolerogenic path, provided T cell help is absent. Finally, it is worthwhile to consider the initial signal transduction steps leading to tolerance. As stated earlier, crosslinking of surface Ig receptors leads to well-described second-messenger events. Two critical steps are the tyrosine phosphorylation of a number of substrates and inositol phospholipid metabolism, leading to increases in intracellular calcium and PKC activation. Recent data suggest that protein tyrosine kinase (PTK) activity is required for tolerance induction in model systems with normal adult B cells and with human B cell lymphomas (Gaur et al., 1993; Beckwith et al., 1991). For example, antiIgM-induced unresponsiveness was prevented in a dose-dependent fashion by pretreatment of adult splenocytes with tyrphostin, a PTK inhibitor (Gaur et al., 1993). Moreover, not only did herbimycin A, another PTK inhibitor, block growth arrest in a human B cell lymphoma, but phosphatase inhibitors augmented the arrest of these cells (Beckwith et al., 1991).These studies implicate tyrosine phosphorylation as a pivotal initial event in tolerance induction, as it is in positive B cell stimulation (Cambier and Campbell, 1992). Whether these two processes ultimately involve different substrate targets is likely, but still unknown. It should, however, be noted that Warner and Scott (1988)could not block unresponsiveness in B cell lymphomas with PKC inhibitors, but rather found that phorbol esters bypassing surface Ig can mimic unresponsiveness in adult B cells. These data imply that PKC activation may play a role in some, but not other forms of tolerance. On the other hand, Warner and Scott (unpublished) found that prolonged treatment with H7 (but not H8) led to enhanced responsiveness to specific antigen. This suggests that PKC activation does function in downregulating antibody responsiveness via the Ig receptor complex. Table I1 summarizes the data from these and other studies, which indicate that
TABLE I1 INHIBITORS THATAFFECTB CELLTOLERANCE INDUCTION in Vitro Effect on in oitro tolerance Inhibitor
Neonatal"
Adule
Interpretation of step blocked
Lanthanum or zinc chloride Chlorpromazine Cytochalasins Sodium wide Tyrphostin, herbimycin H7, H8 Cyclosporin A Cordycepin Antisense myc Puromycin Hydroxyurea, cytosine arabinoside
Not done None None Inhibition Inhibitiond Noned Not done Inhibition Inhibition Inhibition Inhibition
Inhibition Inhibition Inhibition Inhibition Inhibition See text Inhibition Not done Not done Not done Not done
Receptor redistribution Receptor redistribution Receptor distribution Energy 2 receptor redistribution Tyrosine phosphorylation Protein kinase A, C Transcription,G1progression Transcription Translation of myc Translation DNA synthesis
Reference'
1
12
1-2
1,2
3,4
5,6
5 2
7 2
2
Splenic focus technique. Using haptenated polysaccharidesor anti-lgM. (1) Desaymard (1981); (2) Teale and Klinman (1984); (3)Beckwith et al. (1991);(4) Gaur et al. (1993);(5)Warner et 01. (1991a);(6)Warner and Scott( 1988, 1991); (7) Fisher et al. (1993). Assuming growth arrest of B cell lymphomas represents neonatal tolerance.
418
DAVID W. SCOTT
adult tolerance and neonatal tolerance appear to have different requirements and, ultimately, distinct mechanisms.
B. DIRECTANALYSISOF SIGNAL TRANSDUCTION The model systems described earlier have been devised to set the stage for direct analyses of the differences between immunogenic and tolerogenic signals. It is important to remember, however, that conditions that favor tolerance over “positive” immunologic signaling also involve the stage of differentiation of the target cell, form of the antigen (e.g., IgG with Fc feedback interactions), and threshold concentration effects. Hence, we should incorporate these parameters into our considerations before generalizing that a given signal is part of a tolerogenic pathway. For example, with adult B cells, uncoupling of surface IgM signaling via Fc receptor crosslinking leads to abortive PI hydrolysis and a failure to progress in the cell cycle (Bijsterbosch and Klaus, 1985); however, both this treatment and the productive stimulation of adult B cells with F(ab’)zanti-Ig lead to antibody unresponsiveness, at least with LPS challenge. The major factor that distinguishes tolerance from “immunity” in this model is the lack of T cell help, although the importance of Fc receptor feedback regulation is also critical (Sinclair and Panoskaltsis, 1987; Waldschmidt et al., 1983).Second, it is equally important to note that quantitative differences in second messengers may not be apparent beyond certain threshold levels of ligand and that, moreover, biological activity can result under conditions that fail to register in our assays (Brunswick et al., 1990).
Involvement of Calcium Mobilization, Phosphoinositide Hydrolysis, and Tyrosine Phosphorylation With these caveats in mind, several laboratories have concentrated on measuring the direct effects of IgM crosslinking in tolerance (reviewed in Monroe et al., 1992; Scott et al., 1993; Page and DeFranco, 1988). Monroe and colleagues (1992) found that immature B cells showed as much calcium mobilization as mature B cells, but that PI hydrolysis via IgM crosslinking was defective. They were able to show that the signaling defect was likely to be upstream of a membrane IgM complex-associated G-protein (Yellen-Shaw and Monroe, 1991) and to involve the lack of a src-like p56 protein. They conclude that, in the absence of events downstream of p56 (including appropriate PI hydrolysis), the calcium signal promoted tolerance, perhaps via apoptosis. In contrast, we (Scott et al., 1987) found that early calcium changes were not required for anti-IgM-mediated growth arrest in B cell lymphomas. Nonetheless, it is conceivable that subtle changes or late
ANALYSIS OF B CELL TOLERANCE in Vitro
419
alterations in [Ca?'] could occur in these cells. Indeed, Page and DeFranco (1988) reported that calcium ionophores (? a PKC activator) led to arrest in other B cell lymphomas; however, these cells may not arrest at the same point in the cell cycle nor undergo apoptosis (Warner and Scott, 1988). Based on the inhibitor studies described earlier (Beckwith et al., 1991), tyrosine phosphorylation must be one of the first signals in both negative and positive signaling. To date, there are no clear differences in the phosphorylated substrates seen under these different conditions, with one notable exception. A16s-Martinez et al. (1992) studied an IgM+IgD--expressing B cell lymphoma that is arrested by anti-IgM but is not inhibited by anti-IgD (Al6s-Martinez et al., 1988). The patterns of tyrosine phosphorylated substrates were virtually identical with either crosslinking antibody except that the &associated form of mb-1 (IgDa) was phosphorylated with anti-IgD. Determination of the protein substrates associated with IgDa will be important to unravel the downstream events in these processes. The pattern that has emerged is that tyrosine phosphorylation of a number of substrates occurs via signaling through the IgM(D) receptor complex that leads to activation of phospholipase C-7, and so on. In immature B cells, there is an uncoupling of downstream events, because of a lack of p56, that leads to abortive signaling for cell cycle progression. Although it is unknown whether the calcium signal in these cells causes programmed cell death, it is clear that T cell help can rescue these cells. It is not known, however, if provision of T cell help to Ig receptor signaled neonatal B cells leads to egr transcription (Monroe, personal communication). A further difference between tolerogenic events in immature versus mature B cells is that most in witro studies with adult B cells lead to blockade of B cell differentiation, but not proliferation (Warner et al., 1991a). Thus, it is important to note that phorbol esters clearly mimic surface IgM-induced unresponsiveness in adult models, a result suggesting that PKC activation plays an important role in downregulation of p heavy-chain RNA regulation by acting on components of the active transcription complex (Chen et al., 1991; Chiles and Rothstein, 1992).
C. IDENTIFYING THE IMMUNOGLOBULIN COMPLEX AND ASSOCIATED SECOND MESSENGERS It is now clear that the limited cytoplasmic structure of the IgM receptor limits, if not obviates, its function as a signal transduction molecule per se (for review, see Cambier and Campbell, 1992; Reth, 1992). Moreover, the common KVK sequence at the cytoplasmic tail of
420
DAVID W. SCOTT
several membrane Ig isotypes is not only too short to signal, but also does not allow for isotypic differences. Therefore, IgM must be associated with other transmembrane signaling proteins, in analogy to the CD3 complex of the T cell receptor, and isotypic differences must be encoded in the C-terminal and transmembrane domains of each Ig. Extensive review of this subject is beyond the scope of this article. Rather, I summarize data that suggest that members of the src family of nonreceptor PTKs not only associate with IgM, but are required for differential signaling in mature versus immature B cells. At least four members of the src family of PTKs are found in cells of the B lineage, including blk, fyn, lyn, and lck kinases. Of these, only blk is B cell specific (Dymecki et al., 1990)and can be co-immunoprecipitated by anti-IgM (Burkhardt et al., 1992).To determine the role of blk or other src family PTKs in B cell tolerance, we have examined the levels of these kinases and used antisense oligos to attempt to prevent growth arrest in immature B cell lymphomas. Our data (Yao and Scott, 1993) demonstrate that blk protein is present and activatable by antiIgM in all inhibitable lymphomas; in contrast, many resistant lymphomas lack the blk gene product, Moreover, anti-IgM stimulates a marked increase in blk phosphorylation (presumably to activate it), whereas this is quantitatively less in those resistant lymphomas that contain blk (Yao and Scott, 1993). To deplete blk from the sensitive lines, we treated CH31 cells for 48 hours with antisense blk and found that both growth inhibition and apoptosis were almost completely prevented. Thus, the activation of the blk PTK appears to be the initial signaling step required for tolerogenesis, as measured in this lymphoma model. Whether this PTK is required for B cell tolerance in development can not be ascertained at present and presumably will be approachable by studies in blk knockout mice, as well as in transgenics overexpressing inactivated blk products. IV. Conclusions
Analysis of B cell tolerance in vitro has permitted an understanding of the molecular pathways in this process. Although differences have been noted in the signal transduction events of immature, tolerancesusceptible, and more mature B cells, there is no doubt that the presence of T cell help is the single overriding feature preventing tolerogenesis. Fundamentally, however, tolerance is an active process in which initial tyrosine phosphorylation events are coupled to downstream signaling leading to (abortive) cell cycle progression. In both neonatal and adult B cells, receptor crosslinking initiates this cascade,
ANALYSIS OF B CELL TOLERANCE i n Vitro
42 1
although redistribution of receptors is a more stringent requirement in mature B cells. The lack of associated kinases in immature cells (or the dominance of others, like blk) may favor the abortive pathway. Ultimately, active transcriptional events perhaps involving c-myc can lead to the lack of pRB phosphorylation, growth arrest, and apoptosis (deletion) in immature cells or blockade of IgM transcription in mature B cells. Presumably, deletion is the result of arresting “cycling” immature B cells that cannot rearrange their receptors productively again. This is clearly an oversimplification of the fact that multiple pathways must exist for B cell tolerance considering the vast differences in the chemistry, vascular concentration, and developmental appearance of the multitude of self antigens (Scott, 1984). Indeed, although it is likely that tolerance in high-affinity B cells specific for a monovalent self antigen exists, it is not clear whether this is mediated by the same pathway as for multivalent antigens. It is, however, satisfying that after a twentysomething-year investment in tolerance in uitro, the results have been validated in transgenic models in uiuo.
ACKNOWLEDGMENTS This article is Publication 99 of the Immunology Division, University of Rochester Cancer Center. Grant support for the author’s research recorded herein has been generously provided by the U.S. Public Health Service (National Institute of Allergy and Infectious Diseases and National Cancer Institute, National Institutes of Health), the American Cancer Society, and the Council for Tobacco Research. I thank Sally Kent, Lieh-Bang Liou, Garvin Warner, and Elias Zambidis for their comments on this manuscript.
REFERENCES Aldo-Benson, M. (1986).Cell. Immunol. 103,417. Aldo-Benson, M., and Scheiderer, L. (1983)./.E x p . Med. 157,342. Aldo-Benson, M., and Tsao, B. P. (1987).Cell. Immunol. 108,1335. AlCs-Martinez, J.-E., Warner, G. L., and Scott, D. W. (1988).Proc. N d l . A c Q ~Sci. . U.S.A. 85,6519. AlCs-Martinez J. E., Phipps, R. P., Scott, D. W., Casnellie, J., Martinez-A.C., and Pezzi, L. (1992). Eur. J . Immunol. 22,845. Beckwith, M., Urba, W. J., Ferris, D. K., Freter, C. E., Kuhns, D. B., Moratz, C. M., and Longo, D. L. (1991).]. Immunol. 147,2411. Benhamou, L. E., Casenave, P-A., and Sarthou, P. (1990).Eur. J. Immunol. 20,1405. Bijsterbosch, M., and Klaus, G. G. B. (1985).J. E x p . Med. 162,1825. Blackwood, E.M., and Eisenman, R. N.(1991).Science 251,1211. Borel, Y.,and Kilham, L. (1974).Proc. SOC. E r p . Biol. Med. 145,470. Boyd, A., and Schrader, J. (1981).]. Immunol. 126,2466. Brown, D., Warner, G. L., Aks-Martinez, J., Scott, D. W., and Phipps, R. P. (1992).Clin. Immunol. Immunopathol. 63,221. Brunswick, M., Burkhardt, A., Finkelman, F., Bolen, J., and Mond, J. J. (1992).J. Immunol. 149,2249.
422
DAVID W. SCOTT
Buchkovich, K., Duffy, K. A., and Harlow, E. (1989).Cell (Cambridge, Mass.) 58,1097. Burkhardt, A., Brunswick, M.., Bolen, J., and Mond, J. J. (1992).Proc. Natl: Acad. Sci. U.S.A.88,7410. Burnet, F. M. (1959).“The Clonal Selection Theory of Acquired Immunity.” Vanderbilt Univ. Press, Nashville, TN. Cambier, J., and Campbell, K. (1992).FASEB J . 6,3207. Cambier, J. C., Vitetta, E., Uhr, J. W., and Kettman, J. R. (1977).J. E x p . Med. 145,778. Carrel, A,, and Ingebrigtsen, R. (1912).J. Exp. Med. 15,287. Carsetti, R.. Kohler, G., and Lamers, M. (1993).Eur.J. Zmmunol., 23,168. Casali, P., and Notkins, A. L. (1989).Annu. Reo. Zmmunol. 7,513. Chace, J. H., and Scott, D. W. (1988a).In “The Year in Immunology 1988”(J. M. Cruse and R. E. Lewis, Jr., eds.) p. 181.Karger, Basel. Chace, J. H., and Scott, D. W. (1988b).J.Zmmunol. 141,3258. Chen, U.,Scheuermann, R.,Wirth, T., Gerster, T., Roeder, R.,Harshman, K., and Berger, C. (1991).Nucleic Acids Res. 19,5981. Chiles, R.,and Rothstein, T. (1992).J. Zmmunol. 149,825. Chiller, J., Habicht, G., and Weigle, W. 0. (1971).Science 171,813. Cohen, D. P., and Rothstein, T.L. (1991).J. Zmmunol. 146,2921. Cong, Y., Rabin, E., and Wortis, H. H.(1991).Znt. Zmmunol. 3,467. DeCaprio, J. A., Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms, H., Huang, C-M., and Livingstone, D. M. (1989).Cell (Cambridge,Mass.) 58,1085. DeFranco, A. (1987).Annu. Rso. Cell. Biol. 3, 143. DeFranco, A., Davis, M., and Paul, W. E. (1982).In “B & T Cell Tumors” (E. Vitetta et al., eds.), pp. 445-450.Academic Press, New York. Desaymard, C. (1981).Cell. Zmmunol. 62,196. Diener, E., and Armstrong, W. D. (1969).J. E x p . Med. 129,591. Diener, E., and Paetkau, V. H.(1972).Proc. Natl. Acad. Sci. U.S.A. 69,2364. Draetta, G. (1990).Trends Biochem. Sci. 15,378. Dymecki, S . , Niederhuber, J. E., and Desiderio, S. (1990).Science 247,332. Ehrlich, P., and Morgenroth, J. (1901).“On Hemolysins: fifth Communication.” (Reprinted from Bed. Klin. Wochenschr., No. 10.)In Ehrlich, P. (1906).“Collected Studies on Immunity,” Chap. VII, p. 71.Wiley, London. Evan, G., Wyllie, A., Gilbert, C., Littlewood, T., Land, H., Brooks, M., Waters, C., Penn, L., and Hancock, D. (1992).Cell (Cambridge,Mass.)69,119. Fedyk, E., Borrello, M., Brown, D., and Phipps, R. P. (1993).Chem. Immunol., in press. J. Exp. Med. 131,247. Feldmann, M., and Diener, E. (1970). Fischer, G., Kent, S. C., Joseph, L., Bissonette, R.,Green, D. R.,and Scott, D. W. (1993). Submitted for publication. Fishman, M. (1959).Nature (London)183,1200. Flahart, R.,and Lawton, A. (1987a).J. E x p . Med. 166,864. Flahart, R., and Lawton, A. (1987b).J. Mol. Cell. Zmmunol. 3,61. Gaur, A., Yao, X-R., and Scott, D. W. (1993).J. Immunol., 150,1663. Coding, J. W., Scott, D. W., and Layton, J. E. (1977).Zmmunol. Reo. 37,152. Goodnow, C.C. (1992).Annu. Reo. Zmmunol. 10,489. Goodnow, C. C., Crosbie, J., Adelstein, S., Lavoi, T., Smith-Gill, S., Brink, R.,PritchardBriscoe, H., Wotherspoon, J., Loblay, R.,Raphael, K., Trent, R.,and Basten, A. (1988). Nature (London)334,676. Goodnow, C . C., Crosbie, J., Jorgensen, H., Brink, A., and Basten, A. (1989).Nature (London)342,385.
ANALYSIS OF B CELL TOLERANCE in Vitro
423
Goodrich, D. W., Wang, N. P., Qian, Y. W., Lee, E.Y-H. P., and Lee, W-H. (1991). Cell (Cambridge,Mass.) 67,293. Greeley, E. A., Berke, G., and Scott, D. W. (1974).j.lmmunol. 113,1883. Haas, W., and Layton, J. (1975).J.Exp. Med. 141, 1004. Hasbold, J., and Klaus, G. G. B. (1990). Eur. J . Immunol. 20, 1685. Hayakawa, K., Hardy, P. R., Honda, M., Herzenberg, L. A., Steinberg, A. D., and Herzenberg, L. A. (1984). Proc. Natl. Acad. Sci. U.S.A.81,2494. Hekkila, R., Schwab, G., Wickstrom, E., Loke, S. L., Pluznik, D., Watt, R., and Neckers, L. N. (1987). Nature (London) 328,445. Howard, M., Scott, D. W., Johnson, B., and Paul, W. E. (1981).In “B Lymphocytes in the Immune Response: Functional, Developmental and Interactive Properties (N. Klinman et ul., eds.), p. 141. Elsevier/North-Holland, New York. Hunter, T., and Pines, J. (1991). Cell (Cambridge,Mass.) 66, 1071. Ishida, Y., Agata, Y., Shibahara, K., and Honjo, T. (1992). E M B O J . 11,3887. Kearney, J., and Vakil, M. (1986). lmmunol. Reo. 94,39. Kincade, P. (1978).Ado. lrnmunol. 31, 177. Klinman, N. R. (1976).A m . ] . Pathol. 85,694. Laiho, M., DeCaprio, J. A., Ludlow, J. W., Livingston, D., and Massague, J. (1990). Cell (Cambridge,Mass.) 62, 175. Lawton, A., and Cooper, M. D. (1974). Contemp. Top. Immunobiol. 3,193. Lenardo, M., and Baltimore, D. (1990). Cell (Cambridge,Mass.) 59,229. Linton, P. J.. Rudie, A., and Klinman, N. R. (1991).j.Immunol. 146,4099. Liou, L-B., Colosia, A., Corley, R., and Scott, D. W. (1993).J.Immunol. 150, in press. Liou, L-B., Warner, G. L., and Scott, D. W. (1992). Int. lmmunol. 4,5. Ludlow, J. W., Shon, J., Pipas, J. M., Livingston, D. M., and DeCaprio, J. A. (1990). Cell (Cambridge,Muss.) 56,57. Marbrook, J. (1967).Luncet 2, 1279. Martensson, I.-L., Nillson, L., and Leanderson, T. (1989).Eur.j. Immunol. 19,1625. Mashesaran, S., McCormick, J. E., and Sonenshein, G. (1991). Oncogene 6,1965. Massague, J. (19YO).J.Annu. Reo. Cell Biol. 6,597. McCormack, J. E., Pepe, V. H., Kent, B. R., Dea, M., Marshak-Rothstein, A., and Sonenshein, G. (1984). Proc. Natl. Acad. Sci. U.S.A.81,5546. Metcalf, E., and Klinman, N. (1976).J.E x p . Med. 143, 1327. Metcalf, E., Schrater, F., and Klinman, N. R. (1977). Immunol. Rev. 43, 143. Mihara, K., Cao, X-R., Yen, A,, Chandler, S., Driscoll, B., Murphree, A. L., T’Ang, A., and Fung, Y. T. (1990).Science 246, 1300. Mishell, R., and Dutton, R. W. (1967).J.E x p . Med. 126,423. Monroe, J., Yellen-Shaw, A,, and Seyfert, V. (1992). In “Advances in Molecular and Cellular Immunology” (B. Singh, ed.), Vol. 1, p. 1. JAI Press, Greenwich, CT. Murakami, M., Tsubata, T., Okamoto., M, Shimizu, A., Kumagai, S., Imura, H., and Honjo, T. (1992). Nature (London) 357,77. Nemazee, D., and Biirki, K. (1989). Proc. Nutl. Acad. Sci. U.S.A. 86,8039. Nossal, G . J. V. (1992).Ado. Immunol. 52,283. Nossal, G. J. V., and Pike, B. (1978).j.E x p . Med. 148, 1161. Owens, G., Hahn, W., and Cohen, J. J. (1991). Mol. Cell. Biol. 11,4177. Page, D., and DeFranco, A. (1988).]. lmmunol. 140,3717. Pennell, C., and Scott, D. W. (1986). Sum. lmmunol. Res. 5,61. Pike, B., and Nossal, G. J. V. (1979). Eur.1. lmmunol. 9,708. Pike, B., Battye, F., and Nossal, G. J. V. (198l).J.lmmunol. 126,89.
424
DAVID W. SCOTT
Pike, B., Boyd, A., and Nossal, G. J. V. (1982).Proc. Natl. Acad. Sci. U.S.A.79,2013. Pillai, P. S.,and Scott, D. W. (1981).J. Zmmunol. 127,1603. Pillai, P. S.,and Scott, D. W. (1983).Cell. Immunol. 77,69. Raff, M. C., Owen, J. J. T., Cooper, M. D., Lawton, A. R.,111, Megson, M., and Gathings, W. (1975).J . Exp. Med. 142,1052. Ralph, P. (1979).Zmmunol. Reu. 48,107. Reth, (1992).Annu. Reu. Zmmunol. 10,97. Schad, V., and Phipps, R. P. (1988).J. Zmmunol. 141,79. Scott, D. W. (1968).Ph.D. dissertation, Yale University, New Haven, CT. Imrnunol. 111,789. Scott, D. W. (1973).J. Scott, D. W. (1976)Cell. Immunol. 22,311. Scott, D. W. (1984).Zmmunol. Today 5,68. Scott, D. W., and Klinman, N. R. (1987).Immunol. Today 8,105. Scott, D. W., and Waksman, B. H. (1968).J. Immunol. 100,312. Scott, D. W., Layton, J. E., and Nossal, G. J. V. (1977).J. Exp. Med. 146,1473. Scott, D. W., Tuttle, J., and Alexander, C. (1979a).I n “Proceedings of an International Conference on B Lymphocytes in the Immune Response” (M. Cooper et al., eds.), p. 263.Elsevier/North-Holland, New York. Scott, D. W., Venkataraman, M., and Jandinski, J. J. (1979b).Immunol. Reo. 43,241. Scott, D. W., Tuttle, J., Livnat, D., Haynes, W., Cogswell, J., and Keng, P. (1985).Cell. Immunol. 93,124. Scott, D. W., Livnat, D., Pennell, C. A., and Keng, P. (1986).J. Exp. Med. 164,156. Scott, D. W., Livnat, D., Whitin, J., Dillon, S. B., Snyderman, R., and Pennell, C. A. (1987).]. Mol. Cell. Immunol. 3,109. Scott, D. W., A16s-Martinez, J., Chace, J., Silver, L., and Warner, G . (1989). Cold Spring Harbor Symp. Quant. Biol. 54,899. Scott, D. W., Borrello, M., Liou, L-B., Yao, X-R.,and Warner, G . L. (1993a).I n “Advances in Molecular and Cellular Immunology” (B. Singh, ed.), Vol. 1, p. 119.JAI Press, Greenwich, CT. Scott, D. W., Warner, G. L., Yao, X-R.,and Kent, S. (1993b).Chem. Immunol. in press. Seyfert, V., Sukhatme, V., and Monroe, J. (1989).Mol. Cell. Biol. 9,2083. Seyfert, V., McMahon, S., Glenn, W., Yellen, A., Sukhatme, V., Cao, X., and Monroe, J. (1990).Science 250,797. Shi, Y.,Glynn, J. M., Gilbert, L. J., Cotter, T. G., Bissonnette, R. P., and Green, D. R. (1992).Science 257,212. Shull, M., Ormsby, I., Kier, A., Pawlowski, S., Diebold, R.,Yin, M., Allen, R.,Sidman, C., Proetzel, G., Calvin, D., Annunziata, N., and Doetschman, T. (1992).Nature (London) 359,693. Sidman, C. L., and Unanue, E. R. (1975).Nature (London)257,149. Sinclair, N. R. StC.,and Panoskaltsis, A. (1987).Immunol. Today 8,76. Spencer, C. A., and Groudine, M. (1991).Ado. Cancer Res. 56,l. Stein, S . , and Phipps, R. P. (1989).Eur. J . Zmmunol. 20,403. Strasser, A., Whittingham, S., Vaux, D., Bath, M. L., Adams, J. M., Cory,S., and Harris, J. (1991).Proc. Natl. Acad. Sci. U S A . 88,8661. Teale, J., and Klinman, N. R.(1980).Nature (London)288,385. Teale, J., and Klinman, N. R. (1984).J. Zmmunol. 133,1811. Ternyck, T., and Avrameas, S. (1986).Immunol. Reo. 94,99. Tiegs, S . M., Russell, D. M., and Nemazee, D. (1993).J. Exp. Med. 177,1009. Tisch, R., Roifman, C., and Hozumi, N. (1988).Proc. Natl. Acad. Sci. U.S.A.85,6514.
ANALYSIS OF B CELL TOLERANCE in Vitro
425
Uhr, J. W., and Vitetta, E. (1975). Science 189,964. Valentine, M., and Licciardi, K. (1992).Eur.1. Zmmunol. 22,3141. Venkataraman, M., and Scott, D. W. (1977).J.Zmmunol. 119, 1879. Venkataraman, M., and Scott, D. W. (1979). Cell. Zmmunol. 47,323. Venkataraman, M., and Scott, D. W. (1980).J . Zmmunol. 124,607. Vitetta, E., Canibier, J. C., Ligler, F., Kettman, J. R., and Uhr, J. W. (1977).J.E x p . Med. 146,1804. Waldschmidt, T., Borel, Y., and Vitetta, E. (1983)J. Zmmunol. 131,2204. Warner, G . L., and Scott, D. W. (1988). Cell. Zmmunol. 115, 195. Warner, G . L., and Scott, D. W. (1991).J.Zmmunol. 146,2185. Warner, G . L., Davies, S., and Scott, D. W. (1989).J.Zmmunol. 143,458. Warner, G . L., Gaur, A., and Scott, D. W. (1991a). Cell. Zmmunol. 138,404. Warner, G . L., Nelson, D., Ludlow, J., and Scott, D. W. (1991b).Cell Growth D y f . 3,175. Waters, C., and Diener, E. (1983). Eur.1. Zmmunol. 13,928. Werblin, T. P., and Siskind, G . W. (1974). Transplant. Rev. 8, 104. Yao, X-R., and Scott, D. W. (1992). Cell. Immunol. 142,434. Yao, X-R., and Scott, D. W. (1993). Immunol. Rev. 132, in press. Yellen-Shaw, A. J., and Monroe, J. G . (1992).J.E x p . Med. 176, 129. Yuan, D. (1987).J . Mol. Cell. Immunol. 3, 133. This article was accepted for publication on 10 February 1993.
Index A Abortive signaling, and second signals in adult B cells, 407-410 Accessibility model, gennline transcription, 241 Accessory molecules, CD4 or CDS, role in immune response, 124-125 Acquired immune deficiency syndrome, and IL-6,37-39 HIV infection, 37-38 Kaposi’s sarcoma, 38-39 ACTH, see Adrenocorticotropic hormone Acute phase proteins, synthesis, 25-27, 216 Adenocarcinoma, rat pancreatic, 324 Adhesion molecules, parallels with regulatory mechanisms, 308-310 LFA-1,309-310 Mac-l/CRB, 310 platelet gpIIb-IIIa integrin, 309 L-selectin, 310 Adrenocorticotropic hormone, 26-27 AIDS, see Acquired immune deficiency syndrome Alveolar macrophages, IL-Ira production, 185 Alzheimer’s disease, and IL-6,42-43 Amino acid sequence amino-terminal, peptides, 350 from cloned cDNA, 13 Antibodies anti-Ig, induced proliferation of normal human B cells, 363 anti-IgM growth arrest and apoptosis induced by, 410-411 induced unresponsiveness in adult B cells, 404-405 role of apoptosis in neonatal B-cell tolerance, 403-404 surrogate systems, 402-410 tolerance in B-cell subsets, 405-407 anti-IL-6, clinical applications, 50
Antigens, see also Superantigens mls identification as product of MMTV, 100-102 in oitro response to, 124-125 specific activation, molecular basis, 337-338 Antiinflammatory cytokines, IL-6 effect, 27-28 during placental/fetal development, 32 role in embryonic development, 31-32 Apoptosis and growth arrest induced by anti-IgM, 410-411 role in neonatal B-cell tolerance, 403-404 Arthritis, and IL-lra in animal models, IL-lra, 210 inflammatory, see Inflammatory arthritis in patients with arthritis, 210-211 in oitro studies, 209-210 Autocrine effects, IL-1,201-202 Autoimmune diseases, and IL-6 Alzheimer’s disease, 42-43 cachexia, 43 cardiac niyxoma, 39-40 Castleman’s disease, 40-41 and interleukin6,39-43 niesangial proliferative glomerulonephritis, 41 psoriasis, 41 rheumatoid arthritis, 40 streptococcal infection-associated, 140- 141 systemic lupus erythematosus, 41-42
B-cell lymphoma as model for tolerance, 410-415 stimulation modes, 253-254 stimulatory factor 2, 1-3 superantigen expression, 144-145
427
428
INDEX
B cells activation for proliferation and differentiation, 345-346 adult, abortive signaling and second signals, 407-410 chronic lymphocytic leukemia, 50-51 differentiation and expression, 338-343 genomic DNA structure or mb-I gene, 355 human counterpart of IgR-associated proteins, 356-357 IL-6 effects, 21 lineage expression of CD44,293 murine peritoneal, 377-378 neoplasia, and IL-6,33-36 regulatory elements of mb-I gene expression in restricted stage of B-cell differentiation, 355-356 signal transduction through MB-1 molecule, 351-355 specific genes involved in IgR complex, 348-351 specific proteins in IgR complex identified by molecular cDNA cloning, 348-357 B-cell tolerance biochemical analysis of signal transduction pathways, 415-420 direct analysis of signal transduction, 418 identifying Ig complex and associated second messengers, 419-420 use of inhibitors to block in oitro tolerance, 415-418 model systems of analysis, 394-415 anti-IgM surrogate systems, 402-410 B-cell lymphoma model, 410-415 hapten-specific, 395-402 history, 394-395 preliminary studies with transgenic spleen cells i n oitro, 414-415 rationale, 393 definition, 394 value of in oitro studies, 394 Bacteria, see also specific bacteria infection, IL-6 effects, 36-37 superantigen products, 113-115 Clostridium and Pseudomonas toxins, 114-115
Mycoplasmu arthritides mitogen, 113 streptococcal M proteins, 114-1 15 Blood vessels, IL-6 effects, 29-30 Bone marrow cells from, transplantation in mice, 46-47 human, CD44 expression, 293-294 Bone metabolism, IL-6 effects, 28 Borrella burgdorferi, 207 Bowel disease, inflammatory, see Inflammatory bowel disease Brucella abortus, 257,397
C Cachexia, and IL-6,43 Calcium, mobilization response in B lineage cells via MB-1 model, 354 role in signal transduction, 418-419 to signals through IgR, 367-368 Cancer, antitumor activity of IL-6,48-50 Capping, and resynthesis of receptors, 402-403 Castleman’s disease, and IL-6.40-41 CD44 and extracellular matrix, 282-291 ligand recognition, 271-272 and lymphocytic homing, 291-302 molecular isoforms and postranslational modifications, 272-282 regulation of interaction with extracellular matrix, 302-318 role in lymphocyte development, 271 and tumor cell migration, role in metastasis, 318-325 Cell cycle, control mechanisms in tolerance, 411-414 blocking negative signaling with antisense myc, 413-414 myc, transforming growth factor p and pRB, 411-413 Cell lines CD44 expression and hyaluranon binding, 303 CD44-positive, 304 IL-lra production, 190 Chromosomes IL-Ira genes, localization, 177-180
429
INDEX
Clinical disorders, and IL-6,44,45 Cloning B cells, 399-400 B cell-specific proteins in IgR complex, 348-357 IL-9, from human and mouse, 80-82 Clostridium, and Pseudomonas toxins, 114-115 Complementary DNA CD44 constructs, transfection and expression, 286-287 IL-lra, purification, and expression, 171-176 Corticosteroids. and estrogens, 50 Cyclic AMP, intracellular levels, elevation, 373 Cyclosporin A, inhibition of B cells, 408 Cytokines antiinflammatory, see Antiinflammatory cytokines and chemicals inhibiting IL-6 synthesis, 50-51 effects on IL-lra production by monocytes, 182-184 shared signals through c~130,15-19 Cytoplasmic domain, 281-282 regulation through, 311-313 and role of phosphorylation, 311-313 Cytoskeletal interactions, 346
D Development embryonic, see Embryonic development fetal, see Fetal development placenta, role of IL-6,32 regulation of IgR functional components, 375-378 Diabetes mellitus, role of IL-lra, 214 Diseases, see also specific diseases autoimmune, streptococcal infection-associated, 140-141 food poisoning, 128 and IL-6,39-43,44-45 immunodeficiency, 142-143 Kawasaki syndrome, 133-137 lymphoproliferative, 143-145 B-cell lymphoma, 144-145 Epstein-Barr virus, 143-144
rheumatoid arthritis, 137-140 systemic lupus-like autoimmune, 141-142 toxic shock syndrome, 128-133 DNA complementary, see Complementary DNA deletion and looping out, 234 replication, role in immune response, 237-238 transfection, expression of IgR on cell surface, 360-361 Downregulation, 374
E EBV, see Epstein-Barr virus Electrophoretic mobility shift assays, 255 Embryonic development, role of IL-6, 31-32 Endocytosis, and processing of membrane IgRlantigen complex, 348 Endothelial cells, 29-30 Endothelium-derived relaxing factor, 29-30 Endotoxins, and IL-1,31 Epstein-Barr virus, 143-144 and herpesvirus saimiri, 115-116 Escherichia coli, 174,236 expression library, immunoscreening, 13 Extracellular domain, modification, 313-3 15 Extracellular matrix and CD44,282-291 CD44 as receptor for hyaluronan, 284-287 features of CD44 molecule mediating interaction with hyaluronan, 287-290 general features, 282-284 and metastasis, 318-322 protein interaction with CD44, 290-291 regulation, and interaction with CD44, 302-3 18 evidence for regulated hyaluronan binding, 302-308 parallels with regulatory mechanisms of other adhesion molecules, 308-310
430
INDEX
possil)le niechanisnis for regulating CD44 receptor function, 310-318
F F e t d developnient, role of IL-6, 32 Fil)rol)lasts growth factor receptors, 365 I L-1ra prodnction, 188- 189 Fluorescein, specific neonatal B cells, 399 Food poisoning, 128 Friiginent cultures, specific, and liniiting dilution, 395-398 iinportance of atfinity and valence, 396-398 role of B-cell matnrational stage, 395-397
G Genes antisense niyc, blocking of negiitive signiiling, 413-414 Ig, regulation of rearrangenient, 375 IL-6, expression regulation, 8-10 N F 1L-6, 9-10 NF-KB, 10 IL- Ira, chroniosoni;il location iind structure, 177-180 d i - 1 , genoniic structure, 355 niyc, transforniing growth fiictor p ;ind pHB, 411-413 VJD,iund switch recoml)iiiatioii, conipiirison, 235-237 I 240-245 C e r n i h e C ~transcription, GI ycosylation patterns, cell-type specific, 314 Gruft-versris-host disease, IL-Ira etfects, 215 Graiirilocyte-iiiacropliage colony-stiiiirilatiiig factor, 174 Growth arrest, and apoptosis induced by iinti-IgM, 410-411 Growth fiictor p, and 1gA expression, 259,262
H Hairy leukemia, 50-51 Hapten-specific B-cell tolerance, 3 Y5-402 analysis with affinity-purified B cells, 398-40 1 ditferential signaling via IgM and IgD, 401-402 specific fragment cnltiires and limiting dilution, 395-398 Heart, niyxoma, and IL-6,39-40 Helio#osoniiodes pol ygyrus, 94 Heniatopoiesis CD44 activation states, 306 by CD44-specific antibodies in oitro, inhibition, 299-300 1L-6 effects, 22-25 physiological experiments implying role for CD44, 296-302 Hematopoietic cells CD44 expression during hematopoietic development in mouse and hrnniin, 29 1-294 on memory T cells and activated T cells, 294-296 normal, 304-305 progenitor, 1L-6 elkcts, 22-23 Hematopoietic system, 83-84 receptor superhiiiily, 82-83 Hepatocytes, see nlso Liver acute-phase protein synthesis, 25-27 1L-lra production, 188 Hepatocyte-stiinulating tictor, and 1L-6,4 Hernies antigen ~nediatingperipheral lyniph node adhesion, 296-298 Herpesvirns sainiiri, and Epstein-Barr virus, 115-116 Histocompatibility complex, niajor, I)inding, 117-120 HIV, see Hunian iiiii~ii~nodeficieiicy virus Horror ciutotoxicus, 394 HTLV-I, see Hnnian T-cell leiikemia virus type 1 Hunian and nionse heavy chain lociis organization, 23 1-233
43 1
INDEX
IL-9 characterization and cloning, 80-82
T cells expression of IL-9, 85 growth-siiiiulating activity of IL-9, 85-88 response to IL-9.86 Human inimunodeficiency virus, infection, and IL-6,37-38 Human T-cell leukemia virus type 1, trmsforniation o f T cells, 93 H yal uronan binding regulation, evidence, 302-308 CD44 as receptor for evidence, 284-287 features, 287-290 sequence of CD44 i n hyaluronan recognition, 284 CD44-specific antibody binding, in hi hi ti on, 284-286 Hy\)ritloina/plasmacytonia growth factor, 3 Hydrolysis, in signal transduction, 418-419 Hypothalainic-pitiiitary-adrenal axis, 30-31
I IL-Ira, see Interleukin-1 receptor an tagon ist Ininii~neresponse regulation, IL-6 effects H cells, 21 T cells, 21-22 to superantigens, 117-127 I>indingto major histoconipatil~ility colllplex, 117-120 direct non-T cell effects, 127 recognition by T-cell receptor, 121-124 role of';iccessor-yCD4 or CDS niolecules, 124-125 T cell deletion and anergy in iil.jected animals, 125-127 T cell signaling, versus conventional antigens, 127 Innniuie system, effects of 1L-lra, 202-204
Immunodeficiency diseases, 142-143 Immunoglobulin A, transforming, expression, 262 Immunoglobulin D, and IgM, differential signaling, 401-402 I ni in u n oglobul in E , I L-4 sti in d a t e d production, 247-248 Immunoglobulin G complex and associated second messengers, identification,
4 19-420
IL-4-stimulated production, 247-248 induced monocytes, 172, 173 Imniunoglobulin M, and IgD, differential signaling, 401-402 Immunoglobulin receptors molecules and B-cell differentiation and expression, 338-343 B cell-specific proteins, identification,
348-357 fn nct ion s, 345-348 activation of B cells for proliferation and differentiation, 345-346 cytoskeletal interactions, 346 endocytosis and processing of membrane IgR/antigen complex, 348 transportation of assembled IgH structure retaining functions, 347 future perspectives, 379-380 isotypes, 343-344 membrane IgM receptor induction by Ig gene transfection, 357-359 inenihrane and secreted, 342-343 regulation by interaction ofCD44 extracellular domain, 315-317 signal transduction, 359-379 Immunoglobulins, isotype switching history, 229-231 in humans, 260-262 IL-4 role in regulation, 260-262 transforming growth factor /3 and IgA expression, 262 interferon y-induced regulation, 256-257 mechanism of switch recombination, 234-240 organization of heavy chain locus in mouse and human, 231-233 switch regions, 233-234
432
INDEX
interleukin-4-induced regulation enhancment of switching to IgE and IgC, 248-249 molecular mechanisms, 252-256 role in T cell-dependent isotype switching, 249-252 stimulation of IgE and IgG production, 247-248 molecular mechanism, 231-246 directed switching, and germline CII transcription, 240-245 expression of downstream isotypes without switch recombination, 245-246 transforming growth factor P-induced, 257-259 Infection, and IL-lra i n vioo administration, 204-209 in animal models of sepsis, 204-207 in human infections and sepsis, 207-209 Infectious viruses, production of exogenous superantigens, 115-1 17 Inflammatory arthritis, IL-Ira in, 209-2 11 animal models, 210 patients with, 210-21 1 in oitro studies, 209-210 Inflammatory bowel disease, IL-lra in, 214-215 Inflammatory diseases, and IL-6,39-43 Inhibitors IL-lra, 168-169 IL-6, clinical applications, 50-51 use to block in oitro tolerance, 415-4 18 a,&-Integrin, 309-310 Interferon-P2/26-KDa protein, 3 Interferon-y, regulation of isotype switching, 256-257 Interleukin-1, autocrine or paracrine effects, 34-35,201-202 Interleukin-la, binding to type I receptors, 193 Interleukin-1 receptor antagonist binding to soluble IL-1 receptors, 195-197 type I1 IL-1 receptors, 194-195 biological relevance, 217-219 cDNA clones, purification and expression, 171-176
chromosomal localization and gene structure, 177-180 in diabetes mellitus, 214 discovery, 169-171 effects on immune system, 202-204 fiinction as receptor antagonist, 197-200 identification, 167-168 inhibitors, 168-169 in infection and sepsis i n vioo, 204-2W,215-217 in inflammatory arthritis, 209-21 1 in inflammatory bowel disease, 214-215 in malignancies, 213-214 niurine, cloning and expression, 175-176 in nervous system, 211-213 receptor binding of IL-lra, 192-200 regulation of IL-lra production, 180-192 structural variants of IL-lra protein,
176-177
i n oitro and i n vioo effects, 200-202
Interleukin-4 enhancement of isotype switching to IgE and lgC, 248-249 niolecular mechanisms of IL-4-induced switching, 252-256 role in T cell-dependent switching, 249-252 stimulation of IgE and IgC production, 247-248 Interleukin-6 biological function, 20-33 acute-phase protein synthesis in hepatocytes, 25-27 in blood vessels, 29-30 and hone metabolism, 28 effect on skin, 28-29 growth regulatory functions, 20 during hematopoiesis, 22-25 immune regulation, 21-22 i n neuronal cells, 30-31 during placental/fetal development, 32 proinflammatory and antiinflammatory cytokine, 27-28 role in embryonic development, 31-32 cell sources and inducers, 6 clinical applications, 48-50
433
INDEX
description, 1-2
a s diagnostic marker, 47-48
a n d disease AIDS, 37-39 1)acterial and parasite infection, 36-37 H-cell neoplasia, 33-36 IL-6 transgenic mice, 45-47 inflammatory o r autoininiriiie diseases, 39-43 viral infection, 37 historical overview, 1-4 inhil)itors, clinical applic,1' t'ions, 50-5 1 receptors for, 11-19 schematic ternary structure, 5 striictiire and expression inducers and producers, 6-8 regulation of IL-6 expression, 8-10 repression of IL-6 expression, 10-1 1 structure, 4-5 transcriptional regulatory clenieiits, H I n terlerikin-9 in hematopoietic system, 83-84 hnman and mouse, cloning anti characterization, 80-82 and human T cells, 85-88 and mast cells, 93-95 and inotise T cells, tiiniorigenesis, 88-93 receptors for, 82-83 studies with, 79-80 Interlenkin- 11, I)iological function, 16-17 Intr~iperitoncriin,IL-lrii injection, 212 Isotypes downstream, without switch recombination, expression, 245-246 d u d , expression without recom1)ination. 246 IgG receptor, 343-344 Isotype switching directed, and gerniline C,, transcription, 240-245 in humans, 260-262 IL-4 regulation, 260-262 transforming growth thctor p and IgA expression, 262 to IgE and IgC, IL-4 enhancement, 248-249
nondeletianal, 245-246 sequential, 238-239 I)y transtimning growth fictor p, regulation, 257-25H
K Kaposi's sarcoma, and IL-6, 38-38 Kawasaki syndrome, 133-137 ;in;ilysis of T-cell receptors, 135 diagnostic criteria for, 133 Keriitinocytes. IL-Ira production, 190- 192
1 Leiikemia-inhil)itory fictor, 15-16,
24-25
LFA-1, 3OCJ-310 Ligand recognition, a n d CD44,271-272 Lipopol ysaccharides, 373-374 Listeriu ?tiotioc!itoReties,207 stimulated splenic H cells, IL-4 effects, 247-248 Liver, see also Hepatocytes acute phase protein synthesis, 25 Lymphocytes H-cell, see H cells tlevelopnient, C D 4 4 involvement, 271 migration, role o f C D 4 4 in oioo, 298-299 T-cell, .WX T cells Lyniphocytic homing expression of CD44 on hematopoietic cells, 281-2136 mtl heniatopoiesis inhibition in c;itro by CD44-specific antibodies, 2H!J-300 physiological experinients implying role for CD44, 286-302 rrlationship to hernies antigen mediating peripheral lymph node ;itlhesion, 296-298 role in lymphocytic migration i n c;ioo, 298-299 signaling through CD44-ligantl interactions, 300-302 thyniocyte progenitors, migration inhihition, 299
434
INDEX
Lymphoma B-cell, see B-cell lymphoma mouse T cell, 289 Lymphoproliferative diseases B-cell lymphoma, 144-145 Epstein-Barr virus, 143-144
Mac-UCR3, 310 Macrophages alveolar, see Alveolar macrophages differentiation, IL-6 effects, 24-25 peritoneal, IL-lra production, 188 synovial, IL-lra production, 187-188 i n oitro-derived, IL-lra production, 185-186 Malignancies, IL-lra in, 213-214 Malignant diseases, therapy for, 49-50 Mammary tumor virus, superantigens, 102 Masking, in regulation of CD44 receptors, 317-318 Mast cell growth enhancing activity, 93 Mast cells, and IL-9,93-95 Mastocytosis, Nipponstronglyus-induced, 95 MB-1 molecule induction of calcium mobilization in early B cells, 354 signal transduction through, 351-355 Megakaryocytes, IL-6 effects, 23-24 Membranes and secreted Ig molecules, 342-343 IgM receptor expression by Ig gene transfection, 357-359 IgR/antigen complex processing, 348 Mesangial proliferative glomerulonephritis, and IL-6,41 Metastases, role of CD44,318-325 MMTV, see Murine mammary tumor virus Monoclonal antibodies anti-CD3, effect on IL-9 expression in T cells, 85 CD44-specific, in oitro effects, 301 to VSAG, structure-function studies, 105- 106 Monocytes adherent IgC effects on inductor of IL-lra production, 182
differential regulation of IL-lp and IL-lra production, 181-184 effects on cytokines in IL-lra production by, 182-184 purification of IL-lra from IgC-induced supernatants, 172 Mouse acquired immunodeficiency disease syndrome, 115 and human heavy chain locus organization, 231-233 IL-9 characterization and cloning, 80-82 1L-lra, cloning and expression, 175-176 IL-6 transgenic, 45-47 IL-9 receptor cDNA, 82,84 T cells, tumorigenesis, 88-93 transgenic spleen cells, in oitro studies with, 414-415 Murine mammary tumor virus, superantigens cellular expression, 106-107 Mls identification as product, 100-102 role in life cycle of virus and host, 107-108 structure-function studies, 103-106 Mycobacterium avium, 36 Mycobacterium tuberculosis, 36 Mycoplasma arthritides, mitogens, 113, 141 Myelosuppression, IL-6 treatment, 48 Myxoma, cardiac, and IL-6.39-40
N Negative signaling, blocking with antisense myc, 413-414 Nervous system, IL-lra in, 211-213 Neuroendocrine system, 30-31 Neuronal cells, IL-6 effect, 30-31 Neutrophils, IL-lra production, 189-190 Nipponstronglyus brasiliensis, 94-95
Oncostatin M, 15-16 Osteoarthritis, 187
435
INDEX
P Pancreas, rat adenocarcinoma cell line, 324 Paracrine effects, IL-1,201-202 Parasites, infection, and IL-6,36-37 Peritoneum, IL-lra production by macrophages, 188 Phosphatidylinositol 3-kinase. 368-369 Phosphoinositide, role in signal transduction, 4 18-419 Phospholipase C activation, 365-367 IgR complex to, 366 Phosphorylation cytoplasmic tail of CD44,281 possible role, 311-313 Physiological abnormalities, IL-lra blocking, 205 Physiological experiments, implying role for CD44 in hematopoiesis in oitro, inhibition by CD44-specific antibodies, 299-300 inhibition of thymocyte progenitor migration, 299 in lymphocytic migration in uioo, 298-299 relationship to hermes antigen mediating peripheral lymph node adhesion, 296-298 signaling through CD44-ligand interactions, 300-302 Phytohemagglutinin, 85 Placenta, development, role of IL-6.32 Plasmacytoma cell lines, dysregulated expression, 15 plasmacytoma/hybridoma growth factor, 3 Plasmodium berghei, 36 Plasmodium falciparum, 36 Plasmodium yoelii, 36 Platelets, gpIIb-IIIa integrin, 309 Polyclonal plastomycytosis, in patients with Castleman’s disease, 40-41 Polymerase chain reaction, quantitative techniques, 137 Post-transplant lymphoproliferative disorder, 35-36 Pregnancy, and preterm delivery, IL-lra effects, 216-217
Preterm delivery, IL-lra effects, 216-217 Proinflammatory cytokine, IL-6 effects, 27-28 Prostaglandins, production by IL-ha, 201 Protein isoforms, 275-281 Protein kinase C isozymes, 371 pathway, 369-374 plasma membrane kinase, 370 tyrosine kinase, role in IgR-associated transmembrane signaling, 359-364 Proteins, see also specific proteins extracellular matrix, interaction of CD44,290-291 C-binding, involvement, 364-365 human counterpart of IgR-associated, 356-357 IL-ha, structural variants, 176-177 in switch recombination, 239-240 Pseudomonas exotoxin A, 114 Psoriasis, and IL-6,41 Pulmonary diseases, IL-lra effects, 2 15-2 16
RA,see Rheumatoid arthritis Rabies virus, nucleocapsid, 116-1 17 Rat, IL-6 administration in oitro, 25-26 Receptor antagonists, function of IL-lra as, 197-200 Recombination homologous, 242 switch, see Switch recombination Retroviruses, origin of endogenous superantigens, 101-102 Rheumatoid arthritis, 137-140, 187 and IL-6,40 Rheumatoid synovial fibroblast-like cells, 193 Rheumatoid synovial tissue, 170
S Salmonella adelaide, 395 Schematic representation, human CD44 proteins, 275,276 Schistosoma mansoni, 37
436
INDEX
L-Selectin, 310 Sepsis, and IL-lra animal models, 204-207 in human infections, 207-209 Serinekhreonine kinase inhibitors,
18-19
Serum, IL-6 levels, 30 Shedding, in regulation of CD44 receptors, 317-318 Signal transduction through CD44-ligand interactions,
300-302
IgR-mediated cascade functional molecules, 351,352 stimulation, 378-379 IL-6R complex chain, 14,18 through IL-9 receptors, 83 through MB-1 molecule, 351-355 and pathway through IgR transmembrane signaling, 359-378 pathways in, biochemical analysis,
415-420
direct analysis, 418 identification of IgG complex and associated second messengers,
419-420
use of inhibitors to block in oitro tolerance, 415-418 Sister chromatoid exchange, 234-235 Skin, IL-6 effects, 28-29 S region, 236 Staphylococcus aureus, 108,109 and food poisoning, 128 infection-associated autoimmune disease, 140-141 Streptococcal M proteins, 114-115 Streptococcal toxins, 108-113 Streptococcus pyogenes, 109,114 vp specificity for, 109,110-1 12 Stromal cells, uterine, IL-lra production, 188 Structure-function studies, MMTV superantigens, 103-106 Superantigens, see also Antigens classes, 99-100 endogenous murine, 100-108 cellular expression of MMTV antigens, 106-107 identification of MIS antigens as MMTV products, 100-102
role in life cycle of MMTV and host,
107-108
structure-function studies, 103-106 exogenous, 108-117 bacterial products with related function as superantigen,
113-115
produced by infectious viruses,
115-117
staphylococcal and streptococcal toxins, 108-113 and human disease, 128-145 autoimmune disease associated with streptococcal infection, 140-141 food poisoning, 128 immunodeficiency disease, 142- 143 Kawasaki syndrome, 133-137 lymphoproliferative diseases,
143-145
rheumatoid arthritis, 137-140 systemic lupus-like autoimmune disease, 141-142 toxic shock syndrome, 128-133 immune responses to binding to major histocompatibility complex, 117-120 direct non-T cell effects, 127 involvement of accessory CD4 or CDS molecules, 124-125 recognition by T-cell receptors,
121-124 T cell deletion and anergy in adult
animals injected with superantigens, 125-127 T cell signaling in response to superantigens versus conventional antigens, 127 Switch recombination comparisons between VJD and switch recombination, 235-237 in DNA replication, 237-238 experimental systems for studying,
237
looping out and deletion, 234 proteins involved in switch recombination, 239-240 sequential switching, 238-239 sister chromatoid exchange, 234-235 Synovium, IL-lra production in macrophages, 187-188
437
INDEX
Systemic lupus erythematosus and IL-6,
41-42
systemic lupus-like autoimmune disease, 141-142
T T cells
activated, CD44 expression, 294-296 deletion and anergy in adult animals injected with superantigens,
125-127
dependent isotype switching, role of
IL-4,249-252
helper, clones, 89 human expression of IL-9,85 response to IL-9,86,87 IL-6 effects, 21-22 memory, CD44 expression, 294-295 mouse, 88-93 receptor recognition of superantigens,
121-124
models, 122-123 modifications of vp rule, 123-124 signaling in response to superantigens versus conventional antigens,
Transfection cell lines, positive and negative regulation of hyaluronan binding,
306,307
and expression of CD44 cDNA constructs, 286-287 Transgenic spleen cells, preliminary in oitro studies, 414-415 Trichinella spiralis, 94 Trichuris rnuris, 94 TSS, see Toxic shock syndrome Tumorigenesis, role of IL-9 and mouse T cells, 88-93 Tumors, see also specific tumors breast and colon, expression of CD44,
323 CD44 expression and metastasis, 322-325 cell migration, role of CD44 in metastasis, 318-325 chimeric, and IL-6,51
metastasis and extracellular matrix,
3 18-322
Tyrosine kinase developmental expression, 362 protein expression, 416 Tyrosine, phosphorylation role in signal transduction, 418-419
127
supernatants, 247 Thrombocytopenia, IL-6 treatment, 48 Thymic lymphomas, 92 Thymocyte progenitors, inhibition of migration, 299 Tolerance, B-cell, see B-cell tolerance Toxic shock syndrome clinical criteria in classification, 128,
129
fatal, 132-133 Staphylococcus uureus toxins, 129
TSST-1, 129, 131
Transcription, CD44 gene, multiple products, 278
U Uterus, stromal cells, IL-lra production,
188
v vp rule, modifications, 123-124 specificities of some exogenous bacterial superantigens, 109 Viruses, see also specific ciruses and host, role of in life cycle, 107-108 infection, and IL-6,37
CONTENTS OF RECENT VOLUMES
Volume 43
Antinuclear Antibodies: Diagnostic Markers for Autoimmune Diseases and Probes for Cell Biology
The Chemistry and Mechanism of Antibody to Protein Antigens
ENG.M. TAN
ELIZABETH D. GETLOFF, JOHNA. TAINER, RICHARDA. LERNER, AND H. MARIOCEYSEN
Interleukin-1and Its BiologicallyRelated Cytokines
CHARLES A. DINARELLO
Structure of Antibody-AntigenComplexes: Implicationsfor Immune Recognition
Molecular and Cellular Events of T Cell Development
P. M. COLEMAN
B. J. FOWLKES AND DREW M. PARDOLL
The y8 T Cell Receptor
MICHAELB. BRENNER, JACK L. STROMINGER, AND MICHAELS. KRANGEL
Specificity of the T Cell Receptor for Antigen
STEPHEN M. HEDRICK
Transcriptional Controlling Elements in the Immunological and T Cell Receptor Loci
KATHRYNCALAME AND SUZANNE EATON
Molecular Biology and Function of CD4 and CD8
JANER. PARNES
Lymphocyte Homing
TEDA. YEDNOCK AND STEVEN D. ROSEN
INDEX
Molecular Aspects of Receptors and Binding Factors for IgE
HENRY METZGER
INDEX
Volume 45 Cellular Interactions in the Humoral Immune Response
Volume 44 Diversity of the ImmunoglobulinGene Superfamily
ELLENS. VITETTA, RAPAEL FERNANDEZ-BOTRAN, CHRISTOPIiER D. MYERS,AND VIRGINIAM. SANDERS
MHC-Antigen Interactions: What Does the T Cell Receptor See?
PHILIPPE KOURILSKYAND JEAN-MICHELCLAVERIE
TIMHUNKAPILLER AND LEROY HOOD
Geneticolly Engineered Antibody Molecules
SHERIE L. MORRISONA N D VERNONT.
Synthetic T and B Cell Recognition Sites: Implications for Vaccine Development
DAVID R. MILICH
01
439
440
CONTENTS OF RECENT VOLUMES
Rationalefor the Development of an Engineered Sporozoite Malaria Vaccine
The Cellular and Subcellular Bases of lmmunosenescence
Virus-Induced Immunosuppression: Infections with Measles Virus and Human ImmunodeficiencyVirus
Immune Mechanisms in Autoimmune Thyroiditis
VICTORNUSSENZWEIC A N D RUTHS. NUSSENZWEIG
MICHAEL B. MCCHESNEY AND MICHAELB. A. OLDSTONE
The Regulators of ComplementActivation (RCA)Gene Cluster
DENNISHOURCADE, V. MICHAEL HOLERS,AND JOHNP. ATKINSON
Origin and Significance of Autoreactive T Cells
MAURICE ZAUDERER
INDEX
MARILYN L. THOMAN AND WILLIAM 0.WEICLE
JEANNINE CHARREIRE
INDEX
Volume 47 Regulation of lmmunoglobin E Biosynthesis
KIMISHIGE ISHIZAKA
Control of the Immune Responseat the Level of Antigen-Presenting Cells: A Comparison of the Function of Dendritic Cells and B Lymphocytes
JOSHUA P. METLAY,ELLENP u R ~AND , RALPHM. STEINMAN
Volume 46
The CD5 B Cell
Physical Maps of the Mouse and Human Immunoglobulin-likeLoci
Biology of Natural Killer Cells
ERICLAI, RICHARD K. WILSON,AND LEROYE. HOOD
Molecular Genetics of Murine Lupus Models
ARCYRIOS N. THEOFILOPOULOS, RIENHOLD KOPLER,PAULA. SINGER, AND FRANK J. DIXON
Heterogeneity of Cytokine Secretion Patterns and Functions of Helper T Cells
TIMR. MOSMANN A N D ROBERT L. COFFMAN
The Leukocyte lntegrins
TAKASHI K. KISHIMOTO, RICHARD S. LARSON, ANGELL. CORBI,MICHAEL L. DUSTIN,DONALD E. STAUNTON, AND TIMOTHY A. SPRINGER
Structure and Function of the Complement Receptors, CR1 (CD35) and CR2 (CD21) JOSEPHM. AHEARNAND DOUGLAS T.
FEAROW
THOMAS J. KIPPS GIORGIO TRINCHIERI
The lmmunopathogenesisof HIV Infection
ZEDAF. ROSENBERG AND ANTHONYS. FAUCI
The Obeses Strain of Chickens: An Animal Model with Spontaneous Autoimmune Thyroiditis
GEORGEWICK,HANSPETER KAREL HALA,HERMANN BREZINSCHEK, DIETRICH,HUGOWOLF,AND GUIDO KROEMER
INDEX
Volume 48 Internal Movements in Immunoglobulin Molecules
ROALDNEZLIN
CONTENTS OF RECENT VOLUMES Somatic Diversification of the Chicken Immunoglobulin Light-Chain Gene
WAYNE T. MCCORMACK AND CRAIG B. THOMPSON
T Lymphocyte-DerivedColony-Stimulating Factors
ANNEKELSO AND DONALD METCALF
The Molecular Basis of Human Leukocyte Antigen Class II Disease Associations
DOMINQUE CHARRON
Neuroimmunology
E. J. GOETZL,D. C. ADELMAN, AND S. P. SREEDHARAN
Immune Privilege and Immune Regulation in the Eye JERRY
Y. NIEDERKORN
Molecular Events Mediating T Cell Activation
AMNONALTMAN, K. MARK COGGESHALL, AND TOMAS MUSTELIN
INDEX
441
Adoptive T Cell Therapy of Tumors: MechanismsOperative in the Recognition and Elimination of Tumor Cells
PHILIPD. GREENBERG
The Development of Rational Strategies for Selective lmmunotherapy against Autoimmune Demyelinating Disease
LAWRENCE STEINMAN
The Biology of Bone Marrow Transplantation for Severe Combined Immune Deficiency
ROBERTSONPARKMAN
INDEX
Volume 50 Selective Elements for the Vp Region of the T Cell Receptor: MISand the Bacterial Toxic Mitogens
CHARLES A. JANEWAY, JR.
Programmed Cell Death in the Immune System
J. JOHNCOHEN
Volume 49 Human Immunoglobulin Heavy-Chain Variable Region Genes: Organization, Polymorphism, and Expression
VIRGINIA PASCUAL AND J. DONALD CAPRA
Surface Antigens of Human Leucocytes
v. HOfiEJSf
Expression, Structure, and Function of the CD23 Antigen G. DELESPESSE, U. SUTER,D.
B. BETTLER,M. SARFATI, MOSSALAYI, H. HOFFSTETTER, E. KILCHERR, P. DEBRE,AND A. DALLOUL
Immunology and Clinical Importance of Antiphospholipid Antibodies
H. PATRICK MCNEIL,COLINN. CHESTERMAN, AND STEVEN A. KRILIS
Avian T Cell Ontogeny
MAXD. COOPER, CHEN-LOH. CHEN, R. PATBUCY,AND CRAIGB. THOMPSON
Structural and Functional Chimerism Results from Chromosomal Tronslocation in Lymphoid Tumors
T. H. RABBITTSAND T. BOEHM
Interleukin-2, Autotolerance, and Autoimmunity
GUIDOKROEMER, Josd LUISANDREV, Josf ANGEL GONZALO, JosB C. GUTIERREZ-RAMOS, AND CARLOS M ARTfNEZ-A.
Histamine Releasing Factors and Cytokine-Dependent Activation of Basophils and Mast Cells ALLEN P. KAPLAN, SESHAREDDIGARI,
MARIABAEZA, AND PIOTRKUNA
442
CONTENTS OF RECENT VOLUMES
Immunologic Interactionsof T Lymphocytes with Vascular Endothelium JORDAN S. POBERAND RAMZI S.
COTRAN
Adoptive Transfer of Human Lymphoid Cells to Severely Immunodeficient Mice: Models for Normal Human Immune Function, Autoimmunity, Lymphomagenesis, ond AIDS
DONALD E. MOSIER
JONATHAN
W. YEWDELL
BENNICK
AND JACK
R.
Human B Lymphocytes: Phenotype, Proliferation, and Differentiation
BANCHEREAU AND FRANCOISE ROUSSET
Cyiokine Gene Regulation: Regulatory cis-Elements and DNA Binding Factors Involved in the Interferon System
Volume 51 Human Antibody Effector Function
DENNIS R. BURTONAND JENNY M.
WOOF
The Development of Functionally Responsive T Cells
ELLEN V. ROTHENBERC
Role of Perforin in Lymphocyte-Mediated Cytolysis
HIDEOYACITA, MOTOMINAKATA, AKEMIKAWASAKI,YOICHI SHINKAI, AND KO OKUMURA
The Central Role of Follicular Dendritic Cells in Lymphoid Tissues
FOLKE SCHRIEVER AND LEE MARSHALLNADLER
The Murine Autoimmune Diabetes Model: NOD ond Related Strains
HITOSHIKIKUTANIAND S u s u ~ u MAKINO
The Pathology of Bronchial Asthma
INDEX
Cell Biology of Antigen Processing and Presentationto Major Histocompatibility Complex Class I Molecule-RestrictedT Lymphocytes
JACQUES
INDEX
JOHATHAN
Volume 52
P. ARMAND TAKH. LEE
NOBUYUKI TANAKA AND TADATSUGU TANICUCHI
Cellular and Molecular Mechanisms of B Lymphocyte Tolerance G . J. V. NOSSAL Cell Surface Structures on Human Basophils and Mast Cells: Biochemical and Functional Characterization
PETERVALENTAND PETER BE-ITELHEIM
Animal Models for Acquired ImmunodeficiencySyndrome
THOMAS J. KINDT, VANESSA M. HIRSCH,PHILIPR. JOHNSON, AND SANSANA SAWASDIKOSOL
INDEX
Volume 53 Lymphokine and Cytokine Production by FcPRI' Cells WILLIAM E. PAUL, ROBERTA. SEDER, AND
MARSHALLPLAUT
The Leukemia Inhibitory Factor and Its Receptor
DAVID P. GEARING
CONTENTS OF RECENT VOLUMES
The Role of CD4 and CD8 in T Cell Activation and Differentiation
M. CARRIEMICELIAND JANE R. PARNES
B Lymphopoiesis in the Mouse ANTONIUSROLINKAND FRITZ MELCHERS Compartmentalizationof the Peripherol Immune System
GUIDOKROEMER,EDUARDOCUENDE, AND CARLOSMARTINEZ-A.
Immunological Memory
CHARLESR. MACKAY
Recognition of Bacterial Endotoxins by Receptor-DependentMechanisms
RICHARD J. ULEVITCH
Cell Adhesion Molecules as Targets of Autoantibodies in Pemphigus and Pemphigoid, Bullous Diseases Due to Defective Epidetmal Cell Adhesion JOHN
INDEX
R. STANLEY
443
