Advances in Immunology Volume 57

ADVANCES IN

Immunology EDITED BY FRANK J. DIXON The Scripps Research Institute La Jolla, California

ASSOCIATE EDITORS

FREDERICK ALT K. FRANKAUSTEN TADAMITSU KISHIMOTO FRITZMELCHERS JONATHAN W. UHR

VOLUME 57

ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper.

@

Copyright 0 1994 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.

A Division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495 United Kingdom Edition published by Academic Press Limited 24-28Oval Road, London N W l 7 D X International Standard Serial Number: 0065-2776 International Standard Book Number: 0-12-022457-7

PRINTED IN THE UNITED STATES OF AMERICA 94 95 9 6 9 7 98 99

BB

9 8 7 6 5

4 3 2 1

ADVANCES IN IMMUNOLOGY, VOL. 57

Molecular Basis of Fc Receptor Function MARK D. HULEll AND P. MARK HOGARTH The Austin Research Insiiitih, Heidelbeg 3084, AustmIm

1. Introduction

Cell membrane receptors specific for the Fc portion of immunoglobulin (FcR) play an important role in immunity and resistance to infection, providing a system that couples antibody-antigen interaction with cellular effector mechanisms. Distinct cell membrane FcRs have been described for all classes of immunoglobulin, including IgG (FcyR), IgE (FceR),IgD (FcsR), IgM (FcpR), and IgA (FcaR). Of these receptors only the leukocyte FcyR and FceR have been extensively characterized. The FcyRs comprise a multimembered family of structurally homologous but distinct receptors and are expressed on the vast majority of leukocytes. The diversity ofthese receptors is reflected in the wide variety of biological responses mediated upon their binding of IgG-antigen complexes, including phagocytosis, endocytosis, antibody-dependent cell-mediated cytotoxicity, release of inflammatory mediators, and regulation of B-cell function (reviewed in Unkeless et al., 1988; Mellman et al., 1988; Kinet, 1989; Ravetch and Anderson, 1991; Van de Winkel and Anderson, 1991; Ravetch and Kinet, 1991; Van de Winkel and Capel, 1993). In contrast, the FceRs comprise only two members which are structurally unrelated to each other. The highaffinity receptor for IgE is closely related to the FcyR. This receptor is exclusively expressed on mast cells, basophils, Langerhans cells, and eosinophils, and is responsible for triggering the IgE-mediated allergic response (reviewed in Metzger et aZ., 1986; Kinet, 1990; Ravetch and Kinet, 1991). The distinct low-affinity receptor for IgE has a much wider cellular distribution and plays a role in B-cell development and IgE-dependent cytotoxicity against parasites (reviewed in $Dieselbeg, 1984; Conrad, 1990; Delespesse et aZ., 1992). In addition to the membrane-bound FcRs, soluble FcRs or immunoglobulin binding factors ( IBFs) have also been described; however, these are not discussed here (for reviews see Fridman and Sautes, 1990; Fridman et al., 1992,1993). This review focuses on studies of the murine and human leukocyte FcyR and FcsRI, with particular reference to the structural characterization of these receptors, the molecular nature of their interaction with 1

Copyright Q 1904 by Academic Press, Inc. All rights of reproduction in any form reserved.

2

MARK D. HULETT AND P. MARKHOGARTH

Ig, and their mechanisms of signal transduction. In addition we also review aspects of FcpR, FcaR, the poly Ig receptor, the receptor for the transport of Ig in neonatal gut (FcRn), and receptors for IgD. The FcR nomenclature used throughout this review follows that proposed by Ravetch and Kinet (1991), unless stated otherwise.' II. Characterization of FcR

A. FcyR Significant advances have been made in recent years in the characterization of the receptors for the Fc portion of IgG (FcyRs) at the protein, transcript, and gene levels. Three distinct classes of mouse and human FcyR are currently recognized: FcyRI, FcyRII, and FcyRIII. These classes can be distinguished on the basis of a number of serological and biochemical criteria, including specificity and affinity for immunoglobulin, cell distribution, molecular size, and recognition with monoclonal antibodies (mAb). The cell-surface FcyRs are all integral membrane glycoproteins, with the exception of the glycosyl-phosphatidylinositol (GP1)anchored hFcyRIIIb isoform. Molecular cloning and sequence analysis of the cDNAs encoding mouse and human FcyRI (Allen and Seed, 1989; Sears et al., 1990), FcyRII (Ravetch et al., 1986; Lewis et al., 1986; Hibbs et al., 1986,1988; Hogarth et al., 1987; Stuart et al., 1987,1989; Stengelin et al., 1988; Brooks et al., 1989), and FcyRIII (Simmons and Seed, 1988; Ravetch and Perussia, 1989; Peltz et al., 1989) have indicated that they are all structurally related, containing conserved extracellular ligand-binding regions of Ig-like domains and as such belong to the Ig superfamily. FcyRI contains three Ig-like domains, whereas FcyRII and FcyRIII contain two Ig-like domains. The homology in the extracellular regions of the FcyR contrasts to the pronounced sequence differences observed in the transmembrane and cytoplasmic tail domains of these receptors. The divergence in the cytoplasmic regions of the FcyR suggests these domains are involved in triggering unique intracellular signals and, combined with selective cellular expression, presumably accounts for the diverse functions of the different FcyR classes. In the mouse, single genes encode each of the three FcyR classes (Qiu et al., 1990; Kulczycki et al., 1990; Hogarth et al., 1991; Osman et al., 1992); whereas in the human, multiple genes have been described for each class: three FcyRI genes, three FcyRII genes, and The prefixes m, h, and rt will be used to denote mouse, human, and rat, respectively.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

3

two FcyRIII genes, which encode multiple forms of these receptors (Ravetch and Perussia, 1989; Qiu et al., 1990; Van de Winkel et al., 1991; Ernst et al., 1992). The genes encoding the mouse and human FcyR are all located in the same region of chromosome 1 in both species (Sammartino et al., 1988; Grundy et al., 1989; Peltz et al., 1989; Qiu et al., 1990; Ernst et al., 1992; De Wit et al., 1993) with the exception of the mouse FcyRI gene on chromosome 3 (Osman et al., 1992), and have clearly arisen by duplication and divergence of a common ancestral FcyR gene. The following sections describe the properties of the three mouse and human FcyR classes, which are summarized in Tables I and I1 for FcyRI, Tables IV and V for FcyRII, and Tables VI and VII for FcyRIII, aspects of which have been reviewed in Unkeless et al. (1981,1988),Anderson and Looney (1986), Mellman et al. (1988), Van de Winkel and Anderson (1991),Ravetch and Anderson (1991),Ravetch and Kinet (1991), and Van de Winkel and Cape1 (1993). 1. FcyRI a. Biochemical and Molecular Structure

Human FcyRI (CD64)is classically defined as a 72-kDa glycoprotein (Anderson et al., 1982,1986; Frey and Engelhardt, 1987; Dougherty et al., 1987; Peltz et al., 1988a), which following removal of N-linked carbohydrate has a protein core of 55 kDa (Peltz et al. 1988a) (Table I). In contrast, reports of the molecular mass of mouse FcyRI have been conflicting and include the description of a 70-kDa protein (Lane et al., 1980; Lane and Cooper, 1982), a 100-kDa protein of two 50kDa subunits (Fernandez-Botran and Sukuki, 1986; Hirata and Suzuki, 1987; Kagami et al., 1989), and a 57-kDa protein (Loube et al., 1978). However, a recent study has definitively determined the molecular weight of mFcyRI to be similar to that described for its human homologue, identifying the receptor as a 70-kDa phosphoprotein on mFcyRI-transfected Chinese hamster ovary cells and on the myeloid cell line, J774 (Quilliam et al., 1993). The cDNA cloning of human and mouse FcyRI has demonstrated that the above forms of the receptor are structurally unique among the FcyR, containing an extracellular region of three Ig-like domains, in contrast to the two domain structures of FcyRII and FcyRIII (Table 11). The third extracellular domain is distinct, whereas the first two domains are homologous to the extracellular domains of FcyRII and FcyRIII, suggesting that the unique IgG binding characteristics of FcyRI are conferred by domain three. Indeed, this has subsequently been demonstrated (see below and Hulett et al., 1991).

4

MARK D. HULETT AND P. MARKHOGARTH TABLE I CHARACTERISTICS OF FcyRI Genes Human Characteristic

hFcyRIA

Isofoms

hFcyRIa

Alleles Chromosome localization Ig-like domains Receptor Associated subunits Molecular mass (kDa) Apparent Protein backbone Affinity for IgG' (&)

-

Specificity hIgG mIgG Cellular distribution

Regulation of expression

hFcyRIB

Mouse mFcyRI

hFcyRIC

hFcyRIb 1, hFcyRIc mFcyRI hFwRIb2

-

-

lq21.1

3

3 TM y chain FcsRI

2 TM, S

2 S

-

3 TM ?

72 40 108-109

ND 22 ND, 4 M-'

ND 24 ND

72 42 107-108 M -

ND ND

3>1>4>>>2 2a>>>l, 2b, 3 Monocytes Macrophages Neutrophils

ND

IFN-y

lq21.1

lq21.1

M-1

2*,

3>1>4>>>2 ND 2a=3>>>1,2b ND Monocytes Macrophages Neutrophilsb Eosinophils' IFN? t , IL-10 t c ND G-CSF t , IL-4 J.

0 7

-

t

Note. TM, transmembrane; S, soluble; ND, not determined. Monomeric I&. * Expression inducible with 1FN-y. Expression upregulated on monocytes with IL-10 or IFN-y, downregulated with IL-4. Expression also upregulated on neutrophils with G-CSF.

In the human, three distinct hFcyRI cDNA clones were initially isolated using a ligand-mediated expression cloning procedure, designated p90, p135, and p98/X2 (Allen and Seed, 1989).All three clones encode integral membrane glycoproteins with an extracellular region of 292 amino acids comprising three Ig-like domains, a single membrane spanning domain of 21 amino acids, and an intracytoplasmic domain of either 61 (p90 and p135) or 31 amino acids (p98/X2). The p90 and p135 cDNA clones are identical in their coding regions, with the exception of two nucleotide substitutions that result in two amino acid differences in the first extracellular domain, suggesting these two clones represent different polymorphic forms of the receptor. These

5

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION TABLE I1 FcyRI GENES AND TRANSCRIPTS Name h F q RIA

hFcy RIB

hFcy RIC

Gene structure' L1 L2 D1

D2

I

I

D

Transcripts

D3

I

I

L1 L2 D1 I

I

I D2 I

hFcy RIbl [L11L2 D1 h F q RIb2

I

TM/C hFcy RIc ~ - ~ L l l L 2D1

I

TM/C mFcy RI p - - [ L l I L 2 I D1

I D2 I D3 I TM/C 1

-1

I

L1 L2 D1 I

hFcy RIa L11 L2 I D l

TM/C

D2

I

D3

n

I

D2

D3

I TM/C

I TM/C I

1

I

stoD in D3

stop in D3

D2

I A3 I TM/C]

stop

L1 L2 D1 mFcy RI

I

I

D2 I

D3

I

Exons shown as boxes, translated regions shaded, untranslated regions open. L, leader peptide; D, extracellular domain; TM, transmembrane; C, cytoplasmictail coding regions; PA, polyadenylation site. Alternate splicing of D3 exon.

amino acid substitutions do not affect the IgG binding characteristics of the receptors (Allen and Seed, 1989). The sequence of the p98/X2 cDNA clone diverges markedly from the other two cDNAs at its 3' end, becoming a complex pattern of repeats of upstream sequences, which encodes a divergent cytoplasmic tail. The authenticity of this clone has therefore been questioned on the basis of the suggestion that it may have resulted from a cloning artifact (Ravetch and Kinet, 1991). The subsequent cloning of the hFcyRI genes supports this conclusion, as the gene sequences do not account for such a transcript (Ernst et al., 1992). Three hFcyRI genes (designated hFcyRIA, B, and C) have been isolated (Van de Winkel et al., 1991; Ernst et aZ., 1992). The genes demonstrate a high degree of similarity, all containing an identical introdexon structure comprising six exons; two exons encoding the 5'-untranslated region (UTR) and leader sequence, one exon for each Ig-like domain, and a single exon for the transmembrane, cytoplasmic tail and 3' UTR (Table 11). Each gene spans 9.4 kb, and maps to chromosome lq21.1 (Ernst et al., 1992; Osman et al., 1992; De Wit et al., 1993; Dietzsch et al., 1993). Of the three genes, only hFcyRIA encodes an integral membrane receptor with three Ig-like domains

6

MARK D. HULETT AND P. MARKHOGARTH

(FcyRIa, the 72-kDa form), as both the hFcyRIB and IC genes contain translation termination codons in the exon encoding the third extracellular domain. Transcripts derived from the hFcyRIB and IC genes containing these stop codons have been described (hFcyRbl and hFcyRIc, respectively), and these may code for soluble receptors; however, the existence of such receptor proteins has not yet been demonstrated. In addition, an alternatively spliced product from the hFcyRIB gene has been reported (hFcyRIb2), in which the third extracellular exon is spliced out to produce a transcript encoding a twodomain integral membrane receptor (Ernst et al., 1992; Porges et al., 1992). Such a receptor would be expected to bind IgG, however, with low affinity, based on the observation that the first two domains of mouse FcyRI function as a low-affinity receptor (Hulett et al., 1991). Indeed, upon transfection into COS cells hFcyRIb2 binds only IgG complexes and not monomeric IgG (Porges et al., 1992). However, the expression of hFcyRIb2 on the surface of hematopoeitic cells has not been demonstrated. The transcript from the hFcyRIA gene resembles that of the hFcyRI cDNA clones p135 and p90 described previously (Allen and Seed, 1989), with the exception of two amino acid substitutions in the first extracellular domain and a single substitution in the cytoplasmic tail (Ernst et al. 1992). Recently, hFcyRIa has been shown to associate with homodimers of the y-subunit of the high-affinity receptor for IgE, FceRI (Ernst et al., 1993; Masuda and ROOS, 1993). This association was observed in monocytes and the myelomonocytic cell line U937 and could be reconstituted by cotransfection of the hFcyRIa and FceRI-y-subunit cDNAs into COS cells. It should be noted that the cell-surface expression of hFcyRIa is not dependent on association with FceRIy (Allen and Seed, 1989; Ernst et al., 1993). This finding adds FcyRI to the growing list of leukocyte FcR that are known to associate with FceRIy homodimers. The association of y with FceRI and FcyRIII has been well characterized, and more recent studies have now suggested not only hFcyRIa, but hFcaRI (L. Pfefferkorn, personal communication) and possibly some forms hFcyRII (Masuda and Roos, 1993)also associate with the y-subunit (see below). A single mouse FcyRI cDNA has been isolated and encodes a receptor homologous to hFcyRIa, comprising an extracellular region of 273 amino acids containing three Ig-like domains, a single transmembrane region of 23 amino acids, and a cytoplasmic tail of 84 amino acids (Sears et al., 1990).Comparison of the predicted amino acid sequences of mouse and human FcyRIa reveals an overall 75% amino acid identity in the extracellular regions and transmembrane domain, yet diver-

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

7

gence is seen in the cytoplasmic tails which are only 25% identical, with the mouse FcyRI tail containing an additional 23 amino acids (Sears et al., 1990). In contrast to the existence of multiple hFcyRI genes, only a single mouse FcyRI gene has been isolated, the structure of which is conserved with that of the human FcyRI genes, comprising six exons of identical organization spanning 9 kb (Table 11) (Osman et al., 1992). The mFcyRI gene has been mapped to chromosome 3 and is located in a conserved linkage group which contains the genes for CD1, LFAS, CD2, and the ATPase gene, which is syntenic with the region spanning the centromere and the proximal long- and shortarm regions of human chromosome 1 (Osman et al., 1992; Oakey et al., 1992; Dietzsch et al., 1993).

b. Ligand Affinity and Specificity In addition to the unique structure of FcyRI, this class of FcyR is also functionally distinct as it is the only FcyR that binds monomeric IgG with high affinity and as such is referred to as the high-affinity FcyR. Scatchard analysis of the direct binding of monomeric IgG has indicated that human FcyRI displays an equilibrium affinity constant (K,) of -z 108-109 M-' (Anderson and Abraham, 1980; Anderson, 1982; Fries et al., 1982; Kurlander and Batker, 1982; Cohen et al., 1983; Perussia et al., 1983; AlIen and Seed, 1989). Human FcyRI exhibits a specificity for hIgGl and hIgG3, binding monomeric forms of these isotypes. The receptor also binds hIgG4, but with a lower affinity and does not bind hIgG2 (Anderson and Abraham, 1980;Woof et al., 1986). The specificity of hFcyRI for mouse IgG is distinctive among the human FcyR, binding only the IgG2a and IgG3 isotypes (Anderson, 1982; Perussia et al., 1983; Jones et al., 1985; Van de Winkel et al., 1987; Ceuppens et al., 1988). The binding of aggregated IgG by FcyRI is reported to be of a similar affinity to monomeric IgG (Cosio et al., 1981; Kurlander and Batker, 1982; Woof et al., 1986). This raises the question of how FcyRI is able to distinguish IgG-coated particles in vivo as presumably due to the high serum levels of monomeric IgG, the receptor would be expected to be continually saturated with ligand. The reported upregulation of the receptor by IFN-y at sites of inflammation may therefore be crucial in the function of FcyRI (see below). The mouse homologue of hFcyRI exhibits many of the unique IgG binding characteristics of hFcyRI. The affinity of mFcyRI for monomeric IgG is also high; however, it is somewhat lower than that of hFcyRI, with a K, in the order of 107-108M-' (Unkeless and Eisen, 1975; Sears et aE., 1990; Hulett et al., 1991). Mouse FcyRI is unique as

8

MARK D. HULETT AND P. MARKHOGARTH

it is the only FcyR that binds a single mIgG class-mIgG2a-making it distinctive from even its human homologue which also binds mIgG3 (Haeffner-Cavaillon et al., 1979b; Sears et al., 1990; Hulett et al., 1991). However, the binding of hIgG subclasses by mFcyRI is similar to that by hFcyRI, with hIgGl and IgG3 binding preferentially over hIgG4 and no binding of hIgG2 (Haeffner-Cavaillon et al., 1979a). c. Cell Distribution and Monoclonal Antibodies Human FcyRI is constitutively expressed on monocytes and macrophages and can be selectively induced on neutrophils and eosinophils with IFN-y, which also upregulates expression on monocytes and macrophages (Guyre et al., 1983; Perussia, 1987; Shen et al., 1987; Pan et al., 1990; Hartnell et al., 1992). Indeed, IFN-.)Iresponse elements have recently been identified in the promoter of the hFcyRIB gene (Pearse et al., 1991; Benench et al., 1992). In a similar manner to IFNy , IL-10 also enhances hFcyRI expression on monocytes (Te Velde et al., 1992). However, in contrast to both IFN-y and IL-10, IL-4 has been shown to downregulate monocyte FcyRI expression (Te Velde et al., 1990). It has also been reported that G-CSF can upregulate hFcyRI expression on neutrophils (Repp et al., 1991). A number of mAb to hFcyRI have been reported, including 32.2 (Anderson et al., 1986), FR51 (Frey and Engelhardt, 1987), 10.1 (Dougherty et al., 1987), 197.1,22,62, and 44.1 (Guyre et al., 1989) (Table 111). All of the mAb have been shown to recognize epitopes distinct from the IgG binding site, with the exception of the FR51 and 10.1 mAb which block IgG binding to hFcyRI, although 10.1 blocks binding only partially (Frey and Engelhardt, 1987; Dougherty et al., 1987). The nonblocking mAb, 22, 32.2, 44.1, and 62, are specific for epitopes distinct from the ligand binding site, where mAb 22 and 44.1 define one epitope and 32.2 and 62 define a second. Determination of the cellular distribution of mFcyRI has been complicated by the lack of anti-mFcyRI mAb, combined with the overlapping expression of other FcyR classes which also bind mIgG2a (originally thought to bind only FcyRI). However, despite these problems, mFcyRI has been identified umambiguously on macrophages and monocytes (Unkeless and Eisen, 1975; Walker, 1976; Unkeless et al., 1979). IFN-y has also been demonstrated to upregulate the expression of mFcyRI on macrophages (Sivo et al., 1993; and N. Osman, personal communication).

d . Polymorphism A functional polymorphism of hFcyRI has been described with the identification of some members of a Belgium family which appear to

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

9

TABLE 111 FcR MONOCLONAL ANTIBODIES Epitope in relation to binding site Receptor hFcyRI mFcyRI hFcyRII

Blocking Fr51, 10.1

Nonblocking 32, 197, 22,44,62

-

-

CIKM5', 8.2c

mFcyRIII hFcsRIa

IV-3, KuFc79, 41H16"sb 2E14 KB61b, ATIOb 7.3@, 8.7b,8.26 2.4G2, Ly-17.2 3G8,4F7, VEPI3,1D3 GRM-ld, B73.1e, CLB-GRANII" Leulla, Leullb 2.4G2 15A5, 12E7,6F7,4B4

hFcaRI

My43

mFcyRII hFcyRIII

-

BW20912

2237, llB4,5D5,8C8 29C9,39D5,3B4 A3, A59, A62, A77

Specific for hFcyRIIa HR isoform. Preferential binding to B-cell FcyRII. Nonblocking onIy in Fab form. Specific for hFcyRIIIb NA-2 isoform. Specific for hFcyRIIIb NA-1 isoform.

lack hFcyRI on their blood monocytes. This was demonstrated as monocytes from these individuals did not bind mIgC2a or mIgG3 in an FcyRI-dependent anti-CD3-induced T-cell proliferation assay. Furthermore, their monocytes did not bind the anti-hFcyRI mAb 32.2 and 44.1, which detect two distinct epitopes of the receptor (Ceuppens et al., 1988; Ceuppens and van Vaecjk, 1989).In addition, stimulation of both monocytes and neutrophils from these individuals with IFNy , which strongly upregulates FcyRI expression (see above), did not induce FcyRI on these cells (Ceuppens et al., 1988).The absence of hFcyRI on the cells of these individuals did not, however, appear to alter their immune function or increase their susceptibility to infection. This suggests that the functions performed by FcyRI can be compensated for by the other FcyR classes, raising the question as to the functional importance of FcyRI. The molecular basis of this polymorphism has yet to be determined. Comparison of cDNA and genomic sequences also suggests genetic polymorphisms of hFcyRIa, specifically two amino acid substitutions in the first extracellular domain and one in the cytoplasmic tail (see above) (Allen and Seed, 1989; Ernst et al., 1992). A polymorphism has also recently been described for mFcyRI, whereby the nonobese diabetic (NOD) mouse strain was found to

10

MARK D. HULETT AND P. MARKHOGARTH

express a form of FcyRI containing 17 amino acid substitutions and a premature stop codon at position 337 which results in a deletion of 73 amino acids in the cytoplasmic tail (Prins et al., 1993).This mutant form of FcyRI demonstrated a 73% reduction in the turnover of cellsurface receptor-antibody complexes. Interestingly, the mutant FcyRI allele was shown to be tightly linked to a diabetic phenotype, suggesting that defective FcyRI function may play a role in susceptibility to the disease. 2. FcyRll a. Biochemical and Molecular Structure Human FcyRII (CD32)has been characterized as a 40-kDa glycoprotein (Cohen et al., 1983; Rosenfeld et al., 1985; Looney et al., 1986a; Van de Winkel et al., 1989; Ierino et al., 1993), with a putative protein core of 36 kDa, determined following treatment with endoglycosidaseF (Van de Winkel et al., 1989). Mouse FcyRII is more heterogeneous in size, with a molecular weight ranging from 40 to 60 kDa (Mellman and Unkeless, 1980; Hibbs et al., 1985; Holmes et al., 1985) (Table IV). The cDNA cloning of mouse and human FcyRII has demonstrated that this class of FcyR comprises multiple isoforms in both species. The receptor isoforms are all integral membrane glycoproteins, with the exception of a single putative soluble receptor in both the human and mouse, and contain extracellular regions of two Ig-like domains. The extracellular and transmembrane domains are highly conserved, yet their intracytoplasmic regions differ considerably, suggesting that the different isoforms of FcyRII are likely to transduce different signals to mediate different cellular responses (see Section IV). The cloning of human FcyRII cDNAs predicted the presence of multiple receptor isoforms (Stuart et al., 1987,1989; Hibbs et al., 1988; Stengelin et al., 1988, Brooks et al., 1989; Seki, 1989; Warmerdam et al., 1990; Rappaport et al., 1993), and the source of the heterogeneity was subsequently demonstrated at the genomic level with the cloning of three distinct genes, FcyRIIA, IIB, and IIC (formally hFcyRIIa, IIb, and IIa’, respectively), encoding a total of six transcripts (Table V) (Qiu et al., 1990; Warmerdam et al., 1993). The three genes are similar in structure, each comprising eight exons; two exons encode the 5’ UTR and leader sequence, one exon for each of the Ig-like domains and the transmembrane region, and three exons encode the cytoplasmic domain and 3’ UTR (Qiu et al., 1990). The existence of multiple exons encoding the transmembrane, cytoplasmic tail and the 3’ UTR regions of the human FcyRII genes (and the mouse FcyRII

TABLE IV

CHARACTERISTICS OF FcyRII Genes Human Characteristic

hFcyRIIA

Isoforms Alleles Chromosome localization Ig-like domains Receptor topology Associated subunitsb Molecular mass (kDa) Apparent Protein backbone Affinity for I@ (&) Specificity hIgG

hFcyRIIa1, hFcyRIIa2 HR, LR" 1q23-24

mIgG Cellular distributionf

Regulation of expression

hFcyRIIB

Mouse mFcvRII

hFcyRIIC

hFcyRIIb1, b2, b3

hFcyRIIc

1q23-24 2 TM

1q23-24 2 TM ?

mFcyRIIbl, b2, b3 Ly-17.1, Ly-17.2 1 2 TM, TM, S ?

40 31

40 29,27

4 0 7 M-1

4 0 7 M-1

40 31 ND

40-60 33,29 4 0 7 M-'

LR 3>1 = 2>>>4 HR 3>1>>2>4 LR 2a=2b>>l HR2a=lb=1 Monocytes Macrophages Neutrophils Platelets Langerhans cells IL-4 1

3>1>4>2d

ND

3>1>2>>4

2a= 2b>l

ND

1=2a=2b>>>3"

Monocytes (IIbl, IIb2) Macrophages B cells (IIbl, IIb2)

Monocytes Macrophages Neutorphils B cells

Monocytes, mast cells Macrophages, platelets Neutrophils, B cells

ND

ND

ND

2

TM, S ?

?

Notes. TM, transmembrane; S, soluble; ND, not determined. a HR, high responder; LR, low responder. F c E R I reported ~ to associate with FcyRII, but isoform unknown. Monomeric I&. Only determined for hFcyRIIbl isoform. mFcyRII also binds mIgE with low affinity. 'Cellular distribution not completed for all FcyRII isoforms and all cell types.

-

12

MARK D. HULETT AND P. MARKHOGARTH TABLE V FmRII GENESAND TRANSCRIPTS

Name

Transcripts

Gene structurp h F q RIIal

L1 L2 D1

D2

TM Clb

h F q RIIA

I L11 L2 I D1 I D2

ITM IC2 I C31 hFcy RIIa2 L11 L2 I D1 I D2 I C21 C3 h F q RIIbl I L l I L 2 I D l I D2ITMIC11C21C31 h F q RIIb2 I L ~ ~ L ~ I D ~I TI MDI c z~ I c ~ ] h F q RIIb3 [ L l l D1 I D2l TM I C2 I C31

1

L1

L2 D1

D2

TM C1 C2 C3 p~

h F q RIIB

I

-I

hFcy RIIc L l ] L2 I D1 I D2 ITM lC2l ~ 3 1

h F q RIIC

mFcv RIIbl

L1 L2 L3

m F q RII

n

LA D1

~m

D2

TM C1 C2 C3

Exom shown as boxes, translated regions shaded, untranslated regions open. alternate spicing indicated. L, leader peptide: D, extra-

cellular domain, TM, transmembrane;C, cytoplasmic tail coding regions;PA, polyadenylation site. C1 exon is cryptic (always spliced out in hFcyRIIA and hFcyRIIC).

gene, see below) is unique to this FcyR class as the FcyRI and FcyRII genes in both species contain a single exon encoding this region. The human FcyRII genes have been located to band q23-24 on chromosome 1(Sammartino et al., 1988; Grundy et al., 1989; Qiu et al., 1990) and are linked with the human FcyRIII genes on a 200-kb genomic fragment (Peltz et al., 1989). The hFcyRIIA gene (formally hFcyRIIa) encodes three transcripts, two of which arise through the use of alternate polyadenylation sites, producing either a 1.4- or 2.4-kb mRNA (Stuart et al., 1987; Hibbs et al., 1988; Stengelin et al., 1988, Brooks et al., 1989; Qiu et al., 1990; McKenzie et al., 1992), which encode identical integral membrane receptors (hFcyRIIal), and a third transcript (hFcyRIIa2) encoding a putative soluble hFcyRIIa product that is believed to arise from alternate splicing of the transmembrane region encoding exon ( Warmerdam et al., 1990; Rappaport et al., 1993; Astier et al., 1994).The predicted hFcyRIIal receptor contains an extracellular region of two Ig-like domains spanning 178 amino acids, a single transmembrane domain

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

13

of 29 amino acids, and a 76-amino acid cytoplasmic tail. The soluble hFcyRIIa2 receptor would be identical to the hFcyRIIa1 receptor, but lacking the 29-amino acid transmembrane region. The hFcyRIIB gene (formerly hFcyRIIb) encodes three distinct transcripts (bl, b2, and b3), which arise by alternate splicing (Brooks et al., 1989; Qiu et al., 1990). The b l and b2 isoforms are produced as a result of alternate splicing of the first cytoplasmic tail encoding exon (Brooks et al., 1989; Qiu et al., 1990; Hogarth et al., 1991). The hFcyRIIb3 transcript arises through alternate splicing of the second exon encoding the leader peptide cleavage site. The predicted hFcyRIIb1 receptor comprises a two Ig-like domain extracellular region of 179 amino acids, a single transmembrane domain of 23 amino acids, and a cytoplasmic tail of64 amino acids. The hFcyRIIb2 receptor is identical to hFcyRIIbl except for the deletion of 19 amino acids from the cytoplasmic tail as a result of alternate splicing of the first cytoplasmic tail encoding exon. The mature form of the hFcyRIIb3 receptor, if expressed, would be identical to hFcyRIIb1, differing only by a 7 amino acid deletion in the leader sequence corresponding to the second exon (Brooks et al., 1989). A single transcript has been reported from the hFcyRIIC gene (formally hFcyRIIa’) (Brooks et d.,1989),which predicts a receptor almost identical to hFcyRIIa, with an extracellular region of 178 amino acids, a transmembrane domain of 29 amino acids, and a cytoplasmic tail of 75 amino acids. The protein products predicted from the cDNAs of the three hFcyRII genes are all closely related, displacing an overall 85% amino acid identity in their extracellular and transmembrane regions (>95% if only the extracellular regions are compared) and only diverge in their leader sequences and cytoplasmic tail regions. The hFcyRIIa and hFcyRIIc receptors differ only in their leader sequences; the leader sequence of hFcyRIIc is homologous to that of hFcyRIIb, whereas the leader sequence ofhFcyRIIa is related to that of hFcyRII1. The hFcyRIIb receptors differ markedly from both FcyRIIa and IIc in their cytoplasmic tail regions, where after the first 10-12 residues, the hFcyRIIb sequence diverges from that ofthe FcyRIIalIIc receptors (Brooks et al., 1989; Qiu et aZ., 1990). Allelic forms of hFcyRIIa have been described which further increase the diversity of hFcyRII. These allelic variants were identified on the basis of a functional polymorphism for the binding of mIgGl or human IgG2 and termed the high-responder (HR) and lowresponder (LR) isoforms of FcyRIIa. The molecular basis of the polymorphism has been defined through recent cDNA cloning studies

14

MARK D. HULETT AND P. MARKHOGARTH

(Clark et at., 1989; Warmerdam et al., 1990; Tate et al., 1992) (see below). A recent report has suggested that hFcyRII, like hFcyRI and hFcyRIII, associates with the y-subunit ofFcER1; however, this association is not required for celI-surface expression (Masuda and ROOS, 1993). In contrast to hFcyRII, only two distinct murine FcyRII cDNAs encoding integral membrane proteins have been cloned, FcyRIIbl and FcyRIIb2 (formally FcyRIIPl and FcyRIIP2) (Table V) (Ravetch et al., 1986; Lewis et al., 1986; Hogarth et al., 1987). These cDNAs encode identical receptors, with the exception of a 47-amino acid inframe deletion in the cytoplasmic domain of FcyRIIb2. Both receptors comprise an extracellular region of two Ig-like domains spanning 180 amino acids, a single transmembrane domain of 26 amino acids, and cytoplasmic tails of either 94 (mFcyRIIb1) or 47 (mFcyRIIb2) amino acids. A third mFcyRII cDNA has also recently been isolated (designated mFcyRIIb3, but distinct from hFcyRIIb3) and encodes a molecule identical to mFcyRIIb2; however, it lacks the transmembrane region and thus encodes a soluble form of mFcyRII (Tartour et al., 1993). The cloning of a single mouse FcyRII gene indicated that the b l and b2 isoforms arise by differential splicing of the 141-bp first cytoplasmic tail encoding exon (corresponding to the 47 amino acid insertion) in an analogous manner to the two human FcyRIIB gene products: hFcyRIIb1 and b2. The b3 isoform arises by splicing of the transmembrane and first cytoplasmic tail encoding exons (Qiu et al., 1990; Kulczycki et al., 1990; Hogarth et al., 1991).The mFcyRII gene comprises 10 exons spanning 18 kb; 4 exons encode the 5' UTR and leader sequence, single exons encode each of the two Ig-like domains and the transmembrane region, and 3 exons encode the cytoplasmic tail and 3' UTR (see Table V). The mFcyRII gene maps to the Ly-17 locus on chromosome 1 (Davidson et al., 1983; Holmes et al., 1985; Hibbs et al., 1985) and is linked to mFcyRIII on a 160-kb genomic fragment (Kulczycki et al., 1990). The mouse FcyRIIbl, IIb2, and IIb3 receptors demonstrate an overall 60% amino acid identity with the hFcyRII receptors in their extracellular regions. mFcyRIIb1 and IIb2 are clearly most closely related to hFcyRIIbl and b2; the FcyRIIbl receptors exhibit an overall 59% amino acid identity across their entire lengths, and the FcyRIIb2 receptors display 57% identity (Brooks et al., 1989). Two allelic forms have been described for mouse FqRII, identified originally with mouse monoclonal alloantibodies and known as the Ly-17 polymorphism. The molecular basis of the polymorphism has been defined as a two residue difference in the second extracellular domain of mFcyRII (see below).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

15

The original mFcyRII clone (mFcyRIIP) was isolated at the same time as another distinct yet homologous clone that exhibited 95% identity in its extracellular region, yet contained divergent leader peptide, transmembrane, and cytoplasmic tail regions (Ravetch et al., 1986). The receptor encoded by this cDNA bound the anti-mFcyRII mAb 2.4G2 and exhibited identical ligand binding characteristics to mFcyRII and as such was designated an FcyRII isoform, FcyRIIa (Weinshank et al., 1988).This receptor has subsequently been reclassified as mFcyRIII, based on its homology to hFcyRIII in cDNA sequence and gene structure, and is discussed below.. The isolation and determination of the introdexon structure of the human and mouse FcyRII and FcyRIII genes has suggested that these two receptor classes have arisen by two evolutionary pathways from an ancestral FcR gene, producing the low-affinity FcyR receptors (Qiu et al., 1990; Kulcyzcki et al., 1990). Mouse FcyRII and FcyRIII have been proposed as the prototype receptors for these two classes (Qiu et al., 1990). As described, the mouse FcyRIII gene structure differs from the FcyRII gene as it contains a single exon encoding the transmembrane, cytoplasmic tail, and 3’ UTR, in contrast to the four exons encoding these regions in the FcyRII gene (Kulcyzckiet al., 1990).The FcyRII evolutionary stream includes only the FcyRII genes, while the FcyRIII stream also includes the FcERI a-chain gene on the basis of similar introdexon organization (Qiu et al., 1990; Ye et al., 1992).The FcyRI gene is structurally unique as it contains an additional exon encoding a third extracellular domain, although the transmembrane and cytoplasmic tail regions are encoded by one exon, thus the gene structure most closely resembles that of FcyRIII (Allen and Seed, 1989; Sears et al., 1990; Qiu et al., 1990; Osman et al., 1992; Ernst et al., 1992). It has been proposed that the FcyRIIA gene has arisen through a recombination event between the mouse FcyRII and FcyRIII primordial genes, whereby the 5‘ end of the gene has been derived from FcyRIII and the 3’ end from FcyRII (Qiu et al., 1990). Based on the gene structures and cDNA sequences of the mouse and human low-affinity FcyR, an order of human FcyRII gene evolution has been suggested, with the order of homology to mouse FcyRII as IIB > IIC > IIA (Qiu et al., 1990).An alternative theory for the evolution of the human FcyRII genes has also been proposed, which suggests that the human FcyRIIC gene was generated by an unequal crossover event between FcyRIIA and FcyRIIB (Warmerdam et al., 1993).This theory implies the order of human FcyRII gene evolution as FcyRIIB > FcyRIIA > FcyRIIC, which is in contrast to that proposed above (Qiu et d.,1990).However, these findings clearly demonstrate that the multiple human FcyRII genes have arisen via the pro-

16

MARK D. HULETT AND P. MARKHOGARTH

cesses of gene duplication, divergence, and recombination, from a primordial FcyRII gene resembling the single mouse FcyRII gene.

b. Ligand Afinity and Spect$city In contrast to FcyRI, FcyRII demonstrate a significantly lower affinity for ligand and, together with FcyRIII, comprise the low-affinity FcyRs (Table IV). Human FcyRII binds monomeric IgG poorly (K,< lo7A4-') and essentially only interacts with IgG complexes (Cohen et al., 1983; Kurlander et al., 1984; Rosenfeld et al., 1985; Rosenfeld and Anderson, 1989; Van de Winkel et al., 1989). However, it has been reported that the affinity of hFcyRII for IgG can be influenced by proteases, which increase affinity for IgG (Van de Winkel et al., 1989,199013;Tax and Van de Winkel, 1990).Early determinations of the specificity of hFcyRII for different subclasses of IgG produced some conflicting results, which was due in part to the coexpression of FcyRII with other FcyR and also the unrecognized heterogeneity of this receptor class, now known to comprise a number of different isoforms some of which exhibit functional polymorphisms (see below). Experiments performed prior to the cloning of hFcyRII suggested that this class of FcyR preferentially bound hIgGl and hIgG3 and to a lesser extent hIgG2 and hIgG4 (Dickler, 1976; Karas et al., 1982; Anderson and Looney, 1986). Human FcyRII was also shown to bind mouse isotypes IgGl and IgG2b (Abo et al., 1984; Tax et al., 1984; Jones et al., 1985; Looney et al., 1986a).The subsequent cDNA cloning of hFcyRII has enabled examination of the specificity of individual isoforms using transfection systems; however, data on the binding of all the isoforms are still incomplete. The hFcyRIIaLRand hFcyRIIaHRisoforms have been shown to display distinct specificities for both human and mouse IgG isotypes (Warmerdam et al., 1990,1991; Tate et al., 1992). Examination of human IgG isotype binding demonstrates that both the HR and LR isoforms bind hIgG3, hIgG1, but not hIgG4. However, these isoforms differ markedly in their binding of hIgG2, with hFcyRIIam exhibiting strong binding, in contrast to hFcyRIIaHRwhich binds hIgG2 weakly (Warmerdam et al., 1991). Examination of mouse IgG isotype binding indicates that both the HR and LR isoforms bind mIgG2a and mIgG2b, whereas only hFcyRIIaHRbinds mIgGl strongly (Warmerdam et al., 1990; Tate et al., 1992).The molecular basis of the differential binding of mIgGl and hIgG2 by these two isoforms has been determined (see below). The specificity of hFcyRIIb1 has also been defined recently, with the avidity of binding of hIgG istoypes following the order

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

17

hIgG3 = h I g G l > > hIgG2 > hIgG4 and mouse isotypes IgG2a = mIgG2b > mIgG1 (Warmerdam et al., 1992). The hFcyRIIb2 and b3 isoforms have not yet been examined; however, as they contain identical extracellular regions to hFcyRIIbl it is likely that they have similar IgG binding specificities. Interestingly, the binding of mIgG1 and hIgG2 to hFcyRIIbl was shown to be temperature sensitive, with the binding of complexes of these isotypes increasing significantly upon raising the temperature from 4 to 37°C (Warmerdam et al., 1992). Mouse FcyRII, like hFcyRII, binds monomeric IgG with a low affinity that is essentially undetectable using direct binding methods (K,< 10 M - I ) (Unkeless et al., 1988; Mellman et al., 1988; Hulett et al., 1991). A recent study has suggested that the specificity of this receptor for IgG is not absolute, providing evidence to demonstrate that mFcyRII can bind mIgE with low affinity (Takizawa et al., 1992). The specificity of mFcyRII for mouse IgG subclasses is broad, as the receptor binds mIgG1,2b, and 2a; however, mFcyRII does not bind mIgG3 (Unkeless, 1977,1979; Heusser et al., 1977; Haeffner-Cavaillon et al., 1979a; Teillaud et al., 1985; Lopez et al., 1985; Hulett et al., 1991). It has been proposed that a distinct mouse receptor exists specific for IgG3, but has not been characterized either biochemically or by molecular cloning (Diamond and Yelton, 1981). Mouse FcyRII binds hIgG subclasses with a specificity similar to hFcyRII, binding hIgGl and hIgG3 preferentially and hIgG2 and IgG4 less well (Haeffner-Cavaillon et al., 197913).

Cell Distribution and Monoclonal Antibodies Human FcyRII is the most widely distributed class of hFcyR, being expressed on almost all leukocytes, including monocytes, macrophages, neutrophils, basophils, eosinophils, Langerhans cells, platelets, B cells, and some T-cell subclasses, but is absent on NK cells (Vaughn et al., 1985; Looney et al., 1986a7b,1988;Valent et al., 1989; Anselmino et al., 1989; Schmitt et al., 1990; Sandor and Lynch, 1992; Mantzioris et al., 1993). Human FcyRII has also been demonstrated on nonimmune cells including placental trophoblasts (Stuart et al., 1989) and placental endothelial cells (Sedmak et aZ., 1991). The specific cell-type expression of the different hFcyRII isoforms is not welldefined; however, using the polymerase chain reaction (PCR), transcripts of the hFcyRIIA and hFcyRIIC genes have been detected in monocytes, macrophages, and neutrophils, whereas hFcyRIIB gene mRNA has been detected in monocytes, macrophages, and B cells (Brooks et al., 1989). A more recent study using both Northern blot and PCR analysis has further defined the cellular distribution of the individual transcripts of the three hFcyRII genes (Cassel et al., 1993). c.

18

MARK D. HULETT AND P. MARKHOGARTH

FcyRIIA was shown to be expressed in megakaryocytic cells, with both the a1 and a2 transcripts present in comparable amounts. In contrast, B cells express FcyRIIbl, b2, and FcyRIIc mRNA, but not FcyRIIal or a2. Myelomonocytic cells were shown to contain transcripts from all three hFcyRII genes, i.e., FcyRIIal, b l , b2, and c, with FcyRIIal the predominant mRNA species (Cassel et al., 1993). These findings clearly demonstrate that the FcyRIIA, B, and C genes products are differentially expressed in hematopoietic cells. It should be noted that substantial quantities of soluble FcyRIIa2 are present in and secreted from platelets, Langerhans cells, and megakaryocytic cell lines (Rappaport et al., 1993; Cassel et al., 1993; Astier et al., 1994).The levels of hFcyRII expression can be influenced by a number ofcytokines. IFN-y and IL-3 have been shown to upregulate the expression of hFcyRII on eosinophils (Hartnel et al., 1992). In contrast to hFcyRI and hFcyRII1, hFcyRII levels on monocytes and neutrophils appear not to be upregulated by any cytokine; however, IL-4 has been reported to downregulate its expression (Te Velde et al., 1990). In addition, GM-CSF has been shown to enhance cytotoxicityby hFcyRII on eosinophils; however, this enhanced function appears to be a result of increased receptor affinity and not due to increased hFcyRII expression (Graziano et al., 1989; Valerius et al., 1990; Koenderman et al., 1993). The distribution of cellular expression of mFcyRII is similar to its human counterpart, displaying a broad range of expression on hematopoietic cells, including monocytes, macrophages, neutrophils, mast cells, eosinophils, platelets, B cells, and some T cells (Unkeless et al., 1988; Mellman et al., 1988; Ravetch and Anderson, 1991). The mFcyRIIb1 and mFcyRIIb2 isoforms exhibit tissue-specific expression, with mFcyRIIb2 found predominantly in monocytes and macrophages, whereas mFcyRIIb1 is preferentially expressed in B lymphocytes (Ravetch et al., 1986; Lewis et al., 1986; Hibbs et al., 1986). The mFcyRIIb3 transcript has been identified in macrophages, and a soluble product possibly encoded by this mRNA has been detected in macrophage supernatents (Tartour et al., 1993). Numerous mAb recognizing hFcyRII have been described and include IV-3 (Looney et al., 1986a),CIKM5 (Pilkington et al., 1986),KuFc79 (Vaughn et al., 1985), 41H16 (Gosselin et al., 1990), 2E1 (Farace et al., 1988), KB61 (Pulford et al., 1986), AT10 (Greenman et al., 1991), 7.30, 8.2, 8.26, and 8.7 (Ierino et al., 1993) (see Table 111).All of these mAb exhibit the capacity to block Fc binding to hFcyRII, with the exception of mAb 8.2 and CIKM5 which have been shown to bind to an epitope distinct from the Ig binding site; however, these mAb can block IgG

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

19

binding to FcyRII via their Fc portions (Van de Winkle et al., 1987; Ierino et al., 1993). Differences in the reactivity of a number of the hFcyRII mAb are also demonstrated in the ability of some to preferentially bind B cells, which include mAb 2EI (Farace et al., 1988), KB61 (Pulford et al., 1986), 41H16 (Gosselin et al., 1990), AT10 (Greenman et al., 1991), 8.7, and 7.30 (Ierino et al., 1993). The 41H16 mAb is unique in that it is able to discriminate between hFcyRIIaHRand hFcyRIIaLR,selectively binding the HR isoform (see below) (Gosselin et al., 1990).The epitopes of a number of the hFcyRII mAb have been mapped in detail using a combination of crossblocking studies and reactivity with chimeric FcRs. The findings reveal that IV-3, 8.26, 8.7, and 7.30 (blocking mAb) have epitopes located in the second extracellular domain of hFcyRII, whereas CIKM5 and 8.2 (nonblocking mAb) have epitopes that seem to comprise regions from both the first and second extracellular domains. This suggests that the second extracellular domain of hFcyRII is involved in the binding of IgG (Ierino et al., 1993). Two mAb have been described that bind mFcyRII, a rat mAb 2.4G2 (Unkeless et al., 1979) and a mouse mAb anti-Ly-17.2 (Hibbs et al., 1985) that specifically detects the Ly-17.2 polymorphic form of mFcyRII (Holmes et al., 1985; Hibbs et al., 1985).Both mAb can block the binding of IgG to the receptor. The 2.4G2 mAb also recognizes mFcyRII1. d. Polymorphism A number of polymorphisms have been identified in hFcyRII. A genetic polymorphism defined as the high-responder/low-responder polymorphism has been identified in hFcyRIIa. This polymorphism was originally observed in anti-CD3 T-cell mitogenesis assays used to examine the interaction of mIgGl and human monocytes. Monocytes of different individuals were found to stimulate mIgGl anti-CD3 mAb T-cell proliferation in such assays either strongly or weakly, and these individuals were termed high responders (HR)or low responders (LR) (Tax et al., 1983). The distribution of high- and low-responder individuals was shown to be distinct in different races, with the finding that Caucasians are 70% HR, 30% LR, whereas in asians the ratio is reversed with 15% HR, 85% LR (Abo et al., 1984). The involvement of hFcyRII in this polymorphism was directly demonstrated with the finding that the anti-hFcyRII mAb IV-3 could block T-cell proliferation in the assay (Looney et al., 1986b) and was not due to differences in FcR numbers on the monocytes of high- and low-responder individuals (Anderson et al., 1987).This implied that there was an intrinsic differ-

20

MARK D. HULETT AND P. MARKHOGARTH

ence in FcyRII on the monocytes of HR and LR individuals, and, indeed, a structural difference in hFcyRII between HR and LR individuals was suggested, as different isoelectric focusing patterns were observed between receptors from these individuals (Anderson et al., 1987). Complementary DNA cloning of hFcyRII from the peripheral blood mononuclear cells of HR and LR individuals enabled the molecular basis of the polymorphism to be determined. The polymorphism has been defined as a two-residue difference in the extracellular region of the FcyRIIa isoform, with a glutamine or tryptophan at position 27 in the first extracellular domain and an arginine or histidine at position 131in the second extracellular domain (Clark et al., 1989;Warmerdam et al., 1990; Tate et al., 1992). Transfection experiments examining the IgG binding capacity of the various alleles have indicated that residue 131is responsible for the functional polymorphism, with arginine critical for the binding of mIgGl and found in HR isoforms, whereas histidine is present in LR isoforms. The tryptophan or glutamine at position 27 has no effect on the binding of mIgGl, and both residues have been described in HR and LR alleles (Warmerdam et al., 1990; Tate et al., 1992).Residue 131has also been shown as crucial for the binding of human IgG2; however, the amino acid required for binding is the reverse of that observed for mIgG1, with histidine (LR form) and not arginine (HR) promoting strong binding of hIgG2 (see above) (Warmerdam et al., 1991; Parren et al., 1992). The observation that residue 131is important for the binding of both mIgGl and hIgG2 supports the finding that the second extracellular domain is the IgG binding domain of hFcyRII (Hulett et al., 1993; see Section 111). It should be noted that hFcyRIIaLRdoes have the capacity to bind mIgG1, demonstrated if high concentrations of mIgGl anti-CD3 mAb are used in the T-cell proliferation assay (Tax et al., 1984; Most et al., 1992) or with immune complexes sensitized with increasing concentrations of mIgGl (Tate et al., 1992). The HR/LR polymorphism has also been described on other cell types including alveolar macrophages (Kindt et al., 1991), neutrophils (Gosselin et al., 1990),and platelets (Looney et al., 1988; Gosselin et al., 1990), but not B cells (Gosselin et al., 1990). As detailed above, the 41H16 mAb can discriminate between the HR and LR isoforms, specifically recognizing only the HR form (Gosselin et al., 1990). A polymorphism has also been described in hFcyRIIb1, whereby T y P 5in the cytoplasmic tail is substituted with an aspartic acid (Warmerdam et al., 1993). T y P 5has been proposed to form part of a signaling motif required for receptor internalization (Van den Herik-Oudijk et

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

21

al., 1994), and as such this polymorphism may result in defective receptor function. Differences in the levels of expression of hFcyRII on platelets (presumably hFcyRIIa) between individuals have also been described. The quantitative differences in FcyRIIa expression, detected using the mAb IV-3, were shown to correlate with platelet activation in response to IgG immune complexes and as such may result in increased susceptibility to immune complex disease (Rosenfeld et al., 1987). A polymorphic receptor for murine IgG2b has been described on human monocytes and EBV-transformed B lymphocytes (Holtrop et al., 1991). The molecular basis of this polymorphism has yet to be determined; however, the receptor appears to be distinct from FcyRII or FceRII (Holtrop et al., 1993a,b). A genetic polymorphism of mouse FcyRII has also been identified, defined by the Ly-17 alloantigen system, and is comprised of two alleles, Ly-lT and L ~ - l 7 that ~ , encode two polymorphic forms of mFcyRI1, designated the Ly-17.1 and Ly-17.2 antigens, respectively (Shen and Boyse, 1980; Davidson et al., 1983; Hibbs et al., 1985). The molecular basis of the polymorphism has been defined using direct sequencing of PCR-amplified mFcyRII extracellular region sequences, derived from the genomic DNA of Ly-17.1 or Ly-17.2 inbred mouse strains. The two alleles were found to differ genetically in only two codons, encoding amino acids 116 and 161; where Pro'16 and Gln161are found in the Ly-17.1 form and Leu116and LeulG1in the Ly17.2 form (Lah et al., 1990). Both these substitutions are situated in the second extracellular domain of the receptor; as antibodies specific for Ly-17.2 antigen have been demonstrated to completely inhibit the binding of IgG to mFcyRII, this suggests that residues 116 and/or 161 are closely associated with the ligand binding site and provides further evidence implicating domain 2 of FcyRII in the binding of IgG. 3. FcyRZZI a. Biochemical and Molecular Structure Human FcyRIII (CD16) is heterogeneous in size with a molecular weight ranging from 50 to 80 kDa (Fleit et al., 1982; Kulczycki, 1984; Lanier et at., 1988) (Table VI). This heterogeneity is due to extensive N-linked glycosylation of two distinct isoforms, hFcyRIIIa and hFcyRIIIb. In addition, two polymorphic forms of hFcyRIIIb have been described which also differ in N-linked glycosylation. Following deglycosylation, the hFcyRIIIa form has a molecular weight of 3334 kDa, and hFcyRIIIb has two distinct sizes of 29 and 33 kDa, which correspond to the different polymorphic forms (Selvaraj et al., 1989;

22

MARK D. HULE'lT AND P. MARKHOGARTH

TABLE VI CHARACTERISTICS OF FcyRIII Genes Human Characteristic Isoforms Alleles Chromosome localization Ig-like domains Receptor topology Associated subunits Receptor formsc Molecular mass (kDa) Apparent Protein Backbone Affinity for IgG

hFcyRIIIA hFcyRIIIa"

-

1~23-24

2

TM

hFcyRIIIb" NAUNA2 1~23-24

-

2 GPI-anchored

2 TM

y-chain, FcsRI (-chain TCR1CD3 a y e , (UrL

Mouse mFcyRIII

hFcyRIIIB

mFcyRIII

1

y-chain, FcsRI p-chain, F C E R I ~ aYz> a h b

4 2

50-80 33

50-80

29

40-60 33

z x 107 M - 1

4 0 7 M-1

4 0 7 M-1

(Ka)

Specificity hIgG mIgG Cellular distribution Regulation of expression

1=3>>>2=4 ND 3>2a>2b>>l 3>2a>2b>> 1 Macrophages Neutrophils NK cells, y6 T cells Eosinophilse monocytes (subpopulati on) TGF-p t (monocytes) TNF-a .1 (neutrophils) IFN-7, GM-CSF, GIL-4 .1 CSF t

3= 1>2>>4 1= 2a = 2b>>>3d Macrophages NK cells yS T cells IFNy

1

Note. TM, transmembrane; S, soluble; ND, not determined; GPI, glycosylphosphatidylinositol. Soluble hFcyRIIIalb generated by proteolytic cleavage of membrane isofoms. Demonstrated for mFcyRIIIa in mast cells, also for hFcyRIIIa in transfection reconstitution experiments. ay2 form expressed in macrophages q 5 ; ay2,a52 expressed in NK ceIls. mFcyRIII also binds mIgE with low affinity. IFNy induces expression hFcyRIIIb on eosinophils.

Kindt et al., 1991). Mouse FcyRIII is also heterogeneous, with amolecular weight ranging from 40 to 80 kDa (Mellman et aZ.,1988). FcyRIII are structurally similar to FcyRII, containing extracellular regions of two Ig-like domains; however, FcyRIII exhibit unique characteristics in that they differ in their forms ofmembrane anchoring. The hFcyRIIIb isoform is the only membrane FcR that is not an integral

23

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

membrane protein and instead is attached to the outer leaflet of the plasma membrane by a glycosylphosphatidylinositol (GPI) moiety (Kurosaki and Ravetch, 1989; Lanier et al., 1989a; Hibbs et al., 1989). The hFcyRIIIa isoform and mouse FcyRIII are integral membrane proteins, yet are also distinct, requiring association with additional subunits for efficient cell-surface expression (Fleit et al., 1982; Kulczycki, 1984; Lanier et al., 1988, 1989b; Kurosaki and Ravetch, 1989; Ra et al., 1989a; Hibbs et al., 1989; Anderson et al., 199Oc; Letourneur et al., 1991) (see below). As described for hFcyRI and hFcyRII, FcyRIII in humans exists in multiple isoforms where two distinct genes FcyRIIIA and FcyRIIIB (formerly FcyRIII-2 and FcyRIII-1, respectively) have been identified, each producing a single transcript that encodes the hFcyRIIIa and FcyRIIIb isoforms, respectively (Simmons and Seed, 1988; Scallon et al., 1989; Ravetch and Perussia, 1989; Peltz et al., 1989). The two hFcyRIII genes are identical in structure comprising five exons, spanning approximately 8 kb: two exons encoding the 5’ UTR and leader sequence, one exon for each of the Ig-like domains, and a single exon encoding the transmembrane, cytoplasmic tail and 3’ UTR (see Table VII) (Ravetch and Perussia, 1989; Qiu et al., 1990).The human FcyRIII genes have been mapped to the q23-24 region of chromosome 1(Peltz et al., 1989; Qiu et al., 1990) and, as detailed above, are linked to the hFcyRII genes.

TABLE VII FcyRIII GENES AND TRANSCRIPTS Name

Gene structurea L1

~FcyRIIIA

L2 D1

Transcripts D2

hFcy RIIIa

TM/C

f lI L l I L 2 1 D1 I D2 ITMK I --c

;heb

I I I 1 < L1

Fey RIIIB

D2

hFcy RIIIb

TM/C

Ll(L2 D1

D2 ITMK Aerb

L1 iFcy RIII

L2 D1

L2 D1

D2

mFcy RIII

TM/C ----c

ILlIL2

1 Dl 1

D2 lTM/C]

a Exons shown as boxes, translated regions shaded, untranslated regions open. L, leader peptide;D, extracellulardomain; ‘M, transmembrane;C, cytoplasmic tail coding regions; PA, polyadenylation site. b Crucial residue in determinationof membrane anchoring form; Phe directs TM, Ser directs GPI.

24

MARK D. HULE'lT AND P. MARKHOGARTH

The transcripts of the two hFcyRIII genes exhibit 10 nucleotide differences in their coding regions which result in only 6 amino acid differences in the hFcyRIIIa and hFcyRIIIb isoforms. Both transcripts encode receptors with extracellular regions of 191 amino acids comprising two Ig-like domains, single transmembrane domains of 21 amino acids, and cytoplasmic tails of 25 (hFcyRIIIa) or 4 amino acids (hFcyRIIIb). The different length cytoplasmic tails result from a single nucleotide change in the cytoplasmic tail coding exons of the two genes, generating an earlier stop codon in the hFcyRIIIB gene. However, a critical amino acid difference between the two forms is observed at position 203, which results in alternate membrane-anchored receptors. Human FcyRIIIb contains Se?03 which specifies a GPIlinked molecule, whereas hFcyRIIa contains Phe203which disrupts the signal for the formation of a GPI anchor, thus preserving the transmembrane and cytoplasmic tail and producing a transmembrane molecule (Kurosaki and Ravetch, 1989; Lanier et al., 1989a; Hibbs et al., 1989). The hFcyRIIIa transmembrane form is the homologue of the single mouse FcyRIII (see below) (Ravetch and Perussia, 1989). The transmembrane form of hFcyRIII requires coexpression of associated molecules for efficient cell-surface expression. In macrophages, hFcyRIIIa is associated with homodimers of the y-subunit of the highaffinity receptor for IgE (FceRI) (Kurosaki and Ravetch, 1989; Ra et al., 1989a; Lanier et al., 1989a; Hibbs et al., 1989) and in NK cells is associated with homo- and/or heterodimers of the FceRI y-subunit and the ( subunit of the T-cell receptor (TCR)-CD3 complex (Lanier et al., 198913; Anderson et al., 199Oc; Letourneur et al., 1991). These accessory chains form disulfide-linked dimeric complexes which noncovalently associate with the transmembrane region of hFcyRIIIa to enable cell-surface expression (Lanier et al., 1991) and are also important in the signaling of the receptor (see Section IV). The y- and (-chains, together with the 7-chain of the TCR-CD3 complex, are all closely related and form a family of these small subunits (Orloff et al., 1990). cDNA and genomic clones have been described in the human for both the y-subunit (Kuster et al., 1990) and (-subunit (Wiessman et al., 1988). The genes for the y- and (-subunits map to the q23-24 region of human chromosome 1, which also contains the hFcyRII, hFcyRII1, and hFceRI a-chain genes (Weissman et al., 1988; Le Conait et al., 1990). Human FcyRIIIa has also been shown to have the capacity to associate with the p-subunit of FceRI. This was demonstrated upon immunoprecipitation of the reconstituted complex, formed by cotransfection of hFcyRIIIa with the FcsRI y- and FceRI p-subunits into the mouse

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

25

mastocytoma cell line P815 (Kurosakiet al., 1992).Given the similarity in the transmembrane regions of the a-chains of FcyRIII and FcsRI, and that the transmembrane regions appear to be the principal sites where the subunits interact, it is not surprising that the FcyRIIIal y complex associates with p-chains (see below). In addition to the association of hFcyRIIIaly with p, it is also possible that FcyRIIIaly may associate with other p-like molecules such as CD20, CD37, or CD53, in NK cells and macrophages (Stamenkovic and Seed, 1988; Classon et al., 1989; Angelisova et al., 1990). Soluble forms of both hFcyRIIIa and hFcyRIIIb have also been described. These molecules are derived from the membrane receptors following proteolytic cleavage from the cell surface. The release of the GPI-anchored hFcyRIIIb isoform from neutrophils has been shown to involve serine proteases, whereas the transmembrane hFcyRIIIa isoform is released upon cleavage by metalloproteases (Harrison et al., 1991). The soluble form of hFcyRIIIb from neutrophils can be detected at high concentrations in normal human sera (Khayat et al., 1987; Huizinga et al., 1988, 1990d) and as such may have important biological role(s). Two allelic forms of hFcyRIIIb have been identified and designated NA-1 and NA-2 (Tetteroo et al., 1988; Huizinga et al., 198913, 1990a; Trounstine et al., 1990; Salmon et al., 1990; Kindt et al., 1991). The molecular basis of this polymorphism has been determined (see below). In contrast to hFcyRII1, only a single isoform of mouse FcyRIII has been described. Mouse FcyRIII is an integral membrane glycoprotein, comprising an extracellular region of 184 amino acids containing two Ig-like domains, a single transmembrane region of 21 amino acids, and a cytoplasmic tail of 26 amino acids (Ravetch et al., 1986). The extracellular region of mFcyRIII is highly conserved with mFcyRII, exhibiting 95% amino acid identity. This conservation contributed to the early classification of mFcyRIII as an isoform of mFcyRIImFcyRIIa (see above). However, mFcyRII1 diverges markedly from mFcyRII in the leader peptide, transmembrane, and cytoplasmic tail regions; these regions of mFcyRIII display homology to hFcyRII1 and specifically the transmembrane form hFcyRIIIa (Ravetch and Perussia, 1989). Mouse FcyRIII is encoded by a single gene that is structurally related to the hFcyRII1 genes comprising five exons spanning 9 kb, including two exons encoding the 5' UTR and leader sequence, a single exon for each of the Ig-like domains, and a single exon encoding the transmembrane, cytoplasmic regions and 3' UTR (see Table VII). The mFcyRIII gene is linked to the mFcyRII gene on a 160-kb geno-

26

MARK D. HULETT AND P. MARKHOCARTH

mic fragment that maps to the Ly-17 locus of chromosome 1(Kulczycki et al., 1990; Qiu et al., 1990). Mouse FcyRIII also associates with accessory chains which are required for efficient cell-surface expression (Ra et al., 1989a; Kurosaki and Ravetch, 1989) and signaling of the receptor (Amigorena et al., 199213; Bonnerot et al., 1992) (see Section IV). However, in contrast to hFcyRII1, mFcyRIII has only been found to associate with the FcsRI y-subunit, as the mCD3lTCR (-subunit does not promote cellsurface expression of mFcyRIII (Kurosaki and Ravetch, 1989; Ra et aZ., 1989b). Interestingly, mouse FcyRIII will associate with the hCD3/TCR (-subunit (Kurosaki and Ravetch, 1989; Ra et al., 1989b). The mouse y-subunit cDNA has been cloned, and, as for human y, the gene mapped to chromosome 1 in the same region in the lowaffinity FcyR genes (Ra et al., 1989b; Huppi et al., 1989).As described for hFcyRIIIa, mFcyRII1 has also been shown to associate with the p-chain of FcsRI in mast cells. This was demonstrated by immunoprecipitation of an endogenously expressed FcyRIIIalylp complex from the mouse mast cell line, MC9, and by reconstitution of the complex by transfection into COS-7 cells (Kurosaki et al., 1992). Rat FcyRIII cDNAs have also been cloned and form a family of multiple isoforms, in contrast to the single mouse FcyRIII. Several distinct isoforms have been isolated, all encoding transmembrane receptors that require subunit association for expression (Zeger et al., 1990; Farber and Sears, 1991). As described for mouse FcyRIII, only the FcsRI y-subunit and not the endogenous rat CD3lTCR (-subunit (in contrast to hCD3ITCR () promotes efficient expression of rat FcyRIII (Farber and Sears, 1991; Farber et al., 1993). The existence of multiple rat FcyRIII genes has been suggested with an organization similar to the mouse and human FcyRIII genes (Farber and Sears, 1991). Examination of the transmembrane region of hFcyRIIIa, mFcyRIII, and the rat FcyRIII isoforms reveals a conserved stretch of eight amino acids (LFAVDTGL) including a negatively charged aspartic acid residue. This region is unique to these FcyR and the a-chain of the tetrameric FcsRI, all of which associate with accessory subunits. The yand (-subunits also have highly conserved transmembrane regions, suggesting that the interaction between these heterologous proteins involves their transmembrane regions. This possibility has been suggested by examination of the association of hFcyRIIIa and the FcsRI y- and CD3 (-subunits, where truncation of the cytoplasmic tails of these molecules did not effect cell-surface expression of hFcyRIIIa (Lanier et al., 1991). A molecular model of the interaction between

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

27

the conserved transmembrane regions of these FcRs and the associated subunits has been proposed (Farber and Sears, 1991).

b. Ligand Affinity and Specificity Human and mouse FcyRIII comprise a second class of low-affinity FcyR. The two hFcyRIII isoforms have been reported to display slightly different affinities for monomeric IgG, the GPI-anchored form (hFcyRIIIb) has an affinity of K , < lo7 M-' (Fleit et al., 1982; Kulczycki, 1984; Anderson and Looney, 1986; Simmons and Seed, 1988), whereas the transmembrane form (hFcyRIIIa) has a higher affinity of K , = 2 x 107M-' (Vance et al., 1992).As described for hFcyRI1, data on the IgG isotype binding specificity of the hFcyRIII isoforms is incomplete. However, the specificity of the GPI-anchored form of hFcyRIII for hIgG subclasses has been determined using dimeric complexes, and the receptor preferentially binds hIgG3 and hIgG1, but not hIgG2 or hIgG4 complexes (Kulczycki, 1984; Huizinga et al., 1989a). The specificity of the transmembrane form of hFcyRIII for hIgG subclasses has not been determined. Reports on the specificity of binding of mouse IgG isotypes by hFcyRIII are conflicting; however, it appears that both the transmembrane and GPI-anchored forms bind mIgG2a, mIgG3, and to a lesser extent mIgG1, but not IgG2b (Kipps et al., 1985; Anasetti et al., 1987; Simmons and Seed, 1988; Braakman et al., 1993). Human FcyRIIIb has been shown to have the unique ability to interact with lectins, probably via its high mannose containing oligosaccharides. This has been demonstrated as the phagocytosis ofconcanavalin A-treated erythrocytes and nonopsonized Escherichia coli by human neutrophils and can be inhibited with the anti-hFcyRIII mAb 3G8, aggregated IgG, and monosaccharides such as D-mannose (Salmon et al., 1987; Kimberley et al., 1989).Thus hFcyRIIIb appears to be able to bind ligands other than IgG through lectin-carbohydrate interactions. These findings also suggests that oligosaccharides may contribute to the integrity of the IgG binding site on hFcyRIIIb. Mouse FcyRIII exhibits identical ligand binding characteristics to mFcyRI1, displaying a similar low affinity for monomeric IgG (K,< lo7 M - ' ) and the same specificity for both mouse and human IgG isotypes, preferentially binding mouse IgG1, 2a, and 2b and human IgGl and 3 (Unkeless, 1977, 1979; Heusser et al., 1977; HaeffnerCavaillon et al., 1979a; Teillaud et al., 1985; Lopez et al., 1985). A recent study has demonstrated that mFcyRIII also binds mIgE with low affinity, as described for mFcyRII (Takizawa et al., 1992). The

28

MARK D. HULETT AND P. MARKHOGARTH

similar interaction of mFcyRIII and mFcyRII with Ig is not surprising as the extracellular ligand binding regions of these receptors are highly conserved with 95% amino acid identity. c. Cell Distribution and Monoclonal Antibodies The expression of the two different hFcyRII1 isoforms is cell specific (Table VI). The GPI-anchored FcyRIIIb is expressed exclusively on neutrophils, whereas the transmembrane FcyRIIIa is expressed on macrophages and NK cells (Simmons and Seed, 1988; Selvaraj et al., 1988,1989; Peltz et al., 1989; Scallon et al., 189; Ravetch and Perussia, 1989; Hibbs et al., 1989; Lanier et al., 1989a; Edberg et al., 1989; Perussia and Ravetch, 1991). The latter form has also been demonstrated on a small population of freshly isolated monocytes (Passlick et al., 1989; Anderson et at., 1990b) and on some T cells (Braakman et al., 1993). Human FcyRIII has also been observed on mesangial cells of the kidney (Sedmak et al., 1990) and on placental trophoblasts (Sedmak et al., 1991). Immunohistochemical staining of normal lymphoid and nonlymphoid tissues with anti-hFcyRII1 mAb has demonstrated strong staining of mantle zone cells and interfollicular zone cells (Tuijnman et al., 1993). The expression of hFcyRIII has been shown to be influenced by cytokines; the transmembrane form on monocytes was found to be upregulated by TNF-P (Welch et al., 1990; Phillips et at., 1991), and the GPI-anchored form on neutrophils can be upregulated by IFN-y, GM-CSF, and G-CSF (Buckle and Hogg, 1989) and also downregulated by TNF-a, which has no effect on the transmembrane form (Mendel et al., 1988). It has also been reported that the GPI-anchored form can be induced on eosinophils by IFN-y (Hartnell et at., 1992).In addition, IL-4 downregulates monorgte hFcyRII1, as for the other FcyR classes (Te Velde et al., 1990). Mouse FcyRIII has a similar cell distribution to hFcyRIIIa and has been described on macrophages, NK cells, subpopulations of T cells, and mast cells (Ravetch et al., 1986; Weinshank et al., 1988; Katz et al., 1990; Benhamou et al., 199Oa). mFcyRIII has also been described on early fetal thymocytes (Rodewald et al., 1993). The expression of mFcyRIII on macrophages is also modulated by IFN-y (Weinshank et al., 1988). Several anti-hFcyRIII mAb have been described and include 3G8 and 4F7 (Fleit et al., 1982), VEPl3 (Rumpold et al., 1982), Leu l l a and Leu l l b (Lanier et al., 1983, 1985), B73.1 (Perussia et al., 1984), GRM-1 (Phillips and Babcock, 1983), CLB GranII (Werner et al., 1988), 1D3 (Tetteroo et al., 1988), and BW209/2 (Huizinga et at., 1990d) (see Table 111).Most of the mAb block the binding of IgG to

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

29

the receptor; however, mAb BW209/2 binds to hFcyRIII when it is occupied with IgG (Huizinga et al., 1990d).A number ofthe hFcyRIII mAb show differential reactivity with the NAUNA2 polymorphic forms, with mAb B73.1 and CLB GranII being specific for the NA1 form, and mAb GRM-1 for the NA-2 form (Huinzinga et al., 1989b; Trounstine et al., 1990; Salmon et al., 1990) (see below). The only mAb demonstrated to bind mFcyRIII is the mAb 2.4G2, which also detects mFcyRII (Unkeless et al., 1979). d. Polymorphism A genetically determined structual polymorphism has been described for hFcyRIIIb, identified by biochemical analysis and reactivity with mAb (Tetteroo et al., 1988; Huizinga et al., 1989b, 199Oa; Trounstine et al., 1990; Salmon et at., 1990; Kindt et al., 1991). The polymorphism has been designated the neutrophil antigen (NA) system and comprises two allelic forms, hFcyRIIIbNA-'and hFcyRIIIbNA-',which exhibit phenotypic frequencies in Caucasians of 37% and 63%, respectively (Lalezari, 1984). The polymorphism is apparent in the different molecular masses observed for the two allotypes following deglycosylation; the NA-1 form has a mass of 29 kDa and the NA-2 form a mass of 33 kDa (Ory et al., 1989; Huizinga et al., 1990a). The two allotypic forms are also distinguishable with mAb as described above. The molecular basis of this polymorphism has been determined and arises from a four-amino acid difference between the two forms, which results in the loss of two N-linked glycosylation sites in the NA-1 form, which has four sites in contrast to the six sites of NA-2 (Ravetch and Perussia, 1989; Ory et al., 1989). In addition, a Taq-1 restriction length fragment polymorphism is also associated with the two hFcyRIIIb alleles (Ravetch and Perussia, 1989; Ory et al., 1989). Individuals who do not express hFcyRIIIb have also been identified. These include a patient with systemic lupus erythematosus (SLE) (Clark et al., 1990) and two healthy individuals who exhibited no sign of increased susceptibility to infection or elevated levels of circulating immune complexes (Huizinga et al., 1990b). The defect in each case appears to be as a result of a disorganized or absent hFcrRIIIB gene. These findings raise the question of the functional significance of hFcyRIIIb and again suggest that the loss of a single class of FcyR can be compensated by the other FcyR. No polymorphism has been described for mFcyRII1; however, based on the high amino acid identity with mFcyRII (which exhibits the Ly-17 polymorphism), and as the mFcyRIII cDNA was isolated

30

MARK D. HULETT AND P. MARKHOGARTH

from an Ly-17.2 mouse strain, it would be interesting to determine if the mFcyRIII sequence is also different in Ly-17.1 strains. B. FcsR Two distinct classes of receptors for the Fc portion of IgE, FcsRI, and FceRII have been defined on the basis of differential affinity for IgE, reactivity with mAb, cell distribution, biological function, and molecular cloning. The isolation of cDNA clones for these receptors has indicated that FcsRI and FcERIIare structurally unrelated, FcsRI belonging to the Ig superfamily and closely related to the leukocyte FcyR, whereas FcsRII belongs to a family of animal lectins. As this review focuses on the Ig superfamily FcR, only FceRI is discussed in detail (summarized in Tables VIII and IX); however, FceRII is described briefly for completeness (for reviews see Spiegelberg, 1984; Metzger et al., 1986, Kinet, 1990; Ravetch and Kinet, 1991; Conrad, 1990; Metzger, 1992a; Delespesse et al., 1992).

TABLE VIII CHARACTERISTICS OF FcsRI Characteristic Affinity for IgE' (K,) Specificity IgE Associated subunitsb Receptor forms Ig-like domains Receptor topology Molecular mass (kDa) Apparent Protein backbone Chromosomal localization Cellular distribution

Human FcsRI

10'oMul-' hIgE, rtIgE, mIgE a,p, y .BY29

ayzc

2 (a) TM

45-65,32,7-9 26.4,25.9,7-8 lq23, llq13,lq23

Rat FcsRI

Mouse FcERI

10lOM-l 10'0M-1 rtIgE, mIgE only rtIgE, mIgE only Q, P, Y a, Y

cvavz

cvarz

TM

TM

45-65,32,7-9 25.2,27, 7.8 ND

45-65,32,7-9 25.8,25.9,7.8 lq, 19,lq

2 (4

Mast Cells Mast Cells Basophils Basophils Langerhans Cells Eosinophils Monocytes (activated)

2 (4

Mast Cells Basophils

Note. TM, transmembrane; ND, not determined. Affinity of receptors for their species specific ligand. a-chain, ligand binding subunit. hFceRI a-chain only requires y-subunit for expression, thus may also exist in ay2 form.

31

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

TABLE IX FceRI GENESAND TRANSCRIPTS Name rtFceRIa

mFc,RIa

Gene structure'

Ll

L2 D1

D2

TM/C

L1

L2 D1

D2

TM/C

L1

L2 D1

D2

TM/C

hFcERIa

- [LIIL~ -

Transcripts

rFc, RIa

[ L i I u 1 DI I DZ ITMK) mFcE RIa

I

DI

I DZ ITM/C\

hFc, RIa

I L ~ I IDI ~ I DZ ITME]

Exons shown as boxes, translated regions shaded, untranslated regions open. L, leader peptide: D, extracellulardomain: TM, transmembrane; C, cytoplasmic tail coding regions; PA, polyadenylation site.

1. FcsRl

a. Biochemical and Molecular Structure FcsRI, also known as the high-affinity IgE receptor, has been characterized at a molecular level in three different species, mouse, rat and human, and has been defined structurally as a tetrameric complex of three distinct polypeptides, comprising an @-subunit(the IgE binding chain) that is homologous to the FcyR, a @subunit, and a dimer of two y-subunits. The early biochemical characterization focused on rat FcsRI of the rat basophilic leukemia cell line RBL-2H3 (Kulczycki et al., 1974). Initial attempts to purify the receptor identified a single polypeptide with an apparent molecular weight of 50-60 kDa (Conrad and Froese, 1976; Kulczycki et al., 1976; Kanellopolous et al., 1979). However, other studies suggested that the receptor was composed of more than one polypeptide (Holowka et al., 1980). This was subsequently found to be correct, as the polypeptide isolated initially (the a-subunit) was found to be associated with a single p-chain of 33 kDa (Holowka et al., 1980; Holowka and Metzger, 1982; Perez-Montfort et al., 1983), and two disulfide-linked y-subunits each of 7-9 kDa (Perez-Montfort et al., 1983; Alcaraz et al., 1984). The initial failure to detect these additional subunits was due to the sensitivity ofthe noncovalent associ-

32

MARK D. HULETT AND P. MARKHOGARTH

ation between the subunits to mild detergents, with purification of the intact tetrameric complex requiring protective phospholipid or submicellular concentrations of detergent (Rivnay et al., 1982; Kinet et al., 1985). Several mAb were also raised against rat FcsRI and allowed further characterization of the membrane topology of the receptor subunits (Basciano et al., 1986). Based on this biochemical data it was suggested that FceRI was a tetrameric complex comprising noncovalently associated subunits: the a-subunit, a highly glycosylated polypeptide expressed on the outer surface of the cell, and two non-glycosylated intramembrane components, the p-subunit, and a dimer of two disulfide-linked y-subunits (Metzger et al., 1983, 1986). The cDNAs for each of the subunits of FcsRI in the rat, mouse, and human have been cloned, which has enabled their molecular structures to be determined and a model for the topology of the FceRI receptor complex to be proposed (Blank et al., 1989). a-Subunit. cDNA clones for the a-subunit have been isolated from all three species (Kinet et al., 1987; Kochan et al., 1988; Shimizu et al., 1988, Liu and Robertson, 1988; Ra et al., 1989b) and encode integral membrane glycoproteins with extracellular regions of two Ig-like domains spanning 180 (human) or 181 (rat, mouse) amino acids; a single transmembrane region of 21 amino acids; and cytoplasmic tails of 31 (human), 25 (mouse), or 20 (rat) amino acids. A single transcript has been identified in each species. Three additional rat FcsRI a-chain cDNA clones have been reported but have substantial differences from the cloned FceRI a-chain gene and as such are likely to be cloning artifacts (Liu and Robertson, 1988). The predicted human, mouse, and rat a-chain protein products exhibit substantial sequence identity (38%), but are the least conserved of the three FcsRI subunits (Ra et al., 198913).The leader peptide and the cytoplasmic domains are the least conserved, exhibiting 17 and 16% amino acid identity between the species, respectively. However, the extracellular and transmembrane regions are more highly conserved, displaying 42 and 62% amino acid identity, respectively, when sequences are compared between these species (Ra et al., 1989b). The high degree of sequence divergence of the predicted a-chain cytoplasmic tails suggest that the cytoplasmic tail ofthis subunit is not involved in a crucial receptor function. In contrast, the high level of amino acid identity observed in the transmembrane region across the species, which all contain a conserved consecutive 8-amino acid motif (LFAVDTGL), suggests this region performs a specific function (Kinet and Metzger, 1990). The finding that this motif is also conserved in the transmembrane region of mFcyRII1, hFcyRII1, and rat FcyRIII, which like the FcsRI a-chain

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

33

are all receptors that require association with the y-subunit for cellsurface expression (see above), provides strong evidence to suggest this region interacts with the y-subunit (Farber and Sears, 1991). Indeed, as described for hFcyRIIIa, recent mutagenesis experiments indicate this is also the case for the FceRI a-chain (see below) (VarinBlank and Metzger 1990). The predicted FcsRI a-chains are homologous to all the FcyR described above, but are most closely related to the FcyRIII subclass (Ra et al., 1989b; Ravetch and Kinet, 1991) (Table IX). This is demonstrated on comparison of mouse FceRIa with mouse FcyRIII, which exhibit an overall 33% amino acid identity across their entire sequence, with 35 and 48% identity in their extracellular and transmembrane regions, respectively. Furthermore, of the 95 conserved residues in the extracellular regions of the rat-mouse-human FceRI a-chains, 61 are also found in both human and mouse FcyRIII, suggesting the unique 34 residues of the a-chains may be specific for IgE (Ravetch and Kinet, 1991). This high degree of conservation suggests that the FceRI a-chain and FcyRIII genes probably arose from a common ancestor by gene duplication. Indeed, the cloning of the rat and mouse and human FceRI a-chain genes has demonstrated that they share a high degree of structural conservation with each other and the mouse and human FcyRIII genes-containing five exons, two of which encode the 5' UTR and leader sequence, one exon for each of Ig-like domains, and a single exon encoding the transmembrane, cytoplasmic tail and 3' UTR (Tepler et al., 1989; Ye et al., 1992; Pang et al., 1993). The human FceRI a-chain gene has been mapped to the same region on chromosome 1 as the low-affinity hFcyR genes-band lq23 (Le Conait et al., 1990). It has also been demonstrated that the mouse FceRI a-chain gene is linked to the mouse low-affinity FcyR genes on chromosome 1 (Huppi et al., 1988). P-Subunit. cDNA clones of the @subunit of FceRI have been isolated from the rat (Kinet et al., 1988), mouse (Ra et al., 1989b), and human (Kuster et al., 1992). Two mRNA species of both mouse and rat p FceRI (1.75 and 2.7 kb) which arise by alternate polyadenylation have been observed. Two transcripts of human FceRIP have also been described, detected as a doublet around 3.9 kb (Kuster et al., 1990). The predicted rat, mouse, and human p-subunits are 243, 235, and 244 amino acids in length, respectively, and are highly homologous, displaying 69% amino acid identity. Based on hydrophobicity plots and studies with mAbs, it has been proposed that the FceRI P-subunit comprises four membrane spanning regions with both the N- and Ctermini in the cytoplasm (Kinet et al., 1988; Ra et al., 1989b). The

34

MARK D. HULETT AND P. MARKHOGARTH

human FcsRI p-subunit gene has been isolated and appears to be a single copy gene comprising seven exons spanning 10 kb (Kuster et al., 1990). The 5’UTR and part of the N-terminal cytoplasmic tail are encoded by exon 1, the first transmembrane region is encoded by exons 2 and 3, transmembrane 2 by exons 3 and 4, transmembrane 3 by exon 5, transmembrane 4 by exon 6, and the C-terminal cytoplasmic tail and 3’UTRs in exon 7. The mouse and rat FceRI p-subunit genes have not been isolated; however, the mouse gene is believed to be encoded by a single gene that maps to chromosome 19, linked to the Ly-l locus (Huppi et al., 1989). The human FceRI p-gene has been mapped to chromosome l l q 1 3 (Sandford et al., 1993). y Subunit. cDNA clones have been isolated for the FcsRI ysubunit in the rat (Blank et al., 1989), mouse (Ra et al., 1989b) and human (Kuster et al., 1990).The y-subunit in all three species is highly conserved, the predicted polypeptide products exhibiting 86% amino acid identity (Kuster et al., 1990).FcsRIy is a small integral membrane protein, with a single transmembrane spanning region of 21 amino acids, a small extracellular region of only 5 residues, and a larger cytoplasmic tail of 42 amino acids. The y-subunit has been demonstrated to exist in dimeric form through the formation of a disulfide bond between the N-terminal cysteine residue (Varin-Blank and Metzger, 1990) and exhibits homology to the (-and 7-chains of the TCR/ CD3 complex, with which it forms a family of disulfide-linked dimers (Orloff et al., 1990). As described above, the y-subunit of FcERI also associates with mouse and rat FcyRIII, human FqRIIIa, FcyRI and FcyRII. In all these receptors (with the exception of FcyRI and FcyRII) the y-subunit is essential for efficient cell-surface expression and also plays a crucial role in signal transduction. The human ysubunit gene has been isolated and mapped to the chromosome 1923, the same region that contains the FceRI a-chain gene and the lowaffinity FcyR genes (Kuster et al., 1990). The mouse y-subunit gene has also been mapped to a region containing the FceRI a-chain, FcyRII and FcyRIII genes on chromosome 1 (Huppi et al., 1989). This close linkage of the genes encoding the FceRI a- and y-subunits, and the low-affinityIgG receptors, suggests the possibility ofcoordinate regulation of these FcR genes. The cDNA cloning of each of the subunits of FceRI in the rat, mouse, and human has enabled cotransfection studies to be performed to assess the requirements for efficient cell-surface expression of the receptor. Initial experiments with the rat FceRI a-chain demonstrated that this subunit could not be expressed on the cell surface following its transfection in isolation into COS-7 (Kinet et al., 1987; Shimizu et

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

35

al., 1988). Subsequently, this has been shown for both the human and mouse FceRI a-chains; however, it should be noted that cell-surface expression does occur, but with an extremely low efficiency making detection difficult (Ra et al., 198913). Efficient cell-surface expression of the a-chain has been shown to require the coexpression of the yand/or /+subunits, and these requirements are different for the rodent and human receptors. In the rat and mouse systems, both the y- and p-subunits are required for efficient cell-surface expression of the asubunit, in contrast to the human a-subunit which requires coexpression of only the y-subunit (which can be of rat, mouse, or human origin) (Miller et al., 1989; Ra et al., 1989b; Kuster et al., 1990). The cotransfection the mouse, rat, or human @chain with hFceRIa and hFceRIy does not increase expression efficiency (Kuster et al., 1990). These findings raise the possibility that hFceRI can exist as an a ( ~ ) ~ complex in vivo and may therefore have the capacity to mediate a distinct intracellular signal. Recent mutagenesis experiments have been performed on the rat FceRI subunits to assess the roles of different regions of the subunits in association and cell-surface expression (Varin-Blank and Metzger, 1990). Truncation of the cytoplasmic tails of any or all of the subunits had little effect on cell-surface expression of the receptor. However, even minor changes in the transmembrane regions resulted in reduced expression levels. These experiments suggest that the transmembrane regions are critical for optimal expression of rat FceRI, and a model to describe the molecular interaction between the transmembrane regions of the subunits has been proposed (Varin-Blank and Metzger, 1990). An interesting finding from this study was that the human achain, when coexpressed with a rat y-chain lacking a cytoplasmic tail, was not expressed on the cell surface, again suggesting that the human and rodent receptors assemble differently. As described above, hFcyRIIIa has been shown to associate with both FceRI y- and the TCR/CD3 (-subunits, and based on the high homology of these subunits and the conserved nature of the hFcyRIIIa and hFceR1 a-chain transmembrane regions, it might be expected that the (-subunit could also associate with the hFceRI a-chain. Indeed, it has been demonstrated recently that the 6-subunit is able to substitute for the y-subunit in the assembly and functional expression of rat FceRI, in a Xenopus oocyte expression system (Howard et al., 1990). However, such an association would not be expected to occur in vivo, as the (-chain appears not to be coexpressed with the FcERI a-chain. Hamawy et a1 (1992) have also indicated that other molecules may be closely associated with the FceRI on the cell surface. A monoclonal

36

MARK D. HULE'IT AND P. MARKHOGARTH

antibody (BD6) that detects a 40-kDa molecule on the surface of RBL2H3 cells blocks IgE binding. This molecule can be chemically crosslinked to the FcsRI complex indicating its likely proximity to the receptor. It is also of interest that a number of novel proteins can be coprecipitated with the FceRI y-subunit (Schoneich et al., 1992).

b. Ligand Affinityand Specificity FcsRI of rat, mouse, and human all bind monomeric IgE with an affinity of approximately 10" M-' (Kulczycki and Metzger, 1974; Ishizaka et al., 1985; Miller et al., 1989). Although FcsRI of each species specifically binds IgE, the specificity for IgE from different species varies. Human FcsRI binds human, rat, and mouse IgE, although rodent IgE binds less well. In contrast, mouse and rat FcsRI only bind rodent IgE, not hIgE (Conrad et al., 1983). Of interest is the recent finding that FcsRI on rat RBL cells can bind EBP, a P-galactoside binding lectin shown to be identical to Mac-2 (Frigeri et al., 1993). c. Cell Distribution and Monoclonal Antibodies FceRI traditionally has been thought to have a unique cell distribution, being exclusively expressed on mast cells and basophils (Metzger et al., 1986; Metzger, 1988; Kinet and Metzger, 1990; Ravetch and Kinet, 1991). However, more recently it has become apparent that FceRI has a broader distribution, as is also found on Langerhans cells (Bieber et al., 1992; Wang et al., 1992), eosinophils (Abdelillah et al., 1994), and activated monocytes (Maurer et al., 1994). Numerous mAb detecting the rat FcsRI a-chain have been described (Basciano et al., 1986). Recently, a number of anti-human FcsRI achain mAb have also been produced (see Table 111)(Riske et d.,1991; and P. M. Hogarth, unpublished). These mAb have been divided into inhibitory and noninhibitory classes on the basis of their capacity to block the binding of IgE to hFcsRI. The inhibitory mAb included 15A5, 12E7,6F7, and 4B4. MAb 15A5 was specifically mapped to the region comprising amino acids 100-115 in the second extracellular domain, and mAb 12E7,6F7, and 4B4 recognized epitopes that were identical to or overlapping that detected by 15A5. The noninhibitory mAb included 2237,1lB4,5D5,8C8,29C9, and 39D5, of which 2237, 5D5, and 8C8 were shown to have epitopes in the first extracellular domain, as all competed with l l B4 which was shown to recognize the peptide corresponding to residues 18-23 of domain 1 (Riske et al., 1991). MAb 29C9 recognized an epitope that was proposed to comprise regions from both domains, since it was able to block the binding of both 15A5 and 2237. The epitope of mAb 39D5 was not mapped.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

37

These findings strongly suggest that the second extracellular domain of the hFcsR1 a-chain is important in the binding of IgE, which supports the findings of others using direct binding site mapping approaches (Hulett et al., 1993; Robertson, 1993; Mallimaci, 1993).

d. Polymorphism

No polymorphisms have been reported for FcsRI in the rat, mouse, or human.

2. FcsRZZ a. Biochemical and Molecular Structure FcsRII (CD23)is a single-chain type I1 integral membrane glycoprotein with an apparent molecular mass of 45-50 kDa (Conrad, 1990; Delepesse, 1991,1992). Complementary DNA cloning studies suggest FcsRII contains a large C-terminal extracellular region of 277 amino acids, a single transmembrane domain of 20 amino acids, and a short N-terminal intracytoplasmic tail of 23 residues (Kikutani et al., 1986a; Ikuta et al., 1987; Ludin et al., 1987).The extracellular region has been

proposed to comprise a C-terminal C-type (Ca2+-dependent)lectin domain, homologous to a family of proteins including the adhesion proteins termed “selectins” and the asialoglycoprotein receptor (Wong et al., 1991). Located on the C-terminal side of the lectin domain is an “inverted RGD” sequence (Arg-Gly-Asp) which has been proposed to have a role in cell adhesion-in a manner similar to the RGD sequence of the integrins. On the N-terminal side of the lectin domain is a repetitive region containing five heptadic repeats of hydrophobic leucineholeucine residues, predicted to form an a-helical “stalk” region which mediates trimer formation (Beavil et al., 1992; Dierks et al., 1993). A single N-linked glycosylation site is situated on the transmembrane side of the stalk region. A single gene has been isolated for human FcsRII and mapped to chromosome 19 (Suter et al., 1987; Wendel-Hansen et al., 1990). The gene comprises 11 exons spanning 13 kb. The mouse FcsRII gene structure is almost identical to the human, with the exception that it contains an additional exon encoding a fourth repeat region (3 in the human) (Richards et al., 1991). Two transcripts (designated A and B) are encoded by the human FcsRII gene, differing only in their 5’ untranslated and intracytoplasmic tail encoding regions. These transcripts are derived from the use of different promoters which control different first exons (Yokota et al., 1988). Mouse FceRII exhibits 52% amino acid identity with human FcsRII (Delepesse et al., 1992). Similarly, in the mouse two isoforms of FcsRII have been described that

38

MARK D. HULETT AND P. MARKHOGARTH

arise by the same mechanism as that for the human forms (Richards et al., 1991). Soluble forms of human FceRII have been described and arise by proteolytic cleavage of the membrane form. Initially a 37- or 33-kDa fragment is released following cleavage at amino acid 82 in the "stalk" region. Additional soluble forms are derived from these by further proteolysis steps, producing fragments of 29, 25 and 16 kDa. All of these soluble fragments retain the capacity to bind IgE (Letellier et al., 1989, 1990).

b. Ligand AfJinity and Speci3city FcsRII binds monomeric IgE with an affinity of K , < lo7 M-' and is referred to as the low-affinity IgE receptor (Conrad, 1990; Delespesse et al., 1991, 1992). FceRII also binds CR2-a membrane protein found on B cells, follicular dendritic cells, T cells, and basophils (Aubry et al., 1992; Pochon et al., 1992).As such CR2 is referred to as a counter-receptor for FceRII. The IgE and CR2 binding functions of FceRII reside entirely in the lectin domain (reviewed in Sutton and Gould, 1993). c. Cell Distribution and Monoclonal Antibodies In the human, FceRII is expressed on a diverse range of hematopoietic cells including T and B cells, monocytes, eosinophils, platelets, follicular dendritic cells, Langerhans cells, and epithelial cells of the bone marrow and thymus (Conrad, 1990; Delepesse et al., 1991,1992). Expression of the two FceRII forms is regulated in a tissue-specific manner. The FceRIIa form is expressed only on antigen-activated B cells; however, following differentiation into Ig secreting plasma cells expression is lost (Kikutani et al., 1986b; Snapper et al., 1991). FcsRIIb is expressed on all the cell types outlined above following induction with IL-4 (Delespesse et al., 1991, 1992). Mouse FcsRII has been described on B cells, monocytes, and eosinophils (Delepesse et al., 1992). C. FcaR FcaRs have been described on hematopoietic cells in both the human and mouse. The FcaR on human myeloid cells has been most extensively characterized and is a member of the Ig superfamily, structurally related to the FcyR and FceRI. The existence of distinct lymphocyte FcaR has also been suggested; however, lymphocyte FcaR are far less well-defined and remain controversial. This section of the review, therefore, focuses primarily on human FcaRI.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

39

1. FcaRl

a. Biochemical and Molecular Structure Human FccuRI (CD89) is a heavily glycosylated protein of 5575 kDa (Albrechtsen et al., 1988; Monteiro et al., 1990, 1992; Mazengera and Kerr, 1990)(Table X). The observed molecular heterogeneity has been shown to arise from variable glycosylation of a single protein product. The removal of N-linked oligosaccharides by treatment with N-glycanase reveals two molecular species of 32 and 36 kDa (Monteiro et al., 1990, 1992). The 32-kDa form has been proposed to represent the protein core, with the 36-kDa form resulting from incomplete deglycosylation rather than being the product of an alternatively spliced transcript (Monteiro et al., 1992). The mouse homologue of human FccuRI has not been defined. A human FcaRI cDNA clone has been isolated and the predicted amino acid sequence indicates an integral membrane glycoprotein of 287 amino acids comprising an extracellular region of 206 amino acids containing six potential N-linked glycosylation sites, a single transmembrane region of 19 amino acids, and a cytoplasmic tail of 41 amino acids (Maliszewski et al., 1990).The extracellular region of this FcaR is homologous to that of the other Ig superfamily FcRs, FcyRII, FcyRIII, FcsRI a-chain, and the first two domains of FcyRI. However, it is

TABLE X CH~RACTERISTICS OF FcaR Characteristics Chromosome localization Ig-like domain Receptor topology Associated subunit Receptor forms Molecular mass (kDa) Apparent Protein backbone Affinity for IgA Specificity Cellular distribution Regulation of expression

Human FcaRI'

19

2

TM FcERIy-chain a,4

55-756

32

5 x 107M-* Monomeric and polymeric IgAl and IgA2 Monocytes, macrophages, neutrophils, eosinophils PMA t (neutrophils), Ca2+ionophore t (eosinophils)

Human FcaRI only FcaR cloned; mouse FcaRI form not reported. Distinct FcaR on human and murine lymphocytes but not described at biochemical/molecular level. Cell type dependent; eosinophils 70-100 kDa.

40

MARK D. HULETT AND P. MARKHOGARTH

more distantly related to these receptors than the FcyR and FccRI are to each other, suggesting FcaRI diverged from a common ancestor early in the evolution of the Ig superfamily FcR (Maliszewski et al., 1990). The gene encoding hFcaRI has recently been mapped to chromosome 19q3.4 (Kremer et al., 1992) and as such is not linked to the other Ig superfamily FcR, which are all found on chromosome 1q23-24 (with the exception of mFcyRI, see above). Genomic clones encoding hFcaRI have not as yet been isolated. Transfection experiments have demonstrated that hFcaRI does not require accessory subunits for cell-surface expression (Maliszewski et al., 1990). However, recent studies suggest that it does associate with the y-subunit of FcERI (L. Pfefferkorn, personal communication), also known to associate with hFcyRI, FcyRII, and FcyRIII (see above). Interestingly, examination of the amino acid sequence of hFcaRI transmembrane region does reveal some homology to the 8-amino acid motif in the transmembrane region of human and mouse FcyRIII and FcERIa-chain, believed to be crucial for interaction with the y-subunit, including the presence of a charged residue (Arg230)(Maliszewski et al., 1990; Farber and Sears, 1991). As described above, FcaR have also been postulated to occur on subpopulations of human and murine T and B cells (reviewed in Mestecky and McGhee, 1987; McGhee et al., 1989; Kerr, 1990; also Lum et al., 1979; Gupta et al., 1979; Sjoberg, 1980a; Lynch and Sandor, 1990; Millet et al., 1989; Roa et al., 1992; Aicher et al., 1992). The lymphocyte FcaR appear to be structurally distinct from FcaRI; however, their biochemical nature and molecular structures have not yet been determined.

b. Ligand Specificity and Affinity Human FcaRI is specific for IgA, binding both monomeric and polymeric forms of IgAl and IgA2 (Albrechsten et al., 1988; Monteiro et al., 1990; Stewert and Kerr, 1990). The receptor binds IgA with high affinity, binding monomeric IgA with an affinity of K , = 5 x lo' M-' (Mazengera and Kerr, 1990). It should be noted that the binding of IgA by hFcaRI is species specific, binding only human IgA and not mouse IgA (M. Kerr, personal communication). c. Cell Distribution and Monoclonal Antibodies Human FcaRI is expressed on monocytes, macrophages, neutrophils, and eosinophils (Gauldie et al., 1983; Chevailler et al., 1989; Fanger et al., 1983; Maliszewski et al., 1985; Abu-Ghazaleh et al., 1989; Monteiro et al., 1990,1992,1993; Shen et al., 1989; Mazengera and Kerr, 1990; Monteiro et al., 1993). FcaRI has also recently been described on human mesangial cells (Gomez-Guerrero et al., 1993).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

41

There is some evidence to suggest that hFcaRI expression is differentially regulated on myeloid cells, as PMA enhances FcaRI expression on the monocyte cell lines U937 and PLB985 but not on eosinophils, whereas Ca2+ ionophores enhance FcaRI expression on eosinophils but not monocyte cell lines (Monteiro et al., 1993). In support of this, FcaR expression is elevated on eosinophils but not neutrophils of allergic individuals (Monteiro et al., 1993). In addition, the treatment of neutrophils with GM-CSF or G-CSF has been shown to reduce the number of IgA binding sites, while increasing the affinity of the remaining receptors (Weisbart et al., 1988); however, it is not known whether this is due to the induction ofa high-affinity FcaR (i-e.,FcaRI) or the modification of a preexisting low-affinity receptor. A number of monoclonal antibodies have been described that specifically bind hFcaRI, including My43 (Shen et d.,1989) (used in the expression cloning of the receptor), A3, A59, A62, and A77 (Monteiro et al., 1990). Only My43 has been shown to inhibit the binding of IgA to hFcaRI, the remaining mAb appearing to bind determinants located outside the IgA binding site. As mentioned, novel FcaR apparently distinct from FcaRI have been postulated to exist on subpopulations of human and mouse T and B cells; however, no mAb have been described that recognize these proposed receptors. d. Polymorphisms No polymorphisms have been described to date which identify additional isoforms of hFcaRI to the original cDNA clone. However, molecular heterogeneity is apparent which arises by differential glycosylation, as on eosinophils FcaRI appears to have a molecular mass of 70-100 kDa (Monteiro et al., 1992), in contrast to FcaR on other cell types which exhibit a molecular mass of 55-75 kDa (Albrechtsen et al., 1988; Monteiro et al., 1990, 1992; Mazengera and Kerr, 1990). Deglycosylation of FcaRI on these cells in all cases produces a core protein of 32 kDa, which, combined with the apparent existence of only a single FcaRI gene (Maliszewski et al., 1990), suggests that multiple receptor isoforms with different numbers of glycosylation sites are not an explanation for the observed heterogeneity.

D. OTHERFCR 1. FcpR In contrast to the FcyR, FcsR, and FcaR, receptors for the Fc portion of IgM are not well-characterized; however, recent studies have begun to define the biochemical nature of these receptors.

42

MARK D. HULETT AND P. MARKHOGARTH

The existence of IgM binding molecules has been well-documented on subpopulations of human and murine B and T cells using EA rosetting and immunoflourescence techniques (Moretta et al., 1975, 1977; Lamon et al., 1976; Pichler and Knapp, 1977; Ferrarini et al., 1977; Burns et al., 1979; Rudders et al., 1980; Mathur et al., 1988a,b; Lydyard and Fanger, 1982; Anderson et al., 1981) (Table XI). Human NK cells have also been reported to express FcpR (Pricop et al., 1991, 1993).Although these approaches have clearly demonstrated IgM binding function of these cell types, it is only recently that FcpR have been biochemically defined. An FcpR of 58 kDa has been isolated from human B-cell lineages (Ohno et al., 1990)and a similar but apparently distinct molecule of 60 kDa has been isolated from human T cells following their short-term culture (Nakamura et al., 1993). The B-cell FcpR is an 0-glycosylated protein linked to the membrane by a GPI anchor and is inducible following cell activation. In contrast, the Tcell FcpR is resistant to phospholipase C treatment, suggesting it may be an integral membrane protein, and is downregulated following cell activation. The question of whether these FcpR are different forms of the one receptor or entirely different FcpR remains to be determined. FcpR on mouse B and T cells have not yet been isolated. The affinity ofthe described FcpRs appears to be quite low, their detection necessitating the use of IgM complexes (Pichler and Knapp, 1977; Mathur et al., 1988a; Lydyard et al., 1982).The function of FcpR on lymphocytes is interesting, and it is tempting to speculate that these receptors play a similar role in immune regulation as the FcyR and FcsR receptors on lymphocytes. Clearly much further work is needed to understand the biology of FcpB.

TABLE XI CHARACTERISTICS OF FcpR Characteristic

Humana

Mr Cell distribution Receptor topology Specificity Regulation expression*

58 B cell GPl IgM

a

t

No FcpR isolated from murine cells.

* Following cell activation. ND, not determined.

60

T cell ? IgM

t

ND NK cell ND IgM ND'

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

43

2. FcsR Receptors for IgD have been described on murine and human T cells and T-cell clones. The expression of these IgD-R is induced upon exposure to IgD complexes, 11-2,Il-4,or T-cell activating agents (Coico et al., 1985, 1987, 1988, 1990). In mice, IgD-R are expressed only on CD4+ T cells and cloned CD4+ T cells (Coico et al., 1985, 1987),whereas in humans they are expressed on both CD4+ and CD8+ T-cell subsets and T-cell lines (Coico et al., 1990; Tamma and Coico, 1991). However, IgD-R on murine and human T cells have not been biochemically characterized and have only been functionally detected on the surface of T cells by rosetting with IgD-coated erythrocytes. The presence of IgD receptors on human B cells have also reported (Sjoberg 1980b; Rudders and Anderson, 1982). Murine T-cell IgD-R have been shown to recognize N-glycans of murine IgD, as the interaction of these receptors with IgD is inhibited by N-acetlyglucosamine, N-acetylgalactosamine, and galactose. As such these receptors have been suggested to be lectin-like molecules (Amin et al., 1991). In contrast to receptors for the other Ig isotypes, the binding of mIgD to murine IgD-R is not specific to the Fc portion of IgD, as both Fab and Fc fragments can block the binding of IgD. The interactive region of murine IgD has been localized to the first and third constant regions of the heavy-chain domains (Tamma et al., 1991). Recent studies have suggested that the IgD-R on human T cells is also a lectin that binds N-glycans. However, in contrast to the murine IgD-R, the hIgD-R appears to interact with both hIgD and hIgAl (G. Thorbecke, personal communication). Both of these Ig isotypes contain Gal 1-3Gal NAc-rich 0-linked glycans. It should also be noted that murine and human IgD do not exhibit cross-species inhibition as assessed by EA rosetting. Clearly, extensive further studies are required to understand the molecular nature of murine and human IgD-R. 3. Polymeric IgR The polymeric IgA/IgM receptor (poly-IgR) is expressed on the basolateral surface of glandular epithelial cells and is responsible for the transcytosis of these polymeric Igs into external secretions (Table XII). Proteolytic cleavage of the extracellular polymeric Ig binding portion of the receptor produces secretory component (SC).The receptor binds polymeric Ig basolaterally and endocytosis of the receptor-ligand complex is followed by transcytosis to the apical cell surface and proteolytic cleavage of the receptor to release polymeric Ig into the apical medium in association with SC (Mostovet al., 1984; Brandtzaeg, 1985).

44

MARK D. HULETT AND P. MARKHOCARTH

TABLE XI1 CHAFIACTERISTICS OF POLY-IGR Characteristic

Rabbit

Isoforms Molecular mass (LDA) Ig-like domains Receptor topology Specificity

A0 70 3 T M ~sc , PkA, PIgM

Cellular distribution

Glandular epithelial cells NDd

Chromosome localization

Rat B‘ 90-95

5

TM, SC PI&, PI@ Glandular epithelial cells ND

Human

A 120 5 TM, SC PkAC

Glandular epithelial cells ND

Glandular epithelial cells lq31-41

Two forms encoded by differentiallyspliced transcripts,the “A, B” nomenclature is not standard and is used here for the sake of comparison. TM (transmembrane) and SC (secretory component) arise by cleavage of receptor from cell surface in association with polymeric Ig following transcytosis. Polymeric IgM binds weakly. ND, not determined.

*

The poly-IgR has been characterized at a biochemical and molecular level in three species, the rabbit, rat, and human. The receptor has been defined as a membrane glycoprotein of 100 kDa in the human (Brandtzaeg, 1985), 120 kDa in the rat (Banting et al., 1989), and two isoforms of 70 and 90-95 kDa in the rabbit (Mostov et al., 1984) (which are the products of alternatively spliced mRNAs of a single gene, see below). Molecular cloning of cDNAs encoding the poly-IgR in the rabbit (Mostov et al., 1984), rat (Banting et al., 1989), and human (Eiffert et al., 1984, 1989; Krajci et al., 1989, 1991, 1992) indicate the receptor is a member of the Ig superfamily and is structurally conserved in all three species, The receptor is an integral membrane molecule comprising an extracellular region of 5 Ig-like domains, a single transmembrane region, and a cytoplasmic tail. The 5 Ig-like domains are highly conserved and exhibit a significant degree ofhomology with the variable domain of Ig and thus are only distantly related to the leukocyte Ig superfamily FcR which contain Ig-like domains of the C2 set (Williams et al., 1989). Comparison of the predicted amino acid sequences of the rabbit, rat, and human poly-IgR reveals an overall 41% amino acid identity. The extracellular regions exhibit only 36% identity; however, the transmembrane and cytoplasmic tail regions are highly homologous displaying 74 and 60% identity across the three species, respectively (Mostov et al., 1984; Banting et al., 1989; Krajci et al., 1989).The significant conservation of these regions

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

4s

suggests they are important for receptor function. Indeed, the high degree of sequence homology of the cytoplasmic tails presumably reflects conservation of intracellular signals required for correct targeting/sorting/transcytosis (see below). A single mRNA species has been observed of 2.8 kb in the human (Krajci et al., 1989)and 3.5 kb in the rat (Banting et al., 1989),whereas two distinct related mRNAs of 2.6 and 2.8 kb are present in the rabbit which arise by differential splicing (Mostov et al., 1984). The gene encoding the human poly-IgR has been isolated and comprises 11 exons spanning 19 kb (Krajci et al., 1992b). Two exons encode the signal peptide (exons 2 and 3), single exons encode three of the five Ig-like domains (domains 1, 3, and 4 encoded by exons 3,5, and 6, respectively), whereas domain 2 and 3 are encoded by the same exon (exon 4). Exons 8 to 11encode the cytoplasmic tail region, with exon 8 also encoding the transmembrane region. As described above, the rabbit poly-IgR exists in both high- and low-molecular-weight forms which are encoded by two distinct differentially spliced mRNA transcripts (Mostov et al., 1984). The sequence of these transcripts indicates that it is the region encoding domains 2 and 3 which is alternatively spliced, which corresponds precisely to exon 4 of the human gene (Krajci et al., 1992b). The 4 exons encoding the cytoplasmic domains seem to correlate with the regions defining the structural determinants proposed to be responsible for the intracellular sorting of the poly-IgR in the rabbit (Apodaca et al., 1991). These regions include a 14-amino acid segment (residues 655-668 in the rabbit) proposed to direct the receptor to the basolateral surface (Casanova et al., 1991); the corresponding region in the human is exons 8 and 9, exon 9 containing Ser655-the phosphorylation of which appears crucial for receptor transcytmis (Hirt et d,1993). Residues 670-707 of the rabbit poly-IgR encode a region believed to be involved in the protection of receptor from lysosomal degradation (Breitfeld et al., 1990), and the corresponding human region is also found in exon 9. The 30 C-terminal residues of the rabbit poly-IgR have been shown to be responsible for the rapid basolateral endocytosis of the receptor (Breitfeld et al., 1990), and the corresponding human region is found in exon 11. The human poly-IgR gene has been mapped to chromosome lq31-41 by direct and genetic approaches (Davidson et al., 1988; Krajci et al., 1991, 1992a). A recent study has demonstrated that the human poly-IgR mRNA is upregulated in a time- and concentration-dependent manner by IFN-y (Krajci et al., 1993). Other proinflammatory cytokines which increase the epithelial expression of the human poly-IgR include TNF-

46

MARK D. HULETT AND P. MARKHOGARTH

a and IL-4 (Sollid et al., 1987; Kvale et al., 1988; Phillips et al., 1990).

In contrast to the human and rabbit poly-IgRs, which both bind polyIgA and poly-IgM, the rat receptor appears to bind only poly-IgA well and not poly-IgM (Underdown et al., 1992). 4. FcRn

FcRn is a receptor for IgG on intestinal epithelial cells which mediates the transfer of maternal Ig from milk to the bloodstream of newborn mice and rats (Table XIII). The receptor has been defined at a molecular level in both the rat and the mouse and is a heterodimer of an integral membrane glycoprotein similar to MHC class I antigens (IgG binding a-subunit) and @2-microglobulin(Simister and Mostov, 1989). The FcRn a-chain has been described as a 45- to 53- or 50-kDa glycoprotein in the rat (Simister and Mostov, 1989)and mouse (Ahouse et al., 1993),respectively. The p2m component has an apparent molecular mass of 14 kDa in both species. The association of the FcRn achain with p2m has been directly demonstrated in the rat using crosslinking studies of the receptor on brush border epithelial cells (Simister and Mostov, 1989) and is also suggested in the mouse as neonate1 mice homozygous for a targeted disruption of the p2m gene lack the ability to bind IgG through FcRn (Ziljstra et al., 1990). A human form of FcRn has also recently been suggested, as microvilli membranes from the small bowel of fetal intestine exhibit pH-dependent binding of IgG with a dissociation constant of 2 x lo7 M - l , which is similar to that of the rodent FcRn (see below) (Israel et al., 1993). cDNA cloning of the rat (Simister and Mostov, 1989) and mouse TABLE XI11 CHARACTERISTICS OF FcRn Characteristic

Rat

Mouse ~~

Molecular mass (kDa) Ig-like domains Receptor topology Associated subunits Cellular distribution Specificityb Affinity Chromosomal localization

45-53 3

50 3

TM /32ma Intestinal epithelial cells, fetal yolk sac IgG lo8M-'

TM P2m" Intestinal epithelial cells, fetal yolk sac

ND

p2m mouse/rat, 14 kDa. Selectively binds I& at pH 6.4; releases bound Ig at pH 7.4.

I& lo8 M-' ch 7

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

47

(Ahouse et al., 1993) FcRn a-chains indicates they are highly conserved integral membrane proteins with a predicted structure comprising an extracellular region of 3 Ig-like domains homologous to class I MHC antigens (therefore only distantly related to the leukocyte FcR and the poly-Ig receptor), a single transmembrane region, and a cytoplasmic tail. The rat and mouse FcRn exhibit 91% overall amino acid identity, with 84, 88, and 100% identity in the al, a2, and a3 domains, respectively, 91% in the transmembrane regions, and 98% in the cytoplasmic tail regions. Additional support for the similar structures of rat FcRn (and by analogy mouse FcRn) with MHC class I is based on the similarity in the circular dichroism spectra of rat FcRn and HLA-B40 (Gastinel et al., 1992). Comparison of the amino acid sequence of rat FcRn with MHC class I antigens reveals significant homology, the highest homology being in the a3 domains (35-37%), followed by a1 (27-30%), and a2 (22-29%) (Simister and Mostov, 1989). Mouse FcRn also exhibits similar homology with MHC class I, with an average of 27% identity in a l , 23% in a2, and 34% in a3 (Ahouse et al., 1993). The gene encoding mouse FcRn has been mapped to the proximal region of chromosome 7 and is therefore not encoded in the same region as the majority of MHC class I antigens which are found on chromosome 17 (Ahouse et al., 1993).This finding, together with the results of amino acid sequence comparisons, supports a divergence of FcRn from MHC class I early in the mammalian lineage. The binding of IgG by rat and mouse FcRn is of high affinity ( K , = lo7 - lo8 M-I) and is pH dependent with Ig being bound at pH 6.5 in the acidic environment of the gut and released at pH 7.4 in the neutral environment of the bloodstream (Simister and Rees, 1985; Hobbs and Jackson, 1985; Raghavan et al., 1993). A secreted form of rat FcRn has been cocrystallized with its ligand, and although no structural data have been reported, the stoichiometry of the interaction has been determined, with two FcRn molecules binding per Fc portion (Gastinel et al., 1992; Huber et al., 1993). Studies of the tissue distribution of FcRn by Northern blot analysis of mRNA has demonstrated that the receptor is expressed in epithelial cells of the neonate1 rat and mouse small intestine (but not adult intestine) (Simister and Mostov, 1989; Ahouse et al., 1993) and in the yolk sac (Ahouse et al., 1993; Roberts et al., 1990). A single 2.2-kb FcRn a-chain mRNA is present in these tissues of the mouse (Ahouse et al., 1993) and in two mRNAs of 1.7 and 3.1 kb in the rat, which possibly arise by the use of alternate polyadenylation sites (Simister and Mostov, 1989).

48

MARK D. HULETT AND P. MARKHOGARTH

111. Molecular Basis of the FcR-19 Interactions

The main focus of studies on the molecular nature of the FcR-Ig interaction has been the identification of the regions in the Fc portion of Ig involved in binding to FcRs, and little attention has been given to the determination of the sites on FcRs responsible for binding Ig (aspects reviewed in Metzger, 1988; Burton and Woof, 1992; Sutton and Gould, 1993).As a consequence, a distinct bias exists in the understanding of the FcR-Ig interaction. However, recent studies performed by ourselves and others examine this interaction from the receptor side, and significant advances are being made into understanding the molecular basis of the interaction of FcRs with Ig. The current. state of understanding of the FcyR-IgG and FceRI-IgE interactions is presented below and summarized in Table XIV and Fig. 1. A. FcyRI The site(s) of interaction on mouse or human FcyRI with IgG are not well-characterized at present. However, by generating chimeric mFcyRIlmFcyRI1 receptors we have been able to define the Ig binding roles of the extracellular domains of mFcyRI (Hulett et al., 1991). The extracellular region of mFcyRI can be divided into two main regions with distinct Ig bindingroles: (i)the first two domains (homologous to the two domains of the other leukocyte Ig superfamily FcR) responsible for the direct binding of IgG and (ii) the unique domain 3, which confers the distinctive specificity and affinity of the receptor (Table XIV). The first two domains of mFcyRI can bind IgG in their own right. However, the removal of domain 3 converts the Ig binding function of mFcyRI to an “FcyRII-like” receptor, domains 1and 2 of mFcyRI in the absence of domain 3 lose the ability to specifically bind mIgG2a with high affinity and instead exhibit a low affinity and broad specificity for mIgGl, 2a, and 2b, characteristic of mFcyRII and I11 (Hulett et al., 1991). This finding demonstrates that the first two domains of mFcyRI represent an IgG binding motif conserved with the lowaffinity FcyR and that domain 3 of mFcyRI is modifying the binding of IgG by domains 1 and 2. Consistent with this finding is that the two-domain fork of hFcyRIb, which lacks domain 3, also exhibits lowaffinity IgG binding (Porges et al., 1992). In addition, it has been claimed that preliminary studies on hFcyRI, whereby point mutations were introduced into domain 3, also indicate that this domain is important in conferring high-affinity binding, although no data were presented (Allen and Seed, 1989).Based on these observations, it would

TABLE XIV

SUMMARYOF FcR-Ig INTERACTIVEREGION^ Receptor-ligand interactionb D1

FcyRI - IgG

Stability ?

I

D2

I

1

F9RII - IgG

Ig binding regions

Receptor binding regions

,,

AffinitY/S#city

D2

.I

’ ,111-114

\

/

0,

Binding domain

D1

D3

I

Stability/AfEnity ?

1!2%134 1 5 p 1 6 1 5

Binding domain

F3RIII - IgG \

FcERI- IgE

I

D1

Stahility/Minity ?

?

CE2

D2 87-12.W

129-137

Binding domain*

154-161

/

/ \

Binding domain ?

,

Stability ?

I=

CE3

,, =

Binding domain

I

CE4

1 0

/ \

Minity&nd site

Schematic diagram of FcR extracellular domains and Ig Fc portion constant domains. Identified binding regions shaded and key residues indicated. Based on studies of mFcyRI, hFcyRII, rat FcyRIII and hFc,RI. Demonstrated directly for hlgGln and inferred for other IgG subclasses. Site located in hinge proximal region. Direct binding role demonstrated for 15p161 region, secondary or indirect binding contribution by 111-114 and 130-135 regions. Direct binding role for all three regions, three subregions of 87-128, ie., &104,10&115, and 111-125 imphcted in binding (see text). f For binding of rat IgE by rat Fc,RI, domain 1appears to be the crucial domain.

*

A

*

*

*

6

*

*

*

C

*

F G L T A N S - D T H L L O G Q S L T L T L E S - P P G S S P S V Q C R S P R G - - - K 86

110

100

90

120

E W L V L Q T P H L E F Q E G E T I M L R C H S W K D - V K V T F F Q N - G - K S Q

D W L L L Q T P Q L V F L E G E T I T L R C H S W R N K L & N R I S F F H N - E - K S V G W L L L Q A P R W V F ' E E D P I H L R C H S W K N T A L H K V T Y L Q N - 2 - K D R G W L L L Q A P R W V T K P P 3 P I H L R C H S W K N T A L H K V T Y L Q N - D - K D R D W L L L Q T P Q R V F ; E G S T I T L R C H S W R N K L L N R I S F F H N - E - K S V

-

--

- -- ------

-

D W L L L Q T P Q L V F L E ' G S R I T L . R C H G W K S I Q L A R 1 S F L Q N - G - Q F V

D W L L L Q T P Q L V F E E G E T I T L R C H S W K N K Q L T K V L L F Q N - G - K P V D W L L L Q A S R R V L T E G E P L A L R C H G W K N K L V T N V V F Y R N - G - K S F G W L L L Q V S S R V F T E G E P L A L R C H A W K D K L V Y N V L Y Y R N - G - K A F

D W L L L Q A S A E V V M E G O P L F L R C H G W R N W D V Y K V I Y Y K D - G E A L K

C *

*

E

G

F

*

*

*

*

*

*

*

N I Q G - G K T L S V S Q L E L Q D S G T W T C T V L Q - N Q K K V Q F K I D I V V L 130

K R K K R

@ Y Y Y Y

S H F F H

140

150

160

@ L D @ T F S I P Q A N H S H S G D Y H C T G [ N ~ Y T ~ S K P V T I T V Q H Y S S N F S I P K A N H S H S G D Y Y C K G S L G R T L H Q S K P V T I T V H H N S D F ~ I P K A T L K D S G S Y F C R G L ~ G S ~ N V S S E T V N H H N S D F H I P K A T L K D S G S Y F C R G L V G S K N V S S E T V N I T I H Y K S N F S I P K A N H S H S G D Y Y C K G S L G S T Q H Q S K P V T I T V

--------- --

-

170

Q I T Q

G T Q D

P I T Q G L A V S T G L A V S T P A T

S F H P Y N V S Y S I S N A N H S H S G D Y Y C K A Y L G R T E H V S K P V T I T V Q G

R Y Y Y Q S S N F S I P K A N H S H S G N Y Y C K A Y L G R T M H V S K P V T I T V Q G

Q F S - S D S E V A I L K T N L S H S G I Y H C S G T - G R H R Y T S A G V S I T V K E L K F F H W N S N L T I L K T N I S H N G T Y H C S G M - G K H R Y T S A G I S V T V K E L Y W Y E N - H N I S I T N A T V E D S G T Y Y C T G I K V W Q L D Y E S E P L N I T V I

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

51

be expected that the hFcyRIbl and c l transcripts, which contain stop codons in their third extracellular domain coding regions, would encode functional soluble low-affinity FcyRs. However, protein products of these transcripts have not as yet been identified. It should be noted that the modifying effect on Ig binding by domain 3 was found to be specific to mFcyRI, as the linking of domain 3 to domains 1 and 2 of mFcyRII did not produce the specific high-affinity binding of mIgGSa, this receptor retaining the specificity and affinity of mFcyRII (Hulett et al., 1991). Furthermore, the binding of mIgG2a by mFcyR1 appears to be a specialized interaction between domains 2 and 3, as replacement of domain 1 of mFcyR1 with domain 1 of mFcyRII does not alter the specificity of IgG binding or have a major influence on the high-affinity binding of mIgG2a, in contrast to the replacement of both domains 1 and 2 of mFcyRI with domains 1 and 2 of mFcyRII (M. Hulett, unpublished observations). Based on these findings it is tempting to speculate that domain 2 of mFcyRI is the key domain involved in the direct binding of IgG, especially in the light of the observation that it is the homologous second extracellular domain of both human FcyRII and FceRI a-chain that is responsible for the binding of IgG and IgE, respectively (see below). The definition of the Ig binding roles of the extracellular domains of mFcyRI should now enable the fine specificity of the molecular interaction with IgG to be determined. In contrast to the limited information available on the FcyRI binding site for IgG, significant advances have been made into identifying the binding site(s) on IgG for hFcyRI (Burton et al., 1988; Burton and Woof, 1992). Early studies using proteolytic fragments of IgG suggested the C y 3 domain of IgG was crucial for the interaction with human FcyRI (Okafor et al., 1974; Ciccimarra et al., 1975); however, this was subsequently shown to be incorrect with purified IgG fragFIG.1. Alignment of Ig superfamily FcR second extracellular domain amino acid sequences. The positions of the putative p strands are overlined and the core hydrophobic residues are indicated by asterisks and are based on comparison with the solved structure of CD4 domain 2 (Ryu et al., 1990; Wang et al., 1990; Hogarth et al., 1992). Regions implicated in the binding of Ig using chimeric receptor studies are boxed. Specific residues implicated in Ig binding through mutagenesis studies are circled. Polymorphic residues also suggested to play a binding role are underlined. Amino acid differences between rat FcyRIIIA and rat FcyRIIIH or hFcyRIIIa and FcyRIIIb are indicated by lines between the two sequences. Three subregions of the 87-127 IgE binding region of hFceRI have been implicated in binding and are underlined. See text for sequence details. The numbering is based on that for hFcyRIIa, with every 10th residue indicated.

52

MARK D. HULETT AND P. MARKHOGARTH

ments and myeloma proteins containing deleted domains (Woof et al., 1984). It was suggested that the Cy2 domain had an important role in the binding of IgG to FcyRI in experiments where removal of N-linked carbohydrate from Cy2 resuIted in a significant loss in affinity for FcyRI (Leatherbarrow et aZ., 1985, Walker et al., 1989).The role of Cy2 in the binding of IgG to hFcyR1 was demonstrated directly in experiments that examined the capacity of anti-human IgG mAb to inhibit the binding of IgG to FcyRI, as only mAb that recognized the N-terminal portion of Cy2 blocked IgG binding to FcyRI (Partridge et al., 1988). These same mAb could not bind to IgG that was bound to FcyRI. In support of this important binding role of Cy2, recent experiments using chimeric immunoglobulins generated between hIgG1 and mIgE, where Cy2 andlor Cy3 were exchanged with the homologous Cs3 and Cs4, demonstrated that mutant immunoglobulins lacking Cy2 did not bind to FcyRI, whereas those containing Cy2 bound to FcyRI (Shopes et al., 1990). However, Cy3 does seem to play a role in the binding of IgG to FcyRI, as calculations of the relative contributions of each domain to binding reveal that Cy3 contributes 25% of the overall drop in free energy on binding, Cy2 contributing 73%. This contribution of Cy3 to the binding of IgG by FcyRI has been proposed as a stabilizing role on the Fc structure (Shopes et al., 1990). A similar study using chimeric hIgG2 and hIgG3 molecules supports the above findings, as IgG2 (which does not bind to FcyRI) substituted with Cy3 of IgG3 (which binds to FcyRI) did not bind to hFcyRI, whereas IgG3 containing Cy3 from IgG2 did bind to FcyRI (Canfield and Morrison, 1991). The Cy2 domain of hIgGl has also been shown as the principal domain involved in the binding of this isotype to hFcyRI. Using a similar chimeric approach as that described above, but with hIgGl and hIgG2, it was demonstrated that only those chimeric molecules containing Cy2 of IgGl were able to bind to hFcyRI (Chappel et al., 1991). In order to identify the binding site in Cy2 for hFcyRI, experiments were performed where a range of IgGs of different isotypes from different species was tested for their ability to bind to hFcyRI (Woof et al., 1986). Examination of the amino acid sequence of the Cy2 domain of these IgGs enabled the identification to be made of sites potentially involved in the binding to FcyRI. The region Leuw to S e P 9 (Leu-Leu-Gly-Gly-Pro-Ser) in the N-terminal region of Cy2, which forms part of the hinge proximal region, was proposed as crucial for interaction with FcyRI, being present in all IgG isotypes that bound to hFcyRI with high affinity, i.e., hIgG1, hIgG3, mIgG2a, rat IgC2b, and rabbit IgG (Woof et al., 1986) (Table XV). Mouse IgG2b and

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

53

TABLE XV

COMPARISON IgG CH2HINGEPROXIMAL REGIONS IgG" hIgGl hIgG3 rIgGb mIgG2a mIgG2b hIgG2 mIgGl a

Amino acid sequence alignmen9

P P P P P P T

A A P A A A V

P P P P P P P

E L L G G P S V

E E N N P

€3-

L L L L A

L G G P L G G P L G G P E G G P - A G P -- V S

S S S S S S

V V V V V V

h, human; m, mouse; r, rabbit. Corresponding to residues 230 to 240 of hIgG1.

hIgG4, which bind weakly to FcyRI, differ in this region, with mIgG2b containing Glu at position 235 and hIgG4 containing Phe at position 234. The importance of the 234-239 region in the binding of IgG to FcyR has been confirmed using site-directed mutagenesis. Mutagenesis of this region in hIgG3 has demonstrated that substitutions between residues 234 and 237 reduce the binding to hFcyRI. Residue when substituted for Glu had a >100-fold decreased binding affinity, whereas replacement of Leu234,GlyU6,and GlyZ3' with Ala had less effect, with affinities reduced 4-, 4-, and 30-fold, respectively. Using the reverse approach, the weak binding of mIgG2b was converted to high affinity (comparable to hIgGl), following replacement of G ~ with Leu (Duncan et al., 1988; Lund et al., 1991). Similarly, point mutations in this region of hIgGl have been shown to either significantly reduce or abolish its hFcyRI binding activity (Chappel et al., 1991). An independent study examining the 234-237 region by sitedirected mutagenesis supported the above findings and confirmed the importance of this region in the binding of IgG to hFcyRI (Canfield and Morrison, 1991). Replacement of in hIgG3 with Glu also resulted in a >lOO-fold reduction in affinity for hFcyRI. In addition, hIgG4, which binds weakly to hFcyRI and differs from the high-affinity IgG isotypes containing a Phe at position 234, was converted to a highaffinity binding immunoglobulin (3-fold lower than hIgG3 for FcyRI) upon substitution of Phem with Leu. In the reciprocal experiment, replacement of Leu2%in IgG3 with Phe produced a molecule with a low affinity for hFcyRI equivalent to that of hIgG4 (Canfield and Morrison, 1991). The inability to impart full high-affinity binding to hIgG4 by replacing PheZa with Leu led to the proposal that other residues in C72 may be involved in the binding to FcyRI. Indeed, a

u

~

54

MARK D. HULETT AND P. MARKHOGARTH

second region of Cy2 comprising a hinge proximal bend which lies in close proximity to the 234 to 237 region has also been implicated in the binding of IgG by hFcyRI (Canfield and Morrison, 1991). This was demonstrated as substitution of Pro331situated in this loop region (Pro is found in this position in the high-affinity binding hIgGl and IgG3 isotypes in contrast to Ser in hIgG4) and was found to reduce the affinity for hFcyRI 10-fold. It has also been proposed that another bend region lying close to the hinge proximal region may be an important contributor to binding, as aglycosylation of Cy2, which lowers the affinity for hFcyRI, appears to result in structural alterations in this region as assessed by NMR methods, which may in turn effect the structure of the lower hinge region (Matsuda et al., 1990).It should be also noted that the binding of IgG to hFcyRI has been shown to involve only one heavy chain, as the valency of hFcyRI for IgG was determined to be one (O’Grady et al., 1986), and monomeric mIgG2a2b and mIgG2a bind equally well to mFcyRI (Koolwijk et al., 1989).

B. FcyRII Of the three classes of FcyR, FcyRII has the most reported information on the molecular basis of its interaction with IgG (Table XIV, Fig. 1). A contribution to the understanding of how FcyRII binds IgG has been made through the identification and characterization of a functional polymorphism of hFcyRII, the high-responder/low-responder polymorphism. This polymorphism has been described extensively above and identifies residue 131 in the second extracellular domain of hFcyRIIa as important in the binding of IgG. The nature of residue in this position is crucial for the binding of both mIgGl (Warmerdam et al., 1990; Tate et al., 1992) and hIgG2 (Warmerdam et al., 1991). The presence of Arg directs the strong binding of mIgGl, yet results in weak binding of hIgG2; whereas the presence of His promotes the strong binding of hIgG2 and weak binding of mIgG1. These findings indicate that position 131 of hFcyRIIa is probably contributing to the binding site of both mIgGl and hIgG2 and suggest that domain 2 has an important role in the binding of IgG. It should be noted that this polymorphism does not affect the binding of other mouse and human IgG isotypes (Warmerdam et al., 1991), suggesting the existence of additional regions important in the binding of IgG by hFcyRII (see below). The mouse Ly-17 polymorphism of mFcyRII (Shen and Boyse, 1980; Hibbs et al., 1985; Holmes et al., 1985) also implicates the second extracellular domain of this FcyR class in the binding of IgG. The

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

55

polymorphism, described above, has been defined at the molecular level as two allelic variants that differ only in residues 116 and 161, where Pro'16 and Glnl6I are found in the Ly-17.1 form and Leu''6 and Leu'" in the Ly-17.2 form (Lah et al., 1990). MAb specific for the Ly17.2 form inhibit the binding of IgG to the receptor, implying that residues 116 and/or 161are involved in binding themselves, or closely situated to residues crucial in the interaction of FcyRII with IgG. W e have used a chimeric receptor strategy, involving the exchange of homologous segments of hFcyRII and the structurally similar hFcsRI a-chain, to definitively demonstrate that domain 2 of hFcyRII is responsible for the direct binding of IgG (Hulett et al., 1993). A chimeric receptor comprising domain 1 of the hFcsRI a-chain and domain 2 of hFcyRIIa was found to bind only IgG immune complexes, whereas the reciprocal receptor containing domain 1of hFcyRIIa and domain 2 of the hFceRI a-chain did not exhibit any IgG binding and instead bound only IgE complexes. In addition, we have also demonstrated that domain 2 of hFcyRIIa contains the epitopes recognized by anti-hFcyRII mAb which block the binding of IgG to hFcyRII, providing further supporting evidence that domain 2 is the principle domain involved in the binding of IgG (Ierino et al., 1993). Although domain 1 of hFcyRII appears not to be directly involved in the binding of IgG, it does play an important structural role. This is suggested as replacement of hFcyRIIa domain 1 with domain 1 of hFcsRIa reduced the capacity to bind IgG, as shown by the failure of this receptor to bind dimeric human or mouse IgG1, which bind to wild-type hFcyRII (M. D. Hulett, unpublished observations). These data imply that the role of domain 1 in Ig binding is likely to be an influence on receptor conformation, stabilizing the structure of domain 2 to enable efficient IgG binding by hFcyRII. However, the possibility remains that a direct interaction of domain 1with IgG can occur following initial binding to domain 2. The systematic examination of hFcyRII domain 2, again using a chimeric hFcyRII/hFcsRI a-chain strategy, has enabled the localization of the IgG binding region to an 8-amino acid segment contained within residues Asn'% to Ser''l (Table XIV). Site-directed mutagenesis on this region identified residues Ilel5' and Gly'56 as crucial in the binding of both mIgGl and hIgGl by hFcyRII as replacement of these residues with alanine resulted in almost complete loss of binding. The importance of this region in the binding of IgG was further supported with the finding that replacement of Leu159,PhelGO,and SerlG1with Ala substantially increased the affinity of these mutant receptors for mIgGl and/or hIgGl (Hulett et al., 1994) (Fig. 1).

56

MARK D. HULETT AND P. MARKHOGARTH

We have generated a three-dimensional model of hFcyRII domain 2 based on the previously described related structure of CD4 domain 2 (Hogarth et al., 1992; Hulett et aE., 1994) (Fig. 2). The model represents an Ig-like domain of the “truncated C2 set,” comprising two anti-parallel 6 sheets of three and four6 strands, respectively (Williams and Barclay, 1988).The putative eight-residue binding region lies in the F-G loop of domain 2 at the interface with domain 1. The spacial location of the Ile”‘ and Gly’= in the hFcyRII domain 2 model suggests these residues contribute to a possible hydrophobic cleft between the F-G and B-C loops (Fig. 2). Based on these findings, this hydrophobic pocket is postulated to be a critical structure for the binding of IgG by hFcyRII. The similar Ig binding specificity of the FcyR, combined with their high amino acid sequence identity, makes it tempting to speculate that the F-G loop of domain 2 may be a conserved binding region in this class of FcR. With this in mind, it is significant to note that comparison of the F-G loop sequences of the human and mouse FcyR reveals that the putative crucial IgG interactive residues of hFcyRII, i.e., Ile15’ and G1y1%, are the only conserved residues, as Gly’= is found in all FcyR and a hydrophobic residue is present at position 155 in all the low-affinity FcyR (Fig. 1). In light of these observations, it is interesting to note that the F-G loop sequences of the two hFcyRIII isoforms differ only in the nature of the hydrophobic residue at position 155, where FcyRIIIA contains a phenylalanine and hFcyRIIIB a valine (Fig. 1). These hFcyRII1 isoforms exhibit distinct affinities for IgG as hFcyRIIIA has a K , = 2 x lo7M - l whereas hFcyRIIIB a K , < lo7M-‘. (Simmons and Seed, 1988; Vance et al., 1992). Thus, based on the proposed importance of the F-G loop in the binding of Ig, these findings suggest that the hydrophobic residue at position 155 may indeed be playing a crucial role in the binding of IgG by hFcyRIII. Other residues implicated in the binding of IgG by FcyRII through the polymorphism studies are situated in loop regions in close proximity to the identified 154-161 binding region. Residue 131 lies in the C’-E loop, and the human equivalents of the mouse Ly-17 mFcyRII polymorphism, i.e., and Led5’, are located in the adjacent B-C and F-G loops, respectively. These findings suggest that the C’-E and B-C loops of FcyRII also contribute to the binding of IgG. Indeed, site-directed mutagenesis on both of these regions has identified a number of residues which when replaced with Ala substantially effect the binding of IgG. These include Lys113,Pro114,and Leu”‘ of the BC loop and Phe12’, Arg/His13’, and Prola of the C’-E loop (M. D. Hulett, unpublished observations) (Fig. 1).However, based on our

FIG.2. Ribbon diagram of hFcyRII second extracellular domain model. The model is oriented with the C’-E and B-C face at the front and adjoins domain 1at the top of the page and the transmembrane region at the bottom (for details see Hogarth et al., 1992).The B-C, C’-E, and F-G loops are shown in dark blue, and the disulfide bond between Cys’“ and Cys’” in the B and F strands, respectively, in yellow. The F-G loop residues IleIss and Gly’” identified as crucial in the binding of IgG by hFcyRII (see text for details) are shown in magenta. t

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

57

findings from the chimeric receptor studies, the F-G loop of hFcyRII appears to be the major interactive region for mIgG1 and hIgG1. This is clearly evident as the substitution of the 154-161 region of hFcyRII with the corresponding region of hFceRIa totally eliminates IgG binding, whereas insertion of this region into FceRI imparts IgG binding to FceRI (Hulett et al., 1994). This effect on IgG binding was not observed for any other regions of FcyRII domain 2. Thus, although both the C’-E and B-C loops are involved in the binding of IgG, their roles appear to be secondary to that of the F-G loop region. In summary, these findings suggest that the IgG interactive site on hFcyRII is at the interface of domains 1 and 2, with loop regions at the membrane-distal end of domain 2 playing the crucial binding roles. It should be noted that the affinity of hFcyRII for IgG immune complexes has been demonstrated to increase following treatment with proteolytic enzymes such as pronase and elastase (van de Winkel et al., 1989); however, the molecular basis of this observation has yet to be determined. The identification of the FcyRII binding site(s) in IgG has proved more difficult for this receptor compared with FcyRI, due to the low affinity of FcyRII for monomeric IgG. However, significant advances have been made into localizing the region in the Fc portion of IgG important for interaction with hFcyRII (Burton and Woof, 1992). As for the binding of hIgGl and hIgG3 by hFcyRI, aglycosylation of the Cy2 domain of these isotypes resulted in a dramatic loss in their capacity to bind hFcyRII, suggesting Cy2 is important in the binding of IgG by hFcyRII (Walker et al., 1989). Studies have been performed to identify the key domains in the Fc portion of IgG responsible for binding to hFcyRII (Shopes et al., 1990). The experiments were performed using the same chimeric hIgGl/mIgE molecules described in the analogous experiments for hFcyRI (see above). Results indicated that only those chimeric Ig containing both Cy2 and Cy3 were able to mediate rosette formation with K562 cells (FcyRI-, FcyRII+), suggesting both these domains are required for the binding of hIgGl to hFcyRII (Shopes et al., 1990). These findings are therefore in contrast to the requirement of hFcyRI seemingly for only the Cy2 domain for the binding of hIgGl (Weetall et al., 1990). The importance of the Cy 2 domain of hIgG3 in the interaction with h F q R I I has been demonstrated using the panel of hIgG3 mutants in the 234-237 region, as described above for binding to hFcyRI (Lund et al., 1991). These mutant hIgG3 molecules were assessed for their ability to form rosettes with the FcyRII expressing B-cell line Daudi. The number of rosettes formed with hIgG3 substituted with Leu2%to Ala, to Ala,

to Glu, Gly236to Ala, and GlyZ3’to Ala was reduced in each case compared with wild-type hIgG3, suggesting this region is important in the binding of hIgG3 to hFcyRII. Human IgG3 containing replaced with Ala exhibited the lowest binding capacity to hFcyRII, forming
MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

59

served (Hulett et al., 1994). The binding of both hIgGl and mIgGl was shown to be dependent on the F-G loop and specifically IlelS5 and of hFcyRII. The differences in the binding of hIgGl and mIgGl were evident as the substitution of Phe'" in hFcyRII with alanine increased the binding ofhIgGl but not mIgGl, and the substitution of Ser'" with alanine increased the binding of mIgGl but not hIgG1. An additional difference in the binding of mIgGl and hIgGl by hFcyRII has also previously been demonstrated with the finding that residue 131 in the putative C'-E loop of FcyRII influences the binding of mIgGl but not hIgGl (Warmerdam et al., 1991). Residue 131 of hFcyRII appears to comprise a secondary binding site to the F-G loop region for mIgGl (Hulett et al., 1993, 1994). The only other IgG isotype influenced by residue 131 in the binding to hFcyRII is hIgG2, and, like mIgG1, contains a disrupted Leu-Leu-Gly-Gly motif. Based on these observations, it is tempting to speculate that there is a pattern of binding of the different IgG isotypes by hFcyRII, such that the F-G loop is crucial for the binding of all IgG isotypes; however, those with a disrupted Leu-Leu-Gly-Gly motif also interact with residue 131 (C'-E loop). The interaction with the 131 region may therefore be compensating for the loss of the Leu-Leu-Gly-Gly motif in the binding of these IgG isotypes to hFcyRII. As described above, an interesting finding was that the IgG binding specificity of mFcyRI (mIgG2a only) could be converted to that of mFcyRII (mIgG1, mIgG2a, and mIgG2b), upon removal of the third extracellular domain (Hulett et al., 1991). Based on the finding that the F-G loop in domain 2 of hFcyRII is crucial for the binding of IgG (Hulett et al., 1994) and that the first two domains of mFcyRI exhibit the identical IgG binding specificity to mFcyRII, it might be expected that this region is conserved in both mFcyRI and mFcyRII. Comparison of the F-G loop sequences of these two receptors reveals that only the and ArglS7are totally conserved (Fig. 1).The putative crucial hydrophobic residue at position 155 is deleted in mFcyRI (and hFcyRI), implying that the FcyRI-IgG interaction is different from FcyRII-IgG. In light of this observation, the difference in the F-G loop may be consistent with differences seen in the importance of Leu234for FcyRII binding but for FcyRI binding. Fine structure analysis of this region of FcyRI will hopefully resolve this question. C. FcyRIII The interaction of FcyRIII and IgG at the molecular level is currently not well-characterized. No information as yet is available on the receptor binding site(s)for IgG. There are some reports that suggest

60

MARK D. HULETT AND P. MARKHOGARTH

the two human receptor forms, i.e., hFcyRIIIb, the GPI-anchored form on neutrophils, and hFcyRIIIa, the transmembrane-anchored form on NK cells and macrophages, have different affinities for human IgG, implying that they may interact differently with IgG. Scatchard analysis of the direct binding of monomeric IgG indicates hFcyRIIIa has an affinity of K , = 2 x lo7 M-' (Vance et al., 1992) and hFcyRIIIb a significantly lower affinity of K , < lo' M-' (Fleit et al., 1982; Kulczycki, 1984; Anderson and Looney, 1986; Simmons and Seed, 1988). Support for this differential affinity is demonstrated in the finding that the anti-FcyRIII mAb 3G8 can be displaced from FcyRIIIa with monomeric IgG, but not from FcyRIIIb (Anderson et al., 1990b). A polymorphism of the GPI-anchored FcyRIIIb has also been identified, generating two allelic forms, designated NA1 and NA2; which exhibit four amino acid differences resulting in the loss of two N-linked glycosylation sites in the NA1 form. However, despite these differences the two hFcyRIIIb forms appear to interact similarly with IgG (Ory et al., 1989; Ravetch and Perussia, 1989).In contrast, there is some suggestion that different glycosylated forms of the transmembrane-anchored FcyRIIIa may have different affinities for IgG. This has been shown as FcyRIIIa on NK cells, which has a high mannose oligosaccharide content and has a higher affinity for IgG than does the low mannose containing FcyRIIIa on monocytes (Kimberly et al., 1991).This finding is consistent with observations that oligosaccharides may contribute to the integrity of the IgG binding site on the hFcyRII1, suggested through studies which demonstrate that aggregated IgG is able to block the unique ability of hFcyRIIIb to bind lectins through its carbohydrate moities (Salmon et al., 1987; Kimberly et al., 1989). The high amino acid identity of FcyRIII and FcyRII, combined with their similar IgG binding specificities and a n i t i e s , suggests that FcyRIII will interact with IgG in a fashion similar to FcyRII. Indeed, a recent study has provided evidence to suggest that like FcyRII, the second extracellular domain of FcyRIII is the principle Ig interactive domain. Two rat FcyRIII isoforms, IIIA and IIIH with extensive amino acid differences in their second extracellular domains (Fig. l),have been shown to bind rat and mouse IgG subclasses differently (Farber et al., 1993).Both isoforms bind rtIgG1, rtIgG2b, and mIgG1; however, only the rtFcyRII1-A isoform binds rtIgG2b and mIgG2b. Significantly, the amino acid differences between the rat FcyRIIIA and H isoforms are situated predominantly in the predicted B-C and C'-E loops of domain 2 (Fig. 1);and both of these regions have been shown to play roles in the binding of IgG by hFcyRI1. However, it is interesting to

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

61

note that the region of rtFcyRIII corresponding to the proposed major binding site of hFcyRII, i.e., the F-G loop, is almost totally conserved between the two rat isoforms. The conservation of this region is consistent with the idea that the F-G loop of FcyRIII is also the crucial binding site of this receptor class, as the binding of IgG isotypes by rtFcyRIIIA and IIIH are also relatively conserved despite the extensive amino acid differences between the two isoforms. Therefore, these observations suggest that like FcyRII, the F-G loop of FcyRIII may well be the major IgG binding site, with the B-C and C'-E loops playing supporting binding roles. Information of the FcyRIII binding site on IgG is also limited. Some early experiments indicated that the entire Fc portion was required for the binding to hFcyRIII on neutrophils, as a pFc' fragment or a tryptic Cy2 domain could not inhibit the interaction of IgG with the receptor (Barnett-Foster et al., 1978). However, it was suggested that the interface between the Cy2 and Cy3 domains was crucial for binding to FcyRIII, as protein A which binds to this region blocked the binding of IgG to FcyRIII (Barnett-Foster et al., 1982). This finding conflicts with the results of another study, which demonstrated that an anti-IgG mAb known to bind to an epitope in the Cy2/Cy3 interface was unable to block IgG binding to FcyRIII (Sarmay et al., 1985). This study using a panel of anti-IgG mAb also provided evidence to suggest that FcyRIII binds to a region in Cy3, with a second region in Cy2 being important for the triggering of antibody-dependent cellular cytotoxicity (ADCC) by the receptor. However, monocyte-depleted peripheral blood mononuclear cells were used in these experiments and as such these results could be misleading as cross-reactivity with FcyRII cannot be excluded. The glycosylation of the Fc portion of IgG may also be important for binding to FcyRIII, as it has been demonstrated that aglycosylation of hIgG3 results in the inability of NK cells to mediate ADCC via FcyRIII (Lund et al., 1990). D. FceRI There has been a great deal of interest in understanding the molecular basis of the interaction of IgE with FceRI, primarily due to the key role the receptor plays in triggering IgE-mediated allergic reactions. Most attention has focused on the binding sites on IgE for FceRI; however, recent advances have been made in the identification of the receptor binding sites for IgE. As described earlier, FceRI is a tetrameric complex comprising an a-subunit (the IgE binding chain), a p-subunit, and a dimer of two y-subunits. Chimeric a-subunits con-

62

MARK D. HULETT AND P. MARKHOGARTH

sisting of the extracellular regions of the a-chain linked to either the transmembrane and cytoplasmic tails of the p55 IL-2 receptor or the transmembrane and cytoplasmic tail of hFcyRII bind IgE with high affinity (Hakimi et al., 1990; Hulett et al., 1993).These results indicate that the 0- and y-subunits have no direct contribution to the binding of IgE by FceRI. These findings have been confirmed through the generation of a soluble human a-chain molecule, consisting of only the extracellular regions, which was shown to bind IgE with an affinity comparable to that of the wild-type receptor (Blank et al., 1990). We and others have identified the second extracellular domain of the hFceRI a-chain as the principle IgE interactive domain. An early study utilizing anti-hFceRI a-chain mAb demonstrated that mAb recognizing epitopes in domain 2 could inhibit the binding of IgE to hFceRI, whereas mAb recognizing epitopes in domain 1 did not, suggesting that IgE binding was mediated by domain 2 (Riske et al., 1991). However, the blocking of IgE binding by these domain 2 mAb could also result from steric hindrance from a site distant to the actual binding site, or from conformational changes induced in the binding site. We have employed a direct approach utilizing hFceRI a-chainlhFcyRIIa chimeras to definitively demonstrate that the second extracellular domain of the hFceRI a-chain is responsible for the direct binding of IgE and provide evidence to suggest that domain 1 plays a crucial role in maintaining conformational stability of the receptor to permit high-affinity IgE binding. A chimeric receptor comprising domain 1 of hFcyRIIa and domain 2 of the hFceRI a-chain bound IgE, whereas a chimera containing domain 1 of the hFceRI a-chain and domain 2 of hFcyRIJa exhibited no IgE binding (Hulett et al., 1993). Similar studies using interspecies chimeras of rat and human hFcERI a-chains or hFcyRIIIa and the hFcsRI a-chain also demonstrate the direct binding of IgE by domain 2 of the hFcsRI a-chain. The substitution of domain 2 of hFceRIa with domain 2 of hFcyRIII or rat hFcsRIa was found to result in the loss of hIgE binding, whereas when domain 1of hFceRIa was substituted with domain 1of hFcyRIII or rat hFceRI, hIgE binding was maintained (Mallimaci et al., 1993; Robertson, 1993). However, in contrast to our finding that substitution of domain 1 of hFcsRI a with domain 1 of hFcyRIII resulted in loss of highaffinity hIgE binding, the substitution of hFceRIa domain 1 with domain 1of hFcyRIII or rat FceRIa was found to have no effect on highaffinity binding. A possible explanation is that domain 1 of hFcyRIII and rat FcsRIa are better able to substitute for hFceRIa domain 1 than hFcyRII domain 1to maintain correct receptor conformation, as both hFcyRIII and rat FcsRIa domain 1 exhibit significantly higher

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

63

amino acid identity to hFceRIa domain 1. Indeed, the finding that domain 2 of hFceRIa when expressed as a single domain in either a filamentous phage display system (Robertson et al., 1993) or in a transient COS cell system (Mallimaci et al., 1993) exhibited only weak or no IgE binding, respectively, clearly indicates that domain 1 of hFceRIa plays an important structural role to ensure optimal receptor interaction with IgE. Of interest is the observation that rat IgE appears to interact with rat FceRI differently than hIgE does with hFceRI (Mallimaci et al., 1993). A chimera containing domain 1 of hFceRIa and domain 2 of rat FceRIa did not bind rat IgE; however, a chimera containing domain 1 of rat FceRI a and domain 2 of hFceRIa bound rat IgE with higher affinity than wild-type hFcsRIa. This finding, together with the observation that rat FceRI a-chain mAb that inhibit rat IgE binding recognize epitopes localized in domain 1, suggests that domain 1 is the principal intractive domain of rat IgE with rat FcsRIa. However, it is possible that domain 2 of rat FceRIa does play a direct binding role, but was not detected as domain 1of hFcsRIa may not be able to fully substitute for rat domain 1. In addition, rat IgE was found to bind the chimera comprising domain 1 of hFcyRIII and domain 2 of hFceRIa, suggesting that hFcsRIa interacts with rat IgE principally though domain 2, in contrast to that proposed for the binding of rat IgE by rat FceRIa. The different binding of rat and human IgE to their receptors is somewhat surprising in light ofthe high degree of sequence conservation of the rat and human FceRI a-subunits. In two separate studies, domain 2 of hFcsRIa has been found to contain multiple IgE binding regions (Hulett et al., 1994; Mallimaci et al., 1993). By interchanging segments of hFceRIa domain 2 with homologous segments of hFcyRIIa domain 2, we have identified at least three independent regions of hFceRIa domain 2 capable of direct2y binding IgE (Hulett et al., 1993,1994). The insertion of the hFceRIa domain 2 regions encompassed by residues Trp87to Tyr12' to and L Y S 'to~ G1u161 ~ into hFcyRII was found to impart IgE binding to hFcyRIIa (Table XIV, Fig. 1).These chimeric receptors bound IgE only in the form of immune complexes, implying that all three regions (together with domain 1as discussed above) contribute to the formation of a high-affinity IgE binding site. A similar approach ulilizing hFceRIalhFcyRII1 chimeras has also identified multiple IgE binding sites of hFceRIa domain 2 (Mallimaci et al., 1993). However, it should be noted that this approach was somewhat different in that homologous regions of hFcyRIII were inserted into hFceRIa, and the loss of IgE binding function was determined. The substitution of three

64

MARK D. HULETT AND P. MARKHOGARTH

regions of hFcsRIa with the equivalent regions of hFcyRIII was found to result in the complete loss of IgE binding; Ser% to PhelM, Arg"' to and Asp'% to Ser137.These regions correlate with two of the three IgE binding regions defined in our study, as both the Serg3to PhelWand Arg"' to G1i.1'~~ regions are situated in the Trp87to L Y S ' ~ ~ region, and the Asp'23 to region overlaps with our Tyr'29to Asp'45 region (Fig. 1). Substitution of a fourth region of hFceRIa (Lys'% to Ile167)resulted in reduced IgE binding and also the loss of binding of two inhibitory mAb and corresponds with the third of the direct IgE binding regions defined in our study, i.e., Lys'" to G1ulG1.It should be noted that rat IgE maintained binding to hFcsRIa substituted for the Lys'" to Ile167region (Mallimaci et al., 1993), implying that in contrast to hIgE, this region is not involved in the binding of rat IgE by hFcsRIa. Thus the results of these two studies clearly demonstrate that domain 2 of hFcsRIa is the principle IgE interactive domain of hFceRI, with at least four regions contributing to the binding of IgE, SerQ3to Phel', Arg"' to G ~ uTyr12' ~ ~ to~ , and L Y S 'to ~ ~G1u161 (Fig. 1). We have generated a molecular model of hFceRIa domain 2 based on the structure of domain 2 of CD4 (Hogarth et al., 1992). A similar model has also recently been proposed based on the structure of domain 2 of CD2 (Sutton and Gould, 1993). Based on these models, the identified IgE binding regions of hFcsRI domain 2 are situated predominantly in loop regions juxtaposed at the interface with domain 1, specifically the F-G, C'-E, and B-C loops, with contributions also from the B and C strands. The localization of the domain 2 IgE interactive sites to this region of domain 2, together with the finding that domain 1 also plays a key role in structural integrity of the receptor, suggests it is this interdomain region between domains 1 and 2 that comprises the IgE binding site of hFcsRIa. In support of this model, the mAb 15A5 which recognizes an epitope encompassed by residues 100-115 of hFcsRIa (corresponding to the € 3 4 loop and €3 strand) can completely block the binding of IgE to FceRI, suggesting the multiple IgE binding sites are indeed juxtaposed (Riske et al., 1991). The studies performed on hFcsRIa, together with those described earlier for hF&RII and mFcyRI, enable a number of statements to be made with respect to the molecular basis of Ig binding by these receptors in the context of the leukocyte Ig superfamily FcR. The two Iglike domain extracellular structure of the leukocyte FcRs clearly represents a conserved binding motif of this receptor family (Hulett et al., Overlapping region of the 123-137 region identified by Mallimaci e t a!. (1993) and the 129-145 region identified by Hulett et al. (1993).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

65

1991). An insight into the molecular basis of the interaction of this structural “unit” with Ig has been obtained through the studies on hFcsRI and hFcyRII. These functionally distinct receptors demonstrate similar yet distinct molecular bases of Ig binding. The second extracellular domains of both receptors are clearly responsible for the direct binding of Ig, with the first domains playing a crucial role in maintaining the structural integrity of the receptors to ensure optimal binding (Hulett et aZ., 1993; Mallimaci et al., 1993; Robertson, 1993). The mapping of the Ig binding regions in domain 2 of hFcsRI and hFcyRII has indicated that the structural basis of the interaction of these receptors with their Ig ligands is different. Domain 2 of hFcyRII contains a single region directly involved in the binding of IgG, whereas hFcsRI contains at least 3 distinct regions each capable of directly binding IgE and may provide a possible explanation in structural terms for the different affinities of these receptors for their ligands. Based on model domain structures, the Ig binding regions of both receptors are located in putative loop regions of domain 2 juxtaposed at domain 1 interface, suggesting that although fine structure binding differences exist, there is a common region involved in the binding of Ig. The single direct IgG binding region of hFcyRII corresponds to the F-G loop, as does one of the IgE binding regions of hFceRI. The additional binding regions of hFceRI are located in the C’-E loop and contained within the A-B-C strand region (probably the B-C loop) (Hulett et aZ., 1993,1994). Although only the F-G loop of hFcyRII is directly involved in the binding of IgG, the C’-E and BC loop regions also make secondary or indirect contributions to IgG binding, demonstrating the importance of this entire region (as seen for FcsRI) in the interaction with Ig. Furthermore, the influence of domain 1 on the Ig binding of domain 2 in both hFcyRII and hFcsRI is consistent with the location of the binding regions of these receptors in close proximity with domain 1, i.e., the interface of domains 1 and 2. Therefore, the finding that these loop regions are involved in the binding of Ig by two functionally distinct FcRs, in conjunction with the conserved nature of the two-domain extracellular binding “unit” of the leukocyte FcRs, strongly suggests that this region will also comprise the key interactive site of all members of this family. Based on this observation, it can be postulated that the structurally conserved “Ig folds” of the second extracellular domains of the leukocyte FcRs are providing the “scaffolding” to display variable loop regions which contain the determinants directing the specificity of these receptors for their Ig ligands. The binding site for FcsRI on IgE has been investigated extensively. Early studies demonstrated that the receptor binding site was located

66

MARK D. HULETT AND P. MARKHOCARTH

entirely in the constant region of IgE, as Fc fragments bound to mast cells with an affinity comparable to wild-type FcsRI (Stanworth et al.,1968, Ishizaka and Ishizaka, 1975). Attempts to further locate the binding site in the Fc portion of IgE by further fragmentation were unsuccessful, as none of the smaller fragments produced retained any FcsRI binding capacity (Ishizaka et al., 1970),and, indeed, the requirement for native IgE conformation for binding to FceRI was demonstrated (Ishizaka and Ishizaka, 1975; Rousseaux-Prevost et al., 1984). Early experiments using circular dichroism on reduced and heatinactivated IgE also implicated the Cs3 and C.54 domains in receptor binding (Dorrington and Bennich, 1973,1978).However, a later study suggested the principal binding region was the Cs2/Cs3 interface, as this region of rat IgE was protected from proteolysis when bound to rat FcsRI (Perez-Montfort and Metzger, 1982). A number of studies have utilized synthetic peptides representing segments of the Fc portion of IgE, or antibodies against such peptides, in attempts to identify the binding region(s) on IgE by inhibiting the IgE-FcsRI interaction. An early study reported the ability of a pentapeptide covering residues 320-324 in the Cs3 domain of hIgE, to inhibit allergic reactions by competing with IgE for basophil mast cell FcsRI (Hamburger, 1975). However, attempts to reproduce this finding were unsuccessful (Bennich et al., 1977). In another study using synthetic peptides generated to segments of Cs3 and Cs4 of rat IgE, receptor binding activity of several of the peptides was demonstrated, yet the most active of these was 1000-fold less efficient on a molar ratio basis than native IgE in inhibiting the IgE-receptor interaction (Burt and Stanworth, 1987). Experiments examining the inhibitory capacity of polyclonal antibodies raised against peptides representing human, rat, and mouse FCEsegments also produced results that were not definitive, as only partial or inefficient inhibition was obtained (Burt et al., 1987; Robertson and Liu, 1988). The use of recombinant FCEsegments proved to be a more successful approach in determining the region on IgE responsible for binding to FcsRI. Expression of cloned segments of human Fcs in bacterial systems, comprising Cs2, Cs3, and (2.94 domains, produced recombinant molecules that retained FcsRI binding activity (Liu et al., 1984; Kenten et al., 1984; Coleman et al., 1985). Using this approach, a 76-amino acid monomeric recombinant peptide (comprising residues 301-376 which span the C ~ 2 / C s 3interface of hIgE) was reported to bind hFcsRI with an affinity similar to that of native IgE (Helm et al., 1988). Studies with a series of smaller peptides covering this region subsequently indicated that residues 363-376 were not required for

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

67

binding to FcsRI (Helm et aZ., 1989).An octapeptide comprising residues 345-352 situated in Cs3 close to the Ce3/Ce4 interface has recently been reported to be capable of the specific inhibition of histamine release by human basophils (Nio et al., 1990). However, the amount of this peptide required to produce the inhibition was at molar concentrations several orders of magnitude higher than that reported for intact human myeloma IgE. In contrast to these findings, a recent study using recombinant human IgE Fc fragments expressed in mammalian cells suggested that the smallest fragment that retained FceRIa binding activity spanned amino acids 329 to 547, encompassing both the entire Ce3 and Cs4 domains (Basu et al., 1993). However, the binding assay used in this study differed from those experiments described above, in that the 329-547 fragment was shown to directly bind FceRIa, whereas the binding activity of the 301-376 and other fragments (above) were demonstrated only in inhibition assays. The use of resonance energy transfer studies to determine the distance between fluorescence probes placed at specific sites on either IgE or anti-IgE mAb bound to FcsRI, and probes at the surface of the cell membrane, has suggested that the Ce3 and Ce4 domains are positioned closest to the membrane, providingevidence for the involvement of these regions of IgE in the interaction with FcsRI (Holowka and Baird, 1983; Holowka et al., 1985; Zheng, 1991). These studies have also indicated that IgE is actually bent and does not change shape when bound to FceRI (Zheng et aZ., 1991). A model of IgE interaction with FcsRI based on this bent structural arrangement of IgE has recently been proposed, with the concave surface facing the membrane (Zheng et al., 1991; Sutton and Gould, 1993). The results of experiments assessing the capacity of a panel of mutant mouse IgE molecules, including deletions, truncations, and site-specific mutations, to bind to rat FcsRI on RBL cells suggest that the Ce3 domain is the principal domain involved in receptor binding (Schwarzbuam et al., 1989). Several groups have been able to inhibit the IgE-receptor interaction using mAb directed against IgE. The mAb are directed to epitopes localized to the Ce3 domain (Baniyash and Eshhar, 1984; Baniyash et al., 1988) lending further support to the suggestion that the Ce3 domain plays a crucial role in receptor binding. Chimeric IgE molecules have also been employed in attempts to clarify the roles of the Fce domains in binding to FceRI, in the hope that such molecules will have minimal conformational alterations. Chimeric mouse IgE molecules have been generated wherein constant domains were substituted with homologous domains from hIgGl (Weetall et d., 1990). These chimeric immunoglobulins were tested for

their capacity to bind FcsRI on rat RBL cells. The replacement of Ce4 with Cy3 of hIgGl was found not to alter the affinity of binding to FceRI compared with wild-type IgE. Furthermore, only those molecules containing both Ce2 and Ce3 were capable of binding FcsRI, suggesting both of these domains are required for the binding interaction (Weetall et al., 1990).A similar chimeric approach has been used whereby chimeric IgE molecules were produced by replacing heavychain constant domains in hIgE with homologous domains from mouse IgE (Nissim et al., 1991).These molecules were tested for their ability to bind both human FcsRI expressed in COS-7 cells and FcsRI on rat RBL cells. As described, human IgE does not bind to rodent FceRI, whereas rodent IgE will bind to human FceRI (Conrad et al., 1983). Chimeric IgE molecules containing mouse Cs3 were able to bind to both the human and rodent receptors, whereas the replacement of Ce2 of human IgE with Cs2 of mIgE did not confer the ability to bind rat FcsRI, yet the capacity to bind human FceRI was retained (Nissim et al., 1991). These findings provide strong evidence to suggest that Ce3 is the principal domain involved in the interaction of IgE with FceR1, but are in contrast to those described in the previous chimeric Ig experiments (Weetall et al., 1990), which suggested Ce2 was also necessary for IgE binding to FcsRI. The role of CE2 has therefore been proposed by the latter group as one of stabilizing the conformation of Ce3 to produce the correct IgE structure for binding to FcsRI. The inability of the chimeric hIgGl containing a Cs3 domain to bind FceRI (Weetall et al., 1990)may therefore be explained by Cy2 being unable to provide the stabilizing role of C e 2 . However, it should be noted that this suggested role of Ce2 in the binding of IgE to FceRI is also in contrast to the recent finding that a recombinant hFcs fragment comprising Cs3 and Cs4, but lacking Cs 2, retained high-affinity binding (Basu et al., 1993). In summary, analysis of the data outlined above suggests that Ce3 is the principal domain involved in the interaction of Fcs with FceRI. The roles of Ce2 and Cs4 are less clear; however, the weight of evidence suggests that both of these domains may provide supporting structural roles. The N-terminal region of Cs3 appears to contain the FceRIa binding site, with residues Asp33oto implicated as necessary for binding (Table XIV). A recent molecular model of the Fc region of IgE predicts an exposed region between residues Ala329-Gly3z from C E ~which , may represent a similar structure to the hinge proximal region in IgG, identified as crucial for the binding to FcyRI and FcyRII (Helm et al., 1991).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

69

IV. FcR Function

A. BIOLOGICAL FUNCTION FcRs initiate multiple effector functions, including cellular effector function involved in the neutralization and removal of infectious agents, immunomodulatory responses through triggering the release of inflammatory mediators and cytokines, and immunoregulatory signals in lymphocyte proliferation and antibody production. As FcR function has been extensively reviewed, it will only be described briefly herein (for review see Spiegelberg, 1984; Metzger et al., 1986; Unkeless et al., 1988; Mellman et al., 1988; Ravetch and Anderson, 1991; van de Winkel and Anderson, 1991, Fridman et al., 1993), and the remainder of this review focuses on new data that are emerging on how FcRs mediate their functions, with an emphasis on the mechanisms of signal transduction and the structural basis of FcR function. 1. FcyR

FcyRs mediate a large spectrum of biological functions, including phagocytosis of IgG-coated particles, killing of IgG-sensitized target cells by ADCC, triggering of the respiratory burst and release of inflammatory mediators, and regulation of B-cell activation and antibody production (for reviews see Mellman et al., 1988; Unkeless et al., 1988; van de Winkel and Anderson, 1991; Ravetch and Kinet, 1991). In many cases, assigning specific functions to the individual FcyR classes and isoforms has been difficult due to their overlapping expression on hematopoietic cells. However, the use of FcyR mAb has enabled a number of functions to be attributed to specific receptor classes. The recent cloning of individual FcyR isoforms has also provided the opportunity for their functions to be examined in isolation using transfection systems, and the specific functions of these isoforms are now beginning to be elucidated. All three classes of hFcyR participate in protective functions against antibody-coated infectious agents. Human FcyRI on monocytes and macrophages can mediate phagocytosis (Anderson et al., 199Oa) and ADCC (Graziano and Fanger, 1987; Fanger et al., 1989). The crosslinking of monocyte FcyRI during these events triggers superoxide production (Anderson et al., 1986; Pfefferkorn and Fanger, 1989) and also induces the release of IL-6 (Krutmann et al., 1990) and TNFa (Debets et al., 1990).In addition, the phagocytosis of IgG-coated erythrocytes via hFcyRI is accompanied by the release of IL-1 (Simms et az., 1991).

70

MARK D. HULETT AND P. MARKHOGARTH

The cross-linking ofhFcyRII has also been shown to trigger phagocytosis (Huizinga et al., 1989c; Anderson et al., 1990a)and ADCC (Graziano and Fanger, 1987; Fanger et al., 1989) on several cell types and mediate a respiratory burst by neutrophils (Willis et al., 1988; Tosi and Berger, 1988; Huizinga et al., 1989c; Crockett-Torabi and Fantone, 1990). Human FcyRII has been demonstrated to signal the release of TNF-a on freshly isolated monocytes, following the exposure of cells to proteolytic enzymes (Debets et al., 1990). Human FcyRIII can also mediate phagocytosis (Anderson et al., 1990a) and ADCC (Graziano and Fanger, 1987; Fanger et al., 1989) on macrophages and NK cells and generate superoxide production when cross-linked on monocytes and macrophages (Trezzini et al., 1990; Welch et al., 1990).These functions have also been described for the GPI-anchored form of hFcyRIII on neutrophils; however, the data are somewhat conflicting (Tosi and Berger, 1988; Salmon et al., 1987,1990; Selvaraj et al., 1989; Anderson et al., 1990a; Huizinga et a1., 1 9 9 0 ~Crockett-Torabi ; and Fantone, 1990; Kushner and Cheung, 1992; Edberg and Kimberly, 1994). It has been suggested that hFcyRII and FcyRIII may act in a cooperative manner on neutrophils to mediate degranulation (Boros et al., 1991). The mouse FcyRs parallel that of their human homologues in protective functions against antibody-coated infectious agents. Mouse FcyR have been shown to mediate phagocytosis of IgG-sensitized particles (Walker, 1977; Mellman et al., 1983) and endocytosis of immune complexes by macrophages (Ukkonen et al., 1986), as well as ADCC of IgG-sensitized target cells on both macrophages and NK cells (Fanger et al., 1989; Segal and Snider, 1989). The release of a variety of cytotoxic and inflammatory agents is also triggered by cross-Iinking FcyR on macrophages (Unkeless et al., 1981; Mellman et al., 1988). In addition, mFcyRII and FcyRIII on mast cells participate in the triggering of degranulation (Daeron et al., 1980)as well as endocytosis and phagocytosis (Lobell et al., 1994; Daeron et al., 1993,1994). Mouse and human FcyRIIb has also been shown to have the unique function of contributing to the regulation of B-cell activation, where it has been demonstrated that the cross-linking of FcyRII with surface Ig produces a dominant negative signal which inhibits B-cell proliferation and Ig production (Phillips and Parker, 1983; Klaus et al., 1987; Sarmay et al., 1991; Barret et al., 1990).

2. FCER FcsRI has been well-characterized as playing a central role in triggering the IgE-mediated allergic response (for review see Metzger et

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

71

al., 1986; Metzger, 1988; Kinet and Metzger, 1990; Kinet, 1990; Ravetch and Kinet, 1991). This has been most graphically demonstrated in vim by the inactivation of the FceRI a- or y-subunit genes by homologous recombination which results in loss of immediate-type hypersensitivity reactions (Dombrowiczet al., 1993; Takai et al., 1994). Activation of FceRI on the surface of mast cells and basophils, through the cross-linking of receptor bound IgE with a multivalent antigen, triggers immediate cellular events that result in the release of inflammatory agents. These events include cellular degranulation and release of granule contents such as histamine and the biosynthesis and secretion of arachidonic acid metabolites such as leukatrienes (Parker, 1987). A number of cytokines are also synthesized and secreted from mast cells and basophils upon activation of FceRI and include IL-1 to IL-6, TNF-a, IFN-y, GM-CSF, and members of the MIP-1 family (Plaut et al., 1989; Wodnar-Filipowicz et al., 1989; Burd et al., 1989; Ravetch and Kinet, 1991).While the role of FceRI in mast cell activation by IgE and antigen is well-documented the recent finding of FceRI on monocytes (Maurer et al., 1994) and eosinophils (Abdelillah et al., 1994) raises a great many questions in relation to function of those cells. B. MECHANISMS OF SIGNAL TRANSDUCTION Signal transduction by receptors following aggregation has been shown to involve a number of early cellular biochemical changes including accumulation of free intracellular calcium Ca2+, formation of inositolphosphates, and protein phosphorylation. These changes are believed to initiate a cascade of events that ultimately lead to cellular activation (Gardiner, 1989; Sefton and Campbell, 1991; Keegan and Paul, 1992).It is apparent that signal transduction by FcyR (and FcsRI) shares many features with the T- and B-cell antigen receptors. Indeed, FcyRI, FcyRIIIa, and some forms of FcyRII are multisubunit receptors-like the TCR which associates with CD3 chains and (,q,and y chains or surface Ig which associates with Iga and Igp (Gold and de Franco, 1994; Cambier et al., 1994). Like the antigen receptors, aggregation of these FcyR as well as FceRI induces protein tyrosine phosphorylation, PI turnover, calcium mobilization, and activation of protein tyrosine kinases (PTK) and is regulated to some extent by the membrane tyrosine phosphatase, CD45. It is interesting to note that in each case there is the direct or indirect involvement of tyrosine containing ARH motif (see below) in the initial signal transduction events which has been the focus of recent attention in a number of receptor systems. The aggregation of FcRs through the binding of

72

MARK D. HULETT AND P. MARKHOCARTH

immune complexes has been shown to initiate many of these early events, and how these events contribute to the signalingpathways coupled to FcRs is now beginning to be elucidated (for review see van de Winkel and Anderson, 1991; Ravetch and Kinet, 1991;Lin et a,?.,1994). 1. FcyR There are also several levels of complexity with respect FcyR signal transduction. First, there are those FcyR where expression and signaling are dependent on associated subunits, e.g., hFcyRIIIa with the FceRI y-subunit. Second, expression of some FcyR is independent of FceRI y-subunit but signaling is likely to be dependent on this subunit, e.g., hFcyR1. Third, FcR expression and signaling appears independent of association with additional subunits but there is an association with FceRI y, e.g., human FcyRII. This is very interesting in the light of recent data where FcyR+ cells from mice deficient in FceRI y-subunit do not phagocytose. Fourth, FcR function is only apparent after association with a separate functional receptor, i.e., Bcell antigen receptor-mIg and mFcyRIIb1. Fifth, there are multiple Fc receptor classes on myeloid cells; this is further complicated by the presence of multiple isoforms of the same class, e.g., hFcyRIIa, IIb2, IIc. However, under physiological conditions on cells that express multiple receptors it is likely that coaggregation of all receptor classes will take place. It is not known whether this will make any difference to the nature of signals transduced (since there is enormous redundancy in FcR signaling) or whether this initiates unique and complex effects that are not obvious from the investigation of individual receptormolecules. However, cooperation between different receptor classes in the initiation of signal transduction and function has been demonstrated. hFcyRIIa and hFcyRIIIb of neutrophils cooperate in the mobilization of CaZf as well as phagocytosis of Ig-coated particles or in phagocytosis via complement receptors (Naziruddin et al., 1992; Edberg and Kimberly, 1994) and neutrophil degradation (Boros et al., 1991).In addition hFcyRIIIb cross-linking induces phosphorylation of hFcyRII. Furthermore FcyRII of U937 cells modifies FcyRI-induced functions (Koolwijk et al., 1991). Recent studies are clarifying the picture and reveal a previously unrecognized sophistication and complexity in the way FcyR signal transduction is achieved and regulated. a. Second Messengers Early studies on low-affinity mFcyR demonstrated that cross-linking of the receptor could result in membrane depolarization and an increase in intracellular Ca2+ levels, suggesting that mFcyRII (or

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

73

FcyRIII) was able to function as a ligand-dependent ion channel or to activate other ion channels (Young et al., 1983a,b,1984). Membrane depolarization has also been observed following cross-linking of FcyR on human alveolar macrophages with IgG immune complexes (Nelson et aZ., 1985). Numerous studies have demonstrated that Ca2+ mobilization can be triggered by hFcyRI (Macintyre et al., 1989; van de Winkel et al., 1990a; Curnow et al., 1992), hFcyRII (Macintyre et al., 1988; van de Winkel et al., 1990a; Anderson and Anderson, 1990; Curnow et al., 1992), hFcyRIIIa (Cassatella et al., 1989; Curnow et al., 1992), and hFcyRIIIb (Kimberly et al., 1990).Changes in intracellular Ca2+levels have been well-characterized as an important component in cellular signaling mechanisms (Berridge, 1987; van de Winkel et al., 1990a), thus the ability of the FcyRs to trigger changes in intracellular Ca2+, suggests such fluxes may have a role in FcyR-mediated cellular responses. Localized Ca2+ gradients have been hypothesized to play important roles in phagocytosis; however, studies on the role of Ca2+ fluxes in FcyR-mediated phagocytosis have produced conflicting results (Young et al., 1984; Di Vergilio et al., 1988).Recently, it has been demonstrated that the cytoplasmic tail of hFcyRIIa contains distinct functional regions for either (i) initiating the internalization of small complexes or (ii) a Ca2+ flux and the phagocytosis of large particles (Odin et al., 1991). Although FcyRIIbl of B cells cannot induce Ca2+ mobilizations in its own right it can, when coengaged with membrane Ig, downregulate the calcium flux induced via the antigen receptor (Phillips and Parker, 1983; Amigorena et al., 1992a; Muta et al., 1994). The cross-linking of FcyRs has also been shown to stimulate increased phosphoinositide ( PI )turnover from activation of phospholipase C (PLC) which mediates the breakdown of phosphatidylinositol 4,5-bisphosphate (PIP2)into inositol-1,4,5-triphosphate(IP,) and diacylglycerol (DAG) (Berridge, 1987). IP, triggers the release of intracellular Ca2+, and DAG activates protein kinase C. Both hFcyRI and hFcyRII in the human monocyte cell line U937 have been shown to activate the Ca2+/PI signal transduction pathway (van de Winkel et aZ., 1990b), as has the triggering of hFcyRII in platelets (Anderson and Anderson, 1990) and hFcyRIIIa in NK cells (Cassatella et al., 1989). Recently, the co-crosslinking of CD4 and FcyR in U937 has also been shown to activate the Ca2+/PI signal transduction pathway, and, importantly, this may occur in vivo during HIV infection or following treatment with anti-CD4 mAb (Guse et al., 1992). The involvement of GTP binding proteins in the signal transduction

74

MARK D. HULETT AND P. MARKHOGARTH

of FcyR-triggered cellular events has been suggested, as a number of studies have demonstrated that some FcyR functions can be inhibited by treatment with pertussis or cholera toxins (van de Winkel and Anderson, 1991). The phagocytosis of IgG immune complexes by macrophages, but not neutrophils, has been found to be sensitive to pertussis toxin (Brown et al., 1987). In addition, hFcyRII- and hFcyRIIIbtriggered superoxide generation in neutrophils has been shown to be almost totally blocked by pertussis toxin (Feister et al., 1988; CrockettTorabi and Fantone, 1990). Human FcyRII-mediated neutrophil degranulation is also blocked by pertussis toxin but not as efficiently as superoxide generation induced via this receptor (Feister et al., 1988). An interesting development has been the recent observation that hFcyRIII and the fMLP receptor on neutrophils may be associated and share a common signaling pathway involving G proteins (Kew et al., 1992),although the nature of this is not clear and evidence indicates that at least calcium mobilization by fMLP and FcyR of neutrophils may have distinct early pathways (Rosales and Brown, 1992). Regulation of second messenger systems is also likely to play a role in modulating hFcyRI function since adenosine A1 or A2 receptors differently affect hFcyR1-mediated phagocytosis (Salmon et aZ., 1993). FcyR-mediated phagocytosis of IgG-coated targets is also accompanied by the generation of arachidonic acid and can be inhibited by the specific inhibition of phospholipase A2, suggesting that this enzyme may also have a role in signaling via FcyR (Lennartz and Brown, 1991).In addition, levels of cyclic adenosine monophosphate (CAMP) have been shown to increase following the interaction of IgG immune complexes with hFcyR (Ina et al., 1987).

b. Phoaphotylation and Protein Kinases Signal transduction by FcyRs has also been shown to involve phosphorylation, either of the FcyR complex itself or of other cellular proteins. Phagocytosis of IgG-coated erythrocytes by lipopolysaccharide (LPS)-stimulatedperitoneal macrophages was found to be accompanied by phosphorylation of several cellular proteins (Bronza et al., 1988).This pattern of phosphorylation was the same as that induced by phorbol ester treatment, suggesting that FcyR-mediated phagocytosis involves activation of the Ser/Thr protein kinase C. The serine phosphorylation of mFcyRI at serine has also been observed (Quilliam et al., 1993). Also inhibitors of tyrosine phosphorylation interfere with FcyR-dependent functions including immune complex-induced Ca" mobilization in neutrophils (Naziruddin et al., 1992)and phagocytosis by macrophages (Greenberg et al., 1993).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

75

The coengagement of membrane Ig and mFcyRIIb1, which results in an inhibitory signal that downregulates B-cell activation (Phillips and Parker, 1983; Klaus et al., 1987; Lazarus et al., 1991; Wilson et al., 1987), also results in serine and tyrosine phosphorylation of the receptor (Huzinker et al., 1990; Muta et al., 1994). Moreover the phosphoserine residue is located in the 47-amino acid insertion unique to the B-cell FcyRIIbl and probably essential for association with the cytoskeleton. At least one phosphotyrosine residue contained in a 13-amino acid segment common to mFcyRIIbl and mFcyRIIb2 and previously shown to be necessary for basolateral targeting in polarized cells is essential for this effect (Huzinker et al., 1991; Miettinen et al., 1992). While the phosphorylation of mFcyRIIbl is clearly induced and dependent on receptor:Ig coengagement, the mechanism by which downregulation occurs is not known. However, requirement for tyrosine phosphorylation indicates possible involvement of an SH2 containing protein (Sonyang et al., 1993; Koch et al., 1991; Cambier et al., 1994). The “modulation of the B-cell response” effect is dependent on the presence of the Ig-associated molecules-Iga or Igp (Muta et al., 1994)-implying an effect directly on these subunits or the complexes with which they associate and includes src-like kinases, fun, lyn, or blk; the syk kinase; as well as p40 and p42 (Pleiman et al., 1994; Clark et al., 1994; and reviewed in Cambier et al., 1994; Gold and Defranco, 1994). Clearly, FcyRII-associated phosphatases or regulators of srclike kinases, e.g., CSK (phosphorylates the regulatory tyrosine of src kinases), are obvious candidates in this context (Cantley et al., 1991; Okada et al., 1991; Bergman et al., 1992). However, a more distal effect cannot be excluded. Activation of human tonsilar B cells by cross-linking membrane Ig (without FcyR engagement) also results in phosphorylation of FcyRII, indicating that there is intracellular communication between Ig and FcyRII on these cells (Sarmay et al., 1990). In additionfyn and an unindentified serinehhreonine kinase p85-90 were found to be associated with FcyRIIb (Sarmay et al., 1994). Phorbol ester stimulation of human tonsilar B cells or a mouse B-cell line also resulted in serine phosphorylation of FcyRII implying a role for protein kinase C in FcyR function in B cells (Sarmay et al., 1990; Huzinker et al., 1990). Aggregation of FcyRII in platelets, myeloid cells, or neutrophils induces tyrosine phosphorylation of a number of substrates including FcyRII and src-related tyrosine kinases (Scholl et al., 1992; Huang et al., 1992; Ghazizadeh et al., 1994; Edberg and Kimberly, 1994). In

76

MARK D. HULETT AND P. MARKHOGARTH

platelets, FcyRIIa is associated with and activates src (Huang et al., 1992), in the myeloid cell line, THPl, FcyRII (isoform unknown but probably FcyRIIa or c) is associated with hck and Eyn (Ghazizadeh et al., 1994) and in neutrophils FcyRIIa with Fgr (Hamada et al., 1993). Phosphorylation and activation of these kinases is a common feature of signaling in these cells. Moreover tyrosine phosphorylation is important in FcyR-mediated endocytosis (Greenberget aZ., 1993;Ghazizadeh and Fleit, 1994). Aggregation of FcyR in 4937 cells also results in tyrosine phosphorylation of PLCyl which is an essential intermediate in the generation of lipid second messengers (Liao et al., 1992). The association of FcyRI with the y-subunit clearly indicates there will be some similarities between signal transduction through this receptor and FcyRIIIa in macrophages and monocytes. Aggregation of FcyRI induces tyrosine phosphorylation of a number of substrates (Scholl et al., 1992; Rankin et al., 1993)including the FcyRI-associated FcsRI y-subunit chain as well as phosphorylation and association of syk, not ZAP-70, with the activated complex (Agarwal et al., 1993; Duchemin et al., 1994).In addition, downstream events include inductive tyrosine phosphorylation including the tyrosine phosphorylation of phospholipase Cyl (Nishibi et al., 1990; Liao et al., 1992; Scholl et al., 1992, Ghazizadeh and Fleit, 1994). The role of the cytoplasmic tail of FcyRI is gradually being resolved. Mouse FcyRI is serine phosphorylated which is increased after stimulation of cells and probable activation of protein kinase C (Quilliam et al., 1993).The cytoplasmic tail is presumably required for internalization as a mutant FcyRI of NOD mice lacks the tail and cannot internalize (Prins et al., 1993). However the possibility of an alternate defect with the y-subunit was not investigated and therefore cannot be ruled out. Human FcyRI physically associates with the actin cross-linking protein-actin binding protein, an interaction that is reversed by the binding of Ig to FcyRI (Ohta et al., 1991). FcyRIIIa cross-linking also induces a number of events. In macrophages or myeloid cell lines the associated FcsRI y-subunit is phosphorylated in both the human (Masuda et al., 1993) and the mouse (Bonnerot et al., 1992). In natural killer cells hFcyRIIIa cross-linking induces the activation and association of the src-related kinase Zck. The lck is associated with the (-subunit and with the y-subunit (Salcedo et al., 1993; Pignata et al., 1993)and induces tyrosine phosphorylation of the 5- and y-subunits as well as PLC-yl and PLC-y2 (Einspahr et aZ., 1991; O’Shea et al., 1991; Ting et al., 1992; Azzoni et al., 1993; Liao et al., 1993) which is presumably necessary for mobilization of PI hydrolysis and a rise in intracellular calcium (Cassatella et al., 1989).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

77

FcyRIIIb of neutrophils is anchored by a GPI tail and mechanisms of signal transduction by this molecule are poorly defined. Nonetheless association with src kinases has been reported (Stefanova et al., 1991) as has cooperation in the induction of tyrosine phosphorylation of FcyRII (Edberg and Kimberly, 1994). The importance of the FceRI y-subunit to FcyRIII signaling is seen in a number of studies which show that in an analysis of human FcyRIIIa and the y-subunit in transfected cells, the y-subunit sequences are sufficient to allow signal transduction (Wirthmueller et aZ., 1992; Romeo et al., 1992; Bonnerot et al., 1992). Indeed the cytoplasmic tails of [-and y-subunits are sufficient to induce Ca2+mobilization in, and cytolysis by, transfected allospecific CTLs indicating that these tails contain sufficient information for the induction of a potent intracellular response (Romeo and Seed, 1991). Indeed chimaeric FcyRIIIaS or FcyRIIIay receptors are able to induce the expected spectrum of signaling events (Vivier et al., 1991,1992).Moreover, signaling through the associated [-subunit is dependent on the tyrosines within ARH motifs (Romeo et al., 1992). Like the antigen receptors of lymphocytes FcyR function is also modulated by CD45 (Gold and de Franco, 1994; Cambier et al., 1994; Rankin et al., 1993), cocrosslinking of FcyRI or FcyRII with the tyrosine phosphatase CD45 on ThP 1 cells, inhibited inductive tyrosine phosphorylation, as well as calcium mobilization (Rankin et al., 1993). 2. FcsRI

It has been well-documented that FcsRI aggregation is necessary to initiate the signal for exocytosis, and the minimum aggregate size appears to be a dimer (Siraganian et al., 1975; Segal et al., 1977), with the level of degranulation depending on the degree of aggregation (De Lisi and Siraganian, 1979; Fewtrell and Metzger, 1980; and reviewed in Metzger, 1992b). The aggregation of FcsRI on these cells has been shown to mediate a number of cellular events, including a rise in intracellular free Ca2+, hydrolysis of phosphoinositides, and protein phosphorylation, ultimately leading to degranulation and the release of inflammatory mediators of the allergic. reaction (reviewed Oliver et al., 1988; Beaven and Cunha-Melo, 1988; Metzger, 1992b; Benhamous and Siraganian, 1992; Beaven and Metzger, 1993). In conjunction with these events there are changes in cellular architecture and the topology of FcsRI (see below). a. Second Messengers Numerous studies have demonstrated a rise in intracellular Ca2+ upon aggregation of FceRI. This has been shown to result from mobili-

78

MARK D. HULETT AND P. MARK HOGARTH

zation of intracellular stores and also by influx of extracellular Ca2+ (Beavan and Cunha-Melo, 1988; Fewtrell et al., 1989; Millard et al., 1989; Matthews et al., 1989). The signal for the rise in Ca2+has been proposed, at least in part, to involve the second-messenger IPS,following the breakdown of phosphoinositides (see below) (Beaven and Cunha-Melo, 1988; Meyer et al., 1988; Matthews et at., 1989). The precise mechanism of influx of extracellular Ca2+ is not clear; however, it has been reported that cell activation results in opening of large conductive cation channels (Matthews et al., 1989), which may be under the control of GTP-binding proteins (Matthews et al., 1989; Wilson et al., 1989; Narasimhan et al., 1988).The isolation of a channel forming protein that is activated through interaction with aggregated FceRI has been claimed (Corcia et al., 1986; Hemmerich and Pecht, 1988); however, this proposed mechanism remains controversial (Metzger, 1988). The nature of the role of Ca2' in triggering cell exocytosis via FceRI remains unclear, as a number of studies have demonstrated that Ca2+is not essential for degradation (Neher, 1988; Hohman, 1988). While it is clear that a Ca2+ rise is associated with cell activation following receptor aggregation, it has been suggested in light of the above findings that Ca2+is probably acting synergistically with other second-messenger systems in the signal transduction pathways leading to exocytosis (Penner and Neher, 1988). Another well-characterized biological event initiated upon aggregation of FceRI is the hydrolysis of phosphoinositides, mediated by activated PLC (reviewed Beaven and Cuhna-Melo, 1988). Possible mechanisms for the FceRI-mediated activation of PLC include activation via an intermediary GTP-binding proteins, or through activation by protein tyrosine kinases. Studies of possible G protein involvement in the FceRI-mediated activation of PI hydrolysis, examining the effect of guanine nucleotides and pertussislcholera toxins on the FcsRImediated activation of this pathway, have produced 6oniiiEting resuit;; A number of such studies have produced findings consistent with a G protein being coupled to PI hydrolysis (Cockcroft and Gomperts, 1985; Ali et al., 1989), whereas others have suggested that neither a pertussis nor cholera toxin-sensitive G protein is involved (McClosky, 1988; Matthews et al., 1989). The weight of recent evidence supports the latter view, that the activation of PLC does not require a G protein as an intermediary. This has been suggested in studies demonstrating that FceRI-mediated PI hydrolysis in permeabilized cells was neither enhanced nor inhibited by GTP analogues (Saito et al., 1989) and exogenous GTP was not required for PI breakdown in RBL cell ghosts (Dreskin and Metzger, 1991).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

79

In addition FcsRI-induced PLC activity is also augmented by phosophenolpyruvate and creatinine phosphate by an as yet undefined mechanism (Dreskin et al., 1993). Recent experiments have provided strong evidence to suggest nonreceptor PTKs may provide a mechanism by which FceRI aggregation triggers PLC activation (Park et al., 1992; Fukamachi et al., 1992; Li et al., 1992). These studies demonstrated that cross-linking of FceRI in RBL-2H3 cells results in the rapid Tyr phosphorylation of PLC-yl. Furthermore, inhibition of PLC-yl Tyr phosphorylation with the PTK inhibitor Herbimycin A was shown to produce a concomitant inhibition of PI hydrolysis (Park et al., 1992). Thus, taken together with the recent report that PLC-yl is translocated to the cell membrane upon FcERI aggregation and may be the major isoform of PLC in RBL2H3 cells (Atkinson et al., 1992; Park et al., 1992), these indicate the importance of PTK activation of PLC-yl for the triggering of PI hydrolysis by FcsRI. In addition to PLC-y1, phospholipase A2 and D are also activated in RBL cells following FceRI aggregation (Narasimhan et al., 1990; Gruchalla et al., 1990). Phospholipase A2 may play an important role in the production of inflammatory mediators, as the enzyme generates arachidonic acid which is an intermediary in the synthesis of prostaglandins and leukotrienes (Parker, 1987). Phospholipase D, upon activation, generates phosphatidic acid which is then converted to DAG. Recent reports suggest that this mechanism of DAG production may be quantitatively more significant than that via phosphatidyl inositols (Gruchalla et aE., 1990; Kennerly, 1990; Nakashima et al., 1991). DAG is known to activate protein kinase C (PKC); however, the role of PKC in FceRI signal transduction is not certain, as although FcsRI aggregation induces membrane translocation of PKC, experiments depleting PKC do not inhibit degranulation (White and Metzger, 1988). In addition selective inhibitors of PKC enhance IgE-mediated responses (Amon et al., 1992).

b. Phosphorylation and Protein Kinases Like the FcyR above, inductive tyrosine, serine, and threonine phosphorylation of the FceRI complex and a number of intracellular substrates are the earliest signaling event (Quarto and Metzger, 1986; Paolini et al., 1991; Park et al., 1992; Eiseman and Bolen, 1992; Hutchcroft et al., 1992; Li et al., 1992).Although the receptor has no intrinsic protein kinase activity, protein kinases are physically associated with the complex (Quarto and Metzger, 1986; Shimizu et al., 1988; Paolini et al., 1992) and are activated following receptor aggregation. These

80

MARK D. HULETT AND P. MARKHOGARTH

include the src-related tyrosine kinases -lyn in RBL cells and yes in mouse PT18 cells (Eisman and Bolen, 1991; Hutchcroft et al., 1992). In addition to the src-related kinases, the PTK 72 (syk) (Stephen et al., 1992) is also associated with the receptor complex (Hutchcroft et al., 1992; Benhamou et al., 1990b; Benhamous and Siraganian, 1992) and activation of this kinase is independent of G protein activation or PI hydrolysis (Stephan et al., 1992). In addition a serine phosphoprotein pp125 is also associated with the receptor complex (Paolini et al., 1992).Thus it appears that different protein tyrosine kinases associated with FcyeRI and indeed the presence of serinekheronine kinases (Paolini et al., 1992) adds complexity to signaling by this receptor and may have wider implications for signaling in FcyR that share subunits with FceRI. Aggregation of FcERI produces rapid phosphorylation of the psubunit (on Tyr and Ser) and the y-subunit (on Tyr and Thr) (Paolini et al., 1991; Li et aZ., 1992).The phosphorylation is restricted to active receptors and is immediately reversible on receptor disengagement by as yet unidentified phosphatases (Paolini et al., 1991,1992). In addition multiple intracellular substrates are phosphorylated (and dephosphorylated) (Bonhamou et al., 1990b; Paolini et al., 1991,1992; Li et al., 1992; Kawakami et al., 1992)on tyrosine, serine, or threonine residues. The balance between phosphory lation and dephosphorylation contributes to the regulation of receptor function in many receptor systems and recent experiments indicate for FceRI, small stable aggregated receptor complexes continue to signal (Kent et al., 1994).Indeed rather than a brief initiation of the signal transduction cascade, these clusters continue to undergo phosphorylation and dephosphorylation, which as the authors point out has major ramifications for a variety of receptor systems (Kent et at., 1994). While most of the intracellular substrates are yet to be identified a 75-kDa phosphoprotein (inducibly phosphorylated) has been cloned recently. This SPY75, which contains helix-turn-helix motifs of DNA binding proteins as well as an SH3 domain, may be an important intermediary in FceRI and signal transduction to the nucleus (Fukamachi et al., 1994). Other substrates include the protooncogene product uau (Margolis et al., 1992; Boguski et al., 1992), an adaptor Nck (Park and Rhee, 1992), ERK2, a serine/threonine MAP kinase (Fukamachi et al., 1993), and ~ ~ 1 a PTK 2 5found ~ in~ focal ~ adhesions (Hamawy et al., 1993; Schaller et al., 1992) and may play a similar role in mast cells (Pfeiffer and Oliver, 1994). The y-subunit of FceRI, and to a lesser extent the p-subunit, has been found to be crucial structures for the signaling of FceRI. A number of

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

81

studies have suggested that Tyr phosphorylation of, and association with, protein tyrosine kinases is essential for triggering by the y subunit in both FcyRIII and the TCR (Romeo and Seed, 1991;Letourneur et al., 1991), and presumably the same applies for the y-subunit in FcsRI. The /3- and y-subunits also contain ARH motifs (see above) (Reth,1989; Reth et al., 1991) which are likely to be functionally important for these receptor complexes to mediate signal transduction (Eiseman and Bolen, 1992; Daeron et al., 1992). The findings of the studies presented above suggest that protein phosphorylation and particularly Tyr phosphorylation by activated PTKs play an important role in the signaling of FcsRI. Indeed, recent functional studies have demonstrated that many of the early biochemical changes initiated by aggregation of FcsRI are inhibited by Tyr kinase inhibitors and enhanced by tyrosine phosphatase inhibitors (Deanin et al., 1991a,b; Stephan et al., 1992). The observation that activated FcsRI is part of a large complex that includes multiple phosphoproteins and associated kinases indicates that the role ofphosphorylation in signal transduction by FceRI is complex. Clearly, despite these significant advances into the elucidation of the signaling mechanisms of FceRI, further studies are required, as for the FcyRs, to complete the understanding of the complex nature of these patkways. C. STRUCTURAL BASISOF FcR FUNCTION AND SIGNAL TRANSDUCTION I. The Role of the ARH Z Mot$ The signal transducing subunits of the T-cell (TCR-6, 7 , CD3-y, 6, E ) and B-cell antigen receptors (Iga and Igp) contain a homologous segment termed the antigen receptor homology (ARH1) motif also known as the tyrosine-based activation motif (TAM) and signal transduction events by these receptors are dependent on this motif (Reth, 1989, Weiss 1994, Cambier et al., 1994; Gold and de Franco, 1994; Flaswinkel and Reth, 1994). The consensus sequence of this motif is DIE-xs-DIE-xz Y-x-x-LII -X7.12-Y-x-x-LII(where x is any amino acid; reviewed in Cambier et al., 1994). Tyrosine phosphorylation of this motif is a major event in antigen receptor signaling and is associated with the interaction with SH-2 domains (src homology region 2) of signaling effector proteins (Koch et al., 1991). Indeed recognition of the sequence YXXL by SH-2 domains is likely when tyrosine is phosphorylated (Sonyang et al., 1993). In B cells a number of proteins have been shown to complex with the motif in the membrane Ig-associated proteins, Iga or Igp (Clark et al., 1992). In addition to the role of the phosphotyrosines other

82

MARK D. HULETT AND P. MARKHOGARTH

segments of the ARHl motif also influence the specificity of the interaction between the motif and intracellular molecules (Clark et al., 1994). Like the antigen receptors some FcR and their associated subunits contain an ARHl motif (TableXVI). While segments ofthe cytoplasmic tail of FcsRIP and FceRIy and FcyRIIa conform well to the consensus, the FcyRIIb isoforms conform less well but clearly have elements within them that are likely recognition sites for SH-2 containing proteins (Sonyang et d . ,1993).As will be seen from the discussion below sequences in these motifs play major roles in connecting immune complex-induced aggregation of FcR to downstream effector systems.

2. FcyR The molecular cloning of the FcyRs has revealed a family of structurally similar receptors that are homologous in their ligand binding extracellular regions, yet exhibit marked divergence in their transmembrane and cytoplasmic tail regions. The structural heterogeneity in the cytoplasmic tails of the FcyR, combined with their differential cellular expression, is presumably the mechanism by which these receptors upon interaction with a common ligand, can mediate such a diverse range of cellular responses (Table XVII). Structure/function studies of the cytoplasmic tails of specific FcyR isoforms (or their associated subunits in the case of F q R I I I ) are elucidating the structural basis of how the FcyR are able to trigger different intracellular signals and cellular functions. Since the greatest sequence differences 'between FcyKll isoforms are found in cytoplasmic tails it would seem likely that there would be some differences in the nature of the signal transduced by these receptors or their role on the cells that express them. Indeed, considerable functional differences exist between isoforms. Thus aggregation of FcyRIIa induces calcium mobilization (Odin et al., 1991; Kolanus et aZ.,1992;Van den Herik-Oudijk et al., 1994)and tyrosine phosphorylation (Huang et al., 1992) as well as phagocytosis (Tuijnman et al., 1992; Indik et al., 1991). However, mouse and human FcyRIIbl or b2, cannot induce tyrosine phosphorylation or Ca2' mobilisation (Amigorena et d., 1992a; Bonnerot et al., 1992; Kolanus et d.,1992; Van den Herik-Oudijk et al., 1994) and differ in their ability to internalize or cap, as FcyRIIb2 internalizes but does not cap whereas FcyRIIbl caps but does not internalize (Miettinen et al., 1989, 1992). These differences are presumably reflected in the amino acid differences in the cytoplasmic tails of the receptors which affects the nature of signal transduction. The structural roles of the cytoplasmic tails of the two mFcyRIIb isoforms in the function of these receptors have been studied exten-

TABLE XVI COMPARISON OF THE ARHl MOTIF OF FcR AND OTHER MULTISUBUNITRECEITORS" Receptor subunit mFcaR1-y FcsRIP hFcyRIIa mFcyIIb (1 or 2) huFcyRIIb (1 or 2) mMB-1 mB 29 mTcR< mCD3c a

-1 A

Sequence

n n n

A I A S R E K A D A V Y T G L N T R S - - - - - Q E T

E E

See text for references.

V Y - - - - - - - S P I N P R A P T D D D K N I K H P E A D E E T E H D M H P D A L E E P - D D N N N R

L I L K

D D G G

D Q R Q

-

-

-

- - C S M - - C S M - - R E E - - R D : L

n

Y E T -

U Y

Y

S E L T -

Y

Q N H

- Q N R

1 L

L

84

MARK D. HULETT AND P. MARKHOGARTH

TABLE XVII COMPARISON OF FUNCTIONS MEDIATEDBY FcyR ISOFORMS Function

Receptor

Inductive Association tryosine with Cae+ phosphorylation kinases Mobilization Endocytosis Phagocytosis Capping

FcyRI FcyRIIa" FcyRIIbl FcyRIIb2 FcyRIIc FcyRIIIa FcyRIIIb a

+

+ -

+ + +

+ +

?fyn' ? ?+

+

+

i

i

+

i i

+

+ -

+ .d + -

+ + -b

?+"

-

+-

+ -

Human only.

* By transfection only, not normally expressed on cells with this function. No apparent phagocytosis in FceRIy-deficientmice. ., Not known.

sively, enabling different regions to be assigned specific functions. As described earlier, mFcyRIIbl and mFcyRIIb2 are identical, with the exception of an in-frame insertion of 47 amino acids in the mFcyRIIb1 isoform, generated by alternate splicing (see section 11,A72).The regulated cell-type expression of the two isoforms suggests that the 47amino acid insertion plays an important role in the differential function of these receptors on macrophages and lymphocytes. Indeed, experiments involving the transfection of wild-type and cytoplasmic taildeleted mutant forms of mFcyRIIb1 and b2 into fibroblasts or rat RBL mast cells (Miettinen et aZ., 1989; Daeron et al., 1993) indicated that only the mFcyRIIb2 form (lacking the insertion) was able to mediate efficient endocytosis and delivery to lysosomes of ligand-receptor complexes via clathrin-coated pits (Miettinen et aZ., 1989). It has since been demonstrated that the 47-amino acid insertion of mFcyRIIb1 actively prevents coated pit localization by tethering the receptor to the cytoskeleton. The insertion does not simply disrupt a coated-pit localization domain, as the insertion was found to prevent endocytosis irrespective of its position in the cytoplasmic tail (Miettinen et al., 1992). The coated-pit localization domain was mapped using deletion mutants to the distal portion of the cytoplasmic tail of mFcyRIIb2 (Fig.3) and it is interesting to note that this region is the only stretch of conserved amino acids between mFcyRIIb2 and hFcyRIIb (Miettinen et al., 1992). However, the insertion containing the coated-pit localization domain was not required for phagocytosis of Ig-coated

85

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

/v rm

I

Y

A E N T T I T Y 4 S 4 L

k 4

I N S E

R

T

-

tethers to cytoskeleton (position independent) prevents internalisation -does not prevent attenuation of Ca+

+ mobilisation bymlg

setine phosphotylated (in mouse) enables capping

association with coated pits internalisation *targets for antigen processing attenuates Ca++ mobilisation in membranelg signalling targets to basolateral surface of polarised cells tyrosine important in function

H

P d

Q N H

?Phagocytosis

I

FIG. 3. Structure-function relationships of the cytoplasmic tail components of FcyRIIbl and b2. The partial sequences shown are of mouse but the human is highly conserved. The “insert” represents the segment of FcyRIIbl that arises from differential mRNA splicing of the FcyRIIb gene. This contains 47 or 18 amino acids for mouse or human FcyRIIb, respectively. The YXXL motif is circled. (Data derived from Hunziker et QZ., 1990; Englehardt et QZ., 1991; Amigorena et QZ., 1992a; Miettinen et QZ., 1992; Daeron et QZ., 1993; Van den Herik-Oudijk et QZ., 1994.)

86

MARK D. HULETT AND P. MARKHOGARTH

parasites, as both mFcyRIIbl and b2 could internalize these whel transfected into fibroblasts (Joiner et al., 1990). However, using Ig coated erthyrocytes rather than parasites, Daeron et al. (1993) showec: these could be phagocytosed by mFcyRIIb2 but not by mFcyRIIbl moreover, this was dependent on two tyrosine residues contained ir the cytoplasmic tail that are found in the ARH-like segment (Tables XVI).Interestingly there was some compartmentalization ofphagocytosis and endocytosis as mutation of tyrosine in the segment-YQNHI resulted in loss of phagocytosis but not endocytosis which was dependent on tyrosine in the YSLL motif (Daeron et al., 1993; Table XVII). While there is abundant evidence that mFcyRIIb2 can internalize and phagocytose large particles it was interesting to note that macrophages from FceRIy-deficient mice still express FcyRII but do not phagocytose. Since it is likely that the FcyRII expressed is mFcyRIIb2, the results imply an additional level of complexity not previously recognized (Takai et al., 1994). A comparison ofmFcyRIIbl and IIb2 has also been made by transfection of a FcsRII- B-cell line or rat RBL mast cell line (Amigorena et al., 1992a; Fridman et al., 1992; Daeron et al., 1993; Alber et al., 1992). The 47-amino acid insert prevented endocytosis and was required for capping (Amigorena et al., 1992a; Daeron et al., 1993). However, it was not essential for the modulation of calcium mobilization by FcyRI1:mIg coengagement as mFcyRIIbl or b2 were equally effective. This effect was dependent on the sequence AENTITYSLLKHP in which the tyrosine was a crucial residue (Amigorena et al., 1992a; Muta et al., 1994). Based on the above findings, it is apparent that the intracytoplasmic regions of mFcyRIIbl and b2 contain distinct functional regions responsible for capping, endocytosis, phagocytosis, and inhibition of Bcell activation (Table XVII). A functional map of the cytoplasmic tail regions of mFcyRIIbl and b2 has been proposed (Fridman et al., 1992; see Fig. 3 also). Similar studies of human FcyRIIb also indicate that hFcyRIIb2 can internalize whereas hFcyRb 1cannot in transfected fibroblasts (Engelhardt et al., 1991) or the FcyRII- mouse B-cell line (Van den HerikOudijk et al., 1994).Internalization of soluble complexes by hFcyRIIb in kidney fibroblasts (BHK-21) was reduced by deletion of the cytoplasmic tail, The significance of the serine phosphorylation of mFcyRIIbl to hFcyRIIbl is unclear since there are no serine residues present in the insert of hFcyRIIb1 which also only has 17 amino acids (Brooks et al., 1989). However, hFcyRIIb is clearly serine phosphory-

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

87

lated (Sarmay et al., 1991) though the identity of the residue remains to be determined. Human FcyRIIa mediates calcium mobilization and endocytosis (Odin et al., 1991; Van den Herik-Oudjik et al., 1994)and phagocytosis (Odin et al., 1991). Signaling functions have been localized to cytoplasmic segments that contain the ARH motif (Table XVI). Deletion of segments containing the mouse C-terminal tyrosine of the motif resulted in loss of calcium mobilization and phagocytosis of large particles but endocytosis function was still intact. However, deletion of substantial portions of the tail including the entire ARH motif resulted in loss of endocytosis as well (Odin et al., 1991).Using chimeric receptors with CD4 or CD16 extracellular domains fused to the cytoplasmic tails of hFcyRIIa, IIc, IIbl, and IIb2, only FcyRIIa and IIc initiated Ca" mobilization and activation of effector cell function measured by lysis of target cells. These functions in hFcyRIIa were dependent on the ARH motif of the cytoplasmic tail and mutation of either tyrosine in this motif resulted in loss of function (Kolanus et al., 1992). Other studies have also confirmed that hFcyRIIa can trigger the phagocytosis of IgG-sensitized erythrocytes in transfected fibroblasts or in a human erythroleukemic cell line (HEL) (Indik et al., 1991; Tuijnman et al., 1992).This study also demonstrated that hFcyRI could not mediate phagocytosis in COS cells, even though transfected cells avidly bound IgG-sensitized erythrocytes (Indik et al., 1991). As described above, hFcyRIIIa and mFcyRIII are multiple component surface complexes which comprise a ligand binding chain and additional associated subunits. In addition to playing an essential role in cell-surface expression, these subunits are also crucial in the signal transduction of these receptors. Mouse FcyRIII mediates cell activation through its associated ysubunit s. Using the FcyR-, sIgG+ B-cell line llA1.6 mouse FcyRIII containing a complete or truncated FcyRIII a-chain cytoplasmic tail was able to trigger the same cellular events as sIgG, including Ca2+ mobilization, tyrosine phosphorylation, and 11-2 secretion, but only when associated with the y-subunit (Bonnerot et al., 1992). The cytoplasmic domain of the y-subunit was sufficient to trigger cell activation. Similar findings have also been reported from experiments that examined the capacity of mFcyRIIbl, b2, and mFcyRIII to mediate activation events following their cross-linking on transfected rat RBL-2H3 cells (Alber et al., 1992, Latour et al., 1992; Daeron et al., 1993,1994). Of these three mFcyR, only mFcyRIII was able to trigger cell activation, mediating the same cellular events as FceRI. These results imply

88

MARK D. HULETT AND P. MARKHOGARTH

that the y-subunit associated with both mFcyRIII and mFcsRI is responsible for mediating the similar signal transduction pathways observed following activation of these receptors. Experiments examining the activation capacity of endogenous mFcyRIIbl, b2, and mFcyRIII on mast cells also support these findings (Katz et al., 1990). The y-subunit ofmFcyRIII has also been shown to have an important role in determining receptor internalization and phagocytosis (Amigorena et al., 1992b; Daeron et al., 1993,1994). Mouse FcyRIII, when transfected into FcyR- antigen-presenting B lymphoma cells, was able to mediate ligand internalization and increased efficiency of Ag presentation (when Ag was coupled to IgG). In the RBL cell line, FcyRIII or chimeric FcR fused to the cytoplasmic tail of the y-subunit could internalize small complexes and phagocytose large complexes. The signal for these events was found to be independent of the a-chain cytoplasmic domain, but required the y-subunit. Partial or complete deletion of the ARHI motif or mutagenesis of the tyrosines in the y-subunit tail have a profound effect on function. Such mutants fail to trigger cell activation (Bonnerot et al., 1992; Daeron et al., 1994), internalization (Amigorena et al., 1992b; Park et al., 1993a,b; Daeron et at., 1994), and tyrosine phosphorylation (Bonnerot et al., 1992). The y-subunit associated with FcyRIII (and FcERI)forms part of a family of homologous molecules, which include the TCR 5- and qsubunits (Orloff et al., 1990).The findings outlined above, which demonstrate the crucial role of the y-subunit in FcyRIII-mediated cell activation, have been supported in a number of studies that have also examined the role of the related 5- and q-subunits in signaling. In one study, chimeric receptors were generated in which the cytoplasmic domains of the y-, 5-, or 7-chains were linked to the transmembrane and extracellular regions of CD4. These chimeras were transfected into CV1 cells and were demonstrated to have the capacity upon aggregation to signal a rise in intracellular Ca2+and to activate the cells to direct cytolytic responses, providing evidence that the cytoplasmic domains of each ofthese subunits contain region(s)important for receptor signaling (Romeo and Seed, 1991). Similar results have been obtained with chimeric receptors consisting of (i) the transmembrane and extracellular regions of CD8 linked to the cytoplasmic tail of the (-chain in the TCR- Jurkett cell line (Irving and Weiss, 1991; Irving et al., 1993), (ii) the extracellular domain of IL-2a linked to the y or ( transmembrane and cytoplasmic tails in T cells and mast cells (Letourneur and Klausner, 1991), and (iii) the extracellular regions derived from hFcyRII1, transmembrane regions from CD5 or CD7,

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

89

and cytoplasmic tails containing an 18-amino acid motif from the 6chain, in T cells (Romeo et al., 1992). These results indicate that the y, 6, and q proteins comprise a family of “trigger” molecules that can initiate signal transduction events as part of a number of different receptor complexes.

3. FceRl The functional regions of FceRI involved in signaling of the receptor and interaction with the cytoskeleton have been investigated. In these studies, various mutant receptors comprising truncated cytoplasmic domain subunits were examined for their functional capacity to mediate these biological events. The roles of the cytoplasmic tails of the FceRI subunits in mediating cell activation were determined by reconstituting mutant receptors by transfection into the mouse mastocytoma cell line, P815, and assessing their abilities to initiate intracellular signaling (Alber et al., 1992). Previous experiments demonstrated that these cells, when transfected with wild-type FceRI, were capable of expressing early signals of activation upon receptor aggregation, including a rise in Ca2+,hydrolysis of phosphoinositides, and protein Tyr phosphorylation (Miller et al., 1990).FceRI comprising an a-subunit with a truncated cytoplasmic tail functioned as a wild-type receptor. This receptor when aggregated was able to stimulate a rise in Ca2+,PI turnover, and phosphorylation ofTyr; suggesting that this region ofthe FcERIa-chain does not contain crucial site(s) for receptor signaling. Reconstituted receptors lacking the p-subunit, or containing a p-subunit with a truncated NH2-terminal cytoplasmic domain, were also fully functional. However, FceRI containing a @subunit with a truncated COOH-terminal cytoplasmic domain were completely inactive (Alber et al., 1992). This unusual finding is difficult to explain; however, similar results have been observed in studies examining the topology of FceRI in relation to receptor aggregation (see below). These results suggest that such mutants do not aggregate correctly and that the receptor subunits are oriented inappropriately for receptor signaling. The possibility that inefficient aggregation of FcsRI occurs has also been suggested previously (Berenstein et d., 1990). Significantly, the deletion of the cytoplasmic domain of the y-subunit abolished the ability of FceRI to stimulate P815 transfectants (Alber et al., 1992). The deletions of the y-subunit cytoplasmic sequence and the COOH-terminal cytoplasmic sequences of the p-subunit, which ablate function, result in the removal of ARHl motifs from these molecules (Blank et al., 1989). Deletions of the p-subunit NH2-terminal cyto-

90

MARK D. HULETT AND P. MARKHOGARTH

plasmic sequences do not ablate function and do not contain an ARHl motif. The topology of FcERI is also altered upon aggregation, the receptor becoming immobile (Menon et al., 1984; Mao et al., 1991)and redistributed into clusters (Becker et al., 1973; Carson and Metzger, 1974; Mendoza and Metzger, 1976; Seagrave et al., 1991). The aggregated receptors also become associated with the degergent-insoluble “cytoskeletal” fraction (Seagrave and Oliver, 1990; Apgar et al., 1990; Mao et al., 1991) and internalized (Iersky et al., 1983; Pfeiffer et al., 1985). The same approach as above using mutant FcsRI receptors has been employed to define the crucial region(s) of the cytoplasmic domains of the receptor subunits that may interact with cytoskeletal elements and thus influence the topological properties of FcsRI following aggregation (Mao et al., 1991,1992,1993Metzger, 1992a). Experiments have been performed to examine the lateral mobility and endocytosis of mutant FcERI receptors before and after aggregation, in both transiently transfected COS cells and stably transfected P815 cells. Receptors lacking the cytoplasmic domain of the a-chain, the NH,terminal cytoplasmic domain of the P-chain, the entire @chain, or the cytoplasmic domain of the y-chain all exhibited the same topological properties as the wild-type FcsRI. Only those receptors comprising a p-chain with a truncated COOH-terminal cytoplasmic domain were abnormal, exhibiting only partial immobilization, reduced internalization, and impaired association with coated pits (Ma0 et al., 1991,1992, 1993).Based on these results, it has been suggested that the interaction between the receptor and the cellular components responsible for the phenomena does not critically depend on any single cytoplasmic domains of the receptor subunits. The resistance of FcsRI to solubilization by detergents following aggregation was also examined using the above panel of mutant FcsRI receptors, again in transfected COS and P815 cells. These experiments produced results similar to those above, demonstrating that the cytoplasmic domains of the receptor subunits had no effect on insolubilization by aggregation. In addition, a chimeric receptor comprising only the extracellular domain of the FcsRI a-subunit attached to the membrane by a GPI anchor was also resistant to solubilization after aggregation (Ma0 et al., 1992). These results support a model in which the bridging of FcsRI results in receptors becoming enmeshed in a network of skeletal proteins, either on the outside and/or on the inside of the membrane, through a nonspecific physical effect rather than from specific interactions.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

91

V. Concluding Remarks

Clearly, significant advances have been made into understanding the biology of the FcyR and FccR. The characterization of the proteins, transcripts, and genes of the different classes ofboth these FcR families has largely been completed and those for IgM, IgD, and IgA receptors still to be completed. Attention is now turning to understanding the nature of the mechanisms by which these receptors and their subunits are able to mediate their functions. Structural analysis of the functional basis of FcyR and FCERsignaling is beginning to shed some light on how these receptors trigger biological responses, and the molecular dissection of the pathways involved in signaling is also underway. The next few years should see major progress into the unraveling of the complex nature of these mechanisms. Studies examining the molecular basis of the interaction of FcRs and Ig are another important future direction in the FcR field. The understanding of these interactions may provide the means to devise strategies to inhibit pathophysiological effects of FcR function, which would have far reaching implications in the treatment of antibodyinduced hypersensitivity. Indeed, the use of soluble recombinant Fc receptors to treat these conditions has been established in experimental animal models (Ierino et aZ., 1993).

ACKNOWLEDGMENTS We thank all our colleagues who generously contributed their unpublished manuscripts. We also thank Mimi Morgan, Julie van Ballegooyen, Sue Collins, and Toula Athanasiadis for desktop publishing assistance. Special thanks goes to Ross Brinkworth for the molecular model of hGcyRII domain 2.

REFERENCES Abdellilah, S. G., Lamkhoued, B., Ochial, K., Tanaka, Y., Delaporte, E., Capron, A,, Kinet, J-P., and Capron., M. (1994). High-affinity IgE receptor on eosinophils is involved in defence against parasites. Nature (London)367, 183-186. Abo, T., Tilden, A. B., Balch, C. M., Kumagai, K., Troup, G. M., and Cooper, M. D. (1984). Ethnic differences in the lymphocyte proliferative response induced by a murine IgGl antibody Leu4 to T3 molecule. J. E x p . Med. 160,303-309. Abu-Ghazaleh, R. I., Fujisawa, T., Mestecky, J., Kyle, R. A,, and Gleich, G. J. (1989). IgA-induced eosinophil degranulation. J. Immunol. 142,2393-2400. Aganval, A., Salem, P., and Robbins, K. C. (1993). Involvement of p72’Yk a protein tyrosine kinase in Fc gamma receptor signalling. J. Biol. Chem. 268, 15900-15905. Ahouse, J. J., Hagerman, C. L., Mittal, P., Gilbert, D. J., Copeland, N. G., Jenkins, N. A., and Simister, N. E. (1993).Mouse MHC class I-like Fc receptor encoded outside the MHC. J . Immunol. 151,6076-6088. Aicher, W. K., McGhee, M. L., McGhee, J. R., Moldoveanu, Z., Kidd, V. J., Tomana,

M., Mestecky, J., Eldridge, J. H., Meyer, T. F., and Kiyono, H. (1992). Properties of IgA-binding receptors on murine T cells-relative importance of FcaR, P-galactosyltransferase and anti-secretory component reactive proteases (ACSP).Scand.J. Immunol. 35,469-486. Alber, G., Miller, L., Jelsema, C., Varin-Blank, N., and Metzger, H. (1991). Structure/ function relationships in the mast cell high-affinity receptor for IgE (FcsRI): Role of cytoplasmic domains and of the /3 subunit. J . Biol. Chem. 266,22613-22620. Alber, G., Kent, U. M., and Metzger, H. (1992).Functional comparison of FceRI, FcyRII, and FcyRIII in mast cells. J. Immunol. 149,2428-2436. Albrechtsen, M., Yeaman, G. R., and Ken; M. A. (1988). Characterization of the IgA receptor from human polymorphonuclear leukocytes Immunology 64,205-210. Alcaraz, G., Kinet, J-P., Kumar, N., Wank, S. A., and Metzger, H. (1984).Phase separation of the receptor for immunoglobulin E and its subunits in Triton X-114.1. B i d . Chem. 259, 14922-14927. Ali, H., Cuhno-Melo, J. R., and Beaven, M. A. (1989). Receptor-mediated release of inositol 1,4,5-triphosphate and inositol 1,4-biphosphate in rat basophilic leukemia RBL-2H3 cells permeabilized with streptolysis 0. Biochem. Biophys. Acta 1010, 88-89. Allen, J. M., and Seed, B. (1989). Isolation and expresison of functional high afhity Fc receptor complementary DNAs. Science 243,378-381. Amigorena, S., Bonnerot, C., Choquet, D., Fridman, W. H., and Teillaud, J. L.(1989). FcyRII expression in resting and activated B lymphocytes. Eur. 3. Immunol. 19, 1379-1385. Amigorena, S., Bonnerot, C., Drake, J. R., Choquet, D., Huniziker, W., Guillet, J-G., Webster, P., Sautes, C. Mellman, I., and Fridman, W. H. (1992a).Cytoplasmic domain heterogeneity and functions of IgG Fc receptors in B lymphocytes. Science 256, 1808-1812. Amigorena, S., Salamero, J., Davoust, J., and Fridman, W. H. (1992b). Tyrosinecontaining motif that transduces cell activation signals also determines internalization and antigen presentation via type 111receptors for IgG. Nature (London)358,337-341. Amin, A. R., Coico, R. F., and Thorbecke, G. J. (1990).Characterisation and regulation of T cell IgD receptors and IgD binding factors. Res. Immunol. 141, 94-99. Amin, A. R., Tamma, S. M. L., Oppenheim, J. D., Finkelman, F. D., Kieda, C., Coico, R. F., and Thorbecke, G. J. (1991). Specificity of the murine IgD receptor on T cells is for N-linked glycans on IgD molecules. Proc. Natl. Acad. Sci. U.SA. 88,9238-9242. Amon, U., von-Stebut, E., Dietz, K. R., Bauer, F. W., and Wolff, H. H. (1992). Pharmacological studies on the role of protein kinase C in signal transduction of human basophils. Znt. Arch. Allergy Immunol. 98,349-354. Anasetti, C., Martin, P. J., Morishita, Y., Badger, C. C., Bernstein, I. D., and Hansen, J. A. (1987). Human large granular lymphocytes express high affinity receptors for murine monoclonal antibodies of the IgG3 subclass. J. Immunol. 138,2979-2981. Anderson, B., Skoglund, A., Ronnholm, M., Lidsten, T.,Lamon, E. W., Collisson, E. W., and Wdia, A. S. (1981). Functional aspects of IgM and IgG Fc receptors on murine T lymphocytes. Immunol. Rev. 56,5-50. Anderson, C. L. (1982). Isolation of the receptor for IgC from a human monocyte cell line (U937) and from human peripheral b1oodmonocytes.J.Exp. Med. 156,1794-1806. Anderson, C. L.,and Abraham, G. N. (1980). Characterisation of the Fc receptor for IgG on a human macrophage cell line U937. J . Immunol. 125,2735-2741. Anderson, C. L., and Looney, R.J. (1986).Human leukocyte IgG Fc receptors. Immunol. Today 7,264-266.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

93

Anderson, C. L., Guyre, P. M., Whitin, J. C., Ryan, D. H., Looney, R. J., and Fanger, M. W. (1986). Monoclonal antibodies to Fc receptors for IgG on human mononuclear phagocytes: Antibody characterization and induction of superoxide production in a monocyte cell. J. Biol. Chem. 261, 12856-12864. Anderson, C. L., Ryan, D. H., Looney, R. J., and Leary, P. C. (1987).Structural polymorphism of the human monocyte 40 kilodalton Fc receptor for IgG. J . Immunol. 138, 2254-2256. Anderson, C. L., Shen, L., Eicher, D. M., Wewers, M. D., and Gill, J. K. (1990a). Phagocytosis mediated by three distinct Fcy receptor classes on human leukocytes. J. Exp. Med. 171, 1333-1345. Anderson, C. L., Looney, R. J., Culp, D. J., Ryan, D. H., Fleit, H. B., Utell, M. J., Frampton, M. W., Manganiello, P. D., and Guyre, P. M. (1990b). Human aveolar and peritoneal macrophages bear three distinct classes of Fc receptors for 1gG.J. Immunol. 145,196-201. Anderson, P., Caligiuri, M., O’Brien, C., Manley, T., Ritz, J., and Schlossman, S. F. (1990~).FcR receptor type I11 (CD16) is included in the 5 NK receptor complex expressed by human natural killer cells. Proc. Natl. Acad. Sci. U . S A 87,2274-2278. Anderson, G. P., and Anderson, C. L. (1990). Signal transduction by the platelet Fc receptor. Blood 76, 1165-1172. Angelisova, P. C., Vleck, I., Stefanova, M., Lipoldova, M., and Horejsi, V. (1990). The human leukocyte surface antigen CD53 is a protein structurally similar to the CD37 and MRC OX-44 antigens. lmmunogenetics 32,281-285. Anselmino, L. M., Perussia, B., and Thomas, L. L. (1989).Human basophils selectively express the FcyRII (CDw32) subtype of IgG receptor. Allergy Clin. Immunol. 84, 907-914. Apgar, J. R. (1990).Antigen-induced cross-linking of the IgE receptor leads to an association with the detergent-insoluble membrane skeleton of rat basophilic leukemia (RBL2H3) cells. J. Immunol. 145,3814-3822. Apodaca, G., Bomesel, M., Arden, J., Breitfeld, P. P., Tang, K., and Mostov, K. E. (1991). The polymeric immunoglobulin receptor: A model protein to study transcytosis J. Clin. Inuest. 87, 1877-1882. Astier, A., de la Salle, H., de la Salle, C., Bieber, T., Esposito-Farese, M-E., Freund, M., Cazenave, J. P., Fridman, W. H., Teillard, J-L., and Hanau, D. (1994). Human epidermal Langerhans cells secrete a soluble receptor for IgG (FcyRIIICD32) that inhibits the binding of immune complexes to FcyR + ce1ls.J. Immunol. 152,201-212. Atkinson, T. P., Kaliner, M. A., and Hohman, R. J. (1992).Phospholipase C-yl is translocated to the membrane of rat basophilic leukemia cells in response to aggregation of IgE receptors. J. Immunol. 148,2194-2200. Aubry, J-P., Pochon, S., Graber, P., Jansen, K. U., and Bonnefoy, J-Y. (1992). CD21 is a ligand for CD23 and regulates IgE production. Nature (London)358, 505-507. Azzoni, L., Kamoun, M., Salcedo, T. W., Kanakaraj, P., and Perussia, B. (1993). Stimulation of Fc gamma RIIIA results in phosopholipase C-gamma1tyrosine phosphorylation and p56(lck) activation. J . E r p . Med. 176, 1745-1750. Baniyash, M., and Eshhar, Z. (1984).Inhibition ofIgE bindingtomast cells and basophils by monoclonal antibodies to murine IgE. Eur. J . Immunol. 14,799-807. Baniyash, M., Kehry, M., and Eshhar, Z. (1988).Anti-IgE monoclonal antibodies directed at the FCEreceptor binding site. Mol. Immunol. 25, 705-711. Banting, G., Brake, B., Bragetta, P., Luzio, J. P., and Stanley, K. K. (1989). Intracellular targeting signals of polymeric immunoglobulin receptors are highly conserved between species. FEBS Lett. 254,177-183.

94

MARK D. HULETT AND P. MARKHOGARTH

Barnett-Foster, D. E., Dorrington, K. J., and Painter, R. H. (1978).Structure and function of immunoglobulin domains. VII. Studies on the structural requirements of human immunoglobulin G for granulocyte binding. J . Immunol. 120, 1952-1956. Barnett-Foster, D. E.,Sjoquist, J., and Painter, R. H.(1982). The effect of fragment B of staphloccocal protein A on the binding of rabbit IgG to human granulocytes and monocytes. Mol. Immunol. 19,407-412. Barrett, T. B., Shu, G. L.,Draves, K. E., Pezzutto, A., and Clark, E. A. (1990). Signaling through CD19, Fc receptors or transforming growth factor-beta: Each inhibits the activation of resting human B cells differently. Eur. J. Immunol. 30, 1053-1059. Basciano, L. K., Berenstein, E. H., Kmak, L.,and Siraganian, R. P. (1986). Monoclonal antibodies that inhibit IgE binding. 1. Eiol. Chem. 261, 11823-11831. Basu, M., Hakimi, J., Dharm, E., Kondas, J. A,, Tsien, W-E., Pilson, R. S., Lin, P., Gilfillan, A., Haring, P., Braswell, E. H., Nettleton, M. Y., and Kochan, J. P. (1993). Purification and characterization of human recombinant IgE-Fc fragments that bind to the human high f i n i t y IgE receptor. J. Eiol. Chem. 268, 13118-13127. Beaven, M. A,, and Cunha-Melo, J. R. (1988). Membrane phosphoinositide-activated signals in mast cells and basophils. h o g . Allergy 42, 123-184. Beaven, M. A., and Metzger, H. (1993).Signal transduction by Fc receptors: The FcERI case. Immunol. Today 14,222-225. Beavil, A. J., Edmeades, R. K., Gould, H. J., and Sutton, B. J. (1992).a-Helical coiledcoil stalks in the low-affinity receptor for IgE (FcsRIUCD23) and related C-type lectins. Proc. Natl. Acad. Sci. U.S.A. 89,753-757. Becker, K. E., Ishizaka, T., Metzger, H., Ishizaka, K., and Grimley, P. M. (1973).Surface IgE on human basophils during histamine release. J . Exp. Med. 138, 394-409. Benhamou, M., and Siraganian, R. P. (1992).Protein-tyrosine phosphorylation: An essential component of FcsRI signaling. Immunol. Today 13, 195-197. Benhamou, M., Bonnerot, C., Fridman, W. H., and Daeron, M. (1990a). Molecular heterogeneity of murine mast cell Fc gamma receptors. J . Immunol. 144,3071-3077. Benhamou, M., Gutkind, J. S., Robbins, K. C., and Siraganian, R. P. (1990b). Tyrosine phosphorylation coupled to IgE receptor signal transduction and histamine release. Proc. Natl. Acad. Sci. U.S.A.87, 5327-5330. Bennech, P. D., Sastry, K., Iyer, R. R., Eichbaum, Q. G., Raveh, D. P., and Ezekowitz, A. S. (1992). Definition of interferon y-response elements in a novel human Fcy receptor gene (FcyRIb) and characterisation ofthe gene.J. Exxp. Med. 176,1115-1123. Bennich, H., Ragnarsson, U., Johansson, S. G. O., Ishizaka, K., Ishizaka, T., Levy, D. A., and Lichtenstein, L. M. (1977). Failure of the putative IgE pentapeptide to compete with IgE for receptors on basophils and mast cells. Int. Archs. Allergy AppZ. Immunol. 53,459-468. Berenstein, E. H., Koh, C. Y.,Fischler, C., and Siraganian, R. P. (1990). Monoclonal antibodies to the beta subunit of the high affinity FCERdo not release histamine. FASEB]. 4, 1874. Bergman, M., Mustelin, T., Oeken, C., Partanen, J., Flint, N. A., Amrein, K. E., Autero, M., Burn, P., and Alitalo, K. (1992). The human p50esktyrosine kinase phosphorylates p56ICkat Tyr-505 and down regulates its catalytic activity. EMBO. J . 11,2919-2924. Berridge, M. J. (1987).Inositol triphosphate and diacylglycerol: Two interacting second messengers. Annu. Rev. Biochem. 56, 159-193. Bieber, T., de la Selle, H., Woolomberg, A., Hakimi, J., Chizzonite, R., Ring, J., Hanau, D., and de la Selle, C. (1992). Human epithelial Langerhans cells express the high affinity receptor for immunoglobulin E (FcsRI).J. E r p . Med. 175, 1285-1290. Blank, U., Ra, C., Miller, L., White, K., Metzger, H., and Kinet, J-P. (1989). Complete

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

95

structure and expression in transfected cells of high affinity IgE receptor. Nature (London)337,187-189. Blank, U., Ra, C., and Kinet, J-P. (1990).Characterization of truncated a chain products from human, rat and mouse high affinity receptor for immunoglobulin E. J . Biol. Chem. 266,2639-2646. Boguski, M. S., Bairoch, A., Attwood, T. K., and Michaels, G. S. (1992). Proto-vav and gene expression. Nature (London)358, 113. Bonnerot, C., Amigorena, S., Choquet, D., Pavlovich, K., Choukroun, V., and Fridman, W. H. (1992). Role of associated y-chain in tyrosine kinase activation via murine F v R I I I . EMBO J . 11,2747-2757. Boros, P., Odin, J. A., Muryoi, T., Masur, S. K., Bona, C., and Unkeless, J. C. (1991). IgM anti-FcyR autoantibodies trigger neutrophil degranulation. J . Exp. Med. 173, 1473-1482. Braakman, E., van de Winkel, J. G. J., van Krimpen, B. A., Jansze, M., and Bolhuis, R. L. H. (1993). CD16 on human y8 T lymphocytes: Expression, function and specificity for mouse IgG-isotypes. Cell. Immunol. 143,97-107. Brandtzaeg, P. (1985). Role of J chain and secretory component in receptor mediated glandular and hepatic transport of immunoglobulins in man. Scand. J . Immunol. 22, 111-146. Breitfeld, P. P., Casanova, J. E., McKinnon, W. C., and Mostov, K. E. (1990). Deletions in the cytoplasmic domain of the polymeric immunoglobulin receptor differentially effects endocytosis rate and postendocytotic traffic.J. Biol.Chem. 265, 13750-13757. Bronza, J. P., Hauff, N. F., Philips, W. A., and Johnston, R. B., Jr. (1988). Activation of the respiratory burst in macrophages: Phosphorylation specifically associated with Fc receptor-mediated stimulation. J . Immunol. 141, 1642-1647. Brooks, D. G., Qui, W. Q., Luster, A. D., and Ravetch, J. V. (1989). Structure and expression of a human IgG FcRII (CD32): Functional heterogeneity is encoded by the alternatively spliced products of multiple genes. J . Exp. Med. 170, 1369-1385. Brown, E. J., Newell, A. M., and Gresham, J. H. D. (1987). Molecular regulation of phagocytic function: Evidence for involvement of a guanine tri-phosphate binding protein in opsonin-mediated phagocytosis by monocytes. J . Immunol. 139,3777-3782. Buckle, A., and Hoss, N. (1989). The effect of IFN-y and colony stimulating factors on the expression of neutrophil cell membrane receptors. J . Immunol. 143,2295-2301. Burd, P. R., Rodgers, H. W., Gorden, J. R., Martin, C. A., Jayaraman, S., Wilson, S. D., Dvorak, A. M., Galli, S . J., and Dorf, M. E. (1989). Interleukin 3-dependent and -independent mast cells stimulated with IgE and antigen express multiple cytokines. J . E x p . Med. 170,245-257. Burns, G. F., Cawley, J. C., and Barker, C. R. (1979). Characterization of the receptor for IgM present on human B lymphocytes. Immunology 36,569-577. Burt, D. S., and Stanworth, D. (1987). Inhibition of binding of rat IgE to rat mast cells by synthetic IgE peptides. Eur. J . Immunol. 17,437-440. Burt, D. S., Hastings, G. Z., Healy, J., and Stanworth, D. R. (1987). Analysis of the interaction between rat immunoglobulin E and rat mast cells using anti-peptide antibodies. Mol. Immunol. 24,379-389. Burton, D. R., and Woof, J. M. (1992).Human antibody effector function. Adv. Immunol. 51, 1-84. Burton, D. R., Jefferis, R., Partridge, L. J., and Woof, J. M. (1988).Molecular recognition of antibody by cellular Fc receptor (FcRI). Mol. Immunol. 25, 1175-1181. Cambier, J. C., Pleiman, C., and Clark, M. R. (1994). Signal transduction by the B cell antigen receptor and its co-receptors. Annu. Rev. Immunol. 12,457-486.

96

MARK D. HULETT AND P. MARKHOGARTH

Canfield, S. M., and Morrison, S. L. (1991). The binding a n i t y of human IgC for its high affinity Fc receptor is determined by multiple amino acids in the CH2 domain and is modulated by the hinge region.]. Exp. Med. 173, 1483-1491. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A., Kopeller, R., and Soltoff, S. (1991). Oncogenes and signal transduction. Cell (Cambridge, Mass.) 64,281-302. Capron, A., and Dessaint, J. P. (1985).Effector and regulatory mechanisms in immunity to schistosomes: A heuristic view. Annu. Rev. Immunol. 3,455-476. Carson, D. A., and Metzger, H. (1974). Interaction of IgE with rat basophilic leukemia cells. IV. Antibody-induced redistribution of IgE receptors. J. Imrnunol. 113, 1271-1277. Casanova, J. E., Apodaca, G.,and Mostov, K. E. (1991). An autonomous signal for basolateral sorting in the cytoplasmic domain of the polymeric immunoglobulin receptor. Cell (Cambridge, Mass.) 66, 65-75. Cassatella, M. A., Anegnon, I., Cuturi, M. C., Griskey, P., Trinchieri, G., and Perussia, B. (1989). FcyR (CD16) interaction with ligand induces Ca2+mobilization and phosphoinositide turnover in human natural killer cells. J . Exp. Med. 169, 549567. Cassel, D. L., Keller, M. A., Surrey, S., Scwartz, E., Schreiber, A. D., Rappaport, E. F., and McKenzie, S. E. (1993).Differential expression of FcyRlIA, FcyRlIB and FcyRIIC in hematopoeitic cells: Analysis of transcripts. Mol. Immunol. 30, 451-460. Ceuppens, J. L., and van Vaecjk, F. (1989). Human T cell activation induced by a monoclonal mouse IgG3 anti-CD3 antibody (Riv9) requires binding of the Fc part of the antibody to the monocytic 72-kDa high affinity Fc receptor (FcRl).Cell. Immunol. 118,136-146. Ceuppens, J. L., Baroja, M. L., Van Vaecjk, F., and Anderson, C. L. (1988). A defect in the membrane expression of high affinity 72 kD Fc receptors on phagocytic cells in four healthy subjects. J . Clin. Invest. 82,571-578. Chevailler, A., Monteiro, R. C., Kubagawa, H., and Cooper, M. D. (1989). Immunofluorescence analysis of IgA binding by human mononuclear cells in blood and lymphoid tissue,./. Immunol. 142,2244-2249. Ciccimarra, F., Rosen, F. S., and Merler, E. (1975).Localization of the IgG effrector site for monocyte receptors. Proc. Natl. Acad. Sci. U.S.A. 72,2081-2083. Clark, M. R., Clarkson, S. B., Ory, P. A,, Stollman, N., and Goldstein, I. M. (1989). Molecular basis for a polymorphism involving Fc receptor I1 on human monocytes. J. Immunol. 143,1731-1734. Clark, M. R., Liu, L., Clarkson, S. B., Ory, P. A., and Goldstein, 1. M. (1990). An abnormality of the gene that encodes neutrophil Fc receptor 111 in a patient with systemic lupus erythematosus. J . Clin. Inoest. 86,341-346. Clark, M. R., Campbell, K. S., Kazlauskaus, A., Johnson, S. A., Hertz, M., Potter, T. A., Pleiman, C., and Cambier, J. C. (1992).The B cell antigen receptor complex: Association of Iga and Igp with distinct cytoplasmic effectors. Science 258, 123-126. Clark, M. R., Johnson, S. A., and Cambier, J. C. (1994). Analysis of Ig a-tyrosine kinase interaction reveals two levels of binding specificity and tyrosine phosphorylated Iga stimulation of FcyR activity. (1994) E M B O J . 13, 1911-1919. Classon, B. J., Williams, A. F., Willus, A. C., Seed, B., and Stamenkovic, I. (1989). The primary structure of human leukocyte antigen CD37, a species homologue of the rat MRC OX44 antigen. J. Erp. Med. 169,1497-1502. Cockcroft, S., and Gompeks, B. D. (1985). Role of guanine nucleotide binding protein in the activation of polyphosphoinositide phosphodiesterase. Nature (London) 314, 534-536.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

97

Cohen, L., Sharp, S., and Kulczycki, A., Jr. (1983). Human monocytes, B lymphocytes and non-B lymphocytes each have structurally unique Fc receptors. J. Immunol. 131, 378-383. Coico, R. F., Xue, B., Wallace, D., Pernis, B., Siskind, G. W., and Thorbecke, G. J. (1985).T cells with receptors for IgD. Nature (London) 316,744-746. Coico, R. F., Berzofsky, J. A., York-Jolley, J., Ozaki, S., Siskind, G. W., and Thorbecke, G. J. (1987). Physiology of IgD. VII. Induction of receptors for IgD on cloned T cells by IgD and interleukin 2. J . Immunol. 138,4-6. Coico, R. F., Finkelman, F., Swenson, C. D., Siskind, G. W., and Thorbecke, G. J. (1988). Exposure to cross-linked IgD induces receptors for IgD on T cells in vivo and in vitro. Proc. Natl. Acad. Sci. U S A . 85, 559-563. Coico, R. F., Tamma, S. L., Bessler, M., Wei, C-F., and Thorbecke, G. J. (1990). IgDreceptor positive human T lymphocytes. I. Modulation of receptor expression by oligomeric IgD and lymphokines. J. Immunol. 145,3556-3561. Coleman, J. W., Helm, B. A., Stanworth, D. R., and Gould, H. J. (1985). Inhibition of mast cell sensitization in oitro by a human immunoglobulin &-chain fragment synthesized in Escherichia coli. Eur. J . Immunol. 1985,966-969. Conrad, D. H. (1990). FcsRII/CD23: The low affinity receptor for IgE. Annu. Reo. Immunol. 8,623-645. Conrad, D. H., and Froese, A. (1976). A characterization of the target cell receptor for IgE: 11-D polyacrylamide gel analysis of the surface IgE receptor from normal rat mast cells and from rat basophilic leukemia cells. J . Immunol. 116, 319-326. Conrad, D. H., Wingard, J. R., and Ishizaka, T. (1983).The interaction of human and rodent IgE with the human basophil IgE receptor. J. Immunol. 130,327-333. Corcia, A., Schweitzer-Stenner, R., Pecht, I., and Rivnay, B. (1986).Characterization of the ion channel activity in planar bilayers containing IgE-Fc epsilon receptor and the cromoly-binding protein. EMBO J. 5,849-854. Cosio, F. G., Ackerman, S. K.,Douglas, S. D., Friend, P. S., and Michael, A. F. (1981). Soluble immune complexes binding to human monocytes and polymorphonuclear leukocytes. Immunology 44,773-780. Crockett-Torabi, E., and Fantone, J. C. (1990).Soluble and insoluble immune complexes activate human neutrophil NADPH oxidase by distinct Fcy receptor-specific mechanisms. J. Immunol. 145,3026-3032. Curnow, S . J., Glenne, M. J., and Stevenson, G. T. (1992). Cytoplasmic calcium fluxes induced in cytotoxic effector cells by engagement of Fcy receptors I, I1 and 111. Scand. J. Immunol. 36,221-231. Daeron, M., Prouvost-Danon, A., and Voisin, G. (1980). Mast cell antigens and Fc receptors in anaphylaxis. 11. Functionally distinct receptors of IgG and IgE on mouse mast cells. Cell. Immunol. 49, 178-189. Daeron, M., Malbec, O., Latour, S., Bonnerto, C., Segal, D. M., and Fridman, W. H. (1993).Distinct intracytoplasmic sequences are required for endocytosis and phagocytosis via murine FcyRII in mast cells. Int. Immunol. 5, 1393-1401. Daeron, M., Malbec, O., Bonnerot, C., Latour, S., Segal, D. M., and Fridman, W. H. (1994). Tyrosine-containing activation motif-dependent phagocytosis in mast cells. J . Immunol. 152, 1-10. Davidson, M. K., Le Beau, M. M., Eddy, R. L., Shows, T. B., DiPietro, L. A., Kingzette, M., and Hanly, W. C. (1988).Genetic mapping ofthe human polymeric immunoglobulin receptor gene to chromosome region lq31-41. Cytogenet. Cell Genet. 51,107-111. Davidson, W. F., Morse 111, H. C., Mathieson, B. J., Kozak, C. A., and Shen, F. W. (1983). The B-cell alloantigen Ly17.1 is controlled by a gene closely linked to Ly20 and Ly9 on chromosome 1. lmmunogenetics 17,325-329.

98

MARK D. HULETT AND P. MARKHOGARTH

Davies, D. R., and Metzger, H. (1983). Structural basis of antibody function. Annu. Reo. Immunol. 1, 87. Deanin, C . G., Martinez, A. M., Pfieffer, J. R., Gardiner, M. E., and Oliver, J. M. (1991a). Tyrosine kinase-dependent phosphatidylinositol turnover and functional responses in the FcERI signalling pathways. Biochem. Biophys. Res. Commun. 179, 551-557. Deanin, G. G., Pfieffer, J. R., Cutts, J. L., Fore, M. F., and Oliver, J. M. (1991b). Isoprenoid pathway activity is required for IgE receptor-mediated tyrosine kinasecoupled transmembrane signalling in permeabilized RBL-2H3 rat basophilic leukemia cells. Cell Regul. 2, 627-640. Debets, J. M. H., van de Winkel, J. G. J., Ceuppens, J. L., Dieteren, I. E. M., and Buurman, W. A. (1990).Crosslinking of both FcyRI and FcyRII induces secretion of tumor necrosis factor by human monocytes, requiring high-affinity Fc-FcyR interactions. J . Immunol. 144,1304-1310. De Haas, M., Kleijer, M., Minchinton, R. M., Roos, D., and von dem Borne, A. E. G. K. (1994). Soluble FcyRIIIa is present in plasma and is derived from natural killer cells. J . Immunol. 152,900-907. Delespesse, G., Suter, U., Mossalayi, D., Bettler, B., Sarfati, M., Hofstetter, H., Kilcherr, E., Debre, P., and Dalloul, A. (1991). Expression, structure and function of the CD23 antigen. Ado. Immunol. 49, 149-192. Delespesse, G., Sarfati, M., Wu, C. Y., Foumier, S., and Letellier, M. (1992). The low affinity receptor for IgE. Immunol. Reu. 125,77-97. De Lisi, C., and Siraganian, R. P. (1979).Receptor cross-linking and histamine release. I. The quantitative dependence of basophil degradation on the number of receptor doublets. J. Immunol. 122,2286-2292. De Wit, T. P. M., Suijkerbuijk, R. F., Capel, P. J. A., van Kessel, A. G., and van de Winkel, J. G. J. (1993). Assignment of the human high affinity Fcy receptor I genes to chromosome 1, band q21.1. Immunogenetics 38,57-59. Diamond, B., and Yelton, D. E. (1981). A new Fc receptor on mouse macrophages binding IgG3. J . E x p . Med. 153,514-519. Dickler, H. B. (1976). Lymphocyte receptors for immunoglobulin. Ado. Immunol. 24, 167-214. Dierks, S. E., Bartlett, W. C., Edmeades, R. L., Gould, H. J., Roa, M., and Conrad, D. €3. (1993). The oligomeric nature of the murine FcsRIIICD23: Implications for function. Eur. J. Immunol. 150,2372-2382. Dietzsch, E., Osman, N., McKenzie, I. F. C., Garson, 0. M., and Hogarth. P. M. (1993). The human FCGl gene encoding the high-&nity FcyRI maps to chromosome lq21. lmmunogenetics 38,307-309. Di Vergilio, F., Meyer, B. C., Greenberg, S., and Silverstein, S. C. (1988). Fc-receptor mediated phagocytosis occurs in macrophages at exceedingly low cytosolic Ca" levels. J . Cell Biol. 106,657-666. Dombrowicz, D., Flamand, V., Brigman, K. K., Koller, B. H., and Kinet J-P. (1993). Abolition of anaphylaxis by targeted disruption of the high affinity immunoglobulin E receptor a chain gene. Cell (Cambridge, Mass.) 75,969-976. Dorrington, K. J., and Bennich, H. H. (1973). Thermally induced structural changes in immunoglobulin E. J . Biol. Chem. 248, 8378-8384. Dorrington, K. J., and Bennich, H. H. (1978). Structure-function relationship in human immunoglobulin E Immunol. Reu. 41,3-25. Dougherty, G. J., Selvendran, Y., Murdoch, S., Palmer, D. G., and Hogg, N. (1987). The human mononuclear phagocyte high-anity Fc receptor, FcRI, defined by a monoclonal antibody, 10.1 Eur. J. Immunol. 17, 1453-1459.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

99

Dreskin, S . C., and Metzger, H. (1991). FcsRI-mediated hydrolysis of phosphoinositides in ghosts derived from rat basophilic leukemia cells. J. Immunol. 146, 3102-3109. Dreskin, S . C., Kuhn, D. E., and Huang, Y. (1993). Phosphoenolpyruvate and creatine phosphate augment ATP and magnesium-dependent, FcsRI-mediated activation of phospholipase C in RBL cell ghosts. J . Immunol. 151,3199-3205. Duchemin, A-M., Ernst, L. K., and Anderson, C. L. (1994).Clustering ofthe high affinity Fc receptor for IgG (FcyRI) results in phosphorylation of its associated y-chain. J. Biol. Chem. 269,12111-12117. Duncan, A. R., Woof, J. M., Partridge, L. J., Burton, D. R., and Winter, G. (1988). Localization of the binding site for the human high-affinity Fc receptor on IgG Nature (London) 332,563-564. Edberg, J . C., and Kimberly, R. P. (1994). Modulation of Fcy and complement receptor function by the glycosylphosphatidylinositol-anchoredform of FcyRIII. J. Immunol. (in press). Edberg, J. C., Redecha, P. B., Salmon, J. E., and Kimberly, R. P. (1989).Human FcyRIII (CD16): Isoforms with distinct allelic expression, extracellular domains and membrane linkages on polymorphonuclear and natural killer cells. J . Immunol. 143, 1642- 1649. Eiffert, H., Quentin, E., Decker, J., Hillmier, S., Hufschmidt, M., Klingmuller, D., Weber, M. H., and Hilschmann, H. (1984). Die Primarstruktur der menschlichen freien Sekretkomponente and die Anordung der disulfidbruken. Hoppe-Seylers Z . Physiol. Chem. 365, 1489-1495. Eiffert, H., Quentin, E., Weiderhold, M., Hillemeir, S., Decker, J., Weber, M., and Hilschmann, N. (1991). Determination of the molecular structure of the human free secretory component. Biol. Chem. Hoppe Seyler 372, 119-128. Einspahr, K. J., Abraham, R. T., Binstadt, B. A., Uehara, Y.,and Liebson, P. J. (1991). Tyrosine phosphorylation provides an early and requisite signal for the activation of natural killer cell cytotoxic function. Proc. Natl. Acad. Sci. U.S.A.88, 6279-6283. Eiseman, E., and Bolen, J. B. (1992). Engagement of the high-affinity IgE receptor activates src protien-related tyrosine kinases. Nature (London)355, 78-80. Engelhardt, W., Gorczyta, H., Buttenveck, A., Monkemann, H., and Frey, J. (1991). Structural requirements of the cytoplasmic domains of the human macrophage Fcy receptor IIa and B cell Fcy receptor IIb for the endocytosis of immune complexes. Eur. J. Immunol. 21,2227-2238. Emst, L. K., van de Winkel, J. G. J., Chiu, I-M., and Anderson, C. L. (1992). Three genes for the human high affinity Fc receptor for IgG (FcyRI) encode four distinct transcription products. J. Biol. Chem. 267, 15692-15700. Ernst, L. K., Duchemin, A-M., and Anderson, C. L. (1993). Association of the highaffinity receptor for IgG (FcyRI) with the y subunit of the IgE receptor. Proc. Natl. Acad. Sci. U.S.A. 90,6023-6027. Fanger, M. W., Goldstine, S. N., and L. Shen. (1983). Cytofluorographic analysis of receptors for IgA on human polymorphonuclear cells and monocytes and the correlation of receptor expression with phagocytosis. Mol. Immunol. 20, 1019. Fanger, M. W., Shen, L., Graziano, R. F., and Guyre, P. M. (1989).Cytotoxicity mediated by human Fc receptors for IgC. Immunol. Today 10,92-99. Farace, F., Mitjavila, M-T., Betaieb, A., Dokhelar, M. C., Wiels, J., Finale, Y., Kieffer, N., Breton-Gorius, J . , Vainchenker, W., and Tursz, T. (1988). New hematopoeitic differentiation antigens detected by anti-K562 monoclonal antibodies. Cancer Res. 48,5759-5765. Farber, D. L., and Sears, D. W. (1991). Rat CD16 is defined by a family of class 111

100

MARK D. HULETT AND P. MARKHOGARTH

Fcy receptors requiring co-expression of heteroprotein subunits. J. Immunol. 146, 4352-4361. Farber, D. L., Giorda, R., Nettleton, M. Y., TNCCO,M., Kochan, J. P., and Sears, D. W. (1993). Rat class I11 Fcy receptor isoforms differ in IgG subclass-binding specificity and fail to associate productively with rat CD35. J . Immunol. 150,4364-4375. Feister, A. J., Browder, B., Willis, H. E., Mohanakumar, T., and Ruddy, S. (1988). Pertussis toxin inhibits human neutrophil responses mediated by the 42-kilodalton IgG Fc receptor. J. Immunol. 141,228-233. Fernandez-Botran, R., and Suzuki, T. (1986). Biochemical signal transmitted by Fc receptor for immunoglobulin G2a of a murine macrophage-like cell line, P388D1: Mode of activation of adenylate cyclase mediated by immunoglobulin G2a binding proteins. Biochemistry 25,4388-4397. Ferrarini, M., Hoffman, T. S., Fu, M., Winchester, R.,and Kunkel, H. G. (1977).Receptors for IgM on certain human B cell lymphocytes. J. Immunol. 119, 1525-1529. Fewtrell, C., and Metzger, H. (1980). Larger oligomers of IgE are more effective than dimers in stimulating rat basophilic leukemia cells. J. Zmmunol. 125, 701-710. Fewtrell, C., Mohr, F. C., Ryan, T. A., and Millard, P. J. (1989). Calcium: An important second messenger in mast cells. In “IgE, Mast Cells and the Allergic Response” (D. Chadwick, D. Evered, and J. Whelen, eds.), p. 114. Wiley, Chichester, UK. Flaswinkel, H., and Reth, M. (1994). Dual role of the tyrosine activation motif of the Ig-a protein during signal transduction via the B cell antigen receptor. EMBOJ. 13, 83-89. Fleit, H. B., Wright, S. D., and Unkeless, J. C. (1982). Human neutrophil Fcy receptor distribution and structure. Proc. Natl. Acad. Sci. USA. 79, 3275-3279. Frey, J., and Engelhardt, W. (1987).Characterisation and structural analysis of Fc gamma receptors of human monocytes, a monoblast cell line (U937) and a myeloblast cell line (HL-60) by a monoclonal antibody. Eur. J. Immunol. 17,583-591. Fridman, W. H., and Sautes, C. (19W). Immunoglobulin-binding factors. I n “Fc Receptors and the Action ojAntibodies.” (H. Metzger, ed.) pp. 335-355. Am. SOC.Microbiol. Washington, DC. Fridman, W. H., Bonnerot, C., Daeron, M., Amigorena, S., Teillaud, J-L., and Sautes, C. (1992). Structural basis of Fcy receptor functions. Immunol. Reu. 125,49-76. Fridman, W. H., Teillaud, J-L., Bouchard, C., Teillaud, C., Astier, A., Tartour, E., Galon, J., Mathiot, C., and Sautes, C. (1993).Soluble Fcy recept0rs.J. Leuk. Biol. 54,504-512. Fries, L. F., Hall, R. P., Lawley, T. J., Crabtree, G. R., and Frank, M. M. (1982). Monocyte receptors for the Fc portion of IgG studied with monomeric human IgG1: Normal in vitro expression of Fcy receptors in HLA-BNDrw 3 subjects with defective Fcmediated in vivo clearance. J . Immunol. 129, 1041-1049. Frigeri, L. G., Zuberi, R. I., and Liu, F. T. (1993). sBP, a p-galactoside binding animal lectin, recognises IgE receptor ( FcERI) and activates mast cells. Biochemistry 32, 7644-7649. Fukamachi, H., Kawakami, Y., Takei, M., Ishizaka, T., Ishizaka, K., and Kawakami, T. (1992).Associationof protein tyrosine kinase with phospholipase Cyl in bone marrow derived mouse mast cells. Proc. Natl. Acad. Sci. U.S.A.89,9524-9528. Fukamachi, H., Takei, M., and Kawakami, F. (1993). Activation of multiple protein kinases including a MAP kinase upon FcERI cross-linking. Int. Arch. Allergy Immunol. 102, 15-25. Fukamachi, H., Yamada, N., Miura, Kator, T., Ishikawa, M., Gulbins, E., Altman, A., Kawakami, Y., and Kawakami, T. (1994). Identification of a protein, SPY75, with repetitive helix-turn-helix motifs and an SH3 domain as a major substrate for protein tyrosine kinase(s) activated by FcERI cross-Iinking.J. Immunol. 152,642-652.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

101

Gardiner, P. (1989). Calcium and T cell activation. Cell (Cambridge,Mass.) 59,15-20. Gastinel, L. N., Simister, N. E., and Bjorkman, P. J. (1992).Expression and crystallization of a soluble and functional form of an Fc receptor related to class I histocompatibility molecules. Proc. Natl. Acad. Sci. U.S.A.89, 638-642. Gauldie, J., Richards, C., and Lamontagne, L. (1983). Fc receptors for IgA and other immunoglobulins on resident and activated macrophages. Mol. Immunol. 20, 1029. Gerwirtz, A. M., Boghasian-Sell, F. L., Catani, L., Ratajczak, M., Shen, Y. M., and Schreiber, A. D. (1992). Expression of FcsRII and CD4 receptors by normal human megakaryocytes. Erp. Hematol. 20,512-516. Ghazizadeh, S., and Fleit, H. B. (1994).Tyrosine phosphorylation provides an obligatory early signal for FcyRII-mediated endocytosis in the monocytic cell line THP-1. J . Immunol. 152, 30-41. Ghazizadeh, S., Bolen, J. B., and Fleit, H. B. (1994).Physical and functional association of src-related protein tyrosine kinases with FcyRII in monocytic THP-1 cells. J . Biol. Chem. 269,8878-8884. Gold, M. R., and Defranco, A. L. (1994).Biochemistry ofB cell activation.Adu. Imrnunol. 55,221-295. Gomez-Guerrero, C., Gonzalez, E., and Egido, J. (1993). Evidence for a specific IgA receptor in rat and human mesangial cells. J . Immunol. 151, 7172-7181. Gordon, J., Flores-Romo, L., Cairns, J. A., Millsum, M. J., Lane, P. J. Johnson, G. D., and MacLennan, I. C. M. (1989). CD23: A multi-functional receptorllymphokine? Immunol. Today 10,153-157. Gosselin, E. J., Brown, M. F., Anderson, C. L., Zipf, T. F., and Guyre, P. M. (1990). The monoclonal antibody 41H16 detects the Leu4 responder phenotype of human FcyRII. J . Immunol. 144,1817-1822. Goulding, N. J., and Guyre, P. M. (1992).Impairment of neutrophil Fc gamma receptor mediated transmembrane signalling in active rheumatoid arthritis. Ann. Rheum. Dis. 51,594-599. Graziano, R. F., and Fanger, M. W. (1987).FcyRI and FcyRII on monocytes and granulocytes are cytotoxic trigger molecules for tumor cells. J . Imrnunol. 139,3536-3541. Graziano, R. F., Looney, R. J., Shen, L., and Fanger, M. W. (1989). FcyR-mediated killing by eosinophils. J. Zmmunol. 142,230-235. Greenberg, S., Chang, P., and Silverstein, S. C. (1993). Tyrosine phophorylation is required for Fc receptor-mediated phagocytosis in mouse macrophages.J . Exp. Med. 177,529-534. Greenman, J., Tutt, A. L., George, A. J. T., Pulford, K. A. F., Stevenson, G. T., and Glennie, M. J. (1991). Characterisation of a new monoclonal antibody, ATIO, and its incorporation into a bispecific F(ab')2deriviative for recruitment of cytotoxic effectors. Mol. Immunol. 28,1243-1254. Gruchalla, R. S., Dinh, T. T., and Kennedy, D. A. (1990). An indirect pathway of receptor-mediated 1,2-diacylglycerolformationin mast cells. I. IgE receptor-mediated activation of phospholipase D. J. Biol. Chem. 144,2334-2342. Grundy, H. O., Peltz, G., Moore, K. W., Golbus, M. S., Jackson, J. G., and Lebo, R. V. (1989). The polymorphic Fc receptor I1 gene maps to human chromosome lq. Immunogenetics 29,331-339. Gupta, S., Platsoucas, C. D., Schulof, R. S., and R. A. Good. (1979). Receptors for IgA on a subpopulation of human T and B lymphocytes. Cell. Immunol. 45,469-470. Guse, A. H., Roth, E., Broker, B.M., and Emmrich, F. (1992).Complex inositol polyphosphate response induced by co-crosslinking of CD4 and Fcy receptors in the human monocytoid cell line U937. J. Immunol. 149,2452-2458. Guyre, P. M., Morganelli, P. M., and Miller, R. (1983).Recombinant immune interferon

102

MARK D. HULETT AND P. MARKHOGARTH

increases immunoglobulin G Fc receptors on cultured human mononuclear phagocytes. J. Clin. Invest. 72, 393-397. Guyre, P. M., Graziano, R. F., Vance, B. A., Morganelli, P. M., and Fanger, M. W. (1989). Monoclonal antibodies that bind to distinct epitopes on FcyRI are able to trigger receptor function. J . Immunol. 143, 1650-1655. Haeffner-Cavaillon, N., Klein, M., and Dorrington, K. J. (1979a). Studies on the Fey receptor of the murine macrophage-like cell line P388D1. I. The binding of homologous and heterologous immunoglobulin G. J. Immunol. 123, 1905-1913. Haeffner-Cavaillon, N., Dorrington, K. J., and Klein, M. (1979b). Studies on the Fcy receptor of the murine macrophage-like cell line P388D1.11. Binding of human IgG subclass proteins and their proteolytic fragments. J. Immunol. 123, 1914-1919. Hakimi, J., Seals, C., Kondas, J. A., Pettine, L., Danho, W., and Kochan, J. (1990). The a subunit of the human IgE receotor (FcsRI) is sufficient for high affinity IgE binding. J. Biol. Chem. 265,22079-22081. Hamada, F., Aoki, M., Akiyama, T., and Toyoshima, K. (1993). Association of immunoglobulin G Fc receptor I1 with Src-like protein-tyrosine kinase FgR in neutrophils. Proc. Natl. Acad. Sci. USA. 90,6305-6309. Hamawy, M. M., Oliver, C., and Siraganian, R. P. (1992). Inhibition of IgE binding to RBL-2H3 cells by a monoclonal antibody (BD6) to a surface protein other than the high affinity IgE receptor. J. Immunol. 148,524-531. Hamawy, M. M., Mergenhagen, S. E., and Siraganian, R. P. (1993).Tyrosine phosphorylation of ~ ~ 1 by 2the 5aggregation ~ ~ ~of high affinity immunoglobulin E receptors requires cell adherence. ]. Biol. Chem. 268, 6851-6859. Hamburger, R. (1975). Peptide inhibition of the Prausnitz-Kustner reaction. Science 189,389-390. Harrison, D., Philips, J. H., and Lanier, L. L. (1991). Involvement of a metalloprotease in spontaneous and phorbol ester-induced release of natural killer cell-associated FcyRIII (CD16-11).J. Immunol. 147, 3459-3465. Hartnell, A., Kay, A. B., and Wardlaw, A. J. (1992). IFN-y induces expression of FcyRIII (CD16) on human eosinophils. J. Immunol 148, 1471-1478. Helm, B. A., Marsh, P., Vercelli, D., Padlan, E., Gould, H., and Geha, R. (1988). The mast cell binding site on human immunoglobulin E. Nuture (London) 331, 180-183. Helm, B. A., Kebo, D., Vercelli, D., Glovsky, M. M., Gould, H., Ishizaka, K., Geha, R., and Ishizaka, T. (1989). Blocking of passive sensitization of human mast cells and basophil granulocytes with IgE antibodies by a recombinant human e-chain fragment of 76 amino acids. Proc. Natl. Acad. Sci. U.S.A.86,9465-9469. Helm, B. A,, Ling, Y.,Teale, C., Padlan, E. A,, and Bruggemann, N. (1991). The nature ofthe importance ofthe inter-e chain disulphide bonds in human IgE. Eur.].Immunol. 21,1543-1548. Hemmerich, S., and Pecht, I. (1988).Isolation and purification of an Fc epsilon receptor activated ion channel from the rat mast cell line RBL-2H3. Biochemistry 27, 7488-7498. Heuser, C. H., Anderson, C. L., and Grey, H. M. (1977). Receptors for IgG: Subclass specificity of receptors on different mouse cell types and the definition of two distinct receptors on a macrophage cell line. ]. E x p . Med. 145, 1316-1327. Hibbs, M. L., Hogarth, P. M., and McKenzie, I. F. C. (1985). The mouse Ly17 locus identifies a polymophism of the Fc receptor. Immunogenetics 22,335-348. Hibbs, M. L., Walker, I. D., Kirszbaum, L., Pietersz, G. A,, Deacon, N. J., Chambers, G. W., McKenzie, I. F. C., and Hogarth, P. M. (1986). The murine Fc receptor for immunoglobulin: Purification, partial amino acid sequence and isolation of cDNA clones. Proc. Natl. Acad. Sci. U S A . 83,6980-6984.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

103

Hibbs, M. L., Bonadonna, L., Scott, B. M., McKenzie, I. F. C., and Hogarth, P. M. (1988).Molecular cloning of a human immunoglobulin G receptor. Proc. Natl. Acad. Sci. U.S.A.85, 2240-2244. Hibbs, M. L., Sevaraj, P., Carpen, O., Springer, T. A., Kuster, H., Jouvin, M-H., and Kinet, J-P. (1989). Mechanisms for regulating expression of membrane isoforms of FcyRIII (CD16). Science 246, 1608-1611. Hirata, Y.,and Suzuki, T. (1987). Protein kinase activity associated with Fcy2a receptor of a murine macrophage cell line, P388D1. Biochemistry 26,8189-8195. Hirt, R. P., Hughes, G. J., Frutiger, S., Michetti, P., Perregaux, C., Poulain-Godefroy, O., Jeanuenat, N., Neutra, M. R., and Kraehenbuhl, J. P. (1993). Transcytosis of the polymeric Ig receptor requires phosphorylation of serine 664 in the absence but not the presence of dimeric IgA. Cell (Cambridge, Mass.) 74,245-255. Hobbs, S. M., and Jackson, L. E. (1985). Binding of subclasses of rat immunoglobulin G to detergent isolated FcR from neonatal rat intestine.]. Biol. Chem.262,8041-8046. Hogarth, P. M., Hibbs, M. L., Bonadonna, L., Scott, B. M., Witort, E., Pietersz, G. A,, and McKenzie, I. F. C. (1987). The mouse Fc receptor for IgG (Ly-17): Molecular cloning and specificity. lmmunogenetics 26, 161-168. Hogarth, P. M., Witort, E., Hulett, M. D., Bonnerot, C., Even, J., Fridman, W. H., and McKenzie, I. F. C. (1991). Structure of the mouse PFcy receptor I1 gene.]. Zmmunol. 146,369-376. Hogarth, P. M., Hulett, M. D., Ierino, F. L., Tate, B., Powell, M. S., and Brinkworth, R. I. (1992). Identification of the immunoglobulin binding regions (IBR) of FcyRII and FcERI. Immunol. Rev. 125,21-35. Hohman, R. J. (1988). Aggregation of IgE receptors induces degranulation in rat basophilic leukemia cells permeabilized with alpha-toxin from Staphylococcus aureus. P ~ o c .Natl. A d . S C ~U.S.A. . 85, 1624-1628. Holmes, K. L., Palfree, R. G. E., Hammerling, U., and Morse, H. C. (1985).Alleles of Ly17 alloantigen define polymorphism of the murine IgG Fc receptor. Proc. Nutl. Acad. Sci. USA. 82,7706-7710. Holowka, D., and Baird, B. (1983).Structural studies on the membrane-bound immunoglobulin E-receptor complex. 2. Mapping of distances between sites on IgE and the membrane surface. Biochemistry 22,3475-3484. Holowka, D., and Metzger, H. (1982). Further characterization of the beta component of the receptor for IgE. Mol. Immunol. 19,219-227. Holowka, D., Hartmann, H., Kanellopolous, J., and Metzger, H. (1980). Association of the receptor for immunoglobulin E with an endogenous polypeptide on rat basophilic leukemia cells. J. Recept. Res. 1 , 4 1 4 8 . Holowka, D., Conrad, D. H., and Baird, B. (1985). Structural mapping of membranebound immunoglobulin E-receptor complexes: Use of monoclonal anti-IgE antibodies to probe the conformation of receptor-bound IgE. Biochemistry 24,6260-6267. Holtrop, S., Rijke-Schilder, G. P. M., Koene, R. A. P., and Tax,W. J. M. (1991). A polymorphic Fc receptor for mouse IgG2b on human B cells and monocytes. ImmunolOgy 74,613-620. Holtrop, S., Rijke-Schilder, G . P., van Doom, N. E., Jacobs, C. W., Koene, R. A., and Tax, W. J. (1993a). Proteolysis increases the Fc-mediated binding of murine IgG2b to human EBV-transformed B cells, but decreases the expression of FcyRII and FceRI1. Scand. J. Immunol. 38,259-266. Holtrop, S., Rijke-Schilder, G. P., Koene, R. A. and Tax, W. J. (1993b). The human Fc receptor for mouse IgG2b on monocytes and EBV-B cells is functionally inhibited by anti-HLA class I1 antibodies. ScandJ. Immunol. 37, 195-201. Howard, F. D.,Rodewald, H-R., Kinet, J-P., and Reinhen, E. L. (1990). CD3t; subunit

104

MARK D. HULETT AND P. MARKHOGARTH

can substitute for the y subunit of Fcs receptor type I in assembly and functional expression of the high a n i t y IgE receptor: Evidence for interreceptor complementation. Proc. Natl. Acad. Sci. U.S.A. 87,7015-7019. Huang, M-M., Indik, Z., Brass, L. F., Hoxie, J. A., Schreiber, A. D., and Brugge, J. S. (1992). Activation of FcyRII induces tyrosine phosphorylation of multiple proteins including FcyRII. 1.Biol. Chem. 267,5467-5473. Huber, A. H., Kelley, R. F., Gastinel, L. N., and Bjorkman, P. J. (1993). Crystallization and stoichiometry of binding of complex between a rat intestinal Fc receptor and Fc. 1.Mol. Biol. 230, 1077-1083. Huizinga, T. W. J., van der Schoot, C. E., Jost, C., Klaasssen, R., Kleiger, M., Kr. von dem Borne, A. E. G., Roos, D., and Tetteroo, P. A. T. (1988).The PI-linked receptor FcyRIII is released on stimulation of neutrophils. Nature (London)377, 667-669. Huizinga, T. W. J., Kerst, M., Nuyens, J. H., Vlug, A., Kr. von dem Borne, A. E. G., Roos, D., and Tetteroo, P. A. T. (1989a).Binding characteristics ofdimeric IgG subclass complexes to human neutrophils. 1.Immunol. 142,2359-2364. Huizinga, T. W. J., Kleijer, M., Roos, D., and Kr. Von dem Borne, A. E. G. (1989b). Differences between FcRIII of human neutrophils and human K/NK lymphocytes in relation to the NA antigen system. In “Leucocyte typing” (W. Knapp, et al., eds.), pp. 582-585. Oxford Univ. Press, Oxford. Huizinga, T. W. J., van Kemenade, F., Koenderman, L., Dolman, K. M., Kr. Von den The 40 kDa Fcy receptor Borne, A. E. G., Tetteroo, P. A. T., and Roos, D. (1989~). (FcRII) on human neutrophils is essential for the IgC-induced respiratory burst and IgG-induced phagocytosis. 1.Immunol. 142,2365-2269. Huizinga, T. W. J., Kleijer, M., Tetteroo, P. A. T., Roos, D., and Kr. Von dem Borne, A. E. G. (1990a). Biallelic neutrophil NA-antigen system is associated with a polymorphism on the phosphoinositol-linked Fcy receptor 111 (CD16). Blood 75, 213217. Huizinga, T. W. J., Kuijpers, R. W. A. M., Kleijer, M., Schulpen, T. W. J., Cuijpers, H. T. M., Roos, D., and Kr. Von dem Borne, A. E. G. (1990b). Maternal genomic neutrophil FcRII1 deficiency leading to neonatal isoimmune neutropenia. Blood 76, 1927-1932. Huizinga, T. W. J., Dolman, K. M., van der Linden, N. J. M., Kleijer, M., Nuijens, J. H., Kr. Von dem Borne, A. E. G., and Roos, D. (1990~).Phosphatidylinositol-linked FcRIII mediates exocytosis of neutrophil granule proteins, but does not mediate initiation of the respiratory burst. J . Immonol. 144, 1432-1437. Huizinga, T. W. J., De Haas, M., Kleijer, M., Nuijens, J. H., Roos, D., and Kr. Von dem Borne, A. E. G. (1990d). Soluble Fcy receptor 111 in human plasma originates from release by neutrophils. J . Clin. Invest. 86,416-423. Hulett, M. D., Osman, N., McKenzie, I. F. C., and Hogarth, P. M. (1991).Chimeric Fc receptors identify functional domains of the murine FcyRI high a n i t y receptor for IgG. 1.Immunol. 147, 1863-1868. Hulett, M. D., McKenzie, I. F. C., and Hogarth, P. M. (1993).Chimeric receptors identify immunoglobulin binding regions in human FcyRII and FcsRI. Eur. I. Immunol. 23, 640-645. Hulett, M. D., Witort, E., Brinkworth, R. I., McKenzie, I. F. C., and Hogarth, P. M. (1994). Identification of the IgG binding site of the low affinity receptor for IgG FcyRII: Enhancement and ablation of binding by site-directed mutagenesis. J . Biol. Chem. 269,15287-15293. Hunter, S., Huang, M-M., Indik Z. K., and Schreiber, A. D. (1993). FcyRIIIA mediated phagocytosis and receptor phosphorylation does not require the protein thrysine kinase src. Exp. Haematol. 21, 1492-1497.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

105

Hunziker, W., aKoch, T., Whitney, A., and Mellman, I. (1990). Fc receptor phosphorylation during receptor-mediated control of B-cell activation. Nature (London) 345, 628-632. Hunziker, W., Harter, C., Matter, K., and Mellman, I. (1991). Basolateral sorting in MDCK cells requires a distinct cytoplasmic domain determinant. Cell (Cambridge, MUSS.)66, 907-920. Huppi, K., Mock, B. A., Hilgers, J., Kochan, J., and Kinet, J-P. (1988). Receptors for FCE and Fcy are linked on mouse chromosome 1. J . Zmmunol. 141,2807-2810. Huppi, K., Siwarski, D., Mock, B. A., and Kimet, J-P. (1989). Gene mapping of the three subunits of the high afEnity FcR for IgE to mouse chromosome 1 and 19.J. Zmmunol. 143,3787-3791. Hutchcroft, J. E., Geahlen, R. L., Deanin, G. G., and Oliver, J. M. (1992).FcsRI-mediated tyrosine phosphorylation and activation of the 72 kDa protein-tyrosine-kinase, PTK72, in RBL-2H3 rat tumor mast cells. Proc. Natl. Acad. Sci. U S A . 89,9107-9111. Ierino, F. L., Hulett, M. D., McKenzie, I. F. C., and Hogarth, P. M. (1993). Human FcyRII monoclonal antibodies define structural d0rnains.J. Zmmunol. 150,1794-1799. Ierino, F. L., Powell, M. S., McKenzie, I. F. C., and Hogarth, P. M. (1994). Recombinant soluble human FcyRII: Production, characterisation and inhibition of the Arthus reaction. J . Exp. Med. 178, 1617-1628. Iersky, C., Rivera, J., Segal, D. M., and Triche, T. (1983). The fate of IgE bound to’rat basophilic leukemia cells. 11. Endocytosis of IgE oligomers and effect on receptor turnover. J. Zmmunol. 131,388-396. Ikuta, K., Takami, M., Kim, C. W., Honjo, T., Miyoshi, T.,Tagaya, Y., Kawake, T., and Yodoi, J. (1987). Human lymphocyte Fc receptor for IgE: Sequence homology of its cloned cDNA with animal lectins. Proc. Natl. Acad. Sci. U.S.A. 84, 819-823. h a , Y., Koide, Y., Nezu, N., and Yoshida, T. 0. (1987). Regulation of HLA class I1 antigen expression: Intracellular signalling molecules responsible for the regulation of I F N r and crosslinking of Fc receptors in HL-60 cells.]. Zmmunol. 139,17111717. Indik, Z., Kelly, C., Chein, P., Levinson, A. I., and Schreiber, A. D. (1991). Human FcyRII in the absence of other Fcy receptors mediates a phagocytic signal. J . Clin. Znuest. 88, 1766-1771. Irving, B. A., and Weiss, A. (1991). The cytoplasmic domain of the T cell receptor 5 chain is sufficient to couple to receptor associated signal transduction pathways. Cell (Cambridge, Mass.) 64,891-901. Irving, B. A,, Chan, A. C., and Weiss, A. (1993). Functional characterization of a signal transducing motif present in the T cell antigen receptor 4 chain. J . E x p . Med. 177, 1093-1103. Ishizaka, K., Ishizaka, T.,and Lee, E. (1979). Biologic function of the Fc fragment of E myeloma proteins. Immunochemistry 7,687-702. Ishizaka, T., and Ishizaka, K. (1975). Biology of immunoglobulin E. Prog. Allergy 19, 60- 121. Ishizaka, T., Dvorak, A. M., Conrad, D. H., Niebyl, J. R., Marquett, J. P., and Ishizaka, K. (1985).Morphologic and immunological characterization ofhuman basophils developed in cultures of cord blood mononuclear cells. J . Zmmunol. 134,532-540. Ishizaka, T., Helm, B. A., Hakimi, J., Niebly, J., Ishizaka, K., and Could, H. (1986). Biological properties of a recombinant human immunoglobulin E-chain fragment. Proc. Natl. Acad. Sci. U.S.A.83, 8323-8327. Israel, E. J., Simister, N., Freiberg, E., Caplan, A., and Walker, W. A. (1993).Immunoglobulin G binding sites on the human foetal intestine: A possible mechanism for the passive transfer of immunity from mother to infant. Zmmunology 79, 77-81.

106

MARK D. HULETT AND P. MARKHOGARTH

Joiner, K. A., Fuhram, S. A., Miettinen, H. M., Kasper, L. H., and Mellman, I. (1990). Toroplasma gondii: Fusion competence of parasitophorous vacuoles in Fc receptortransfected fibroblasts. Science 249, 641-646. Jones, D. H., Looney, R. J., and Anderson, C. L. (1985).Two distinct classes of IgG Fc receptors on a human monoctye cell line (U937) defined by differences in binding of murine IgG subclasses at low ionic strength. J . Immunol. 135,3348-3353. Kagami, M., Funatsu, Y., and Suzuki, T. (1989). Production and characterization of monoclonal antibodies to Fcy2a-binding protein isolated from the detergent lysate of a murine macrophage-like cell line, P388D1. J. Leuk. B i d . 45,311-321. Kanellopolous, J., Rossi, G., and Metzger, H. (1979). Preparative isolation of the cell receptor for immunoglobulin E. J . Biol. Chem. 254,7691-7697. Karas, S. P., Rosse, W. F., and Kurlander, R. J. (1982). Characterisation of the IgG-Fc receptor on human platelets. Blood 60, 1277-1282. Katz, H. R., Arm, J. P., Benson, A. C., and Austin, K. F. (1990). Maturation-related changes in the expression of FcyRII and FcyRIII on mouse mast cells derived in vitro and in vioo. J . Immunol. 145,3412-3417. Katz, H. R., Raizman, M. B., Gartner, C. S., Scott, H. C., Benson, A. C., and Austen, K. F. (1992). Secretory granule mediator release and generation of oxidative metabolites of arachidonic acid via Fc-IgG receptor bridging in mouse mast cells.J . Immunol. 148,868-871. Kawakami, T., Inagaki, N., Takei, M., Fukamachi, H., Coggeshall, K. M., Ishizaka, K., and Ishizaka, T. (1992).Tyrosine phosphorylation is required for mast cell activation by FcsRI cross-linking. J . Immunol. 148,3513-3519. Keegan,A. D., and Paul, W. E. (1992).Multichain immune recognitionreceptors: Similarities in structure and signalling pathways. Immunol. Today 13,63-68. Keegan, A. D., Pierce, J. H., Avtrip, J., Plant, M., and Paul, W. E. (1991). Liquid stimulation of transfected and endogenous growth factor receptors enhance cytokine production by mast cells. EMBO J. 10,3675-3682. Kennerly, D. A. (1990). Phosphatidylcholine is a quantitatively more important source of increased 1,2-diacylglycerolthan is phosphatidylinositol in mast cells. J . Immunol. 144,3912-3919. Kent, U. M., Mao, S-Y., Wofsy, C., Goldstein, B., Ross, S., and Metzger, H. (1994). Dynamics of signal transduction after aggregation of cell-surface receptors: Studies on the type I receptor for IgE. Proc. Natl. Acad. Sci. U.S.A.(in press). Kenten, J., Helm, B., Ishizaka, T., Cattini, P., and Gould, H. (1984). Properties of a human immunoglobulin 8-chain fragment synthesized in Eschericiha coli. Proc. Natl. Acad. Sci. U.S.A. 81,2955-2959. Kerr, M. A. (1990). The structure of function of human IgA. J. Biochem. 271,285-296. Kew, R. R., Grimaldi, C. M., Furie, M. B., and Fleit, H. B. (1992). Human neurophi1 FcyRIIIB and formyl peptide receptors are functionally linked during formylmethionyl-leucyl-phenylalanine-inducedchemotaxis. J . Immunol. 149,989-997. Khayat, D., Geffrier, C., Yoon, S., Scigliano, E., Soubrane, C., Wiel, M., Unkeless, J. C., and Jacquille. C. (1987). Soluble circulating Fcy receptors in human serum-a new ELISA for specific and quantiative detection. J. Imrnunol. 100,235-241. Kikutani, H., Inui, S., Sato, R., Barsumian, E. L., Owaki, H., Yamasaki, K., Kaisho, T., Uchibayashi, N., Hardy, R. R., Hirano, T., Tsunasawa, S., Sakiyama, F., Suemura, F. M., and Kishimoto, T. (1986a).Molecular structure of human lymphocyte receptor for immunoglobulin E. Cell (Cambridge, Mass.) 47,657-665. Kikutani, H., Suemura, N. M., Owaki, H., Nakamura, H., Saito, R., Yamasaki, K. Barsumina, E. L., Hardy, R. R., and Kishimoto, T. (198613). Fc epsilon receptor, a specific

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

107

differentiation marker transiently expressed on mature B cells before isotype switching. ]. Exp. Med. 164, 1455-1469. Kimberly, R. P., Tappe,N. J., Merriam, L.T.,Redecha, P. B., Edberg, J. C., Schwartzman, S., and Valinsky, J. E. (1989). Carbohydrates on human Fcy receptors: Interdependance on the classical IgG and nonclassical lectin binding sites on human FcyRIII expressed on neutrophils. 1.Immunol. 142,3923-3930. Kimberly, R. P., Ahlstrom, J. W., Click, M. E., and Edberg, J. C. (1990).The glycosyl phosphatidylinositol-linkedFcyRIIIpmn mediates transmembrane signalling events distinct from FcyRIJ. 1.E x p . Med. 171, 1239-1255. Kimberly, R. P., Edberg, J. C., Redecha, P. B., and Barinsky, M. (1991). FcyRIII-A glycoforms have distinct affinities for ligand. FASEB 1.5, 1346. Kindt, G. C., van de Winkel, J. G. J., Moore, S. A,, and Anderson, C. L. (1991).Identification and structural characterization of Fcy receptors on pulmonary alveolar macrophages. Am. ]. Physiol. 260,403-411. Kinet, J-P. (1989).Antibody-cell interactions: Fc receptors. Cell (Cambridge, Mass.)57, 351-354. Kinet, J-P. (1990). The high affinity receptor for IgE. Curr. Opinion Immunol. 2, 499-505. Kinet, J-P., and Metzger, H. (1990).Genes, structure and action of the high affinity Fc receptor for immunoglobulin E. In “Fc Receptors and the Action of Antibodies” (H. Metzger, ed.), pp. 239-259. Am. SOC. Microbiol., Washington, DC. Kinet, J-P., Alcaraz, G., Leonard, A., Wank, S., and Metzger, H. (1985). Dissociation of the receptor for immunoglobulin E in mild detergents. Biochemistry 24,4117-4124. Kinet, J-P., Metzger, H., Hakimi, J., and Kochan, J. (1987). A cDNA presumptively coding for the alpha subunit of the receptor with high a n i t y for immunoglobulin E. Biochemistry 26,4605-4610. Kinet, J-P., Blank, U., Ra, C., White, K., Metzger, H., and Kochan, J. (1988). Isolation and characterization of cDNAs coding for the p subunit of the high-affinity receptor for immunoglobulin E. Proc. Natl. Acad. Sci. U S A . 85,6483-6487. King, M., McDermott, P., and Schreiber. A. D. (1990).Characterization of the FcyR on human platelets. Cell Immunol. 128,462-479. Kipps, T . J., Parham, P., Punt, J., and Herzenberg, L. A. (1985).Importance of immunoglobulin isotype in human antibody dependent, cell-mediated cytotoxicity directed by murine monoclonal antibodies.]. E x p . Med. 161, 1-17. Klaus, G. G. B., Bijsterbosch, M. K., O’Garra, A., Harnet, M. M., and Rigley, K. P. (1987). Receptor signaling and crosstalk in B lymphocytes. Immunol. Reu. 99, 19-38. Koch, C. A,, Anderson, D., Moran, M. F.,Ellis, C., and Pawson, T. (1991). SH-2 and SH-3 domains: Elements that control interactions of cytoplasmic signalling proteins. Science 252,668-674. Kochan, J., Pettine, L. F., Hakimi, J., Kishi, K., and Kinet, J-P. (1988). Isolation of the gene coding for the a subunit of the human high affinity IgE receptor. Nucleic Acid. Res. 16,3584. Koenderman, L., Hermans, S. W., Capel, P. J., and van de Winkel, J. G. (1993). Granulocyte-macrophage colony-stimulating factor induces sequential activation and deactivation of binding via a low-dnity IgC Fc receptor, hFcyRII, on human eosinophils. Blood 81,2413-2419. Kolanus, W., Romeo, C., and Seed, B. (1992).Lineage-independent activation ofimmune system effector function by myeloid Fc receptors. EMBO 1.11,4861-4868. Koolwijk, P., Spierenburg, G. T. H., Frasa, T., Boot, J. H. A., van de Winkel, J. G. J., and Bast, E. J. E. G. (1989).Interaction between hybrid mouse monoclonal antibodies

108

MARK D. HULETT AND P. MARKHOGARTH

and the human high-affinity IgG FcR, huFcyRI, on U937: Involvement of only one of the mIgC heavy chains in receptor binding. 1.Immunol. 143, 1656-1662. Koolwijk, P., van de Winkel, J. C. J., Pfefferkom, L. C., Jacobs, C. W. M., Otten, I., Spierenburg, G. T., and Bast, B. J. E. G. (1991). Induction of intracellular Ca2' mobilization and cytotoxicity by hybrid mouse monoclonal antibodies. 1.Immunol. 147,595-602. Krajci, P., Solberg, R., Sandberg, M., Oyen, O., Jahnsen, T., and Brandtzaeg, P. (1989). Molecular cloning of the human transmembrane secretory component (poly-Ig receptor) and its mRNA expression in human tissues. Biochem. Biophys. Res. Commun. 158,783-789. Krajci, P., Grzeschik, K-H., Guerts van Kessel, A. H. M., Olaisen, B., and Brandtzaeg, P. (1991).The human transmembrane secretory component (poy-Ig receptor): Molecular cloning, restriction fragment length polymorphism and chromosomal sublocalization. Hum. Genet. 87,642-648. Krajci, P., Gedde-Dahl, T., Jr., Hoyheim, B., Rodge, S., Olaisen, B., and Brandtzaeg, P. (1992a).The gene encoding human transmembrane secretory component (locus PIGR) is linked to DlS58 on chromosome 1. Hum. Genet. 90,215-219. Krajci, P., Kvale, D., Tasken, K., and Brandtzae, P. (199%). Molecular cloning and characterization of the gene encoding the human transmembrane secretory component (the poly-Ig receptor). Eur. 1.Immunol. 22,2309-2315. Krajci, P., Tasken, K., Kvale, D., and Brandzaeg, P. (1993).Interferon-gamma stimulation of messenger RNA for human secretory component (poly-Ig receptor) depends on continuous intermediate protein synthesis. Scand. 1.Immunol. 37,251-256. Kremer, E. J,, Kalatzis, V., Baker, E., Callen, D. F., Sutherland, G. R., and Maliszewski, C. R. (1992).The gene for the human IgA Fc receptor maps to 19q13.4.Hum. Genet. 89,107-108. Krutmann, J., Kirnbauer, R., Kock, A., Schwarz, T., Schopf, E., May, L. T., Sehgal, P. B., and Luger, T. A. (1990).Crosslinking Fc receptors on monocytes triggers IL6 production. 1.Immunol. 145, 1337-1342. Kulczycki, A., Jr. (1984).Human neutrophils and eosinophils have structurally distinct Fc y receptors. 1.Immunol. 133, 849-854. Kulczychi, A., Jr., and Metzger, H. (1974).The interaction of IgE with rat basophilic leukemia cells. 11. Quantitative aspects of the binding reaction. 1.E x p . Med. 140, 1676- 1695. Kulczycki, A,, Jr., Iersky, C., and Metzger, H. (1974). The interaction of IgE with rat basophilic leukemia cells. I. Evidence for specific binding of IgE.]. E x p . Med. 139, 600-616. Kulczycki, A,, Jr., McNeamey, T. A,, and Parker, C. W. (1976). The rat basophilic leukemia cell for IgE. I. Characterization as a glycoprotein.]. Immunol. 117,661-665. Kulczycki, A., Jr., Webber, J., Soares, H. A,, Onken, M. D., Thompson, J. A., Chaplin, D. D., Loh, D. Y., and Tillinghast, J. P. (1990).Genomic organization of mouse FcyR genes. Proc. Natl. Acad. Sci. U.S.A.87,2856-2860. Kurlander, R. J., and Batker, J. (1982). The binding of human immunoglobulin G1 monomer and small covalently cross-linked polymers of immunoglobulin G1 to human peripheral blood monocytes and polymorphonuclear leukocytes. 1.Clin. Inoest. 69, 1-8. Kurlander, R. J., Haney, A. F., and Gartrell, J. (1984).Human peritoneal macrophages possess two populations of IgG Fc receptors. Cell. Immunol. 86,479-490. Kurosaki, T., and Ravetch, J. V. (1989). A single amino acid in the GPI attachment domain determines the membrane topology of FcyRIII. Nature (London) 342, 805-807.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

109

Kurosaki, T., Gander, I., Wirthmueller, U., and Ravetch, J. V. (1992). The P subunit of FcERI is associated with the FcyRIII on mast cells. J . E x p . Med. 175,447-451. Kushner, B. H., and Cheung, N. K. V. (1992). Absolute requirement of CD11/CD18 adhesion molecules, FcRII, and the phosphatidylinositol-linked FcRIII for monoclonal antibody-mediated neutrophil anti human tumor cytotoxicity. Blood 79, 1484-1490. Kuster, H., Thompson, H., and Kinet, J-P. (1990). Characterization and expression of the gene for the human Fc receptor y subunit: Definition of a new gene family. J . Biol. Chem. 265,6448-6452. Kvale, D., Levhaug, D., Sollid, L. M., and Brandtzaeg, P. (1988).Tumor necrosis factoralpha up-regulates expression of secretory component, the epithelial receptor for polymeric Ig. J . Immunol. 140, 3086-3089. Lah, M., Quelch, K., Deacon, N. J., McKenzie, I. F. C., and Hogarth, P. M. (1990). Identification of the mouse PFcyRII polymorphism by direct sequencing of amplified genomic DNA. lmmunogenetics 31,202-206. Lalezari, P. (1984). Granulocyte antigen systems. In “Zmmunohaematology” (C. P. Engelfriet, J. J. Van Loghem, and A. E. G. Kr., eds.), p. 33. Elsevier, Amsterdam. Lamon, E. W., Anderson, B., Whitten, H. D., Hurst, M. M., and Ghanta V. (1976). IgM complex receptors on subpopulations of murine lymphocytes. J . Zmmunol. 116, 1199-1203. Lane, B. C., and Cooper, S. M. (1982). Fc receptors of mouse cell lines I. Distinct proteins mediate the IgG subclass-specific Fc binding activities of macrophages. J . Zmmunol. 128, 1819-1824. Lane, B. C., Mitchell, J. K., Mitchell, M. S., andcooper, S. M. (1980).Structural evidence for distinct subclass-specific Fc receptors on mouse peritoneal macrophages. J . E x p . Med. 152,1147-1161. Lanier, L. L., Le, A. M., Phillips, J. H., Warner, N. L., and Babcock, G. F. (1983). Subpopulations of human natural killer cells defined by expression of the Leu-7 (HNK-1) and Leu-11 (NK-15) antigens. J. Zmmunol. 131, 1789-1796. Lanier, L. L., Kipps, T. J., and Phillips J. H. (1985). Functional properties of a unique subset of cytotoxic CD3’ T lymphocytes that express Fc receptor for IgG (CD16/ Leu 11 antigen). J . Exp. Med. 162,2089-2106. Lanier, L. L., Ruitenberg, J. J., and Phillips J. H. (1988). Functional and biochemical analysis of CD16 antigen on natural killer cells and granulocytes. J . Zmmunol. 141, 3478-3485. Lanier, L. L., Cwirla, S., Yu, G., Testi, R., and Phillips J. H. (1989a).Membrane anchoring of a human IgG Fc receptor (CD16) determined by a single amino acid. Science 246, 1611-1613. Lanier, L. L., Yu, G., and Phillips J. H. (198913).Co-association of CD3-5 with receptor (CD16) for IgG Fc on human natural killer cells. Nature (London)342, 803-805. Lanier, L. L., Yu, G., and Phillips J. H. (1991).Analysis of FcyRIII (CD16) membrane expression and association with CD35 and FcsRI-y by site-directed mutation.]. Immunol. 146,1571-1576. Latour, S., Bonnerot, C., Fridman, W. H., and Daeron, M. (1992). Induction of tumor necrosis factor-a production by mast cells via FcyR: Role of the FcyRIII y subunit. J . Zmmunol. 149,2155-2162. Lazarus, A. H., Mills, G. B., Crow, A. R., and Delovitch, T. L. (1991). Antigen induced Fc receptor-dependent and independent B cell densitization: An elevation in (Ca2‘)i is not sufficient and protein kinase C activation is not required for these pathways of surfac IgM-mediated desensitization. J. Zmmunol. 147, 1739-1745. Leatherbarrow, R. J., Rademacher, T. W., Dwek, R. A., Woof, J. M., Clark, A,, Burton,

110

MARK D. HULETT AND P. MARKHOGARTH

D. R., Richardson, N., and Feinstein, A. (1985). Effector functions of a monoclonal aglycosylated mouse IgG2a: Binding and activation of complement component C1 and interaction with human monocyte Fc receptor. Mol. Immunol. 22,407-415. Le Conait, M., Kinet, J-P., and Berger R. (1990). The human genes for the a and y subunits of the mast cell receptor for immunoglobulin E are located on human chromosome band lq23. Zmmunogenetics 32, 183-186. Lennartz, M. R., and Brown E. J. (1991). Arachidonic acid is essential for IgG Fc receptor-mediated phagocytosis by human monocytes. J . Zmmunol. 147,621-626. Letellier, M. M., Sarfati, M., and Delespesse, G. (1989). Mechanisms of formation of IgE-binding factors (soluble CD23)-I. Fc epsilon RII bearing B cells generate IgEbinding factors of different molecular weights. Mol. Zmmunol. 26, 1105-1112. Letellier, M. M., Nakajima, T., Pulido-Cejudo, G., Hofstetter, H., and Delespesse, G. (1990). Mechanisms of formation of human IgE-binding factors (soluble CD23). 111. Evidence for a receptor (FcsRI1)-associatd proteolytic activity. J . Exp. Med. 172, 693-700. Letourneur, F., and Klausner R. D. (1991). T cell and basophil activation through the cytoplasmic tail of T-cell-receptor 5 chain family protein. Proc. Natl. Acad. Sci. U.S.A. 88,8905-8909. Letourneur, O., Kennedy, I. C. S., Brini, A. T., Ortaldo, J. R., O’Shea, J. J., and Kinet, J-P. (1991).Characterization of the family of dimers associated with Fc receptors (FcsRI and FcyRIII). J . Immunol. 147,2652-2656. Lewis, V. A., Koch, T., Plutner, H., and Mellman I. (1986). A complementary DNA clone for a macrophage-lymphocyte Fc receptor. Nature (London)324,372-375. Li, W., Deanin, G. G., Margolis, B., Schlessinger, J., and Oliver, J. M. (1992). FcsRImediated tyrosine phosphorylation of multiple proteins, including phospholipase Cyl and the receptor PyRI complex, in RBL-2H3 rat basophiiic leukemia cells. Mol. Cell. Biol. 12,3176-3182. Liao, F., Shin, H. S., and Rhee, S. G. (1992).Tyrosine phosphorylation ofphospholipase Cy-1 induced by crosslinking of the high affinity or low-dnity Fc receptor for IgG in U937 cells. Proc. Natl. Acad. Sci. U.S.A. 89,3659-3663. Liao, F., Shin, H. S., and Rhee, S. G. (1993). Cross-linking of FcyRIIIA on natural killer cells results in tyrsoine phosphorylation of PLC-71 and PLC-72. J. Zmmunol. 150, 2668-2674. Lin, C-T., Shen, Z., Boros, P., and Unkeless, J. C. (1994). Fc receptor-mediated signal transduction. J . Clin. Immunol. 14, 1-13. Liu, F-T., and Robertson, M. W. (1988).cDNA heterogeniety suggests structural variants related to the high afhity IgE receptor. Proc. Natl. Acad. Sci. U.S.A. 85,5639-5643. Liu, F-T., Albrandt, K., Bry, C. G., and Ishizaka T. (1984).Expression of a biologically active fragment of human IgE E chain in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 81,5369-5373. Lobell, R. B., Austen, K. F., and Katz, H. R. (1994). FcyR-mediated endocytosis and expression of cell surface FcyRIIbl and FcyRIIb2 by mouse bone marrow culturederived progenitor mast cells. J. Immunol. 152,812-818. Looney, R. J., Abraham, G. N., and Anderson, C. L. (1986a). Human monocytes and U937 bear two distinct Fc receptors for IgG. J . Immunol. 136, 1641-1647. Looney, R. J., Ryan, D. H., Takahashi, K., Fleit, H. B., Cohen, H. J., Abraham, G. N., and Anderson, C. L. (1986b). Identification of a second class of IgG Fc receptors on human neutrophils: A 40kD molecule also found on eosinophils. J . Exp. Med. 163, 826-836. Looney, R. J., Anderson, C. L., Ryan, D. H., and Rosenfeld, S. I. (1988). Structural polymorphism of the human platelet Fcy receptor. J . Immunol. 141,2680-2683.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

111

Lopez, A. F., Battye, F. L., and Vadas, M. A. (1985).Fc receptors on mouse neutrophils and eosinophils: Antigenic characteristics, isotype specificity and relative cell membrane density measured by flow cytometry. Immunology 55,125-133. Loube, S . R., McNabb, T. C., and Dorrington, K. J. (1978).Isolation of an Fey-binding protein from the cell membrane of a macrophage-like cell line (P388D1)after detergent solubilization. J . Immunol. 120,709-715. Ludin, C., Hoftstetter, H., Sarfati, M., Levy, C. A., Suter, U., Alaimo, D., Kichherr, E., Frost, H., and Delespesse, G. (1987). Cloning and expression of the cDNA coding for a human lymphocyte IgE receptor. E M B O J . 6, 109-114. Lum, L. G., Muchmore, A. V., Keren, D., Decker, J., Koski, I., Strober, W., and Blaese, R. M. (1979). A receptor for IgA on human T lymphocytes. J . Immunol. 122, 6569. Lund, J., Tanaka, T., Takahashi, N., Sarmay, G., Arata, Y., and Jefferis, R. (1990). A protein structural change in aglycosylated IgG3 correlates with loss of huFcyRI and huFcyRII1 binding andlor activation. Mol. Immunol. 27, 1145-1153. Lund, J., Winter, G., Jones, P. T., Pound, J. D., Tanaka, T., Walker, M. R., Artymuik, P. J., Arata, Y., Burton, D. R., Jefferis, R., and Woof, J. M. (1991).Human FcyRI and FcyRII interact with distinct but overlapping sites on human IgG. J. Immunol. 147, 2657-2662. Lydyard, P. M., and Fanger, M. W. (1982).Characteristics and function of Fc receptors on human lymphocytes. Immunology 47, 1. Lynch, R. G., and Sandor, M. (1990). Fc receptors on T and B lymphocytes. In “Fc Receptors and the Action of Antibodies” (H. Metzger, ed.), pp. 305-334. American Society of microbiology, Washington, DC. Macintyre, E. A., Abdul-Gaffer, R., O’Flynn, K., Pilkington, G. R., Farace, F., Morgan, J., and Linch, D. C. (1988).Mechanism of human monocyte activation via the 40-kD Fc receptor for IgG. J. Immunol. 141,4333-4343. Macintyre, E. A., Roberts, P. J., Jones, M., van der Schoot, C. E., Favalaro, E. J., Tidman, N., and Linch, D. C. (1989).Activation of human monocytes occurs on crosslinking monocytic antigens to an Fc receptor. J . Immunol. 142,2377-2383. Maliszewski, C. R., Shen, L., and Fanger, M. W. (1985). The expression of receptors for IgA on human monocytes and calcitriol-treated HL-60 cells. J. Immunol. 135, 3878-3881. Maliszewski, C. R., March, C. J., Shoenborn, M. A,, Gimpel, S., and Li, S. (1990). Expression cloning of a human Fc receptor for IgA. J . Erp. Med. 172, 1665-1672. Mallamaci, M. A., Chizzonite, R., Griffen, M., Nettleton, M., Hakimi, J., Tsien, W-H., and Kochan, J. P. (1993).Identification of sites on the human FCERICY subunit which are involved in binding human and rat IgE. J. Biol. Chem. 268,22076-22083. Mantzioris, B. X.,Berger, M. F., Sewell, W., and Zola. H. (1993).Expression of the Fc receptor for IgG (FeyRIIKD32) by human circulating T and B lymphocytes. J. Immunol. 150,5175-5184. Mao, S-Y.,Varin-Blank, N., Edidin, M., and Metzger, H. (1991). Immobilization and internalization of mutated IgE receptors in transfected cells. J. Immunol. 146, 958-966. Mao, S-Y., Alber, G., Rivera, J., Kochan, J., and Metzger, H. (1992). Interaction of aggregated native and mutant IgE receptors with the cellular skeleton. Proc. Natl. Acad. Sci. U.S.A. 89,222-226. Mao, S-Y., Pfeiffer, J. R., Oliver, J. M., and Metzger, H. (1993). Effects of subunit mutation on the localization to coated pits and internalization of cross-linked Ig-E receptor complexes. J. Immunol. 151,2760-2774. Margolis, B., Hu, P., Katzav, S., Li, W., Oliver, J. M., Ullrich, A., Weiss, A., and Sches-

112

MARK D. HULETT AND P. MARKHOGARTH

singer, J. (1992). Tyrosine phosphorylation of vav, protooncogene product containing SH2 domain and transcription factor motifs. Nature (London)356, 71-74. Masuda, M., and Roos, D. (1993). Association of all three types of FcyR (CD64, CD32, and CD16) with a y-chain homodimer in cultured human monocytes. 1. Immunol. 151,7188-7195. Masuda, M., Verhoeven, A. J., and Roos, D. (1993).Tyrosine phosphorylation of a ychain homodimer associated with FcyRIII (CD16) in cultured human monocytes. J . Immunol. 151,6382-6388. Mathur, A., Lynch, R. G., and Koller, G. (1988a).The contribution of constant region domains to the binding of murine IgM to F c receptors. ~ J. Immunol. 140, 143-147. Mathur, A., Lynch, R.G., and Koller, G. (1988b).Expression distribution and specificity of Fc receptors for IgM on murine B cells. J . Immunol. 141,1855-1862. Matsuda, H., Nakamura, S., Ichikawa, Y., Kozai, K., Takano, R.,Nose, M., Endo, S., Nishimura, Y., and Arata, Y. (1990). Protein NMR studies of the structure of the Fc fragment of human immunoglobulin G1: Comparisons of native and recombinant proteins. Mol. Immunol. 27, 571-579. Matthews, G., Neher, E., and Penner, R. (1989). Second messenger-activated calcium influx in rat peritoneal mast cells. J. Physiol. 418, 105-130. Maurer, D., Fiebiger, E., Reininger, B., Wolffwiniski, B., Jouvin, M. H., Kilgus, O., Kinet, J-P., and Stingl, G. (1994). Expression of a functional high affinity immunoglobulin E receptors (FceRI) on monocytes of atopic individuals. J . E r p . Med. 179, 745-750. Mazengera, R. L., and Kerr, M. A. (1990). The specificity of the IgA receptor purified from human neutrophils. J . Biochem. 272, 159-165. McCloskey, M. A. (1988). Cholera toxin potentiates IgE-coupled inositol phospholipid hydrolysis and mediator secretion by RBL-2H3 cells. Proc. Natl. Acad. Sci. U.S.A. 85,7260-7264. McCloskey, M. A. (1993). Immobilization of Fce receptors by wheat germ agglutinin. J . Immunol. 151,3237-3251. McGhee, J. R., Mestecky, J., Elson, C. O., and Kiyono, H. (1989). Regulation of IgA and immune response by T cells and interleukins. J. Clin. Immunol. 9, 175. McKenzie, S., Keller, M., Cassel, D., Schreiber, A,, Schwartz, S., Surrey, S., and Rappaport, E. (1992). Human FcyRIIA gene: Genomic sequence and transcription start sites. Mol. Immunol. 29, 1165-1174. Mellman, I., and Unkeless, J. C. (1980).Purification of a functional Fc receptor through the use of a monoclonal antibody. J . E x p . Med. 152,1048-1069. Mellman, I., Plutner., H., Steinman, R. M., Unkeless, J. C., and Cohn, Z. A. (1983). Intemalisation and degradation of macrophage Fc receptors during receptor-mediated phagocytosis. J. Cell. Biol. 96, 887-895. Mellman, I., Koch, T., Healey, G., Hunziker, W., Lewis, V., Plutner, H., Miettinen, H., Vaux, D., Moore, K., and Stuart, S. (1988). Structure and function of Fc receptors on macrophages and lymphocytes. J . Cell. Sci. Suppl. 9,45-65. Mendel, D. B., Shen, L., and Guyre, P. M. (1988).The effect of tumor necrosis factor on three IgG Fc receptors of human neutrophils. Clin. Res. 35,460. Mendoza, G. R., and Metzger, H. (1976). Distribution and valency of receptor IgE on rodent mast cells and related tumor cells. Nature (London)264,548-550. Menon, A. K., Holowka, D., and Baird, B. (1984). Small oligomers of immunoglobulin E (IgE) cause large scale clustering of IgE receptors on the surface of rat basophilic leukemia cells. J. Cell Biol. 98,577-583. Mestecky, J., and McGhee, R. (1987). Immunoglobulin A (IgA): Molecular and cellular interactions involved in IgA biosynthesis and immune response. Ado. Immunol. 40, 153-245.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

113

Metzger, H. (1988). Molecular aspects of receptors and binding factors for IgE. Adu. Immunol. 43,277-312. Metzger, H. (1991). The high aflinity receptor for IgE on mast cells. Clin. Exp. Allergy 21,269-279. Metzger, H. (1992a).The receptor with high affinity for IgE. Immunol. Reo. 125,37-48. Metzger, H. (1992b). Transmembrane signaling: The joy of aggregation. J. Immunol. 149,1477-1487. Metzger, H., Kinet, J-P., Perez-Montfort, R., Rivnay, B., and Wank, S. A. (1983). A tetrameric model for the structure of the mast cell receptor for immunoglobulin E. Prog. Immunol. 5,493-501. Metzger, H., Alcaraz, G., Hohman, R., Kinet, J-P., Pribluda, V., and Quarto, R. (1986).The receptor with high af€inity for immunoglobulin E. Annu. Reo. Immunol. 4,419-470. Meyer, T., Holowka, D., and Stryer, L. (1988).Highly co-operative opening of calcium channels by inositol 1,4,5-triphosphate. Science 240,653-656. Miettinen, H. M., Rose, J. K., and Mellman, I. (1989). Fc receptor isoforms exhibit distinct abilities for coated pit localization as a result of cytoplasmic domain heterogeneity. Cell (Cambridge, Mass.) 58,317-327. Miettinen, H. M., Matter, K., Hunziker, W., Rose, J. K., and Mellman, I. (1992). Fc receptor endocytosis is controlled by a cytoplasmic domain determinant that actively prevents coated pit localization. J. Cell Biol. 116, 875-888. Millard, P. J., Ryan, T. A., Webb, W. W., and Fewtrell, C. (1989). Immunoglobulin E receptor crosslinking induces oscillations in intracellular free ionized calcium in individual tumor mast cells. J . Biol. Chem. 264, 19730-19739. Miller, L., Blank, U., Metzger, H., and Kinet, J-P. (1989). Expression of high-anity binding of human immunoglobulin E by transfected cells. Science 244, 334-337. Miller, L., Alber, G., Varin-Blank, N., Ludowyke, R., and Metzger, H. (1990).Transmembrane signalling in P815 mastocytoma cells by transfected IgE receptors. J. Biol. Chem. 265,12444-12453. Millet, I., Briere, F., Vincent, C., Rousset, F., Andreoni, C., de Vries, J. E., and Revillard, J. P. (1989). Spontaneous expression of a low affinity receptor for IgA (FcaR) on human B cell lines. Clin. Exp. Immunol. 76,268-273. Monteiro, R. C., Kubagawa, H., and Cooper, M. D. (1990).Cellular distribution, regulation, and biochemical nature of an Fca receptor in humans.]. E x p . Med. 171,597-613. Monteiro, R. C., Cooper, M. D., and Kubagawa, H. (1992). Molecular heterogeneity of Fca receptors detected by receptor-specific monoclonal antibodies. J . Immunol. 148, 1764-1770. Monteiro, R. C., Hostoffer, R. W., Cooper, M. D., Bonner, J. R., Gartland, G. L., and Kubagawa, H. (1993). Definition of IgA receptors on eosinophils and their enhanced expression in allergic individuals. J. Clin. Inuest. 92, 1681-1685. Moretta, L., Ferrarini, M., Durante, M. L., and Mingari, M. C. (1975). Expression of a receptor for IgM by human T cells in vitro. Eur. J . Immunol. 5,565. Moretta, L., Webb, S. R., Grossi, C. E., Lydyard, P. M., and Cooper, M. D. (1977). Functional analysis of two human T-cell populations: Help and suppression of B-cell responses by T cells bearing receptors for IgM or IgG. J . Exp. Med. 146, 184-200. Most, J., Gattringer, C., Neumayer, H. P., Schulz, T. F., Dierch, M. P., and Huber, H. (1992). Direct demonstration of binding of aggregated mouse IgGl to the 40kDa Fc receptor for IgG (FcyRII) in both high and low responders. Immunol. Lett. 32, 159- 165. Mostov, K. E., Friedlander, M., and Blobel, G. (1984). The receptor for transepithelial transport of IgA and IgM contains multiple immunoglobulin-like domains. Nature (London) 308,37-43.

114

MARK D. HULETT AND P. MARKHOGARTH

Muta, T., Kurosaki, T., Misulovin, Z., Sanchez, M., Nussenzweig, M. C., and Ravetch, J. V. (1994).A 13-amino acid motif in the cytoplasmic domain of FcyRIIB modulates B-cell receptor signalling. Nature (London)368, 70-72. Nakamura, T., Kubagawa, H., Ohno, T., and Cooper, M. D. (1993). Characterization of an IgM Fc-binding receptor on human cells. J . Immunol. 151,6933-6941. Nakashima, S . , Fujimiya, H., Miyata, H., and Nozawa, Y. (1991).Antigen induced biphasic diaglycerol formation in RBL-2H3 cells: The late sustained phase due to phosphatidylcholine hydrolysis is dependent on protein kinase C. Biochem. Biophys. Res. Commun. 177,336-342. Narasimhan, V., Holowka, D., Fewtrell, C., and Baird, B. (1988).Cholera toxin increases the rate of antigen-stimulated calcium influx in rat basophilic leukemia cells. J . Biol. Chem. 263,19626-19632. Narasimhan, V., Holowka, D., and Baird, B. (1990).A guanine nucleotide-binding protein participates in IgE receptor-mediated activation of endogenous and reconstituted phospholipase A2 in a permeabilized cell system. J. Biol. Chem. 265, 1459-1464. Naziruddin, B., Duffy, B. F., Tucker, J., and Mohanakumar, T. (1992). Evidence for cross-regulation of FcyRIIIB (CD16) receptor-mediated signaling by FcyRII (CD32) expressed on polymorphonuclear neutrophils. J. Immunol. 149,3702-3709. Neher, E. (1988).The influence of intracellular calcium concentration on degranulation of dialysed mast cells from rat peritoneum. J . Physiol. 395, 193-214. Nelson, D. J., Jacobs, E. R.,Tang, J. M., Zeller, J. M.,andBone, R. C. (1985).Immunoglobulin G-induced single ionic channels in human aveolar macrophage membranes. J . Clin. Invest. 76,500-507. Nio, N., Seguro, K., Ariyoshi, Y., Nakanishi, K., Kita, A., Ishiii, K., and Nakamura, H. (1990). Inhibition of histamine release by synthetic human IgE peptide fragments: Structure-activity studies. In “Peptide Chemistry 1989” (N. Yanaihara, ed.), pp. 204-208. Protein Res. Found., Osaka, Japan. Nishibe, S., Wahl, M. I., Sotomajor, S. M., Tonks, N. K., Rhee, S. G., and Carpenter, G. (1990). Increase of the catalytic activity of phospholipase C-gamma-1 by tyrosine phosphorylation. Science 250, 1253-1256. Nissim, A., Jouvin, M-H E., and Eshhar, Z. (1991). Mapping of the high affinity FCE receptor binding site to the third constant region domain of IgE. EMBOJ. 10,101-107. Nissim, A. S., Schwarzbaum, R., Siraganian, R., and Eshhar, Z. (1993). Fine specificity of the IgE interaction with the low and high affinity Fcs receptors. J . Immunol. 150, 1365-1374. Oakey, R. J., Howard, T. A., Hogarth, P. M., Tani, K., and Seldin, M. F. (1992).Chromosomal mapping of the high affinity Fcy receptor gene. Immunogenetics 33,279-282. Odin, J. A., Edberg, J. C., Painter, C. J., Kimberly, R. P., and Unkeless, J. C. (1991). Regulation of phagocytosis and (Ca2’), flux by distinct regions of an Fc receptor. Science 254,1785-1788. O’Grady, J. H., Looney, R. J., and Anderson, C. L. (1986). The valence for ligand of the human monocyte phagocyte 72 kD high-affinity IgG Fc receptor is 0ne.J. Immunol. 137,2307-2310. Ohno, T., Kubagawa, H., Sanders, S. K., and Cooper, M. D. (1990). Biochemical nature of a Fcp receptor on human B-lineage cells. J. Exp. Med. 172,1165-1175. Ohta, Y., Stossel, T. P., and Hartwig, J. H. (1991). Ligand-sensitive binding of actinbinding protein to immunoglobulin G Fc receptor I (FcyRI). Cell 67, 275-282. Okada, M., Nada, S., Yamanashi, Y., Yamamoto, T., and Makagawa, H. (1991).CSK: A protein tyrosine kinase involved in regulation of STC family kinases. J . Biol. Chem. 266,24249-24252.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

115

Okafor, G. O., Turner, M. W., and Hay, F. C. (1974). Localization of monocyte binding site of human immunoglobulin G. Nature (London)248,228-230. Oliver, J. M., Seagrave, J., Stump, R. F., Pfeiffer, J. R., and Deanin, G. G. (1988). Signal transduction and cellular response in RBL-2H3 mast cells. Prog. Allergy 42,185-245. Oliver, C., Sahara,N., Kitani, S., Robbins, A. R., Mertz, L. M., and Siraganian, R. P. (1992). Binding of monoclonal antibody AA4 to gangliosides on rat basophalic leukemia cells produces changes similar to those seen with Fc epsilon receptor activation. J . Cell Biol. 116,635-646. Orloff, D. G., Ra, C., Frank, S. J., Klausner, R. D., and Kinet, J-P. (1990). Family of disulphide-linked dimers containing the 5 and p chains of the T cell receptor and the y chain of Fc receptor. Nature (London)347, 189-191. Ory, P. A., Goldstein, I. A., Kwoh, E. E., and Clarkson, S. B. (1989). Characterization of polymorphic forms of FcRIII on human neutrophi1s.J. Clin. Inuest. 83,1676-1681. O’Shea, J. J., Wiessman, A. M., Kennedy, I. C. S., and Ortaldo, J. R. (1991).Engagement of the natural killer cell IgG Fc receptor results in tyrosine phosphorylation of the 5 chain. Proc. Natl. Acad. Sci. U.S.A.88, 350-354. Osman, N., Kozak, C. A., McKenzie, I. F. C., and Hogarth, P. M. (1992). Structure and mapping of the gene encoding the mouse high affinity FcyRI gene and chromosomal location of the human FcyRI gene. J . lmmunol. 148,1570-1575. Pan, L., Mendel, D. B., Zurlo, J., and Guyre, P. M. (1990). Regulation of the steady state level of FcyRI mRNA by IFN-gamma and dexamethasone in human monocytes, neutrophils, and U937 cells. J . Immunol. 145,267-275. Pang, J., Taylor, G. R., Munroe, D. G., Ishaque, A., Fung-Leung, W-P., Lau, C. Y., Liu, F-T., and Zhou, L. (1993). Characterization of the gene for the human high affinity receptor (FcsRI) a-chain. J. lmmunol. 151,6166-6174. Paolini, R., Jouvin, M-H., and Kinet, J-P. (1991).Phosphorylation and dephosphorylation of the high affinity receptor for immunoglobulin E immediately after engagement and disengagement. Nature (London) 353,855-858. Paolini, R., Numerof, R., and Kinet, J-P. (1992). Phosphorylation / dephosphorylation of high-affinity IgE receptors: A mechanism for couplingluncoupling a large signalling complex. Proc. Natl. Acad. Sci. U.S.A.89, 10733-10737. Park, D., and Rhee, S. G. (1992). Phosphorylation of Nck in response to a variety of receptors, phorbol mysistate acetate and cyclic AMP. Mol. Cell B i d . 12,5816-5823. Park, D. J., Min, H. H., and Rhee, S. G. (1992). IgE-induced tyrosine phosphorylation of phospholipase C-yl in rat basophilic leukemia cells. J . Biol. Chem. 266, 2423724240. Park, J-G., Isaacs, R. E., Chien, P., and Schreiber, A. D. (1993a).In the absence of other Fc receptors, FcyRIIIA transmits a phagocytic signal which requires the cytoplasmic domain of its y subunit. J . Clin. Invest. 92, 1967-1973. Park, J-G., Murray, R.K., Chien, P., Darby, C., and Schreiber, A. D. (1993b).Conserved cytoplasmic tyrosine residues of the y subunit are required for a phagocytic signal mediated by FqRIlIA. J. Clin. Invest. 92, 2073-2079. Parker, C. W. (1987). Lipid mediators produced through the lipoxygenase pathway. Annu. Reu. lmmunol. 5,65-84. Parren, P. W., Warmerdam, P. A. M., Boeije, L. C., Capel, P. J., van de Winkel, J. G. J., Aarden, L. A. (1992). Characterization of IgG FcR-mediated proliferation of human T cells induced by mouse and human anti-CD3 monoclonal antibodies: Identification of a functional polymorphism to human IgG2 anti-CD3. J . lmmunol. 148,695-701. Partridge, L. J.. Woof, J. M., Jefferies, R.,and Burton, D. R. (1988).The use of anti-IgG

116

MARK D. HULETT AND P. MARKHOGARTH

monoclonal antibodies in mapping the monocyte subpopulation in human peripheral blood. Mol. Immunol. 23, 1365-1372. Passlick, B., Flieger, D., and Ziegler-Heitbrock, H. W. L. (1989). Identification and characterisation of a novel monocyte subpopulation in human peripheral blood. Blood 74,2527-2534. Pearse, R. N., Feinman, R., and Ravetch, J. V. (1991). Characterization of the promoter of the human gene encoding the high affinity IgG receptor: Transcriptional induction by y-interferon is mediated through common DNA response elements. Proc. Nutl. Acad. Sci. U.S.A. 88, 11305-11309. Peltz, G. A., Frederick, K., Anderson, C. L., and Peterlin, B. M. (1988a).Characterisation of the human monocyte high affinity Fc receptor (huFcR1). Mol. Immunol. 25, 243-250. Peltz, G. A., Troustine, M. L., and Moore, K. (1988b).Cloned and expressed human Fc receptor for IgG mediates anti-CD3 dependent 1ymphoproliferation.J.Immunol. 141, 1891-1896. Peltz, G. A., Grundy, H. O., Labo, R. V., Yssel, H., Barsh, G. S., and Moore, K. W. (1989). Human FcyRIII, cloning, expression, and identification of the chromosomal locus of two Fc receptors for IgG. Proc. N a t l . Acad. Sci. U.S.A. 86, 1013-1017. Penner, R., and Neher, E. (1988). Secretory responses of rat peritoneal mast cells to high intracellular calcium. FEBS Lett. 226,307-313. Perez-Montfort, R.,and Metzger, H. (1982). Proteolysis of soluble IgE-receptor complexes: Localization of sites on IgE which interact with the Fc receptor. Mol. Immunol. 19,1113-1125. Perez-Montfort, R., Kinet, J-P., and Metzger, H. (1983). A previously unrecognised subunit of the receptor for immunoglobulin E. Biochemistry 22,5722-5728. Perussia, B., and Ravetch, J. V. (1991). FcyRIII (CD16) on human macrophages is a functional product of the FcyRIII-2 gene. Eur. J. Immunol. 21,425-429. Perussia, B., Starr, S., Abraham, S., Fanning, V.,and Trinchieri, G. (1983). Human natural killer cells analyzed by B73.1, a monoclonal antibody blocking Fc receptor function. J. Immunol. 130,2133-2141. Perussia, B., Trinchieri, G., Jackson, A., Warner, J. N. L., Faust, J., Rumpold, H., Kraft, D., and Lanier, L. L. (1984). The Fc receptor for IgG on human natural killer cells: Phenotypic, functional, and comparative studies with monoclonal antibodies.]. Immunol. 133, 180-189. Perussia, B., M. Kobayashi, M., Rossi, M. E., Anegon, I., and Trinchieri, G. (1987). Immune interferon enhances functional properties of human granulocyte: Role of Fc receptors and effect of lymphotoxin, tumor necrosis factor and granulocytemacrophage colony-stimulating factor. J . Immunol. 138, 765-774. Pfefferkorn, L. C., and Fanger, M. W. (1989). Cross-linking of the high &nity Fc receptor for human immunoglobulin G1 triggers transient activation of NADPH oxidase activity. J . Biol. Chem. 264, 14112-14120. Pfeiffer, J. R., and Oliver, J. M. (1994). Tyrosine kinase-dependent assembly of actin plaques linking FceR cross-linking to increased cell substrate adhesion in RBL-2H3 tumor mast cells. J . Immunol. 152,270-279. Pfeiffer, J. R., Seagrave, J. C., Davis, B. H., Deanin, G. G., and Oliver, J. M. (1985). Membrane and cytoskeletal changes associated with IgE-mediated serotonin release from rat basophilic leukemia cells. J . Cell Bid. 101,2145-2155. Phillips, J. O., Everson, M. P., Moldoveanu, Z., Lue, C., and Mestecky, J. (1990). Synergistic effect of IL-4 and IFN-y on the expression of polymeric Ig receptor (secretory componant) and IgA binding by human epithelial cells. J . Immunol. 145, 1740-1744.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

117

Phillips, J. G., and Babcock, G. F. (1983). NKP-15: A monoclonal antibody reactive against purified human natural killer cells and granulocytes. Immunol. Lett. 6, 143-149. Phillips, J. H., Chang, C., and Lanier, L. L. (1991). Platelet-induced expression of FcyRIII (CD16) on human monocytes. Eur. J . Immonol. 21,895-899. Phillips, N. E., and Parker, D. C. (1983). Fc-dependent inhibition of mouse B cell activation by whole anti-p antibodies. J . Immunol. 130,602-606. Pichler, W. J., and Knapp, W. (1977). Receptors for IgM-coated erythrocytes on chronic lymphocytic leukemia cells. J. Immunol. 118, 1010-1015. Pignata, C., Prasad, K. V. S., Robertson, M. J., Levine, H., Rudd, C. E., and Ritz, J. (1993). FcyRIIIA-mediated signaling involves src-family lck in human natural killer cells. J. Immunol. 151,6794-6800. Pilkington, G. R., Lee, G. T. H., Michael, P. M.,Carson, 0. M., Kraft, N., Atkins, R. C., and Jose, D. G. (1986).Subclassificationof leukemia using monoclonal antibodies. In “Leukocyte Typing 11” (E. L. Reinherz, et al., eds.), p. 353. Springer-Verlag, New York. Plaut, M.,Pierce, J. H., Watson, C. J., Hanley-Hyde, J., Nordan, R. P., and Paul, W. E. (1989). Mast cell lines produce lymphokines in response to cross-linkage of FcsRI or to calcium ionophores. Nature (London) 339,64-67. Pleiman, C. M., Abrams, C., Timson-Gamen, L., Beedzyk, W., Jongstra, J., Shaw, A. S., and Cambier, J. C. (1994).Distinct p53/56lUn and p59fundomains associate with nonphosphorylated and phosphorylated Ig-a. (in press). Pochon, S., Graber, P., Yeager, M., Jansen, K., Bernard, A. R., Aubrey, J-P., and Bonnefoy, J-Y. (1992). Demonstration of a second ligand for the low affinity receptor for immunoglobulin E (CD23) using recombinant CD23 reconstituted into fluorescent liposomes. J . Exp. Med. 176,389-397. Porges, A. J., Redecha, P. B., Doebele, R., Pan, L. C., Salmon, J. E., and Kimberly, R. P. (1992). Novel Fcy receptor I family gene products in human mononuclear cells. J . Clin. Invest. 90,2101-2109. Pound, J. D., and Walker, M. R. (1990). Membrane Fc receptors for IgG subclasses. I n “The Human IgG Subclasses: Molecular Analysis of Structure, Function and Regulation” (F. Shakib, ed.), Chap. 6, pp. 111-133. Pergamon, Oxford. Pricop, L., Galatiuc, C., Manciulea, M., DeLeo, A., Sulica, A., Herberman, R. B., and Whiteside, T. L. (1991). Expression of Fcp receptors on human natural killer cells. Clin. Immunol. Immunopathol. 59,355-367. Pricop, L., Rabinowich, H., Morel, P. A., Sulica, A., Whiteside, T. L., and Herberman, R. B. (1993).Characterization of the Fcp Receptor on human killer cells. J . Immunol. 151,3018-3029. Prins, J-B., Todd, J. A,, Rodrigues, N. R., Ghosh, S., Hogarth, P. M., Wicker, L. S., Gaffrey, E., Podolin, P. L., Fischer, P. A,, Sirotina, A., and Peterson, L. B. (1993). Linkage on chromosome 3 of autoimmune diabetes and defective Fc receptor for IgG in NOD mice. Science 260,695-698. Pulford, K., Ralfkiaer, E., Macdonald, S. M., Erber, W. N., Falini, B., Gatter, K. C., and Mason, D. Y. (1986). A new monoclonal antibody (KB61) recognising a novel antigen which is selectively expressed on a subpopulation of human B lymphocytes. Immunology 57,71-76. Qiu, W. Q., De Bruin, D., Brownstein, B. H., Pearse, R., and Ravetch, J. V. (1990). Organization of the human and mouse low-affinity FcyR genes: Duplication and recombination. Science 248, 732-735. Quarto, R., and Metzger, H. (1986). The receptor for immunoglobulin E: Examination for kinase activity as a substrate for kinases. Mol. Immunol. 23, 1215-1223.

118

MARK D. HULETT AND P. MARK HOGARTH

Quilliam, A. L., Osman, N., McKenzie, I. F. C., and Hogarth, P. M. (1993). Biochemical characterisation of murine FcyRI. Immunology 78,358363, Ra, C., Jouvin, M-H., Blank, U., and Kinet, J-P. (1989a). A macrophage Fcy receptor and the mast cell receptor for immunglobulin E share an identical subunit. Nature (London)341,752-754. Ra, C., Jouvin, M-H. E., and Kinet, J-P. (1989b). Complete structure of the mouse mast cell receptor for IgE (FcERI)and surface expression of chimeric receptors (rat-mousehuman) on transfected cells. J. Biol. Chem. 264, 15323-15327. Raghavan, M., Gastinel, L. M., and Bjorkman, P. J. (1993).The class I major histocompatibility complex related Fc receptor shows pH-dependent stability differences correlating with immunoglobulin binding and release. Biochemistry 32,8654-8660. Rankin, B. M., Yocum, S. A., Mittler, R. S., and Kiener, P. A. (1993). Stimulation of tyrosine phosphorylation and calcium mobilization by Fcy receptor cross-linking: Regulation by the phosphotyrosine phosphatase CD45.j. Zmmunol. 150, 605-616. Rappaport, E. F., Cassel, D. L., Walterhouse, D. O., McKenzie, S. E., Surrey, S., Keller, M. A., Scrieber, A.D., and Schwartz, E. (1993).A soluble form ofthe human Fc receptor FcyRIIA: Cloning, transcript analysis and detection. E x p . Hematol. 21, 689-696. Ravetch, J. V., and Anderson, C. L. (1991). FcyR family: Proteins, transcripts and genes. In “Fc Receptors and the Action ofdntibodies,” (H. Metzger, ed.), chap. VI. Am. SOC.Microbiol., Washington, DC. Ravetch, J. V., and Kinet, J-P. (1991). Fc Receptors. Annu. Rev. Zmmunol. 9,457-492. Ravetch, J. V., and Perussia, B. (1989). Alternative membrane forms of FcyRIII (CD16) on human NK cells and neutrophils: Cell-type specific expression of two genes which differ in single nucleotide substitutions. J. Exp. Med. 170,481-491. Ravetch, J. V., Luster, A. D., Weinshank, J. K., Kochan, J., Pavlovec, A., Portnoy, D. A., Hulmes, J,, Pan, Y-C. E., and Unkeless, J. C. (1986). Structural heterogeneity and functional domains of murine immunoglobulin G Fc receptors. Science 234,718-725. Repp, R., Valerius, T., Sendler, A., Gramatzki, M., Iro, H., Kalden, J. R., and Platzer, E. (1991). Neutrophils express the high-affinity receptor for IgG (FcyRI, CD64) after in vivo application of recombinant human granulocyte colony-stimulating factor. Blood 78,885-889. Reth, M. (1989). Antigen receptor tail clue. Nature (London)338, 383-384. Reth, M., Hombach, J., Wienards, J., Cambell, K. S., Chien, N., Justement, L. B., and Cambier, J. C. (1991). The B cell antigen receptor complex. Zmmunol. Today 12, 196-201. Richards, M. L., Katz, H., and Liu, F. T. (1991). Complete genomic sequence of the murine low affinity Fc receptor for IgE. J. Zmmuunol. 147, 1067-1074. Riske, F., Hakimi, J., Mallamaci, M., Griffin, M., Pilson, B., Tobkes, N., Lin, P., Danho, W., Kochan, J., and Chizzonite, R. (1991). High a f k i t y human IgE receptor (FcERI): Analysis of functional domains of the a-subunit with monoclonal antibodies. j . Biol. Chem. 266,11245-11251. Rivnay, B., Wank, S. A., Poy, G., and Metzger, H. (1982). Phospholipids stabilize the interaction between the alpha and beta subunits of the solubilized receptor for immunoglobulin E. Biochemistry 21,6922-6927. Roa, T. D., Maghazachi, A. A., Gonzalez, A. V., and Phillips-Quagliata, J. M. (1992). A novel IgA receptor expressed on amurine B cell 1ymphoma.J. Zmmunol. 149,143-153. Roberts, D. M., Guenthert, M., and Rodewald, R. (1990). Isolation and characterisation of the Fc receptor from the fetal yolk sac of the rat. J. Cell B i d . 111, 1867-1876. Robertson, M. W. (1993). Phage and Escherichia coli expression of the human high affinity immunoglobulin E receptor a-subunit ectodomain. j . B i d . Chem. 268, 12736-12743.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

119

Robertson, M. W., and Liu, F-T. (1988). IgE structure-function relationships defined by sequence directed antibodies induced by synthetic peptides. Mol. Zmmunol. 25, 103-113. Rodewald, H-R., Awad, K., Moingeon, P., D’Adamio, L., Rabinowitz, D., Shinkai, Y., Alt, F. W., and Reinherz, E. L. (1993). FcyRIIIIII and CD2 expression mark distinct subpopulations of immature CD4- CD8-murine thymocytes: In vivo developemental kinetics and T cell receptor /3 chain rearrangement status.J. Exp. Med. 177,1079-1092. Romeo, C., and Seed, B. (1991). Cellular immunity to HIV activated CD4 fused to T cell or Fc receptor polypeptides. Cell (Cambridge, Mass.) 64, 1037-1046. Romeo, C., Amiot, M., and Seed, B. (1992). Sequence requirements for induction of cytolysis by the T cell antigen/Fc receptor 5 chain. Cell (Cambridge, Mass.) 68, 889-897. Rosales, C., and Brown, E. J. (1992).Calcium channel blockers nifedipine and diltiazem inhibit Ca2+ release from intracellular stores in neutrophils. J . Biol. Chem. 25, 1443-1448. Rosenfeld, S. I., and Anderson, C. L. (1989).Fc receptors on human platelets. In “Platelet Immunology” (T. J. Kunicki and J. N. George, eds.), pp. 337-353. Lippincott, Philadelphia. Rosenfeld, S. I., Looney, R. J., Leddy, J. P., Phipps, D. C., Abraham, G. N., and Anderson, C. L. (1985). Human platelet Fc receptor for immunoglobulin G: Identification as a 40,000 molecular weight membrane protein shared by monocytes./. Clin. Inuest. 76, 2317-2322. Rosenfeld, S. I., Ryan, D. H., Looney, R. J., Anderson, C. L., Abraham, G. N., and Leddy, J. P. (1987). Human Fcy receptors stable inter-donor variation in quantitative expression on platelets correlates with functional responses. J . Immunol. 138, 2869-2873. Rosseaux-Prevost, R., Rosseaux, J,, Bazin, H., and Bisert, G. (1984).Differential reduction of the inter-chain disulphide bond of rat immunoglobulin E: Relation to biological activity. Mol. Zmmunol. 21,233-241. Rudders, R. A., and Anderson, J. (1982). IgD-Fc receptors on normal and neoplastic human B cell lymphocytes. Clin. E x p . Immunol. 50,579-586. Rudders, R. A., Andersen, J., Howard, J. P., and Freid, R. (1980).The binding of IgM to B lymphocytes. I. Studies of the binding of aggregated IgM to normal and leukemic human B lymphocytes. Clin. Zmmunol. Zmmunopathol. 15,673-686. Rumpold, H., Kraft, D., Obexer, G., Bock, G., and Gebhart, W. (1982). A monoclonal antibody against a surface antigen shared by human large granular lymphocytes and granulocytes. J . Immunol. 129, 1458-1464. Ryu, S-E., Kwong, P. D., Truneh, A., Porter, T. G., Arthos, J., Rosenberg, M., Dai, X., Xuong, N-H., Axel, R., Sweet, R. W., and Hendrikson, W. A. (1990).Crystal structure of an HIV-binding recombinant fragment of CD4. Nature (London)348,418-426. Saito, H., Ishizaka, K., and Ishizaka, T. (1989). Effect of nonhydrolizable guanine phosphate on IgE-mediated activation of phosphilipase C and histamine release from rodent mast cells. J. Zmmunol. 143,250-258. Salcedo, T. W., Kurosaki, T., Kanakaraj, P., Ravetch, J., and Perussia, R. (1993).Physical and functional association of ~56“’with FcyRIIIA (CD16) in natural killer cells. /. E x p . Med. 177,1475-1480. Salmon, J. E., Kapur, S., and Kimberly, R. P. (1987). Opsonin-independent ligation of Fc gamma receptors: The 3G8-bearing receptors on neutrophils mediate the phagocytosis of concanavalin A-treated erythrocytes and nonopsonized Escherichia coli. J . E x p . Med. 166,1798-1813. Salmon, J. E., Edberg, J. C., Kimberly, R. P., Mensa, E., and Ryan, R. (1990).Fcy receptor

120

MARK D. HULETT AND P. MARKHOGARTH

I11 on human neutrophils: Allelic variants have functionally distinct capabilities. J. Clin. Invest. 85,1287-1295. Salmon, J. E., Brogle, N., Brownlie, C., Edberg, J. C., Kimberly, R. P., Chen, B-X., and Erlanger, B. F. (1993). Human mononuclear phagocytes express adenosine A 1 receptors: A novel mechanism for differential regulation of Fcy receptor function. J. Immunol. 151,2775-2785. Sammartino, L., Webber, L. M., Hogarth, P. M., McKenzie, I. F. C., and Garson, 0. M. (1988). Assignment of the genes coding for human FcyRII (CD32) to band q23q24 on chromosome I. Immunogenetics 28,380-381. Sandford, A. J., Shirakawa, T., Moffat, M. F., Daniels, S. E., Ra, C., Faux, J. A., Young, R. P., Nakamura, Y., Lathrop, G. M., Cookson, W. 0. C. M., and Hopkin, J. M. (1993). Localisation of atopy and fl subunit of high-affinity IgE receptor (FcyRI) on chromosome l l q . Lancet 341,332-334. Sandor, M., and Lynch, R. G. (1992). Lymphocyte Fc receptors: The special case of T cells. Immunol. Today 14,227-231. Sarmay, G., Jefferies, R., Klein, E., Benczur, M., and Gergely, J. (1985). Mapping of the functional topography of IgG Fc y with monoclonal antibodies: Localization of epitopes interacting with the binding sites of Fc receptor on human K cells. Eur. J. Immunol. 15, 1037-1042. Sarmay, G., Pecht, I., and Gergely, J. (1990). Phosphorylation of type I1 Fcy receptor on activated human B lymphocytes. Int. Immunol. 12, 1235-1243. Sarmay, G., Rozsnyay, Z., Szabo, I., Biro, A., and Gergely, J. (1991).Modulation of type I1 Fcy receptor expression on activated human B lymphocytes. Eur. J . Immunol. 21, 541-549. Sarmay, G., Pecht, I., and Gergely, J. (1994). Protein tyrosine kinase activity tightly associated with human type 11 Fcy receptors. Proc. Natl. Acad. Sci. 91, 4140-4144. Scallon, B. J., Scigliano, E., Freedman, V. H., Miedel, M. C., Pan, Y-C. E., Unkeless, J. C., and Kochan, J. P. (1989). A human immunoglobulin G receptor exists in both polypeptide-anchored and phosphoinositol-glycan-anchoredforms. Proc. Natl. Acad. Sci. U.S.A. 86, 5079-5083. Schaller, M. D., Borgman, C. A., Cobb, B. S., Vines, R. R., Reynolds, A. B., and Parsons, J. T. (1992). p ~ 1 2 5 ~ a* structurally ~, distinctive protein-tyrosine kinase associated with focal adhesions. Proc. Natl. Acad. Sci. U.S.A. 89,5129-5133. Schmitt, D. A., Hanau, D., Bieber, T., Dezutter-Dambuyant, C., Schmitt, D., Fabre, M., Pauly, G., and Cazenave, J-P. (1990). Human epidermal Langerhans cells express only the 40-kilodalton Fcy receptor (FcRII).J. Immunol. 144,4284-4290. Schneider, H. A., Cohen-Dayag, A., and Pecht, I. (1992). Tyrosine phosphorylation of phospholipase Cyl complexes FCEreceptor mediated signal to mast cell secretion. Int. Immunol. 4,447. Scholl, P. R., Ahern, D., and Geha, R. S. (1992).Protein tyrosinephosphorylation induced via the IgG receptors FcyRI and FcyRII in the human monocytic cell line THP-1. J. Immunol. 149, 1751-1757. Schoneich, J. T., Wilkinson, V. L., Kado-Fong, H., Presky, D. H., and Kochan, J. P. (1992).Association of the human Fc epsilon RI gamma subunit with novel cell surface polypeptides. J. Immunol. 148,2181-2185. Schwarzbaum, S., Nissim, A., Alkalay, I., Ghozi, M. C., Schindler, D. G., Bergman, Y., and Eshhar, Z. (1989).Mapping ofmurine IgE epitopes involved in IgE Fcs receptor interactions. J. Immunol. 19, 1015-1023. Seagrave, J., and Oliver, J. M. (1990).Antigen-dependent transition of IgE to a detergentinsoluble form is associated with reduced IgE- receptor-dependent secretion from RBL-2H3 mast cells. J. Cell. Physiol. 144, 128-136.

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

12 1

Seagrave, J., Pfeiffer, J. R.,Wofsy, C., and Oliver, J. M. (1991). Relationship of IgE receptor topography to secretion in RBL-2H3 mast cells. 1.Cell Physiol. 148, 139151. Sears, D. W., Osman, N., Tate, B., McKenzie, I. F. C., and Hogarth, P. M. (1990). Molecular cloning and expression of the mouse high A n i t y Fc receptor for IgG. 1.Immunol. 144,371-378. Sedmak, D. H, Davis, H. H., Singh, U., van de Winkel, J. G. J., and Anderson, C. L. (1991). Distribution of IgG Fc receptor antigens in placenta and endothelial cells in man: An immunohistochemical study. Am. 1.Pathol. 138,175-181. Sefton, B. M., and Campbell, M. A. (1991).The role oftyrosine protein phosphorylation in lymphocyte activation. Annu. Rev. Cell Biol. 7,257-274. Segal, D. M., and Snider, D. P. (1989).Targetingand activation ofcytotoxic lymphocytes. Chem. Immunol. 47,179-213. Segal, D. M., Taurog, J. D., and Metzger, H. (1977). Dimeric immunoglobulin E serves as a unit signal for mast cell degranulation. Proc. Natl. Acad. Sci. U S A . 74,2993-2997. Seki, T. (1989). Identification of multiple isoforms of the low-Anity human IgG Fc receptor. lmmunogenetica 30,5-12. Selvaraj, P., Rosse, W. F., Silber, R.,and Springer, T. A. (1988).The major Fc receptor in blood has a phosphatidylinositol anchor and is deficient in paroxymal nocturnal haemoglobinuria. Nature (London)333,564-567. Selvaraj, P., Carpen, O., Hibbs, M. L., and Springer, T. A. (1989). Natural killer cell and granulocyte Fcy receptor I11 (CD16) differ in membrane anchor and signal transduction. ]. Immunol. 143,3283-3288. Shen, F. W., and Boyse, E. A. (1980). An alloantigen selective for B cells Ly17.1. Immunogenetics 11,315-317. Shen, L., Guyre, P. M., Anderson C. L., and Fanger, M. W. (1986). Heteroantibodymediated cytotoxicity: Antibody to the high &nity Fc receptor for IgG mediates cytotoxicity by human monocytes that is enhanced by interferon? and is not blocked by human IgG. 1.Immunol. 137,3378-3382. Shen, L., Guyre, P. M., and Fanger, M. W. (1987).Polymorphonuclear leukocyte function triggered through the high affinity Fc receptor for monomeric IgG.]. Immunol. 139, 534-538. Shen, L., Lasser, R.,and Fanger, M. W. (1989).My43, a monoclonal antibody that reacts with human myeloid cells inhibits monocyte IgA binding and triggers function. 1.Immunol. 143,4117-4122. Shimizu, A., Tepler, I., Bemfey, E. H., Berenstein, R. P., and Leder, P. (1988). Human and rat mast cell high affinity immunoglobulin E receptor: Characterization of putative a chain gene products. Proc. Natl. Acad. Sci. U.S.A.85, 1907-1911. Shopes, B., Weetall, M., Holowka, D., and Baird, B. (1990).Recombinant human IgGlmurine IgE chimeric Ig: Construction, expression and binding to human Fcy receptors. 1.Immunol. 145, 3842-3848. Simister, N. E., and Rees, A. R. (1985). Isolation and characterisation of a Fc receptor from neonatal rat small intestine. Eur. 1.Immunol. 15, 733-738. Simister, N. E., and Mostov, K. E. (1989). An Fc receptor structurally related to MHC Class I antigen. Nature (London)337, 184-187. Simmons, D., and Seed, B. (1988). The Fcy receptor of natural killer cells is a phospholipid-linked membrane protein. Nature (London)333,568-570. Simms, H. H., Gaither, T. A., Fries, L. F., and Frank, M. M. (1991).Monokines released during short term Fc/ receptor phagocytosis up-regulate polymorphonuclear leukocytes and mono-phagocytis function. J . Immunol. 147,265-272. Siraganian, R. P., Hook, W. A., and Levine, B. B. (1975). Specific in vitro histamine

122

MARK D. HULETT AND P. MARKHOGARTH

release from basophils by bivalent haptens: Evidence for activation by simple bridging of membrane bound antibodies. lmmunochemistry 12,149-157. Sivo, J., Politis, A. D., and Vogel, S. N. (1993). Differential effects of interferon-y and glucocorticoids on FcyR gene expression on murine macrophages. 1.Leuk. Biol. 54, 451-457. Sjoberg, 0. (1980a). Presence of receptors for IgA on human T and non-T lymphocytes. Eur. J . lmmunol. 10,226-228. Sjoberg, 0.(1980b). Presence of receptors for IgD on human T and non-T lymphocyes. Scund.]. Immunol. 11,3842-3848. Snapper, C. M., Hooley, J. J., Urban, J. F., Jr., and Finkelman, F. D. (1991). Lack of FceRII expression by murine B cells after in vivo immunization is directly associated with Ig secretion and not isotype switching.]. lmmunol. 146,2161-2168. Sollid, L. M., Kvale, D., Brandtzaeg, P., Markussen, G., and Thorsby, E. (1987). Interferon-y enhances expression of secretory component, the epithelial receptor for polymeric Ig. 1.lmmunol. 138,4303-4306. Sonyang, Z., Shoelson, S. E., Chaudhuri, T., Gush, G., Pawson, T., Hasert, W., King, G., Roberts, T., Ratnofshy, S., Lechneide, R., Neel, B.G., Birge, R. B., Fajardo, J. E., Chou, M., Hanafusa, H., Schaffhausan, B., and Chantely, L. I. (1993). SH-2 domains recognise specific phosphopeptide sequences. Cell (Cumbridge,Muss.) 72,767-778. Spiegelberg, H. L. (1984).Structure and function of Fc receptors for IgE on lymphocytes, monocytes and macrophages. Ado. lmmunol. 35,6148. Stamenkovic, I., and Seed, B. (1988). Analysis of two cDNA clones encoding the B lymphocyte antigen CD20 (Bl, Bp35), a type 111 integral membrane protein.]. E x p . Med. 167, 1975-1980. Stanworth, D. R., Humphrey, J. H., Bennich, H., and Johansson, S. G. 0. (1968).Inhibition of Prausnitz-Kustner reaction by proteolytic cleavage fragments of a human myeloma protein Gf imiimnoglebulin class E. Lancet. 11, 17-18. Stefanova, I., Horejsi, V., Ansotegui, I. J., Knapp, W., and Stockinger, H. (1991). GPIanchored cell-surface molecules complexed to protein tyrosine kinases. Science 245, 1016-1019. Steinmetz, M., Winoto, A., Minard, K., and Hood, L. (1982).Clusters of genes encoding mouse transplantation antigens. Cell (Cumbridge, Muss.)28,489-498. Stengelin, S., Stamenkovic, I., and Seed, B. (1988). Isolation of cDNAs for two distinct human Fc receptors by ligand affinity cloning. EMBOJ. 7, 1053-1059. Stephan, V., Benhamou, M., Gutkind, J. S., Robbins, K. C., and Siraganian, R. P. (1992). FceRI-induced protein tyrosine phosphorylation of pp72 in rat basophilic leukemia cells (RBL-2H3).J. Biol. Chem. 267, 5434-5441. Stewart, W. W., and Ken, M. A. (1990). The specificity of the human neutrophil IgA receptor (FcaR) determined by measurement of chemiluminescence induced by serum or secretory IgAl or IgA2. lmmunology 71,328-334. Stuart, S . G., Troustine, M., Vaux, D., Koch, T., Martens, C.,Mellman, I., and Moore, K. W. (1987). Isolation and expression of cDNA clones encoding a human receptor for IgG (FcyRII). J. E r p . Med. 166, 1668-1684. Stuart, S. G., Simister, N. E., Clarkson, S. B., Shapino, M., and Mellman, I. (1989).The low affinity Fc receptor for human IgG (hFcyRII) exists as multiple isoforms in macrophages, lymphocytes and IgG-transporting placental epithelium. EMBO J. 8, 3657-3666. Suter, U., Bastos, R., and Hofstetter, H. (1987).Molecular structure ofthe gene and the 5’ flanking region of the human lymphocyte immunoglobulin E receptor. Nucleic Acids Res. 15, 7295.

MOLECULAR BASIS O F Fc RECEPTOR FUNCTION

123

Sutton, B. J., and Gould, H. J. (1993). The human IgE network. Nature (London) 366, 42 1-428. Suzuki, T. (1991). Signal transduction mechanisms through Fcy receptors on the mouse macrophage surface. FASEB J. 5, 187-193. Takai, T., Li, M., Sylvestre, D., Clynes, R., and Ravetch, J. V. (1994).Fcy chain deletion results in pleiotrophic effector cell defects. Cell (Cambridge, Mass.) 76, 519-529. Takizawa, F., Adamczewski, M., and Kinet J-P. (1992). Identification ofthe low affinity receptor for immunoglobulin E on mouse mast cells and macrophages as FcyRII and FcyRIII. J. E x p . Med. 176, 469-475. Tamma, S. M. L., and Coico, R. F. (1991). Release of human IgD-R+ T cells. FASEB J. 5, A1718. Tamma, S. M. L., Amin, A. R., Finkelman, F. D., Chen, Y-W., Thorbecke, G. J., and Coico, R. F. (1991). IgD receptors on murine T-helper cells bind to Fd and Fc regions of immunoglobulin D. Proc. Natl. Acad. Sci. U.S.A. 88,9233-9237. Tartour, E., de la Salle, H., de la Salle C., Teillaud, C., Camoin, L., Galinha, A., Latour, S., Hanau, D., Fridman, W. F., and Sautes C. (1993). Identification, in mouse macrophages and in serum, of a soluble receptor for the Fc portion of IgG (FcyR) encoded by an alternatively spliced transcript of the FcyRII gene. Int. Immunol. 5,859-868. Tate, B. J., Witort, E., McKenzie, I. F. C., and Hogarth P. M. (1992). Expression of the high responderhon-responder human FcyRII: Analysis by PCR and transfection into FcR-COS cells. Immunol. Cell Biol. 70, 79-87. Tax, W. J. M., and van de Winkel, J. G. J. (1990). Human Fcy receptor 11: A standby receptor activated by proteolysis? Immunol. Today 11,308-310. Tax, W. J. M., Willems, H. W., Reekers, P. P. M., Capel, P. J. A,, and Koene, R. A. P. (1983). Polymorphism in mitogenic effect of IgGl monoclonal antibodies against T3 antigen on human T cells. Nature (London)304,445-447. Tax, W. J. M., Hermes, F. F. M., Willems, R. W., Capel, P. J. A,, and Koene, R. A. (1984). Fc receptors for mouse IgGl on human monocytes, polymorphism and role in antibody induced T cell proliferation. J . Immunol. 133, 1185-1189. Teillaud, J-L., Diamond, B., Pollock, R. R., Fajtova, V., and Scharff, M. D. (1985). Fc receptors on cultured myeloma and hybridoma cells. J . Immunol. 134, 1774-1779. Tepler, I., Shimizu, A., and Leder, P. (1989). The gene for the rat mast cell high a n i t y IgE receptor a chain: Structure and alternate mRNA splicing patterns.J. Biol. Chem. 264,5912-5915. Tetteroo, P. A. T., van der Scoot, C. E., Visser, F. J., Bos, M. J. E., and Kr. von dem Borne, A. E. G. (1988). Three different types of Fcy receptors on human leucocytes defined by workshop antibodies: FcyRlow of neutrophils, FcyRlow of KINK lymphocytes and FcyRII. In “Leukocyte Typing 111.” Vol. 3. “Human Myeloid and Hematopoetic Cells” (A. McMichael, ed.), pp. 702-706. Oxford Univ. Press, Oxford. Te Velde, A. A., Huijbens, J. F., De Vries J. E., and Figdor, C. G. (1990). IL-4 decreases FcyR membrane expression and FcyR-mediated cytotoxic activity of human monocytes. ]. Immunol. 144,3046-3051. Te Velde, A. A., de waal Malefijt, R., Huijbens, R. J., de Vries, J. E., and Figdor, C. G. (1992). IL-10 stimulates monocyte FcyR surface expression and cytotoxic activity: Distinct regulation of antibody-dependent cellular cytotoxicity by IFN-7, IL-4, and IL-10. J . Immunol. 149,4048-4052. Ting, A. T., Karnitz, L. M., Schoon, R. A., Abraham, R. T., and Leibson, P. J. (1992). Fc gamma receptor induces the tyrosine phosphorylation of both phospholipase C (PLC)-gamma 1 and PLC-gamma 2 in natural killer cel1s.J. Erp. Med. 176,1751-1755. Tosi, M. F., and Berger, M. (1988). Functional differences between the 40 kDa and 50

124

MARK D. HULETT AND P. MARKHOGARTH

to 70 kDa IgG Fc receptors on human neutrophils revealed by elastase treatment and antireceptors antibodies. J . Immunol. 141,2097-2103. Trezzini, C., Jungi, T. W., Spycher, M. O., Maly F. E., and Rao, P. (1990) Human monocyte CD36 and CD16 are signalling molecules. Immunology 71,29-37. Trounstine, M. L., Peltz, G. A,, Yssei, H., Huizinga, T. W. J., Kr. von dem Borne, A. E. G., Spits, H., and Moore, K. W. (1990). Reactivity of cloned, expressed human FcyRIII isoforms with monoclonal antibodies which distinguish cell-type-specific and allelic forms of FcyRIII. I n t . Immunol. 2, 303-310. Tuijnman, W. R., Capel, P. J., and van de Winkel, J. G. (1992). Human low-affinity IgG receptor Fc gamma RIIa (CD32) introduced into mouse fibroblasts mediates phagocytosis of sensitized erythrocytes. Blood 79,1651-1656. Ukkonen, P., Lewis, V., March, M., Helenius, A., and Mellman, 1. (1986).Transport of macrophage Fc receptors and Fc receptor-bound ligands to lysosomes. J . Exp. Med. 163,952-971. Underdown, B. J., Switzer, I., and Jackson, G. D. (1992). Rat secretory component binds poorly to rodent IgM. J . Immunol. 149,487-491. Unkeless, J. C. (1977). The presence of two Fc receptors on mouse macrophages: Evidence from a variant cell line and differential trypsin sensitivity.J. Exp. Med. 145, 931-947. Unkeless, J. C. (1979).Characterisation of a monoclonal antibody directed against mouse macrophage and lymphocyte Fc receptors. J . Exp. Med. 150,580-596. Unkeless, J. C., and Eisen H. N. (1975). Binding of monomeric immunoglobulins to Fc receptors of mouse macrophages. J . E x p . Med. 142,1520-1533. Unkeless, J. C., Kaplan, G., Plutner, H., and Cohn, Z. A. (1979). Fc-receptor variants of a mouse macrophage cell line. Proc. Natl. Acad. Sci. U.S.A.76, 1400-1404. Unkeless, J. C., Fleit, H., and Mellman, I. (1981). Structural aspects and heterogeneity of immunoglobulin Fc receptors. Adu. Immunol. 31,247-270. Unkeless, J. C., Scigliano, E., and Freedman, V. H. (1988). Structure and function of human and murine receptors for IgG. Annu. Reu. Immunol. 6,251-281. Valent, P., Ahsman, L. K., Hinterberger, W., Eckersberger, F., Majdic, O., Lechner, K., and Bettelhiem, P. (1989).Mast cell typing: Demonstration of a distinct hematopoeitic cell type and evidence for immunophenotypic relationship to mononuclear phagocytes. Blood 73, 1778-1785. Valerius, T. R., Repp, Kalden, J. R., and Platzer., E. (1990). Effects of IFN on human eosinophils in comparison with other cytokines. J. Immunol. 145,2950-2958. Vance, B. A., Huizinga, T. W. J., and Guyre, P. M. (1992). Functional polymorphism of FqRIII on human LGLlNK cells. FASEB J. 6,1620A. Van Den Herik-Oudijk, I. E., Westerdaal, N. A. C., Henriquez, N. V., Capel, P. J. A., and Van De Winkel, J. G. J. (1994). Functional analysis of human FcyRII (CD32) isoforms expressed in B lymphocytes. J . Immunol. 152,574-585. van de Winkel, J. G. J., and Anderson, C. L. (1991).Biology ofhuman IgG Fc receptors. J . Leuk. Biol. 49,511-524. van de Winkel, J. C. J., and Capel, P. J. A. (1993).Human IgG Fc receptor heterogeniety: Molecular aspects and clinical applications. Immunol. Today 14,215-221. van de Winkel, J. G. J,, Tax, W. J. M., van Bruggen, M. J. C, Van Roozendaal, C. E. P., Willems, H.W., Vlug, A., Capel, P. J. A., and Koene, R. A. P. (1987). Characterisation of two Fc receptors for mouse immunoglobulins on human monocytes and cell lines. Scand. J . Immunol. 26,663-672. van de Winkel, J. G. J., Van Ommen, T. W. J., Huizinga, M. A. H. V. M., De Raad, W. B., Tuijnman, P. J. T. A., Groenen, P. J. A., Capel, R. A. Koene, P., and Tax,

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

125

W. J. M. (1989). Proteolysis induces increased binding affinity of the monocyte type I1 FcR for human IgG. J. Immunol. 143,571-578. van de Winkel, J. G. J., Tax, W. J. M., Jacobs, C. W. M., Huizinga, T.W. J., and Willems, P. H. G. M. (1990a). Cross-linking of both types of IgG receptors, FcyRI and FcyRII, enhances intracellular free Ca2+in the monocytic cell line U937. Scand.J . Immunol. 31,315-325. van de Winkel, J. G. J., Jansze, M., and Capel, P. J. A. (1990b). Effects of protease inhibitors on human monocyte IgG Fc receptor 11.J. Immunol. 145, 1890-1896. van de Winkel, J. G. J., Emst, L. K., Anderson, C. L., and Chiu, I-M (1991). Gene organization of the human high affinity receptor for IgG, FcyRI (CD64):Characterization and evidence for a second gene. J. Biol. Chem. 266,13449-13455. Van Doren, K., Hanahan, D., and Gluzman, Y. (1984). Infection of eukaryotic cells by helper-independent recombinant adenovirus: Early region I is not obligatory for integration of viral DNA. J. Virol. 50,606. Varin-Blank, N., and Metzger H. (1990). Surface expression of mutated subunits of the high affinity mast cell receptor for IgE. J. Biol. Chem. 265, 15685-15694. Vaughn, M., Taylor, M., and Mohanakumar, T. (1985).Characterisation of human IgG Fc receptors. J. Immunol. 135,4059-4065. Vercelli, D., Helm, B., Marsh, E., Padlam, E., Geha, R. S., and Gould, H. (1989). The B cell binding site on human immunoglobulin E. Nature (London) 338,649-651. Vivier, E., Moroin, P., O'Brien, C., Druker, B., Schlossman, S. F., and Anderson, P. (1991). Tyrosine phosphorylation of the FcyRIII (CD16): 5 complex in human killer cells. J. Immunol. 146,206-210. Vivier, E., Rochet, N., Ackerly, M., Petrini, J., Levine, H., Daley, J., and Anderson, P. (1992). Signaling function of reconstituted CD16: zeta:gamma receptor complex isoforms. Int. Immunol. 4, 13-23. Walker, W. S. (1976). Seperate Fc receptors for immunoglobulins IgG2a and IgG2b on an established cell line of mouse macrophages. J. Immunol. 116,911-914. Walker, W. S. (1977). Mediation of macrophage cytolyic and phagocytic activities by antibodies of different classes and class-specific Fc-receptors. J. Immunol. 119, 367-373. Walker, B. A. M., Hagenlocker, B. E., Jr., Stubbs, E. B., Sandborg, R. R., Agranoff, B. W., and Ward, P. A. (1991). Signal transduction events and FcyR engagement in human neutrophils stimulated with immune complexes. J . Immunol. 146,735-741. Walker, M. R., Lund, J., Thompson, K. M., and Jefferies, R. (1989). Aglycosylation of human IgGl and IgG3 monoclonal antibodies can eliminate recognition by human cells expressing FcyRI and/or FcyRII receptors. Biochem. J. 259, 1347-1353. Wang, B., Rieger, A., Kilgus, O., Ochiai, K., Maurer, D., Fodinger, D., Kinet, J-P., and Stingl, G. (1992). Epidermal Langerhans cells from normal human skin bind monomeric IgE via FcsRI. J . Exp. Med. 175, 1353-1365. Wang, J., Yan, Y., Garret, T. P. J., Liu, J., Rodgers, D. W., Garlick, R. L., Tarr, G. E., Husain, Y., Reinhen, E. L., and Harrison, S. C. (1990).Atomic structure of a fragment of human CD4 containing two immunoglobulin-like domains. Nature (London) 348, 411-418. Warmerdam, P. A., van de Winkel, J. G. J., Grosselin, E. J., and Capel, P. J. A. (1990). Molecular basis for a polymorphism of human FcyRII (CD32). J. Exp. Med. 172, 19-25. Warmerdam, P. A. M, van de Winkel, J. G. J., Vlug, A., Westerdaal, N. A. C., and Capel, P. J. A. (1991). A single amino acid in the second Ig-like domain of the human Fcy receptor I1 is critical in human IgG2 binding. J . Immunol. 147, 1338-1343.

126

MARK D. HULETT AND P. MARKHOCARTH

Wamerdam, P. A., Nabben, N. M. J. M., Grad, S. A. R., van d e Winkel, J. G. J., a Capel, P. J. A. (1992). The human low-affinity IgG receptor IIC is a result of unequal crossover event. J. Biol. Chem. 268,7346-7349. Warmerdam, P. A. M., van den Herik-Oudijk, I. E., Parren, P. W., Westerdaal, N. i van de Winkel, J. G. J., and Capel, P. J. (1993). Interaction ofhuman FcgRIIbl (CD3 with murine and human IgG subclasses. Int. Immunol. 5,239-247. Wendel-Hansen, V., Riviere, M., Uno, M., Jannson, I., Szirer, J., Islam, M. Q., Lev G., Klien, G., Yodoi, J., and Rosen. A. (1990). The gene encoding CD23 leukocy 1 antigen (FcsRII) is located on human chromosome 19. Somat. Cell. Mol. Genet. 1 283. Weetall, M., Shopes, B., Holowka, D., and Baird, B. (1990). Mapping of the site 1 interaction between murine IgE and its high affinity receptor with chimeric 11 J . Immunol. 145,3849-3854. Weetall, M., Holowka, D., and Baird, B. (1993). Heterologous desensitization of th high affinity receptor for IgE (FcyR1) on RBL cells. J . Immunol. 150,4072-4083. Weinshank, R. L., Luster, A. D., and Ravetch, J. V. (1988). Function and regulation c a murine macrophage specific IgG Fc receptor, FcyRa.1. Exp. Med. 167,1909-1925 Weisbart, R. H., Kacena, A., Schuh, A., and Glode, D. W. (1988).GM-CSF induces humai neutrophil IgA-mediated phagocytosis by an IgA Fc receptor activation mechanism Nature (London)332,647-649. Weiss, A. (1994). T cell antigen receptor signal transduction. A tale of tails and cyto plasmic protein tyrosine kinases. Cell 73,209-212. Weissman, A. M., Hon, D., Orlaff, D. G., Modi, W. S., Seuanez, H., O’Brien, S. J., anc Klausner, R. D. (1988).Molecular cloning and characterization of the human T-celi receptor 6 chain: Distinctive from the CD3 complex. Proc. Natl. Acad. Sci. U.S.A. 85. 9709-9713. Welch, G. R., Wong, H. L., and Wahl, S. M. (1990). Selective induction of FcyRIII on human monocytes by transforming growth factor-p. ]. Immunol. 144,3444-3448. Werner, G., von dem Borne, A. E. G. Kr., Tromp, J. F., van der Plas-van Dalen, C. M., Visser, F. J., Engelfriet, C. P., and Tetteroo, P. A. T. (1988). Localization of the human NA1 alloantigen on neutrophil Fc-gamma receptors. In “Leukocyte Typing 111.” Vol. 3. “Human Myeloid and Hematopoetic Cells” (E. L. Reinhertz, B. F. Haynes, L. M. Nadler, and I. D. Bernstein, eds.), pp. 109-121. Springer-Verlag, New York. White, K. N., and Metzger, H. (1988). Translocation of protein kinase-C in rat basophilic leukemia cells induced by phorbol ester or by aggregation of IgE receptors.]. Immunol. 141, 942-947. Williams, A. F. (1985). Immunoglobulin-related domains for cell surface recognition. Nature (London) 314,579-580. Williams, A. F., and Barclay, A. N. (1987). A year in the life of the immunoglobulin superfamily. Immunol. Today 8,298-303. Williams, A. F., and Barclay, A. N. (1988). The immunoglobulin superfamily-domains for cell surface recognition. Annu. Reu. Immunol. 6, 381-405. Williams, A. F., Davis, S. J., He, Q., and Barclay, A. N. (1989). Structural diversity in domains of the immunoglobulin gene superfamily. Cold Spring Harbor Symp. Quant. Biol. 54,637-647. Willis, H. E., Browder, B., Feister, A. J., Mohanakumar, T., and Ruddy, S. (1988). Monoclonal antibody to human IgG Fc receptors: Crosslinking of receptors induces lysosomal enzyme release and superoxide generation by neutrophils. J . Immunol. 140, 234-239. Wilson, B. S . , Deanin, G. G., Standefer, J. C., Vandejaat, D., and Oliver J. M. (1989).

MOLECULAR BASIS OF Fc RECEPTOR FUNCTION

127

Depletion of guanine nucleotides with mycophenolic acid suppresses IgE receptormediated degradation in rat basophilic leukemia cells. ]. Immunol. 143, 259-265. Wilson, H. A., Greenblatt, D., Taylor, C. W., Putney, J. W., Tsien, R. Y., Finkelman, F. D., and Chused, T. M. (1987). The B lymphocyte calcium response to anti-Ig is diminished by membrane immunoglobulin cross-linkage to the Fcy receptor. J. Immunol. 138, 1712-1718. Wirthmueller, U., Kurosaki, T., Murakami, M. S., and Ravetch, J. V. (1992). Signal transduction by FcyRIII (CD16) is mediated through the y chain. J. E x p . Med. 175, 1381-1390. Wodnar-Filipowicz, A,, Heusser, C. H., and Moroni, C. (1989).Production ofthe haemopoetic growth factors GM-CSF and interleukin-3 by mast cells in response to IgE receptor mediated activation. Nature (London)339, 150-152. Wong, S., Freeman, J. D., Kelleher, C., Mager, D., and Takei, F. (1991).Ly-49 multigene family. J. Immunol. 147, 1417-1423. Woof, J. M., Nik Jaafar, M. I., Jefferies, R., and Burton, D. R. (1984). The monocyte binding domain(s) on human immunoglobulin G. Mol. Immunol. 21, 523-527. Woof, J. M., Partridge, L. J., Jefferies, R., and Burton, D. R. (1986). Localization of the monocyte-binding region on human immunoglobulin G. Mol. Immunol. 23,319-330. Ye, Z-S., Kinet, J-P, and Paul W. E. (1992).The gene for the a chain of the mouse high affinity receptor for IgE (FcsRI).J. Immunol. 149,897-900. Yokota, A., Kikutani, H., Tanaka, T., Sato, R., Barsumian, E. L., Suemura, M., and Kishimoto, T. (1988). Two species of human FCEreceptor I1 (FcsRUCD23): Tissue specific and 11-4 specific regulation of gene expression. Cell (Cambridge,Mass.) 55, 611-618. Young, J. D-E., Unkeless, J. C., Kaback, H. R., and Cohn, Z. A. (1983a). Macrophage membrane potential changes associated with y2blyl Fc receptor-ligand binding. Proc. Natl. Acad. Sci. USA. 80, 1357-1361. Young, J. D-E., Unkeless, J. C., Young, T. M., Mauro, A., and Cohn, Z. A. (1983b). Role for mouse macrophage IgG Fc receptor as ligand-dependent ion channel. Nature (London)306,186-189. Young, J. D-E., KO, S. S., and Cohn, Z. A. (1984). The increase in intracellular free calcium associated with IgG y2blyl Fc receptor-ligand interactions: Role in phagocytosis. Proc. Natl. Acad. Sci. U.S.A. 81,5430-5434. Zeger, D. L., Hogarth, P. M., and Sears, D. W. (1990).Characterization and expression of an Fcy receptor cDNA cloned from rat natural killer cells. Proc. Natl. Acad. Sci. U.S.A. 87,3425-3429. Zheng, Y., Shopes, B., Holowka, D., and Baird, B. (1991). Conformations of IgE bound to its receptor FcsRI and in solution. Biochemistry 30,9125-9132. Zheng, Y., Shopes, B., Holowka, D., and Baird, B. (1992). Dynamic conferences compared for IgE and IgGl in solution and bound to receptors. Biochemistry 31, 7446-7456. Ziljstra, M., Bix, M., Simister, N. E., Loring, J. M., Raulet, D. H., and Jaenisch, R. (1990). p2-microglobulin deficient mice lack CD4-8 + cytolytic T cells. Nature (London)344, 472-476. Note added in proof: A recent study using both chimeric and mutant receptors has identified domain 2 of human FcyRIII as important in the binding of IgG (Hibbs et al. (1994).J. Immunol. 152,4466-4474).

ADVANCES IN IMMUNOLOGY, VOL. 57

Fas and Fas Ligand: A Death Factor and Its Receptor SHIGEKAZU NAGATA Owka Biascience Institute, 6-2-4 F u d a i , Suihr, Osaka 565, Japan

1. Introduction

Homeostasis ofmammalian tissues is controlled not only by proliferation and differentiation of cells but also by cell death (Raff, 1992). There are two death processes, apoptosis and necroiis (Walker et al., 1988; Wyllie et al., 1980). The death of cells during embryogenesis, metamorphosis, endocrine-dependent tissue atrophy, and normal tissue turnover is called programmed cell death. Most of the programmed cell death which occurs during mammalian development proceeds by apoptosis. Apoptosis can be morphologically and biochemically distinguished from necrosis which occurs during pathological cell death as a result of injury, complement attack, severe hypoxia, hyperthermia, lytic viral infection, and exposure to a variety of toxins. Apoptosis is accompanied by condensation and segmentation of nuclei, loss of plasma membrane microvilli, and extensive degradation of the chromosomal DNA into nucleosome units. In addition to apoptosis during development (programmed cell death), apoptosis occurs in other systems. For example, in the immune system, the death of thymocytes induced through their antigen-receptor complex or by glucocorticoid occurs by an apoptotic process (Golstein et al., 1991). Tumor regression by the immune system is also mediated by apoptosis; that is, cytotoxic T lymphocytes (CTL) or natural killer cells (NK) as well as tumor necrosis factor (TNF) or lymphotoxin (LT) induce apoptosis in the target cells. Furthermore, low doses of UV or y-ray irradiation or antitumor chemical drugs cause apoptosis of tumor cells (Hickman, 1992; Wyllie et al., 1980). Programmed cell death has been extensively studied in the live nematode, Caenorhabditis elegans (Ellis et al., 1991), in which the division and death of cells can be followed under the microscope. Many mutants of the death process have been identified, and their molecular analyses have indicated that many gene products are involved in various aspects of cell death in C. elegans. On the other hand, the molecular mechanism of cell death in the mammalian system is poorly understood despite its importance during development. The Fas antigen is expressed on the surface of various mammalian 129

Copyright 0 1994 by Academic Press,Inc. All rights of reproduction in any form reserved.

130

SHIGEKAZU NAGATA

cells. It is a member of the TNF/nerve growth factor (NGF) receptor family and transduces the apoptotic signal (Itoh et al., 1991; WatanabeFukunaga et al., 1992b). Molecular analysis of the Fas gene has indicated that it is the structural gene for the mouse lpr (lymphoproliferation) mutation (Adachi et al., 1993; Watanabe-Fukunaga et al., 1992a). We have identified a natural Fas ligand in a CTL cell line and showed that it is a member of the TNF family (Suda et al., 1993). Here, the Fas/Fas ligand system is summarized and its physiological role is discussed. II. Far Antigen

The Fas antigen (Fas) (Yonehara et al., 1989), also called APO-1 antigen (Trauth et al., 1989),is a cell-surface protein belonging to the TNF/NGF receptor family (Itoh et al., 1991; Oehm et al., 1992; Nagata, 1993). As shown in Fig. 1, the members of this family include two TNF receptors (types I or 55K and type I1 or 75K, respectively) (Schall et al., 1990; Smith et al., 1990);the low-affinity NGF receptor ( Johnson et al., 1986); the B-cell antigen CD40 (Stamenkovic et al., 1989); the T-cell antigen OX40 (Mallettet al., 1990),CD27 (Camerini et al., 1991), 4-1BB (Kwon and Weissman, 1989);and the Hodgkin’s lymphoma cellsurface antigen CD30 (Durkop et al., 1992).The extracellular regions of members in this family are rich in cysteine residues, and they can be divided into three to six subdomains. The amino acid sequence of the extracellular region is relatively conserved (about 24-30% homology), whereas the cytoplasmic region is not, except for some similarity between Fas and the TNF type I receptor (Itoh et al., 1991). The TNF and NGF receptors were identified as cytokine receptors. Fas, CD40, CD27, and CD30 are proteins which are recognized by specific monoclonal antibodies. Molecular cloning of the ligands for CD40, CD27, CD30, and 4-1BB (Armitage et al., 1992; Goodwin et al., 1993a,b; Smith et al. 1993) indicated that they are TNF-related type I1 membrane proteins and constitute a novel cytokine family (Farrah and Smith, 1992). As described below, the Fas ligand also turns out to be a member of the TNF family. 111. Expression of Fas

Activated human T and B cells abundantly express Fas (Trauth et al., 1989). Lymphoblastoid cells transformed with human T-cell leukemia virus (HTLV)-1 (Debatin et al., 1990), human immunodeficiency virus (HIV)(Kobayashiet al., 1990),or Epstein-Barrvirus (EBV) (Falk et a,?., 1992) highly express Fas. Some other tumor cell lines

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

131

FIG.1. The FaslTNFINGF receptor family. Members of the TNFlNCF receptor family are schematically shown. These include Fas; TNF type I and type I1 receptors; low-affinity NCF receptor; B-cell antigen CD40; T-cell antigens 0x40, 4-1BB, and CD27; Hodgkin’s lymphoma antigen CD30; and the soluble protein coded by Shope fibroma virus (SFV-T2). The shaded regions represent cysteine-rich subdomains, of which each member of the family contains three to six. A domain of about 80 amino acids in the cytoplasmic regions of Fas and the type I TNF receptor has some similarity, and it is shown as a bold line. indicates N-glycosylation sites.

such as human myeloid leukemia U937 (Yonehara et al., 1989), human squamous carcinoma CHU-2 (Itoh et al., 1991), and SV40-transformed mouse macrophage BAM3 cells ( Watanabe-Fukunaga et al., 1992b) express Fas, although the expression level is low compared with that of the lymphoblastoid cell lines. The expression of Fas is upregulated by interferon-y (IFN-y)in the mouse macrophage BAM3, human adenocarcinoma HT-29 and mouse fibroblast L929 cell lines (Itoh et al., 1991; Watanabe-Fukunaga et al., 1992b), or by a combination of IFNy and TNF-(r in human tonsillar B cells (Moller et al., 1993). The tissue distribution of the Fas mRNA in the mouse has been examined (Watanabe-Fukunaga et al., 199213). The Fas mRNA was detected abundantly in the thymus, heart, liver, and ovary of 8-weekold adult mice, but not in the brain, bone marrow, and spleen. In thymocytes, Fas is expressed in almost all populations except for double-negative (CD4- CD8-) thymocytes (Drappa et al., 1993; Oga-

132

SHIGEKAZU NAGATA

sawara et al., 1993; J . Ogasawara, T. Suda, and S. Nagata, unpublished results). IV. Mutation of the Fas Gene in Ipr Mice

Southern hybridization of genomic DNA indicated that there is only one chromosomal gene for Fas in human and mouse chromosomes (Adachi et al., 1993). In situ hybridization localized the human gene on chromosome 10q24.1 (Inazawa et aZ., 1992), and interspecific backcross analysis indicated that the mouse Fas gene is in the region of chromosome 19, which is homologous to human 10q24.1 (WatanabeFukunaga et al., 1992b). Referring the location of the mouse Fas gene to the mouse genomic database (GBASE), it was found that the Fas gene is close to the l p r locus (Watanabe et al., 1991). There are two known allelic mutations, l p r and Z p F , at the Zpr locus. These mutants have a similar phenotype, but ZprCg slightly complements the gld mutation in double-heterozygotes between Zpr and gld mutations (Matsuzawa et al., 1990). Northern hybridization of the thymus and liver from l p r mice showed little expression of the Fas mRNA (WatanabeFukunaga et al., 1992a).Accordingly, flow cytometry using anti-mouse Fas antibody hardly detected the Fas protein on the thymocytes from Zpr mice (Drappa et al., 1993; Ogasawara et al., 1993). Since Southern hybridization of the chromosomal DNA suggested a distinct rearrangement of the Fas gene in lpr mice, the chromosomal gene was molecularly cloned from the wild-type and lpr mice (Adachi et al., 1993). The mouse Fas gene consists of over 70 kb and is split by 9 exons (R. Watanabe-Fukunaga and S. Nagata, unpublished results). Restriction enzyme mapping of the Fas gene from lpr mice indicated that the promoter and exons of the Fas gene in this mouse are intact. However, an early transposable element (ETn) of 5.4 kb was inserted in intron 2 ofthe Fas gene. The ETn is a mouse endogenous retrovirus, of which about 1000 copies can be found in the mouse genome (Brulet et al., 1983). Although the ETn does not carry a meaningful open reading frame, it has long terminal repeat (LTR) sequences (about 300 bp) at both the 5’ and 3‘ termini. This LTR sequence contains a poly(A) adenylation signal (AATAAA) which terminates the transcription at this region. In fact, short mRNA’s of about 1.0 kb coding for exons 1 and 2 of the Fas gene were abundant in the thymus and liver of the Zpr mice (Adachi et aZ., 1993). Furthermore, inserting the ETn into an intron of a mammalian expression vector dramatically, but not completely, reduced the expression efficiency in mammalian cells. These results indicate that, in Zpr mice, an insertion of an ETn into an intron of the Fas gene greatly reduces the expression of the func-

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

133

tional Fas mRNA, but its mutation is leaky. Later, several other groups reached the same conclusion by analyzing the Fas transcript in l p r mice by means of the reverse polymerase chain reaction (Chu et al., 1993; Kobayashi et al., 1993; Wu et al., 1993). In contrast to the lpr mice, Z p F mice express the Fas mRNA of normal size as abundantly as the wild type (Watanabe-Fukunaga et al., 1992a). However, this mRNA carries a point mutation of T to A, which causes the replacement of isoleucine with asparagine in the Fas cytoplasmic region. This mutation is in the domain which has similarity with the TNF type I receptor (see below), and it abolishes the ability of Fas to transduce the apoptotic signal (WatanabeFukunaga et al., 1992a). Furthermore, when the corresponding amino acid (valine-238) of the human Fas was mutated to asparagine, it could not transduce the apoptotic signal into cells (Itoh and Nagata, 1993). V. Fas-Mediated Apoptosis in Vitro and in Vivo

To assess the function of Fas, mouse cell transformants constitutively expressing human Fas were established using various mouse cell lines as host cells (Itoh et al., 1991).When the transformed cells were treated with anti-human Fas antibody, cells expressing human Fas, but not the parental mouse cells, died within 5 hr. Examination of the dying cells under an electron microscope revealed extensive condensation and fragmentation of the nuclei, which is characteristic of apoptosis. The chromosomal DNA started to degrade in a laddered fashion after a 2-hr incubation with the anti-Fas antibody. A human Fas expression plasmid has also been introduced into a mouse interleukin3( IL-3)dependent myeloid leukemia FDC-P1 cell line (Itoh et al., 1993). Although the transformed cells died due to IL-3 depletion, they did so over 36 hr, as observed with the parental FDC-P1 cells. On the other hand, exposure to the anti-human Fas antibody killed the cells within 5 hr in the presence of IL-3. From these results, we concluded that Fas actively mediates the apoptotic signal into cells, and the antiFas antibody works as agonist. The anti-Fas antibody has lethal activity in uiuo (Ogasawara et aZ., 1993). We established several hamster monoclonal antibodies against mouse Fas. One of them had cytolytic activity in uitro. When this antibody is intraperitoneally injected into mice, the wild-type mice but neither lpr nor l p f g mice died within 5-6 hr. These results clearly indicate that the lethal effect of the anti-Fas antibody is due to binding of the antibody to Fas to activate the death pathway and not due to a substance(s) such as endotoxin contaminated with the antibody. Furthermore, the fact that lpl"g mice expressing the nonfunctional Fas are

134

SHIGEKAZU NAGATA

resistant to the lethal effect of the antibody suggests little involvement of the complement system in this killing process. Biochemical analysis of sera from the dying mice showed a specific and dramatic increases of glutamic oxaloacetic transaminase (GOT)and glutamic pyruvic transaminase (GPT)levels shortly after injection of the antibody, suggesting liver injury. Accordingly, histological and electron microscope analyses of the tissues indicated that hepatocytes were killed by apoptosis (Fig. 2). The effect of the anti-Fas antibody in vivo seems to be a direct effect on the liver because the anti-Fas antibody also caused apoptosis in primary cultures of hepatocytes (R. Ni, Y. Tomita, A. Ichihara, K. Ishimura, J. Ogasawara, and S. Nagata, unpublished results). These results indicate that the Fas expressed in mouse tissues (at least in the liver) is competent in transducing the apoptotic signal into cells. VI. Signal Transduction Mediated by Fas

The apoptotic signal is induced by the binding of anti-Fas or antiAPO-1 antibody, or the Fas ligand, to Fas. The anti-human Fas antibody is an IgM class antibody which is an immunoglobulin pentamer, whereas the anti-APO-1antibody is an IgG, class antibody which tends to aggregate. The F(ab’)z fragment or other isotypes of the anti-APO1 antibody hardly induce apoptosis of cells expressing Fas (Dhein et al., 1992). On the other hand, the cytotoxic activity of the inactive antiAPO-1 antibody was reconstituted by cross-linking the antigen with a second antibody or with protein A. These results indicate that Fas dimerization alone is not sufficient to transduce the apoptotic signal. It seems that the oligomerization of at least three Fas molecules is a biologically relevant complex in generating an intracellular signal. As described below, the fact that Fas ligand is a TNF-related molecule which exists as a trimer (Smith and Baglioni, 1987) agrees with this hypothesis. The cytoplasmic domain of Fas consists of 145amino acids, in which no motif for enzymatic activity such as kinases or phosphatase can be found. However, about 70 amino acids in this region have significant similarity with a part of the cytoplasmic region of the type I, but not the type 11, TNF receptor (Itoh et al., 1991). TNF has numerous biological functions, including cytotoxic and proliferative activities (Old, 1985). Tartaglia et al. (1991) have shown that the type I receptor is mainly responsible for the cytotoxic activity of TNF, while the type I1 receptor mediates the proliferation signal in thymocytes. The similarity of Fas and the type I TNF receptor in their cytoplasmic regions therefore suggests an important role of this domain for apop-

FIG.2. The Fas-mediated apoptosis of hepatocytes in vivo. The purified anti-mouse Fas antibody (100 pg) was subcutaneouslyinjected into mice. At 2 hr after injection, the liver section was stained with hematoxylin and eosin (A), which shows focal hemorrhage and necrosis. Only a few normal hepatocytes remained, and most hepatocytes carry pyhotic nuclei. (B) A liver section examined

under a transmission electron microscope.The affected hepatocytes show the condensed and fragmented nuclei characteristic of apoptosis.

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

135

totic signal transduction. In fact, analyses of serial deletions and point mutations in the Fas protein have indicated that the domain conserved between Fas and the type I TNF receptor is essential for the function of Fas (Itoh and Nagata, 1993). Observations of the human type I TNF receptor have indicated that the domain homologous to Fas is responsible and sufficient for TNF-induced cytolytic activity (Tartaglia et al., 1993),which agrees with our conclusion. Furthermore, the mutational analysis of Fas revealed an inhibitory domain for apoptosis in the C-terminus. That is, a Fas mutant lacking 15amino acids from the C-terminus was an upmutant, in which about 10 times less anti-Fas antibody than that required for the wild-type Fas was sufficient to induce apoptosis (Itoh and Nagata, 1993).It is possible that association of accessory molecule(s) or modification of Fas at this region downregulates the activity of Fas to transduce the apoptotic signal. VII. Fas Ligand

As described above, the structure of Fas suggested that it is a receptor for an unknown cytokine. Rouvier et al. (1993) have established a CTL hybridoma cell line (PCGO-dlOS, abbreviated dlOS) which has cytotoxic activity against thymocytes from wild-type, but not lpr mice, suggesting the presence of Fas ligand on its surface. To confirm the expression of Fas ligand in this cell line, we prepared a soluble form (Fas-Fc) of Fas by fusing the extracellular region of Fas to the Fc region of human IgG. The fusion protein inhibited the Fas-dependent CTL activity of dlOS cells in a dose-dependent manner, and the Fas ligand was detected by FACS on the cell surface of dlOS cells using labeled Fas-Fc (Suda et al., 1993).The Fas ligand was then purified to homogeneity by affinity chromatography using Fas-Fc, and we showed that the purified protein had cytolytic activity against cells expressing Fas (Suda and Nagata, 1994). We then isolated the Fas ligand cDNA from the dlOS cell line using the panning procedure (Suda et al., 1993).The recombinant Fas ligand expressed in COS cells induced apoptosis of cells expressing Fas. The amino acid sequence deduced from the nucleotide sequence of the cDNA indicated that the Fas ligand is a TNF-related type I1 membrane protein (Suda et d.,1993). As shown in Fig. 3, members of the TNF family include Fas ligand, TNF, LT, and ligands for CD40, CD30, and CD27. TNF was originally identified as a soluble cytokine (Pennica et al., 1984), which works as a trimer (Smith and Baglioni, 1987). However, it was later shown that TNF is synthesized as a type I1 membrane protein which can be cleaved to produce a soluble form (Kriegler et al., 1988). LT consists of LTa and LTP and is expressed in certain CTL (Androlewicz et al.,

136

SHIGEKAZU NAGATA

FasL FIG.3. The TNF family. The members of the TNF family are schematically shown. The members include the Fas ligand (FasL), membrane-boundTNF, lymphotoxin (LT) which consists of LTa and LTP, CD40 ligand (CD40L), CD30 ligand (CDSOL), and CD27 ligand (CD27L). The shaded regions have significant similarity. The numbers indicate the amino acid number of the conserved, the spacer, and intracellular regions.

1992). LTa, also called TNF-P, is produced as a soluble cytokine with a signal sequence (Gray et at., 1984),while LTP is a type I1 membrane protein (Browning et al., 1993). LTa and LTP associate on the cell surface probably as a trimer (Androlewicz et al., 1992) and bind to a newly identified receptor which belongs to the TNF/NGF receptor family (Crowe et al., 1994).The ligands for CD40, CD30, CD27, and 4-1BB are type I1 membrane proteins expressed in activated T cells (Armitage et al., 1992; Goodwin et al., 1993a,b; Smith et al., 1993). When the Fas ligand is overproduced in COS cells, the soluble form of the Fas ligand which can actively induce apoptosis can be found in supernatant (Suda et al., 1993; Suda and Nagata, 1994). These results suggest that under abnormal conditions, the soluble form of the Fas ligand can be produced in the body as found in the TNF system (Old, 1985). The tertiary structure of TNF has been extensively studied. It forms an elongated, antiparallel p-pleated sheet sandwich with a jellyroll topology (Banner et al., 1993; Eck and Sprang, 1989; Eck et al., 1992). The significant conservation of the amino acid sequence among mem-

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

137

bers suggests that others of the family, including the Fas ligand, have a structure similar to that TNF and work as a trimer. However, despite the high similarity between Fas ligand and TNF (about 30% identical at the amino acid sequence level), Fas ligand does not bind to the TNF receptor (Suda et al., 1993). VIII. Physiological Roles of the Fas System

Since Fas was identified as a cell-surface protein which mediates apoptosis (Itoh et al., 1991), considerable progress has been made regarding its physiological role. Our finding that the Fas gene is the structural gene for the Zpr mutation pointed to the important role of Fas in the development of T cells. However, it remains controversial at which step of T-cell development Fas is involved. Immature T cells are killed by apoptosis in at least two steps during development in the thymus (Ramsdell and Fowlkers, 1990). Those T cells carrying Tcell receptors which do not recognize self-MHC antigens as a restriction element are killed or “neglected,” while the T cells recognizing the self antigens are killed by a process called “negative selection.” Analysis of thymic T-cell development in wild-type and l p r mice has suggested that the neglected thymocytes escape from apoptosis in the thymus of Zpr mice and then migrate to the periphery (Zhou et al., 1993).On the other hand, Herron et al. (1993)reported that the development of T cell in the thymus is relatively normal in Zpr mice. These different observations may be partly due to the leakiness of the l p r mutation as mentioned above. In addition to being expressed in thymocytes, Fas is expressed in activated mature T cells (Trauth et aZ., 1989), and the prolonged activation of T cells leads the cells susceptible to cytolytic activity of anti-Fas antibody (Klas et aZ., 1993; Owen-Schaub et al., 1992). Since mature T cells from lpr mice are resistant to antiCDSstimulated suicide, Russell et al. (1993) suggested a role of Fasmediated apoptosis in the induction of peripheral tolerance and/or in the antigen-stimulated suicide of mature T cells. Fas is expressed in other tissues such as the liver, heart, and lung (Watanabe et al., 1991). Although these organs are rather stable, and no apparent abnormal phenotypes are seen in these tissues of Zprmice, Fas may also be involved in development and/or turnover in these tissues. Since abnormal activation of Fas causes severe tissue damage (Ogasawara et al., 1993), it is possible that the Fas system is involved in many human autoimmune diseases such as fulminant hepatitis. In this regard, it is notable that a particular CTL cell line induces apoptosis in hepatocytes and causes fulminant hepatitis (Ando et al., 1993; Chisari, 1992). If involvement of the Fas system in human dis-

138

SHIGEKAZU NAGATA

eases is proven, antagonistic antibodies against Fas or Fas ligand, or the soluble form of Fas, could be used in a clinical setting. The Fas ligand is expressed in some CTL cell lines and in activated splenocytes (Suda et aZ., 1993), suggesting an important role of the Fas system in CTL-mediated cytotoxicity. Two mechanisms for CTLmediated cytotoxicity are known (Apasov et al., 1993; Golstein et al., 1991; Podack et al., 1991). One is a Ca2+-dependentpathway in which perforin plays an important role. The other pathway is a Ca2+independent pathway. In the perforin-knocked out mice, the spleen cells still showed some Ca2+-independentCTL activity (Ktigi et aZ., 1994a). Fas ligand can kill the cells independently of Ca2' and Kagi et al. (199413)recently showed that the residual CTL activity remaining in the perforin-knocked out mice is due to the Fas ligand expressed in CTL. Mice carrying the gld mutation show phenotypes similar to Ipr (Cohen and Eisenberg, 1991). Allen et al. (1990) suggested that gld and Zpr are mutations of an interacting pair of molecules. As shown above, the Fas gene is the structural gene for Zpr, and Fas is the receptor for Fas ligand. Recently, we have found that gld mice carry a mutation in the Fas ligand gene (Takahashi et al., 1994). The mutations in Fas (Zpr mutation) or Fas ligand (gld mutation) causes lymphadenopathy and autoimmune disease, and the Fas ligand was found in CTL. These results imply that the FaslFas ligand system involved in the T-cell development plays an important role in CTL-mediated cytotoxicity. It is possible that the killing process of autoreactive T cells in Tcell development and the killing process of tumor cells by CTL may proceed by a similar mechanism. As shown schematically in Fig. 4, self, tumor, or virus antigens in the target cells may activate effector cells (CTL) through the T-cell receptor to induce the expression of the Fas ligand gene. Fas ligand then binds to Fas on the target cells, causing apoptosis. In addition to lymphocytes, Fas and the Fas ligand are expressed in other tissues such as the liver, lung, and heart, suggesting that a similar mechanism operates to remove unnecessary or toxic cells from these tissues during development. IX. Perspectives

We demonstrated that Fas ligand is a death factor, and Fas is its receptor. These results indicate that just as growth factor and its receptor regulate cell proliferation, cell death or apoptosis is regulated by a death factor and its receptor (Fig. 5). The growth and differentiation of cells are controlled by signals such as activation of kinases, Ca2+ mobilization, or CAMP formation, which are stimulated by growth

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

139

Effector Cells (CTL)

tumor Ag

I

tApoptosis Target Cells

FIG.4. A model for the Fas-mediated cytotoxicity of CTL. A proposed mechanism for the Fas-mediated cytotoxicity in the CTL system is schematically shown. The target cells express the self, tumor, or virus antigen as a complex with MHC, which interacts with the T-cell receptor (TCR) on CTL. This interaction activates the CTL and induces the expression of the Fas ligand (Fas-L) gene. The Fas-L expressed on the cell surface of the CTL then binds to Fas on the target cells and induces its apoptosis.

Fas ligand

TNFINGF-R homologous domain

Signal transducing domain Inhibitory domain

t

+ ?

Apoptosis

FIG.5. Fas-mediated apoptosis. Fas and the Fas ligand are schematically shown. The Fas ligand binds to Fas on the cell surface probably as a trimer and activates apoptotic signal transduction. In the cytoplasmic region of Fas, a region of about 80 amino acids is responsible for the signal transduction, while the C-terminal domain (about 15 amino acids) inhibits apoptosis.

140

SHlGEKAZU NAGATA

and differentiation factors. Currently, the kinds of signaling molecules involved in Fas-mediated apoptosis are unknown. Fas may activate a similar signal transducer, or utilize a completely different set of molecules. Since overexpression of the bcl-2 oncogene product partially inhibits Fas-mediated apoptosis (Itoh et al., 1993), bcl-2 should interact somewhere in the signal-transducing pathway activated by the Fas system. Elucidation of the apoptotic signal transduction mechanism mediated by Fas may reveal a novel mechanism. The gain-of-function mutation of the growth factor system causes cellular transformation, whereas the loss-of-function mutation of the Fas system (Zpr mutation) causes lymphadenopathy. In this regard, Fas and the Fas ligand may be considered as tumor suppressor genes. The loss-of-function mutation in the growth factor system causes the disappearance or dysfunction of specific cells. As pointed out above, it is possible that abnormal activation (gain of function mutation) of the Fas or Fas ligand causes fulminant hepatitis or other diseases such as CTL-mediated autoimmune diseases.

ACKNOWLEDGMENTS I thank Drs. 0. Hayaishi and C. Weissmann for encouragement and discussion. The work was camed out in a collaboration with Drs. T. Suda, M. Adachi, J. Ogasawara, T. Takahashi, N. Itoh, and R. Watanabe-Fukunaga and supported in part by Grants-in-Aid from the Ministry of Education, Science and Culture of Japan. I also thank Ms. K. Mimura for secretarial assistance.

REFERENCES Adachi, M., Watanabe-Fukunaga, R., and Nagata, S. (1993).Aberrant transcription caused by the insertion of an early transposable element in an intron of the Fas antigen gene of lpr mice. Proc. Nutl. Acad. Sci. U.S.A. 90, 1756-1760. Allen, R. D., Marshall, J. D., Roths, J. B., and Sidman, C. L. (1990). Differences defined by bone marrow transplantation suggest that lpr that gld are mutations of genes encoding an interacting pair of molecules. J . E x p . Med. 172, 1367-1375. Ando, K., Moriyama, T., Guidotti, L. G., Wirth, S., Schreiber, R. D., Schlicht, H. J., Huang,S., and Chisari, F. V. (1993). Mechanisms of class I restricted immunopathology: A transgenic mouse model of fulminant hepatitis. J . Exp. Med., in press. Androlewicz, M. J., Browning, J. L., and Ware, C. F. (1992). Lymphotoxin is expressed as a heteromeric complex with a distinct 33-kDa glycoprotein on the surface of an activated human T cell hybridoma. J . Biol. Chem. 267,2542-2547. Apasov, S., Redegeld, F., and Sitkovsky, M. (1993). Cell-mediated cytotoxicity: Contact and secreted factors. Cum. Opinion lmmunol. 5,404-410. Armitage, R. J., Fanslow, W. C., Strockbine, L., Sato, T. A., Clifford, K. N., Macduff, B. M., Anderson, D. M., Gimpel, S. D., Davis-Smith, T., Maliszewski, C. R., Clark, E. A., Smith, C. A., Grabstein, K. H., Cosman, D., and Spriggs, M. K. (1992). Molecular and biological characterization of a murine ligand for CD40. Nature (London) 357, 80-82. Banner, D. W., D'Arcy, A., Janes, W., Gentz, R., Schoenfeld, H.-J.,Broger, C., Loetscher,

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

141

H., and Lesslauer, W. (1993). Crystal structure of the soluble human 55 kd TNF receptor-human TNFP complex: Implication for TNF receptor activation. Cell (Cambridge, Mass.) 73, 431-445. Browning, J. L., Ngam-ek, A., Lawton, P., DeMarinis, J., Tizard, R., Chow, E. P., Hession, C., O’Brine-Greco, B., Foley, S. F., and Ware, C. F. (1993).Lymphotoxin /3, a novel member of the TNF family that forms a heteromeric complex with lymphotoxin on the cell surface. Cell (Cambridge, Mass.) 72,847-856. Brulet, P., Kaghad, M., Xu,Y.-X., Croissant, O., and Jacob, F. (1983).Early differential tissue expression of transposon-like repetitive DNA sequences of the mouse. Proc. Natl. Acad. Sci. USA. 80,5641-5645. Camerini, D., Walz, G., Loenen, W. A. M., Borst, J., and Seed, B. (1991). The T cell activation antigen CD27 is a member of the nerve growth factor/tumor necrosis factor receptor gene family. J . Immunol. 147,3165-3169. Chisari, F. V. (1992).Hepatitis B virus biology and pathogenesis. Mol. Genet. Med. 2, 67-103. Chu, B. J.-L., Drappa, J., Pamassa, A., and Elkon, K. B. (1993).The defect in Fas mRNA expression in MRL/lpr mice is associated with insertion of the retrotransposon, ETn. J . Exp. Med. 178,723-730. Cohen, P. L., and Eisenberg, R. A. (1991).Lpr and gld: Single gene models of systemic autoimmunity and lymphoproliferative disease. Annu. Rev. Immunol. 9,243-269. Crowe, P. D., VanArsdale, T. L., Walter, B. N., Ware, C. F., Hession, C., Ehrenfels, B., Browning, J. L., Din, W. S., Goodwin, R. G., and Smith, C. A. (1994).A lymphotoxinP-specific receptor. Science 264,707-710. Debatin, K.-M., Goldmann, C. K., Bamford, R., Waldmann, T. A., and Krammer, P. H. (1990). Monoclonal-antibody-mediated apoptosis in adult T-cell leukaemia. Lancet 335,497-500. Dhein, J., Daniel, P. T., Trauth, B. C., Oehm, A., Moller, P., and Krammer, P. H. (1992). Induction of apoptosis by monoclonal antibody anti-APO-1 class switch variants is dependent on cross-linking of APO-1 cell surface antigens. J . Immunol. 149, 3166-3173. Drappa, J., Brot, N., and Elkon, K. B. (1993).The Fas protein is expressed at high levels on CD4+CD8+ thymocytes and activated mature lymphocytes in normal mice but not in the lupus-prone strain, MRL lpr/lpr. Proc. Natl. Acad. Sci. U S A . 90,10,34010,344. Durkop, H., Latza, U., Hummel, M., Eitelbach, F., Seed, B., and Stein, H. (1992). Molecular cloning and expression of a new member of the nerve growth factor receptor family that is characteristic for Hodgkin’s disease. Cell (Cambridge, Mass.) 68, 421-427. Eck, M. J., and Sprang, S. R. (1989).The structure of tumor necrosis factor-a at 2.6 hi resolution. J . Biol. Chem. 264, 17,595-17,605. Eck, M. J., Ultsch, M., Rinderknecht, E., de Vos, A:M., and Sprang, S. R. (1992).The structure of human lymphotoxin (tumor necrosis factor$) at 1.9-hiresolution.]. Biol. Chem. 267,2119-2122. Ellis, R. E., Yuan, J., and Horvitz, H. R. (1991).Mechanisms and functions of cell death. Annu. Rev. Cell Biol. 7,663-698. Falk, M. H., Trauth, B. C., Debatin, K.-M., Klas, C., Gregory, C. D., Rickinson, A. B., Calender, A,, Lenoir, G. M., Ellwart, J. W., Krammer, P. H., and Bornkamm, G. W. (1992). Expression of the APO-1 antigen in Burkitt lymphoma cell lines correlates with a shift towards a lymphoblastoid phenotype. Blood 79,3300-3306. Farrah, T., and Smith, C. A. (1992). Emerging cytokine family. Nature (London) 358,26.

142

SHIGEKAZU NAGATA

Golstein, P., Ojcius, D. M., and Young, J. D.-E. (1991). Cell death mechanisms and the immune system. Immunol. Reo. 121,29-65. Goodwin, R. G., Alderson, M. R., Smith, C. A., Armitage, R. J., VandenBos, T., Jerzy, R., Tough, T. W., Schoenbom, M. A., Davis-Smith, T., Hennen, K., Falk, B., Cosman, D., Baker, E., Sutherland, G. R., Grabstein, K. H., Farrah, T., Giri, J. G., and Beckman, M. P. (1993a). Molecular and biological characterization of a ligand for CD27 defines a new family of cytokines with homology to tumor necrosis factor. Cell (Cambridge, MUSS.)73,447-456. Goodwin, R. G., Din, W. S., Davis-Smith, T., Anderson, D. M., Gimpel, S. D., Sato, T. A., Maliszewski, C. R., Brannan, C. I., Copeland, N. G., Jenkins, N. A., Farrah, T., Armitage, R. J.,Fanslow, W. C., and Smith, C. A. (199313). Molecular cloning of a ligand for the inducible T cell gene 4-1BB: A member of an emerging family of cytokines with homology to tumor necrosis factor a. Eur. J. Immunol. 23,2631-2641. Gray, P. W., Aggarwal, B. B., Benton, C. V., Bringman, T.S., Henzel, W. J., Jarrett, J. A., Leung, D. W., Moffat, B., Ng, P., Svedersky, L. P., Palladino, M. A., and Nedwin, G. E. (1984).Cloning and expression of cDNA for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature (London) 312,721-724. Herron, L. R., Eisenberg, R. A., Roper, E., Kakkanaiah, V. N., Cohen, P. L., and Kotzin, B. L. (1993). Selection of the T cell receptor repertoire in Lpr mice'. J . Immunol. 151,3450-3459. Hickman, J. A. (1992). Apoptosis induced by anticancer drugs. Cancer Metast. Reo. 11, 121-139. Inazawa, J., Itoh, N., Abe, T., and Nagata, S. (1992). Assignment of the human Fas antigen gene (FAS) to 1Oq24.1. Genomics 14,821-822. Itoh, N., and Nagata, S. (1993).A novel protein domain required for apoptosis: Mutational analysis of human Fas antigen. J . Biol. Chem. 268, 10,932-10,937. Itoh, N., Yonehara, S., Ishii, A,, Yonehara, M., Mizushima, S., Sameshima, M., Hase, A., Seto, Y., and Nagata, S. (1991).The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis. Cell (Cambridge, Mass.) 66, 233243. Itoh, N., Tsujimoto, Y., and Nagata, S. (1993). Effect of bcl-2 on Fas antigen-mediated cell death, J . Immunol. 151,621-627. Johnson, D., Lanahan, A., Buck, C. R., Sehgal, A,, Morgan, C., Mercer, E., Bothwell, M., and Chao, M. (1986).Expression and structure of the human NGF receptor. Cell (Cambridge, Mass.) 47,545-554. Kiigi, D., Ledermann, B., Burki, K., Seiler, P., Odermatt, B., Olsen, K. J., Podack, E. R., Zinkemagel, R. M., and Hengartner, H. (1994a). Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature (London) 369,31-37. Kiigi D., Vignaux, F., Ledermann, B., Biirki, K., Depraetere, V., Nagata, S., Hengartner, H., and Golstein, P. (1994b). Fas and perforin pathways as major mechanisms of T cell-mediated cytotoxicity. Science, (in press). Has, C., Debatin, K.-M., Jonker, R. R., and Gammer, P. H. (1993).Activation interferes with the APO-1 pathway in mature human T cells. Znt. Immunol. 5,625-630. Kobayashi, N., Hamamoto, Y., Yamamoto, N., Ishii, A., Yonehara, M., and Yonehara, S. (1990).Anti-Fas monoclonal antibody is cytocidal to human immunodeficiency virusinfected cells without augmenting viral replication. Proc. Natl. Acad. Sci. U S A . 87, 9620-9624. Kobayashi, S., Hirano, T., Kakinuma, M., and Uede, T. (1993).Transcriptional repression and differential splicing of Fas mRNA by early transposon (ETn)insertion in autoimmune LPR mice. Biochem. Biophys. Res. Commun. 191,617-624.

FAS AND FAS LIGAND: A DEATH FACTOR AND ITS RECEPTOR

143

Kriegler, M., Perez, C., DeFay, K., Albert, I., and Lu, S. D. (1988). A novel form of TNFEachectin is a cell surface cytotoxic transmembrane protein: Ramifications for the complex physiology of TNF. Cell (Cambridge, Mass.) 53,45-53. Kwon, B. S., and Weissman, S. M. (1989). cDNA sequences of two inducible T-cell genes. Proc. Natl. Acad. Sci. U.S.A. 86, 1963-1967. Mallett, S., Fossum, S., and Barclay, A. N. (1990).Characterization of the MRC OX40 antigen of activated CD4 positive T lymphocytes-A molecule related to nerve growth factor receptor. EMBO J . 9, 1063-1068. Matsuzawa, A., Moriyama, T., Kaneko, T., Tanaka, M., Kimura, M., Ikeda, H., and Katagiri, T. (1990).A new allele of the lpr locus, lpl“g, that complements the gld gene in induction of lymphadenophaty in the mouse. J . E x p . Med. 171,519-531. Moller, P., Henner, C., Leithauser, F., Eichelmann, A,, Schmidt, A., Bruderlein, S., Dhein, J., and Krammer, P. H. (1993). Co-regulation of the APO-1 antigen with ICAM-1 (CD45) in tonsillar B cells and coordinate expression in follicular center B cells and in follicle center and mediastinal B cell lymphomas. Blood 81, 20672075. Nagata, S. (1993). Apoptosis-mediating Fas antigen and its natural mutation. In “Apoptosis 11” (L. D. Tomei and F. C. Cope, eds.) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 313-326. Oehm, A., Behrmann, I., Falk, W., Pawlita, M., Maier, G., Has, C., Li-Weber, M., Richards, S., Dhein, J., Trauth, B. C., Ponstingl, H., and Krammer, P. H. (1992). Purification and molecular cloning of the APO-1 cell surface antigen, a member of the tumor necrosis factorherve growth factor receptor superfamily: Sequence identity with the Fas antigen. J. B i d . Chem. 267, 10,709-10,715. Ogasawara, J., Watanabe-Fukunaga, R., Adachi, M., Matsuzawa, A., Kasugai, T., Kitamura, Y., Itoh, N., Suda, T., and Nagata, S. (1993).Lethal effect ofthe anti-Fas antibody in mice. Nature (London)364,806-809. Old, L. J. (1985).Tumor necrosis factor (TNF). Science 230,630-632. Owen-Schaub, L. B., Yonehara, S., Crump 111, W. L., and Grimm, E. A. (1992). DNA fragmentation and cell death is selectively triggered in activated human lymphocytes by Fas antigen engagement. Cell. Immunol. 140,197-205. Pennica, D., Nedwin, G. E., Hayflick, J. S., Seeburg, P. H., Derynck, R., Palladino, M. A., Kohr, W. J., Aggarwal, B. B., and Goeddel, D. V. (1984).Human tumour necrosis factor: Precursor structure, expression and homology to lymphotoxin. Nature (London) 312,724-728. Podack, K. R., Hengartner, H., and Lichtenheld, M. G. (1991).A central role of perforin in cytolysis? Annu. Reu. Immunol. 9, 129-157. Raff, M. C. (1992).Social controls on cell survival and cell death. Nature (London)356, 397-400. Ramsdell, F., and Fowlkers, B. J. (1990).Clonal deletion versus clonal anergy: The role of the thymus in inducing self tolerance. Science 248, 1342-1348. Rouvier, E., Luciani, M.-F., and Golstein, P. (1993). Fas involvement in CaZfindependent T cell-mediated cytotoxicity. J. Exp. Med. 177, 195-200. Russell, J. H., Rush, B., Weaver, C., and Wang, R. (1993).Mature T cells of autoimmune lprllpr mice have a defect in antigen-stimulated suicide. Proc. Natl. Acad. Sci. U.S.A. 90,4409-4413. Schall, T. J., Lewis, M., Koller, K. J., Lee, A., Rice, G. C., Wong, G. H. W., Gatanaga, T., Granger, G. A,, Lentz, R., Raab, H., Kohr, W. J., and Goeddel, D. V. (1990). Molecular cloning and expression of a receptor for human tumor necrosis factor. Cell (Cambridge, Mass.) 61,361-370. Smith, C. A,, Davis, T., Anderson, D., Solam, L., Beckman, M. P., Jerzy, R., Dower,

144

SHIGEKAZU NAGATA

S. K., Cosman, D., and Goodwin, R. G. (1990). A receptor for tumor necrosis factor defines an unusual family of cellular and viral proteins. Science 248, 1019-1023. Smith, C. A., Gruss, H.-J., Davis, T., Anderson, D., Farrah, T., Baker, E., Sutherland, G. R., Brannan, C. I., Copeland, N. G., Jenkins, N. A., Grabstein, K. H., Gliniak, B., McAlister, I. B., Fanslow, W., Alderson, M., Falk, B., Gimpel, S., Gillis, S., Din, W. S., Goodwin, R. G., and Armitage, R. J. (1993). CD30 antigen, a marker for Hodgkin’s lymphoma, is a receptor whose ligand defines an emerging family of cytokines with homology to TNF. Cell (Cambridge, Mass.) 73,1349-1360. Smith, R. A., and Baglioni, C. (1987).The active form of tumor necrosis factor is a trimer. J. Biol. Chem. 262,6951-6954. Stamenkovic, I., Clarck, E. A., and Seed, B. (1989). A B-lymphocyte activation molecule related to the nerve growth factor receptor and induced by cytokines in carcinomas. EMBO J . 8, 1403-1410. Suda, T., and Nagata, S. (1994). Purification and characterization of the Fas-ligand that induces apoptosis. J . Exp. Med. 179, 873-878. Suda, T., Takahashi, T., Golstein, P., and Nagata, S. (1993). Molecular cloning and expression of the Fas ligand: A novel member of the tumor necrosis factor family. Cell (Cambridge, Mass.) 75, 1169-1178. Takahashi, T., Tanaka, M., Brannan, C. I., Jenkins, N. A., Copeland, N. G., Suda, T., and Nagata, S. (1994).Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76, 969-976. Tartaglia, L. A., Weber, R. F., Figari, I. S., Reynolds, C., Palladino Jr., M. A., and Goeddel, D. V. (1991). The two different receptors for tumor necrosis factor mediate distinct cellular responses. Proc. Natl. Acad. Sci. U.S.A. 88,9292-9296. Tartaglia, L. A., Ayres, T. M., Wong, G. H. W., and Goeddel, D. V. (1993). A novel domain within the 55 kd TNF receptor signals cell death. Cell (Cambridge, Mass.) 74,845-853. Trauth, B. C., Klas, C., Peters, A. M. J., Matzuku, S., Moller, P., Falk, W., Debatin, K.M., and Krammer, P. H. (1989). Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245, 301-305. Walker, N. I., Harmon, B. V., Gobe, G. C., and Kerr, J. F. R. (1988). Patterns of cell death. Methods Achieu. Erp. Pathol. 13, 18-54. Watanabe, T., Sakai, Y., Miyawaki, S., Shimizu, A., Koiwai, O., and Ohno, K. (1991).A molecular genetic linkage map of mouse chromosome-19, including the lpr, Ly-44, and TdT genes. Biochem. Genet. 29,325-336. Watanabe-Fukunaga, R., Brannan, C. I., Copeland, N. G., Jenkins, N. A., and Nagata, S. (1992a).Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature (London)356,314-317. Watanabe-Fukunaga, R., Brannan, C. I., Itoh, N., Yonehara, S.,Copeland, N. G., Jenkins, N. A., and Nagata, S. (1992b). The cDNA structure, expression, and chromosomal assignment of the mouse Fas antigen. J. Immunol. 148,1274-1279. Wu, J., Zhou, T., He, J., and Mountz, J. D. (1993). Autoimmune disease in mice due to integration of an endogenous retrovirus in an apoptosis gene. J. Erp. Med. 178, 461-468. Wyllie, A. H., Kerr, J. F. R., and Currie, A. R. (1980). Cell death: The significance of apoptosis. Int. Reu. Cytol. 68,251-306. Yonehara, S., Ishii, A., and Yonehara, M. (1989). A cell-killing monoclonal antibody (anti-Fas) to a cell surface antigen co-downregulated with the receptor of tumor necrosis factor. J. Exp. Med. 169, 1747-1756. Zhou, T., Bluethmann, H., Eldridge, J., Berry, K., and Mountz, J. D. (1993). Origin of CD4-CD8-B220t T cells in MRL-lpr/lpr mice. J . Immunol. 150,3651-3667.

ADVANCES IN

IMMUNOLOGY,VOL. 51

Interleukin-5 and Its Receptor System: Implications in the Immune System and Inflammation KlYOSHl TAKATSU, SATOSHI TAKAKI, AND YASUMlCHl HlTOSHl Department of Immunology, Insfiiute of Medico1 Science, Univemify of Tokyo, Tokyo 108, Japan

I. Introduction

The immune system to infectious microbes is regulated by a series of interactions among T cells, B cells, and macrophages. During this process, B cells proliferate and differentiate into plasma cells which produce antibodies against distinct antigenic determinants of the antigen, and the antibodies produced play a key role in the humoral immune response against invading microorganisms. The B-cell response to an antigen is regulated by a helper T cell responding to, and specific for, the same antigen molecule. Helper T cells recognize antigenic peptide in the context of class I1 major histocompatibility complex (MHC)molecules on accessory cells and/or B cells (Brown et al., 1993) and secrete several soluble factors including interleukin-4 (IL-4), IL5, and IL-6 which can induce B-cell growth and maturation of B cells (reviewed by Howard and Paul, 1983; Melchers and Anderson, 1986; Kishimoto and Hirano, 1988; Takatsu, 1988; Paul and Ohara, 1987; Vitetta et al., 1984). Mouse interleukin-5 (mIL-5) is a glycoprotein induced in T cells after stimulation with an antigen, such as Mycobacterium tuberculosis (Tominaga et al., 1988)or Toxocara canis (Y. Yamaguchi et al., 1990a), and in mast cells upon stimulation with allergen/IgE complex or calcium ionophores (Plaut et al., 1989). The study of mIL-5 originated from the search for the B-cell differentiation factor that induces antigen-primed B cells to differentiate into antigen-specific antibody-producing cells or proliferation of BCLl B-cell tumor cells (Takatsu et al., 1980a; reviewed by Takatsu et al., 1988).This molecule was identified as a cytokine that has pleiotropic activities on various target cells including B cells, T cells, eosinophils, and basophils by the use of recombinant IL-5 and monoclonal antibody (mAb) to IL-5 (Harada et al., 1987a; Schumacher et al., 1988). Two mAbs, NC17 and TRF4, have been widely used because of their ability to neutralize mIL-5 function both in vitro and in vivo and hIL-5 function in vitro (Abrams et al., 1992; Coffman et al., 1989a; Hitoshi et al., 1991; Mita et al., 198913). A number of mIL-5-dependent mouse B-cell lines have 145

Copyright 0 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.

146

KIYOSHI TAKATSU ET AL.

now been isolated and provided a convenient biological assay for IL-5 (Tominaga et al., 1989). In humans, IL-5 induces the production of eosinophils from bone marrow progenitors and works for survival and priming of eosinophils in vitro. Human IL-5 expression was observed in many diseases with eosinophilia, suggesting that IL-5 plays important roles in promoting production and function of eosinophils in vivo (reviewed by Sanderson, 1992). A key question regarding the action of IL-5 in responsive cells has been the molecular mechanism of signal transduction cascade after IL-5 binding to the functional IL-5 receptor (IL-5R). The IL-5 signals can be transduced through the high-affinity 1L-5R that consists of two different polypeptide subunits: a and p. The cDNAs encoding both a- and P-subunits for mIL-5R have been isolated; the a-subunit was found to be a 60 kDa (p60) protein and the p-subunit (a 130-kDa protein, p130) was identified as the protein, AICBB, identical to the p-subunit for mIL-3R and mGM-CSFR, which can convert the lowaffinity m I L J R a into a high-affinity receptor (Devos et al., 1991; Takaki et al., 1990,1991,1993). It is not clear, however, how IL-5 unique signals can be transduced through IL-5R into the cell interior. Although we and others have clarified that the tyrosine phosphorylation of cellular proteins is required for IL-5R-mediated signaling, the precise molecular mechanisms of the IL-5R-mediated signaling cascade are not clear. In this review, we summarize advances in IL-5 and IL-SR research in the study of structure, physiologic functions, and unique modes of receptor-mediated signaling. We also discuss pathophysiology of aberrant expression of IL-5 and its receptor. II. Interleukin-5

A. HISTOLOGICAL BACKGROUND IL-5 was described initially as a factor that induces terminal differentiation of B cells to Ig-secreting cells and originally designated as Tcell-replacing factor (TRF, Dutton et al., 1971; Schimpl and Wecker, 1972). TRF activity was originally screened by its ability to support the IgM hemolytic plaque-forming cell response of T-cell-depleted mouse B cells to sheep red blood cells, and it was found in supernatants of mixed lymphocyte cultures or of concanavalin A (Con-A)-stimulated T cells, and these TRF preparations were found to exert remarkable effects in a number of different assays of T- and B-cell function. Biochemical characterization of TRF took a long time because there was

IL-5 AND ITS RECEPTOR SYSTEM

147

a popular belief that the different activities should be ascribed to separate biochemical entities and because purification of TRF was difficult. We found TRF activity by using an adoptive cell transfer system in supernatants of lymph node cell culture of M . tuberculosisprimed mice after stimulation with purified protein derivative (PPD)-presenting cells (Takatsu et al., 1974,1979). TRF activity was assessed by its ability to induce anti-2,Pdinitrophenyl (DNP) IgG antibody-secreting cells from T-cell-depleted DNP-primed B cells. Then we established an in vitro culture system to assess TRF activity and examined TRF-producing T cell subsets (Takatsu et al., 1980a; Tominaga et al., 1980; Sano et al., 1984,1985). We also reported a strain of mice, DBA/2Ha, as a model of X-linked immunodeficiency which is selectively defective in the expression of putative TRF acceptor site(s) on B cells (Takatsu et al., 1981a,b,1982; Takatsu and Hamaoka, 1982; Tominaga et al., 1982). The establishment of a TRFproducingT cell hybrid B151K12 (B151),which does not secrete detectable levels of other lymphokines affecting B cells, demonstrated that TRF is a novel lymphokine distinct from other lymphokines and cytokines (Takatsu et al., 1980b). Subsequently, in vivo growing murine chronic B-cell leukemia (BCLJ was shown to differentiate into IgMsecreting cells by stimulation with TRF-containing B151 supernatant (Pure et al., 1981). A homogeneous TRF preparation purified from B151 supernatants appeared to be an acidic glycoprotein with a molecular mass of 50 to 60 kDa which had a smaller mass (25 to 30 kDa) under reducing conditions (Takatsu et al., 1985). We deduced from these observations that secreted forms of B151-TRF consist of dimers. This purified TRF preparation contained undetectable IL-1, IL-2, IL-3, IL-4, and interferon-y (IFN-y) activities, but did contain the growth-promoting activity (BCGFII activity) (Swain et al., 1982,1983) of BCL, cells as well as dextran sulfate-stimulated normal B cells (Harada et al., 1985). BCGFII activity always resided in the same fraction in which TRF activity was detected, suggesting that a single molecule is able to induce B-cell growth and differentiation. The molecular cloning of complementary DNA (cDNA)-encoding mouse TRF and expression and functional studies have convincingly demonstrated that a single molecule is responsible for both TRF and BCGF I1 activity (Kinashi et al., 1986; Azuma et al., 1986). Furthermore, TRF has been shown to exert pleiotropic activities on various target cells other than B cells such as T cells and eosinophils (reviewed by Takatsu, 1992). We therefore proposed that TRF be called IL-5. Mouse IL-5 stimulates proliferation and differentiation of activated B

148

KIYOSHI TAKATSU ET AL.

cells and regulates the production and function of some other hematopoietic cells, such as Ly-l(CD5)+ B cells (also known as B-1 cells) (Hardy and Hayakawa, 1986; Hayakawa et al., 1986; Hayakawa and Hardy, 1988; Kantor, 1991), eosinophils, and basophils. In humans, IL-5 acts mainly to induce production of eosinophils from bone marrow progenitors and to prolong survival in uitro. IL-5 expression was observed in many diseases with eosinophilia.

B. MOLECULARSTRUCTURE OF IL-5 AND ITSGENE 1 . cDNA and Organization of the Gene Since neither structural characterization of IL-5 nor mAb against IL-5 was available at that time, the best strategy for cDNA cloning for IL-5 was to use an expression vector system that requires only a limited amount of mRNA. We used the expression vector system (pSP6K) containing the SP6 promoter that was established by Noma et al., (1986) (reviewed by Honjo and Takatsu, 1990). Mouse IL-5 cDNA (pSP6K-mTRF) was isolated from 5 x lo4 independent cDNA libraries that were constructed from poly(A)+ RNA of the alloreactive Tcell clone, 2.19 (Kinashi et al., 1986). Mouse IL-5 cDNA encodes a mature polypeptide of 113 amino acids with a calculated molecular mass of 13kDa (Kinashi et al., 1986). N-terminal amino acid sequencing of mouse T-cell-derived IL-5 revealed a sequence identical to that deduced from the nucleotide sequence of the cDNA and confirms the inferred amino-terminus of the mature protein (Takahashi et aE., 1990). Mouse IL-5 contains three potential N-linked glycosylation sites and three cysteine residues. Using the mIL-5 cDNA as a probe, cDNAencoding hIL-5 was also isolated (Azuma et al., 1986; Campbell et al., 1987; Yokota et al., 1987). cDNA of hIL-5 codes for a mature polypeptide of 115 amino acids. Molecular structure and biological activities of IL-5 are summarized in Table 1. The chromosomal genes for mIL-5 and hIL-5 were isolated using IL-5 cDNAs as probes (Campbell et al., 1988; Tanabe et al., 1987; Mizuta et al., 1989). The IL-5 genes consist of four exons and three introns. Conserved TATA-like and lymphokine consensus sequences are found at about 30 and 70 base pairs (bp), respectively, upstream from the transcription initiation sites. The exon-intron organization and the location of the cysteine codons of the IL-5 genes resemble those of the granulocyte-macrophage colony-stimulating factor (GMCSF), IL-2, and IL-4 genes, but are quite distinct from the G-CSF gene. Within this region, the IL-5 gene is associated with short stretches of conserved sequence motifs, designated CLE (conserved lymphokine

149

IL-5 AND ITS RECEPTOR SYSTEM

TABLE I STRUCTURE AND BIOLOGICAL ACTIVITYOF IL-5 Mouse (1)Apparent MW (2) Number of amino acids Total Mature (3)N-Glycosylation sites (4) Chromosome localization (5)Genomic structure (6) Producer cells

Human

40-50 kDa

30-40 kDa

133 113 3 11 4 exons T cells Mast cells

134 115 2 5 4 exons T cells Reed Sternburg cells Eosinophils EBV-transformed B cells

(7) Biological activity (1)Induction of differentiation (a) B cell differentiation (b) Induction of IL-OR expression on B and T cells (c) Cofactor of cytotoxic T cell differentiation (d) Differentiation of eosinophils (e) Enhancement of histamine release by basophils (2) Stimulation of cell growth (a) Promotion of Ly-1 + early B cell growth (b) Ly-1 + B cell proliferation (c) Clonal expansion of resting B cells (d) Eosinophil precursor cell growth

element) 0, CLE1, and CLE2, which are also found in the 5' upstream regions of the IL-3, IL-4, and GM-CSF genes (Arai et al., 1990; Miyatake et al., 1991).The hIL-5 gene was mapped on chromosome 5q23.331.1 on which the IL-4, GM-CSF, and IL-3 genes are mapped (Huebner et al., 1985; Le Beau et al., 1986, 1987; Takahashi et al., 1988). The corresponding murine genes were shown to cluster within a 230-kilobase (kb) region of chromosome 11(Barlow et al., 1987).The IL-4, IL-5, GM-CSF, and IL-3 genes might be derived by duplications of a common ancestral gene. 2. Polypeptide Structures A single polypeptide with a molecular mass of about 14 kDa is detected when mIL-5 and RNA is transiently translated in rabbit reticulocyte lysates. Both mIL-5 and hIL-5 are in the range of 45-60 kDa when expressed in mammarian cells, but this changes to 22-30 kDa after treatment with a reducing agent (Tominaga et al., 1990; Karasuyama et al., 1988; Tsujimoto et al., 1989; Mita et al., 1989b). Thus

150

KIYOSHI TAKATSU ET AL.

IL-5 is a disulfide-linked dimer; it has been shown that conserved cysteine residues cross-link the monomer in an antiparallel arrangement (Minamitake et d.,1990).The large variation in molecular weight of mIL-5 as well as hIL-5 is predominantly a result of the heterogeneous addition of carbohydrate which can be removed without reducing activity (Tominaga et al., 1990; Proudfoot et al., 1990,1991). The dimer formation is essential for expressing biological activity since the monomeric IL-5 produced by reduction and alkylation of IL-5 (Takahashi et al., 1990; Minamitake et al., 1990) or by mutating the cysteine residue to threonine (McKenzie et al., 1991a) failed to exert its activity. Importance of the dimer formation for expressing IL-5 activity was further elucidated by analysis of the crystal structure of hIL-5. The crystal structure of hIL-5 at 2.4 %, resolution has been reported (Milburn et al., 1993). Human IL-5 has been revealed to be a novel two-domain structure, with each domain showing a similarity to the cytokine fold found in granulocyte-macrophage colony-stimulating factor (GM-CSF), macrophage colony-stimulating factor (M-CSF), IL-2, IL-4, and growth hormone receptor: four a-helices and two p-sheets. The importance of the cytokine fold in expressing biological activities has been proposed (Bazan, 1989,1990).Human IL-5 (hIL-5) also has a unique intimate dimer configuration in which each domain contains secondary structural elements from both chains. The core of hIL-5 is two left-handed bundles of four a-helices, laid end to end, and two short antiparallel p-strands packed on opposite sides of the dimer (Fig. 1).The C-terminal strand and helix of one chain complete a bundle of four a-helices and a /3-sheet with the N-terminal three ahelices and one strand of the other chain. The tertiary structure of hIL-5 explains that IL-5 absolutely requires dimerization in expressing its biological activities, in agreement with the results obtained previously. Human IL-5 and mIL-5 share 70%sequence homology at the amino acid level. Whereas mIL-5 and hIL-5 are equally active in human eosinophil assays, hIL-5 is 100-fold less active than mIL-5 in mouse cell assay. Functional analysis of human/mouse IL-5 chimeric proteins showed that the replacement of the last 36 residues in the carboxyterminus of mIL-5 produced a hybrid with biological activity comparable to mIL-5 (McKenzie et al., 1991b). In this domain, there are only eight residues that differ between hIL-5 and mIL-5 with only three nonconservative amino acid substitutions. Cyanogen bromide treatment of hIL-5, which causes the removal of eight amino acids from the carboxy-terminus of hIL-5, leads to a complete loss of biological activity (Kodama et al., 1991). These results indicate that the IL-5R

151

IL-5 AND ITS RECEPTOR SYSTEM

IL-5

GM-CSF

hGH

FIG.1. Schematic illustration oftertiary structure ofhuman IL-5 molecule in comparison to hGM-GSF and human growth hormone (hGH). Boxes and arrows indicate ahelices and 8-strands, respectively. Drawing was prepared based on the analysis of the crystal structure of hIL-5 (Milburn et al., 1993).

a-chain binds IL-5 via a site located at the carboxyl terminus of each monomer. Localization of unique receptor-binding domains on a series of human-mouse hybrid GM-CSF and IL-5 has been determined by Shanafelt et al., (1991). Their results show that the presence of a few key residues in the amino-terminal a-helix of each ligand is sufficient to confer specificity to the interaction. Essentially similar results have been shown by different approaches (Lopes et al., 1991,1992). C. FUNCTIONAL PROPERTIES OF IL-5 1 . Regulation of Ig Production The effect of IL-5 on Ig production has been investigated in detail in the mouse. Mouse IL-5 induces polyclonal IgM, IgG1, and IgA production with different mechanisms (Koyama et al., 1988; Rasmussen et al., 1988; Takatsu et al., 1988). IL-5 stimulates polyclonal IgM production in BCL, and in in uiuo-activated B-cell blasts, by inducing an increase in the level of mRNA of secreted forms of the p-chain (Matsumoto et al., 1987; Migita et al., 1991).Relatively low concentrations (30 ng/ml) of cyclosporin A can significantly inhibit IL-5-induced IgM production by BCL, cells (Mita et al., 1991).Mouse IL-5 markedly increases the frequencies of both proliferating and antibody-secreting clones (Alderson et al., 1987; Karasuyama et al., 1988). Webb et al. (1991), using the BCg3R-ld transfected cell line which produces

152

KIYOSHI TAKATSU ET AL.

antibody of a T15 idiotype. A region of between -250 and -125 base pairs, lying 5' of the Ig transcription start site has been identified that is necessary for the induction of increased p-mRNA levels by phosphocholine-coupled antigen plus IL-5. These results suggest that IL-5 plus Ag induces several DNA-binding proteins. The regulation of IgA expression in B cells is influenced by Tcell-derived lymphokines. Increased IgA synthesis in committed B cells in the absence of proliferation has been reported. Mouse IL-5 can induce antigen-specific and polyclonal IgA production by antigenprimed B cells and lipopolysaccharide (LPS)-stimulated B cells, respectively (Beagley et al., 1988; Bond et al., 1987; Coffman et al., 1987; Harriman et al., 1988; Kim and Kagnoff, 1990; Li et al., 1992; McKenzie et al., 1987; Matsumoto et al., 1989; Murray et al., 1987; Sonoda et al., 1989,1992). IL-5 acts on surface IgA-positive (sIgA+)B ceIls, but not on sIgA- B cells, to induce IgA production (Matsumoto et al., 1989). Thus mIL-5 appears to act on B cells committed to becoming IgA-secreting cells. Mouse IL-5 augments IgA secretion by sIgA+ B cells through the induction of IgA secretion, but not through the proliferation of sIgA+ cells (Harriman et al., 1988; Schoenbeck et al., 1992; Beagley et al., 1988). Intriguingly, IL-5 slightly enhanced 3H-thymidine incorporation by sorted sIgA' B cells compared to control cultures, and cell recoveries in IL-5-containing cultures are always better than those in control cultures irrespective of the presence of LPS (Sonoda et al., 1992), suggesting that IL-5 may support prolonged survival of B cells. Transforming growth factor-pl (TGFP) was also shown to enhance IgA production by LPS-stimulated mouse B cells (Chen and Li, 1990; Coffman et al., 1989b; Kim and Kagnoff, 1990; Sonoda et al., 1989) and synergized with IL-5 to induce IgA synthesis (Sonoda et al., 1989). TGF-P acts on sIgA- cells to induce class-switching from p- to achains determined by expression of sterile a-chain transcripts andsIgA expression and also enhances ILdR expression, whereas IL-5 induces B cells, in which the p-chains have already switched to a-chains, to differentiate into IgA-secreting cells (Lebman et al., 1990a, b; Sonoda et al., 1989,1992).TGFPILPS stimulation-induced switching from pto a-chains has been confirmed by the appearance of germline Ca transcripts, by the analysis of IgA-specificswitch circular DNA (Matsuoka et al., 1990; Iwasato et al., 1990), and by the frequency analysis of IgA-secreting cells (Kim and Kagnoff, 1990). IL-2 also acts on LPSstimulated B cells synergistically with TGFp to augment IgA production without affecting the total number of sIgA+ cells (Lebman et al., 1990; Lebman and Coffman, 1990). This is in sharp contrast to the

IL-5AND ITS RECEPTOR SYSTEM

153

observation that IL-5 causes a significant increase in the number of sIgA+ B cells, up to 17%, in LPS-stimulated B cell cultures with TGF6. IL-5 induces neither a germline Ca transcript nor the formation of IgA-specific switch circular DNA (Yoshida et al., 1990; Sonoda et al., in preparation). TGF-/3 enhances not only IgA but also IgG2b and to a lesser extent IgG3 production by LPS-stimulated B cells at optimum conditions (Sonoda et al., 1992; McIntyre et al., 1993). In this case, IL-5 does not synergize with TGFp for IgG2b or IgG3 synthesis. The addition of anti-TGF-P antibody to LPS-stimulated culture decreased the production of IgG2b and IgG3 (Snapper et al., 1993; Sonoda et al., 1992),the major isotypes induced by the stimulation of LPS. These results indicate that TGF-P is class-switching factor for p- to a-chains and p- to y2b-chains as well, and IL-5 can synergize with TFGp only for IgA synthesis. IL-5 also promotes IgGl and IgE secretion in the presence of IL4. Although IL-4 plays an important role in the regulation of p- to ylchain or p- to &-chainswitching, in combination with IL-4, IL-5 induces marked accumulation of productive yl- and &-chaintranscripts (Purkerson and Isakson, 1992). These observations suggest that IL-5 may promote switch recombination, enhance transcription of rearranged yl- and &-chainloci, or increase the stability of the VDJ-Cyl and VDJ-CE mRNA. It has been controversial whether hIL-5 is able to induce Igsecretion of human peripheral B cells stimulated with Staphylococcus aureus Cowan I (SAC) or pokeweed mitogen (Azuma et al., 1986; Yokota et al., 1987;. Clutterbruck et al., 1987). It has been clearly demonstrated that hIL-5 augments Ig production of SAC- or pokeweed mitogen-stimulated human B blasts in the presence of IL-2 and also increases the secretory form of p-chain mRNA expression (Morikawa et al., 1993).In the system of Morikawa et al., (1993),IL-2 is indispensable for IL-5 activity. Similar results were shown by Bertolini et al. (1993). They demonstrated that hIL-4 synergizes with hIL-5 in augmenting IgM secretion, while hIL-5 synergizes with hIL-2 to augment IgM, IgG, and IgA secretion, and that human B-cell responsiveness to IL-5 is not long lived after SAC stimulation. This observation may explain the negative results described above, because the authors added human IL-5 to B cells 3 days after SAC activation, when human IL-5 responsiveness had diminished to low levels. Huston et al. (1993) established an elegant system for assaying for hIL-5 on human B cells in the absence of hIL-2. They stimulated purified peripheral human B cells with Moraxella catarrhalis (Mcat) or SAC and then cultured with hIL-5. Mcat but not SAC stimulation induced B-cell respon-

154

KIYOSHI TAKATSU ET AL.

siveness to hIL-5 for IgM synthesis. Their results demonstrate differential responsiveness to IL-5 by SAC- or Mcat-activated human B cells and provide evidence of a direct role for hIL-5 in normal human Bcell terminal differentiation. The inability of freshly isolated B cells to proliferate or produce Ig with hIL-5 alone may suggest the need for a prior activation signal to induce hIL-5 responsiveness. The failure of SAC to induce hIL-5 responsiveness suggests that, as with mouse B cells, the nature of the activation signal is critical in determining human B-cell responsiveness to hIL-5. 2. Regulation of B-Cell Development B cells are generated from pluripotent hematopoietic stem cells in bone marrow that also generate myeloid and erythroid elements (Kincade, 1981), but decisive commitment to the B-cell lineage is championed by the V-D-J recombination that brings variable (V), diversity (D), and joining ( J ) segments of Ig-heavy-chain (IgH) genes together to create a complete V gene (Tonegawa, 1983).Pro-B cells, the earliest stage of B-cell development, contain the germ-line pattern of IgH genes and express B-cell lineage characteristic surface phenotypes (Rolink and Melchers, 1991; Ehlich et al., 1993).Pro-B cells can differentiate into pre-B cells accompanied by the rearrangement of IgH genes and expression of cytoplasmic p-chain. Pre-B cells proliferate, rearrange light (L)-chaingenes, and express surface IgM (sIgM) giving rise to mature B cells. Eighteen different IL-5- and stromal-cell-dependent Ly-l+ precursor B-cell lines have been established from long-term cultures of young bone marrow cells on a line of stromal cells (ST2) in the presence of IL-5 (Takatsu and Tominaga, 1991; Takatsu et al., 1992; Tominaga et al., 1989). They show a germ-line configuration of the IgH genes and their surface phenotype is B220+, Lyb-2+, cIgM-, Mac-1-, Thy-1-, and IL-BRa+, indicating that they are pro-B cells. Expression of the high-affinity IL-5R by these lines indicates that IL-5 can act directly on their progenitors. After treatment with 5-azacytidine, followed by coculture with ST2 cells and IL-5, they differentiate to Ly-l+ cells expressing either sIgM (indicating that they have become B cells) or Mac-1 (indicating that they have become macrophages) (Katoh et al., 1991). When maintained by coculture with ST2 cells and IL-5, pro-B cells differentiate to pre-B cells, with rearranged IgH genes, that express Cp-, A5-, and c-fms-mRNA. Culture of these pre-B cell lines with ST2 cells and GM-CSF, however, results in the development of Ly1 and Mac-1+ macrophages which had lost the expression of surface p-chain (Katoh et al., 1990). They are nonspecific esterase-positive +

IL-5 AND ITS RECEPTOR SYSTEM

155

and can ingest latex beads. A converted myeloid line had the identical rearrangement pattern as its lymphoid parent and adhered to culture flasks. These results emphasize the close relationship between the myeloid and the pre-B lymphoid Ly-l+ pathways. Our results also suggest that B-lineage cells maintained on ST2 in the presence of IL5 may have a potential distinct from those maintained in the absence of IL-5. Several pre-B-cell lines have been recognized to differentiate either Ly-1 B cells or macrophage/monocytes (Boyd and Schrader, 1982; Klinken et al., 1988). These observations support evidence for a close relationship between the myeloid and Ly-1+ B-cell pathways of differentiation and indicate that ILS-dependent clones are Ly-1+ multipotential intermediates in the differentiation process from pro-B to B cells and macrophages. We have also established an IL-5-dependent and stromal cell-independent pre-B-cell line, T88-M, whose growth is sensitive to IFN-y, and TGFP, but not to cyclosporine A (Hitoshi et al., 1989; Mita et al., 1991b; Tominaga et al., 1989). JH

+

3. Znduction of ZL-2 Receptors It was long debated whether there exists a cytokine to induce IL2R on B cells as well as on T cells. When DNP-primed B cells are cultured in the presence of IL-5, they express high-affinity IL-2R with increases in the steady-state level of mRNA for IL-2Ra and become responsive to IL-2 for Ig secretion (Harada et al., 1987b). Similar findings are reported independently by Nakanishi et al. (1988) using a monoclonal B-cell line (BCL1-CL-3) and by Loughnan et al. (1987) using resting B cells. IL-5 does not induce the increase of IL-2Rp expression on B cells, whereas anti-Ig stimulation does (Nakanishi et al., 1990). IL-5 has an activity for killer helper factor by which generation of cytotoxic T lymphocytes (CTL) from their precursors in peanut agglutinin-binding thymocytes is induced in the presence of stimulator cells and suboptimal doses of IL-2 (Takatsu et al., 1987). IL-5 also markedly increases the expression of IL-2Ra mRNA in thymocytes which had been stimulated with hapten-modified spleen cells and IL-2. At least four different lymphokines seem to have killer helper activity; IL-5, factor from a T-cell hybridoma, 2Y4, IL-4, and IL-6 (Takatsu and Kikuchi, 1988). Human IL-5 also induces an increase in the levels of expression of IL-2Ra on human T cells (Noma et al., 1987), enhances generation of human lymhokine-activated killer cells (Aoki et al., 1989) together with IL-2, and augments CTL generation in the presence of IL-4 (Nagasawa et al., 1991).

156

KIYOSHI TAKATSU ET AL.

4 . E osinophil Production and Basophil Function Eosinophil differentiation factor (EDF)was shown by cDNA production and cloning to be IL-5 (Campbell et aE., 1987); this explains its novel property of promoting B-cell growth and differentiation activity (reviewed by Sanderson, 1992). Eosinophil production can be induced in vitro by IL-3, GM-CSF, and IL-5. Both IL-3 and GM-CSF have activities on other hematopoietic lineages, whereas IL-5 is specific for the eosinophil/basophil lineage (Denburg, 1992). Mouse IL-5 acts on eosinophilic progenitor cells in bone marrow for their proliferation and maturation (Y. Yamaguchi et al., 1988a). Primitive mouse stem cells can also differentiate into mature eosinophils in the presence of IL-5 together with G-CSF. Similar synergy of hIL-5 and G-CSF was shown for eosinophil colony formation from human bone marrow cells and human cord blood cells (Enokihara et al., 1989; Saito et al., 1988). Human IL-7 was reported to support eosinophil colony formation from human bone marrow cells that can be abrogated by the addition of anti-IL-5 mAb (Vellenga et al., 1992), indicating that the supportive effects of IL-7 on eosinophil precursors are mediated by the endogeneous release of IL-5. Mouse IL-5 supports the viability of mature eosinophils in vitro and the production of superoxide anion and has chemotactic activity for and activates eosinophils (Y.Yamaguchi et al., 1988bJ991). Human IL-5 also has chemotactic activity for eosinophils (Lopez et al., 1988;Enokihara et al., 1989).The hIL-5-cultured eosinophils are activated and degranulated more readily than freshly isolated eosinophils (Kita et al., 1992). Human IL-5 also acts on basophils to enhance the release of mediators such as histamine and leukotriene (Bischoff et al., 1990; Hirai et al., 1990). It was shown that IL-5 increased the expression of the integrin C D l l b on human eosinophils (Walsh et al., 1990), and this increased expression was accompanied by an increased adhesion to endothelial cells. Adhesion was inhibited by antibody to C D l l b or CD18, suggesting that the integrins are involved in eosinophil adhesion to endothelial cells. These functions of IL-5 are reported to contribute to the urgent mobilization of eosinophils during helminthic infection or an allergic response. D. EXPRESSION OF IL-5 mRNA Lymphokine genes such as IL-2, IL-3, IL-4, IL-5, and GM-CSF are coordinately expressed during T-cell activation. The recognition by the T-cell receptor of antigenic peptide associated with the MHC on antigen-presenting cells or on B cells triggers a series of biochemical events, including elevation of the intracellular calcium ion concentra-

IL-5 AND ITS RECEPTOR SYSTEM

157

tion and activation of protein kinase C (PKC). These processes can be mimicked in part by phobol myristate acetate (PMA) and the calcium ionophore (Crabtree, 1992). Many, if not all, lymphokine genes are induced by phorbol ester plus calcium ionophore; however, detailed studies revealed that the signals for optimal induction of the lymphokine genes are not exactly the same for each gene. In mouse, constitutive expression of 1.7-kb IL-5 mRNA is detected in an IL-5-producing T-cell hybridoma and is augmented by the stimulation with PMA plus calcium ionophore (Tominaga et al., 1988). Among several cell lines initially analyzed for IL-5 mRNA, the EL4 (thymoma of TH%type) and D9 (cloned T-cell line ofTH%type) express mIL-5 mRNA with IL-5 production upon stimulation with PMA and Con A, respectively (Tominaga et al., 1988). None of murine tumor cell lines such as macrophage monocytic cell line, myelomonocytic cell line, BCL1, or myeloma cell line express mIL-5 mRNA. T cells from M. tuberculosis-primed mice and nematode-infected mice can express mIL-S mRNA, with IL-5 production, upon stimulation with Con A or FPD and nematode antigens, respectively (Tominaga et al., 1988; Y. Yamaguchi et al., 1990~). In the former system, IL-1 is also required for inducing IL-5 production by T cells (Tominaga et al., unpublished observation). Normal spleen cells stimulated with Con A show a little expression of the mRNA. An undetectable level of mIL-5 mRNA expression is observed in PMA plus calcium ionophorestimulated nonprimed spleen cells in which remarkable mIL-2 mRNA expression is observed. These results imply that only a subset of T H cells expresses mIL-5 mRNA. Once T cells are activated, IL-2 induces mIL-5 mRNA expression (Y. Yamaguchi et al., 1990b). It has been clearly demonstrated by Chenvinski et al., and Mossman et al., that IL-5, IL-4, and IL-10 can be produced by a subset of T cells (TH2)distinct from that of T cells ( T H 1 ) which can produce IL2, IFN-7, and TNF-/3 (Chenvinski et al., 1987; Mossman et al., 1986; Mossman and Coffman, 1989). Our results appear to fit with their hypothesis, although alternative explanation cannot be ruled out. Several lines of evidence indicate that TH1 and TH2 cells use different signal transduction pathways resulting in the characteristic pattern of lymphokine production. For example, TH1and TH2cells exhibit distinct responses to prostaglandin E2 (Betz and Fox, 1991) and cholera toxin (Munoz et at., 1990) which elevate intracellular CAMP.Recent results support the notion that IL-12 and IL-4 play important roles for TH1 and TH2 development, respectively (Hsieh et d., 1993; Paul et al., 1994). As mentioned, IL-4 and IL-5 are often coexpressed in clones desig-

158

KIYOSHI TAKATSU ET AL.

nated TH2.However, anti-CD3 induces the expression of mIL-4, mIL-

5, and mGM-CSF mRNA expression in mouse T cells, whereas treat-

ment with mIL-2 induced mIL-5 mRNA expression but did not induce detectable mIL-4 or mGM-CSF (Bohjanen et al., 1990).In this respect, it is interesting that 8-bromo CAMP, IL-1, or prostaglandin E2 enhances IL-5 production in TH2 cells (Betz and Fox, 1991; Munoz et al., 1990). Lee et al. (1993) have assayed the mIL-5 promoter activity using a transient transfection system and clarified that the mIL-5 production at the mRNA and protein levels can be increased more than 10-fold in EL-4 cells upon stimulation with PMA plus dibutyryl CAMP. Moreover, dibutyryl CAMP activated the mIL-5 promoter more than 10-fold in a PMA-dependent manner. Two signals, PMA and CAMP, may be required for optimal activation of the mIL-5 promoter in T,2 cells. It should also be noted that cultured mast cell lines expressed mIL-5 mRNA besides IL-3 and GM-CSF mRNA by cross-linkage of FcsRI or upon stimulation with calcium ionophore (Plaut et al., 1989). The hIL-5 mRNA expression is reported in HTLV-1-infected T cells (Noma et al., 1987), in Reed-Sternberg cells, and in Hodgkin’s disease with eosinophiIia (Samoszuk and Nansen, 1990). Human B cells that have been transformed by the Epstein-Barr (EB) virus constitutively produce IL-5 (Paul et al., 1990). Furthermore, hIL-5-producing EB virus-transformed B cells specifically bind hIL-5, are growth-inhibited by anti-IL-5 mAb, and are capable of responding to IL-5 by augmented proliferation (Bauman and Paul, 1992). In contrast, no stimulatory effect of hIL-5 is noted on Burkitt’s lymphoma cell lines. These results strongly support a role for hIL-5 in autocrine growth factor of EB virusinfected B cells. Eosinophils infiltrating into the mucosa of patients with active coeliac disease have been shown to express the hIL-5 mRNA (Desreumaux et al., 1992). No positive signal are reported in normal duodenum tissues. Moreover, purified blood eosinophils from some but not all patients with eosinophilia are also shown to express the IL-5 mRNA (Desreumaux et al., 1992). These results also suggest that eosinophils may have the capacity to synthesize IL-5 and IL-5 plays a role in autocrine growth factor of eosinophils. There is ample evidence showing that disregulated IL-5 production is involved in eosinophilia in patients with idiopathic hypereosinophilic syndrome. For example, significant levels of hIL-5 are detected in the serum of patients with idiopathic hypereosinophilia (Owen et aZ., 1989) and patients with the eosinophilic myalgia syndrome resulting from the ingestion of L-tryptophan (Owen et al., 1990). Patients with eosinophilia associated with Hodgkin’s disease are found to have IL5 mRNA in the tumor cells (Samoszuk and Nansen, 1990),and a patient

IL-5 AND ITS RECEPTOR SYSTEM

159

with angiolymphoid hyperplasia with eosinophilia (Kimura’s disease) is also found to have constitutive expression of IL-5 mRNA in lymph node tissue Inoue et al., 1990). In patients with the syndrome of episodic angioedema and eosinophilia, IL-5 levels peaked several days before maximal eosinophilia and then decline (Butterfield et al., 1992). In some patients, elevated serum IL-5 was found during attack. So production ofIL-5 is likely an important determinant ofthe pathophysiology of episodic angioedema and eosinophilia. In many patients suffering from diseases with eosinophilia, serum IL-5 levels were elevated (Matsumoto et al., 1992; S . Yamaguchi et al., 1994). 111. Receptors for 11-5

A. STRUCTURE OF THE IL-5 RECEPTOR The IL-5R is expressed mainly on naturally activated B cells, eosinophils as well as ILJ-dependent cell lines (Mita et al., 1988,1989a; Rolink et al., 1989). In mouse cells, IL-5-responsive cell types express relatively small numbers of high-affinity IL-5R (Kd of 10- 150 pM) and large numbers of low-affinity ILdR ( K d of 2- 10 nM). The concentration of mIL-5 necessary to elicit a response and the detection of the high-affinity ILdR on IL-5-dependent cell lines indicate that biological responsiveness to mIL-5 depends on interaction with the highaffinity IL-5R. Cross-linking of sIgM of mouse B cells with anti-Ig enhances the expression of the high-affinity IL-5R (Allison et al., 1991; Rolink et al., 1990). Freshly prepared human eosinophils and eosinophilic cell lines express a single class (high affinity, Kd of 170330 pM) of hIL-5R (Chihara et al., 1990; Ingly and Young, 1991; Migita et al., 1991; Plaetinck et al., 1990),which is enhanced on eosinophilic cell lines by treatment with sodium butylate (Plaetinck et al., 1990). Chemical cross-linking of mIL-5R with IL-5 reveals that the highaffinity mIL-5R consists of two distinct polypeptide chains (a,p60, and p, p130). Affinity cross-linking experiments revealed that hIL-5 binds 55- to 60-kDa proteins on eosinophils. Anti-mIL-5R mAb’s H7 and R52.120 that recognize a and p, respectively, have been prepared (Hitoshi et al., 1990; Rolink et al., 1989; N. Yamaguchi et al., 1990). H7 mAb specifically inhibits IL-5-induced proliferation, while 1152.120 mAb partially inhibits both IL-5- and IL-3-induced proliferation of a pre-B-cell line, B13, that responds to both mIL-5 and mIL3. The surface-staining analysis using H7 and R52.120 mAb’s revealed that mIL-5R is expressed on more than 70% of peritoneal B cells, most of which are Ly-l+ and on 4-8% of splenic B cells, most of which are

160

KIYOSHI TAKATSU ET AL.

Ly-1- (Hitoshi et al., 1990, 1992, 199313; Wetzel, 1989). In contrast, the &chain is expressed not only on IL-5-dependent cell lines, but also on IL-3-dependent cell lines (Mita et al., 1991a). 1 . Structure of the IL-5 Receptor a-Chain cDNAs encoding mIL-5Ra have been isolated from a mouse IL-5dependent early B cell line, Y16 (Takaki et al., 1990). The mIL-5Ra chain is a type-I membrane protein of 415 amino acids, including an amino-terminal signal peptide, a glycosylated extracellular domain, a single transmembrane segment, and a cytoplasmic tail. We purified mIL-5Ra from cell lysates of IL-5-responsive Y16 cells by using an anti-mIL-5Ra mAb-coupled affinity column and determined its aminoterminal sequence of 17 residues (N. Yamaguchi et al., 1991). The results revealed that N-terminal amino acid sequencing of mIL-5Ra revealed a sequence identical to that deduced from the nucleotide sequence of the cDNA and confirms the inferred amino-terminus of the mature protein (N. Yamaguchi et al., 1991). The extracellular domain of the m I L J R a contains two motifs that are characteristic of a set of a receptor for cytokines, growth hormone and prolactin (Bazan et al., 1990; reviewed by Miyajima et al., 1992; Nicola and Metcalf, 1991; Taga and Kishimoto, 1992), a particular spacing of four cysteine residues in the amino-terminal half of the region and the tryptophanserine-X-tryptophan-serine (where X is any amino acid) motif located close to the transmembrane domain. In addition, the extracellular region comprises three tandemly repeated sets of a fibronectin type I11 domain (Patthy, 1990). The cytoplasmic domain does not contain the consensus sequences for either a tyrosine kinase domain or a catalytic domain of protein kinases, but shows homology with a part of the actin-binding domain of P-spectrin (N. Yamaguchi et al., 1991). The cytoplasmic domain has regions rich in proline following the transmembrane domain that are well conserved between IL-5Ra as well as among receptors for IL-3, GM-CSF, prolactin, and growth hormone (Takaki et al., 1991) (Figs. 2 and 3). This region’s cDNAs of hILSRa also have been isolated from human eosinophils and a cell line, HL60 (Murata et al., 1992; Tavernier et al., 1991). The entire nucleotide sequence of hIL-5Ra chain cDNA shows considerable similarity to the coding sequence of the mIL-5Ra chain, and the amino acid sequence has about 70% homology with the mIL-5Ra chain and retains features common to the cytokine receptor superfamily. The cytoplasmic regions rich in proline following the transmembrane domain are also well conserved. cDNAs coding for soluble forms of both mIL-5Ra and hIL-5Ra chains have also been isolated (Murata et al., 1992; Takaki et al., 1990; Tavernier et al., 1991).

"

IL-5 AND ITS RECEPTOR SYSTEM

Mouse IL-5R

I'

Human

IL-5R

IL-3R

GM-CSFR

GM-CSFR

161

["I IL-3R

FIG.2. Schematic illustration of molecular components of ILdR, IL3R, and GMCSFR. Conserved cytokine receptor superfamily motifs, four cysteine residues and WSXWS in the extracellularregions and proline-rich residues in the cytoplasmic portion are shown in the figures. Detailed are described in the text.

COS7 cells transfected with the mIL-5Ra cDNA express p60, which binds IL-5 with low affinity.Interestingly,however, the IL-5 nonbinding but IL-3-dependent cell line FDC-P1, binds IL-5 with both high and low affinity and proliferates in response to IL-5 when transfected with mIL-5Ra chain cDNA (Takaki et al., 1990); this indicates that

162

KIYOSHI TAKATSU ET AL.

IL-5R

mOuse~HLWTRLFPPVPAPKSNIKDLPWTEYrKPSNETKIEWHCVEEVGFEVMGNSTF" human~HLWIKLFPPIPArKSNIKDLPVTTNYEKAGSSETEIEVICYIEKP~VETLE~SV~~

IL-3R

mouse RKSLLYWPPIPRLRLPLGEIWVh'EPALEDCEVTPVTllA* human ~ R R Y L V M Q R L F P R I P H M K D P I G D S ~ Q N D K L W W E A G K A G L E ~ C L V T E V Q W Q K ~ "

GM-CSFR mouse ~ R R F E V T R R L F P P I P G I R D K V S D D V R V N P E T L R K D L L O P '

human~LRFLRIQRLFPPVPQIKDKLNDN!iEVEDEIIWEEFTPEEGKGYREEVLTVKEIT'

PRL-R

mouse~KGYSMMTC1FPPVPGPKIKGFDTHLLEKGKSEELLSALGCQDFPPTSDCEDI.LVFFLEVD ~ . . h u m a n ~ K G Y S M V T C I F P P V P G P K i K G F D A l i L L E K G K S E E L L S A L G C ~ D F P P T S D Y E D L L V E Y L ~ V .~ .r J.

GH-R

m o u s e ~ K ~ O R I K M L 1 L P P V P V P K i K G I D P D L L K E G K L E E V N T I L t i I H D N Y K P D F Y N D D S W V E F : E L D. . . h u m a n ~ K Q ~ R I K M L I L P P V P V P K I K G I D P I ) L L K E G K L E E V N T I L A I l l D S Y K P E F H S D D S ~ E F.~.~.! D

FIC.3. Conserved regions rich in proline residues in the cytoplasmic domain of the IL-5Ra chain. Conserved regions rich in proline residues and basic and acidic amino acids were depicted as bold capital. TM, transmembrane domain; PRL-R, prolactin receptor; GH-R, growth hormone receptor.

mIL-5Ra-negative FDC-P1 cells constitutively express the additional I L d R component ( IL-5RP chain) that confers high-&nity binding when expressed with mIL-5Ra chain. Recombinant hIL-5Ra chain expressed on COS7 cells has a molecular weight of approximately 60 kDa and binds hIL-5 and mIL-5 with a relatively high affinity (Kd of 300-450 pM), characteristics that are similar to those of the hIL5Ra chain expressed on eosinophils in peripheral blood.

2. Structure of the IL-5Ra Gene The chromosomal gene of the mIL-5Ra has been cloned and analyzed and its genomic structure and chromosomal localization determined (Imamura et al., 1994).The mIL-5Ra genomic gene is divided into 11exons and 10 introns and spans more than 35 kb. The nucleotide sequences of all exons are in agreement with that of the mIL-5Ra cDNA. Exon 2 encodes 5'-untranslated sequences, signal peptides, and NH2-terminal amino acids of mature protein. Exon 3 to exon 8 encode the extracellular region that is composed of three fibronectin type I11 (FBN 111)-likedomains each of which is composed of a pair of two exons. Exons 5 and 6 encode a pair of cysteine residues which is conserved in a cytokine receptor family. Exon 8 contains the "WSXWS" motif. The transmembrane domain is in exon 9. Exons 10 and 11 encode the cytoplasmic domain. This gene organization revealed a pattern of considerable structural homology with genomic genes for IL-2RP (Shibuya et at., 1990), AIC2A and AIC2B (Gorman et al., 1992), IL4R (Wrighton et al., 1992), IL-7R (Pleiman et al., 1991), and EpoR (Youssoufianet al., 1990).The soluble forms of mIL5Ra (s-mIL9Ra) appear to be produced by lacking exons 9 and 10 (exon 8 spliced to exon 11)or exon 9 (exon 8 spliced to exon 10) by alternative splicing.

.

IL-5 AND ITS RECEPTOR SYSTEM

163

Several potential factor-binding sites in the mIL-5Ra gene can be assigned such as an AP-1-binding site (TGACTCA), a Spl site (CCCGCC), a NF-kB site (GGGAATCTTC),and a NF-IL-6 sequence (TGATGAAAG). Two GATA-1 recognition sites (AGATAA/G), two PU. 1 sites (GAGGAA), and six IFN-y-responsive elements (C(T/A) (G/T)(G/T)ANN (C/T) are also notable. Chromosomal localization of the mILSRa gene was mapped on the distal half of chromosome 6 (Gough and Raker 1992; Imamura et al., 1994). The chromosomal localization of the hILdRa gene has also been mapped on the human (Isobe et al., 1992).The results suggest that chromosome 3 (3~24-3p26) the IL-5Ra gene is unlinked to other members of the hematopoietic receptor family. 3. Soluble Forms of ZL-5Ra Chain The mRNA expression for s-mIL-5% is detected in spleen cells, peritoneal exudate cells, and cell lines bearing mIL-5R (Takaki et aZ., 1990; Kikuchi et al., 1993). COS7 cells transfected with the s-mIL5Ra cDNA secreted the s-mIL-5Ra protein into media, indicating that s-mIL-5Ra would be secreted by cells expressing the mRNA coding for s-mIL-5Ra in uiuo. To investigate the immunological role of s-mIL-5Ra7expression of large amounts of s-mIL-5Ra in insect cells, by using baculovirus expression system, have been completed and established the ELISA system for detection of s-mIL-5Ra (Kikuchi et al., 1993). Purified recombinant s-mIL5Ra retains the IL-5 binding ability with low affinity, possibly comparable to that of membranebound mILSRa (Y. Yamaguchi et al., 1991; Kikuchi et al., 1993). However, it fails to inhibit the proliferation of IL-5-responsive Y 16 cells which could be triggered by the signal generated through the high-affinity mIL-5R. FDC-P1 which expresses mIL-5RP but not mIL5Ra never responded to IL-5 in the presence of s-rnILSRa, suggesting that the s-mILdRa/mIL-5 complexes can neither interact with mIL5RP nor trigger biological responses. These results are sharp contrast to those obtained with the IL-G/IL-GRa (p80) system in which the IL-G/sIL-GRa complex can bind to its signal transducer, gp130, and trigger biological responses (Hibi et al., 1990; reviewed by Kishimoto et al., 1992). No measurable levels of s-mIL-5Ra (less than 0.5 ng/ml) were detected in normal mice sera, while significant s-mIL-5Ra was detected in the sera from NZB, NZW, and BCL,-bearing mice. Soluble forms of IL-2R, IL-4R7 IL-GR, IL-7R, IFN-yR, and TNFR have also been detected in biological fluids in humans and animals (reviewed by Maliszerski and Fanslow, 1990). The secreted forms of these cytokine receptors are thought to be generated by proteolytic cleavage of relevant membrane-bound receptors or by the translation of alternatively

164

KIYOSHI TAKATSU ET AL.

spliced mRNAs. While the mechanisms for the production of s-mIL5Ra are not clear, the gene structure of mIL5Ra may support the alternative splicing pathway. The role of s-mIL-5Ra in oioo, whether it would work positively or negatively for these cells, is not clear. The s-mIL-5Ra might modulate IL-5 effects on hematopoiesis and immune responses or might protect IL-5 from proteolytic cleavage and prolong their half-life in uiuo. In human, Tavernier et al. (1992) have shown that two forms of s-hIL-5Ra (slhIL-5Rh and s2hIL-5Ra) are generated by a mechanism different from that in the mouse. One is generated by a normal splicing event using the first splice acceptor site coupled with the extaractcellular domain to a specific small exon and the other is generated without splicing (Tavernier et al., 1992).In contrast to s-mIL-5Rq the recombinant s-hIL-5Ra and the fusion proteins of hIL-SRcu and Fc domain of immunoglobulin (hILdRa-Fc) bind hIL-5 with high affinity and inhibit the biological activities of hIL-5 (Devos et al., 1993). Devos et al. also showed that s-hILdRa can bind dimerized hIL-5 in a 1: 1 ratio. These observations support the possibility that s-hIL-5Ra may act as a potent regulator for hIL-5, if the protein can be secreted enough to compete with membrane-bound IL-5Ra. However, there are no reports whatsoever to show the existence of s-hILdRa in human biological fluids. The differences between s-mILdRa and s-hIL-5Ra in the mechanism for production and ligand-binding affinity might suggest a different role of s-hIL-SRa from s-mIL-5Ra in function. 4 . The Structure of the IL-SRP Chain The anti-mIL-EiRP chain mAb (R52.120)established by Rolink et al. (1989) inhibited IL-5-induced proliferation by downregulation of the number and Kd of the high-affinity mIL-5R (Mita et al., 1991a). The mAb also inhibited IL-3-driven proliferation of 11-5- and IL-3-dependent B13 and IL-3-dependent FDC-P1. The anti-mIL3R (anti-Aic-2) mAb (Yonehara et al., 1990) was shown to react with IL-5-responsive cell lines and down modulate the ILJ-binding (Takaki et al., 1991). Both R52.120 and anti-mIL-3R mAb’s can immunoprecipitate a similar polypeptide (p130/ p140), whereas they do not react with mIL-5Ra chain, indicating that the p130/p140 involved in the mIL-3R system can associate with the mILdRa chain for the formation of the highaffinity mIL-SR. The protein turned out to be identical to AIC2 protein (Devos et al., 1991; Takaki et al., 1991) recognized by the antimIL-3R mAb. The anti-mIL3R mAb has been shown to recognize both the low-affinity mIL-3R (AIC2A) (Itoh et al., 1990)and its homologue (AIC2B) (Gorman et al., 1990). Finally, the high-affinity IL-5R can be

IL-5 AND ITS RECEPTOR SYSTEM

165

reconstituted by combining mIL-5Ra and AICSB, but not by mIL-5Ra and AIC2A protein on L cells (Takaki et al., 1991). AIC2B protein (mIL-5RP chain) by itself does not show any specific binding for IL-5. Further, CTLL transfectants, which express recombinant mIL-5Ra and rnIL-5RP7but not mIL-5a or mIL-5P7can proliferate in response to IL-5 (Takaki et al., 1993), indicating that the @chain is indispensible for the IL-5 signal transduction. The AIC2B protein contains motifs conserved among cytokine receptor families (Gorman et aZ., 1990) and also contribute to the /3-chain of mGM-CSFR (Kitamura et aZ., 1990, 1991; Park et al., 1992) and mIL-3R (Hara and Miyajima, 1992). COS7 transfectants expressing hIL-5Ra show relatively high-affinity IL-5 binding ( K d of -300 pM), whereas they do not respond to hIL-5 for proliferation. However, FDC-P1 transfected with the hIL-5Ra cDNA can proliferate in response to hIL-5 (Murata et al., 1992), implying, as in the mouse, the functional hIL-5R also consists of the hIL-5RaP heterodimer; however, the /?-subunit may be engaged in a way that does not contribute to its affinity as much as in the mouse system. In the human, there is only one AIC2 (mIL-5Rp) homologue, KH97 (Hayashida et al., 1990), that has been shown to be the @chain (p,) of GM-CSFR (Hayashida et al., 1990) and IL-3R (Kitamura et al., 1991). KH97 was clearly shown by cotransfection experiments to be hIL-5R&. Only CTLL-h5Ra-PCresponded to hIL-5 in DNA synthesis (Takaki et al., 1993). Neither CTLL-h5Ra nor CTLL-P, responds to hIL-5. From these results, it is clear that p, is indispensable for the function of hIL-5R. In contrast to mIL-5R, p, does not contribute much to hIL-5 binding affinity. It is supposed that binding of hIL-5 is irrelevant to the existence of p,, as the binding affinity of hIL-5Ra chain alone does not differ so much from that of hIL-5Ra.P complex in contrast to hIL-3R and hGM-CSFR systems. Tavernier et al. (1991) reached the same conclusions as ours by using KH97 and humanmouse chimeric a-subunit which has transmembrane and cytoplasmic domains of mILSRa. It is still not clear how the cross-competition of hIL-5 binding by hIL-3 or hGM-CSF occurs (Lopez et al., 1990,1991,1992). OF THE IL-5R B. FUNCTION As described in the above section, the AIC2B in mouse and p, (KH97) in human are indispensable to the construction of functional receptors for IL-5, IL-3, and GM-CSF as a signal transducer. This may account for the many overlapping biological activities of these cytokines on eosinophil production and activation. However, IL-5 has

166

KIYOSHI TAKATSU ET AL.

unique functions different from these of IL-3 and GM-CSF. We hypothesized that signals specific for IL-5 may exist and be mediated by the IL-5Ra chain. The approach we took was to elucidate a role of cytoplasmic domain of mIL-5Ra chain in biological activity of IL-5. FDC-P1 transfectants expressing mutant mIL-5Ra chain, which lacks the total cytoplasmic region, bound mIL-5 with high affinity ( K d of 10 pM ) like transfectants expressing wild-type mILdRa. However, they did not respond to mIL-5 in contrast to transfectants expressing the wild-type mIL-5Ra chain which could proliferate in response to mIL-5 (Takaki et al., 1993).Essentially identical results were obtained by using CTLL transfectants coexpressing mutant mIL-5Ra and the @-chain. The mIL-5-mediated ,growth signal is never transmitted through the high-affinity mIL-5R that consisted of mutant a and the wild-type @-chain,indicating that the cytoplasmic domain of mIL-5Ra has critical roles in transmitting IL-5 signals. Using deletion mutants of a particular cytoplasmic portion of mIL-5Ra chain, regions rich in proline residues of the cytoplasmic domain of IL-5Ra are shown to be essential for mediating IL-5 signaling (Takaki et al., submitted). The regions rich in proline rsidues of the cytoplasmic domain of mIL5Ra chain may play a crucial role in interacting with the @-chainor association with cellular proteins. The roles of mIL-5Ra and mIL-5RP on the association, dissociation, and internalization of IL-5 were analyzed using IL-5-responsive cells and transfectants expressing the mIL5Ra chain alone, @-chain alone, or both a-and @-chains(Mita et aE., 1993). The maximal binding of mIL-5 to both high- and low-affinity ILdR is rapid (within 10 min). The dissociation of mIL-5 from low-affinity IL-5R is rapid (tl12< 30 min), but that from the high-affinity mIL-5R is remarkably slower (t,/z> 120 min). Internalization of IL-5 appears to be mediated through high-affinity IL-5R. The IL-5Ra chain binds IL-5 and may be involved, in part, in the internalization of IL-5, whereas the mIL5R@chain is responsible for slowing the dissociation and the efficient internalization of IL-5 by stabilizing the ligand-receptor complex.

c. IL-SR-MEDIATED SIGNALING PATHWAY

Although the ability of IL-5 to promote growth and differentiation of B-cells and eosinophils has been known for many years, the molecular mechanism by which IL-5 binding to its receptor elicits diverse responses has remained an enigma. Accumulating evidence suggests that phosphatidyl inositol turnover, Ca2' mobilization, and activation of A- or C-kinase appear unlikely to be essential for IL-5-induced differentiation signals (N. Yamaguchi et al., 1989).Tyrosine phosphory-

IL-5 AND ITS RECEPTOR SYSTEM

167

lation is usually one of the earliest biochemical events in the signal transduction cascade induced by growth factors such as epidermal growth factor, platelet-derived growth factor, fibroblast growth factor, and colony-stimulating factor-1; in these cases, the receptor itself also has enzymatic activity (Ullrich and Schlessinger, 1990). Like other cytokine receptors, the intracellular signal transduction element of IL5Ra and p-chain lacks intrinsic kinase activity. We have reported the predominant tyrosine phosphorylation of several cellular proteins (p130, p92, p53, p45) within 5 min of exposure to IL-5 (Murata et al., 1990).Furthermore, the inhibition of tyrosine kinases with herbimycin A suppressed the IL-5 responsiveness, indicating that tyrosine phosphorylation of cellular proteins and activation of tyrosine kinases are involved in the signal-transducing pathways of IL-5. A similar set of protein phosphorylations has been reported upon stimulation with IL-3 and GM-CSF (Duronio et al., 1992; Hanazono et al., 1993; Kanakura et al., 1990; Sakamaki et al., 1992; Watanabe et al., 1993). Thus, intracellular signals induced by IL-5, IL-3, GM-CSF, and probably other cytokines are likely to use the same or similar pathways. To further envisage the IL-5-mediated signaling pathway, we evaluated the tyrosine phosphorylation of several proteins that are predicted to play a role in growth factor-mediated signal transductions. Immunoprecipitations by various mAb's followed by the Western blotting analysis using antiphosphotyrosine mAb revealed that mIL-5RP is phosphorylated within 5 min after the IL-5 stimulation of IL-5responsive cells (Satoh et al., submitted). We next investigated tyrosine phosphorylation of the molecules such as phospholipase C-y, Ras GTPase-activating protein (GAP), PI3 kinase, ~95"" (Vav), ~ 7 5 (HSl), ~ ' ~and Shc proteins. All of the molecules have SH2 (Src homology 2) and/or SH3 domains in their structure. Among these, we detected tyrosine phosphorylation of the PI3 kinase and Vav proteins 5 min after IL-5 stimulation of the IL-5-dependent pre-B-cell line, Y16. Since the product of the PI3 kinase has been reported to activate protein kinase. C-6 which is crucial for mitosis, there should be a PI3 kinase-mediated pathway for cell growth in IL-5 signaling. Tyrosine-phosphorylated Vav has been reported to function as a guanine nucleotide-releasing factor (GRF) for Ras upon T-cell receptor linkage. Furthermore, HS1 protein (D. Kitamura et al., 1990), which has been shown to be a substrate of lyn kinase and be phosphorylated by stimulation of B cells with anti-Ig antibody (Yamanashi et al., 1993), was also phosphorylated within 5 min upon stimulation with IL-5. Moreover, Shc was also phosphorylated in response to IL-5. Tyrosinephosphorylated Shc has been found to activate mSos, another GRF

168

KIYOSHI TAKATSU ET AL.

for Ras. Therefore, both Vav and Shc tyrosine phosphorylation may cause the activation of Ras after IL-5 stimulation. Intriguingly, JAK2, a 130-kDa tyrosine kinase, which has been shown to be a growth hormone receptor- and erythropoietin-associated kinase (Argetsinger et al., 1993; Witthuhn et al., 1993), was also phosphorylated (Takaki et al.,submitted).JAK2-IL-5R interaction may suggest a shared mechanism among other members of the cytokine receptor family and proteins involved in IFN signaling, for gene expressions may transduce signals for IL-5-induced cell growth. We then examined tyrosine kinases involved in IL-5 signaling. Since we could not detect tyrosine kinase activities in the immunoprecipitates obtained by anti-IL5Ra or anti-IL-5RP mAb’s, we focused on particular tyrosine kinases such as Btk (Bruton’s tyrosine kinase), Lyn, and Fyn and tried to investigate their tyrosine kinase activity and tyrosine phosphorylation upon IL-5 stimulation. We detected a two to threefold increase in the tyrosine kinase activity of the Btk 5 to 20 min after IL-5 stimulation of Y16 cells. Both Lyn and Fyn, src family tyrosine kinases, have been reported to be associated with B-cell antigen receptor (BCR) and play critical roles in BCR-mediated signaling. However, we could not observe the change of their activity after IL-5R ligation. Instead, tyrosine phosphorylation of the Lyn and the Lyn-associated 45-kDa protein was detected after the IL-5 stimulation. By using T88-MP1, which expresses high-affinity receptors for both IL-2 and IL-5, we demonstrated that IL-5 stimulation induces phosphorylation of the IL-SRP chain in a manner similar to IL-2 stimulation (Ohbo et al., unpublished observation); therefore IL-2RP functions as a substrate for kinases activated by IL-5 and can be phosphorylated in a manner similar to that induced by IL-2. IL-5 and IL-2 also stimulated similar tyrosine phosphorylation of various proteins. Interestingly, IL-5RP is not phosphorylated upon stimulation of CTLL transfectants expressing mIL-5RaP with IL-2 (Takaki et al., submitted). It remains to be determined whether the IL-2 and IL-5 signal transduction pathways utilize identical kinases or whether they utilize distinct kinases with overlapping specificities.

D. EXPRESSKON OF IL-5R Northern blot analysis revealed that the expression of two types of murine IL-5Ra mRNA’s (approximately) 5.0 kb and 5.8 kb) is detected only in IL-5R-bearing murine cell lines. Overall, the level of mRNA expressed in particular cells correlates well with the number of lowaffinity IL-5R. Analysis of the gene amplification of IL-5R cDNA by

IL-5 AND ITS RECEPTOR SYSTEM

169

PCR technique reveals the existence of variant transcripts corresponding to both membrane-bound form and soluble forms of mIL-5Ra in spleen cells, peritoneal exudate cells, and cell lines bearing IL-5R (Kikuchi et al., 1993; Takaki et al., 1990). Northern blot analysis revealed that hIL5Ra mRNA was exclusively expressed in peripheral blood eosinophils and eosinophilic cell lines (Murata et al., 1992; Tavemier et aZ., 1991). The s-hIL-5Ra mRNA was also expressed in eosinophils and its expression is much more abundant than the mRNA expression for the membrane-anchored form (Tavernier et al., 1991; Murata et al., 1992). As described, Huston et al. (1993) demonstrated that Mcatstimulated human B cells respond to IL-5 for IgM secretion. They also analyzed the human IL-5R mRNA expression. Kinetic analysis of human IL-5Ra and P-message by Mcat-stimulated B cells demonstrated upregulation af the human IL-5RP, but not human IL-5Ra, message over 48 hr. IV. Pathophysiology of Disregulated 11-5 Production and 11-5 Receptor Expression

A. ANIMALMODEL 1 . IL-5 Transgenic Mice To envisage the possible engagement of IL-5 in the development of Ly-l+ B cells and eosinophils in uiuo, two different approaches were taken. First, transgenic mice carrying the mIL-5 gene ligated with a metallothionein promoter (Tominaga et al., 1991) or human CD2 promoter (Dent et al., 1990) are produced. Second, bone marrow cells transfected with the IL-5 gene ligated with retrovirus vector are transplanted into lethally X-irradiated syngenic mice (Vaux et al., 1990). Results obtained from the two different approaches are essentially the same. The IL-5 transgenic (IL-5-Tg)mice, which have constitutive production of mIL-5, exhibit the elevated levels of serum IgM, IgE, and IgA and clearly show the induction of Ly-1 B-cell population in spleen and polyreactive IgM autoantibody production against ssDNA, dsDNA, cardiolipin, poly(ADP-ribose), and TNP. Ly-1+, IL-5Ra' B cells in IL-5-Tg mice secrete fewer autoantibodies and have fewer N nucleotides at the 3' end of the D elements compared with Ly-1- B cells (Katoh et al., 1993). The reduction in nucleotides, along with the finding that Ly-1+, IL-5Ra' B cells in IL-5-Tg mice use 452 families more frequently than age-matched control B cells, suggests that IL-5 maintains Ly-l+ B cells that have a fetal-type Ig

170

KIYOSHI TAKATSU ET AL.

gene usage and could be responsible for prolonging the life span of immature Ly-1' B cells, which subsequently mature to Ly-1- B cells that secrete polyreactive natural antibodies. Another striking feature of the aberrant expression of IL-5 is a marked increase in the number of peripheral white blood cells, spleen cells, and peritoneal cavity cellularity. Particularly, the increase in the numbers of eosinophils in PBL is enormous. Infiltration of eosinophils into various tissues including bone marrow, spleen, liver, lung, and muscle are obvious. Eosinophils are also present in lymph nodes, particularly those along the trachea. Passive administration of antimIL-5 or anti-mIL-5Ra chain mAb's causes a decrease in the levels of peripheral blood eosinophils in the IL-5-Tg mice (Hitoshi et al., 1991). Mature eosinophils in the peritoneal cavity of the IL-5 transgenic mice or of Strongyloides ratti-infected mice also express IL-5Ra on their surface (Hitoshi et aZ., 1992; Korenaga et aZ., 1991). Passive administration of anti-IL-5Ra mAb intraperitoneally into the IL-5 transgenic mice induces a decrease in the number of recognizable eosinophils in peripheral blood within 5 days to normal levels (Hitoshi et al., 1991),clearly demonstrating that anti-IL-SR mAb can antagonize to IL-5 in vivo. These effects and their kinetics are similar to those induced by the treatment with anti-IL-5 mAb, which was also reported by Coffman et al. (1989a). Taking all of the results together, aberrant expression of the IL-5 gene induces accumulation of Ly-1' B-lineage cells and eosinophils, and persistent production of these cell types is not pathogenic. These animal models will give us feasible tools for analyzing certain diseases such as myositis, fascitis, or asthma with eosinophils.

2. IL-SRa Transgenic Mice To investigate the role of IL-5 in B-cell development, the transgenic mice carrying the MIL-5Ra gene ligated with human IgH promoter and mouse VH enhancer were generated (Koike et al., 1993). In these transgenic mice, mIL-5Ra was expressed on peripheral B cells and/ or some T cells. The transgenic mice (5Ra-Tg) were healthy and showed no changes in the cellularity of peripheral blood cells including B cells and eosinophils and in the expression of lineage-specific marker on the cell surface. Splenic B cells from 5Ra-Tg mice showed increased responsiveness to mIL-5 in Ig production and proliferation, but no drastic change in other immunological characteristics was observed. As described above, the functional mIL-5R is composed of a- and @-chains,so the responsiveness of B cells in 5Ra-Tg mice to IL-5 may be restricted by the number of endogeneous &chain expressed on B cells.

IL-5 AND ITS RECEPTOR SYSTEM

171

3. Defect of B-Cell-Speci$c IL-5-Mediated Signaling in XZD Mice

The CBA/N mice, which bear X chromosome-linked immunodeficiency (XID), are known to have defects manifested by a decrease in the overall number of B cells, low levels of circulating IgM and IgG3, and failure in responding to type I1 thymus-independent antigen (Scher, 1982).These defects could be accounted for by the absence of subpopulations of mature or late-developing B cells. Particularly, Ly-1 B cells are reported to be absent in the peritoneal cavity of XID mice (Hayakawa et al., 1986). Low responsiveness of XID B cells to IL-4, IL-10, and anti-IgM stimulation has been reported (Go et al., 1990; Howard et al., 1986). XID mice also showed delayed progression of murine acquired immunodeficiency syndromes (MAIDS) induced by murine leukemic virus (LP-BM5 MuLV) infection (Hitoshi et al., 1993a). As we demonstrated (Hitoshi et al., 1990), most peritoneal Ly-1 B cells coexpress IL-5R. We have therefore been interested in the ILSR expression on and the IL-5 responsiveness of B cells in XID mice. The IL-5R expression analyzed by using anti-mIL-5Ra and antimIL-5RP mAb's revealed that XID mice have fewer IL-5R' B cells than normals. In particular, a decrease in the number of peritoneal IL-5R' B cells among Ly-1 B cells is remarkable. Furthermore, the frequency of precursors of IL-5-responsive B cells in XID mice is approximately 100-fold lower than that of normal mice. Interestingly, sorted IL-5R' peritoneal B cells from XID mice displayed a low response to IL-5. Furthermore, intraperitoneal injection of IL-5 into normal mice induces polyclonal IgM production. The same regimen fails to induce an increase in the same parameters in XID mice. However, XID mice show mIL-5-induced eosinophilia in peripheral blood to a similar extent as normal mice. Eosinophils from mIL-5-injected XID mice express both a- and p-chains of IL-5R and respond to IL-5 with prolonged in vitro survival. These results indicate that a global defect observed in XID mice, in terms of growth and differentiation in response to IL-5, appears to be restricted in B cells. The low responsiveness of XID IL-5R' B cells to IL-5 may be explained by two possibilities. First, XID mice may have extremely low binding affinity to mIL-5. Our experiments may not support this possibility, because B cells in XID mice bear both IL-5Ra and IL-SRp on their surface and do not respond well to even high concentrations (-10,000 U/ml) of mIL-5. Second, XID B cells may have defects in IL-5-mediated signaling. It has been reported that anti-Ig-induced inositol phospholipid hydrolysis and Ca2+mobilization of XID B cells are only 40-50% of those of normal B cells (Rabinowitz et al., 1990). And Chen and Li (1990) reported that some Ig-associated glycopro-

172

KIYOSHI TAKATSU ET AL.

teins, which might be important for signal transduction of antigen receptor, differ between normal and XID mice. It should be stressed that eosinophils of XID mice do not have abnormalities in expression of IL-5R and their IL-5 responsiveness. There may be a defect in IL-5-mediated signaling, which is specific for B cells, not eosinophils. To get further insight into the B-cell defect in XID mice, we obtained XID-5Ra-Tg mice, in which all B cells constitutively express recombinant mIL-5Ra chain, and analyzed their B cells for responsiveness to IL-5. B cells from XID-5Ra-Tg mice did not respond to either TNPFicoll or IL-5 (Koike et al., 1993), indicating that B cells in XID mice have a defect at least in IL-5R-mediated signaling. B-cell-specific tyrosine kinase (Btk, previously called B-cell progenitor kinase, Bpk, or agammaglobulinemia tyrosine kinase, Atk) was isolated (Tsukada et al., 1993; Vetrrie et al., 1993).A single conserved residue substitution (Arg Cys) at position 28 within the amino-terminal unique region of Btk was shown to be mutated in XID mice (Rawlings et al., 1993; Thomas et al., 1993). This suggests that the point mutation of Btk in XID mice may be coupled with an impaired IL-5 responsiveness. Actually we found significantly enhanced kinase activities of Btk in IL-5-induced proliferation (Satoh et al., submitted). We believe that it is important therefore to clarify differences of Btk-associated molecule(s) in B cells between normal and XID mice and to envisage differences of IL-5-mediated signaling between B cells and eosinophils.

4. Parasite Znfection The development of eosinophilia in mice infected with T. canis is associated with the appearance of mIL-5 mRNA in the spleen (Y. Yamaguchi et al., 1990b). The development of eosinophilia in mice infected with nematodes other than T. canis was also reported by other investigators (Coffman et al., 1989a; Herndon and Kayes, 1992; Korenaga et al., 1991; Rennick et al., 1990; Sher et al., 1990). In all cases, the administration of anti-IL-5 mAb to parasite-infected mice totally blocked the development of eosinophilia. The role of the eosinophils induced by IL-5 in parasite rejection is, however, controversial. Depletion of eosinophils by anti-IL-5 mAb does not alter parasite burden or granuloma formation in mice infected with TrichineZZa spiralis and Shistosoma mansoni, respectively (Herndon and Kayes, 1992; Sher et al., 1990). Korenaga et al. (1991) reported that by using Strongyloides venezuelensis-infected mice passive administration of anti-IL-5 or anti-IL-5R mAb impaired the host’s protective immunity against larvae in a secondary infection, suggesting the IL-5 depen-

IL-5 AND ITS RECEPTOR SYSTEM

173

dency ofthe host's protective response to stage-specific immune mechanisms in the parasite infection.

5. IL-5and Experimental Eosinophilia The factors responsible for in vivo eosinophil accumulation at inflammatory sites were poorly defined, although T cells appear to be involved in the etiology of eosinophilia. IL-5 was shown to play an important role in local accumulation of eosinophils in mouse allergic peritonitis, which was induced by the injection of ragweed pollen extracts into the peritoneal cavity (Kaneko et al., 1991).Marked eosinophilia in the bronchoalveolar lavage fluid was produced by transnasal administration of an extract of the parasite Ascaris suum (Okudaira et al., 1991). Passive administration of anti-IL-5 mAb clearly inhibited the infiltration of eosinophils in the lung, suggesting that T-cellderived IL-5 is essential. It has been shown that the eosinophil recruitment into the site of cutaneous late-phase reaction (LPR)is dependent on IgE antibody and mast cells. A role of T cells in causing antigeninduced eosinophil recruitment in the LPR in mouse skin and trachea was indicated. Further it was found that CD4' T cells, but not CD8' T cells, caused the second peak of antigen-induced eosinophil recruitment of cutaneous LPR and that IL-5 mediates the eosinophil recruitment (Iwamoto et al., 1992; Nakajima et al., 1992). The effect of pretreatment with anti-IL-5 mAb on antigen-induced allergic bronchial eosinophilia and bronchial reactivity to acetylcholin in mice was also studied (Nagai et al., 1993). Consecutive inhalations of an antigen by actively sensitized animals resulted in an increase in airway reactivity to acetylcholine. Interestingly, anti-IL-5 mAb inhibited antigeninduced increase of eosinophils with little effect on bronchial hyperreactivity.

6. Airway Hyperreactivity and Eosinophils in Guinea Pigs The ovalbumin-sensitized guinea pig is a well-known model for the study of allergic asthma. In guinea pigs ovalbumin challenge induces early- and late-phase asthmatic reactions, airway eosinophilia, and bronchial hyperreactivity: It was demonstrated that anti-IL-5 antibody inhibits airway eosinophilia after a single ovalbumin challenge in sensitized guinea pigs (Chand et al., 1992; Gulbenkian et al., 1992). These observations are also confirmed by chronically challenged ovalbumin-sensitized guinea pigs (van Oosterhout et al., 1993). Bronchoalveolar lavage (BAL) eosinophilia and tracheal hyperreactivity observed in chronic ovalbumin challenge of sensitized guinea pigs are markedly inhibited by the anti-IL-5 mAb treatment. In contrast,

174

KIYOSHI TAKATSU ET AL.

the number of neutrophils is not affected by anti-IL-5 treatment. In guinea pigs treated with anti-IL-5 mAb, the development of hyperreactivity to histamine and aerocholin after ovalbumin challenge is completely inhibited (van Oosterhout et al., 1993; Akutsu et al., unpublished observation). These results suggest that IL-5 appears to be involved in airway eosinophilia and in the development of hyperreactivity in the guinea pig model but involvement of other cytokines is not excluded. It is not clear at this moment whether IL-5 directly increases airway hyperreactivity. Development of inhibitors for IL-5 synthesis and/or I L J R antagonists may provide a new type of antiasthmatic drugs.

7 . Tumor Rejection An intriguing experimental system designed to assess the role of cytokines in tumor rejection uses tumor cells transfected with an expression vector containing the cDNA sequence of a cytokine. The tumor cells produce the cytokine constitutively. When tumor cells secreting IL-4 or IL-2 were introduced into mice, they were rejected faster than the parent tumor cells. (Tepper et al., 1989,1992). In both cases tumors were infiltrated with eosinophils. Passive administration of anti-granulocyte and anti-IL-5 mAb completely and partially, respectively, suppressed the above-mentioned anti-tumor activity (Tepper et al., 1992), suggesting that IL-5 plays an important role in tumor rejection mediated by IL-4. Anti-tumor effect of IL-5 in uiuo has been also shown in two different tumor-implanting systems in mouse (Wu et al., 1992; Nakajima et al., 1992). B. HUMAN IL-5 IN DISEASES 1 . IL-5 and Asthma Asthma is associated with an inflammatory reaction involving the accumulation and activation of granulocytes, particularly eosinophils. The central role of eosinophils as possibly the most important inflammatory cells in asthma is becoming more wideG accepted (Kay 1991; Kay et al., 1991). Antigen challenge in asthmatic patients results in the accumulation of eosinophils in the lung (DeMonchy et al., 1985), and the number of eosinophils correlates with the severity of the late asthmatic reaction (Bousquet et al., 1990). Immunocytochemical studies of bronchial biopsies taken from patients with asthma (Bentley et al., 1991; Azzawi et al., 1990) have also shown that activated T cells can be detected in the bronchial mucosa, and their numbers can be correlated both with the numbers of local activated eosinophils and

IL-5 AND ITS RECEPTOR SYSTEM

175

with disease severity (Bentley et al., 1992). Activated CD25+, HLADR+ T cells are also detected in the peripheral blood of subjects with acute severe asthma (Corrigan et al., 1988,1990), and their numbers are reduced after therapy to a degree that correlates with the extent of clinical improvement. Of the lymphokines secreted by activated T cells, IL-5 is particularly implicated in the pathogenesis of asthma because it specifically enhances T-cell adhesion to vascular endothelial cells (Walsh et al., 1990), an important initial process in tissue emigration. It also selectively promotes the chemotaxis, proliferation, differentiation, chemiluminescence, cytotoxic cationic protein release, and survival of eosinophils in vitro. Messenger RNA encoding IL-5 has been detected both in bronchial mucosal biopsies (Hamid et al., 1991)and bronchoalveolar lavage T cells of mild atopic asthmatic subjects (Robinson et al., 1992) by in situ hybridization. IL-5, IL-3, and GM-CSF gave a weak locomotory response for eosinophilia from normal nonatopic subjects, but not for eosinophils from subjects with eosinophilia with asthma and/or allergic rhinitis. In contrast, IL-5 and IL-3, but not GM-CSF, had no effect on neutrophils. Eosinophils from normal subjects precultured with IL-5 showed enhanced locomotory response to platelet-activating factor (PAF), leukotrien B4 (LTB4), and formyl-methionyl-leucyl phenylalanine (FMLP) (Sehmi et al., 1992).Selective priming of eosinophil but not neutrophil locomotion by IL-5 suggests that IL-5 may play a significant role in the preferential accumulation of eosinophils at sites of allergic inflammation. Corrigan et al. measured the concentrations of IL-5 in the peripheral blood of patients with moderate to severe exacerbations of asthma and studied the effects of glucocorticoid therapy on these concentrations (1993). Their results revealed that IL-5 was detectable in the serum of 8 of 15 of the asthmatic patients before glucocorticoid therapy but in none of these patients 7 days after the therapy. The therapy is also associated with a significant fall in the peripheral blood eosinophil counts. They proposed the hypotheses that exacerbations of asthma are associated with activation of T cells secreting IL-5, which regulates eosinophilia and that glucocorticoid therapy results in reduction of the activation status of these cells concomitant with inhibition of IL-5 secretion. Further studies are required to confirm the T-cell origin of IL-5 in patients with asthma and to examine the effects of glucocorticoids and other drugs that inhibit T cells for their influences on IL-5 synthesis and release. It has been shown that incubation of eosinophils with serum from patients with asthmatic or the idiopathic hypereosinophilic syndrome

176

KIYOSHI TAKATSU ET AL.

prolonged their survival in uitro, and that this effect is partly abrogated in the presence of anti-IL-5 antibody (Owen et al., 1989; Walker et al., 1991); similarly, anti-IL-5 antibody partially abrogates eosinophil colony-stimulating activity in the serum of patients with peripheral blood eosinophilia (Enokihara et al., 1990).The accumulation of eosinophils was found to be selective after specific allergen inhalation, because the numbers of macrophages, neutrophils, and lymphocytes were not significantly increased. Patients with a late response showed a significant increase in eosinophils by 4 hr with higher numbers at 24 hr (Rossi et al., 1991). Endobronchial allergen challenge in asthmatic patients induced a significant airway eosinophilia and postallergen challenge eosinophils expressed IL-5 and GM-CSF mRNA, but not IL-lP or IL-2 mRNA (Broide et al., 1992). In situ hybridization on mucosal bronchial biopsies indicated local expression of IL-5, which correlated with the number of infiltrating eosinophils (Hamid et al., 1991).The expression of IL-5 by eosinophils at sites of allergic inflammation in asthmatics may provide an important autocrine pathway, maintaining the viability and effector function of the recruited eosinophils. Evidence that eosinophils are producing tissue damage in the asthmatic lung is based on the demonstration of eosinophils major basic protein and eosinophil-derived neurotoxin, indicating degranulation at sites of injury. Furthermore, the concentrations of these toxic products in asthmatic sputum are similar to the concentrations that are cytotoxic in uitro (Gleich and Adolphson, 1986). IL-5 can prime human eosinophils to release large amounts of cytotoxic cationic pro1989; Fujisawa et d . ,1990; Takateins and leukotrienes (Lopez et d., fuji et al., 1991). 2. IL-5 and Helminth Infections Human helminth infections are commonly accompanied by elevation of serum IgE and peripheral blood eosinophilia. It is well known that in helminth-infected individuals, an expansion of a TH2-likepopulation results in increased production of IL-4 and IL-5 (Mahanty et al., 1992). The regulation of IgE production in humans is mediated primarily by IL-4. However, eosinophilia appears to be regulated by IL-5, IL-3, and GM-CSF. A comparison of the production of cytokines by normal individuals and by eosinophilic patients infected with the filarial parasite Lou loa reveals that both groups produced similar levels of IL-3 and GM-CSF, but the normals gave relatively little IL-5, while cells from infected patients produced high levels of IL-5 (Limaya et al., 1990, 1991). Because of their prominence in immunopathological lesions, it has

IL-5 AND ITS RECEPTOR SYSTEM

177

been hypothesized that eosinophils play a role both in pathogenesis of egg granulomas as well as in protective immunity itself. It has been more difficult to obtain convincing in uiuo evidence supporting effector function for eosinophils in immunity to parasites. The association between eosinophilia and resistance to shistosomiasis in two studies in Africa remain the best evidence that eosinophils are involved in immunity to parasites (Sturrock et al., 1983; Hagan et al., 1985). It is possible that resistance to parasitic infections has provided the selective pressure for the evolution of 1L-5. 3. IL-5 and Graft Rejection

The local mechanisms of allograft rejection are thought to be regulated by cytokines produced by graft-infiltrating cells. Martinez et al. (1993) have reported direct evidence for IL-lP, IL-2, IL-4, IL-5, IL-6, and TNF-a mRNA in human allografts by PCR. Eosinophils in liver grafts were reported to be a sensitive and specific indicator for acute rejection (Foser et al., 1989). The role of IL-5 and eosinophils in allograft rejection was studied in human liver allograft recipients (Martinez et al., 1993).Liver allograft biopsies were analyzed for intragraft IL-5 gene expression, and the percentages of eosinophils and plasma cells within the portal infiltrate as well as peripheral eosinophil levels were determined. The majority of allografts with evidence of rejection had concomitant IL-5 mRNA and eosinophilia, while no resolving or nonrejecting allografts had simultaneous IL-5 mRNA and eosinophilia. In fact, rejecting liver allografts that contain IL-5 mRNA and eosinophils also contain infiltrating cells that produce the cytotoxic mediator major basic proteins. In contrast, intragrafi plasma cell and peripheral eosinophil levels did not correlate with the his topathologic status of the allograft. Cyclosporin and FK506 had similar effects on the frequency of IL-5 gene expression in rejecting and nonrejecting allografts. These observations raise the possibility of a cellular pathway of liver allograft rejection involving by IL-5-mediated eosinophils. 4. IL-5 and Tumors The presence of eosinophils in association with tumors has been noted for many years (Lowe et al., 1981). Histologic studies showed that a positive correlation between eosinophil infiltration and survival rates could be observed in carcinoma of stomach, lung, colon, and uterine cervix (Iwasaki et al., 1986; Lowe et al., 1981; Pasternark and Hansa, 1984; Pretlow et al., 1983). Patients receiving IL-2 as an antitumor treatment frequently develop eosinophilia. This appears to be because IL-2 induces the production of IL-5 by T cells (MacDonald

178

KIYOSHI TAKATSU ET AL.

et al., 1990; Y. Yamaguchi et al., 1990b). In view of the observations that infiltration of eosinophils in tumor-growing sites correlates with a positive prognosis, it may be useful to analyze the number and distribution of eosinophils for a correlation with the antitumor activity of IL-2. Rivoltini et al. (1993) evaluated the existence of in vitro eosinophil-mediated tumor cell killing and its in uiuo induction in neoplastic patients treated with IL-2 and found that a marked increase of cytotoxicity of eosinophils was observed in IL-2-treated patients who showed the increase in the serum IL-5 level. These results support the hypotheses that eosinophils whose production is increased by IL-5 play a positive role in the antitumor effect in humans. V. Future Perspectives

IL-5 functions as a B-cell stimulatory factor and an eosinophilopoietic factor. In addition, mIL-5 induces not only early B-cell development of Ly-l+ lineage, but also differentiation of mature B cells. If we consider that IL-5, IL-3, and GM-CSF have some different functions, we have to ask how the signal of IL-5 is transduced differently from that of GM-CSF or IL-3. The signals generated by these cytokines may be equivalent, and different functions of these cytokines are due to the stage of development of cells expressing the receptors. Alternatively, different signal-transducing molecules, which generate a specific signal for each cytokine, may be associated with respective ligand-binding a-subunits of each receptor. These possibilities are not mutually exclusive. It is not clear what determines whether a responding cell proliferates, differentiates, or simply increases functional activity in response to IL-5. Unlike the receptor tyrosine kinase, the cytoplasmic portions of cytokine receptors do not contain kinase domains or any other sequences with recognizable catalytic function. Some of the cytokines can be grouped into subfamilies that share receptor components. IL-3, IL-5, and GM-CSF make up a subfamily of cytokines displaying a variety of overlapping actions during hematopoiesis. These three hematopoietic factors bind to distinct specificity-determining a-receptor components, but share a common signal-transducing P-receptor component; the unique distributions of the individual asubunits determine which hematopoietic cells are capable of responding to each of the factors. Each a-subunit of IL-5R7IL3R, and GM-CSFR has a conserved region (RLFPPI/VPxxxK/RxxL/IxD) close to the transmembrane domain. This region is important for interaction with the cytoplasmic domain of the common p-subunit or with a certain

IL-5 AND ITS RECEPTOR SYSTEM

179

common signal-transducing molecule. The carboxy-terminal regions of respective a-subunit or relatively diversified from each other. The specific function of the respective cytokine may depend on the structure of the distal part of the cytoplasmic portion of the a-subunits. Or, each a-subunit may simply support the function of the common psubunit by inducing conformational change or dimerization of the psubunit. These issues should be resolved in the future. There is considerable evidence that homodimerization of the cytoplasmic domain of the common @chain, signal transducer may be required to transmit cytokine-mediated signals (Murakami et al., 1993). It is supposed that a similar dimerization mechanism may be involved in IL-5-mediated signal transduction. However, this may only be the case in the signal transduction for cell growth. In other cases, such as signals for B cells to induce differentiation to immunoglobulinsecreting cells, the cytoplasmic domain of the IL-5Ra chain may have some important roles in transmitting IL-5-specific signal. It is possible that there are several signal-transducing machineries which connect with the membrane receptors to the intracellular device and that some machineries function only in a limited cell type. We are studying this possibility. ACKNOWLEDGMENTS The authors are grateful to Drs. A. Tominaga, Y. Sano, N. Harada, N. Yamaguchi, M. Matsumoto, N. Koyama, Y. Kikuchi, R. Matsumoto, S. Mita, M. Migita, E. Sonoda, S. Katoh, Y. Murata, F. Imamura, J. Miyazaki, T. Kitamura, A. Miyajima, S. Yonehara, T. Katagiri, M. Koike, S. Tsukada, T. Honjo, T. Kinashi, Y. Yamaguchi, T. Suda, S. Nishikawa, A. Rolink, and K. Yamamura for their tremendous contribution during the course of these studies. We also thank Drs. H. Hayashi, K. Onoue, K. Ishizaka, T. Hamaoka, T. Kishimoto, K. A. Smith, and F. Melchers for their helpful suggestions and encouragement during the course of these studies. This work was supported in part by a Grant-in-Aid for Scientific Research for Special Project Research, Cancer Bioscience from the Ministry of Education, Science, and Culture; by Special Coordination Funds for promoting Science and Technology from the Japanese Ministry of Science and Technology; by a grant from Takeda Medical Science Foundation; by a grant from Mochida Memorial Foundation of Medical and Pharmaceutical Science; and by a Uehara Foundation of Medical Research.

REFERENCES Abrams, J. S., Roncarolo, M. G., Yssel, H., Anderson, U.,Gleich, G. J., and Silver, J. E. (1992). Immunol. Reu. 127,5-24. Alderson, M. R., Pike, B. L., Harada, N., Tominaga, A., Takatsu, K., and Nossal, G . J. V. (1987).J . Immunol. 139,2656-2660. Allison, K. C., Strober, W., and Harriman, G. R. (1991).J.Immunol. 146,4197-4203. Aoki, T., Kikuchi, H., Miyatake, S-I., Oda, Y., Iwasaki, K., Yamasaki, T., Kinashi, T., and Honjo, T. (1989). 170,583-588.

180

KIYOSHI TAKATSU ET AL.

Arai, K. I., Lee, F., Miyajima, A., Miyatake, S., Arai, N., and Yokota, T. (1990).Annu. Rev. Biochem. 59,783-836. Argetsinger, L. S., Campbell, G. S., Yang, X., Witthuhn, B. A., Silvennoinen, O., Ihle, J. N., and Carter&, C. (1993).Cell (Cambridge, Mass.) 74,237-244. Azzawi, M., Bradley, B., Jeffery, P. K., Frew, A. J., Wardlaw, A. J., Knowles, G., Assoufi, B., Collins, J. V., Durham, S. R., and Kay, A. B. (1990).Am. Reu. Respir. Dis. 142, 1410-1413. Azuma, C., Tanabe, T., Konishi, M., Kinashi, T., Noma, T., Matsuda, R., Yaoita, Y., Takatsu, K., Hammarstrom, L., Severinson, E., Smith, L. I. E., and Honjo, T. (1986). Nucleic Acids Res. 14,9149-9158. Barlow, D. P., Bucan, M.,Lehrath, H., Hogan, M.,and Gough, N. M. (1987).EMBOJ. 6,617-623. Bauman, A. A., and Paul, C. C. (1992). Blood 79,1763-1767. Bazan, J. F. (1989).Biochem. Biophys. Res. Commun. 164,788-795. Bazan, J. F. (1990).Proc. Natl. Acad. Sci. U.S.A. 87,6934-6938. Beagley, K. W., Eldridge, J. H., Kiyono, H., Everson, M. P., Koopman, W. J., Honjo, T., and McGhee, J. R. (1988).]. Immunol. 141,2035-2042. Bentley, A. M., Menz, G., Storz, C., Robinson, D. S., Bradley, B., Jeffery, P. K., Durham, S. P., and Kay, A. B. (1992).Am. Rev. Respir. Dis. 146,500-506. Bertolini, J. N., Sanderson, C. J., and Benson, E. M. (1993).Eur.J.Immunol. 23,398-402. Betz, M., and Fox, B. S. (1991).]. Immunol. 146, 108-113. Bischoff, S. C., Brunner, T., De Weck, A., and Dahinden, C. A. (1990).J. E x p . Med. 172, 1577-1582. Bohjanen, P. R., Okajima, N., and Hodes, R. J. (1990).Proc. Natl. Acad. Sci. U.S.A.87, 5283-5287. Bond, M. W., Shrader, B., Mosmann, T. R., and Coffman, R. L. (1987).J.Immunol. 139, 369 1-3696. Bousquet, J., Chanez, P., Lacoste, J. Y., Barneon, G., Ghavanian, N., Enander, J., Venge, P., Ahlstedt, S., Simony Lafontaine, J., Godrad, P., and Francois-Bernard, M. (1990). N. Engl.]. Med. 323,1033-1036. Boyd, A. W., and Schrader, J. W. (1982).Nature (London)297,691-693. Broide, D. H., Paine, M. M., and Firestein, G. S. (1992).]. Clin. Invest. 90,1414-1424. Brown, J. H., Jardetzky, T. S., Gorga, J. C., Stern, L. J., Urban, R. G., Strominger, J. L., and Wiley, D. C. (1993).Nature (London) 364,33-39. Butterfield, J. H., Leiferman, K. M., Abrams, J., Silver, J. E., Bower, J., Gonchoroff, N., and Gleich, G. J. (1992). Blood 79,688-692. Campbell, H. D., Tucker, W. Q . J., Hort, Y., Martinson, M. E., Mayo, G., Clutterbuck, E. J., Sanderson, C. J., and Young, I. G. (1987). Proc. Natl. Acad. Sci., U.S.A. 84, 6626-6633. Campbell, H. D., Sanderson, C. J., Wang, Y., Hort, Y., Martinson, M. E., Tucker, W. Q. J., Stellwagen, A,, Strath, M., and Young, I. G. (1988).Eur. J . Biochem. 174, 345-352. Chand, N., Harrison, J. E., Rooney, S., Pillar, J., Jakubicki, R., Nolan, K., Diamantis, W., and Sofia, R. D. (1992).Eur. J . Phawacol. 211,121-123. Chen, S.-S., and Li, Q. (1990).Cell Immunol. 128,353-361. Cherwinski, H. M., Schumacher, J. H., Brown, K. D., and Mosmann, T. R. (1987). J. Erp. Med. 166,1229-1239. Chihara, J., Piumas, J., Gruart, V., Tavemier, J., Prin, L., Capron, A., and Capron, M. (1990).]. Exp. Med., 172, 1347-1351.

IL-5AND ITS RECEPTOR SYSTEM

181

Clutterbuck, E., Shields, J. G., Gordon, J., Smith, S. H., Boyd, A., Callard, R. E., Campbell, H. D., Young, I. G., and Sanderson, C. J. (1987). Eur.1. Immunol. 17,1743-1750. Coffman, R. L., Shrader, B., Carty, J., Mosmann, T. R., and Bond, M. W. (1987). J, Immunol. 139,3685-3690. Coffman, R. L., Seymour, B. W., Lebman, D. A., Hiraki, D. D., Christiansen, J. A., Shrader, B., Chenvinski, H. M., Savelkoul, H. F. J., Finkelman, F. D., Bond, M. W., and Mosman, T. R. (1988). Immunol. Rev. 102,5-28. Coffman, R. L., Seymour, B. W. P., Hudak, S., Jackson, J., and Rennick, D. (1989a). Science 245,308-310. Coffman, R. L., Lebman, D. A., and Shrader, B. (1989b).J . Exp. Med. 170,1039-1043. Corrigan, C. J., and Kay, A. B. (1990). Am. Reu. Respir. Dis. 141,970-977. Corrigan, C. J., Hartnell, A,, and Kay, A. B. (1988). Lancet 1, 1129-1131. Corrigan, C. J., Haczku, A., Gemou, E. V., Doi, S., Kikuchi, Y., Takatsu, K., Durham, S. R., and Kay, A. B. (1993). Am. Rev. Respir. Dis. 147,540-547. DeMonchy, J. G., Kauffman, H. F., Venge, P., Koeter, G. H., Jansen, H. M., Sliter, H. J., and De Vries, K. (1985). Am. Rev, Respir. Dis. 131,373-376. Denburg, J. A. (1992). Blood 79,846-860. Dent, L. A., Strath, M., Mellor, A. L., and Sanderson, C. J. (1990). J . Exp. Med. 172, 1425-1431. Desreumaux, P., Janin, A,, Colombel, J. E., Prin, L., Pulmas, J., Emilie, D., Torpier, G., Capron, A,, and Capron, M. (1992).J. E x p . Med. 175,293-296. Devos, R., Plaetinck, G., Van der Heyden, J., Cornelis, S., Vandekerckhove, J., Fiers, W., and Tavernier, J. (1991). EMBOJ. 10,2133-2137. Devos, R., Guisez, Y., Cornelis, S . , Verhee, A., Van der Heyden, J., Manneberg, M., Lahn, H. W., Fiers, W., Tavernier, J., and Plaetinck, G. (1993).J. Biol. Chem. 268, 6581-6587. Duronio, V., Welham, M. J., Abraham, S., Dryden, P., and Schrader, J. W. (1992). Proc. Natl. Acad. Sci. U S A . 89, 1587-1591. Dutton, R. W., Falkoff, R., Hirst, J. A,, Hoffman, M., Kappler, J. W., Kettman, J. R., Lesley, J. F., and Van, D. (1971). Prog. Immunol. 1, 355-368. Ehlich, A., Schaal, S., Gu, H., Kitamura, D., Muller, W., and Rajewsky, K. (1993). Cell (Cambridge, Mass.) 72,695-704. Enokihara, H., Furusawa, S., Nakakubo, H., Kajitani, H., Nagashima, S., Saito, K., Shishido, H., Hitoshi, Y., Takatsu, K., Noma, T., Shimizu, A., and Honjo, T. (1989). Blood 73,1809-1813. Enokihara, H., Kajitani, H., Nagashima, S., Tsunogake, S., Takano, N., Saito, K., Furusawa, S., Shishido, H., Hitoshi, Y., and Takatsu, K. (1990). Brit. J . Haematol. 75, 458-462. Foser, P. F., Sankary, H. N., Hart, M., Ashman, M., and Williams, J. W. (1989). Transplantation 45,537-540. Fujisawa, T., Abu-Ghazaleh, R., Kita, Sanderson, C. J., Gleich G. J., Young, I. G., and Vadas, M. S. (1990).]. Immunol. 144,642-646. Gleich, G. J., and Alderson, C. R. (1986). Ado. Immunol. 39, 177-253. Go, N. F., Castle, B. E., Barrett, R., Kastelein, R., Dang, W., Mosmann, T. R., Moore, K. W., and Howard, M. (199O).J.Exp. Med. 172, 1625-1631. Gorman, D. M., Itoh, N., Kitamura, T., Schreurs, J., Yonehara, S., Yahara, I., Arai, K-I., and Miyajima, A. (1990). Proc. Natl. Acad. Sci., U.S.A. 87, 5459-5463. Gorman, D. M., Itoh, N., Jenkins, N. A., Gilbert, D. J., Copeland, N. G., and Miyajima, A. (1992).J. Biol. Chem. 267, 15,842-15,848.

182

KIYOSHI TAKATSU ET AL.

Cough, N. M., and Raker, S. (1992). Genomics 12,855-856. Grabtree, G . R. (1992). Science 243, 355-360. Gulbenkian, A. R., Egan, R. W., Femandez, X., Jones, H., Kreutner, W., Kung, T., Payvandi, F., Sullivan, L., Zurcher, J. A., and Watnick, A. S. (1992).Am. Reu. Respir. Dis. 146, 263-265. Hagan, P., Wilkins, H. A., Blumental, U. J., Hayes, R. J., and Greenwood, B. M. (1985). Parasite lmmunol. 7,625-632. Hamid, Q., Azzawi, M., Ying, S., Moqbel, R., Wardlaw, A. J., Corrigan, C. J., Bradley, B., Durham, S. R., Collins, J. V., Jeffery, P. K., Quint, D. J., and Kay, A. B. (1991). J . Clin. Invest. 87, 1541-1550. Hanazono, Y., Chiba, S., Sasaki, K., Mano, H., Miyajima, A., Arai, K., and Hirai, H. (1993). EMBO J. PA,1641-1646. Hara, T., and Miyajima, A. (1992). E M B O J . 11, 1875-1884. Harada, N., Kikuchi, Y., Tominaga, A., Takaki, S., and Takatsu, K. (1985).J. lmmunol. 134,3944-3951. Harada, N., Takahashi, T., Matsumoto, M., Kinashi, T., Kikuchi, Y.,Koyama, N., Severinson, E., Yaoita, Y., Honjo, T., Yamaguchi, N., Tominaga, A., and Takatsu, K. (1987a). Proc. Natl. Acad. Sci. U.S.A. 84,4581-4585. Harada, N., Matsumoto, M., Koyama, N., Honjo, T., Shimizu, A., Tominaga, A., and Takatsu, K. (1987b). Immunol. Lett. 15,205-215. Hardy, R. R., and Hayakawa, K. (1986). lmmunol. Reo. 93,53-79. Harriman, G. R., Kunimoto, D. Y., Elliott, J. F., Paetkau, V., and Strober, W. (1988). J . lmmunol. 140,3033-3039. Hayakawa, K., and Hardy. (1988). Annu. Rev. lmmunol. 6, 197-218. Hayakawa, K., Hardy, R. R., and Herzenberg, L. E. (1986).Eur.J.lmmunol. 16,450-456. Hayashida, K., Kitamura, T., Gorman, D. M., Arai, K-I., Yokota, T., and Miyajima, A. (1990). Proc. Natl. Acad. Sci. U S A . 87,9655-9659. Herndon, F. J., and Kayes, S. G. (1992).J. lmmunol. 149,3642-3647. Hibi, M., Murakami, M., Saito, M., Hirano, T., Taga, T., and Kishimoto, T. (1990). Cell (Cambridge, Mass.) 63, 1149-1157. Hirai, K., Yamaguchi, M., Misaki, Y., Takahashi, T., Ohta, T., Morita, Y., Ito, K., and Miyamoto, T. (1990).J . E x p . Med. 172, 1525-1528. Hitoshi, Y., Mita, S., Tominaga, A., Kikuchi, Y., Sonoda, E., Watanabe, Y., and Takatsu, K. (1989). Znt. Immunol. 1, 185-190. Hitoshi, Y., Yamaguchi, N., Mita, S., Takaki, S., Tominaga, A,, and Takatsu, K. (1990). J . lmmunol. 144,4218-4225. Hitoshi, Y., Yamaguchi, N., Korenaga, M., Mita, M., Tominaga, A., andTakatsu, K. (1991) lnt. lmmunol. 3,135-139. Hitoshi, Y., Yamaguchi, N., Tominaga, A., and Takatsu, K. (1992).Ann. N.Y. Acad. Sci. U.S.A.651,261-263. Hitoshi, Y., Okada, Y., Sonoda, W., Tominaga, A., Makino, M., Suzuki, K., Kinoshita, J., Komura, K., Mizuochi, T., and Takatsu, K. (1993a).J . E x p . Med. 177, 621-626. Hitoshi, Y., Sonoda, E., Kikuchi, Y., Yonehara, S., Nakauchi, H., Tominaga, A., and Takatsu, K. (1993b). Znt. lmmunol. 5, 1183-1190. Honjo, T., and Takatsu, K. (1990). In “Peptide Growth Factors and Their Receptors (Handbook of Experimental Pharmacology, Vol 95/I, Peptide Growth Factors and Their Receptors I)” (M. A. Sporn and A. B. Roberts, eds.), pp. 609-632. SpringerVerlag, New York/Berlin. Howard, M., and Paul, W. E. (1983). Annu. Reu. immunol. 1,307-333. Howard, M., Pesavento, P., and Stein, P. (1986).J . lmmunol. 136,4531-4537.

IL-5 AND ITS RECEPTOR SYSTEM

183

Hsieh, C-S., Macatonia, S. E., Tripp, C. S., Wolf, S, F., O’Garra, A., and Murphy, K. M. (1993). Science 260,547-549. Huebner, K., Isobe, M., Croce, C. M., Golde, D. W., Kaufman, S. E., and Gasson, J. C. (1985). Science 230, 1282-1285. Huston, D. P., Moore, J. P., and Huston, M. M. (1993).J . Immunol. 150, 182A. Imamura, F., Takaki, S., Akagi, K., Ando, M., Yamamura, K.-I., Takatsu, K., and Tominaga, A. (1994). DNA Cell Biol., in press. Ingley, E., and Young, I. G. (1991). Blood 78,339-344. Inoue, C., Ichikawa, A., Hotta, T., and Saito, H. (1990). Br. J . Haematol. 76, 554555. 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). Cell (Cambridge, Mass.) 69,1121-1132. Isobe, M., Kumura, Y., Murata, Y., Takaki, S., Tominaga, A., Takatsu, K., and Ogita, Z. (1992). Genomics 14, 755-758. Itoh, N., Yonehara, S., Schreurs, J., Gorman, D. M., Maruyama, K., Ishii, A., Yahara, I., Arai, K., and Miyajima, A. (1990). Science 247,324-327. Iwamoto, I., Tomoe, S., Tomioka, H., Takatsu, K., and Yoshida, S. (1992).J . Leukocyte Biol. 52,572-578. Iwasaki, K., Torisu, M., and Fujimura, T. (1986). Cancer 58, 1321-1326. Iwasato, T., Shimizu, A,, Honjo, T., and Yamagishi, H. (1990). Cell (Cambridge, Mass.) 62,135-142. Kaneko, M., Hitoshi, Y.,Takatsu, K., and Matsumoto, M. (1991). Znt. Arch. Allergy Appl. Immunol. 96,41-45. Kanakura, Y., Druker, B., Cannistra, S. A., Furukawa, Y., Tomimoto, Y., and Griffin, J. D. (1990). Blood 76, 709-715. Kantor, A. B. (1991). Immunol. Today, 12,389-391. Karasuyama, H., Rolink, A., and Melchers, F. (1988).J . Exp. Med. 167, 1377-1390. Katoh, S., Tominaga, A., Migita, M., Kudo, A., and Takatsu, K. (1990). Dew. Immunol. 1,113-125. Katoh, S., Bendig, M. M., Kanai, Y., Schultz, L. D., Hitoshi, Y., Takatsu, K., and Tominaga, A. (1993). DNA Cell. Biol. 12,481-491. Kay, A. B. (1991).J . Allergy Clin. Immunol. 87,893-910. Kay, A. B., Ying, S., Vaney, V., Gage, M., Durham, S. R., Moqbel, R., Wardlaw, A. J., and Hamid, Q. (1991).J . Exp. Med. 173, 775-778. Kikuchi, Y., Migita, M., Takaki, S., Tominaga, A., and Takatsu, K. (1994).J . Immunol. Methods, 167,289-298. Kim, P.-H., and Kagnoff, M. F. (1990).J. Immunol. 145,3774-3778. Kinashi, T., Harada, N., Severinson, E., Tanabe, T., Sideras, P., Konishi, M., Azuma, C., Tominaga, A., Bergstedt-Lindqvist, S., Takahashi, M., Matsuda, F., Yaoita, Y., Takatsu, K., and Honjo, T. (1986). Nature (London) 324, 70-73. Kincade, P. W. (1981).Ado. Immunol. 31, 177-245. Kishimoto, T., and Hirano, T. (1988). Annu. Reo. Immunol. 6,485-512. Kishimoto, T., Akira, S., and Taga, T. (1992). Science 258,593-597. Kita, H., Weiler, D. A., Abu, G. R., Sanderson, C. J., and Gleich, G. J. (1992).J.Immunol. 149,629-635. Kitamura, D., Kaneko, H., Miyagoe, Y., Ariyasu, T., and Watanabe, T. (1989). Nucleic Acids Res. 17,9367-9379. Kitamura, T., Hayashida, K., Sakamaki, K., Yokota, T., Arai, K., and Miyajima, A. (1990). Proc. Natl. Acad. Sci. U.S.A. 88, 5082-5086.

184

KIYOSHI TAKATSU ET AL.

Kitamura, T., Sato, N., Arai, K., and Miyajima, A. (1991). Cell (Cambridge, Mass.) 66, 1165-1174. Klinken, S. P., Alexander, W. S., and Adams, J. M. (1988). Cell (Cambridge,Mass.) 53, 857-867. Kodama, S . , Tsuruoka, N., and Tsujimoto, M. (1991). Biochem. Biophys. Res. Commun. 178,514-519. Koike, M., Kikuchi, Tominaga, A,, Takaki, S., Akagi, K., Miyazaki, J., Yamamura, K., and Takatsu, K. (1993).J . Zmmunol. 150,206A. Korenaga, M., HItoshi, Y., Yamaguchi, N., Sato, Y., Takatsu, K., and Tada, I. (1991). Immunology 72,502-507. Koyama, N., Harada, N., Takahashi, T., Mita, S., Okamura, H.,Tominaga, A., and Takatsu, K. (1988). Immunology 63,277-283. Le Beau, M. M., Westbrook, C. A., Diaz, M. O., Larson, R. A., Rowley, J. D., Gasson, J. C., Golde, D. W., and Sherr, C. J. (1986). Science 231, 984-987. Le Beau, M. M., Epstein, D. D., OBrien, S. J., Nienhuis, A. W., Yang, Y. C., Clark, S. C., and Rowley, J. D. (1987). Hum. Gene Mapping 9, A342. Lebman, D. A., and Coffman, R. L. (1990).J. Immunol. 141,2050-2056. Lebman, D. A., Nomura, D. Y.,C o h a n , R. L., and Lee, F. D. (1990).Proc. Natl. Acad. Sci. U.S.A. 87,3692-3696. Lee, H. J., Nakagawa, N. K., Naito, Y., Nishida, J., Arai, N., Arai, K. I., and Yokota, T. (1993). J . Immunol. 151. Li, Y. S., Gimond, C., and Revillard, J. P. (1992). Eur. Cytokine Network 3, 103-107. Limaye, A. P., Abrams, J. S., Silver, J. E., Ottesen, E. A., and Nutman, T. B. (1990). J . E x p . Med. 172,399-402. Limaye, A. P., Abrams, J. S., Silver, J. E., Awadzi, K., Fransis, H. F., Ottesen, E. A., and Nutman, T. B. (1991).J . Clin. Inuest. 88, 1418-1421. Lopez, A. F., Sanderson C. J., Gamble, J. R., Campbell, H. D., Young, I. G., and Vadas, M. A. (1988).J . E x p . Med. 167,219-224. Lopez, A. F., Eglington, J. M., Gillis, D., Park, L. S., Clark, C. S., and Vadas, M. A. (1989). Proc. Natl. Acad. Sci. U.S.A. 86,7022-7026. Lopez, A. F., Eglington, J, M., Lyons, A. B., Tapley, P. M., To, L. B., Park, L. S., Clark, C. S., and Vadas, M. A. (1990).J . Cell. Physiol. 145, 69-77. Lopez, A. F., Vadas, M. A,, Woodstock, J. M., Milton, S. E., Lewis, A., Elliott, M. J., Gillis, D., Lreland, R., Olwell, E., and Park, L. S. (1991).J . Biol. Chem. 266, 24,741-24,747. Lopez, A. F., Elliott, M. J., Woodstock, J. M., andVadas, M. A. (1992).Immunol. Today 13,495-500. Loughnan, M. S., Takatsu, K., Harada, N., and Nossal, G. J. V. (1987).Proc. Natl. Acad. Sci. U.S.A.84,5399-5403. Lowe, D., Jorizzo, J., and Hunt, M. S. R. (1981).J. Clin. Pathol. 34, 1343-1348. MacDonald, D., Gordon, A. A,, Kajitani, H., Enokihara, H., and Barrett, A. J. (1990). Br. J . Haematol. 76, 168-173. Mahanty, S., Abrams, J. S., King, C. L., Limaye, A. P., and Nutman, T. B. (1992). J . Immunol. 148,3567-3571. Maliszewski, C. R., and Fanslow, W. C. (1990). Trends Biotechnol. 8, 324-329. Martinez, 0. M., Krams, S. M., Villanueva, J. C., Ferrell, L., Lake, J., Robberts, J. P., and Ascher, N. L. (1993). Transplant. Proc. 25, 126-127. Matsumoto, M., Tominaga, A., Harada, N., and Takatsu, K. (1987). J. Immunol. 138, 1826-1833. Matsumoto, R., Matsumoto, M., Mita, S., Hitoshi, Y., Ando, M., Araki, S., Yamaguchi, N., Tominaga, A., and Takatsu, K. (1989). Immunology 66,32-38.

IL-5 AND ITS RECEPTOR SYSTEM

185

Matsumoto, R.,Ando, M., Kohrogi, H., Araki, S., and Takatsu, K. (1992). Chest 102, 1296-1298. Matsuoka, M.,Yoshida,K., Maeda,T., Usuda, S.,andSakano, H. (1990).Cell(Cambridge, MUSS.)62, 135-142. McIntyre, T. M., Klinman, D. R.,Rothman, P., Lugo, M., Dasch, J. R.,Mond, J. J., and Snapper, C. M. (1993).J.E r p . Med. 177, 1031-1037. McKenzie, D. T., Filutowib, H., Swain, S. L., and Dutton, R. W. (1987).J . Immunol. 139,2661-2668. McKenzie, A. N. J., Ely, B., and Sanderson, C. J. (1991a).Mol. Immunol. 28, 155-158. McKenzie, A. N. J., Barry, S. C., Strath, M., and Sanderson, C. J. (1991b).EMBOJ. 10, 1193-1 199. Melchers, F., and Anderson, J. (1986).Annu. Reu. Immunol. 4, 13-36. Migita, M., Yamaguchi, N., Kato, S., Mita, S., Matsumoto, R., Sonoda, E., Tsuchiya, H., Matsuda, I., Tominaga, A., and Takatsu, K. (1990).Microbiol. Immunol. 34,937-952. Migita, M., Yamaguchi, N., Mita, S., Higuchi, S., Hitoshi, Y., Yoshida, Y., Tomonaga, M., Matsuda, I., Tominaga, A,, and Takatsu, K. (1991).Cell. Zmmunol. 133,484-497. Milburn, M. V., Hassel1,A. M., Lambert, M. H., Jordan, S. R., Proudfoot,A. E. I., Graber, P., and Wells, T. N. C. (1993).Nature (London) 363, 172-176. Minamitake, Y., Kodama, S., Katayama, T., Adachi, H., Tanaka, S., and Tsujimoto, M. (1990).J . Biochem. (Tokyo) 107,292-297. Mita, S., Harada, N., Naomi, N., Hitoshi, Y., Sakamoto, K., Akagi, M., Tominaga, A., and Takatsu, K. (1988).J. E r p . Med. 168,863-878. Mita, S., Tominaga, A., Hitoshi, Y., Sakamoto, K., Honjo, T., Akagi, M., Kikuchi, Y., Yamaguchi, Y., and Takatsu, K. (1989a). Proc. Natl. Acad. Sci. U S A . 86, 23112315. Mita, S., Hosoya, Y., Kubota, I., Nishihara, T., Honjo, T., Takahashi, T., and Takatsu, K. (198913).J . Immunol. Methods 125,233-241. Mita, S . , Takaki, S., Hitoshi, Y., Rolink, A. G., Tominaga, A., Yamaguchi, N., and Takatsu, K. (1991a).Int. Immunol. 3,665-672. Mita, S., Kikuchi, Y., Hitoshi, Y., Sakamoto, K., Akagi, M., Tominaga, A,, and Takatsu, K. (1991b).Kumamoto Med.]. 42,63-72. Mita, S., Takaki, S., Tominaga, A., and Takatsu, K. (1993).J.Immunol. 151,6924-6932. Miyajima, A., Kitamura, T., Harada, N., Yokota, T., and Arai, K-I. (1992).Annu. Reo. Immunol. 10,295-311. Miyatake, S., Shlomai, J., Arai, K., and Arai, N. (1991).Mol. Cell. Biol. 11,5894-5898. Mizuta, T. R., Tanabe, T., Nakakubo, H., Noma, T., and Honjo, T. (1989). Growth Factors 1,51-57. Morikawa, K., Oseki, F., Morikawa, S., Imai, K., and Sawada, M. (1993).Cell. Immunol. 149,390-401. Mosmann, T. R., and Cofhan, R. L. (1989).Annu. Reo. Immunol. 7, 145-173. Mosmann, T. R., Chenvinski, H., Bond, M. W., Giedlin, M. A., and C o h a n , R. L. (1986).J . Immunol. 136,2348-2357. Munoz, E., Beutner, U., Zubiaga, A., and Huber, B. T. (1990).J.Immunol. 144,964-969. Murakami, M., Hibi, M., Nakagawa, N., Nakagawa, T., Yasukawa, K., Yamanishi, K., Taga, T., and Kishimoto, T. (1993).Science 260, 1808-1810. Murata, Y., Yamaguchi, N., Hitoshi, Y., Tominaga, A., and Takatsu, K. (1990).Biochem. Biophys. Res. Commun. 173, 1102-1108. Murata, Y., Takaki, S., Migita, M., Kikuchi, Y., Tominaga, A., and Takatsu, K. (1992). J . E r p . Med. 175,341-351. Murray, P. D., McKenzie, D. T., Swain, S. L., and Kagnoff, M. T. (1987).J . Immunol. 139,2669-2674.

186

KIYOSHI TAKATSU ET AL.

Nagai, H., Yamaguchi, S., Inagaki, N., Tsuruoka, N., Hitoshi, Y., and Takatsu, K. (1993). Life Sci. 53,243-247. Nagasawa, M., Ohshiba, A., and Yata, J. (1991).Cell. Immunol. 133,317-326. Nakajima, H., Iwamoto, I., Tomoe, S., Matsumura, R., Tomioka, H., Takatsu, K., and Yoshida, S. (1992).Am. Reo. Respir. Dis. 146,374-377. Nakanishi, K., Yoshimoto, T., Katoh, S., Matsui, K., and Hamaoka, T. (1988).J. Immunol. 144,183-192. Nakanishi, K., Matsui, S., Hirose, S., Yoshimoto, T., Hiroishi, K., Hada, T., Hamaoka, T., and Higashino, K. (1990).J. Immunol. 145, 1423-1429. Nakashima, Y., Mita, S., Takatsu, K., and Ogawa, M. (1993). Cancer Immunol. Immunother. 37,245-250. Nicola, N. S., and Metcalf, D. (1991).CeEl (Cambridge, Mass.) 67, 1-4. Noma, T., Sideras, P., Naito, T., Bergstedt-Lindqvist, S., Azuma, C., Severinson, E., Tanabe, T., Kinashi, T., Matsuda, F., Yaoita, Y., and Honjo, T. (1986).Nature (London) 319,640-646. Noma, T., Mizuta, T., Rosen, A., Hirano, T., Kishimoto, T., and Honjo, T. (1987).Immunol. Lett. 15, 249-253. Okudaira, H., Nogami, M., Matsuzaki, G., Dohi, M., Suko, M., Kasuya, S., and Takatsu, K. (1990). lnt. Arch. Allergy Appl. Immunol. 94, 171-173. Owen, W. F., Rothenberg, M. E., Petersen, J., Weller, P. F., Silberstein, D., Sheffer, A. L., Stevens, R. L., Soberman, R. J., and Austen, K. F. (1989).J . Erp. Med. 170, 343-348. Owen, W. F., Peterson, J., Sheff, D. M., Folkerth, R. D., Anderson, R. J., Sheffer, A. L., and Austen, K. F. (1990).Proc. Natl. Acad. Sci. U.S.A.87, 8647-8651. Park, L. S., Martin, U., Sorensen, R., Luhr, S., Morrissey, P. J., Cosman, D., and Larsen, A. (1992).Proc. Natl. Acad. Sci. U S A . 89,4295-4299. Pasternark, A., and Jansa, P. (1984).Neoplasm 31,323-326. Patthy, L. (1990).Cell (Cambridge, Mass.) 61, 13-14. Paul, W. E., and O'Hara, J. (1987).Annu. Rew. Immunol. 5,429-460. Paul, C. C., Kellor, J. R., Armpriester, J. M., and Bauman, M. A. (1990). Blood 75, 1400-1403. Plaetinck, G., Van der Hyden, J., Tavemier, J., Fache, I., Tuypens, T., Fischkoff, S., Fiers, W., and Devos, R. (1990).J . E x p . Med. 172, 683-691. Plaut, M., Pierce, J. H., Watson, C. J., Hanley-Hyde, J., Nordan, R. P., and Paul, W. E. (1989). Nature (London) 339,64-67. Pleiman, C. M., Gimpel, S. D., Park, L. S., Harada, H., Tanigudhi, T., and Ziegler, S. F. (1991).Mol. Cell. Biol. 11,3052-3059. Pretlow, T. P., Keith, E. F., Cryar, A. K., Bartolucci, A. A,, Pitts, A. M., Pretlow 11, T. G., Kimball, P. M., and Boohaker, E. A. (1983).Cancer Res. 43,2997-3000. Proudfoot, A. E., Fattah, D., Kawashima, E. H., Bernard, A., and Wingfield, P. T. (1990). Biochem. J . 270,357-361. Proudfoot, A. E., Davies, J. G., Turcatti, G., and Wingfield, P. T. (1991). FEBS Lett. 283,61-64. Pure, E., Isakson, P. C,, Takatsu, K., Hamaoka, T., Swain, S. L., Dutton, R. W., Dennert, G., Uhr, J. W., and Vitteta, E. S. (1981).J . Immunol. 126, 1953-1958. Purkerson, J., and Isakson, P. (1992).FASEB J. 6, 3245-3252. Paul, W. E., and Sedar, R. A. (1994).Cell 76,241-256. Rabinowitz, J. L., Tsiagbe, V. K., Nicknam, M. H., and Thorbecke, G. J. 1990). J. Immunol. 145,2440-2447. Rasmussen, R., Takatsu, K., Harada, N., Takahashi, T., and Bottomly, K. 1988). J . Immunol. 140,705-712.

IL-5 AND ITS RECEPTOR SYSTEM

187

Rawlings, D., Saffran, D. C., Tsukada, S., Largaespada, D. A,, Grimaldi, J. C., Cohen, L., Mohr, R. N., Bazan, J. F., Howard, M., Copeland, N. G., Jenkins, N. A., and Witte, 0. N. (1993).Science 261,358-360. Rennick, D. N., Thompson-Snipes, L., Coffman, R. L., Seymour, B. W. P., Jackson, J. D., and Hudak, S. (1990).Blood 76,312-316. Rivoltini, L., Viggiano, V., Spinazze, S., Santoro, A., Colombo, M. P., Takatsu, K., and Parmiani, G. (1993).Int . J . Immunol. 54,8-15. Robinson, D. S., Hamid, Q , , Ying, S., Tsicopoulos, A,, Barkans, J., Bentley, A. M., Corrigan, C. J., Durham, S . R., and Kay, A. B. (1992).N . Engl. J . Med. 326,298-304. Rolink, A., and Melchers, F. (1991).Cell (Cambridge, Mass.) 66, 1081-1094. Rolink, A. G., Melchers, F., and Palacios, R. (1989).J. E x p . Med. 169, 1693-1701. Rolink, A. G., Thalman, P., Kikuchi, Y., and Erdei, A. (1990). Eur. J. lmmunol. 20, 1949-1956. Rossi, G. A,, Crimi, E., Lantero, S., Gianiorio, P., Oddera, S., Crimi, P., and Brusasco, V. (1991).Am. Reo. Respir. Dis. 144, 379-383. Saito, H., Hatake, K., Dvorak, A. M., Leifeman, K. D., Donnenberg, A. D., Arai, N., Ishizaka, K., and Ishizaka, T. (1989).Proc. Natl. Acad. Sci. U.S.A. 85,2288-2292. Sakamaki, K., Miyajima, I., Kitamura, T., and Miyajima, A. (1992). EMBO J. 11, 3541-3549. Samoszuk, M., and Nansen, L. (1990).Blood 75, 13-16. Sanderson, C. J. (1992).Blood 79,3101-3109. Sano, Y., Yamada, G., Dobashi, K., Mizuochi, T., Hamaoka, T., and Takatsu, K. (1984). 1.Immuol. 133,629-635. Sano, Y., Harada, N., and Takatsu, K. (1985).Immunology 56, 585-596. Scher, I. (1982).lmmunol. Reo. 64, 117-136. Schimpl, A., and Wecker, E. (1972).Nature (New Biol. ) 237, 15-17. Schoenberck, S., McKenzie, D. T., and Kagnoff, M. F. (1992).Eur. J . Immunol, 19, 965-969. Schumacher, J, H., O'Garra, A., Shrader, B., van Kimmenade, A., Bond, M. W., Mosmann, T.R., and Coffman, R. L. (1988).J . Immunol. 141, 1576-1581. Sehmi, R., Wardlaw, A. J., Cromwell, O., Kurihara, K., Waltmann, P., and Kay, B. A. (1992).Blood 79,2952-2959. Shanafelt, A. B., Miyajima, A., Kitamura, T. and Kastelein, R. A. (1991).EMBOJ. 10, 4 105-4 112. Sher, A., Coffman, R. L., Hieny, S., Scott, P,, and Cheever A. W. (1990). Proc. Nutl. Acad. Sci. U S A . 87, 61-65. Shibuya, H., Yoneyama, M., Nakamura, Y., Harada, H., Hatakeyama, M., Minamoto, S . , Kono, T., Doi, T., White, R.,and Taniguchi, T. (1990).Nucleic Acids Res. 18, 3697-3703. Snapper, C . M., Waegell, W., Beernink, H., and Dasch, J. R. (1993).J. Immunol. 151, 4625-4636. Sonoda, E., Matsumoto, R., Hitoshi, Y., Mita, S., Ishii, N., Sugimoto, M., Araki, S., Tominaga, A., Yamaguchi, N., and Takatsu, K. (1989).J . E x p . Med. 170, 14151420. Sonoda, E., Hitoshi, Y., Yamaguchi, N., Ishii, T., Tominaga, A., Araki, S., and Takatsu, K. (1992).Cell. lmmunol. 140, 158-172. Stahl, N., and Yancopoulos, G. D. (1993).Cell (Cambridge,Muss.) 74,587-590. Sturrock, R. F., Kimani, R., Cottrell, B. J., Buttenvorth, A. E., Seik, H. M., Siongok, T. K., and Houba, V. (1983). Trans. R. Soc. Trop. Med. H y g . 77,363-368. Swain, S. L., Wetzel, G. D., Soubian, P., and Dutton, R. W. (1982).Immunol. Reo. 63, 77-105.

188

KIYOSHI TAKATSU ET AL.

Swain, S. L., Howard, M., Kappler, J., Marrack, P., Watson, J., Booth, R., Wetzel, G. D., and Dutton, R. W. (1983).J . E x p . Med. 158,822-835. Taga, T., and Kishimoto, T. (1992).FASEB J. 6,3387-3396. Takahji, S., Bischoff, S. C., De Weck, A. L., and Dahinden, C. A. (1991).J . Immunol. 147,3855-3861. Takahashi, M., Yoshida, M. C., Satoh, H., Hilgers, J., Yaoita, Y., and Honjo, T. (1988). Genomics 4,49-52. Takahashi, T., Yamaguchi, N., Mita, S., Yamaguchi, Y., Suda, T., Tominaga, A., Kikuchi, Y., Miura, Y., and Takatsu, K. (1990).Mol. Immunol. 9,911-920. Takaki, S., Tominaga, A., Hitoshi, Y., Mita, S., Sonoda, E., Yamaguchi, N., and Takatsu, K. (1990).EMBO J. 9,4367-4374. Takaki, S., Mita, S., Kitamura, T., Yonehara, S., Yamaguchi, N., Tominaga, A., Miyajima, A., and Takatsu, K. (1991).EMBOJ. 10,2833-2838. Takaki, S., Murata, Y., Kitamura, T., Miyajima, A., Tominaga, A,, and Takatsu, K. (1993). J . E x p . Med. 177,1523-1529. Takatsu, K. (1988).Proc. SOC. E z p . Biol. Med. 188,243-258. Takatsu, K. (1992). Cuw. Opinion Immunol. 4,299-306. Takatsu, K., and Hamaoka, T. (1982).Immunol. Reu. 64,25-55. Takatsu, K., and Kikuchi, Y. (1988).Gann Monogr. Cancer Res. 34, 143-154. Takatsu, K., and Tominaga, A. (1991).Prog. Growth Factor Res. 3,87-102. Takatsu, K., Haba, S., Aoki, T., and Kitagawa, K. (1974).lmmunochemistry 11,107-109. Takatsu, K., Tominaga, A., Hamaoka, T., and Kitagawa, M. (1979). Microbiol. Immunol. 23,1109-1125. Takatsu, K., Tominaga, A., and Hamaoka, T. (1980a).J . Immunol. 124,2414-2422. Takatsu, K., Tanaka, K., Tominaga, A,, Kumahara, Y., and Hamaoka, T. (1980b)~. Immunol. 125,2646-2653. Takatsu, K., Sano, Y., Tomita, S., Hashimoto, N., and Hamaoka, T. (1981a). Nature (London)292,360-362. Takatsu, K., Tominaga, A., and Hamaoka, T. (1981b).Immunol. Lett. 3, 137-143. Takatsu, K., Sano, Y., Tomita, S., Hashimoto, N., and Hamaoka, T. (1982).J . Immunol. 128,2575-2580. Takatsu, K., Harada, N., Hara, Y., Takahama, Y., Yamada, G., Dobashi, K., and Hamaoka, T. (1985).J . Immunol. 134,382-389. Takatsu, K., Kikuchi, Y., Matsumoto, M., Harada, N., Yamaguchi, N.. and Tominaga, A. (1987).Proc. Natl. Acad. Sci. U S A . 84,4234-4238. Takatsu, K., Tominaga, A., Harada, N., Mita, S., Matsumoto, M., Takahashi, T., Kikuchi, Y., and Yamaguchi, N. (1988).Immunol. Aeu. 102, 107-135. Takatsu, K., Takaki, S . , Hitoshi, Y., Katoh, S., Yamaguchi, N., and Tominaga, A. (1992). Ann. N.Y. Acad. Sci. 651,229-241. Tanabe, T., Konishi, M., Mizuta, T., Noma, T., and Honjo, T. (1987).J.Biol. Chem. 262, 16,580-16,584. Tavemier, J., Devos, R., Cornelis, S., Tuypens, T., Van der Heyden, J., Fiers, W., and Plaetinck, G. (1991).Cell (Cambridge,Mass.)66, 1175-1184. Tavemier, J., Tuypens, T., Plaetinck, G., Verhee, A., Fiers, W., and Devos, R. (1992). Proc. N d . Acad. Sci. U S A . 89, 7041-7045. Tepper, R. I., Pattengale, P. K., and Leder, P. (1989). Cell (Cambridge, Mass.) 57, 503-512. Tepper, R. I., C o h a n , R. L., and Leder, P. (1992). Science 257,548-551. Thomas, J. D., Sideras, P., Smith, C. I. E., Vorechovsky, I., Chapman, V., and Paul, W. E. (1993).Science 261,355-358. Tominaga, A., Takatsu, K., and Hamaoka, T. (1980).J . Immunol. 124,2423-2431.

IL-5 AND ITS RECEPTOR SYSTEM

189

Tominaga, A., Takatsu, K., and Hamaoka, T. (1982).J . Immunol. 128,2581-2585. Tominaga, A., Matsumoto, M., Harada, N., Takahashi, T., Kikuchi, Y., and Takatsu, K. (1988).J . Immunol. 140, 1175-1181. Tominaga, A., Mita, S., Kikuchi, Y., Hitoshi, Y., Takatsu, K., Nishikawa, S-I., and Ogawa, M. (1989). Growth Factors 1, 135-146. Tominaga, A., Takahashi, T., Kikuchi, Y., Mita, S., Naomi, S., Harada, N., Yamaguchi, N., and Takatsu, K. (1990).J . Immunol. 144, 1345-1352. Tominaga, A., Takaki, S., Koyama, N., Katoh, S., Matsumoto, R., Migita, M., Hitoshi, Y., Hosoya, Y., Yamauchi, S., Kanai, Y., Miyazaki, J-I., Usuku, G., Yamamura, K-I., and Takatsu, K. (1991).J . Erp. Med. 173,429-437. Tonegawa, S. (1983).Nature (London) 302,575-581. Tsujimoto, M., Adachi, H., Kodama, S., Tsruoka, N., Yamada, Y., Tanaka, S., Mita, S., and Takatsu, K. (1989).J.Biochem. (Tokyo) 106,23-28. Tsukada, S., SaEran, D. C., Rawling, D. J., Parolini, O., Allen, R. C., Klisak, I., Sparkes, R. S., Kubagawa, H., Mohandas, T., Quan, S., Belmont, J. W., Cooper, M. D., Conley, M. E., and Witte, N. (1993).Cell (Cambridge, Mass.) 72,279-290. Ullrich, A., and Schlessinger, J. (1990). Cell (Cambridge,Mass.) 61, 203-211. Van Oosterhout, A. J. M., Ladenius, A. R. C., Savelkoul, H. F. J., van Ark, I., Delsman, K. C., and Nijkamp, F. P. (1993).Am. Rev. Respir. Dis.147,548-552. Vaux, D. L., Lalor, P. A., and Johnson, G. R. (1990).Int. Zmmunol. 2,965-971. Vellenga, E., Esselink, M. T., Straaten, J., Stulp, B. K., d e Wolf, J. T. M., Brons, R., Gliannotti, J., Smit, J. W., and Halie, M. R. (1992).J . Immunol. 149,2992-2995. Vetrrie, D., Vorechovsky, I., Sideras, P., Holland, J., Davies, A., Flinter, F., Hammerstrome, L., Kinnon, C., Levinsky, R., Bobrow, M., Smith, C. I. E., and Bentley, D. R. (1993).Nature (London) 361,226-233. Vitteta, E. S., Brooks, K., Chen, Y. W., Isakson, P., Jones, S., Layton, J. E., Mishra, G. C., Pure, E., Weiss, E., Word, C., Yuan, D., Tucker, P., Uhr, J. W., and Krammer, P. H. (1984). Immunol. Reu. 78,137-158. Walker, C., Virchow, J. C. Jr., Bruijnzeel, P. L., and Blaser, K. (1991).J.Immunol. 146, 1829-183s \ Walsh, G. M., Hartnell, A., Wardlaw, A. J., Kurihara, K., Sanderson, C. J., and Kay, A. B. (1990). Immunology 71,258-265. Watanabe, S., Mui, A. L.-F., Muto, A., Chan, J. X., Hayashida, K., Yokota, T., Miyajima, A., and Arai, K. (1993).Mol. Cell. B i d . 13, 1440-1448. Webb, S. F., Das,C., Eaton, S., Calame, K., and Tucker, P. W. (1991).Mol. Cell. Biol. 115,5197-5205. Wetzel, G. D. (1989).Eur. J . Immunol. 19, 1701-1707. Witthuhn, B. A., Quelle, F. W., Silvennoinen, O., Yi, T., Tang, B., Miura, O., and Ihle, J. N. (1993). Cell (Cambridge,Muss.) 74,227-236. Wrighton, N., Campbell, L. A., Harada, N., Miyajima, A., and Lee, F. (1992). Growth Factors 6, 103-118. Wu, H.-K., Hirai, H., Inamori, K., Kitamura, K., and Takaku, F. (1992).Jpn. J . Cancer Res. 83,200-210. Yamaguchi, N., Takahashi, T., Harada, N., and Takatsu, K. (1989).J. Biochem. (Tokyo) 106,837-843. Yamaguchi, N., Hitoshi, Y., Mita, S., Hosoya, Y., Kikuchi, Y., Tominaga, A., and Takatsu, K. (1990).Int. Immunol. 2, 181-187. Yamaguchi, N., Hitoshi, Y., Takaki, S., Murata, Y., Migita, M., Kamiya, T., Minowada, J., Tominaga, A., and Takatsu, K. (1991). Int. Immunol. 3,889-898. Yamaguchi, S.,Okubo, Y.,Hossain, M., Fujimoto, K., Honda, T., Kobayashi, T., Sekiguchi, M., and Takatsu, K. (1994).Am. Rev. Respir. Dis.,in press.

190

KIYOSHI TAKATSU ET AL.

Yamaguchi, Y., Suda, T., Suda, J., Eguchi, M., Miura, Y., Harada, N., Tominaga, A., and Takatsu, K. (1988a). J. Exp. Med. 167, 43-56. Yamaguchi, Y., Hayashi, Y., Sugama, Y., Miura, Y., Kasahara, T., Torisu, T., Mita, S., Tominaga, A., Takatsu, K., and Suda, T. (1988b).J. Exp. Med. 167,1737-1742. Yamaguchi, Y., Matsui, T., Kasahara, T., Etoh, S., Tominaga, A., Takatsu, K., Miura, Y., and Suda, T. (l990a). Exp. Hematol. 18, 1152-1157. Yamaguchi, Y., Suda, T., Nakano, K., Shiozaki, H., Miura, Y., Hitoshi, Y., Tominaga, A., Takatsu, K., and Kasahara, T. (1990b).J. Zmmunol. 145,873-877. Yamaguchi, Y., Suda, T., Suda, J., Miura, Y., Mita, S., Hitoshi, Y., Tominaga, A., and Takatsu, K. (1990~).Acta. Hematol. Jpn. 144, 688-698. Yamaguchi, Y., Suda, T., Ohta, S., Tominaga, A., Miura, Y., and Kasahara, T. (1991). B10c)d 78, 2542-2547. Yamanashi, Y., Okada, M., Sembe, T., Yamori, T., Umemori, H., Tsunasawa, S., Toyoshima, K., Kitamura, D., Watamabe, T., and Yamamoto, T. (1993). Proc. Natl. Acad. Sci. U.S.A.90, 3631-3635. Yokota, T., Coffman, H. L., Hagiwara, H., Rennick, D. M., Takebe, Y., Yokota, N., Cemmell, L., Shrader, B., Yang, G., Meyerson, P., Luch, J., Hoy, P., Pene, J., Briere, F., Spitz, € I . , Bancherean, J., de Vries, J . E., Lee, F. D., Arai, N., and Arai, K. (1987). Proc. N a t l . Acad. Sci. U.S.A.84, 7388-7392. Yonchara, S.,Ishii, S., Yonehara, M., Koyasu, S., Miyajima, A., Scheurs, J., Arai, K., and Ynhara, 1. (1990).Znt. Zmmunol. 2, 143-150. Yoshida, K., Matsuoka, M., Usuda, S., Mori, A., Ishizaka, K., and Sakano, H. (1990). Proc. N a t l . Acad. Sci. U.S.A. 87, 7829-7833. Youssoiifian, H., Zon, L. I., Orkin, S . H., D’Andrea, A. D., and Lodish, H. F. (1990). M o l . Cell. Biol. 10, 3675-3682.

ADVANCES IN IMMUNOLOGY, VOL. 57

Human Antibodies from Combinatorial libraries DENNIS R. BURTON AND CARLOS F. BARBAS 111 Departments of Immunology and Moluular Biology, lhe Scripps Research Institute, 10666 North Tomy Pines Rwd, La Jolla, California 92037

1. Introduction

The survival ofanimals in a hostile environment teeming with pathogens has necessitated the evolution of complex systems of immune defense. These systems contain molecules, primarily antibodies, and T-cell receptors, with exquisite specificities which have naturally attracted attention as potential therapeutic agents. Antibodies are clearly more straightforward to use in this role since they are soluble protein molecules whereas T-cell receptors are of course cell associated and require MHC compatibility for recognition. Indeed over 100 years ago, Emil von Behring demonstrated the efficacy of hyperimmune horse serum containing high titers of antibody to diptheria in combating an outbreak of disease in humans (Behring, 1893). However, despite this early success, the use of passive immunization through the 20th century has been very limited. This probably reflects the great success of antibiotics in combating bacterial disease, of vaccines in the prevention of viral disease, and the difficulties of obtaining antibodies useful for passive immunization. The emergence of AIDS in the 1980s and of a spectrum of infections in a large pool of immunocompromised people has once more focused attention on infectious disease. This in turn has renewed interest in passive immunization at a time, we will argue, of rapid progress in the generation and design of antibodies for medicine. Antibodies are generally thought to have evolved to recognize and eliminate foreign pathogens. However, the ability to target antibodies to self-molecules provides new therapeutic opportunities. These include immunomodulation, e.g., immunosuppression, antiinflammation, and cancer. Further human antibodies could be used to explore the function of certain molecules in humans by neutralizing such molecules in v i m . Immunotherapy prefers a ready supply of specific human antibodies. Hyperimmune human serum preparations have proved useful in a number of instances, notably to achieve a great decrease in the developed countries in incidence of hemolytic disease of the newborn through anti-rhesus D prophylaxis. The disadvantages of hyperim191

Copyright 8 1994 by Academic Press, Inc

All rights of reproduction In any form reserved.

192

DENNIS R. BURTON AND CARLOS F. BARBAS I I I

mune serum are that it is only practically available in limited instances and that it is inherently extremely inefficient since only a small proportion of the preparation constitutes specific functional antibody. Hybridoma technology offers unlimited supplies of specific rodent monoclonal antibodies but has been much less successful for human monoclonal antibodies ( James and Bell, 1987). Epstein-Barr virus transformation of human B cells (Casali and Notkins, 1989) has met with some success but again has major limitations. “Humanization” of existing mouse monoclonal antibodies by variable domain replacement (Boulianne et d.,1984; Morrison et d.,1984; Neuberger et d., 1985) or by CDR grafting (Jones et al., 1986) has been achieved and this allows for the generation of anti-self-antibodies. The chimeric antibodies are easily produced but retain the mouse variable domains with possible consequences for immunogenicity. The CDR-grafted antibodies are more nearly human but take longer to produce and may require adjustments to the human frameworks. A new technology for the generation of human monoclonal antibodies has been developed over the past 5 years based on cloning antibody fragments in phage libraries. The vectors used were initially A phage (Huse et al., 1989) but these have been superseded by filamentous phage display vectors (McCafferty et al., 1990; Kang et aZ., 1991a; Barbas et al., 1991; Clackson et al., 1991). II. Principles of Combinatorial libraries

A. THECOMBINATORW APPROACH The “combinatorial approach” of cloning antibody repertories was initially developed for application in the field of catalytic antibodies. Traditional hybridoma methodologies allow for a limited sampling of the immune repertoire. Cloning the repertoire in Escherichia coli would allow a much more extensive survey of the immune response and might be essential for screening or selecting for rare catalytic activities. This approach became possible due to two developments. First, it was demonstrated in 1988 that the antigen binding fragments Fv (variable heavy, VH,variable light, V,, Fig. 6) and Fab could be expressed and functionally assembled in E. coli (Skerra and Pluckthun, 1988; Better et al., 1988). These experiments were successful because the antibody fragments were secreted from the cytoplasm to the oxidizing environment of the periplasm in E. coli under the guidance of bacterial leader sequences. The oxidizing environment and possibly the secretory event were necessary for disulfide bond forma-

ANTIBODIES FROM COMBINATORIAL LIBRARIES

193

tion and proper folding of antibody domains. This is directly analogous to the production of antibodies in eukaryotic cells where the two chains are transported from the cytoplasm to the lumen of the endoplasmic reticulum. Second, the development of the polymerase chain reaction (PCR)allowed for the rapid cloning of antibody genes from hybridomas (Larrick et aZ., 1989; Orlandi et al., 1989) and mixed populations of antibody producing cells (Sastry et al., 1989). Within the year, these two developments were utilized to produce the first combinatorial antibody library (Huse et al., 1989). At the same time an alternative approach consisting of single heavy-chain variable domains (dAbs) was reported (Ward et al., 1989).This dAb approach makes the assumption that the role of the light chain in antigen binding is negligible. Crystallographic studies indicate that light chains often make considerable contact with antigen and can play an important role both in determining affinity and in specificity. Furthermore and equally important, single VH domains are not physicochemically well-behaved due to exposure of a large hydrophobic patch on the domain usually involved in light-chain association. Indeed, the dAb approach has not proven to be generally applicable. The first combinatorial library experiment reported in 1989 involved screening antibody Fab fragments for binding of labeled antigen. In this case the libraries were expressed in X phage, a lytic phage which upon lysis releases the periplasmically sequestered Fab (Huse et al., 1989). In these experiments a mouse was immunized with the hapten designed to elicit catalytic antibodies and RNA was prepared from the spleen. After reverse transcription, the cDNA of antibody heavy chains (Fd part of IgG1) and light chains were amplified by the PCR reaction and ligated into modified X phage vectors to give libraries of heavy and light chains. The two libraries were then combined by digestion of opposite arms of the vectors and religation to generate a random combinatorial library containing the genetic information for the production of Fab fragments. The library was screened by transfer of Fabs produced by A phage-lysed E. coZi onto nitrocellulose filters. The filters were then probed with '251-labeled hapten conjugate (Fig. 1)and revealed a high frequency of positives in the library (about 1 in 5000) which allowed 200 monoclonal Fab fragments to be identified following an examination of lo6 Fabs. The examination of this library of a million clones, which as we see later is a relatively modest library size, required 20 filter lifts. Analysis of 22 of the positive Fabs showed sequence diversity and apparent binding affinities of the order of 107 M - 1. With the demonstration of the applicability of the system for the

FIG. 1. Screening the first combinatonal library for antigen binding (Huse et al., 1989).A mouse was immunized with the hapten NPN on the carrier KLH and a spleen library prepared in A phage. The Fab library (A and B), heavy-chain alone library (E and F), and light-chain alone library (G and H) were screened in duplicate on nitrocellulose filters at approximately 30,000 plaques per filter using a radiolabeled

ANTIBODIES FROM COMBINATORIAL LIBRARIES

195

cloning of mouse antibodies, the next step required a model system for cloning human antibodies. Peripheral blood lymphocytes from an individual recently boosted with tetanus toxoid (Person et al., 1991) were utilized to construct an IgGlK library. The antibodies generated in this case showed considerable sequence diversity and apparent affinities in the range 107-109 M-'. Similar studies have also been reported elsewhere (Mullinax et al., 1990). Recent boosting was very important in that it was not possible to isolate antigen binding Fabs from an individual with a high anti-tetanus toxoid titer who had not been boosted. This probably reflects the presence of antigen-specific plasma cells with their high concentration of specific mRNA in the peripheral blood of boosted subjects as compared to the low resting level of plasma cells in PBLs. The method has also been applied to the generation of human autoantibodies (e.g., Portolano et al., 1991; Hexham et al., 1991).

B. PHAGEDISPLAY OF ANTIBODYFRAGMENTS The screening procedure limits the size of the library which may be examined in the A phage system. For example, the screening of a diverse library of 5 x lo8 antibodies would require an examination of a minimum of 10,000 filter lifts (at 50,000 plaques per plate). Furthermore, the screening procedure places restrictions on the antigens which are being examined in that the antigen must be available in significant quantities in purified form, be amenable to labeling with lZ5I or enzymes, and should not stick significantly to filters in the absence of antibody. This is very restrictive ifthe interest is in isolating antibodies against proteins which have yet to be identified or characterized, found for example on the surface of a cell or in a crude protein extract such as a viral lysate. Selection is inherently a more efficient method for deriving positive clones from a library than screening since virtually no effort is spent examining the vast majority of negative clones. Selection is practiced very effectively by the immune system through the expression ofimmunoglobulin on the surface of B cells. The selected clones are then amplified by the linkage of this recognition device to the replication of the genetic information within the B cell and clonal expansion. It BSA-NPN conjugate. Filters C and D illustrate the duplicate secondary screening of a positive from a primary screening (e.g., arrows on filters A and B). As expected for a bona fide positive, the frequency of positives in the secondary screening was much higher than that in the primary screening. No positives were identified in either the heavy- or light-chain alone libraries.

196

DENNIS R. BURTON AND CARLOS F. BARBAS

111

was therefore logical to seek a selection system in which recognition and replication could be linked. George Smith (1985) had shown that peptides could be expressed on the surface of filamentous phage indicating how this linkage might be achieved. The concept of selectable phage display libraries had been established in 1990 with peptide libraries (Scott and Smith, 1990; Devlin et al., 1990; Cwirla et al., 1990). Fragments of P-galactosidase had previously been displayed on phage (Parmley and Smith, 1988). The concept needed only be extended to libraries of proteins. This would allow selection of specific antibodies based on their ability to bind to immobilized antigen largely circumventing the problems associated with screening. An outline of the strategy is presented in Fig. 2. For a general review of phage display see Barbas, 1993. A brief introduction to filamentous (Ff) phage biology makes their use in cloning or creating antibodies more obvious (reviewed in Model and Russel, 1988).Ff phage, most notably fl, fd, and M13 (the three are

PELS, Bone Marrow, Spleen

Discard Nonspeciflc Binders

Semi-Synthetic Antibody Genee

/Infect Bacteria Converslon to Soluble Fab Expressing Phagemid

1

Cheracteriratlon of Fab

FIG.2. Strategy for cloning human monoc-mi Fab fragments from combinatorial libraries on the surface of phage. The strategy has also been applied to single-chain Fv fragments (scFv). The antibody fragment is shown as a fusion with the capping protein 3 (g3p or cpIII) in monovalent display. Multivalent display using g3p has also been described. Fusion with phage coat protein 8 (g8p or cpVIII) displaying the antibody fragment multivalently along the phage surface is another format which has been adopted. After Burton and Barbas (1993a).

ANTIBODIES FROM COMBINATORIAL LIBRARIES

197

almost identical) infect gram-negative bacteria by virtue of a specific interaction between one end of the phage and the bacterial pilus. Ff phage are long and thin (ca. 900 x 6-10 nm, Fig. 3); essentially they are tubes of protein encapsulating a single-stranded closed circular DNA genome. The mature Ff phage consists of five proteins and a covalently closed genome of approximately 6.4 kb. Approximately 2700 copies of the gene 8 protein (g8p or cpVIII) form all but the end structures. Minor coat proteins (g3p or cpIII, g6p, g7p, g9p) are present in about 5 copies each. The proteins are arranged so that g3p and g6p are expressed at the tail of the phage, g7p and g9p at the head. g3p is the phage protein that makes specific contact with the bacterial receptor. Rough genetic mapping demonstrated that gp3 has two domains, an amino-terminal domain important for phage infectivity and a carboxy-terminal domain important for closing or capping the phage tube. In addition to the coat proteins are an endonuclease/topoisomerase (gzp), a single-stranded DNA binding protein (g5p), and several less-studied proteins: glp, gXp, g4p. An important regulatory element, the intergenic region (IR), is also present. The IR contains a DNA origin of replication and a DNA packaging signal. A plasmid with a Ff IR, most commonly fl, is known as a phagemid. The IR is sufficient to allow helper phagemediated replication and packaging of the phagemid single-stranded DNA. Packaged phagemids can infect bacteria and are known as transducing particles. Unlike A phage in which the first repertoire cloning experiments were performed, the Ff phage are not assembled in the cytoplasm and are not released by cell lysis. They are instead extruded through the outer membrane leaving the cell unharmed. The Ff phage are extruded as they are assembled in the bacterial membrane. Assembly begins with the coating of genomic (or phagemid) single-stranded DNA by g5p. This nucleoprotein complex then migrates to the inner membrane. Coat proteins then replace g5p in a vectorial fashion-g7p and g9p are first, followed by polymerization of g8p onto the DNA rod, and finally g6p and g3p cap the particle. Since g3p and g8p are secreted and anchored in the periplasmic space of E . coli and are assembled to form the surface of the phage, they are ideal targets for fusion with antibody domains. Note that secretion into the periplasmic space is a prerequisite for proper folding and assembly of antibodies in all but specifically engineered E. coli (C.F.B., unpublished data). Phage displaying antibody fragments can then be sorted by a process known as panning to isolate those which encode the desired specificity and highest affinity as described below.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

199

C. PHAGEDISPLAY VECTORSYSTEMS The phage genome itself may serve as a cloning vehicle, i.e., genes can be directly cloned into the phage genome. Genomic g3p fusion systems allow the polyvalent display of foreign peptides. In these systems every copy of the three to five copies of g3p is fused with the displayed protein or peptide. With multivalent display of antibodies on the phage surface, avidity effects minimize the selection efficiencies between clones which bear antibodies even if they differ by two orders of magnitude in affinity. The standard genomic fusion phage vector is that first described by Smith, fd-tet. With this system it is difficult to prepare the double-stranded DNA necessary for cloning. Genomic g3p fusions have been highly successful in the area of peptide libraries. The display of proteins has been described using this system and includes the display of fragments of P-galactosidase and murine singlechain antibodies. Genomic cloning need not be limited to fusions with the native coat protein. As an alternative, a second coat protein gene may be introduced as the carrier of the fusion. This type of system would have a decreased valency of display but has yet to be utilized for antibody expression. Phagemid systems offer an attractive alternative to cloning directly into the phage genome. Phagemids are simply plasmids which bear the IR region of an Ff phage. The genomes of phagemids, when propagated in cells superinfected with a helper phage, will be packaged as phage particles in a fashion identical to the phage itself. Phagemid systems have two distinct advantages over cloning into the phage genome. First, double-stranded DNA is easily obtained making library construction less tedious and, second, the valency of the displayed protein may be controlled. Phagemid systems have been described using both g8p and g3p fusions (Fig. 4). For display purposes the phagemid also carries either g3 or g8 on which the gene of the protein to be displayed is fused. Subsequently the use of helper phage superinfection leads to the expression of two forms of the coat protein, one FIG.3. (a) Electron microscopy of hepatitis B surface antigen particles and filamentous phage. Fresh phage were prepared from the third pan stage of a library from a donor recently boosted with the hepatitis antigen. The phage were incubated with antigen particles, added to grids coated with antibody against the hepatitis antigen, and negatively stained. Magnification 28,OOOx. From Zebedee et al., 1992; figure kindly provided by Suzanne Zebedee. (b) Transmission electron microscopic images of antirhesus D antigen (Rh(D))expressing phage bound to the surface of an Rh(D)-positive red cell. Magnification 44,OOOx. From Siegel and Silberstein (1994); figure kindly provided by Don Siegel.

Periplasm CHI&

CH1

'\oat

Protein

Ill

X

** c

pComb 3 h , 11

b -

lnfectlon

Extruslon Outer Membrane Inner Membrane

ds. Phage DNA

pComb3 Phagemld

S.S.

Phagemid DNA

ANTIBODIES FROM COMBINATORIAL LIBRARIES

20 1

native and one fusion protein. On phage assembly there is a competition between these proteins for incorporation into the virion. The g8p system like the genomic g3p system is multivalent. The display of Fab fragments as g8p fusions was described initially by Kang et al. (1991a) and later by Chang et al. (1991). Sorting of antibodies of different specificities was demonstrated in one case with a 1000-foldenrichment in a single panning step. Examination of packaged phagemid displaying antibodies revealed the presence of a variable number of Fab fragments ranging from 1to 24. Further examination of this system demonstrated that the avidity effect of multiple copies of Fab on the phage surface does not allow for the selection of antibodies of the highest affinity. In panning experiments with human anti-tetanus toxoid antibodies which differed by 100-fold in affinity only a 5-fold enrichment could be demonstrated in a single selection (Barbas et al., 1991). In contrast to g8p fusions, g3p fusions can be made monovalent. In the virion there are normally three to five copies of g3p and phage morphogenesis leads to incorporation of the Fab-g3p fusion and the native helper phage produced g3p into the virion (Fig. 5). Native g3p is necessary for infection as the infectivity domain should not be present in the Fab-g3p fusion (vida infra). The fusion, generally one

FIG.4. (a) The composition of a phage display vector, pComb3, and the proposed pathway for Fab assembly. Expression of Fd (Hc, heavy chain)/cpIII (g3p) fusion and light chain (Lc) is controlled by lac promoter/operator sequences. The chains are directed to the periplasmic space by pelB signal sequences which are subsequently cleaved. The heavy chain is anchored to the membrane by the cpIII fusion whereas the light chain is secreted into the periplasm. The two chains then assemble on the membrane. (b) Helper phage rescue of phagemid to give a phage display library. Helper phage, a single-stranded DNA virus in a protein coat, infects E. coli and the singlestranded DNA is converted to double-stranded DNA. This codes for a number of phage proteins. Some of these are coat proteins which accumulate in the inner membrane of the cell. Others act on the pComb3 phagemid DNA, because of the f l sequence (a), causing the packaging of a single-stranded copy of the phagemid DNA. This is packaged in preference to the helper phage DNA which has been mutated to decrease packaging efficiency. As the assembling phage is extruded from the cell it is “capped” by coat protein 111 which includes native cpIII and the cpIII-Fab complex. Thus a phage results with an Fab displayed on the surface and the phagemid DNA, containing the corresponding antibody genes, inside. The advantage ofphagemid rescue as opposed to directly cloning antibody genes into phage are twofold. Direct cloning would mean that every cpIII molecule would carry Fab. This would greatly reduce infectivity since cpIII is involved in entry to E. coli. Further multivalent display is expected to hinder selection of binders on the basis of a n i t y in the panning process because of chelation effects. Phagemid rescue probably leads to primarily monovalent Fab display.

202

DENNIS R. BURTON AND CARLOS F. BARBAS I11

Monovalent Display

Multivalent Display

FIG.5. Comparison of antibody display on coat protein I11 (g3p) produced by direct cloning into the phage genome or via phagemid rescue. In the case of direct cloning, the phage contains phage DNA and each molecule ofcoat protein 111 carries an antibody fragment leading to multivalent display. In the case of phagemid rescue, the phage contains phagemid DNA and display will be statistically determined. Monovalent display will likely predominate.

copy per phage particle, is displayed in functional form on the surface of the phage and is therefore available for antigen selection. Fusions with g3p should respect the function of this 406-residue protein. Studies of g3p have revealed the two functional properties, infectivity and normal (nonpolyphage) morphogenesis, which map to roughly the first and second half of the gene, respectively. The N-terminal domain of g3p binds to the F’ pili, allowing for infection of E . coli, whereas the membrane-bound C-terminal domain, P198-S406, serves the morphogenic role of capping the trailing end of the filament according to the vectorial polymerization model. Another important biological feature of the N-terminal domain of g3p is its role in providing immunity to an Ff-infected cell to superinfection by other Ff phage. This fact has important consequences in the design of phagemid vectors which utilize g3p fusions. Since superinfection with helper phage is required in this approach, the N-terminal domain should be deleted in the fusion provided by the phagemid. Otherwise, superinfection will be prevented. This feature was incorporated in the first vector designed for Fab display, pComb3 (Barbas et al., 1991), and is also found in pDH188 (Garrard et al., 1991) and pEXmide3 (Soderlind et al., 1992), which are also designed for Fab display. Previously this same design was utilized in the display of growth hormone (Bass et al., 1990). Several g3p phagemid systems, pHEN (Hoogenboom et al, 1991), pSEX (Dubel, 1993; single-chain antibody display only), and pCAN-

ANTIBODIES FROM COMBINATORIAL LIBRARIES

203

TAB (single-chain antibody display only, Pharmacia), utilize the entire g3p as a fusion partner. This approach necessitates additional steps in the selection process, first to shut down expression to allow superinfection and then to induce expression to allow for display. Monovalency in g3p phagemid systems has important consequences in selection of the highest affinity antibody. The only study of the role of valency in the separation of clones of defined affinities has been reported with pComb3 (Barbas et al., 1991). In this study, phagemid vectors were constructed which where identical except for their antibiotic resistance. Enrichment could be monitored by simply plating on chloramphenicol and ampicillin plates. Enrichment of 253-fold with a single round of selection was observed in the separation of two human anti-tetanus toxoid Fabs with affinities that differed by two orders of magnitude. In an identical experiment, the multivalent g8p system, pComb8, yielded only a 5-fold enrichment. Note that monovalency in the g3p system can only be ensured when Fab fragments are displayed. In contrast, single-chain antibodies, scFv, have been found in many cases to spontaneously dimerize (Holliger et al., 1993).

D. CLONING STRATEGIES Library construction begins with mRNA that is used as a template for cDNA synthesis. cDNA then becomes the template for the polymerase chain reaction. Two types of antibody fragments may be chosen for construction, Fab or scFv. Fab fragments, the natural antigen binding fragments of whole antibodies, retain the binding characteristics of the whole antibody with the exception of avidity. Single-chain antibodies reproduce the binding characteristics of the whole antibody to variable extents and may have greatly reduced binding affinity for antigen as compared to the whole antibody (Bird and Walker, 1991; Whitlow and Filpula, 1991). Detection and purification of the fragments should also be considered. Fab fragments may be detected with many commercially available reagents (anti-Fab, anti-rc etc.) with many reporter enzyme conjugates. scFv may only be detected by including a peptide tag at the tail of the protein and detection is limited by the anti-peptide antibody which must be used. Purification regimes are also more limited with scFv's. The properties of increased tissue penetration may, however, favor the use of scFv for particular applications. Thorough consideration of the end use of the antibody fragments should dictate the strategy to be taken. The protocols for the construction of Fab or scFv libraries are quite distinct. Consider the construction of IgGlK and IgGlh Fab libraries (Fig 6). For IgGl PCR, we commonly use eight 5 ' - v H and one 3'-1gG1 and one 3 ' - K primer primers to amplify VHgenes. For K PCR, five ~ ' - V K

204

DENNIS R. BURTON ANT, CARLOS F. BARBAS I11

FIG.6. The human IgGl molecule. For generation of Fab libraries, PCR primers are chosen to amplify DNA corresponding to the Fd part (VHCHI) of the heavy chain and the whole of the light chain. Equivalent positions on the protein are represented by arrows in the figure. The 3’ primer for the heavy chain hybridizes to a hinge region sequence to include the cys involved in the heavy-light chain disulfide bridge in IgG1. Other antibody isotypes have been amplified by a similar strategy. For generation of single-chain Fv libraries, the 3’ primers are placed at positions corresponding to the C-termini of the VH and V, domains.

are used. For A, eight 5’-VA and one 3’-A primer are used. Following PCR, all IgGl products are pooled as well as all K and A products. The three pools are prepared for cloning by restriction digest. The prepared K and A products may be mixed and ligated with cut phagemid. Electroporation then allows for the efficient introduction of ligated products to form the light-chain library. From this library, phagemid is prepared for heavy-chain cloning and the combinatorial library is formed and ready for selection following helper phage rescue. A total of 21 PCR primers are required for completeness, a single PCR step is performed, and two ligation and transformation steps are performed. Libraries of about 10’ transformants can be economically prepared using electroporation in the transformation steps. Now let’s consider the construction of scFv libraries of IgGlK and IgGlh (e.g., Marks et al., 1991).A major difference in the construction of scFv libraries is that there is no antibody constant region in this construct. This fact requires the utilization of multiple J region primers while for Fab cloning a single primer would suffice. For SCFV,VH construction requires six 5’-vH primers and four 3’-JHprimers (JH 2 and 5 not reported). VK requires six 5 l - V ~and five ~ ’ - J Kprimers. VA requires seven 5‘-VA and three 3’-JAprimers. PCR is performed with these sets of primers. Subsequently another round of PCR to create the linker regions is required with an additional four JH, six VK, and seven VA-linker primers in 52 separate PCR reactions. Now in a third

ANTIBODIES FROM COMBINATORIAL LIBRARIES

205

round of PCR, VH is fused with VKand VA and the linkers first without addition of primers and then in another step with the addition of primers. This fused product is then subjected to a fifth round of PCR with an additional new set of six V,, five JK, and three JA primers, which add restriction sites. The products are subjected to a restriction digest and ligated and cloned in one step. In five PCR steps with 62 different PCR primers and one ligation and transformation step, the library is constructed. In our estimation, the difficulty and cost of constructing representative scFv libraries can only be justified when scFv is required as the end product. Indeed, it may be more effective to first construct and select clones from a Fab library and convert the desired clones into scFv. This two-step strategy also circumvents the multivalency problem in the selection of clones of the highest affinity as the library is selected by way of Fab display. An alternative approach to the construction of Fab libraries has been suggested (Hogrefe et al., 1993). This approach combines aspects of the original A phage approach with phagemid Fab and scFv protocols. Light-chain amplification is performed as described in the Fab construction protocol with an additional PCR step to fuse the heavy-chain leader sequence to the light chain. Next, V, regions are amplified requiring the multitude of primers required in scFv construction. The two fragments are then cut with restriction enzymes, ligated, and subject to another round of PCR after which the single light-chain leader V, fragment is again cut with restriction enzymes. This product is then ligated with a ImmunoZap 13 vector encoding a phagemid which is a close copy of pComb 3 containing a CHI region. I n vitro packaging then allows for the construction of a A library. To obtain Ff phage display, helper phage is used in a mass excision step resulting in packaged phagemid. The rationale of this approach is to simplify library construction by taking advantage of the high efficiency of in vitro packaging of A phage as an alternative to electroporation. The strategy assumes that PCR product is limiting in library construction although we would suggest that more product is easily produced with additional PCR reactions. Furthermore in vitro packaging, though efficient, is extremely expensive and limited to a small scale making the construction of libraries of 10' clones tedious, if not economically unfeasible. There is no report in the literature, to our knowledge, of the construction of A libraries of this magnitude. The utilization of a VH cloning strategy further complicates library construction by necessitating more PCR primers and additional steps. An additional concern is the lytic nature of A phage which frequently leads to serious problems of contamination in the laboratory.

206

DENNIS R. BURTON AND CARLOS F. BARBAS 111

Huse et al. (1992) suggest a modification of the original lambda phage approach for library screening. In this report M13 Ff phage is utilized with a second copy of g8p to which a Fab is fused. The library is then screened utilizing essentially the same protocol described for A phage. No selection was attempted and the g8p is utilized only to expose the Fab as it is secreted with the phage. A limitation of this approach is its requirement for screening.

E. SELECTION STRATEGIES Phage systems readily allow sorting of enormous libraries. This rapid sorting is a result of the linkage of recognition and replication. Only those phage that bear a functional peptide or protein expressed on their surface will be permitted to propagate through a procedure known as panning (Fig. 7). The display of the antibody on the surface of the phage allows for the selection of clones by panning against antigen in ELISA wells. This is analogous to an affinity chromatography step. Generally a single well of a 96-well ELISA plate is sufficient for selection making several micrograms of protein sufficient for multiple selection steps and subsequent binding assays to verify success. The efficiency of the selection process is due in part to oversampling of the library. This is a distinct advantage of phage since 10l2phage in a volume of 50 pI may be applied to a single ELISA well. Thus for a library of a million clones, each type will be present on a million phage. Phage bind by way of the displayed antibody fragment. The approach is not limited by a requirement for purified antigen and is sensitive to rather rare components, for example, in viral lystaes, allowing for the discovery in principle of biologically active antibodies to unknown antigens. Furthermore whole cells may be utilized for selection, Antigen immobilized on beads or whole cells may also be utilized. Chromatographic procedures may also be used for phage selection; however, high levels of background binding of phage make panning in microtiter plates the preferred mode of selection. After vigorous washing, the bound phage is enriched for antigenspecific Fabs. Elution of the specifically bound phage is most generally achieved by acid elution. Alternatives to acid elution include base elution, antigen elution, direct infection by simple addition of E. coli, or reductive elution. This phage is then amplified and reselected by further rounds of panning. Each step selects for antigen-specific clones as well as for clones of the highest affinity, at least in the monovalent Fab display. In this way one can rapidly generate a panel of antigenspecific Fabs/scFvs.

E

ANTIBODIES FROM COMBINATORIAL LIBRARIES

207

C. Elute e.g. at low pH

D.

Isolated specific phage-Fab

y/

E. Reconstruct phagemid; Fab phage display+soluble expression

0

c 3

C

c 3,

Soluble Fab

FIG.7. Panning for selection of specific Fabs from the combinatorial phage display library. Antigen immobilized e.g. on a plastic surface, is exposed to the phage display library (A) when specific phage-Fabs bind and most of the rest of the phage are removed by washing (B). Specific phage-Fabs are then eluted by, e.g., low pH or excess soluble antigen (C) and isolated (D). In practice aRer one round of panning many irrelevant phage-Fabs are present at (D) and the process (A-D) is repeated as several rounds of panning. After each round of panning the eluted phage-Fabs are amplified to minimize the chances of losing important phage. Amplification is carried out by infecting bacterial cells with the phage (which are then converted from single-stranded phage to doublestranded phagemid DNA), growing the cells, and then rescuing phage-Fabs using helper phage as in Fig. 4b. Finally the phage-Fabs from (D) are converted to the phagemid form, the DNA is prepared, and the gene for cpIII is excised. Religation then gives a reconstructed phagemid which can be used to transform bacterial cells for the production of soluble Fab fragment. After Burton, 1992.

F. EXPRESSION OF ANTIBODYFRAGMENTS Selected clones must be characterized individually to assess their binding characteristics. The simplest route to soluble Fab is to rely on the proteases present in E. coli to cleave the g3p fusion protein to release Fab. Simple induction of expression overnight yields sufficient protein in the supernatant for specificity and competitive studies. For the production and purification of larger amounts of protein, the gene encoding the g3p may be removed in the pComb3 vector by restriction digest with NheI and SpeI followed by religation and transformation (Barbas et al., 1991). The removal and religation may be performed in a single step as religation destroys the restriction site. Purification

208

DENNIS R. BURTON AND CARLOS F. BARBAS 111

is performed most readily on a large number of clones by using an anti-Fab affinity column. As an alternative to removal of the g3p gene an amber mutation may be placed at the scFv or Fab/g3p junction (Clackson et al., 1991). Isolation of DNA and retransformation into a nonsuppressing strain of E . coli allow for the amber codon to be read as a stop and soluble protein to be produced. The amber codon strategy is a viable approach when cloning natural genes; however, when the libraries are constructed with random synthetic segments which encode 1part in 32 as an amber codon this approach becomes less reliable (vida infra). As noted previously, the options for purification of scFv are more limited, although affinity chromatography is an option if large amounts of the anti-peptide monoclonal antibody are available. Fabs are purified more economically with polyclonal reagents. G. WHOLEANTIBODYMOLECULES For most applications of antibodies in therapy, the whole antibody molecule is preferred. Expression of whole antibody molecules, at least in part because of the glycosylation requirement of the Fc part of the molecule, demands the use of eukaryotic cell lines. The favorites in this regard have been myeloma (Wright and Shin, 1991)or Chinese hamster ovary (CHO) cells (Bebbington, 1991) although baculovirus has also been used (Haseman and Capra, 1991). Modification of existing vectors enables the facile expression of whole antibodies utilizing Fab cassettes from the phage system. Using such an approach we have expressed a whole IgGl molecule in CHO cells using the Fd and light chains derived from a phage clone (Bender et al., 1993). In principle any Fc can be linked to the Fd of the heavy chain from the phage system. This allows the experimenter to choose a suitable isotype for a given task. The reader is referred to Burton and Woof (1992) for a more complete discussion of human antibody effector function. 111. Antibodies to Non-Self-Antigens from Immune Donors

Humans typically have serum antibody titers to a range of nonself-molecules. Clinically the most significant are those to infectious agents, to vaccines, and to allergens. There is great interest in cloning antibodies to these antigens for prophylaxis and therapy and to monitor humoral responses. In adopting a library approach, a number of choices need to be made including donor, tissue source, and antibody isotype. The choice of donor and isotype is dictated primarily by the serum titer against the antigen to be studied. A high serum titer is presumed to reflect rela-

ANTIBODIES FROM COMBINATORIAL LIBRARIES

209

tively brisk antibody production and higher levels of specific mRNA with which to begin the cloning process. In practice, we prepared a human Fab to measles virus nucleoprotein using bone marrow from a donor who had a serum titer (defined as the last serum dilution significantly above background) to measles antigens of approximately 1: 100 and had been vaccinated some 20 years previously (Bender et al., 1994). However, only a single specific Fab was generated in contrast to other instances where arrays of Fabs were derived from donors with higher serum titers as described below. Most studies have concentrated on the IgGl isotype although IgA and IgM have also been investigated. Choice of tissue source is crucial in the library approach. Ideally one begins with the source richest in plasma cells secreting antibodies against the antigen of interest since this will contain the highest levels of specific mRNA. Richest can refer to numbers relative to irrelevant specificities and to diversity of clones secreting antibody to the given antigen. In practice the two probably go together. It appears from studies on tetanus toxoid boosting of humans that, immediately following secondary antigen contact, i.e., 3-10 days, the number of specific antibody secreting cells in the peripheral blood is high (Stevens et al., 1979; Thiele et al., 1981; Ershler et al., 1982; Volkman et al., 1982). Indeed human Fabs to tetanus toxoid have been derived from PBL libraries from boosted donors (Mullinax et al., 1990; Persson et al., 1991). However, the number of antibody secreting cells :n the periphery declines very rapidly following the antigen boost to a low resting level (Stevens et al., 1979; Ershler et al., 1982). Consistent with this was our failure to detect anti-tetanus toxoid antibodies in a library prepared from an individual with a high serum titer to toxoid but no recent boost (Persson et al., 1991). Bone marrow has been shown in humans to be a major repository of antibody producing cells (Lum et al., 1990) and is the source we have used most frequently for libraries. We have shown that antibodies to many different pathogens can be derived from bone marrow libraries constructed from a single donor (Williamson et al., 1993). Other tissue sources from which we have prepared libraries are spleen, lymph node, and tonsil. Antibodies to respiratory syncytial virus (RSV) have been selected from spleen and tonsil libraries (unpublished results). A. ANTIBODIES TO VIRUSES Specific human antibodies have been shown to prevent disease caused by a wide variety of viruses belonging to diverse RNA or DNA virus families that include the orthomyxoviruses, paramyxoviruses,

2 10

DENNIS R. BURTON AND CARLOS F. BARBAS I11

alphaviruses, flaviviruses, arenaviruses, lentiviruses, picornaviruses, hepadnaviruses, and herpesviruses (reviewed in Chanock et al., 1993). Examples from clinical medicine include the hepatitis A virus, measles virus, and poliovirus. Commercial human gamma globulin, which usually contains 16 to 18%IgG, of which only a very small proportion is specific for any given antigen, is highly effective in the prevention of hepatitis A disease and has been widely used for that purpose during the past 40 years (Krugman et al., 1960).A total of 5-10 ml of commercial human gamma globulin (hereafter referred to as IgG) provides effective protection for adults for about 6 months. Prior to licensing of the live measles virus vaccine, human IgG was successfully used for passive prophylaxis and was the mainstay for prevention of the disease ( Janeway, 1945).A double-blind prospective clinical trial performed in 1951-1952 demonstrated that human IgG was also effective in preventing paralytic disease caused by poliovirus (Hammon et al.,

1953).

One of the more dramatic effects of viral antibodies was observed during a study (Beasley et al., 1983) in which human IgG was shown to be effective in preventing chronic hepatitis B virus (HBV) infection in high-risk infants. The IgG preparation, selected for having a high titer of antibodies to HBV, was administered to infants born to mothers chronically infected with HBV. In the absence of any treatment, most such infants (>go%)become chronic HBV carriers within the first few months of life. IgG given in three 0.5-ml doses at birth and 3 and 6 months was 71%effective in preventing the development of persistent HBV infection. A single dose given at birth was 41%effective. In the group of successfully treated infants, 75%were in fact initially infected with virus but they resolved their infection and developed protective antibodies. The prophylactic value of antibodies to cytomegalovirus (CMV) and varicella Zoster virus (VZV) is discussed below. Conventional wisdom, while recognizing the prophylactic role of antibodies, tends to discount their importance in the resolution of established viral infections. Instead, clearance of virus is thought to be mediated primarily by CTLs. Several recent observations cast doubt on this generalization and suggest that in some situations antibody can independently bring about resolution of infection. The most dramatic clinical therapeutic effect of viral antibodies has been observed in patients with Argentine hemorrhagic fever which is caused by Junin virus, an arenavirus (Enria et al., 1984).The disease has a high mortality but death can be prevented when a preparation of pooled human sera with a high titer of Junin virus neutralizing antibodies is administered within 8 days of the onset of symptoms.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

211

Mucosal virus infections that are limited to the cells that line the lumen of the respiratory tract can also be cleared by specific antibodies that are delivered by parenteral inoculation or by direct instillation into the lungs. High therapeutic efficiency has been demonstrated for RSV neutralizing antibodies delivered to the lungs of cotton rats at the height of infection (Prince et al., 1987; see also our studies described below). In the case of influenza A virus, viral hemagglutinin-specific antibodies can clear the lungs of infected mice in the absence of other immune functions (Scherle et al., 1992). Complete clearance of virus from persistently infected SCID mice was achieved using physiologic amounts of hemagglutinin-specific antibodies inoculated parenterally. Therapeutic efficacy of antibodies has been described for a number of neurotropic viruses. In SCID mice, persistent infection of neurons by Sindbis virus can be cleared rapidly by parenteral inoculation of Sindbis virus envelope glycoprotein-specific antibodies (Levine et al., 1991). Such neurons are not recognized by CD8' CTLs because these cells are deficient in class I MHC molecules. Surprisingly this clearance appears to occur without causing obvious cell damage. The mechanism responsible for this dramatic effect is not understood. Other studies in mice indicate that antibody directed against the a1 outer capsid protein of reovirus (the viral attachment protein) can interrupt the spread of established reovirus infection within the central nervous system (Virgin et al., 1988; Tyler et al., 1989). Monoclonal antibody against the reovirus a l , 03, or p l outer capsid proteip can also inhibit the neural spread of virus from the brain to the eye. Therapeutic effects have also been described for antibodies to rabies and measles viruses in rodent systems. IgG has been used in the successful treatment of chronic enteroviral meningoencephalitis (Mease et al., 1981). For some of the reasons outlined above and because current antiviral strategies have met with only limited success, we have put some effort into the generation of human antibodies to viruses as described below. Most of the published work to date has originated from our laboratories, although several groups are now working successfully in this area.

1 . Antibodies to Human Immunodeficiency Virus Type 1 (HIV-I) There are three principal reasons for producing human monoclonal antibodies from HIV-l-seropositive individuals. The first is to explore the humoral response to the virus and thereby increase understanding of the nature of infection and disease. The second is to aid in vaccine design. One of the goals of a vaccine is to elicit potent neutralizing antibodies and therefore study of such antibodies and the epitopes they recognize is of clear value. The third reason is to produce reagents

2 12

DENNIS R. BURTON AND CARLOS F. BARBAS I11

for passive immunization. The variability of the HIV-1 virus means that a cocktail of human monoclonal antibodies rather than a single antibody may be appropriate here. The ability of the library approach to provide large numbers of antibodies for evaluation may then be highly beneficial. Passive immunization is likely to find immediate application in the treatment of individuals accidentally infected with HIV-1, e.g., needle sticks and in prophylaxis of infected mothers to hinder transmission of virus to the fetus or newborn. Given the complexity of AIDS, the value of antibodies in therapy will remain uncertain until it is tried. However, it seems reasonable to suppose that reducing the viral load will always be a desirable goal and, even if antibodies alone do not arrest the disease, they may be valuable in conjunction with new generations of anti-retroviral drugs. All of the antibody-mediated HIV-1 neutralizing activity in human sera has been associated with reactivity to the envelope glycoproteins gp120 and gp41 and in particular to the third hypervariable domain (V3 loop) and CD4 binding site of gp120 (for reveiws see Nara et al., 1991; Moore and Nara, 1991; Laal and Zolla-Pazner, 1993; Moore and Ho, 1993; Neurath, 1993). The importance of antibodies to the V2 loop of gp120 in the neutralizing response has also been reported (McKeating et al., 1993; Moore et al., 1993).It is these epitopes which figure in the discussion below. The first phage display library in these studies was prepared from a 31-year-old homosexual male who had been HIV positive for 6 years but had no symptoms of disease (Burton et al., 1991). Serological studies showed the presence of a high ELISA titer against the HIV-1 surface glycoprotein, gp120 (LA1 strain). After securing informed consent, bone marrow cells were obtained by aspiration. Amplified antibody genes were then cloned into pComb3 to give a library of lo' members. The phage surface expression library was panned against recombinant gp120 (strain LAI) coated on ELISA wells. This choice of antigen was reasoned to encourage strain cross-reactivity because the library donor was of United States origin, where MN-like strains predominate, whereas the antigen was derived from the LA1 strain which is very rare in the United States. Four rounds of panning produced an amplification in the number of eluted phage of a factor of about 100 compared to the first round of panning, indicating enrichment for specific antigen binding clones. Forty reconstructed clones secreting soluble Fab fragments were grown up and the supernates screened in an ELISA assay for reactivity with recombinant gp120.The supernates from 33 clones showed clear reactivity. The supernates did not react with BSA-coated

213

ANTIBODIES FROM COMBINATORIAL LIBRARIES

wells and anti-tetanus toxoid Fab supernates did not react with gp120coated wells. DNA from the 33 clones was used as templates for sequencing of thymidine nucleotides of the V, and VL regions to reveal that at least 10 clones had unique heavy chains and 20 clones unique light chains. A representative number of chains were then sequenced. Figures 15 and 16 indicate the diversity of the panel of antibodies cloned. The heavy chains fell into a number of groups with evidence of somatic mutation within a given group. The light chains showed even greater diversity. Chain promiscuity was observed in the sense that very similar or identical heavy chains were paired with different light chains. The sequences are discussed in detail below. To estimate the affinities of the Fab fragments for gp120, inhibition ELISAs using soluble gp120 were performed (Fig. 8).The examination of 15 clones showed that most inhibition constants were less then lo-' M implying monomer Fab-gpl2O apparent binding constants of the order of or greater than lo8 M-'. These values have now been confirmed by surface plasmon resonance using the Pharmacia BiaCore machine (Table I). Generally we find a good correlation between apparent binding constants deduced from inhibition ELISA data and binding constants determined by surface plasmon resonance. The phage surface library was also panned against gp160 (LAI), gp 120 (SF2), and a constrained peptide having the central part of the

* 6 029 - 2 0 3 x Imp 2

"

-12

-11

-10 -9 -8 log [competing gp1201

-7

-6

FIG.8. Relative affinities of Fab fragments for gp120 (LAI)as illustrated by inhibition ELISA. Fabs 12,27,6,29,2, and 3 are all prototype members of groups of CD4 binding site Fabs discussed in the text. Loop 2 is a Fab fragment selected from the same library as the other Fabs but which recognizes the V3 loop. In this case competition was carried out with gp120 from the SF2 strain.

214

DENNIS A. BURTON AND CARLOS F. BARBAS I11

TABLE I BINDING CONSTANTS FOR RECOMBINANT Fabs TO a 1 2 0 DETERMINED BY BIA CORE

b3

b6 b12 b13

9.6 x 1.6 x 4.5 x 1.1 x

103 104 104 104

1.8 X lo4 1.6 x lo4 4.3 x 104 1.4 x 104

5.1 x 9.7 x 1.1 x 7.9 x

lo' 10'

10s 107

Note. Kinetic constants and calculated affinity constants for the binding of selected Fabs to LA1 gp120 measured by BiaCore (from Roben e t al., 1994).

MN V3 loop sequence. Fabs isolated by panning against gp160, and showing a strong ELISA signal with gp160, also cross-reacted strongly with gp120. The sequences of the Fabs obtained by panning against gp160 (LAI) or gp120 (SF2) were mostly closely related to those described above from panning against gpP20 (LAI). Indeed regardless of the panning antigen (gp120 (LAI), gp120 (SF2), or gp160 (LAI)), the Fabs selected reacted with both gp120(LAI) and gp120 (SF2). Several Fabs were obtained by panning against the constrained peptide but only one also reacted with gp120. In fact this Fab reacted with gp120 (SF2) but only weakly with gp120 (LAI). The ability of Fabs and soluble CD4 to compete for gp120 was investigated in competition ELISAs. All of the Fabs obtained b y panning against gp120/160 were found to be completed by CD4 for binding to gp120. The Fab specific for the V3 loop was not completed. Therefore the predominant Fabs isolated from this donor by selection against gp120 (LAI) appear to be strain cross-reactive and CD4 inhibited. This is consistent with the observation that more than 50% of the reactivity of the donor serum with gp120 (LAI) is inhibitable by CD4. Further, the Fabs appear to be directed to major epitopes on gp120, in that a cocktail of three of the Fabs was able to inhibit >50% of the serum reactivity with gp120 (LAI) of more than 90% of a selection of seropositive donors (R. Burioni, private communication). A major question in relation to Fabs isolated from libraries is their functionality. To this end, a panel of 20 recombinant Fab fragments reacting with the surface glycoprotein gp120 of HIV-1 was examined

215

ANTIBODIES FROM COMBINATORIAL LIBRARIES

for their ability to neutralize MN and LA1 strain of the virus (Barbas et al., 199213). Neutralization was determined as the ability of Fab supernates to inhibit infection as measured in both p24 ELISA and syncytia assays. One group of closely sequence-related Fabs was found to neutralize virus in both assays (Table 11, compare Figs. 15 and 16). The corresponding purified Fabs neutralized virus with a titer (50% neutralization) at approximately 1pg/ml. The prototype of this group, designated Fab b12, has now been shown in a number of different laboratories to be, as the Fab fragment, a more potent neutralizer than most CD4 site whole antibodies. The spectrum of neutralizing ability of a set of Fabs to the CD4 binding site of gp120 (Fig. 9) provides an opportunity to explore the TABLE I1 NEUTRALIZATION OF HIV-1 BY RECOMBINANT Fab SUPERNATANTS

Clone

Fab concn. (pg/ml)

gp120 ELISA titer

1 2 3 4 5 6 7 8 11 12 13 18 20 21 22 24 27 29 31 35

1.8 3.1 4.1 25.0 2.4 4.0 9.6 14.0 11.0 6.0 6.1 0.9 6.9 8.5 8.6 0.7 10.0 16.0 9.3 8.9

1:8 1:64 1:32 1:16 1:128 1:64 1:256 1:256 1:128 1:64 1:128 1:128 1:256 1:32 1:64 1:32 1:64 1:1024 1:128 1:64

p24assay

MN

IIIB

40 -

-

80 -

-

-

20 20

Syncytia assay MN

40 80 20

>128

20 20

20 20

-

20

20

-

-

-

-

-

10,708

1328 184

-

80 80

-

IIIB

-

>32

-

MN

>128

20 20

-

Plaque assay

-

-

-

32 32

-

32

Note: From Barbas et al., 1992b. In this experiment aliquots of the same supernate preparations were used in p24 and syncytia assays. Figures indicate neutralization titers. For the p24 assay the titer corresponds to the greatest dilution producing >SO% reduction in absorbance in ELISA. For the syncytia assay, Fabs 4 and 12 produced >95% neutralization at a 1:4 dilution of supernate and 80 and 70% reduction at 1:128 dilution respectively.-indicates no neutralization at 1:20 dilution in the p24 assay and 1:16 in the syncytialassay (with most clones showing no detectable neutralization at a 1:4 dilution). The plaque assay data were kindly provided by Michael Hendry.

216

DENNIS R. BURTON AND CARLOS F. BARBAS 111

U

b3 b6

-M- b l l

-& b12 U b13

T

60

"

0.1

0.2

FIG.9. Neutralization of HIV-1 by a panel of recombinant Fabs using an envelope glycoprotein complementation assay (from Roben et al., 1994). The ability of recombinant Fabs to neutralize the HXBc2 molecular clone of the HTLV-IIIB (LAI)isolate was assessed in an envelope complementation assay (Helseth et al., 1991). Briefly, COS-1 cells were cotransfected with a plasmid expressing envelope glycoproteins and a plasmid containing an env-defective HIV-1 virus encoding the bacterial chloramphenicol acetyltransferase (CAT) gene. Equal fractions of the cell supernates containing recombinant virions were incubated at 37" C for 1 hr with varying concentrations of Fab prior to incubation with Jurkat cells. Three days postinfection; Jurkat cells were lysed and CAT activity measured.

mechanism(s) of virus neutralization. Clearly neutralization does not require either cross-linking of virion particles or cross-linking of gp120 molecules on the surface of the virion since the Fab is monovalent. Neutralization is not correlated either with Fab affinity for recombinant gp120 or with the ability of Fabs and soluble CD4 to compete for binding to recombinant gp120.However, there is evidence from radioimmunoprecipitation studies that the most potent Fab recognizes a native presentation of gp120 on the surface of infected cells more effectively than less potent Fabs (Roben et al., 1994). This finding may have important consequences for vaccine design. Fab b12 has been linked to Fc to generate a whole IgGl molecule using methods outlined above. IgG b12 is an outstanding neutralizer, clearly 2-3 orders more effective than other CD4 site antibodies sub-

ANTIBODIES FROM COMBINATORIAL LIBRARIES

217

mitted to a recent NIH workshop (D’Souza et al., 1994). Furthermore the antibody has been shown to be capable of neutralizing a range of primary isolates of HIV-1 (D. D. Ho, unpublished observations; R. M. Hendry, unpublished observations). Neutralizing antibodies to other epitopes on the HIV-1 envelope have been generated. These are generally broadly strain cross-reactive because of the design of the selection procedure. Antibodies to the tip of the V3 loop have been derived by panning libraries against a constrained peptide having a consensus sequence (GPGRAF) corresponding to this region. Neutralizing antibodies to an epitope dependent on the V2 loop of gp120 have been derived by a novel epitope masking procedure (Ditzel et al., manuscript submitted). Antibodies to the CD4bs predominate in sera from HIV-1-infected humans (Moore and Ho, 1993)and antibodies to the V2 loop are relatively rare (Moore et al., 1993). Therefore, panning was carried out after the CD4bs epitope(s) was masked with two CD4bs-directed Fabs. Three novel Fabs were derived, which did not compete with the CD4 bs antibodies and were sensitive to mutations in the V2 loop. This demonstrates the feasibility of selecting minor specificities from libraries by masking major specificities before panning. Neutralizing antibodies to gp41 have been derived by panning a library prepared from a gp160immunized donor with a peptide corresponding to a known neutralizing epitope (M. A. A. Persson, personal communication). 2 . Antibodies to Respiratory Syncytial Virus RSV is the major pediatric viral respiratory tract pathogen outranking all others as a cause of pneumonia and bronchiolitis in infants under 1 year of age (reviewed in McIntosh and Chanock, 1990). It has been concluded that RSV infection in bone marrow transplant patients is a serious and life threatening infection with a high mortality rate once pneumonia develops (Harrington et al., 1992). Several lines of evidence indicate that antibodies mediate resistance to RSV infection and illness including clinical studies of pooled human IgG containing a high titer of RSV neutralizing antibodies which provided indications that these antibodies can exert a therapeutic effect in serious RSV infection in infants and young children (Hemming et al., 1987) and a prophylactic effect in high-risk infants and children (Groothuis et al., 1993). Given this evidence, there is considerable interest in having available human monoclonal neutralizing antibodies to RSV for prophylaxis in protecting infants at high risk of serious disease and for therapy in cases of serious RSV lower respiratory tract disease. Since ELISA analysis of the serum of the donor used above also

218

DENNIS R. BURTON AND CARLOS F. BARBAS I11

indicated a high titer to RSV surface FG glycoprotein, it was decided to pan the same library against FG. Two classes of Fab fragments interacting with the F glycoprotein were identified (Barbas et al., 1992~).The predominant Fab neutralized the virus with a titer of approximately 1 pg/ml, The less abundant Fab (about 5% ofthe clones) neutralized both A2 and B strains of the virus highly effectively with a titer around 0.1 pg/ml. Furthermore, this Fab neutralized 19 field isolates of virus of various geographical and temporal origins with titers of approximately 0.1-1.0 pg/ml (Table 111). Most recently, we have demonstrated a marked therapeutic efficacy for this Fab i n vivo (Crowe et al., 1994).A single dose of 12.9 p g of Fab (RSV 19) administered to mice at the height of RSV infection is able to reduce the viral titer in the lungs of the mice by a factor of 5000 (Table IV). A nonneutralizing anti-F Fab (RSV 126)was ineffective. The therapeutic effect of RSV19 is not fully sustained and there is some rebound in pulmonary virus titer a day after Fab administration. However, the rebound could be prevented by two further doses of RSV19 (Fig. 10). The results are very promising for the development of an aerosol of specific recombinant Fabs for the treatment of RSV disease and possibly for virus-mediated respiratory disease generally. TABLE I11 NEUTFIALIZING ACTIVITYOF Fab RSV 19 AGAINST DIVERSE RSV ISOLATES BELONGING TO ANTIGENIC SUBGROUP A OR B

RSV Isolates Tested Antigenic subgroup

No. isolates

Temporal distribution

A

10

1959-1984

B

9

1962-1990

Specific neutralizing activity of Fab RSV 19 (conc. of Fab(Nglm1) needed for 60%plaque reduction)

0.3, 0.3, 0.4, 0.7, 1.0, 1.1, 1.2, 1.2, 1.7, 3.0" <0.2, 0.3, 0.4, 0.4, 0.4, 0.4, 0.5. 0.6, 0.8'

Note. Neutralizing activity of RSV Fab 19 against diverse RSV isolates belonging to antigenic subgroup A or B. A purified preparation of Fab was used (from Barbas et al., 1892~). Washington/Eiern/65, St. LouisllO865/84, AustralialAZ/GI, S t Louis/ 863/84, Washington/343/67, Australia/A1/61, Washington/11657/60, St. Louis/10849/84, Washington/3LQ9/66,Sweden/669/59, respectively. West Virginia (WV)114617/85, WV/17154/85, WV/4843/81, WV/ 20323187, WV/401R/90, Washington/18537/62, WV/474/r/90, WV1285Rl 90,WV/1293/79, respectively. (West Virginia strains kindly provided by Maurice A. Mufson, MD.)

TABLE IV THERAPEUTIC EFFECTOF Fab RSVl9 ADMINISTEREDINTRANASALLY TO MICEAT 3 DAYSPOSTINFECTION

Fab RSV19

RSV126

HIV DL21 None

Dose (&

Virus titer (log,,pfu/g tissue) (six animals)

13 6.5 3.2 1.6 0.8 14 6.9 15 0

Note. From Crowe et QZ. (1994).

2.4 f 0.3 4.2 f 0.5 4.8 f 0.2 5.5 f 0.1 6.0 2 0.1 5.6 f 0.1 5.9 f 0.1 5.9 f 0.05 6.1 f 0.1

*----* -A A-

Untreated fab19 on Day 4 Fabl9onDays4,5and6

6:

1 .

5r

:5

4;

14

3'

13

2

1

i

:2

i

tl

FIG.10. Rebound of virus replication following administration of RSV Fab 19 on Day 4 postinfection is reduced by additional daily instillation of the Fab (from Crowe et al., 1994). BALB/c mice were infected intranasally with 106.3pfu of wild-type RSV on day 0, then were treated intranasally with 50 pg RSV Fab 19 on either Day 4 alone (a), Days 4,5, and 6 (A), or were not treated (m). The titers of virus in the lungs of RSVinfected mice treated on days 4, 5, and 6 with a control Fab (HIV-12)did not differ from that of untreated mice (data not shown). Four animals were sacrificed at each time indicated in the figure and lung tissue was obtained for determination of virus titer on the indicated day.

220

DENNIS R. BURTON AND CARLOS F. BARBAS I11

3. Antibodies to Hepatitis B Virus Hepatitis B affects more than 200 million people worldwide (Hollinger, 1990). A dramatic example of the protective effect of antiviral antibodies is described above although any therapeutic effect in humans remains largely unexplored. In a library study (Zebedee et al., 1992), two individuals were vaccinated with recombinant hepatitis B surface antigen and PBLs extracted 7-14 days later (although 4-5 days is probably optimal). Libraries (IgGlK and IgGlA) were constructed on the surface of phage and panned against the hepatitis B surface antigen. Specific Fabs were identified and in this case the binding of phage expressing specific Fabs to virus antigen particles could be clearly visualized in the electron microscope (Fig. 3a). Sequencing of positive clones showed a limited diversity with a remarkable example of promiscuity in that a given heavy chain was shown to bind antigen with either a K or A light chain. Further, the light-chain partner appeared to affect specificity as measured by the ability of excess Fab to compete with mouse monoclonal antibodies for virus antigen. 4 . Antibodies to Human C ytomegalovirus (HCMV) CMV disease may affect virtually every organ system and is frequently the causative agent in retinitis and gastroenteritis in AIDS patients. Essentially all HIV-1-positive patients are at risk from both reactivation of endogenous infection as well as reinfection with a second strain. Similarly, immunosuppressed patients who have not been infected previously with CMV are at extremely high risk of developing severe life threatening disease when they receive an organ transplant from an infected donor. Studies by Snydman et al., (1987,1990) show that administration of human IgG with high anti-CMV titer prevents approximately one-half of these serious CMV illnesses, but does not reduce the incidence of infection. As CMV is thought to be strongly cell associated during infection, control of infection is usually ascribed to class I MHC-restricted CD8' CTLs. As a consequence, prevention of disease in the high-risk group by IgG is of some interest because it suggests that antibodies can also suppress infection by their action on CMV-infected cells. Two major complement-independent neutralization antigenic targets have been identified on the viral envelope of CMV, i.e., glycoproteins gB and gH. gH, and more particularly gB in some individuals, has been determined to be the most immuno- and neutralization dominant in studies with convalescent human sera (Marshall et al., 1992; Britt et aE., 1990). These molecules have been implicated in viral attachment to host cells and in membrane fusion and penetration. A

ANTIBODIES FROM COMBINATORIAL LIBRARIES

221

number of groups have described gB- and gH-specific mouse and human monoclonal antibodies capable of neutralizing virus (Utz et al., 1989; Urban et al., 1992; Meyer et al., 1990; Ohizumi et al., 1992; Aulitzky et al., 1991; Ohlin et al., 1993). We have constructed two libraries from HIV-1-seropositive individuals with significant antibody titers against HCMV and panned them against lysates of HCMV-infected cells (Williamson et al., 1993). Of 25 clones examined from a single library, 6 different heavy- and lightchain sequence pairings were observed (Williamson et al., unpublished observations). These clones recognized infected cells and were variously studied in immunoprecipitation and Western blotting assays. One clone immunoprecipitated what we believe to be a novel banding pattern. The remaining clones were found to recognize either the 65kDa lower matrix phosphoprotein or the 52-kDa nuclear DNA binding phosphoprotein. As both of these antigens have been shown to be strongly immunogenic in humans and are present in large quantities in viral lysates, it is not surprising that antibodies with these specificities should be recovered. Indeed these antibodies were able to strongly compete with donor serum for binding to viral lysate, confirming their high serum levels in this individual and the ability of library technology to reflect the donor’s antibody response.

5. Antibodies to Varicella Zoster Virus VZV is recognized as one of the most frequent viral opportunistic infections in HIV-infected individuals, classically presenting as reactivated multidermatomal zoster, with or without dissemination, although a recently recognized chronically progressive syndrome is increasingly being identified. Immune prophylaxis with large doses of human globulin is currently used in the clinic for the prevention of VZV infection in immunocompromised patients. Specifically, the administration of zoster immune globulin within 4 days after exposure to the virus will prevent or reduce the severity of subsequent disease (Zaia et al., 1983; Balfour, 1988). Moreover, Larkin et al. (1985) determined that disseminated herpes zoster is associated with significantly low levels of antibody to VZV glycoproteins, implicating antibodies recognizing particular viral determinants in the containment of virus postinfection. The envelope glycoproteins I through IV have been shown to be powerful immunogens and have been shown to elicit antibodies capable of neutralizing virus in vitro (Forghani et al., 1984; Grose et al., 1983; Keller et al., 1984; Vafai et al., 1984). Libraries from two HIV-1-positive individuals with significant anti-

222

DENNIS R. BURTON AND CARLOS F. BARBAS 111

body titers against VZV have been independently panned against VZVinfected cell lysate (Williamson et al., 1993). Amino acid sequences of the clones revealed that panning had isolated only four different sequences from the two libraries. None of these clones demonstrated any significant neutralization of virus in inhibition of plaque development assays. 6. Antibodies to Herpes Simplex Viruses Types 1 and 2 Herpes simplex virus (HSV) remains one of the most common viral maladies in man, achieving a worldwide distribution and causing a variety of infections (reviewed in Whitley, 1990). Two forms of the virus, HSV-1 and HSV-2, have been distinguished by clinical manifestations and biochemical and serological criteria. HSV-2 is more commonly implicated in genital infection, while HSV-1 is associated with oral and ocular disease. Both types of the virus may become latent following travel intraaxonally to sensory ganglia. Primary and secondary herpes infections in those immunocompromised by underlying disease or immunosuppressive drugs are often more severe than in the healthy host. Such individuals, which include AIDS patients, those with hematological or lymphoreticular neoplasms, and organ and bone marrow transplant recipients, are also prone to increased frequency of secondary herpes episodes. In the transplant recipients, the severity of herpes infection correlates with the degree of immunosuppressive therapy employed. Devastating illness may also result from HSV infection of the neonate. In the United States such infections are encountered in 1 in 2500 to 1 in 5000 deliveries per year, most being acquired following intrapartum contact with infected genital secretions. As compared to a recurrent episode, primary infection occurring late in pregnancy generally produces more frequent and more severe disease in the newborn. This correlates with a greater maternal viral load at delivery (Corey et al., 1983), probably arising because the immune response to the virus is only in its early stages. Current therapy of herpes infections is limited, although acyclovir, vidarabine, and related drugs have proven useful for the management of specific infections such as mucocutaneous herpes infections in the immunocompromised host, herpes simplex encephalitis, and neonatal herpes. Recurrent episodes, however, are less responsive. Moreover, viral strains resistant to these drugs have been isolated from immunocompromised patients. There is evidence of a significant protective role for antibody in human infection in uiuo. The presence of neutralizing antibody in

ANTIBODIES FROM COMBINATORIAL LIBRARIES

223

acute phase serum during primary infections has been associated with reduced severity and duration of the primary genital herpes episode (Corey, 1982). Further, it has been shown that the development of recurrent genital herpes in an individual following a primary infection with homologous virus is inversely correlated with the titer of HSV2 neutralizing antibody in the convalescent serum (Reeves et al., 1981). Moreover, the titer of anti-HSV antibodies in bone marrow transplant recipients is predictive of the risk of infection (Pass et aZ., 1979). In uitro, human serum antibody has additionally been shown to neutralize extracellular virus and lyse certain HSV-infected cells. HSV is a complex virus: over 50 virus-encoded polypeptides including both structural (envelope and core) and regulatory proteins have been identified in infected cells. Analysis of the humoral response against HSV and the identification and characterization of potentially protective antigens have been undertaken largely using human sera and mouse monoclonal antibodies. The main targets of the humoral and cellular responses appear to be the seven well-characterized HSV envelope glycoproteins, gB, gC, gD, gE, gG, gH, and gI, which are found both on the virion and on the infected cell surface where they are thought to act to promote viral attachment and penetration through multiple interactions between themselves and the cell membrane. Only gB, gD, and gH have been found to be indispensable for viral growth in cell culture. Virus mutants defective in these molecules will bind to the host cell surface but are unable to penetrate into the cytoplasm. Glycoproteins mandatory for virus attachment have not yet been identified. Each envelope glycoprotein is capable of eliciting mouse monoclonal antibodies able to neutralize virus in vitro. Passive immunization with monoclonal antibodies specific for gB, gC, gD, gE, and gH has also been shown to protect animals from infection (Kumel et al., 1985; Simmons and Nash, 1985; Balachandran et aZ., 1982; Dix et al., 1981). In addition, polyclonal immune sera and gD-specific monoclonal antibodies have been shown to protect mice from recurrent disease (Simmons and Nash, 1985). It has also been demonstrated that monoclonal antibodies against gB and gE suppress the replication of HSV-1 in trigeminal ganglia (Oakes and Lausch, 1984).Glycoproteins B and D appear to elicit a major part of the antibody response to virus in humans and also appear to be the major targets of neutralizing antibodies (Kuhn et al., 1987). A combinatorial library constructed from an HIV-l-seropositive individual with serum antibody titers against HSV-1 and -2 was independently probed using lysates of cells infected with each of these viruses

224

DENNIS R. BURTON AND CARLOS F. BARBAS 111

(Burioni et al., 1994). The antibodies were sequenced and characterized by their ability to react in ELISA with the antigen against which they were panned and in immunofluorescence studies with infected cells. A diverse panel of Fab clones was derived from the two pannings, with 9 of 18 heavy-chain sequences obtained from the HSV-2 panning being largely unrelated. Five of eight heavy chains taken from the HSV-1 panning were different from each other. However, although each lysate selected two diverse and largely distinct (only one heavychain sequence was common to both pannings) sets of clones, all of the antibodies examined were found to bind to both HSV-1 and -2. The generation of these different yet cross-reactive sets of clones using two similar virions probably arises from differences in affinity for the two virions, reflecting the sensitivity of the panning procedure to antibody affinity. Of 17 different clones tested for their ability to’neutralize HSV-1 and -2 in plaque reduction and inhibition of plaque development assays only one was able to significantly interfere with virus infectivity when tested as a crude bacterial supernatant. Accordingly this antibody, HSV Fab8, has been thoroughly characterized. We have determined the antibody binds specifically to the envelope glycoprotein D and is able to neutralize HSV-2 (50% plaque reduction at 0.05 pg/ml and 80% inhibition at 0.1 pg/ml) somewhat more efficiently than HSV1 (50% inhibition at 0.25 pg/ml and 80% inhibition at 0.6 pg/ml) (Fig. 11). Figures comparable to these have been obtained in the neutralization of several clinical isolates of both HSV-1 and -2. HSV Fab8 has further been shown to completely abolish HSV-2 plaque development in virus-infected monolayers at a concentration of 25 Fg/ml, illustrating that Fab efficiently prevents cell-to-cell transmission of the virus (Fig. 12). In addition this antibody strongly reduced infectivity after attachment of the virus suggesting its inhibitory action takes place either at the level of membrane fusion or during virus penetration or uncoating and not by blocking attachment of virus to the target cell. Consistent with this supposition, we have demonstrated that neutralization obeys single-hit kinetics. The data indicate that Fab8 is approximately 1order of magnitude more potent than most murine whole antibodies described so far, and further underlines the potential of Fab in the containment of virus infectivity. It is, however, significant that all but two of the antibodies isolated above by probing of libraries using HSV-1 and -2, HCMV and VZV viral lysates were found to be nonneutralizing. In using whole viral lysate as a selecting agent it may be expected that the breadth of the humoral response to the pathogen will be accessed, of which high-

ANTIBODIES FROM COMBINATORIAL LIBRARIES

225

B

.001

.01

.1

1

10

100

kglml FIG.11. Neutralizing activity of HSV Fab8, as measured by plaque reduction (from Burioni et al., 1994), against (A) HSV-1 and (B) HSV-2. Purified Fab8 neutralized HSV-1 with a 50% inhibition at about 0.25 pg/ml and with an 80% inhibition at 0.6 Wglml, while HSV-2 was neutralized with a 50%inhibition at about 0.05 pg/ml and an 80%inhibition at about 0.1 pg/ml.

affinity neutralizing antibodies may typically be only a small part. In response to our experiences with these and other herpesviruses we have successfully developed a simple “antigen-capture” technique to overcome this problem by more accurately selecting Fab specifically binding to antigens known to be important targets for neutralizing antibodies, even when a recombinant form of the molecule is unavailable (Fig. 13).Here, antibody specific for nonneutralizing epitopes of a chosen antigen is attached to a solid surface and used to “capture” (typically from a viral lysate) that antigen. Unbound material is then washed away and the panning procedure is performed over the captured antigen. This method has enabled us to more accurately direct the panning procedure toward isolation of antibodies against a number of virus-neutralizing epitopes, including glycoproteins D and B of HSV (Sanna et al., unpublished data). In the case of gD, ELISA wells

D

E HSVZ control

1 pg/rnl

5pg/ml

25pg/rnl

uninfected control

FIG. 12. Inhibition of HSV plaque development by HSV Fab8 (from Burioni et al., 1994).Purified Fab8 inhibited the development ofplaques when applied 4 hr postinfection (hpi) on monolayers infected with HSV-1 (A, B ) or HSV-2 (C, D). (A) A statistically significant reduction in plaque size was observed at concentrations of 5 and 1 pg/ml (* = P< 0.01), with an approximate 50% reduction in plaque size at 5 pglrnl. The number of plaques was also dramatically reduced at Fab concentrations of 5 and 25 pg/ml (B, D). At 25 pglml and 72 hr hpi plaque development in HSV-2-infected monolayers was completely inhibited (C, D). (E) Inhibition of plaque development assay with HSV-2-infected monolayers at a number of different Fab concentrations 86 hpi.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

227

Protein C

non-neutralizing epitope

epitope

Mixture e.g. viral lysate Capture with mouse or rFab anti-D (NNE)

Pan human library Human anti-D (NE)

FIG.13. Directed antigen capture for library panning. In order to focus selection to a neutralizing epitope on a given protein, the protein is captured by a mouse antibody or recombinant Fab to a nonneutralizing epitope prior to panning. The method has the advantage that it does not require recombinant or purified antigen.

were independently coated with recombinant Fab8 and a nonneutralizing gD-specific mouse monoclonal antibody. HSV-2-infected cell lysate was then overlayed into the wells and the library panned. When gD was captured from infected cell lysate using the nonneutralizing murine antibody, all 10 clones examined were of the same sequence as Fab8. However, when antigen was captured by Fab8, the antibodies derived possessed a different sequence and did not neutraIize the virus. This example clearly shows the ability to fine tune antibody selection against chosen regions of a given antigen by capturing and masking epitopes using existing antibodies.

7. Antibodies to Measles Virus Currently there is a renewed interest in the characterization of measles virus immunobiology. There are two principal reasons for this.

228

DENNIS R. BURTON AND CARLOS F. BARBAS I11

The first is the incomplete success of some vaccination programs in industrialized countries. To a major extent, the problems in this situation are of a logistic nature. However, there is a need to understand vaccine failures and to interpret the effect of revaccinations (a twodose scheme of immunization is used in some countries). The second reason is the problem of using live vaccines for vaccination at the age of 3-6 months when infants can be at risk of infection. There is a need to define the possibility of generating an attenuated variant of the virus that displays an improved capacity to replicate in the presence of a maternal immune response. These problems would clearly benefit from enhanced understanding of human antibody responses to virus and vaccine. We have carried out a single study on a 22-year-old donor who was vaccinated against measles in his childhood (Bender et al., 1994). An IgGlK bone marrow library was generated and panned against a measles antigen preparation. A single high-affinity Fab was recovered which was subsequently shown to be specific for measles virus nucleoprotein. Interestingly, two other Fabs were recovered which showed lower affinity for the measles antigen preparation and polyreactivity. The heavy chains of both of these Fabs belonged to the VH6 family which has been associated with polyreactivity.

B. ANTIBODIESTO BACTERIA The only reports of which we are aware describe the generation of human Fabs to tetanus toxoid (TT) from libraries using peripheral blood lymphocytes of recently boosted donors. In one study (Mullinax et al., 1990),the frequency of TT binding clones in a A phage library was 1:500and the apparent antigen binding affinities were in the range 107-109 A4-I. In our study (Persson et al., 1991), a A phage library prepared from a donor with a serum titer of 1:14,000seven days after a boost showed antigen-specific clones at a frequency of 1:SOOO. Apparent affinities for antigen were in the range of 107-109M-' and sequence diversity was observed in that six of eight heavy chains from binders were clearly clonally unrelated.

C. ANTIBODIESTO ALLERGENS One report has described the PCR amplification of human IgE heavy chains (Walker et al., 1992).It was not found possible to obtain amplification from PBL mRNA directly. Rather it was found necessary to first enrich for IgE-positive B cells by a sorting procedure using an antiIgE reagent. We are not aware of any specific IgE antibodies prepared by the library approach at this time.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

229

D. ANTIBODIESTO UNCONVENTIONAL ANTIGENS The binding sites on antibodies for certain antigens occur with a single type of VH gene segment with little conservation of other elements in the V regions. For one of these, staphylococcal protein A (SPA), binding clearly occurs at a site distinct from the conventional antigen binding pocket since SpA binding is not inhibited by binding of the appropriate conventional antigen. Molecules of this type have been dubbed unconventional antigens by Silverman (1992,1994).HIV1 gp120 is another molecule which may fall into this category (Berberian et al., 1993). The interaction of SpA and Fab has been explored using the library approach (Sasano et al., 1993).An IgGKh library was constructed from the peripheral blood lymphocytes of a healthy donor with no history of staphylococcal infection. About 17%of antibody producing clones in the unselected library were found to bind SPA. SPA binding was completely restricted to Fab with vH3 heavy chains and about 60% of the V,3 Fab in the unselected library had SPA binding capacity. Analysis of 21 V, and 6 V, sequences showed that the SpA binders used diverse vH3 genes and L chains deriving from a variety of VK and Vh gene families. Chain shuffling experiments showed that capacity to bind SpA was dictated by the heavy chain but different light chain usage could result in up to a fourfold change in affinity. The apparent affinities of the different Fabs for SpA varied from 4 x lo6 M-' to
230

DENNIS R. BURTON AND CARLOS F. BARBAS I11

of antigens (James and Bell, 1987). Such lines can be unstable or low secretors of antibody. Alternatively, the antibody may not be of the isotype desired. In any of these cases, it may be appropriate to rescue heavy- and light-chain genes from mRNA from the cell lines and express the Fab in bacteria. We have described this process for the cloning and expression of a human anti-rhesus D antibody from an EBV-transformed cell line using a A phage system (Williamson et aZ., 1991). In this case five of five clones examined had the correct heavy and light chains. This may not be the case for all hybridomas which can contain mRNA from other chains which are PCR amplified and cloned. For instance, for one mouse hybridoma only 1in 1000recombinants had antigen binding activity (L. Sastry, private communication). Therefore, the most prudent general strategy may be to clone from the cell line into a phage display vector such as pComb3 and then pan against antigen to select positive clones. More recently, plasmids derived from the A phage vectors were used to rescue human Fabs to rabies virus (Cheung et aZ., 1992)and to HIV1 gp120 and gp41 (Takeda et al., 1994) from hybridoma lines. The recombinant Fab to rabies was able to neutralize the virus about as effectively as the whole hybridoma-derived antibody. The anti-D cell line described above has also been rescued using the pComb3 phage display system (Siege1and Silberstein, 1994).In this case, it was shown that the anti-D specificity could be retrieved from a library comprising anti-D Fab expressing phage and anti-tetanus toxoid expressing phage at a ratio of 1:104by affinity selection using intact Rh(D)-positive red blood cells.

F. RESCUE OF HUMAN ANTIBODIESFROM HuSCID MICE SCID mice populated with human cells offer the possibility of antigen boosting of human responses outside the human body. We have shown that SCID mice can be used in conjunction with the combinatorial library approach to give human monoclonal Fabs (Duchosal et al., 1992). PBLs taken from donors who had not been boosted with tetanus toxoid for very long periods (15-20 years) could be boosted in vitro, introduced into SCID mice, and reboosted, and strong secondary responses could be observed in the mice (Fig. 14). The spleen from such a mouse was used as a source of RNA from which libraries were constructed in A phage and on M13 phages. Antigen-specific highaffinity human Fab fragments were readily derived from these libraries. For seropositive donors this sort of approach might be useful in stimulating antibodies essentially relegated to the memory compartment. One also has the advantage of being able to boost with smaller

ANTIBODIES FROM COMBINATORIAL LIBRARIES

231

0

r

d

Q

0

Serum Dilution FIG.14. Generation of secondary antibody responses in hu-PBL-SCID mice. ELISA comparison of tetanus toxoid (TT) serum titer for a donor not boosted for 20 years (A) with that (+SEM) obtained by boosting the donor’s PBLs in six Hu-PBL-SCID mice ( 0 ) . The mean values for three similarly populated but nonboosted mice are shown (0).

units than the original immunogen so for instance a response to a particular peptide could be stimulated and human monoclonal Fab fragments rescued. In the future it may also become possible to create secondary responses from seronegative donors.

G. ANTIBODIESFROM PRIMATES The constraints applied to immunization of primates are less than those applied to humans. However, antibody sequences are very similar suggesting that immunized primates may be an attractive source for the library approach. Chimpanzees are especially close to humans; differences between the two species in the constant domains are of the order of the differences found between human allotypes (Ehrlich et al., 1990,1992).We have indeed used PCR primers designed for amplification of human sequences to prepare IgGlK display libraries from HIV-1-infected chimpanzees (M. Zeidel et al., unpublished). The libraries have been panned against gp120 to yield antigen-specific high-affinity chimp Fabs. There are greater sequence differences between macaques and humans necessitating some modification to the

232

DENNIS R. BURTON AND CARLOS F. BARBAS I11

human PCR primers (M. A. A. Persson et al., private communication). Libraries have been constructed from simian immunodeficiency virus (S1V)-infected macaques and high-affinity Fabs specific for envelope prepared (Samuelsson et al., 1994).

H. THEUSEOF THE LIBRARY APPROACH TO STUDY ANTIBODYRESPONSES A major issue in library-derived antibodies from immune donors is their relevance to the in vivo response. Two studies have attempted to directly evaluate the relation of cloned to natural response, both in the mouse. The results led the authors to different conclusions. In the first study (Caton and Koprowski, 1990), Fab fragments-specific for influenza virus hemagglutinin were isolated from a A phage library prepared from an immunized donor mouse. Sequence analysis indicated that the VH regions were derived from members of an expanded hemagglutinin-specific B-cell clone, in conjunction with one of two V, regions. The most frequently identified V,/V, combination was very similar to aVH/V, combination that had been identified previously among hemagglutinin-specific hybridoma antibodies. The authors concluded that the antibodies isolated bore a close relationship to the immune status of the donor mouse. In the second study, Gherardi and Milstein (1992)compared antibodies to 2-phenyloxazolone (phOx) obtained by hybridomas with those obtained from a single-chain Fv phage display library using the same pool of spleen cells. The library approach yielded binders arising from eight individual V, genes and seven V, genes. The V, and V, genes which dominate the response as assessed by the hybridoma approach were found in only two cases in the library approach and then not in combination. The authors concluded that the antibodies isolated from the library were a poor guide to the natural response. One consideration here is that different approaches are sensitive to different parameters and so will report on a different facet of an antibody response. The library approach begins with mRNA and is therefore probably reflecting activated and differentiated B cells, i.e., plasma cell populations. In contrast, hybridomas are thought to reflect activated but not terminally differentiated B-cell populations and EBV transformation to reflect resting B-cell populations. Therefore even experiments carried out on the same human tissue using the library approach and, for example, EBV transformation will not reveal how faithfully the former is reporting on the natural response. A possible alternative to at least maintain H-L pairings is to carry out the PCR ofthe library approach in situ, “in cell-PCR,” as described for a mixture

ANTIBODIES FROM COMBINATORIAL LIBRARIES

233

of two hybridomas (Embleton et al., 1992). In this study, the authors fixed and permeabilized cells from the two hybridomas, reverse transcribed mRNA to cDNA, and amplified by PCR. After cloning, it could be shown that the original hybridoma VH-VL combinations had been maintained. At this time, we are not aware of an extension of this approach to a diverse mixture of B cells. It is instructive in relation to the authenticity of cloned antibodies from libraries to consider some of the properties of these antibodies. In principle, heavy-light chain combinations responsible for binding an antigen could arise fortuitously during library construction, i.e., neither chain is involved in binding antigen in uiuo but the (random) combinations bind antigen in uitro. The difficulties of obtaining highaffinity antibodies from nonimmune compared to immune libraries of comparable size indicate this is unlikely to be a common occurrence. Furthermore, where it has been examined (Barbas et al., 1993b; Williamson et d., 1993; Portolano et al., 1992; Hexham et d., 1994), a single or a small number of cloned antibodies have been able to inhibit antigen binding of a large proportion of serum antibodies from the donor. This implies that broad features of the in uiuo response are reproduced in the cloned repertoires in these instances. From detailed analyses of cloned Fabs from a single library, we concluded that the cloned heavy chains probably corresponded to those used in uiuo, whereas the status of the light chain was unclear (Barbas et al., 199313).Figures 15and 16 show the V, and V, sequences of 33 Fabs derived from panning the library against recombinant gp120 from LA1 or SF2 strains of HIV-1 or gp160 from LA1 HIV-1. All of the Fabs were shown to be directed to the CD4 binding site of gp120 and most had apparent affinities of the order of lo8 M-'. The analysis immediately established that similar or identical antibodies were derived from panning against the different presentations of gp120. The heavy chains could be organized into seven groups based on sequence similarities. For some of these groups, e.g., the b8 group, differences between members of the group were so small as to be attributable to PCR or reverse transcription artifacts, i.e., the clones could have arisen from a single B cell in uiuo. For others, e.g., the b3 group, the differences were more compatible with the cloning of somatic variants arising in uivo. For one group, the b l group, the CDR3 region was conserved against a background of considerable variation in the VH gene segment. It is unclear whether the pattern in this group has arisen from somatic mutation, convergent evolution, or multiple cross-overs (in uiuo or in uitro during PCR). The light chains in the antigenselected Fabs also contained some groups of related sequences. In

' to

Clone

FRI

FR2

CDRl

bl _. L E E S C T E F K P P C S S i l W ~ DYNNYAIS

CDR2

WVRQAPGQILEYIG GIWTSGSADYAQKFQG .w. T.T.S.....

830

....... .....AAWK .....R...Q...... D SF.... .v. .......... ......... .............. ................. ........................... .....A.V.X ..............1.S .F.. ..... ............M. ................. .... .v. ...........wH . ....... .....AAWK .....R ...Q . . . . . .D w.... . H . . . ............M. .......T.T.S..... T ........A ..Q ..A.V.K..........T..... 5

b3

LF.ESG(jRLVIBUjSZRLSCEOYjFTFS

b2 b14 b24 82

N A W

.....

57

..P...G.................... ......c .................... .................... A...... ......c... ..........s...... ........................... ......G ............A.......

M

L F i a m m ~ s c O A S G Y R F SNFVIH

M

820

s2 s3

ss

bl

bl 2 b2 1 t% b20 s6

b8 b13 b18 b22 b27 826

BB

035

s4

..... .s... ..... ..... .....

.............. .............. .............. .............. .............. ..............

W V R P S P G K G ~ SIKSKFDGGSPHYMPVEG

................... ................... ... R ............... ................... ................... ..........S..PG....

..... ..... ..... LF.EXiCCLVWa;sLRL-StFTFS ..........A ................ ..... ........ I .................. ....T

mw(yIFE(McW I N P ~ F S A K F Q D

L E E S G F A W ~ L I F RN Y M Q T...........T.. 9 T T.

~

........................... ........................... ...........................

SAM

.. ......... .. ........................ ............ ........... . ........................... ..9 ..................... T.. ........................... ..Q ........................ ...........................

..... ..... .....

..... ..... ..... ..... .....

.....

bll b29

A m L E Q s G c G v v K p u i s ~ c E G ~ mN

S8

LFESGGt L V G f G R s L R v s c ~ m s S Y W

D D M

..E ........................

.............. ................. .............. ................. .............. ................. LIKSKADGETIDYAWVKG .. ........... ................... .......K....I. ................... G

LIKYDGRNKYYADSVKG m

.............. ................. .............. ................. .............. ................. .............. ................. .............. ................. .............. ................. .............. ................. .............. ................. .............. S1KSKFDGGSPHYAWVE.G ................... w

K

G

FR3

FR4

CDR3

RVTISADRFCPILYELWLRIEDTAIYYCAR E R R E E U X 4 P m D F

............v....v .................. ................ ............v....vEv ......HE.... V ........SWH.T.... T ................. I RFSISRN)~I@€LUEiGLKhEDTCWIW\T KYPRYSLMVXC+JHFYMV ..T .........L......... .......... .....F. .......... .. T .........L....... ............ L.... ................................ ........M ........... .. T .........L......... .......... ........m.......... ................................ ........M ........... ..T ....YI...L......... .......... .....Y ..M.....Y.... ......APL ...I........DD... ..... ................................ .......M..RV........S....V.F... ......W L...I........DD...V..... V

.pR

.....,..pa.....

.... ........... ..... .....TT.I... .I..S .....T.I..S

WGPGTRVMP .T.

WGKGITvlvSS

..T ........ ........... ........... ........... ........... ...........

........D .................I..... ................................ ........D .............T...I.....

....T ............. .................. ....T .............

WGKFIIVIVSS

RFSIS R E E ( L E D I W U E D S L Y C A T

................................

QKPRYFDLLSMYRRVlrCAFW

WGffinVNsP

RFTI SPDNSXNl'LY IJ3NSLRAWTAVYYCAR 5 E................

DIGLKGEHYDILTAYGPDY

wGocRvNss A......

RVFTADT~AWIRSLRSADTAVYYCAR

..T .......N................

..... K..... ........... .....K.....

-c

...................... ........... ..... ...S.W. .............. ...........

................................

................... .................... ................... ................... .....A ............. ................... ................... ................... RYPRYS p m c v I M H F y p O V ....................

QISSSGSRTtYADSVKG

RFTIS R D I W ( N S L Y L E M I S L Y C A R

GRRLVIFGGWSGGNI

VISYEGSDXYYADSYXG

R F T 1 S R W S ~ L Y U ) M ) S L T A L Y Y C A R NIRENIEMIGTAYYSYIWW

.. ............ ................................ ..5 . ...........E.. .............. ................................ ................................ ................................

................................ ................................

RFTISRPDLEDKWUTGVfYCAT

.... ........... .... A...... ........... ........... ........... ........... ...........

HiKGTzvSVSS

...........

WGQiRulVsS

WGKGTIVIVSS

FIG.15. Amino acid sequences of variable heavy (VH) domains of Fabs binding to a 1 2 0 (from Barbas et al., 1993b with corrections). Identity with the first sequence in a group is indicated by dots. The N-terminal sequence LEQSG arises from the VHla and LEESG from the VH3a primers. The b l l and b29 sequences are very similar to the b3 group and could be argued to be intraclonal variants within that group; they are placed in their own group because of differences at the V-D and D-J interface.

Clone

FR1

CDRl

............ ..................... ............ I........ ......GT.. L.P.E.A.U.

...... ........... ...... N.... .......SVISNYU .svsmaLA

FRZ

ELTQSPSSUASWXJRVWK WSISNYU I........ N....

b3 b5 820 s2 s3 55

ELlL?SGTLSUFGERATKSC RILsElRMwxa WYPCI1PG(WRUIS E L ' I D S P A S V S A S W I T C WcOIHHlu WYWODG1UPKUIY ELTQSPGTLSUPGERATLSC w u m y u WPQOAPRLLIY QSPDTUINPGEPATLSC RASHRISSKRU lwwiwmWRLLIY wQoRpG(wRLyIy Q s P s H u u ~ ~ I xRAsoRvsApm : QSPCTLSUPGDRATLSC ~ 1 5 s s F L AwYwBG(wRLLIY QSPGTLSUPGERATLSC RASCSFSSNFU HYQcpKwRLLIY

51

w

b7 b12 b21

......GT..

L.P.E.A.15.

CDW

w ~ ~ R L L IMSTLQP Y R...A.N....

GAS-

AAssm

G S S W

VCPNRAG GASlTlAT SASW

VHPNRAT

A.......

ba bl8 b22 b27 835

ELTVSPSSLSASKDRmITC RILsCSIsNYIN N.N....

s6 b20

S4

b13 82 6 00

bl 1 b29 S8

loop35

............. ......

..................... ..................... ..................... ..................... .................. ..................... ..................... ................I....

..

.... ....... s... ....... s.. ....... s... ....T..S...

............... ....... .... WYWIBGWKLLIY T.M. ....... E......H .: ....s ............... ......s ............... ......s ............... ......s ............... D. .NSET ............... D. .N.ET ............... ............... G..N..S MSSIx)R .S

rLoena

.............. .....V.H.

FGoGmyEIKRI

.. ............ .. G........ ............

JK1 JK4

FGOGTI(LE1KRT FGPGlWDIKR

JK2

GIPDRFSGSGSGIDFTLTISRLEPDDFAWYC PPYGDSPLYS GVPSRFSGRGSGIUFTLTISSW!ZDFA~YC PPevSFPK GIPDRFSGGGXilDtTLTISRLEPEDFAVYYC WYGNSWT GVPDRFSGSGSGIDFTLTYSRLEPEDFMYC OYYGGSSYT DIPDRFSGSCSGTIXTLTISRLEPEDFAIYYC QWGQSPVI. GIPDWRGSVSGIDFTLTITFAWYC OWGTSPLM GVPDRFSGSGSGIUFTLTIRWFAVYYC PPYGASLVS GISDRFSGSGSGIDFTLTXTEDFALYYC QWGASSYT

L..........

N.......

ELTQSPGl'LSLSPGEPATLSC RAGQSISRFIU WM)KPG(HpRLLIY GASNRAT A.... .S.T... S..L.N.... ...TlQs .TS.G. .NYU SS..A.V.D.V.IT. wv.x

FR4

.o.GTs.w. .O.H.S.Y.

... ...... ...... ..P................S.L......V... ............ ............... ............... .............. ................................ ..................... ....

b6

CDR3

GVPSRFSGSGSGRIITLTISSWWWSYYC WYNSAPRT T............................ .v..

GV.NRAT G..NRAT

ELTQSPCTLSUFGEPATFSC RSSHSIRSRRVA WYOIIXFGQWRLVIH GVSNPAS V.G L.Y G .T........L. T......L..

........... ..................... .... ..D......

FR3

.... R.......... ............ss ... .......... K.. .... ......s ................................ ... P............................ ....... m...... .I.D.............-..R.E...F.V.S. ....R..M...... .I.D.........A..............I...

bl b2 b14 82 b24 830

GIPDWSGSGSGIDFTLSISRLEPWFAWC

........ G.... ....T.............. .G.S.............T.NS.Q...V.T.S.

...Q......S...R............ PPYGTSPYT .... .N.DSA.W. N.V..

........TA...E...T.R............ .................T.......... .................T..........

SPGEPATUC WCSVG'I?3U

WQlXFGQAPRLLIF

L....

FGQG?xLEIKR FGOCIXVEITR FCQ.R(WR FG(X;IXWR FGPGIMMIKR

FGPGTKIEIIKRT I...

JK1 JK4 JKl

JK2

JK3 JK2

JK2 JK2

JK2 JK3

JK2 JK2

........ ........... ....... DF...

JKZ JK2

..... K.E... .....Kv....

JK2 5x2 JKl

FCoG~WIKRr

.....T.YT ..o......... .... T.YT. ..O...L.. ... .... ..... ...T.QT ..Q...L ..... ...... .T.YT ..Q...L.....

m 4

............ ..Q . . . . . . . . .

JK4

FGGGTY(LEIKP3

...........

JKZ

FGUXKVEIKRT

JK4

FGQjlxLEvxRT

Jxz

woQKpGQlpRwIY

RAsQ)xmw WYHQrnWKLLIS

......

GVPSRFSGSGSGIDFTLSISSLQPWFATYYC PPSYSIPPLT FGGGTKVEIKRT

................. T.............. ........ G........T.....;........ .....T.M ............ R...FT........V..... ..HQMI.LT Q...D.R.... ..yDNL.LT ................FT........ I..... Q...D...H.. .................T............F. ....w. .wT ....T.N.... GVPDRFSGSGSG'IU~LTISRLEPEDFAVYYC OPFGMQYT ELTQSPGIUUPGEPATLSC R A S m m ....... ................................ ......... ....... ............ ............... STSPRAT lL?SPSSVSASKc7TvrrK

G........

PPYADLXT

DASDLEI

GVPSWSGSGSATYFSFTISSLQPWIGIYYC

DAS1RDT

Y I P D ~ S G ~ S G I D F A L T I S S ~ W f f i F Y Y Cwmwm

.. O...L..... ............

FIG.16. Amino acid sequences of variable light (V,) domains of Fabs binding to gp120.

JK2

JK2

JK2

JK2 JK2

JK4 JK1 JKZ

236

DENNIS R. BURTON AND CARLOS F. BARBAS I11

many cases, the light chains in a group were each paired with heavy chains from the same group and defined predominant H-L combinations, e.g., the b4 group and most of the b8 group (Fig. 16). However, many of the light chains were less easily grouped together on the basis of sequence. If constrained into the groupings defined by the heavychain sequence similarities, evidence of heavy-chain promiscuity was clearly revealed. For instance the b3 group shows seven closely related heavy chains (Fig. 15) paired with seven very different light chains (Fig. 16). Such extensive promiscuity is suggestive of some preeminence of the heavy chain in determining antigen binding in these cases. To further explore chain pairing, the heavy chain from a single clone (b12, a potent neutralizing Fab) was recombined with the entire original light-chain library and the resulting library (H12-Ln) panned against gp120 LAI. Similarly the L12-Hn library was constructed and panned. Figures 17 and 18 show sequences of resulting binders. The apparent affinities of these binders were roughly similar to those for the original Fab (b12 = H12 L12). Among the light chains selected using H12, three are very similar to the original light chain whereas the rest arise from several different V-J rearrangements. In contrast, the shuffled heavy chains probably all arise from the same V-D-J rearrangement although there are several differences in the V gene segments. Therefore the shuffling data suggest some preeminence of the heavy chain in antigen binding for b12 Fab. Chain shuffling enables pairings to be evaluated in a competitive situation. Pairing has also been examined in a forced situation using a binary plasmid system (Collet et al., 1992).A11441 H-L combinations arising from 21 anti-gpl20 Fabs (from Figs. 15 and 16) were examined for reactivity in ELISA with gp120. The results are shown in Table V and indicate extensive chain promiscuity. The degree to which a given heavy chain paired productively with any light chain to bind antigen varied from 43 to 100% for antibodies selected on gp120 LAI. It should be borne in mind that all of these antibodies bind to the CD4 binding site of gp120. For two clones selected on other antigens, gp120 SF2 (s8) and a V3 loop peptide (p35), heavy-chain promiscuity is notably absent. This topic was further probed by directed crosses of heavy and light chains originating from several of the anti-gpl20 Fabs with an anti-TT Fab. The heavy chains from several of the antigp120 Fabs in combination with the light chain from the anti-TT Fab retained binding to gp120. The heavy chain from the anti-TT Fab retained toxoid binding in combination with several of the light chains from anti-gpl20 Fabs. None of the Iight chains from the gp120 binders

CDRl

FR2

CDRZ

Clone

FR1

HIV-H12/L12

ELTQSPGTLSLSPGERATFSC RSSHSIRSRRVA UYQHKPGQAPRLVIH GVSNRAS

FR3

CDR3

GISDRFSGSGSGTDFTLTITRVEPEDFALYYC QWGASSYT FGQGI1(LERKR

HIV-H12/LC25

.............. .... ........................................... L.......... .............. D...... ....NNN.................... ...................... ........................................... L.......... ......................... T ..................... L.......... ..... R......L.Y ... S..G .vP............ S...S.L......M... .A..R.S...L. D....NV.....L.. ...... ..... R......L.Y ...... G .vP............ S...S.L......I... D....NA.....L.. .A..R.S...L. ...... ....RR...... L.Y ......G .vP................ S.L......I... .A..R.G...L. ......D....NT.....L.. ...K..G ..P........A.......S........V... T.....IL.. KT..N.W...L . . . .LL..SS............LL..YY ........... ...KK....GG .. ..PP...............A..A..........S......S.........V.......V... ...L.S......L.Y ...........TT..........ILI.L... KKTT....NN..WW......LL. .... ... ........... ......~... ST.....IL.. KT..N.W...L. ...V.S.LP...L.. ...R..G ..P........AR......S.L..A...V... ...........N.....VL.. .T.RN.W...L. L.. ...K..C .vP........AR...... S.L......V.F.

HIV-H12/L12

ELTQAPGTLSLSPGERATFSC RSSHSIRSRRVR UYQHKPGQAPRLVIH GVSNRAK GISDRFSGSGSGTDFTLTITRVEPEDFALYYC QWGASSYT

HIV-HlZ/LCll HIV-H1Z/LC24 HIV-H12/LC22 HIV-H1Z/LCl HIV-HlZILC’I HIV-H12/LC28 HIV-H12/LC13 HIV-H12/LC3 HIV-HlZILCS HIV-H12/LC26

D......

...VRR......

FR4

................ DF.. ................ ................DF.. ...... VDI.. .T..G.... ..... .T..G.... ......TVDI.. VDI.. .T..G.... .T..G.A.. ....... DI.. ....... .T..G.A.. .T..G.A.. ........DII.... .T..G...S .......DFN. .T..G.... ....N..DIR. DF..

FGQGTKLERKR

FIG. 17. Amino acid sequences of VL domains from Fabs binding to gp120 and generated by shuffling the heavy chain from clone b12 against a library of light chains (H12-LCn Fabs).

lone

F7u

w-wL12-HC1 pIKL........................... L12-HC2 QVKL ........................... L12-HC3 Q\IKL ........................... L12-HC7 QVKL................ I......... T L12-HC9 QVKL .................... L12-HC10 QVKL................ L.......... L12-HC11 QVKL.....T..........I..K...... T .L12-HC12 QVKL ....................... :... L12-HC13 QVKL ........................... L12-HC14 QVKL ........................... L12-Hl2

.2-H12

-

S

w

W

CDP3

CDRl

FRZ

CDR2

FR3

NWIH

wvRI;v\pGQRFEwH:

W V A X E Q D

RVlFTAWTSAMlYHELRSLRSADTAVWCAR

.. .L.

.A .....H.P.... .A H.P .... .A.. .H.P.... .A.......P.... .A ..... H.P .... .A .....H.P.... ......... P....

....A.

..%.R ....G.V.L.

. ..L.

.. .L. .. .L.

.. .L.

.. .L. . .PL.

..... .. .L. .. .L. NNIH

..... ..

.VP. IPP....

....A..VT.I.P....

....A. .VP.I.P. . . .

A . . .H.P....

.F..A..I..I.P.... ....A..W.I.P.... ....A..VT.I.P.... ..KIV..E.KY.Q..V. ............... R. . _ . . A . .V..I.P ....A..W.IPP....

W V R o A m -

WI-?+K€QD

..............

.A....T.DL....

.

....

FR4

~KGTlVIVssAsIx

..

..Em.............

.......T.

..SL.G....S.V.L ......F.......... ..S ..G....S...V...N............. ..SL.G:. ..S.V.L.. .N..F.......... ..SL.G ....S.V.L ...N ..F.......... .....T..........V.G ....... T..... ........D ................. I.....

..Em.............

....... ....... ....... .......T ...... .T

.'IN.. F........

..SL.G ....S.V.L ...N ..F..........

..SL.G....S.V.L ...... F.. ........ . .SL.R ....G.V.L..W ..F..........

RVlFTADlSAWAYWLRSLRSADTAVYYCAR

..m............. ..m.............

. . m . .F.. ........

..m.............

..Em..MD..A . . . . . .

.... T............. . .m. . . . . . . . . . . . .

..Em.............

..R.. ..T

....... .......

.......

WGrnIVSSAsTK

FIG.18. Amino acid sequences of VH domains from Fabs binding to gp120 and generated by shuffling the light chain from clone b12 against a library of heavy chains (L12-HCn Fabs).

ANTIBODIES FROM COMBINATORIAL LIBRARIES

239

TABLE V CHAIN PROMISCUITY REVEALEDBY HEAVY-LIGHT CHAINRECOMBINATIONS I N A BINAHY SYSTEM^ PLASMID

-

.

*

bll b6 b4 b12 b7 b21 b: 511

+

b6

+

b4

+

512

+

b7

+

321

+

b3

+

313 w

b22

+

326 w b8

+

318

+

327

+

B8

+

B35

+

s4

+

bl

+

b14

w

b24

+

s8

w

p35

+

+ + -

'Ilj

+ + + + + + -

+ +

+

+

-I+

W

+ + + + + + + +

t

+ + + -

+

t t 1 7 $ $ $ $ 9 $ 113 b22 b26 b8 b18 b27 B8 B35 s4 b l b14 b 2 ~

+

+

+

+

+

+

-I-

-I+

+ +

+

-

-

+

-

-

-

-

-

+

-

-

+

-

-

-

-

-

+

-

-

+

-

-

-

-

-

+

-

-

+

-

-

-

-

-

+

-

-

+

-

-

-

-

-

-

-

-

+

-

-

-

-

-

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

w

-

+

w

-

-

-

-

+

-

-

+

-

-

-

-

-

-

-

-

+

-

-

-

-

-

+

+

w

w

w

w

+ +

+ w

+ w

+ w

+ w

+

+

+ +

From Collet et al., 1992. Directed crosses between heavy and light chains of Fab fragments isolated from a single library by panning with glycoprotein gp120(bl-b27) or gp160(B8-B35)of HIV-1 strain IIIB, gp120 (s4-s.8) of HIV-1 strain SF2, ind the loop peptide (p35), assayed by ELISA against IIIB gp120. Heavy chains are listed horizontally and ight chains are listed vertically. Clones are sorted according to the grouping established in Barbas et aE. 1993b). Different groups are separated by horizontal and vertical lines. ELISA results: -, negative (a signal If three times background or less); +, positive (comparableto the original heavy- and light-chain combination); w, intermediate value; *, the HCp35/LCp35 combination was negative with IIIB gp120 but positive with SF2 ~ 1 2 0Identical . chains cany the same identifier (either Ir, II,5, or $).

240

DENNIS R. BURTON AND CARLOS F. BARBAS I11

were able to confer gp120 specificity in combination with the anti-TT heavy chain. Similarly, the light chain from the anti-TT Fab was unable to generate TT specificity in combination with any of the heavy chains originating from the gp120 binders. These results again support heavychain dominance in antigen binding. Heavy-chain promiscuity has now been described in several reports using combinatorial libraries (e.g., Caton and Koprowski, 1990; PersRemarkable son et al., 1991; Clackson et al., 1991; Kang et al., 1991~). examples of chain promiscuity include those where K and A light chains combine with the same heavy chain with maintenance of antigen binding (e.g., Zebedee et al., 1992; unpublished. observations). The negative side of this promiscuity with respect to the ability of libraries to report on antibody responses is that the in viuo light chain can probably not generally be identified. The positive side is that this assumes rather less importance if the heavy chain is dominating antigen binding. Light-chain promiscuity, i.e., the ability of the same light chain to pair with different heavy chains with retention of antigen binding, has also been described. We argue that, in most cases, this behavior is most appropriately viewed as light-chain plasticity, rather than lightchain dominance. The reason is that the evidence indicates that the heavy chains in these cases are dominating the binding specificity with the light chain acting as a pluripotent or plastic partner. For instance, a comparison of sequences from B-cell lines (Kabat and Wu, 1991) shows identical light chains pairing with different heavy chains in antibodies of widely varying specificities. This phenomenon is a relatively common occurrence in library studies. K light chains deriving from two germline genes, kv325 and vk02/012, are particularly prominent. The kv325 light chain in a near germline configuration has been described in library-derived antibodies to thyroglobulin (Hexham et al., 1991,1992), hepatitis B surface antigen (Zebedee et al., 1992), HIV-1 gp120 (Barbas et al., 1993b), and thyroid peroxidase (Hexham et al., 1994). A dramatic example of kv325 light-chain plasticity is provided by changing the specificity of an antibody from tetanus toxoid to fluorescein solely through changes in the heavy-chain CDR3 region (Barbas et al., 1992a). We would similarly tend to invoke plasticity as the most likely cause of the involvement of a similar light chain with different heavy chains in binding the same antigen. For instance the light chains B20 and s6 in Fig. 16 are virtually identical. One interpretation would be that this light chain is dominating binding to gp120 and the two heavy chains are of lesser importance. However, when the B20 heavy chain is paired with six other distinct light chains, the retention of gp120 binding

ANTIBODIES FROM COMBINATORIAL LIBRARIES

24 1

observed strongly implies the preeminence of the heavy chain. The B20/s6 light chain is likely relatively passive in gp120 binding and has plasticity to allow the heavy chain specificity to be expressed. Precisely the opposite interpretation of this sort of data has been made from library studies on antibodies to thyroid peroxidase. This set of studies is taken up in more detail below. In one chain shuffling investigation (Portolano et al., 1993a), the authors found the same lightchain (vk02/12 derived) pairing to two different heavy chains with maintenance of antigen binding and suggested that the light chain may be defining epitope specificity. Our feeling is that a “plastic” light chain is a good tool to go fishing for heavy chains in the chain shuffling experiment. It is worth emphasizing strongly that we do not suggest that heavychain dominance, heavy-chain promiscuity, and light-chain plasticity are attributes which will apply to every heavy-light chain combination. For instance, there will undoubtedly be cases where light chains dominate antigen binding. Further, plastic behavior strictly requires an identical light chain in different antibody specificities although “very similar” (typically 95-99% identity) light chains may merit this description with’the caveat that single amino acid substitutions can have profound effects on binding affinity and specificity. We also emphasize that heavy-chain dominance does not imply that heavy chains alone could function as isolated binding units or that light chains in such cases make no contribution to binding. Since libraries are random combinations of heavy and light chains, it can be argued that combinations using promiscuous heavy chains and/or plastic light chains will be recovered from libraries with greater frequency than their occurrence in natural responses. Similarly, antigen binding sites dependent on very restricted heavy-light chain pairings may be underrepresented in antibodies obtained from the library approach. Shuffling experiments using heavy chains isolated by conventional means would be of interest here. For K light chains, there is a growing body of evidence to indicate that certain light chains are overrepresented in the expressed repertoire. Using our PCR primers, we find a strong overrepresentation of K chains with kv325 or vk02/ 012 as the closest germline gene in unselected libraries from peripheral blood lymphocytes of a healthy donor (G. R. Pilkington, unpublished observations) or from thyroid lymphocytes from a donor with autoimmune disease (Hexham et al., 1994). This overrepresentation is reflected in the light chains from Fabs binding gp120 in Fig. 16. A number of factors, apart from considerations of chain pairing, might also skew the cloned response relative to the natural response. These

242

DENNIS R. BURTON AND CARLOS F. BARBAS I11

include unequal PCR priming, the presence of restriction sites in certain antibody genes, or the preferential expression of certain antibodies on the phage surface or in E . coli. The more PCR primers are used the more complete the captured repertoire is likely to be. However, using a relatively small and manageable set of primers (Persson et al., 1991; Kang et al., 1991b; Williamson et al., 1991),we have obtained members of most of the heavy- and light-chain families and a wide diversity of genes. As regards restriction sites, one systematic loss is currently incurred in our system, for example: a family of A light chains possesses a restriction site used in the cloning process in the constant domain exon. From known sequences we estimate we presently lose about 5% of the total repertoire because of internal cloning sites. The effects of antibody sequence on bacterial expression levels are great ranging over at least 2 orders of magnitude and so this is clearly a factor which could skew cloned repertoires. IV. Antibodies to Self-Antigens from Human Donors

Combinatorial libraries have been applied to human thyroid disease, primary biliary cirrhosis, and Graves' ophthalmopathy. They have also been applied to the autoimmune repertoire in HIV-1 infection. These cases are now considered in turn.

A. THYROID DISEASE Autoantibodies to thyroid peroxidase (TPO), to thyroglobulin (Tg), and to the thyroid stimulating hormone (TSH) receptor are important features of thyroid autoimmunity (forreviews see Funnaniak and ReesSmith, 1990; McLachlan and Rapoport, 1992; Bigazzi, 1993; Kendler and Davies, 1993).Autoantibodies to TPO and Tg are associated with hypothroidism while anti-TSH receptor antibodies generally cause hyperthroidism. A set of pioneering studies have used the library approach to probe the autoimmune repertoire to TPO in Graves' disease. Prior to these studies, despite numerous attempts using EBV transformation and/or cell fusion, only one cell line secreting an IgG anti-TPO antibody had been produced and this human-mouse hybridoma was unstable (Portolano et al., 1991; Fukuma et al., 1990). The studies have used thyroid tissue from Graves' patients as the source of RNA. Initially an I g G k Fab library was constructed in A phage and this was screened to identify one (Portalano et al., 1991) and then three (Portolano et al., 1992) Fabs with high apparent affinities (around 10" M-')for TPO. The heavy-chain sequences of the three Fabs were identical. Two of

ANTIBODIES FROM COMBINATORIAL LIBRARIES

243

the light chains were very similar and a third was more distant. The closest germline gene to all three light chains was vk02/012, although for the more distant clone homology was only 92%. These results therefore again indicate heavy-chain promiscuity. Importantly, the Fabs were shown capable of inhibiting a high proportion (36-72%) of serum TPO autoantibodies from 11patients with autoimmune thyroid disease, implying the epitope recognized is of major significance. This finding also applied to patients whose response was predominantly IgG4 as well as those for which IgGl antibodies were prevalent. The epitope recognized was shown to be conformational. The heavy and light chains from one of the above Fabs were then used in chain shuffling experiments (described as “roulette” by the authors) (Portolano et al., 1993a). The fixed heavy chain found a set of light chains with which antigen binding could be generated, all of which were most closely (92-95%) related to the vk02/012 gene. However, there were many differences between some of the chains in the VKgene and at the VK-Jjoining region and at least two J K regions were used. The fixed light chain found five heavy chains closely related to the original and one new heavy chain which clearly arose from a different V-D-J rearrangement. The Fab with the new heavy chain competed with the parent Fab (which has an identical light chain). The authors therefore raised the possibility that the light chain was dominating binding and defining epitope specificity. We prefer an alternative explanation. The original heavy chain did pair with several quite different light chains, albeit that these were most readily related to the same germline gene, suggestive of heavy-chain dominance. The competition of two Fabs with different but dominant heavy chains is common (compare Fig. 15 where all of the Fabs except one are crosscompetitive). As above, we suggest light-chain plasticity or pluripotency is likely to be a better explanation of the finding of the same light chain in two antibodies than light-chain dominance. In their next study, the investigators identified TPO binders in phage A libraries prepared from three further Graves’ disease donors to yield a panel of 30 cloned high-affinity IgG Fabs (Chazenbalk et al., 1993). The heavy chains could be organized into four groups with many closely sequence-related chains within a group as described above for the cloned response to gp120. Here, however, closely related heavy chains were derived from libraries from different donors. This is a potentially very important finding. However, the possibility of crosslibrary phage contamination should be borne in mind (we have found this to be a considerable problem) although considerable precautions were taken in library construction (S. M. McLachlan, private communi-

244

DENNIS R. BURTON AND CARLOS F. BARBAS 111

cation). There was a very strong bias among the binders toward light chains having vk02/012 as the closest germline gene. Some responsibility for this bias may lie in the fact that unselected libraries have a very high (85%) representation of this light chain (Jaume et al., 1994). It is unclear at this time whether this is due to a real skewing of the repertoire in the tissue source or to the PCR. Next the investigators subcloned heavy and light chains, maintaining pairing, from one of the A phage libraries to the phage display vector pComb3 (Portolano et al., 1993b). This library was panned against viable, stably transfected CHO cells expressing human TPO on their surface. Several high-affinity Fabs were derived. The heavy chains fell into four groups, two corresponding to rearrangements described previously using phage A and two to apparently different rearrangements. Two rearrangements found in the A phage system were not found using the phage display system. The light chains again showed a bias toward vk02/012 and a tendency to be closer to the germline than in Fabs derived from the A phage library. Therefore the phage display system recapitulated some of the features of the A phage system but there were important differences. Very recently, directed crosses have been used to explore the promiscuity of heavy and light chains in binding TPO (Constante et al., 1994). Three heavy chains clearly originating from different germline genes were combined with one or two light chains clearly originating from germline genes different from that of the original partner. Five new combinations were produced. None of these bound TPO providing no evidence of promiscuity among the chains examined. A second group of investigators have carried out a more limited set of studies using libraries on anti-Tg and anti-TPO antibodies isolated from donors with Hashimoto’s disease. Initially a high-affinity (apparent K , = 5 X lo9 M - l ) antibody to Tg was isolated from an IgGK A phage library (Hexham et al., 1991,1992).The light-chain V gene from this Fab was identical to the kv325 germline. Using the pComb3 phage display system, the investigators have reported three Fabs with high affinity (apparent K, = lo9 M - ’ ) for TPO (Hexham et al., 1994). The heavy chains were unique while two of the light chains appeared derived from the vkOW012 germline. In this case the unselected library had 18% vk02/012 and 36% kv325. B. PRIMARYBILIARY CIRRHOSIS (PBC) PBC is a chronic autoimmune disease of the liver of unknown etiology and is characterized by the presence of high-titer antimitochondrial antibodies, inflammation of the septa1and interlobular bile ducts,

ANTIBODIES FROM COMBINATORIAL LIBRARIES

245

followed by necrosis, and ultimately cirrhosis (reviewed in Rowley et al., 1993). In recent years, the mitochondrial autoantigens have been identified as components of the 2-0x0 acid dehydrogenase enzymes including the E2 subunits of the pyruvate dehydrogenase complex (PDC), the branched-chain 0x0 acid.dehydrogenase complex (BCKD), the 2-0x0 acid glutarate dehydrogenase complex, and other protein components. Murine monoclonal antibodies to the mitochondrial autoantigen PDC-E2 have been isolated. Although these monoclonal antibodies were able to inhibit PDC-E2 enzymatic activity, they recognized a different region of the target autoantigen than did patient sera (Surh et al., 1990), suggesting that monoclonal antibodies of rodent origin differ in fine specificity from human autoantibodies. mRNA from a regional lymph node of a patient with PBC has been used to construct a combinatorial immunoglobulin library in the A phage vector system (Cha et al., 1993). Six human monoclonal IgG Fabs specific for the major autoantigen of PBC, PDC-E2, were isolated, appearing at a frequency of 0.01% in the combinatorial library. These Fabs recognize human PDC-E2 with high affinity (K,= 107-109M- 1). Using both immunoblotting and ELISA, the Fabs showed little crossreactivity to any of the other autoantigens commonly recognized by PBC sera or to other antigens commonly recognized by PBC sera or to other antigens. The Fabs showed a typical antimitochondrial staining pattern in HEp-2 cells but reacted strongly with the luminal aspect of biliary epithelial cells of patients with PBC. It was argued that the Fabs have similar specificities to those found in human PBC sera and in human monoclonal antibodies prepared by conventional means. OPHTHALMOPATHY C. GRAVES’ Graves’ opthalmopathy is a common disfiguring autoimmune disorder of unknown etiology. Despite extensive investigations over many years, there is still no consensus on the target cell(s) or autoantigeds) involved (Weetman, 1992; Wall et al., 1993). The inflammatory cells infiltrating Graves’ orbital muscle and fat/connective tissue are predominantly T cells, but plasma cells are also present. Hypothesizing that antibodies generated within Graves’ orbital tissue may include autoantibodies against orbital antigen(s), Jaume et al., 1994 constructed an I g G k library in A phage from the orbital tissue of a patient with active Graves’ ophthalmopathy. Sequencing of 15 unselected heavy and light chains revealed an extreme bias (14/15) toward the vk02/012 germline among the light chains and some restriction in heavy chain gene usage. The authors raised the possibility that particular germline genes may be associated with autoimmunity in humans.

246

DENNIS R. BURTON AND CARLOS F. BARBAS 111

D. AUTOANTIBODIESIN HIV-1 INFECTION Although most interest concerning the pathogenesis of HIV-1 infection has been focused on T cells, abnormal B-cell function is also a feature of the infection. Individuals infected with HIV-1 have serum antibody titers to a wide range of self and exogenous antigens. The origin of the autoantibodies and their effects are unknown, although speculations as to a pathogenic role have been made. The autoantibodies in HIV-1 infection appear early postinfection and persist throughout the disease course. In general, HIV-positive patients and in particular patients with AIDS have elevated numbers of spontaneous immunoglobulin secreting B cells in peripheral blood giving rise to h y pergammaglobulinemia. Several hypotheses for the retroviral induction of autoantibodies have been suggested (reviewed in Amadori, 1993). One hypothesis is that the virus or viral proteins themselves, directly or indirectly, induce generalized polyclonal B-cell activation that leads to elevated levels of all antibodies, including those to self-antigens. Another hypothesis invokes molecular mimickry whereby a viral epitope elicits antibodies that cross-react with a self-antigen. Sequence similarities have been reported for MHC class 11, particular HLA-DR and HLA-DQ, and gp120, HIV-1 Nef, and gp41, and between the major retroviral capsid protein (CA) and regions in the SM-BB' autoantigen, the SSB/La antigen, the 70-kDa ribonucleoprotein, topoisomerase I, and the acetylchoIine receptor. Finally, an autoimmune model of AIDS pathogenesis that involves both an immune response to HIV and to allogeneic stimuli has been proposed. A synergistic effect of the two responses is postulated to cause collapse of the immune system. To characterize the autoantibodies seen in asssociation with HIV-1 infection, we decided to use several libraries to select a series ofantibodies against a range of autoantigens (Ditzel et al., 1994). The sera from eight HIV-1-seropositive male donors were first tested in ELISA for binding to the purified autoantigens: MHC class 11, double-stranded DNA, CD14, EGF receptor, and the ganglioside GD2. As shown in Fig. 19, higher average serum IgG titers against all five antigens were found in the eight HIV-1-positive donors compared to eight healthy seronegative control male donors. Serum IgG responses to the autoantigens varied significantly between HIV-1 donors, and some had serum levels comparable to those of healthy controls. Bone marrow IgG combinatorial libraries were constructed from the eight HIV-seropositive donors and selected libraries panned against antigens from the set above in order to retrieve a panel of 38 monoclonal autoantibodies.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

247

u

el 2-

MHCclassll dsDNA EGFFI CDt4 ganglioside GD2

v)

0

d

.

n O

1-

0-

HIV-donors (n:8)

normals (n:8)

HIV-donor L

FIG. 19. Reactivity of sera (diluted 1:lOO) from eight asymptomatic HIV-1seropositive male donors and eight healthy males against five selected autoantigens individually assessed by ELISA and then averaged as shown (from Ditzel et al., 1994): Several HIV-1-positive donors (e.g., donor L) had strong serum responses to all of the autoantigens tested compared to those ofthe healthy donors. Serum samples were taken concomitant with bone marrow aspiration for library construction.

The specificities and affinities of the Fabs for a panel of autoantigens were determined using ELISA. As exemplified in Fig. 20, the selected Fabs showed moderate affinity for antigen and cross-reactivity with several other autoantigens as well as exogenous antigens. The pattern of cross-reactivity was individual to a given Fab. These findings contrasted with antibodies selected to infectious agents from the same libraries. In a number of cases, the serum titers to the exogenous antigens were very similar to those to self-antigens. For example, the serum from one donor had titers against double-stranded DNA and gp120 of 1:500 and 1:400, respectively. However, as depicted in Fig. 20B, two human anti-gpl20 Fabs retrieved from this library were of high affinity and monospecific, in contrast to the autoantibodies selected by panning against DNA, which were polyspecific and of lower affinity. Similarly, there was a marked contrast between the antibodies derived from HIV-1-seropositive donors with those derived from donors with established autoimmune disease. From a library constructed from a systemic lupus erythematosus (SLE) patient (S. M. Barbas, E. M. Salonen, H. J. Ditzel, G. J. Silverman, and D. R. B., unpublished data, Fig. 20A), we have isolated high-affinity

248

DENNIS R. BURTON AND CARLOS F. BARBAS I11

.01

.1

1

10

100

1000

1

10

100

1000

Fab conc pglml

. B 2.0

1.5 v)

0

d

8

1.0

0.5

0.0

.01

.1

Fa b conc pglml FIG. 20. The polyspecific nature of autoantibodies isolated from an HIV-1seropositive donor by the library approach (from Ditzel et al., 1994). (A) Comparison of monospecific anti-DNA antibodies from an SLE donor library (SI) and polyspecific anti-DNA antibodies from an HIV-1-seropositive donor library (donor L). The solid lines indicate the binding properties of a Fab [LNA3; donor L, antigen DNA (NA), clone 31 selected by panning the L library against human placental double-stranded DNA (dsDNA). LNA3 was assayed for binding to solid phase dsDNA (0),ovalbumin (a), human transferrin (+), BSA (m), human IgG Fc (A), and ganglioside GTI (a). The dashed lines indicate the binding properties of two Fabs (ds3-40and ds3-32)selected by

ANTIBODIES FROM COMBINATORIAL LIBRARIES

249

anti-DNA antibodies using the library approach. These Fabs showed no cross-reactivity with the same panel of autoantigens. The polyreactivity of the Fabs from the HIV-positive donors was confirmed in ELISA competitive inhibition assays in which the binding of a given Fab to a solid phase antigen was tested in the presence of both homologous and heterologous soluble antigens. The apparent affinities of these antibodies were of the order of 106-107A4-l. Clearly it is not possible to examine all autoantigens and therefore the study described could not unequivocally eliminate the presence of specific autoantibodies in HIV-1 infection. However no support for this hypothesis was found, particularly in cases where sequence similarities between HIV-1 proteins and self-antigens have been reported. Instead, the results suggest that an increased level of polyreactive antibodies is probably responsible for the elevated serum titers to autoantigens in the serum of HIV-1-seropositive donors. A role for polyreactive antibodies in the pathogenesis of HIV-1 infection is unclear. However, the affinities of the monovalent Fab fragments described were of the order of 106-107M-' for arange of antigens. As bivalent IgG molecules, the affinity for antigen (avidity) may well be an order of magnitude higher, which would be expected to permit in vivo antibody-antigen interactions. The polyreactivity of the combining site may introduce novel features to the modus operandi of the panning the SI library against dsDNA. The steep dashed curves show the concentration dependence of ds3-40 (A)and ds3-32 (0)binding to dsDNA. Also shown is the binding of ds3-32 to human IgG Fc (x), ovalbumin (+), and human transferrin (H).The binding of Fab ds3-40 to these latter antigens is essentially identical to that of Fab ds3-32. The figure illustrates the much higher affinity (half-maximalbinding at 10" M compared to 10bM) and lower cross-reactivityof the anti-DNA Fabs from the SLE library compared to the Fab from the HIV-1-seropositivedonor. (B)Comparison of mono- and polyspecific antibodies isolated from the same HIV-1 library (L).The solid lines indicate the binding properties of a Fab (LNA12) selected by panning the library against dsDNA. LNAl2 was assayed for binding to the same panel of antigens as in A. The dashed lines indicate the binding properties of two Fabs (L21and L41) selected by panning the library against gp120. The steep dashed curves show the concentration dependence of Fab L21 (A) and Fab L41 (0)binding to gp120. Also shown is the binding of Fab L21 to human dsDNA (x), ovalbumin (+), and human transferrin (H).The binding of Fab L41 to these latter antigens is essentially identical to that of Fab L21. The figure illustrates the much higher affinity (approximately lW-fold, half-maximal binding at 10" M compared to 10' M ) and lower cross-reactivity of the Fabs selected by panning against gp120 compared to the Fabs selected by panning against dsDNA. The IgG titers in the serum of the donor to gp120 and dsDNA were essentially equivalent (1:500) at the time of bone marrow aspiration for library construction. Comparison of Figs. 2A and 2B also shows that polyreactivity is more pronounced for Fab LNA3 and LNAl2.

250

DENNIS R. BURTON AND CARLOS F. BARBAS I11

antibody molecule, e.g., self-association, cross-linking of unrelated antigen molecules, and focusing of diverse species to the surface of Fc receptor bearing cells. Given considerations such as these, the study concluded that the effects of high serum levels of antibodies capable of interacting with diverse molecular species were difficult to predict and warranted further investigation. V. Antibodies without Immunization

A direct consequence of the ability to sort through vast libraries of antibodies is the isolation of antibodies without immunization (Lerner et al., 1992; Marks et al., 1992). Phage display allows for the construction and sorting of libraries whose size matches or exceeds those displayed by an animal at a given moment. Two such approaches have been described. The first takes advantage of the combinatorial strategy and the observation made with hybridomas that naive IgM antibodies can recognize a variety of antigens. Prior to the development of hybidoma technology, myelomas had been screened for their reactivity with small molecules. An extension of this approach is to search through natural libraries and to use phage display to select antibodies against a defined antigen. This strategy, termed the naive repertoire approach, has been demonstrated for both human (Marks et aZ., 1991) and murine (Gram et al., 1992) antibodies. The second strategy involves introducing synthetic diversity into antibody CDR regions (Barbas et d . ,1992a; Hoogenboom and Winter, 1992)and takes advantage of our more sophisticated understanding of antigen-antibody interactions. A. NAIVEREPERTOIREAPPROACH The term “naive” in this sense refers to a library of antibodies which has not been educated or biased by the immune system toward the recognition of any particular antigen or antigens. Preparation of naive libraries involves the use of RNA from nonimmune sources and amplification of p or 6 heavy chains which are the starting point in the natural response. Naive libraries have been constructed from both humans (Marks et al., 1991)and mice (Gram et aZ., 1992)and selection attempts have had some success for the production of low or medium affinity antibodies. Marks et al. (1991) have reported the construction of naive scFv libraries. In this report PBLs from two donors served as the source of mRNA which was used to prepare cDNA using IgM, IgG1, K , and A oligonucleotides for first strand synthesis. This cDNA was then used

ANTIBODIES FROM COMBINATORIAL LIBRARIES

251

as the template for scFv construction. The construction of this library has already been considered (vida supra). From libraries of 2.9 x lo7 VHM-VL and 1.6 X lo8 VHG-VL scFv clones, several scFv's were selected. Clones which had affinities amenable to ELISA studies bound turkey egg white lysozyme, TEL, (3 clones), bovine serum albumin, BSA, (I clone), and the hapten 2-phenyloxazol-5-one, phOX, (1 clone). The BSA binding clone was an artifact from the selection of phOX clones as BSA had been used as a carrier for this hapten. All clones were tested for cross-reactivity with a panel of other antigens and demonstrated good selectivity abeit good to poor signal on ELISA. The affinites of the best TEL clone and the phOX clone were 1.2 x 107M-' and 1.8 x 106M-', respectively. Two subsequent papers have appeared since this initial report and describe additional antibodies isolated from these libraries. Griffiths et al. (1993) describe the isolation of antibodies reactive with selfantigens. The antigens chosen for investigation were thyroglobulin, a human antibody, TNF-a, CEA, MUC1-peptide, and rsCD4. The scFv's selected were of such low affinity only those which dimerized on the phage and subsequently in solution were selected. Affinities of multimeric forms were determined and in one case monomer could be isolated by gel filtration and independently examined. Avidity constants for one anti-TNF-a, three anti-human antibodies, and one antithyroglobulin were reported and range from 0.6 to 1.4 x 107M-'. A fivefold effect was reported to be due to avidity yielding a corrected monovalent affinity of 1.3-2.9 X lo6 M-'. ELISA signals for clones selected against other antigens were extremely weak and in one case did not exceed an OD405 of 0.12. To select these clones a helper phage which was deleted in gIII was utilized in the phagemid system. This resulted in a multivalent system to aid the capture of these very low-affinity clones. Additional valency is added by dimerization g3pfused scFv and soluble scFv released by proteolysis. This suggests that Fab fragments which are clearly monomeric in nature (according to all published reports) may be superior to scFv in phage display for the isolation of high-affinity clones. Alternatively, this may have resulted from selecting from a library which,contained only very lowaffinity clones. Marks et al. (1993) later reported the isolation of six scFv's selected to bind blood group antigens by way of selecting on fixed and suspended cells. This report demonstrates the sensitivity of the phage display approach to select clones from mixed antigen sources as had been done with viral lysates and in particular extends the approach to the cell surface. The isolated scFv's were functional in immunofluo-

252

DENNIS R. BURTON AND CARLOS F. BARBAS 111

rescence; however, no affinity measurements were reported. Characterization of conditions for the selection of cell-surface antigens has also been performed by Portolano et al. (1993b) and by Siege1 and Silberstein (1994). Thus several reports have demonstrated that scFv of modest to low affinity can be isolated from naive libraries. The main problem with the naive approach is the difficulty in ensuring the library is diverse and unbiased. This is due to the fact that the libraries will be influenced or edited in the human and that library construction is extremely susceptible to contamination from mRNA derived from activated B cells or plasma cells. B. SYNTHETIC REPERTOIREAPPROACH This approach allows for the construction of more diverse and defined libraries. The ability to design and construct synthetic antibody repertoires reflects our increased understanding of antibody/antigen interactions and in particular the contributions to binding of the complementarity determining regions (CDRs). In 1970, the CDRs were predicted by Wu and Kabat (1970) to be the regions responsible for antigen recognition. Numerous crystallographic and protein engineering studies have since supported this hypothesis. In the first synthetic antibody repertoire experiment (Barbas et al., 1992a) a single human anti-tetanus toxoid binding antibody was utilized as the backbone of an antibody library which was varied in sequence only over the heavy-chain CDR3 region. The variants contained an HCDR3 of 16 amino acids which was the length in the parent antibody. This region was completely randomized using a synthetic oligonucleotide with an NNS doping strategy, where N is any of the four nucleotides and S is G or C. NNS encodes all 20 amino acids and a single amber stop codon in a total of 32 codons. The rationale for limiting diversity to this region was based on the observation that this loop contributes most in terms of molecular diversity in the antibody combining site. Genetic considerations in the human system have led to estimates of loi4 naturally occurring sequences in this region (Sanz, 1991). Diversity in this region is not limited to sequence but includes length which may vary from 2 to more than 26 residues (Wu et al., 1993). In this case the extended length of 16 residues was utilized to ensure structural diversity even though such a library would be incomplete as more than lomclones would be required for each possible sequence to be represented. The synthetic genes were constructed by PCR and cloned into pComb3 to yield a library of 5 x lo7clones, a size on the order of the number of B cells in a mouse. The library

253

ANTIBODIES FROM COMBINATORIAL LIBRARIES

of Fab fragments was selected to bind either fluorescein-BSA conjugate or free fluorescein. Selective pressure for binding free fluorescein was provided by eluting bound phage with fluorescein. Seven unique clones from the conjugate selection were obtained. The affinities of these clones were in the lo6M-' range for free flourescein and as high as 2 x 107M-' for the conjugate on which they were selected. Three unique clones were obtained in the selection for binding free fluorescein and the affinities of these were approximately 107M-' for free fluorescein and approximately 5 x lo7A4-l for the conjugate (Fig. 21). The most striking observation was the appearance of a Ser-Arg-Pro sequence near the central position of the HCDR3 of these three clones. Codon usage suggested that this consensus was the result of selection at the level of the protein and not due to bias in the initial oligonucleotide sequence. Another structurally significant observation was the selection of Asp-101 in the HCDR3. In naturally occurring antibodies Asp-101 has been shown to participate in a structurally important salt bridge with Arg-94, the last residue in FR3. Thus, these synthetic antibodies have recapitulated a structurally conserved feature of natuClone

HCDRJ Sequence

K d [FI] M

Kd [FI-BSA] M <1o-5

I x 16'

2 x 10-a

F31 FBSA-8 FBSA-11 FBSA-13

V A S Q V P Q R A K R P M F WD F LA F R LY RKP L PRAGL G L P HG R GWS F T R QA P S

FBSA-48 FBSA-55

L R T LOR SQG R F A F R N A

SARKPFRWSTPFPSTL-

- 1.5 x 1 6 ' 1.61~10~

3 x

rod

5 x

lo-%

1 x lo-' 8 x 10-a 1 lo-'

FIG.21. Amino acid sequences of the heavy-chain CDR3 of a tetanus toxoid binding Fab (startingclone, 7E) and fluorescein binding Fabs derived mutagenesis of H-CDR3 and selection on BSA fluorescein. K i s for binding to fluorescein (FI) were determined by fluorescence quenching and apparent affinities for binding to FI BSA by inhibition ELISA. From Barbas et al., 1992a.

254

DENNIS R. BURTON AND CARLOS F. BARBAS 111

raI antibodies supporting the hypothesis for the role of this conserved sequence. This leads to the suggestion that synthetic antibodies will not be artificial in their character and will be constrained in some extent to sequences found in natural antibodies. Subsequent reports have investigated the extent to which the introduction of structural diversity into the antibody combining site is more important than simple sequence diversity as well as the role of Asp101 in the HCDR3. For practical reasons it is very difficult to prepare libraries containing more than lo9heavy or light chains. This imposes a practical limit to the sequence diversity which may be examined. For example, to survey a library with 99% confidence, a library with six amino acids randomized requires >lo9 clones. Randomization of six residues results in a library diverse in sequences but constrained in structure. The remodeling of a single combining site to bind a variety of antigens might best be achieved by focusing on structural diversity, i.e., randomizing stretches which are longer than can be completely surveyed and display a greater degree of structural freedom. This hypothesis is testable using a competetive selection scheme. Barbas et al. (1993d) utilized three HCDR3 libraries of length 5, 10, and 16. This report focused on developing a strategy for the iterative selection of catalytic antibodies which utilized metal cofactors. Fabs were selected to bind a variety of metal ions by immobilizing the metal with an iminodiacetic acid support. Success of the selection strategy could be verified with some of the metals, such as Cu2+and Zn2+,by knowledge of side chains which have been shown to be important in chelating these metals. The Cu2+and Zn2+selected clones were abundant in His, Cys, Met, and Asp as would be predicted. Binding to free Cu2+was verified by fluorescence quench experiments. The selective panning was also applied to the selection of Fab which could bind the surface of the metal oxide magnetite. This experiment produced four clones which were similar to other loop sequences which had previously been reported to bind magnetite. No clones from the five residue-length HCDR3 library were observed following analysis of 49 selected clones. In another report (Barbas et al., 1993a),libraries of HCDR3, HCDR3 and LCDR3, and LCDR3 alone were examined using a competetive selection scheme for the selection of anti-hapten Fab. In this report, selection for binding three different haptens resulted in 18 antihapten antibodies; additional clones were available but were not sequenced. The specificities of a number of these were examined and are shown in Fig. 22. Various degrees of specificity were observed with binding constants in the range 1-3 x 1O'M-l as determined by surface plasmon resonance. Again no clones from the five-residue HCDR3 were se-

255

ANTIBODIES FROM COMBINATORIAL LIBRARIES

100

' s v)

8

80 60

40

20 0

P2

F22

s2

s4

s10

C15

om rn ""-8 Tetanus toxoid

1

Bovine serum albumin FIG.22. The specificity of Fabs selected from semisynthetic libraries by panning against various haptens. F22 and P2 are Fabs selected with conjugate 1; S2, S4, and S10 selected with conjugate 2; and C15 selected with conjugate 3. From Barbas et al., 1993a).

lected with the competitive selection scheme. Additionally none were selected from the LCDR3 only randomized libraries where a maximum of six residues were randomized. The selected clones were derived mostly from HCDR3/LCDR3 libraries. Furthermore, all clones selected were derived from HCDRS libraries wherein Asp-101 was fixed. These results support the notion that structurally diverse libraries are a requirement for remodeling a given framework. Extrapolation of these results may be used to rationalize the utilization of shorter HCDRS lengths in mice as compared with humans. Structural diversity lacking in HCDR3 of the mouse may be compensated by the use of a far greater number of V segments. Synthetic libraries can be utilized as probes for molecular recognition of antigens from which it is difficult to prepare antibodies for reasons of tolerance, toxicity, or reactivity. To this end, S. M. Barbas et al. (1994) have utilized the aforementioned libraries to study the recognition of double-stranded DNA. Two Fabs were isolated which bound double- and single-stranded DNA with moderate to high affin-

256

DENNIS R. BURTON AND CARLOS F. BARBAS 111

ity. The Fabs did not bind the negatively charged polyelectrolyte dextran sulfate or lipid A. The ability of one of the Fabs to perform as a naturally occurring DNA binding protein was investigated in electrophoretic mobility shift assays. The ability to form nucleoprotein complexes was clearly demonstrated (Fig. 23). The two Fabs isolated contained HCDR3s of 10 and 16 residues in length. Interestingly, although the CDRs differed in length, the amino- and carboxy-ends of the HCDR3 were virtually identical (Fig. 24). The clone with the 16-residue HCDR3 demonstrated a clear preference for poly(dGdC).poly(dGdC),whereas the other clone bound different oligonucleotides with similar affinity (Fig. 25). This observation and the fact that the two clones differ only in the central portion of HCDR3 suggest this region is crucial for the sequence preference displayed by one of the clones and may provide the basis for the design and selection of antibodies capable of sequence-specific recognition. In unpublished studies, these synthetic libraries have further been utilized in the selection of Fabs which bind a number of protein antigens and in one case neutralize HIV. Fabs with catalytic activity may also be selected from the aforementioned synthetic antibody libraries. The direct covalent selection of Fabs from phage display libraries was proposed in the original Fab display report of Barbas et al. (1991). The strategy proposed was to use mechanism-based inhibitors or affinity labels to select for appropriately positioned functionalities within the combining site of an antibody. These functionalities, amino acid side chains with the appropriate chemical characteristics and geometries, would then catalyze a chemical reaction with the appropriate substrate. To this end, Janda et al. (1994) have utilized an pyridyl disulfide affinity label to trap an appropriately positioned thiol in the active site of an synthetic antibody. Phage which covalently bind to the support via a disulfide bond are then selectivly released by reduction of the disulfide following elution with acid of noncovalently attached phage (Fig. 26). Sequencing of 10 of the selected Fabs revealed 2 with unpaired cysteines. One of these was examined for its ability to hydrolyze a thiol ester substrate designed to place the electrophilic center of the carbonyl in the position of original active disulfide ofthe affinity label. The Fab was shown FIG.23. (a) Interaction between the synthetic Fab SD1 and DNA in an electrophoretic mobility shift assay (EMSA). Lanes 1-6 contain 0.08 pmol 32P-labeleddoublestranded probe (5'-AAT-GTA-TGC-GCG-CGC-GCT-TTA-GGC-GCC-CC-3') with 0.2 pmol Fab SD1 (lanes 2, 3, and 4). Lanes 3 and 6 included a preincubation with an anti-Fab reagent (a-Fab IgG). Lanes 4 and 5 included 0.2 pmol of a control Fab reactingwith HIV-1 surface glycoprotein gp120 (HIV Fab). The thick arrow indicates the

257

ANTIBODIES FROM COMBINATORIAL LIBRARIES

a

- - - +

HIVFab a-FablgG SD1 Fab

- - + + - -

- + +

+

+

-0

as-

--+

-

1 2 3 4 5 6

-FP

b 0 0 0 0

0

0

0 0

08RszO88s$

1

2 3 4 5 6 7 8 9 10 11

0

0

0 0 0 0 8 8 8 0 0 0 0 oV)6lV)roV)OIV)r

12 13 14 15 16 17 18 19 20 21 22

position of the specific Fab SDl nucleoprotein complex and the a,arrow indicates the position of the supershifted Fab SD1 nucleoprotein complex. 0 marks the origin of migration and FP the position of the free probe. (b) Specificity of binding of Fab SD1 to DNA. Lanes 1-22 contain 0.08 pmol %P-labeled probe and 0.2 pmol Fab SDl in lanes 2-11 and 13-22, respectively. In addition increasing concentrations of competitor oligonucleotide are included as shown. Numbers refer to molar excess of competitor oligonucleotide relative to the probe. Arrows indicate the position of the Fab nucleoprotein complex. From S. M. Barbas et al. (1994).

258

DENNIS R. BURTON AND CARLOS F. BARBAS I11

Library E

GxxxxxxxxxxxxxDx

QQM;GsRN

Library F

GxxxxxxxDx

QQyI;Gspw

Library ~1 O I E

GxxxxxxxxxxxxxDx

QQYxxxxxxT

Fab SD1

GRAYGGWMSLDN

QwGGsRN

Fab SD2

GRGWSGSLDI

QQyt;Gspw

FIG.24. Amino acid sequences of the variable domain CDR3 regions of the heavy chains (HC) and light chains (LC) from the DNA binding antibodies SD1 and SD2 and the semisynthetic libraries from which they were selected. x refers to a position which was randomized. From S. M. Barbas et al. (1994).

to be catalytic in its hydrolysis of the ester and was futher characterized to proceed through a covalent intermediate. Thus, the catalytic activity had been directly selected to proceed through the mechanism dictated by the design of the affinity label/substrate pair. This is the first report of the selection of a catalytic protein, in this case a Fab, from a random protein library and will likely have a major impact on the field of catalytic antibodies. Garrard and Henner (1993) have constructed a synthetic antibody library with the introduction of limited diversity over four CDRs of a humanized anti-HER-2 Fab. The doping strategy was designed to incorporate mostly amino acids which are naturally found in the CDRs. The library of Fab fragments was selected to bind rsCD4, insulin-like growth factor 1 (IGF-1), and tissue plasminogen activator. A single Fab was isolated which bound IGF-1 with an affinity of 3 X lo5M-’. Two other reports, Hoogenboom and Winter (1992) and Akamatsu et al. (1993), have utilized combinations of genomic V segments with synthetic CDR3 segments as had been previously suggested (Barbas et al., 1992a).The report of Hoogenboom and Winter utilized a defined collection of 49 V, genes and a single light chain. Both groups constructed scFv libraries as opposed to Fab as discussed above. The former report combined a HCDR3 segment randomized over five amino acids with V, genes and selected the library to bind phOX-BSA, 3-iodo-4-hydroxy-5-nitrophenylacetate-BSA (NIP-BSA), BSA, turkey egg white lysozyme, TNF-a, and human thyroglobulin. Several scFv’s were shown to bind phOX and NIP while only a single clone was found to bind weakly to one of the protein antigens, TNF-a. The affinities of the phOX and NIP binding scFv’s were reported to be 105-106 M - 1 . Akamatsu et al. introduced,a biased set of amino acids over portions

259

ANTIBODIES FROM COMBINATORIAL LIBRARIES

a 3

2 In 0 Q

a 1

0

-1

0

1

log Fab conc. (Ug/ml)

.

-

poly(dA).poly(dT) poly(dAdT).poly(dAdT) poly(dG).poly(dC) poly(dGdC).poly(dGdC)

2

260

DENNIS R. BURTON AND CARLOS F. BARBAS 111

FIG.26. Scheme illustrating the selection of Fab fragments containing an unpaired cysteine from a phage display library. From Janda et al. (1994).Nonbinders are eluted in detergent and noncovalent binders in acid before covalent binders (via a disulfide linkage) are eluted with dithiothreitol (DTT). Panning steps are repeated as in Fig. 7.

of both HCDRS and LCDRS to a set of genomic V genes. The library was selected to bind ConA. Six scFv's were characterized to bind ConA with affinities of 5 X 104-105M-'. The ConA scFv's demonstrated good specificity in binding the target antigen. Targeted CDR mutagenesis can also be utilized to make very modest specificity changes in an antibody. An example of altering the reactivity of an antibody to its anti-Ids has been reported by Glaser et al. (1992). In this report LCDR 1 and 2 were targeted for mutagenesis using a minimal doping scheme called codon-based mutagenesis. This restricted doping scheme employs oligonucleotide synthesis using two columns and the opening, mixing an exchange of resin between the columns during synthesis. This synthesis scheme was necessitated by the use of screening rather than selection in the examination of the library. The system utilized for library screening has been discussed (vida supra, Huse et al., 1992).Screening required that a larger number of positive clones would be represented in the library as only 1000clones would be screened for reactivity. The goal was to moderate the binding of a number of anti-Ids which had been previously characterized as reacting with the LCDRs and the experiment was successful in this regard.

C. COMBINING DESIGNAND SELECTION IN SYNTHETIC HUMAN ANTIBODIES An alternative to a random search of binding sites for the desired specificity is a directed search that incorporates information known to be relevant to a given binding problem. For anti-receptor antibodies

ANTIBODIES FROM COMBINATORIAL LIBRARIES

261

information relevant to the binding of the ligand to the receptor may be utilized to construct specialized libraries directed to bind a given receptor or receptor family. The minimal binding sequences of a number of ligands for their receptors have been characterized. Barbas et al. (1993~) described a methodology which could allow for the direct design and selection of human antibodies reactive with human receptors. The overall strategy is outlined in Fig. 27. A minimal sequence known to be important for the binding of the ligand to the receptor is transplanted into an antibody CDR. Since the correct conformational display of the sequence will, in almost all cases, be critical for interaction with the receptor, the transplanted sequence is flanked by randomized segments which form the elements of diversity within the library. The randomized elements also allow for the selection of additional contact residues. As a demonstration of the validity of this approach a member of the Asp-Gly-Arg (RGD)binding integrin receptor family, aVp3was chosen as a target receptor. The simple tripeptide RGD formed the basis for the construction of the library. In this case, the RGD sequence was placed near the apex of the extended hairpin loop of the HCDR3 of a human anti-HIV Fab. The HCDR3 had a moderately long sequence, 18 residues. Residues at the N- and C-terminal of the CDR were conserved to retain the stem of the loop. On each flank of the RGD were three randomized residues. Following selection of the 3 X lo7 member library for binding to avp3 5 Fab clones were characterized. The selected flanking regions showed some homology with known integrin binding peptides; however, some elements were quite unique. The affinities of the Fabs were astonishingly high, 10" it4-' (Table VI). Specificity characterization revealed the Fabs also bound aI&3 but not avp5, two integrins highly related to a,p3. These three integrins can share the same high-affinity ligand vitronectin and all bind RGD containing peptides. Furthermore, aVp3and aVp5bind with the same affinity to some RGD containing peptides. Fabs were shown to compete with RGD peptides for the integrin ligand binding site as designed. Further functional characterization demonstrated the Fabs to be potent in adhesion assays and subsequently (Smith et al., 1994) in inhibiting platelet aggregation (Fig. 28). Collectively, antagonism of these integrins could have potential in the treatment of osteoporosis and as anti-metastatic and anti-thrombotic agents. Synthetic antibodies can also be utilized to derive novel minimal ligands for receptors. Smith et al. (1994) demonstrate that randomization of the RGDX sequence within the context of the optimized Fab-9 (Barbas et al., 1993c) architecture and reselection for aII& binding leads to the identification of a number of novel non-RGD ligand sequences within the HCDR3. This experiment led to the sug-

262

DENNIS R. BURTON AND CARLOS F. BARBAS I11

IQ

Receptor

L

/ptitnize

Minimal Ligand

Ligand in an Antibody

FIG.27. The design and selection of human anti-receptor antibodies is aided by a knowledge of residues within the ligand that are involved in binding to the receptor. The first step involves the characterization of this interaction by mapping regions of the ligand involved in binding or by selecting peptides from linear or constrained peptide libraries to bind the receptor. The minimal ligand is then transplanted into a semisynthetic antibody library. The random sequence is provided to optimize the conformational display of the transplanted sequence which will be dependent on constaints imposed by the antibody. Random sequence also allows for selection ofadditional contact residues. Selection horn a vast library of variants provides the optimal antireceptor antibody. For integrins which bind RGD containing peptides, the simple RGD sequence is transplanted. For this case we have chosen the heavy-chain CDR3 for insertion of the sequence and optimization. The sequence of the starting a n t i - e l 2 0 antibody HCDR3 was VGPYSWDDSPQDNYYMDV. Following transplantation into a synthetic version of HCDRB the library consisted of variants where the HCDR3 sequences are VGCXXXRGDXXXCYYMDV,where X represents a mixture ofall 20 amino acids. Following selection for binding to the integrin a& the sequence of the HCDR3 of the highest A n i t y antibody, Fab-9, was VGCSFGRGDIRNCYYMDV. From Barbas et al. (1993~).

gestion of a new minimal sequence consensus for a I I b p 3 , namely, R/K-X-D, where Xis virtually any amino acid except proline. Several of these novel non-RGD Fabs demonstrated the capacity to discriminate between avp3and a I I b p 3 with a maximal difference of 100-fold preference (affinity) for aIIbp3.Prospects for the use of these antibodies as lead compounds were examined through the characterization of HCDR3 peptides. Peptides were shown to maintain receptor specificity albeit at reduced affinity.

D. IMPROVING THE AFFINITYOF HUMAN ANTIBODIES The systems and mutagenesis strategies described for producing new antibodies can also be utilized to improve the properties of ex-

ANTIBODIES FROM COMBINATORIAL LIBRARIES

263

TABLE VI AND AMINOACID SEQUENCES OF RGD CONTAINING SYNTHETIC INHIBITION CONSTANTS HUMAN ANTIBODIES ICa (Molar) Antibody No. 4

7

8 9 10

d 3 3

2.5 X 2.0 x 2.0 x 1.0 x 2.5 x

10-l0 10-10 10-10 10-lo

%BP3

2.5 X 5.0 x 1Olo 3.5 x 10-lo 1.0 x 10-lo 2.5 x 10-lo

4

5

5X NI NI NI NI

lo-'

Sequence TQG-RGD-WRS TYG-RGD-TRN PIP-RGD-WRE SFG-RGD-IRN TWG-RGD-ERN

Note. The ability of the recombinant antibodies to block vitronectin binding to avb3 and avb5 and fibrinogen binding to was determined with a purified receptor binding assay. The sequences flanking the RGD motif of each antibody are shown. NI, no inhibition at concentrations of up to 5 x I@'M. From Barbas et al. (1993~).

isting antibodies. Several strategies have been applied to improve the affinity of antibody fragments. For all of these studies it is important to consider the affinity range of the antibody to be improved. It is expected, and indeed observed, that greater relative improvements are more readily obtained with low-affinity antibodies as starting points. High-affinity antibodies require more modest adjustments of an already tight antigedantibody interface and are therefore less readily improved.

20xl09M JOxlO@M

Fab-9 FIG.28. The synthetic human antibody fragment Fab-9 blocks platelet aggregation. Human platelets (1 x 108) were mixed with 100 pg/ml of purified fibrinogen and 2.0 mM Ca2' in Tyrodes buffer. These were placed in a glass aggregation tube. The indicated concentration of Fab was added and the platelets were stimulated with 20 p M ADP. Aggregation was measured as light transmission through the platelet suspension using an aggregometer. From Smith et al. (1994).

264

DENNIS R. BURTON AND CARLOS F. BARBAS I11

Gram et al. (1992) have utilized error-prone PCR conditions to construct libraries of point mutations. This strategy is interesting since it mimics the random point mutations generated in natural somatic mutation. This approach was utilized to improve a Iow-affinity mouse scFv selected from a naive library to bind progesterone. The monovalent display system pComb3 allowed for the selection of an scFv with 30-fold improved affinity. The final scFv had an affinity of ca. lo6M-’. Hawkins et al. (1992) have utilized a similar strategy to improve a mouse anti-NIP scFv. In this case the multivalent display system fdtet-DOG1 was utilized. In order to select for higher affinity variants from this multivalent display system, the antigen was biotinylated and the selection performed in solution. Following several rounds of selection a clone was isolated which was improved fourfold in NIP binding to a final affinity of lo8M’. In a subsequent report, Hawkins et al. (1993) have utilized the same strategy and vector with a highaffinity mouse anti-lysozyme scFv the structure of which had been previously determined. The library was selected in two different fashions. The first experiment entailed 13rounds ofbiotin capture panning. The second experiment was performed in a competitive selection strategy where biotin-captured lysozyme was used with nonlabeled lysozyme as a competitor for five rounds of selection. The second selection was performed to bias the selection for the capture of clones with slower off-rates. These two experiments yielded two clones which were improved threefold in affinity. The two sets of mutations found in these clones were combined by site-directed mutagenesis to yield an scFv with an affinity of 2 x lo9M-‘,a fivefold overall improvement in affinity. One of the potential problems with random mutagenesis of the entire gene is that many of the mutations occur outside the CDRs and thus might more easily generate an antigenic antibody. Chain shuffling as proposed by Huse et al. (1989)may be used with great success when starting with a low-affinity clone (Marks et al., 1992). Starting with low-affinity clone and shuffling against a library of light chains allowed for the selection of a clone with 20-fold improved affinity. This clone was then combined with a V, library and selected for an additional improvement of 18fold. The final clone had an affinity of 109M-’ for phOX. This approach may be of lesser utility for the improvement of the high-affinity antibodies (Barbas et al., 1993b) as the sequence changes are too great. A final approach termed “CDR walking” (C. F. Barbas et aZ., 1994) may prove to be a general method for the improvement of antibody affinity. This approach is a variant of the synthetic antibody approach with one important difference. In this case library completeness is stressed over structural diversity and randomization is limited to six

ANTIBODIES FROM COMBINATORIAL LIBRARIES

265

residues or less. The approach may be applied in a parallel fashion where libraries are constructed in different CDRs. Following selection, the improved CDRs are then assembled to give the best antibodies, assuming additivity. The approach may also be applied in a sequential fashion. C. F. Barbas et al. (1994) applied the sequential approach for the improvement of a human anti-HIV-1 gp120 Fab. The Fab chosen (designated here as HIV-4 has an identical sequence to Fab b12 discussed above) had already been shown to potently neutralize HIV1.In this case a library of HCDRl variants with five residues randomized was selected by four rounds ofpanning against rgpl20. The collection of clones which resulted were then used in the construction of a HCDRS library where four additional residues were randomized. The resulting library was selected by an additional six rounds of panning. A clone was isolated with an 8-fold improvement of affinity for a final affinity of 1.4 x 1O’M-l. The epitope recognized by this antibody is the CDCbinding site on gp120. As this functional feature, which is not formed by a linear sequence, is retained by all variants of HIV-1, it was proposed that a f h i t y should also be increased for divergent isolates. This was tested by determining the affinities of several clones 15

EL g 5

h

5-

!2

3B4 HIV4 0

.:I :;; 5

0

2 4 6 8 1 0 1 2 1 4 K. (MN) [wl] (x 10’) CDRl CDRS

383

.-

3B9

384 HIV-4

-

NFTVH NFTLM NFTVH NYTLl NFV I H

EWGW QWNW PWRW PWNW P Y SW

MAA A6AA A4AA A6AA

FIG.29. The affinity increases of evolving Fabs for binding the divergent envelope proteins gp120 IIIB (LAI)andgpl20 MN are well-correlated.Affinities were determined using the surface plasmon resonance technique. The sequences of evolved clones are ranked as compared to the parent and changes in the amino acid sequence from the parent are shown as A M . The complete VH sequence of HIV-4 is given in Fig. 15 as Fab b4 and is identical with Fab b12. From C. F. Barbas et al. (1994).

266

DENNIS R. BURTON AND CARLOS F. BARBAS Ill

for gp120 derived from MN and IIIB strains. These proteins differ in over 80 amino acids and are highly divergent. As shown in Fig. 29 the affinity increases for both gpl20’s are well correlated. Functional improvement was assayed in virus neutralization studies with laboratory isolates and a 54-fold improvement was determined for the highest affinity clone. Furthermore, neutralization studies with primary clinical isolates demonstrated that the improved Fab acquired the ability to neutralize additional variants not neutralized by the parent. VI. Conclusions

Since its conception 5 years ago, the combinatorial approach has allowed unprecedented access to the human antibody response. The cloning of antibodies from preimmune, immune, and memory compartments ofthe human immune system has been demonstrated. Combinatorial antibodies have been shown to provide an accurate functional reflection of the natural response as demonstrated by the ability of cloned antibodies to compete with serum antibodies for binding antigens. Combinatorial antibodies also provide a useful (if somewhat incomplete) guide to the molecular biology of the response. The ability to select large numbers of antibodies against the same antigen and even the same epitope has led to a greater understanding ofthe molecular and structural biology ofthe immune response. Examination of natural libraries has highlighted the functional significance of HCDR3 and its prominent role in recognition. This observation coupled with the observation of light-chain plasticity characterizes an immune response which weights HCDR3 diversity over other elements. The ability to select antibodies to a multiplicity of antigens from synthetic libraries which differ only in the HCDR3 supports this notion. Furthermore competitive selections with synthetic libraries demonstrate that limited structural diversity within the rest of the binding site can be compensated for by increased structural diversity within the HCDR3. This observation may be important in rationalizing the differences between murine and human repertoires as reflected in HCDR3 length and the number of expressed V genes. These studies complement the recent increased appreciation of the role of HCDR3 gained through structural studies (Wilson and Stanfield, 1993). With its ability to provide large numbers of human antibodies directed against a single antigen, the combinatorial approach allows for the rapid assessment of immunodominant as well as neutralizing epitopes in the context ofthe human response. This information should be utilized in the future to guide the design of more effective vaccines. Antibodies neutralize viruses by mechanisms which in most cases

ANTIBODIES FROM COMBINATORIAL LIBRARIES

267

are not yet defined. Mechanistic investigation of anti-viral antibodies should allow for the elucidation of novel pathways which might be targeted by small-molecule pharmaceuticals. Thus we suggest that combinatorial antibodies will play a significant role in the design of vaccines and new anti-viral agents. Antibodies may also provide a way of determining receptor function in vivo and serve as templates for the design of small molecules. One hundred years ago von Behring suggested, “Considering that antitoxin is an inanimate chemical substance, the possibility cannot be discounted that it may, at a later date, be able to be produced without the aid of an animal body,” (as translated in Gronski et al., 1991). Indeed this has now been realized. Our new-found ability to generate human antibodies and to evolve their specificities and affinities ex vivo promises increased use of this class of molecules in the service of human health. ACKNOWLEDGMENTS We acknowledge the enthusiasm and support of Richard Lemer for our efforts in the combinatorial antibody area. We are grateful to Gregg Silverman for a critical review of the manuscript and to members of our laboratories for valuable comments and suggestions including Shana Barbas, James Binley, Henrik Ditzel, Paul Parren, Jonathan Rosenblum, Pietro Sanna, and Anthony Williamson. We thank Joanne Marshall for help in preparing the manuscript. Some of our work described is supported by NIH (A133292 and A135165) and by Johnson and Johnson. C.F.B. is a Scholar of the American Foundation for AIDS Research and the recipient of an Investigator Award from the Cancer Research Institute.

REFERENCES Akamatsu, Y., Cole, M.S., Tso, J. Y., and Tsurushita, N. (1993). Construction o f a human Ig combinatorial library from genomic V segments and synthetic CDR3 fragments. J. Immunol. 151,4651-4659. Amadori, A. (1993). AIDS and autoimmunity. I n “The Molecular Pathology of Autoimmune Diseases” (C. A. Bona, K. A. Siminovitch, M. Zanetti, and A. N.Theofilopoulos, eds.), pp. 727-750. Hanvood Academic Publishers, Chur, Switzerland. Aulitzky, W. E., Schulz, T. F., Tilig, H., Niederwieser, D., Larcher, K., Ostberg, L., Scriba, M., Martindale, J., Stern, A. C., Grass, P., Mach, M., Dierich, M. P., and Huber, C. (1991). Human monoclonal antibodies neutralizing cytomegalovirus (CMV) for prophylaxis of CMV disease:Report of a phase I trial in bone marrow transplant recipients. J. Infect. Dis. 163, 1344-1347. Balachandran, N., Bacchetti, S., and Rawls, W. E. (1982). Protection against lethal challenge of Balb/c mice by passive transfer of monoclonal antibodies to five glycoproteins of herpes simplex virus type 2. Infect. Imrnun. 37,1131-1137. Balfour, H. H. (1988).Varicella zoster virus infection immunocompromised hosts. Am. J . Med. 85,68-73. Barbas 111, C. F. (1993). Recent advances in phage display. Cum. Opinion Biotechnol. 4,526-530. Barbas 111, C. F. (1994). The combinatorial approach to human antibodies: The pharmacology of monoclonal antibodies. Handb. Erp. Phannacol. 13, in press.

268

DENNIS R. BURTON AND CARLOS F. BARBAS 111

Barbas 111, C. F., and Lerner, R A. (1991). Combinatorial immunoglobulin libraries on the surface of phage (Phabs): Rapid selection of antigen-specific Fabs. In “Methods: A Companion to Methods in Enzymology” (R. A. Lerner and D. R. Burton, eds.), Vol. 2, pp. 119-124. Academic Press, Orlando. Barbas 111, C. F., Kang, A. S., Lerner, R. A., and Benkovic, S . J. (1991). Assembly of combinatorial antibody libraries on phage surfaces: The gene 111 site. Proc. Natl. A d . Sci. U.S.A.88, 7978-7982. Barbas 111, C. F., Bain, J. D., Hoekstra, D. M., and Lerner, R. A. (1992a).Semisynthetic combinatorial antibody libraries: A chemical solution to the diversity problem. Proc. Natl. Acad. Sci. U S A . 89,4457-4461. Barbas 111, C. F., Bjorling, E.,Chiodi, F., Dunlop, N., Cababa, D., Jones, T. M., Zebedee, S. L., Persson, M. A. A,, Nara, P. L., Norrby, E.,andBurton, D. R. (199213).Recombinant human Fab fragments neutralize human type 1immunodeficiencyvirus in vitro. Proc. Natl. A d . Sci. U.S.A. 89, 9339-9343. Barbas 111, C. F., Crowe, J. E., Jr., Cababa, D., Jones, T. M., Zebedee, S. L., Murphy, B. R., Chanock, R. M., and Burton, D. R. (1992~).Human monoclonal Fab fragments derived from a combinatorial library bind to respiratory syncytial virus F glycoprotein and neutralize infectivity. Proc. Natl. Acad. Sci U.S.A.89, 10,164-10,168. Barbas 111, C. F., Amberg, W., Simoncsits, A., Jones, T. M., and Lemer, R. A. (1993a). Selection of human anti-hapten antibodies from semisynthetic libraries. Gene 137, 57-62. Barbas 111, C. F., Collet, T. A., Amberg, W., Roben, P., Binley, J. M., Hoekstra, D., Cababa, D., Jones, T. M., Williamson, R. A., Pilkington, G. R., Haigwood, N. L., Satterthwait, A. C., Sanz, I., and Burton, D. R. (1993b).Molecular profile ofan antibody response to HIV-1 as probed by combinatorial libraries. J . Mol. Biol. 230,812-823. Barbas 111, C. F., Languino, L. R., and Smith, J. W. (1993~).High-affinity self-reactive human antibodies by design and selection: Targetingthe integrin ligand binding site. Proc. Natl. Acad. Sci. U.S.A.90, 10,003-10,007. Barbas 111, C. F., Rosenblum, J. S., and Lemer, R. A. (1993d). Direct selection of antibodies which coordinate metals from semisynthetic combinatorial libraries. Proc. Natl. Acad. Sci. U.S.A. 14,6385-6389. Barbas 111, C. F., Hu, D., Dunlop, N., Sawyers, L., Cababa, D., Hendry, R. M., Nara, P. L., and Burton, D. R. (1994a).In vitro evolution of a neutralizing human antibody to HIV-1 to enhance affinity and broaden strain cross-reactivity. Proc. Natl. Acad. Sci. U.S.A., 91, 3809-3813. Barbas, S. M., Ghazal, P., Barbas 111, C. F., and Burton, D. R. (1994b). Recognition of DNA by synthetic antibodies. J. Am.Chem. SOC. 116,2161-2162. Bass, S., Greene, R., and Wells, J. A. (1990).Hormone phage: An enrichment method for variant proteins with altered binding properties. Proteins Struct. Funct. Genet. 8,309-314. Beasley, R. P., Hwang, L. Y.,Stevens, C. E., Lin, C. C., Hsieh, F. J., Wang, K. Y.,Sun, T. S., and Szmuness,W. (1983).Efficacy of hepatitis B immune globulin for prevention of perinatal transmission of the hepatitis B virus carrier state: Final report of a randomized double-blind, placebo-controlled trial. Hepatology 3, 135-141. Bebbington, C. R. (1991). Expression of antibody genes in nonlymphoid mammalian cells. In “Methods: A Companion to Methods in Enzymology” (R. A. Lerner and D. R. Burton, eds.), Vol. 2, pp. 136-145, Academic Press, Orlando. Behring, E. A. (1893). “Die Gesehichte der Diphtherie,” p. 186. Thieme, Leipzig. Behring, E. A. (1894).“Das neue Diphtherieheilmittel,” p. 40. 0. Hering, Berlin. Bender, E., Woof, J. M., Atkin, M. D., Barker, M. D., Bebbington, C. R., and Burton, D. R. (1993). Recombinant human antibodies: Linkage of an Fab fragment from a

ANTIBODIES FROM COMBINATORIAL LIBRARIES

269

combinatorial library to an Fc fragment for expression in mammalian cell culture. Hum. Antibod. Hybridomas 4,74-79. Bender, E., Pilkington, G. R., and Burton, D. R. (1994). Human monoclonal Fab fragments from a combinatorial library prepared from an individual with a low serum titer to a virus. Hum. Antibod. Hybridomas, in press. Berberian, L., Goodglick, L., Kipps, T. J., and Braun, J. (1993). Immunoglobulin VH3 gene products: Natural ligands for HIV gp120. Science 261, 1588-1591. Better, M., Chang, C. P., Robinson, R. R. et al. (1988).Escherichia co2i secretion of an active chimeric antibody fragment. Science 240, 1041-1043. Bigazzi, P. E. (1993).Autoimmunity in Hashimoto’s disease. In “The Molecular Pathology of Autoimmune Diseases” (C. A. Bona, K. A. Siminovitch, M. Zanetti, and A. N. Theofilopoulos, eds.), pp. 493-510. Hanvood Academic Publishers, Chur, Switzerland. Binley, J. M., Ditzel, H. J., Hendry, M., Sawyer, L., Dunlop, N., Nara, P. L., Barbas 111, C. F., and Burton, D. R. (1994). Generation of neutralizing human antibodies reactive with conformational epitopes of gp41, the transmembrane glycoprotein of HIV-1. Submitted for publication. Bird, R. E., and Walker, B. W. (1991). Single chain antibody variable regions. Trends Biotechnol. 9, 132-137. Boulianne, G. L., Hozumi, N., and Schulman, M. J. (1984). Production of functional chimaeric mouse/human antibody region domains. Nature (London)312,643-646. Britt, W . J., Vugler, L., Buffiloski, E. J.,and Stephens, E. B. (1990).Cell surface expression of human cytomegalovirus (HCMV) gp55-116(gB): Use of HCMV-recombinant vaccinia virus-infected cells in analysis of the human neutralizing antibody response. J . Virol. 64, 1079-1085. Burioni, R., Williamson, R. A., Sanna, P. P., Bloom, F. E., and Burton, D. R. (1994). Recombinant human Fab to glycoprotein D neutralizes infectivity and prevents cellto-cell transmission of herpes simplex virus type-1 and -2 in uitro. Proc. Natl. Acad. Sci. U.S.A.91,355-359. Burton, D. R. (1990). The conformation of antibodies. In “Fc receptors and the action of antibodies” (H. Metzger, ed.), pp. 31-54. American Society of Microbiology,Washington, DC. Burton, D. R. (1991). Human and mouse monoclonal antibodies by repertoire cloning. Trends Biotechnol. 9, 169-175. Burton, D. R. (1992).Human monoclonal antibodies: Achievement and potential. Hospital Practice. 27, 67-74. Burton, D. R. (1993a). Human monoclonal antibodies to viruses from combinatorial libraries. In “Vaccines 93,” pp. 1-5. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Burton, D. R. (1993b). Monoclonal antibodies from combinatorial libraries. Acc. Chem. Res. 26,405-411. Burton, D. R.,and Barbas 111, C. F. (1992). Antibodies from libraries. Nature (London) 359,782-783. Burton, D. R., and Barbas 111, C. F. (1993a). Human antibodies from combinatorial libraries. In “Protein Engineering of Antibody Molecules for Prophylactic and Therapeutic Applications in Man” (M. Clark, ed.), pp. 65-82, Academic Titles, Nottingham, England. Burton, D. R., and Barbas 111, C. F. (1993b).Monoclonal Fab fragments from combinatorial libraries displayed on the surface of phage. Zmmunomethods 3, 1-9. Burton, D. R., and Woof, J. M. (1992).Human antibody effector function. Adu. Zmmunol. 51, 1-86.

270

DENNIS R. BURTON AND CARLOS F. BARBAS 111

Burton, D. R., Barbas, C. F., Persson, M. A. A., Koenig, S., Chanock, R. M., and Lerner, R. A. (1991).A large array ofhuman monoclonalantibodies to type 1 human immunodeficiency virus from combinatorial libraries of asymptomactic seropositive individuals. Proc. Natl. Acad. Sci. U.S.A.88, 10,134-10,137. Casali, P., and Notkins, A. L. (1989). Probing the human B-cell repertoire with EBV: Polyreactive antibodies and CD5+ B lymphocytes. Annu. Reu. Immunol. 7,512-536. Caton, A. J.. and Koprowski, H. (1990).Influenza virus hemagglutinin-specificantibodies isolated from a combinatorial expression library are closely related to the immune response of the donor. Proc. Natl. Acad. Sci. U.S.A.87,6450-6454. Cha, S., Leung, P. S. C., Gershwin, M. E., Fletcher, M. P., Ansari, A. A., and Coppel, R. L. (1993). Combinatorial autoantibodies to dihydrolipoamide acetyltransferase, the major autoantigen of primary biliary cirrhosis. Pro. Natl. Acad. Sci. U.S.A. 90, 2527-2531. Chang, C. N., Landolfi, N. F., and Queen, C. (1991).Expression ofantibody Fab domains on bacteriophage surfaces: Potential use for antibody selection. J . Immunol. 147, 3610-36 14. Chanock, R. M., Crowe, J. E., Jr., Murphy, B. R., and Burton, D. R. (1993). Human monoclonal antibody Fab fragments cloned from combinatorial libraries: Potential usefulness in prevention and/or treatment of major human viral disease. Infect. Agents Dis. 2, 118-131. Chazenbalk, G. D., Portolano, S., Russo, D., Hutchison, J. S., Rapoport, B., and McLachIan, S. (1993). Human organ-specific autoimmune disease: Molecular cloning and expression of an autoantibody gene repertoire for a major autoantigen reveals an antigenic immunodominant region and restricted immunoglobulin gene usage in the target organ. J . Clin. Inuest. 92, 62-74. Chetrit, E. B., Dunsky, E. H., Wollner, S., and Eilat, D. (1985).In vivo clearance and tissue uptake of an anti-DNA monoclonal antiody and its complexes with DNA. Clin. Exp. Immunol. SO, 159-168. Cheung, S. C., Dietzschold, B., Koprowski, H., Notkins, A. L., and Rando, R. F. (1992). A recombinant human Fab expressed in Escherichia coli neutralizes rabies virus. J . Virol. 66,6714-6720. Clackson, T., Hoogenboom, H. R., Griffiths, A. D., and Winter, G. (1991). Making antibody fragments using phage display libraries. Nature (London) 352, 624628. Collet, T. A., Roben, P., O’Kennedy, R., Barbas, C. F., Burton, D. R., and Lerner, R. A. (1992).A binary plasmid system for shuffling combinatorial antibody libraries. P~ocN . d . Acad. Sci. U.S.A. 89, 10,026-10,030, Constante, G., Portolano, S., Nishikawa, T., Jaume, J. C., Chazenblak, G. D., Rapoport, B., and McLachlan,S. M. (1994).Recombinantthyroid peroxidase-specificautoantibodies. 11. Role of individual heavy and light chains in determining epitope recognition. Endocrinology. jp press. Corey, L. (1982). Diagnosis and treatment of genital herpes. J . Am. Med. Assoc. 248, 1041- 1049. Corey, L., Adams, H. G., Brown, Z. A., and Holmes, K. K. (1983).Genital herpes simplex virus infections, clinical manifestations, course and complications. Ann. Intern. Med. 98,958-972. Crowe, J. E., Jr, Murphy, B. R., Chanock, R. M., Williamson, R. A., Barbas 111, C. F., and Burton, D. R. (1994). Recombinant human RSV monoclonal antibody Fab is effective therapeutically when introduced directly into the lungs of respiratory syncytial virus infected mice. Proc. Natl. Acad. Sci. U.S.A. 91, 1386-1390.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

271

Current, W. L., and Bick, P. H. (1989).Immunobiology of Cryptosporidium spp. Pathol. Imunopathol. Res. 8, 141-160. Cwirla, S. E., Peters, E. A., Barrett, R. W., and Dower, E. J. (1990).Peptides on phage: A vast library of peptides for identifying ligands. Proc. Natl. Acad. Sci. U.S.A. 87, 6378-6382. Devlin, J. J., Panganiban, L. C., and Devlin, P. E. (1990).Random peptide libraries: A source of specific protein binding molecules. Science 249,404-406. Ditzel, H. J., Barbas, S. M., Barbas 111, C. F., and Burton, D. R. (1994).The nature of the autoimmune antibody repertoire in HIV-1 infection. Proc. Natl. Acad. Sci. U.S.A., 91,3710-3714. Dix, R. D., Pereira, L., and Baringer, J. R. (1981).Use of monoclonal antibody directed against herpes simplex virus glycoproteins to protect mice against acute virus-induced neurological disease. Infect. Immun. 34, 192-199. D'Souza, M. P., Geyer, S . J., Hanson, C. V., Hendry, R. M., Milman, G., and collaborating investigators. (1994). Evaluation of monoclonal antibodies to HIV-1 envelope by neutralization and binding assays: An international collaboration. AIDS 8, 169-181. Dube, M. P., and Sattler, F. R. (1993). Prevention and treatment of opportunisitic infections. Curr. Opinion Infect. Dis. 6,230-236. Dubel, S., Breitling, F., Fuchs, P., Braunagel, M., Klewinghaus, I., and Little, M. (1993). A family ofvectors for surface display and production ofantibodies. Gene 128,97-101. Duchosal, M. A., Eming, S., Fischer, P., Leturcq, D., Barbas 111, C. F., McConahey, P. J., Caothien, R. H., Thornton, G. B., Dixon, F. J., and Burton, D. R. (1992).Immunization of hu-PBL-SCID mice and the rescue of human monoclonal Fab fragments through combinatorial libraries. Nature (London)355,258-262. Ehrlich, P. H., Moustafa, Z. A., Harfeldt, K. E., Isaacson, C., and Ostberg, L. (1990). Potential ofprimate monoclonal antibodies to substitute for human antibodies: Nucleotide sequence of chimpanzee Fab fragments. Hum. Antibod. Hybridomas 1,23-26. Ehrlich, P. H., Moustafa, Z. A., and Osterberg, L. (1991).Nucleotide sequence of chimpanzee Fc and hinge regions. Mol. Immuonol. 28,319-322. Eichelberger, M., Allan, W., Zijlstra, M., Jaenisch, R., and Doherty, P. C. J. (1991). Clearance of influenza virus respiratory infection in mice lacking class 1 MHCrestricted CD8' T cells. J. Exp. Med. 174, 875-880. Embleton, M. J., Gorochov, G., Jones, P. T., and Winter, G. (1992).In-cell PCr from mRNA: Amplifying and linking the rearranged immunoglobulin heavy and light chain V-genes within single cells. Nucleic Acids Res. 20,3831-3837. Enria, D. A., Femandez, N. J,, Briggler, A. M., Levis, S. C., and Maiztegni, J. I. (1984). Importance of dose of neutralizing antibodies in treatment of Argentine haemorrhagic fever with immune plasma. Lancet 4,255-256. Ershler, W. B., Moore, A. L., and Hacker, M. P. (1982).Specific in vivo and in vitro antibody response to tetanus toxoid immunization. Clin. Exp. Immunol. 49,552-558. Forghani, B., Dupuis, K.W., and Schmidt, N. J. (1984).Varicella-zoster viral glycoproteins analyzed with monoclonal antibodies. J. Virol. 55, 55-62. Fukuma, N., Sarsero, D., Furmaniak, J., Pegg, C. A. S., McLachlan, S. M., and ReesSmith, B. (1990).B and T cell epitopes on thyroid peroxidase (P. Carayon and J. Ruf, eds.), pp. 195-201. Colloque INSERM/John Libbey Eurotext Ltd. Fulpius, T., Spertini, F., Reininger, L., and Izui, S. (1993).Immunoglobulin heavy chain constant region determines the pathogenicity and the antigen-binding activity of rheumatoid factor. Proc. Natl. Acad. Sci. U.S.A. 90,2345-2349. Furmaniak, J., and Rees-Smith, B. (1990).The structure of thyroid autoantigens. Autoimmunity 7 , 63-80.

272

DENNIS R. BURTON AND CARLOS F. BARBAS 111

Garrard, L. J., and Henner, D. J. (1993). Selection of an anti-IGF-1 Fab from a Fab phage library created by mutagenesis of multiple CDR loops. Gene 128, 103-109. Garrard, L. J., Yang, M., O’Connell, M. P., Kelley, R. F., and Henner, D. J. (1991). Fab assembly and enrichment in a monovalent phage display system. Biotechnology 9, 1373-1377. Gherardi, E., and Milstein, C. (1992).Original and artificial antibodies. Nature (London) 357,201-202. Glaser, S. M., Yelton, D. E., and Huse, W. D. (1992). Antibody engineering by codonbased mutagenesis in a filamentous phage vector system. J. of Immunol. 149, 3903-3913. Gram, H., Marconi, L.-A., Barbas 111, C. F. (1992). In vitro selection and affinity maturation of antibodies from a naive combinatorial immunoglobulin library. Proc. Nutl. Acad. Sci. USA. 89,3576-3580. Griffiths, A. D., Malmqvist, M., Marks, J. D., Bye, J. M., Embleton, M. J., McCafferty, J., Baier, M., Holliger, K. P., Gorick, B. D., Hughes-Jones, N. C., Hoogenboom, H. R., and Winter, G. (1993). Human anti-self antibodies with high specificity from phage display libraries. EMBO J. 12,725-734. Gronski, P., Seiler, F. R., and Schwick, H. G .(1991).Discoveryofantitoxins anddevelopment of antibody preparations for clinical uses from 1890 to 1990. Mol. Immunol. 28, 1321-1332. Groothuis, J. R., Simoes, E. A. F., Levin, M. J., et al. and the RSVIG Study Group (1993).Respiratory syncytial virus (RSV)immune globulin (IG) prevents severe lower respiratory infection (LRI) in high risk children. Pediatr. Res. 33, 169A. Grose, C., Edwards, D. P., Friedrich, W. E., Weigle, K. A., and McGuire, W. L. (1983). Monoclonal antibodies against three major glycoproteins of varicella-zoster virus. Infect. Immun. 40,381-388. Hammon, W. M., Coriell, L. L., Wehrie, P. F., Stokes, J., Jr. (1953). Evaluation of Red Cross gamma globulin as a prophylactic agent of poliomyelitis: Final report of results based on clinical diagnoses. J. Am. Med. Assoc. 151, 1272-1285. Hanson, C. V., Crawford-Miksza, L., and Sheppard, H. W. (1990).Application of a rapid micrcplaque assay for determination of human immunodeficiency virus neturalizing antibody titers. J. Clin. Microbiol. 20,2030-2034. Harrington, R. D., Hooton, T. M., Hackman, R. C., Storch, G. A., Osborne, B., Cleaves, C. A., Benson, A,, and Meyers, J. D. (1992). An outbreak of respiratory syncytial virus in a bone marrow transplant center. J . Infect. Dis. 165,987-993. Hasemann, C. A., and Capra, J. D. (1991). Baculovirus expression of antibodies: A method for the expression of complete immunoglobulins in a eukaryotic host. In “Methods: A Companion to Methods in Enzymology” (R. A., Lerner and D. R. Burton, eds.), Vol. 2, pp. 146-158. Academic Press, Orlando. Hawkins, R. E., Russell, S. J., and Winter, G. (1992). Selection of phage antibodies by binding af€inity mimicking affinity maturation. J. Mot. Biot. 226,889-896. Hawkins, R. E., Russell, S. J., Baier, M., and Winter, G. (1993). The contribution of contact and non-contact residues of antibody in the affinity of binding to antigen. The interaction of mutant D1.3 antibodies with lysozyme. J. Mol. Biol. 234,958-964. Helseth, E., Olshevsky, U., Furman, C., and Sodoroski, J. (1991). Human immunodeficiency virus type 1 gp120 envelope glycoprotein regions important for association with the gp41 transmembrane glycoprotein. J . Virol. 65,2119-2123. Hemming, V. G., Rodriguez, W., Kim, H. W., et al. (1987). Intravenous immunoglobulin treatment of respiratory syncytial virus infections in infants and young children. Antimicrob. Agents Chemother. 31,1882-1886.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

273

Hexham, J. M., Persson, M.A. A., Pegg, C., Burton, D. R., Furmaniak, J., and Smith, B. R. (1991).Cloning and expression of a human thyroglobulin autoantibody. Autoimmunity 11, 69-70. Hexham, J. M., Furmaniak, J., Pegg, C.,Burton, D. R., and Smith, B. R. (1992).Cloning of a human autoimmune response: Preparation and sequencing of a human antithyroglobulin autoantibody using a combinatorial approach. Autoimmunity 12, 135- 141. Hexham, J. M., Partridge, L. J., Furmaniak, J., Petersen, V. B., Colls, J. C., Pegg, C. A. S., Smith, B. R., and Burton D. R. (1994).Probing the human anti thyroid peroxidase autoantibody repertoire from a Hashimoto’s thyroiditis patient using combinatorial phage display libraries. Autoimmunity, in press. Hirsch, M. S., and Schooley, R. T. (1989).Resistance to antiviral drugs: The end of innocence [editorial]. N. Engl. J. Med. 320,313. Hogrefe, H. H., Mullinax, R. L., Lovejoy, A. E., Hay, B. N., and Sorge, J. A. (1993).A bacteriophage lambda vector for the cloning and expression of immunoglobulin Fab fragments on the surface of filamentous phage. Gene 128, 119-126. Holliger, P., Prospero, T., and Winter, G. (1993).Diabodies: Small bivalent and bispecific antibody fragments. Proc, Natl. Acad. Sci. U.S.A. 90,6444-6448. Hollinger, F. B.(1990).Hepatitis B virus. In “Virology” (B.N. Fields and D. M. Knipe, eds.), Vol. 2,pp. 2171-2236.Raven Press, New York. Hoogenboom, H. R., and Winter, G. (1992).By-passing immunisation: Human antibodies from synthetic repertoires of germline VHgene segments rearranged in uitro.J . Mol. Biol. 227,381-388. Hoogenboom, H. R., Griffiths, A. D., Johnson, K. S., Chiswell, D. J., Hudson, P., and Winter, G. (1991).Multi-subunit proteins on the surface of filamentous phage: Methodologies for displaying antibody (Fab) heavy and light chains. Nucleic Acids Res. 19, 4133-4 137. Huse, W. D., Sastry, L., Iverson, S. A,, Kang, A. S., Alting-Mees, M., Burton, D. R., Benkovic, S . J., and Lerner, R. A. (1989).Generation of a large combinatorial library of the immunoglobulin repertoire in phage lambda. Science 246,1275-1281. Huse, W. D., Stinchcombe, T. J., Glaser, S. M., Starr, L., MacLean, M., Hellstrom, K. E., Hellstrom, I., and Yelton, D. E. (1992).Application ofa filamentous phage pVII1 fusion protein system suitable for efficient production, screening, and mutagenesis of F(ab) antibody fragments. J. Immunol. 149,3914-3920. James, K., and Bell, G. T. (1987).Human monoclonal antibody production: Current status and future prospects. J . Immunol. Methods 100,15-40. Janda, K.D., Lo, C.-H. L., Li, T., Barbas 111, C. F., Winching, P., and Lerner, R. A. (1994). Direct selection for a catalytic mechanism from combinatorial antibody libraries. Proc. Natl. Acad. Sci. U.S.A., 91,2532-2536. Janeway, C. A. (1945).Use of concentrated human serum gamma globulin in the prevention and attenuation of measles. Bull. N.Y. Acad. Med. 21,202. Jaume, J. C., Portolano, S., Prummel, M. F., McLachlan, S. M., and Rapoport, B. (1994). Molecular cloning and characterization of genes for antibodies generated by orbital tissue-infiltrating B-cells in Graves’ ophthalmopathy. J. Clin. Endocrinol. Metabol., 78,348-352. Jones, P. T., Dear, P. H., and Foote, J. (1986).Replacing the complementarity-determining regions in a human antibody with those from a mouse. Nature (London) 321, 522-525. Jansson, U.,and Malmqvist, M. (1992).In “Advances in Biosensors” (Turner, ed.), pp. 291-336.JAI Press, San Diego, CA.

274

DENNIS R. BURTON AND CARLOS F. BARBAS I11

Kabat, E. A., and Wu, T. T. (1991).Identical V region amino acid sequences and segments of sequences in antibodies of different specificities: Relative contributions of VH and VL genes, minigenes, and complementarity-determining regions to binding of antibody-combining sites. J. Zmmunol. 147, 1709-1719. Kang, A. S., Barbas 111, C. F., Janda, K. D., Benkovic, S. J., and Lerner, R. A. (1991a). Linkage of recognition and replication functions by assemblying combinatorial antibody Fab libraries along phage surfaces. Proc. Natl. Acad. Sci. U.S.A.88,4363-4366. Kang, A. S., Burton, D. R., and Lerner, R. A. (1991b). Combinatorial immunoglobulin libraries in phage lambda. In “Methods: A Companion to Methods in Enzymology” (R. A. Lerner and D. R. Burton, eds.), Vol. 2, pp. 111-118. Academic Press, Orlando. Antibody redesign by chain shuffling Kang, A. S., Jones, T. M., and Burton, D. R. (1991~). from random combinatorial immunoglobulin libraries. Proc. Nalt. Acad. Sci. U.S.A. 88, 11,120-1 1,123. Keller, P. M., Neff, B. J., and Ellis, R. W. (1984). Three major glycoprotein genes of varicella-zoster virus whose products have neutralization epitopes. J . Virol. 52, 293-297. Kendler, D. L., and Davies, T. F. (1993).Immunological mechanisms in Graves’disease. In “The Molecular Pathology of Autoimmune Diseases” (C. A. Bona, K. A. Siminovitch, M. Zanetti, and A. N. Theofilopoulos, eds.), pp. 511-540. Harwood Academic Publishers, Chur, Switzerland. Krugman, S. (1976). Effect of human immune serum globulin on infectivity of hepatitis A virus. J . Infect. Dis. 134, 70-74. Krugman, S., Ward, R., Giles, J. P., and Jacobs, A. M. (1960).Infectious hepatitis: Studies on the effect of gamma globulin and on the incidence of inapparent infection. J . Am. Med. Assoc. 174,83-90. Kiihn, J. E., Dunkler, G., Munk, K., and Braun, R. W. (1987). Analysis of the IgM and IgG antibody response against herpes simplex virus type 1 (HSV-1) structural and nonstructural protein. J . Med. Virol. 23, 135-150. Kiimel, G., Kaerner, H. C., Levine, M., Schroder, C. H., and Glorioso, J. C. (1985). Passive immune protection by herpes simplex virus-specific monoclonal antibodies and monoclonal antibody-resistant mutants altered in pathogenicity. J . Virol. 56, 930-937. Laal, S., and Zolla-Pazner, S. (1993). Epitopes of HIV-1 glycoproteins recognized by the human immune system. Chem. Zmmunol. 56,91-111. Larkin, M., Heckels, J. E., and Ogilvie, M. M. (1985).Antibody response to a varicellazoster virus surface glycoproteins in chickenpox and shingles. J . Gen. Virol. 66, 1785-1793. Larrick, J. W., Danielsson, L., Brenner, C. A., et al. (1989). Rapid cloning of rearranged immunoglobulin genes from human hybridoma cells using mixed primers and the polymerase chain reaction. Biochem. Biophys. Res. Commun. 160, 1250-1256. Laughon, B. E., Allaudeen, H. S., Becker, J. M., Current, W. L., Feinberg, J., Frenkel, J. K., Hafner, R., Hughes, W. T., Laughlin, C. A., Meyers, J. D., Schrager, L. K., and Young, L. S. (1991). Summary of the workshop on future directions in discovery and development of therapeutic agents for opportunistic infections associated with AIDS. J. Infect. Dis. 164,244-251. Lerner, R. A., Kang, A. S., Bain, J. D., Burton, D. R. and Barbas 111, C. F. (1992). Antibodies without immunization. Science 258, 1313-1314. Leung, D. W., Chen, E., and Goeddel, D. V . (1989). A method for random mutagenesis of a defined DNA segment using a modified polymerase chain reaction. Tech. J. Methods Cell Mol. Biol. 1, 11-15.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

275

Levine, B., Hardwick, J. M., Trapp, B. D., Crawford, T. O., Bollinger, E. C., and Griffin, D. E. (1991). Antibody-mediated clearance of alphavirus infection from neurons. Science 254,856-860. Lowman, H. B., Bass, S. H., Simpson, N., and Wells, J, A. (1991). Selecting high-affinity binding proteins by monovalent phage display. Am. Chem. SOC.30, 10,832-10,838. Lum, L. G., Bums, E., and Janson, M. M. (1990). IgG anti-tetanus toxoid antibody synthesis by human bone marrow. I. Two distinct populations of marrow B cells and functional differences between marrow and peripheral blood B ce1ls.J. Clin. Immunol. 10,255-264. Marks, j . D., Hoogenboom, H. R., Bonnert, T. P., McCafferty, j . , Griffiths, A. D., and Winter, G . (1991). By-passing immunization: Human antibodies from V-gene libraries displayed on phage.]. Mol. Biol. 222, 581-597. Marks, J. D., Griffiths, A. D., Malmqvist, M., Clackson, T. P., Bye, J. M., and Winter, G. (1992).By-passing immunization: Building high affinity human antibodies by chain shuffling. Biotechnology 10,779-783. Marks, J. D., Ouwehand, W. H., Bye, J. M., Finnern, R., Gorick, B. D., Voak, V., Thorpe, S. J,, Hughes-Jones, N. C., and Winter, G. (1993).Human antibody fragments specific for human blood group antigens from phage display library. Biotechnology 11, 1145-1148. Marshall, G. S., Rabalais, G . P., Stout, G. G., and Waldeyer, S. L. (1992).Antibodies to recombinant-derived glycoprotein B after natural human cytomegalovirus infection correlate with neutralizing activity. J. Infect. Dis. 165,381-384. Masur, H. (1990). Proceedings of a symposium: Clinical implications of herpesvirus infections in patients with AIDS. Am. J. Med. 92, 2A-1s. McCafferty, J., Griffiths, A. D., Winter, G., and Chiswell, D. J. (1990).Phage antibodies: Filamentous phage displaying antibody variable domains. Nature (London) 348, 552-554. McIntosh, K., and Chanock, R. M. (1990). Respiratory syncytial viruses. In “Virology” (B. N. Fields and D. M. Knipe, eds.), Vol. 1, pp. 1045-1072. Raven Press, New York. McKeating, J. A., Shotton, C., Cordell, J., Graham, S., Balfe, P., Sullivan, N., Charles, M., Page, M., Blomstedt, A., Olofsson, S., Kayman, S. C., Wu, Z., Pinter, A., Dean, C., Sodroski, J., and Weiss, R. A. (1993).Characterization of neutralizing monoclonal antibodies to linear and conformation-dependent epitopes within the first and second variable domains of human immunodeficiency virus type 1 gp120. J. Virol. 67, 4932-4944. McLachlan, S. M., and Rapoport, B. (1992).The molecular biology ofthyroid peroxidase: Cloning, expression and role as autoantigen in autoimmune thyroid disease. Endocr. Reu. 13, 192-206. Mease, P. J., Ochs, H. D., and Wedgwood, R. J. (1981).Successful treatment ofechovirus 11 meningoencephalitis and myositisfascitis with intravenous immune globulin therapy in a patient with X-linked agammaglobulinemia. N . Engl. J. Med. 304,12781281. Meyer, H., Masuho, Y., and Mach, M. (1990). The gp116 of the gp58/116 complex of human cytomegalovirus represents the amino terminal part of the precursor molecule and contains a neutralizing epitope. J . Gen. Virol. 71,2443-2450. Model, P., and Russel, M. (1988). Filamentous bacteriophage. In “The Bacteriophages” (R. Calendar, ed.), Vol. 2, pp. 375-456. Plenum, New York. Mofenson, L. M., and Moye, J. (1953). Intraveneous immune globulin for the prevention of infections in children with symptomatic human immunodeficiency virus infection. Pediatr. Res. 33, (1 suppl.), S80. Moore, J. P., and Ho, D. D. (1993). Antibodies to discontinuous or conformationally-

276

DENNIS R. BURTON AND CARLOS F. BARBAS I11

sensitive epitopes on the gp120 glycoprotein of human immunodeficiency virus type 1 are higly prevalent in sera of infected humans. J. Virol. 67, 863-875. Moore, J. P., and Nara, P. L. (1991).The role of the V3 loop in HIV infection. AIDS 5 (Suppl. 2), S21-S33. Moore, J. P., Sattentau, Q. J., Yoshiyama, H., Thali, M., Charles, M., Sullivan, N., Poon, S.-W., Fung, M. S., Traincard, F., Pinkus, M., Robey, G., Robinson, J. E., Ho, D. D., and Sodroski, J. (1993).Probing the structure of the V2 domain ofhuman immunodeficiency virus type 1 surface glycoprotein gp120 with a panel of eight monoclonal antibodies: Human immune response to the V1 and V2 domains. J. Virol. 67, 6136-6151. Moore, P. M. (1990). Immunoglobulin binding to neuronal cell surface epitopes in murine systemic lupus erythematosus. J . Neuroimmunol. 30, 101-109. Momson, S. L., Johnson, M. J., and Herzenberg, L. A. (1984).Chimeric human antibody molecules: Mouse antigen-binding domains with human constant domains. Proc. Notl. Acad. Sci. U.S.A. 81,6851-6855. Mosca, J. D., Bednarik, D. P., Raj, N. B. K., Rosen, C. A., Sodroski, J. G., Haseltine, W. A., and Pitha, P. M. (1987).Herpes simplex virus type-1 can reactivate transcription of latent human immunodeficiency virus. Nature (London) 325,67-70. Mullinax, R. L., Gross, E. A., Amberg, J. F., Hay, B. N., Hogrefe, H. H., Kubitz, M. M., Greener, A., Alting-Mees, M., Ardourel, D., Short, J. M., Sorge, J. A., and Shopes, B. (1990). Identification of human antibody fragment clones specific for tetanus toxoid in a bacteriophage lambda immunoexpression library. Proc. Natl. Acad. Sci. U.S.A. 87,8095-8099. Myers, R. M., Lerman, L. S., and Maniatis, T. (1985).A general method for saturation mutagenesis of cloned DNA fragments. Science 229,242-247. Nara, P. L., Hatch, W. C., Dunlop, N. M., Robey, W. G., Arthur, L. O., Gonda, M. A., and Fischinger, P. J. (1987).Simple, rapid, quantitative syncytium-forming microassay for the detection of human immunodeficiency virus neutralizing antibody. AIDS Res. Hum. Retroviruses. 3,283-302. Nara, P. L., Garrity, R. R., and Goudsmit, J. (1991).Neutralization of HIV-1: A paradox of humoral proportions. FASEB J . 5,2437-2455. Neuberger, M. S., Williams, G. T., and Mitchell, E. B. (1985).A hapten-specific chimeric immunoglobulin E antibody which exhibits human physiological effector function. Nature (London) 314,268-271. Neurath, A. R. (1993).B cell antigenic mapping of HIV-1 glycoproteins. Chem. Immunol. 56,34-60. Oakes, J. E., and Lausch, R. N. (1984).Monoclonal antibodies suppress replication of herpes simplex virus type 1 in trigeminal ganglia. J . Virol. 51,656-661. Ohizumi, Y., Suzuki, H., Matsumoto, Y.-I., Masuho, Y., and Numazaki, Y. (1992).Neutralizing mechanisms oftwo human monoclonal antibodies against human cytomegalovirus glycoprotein 130/55.J. Gen. Virol.'73,2705-2707. Ohlin, M., Sundqvist, V.-A., Mach, M., Wahren, B., and Borrebaeck, C. A. K. (1993). Fine specificity of the human immune response to the major neutralization epitopes expressed on cytomegalovirus gp58/116 (gB) as determined with human monoclonal antibodies. J . Virol. 67,703-710. Orlandi, R., Gussow, D. H., Jones, P. T., et al. (1989).Cloning immunoglobulin variable domains for expression by the polymerase chain reaction. Proc. Natl. Acad. Sci. U S A . 86,3833-3837. Pantaleo, G., Graziosi, C., and Fauci, A. S. (1993).The immunopathogenesis of human immunodeficiency virus infection. N. Engl. J. Med. 328,327-335.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

277

Parmley, S. F., and Smith, G. P. (1988). Antibody selectable filamentous fd phage vectors: Affinity purification of target genes. Gene 73,305-318. Pass, R. F., Whitley, R. J., Welchel, J. D., Diethelm, A. G., Reynolds, D. W., and Alford, C. A. (1979). Identification of patients with increased risk of infection with herpes simplex after renal transplantation. J . Infect. Dis. 140,487-492. Persson, M. A., Caothien, R. H., and Burton, D. R. (1991). Generation of diverse highaffinity human monoclonal antibodies by repertoire cloning. Proc. Natl. Acad. Sci. U.S.A. 88,2432-2436. Portolano, S., Seto, P., Chazenbalk, G. D., Nagayama, Y.,McLachlan, S. M., and Rapoport, B. (1991). A human Fab fragment specific for thyroid peroxidase generated by cloning thyroid lymphocyte derived immunoglobulin genes in a bacteriophage lambda library. Biochem. Biophys. Res. Commun. 179,372-377. Portolano, S., Chazenblak, G. D., Seto, P., Hutchison, J. S., Rapoport, B., and McLachlan, S. M. (1992). Recognition by recombinant autoimmune thyroid disease-derived Fab fragments of a dominant conformational epitope on human thyroid per0xidase.J. Clin. Invest. 90,720-726. Portolano, S., Chazenbalk, G. D., Hutchison, J. S., McLachlan, S. M., and Rapoport, B. (1993a). Lack of promiscuity in autoantigen-specific H and L chain combinations as revealed by human H and L chain “roulette.”J. Immunol. 150,880-887. Portolano, S., McLachlan, S. M., and Rapoport, B. (1993b).High affinity, thyroid specific human autoantibodies displayed on the surface of filamentous phage use V genes similar to other autoantibodies. J. Immunol. 151,2839-2851. Prince, G. A., Hemming, V. G., Horswood, R. L., Baron, P. A., and Chanock, R. M. (1987). Effectiveness of topically administered neutralizing antibodies in experimental immunotherapy of respiratory syncytial virus infection in cotton rats. J . Virol. 61, 1851-1854. Radic, M. Z., Mascelli, M. A., Erikson, J., Shan, H., and Weigert, M. (1991). Ig H and L chain contributions to autoimmune specificities. J . Immunol. 146, 176-82. Rasmussen, L. (1993).Progress in the development ofa human cytomegalovirusglycoprotein subunit vaccine. In “Multidisciplinary approach to understanding cytomegalovirus disease” (S. Michelson and S. A. Platkin, eds.), pp. 311-320. Excerpta Medica, Amsterdam. Reeves, W. C., Corey, L., Adams, H. G., Vontver, L. A., and Holmes, K. K. (1981).Risk of recurrence after first episodes of genital herpes: Relation to HSV type and antibody response. N. Engl. J . Med. 305,315-319. Riggs, M. W., and Perryman, L. E. (1987).Infectivity and neutralization of Cryptosporidium paroum sporozoites. Infect. Immun. 55,2081-2087. Riggs, M. W., McGuire, T. C., Mason, P. H., and Perryman, L. E. (1989).Neutralizationsensitive epitopes are exposed on the surface of infectious Cryptosporidium paroum sporozoites. J . Immunol. 143, 1340-1345. Roben, P., Moore, J. P., Thali, M., Sodroski, J., Barbas 111, C. F., and Burton, D. R. (1994). Recognition properties of a panel of human recombinant Fab fragments to the CD4 binding site of GP120 showing differing ability to neutralise human immunodeficiency virus type 1. Submitted for publication. Rosenblum, J. S., and Barbas 111, C. F. (1994). Synthetic antibodies. Antibody Eng., in press. Rowley, M. J., Gershwin, M. E., and Mackay, I. R. (1993).Molecular biology ofautoantigens in primary biliary cirrhosis. In “The Molecular Pathology of Autoimmune Diseases’’ (C. A. Bona, K. A., Siminovitch, M. Zanetti, and A. N. Theofilopoulos, eds.), pp. 627-642. Hanvood Academic, Chur, Switzerland.

278

DENNIS R. BURTON AND CARLOS F. BARBAS 111

Samuelsson, A., Chiodi, F., Norrby, E., and Persson, M. A. A. (1994). Macaque monoclonal Fab molecules derived from a combinatorial library are specific for the SVI envelope glycoprotein. In “Vaccines 94.” Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sanz, I. (1991).Multiple mechanisms participate in the generation of diversity of human H chain CDR3 regions. J . Immunol. 147, 1720-1729. Sasano, M., Burton, D. R., and Silverman, G. J. (1993). Molecular selection of human antibodies with unconventional Bacterial B cell antigen. J . Immunol. 151,5822-5839. Sastry, L. Alting-Mees, M., Huse, W. D., et al. (1989). Cloning of the immunological repertoire in Escherichia coli for generation of monoclonal catalytic antibodies: Construction of a heavy chain variable region-specific cDNA library. Proc. Natl. Acad. Sci. U.S.A. 86,5728-5732. Schaaper, R. M. (1988). Mechanisms of mutagenesis in the Escherichia coli mutator mutD5: Role of DNA mismatch repair. Proc. Natl. Acad. Sci. U.S.A. 8,8126-8130. Scherle, P. A., Palladino, G., and Gerhard, W. (1992). Mice can recover from pulmonary influenza virus infection in the absence of class 1-restricted cytotoxic T cells. J . lmmunol. 148,212-217. Scott, J. K., and Smith, G. P. (1990). Searching for peptide ligands with an epitope library. Science 249,386-390. Siegel, D. L., and Silberstein, L. E. (1994). Expression and characterization of recombinant anti-Rh(D) antibodies on filamentous phage: A model system for isolating human red cell antibodies by repertoire cloning. Blood, in press. Silverman, G. J. (1992). Human antibody responses to bacterial antigens: Studies of a model conventional antigen and a proposed model B cell superantigen. Int. Reo. lmmunol. 9,57-78. Silverman, G. J. (1994). Superantigens and the spectrum of unconventional B cell antigens. Immunologist, in press. Silverman, G. J., Sasano, M., and Wormsley, S. B. (1993). Age-associated changes in binding of human B lymphocytes to a VH3-restrictedunconventional bacterial antigen. J . Immunol. 151,5840-5855. Simmons, A., and Nash, A. A. (1985). Role of antibody in primary and recurrent herpes simplex virus infection. J . Virol. 53,944-948. Skerra, A., and Pluckthun, A. (1988). Assembly of a functional immunoglobulin Fv fragment in Escherichia coli Science 240, 1038-1041. Smith, G. P. (1985). Filamentous fusion phage: Novel expression vectors that display cloned antigens on the virion surface. Science 228, 1315-1316. Smith, J. W., Languino, L. R., and Barbas 111, C. F. (1993). High affinity self-reactive human antibodies by design and selection: Targeting the integrin ligand binding site. Proc. Natl. Acad. Sci. U.S.A. 90, 10,003-10,007. Smith, J. W., Hu, D., Satterthwait, A. C., Pinz-Sweeney, S., and Barbas 111, C. F. (1994). Building synthetic antibodies as adhesive ligands for integrins. Submitted for publication. Syndman, D. R. (1990). Cytomegalovirus immunoglobulins in the prevention and treatment of cytomegalovirus disease. Reo. Infect. Dis. 12, S8394848. Snydman, D. R., Werner, B. G., Heinze-Lacey, B., Beredi, B. P., Tilney, N. L., Kirkman, R. L., Milford, E. L., Cho, S. I., Bush, H. L., Levey, A. S., Strom, T. B., Carpenter, C. B., Levey, R. H., Harman, W. E., Zimmerman, Z. E., Shapiro, M. E., Steinman, T., LoGerfo, F., Idelson, B., Schroter, G. P. J., Levin, M. J., McIver, J., Leszczynski, J., and Grady, G. F. (1987). Use of cytomegalovirus immune globulin to prevent cytomegalovirus disease in renal-transplant recipients. N . Engl. J. Med. 317, 1049-1054.

ANTIBODIES FROM COMBINATORIAL LIBRARIES

279

Soderlind, E, Simonsson, A. C., and Borrebaeck, C. A. K. (1992). Phage display technology in antibody engeering: Design of phagemid vectors and in oitro maturation systems. Immunol. Reo. 130,109-124. Soderlind, E., Lagerkvist, A. C. S., Duetias, M., Malmborg, A.-C., Ayala, M., Danielsson, L., and Borrebaeck, C. A. K. (1993). Chaperonin assisted phage display of antibody fragments on filamentous bacteriophages. Biotechnology 11,503-507. Stanfield, R. L., Takimoto-Kamimura, M., Rini, J. M., Profy, A. T., and Wilson, I. A. (1993).Major antigen induced domain rearranging in an antibody. Structure 1,83-93. Stevens, R. H., Macy, E., Morrow, C., and Saxon,A. (1979).Characterization ofa circulating subpopulation of spontaneous antitetanus toxoid antibody producingB cells following in vivo booster immunization. ]. Immunol. 122,2498-2504. Stewart, A. K., Huang, C., Long, A. A., Stollar, B. D., and Schwartz, R. S. (1992). VH-representation in autoantibodies reflects the normal human B-cell repertoire. Immunol. Reo. 128,101-122. Surh, C. D., Ahmed-Ansari, A., and Gershwin, M. E. (1990).Comparative epitope mapping of murine monoclonal and human autoantibodies to human PDH-E2, the major mitochondria1 autoantigen of primary biliary cirrhosis. ]. Immunol. 144, 26472652. Takeda, S . , Dorfman, N. A., Robert-Guroff, M., Notkins, A. L., and Rando, R. F. (1994). Two-phase approach for the expression of high affinity human anti-hiv immunoglobulin Fab domains in E . Coli. AIDS Res. Hum. Retroviruses, in press. Thiele, C. J., Morrow, C. D., and Stevens, R. H. (1981). Multiple subsets of anti-tetanus toxoid antibody-producing cells in human peripheral blood differ by size, expression of membrane receptors, and mitogen reactivity. J. Immunol. 126, 1146-1153. Tpizori, S . , Roberton, D., and Chapman, C. (1986).Remission of diarrhea due to cryptosporidiosis in an immunodeficient child treated with hyperimmune bovine colostrum. Br. Med. J. 293,1276-1277. Tyler, K. I., Virgin IV, H. W., Bassel-Duby, R., and Fields, B. N. (1989). Antibody inhibits defined stages in the pathogenesis of reovirus serotype 3 infection of the centyl nervous system. J . Erp. Med. 170,887-900. Uhl, E. W., O’Connor, R. M., Perryman, L. E., and Riggs, M. W. (1992).Neutralizationsensitive epitopes are conserved among geographically diverse isolates of Cryptosporidium paruum. Infect. Immun. 60, 1703-1706. Ungar, B. L. P. (1990). Cryptosporidiosis in humans (Homo sapiens). In “Cryptosporidiosis of Man and Animals” (J. P. Dubey, C. A. Speer, and R. Fayer, eds.), pp. 66-75. CRC Press, Boca Raton, FL. Urban, M., Britt, W., and Mach, M. (1992).The dominant linear neutralizing antibodybinding site of glycoprotein gp86 of human cytomegalovirus is strain specific.]. Viro. 66,1303-1311. Utz, U., Brit, W., Vugler, L., and Mach, M. (1989).Identification of a neutralizing epitope on glycoprotein gp58 of human cytomegalovirus.J. Virol. 63,1995-2001. Vafai, A., Wroblewska, Z., Wellish, M., Green, M., and Gilden, D. (1984). Analysis of three late varicella-zoster virus proteins, a 125,000-molecular-weightprotein and gpl and gp3. J . Virol. 52,953-959. Virgin IV, H. W., Bassel-Duby, R., Fields, B. N., and Tyler, K. L. (1988). Antibody protects against lethal infection with the neurally spreading reovirus type 3 (Dearing). J. Virol. 62,4594-4604. Volkman, D. J,, Allyn, S . P., and Fauci, A. S . (1982).Antigen-induced in vitro antibody production in humans: Tetanus toxoid specific antibody synthesis. J . Immunol. 107-112. Walker, M. R., Bevan, L. J., Daniels, J., et al. (1992). Isolation and amplification of

280

DENNIS R. BURTON AND CARLOS F. BARBAS I11

human IgE Fd encoding mRNA from human peripheral blood lymphocytes.]. Immunol. Methods 149,77-85. Wall, J. R., Bernard, N., and Boucher, A. (1993). Pathogenesis of thyroid-associated ophthalmopathy: An autoimmune disorder of the eye muscle associated with Graves’ hyperthyroidism and Hashimoto’s thyroiditis. Clin. Immunol. Immunopathol. 68, 1-8. Ward, E. S., Giissow, D., Griffiths, A. D., Jones, P. T., and Winter, G. (1989). Binding activities of a repertoire of single immunoglobulin variable domains secreted from Escherichia coli Nature (London) 341,544-546. Weetman, A. P. (1992).Thyroid associated ophthalmopathy. Autoimmunity 12,215-222. Whitley, R. J. (1990). Herpes simplex viruses. In “Virology” (B. N. Fields and D. M. Knipe, eds.), 2nd Ed., pp. 1843-1887. Raven, New York. Whitlow, M., and Filpula, D. (1991).Single-chain Fv proteins and their fusion proteins. In “Methods: A Companion to Methods in Enzymology” (R. A. Lerner and D. R. Burton, eds.), Vol. 2, pp. 97-105. Academic Press, Orlando. Whittle, N., Adari, J., Lloyd, C., Jenkins, L., Devine, J., Schlom, J., Raubitschek, A., Colcher, D., and Bodmer, M. (1987). Expression in COS cells of a mouse-human chimaeric B72.3 antibody. Protein Eng. 1,499-506. Williamson, R. A., Persson, M. A. A., and Burton, D. R. (1991). Expression of a human monoclonal anti(rhesus D) Fab fragment in Escherichia coli with the use of bacteriophage lambda vectors. Biochem. J . 277,561-563. Williamson, R. A., Burioni, R., Sanna, P. P., Partridge, L. J., Barbas 111, C. F., and Burton, D. R. (1993).Human monoclonal antibodies against a plethora of viral pathogens from single combinatorial libraries. Proc. Natl. Acad. Sci. U S A . 90,4141-4145. Wilson, I. A., and Stanfield, R. L. (1993).Antibody-antigen ineractions. Curr. Opinion Struct. Biol. 3, 113-118. Winkler, T. H., Jahn, S., and Kalden, J. R. (1991). IgG human monoclonal anti-DNA autoantibodies from patients with systemic lupus erythematosus. Clin. E x p . Immunol. 85,379-385. Winkler, T. H., Fehr, H., and Kalden, J. R. (1992).Analysis of immunoglobulin variable region genes from human IgG anti-DNA hybridomas. Eur.]. Immunol. 22,1719-1728. Winter, G., and Milstein, C. (1991). Man-made antibodies. Nature (London) 349, 293-299. Wright, A., and Shin, S.-U.(1991).Production of genetically engineered antibodies in myeloma cells: Design, expression and applications. In “Methods: A Companion to Methods in Enzymology” (R. A. Lerner and D. R. Burton, eds.), Vol. 2, pp. 125-135. Academic Press, Orlando. Wu, T. T., and Kabat, E. A. (1970).An analysis of the sequences of the variable regions of Bence Jones proteins and myeloma light chains and their implications for antibody complementarity. J . E r p . Med. 132,211-250. Wu, T. T., Johnson, G., and Kabat, E. A. (1993). Length distribution of CDRH3 in antibodies. Proteins Struct. Funct. Genet. 16, 1-7. Zaia, J. A., Levin, M. J,, Preblud, S. R., Leszezynski, J., Wright, G. G., Ellis, R. J., Curtis, A. C., Valerio, M. A., and LeCore, J. (1983). Evaluation of varicella-zoster immune globulin: Protection of immunosuppressed children after household exposure to varicella. J . Infect. Dis. 147, 737-743. Zebedee, S. L., Barbas, C. F., Hom, Y.-L., Caothien, R., Graff, R., DeGraw, J., Pyati, J,, LaPolla, R., Burton, D. R., Lerner, R. A,, and Thornton, G. B. (1992). Human combinatorial antibody libraries to hepatitis B surface antigen. Proc. Natl. Acad. Sci. U.S.A.89,3175-3179.

ADVANCES IN IMMUNOLOGY,VOL. 57

Immune Response against Tumors CLAUDE ROTH, CHRISTOPH ROCHLITZ,’ AND PHlLlPPE KOURILSKY Unit6 de Biologie Mdkulrrirr du Gkre, 0.277 Insem, lnstht Pasteur, 25, rue du Dr. Rou~r,75724 Pans c6dex 15, France

1. Introduction

In the past few years, tumor immunology has made considerable progress and is receiving more and more attention. This increasing interest is justified by several reasons. First of all, the notion of tumorspecific antigens, which initially involved integral cell-surface proteins recognized by antibodies, has been extended to T-cell epitopes presented by MHC molecules and recognized by specific T cells. The field of tumor immunology has then largely shifted toward the analysis and manipulation of T-cell responses. A new conceptual framework has opened up new experimental approaches. Results now obtained in animal models have become so spectacular that expectations have been raised concerning the prospects for the immunotherapy of human cancers. A major transition can be traced back to the mid 1980s, when Townsend et aZ. (1985,1986a)demonstrated that the influenza nucleoprotein is presented to specific cytolytic T cells (CTL) in the form of a small peptide carried by a given class I MHC molecule. It followed that intracellular proteins of the host might also be displayed as processed peptides at the cell surface (Townsend et d.,1985; Kourilsky and Claverie, 1986). The notion that any class I MHC-positive cell could permanently expose on its surface thousands of self-peptides led to a broadening of the concept of immunosurveillance (Kourilsky and Claverie, 1986). It was thus shown that dysregulations in tumor cells might become visible by the immune system, either as peptides derived from unmutated, upregulated self-proteins, or as mutant peptides derived from mutated self-proteins (Kourilsky and Claverie, 1986; Kourilsky et al., 1987,1991).The pioneering work of Boon and his co-workers, started in the mid 1970s, has provided, in 1988-1991, the experimental foundation for the important concept of T-celldefined specific tumor epitopes, the latter being actually derived from Present address: Department Innere Medizin, Abteilung fur Onkologie, Kantosspital Basel am Petergsraben 4, CH-4031 Basel, Switzerland

281

Copyright 0 1994 by Academic Press, Inc. All rights of reproduction in any form reserved.

282

CLAUDE ROTH ET AL

mutant self- or upregulated unmutated self-proteins (reviewed in Boon, 1992). A growing body of basic immunology is developing around the fundamental notion that not all potentially presentable self-peptides are actually presented by MHC molecules in a way that triggers T-cell tolerance (Cibotti et al., 1991; Milich et aZ., 1991; reviewed in Sercarz et al., 1993). It is expected that these peptides may serve as a reservoir for some tumor-specific antigens and that understanding the rules that place them in this category will ultimately be useful to tumor immunology. The realization that tumor epitopes can be recognized by T cells opened new avenues for immunointervention. Among these, the transfection of genes which could possibly increase the immunogenicity of tumor cells, such as cytokine genes, was particularly attractive, because immunogenicity could be increased without even knowing the molecular nature of the tumor-specific antigens. The first results obtained in mouse tumor models between 1988 and 1991 were indeed remarkably successful (Bubenik et al., 1988; Fearon et ul., 1990; Gansbacher et uZ., 1990a; Ley et al., 1990,1991; Russel et al., 1991). At the least, they could sustain the claim that, even if natural immune responses are often weakly relevant to tumor control in uiuo, it may be possible to manipulate immune responses in order to make them efficient enough to control tumors. The debate over the relevance of immune responses against tumors was thus muted by the notion that immunointervention might well do the job, provided that the appropriate T-cell responses are activated. In the past few years, the flurry of spectacular results in mouse tumor models has encouraged optimism as to their applicability in man. Dozens of clinical trials are currently under way and dozens are being planned. An understandable trait of tumor immunology is that every finding is almost immediately framed within a therapeutic perspective. This may explain the confusion, commonly encountered in the field, between an antigen and an immunogen-because any antigen is immediately used as an immunogen to try to control tumor growth. In this context, we have attempted to review here a number of aspects related both to the antigenicity and the immunogenicity of tumors. We could not cover the entire field and have selected what we felt are the most relevant trends in today’s research. In doing so, we hope to have reviewed a sufficiently broad area to substantiate the conclusion that the complexity of the field should temper some of the ongoing, naive optimism for human immunotherapy. This and connected fields have been reviewed by others (reviewed in Klein and Boon, 1993), and we apologize in advance for any omission which, in this relatively large survey, we might have unwillingly made.

IMMUNE RESPONSE AGAINST TUMORS

283

It. Tumor Antigens

A. ANTIGENS D E ~ ~ NSEROLOGICALLY ED

It was initially assumed that tumor antigens would be unique to malignancy and not found in normal cells and that they might be specific to the tissue of origin and thus reflect the state of differentiation of the cells within this tissue. Antigens expressed on the cell surface of tumor cells were thought to be the main components detected by the immune system. Specific antisera and, subsequently, monoclonal antibodies were developed to identify them. Tumors often express high levels of differentiation antigens, usually corresponding to the differentiation stage of the cell in which the malignant transformation occurred. An example is provided by the oncofetal antigens, such as a-feto-protein (AFP) on human hepatomas and carcinoembryonic antigen (CEA)on colon carcinomas. These markers were the first human tumor-associated antigens to be discovered. While they are not as tumor specific as originally believed, they still continue to be promising for the purpose of diagnosis and therapy (Mach et al., 1993). On B and T lymphocytes, growth factor receptors, such as receptors for transferrin, interleukin-2, or insulin, also belong to this category of normal differentiation antigens, since they are absent or barely detectable on resting lymphoid cells, but abundant on stimulated and dividing cells. In human melanomas, the gp95/p97 antigen studied by several groups provides a clear example of a tumor-associated antigen shared with normal tissues. The initial biochemical characterization of the p97 antigen from human melanoma cells showed that the parital amino acid sequence of this phosphorylated sialoglycoprotein is structually related to that of transferrin (Brown et al., 1981,1982). The entire gene was then cloned (Rose et al., 1986). Immunization of mice with recombinant vaccinia virus expressing p97 antigen (Hu et al., 1988) induced antibody production and T-cell proliferation. More strikingly, immunized mice rejected melanoma cells expressing p97 while they allowed the outgrowth of p97-negative tumors (Estin et al., 1988). Whether this approach is valid for human therapy is not yet clear, as normal tissues expressing trace amounts of p97 could constitute targets for immune attack. At about the same time, other investigators observed that sera isolated from patients with melanoma bound only to the autologous melanoma, while a mouse monoclonal antibody raised against the purified gp95/p97 antigen could also bind to allogeneic melanomas (Shiku et al., 1976; Real et al., 1984). By immunoprecipitation, this monoclonal antibody was shown to detect a 95-kDa molecule on other melanomas

284

CLAUDE ROTH ET AL.

and on some normal tissues. It was thus postulated that the patients’ antibodies were able to detect a new epitope on a common molecule, originally identified as a melanoma antigen, and designated gp95 (Dippold et al., 1980) or p97 (Woodbury et al., 1980; Brown et al., 1980). These results were the first to show that a common melanoma antigen bearing specific mutations can behave as an autoimmunogenic epitope that elicits a specific antibody response (Furukawa et al., 1989). An additional example of tumor-associated antigens expressed in normal tissues at a given stage of differentiation is provided by the tyrosinase-related melanosomal antigens and by the cell-surface glycolipids whose expression increases during malignant transformation. The glycoprotein gp75, related to tyrosinase, an enzyme that catalyzes the formation of melanin, was originally identified by autoantibodies in patients with metastatic melanoma (Vijayasaradhi and Houghton, 1991). Using a mouse monoclonal antibody against gp75, it was shown that the passive transfer of this autoantibody into mice bearing B16 melanoma tumor cells induced the rejection of established tumors, but also produced autoimmunity against regenerating melanocytes (Houghton, 1992). Gangliosides have also been identified as target structures for antibodies in cancer patients. In passive immunotherapy, tumor remissions were observed with antibodies directed against gangliosides in melanomas and neuroblastomas (Houghton et al., 1985; Cheung et al., 1989). For active immunotherapy, the main problem is still to induce in vivo a strong enough antibody response, as discussed by Livingston (1991). Various immunization protocols are currently being investigated, using either purified or modified glycolipid antigens (Ritter et al., 1991) or antiidiotypic antibodies against the GD3 ganglioside as the immunogen (Chapman and Houghton, 1991). In mice, the p53 tumor suppressor molecule is another case of a serologically detected tumor antigen. Antibodies reacting against p53 molecules have been found in the sera of animals inoculated with methylcholantrene-induced sarcomas (Deleo et al., 1979). Biochemical analysis of p53 tryptic peptides demonstrated that mutations in p53 might be involved in the individually distinct immunological characteristics of methylcholantrene-induced tumors ( Jay et d., 1979). From these observations, it appears that tumor antigens, recognizable by antibodies, derive either from mutant self-proteins or from unmutated upregulated or modified components (e.g., at the level of carbohydrates) of the cell surface. The concept thus emerged that tumors may express specific epitopes, rather than specific molecules. This became even more evident when T-cell responses against tumor cells were analyzed in detail.

IMMUNE RESPONSE AGAINST TUMORS

285

B. EARLYEVIDENCE FOR TUMOR-SPECIFIC ANTIGENSRECOGNIZED BY T CELLS Evidence for the existence of specific tumor rejection antigens can be traced back to the pioneering work of Gross (1943), Foley (1953), and Prehn and Main (1957),who showed that inbred mice immunized with a syngeneic chemically induced tumor could reject a graft of the same tumor. The antigens were unique for each individual chemically induced tumor, since specific protection was obtained against the' immunizing but not against other tumors, even if induced by the same carcinogen (methycholantren), and in mice of the same strain (Klein et al., 1960). The antigens responsible for tumor rejection were then referred to as tumor-specific transplantation antigens (TSTAs).Several major characteristics could be defined: 1. Each tumor had individually distinct antigens. 2. These antigens were stably expressed during successive passages in vivo and in uitro. They were usually not lost after passage in immunized animals. 3. There was a relationship between the dose of carcinogen used to induce the tumor, the latency period before tumor growth, and the immunogenicity of the tumor. In tumors induced by high doses of carcinogen, the latency period was short and the immunogenicity was generally high (Prehn, 1975; Prehn and Karcher, 1983; Prehn and Prehn, 1987). 4. TSTAs were not restricted to tumors induced by chemical carcinogens but were also found expressed in UV-induced tumors (Kripke, 1981; Ward et al., 1989),mammary tumors induced by mammary tumor virus (Morton et al., 1969),and tumors of spontaneous origin (Carswell et al., 1970). 5. Immunity could be transferred to naive mice with lymphocytes from immunized animals (reviewed in Melief, 1992), strongly implicating T cells.

Extensive biochemical studies on chemically induced murine tumors were undertaken to determine the molecular nature of these antigens. The diversity of the results diminished the hope of identifying common structures that could be used for diagnostic or therapeutic purposes. In some cases, TSTAs were identified as deregulated embryonic antigens (Coggin and Anderson, 1974; Medawar and Hunt, 1983),while in other cases they were characterized as allo- or mutatedMHC molecules (Invernizzi and Parmiani, 1975; Philipps et al., 1985; Stauss et al., 1986a; Lee et al., 1988), as viral antigens derived from

286

CLAUDE ROTH ET AL.

recombinant murine virus (Roman et al., 1981; Lennox et al., 1981), as antigens linked to the Ig heavy-chain locus (Pravtcheva et al., 1981), or as antigens related to heat shock proteins (Ullrich et al., 1986; Moore et al., 1987). Several hypothesis were proposed to account for this antigenic diversity (Schreiber et al., 1988).First, individual transplantation antigens could result from a direct mutagenic activity of chemical and physical carcinogens in a cellular gene, leading to the expression of a neo T-cell epitope. Alternatively, carcinogens could only lead to clonal amplification of one single cell (Burnet, 1970), expressing a particular antigen, but usually present at insufficient levels on normal cells to be recognized by the immune system. This is observed with malignancies of the B-cell lineage expressing a private idiotope (Lynch et al., 1972). It is now well-accepted that carcinogens can induce the appearance of neoantigens, but it is not always clear whether the latter result from mutated cellular genes or from activated preexisting silent genes (reviewed in Schreiber et al., 1988). The methylcholanthrene-induced tumor MethA has been extensively analyzed (Deleo et d.,1977). In this case, it was not possible to generate tumor-specific CTLs which could select antigen-loss variants, and the immunity involves mainly helper T cells. While initial studies were concerned with serological specificities (Pravtcheva et al., 1981), further work used the transplantation assay to identify tumor rejection antigens. These investigations led to the characterization of several molecules that could be used to specifically inhibit the growth of the MethA tumor in immunized mice. From plasma membrane, or solubilized extracts, two groups of rejection antigens were identified, the 96-kDa (gp96) cell-surface glycoprotein (Srivastava et al., 1986) and the 84- to 86-kDa (p84/86) intracellular antigen (Ullrich et al., 1986). Interestingly, another methylcholanthrene-induced tumor (CMS.5) was shown to contain a tumor rejection activity associated with the 96-kDa fraction of the plasma membrane. The gp96 from these various sources could not be distinguished on the basis of biochemical characteristics. Furthermore, molecular cloning of the cDNA encoding gp96 revealed exactly the same sequence in MethA and CMS.5 tumors and in normal liver (Srivastava and Old, 1988).However, major regions of the gp96 gene still remain to be explored. The other proteins that conferred protection against tumor growth were two polypeptide isoforms of 84 and 86 kDa, displaying considerable homology with each other and with certain stress-induced or heat shock proteins (hsp), namely, the hsp90 family (Ullrich et al., 1986; Moore et al., 1987) or the hspl08 family (Srivastava et al., 1986,1987). The possibility has been raised that these molecules may not be

IMMUNE RESPONSE AGAINST TUMORS

287

immunogenic by themselves but rather serve as immunogenic peptide carriers (Srivastava and Maki, 1991). This was suggested for the gp96 molecules in relation with their capacity to load MHC class I antigens with peptides in the endoplasmic reticulum (Srivastava and Heike, 1991; Li and Srivastava, 1993). The identification of the human homologue of murine gp96 and its homology with hsp.lOO and 108reinforces this hypothesis (Maki et al., 1990). Interestingly, other members of hsp that belong to the hsp.70 family could also contribute, albeit to a lesser extent, to the immunogenicity of various tumors. By interfering with the antigen processing and presentation pathway and facilitating peptide/MHC class I1 interactions, they could enhance tumor antigen presentation and strengthen immune recognition of specific target molecules (Young, 1990; Kaufmann, 1990; De Nagel and Pierce, 1992,1993). Moreover, hsp7O was shown to play a key role in the nuclear transport of various oncogenes and antioncogenes, as mentioned for the T antigen of SV40 (Sawai and Butel, 1989),the adenovirus E1A proteins (White et al., 1988), and the p53 tumor suppressor molecule (Pinhasi-Kimhi et al., 1986). In this situation, stable complexes of mutated p53 and hsp7O (Lam and Calderwood, 1992) could lead to an increased half-life and immunogenicity of this tumor antigen (Davidoff et al., 1992). The antigenicity of the hsp molecules could also be accounted for by the existence of mutated epitopes. At any rate, it is remarkable that these highly conserved proteins should constitute target molecules of immune responses in both healthy and unhealthy individuals (reviewed in Young, 1990; Cohen, 1991; Kaufmann, 1990). In contrast to the MethA tumor, which induces only a relative and dose-dependent protection, the UV-induced tumor 1591 is much more immunogenic and is rejected by normal young mice, even after inoculation of large tumor fragments. Also, in contrast to the MethA tumor, CTL clones could be generated against the UV-induced tumor. This allowed for a characterization of the tumor antigens, by sequential in vitro selection of tumor variants having lost the ability to be lysed by given CTL clones. It was thus demonstrated that a single tumor may express several tumor-specific antigens that can be lost independently of each other (Wortzel et al., 1983,1984; Uytenhove et al., 1983). Two additional main features characterize the 1591 UV-induced tumor. First, there is a hierarchy in the antigenicity of the multiple antigens expressed by the same tumor cell. This results in antigenic changes and loss in vivo and in the appearance of more malignant variants leading to tumor progression (Urban et al., 1984). Second, tumor variants which could not retain a CTL-defined tumor specificity, nor induce protective immunity to itself, could still present a tumor-specific

288

CLAUDE ROTH ET AL.

antigen defined by T helper cells and cause protective immunity to any variant that expresses a CTL-defined antigen (Van Waes et al., 1986). One antigen of the 1591 tumor has been defined by monoclonal antibodies that recognize a unique specificity on “regressor” tumors. The gene was cloned and found to encode a novel class I molecule, with three antigenic specificities shared with other MHC molecules, namely Lq, Dq and K216 (Stauss et al., 1986a; Lee et al., 1988). While the Lq and Dq specificities were shown to be 100% homologous to the known Lq and Dq specificities at the gene level, the origin of the K216 specificity remains elusive, but it encodes an antigen which is sufficiently immunogenic to induce rejection by normal mice (Stauss et al., 198613). C. TUMOR-SPECIFIC PEPTIDES IDENTIFIED BY T-CELLRECOGNITION Tumor immunology then made considerable progress and this is largely due to the better understanding of the mechanisms of antigen presentation by MHC molecules and of T-cell recognition of peptide-MHC complexes. First, T cells were shown to recognize antigen only in the context of MHC products at the cell surface (Zinkernagel and Doherty, 1974,1979; Bevan, 1975) as small antigenic peptides located in a groove of the MHC molecules (Babbit et al., 1985; Townsend et al., 1985). The three-dimensional structures of human and mouse MHC molecules are now known (Bjorkman et al., 1987; Ejorkman and Parham, 1990)and it appears that class I and class I1 products share very similar structures (Brown et al., 1988,1993). Functional studies demonstrated that both CD4+ helper (TH) and CD8 + cytotoxic T lymphocytes (CTL) recognize short peptides derived from fragmented proteins which are presented by MHC class I1 or class I molecules (Babbit et al., 1985; Maryanski et al., 1986;Townsend et al., 1986a,1989; Buus et al., 1987; Schumacher et al., 1990; Van Bleek and Nathenson, 1990; Falk et al., 1991; Rudensky et al., 1991). Two major pathways of antigen processing and presentation for Tcell recognition have been identified. Antigens of exogenous proteins are usually presented to T cells by professional antigen-presenting cells, such as macrophages, B cells, bone marrow-derived dendritic cells (reviewed in Unanue and Allen, 1987), and other cells that express class I1 molecules (Rock et al., 1984; Lanzavecchia, 1985; Boog et al., 1988; Melief et al., 1988).This “exogenous” presentation pathway involves endocytosis of the foreign protein, followed by its degradation into short peptides in the endocytic compartments (reviewed in Brod-

IMMUNE RESPONSE AGAINST TUMORS

289

sky and Guagliardi, 1991).In general, antigenic peptides bind to class 11 molecules when the invariant chain ( Ii) is removed at low pH in the late endosome (Babbit et al., 1985; Buus et al., 1986,1987; Roche and Cresswell, 1990; Teyton et al., 1990). Class I1 MHC-peptide complexes then reach the cell surface, where they are specifically recognized by the receptor of T lymphocytes expressing the CD4 molecule, which binds to a constant region of class I1 product (Sprent and Webb, 1987). The other processing pathway concerns mostly endogenous proteins, peptides of which are loaded into the peptide presenting groove of MHC class I molecules in the endoplasmic reticulum. In this scheme, proteins originating from either virus infection (Townsend et al., 1985,1986a; Morrison et al., 1988; Yewdell et al., 1988), or osmotic lysis of pinosomes (Moore et al., 1988)or normal cellular components, which are present mostly in the cytoplasm or nucleus, are fragmented into short peptides, ususally eight to nine amino acids length, that are loaded onto MHC class I antigens after transport into the endoplasmic reticulum by MHC-encoded transport molecules (Powis et al., 1991,1992). MHC class I molecules are quite unstable if they are not loaded with a peptide, as shown with mutant cell lines altered in their transporter genes (Powis et al., 1991; Spies et al., 1992).T cells which recognize peptides presented by class I molecules express CD8, which binds to a constant region of class I MHC molecules (reviewed in Parnes, 1989). As a consequence, endogenous proteins are preferentially recognized by CD8+ CTL in the context of MHC class I molecules, whereas exogenous antigens are mainly recognized by CD4' TH cells in the context of MHC class I1 molecules (Morrison et al., 1988). However, there are also T helper cells specific for class I molecules (reviewed in Singer et al., 1987)and CTL specific for class I1 molecules (reviewed in Mills, 1986; Morrison et al., 1986). Likewise, in certain circumstances, exogenously administered peptides can induce MHC class I restricted CTL responses (Carbone and Bevan, 1989; Deres et al., 1989; Aichele et al., 1990; Schultz et al., 1991; Kast et al., 1991), and endogenous proteins can prime class 11-restricted helper T cells, as shown, for example, in a B-cell line transfected with a A light-chain gene (Bogen and Weiss, 1988).Peptides eluted from class I molecules are usually about nine amino acids long (van Bleek and Nathenson, 1990; Rotzchke et al., 1990; Schumacher et al., 1991), while peptides eluted from class I1 molecules are usually longer and heterogeneous in size. The peptide binding pocket of class I molecules, being closed on both sides, is believed to display more stringent peptide length

290

CLAUDE ROTH ET AL.

requirements than that of class I1 molecules (Falk et al., 1991; Brown et al., 1993). The first antigenic structures to be regularly detected on tumor cells as targets for T-cell immunity were viral antigens on virus-induced murine neoplasms (reviewed in Melief, 1992; Klein, 1991). Subsequently, two additional types oftumor-specific antigens, whose molecular nature had long remained elusive, could be distinguished: the first one comprises antigens derived from structurally abnormal proteins, while the other includes antigens resulting from the abnormal expression of a normal protein (reviewed in van der Bruggen and van den Eynde, 1992). D. TUMOR ANTIGENSON VIRUS-INDUCED TUMORS The first evidence of rejection antigens expressed on tumors induced by oncogenic viruses was based on early work on the polyoma virus (Sjogren et al., 1961; Habel, 1961).These studies demonstrated that, in contrast to the transplantation antigens of chemically induced tumors, polyoma virus-related-specific antigens have identical or crossreactive specificities, even if they are induced in different species (Hellstrom and Sjogren, 1966). For polyoma, for instance, this specific antigen was initially thought to play a key role in maintaining the tumorigenicity of tumor cells, as suggested from the inability to isolate polyoma antigen-negative variants (Sjogren, 1964). However, when these antigens are coexpressed on other tumors, such as Moloney virus-induced lymphoma, the polyoma antigen is lost without affecting the growth potential of the lymphoma. This demonstrated that a given tumor could acquire several rejection antigens independently of each other and that some of the antigens could be lost following passage of tumors in mice. These antigen-loss variants still maintain their malignant phenotype if the disappearing antigen is unrelated to the malignancy of the cell. For polyoma-induced tumors, two antigens have been identified, the middle T antigen (a 56-kDa protein) and the small T antigen (a 22-kDa protein), both of which are localized in the nucleus. In two sets of experimental immunization protocoles, it was observed that middle T antigen can act as a TSTA, since it causes the rejection of transplanted syngeneic polyoma tumors (Lathe et al., 1987; Ramquist et al., 1988). Experimental tumors induced by Gross virus have been shown to stimulate an MHC-restricted CTL response in grafted mice. An epitope was traced down to the gag polyprotein, and it could be demonstrated that mouse L cells transfected with the gag gene and the proper restriction elements (H-2Kb)could vaccinate against the graft (Abas-

IMMUNE RESPONSE AGAINST TUMORS

291

tad0 et al., 1985; Plata et al., 1986,1987).In the case of Friend leukemia virus, the virus-derived FBL-3 tumors were shown to induce a T-cell response against the gag or env viral components in C57B1/6 mice. Both helper T lymphocytes and cytotoxic T cells participated in the rejection of these antigenic tumors (Leclerc and Cantor, 1980; Greenberg et al., 1981). CD8+ cytotoxic T cells recognize a viral gagencoded epitope restricted by H-2Db class I molecules, whereas CD4+ helper T cells recognize a viral env-encoded epitope in the context of I-Ab class I1 molecules (Klarnet et al., 1989). Furthermore, studies using recombinant vaccines demonstrated that both antigens can prime T helper cells, while only the gag nuclear antigen can induce a CTL response (Earl et al., 1986; Klarnet et al., 1989),and both antigens can induce protection against leukemia. Likewise, in an adoptive therapy model, CD4+ and CD8+ populations directed against FBL-3 can eradicate disseminated disease, by a mechanism which involves the induction of tumoricidal macrophages activated by IFN-y (reviewed in Greenberg, 1991). As in the polyoma virus tumor system, the tumor antigens encoded by the genome of the simian virus 40 (SV40)are highly immunogenic. One of these virally encoded proteins, the large T antigen, has been involved in the transformation process leading to malignancy, presumably by disrupting the interaction between transcription factor E2F and the retinoblastoma tumor suppressor gene product (Chellapan et al., 1992). In C57B1/6 mice (H-2bhaplotype), five antigenic sites have been identified so far in large T, on the basis of their recognition by MHC class I-restricted CTLs (Tanaka et al., 1988; Tevethia et al., 1990). Four of the sites have been mapped using synthetic peptides and are H-2Db restricted while one site is H-2Kb restricted (Tanaka et al., 1988; Deckhut et al., 1992). To identify the molecular basis of tumor antigen loss variants, an SV40-transformed B6 mouse kidney cell line (K-0) was cocultivated with SV40 T-antigen site-specific CTL clones (Tanaka and Tevethia, 1988,1990).The resulting epitope-loss variants that were selected for their resistance against CTL clones contained amino acid substitutions which were still presented by MHC molecules but were no longer recognized by the CTL clones (Lill et al., 1992; Deckhut and Tevethia, 1992). These specific amino acid changes did not affect the MHC class I binding motifs, but disrupted the ability to provide a target for specific CTL, which could explain the escape of tumors from immunosurveillance. One could argue against this hypothesis, though, due to the existence of other T cells in the in vivo repertoire which might recognize the mutant tumor cells.

292

CLAUDE ROTH ET AL.

In other murine or human virus-induced tumors, various T-cell epitopes, which may be potential targets for immune intervention, were defined using CD8+ CTL clones. In the case of human adenovirus early region 1 (AdEQtransformed murine cells, H-2b-restricted CTL clones were found to be directed against a peptide sequence of the viral nuclear oncogene product ElA, presented in the context of the H-2Db class I molecule (Kast et al., 1989; Kast and Melief, 1991). The Ad5-E 1-transformed murine cells, which express this H-2Db-restricted peptide, are, however, incapable of causing tumors in immunocompetent animals, due to efficient surveillance by CD8+ CTLs (reviewed in Melief et al., 1989; Melief and Kast, 1990). No antigen-loss variants were isolated, suggesting that the expression of E1A is essential to maintain the transformed state. Moreover, in this experimental system, CTLs are highly specific, as shown by tumor regression, and longterm memory was induced in immunoincompetent mice after adoptive immunotherapy (Kast et al., 1989). In the case of the human Epstein-Barr virus (EBV), one peptide derived from the nuclear antigen 3 (EBNA-3) was originally characterized by CD8+ CTL clones in the context of HLA-B8 molecules (Burrows et al., 1990).Afterward, additional target antigens for CTL recognition were identified on transformed cells, using recombinant viruses that carry EBV genes. These strategies allowed the identification of EBNA-4- and EBNA-6-derived peptides, presented by HLA-A2 molecules (Gavioli et al., 1992,1993). A hierarchy was established with respect to the immunogenicity of these epitopes (Murray et al., 1992) and for HLA-A2-restrictedCTL peptides, only a few of which appeared to be immunodominant (Gavioli et al., 1993). A main feature of this oncogenic virus is related to its multiple escape mechanisms, which contribute to counteract an effective immune response, as shown by the high frequency of EBV-specific CTL precursors and their remarkable stability. First, it encodes for aprotein sharing many ofthe biological activities of IL-10 (Vieira et al., 1991),which can suppress lymphokine production by the TH-1 cells (Fiorentino et aE., 1989) and which is a potent stimulator for B-cell proliferation (Defrance et al., 1992; Rousset et al., 1992). Second, the emergence of variants having lost the immunodominant A2-restricted CTL epitope (De Campos-Lima et al., 1993) has been observed. This could result in a weaker T-cell response against the infected cells. In the case of papillomaviruses, which cause epithelial proliferative diseases in animals and in man, several epitopes, which induce a CTL response against tumor cells in viuo, have been identified. Recombinant vaccinia viruses expressing the E5, E6, or E7 early region proteins

IMMUNE RESPONSE AGAINST TUMORS

293

of the bovine papillomavirus BPVl were shown to elicit antitumor immunity in animals inoculated with transformed cells (Meneguzzi et al., 1990). In humans, recombinant vaccinia virus expressing E6 or E7, then peptides derived from the HPV type 16 oncogenes E6 and E7, selected for their binding motifs for H-2Kb and H-2Db (Falk et al., 1991), were used to identify a CTL epitope able to protect against a tumor induced by human papillomavirus type 16-transformed cells (Meneguzzi et al., 1991; Feltkamp et al., 1993). Interestingly, in this study, successful antitumor vaccination was achieved with a nonimmunodominant CTL epitope. This result underlines the necessity for identifying peptides which are not necessarily immunodominant (as reviewed by Sercarz et al., 1993),for antigenic determinants presented by MHC class I1 molecules) but which are obvious candidates for inducing an in vivo CTL response. Other virus-induced tumors in humans have been studied and are currently under active investigation. Human T-cell leukemia virus-1, which is, the causative agent of adult T-cell leukemia, does not require additional immunosuppression to be leukemogenic. Hepatitis B virus (HBV)seems to play an important role in the genesis ofprimary hepatomas (Feitelson et al., 1993; reviewed in Buendia, 1992). In this human tumor virus system, target cells have been produced by expression of the viral protein in immortalized human B-cell lines that can stimulate CTLs in vitro (Guilhot et al., 1992; Bertoletti et al., 1993).

E. ANTIGENSDERIVED FROM STRUCTURALLY ABNORMALPROTEINS 1. TurnWhile viral antigens were relatively easy to identify, it has been much more difficult to characterize those tumor rejection antigens which are encoded by the cellular genome. Originally, clonal mouse and tumors were mutagenized with N-methyl-N’-nitrosoguanidine, variants that were no longer tumorigenic in syngeneic animals were isolated (Boon, 1983). Such tum- variants have been obtained in various tumor models, including a teratocarcinoma (Boon and Kellerman, 1977), Lewis lung carcinoma (van Pel et al., 1979),mastocytoma P815 (Uyttenhove et a1., 1980),and radiation-induced or spontaneous leukemia (van Pel and Boon, 1982; van Pel et al., 1983). In the P815 model, the tumor antigens are recognized by CD8+ cytotoxic T lymphocytes, leading to the rejection of the tum- tumors (reviewed in Boon, 1992). Anti-P815 CTL clones were used to define the antigens expressed by the original tum+ tumor (Pl).By transfection of the highly transfectable P1 tum+ cell line (P1 HTR) with the DNA

294

CLAUDE ROTH ET AL.

of the tum variant, it was found that tum antigens could be expressed on the transfected target cells, thus rendering them sensitive to lysis by specific CTLs (DePlaen et al., 1988).This led to the identification ofthe gene coding for the tum antigen, P91A, and it was demonstrated that a single point mutation is responsible for the antigenicity of the tum- peptide (Lurquin et al., 1989). Using a similar strategy, the gene encoding the tum- antigen specific for the P35B tum- variant was cloned. Again, it appeared to be a mutant differing from the wild-type gene by a point mutation in exon 5 (Szikora et al., 1990). For the P91A epitope, the mutation allowed the peptide to bind to the Ld molecule by generating an agretope. This was not the case for the P35B antigen, since both the wild type and mutated peptides were able to bind to Ddmolecules and to render target cells sensitive to lysis by the specific CTL (Szikora et al., 1990).A third tum- antigen, P198, was identified and shown to be derived from a normal gene bearing a point mutation in one of the exons. Both peptides corresponding to the normal and the mutated sequences could bind to the Kd molecule, but only the tum - P198 mutated antigen was recognized by anti-P198-specific CTLs (Sibille et al., 1990). These results indicated that many cellular genes can mutate to generate tum- variants, which explained their high rate of appearance (about This was the first rigorous demonstration that mutations induced by a mutagenic treatment can generate tumor rejection antigens, specific for each tumor, either by creating an agretope that allows apeptide to bind to MHC class I molecules or by creating a new epitope that competes with the normal peptide for binding to MHC class I molecules (reviewed in Boon, 1992). In mice and in humans, several other tumor antigens that derive from mutated proteins and are recognized by T cells have now been identified. Indeed, there is now mounting evidence that peptides derived from the protein products of altered oncogenes and tumor suppressor genes are associated with the MHC class I or class I1 molecules and thus may provide potential candidates for inducing specific immunity against cancer cells. The protein products of altered oncogenes and tumor suppressor genes (TSG) are interesting potential targets for a T-cell-mediated immune response for at least three reasons. First, alterations in these genes are obligatory events in the carcinogenic process and are thus present in every malignant cell. It has been estimated that, in the development of colorectal cancer, a total of five to eight different “genetic hits” affecting oncogenes and/or tumor suppressor genes have to occur in the same cell to induce the complete transformation from normal mucosa to malignant carcinoma (Reviewed

IMMUNE RESPONSE AGAINST TUMORS

295

in Fearon and Vogelstein, 1990). The emergence of altered peptides derived from the protein products of oncogenes and tumor suppressor genes and potentially presentable in an MHC class I or I1 context thus appears as a general phenomenon in cancer. Second, some of the most common mechanisms of oncogene activation, such as point mutation and chromosomal rearrangement, create protein sequences that do not exist in nontransformed cells (Bishop, 1991). Similarly, the common mechanism of inactivation of a tumor suppressor gene is chromosomal deletion of one allele and point mutation of the second allele, thus creating sequences that are specific of the cancer cells (Marshall, 1991). The induction of a strong immune response against these “foreign” sequences in a therapeutic context should thus not face problems associated with tolerance or autoimmunity. Finally, since the expression of altered oncogene and tumor suppressor gene proteins confers a selective growth advantage to the cancer cell, one escape mechanism in response to immune intervention, namely the downregulation of the expression of the altered gene, might lead to the same therapeutic effect as successful targeting by the Tcell response. Although more than 10tumor suppressor genes and almost 100 oncogenes have been defined (Bishop, 1991), relatively little information is currently available on immune responses directed against these cancer-specific proteins, or on the induction of specific immunity against oncogene and tumor suppressor gene proteins. There is, however, a growing interest in possible immune intervention directed against these “cancer genes”; naturally occurring or induced T-cellmediated immunity against the p53 tumor suppressor gene, the bcr/ abl tyrosine kinase, and the ras-proteins has been reported.

2 . Mutated p53 Proteins Sporadic mutations in the p53 tumor suppressor gene are the most common genetic alterations observed in human cancers (Levine et al., 1991) and approximately 70% of colon cancers, 30 to 50% of breast cancers, 50%of lung cancers, and almost 100% of small-cell carcinomas of the lung harbor these mutations (Hollstein et al., 1991). Germline mutations of the p53 gene have been described in “hereditary cancers” such as in patients with Li-Fraumeni syndrome (Malkin et al., 1990; Srivasta et al., 1990; Santibanez-Koref et al., 1991), in young patients with more than one malignant neoplasm (Malkin et al., 1992), and in adults with sarcoma with a family history of cancer (Toguchida et al., 1992). A computerized scoring system for predicting peptide binding to

296

CLAUDE ROTH ET AL.

HLA-A2.1 molecules has been applied to the p53 gene (Houbiers et al., 1993). The 393 amino acids of wild-type p53 and the 20 amino acids around each of the 32 published point mutations in colorectal and ovarian cancer were scanned, which resulted in a set of 63 peptides that scored above a predefined level indicating presumptive binding to HLA-A2.1. Peptides were synthesized and about 40% of them actually bound. The human processing defective cell line 174CEM.T2 was then loaded with some of these peptides and utilized as antigenpresenting cells to induce CD8+ CTL clones. These clones were capable of specifically lysing target cells loaded with the respective wildtype or tumor-specific mutant p53 peptides (Houbiers et al., 1993).The authors concluded that p53-specific CTL clones might be potentially useful for cellular immunotherapy of cancer. Yanuck et all. (1993) used another approach to demonstrate the immunogenicity of peptides derived from mutated p53 sequences. Spleen cells of BALB/c (H-2d)mice were pulsed with a 21-amino acid peptide encompassing a point mutation (135Cys to Tyr) in the mutant p53 from a human lung carcinoma and then injected intravenously into syngeneic animals. This immunization protocol generated CD8+ CTL that were capable of specifically killing mouse fibroblasts transfected with the complete mutated human p53 gene, whereas fibroblasts expressing the nonmutated human p53 were not recognized (Yanuck et al., 1993). Since the level of expression of the transfected p53 gene was comparable to that seen in human tumors and leukemias, it was proposed that peptide immunization against mutant p53 could be used for cancer immunotherapy. 3. The bcrlabl Oncogene

More than 95% of patients with chronic myelogenic leukemia (CML) carry the t(9;22)(q34;qll)translocation. This translocation of the c-abl protooncogene (abl) on chromosome 9 to the breakpoint cluster region (bcr) on chromosome 22 produces the Philadelphia chromosome and leads to the formation of a fusion gene termed bcr/abl, which encodes a 210-kDa chimeric protein with abnormal tyrosine kinase activity (Shtivelman et al., 1985; Hermans et al., 1987). In 10% of children and approximately 25% of adults with acute lymphoblastic leukemia (ALL), the abl gene is also translocated to chromosome 22, but to a different region of the bcr gene, and detection of the bcr/abl fusion mRNA in ALL has been associated with poor prognosis (Maurer et al., 1991). The abl and bcr genes are expressed by norma1 cells and thus the encoded proteins are probably nonimmunogenic. However, the join-

IMMUNE RESPONSE AGAINST TUMORS

297

ing region segment of the bcr/abl chimeric protein is composed of unique sequences of abl amino acids joined to bcr amino acids that are expressed only by malignant cells. Synthetic peptides corresponding to the bcr/abl joining region have been used to immunize BALB/c and C57B1/6 mice by repeated subcutaneous injections together with complete Freud's adjuvant (Chen, W. et al., 1992). The immunization protocol elicited peptide-specific, CD4+, class I1 MHC-restricted T cells that recognized only the combined bcr/abl sequences but not bcr or abl sequences alone. The proliferative response of these T-cell clones to the whole bcr/abl protein presented by syngeneic spleen cells demonstrates that processing of the fusion protein and binding of joining-region peptides to MHC molecules also occurs in a less-artificial system (Chen, W. et al., 1992). The bcr/abl protein thus represents a potential tumor-specific antigen related to the transforming event and expressed in the large majority of patients with CML. Further studies will indicate if an immune reaction against bcr/abl dependent on cytotoxic CD8+ T lymphocytes can also be induced. 4 . The ras Oncogenes Mutations of the ras protooncogenes are among the most frequent alterations found in human malignancies. Thus, mutations of the Kiras, N-ras, or Ha-ras genes are found in approximately 90% of pancreatic carcinomas (Almoguera et al., 1988),50% of colorectal carcinomas (Bos et al., 1987a; Forrester and Almoguera, 1987), and 25% of acute myelogenic leukemias (Bos et al., 198713; Farr et al., 1988), but only in less than 5% of breast carcinomas (Rochlitz et al., 1989). Ras oncogenes are activated by point mutations that occur almost exclusively in codons 12, 13, and 61 (Barbacid, 1987). These welldefined and easy to diagnose genetic alterations can be conveniently included in immunization strategies, and this is probably the reason why the ras genes were the first oncogenes whose role as potential targets for immunotherapy of cancer was evaluated. Using a synthetic peptide corresponding to amino acids 5-16 of a mutated ras protein with an exchange of the normal Gly at position 12 by Val, Jung and Schluesener (1991)first demonstrated that specific CD4+ T-cell lines could be derived from the blood of two of four healthy donors analyzed. These T-cell lines were generated by continuous in vitro stimulation of peripheral blood cells and showed no crossreactivity to normal, nonmutated p2lras proteins. Similarly, Peace et al. (1991) showed the presence of class I1 restricted, specific T cells in the repertoire of healthy C57/B16 mice immunized with Freund's

298

CLAUDE ROTH ET AL.

adjuvant plus mutated ras peptides with an Arg substituted for Gly at position 12 (Peace et al., 1991). Several other mutated ras peptides and mouse strains with different H-2 backgrounds have been evaluated. Some, but not all, are immunogenic in individual strains of mice, presumably reflecting the ability of individual peptides to be efficiently presented by the MHC molecules ofthe host. The same authors have reported preliminary results on the induction of class I MHCrestricted CD8+ CTL, specific for mutated ras peptides, using the B6 leukemia cell line of the C57/B16 (H-zb)mouse (Peace et al., 1992). In line with the results of Jung and Schluesener (1991),Gedde-Dahl et al. (1992a) have shown that the repertoire of a single healthy donor contained T-cell clones specific for six of nine mutated peptides tested. All of these peptides were capable of binding to HLA-DQ molecules and several to HLA-DR molecules (Gedde-Dahl et al., 1992aJ993b). Four T-cell clones were established and their fine specificity determined: two of the clones recognized peptides presented by DR2, and the other two recognized peptides presented by DQ6 (Gedde-Dahl et al., 1993a).An interesting observation was the occurrence of an HLADQ8-restricted clone, specific for a codon 61 Gln to Leu mutation of p21 ras, in a patient with thyroid cancer (Gedde-Dahl et al., 1992b). The authors could not find the corresponding codon 61 mutation in the thyroid carcinoma cells, and they speculate that this might be due to the prior elimination of the mutation-bearing cells during tumor progression. Another interpretation, however, is that the isolation of the clone specific for the mutated codon 61 is accidental and that it belongs to the normal repertoire of the cancer patient, as shown for healthy donors in the studies mentioned above. An encouraging study has involved immunization of C57/B16 mice by intraperitoneal injection of mutant H-ras proteins. A CD8+ CTL activity was demonstrated in the splenocytes of the treated animals against cells carrying Arg instead of Gly in Ha-ras position 12. More importantly, 9110 animals immunized with the mutant protein were protected against subsequent inoculation of lo5 Ha-Balb fibroblasts, a malignant cell line that expresses an activated Ha-ras gene with the same Gly to Arg mutation in position 12. Only 2/10 mice immunized with nonmutated ras protein and 0/10 immunized with ovalbumin were protected against the Ha-Balb cells (Fenton et al., 1993).A drawback of this study was the fact that the same experiments did not work when Ha-ras with Val instead of Arg at position 12 was used to induce immunity. In conclusion, altered oncogenes and tumor suppressor genes are potentially important targets for the immunotherapy of cancer. Not all

IMMUNE RESPONSE AGAINST TUMORS

299

of the mutated oncogenehmor suppressor gene peptides analyzed so far seem to be immunogenic in all circumstances, but strategies to increase immunogenicity might overcome this problem. Clinical studies exploiting the immunogenicity of these cancer-inducing genes are currently under way.

F. ANTIGENSDERIVEDFROM THE OVEREXPRESSION OF NORMALPROTEINS Another mechanism may account for the expression of a tumor antigen on transformed cells. If a gene is silent, or expressed at a low level, in the normal tissue, and if it is activated in tumor cells, the overexpression of the unmutated self-proteins may lead to the production of a set of presentable self-peptides which are potential targets for antitumor CTLs. In the murine mastocytoma P815, comparison of the specificities expressed by tum- variants that are selected either in vitro by coculture with CTL clones or in vivo by passage in syngeneic DBA/2 mice led to the identification of five distinct rejection antigens, named P815 A, B, C, D, and E. Because antigens A and B are usually lost together by tumor cells that escape tumor rejection in vivo, they were identified as dominant antigens that play a significant role in the antitumoral response. Using the same strategy as for molecular cloning of the tumantigens (i-e., transfection of the P1A gene in P1A- B - antigen-loss variants), transfectants were isolated that express both A and B rejection antigens, recognized by CTL clone. This strategy allowed the identification of the P1A gene which encodes both P1A and P1B antigens presented by the Ld restriction element (van den Eynde et al., 1991). The sequence of the P1A gene, which differs completely from the genes coding for Tum- specificities, is identical in tumor cells and in normal mouse tissues. A nonapeptide was synthesized from the coding sequence of P1A and shown to carry specificities for P1A and PlB. Antigen-loss variants of P815 having one mutation were shown to express only one of the two epitopes (Lethe et aZ., 1992). The PlA gene is silent in precursors of most cell lines but is transcribed at a very low level in normal adult tissues. Several hypotheses have been proposed to account for the antitumor immune response, which occurs independently of any observable autoimmune reaction (reviewed in Boon, 1992). P l A might be expressed at an early stage in embryonic tissues before the establishment of natural tolerance. Alternatively, it could be transiently expressed by some cell precursors at a given stage of their differentiation. In any case, normal cells by themselves cannot induce an immune response, and it cannot be ruled

300

CLAUDE ROTH ET AL.

out that the antitumoral reaction does not eliminate precursor cells that express the antigen, without visible damage to the organism. Similar experiments, using tumorigenic cell lines obtained by transfection with oncogenes, are in agreement with this observation and confirm that activation of a normally silent gene leads to the expression of a potential tumor rejection antigen (Torigoe et al., 1991). The neu/erbB2 oncogene is a transmembrane receptor with a high degree of homology to the epidermal growth factor receptor EGF-R (Schechter et al., 1985). Amplification and/or overexpression of neu/ erbB2 occurs in approximately 20-30% of human ovarian and breast carcinomas and has been associated with poor prognosis in both diseases (Slamon et aZ., 1987,1989; Allred et al., 1992; Gasparini et al., 1992). Preliminary data indicate that tumor-infiltrating lymphocytes isolated from human ovarian carcinoma can recognize peptides corresponding to neu/erbB2 epitopes and’lyse targets pulsed with these peptides (Ioannides et al., 1992). Similarly, CTL have been cloned from the peripheral blood of patients with ovarian and breast cancer that seem to be specific for peptides derived from the neulerbB2 oncogene (Tim Eberlein, Harvard Medical School, personal communication). Thus, the mechanism of activation of the neu/erbB2 oncogene is overexpression of an unaltered protein, and specific T-cell reactions can be raised against it. As above with the P815 antigen, the unresponsiveness against the “self-antigen” in normal cells might result from anergy or suppression of established specific CTL, absent or insufficient antigen processing of the antigen in low expressors, or clonal deletion of neu/erbB2 reactive T cells. One should bear in mind that attempts to induce a specific immune response against malignant cells overexpressing neu/erbB2 do, of course, carry a certain risk of also inducing an autoimmune response. In human tumors, several groups have reported the existence of such specific tumor antigens that are recognized by autologous CTL clones (Hainaut et al., 1990; Knuth et al., 1992). By immunoselection with these CTLs, variants having lost several antigens have been prepared, which allowed the definition of some antigenic specificities in human melanomas (Knuth et al., 1989; van den Eynde et al., 1989; Wolfel et al., 1989; Topalian et al., 1990). Three different epitopes presented by HLA-A2 molecules were identified in a human melanoma tumor. HLA-A2 is the most widespread MHC class I allele in Caucasian populations, and the HLA-A2 molecule is a common restriction element for melanoma-specific effector T cells (Kawakami et al., 1992; Topalian et al., 1989; Slingluff et al., 1993; Viret et al., 1993).Furthermore, it was shown in a study that the antigen expressed

IMMUNE RESPONSE AGAINST TUMORS

301

by melanoma cells and restricted by HLA-A2 molecules is also expressed by HLA-A2 melanocytes derived from normal tissues (Anichini et al., 1993). Biochemical studies have led to the identification of a tyrosinasederived peptide as one of several peptides isolated from an HLAA2 melanoma (Slingluff et al., 1993). Using the previously described strategy which allowed the identification of the P1A gene, T. Boon and his co-workers have prepared a genomic library from DNA of a human melanoma (MZ2-MEL).The library was transfected into a variant which had lost one of the six distinct antigens on the original tumor and was recognized by autologous CTL clones. Transfectants expressing antigen MZ2-E were selected based on their ability to induce the production of tumor necrosis factor by an anti-E CTL clone (Traversari et al., 1992). This led to the isolation of the MAGE-1 gene, which encodes the MZ2-E antigen (van der Bruggen et al., 1991). The sequence of the MAGE-1 gene does not correspond to any previously known protein. It encodes a 30-kDa protein that contains a nonapeptide that binds to the HLA-A1 molecule, and this is recognized by the anti-E CTL clones. While not expressed by normal tissues, except sperm, this gene is expressed by different tumors, such as melanomas (40%),lung cancer, head and neck tumors, sarcomas, and breast tumors. Since about 10% of melanoma patients should express the antigen MZ2-E restricted by the HLA-A1 alIele, vaccination protocols are currently being assessed, using either irradiated cells expressing the MZ2E antigen or genetic constructs expressing the antigen, the HLA-A1 molecule, and/or interleukine, that could improve the antitumor immunity in these melanoma patients. Subsequently, a similar approach has been used to define antigens recognized by T cells in the context of HLA-A2 molecules. At least two antigens have been isolated, a tyrosinase or tyrosinase-associated molecule (Ab) and a molecule termed Aa (Brichard et al., 1993; Wolfel et al., 1993). In one study, another strategy was used to identify human melanoma peptides that are recognized by class I-restricted tumor-infiltrating T lymphocytes (Storkus et al., 1993). Using elution from HLA-A2+ or HLA-A2- melanoma lines by repeated acid treatment, different tumor peptides isolated by fractionation on HPLC were shown to sensitize HLA-A2 nonmelanoma targets for lysis by melanoma-specific CTLs. The relationship between these peptides and the peptide isolated by Slingluff et al. from melanoma cells by affinity chromatography on HLA-A2 molecules still remains to be determined (Slingluff et al., 1993; Storkus et al., 1993). Interestingly, some of these peptides were

302

CLAUDE ROTH ET AL.

shown to bear the HLA-A2 binding motif, as already predicted from the sequence of self-peptides eluted from MHC molecules (Falk et al., 1991). In other human tumors, there is mounting evidence that autologous CTLs can recognize antigens on tumor cells, such as squamous cell carcinoma of the head and neck (Yasumura et d . ,1993), or renal-cell carcinomas bearing the HLA-A2 restriction element (Schendel et al., 1993)or ovarian tumors bearing the HLA-B5 molecule (Wang et al., 1992). G. THECASEOF MUCIN Several groups, looking for antigens preferentially expressed by malignant cells, have generated tumor-specific monoclonal antibodies that recognize an epitope on the mucin, an epithelial cell glycoprotein, produced by breast and pancreatic carcinomas, but not by the corresponding normal tissues (Girling et al., 1989; Barnd et al., 1989). The MUC-1 antigen, which corresponds to the heavily glycosylated mucin, consists of multiple tandem repeats of a 20-amino acid sequence with abundant 0-linked carbohydrate side chains, the protein core being identical to the one expressed on the normal tissues (Gendler et al., 1988; Burchell et al., 1989; Wreshner et al., 1990; Hareuverni et al., 1990a). The antigenic epitope expressed as the epithelial tumor antigen in a recombinant vaccinia virus was shown to provide a true target molecule for immune responses against tumor cells bearing this antigen (Hareuveni et al., 1990b). Furthermore, the originality of this tumor antigen resides in its capacity to stimulate specific CTLs in an MHC-unrestricted manner, a finding attributed to its highly repetitive nature (Finn, 1992a). In a first attempt, human CTLs were generated from lymph nodes of patients with pancreatic carcinomas (Barnd et al., 1989; Finn, 1992b) and breast carcinomas (Jerome et al., 1991), using allogeneic mucin-producing tumor cell lines as stimulators. These data were then repeated using presenting cells transfected with MUC-1 cDNA as stimulators and targets for mucin-specific CTLs (Jerome and Finn, 1992; Jerome et al., 1993). The results confirmed original observations and demonstrated that both syngeneic- and allogeneic-presenting cells can stimulate mucin-specific CTLs. Furthermore, to determine the precise peptide specificity of TCRs on cloned CTLs, an expression vector that contains only two repeats was constructed, and transfected cells with this vector were shown to be recognized efficiently by specific CTLs (Bu et al., 1993). It was thus postulated that target cells would express a two-repeat molecule that is not extensively glycosylated, leading to a more effective presentation of the CTL epitope. This is in agreement with the observation that a

IMMUNE RESPONSE AGAINST TUMORS

303

major group of T-cell clones requires inhibition of mucin glycosylation for target cell recognition (Jerome et al., 1993). In order to evaluate the immunogenicity of mucin epitopes, peptides derived from the mucin tandem repeat or the whole mucin molecule were assayed for their ability to induce a protective immunity against tumor growth (Ding et al., 1993; Acres et al., 1993). Since mucin is a self-antigen that is preferentially recognized on tumor cells, it should be a good candidate for a tumor-rejection antigen in vaccines, which could be further modified to secrete locally cytokines that stimulate effective helper and cytotoxic T-cell responses. 111. Why Are Tumors Poorly Immunogenic?

A. Low ANTIGENOR MHC CLASS I EXPRESSION Downregulation of MHC class I expression has been found to occur in many animal and human leukemias, lymphomas, and solid tumors, as extensively reviewed elsewhere (Browning and Bodmer, 1992; Moller and Hammerling, 1992; Schrier and Peltenburg, 1993). The notion is generally accepted that loss of MHC expression and the subsequent inability of the cells to present tumor-specific peptides to CTL clones can lead to an escape of the respective tumor cells from immunological control. This notion is supported by the observation that transfection of MHC class I genes into aggressive mouse tumor cell lines with low or no endogenous expression of these genes leads to tumor rejection when mice are transplanted with the transfected cells (Hui et al., 1984; Tanaka et al., 1985). A reciprocal experiment in the murine thymoma AKR showed that tumor growth was promoted by transfection of antisense DNA suppressing endogenous H-2KkMHC expression (Hui et al., 1991). These and similar studies are discussed in more detail in Section IV,A of this review. A variety of mechanisms through which tumor cells can lose MHC expression have been discussed. They include loss of peptide transporter genes in murine RMA-S cells (Franksson et al., 1993) and several human cancers (Restifo et al., 1993),altered binding of regulatory factors to HLA class I enhancer sequences (Blanchet et al., 1992), a frameshift mutation of the P2-microglobulin gene in the class Inegative melanoma cell line SK-MEL-33 (Wang et d . , 1993), among other mechanisms (Chapman and Houghton, 1993). The “immune evasion” theory, however, is not the only possible explanation for MHC loss by tumor cells, and other interpretations

304

CLAUDE ROTH ET AL.

should be considered. For example, nonimmunological, selective advantages may accompany low class I expression, such as an intrinsic growth advantage for low expressors (Haliotis et al., 1990).Also, MHC class I genes are expressed at different levels in various cells and tissues, so that their level of expression can be taken as a differentiation trait. In a study of Burkitt lymphomas, loss of HLA class I expression was strictly correlated with B-cell differentiation and could thus be considered as a phenotypical trait of the tumor rather than a means to escape immune surveillance (Anderson et al., 1991). Similar findings have been reported in human urothelial tumors (Ottesen et al., 1987; Tomita et al., 1990). Finally, MHC loss might be a consequence of the oncogenic process itself, i.e., of alterations in oncogenes and tumor suppressor genes. This has been shown for the transformation of rodent cells by the adenovirus E1A oncogene, whose expression is responsible for MHC class I downregulation. Similarly, the c-myc and N-myc genes are capable of switching off the expression of MHC class I genes in human melanomas and neuroblastomas, respectively (reviewed in Schrier and Peltenburg, 1993). In several tumors, low MHC expression was found more frequently in metastatic than in primary lesions (Cordon-Cardo et aE., 1991; Pantel et al., 1991), or was associated with bad prognosis, although the evidence on this last point is controversial (Lopez-Nevot et al., 1989; Moller and Hammerling, 1992). These findings can be interpreted as arguments in favor of the “immune evasion” theory but would also fit with m y of the aforementioned explanations for MHC loss. In contrast to complete loss of HLA expression on human tumors, loss of the expression of individual HLA alleles argues in favor of immunoselection of a tumor cell clone (Browning and Bodmer, 1992; Schrier and Peltenburg, 1993) and has been detected in a large percentage of colorectal carcinoma and other epithelial and nonepithelial neoplasms (Momburg et al., 1989; Natali et al., 1989; Smith et al., 1989; Pandolfi et al., 1991; Kaklamanis et al., 1992). Furthermore, there is growing evidence that different MHC alleles either predispose to or protect against the development of certain virally induced tumors (Klitz, 1992). Such MHC associations have been found in Hodgkin’s lymphoma (Bodmer et al., 1989), nasopharyngeal carcinoma (Lu et al., 1990), squamous cell carcinoma of the cervix (Wank and Thomssen, 1991), and rabbit viral papillomas (Han et al., 1992). Similarly, renal transplant recipients treated with long-term immunosuppressive therapy have an increased risk of skin cancer only when they carry certain HLA alleles, while others (HLA-All) seem to partially protect against the development of skin tumors (Bouwes Bavinck et al., 1991; Browning and Bodmer, 1992). It is possible that the ability of a given

IMMUNE RESPONSE AGAINST TUMORS

305

HLA allele to present certain viral immunogenic peptides to stimulate and/or be recognized by CTL clones determines whether a transformed virus-infected cell develops into a tumor or is attacked and removed by cytotoxic cells. Loss of MHC class I expression by tumor cells may be a major obstacle in immune therapy trials designed to activate CTL responses against the tumor. The question of whether MHC class I gene expression remains activatable by IFN-y, TNF-a, or other agents therefore becomes a significant issue. It has also been observed in several mouse models that even modest levels of MHC class I-petide complexes, which do not allow the induction of specific CTL, are nevertheless sufficient to permit specific lysis by activated CTL (Shastri and Gonzalez, 1993). B. LACKOF COSTIMULATORY MOLECULES (B7-CD28/CTLA-4) Maximal stimulation of a T-cell response to a peptide/MHC complex, requires a second, costimulatory signal, usually delivered by antigen-presenting cells, i.e., activated B cells, macrophages, or dendritic cells. A suboptimal proliferative T-cell response can be enhanced by different soluble or membrane-bound molecules, such as nutrients, hormones, trophic factors, interleukins, and cell adhesion molecules (reviewed by Schwartz, 1990). Occupancy of the TCR in the absence of a costimulatory signal induces a state of unresponsiveness in the T cell, characterized by a lack of IL-2 production and proliferation in response to subsequent exposure to antigen, even in the presence of a costimulatory signal. This long-lasting, antigenspecific state of unresponsiveness has been termed anergy (Schwartz, 1990). Although the molecular basis ofT-cell anergy is not fully understood, evidence suggests that one such costimulatory pathway involves the interaction of the T-cell surface antigens CD28 and CTLA-4 with their ligand B7, expressed on APCs (Freeman et al., 1991; Gimmi et al., 1991; Linsley et al., 1991). CD28 is constitutively expressed on 95% of human CD4+ T cells, 50% of CD8+ T cells, and thymocytes coexpressing CD4 and CD8, and its regulates T-cell cytokine production by transcriptional and post-transcriptional mechanisms ( Jenkins and Johnson, 1993). CTLA-4 is a second ligand, with approximately 20fold higher avidity than CD28 for B7, and is virtually undetectable on resting T cells but is expressed on activated CD4+ and CD8+ T-cell subsets at -1/30-50 the levels of CD28 (Linsley et al., 1992). B7, like CD28 and CTLA-4, is a member of the immunoglobulin gene superfamily and is expressed on activated B cells and macrophages and Constitutively on dendritic cells (Schwartz, 1992). The function

306

CLAUDE ROTH ET AL.

of CTLA-4 on activated T cells may be to promote effective interaction between CD28 and B7. Low-abundance, high-avidity CTLA-4 molecules might facilitate interactions between higher abundance, lower avidity CD28 molecules and the B7 counter receptor (Linsley et al., 1992). Blocking costimulation of the CD28 receptor in the continued presence of a foreign antigen can lead to T-cell unresponsiveness. This has been shown in experiments achieving long-term survival of xenogeneic pancreatic islets when B7 was blocked by the high-affinity inhibitor CTLA-41g (Lenschow et al., 1992). The induction of T-cell activation and proliferation is thus a twostep process. First, cognate recognition of a specific peptidelMHC complex by the T-cell receptor leads to the activation of tyrosine kinases in the T cell and to signal transmission by the cytoplasmic tail of the MHC molecule, which causes increased expression of B7 on the antigen-presenting cell (Nabavi et al., 1992; Baskar et al., 1993). Second, B7 interacts with its natural counterreceptors, CD28lCTLA-4, on the T cell; induces an independent stimulus, probably uia tyrosine phosphorylation of specific substrates, including phopholipase C-1; and triggers calcium-dependent and calcium-independent signals. This subsequently leads to the production of IL-2 and other cytokines and to proliferation of the specific T-cell clone (Freeman et al., 1991; Gimmi et al., 1991; Linsley et al., 1991; Reiser et al., 1992; reviewed in: Jenkins and Johnson, 1993; Linsley and Ledbetter, 1993). In contrast, when peptide antigen is presented by MHC in the absence of €37, T-cell clones are unable to secrete detectable levels of IL-2 or proliferate (Gimmi et al., 1993). In model systems, this anergy can be reversed, even after its full induction, by IL-2 at very high doses, although four or more division cycles are required for complete reversal (Schwartz, 1992). CD28 is not required, however, for all T-cell responses in uivo, since CD28-deficient mice were capable of mounting a CTL response and delayed-type hypersensitivity after being infected with lymphocytic choriomeningitis virus (Shahinian et al., 1993), suggesting that alternative costimulatory pathways may exist, at least in this model. More recently, in agreement with this hypothesis, another surface molecule, that is distinct from B7 and abundantly expressed on activated B.cells, was identified as the predominant ligand for the T-cell activation molecule CTLA.4 (Freeman et al., 1993a,b; Hathcock et al., 1993). This CTLA.4 counterreceptor named B7.2 seems to provide a critical early costimulatory signal for T cells. The importance of the discovery of the B7,CD28/CTLA-4 pathway for tumor immunology lies in the fact that most tumors are derived

IMMUNE RESPONSE AGAINST TUMORS

307

from parenchymal or mesenchymal cells that do not express B7. It was thus speculated that the ability of tumor cells to effectively present tumor-specific antigens (TSA) might be severely impaired and that lack of costimulation could be an important mechanism confering low immunogenicity even to tumor cells expressing MHC and TSA (Schwartz, 1992). Tumor reactive T-cell clones might thus receive inadequate costimulation and enter a state of anergy rather than eradicating the tumor. The introduction of the B7 gene into tumors might result in effective costimulation and tumor immunogenicity in the immunocompetent host. This has been shown in several examples (Chen, L. et al., 1992; Baskar et al., 1993; Harding and Allison, 1993; Townsend and Allison, 1993) and is discussed below (Section IV,B). C. IMMUNE SUPPRESSIVE FACTORS (TGFP AND IL-10) Besides changes in the tumor cells themselves that lead to decreased immunogenicity, as discussed above, it has long been postulated that malignant cells may secrete immunosuppressive factors that could contribute to immune evasion. A variety of immunosuppressive molecules have been found in the culture supernatant of tumor cells, as well as in the sera and effusions of animals and patients with cancer. Although the active molecules have frequently not been purified to homogeneity, they were shown to inhibit a variety of immune reactions: delayed-type hypersensitivity (DTH); macrophage accumulation at sites of inflammation; macrophage chemotaxis, phagocytosis, and cytotoxicity; skin graft rejection; lymphocyte proliferation in response to mitogenic stimuli; antibody synthesis; generation of LAK, TIL, and CTL cells; and lymphokine production (reviewed in Sulitzeanu, 1993). Probably the best characterized among cancer-related immunosuppressive molecules is the transforming growth factor (TGFP). It was originally identified due to its ability to confer a transformed phenotype to normal fibroblasts, but it also exerts a strong inhibitory effect on malignant and nonmalignant epithelial cells (Barnard et al., 1990) and potent immunosuppressive activity. TGFP inhibits the activity of a number of other cytokines, including IL-2, IL-4, IFN-y, and TNFa.It inhibits the proliferation of T and B cells as well as the generation of LAK and CTL. In addition, TGFP blocks NK cytolytic activity and downregulates the expression of the IL-2 receptor and of MHC molecules (Sulitzeanu, 1993).TGFP may thus be responsible for many of the immunosuppressive activities that have been initially attributed to “suppressor” lymphocytes and macrophages (Sulitzeanu, 1993). TGFP was originally identified as an immunosuppressive factor in

308

CLAUDE ROTH ET AL.

the sera of patients with glioblastoma (reviewed in Siepl et al., 1988), but its mRNA was later detected in a large number of tumors (Derynck et al., 1987). The TGFP protein could be detected by immunocytochemistry or Western blot analysis in cell lines established from colorectal cancer (Coffey et al., 1986), breast carcinoma (Knabbe et al., 1987), endometrial cancer (Boyd and Kaufman, 1990), and other cancers. TGFP was also found in the ascitic effusion of ovarian (Hirte and Clark, 1991) and breast cancer (Sulitzeanu, 1993) patients, where it was at least partially produced by metastatic cancer cells. The transfection of TGFP cDNA into a highly immunogenic UVinduced tumor leads to reduction in immunogenicity and progressive tumor growth in one model (Torre-Amione et al., 1990). In a mouse plasmocytoma, TGFP was secreted in large amounts by the malignant cells and consequently blocked the expression of the surface markers IgM, the CD23 receptor, and the transferrin receptor (Berg and Lynch, 1991). One may speculate that in humans, immunosuppression and increased susceptibility to infection might be caused by similar TGFPassociated mechanisms. An important regulatory cytokine, initially named CSIF (cytokine synthesis inhibitory factor) but now known as IL-10, has been cloned (Moore et al., 1990). In the mouse, CD4 T-cell clones have been classified in three categories, TH-0, TH-1, and TH-2 (reviewed in Mosmann and Coffman, 1989).Even if questions remain on the validity of the classification in wiwo and to its extrapolation to human T cells, it is well-established that IL-10 is secreted by the cells of the mouse TH-2 subset and suppresses the production of IL-2, IFN-.)I,and TNFa by the TH-1 subset (reviewed by Moore et al., 1993).The production of the NK cell stimulatory factor (NKSF or IL-12) by human peripheral mononuclear cells, and consequently the proliferation of NK cells, has also been shown to be blocked by IL-10 (D'Andrea et al., 1993). The inhibition of cytokine synthesis and of several accessory functions of macrophages thus renders IL-10 a potent suppressor of T cells, NK cells, and macrophages. In analogy to the classification of CD4+ helper cells into TH-1 or TH-2 subsets, it has been proposed that murine CD8+ T-cell clones could be subdivided into two groups, the cytotoxic (CTL) and the suppressor (Ts) subsets. As opposed to CTLs, all the Ts clones produce IGlO after stimulation with anti-CD3 mAb and suppress the proliferative activity of both TH-1- and TH-2-type CD4+ clones (Inoue et al., 1993). In a study of 48 human cancer cell lines established from different carcinomas, malignant melanomas, and neuroblastomas, IL-10 was

IMMUNE RESPONSE AGAINST TUMORS

309

found in the supernatant of 15 cell lines, mostly those derived from colorectal cancers (Gastl et al., 1993). In addition, as described for TGFp, IL-10 can be found in the peritoneal fluid and serum of patients with ovarian and other intraperitoneal cancers (Gotlieb et al., 1992), suggesting a role for this cytokine in solid tumor development and antitumor immunity. In addition to TGFp and IL-10, a large number of soluble immunosuppressive factors secreted by tumor cells have been described. These include the lymphocyte blastogenesis-inhibiting factor (LBIF) produced by a macrophage cell line (Fujiwara and Ellner, 1986); a transmembrane retroviral envelope protein, p15E, shown to diminish tumor immunity in vivo (Cianciolo et al., 1985); insulin-like growth factor 1(IGF-1) (Trojan et al., 1993);the suppressive E-receptor (SER), a molecule isolated from malignant effusions of patients with head and neck, ovarian, and lung cancer (Oh et al., 1987); different colonystimulating factors (CSF); and many other molecules (reviewed in Sulitzeanu, 1993). It must be kept in mind, however, that for most of these factors found in sera and effusions, it has not been clearly demonstrated that they are tumor-derived rather than tumor-associated molecules. Many of them have also been detected in the sera of normal patients and they might represent growth regulatory rather than immunosuppressive molecules. Nevertheless, the modulation of “immune suppressive factors” by antisense oligonucleotides, monoclonal antibodies, and other reagents remains an attractive target for future attempts at cancer immunotherapy.

D. MODIFICATION OF THE TUMOR ENVIRONMENT Modifications in the environment of a developing tumor can occur during the entire process of oncogenesis, from the preneoplastic state to the highly metastatic, autonomously growing cancer cell. These changes of extratumoral factors can cause growth-promoting or -inhibiting alterations in the complex interaction between the tumor and its environment. The process involved in tumor spread can be subdivided into the following steps: transformation, growth, angiogenesis, detachment, invasion, intravasation/release, survival in the blood stream, arrest, extravasation/invasion, and growth/angiogenesis (Hart and Saini, 1992). During each ofthese phases, immunological as well as nonimmunological mechanisms can influence the progression of the malignant disease. Since the biology of metastasis and the importance of host-tumor interactions during progression have been extensively reviewed elsewhere (Killion and Fidler, 1989; Hart and Easty, 1990; Liotta et al.,

310

CLAUDE ROTH ET AL.

1991a,b; Frost and Levin, 1992; Hart and Saini, 1992), only a few important findings are discussed in this section. During progression, tumor cells interact with countless other host cells such as lymphocytes, monocytes, and epithelial and endothelial cells. The regulation of many of these intercellular interactions is mainly dependent on integrins, a class of heterodimeric receptor molecules by which cells attach to extracellular matrices and adhere to other cells. Ligands and counterreceptors of the integrins are cellular adhesion molecules such as JCAM-1 and -2, and VCAM-1, but also molecules like laminin, collagens, fibronectin, or vitronectin (Hynes, 1987,1992). The analysis of ICAM-1 on human melanoma clones demonstrated that resistance to lysis by monocytes was associated with reduced expression of the adhesion molecules, and sensitivity to monocytes could be restored by transfection of the ICAM-1 gene into melanoma cells (Jonjic et al., 1992). On the other hand, increased expression of the cellular adhesion molecules MUC18 and ICAM-1 was correlated with metastasis in other studies (Johnson et aZ., 1989; Lehmann et al., 1989), and human melanoma as well as murine rhabdomyosarcoma cells expressing the integrins VLA-2,5, and 6 show increased adherence to the basement membrane glycoproteins, laminin, and collagen, and consequently display enhanced metastatic properties (Chan et al., 1991; Mortarini et al., 1991). Another study reported that reduced expression of the adhesion molecule cadherin was associated with increased tumor invasion (Takeichi, 1991),and the work of numerous investigators on basement membranes in malignant tissues demonstrates that a dysregulation of laminin expression leads to modifications of the interaction between tumor cells and the matrix and can either increase or decrease metastasis (reviewed in Liotta et al., 1986). An important role in invasion and intravasation is also played by macrophages and monocytes. It has been observed that infiltration of tumors by macrophages could result in an increased release of tumor cells by induction of lysosomes, causing necrosis and consequently an increase in tumor motility (Turner and Weiss, 1980). Similarly, the coculture of rat hepatoma cells with macrophages in vitro increased the invasive potential of the tumor cells in uitro and in vivo (Mukai et at., 1987). Finally, several cytokines such as TNF-a, TGFP, and IL6 can enhance the invasive potential of tumor cells, as measured in different experimental models (Miller, 1993). It has to be emphasized that, in many of the aforementioned studies on tumor invasion, immunological as well as nonimmunological mechanisms might have been responsible for the observed phenomena, and it is not clear whether

IMMUNE RESPONSE AGAINST TUMORS

311

the tumor cells or the tumor environment had undergone the primary alterations. Proteolysis of tissue barriers is another essential component of the invasive/metastatic process and has been linked to several enzymes degrading collagens, proteoglycans, or gelatin. The most extensively studied of these enzymes are the metalloproteinases, of which at least seven are known (Liotta et al., 1991a). These enzymes are produced by tumor cells and have been associated with metastasis in many model systems analyzed. Their function is counteracted by tissue inhibitors of metalloproteinases (TIMP) that are produced by malignant and normal cells, and the balance between the TIMPs and metalloproteinases apparently determines the outcome (Hart and Saini, 1992). A metalloproteinase gene that is specifically expressed in stromal cells of breast carcinomas has been cloned by Basset et al. (1990). The product of this gene, named stromelysin-3 (ST3), could be detected in 30/30 breast carcinomas evaluated and was always found in invasive but never in in situ components of the tumors (Basset et al., 1990). The authors speculated that ST3 might be instrumental in the lytic process leading to cancer invasion and metastasis, while participating at the same time in the host reaction to prevent further spread of the tumor, since ST3 is also associated with desmoplasia, a tissue reaction known to be directed against the most invasive breast cancer lesions (Basset et al., 1990). The formation of new blood vessels is a prerequisite for tumor growth in three dimensions beyond approximately 2 mm ( Folkman et al., 1989). The regulation of angiogenesis is a balance between a variety of angiogenic peptides, such as fibroblast growth factors (FGF), TGFP, and other cytokines. These regulatory factors may be produced by either normal or malignant cells (Mahadevan and Hart, 1990) and represent a further means by which modification of the tumor environment can influence tumor progression and metastasis. In the immunological context, it is important to note that several cytokines that block angiogeneisis do so by affecting the proteolytic activity (Liotta et al., 1991a). Thus, monocytes produce several cytokines, such as TNF-a, TGFP, and IL-1, that can have angiogenic activity (Roberts et al., 1986; reviewed in Miller, 1993). As in the case of adhesion of tumor cells to matrices, proteolysis of membrane barriers, or angiogenesis, other steps involved in the progression of a malignant disease are also subject to regulatory mechanisms imposed by the host and possibly counteracted by the tumor in its attempts to evade “surveillance.” These include factors that regulate survival of tumor cells in the bloodstream, early postarrest events,

312

CLAUDE ROTH ET AL.

and similar other steps (reviewed by Miller, 1993). These multiple alterations of the tumor environment evoke a number of possible targets for therapeutic interventions. It would be attractive to target the tumor environment rather than the tumor itself because, in contrast to many tumor-directed immunotherapeutical approaches, it is probably unnecessary to modify the entire tissue in order to achieve a beneficial effect, which could also be reached by systemic treatment. IV. Overcoming the Poor Immune Responses Elicited by Tumors

A. INCREASING MHC EXPRESSION ON TUMOR CELLS

1 . Transfection of Allo-MHC into Tumor Cells Attempts to increase the antigenicity and immunogenicity of tumors by introducing foreign genes have been made for more than 20 years. This modification has been called the “xenogenization of tumor cells” (Kobayashi et al., 1969) and can be achieved by the use of viruses (Lindenman and Klein, 1967; Kuzumaki et al., 1978), chemical coupling (Lachmann and Sikora, 1978), enzyme treatment (Currie and Bagshawe, 1969; Bekesi et al., 1971), somatic hybridization (Watkins and Chen, 1969), or transfection of allo-MHC molecules (Itaya et al., 1987). Xenogenized tumors are able to elicit immune responses not only to the “neoantigen,” but also to the TAA of the parental cells, most likely through helper antigen mechanisms (Boone et aE., 1974). Infection of a mixture of four human tumor cell lines by vaccinia virus, in order to potentiate their antigenicity, has been used to produce a lysate which is currently under clinical investigation (Wallack and Sivanandham, 1993). Itaya et al. transfected an allogeneic H-2Ld gene into the Lewis lung carcinoma 3LL/3 derived from the C57B1/6 strain (H-2b).The antigenic expression of the Ld-positive clones was approximately 20-40% of that observed with MethA tumor cells of BALB/c mice (H2-d), and the transfected cells also expressed constant amounts of their native Db antigens (Itaya et al., 1987). The tumorigenicity of these clones was reduced to less than 1/40 of that of the parent tumor cells. The mice also acquired transplantation resistance against challenge with the parental cells after inoculation and regression of viable Ld-positive cells. Surprisingly, the immunogenicity of the Ld-transfected clones was not different from Ld-negative parental cells. The latter was determined by measuring the resistance of C57B1/6 mice against parental 3LL/3 cells after pretreatment with Ld-positive and Ld-negative ceIls rendered nonviable by mitomycin C. The authors speculate that the

IMMUNE RESPONSE AGAINST TUMORS

313

antigenic expression of the allo-MHC molecule might be too strong to enhance the relatively weak TAA, as has been shown for xenogenization by viruses (Yamaguchi et al., 1982).Alternatively, the immunogenicity of the TAA of 3LL may be too low to be recognized by the mice, even after immunogenization by allo-MHC, a mechanism that also has been described in virally xenogenized tumors (Kobayashi et al., 1978). More recently, Plautz et al. (1993) reported on the in uiuo transfer of an H2-KS gene into the colon adenocarcinoma cell line CT26 (H2Kd) and the fibrosarcoma cell line MCA 106 (H2-Kb).This was achieved by surgical exposure of tumors and injection with either retroviral or DNA liposome vectors. In contrast with the above results (Itaya et al., 1987), a cytotoxic T-cell response to H-2Ks and other antigens present on unmodified tumor cells was induced. This immune response attenuated tumor growth in more than 70%ofthe treated animals and resulted in complete cure in approximately 20%. It is important to note that these results were achieved in tumor cell lines that are known to be poorly immunogenic.

2. Transfection of Self-MHC Class ZZ into Tumor Cells The introduction of INF-y and other cytokine genes into tumor cells can upregulate the expression of self-MHC class I and I1 molecules. This aspect is dealt with in more detail in Section IV,C. Only the effects of transfection of the MHC genes themselves are therefore described below. In some murine tumors, introduction of self-MHC class I1 genes into tumor cells leads to the rejection of the transfected cells and can even cause the rejection of nonmodified parental cells. OstrandRosenberg et a2. immunized syngeneic A/J mice with Sal sarcoma cells transfected with self-Ak molecules and showed that the MHC class 11-expressing cells stimulated an improved tumor-specific immune response and that the immunity was cross-reactive with the class 11-negative tumor. Since the transfected MHC molecule could not function as a target molecule for autologous tumor rejection, the increased immunity of Sal/Akcells was probably due to stimulation of a tumor-specific T helper cell population (Ostrand-Rosenberg et al., 1990). In another experimental system, increased immunogenicity after transfection of autologous MHC class I1 molecules was found to be dependent on their cytoplasmic tail. This suggests that enhanced presentation of peptides to CD4+-positive T cells is not the only mechanism by which the transduced MHC molecule enhanced immunogenicity (Baskar et al., 1993).

314

CLAUDE ROTH ET AL.

3. Transfection of SelfMHC Class I into Tumor Cells Most of the experiments aimed at increasing antitumor immunity by the transfection of tumor cells with MHC genes involved selfMHC class I molecules. Increasing MHC class I expression by gene transfection usually results in increased immunogenicity and decreased tumorigenicity and metastatic capacity in murine models (Hui et al., 1984; Tanaka et al., 1985; Wallich et at., 1985). The decreased tumorigenicity is postulated to be due to enhanced presentation of tumor-specific peptides to CD8+-positive CTLs in viuo. However, increased expression of MHC does not necessarily lead to an increase in the immunogenicity of a tumor. It can also lead to a decrease in NK-mediated cell killing and thus to an increase in the tumorigenicity of the transfected cells (Kame et al., 1986; Glas et al., 1992; Franksson et al., 1993). In addition, several nonimmunological mechanisms are apparently involved in the alterations of tumorigenicity after MHC transfection (Gorelik et al., 1990,1991; Kaufman et al., 1993). a. Zmmunological Effects i. Recognition by CTL Effector Cells The most straightforward consequence of self-MHC class I gene transfection may be to induce or increase the capacity of the transfected tumor cells to present peptides derived from putative tumor-specific antigens. This concept was supported by early experiments in which transfection of H-2Kkgenes into the AKR leukemia led to an increase of the immunogenic properties of these cells and, consequently, to a suppression of growth in vivo (Hui et al., 1984). Similar findings have been reported in several other tumor systems, such as T10 sarcoma, 3LL lung carcinoma, or B16 melanoma transfected with the H-2Kb gene (Wallich et al., 1985; Bahler et al., 1987; Plaksin et al., 1988; Porgador et at., 1989). As described in more detail below (Section IV,A,3,c), in several cases even the metastatic phenotype of the tumor cells could be diminished or completely abrogated by transfection of autologous MHC class I molecules (reviewed in Feldman and Eisenbach, 1991). In some of the tumor models, such as the 3LL lung carcinoma (Plaskin et al., 1988), the AKR leukemia (Hui et al., 1984) and the B16 melanoma (Porgador et al., 1989), preimmunization with MHCpositive cells decreased dramatically the tumorigenicity and/or metastatic spread of the parental MHC-negative cells. Therefore, the very low density of the MHC class I molecule on the parental cells did not confer immunogenic competence per se, but seemed to be sufficient to make these cells susceptible to lysis in uitro by MHC class Irestricted CTLs.

IMMUNE RESPONSE AGAINST TUMORS

315

In contrast, in the T10 sarcoma system, such cross-reactivity was not observed in vitro. CTLs induced by Kb- or Kk-transfected IE7 or IC9 cells did not kill the parental lines, most probably because the parental T10 cells did not express any cell-surface H-2K at all (Wallich et al., 1985). Since an increasing number of human tumors have been found to express very low but detectable levels of MHC class I molecules (see Section III,A, above), antimetastatic immunotherapy with MHC class I-transfected primary tumor cells deserves further investigation. ii. Recognition b y NK Effector Cells In a number of experimental tumor models, the sensitivity of tumor cells to NK cytotoxicity is determined by the level of MHC class I expression on the tumor cells (reviewed in Ljunggren and Karre, 1990; Schrier an& Peltenburg, 1993). Thus, in the murine lymphoma RBL-5, tumor cel(s selected for loss of H-2 expression were less malignant than wild type cells after low-dose injection in syngeneic hosts (Karre et al., 1986). In a P2microglobulin (P2-m)-deficientvariant of the murine lymphoma EL4,which does not express stable MHC class I on the cell surface, a marked reduction in tumorigenicity in syngeneic C57B1/6 mice was also observed when compared to normal EL-4 cells. Transfection of a functional P2-m gene into the variant cells reestablished normal tumorigenicity. In addition, in athymic B6 nude mice, the P2-mtransfected cells were much more tumorigenic than nontransfected variant cells, whereas depletion of NK cells in these mice restored normal tumorigenicity of &-m-deficient cells (Glas et aZ., 1992).Induction of MHC class I expression can thus lead to escape from NK cells in vivo. Similarly, murine RMA-S lymphoma cells, which are defective in the peptide transporter TAP-2 gene, do not express significant levels of MHC class I and are sensitive to NK lysis. Transfection of RMA-S with the missing transporter gene led to a marked increase in tumor outgrowth potential in uivo, despite the fact that the transfected cells had restored antigen presentation and would have been expected to be more immunogenic and thus less tumorigenic (Franksson et al., 1993). The increase in tumorigenicity was shown to be due to a loss of sensitivity to NK lysis. The “missing self” hypothesis (Ljunggren and Eirre, 1990) states that somatic ceIls are permanently scanned for MHC class I expression by surveying NK cells. Cells that have lost MHC will be recognized as targets and killed by NK cells. The authors have proposed two distinct, albeit not mutually exclusive, models to explain this mechanism. According to the “target interference” model, class I MHC molecules prevent the interaction between the tentative NK target and the

316

CLAUDE ROTH ET AL.

NK cells. The “effector inhibition” model postulates that the delivery of the lytic hit is blocked after interaction of NK cell and target cell and recognition of MHC class I. Thus, if class I MHC molecules are absent, NK cells kill the target. The latter model is supported by the observations that NK cells of normal mice can reject bone marrow from syngeneic mice deficient in &-m expression (Bix et al., 1991) and that normal T cell blasts from MHC class-I-deficient mice can serve as targets for syngeneic NK cells in vitro (Liao et al., 1991). Further support in favor of the “effector inhibition” model was provided by the finding that the Ly-49 molecule, which is expressed on a subset of murine NK cells, can mediate an inhibitory signal upon recognition of certain MHC class I molecules (Karlhofer et al., 1992). In addition, it has been demonstrated that proximal signaling events, such as phospholipase C-mediated hydrolysis of membrane phophoinositides or calcium signaling, are not decreased in NK cells after interaction with MHC class I molecules in transfected target cells. On the contrary, addition of anti-HLA mAb leads to increased lysis of the class I-transfected targets (Kaufman et al., 1993).These results suggest that MHC class I expression on target cells does not block the access of NK cells to their cellular target but that MHC molecules initiate delivery of inhibitory signals in the NK cells once they have bound to their putative receptors. An inverse correlation between NK susceptibility and expression of MHC class I has not, hgwever, been found ubiquitously (Gorelik et al., 1990; Ljunggren and Karre, 1990). Other factors determining NK sensitivity of a tumor must, therefore, be involved. In agreement with this hypothesis, results suggested both MHC-dependent and MHCindependent cytotoxic mechanisms, may influence NK susceptibility, depending on the nature of the target cell (Litwin et al., 1993). The question of whether MHC class I upregulation is systematically beneficial or not thus has no general answer. Opposite results have been reported with regard to tumorigenicity even in H-2-positive and H-2-negative variants ofthe same tumor if different experimental protocols are used (Ljunggren and E r r e , 1990). Most of the available data can be rationalized as follows: the effects of variation in class I expression on tumorigenicity will depend on the dominating immunosurveillance system in the model studied. If NK cells (eliminating class Ideficient cells) play the major role, upregulation of MHC class I will lead to an increase in tumorigenity of the malignant cells. If CTLs (eliminating class I-positive cells that present tumor-specific peptides) dominate, an increase of MHC class I will rather lead to enhanced immunogenicity and thus decreased tumorigenicity (Ljunggren and

IMMUNE RESPONSE AGAINST TUMORS

317

Karre, 1990).What determines which immunosurveillance system dominates is not known. It appears as if domination of NK cells is more prominent in tumors of low inherent immunogenicity, whereas in highly immunogenic tumors, CD8+-positive CTLs play the dominant role (Elliot et al., 1989).

b. Nonimmunological Effects It is generally accepted that the main importance of class I MHC molecules in oncogenesis lies in their key role in eliciting immune responses against the peptides which they present. However, a variety of nonimmunological effects by which MHC molecules may interfere with the development of tumors have also been proposed. It has been reported that the transfection of class I H-2Kb but not of class I1 H-21Akgenes into a BL6 melanoma clone led to a number of alterations in the tumor cells which were potentially irrelavant to their immunological functions. Thus, clones which expressed high levels of H-2Kbdisplayed elevated levels of soybean agglutinin (SBA), Griffonia simplicifolia I-B4 (GSIB4), and peanut agglutinin (PNA) lectin-binding sites (Gorelik et al., 1991). In parallel, these cells had lost the expression of the melanoma-associated antigen (MAA) and became sensitive to natural cell-mediated cytolysis (NCMC) and to lysis by recombinant TNF-a. Natural cytotoxicity (NC) sensitivity of these tumor cells was blocked by anti-TNF-a antibodies, suggesting that expression of the transfected H-2Kb gene resulted in increased sensitivity to TNF-a-mediated cytotoxicity (Gorelik et al., 1990). Since it seemed unlikely that TNF-a could directly recognize class I antigens, it was speculated that transfection of the H-2 gene may have an indirect effect on TNF-a sensitivity. In fact, four H-2Kb-transfected clones that subsequently lost H-2Kbexpression but retained their ability to bind to SBA agarose beads still displayed the modified phenotype described above, namely, an increase in SBA and GS1B4 lectin binding, the loss of MAA, and sensitivity to TNF lysis (Kim et al., 1993). Transfection of BL6 clones with H-2Kbhas also been reported to lead to the inhibition of the production of the endogenous A- and C-type retroviruses found in these melanoma cells and to the concomitant expression of some cellular genes that are usually repressed by these retroviruses (Gorelik et al., 1994). The phenotypic changes in the BL6 melanoma cells were probably induced by the transfection of the H2Kbgene that was later lost or downregulated during the development of the malignant clone. This assumption was based on the observation that ecotropic retrovirus production was lost because of induced stable rearrangements in proviral DNA that were sustained in the absence

318

CLAUDE ROTH ET AL.

of H-2Kb gene expression (Muller et al., 1994). Electron microscopic analysis of the BL6 melanoma clones expressing transfected or endogenous H-2Kbmolecules revealed that all clones that had lost cell-surface expression of MAA had also lost budding virus particles. This loss of budding particles was associated with DNA rearrangements of proviral DNA in all samples analyzed. The postulated relationship between retrovirus expression and TNF sensitivity is supported by findings indicating that human adenovirus gene products could also control tumor cell sensitivity to TNF lysis (Chen et al., 1987; Gooding et al., 1990). The reversal of TNF resistance in H-2Kb-transfectedBL6 melanoma cells was associated with an increase in p55 TNF receptor expression and enhanced TNF internalization and degradation. TNF-induced activation of phospholipase A2 and release of arachidonic acid metabolites was observed and might have been responsible for the reversion of a block in transduction of the lytic signal present in TNF-resistant parental BL6 cells (Kim et al., 1993).

c. Effects on the Metastatic Phenotype The emergence of a metastatic phenotype is a multistep process. The tumor cell must express a variety of genes and acquire a number of specific capabilities to successfully go through the selective process that finally leads to metastasis. These include the expression of proteolytic enzymes allowing penetration through intercellular matrices and blood capillaries, the matrix-independent growth in serum, synthesis of angiogenic factors controlling metastatic spread, and many others (Hart and Easty, 1990; Liotta et al., 1991b). At each of these steps, interaction with the immune system is possible, and a number of experimental data suggest that MHC class I molecules may play an important role in the suppression of metastatic spread by immune mechanisms (Feldman and Eisenbach, 1991). The first experiments evaluating the association between metastasis and MHC expression were carried out more than 10 years ago and demonstrated a clear dependence of the metastatic properties of a tumor cell on the differential expression of H-2 genes. While some genes (H-2Kband H-2Kk)caused abolition of the metastatic phenotype, others (H-2Db)increased the metastatic competence of the cells (De Baetselier et al., 1980; Katzav et al., 1984). Later, the abrogation of metastatic properties of tumor cells by the transfection of H-2 genes was also demonstrated (Wallich et al., 1985). The expression of H2Db but not H-2Kb was found to be associated with the suppression of metastases in an allogeneic mouse model of metastasis (Lewis lung

IMMUNE RESPONSE AGAINST TUMORS

319

carcinoma), while local proliferation was unhampered by MHC expression (Isakov et al., 1983). Metastatic cells, therefore, appeared more susceptible to allograft rejection than locally proliferating cells. This might be explained by the fact that a metastasizing cell is individually exposed to CTLs, while the primary tumor grows in a bulk where individual cells are less exposed to cytotoxic cells and can survive despite the presentation ofallo-antigens. Similar studies in other allogeneic systems have confirmed the inverse correlation between metastatic phenotype and expression of allo-MHC (Gelber et al., 1989). In another syngeneic model, metastasis was inversely correlated with the level of expression of H-2Kb (Eisenbach et al., 1983,1984). The introduction of the H-2Kb gene into highly metastatic cells led to either a complete loss or a significant reduction of their metastatic competence (Plaskin et al., 1988). The suppression of the metastatic phenotype was associated with H-2Kb-controlled immunogenicity, since the H-2Kb-transfectedcells generated spontanous metastases in immune-suppressed animals, and CTLs specific for the transfected but not the parental cells appeared in immunized immunocompetent mice. The nontransfected, parental cells expressed low amounts of H-2Kb, and it proved possible to protect animals against injection of large numbers of the parental cells by preimmunization with irradiated or mitomycin C-treated H-2Kbtransfectants. In addition, the treatment of established lung metastases from parental cells with H-2Kb-expressing transfectants resulted in cure of some animals and significant tumor reduction in many others (Plaksin et al., 1988). In the B16 melanoma cell line, MHC class I expression does not appear to be an important factor in the control of the metastatic phenotype, since cells of both high and low metastatic competence express virtually no MHC molecules. Metastasis seems instead to be controlled by different levels of collagenase activity, fibrinolytic activity, and adhesiveness to lung cells (Feldman and Eisenbach, 1991). Nonetheless, transfection of syngeneic H-2Kbmolecules into cells of high metastatic competence leads to a very significant fall in the number of metastases formed and an induction of CTLs restricted by H-2Kb. These CTLs manifest high cytotoxicity not only against Kb-transfected targets but also against parental cells expressing very low amounts of Kb (Porgador et al., 1989). It was later shown that the c-fos oncogene, whose expression can be induced by IFN-7, is involved in the control of MHC expression and is differentially expressed in low versus highly metastatic clones (Kushtai et al., 1988). Accordingly, transfection of cells with c-fos

320

CLAUDE ROTH ET AL.

turned on MHC expression and consequently reduced the metastatic phenotype (Kushtai et al., 1990). It may be concluded that downregulation of MHC class I expression seems to be an important mechanism by which potentially metastatic cells evade immune surveillance during the sequential steps of the metastatic process. This notion is also suworted by the fact that. in humans, metastatic tumor cells frequently express lower levels of MHC class I than their parental primary tumor cells (Cordon-Cardo et al., 1991; Pantel et al., 1991). It has to be kept in mind, however, that the loss of MHC expression was found to lead to increased tumorigenicity in several primary tumor models, as discussed above (Section IV,A). It is conceivable that in appropriate circumstances similar effects might also play a role in metastasis and thus lead to stronger metastatic competence in MHC nonexpressors, as has been observed with T10 cells transfected with the H-2Dk gene (Gopas et al., 1989). B. INTRODUCTION OF THE COSTIMULATORY MOLECULEB7 INTO TUMORS The poorly immunogenic, B7-negative, murine melanoma cell line K1735 expresses both MHC class I and class I1 molecules and stimulates a specific, albeit ineffective, immune response in uiuo. In an attempt to provide the tumor cells with costimulatory activity, K1735 was transfected with the B7 gene (Chen, L. et al., 1992; Townsend and Allison, 1993). Expression of B7 on transfected K1735 cells led to a CD8+ T-cell-mediated immune response in syngeneic mice inoculated with these cells. Tumor rejection was independent of CD4+ helper cells and immunized mice rejected transfected as well as nontransfected K1735 cells when challenged 25 days after vaccination (Townsend and Allison, 1993). In the study by Chen et al., cotransfection with B7 and a strong viral tumor rejection antigen, the E7 gene of human papilloma virus 16, was necessary to achieve successful vaccination in the same murine melanoma model (Chen, L. et al., 1992). Further work is needed to explain this discrepancy. It is interesting to note, however, that, in the experiments of Chen et al., even mice with established micrometastases of E7+B7- K1735 cells could be cured by injection of E7+B7+cells, suggesting that CTL precursors directed against the E7 antigen were not anergic, or that anergy could be reversed by the double-transfectants. In both cases, tumor growth of B7-transfected cells was unaltered in nude mice and in immunocompetent mice deprived of CD8+ cells. Since depletion of CD4+ cells did not impair the animals’ capacity to reject the tumor, it was suggested that, following B7-mediated costimu-

IMMUNE RESPONSE AGAINST TUMORS

321

lation, tumor-specific CD8' CTL clones could make enough IL-2 on their own to become fully activated and expand (Schwartz, 1992). In the two studies on K1735 and in a similar model of B7-transfected P815 mastocytoma cells, it was shown that costimulation is important during the inductive but not the effector phase of the response, since tumor-specific CTL effectors were equally effective in lysing B7transfected and nontransfected cells in uitro or in viuo (Chen L. et al., 1992; Harding and Allison, 1993; Townsend and Allison, 1993). Baskar et al., transfected the very weakly immunogenic murine sarcoma SaI with the B7 gene and achieved complete protection against high doses ofparental tumor cells in mice immunized with the transfectants. Interestingly, CD4+ effector cells were responsible for the observed tumor rejection, suggesting that, in the sarcoma model, the immunogenic tumor-specific antigens were presented by MHC class I1 rather than class I molecules (Baskar et al., 1993). Taken together, these studies demonstrate that under appropriate conditions, coexpression of B7 can stimulate both CD4+ and CD8+ T cells, thereby enhancing the tumor-specific response in both T-cell compartments.

C. ENGINEERING OF TUMOR CELLSWITH CYTOKINE GENES In murine tumor models, it is now generally accepted that the failure of an antitumor immune response is often due not to the absence of tumor specific antigens, but rather to defects in immune regulation. One goal, then, is to overcome these defects by modifying the local immunological environment of the tumor cell in such a way as to enhance the presentation of tumor-specific antigens and/or the activation of tumor-specific lymphocytes. The first successful attempts to stimulate specific effector cells in the vicinity of the tumor were performed in murine tumor models, using malignant cells engineered with the IL-2 gene (Bubenik et al., 1988,1990; Fearon et al., 1990; Gansbacher et al., 1990a; Ley et al., 1990,1991; Russel et al., 1991). In all the experimental systems, the expression of IL-2 by the weakly immunogenic tumor cells resulted in growth inhibition of the modified tumor cells. Furthermore, in several cases, a systemic immune response against the parental tumor was observed, as a consequence ofthe activation of specific T lymphocytes, leading to an immune memory against a challenge with the parental tumor cells (Fearon et aE., 1990; Ley et al., 1991; Gansbacher et al., 1990a). A parallel approach was tried with tumor cells producing IL4, another helper lymphokine that induces a local inflammatory re-

322

CLAUDE ROTH ET AL.

sponse. All the animals inoculated with the IL-4-producing cells developed an antitumor response, leading to the rejection of genetically modified tumor cells (Tepper et al., 1989;Li et al., 1990; Blankenstein et al., 1990; Golumbek et aZ., 1991).Local immune reactions, evaluated by histological analysis, revealed the recruitment of activated macrophages and eosinophils at the site of tumor rejection (Tepper et al., 1992). In the latter study, the antitumor effect was not mediated by T cells and was also present in nude mice, but subsequent experiments using a renal cell carcinoma transduced with IL-4 revealed a systemic immune response generated against the parental tumor. Similar experiments were performed using IL-7-transfected plasmocytoma 5558. Tumor rejection and generation of a systemic immune response was reported only in syngeneic but not in nude mice (Hock et al., 1991; McBride et al., 1992).In this experimental system, the antitumor effect was shown to be dependent on CD4 T cells, while in other tumor models, the rejection of modified tumor cells involved mainly CD8+ and, to alesser extent, CD4+ T lymphocytes (Aokiet al., 1992).Discrepancies between the effector cell populations required for tumor rejection in the different systems may be due to heterogenous levels of class I or class I1 molecules expressed on the target tumor cells. For example, unlike most tumor models, the 5558 plasmocytoma is a class II-positive cell line that could present tumor antigens to class IIrestricted CD4' T lymphocytes. Other cytokines, such as IFN-y and GM-CSF, might locally enhance the level of antigen presentation by the tumor cells and/or antigenpresenting cells in their vicinity. In various systems, it was shown that local secretion of IFN-y by the tumor cells can protect the mice against tumor growth and induce a long-lasting specific immunity mediated by CD8+ T cells (Watanabe et al., 1989; Cansbacher et al., 1990b; Restifo et al., 1992). It was postulated that loss of tumorigenicity in animals inoculated with IFN-y-secreting tumors is due to enhanced class I expression and antigen presentation (Restifo et al., 1992). However, MHC class I expression, while necessary, was shown to be insufficient by itself to inhibit tumorigenicity, supporting the notion that additional factors are needed (Esumi et al., 1991). Indeed, IFN-y is also an activator of CTL (Hayashi et al., 1985) and NK cells (Weigert et al., 1983), and it promotes the tumoricidal activity of macrophages (Pace et aZ., 1983). More recently, the beneficial effect of another cytokine, GM-CFS, which is required for the differentiation of hematopoietic progenitors and the maturation of specialized antigenpresenting cells (Steinman, 1991),was tested in the highly tumorigenic B16 melanoma model. Transfected tumor cells producing GM-CSF are rejected in syngeneic animals by stimulating specific cytotoxic T

IMMUNE RESPONSE AGAINST TUMORS

323

cells (Dranoff et al., 1993). It was hypothesized that dendritic cells, as they exhibit increased antigen presentation and accessory functions, could be potent stimulators of helper and cytotoxic activities, thus reversing the T-cell anergy to tumor cells (Nonacs et al., 1992).Further experiments in induced tumor models as well as in spontaneous tumors will be required to compare the beneficial effect of these various cytokines in inducing tumor growth inhibition and rejection of preestablished tumors. Interestingly, it was shown that cytokines may not have to be secreted by the tumor cells themselves to deliver their beneficial effect. Using IL-2-secreting allogeneic or even xenogeneic cells as vectors, protection against tumor growth was obtained in two different experimental tumor systems (Roth et al., 1992,1994).In this situation, stimulation of nonspecific (NK) and/or specific (CTL)effector cells was shown to depend on the tumor model. While CTL were shown to be responsible for immune rejection of P815 tumor cells, NK cells were the major effector cells involved in the protective effect against Lewis tumor growth (Roth et al., 1992). Finally, in the mouse models, it appears that the immune reaction against tumor cells can be described as a two-step reaction. In the initial phase, nonspecific effector cells, such as macrophages, activated eosinophils, or neutrophils, and NK cells are recruited in the local inflammatory reaction observed (Cavallo et al., 1992; Hock et al., 1993a,b). It appears that the immune rejection of IL-4-producing transfectants involves two waves of effector cell types. In the early phase (within 18hr), macrophages and activated eosinophils are associated with the local inflammatory reaction observed and play a dominant role in the tumoricidal effect initiated by IL-4, while, in the later phase, T lymphocytes can be stimulated, depending on the inherent immunogenicity of the tumor cell line, which can then lead to systemic immunity (Tepper et al., 1992). In general, then, it appears that the first nonspecific reaction enables the secondary specific immune reaction to take place, by slowing tumor growth and enhancing the level of tumor antigen presentation by the tumor and/or specialized antigenpresenting cells, through MHC class I and class I1 molecules. In agreement with this hypothesis, it has been repeatedly observed that tumors start growing in T-cell-deficient mice only after a latency period that corresponds to the first wave of non-T effectors that slow down tumor growth (Hock et al., 1993a). In immunocompetent animals, this gives T cells the time to proliferate and differentiate into effector cytolytic cells. In the latter phase, CD8+ tumor-specific T cells can arise, leading to long-term immune protection against the parental tumor growth (Colombo et al., 1992).

324

CLAUDE ROTH ET AL.

D. ESCAPE MECHANISMS In order to evaluate the potential of the above immune manipulations for future therapeutic procedures, it is important to consider some of the known immune escape mechanisms. Initially, some tumor cells enter a prolonged, quiescent state, termed tumor dormancy,” a situation in which tumor cells are present but tumor progression is not clinically apparent. While observed in various human tumors, such as lymphomas, leukemias, carcinomas, and melanomas (Stewart et al., 1991), this phenomenon has been mainly documented in the experimental model of antibody-induced dormancy (Yefenof et al., 1993a,b). In this model, the interaction of antibodies (antiidiotypes) with surface IgM (idiotypes) on BCLl tumor cells is sufficient to stop proliferation of the growing malignant cells. Strikingly, these dormant cells, whose growth was arrested, retained their malignant phenotype, since their injection into a naive animal resulted in the progressive growth of BCLl tumor cells in the adoptive recipient (George et al., 1987). Furthermore, a sporadic loss of dormancy has been observed in immune mice, and the rate of dormancy loss suggests that a single mutation event plays a critical role in the escape process (Yefenof et al., 1993b). This phenomenon is of great interest in dissecting the cellular and molecular mechanisms leading to the induction, maintenance, and termination of the dormant state in various cancers. Tumor escape may also be associated with suppressive effects mediated by CD4+ T cells. In a highly immunogenic UV-induced tumor, in vivo treatment with anti-CD4 antibodies can prevent or slow tumor growth. Suppression of CTL responses can in fact be mediated by CD4’ T cells through immunosuppressive factors such as interleukin10 secreted by the TH-2 subset, or through antibodies or other B-cell products, whose secretion requires T-cell help (Schreiber et al., Fortyfifth Annual Symposium on Fundamental Cancer Research, October 1992, Houston, Texas; Monach et al., 1993). Furthermore, escape from the immune response could result from the selection of tumor variants that are no longer recognized by the specific CTLs. This has been shown in SV40-induced tumors in mice (Lill et al., 1992). Five antigenic sites were identified on the 94-kDa large T antigen of SV40 tumor virus (Tanaka et al., 1988; Deckhut et al., 1992), and CTL-resistant variants were selected by in vitro cocultivation of an SV40-transformedmouse kidney cell line and CTL clones recognizing the various epitopes. Identification of variants lacking sites I, I1 and I11 coding sequences revealed that point mutations ‘I

IMMUNE RESPONSE AGAINST TUMORS

325

were responsible for the loss of the relevant CTL recognition epitope. These results reinforce the view that minimal genetic changes can generate new epitopes that still bind to MHC class I molecule but have lost the CTL recognition sites from tumor antigens, allowing transformed cells to escape immunosurveillance. Finally, an important observation has been made recently. In tumorbearing mice as well as in a number of cancer patients (Mizoguchi et al., 1992; Finke et al., 1993; Nakagoni et al., 1993) CD8' T cells which infiltrate the tumors show a defect in signal transduction, associated with alterations in the structure of the CD3 complex and more particularly of its 5 chain. This might be a major cause for immune defects in tumor-bearing hosts. Whether or not such modifications of the CD3 complex are induced by some diffusable immunosuppressive factors produced by the tumor cells remains opened to investigation. V. Conclusion

We have reviewed here a number of aspects dealing with immune responses directed against and stimulated by tumor cells. As mentioned in the Introduction, the antigenic and immunogenic properties of tumors are intimately related, inasmuch as everything learned about tumor cells is very rapidly applied to fight them. We have certainly not covered the entire field. In particular, there are many other ways to try and overcome the poor immunogenicity of tumor cells than described here, especially with respect to modifications of antigenpresenting cells (as opposed to the tumor cells) and to the use of various gene therapy protocols. The results obtained in animal models are indeed extremely encouraging, but the growing enthusiasm about the applicability of in vivo or ex vivo immunotherapy protocols in humans should be tempered for a variety of reasons. One is that there are very large gaps in our basic knowledge and much remains to be learned about the antigenicity and the immunogenicity of tumor cells. A second reason is that oncologists will in general agree that there are no really accurate animal models for human cancers, and that work performed in the mouse, while necessary and informative, may not yield results directly transposable to man. In this context, it is worthwhile mentionning that preclinical work performed with spontaneous tumors in domestic animals might be extremely useful in approaching the high complexity of the situations which must be faced in human beings. A third reason pertains to the clinical trials themselves: first, a very large number of trials are going to be necessary in order to evaluate the large number of new

326

CLAUDE ROTH ET AL.

procedures (and of combinations of them) which are being suggested for immune therapy. Second, many such trials will be performed first with patients with advanced diseases, where they are less likely to be efficient. Confidence will have to be gradually gained in a selected number of procedures to permit more extensive and, perhaps, more significant trials to be performed. Nevertheless, it is a fair assessment of the current situation that the impressive progess made in immunology and basic tumor immunology is likely to prove useful in the relatively near future.

ACKNOWLEDGMENTS The authors thank many colleagues who have sent us preprints and unpublished information. We acknowledge Drs. J.-P. Abastado, R. Cohen, J. Even, and D. Ojcius for fruitful discussions. We are especially indebted to Dr. D. Ojcius for critical reading of the review and to Ms. V. Caput for editorial assistance in the preparation of the manuscript.

REFERENCES Abastado, J.-P., Plata, F., Morello, D., Daniel-Vedele, F., and Kourilsky, P. (1985). H2 restricted cytolysis of L cells doubly transformed with a cloned H-2Kb gene and cloned retroviral DNA. J . Immunol. 135, 3512-3519. Acres, R. B., Hareuveni, M., Balloul, J.-M., and Kieny, M.-P. (1993). Vaccinia virus MUCl immunization of mice: Immune response and protection against the growth of murine tumours bearing the MUCl antigen. J . Immunother. 14, 136-143. Aichele, P., Hengartner, H., Zinkernagel, R. M., and Schulz, M. (1990).Antiviral cytotoxic T cell response induced by in vivo priming with a free synthetic peptide. J . E x p . Med. 171, 1815-1820. Allred, D. C., Clark, G. M., Tandon, A. K., Molina, R., Tormey, D. C., Osborne, C. K., Gilchrist, K. W., Mansour, E. G., Abeloff, M., Eudey, L., et 41. (1992). HER-2heu in node-negative breast cancer: Prognostic significance of overexpression influenced by the presence of in situ carcinoma. J . Clin. Oncol. 10, 599-605. Almoguera, C., Shibata, D., Forrester, K., Martin, J., Arnhem, N., and Perucho, M. (1988).Most human carcinomas ofthe exocrine pancreas contain mutant c-K-ras genes. Cell (Cambridge, Mass.) 53,549-554. Anderson, M. L., Stam, N. J., Klein, G., Ploegh, H. L., and Masucci, M. G. (1991). Aberrant expression of HLA class I antigens in Burkitt lymphoma cells. Int.1. Cancer 47,544-550. Anichini, A., Maccalli, C., Mortarini, R., Salvi, S., Mazzocchi, A., Squarcina, P., Herlyn, M., and Parmiani, G. (1993).Melanoma cells and normal melanocytes share antigens recognized by HLA-A2-restricted cytotoxic T cell clones from melanoma patients. 1.Exp. Med. 177,989-998. Aoki, T., Tashiro, K., Miyatake, S.-I., Kinashi, T., Nakano, T., Oda, Y.,Kikuchi, H., and Honjo, T. (1992). Expression of murine interleukin 7 in a murine glioma cell line results in reduced tumorigenicity in uiuo. Proc. Natl. Acad. Sci. U.S.A.89,3850-3854. Babbit, B. P., Allen, P. M., Matsueda, G., Haber, E., and Unanue, U. E. (1985).Binding of immunogenic peptides to Ia histocompatibility molecules. Nature (London)317, 359-361. Bahler, D. W., Frelinger, J. G., Hanvell, L. W., and Lord, E. M. (1987). Reduced

IMMUNE RESPONSE AGAINST TUMORS

327

tumorigenicity of a spontanous mouse lung carcinoma following H-2 gene transfection. Proc. Natl. Acad. Sci. U.S.A.84,4562-4566. Barbacid, M. (1987). Ras genes. Annu. Reu. Biochem. 56,779-827. Barnard, J. A,, Lyons, R. M., and Moses, H. L. (1990). The cell biology of transforming growth factor beta. Biochim. Biophys. Acta 1032, 79-87. Barnd, D. L., Lan, M., Metzgar, R., and Finn, 0. J. (1989). Specific MHC-unrestricted recognition of tumor-associated mucins by human cytotoxic T cells. Proc. Natl. Acad. Sci. U.S.A.86,7159-7163. Baskar, S., Ostrand-Rosenberg, S., Nabavi, N., Nadler, L., Freeman, G., and Glimcher, L. (1993). Constitutive expression of B7 restores immunogenicity of tumor cells expressing truncated MHC class I1 molecules. Proc. Natl. Acad. Sci. U.S.A.90, 5687-5690. Basset, P., Bellocq, Wolf, C., Stoll, I., Hutin, P., Limacher, J. M., Podhajcer, 0. L., Chenard, M. P., Rio, M. C., and Chambon, P. (1990).A novel metalloproteinase gene specifically expressed in stromal cells of breast carcinomas. Nature (London) 348, 699-704. Bekesi, J. G., Saint-Arneault, G., and Holland, J. E. (1971). Increase of leukemia L1210 immunogenicity by Vibrio cholate neuraminidase treatment. Cancer Res. 31, 2130-2132. Berg, D. J., and Lynch, R. G. (1991). Immune dysfunction in mice with plasmacytomas. I. Evidence that transforming growth factor-beta contributes to the altered expression of activation receptors on host B lymphocytes. J. Immunol. 146,2865-2872. Bertoletti, A., Chisari, F. V., Penna, A., Guilhot, S., Galati, L., Missale, G., Fowle, P., Schlicht, H.-J., Vitiello, A., Chesnut, R. C., Fiaccadori, F., and Ferrari, C. (1993). Definition of a minimal optimal cytotoxic T-cell epitope with the hepatitis B virus nucleocapsid protein. J. Virol. 67,2376-2380. Bevan, M. J . (1975). Interaction antigens detected by cytotoxic T cells with the major histocompatibility complex as modifier. Nature (London)256,419-421. Bishop, M. (1991). Molecular themes in oncogenesis. Cell (Cambridge, Mass.) 64, 235-248. Bix, M., Liao, N.-S., Zijlstra, M., Loring, J., Jaenisch, R., and Raulet, D. (1991).Rejection of class I MHC-deficient haematopoietic cells by irradiated MHC-matched mice. Nature (London)349,329-331. Bjorkman, P. J., and Parham, P. (1990). Structure, function and diversity of class I major histocompatibility complex molecules. Annu. Rev. Biochem. 59,253-288. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C. (1987). Structure of the human class I histocompatibility antigen, HLA-A2. Nature (London) 329,506-512. Blanchet, O., Borge, J., Zinszner, H., Israel, A., Kourilsky, P., Dausset, J., Degos, L., and Paul, P. (1992). Altered binding of regulatory factors to HLA class I enhancer sequence in human tumor cell lines lacking class I antigen expression. Proc. Natl. Acad. Sci. USA.89,3488-3492. Blankenstein, T., Li, W., Muller, W., and Diamantstein, T. (1990).Retroviral interleukin 4 gene transfer into interleukin 4-dependent cell line results in autocrine growth but not in tumorigenicity. Eur. J. Immunol. 20, 935-938. Bodmer, J. G., Tonks, S., Oza, A. M., Lister, T. A,, and Bodmer, W. F. (1989). HLA-DP based resistance to Hodgkin’s disease. Lancet 1, 1455-1456. Bogen, B., and Weiss, S. (1988).B lymphoma cells process and present their endogenous IG: Implications for network theory. In “Lymphatic Tissues and Germinal IG Centres in Immune Reactions” (S. Fossum and B. Rotstad, eds.). Plenum Press, New York. Boog, C. J. P., Boes, J., and Melief, C. J. M. (1988). Role of dendritic cells in the

328

CLAUDE ROTH ET AL.

regulation of class I restricted cytotoxic T lymphocyte responses. J . Immunol. 140, 3331-3337. Boon, T. (1983). Antigenic tumor variants obtained with mutagens. Ado. Cancer Res. 39,121-151. Boon, T. (1992). Toward a genetic analysis of tumor rejection antigens. Ado. Cancer Res. 58, 177-210. Boon, T., and Kellermann, 0. (1977). Rejection by syngeneic mice of cell variants obtained by mutagenesis of a malignant teratocarcinoma cell line. Proc. Natl. Acad. Sci. U.S.A.74, 272-275. Boone, C. W., Paranjpe, M., Orme, T., and Gillette, R. (1974). Virus-augmented tumor transplantation antigens: Evidence for a helper antigen mechanism. Int. J. Cancer 13,543-551. Bos, J. L., Fearon, E. R., Hamilton, S. R., Verlaan-de-vries, M., van Boom, J. H., van der Eb, A. J., and Vogelstein, B. (1987a). Prevalence of ras gene mutations in human colorectal cancers. Nature (London)327,293-297. Bos, J. L., Verlaan-de-vries, M., van der Eb, A. J., Janssen, R., DeIwel, B., Lowenberg, B., and Colly, L. P. (1987b). Mutations in N-ras predominate in acute myeloid leukemia. Blood 69, 1237-1241. Bouwes Bavinck, J. M., Vermeer, B. J., Van der Voude, F. J., Vandenbroucke, J. P., Schreuder, G. M. T., Thorogood, J., Persijn, G. G. and Claas, F. H. J. (1991). Relation between skin cancer and HLA antigens in renal-transplant recipients. N . Engl. J . Med. 325,843-848. Boyd, J. A., and Kaufman,D. G. (1990). Expression of transforming growth factor beta 1 by human endometrial carcinoma cell lines: Inverse correlation with effects on growth rate and morphology. Cancer Res. 50,3394-3399. Brichard, V., van Pel, A., Wolfel, T., Wolfel, C., DePlaen, E., Lethb, B., Coulie, P., and Boon, T. (1993). The tyrosinase gene codes for an antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanomas. J . Erp. Med. 178,489-495. Brodsky, F. M., and Guagliardi, L. E. (1991). The cell biology of antigen processing and presentation. Annu. Reu. Immunol. 9, 707-744. Brown, J. H., Jardetzky, T. S., 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 histocompatibility molecules. Nature (London)352,845-850. Brown, J. H., Jardetzky, T. S., Gorga, J. C., Stern, L. J., Urban, R. G., Strominger, J. L., and Wiley, D. C. (1993). Three-dimensional structure ofthe human class I1 histocompatibility antigen HLA-DR1. Nature (London)364,33-34. Brown, J. P., Wright, P. W., Hart, C. E., Woodbury, R. G., Hellstrom, K. E., and Hellstrom, I. (1980). Protein antigens ofnormal and malignant human cells identified by immunoprecipitation with monoclonal antibodies. J. B i d . Chem. 255,4980-4983. Brown, J. P., Woodbury, R. G., Hart, C., Hellstrom, I., and Hellstrom, K. E. (1981). Quantitative analysis of melanoma-associated antigen p97 in normal and neoplastic tissues. Proc. Natl. Acad. Sci. U.S.A.78, 539-543. Brown, J. P., Hewick, R. M., Hellstrom, I., Doolittle, R. F., and Dreyer, R. F. (1982). Human melanoma-associated antigen p97 is structurally and functionally related to transferrin. Nature (London)296, 171-173. Browning, M. J., and Bodmer, W. F. (1992). MHC antigens and cancer: Implications for T-cell surveillance. Curr. Opinion Immunol. 4, 613-618. Bu, D., Domenech, N., Lewis, J., Taylor-Papadimitriou, J., and Finn, 0. J. (1993). Recombinant vaccinia mucin vector: In uitro analysis of expression of tumorassociated epitopes for antibody and human cytotoxic T-cell recognition. J . Immunother. 14, 127-135.

IMMUNE RESPONSE AGAINST TUMORS

329

Bubenik, J., Voitenok, N. N., Kieler, J., Prassolov, V., Chumakov, P. M., Bubenikova, D., Simova, J., and Jandlova, T. (1988). Local administration of cells containing an inserted IL-2 gene and producing IL-2 inhibits growth of human tumours in nulnu mice. Immunol. Lett. 19,279-282. Bubenik, J., Simova, J., and Jandlova, T. (1990). lmmunotherapy of cancer using local administration of lymphoid cells expression IL-2 cDNA and constitutively producing IL-2. Immunol. Lett. 23,287-292. Buendia, M. A. (1992). Mammalian hepatitis B viruses and primary liver cancer. I n “Seminars in Cancer Biology: Viruses and Human Cancer” (A. B. Rickinson, ed.), pp. 309-320. Saunders Scientific Publications/Academic Press, London. Burchell, J., Taylor-Papadimitriou, J., Boshell, M., Gendler, S., and Duhig, T. (1989). Short sequence, within the amino acid tandem repeat of a cancer-associated mucin contains immunodominant epitopes. Int. J. Cancer 44,691-696. Burnet, F. M. (1970). A certain symmetry: Histocompatibility antigens compared with immune receptors. Nature (London)226,123-126. Burrows, S. R., Sculley, T. B., Misko, I. S., Schmidt, C., and Moss, D. J. (1990). An Epstein-Barr virus-specific T cell epitope in EBNA-3. J. Exp. Med. 171,345-349. Buus, S., Colon, S., Smith, C., Freed, J. H., Miles, C., and Grey, H. (1986). Interaction between a “processed” ovalbumin peptide and Ia molecules. Proc. Natl. Acad. Sci. U.S.A. 83,3968-3971. Buus, S., Sette, A., Colon, S. M., Miles, C., and Grey, H. (1987). The relation between major histocompatibility complex (MHC) restriction and the capacity of Ia to bind immunogenic peptides. Science 235,1353-1358. Carbone, F., and Bevan, M. J. (1989).Induction of ovalbumin-specific cytotoxic T cells by in vivo peptide immunization. J. Exp. Med. 169,603-612. Carswell, E. A., Wanebo, H. J., Old, L. J., and Boyse, E. A. (1970). Immunogenic properties of reticulum cell sarcomas of SJL/J mice. J . Natl. Cancer Inst. 44, 1281-1288. Cavallo, F., Giovarelli, M., Gulino, A., Vacca, A., Stoppacciaro, A., Modesti, A., and Forni, G. (1992). Role of neutrophils and CD4 + T lymphocytes in the primary and memory response to nonimmunogenic murine mammaryadenocarcinoma made immunogenic by IL-2 gene. J. Immunol. 149,3627-3635. Chan, B., Matsuura, N., Takada, Y., et al. (1991).In uitro and in uiuo consequences of VLA-2 expression on rhabdomyosarcoma cells. Science 251, 1600-1602. Chapman, P. B., and Houghton, A. N. (1991). Induction of IgG antibodies against GD3 ganglioside in rabbits by an anti-idiotypic monoclonal antibody. J . Clin. Znuest. 88, 186-192. Chapman, P. B., and Houghton, A. N. (1993). Non-antibody immunotherapy of cancer. Curr. Opin. Immunol. 5,726-731. Chellapan, S., Kraus, V. B., Kroger, B., Munger, K., Howley, P. M., Phelps, W. L., and Nevins (1992).Adenovirus ElA, simiavirus 40tumor antigen, and human papillomavirus E7 protein share the capacity to disrupt the interaction between transcription factor E2F and the retinoblastoma gene product. Proc. Natl. Acad. Sci. U.S.A. 89, 4549-4553. Chen, L., Ashe, S., Brady, W. A,, Hellstrom, I., Hellstrom, K. E., Ledbetter, J. A., McCowan, P., and Linsley, P. S. (1992). Costimulation of antitumor immunity by the B7 counterreceptor for the T lymphocyte molecules CD28 and CTLA-4. Cell (Cambridge, Mass.) 71,1093-1102. Chen, M., Holskin, B., Stickler, J., Gorniak, J., Clark, M. A,, Johnson, P. J., Mitcho, M., and Shalloway, D. (1987). Induction by E1A oncogene expression of cellular susceptibility to lysis by TNF. Nature (London)330, 581-583.

330

CLAUDE ROTH ET AL

Chen, W., Peace, D. J., Rovira, D. K., You, S.-G., and Cheever, M. A. (1992). T-cell immunity to the joining region of p21OBCR-ABL protein. Proc. Natl. Acad. Sci. U.S.A. 89,1468-1472. Cheung, N. K. V., Munn, D., Kushner, B. H., Usmani, N., and Yeh, S. D. J. (1989) Targeted radiotherapy and immunotherapy of human neuroblastoma with GD2 specific monoclonal antibodies. Nucl. Med. Biol. 16, 111-120. Cianciolo, G. J., Copeland, T. D., Oroszlan, S., and Snyderman, R. (1985). Inhibition of lymphocyte proliferation by a synthetic peptide homologous to retroviral envelope proteins. Science 230, 453-455. Cibotti, R., Kanellopoulos, J. M., Cabaniols, J.-P., Halle-Pannenko, O., Kosmatopoulos, K., Seercarz, E., and Kourilsky, P. (1992). Tolerance to a self-protein involves its immunodominant but not subdominant determinants. Proc. Natl. Acad. Sci., U.S.A. 89,416-420. Coffey, R. J., Jr., Shipley, G. D., and Moses, H. L. (1986). Production of transforming growth factors by human colon cancer lines. Cancer Res. 46, 1164-1169. Coggin, J. H., and Anderson, N. G. (1974). Cancer, differentiation and embryonic antigens: Some central problems. Ado. Cancer Res. 19,105-165. Cohen, I. R. (1991). Autoimmunity to chaperonins in the pathogenesis of arthritis and diabetes. Annu. Rev. lmmunol. 9, 567. Colombo, M. P., Modesti, A., Parmiani, G., and Forni, G. (1992). Local cytokine availability elicits tumor rejection and systemic immunity through granulocyte-T-lymphocyte cross-talk. Cancer Res. 52, 4853-4857. Cordon-Cardo, C., Fuks, Z., Drobnjak, M., Moreno, C., Eisenbach, L., and Feldman, M. (1991). Expression of HLA-A,B,C antigens on primary and metastatic tumor cell populations of human carcinomas. Cancer Res. 51,6372-6380. Currie, G. A., and Bagshawe, K. D. (1969).Tumor specific immunogenicity ofmethylcholanthrene-induced sarcoma cells after incubation in neuraminidase. Br. J . Cancer 23, 141- 149. DAndrea, A., Aste-Amezaga, M., Valiante, N. M., Ma, X., Kubin, M., and Trinchieri, G. (1993). Interleukin 10 (IL-10) inhibits human lymphocyte interferon gammaproduction by suppressing natural killer cell stimulatory factorlIL-12 synthesis in accessory cells. J . E x p . Med. 178, 1041-1048. Davidoff, A., Iglehardt, J. D., and Marks, J. R. (1992). Immune response to p53 is dependent upon p53/HSP70 complexes in breast cancers. Proc. Natl. Acad. Sci. U.S.A. 89,3439-3442. De Baetselier, P., Katzav, S., Gorelik, E., Feldman, M., and Segal, S. (1980). Differential expression of H-2 gene products in tumor cells is associated with their metastatic properties. Nature (London)288,179-181. De Campos-Lima, P. O., Gavioli, R., Zhang, Q.-J., Wallace, L., Rowe, M., Rickinson, A. B., and Masucci, M. G. (1993). HLA-A11 epitope loss isolates of Epstein-Barr virus from a highly A l l + population. Science 260,98-100. Deckhut, A. M., and Tevethia, S. S. (1992). Effect of point mutations in the native simian virus 40 tumor antigen and in synthetic peptides corresponding to the H-2Dbrestricted epitopes, on antigen presentation and recognition by cytotoxic T lymphcoyte clones. J . Immunol. 146,3012-3020. Deckhut, A. M., Lippolis, J. D., and Tevethia, S. S. (1992). Comparative analysis of core amino acid residues of H-2Db-restricted cytototix T-lymphocyte recognition epitopes in SV40 T antigen. J . Virol. 66,440-447. Defrance, T., Vanbervliet, B., Briere, F., Durand, I., Rousset, F., and Banchereau, J. (1992). Interleukin 10 and transforming growth factor B cooperate to induce anti-

IMMUNE RESPONSE AGAINST TUMORS

33 1

CD4O-activated naive human B cells to secrete immunoglobulin A. J. Exp. Med. 175, 671-682. Deleo, A. B., Shiku, H., Takahashi, T., John, M., and Old, L. J. (1977). Cell surface antigens of chemically-induced sarcomas of the mouse. I. Murine leukemia virusrelated antigens and alloantigens on cultured fibroblasts and sarcoma cells: Description of a unique antigen on BALBIc meth A sarcoma. J. Exp. Med. 146, 720-734. Deleo, A. B., Jay, G., Appella, E., Dubois, G. C., Law, L. W., and Old, L. J. (1979). Detection of a transformation-related antigen in chemically induced sarcomas and other transformed cells of the mouse. Proc. Natl. Acad. Sci. U.S.A. 76,2420-2424. De Nagel, D. C., and Pierce, S. K. (1992). A case for chaperones in antigen processing. Immunol. Today 13,86-89. De Nagel, D. C., and Pierce, S. K. (1993). Heat-shock proteins in immune responses. Crit. Rev. Immunol. 13,71-81. DePlaen, E., Lurquin, C., van Pel, A., Mariame, B., Szikora, J. P., Wolfel, T., Sibille, C., Chomez, P., and Boon, T. (1988). Immunogenic (turn-) variants of mouse tumor p815: Cloning ofthe gene oftum- antigen P91 and identification ofthe tum- mutation. PTOC.Natl. Acad. Sci. U.S.A. 85,2274-2278. Deres, K., Schild, H., Wiesmiiller, K. H., Jung, G., and Rammensee, H.-G. (1989). In vivo priming of virus-specific cytotoxic T lymphoeytes with synthetic lipopeptide vaccine. Nature (London)342, 561-564. Derynck, R., Goeddel, D. V., Ullrich, A., Gutterman, J. U., Williams, R. D., Bringman, T. S., and Berger, W. H. (1987). Synthesis ofmessenger RNAs for transforming growth factors alpha and beta and the epidermal growth factor receptor by human tumors. Cancer Res. 47,707-712. Ding, L., Lalani, E. N., Reddish, M., Koganty, R., Wong, T., Samuel, J., Yacyshyn, M. B., Meikle, A., Fung, P. Y. S., Taylor-Papadimitriou, J., and Longenecker, B. M. (1993). Immunogenicity of synthetic peptides related to the core peptide sequence encoded by the human MUCl mucin gene: Effect of immunization on the growth of murine mammary adenocarcinoma cells transfected with the human MUCl gene. Cancer Immunol. Immunother. 36,9-17. Dippold, W. G., Lloyd, K. O., Li, L. T. C., Ikeda, H., Oettgen, H. F., and Old, L. J. (1980). Cell surface antigens of human malignant melanoma: Definition of six antigenic systems with mouse monoclonal antibodies. Proc. Natl. Acad. Sci. U.S.A.77, 6114-61 18. Dranoff, G., Jaffee, E., Lazenby, A., Golumbek, P., Levitsky, H., Brose, K., Jackson, V., Hamada, H., Pardoll, D., and Mulligan, R. C. (1993). Vaccination with irradiated tumor cells engineered to secrete murine GM-CSF stimulates potent, specific, and long lasting anti-tumor immunity. Proc. Natl. Acad. Sci. U.S.A.90,3539-3543. Earl, P. L., Moss, B., Morrisson, R. P., Wehrly, K., Nishio, J., and Chesebro, B. (1986). T lymphocyte priming and protection against Friend leukemia by vaccinia-retrovirus env gene recombinant. Science 234,728-731. Eisenbach, L., Segal, S., and Feldman, M. (1983). MHC imbalance and metastatic spread in Lewis lung carcinoma clones. Int. J . Cancer 32, 113-120. Eisenbach, L., Hollander, N., Greenfeld, L., Yakor, H., Segal, S., and Feldman, M. (1984). The differential expression of H2-K versus H2-D antigens, distinguishing low metastatic from high metastatic clones, correlated with the immunogenic properties of the tumor cells. Int. J . Cancer 34,567-573. Elliot, B. E., Carlow, D. A,, Rodrick, A. M., and Wade, A. (1989). Perspectives on the role of MHC antigens in normal and malignant cell development. Adu. Cancer Res. 53,181-244.

332

CLAUDE ROTH ET AL.

Estin, C. D., Stevenson, U. S., Plowman, G. D., Hu, S . L., Sridhar, P., Hellstrom, I., Brown, J. P., and Hellstrom, K. E. (1988). Recombinant vaccinia vaccine against the human melanoma antigen p97 for us in immunotherapy. Proc. Natl. Acad. Sci. U.S.A. 85,1052-1056. Esumi, N., Hunt, B., Itaya, T., and Frost, P. (1991). Reduced tumorigenicity of murine tumor cells secreting gamma-interferon is due to nonspecific host responses and is unrelated to class I major histocompatibility complex expression. Cancer Res. 51, 1185-1 189. Falk, K., Rotzschke, O., Stevanovic, S., Jung, G., and Rammensee, H.-G. (1991).Allelespecific motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature (London)351,290-296. Farr, C., Saiki, R., Erlich, H. A., McCormick, F., and Marshall, C. J. (1988). Analysis of ras gene mutations in acute myeloid leukemia by polymerase chain reaction and oligonucleotide probes. Proc. Natl. Acad. Sci. U.S.A.85, 1629-1633. Fearon, E. R., Pardoll, D. M., Itaya, T., Golumbek, P., Levitsky, H. I., Simons, J. W., Karasuyama, H., Vogelstein, B., and Frost, P. (1990). Interleukin-2 production by tumor cells bypasses T helper function in the generation of an antitumor response. Cell (Cambridge, Mass.)60, 397-403. Fearon, E. R.,and Vogelstein, B. (1990). A genetic model for colorectal tumorigenesis. Cell 61,767-769. Feitelson, M. A., Zhu, M., Duan, L.-X., and London, W. T. (1993). Hepatitis Bx antigen and p53 are associated in uitro and in liver tissues from patients with primary hepatocellular carcinoma. Oncogene 8, 1109-1 117. Feldman, M., and Eisenbach, L. (1991). MHC class I genes controlling metastatic phenotype of tumor cells. Semin. Cancer Biol. 2,337-346. Feltkamp, M. C. W., Smits, H. L., Vierboom, M. P. M., Minnaar, R. P., De Jongh, B. M., Drijfhout, J. W., Ter Schegget, J., Melief, C. J. M., and Kast, W. M. (1993). Vaccination with cytotoxic T lymphocyte epitope-containing peptide protects against a tumor induced by human papillomavirus type 16-transformedcells. Eur.J. lmmunol. 23,2242-2249. Fenton, R. G., Taub, D. T., Kwak, L. W., Smith, M. R.,and Longo, D. L. (1993).Cytotoxic T-cell response and in uioo protection against tumor cells harboring activated ras proto-oncogenes. J. Natl. Cancer lnst. 85, 1294-1302. Finke, J. H., Zea, A. H., Stanley, J., Longo, D. L., Mizoguchi, H., Tubbs, R. R., Wiltrout, R. H., O’Shea, J. J., Kudoh, S., Klein, E., Bukowski, R. M., and Ochoa, A. C. (1993). ‘ ~T-cells infiltrating human renal cell Loss of T-cell receptor 1; chain and ~ 5 6 ’ in carcinoma. Cancer Res. 53,5613-5616. Finn, 0.J. (1992a).Antigen-specific MHC-unrestricted T cells. Biotherapy 4,159-170. Finn, 0. J. (1992b). Pancreatic tumor antigens: Diagnostic markers and targets for immunotherapy. In “Important Advances in Oncology”, (V. T. De Vita, S. Hellman, and S . A. Rosenberg, eds.), pp. 61-77. Lippincott, Philadelphia. Fiorentino, D., Bond, M., and Mosmann, T. (1989). Two types of mouse T helper cells. IV. Th2 clones secrete a factor that inhibits cytokine production by T h l clones. J . E r p . Med. 170,2081-2095. Foley, E.J. (1953). Antigenic properties of a methyl cholanthrene-induced tumors in mice of the strain origin. Cancer Res. 13,835-837. Folkman, J., Watson, K., Ingber, D., and Hanahan, D. (1989). Induction of angiogenesis during transition from hyperplasia to neoplasia. Nature (London) 339,58-61. Forrester, K., Almoguera,C., Han, K., Grizzle, W. E., and Perucho, M. (1987).Detection of high incidence of K-ras oncogenes during human colon tumorigenesis. Nature (London)327,298-303.

IMMUNE RESPONSE AGAINST TUMOR3

333

Franksson, L., George, E., Powis, S., Butcher, G., Howard, J., and Karre, K. (1993). Tumorigenicity conferred to lymphoma mutant by major histocompatibility complexencoded transporter gene. ]. E x p . Med. 177,201-205. Freeman, G. J., Gray, G. S., Gimmi, C. D., Lombard, D. B., Zhou, L.-J., White, M., Fingeroth, J. D., Gribben, J. G., and Nadler, L. M. (1991). Structure, expression, and T cell costimulatory activity of the murine homologue of the human B lymphocyte activation antigen B7. j . Exp. Med. 174,625-631. Freeman, G. J., Borriello, F., Hodes, R. J., Reiser, H., Hathcock, K. S., Laszlo, G., McKnight, A. J., Kim, J., Du, L., Lombard, D. B., Gray, G. S., Nadler, L. M., and Sharpe, A. H. (1993a). Uncovering of functional alternative CTLA-4 counter-receptor in B7-deficient mice. Science 262,907-909. Freeman, G. J., Gribben, J. G., Boussiotis, V. A,, Ng, J. W., Restivo, Jr., V. A., Lombard, L. A., Gray, G. S., and Nadler, L. M. (1993b). Cloning of B7-2: A CTLA-4 counterreceptor that costimulates human T cell proliferation. Science 262,909-91 1. Frost, P., and Levin, B. (1992). Clinical implications of the metastatic process. Lancet 339,1458-1461. Fujiwara, H., and Ellner, J. J. (1986). Spontaneous production of a suppressor factor by the human macrophage-like cell line U937. I. Suppression of interleukin 1, interleukin 2, and mitogen-induced blastogenesis in mouse thymocytes. j . Immunol. 136, 181-185. Furukawa, K. S., Furukawa, R., Real, F. X., Old, L. J., and Old, K. 0. (1989). A unique antigenic epitope of human melanoma is carried on the common melanoma glycoprotein gp95lp97. j . Ex p . Med. 169,585-590. Gansbacher, B., Zier, K., Daniels, B., Cronin, K., Banneji, R., and Gilboa, E. (1990a). Interleukin 2 gene transfer into tumor cells abrogates tumorigenicity and induces protective immunity. J. E x p . Med. 172, 1217-1224. Gansbacher, B., Banneji, R., Daniels, B., Zier, K., Cronin, K., and Gilboa, E. (1990b). Retroviral vector-mediated gamma-interferon gene transfer into tumor cells generates potent and long lasting antitumor immunity. Cancer Res. 50,7820-7825. Gasparini, G., Gullick, W. J., Bevilacqua, P., Sainsbury, J. R., Meli, S., Boracchi, P., Testolin, A., La-Malfa, G., and Pozza, F. (1992). Human breast cancer: Prognostic significance ofthe c-erbB-2 oncoprotein compared with epidermal growth factor receptor, DNA ploidy, and conventional pathologic features [see comments]. j . Clin. Oncol. 10,686-695. Gastl, G. A., Abrams, J. S., Nanus, D. M., Oosterkamp, R., Silver, J., Liu, F., Chen, M., Albino, A. P., and Bander, N. H. (1993). Interleukin-10 production by human carcinoma cell lines and its relationship to interleukin-6 expression. Int. J . Cancer 55, 96- 101. Gavioli, R., De Campos-Lima, P. O., Kurilla, M.G., Kieff, E., Klein, G., and Masucci, M. G. (1992). Recognition of the EBV encoded nuclear antigens EBNA-4 and EBNA6 by HLA-A11 restricted cytotoxic T lymphocytes: Implications for the down regulation of HLA-All in Burkitt’s lymphoma. Proc. Natl. Acad. Sci. U S A . 89, 58625866. Gavioli, R., Kurilla, M. G., De Campos-Lima, P. O., Wallace, L.E., Dolcetti, R., Murray, R. J., Rickinson, A. B., and Masucci, M. G. (1993). Multiple HLA-A11 restricted CTL epitopes of different immunogenicity in the Epstein-Barr virus (EBV)encoded nuclear antigen-4 (EBNA4).1.Virol. 67, 1572-1578. Gedde-Dahl 111, T., Eriksen, J. A,, Thorsby, E., and Gaudemack, G. (1992a). T-cell responses against products of oncogenes: Generation and characterization of human T-cell clones specific for p21 ras-derived synthetic peptides. Hum. Immunol. 33, 266-274.

334

CLAUDE ROTH ET AL.

Gedde-Dahl 111, T., Spurkland, A., Eriksen, J. A., Thorsby, E., and Gaudernack, G. (199213). Memory T cells of a patient with follicular thyroid carcinoma recognize peptides derived from mutated p21 ras (Gln-->LeuGl). Znt. Zmmunol. 4, 1331-1337. Gedde-Dahl 111, T., Fossum, B., Eriksen, J. A., Thorsby, E., and Gaudernack, G. (1993a). T cell clones specific for p21 ras-derived peptides: Characterization of their fine specificity and HLA restriction. Eur.]. Zmmunol. 23,754-760. Gedde-Dahl 111, T., Olsen, B. H., Thorsby, E., and Gaudernack, G. (1993b).Oncogenederived peptides: A new class of tumor rejection antigens? Trunsplant. Proc. 25, 1385-1386. Gelber, C., Plaksin, D., Vaday, E., Feldman, M., and Eisenbach, L. (1989).The abolishment of metastasis formation by tumor cells transfected with “foreign” H-2K genes. Cancer Res. 49,2366-2372. Gendler, S., Taylor-Papadimitriou, J., Duhlig, T., Rothbard, J., and Burchell, J. (1988). A highly immunogenic region of a human polymorphic epithelial mucin expressed by carcinomas is made up of tandem repeats. J . B i d . Chem. 263, 12820-12823. George, A. J. T., Tutt, A. L., and Stevenson, F. K. (1987). Anti-idiotypic mechanisms involved in suppression ofamouse B cell lymphoma, BCL1.J. Immunol. 138,628-634. Gimmi, C. D., Freeman, G. J., Gribben, J. G., Sugita, K., Freedman, A. S., Morimoto, C., and Nadler, L. M. (1991).B-cell surface antigen B7 provides a costimulatory signal that induces T cells to proliferate and secrete interleukin 2. Proc. Natl. Acad. Sci. U.S.A.88, 6575-6579. Gimmi, C. D., Freeman, G. J., Gribben, J. G., Gray, G. S., and Nadler, L. M. (1993). Human T-cell clonal anergy is induced by antigen presentation in the absence of B7 costimulation. Proc. Natl. Acad. Sci. U.S.A.90, 6586-6590. Girling, A., Bartkova, J., Burchell, J., Gendler, S., Gillett, C., and Taylor-Papadimitriou, J. A. (1989). Core protein epitope of the polymorphic epithelial mucin detected by the monoclonal antibody SM-3 is selectively exposed in a range of primary carcinomas. Znt. J. Cancer 43, 1072-1076. Glas, R., Sturmhofel, K., Hammerling, G., Karre, K., and Ljunggren, H. (1992).Restoration of a tumorigenic phenotype by p2-microglobulin transfection to EL-4 mutant cells. J. Exp. Med. 175,843-846. Golumbek, P. T., Lazenby, A. J., Levitsky, H. I., Jaffee, E. M., Karasuyama, H., Baker, M., and Pardoll, D. M. (1991).Treatment of established renal cancer by tumor cells engineered to secrete interleukin-4. Science 254,713-716. Gooding, L., Sofola, I., Tollefson, A., Duerksen-Hughes, P., and Wold, W. (1990). The adenovirus E3-14.K protein is a general inhibitor of tumor necrosis factor-mediated cytolysis. J. Immunol. 145,3080-3086. Gopas, J., Rager-Zisman, B., Har-Vardi, I., Hammerling, H.-G., Bar-Eli, M., and Segal, S. (1989). NK sensitivity, H-2 expression and metastatic potential: Analysis of H-2Dk gene transfected fibrosarcoma cells. J. lmmunogenet. 16, 305-313. Gorelik, E., Jay, G., Kwiatkowski, B., and Herberman, R. B. (1990).Increased sensitivity to MHC-nonrestricted lysis of BL6 melanoma cells by transfection with class I H2Kb gene. J. Immunol. 145, 1621-1632. Gorelik, E., Jay, G., Kim, M., Hearing, V. J., Deleo, A., and McCoy, J. P. (1991). Effects of H-2Kb gene on expression of melanoma-associated antigen and lectin-binding sites on BL6 melanoma cells. Cancer Res. 51,5212-5218. Gorelik, E., Kim, M., Duty, L., and Herberman, R. B. (1994). Nonimmunological effects of MHC class I genes in BL6 melanoma cells: Comparative effects of H-2K, H-2D and H-2L genes. Submitted for publication. Gotlieb, W. H., Abrams, J. S., Watson, J. M., Velu, T. J., Berek, J. S., and Martinez-

IMMUNE RESPONSE AGAINST TUMORS

335

Maza, 0. (1992). Presence of interleukin 10 (IL-10) in the ascites of patients with ovarian and other intra-abdominal cancers. Cytokine 4,385-390. Greenberg, P. D. (1991). Adoptive T cell therapy of tumors: Mechanisms operative in the recognition and elimination of tumor cells. Adu. Immunol. 49,281-355. Greenberg, P. D., Cheever, M. A., and Feffer, A. (1981). Eradication of disseminated murine leukemia by chemoimmunotherapy with cyclophosphamide and adoptively transferred immune syngeneic Lyt-1'2- lymphocytes. J. E x p . Med. 154,952-963. Gross, L. (1943). Interdermal immunization of C3H mice against a sarcoma that originated in an animal of the same line. Cancer Res. 3,326-333. Guilhot, S., Fowler, P., Portillo, G., Margolskee, R. T., Ferrari, C., Bertoletti, A., and Chisari, F. V. (1992). Hepatitis B virus (HBV)-specific cytotoxic T-cell responses in human: Production of target cells by stable expression of HBV-encoded protein in immortalized human B cell lines. J. Virol. 66,2670-2678. Habel, K. (1961). Immunological determinants of polyoma virus oncogenesis. J. E x p . Med. 115,181-193. Hainaut, P., Weynants, P., Coulie, P., and Boon, T. (1990). Anti-tumor T lymphocyte responses: Immunology and allergy clinics of North America. Human Cancer Irnmunol. 10,639-662. Haliotis, T., Carlow, D. A., and Elliott, B. E. (1990). Non-immunological aspects of MHC-function in the regulation of cell proliferation and the malignant phenotype. Cancer Cells 2, 86-90. Han, R., Breitburd, F., Marche, P. N., and Orth, G. (1992). Linkage of regression and malignant conversion of rabbit viral papillomas to MHC class I1 genes. Nature (London) 356,66-68. Harding, F. A., and Allison, J. P. (1993). CD28-B7 interactions allow the induction of CD8 + cytotoxic T lymphocytes in the absence of exogenous help. J. Exp. Med. 177, 1791-1796. Hareuveni, M., Tsarfaty, I., Zaretsky, J., Kotkes, P., Horev, J., Srihan, S., Weiss, M., Green, S., Lath6, R., Keydar, I., and Wreschner, D. H. (1990a). A transcribed gene, containing a variable number of tandem repeats, codes for a human epithelial tumor antigen: cDNA cloning, expression of the transfected gene and over-expression in breast cancer tissue. Eur. J. Biochem. 189,475-486. Hareuveni, M., Gautier, C., Kieny, M.-P., Wreschner, D. H., Chambon, P., and Lathe, R. (1990b). Vaccination against tumor cells expressing breast cancer epithelial tumor antigen. Proc. Natl. Acad. Sci. U.S.A. 87,9498-9502. Hart, I. R., and Easty, D. (1990). Identification of genes controlling metastatic behaviour. Br. J . Cancer 63,9-12. Hart, I. R., and Saini, A. (1992). Biology of tumor metastasis. Lancet 339, 1453-1457. Hathcock, K. S., Laszlo, G., Dickler, H. B., Bradshaw, J., Linsley, P., and Hodes, R. J. (1993). Identification of an alternative CTLA-4 ligand costimulatory for T cell activation. Science 262,905-907. Hayashi, H., Tanaka, K., Jay, F., Khoury, G., and Jay, G. (1985). Modulation of the tumorigenicity of human adenovirus-12-transformed cells by interferon. Cell (Cambridge, Mass.) 43,263-267. Hellstrom, I., and Sjogren, H. 0. (1966). Demonstration of common specific antigen(s) in mouse and hamster polyoma tumors. Int. J. Cancer 1,481-489. Hermans, A., Heisterkamp, N., von Lindern, M., van Baals, S., Meijer, D., van der Plas, D., Wiedemann, L. M., Groffen, J., Bootsma, D., and Grosveld, D. (1987). Unique fusion of bcr and c-abl genes in Philadelphia chromosome positive acute lymphoblastic leukemia. Cell (Cambridge, Mass.) 51, 33-37.

336

CLAUDE ROTH ET AL.

Hirte, H., and Clark, D. A. (1991). Generation of lymphokine-activated killer cells in human ovarian carcinoma ascitic fluid: Identification of transforming growth factorbeta as a suppressive factor. Cancer Immunol. Immunother. 32,296-302. Hock, H., Dorsch, M., Diamantstein T., and Blankenstein, T. (1991). Interleukin 7 induces CD4 + T cell-dependent tumor rejection. J . Erp. Med. 174, 1291-1298. Hock, H., Dorsch, M., Kunzendorf, U., Qin, Z., Diamantstein, T., and Blankenstein, T. (1993a). Mechanisms of rejection induced by tumor cell-targeted gene transfer of interleukin 2, interleukin 4, interleukin 7, tumor necrosis factor, or interferon gamma. Proc. Natl. Acad. Sci. U.S.A.90,2774-2778. Hock, H., Dorsch, M., Kunzendorf, U., Uberla, K., Qin, Z., Diamantstein, T., and Blankenstein, T. (1993b).Vaccinations with tumor cells genetically engineered to produce different cytokines: Effectivity not superior to a classical adjuvant. Cancer Res. 53, 714-716. Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C. (1991). p53 mutations in human cancer. Science 253,49-53. Houbiers, J. G. A., Nijman, H. W., van der Burg, S. H., Drijfhout, J. W.,‘Kenemans, P., van de Velde, C. J. H., Brand, A., Momburg, F., Kast, W. M., and Meliekc. J. M. (1993).In uitro induction of human cytotoxic T lymphocyte responses against peptides of mutant and wild-type p53. Eur. J . Immunol. 23,2072-2077. Houghton, A. N. (1992).In “Immunobiology of Cancer: Cellular and Molecular Mechanisms.” Forty-fifth Annual Symposium on Fundamental Cancer Research, Houston, TX. Houghton, A. N., Mintzer, D., Cordon-Cardo, C., Welt, S., Fliegel, B., Vadhan, S., Carswell, E., Melamed, M. R., Oettgen, H. F., and Old, L. J. (1985).Mouse monoclonal IgG3 antibody detecting GD3 ganglioside: A phase I trial in patients with malignant melanoma. Proc. Natl. Acad. Sci. U.S.A. 82, 1242-1246. Hu, S. L., Plowman, G. D., Sridhar, P., Stevenson, U.S., Brown, J. P., and Estin, C. D. (1988).Characterization of a recombinant vaccinia virus expressing human melanomaassociated antigen p97. ]. Virol. 62, 176-180. Hui, K., Grosveld, F., and Festenstein, H. (1984). Rejection of transplantable AKR leukaemic cells following MHC DNA-mediated cell transformation.Nature (London) 311,750-752. Hui, K. M., Sim, B. C., Foo, T. T., and Oei, A. A. (1991). Promotion of tumor growth by transfecting antisense DNA to suppress endogenous H-2Kk MHC gene expression in AKR mouse thymoma. Cell. Immunol. 136,80-94. Hynes, R. 0. (1987). Integrins: A family of cell surface receptors. Cell (Cambridge, MUSS.)48, 549-554. Hynes, R. 0. (1992). Integrins: Versatility, modulation, and signaling in cell adhesion. Cell (Cambridge,Mass.) 69, 11-25. Inoue, T.,Asano, Y.,Matsuoka, S., Furutani-Seiki, M., Aizawa, S., Nishimura, H., Shirai, T., and Tada, T. (1993).Distinction of mouse CD8 + suppressor effector T cell clones from cytotoxic T cell clones by cytokine production and CD45 isoforms.J. Immunol. 150,2121-2128. Invernizzi, G., and Parmiani, G. (1975). Tumor-associated transplantation antigens of chemically-induced sarcomata cross-reacting with allogeneic histocompatibility antigens. Nature (London)254,713-714. Ioannides, C . G., Freedman, R. S., Ioannides, M. G., Fisk, B., and O’Brian, C. A. (1992).CTL clones isolated from ovarian tumor infiltrating lymphocytes can recognize peptides with sequences corresponding to the HER2Ineu gene product. FASEB J . 6, 2711.

IMMUNE RESPONSE AGAINST TUMORS

337

Isakov, N., Katzav, S., Feldman, M., and Segal, S. (1983).Loss ofexpressionoftransplantation antigens encoded by H2-K locus on Lewis lung carcinoma cells and its relevance to the tumor’s metastatic properties. J. Natl. Cancer Znst. 71, 139-145. Itaya, T., Yamagiwa, S., Okada, F., Oikawa, T., Kuzumaki, N., Takeichi, N., Hosokawa, M., and Kobayashi, H. (1987).Xenogenization of a mouse lung carcinoma (3LL) by transfection with an allogeneic class I Major Histocompatibility Complex gene (H2Ld). Cancer Res. 47,3136-3140. Jay, G., De Leo, A. B., Appella, E. Dubois, G. C., Law, L. W., Khoury, G., and Old L. J. (1979).A common transformation-related proteins in murine sarcomas and leukemia. Cold Spring Harbor Symp. Quant. Biol. 44,659-664. Jenkins, M. K., and Johnson, J. G. (1993). Molecules involved in T-cell costimulation. Curr. Opinion Zmmunol. 5, 361-367. Jerome, K., Bu, D., and Finn, 0. J. (1992). Expression oftumor-associated epitopes on Epstein-Barr virus-immortalized B-cells and Burkitt’s lymphomas transfected with epithelial mucin complementary DNA. Cancer Res. 52,985-990. Jerome, K. R., Barnd, D. L., Bendt, K. M., Boyer, C. M., Taylor-Papadimitriou, J., McKenzie, I. F. C., Bast, R. C. J., and Finn, 0. J. (1991). Cytotoxic T lymphocytes derived from patients with breast adenocarcinoma recognize an epitope present on the protein core of a mucin molecule preferentially expressed by malignant cells. Cancer Res. 51,2908-2916. Jerome, K. R., Domenech, N., and Finn, 0. J. (1993). Tumor-specific cytotoxic T cell clones from patients with breast and pancreatic adenocarcinoma recognize EBVimmortalized B cells transfected with polymorphic epithelial mucin complementary DNA. J. Zmmunol. 151,1654-1662. Johnson, J., Lehmann, J., Stade, B., et al. (1989). Functional aspects of three molecules associated with metastasis development in human malignant melanoma. Znuasion Metastasis, 9,338. Jonjic, N., Alberti, S.,Bernasconi, S., Peri, G., Jilek, P., Anichini, A., Parmiani, G., and Mantovani, A. (1992). Heterogenous sdsceptibility of human melanoma clones to monocyte cytotoxicity: Role of ICAM-1 defined by antibody blocking and gene transfer. Eur. J. Zmmunol. 22,2255-2260. Jung, S., and Schluesener, H. J. (1991). Human T lymphocytes recognize a peptide of single point-mutated, oncogenic ras proteins. J. E x p . Med. 173,273-276. Kaklamanis, L., Gatter, K. C., Hill, A. B., Mortensen, N., Harris, A. L., Krausa, P., McMichael, A., Bodmer, J. G., and Bodmer, W. F. (1992).Loss of HLA class4 alleles, heavy chains and /32-microglobulin in colorectal cancer. Znt. J. Cancer 51, 379385. Karlhofer, F. M., Ribaudo, R. K., and Yokoyama, W. M. (1992).MHC class I alloantigen specificity of Ly-49 + IL-2-activated natural killer cells. Nature (London)358, 66. Grre, K., Ljunggren, H. G., Piontek, G., and Kiessling, R. (1986). Selective rejection of H-2-deficient lymphoma variants suggests alteranative immune defense strategy. Nature (London)319,675-678. Kast, W . M., and Melief, C. J. M. (1991). Fine specificity of cytotoxic T lymphocytes directed against adenovirus induced tumors and peptide MHC binding. Znt.J. Cancer (Suppl.) 6,90-94. Kast, W. M., OEringa, R., Peters, P. J., Voordouw, A. C., Meloen, R. H., van der Eb, A. J., and Melief, C. J. M. (1989). Eradication of adenovirus El-induced tumors by E1A-specific cytotoxic T lymphcoytes. Cell (Cambridge,Mass.) 59,603-614. Kast, W . M., Roux, L.,Curran, J., Blom, H. J. J.,Voordouw,A., Meloen, R. H.,Kolakofsky, D., and Melief, C. J. M. (1991). Protection against lethal Sendai virus infection by in

338

CLAUDE ROTH ET AL.

uiuo priming of virus-specific cytotoxic T lymphocytes with a free synthetic peptide. Proc. Natl. Acad. Sci. U.S.A. 88,2283-2287. Katzav, S., Segal, S., and Feldman, M. (1984).Immunoselection in oioo of H-2D phenotypic variants from a metastatic clone of sarcoma cells results in cell lines of altered metastatic competence. Int. J . Cancer 33,407-415. Kaufmann, S . H. E. (1990). Heat shock protein and the immune response. ImmunoE. Today 11,129. Kaufman, D. S., Schoon, R. A., and Leibson, P. J. (1993). MHC class I expression on tumor targets inhibits natural killer cell-mediated cytotoxicity without interfering with target recognition. J. Immunol. 150, 1429-1436. Kawakami, Y., Zakut, R., Topalian, S. I., Stotter, H., and Rosenberg, S. A. (1992).Shared human melanoma antigens: Recognition by tumor-infiltrating lymphocytes in HLAA2.1-transfected melanomas. J. Immunol. 148, 638-643. Killion, J. J., and Fidler, I. J. (1989).The biology of tumor metastasis. Semin. Oncol. 16, 106-115. Kim, M., Herbennan, R., and Gorelik, E. (1993).Increased sensitivity to TNF-mediated cytotoxicity of BL6 melanoma cells after H-2Kb gene transfection. J . Immunol. 151, 1-11. Klarnet, J. P., Kern, D. E., Okuno, K., Holt, C., Lilly, F., and Greenberg, P. D. (1989). FBL-reactive CD8 cytotoxic and CD4 + helper T lymphocytes recognize distinct Friend murine leukemia virus-encoded antigens. J . E x p . Med. 169,457-467. Klein, G. (1991). Immunovirology of transforming viruses. Curr. Opinion Immunol. 3, 665-673. Klein, G., and Boon, T. (1993).Tumor immunology: Present perspectives. Curr. Opinion Immunol. 5,687-692. Klein, G., Sjogren, H. O., Klein, E., and Hellstrom, K. E. (1960). Demonstration of resistance against methylcholanthrene-inducedsarcomas in the primary autchthonous host. Cancer Res. 20, 1561-1572. Klitz, W. (1992).Viruses, cancer and the MHC. Nature (London)356,17-18. Knabbe, C., Lippman, M. E., Wakefield, L. M., Flanders, K. C., Kasid, A., Derynck, R., and Dickson, R. B. (1987).Evidence that transforming growth factor-beta is a hormonally regulated negative growth factor in human breast cancer cells. Cell (Cambridge, Mass.) 48,417-428. Knuth, A., Wolfel, T., Klehmann, E., Boon, T., and Meyer Zum Bueschenfelde, K. H. (1989). Cytolytic T-cell clones against an autologous human melanoma: Specificity study and definition of three antigens by immunoselection. Proc. Natl. Acad. Sci. U.S.A.86,2804-2808. Knuth, A., Wolfel, T., and Meyer Zum Bueschenfelde, K. H. (1992).T cell response to human malignant tumors. Cancer Sum. 13,39-52. Kobayashi, H., Sendo, F., Shirai, T., Kajo, H., Kodama, T., and Saito, H. (1969).Modification in growth of transplantable rat tumors exposed to Friend virus. J . Notl. Cancer Inst. 42, 413-419. Kobayashi, H., Takeichi, N., and Kuzumaki, N. (1978).Xenogenization of tumor cells in xenogenization of lymphocytes, erythroblasts and tumor cells. Hokaido Unio.Med. Library Ser. 10, 117-163. Kourilsky, P., and Claverie, J.-M. (1986).The peptidic self-model: A hypothesis in the molecular nature of the immunological self. Ann. fnst.Pasteurllmmunol. 137D, 3-21. Kourilsky, P., Chaouat, G., Rabourdin-Combe, C., and Claverie, J.-M. (1987). Working principles in the immune system implied by the "peptidic self" model. Proc. Notl. Acad. Sci. U.S.A. 84, 3400-3404.

+

IMMUNE RESPONSE AGAINST TUMORS

339

Kourilsky, P., Jaulin, C., and Ley, V. (1991). The structure and function of MHC molecules: Possible implication for the control oftumor growth by MHC-restricted T cells. Cancer Biol. 2,275-282. Kripke, M. L. (1981). Immunology mechanisms in U.V. radiation carcinogenesis. Adu. Cancer Res. 34,69-106. Kushtai, G., Barzilay, J., Feldman, M., and Eisenbach, L. (1988).The c-fos protooncogene in murine 3LL carcinoma clones controls the expression of MHC genes. Oncogene 2, 119-128. Kushtai, G., Feldman, M., and Eisenbach, L. (1990). c-fos transfection of 3LL tumor cells turns on MHC gene expression and consequently reduces their metastaticcompetence. Int. J. Cancer 45,1131-1136. Kuzumaki, N., Fenyo, E. M., Giovanella, B., and Klein, G. (1978).Increased immunogenicity of low-antigenic rat tumors after superinfection with endogenous murine Ctype virus in nude mice. Int. J. Cancer 21,62-66. Lachmann, P. J., and Sikora, K. (1978). Coupling PPD to tumor cells enhances their antigenicity in BCG-primed mice. Nature (London)271,463-464. Lam, K. T., and Caldenvood, S. K. (1992). HSP70 bind specifically to a peptide derived from the highly conserved domain (I)region of p53. Biophys. Biochem. Res. Commun. 184,167-174. Lanzavecchia, A. (1985). Antigen-specific interaction between T and B cells. Nature (London) 314,537-539. Lathe, R., Kieny, M.-P., Gerlinger, P., Clertant, P., Guizani, I., Cuzin, F., and Chambon, P. (1987). Tumor prevention and rejection with recombinant vaccinia. Nature (London) 326,878-880. Leclerc, J. C., and Cantor, H. (1980).T cell-mediated immunity to oncoma-virus-induced tumors. 11. Ability of different T cell sets to prevent tumor growth in uiuo.J . Immunol. 124,851-854. Lee, D. R., Rubocki, R. J., Lie, W. R., and Hansen, T. H. (1988). The murine MHC class I genes, H-2Dq and H-2L9 are strikingly homologous to each other, H-2Ld and two genes reported to encode tumor-specific antigens. J. E x p . Med. 168,1719-1739. Lehmann, J., Riethmuller, G., and Johnson, J. (1989). MUC 18, a marker of tumour progression in human melanoma, shows sequence similarity to the neural cell adhesion molecules of the immunoglobulin superfamily. Proc. Natl. Acad. Sci. U.S.A.86, 9891. Lennox, E. S., Lowe, A. D., Cohn, J., and Evan, G. (1981). Specific antigens on methylcholanthrene-induced tumors of mice. Transplant. Proc. 13,1759-1761. Lenschow, D. W.,Zeng, Y.,Thistlewaite, J. R., Montag, A., Brady, W., Gibson, M. G., Linsley, P. S., and Bluestone, J. A. (1992).Long term survival of xenogeneic pancreatic islets induced by CTLA-41g. Science 257,789-792. Lethe, B., van den Eynde, B., van Pel, A., Corradin, G., and Boon, T. (1992). Mouse tumor rejection antigens P815 A and B: Two epitopes carried by a single peptide. Eur. J. Immunol. 22,2283-2288. Levine, A. J., Momand, J., and Finlay, C. A. (1991). The p53 tumour-suppressor gene. Nature (London)351,453-456. Ley, V., Roth, C., Langlade-Demoyen, P., Larsson-Sciard,E.-L., and Kourilsky, P. (1990). A novel approach to the induction of specific cytolytic T cells in uiuo. Res. Immunol. 141,855-863. Ley, V., Langlade-Demoyen, P., Kourilsky, P., and Larsson-Sciard,E.-L. (1991).Interleukin 2-dependent activation of tumor-specific cytotoxic T lymphocytes in uiuo. Eur. J . Immunol. 21,851-854.

340

CLAUDE ROTH ET AL.

Li, W., Diamantstein, T., and Blankenstein, T. (1990).Lack of tumorigenicity of interleukin 4 autocrine growing cells seems related to the anti-tumor function of interleukin 4. Mol. Immunol. 27, 1331-1337. Li, Z., and Srivastava, P. K. (1993). Tumor rejection antigen gp96/grp94 is an ATPase: Implications for protein folding and antigen presentation. EMBO J. 12, 3143-3151. Liao, N.-S., Bix, M., Zijlstra, M., Jaenisch, R., and Raulet, D. (1991). MHC class I deficiency: Susceptibility to natural killer (NK)cells and imparied NK activity. Science 253,199-202. Lill, N. L., Tevethia, M. J., Hendrickson, W. G., and Tevethia, S. S. (1992). Cytotoxic T lymphocytes (CTL) against a transforming gene product select for transformed cells with point mutations within sequences encoding CTL recognition epitopes. J . E x p . Med. 176,449-457. Lindenman, J., and Klein, P. (1967).Viral oncolysis: Increased immunogenicity of host cell antigen associated with influenza virus. J. E x p . Med. 126,93-108. Linsley, P. S., and Ledbetter, D. H. (1993). The role of the CD28 receptor during T cell responses to antigen. Annu. Reu. Immunol. 11, 191-212. Linsley, P. S., Brady, W., Grosmaire, L., Aruffo, A., Damle, N., and Ledbetter, J. A. (1991).Binding of the B-cell activation antigen B7 to CD28 costimulates T cell proliferation and interleukin-2 mRNA accumulation. J . Exp. Med. 173, 721-730. Linsley, P. S., Greene, J. L., Tan, P., Bradshaw, J., Ledbetter, D. H., Anasetti, C., and Damle, N. K. (1992).Coexpression and functional cooperation of CTLA-4 and CD28 on activated T lymphocytes. J . E x p . Med. 176, 1595-1604. Liotta, L. A., Rao, C. N., and Wewer, U. M. (1986). Biochemical interactions of tumor cells with the basement membrane. Annu. Rev. Biochem. 55, 1037-1057. Liotta, L. A., Steeg, P. S., and Stetler-Stevenson, W. G. (1991a).Cancer metastasis and angiogenesis: An imbalance of positive and negative regulation. Cell (Cambridge, MUSS.)64,327-336. Liotta, L. A., Stetler-Stevenson, W. G., and Steeg, P. S. (1991b). Cancer invasion and metatasis: Positive and negative regulatory elements. Cancer lnuest. 9, 543-551. Litwin, V., Gumpen, J., Parham, P., Phillips, J. H., and Lanier, L. L. (1993). Specificity of HLA class I antigen recognition by human NKclones: Evidence for clonal heterogeneity, protection by self and non-self alleles, and influence of the target cell type. J. E x p . Med. 178, 1321-1326. Livingston, P. 0. (1991).Active specific immunotherapy in the treatment of cancer. In “Immunology and Allegy Clinics of America: Human Cancer Immunology 11” (H. F. Oettgen, ed.) Vol. 11,pp. 402-423. Saunders, London. Ljunggren, N., and Karre, K. (1990). In search of the “missing self”: MHC molecules and NK recognition. Immunol. Today 11,237-244. Lopez-Nevot, M. A., Esteban, F., Ferron, A,, Gutierrez, J., Oliva, M. R., Romero, C., Huelin Ruiz-Cabello, F., and Carrido, F. (1989). HLA class I gene expression on human primary tumours and autologous metastases: Demonstration of selective losses of HLA antigens on colorectal, gastric and laryngeal carcinomas. Br. J. Cancer 59, 221-222. Lu, S.-J., Day, N. E., Degos, L., Lepage, V., Wang, P.-C., Chan, S.-H., Simons, M., McNight, B., Easton, D., Zeng, Y.,and De ThB, G. (1990).Linkage of a nasopharyngeal carcinoma susceptibility locus to the HLA region. Nature (London)346,470-471. Lurquin, C. A., van Pel, A,, Mariame, B., DePlaen, E., Szikora, J. P., Janssens, C., Reddehase, I., Lejeune, J., and Boon, T. (1989). Structure of the gene coding for Tum- transplantation antigen P91A: A peptide encoded by the mutated exon is recognized with Ld cytolytic T cells. Cell (Cambridge, Mass.) 58,293-303.

IMMUNE RESPONSE AGAINST TUMORS

341

Lynch, R. G., Graft, R. K., Sirisinha, S., Simms, E. S., and Eisen, H. N. (1972).Myeloma proteins as tumor-specific transplantation antigens. Proc. Natl. Acad. Sci. U.S.A. 69, 1540-1544. Mach, J. P., Folli, S., Pelegrin, A., Wagnieres, G., Calmes, J. M., Braichotte, D., Buchegger, F., Hardman, N., Heusser, C. H., Givel, J. C., Chapuis, G., Chatelain, A,, and Van Den Bergh, H. (1993). In “Frontiers of Immunology and Cancer Immunology.” William B. Coley Symposium-New York. Mahadevan, V., and Hart, I. R. (1990). Metastasis and angiogenesis. Acta Oncol. 29, 97-103. Maki, R. G., Old, L. J., and Srivastana, P. K. (1990).Human homologue of murine tumor rejection antigen gp96: 5’ regulatory and coding regions and relationships to stressinduced proteins. Proc. Natl. Acad. Sci., U.S.A.87, 5658-5662. Malkin, D., Li, F. P., Strong, L. C. Fraumeni, J. F., et al. (1990).Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250, 1233-1238. Malkin, D., Jolly, K. W., Barbier, N., Look, T., Friend, S. H., Gebhardt, M. C., Andersen, T. I., Borresen, A. L., Li, F. P., Garber, J., and Strong, L. C. (1992).Germline mutations of the p53 tumour-suppressor gene in children and young adults with second malignant neoplasms. N . Engl. J . Med. 326,1309-1315. Marshall, C. J. (1991). Tumor suppressor genes. Cell (Cambridge, Mass) 64, 313-326. Maryanski, J. L., Pala, P., Corradin, G. Jordan, B. R., and Cerottini, J.-C. (1986). H-2restricted cytolytic T cells specific for HLA can recognize a synthetic HLA peptide. Nature (London)324,578-579. Maurer, J., Jannsen, J. W., Thiel, E., van Denderen, J., Ludwig, W. D., Aydemir, U., Heinze, B., Fonatsch, C., Harbott, J., and Reiter, A. (1991). Detection of chimeric bcr-able genes in actue lymphoblastic leukemia by the polymerase chain reaction. Lancet 337, 1055-1058. McBride, W. H., Thacker, J. D., Comora, S., Economou, J. S., Kelley, D., Hogge, D., Duhinett, S. M., and Dougherty, G. J. (1992). Genetic modification of a murine fibrosarcoma to produce interleukin 7 stimulates host cell infiltration and tumor immunity. Cancer Res. 52,3931-3937. Medawar, P. B., and Hunt, R.(1983).Can fetal antigens be used for prophylactic immunization? CIBA Found. Symp. 96,160-180. Melief, C. J. M. (1992). Tumor eradication by adoptive transfer of cytotoxic T lymphocytes. Adv. Cancer. Res. 58, 143-175. Melief, C. J. M., and Kast, W. M. (1990). Efficacy of cytotoxic T lymphocytes against virus-induced tumors. Cancer Cells 2, 116-120. Melief, C. J. M., Boog, C. J. P., Vasmel, W. L. E., Boes, J., Voordouw, A. C., and Kast, W. M. (1988). Dendritic cells and antigen presentation in the regulation of cytotoxic T lymphocyte responses against viruses and transplantation antigens. Ann. N.Y. Acad. Sci. 532,280-291. Melief, C. J. M., Vasmel, W. L. E., Offringa, R., Sijts, E. J. A. M., and Matthews, W. M. (1989). Immunosurveillance of virus-induced tumors. Cold Spring Harbor Symp. Quant. Biol. 54, 597-603. Meneguizzi, G., Kieny, M.-P., Lecocq, J.-P,, Chambon, P., Cuzin, F., and Lath6 R. (1990). Vaccinia recombinants expressing early bovine papilloma virus (BPV1) proteins: retardation of BPVl tumour development. Vaccine, 8, 199-204. Meneguzzi, G., Cerni, C., Kieny, M.-P., and LathB, R. (1991). Immunization against human papillomavirus type 16tumor cells with recombinant vaccinia viruses expressing E6 and E7. Virology 181,62-69.

342

CLAUDE ROTH ET AL.

Milich, D. R., McLachlan, A., Raney, A. K., Houghten, R., Thornton, G. B., Maruyama, T., Hughes, J. L., and Jones, J. E. (1991). Autoantibody production in hepatitis B antigen transgenic mice elicited with a self T-cell peptide and inhibited with nonself peptides. Proc. Natl. Acad. Sci. U.S.A.88,4348-4352. Miller, F. R. (1993). Immune mechanisms in the sequential steps of metastasis. Grit. Rev. Oncogen. 4,293-311. Mills, K. H. G. (1986).Processing of viral antigens and presentation to class 11-restricted T cells. lmmunol. Today 7,260-263. Mizoguchi, H., O’Shea, J., Longo, D. L., LoefAer, C. M., McVicar, D. W., and Ochoa, A. C. (1992). Alterations in signal transduction molecules in T lymphocytes from tumor-bearing mice. Science 258, 1795-1798. Moller, P., and Hammerling, G. (1992). The role of surface HLA-A,B,C molecules in tumor immunity. Cancer Sum. 13,101-128. Momburg, F., Ziegler, A,, Harpprecht, J., Moller, P., Moldenhauer, G., and Hammerling, G. (1989). Selective loss of HLA-A or HLB-B antigen expression in colon carcinoma. J. Immunol. 142,352-358. Monach, P. A., Schreiber, H., and Rowley, D. A. (1993). CD4+ and B lymphocytes in transplantation immunity. 11. Augmented rejection of tumor allograftsby mice lacking B cells. Transplantation 55, 1356-1361. Moore, K. W., Vieira, P., Fiorentino, D. F., Trounstine, M. L., Khan, T. A., and Mosmann, T. R. (1990). Homology of cytokine synthesis inhibitory factor (IL-10) to the EpsteinBarr virus gene BCRFI. Science 248,1230-1234. Moore, K. W., O’Garra, A., De Waal-Malefyt, R., Vieira, P., and Mosmann, T. R. (1993). Interleukin-10. Annu. Rev. lmmunol. 11, 165-190. Moore, M. W., Carbone, F. R., and Bevan, M. J. (1988). Induction of soluble protein into the class I pathway of antigen processing and presentation. Cell (Cambridge, MUSS.)54,777-785. Moore, S. K., Kizak, C., Robinson, E. A., Ullrich, S. J., and Appella, E. (1987). Cloning and nucleotide sequence of the murine hsp84 cDNA and chromosome assignment of related sequenes. Gene 56, 29-40. Morrison, L. A., Lukather, A. E., Braciale, V. L., Fan, D. P., and Braciale, T. J. (1986). Differences in antigen presentation to MHC class I- and class 11-restrictedinfluenza virus-specific cytolytic T lymphocyte clones. /. E r p . Med. 163, 903-921. Morrison, L. A,, Braciale, V. L., and Braciale, T. J. (1988). Antigen form influences induction and frequency of influenza-specific class I and class I1 MHC-restricted cytolytic T lymphocytes. /. lmmunol. 141,363-368. Mortarini, R., Anichini, A., and Parmiani, G. (1991). Heterogeneity for integrin expression and cytokine mediated VLA modulation can influence the adhesion of human melanoma cells to extracellular matrix proteins. Int. J. Cancer 47, 551-559. Morton, D. L., Miller, G. F., and Wood, D. A. (1969). Demonstration of tumor-specific immunity against antigens unrelated to the mammary tumor virus in spontaneous mammary adenocarcinomas. J . Natl. Cancer Inst. 42,289-301. Mosmann, R. T., and Cotban, R. L. (1989).Heterogeneity of cytokine secretion patterns and functions of helper T cells. Adu. lmmunol. 46, 111-147. Mukai, M., Shinkai, R., Tateishi, R., Mori, Y.,and Akedo, H. (1987).Macrophagepotentiation of invasive capacity of rat ascites hepatoma cells. Cancer Aes. 47,2167-2171. Muller, J . , Khan, A. S., Rao, V., Hearing, V. J., Kim, S., and Gorelik, E. (1994). Loss of A- and C-type retrovirus particles induced by H-2Kb gene expression in B16 BL6 melanoma cells. Submitted for publication. Murray, R., Kurilla, M., Brooks, J., Thoams, W., Rowe, M., Kieff, E., and Rickinson, A.

IMMUNE RESPONSE AGAINST TUMORS

343

(1992). Identification of target antigens for the human cytotoxic T cell response to Epstein-Barr virus (EBV):Implications for the immune control of EBV-positivemalignancies. J . Exp. Med. 176, 157-168. Nabavi, N., Freeman, G. J., Gault, A., Godfrey, D., Nadler, L. M., and Glimcher, L. H. (1992). Signalling through the MHC class I1 cytoplasmic domain is required for antigen presentation and induces B7 expression. Nature (London) 360, 266268. Nakagoni, H., Peterson, M., Magnusson, I., Juhlin, C., Matsuda, M., Mellstedt, H., Taupin, J.-L., Vivier, E., Anderson, P., and Kiessling, R. (1993).Decreased expression of the signal-transducing 5 chains in tumor-infiltrating T-cells and NK cells of patients with colorectal carcinoma. Cancer Res. 53,5610-5612. Natali, P., Nicotra, M., Bigotti, A., Venturo, I., Marcenaro, L., Gracomini, P., and Runo, C. (1989). Selective changes in expression of HLA class I polymorphic determinants in human solid tumors. Proc. Natl. Acad. Sci. U.S.A.86,6719-6723. Nonacs, R., Humborg, C. N., Tam, J. P., and Steinman, R. M. (1992). Mechanisms of mouse spleen dendritic cell function in the generation of influenza-specific cytolytic T lymphocytes. J . Exp. Med. 176,519-529. Oh, S. K., Very, D. L., Walker, J., Raam, S., and Ju, S. T. (1987).An analogy between fetal haptoglobin and a potent immunosuppressant in cancer. Cancer Res. 47,5120-5126. Ostrand-Rosenberg, S., Thakur, A,, and Clements, V. (1990).Rejection of mouse sarcoma cells after transfection of MHC class I1 genes. J . Imunol. 144,4068-4071. Ottesen, S. S., Kieler, J., and Christensen, B. (1987).Changes in HLA-A,B,C expression during spontanous transformation of human urothelial cells in oitro. Eur. J . Clin. Oncol. 23,991-995. Pace, J. L., Russel, S. W., Torres, B. A., Johnson, H. M., and Gray, P. W. (1983).Recombinant mouse y-interferon induces the priming step in macrophage activation for tumor cell killing. J . Immunol. 130, 2011-2013. Pandolfi, F., Boyle, L. A., Trentin, L., Kurnick, J. T., Isselbacher, K. J., and GattoniCelli, S. (1991). Expression of HLA-A2 antigen in human melanoma cell lines and its role in T-cell recognition. Cancer Res. 51,3164-3170. Pantel, K., Schlimok, G., Kutter, D., Schaller, G., Genz, T., Wiebecke, B., Backmann, R., Funke, I., and Riethmuller, G. (1991).Frequent down-regulation ofmajor histocompatibility class I antigen expression on individual micrometastatic carcinoma cells. Cancer Res. 51,4712-4715. Parnes, J. R. (1989). Molecular biology and function of CD4 and CD8. Ado. Immunol. 44,265-3 11. Peace, D. J., Chen, W., Nelson, H., and Cheever, M. A. (1991). T cell recognition of transforming proteins encoded by mutated ras proto-oncogenes. J. Immunol. 146, 2059-2065. Peace, D. J,, Smith, J,, Disis, N., Cosand, W., Blake, J., You, S. G., Chen, W., and Cheever, M. A. (1992). Class I MHC-restricted T cell recognition of the mutated segment of activated pel-ras protein. FASEB J . 6,2710. Philipps, C., McMillan, M., Flood, P. M., Murphy, D. B., Forman, J., Lancki, D., Womack, J. E., Goodenow, R. S., and Schreiber, H. (1985). Identification of a unique tumor-specific antigen as a novel class I major histocompatibility molecule. Proc. Natl. Acad. Sci. U.S.A.82,5140-5144. Pinhasi-Kimhi, O., Michalovitz, D., Ben-Zeev, A., and Oren, M. (1986).Specific interaction between the p53 cellular tumor antigen and major heat shock proteins. Nature (London)320,182-184. Plaksin, D., Gelber, C., Feldman, M., and Eisenbach, L. (1988). Reversal of the meta-

344

CLAUDE ROTH ET AL.

static phenotype in Lewis lung carcinoma cells after transfection with syngeneic H2Kb gene. Proc. Natl. Acad. Sci. U S A . 85,4463-4467. Plata, F., Langlade-Demoyen, P., Abastado, J.-P., Berbar, T., and Kourilsky, P. (1986). Molecular definition of retrovirus-induced antigens recognized by tumor-specific H2 restricted cytolytic T lymphocytes. J . Zmmunogenet. 13,263-268. Plata, F., Langlade-Demoyen, P., Abastado, J.-P., Berbar, T., and Kourilsky, P. (1987). Retrovirus antigens recognized by cytolytic T lymphocytes activate tumor rejection in oioo. Cell (Cambridge, Mass.) 48, 231-240. Plautz, G. E., Yang, Z.-Y., Wu, B.-Y., Gao, X., Huang, L., and Nabel, G. J. (1993). Immunotherapy of malignancy by in oioo gene transfer into tumors. Proc. Natl. Acad. Sci. U.S.A.90,4645-4649. Porgador, A., Feldman, M., and Eisenbach, L. (1989). H-2Kb transfection of B16 melanoma cells results in reduced tumourigenicity and metastatic competence.]. Immunogenet. 16,291-303. Powis, S . J., Townsend, A. R. M., Deverson, E. V., Bastin, J., Butcher, G. W., and Howard, J. C. (1991).Restoration of antigen presentation of the mutant cell line RMAS by an MHC-linked transporter. Nature (London)354,528-531. Powis, S., Deverson, E., Coadwell, W., Ciruela, A., Huskisson, N., Smith, H., Butcher, G., and Howard, J. (1992). Effect of polymorphism of an MHC-linked transporter on the peptides assembled in a class I molecule. Nature (London)357,211-215. Pravtcheva, D. D., De Leo, A. B., Ruddle, F. H., and Old, L. J. (1981). Chromosome assignment of the tumor-specific antigen of a 3-methylchloranthrene-induced mouse sarcoma. J . E x p . Med. 154, 964-977. Prehn, R. T. (1975). Relationship of tumor immunogenicity to concentration of the oncogen. J . Natl. Cancer Inst. 55,189-190. Prehn, R. T., and Karcher, C. A. (1983). Immunodepression by the oncogen does not determine the relationship between tumor latency and immunogenicity. Int.].Cancer 31,227-229. Prehn, R.T., and Main, J. M. (1957).Immunity of methylcholanthene induced sarcomas. J . Natl. Cancer Inst. 18, 69-78. Prehn, R. T., and Prehn, L. M. (1987). The autoimmune nature of cancer. Cancer Res. 47,927-932. Ramquist, T., Pallas, D. O., Deanda, J., Ahrlund-Richter, L., Reinholdsson, G., Roberts, T. M., Schaflhausen, B. S., and Dalianis, T. (1988).Immunization against the polyoma tumor-specific transplantation antigen (TSTA with polyoma T-antigens). Int.J . Cancer 42,123-128. Real, F. X., Mattes, M. J., Houghton, A. N., Oettgen, H. F., Lloyd, K. O., and Old, L. J. (1984). Class I (unique) tumor antigens of human melanoma: Identification of a 90,000 dalton cell surface glycoprotein by autologous antibody. J. E r p . Med. 160, 1219-1233. Reiser, H., Freeman, G. J., Razi-Wolf, Z., Gimmi, C. D., Benaceraf, B., and Nadler, L. M. (1992). Murine B7 antigen provides an efficient costimulatory signal for activation of murine T lymphocytes via the T-cell receptodCD3 complex. Proc. Natl. Acad. S C ~ U. . S A . 89,271-275. Restifo, N. P., Spiess, P. J., Karp, S. E,, MulB, J. J., and Rosenberg, S. A. (1992). A nonimmunogenic sarcoma transduced with the cDNA for interferon gamma elicits CD8+ T cells against the wild-type tumor: Correlation with antigen presentation capability. J . Exp. Med. 175, 1423-1431. Restifo, N., Esquivel, F., Kawakami, Y., Yewdell, J., Mule, J., Rosenberg, S., and Bennink, J. (1993). Identification of human cancers deficient in antigen processing. J . E x p . Med. 177,265-272.

IMMUNE RESPONSE AGAINST TUMORS

345

Ritter, G., Boosfeld, E., Adluri, R., Calves, M., Oettgen, H. F., Old, L. J., and Livingston, P. 0. (1991). Antibody response to immunization with ganglioside GD3 and GD3 congeners (lactones, amide and gangliosidol) in patients with malignant melanoma. lnt. J. Cancer, 48,379-385. Roberts, A. B., Spom, M. B., Assoian, R. K., Smith, J. M., Roche, N. S., Wakefield, L. M., Heine, U. I., Liotta, L. A., Falanga, V., Kehrl, J. H., and Fauci, A. S. (1986). Transforming growth factor type p: Rapid induction of fibrosis and angiogenesis in oioo and stimulation of collagen formation in oitro. Proc. Natl. Acad. Sci. U S A . 83, 4167-4171. Roche, P. A., and Cresswell, P. (1990). Invariant chain association with HLA-DR molecules inhibits immunogenic peptide binding. Nature (London)345, 615-617. Rochlitz, C. F., Scott, G. K., Dodson, J. M., Liu, E., Dollbaum, C., Smith, H. S., and Benz, C. C. (1989). Incidence of activating ras oncogene mutations associated with primary and metastatic human breast cancer. Cancer Res. 49,357-360. Rock, K. L., Benacerraf, B., and Abbas, A. K. (1984). Antigen presentation by haptenspecific B lymphocytes. J . E x p . Med. 160, 1102-1113. Roman, J. M., Hirsch, J., Readhead, C., Levy, D., De Ogny, L., and Dreyer, W. J. (1981). Heritable differences among gp70-like molecules on C3H ultraviolet light-induced fibrosarcomas. Transplant. Proc. 13, 1782-1786. Rose, T. M., Plowman, G. D., Teplow, D. B., Dreyer, W. J., Hellstrom, K. E., and Brown, J. P. (1986).Primary structure of the human melanoma-associated antigen p97 (melanoma transferrin) deduced from the mRNA sequence. Proc. Natl. Acad. Sci. U.SA. 83,1261-1265. Roth, C., Mir, L. M., Cressent, M., Quintin-Colonna, F., Ley, V., Fradelizi, D., and Kourilsky, P. (1992).Inhibition of tumor growth by histoincompatible cells expressing interleukin-2. Int. Immunol. 4, 1429-1436. Roth, C., Delassus, S., Even, J., and Kourilsky, P. (1994). Inhibition of tumor growth by histoincompatible cells expressing IL-2. In “Cytokine Induced Tumour Immunogenicity” (R. Fao and G. Forni, eds.) Academic Press, in press. Rotzschke, O., Falk, F., Deres, K., Schild, H., Norda, M., Metzger, J., Jung, G., and Rammensee, H.-G. (1990).Isolation and analysis of naturally processed viral peptides as recognized by cytotoxic T cells. Nature (London)348,252-254. Rousset, F., Garcia, E., Defrance, T., Peronne, C., Vezzio, N., Hsu, D. H., Kastelein, R., Moore, K., and Banchereau, J. (1992). Interleukin 10 is a potent growth and differentiation factor for activated human B lymphocytes. Proc. Natl. Acad. Sci. U S A . 89,1890-1893. Rudensky, A,, Preston, H. P., Hong, S. C., Barlow, A.,and Janeway, C. J. (1991).Sequence analysis of peptides bound to MHC class I1 molecules. Nature (London)353,622-627. Russell, S. J., Eccles, S. A., Flemming, C. L., Johnson, C. A., and Collins, M. K. L. (1991). Decreased tumorigenicity of a transplantable rat sarcoma following transfer and expression of an IL-2 cDNA. lnt. J . Cancer 47,244-251. Santibanez-Koref, M., Birch, J. M., Hartley, A. L., Morris Jones, P. H., Craft, A. W., Eden, T., Crowther, D., Kelsey, A. M., and Harris, M. (1991).p53 germline mutations in Li-Fraumeni syndrome. Lancet 338, 1490-1491. Sawai, E. T., and Butel, J. S. (1989). Association of a cellular heat shock protein with simian virus 40 large T antigen in transformed cells. J. Virol. 63,3961-3973. Schechter, A., Hung, M., and Coussens, L. (1985).The new gene: An erbB-homologous gene distinct from and unlinked to the gene encoding the EGF receptor. Science 229,976-978. Schendel, D. J., Gansbacher, B., Oberneder, R., and Segurado, 0. G. (1993). HLA-A2 restricted lysis of human renal cell carcinoma by tumor-infiltrating lymphocytes.

346

CLAUDE ROTH ET AL.

Abstract No. 2218. Keystone Symposium on “Cellular Immunity and Immunotherapy of Cancer,” Taos, NM, March 1993. Schreiber, H., Ward, P. L., Rowley, D. A., and Straws, H. (1988). Unique tumor-specific antigens. Annu. Rev. Immunol. 6, 465-483. Schrier, P. I., and Peltenburg, L. T. C. (1993). Relationship between myc oncogene activation and MHC class I expression. Adv. Cancer Res. 60, 181-246. Schultz, M., Zinkernagel, R. M., and Hentgartner, H. (1991). Peptide-induced antiviral protection by cytotoxic T cells. Proc. Natl. Acad. Sci. U.S.A. 88,991-993. Schumacher, T. N. M., Heemels, M. T., Neefjes, J. J., Kast, W. M., Melief, C. J. M., and Ploegh, H. L. (1990). Direct binding of peptide to empty MHC class I molecules in intact cells and in uitro. Cell (Cambridge, Mass.) 62,563-567. Schumacher, T. N. M., De Bruijn, M. L. H., Vernie, L. N., Kast, W. M., Melief, C. J. M., Neefjies, J. J., and Ploegh, H. L. (1991). Peptide selection by MHC class I molecules. Nature (London)350,703-706. Schwartz, R. H. (1990). A cell culture model for T lymphocyte clonal anergy. Science 248,1349-1356. Schwartz, R. H. (1992). Costimulation of T lymphocytes: the role of CD28, CTLA-4, and B7/BB1in interleukin-2 production and immunotherapy. Cell (Cambridge,Mass.) 71, 1065-1068. Sercarz, E. E., Lehmann, P. V., Ametani, A., Benichou, G., Miller, A., and Moudgil, K. (1993). Dominance and crypticity of T cell antigenic determinants. Adv. Rev. Immun01. 11, 729-766. Shahinian, A,, Pfeffer, K., Lee, K. P., Kiindig, T. M., Kishihara, K., Wakeham, A., Kawai, K., Ohashi, P. S., Thompson, C. B., and Mak, T. W. (1993).Differential costimulatory requirements in CD28-deficient mice. Science 261,609-612. Shastri, N., and Gonzalez, F. (1993). Endogenous generation and presentation of the ovalbumin peptide/Kb complex to T cells. J . Immunol. 150,2724-2736. Shiku, H., Takahashi, T., Oettgen, H. F., and Old, L. J. (1976). Cell surface antigens of human malignant melanoma. 11. Serological typing with immune adherence assays and definition of two new surface antigens. J . Erp. Med. 144,873-881. Shtivelman, E. B., Lifshitz, R. P., Gale, R. P., and Canaani, E. (1985). Fused transcript of abl and bcr genes in chronic myelogenous leukemia. Nature (London)315,550-554. Sibille, C., Chomez, P., Wildmann, C., van Pel, A., DePlaen, E., Maryanski, J. L., De Bergeyhck, V., and Boon, T. (1990). Structure of the gene of Tum- transplantation antigen P198: A point mutation generates a new antigenic peptide.J. Erp. Med. 172, 35-45. Siepl, C., Bodmer, S., Frei, K., MacDonald, H. R., De-Martin, R., Hofer, E., and Fontana, A. (1988). The glioblastoma-derived T cell suppressor factodtransforming growth factor-beta 2 inhibits T cell growth without affecting the interaction of interleukin 2 with its receptor. Eur. J . Immunol. 18,593-600. Singer, A., Muitz, T. I., Goldin, H., Rosenberg, A. S.,and Mizvochi, T. (1987).Recognition requirements for the activation, differentiation and function of T-helper cells specific for class I MHC alloantigens. lmmunol. Rev. 98, 143-170. Sjogren, H. 0. (1964). Studies on specific transplantation resistance to polyoma-virusinduced tumors. IV. Stability of the polyoma cell antigen. J. Natl. Cancer Inst. 32, 661-666. Sjogren, H. O., Hellstrom, I., and Klein, G. (1961). Transplantation of polyoma virus induced tumor in mice. Cancer Res. 21,329-337. Slamon, D., Clark, G., Wong, S., Levin, W., Ullrich, A., and McGuire, W. (1987). Human breast cancer: Correlation of relapse and survival with amplification of the HER-21 neu oncogene. Science 235,177-182.

IMMUNE RESPONSE AGAINST TUMORS

347

Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. F., Levin, W. J., Stuart, S. G., Udove, J., Ullrich, A., and Press, M. F. (1989). Studies of the HER-2heu protooncogene in human breast and ovarian cancer. Science 244, 707-712. Slingluff C. L., Jr., Cox, A. L., Henderson, R. A,, Hunt, D. F., and Engelhard, V. H. (1993). Recognition of human melanoma cells by HLA-A2.1 restricted cytotoxic T lymphocytes is mediated by at least six shared peptide epitopes. J. Immunol. 150, 2955-2963. Smith, M. E. F., Marsh, S. G. E., Bodmer, J. G., Gelsthorpe, K., and Sodmer, W. F. (1989). Loss of HLA-A,B,C allele products and lymphocyte function-associated antigen 3 in colorectal neoplasia. Proc. Natl. Acad. Sci. U S A . 86,5557-5561. Spies, T., Cerundolo, V., Colonna, M., Cresswell, P., Townsend, A., and Demars, R. (1992). Presentation of viral antigen by MHC class-I molecules is dependent on a putative peptide transporter heterodimer. Nature (London) 355,644. Sprent, J., and Webb, S. R. (1987). Function and specificity of T cell subsets in the mouse. Adv. Immunol. 41,39-133. Srivasta, S . , Zou, Z. Q., Pirollo, K.,Blattner, W. A., and Chang, E. H. (1990). Germline transmission of a mutated p53 gene in a cancer-prone family with Li-Fraumeni syndrome. Nature (London) 348,747-749. Srivastava, P. K., and Heike, M. (1991). Tumor-specific immunogenicity of stressinduced proteins: Convergence of two evolutionary pathways of antigen presentation? Semin. Immunol. 3,57-64. Srivastava, P. K.,and Maki, R. G. (1991). Stress-induced proteins in immune response to cancer. C u m Topics Microbiol. Immunol. 167, 109-123. Srivastava, P. K., and Old, L. J. (1988). Individually distinct transplantation antigen of chemically induced mouse tumors. Imunol. Today 9,78-83. Srivastava, P. K., De Leo A. B., and Old, L. J. (1986). Tumor rejection antigens of chemically induced tumors of inbred mice. Proc. Natl. Acad. Sci. U.S.A. 83, 3407-34 11. Srivastava, P. K., Chen, Y. T., and Old, L. J. (1987). 5’-structural analysis of genes encoding polymorphic antigens of chemically induced tumors. Proc. Natl. Acad. Sci. U.S.A. 84,3807-3811. Stauss, H. J., Linsk, R., Fischer, A,, Watts, S., Banasiak, D., Haberman, A,, Clark, I., Forman, J., McMillan, M., Schreiber, H., and Goodenow, R. S. (1986a). Isolation of the MHC genes encoding the tumor-specific class I antigens expressed on a murine fibrosarcoma. J. Immunogenet. 13,101-111. Stauss, H. J., van Waes, C., Fink, M. A., Starr, B., and Schreiber, H. (198613).Identification of a unique tumor antigen as rejection antigen by molecular cloninc and gene transfer. J . Erp. Med. 164,1516-1530. Steinman, R. M. (1991).The dendritic cell system and its role in immunogenicity. Annu. Rev. Immunol. 9,271-296. Stewart, T. H., Hollinshead, A. C., and Raman, S. (1991). Tumor dormancy: Initiation, maintenance and termination in animals and humans. Can. J. Surg. 34,321-325. Storkus, W. J., Zeh 111, H. J., Maeurer, M. J., Salter, S. D., and Lotze, M. R. (1993). Isolation of human melanoma peptides recognized by class I restricted tumor infiltrating T lymphocytes. 1.Immunol. 151,3719-3727. Sulitzeanu, D. (1993). Immunosuppressive factors in human cancer. Ado. Cancer Res. 60,247-267. Szikora, J. P., Van Pel, A., Brichard, V., Andre, M., Vanbaren, N., Henry, P., and Boon, T. (1990). Structure of the gene of Tum- transplantation antigen p35B: Presence of a point mutation in the antigenic allele. E M B O J . 9, 1041-1050.

348

CLAUDE ROTH ET AL.

Takeichi, M. (1991). Cadherin cell adhesion receptors as a morphogenetic regulator. Science 251,1451-1455. Tanaka, K., Isselbacher, K., Khouvy, G., and Jay, G. (1985).Reversal of oncogenesis by the expression of a major histocompatibility complex gene. Science 228,26-30. Tanaka, Y., and Tevethia, S. S. (1988). Zn uitro selection of SV40 T antigen epitope loss variants by site-specific cytotoxic T lymphocyte clones. 1.Zmmunol. 140,4348-4354. Tanaka, Y., and Tevethia, S. S. (1990). Loss of immunorecessive cytotoxic T lymphocyte determinant V on SV40 T antigen following cocultivation with site specific cytotoxic T lymphocyte clone Y-5 Znteruirology 31, 197-202. Tanaka, Y., Tevethia, M. J., Kalderon, D., Smith, A. E., and Tevethia, S. S. (1988). Clustering of antigenic sites recognized by cytotoxic T lymphocyte clones in the amino terminal half of SV40 T antigen. Virology 162,427-436. Tepper, R. I., Pattengale, P. K., and Leder, P. (1989). Murine interleukin-4 displays potent anti-tumor activity in uiuo. Cell (Cambridge, Muss.)57,503-512. Tepper, R. I., Coffman, R. L., and Leder, P. (1992).An eosinophil-dependent mechanism for the antitumor effect of interleukin-4. Science 257, 548. Tevethia, S. S., Lewis, M., Tanaka, Y.,Milici, J., Knowles, B., Maloy, W. L., and Anderson, R. (1990). Dissection on H-2Db-restricted cytotoxic T-lymphocyte epitopes on Simian virus 40 T antigen by the use of synthetic peptidec and H-2Db" mutants. /. V i d . 64, 1192-1200. Teyton, L., OSullivan, D., Dickson, P. W., Lotteau, V., Sette, A., Fink, P., and Peterson, P. (1990). Invariant chain distinguishes between the exogenous and endogenous antigen presentation pathways. Nature (London)348,39-44. Toguchida, J., Yamaguchi, T., Dayton, S. H., Beauchamp, R. L., Herrera, G. E., Ishizaki, K., Yamamuro, T., Meyers, P. A., Little, J. B., Sasaki, M. S., Weichselbaum, R. R., and Yandell, D. W. (1992). Prevalence and spectrum of germline mutations of the p53 gene among patients with sarcoma. N . Engl. J . Med. 326, 1301-1308. Tomita, Y., Matsumoto, Y.,Nishiyama, T., and Fujiwara, M. (1990). Reduction of major histocompatibility complex class I antigens on invasive and high grade transitional cell carcinoma. /. Pathol. 162, 157-164. Topalian, S . L., Solomon, D., and Rosenberg, S. A. (1989).Tumor-specific cytolysis by lymphocytes infiltrating human melanomas. J . Zmmunol. 142,3714. Topalian, S . L., Kasid, A., and Rosenberg, S. A. (1990). Immunoselection of a human melanom resistant to specific lysis by autologoustumor-infiltratinglymphocytes: Possible mechanisms for immunotherapeutic failures. J . lmmunol. 144,4487. Torigoe, T., Sato, N., Takashima, T., Joong-Moon,C., Tsuboi, N., Qi, W., Hara, I., Wada, Y., Takahashi, N., and Kikuchi, K. (1991).Tumor rejection antigens on BALB3T3 cells transformed by activated oncogenes. J . lmmunol. 147,3251-3258. Torre-Amione, G., Beauchamp, R. D., Koeppen, H., Park, B. H., Schreiber, H., Moses, H. L., and Rowley, D. A. (1990). A highly immunogenic tumor transfected with a murine transforming growth factor type beta 1cDNA escapes immune surveillance. Proc. Natl. Acad. Sci. U.S.A.87, 1486-1490. Townsend, S. E., and Allison, J. P. (1993). Tumor rejection after direct costimulation of CD8+ T cells by B7-transfected melanoma cells. Science 259,368-370. Townsend, A. R., Gotch, F. M., and Davey, J. (1985). Cytotoxic T cells recognize fragments of the immune influenza nucleoprotein. Cell (London)42,457-467. Townsend, A., Rothbard, J., Gotsch, F., Bahadur, G., Wraith, D., and McMichel, J. (1986a).Epitopes of influenza nucleoprotein recognized by CTLs can be defined with short synthetic peptides. Cell (Cambridge, Mass.) 44,959-968. Townsend, A. R., Bastin, J., Could, K., and Brownlee, G . G. (1986b).CytotoxicT lympho-

IMMUNE RESPONSE AGAINST TUMORS

349

cytes recognize influenza hemagglutinin that lacks a signal sequence. Nature (London) 324,575. Townsend, A., Ohlen, C., Bastin, J., Lundgren, H. G., Foster, L., and Ktirre, K. (1989). Association of class I MHC heavy and light chains induced by viral peptides. Nature (London)340,443-448. Traversari, C., Van Der Brugen, P., Luesher, I. F., Lurquin, C., Chomez, P., Van Pel, A., DePlaen, E., Amar-Costersec, A., and Boon, T. (1992). A nonapeptide encoded by human gene MAGE-I is recognized on HLA-A1 by cytolytic T lymphocytes directed against tumor antigen MZ3-3. J. E x p . Med. 176, 1453-1457. Trojan, J., Johnson, T. R., Rudin, S. D., Ilan, J., and Tykocinski, M. L. (1993). Treatment and prevention of rat glioblastoma by immunogenic C6 cells expessing antisense insulin-like growth factor. Science 259, 94-96. Turner, G. A., and Weiss, L. (1980). Some effects of products from necrotic regions of tumours on the in vitro migration of cancer and peritoneal exudate cells. 1nt.J.Cancer 26,247-257. Ullrich, S. E., Robinson, E. A., Law, L. W., Willingham, W., and Appella, E. (1986). A mouse tumor-specific transplantation antigen is a heat shock related protein. Proc. Natl. Acad. Sci. U.S.A. 83, 3121-3125. Unanue, E. R., and Allen, P. M. (1987). The basis for the immunoregulatory role of macrophages and other accessory cells. Science 236,551-557. Urban, J. L., Van Waes, C., and Schreiber, H. (1984). Pecking order among tumorspecific antigens. Eur. J . Immunol. 14,181-187. Uyttenhove, C., Van Snick, J., and Boon, T. (1980). Immunogenic variants obtained by mutagenesis of mouse mastocytoma P815. I. Rejection by syngeneic mice. J. E x p . Med. 152,1175-1183. Uyttenhove, C., Maryanski, J., and Boon, T. (1983). Escape of mouse mastocytoma P815 after nearly complete rejection is due to antigen loss variants rather than immunosuppression. J . E x p . Med. 157, 1040-1052. Van Bleek, G., and Nathenson, S. G. (1990).Isolation of an endogenously processed immunodominant viral peptide from the class I H-2Kb molecule. Nature (London) 348,213-216. van den Eynde, B., Hainaut, P., Herin, M., Knuth, A., Lemoine, C., Weynants, P., van der Bruggen, P., Fauchet, R., and Boon, T. (1989). Presence of a human melanoma of multiple antigens recognized by autologous CTL. Int. J . Cancer 44, 634640. van den Eynde, B., Lethe, B., van Pel, A., DePlaen, E., and Boon, T. (1991). The gene coding for a major tumor rejection antigen of tumor P815 is identical to the normal gene of syngenic DBAl2 mice. I. Exp. Med. 173,1373-1384. van Der Bruggen, P., and van den Eynde, B. (1992). Molecular definition of tumor antigens recognized by T lymphocytes. Curr. Opinion Immunol. 4, 608-612. van Der Bruggen, P., Traversari, C., Chomez, P., Lurquin, C., DePlaen, E., Van Den 'Eynde, B., Knuth, A,, and Boon, T. (1991). A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma. Science 254,1643-1647. van Pel, A., and Boon, T. (1982). Protection against a non-immunogenic mouse leukemia by an immunogenic variant obtained by mutagenesis. Proc. Natl. Acad. Sci. U.S.A. 79,4718-4722. van Pel, A,, Georlette, K., and Boon, T. (1979).Tumor cell variants obtained bymutagenesis of a Lewis lung carcinoma cell line: Immune rejection by syngeneic mice. Proc. Natl. Acud. Sci. U.S.A. 76,5282-5285. van Pel, A., Vessiere, F., and Boon, T. (1983). Protection against two spontaneous mouse

350

CLAUDE ROTH ET AL.

leukemias conferred by immunogenic variants obtained by mutagenesis.J. Erp. Med. 157,1192-2001. Van Waes, C., Urban, J. L., Rothstein, J. L., Ward, P. L., and Schreiber, H. (1986). Highly malignant tumor variants retain tumor-specific antigens recognized by helper T cells. J . Erp. Med. 164, 1547-1565. Vieira, P., Waal-Malefyt, R. D., Dang, M.-N., Johnson, K., Kastelein, R., Fiorentino, D., Vries, J. D., Roncarolo, M.-G., Mosmann, T., and Moore, K. (1991). Isolation and expression of human cytokine synthesis inhibitory factor cDNA clones; Homology to Epstein-Barr virus open reading frame BCRFI. Proc. Natl. Acad. Sci. U.S.A. 88, 1172-1 176. Vijayasaradhi, S., and Houghton, A. N. (1991). Purification of an autoantigenic 75 KDa human melanosomal glycoprotein. Int. J. Cancer 47,298-303. Viret, C., Davodeau, F., Guilloux, Y.,Bignon, J. D., Semana, G., Breathnach, R., and Jotereau, R. (1993). Recognition of shared melanoma antigen by HLA-A2-restricted cytolytic T cell clones derived from human tumor-infiltrating lymphocytes. Eur. J . Immunol. 23,141-146. Wallack, M. K., and Sivanadham, M. (1993). Clinical trials with VMO for melanoma. Ann. N . Y. Acad. Sci. 690,178-189. Wallich, R.,Bulbuc, N., Hammerling, G., Katzav, S., Segal, S., and Feldman, M. (1985). Abrogation of metastatic properties of tumor cells by de novo expression of H-2K antigens following H-2 gene transfection. Nature (London)315,301-305. Wang, P., Vegh, Z., Vanki, F., and Klein, E. (1992). HLA-B5-restricted auto-tumorspecific cytotoxic T cells generated in mixed lymphocyte-tumour-cell culture. Znt. J. Cancer 52,517-522. Wang, Z., Cao, Y.,Albino, A., Zeff, R., Houghton, A., and Ferrone, S. (1993). Lack of HLA class I antigen expression by melanoma cells SK-MEL-33caused by a reading frameshift in p2-microglobulin mRNA. J. Clin. Inoest. 91,684-692. Wank, R., and Thomssen, C. (1991). High risk of squamous cell carcinoma of the cervix for women with HLA-DQw3. Nature (London)352,723-725. Ward, P. L., Koeppen, H., Hurteau, T., and Schreiber, H. (1989).Tumor antigens defined by cloned immunological probes are highly polymorphic and are not detected on autologous normal cells. J . Exp. Med. 170,217. Watanabe, Y., Kuribayashi, K., Miyatake, S.,Nishihara, K., Nakayama, E.-I., Taniyama, T., and Sakata, T. A. (1989).Exogenous expression of mouse interferon gamma cDNA in mouse neuroblastoma C1300 cells results in reduced tumorigenicity by augmented anti-tumor immunity. Proc. Natl. Acad. Sci. U S A . 86,9456-9460. Watkins, J. F., and Chen, L. (1969). Immunization of mice against Ehrlich ascites tumor using a HamsterlEhrlich ascites tumor hybrid cell line. Nature (London)223, 1018-1022. Weigert, D. A., Stanton, G. J., and Johnson, H. M. (1983).Recombinant gamma interferon enhances natural killer cell activity similar to natural gamma interferon. Biochem. Biophys. Res. Commun. 111,525-529. White, E., Spector, D., and Welch, W. J. (1988). Differential distribution of the adenovirus ElA proteins and colocalization of ElA with the 7Gkilodalton heat shock protein in infected cells. J . Virol. 62,4153-4166. Wolfel, T., Klehmann, E., Muller, C., Schutt, K. H., Meyer Zum Buschenfelde, K. H., and Knuth, A. (1989). Lysis of human melanoma cells by autologous cytolytic T cell clones: Identification of human histocompatibility leukocyte antigen A2 as a restrictrion element for three different antigens. J. Erp. Med. 170, 797-810. Wolfel, T.,Hauer, M., Klehmann, E., Brichard, V., Ackermann, B., Knuth, A., Boon, T.,

IMMUNE RESPONSE AGAINST TUMORS

351

and Meyer Zum Buschenfelde, K. H. (1993). Analysis of antigens recognized on human melanoma cells by A2-restricted cytolytic T lymphocytes (CTL). Int.J.Cancer 55,237-244. Woodbury, R. C., Brown, J. P. Yeh, M. Hellstrom, I., and Hellstrom, K. E., (1980). Identification of a cell surface protein p97 in human melanoma and certain other neoplasms. Proc. Natl. Acad. Sci. U.S.A. 77,2183-2187. Wortzel, R. D., Philipps, C., and Schreiber, H. (1983). Multiplicity of unique tumorspecific antigens expressed on a single malignant cell. Nature (London)304,165-167. Wortzel, R. D., Urban, J. L., and Schreiber, H. (1984). Malignant growth in the normal host after variant selection in oitro with cytolytic T cell lines. Proc. Natl. Acad. Sci. U.S.A. 81,2186-2190. Wreschner, D. H., Hareuveni, M., Tsarfaty, I., Smorodinsky, N., Horev, J., Zaretsky, J., Kotkes, P., Weiss, M., Lathe, R., Dion, A,, and Keydar, I. (1990). Human epithelial tumor antigen cDNA sequences: Differential splicing may generate multiple protein forms. Eur. J . Biochem. 189,463-473. Yamaguchi, H., Moriuchi, T., Hosokawa, M., and Kobayashi, H. (1982). Increased or decreased immunogenicity of tumor-associated antigen according to the amount of virus-associated antigen in rat tumor cells infected with Friend virus. Cancer Immunol. Immunother. 12,119-123. Yanuck, M., Carbone, D. P., Pendleton, C. D., Tsukui, T., Winter, S. F., Minna, J. D., and Berzofsky, J. A. (1993). A mutant p53 tumor suppressor protein is a target for peptide-induced CD8 + cytotoxic T-cells. Cancer Res. 53, 3257-3261. Yasumura, S., Hirabayashi, H., Schwartz, D. R., Toso, J. F., Johnson, J. T., Herberman, R. B., and Whisteside, T. L. (1993) Human cytotoxic T-cell lines with restricted specificity for squamous cell carcinoma of the head and neck. Cancer Res. 53, 1461-1468. Yefenof, E., Picker, L. J., Schevermann, R. H., Tucker, T. F., Vitetta,E. S., and Uhr, J. (1993a).Cancer dormancy: Isolation and characterization of dormant lymphoma cells. Proc. Natl. Acad. Sci. U.S.A. 90, 1829-1833. Yefenof, E., Picker, L. J., Schevermann, R. H., Vitetta, E. S., Street, N. E., Tucker, T. F., and Uhr, J. (1993b). Induction of B cell tumor dormancy by anti-idiotypic antibodies. Curr. Opinion Immunol. 5, 740-744. Yewdell, J. W., Bennink, J. R., and Hosaka, Y. (1988).Cells process exogenous proteins for recognition by cytotoxic T lymphocytes. Science 239,637-640. Young, R. A. (1990). Stress proteins and immunology. Annu. Reo. Immunol. 8,401. Zinkernagel, R. M., and Doherty, P. C. (1974). Immunological surveillance against altered self components by sensitised T lymphocytes in lymphocytic choriomeningitis. Nature (London)251, 547. Zinkernagel, R. M., and Doherty, P. C. (1979).MHC-restricted cytotoxic T cells: Studies on the biological role of polymorphic major transplantation antigens determining T cell restriction-specificity, function and responsiveness. Ado. Immunol. 27, 51.

ADVANCES IN IMMUNOLOGY, VOL. 57

Formation of the Chicken B-Cell Repertoire: Ontogenesis, Regulation of Ig Gene Rearrangement, and Diversification by Gene Conversion CLAUDE-AGNES INSERM U373, In&t

REYNAUD, BARBARA BERTOCCI, AURIEL DAHAN, AND JEAN-CLAUDE WEILL Nockor, 156 we de Vaugimd, 75730 Pans Codex 05, Fmnco

1. Introduction

Past years have seen the description of immune systems from several species, for which the mechanism of B-cell repertoire formation differs fundamentally from what has been established for the mouse. In the mouse, rearrangement of Ig genes takes place continuously in the bone marrow during the lifelong differentiation of B cells from uncommitted progenitors. It provides the opportunity for random assortment of V-encoding elements (V, D, and J) and for the linked junctional diversification processes in each newly formed B cell (Tonegawa, 1983; Alt et al., 1987). In chickens, sheep, and rabbits, Ig gene rearrangement is not the key event for Ig diversity: postrearrangement diversification processes taking place during an early phase of B-cell amplification generate the B-cell repertoire in these three species. The molecular mechanisms differ however among them: in the chicken, gene conversion diversifies a unique rearranged gene at both heavy- and light-chain loci by recombination with a pool a pseudogene elements (Reynaud et al., 1985,1987,1989);in the sheep, several functional light-chain V genes undergo extensive modification by untemplated somatic mutations (Reynaud et al., 1991a); for the rabbit heavy chain, a major rearranged gene undergoes gene conversion, with possibly extensive somatic mutation of the D region (Becker and Knight, 1990; discussed by Knight and colleagues in the previous volume). These postrearrangement diversification processes take place in primary lymphoid organs in which considerable B-cell proliferation occurs and allow modifications to accumulate as further cell division proceeds: the bursa of Fabricius has been known for a long time to be the primary site of B-cell formation in the chicken; in sheep, ileal Peyer’s patches fulfill a similar function (Reynolds and Morris, 1983), the bone marrow having no B lymphopoietic activity in these two species. The question turns out to be more complex in the rabbit; 353

Copyright 8 1994 by Academic Press, Inc

All rights of reproduction In any form reserved

354

CLAUDE-AGN~SREYNAUD ET AL.

whether gut-associated lymphoid tissues (i.e., appendix and sacculus rotondus) are sites where such B-cell diversification takes place, as originally proposed by Archer et al. (1963),and whether this is superimposed to a mouse-type bone marrow lymphopoiesis are still open issues. Among these three species, the chicken immune system is so far the model for which the more complete description has been achieved since the initial description of the bursa-derived B-cell lineage (Cooper et al., 1965).We will summarize the actual knowledge of the formation of the B-cell repertoire, including data on early B-cell commitment and regulation of rearrangement obtained with chicken substrates in transgenic mice. II. Organization of Ig Genes in the Chicken

A. LIGHTAND HEAVY-CHAIN LOCI Both heavy- and light-chain loci share a similar gene organization with unique functional V and J elements, indicative of a striking coevolution (Fig. 1). For the light-chain locus, which is a A isotype, the unique Vhl gene is 1.8 kb upstream of a single J-C unit. This V gene has all the characteristics (promoter with octamer sequence, leader peptide, leader intron, recombination signals) of mouse V elements. Upstream of Vhl are 25 V pseudogenes clustered in 19 kb of DNA with alternate polarities (16 with the same polarity as Vhl and 9 with a reverse orientation) (Reynaud et al., 1985,1987). Most of these pseudogenes (but not all) lack recombination signal sequences. The homology with the VX1 gene never extends farther 5' than approximately 40 bp in the leader intron. None of them have upstream regulatory elements, i.e., transcription signals and leader sequence, some of them being even truncated genes with only part of the V coding sequence. For the heavy chain, the unique vH1 and J H elements are 15 kb apart, and JH is approximately 15 kb upstream of the C p gene. In between vH1 and JH are 16 D elements (Reynaud et d., 1989,1991b). A pool of VHpseudogenes, larger than that for the light-chain locus, has been analyzed but not completely sequenced; it covers 60-80 kb of DNA, with an average spacing of V elements of 0.8 kb (i.e., a total of 80 to 100 pseudogenes). The alternance of polarities is quasisystematic. Like that for Vh pseudogenes, homology with vH1 is restricted in 5' to ca. 80 bp in the leader intron, with no leader sequence nor

v25

liaht chain locus yr V). cluster

*-

**+++*+*v23v22v21

vlB

v17

v15 v13 v 1 2 v l l

c

Vev6

t * f

v4 vS

v6

vl I

v24

c-

v20

v16

v16 v14

e-4-

VlO

c

v7

c

VS

+

vH1

-

4

-

J

C).

I

II

II

+ I

v2

1 kb

c

d

D cluster (16 0)

yr VH cluster (80-100 yr VH)

4

4

LVk1

1

,

JH

CF

4

FIG.1. Organization of the chicken light- and heavy-chain loci. The chicken light-chain locus contains a single functional V gene (VAl), a single J-C unit, and a cluster of 25 pseudogenes (Reynaud et al., 1985,1987).A twenty-sixth JIVA element, located between JIV7 and JIVS, has been reported by Kondo et al. (1993) in the chicken H-B15 strain. The heavy-chain locus contains single Cp, JH, and functional VH ( v H 1 ) elements, a cluster of 16 D, and a group of pseudogenes (80-100 JIV, in 60-80 kb of DNA) (Reynaud et d., 1989,199113).Horizontal arrows indicate transcriptional polarities.

356

CLAUDE-AGNl?S REYNAUD ET AL.

transcription signals. In 3’, they have no recombination signals, but surprisingly a “D-like” segment fused to the V element; this D-like sequence is even terminated in some pseudogenes by a few nucleotide homology with JH,which has functional relevance to the diversification process (see below). The presence of leader intronic sequence in some of these pseudogenes argues against a processed gene origin for these fused V-D structures. B. DELEMENTS D elements are the only functional V-encoding elements present in multiple copies in the chicken genome. Sixteen D elements exist, with 15 of them being extremely homologous (some even in several identical copies) and 1 being rather different (Dx) (Fig. 2); this last one is however poorly functional, since it has a low rearrangement frequency and is selectively counterselected during B-cell expansion in the bursa (see below) (Reynaud et al., 1991b). All D elements encode an exact number of amino acids in reading frame 1: mostly 10, but also 9 or 8. The amino acid composition of the three reading frames is spectacularly different: Gly-Ser-Ala-Tyr-Cys, i.e., hydrophilic and aromatic residues, in reading frame 1; hydrophobic amino acids (Leu, Val) in reading frame 2; and one or two stop codons in reading frame 3 (Table I). This amino acid composition resembles that of the mouse Dfl16 and Dsp families, and a bias for usage of reading frame 1 is observed in the chicken as in the mouse (cf. section 1II.C). The germline-encoded D elements are only minor contributors to the overall heavy-chain repertoire, which results mainly from gene conversion (cf. 1I.A). What is thus the function of this pool of genomic D elements? Its main contribution could be the formation of D-D junctions: they amount to 25% among D-J alleles and are still maintained around 15% in “functional” VDJ sequences, indicating that such large CDR3 domains are not detrimental for the chicken Ig molecules.’ This indicates also that no specific length matching is required between heavy and light chains in the chicken, the heavy-chain CDRS varying between 15 and 30 amino acids, whereas light-chain CDRS length heterogeneity is only moderate (McCormack et aE., 1989a; Considering a homology-mediatedjunction leading to 50% D-D joining in reading frame 1 and selection of D reading frame 1 in functional VDJ sequences (Reynaud et al., 1991b), the 25%D-D junctions among V-D-J alleles are expected to result after selection in 12.5%V-D-D-J sequences with both D’s in reading frame 1 (see section 1II.C and Table 111).

D6

GGT AGT GGT TAC TGT GGT AGT GGT GCT TAT

D1 D2 D3 D4 D5 D7 D8 D9 D10 D11 D12 D13 D14 D15

GG GGT GGT GGT GGT CGT CGT GGT GGT GGT GGT GGT GGT GGT

FAGT GCT TAC GGT AGT TGT TGT

_ _ _ _ _ _A-_ _ - -

( )

GGT GCT TAT

GCT ( ) ( ) GGT CCT TAT AGT GCT TAC TGT TGT AGT GGT GCT TAT

AGT AGT AGC AGT AGT AGT AGT AGT AGT AGT AGT

GCT TAC TGT ( ) GCT TAC TGT GGT GCT TAC TGT ( ) GCT TAC TGT ( ) GGT TAC TGT GGT GGT TAC TGT GGT GCT TAC TGT ( ) GGT TAC TGT GGT GGT TAC TGT GGT GGT TAC TGT GGT GGT TAC TGT GGT

TGG GAT AGT GGT TGG TAT TGG GAT AGT GCT TGG GGT TGG GAT AGT GCT AGT GCT TGG AGT AGT GGT

GCT GAT GCT TAT GCT GAT GCT GAT GCT TAT GCT TAT GCT GAT GCT TAT GCT TAT GCT TAT GCT GAT

Dx

FIG.2. Genomic D sequences. Genomic D sequences are shown, including recombination signal sequences (Reynaud et al., 1991b). D6 has been chosen arbitrarily to be compared to, with dashes for nucleotide identity and brackets and mows for, respectively, gaps and insertions introduced to maximize homology. The D coding part is written in a triplet mode in reading frame 1 and indicated in full to facilitate comparison. Heptamer-nonamer signal sequences are boxed.

358

CLAUDE-AGN~S REYNAUD ET AL.

TABLE I THETHREEREADINGFRAMES OF THE CHICKEN D ELEMENTS HAVEA MARKEDLY DIFFERENT AMINOACID COMPOSITION’ ~~

Frame 1

I

D1 D2 D3 D5 D6 D9/12/13 D10 D14 D15 D4/8/11 D7

GSAYGC-GAY GSA-CC-GPY GSAYCCSGAY GSAYCGSGAY GSGYCGSGAY GSGYCGSAAY GSGYCGWGAY GSGYCGWSAY GSGYCGSGAD GSAYC-h”3 GSAYC-WAD

consensus

G S F F D

Frame 2

Frame 3 L*CLRLFFCL *CL-w-SL *CLLL*WcL *CLLw*WcL *wLIw*WcL *W*CCL *wwJLGcL *wLLWLECL *W*Wc* *CLL-LGC* *W-LVC*

a The amino acids encoded by the three reading frames of the D elements are shown in the one-letter code with asterisks for stop codons, dashes for gaps, and arrows for insertions introduced as in Fig. 2. Frames 1, 2, and 3 refer to sequences beginning at the first, second, and third nucleotide of the D’s, respectively. A consensus sequence is shown for the reading frame 1 (Xstanding for a nonconsewed residue).

Reynaud et al., 1991b). Surprisingly, this D-D joining requires rearrangement between two signals with the same 12-bp spacing, a mechanism that would be expected from model systems to be extremely inefficient. 111. Generation of the Chicken B-Cell Repertoire by Gene Conversion

A. A HYPERCONVERSION MECHANISMDIVERSI~~ES THE UNIQUE FUNCTIONAL VL AND VH GENES The initial observation that most chicken B cells harbor the same Vh gene rearrangement made the question of the origin of the Ig repertoire in this species an obvious one. Light-chain cDNA sequences isolated from splenic cells displayed striking blocks of homologies with the first three pseudogenes analyzed, which led us to propose a gene conversion mechanism between V elements (Reynaud et al., 1985), analogous to the recombination models proposed 25 years ago to explain the generation of diversity, models that turned out to be irrelevant to the mouse immune system (Smithies, 1967; Edelman and Gally, 1967).

359

CHICKEN B-CELL REPERTOIRE FORMATION

The systematic analysis of rearranged genes during bursal development together with the complete sequence of the V-encoding elements of the locus confirmed that a gene conversion-like mechanism was responsible for generating diversity in this species (Reynaud et al., 1987). The recombination process transfers in a nonreciprocal manner blocks of sequence from the pseudogene pool acting as donor into the unique functional gene acting as acceptor. Ongoing diversification of the V h l gene was also demonstrated by Thompson and Neiman (1987), using restriction sites located in the V sequence. The number of converted segments increases with time, from an average of one to three at Day 18 of embryonic development to four to six at 3 weeks after hatching (Fig. 3); this figure represents an underestimate since conversion tracks may superimpose on each other. Bursa1 cells divide every 8-10 hr during the embryonic period, and a high division rate is probably maintained after hatching. A frequency of one gene conversion event per V sequence every 10-20 cell divisions can thus be estimated (Reynaud et al., 1987).

I

I I I

I

I

I

I

1

I I

I I

I

I

I

I I

I

I

I

I

I

I

I

I

I

I

FIG.3. Diversification of bursal VX1 sequences through multiple gene conversion events. Six rearranged bursal V h l sequences isolated from either Day 18 embryos or a 3-week-old chicken are compared to the original VX1 sequence (Reynaud et al., 1987). The various parts of the variable region are delineated: leader (L), framework (FR), complementarity-determining regions (CDR), and J segment ( J). Domains in white indicate regions of nonmodified VX1 sequence; domains in gray represent approximate borders of gene-converted segments, with the name of the putative pseudogene donor (several numbers in a box refer to multiple equivalent donors). Dots represent single untemplated nucleotide modifications.

360

CLAUDE-AGNES AEYNAUD ET AL.

The length and position of converted segments vary widely. Length can be from 10 to more than 100 bp (a 250-bp event was reported) (Reynaud et al., 1987; McCormack and Thompson, 1990). No precise borders are observed for the conversion tracks that affect all parts of the V sequence including the D region. D reading frame 1 is already largely selected among bursa1 VDJ sequences when incidence of gene conversion can be monitored (cf. section 111.C);the D-like sequence is fused to the $VH elements in reading frame 1 and therefore does not modify the overall reading frame of the Ig molecule when it is transferred by gene conversion. Due to the particular structure of several $VH and $Vh elements that end in 3’ with a few nucleotide homology with the J H and Jh segments, gene conversion is moreover able to extend over the V-J (V-D-J) junction. This particular feature of the gene conversion process probably dictates the particular Ig gene configuration of chicken B cells: only one allele is rearranged for both heavy- and light-chain loci, a configuration which avoids the presence of abortive rearrangements for the other allele that would be susceptible to being put back in frame by gene conversion and would result in allelic inclusion of the cell. This point, originally raised by M. Cohn, is discussed at length in a Forum in Immunology (Langman and Cohn, 1993). Differential pseudogene usage has been observed (Reynaud et aE., 1987; McCormack and Thompson, 1990): proximal genes were used more frequently, and the pseudogenes with higher homology to V h l were similarly favored. This homology-biased exchange may vary with further diversification of the V h l gene, possibly restricting pseudogene usage. In the DT40 cell line which undergoes gene conversion in culture (cf section IIB), the very particular CDRl structure of the Vh1 gene (with $VS sequence) biases further gene conversion toward the use of the homologous $Vl8 gene (Buerstedde et al., 1990; Kim et al., 1990). Such sequential pseudogene usage has been described during the gene conversion process leading to surface antigen variation in the trypanosome (Roth et aZ., 1986). The third type of pseudogene usage preference concerns their orientation, the ones with an inverted polarity compared to V h l being more frequent donors (McCormack and Thompson, 1990). By studying cell lines with restriction enzyme polymorphism in the $V cluster, Carlson et aZ. (1990) have brought evidence that the mechanism used was indeed intrachromosomal gene conversion, i.e., a mechanism operating in cis with nonreciprocal exchange of DNA. McCormack and Thompson (1990) observed furthermore that, on the average, a longer stretch of homology flanks gene conversion tracks

CHICKEN B-CELL REPERTOIRE FORMATION

36 1

in 5’ than in 3‘ (referred to V coding sequence), which led them to propose a polarity in the sequence transfer mechanism with initiation on the 5’ side. Untemplated mutations are also observed (8 among 214 mutations collected from bursal sequences before polymerase chain reaction times (Reynaud et aZ., 1987)) frequently located at the border of a conversion event. Such mutations might be intrinsincally linked to the recombination process. The occurrence of further somatic mutation in peripheral B cells has been suggested by Parvari et al. (1990). However, a systematic study of somatic mutations of IgG versus IgM molecules during the course of an immune response has not been performed so far. B. DT40: AN ALV-INDUCED CELLLINEUNDERGOING IG GENE CONVERSION CULTURE Several transformed cell lines from bursal origin have been tested for their capacity to undergo gene conversion of their Ig gene in culture. Bursa1 cells immortalized in uitro by the REV-T virus (reticuloendotheliovirus) have no gene conversion activity in culture (Barth and Humphries, 1988). Avian leukosis virus (ALV)-induced tumors clonal for the virus integration site have heterogeneous diversified lightchain sequences, indicating that gene conversion was active during tumor cell proliferation in the animal, irrespective of the anatomical localization of the tumor; however, among those that have been established as cell lines in uitro, only one (DT40) has shown ongoing gene conversion activity in culture (Thompson and Neiman, 1987; Buerstedde et al., 1990; Kim et al., 1990). There are however major differences in the gene conversion activity of this cell line compared to that of normal bursal cells. The rate of gene conversion appears to be much slower, with small changes of one or two nucleotides being the most frequent event. It is unclear so far whether this represents a trend specific to the DT40 toward recombination between regions of high homology (i.e., exchanging few nucleotides) or whether such small changes are an integral part of the normal process that is wiped out by longer conversion tracks occurring at higher frequency in the chicken bursa. Spontaneous IgM-negative variants of the DT40 cell line were analyzed to characterize possible errors in the gene conversion mechanism; such variants occurred at a low frequency (less than 1%of the total population) and were due to a frameshift mutation in the V h l sequence, thus indicating a low error rate in the gene conversion process. Such negative variants could give rise upon culture to IgM-

362

CLAUDE-AGNh REYNAUD ET AL.

positive subclones, the frameshift mutation being corrected by a superimposed gene conversion (Buerstedde et al., 1990). It is unlikely that such a correction process plays a major role in the bursa, the lifespan of B cells with nonfunctional IgM being probably too short for them to be rescued. A high frequency of gene targeting is obtained upon transfection of the DT40 cell line (Buerstedde and Takeda, 1991). This feature is however shared by other chicken B-cell lines, whether gene conversion active or not, and is not specific for the Ig loci; it might thus reflect some general recombination activity that can be dissociated from the gene conversion process. This observation allows nevertheless specific genes involved in the gene conversion machinery to be sought by knocking them out on both alleles in the DT40 cell line: such experiments have already suggested that the high RAG-2 gene expression observed in the bursa was unnecessary for the maintenance of the gene conversion activity (see below) (Takeda et al., 1992). C. RECOMBINATION MODELSFOR IG GENECONVERSION IN THE CHICKEN: Is THERE FORMATION OF HOLLIDAY JUNCTIONS? Models of gene conversion involve the formation of DNA single- or double-strand breaks as initiator sites ofrecombination, strand displacement, gap repair, and resolution of crossed DNA strands (Holliday junctions). The closest related model is the yeast-mating type switch: this event consists of a programmed gene conversion whereby sequences at the MAT (mating-type specific) locus are replaced using either of two possible donors, which can occur as frequently as once per generation. The whole process is induced by a double-strand break created by the site specific HO endonuclease that results in an exchange of ca. 700 bp. The directionality of the exchange is ensured through the silencing of the donor genes that prevents recognition and cutting by the HO endonuclease (reviewed by Haber, 1992). No such DNA breaks have been reported so far for chicken V genes in the developing bursa (Thompson, 1992),but we are still investigating the question using more sensitive techniques. Heptamer sequences specific for the rearrangement process exist in two locations within the V h l gene (and in the VH1 and half of the D elements as well). Although it was tempting to propose that they could be recognized by some lymphoid-specific, heptamer-nicking activity (Reynaud et al., 1987), it was nevertheless difficult to envision how so few entry sites could generate gene conversion tracks with such diverse end points. This proposal was raised again by Carlson et al. (1991), who observed high RAG-2 gene expression in the bursa and proposed that RAG-2 might represent this heptamer-nicking activity. Correlation of

CHICKEN B-CELL REPERTOIRE FORMATION

363

RAG-2 expression with gene conversion was however questioned due to the observation that disruption of both RAG-2 alleles in the DT40 cell line did not modify its gene conversion activity (Takeda et al., 1992). The role of RAG-2 in the bursa and the existence of specific DNA breaks promoting gene conversion in the V sequence are thus open issues. McCormack and Thompson (1990)proposed a model of recombination to account for their observation ofpolarity in the conversion tracks, which involves strand displacement following the direction of transcription that would be resolved by strand dissociation. In their model, the authors exclude the formation of Holliday junctions, arguing that no crossing over associated with their resolution has been reported. Such structures however would result in nonfunctional V sequences that would be rapidly eliminated in the bursa; crossing over in a gene conversion involving a pseudogene donor with the same orientation as VA1 would result in a fused 5’ ($V-Vhl-J) 3‘ sequence with excision of the DNA sequence in between (Fig. 4a); when involving a pseudogene donor with an opposite polarity, crossing over then results in an inversion that would produce again similar nonfunctional hybrid $V-Vhl structures (Fig. 4b). Only in the DT40 cell line, whose proliferation does not depend upon a functional Ig molecule, would it be possible to observe them. It is nevertheless anticipated that associated crossing overs are rare events in such intrachromosomal gene conversions, and this is also the case in the mating-type switch (Klar and Strathern, 1984; Ray et al., 1988). Kondo et al. (1993)have described circular DNA structures isolated from the bursa. Two circular DNA with an “abnormal” V gene configuration were reported, both containing a hybrid V gene formed by the 5’ end of the VA1 gene and the 3‘ end of a pseudogene having the same polarity as VX1. Such structures represent strikingly the DNA product expected from a crossing over associated with a gene conversion event (cf. Fig. 4a), although obviously other explanations exist for their formation. Whether such rare structures reflect the normal pathway of gene conversion, i.e., resolution of Holliday junctions, with few percent of associated crossing over due to topological constraints or whether they are accidental, atypical structures remains to be determined. If a systematic link could be established between the presence of such hybrid V genes in circular DNA and the gene conversion activity of a cell, their detection could represent a “gene conversion assay,” analogous to the gene rearrangement assay constituted by the detection of the circular DNA region excised upon rearrangement (McCormack et al., 1989a). When considering an activity or a signal that would activate the

364 a

CLAUDE-AGNfb REYNAUD ET AL.

L Vhl J

wV6

nrft t

b

ww

L Vhl J

t

wEIl-L Vhl J

t

t

6@ I I Vh1 J

-BEG+

Q ~

L Vhl lwV6

FIG.4. Bursa1 circular DNAs with hybrid V genes: Possible products of the Resolution of HoIliday Junctions? The DNA products expected from a crossing-over associated with the resolution of a Holliday junction are depicted, the gene conversion event initiating the strand exchange not being represented. Two possible cases exist, depending upon the orientation of the pseudogene participating in the gene conversion: (a) when a pseudogene with the same orientation as the VAl gene is involved, excision of a circular DNA molecule occurs; (b)with a +V in the reverse orientation, an inversion occurs. Circular DNAs containing hybrid 5' VA1/3' +V sequences as described in (a) have been isolated from bursa1 cells by Kondo et al. (1993).

gene conversion program in chicken B cells, it is striking that the same question pertains both to gene conversion of V genes during chicken B cell development and to somatic mutation of V genes during the secondary immune response. Both" are basic, nonspecific mechanisms, involved in the evolution of genes, that become specifically activated at a high rate at a precise developmental stage and are targeted exclusively to both VH and VL genes without affecting other genomic sites.

CHICKEN B-CELL REPERTOIRE FORMATION

365

How the specific recognition of the V sequence is achieved remains the key question in both of these processes. IV. Chicken B-Cell Development

A. N o ONGOING IG GENEREARRANGEMENT IS OCCURRING IN THE CHICKEN BURSA The bursa of Fabricius is the primary organ of B-cell formation in the chicken; bursectomy in the embryo results in a severe agammaglobulinemia, a discovery that allowed the B (bursa-derived)-cell lineage to be identified (Glick et al., 1956; Mueller et al., 1959; Cooper et al., 1965,1969; Warner et al., 1969). The stem cells (prebursal stem cells) that give rise to B lymphocytes colonize the bursa from the general circulation between Days 8 and 14 of embryonic development (Moore and Owen, 1967; Le Douarin et al., 1975; Houssaint et al., 1976).When they reach the bursal epithelium, they induce the formation of lymphoid follicles and, as bursal stem cells, differentiate into B lymphocytes. The bursa is composed of lo4 follicles, colonized by a few prebursal stem cells (two to three on average), these lymphoid follicles reaching a size of 2-5 x lo5cells at maximal bursal size. B cells start to migrate out of the bursa at Day 18 of embryonic development, and install the peripheral lymphoid compartment as postbursal stem cells. Three weeks after hatching, removal of the bursa has no incidence on the further development of the B-cell repertoire and by 6 months of age, the bursa has completely involuted (reviewed by Pink, 1986). The terms prebursal, bursal, and postbursal stem cells were coined by Toivanen and Toivanen (1973);they refer not simply to their successive localization during the development of the animal, but rather to the fact that they either need or do not need to migrate in the bursa to restore the B-cell compartment of a depleted host in adoptive cell transfer experiments. However, despite this terminology, it is still totally unclear for the moment whether there exists a cell population with true “stem cell” characteristics. It remains equally possible that every chicken B cell at both bursal and postbursal stages is endowed with a large proliferative capacity, the B-cell population ensuring as a whole its long-term maintenance. When first described, B-cell development in the bursa appeared as a typical instructive model, wherein a multipotent progenitor would be driven into the lymphoid pathway through the induction of a specific microenvironment. One would thus expect Ig gene rearrangements to

366

CLAUDE-AGNRS REYNAUD ET AL.

be constantly produced in each newly formed B cell. This turned out to be doubly wrong; first, stem cells migrating into the bursa are already committed to the B-cell lineage (as discussed in the following section); second, Ig gene rearrangement is a unique event, occurring during a short time and does not take place continuously during the bursal phase of B-cell development. The study by Pink and his colleagues established at first that the bursal stem cell, responsible for restoring the immune capacity of a B-cell-depleted host, was a surface IgM-positive cell (Pink et at., 1985). Accordingly, long-lasting allotype suppression could be achieved by a single anti-IgM anti-allotype injection at Day 13 of development, a situation in striking contrast to the one in the mouse (Ratcliffe and Ivanyi, 1981). Study of Ig gene configuration in individual lymphoid follicles established that rearrangement is not an ongoing event in the bursa; using a restriction enzyme polymorphism at the light-chain locus, individual follicles were shown to be oligoclonal for rearrangement, the number of rearrangement events (two to three on average) corresponding to the number of progenitors that colonized each follicle (Weill et al., 1986). A similar conclusion was reached from the analysis of IgM allotype distribution in bursal follicles (Ratcliffe et al., 1986). IgMcommitted stem cells are thus formed at the very beginning of the colonization process, or even slightly before. This last point was confirmed using systematic polymerase chain reaction (PCR) analysis of Ig gene rearrangement during early development (cf. section 111,B). More recently, the issue of ongoing rearrangement in the bursa was addressed by McCormack et al. (1989a), who monitored by PCR the formation of the circular DNA excised during light-chain rearrangement. This elegant “rearrangement assay,” which is of general relevance for rearrangement of V elements with the same polarity, failed to detect such circular structures at late embryonic and posthatching stages, in agreement with previous conclusions. Analysis of Ig gene rearrangement in the total bursa and in individual follicles has revealed an Ig gene configuration particular to chicken B cells: only one allele is rearranged for both heavy- and light-chain loci; the other allele remains in germline configuration for the light chain and mostly DJ for the heavy chain (with even 5% of cells with a heavy-chain locus in germline configuration). About 1-2% cells were estimated to have both alleles rearranged. A concordant, but slightly higher value was reported from analysis of chicken B-cell lines (6% of cell lines with both light-chain alleles rearranged) (Reynaud et al., 1985,1989;Weill et al., 1986; McCormack et al., 1989b). We proposed

CHICKEN B-CELL REPERTOIRE FORMATION

367

that this particular feature is caused by rearrangement being attempted only once on a chromosome, the vast majority of cells with unproductive rearrangements being simply discarded (Reynaud et al., 1987). Such a restricted rearrangement would be achieved through a short window of time during which access to the genes would be allowed. A possible regulation for this phenomenon is described in section 1II.D. The most striking feature of B-cell development in the chicken is thus the existence of a unique “wave” of Ig gene rearrangement in B-cell progenitors. The entire B-cell compartment is thus derived from the 2-3 x lo4 productively rearranged bursal stem cells that develop within bursal follicles. Contrary to previous expectations, the main function of the bursa therefore appears to be the expansion and the diversification of the B-cell population (Weill and Reynaud, 1987; Pink and Lassila, 1987). B. EARLY B-CELLCOMMITMENT IN THE DEVELOPING CHICKEN EMBRYO As discussed in the previous section, committed B-cell progenitors have a unique phenotype in the chicken, having their Ig genes in a rearranged configuration and presenting a surface IgM molecule. This unique property allows us to ask where and when these progenitors emerge during embryonic development by following at the singlecell level the rearrangement pattern of embryonic cells. This can be done at all stages of development, taking advantage of the easy access to the embryo; the simplicity of chicken Ig loci with single VH,V,, and J elements; and the use of the PCR technique allowing the detection of one rearrangement event in lo5 cells (Reynaud et al., 1992a). Based on these and on previous results (Moore and Owen, 1967; Dieterlen-Lihvre and Martin, 1981; Lassila et al., 1978,1982;Ratcliffe et al., 1986; Houssaint et al., 1991),one can recapitulate the development of the B-cell system in the early chicken embryo (Fig. 5).Hematopoietic precursors can first be detected in the intraembryonic paraaortic region at Day 3-4. These progenitors then migrate to the yolk sac at Day 5-6 at which stage the specific B-cell progenitors start to perform D to JH rearrangement, thus segregating from the other hematopoietic lineages (chicken T cells as opposed to mouse T cells do not show any DJH rearrangement). This DJH rearrangement could be triggered either by the yolk sac environment or by the removal of a differentiation block as cells migrate out from intraembryonic sites. DJH rearrangement goes on as B-cell progenitors circulate back to the embryo and seed the various lymphoid organs. One or two days later, whether

368

CLAUDE-AGNh REYNAUD ET AL.

I "

\ DJ+ day 10

t J

FIG.5. A proposed scheme for the emergence of B-cell progenitors in the chicken embryo (Reynaud et al., 1992a). Hematopoietic progenitors are first detected within the embryo in the paraaortic region (Lassila et al., 1978; Dieterlen-LiBvre and Martin, 1981); they further migrate through the yolk sac, where DJH-committed progenitors start to segregate from the other cell lineages and seed via the general circulation the various lymphoid organs. These progenitor populations regress thereafter in spleen and bone marrow (marked with a t symbol) and only expand in the bursa ( t t t ) (cf. Table 11). The first day of detection of DJH rearrangement is indicated for each compartment.

in the blood or in a lymphoid organ, they start to rearrange their heavyand light-chain V genes, this event being most probably part of an intrinsic cellular program rather than being induced in any particular site. Once in spleen or in bone marrow the B-cell progenitors, whether they have rearranged their Ig genes productively or not, cease to divide and slowly disappear. To support this view we have shown that there is neither selection for functional sequences in the spleen nor B-cell proliferation (as monitored by the ratio of rearranged cells to light chain circular DNA excision product) (Reynaud et al., 1991b71992a). When put in quantitative terms, the picture indeed favors a role of the general circulation in distributing DJH progenitors to the various organs (Table 11);approximately 1000 DJH progenitors are present at Day 8 in the yolk sac, and then DJHprogenitors dominate in the blood at Day 10 (125,000-250,000 cells). This population remains very high in blood at Day 13 as well as in the spleen and bone marrow

CHICKEN B-CELL REPERTOIRE FORMATION

369

TABLE I1 NUMBER OF DJH-COMMITTED PROGENITORS IN VARIOUS EMBRYONIC COMPARTMENTS“ Days of Development: Yolk sac

Blood

Spleen Bone marrow Bursa Thymus

Number of DJH Progenitors Day 8

1000-2000

Day 10

Day 13

Day 17

125,OOO-250,000 2,000-5,000 10-20 50-200 100

250.000-500,000 500,000 100,OOO 30,000 400-1,000

500-1,OOO 20,000-40,OOO 2,000-10,000 > 1-2 x 106 1,6bO-4,000

a The total number of DJH-committedprogenitors in various compartments at four developmental stages was estimated from a quantitative PCR analysis reported to the size of each organ at each stage (Reynaud et al., 1992a). Figures underlined represent the major DJH progenitor compartment(s) at a given stage.

(100,000-500,000 cells), while it starts to accumulate in the bursa. At Day 17 these populations have declined in blood, spleen, and bone marrow but expand strikingly in the bursa. Our data do not disagree with previous reports showing commitment of B-cell progenitors in nonbursal sites, these cell populations being able moreover to restore a B-cell-depleted animal in cell transferexperiments (Ratcliffe et al., 1986; Houssaint et al., 1989,1991).We estimate however that such sites represent dead ends rather than functional intermediates for further bursal development. A large excess of cells (1-2 x lo6)is thus engaged in the B-lymphoid lineage, compared to the approximately 2-3 X 104 productively rearranged committed B-cell progenitors that are needed to generate the B-cell compartment in the bursa. IG SEQUENCES ARE SELECTED DURING B-CELL C. FUNCTIONAL PROLIFERATION I N THE BURSA After bursal colonization by B-cell progenitors, there is a defined period (Day 10-18) during which selection for “functional” Ig sequences takes place, before and independent of the occurrence of gene conversion; this selection bears on both in-frame sequences and on the D reading frame and can be easily monitored on the total bursal population since only one Ig allele is rearranged per cell (McCormack et al., 1989b; Reynaud et al., 1991b). Heavy- and light-chain gene rearrangement in the bursa, analyzed at the time of DJ progenitor colonization (Day 10-12), represents the

370

CLAUDE-AGNES REYNAUD ET AL.

outcome of the joining process with characteristics similar to those described for the mouse fetal repertoire: absence of N additions, biased joining at both V-D and D- J junctions due to terminal homology, and favoring the D reading frame 1; 60% of VD and 50% of DJ result in D reading frame 1, which makes about 113 of “functional” VDJ sequences (i.e., in frame with a D in reading frame l), as opposed to 119 expected from random joining (Table 111).The proportion of the three reading frames on the DJ allele does not change with time, which suggests that the selection for the D reading frame is produced on the assembled VDJ structure presented at the cell surface (Reynaud et al., 1991b). For the light chain, the joining appears to be roughly random (McCormack et al., 1989b). A total of 80-90% of Ig sequences is already in frame (with a D in reading frame 1)at Day 15, and this amounts to more than 95% at Day 18 of development (Table 111). What is thus the structure signaling for cell proliferation and selecting for “functional” Ig molecules? It has been proposed that the germline-encoded specificity (i.e., the VHl-VA1 pair) is recognizing some bursal determinant that would induce further development (Reynaud et al., 1989; McCormack et al., 1989b; see also the discussion in Langman and Cohn, 1993). Alternatively, the mere presence of an IgM molecule at the cell surface could be signaling for proliferation and would result in selection of in-frame sequences with a D region ensuring the proper folding of the Ig molecule, i.e., in reading frame 1 (cf. Table 1). The hypothesis of the recognition of a bursal ligand by the germline VH1-VAl specificity has been further developed to propose that abolition of this germline recognition by gene conversion would signal the arrest of proliferation and allow the mutated cell to leave the bursa (McCormack et al., 1989b). It is however difficult to envision, since light (and heavy)-chain sequences can accumulate six to eight gene conversion tracks, that only the very last event abolishes the binding of the bursal ligand, all the previous events having maintained the germline-encoded specific recognition. Is there also a role for external antigens in selection of B cells as further diversification proceeds in the developing bursa? This question was raised, since antigen uptake by specialized bursal epithelial cells has been described (Toivanen et al., 1987). Moreover, bursal duct ligation, isolating the bursa from contact with gut-associated antigens, was proposed to result in reduced B-cell diversification. We do not think however that, since the initial description of B-cell development in the bursa as an antigen-independent process (Lydyard et al., 1976), any function other than a nonspecific mitogenic signal has been estab-

TABLE 111 SELECTIONOF HEAVY-CHAIN REARRANGEMENTS IN BURSALVERSUS NONBWRSAL SITES"

DJ VDJ bursa Day 13 VDJ bursa Day 15 VDJ bursa Day 18 VDJ spleen Day 15

Number of Sequences

In-frame Sequences

48 28 12 45 22

78% 92% 96%

50%

D Reading Frame Frame 1

Frame 2

Frame 3

D-D Junctions

50% 67% 92% 95% 59%

31% 30% 8% 5% 27%

19% 3% 0% 0% 13%

25% 21% 17% 15% 18%

Heavychain rearranged sequences have been analyzed concerning the overall reading frame, the frame ofthe D elements, and the incidence of D-D junctions, at Days 13, 15, and 18 of development in the bursa and at Day 15 in the spleen. The reading frame of the D is referred to J for DJ sequences and to V for VDJ sequences (taken from Reynaud et al., 1991b).

372

CLAUDE-AGNES REYNAUD ET AL.

lished for antigens present in the bursa (Ekino, 1993).A similar question on the role of gut-associated antigens in diversification of B cells in sheep ileal Peyer’s patches has a very clear answer: the rate of Ig diversification is unchanged in genn-free animals or in segments of the ileum isolated from the intestinal tract (analogous to bursa1 duct ligation) (Reynaud et aE., 199210; Reynaud et al., in preparation); however, cell proliferation is impaired after some weeks without external antigen contact.

D. ALLELICEXCLUSION IN CHICKEN B CELLS:A SILENCER/ ANTISILENCERREGULATION Regulation of rearrangement at the Ig locus operates in the chicken as well as in the mouse: Ig genes are rearranged in B cells only, and allelic exclusion is performed for both heavy- and light-chain gene rearrangement. The main difference between the two systems comes from the short time during which rearrangement takes place, with two main consequences: first, the concommitant rearrangement of both heavy- and light-chain genes (no heavy-, then light-chain regulation, possibly mediated by a pre-B receptor), which makes chicken rearrangement closer to a “stochastic” process (Benatar et al., 1992; Reynaud et al., 1992a); second, the fact that only one allele is rearranged, no “second trial” being allowed in case of abortive joining (this precludes any regulatory function of the assembled Ig molecule in the process of allelic exclusion). The chicken light-chain locus with its natural compaction is thus an ideal substrate for investigating regulatory elements giving access to the rearrangement enzymes. Rearrangement of such a transgene with 11.5 kb of DNA is observed in mouse B cells (Bucchini et al., 1987). Further deletions/mutations of this rearrangement substrate have defined four regions involved in the regulation of rearrangement (Lauster et al., 1993) (Fig. 6). Positive regulatory elements map to promoter and enhancer elements, the promoter region necessary for efficient rearrangement being larger than the sole octamer motif. Such a role of enhancer elements in the control of Ig and TCR rearrangement has also been shown in the mouse, by both transgenic and knock-out experiments (Ferrier et al., 1990; reviewed by Chen and Alt, 1993). A negative regulatory element is present in the V-J intervening sequence, the region excised upon light-chain rearrangement, this element showing strong transcriptional silencing activity in a CAT assay in uitro. One or two sites flank this region that have no positive effect on rearrangement by themselves, but only in conjunction with the V-J intervening sequence; if these sites are mutated in the context of

CHICKEN B-CELL REPERTOIRE FORMATION

IH 1H

B PROGENITOR

V

“UO“

V

V

C

J

C

J

J” Q

B CELL

J

373

C

Antisilencer binding gives access to the recombinasemachinery

Silent allele V J

C

Transcription

FIG.6. A silencerlantisilencer regulation of chicken light-chain gene rearrangement. Four DNA elements regulating rearrangement of the chicken light-chain locus in transgenic mice have been described (Lauster et al., 1993): two positive regulatory elements, the promoter and the enhancer regions (the enhancer is located 3’ of CA, as described by Hagman et al. (1990)for the mouse A locus); one negative control element, corresponding to a strong transcriptional silencer, located in the V-J intervening sequence excised upon rearrangement (“Uo segment”); one (or two) putative elements located on one (orboth) side(s)ofthe Uo segment, antagonizing the effect ofthe silencer. It is proposed that the antisilencer factors would be present transiently in chicken Bcell progenitors, removing the silencer and allowing rearrangement to be performed on one allele (see Section 111,D). In the mature B cell, the remaining silencing element maintains the unrearranged allele in a silent configuration,the other allele being actively transcribed. Binding of the comesponding promoter (P), enhancer (E), silencer (S),and antisilencer (AS) factors is represented.

374

CLAUDE-AGNES REYNAUD ET AL.

the normal V-J negative element, rearrangement is abolished in such a transgenic construct; on the contrary, these mutations have no negative effect if the V-J intervening sequence is replaced by a “neutral” segment of DNA (Lauster et al., 1993). We suggest that they act as antisilencer elements that would function by counteracting the silencing effect of the V-J segment (Fig. 6). We have proposed, to explain the particular configuration of Ig genes in chicken B cells with one allele rearranged and one allele in germline (or DJ) configuration, that a short window oftime allows rearrangement to be attempted only once on a chromosome; a transient antisilencer expression would give access to the Ig gene during a short time, after which silencer activity would dominate to ensure the exclusion of the other allele, making it unnecessary to coordinate the shut-off of the RAG genes with the expression of a functional Ig. Indeed, not only is RAG-2 expressed at a high level in the bursa (Carlson et al., 1991), but we also found RAG-1 transcription (at a lower level, but still detectable by Northern blot analysis (Reynaud et al., 1992a)), which might reflect some “leakiness” in the regulation of its expression, once Ig rearrangement is performed. Silencing activity through the V-J excision segment might also have relevance for gene conversion, maintaining the unrearranged allele in a silent inaccessible configuration, despite the presence of transcription factors and the small distance between promoter and enhancer regulatory elements in the chicken Ig locus in its germline configuration. Such a regulation is observed on a chicken transgene in mouse B cells, suggesting that it has been conserved between these two species: since activtors of rearrangement largely coincide with transcription factors, a dominant negative regulation exerted by the region excised upon rearrangement would be crucial to maintain an allele unrearranged in a recombination-active cell (e.g., a VDJ/DJ mouse pre-B cell proceeding to light rearrangement). V. Concluding Remarks

The chicken B-cell immune system is attractive because of its apparent simplicity: one can easily follow and count the different actors of the play. A few million cells are enrolled at the very beginning of embryonic development to build up the system. Thereafter, 20,000 to 30,000 B-cell progenitors having rearranged productively their unique V, and VL genes and having colonized the 10,000 bursa1 follicles produce the B-cell lineage. During B-cell expansion in the bursa, gene conversion generates a

CHICKEN B-CELL REPERTOIRE FORMATION

37s

diversified B-cell repertoire. After a few months, the bursa involutes completely, the animal maintaining lifelong immunity with this established peripheral B-cell compartment. Strikingly, the chicken model appears today less and less evolutionarily distant from mammalian immune systems. Rabbits use gene conversion to generate B-cell diversity and this process may well occur during a short period of development in gut-associated lymphoid tissues (GALT). Sheep, and probably ruminants in general, use GALT that are only present during early development to generate their Bcell repertoire. Surprisingly in this species, the molecular mechanism used is an antigen-independent hypermutation process. In this new outlook, bone marrow lymphopoiesis involving ongoing Ig gene rearrangement seems mainly a characteristic of rodents and men. Many questions remain unsolved, and they appear to be common to the various B-cell systems which have been studied. These questions concern the control of Ig gene rearrangement and therefore of allelic exclusion, the processes of selection occurring during the different stages of B-cell formation, and the homeostasis between the different B-cell subpopulations and the relative contribution of de novo B-cell production versus expansion of preexisting clones as the animal ages.

REFERENCES Alt, F. W., Blackwell, T. K., and Yancopoulos, G. D. (1987).Development of the primary antibody repertoire. Science 238, 1079-1084. Archer, 0. K.,Sutherland, D. E. R., and Good, R. A (1963).Appendix of the rabbit: A homologue of the bursa in the chicken. Nature (London)200,337-339. Barth, C. F., and Humphries, E. H. (1988).Expression of u-re1 induces mature B cell lines that reflect the diversity of avian immunoglobulin heavy and light chain rearrangements. M o l . Cell. Biol. 8,5358-5368. Becker, R. S., and Knight, K. L. (1990).Somatic diversification of immunoglobulin heavy chain VDJ genes: Evidence for somatic gene conversion in rabbits. Cell (Cambridge, Mass.) 63,987-997. Benatar, T., Tkalec, L., and Ratcliffe, J. H. (1992).Stochastic rearrangement of immunoglobulin variable-region genes in chicken B-cell development. Proc. Natl. Acad. Sci.

U.S.A.89,7615-7619.

Bucchini, D., Reynaud, C.-A., Ripoche, M. A,, Grimal, H., Jami, J., and Weill, J.-C. (1987).Rearrangement of the chicken immunoglobulin gene occurs in the lymphoid lineage of transgenic mice. Nature (London) 326,409-411. Buerstedde, J. M., and Takeda, S. (1991).Increased ratio of targeted to random integration after transfection of chicken B cell lines. Cell (Cambridge, Mass.) 67, 179-188. Buerstedde, J. M.,Reynaud, C. A., Humphries, H. E., Olson, W., Ewert, D. L., and Weill, J. C. (1990).Light chain gene conversion continues at high rate in an ALVinduced cell line. E M B O J . 9,921-927. Carlson, L. M., McCormack, W. T., Postema, C. E., Humphries, E. H., and Thompson, C. B. (1990).Templated insertions in the rearranged chicken IgL V gene segment arise by intrachromosomal gene conversion. Genes Deu. 4,536-547.

376

CLAUDE-AGNES REYNAUD ET AL.

Carlson, L. M., Oettinger, M. A., Schatz, D. G., Masteller, E. L., Hurley, E. A., McCormack, W. T., Baltimore, D., and Thompson, C. B. (1991).Selective expression of RAG2 in chicken B cells undergoing immunoglobulin gene conversion. Cell (Cambridge, Mass.) 64,201-208. Chen, J., and Alt, F. W. (1993). Gene rearrangement and B-cell development. Curr. Opinion Immunol. 5, 194-200. Cooper, M. D., Peterson, R. D. A., and Good, R. A. (1965). Delineation of the thymic and bursal lymphoid systems in the chicken. Nature (London) 205,143-146. Cooper, M. D., Cain, W. A., Van Alten, P. J., and Good, R. A. (1969).Development and function of the immunoglobulin producing system. I. Effect of bursectomy at different stages of development on germinal centers, plasma cells, immunoglobulins, and antibody production. lnt. Arch. Allergy 35,242-252. Dieterlen-LiBvre, F., and Martin, C. (1981). Diffuse intraembryonic hemopoiesis in normal and chimeric avian development. Deu. Biol. 88, 180-191. Edelman, G. M., and Gally, J. A. (1967).Somatic recombination of duplicated genes: An hypothesis on the origin of antibody diversity. Proc. Natl. Acad. Sci. U.S.A. 57, 353-358. Ekino, S. (1993).Role of environmental antigens in B cell proliferation in the bursa of Fabricius at neonatal stage. Eur. J . Immunol. 23, 772-775. Ferrier, P., Krippel, B., Blackwell, T. K., Furley, A. J. W., Suh, H., Winoto, A., Cook, W. D., Hood, L., Costantini, F., and Alt, F. W. (1990). Separate elements control DJ and VDJ rearrangements in atransgenic recombination substrate. EMBO J . 9,117-125. Glick, G., Chang, T. S., and Jaap, R. G. (1956).The bursa of Fabricius and antibody production. Poult. Sci. 35,224-225. Haber, J. E. (1992).Mating-type gene switching in Saccharomyces cereuisiae. Trends Genet. 8,446-452. Hagman, J., Rudin, C. M., Haasch, D., Chaplin, D., and Storb, U. (1990). A novel enhancer in the immunoglobulin lambda locus is duplicated and functionally independent of NFKB. Genes Deu. 4,978-992. Hesse, J. E., Lieber, M. R., Mizuuchi, K., and Gellert, M. (1989).V(D)J recombination: A functional definition of the joining signals. Gene Dew. 3, 1053-1061. Houssaint, E., Belo, M., and LeDouarin, N. M. (1976). Investigations on cell lineage and tissue interactions in the developing bursa of Fabricius through interspecific chimeras. Deu. Biol. 53,250-264. Houssaint, E., Lassila, O., and Vainio, 0. (1989). Bu-1 antigen expression as a marker for B cell precursors in chicken embryos. Eur. J . Immunol. 19,239-243. Houssaint, E., Mansikka, A., and Vainio, 0. (1991).Early separation of B and T lymphocyte precursors in chick embryo. J . Exp. Med. 174,397-406. Kim, S., Humphries, E. H., Tjoelker, L., Carlson, L., and Thompson, C. B. (1990). Ongoing diversification of the rearranged immunoglobulin light chain gene in a bursal lymphoma cell line. Mol. Cell. Biol. 10, 3224-3231. Klar, A. J., and Strathem, J. N. (1984).Resolution of recombination intermediates generated during yeast mating type switching. Nature (London) 310,744-753. Kondo, T., Arakawa, H., Kitao, H., Hirota, Y.,and Yamagishi, H. (1993).Signal joint of immunoglobulin VAl-JA and novel joints of chimeric V pseudogenes on extrachromosoma1 circular DNA from chicken bursa. Eur. J. Imunol. 23,245-249. Langman, R. E., and Cohn, M. (1993).A theory of the ontogeny of the chicken humoral immune system: The consequences of diversification by gene hyperconversion and its extension to rabbit. Res. Immunol. 144,422-446. Lassila, O., Eskola, J., Toivanen, P., Martin, C., and Diterlen-LiBvre, F. (1978). The

CHICKEN B-CELL REPERTOIRE FORMATION

377

origin of lymphoid stem cells studied in chick yolk sac-embryo chimeras. Nature (London)272,353-354. Lassila, O., Martin, C., Dieterlen-LiBvre, F., Gilmour, D., Eskola, J., and Toivanen, P. (1982).Migration of prebursal stem cells from the early chicken embryo to the yolk sac. Scand. J . Immunol. 16,265-268. Lauster, R., Reynaud, C.-A., Mirtensson, L., Peter, A,, Bucchini, D., Jami, J., and Weill, J.-C. (1993).Promoter, enhancer and silencer elements regulate rearrangement of an immunoglobulin transgene. EMBO J . 12,4615-4623. Le Douarin, N. M., Houssaint, E., Jotereau, F. V., and Belo, M. (1975). Origin of hemopoietic stem cells in embryonic bursa of Fabricius and bone marrow studied through interspecific chimeras. Proc. Natl. Acad. Sci. U.S.A. 72,2701-2705. Lydyard, P. M., Grossi, C. E., and Cooper, M. D. (1976). Ontogeny of B cells in the chicken. I. Sequential development of clonal diversity in the Bursa.]. Erp. Med. 144, 79-97. McCormack, W. T., and Thompson, C. B. (1990). Chicken IgL variable region gene conversions display pseudogene donor preference and 5’ to 3’ polarity. Genes Deu. 4,548-558. McCormack, W. T., Tjoelker, L. W., Carlson, L. M., Petryniak, J. B., Barth, C. F., Humphries, E. H., and Thompson, C. B. (1989a).Chicken Ig, gene rearrangement involves deletion of a circular episome and addition of single nonrandom nucleotides to both coding segments. Cell (Cambridge, Mass.) 56, 785-791. McCormack, W. T., Tjoelker, L. W., Barth, C . F., Carlson, L. M., Petryniak, B., Humphries, E. H., and Thompson, C. B. (1989b).Selection for B cells with productive IgL gene rearrangements occurs in the bursa of Fabricius during chicken embryonic development. Genes Deu. 3,838-847. Moore, M. A., and Owen, J. J. T. (1967).Chromosome marker studies in the irradiated chick embryo. Nature (London)215,1081-1082. Mueller, A. P., Wolfe, H. R.,and Meyer, J. (1959).Precipitin production in chickens. XXI. Antibody production in bursectomized chickens and in chickens injected with 19-nortestosterone on the fifth day of incubation.]. Immunol. 83,507-510. Parvari, R.,Ziv, E., Lantner, F., Heller, D. K., and Schechter, I. (1990).Somatic diversification of chicken immunoglobulin light chains by point mutations. Proc. Natl. Acad. Sci. U S A . 87,3072-3076. Pink, J. R. L. (1986).Counting components of the chicken’s B cell system. Immunol. Reu. 91, 115-128. Pink, J. R. L., and Lassila, 0.(1987).B-cell commitment and diversification in the bursa of Fabricius. Curt-. Topics Microbiol. Immunol. 135,57-64. Pink, J. R. L., Ratcliffe, M. J. H., and Vainio, 0. (1985). Immunoglobulin-bearing stem cells for clones of B (bursa-derived) lymphocytes. Eur. 1. Immunol. 15, 617620. Ratcliffe, M. J. H., and Ivanyi, J. (1981).Allotype suppression in the chicken. IV. Deletion of B cells and lack of suppressor cells during chronic suppression. Eur.]. Immunol. 11,306-310. Ratcliffe, M. J. H., Lassila, O., Pink, J. R. L., and Vainio, 0. (1986). Avian B cell precursors: Surface immunoglobulin expression is an early, possibly bursaindependent event. Eur. ]. Immunol. 16,129-133. Ray, A,, Siddiqi, I., Kolodkin A. L., and Stahl, F. W. (1988).Intra-chromosomal gene conversion induced by a DNA double-strand break in Saccharomyces cereoisiae. 1.Mol. Biol. 201,247-260. Reynaud, C. A., Anquez, V., Dahan, A., and Weill, J. C. (1985).A single rearrangement

378

CLAUDE-AGNkS REYNAUD ET AL.

event generates most of the chicken immunoglobulin light chain diversity. Cell (Cambridge, Mass.)40,283-291. Reynaud, C. A., Anquez, V., Grimal, H., and Weill, J. C. (1987). A hyperconversion mechanism generates the chicken light chain preimmune repertoire. Cell (Cambridge, Mass.) 48,379-388. Reynaud, C. A., Dahan, A., Anquez, V., and Weill, J. C. (1989).Somatic hyperconversion diversifies the single V, gene of the chicken with a high incidence in the D region. Cell (Cambridge, Mass.) 59, 171-183. Reynaud, C. A., Mackay, C. R., Miiller, R. G., and Weill, J. C. (1991a).Somatic generation of diversity in a mammalian primary lymphoid organ: The sheep ileal Peyer’s patches. Cell (Cambridge, Mass.) 64,995-1005. Reynaud, C. A., Anquez, V., and Weill, J. C. (1991b). The chicken D locus and its contribution to the immunoglobulin heavy chain repertoire. Eur. J. Zmmunol. 21, 2661-2670. Reynaud, C. A., Imhof, B. A., Anquez, V., and Weill, J. C. (1992a). Emergence of committed B lymphoid progenitors in the developing chicken embryo. EMBOJ. 12, 4349-4358. Reynaud, C. A., Hein, W. R., Imhof, B. A., and Weill, J. C. (1992b).Diversity is generated with diversity. In “Progress in Immunology,” Vol. VIII, pp. 121-128. Springer-Verlag, Heidelberg/Berlin. Reynolds, J. D., and Morris, B. (1983).The evolution and involution of Peyer’s patches in fetal and postnatal sheep. Eur. J. Zmmunol. 13,627-635. Roth, C. W., Longacre, S., Raibaud, A., Baltz, T., and Eisen, H. (1986). The use of incomplete genes for the construction of a Trypanosoma equiperdum. EMBO J . 5, 1065-1070. Smithies, 0. (1967). The genetic basis of antibody variability. Cold Spring Harbor Symp. Quant. Biol. 32, 161-166. Takeda, S., Masteller, E. L., Thompson, C. B., and Buerstedde, J. M. (1992). RAG-2 expression is not essential for chicken immunoglobulin gene conversion. Proc. Natl. Acad. Sci. U.S.A. 89,4023-4027. Thompson, C. B. (1992). Creation of immunoglobulin diversity by intrachromosomal gene cohversion. Trends Genet. 8,416-422. Thompson, C. B., and Neiman, P. (1987).Somaticdiversification ofthe chicken immunoglobulin light chain gene is limited to the rearranged variable gene segment. Cell (Cambridge, Mass.)48,369-378. Toivanen, P., and Toivanen, A. (1973).Bursa1 and post-bursa1 stem cells in the chicken: Functional characteristics. Eur. J. Immunol. 3,585-595. Toivanen, P., Naukkarinen, A., and Vainio, 0.(1987).What is the function of the bursa of Fabricius? In “Avian Immunology: Basis and Practice” (A. Toivanen and P. Toivanen, eds.), Vol. 1, pp. 79-99. CRC Press, Boca Raton, FL. Tonegawa, S. (1983). Somatic generation of antibody diversity. Nature (London)302, 575-581. Warner, J., Uhr, W., Thorbecke, G. J., and Ovary, Z. (1969).Immunoglobulins, antibodies and the bursa of Fabricius: Induction of agammaglobulinemia and the loss of all antibody-forming capacity by hormonal buresectomy. J. Zmmunol. 103, 1317-1330. Weill, J.-C., and Reynaud, C.-A. (1987). The chicken B cell compartment. Science 238, 1094-1098. Weill, J. C., Reynaud, C. A., Lassila, O., and Pink, J. R. L. (1986). Rearrangement of chicken immunoglobulin genes is not an ongoing process in the embryonic bursa of Fabricius. Proc. Natl. Acad. Sci. U.S.A. 83,3336-3340.

INDEX

A

phage display antibody fragments, 195-198 vector systems, 199-203 to primary biliary cirrhosis selfantigens from human donors, 244-245 primate-derived, 231-232 principles of combinatorial approach, 192-195,266-267 selection strategies, 206-207,261-263 staphylococcal protein A, 229 study of responses, 232-242 synthetic repertoire approach, 250, 252-260 to thyroid disease self-antigens from human donors, 242-244 viruses, 209-228 cytomegalovirus, 220-221 hepatitis B virus, 220 herpes simplex virus type 1,222-227 herpes simplex virus type 2, 222-227 human immunodeficiency virus type 1,211-217 measles, 227-228 respiratory syncytial virus, 217-219 varicella Zoster virus, 221-222 whole antibody molecules, 208 Antibodies, monoclonal, see Monoclonal antibodies Antigen receptor, homology motif, signal transduction role, 81-83 Antigens Fas, 129-131 MHC, tumor immune response to increased expression of allo-MHC, 312-313

Airway hypersensitivity, guinea pig, interleukin-5 role, 173-174 Alleles, exclusion in chicken B cell development, 372-374 Allergens, antibodies from combinatorial libraries, 228 Amino acid sequences, antibodies from combinatorial libraries analysis, 233-238 RGD, 261-263 Antibodies, human, from combinatorial libraries affinity, strategy for improvement, 263-266 allergens, 228 amino acid sequences analysis, 233-238 RGD, 261-263 bacteria, 228 cloning strategies, 203-206 design, 261-263 Epstein-Ban virus-transformed cell line-derived, 229-230 expression of antibody fragments, 207-208 to Graves' ophthalmopathy selfantigens from human donors, 245 to human immunodeficiency virus type 1 self-antigens from human donors, 246-250 HuSCID mice-derived, 230-231 hybridomas, 229-230 immune donors, 208-242 naive repertoire approach, 250-252 overview, 191-192 379

380

INDEX

MHC class I self-, 314-320 metastatic phenotype affected by, 318-320 nonimmunological effects on tumor cells, 317-318 recognition by cytotoxic T lymphocyte effector cells, 314-315 recognition by natural killer effector cells, 315-317 tumor immunogenicity affected by low level of expression, 303-305 MHC class 11, self-, tumor immune response to increased expression, 313 non-self, antibodies to allergens, 228 bacteria, 228 Epstein-Ban-virus-transformed cell lines, 229-230 HuSCID mice, 230-231 hybridomas, 229-230 overview, 208-209 primates, 231-232 staphylococcal protein A, 229 study of responses, 232-242 viruses, 209-228 self-, from human donors, antibodies to Graves’ ophthalmopathy, 245 human immunodeficiency virus type 1,246-250 primary biliary cirrhosis, 244-245 thyroid disease, 242-244 staphylococcal protein A, antibodies to, 229 tumor, see Tumor antigens tumor immunogenicity affected by low level of expression, 303-305 tumor-specific transplantation, characteristics, 285 Apoptosis, Fas-mediated, 129, 133-134, 140 Asthma, interleukin-5 role, 174-176 Autoimmune diseases, Fas role, 138, 140 Avian leukosis virus, in chicken B cell DT40 induction, 361-362

B Bacteria, antibodies from combinatorial libraries, 228 Basophils, interleukin-5-mediated activity, 156 B cells, chicken, repertoire formation in, 353-375 development allelic exclusion, 372-374 antisilencer regulation, 372-374 bursa role, 364-367 embryonic, 367-369,374 immunoglobulin sequence selection, 369-372 silencer regulation, 372-374 gene conversiop avian leukosis virus induction of DT40 cell line, 361-362 DT40,361-362 Holliday junction, possible formation, 362-365 hyperconversion mechanism, 358-361 recombination models, 362-365 immunoglobulin gene organization amino acid composition, 356, 358 D elements, 356-358 heavy chain loci, 354-356 light chain loci, 354-356 B cells, interleukin-5 effects regulation of development, 154-155 signaling in X chromosome-linked immunodeficient mice, 171-172

C Cell death, Fas-mediated, 129, 133-134, 140 Chicken B cells, repertoire formation in, see B cells Cirrhosis, primary biliary, and antibodies to self-antigens from human donors, 244-245 Cloning, antibodies from combinatorial libraries, 203-206 Combinatorial libraries, human antibodies from, see Antibodies, human, from combinatorial libraries

INDEX

Cytokine genes, in tumor cell engineering, 321-323 Cytomegalovirus, human, antibodies from combinatorial libraries, 220-221 Cytotoxic T lymphocytes Fas expression in, 138 self-MHC class I recognition by effector cells, 314-315

D Death factor, Fas as, 129-140 Diseases autoimmune, Fas role in, 138, 140 human, interleukin-5 role in asthma, 174-176 graft rejection, 177 helminth infections, 176-177 tumors, 177-178 DNA, complementary, interleukin-5, organization, 148-149

E Embryo, chicken, B cell, early commitment in, 367-369,374 Eosinophilia, interleukin-5 role asthmatic patients, 174-176 experimental, ,173 guinea pig, 173-174 mRNA expression in patients, 158-159 parasite infection association, 172 tumors associated with, 177-178 Eosinophils, interleukin-5-mediated production, 156 Epstein-Barr virus antibodies from transformed cell lines, 229-230 tumor antigen induction, 292 Experimental eosinophilia, interleukin-5 role, 173

F Fas, 129-140 antigen, 129-131

38 1

apoptosis role, 129, 133-134, 140 in oitro, 133-134 in oioo, 133-134 autoimmune diseases associated with, 138,140 expression, 130-132 gene, mutation in Zpr mice, 132-133 loss of function mutation, 139 physiologic roles, 137-139 signal transduction role, 134-135,139 T cell development role, 137 Fas ligand, 130, 135-139 characteristics, 135-137 expression in cytotoxic T lymphocytes, 138 Fc receptors characteristics, 1 FcaR, characteristics, 38-39 FcaRI biochemical structure, 39-40 cell distribution, 40-41 characteristics, 39 ligand properties, 40 molecular structure, 39-40 monoclonal antibodies, 9, 40-41 polymorphisms, 41 FcyR biological function, 69-70 characteristics, 2-3 signal transduction role mechanisms, 72-77 phosphorylation, 74-77 second messenger interactions, 72-74 structural factors, 82, 84-89 FcyRI biochemical structure, 3-7 biological function, 69 cell distribution, 8 characteristics, 2-4 gene structure, 4-5,7 immunoglobulin interactions, molecular basis, 48-54 ligand properties, 7-8 molecular structure, 3-7 monoclonal antibodies, 8-9 polymorphisms, 8-10 signal transduction role phosphorylation, 76-77 structural factors, 84

382 FcyRII biochemical structure, 10-16 biological function, 70 cell distribution, 17-19 characteristics, 2-3, 11 gene structure, 11-13,15-16 immunoglobulin interactions, molecular basis, 49-50,54-59 ligand properties, 16-17 molecular structure, 10-16 monoclonal antibodies, 9, 17-19 polymorphisms, 19-21 signal transduction role phosphorylation, 75-77 structural factors, 82, 84-87 transcripts, 13 FcyRIII biochemical structure, 21-27 biological function, 70 cell distribution, 28-29 characteristics, 2-3, 22 gene structure, 22-26 immunoglobulin interactions, molecular basis, 49-50,59-61 ligand properties, 27-28 molecular structure, 21-27 monoclonal antibodies, 9,28-29 polymorphisms, 29-30 signal transduction role phosphorylation, 7 6 7 7 structural factors, 84, 87-88 transcripts, 23 FcsR, characteristics, 43 FccR, characteristics, 30 FceRI biochemical structure, 31-36 biological hnction, 70-71 cell distribution, 36-37 characteristics, 30 gene structure, 31-34 immunoglobulin interactions, molecular basis, 49-50,61-68 ligand properties, 36 molecular structure, 31-36 monoclonal antibodies, 9,36-37 polymorphisms, 37 signal transduction role mechanisms, 77-81 phosphorylation in, 79-81 second messenger interactions, 77-79

INDEX

structural factors, 89-90 a-subunit, 32-33 @subunit, 33-34 y-subunit, 34-36 transcripts, 31 FceRII biochemical structure, 37-38 cell distribution, 38 characteristics, 30 gene structure, 37 ligand properties, 38 molecular structure, 37-38 monoclonal antibodies, 38 FcpR, characteristics, 41-42 FcRn, characteristics, 46-47 function biological, 69-71 FwR, 69-70 FcERI, 70-71 signal transduction, mechanism, 71-81 signal transduction, structural basis, 81-90 genes, see Genes immunoglobulin interactions, molecular basis of FcyRI, 48-54 Fc/RII, 49-50,54-59 FcyRIII, 49-50,59-61 FcERI, 49-50,61-68 overview, 48-50 polymeric IgR, characteristics, 43-46 research directions, 91 signal transduction, mechanism, 71-81 phosphorylation role, 74-77,7931 protein kinase role, 74-77, 79-81 second messenger interactions, 72-74,77-79 signal transduction, structural basis, 81-90 antigen receptor homology motif role, 81-83 F q R role, 82,84-89

G Genes, see also Oncogenes chicken B cell, conversion DT40,361-362 Holliday junction, possible formation of, 362-365

INDEX

hyperconversion mechanism, 358-361 recombination models, 362-365 cytokine, in tumor cell engineering, 321-323 Fas, mutation in Zpr mice, 132-133 Fc receptor FcyRI characteristics, 4 structure, 5, 7 FcyRII characteristics, 11 structure, 12-13, 15-16 FcyRIII characteristics, 22 structure, 23-26 FcERI a-chain, 32-33 structure, 31-34 FcERII, structure, 37 immunoglobulin, see Immunoglobulin genes interleukin-5, organization, 148-149 interleukin-5Rq structure, 162-163 V, chicken B cell diversification, 358-361 recombination models, 362-365 Graft rejection, interleukin-5 role, 177 Graves’ ophthalmopathy, antibodies to self-antigens from human donors, 245

H Helminth infections, interleukin-5 role, 176-177 Hepatitis B virus, antibodies from combinatorial libraries, 220 Herpes simplex virus type 1, antibodies from combinatorial libraries, 222-227 Herpes simplex virus type 2, antibodies from combinatorial libraries, 222-227 Hodgkin’s disease, interleukin-5 mRNA expression in patients, 158 Human antibodies, from combinatorial libraries, see Antibodies, human, from combinatorial libraries

383

Human cytomegalovirus, antibodies from combinatorial libraries, 220-221 Human diseases, see Diseases Human immunodeficiency virus type 1, antibodies from combinatorial libraries, 211-217 to self-antigens from human donors, 246-250 Hybridomas, antibodies from, 229-230

Immune response, against tumors, see Tumors, immune response against Immune suppressive factors, tumor immunogenicity affected by, 307-309 Immunodeficient mouse HuSCID-derived antibodies, 230-231 interleukin-5 production in, 171-172 Immunoglobulin genes, in chicken B cells conversion allelic exclusion, 372, 374 avian leukosis virus induction of DT40 cell line, 361-362 rearrangement in bursa, 365-367 recombination models, 362-365 organization D elements, 356-358 heavy chain loci, 354-356 light chain loci, 354-356 Immunoglobulins Fc receptors for, see Fc receptors interleukin-5-regulated production, 151-154 IgA, 152-153 IgE, 153 IgG, 153 sequence selection in chicken B cell development, 369-372 Infections, interleukin-5 role helminth, 176-177 parasite, 172-173 Inflammation, interleukin-5-associated, in asthmatic patients, 174-176 Interleukin-5 animal models of production airway hypersensitivity, 173-174 eosinophilia in guinea pigs, 173-174 experimental eosinophilia, 173

384

INDEX

parasite infection, 172-173 transgenic mouse, 169-170 tumor rejection, 174 Major histocompatibility complex, see X chromosome-linked Antigens immunodeficient mouse, Measles virus, antibodies from 171-172 combinatorial libraries, 227-228 characteristics, 145-146 MHC, see Antigens functional properties Monoclonal antibodies, Fc receptor FcaRI, 9 , 4 0 4 1 basophil regulation, 156 B cell developmentregulation, 154-155 FwRI, 8-9 eosinophil production, 156 FwRII, 9, 17-19 immunoglobulin production F q R I l l , 9, 29-30 regulation, 151-154 FcERI, 9,36-37 interleukin-2 receptor induction, 155 FcsRII, 38 future perspectives, 178-179 Mouse HuSCID, antibodies derived from, histological background, 146-148 in human disease 230-231 asthma, 174-176 immunodeficient, interleukin-5 graft rejection, 177 production, 171-172 lpr, Fas gene mutation in, 132-133 helminth infections, 176-177 tumors, 177-178 transgenic, interleukin-5 production messenger RNA expression, 156-159 models, 169-170 molecular structure X chromosome-linked biological activity, 149 immunodeficient, interleukin-5 cDNA organization, 148-149 production models, 171-172 Mucin, tumor antigen recognition of, gene organization, 148-149 302-303 polypeptides, 149-151 Mutations Interleukin-10, tumor immunogenicity Fas reduced by, 307-309 genes in lpr mice, 132-133 Interleukin-2 receptor, induction by interleukin-5, 155 loss of function, 139 Interleukin-5 receptor p53 protein-derived tumor antigens, expression 295-296 aberrations in mice, 170-172 analysis, 168-169 function, 165-166 IL-5Ra chain N soluble forms, 163-164 structure, 160-164 Natural killer cells, self-MHC class I IL-5Ra gene structure,. 162-163 recognition in tumor cells, 315-317 IL-5RP chain structure, 164-165 signaling pathway, 166-168 signal transduction, 146 structure, 159-165 0

L Zpr mouse, Fas gene mutation in,

132-133 Lymphocytes, T, see T cells

Oncogenes, tumor antigens derived from abl, 296-297 bcr, 296-297 ras, 297-299 Ophthalmopathy, antibodies to selfantigens from human donors, 245

385

INDEX

P Papillomavirus, tumor antigen induction, 292-293 Parasite infection, interleukin-5 production in animal models, 172-173 Peptides, tumor-specific, T cell recognition, 288-290 Phosphorylation, in Fc receptormediated signal transduction FwR, 74-77 FcsRI, 79-81 Polypeptides, interleukin-5, molecular structure, 149-151 Primary biliary cirrhosis, antibodies to self-antigens from human donors, 244-245 Primates, antibodies derived from, 231-232 Protein kinases, in Fc receptor-mediated signal transduction F v R , 74-77 FcsRI, 79-81 Proteins staphylococcal protein A, antibodies from combinatorial libraries, 229 tumor antigens derived from abl oncogenes, 296-297 bcr oncogenes, 296-297 mutated ~53,295-296 overexpression role, 299-302 ras oncogenes, 297-299 tum-, 293-295

R Respiratory syncytial virus, antibodies from combinatorial libraries, 217-219 RNA messenger, interleukin-5 expression, 156-159

S Second messengers, in Fc receptormediated signal transduction FwR, 72-74 FcsRI, 77-79

Signal transduction Fas-mediated, 134-135, 139 Fc receptor-mediated antigen receptor homology motif, 81-83 FcyR, mechanisms of activity, 72-77 FcyR, structural factors, 82, 84-89 FcERI, 77-81,89-90 mechanism, 71-81 phosphorylation, 74-77,79-81 protein kinase involvement, 74-77, 79-81 second messenger interactions, 72-74,77-79 structural factors, 81-90 Staphylococcal protein A, antibodies from combinatorial libraries, 229

T T cells Fas effects development role, 137 expression in cytotoxic T lymphocytes, 138 self-MHC class I recognition by cytotoxic T lymphocyte effector cells, 314-315 tumor antigen interactions costimulatory pathway-determined immogenicity, 288-290 early evidence of recognition, 285-288 tumor-specific peptide recognition, 288-290 Thyroid disease, antibodies to selfantigens from human donors, 242-244 Transforming growth factor+, tumor immunogenicity affected by, 307-309 Tumor antigens B7, costimulatory effects on immunogenicity, 305-307 CD28, costimulatory effects on immunogenicity, 305-307 costimulatory pathway, 305-307 CTLA-4, costimulatory effects on immunogenicity, 305-307 mucin recognition, 302-303

386

INDEX

protein-derived abl oncogenes, 296-297 bcr oncogenes, 296-297 mutated p53,295-296 overexpression resulting in, 299-302 ras oncogenes, 297-299 tum-, 293-295 serological detection, 283-284 T cell recognition early evidence, 285-288 tumor-specific peptides, 288-290 virus-induced tumors, 290-293 Tumors, immune response against, 281-326 antigens, see Tumor antigens B7-transfected cells, 320-321 cytokine genes in engineering of tumor cells, 321-323 escape mechanisms, 324-325 immunogenicity, factors contributing to low level antigen expression, low, 303-305 costimulatory molecules, lack of, 305-307 immune suppressive factors, 307-309 interleukin-10, 307-309 MHC class I expression, low, 303-305 transforming growth factor-@, 307-309 tumor environment modification, 309-312 MHC expression increased by transfection

allO-MHC, 312-313 self-MHC class I, 314-320 self-MHC class 11,313 tumor antigens, see Tumor antigens Tumors, interleukin-5 effects, 177-178 rejection, 174 Tumor-specific transplantation antigens, characteristics, 285

V Varicella Zoster virus, antibodies from combinatorial libraries, 221-222 Viruses antibodies, from combinatorial libraries cytomegalovirus, human, 220-221 hepatitis B virus, 220 herpes simplex virus type 1, 222-227 herpes simplex virus type 2, 222-227 human immunodeficiency virus type 1,211-217 measles, 227-228 respiratory syncytial virus, 217-219 varicella Zoster virus, 221-222 avian leukosis, in chicken B cell DT40 induction, 361-362 tumor antigen induction by, 290-293 Epstein-Barr virus, 292 papillomavirus, 292-293